A five antibody IHC test for histotype subclassification of non-small cell lung carcinoma.

Poster at  ASCO-NCI-EORTC Annual Meeting on Molecular Markers in Cancer, Oct 2008

R. Seitz¹, R. A. Beck¹, W. Shasteen¹, B. Z. Ring², D. T. Ross², A. Soltermann³, S. Arbogast³, F. Robert⁴, M. T. Schreeder⁵

¹ Applied Genomics, Inc, Huntsville, AL and ² Burlingame, CA, ³ Universitätsspital Zürich, Zürich, Switzerland, ⁴ University of Alabama Birmingham, Birmingham, AL,  ⁵ Clearview Cancer Center, Huntsville, AL

Abstract

Background: Non-small cell lung cancer (NSCLC) can be histologically subclassified into three main tumor entities: adenocarcinoma, squamous cell and large cell carcinomas. The clinical importance of histologic classification has become more significant with the advent of novel treatments targeting NSCLC patients.

Methods: We used a 551 patient NSCLC retrospective cohort from a regional hospital to assess the association of a large number of IHC markers with histotype. Five of these markers targeting TRIM29, CEACAM5, SLC7A5, MUC1, and CK5/6 were combined into one test using a weighted algorithm. We then validated the association of this prospectively defined classifier with histotype in three independent cohorts. We compare our results to a classifier using TTF1 and P63.

Results: The five antibody test was able to robustly distinguish adenocarcinoma from squamous cell carcinomas in all the test cohorts. On average across the test cohorts only 14% of adenocarcinoma patients were misclassified as potentially squamous cell carcinoma (3% classified as squamous, 11% unclassifiable). In comparison the two antibody test misclassified 37% of adenocarcinoma patients (3% classified as squamous, 35% unclassifiable). Each test was similarly strong in correctly classifying squamous cell carcinomas (5 antibody test: 97% properly classified, 2 antibody test: 98%).

Conclusions: We validated a five antibody IHC classifier of NSCLC tumor histotypes on three independent cohorts. The need for a uniform and accepted method of classifying lung carcinomas is growing as therapies come to target different tumor subtypes. The results of this study suggest that prospectively defined clinical trials combining this classifier with these therapies are warranted.

	(2 Mb PDF)