On-treatment modified Glasgow prognostic score (mGPS) improves imaging-only prediction of response and outcomes in metastatic renal cell carcinoma

Abstract

Introduction & Objectives: Reliable prediction of response to treatment is crucial for optimal patient management. However, in routine clinical practice, we have not yet progressed beyond imaging-based staging with its vague estimation of tumor volume. Especially in the era of immuno-oncology, imaging alone seems to be insufficient to capture treatment response. In particular, patients with disease control (DC) treated with immunotherapy have a wide range of clinical outcomes. The modified Glasgow prognostic score (mGPS) is a simple score based on the two serum markers C-reactive protein (CRP) and albumin. Recently, we demonstrated that this simple score was superior to the IMDC score in patients with metastatic renal cell carcinoma (mRCC) treated in the phase 3 IMmotion151 trial. However, the question of whether longitudinal changes in mGPS can predict responses and outcomes has not been systematically investigated. Materials & Methods: We evaluated the prognostic and predictive performance of on-treatment mGPS in patients with mRCC treated with atezolizumab (plus bevacizumab) or sunitinib in two independent clinical trials, namely the phase 3 IMmotion151 study (discovery cohort) and the phase 2 IMmotion150 study (validation cohort) comprising in total n=890 patients. To compare the prognostic value of the on-treatment mGPS with radiological staging, we used RECIST assessed by the Independent Review Committee (IRC-RECIST) to ensure high data quality. IRC-RECIST was available for n=611 patients. Acknowledgment: This publication is based on research using data from Roche that has been made available through Vivli, Inc. Vivli has not contributed to or approved, and is not in any way responsible for, the contents of this publication. Results: In both cohorts, longitudinal changes in baseline mGPS provided valuable prognostic information. In the IMmotion150 study, on-treatment mGPS predicted outcomes as early as 6 weeks following therapy initiation. Of note, in both clinical trials, on-treatment mGPS provides additive prognostic information regardless of imaging-assessed treatment response at first staging. Of note, in the DC subgroup, comprising about 80 % of both cohorts, respectively, on-treatment mGPS exhibited superior prognostic information compared to IRC-RECIST: Univariable Cox: Hazard ratio (HR) for death 2.71/ 7.54 for mGPS intermediate/ high-risk compared to mGPS low-risk vs. HR for death 2.05 for SD compared to PR/CR subgroup; C-index of on-treatment mGPS was 0.651 (0.588 – 0.714) vs. IRC-RECIST: 0.574 (0.528 – 0.619). Conclusions: The simple and inexpensive determination of on-treatment mGPS provides valuable prognostic information for patients with mRCC. Our data support the idea of integrating on-treatment mGPS as a simple approach for more patient-centered therapy monitoring in addition to radiological staging alone for patients with mRCC.