Clinical significance of both tumor and stromal expression of components of the IL-1 and TNF-α signaling pathways in prostate cancer

IL-1 and TNF-alpha, the two major proinflammatory cytokines, have been involved in initiation and progression of several malignancies. They could influence the biological behavior of prostatic tumors and patient outcome, and could be useful as prognostic factors. This study evaluated the prognostic capability for biochemical progression after radical prostatectomy of expression of IL-1, TNF-alpha and related signaling components, in the tumor and surrounding stroma, as well as its correlation with other clinicopathological features. Expression of IL-1 alpha, IL-1 beta, IL-1Ra, IL-1RI, IL-1RII, IRAK-1, TRAF6, TNF-alpha, TNFRI and TRAF2 was analyzed by immunohistochemistry in radical prostatectomy samples from 93 prostate cancer patients. Spearman's test, Kaplan-Meier curves, and univariate and multivariate Cox proportional hazard regression analyses were performed. Expression of TNF-alpha, TNFRI, TRAF2, ILRI, IRAK-I and TRAF6 correlated with at least one clinicopathological feature (clinical T stage, pathological T stage, preoperative serum PSA or Gleason score). Increased tumor expression of TNF-alpha, TNFRI and IL-1RI, and reduced tumor expression of IRAK-1 were significantly correlated with a poor prognosis in univariate analysis. Reduced stromal expression of IL-1 beta and IL-1RII, and increased stromal expression of IRAK-1 were also adverse prognostic factors in univariate analysis. Remarkably, tumor IL-1 beta and stromal IL-1RII and IRAK-1 remained as independent prognostic factors after adjustment for preoperative serum PSA, pathological T stage and Gleason score in multivariate Cox models. Our results suggest that prostatic expression of TNF-alpha, IL-1 beta and related signaling proteins (TNFRI, IL-1RI, IL-1RII and IRAK-1) predicts clinical outcome in prostate cancer, and support the involvement of TNF-alpha and IL-1 beta signaling in prostate cancer progression. (C) 2013 Elsevier Ltd. All rights reserved.