Immunohistochemistry to enhance prognostic allocation and guide decision-making of patients with advanced urothelial cancer receiving first-line chemotherapy

Background Knowledge of the expression of molecular drivers and potentially druggable targets might enhance prognostic classification of M UC. Materials and Methods We analyzed archival tissue from patients with UC who underwent first-line chemotherapy for locally advanced (LA) and M disease between the years 2000 and 2013. The following biomarkers were evaluated using IHC: excision repair cross complementation (ERCC) group 1 (ERCC1), epidermal growth factor receptor (EGFR), HER2, VEGFR-3, PDGFRα, p53, and p63. Expression of ERCC1, EGFR, and HER2 was dichotomized as positive (2+, 3+) or negative (≤ 1+). Cox regression models were used to evaluate the association of biomarker expression with progression-free (PFS) and overall survival (OS), after controlling for known prognostic factors. Results Since June of 2009, tissues of 88 cases (27 LA, 61 M) were stained. Rates of positive IHC/number evaluable were as follows: ERCC1: 30 of 66 (45%); HER2: 24 of 52 (46%); EGFR: 31 of 54 (57%); VEGFR-3: 50 of 66 (76%); PDGFRα: 10 of 63 (16%); p53: 25 of 56 (45%); and p63: 46 of 53 (87%). In the multivariable model, PDGFRα was significantly prognostic for poorer PFS (hazard ratio [HR], 2.53; 95% confidence interval [CI], 1.01-6.37; P =.047) and trended to significance for poorer OS (HR, 2.66; 95% CI, 0.96-7.42; P =.060) and VEGFR-3 was significantly prognostic for better PFS (HR, 0.33; 95% CI, 0.15-0.74; P =.007) and OS (HR, 0.36; 95% CI, 0.15-0.85; P =.019). The c-index of the model was 0.67 and 0.68 for the 2 end points, respectively. Conclusion Tumor VEGFR-3 and PDGFRα expression appeared to confer a divergent prognostic effect. These data underscore the hurdles in defining the role of angiogenesis as a molecular driver and therapeutic target, and the controversial role of IHC to guide therapeutic decision-making.