Home Chest Diseases Lung Cancer Lung Cancer Screening - Raising the Threshold
Lung Cancer Screening - Raising the Threshold PDF Print E-mail
Written by Leong Oon Keong   
Wednesday, 27 February 2013 01:12

Article from Ann Intern Med. 2013;158: 246-252, 289-290.

When screening patients for lung cancer with low-dose computed tomography (CT), raising the nodule size threshold from 5 mm to 7 or 8 mm before starting a more intensive workup could result in fewer false-positive results.

This conclusion, from a study published in the February 19 issue of the Annals of Internal Medicine, should be prospectively evaluated, say the authors, but an editorialist points out that such a study would be "extremely expensive."

"The definition of a positive result is an important part of a screening program. As technology improves, it needs to be reevaluated," lead author Claudia I. Henschke, PhD, MD, from the Mount Sinai School of Medicine in New York City, told Medscape Medical News.

In 1993, any nodule identified with low-dose CT underwent a workup. By 2000, only those larger than 5 mm underwent a workup. "Technology has advanced, and we've looked at this threshold again," Dr. Henschke explained.

Raising the Threshold

Dr. Henschke and her colleagues from the International Early Lung Cancer Action Program (I-ELCAP) retrospectively analyzed how raising the threshold for workup would affect the frequency of positive results among the 21,136 ELCAP participants who had baseline CT performed from 2006 to 2010.

A positive result was defined as any parenchymal, solid, or part-solid noncalcified nodule that met the threshold.

Table. Percent of ELCAP Participants With Positive Results on CT Screening By Threshold

Threshold Positive Screening Result, %
5 mm 16.0
6 mm 10.2
7 mm 7.1
8 mm 5.1
9 mm 4.0

 



Subsequent workups were reduced by 36% with a 6 mm instead of the 5 mm threshold, by 56% with a 7 mm threshold, by 68% with an 8 mm threshold, and by 75% with a 9 mm threshold.

The analysis showed that lung cancer diagnoses would have been delayed by 9 months at most.

"Of course, to delay any diagnosis in lung cancer is significant, but those cancers that were found because of the workup still needed 1 or 2 CT scans to determine that they were really cancers," Dr. Henschke explained. "All of them were still in stage I when they were diagnosed," she reported.

The definition of a positive result always needs to strike a balance between the very large number of workups for benign lesions and the possibility of a delay in the diagnosis of a small number of cancers, she noted.

"That's really the key message. Each screening program or society or whoever comes up with a decision of what that cut-off is has to think about that balance," Dr. Henschke said. Cost effectiveness and the anxiety created also need to be considered. "We are now suggesting that we should prospectively evaluate the cut-off size, just like we did in the early 2000s for the 5 mm threshold," she asserted.

Very Expensive Proposition

However, re-evaluating this threshold in a prospective study would be extremely expensive, counters Stephen Lam, MD, chair of the Lung Tumor Group at the British Columbia Cancer Agency and professor of medicine at the University of British Columbia in Vancouver, Canada, in an accompanying editorial.

"You really need to have a large sample size if you want to randomly look at different thresholds. This is the problem," Dr. Lam told Medscape Medical News.

This study, although provocative, looks only at the size of the nodule, Dr. Lam points out. "Size is important, but it is not the only criteria that would determine whether a nodule is cancerous," he writes.

"Dr. Henschke and others have shown that semisolid nodules have a higher probability of cancer than solid nodules, so simply applying the size criteria to different types of nodules may not accurately reflect the risk for cancer," he said. It has been suggested that "other variables, such as the presence or absence of emphysema or nodule speculation...predict cancer," he explained.

Screening for lung cancer with CT scans has been gaining acceptance, and is now recommended by a number of professional bodies, including the National Comprehensive Cancer Network. These recommendations are based on positive results from the National Lung Screening Trial (NLST), which involved more than 50,000 people who were randomized to undergo either low-dose CT or a chest x-ray. The trial showed a 20% reduction in lung cancer mortality with CT screening, compared with chest x-ray, and a 7% reduction in all-cause mortality.

"The NLST used 4 mm as the cut-point to determine whether you need to have additional investigation before the next annual repeat CT. Dr. Henschke and colleagues want to change that to 7 or 8 mm," Dr. Lam noted. If you have a higher threshold and only investigate people with larger nodules, the number of false-positive results would be much lower, but you would also miss cancers that are smaller, he explained.

"To evaluate whether to take 4 mm or 8 mm, you need to have some kind of a randomized study to ensure that the results are the same as those found with the NLST. It cost $250 million for the NCI to do that study, so it is a very expensive proposition," Dr. Lam noted.

The study was funded in part by the Flight Attendant Medical Research Institute and by the American Legacy Foundation. Dr. Henschke reports financial relationships with the Flight Attendant Medical Research Institute and the American Legacy Foundation. Dr. Lam has disclosed no relevant financial relationships.

Last Updated on Friday, 27 December 2013 07:17