From the AGA Journals

Improved follow-up needed to find late-stage pancreatic cancers

View on the News

Finding factors may make a difference

Surveillance of individuals at increased risk of pancreatic ductal adenocarcinoma (PDAC) offers an opportunity to improve disease mortality through detection of premalignant lesions and earlier stage PDAC. Emerging data suggest that outcomes in surveillance-detected PDAC are superior to those diagnosed after onset of signs and symptoms. This study by Chhoda et al. highlights a potential quality gap in current surveillance programs, namely the diagnosis of interval cancers and late-stage metastatic PDAC.

Investigators report a cumulative incidence of late-stage PDAC of 1.7 per 1,000 patient-years in surveillance, while the incidence of any advanced neoplasia (high-grade PanIN, high-grade IPMN, NET > 1 cm and any stage PDAC) was 3.3 per 1,000 patient-years. Importantly, late-stage PDAC was defined as T2-4N0-1M0-1 in this study. This is based on the 2013 International Cancer of the Pancreas Screening definition of “success of a screening program” as treatment of T1N0M0 PDAC, which was later updated to include a resected PDAC confined to the pancreas. The cumulative incidence of resectable lesions was 2.2 per 1,000 patient-years, while the incidence of unresectable PDAC was 0.6 per 1,000 patient-years in surveillance. Unfortunately, clinical features were unable to predict the onset of these 11 unresectable PDACs.

Aimee Lucas, MD, Associate Professor of Medicine, Division of Gastroenterology, Icahn School of Medicine at MountSinai, New York

Dr. Aimee Lucas

Given data reporting limitations, it is uncertain how many advanced PDACs were a result of delayed surveillance, diagnostic errors, or other preventable factors. Addressing these contributing factors as well identifying clinical indicators that may improve the efficacy of existing regimens (such as new onset diabetes, worsening in glycemic control in a person with diabetes, weight loss, and incorporation of novel biomarkers) will be critical to optimizing PDAC surveillance outcomes in in high-risk individuals.

Aimee Lucas, MD, is associate professor of medicine, division of gastroenterology, Icahn School of Medicine at Mount Sinai, New York. She reports receiving research support and consulting from Immunovia, who developed a blood-based biomarker for early PDAC detection.



A relatively large number of late-stage pancreatic ductal adenocarcinomas (PDACs) are detected during follow-up surveillance, yet no single patient- or protocol-specific factor appears to be significantly associated with detecting late-stage disease during this period, according to a new systematic literature review and meta-analysis.

The researchers, led by Ankit Chhoda, MD, of Yale University, New Haven, Conn., wrote in Gastroenterology that interval progression in high-risk individuals “highlights the need for improved follow-up methodology with higher accuracy to detect prognostically significant and treatable lesions.”

Individuals at high risk for PDAC are encouraged to undergo routine surveillance for the disease because early detection and resection of T1N0M0 PDAC and high-grade precursors may improve survival outcomes. According to Dr. Chhoda and colleagues, challenges of interval progression of cancers during the surveillance period for gastrointestinal malignancies have been well described in the general and at-risk patient populations. Previous studies, the authors explained, have not scrutinized the issues associated with late-stage PDACs detected during follow-up surveillance.

“Late-stage PDACs necessitate critical appraisal of current follow-up strategies to detect successful targets and perform timely resections,” the authors wrote. The researchers added that the diagnosis of late-stage PDACs during follow-up emphasizes the need for implementing “quality measures to avoid preventable causes, including surveillance adherence and diagnostic errors.”

To understand the incidence rates of late-stage PDACs during follow-up in high-risk individuals, Dr. Chhoda and researchers performed a systematic literature review and meta-analysis of data that included follow-up strategies for early PDAC detection among a high-risk population.

Outcomes of interest for the analysis included the overall diagnosis of advanced neoplasia as well as surveillance-detected/interval late-stage PDACs (T2–4N0M0/metastatic stage PDAC) during follow-up. The investigators defined surveillance-detected and interval late-stage PDACs as late-stage PDACs that were detected during surveillance and as those presenting symptomatically between visits, respectively.

The researchers also performed metaregression of the incidence rates of late-stage PDACs to examine the relationship with clinicoradiologic features in high-risk individuals.

A total of 13 studies on surveillance in 2,169 high-risk individuals were included in the systematic review, while 12 studies were included in the meta-analysis. Across studies, high-risk individuals were followed for over 7,302.72 patient-years for the purposes of detecting incident lesions or progression of preexisting pancreatic abnormalities.

In all high-risk individuals who underwent follow-up, the investigators identified a total yield of advanced neoplasia of 53. This total yield consisted of 7 high-grade pancreatic intraepithelial neoplasms, 7 high-grade intraductal papillary mucinous neoplasms, and 39 PDACs. According to the meta-analysis, the cumulative incidence of advanced neoplasia was 3.3 (95% confidence interval, 0.6-7.4; P < .001) per 1,000 patient-years. During follow-up, the cumulative incidence of surveillance-detected/interval late-stage PDACs was 1.7 per 1,000 patient-years (95% CI, 0.2-4.0; P = .03).

In a separate analysis, the investigators sought to identify the relationship between the modality of follow-up imaging and late-stage PDAC incidence. Imaging modalities used during follow-up were mostly cross-sectional imaging, such as computed tomography or magnetic resonance imaging with cholangiopancreatography (n = 4) or endoscopic ultrasound and cross-sectional modalities (n = 8).

The investigators found no significant associations between late-stage PDACs and surveillance imaging, baseline pancreatic morphology, study location, genetic background, gender, or age. Incidence of late-stage PDACs in studies with mostly cross-sectional imaging was 0.7 per 1,000 patient-years (95% CI, 0.0-8.0). This incidence rate was lower than that reported with EUS and cross-sectional modalities (2.5 per 1,000 patient-years; 95% CI, 0.6-5.4), but this difference was not statistically significant (P = .2).

No significant difference was found during follow-up in the incidence of late-stage PDACs between high-risk individuals with baseline pancreatic abnormalities (0.0 no significant difference; 95% CI, 0.0-0.3) vs. high-risk individuals with normal baseline (0.9 per 1,000 patient-years; 95% CI, 0.0-2.8) (P = .9).

Most studies included in the analysis did not report on diagnostic errors and surveillance adherence, the researchers wrote. Nonadherence to surveillance as well as delays in surveillance accounted for four late-stage PDACs, and surveillance cessation and/or delays were reported in 4 out of 19 high-risk individuals. There was limited information on symptoms, presentation timing, site of lesion, and surveillance adherence, which the investigators indicated prevented a formal meta-analysis.

In their summary, the study authors noted that in clinical practice there is a need for improved quality measures and adherence to surveillance programs to reduce the risk of diagnostic errors. The authors stated that evidence on the impact of these quality measures “on surveillance outcomes will not only improve quality of surveillance practices, but also enrich our communication with patients who undergo surveillance.”

The researchers reported no conflicts of interest with the pharmaceutical industry, and the study did not receive any funding.

Next Article: