Differences in prostate cancer survival between Denmark, Iceland, and Sweden: A high-resolution study.

Abstract

159 Background: Register-based studies have demonstrated large differences in relative survival and excess mortality following the diagnosis of PC in the Nordic countries. These differences may reflect differences in patient characteristics, diagnostics (PSA-use) and treatment strategies. This high-resolution study explores the background for the differences in PC survival between Denmark, Iceland and Sweden. Methods: Patients with newly diagnosed PC in 1997 were identified through population-based national cancer registers in Denmark and Iceland. Information on clinical findings was retrieved by reviewing hospital files. In Sweden information was gathered from two regional population-based prostate cancer registers providing both time of diagnosis and clinical information. Country specific excess mortality rates were compared adjusting for available information on known prognostic factors. Results: The overall relative survival in the cohorts was comparable to population-based results previously published. Across countries significant differences in excess mortality rates were seen. These differences were largely explained by differences in patient characteristics at diagnosis, and when adjusting for differences in patient characteristics i.e. metastatic / non-metastatic disease, clinical T-stage, and PSA level at diagnosis, the differences in excess mortality diminished or disappeared. The difference in percentage of patients with metastatic disease at diagnosis was the one factor responsible for the major differences in mortality-rates across countries. Conclusions: Register-based studies on relative survival and excess mortality following a PC diagnosis may be influenced by national differences in clinical presentation at diagnosis. Differences between countries in the proportion of patients with metastatic spread at diagnosis apparently explains most of the difference in relative survival previously reported. Further studies and cross country comparisons of survival and excess mortality for PC should adjust for differences in patient characteristics, mainly TNM.