METHODS: Data for all non-metastatic prostate cancer cases diagnosed 1993-2002 and followed to the end of 2007 were analysed. The outcome was progression to metastatic disease, identified by metastatic episode notifications in the CCR or by prostate cancer death. Factors associated with metastatic disease progression were identified using Cox regression models.
RESULTS: Of the 32,643 men with non-metastatic prostate cancer at diagnosis 43.1% had localised disease, 5.1% had regional spread and 51.9% had unknown stage. After a median of 6.8 years of follow-up 6708 cases (20.6%) had developed distant metastases. The risk of developing metastatic disease was significantly higher for those with regional (adjusted HR=2.65, 95% CI: 2.40-2.93) or unknown initial stage (adjusted HR=1.70, 95% CI: 1.61-1.80), for older men (65-74 years: HR=1.43, 95% CI: 1.33-1.53; >74 years: HR=2.73, 95% CI: 2.55-2.93), and those living in inner regional (HR=1.11, 95% CI: 1.04-1.18) or rural areas (HR=1.24, 95% CI: 1.14-1.36) or more disadvantaged areas (middle tertile: HR=1.09, 95% CI: 1.02-1.16; most disadvantaged: HR=1.12, 95% CI: 1.04-1.19). The risk of developing metastatic disease decreased over calendar time (adjusted HR=0.98, 95% CI: 0.97-0.99 per year).
CONCLUSIONS: After a median follow-up of 6.8 years more than 1 in 5 men diagnosed with non-metastatic prostate cancer developed distant metastases. This estimate of the overall risk of developing metastatic disease in the population, and the geographical disparities identified, can inform the planning of required cancer services.