Introduction & Objectives
Since the introduction of PSA in clinical practice, prostate cancer (PCa) has faced a stage migration towards more localized disease. However, in patients with a long life expectancy with high-risk PCa (HRPCa) the risk to die from PCa is not negligible and may change significantly according to the time elapsed from surgery. The aim of the study was to evaluate long term survival patterns of young patients treated with radical prostatectomy (RP) for HRPCa.
Material & Methods
We evaluated data of 7,650 patients treated with RP at 14 tertiary care centers between 1987 and 2013 for HRPCa defined as at least one of the following adverse characteristics: PSA>20, cT3 or higher, biopsy Gleason sum 8-10. For the purpose of these analyses, only patients aged 60 years or less were included (n=612). First, we estimated cancer specific survival (CSS) and overall survival (OS) rates using the Kaplan-Meier method. Subsequently, univariable and multi-variable competing risk analyses were employed to identify predictors of cancer specific (CSM) and other cause mortality (OCM). Covariates consisted of age at surgery, Charlson comorbidity index (0 vs. >=1), year of surgery, pathological Gleason score (GS; 6 vs. 7 vs. 8-10), surgical margins status (SM), lymph-node invasion (LNI), pT stage (pT2 vs. pT3a vs. pT3b-pT4) and adjuvant radiotherapy. Third, competing risks Poisson regression methodology was used to assess CSM and OCM rates at 5 and 10 years after RP. Finally, the same analyses were repeated in order to assess the probability of surviving additional 5 years in patients who survived 5, 8 and 10 years after RP.
Results
Mean age at diagnosis was 56 years (median 57 years; range 39-60). Mean follow-up time was 90 months (median 89). Overall, 57 (9.3%) and 37 (6.0%) patients succumbed to CSM and OCM, respectively. Within the overall cohort, CSS rates at 5, 10 and 15 years were 93.9, 87.0 and 82.2, respectively. The OS rates at 5, 10, and 15 years were 91.0, 82.1 and 69.6%, respectively. At multi-variable competing risks analyses predicting CSM, year of surgery, GS, pT stage, SM status and LNI emerged as significant predictors of PCa death, after adjusting for other cause mortality (all p≤0.02). Conversely, none of the covariates was significantly associated with OCM (all p≥0.1). Within the overall population, the baseline 5 and 10 years competing risk CSM and OCM rates were 5.9 and 12.5% vs. 3 and 5.4%, respectively. Among those patients who survived 5 and 8 years after surgery, the probability of succumbing to PCa within the next 5 years prevailed that of experiencing OCM. Specifically, the 5-years CSM and OCM rates given a 5 and 8 years of survivor ship after RP were 7.3 and 6.7% vs. 2.6 and 5.8%, respectively. Conversely, when a 10-years survivor ship after RP was achieved, OCM became the main cause of death during the next 5 years (9.9 vs. 5.3 for CSM).
Conclusions
In young patients with high risk PCa PCa represents the main cause of death during the first 10 years after RP. Even though no predictors of OCM were identified at MVA, mortality not related to PCa becomes the main cause of death after 10 years of survivor ship. Accordingly young patients with HRPCa should be strictly followed-up for the first 10 years after RP. However, a comorbidity profile reassessment should be suggested after 10 years from RP in order to better stratify patient prognosis
Funding: No financial aid was received for this research.