Cleveland Clinic Journal of Medicine

To the Editor: We read with great interest the excellent review by Dore¹ on the prevention of glucocorticoid-induced osteoporosis. As indicated by the author, bone loss is one of the most serious complications of corticosteroid therapy, causing significant costs, morbidity, and mortality related to vertebral and hip fractures. Therefore, prevention of bone loss is mandatory, and several drugs are available.

However, the author does not mention strontium ranelate in the armamentarium for this preventive treatment. Strontium ranelate is an orally administered treatment of postmenopausal osteoporosis, reducing the risk of vertebral and hip fractures, and its efficacy has been demonstrated in clinical and histologic studies.^2,3 It has a particular mode of action, since it simultaneously inhibits bone resorption and stimulates bone formation.^2,3 Only minor adverse effects have been reported, including gastrointestinal signs such as nausea and diarrhea (only during the first 3 months), headache, and skin lesions. Strontium ranelate is currently licensed for the treatment of postmenopausal osteoporosis, but it appears to be an effective solution for diverse fracture risks, including the treatment of glucocorticoid-induced osteoporosis.

In a 2-year observational, controlled study that included 107 patients with glucocorticoid-induced osteoporosis treated with strontium ranelate or risedronate, there was a significantly higher increase in lumbar spine and total hip bone mineral density and a stronger reduction in back pain in the group of patients treated with strontium ranelate than in the group of patients under risedronate therapy, but the number of patients with no new fractures was similar in both treatment groups.⁴

In an animal model, strontium ranelate was significantly superior to alendronate in the prevention of glucocorticoid-induced osteopenia according to bone mineral density and histomorphometric analysis.⁵

Therefore, we consider that strontium ranelate could also be effective in glucocorticoid-induced osteopenia prevention, but prospective studies are required.

To the Editor: We read with great interest the excellent review by Dore¹ on the prevention of glucocorticoid-induced osteoporosis. As indicated by the author, bone loss is one of the most serious complications of corticosteroid therapy, causing significant costs, morbidity, and mortality related to vertebral and hip fractures. Therefore, prevention of bone loss is mandatory, and several drugs are available.

However, the author does not mention strontium ranelate in the armamentarium for this preventive treatment. Strontium ranelate is an orally administered treatment of postmenopausal osteoporosis, reducing the risk of vertebral and hip fractures, and its efficacy has been demonstrated in clinical and histologic studies.^2,3 It has a particular mode of action, since it simultaneously inhibits bone resorption and stimulates bone formation.^2,3 Only minor adverse effects have been reported, including gastrointestinal signs such as nausea and diarrhea (only during the first 3 months), headache, and skin lesions. Strontium ranelate is currently licensed for the treatment of postmenopausal osteoporosis, but it appears to be an effective solution for diverse fracture risks, including the treatment of glucocorticoid-induced osteoporosis.

In a 2-year observational, controlled study that included 107 patients with glucocorticoid-induced osteoporosis treated with strontium ranelate or risedronate, there was a significantly higher increase in lumbar spine and total hip bone mineral density and a stronger reduction in back pain in the group of patients treated with strontium ranelate than in the group of patients under risedronate therapy, but the number of patients with no new fractures was similar in both treatment groups.⁴

In an animal model, strontium ranelate was significantly superior to alendronate in the prevention of glucocorticoid-induced osteopenia according to bone mineral density and histomorphometric analysis.⁵

Therefore, we consider that strontium ranelate could also be effective in glucocorticoid-induced osteopenia prevention, but prospective studies are required.

To the Editor: We read with great interest the excellent review by Dore¹ on the prevention of glucocorticoid-induced osteoporosis. As indicated by the author, bone loss is one of the most serious complications of corticosteroid therapy, causing significant costs, morbidity, and mortality related to vertebral and hip fractures. Therefore, prevention of bone loss is mandatory, and several drugs are available.

However, the author does not mention strontium ranelate in the armamentarium for this preventive treatment. Strontium ranelate is an orally administered treatment of postmenopausal osteoporosis, reducing the risk of vertebral and hip fractures, and its efficacy has been demonstrated in clinical and histologic studies.^2,3 It has a particular mode of action, since it simultaneously inhibits bone resorption and stimulates bone formation.^2,3 Only minor adverse effects have been reported, including gastrointestinal signs such as nausea and diarrhea (only during the first 3 months), headache, and skin lesions. Strontium ranelate is currently licensed for the treatment of postmenopausal osteoporosis, but it appears to be an effective solution for diverse fracture risks, including the treatment of glucocorticoid-induced osteoporosis.

In a 2-year observational, controlled study that included 107 patients with glucocorticoid-induced osteoporosis treated with strontium ranelate or risedronate, there was a significantly higher increase in lumbar spine and total hip bone mineral density and a stronger reduction in back pain in the group of patients treated with strontium ranelate than in the group of patients under risedronate therapy, but the number of patients with no new fractures was similar in both treatment groups.⁴

In an animal model, strontium ranelate was significantly superior to alendronate in the prevention of glucocorticoid-induced osteopenia according to bone mineral density and histomorphometric analysis.⁵

Therefore, we consider that strontium ranelate could also be effective in glucocorticoid-induced osteopenia prevention, but prospective studies are required.

In Reply: I thank Drs. Bachmeyer and Gauthier for their kind comments. My review was limited to therapies currently available by prescription in the United States; therefore, strontium ranelate was not included. I agree with their comment that prospective studies are required to consider strontium ranelate as an effective therapy for glucocortocoid-induced osteoporosis.

In Reply: I thank Drs. Bachmeyer and Gauthier for their kind comments. My review was limited to therapies currently available by prescription in the United States; therefore, strontium ranelate was not included. I agree with their comment that prospective studies are required to consider strontium ranelate as an effective therapy for glucocortocoid-induced osteoporosis.

In Reply: I thank Drs. Bachmeyer and Gauthier for their kind comments. My review was limited to therapies currently available by prescription in the United States; therefore, strontium ranelate was not included. I agree with their comment that prospective studies are required to consider strontium ranelate as an effective therapy for glucocortocoid-induced osteoporosis.

To the Editor: Like Dr. Hanlon (Cleve Clin J Med 2010; 77:408–411), I too am alarmed by the inability of electronic medical records to incorporate whole language. Physicians can make treatment errors when they fail to include contextual factors in their diagnosis and treatment plans. The social and circumstantial complexities of a patient’s life cannot be parsed by computer systems that can only “search” bullet points. The current template-driven systems were originally designed for billing and now are touted for “quality measurements.” They could tell us whether a patient’s hemoglobin A_1c was at goal, or if she was “noncompliant” and hadn’t filled a prescription; they could not tell us that a psychologically abusive husband would not allow her to purchase her diabetes medications (this actually happened to one of my patients). I would argue that addressing the abuse is more important to her health. Yet we are all being pushed, like teachers teaching to a standardized test, to hit certain “benchmarks,” in order to be called “quality” physicians.

Since it is unlikely that the tide will turn back to a written record, physicians should be demanding rapid deployment of computer systems, now in development, that can analyze whole language and find information in context. This technology is out there and needs aggressive support.

Texting contractions, Twitter, and the rest are chipping away at the concept of narrative. Our patients’ lives are worthy of a narrative, not the bullet points and cut-and-paste we are forcing their lives and health into.

To the Editor: Like Dr. Hanlon (Cleve Clin J Med 2010; 77:408–411), I too am alarmed by the inability of electronic medical records to incorporate whole language. Physicians can make treatment errors when they fail to include contextual factors in their diagnosis and treatment plans. The social and circumstantial complexities of a patient’s life cannot be parsed by computer systems that can only “search” bullet points. The current template-driven systems were originally designed for billing and now are touted for “quality measurements.” They could tell us whether a patient’s hemoglobin A_1c was at goal, or if she was “noncompliant” and hadn’t filled a prescription; they could not tell us that a psychologically abusive husband would not allow her to purchase her diabetes medications (this actually happened to one of my patients). I would argue that addressing the abuse is more important to her health. Yet we are all being pushed, like teachers teaching to a standardized test, to hit certain “benchmarks,” in order to be called “quality” physicians.

Since it is unlikely that the tide will turn back to a written record, physicians should be demanding rapid deployment of computer systems, now in development, that can analyze whole language and find information in context. This technology is out there and needs aggressive support.

Texting contractions, Twitter, and the rest are chipping away at the concept of narrative. Our patients’ lives are worthy of a narrative, not the bullet points and cut-and-paste we are forcing their lives and health into.

To the Editor: Like Dr. Hanlon (Cleve Clin J Med 2010; 77:408–411), I too am alarmed by the inability of electronic medical records to incorporate whole language. Physicians can make treatment errors when they fail to include contextual factors in their diagnosis and treatment plans. The social and circumstantial complexities of a patient’s life cannot be parsed by computer systems that can only “search” bullet points. The current template-driven systems were originally designed for billing and now are touted for “quality measurements.” They could tell us whether a patient’s hemoglobin A_1c was at goal, or if she was “noncompliant” and hadn’t filled a prescription; they could not tell us that a psychologically abusive husband would not allow her to purchase her diabetes medications (this actually happened to one of my patients). I would argue that addressing the abuse is more important to her health. Yet we are all being pushed, like teachers teaching to a standardized test, to hit certain “benchmarks,” in order to be called “quality” physicians.

Since it is unlikely that the tide will turn back to a written record, physicians should be demanding rapid deployment of computer systems, now in development, that can analyze whole language and find information in context. This technology is out there and needs aggressive support.

Texting contractions, Twitter, and the rest are chipping away at the concept of narrative. Our patients’ lives are worthy of a narrative, not the bullet points and cut-and-paste we are forcing their lives and health into.

Children born with tetralogy of Fallot and other congenital heart defects are living longer—long enough for new problems to arise, and, eventually, to present to your clinic. In primary care, the presentation of tetralogy of Fallot is still rare, but it is becoming more common.

Congenital heart disease was once solely a pediatric specialty, but adults who have been treated for these conditions now outnumber children with congenital heart conditions.^1–4 More than 85% of infants with congenital heart disease are now expected to reach adulthood.^5,6 For those with tetralogy of Fallot, the most common form of cyanotic congenital heart disease, the 40-year survival rate is now at least 90%.⁵

But these former “blue babies” eventually have serious problems. Most develop pulmonary valve insufficiency (regurgitation), which, over time, can result in right ventricular volume overload, enlargement, and dysfunction. ^7–10 These problems lead to arrhythmias, the most significant cause of illness and death in these patients.^11–13 Ventricular and atrial arrhythmias occur in up to 35% of patients with tetralogy of Fallot, and over a follow-up period of up to 30 years the incidence of sudden cardiac death is 6%.¹⁴

Furthermore, because many patients have no symptoms in early adulthood, they are often lost to follow-up, potentially missing the opportunity to have complications treated before they become irreversible. Recent data suggest that most patients who present with symptoms had stopped seeing a cardiologist about 10 years before.¹⁵

The challenge for primary care clinicians is to identify these patients in their practice, to recognize the early signs and symptoms of a worsening condition, and to refer and treat before cardiac damage becomes irreversible.

ONE IN 3,600 LIVE BIRTHS

Tetralogy of Fallot occurs in approximately 1 in 3,600 live births or 3.5% of infants born with congenital heart disease.⁶ It is the most common type of cyanotic congenital heart disease, accounting for 10% of all cases.¹⁶ Patients in whom it has been repaired are the biggest group of adults with complex congenital heart disease. At Cleveland Clinic, it is the reason for 23% of new referrals to our adult congenital cardiology clinic, second only to atrial septal defects (33%).

FOUR DISTINCT FEATURES

• Pulmonary stenosis (subvalvar, valvar, or both subvalvar and valvar)
• Ventricular septal defect
• Hypertrophy of the right ventricle
• Rightward deviation of the aortic valve, so that it overrides the ventricular septum; this can range from minimal overriding of the aorta and trivial pulmonary stenosis to up to 90% override and frank pulmonary atresia.

The aorta, receiving blood from both ventricles, is usually dilated. It arises from a right-sided arch in about 25% of patients and may override the septum so much that more than 50% of the blood flow comes from the right ventricle.¹⁷ In such cases, whether the patient has true tetralogy of Fallot or a double-outlet right ventricle with pulmonic stenosis may be ambiguous. Though controversial, the latter condition is generally distinguished by a ventricular septal defect that is integral to the left ventricular outflow tract and by lack of fibrous continuity between the aortic and mitral valves.¹⁸

SURGERY HAS EVOLVED

Surgical repair has been performed since the 1950s, and the perioperative death rate has fallen to less than 1% at most experienced centers.¹⁹

In the past, surgeons often placed a shunt between a systemic artery and the pulmonary artery as a palliative measure to improve oxygenation in infants with tetralogy of Fallot, waiting until the child was older to remove the shunt and repair the defects definitively.

Now, however, they generally favor repairing the heart in the initial procedure. This involves patching the ventricular septal defect, widening the infundibulum, and repairing the pulmonary valve or patching the annulus. Transannular patching opens the entire right ventricular outflow tract, but it crosses the pulmonary valve, and this is what eventually results in severe pulmonary insufficiency and its complications.²⁰ For this reason, surgeons at most institutions now favor valve-sparing procedures rather than transannular patching, whenever possible.

WHAT HAPPENS YEARS AFTER SURGICAL REPAIR?

Surgery used to be considered the definitive cure for tetralogy of Fallot. However, problems that arise years later include chronic pulmonary valve insufficiency, obstruction of the right ventricular outflow tract, depressed right ventricular function, residual ventricular septal defect leaks, and arrhythmias.^17,21,22 For these reasons, many experts have abandoned the notion that surgical repair is definitive. ^23,24

Pulmonary valve insufficiency leads to right ventricular systolic dysfunction

In the past, pulmonary insufficiency was considered relatively benign because most patients tolerate it well for a long time. As these patients age, however, it becomes the core of their problems.¹ If severe, it may result in right ventricular volume overload and dilatation, fibrosis, arrhythmia, and myocardial damage, all of which are cumulatively detrimental.²⁵ Right ventricular function and exercise capacity deteriorate, and the tendency toward ventricular arrhythmias develops.²⁶

If the problem is chronic, right ventricular systolic function may remain normal for years, during which most patients remain relatively free of symptoms. In time, however, the compensatory mechanisms of the right ventricular myocardium fail, the right ventricular wall stress (afterload) increases, while the right ventricular ejection fraction decreases. Patients begin to experience symptoms, and if the volume load is not reduced, the dysfunction may become irreversible.²⁷

PULMONARY INSUFFICIENCY PREDISPOSES TO ARRHYTHMIAS, SUDDEN CARDIAC DEATH

Pulmonary insufficiency predisposes to atrial and ventricular arrhythmias, presumably due to progressive enlargement and stretching of the right atrium and ventricle.

Clinically significant atrial arrhythmias, predominantly intra-atrial reentrant tachycardia but also atrial tachycardia and atrial fibrillation, occur in 12% to 35% of patients with repaired tetralogy of Fallot.^11,28–31

Ventricular arrhythmias and sudden cardiac death also occur. In one study,¹ 100% of patients who died suddenly had moderate or severe pulmonary insufficiency, and 94% with ventricular tachycardia had significant pulmonary insufficiency. In contrast, only 49% of patients who were arrhythmia-free had significant pulmonary insufficiency. None of the patients with late sudden death or ventricular tachycardia had undergone late pulmonary valve replacement. This is further supported by a multicenter analysis of patients with repaired tetralogy, which demonstrated that moderate or severe pulmonary insufficiency was the main hemodynamic abnormality in patients with ventricular tachycardia and sudden death.¹¹

In general, the risk of late sudden death is 25 to 100 times higher in patients who survive surgery for congenital heart disease than in age-matched controls, and the risk is even higher for those with cyanotic conditions such as tetralogy of Fallot. In fact, one-third to one-half of deaths in adults with tetralogy of Fallot are sudden.^25,32

FINDINGS ON ASSESSMENT

Most patients with tetralogy of Fallot remain free of symptoms for many years. While individual responses to pulmonary insufficiency vary, symptoms generally get worse as the pulmonary insufficiency gets worse. Patients present with a spectrum of complaints, from palpitations to a general decline in function. Late symptoms include exertional dyspnea, palpitations, right heart failure, and syncope.¹⁷

Signs of right ventricular failure can include elevated jugular venous pressure, peripheral edema, hepatomegaly, ascites,³³ and jugular venous distention with a large a wave.

Heart murmurs

Pulmonary insufficiency causes a low-pitched, brief diastolic murmur. Although often present, it may be short or difficult to hear, even if the regurgitation is severe, because this is “low-pressure” pulmonary insufficiency as opposed to the regurgitation that can occur in patients with pulmonary hypertension. Therefore, this murmur is often missed on physical examination.

There may be an ejection click due to a dilated aorta. An aortic insufficiency murmur may also be present.

A right ventricular outflow murmur is generally audible, along with a pansystolic murmur if a residual ventricular septal defect is also present.

A right-sided aortic arch, present in about 25% of patients with tetralogy of Fallot, may cause a lift below the right sternoclavicular junction.¹⁷

Electrocardiographic findings

Electrocardiography commonly shows right ventricular hypertrophy with a right bundle branch block. The longer the QRS duration, the greater the right ventricular volume and mass. Furthermore, a QRS duration greater than 180 ms is a significant marker of risk of ventricular arrhythmias and sudden death.^22,34–37

Another feature strongly associated with ventricular arrhythmias and sudden death is the rate of change in the QRS duration. A relatively rapid increase (> 3.5 ms/year) is associated with a significantly higher risk.¹ A rapid rate of change may be meaningful even if the QRS duration is not markedly prolonged.¹¹

Reduced heart rate variability also appears to be a marker of risk of sudden cardiac death in these patients.^38,39

Chest radiography typically shows a prominent right ventricular shadow and cardiomegaly.¹⁷

Many centers specializing in congenital heart disease therefore recommend baseline cardiac MRI, even for patients without symptoms.³³

PULMONARY VALVE REPLACEMENT IS THE ONLY PROVEN TREATMENT

No study has yet shown that drug therapy alone slows the progression of complications.¹ Pulmonary valve replacement is the only treatment proven to reduce right ventricular size and improve right ventricular function in the long term.

The risks of surgery, including the need for repeat operations, must be balanced against the risk of irreversible right ventricular dysfunction and its associated complications. The operative death rate is low, as is the long-term risk of death afterward. Therrien et al¹² reported that, in a series of 70 patients who underwent pulmonary valve replacement, the probability of survival was 92% at 5 years and 86% at 10 years.

Surgery appears to reverse or at least arrest the progression of many of the complications associated with pulmonary insufficiency, including tricuspid regurgitation and diastolic dysfunction.¹⁷ Its utility in ameliorating ventricular tachycardia, however, remains controversial. One series showed a lower prevalence of tachycardia after pulmonary valve replacement (9% after surgery vs 22% before), but later studies have had more equivocal results.¹⁷

When should surgery be done?

There is little controversy about the eventual need for pulmonary valve replacement in most patients. What is controversial is the timing.^12,44–47

This issue has been hotly debated. Some believe that pulmonary valve replacement should be done only if evidence of right ventricular dysfunction has developed.¹⁷ Others suggest that it be considered earlier and that the onset of symptoms may be a late and suboptimal indication for it.^6,8,48,49 Many experts now recommend surgery early, before symptoms of heart failure develop.¹⁷ Though surgery has traditionally been recommended if the QRS duration is longer than 180 ms, some believe it should be done before this occurs.¹¹

Arguments for early surgery. In one study, in no patient who had a right ventricular end-diastolic volume greater than 170 mL/m² (normal ≤ 108) or a right ventricular end-systolic volume greater than 85 mL/m² (normal ≤ 47) did these numbers return to normal after pulmonary valve replacement.^45,50 This suggests a point of irreversible dilatation and a volume threshold beyond which right ventricular function is unlikely to completely improve. Normalization of right ventricular volumes was shown to occur when pulmonary valve replacement was performed before the right ventricular end-diastolic volume reached 160 mL/m² or the right ventricular end-systolic volume reached 82 mL/m².^47,51

Delaying surgery until symptoms occur may be unfavorable because the long-term outcomes of increased right ventricular volumes and decreased right ventricular ejection fractions after surgery are not known.

Arguments for watchful waiting. There does not seem to be a threshold above which right ventricular volumes do not decrease after surgery—although they may not decrease to the normal range. Pulmonary valve replacement substantially reduced right ventricular dilatation even in patients with very high right ventricular volumes and right ventricular dysfunction, and resulted in an overall improvement in function (measured by New York Heart Association class).⁴⁷

Late pulmonary valve replacement rapidly improves right ventricular volumes and improves the effective ejection fraction, although its impact on absolute right ventricular function is not as pronounced. The QRS duration shortened after surgery in those in whom it was 180 ms or longer before surgery, although this appeared to be a transient change.⁵² The prevalence of ventricular tachycardia declined from 22% to 9% and that of atrial fibrillation or flutter declined from 17% to 12%.^17,48

A recent study with long-term follow-up has raised questions about the necessity of aggressive early intervention in tetralogy of Fallot. Sixty-seven patients were followed for as long as 27 years after surgery. Forty-five had severe pulmonary insufficiency and severe right ventricular dilatation, and of those, 28 remained free of symptoms and did not undergo pulmonary valve replacement. The authors found that refraining from pulmonary valve replacement in asymptomatic patients with severe pulmonary insufficiency led to no measurable deterioration in 25 of 28 patients.⁵³

The available data do not support pulmonary valve replacement in young patients with mild or moderate right ventricular dilatation, normal right ventricular systolic function, and no additional risk factors.²⁷

Mechanical vs bioprosthetic replacement valves

Once the decision is made to proceed to surgery, the next step is choosing the type of prosthetic valve.

Mechanical valves pose a risk of thrombosis, requiring life-long anticoagulation. To give warfarin (Coumadin) to younger, active people exposes them to the risk of potentially catastrophic bleeding if trauma were to occur. Women who become pregnant are generally at an increased risk of thrombotic complications due to the hypercoagulable state of pregnancy, but the risk of fetal defects is considerable if they receive warfarin.^54–56

Bioprosthetic valves generally come in two varieties: preserved and treated human tissue (homografts) and animal tissue (bovine pericardial or porcine, depending on the size required). These can be implanted as isolated valves or as part of a conduit (valve and surrounding tissue).

Bioprosthetic valves eliminate the need for anticoagulation. However, they are not very durable, especially in younger patients, which is worrisome. An estimated 45% of bioprosthetic valves fail by 10 years,⁵⁷ thus nearly guaranteeing that an otherwise healthy 40-year-old, for example, will need to undergo at least one repeat surgery, and very likely more.

NOVEL THERAPIES

Percutaneous valve replacement

The future of pulmonary valve replacement may lie in percutaneous procedures.

The Melody valve is now approved through a humanitarian device exemption (ie, based on demonstrated safety without proven efficacy) for patients who have a prior pulmonary conduit now complicated by either stenosis or regurgitation.

If percutaneous pulmonary valve replacement proves to have reasonable long-term durability, it has the potential to dramatically shift the balance toward earlier intervention.

Pulmonary vasodilator drugs

Our group is examining whether pharmacologic therapy can alter the clinical outcome in patients with pulmonary insufficiency (due to either tetralogy of Fallot or valvotomy done to treat remote pulmonary stenosis). Specifically, we are using MRI to examine the effects of inhaled nitric oxide, a selective pulmonary vasodilator. Preliminary results suggest that such a strategy may work, and we are designing a trial to examine the longer-term benefit of using an oral drug with similar properties.

Children born with tetralogy of Fallot and other congenital heart defects are living longer—long enough for new problems to arise, and, eventually, to present to your clinic. In primary care, the presentation of tetralogy of Fallot is still rare, but it is becoming more common.

Congenital heart disease was once solely a pediatric specialty, but adults who have been treated for these conditions now outnumber children with congenital heart conditions.^1–4 More than 85% of infants with congenital heart disease are now expected to reach adulthood.^5,6 For those with tetralogy of Fallot, the most common form of cyanotic congenital heart disease, the 40-year survival rate is now at least 90%.⁵

But these former “blue babies” eventually have serious problems. Most develop pulmonary valve insufficiency (regurgitation), which, over time, can result in right ventricular volume overload, enlargement, and dysfunction. ^7–10 These problems lead to arrhythmias, the most significant cause of illness and death in these patients.^11–13 Ventricular and atrial arrhythmias occur in up to 35% of patients with tetralogy of Fallot, and over a follow-up period of up to 30 years the incidence of sudden cardiac death is 6%.¹⁴

Furthermore, because many patients have no symptoms in early adulthood, they are often lost to follow-up, potentially missing the opportunity to have complications treated before they become irreversible. Recent data suggest that most patients who present with symptoms had stopped seeing a cardiologist about 10 years before.¹⁵

The challenge for primary care clinicians is to identify these patients in their practice, to recognize the early signs and symptoms of a worsening condition, and to refer and treat before cardiac damage becomes irreversible.

ONE IN 3,600 LIVE BIRTHS

Tetralogy of Fallot occurs in approximately 1 in 3,600 live births or 3.5% of infants born with congenital heart disease.⁶ It is the most common type of cyanotic congenital heart disease, accounting for 10% of all cases.¹⁶ Patients in whom it has been repaired are the biggest group of adults with complex congenital heart disease. At Cleveland Clinic, it is the reason for 23% of new referrals to our adult congenital cardiology clinic, second only to atrial septal defects (33%).

FOUR DISTINCT FEATURES

• Pulmonary stenosis (subvalvar, valvar, or both subvalvar and valvar)
• Ventricular septal defect
• Hypertrophy of the right ventricle
• Rightward deviation of the aortic valve, so that it overrides the ventricular septum; this can range from minimal overriding of the aorta and trivial pulmonary stenosis to up to 90% override and frank pulmonary atresia.

The aorta, receiving blood from both ventricles, is usually dilated. It arises from a right-sided arch in about 25% of patients and may override the septum so much that more than 50% of the blood flow comes from the right ventricle.¹⁷ In such cases, whether the patient has true tetralogy of Fallot or a double-outlet right ventricle with pulmonic stenosis may be ambiguous. Though controversial, the latter condition is generally distinguished by a ventricular septal defect that is integral to the left ventricular outflow tract and by lack of fibrous continuity between the aortic and mitral valves.¹⁸

SURGERY HAS EVOLVED

Surgical repair has been performed since the 1950s, and the perioperative death rate has fallen to less than 1% at most experienced centers.¹⁹

In the past, surgeons often placed a shunt between a systemic artery and the pulmonary artery as a palliative measure to improve oxygenation in infants with tetralogy of Fallot, waiting until the child was older to remove the shunt and repair the defects definitively.

Now, however, they generally favor repairing the heart in the initial procedure. This involves patching the ventricular septal defect, widening the infundibulum, and repairing the pulmonary valve or patching the annulus. Transannular patching opens the entire right ventricular outflow tract, but it crosses the pulmonary valve, and this is what eventually results in severe pulmonary insufficiency and its complications.²⁰ For this reason, surgeons at most institutions now favor valve-sparing procedures rather than transannular patching, whenever possible.

WHAT HAPPENS YEARS AFTER SURGICAL REPAIR?

Surgery used to be considered the definitive cure for tetralogy of Fallot. However, problems that arise years later include chronic pulmonary valve insufficiency, obstruction of the right ventricular outflow tract, depressed right ventricular function, residual ventricular septal defect leaks, and arrhythmias.^17,21,22 For these reasons, many experts have abandoned the notion that surgical repair is definitive. ^23,24

Pulmonary valve insufficiency leads to right ventricular systolic dysfunction

In the past, pulmonary insufficiency was considered relatively benign because most patients tolerate it well for a long time. As these patients age, however, it becomes the core of their problems.¹ If severe, it may result in right ventricular volume overload and dilatation, fibrosis, arrhythmia, and myocardial damage, all of which are cumulatively detrimental.²⁵ Right ventricular function and exercise capacity deteriorate, and the tendency toward ventricular arrhythmias develops.²⁶

If the problem is chronic, right ventricular systolic function may remain normal for years, during which most patients remain relatively free of symptoms. In time, however, the compensatory mechanisms of the right ventricular myocardium fail, the right ventricular wall stress (afterload) increases, while the right ventricular ejection fraction decreases. Patients begin to experience symptoms, and if the volume load is not reduced, the dysfunction may become irreversible.²⁷

PULMONARY INSUFFICIENCY PREDISPOSES TO ARRHYTHMIAS, SUDDEN CARDIAC DEATH

Pulmonary insufficiency predisposes to atrial and ventricular arrhythmias, presumably due to progressive enlargement and stretching of the right atrium and ventricle.

Clinically significant atrial arrhythmias, predominantly intra-atrial reentrant tachycardia but also atrial tachycardia and atrial fibrillation, occur in 12% to 35% of patients with repaired tetralogy of Fallot.^11,28–31

Ventricular arrhythmias and sudden cardiac death also occur. In one study,¹ 100% of patients who died suddenly had moderate or severe pulmonary insufficiency, and 94% with ventricular tachycardia had significant pulmonary insufficiency. In contrast, only 49% of patients who were arrhythmia-free had significant pulmonary insufficiency. None of the patients with late sudden death or ventricular tachycardia had undergone late pulmonary valve replacement. This is further supported by a multicenter analysis of patients with repaired tetralogy, which demonstrated that moderate or severe pulmonary insufficiency was the main hemodynamic abnormality in patients with ventricular tachycardia and sudden death.¹¹

In general, the risk of late sudden death is 25 to 100 times higher in patients who survive surgery for congenital heart disease than in age-matched controls, and the risk is even higher for those with cyanotic conditions such as tetralogy of Fallot. In fact, one-third to one-half of deaths in adults with tetralogy of Fallot are sudden.^25,32

FINDINGS ON ASSESSMENT

Most patients with tetralogy of Fallot remain free of symptoms for many years. While individual responses to pulmonary insufficiency vary, symptoms generally get worse as the pulmonary insufficiency gets worse. Patients present with a spectrum of complaints, from palpitations to a general decline in function. Late symptoms include exertional dyspnea, palpitations, right heart failure, and syncope.¹⁷

Signs of right ventricular failure can include elevated jugular venous pressure, peripheral edema, hepatomegaly, ascites,³³ and jugular venous distention with a large a wave.

Heart murmurs

Pulmonary insufficiency causes a low-pitched, brief diastolic murmur. Although often present, it may be short or difficult to hear, even if the regurgitation is severe, because this is “low-pressure” pulmonary insufficiency as opposed to the regurgitation that can occur in patients with pulmonary hypertension. Therefore, this murmur is often missed on physical examination.

There may be an ejection click due to a dilated aorta. An aortic insufficiency murmur may also be present.

A right ventricular outflow murmur is generally audible, along with a pansystolic murmur if a residual ventricular septal defect is also present.

A right-sided aortic arch, present in about 25% of patients with tetralogy of Fallot, may cause a lift below the right sternoclavicular junction.¹⁷

Electrocardiographic findings

Electrocardiography commonly shows right ventricular hypertrophy with a right bundle branch block. The longer the QRS duration, the greater the right ventricular volume and mass. Furthermore, a QRS duration greater than 180 ms is a significant marker of risk of ventricular arrhythmias and sudden death.^22,34–37

Another feature strongly associated with ventricular arrhythmias and sudden death is the rate of change in the QRS duration. A relatively rapid increase (> 3.5 ms/year) is associated with a significantly higher risk.¹ A rapid rate of change may be meaningful even if the QRS duration is not markedly prolonged.¹¹

Reduced heart rate variability also appears to be a marker of risk of sudden cardiac death in these patients.^38,39

Chest radiography typically shows a prominent right ventricular shadow and cardiomegaly.¹⁷

Many centers specializing in congenital heart disease therefore recommend baseline cardiac MRI, even for patients without symptoms.³³

PULMONARY VALVE REPLACEMENT IS THE ONLY PROVEN TREATMENT

No study has yet shown that drug therapy alone slows the progression of complications.¹ Pulmonary valve replacement is the only treatment proven to reduce right ventricular size and improve right ventricular function in the long term.

The risks of surgery, including the need for repeat operations, must be balanced against the risk of irreversible right ventricular dysfunction and its associated complications. The operative death rate is low, as is the long-term risk of death afterward. Therrien et al¹² reported that, in a series of 70 patients who underwent pulmonary valve replacement, the probability of survival was 92% at 5 years and 86% at 10 years.

Surgery appears to reverse or at least arrest the progression of many of the complications associated with pulmonary insufficiency, including tricuspid regurgitation and diastolic dysfunction.¹⁷ Its utility in ameliorating ventricular tachycardia, however, remains controversial. One series showed a lower prevalence of tachycardia after pulmonary valve replacement (9% after surgery vs 22% before), but later studies have had more equivocal results.¹⁷

When should surgery be done?

There is little controversy about the eventual need for pulmonary valve replacement in most patients. What is controversial is the timing.^12,44–47

This issue has been hotly debated. Some believe that pulmonary valve replacement should be done only if evidence of right ventricular dysfunction has developed.¹⁷ Others suggest that it be considered earlier and that the onset of symptoms may be a late and suboptimal indication for it.^6,8,48,49 Many experts now recommend surgery early, before symptoms of heart failure develop.¹⁷ Though surgery has traditionally been recommended if the QRS duration is longer than 180 ms, some believe it should be done before this occurs.¹¹

Arguments for early surgery. In one study, in no patient who had a right ventricular end-diastolic volume greater than 170 mL/m² (normal ≤ 108) or a right ventricular end-systolic volume greater than 85 mL/m² (normal ≤ 47) did these numbers return to normal after pulmonary valve replacement.^45,50 This suggests a point of irreversible dilatation and a volume threshold beyond which right ventricular function is unlikely to completely improve. Normalization of right ventricular volumes was shown to occur when pulmonary valve replacement was performed before the right ventricular end-diastolic volume reached 160 mL/m² or the right ventricular end-systolic volume reached 82 mL/m².^47,51

Delaying surgery until symptoms occur may be unfavorable because the long-term outcomes of increased right ventricular volumes and decreased right ventricular ejection fractions after surgery are not known.

Arguments for watchful waiting. There does not seem to be a threshold above which right ventricular volumes do not decrease after surgery—although they may not decrease to the normal range. Pulmonary valve replacement substantially reduced right ventricular dilatation even in patients with very high right ventricular volumes and right ventricular dysfunction, and resulted in an overall improvement in function (measured by New York Heart Association class).⁴⁷

Late pulmonary valve replacement rapidly improves right ventricular volumes and improves the effective ejection fraction, although its impact on absolute right ventricular function is not as pronounced. The QRS duration shortened after surgery in those in whom it was 180 ms or longer before surgery, although this appeared to be a transient change.⁵² The prevalence of ventricular tachycardia declined from 22% to 9% and that of atrial fibrillation or flutter declined from 17% to 12%.^17,48

A recent study with long-term follow-up has raised questions about the necessity of aggressive early intervention in tetralogy of Fallot. Sixty-seven patients were followed for as long as 27 years after surgery. Forty-five had severe pulmonary insufficiency and severe right ventricular dilatation, and of those, 28 remained free of symptoms and did not undergo pulmonary valve replacement. The authors found that refraining from pulmonary valve replacement in asymptomatic patients with severe pulmonary insufficiency led to no measurable deterioration in 25 of 28 patients.⁵³

The available data do not support pulmonary valve replacement in young patients with mild or moderate right ventricular dilatation, normal right ventricular systolic function, and no additional risk factors.²⁷

Mechanical vs bioprosthetic replacement valves

Once the decision is made to proceed to surgery, the next step is choosing the type of prosthetic valve.

Mechanical valves pose a risk of thrombosis, requiring life-long anticoagulation. To give warfarin (Coumadin) to younger, active people exposes them to the risk of potentially catastrophic bleeding if trauma were to occur. Women who become pregnant are generally at an increased risk of thrombotic complications due to the hypercoagulable state of pregnancy, but the risk of fetal defects is considerable if they receive warfarin.^54–56

Bioprosthetic valves generally come in two varieties: preserved and treated human tissue (homografts) and animal tissue (bovine pericardial or porcine, depending on the size required). These can be implanted as isolated valves or as part of a conduit (valve and surrounding tissue).

Bioprosthetic valves eliminate the need for anticoagulation. However, they are not very durable, especially in younger patients, which is worrisome. An estimated 45% of bioprosthetic valves fail by 10 years,⁵⁷ thus nearly guaranteeing that an otherwise healthy 40-year-old, for example, will need to undergo at least one repeat surgery, and very likely more.

NOVEL THERAPIES

Percutaneous valve replacement

The future of pulmonary valve replacement may lie in percutaneous procedures.

The Melody valve is now approved through a humanitarian device exemption (ie, based on demonstrated safety without proven efficacy) for patients who have a prior pulmonary conduit now complicated by either stenosis or regurgitation.

If percutaneous pulmonary valve replacement proves to have reasonable long-term durability, it has the potential to dramatically shift the balance toward earlier intervention.

Pulmonary vasodilator drugs

Our group is examining whether pharmacologic therapy can alter the clinical outcome in patients with pulmonary insufficiency (due to either tetralogy of Fallot or valvotomy done to treat remote pulmonary stenosis). Specifically, we are using MRI to examine the effects of inhaled nitric oxide, a selective pulmonary vasodilator. Preliminary results suggest that such a strategy may work, and we are designing a trial to examine the longer-term benefit of using an oral drug with similar properties.

Children born with tetralogy of Fallot and other congenital heart defects are living longer—long enough for new problems to arise, and, eventually, to present to your clinic. In primary care, the presentation of tetralogy of Fallot is still rare, but it is becoming more common.

Congenital heart disease was once solely a pediatric specialty, but adults who have been treated for these conditions now outnumber children with congenital heart conditions.^1–4 More than 85% of infants with congenital heart disease are now expected to reach adulthood.^5,6 For those with tetralogy of Fallot, the most common form of cyanotic congenital heart disease, the 40-year survival rate is now at least 90%.⁵

But these former “blue babies” eventually have serious problems. Most develop pulmonary valve insufficiency (regurgitation), which, over time, can result in right ventricular volume overload, enlargement, and dysfunction. ^7–10 These problems lead to arrhythmias, the most significant cause of illness and death in these patients.^11–13 Ventricular and atrial arrhythmias occur in up to 35% of patients with tetralogy of Fallot, and over a follow-up period of up to 30 years the incidence of sudden cardiac death is 6%.¹⁴

Furthermore, because many patients have no symptoms in early adulthood, they are often lost to follow-up, potentially missing the opportunity to have complications treated before they become irreversible. Recent data suggest that most patients who present with symptoms had stopped seeing a cardiologist about 10 years before.¹⁵

The challenge for primary care clinicians is to identify these patients in their practice, to recognize the early signs and symptoms of a worsening condition, and to refer and treat before cardiac damage becomes irreversible.

ONE IN 3,600 LIVE BIRTHS

Tetralogy of Fallot occurs in approximately 1 in 3,600 live births or 3.5% of infants born with congenital heart disease.⁶ It is the most common type of cyanotic congenital heart disease, accounting for 10% of all cases.¹⁶ Patients in whom it has been repaired are the biggest group of adults with complex congenital heart disease. At Cleveland Clinic, it is the reason for 23% of new referrals to our adult congenital cardiology clinic, second only to atrial septal defects (33%).

FOUR DISTINCT FEATURES

• Pulmonary stenosis (subvalvar, valvar, or both subvalvar and valvar)
• Ventricular septal defect
• Hypertrophy of the right ventricle
• Rightward deviation of the aortic valve, so that it overrides the ventricular septum; this can range from minimal overriding of the aorta and trivial pulmonary stenosis to up to 90% override and frank pulmonary atresia.

The aorta, receiving blood from both ventricles, is usually dilated. It arises from a right-sided arch in about 25% of patients and may override the septum so much that more than 50% of the blood flow comes from the right ventricle.¹⁷ In such cases, whether the patient has true tetralogy of Fallot or a double-outlet right ventricle with pulmonic stenosis may be ambiguous. Though controversial, the latter condition is generally distinguished by a ventricular septal defect that is integral to the left ventricular outflow tract and by lack of fibrous continuity between the aortic and mitral valves.¹⁸

SURGERY HAS EVOLVED

Surgical repair has been performed since the 1950s, and the perioperative death rate has fallen to less than 1% at most experienced centers.¹⁹

In the past, surgeons often placed a shunt between a systemic artery and the pulmonary artery as a palliative measure to improve oxygenation in infants with tetralogy of Fallot, waiting until the child was older to remove the shunt and repair the defects definitively.

Now, however, they generally favor repairing the heart in the initial procedure. This involves patching the ventricular septal defect, widening the infundibulum, and repairing the pulmonary valve or patching the annulus. Transannular patching opens the entire right ventricular outflow tract, but it crosses the pulmonary valve, and this is what eventually results in severe pulmonary insufficiency and its complications.²⁰ For this reason, surgeons at most institutions now favor valve-sparing procedures rather than transannular patching, whenever possible.

WHAT HAPPENS YEARS AFTER SURGICAL REPAIR?

Surgery used to be considered the definitive cure for tetralogy of Fallot. However, problems that arise years later include chronic pulmonary valve insufficiency, obstruction of the right ventricular outflow tract, depressed right ventricular function, residual ventricular septal defect leaks, and arrhythmias.^17,21,22 For these reasons, many experts have abandoned the notion that surgical repair is definitive. ^23,24

Pulmonary valve insufficiency leads to right ventricular systolic dysfunction

In the past, pulmonary insufficiency was considered relatively benign because most patients tolerate it well for a long time. As these patients age, however, it becomes the core of their problems.¹ If severe, it may result in right ventricular volume overload and dilatation, fibrosis, arrhythmia, and myocardial damage, all of which are cumulatively detrimental.²⁵ Right ventricular function and exercise capacity deteriorate, and the tendency toward ventricular arrhythmias develops.²⁶

If the problem is chronic, right ventricular systolic function may remain normal for years, during which most patients remain relatively free of symptoms. In time, however, the compensatory mechanisms of the right ventricular myocardium fail, the right ventricular wall stress (afterload) increases, while the right ventricular ejection fraction decreases. Patients begin to experience symptoms, and if the volume load is not reduced, the dysfunction may become irreversible.²⁷

PULMONARY INSUFFICIENCY PREDISPOSES TO ARRHYTHMIAS, SUDDEN CARDIAC DEATH

Pulmonary insufficiency predisposes to atrial and ventricular arrhythmias, presumably due to progressive enlargement and stretching of the right atrium and ventricle.

Clinically significant atrial arrhythmias, predominantly intra-atrial reentrant tachycardia but also atrial tachycardia and atrial fibrillation, occur in 12% to 35% of patients with repaired tetralogy of Fallot.^11,28–31

Ventricular arrhythmias and sudden cardiac death also occur. In one study,¹ 100% of patients who died suddenly had moderate or severe pulmonary insufficiency, and 94% with ventricular tachycardia had significant pulmonary insufficiency. In contrast, only 49% of patients who were arrhythmia-free had significant pulmonary insufficiency. None of the patients with late sudden death or ventricular tachycardia had undergone late pulmonary valve replacement. This is further supported by a multicenter analysis of patients with repaired tetralogy, which demonstrated that moderate or severe pulmonary insufficiency was the main hemodynamic abnormality in patients with ventricular tachycardia and sudden death.¹¹

In general, the risk of late sudden death is 25 to 100 times higher in patients who survive surgery for congenital heart disease than in age-matched controls, and the risk is even higher for those with cyanotic conditions such as tetralogy of Fallot. In fact, one-third to one-half of deaths in adults with tetralogy of Fallot are sudden.^25,32

FINDINGS ON ASSESSMENT

Most patients with tetralogy of Fallot remain free of symptoms for many years. While individual responses to pulmonary insufficiency vary, symptoms generally get worse as the pulmonary insufficiency gets worse. Patients present with a spectrum of complaints, from palpitations to a general decline in function. Late symptoms include exertional dyspnea, palpitations, right heart failure, and syncope.¹⁷

Signs of right ventricular failure can include elevated jugular venous pressure, peripheral edema, hepatomegaly, ascites,³³ and jugular venous distention with a large a wave.

Heart murmurs

Pulmonary insufficiency causes a low-pitched, brief diastolic murmur. Although often present, it may be short or difficult to hear, even if the regurgitation is severe, because this is “low-pressure” pulmonary insufficiency as opposed to the regurgitation that can occur in patients with pulmonary hypertension. Therefore, this murmur is often missed on physical examination.

There may be an ejection click due to a dilated aorta. An aortic insufficiency murmur may also be present.

A right ventricular outflow murmur is generally audible, along with a pansystolic murmur if a residual ventricular septal defect is also present.

A right-sided aortic arch, present in about 25% of patients with tetralogy of Fallot, may cause a lift below the right sternoclavicular junction.¹⁷

Electrocardiographic findings

Electrocardiography commonly shows right ventricular hypertrophy with a right bundle branch block. The longer the QRS duration, the greater the right ventricular volume and mass. Furthermore, a QRS duration greater than 180 ms is a significant marker of risk of ventricular arrhythmias and sudden death.^22,34–37

Another feature strongly associated with ventricular arrhythmias and sudden death is the rate of change in the QRS duration. A relatively rapid increase (> 3.5 ms/year) is associated with a significantly higher risk.¹ A rapid rate of change may be meaningful even if the QRS duration is not markedly prolonged.¹¹

Reduced heart rate variability also appears to be a marker of risk of sudden cardiac death in these patients.^38,39

Chest radiography typically shows a prominent right ventricular shadow and cardiomegaly.¹⁷

Many centers specializing in congenital heart disease therefore recommend baseline cardiac MRI, even for patients without symptoms.³³

PULMONARY VALVE REPLACEMENT IS THE ONLY PROVEN TREATMENT

No study has yet shown that drug therapy alone slows the progression of complications.¹ Pulmonary valve replacement is the only treatment proven to reduce right ventricular size and improve right ventricular function in the long term.

The risks of surgery, including the need for repeat operations, must be balanced against the risk of irreversible right ventricular dysfunction and its associated complications. The operative death rate is low, as is the long-term risk of death afterward. Therrien et al¹² reported that, in a series of 70 patients who underwent pulmonary valve replacement, the probability of survival was 92% at 5 years and 86% at 10 years.

Surgery appears to reverse or at least arrest the progression of many of the complications associated with pulmonary insufficiency, including tricuspid regurgitation and diastolic dysfunction.¹⁷ Its utility in ameliorating ventricular tachycardia, however, remains controversial. One series showed a lower prevalence of tachycardia after pulmonary valve replacement (9% after surgery vs 22% before), but later studies have had more equivocal results.¹⁷

When should surgery be done?

There is little controversy about the eventual need for pulmonary valve replacement in most patients. What is controversial is the timing.^12,44–47

This issue has been hotly debated. Some believe that pulmonary valve replacement should be done only if evidence of right ventricular dysfunction has developed.¹⁷ Others suggest that it be considered earlier and that the onset of symptoms may be a late and suboptimal indication for it.^6,8,48,49 Many experts now recommend surgery early, before symptoms of heart failure develop.¹⁷ Though surgery has traditionally been recommended if the QRS duration is longer than 180 ms, some believe it should be done before this occurs.¹¹

Arguments for early surgery. In one study, in no patient who had a right ventricular end-diastolic volume greater than 170 mL/m² (normal ≤ 108) or a right ventricular end-systolic volume greater than 85 mL/m² (normal ≤ 47) did these numbers return to normal after pulmonary valve replacement.^45,50 This suggests a point of irreversible dilatation and a volume threshold beyond which right ventricular function is unlikely to completely improve. Normalization of right ventricular volumes was shown to occur when pulmonary valve replacement was performed before the right ventricular end-diastolic volume reached 160 mL/m² or the right ventricular end-systolic volume reached 82 mL/m².^47,51

Delaying surgery until symptoms occur may be unfavorable because the long-term outcomes of increased right ventricular volumes and decreased right ventricular ejection fractions after surgery are not known.

Arguments for watchful waiting. There does not seem to be a threshold above which right ventricular volumes do not decrease after surgery—although they may not decrease to the normal range. Pulmonary valve replacement substantially reduced right ventricular dilatation even in patients with very high right ventricular volumes and right ventricular dysfunction, and resulted in an overall improvement in function (measured by New York Heart Association class).⁴⁷

Late pulmonary valve replacement rapidly improves right ventricular volumes and improves the effective ejection fraction, although its impact on absolute right ventricular function is not as pronounced. The QRS duration shortened after surgery in those in whom it was 180 ms or longer before surgery, although this appeared to be a transient change.⁵² The prevalence of ventricular tachycardia declined from 22% to 9% and that of atrial fibrillation or flutter declined from 17% to 12%.^17,48

A recent study with long-term follow-up has raised questions about the necessity of aggressive early intervention in tetralogy of Fallot. Sixty-seven patients were followed for as long as 27 years after surgery. Forty-five had severe pulmonary insufficiency and severe right ventricular dilatation, and of those, 28 remained free of symptoms and did not undergo pulmonary valve replacement. The authors found that refraining from pulmonary valve replacement in asymptomatic patients with severe pulmonary insufficiency led to no measurable deterioration in 25 of 28 patients.⁵³

The available data do not support pulmonary valve replacement in young patients with mild or moderate right ventricular dilatation, normal right ventricular systolic function, and no additional risk factors.²⁷

Mechanical vs bioprosthetic replacement valves

Once the decision is made to proceed to surgery, the next step is choosing the type of prosthetic valve.

Mechanical valves pose a risk of thrombosis, requiring life-long anticoagulation. To give warfarin (Coumadin) to younger, active people exposes them to the risk of potentially catastrophic bleeding if trauma were to occur. Women who become pregnant are generally at an increased risk of thrombotic complications due to the hypercoagulable state of pregnancy, but the risk of fetal defects is considerable if they receive warfarin.^54–56

Bioprosthetic valves generally come in two varieties: preserved and treated human tissue (homografts) and animal tissue (bovine pericardial or porcine, depending on the size required). These can be implanted as isolated valves or as part of a conduit (valve and surrounding tissue).

Bioprosthetic valves eliminate the need for anticoagulation. However, they are not very durable, especially in younger patients, which is worrisome. An estimated 45% of bioprosthetic valves fail by 10 years,⁵⁷ thus nearly guaranteeing that an otherwise healthy 40-year-old, for example, will need to undergo at least one repeat surgery, and very likely more.

NOVEL THERAPIES

Percutaneous valve replacement

The future of pulmonary valve replacement may lie in percutaneous procedures.

The Melody valve is now approved through a humanitarian device exemption (ie, based on demonstrated safety without proven efficacy) for patients who have a prior pulmonary conduit now complicated by either stenosis or regurgitation.

If percutaneous pulmonary valve replacement proves to have reasonable long-term durability, it has the potential to dramatically shift the balance toward earlier intervention.

Pulmonary vasodilator drugs

Our group is examining whether pharmacologic therapy can alter the clinical outcome in patients with pulmonary insufficiency (due to either tetralogy of Fallot or valvotomy done to treat remote pulmonary stenosis). Specifically, we are using MRI to examine the effects of inhaled nitric oxide, a selective pulmonary vasodilator. Preliminary results suggest that such a strategy may work, and we are designing a trial to examine the longer-term benefit of using an oral drug with similar properties.

Fortunately, the 2009 pandemic of influenza A (H1N1) was less severe than some earlier pandemics, in part thanks to advances in our ability to diagnose influenza, to treat it, and to quickly activate the public health and industry infrastructures to mitigate such a pandemic.

In this article, we present lessons learned from the 2009 pandemic, which may allow clinicians to better prepare for the upcoming influenza seasons.

FOUR PANDEMICS IN THE LAST 100 YEARS

Influenza causes annual epidemics of varied severity and risk of death. In the United States, these seasonal epidemics are estimated to account for more than 200,000 hospitalizations¹ and 1.4 to 16.7 deaths per 100,000 persons (3,349 to 48,614 deaths) each year, mostly in the elderly.²

The past 100 years have seen four influenza pandemics^3,4: H1N1 in 1918, H2N2 in 1957, H3N2 in 1962, and H1N1 in 2009. With each pandemic came a spike in hospitalization and death rates in addition to a higher proportion of deaths in people under the age of 65,³ although the relative impact varied widely with the different viruses.^3,5

After the 1918, 1957, and 1962 pandemics, the rates of hospitalization and death decreased, although still varying from year to year, and the pattern of who developed serious disease returned to normal, with the very young, those with underlying medical conditions, pregnant women, and those age 65 and older being at risk.^3,5,6 Whether the situation in the current postpandemic period will evolve similarly remains uncertain; however, it is believed that the 2009 H1N1 virus will continue to circulate among other established viruses in the community.

THE 2009 PANDEMIC H1N1 VIRUS CAME FROM PIGS, NOT BIRDS

In the late winter and early spring of 2009, H1N1, a novel strain of influenza A, was recognized to have caused outbreaks of respiratory illness in Mexico and southern California. ^7,8 The virus spread rapidly, and with the aid of global air travel it reached nearly every country in the world within several weeks.^4,9

The virus was of swine origin, having six genes of North American swine virus lineage and two genes of Eurasian swine virus lineage. ¹⁰ Although classic teaching suggested that pandemics were caused by “new” viruses, typically of avian origin,¹¹ antigen mapping has clearly shown that swine viruses are antigenically significantly divergent from human viruses,¹⁴ but are more adapted than avian viruses for human transmission.^10,12,13

Little antigenic drift has occurred since the beginning of the outbreak. Nearly all isolates seen to date are antigenically similar to the A/California/7/2009 strain that was selected for pandemic influenza vaccines worldwide and that is now included in the vaccine for seasonal influenza for 2010–2011.^4,6,15

The virus appears to replicate more efficiently in the lungs and lower airways than seasonal H1N1 and H3N2 viruses, but generally lacks many of the mutations associated with greater pathogenicity in other influenza viruses.^4,10,16

PANDEMIC H1N1 DISPROPORTIONATELY AFFECTED THE YOUNG

Most infections caused by the 2009 influenza A (H1N1) pandemic virus were acute and self-limited, similar to seasonal influenza.⁴ Asymptomatic infection has been demonstrated from serologic surveys.^17,18

Notably, many older people had preexisting antibodies that cross-reacted with the novel 2009 pandemic virus, which is antigenically related to but highly divergent from the 1918 pandemic H1N1 virus.¹⁴ This phenomenon may explain why older people were relatively protected against contracting the virus, while younger people, who lacked these antibodies, were more likely to be infected.

A number of studies, using various methods, suggest that each person infected goes on to infect 1.3 to 1.7 other people, a rate called the basic reproduction number or R0. This rate is comparable to that for seasonal influenza and is higher in more crowded settings.^4,19 Seroprevalence studies suggest that there was significant geographic variability in the proportion of the population affected during the first and second waves of the pandemic.^4,20,21

Risk factors for complications or severe illness include age younger than 5 years, pregnancy, morbid obesity, and chronic medical conditions. Interestingly, although people 65 years of age and older had the lowest rate of infection, they had high case-fatality rates if they became sick.^4,22–25 However, in up to 50% of patients with severe disease, no conventional risk factor could be identified.^4,22

Hospitalization rates varied widely by country but were generally highest in those under the age of 5; 9% to 31% of hospitalized patients required intensive care, and 14% to 46% of those receiving intensive care died.⁴

Overall, the case-fatality rate was less than 0.5%, but ranged from 0.0004% to 1.47%.⁴ The lowest case-fatality rates were in Japan, where early diagnosis and treatment are credited, in large part, for such exceptional outcomes.²⁶

The incubation period of pandemic H1N1 influenza is 1.5 to 3 days but may be as long as 7 days.⁴ This virus causes a spectrum of clinical syndromes that range from afebrile upper respiratory illness to fulminant viral pneumonia. ⁴ As with seasonal influenza, most patients present with fever, sore throat, and cough. Gastrointestinal symptoms including nausea, vomiting, and diarrhea are more common than with seasonal influenza.^4,27,28

The viral kinetics of H1N1 are similar to those of seasonal influenza in ambulatory patients, although some reports suggest that the duration of viral shedding may be slightly longer.²⁸

Most patients who needed to be hospitalized presented late after symptom onset with viral pneumonia, which was sometimes may be accompanied by severe hypoxemia, acute respiratory distress syndrome, shock, and renal failure.^29,30 Viral loads were very high in those needing intensive care, and virus shedding longer than 5 days, particularly in the lower respiratory tract, was documented despite antiviral therapy.²⁹ Fewer patients were hospitalized for other indications, including exacerbation of underlying medical conditions (especially asthma or chronic obstructive pulmonary disease) and bacterial pneumonia, which might be explained by the different profiles of patients with pandemic vs seasonal influenza.^4,31–33

In severe cases, a number of laboratory abnormalities were common at presentation, including lymphopenia and elevations in levels of serum aminotransferases, lactate dehydrogenase, creatine kinase, and creatinine.⁴

SEASONAL INFLUENZA: USUALLY ACUTE AND SELF-LIMITED

Most seasonal influenza infections are acute and self-limited. Risk factors for complications or severe illness include age 2 years or younger, age 65 years or older, pregnancy, and chronic medical conditions.^5,30,34

Secondary bacterial infections occur at a rate similar to that during the pandemic.^4,19 The prevalence of bacterial superinfection is about 5% to 15%, depending on the virus, the local prevalence of bacterial pathogens, and the tests used to diagnose the infections.

Hospitalization rates in the United States average 0.052% but range from 0.0115% for ages 5 to 49 to 0.773% for ages 85 and older. ³⁵ Death rates range from year to year from 0.0014% to 0.0167%.² Indications for hospital admission include viral pneumonia, bacterial pneumonia, and exacerbation of underlying medical conditions, especially asthma or chronic obstructive pulmonary disease. Exacerbation of underlying lung disease appears to be a more common indication for admission in patients with seasonal infection than with pandemic infection.^5,30–34

CLINICAL DIAGNOSIS OF INFLUENZA IS UNRELIABLE

Clinical diagnosis of influenza is unreliable, particularly in patients requiring hospitalization. ³⁶ The wide clinical spectrum of influenza infection overlaps with those of other common respiratory viral or bacterial infections. In hospitalized patients, the diagnosis is further confounded by underlying conditions, immunosuppression, and extrapulmonary complications.

Thus, up to half of cases may go unrecognized. ^31,33,36 Clinicians should consider influenza as a potential cause of or contributor to any hospitalization whenever influenza is circulating in the community (ie, during seasonal peaks or pandemics).

Diagnostic tests

Several diagnostic assays are commonly used.^37,38

Rapid antigen tests generally have low sensitivity, in the range of 50% to 60%, particularly for the 2009 A (H1N1) virus. Therefore, a negative test result does not exclude infection and should be interpreted with caution. Newer technologies are being developed that may improve the diagnostic yield of these assays.^4,37–39

Immunofluorescence antigen tests, when performed on nasopharyngeal aspirates or on flocked swabs, are very sensitive for seasonal influenza. However, their sensitivity is lower for 2009 H1N1 influenza.⁴⁰

In general, the sensitivity of antigen assays depends on where the specimen is collected (nose, throat, or lower respiratory tract—eg, tracheal aspirates, bronchoalveolar lavage), the collection method (conventional vs flocked swabs, nasopharyngeal aspirate and wash, bronchoalveolar lavage), the assay type, the virus, and the viral burden at the time of testing (the longer the time, the lower the viral load).^40,41

Viral culture is 100% specific and more sensitive than antigen assays, but it takes 2 to 3 days to run, limiting its usefulness in guiding patient management.

Polymerase chain reaction (PCR) is highly sensitive and specific and, where available, is now the test of choice.⁴⁰ In addition, it can be performed on a wide range of specimens, and subtype-specific PCR assays may provide immediate information on virus subtypes, which may have therapeutic implications. Expanded assays can detect a wider range of pathogens, such as respiratory syncytial virus, although these assays are typically used in selected patients, such as those requiring intensive care or those who are immunocompromised.

Consider sampling the lower airway

In patients with 2009 H1N1 viral pneumonia, up to 19% may have had negative upper respiratory tract samples but detectable virus in the lower airways. Therefore, obtaining a lower respiratory tract specimen for testing should be considered, whenever possible, in cases of suspected influenza pneumonia.^4,42,43

Similarly, when monitoring clearance of the virus in cases of influenza pneumonia, clinicians should remember that the upper respiratory tract may become negative earlier than the lower airways. Active viral replication may continue in the lungs despite apparent clearance in the upper airways.^29,43,44

Relapsed disease and viral replication have been documented when antiviral drugs are discontinued early, even when upper tract shedding is no longer measurable.^29,45,46 Nonetheless, no study has compared the risk of transmission in patients who remain PCR- or culture-positive for a prolonged time. In theory, those who are culture-positive could transmit infection. Clinicians should consult with local infection-control clinicians to determine the duration of isolation for individual patients.

DRUG THERAPY FOR INFLUENZA INFECTION

Antiviral drugs that are active against influenza are:

• The neuraminidase inhibitors oseltamivir (Tamiflu), zanamivir (Relenza), and peramivir (commercially available only in Japan and South Korea)⁴⁷
• The adamantanes amantadine (Symmetrel) and rimantadine (Flumadine)⁴⁸
• Ribavirin (Rebetol).⁴⁹

The neuraminidase inhibitors and adamantanes are generally well tolerated. These classes of drugs have been reviewed extensively elsewhere.^47,48 The oral agents may be challenging to administer to patients who cannot swallow and in those with critical illness or gastrointestinal dysfunction. Some studies have shown reasonable absorption of oseltamivir given by nasogastric tube in critically ill patients.⁵⁰

Inhaled zanamivir, taken via a proprietary “Diskhaler” device, requires the patient to inspire deeply and may induce bronchospasm, which could be problematic in those with underlying airway diseases such as chronic obstructive pulmonary disease or asthma.⁴⁷ Nebulization of the commercially available preparation has been reported to cause ventilator dysfunction and even death, so this should not be done.⁵¹

Several large prospective studies in ambulatory adult and pediatric patients have clearly shown that antiviral therapy can reduce the duration of symptomatic illness due to influenza by up to 2 days if started within 48 hours of symptom onset.^47,52 In fact, the earlier these drugs are started, the better the clinical outcome. ⁴⁷ Further, starting antiviral therapy early is associated with lower rates of hospitalization, death, and complications requiring antibiotics.⁴⁷ Recent data from Japan also suggest that such early therapy may be partially responsible for the low death rate in that country during the recent pandemic.²⁶

Given the evidence of efficacy, antiviral drugs should be considered in all patients with risk factors for severe disease. Antiviral drugs are also appropriate in patients without specific risk factors because of the risk of progression to severe disease in these patients, especially in the context of pandemic H1N1 influenza.^38,53,54 Further, therapy is associated with symptomatic improvement and reduced infectious complications even in patients without risk factors for severe disease.⁵⁵ Such early therapy may also have a positive impact on secondary infections among contacts.⁵⁵

Antiviral therapy recommended in hospitalized patients with influenza

Clinical studies of the treatment of hospitalized influenza patients are limited, with few prospectively conducted studies. Because of differences in clinical course and viral kinetics in hospitalized patients and emerging data in these patients, the ambulatory treatment data and paradigms likely do not apply to hospitalized adults.^{29,43,44,56–59}