Predictive value of kallikrein forms and β-microseminoprotein in blood from patients with evidence of detectable levels of PSA after radical prostatectomy

Purpose To determine whether β-microseminoprotein or any of the kallikrein forms in blood-free, total or intact PSA or total hK2-predict metastasis in patients with evidence of detectable levels of PSA in blood after radical prostatectomy. Method We determined marker concentrations in blood from 173 men treated with radical prostatectomy and evidence of detectable levels of PSA in the blood (PSA ≥ 0.05) after surgery between 2014 and 2015 and at least 1 year after any adjuvant therapy. We used Cox regression to determine whether any marker was associated with metastasis using both univariate and multivariable models that included standard clinical predictors. Results Overall, 42 patients had metastasis, with a median follow-up of 67 months among patients without an event. The levels of intact and free PSA and free-to-total PSA ratio were significantly associated with metastasis. Discrimination was highest for free PSA (c-index: 0.645) and free-to-total PSA ratio (0.625). Only free-to-total PSA ratio remained associated with overall metastasis (either regional or distant) after including standard clinical predictors ( p = 0.025) and increased discrimination from 0.686 to 0.697. Similar results were found using distant metastasis as an outcome ( p = 0.011; c-index increased from 0.658 to 0.723). Conclusion Our results provide evidence that free-to-total PSA ratio can risk stratifying patients with evidence of detectable levels of PSA in blood after RP. Further research is warranted on the biology of prostate cancer markers in patients with evidence of detectable levels of PSA in blood after radical prostatectomy. Our findings on the free-to-total ratio for predicting adverse oncologic outcomes need to be validated in other cohorts.