Early detection in lung cancer is strongly linked to better survival. In a landmark decision the CMS recently approved national coverage for lung cancer screening with LDCT for high-risk smokers.
Eli Glatstein, MD
Early detection in lung cancer is strongly linked to better survival.1In a landmark decision the Centers for Medicare & Medicaid Services (CMS) recently approved national coverage for lung cancer screening with low-dose computed tomography (LDCT) for high-risk smokers.
The CMS approval for screening was based on a favorable recommendation by the United States Preventive Services Task Force (USPSTF), an advocacy group.2The USPSTF recommendation, in turn, was based on the successful completion of National Lung Cancer Screening Trial (NLST), which was initiated to review mortality reduction benefits associated with LDCT screening.3Data were collected from 53,454 patients at high-risk for lung cancer between 2002 and 2004, and patients were randomly assigned to either a test group (receiving 3 annual LDCT screenings, 26,722) or a control group (receiving chest radiography, 26,732). Morbidity data collected until December 2009 showed a 20% reduction of mortality in the LDCT group. Simultaneously, a high incidence of false-positives (96.4%) was also observed.
While the reduction in mortality was a huge strength, high incidence of false-positives was a major weakness in the study, along with the arbitrary cut-off for age selection. Conservatives felt the false-positives could trigger a potential cascade of misdiagnosis, overtreatment, and unintended harm to patients.
Mary E. Reid, PhD, a cancer epidemiologist and professor of oncology at Roswell Park Cancer Institute in Buffalo, New York, is not worried about the high false-positive rates at this stage.
“Ninety-five percent of polyps in colon cancer screening are noncancerous, but screening still results in a 50% decrease of mortality rates,” she said. “We all are still relatively new to lung cancer screening. Seventy percent of lung cancer patients report with late-stage cancers that are inoperable. We have never had the opportunity to look at precursor lesions like nodules and understand their molecular progression to tumors. Since the screening now requires submission of data to a national registry, we will be in a position to mine the data to identify differences in lesions,” said Reid.
The current CMS guidelines have 4 broad requirements:
Before CMS guidelines became available, multiple professional society guidelines were provided for identification of lung cancer screening eligibility. Selection of patients largely depended on what guidelines were adopted by the institutions. In 2015, CMS finally approved national screening for lung cancer high-risk populations.A major
advantage of the CMS coverage is that most organizations now follow similar guidelines and will be submitting screening data to a national registry. A major drawback is the strict age cut-off criteria.
Eli Glatstein, MD, professor and vice chair, department of radiation oncology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, said, “At Penn we will continue to do what we have been doing for some time: screen high-risk patients.” Glatstein stressed that these are guidelines and not rules. Strict adherence equates it to a ‘one-size-fits-all’ approach, which is not practical in the real world. All treatment decisions fundamentally are about risk/benefit ratio, and it is ultimately the physician’s job to identify the patients who qualify for screening, according to Glatstein.
Reid also disagrees with the concept of stopping selection at arbitrary age numbers. “We just don’t understand the molecular signatures that differentiate a precancerous nodule from a cancerous one,” she said. Ultimately, the expertise of the radiologist will dictate which path to follow, according to Reid.Strategies are being developed to minimize the 23.5% false-positive rate associated with LDCT Lung Imaging Reporting and Data System (Lung-RADS) is one such lung cancer screening and reporting tool, which was developed by the American College of Radiology.
McKee et al retrospectively reanalyzed NLST data from a single center with Lung-RADS.4. Lung-RADS reduced the overall positive rate from 27.6% to 10.6% and increased the positive predictive value for diagnosed malignancy from 6.9% to 17.3%, supporting its use for screening analysis.
Another retrospective assessment by Pinsky et al also reported a decrease in false-positives using Lung-RADS.5False-positive result rates for Lung-RADS was lower than NLST, 12.8% and 26.6 % at baseline and subsequently 5.3% and 21.8%, respectively. Baseline sensitivity went down, 84.9% and 93.5 % for Lung-RADS and NLST and after baseline, 78.6% and 93.8 %, respectively. The authors, however, cautioned to carefully consider the effect of using Lung-RADs criteria in clinical practice.
“It’s the tiny lesions that are the big problem,” wrote Glatstein. An abnormality <1 cm is followed up with a new scan in 6 months. No change in size after 6 months generally indicates nonmalignancy (false-positive). To reduce the rate of false-positives, clinicians need an additional screening procedures that in combination with CT can enhance the correct detection of abnormalities. That’s where techniques like Lung-RADS need to be further considered. Imaging studies help in distinguishing normal from abnormal, but the smaller the abnormality, the more difficult it is to interpret with any confidence.
Reid added that reduction of false-positives at the expense of sensitivity is not recommended. It is premature at this point to adopt such tools, because there is not enough longitudinal experience to validate the conclusions.
Collectively, the LDCT trial has shown mortality reduction benefit. High false-positives and strict age eligibility standards stand out as weaknesses that can lead to overdiagnosis and unnecessary treatment. The CMS coverage, however, has set the ball rolling to generate data from the screenings that will eventually help clinicians gain in-depth understanding of lung cancer.