Higher T-Cell Inflamed GEP Values Are Positively Associated with Outcomes For Pembrolizumab/Axitinib in ccRCC

Brian Rini, MD

Higher T-cell inflamed gene expression profile (GEP) values were positively associated with clinical outcomes for pembrolizumab (Keytruda) plus axitinib (Inlyta), and angiogenesis had a positive association with clinical outcomes for sunitinib (Sutent), in patients with clear cell renal cell carcinoma (RCC), according to an exploratory biomarker analysis of the phase 3 KEYNOTE-426 trial (NCT02853331) presented at the 2024 ASCO Annual Meeting.

There was a significant positive association of a higher T-cell inflamed GEP with objective response rate (ORR; P < .0001), progression-free survival (PFS; P < .0001), and overall survival (OS; P = .002) for pembrolizumab plus axitinib; there was no significant association of higher T-cell inflamed gene expression score with outcomes for sunitinib. Higher angiogenesis was positively associated with OS (P = .004) but not PFS or ORR for pembrolizumab/axitinib and was positively associated with ORR (P = .002), PFS (P < .001), and OS (P < .0001) for sunitinib. Additionally, PD-L1 combined positive score (CPS) was negatively associated with OS (P = .013) for sunitinib and not significant for the other outcome measures or with any outcome measures for the doublet therapy.

“PD-L1 continues to underperform as a biomarker in kidney cancer [and] was not associated with clinical outcomes for pembrolizumab plus axitinib,” Brian Rini, MD, said in a presentation of the data. “We saw an improved response rate across molecular subtypes for the combination regimen and, more interestingly, the highest response rate within the immune/proliferative subtype for pembrolizumab plus axitinib and the highest response rate within the angiogenetic subtype for sunitinib.”

Patients with angiogenesis below the median in the pembrolizumab plus axitinib arm (n = 190) experienced a median PFS of 15.2 months vs 9.5 months for those in the sunitinib arm (n = 175; HR, 0.62; 95% CI, 0.48-0.79); the median OS was 42.9 months vs 32.6 months, respectively (HR, 0.69; 95% CI, 0.54-0.89). In patients with angiogenesis at or above the median who received the combination therapy (n = 179) the median PFS was 18.0 months vs 15.0 months for those treated with sunitinib (n = 186; HR, 0.73; 95% CI, 0.57-0.94). The median OS was 57.6 months vs 54.2 months, respectively (HR, 0.94; 95% CI, 0.71-1.24).

The median PFS for patients with a low T-cell inflamed GEP in the pembrolizumab plus axitinib arm (n = 129) was 12.4 months compared with 16.3 months in the sunitinib arm (n = 114; HR, 0.91; 95% CI, 0.67-1.23); the median OS was 44.3 months vs 47.1 months, respectively (HR, 0.90; 95% CI, 0.65-1.23).
Among patients with a non-low T-cell inflamed GEP, those treated with pembrolizumab plus axitinib (n = 240) experienced a median PFS of 20.2 months vs 9.8 months for those given sunitinib (n = 247; HR, 0.58; 95% CI, 0.47-0.72). The median OS was 53.3 months compared with 40.8 months, respectively (HR, 0.77; 95% CI, 0.61-0.96).

Rini, who is chief of Clinical Trials, Ingram Professor of Cancer Research, and a professor of medicine at Vanderbilt Ingram Cancer Center in Nashville, Tennessee, also added that “Most importantly, additional correlative data is needed. We need to translate these findings and [conduct] prospective clinical investigations to turn these biomarkers into useful biomarkers to select treatment for patients.”

Enrollment Criteria and Baseline Patient Characteristics

KEYNOTE-426 enrolled patients with newly diagnosed or recurrent stage IV clear cell RCC who did not receive prior systemic treatment for advanced disease and stratified them by International Metastatic RCC Database Consortium (IMDC) risk group (favorable vs intermediate vs poor) and geographic region (North America vs Western Europe vs rest of world). Patients who received at least 1 dose of treatment and had an evaluable RNA-sequencing or whole-exome sequencing sample were included in the biomarker-analysis. In KEYNOTE-426 patients were randomly assigned 1:1 to receive pembrolizumab 200 mg every 3 weeks for up to 35 cycles plus axitinib 5 mg twice daily or sunitinib 50 mg daily for the first 4 weeks of each 6-week cycle. The dose of axitinib could be increased to 7 mg then 10 mg twice daily if safety criteria were met and could also be reduced to 3 mg then 2 mg twice daily to manage toxicities; sunitinib could be decreased to 37.5 mg then 25 mg once daily for the first 4 weeks of each 6-week cycle.

Angiogenesis, 18-gene T-cell inflamed GEP, non-T-cell inflamed GEP signatures, and molecular clusters were evaluated via RNA sequencing. VHL, PBRM1, SETD2, and BAP1 were examined via whole-exome sequencing, and PD-L1 CPS was evaluated via the PD-L1 IHC 22C3 pharmDx assay. Rini noted that angiogenesis, 18-gene T-cell inflamed GEP, and PD-L1 CPS were the primary end points of the biomarker analysis.

“We used standard statistical methods to look at the association between each of these biomarkers and clinical outcomes both within and across the pembrolizumab/axitinib and sunitinib arms, importantly, with adjustment for IMDC criteria and adjustment for multiplicity. These translational data are based on the last clinical cutoff of the minimum 5-year follow-up—the most recent clinical data that have been presented,” Rini said.

Baseline characteristics were well balanced across the doublet and monotherapy arms for those who received whole-exome sequencing as well as RNA sequencing. The median age of those who received RNA sequencing was 62 years (IQR, 14) in the pembrolizumab plus axitinib arm (n = 369) and 61 years (IQR, 15) in the sunitinib arm (n = 361). Most patients were male (70.7% vs 74.5%), had intermediate/poor disease by IMDC risk category (69.9% vs 70.4%), and a PD-L1 CPS of at least 1 (61.2% vs 62.6%), respectively. Patients included in the biomarker analysis also had sarcomatoid features (12.5% vs 13.0%) and were enrolled in North America (25.5% vs 25.2%), Western Europe (27.1% vs 26.0%), and the rest of the world (47.4% vs 48.8%), respectively.

Additional Findings From the Biomarker Analysis

“Pembrolizumab/axitinib had a superior ORR over sunitinib across molecular subtypes,” Rini explained. “The angiogenic [subtype had] the least difference in response rates between the arms and that’s because that’s the molecular cluster in which sunitinib performed the best in terms of response rate. If you then look at cluster 4, you see the greatest difference between the arms and that’s because the pembrolizumab/axitinib response rate in this subset was nearly 80%. Differential response according to molecular cluster [is seen as well].”
Additionally, the only significant DNA result was seen with the PBRM1 subtype, according to Rini. The ORR for patients with mutated PBRM1 treated with the combination (n = 154) was 71.4% compared with 52.3% for those treated with sunitinib (n = 193; P = .002).

When examining non–T-cell inflamed GEP signatures and clinical outcomes without adjustment for T-cell inflamed GEP, investigators found that myeloid derived suppressor cell (MDSC), glycolysis, RAS, stomal/EMT/TGF-β, and WNT did not show significance in either treatment arm.

Rini explained that “There was some association with the MDSC signature and clinical outcomes in the pembrolizumab/axitinib arm. However, this was not significant after adjusting for the T-cell [inflamed] GEP. The other non–T-cell GEP signatures are generally not significant, although there were some negative associations between the MYC and proliferative subsets and some of the clinical outcomes in the sunitinib arm.”

Reference

Rini BI, Plimack ER, Stus V, et al. Biomarker analysis of the phase 3 KEYNOTE-426 study of pembrolizumab (P) plus axitinib (A) versus sunitinib (S) for advanced renal cell carcinoma (RCC). J Clin Oncol. 2024;42(suppl 16):4505. doi:10.1200/JCO.2024.42.16_suppl.4505