Khor LY et al. (2007) COX-2 expression predicts prostate-cancer outcome: analysis of data from the RTOG 92-02 trial. Lancet Oncol 8: 912–920

Cyclo-oxygenase-2 (COX-2) is overexpressed in some types of human cancer, including prostate cancer, and the overexpression of COX-2 might also be involved in tumor resistance to chemotherapy and radiation therapy. Khor and colleagues analyzed data from the Radiation Therapy Oncology Group (RTOG) 92-02 trial to identify the association between COX-2 and prostate cancer outcomes.

The study population included 586 patients (median age 70 years, range 43–88 years) with locally advanced (T2c–T4) prostate cancer who were randomly assigned to receive either long-term (28 months; n = 316) or short-term (4 months; n = 270) neoadjuvant, concurrent and adjuvant androgen deprivation therapy (ADT) plus external-beam radiation therapy. Tissue samples, obtained from either transurethral resection or needle-core biopsy, were analyzed for COX-2 staining intensity.

On univariate analysis, increased COX-2 staining intensity independently predicted distant metastasis (hazard ratio [HR] 1.181, P = 0.0004), biochemical failure (for both the ASTRO definition [HR 1.073, P = 0.008] and Phoenix definition [HR 1.073, P = 0.014]), and any failure (defined as “a first event from ASTRO biochemical failure, local failure, distant metastasis, or cause-specific mortality”; HR 1.068, P = 0.011). COX-2 overexpression seemed to be most predictive of outcomes in patients who underwent short-term ADT.

The authors conclude that COX-2 staining intensity could be used to select patients who might benefit from long-term ADT rather than short-term ADT. Furthermore, the authors suggest that COX-2 inhibitors might improve patient response to radiation therapy, with or without ADT.