Original Article: Lung Cancer Screening: Pros and Cons [Lown Right Care]
Issue Date: June 15, 2019
To the Editor: We read with interest the article by Drs. Lazris and Roth. They report that no other study besides the National Lung Screening Trial (NLST) has shown a benefit from lung cancer screening. Two other randomized trials (Multicentric Italian Lung Detection [MILD], Dutch-Belgian Randomized Lung Cancer Screening Trial) have reported lung cancer mortality reductions greater than were found in the NLST.1,2 A third trial (German Lung Cancer Screening Intervention Trial) demonstrated the same finding in women. The greater reduction in lung cancer mortality found in these studies may be because of the longer screening protocols compared with the 24 months in the NLST. The MILD trial continued screening for up to 10 years and showed the largest reduction in lung cancer (39%) and overall mortality (20%).1
The authors highlight important risks including false-positive findings and overdiagnosis. Compared with the NLST, the widely implemented American College of Radiology Lung Imaging Reporting and Data System (Lung-RADS) has lowered false-positive rates to 10% to 12% on the initial screening examination and 5% on subsequent scans.3 Using the original Lung-RADS system would have reduced the number of false-positive findings in the NLST by more than 50% and corresponding invasive procedures by 23%. The Veterans Health Administration study cited in the Lazris and Roth article used liberal criteria for positive examinations (i.e., any nodule larger than 2 mm), which contributed to the high false-positive rate.4
Long-term follow-up in the NLST suggests that overdiagnosis does not occur other than for carcinoma in situ. Under the current standard of care, these patients return to annual surveillance with no further interventions unless the lesion grows aggressively.5 The recent update to Lung-RADS increased the size threshold for such lesions from 20 mm to 30 mm, further reducing the risk of overdiagnosis.5
Real-world evidence highlights the effectiveness of well-organized screening programs. A survey of 165 lung cancer screening programs in the United States, most in community settings, demonstrated similar outcomes as those in academic centers.6 Lung cancer screening can successfully take place in the real world and is increasingly being performed, with a survey of 10 states finding that 14.4% of eligible patients had a low-dose computed tomography scan within the past 12 months.
Lung cancer screening is effective for reducing the number one cause of cancer deaths for men and women. The time has come to make lung cancer screening part of routine primary care for patients who are at risk.
In Reply: We appreciate the letter from Drs. Lim, Kitts, and Tremblay. After our article was sent to press, several new studies were published that further illuminated the potential benefits of annual screening. However, these new studies do not fundamentally alter our conclusion that there are both risks and benefits to low-dose computed tomography screening, and persistent uncertainty around how effective screening is for different populations.
The authors cite a 39% reduction in lung cancer death among participants screened for 10 years in the MILD trial. Although technically correct, relative numbers (39%) do not help physicians and patients make shared decisions because these numbers do not tell us the absolute benefit of screening. In the MILD trial, 1.7% of participants died of lung cancer over 10 years in the screened arm, and 2.3% died in the control arm, leading to a reduction of six lung cancer deaths out of 1,000 people screened. This is similar to the NLST findings, showing that long-term screening can slightly improve benefits, with persistently high false-positive rates. The German Lung Cancer Screening Intervention Trial did show reduced lung cancer mortality in women (seven out of 1,000 fewer lung cancer deaths after five years of screening), but no survival advantage for men. Finally, Lung-RADS does reduce the false-positive rate; however, the authors noted a decrease in sensitivity of lung cancer detection, which may lead to a lower survival advantage of screening.1
There are still many questions left unanswered. Why do men have improved survival in some studies and not others? Do survival advantages vary based on other characteristics? The Centers for Medicare and Medicaid Services is organizing a real-world survey to assess the prevalence, survival advantage, and risks of screening in the community, and many physicians and organizations are waiting for those data before pursuing an organized screening program. Until then, the results from the NLST and the newer trials do provide some information that will help physicians and patients reach a decision as to whether to pursue lung cancer screening.