Diagnostic sub-classification of non-small cell lung cancer: Importance in clinical therapeutics and prognostication

M. J. Donovan, A. Kotsianti, A. Colomer, M. Verdú, M. Clayton, H. Pang, S. Hamann, C. Cordón-Cardó, X. Puig.

Background: The association of EGFR mutations with bronchoalveolar carcinoma (BAC) / non-small cell lung cancer (NSCLC) and the therapeutic and prognostic variability reported for lung cancer has reinforced the need for more accurate NSCLC sub-classification. The present study utilizes clinical data, immunohistochemistry (IHC), molecular analyses, and quantified immunofluorescent multiplexing (QIFM) to develop a systems pathology model for NSCLC. Methods: 73 NSCLC cases; 64 males / 9 females were compiled (including whole sections) for Tissue Micro Array (TMA) generation: 34 adenocarcinoma (ACA), 27 squamous cell carcinoma (SCC), 3 adenosquamous carcinoma (ASC), 5 large cell undifferentiated carcinoma (LCC), and 1 BAC. Diagnostic assessment, for both whole sections and TMA, included histology and IHC with a panel of antibodies: EGFR (clones E30 and 31G7), cytokeratins (CK7, 20 and 5/6), and thyroid transcription factor (TTF1). In addition, EGFR mutation analysis was performed utilizing previously published primers. Data from QIFM with CK 18, pKDR and pERK was also collected. Results: Twenty out of 70 cases (29%) required re-categorization, 19 reclassified according to the IHC profiles and one based on morphology. EGFR IHC with clone 31G7 was more robust than E30, SCC tumors exhibiting a higher score (2-3+ overall) compared to AC. EGFR mutation in exon 19 was identified in one BAC tumor sample (female) out of 31 cases evaluated. Two tumor samples (ASC and AC) contained intronic point mutations. Conclusion: The integration of clinical data, IHC profiles, EGFR status and additional marker studies (QIFM) has generated an improved model for NSCLC classification. This further demonstrates the benefit of a systems pathology approach, integrating biomarker analysis, clinical data, molecular assays and image based IF, to refine diagnostic pathology as it is performed today. Individualized patient management requires a more comprehensive molecular profile of the actual tumor specimen in order to select appropriate therapeutics and define relevant parameters of prognosis.

Abstract presentado en: