Evaluation of PBRM1, PD-L1, CD31, and CD4/CD8 ratio as a predictive signature of response to VEGFR-TKI–based therapy in patients with metastatic renal cell carcinoma (mRCC) with IMDC intermediate prognosis: Results from the APAChE-I Study

Background: Intermediate IMDC group is the largest and most heterogeneous group of mRCC. Current first-line (1L) therapy options for these patients (pts) are based on either an anti-angiogenic agent (VEGFR-TKI) combined with immunotherapy (IO), or a combo of IO (ipilimumab+nivolumab [I/N]). No biomarkers (BM) for selecting the most effective regimen have been identified so far. Methods: Immunohistochemical expression of PBRM1, PD-L1, CD31, and CD4/CD8 ratio was evaluated on histological samples of intermediate-risk mRCC pts treated with VEGFR-TKI monotherapy, and then in pts receiving a VEGFR-TKI-based therapy or the immune doublet I/N. PBRM1 positivity score was based on the percentage of positive cells and on the intensity of nuclear expression; PD-L1 positivity was defined as CPS≥10; CD31 high-density had moderate to strong nuclear staining; and the CD4/CD8 ratio cut-off for positivity was >0.2. Cox model was used to assess the correlation between BM and outcomes; PFS and OS were estimated by Kaplan-Meier method. Results: After screening of tumor tissues from 150 pts, a total of 111 were included in the final analysis (Table). In pts treated with VEGFR-TKI monotherapy, a significant correlation with PFS was observed with loss of PBRM1 expression (HR 0.58, p=0.035), PD-L1 negativity (HR 0.44, p=0.048), and high CD4/CD8 ratio (HR 0.62, p=0.073). CD31 density did not significantly correlate with PFS. A profile potentially predictive of angiogenesis (AP+) was defined based on the PBRM1 loss, PD-L1 negative, and high CD4/CD8. In pts treated with VEGFR-TKI monotherapy, tumors with the AP+ (43% of all cases) had a significantly longer median PFS (mPFS 23.8 vs. 11.8 months, p=0.003) and mOS (41.5 vs. 26.9 months, p=0.024) compared to the others. The AP+ retained its significant correlation with PFS (mPFS 23.8 vs. 11.1 months, p<0.001) and OS (41.5 vs. 24.9, p=0.006) in pts receiving VEGFR-TKI-based therapies. The rate of AP+ tumors was 55.6% and 32.7% in pts with one or two IMDC risk factors, respectively (p=0.022). In the small cohort of pts treated with I/N, no differences were observed in PFS (p=0.64) and OS (p=0.75) between AP+ and AP-negative. Conclusions: The AP+ signature (loss of PBRM1, PD-L1 negative, and CD4/CD8 high ratio) was associated with improved clinical outcomes in mRCC pts at IMDC intermediate prognosis treated with VEGFR-TKI-based therapy; this correlation was significant regardless from the addition of IO to VEGFR-TKI monotherapy. Prospective validation of this signature is required for guiding the selection of the most appropriate 1L therapy.