Starting a new discussion about screening for lung cancer.

Abstract

Screening fordisease is conceptually simple: identifydiseases inan early stage increasing the likelihoodof cure.However, the trade-offs involved, including false-positive test results and diagnosing clinicallyunimportant lesions,make thedecision toscreencomplex.Perhaps that iswhyphysiciansandpatients find itdifficult to resolve the trade-offs inherent in screening programs and to find the appropriateplaceofscreening inprevention.Publicdiscussionsof thesetradeoffs in screening for breast, cervical, and prostate cancers have increased awareness of harms and uncertainty about benefits and are informing the decisionmaking about less intensive screening. Now, for lung cancer, it is time to embark on a new discussion of both the trade-offs and the appropriate role of screening when alternate approaches to reducing the societal disease burden are available. Although earlier studies suggested that chest radiography screening for lung cancerwas not effective, the 2011 National Lung ScreeningTrial (NLST) foundthat screeninghigher-riskpatientswith 3annual low-dosecomputedtomography(LDCT)scansreduced lung cancermortality by 16% (risk ratio, 0.84; 95%CI, 0.75-0.95; absolute risk reduction from21per 1000to18per 1000over6.5years).1,2 The NLST participants were generally healthy adults aged 55 to 74 yearswith a 30-pack-year history of cigarette smoking.1,2 These results have spawned multiple recommendations for lung cancer screening. Relying on models from the National Cancer Institute’s Cancer InterventionandSurveillanceModelingNetwork, theUSPreventive Services Task Force (USPSTF) recommended more intensive screening for adults with more than a 30-pack-year history of smoking: they should undergo LDCTevery year fromages 55 to80 years rather than only 3 total annual screens for people aged 55 to 74 years in NLST.3 Before screening can be recommended, the trade-offs betweenthebenefitsof screeningwithharmsandcosts shouldbeconsidered.TheharmsofLDCTscreeningareonlypartially known.4The NLSTfoundahigh rateof false-positivescreening tests:39.1%ofparticipants in the LDCT group had at least 1 false-positive scan, almost all requiring some diagnostic evaluation. Although most of theseevaluations involvedonly further imaging, thenumberofstudy participantswithout lungcancerundergoing invasiveprocedures far outnumbered the number whose lives were extended by screening.Manyof theLDCTparticipantswithpositive screening testshad indeterminatediagnostic evaluations, resulting inextendedsurveilJAMA INTERNALMEDICINE