UTILITY OF C-REACTIVE PROTEIN LEVELS IN IMPROVING RISK STRATIFICATION IN METASTATIC RENAL CELL CARCINOMA

Abstract

Introduction

The International Metastatic Renal Cell Carcinoma Database Consortium (IMDC) serves as the main prognostic model for metastatic Renal Cell Carcinoma (mRCC). C-reactive Protein (CRP), an acute phase inflammatory marker, is associated with outcomes in mRCC. We sought to evaluate utility of addition of CRP to the current IMDC model and determine whether such additions may improve predictive capability of this model.

Methods

We conducted a retrospective analysis utilizing the International Marker Consortium for Renal Cancer registry. Inclusion criteria were mRCC at presentation or mRCC progression/recurrence. Main outcome was overall survival (OS)/all-cause mortality (ACM). Cox proportional hazard multivariable analysis (MVA) was conducted to elucidate independent predictors for ACM including preoperative CRP and current IMDC strata. Kaplan-Meier analysis (KMA) was conducted to evaluate for survival outcomes stratified by IMDC category, which were further subdivided based on normal or elevated preoperative CRP and used to propose new inclusion criteria for each IMDC category. To compare predictive capability of the proposed IMDC stratification to the current stratification, receiver operating characteristic/area under the curve (ROC/AUC) analysis was conducted.

Results

515 patients met inclusion criteria (median follow up 33 months), which were stratified into 156 favorable, 345 intermediate, and 14 poor-risk. IMDC strata were subdivided based on preoperative CRP. MVA for ACM revealed elevated CRP (CRP ≥ 5 ng/mL, HR=2.1, p=0.01) and worsening IMDC status (HR=2.2-2.5; p-value=0.006-0.03) associated with ACM. KMA comparing IMDC favorable/intermediate/poor-risk revealed 3-year OS of 60%/42%/50%, p=0.001; further subdivision by CRP revealed alignment between favorable/elevated-CRP with intermediate/normal-CRP and intermediate/elevated-CRP with poor-risk. Proposed realignment of new favorable (current favorable/normal-CRP), new intermediate (current favorable/elevated-CRP and current intermediate/normal-CRP) and new poor-risk (current intermediate/elevated-CRP and poor-risk) revealed 3-year OS of 53%/50%/40%, p=0.001 with improved predictive capability of the CRP-augmented model compared to the current IMDC (ROC/AUC of 0.677 vs. 0.649, respectively).

Conclusions

Addition of CRP to the IMDC criteria can improve patient stratification and thereby more accurately stratify patients based on their probability of survival. Validation of our findings is requisite.