Meeting Abstracts
» Pathologic Review of the Mayo Lung Project

(CANCER 95:2361, 2002)
Thomas V. Colby, M.D.
Los Angeles, CA, April 2003

In the Mayo Lung Project screening trial there were more carcinomas identified in the screened group compared with the control group and the patients with cancer in the screened group had a better survival than the control group but there was no difference in lung cancer mortality between the two groups. All available cases were pathologically reviewed to address the questions of pathologic misdiagnosis and overdiagnosis (carcinomas that do not result in the death of the patient) to see if these might explain the discrepancy.

Results: Slides from 105 of the 106 patients who underwent surgery at Mayo Clinic were available for review. The original diagnosis of carcinoma was confirmed in all cases with minor interobserver disagreements as to exact pathologic terms used (among the four pathologists). There was thus no case for misdiagnosis.

Depending on the observer, there were between 8 and 15 cases considered pre-invasive (carcinoma in situ). All but one of these cases occurred in the screened group. Most of these were postsurgical stage I and II and had been resected. It could be suggested that the inclusion of these in situ carcinomas in the screened group represents overdiagnosis since current data suggests these tumors may not be fully malignant in the usual sense of the term implied with a diagnosis of "lung cancer." Statistical analysis of the data after removing these cases however did not fully account for the differences in the number of diagnoses of carcinoma between the screened group and the control group. Nevertheless, it probably was a factor and overdiagnosis could not be ruled out.

This review may have implications for current HRCT screening trials. HRCT screening for lung cancer allows the recognition of small peripheral nodules, some of which may manifest entirely as a ground glass opacity. Excision of these lesions has shown that some represent atypical adenomatous hyperplasia (AAH), some represent bronchioloalveolar carcinoma (BAC), an in situ carcinoma, and some represent invasive cancers; a portion are different lesions entirely. AAH is a presumed precursor of non-mucinous BAC.

While most pathologists readily recognize carcinomas that have demonstrable invasion, the distinction between AAH and BAC is not at all clearcut and the criteria in the literature are vague. One pathologist’s AAH is another’s BAC. Thus if one of these lesions is identified by screening it is unnecessary surgery for a benign lesion according to one pathologist’s interpretation and a screening success and "cure" according to another pathologist’s interpretation.

The AAH-BAC continuum may also have some implications similar to those of in situ squamous carcinomas of the large airways. Just as some in situ carcinomas of the large airways may be slowly evolving and not full-fledged cancers the same may be possible for lesions in the AAH-BAC spectrum. While AAH is the presumed precursor for some BACs, there is no prospective study that proves the risk of progression of an individual lesion or the rate at which such a lesion might progress to BAC. There are anecdotal CT studies that confirms that this certainly does occur (GGO’s get solid foci that correspond to invasion) but it is not clear when this lesion should be considered a "full-fledged cancer."

Thus, I think one could say its déjà vu all over again.

This page last modified on 3 February 2004 by Society Webmaster.