Screening for pancreatic and lung cancers


In this edition of “How I will treat my next patient,” I examine the U.S. Preventive Services Task Force Reaffirmation Recommendation Statement regarding screening for pancreatic cancer in normal-risk populations. I also review newly published information regarding low-dose CT screening (LDCT) for lung cancer in a commonly screened population – individuals with chronic obstructive pulmonary disease (COPD). Both publications highlight the complexity of implementing shared decision making in clinicians’ efforts to find these highly lethal cancers in their earliest, most curable stages.

Dr. Alan P. Lyss has been a community-based medical oncologist and clinical researcher for more than 35 years, practicing in St. Louis.

Dr. Alan P. Lyss

Pancreatic cancer screening

In their recommendation, the USPSTF considered data relevant to the benefits and harms (exclusive of costs) of screening for pancreatic cancer in the 85%-90% of individuals who are at normal risk because they lack a known familial or genetic syndrome and do not have at least two affected relatives or one first-degree affected relative (JAMA. 2019;322[5]:438-44).

After reviewing 13 cohort studies employing image-based technologies (CT, MRI, endoscopic ultrasound) and biomarkers for screening, the USPSTF reaffirmed its 2004 recommendation against pancreatic cancer screening. They found no new evidence of sufficient strength and quality to alter their previous “D grade” for screening (i.e., “Don’t do it.”). There were at least moderate harms of screening and subsequent treatment in normal-risk populations. These recommendations apply to asymptomatic individuals with new-onset diabetes mellitus, smokers, older adults, obese patients, and patients with a history of chronic pancreatitis.

What this means in practice

In the nicely written and comprehensive recommendation statement and in the two accompanying editorials (JAMA Surg. 2019 Aug 6. doi: 10.1001/jamasurg.2019.2832; JAMA. 2019;322[5]:407-8), the authors were explicit that the “D grade” for pancreatic cancer screening did not apply to individuals from familial pancreatic cancer kindreds and those with germline mutations and Lynch syndrome mismatch repair genes. For them, the relative risk of pancreatic cancer (greater than 5%) may justify the morbidity of available surveillance technologies, especially since U.S. and International screening studies in these high-risk individuals have generated data suggesting a benefit for treatment of screen-detected cancers.

The USPSTF and the editorial authors were strongly supportive of, and enthusiastic about, ongoing research efforts. Recently, a joint effort at the National Institutes of Health began recruiting centers for a study to assess the sensitivity of novel biomarkers in detecting pancreatic cancer among adults with new-onset diabetes (A211701).

To me, it is clear that the pathway to identifying effective screening for pancreatic cancer – which is forecast to become the second leading cause of cancer death in the United States by 2020 – will focus on high-risk populations first, enabling accurate determination of sensitivity and specificity before being applied to the general population. This is as it should be.

Lung cancer screening

Currently, guidelines from the National Comprehensive Cancer Network recommend LDCT screening annually for high-risk smokers, former smokers, and individuals with additional risk factors aged 55-77 years. The National Lung Screening Trial indicated that LDCT screening achieved a 20% relative reduction in lung cancer mortality and 6.7% relative reduction in overall mortality in a similar population. NCCN guidelines stress the importance of shared decision making and include a table of risks and benefits that should be considered.


Next Article: