Intrinsic markers of tumour hypoxia and angiogenesis in localised prostate cancer and outcome of radical treatment: a retrospective analysis of two randomised radiotherapy trials and one surgical cohort study.

BACKGROUND: Expression of intrinsic markers of tumour hypoxia and angiogenesis are important predictors of radiotherapeutic, and possibly surgical, outcome in several cancers. Extent of tumour hypoxia in localised prostate cancer is comparable to that in other cancers, but few data exist on the association of extent of tumour hypoxia with treatment outcome. We aimed to study the predictive value of intrinsic markers of tumour hypoxia and angiogenesis in localised prostate cancer, both in patients treated with radiotherapy and in those treated surgically.
METHODS: We applied a new, needle biopsy tissue microarray (TMA) technique to study diagnostic samples from men with localised, previously untreated prostate cancer treated in two randomised controlled trials of radiotherapy-dose escalation. Multivariate analysis by Cox proportional hazards was done to assess the association between clinical outcome, in terms of biochemical control, and immunohistochemical staining of hypoxia inducible factor-1 alpha (HIF-1 alpha), vascular endothelial growth factor (VEGF), and osteopontin expression. The analysis was repeated on an independent series of men with localised, previously untreated prostate cancer treated by radical prostatectomy. The main outcome was time to biochemical (ie, prostate-specific antigen [PSA]) failure.
FINDINGS: Between Oct 12, 1995, and Feb 5, 2002, 308 patients were identified from two prospective, randomised trials at the Royal Marsden Hospital, London and Sutton, UK, for the radiotherapy cohort and diagnostic biopsies were available for 201 of these patients. Between June 6, 1995, and Nov 4, 2005, 329 patients were identified from the Aarhus University Hospital, Skejby, Denmark, for the prostatectomy cohort; of these, 40 patients were excluded because the tumour was too small to sample (19 patients), because the paraffin block was too thin (19 patients), or because the blocks were missing (two patients), leaving 289 patients for analysis. For patients treated with radiotherapy, increased staining for VEGF (p=0.008) and HIF-1 alpha (p=0.02) expression, but not increased osteopontin expression (p=0.978), were significant predictors of a shorter time to biochemical failure on multivariate analysis, independent of clinical tumour stage, Gleason score, serum PSA concentration, and dose of radiotherapy. For patients treated with surgery, increased staining for VEGF (p<0.0001) and HIF-1 alpha (p<0.0001) expression, and increased osteopontin expression (p=0.0005) were each significantly associated with a shorter time to biochemical failure on multivariate analysis, independent of pathological tumour stage, Gleason score, serum PSA concentration, and margin status.
INTERPRETATION: To our knowledge, this is the largest study of intrinsic markers of hypoxia and angiogenesis in relation to the outcome of radical treatment of localised prostate cancer. Increased expression of VEGF, HIF-1 alpha, and, for patients treated with surgery, osteopontin, identifies patients at high risk of biochemical failure who would be suitable for enrolment into trials of treatment intensification.