ΔNp63 (p40) distribution inside lung cancer: a driver biomarker approach to tumor characterization

ΔNp63 (henceforth simply p40) is a squamous/basal-type biomarker corresponding to nontransactivating (non-TA) isoforms of p63 gene. Its prospective relevance as driver biomarker in lung cancer has not yet been thoroughly investigated. In all, 72 adenocarcinomas (ADs), 27 squamous cell carcinomas (SQCs), 13 pleomorphic carcinomas (PLCs), 10 small-cell lung carcinomas (SCLCs), 5 large-cell neuroendocrine carcinomas (LCNECs), 5 adenosquamous carcinomas (ADSQCs), 3 large-cell carcinomas with basaloid features (B-LCC), 2 carcinoids, 2 carcinosarcomas (CS), 2 salivary-gland type tumors (SGTTs) of the lung, and 5 pleural malignant epithelioid mesotheliomas (MEMs) were prospectively diagnosed by morphology and verified by immunohistochemistry for p40, p63, and thyroid transcription factor 1 (TTF1). Histological scores (HS) were devised by multiplying the percentage of immunoreactive cells (0 to 100%) by immunostaining intensity (low = 1 vs strong = 2, according to internal controls). There was a nonrandom distribution of p40 across the diverse tumor groups and cell differentiation lineages, with p40-HS > 100 closely paralleling squamous or myoepithelial carcinomas (SQC, B-LCC, SQC-containing PLC, ADSQC with predominant SQC, SGTT), and p40-HS ≤ 10 pinpointing AD, AD-containing PLC, or CS and neuroendocrine (NE) tumors. At variance, p63-HS was significantly higher than p40 in AD (P < .0001) and NE tumors (P = .0156), with positive predictive value being 83% and 95% and overall accuracy being 95% and 99%, respectively. Also, TTF1 was shared by gland-differentiated and NE tumors. MEM cases were always negative for all biomarkers. The HS-guided assessment of p40 allowed an effective orientation among thoracic malignancies at the level of individual tumor patients thereby contributing to prospectively realize a driver, holistic approach to cancer characterization.