Cabozantinib Improves Survival Over Everolimus or Sunitinib in PD-L1 Unselected Patients with mRCC

Cabozantinib treatment in two randomised controlled clinical trials improved progression-free and overall survival consistently over everolimus and sunitinib in patients with metastatic, clear cell, renal cell carcinoma (mRCC), investigators reported at the ESMO 2018 Congress in Munich, Germany. Toni K. Choueiri, Division Of Genitourinary Medical Oncology, Dana-Farber Cancer Institute in Boston, United States of America and colleagues investigated the utility of PD-L1 status assessed by immunohistochemistry as a potential prognostic and/or predictive biomarker of cabozantinib activity in mRCC. Dr. Sabina Signoretti, is the senior author of the research. 

Professor Choueiri noted that based on CheckMate 214 study, PD-L1 expression on tumour cells was associated with improved outcomes in patients receiving nivolumab and ipilimumab compared to sunitinib (Motzer, ESMO 2017 Presidential session and NEJM 2018). Prior studies from Drs. Choueiri and Signoretti showed that PD-L1 expression is associated with worse outcomes to sunitinib in metastatic RCC patients (Choueiri, Clin Cancer Res 2015). Cabozantinib is an oral kinase inhibitor that is an option for first- and second-line treatment of patients with mRCC, based upon anti-tumour activity demonstrated in the CABOSUN (Alliance A031203; NCT01865747) and METEOR (NCT01835158) randomised clinical trials. 

Using formalin-fixed paraffin-embedded baseline tumour tissue obtained from 110 patients participating in CABOSUN and 306 patients from METEOR, the investigators assessed PD-L1 expression in both tumour cell and immune cell by performing immunohistochemical double-staining for PD-L1 and CD45/CD163 (immune cell markers). 

The percentages of PD-L1-positive tumour or immune cells were assessed by image analysis. Comparison of the overall response rate (ORR) by RECIST between PD-L1-positive, using ≥1% as cut-off, versus PD-L1-negative tumours was done by Fisher’s exact test and Cox regression was used to correlate progression-free survival (PFS) and overall survival (OS) with tumour PD-L1 expression across each treatment arm. The ORR, PFS and OS were per independent central review (ICR). 

Cabozantinib was associated with improved PFS and OS versus everolimus or sunitinib that was irrespective of PD-L1 expression

Twenty-nine percent of METEOR and 23% of CABOSUN samples contained PD-L1-positive tumour cells. 

With cabozantinib versus everolimus treatment in METEOR, the median PFS per IRC was 8.5 versus 5.6 months (p = 0.027), and median OS was 21.3 versus 15.1 months, p = 0.003, respectively. Median PFS per IRC with cabozantinib versus sunitinib in CABOSUN was 8.3 versus 5.5 months (p = 0.059); and median OS was 28.1 versus 20.8 months (p = 0.05). 

Treatment comparison of PFS according to PD-L1 expression revealed consistent results across PD-L1 measures, including immune cell PD-L1, combined PD-L1 score, and using different PD-L1 cut-offs. 

By univariate analysis, patients with PD-L1 levels < 1% on tumour cells had better PFS and OS than patients with PD-L1-positive tumour cells in both trials independent of therapy. 

Median PFS with cabozantinib versus everolimus in patients with tumour cell PD-L1 expression < 1% was 8.5 (95% confidence interval [CI], 7.2 - 13.5) versus 4.1 (95% CI, 3.7 – 6.0) months, hazard ratio (HR) 0.46 (95% CI, 0.32 – 0.66). Median PFS with cabozantinib versus everolimus in patients with PD-L1 expression ≥ 1% was 5.6 (95% CI, 4.5 – 7.4) versus 3.7 (95% CI, 2.5 – 5.3) months, HR 0.66 (95% CI, 0.40 – 1.11), respectively. 

In the CABOSUN trial, median PFS with cabozantinib versus sunitinib in patients with PD-L1 expression < 1% was 11.0 (95% CI, 6.8 -15.6) versus 5.0 (95% CI, 3.0 – 12.9) months, HR 0.47 (95% CI, 0.26 - 0.86). Median PFS with the respective treatments in patients with PD-L1 expression ≥ 1% was 8.4 (95% CI 1.1 – 16.6) versus 3.1 (95% CI, 1.6-10.1) months, HR 0.46 (95% CI 0.18 - 1.21).

Differences were not statistically significant in multivariable analyses adjusting for International Metastatic Renal Cell Carcinoma Database Consortium (IMDC) prognostic risk groups and bone metastasis, in either trial. 

Discussant Points

Dr Cristina Suárez of the Vall d´Hebron Institute of Oncology, Barcelona, Spain who discussed the study findings said that in univariable analysis, patients with positive PD-L1 expression in TCs had poorer PFS and OS. The association between PD-L1 expression on TCs and OS was statistically significant in the multivariable analysis combining both trials. Cabozantinib had superior clinical efficacy compared to both everolimus and sunitinib, in both PD-L1 positive and PD-L1 negative patients. The dual-IHC staining is a robust and efficient to characterise PD-L1 status on TCs and ICs.

Conclusions 

Professor Choueiri and colleagues concluded that these data support use of cabozantinib in a PD-L1 unselected population and, possibly, in combination with checkpoint blockers irrespective of PD-L1 status.

Reference

LBA 34 -Choueiri TK, Flaifel A, Xie W, et al. PD-L1 status and clinical outcomes to cabozantinib, sunitinib and everolimus in patients with metastatic clear-cell RCC treated on CABOSUN and METEOR clinical trials.