Greater understanding of molecular pathways important in cell growth and proliferation of thoracic malignancies, particularly non-small cell lung cancer (NSCLC), has resulted in intense clinical and translational research. There is now considerable interest in personalizing treatment based on an understanding of tumor histology and molecular abnormalities. However, there is a multiplicity of data, often with discordant results resulting in confusion and uncertainty among clinicians.
We conducted a systematic review and a consensus meeting of Canadian lung cancer oncologists and pathologists to make recommendations on the use of biomarkers in NSCLC. PubMed covering 2005 to March 2010 was searched using MESH terms for NSCLC and randomized trials, plus text words for the biomarkers of interest. Conference proceedings from 2005 to 2009 ASCO, ESMO, IASLC, and USCAP were also searched. The articles were reviewed by pairs of oncologists and pathologists to determine eligibility for inclusion.
Ten oncologists and pathologists reviewed and summarized the literature at a meeting attended by 37 individuals. Draft recommendations were formulated and agreed upon by consensus process. There is some evidence that histology is prognostic for survival. There is evidence from multiple randomized clinical trials to recommend the following: histologic subtype is predictive of treatment efficacy and for some agents toxicity. Immunohistochemistry testing should be performed on NSCLC specimens that cannot be classified accurately with conventional H&E staining. As EGFR mutations are predictive of benefit from tyrosine kinase inhibitors, diagnostic NSCLC samples should be routinely tested for EGFR-activating mutations. Clinical data on K-RAS mutations are inconsistent, therefore testing is not recommended. There is insufficient evidence to recommend other biomarker testing. No biomarkers to date reliably predict improved efficacy for anti-VEGF therapy. Routine assessment for EML4/ALK mutations is not recommended at present, although emerging data suggest that it may become valuable in the near future.
Assessment of NSCLC biomarkers is becoming increasingly important. Therefore, adequate diagnostic material must be obtained for accurate histologic subtyping and relevant molecular biology assays.