A new randomized study has confirmed that many deaths from lung cancer could be prevented by continuing and augmenting the global rollout of computed tomography (CT) screening for long-term heavy smokers who are at high risk of the disease.
Details of the Nederlands-Leuvens Longkanker Screenings Onderzoek (NELSON) population-based, randomized, controlled trial are published in the New England Journal of Medicine in an article that begins by reminding readers that lung cancer is the leading cause of death from cancer worldwide, accounting for 18.4 percent of all cancer deaths and causing more deaths than breast, colorectal, and cervical cancers combined (2020;382:503-513).
“At 10 years of follow-up, we saw a 24 percent difference in the lung cancer mortality reduction [among men] in the screening arm as compared to the control arm,” said first author of the study, Harry de Koning, MD, PhD, Professor of Public Health at Erasmus MC University Medical Centre in Rotterdam, the Netherlands. “And in the female subgroups, we found a 33 percent reduction at 10 years. But in the years before that: years 7, 8 and 9, we found an even larger reduction that was statistically significant in the females,” he told Oncology Times.
Although smoking is decreasing among adults in western countries, its popularity is still growing globally and many youngsters—even in the west—are being recruited to it. But only 15 percent of patients with lung cancer are alive 5 years after diagnosis because their disease is detected too late. The NELSON study confirms and reinforces findings from the US-based National Lung Screening Trial (NLST) that cases can be found early, and lives saved, by using CT screening.
“We randomized about 14,000 smokers and ex-smokers between two arms: screening or no screening. The screening arm got four CT scans,” said de Koning. “You then try to find cancer earlier in the screening group, treat them earlier, and then basically hope that this is more effective.” To get proof of a benefit from screening, the researchers followed patients for about 10 years and looked for lung cancer deaths in both arms, he said.
NELSON Study Findings
The NELSON study randomized 13,195 men in the primary analysis and 2,594 women in the subgroup analysis (all of them between the ages of 50 and 74) either to CT screening on recruitment to the study and additionally after 1, 3, and 5.5 years or to no screening. Data on cancer diagnosis and the date and cause of death were taken from national registries in the Netherlands and Belgium, and a committee confirmed lung cancer as the cause of death when possible. The minimum follow-up was 10 years.
Ninety percent of the men on average adhered to CT screening. And one in 10 of them had at least one extra scan in addition to the scheduled scans because of borderline findings. Two percent of patients were then referred for “suspicious nodules.” At 10 years of follow-up, there had been 5.58 cases of lung cancer per 1,000 person-years in the screening group and 4.91 cases per 1,000 person-years in the control group. Lung cancer mortality was 2.50 deaths per 1,000 person-years in screened patients, but 3.30 deaths per 1,000 person-years in the unscreened control arm in the male patients. This amounted to a risk ratio for death from lung cancer at 10 years of 0.76 (P=0.01) in favor of the screening group as compared with the control group,
In men, this was similar to the values at years 8 and 9. Among women, the ratio was 0.67 at 10 years of follow-up, with values of 0.41 to 0.52 in years 7 through 9.
The NELSON investigators concluded that lung cancer mortality had been significantly lower among those who underwent volume CT screening than those who were not screened.
Koning said the new data from NELSON were welcome because the volumetric CT method they used had yielded more precise prediction of lung cancer, with fewer referrals of patients who turned out to have no malignancy than the previous U.S. study which used nodule diameter as the metric.
“There really is a difference actually. The point is: the American trial used a certain cutoff. Everyone above the 4 millimeters was found positive and referred. But that led to more referrals. We looked at the volumes—in a three-dimensional way—so we only referred 2 percent for further investigation,” he noted. “Volumetric measurements are simply much more accurate. So, to see a difference in the growth rate with a volume (versus the diameter) is a really big difference. We think the quality is simply better.”
Commenting on the NELSON findings, David Baldwin, MD, Consultant Respiratory Physician and Honorary Professor of Medicine at the University of Nottingham, and the Nottingham University Hospitals, told Oncology Times that the issue with screening in general, and lung cancer screening specifically, was that “you needed to do a randomized controlled trial of the screening tests versus either an alternative that is established or nothing at all.”
He said there had been concern that, although the one very large study from the U.S. published in 2011 had indeed shown the benefit of low-dose CT screening over the control arm (which was chest X-ray), the worry had been that chest X-ray had not been the established standard. And, indeed, the earlier trials of chest X-ray (added to other tests like sputum analysis) had not shown any benefit from such screening at all.
The choice of chest X-ray as control had even reduced the apparent benefit of screening revealed by the powerful NLST study. “At the time that they published the results, it looked as if the chest X-ray arm had a reduction in overall mortality of about 9 percent. So, it was actually taking off the difference between the chest X-ray and the CT,” Baldwin noted.
“And then, furthermore, the inclusion of patients who had a very low risk of lung cancer in the American trial—and during that trial period had almost no benefit from the screening—meant that the results were diluted. So, we were all saying that actually we can expect a much bigger effect,” he added.
This interpretation of the NLST findings had already prompted the English National Health Service to launch a pioneering European initiative by implementing CT screening for lung cancer. The NELSON study findings now strongly endorsed this.
Baldwin pointed out that the American trial in 2011 had been the first to show there had been a significant advantage from using CT screening. “But what we haven't had is another trial. People are very skeptical about using just one single trial to drive a potentially very effective, but also very expensive, screening program. So, we were waiting for this second trial to see what the final results were going to be. And I'm glad to say that they were acceptable.”
He was also impressed by the added value of the NELSON study screening schedule. “Because they screened for 5 ½ years, as opposed to the American trial where they screened just for 2 years, it is almost three times a difference in the length of time people are screened. [This] resulted in pretty much three times the number of deaths that were prevented in the trial. So very, very, encouraging: showing the power of screening for longer and also the power of using the volume approach for managing pulmonary nodules,” he said.
From Baldwin's perspective, the most important contribution NELSON had added to the picture was to provide a published document. “We can now quote to policymakers that this has confirmed that CT scanning for lung cancer works.”
He was pleased to add that the new results demonstrated that CT screening worked better than had been shown in the American trial: “Substantially better.”
Although the U.S. had already approved screening for lung cancer, there had been issues about delivery of the screening and public participation. “One of the very significant challenges is getting patients into these programs. And part of that is about giving them information about how effective the program [is],” Baldwin said, noting there were a number of factors that stopped people from entering programs.
“One of them in the U.S. is the requirement that has been stipulated for a detailed shared decision-making process that occurs before the consent to the CT scanning. And this has certainly been identified as being quite difficult for patients to go through, and reducing their participation rates,” he noted.
“We are not seeing good participation rates in the U.S. at the moment. The additional data from the NELSON, showing just how much more effective CT screening can be, will help that process to some extent. Globally, many countries have gone ahead with screen programs,” Baldwin stated, which was a process that could now be galvanized by the new data.
De Koning said the logical next step was to push ahead with implementing CT screening globally, following the lead from countries like the U.S. and the U.K. “I think most of the European countries have really been waiting for this second trial to show something, and to be the trial with a low referral rate. So, we have started a project of implementation this year. But I think many European countries are now about to think about what they can do. And, in the meantime, the U.K. has really stood out with a (sort of) pilot project throughout the country by the NHS. So, the U.K. really is stepping up as the one of the first European countries,” he said.
When he was asked about the cost, which can run to up to around $200 per screen, de Koning said CT could be highly cost-effective. “If you start off with stringent criteria to get the CT scanning selected group, we do think it's cost-effective. So far, we have done analyses for Switzerland and Canada based on the American trial. And because the European trial is even more positive, we do think it can be cost-effective. But it is about the benefit you get. So, it is possible to come up with the program that maybe costs as much as breast cancer screening, but it gets actually more deaths prevented per year.”
Peter M. Goodwin is a contributing writer.