Reviews

The Susan G. Komen Foundation estimates that 84% of breast cancers are found through mammography.¹ Clearly, the value of mammography is proven. But controversy and confusion abound on how much mammography, and beginning at what age, is best for women.

Currently, the United States Preventive Services Task Force (USPSTF), the American Cancer Society (ACS), and the American College of Obstetricians and Gynecologists (ACOG) all have differing recommendations about mammography and about the importance of clinical breast examinations. These inconsistencies largely are due to different interpretations of the same data, not the data itself, and tend to center on how harm is defined and measured. Importantly, these differences can wreak havoc on our patients’ confidence in our counsel and decision making, and can complicate women’s access to screening. Under the Affordable Care Act, women are guaranteed coverage of annual mammograms, but new USPSTF recommendations, due out soon, may undermine that guarantee.

On October 20, ACOG responded to the ACS’ new recommendations on breast cancer screening by emphasizing our continued advice that women should begin annual mammography screening at age 40, along with a clinical breast exam.²

Consensus conference plansIn an effort to address widespread confusion among patients, health care professionals, and payers, ACOG is convening a consensus conference in January 2016, with the goal of arriving at a consistent set of guidelines that can be agreed to, implemented clinically across the country, and hopefully adopted by insurers, as well. Major organizations and providers of women’s health care, including ACS, will gather to evaluate and interpret the data in greater detail and to consider the available data in the broader context of patient care.

Without doubt, guidelines and recommendations will need to evolve as new evidence emerges, but our hope is that scientific and medical organizations can look at the same evidence and speak with one voice on what is best for women’s health. Our patients would benefit from that alone.

ACOG’s recommendations, summarized

• Clinical breast examination every year for women aged 19 and older.
• Screening mammography every year for women aged 40 and older.
• Breast self-awareness has the potential to detect palpable breast cancer and can be recommended.²

Share your thoughts! Send your Letter to the Editor to rbarbieri@frontlinemedcom.com. Please include your name and the city and state in which you practice.

The Susan G. Komen Foundation estimates that 84% of breast cancers are found through mammography.¹ Clearly, the value of mammography is proven. But controversy and confusion abound on how much mammography, and beginning at what age, is best for women.

Currently, the United States Preventive Services Task Force (USPSTF), the American Cancer Society (ACS), and the American College of Obstetricians and Gynecologists (ACOG) all have differing recommendations about mammography and about the importance of clinical breast examinations. These inconsistencies largely are due to different interpretations of the same data, not the data itself, and tend to center on how harm is defined and measured. Importantly, these differences can wreak havoc on our patients’ confidence in our counsel and decision making, and can complicate women’s access to screening. Under the Affordable Care Act, women are guaranteed coverage of annual mammograms, but new USPSTF recommendations, due out soon, may undermine that guarantee.

On October 20, ACOG responded to the ACS’ new recommendations on breast cancer screening by emphasizing our continued advice that women should begin annual mammography screening at age 40, along with a clinical breast exam.²

Consensus conference plansIn an effort to address widespread confusion among patients, health care professionals, and payers, ACOG is convening a consensus conference in January 2016, with the goal of arriving at a consistent set of guidelines that can be agreed to, implemented clinically across the country, and hopefully adopted by insurers, as well. Major organizations and providers of women’s health care, including ACS, will gather to evaluate and interpret the data in greater detail and to consider the available data in the broader context of patient care.

Without doubt, guidelines and recommendations will need to evolve as new evidence emerges, but our hope is that scientific and medical organizations can look at the same evidence and speak with one voice on what is best for women’s health. Our patients would benefit from that alone.

ACOG’s recommendations, summarized

• Clinical breast examination every year for women aged 19 and older.
• Screening mammography every year for women aged 40 and older.
• Breast self-awareness has the potential to detect palpable breast cancer and can be recommended.²

Share your thoughts! Send your Letter to the Editor to rbarbieri@frontlinemedcom.com. Please include your name and the city and state in which you practice.

The Susan G. Komen Foundation estimates that 84% of breast cancers are found through mammography.¹ Clearly, the value of mammography is proven. But controversy and confusion abound on how much mammography, and beginning at what age, is best for women.

Currently, the United States Preventive Services Task Force (USPSTF), the American Cancer Society (ACS), and the American College of Obstetricians and Gynecologists (ACOG) all have differing recommendations about mammography and about the importance of clinical breast examinations. These inconsistencies largely are due to different interpretations of the same data, not the data itself, and tend to center on how harm is defined and measured. Importantly, these differences can wreak havoc on our patients’ confidence in our counsel and decision making, and can complicate women’s access to screening. Under the Affordable Care Act, women are guaranteed coverage of annual mammograms, but new USPSTF recommendations, due out soon, may undermine that guarantee.

On October 20, ACOG responded to the ACS’ new recommendations on breast cancer screening by emphasizing our continued advice that women should begin annual mammography screening at age 40, along with a clinical breast exam.²

Consensus conference plansIn an effort to address widespread confusion among patients, health care professionals, and payers, ACOG is convening a consensus conference in January 2016, with the goal of arriving at a consistent set of guidelines that can be agreed to, implemented clinically across the country, and hopefully adopted by insurers, as well. Major organizations and providers of women’s health care, including ACS, will gather to evaluate and interpret the data in greater detail and to consider the available data in the broader context of patient care.

Without doubt, guidelines and recommendations will need to evolve as new evidence emerges, but our hope is that scientific and medical organizations can look at the same evidence and speak with one voice on what is best for women’s health. Our patients would benefit from that alone.

ACOG’s recommendations, summarized

• Clinical breast examination every year for women aged 19 and older.
• Screening mammography every year for women aged 40 and older.
• Breast self-awareness has the potential to detect palpable breast cancer and can be recommended.²

Share your thoughts! Send your Letter to the Editor to rbarbieri@frontlinemedcom.com. Please include your name and the city and state in which you practice.

Over the past 20 years, substantial progress has been achieved in our understanding of breast cancer and in breast cancer treatment, with mortality from breast cancer declining by more than 25% over this time. This progress has been characterized by a greater understanding of the molecular biology of breast cancer, rational drug design, development of agents with specific cellular targets and pathways, development of better prognostic and predictive multigene assays, and marked improvements in supportive care.

To read the full article in PDF:

Click here

Over the past 20 years, substantial progress has been achieved in our understanding of breast cancer and in breast cancer treatment, with mortality from breast cancer declining by more than 25% over this time. This progress has been characterized by a greater understanding of the molecular biology of breast cancer, rational drug design, development of agents with specific cellular targets and pathways, development of better prognostic and predictive multigene assays, and marked improvements in supportive care.

To read the full article in PDF:

Click here

Over the past 20 years, substantial progress has been achieved in our understanding of breast cancer and in breast cancer treatment, with mortality from breast cancer declining by more than 25% over this time. This progress has been characterized by a greater understanding of the molecular biology of breast cancer, rational drug design, development of agents with specific cellular targets and pathways, development of better prognostic and predictive multigene assays, and marked improvements in supportive care.

To read the full article in PDF:

Click here

Pancreas transplant is the only long-term diabetes treatment that consistently results in normal hemoglobin A_1c levels without the risk of severe hypoglycemia. Additionally, pancreas transplant may prevent, halt, or even reverse the complications of diabetes.

Here, we explore the indications, options, and outcomes of pancreas transplant as a treatment for diabetes mellitus.

DIABETES IS COMMON, AND OFTEN NOT WELL CONTROLLED

Diabetes mellitus affects more than 25 million people in the United States (8.3% of the population) and is the leading cause of kidney failure, nontraumatic lower-limb amputation, and adult-onset blindness. In 2007, nearly $116 billion was spent on diabetes treatment, not counting another $58 billion in indirect costs such as disability, work loss, and premature death.¹

Only about half of patients achieve hemoglobin A_1c < 7% with medical therapy

Despite the tremendous expenditure in human, material, and financial resources, only about 50% of patients achieve their diabetes treatment goals. In 2013, a large US population-based study­² reported that 52.2% of patients were achieving the American Diabetes Association treatment goal of hemoglobin A_1c lower than 7%. A similar study in South Korea³ found that 45.6% were at this goal.

Most of the patients in these studies had type 2 diabetes, and the data suggested that attaining glycemic goals is more difficult in insulin-treated patients. Studies of patients with type 1 diabetes found hemoglobin A_1c levels lower than 7% in only 8.1% of hospitalized patients with type 1 diabetes, and in only 13% in an outpatient diabetes clinic.^4,5

YET RATES OF PANCREAS TRANSPLANT ARE DECLINING

Pancreas transplant was first performed more than 40 years ago at the University of Minnesota.⁶ Since then, dramatic changes in immunosuppression, organ preservation, surgical technique, and donor and recipient selection have brought about significant progress.

Currently, more than 13,000 patients are alive with a functioning pancreas allograft. After reaching a peak in 2004, the annual number of pancreas transplants performed in the United States has declined steadily, whereas the procedure continues to increase in popularity outside North America.⁷ The primary reason for the decline is recognition of donor factors that lead to success—surgeons are refusing to transplant organs they might have accepted previously, because experience suggests they would yield poor results. In the United States, 1,043 pancreas transplants were performed in 2012, and more than 3,100 patients were on the waiting list.⁸

Islet cell transplant—a different procedure involving harvesting, encapsulating, and implanting insulin-producing beta cells—has not gained widespread application due to very low long-term success rates.

THREE CATEGORIES OF PANCREAS TRANSPLANT

Pancreas transplant can be categorized according to whether the patient is also receiving or has already received a kidney graft (Table 1).

Simultaneous kidney and pancreas transplant is performed in patients who have type 1 diabetes with advanced chronic kidney disease due to diabetic nephropathy. This remains the most commonly performed type, accounting for 79% of all pancreas transplants in 2012.⁸

Pancreas-after-kidney transplant is most often done after a living-donor kidney transplant. This procedure accounted for most of the increase in pancreas transplants during the first decade of the 2000s. However, the number of these procedures has steadily decreased since 2004, and in 2012 accounted for only 12% of pancreas transplants.⁸

Pancreas transplant alone is performed in nonuremic diabetic patients who have labile blood sugar control. Performed in patients with preserved renal function but severe complications of “brittle” diabetes, such as hypoglycemic unawareness, this type accounts for 8% of pancreas transplants.⁹

Indications for pancreas transplant

A small number of these procedures are done for indications unrelated to diabetes mellitus. In most of these cases, the pancreas is transplanted as part of a multivisceral transplant to facilitate the technical (surgical) aspect of the procedure—the pancreas, liver, stomach, gallbladder, and part of the intestines are transplanted en bloc to maintain the native vasculature. Very infrequently, pancreas transplant is done to replace exocrine pancreatic function.

A small, select group of patients with type 2 diabetes and low body mass index (BMI) may be eligible for pancreas transplant, and they accounted for 8.2% of active candidates in 2012.⁸ However, most pancreas transplants are performed in patients with type 1 diabetes.

WHAT MAKES A GOOD ALLOGRAFT?

Pancreas allografts are procured as whole organs from brain-dead organ donors. Relatively few pancreas allografts (3.1% in 2012) are from cardiac-death donors, because of concern about warm ischemic injury during the period of circulatory arrest.⁸

Proper donor selection is critical to the success of pancreas transplant, as donor factors including medical history, age, BMI, and cause of death can significantly affect the outcome. In general, transplant of a pancreas allograft from a young donor (age < 30) with excellent organ function, low BMI, and traumatic cause of death provides the best chance of success.

The Pancreas Donor Risk Index (PDRI)¹⁰ was developed after analysis of objective donor criteria, transplant type, and ischemic time in grafts transplanted between 2000 and 2006. One-year graft survival was directly related to the PDRI and ranged between 77% and 87% in recipients of “standard” pancreas allografts (PDRI score of 1.0). Use of grafts from the highest (worst) three quintiles of PDRI (PDRI score > 1.16) was associated with 1-year graft survival rates of 67% to 82%, significantly inferior to that seen with “higher- quality” grafts, again emphasizing the need for rigorous donor selection.¹⁰

In addition to these objective measures, visual assessment of pancreas quality at the time of procurement remains an equally important predictor of success. Determination of subjective features, such as fatty infiltration and glandular fibrosis, requires surgical experience developed over several years. In a 2010 analysis, dissatisfaction with the quality of the donor graft on inspection accounted for more than 80% of refusals of potential pancreas donors.¹¹ These studies illustrate an ill-defined aspect of pancreas transplant, ie, even when the pancreas donor is perceived to be suitable, the outcome may be markedly different.

SURGICAL COMPLICATIONS

Surgical complications have long been considered a limiting factor in the growth of pancreas transplant. Technical failure or loss of the graft within 90 days is most commonly due to graft thrombosis, leakage of the enteric anastomosis, or severe peripancreatic infection. The rate of technical failure has declined across all recipient categories and is currently about 9%.⁸

DO RECIPIENT FACTORS AFFECT OUTCOMES?

As mentioned above, the PDRI identifies donor factors that influence the 1-year graft survival rate. Recipient factors are also thought to play a role, although the influence of these factors has not been consistently demonstrated.

Humar et al¹⁵ found that recipient obesity (defined in this study as BMI > 25 kg/m²) and donor age over 40 were risk factors for early laparotomy after pancreas transplant.¹⁵ Moreover, patients undergoing early laparotomy had poorer graft survival outcomes.

This finding was reinforced by an analysis of 5,725 primary simultaneous pancreas-kidney recipients between 2000 and 2007. Obesity (BMI 30 ≥ kg/m²) was associated with increased rates of patient death, pancreas graft loss, and kidney graft loss at 3 years.¹⁶

More recently, Finger et al¹⁷ did not find a statistically significant association between recipient BMI and technical failure, but they did notice a trend toward increased graft loss with a BMI greater than 25 kg/m². Similarly, others have not found a clear adverse association between recipient BMI and pancreas graft survival.

Intuitively, obesity and other recipient factors such as age, vascular disease, duration of diabetes, and dialysis should influence pancreas graft survival but have not been shown in analyses to carry an adverse effect.¹⁸ The inability to consistently find adverse effects of recipient characteristics is most likely due to the relative similarity between the vast majority of pancreas transplant recipients and the relatively small numbers of adverse events. In 98 consecutive pancreas transplants at our center between 2009 and 2014, the technical loss rate was 1.8% (unpublished data).

Acute rejection most commonly occurs during the first year and is usually reversible. More than 1 year after transplant, graft loss is due to chronic rejection, and death is usually from underlying cardiovascular disease.

The immunosuppressive regimens used in pancreas transplant are similar to those in kidney transplant. Since the pancreas is considered to be more immunogenic than other organs, most centers employ a strategy of induction immunosuppression with T-cell–depleting or interleukin 2-receptor antibodies. Maintenance immunosuppression consists of a calcineurin inhibitor (tacrolimus or cyclosporine), an antimetabolite (mycophenolate), and a corticosteroid.⁸

Immunosuppressive complications occur at a rate similar to that seen in other solid-organ transplants and include an increased risk of opportunistic infection and malignancy. The risk of these complications must be balanced against the patient’s risk of health decline with dialysis and insulin-based therapies.

OVERALL OUTCOMES ARE GOOD

The success rate of pancreas transplant is currently at its highest since the inception of the procedure. The unadjusted patient survival rate for all groups is over 96% at 1 year, and over 80% at 5 years.⁸ One-year patient survival after pancreas transplant alone, at better than 96%, is the highest of all organ transplant procedures.⁹

Patient survival 1 year after pancreas-alone transplant is > 96%

Several recently published single-center reviews of pancreas transplant since 2000 report patient survival rates of 96% to 100% at 1 year and 88% to 100% at 5 years.^19–22 This variability is likely closely linked to donor and recipient selection, as centers performing smaller numbers of transplants tend to be more selective and, in turn, report higher patient survival rates.^19,21

Long-term patient survival outcomes can be gathered from larger, registry-based reviews, accepting limitations in assessing causes of patient death. Siskind et al²³ analyzed the outcomes of 20,854 US pancreas transplants done between 1996 and 2012 and found the 10-year patient survival rate ranged from 43% to 77% and was highly dependent on patient age at the time of the procedure.²³ Patient survival after transplant must be balanced against the generally poor long-term survival prospects of diabetic patients on dialysis.

By type of transplant, pancreas graft survival rates at 1 year are 89% for simultaneous pancreas-kidney transplant, 86% for pancreas-after-kidney transplant, and 84% for pancreas-alone transplant. Graft survival rates at 5 years are 71% for simultaneous pancreas-kidney transplant, 65% for pancreas-after-kidney transplant, and 58% for pancreas-alone transplant.^8,9

Simultaneous pancreas-kidney transplant has been shown to improve the survival rate compared with cadaveric kidney transplant alone in patients with type 1 diabetes and chronic kidney disease.^24,25 The survival benefit of isolated pancreas transplant (after kidney transplant and alone) is not evident at 4-year follow-up compared with patients on the waiting list. However, the benefit for the individual patient must be considered by weighing the incapacities experienced with insulin-based treatments against the risks of surgery and immunosuppression.^26,27 For patients who have experienced frequent and significant hypoglycemic episodes, particularly those requiring third-party assistance, pancreas transplant can be a lifesaving procedure.

Effects on secondary diabetic complications

Notwithstanding the effect on the patient’s life span, data from several studies of long-term pancreas transplant recipients suggest that secondary diabetic complications can be halted or even improved. Most of these studies examined the effect of restoring euglycemia in nephropathy and the subsequent influence on renal function.

Effect on renal function. Kleinclauss et al²⁸ examined renal allograft function in type 1 diabetic recipients of living-donor kidney transplants. Comparing kidney allograft survival and function in patients who received a subsequent pancreas-after-kidney transplant vs those who did not, graft survival was superior after 5 years, and the estimated glomerular filtration rate was 10 mL/min higher in pancreas-after-kidney recipients.²⁸ This improvement in renal function was not seen immediately after the pancreas transplant but became evident more than 4 years after establishment of normoglycemia. Somewhat similarly, reversal of diabetic changes in native kidney biopsies has been seen 10 years after pancreas transplant.²⁹

Effect on neuropathy. In other studies, reversal of autonomic neuropathy and hypoglycemic unawareness and improvements in peripheral sensory-motor neuropathy have also been observed.^30–32

Effect on retinopathy. Improvements in early-stage nonproliferative diabetic retinopathy and laser-treated proliferative lesions have been seen, even within short periods of follow-up.³³ Other groups have shown a significantly higher proportion of improvement or stability of advanced diabetic retinopathy at 3 years after simultaneous pancreas-kidney transplant, compared with kidney transplant alone in patients with type 1 diabetes.³⁴

Effect on heart disease. Salutary effects on cardiovascular risk factors and amelioration of cardiac morphology and functional cardiac indices have been seen within the first posttransplant year.³⁵ Moreover, with longer follow-up (nearly 4 years), simultaneous pancreas-kidney recipients with functioning pancreas grafts were found to have less progression of coronary atherosclerosis than simultaneous pancreas-kidney recipients with early pancreas graft loss.³⁶ These data provide a potential pathophysiologic mechanism for the long-term survival advantage seen in uremic type 1 diabetic patients undergoing simultaneous pancreas-kidney transplant.

In the aggregate, these findings suggest that, in the absence of surgical and immunosuppression-related complications, a functioning pancreas allograft can alter the progress of diabetic complications. As an extension of these results, pancreas transplant done earlier in the course of diabetes may have an even greater impact.

Pancreas transplant is the only long-term diabetes treatment that consistently results in normal hemoglobin A_1c levels without the risk of severe hypoglycemia. Additionally, pancreas transplant may prevent, halt, or even reverse the complications of diabetes.

Here, we explore the indications, options, and outcomes of pancreas transplant as a treatment for diabetes mellitus.

DIABETES IS COMMON, AND OFTEN NOT WELL CONTROLLED

Diabetes mellitus affects more than 25 million people in the United States (8.3% of the population) and is the leading cause of kidney failure, nontraumatic lower-limb amputation, and adult-onset blindness. In 2007, nearly $116 billion was spent on diabetes treatment, not counting another $58 billion in indirect costs such as disability, work loss, and premature death.¹

Only about half of patients achieve hemoglobin A_1c < 7% with medical therapy

Despite the tremendous expenditure in human, material, and financial resources, only about 50% of patients achieve their diabetes treatment goals. In 2013, a large US population-based study­² reported that 52.2% of patients were achieving the American Diabetes Association treatment goal of hemoglobin A_1c lower than 7%. A similar study in South Korea³ found that 45.6% were at this goal.

Most of the patients in these studies had type 2 diabetes, and the data suggested that attaining glycemic goals is more difficult in insulin-treated patients. Studies of patients with type 1 diabetes found hemoglobin A_1c levels lower than 7% in only 8.1% of hospitalized patients with type 1 diabetes, and in only 13% in an outpatient diabetes clinic.^4,5

YET RATES OF PANCREAS TRANSPLANT ARE DECLINING

Pancreas transplant was first performed more than 40 years ago at the University of Minnesota.⁶ Since then, dramatic changes in immunosuppression, organ preservation, surgical technique, and donor and recipient selection have brought about significant progress.

Currently, more than 13,000 patients are alive with a functioning pancreas allograft. After reaching a peak in 2004, the annual number of pancreas transplants performed in the United States has declined steadily, whereas the procedure continues to increase in popularity outside North America.⁷ The primary reason for the decline is recognition of donor factors that lead to success—surgeons are refusing to transplant organs they might have accepted previously, because experience suggests they would yield poor results. In the United States, 1,043 pancreas transplants were performed in 2012, and more than 3,100 patients were on the waiting list.⁸

Islet cell transplant—a different procedure involving harvesting, encapsulating, and implanting insulin-producing beta cells—has not gained widespread application due to very low long-term success rates.

THREE CATEGORIES OF PANCREAS TRANSPLANT

Pancreas transplant can be categorized according to whether the patient is also receiving or has already received a kidney graft (Table 1).

Simultaneous kidney and pancreas transplant is performed in patients who have type 1 diabetes with advanced chronic kidney disease due to diabetic nephropathy. This remains the most commonly performed type, accounting for 79% of all pancreas transplants in 2012.⁸

Pancreas-after-kidney transplant is most often done after a living-donor kidney transplant. This procedure accounted for most of the increase in pancreas transplants during the first decade of the 2000s. However, the number of these procedures has steadily decreased since 2004, and in 2012 accounted for only 12% of pancreas transplants.⁸

Pancreas transplant alone is performed in nonuremic diabetic patients who have labile blood sugar control. Performed in patients with preserved renal function but severe complications of “brittle” diabetes, such as hypoglycemic unawareness, this type accounts for 8% of pancreas transplants.⁹

Indications for pancreas transplant

A small number of these procedures are done for indications unrelated to diabetes mellitus. In most of these cases, the pancreas is transplanted as part of a multivisceral transplant to facilitate the technical (surgical) aspect of the procedure—the pancreas, liver, stomach, gallbladder, and part of the intestines are transplanted en bloc to maintain the native vasculature. Very infrequently, pancreas transplant is done to replace exocrine pancreatic function.

A small, select group of patients with type 2 diabetes and low body mass index (BMI) may be eligible for pancreas transplant, and they accounted for 8.2% of active candidates in 2012.⁸ However, most pancreas transplants are performed in patients with type 1 diabetes.

WHAT MAKES A GOOD ALLOGRAFT?

Pancreas allografts are procured as whole organs from brain-dead organ donors. Relatively few pancreas allografts (3.1% in 2012) are from cardiac-death donors, because of concern about warm ischemic injury during the period of circulatory arrest.⁸

Proper donor selection is critical to the success of pancreas transplant, as donor factors including medical history, age, BMI, and cause of death can significantly affect the outcome. In general, transplant of a pancreas allograft from a young donor (age < 30) with excellent organ function, low BMI, and traumatic cause of death provides the best chance of success.

The Pancreas Donor Risk Index (PDRI)¹⁰ was developed after analysis of objective donor criteria, transplant type, and ischemic time in grafts transplanted between 2000 and 2006. One-year graft survival was directly related to the PDRI and ranged between 77% and 87% in recipients of “standard” pancreas allografts (PDRI score of 1.0). Use of grafts from the highest (worst) three quintiles of PDRI (PDRI score > 1.16) was associated with 1-year graft survival rates of 67% to 82%, significantly inferior to that seen with “higher- quality” grafts, again emphasizing the need for rigorous donor selection.¹⁰

In addition to these objective measures, visual assessment of pancreas quality at the time of procurement remains an equally important predictor of success. Determination of subjective features, such as fatty infiltration and glandular fibrosis, requires surgical experience developed over several years. In a 2010 analysis, dissatisfaction with the quality of the donor graft on inspection accounted for more than 80% of refusals of potential pancreas donors.¹¹ These studies illustrate an ill-defined aspect of pancreas transplant, ie, even when the pancreas donor is perceived to be suitable, the outcome may be markedly different.

SURGICAL COMPLICATIONS

Surgical complications have long been considered a limiting factor in the growth of pancreas transplant. Technical failure or loss of the graft within 90 days is most commonly due to graft thrombosis, leakage of the enteric anastomosis, or severe peripancreatic infection. The rate of technical failure has declined across all recipient categories and is currently about 9%.⁸

DO RECIPIENT FACTORS AFFECT OUTCOMES?

As mentioned above, the PDRI identifies donor factors that influence the 1-year graft survival rate. Recipient factors are also thought to play a role, although the influence of these factors has not been consistently demonstrated.

Humar et al¹⁵ found that recipient obesity (defined in this study as BMI > 25 kg/m²) and donor age over 40 were risk factors for early laparotomy after pancreas transplant.¹⁵ Moreover, patients undergoing early laparotomy had poorer graft survival outcomes.

This finding was reinforced by an analysis of 5,725 primary simultaneous pancreas-kidney recipients between 2000 and 2007. Obesity (BMI 30 ≥ kg/m²) was associated with increased rates of patient death, pancreas graft loss, and kidney graft loss at 3 years.¹⁶

More recently, Finger et al¹⁷ did not find a statistically significant association between recipient BMI and technical failure, but they did notice a trend toward increased graft loss with a BMI greater than 25 kg/m². Similarly, others have not found a clear adverse association between recipient BMI and pancreas graft survival.

Intuitively, obesity and other recipient factors such as age, vascular disease, duration of diabetes, and dialysis should influence pancreas graft survival but have not been shown in analyses to carry an adverse effect.¹⁸ The inability to consistently find adverse effects of recipient characteristics is most likely due to the relative similarity between the vast majority of pancreas transplant recipients and the relatively small numbers of adverse events. In 98 consecutive pancreas transplants at our center between 2009 and 2014, the technical loss rate was 1.8% (unpublished data).

Acute rejection most commonly occurs during the first year and is usually reversible. More than 1 year after transplant, graft loss is due to chronic rejection, and death is usually from underlying cardiovascular disease.

The immunosuppressive regimens used in pancreas transplant are similar to those in kidney transplant. Since the pancreas is considered to be more immunogenic than other organs, most centers employ a strategy of induction immunosuppression with T-cell–depleting or interleukin 2-receptor antibodies. Maintenance immunosuppression consists of a calcineurin inhibitor (tacrolimus or cyclosporine), an antimetabolite (mycophenolate), and a corticosteroid.⁸

Immunosuppressive complications occur at a rate similar to that seen in other solid-organ transplants and include an increased risk of opportunistic infection and malignancy. The risk of these complications must be balanced against the patient’s risk of health decline with dialysis and insulin-based therapies.

OVERALL OUTCOMES ARE GOOD

The success rate of pancreas transplant is currently at its highest since the inception of the procedure. The unadjusted patient survival rate for all groups is over 96% at 1 year, and over 80% at 5 years.⁸ One-year patient survival after pancreas transplant alone, at better than 96%, is the highest of all organ transplant procedures.⁹

Patient survival 1 year after pancreas-alone transplant is > 96%

Several recently published single-center reviews of pancreas transplant since 2000 report patient survival rates of 96% to 100% at 1 year and 88% to 100% at 5 years.^19–22 This variability is likely closely linked to donor and recipient selection, as centers performing smaller numbers of transplants tend to be more selective and, in turn, report higher patient survival rates.^19,21

Long-term patient survival outcomes can be gathered from larger, registry-based reviews, accepting limitations in assessing causes of patient death. Siskind et al²³ analyzed the outcomes of 20,854 US pancreas transplants done between 1996 and 2012 and found the 10-year patient survival rate ranged from 43% to 77% and was highly dependent on patient age at the time of the procedure.²³ Patient survival after transplant must be balanced against the generally poor long-term survival prospects of diabetic patients on dialysis.

By type of transplant, pancreas graft survival rates at 1 year are 89% for simultaneous pancreas-kidney transplant, 86% for pancreas-after-kidney transplant, and 84% for pancreas-alone transplant. Graft survival rates at 5 years are 71% for simultaneous pancreas-kidney transplant, 65% for pancreas-after-kidney transplant, and 58% for pancreas-alone transplant.^8,9

Simultaneous pancreas-kidney transplant has been shown to improve the survival rate compared with cadaveric kidney transplant alone in patients with type 1 diabetes and chronic kidney disease.^24,25 The survival benefit of isolated pancreas transplant (after kidney transplant and alone) is not evident at 4-year follow-up compared with patients on the waiting list. However, the benefit for the individual patient must be considered by weighing the incapacities experienced with insulin-based treatments against the risks of surgery and immunosuppression.^26,27 For patients who have experienced frequent and significant hypoglycemic episodes, particularly those requiring third-party assistance, pancreas transplant can be a lifesaving procedure.

Effects on secondary diabetic complications

Notwithstanding the effect on the patient’s life span, data from several studies of long-term pancreas transplant recipients suggest that secondary diabetic complications can be halted or even improved. Most of these studies examined the effect of restoring euglycemia in nephropathy and the subsequent influence on renal function.

Effect on renal function. Kleinclauss et al²⁸ examined renal allograft function in type 1 diabetic recipients of living-donor kidney transplants. Comparing kidney allograft survival and function in patients who received a subsequent pancreas-after-kidney transplant vs those who did not, graft survival was superior after 5 years, and the estimated glomerular filtration rate was 10 mL/min higher in pancreas-after-kidney recipients.²⁸ This improvement in renal function was not seen immediately after the pancreas transplant but became evident more than 4 years after establishment of normoglycemia. Somewhat similarly, reversal of diabetic changes in native kidney biopsies has been seen 10 years after pancreas transplant.²⁹

Effect on neuropathy. In other studies, reversal of autonomic neuropathy and hypoglycemic unawareness and improvements in peripheral sensory-motor neuropathy have also been observed.^30–32

Effect on retinopathy. Improvements in early-stage nonproliferative diabetic retinopathy and laser-treated proliferative lesions have been seen, even within short periods of follow-up.³³ Other groups have shown a significantly higher proportion of improvement or stability of advanced diabetic retinopathy at 3 years after simultaneous pancreas-kidney transplant, compared with kidney transplant alone in patients with type 1 diabetes.³⁴

Effect on heart disease. Salutary effects on cardiovascular risk factors and amelioration of cardiac morphology and functional cardiac indices have been seen within the first posttransplant year.³⁵ Moreover, with longer follow-up (nearly 4 years), simultaneous pancreas-kidney recipients with functioning pancreas grafts were found to have less progression of coronary atherosclerosis than simultaneous pancreas-kidney recipients with early pancreas graft loss.³⁶ These data provide a potential pathophysiologic mechanism for the long-term survival advantage seen in uremic type 1 diabetic patients undergoing simultaneous pancreas-kidney transplant.

In the aggregate, these findings suggest that, in the absence of surgical and immunosuppression-related complications, a functioning pancreas allograft can alter the progress of diabetic complications. As an extension of these results, pancreas transplant done earlier in the course of diabetes may have an even greater impact.

Pancreas transplant is the only long-term diabetes treatment that consistently results in normal hemoglobin A_1c levels without the risk of severe hypoglycemia. Additionally, pancreas transplant may prevent, halt, or even reverse the complications of diabetes.

Here, we explore the indications, options, and outcomes of pancreas transplant as a treatment for diabetes mellitus.

DIABETES IS COMMON, AND OFTEN NOT WELL CONTROLLED

Diabetes mellitus affects more than 25 million people in the United States (8.3% of the population) and is the leading cause of kidney failure, nontraumatic lower-limb amputation, and adult-onset blindness. In 2007, nearly $116 billion was spent on diabetes treatment, not counting another $58 billion in indirect costs such as disability, work loss, and premature death.¹

Only about half of patients achieve hemoglobin A_1c < 7% with medical therapy

Despite the tremendous expenditure in human, material, and financial resources, only about 50% of patients achieve their diabetes treatment goals. In 2013, a large US population-based study­² reported that 52.2% of patients were achieving the American Diabetes Association treatment goal of hemoglobin A_1c lower than 7%. A similar study in South Korea³ found that 45.6% were at this goal.

Most of the patients in these studies had type 2 diabetes, and the data suggested that attaining glycemic goals is more difficult in insulin-treated patients. Studies of patients with type 1 diabetes found hemoglobin A_1c levels lower than 7% in only 8.1% of hospitalized patients with type 1 diabetes, and in only 13% in an outpatient diabetes clinic.^4,5

YET RATES OF PANCREAS TRANSPLANT ARE DECLINING

Pancreas transplant was first performed more than 40 years ago at the University of Minnesota.⁶ Since then, dramatic changes in immunosuppression, organ preservation, surgical technique, and donor and recipient selection have brought about significant progress.

Currently, more than 13,000 patients are alive with a functioning pancreas allograft. After reaching a peak in 2004, the annual number of pancreas transplants performed in the United States has declined steadily, whereas the procedure continues to increase in popularity outside North America.⁷ The primary reason for the decline is recognition of donor factors that lead to success—surgeons are refusing to transplant organs they might have accepted previously, because experience suggests they would yield poor results. In the United States, 1,043 pancreas transplants were performed in 2012, and more than 3,100 patients were on the waiting list.⁸

Islet cell transplant—a different procedure involving harvesting, encapsulating, and implanting insulin-producing beta cells—has not gained widespread application due to very low long-term success rates.

THREE CATEGORIES OF PANCREAS TRANSPLANT

Pancreas transplant can be categorized according to whether the patient is also receiving or has already received a kidney graft (Table 1).

Simultaneous kidney and pancreas transplant is performed in patients who have type 1 diabetes with advanced chronic kidney disease due to diabetic nephropathy. This remains the most commonly performed type, accounting for 79% of all pancreas transplants in 2012.⁸

Pancreas-after-kidney transplant is most often done after a living-donor kidney transplant. This procedure accounted for most of the increase in pancreas transplants during the first decade of the 2000s. However, the number of these procedures has steadily decreased since 2004, and in 2012 accounted for only 12% of pancreas transplants.⁸

Pancreas transplant alone is performed in nonuremic diabetic patients who have labile blood sugar control. Performed in patients with preserved renal function but severe complications of “brittle” diabetes, such as hypoglycemic unawareness, this type accounts for 8% of pancreas transplants.⁹

Indications for pancreas transplant

A small number of these procedures are done for indications unrelated to diabetes mellitus. In most of these cases, the pancreas is transplanted as part of a multivisceral transplant to facilitate the technical (surgical) aspect of the procedure—the pancreas, liver, stomach, gallbladder, and part of the intestines are transplanted en bloc to maintain the native vasculature. Very infrequently, pancreas transplant is done to replace exocrine pancreatic function.

A small, select group of patients with type 2 diabetes and low body mass index (BMI) may be eligible for pancreas transplant, and they accounted for 8.2% of active candidates in 2012.⁸ However, most pancreas transplants are performed in patients with type 1 diabetes.

WHAT MAKES A GOOD ALLOGRAFT?

Pancreas allografts are procured as whole organs from brain-dead organ donors. Relatively few pancreas allografts (3.1% in 2012) are from cardiac-death donors, because of concern about warm ischemic injury during the period of circulatory arrest.⁸

Proper donor selection is critical to the success of pancreas transplant, as donor factors including medical history, age, BMI, and cause of death can significantly affect the outcome. In general, transplant of a pancreas allograft from a young donor (age < 30) with excellent organ function, low BMI, and traumatic cause of death provides the best chance of success.

The Pancreas Donor Risk Index (PDRI)¹⁰ was developed after analysis of objective donor criteria, transplant type, and ischemic time in grafts transplanted between 2000 and 2006. One-year graft survival was directly related to the PDRI and ranged between 77% and 87% in recipients of “standard” pancreas allografts (PDRI score of 1.0). Use of grafts from the highest (worst) three quintiles of PDRI (PDRI score > 1.16) was associated with 1-year graft survival rates of 67% to 82%, significantly inferior to that seen with “higher- quality” grafts, again emphasizing the need for rigorous donor selection.¹⁰

In addition to these objective measures, visual assessment of pancreas quality at the time of procurement remains an equally important predictor of success. Determination of subjective features, such as fatty infiltration and glandular fibrosis, requires surgical experience developed over several years. In a 2010 analysis, dissatisfaction with the quality of the donor graft on inspection accounted for more than 80% of refusals of potential pancreas donors.¹¹ These studies illustrate an ill-defined aspect of pancreas transplant, ie, even when the pancreas donor is perceived to be suitable, the outcome may be markedly different.

SURGICAL COMPLICATIONS

Surgical complications have long been considered a limiting factor in the growth of pancreas transplant. Technical failure or loss of the graft within 90 days is most commonly due to graft thrombosis, leakage of the enteric anastomosis, or severe peripancreatic infection. The rate of technical failure has declined across all recipient categories and is currently about 9%.⁸

DO RECIPIENT FACTORS AFFECT OUTCOMES?

As mentioned above, the PDRI identifies donor factors that influence the 1-year graft survival rate. Recipient factors are also thought to play a role, although the influence of these factors has not been consistently demonstrated.

Humar et al¹⁵ found that recipient obesity (defined in this study as BMI > 25 kg/m²) and donor age over 40 were risk factors for early laparotomy after pancreas transplant.¹⁵ Moreover, patients undergoing early laparotomy had poorer graft survival outcomes.

This finding was reinforced by an analysis of 5,725 primary simultaneous pancreas-kidney recipients between 2000 and 2007. Obesity (BMI 30 ≥ kg/m²) was associated with increased rates of patient death, pancreas graft loss, and kidney graft loss at 3 years.¹⁶

More recently, Finger et al¹⁷ did not find a statistically significant association between recipient BMI and technical failure, but they did notice a trend toward increased graft loss with a BMI greater than 25 kg/m². Similarly, others have not found a clear adverse association between recipient BMI and pancreas graft survival.

Intuitively, obesity and other recipient factors such as age, vascular disease, duration of diabetes, and dialysis should influence pancreas graft survival but have not been shown in analyses to carry an adverse effect.¹⁸ The inability to consistently find adverse effects of recipient characteristics is most likely due to the relative similarity between the vast majority of pancreas transplant recipients and the relatively small numbers of adverse events. In 98 consecutive pancreas transplants at our center between 2009 and 2014, the technical loss rate was 1.8% (unpublished data).

Acute rejection most commonly occurs during the first year and is usually reversible. More than 1 year after transplant, graft loss is due to chronic rejection, and death is usually from underlying cardiovascular disease.

The immunosuppressive regimens used in pancreas transplant are similar to those in kidney transplant. Since the pancreas is considered to be more immunogenic than other organs, most centers employ a strategy of induction immunosuppression with T-cell–depleting or interleukin 2-receptor antibodies. Maintenance immunosuppression consists of a calcineurin inhibitor (tacrolimus or cyclosporine), an antimetabolite (mycophenolate), and a corticosteroid.⁸

Immunosuppressive complications occur at a rate similar to that seen in other solid-organ transplants and include an increased risk of opportunistic infection and malignancy. The risk of these complications must be balanced against the patient’s risk of health decline with dialysis and insulin-based therapies.

OVERALL OUTCOMES ARE GOOD

The success rate of pancreas transplant is currently at its highest since the inception of the procedure. The unadjusted patient survival rate for all groups is over 96% at 1 year, and over 80% at 5 years.⁸ One-year patient survival after pancreas transplant alone, at better than 96%, is the highest of all organ transplant procedures.⁹

Patient survival 1 year after pancreas-alone transplant is > 96%

Several recently published single-center reviews of pancreas transplant since 2000 report patient survival rates of 96% to 100% at 1 year and 88% to 100% at 5 years.^19–22 This variability is likely closely linked to donor and recipient selection, as centers performing smaller numbers of transplants tend to be more selective and, in turn, report higher patient survival rates.^19,21

Long-term patient survival outcomes can be gathered from larger, registry-based reviews, accepting limitations in assessing causes of patient death. Siskind et al²³ analyzed the outcomes of 20,854 US pancreas transplants done between 1996 and 2012 and found the 10-year patient survival rate ranged from 43% to 77% and was highly dependent on patient age at the time of the procedure.²³ Patient survival after transplant must be balanced against the generally poor long-term survival prospects of diabetic patients on dialysis.

By type of transplant, pancreas graft survival rates at 1 year are 89% for simultaneous pancreas-kidney transplant, 86% for pancreas-after-kidney transplant, and 84% for pancreas-alone transplant. Graft survival rates at 5 years are 71% for simultaneous pancreas-kidney transplant, 65% for pancreas-after-kidney transplant, and 58% for pancreas-alone transplant.^8,9

Simultaneous pancreas-kidney transplant has been shown to improve the survival rate compared with cadaveric kidney transplant alone in patients with type 1 diabetes and chronic kidney disease.^24,25 The survival benefit of isolated pancreas transplant (after kidney transplant and alone) is not evident at 4-year follow-up compared with patients on the waiting list. However, the benefit for the individual patient must be considered by weighing the incapacities experienced with insulin-based treatments against the risks of surgery and immunosuppression.^26,27 For patients who have experienced frequent and significant hypoglycemic episodes, particularly those requiring third-party assistance, pancreas transplant can be a lifesaving procedure.

Effects on secondary diabetic complications

Notwithstanding the effect on the patient’s life span, data from several studies of long-term pancreas transplant recipients suggest that secondary diabetic complications can be halted or even improved. Most of these studies examined the effect of restoring euglycemia in nephropathy and the subsequent influence on renal function.

Effect on renal function. Kleinclauss et al²⁸ examined renal allograft function in type 1 diabetic recipients of living-donor kidney transplants. Comparing kidney allograft survival and function in patients who received a subsequent pancreas-after-kidney transplant vs those who did not, graft survival was superior after 5 years, and the estimated glomerular filtration rate was 10 mL/min higher in pancreas-after-kidney recipients.²⁸ This improvement in renal function was not seen immediately after the pancreas transplant but became evident more than 4 years after establishment of normoglycemia. Somewhat similarly, reversal of diabetic changes in native kidney biopsies has been seen 10 years after pancreas transplant.²⁹

Effect on neuropathy. In other studies, reversal of autonomic neuropathy and hypoglycemic unawareness and improvements in peripheral sensory-motor neuropathy have also been observed.^30–32

Effect on retinopathy. Improvements in early-stage nonproliferative diabetic retinopathy and laser-treated proliferative lesions have been seen, even within short periods of follow-up.³³ Other groups have shown a significantly higher proportion of improvement or stability of advanced diabetic retinopathy at 3 years after simultaneous pancreas-kidney transplant, compared with kidney transplant alone in patients with type 1 diabetes.³⁴

Effect on heart disease. Salutary effects on cardiovascular risk factors and amelioration of cardiac morphology and functional cardiac indices have been seen within the first posttransplant year.³⁵ Moreover, with longer follow-up (nearly 4 years), simultaneous pancreas-kidney recipients with functioning pancreas grafts were found to have less progression of coronary atherosclerosis than simultaneous pancreas-kidney recipients with early pancreas graft loss.³⁶ These data provide a potential pathophysiologic mechanism for the long-term survival advantage seen in uremic type 1 diabetic patients undergoing simultaneous pancreas-kidney transplant.

In the aggregate, these findings suggest that, in the absence of surgical and immunosuppression-related complications, a functioning pancreas allograft can alter the progress of diabetic complications. As an extension of these results, pancreas transplant done earlier in the course of diabetes may have an even greater impact.

The immunosuppressed state of liver transplant recipients makes them vulnerable to infections after surgery.¹ These infections are directly correlated with the net state of immunosuppression. Higher levels of immunosuppression mean a higher risk of infection, with rates of infection typically highest in the early posttransplant period.

Common infections during this period include operative and perioperative nosocomial bacterial and fungal infections, reactivation of latent infections, and invasive fungal infections such as candidiasis, aspergillosis, and pneumocystosis. Donor-derived infections also must be considered. As time passes and the level of immunosuppression is reduced, liver recipients are less prone to infection.¹

The risk of infection can be minimized by appropriate antimicrobial prophylaxis, strategies for safe living after transplant,² vaccination,³ careful balancing of immunosuppressive therapy,⁴ and thoughtful donor selection.⁵ Drug-drug interactions are common and must be carefully considered to minimize the risk.

This review highlights common infectious complications encountered after liver transplant.

INTRA-ABDOMINAL INFECTIONS

Intra-abdominal infections are common in the early postoperative period.^6,7

Risk factors include:

• Pretransplant ascites
• Posttransplant dialysis
• Wound infection
• Reoperation⁸
• Hepatic artery thrombosis
• Roux-en-Y choledochojejunostomy anastomosis.⁹

Signs that may indicate intra-abdominal infection include fever, abdominal pain, leukocytosis, and elevated liver enzymes. But because of their immunosuppressed state, transplant recipients may not manifest fever as readily as the general population. They should be evaluated for cholangitis, peritonitis, biloma, and intra-abdominal abscess.

Organisms. Intra-abdominal infections are often polymicrobial. Enterococci, Staphylococcus aureus, gram-negative species including Pseudomonas, Klebsiella, and Acinetobacter, and Candida species are the most common pathogens. Strains are often resistant to multiple drugs, especially in patients who received antibiotics in the weeks before transplant.^8,10

Liver transplant recipients are also particularly susceptible to Clostridium difficile-associated colitis as a result of immunosuppression and frequent use of antibiotics perioperatively and postoperatively.¹¹ The spectrum of C difficile infection ranges from mild diarrhea to life-threatening colitis, and the course in liver transplant patients tends to be more complicated than in immunocompetent patients.¹²

Diagnosis. Intra-abdominal infections should be looked for and treated promptly, as they are associated with a higher mortality rate, a greater risk of graft loss, and a higher incidence of retransplant.^6,10 Abdominal ultrasonography or computed tomography (CT) can confirm the presence of fluid collections.

Treatment. Infected collections can be treated with percutaneous or surgical drainage and antimicrobial therapy. In the case of biliary tract complications, retransplant or surgical correction of biliary leakage or stenosis decreases the risk of death.⁶

Suspicion should be high for C difficile-associated colitis in cases of posttransplant diarrhea. C difficile toxin stool assays help confirm the diagnosis.¹² Oral metronidazole is recommended in mild to moderate C difficile infection, with oral vancomycin and intravenous metronidazole reserved for severe cases. Colectomy may be necessary in patients with toxic megacolon.

CYTOMEGALOVIRUS INFECTION

Cytomegalovirus is an important opportunistic pathogen in liver transplant recipients.¹³ It causes a range of manifestations, from infection (viremia with or without symptoms) to cytomegalovirus syndrome (fever, malaise, and cell-line cytopenias) to tissue-invasive disease with end-organ disease.¹⁴ Without preventive measures and treatment, cytomegalovirus disease can increase the risk of morbidity, allograft loss and death.^15,16

Risk factors for cytomegalovirus infection (Table 1) include:

• Discordant serostatus of the donor and recipient (the risk is highest in seronegative recipients of organs from seropositive donors)
• Higher levels of immunosuppression, especially when antilymphocyte antibodies are used
• Treatment of graft rejection
• Coinfection with other human herpesviruses, such as Epstein-Barr virus.^4,17

Preventing cytomegalovirus infection

The strategy to prevent cytomegalovirus infection depends on the serologic status of the donor and recipient and may include antiviral prophylaxis or preemptive treatment (Table 2).¹⁸

Prophylaxis involves giving antiviral drugs during the early high-risk period, with the goal of preventing the development of cytomegalovirus viremia. The alternative preemptive strategy emphasizes serial testing for cytomegalovirus viremia, with the goal of intervening with antiviral medications while viremia is at a low level, thus avoiding potential progression to cytomegalovirus disease. Both strategies have pros and cons that should be considered by each transplant center when setting institutional policy.

A prophylactic approach seems very effective at preventing both infection and disease from cytomegalovirus and has been shown to reduce graft rejection and the risk of death.¹⁸ It is preferred in cytomegalovirus-negative recipients when the donor was cytomegalovirus-positive—a high-risk situation.¹⁹ However, these patients are also at higher risk of late-onset cytomegalovirus disease. Higher cost and potential drug toxicity, mainly neutropenia from ganciclovir-based regimens, are additional considerations.

Preemptive treatment, in contrast, reserves drug treatment for patients who are actually infected with cytomegalovirus, thus resulting in fewer adverse drug events and lower cost; but it requires regular monitoring. Preemptive methods, by definition, cannot prevent infection, and with this strategy tissue-invasive disease not associated with viremia does occasionally occur.²⁰ As such, patients with a clinical presentation that suggests cytomegalovirus but have negative results on blood testing should be considered for tissue biopsy with culture and immunohistochemical stain.

The most commonly used regimens for antiviral prophylaxis and treatment in liver transplant recipients are intravenous ganciclovir and oral valganciclovir.²¹ Although valganciclovir is the most commonly used agent in this setting because of ease of administration, it has not been approved by the US Food and Drug Administration in liver transplant patients, as it was associated with higher rates of cytomegalovirus tissue-invasive disease.^22–24 Additionally, drug-resistant cytomegalovirus strains have been associated with valganciclovir prophylaxis in cytomegalovirus-negative recipients of solid organs from cytomegalovirus-positive donors.²⁵

Prophylaxis typically consists of therapy for 3 months from the time of transplant. In higher-risk patients (donor-positive, recipient-negative), longer courses of prophylaxis have been extrapolated from data in kidney transplant recipients.²⁶ Extension or reinstitution of prophylaxis should also be considered in liver transplant patients receiving treatment for rejection with antilymphocyte therapy.

Routine screening for cytomegalovirus is not recommended while patients are receiving prophylaxis. High-risk patients who are not receiving prophylaxis should be monitored with nucleic acid or pp65 antigenemia testing as part of the preemptive strategy protocol.

Treatment of cytomegalovirus disease

Although no specific threshold has been established, treatment is generally indicated if a patient has a consistent clinical syndrome, evidence of tissue injury, and persistent or increasing viremia.

Treatment involves giving antiviral drugs and also reducing the level of immunosuppression, if possible, until symptoms and viremia have resolved.

The choice of antiviral therapy depends on the severity of disease. Intravenous ganciclovir (5 mg/kg twice daily adjusted for renal impairment) or oral valganciclovir (900 mg twice daily, also renally dose-adjusted when necessary) can be used for mild to moderate disease if no significant gastrointestinal involvement is reported. Intravenous ganciclovir is preferred for patients with more severe disease or gastrointestinal involvement. The minimum duration of treatment is 2 weeks and may need to be prolonged until both symptoms and viremia completely resolve.¹⁸

Drug resistance can occur and should be considered in patients who have a history of prolonged ganciclovir or valganciclovir exposure who do not clinically improve or have persistent or rising viremia. In such cases, genotype assays are helpful, and initiation of alternative therapy should be considered. Mutations conferring resistance to ganciclovir are often associated with cross-resistance to cidofovir. Cidofovir can therefore be considered only when genotype assays demonstrate specific mutations conferring an isolated resistance to ganciclovir.²⁷ The addition of foscarnet to the ganciclovir regimen or substitution of foscarnet for ganciclovir are accepted approaches.

Although cytomegalovirus hyperimmunoglobulin has been used in prophylaxis and invasive disease treatment, its role in the management of ganciclovir-resistant cytomegalovirus infections remains controversial.²⁸

EPSTEIN-BARR VIRUS POSTTRANSPLANT LYMPHOPROLIFERATIVE DISEASE

Epstein-Barr virus-associated posttransplant lymphoproliferative disease is a spectrum of disorders ranging from an infectious mononucleosis syndrome to aggressive malignancy with the potential for death and significant morbidity after liver transplant.²⁹ The timeline of risk varies, but the disease is most common in the first year after transplant.

Risk factors for this disease (Table 1) are:

• Primary Epstein-Barr virus infection
• Cytomegalovirus donor-recipient mismatch
• Cytomegalovirus disease
• Higher levels of immunosuppression, especially with antilymphocyte antibodies.³⁰

The likelihood of Epstein-Barr virus playing a contributing role is lower in later-onset posttransplant lymphoproliferative disease. Patients who are older at the time of transplant, who receive highly immunogenic allografts including a liver as a component of a multivisceral transplant, and who receive increased immunosuppression to treat rejection are at even greater risk of late posttransplant lymphoproliferative disease.³¹ This is in contrast to early posttransplant lymphoproliferative disease, which is seen more commonly in children as a result of primary Epstein-Barr virus infection.

Recognition and diagnosis. Heightened suspicion is required when considering posttransplant lymphoproliferative disease, and careful evaluation of consistent symptoms and allograft dysfunction are required.

Clinically, posttransplant lymphoproliferative disease should be suspected if a liver transplant recipient develops unexplained fever, weight loss, lymphadenopathy, or cell-line cytopenias.^30,32 Other signs and symptoms may be related to the organ involved and may include evidence of hepatitis, pneumonitis, and gastrointestinal disease.³¹

Adjunctive diagnostic testing includes donor and recipient serology to characterize overall risk before transplantation and quantification of Epstein-Barr viral load, but confirmation relies on tissue histopathology.

Treatment focuses on reducing immunosuppression.^30,32 Adding antiviral agents does not seem to improve outcome in all cases.³³ Depending on clinical response and histologic classification, additional therapies such as anti-CD20 humanized chimeric monoclonal antibodies, surgery, radiation, and conventional chemotherapy may be required.³⁴

Preventive approaches remain controversial. Chemoprophylaxis with an antiviral such as ganciclovir is occasionally used but has not been shown to consistently decrease rates of posttransplant lymphoproliferative disease. These agents may act in an indirect manner, leading to decreased rates of cytomegalovirus infection, a major cofactor for posttransplant lymphoproliferative disease.²⁴

Although oral valganciclovir is used more than intravenous ganciclovir, it is not approved for liver transplant patients

Passive immunoprophylaxis with immunoglobulin targeting cytomegalovirus has shown to decrease rates of non-Hodgkin lymphoma from posttransplant lymphoproliferative disease in renal transplant recipients in the first year after transplant,³⁵ but data are lacking regarding its use in liver transplant recipients. Monitoring of the viral load and subsequent reduction of immunosuppression remain the most efficient measures to date.³⁶

FUNGAL INFECTIONS

Candida species account for more than half of fungal infections in liver transplant recipients.³⁷ However, a change has been noted in the past 20 years, with a decrease in Candida infections accompanied by an increase in Aspergillus infections.³⁸ Endemic mycoses such as coccidioidomycosis, blastomycosis, and histoplasmosis should be considered with the appropriate epidemiologic history or if disease develops early after transplant and the donor came from a highly endemic region.³⁹Cryptococcus may also be encountered.

Diagnosis. One of the most challenging aspects of fungal infection in liver transplant recipients is timely diagnosis. Heightened suspicion and early biopsy for pathological and microbiological confirmation are necessary. Although available noninvasive diagnostic tools often lack specificity, early detection of fungal markers may be of great use in guiding further diagnostic workup or empiric treatment in the critically ill.

Noninvasive tests include galactomannan, cryptococcal antigen, histoplasma antigen, (1-3)-beta-D-glucan assay and various antibody tests. Galactomannan testing has been widely used to aid in the diagnosis of invasive aspergillosis. Similarly, the (1-3)-beta-D-glucan assay is a non–culture-based tool for diagnosing and monitoring the treatment of invasive fungal infections. However, a definite diagnosis cannot be made on the basis of a positive test alone.⁴⁰ The complementary diagnostic characteristics of combining noninvasive assays have yet to be fully elucidated.⁴¹ Cultures and tissue histopathology are also used when possible.

Treatment is based on targeted specific antifungal drug therapy and reduction of immunosuppressive therapy, when possible. The choice of antifungal agent varies with the pathogen, the site of involvement, and the severity of the disease. A focus on potential drug interactions, their management, and therapeutic drug monitoring when using antifungal medications is essential in the posttransplant period. Combination therapy can be considered in some situations to enhance synergy. The following sections discuss in greater detail Candida species, Aspergillus species, and Pneumocystis jirovecii infections.

Candida infections

Candidiasis after liver transplant is typically nosocomial, especially when diagnosed during the first 3 months (Table 3).³⁷

Risk factors for invasive candidiasis include perioperative colonization, prolonged operative time, retransplant, greater transfusion requirements, and postoperative renal failure.^37,42,43 Invasive candidiasis is of concern for its effects on morbidity, mortality, and cost of care.^43–46

Organisms. The frequency of implicated species, in particular those with a natural resistance to fluconazole, differs in various reports.^37,45,46Candida albicans remains the most commonly isolated pathogen; however, non-albicans species including those resistant to fluconazole have been reported more frequently and include Candida glabrata and Candida krusei.^47,48

Signs and diagnosis. Invasive candidiasis in liver transplant recipients generally manifests itself in catheter-related blood stream infections, urinary tract infections, or intra-abdominal infections. Diagnosis can be made by isolating Candida from blood cultures, recovering organisms in culture of a normally sterile site, or finding direct microscopic evidence of the fungus on tissue specimens.⁴⁹

Disseminated candidiasis refers to the involvement of distant anatomic sites. Clinical manifestations may cause vision changes, abdominal pain or skin nodules with findings of candidemia, hepatosplenic abscesses, or retinal exudates on funduscopy.⁴⁹

Treatment of invasive candidiasis in liver recipients often involves antifungal therapy and reduction of immunosuppression. Broad-spectrum antifungals are initially advocated in an empirical approach to cover fluconazole-resistant strains of the non-albicans subgroups.⁵⁰ Depending on antifungal susceptibility, treatment can later be adjusted.

Fluconazole remains the agent of choice in most C albicans infections.⁴⁷ However, attention should be paid to the possibility of resistance in patients who have received fluconazole prophylaxis within the past 30 days. Additional agents used in treatment may include echinocandins, amphotericin, and additional azoles.

Antifungal prophylaxis is recommended in high-risk liver transplant patients, although its optimal duration remains undetermined.⁴⁴ Antifungal prophylaxis has been associated with decreased incidence of both superficial and invasive candidiasis.⁵¹

Aspergillus infection

Aspergillus, the second most common fungal pathogen, has become a more common concern in liver transplant recipients. Aspergillus fumigatus is the most frequently encountered species.^38,52

Risk factors. These infections typically occur in the first year, during intense immunosuppression. Retransplant, renal failure, and fulminant hepatic failure are major risk factors.⁵² In the presence of risk factors and a suggestive clinical setting, invasive aspergillosis should be considered and the diagnosis pursued.

Diagnosis is suggested by positive findings on CT accompanied by lower respiratory tract symptoms, focal lesions on neuroimaging, or demonstration of the fungus on cultures.⁴⁹ However, Aspergillus is rarely grown in blood culture. The galactomannan antigen is a noninvasive test that can provide supporting evidence for the diagnosis.^41,52 False-positive results do occur in the setting of certain antibiotics and cross-reacting fungi.⁵³

Treatment consists of antifungal therapy and immunosuppression reduction.⁵²

Candida accounts for more than half of fungal infections in liver transplant recipients, but Aspergillus is gaining

Voriconazole is the first-line agent for invasive aspergillosis. Monitoring for potential drug-drug interactions and side effects is required.^54,55 Amphotericin B is considered a second-line choice due to toxicity and lack of an oral formulation. In refractory cases, combined antifungal therapy could be considered.⁵² The duration of treatment is generally a minimum of 12 weeks.

Prophylaxis. Specific prophylaxis against invasive aspergillosis is not currently recommended; however, some authors suggest a prophylactic approach using echinocandins or liposomal amphotericin B in high-risk patients.^51,52 Aspergillosis is associated with a considerable increase in mortality in liver transplant recipients, which highlights the importance of timely management.^52,56

Pneumocystis jirovecii

P jirovecii remains a common opportunistic pathogen in people with impaired immunity, including transplant and human immunodeficiency virus patients.

Prophylaxis. Widespread adoption of antimicrobial prophylaxis by transplant centers has decreased the rates of P jirovecii infection in liver transplant recipients.^57,58 Commonly used prophylactic regimens after liver transplantation include a single-strength trimeth­oprim-sulfamethoxazole tablet daily or a double-strength tablet three times per week for a minimum of 6 to 12 months after transplant. Atovaquone and dapsone can be used as alternatives in cases of intolerance to tri­methoprim-sulfamethoxazole (Table 2).

Inhaled pentamidine is clearly inferior and should be used only when the other medications are contraindicated.⁵⁹

Signs and diagnosis. P jirovecii pneumonia is characterized by fever, cough, dyspnea, and chest pain. Insidious hypoxemia, abnormal chest examination, and bilateral interstitial pneumonia on chest radiography are common.

CT may be more sensitive than chest radiography.⁵⁷ Findings suggestive of P jirovecii pneumonia on chest CT are extensive bilateral and symmetrical ground-glass attenuations. Other less-characteristic findings include upper lobar parenchymal opacities and spontaneous pneumothorax.^57,60

The serum (1,3)-beta-D-glucan assay derived from major cell-wall components of P jiro­vecii might be helpful. Studies report a sensitivity for P jirovecii pneumonia as high as 96% and a negative predictive value of 99.8%.^61,62

Definitive diagnosis requires identification of the pathogen. Routine expectorated sputum sampling is generally associated with a poor diagnostic yield. Bronchoscopy and bronchoalveolar lavage with silver or fluorescent antibody staining of samples, polymerase chain reaction testing, or both significantly improves diagnosis. Transbronchial or open lung biopsy are often unnecessary.⁵⁷

Treatment. Trimethoprim-sulfamethoxazole is the first-line agent for treating P jirovecii pneumonia.⁵⁷ The minimum duration of treatment is 14 days, with extended courses for severe infection.

Intravenous pentamidine or clindamycin plus primaquine are alternatives for patients who cannot tolerate trimethoprim-sulfamethoxazole. The major concern with intravenous pentamidine is renal dysfunction. Hypoglycemia or hyperglycemia, neutropenia, thrombocytopenia, nausea, dysgeusia, and pancreatitis may also occur.⁶³

Atovaquone might also be beneficial in mild to moderate P jirovecii pneumonia. The main side effects include skin rashes, gastrointestinal intolerance, and elevation of transaminases.⁶⁴

A corticosteroid (40–60 mg of prednisone or its equivalent) may be beneficial in conjunction with antimicrobial therapy in patients with significant hypoxia (partial pressure of arterial oxygen < 70 mm Hg on room air) in decreasing the risk of respiratory failure and need for intubation.

With appropriate and timely antimicrobial prophylaxis, cases of P jirovecii pneumonia should continue to decrease.

TUBERCULOSIS

Development of tuberculosis after transplantation is a catastrophic complication, with mortality rates of up to 30%.⁶⁵ Most cases of posttransplant tuberculosis represent reactivation of latent disease.⁶⁶ Screening with tuberculin skin tests or interferon-gamma-release assays is recommended in all liver transplant candidates. Chest radiography before transplant is necessary when assessing a positive screening test.⁶⁷

The optimal management of latent tuberculosis in these cases remains controversial. Patients at high risk or those with positive screening results on chest radiography warrant treatment for latent tuberculosis infection with isoniazid unless contraindicated.^67,68

The ideal time to initiate prophylactic isoniazid therapy is unclear. Some authors suggest delaying it, as it might be associated with poor tolerance and hepatotoxicity.⁶⁹ Others have found that early isoniazid use was not associated with negative outcomes.⁷⁰

Risk factors for symptomatic tuberculosis after liver transplant include previous infection with tuberculosis, intensified immunosuppression (especially anti-T-lymphocyte therapies), diabetes mellitus, and other co-infections (Table 1).⁷¹

The increased incidence of atypical presentations in recent years makes the diagnosis of active tuberculosis among liver transplant recipients challenging. Sputum smears can be negative due to low mycobacterial burdens, and tuberculin skin testing and interferon-gamma-release assays may be falsely negative due to immunosuppression.⁶⁷

Treatment of active tuberculosis consists initially of a four-drug regimen using isoniazid, rifampin, pyrazinamide, and ethambutol for 2 months. Adjustments are made in accordance with culture and sensitivity results. Treatment can then be tapered to two drugs (isoniazid and rifampin) for a minimum of 4 additional months. Prolonged treatment may be required in instances of extrapulmonary or disseminated disease.^65,72

Tuberculosis treatment can be complicated by hepatotoxicity in liver transplant recipients because of direct drug effects and drug-drug interactions with immunosuppressive agents. Close monitoring for rejection and hepatotoxicity is therefore imperative while liver transplant recipients are receiving antituberculosis therapy. Drug-drug interactions may also be responsible for marked reductions in immunosuppression levels, especially with regimens containing rifampin.⁷¹ Substitution of rifabutin for rifampin reduces the effect of drug interactions.⁶⁶

VIRAL HEPATITIS

Hepatitis B virus

Hepatitis B virus-related end-stage liver disease and hepatocellular carcinoma are common indications for liver transplant in Asia. It is less common in the United States and Europe, accounting for less than 10% of all liver transplant cases. Prognosis is favorable in recipients undergoing liver transplant for hepatitis B virus, with excellent survival rates. Prevention of reinfection is crucial in these patients.

Treatment with combination antiviral agents and hepatitis B immunoglobulin (HBIG) is effective.⁷³ Lamivudine was the first nucleoside analogue found to be effective against hepatitis B virus. Its low cost and relative safety are strong arguments in favor of its continued use in liver transplant recipients.⁷⁴ In patients without evidence of hepatitis B viral replication at the time of transplant, monotherapy with lamivudine has led to low recurrence rates, and adefovir can be added to control resistant viral strains.⁷⁵

Widespread adoption of prophylaxis has decreased the rate of P jirovecii infection in liver transplant recipients

The frequent emergence of resistance with lamivudine favors newer agents such as entecavir or tenofovir. These nucleoside and nucleotide analogues have a higher barrier to resistance, and thus resistance to them is rare. They are also more efficient, potentially allowing use of an HBIG-sparing protocol.⁷⁶ However, they are associated with a higher risk of nephrotoxicity and require dose adjustments in renal insufficiency. Data directly comparing entecavir and tenofovir are scarce.

Prophylaxis. Most studies support an individualized approach for prevention of hepatitis B virus reinfection. High-risk patients, ie, those positive for HBe antigen or with high viral loads (> 100,000 copies/mL) are generally treated with both HBIG and antiviral agents.⁷⁷ Low-risk patients are those with a negative HBe antigen, low hepatitis B virus DNA levels, hepatitis B virus-related acute liver failure, and cirrhosis resulting from coinfection with both hepatitis B and hepatitis D virus.⁷⁵ In low-risk patients, discontinuation of HBIG after 1 to 2 years of treatment is appropriate, and long-term prophylaxis with antiviral agents alone is an option. However, levels of hepatitis B DNA should be monitored closely.^78,79

Hepatitis C virus

Recurrence of hepatitis C virus infection is the rule among patients who are viremic at the time of liver transplant.^80,81 Most of these patients will show histologic evidence of recurrent hepatitis within the first year after liver transplant. It is often difficult to distinguish between the histopathological appearance of a recurrent hepatitis C virus infection and acute cellular rejection.

Progression to fibrosis and subsequently cirrhosis and decompensation is highly variable in hepatitis C virus-infected liver transplant recipients. Diabetes, insulin resistance, and possibly hepatitis steatosis have been associated with a rapid progression to advanced fibrosis. The contribution of immunosuppression to the progression of hepatitis C virus remains an area of active study. Some studies point to antilymphocyte immunosuppressive agents as a potential cause.⁸² Liver biopsy is a useful tool in this situation. It allows monitoring of disease severity and progression and may distinguish recurrent hepatitis C virus disease from other causes of liver enzyme elevation.

The major concern with the recurrence of hepatitis C virus infection after liver transplant is allograft loss. Rates of patient and graft survival are reduced in infected patients compared with hepatitis C virus-negative patients.^83,84 Prophylactic antiviral therapy has no current role in the management of hepatitis C virus disease. Those manifesting moderate to severe necroinflammation or mild to moderate fibrosis indicative of progressive disease should be treated.^81,85

Sustained viral clearance with antiviral agents confers a graft survival benefit.

The combination of peg-interferon and weight-based ribavirin has been the standard of treatment but may be associated with increased rates of rejection.^86,87 The sustained virologic response rates for hepatitis C virus range from 60% in genotypes 4, 5, and 6 after 48 weeks of treatment to 60% to 80% in genotypes 2 and 3 after 24 weeks, but only about 30% in genotype 1.⁸⁸

The major concern with hepatitis C recurrence after liver transplant is allograft loss

Treatment with the newer agents, especially protease inhibitors, in genotype 1 (peg-interferon, ribavirin, and either telaprevir or boceprevir) has been evaluated. Success rates reaching 70% have been achieved.⁸⁹ Adverse effects can be a major setback. Serious complications include severe anemia, renal dysfunction, increased risk of infection, and death.

Triple therapy should be carefully considered in liver transplant patients with genotype 1 hepatitis C virus.⁹⁰ Significant drug-drug interactions are reported between hepatitis C virus protease inhibitors and immunosuppression regimens. Additional new oral direct- acting antivirals have been investigated. They bring promising advances in hepatitis C virus treatment and pave the way for interferon-free regimens with pangenotypic activity.

IMMUNIZATION

Immunization can decrease the risk of infectious complications in liver transplant recipients, as well as in close contacts and healthcare professionals.³

Influenza. Pretransplant influenza vaccine and posttransplant annual influenza vaccines are necessary.

Pneumococcal immunization should additionally be provided prior to transplant and repeated every 3 to 5 years thereafter.^3,91

A number of other vaccinations should also be completed before transplant, including the hepatitis A and B vaccines and the tetanus/diphtheria/acellular pertussis vaccines. However, these vaccinations have not been shown to be detrimental to patients after transplant.⁹¹

Varicella and zoster vaccines should be given before liver transplant—zoster in patients over age 60, and varicella in patients with no immunity. Live vaccines, including varicella and zoster vaccines, are contraindicated after liver transplant.³

Human papillomavirus. The bivalent human papillomavirus vaccine can be given before transplant in females ages 9 to 26; the quadrivalent vaccine is beneficial in those ages 9 to 26 and in women under age 45.^3,91

IMMUNOSUPPRESSION CARRIES RISK OF INFECTION

Most liver transplant patients require prolonged immunosuppressive therapy. This comes with an increased risk of new or recurrent infections, potentially causing death and significant morbidity.

Evaluation of existing risk factors, appropriate prophylaxis and immunization, timely diagnosis, and treatment of such infections are therefore essential steps for the successful management of liver transplant recipients.

The immunosuppressed state of liver transplant recipients makes them vulnerable to infections after surgery.¹ These infections are directly correlated with the net state of immunosuppression. Higher levels of immunosuppression mean a higher risk of infection, with rates of infection typically highest in the early posttransplant period.

Common infections during this period include operative and perioperative nosocomial bacterial and fungal infections, reactivation of latent infections, and invasive fungal infections such as candidiasis, aspergillosis, and pneumocystosis. Donor-derived infections also must be considered. As time passes and the level of immunosuppression is reduced, liver recipients are less prone to infection.¹

The risk of infection can be minimized by appropriate antimicrobial prophylaxis, strategies for safe living after transplant,² vaccination,³ careful balancing of immunosuppressive therapy,⁴ and thoughtful donor selection.⁵ Drug-drug interactions are common and must be carefully considered to minimize the risk.

This review highlights common infectious complications encountered after liver transplant.

INTRA-ABDOMINAL INFECTIONS

Intra-abdominal infections are common in the early postoperative period.^6,7

Risk factors include:

• Pretransplant ascites
• Posttransplant dialysis
• Wound infection
• Reoperation⁸
• Hepatic artery thrombosis
• Roux-en-Y choledochojejunostomy anastomosis.⁹

Signs that may indicate intra-abdominal infection include fever, abdominal pain, leukocytosis, and elevated liver enzymes. But because of their immunosuppressed state, transplant recipients may not manifest fever as readily as the general population. They should be evaluated for cholangitis, peritonitis, biloma, and intra-abdominal abscess.

Organisms. Intra-abdominal infections are often polymicrobial. Enterococci, Staphylococcus aureus, gram-negative species including Pseudomonas, Klebsiella, and Acinetobacter, and Candida species are the most common pathogens. Strains are often resistant to multiple drugs, especially in patients who received antibiotics in the weeks before transplant.^8,10

Liver transplant recipients are also particularly susceptible to Clostridium difficile-associated colitis as a result of immunosuppression and frequent use of antibiotics perioperatively and postoperatively.¹¹ The spectrum of C difficile infection ranges from mild diarrhea to life-threatening colitis, and the course in liver transplant patients tends to be more complicated than in immunocompetent patients.¹²

Diagnosis. Intra-abdominal infections should be looked for and treated promptly, as they are associated with a higher mortality rate, a greater risk of graft loss, and a higher incidence of retransplant.^6,10 Abdominal ultrasonography or computed tomography (CT) can confirm the presence of fluid collections.

Treatment. Infected collections can be treated with percutaneous or surgical drainage and antimicrobial therapy. In the case of biliary tract complications, retransplant or surgical correction of biliary leakage or stenosis decreases the risk of death.⁶

Suspicion should be high for C difficile-associated colitis in cases of posttransplant diarrhea. C difficile toxin stool assays help confirm the diagnosis.¹² Oral metronidazole is recommended in mild to moderate C difficile infection, with oral vancomycin and intravenous metronidazole reserved for severe cases. Colectomy may be necessary in patients with toxic megacolon.

CYTOMEGALOVIRUS INFECTION

Cytomegalovirus is an important opportunistic pathogen in liver transplant recipients.¹³ It causes a range of manifestations, from infection (viremia with or without symptoms) to cytomegalovirus syndrome (fever, malaise, and cell-line cytopenias) to tissue-invasive disease with end-organ disease.¹⁴ Without preventive measures and treatment, cytomegalovirus disease can increase the risk of morbidity, allograft loss and death.^15,16

Risk factors for cytomegalovirus infection (Table 1) include:

• Discordant serostatus of the donor and recipient (the risk is highest in seronegative recipients of organs from seropositive donors)
• Higher levels of immunosuppression, especially when antilymphocyte antibodies are used
• Treatment of graft rejection
• Coinfection with other human herpesviruses, such as Epstein-Barr virus.^4,17

Preventing cytomegalovirus infection

The strategy to prevent cytomegalovirus infection depends on the serologic status of the donor and recipient and may include antiviral prophylaxis or preemptive treatment (Table 2).¹⁸

Prophylaxis involves giving antiviral drugs during the early high-risk period, with the goal of preventing the development of cytomegalovirus viremia. The alternative preemptive strategy emphasizes serial testing for cytomegalovirus viremia, with the goal of intervening with antiviral medications while viremia is at a low level, thus avoiding potential progression to cytomegalovirus disease. Both strategies have pros and cons that should be considered by each transplant center when setting institutional policy.

A prophylactic approach seems very effective at preventing both infection and disease from cytomegalovirus and has been shown to reduce graft rejection and the risk of death.¹⁸ It is preferred in cytomegalovirus-negative recipients when the donor was cytomegalovirus-positive—a high-risk situation.¹⁹ However, these patients are also at higher risk of late-onset cytomegalovirus disease. Higher cost and potential drug toxicity, mainly neutropenia from ganciclovir-based regimens, are additional considerations.

Preemptive treatment, in contrast, reserves drug treatment for patients who are actually infected with cytomegalovirus, thus resulting in fewer adverse drug events and lower cost; but it requires regular monitoring. Preemptive methods, by definition, cannot prevent infection, and with this strategy tissue-invasive disease not associated with viremia does occasionally occur.²⁰ As such, patients with a clinical presentation that suggests cytomegalovirus but have negative results on blood testing should be considered for tissue biopsy with culture and immunohistochemical stain.

The most commonly used regimens for antiviral prophylaxis and treatment in liver transplant recipients are intravenous ganciclovir and oral valganciclovir.²¹ Although valganciclovir is the most commonly used agent in this setting because of ease of administration, it has not been approved by the US Food and Drug Administration in liver transplant patients, as it was associated with higher rates of cytomegalovirus tissue-invasive disease.^22–24 Additionally, drug-resistant cytomegalovirus strains have been associated with valganciclovir prophylaxis in cytomegalovirus-negative recipients of solid organs from cytomegalovirus-positive donors.²⁵

Prophylaxis typically consists of therapy for 3 months from the time of transplant. In higher-risk patients (donor-positive, recipient-negative), longer courses of prophylaxis have been extrapolated from data in kidney transplant recipients.²⁶ Extension or reinstitution of prophylaxis should also be considered in liver transplant patients receiving treatment for rejection with antilymphocyte therapy.

Routine screening for cytomegalovirus is not recommended while patients are receiving prophylaxis. High-risk patients who are not receiving prophylaxis should be monitored with nucleic acid or pp65 antigenemia testing as part of the preemptive strategy protocol.

Treatment of cytomegalovirus disease

Although no specific threshold has been established, treatment is generally indicated if a patient has a consistent clinical syndrome, evidence of tissue injury, and persistent or increasing viremia.

Treatment involves giving antiviral drugs and also reducing the level of immunosuppression, if possible, until symptoms and viremia have resolved.

The choice of antiviral therapy depends on the severity of disease. Intravenous ganciclovir (5 mg/kg twice daily adjusted for renal impairment) or oral valganciclovir (900 mg twice daily, also renally dose-adjusted when necessary) can be used for mild to moderate disease if no significant gastrointestinal involvement is reported. Intravenous ganciclovir is preferred for patients with more severe disease or gastrointestinal involvement. The minimum duration of treatment is 2 weeks and may need to be prolonged until both symptoms and viremia completely resolve.¹⁸

Drug resistance can occur and should be considered in patients who have a history of prolonged ganciclovir or valganciclovir exposure who do not clinically improve or have persistent or rising viremia. In such cases, genotype assays are helpful, and initiation of alternative therapy should be considered. Mutations conferring resistance to ganciclovir are often associated with cross-resistance to cidofovir. Cidofovir can therefore be considered only when genotype assays demonstrate specific mutations conferring an isolated resistance to ganciclovir.²⁷ The addition of foscarnet to the ganciclovir regimen or substitution of foscarnet for ganciclovir are accepted approaches.

Although cytomegalovirus hyperimmunoglobulin has been used in prophylaxis and invasive disease treatment, its role in the management of ganciclovir-resistant cytomegalovirus infections remains controversial.²⁸

EPSTEIN-BARR VIRUS POSTTRANSPLANT LYMPHOPROLIFERATIVE DISEASE

Epstein-Barr virus-associated posttransplant lymphoproliferative disease is a spectrum of disorders ranging from an infectious mononucleosis syndrome to aggressive malignancy with the potential for death and significant morbidity after liver transplant.²⁹ The timeline of risk varies, but the disease is most common in the first year after transplant.

Risk factors for this disease (Table 1) are:

• Primary Epstein-Barr virus infection
• Cytomegalovirus donor-recipient mismatch
• Cytomegalovirus disease
• Higher levels of immunosuppression, especially with antilymphocyte antibodies.³⁰

The likelihood of Epstein-Barr virus playing a contributing role is lower in later-onset posttransplant lymphoproliferative disease. Patients who are older at the time of transplant, who receive highly immunogenic allografts including a liver as a component of a multivisceral transplant, and who receive increased immunosuppression to treat rejection are at even greater risk of late posttransplant lymphoproliferative disease.³¹ This is in contrast to early posttransplant lymphoproliferative disease, which is seen more commonly in children as a result of primary Epstein-Barr virus infection.

Recognition and diagnosis. Heightened suspicion is required when considering posttransplant lymphoproliferative disease, and careful evaluation of consistent symptoms and allograft dysfunction are required.

Clinically, posttransplant lymphoproliferative disease should be suspected if a liver transplant recipient develops unexplained fever, weight loss, lymphadenopathy, or cell-line cytopenias.^30,32 Other signs and symptoms may be related to the organ involved and may include evidence of hepatitis, pneumonitis, and gastrointestinal disease.³¹

Adjunctive diagnostic testing includes donor and recipient serology to characterize overall risk before transplantation and quantification of Epstein-Barr viral load, but confirmation relies on tissue histopathology.

Treatment focuses on reducing immunosuppression.^30,32 Adding antiviral agents does not seem to improve outcome in all cases.³³ Depending on clinical response and histologic classification, additional therapies such as anti-CD20 humanized chimeric monoclonal antibodies, surgery, radiation, and conventional chemotherapy may be required.³⁴

Preventive approaches remain controversial. Chemoprophylaxis with an antiviral such as ganciclovir is occasionally used but has not been shown to consistently decrease rates of posttransplant lymphoproliferative disease. These agents may act in an indirect manner, leading to decreased rates of cytomegalovirus infection, a major cofactor for posttransplant lymphoproliferative disease.²⁴

Although oral valganciclovir is used more than intravenous ganciclovir, it is not approved for liver transplant patients

Passive immunoprophylaxis with immunoglobulin targeting cytomegalovirus has shown to decrease rates of non-Hodgkin lymphoma from posttransplant lymphoproliferative disease in renal transplant recipients in the first year after transplant,³⁵ but data are lacking regarding its use in liver transplant recipients. Monitoring of the viral load and subsequent reduction of immunosuppression remain the most efficient measures to date.³⁶

FUNGAL INFECTIONS

Candida species account for more than half of fungal infections in liver transplant recipients.³⁷ However, a change has been noted in the past 20 years, with a decrease in Candida infections accompanied by an increase in Aspergillus infections.³⁸ Endemic mycoses such as coccidioidomycosis, blastomycosis, and histoplasmosis should be considered with the appropriate epidemiologic history or if disease develops early after transplant and the donor came from a highly endemic region.³⁹Cryptococcus may also be encountered.

Diagnosis. One of the most challenging aspects of fungal infection in liver transplant recipients is timely diagnosis. Heightened suspicion and early biopsy for pathological and microbiological confirmation are necessary. Although available noninvasive diagnostic tools often lack specificity, early detection of fungal markers may be of great use in guiding further diagnostic workup or empiric treatment in the critically ill.

Noninvasive tests include galactomannan, cryptococcal antigen, histoplasma antigen, (1-3)-beta-D-glucan assay and various antibody tests. Galactomannan testing has been widely used to aid in the diagnosis of invasive aspergillosis. Similarly, the (1-3)-beta-D-glucan assay is a non–culture-based tool for diagnosing and monitoring the treatment of invasive fungal infections. However, a definite diagnosis cannot be made on the basis of a positive test alone.⁴⁰ The complementary diagnostic characteristics of combining noninvasive assays have yet to be fully elucidated.⁴¹ Cultures and tissue histopathology are also used when possible.

Treatment is based on targeted specific antifungal drug therapy and reduction of immunosuppressive therapy, when possible. The choice of antifungal agent varies with the pathogen, the site of involvement, and the severity of the disease. A focus on potential drug interactions, their management, and therapeutic drug monitoring when using antifungal medications is essential in the posttransplant period. Combination therapy can be considered in some situations to enhance synergy. The following sections discuss in greater detail Candida species, Aspergillus species, and Pneumocystis jirovecii infections.

Candida infections

Candidiasis after liver transplant is typically nosocomial, especially when diagnosed during the first 3 months (Table 3).³⁷

Risk factors for invasive candidiasis include perioperative colonization, prolonged operative time, retransplant, greater transfusion requirements, and postoperative renal failure.^37,42,43 Invasive candidiasis is of concern for its effects on morbidity, mortality, and cost of care.^43–46

Organisms. The frequency of implicated species, in particular those with a natural resistance to fluconazole, differs in various reports.^37,45,46Candida albicans remains the most commonly isolated pathogen; however, non-albicans species including those resistant to fluconazole have been reported more frequently and include Candida glabrata and Candida krusei.^47,48

Signs and diagnosis. Invasive candidiasis in liver transplant recipients generally manifests itself in catheter-related blood stream infections, urinary tract infections, or intra-abdominal infections. Diagnosis can be made by isolating Candida from blood cultures, recovering organisms in culture of a normally sterile site, or finding direct microscopic evidence of the fungus on tissue specimens.⁴⁹

Disseminated candidiasis refers to the involvement of distant anatomic sites. Clinical manifestations may cause vision changes, abdominal pain or skin nodules with findings of candidemia, hepatosplenic abscesses, or retinal exudates on funduscopy.⁴⁹

Treatment of invasive candidiasis in liver recipients often involves antifungal therapy and reduction of immunosuppression. Broad-spectrum antifungals are initially advocated in an empirical approach to cover fluconazole-resistant strains of the non-albicans subgroups.⁵⁰ Depending on antifungal susceptibility, treatment can later be adjusted.

Fluconazole remains the agent of choice in most C albicans infections.⁴⁷ However, attention should be paid to the possibility of resistance in patients who have received fluconazole prophylaxis within the past 30 days. Additional agents used in treatment may include echinocandins, amphotericin, and additional azoles.

Antifungal prophylaxis is recommended in high-risk liver transplant patients, although its optimal duration remains undetermined.⁴⁴ Antifungal prophylaxis has been associated with decreased incidence of both superficial and invasive candidiasis.⁵¹

Aspergillus infection

Aspergillus, the second most common fungal pathogen, has become a more common concern in liver transplant recipients. Aspergillus fumigatus is the most frequently encountered species.^38,52

Risk factors. These infections typically occur in the first year, during intense immunosuppression. Retransplant, renal failure, and fulminant hepatic failure are major risk factors.⁵² In the presence of risk factors and a suggestive clinical setting, invasive aspergillosis should be considered and the diagnosis pursued.

Diagnosis is suggested by positive findings on CT accompanied by lower respiratory tract symptoms, focal lesions on neuroimaging, or demonstration of the fungus on cultures.⁴⁹ However, Aspergillus is rarely grown in blood culture. The galactomannan antigen is a noninvasive test that can provide supporting evidence for the diagnosis.^41,52 False-positive results do occur in the setting of certain antibiotics and cross-reacting fungi.⁵³

Treatment consists of antifungal therapy and immunosuppression reduction.⁵²

Candida accounts for more than half of fungal infections in liver transplant recipients, but Aspergillus is gaining

Voriconazole is the first-line agent for invasive aspergillosis. Monitoring for potential drug-drug interactions and side effects is required.^54,55 Amphotericin B is considered a second-line choice due to toxicity and lack of an oral formulation. In refractory cases, combined antifungal therapy could be considered.⁵² The duration of treatment is generally a minimum of 12 weeks.

Prophylaxis. Specific prophylaxis against invasive aspergillosis is not currently recommended; however, some authors suggest a prophylactic approach using echinocandins or liposomal amphotericin B in high-risk patients.^51,52 Aspergillosis is associated with a considerable increase in mortality in liver transplant recipients, which highlights the importance of timely management.^52,56

Pneumocystis jirovecii

P jirovecii remains a common opportunistic pathogen in people with impaired immunity, including transplant and human immunodeficiency virus patients.

Prophylaxis. Widespread adoption of antimicrobial prophylaxis by transplant centers has decreased the rates of P jirovecii infection in liver transplant recipients.^57,58 Commonly used prophylactic regimens after liver transplantation include a single-strength trimeth­oprim-sulfamethoxazole tablet daily or a double-strength tablet three times per week for a minimum of 6 to 12 months after transplant. Atovaquone and dapsone can be used as alternatives in cases of intolerance to tri­methoprim-sulfamethoxazole (Table 2).

Inhaled pentamidine is clearly inferior and should be used only when the other medications are contraindicated.⁵⁹

Signs and diagnosis. P jirovecii pneumonia is characterized by fever, cough, dyspnea, and chest pain. Insidious hypoxemia, abnormal chest examination, and bilateral interstitial pneumonia on chest radiography are common.

CT may be more sensitive than chest radiography.⁵⁷ Findings suggestive of P jirovecii pneumonia on chest CT are extensive bilateral and symmetrical ground-glass attenuations. Other less-characteristic findings include upper lobar parenchymal opacities and spontaneous pneumothorax.^57,60

The serum (1,3)-beta-D-glucan assay derived from major cell-wall components of P jiro­vecii might be helpful. Studies report a sensitivity for P jirovecii pneumonia as high as 96% and a negative predictive value of 99.8%.^61,62

Definitive diagnosis requires identification of the pathogen. Routine expectorated sputum sampling is generally associated with a poor diagnostic yield. Bronchoscopy and bronchoalveolar lavage with silver or fluorescent antibody staining of samples, polymerase chain reaction testing, or both significantly improves diagnosis. Transbronchial or open lung biopsy are often unnecessary.⁵⁷

Treatment. Trimethoprim-sulfamethoxazole is the first-line agent for treating P jirovecii pneumonia.⁵⁷ The minimum duration of treatment is 14 days, with extended courses for severe infection.

Intravenous pentamidine or clindamycin plus primaquine are alternatives for patients who cannot tolerate trimethoprim-sulfamethoxazole. The major concern with intravenous pentamidine is renal dysfunction. Hypoglycemia or hyperglycemia, neutropenia, thrombocytopenia, nausea, dysgeusia, and pancreatitis may also occur.⁶³

Atovaquone might also be beneficial in mild to moderate P jirovecii pneumonia. The main side effects include skin rashes, gastrointestinal intolerance, and elevation of transaminases.⁶⁴

A corticosteroid (40–60 mg of prednisone or its equivalent) may be beneficial in conjunction with antimicrobial therapy in patients with significant hypoxia (partial pressure of arterial oxygen < 70 mm Hg on room air) in decreasing the risk of respiratory failure and need for intubation.

With appropriate and timely antimicrobial prophylaxis, cases of P jirovecii pneumonia should continue to decrease.

TUBERCULOSIS

Development of tuberculosis after transplantation is a catastrophic complication, with mortality rates of up to 30%.⁶⁵ Most cases of posttransplant tuberculosis represent reactivation of latent disease.⁶⁶ Screening with tuberculin skin tests or interferon-gamma-release assays is recommended in all liver transplant candidates. Chest radiography before transplant is necessary when assessing a positive screening test.⁶⁷

The optimal management of latent tuberculosis in these cases remains controversial. Patients at high risk or those with positive screening results on chest radiography warrant treatment for latent tuberculosis infection with isoniazid unless contraindicated.^67,68

The ideal time to initiate prophylactic isoniazid therapy is unclear. Some authors suggest delaying it, as it might be associated with poor tolerance and hepatotoxicity.⁶⁹ Others have found that early isoniazid use was not associated with negative outcomes.⁷⁰

Risk factors for symptomatic tuberculosis after liver transplant include previous infection with tuberculosis, intensified immunosuppression (especially anti-T-lymphocyte therapies), diabetes mellitus, and other co-infections (Table 1).⁷¹

The increased incidence of atypical presentations in recent years makes the diagnosis of active tuberculosis among liver transplant recipients challenging. Sputum smears can be negative due to low mycobacterial burdens, and tuberculin skin testing and interferon-gamma-release assays may be falsely negative due to immunosuppression.⁶⁷

Treatment of active tuberculosis consists initially of a four-drug regimen using isoniazid, rifampin, pyrazinamide, and ethambutol for 2 months. Adjustments are made in accordance with culture and sensitivity results. Treatment can then be tapered to two drugs (isoniazid and rifampin) for a minimum of 4 additional months. Prolonged treatment may be required in instances of extrapulmonary or disseminated disease.^65,72

Tuberculosis treatment can be complicated by hepatotoxicity in liver transplant recipients because of direct drug effects and drug-drug interactions with immunosuppressive agents. Close monitoring for rejection and hepatotoxicity is therefore imperative while liver transplant recipients are receiving antituberculosis therapy. Drug-drug interactions may also be responsible for marked reductions in immunosuppression levels, especially with regimens containing rifampin.⁷¹ Substitution of rifabutin for rifampin reduces the effect of drug interactions.⁶⁶

VIRAL HEPATITIS

Hepatitis B virus

Hepatitis B virus-related end-stage liver disease and hepatocellular carcinoma are common indications for liver transplant in Asia. It is less common in the United States and Europe, accounting for less than 10% of all liver transplant cases. Prognosis is favorable in recipients undergoing liver transplant for hepatitis B virus, with excellent survival rates. Prevention of reinfection is crucial in these patients.

Treatment with combination antiviral agents and hepatitis B immunoglobulin (HBIG) is effective.⁷³ Lamivudine was the first nucleoside analogue found to be effective against hepatitis B virus. Its low cost and relative safety are strong arguments in favor of its continued use in liver transplant recipients.⁷⁴ In patients without evidence of hepatitis B viral replication at the time of transplant, monotherapy with lamivudine has led to low recurrence rates, and adefovir can be added to control resistant viral strains.⁷⁵

Widespread adoption of prophylaxis has decreased the rate of P jirovecii infection in liver transplant recipients

The frequent emergence of resistance with lamivudine favors newer agents such as entecavir or tenofovir. These nucleoside and nucleotide analogues have a higher barrier to resistance, and thus resistance to them is rare. They are also more efficient, potentially allowing use of an HBIG-sparing protocol.⁷⁶ However, they are associated with a higher risk of nephrotoxicity and require dose adjustments in renal insufficiency. Data directly comparing entecavir and tenofovir are scarce.

Prophylaxis. Most studies support an individualized approach for prevention of hepatitis B virus reinfection. High-risk patients, ie, those positive for HBe antigen or with high viral loads (> 100,000 copies/mL) are generally treated with both HBIG and antiviral agents.⁷⁷ Low-risk patients are those with a negative HBe antigen, low hepatitis B virus DNA levels, hepatitis B virus-related acute liver failure, and cirrhosis resulting from coinfection with both hepatitis B and hepatitis D virus.⁷⁵ In low-risk patients, discontinuation of HBIG after 1 to 2 years of treatment is appropriate, and long-term prophylaxis with antiviral agents alone is an option. However, levels of hepatitis B DNA should be monitored closely.^78,79

Hepatitis C virus

Recurrence of hepatitis C virus infection is the rule among patients who are viremic at the time of liver transplant.^80,81 Most of these patients will show histologic evidence of recurrent hepatitis within the first year after liver transplant. It is often difficult to distinguish between the histopathological appearance of a recurrent hepatitis C virus infection and acute cellular rejection.

Progression to fibrosis and subsequently cirrhosis and decompensation is highly variable in hepatitis C virus-infected liver transplant recipients. Diabetes, insulin resistance, and possibly hepatitis steatosis have been associated with a rapid progression to advanced fibrosis. The contribution of immunosuppression to the progression of hepatitis C virus remains an area of active study. Some studies point to antilymphocyte immunosuppressive agents as a potential cause.⁸² Liver biopsy is a useful tool in this situation. It allows monitoring of disease severity and progression and may distinguish recurrent hepatitis C virus disease from other causes of liver enzyme elevation.

The major concern with the recurrence of hepatitis C virus infection after liver transplant is allograft loss. Rates of patient and graft survival are reduced in infected patients compared with hepatitis C virus-negative patients.^83,84 Prophylactic antiviral therapy has no current role in the management of hepatitis C virus disease. Those manifesting moderate to severe necroinflammation or mild to moderate fibrosis indicative of progressive disease should be treated.^81,85

Sustained viral clearance with antiviral agents confers a graft survival benefit.

The combination of peg-interferon and weight-based ribavirin has been the standard of treatment but may be associated with increased rates of rejection.^86,87 The sustained virologic response rates for hepatitis C virus range from 60% in genotypes 4, 5, and 6 after 48 weeks of treatment to 60% to 80% in genotypes 2 and 3 after 24 weeks, but only about 30% in genotype 1.⁸⁸

The major concern with hepatitis C recurrence after liver transplant is allograft loss

Treatment with the newer agents, especially protease inhibitors, in genotype 1 (peg-interferon, ribavirin, and either telaprevir or boceprevir) has been evaluated. Success rates reaching 70% have been achieved.⁸⁹ Adverse effects can be a major setback. Serious complications include severe anemia, renal dysfunction, increased risk of infection, and death.

Triple therapy should be carefully considered in liver transplant patients with genotype 1 hepatitis C virus.⁹⁰ Significant drug-drug interactions are reported between hepatitis C virus protease inhibitors and immunosuppression regimens. Additional new oral direct- acting antivirals have been investigated. They bring promising advances in hepatitis C virus treatment and pave the way for interferon-free regimens with pangenotypic activity.

IMMUNIZATION

Immunization can decrease the risk of infectious complications in liver transplant recipients, as well as in close contacts and healthcare professionals.³

Influenza. Pretransplant influenza vaccine and posttransplant annual influenza vaccines are necessary.

Pneumococcal immunization should additionally be provided prior to transplant and repeated every 3 to 5 years thereafter.^3,91

A number of other vaccinations should also be completed before transplant, including the hepatitis A and B vaccines and the tetanus/diphtheria/acellular pertussis vaccines. However, these vaccinations have not been shown to be detrimental to patients after transplant.⁹¹

Varicella and zoster vaccines should be given before liver transplant—zoster in patients over age 60, and varicella in patients with no immunity. Live vaccines, including varicella and zoster vaccines, are contraindicated after liver transplant.³

Human papillomavirus. The bivalent human papillomavirus vaccine can be given before transplant in females ages 9 to 26; the quadrivalent vaccine is beneficial in those ages 9 to 26 and in women under age 45.^3,91

IMMUNOSUPPRESSION CARRIES RISK OF INFECTION

Most liver transplant patients require prolonged immunosuppressive therapy. This comes with an increased risk of new or recurrent infections, potentially causing death and significant morbidity.

Evaluation of existing risk factors, appropriate prophylaxis and immunization, timely diagnosis, and treatment of such infections are therefore essential steps for the successful management of liver transplant recipients.

The immunosuppressed state of liver transplant recipients makes them vulnerable to infections after surgery.¹ These infections are directly correlated with the net state of immunosuppression. Higher levels of immunosuppression mean a higher risk of infection, with rates of infection typically highest in the early posttransplant period.

Common infections during this period include operative and perioperative nosocomial bacterial and fungal infections, reactivation of latent infections, and invasive fungal infections such as candidiasis, aspergillosis, and pneumocystosis. Donor-derived infections also must be considered. As time passes and the level of immunosuppression is reduced, liver recipients are less prone to infection.¹

The risk of infection can be minimized by appropriate antimicrobial prophylaxis, strategies for safe living after transplant,² vaccination,³ careful balancing of immunosuppressive therapy,⁴ and thoughtful donor selection.⁵ Drug-drug interactions are common and must be carefully considered to minimize the risk.

This review highlights common infectious complications encountered after liver transplant.

INTRA-ABDOMINAL INFECTIONS

Intra-abdominal infections are common in the early postoperative period.^6,7

Risk factors include:

• Pretransplant ascites
• Posttransplant dialysis
• Wound infection
• Reoperation⁸
• Hepatic artery thrombosis
• Roux-en-Y choledochojejunostomy anastomosis.⁹

Signs that may indicate intra-abdominal infection include fever, abdominal pain, leukocytosis, and elevated liver enzymes. But because of their immunosuppressed state, transplant recipients may not manifest fever as readily as the general population. They should be evaluated for cholangitis, peritonitis, biloma, and intra-abdominal abscess.

Organisms. Intra-abdominal infections are often polymicrobial. Enterococci, Staphylococcus aureus, gram-negative species including Pseudomonas, Klebsiella, and Acinetobacter, and Candida species are the most common pathogens. Strains are often resistant to multiple drugs, especially in patients who received antibiotics in the weeks before transplant.^8,10

Liver transplant recipients are also particularly susceptible to Clostridium difficile-associated colitis as a result of immunosuppression and frequent use of antibiotics perioperatively and postoperatively.¹¹ The spectrum of C difficile infection ranges from mild diarrhea to life-threatening colitis, and the course in liver transplant patients tends to be more complicated than in immunocompetent patients.¹²

Diagnosis. Intra-abdominal infections should be looked for and treated promptly, as they are associated with a higher mortality rate, a greater risk of graft loss, and a higher incidence of retransplant.^6,10 Abdominal ultrasonography or computed tomography (CT) can confirm the presence of fluid collections.

Treatment. Infected collections can be treated with percutaneous or surgical drainage and antimicrobial therapy. In the case of biliary tract complications, retransplant or surgical correction of biliary leakage or stenosis decreases the risk of death.⁶

Suspicion should be high for C difficile-associated colitis in cases of posttransplant diarrhea. C difficile toxin stool assays help confirm the diagnosis.¹² Oral metronidazole is recommended in mild to moderate C difficile infection, with oral vancomycin and intravenous metronidazole reserved for severe cases. Colectomy may be necessary in patients with toxic megacolon.

CYTOMEGALOVIRUS INFECTION

Cytomegalovirus is an important opportunistic pathogen in liver transplant recipients.¹³ It causes a range of manifestations, from infection (viremia with or without symptoms) to cytomegalovirus syndrome (fever, malaise, and cell-line cytopenias) to tissue-invasive disease with end-organ disease.¹⁴ Without preventive measures and treatment, cytomegalovirus disease can increase the risk of morbidity, allograft loss and death.^15,16

Risk factors for cytomegalovirus infection (Table 1) include:

• Discordant serostatus of the donor and recipient (the risk is highest in seronegative recipients of organs from seropositive donors)
• Higher levels of immunosuppression, especially when antilymphocyte antibodies are used
• Treatment of graft rejection
• Coinfection with other human herpesviruses, such as Epstein-Barr virus.^4,17

Preventing cytomegalovirus infection

The strategy to prevent cytomegalovirus infection depends on the serologic status of the donor and recipient and may include antiviral prophylaxis or preemptive treatment (Table 2).¹⁸

Prophylaxis involves giving antiviral drugs during the early high-risk period, with the goal of preventing the development of cytomegalovirus viremia. The alternative preemptive strategy emphasizes serial testing for cytomegalovirus viremia, with the goal of intervening with antiviral medications while viremia is at a low level, thus avoiding potential progression to cytomegalovirus disease. Both strategies have pros and cons that should be considered by each transplant center when setting institutional policy.

A prophylactic approach seems very effective at preventing both infection and disease from cytomegalovirus and has been shown to reduce graft rejection and the risk of death.¹⁸ It is preferred in cytomegalovirus-negative recipients when the donor was cytomegalovirus-positive—a high-risk situation.¹⁹ However, these patients are also at higher risk of late-onset cytomegalovirus disease. Higher cost and potential drug toxicity, mainly neutropenia from ganciclovir-based regimens, are additional considerations.

Preemptive treatment, in contrast, reserves drug treatment for patients who are actually infected with cytomegalovirus, thus resulting in fewer adverse drug events and lower cost; but it requires regular monitoring. Preemptive methods, by definition, cannot prevent infection, and with this strategy tissue-invasive disease not associated with viremia does occasionally occur.²⁰ As such, patients with a clinical presentation that suggests cytomegalovirus but have negative results on blood testing should be considered for tissue biopsy with culture and immunohistochemical stain.

The most commonly used regimens for antiviral prophylaxis and treatment in liver transplant recipients are intravenous ganciclovir and oral valganciclovir.²¹ Although valganciclovir is the most commonly used agent in this setting because of ease of administration, it has not been approved by the US Food and Drug Administration in liver transplant patients, as it was associated with higher rates of cytomegalovirus tissue-invasive disease.^22–24 Additionally, drug-resistant cytomegalovirus strains have been associated with valganciclovir prophylaxis in cytomegalovirus-negative recipients of solid organs from cytomegalovirus-positive donors.²⁵

Prophylaxis typically consists of therapy for 3 months from the time of transplant. In higher-risk patients (donor-positive, recipient-negative), longer courses of prophylaxis have been extrapolated from data in kidney transplant recipients.²⁶ Extension or reinstitution of prophylaxis should also be considered in liver transplant patients receiving treatment for rejection with antilymphocyte therapy.

Routine screening for cytomegalovirus is not recommended while patients are receiving prophylaxis. High-risk patients who are not receiving prophylaxis should be monitored with nucleic acid or pp65 antigenemia testing as part of the preemptive strategy protocol.

Treatment of cytomegalovirus disease

Although no specific threshold has been established, treatment is generally indicated if a patient has a consistent clinical syndrome, evidence of tissue injury, and persistent or increasing viremia.

Treatment involves giving antiviral drugs and also reducing the level of immunosuppression, if possible, until symptoms and viremia have resolved.

The choice of antiviral therapy depends on the severity of disease. Intravenous ganciclovir (5 mg/kg twice daily adjusted for renal impairment) or oral valganciclovir (900 mg twice daily, also renally dose-adjusted when necessary) can be used for mild to moderate disease if no significant gastrointestinal involvement is reported. Intravenous ganciclovir is preferred for patients with more severe disease or gastrointestinal involvement. The minimum duration of treatment is 2 weeks and may need to be prolonged until both symptoms and viremia completely resolve.¹⁸

Drug resistance can occur and should be considered in patients who have a history of prolonged ganciclovir or valganciclovir exposure who do not clinically improve or have persistent or rising viremia. In such cases, genotype assays are helpful, and initiation of alternative therapy should be considered. Mutations conferring resistance to ganciclovir are often associated with cross-resistance to cidofovir. Cidofovir can therefore be considered only when genotype assays demonstrate specific mutations conferring an isolated resistance to ganciclovir.²⁷ The addition of foscarnet to the ganciclovir regimen or substitution of foscarnet for ganciclovir are accepted approaches.

Although cytomegalovirus hyperimmunoglobulin has been used in prophylaxis and invasive disease treatment, its role in the management of ganciclovir-resistant cytomegalovirus infections remains controversial.²⁸

EPSTEIN-BARR VIRUS POSTTRANSPLANT LYMPHOPROLIFERATIVE DISEASE

Epstein-Barr virus-associated posttransplant lymphoproliferative disease is a spectrum of disorders ranging from an infectious mononucleosis syndrome to aggressive malignancy with the potential for death and significant morbidity after liver transplant.²⁹ The timeline of risk varies, but the disease is most common in the first year after transplant.

Risk factors for this disease (Table 1) are:

• Primary Epstein-Barr virus infection
• Cytomegalovirus donor-recipient mismatch
• Cytomegalovirus disease
• Higher levels of immunosuppression, especially with antilymphocyte antibodies.³⁰

The likelihood of Epstein-Barr virus playing a contributing role is lower in later-onset posttransplant lymphoproliferative disease. Patients who are older at the time of transplant, who receive highly immunogenic allografts including a liver as a component of a multivisceral transplant, and who receive increased immunosuppression to treat rejection are at even greater risk of late posttransplant lymphoproliferative disease.³¹ This is in contrast to early posttransplant lymphoproliferative disease, which is seen more commonly in children as a result of primary Epstein-Barr virus infection.

Recognition and diagnosis. Heightened suspicion is required when considering posttransplant lymphoproliferative disease, and careful evaluation of consistent symptoms and allograft dysfunction are required.

Clinically, posttransplant lymphoproliferative disease should be suspected if a liver transplant recipient develops unexplained fever, weight loss, lymphadenopathy, or cell-line cytopenias.^30,32 Other signs and symptoms may be related to the organ involved and may include evidence of hepatitis, pneumonitis, and gastrointestinal disease.³¹

Adjunctive diagnostic testing includes donor and recipient serology to characterize overall risk before transplantation and quantification of Epstein-Barr viral load, but confirmation relies on tissue histopathology.

Treatment focuses on reducing immunosuppression.^30,32 Adding antiviral agents does not seem to improve outcome in all cases.³³ Depending on clinical response and histologic classification, additional therapies such as anti-CD20 humanized chimeric monoclonal antibodies, surgery, radiation, and conventional chemotherapy may be required.³⁴

Preventive approaches remain controversial. Chemoprophylaxis with an antiviral such as ganciclovir is occasionally used but has not been shown to consistently decrease rates of posttransplant lymphoproliferative disease. These agents may act in an indirect manner, leading to decreased rates of cytomegalovirus infection, a major cofactor for posttransplant lymphoproliferative disease.²⁴

Although oral valganciclovir is used more than intravenous ganciclovir, it is not approved for liver transplant patients

Passive immunoprophylaxis with immunoglobulin targeting cytomegalovirus has shown to decrease rates of non-Hodgkin lymphoma from posttransplant lymphoproliferative disease in renal transplant recipients in the first year after transplant,³⁵ but data are lacking regarding its use in liver transplant recipients. Monitoring of the viral load and subsequent reduction of immunosuppression remain the most efficient measures to date.³⁶

FUNGAL INFECTIONS

Candida species account for more than half of fungal infections in liver transplant recipients.³⁷ However, a change has been noted in the past 20 years, with a decrease in Candida infections accompanied by an increase in Aspergillus infections.³⁸ Endemic mycoses such as coccidioidomycosis, blastomycosis, and histoplasmosis should be considered with the appropriate epidemiologic history or if disease develops early after transplant and the donor came from a highly endemic region.³⁹Cryptococcus may also be encountered.

Diagnosis. One of the most challenging aspects of fungal infection in liver transplant recipients is timely diagnosis. Heightened suspicion and early biopsy for pathological and microbiological confirmation are necessary. Although available noninvasive diagnostic tools often lack specificity, early detection of fungal markers may be of great use in guiding further diagnostic workup or empiric treatment in the critically ill.

Noninvasive tests include galactomannan, cryptococcal antigen, histoplasma antigen, (1-3)-beta-D-glucan assay and various antibody tests. Galactomannan testing has been widely used to aid in the diagnosis of invasive aspergillosis. Similarly, the (1-3)-beta-D-glucan assay is a non–culture-based tool for diagnosing and monitoring the treatment of invasive fungal infections. However, a definite diagnosis cannot be made on the basis of a positive test alone.⁴⁰ The complementary diagnostic characteristics of combining noninvasive assays have yet to be fully elucidated.⁴¹ Cultures and tissue histopathology are also used when possible.

Treatment is based on targeted specific antifungal drug therapy and reduction of immunosuppressive therapy, when possible. The choice of antifungal agent varies with the pathogen, the site of involvement, and the severity of the disease. A focus on potential drug interactions, their management, and therapeutic drug monitoring when using antifungal medications is essential in the posttransplant period. Combination therapy can be considered in some situations to enhance synergy. The following sections discuss in greater detail Candida species, Aspergillus species, and Pneumocystis jirovecii infections.

Candida infections

Candidiasis after liver transplant is typically nosocomial, especially when diagnosed during the first 3 months (Table 3).³⁷

Risk factors for invasive candidiasis include perioperative colonization, prolonged operative time, retransplant, greater transfusion requirements, and postoperative renal failure.^37,42,43 Invasive candidiasis is of concern for its effects on morbidity, mortality, and cost of care.^43–46

Organisms. The frequency of implicated species, in particular those with a natural resistance to fluconazole, differs in various reports.^37,45,46Candida albicans remains the most commonly isolated pathogen; however, non-albicans species including those resistant to fluconazole have been reported more frequently and include Candida glabrata and Candida krusei.^47,48

Signs and diagnosis. Invasive candidiasis in liver transplant recipients generally manifests itself in catheter-related blood stream infections, urinary tract infections, or intra-abdominal infections. Diagnosis can be made by isolating Candida from blood cultures, recovering organisms in culture of a normally sterile site, or finding direct microscopic evidence of the fungus on tissue specimens.⁴⁹

Disseminated candidiasis refers to the involvement of distant anatomic sites. Clinical manifestations may cause vision changes, abdominal pain or skin nodules with findings of candidemia, hepatosplenic abscesses, or retinal exudates on funduscopy.⁴⁹

Treatment of invasive candidiasis in liver recipients often involves antifungal therapy and reduction of immunosuppression. Broad-spectrum antifungals are initially advocated in an empirical approach to cover fluconazole-resistant strains of the non-albicans subgroups.⁵⁰ Depending on antifungal susceptibility, treatment can later be adjusted.

Fluconazole remains the agent of choice in most C albicans infections.⁴⁷ However, attention should be paid to the possibility of resistance in patients who have received fluconazole prophylaxis within the past 30 days. Additional agents used in treatment may include echinocandins, amphotericin, and additional azoles.

Antifungal prophylaxis is recommended in high-risk liver transplant patients, although its optimal duration remains undetermined.⁴⁴ Antifungal prophylaxis has been associated with decreased incidence of both superficial and invasive candidiasis.⁵¹

Aspergillus infection

Aspergillus, the second most common fungal pathogen, has become a more common concern in liver transplant recipients. Aspergillus fumigatus is the most frequently encountered species.^38,52

Risk factors. These infections typically occur in the first year, during intense immunosuppression. Retransplant, renal failure, and fulminant hepatic failure are major risk factors.⁵² In the presence of risk factors and a suggestive clinical setting, invasive aspergillosis should be considered and the diagnosis pursued.

Diagnosis is suggested by positive findings on CT accompanied by lower respiratory tract symptoms, focal lesions on neuroimaging, or demonstration of the fungus on cultures.⁴⁹ However, Aspergillus is rarely grown in blood culture. The galactomannan antigen is a noninvasive test that can provide supporting evidence for the diagnosis.^41,52 False-positive results do occur in the setting of certain antibiotics and cross-reacting fungi.⁵³

Treatment consists of antifungal therapy and immunosuppression reduction.⁵²

Candida accounts for more than half of fungal infections in liver transplant recipients, but Aspergillus is gaining

Voriconazole is the first-line agent for invasive aspergillosis. Monitoring for potential drug-drug interactions and side effects is required.^54,55 Amphotericin B is considered a second-line choice due to toxicity and lack of an oral formulation. In refractory cases, combined antifungal therapy could be considered.⁵² The duration of treatment is generally a minimum of 12 weeks.

Prophylaxis. Specific prophylaxis against invasive aspergillosis is not currently recommended; however, some authors suggest a prophylactic approach using echinocandins or liposomal amphotericin B in high-risk patients.^51,52 Aspergillosis is associated with a considerable increase in mortality in liver transplant recipients, which highlights the importance of timely management.^52,56

Pneumocystis jirovecii

P jirovecii remains a common opportunistic pathogen in people with impaired immunity, including transplant and human immunodeficiency virus patients.

Prophylaxis. Widespread adoption of antimicrobial prophylaxis by transplant centers has decreased the rates of P jirovecii infection in liver transplant recipients.^57,58 Commonly used prophylactic regimens after liver transplantation include a single-strength trimeth­oprim-sulfamethoxazole tablet daily or a double-strength tablet three times per week for a minimum of 6 to 12 months after transplant. Atovaquone and dapsone can be used as alternatives in cases of intolerance to tri­methoprim-sulfamethoxazole (Table 2).

Inhaled pentamidine is clearly inferior and should be used only when the other medications are contraindicated.⁵⁹

Signs and diagnosis. P jirovecii pneumonia is characterized by fever, cough, dyspnea, and chest pain. Insidious hypoxemia, abnormal chest examination, and bilateral interstitial pneumonia on chest radiography are common.

CT may be more sensitive than chest radiography.⁵⁷ Findings suggestive of P jirovecii pneumonia on chest CT are extensive bilateral and symmetrical ground-glass attenuations. Other less-characteristic findings include upper lobar parenchymal opacities and spontaneous pneumothorax.^57,60

The serum (1,3)-beta-D-glucan assay derived from major cell-wall components of P jiro­vecii might be helpful. Studies report a sensitivity for P jirovecii pneumonia as high as 96% and a negative predictive value of 99.8%.^61,62

Definitive diagnosis requires identification of the pathogen. Routine expectorated sputum sampling is generally associated with a poor diagnostic yield. Bronchoscopy and bronchoalveolar lavage with silver or fluorescent antibody staining of samples, polymerase chain reaction testing, or both significantly improves diagnosis. Transbronchial or open lung biopsy are often unnecessary.⁵⁷

Treatment. Trimethoprim-sulfamethoxazole is the first-line agent for treating P jirovecii pneumonia.⁵⁷ The minimum duration of treatment is 14 days, with extended courses for severe infection.

Intravenous pentamidine or clindamycin plus primaquine are alternatives for patients who cannot tolerate trimethoprim-sulfamethoxazole. The major concern with intravenous pentamidine is renal dysfunction. Hypoglycemia or hyperglycemia, neutropenia, thrombocytopenia, nausea, dysgeusia, and pancreatitis may also occur.⁶³

Atovaquone might also be beneficial in mild to moderate P jirovecii pneumonia. The main side effects include skin rashes, gastrointestinal intolerance, and elevation of transaminases.⁶⁴

A corticosteroid (40–60 mg of prednisone or its equivalent) may be beneficial in conjunction with antimicrobial therapy in patients with significant hypoxia (partial pressure of arterial oxygen < 70 mm Hg on room air) in decreasing the risk of respiratory failure and need for intubation.

With appropriate and timely antimicrobial prophylaxis, cases of P jirovecii pneumonia should continue to decrease.

TUBERCULOSIS

Development of tuberculosis after transplantation is a catastrophic complication, with mortality rates of up to 30%.⁶⁵ Most cases of posttransplant tuberculosis represent reactivation of latent disease.⁶⁶ Screening with tuberculin skin tests or interferon-gamma-release assays is recommended in all liver transplant candidates. Chest radiography before transplant is necessary when assessing a positive screening test.⁶⁷

The optimal management of latent tuberculosis in these cases remains controversial. Patients at high risk or those with positive screening results on chest radiography warrant treatment for latent tuberculosis infection with isoniazid unless contraindicated.^67,68

The ideal time to initiate prophylactic isoniazid therapy is unclear. Some authors suggest delaying it, as it might be associated with poor tolerance and hepatotoxicity.⁶⁹ Others have found that early isoniazid use was not associated with negative outcomes.⁷⁰

Risk factors for symptomatic tuberculosis after liver transplant include previous infection with tuberculosis, intensified immunosuppression (especially anti-T-lymphocyte therapies), diabetes mellitus, and other co-infections (Table 1).⁷¹

The increased incidence of atypical presentations in recent years makes the diagnosis of active tuberculosis among liver transplant recipients challenging. Sputum smears can be negative due to low mycobacterial burdens, and tuberculin skin testing and interferon-gamma-release assays may be falsely negative due to immunosuppression.⁶⁷

Treatment of active tuberculosis consists initially of a four-drug regimen using isoniazid, rifampin, pyrazinamide, and ethambutol for 2 months. Adjustments are made in accordance with culture and sensitivity results. Treatment can then be tapered to two drugs (isoniazid and rifampin) for a minimum of 4 additional months. Prolonged treatment may be required in instances of extrapulmonary or disseminated disease.^65,72

Tuberculosis treatment can be complicated by hepatotoxicity in liver transplant recipients because of direct drug effects and drug-drug interactions with immunosuppressive agents. Close monitoring for rejection and hepatotoxicity is therefore imperative while liver transplant recipients are receiving antituberculosis therapy. Drug-drug interactions may also be responsible for marked reductions in immunosuppression levels, especially with regimens containing rifampin.⁷¹ Substitution of rifabutin for rifampin reduces the effect of drug interactions.⁶⁶

VIRAL HEPATITIS

Hepatitis B virus

Hepatitis B virus-related end-stage liver disease and hepatocellular carcinoma are common indications for liver transplant in Asia. It is less common in the United States and Europe, accounting for less than 10% of all liver transplant cases. Prognosis is favorable in recipients undergoing liver transplant for hepatitis B virus, with excellent survival rates. Prevention of reinfection is crucial in these patients.

Treatment with combination antiviral agents and hepatitis B immunoglobulin (HBIG) is effective.⁷³ Lamivudine was the first nucleoside analogue found to be effective against hepatitis B virus. Its low cost and relative safety are strong arguments in favor of its continued use in liver transplant recipients.⁷⁴ In patients without evidence of hepatitis B viral replication at the time of transplant, monotherapy with lamivudine has led to low recurrence rates, and adefovir can be added to control resistant viral strains.⁷⁵

Widespread adoption of prophylaxis has decreased the rate of P jirovecii infection in liver transplant recipients

The frequent emergence of resistance with lamivudine favors newer agents such as entecavir or tenofovir. These nucleoside and nucleotide analogues have a higher barrier to resistance, and thus resistance to them is rare. They are also more efficient, potentially allowing use of an HBIG-sparing protocol.⁷⁶ However, they are associated with a higher risk of nephrotoxicity and require dose adjustments in renal insufficiency. Data directly comparing entecavir and tenofovir are scarce.

Prophylaxis. Most studies support an individualized approach for prevention of hepatitis B virus reinfection. High-risk patients, ie, those positive for HBe antigen or with high viral loads (> 100,000 copies/mL) are generally treated with both HBIG and antiviral agents.⁷⁷ Low-risk patients are those with a negative HBe antigen, low hepatitis B virus DNA levels, hepatitis B virus-related acute liver failure, and cirrhosis resulting from coinfection with both hepatitis B and hepatitis D virus.⁷⁵ In low-risk patients, discontinuation of HBIG after 1 to 2 years of treatment is appropriate, and long-term prophylaxis with antiviral agents alone is an option. However, levels of hepatitis B DNA should be monitored closely.^78,79

Hepatitis C virus

Recurrence of hepatitis C virus infection is the rule among patients who are viremic at the time of liver transplant.^80,81 Most of these patients will show histologic evidence of recurrent hepatitis within the first year after liver transplant. It is often difficult to distinguish between the histopathological appearance of a recurrent hepatitis C virus infection and acute cellular rejection.

Progression to fibrosis and subsequently cirrhosis and decompensation is highly variable in hepatitis C virus-infected liver transplant recipients. Diabetes, insulin resistance, and possibly hepatitis steatosis have been associated with a rapid progression to advanced fibrosis. The contribution of immunosuppression to the progression of hepatitis C virus remains an area of active study. Some studies point to antilymphocyte immunosuppressive agents as a potential cause.⁸² Liver biopsy is a useful tool in this situation. It allows monitoring of disease severity and progression and may distinguish recurrent hepatitis C virus disease from other causes of liver enzyme elevation.

The major concern with the recurrence of hepatitis C virus infection after liver transplant is allograft loss. Rates of patient and graft survival are reduced in infected patients compared with hepatitis C virus-negative patients.^83,84 Prophylactic antiviral therapy has no current role in the management of hepatitis C virus disease. Those manifesting moderate to severe necroinflammation or mild to moderate fibrosis indicative of progressive disease should be treated.^81,85

Sustained viral clearance with antiviral agents confers a graft survival benefit.

The combination of peg-interferon and weight-based ribavirin has been the standard of treatment but may be associated with increased rates of rejection.^86,87 The sustained virologic response rates for hepatitis C virus range from 60% in genotypes 4, 5, and 6 after 48 weeks of treatment to 60% to 80% in genotypes 2 and 3 after 24 weeks, but only about 30% in genotype 1.⁸⁸

The major concern with hepatitis C recurrence after liver transplant is allograft loss

Treatment with the newer agents, especially protease inhibitors, in genotype 1 (peg-interferon, ribavirin, and either telaprevir or boceprevir) has been evaluated. Success rates reaching 70% have been achieved.⁸⁹ Adverse effects can be a major setback. Serious complications include severe anemia, renal dysfunction, increased risk of infection, and death.

Triple therapy should be carefully considered in liver transplant patients with genotype 1 hepatitis C virus.⁹⁰ Significant drug-drug interactions are reported between hepatitis C virus protease inhibitors and immunosuppression regimens. Additional new oral direct- acting antivirals have been investigated. They bring promising advances in hepatitis C virus treatment and pave the way for interferon-free regimens with pangenotypic activity.

IMMUNIZATION

Immunization can decrease the risk of infectious complications in liver transplant recipients, as well as in close contacts and healthcare professionals.³

Influenza. Pretransplant influenza vaccine and posttransplant annual influenza vaccines are necessary.

Pneumococcal immunization should additionally be provided prior to transplant and repeated every 3 to 5 years thereafter.^3,91

A number of other vaccinations should also be completed before transplant, including the hepatitis A and B vaccines and the tetanus/diphtheria/acellular pertussis vaccines. However, these vaccinations have not been shown to be detrimental to patients after transplant.⁹¹

Varicella and zoster vaccines should be given before liver transplant—zoster in patients over age 60, and varicella in patients with no immunity. Live vaccines, including varicella and zoster vaccines, are contraindicated after liver transplant.³

Human papillomavirus. The bivalent human papillomavirus vaccine can be given before transplant in females ages 9 to 26; the quadrivalent vaccine is beneficial in those ages 9 to 26 and in women under age 45.^3,91

IMMUNOSUPPRESSION CARRIES RISK OF INFECTION

Most liver transplant patients require prolonged immunosuppressive therapy. This comes with an increased risk of new or recurrent infections, potentially causing death and significant morbidity.

Evaluation of existing risk factors, appropriate prophylaxis and immunization, timely diagnosis, and treatment of such infections are therefore essential steps for the successful management of liver transplant recipients.

Botulinum toxin is commonly used to treat movement disorders of the head and neck. It was first used to treat focal eye dystonia (blepharospasm) and laryngeal dystonia (spasmodic dysphonia) and is now also used for other head and neck dystonias, movement disorders, and muscle spasticity or contraction.

This article reviews the use of botulinum toxin for primary disorders of the laryngopharynx—adductor and abductor spasmodic dysphonias, laryngopharyngeal tremor, and cricopharyngeus muscle dysfunction—and its efficacy and side effects for the different conditions.

ABNORMAL MUSCLE MOVEMENT

Dystonia is abnormal muscle movement characterized by repetitive involuntary contractions. Dystonic contractions are described as either sustained (tonic) or spasmodic (clonic) and are typically induced by conscious action to move the muscle group.^1,2 Dystonia can be categorized according to the amount of muscle involvement: generalized (widespread muscle activity), segmental (involving neighboring groups of muscles), or focal (involving only one or a few local small muscles).³ Activity may be associated with gross posturing and disfigurement, depending on the size and location of the muscle contractions, although the muscle action is usually normal during rest.

The cause of dystonia has been the focus of much debate and investigation. Some types of dystonia have strong family inheritance patterns, but most are sporadic, possibly brought on by trauma or infection. In most cases, dystonia is idiopathic, although it may be associated with other muscle group dystonias, tremor, neurologic injury or insults, other neurologic diseases and neurodegenerative disorders, or tardive syndromes.¹ Because of the relationship with other neurologic diseases, consultation with a neurologist should be considered. 

Treatment of the muscle contractions of the various dystonias includes drug therapy and physical, occupational, and voice therapy. Botulinum toxin is a principal treatment for head and neck dystonias and works by blocking muscular contractions.⁴ It has the advantages of having few side effects and predictable results for many conditions, although repeat injections are usually required to achieve a sustained effect.

LARYNGEAL DYSTONIAS CAUSE VOICE ABNORMALITIES

Dystonia is most often idiopathic

The most common laryngeal dystonia is spasmodic dysphonia, a focal dystonia of the larynx. It is subdivided into two types according to whether spasm of the vocal folds occurs during adduction or abduction.

Adductor spasmodic dysphonia accounts for 80% to 90% of cases. It is characterized by irregular speech with pitch breaks and a strained or strangulated voice. It was formerly treated by resection of the nerve to the vocal folds, but results were neither consistent nor persistent. Currently, the primary treatment is injection of botulinum toxin, which has a high success rate,⁵ with patients reporting about a 90% return of normal function.

Abductor spasmodic dysphonia accounts for 10% to 20% of cases.⁶ Patients have a breathy quality to the voice with a short duration of vocalization due to excessive loss of air on phonation. This is especially noticeable when the patient speaks words that begin with a voiceless consonant followed by a vowel (eg, pat, puppy). Response to botulinum toxin injection is more variable,⁶ possibly because of the pathophysiology of the disorder or because of the technical challenges of administering the injection.

Fewer than 1% of patients have both abductor and adductor components, and their treatment can be particularly challenging.

Adductor spasmodic dysphonia: Treatment usually successful

Botulinum toxin can be injected for adductor spasmodic dysphonia via a number of approaches, the most common being through the cricothyroid membrane (Figure 1). Injections can be made into one or both vocal folds and can be performed under guidance with laryngeal electromyography or with a flexible laryngoscope to visualize the larynx.

Patients typically experience breathiness beginning 1 or 2 days after the injection, and this effect may last for up to 2 weeks. During that time, the patient may be more susceptible to aspiration of thin liquids and so is instructed to drink cautiously. Treatment benefits typically last for 3 to 6 months. As the botulinum toxin wears off, the patient notices a gradual increase in vocal straining and effort.

Dosages of botulinum toxin for subsequent treatments are adjusted by balancing the period of benefit with postinjection breathiness. The desire of the patient should be paramount. Some are willing to tolerate more side effects to avoid frequent injections, so they can be given a larger dose. Others cannot tolerate the breathiness but are willing to accept more frequent injections, so they should be given a smaller dose. In rare cases, patients have significant breathiness from even small doses; they may be helped by injecting into only one vocal fold or, alternatively, into a false vocal fold, allowing diffusion of the toxin down to the muscle of the true vocal fold.

Abductor spasmodic dysphonia: Treatment more challenging

The success of botulinum toxin treatment for abductor spasmodic dysphonia is more variable than for the adductor type. The injections are made into the posterior cricoarytenoid muscle (Figure 2); because this muscle cannot be directly visualized, this procedure requires guidance with laryngeal electromyography. Most patients note improvement, and about 20% have a good response.⁶ Most require a second injection about 1 month later, often on the other side. Bilateral injections at one sitting may compromise the airway, and vocal fold motion should be evaluated at the time of the contralateral injection to assess airway patency. Interest has increased in simultaneous bilateral injections with lower doses of botulinum toxin, and this approach has been shown to be safe.⁷

ESSENTIAL TREMOR OF THE VOICE

Essential tremor is an action tremor that can occur with voluntary movement. It can occur anywhere in the body, often the head or hand, but the voice can also be affected. About half of cases are hereditary. Essential tremor of the voice causes a rhythmic oscillation of pitch and intensity.

Consultation with a neurologist is recommended to evaluate the cause, although voice tremor is often idiopathic and occurs in about 30% of patients with essential tremor in the arms or legs, as well as in about 30% of patients with spasmodic dysphonia. Extremity tremor can usually be successfully managed medically, but this is not true for voice tremor.

Botulinum toxin injection is the mainstay of treatment for essential tremor of the voice, although its success is marginal. About two-thirds of patients have some degree of improvement from traditional botulinum toxin injections in the true vocal fold.⁸

Patients almost always require repeat injections to obtain a sustained effect

The results of treatment are likely to be inconsistent because tremor tends to involve several different muscles used in voice production, commonly in the soft palate, tongue base, pharyngeal walls, strap muscles, false vocal folds, and true vocal folds. A location-oriented tremor scoring system⁹ can help identify the involved muscles to guide injections. Treatment is less likely to be successful in patients with multiple sites of voice tremor. Injection into the false vocal fold, true vocal fold, and interarytenoid muscle¹⁰ can safely be performed; injections into the palate, tongue base, and strap muscles are to be avoided because of the high risk of postinjection aspiration.

Patients who have good results can have repeat treatments as needed. The dosage of botulinum toxin is adjusted according to response, side effects (eg, breathy voice, dysphagia), and patient preference.

CRICOPHARYNGEUS MUSCLE DYSFUNCTION: TROUBLE SWALLOWING

Dysfunction of the cricopharyngeus muscle causes difficulty swallowing, especially swallowing solid foods. It can be attributed to a mechanical stricture or to hyperfunction (spasm).

Mechanical stricture at the esophageal inlet frequently occurs in patients who have had a total laryngectomy for advanced laryngeal cancer. Fibrosis tends to be worse in patients who have also undergone radiation therapy.

Stricture can be treated with botulinum toxin injections and dilation. Conservative treatment is preferred to surgical myotomy for patients with complex postlaryngectomy anatomy and scarring from radiation therapy.

Cricopharyngeus muscle spasm or hyperfunction can be an important cause of dysphagia, especially in the elderly. Patients should be evaluated with barium esophagography or a modified barium swallow. The finding of a cricopharyngeal “bar” provides evidence of contraction of the muscle that impedes the passage of food.

Botulinum toxin injections for cricopharyngeus muscle dysfunction (Figure 2) can be effective in some cases, especially if the toxin is injected bilaterally. However, because the cricopharyngeus muscle plays an important role in preventing esophageal reflux into the laryngopharynx, botulinum toxin injection in patients with substantial hiatal hernia or laryngopharyngeal reflux disease should only be done with caution. In addition, treatment of reflux disease should be considered in any patient undergoing botulinum toxin injection for cricopharyngeus muscle dysfunction.

Most patients require repeat injections when the toxin wears off, although occasionally one or two injections provide long-term or permanent relief. Dosages are adjusted for the patient’s age, the presence of other swallowing problems, and reflux. Patients may experience increased difficulty swallowing for 1 or 2 weeks after the procedure and so should be counseled to eat slowly and carefully.

Botulinum toxin is commonly used to treat movement disorders of the head and neck. It was first used to treat focal eye dystonia (blepharospasm) and laryngeal dystonia (spasmodic dysphonia) and is now also used for other head and neck dystonias, movement disorders, and muscle spasticity or contraction.

This article reviews the use of botulinum toxin for primary disorders of the laryngopharynx—adductor and abductor spasmodic dysphonias, laryngopharyngeal tremor, and cricopharyngeus muscle dysfunction—and its efficacy and side effects for the different conditions.

ABNORMAL MUSCLE MOVEMENT

Dystonia is abnormal muscle movement characterized by repetitive involuntary contractions. Dystonic contractions are described as either sustained (tonic) or spasmodic (clonic) and are typically induced by conscious action to move the muscle group.^1,2 Dystonia can be categorized according to the amount of muscle involvement: generalized (widespread muscle activity), segmental (involving neighboring groups of muscles), or focal (involving only one or a few local small muscles).³ Activity may be associated with gross posturing and disfigurement, depending on the size and location of the muscle contractions, although the muscle action is usually normal during rest.

The cause of dystonia has been the focus of much debate and investigation. Some types of dystonia have strong family inheritance patterns, but most are sporadic, possibly brought on by trauma or infection. In most cases, dystonia is idiopathic, although it may be associated with other muscle group dystonias, tremor, neurologic injury or insults, other neurologic diseases and neurodegenerative disorders, or tardive syndromes.¹ Because of the relationship with other neurologic diseases, consultation with a neurologist should be considered. 

Treatment of the muscle contractions of the various dystonias includes drug therapy and physical, occupational, and voice therapy. Botulinum toxin is a principal treatment for head and neck dystonias and works by blocking muscular contractions.⁴ It has the advantages of having few side effects and predictable results for many conditions, although repeat injections are usually required to achieve a sustained effect.

LARYNGEAL DYSTONIAS CAUSE VOICE ABNORMALITIES

Dystonia is most often idiopathic

The most common laryngeal dystonia is spasmodic dysphonia, a focal dystonia of the larynx. It is subdivided into two types according to whether spasm of the vocal folds occurs during adduction or abduction.

Adductor spasmodic dysphonia accounts for 80% to 90% of cases. It is characterized by irregular speech with pitch breaks and a strained or strangulated voice. It was formerly treated by resection of the nerve to the vocal folds, but results were neither consistent nor persistent. Currently, the primary treatment is injection of botulinum toxin, which has a high success rate,⁵ with patients reporting about a 90% return of normal function.

Abductor spasmodic dysphonia accounts for 10% to 20% of cases.⁶ Patients have a breathy quality to the voice with a short duration of vocalization due to excessive loss of air on phonation. This is especially noticeable when the patient speaks words that begin with a voiceless consonant followed by a vowel (eg, pat, puppy). Response to botulinum toxin injection is more variable,⁶ possibly because of the pathophysiology of the disorder or because of the technical challenges of administering the injection.

Fewer than 1% of patients have both abductor and adductor components, and their treatment can be particularly challenging.

Adductor spasmodic dysphonia: Treatment usually successful

Botulinum toxin can be injected for adductor spasmodic dysphonia via a number of approaches, the most common being through the cricothyroid membrane (Figure 1). Injections can be made into one or both vocal folds and can be performed under guidance with laryngeal electromyography or with a flexible laryngoscope to visualize the larynx.

Patients typically experience breathiness beginning 1 or 2 days after the injection, and this effect may last for up to 2 weeks. During that time, the patient may be more susceptible to aspiration of thin liquids and so is instructed to drink cautiously. Treatment benefits typically last for 3 to 6 months. As the botulinum toxin wears off, the patient notices a gradual increase in vocal straining and effort.

Dosages of botulinum toxin for subsequent treatments are adjusted by balancing the period of benefit with postinjection breathiness. The desire of the patient should be paramount. Some are willing to tolerate more side effects to avoid frequent injections, so they can be given a larger dose. Others cannot tolerate the breathiness but are willing to accept more frequent injections, so they should be given a smaller dose. In rare cases, patients have significant breathiness from even small doses; they may be helped by injecting into only one vocal fold or, alternatively, into a false vocal fold, allowing diffusion of the toxin down to the muscle of the true vocal fold.

Abductor spasmodic dysphonia: Treatment more challenging

The success of botulinum toxin treatment for abductor spasmodic dysphonia is more variable than for the adductor type. The injections are made into the posterior cricoarytenoid muscle (Figure 2); because this muscle cannot be directly visualized, this procedure requires guidance with laryngeal electromyography. Most patients note improvement, and about 20% have a good response.⁶ Most require a second injection about 1 month later, often on the other side. Bilateral injections at one sitting may compromise the airway, and vocal fold motion should be evaluated at the time of the contralateral injection to assess airway patency. Interest has increased in simultaneous bilateral injections with lower doses of botulinum toxin, and this approach has been shown to be safe.⁷

ESSENTIAL TREMOR OF THE VOICE

Essential tremor is an action tremor that can occur with voluntary movement. It can occur anywhere in the body, often the head or hand, but the voice can also be affected. About half of cases are hereditary. Essential tremor of the voice causes a rhythmic oscillation of pitch and intensity.

Consultation with a neurologist is recommended to evaluate the cause, although voice tremor is often idiopathic and occurs in about 30% of patients with essential tremor in the arms or legs, as well as in about 30% of patients with spasmodic dysphonia. Extremity tremor can usually be successfully managed medically, but this is not true for voice tremor.

Botulinum toxin injection is the mainstay of treatment for essential tremor of the voice, although its success is marginal. About two-thirds of patients have some degree of improvement from traditional botulinum toxin injections in the true vocal fold.⁸

Patients almost always require repeat injections to obtain a sustained effect

The results of treatment are likely to be inconsistent because tremor tends to involve several different muscles used in voice production, commonly in the soft palate, tongue base, pharyngeal walls, strap muscles, false vocal folds, and true vocal folds. A location-oriented tremor scoring system⁹ can help identify the involved muscles to guide injections. Treatment is less likely to be successful in patients with multiple sites of voice tremor. Injection into the false vocal fold, true vocal fold, and interarytenoid muscle¹⁰ can safely be performed; injections into the palate, tongue base, and strap muscles are to be avoided because of the high risk of postinjection aspiration.

Patients who have good results can have repeat treatments as needed. The dosage of botulinum toxin is adjusted according to response, side effects (eg, breathy voice, dysphagia), and patient preference.

CRICOPHARYNGEUS MUSCLE DYSFUNCTION: TROUBLE SWALLOWING

Dysfunction of the cricopharyngeus muscle causes difficulty swallowing, especially swallowing solid foods. It can be attributed to a mechanical stricture or to hyperfunction (spasm).

Mechanical stricture at the esophageal inlet frequently occurs in patients who have had a total laryngectomy for advanced laryngeal cancer. Fibrosis tends to be worse in patients who have also undergone radiation therapy.

Stricture can be treated with botulinum toxin injections and dilation. Conservative treatment is preferred to surgical myotomy for patients with complex postlaryngectomy anatomy and scarring from radiation therapy.

Cricopharyngeus muscle spasm or hyperfunction can be an important cause of dysphagia, especially in the elderly. Patients should be evaluated with barium esophagography or a modified barium swallow. The finding of a cricopharyngeal “bar” provides evidence of contraction of the muscle that impedes the passage of food.

Botulinum toxin injections for cricopharyngeus muscle dysfunction (Figure 2) can be effective in some cases, especially if the toxin is injected bilaterally. However, because the cricopharyngeus muscle plays an important role in preventing esophageal reflux into the laryngopharynx, botulinum toxin injection in patients with substantial hiatal hernia or laryngopharyngeal reflux disease should only be done with caution. In addition, treatment of reflux disease should be considered in any patient undergoing botulinum toxin injection for cricopharyngeus muscle dysfunction.

Most patients require repeat injections when the toxin wears off, although occasionally one or two injections provide long-term or permanent relief. Dosages are adjusted for the patient’s age, the presence of other swallowing problems, and reflux. Patients may experience increased difficulty swallowing for 1 or 2 weeks after the procedure and so should be counseled to eat slowly and carefully.

Botulinum toxin is commonly used to treat movement disorders of the head and neck. It was first used to treat focal eye dystonia (blepharospasm) and laryngeal dystonia (spasmodic dysphonia) and is now also used for other head and neck dystonias, movement disorders, and muscle spasticity or contraction.

This article reviews the use of botulinum toxin for primary disorders of the laryngopharynx—adductor and abductor spasmodic dysphonias, laryngopharyngeal tremor, and cricopharyngeus muscle dysfunction—and its efficacy and side effects for the different conditions.

ABNORMAL MUSCLE MOVEMENT

Dystonia is abnormal muscle movement characterized by repetitive involuntary contractions. Dystonic contractions are described as either sustained (tonic) or spasmodic (clonic) and are typically induced by conscious action to move the muscle group.^1,2 Dystonia can be categorized according to the amount of muscle involvement: generalized (widespread muscle activity), segmental (involving neighboring groups of muscles), or focal (involving only one or a few local small muscles).³ Activity may be associated with gross posturing and disfigurement, depending on the size and location of the muscle contractions, although the muscle action is usually normal during rest.

The cause of dystonia has been the focus of much debate and investigation. Some types of dystonia have strong family inheritance patterns, but most are sporadic, possibly brought on by trauma or infection. In most cases, dystonia is idiopathic, although it may be associated with other muscle group dystonias, tremor, neurologic injury or insults, other neurologic diseases and neurodegenerative disorders, or tardive syndromes.¹ Because of the relationship with other neurologic diseases, consultation with a neurologist should be considered. 

Treatment of the muscle contractions of the various dystonias includes drug therapy and physical, occupational, and voice therapy. Botulinum toxin is a principal treatment for head and neck dystonias and works by blocking muscular contractions.⁴ It has the advantages of having few side effects and predictable results for many conditions, although repeat injections are usually required to achieve a sustained effect.

LARYNGEAL DYSTONIAS CAUSE VOICE ABNORMALITIES

Dystonia is most often idiopathic

The most common laryngeal dystonia is spasmodic dysphonia, a focal dystonia of the larynx. It is subdivided into two types according to whether spasm of the vocal folds occurs during adduction or abduction.

Adductor spasmodic dysphonia accounts for 80% to 90% of cases. It is characterized by irregular speech with pitch breaks and a strained or strangulated voice. It was formerly treated by resection of the nerve to the vocal folds, but results were neither consistent nor persistent. Currently, the primary treatment is injection of botulinum toxin, which has a high success rate,⁵ with patients reporting about a 90% return of normal function.

Abductor spasmodic dysphonia accounts for 10% to 20% of cases.⁶ Patients have a breathy quality to the voice with a short duration of vocalization due to excessive loss of air on phonation. This is especially noticeable when the patient speaks words that begin with a voiceless consonant followed by a vowel (eg, pat, puppy). Response to botulinum toxin injection is more variable,⁶ possibly because of the pathophysiology of the disorder or because of the technical challenges of administering the injection.

Fewer than 1% of patients have both abductor and adductor components, and their treatment can be particularly challenging.

Adductor spasmodic dysphonia: Treatment usually successful

Botulinum toxin can be injected for adductor spasmodic dysphonia via a number of approaches, the most common being through the cricothyroid membrane (Figure 1). Injections can be made into one or both vocal folds and can be performed under guidance with laryngeal electromyography or with a flexible laryngoscope to visualize the larynx.

Patients typically experience breathiness beginning 1 or 2 days after the injection, and this effect may last for up to 2 weeks. During that time, the patient may be more susceptible to aspiration of thin liquids and so is instructed to drink cautiously. Treatment benefits typically last for 3 to 6 months. As the botulinum toxin wears off, the patient notices a gradual increase in vocal straining and effort.

Dosages of botulinum toxin for subsequent treatments are adjusted by balancing the period of benefit with postinjection breathiness. The desire of the patient should be paramount. Some are willing to tolerate more side effects to avoid frequent injections, so they can be given a larger dose. Others cannot tolerate the breathiness but are willing to accept more frequent injections, so they should be given a smaller dose. In rare cases, patients have significant breathiness from even small doses; they may be helped by injecting into only one vocal fold or, alternatively, into a false vocal fold, allowing diffusion of the toxin down to the muscle of the true vocal fold.

Abductor spasmodic dysphonia: Treatment more challenging

The success of botulinum toxin treatment for abductor spasmodic dysphonia is more variable than for the adductor type. The injections are made into the posterior cricoarytenoid muscle (Figure 2); because this muscle cannot be directly visualized, this procedure requires guidance with laryngeal electromyography. Most patients note improvement, and about 20% have a good response.⁶ Most require a second injection about 1 month later, often on the other side. Bilateral injections at one sitting may compromise the airway, and vocal fold motion should be evaluated at the time of the contralateral injection to assess airway patency. Interest has increased in simultaneous bilateral injections with lower doses of botulinum toxin, and this approach has been shown to be safe.⁷

ESSENTIAL TREMOR OF THE VOICE

Essential tremor is an action tremor that can occur with voluntary movement. It can occur anywhere in the body, often the head or hand, but the voice can also be affected. About half of cases are hereditary. Essential tremor of the voice causes a rhythmic oscillation of pitch and intensity.

Consultation with a neurologist is recommended to evaluate the cause, although voice tremor is often idiopathic and occurs in about 30% of patients with essential tremor in the arms or legs, as well as in about 30% of patients with spasmodic dysphonia. Extremity tremor can usually be successfully managed medically, but this is not true for voice tremor.

Botulinum toxin injection is the mainstay of treatment for essential tremor of the voice, although its success is marginal. About two-thirds of patients have some degree of improvement from traditional botulinum toxin injections in the true vocal fold.⁸

Patients almost always require repeat injections to obtain a sustained effect

The results of treatment are likely to be inconsistent because tremor tends to involve several different muscles used in voice production, commonly in the soft palate, tongue base, pharyngeal walls, strap muscles, false vocal folds, and true vocal folds. A location-oriented tremor scoring system⁹ can help identify the involved muscles to guide injections. Treatment is less likely to be successful in patients with multiple sites of voice tremor. Injection into the false vocal fold, true vocal fold, and interarytenoid muscle¹⁰ can safely be performed; injections into the palate, tongue base, and strap muscles are to be avoided because of the high risk of postinjection aspiration.

Patients who have good results can have repeat treatments as needed. The dosage of botulinum toxin is adjusted according to response, side effects (eg, breathy voice, dysphagia), and patient preference.

CRICOPHARYNGEUS MUSCLE DYSFUNCTION: TROUBLE SWALLOWING

Dysfunction of the cricopharyngeus muscle causes difficulty swallowing, especially swallowing solid foods. It can be attributed to a mechanical stricture or to hyperfunction (spasm).

Mechanical stricture at the esophageal inlet frequently occurs in patients who have had a total laryngectomy for advanced laryngeal cancer. Fibrosis tends to be worse in patients who have also undergone radiation therapy.

Stricture can be treated with botulinum toxin injections and dilation. Conservative treatment is preferred to surgical myotomy for patients with complex postlaryngectomy anatomy and scarring from radiation therapy.

Cricopharyngeus muscle spasm or hyperfunction can be an important cause of dysphagia, especially in the elderly. Patients should be evaluated with barium esophagography or a modified barium swallow. The finding of a cricopharyngeal “bar” provides evidence of contraction of the muscle that impedes the passage of food.

Botulinum toxin injections for cricopharyngeus muscle dysfunction (Figure 2) can be effective in some cases, especially if the toxin is injected bilaterally. However, because the cricopharyngeus muscle plays an important role in preventing esophageal reflux into the laryngopharynx, botulinum toxin injection in patients with substantial hiatal hernia or laryngopharyngeal reflux disease should only be done with caution. In addition, treatment of reflux disease should be considered in any patient undergoing botulinum toxin injection for cricopharyngeus muscle dysfunction.

Most patients require repeat injections when the toxin wears off, although occasionally one or two injections provide long-term or permanent relief. Dosages are adjusted for the patient’s age, the presence of other swallowing problems, and reflux. Patients may experience increased difficulty swallowing for 1 or 2 weeks after the procedure and so should be counseled to eat slowly and carefully.

Clinicians may be encountering more cannabis users than before, and may be encountering users with complications hitherto unseen. Several trends may explain this phenomenon: the legal status of cannabis is changing, cannabis today is more potent than in the past, and enthusiasts are conjuring new ways to enjoy this substance.

This article discusses the history, pharmacology, and potential complications of cannabis use.

A LONG AND TANGLED HISTORY

Cannabis is a broad term that refers to the cannabis plant and its preparations, such as marijuana and hashish, as well as to a family of more than 60 bioactive substances called cannabinoids. It is the most commonly used illegal drug in the world, with an estimated 160 million users. Each year, about 2.4 million people in the United States use it for the first time.^1,2

Cannabis has been used throughout the world for recreational and spiritual purposes for nearly 5,000 years, beginning with the fabled Celestial Emperors of China. The tangled history of cannabis in America began in the 17th century, when farmers were required by law to grow it as a fiber crop. It later found its way into the US Pharmacopeia for a wide range of indications. During the long prelude to Prohibition in the latter half of the 19th century, the US government became increasingly suspicious of mind-altering substances and began restricting its prescription in 1934, culminating in its designation by the US Food and Drug Administration as a schedule I controlled substance in 1970.

Investigation into the potential medical uses for the different chemicals within cannabis is ongoing, as is debate over its changing legality and usefulness to society. The apparent cognitive dissonance surrounding the use and advocacy of medical marijuana is beyond the scope of this review,³ which will instead restrict itself to what is known of the cannabinoids and to the recreational use of cannabis.

THC IS THE PRINCIPAL PSYCHOACTIVE MOLECULE

Delta-9 tetrahydrocannabinol (THC), first isolated in 1964, was identified as the principal psychoactive constituent of cannabis in 2002.⁴

Two G-protein–linked cannabinoid receptors cloned in the 1990s—CB1 and CB2—were found to be a part of a system of endocannabinoid receptors present throughout the body, from the brain to the immune system to the vas deferens.⁵ Both receptors inhibit cellular excitation by activating inwardly rectifying potassium channels. These receptors are mostly absent in the brainstem, which may explain why cannabis use rarely causes life-threatening autonomic dysfunction. Although the intoxicating effects of marijuana are mediated by CB1 receptors, the specific mechanisms underlying the cannabis “high” are unclear.⁶

CANNABINOIDS ARE LIPID-SOLUBLE

The rate of absorption of cannabinoids depends on the route of administration and the type of cannabis product used. When cannabis products are smoked, up to 35% of THC is available, and the average time to peak serum concentration is 8 minutes.⁷ The peak concentration depends on the dose.

On the other hand, when cannabis products (eg, nabilone, dronabinol) are ingested, absorption is unpredictable because THC is unstable in gastric acid and undergoes first-pass metabolism in the liver, which reduces the drug’s bioavailability. Up to 20% of an ingested dose of THC is absorbed, and the time to peak serum concentration averages between 2 and 4 hours. Consequently, many users prefer to smoke cannabis as a means to control the desired effects.

Cannabinoids are lipid-soluble. They accumulate in fatty tissue in a biphasic pattern, initially moving into highly vascularized tissue such as the liver before accumulating in less well-vascularized tissue such as fat. They are then slowly released from fatty tissue as the fat turns over. THC itself has a volume of distribution of about 2.5 to 3.5 L/kg. It crosses the placenta and enters breast milk.⁸

THC is metabolized by the cytochrome P450 system, primarily by the enzymes CYP­2C9 and CYP3A4. Its primary metabolite, 11-hydroxy-delta-9 THC, is also active, but subsequent metabolism produces many other inactive metabolites. THC is eliminated in feces and urine, and its half-life ranges from 2 to nearly 60 hours.⁸

A LITTLE ABOUT PLANTS AND STREET NAMES

The plant from which THC and nearly a hundred other chemicals, including cannabinoids, are derived has been called many things over the years:

Hemp is a tall fibrous plant grown for rope and fabric that was used as legal tender in early America. In the mid-19th century, there were over 16 million acres of hemp plantations. Hemp contains very low THC concentrations.

Cannabis is an annual flowering herb that is predominantly diecious (ie, there are male and female plants). After a centuries-long debate among taxonomists, the two principal species are considered to be C sativa and C indica, although today many cannabis cultivars are grown by a great number of breeding enthusiasts.

THC levels in marijuana have increased from about 5% historically to over 30% in some samples today

Concentrations of THC vary widely among cannabis cultivars, ranging historically from around 5% to today’s highly selectively bred species containing more than 30%. Concentrations in seized cannabis have been measured as high as 37%, although the average is around 11%.⁹ This concentration is defined by the percent of THC per dried mass of plant material tested, usually via gas chromatography.

Hashish is a solid or resinous preparation of the trichomes, or glandular hairs, that grow on the cannabis plant, chiefly on its flowers. Various methods to separate the trichomes from the rest of the plant result in a powder called kief that is then compressed into blocks or bricks. THC concentrations as high as 66% have been measured in nondomestic sources of hashish.⁹

Hash oil is a further purification, produced by using solvents to dissolve the resin and by filtering out remaining plant material. Evaporating the solvent produces hash oil, sometimes called butane hash oil or honey oil. This process has recently led to an increasing number of home explosions, as people attempt to make the product themselves but do not take suitable precautions when using flammable solvents such as butane. THC concentrations as high as 81% have been measured in nondomestic sources of hash oil.⁹

Other names for hash oil are dab, wax, and budder. Cannabis enthusiasts refer to the use of hash oil as dabbing, which involves heating a small amount (dab) of the product using a variety of paraphernalia and inhaling the vapor.

IT’S ALL ABOUT GETTING HIGH

One user’s high is another user’s acute toxic effect

For recreational users, the experience has always been about being intoxicated—getting high. The psychological effects range broadly from positive to negative and vary both within and between users, depending on the dose and route of administration. Additional factors that influence the psychological effects include the social and physical settings of drug use and even the user’s expectations. One user’s high is another user’s acute toxic effect.

Although subjective reports of the cannabis experience vary greatly, it typically begins with a feeling of dizziness or lightheadedness followed by a relaxed calm and a feeling of being somewhat “disconnected.” There is a quickening of the sense of humor, described by some as a fatuous euphoria; often there is silly giggling. Awareness of the senses and of music may be increased. Appetite increases, and time seems to pass quickly. Eventually, the user becomes drowsy and experiences decreased attention and difficulty maintaining a coherent conversation. Slowed reaction time and decreased psychomotor activity may also occur. The user may drift into daydreams and eventually fall asleep.

Common negative acute effects of getting high can include mild to severe anxiety and feeling tense or agitated. Clumsiness, headache, and confusion are also possible. Lingering effects the following day may include dry mouth, dry eyes, fatigue, slowed thinking, and slowed recall.⁶

ACUTE PHYSICAL EFFECTS

Acute physical effects of cannabis use include a rapid onset of increased airway conductance, decreased intraocular pressure, and conjunctival injection. A single cannabis cigarette can also induce cardiovascular effects including a dose-dependent increase in heart rate and blood pressure. Chronic users, however, can experience a decreased heart rate, lower blood pressure, and postural hypotension.

In a personal communication, colleagues in Colorado—where recreational use of cannabis was legalized in 2012—described a sharp increase (from virtually none) in the number of adults presenting to the emergency department with cannabis intoxication since 2012. Their patients experienced palpitations, light-headedness, and severe ataxia lasting as long as 12 hours, possibly reflecting the greater potency of current cannabis products. Most of these patients required only supportive care.

Acute effects of cannabis include increased airway conductance, decreased intraocular pressure, and conjunctival injection

Other acute adverse cardiovascular reactions that have been reported include atrial fibrillation, ventricular tachycardia, and a fivefold increased risk of myocardial infarction in the 60 minutes following cannabis use, which subsequently drops sharply to baseline levels.¹⁰ Investigations into the cardiovascular effects of cannabis are often complicated by concurrent use of other drugs such as tobacco or cocaine. Possible mechanisms of injury include alterations in coronary microcirculation or slowed coronary flow. In fact, one author found that cannabis users with a history of myocardial infarction had a risk of death 4.2 times higher than users with no history of myocardial infarction.^11,12

In children, acute toxicity has been reported from a variety of exposures to cannabis and hashish, including a report of an increase in pediatric cannabis exposures following the changes in Colorado state laws.¹³ Most of these patients had altered mental status ranging from drowsiness to coma; one report describes a child who experienced a first-time seizure. These patients unfortunately often underwent extensive evaluations such as brain imaging and lumbar puncture, and mechanical ventilation to protect the airway. Earlier consideration of cannabis exposure in these patients might have limited unnecessary testing. Supportive care is usually all that is needed, and most of these patients fully recover.^13–17

CHRONIC EFFECTS

Cannabinoids cause a variety of adverse effects, but the ultimate risk these changes pose to human health has been difficult to calculate. Long-term studies are confounded by possible inaccuracies of patient self-reporting of cannabis use, poor control of covariates, and disparate methodologies.

For more than a century, cannabis use has been reported to cause both acute psychotic symptoms and persistent psychotic disorders.¹⁸ But the strength of this relationship is modest. Cannabis is more likely a component cause that, in addition to other factors (eg, specific genetic polymorphisms), contributes to the risk of schizophrenia. Individuals with prodromal symptoms and those who have experienced discrete episodes of psychosis related to cannabis use should be discouraged from using cannabis and cannabinoids.^19–21

Mounting evidence implicates chronic cannabis use as a cause of long-term medical problems

Mounting evidence implicates chronic cannabis use as a cause of long-term medical problems including chronic bronchitis,²² elevated rates of myocardial infarction and dysrhythmias,¹¹ bone loss,²³ and cancers at eight different sites including the lung, head, and neck.²⁴ In view of these chronic effects, healthcare providers should caution their patients about cannabis use, as we do about other drugs such as tobacco.

WITHDRAWAL SYNDROME RECOGNIZED

Until recently, neither clinicians nor users recognized a withdrawal syndrome associated with chronic use of cannabis, probably because this syndrome is not as severe as withdrawal from other controlled substances such as opioids or sedative-hypnotics. A number of studies, however, have reported subtle cannabis withdrawal symptoms that are similar to those associated with tobacco withdrawal.

As such, the fifth and latest edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5)²⁵ characterized withdrawal from cannabis use in 2013. The DSM-5 criteria require cessation of heavy or prolonged use of cannabis (ie, daily or almost daily over a period of at least a few months) and three or more of the following withdrawal symptoms:

• Irritability and anger
• Nervousness
• Sleep difficulty or insomnia
• Decreased appetite or weight loss
• Restlessness
• Depressed mood
• Physical symptoms causing discomfort.

Medical treatment of cannabis withdrawal has included a range of antidepressants, mood stabilizers, and alpha-2-adrenergic agonists, all of which have limited success.²⁶ Symptoms of cannabis withdrawal tend to be most intense soon after cessation and decline over the next few weeks.²⁷

CANNABINOID HYPEREMESIS SYNDROME

First reported in 2004,²⁸ cannabinoid hyperemesis syndrome is a recurrent disorder, the pathophysiology of which is poorly understood. It has three phases.

The first phase is a prodrome that may last months or years and is characterized by morning nausea, fear of vomiting, and abdominal discomfort. During this phase, the patient maintains normal eating patterns and may well increase his or her cannabis use due to its well-known antiemetic effects.

The second phase is the hyperemetic phase, characterized by intense, incapacitating emesis with episodes of vomiting throughout the day. These symptoms can be relieved only with frequent hot baths, a feature that distinguishes cannabinoid hyperemesis syndrome from other vomiting syndromes. Hot-water bathing is reported to be a compulsive but learned behavior in which the patient learns that only hot water will provide relief. The extent of relief depends on the temperature of the water—the hotter, the better. Symptoms recur as the water cools.²⁸ Patients often present to the emergency department repeatedly with recurrent symptoms and may remain misdiagnosed or subjected to repeated extensive evaluation including laboratory testing and imaging, which are usually not revealing. If the patient has not been accurately diagnosed, there may be reported weight loss of at least 5 kg.

The third phase, recovery, may take several months to complete, possibly because of the prolonged terminal elimination time of cannabinoids. Complete cessation of cannabis use, including synthetic cannabinoids, is usually necessary.²⁹

Diagnostic criteria for cannabinoid hyperemesis syndrome have been suggested, based on a retrospective case series that included 98 patients.³⁰ The most common features of these affected patients were:

• Severe cyclical vomiting, predominantly in the morning
• Resolution of symptoms with cessation of cannabis use
• Symptomatic relief with hot showers or baths
• Abdominal pain
• At least weekly use of cannabis.

Interestingly, long-term cannabis use has been cited as a critical identifying feature of these patients, with the duration of cannabis use ranging from 10 to 16 years.^31,32 Other reports show greater variability in duration of cannabis use before the onset of cannabinoid hyperemesis syndrome. In the large study noted above,³⁰ 32% of users reported their duration of cannabis use to be less than 1 year, rendering this criterion less useful.

How can cannabis both cause and prevent vomiting?

The body controls nausea and vomiting via complex circuitry in the brain and gut that involves many neurotransmitters (eg, dopamine, serotonin, substance P) that interact with receptors such as CB1, 5-HT_1–4, alpha adrenergic receptors, and mu receptors. Interestingly, cannabis use has antiemetic properties mediated by CB1 with a still unclear additional role of CB2 receptors. Data point to the existence of an underlying antiemetic tone mediated by the endocannabinoid system.

Unfortunately, the mechanism by which cannabinoid hyperemesis syndrome occurs is unknown and represents a paradoxical effect against the otherwise antiemetic effects of cannabis. Several theories have been proposed, including delayed gastric emptying, although only a third of patients demonstrated this on scintigraphy in one study.³⁰ Other theories include disturbance of the hypothalamic-pituitary axis, a buildup of highly lipophilic THC in the brain, and a down-regulation of cannabinoid receptors that results from chronic exposure.³⁰ Given that this syndrome has been recognized only relatively recently, one author has suggested the cause may be recent horticultural developments.⁵

Treating cannabinoid hyperemesis syndrome is difficult

Treatment of cannabinoid hyperemesis syndrome is notoriously difficult, with many authors reporting resistance to the usual first-line antiemetic drugs. Generally, treatment should include hydration and acid-suppression therapy because endoscopic evaluation of several patients has revealed varying degrees of esophagitis and gastritis.²⁹

Antiemetic therapy should target receptors known to mediate nausea and vomiting. In some cases, antiemetic drugs are more effective when used in combination. Agents include the serotonergic receptor antagonists ondansetron and granisetron, the dopamine antagonists prochlorperazine and metoclopramide, and even haloperidol.^33,34 Benzodiazepines may be effective by causing sedation, anxiolysis, and depression of the vomiting center.^34,35 Two antihistamines—dimenhydrinate and diphenhydramine—have antiemetic effects, perhaps by inhibiting acetylcholine.³⁴

Aprepitant is a neurokinin-1 antagonist that inhibits the action of substance P. When combined with a corticosteroid and a serotonin antagonist, it relieves nausea and vomiting in chemotherapy patients.^34,36

Corticosteroids such as dexamethasone are potent antiemetics thought to inhibit prostaglandin synthesis.³⁴

Capsaicin cream applied to the abdomen has also been reported to relieve symptoms, possibly through an interaction between the TRPv1 receptor and the endocannabinoid system.^37,38

DIAGNOSTIC TESTING

Cannabinoids are detectable in plasma and urine, with urine testing being more common.

Common laboratory methods include the enzyme-multiplied immunoassay technique (EMIT) and radioimmunoassay. Gas chromatography coupled with mass spectrometry is the most specific assay; it is used for confirmation and is the reference method.

EMIT is a qualitative urine test that detects 9-carboxy-THC as well as other THC metabolites. These urine tests detect all metabolites, and the result is reported as positive if the total concentration is greater than or equal to a prespecified threshold level, such as 20 ng/mL or 50 ng/mL. A positive test does not denote intoxication, nor does the test identify the source of THC (eg, cannabis, dronabinol, butane hash oil). EMIT does not detect nabilone. The National Institute on Drug Abuse guidelines for urine testing specify a test threshold concentration of 50 ng/mL for screening and 15 ng/mL for confirmation.

Many factors affect the detection of THC metabolites and their presence and duration in urine: dose, duration of use, route of exposure, hydration status, urine volume and concentration, and urine pH. THC metabolites have been detected in urine using gas chromatography-mass spectrometry for up to 7 days after smoking one marijuana cigarette.⁷ Chronic users have also been reported to have positive urine EMIT tests for up to 46 days after cannabis cessation.³⁹ Detection may be further complicated in chronic users: in one study, users produced both negative and positive specimens over 24 days, suggesting that diet and exercise may influence clearance.⁴⁰ Also, many factors are known to produce false-positive and false-negative results for these immunoassays (Table 1).^39,41

In the United States, penalties for driving under the influence of cannabis vary from state to state, and laws specify plasma testing for quantitative analysis. Some states use a threshold of 5 ng/mL in plasma to imply driving under the influence, whereas others use any detectable amount. Currently, there are no generally accepted guidelines for storage and testing of blood samples, despite the known instability of analytes.⁴²

Saliva, hair, and sweat can also be used for cannabinoid testing. Saliva is easy to collect, can be tested for metabolites to rule out passive cannabis exposure, and can be positive for up to 1 day after exposure. Calculating a blood or plasma concentration from a saliva sample is not possible, however.

Hair testing can also rule out passive exposure, but THC binds very little to melanin, resulting in very low concentrations requiring sensitive tests, such as gas chromatography with tandem mass spectrometry.

Only one device is commercially available for sweat testing; further work is needed to elucidate sweat excretion pharmacokinetics and the limitations of the collection devices.⁴³

CLINICAL MANAGEMENT IS GENERALLY SUPPORTIVE

Historically, clinical toxicity from recreational cannabis use is rarely serious or severe and generally responds to supportive care. Reports of cannabis exposure to poison centers are one-tenth of those reported for ethanol exposures annually.⁴⁴ Gastrointestinal decontamination with activated charcoal is not recommended, even for orally administered cannabis, since the risks outweigh the expected benefits. Agitation or anxiety may be treated with benzodiazepines as needed. There is no antidote for cannabis toxicity. The ever-increasing availability of high-concentration THC preparations may prompt more aggressive supportive measures in the future.

SYNTHETIC MARIJUANA ALTERNATIVES

Available since the early 2000s, herbal marijuana alternatives are legally sold as incense or potpourri and are often labeled “not for human consumption.” They are known by such brand names as K2 and Spice and contain blends of herbs adulterated with synthetic cannabinoid chemicals developed by researchers exploring the receptor-ligand binding of the endocannabinoid system.

Clinical effects, generally psychiatric, include paranoia, anxiety, agitation, delusions, and psychosis. There are also reports of patients who arrive with sympathomimetic toxicity, some of whom develop bradycardia and hypotension, and some who progress to acute renal failure, seizures, and death. Detection of these products is difficult as they do not react on EMIT testing for THC metabolites and require either gas chromatography-mass spectrometry or liquid chromatography with tandem mass spectrometry.^45–48

Clinicians may be encountering more cannabis users than before, and may be encountering users with complications hitherto unseen. Several trends may explain this phenomenon: the legal status of cannabis is changing, cannabis today is more potent than in the past, and enthusiasts are conjuring new ways to enjoy this substance.

This article discusses the history, pharmacology, and potential complications of cannabis use.

A LONG AND TANGLED HISTORY

Cannabis is a broad term that refers to the cannabis plant and its preparations, such as marijuana and hashish, as well as to a family of more than 60 bioactive substances called cannabinoids. It is the most commonly used illegal drug in the world, with an estimated 160 million users. Each year, about 2.4 million people in the United States use it for the first time.^1,2

Cannabis has been used throughout the world for recreational and spiritual purposes for nearly 5,000 years, beginning with the fabled Celestial Emperors of China. The tangled history of cannabis in America began in the 17th century, when farmers were required by law to grow it as a fiber crop. It later found its way into the US Pharmacopeia for a wide range of indications. During the long prelude to Prohibition in the latter half of the 19th century, the US government became increasingly suspicious of mind-altering substances and began restricting its prescription in 1934, culminating in its designation by the US Food and Drug Administration as a schedule I controlled substance in 1970.

Investigation into the potential medical uses for the different chemicals within cannabis is ongoing, as is debate over its changing legality and usefulness to society. The apparent cognitive dissonance surrounding the use and advocacy of medical marijuana is beyond the scope of this review,³ which will instead restrict itself to what is known of the cannabinoids and to the recreational use of cannabis.

THC IS THE PRINCIPAL PSYCHOACTIVE MOLECULE

Delta-9 tetrahydrocannabinol (THC), first isolated in 1964, was identified as the principal psychoactive constituent of cannabis in 2002.⁴

Two G-protein–linked cannabinoid receptors cloned in the 1990s—CB1 and CB2—were found to be a part of a system of endocannabinoid receptors present throughout the body, from the brain to the immune system to the vas deferens.⁵ Both receptors inhibit cellular excitation by activating inwardly rectifying potassium channels. These receptors are mostly absent in the brainstem, which may explain why cannabis use rarely causes life-threatening autonomic dysfunction. Although the intoxicating effects of marijuana are mediated by CB1 receptors, the specific mechanisms underlying the cannabis “high” are unclear.⁶

CANNABINOIDS ARE LIPID-SOLUBLE

The rate of absorption of cannabinoids depends on the route of administration and the type of cannabis product used. When cannabis products are smoked, up to 35% of THC is available, and the average time to peak serum concentration is 8 minutes.⁷ The peak concentration depends on the dose.

On the other hand, when cannabis products (eg, nabilone, dronabinol) are ingested, absorption is unpredictable because THC is unstable in gastric acid and undergoes first-pass metabolism in the liver, which reduces the drug’s bioavailability. Up to 20% of an ingested dose of THC is absorbed, and the time to peak serum concentration averages between 2 and 4 hours. Consequently, many users prefer to smoke cannabis as a means to control the desired effects.

Cannabinoids are lipid-soluble. They accumulate in fatty tissue in a biphasic pattern, initially moving into highly vascularized tissue such as the liver before accumulating in less well-vascularized tissue such as fat. They are then slowly released from fatty tissue as the fat turns over. THC itself has a volume of distribution of about 2.5 to 3.5 L/kg. It crosses the placenta and enters breast milk.⁸

THC is metabolized by the cytochrome P450 system, primarily by the enzymes CYP­2C9 and CYP3A4. Its primary metabolite, 11-hydroxy-delta-9 THC, is also active, but subsequent metabolism produces many other inactive metabolites. THC is eliminated in feces and urine, and its half-life ranges from 2 to nearly 60 hours.⁸

A LITTLE ABOUT PLANTS AND STREET NAMES

The plant from which THC and nearly a hundred other chemicals, including cannabinoids, are derived has been called many things over the years:

Hemp is a tall fibrous plant grown for rope and fabric that was used as legal tender in early America. In the mid-19th century, there were over 16 million acres of hemp plantations. Hemp contains very low THC concentrations.

Cannabis is an annual flowering herb that is predominantly diecious (ie, there are male and female plants). After a centuries-long debate among taxonomists, the two principal species are considered to be C sativa and C indica, although today many cannabis cultivars are grown by a great number of breeding enthusiasts.

THC levels in marijuana have increased from about 5% historically to over 30% in some samples today

Concentrations of THC vary widely among cannabis cultivars, ranging historically from around 5% to today’s highly selectively bred species containing more than 30%. Concentrations in seized cannabis have been measured as high as 37%, although the average is around 11%.⁹ This concentration is defined by the percent of THC per dried mass of plant material tested, usually via gas chromatography.

Hashish is a solid or resinous preparation of the trichomes, or glandular hairs, that grow on the cannabis plant, chiefly on its flowers. Various methods to separate the trichomes from the rest of the plant result in a powder called kief that is then compressed into blocks or bricks. THC concentrations as high as 66% have been measured in nondomestic sources of hashish.⁹

Hash oil is a further purification, produced by using solvents to dissolve the resin and by filtering out remaining plant material. Evaporating the solvent produces hash oil, sometimes called butane hash oil or honey oil. This process has recently led to an increasing number of home explosions, as people attempt to make the product themselves but do not take suitable precautions when using flammable solvents such as butane. THC concentrations as high as 81% have been measured in nondomestic sources of hash oil.⁹

Other names for hash oil are dab, wax, and budder. Cannabis enthusiasts refer to the use of hash oil as dabbing, which involves heating a small amount (dab) of the product using a variety of paraphernalia and inhaling the vapor.

IT’S ALL ABOUT GETTING HIGH

One user’s high is another user’s acute toxic effect

For recreational users, the experience has always been about being intoxicated—getting high. The psychological effects range broadly from positive to negative and vary both within and between users, depending on the dose and route of administration. Additional factors that influence the psychological effects include the social and physical settings of drug use and even the user’s expectations. One user’s high is another user’s acute toxic effect.

Although subjective reports of the cannabis experience vary greatly, it typically begins with a feeling of dizziness or lightheadedness followed by a relaxed calm and a feeling of being somewhat “disconnected.” There is a quickening of the sense of humor, described by some as a fatuous euphoria; often there is silly giggling. Awareness of the senses and of music may be increased. Appetite increases, and time seems to pass quickly. Eventually, the user becomes drowsy and experiences decreased attention and difficulty maintaining a coherent conversation. Slowed reaction time and decreased psychomotor activity may also occur. The user may drift into daydreams and eventually fall asleep.

Common negative acute effects of getting high can include mild to severe anxiety and feeling tense or agitated. Clumsiness, headache, and confusion are also possible. Lingering effects the following day may include dry mouth, dry eyes, fatigue, slowed thinking, and slowed recall.⁶

ACUTE PHYSICAL EFFECTS

Acute physical effects of cannabis use include a rapid onset of increased airway conductance, decreased intraocular pressure, and conjunctival injection. A single cannabis cigarette can also induce cardiovascular effects including a dose-dependent increase in heart rate and blood pressure. Chronic users, however, can experience a decreased heart rate, lower blood pressure, and postural hypotension.

In a personal communication, colleagues in Colorado—where recreational use of cannabis was legalized in 2012—described a sharp increase (from virtually none) in the number of adults presenting to the emergency department with cannabis intoxication since 2012. Their patients experienced palpitations, light-headedness, and severe ataxia lasting as long as 12 hours, possibly reflecting the greater potency of current cannabis products. Most of these patients required only supportive care.

Acute effects of cannabis include increased airway conductance, decreased intraocular pressure, and conjunctival injection

Other acute adverse cardiovascular reactions that have been reported include atrial fibrillation, ventricular tachycardia, and a fivefold increased risk of myocardial infarction in the 60 minutes following cannabis use, which subsequently drops sharply to baseline levels.¹⁰ Investigations into the cardiovascular effects of cannabis are often complicated by concurrent use of other drugs such as tobacco or cocaine. Possible mechanisms of injury include alterations in coronary microcirculation or slowed coronary flow. In fact, one author found that cannabis users with a history of myocardial infarction had a risk of death 4.2 times higher than users with no history of myocardial infarction.^11,12

In children, acute toxicity has been reported from a variety of exposures to cannabis and hashish, including a report of an increase in pediatric cannabis exposures following the changes in Colorado state laws.¹³ Most of these patients had altered mental status ranging from drowsiness to coma; one report describes a child who experienced a first-time seizure. These patients unfortunately often underwent extensive evaluations such as brain imaging and lumbar puncture, and mechanical ventilation to protect the airway. Earlier consideration of cannabis exposure in these patients might have limited unnecessary testing. Supportive care is usually all that is needed, and most of these patients fully recover.^13–17

CHRONIC EFFECTS

Cannabinoids cause a variety of adverse effects, but the ultimate risk these changes pose to human health has been difficult to calculate. Long-term studies are confounded by possible inaccuracies of patient self-reporting of cannabis use, poor control of covariates, and disparate methodologies.

For more than a century, cannabis use has been reported to cause both acute psychotic symptoms and persistent psychotic disorders.¹⁸ But the strength of this relationship is modest. Cannabis is more likely a component cause that, in addition to other factors (eg, specific genetic polymorphisms), contributes to the risk of schizophrenia. Individuals with prodromal symptoms and those who have experienced discrete episodes of psychosis related to cannabis use should be discouraged from using cannabis and cannabinoids.^19–21

Mounting evidence implicates chronic cannabis use as a cause of long-term medical problems

Mounting evidence implicates chronic cannabis use as a cause of long-term medical problems including chronic bronchitis,²² elevated rates of myocardial infarction and dysrhythmias,¹¹ bone loss,²³ and cancers at eight different sites including the lung, head, and neck.²⁴ In view of these chronic effects, healthcare providers should caution their patients about cannabis use, as we do about other drugs such as tobacco.

WITHDRAWAL SYNDROME RECOGNIZED

Until recently, neither clinicians nor users recognized a withdrawal syndrome associated with chronic use of cannabis, probably because this syndrome is not as severe as withdrawal from other controlled substances such as opioids or sedative-hypnotics. A number of studies, however, have reported subtle cannabis withdrawal symptoms that are similar to those associated with tobacco withdrawal.

As such, the fifth and latest edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5)²⁵ characterized withdrawal from cannabis use in 2013. The DSM-5 criteria require cessation of heavy or prolonged use of cannabis (ie, daily or almost daily over a period of at least a few months) and three or more of the following withdrawal symptoms:

• Irritability and anger
• Nervousness
• Sleep difficulty or insomnia
• Decreased appetite or weight loss
• Restlessness
• Depressed mood
• Physical symptoms causing discomfort.

Medical treatment of cannabis withdrawal has included a range of antidepressants, mood stabilizers, and alpha-2-adrenergic agonists, all of which have limited success.²⁶ Symptoms of cannabis withdrawal tend to be most intense soon after cessation and decline over the next few weeks.²⁷

CANNABINOID HYPEREMESIS SYNDROME

First reported in 2004,²⁸ cannabinoid hyperemesis syndrome is a recurrent disorder, the pathophysiology of which is poorly understood. It has three phases.

The first phase is a prodrome that may last months or years and is characterized by morning nausea, fear of vomiting, and abdominal discomfort. During this phase, the patient maintains normal eating patterns and may well increase his or her cannabis use due to its well-known antiemetic effects.

The second phase is the hyperemetic phase, characterized by intense, incapacitating emesis with episodes of vomiting throughout the day. These symptoms can be relieved only with frequent hot baths, a feature that distinguishes cannabinoid hyperemesis syndrome from other vomiting syndromes. Hot-water bathing is reported to be a compulsive but learned behavior in which the patient learns that only hot water will provide relief. The extent of relief depends on the temperature of the water—the hotter, the better. Symptoms recur as the water cools.²⁸ Patients often present to the emergency department repeatedly with recurrent symptoms and may remain misdiagnosed or subjected to repeated extensive evaluation including laboratory testing and imaging, which are usually not revealing. If the patient has not been accurately diagnosed, there may be reported weight loss of at least 5 kg.

The third phase, recovery, may take several months to complete, possibly because of the prolonged terminal elimination time of cannabinoids. Complete cessation of cannabis use, including synthetic cannabinoids, is usually necessary.²⁹

Diagnostic criteria for cannabinoid hyperemesis syndrome have been suggested, based on a retrospective case series that included 98 patients.³⁰ The most common features of these affected patients were:

• Severe cyclical vomiting, predominantly in the morning
• Resolution of symptoms with cessation of cannabis use
• Symptomatic relief with hot showers or baths
• Abdominal pain
• At least weekly use of cannabis.

Interestingly, long-term cannabis use has been cited as a critical identifying feature of these patients, with the duration of cannabis use ranging from 10 to 16 years.^31,32 Other reports show greater variability in duration of cannabis use before the onset of cannabinoid hyperemesis syndrome. In the large study noted above,³⁰ 32% of users reported their duration of cannabis use to be less than 1 year, rendering this criterion less useful.

How can cannabis both cause and prevent vomiting?

The body controls nausea and vomiting via complex circuitry in the brain and gut that involves many neurotransmitters (eg, dopamine, serotonin, substance P) that interact with receptors such as CB1, 5-HT_1–4, alpha adrenergic receptors, and mu receptors. Interestingly, cannabis use has antiemetic properties mediated by CB1 with a still unclear additional role of CB2 receptors. Data point to the existence of an underlying antiemetic tone mediated by the endocannabinoid system.

Unfortunately, the mechanism by which cannabinoid hyperemesis syndrome occurs is unknown and represents a paradoxical effect against the otherwise antiemetic effects of cannabis. Several theories have been proposed, including delayed gastric emptying, although only a third of patients demonstrated this on scintigraphy in one study.³⁰ Other theories include disturbance of the hypothalamic-pituitary axis, a buildup of highly lipophilic THC in the brain, and a down-regulation of cannabinoid receptors that results from chronic exposure.³⁰ Given that this syndrome has been recognized only relatively recently, one author has suggested the cause may be recent horticultural developments.⁵

Treating cannabinoid hyperemesis syndrome is difficult

Treatment of cannabinoid hyperemesis syndrome is notoriously difficult, with many authors reporting resistance to the usual first-line antiemetic drugs. Generally, treatment should include hydration and acid-suppression therapy because endoscopic evaluation of several patients has revealed varying degrees of esophagitis and gastritis.²⁹

Antiemetic therapy should target receptors known to mediate nausea and vomiting. In some cases, antiemetic drugs are more effective when used in combination. Agents include the serotonergic receptor antagonists ondansetron and granisetron, the dopamine antagonists prochlorperazine and metoclopramide, and even haloperidol.^33,34 Benzodiazepines may be effective by causing sedation, anxiolysis, and depression of the vomiting center.^34,35 Two antihistamines—dimenhydrinate and diphenhydramine—have antiemetic effects, perhaps by inhibiting acetylcholine.³⁴

Aprepitant is a neurokinin-1 antagonist that inhibits the action of substance P. When combined with a corticosteroid and a serotonin antagonist, it relieves nausea and vomiting in chemotherapy patients.^34,36

Corticosteroids such as dexamethasone are potent antiemetics thought to inhibit prostaglandin synthesis.³⁴

Capsaicin cream applied to the abdomen has also been reported to relieve symptoms, possibly through an interaction between the TRPv1 receptor and the endocannabinoid system.^37,38

DIAGNOSTIC TESTING

Cannabinoids are detectable in plasma and urine, with urine testing being more common.

Common laboratory methods include the enzyme-multiplied immunoassay technique (EMIT) and radioimmunoassay. Gas chromatography coupled with mass spectrometry is the most specific assay; it is used for confirmation and is the reference method.

EMIT is a qualitative urine test that detects 9-carboxy-THC as well as other THC metabolites. These urine tests detect all metabolites, and the result is reported as positive if the total concentration is greater than or equal to a prespecified threshold level, such as 20 ng/mL or 50 ng/mL. A positive test does not denote intoxication, nor does the test identify the source of THC (eg, cannabis, dronabinol, butane hash oil). EMIT does not detect nabilone. The National Institute on Drug Abuse guidelines for urine testing specify a test threshold concentration of 50 ng/mL for screening and 15 ng/mL for confirmation.

Many factors affect the detection of THC metabolites and their presence and duration in urine: dose, duration of use, route of exposure, hydration status, urine volume and concentration, and urine pH. THC metabolites have been detected in urine using gas chromatography-mass spectrometry for up to 7 days after smoking one marijuana cigarette.⁷ Chronic users have also been reported to have positive urine EMIT tests for up to 46 days after cannabis cessation.³⁹ Detection may be further complicated in chronic users: in one study, users produced both negative and positive specimens over 24 days, suggesting that diet and exercise may influence clearance.⁴⁰ Also, many factors are known to produce false-positive and false-negative results for these immunoassays (Table 1).^39,41

In the United States, penalties for driving under the influence of cannabis vary from state to state, and laws specify plasma testing for quantitative analysis. Some states use a threshold of 5 ng/mL in plasma to imply driving under the influence, whereas others use any detectable amount. Currently, there are no generally accepted guidelines for storage and testing of blood samples, despite the known instability of analytes.⁴²

Saliva, hair, and sweat can also be used for cannabinoid testing. Saliva is easy to collect, can be tested for metabolites to rule out passive cannabis exposure, and can be positive for up to 1 day after exposure. Calculating a blood or plasma concentration from a saliva sample is not possible, however.

Hair testing can also rule out passive exposure, but THC binds very little to melanin, resulting in very low concentrations requiring sensitive tests, such as gas chromatography with tandem mass spectrometry.

Only one device is commercially available for sweat testing; further work is needed to elucidate sweat excretion pharmacokinetics and the limitations of the collection devices.⁴³

CLINICAL MANAGEMENT IS GENERALLY SUPPORTIVE

Historically, clinical toxicity from recreational cannabis use is rarely serious or severe and generally responds to supportive care. Reports of cannabis exposure to poison centers are one-tenth of those reported for ethanol exposures annually.⁴⁴ Gastrointestinal decontamination with activated charcoal is not recommended, even for orally administered cannabis, since the risks outweigh the expected benefits. Agitation or anxiety may be treated with benzodiazepines as needed. There is no antidote for cannabis toxicity. The ever-increasing availability of high-concentration THC preparations may prompt more aggressive supportive measures in the future.

SYNTHETIC MARIJUANA ALTERNATIVES

Available since the early 2000s, herbal marijuana alternatives are legally sold as incense or potpourri and are often labeled “not for human consumption.” They are known by such brand names as K2 and Spice and contain blends of herbs adulterated with synthetic cannabinoid chemicals developed by researchers exploring the receptor-ligand binding of the endocannabinoid system.

Clinical effects, generally psychiatric, include paranoia, anxiety, agitation, delusions, and psychosis. There are also reports of patients who arrive with sympathomimetic toxicity, some of whom develop bradycardia and hypotension, and some who progress to acute renal failure, seizures, and death. Detection of these products is difficult as they do not react on EMIT testing for THC metabolites and require either gas chromatography-mass spectrometry or liquid chromatography with tandem mass spectrometry.^45–48

Clinicians may be encountering more cannabis users than before, and may be encountering users with complications hitherto unseen. Several trends may explain this phenomenon: the legal status of cannabis is changing, cannabis today is more potent than in the past, and enthusiasts are conjuring new ways to enjoy this substance.

This article discusses the history, pharmacology, and potential complications of cannabis use.

A LONG AND TANGLED HISTORY

Cannabis is a broad term that refers to the cannabis plant and its preparations, such as marijuana and hashish, as well as to a family of more than 60 bioactive substances called cannabinoids. It is the most commonly used illegal drug in the world, with an estimated 160 million users. Each year, about 2.4 million people in the United States use it for the first time.^1,2

Cannabis has been used throughout the world for recreational and spiritual purposes for nearly 5,000 years, beginning with the fabled Celestial Emperors of China. The tangled history of cannabis in America began in the 17th century, when farmers were required by law to grow it as a fiber crop. It later found its way into the US Pharmacopeia for a wide range of indications. During the long prelude to Prohibition in the latter half of the 19th century, the US government became increasingly suspicious of mind-altering substances and began restricting its prescription in 1934, culminating in its designation by the US Food and Drug Administration as a schedule I controlled substance in 1970.

Investigation into the potential medical uses for the different chemicals within cannabis is ongoing, as is debate over its changing legality and usefulness to society. The apparent cognitive dissonance surrounding the use and advocacy of medical marijuana is beyond the scope of this review,³ which will instead restrict itself to what is known of the cannabinoids and to the recreational use of cannabis.

THC IS THE PRINCIPAL PSYCHOACTIVE MOLECULE

Delta-9 tetrahydrocannabinol (THC), first isolated in 1964, was identified as the principal psychoactive constituent of cannabis in 2002.⁴

Two G-protein–linked cannabinoid receptors cloned in the 1990s—CB1 and CB2—were found to be a part of a system of endocannabinoid receptors present throughout the body, from the brain to the immune system to the vas deferens.⁵ Both receptors inhibit cellular excitation by activating inwardly rectifying potassium channels. These receptors are mostly absent in the brainstem, which may explain why cannabis use rarely causes life-threatening autonomic dysfunction. Although the intoxicating effects of marijuana are mediated by CB1 receptors, the specific mechanisms underlying the cannabis “high” are unclear.⁶

CANNABINOIDS ARE LIPID-SOLUBLE

The rate of absorption of cannabinoids depends on the route of administration and the type of cannabis product used. When cannabis products are smoked, up to 35% of THC is available, and the average time to peak serum concentration is 8 minutes.⁷ The peak concentration depends on the dose.

On the other hand, when cannabis products (eg, nabilone, dronabinol) are ingested, absorption is unpredictable because THC is unstable in gastric acid and undergoes first-pass metabolism in the liver, which reduces the drug’s bioavailability. Up to 20% of an ingested dose of THC is absorbed, and the time to peak serum concentration averages between 2 and 4 hours. Consequently, many users prefer to smoke cannabis as a means to control the desired effects.

Cannabinoids are lipid-soluble. They accumulate in fatty tissue in a biphasic pattern, initially moving into highly vascularized tissue such as the liver before accumulating in less well-vascularized tissue such as fat. They are then slowly released from fatty tissue as the fat turns over. THC itself has a volume of distribution of about 2.5 to 3.5 L/kg. It crosses the placenta and enters breast milk.⁸

THC is metabolized by the cytochrome P450 system, primarily by the enzymes CYP­2C9 and CYP3A4. Its primary metabolite, 11-hydroxy-delta-9 THC, is also active, but subsequent metabolism produces many other inactive metabolites. THC is eliminated in feces and urine, and its half-life ranges from 2 to nearly 60 hours.⁸

A LITTLE ABOUT PLANTS AND STREET NAMES

The plant from which THC and nearly a hundred other chemicals, including cannabinoids, are derived has been called many things over the years:

Hemp is a tall fibrous plant grown for rope and fabric that was used as legal tender in early America. In the mid-19th century, there were over 16 million acres of hemp plantations. Hemp contains very low THC concentrations.

Cannabis is an annual flowering herb that is predominantly diecious (ie, there are male and female plants). After a centuries-long debate among taxonomists, the two principal species are considered to be C sativa and C indica, although today many cannabis cultivars are grown by a great number of breeding enthusiasts.

THC levels in marijuana have increased from about 5% historically to over 30% in some samples today

Concentrations of THC vary widely among cannabis cultivars, ranging historically from around 5% to today’s highly selectively bred species containing more than 30%. Concentrations in seized cannabis have been measured as high as 37%, although the average is around 11%.⁹ This concentration is defined by the percent of THC per dried mass of plant material tested, usually via gas chromatography.

Hashish is a solid or resinous preparation of the trichomes, or glandular hairs, that grow on the cannabis plant, chiefly on its flowers. Various methods to separate the trichomes from the rest of the plant result in a powder called kief that is then compressed into blocks or bricks. THC concentrations as high as 66% have been measured in nondomestic sources of hashish.⁹

Hash oil is a further purification, produced by using solvents to dissolve the resin and by filtering out remaining plant material. Evaporating the solvent produces hash oil, sometimes called butane hash oil or honey oil. This process has recently led to an increasing number of home explosions, as people attempt to make the product themselves but do not take suitable precautions when using flammable solvents such as butane. THC concentrations as high as 81% have been measured in nondomestic sources of hash oil.⁹

Other names for hash oil are dab, wax, and budder. Cannabis enthusiasts refer to the use of hash oil as dabbing, which involves heating a small amount (dab) of the product using a variety of paraphernalia and inhaling the vapor.

IT’S ALL ABOUT GETTING HIGH

One user’s high is another user’s acute toxic effect

For recreational users, the experience has always been about being intoxicated—getting high. The psychological effects range broadly from positive to negative and vary both within and between users, depending on the dose and route of administration. Additional factors that influence the psychological effects include the social and physical settings of drug use and even the user’s expectations. One user’s high is another user’s acute toxic effect.

Although subjective reports of the cannabis experience vary greatly, it typically begins with a feeling of dizziness or lightheadedness followed by a relaxed calm and a feeling of being somewhat “disconnected.” There is a quickening of the sense of humor, described by some as a fatuous euphoria; often there is silly giggling. Awareness of the senses and of music may be increased. Appetite increases, and time seems to pass quickly. Eventually, the user becomes drowsy and experiences decreased attention and difficulty maintaining a coherent conversation. Slowed reaction time and decreased psychomotor activity may also occur. The user may drift into daydreams and eventually fall asleep.

Common negative acute effects of getting high can include mild to severe anxiety and feeling tense or agitated. Clumsiness, headache, and confusion are also possible. Lingering effects the following day may include dry mouth, dry eyes, fatigue, slowed thinking, and slowed recall.⁶

ACUTE PHYSICAL EFFECTS

Acute physical effects of cannabis use include a rapid onset of increased airway conductance, decreased intraocular pressure, and conjunctival injection. A single cannabis cigarette can also induce cardiovascular effects including a dose-dependent increase in heart rate and blood pressure. Chronic users, however, can experience a decreased heart rate, lower blood pressure, and postural hypotension.

In a personal communication, colleagues in Colorado—where recreational use of cannabis was legalized in 2012—described a sharp increase (from virtually none) in the number of adults presenting to the emergency department with cannabis intoxication since 2012. Their patients experienced palpitations, light-headedness, and severe ataxia lasting as long as 12 hours, possibly reflecting the greater potency of current cannabis products. Most of these patients required only supportive care.

Acute effects of cannabis include increased airway conductance, decreased intraocular pressure, and conjunctival injection

Other acute adverse cardiovascular reactions that have been reported include atrial fibrillation, ventricular tachycardia, and a fivefold increased risk of myocardial infarction in the 60 minutes following cannabis use, which subsequently drops sharply to baseline levels.¹⁰ Investigations into the cardiovascular effects of cannabis are often complicated by concurrent use of other drugs such as tobacco or cocaine. Possible mechanisms of injury include alterations in coronary microcirculation or slowed coronary flow. In fact, one author found that cannabis users with a history of myocardial infarction had a risk of death 4.2 times higher than users with no history of myocardial infarction.^11,12

In children, acute toxicity has been reported from a variety of exposures to cannabis and hashish, including a report of an increase in pediatric cannabis exposures following the changes in Colorado state laws.¹³ Most of these patients had altered mental status ranging from drowsiness to coma; one report describes a child who experienced a first-time seizure. These patients unfortunately often underwent extensive evaluations such as brain imaging and lumbar puncture, and mechanical ventilation to protect the airway. Earlier consideration of cannabis exposure in these patients might have limited unnecessary testing. Supportive care is usually all that is needed, and most of these patients fully recover.^13–17

CHRONIC EFFECTS

Cannabinoids cause a variety of adverse effects, but the ultimate risk these changes pose to human health has been difficult to calculate. Long-term studies are confounded by possible inaccuracies of patient self-reporting of cannabis use, poor control of covariates, and disparate methodologies.

For more than a century, cannabis use has been reported to cause both acute psychotic symptoms and persistent psychotic disorders.¹⁸ But the strength of this relationship is modest. Cannabis is more likely a component cause that, in addition to other factors (eg, specific genetic polymorphisms), contributes to the risk of schizophrenia. Individuals with prodromal symptoms and those who have experienced discrete episodes of psychosis related to cannabis use should be discouraged from using cannabis and cannabinoids.^19–21

Mounting evidence implicates chronic cannabis use as a cause of long-term medical problems

Mounting evidence implicates chronic cannabis use as a cause of long-term medical problems including chronic bronchitis,²² elevated rates of myocardial infarction and dysrhythmias,¹¹ bone loss,²³ and cancers at eight different sites including the lung, head, and neck.²⁴ In view of these chronic effects, healthcare providers should caution their patients about cannabis use, as we do about other drugs such as tobacco.

WITHDRAWAL SYNDROME RECOGNIZED

Until recently, neither clinicians nor users recognized a withdrawal syndrome associated with chronic use of cannabis, probably because this syndrome is not as severe as withdrawal from other controlled substances such as opioids or sedative-hypnotics. A number of studies, however, have reported subtle cannabis withdrawal symptoms that are similar to those associated with tobacco withdrawal.

As such, the fifth and latest edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5)²⁵ characterized withdrawal from cannabis use in 2013. The DSM-5 criteria require cessation of heavy or prolonged use of cannabis (ie, daily or almost daily over a period of at least a few months) and three or more of the following withdrawal symptoms:

• Irritability and anger
• Nervousness
• Sleep difficulty or insomnia
• Decreased appetite or weight loss
• Restlessness
• Depressed mood
• Physical symptoms causing discomfort.

Medical treatment of cannabis withdrawal has included a range of antidepressants, mood stabilizers, and alpha-2-adrenergic agonists, all of which have limited success.²⁶ Symptoms of cannabis withdrawal tend to be most intense soon after cessation and decline over the next few weeks.²⁷

CANNABINOID HYPEREMESIS SYNDROME

First reported in 2004,²⁸ cannabinoid hyperemesis syndrome is a recurrent disorder, the pathophysiology of which is poorly understood. It has three phases.

The first phase is a prodrome that may last months or years and is characterized by morning nausea, fear of vomiting, and abdominal discomfort. During this phase, the patient maintains normal eating patterns and may well increase his or her cannabis use due to its well-known antiemetic effects.

The second phase is the hyperemetic phase, characterized by intense, incapacitating emesis with episodes of vomiting throughout the day. These symptoms can be relieved only with frequent hot baths, a feature that distinguishes cannabinoid hyperemesis syndrome from other vomiting syndromes. Hot-water bathing is reported to be a compulsive but learned behavior in which the patient learns that only hot water will provide relief. The extent of relief depends on the temperature of the water—the hotter, the better. Symptoms recur as the water cools.²⁸ Patients often present to the emergency department repeatedly with recurrent symptoms and may remain misdiagnosed or subjected to repeated extensive evaluation including laboratory testing and imaging, which are usually not revealing. If the patient has not been accurately diagnosed, there may be reported weight loss of at least 5 kg.

The third phase, recovery, may take several months to complete, possibly because of the prolonged terminal elimination time of cannabinoids. Complete cessation of cannabis use, including synthetic cannabinoids, is usually necessary.²⁹

Diagnostic criteria for cannabinoid hyperemesis syndrome have been suggested, based on a retrospective case series that included 98 patients.³⁰ The most common features of these affected patients were:

• Severe cyclical vomiting, predominantly in the morning
• Resolution of symptoms with cessation of cannabis use
• Symptomatic relief with hot showers or baths
• Abdominal pain
• At least weekly use of cannabis.

Interestingly, long-term cannabis use has been cited as a critical identifying feature of these patients, with the duration of cannabis use ranging from 10 to 16 years.^31,32 Other reports show greater variability in duration of cannabis use before the onset of cannabinoid hyperemesis syndrome. In the large study noted above,³⁰ 32% of users reported their duration of cannabis use to be less than 1 year, rendering this criterion less useful.

How can cannabis both cause and prevent vomiting?

The body controls nausea and vomiting via complex circuitry in the brain and gut that involves many neurotransmitters (eg, dopamine, serotonin, substance P) that interact with receptors such as CB1, 5-HT_1–4, alpha adrenergic receptors, and mu receptors. Interestingly, cannabis use has antiemetic properties mediated by CB1 with a still unclear additional role of CB2 receptors. Data point to the existence of an underlying antiemetic tone mediated by the endocannabinoid system.

Unfortunately, the mechanism by which cannabinoid hyperemesis syndrome occurs is unknown and represents a paradoxical effect against the otherwise antiemetic effects of cannabis. Several theories have been proposed, including delayed gastric emptying, although only a third of patients demonstrated this on scintigraphy in one study.³⁰ Other theories include disturbance of the hypothalamic-pituitary axis, a buildup of highly lipophilic THC in the brain, and a down-regulation of cannabinoid receptors that results from chronic exposure.³⁰ Given that this syndrome has been recognized only relatively recently, one author has suggested the cause may be recent horticultural developments.⁵

Treating cannabinoid hyperemesis syndrome is difficult

Treatment of cannabinoid hyperemesis syndrome is notoriously difficult, with many authors reporting resistance to the usual first-line antiemetic drugs. Generally, treatment should include hydration and acid-suppression therapy because endoscopic evaluation of several patients has revealed varying degrees of esophagitis and gastritis.²⁹

Antiemetic therapy should target receptors known to mediate nausea and vomiting. In some cases, antiemetic drugs are more effective when used in combination. Agents include the serotonergic receptor antagonists ondansetron and granisetron, the dopamine antagonists prochlorperazine and metoclopramide, and even haloperidol.^33,34 Benzodiazepines may be effective by causing sedation, anxiolysis, and depression of the vomiting center.^34,35 Two antihistamines—dimenhydrinate and diphenhydramine—have antiemetic effects, perhaps by inhibiting acetylcholine.³⁴

Aprepitant is a neurokinin-1 antagonist that inhibits the action of substance P. When combined with a corticosteroid and a serotonin antagonist, it relieves nausea and vomiting in chemotherapy patients.^34,36

Corticosteroids such as dexamethasone are potent antiemetics thought to inhibit prostaglandin synthesis.³⁴

Capsaicin cream applied to the abdomen has also been reported to relieve symptoms, possibly through an interaction between the TRPv1 receptor and the endocannabinoid system.^37,38

DIAGNOSTIC TESTING

Cannabinoids are detectable in plasma and urine, with urine testing being more common.

Common laboratory methods include the enzyme-multiplied immunoassay technique (EMIT) and radioimmunoassay. Gas chromatography coupled with mass spectrometry is the most specific assay; it is used for confirmation and is the reference method.

EMIT is a qualitative urine test that detects 9-carboxy-THC as well as other THC metabolites. These urine tests detect all metabolites, and the result is reported as positive if the total concentration is greater than or equal to a prespecified threshold level, such as 20 ng/mL or 50 ng/mL. A positive test does not denote intoxication, nor does the test identify the source of THC (eg, cannabis, dronabinol, butane hash oil). EMIT does not detect nabilone. The National Institute on Drug Abuse guidelines for urine testing specify a test threshold concentration of 50 ng/mL for screening and 15 ng/mL for confirmation.

Many factors affect the detection of THC metabolites and their presence and duration in urine: dose, duration of use, route of exposure, hydration status, urine volume and concentration, and urine pH. THC metabolites have been detected in urine using gas chromatography-mass spectrometry for up to 7 days after smoking one marijuana cigarette.⁷ Chronic users have also been reported to have positive urine EMIT tests for up to 46 days after cannabis cessation.³⁹ Detection may be further complicated in chronic users: in one study, users produced both negative and positive specimens over 24 days, suggesting that diet and exercise may influence clearance.⁴⁰ Also, many factors are known to produce false-positive and false-negative results for these immunoassays (Table 1).^39,41

In the United States, penalties for driving under the influence of cannabis vary from state to state, and laws specify plasma testing for quantitative analysis. Some states use a threshold of 5 ng/mL in plasma to imply driving under the influence, whereas others use any detectable amount. Currently, there are no generally accepted guidelines for storage and testing of blood samples, despite the known instability of analytes.⁴²

Saliva, hair, and sweat can also be used for cannabinoid testing. Saliva is easy to collect, can be tested for metabolites to rule out passive cannabis exposure, and can be positive for up to 1 day after exposure. Calculating a blood or plasma concentration from a saliva sample is not possible, however.

Hair testing can also rule out passive exposure, but THC binds very little to melanin, resulting in very low concentrations requiring sensitive tests, such as gas chromatography with tandem mass spectrometry.

Only one device is commercially available for sweat testing; further work is needed to elucidate sweat excretion pharmacokinetics and the limitations of the collection devices.⁴³

CLINICAL MANAGEMENT IS GENERALLY SUPPORTIVE

Historically, clinical toxicity from recreational cannabis use is rarely serious or severe and generally responds to supportive care. Reports of cannabis exposure to poison centers are one-tenth of those reported for ethanol exposures annually.⁴⁴ Gastrointestinal decontamination with activated charcoal is not recommended, even for orally administered cannabis, since the risks outweigh the expected benefits. Agitation or anxiety may be treated with benzodiazepines as needed. There is no antidote for cannabis toxicity. The ever-increasing availability of high-concentration THC preparations may prompt more aggressive supportive measures in the future.

SYNTHETIC MARIJUANA ALTERNATIVES

Available since the early 2000s, herbal marijuana alternatives are legally sold as incense or potpourri and are often labeled “not for human consumption.” They are known by such brand names as K2 and Spice and contain blends of herbs adulterated with synthetic cannabinoid chemicals developed by researchers exploring the receptor-ligand binding of the endocannabinoid system.

Clinical effects, generally psychiatric, include paranoia, anxiety, agitation, delusions, and psychosis. There are also reports of patients who arrive with sympathomimetic toxicity, some of whom develop bradycardia and hypotension, and some who progress to acute renal failure, seizures, and death. Detection of these products is difficult as they do not react on EMIT testing for THC metabolites and require either gas chromatography-mass spectrometry or liquid chromatography with tandem mass spectrometry.^45–48

Women's health encompasses a broad range of issues unique to the female patient, with a scope that has expanded beyond reproductive health. Providers who care for women must develop cross-disciplinary competencies and understand the complex role of sex and gender on disease expression and treatment outcomes. Staying current with the literature in this rapidly changing field can be challenging for the busy clinician.

This article reviews recent advances in the treatment of depression in pregnancy, nonhormonal therapies for menopausal symptoms, and heart failure therapy in women, highlighting notable studies published in 2014 and early 2015.

TREATMENT OF DEPRESSION IN PREGNANCY

A 32-year-old woman with well-controlled but recurrent depression presents to the clinic for preconception counseling. Her depression has been successfully managed with a selective serotonin reuptake inhibitor (SSRI). She and her husband would like to try to conceive soon, but she is worried that continuing on her current SSRI may harm her baby. How should you advise her?

Concern for teratogenic effects of SSRIs

Depression is common during pregnancy: 11.8% to 13.5% of pregnant women report symptoms of depression,¹ and 7.5% of pregnant women take an antidepressant.²

SSRI use during pregnancy has drawn attention due to mixed reports of teratogenic effects

SSRI use during pregnancy has drawn attention because of mixed reports of teratogenic effects on the newborn, such as omphalocele, congenital heart defects, and craniosynostosis.³ Previous observational studies have specifically linked paroxetine to small but significant increases in right ventricular outflow tract obstruction^4,5 and have linked sertraline to ventricular septal defects.⁶

However, reports of associations of congenital malformations and SSRI use in pregnancy in observational studies have been questioned, with concern that these studies had low statistical power, self-reported data leading to recall bias, and limited assessment for confounding factors.^3,7

Recent studies refute risk of cardiac malformations

Several newer studies have been published that further examine the association between SSRI use in pregnancy and congenital heart defects, and their findings suggest that once adjusted for confounding variables, SSRI use in pregnancy may not be associated with cardiac malformations.

Huybrechts et al,⁸ in a large study published in 2014, extracted data on 950,000 pregnant women from the Medicaid database over a 7-year period and examined it for SSRI use during the first 90 days of pregnancy. Though SSRI use was associated with cardiac malformations when unadjusted for confounding variables (unadjusted relative risk 1.25, 95% confidence interval [CI] 1.13–1.38), once the cohort was restricted to women with a diagnosis of only depression and was adjusted based on propensity scoring, the association was no longer statistically significant (adjusted relative risk 1.06, 95% CI 0.93–1.22).

Additionally, there was no association between sertraline and ventricular septal defects (63 cases in 14,040 women exposed to sertraline, adjusted relative risk 1.04, 95% CI 0.76–1.41), or between paroxetine and right ventricular outflow tract obstruction (93 cases in 11,126 women exposed to paroxetine, adjusted relative risk 1.07, 95% CI 0.59–1.93).⁸

Furu et al⁷ conducted a sibling-matched case-control comparison published in 2015, in which more than 2 million live births from five Nordic countries were examined in the full cohort study and 2,288 births in the sibling-matched case-control cohort. SSRI or venlafaxine use in the first 90 days of pregnancy was examined. There was a slightly higher rate of cardiac defects in infants born to SSRI or venlafaxine recipients in the cohort study (adjusted odds ratio 1.15, 95% CI 1.05–1.26). However, in the sibling-controlled analyses, neither an SSRI nor venlafaxine was associated with heart defects (adjusted odds ratio 0.92, 95% CI 0.72–1.17), leading the authors to conclude that there might be familial factors or other lifestyle factors that were not taken into consideration and that could have confounded the cohort results.

Bérard et al⁹ examined antidepressant use in the first trimester of pregnancy in a cohort of women in Canada and concluded that sertraline was associated with congenital atrial and ventricular defects (risk ratio 1.34; 95% CI 1.02–1.76).⁹ However, this association should be interpreted with caution, as the Canadian cohort was notably smaller than those in other studies we have discussed, with only 18,493 pregnancies in the total cohort, and this conclusion was drawn from 9 cases of ventricular or atrial septal defects in babies of 366 women exposed to sertraline.

Although at first glance SSRIs may appear to be associated with congenital heart defects, these recent studies are reassuring and suggest that the association may actually not be significant. As with any statistical analysis, thoughtful study design, adequate statistical power, and adjustment for confounding factors must be considered before drawing conclusions.

SSRIs, offspring psychiatric outcomes, and miscarriage rates

Clements et al¹⁰ studied a cohort extracted from Partners Healthcare consisting of newborns with autism spectrum disorder, newborns with attention-deficit hyperactivity disorder (ADHD), and healthy matched controls and found that SSRI use during pregnancy was not associated with offspring autism spectrum disorder (adjusted odds ratio 1.10, 95% CI 0.7–1.70). However, they did find an increased risk of ADHD with SSRI use during pregnancy (adjusted odds ratio 1.81, 95% CI 1.22–2.70).

Andersen et al¹¹ examined more than 1 million pregnancies in Denmark and found no difference in risk of miscarriage between women who used an SSRI during pregnancy (adjusted hazard ratio 1.27) and women who discontinued their SSRI at least 3 months before pregnancy (adjusted hazard ratio 1.24, P = .47). The authors concluded that because of the similar rate of miscarriage in both groups, there was no association between SSRI use and miscarriage, and that the small increased risk of miscarriage in both groups could have been attributable to a confounding factor that was not measured.

Should our patient continue her SSRI through pregnancy?

Our patient has recurrent depression, and her risk of relapse with antidepressant cessation is high. Though previous, less well-done studies suggested a small risk of congenital heart defects, recent larger high-quality studies provide significant reassurance that SSRI use in pregnancy is not strongly associated with cardiac malformations. Recent studies also show no association with miscarriage or autism spectrum disorder, though there may be risk of offspring ADHD.

She can be counseled that she may continue on her SSRI during pregnancy and can be reassured that the risk to her baby is small compared with her risk of recurrent or postpartum depression.

NONHORMONAL TREATMENT FOR VASOMOTOR SYMPTOMS OF MENOPAUSE

You see a patient who is struggling with symptoms of menopause. She tells you she has terrible hot flashes day and night, and she would like to try drug therapy. She does not want hormone replacement therapy because she is worried about the risk of adverse events. Are there safe and effective nonhormonal pharmacologic treatments for her vasomotor symptoms?

Paroxetine 7.5 mg is approved for vasomotor symptoms of menopause

As many as 75% of menopausal women in the United States experience vasomotor symptoms related to menopause, or hot flashes and night sweats.¹² These symptoms can disrupt sleep and negatively affect quality of life. Though previously thought to occur during a short and self-limited time period, a recently published large observational study reported the median duration of vasomotor symptoms was 7.4 years, and in African American women in the cohort the median duration of vasomotor symptoms was 10.1 years—an entire decade of life.¹³

In 2013, the US Food and Drug Administration (FDA) approved paroxetine 7.5 mg daily for treating moderate to severe hot flashes associated with menopause. It is the only approved nonhormonal treatment for vasomotor symptoms; the only other approved treatments are estrogen therapy for women who have had a hysterectomy and combination estrogen-progesterone therapy for women who have not had a hysterectomy.

Further studies of paroxetine for menopausal symptoms

Since its approval, further studies have been published supporting the use of paroxetine 7.5 mg in treating symptoms of menopause. In addition to reducing hot flashes, this treatment also improves sleep disturbance in women with menopause.¹⁴

Pinkerton et al,¹⁴ in a pooled analysis of the data from the phase 3 clinical trials of paroxetine 7.5 mg per day, found that participants in groups assigned to paroxetine reported a 62% reduction in nighttime awakenings due to hot flashes compared with a 43% reduction in the placebo group (P < .001). Those who took paroxetine also reported a statistically significantly greater increase in duration of sleep than those who took placebo (37 minutes in the treatment group vs 27 minutes in the placebo group, P = .03).

Some patients are hesitant to take an SSRI because of concerns about adverse effects when used for psychiatric conditions. However, the dose of paroxetine that was studied and approved for vasomotor symptoms is lower than doses used for psychiatric indications and does not appear to be associated with these adverse effects.

Portman et al¹⁵ in 2014 examined the effect of paroxetine 7.5 mg vs placebo on weight gain and sexual function in women with vasomotor symptoms of menopause and found no significant increase in weight or decrease in sexual function at 24 weeks of use. Participants were weighed during study visits, and those in the paroxetine group gained on average 0.48% from baseline at 24 weeks, compared with 0.09% in the placebo group (P = .29).

Sexual dysfunction was assessed using the Arizona Sexual Experience Scale, which has been validated in psychiatric patients using antidepressants, and there was no significant difference in symptoms such as sex drive, sexual arousal, vaginal lubrication, or ability to achieve orgasm between the treatment group and placebo group.¹⁵

Paroxetine inhibits CYP2D6 and thus decreases tamoxifen activity

Of note, paroxetine is a potent inhibitor of the cytochrome P-450 CYP2D6 enzyme, and concurrent use of paroxetine with tamoxifen decreases tamoxifen activity.^12,16 Since women with a history of breast cancer who cannot use estrogen for hot flashes may be seeking nonhormonal treatment for their vasomotor symptoms, providers should perform careful medication reconciliation and be aware that concomitant use of paroxetine and tamoxifen is not recommended.

Other antidepressants show promise but are not approved for menopausal symptoms

In addition to paroxetine, other nonhormonal drugs have been studied for treating hot flashes, but they have been unable to secure FDA approval for this indication. One of these is the serotonin-norepinephrine reuptake inhibitor venlafaxine, and a 2014 study¹⁷ confirmed its efficacy in treating menopausal vasomotor symptoms.

Joffe et al¹⁷ performed a three-armed trial comparing venlafaxine 75 mg/day, estradiol 0.5 mg/day, and placebo and found that both of the active treatments were better than placebo at reducing vasomotor symptoms. Compared with each other, estradiol 0.5 mg/day reduced hot flash frequency by an additional 0.6 events per day compared with venlafaxine 75 mg/day (P = .09). Though this difference was statistically significant, the authors pointed out that the clinical significance of such a small absolute difference is questionable. Additionally, providers should be aware that venlafaxine has little or no effect on the metabolism of tamoxifen.¹⁶

Shams et al,¹⁸ in a meta-analysis published in 2014, concluded that SSRIs as a class are more effective than placebo in treating hot flashes, supporting their widespread off-label use for this purpose. Their analysis examined the results of 11 studies, which included more than 2,000 patients in total, and found that compared with placebo, SSRI use was associated with a significant decrease in hot flashes (mean difference –0.93 events per day, 95% CI –1.49 to –0.37). A mixed treatment comparison analysis was also performed to try to model performance of individual SSRIs based on the pooled data, and the model suggests that escitalopram may be the most efficacious SSRI at reducing hot flash severity.

These studies support the effectiveness of SSRIs¹⁸ and venlafaxine¹⁷ in reducing hot flashes compared with placebo, though providers should be aware that they are still not FDA-approved for this indication.

Nonhormonal therapy for our patient

We would recommend paroxetine 7.5 mg nightly to this patient, as it is an FDA-approved nonhormonal medication that has been shown to help patients with vasomotor symptoms of menopause as well as sleep disturbance, without sexual side effects or weight gain. If the patient cannot tolerate paroxetine, off-label use of another SSRI or venlafaxine is supported by the recent literature.

HEART DISEASE IN WOMEN: CARDIAC RESYNCHRONIZATION THERAPY

A 68-year-old woman with a history of nonis­chemic cardiomyopathy presents for routine follow-up in your office. Despite maximal medical therapy on a beta-blocker, an angiotensin II receptor blocker, and a diuretic, she has New York Heart Association (NYHA) class III symptoms. Her most recent studies showed an ejection fraction of 30% by echocardiography and left bundle-branch block on electrocardiography, with a QRS duration of 140 ms. She recently saw her cardiologist, who recommended cardiac resynchronization therapy, and she wants your opinion as to whether or not to proceed with this recommendation. How should you counsel her?

Which patients are candidates for cardiac resynchronization therapy?

Heart disease continues to be the number one cause of death in the United States for both men and women, and almost the same number of women and men die from heart disease every year.¹⁹ Though coronary artery disease accounts for most cases of cardiovascular disease in the United States, heart failure is a significant and growing contributor. Approximately 6.6 million adults had heart failure in 2010 in the United States, and an additional 3 million are projected to have heart failure by 2030.²⁰ The burden of disease on our health system is high, with about 1 million hospitalizations and more than 3 million outpatient office visits attributable to heart failure yearly.²⁰

Patients with heart failure may have symptoms of dyspnea, fatigue, orthopnea, and periph­eral edema; laboratory and radiologic findings of pulmonary edema, renal insufficiency, and hyponatremia; and electrocardiographic findings of atrial fibrillation or prolonged QRS.²¹ Intraventricular conduction delay (QRS duration > 120 ms) is associated with dyssynchronous ventricular contraction and impaired pump function and is present in almost one-third of patients who have advanced heart failure.²¹

Heart disease continues to be the number one cause of death in both men and women

Cardiac resynchronization therapy, or biventricular pacing, can improve symptoms and pump function and has been shown to decrease rates of hospitalization and death in these patients.²² According to the joint 2012 guidelines of the American College of Cardiology Foundation, American Heart Association, and Heart Rhythm Society,²² it is indicated for patients with an ejection fraction of 35% or less, left bundle-branch block with QRS duration of 150 ms or more, and NYHA class II to IV symptoms who are in sinus rhythm (class I recommendation, level of evidence A).

Studies of cardiac resynchronization therapy in women

Recently published studies have suggested that women may derive greater benefit than men from cardiac resynchronization therapy.

Zusterzeel et al²³ (2014) evaluated sex-specific data from the National Cardiovascular Data Registry, which contains data on all biventricular pacemaker and implantable cardioverter-defibrillator implantations from 80% of US hospitals.²³ Of the 21,152 patients who had left bundle-branch block and received cardiac resynchronization therapy, women derived greater benefit in terms of death than men did, with a 21% lower risk of death than men (adjusted hazard ratio 0.79, 95% CI 0.74–0.84, P < .001). This study was also notable in that 36% of the patients were women, whereas in most earlier studies of cardiac resynchronization therapy women accounted for only 22% to 30% of the study population.²²

Goldenberg et al²⁴ (2014) performed a follow-up analysis of the Multicenter Automatic Defibrillator Implantation Trial With Cardiac Resynchronization Therapy. Subgroup analysis showed that although both men and women had a lower risk of death if they received cardiac resynchronization therapy compared with an implantable cardioverter-defibrillator only, the magnitude of benefit may be greater for women (hazard ratio 0.48, 95% CI 0.25–0.91, P = .03) than for men (hazard ratio 0.69, 95% CI 0.50–0.95, P = .02).

In addition to deriving greater mortality benefit, women may actually benefit from cardiac resynchronization therapy at shorter QRS durations than what is currently recommended. Women have a shorter baseline QRS than men, and a smaller left ventricular cavity.²⁵ In an FDA meta-analysis published in August 2014, pooled data from more than 4,000 patients in three studies suggested that women with left bundle-branch block benefited from cardiac resynchronization therapy more than men with left bundle-branch block.²⁶ Neither men nor women with left bundle-branch block benefited from it if their QRS duration was less than 130 ms, and both sexes benefited from it if they had left bundle-branch block and a QRS duration longer than 150 ms. However, women who received it who had left bundle-branch block and a QRS duration of 130 to 149 ms had a significant 76% reduction in the primary composite outcome of a heart failure event or death (hazard ratio 0.24, 95% CI 0.11–0.53, P < .001), while men in the same group did not derive significant benefit (hazard ratio 0.85, 95% CI 0.60–1.21, P = .38).

Despite the increasing evidence that there are sex-specific differences in the benefit from cardiac resynchronization therapy, what we know is limited by the low rates of female enrollment in most of the studies of this treatment. In a systematic review published in 2015, Herz et al²⁷ found that 90% of the 183 studies they reviewed enrolled 35% women or less, and half of the studies enrolled less than 23% women. Furthermore, only 20 of the 183 studies reported baseline characteristics by sex.

Recognizing this lack of adequate data, in August 2014 the FDA issued an official guidance statement outlining its expectations regarding sex-specific patient recruitment, data analysis, and data reporting in future medical device studies.²⁸ Hopefully, with this support for sex-specific research by the FDA, future studies will be able to identify therapeutic outcome differences that may exist between male and female patients.

Should our patient receive cardiac resynchronization therapy?

Regarding our patient with heart failure, the above studies suggest she will likely have a lower risk of death if she receives cardiac resynchronization therapy, even though her QRS interval is shorter than 150 ms. Providers who are aware of the emerging data regarding sex differences and treatment response can be powerful advocates for their patients, even in subspecialty areas, as highlighted by this case. We recommend counseling this patient to proceed with cardiac resynchronization therapy.

Women's health encompasses a broad range of issues unique to the female patient, with a scope that has expanded beyond reproductive health. Providers who care for women must develop cross-disciplinary competencies and understand the complex role of sex and gender on disease expression and treatment outcomes. Staying current with the literature in this rapidly changing field can be challenging for the busy clinician.

This article reviews recent advances in the treatment of depression in pregnancy, nonhormonal therapies for menopausal symptoms, and heart failure therapy in women, highlighting notable studies published in 2014 and early 2015.

TREATMENT OF DEPRESSION IN PREGNANCY

A 32-year-old woman with well-controlled but recurrent depression presents to the clinic for preconception counseling. Her depression has been successfully managed with a selective serotonin reuptake inhibitor (SSRI). She and her husband would like to try to conceive soon, but she is worried that continuing on her current SSRI may harm her baby. How should you advise her?

Concern for teratogenic effects of SSRIs

Depression is common during pregnancy: 11.8% to 13.5% of pregnant women report symptoms of depression,¹ and 7.5% of pregnant women take an antidepressant.²

SSRI use during pregnancy has drawn attention due to mixed reports of teratogenic effects

SSRI use during pregnancy has drawn attention because of mixed reports of teratogenic effects on the newborn, such as omphalocele, congenital heart defects, and craniosynostosis.³ Previous observational studies have specifically linked paroxetine to small but significant increases in right ventricular outflow tract obstruction^4,5 and have linked sertraline to ventricular septal defects.⁶

However, reports of associations of congenital malformations and SSRI use in pregnancy in observational studies have been questioned, with concern that these studies had low statistical power, self-reported data leading to recall bias, and limited assessment for confounding factors.^3,7

Recent studies refute risk of cardiac malformations

Several newer studies have been published that further examine the association between SSRI use in pregnancy and congenital heart defects, and their findings suggest that once adjusted for confounding variables, SSRI use in pregnancy may not be associated with cardiac malformations.

Huybrechts et al,⁸ in a large study published in 2014, extracted data on 950,000 pregnant women from the Medicaid database over a 7-year period and examined it for SSRI use during the first 90 days of pregnancy. Though SSRI use was associated with cardiac malformations when unadjusted for confounding variables (unadjusted relative risk 1.25, 95% confidence interval [CI] 1.13–1.38), once the cohort was restricted to women with a diagnosis of only depression and was adjusted based on propensity scoring, the association was no longer statistically significant (adjusted relative risk 1.06, 95% CI 0.93–1.22).

Additionally, there was no association between sertraline and ventricular septal defects (63 cases in 14,040 women exposed to sertraline, adjusted relative risk 1.04, 95% CI 0.76–1.41), or between paroxetine and right ventricular outflow tract obstruction (93 cases in 11,126 women exposed to paroxetine, adjusted relative risk 1.07, 95% CI 0.59–1.93).⁸

Furu et al⁷ conducted a sibling-matched case-control comparison published in 2015, in which more than 2 million live births from five Nordic countries were examined in the full cohort study and 2,288 births in the sibling-matched case-control cohort. SSRI or venlafaxine use in the first 90 days of pregnancy was examined. There was a slightly higher rate of cardiac defects in infants born to SSRI or venlafaxine recipients in the cohort study (adjusted odds ratio 1.15, 95% CI 1.05–1.26). However, in the sibling-controlled analyses, neither an SSRI nor venlafaxine was associated with heart defects (adjusted odds ratio 0.92, 95% CI 0.72–1.17), leading the authors to conclude that there might be familial factors or other lifestyle factors that were not taken into consideration and that could have confounded the cohort results.

Bérard et al⁹ examined antidepressant use in the first trimester of pregnancy in a cohort of women in Canada and concluded that sertraline was associated with congenital atrial and ventricular defects (risk ratio 1.34; 95% CI 1.02–1.76).⁹ However, this association should be interpreted with caution, as the Canadian cohort was notably smaller than those in other studies we have discussed, with only 18,493 pregnancies in the total cohort, and this conclusion was drawn from 9 cases of ventricular or atrial septal defects in babies of 366 women exposed to sertraline.

Although at first glance SSRIs may appear to be associated with congenital heart defects, these recent studies are reassuring and suggest that the association may actually not be significant. As with any statistical analysis, thoughtful study design, adequate statistical power, and adjustment for confounding factors must be considered before drawing conclusions.

SSRIs, offspring psychiatric outcomes, and miscarriage rates

Clements et al¹⁰ studied a cohort extracted from Partners Healthcare consisting of newborns with autism spectrum disorder, newborns with attention-deficit hyperactivity disorder (ADHD), and healthy matched controls and found that SSRI use during pregnancy was not associated with offspring autism spectrum disorder (adjusted odds ratio 1.10, 95% CI 0.7–1.70). However, they did find an increased risk of ADHD with SSRI use during pregnancy (adjusted odds ratio 1.81, 95% CI 1.22–2.70).

Andersen et al¹¹ examined more than 1 million pregnancies in Denmark and found no difference in risk of miscarriage between women who used an SSRI during pregnancy (adjusted hazard ratio 1.27) and women who discontinued their SSRI at least 3 months before pregnancy (adjusted hazard ratio 1.24, P = .47). The authors concluded that because of the similar rate of miscarriage in both groups, there was no association between SSRI use and miscarriage, and that the small increased risk of miscarriage in both groups could have been attributable to a confounding factor that was not measured.

Should our patient continue her SSRI through pregnancy?

Our patient has recurrent depression, and her risk of relapse with antidepressant cessation is high. Though previous, less well-done studies suggested a small risk of congenital heart defects, recent larger high-quality studies provide significant reassurance that SSRI use in pregnancy is not strongly associated with cardiac malformations. Recent studies also show no association with miscarriage or autism spectrum disorder, though there may be risk of offspring ADHD.

She can be counseled that she may continue on her SSRI during pregnancy and can be reassured that the risk to her baby is small compared with her risk of recurrent or postpartum depression.

NONHORMONAL TREATMENT FOR VASOMOTOR SYMPTOMS OF MENOPAUSE

You see a patient who is struggling with symptoms of menopause. She tells you she has terrible hot flashes day and night, and she would like to try drug therapy. She does not want hormone replacement therapy because she is worried about the risk of adverse events. Are there safe and effective nonhormonal pharmacologic treatments for her vasomotor symptoms?

Paroxetine 7.5 mg is approved for vasomotor symptoms of menopause

As many as 75% of menopausal women in the United States experience vasomotor symptoms related to menopause, or hot flashes and night sweats.¹² These symptoms can disrupt sleep and negatively affect quality of life. Though previously thought to occur during a short and self-limited time period, a recently published large observational study reported the median duration of vasomotor symptoms was 7.4 years, and in African American women in the cohort the median duration of vasomotor symptoms was 10.1 years—an entire decade of life.¹³

In 2013, the US Food and Drug Administration (FDA) approved paroxetine 7.5 mg daily for treating moderate to severe hot flashes associated with menopause. It is the only approved nonhormonal treatment for vasomotor symptoms; the only other approved treatments are estrogen therapy for women who have had a hysterectomy and combination estrogen-progesterone therapy for women who have not had a hysterectomy.

Further studies of paroxetine for menopausal symptoms

Since its approval, further studies have been published supporting the use of paroxetine 7.5 mg in treating symptoms of menopause. In addition to reducing hot flashes, this treatment also improves sleep disturbance in women with menopause.¹⁴

Pinkerton et al,¹⁴ in a pooled analysis of the data from the phase 3 clinical trials of paroxetine 7.5 mg per day, found that participants in groups assigned to paroxetine reported a 62% reduction in nighttime awakenings due to hot flashes compared with a 43% reduction in the placebo group (P < .001). Those who took paroxetine also reported a statistically significantly greater increase in duration of sleep than those who took placebo (37 minutes in the treatment group vs 27 minutes in the placebo group, P = .03).

Some patients are hesitant to take an SSRI because of concerns about adverse effects when used for psychiatric conditions. However, the dose of paroxetine that was studied and approved for vasomotor symptoms is lower than doses used for psychiatric indications and does not appear to be associated with these adverse effects.

Portman et al¹⁵ in 2014 examined the effect of paroxetine 7.5 mg vs placebo on weight gain and sexual function in women with vasomotor symptoms of menopause and found no significant increase in weight or decrease in sexual function at 24 weeks of use. Participants were weighed during study visits, and those in the paroxetine group gained on average 0.48% from baseline at 24 weeks, compared with 0.09% in the placebo group (P = .29).

Sexual dysfunction was assessed using the Arizona Sexual Experience Scale, which has been validated in psychiatric patients using antidepressants, and there was no significant difference in symptoms such as sex drive, sexual arousal, vaginal lubrication, or ability to achieve orgasm between the treatment group and placebo group.¹⁵

Paroxetine inhibits CYP2D6 and thus decreases tamoxifen activity

Of note, paroxetine is a potent inhibitor of the cytochrome P-450 CYP2D6 enzyme, and concurrent use of paroxetine with tamoxifen decreases tamoxifen activity.^12,16 Since women with a history of breast cancer who cannot use estrogen for hot flashes may be seeking nonhormonal treatment for their vasomotor symptoms, providers should perform careful medication reconciliation and be aware that concomitant use of paroxetine and tamoxifen is not recommended.

Other antidepressants show promise but are not approved for menopausal symptoms

In addition to paroxetine, other nonhormonal drugs have been studied for treating hot flashes, but they have been unable to secure FDA approval for this indication. One of these is the serotonin-norepinephrine reuptake inhibitor venlafaxine, and a 2014 study¹⁷ confirmed its efficacy in treating menopausal vasomotor symptoms.

Joffe et al¹⁷ performed a three-armed trial comparing venlafaxine 75 mg/day, estradiol 0.5 mg/day, and placebo and found that both of the active treatments were better than placebo at reducing vasomotor symptoms. Compared with each other, estradiol 0.5 mg/day reduced hot flash frequency by an additional 0.6 events per day compared with venlafaxine 75 mg/day (P = .09). Though this difference was statistically significant, the authors pointed out that the clinical significance of such a small absolute difference is questionable. Additionally, providers should be aware that venlafaxine has little or no effect on the metabolism of tamoxifen.¹⁶

Shams et al,¹⁸ in a meta-analysis published in 2014, concluded that SSRIs as a class are more effective than placebo in treating hot flashes, supporting their widespread off-label use for this purpose. Their analysis examined the results of 11 studies, which included more than 2,000 patients in total, and found that compared with placebo, SSRI use was associated with a significant decrease in hot flashes (mean difference –0.93 events per day, 95% CI –1.49 to –0.37). A mixed treatment comparison analysis was also performed to try to model performance of individual SSRIs based on the pooled data, and the model suggests that escitalopram may be the most efficacious SSRI at reducing hot flash severity.

These studies support the effectiveness of SSRIs¹⁸ and venlafaxine¹⁷ in reducing hot flashes compared with placebo, though providers should be aware that they are still not FDA-approved for this indication.

Nonhormonal therapy for our patient

We would recommend paroxetine 7.5 mg nightly to this patient, as it is an FDA-approved nonhormonal medication that has been shown to help patients with vasomotor symptoms of menopause as well as sleep disturbance, without sexual side effects or weight gain. If the patient cannot tolerate paroxetine, off-label use of another SSRI or venlafaxine is supported by the recent literature.

HEART DISEASE IN WOMEN: CARDIAC RESYNCHRONIZATION THERAPY

A 68-year-old woman with a history of nonis­chemic cardiomyopathy presents for routine follow-up in your office. Despite maximal medical therapy on a beta-blocker, an angiotensin II receptor blocker, and a diuretic, she has New York Heart Association (NYHA) class III symptoms. Her most recent studies showed an ejection fraction of 30% by echocardiography and left bundle-branch block on electrocardiography, with a QRS duration of 140 ms. She recently saw her cardiologist, who recommended cardiac resynchronization therapy, and she wants your opinion as to whether or not to proceed with this recommendation. How should you counsel her?

Which patients are candidates for cardiac resynchronization therapy?

Heart disease continues to be the number one cause of death in the United States for both men and women, and almost the same number of women and men die from heart disease every year.¹⁹ Though coronary artery disease accounts for most cases of cardiovascular disease in the United States, heart failure is a significant and growing contributor. Approximately 6.6 million adults had heart failure in 2010 in the United States, and an additional 3 million are projected to have heart failure by 2030.²⁰ The burden of disease on our health system is high, with about 1 million hospitalizations and more than 3 million outpatient office visits attributable to heart failure yearly.²⁰

Patients with heart failure may have symptoms of dyspnea, fatigue, orthopnea, and periph­eral edema; laboratory and radiologic findings of pulmonary edema, renal insufficiency, and hyponatremia; and electrocardiographic findings of atrial fibrillation or prolonged QRS.²¹ Intraventricular conduction delay (QRS duration > 120 ms) is associated with dyssynchronous ventricular contraction and impaired pump function and is present in almost one-third of patients who have advanced heart failure.²¹

Heart disease continues to be the number one cause of death in both men and women

Cardiac resynchronization therapy, or biventricular pacing, can improve symptoms and pump function and has been shown to decrease rates of hospitalization and death in these patients.²² According to the joint 2012 guidelines of the American College of Cardiology Foundation, American Heart Association, and Heart Rhythm Society,²² it is indicated for patients with an ejection fraction of 35% or less, left bundle-branch block with QRS duration of 150 ms or more, and NYHA class II to IV symptoms who are in sinus rhythm (class I recommendation, level of evidence A).

Studies of cardiac resynchronization therapy in women

Recently published studies have suggested that women may derive greater benefit than men from cardiac resynchronization therapy.

Zusterzeel et al²³ (2014) evaluated sex-specific data from the National Cardiovascular Data Registry, which contains data on all biventricular pacemaker and implantable cardioverter-defibrillator implantations from 80% of US hospitals.²³ Of the 21,152 patients who had left bundle-branch block and received cardiac resynchronization therapy, women derived greater benefit in terms of death than men did, with a 21% lower risk of death than men (adjusted hazard ratio 0.79, 95% CI 0.74–0.84, P < .001). This study was also notable in that 36% of the patients were women, whereas in most earlier studies of cardiac resynchronization therapy women accounted for only 22% to 30% of the study population.²²

Goldenberg et al²⁴ (2014) performed a follow-up analysis of the Multicenter Automatic Defibrillator Implantation Trial With Cardiac Resynchronization Therapy. Subgroup analysis showed that although both men and women had a lower risk of death if they received cardiac resynchronization therapy compared with an implantable cardioverter-defibrillator only, the magnitude of benefit may be greater for women (hazard ratio 0.48, 95% CI 0.25–0.91, P = .03) than for men (hazard ratio 0.69, 95% CI 0.50–0.95, P = .02).

In addition to deriving greater mortality benefit, women may actually benefit from cardiac resynchronization therapy at shorter QRS durations than what is currently recommended. Women have a shorter baseline QRS than men, and a smaller left ventricular cavity.²⁵ In an FDA meta-analysis published in August 2014, pooled data from more than 4,000 patients in three studies suggested that women with left bundle-branch block benefited from cardiac resynchronization therapy more than men with left bundle-branch block.²⁶ Neither men nor women with left bundle-branch block benefited from it if their QRS duration was less than 130 ms, and both sexes benefited from it if they had left bundle-branch block and a QRS duration longer than 150 ms. However, women who received it who had left bundle-branch block and a QRS duration of 130 to 149 ms had a significant 76% reduction in the primary composite outcome of a heart failure event or death (hazard ratio 0.24, 95% CI 0.11–0.53, P < .001), while men in the same group did not derive significant benefit (hazard ratio 0.85, 95% CI 0.60–1.21, P = .38).

Despite the increasing evidence that there are sex-specific differences in the benefit from cardiac resynchronization therapy, what we know is limited by the low rates of female enrollment in most of the studies of this treatment. In a systematic review published in 2015, Herz et al²⁷ found that 90% of the 183 studies they reviewed enrolled 35% women or less, and half of the studies enrolled less than 23% women. Furthermore, only 20 of the 183 studies reported baseline characteristics by sex.

Recognizing this lack of adequate data, in August 2014 the FDA issued an official guidance statement outlining its expectations regarding sex-specific patient recruitment, data analysis, and data reporting in future medical device studies.²⁸ Hopefully, with this support for sex-specific research by the FDA, future studies will be able to identify therapeutic outcome differences that may exist between male and female patients.

Should our patient receive cardiac resynchronization therapy?

Regarding our patient with heart failure, the above studies suggest she will likely have a lower risk of death if she receives cardiac resynchronization therapy, even though her QRS interval is shorter than 150 ms. Providers who are aware of the emerging data regarding sex differences and treatment response can be powerful advocates for their patients, even in subspecialty areas, as highlighted by this case. We recommend counseling this patient to proceed with cardiac resynchronization therapy.

Women's health encompasses a broad range of issues unique to the female patient, with a scope that has expanded beyond reproductive health. Providers who care for women must develop cross-disciplinary competencies and understand the complex role of sex and gender on disease expression and treatment outcomes. Staying current with the literature in this rapidly changing field can be challenging for the busy clinician.

This article reviews recent advances in the treatment of depression in pregnancy, nonhormonal therapies for menopausal symptoms, and heart failure therapy in women, highlighting notable studies published in 2014 and early 2015.

TREATMENT OF DEPRESSION IN PREGNANCY

A 32-year-old woman with well-controlled but recurrent depression presents to the clinic for preconception counseling. Her depression has been successfully managed with a selective serotonin reuptake inhibitor (SSRI). She and her husband would like to try to conceive soon, but she is worried that continuing on her current SSRI may harm her baby. How should you advise her?

Concern for teratogenic effects of SSRIs

Depression is common during pregnancy: 11.8% to 13.5% of pregnant women report symptoms of depression,¹ and 7.5% of pregnant women take an antidepressant.²

SSRI use during pregnancy has drawn attention due to mixed reports of teratogenic effects

SSRI use during pregnancy has drawn attention because of mixed reports of teratogenic effects on the newborn, such as omphalocele, congenital heart defects, and craniosynostosis.³ Previous observational studies have specifically linked paroxetine to small but significant increases in right ventricular outflow tract obstruction^4,5 and have linked sertraline to ventricular septal defects.⁶

However, reports of associations of congenital malformations and SSRI use in pregnancy in observational studies have been questioned, with concern that these studies had low statistical power, self-reported data leading to recall bias, and limited assessment for confounding factors.^3,7

Recent studies refute risk of cardiac malformations

Several newer studies have been published that further examine the association between SSRI use in pregnancy and congenital heart defects, and their findings suggest that once adjusted for confounding variables, SSRI use in pregnancy may not be associated with cardiac malformations.

Huybrechts et al,⁸ in a large study published in 2014, extracted data on 950,000 pregnant women from the Medicaid database over a 7-year period and examined it for SSRI use during the first 90 days of pregnancy. Though SSRI use was associated with cardiac malformations when unadjusted for confounding variables (unadjusted relative risk 1.25, 95% confidence interval [CI] 1.13–1.38), once the cohort was restricted to women with a diagnosis of only depression and was adjusted based on propensity scoring, the association was no longer statistically significant (adjusted relative risk 1.06, 95% CI 0.93–1.22).

Additionally, there was no association between sertraline and ventricular septal defects (63 cases in 14,040 women exposed to sertraline, adjusted relative risk 1.04, 95% CI 0.76–1.41), or between paroxetine and right ventricular outflow tract obstruction (93 cases in 11,126 women exposed to paroxetine, adjusted relative risk 1.07, 95% CI 0.59–1.93).⁸

Furu et al⁷ conducted a sibling-matched case-control comparison published in 2015, in which more than 2 million live births from five Nordic countries were examined in the full cohort study and 2,288 births in the sibling-matched case-control cohort. SSRI or venlafaxine use in the first 90 days of pregnancy was examined. There was a slightly higher rate of cardiac defects in infants born to SSRI or venlafaxine recipients in the cohort study (adjusted odds ratio 1.15, 95% CI 1.05–1.26). However, in the sibling-controlled analyses, neither an SSRI nor venlafaxine was associated with heart defects (adjusted odds ratio 0.92, 95% CI 0.72–1.17), leading the authors to conclude that there might be familial factors or other lifestyle factors that were not taken into consideration and that could have confounded the cohort results.

Bérard et al⁹ examined antidepressant use in the first trimester of pregnancy in a cohort of women in Canada and concluded that sertraline was associated with congenital atrial and ventricular defects (risk ratio 1.34; 95% CI 1.02–1.76).⁹ However, this association should be interpreted with caution, as the Canadian cohort was notably smaller than those in other studies we have discussed, with only 18,493 pregnancies in the total cohort, and this conclusion was drawn from 9 cases of ventricular or atrial septal defects in babies of 366 women exposed to sertraline.

Although at first glance SSRIs may appear to be associated with congenital heart defects, these recent studies are reassuring and suggest that the association may actually not be significant. As with any statistical analysis, thoughtful study design, adequate statistical power, and adjustment for confounding factors must be considered before drawing conclusions.

SSRIs, offspring psychiatric outcomes, and miscarriage rates

Clements et al¹⁰ studied a cohort extracted from Partners Healthcare consisting of newborns with autism spectrum disorder, newborns with attention-deficit hyperactivity disorder (ADHD), and healthy matched controls and found that SSRI use during pregnancy was not associated with offspring autism spectrum disorder (adjusted odds ratio 1.10, 95% CI 0.7–1.70). However, they did find an increased risk of ADHD with SSRI use during pregnancy (adjusted odds ratio 1.81, 95% CI 1.22–2.70).

Andersen et al¹¹ examined more than 1 million pregnancies in Denmark and found no difference in risk of miscarriage between women who used an SSRI during pregnancy (adjusted hazard ratio 1.27) and women who discontinued their SSRI at least 3 months before pregnancy (adjusted hazard ratio 1.24, P = .47). The authors concluded that because of the similar rate of miscarriage in both groups, there was no association between SSRI use and miscarriage, and that the small increased risk of miscarriage in both groups could have been attributable to a confounding factor that was not measured.

Should our patient continue her SSRI through pregnancy?

Our patient has recurrent depression, and her risk of relapse with antidepressant cessation is high. Though previous, less well-done studies suggested a small risk of congenital heart defects, recent larger high-quality studies provide significant reassurance that SSRI use in pregnancy is not strongly associated with cardiac malformations. Recent studies also show no association with miscarriage or autism spectrum disorder, though there may be risk of offspring ADHD.

She can be counseled that she may continue on her SSRI during pregnancy and can be reassured that the risk to her baby is small compared with her risk of recurrent or postpartum depression.

NONHORMONAL TREATMENT FOR VASOMOTOR SYMPTOMS OF MENOPAUSE

You see a patient who is struggling with symptoms of menopause. She tells you she has terrible hot flashes day and night, and she would like to try drug therapy. She does not want hormone replacement therapy because she is worried about the risk of adverse events. Are there safe and effective nonhormonal pharmacologic treatments for her vasomotor symptoms?

Paroxetine 7.5 mg is approved for vasomotor symptoms of menopause

As many as 75% of menopausal women in the United States experience vasomotor symptoms related to menopause, or hot flashes and night sweats.¹² These symptoms can disrupt sleep and negatively affect quality of life. Though previously thought to occur during a short and self-limited time period, a recently published large observational study reported the median duration of vasomotor symptoms was 7.4 years, and in African American women in the cohort the median duration of vasomotor symptoms was 10.1 years—an entire decade of life.¹³

In 2013, the US Food and Drug Administration (FDA) approved paroxetine 7.5 mg daily for treating moderate to severe hot flashes associated with menopause. It is the only approved nonhormonal treatment for vasomotor symptoms; the only other approved treatments are estrogen therapy for women who have had a hysterectomy and combination estrogen-progesterone therapy for women who have not had a hysterectomy.

Further studies of paroxetine for menopausal symptoms

Since its approval, further studies have been published supporting the use of paroxetine 7.5 mg in treating symptoms of menopause. In addition to reducing hot flashes, this treatment also improves sleep disturbance in women with menopause.¹⁴

Pinkerton et al,¹⁴ in a pooled analysis of the data from the phase 3 clinical trials of paroxetine 7.5 mg per day, found that participants in groups assigned to paroxetine reported a 62% reduction in nighttime awakenings due to hot flashes compared with a 43% reduction in the placebo group (P < .001). Those who took paroxetine also reported a statistically significantly greater increase in duration of sleep than those who took placebo (37 minutes in the treatment group vs 27 minutes in the placebo group, P = .03).

Some patients are hesitant to take an SSRI because of concerns about adverse effects when used for psychiatric conditions. However, the dose of paroxetine that was studied and approved for vasomotor symptoms is lower than doses used for psychiatric indications and does not appear to be associated with these adverse effects.

Portman et al¹⁵ in 2014 examined the effect of paroxetine 7.5 mg vs placebo on weight gain and sexual function in women with vasomotor symptoms of menopause and found no significant increase in weight or decrease in sexual function at 24 weeks of use. Participants were weighed during study visits, and those in the paroxetine group gained on average 0.48% from baseline at 24 weeks, compared with 0.09% in the placebo group (P = .29).

Sexual dysfunction was assessed using the Arizona Sexual Experience Scale, which has been validated in psychiatric patients using antidepressants, and there was no significant difference in symptoms such as sex drive, sexual arousal, vaginal lubrication, or ability to achieve orgasm between the treatment group and placebo group.¹⁵

Paroxetine inhibits CYP2D6 and thus decreases tamoxifen activity

Of note, paroxetine is a potent inhibitor of the cytochrome P-450 CYP2D6 enzyme, and concurrent use of paroxetine with tamoxifen decreases tamoxifen activity.^12,16 Since women with a history of breast cancer who cannot use estrogen for hot flashes may be seeking nonhormonal treatment for their vasomotor symptoms, providers should perform careful medication reconciliation and be aware that concomitant use of paroxetine and tamoxifen is not recommended.

Other antidepressants show promise but are not approved for menopausal symptoms

In addition to paroxetine, other nonhormonal drugs have been studied for treating hot flashes, but they have been unable to secure FDA approval for this indication. One of these is the serotonin-norepinephrine reuptake inhibitor venlafaxine, and a 2014 study¹⁷ confirmed its efficacy in treating menopausal vasomotor symptoms.

Joffe et al¹⁷ performed a three-armed trial comparing venlafaxine 75 mg/day, estradiol 0.5 mg/day, and placebo and found that both of the active treatments were better than placebo at reducing vasomotor symptoms. Compared with each other, estradiol 0.5 mg/day reduced hot flash frequency by an additional 0.6 events per day compared with venlafaxine 75 mg/day (P = .09). Though this difference was statistically significant, the authors pointed out that the clinical significance of such a small absolute difference is questionable. Additionally, providers should be aware that venlafaxine has little or no effect on the metabolism of tamoxifen.¹⁶

Shams et al,¹⁸ in a meta-analysis published in 2014, concluded that SSRIs as a class are more effective than placebo in treating hot flashes, supporting their widespread off-label use for this purpose. Their analysis examined the results of 11 studies, which included more than 2,000 patients in total, and found that compared with placebo, SSRI use was associated with a significant decrease in hot flashes (mean difference –0.93 events per day, 95% CI –1.49 to –0.37). A mixed treatment comparison analysis was also performed to try to model performance of individual SSRIs based on the pooled data, and the model suggests that escitalopram may be the most efficacious SSRI at reducing hot flash severity.

These studies support the effectiveness of SSRIs¹⁸ and venlafaxine¹⁷ in reducing hot flashes compared with placebo, though providers should be aware that they are still not FDA-approved for this indication.

Nonhormonal therapy for our patient

We would recommend paroxetine 7.5 mg nightly to this patient, as it is an FDA-approved nonhormonal medication that has been shown to help patients with vasomotor symptoms of menopause as well as sleep disturbance, without sexual side effects or weight gain. If the patient cannot tolerate paroxetine, off-label use of another SSRI or venlafaxine is supported by the recent literature.

HEART DISEASE IN WOMEN: CARDIAC RESYNCHRONIZATION THERAPY

A 68-year-old woman with a history of nonis­chemic cardiomyopathy presents for routine follow-up in your office. Despite maximal medical therapy on a beta-blocker, an angiotensin II receptor blocker, and a diuretic, she has New York Heart Association (NYHA) class III symptoms. Her most recent studies showed an ejection fraction of 30% by echocardiography and left bundle-branch block on electrocardiography, with a QRS duration of 140 ms. She recently saw her cardiologist, who recommended cardiac resynchronization therapy, and she wants your opinion as to whether or not to proceed with this recommendation. How should you counsel her?

Which patients are candidates for cardiac resynchronization therapy?

Heart disease continues to be the number one cause of death in the United States for both men and women, and almost the same number of women and men die from heart disease every year.¹⁹ Though coronary artery disease accounts for most cases of cardiovascular disease in the United States, heart failure is a significant and growing contributor. Approximately 6.6 million adults had heart failure in 2010 in the United States, and an additional 3 million are projected to have heart failure by 2030.²⁰ The burden of disease on our health system is high, with about 1 million hospitalizations and more than 3 million outpatient office visits attributable to heart failure yearly.²⁰

Patients with heart failure may have symptoms of dyspnea, fatigue, orthopnea, and periph­eral edema; laboratory and radiologic findings of pulmonary edema, renal insufficiency, and hyponatremia; and electrocardiographic findings of atrial fibrillation or prolonged QRS.²¹ Intraventricular conduction delay (QRS duration > 120 ms) is associated with dyssynchronous ventricular contraction and impaired pump function and is present in almost one-third of patients who have advanced heart failure.²¹

Heart disease continues to be the number one cause of death in both men and women

Cardiac resynchronization therapy, or biventricular pacing, can improve symptoms and pump function and has been shown to decrease rates of hospitalization and death in these patients.²² According to the joint 2012 guidelines of the American College of Cardiology Foundation, American Heart Association, and Heart Rhythm Society,²² it is indicated for patients with an ejection fraction of 35% or less, left bundle-branch block with QRS duration of 150 ms or more, and NYHA class II to IV symptoms who are in sinus rhythm (class I recommendation, level of evidence A).

Studies of cardiac resynchronization therapy in women

Recently published studies have suggested that women may derive greater benefit than men from cardiac resynchronization therapy.

Zusterzeel et al²³ (2014) evaluated sex-specific data from the National Cardiovascular Data Registry, which contains data on all biventricular pacemaker and implantable cardioverter-defibrillator implantations from 80% of US hospitals.²³ Of the 21,152 patients who had left bundle-branch block and received cardiac resynchronization therapy, women derived greater benefit in terms of death than men did, with a 21% lower risk of death than men (adjusted hazard ratio 0.79, 95% CI 0.74–0.84, P < .001). This study was also notable in that 36% of the patients were women, whereas in most earlier studies of cardiac resynchronization therapy women accounted for only 22% to 30% of the study population.²²

Goldenberg et al²⁴ (2014) performed a follow-up analysis of the Multicenter Automatic Defibrillator Implantation Trial With Cardiac Resynchronization Therapy. Subgroup analysis showed that although both men and women had a lower risk of death if they received cardiac resynchronization therapy compared with an implantable cardioverter-defibrillator only, the magnitude of benefit may be greater for women (hazard ratio 0.48, 95% CI 0.25–0.91, P = .03) than for men (hazard ratio 0.69, 95% CI 0.50–0.95, P = .02).

In addition to deriving greater mortality benefit, women may actually benefit from cardiac resynchronization therapy at shorter QRS durations than what is currently recommended. Women have a shorter baseline QRS than men, and a smaller left ventricular cavity.²⁵ In an FDA meta-analysis published in August 2014, pooled data from more than 4,000 patients in three studies suggested that women with left bundle-branch block benefited from cardiac resynchronization therapy more than men with left bundle-branch block.²⁶ Neither men nor women with left bundle-branch block benefited from it if their QRS duration was less than 130 ms, and both sexes benefited from it if they had left bundle-branch block and a QRS duration longer than 150 ms. However, women who received it who had left bundle-branch block and a QRS duration of 130 to 149 ms had a significant 76% reduction in the primary composite outcome of a heart failure event or death (hazard ratio 0.24, 95% CI 0.11–0.53, P < .001), while men in the same group did not derive significant benefit (hazard ratio 0.85, 95% CI 0.60–1.21, P = .38).

Despite the increasing evidence that there are sex-specific differences in the benefit from cardiac resynchronization therapy, what we know is limited by the low rates of female enrollment in most of the studies of this treatment. In a systematic review published in 2015, Herz et al²⁷ found that 90% of the 183 studies they reviewed enrolled 35% women or less, and half of the studies enrolled less than 23% women. Furthermore, only 20 of the 183 studies reported baseline characteristics by sex.

Recognizing this lack of adequate data, in August 2014 the FDA issued an official guidance statement outlining its expectations regarding sex-specific patient recruitment, data analysis, and data reporting in future medical device studies.²⁸ Hopefully, with this support for sex-specific research by the FDA, future studies will be able to identify therapeutic outcome differences that may exist between male and female patients.

Should our patient receive cardiac resynchronization therapy?

Regarding our patient with heart failure, the above studies suggest she will likely have a lower risk of death if she receives cardiac resynchronization therapy, even though her QRS interval is shorter than 150 ms. Providers who are aware of the emerging data regarding sex differences and treatment response can be powerful advocates for their patients, even in subspecialty areas, as highlighted by this case. We recommend counseling this patient to proceed with cardiac resynchronization therapy.