Combination of C-reactive protein/albumin ratio and time to castration resistance enhances prediction of prognosis for patients with metastatic castration-resistant prostate cancer

Abstract

Objective: This study aimed to identify the prediction accuracy of the combination of C-reactive protein (CRP) albumin ratio (CAR) and time to castration resistance (TTCR) for overall survival (OS) following development of metastatic castration-resistant prostate cancer (mCRPC).

Methods: Clinical data from 98 mCRPC patients treated at our institution from 2009 to 2021 were retrospectively evaluated. Optimal cutoff values for CAR and TTCR to predict lethality were generated by use of a receiver operating curve and Youden's index. The Kaplan-Meier method and Cox proportional hazard regression models for OS were used to analyze the prognostic capabilities of CAR and TTCR. Multiple multivariate Cox models were then constructed based on univariate analysis and their accuracy was validated using the concordance index.

Results: The optimal cutoff values for CAR at the time of mCRPC diagnosis and TTCR were 0.48 and 12 months, respectively. Kaplan-Meier curves indicated that patients with CAR >0.48 or TTCR <12 months had a significantly worse OS (both p < 0.005). Univariate analysis also identified age, hemoglobin, CRP, and performance status as candidate prognostic factors. Furthermore, a multivariate analysis model incorporating those factors and excluding CRP showed CAR and TTCR to be independent prognostic factors. This model had better prognostic accuracy as compared with that containing CRP instead of CAR. The results showed effective stratification of mCRPC patients in terms of OS based on CAR and TTCR (p < 0.0001).

Conclusion: Although further investigation is required, CAR and TTCR used in combination may more accurately predict mCRPC patient prognosis.

Keywords: C-reactive protein albumin ratio; biomarker; metastatic castration-resistant prostate cancer (mCRPC); metastatic hormone sensitive prostate cancer; time to castration resistance.

Figure 1

Flowchart showing patient eligibility, study design, and statistical methods. CRPC, castration-resistant prostate cancer; mCRPC, metastatic CRPC; ADT, androgen deprivation therapy; HSPC, hormone-sensitive prostate cancer; CAR, C-reactive protein albumin ratio; TTCR, time to castration resistance; ROC, receiver operating curve; OS, overall survival; PSA, prostate-specific antigen.

Figure 2

ROC curves for overall survival after castration resistance shown by CAR (CRP/Alb ratio) or TTCR (time to castration resistance). Optimal cutoff values for CAR and TTCR were determined to be 0.48 (area under the curve 0.637, sensitivity 0.481, and specificity 0.783) and 12.2 months (area under the curve 0.609, sensitivity 0.577, and specificity 0.630), respectively. Comparisons of these values with the cutoff value defined by the median confirmed the superiority of values determined with use of the Youden index.

Figure 3

Kaplan–Meier analysis of overall survival after castration resistance, and PSA progression-free survival following first-line treatment for mCRPC based on CAR (CRP/Alb ratio) and TTCR (time to castration resistance). (A) Kaplan–Meier curves for OS based on CAR and TTCR, and those in combination. OS for the high CAR and TTCR <12-month groups was significantly worse than for the low CAR and TTCR ≥12-month groups, respectively. Risk stratification according to values for CAR and TTCR effectively stratified the prognosis of mCRPC patients. (B) Kaplan–Meier curves for PSA progression-free rate after first-line treatment against mCRPC based on CAR and TTCR, and those in combination. Similar to the results seen in the OS analysis, CAR, TTCR, and the combination of both factors provided correct risk classification regarding the duration of first-line treatment response.

Figure 4

Efficacy and impact on overall survival and PSA progression-free survival of different first-line agents for mCRPC. (A) PSA responses for patients with ARAT, AA, and DTX treatment were 82.0%, 37.8%, and 34%, respectively. (B) The PSA progression-free survival rate was significantly better in ARAT patients, whereas OS was not significantly different among the three treatments. (C) Kaplan–Meier curve showing PSA progression-free rate after first-line treatment for mCRPC with the three treatments. The duration of PSA response in ARAT- and DTX-treated patients was significantly different among the three groups classified by CAR and TTCR. (D) Kaplan–Meier curve showing OS after first-line treatment for mCRPC with the three treatments. CAR- and CRP-based risk categorization effectively stratified the respective OS of mCRPC patients treated with the three different agents.