Pancreatic Cancer Screening in High-Risk Individuals

Helen Leask

August 19, 2019

Although the "do not screen" recommendation for pancreatic cancer in the general population continues to hold, experts point to recent data that support surveillance for high-risk individuals.

The "do not screen" recommendation comes from the US Preventive Services Taks Force (USPSTF), which announced recently that the recommendation — issued in 2004 — still stands. The decision, published online August 6 in the Journal of the American Medical Association, was not unexpected because there have been no prospective studies supporting screening for pancreatic cancer in asymptomatic persons since that time.

However, in three accompanying editorials, clinicians and researchers argue the case for screening in high-risk individuals.

All the editorialists make a plea that the USPSTF "do not screen" recommendation should not obscure the great strides that have been made since 2004 in identifying these at-risk individuals who could benefit from surveillance.

People at high risk for pancreatic ductal adenocarcinoma are "outside the purview of the USPSTF," according to the statement's evidence report.

All the data that support pancreatic screening are in regard to high-risk individuals. The USPSTF assessed the 13 prospective studies that were available at the time of their first data-review cutoff of April 27, 2018.

Although the USPSTF maintained its "do not screen" recommendation for asymptomatic adults because of the lack of evidence, it indicated that  "this recommendation does not apply to these high-risk populations."

The USPSTF did not make any recommendations for screening these individuals, but the authors of the three editorials stepped up to fill the gap.

Third Leading Cause of Cancer Death

Pancreatic cancer is currently the third leading cause of cancer death in the United States. It is predicted to move to second place by 2030. Approximately 90% of people with the disease die within 5 years of diagnosis. USPSTF coauthor Chyke Doubeni, MD, MPH, in an interview on the JAMA website, stated, "It really is a lethal disease and one we're concerned about."

Although there is some evidence that early treatment leads to better outcomes, the window of opportunity is small. Because of the location of the pancreas, pancreatic cancer usually presents at an advanced stage. It also progresses rapidly: patients with stage IV pancreatic cancer were only 1.3 years older on average than patients with stage I disease, according to a recent analysis of the National Cancer Institute's Surveillance, Epidemiology and End Results database.

At diagnosis, more than half the patients present with metastatic disease. The only treatment that offers a potential cure is pancreatoduodenectomy, which carries perioperative mortality risks of 1% to 8%, according to the USPSTF evidence review.

In an accompanying JAMA editorial, Aimee Lucas, MD, of Icahn School of Medicine at Mount Sinai, and Fay Kastrinos, MD, MPH, of Columbia University, both in New York City, acknowledge that "it could be anticipated that screening for pancreatic cancer in asymptomatic, average-risk individuals would generate more harm than good." They cite the low prevalence, the potential harms of false positive screening results, and the risks associated with surgery.

However, they argue that individuals who have a fivefold increase in relative risk might benefit from surveillance.

These at-risk individuals include those with a family history of pancreatic cancer (specifically, those for whom two or more blood relatives and at least one first-degree relative were affected); those with germline mutations in ATM, BRCA1, BRCA2, CDKN2A, PALB2, PRSS1, STK11, TP53, and the Lynch-syndrome mismatch repair genes; and patients who have been recently diagnosed with new-onset diabetes (which confers an eightfold greater risk).

Timing of surveillance for such people can be guided by the natural history of the disease, Lucas and Kastrinos suggest. For example, even in families with multiple affected relatives, the mean age at diagnosis is 68 years, "which may inform the starting age for surveillance of high-risk individuals," they write.

In an editorial in JAMA Surgery, Ralph Hruban, MD, of Johns Hopkins School of Medicine, Baltimore, Maryland, and Keith Lillemoe, MD, of Massachussetts General Hospital, Boston, Massachusetts, highlight early results from the Cancer of the Pancreas Screening (CAPS) trial, which were published last year. Hruban was involved in that study.

The CAPS trial followed 354 individuals at high genetic risk for pancreatic cancer for a mean of 5.6 years. In the study, 9 of the 10 pancreatic cancers detected during surveillance were resectable, and 85% of these patients survived 3 years. By contrast, four symptomatic cancers that emerged in patients who were not under surveillance for pancreatic cancer were all unresectable, and only one of these people survived 3 years.

"This is a critical advance because it suggest that down staging, and therefore real benefit, is possible," Hruban and Lillemoe comment.

Lucas and Kastrinos, in their JAMA editorial, also cite the results of CAPS but acknowledge that the study population extended "beyond the task force's population of interest." Although CAPS was not considered by the USPSTF, all 13 cohort studies that were considered were conducted in "persons at high familial risk."

CAPS may have been published too late to be discussed by the USPSTF. The study results were published in September 2018, 6 months after the USPSTF data search cutoff of April 27, 2018, but within the task force's ongoing literature surveillance period, which continued until March 22, 2019. The study is listed in the references of the evidence report but is not mentioned in the body of the report.

In their JAMA editorial, Lillemoe and Hruban emphasize the positive. They describe the "extraordinarily bright" future that surveillance offers for individuals at high risk for pancreatic cancer. They say that, despite the rarity of the disease, at-risk people can now be identified and their risk quantified — and that they can be targeted for screening. This has become possible because of such discoveries as the risk-carrying germline variants and the link with new-onset diabetes, they argue. Molecular drivers of the high-grade precancers and early cancers are also now understood and provide targets for early detection technology, they add.

In the third editorial, published in JAMA Internal Medicine, Hruban and colleagues enlarge on promising technologies that could transform the surveillance landscape if validated. For example, studies are being conducted regarding several highly sensitive liquid biopsies for pancreatic tumor DNA and cancer-related protein markers. They also cite a multicenter study of 221 patients in which a multimarker blood test identified 64% of patients with early pancreatic cancer with high specificity.

These editorialist also note that the International Cancer of the Pancreas Screening Consortium has published guidelines for screening high-risk individuals.

JAMA. Published online August 6, 2019. Full recommendation; Evidence report; Lucas and Kastrinos editorial

JAMA Surg. Published online August 6, 2019. Hruban and Lillemoe editorial

JAMA Intern Med. Published online August 6, 2019. Lennon et al editorial

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.