Abstract: PD51-03
Sources of Funding: none

Introduction

Estimating the risk of competing mortality is of importance in tailoring optimal individual management strategies in patients with early prostate cancer. Patients selected for radical prostatectomy at an age of 70 years or older, a stricter selection for good risks may influence the prognostic significance of individual risk factors for competing mortality. No generally accepted tool is available in order to predict competing mortality in in this particular population.

Methods

Using proportional hazard models for the subdistribution of competing risks according to Fine and Gray, we studied 2961 consecutive patients treated at our institution between 1992 and 2007 in order to determine which parameters predict competing mortality in patients selected for radical prostatectomy at an age of 70 years or older and compared the prognostic impact of individual parameters with that in their younger counterparts. The mean follow-up was 11.2 years.

Results

Three common diseases (diabetes mellitus, chronic lung disease and other cancer) which predicted competing mortality in younger men were no predictors of competing mortality in men selected for radical prostatectomy at an age of 70 years or older (hazard ratios, HRs, lower than 1). Beside age (HR per year 1.08, p=0.0255), peripheral vascular disease (HR 2.33, p=0.0195), cerebrovascular disease (HR 2.23, p=0.0242), American Society of Anesthesiologists (ASA) physical status class 3 (HR 2.19, p<0.0001), current smoking (HR 2.18, p=0.0098) and lower or unknown level of education (HR 2.07, p=0.0002) were independent predictors of competing mortality in patients aged 70 years or older. With adding one risk point for each of these parameters, the resulting score compared favorably with five conventional comorbidity measures (Akaike information criterion 1425 versus a range between 1452 for the unweighed Charlson score and 1460 for the modified Lee mortality index).

Conclusions

Conclusions: Combining these five conditions in a score might provide a superior comorbidity measure in patients undergoing radical prostatectomy at an age of 70 years or older.

Funding

none