CT Screening Cuts Lung Cancer Mortality; Raises Policy Questions

Jun 30, 2011

Bookmark and Share  | 

Final results from the National Lung Screening Trial show a significant reduction in lung cancer mortality with the use of annual low-dose CT screening, compared with standard chest x-rays among former heavy smokers at high risk for lung cancer.

Low-dose CT screening led to a relative reduction of 20% in the rate of death from lung cancer, according to findings released online by the New England Journal of Medicine on June 29 (doi: 10.1056/NEJMoa1102873).The number needed to screen with low-dose CT to prevent one death from lung cancer was 320.

Although preliminary study results were announced in November 2010, the article by the National Lung Screening Trial (NLST) research team marks the first time that the results appear in a peer-reviewed journal. Acknowledging that the earlier announcement has led to calls for lung cancer screening, the authors urge rigorous analysis of cost-effectiveness before public policy recommendations are made.

"The reduction in lung-cancer mortality must be weighed against the harms from positive screening results and overdiagnosis, as well as the costs," they wrote.

In the study, 53,454 men and women aged 55-74 years – who were current or former smokers with a smoking history of at least 30 pack-years – were recruited at 33 U.S. medical centers. A total of 26,722 participants were randomized to receive three annual screens with low-dose helical CT; 26,732 were randomized to three annual screens using chest x-ray. The two groups were virtually identical in demographics and smoking history.

In all three screening rounds, positive screening tests were substantially more common in the low-dose CT group than in the radiography group (27.3% vs. 9.2% in the first round; 27.9% vs. 6.2% in the second; and 16.8% vs. 5% in the third). All told, 39.1% of the CT group and 16% of the radiography group had at least one positive result.

The percentage of screening tests that identified a clinically significant abnormality -- other than an abnormality suspicious for lung cancer – also was more than three times as high in the low-dose CT group as in the radiography group (7.5% vs. 2.1%).

More than 90% of positive screenings in the first round of the study led to a diagnostic evaluation, though the follow-up rates were lower in the later rounds. Diagnostic evaluation most often consisted of additional imaging with invasive procedures being performed infrequently.

Across the three screenings, most of the positive results were false positives – 96.4% in the CT group and 94.5% in the radiography group. Of the total number of low-dose CT screening tests, 24.2% were classified as positive and 23.4% had false-positive results; of the total number of radiographic screening tests, 6.9% were classified as positive and 6.5% were false-positive results.

In all, 1,060 lung cancers were diagnosed in the low-dose CT group (645/100,000 person-years) vs. 941 in the radiography group (572/100,000 person-years). Of these cancers, 649 in the low-dose CT group were diagnosed after a positive screening test and 44 were diagnosed after a negative screening test. In the radiography group, 279 cancers were diagnosed after a positive screening test and 137 were diagnosed after a negative screening test.

In both groups, the remaining cases were among participants who missed screening or were diagnosed after their trial screening phase was over

Analysis of lung cancer-specific mortality showed that in the CT group 356 lung cancer deaths occurred after 144,103 person-years; in the radiography group 443 lung cancer deaths occurred after 143,368 person-years. This corresponded to 247 and 309 lung cancer deaths, respectively, per 100,000 person-years in the CT and radiography groups.

There were 1,877 and 2,000 deaths from all causes in the CT and radiography groups, respectively, "representing a significant reduction with low-dose CT screening of 6.7% ... in the rate of death from any cause," the investigators wrote. While lung cancer accounted for 24.1% of all the deaths in the trial, 60.3% of the excess deaths in the radiography group were due to lung cancer.

The authors concluded that "although some agencies and organizations are contemplating the establishment of lung-cancer screening recommendations on the basis of the findings of the NLST, the current NLST data alone are, in our opinion, insufficient to fully inform such important decisions."

They noted that "The observation that low-dose CT screening can reduce the rate of death from lung cancer has generated many questions." Among these they listed: Will populations with different risk profiles benefit from screening? Could less frequent screening programs be equally effective? Would the use of different criteria for a positive screening result translate to similar benefit? For how long should people be screened?

In an accompanying editorial, Dr. Harold C. Sox, professor of medicine at the Dartmouth Institute in Hanover, N. H., agreed with the investigators’ reservations. In particular, the cost effectiveness of low-dose CT screening for lung cancer must be analyzed, he said: "Policy makers should wait for cost-effectiveness analyses to determine the amount of overdiagnosis in the NLST, and, perhaps identification of biologic markers of cancers that do not progress."

In addition, "it may be possible to define subgroups of smokers who are at higher or lower risk for lung cancer and tailor the screening strategy accordingly," he said. "The findings of the NLST regarding lung-cancer mortality signal the beginning of the end of one era of research on lung-cancer screening and the start of another. The focus will shift to informing the difficult patient-centered and policy decisions that are yet to come."

Dr. Sox also noted that "overdiagnosis is a problem because predicting which early-stage cancers will not progress is in an early stage of development, so that everyone with screen-detected cancer receives treatment that some do not need," he wrote in an accompanying editorial (doi: 10.1056/NEJME1103776).

All but two of the NLST study authors reported that they have no relevant financial relationships. Jonathan D. Clap reported having financial interest in Human Genome Sciences. Constantine Gatsonis, Ph.D., is a consultant for Wilex AG, Mela Sciences, and Endocyte Inc., has received speaker fees from Bayer Health and payment for education development by the Radiologic Society of North America. He also has invested in the Vanguard Health Fund. Dr. Sox had no conflicts.