Indolent tumors on lung CT an "important potential harm"

By Megan Brooks

NEW YORK (Reuters Health) - Nearly one in five lung cancers detected by low-dose computed tomography (LDCT) may be indolent, resulting in overdiagnosis - and patients should be told this, researchers advise.

In addition to detecting aggressive lung tumors and potentially saving lives, LDCT screening also detects slow-growing or clinically insignificant tumors that otherwise may not cause clinical symptoms.

These cases of overdiagnosis represent "an important potential harm of screening because they incur additional cost, anxiety, and morbidity associated with cancer treatment," Dr. Edward F. Patz, of Duke University Medical Center in Durham, North Carolina and colleagues say.

They estimated the excess number of lung cancers detected with LDCT (vs chest radiography) among more than 53,000 patients at high risk for lung cancer who participated in the National Lung Cancer Screening Trial (NLST).

Among 1,089 lung cancers reported in the LDCT group during 6.4 years follow-up, the investigators estimated that 18.5% were indolent (representing overdiagnosis).

They also estimated that 22.5% of non-small cell lung cancer detected by LDCT represented an overdiagnosis, and that 78.9% of bronchioalveolar lung cancers seen on LDCT represented an overdiagnosis.

"These data provide critical information about one of the limitations of CT screening for lung cancer," Dr. Patz told Reuters Health by email. "While I do not believe that clinical practice will or should change right now, it is important to understand these findings, so that the risks and benefits of screening can be fully disclosed, and more efficient screening strategies identified."

The study is online today in JAMA Internal Medicine.

A related study on lung cancer screening suggests a "lack of uniformity" in CT screening practices, particularly technical aspects of screening.

Dr. Phillip Boiselle, Department of Radiology, Beth Israel Deaconess Medical Center in Boston and colleagues surveyed 15 thoracic radiology division chiefs at leading US academic medical centers.

They learned that most sites had similar patient selection and referral policies. Eleven of 15 active screening sites (73%) use the NLST entry criteria, one site uses expanded selection criteria, and, "interestingly," the remaining three offer screening to any patient who has participated in shared decision making with their physician.

For nodule management, 10 sites (67%) use the Fleischner Society guidelines, two (13%) use the NCCN guidelines, and one (7%) uses NLST guidelines. The remaining 2 sites use a hybrid approach. All but one site (93%) include a smoking cessation program, which is mandatory at three sites for current smokers.

"In my opinion, it is noteworthy that most leading academic medical centers had CT screening programs in place and that almost all programs incorporated smoking cessation for current smokers. This is significant because smoking cessation is the single most important step that an active smoker can take to reduce his or her chance of developing lung cancer," Dr. Boiselle told Reuters Health by email.

Most of the sites (13, 87%) scanned one to five patients each week. At five sites (33%), the estimated dose used was less than 1 mSv, 1 to 2 mSv at seven sites (47%), and 2 to 3 mSv at two sites (13%). One respondent didn't know the dose.

There was "less uniformity among active CT screening programs for some of the technical and logistical aspects of screening," Dr. Boiselle said, "and this may be explained by the fact these aspects of screening have received relatively less attention in most published clinical practice guidelines to date."

Overall, the findings highlight the need for formalized radiology guidelines for CT screening for lung cancer, Dr. Boiselle and colleagues say in their Research Letter. Based on their findings, they say these guidelines should place "primary emphasis" on the technical and logistical aspects of screening that are not covered by current guidelines.

"Importantly, forthcoming joint guidelines for screening from the American College of Radiology and the Society of Thoracic Radiology will provide necessary guidance in this area once they are finalized and disseminated this spring," Dr. Boiselle said.

The authors of a separate report in JAMA Internal Medicine today propose a coherent framework, or taxonomy, for thinking about the potential physical, psychological and financial harms of lung cancer screening.

"Harms can occur at any step of the screening cascade. The taxonomy provides a systematic way to conceptualize harms as experienced by patients," Dr. Russell Harris of the Center for Health Services Research, University of North Carolina, Chapel Hill and colleagues say.

They add, "Our purpose in developing a taxonomy of the potential harms of screening is to help investigators, policy makers, clinicians, and the public think more clearly and systematically about harms and to consider harms equally with benefits in decisions about screening. We do not assert that harms always outweigh benefits, only that it is always necessary to weigh the two. We consider this a draft taxonomy, a work in progress that could contribute to our current public discussion about screening. In the end, we hope that a widely agreed-on taxonomy will eventually lead to more balanced decision making about the wisdom of screening."

SOURCES: http://bit.ly/18j0sR3, http://bit.ly/1gkug2J, and http://bit.ly/1hIWJ2v

JAMA Intern Med 2013.

(c) Copyright Thomson Reuters 2013. Click For Restrictions - http://about.reuters.com/fulllegal.asp