Presentation Authors: Michael Froehner, Ulrike Heberling*, Christoph Heitmann, Samaneh Farahzadi, Marcus Lindner, Angelika Borkowetz, Rainer Koch, Matthias Hübler, Manfred P. Wirth, Dresden, Germany
Introduction: In the Standard Set for localized prostate cancer of the International Consortium for Health Outcomes Measurement (ICHOM), an easily applicable comorbidity index containing 12 unweighed conditions was recommended [Martin et al., Eur Urol 2015]. To our knowledge, this comorbidity index has, however, not yet been validated in patients with prostate cancer.
Methods: In 5050 consecutive patients who underwent radical prostatectomy at our institution between 1992 and 2012 the parameters heart disease, hypertension, peripheral vascular disease, lung disease, diabetes, renal disease, liver disease, neurologic disease, tumor, depression, connective tissue disease and AIDS were recorded. The mean patient age was 65 years and the mean follow-up was 10 years. Univariate and multivariable competing risk analyses were used in order to investigate the prognostic significance of single conditions and of the comorbidity index with non-prostate cancer and overall mortality as endpoints.
Results: Eight out of twelve parameters were significant predictors of mortality both in univariate and multivariable analyses. In a six-sided stratification of the index with overall mortality as endpoint, there was a dose response pattern was seen (hazard ratios 1 - 1.4 -2.3 - 3.8 - 79.0 - >1000). With one exception, the neighboring mortality curves differed significantly from each other. In the subgroup of patients with 6 or more concomitant diseases (n=72, 1.4% of patients) the 10-year non-prostate cancer mortality superseded the 50% level. Only two conditions (heart disease, hazard ratio 1.8, p=0.0011 and peripheral vascular disease, hazard ratio 2.0, p=0.0493) were predictors of non-prostate cancer mortality in patients aged 70 years or older.
Conclusions: The easily applicable unweighed comorbidity index recommended in the ICHOM Standard Set was well suited to predict mortality after radical prostatectomy in the whole sample but was of limited value in patients aged 70 years or older.