Background
Men with biochemical recurrence (BCR) after radical prostatectomy (RP) need information on competing risks of mortality to inform prognosis and guide treatment. We sought to quantify the risk of prostate cancer metastasis and mortality (PCSM) and other-cause mortality (OCM) across key clinical predictors.
Methods
We analyzed 1,225 men with BCR after RP from 2001–2017 in the VA SEARCH database. Multivariable competing risks regression was used to identify predictors and quantify cumulative incidence of metastasis, PCSM, and OCM. Recursive partitioning analysis (RPA) was used to identify optimum variable cutpoints for prediction of PCSM and OCM.
Results
Over a median follow up of 5.6 years after BCR(IQR 2.7,9.1), 243 (20%) men died of other causes and 68 (6%) died of prostate cancer. Multivariable competing risks regression showed that high D’Amico tumor risk and PSA doubling time (PSADT) at BCR<9 months were associated with metastasis and PCSM(p≤0.001); 10-year PCSM was 14% and 9% for those with high-risk tumors and PSADT<9 mos, respectively. Advanced age and worse comorbidity were associated with OCM(p≤0.001); 10-year OCM was higher among men ≥70 years with any Charlson comorbidity (1–3+) (40–49%) compared with those with none (20%). RPA identified optimal variable cutpoints for prediction of PCSM and OCM, with 10-year PCSM ranging from 3–59% and 10-year OCM ranging from 17–50% across risk subgroups.
Conclusions
Among men with BCR after RP, there is significant heterogeneity in prognosis that can be explained by available clinical variables. Men in their 70s with any major comorbidity are 2-10 times more likely to die of other causes than prostate cancer.