Gene Expression Data Validated for Treatment Benefit in RCC

Distinct subgroups of patients with renal cell carcinoma defined by gene expression signatures were correlated with improved progression-free survival with immunotherapy or a targeted agent, an analysis of the phase II&nbsp;IMmotion151 trial showed.<br /> &nbsp;

Brian Rini, MD

Distinct subgroups of patients with renal cell carcinoma defined by gene expression signatures were correlated with improved progression-free survival (PFS) with immunotherapy or a targeted agent, an analysis of the phase II IMmotion151 trial showed.

According to the findings, presented at the 2018 ESMO Congress, signature reflecting high T-effector cell expression (TeffHigh) was associated with significant improvement in PFS with the combination of atezolizumab (Tecentriq) and bevacizumab (Avastin) versus sunitinib (Sutent). The combination also led to a numerically better PFS in tumors that had a signature reflecting low expression of genes associated with angiogenesis (AngLow). Sunitinib performed better in tumors with the angiogenesis (AngHigh) signature.

Atezolizumab/bevacizumab significantly improved median PFS versus sunitinib in the AngLow (8.94 vs 5.95 months; HR, 0.68; 95% CI, 0.52-0.88). The combination led to a smaller improvement in PFS in the AngHigh subgroup (12.45 vs 10.12 months; HR, 0.95; 95% CI, 0.76-1.19).

The combination of atezolizumab and bevacizumab also improved PFS versus sunitinib in the TeffHigh subgroup (12.45 vs 8.34 months; HR, 0.76; 95% CI, 0.59-0.99). A smaller, nonsignificant increase in PFS was observed with the combination in patients with TeffLow tumors (9.72 vs 8.41 months; HR, 0.91; 95% CI, 0.73-1.14).

The previously reported primary analysis from IMmotion151 showed that the atezolizumab/bevacizumab combination significantly improved the median PFS versus sunitinib in both the intention-to-treat population (11.2 vs 8.4 months; HR, 0.83; 95% CI, 0.70-0.97) and in the subgroup of patients with PD-L1—positive tumors (11.2 vs 7.7 months; HR, 0.74; 95% CI, 0.57-0.96,P= .02).

Those earlier findings from IMmotion151 validated gene signatures initially identified in the randomized phase II IMmotion150 trial that compared the atezolizumab/bevacizumab combination versus atezolizumab monotherapy versus sunitinib monotherapy in patients with untreated metastatic RCC.

“This was an important finding, that the gene signatures were indeed validated,” Brian I. Rini, MD, of the Cleveland Clinic, said. “It’s not often that we get a chance to validate gene expression data like this in two datasets.”

“Findings from this study further the understanding of RCC biology and inform future strategies to enable personalized therapy,” he added.

The IMmotion150 trial (N = 305) demonstrated a 7-month improvement in median PFS versus sunitinib. As part of the study, investigators developed a transcriptome map of tumors for analysis of baseline gene expression profiles.

That analysis identified distinct subgroups defined by gene expression signatures, including AngHigh and TeffHigh, associated with genes involved in immunity and antigen presentation. The TeffHigh subgroup was further subdivided into myeloidHigh and myeloidLow groups.

Comparison of PFS and gene expression signatures in IMmotion150 showed that sunitinib improved PFS in the AngHigh subset. The atezolizumab/bevacizumab combination improved PFS versus sunitinib in patients with TeffHigh and AngLow tumors and improved PFS versus atezolizumab monotherapy in TeffHigh/MyeloidHigh.

Rini and colleagues developed a transcriptome map for the 823 patients in IMmotion151 to validate the signatures. Using cutoffs established in IMmotion150, investigators performed a prespecified analysis of the association between gene expression signatures and PFS, focusing only on AngHigh and TeffHigh without the myeloid subdivision.

Sunitinib was associated with a significantly better median PFS in patients with AngHigh versus AngLow (10.12 vs 5.95 months; HR, 0.59; 95% CI, 0.47-0.75). In contrast, the median PFS did not differ significantly between patients with AngHigh or AngLow tumors (12.45 vs 8.94 months; HR, 0.86; 95% CI, 0.67-1.1).

Investigators performed analyses involving Memorial Sloan Kettering Cancer Center (MSKCC) risk status (favorable and intermediate/poor risk) and sarcomatoid histology to gain insight into the interaction between expression and other RCC parameters.

Among patients with PD-L1—positive tumors, atezolizumab/bevacizumab consistently improved PFS versus sunitinib, achieving hazard ratios of 0.70 for MSKCC favorable (n = 61), 0.74 for MSKCC intermediate/poor (n = 291), 0.48 for sarcomatoid tumors (n = 84), and 0.86 for nonsarcomatoid tumors (n = 267).

An analysis of all evaluable patients showed smaller but consistent benefits for the combination, which led to hazard ratios of 0.94 for MSKCC favorable (n = 156), 0.82 for MSKCC intermediate/poor (n = 667), 0.57 for sarcomatoid tumors (n = 134), and 0.92 for nonsarcomatoid tumors (n = 688).

An analysis of gene signatures by MSKCC risk status showed that a significantly higher proportion of favorable-risk patients had high angiogenesis gene expression as compared with intermediate- and poor-risk patients (74% vs 57%;P= 8.26e-05). Teff gene expression and PD-L1 status did not differ significantly by MSKCC risk category.

Patients with sarcomatoid tumors derived substantial benefit from the atezolizumab/bevacizumab combination. Investigators found that sarcomatoid tumors had significantly lower angiogenesis gene expression as compared with nonsarcomatoid tumors (34% vs 65%; P = 1.12e-11). T-effector gene expression was increased in sarcomatoid tumors (54% vs 40%;P= 1.90e-03). Sarcomatoid tumors also were significantly more likely to be PD-L1—positive (63% vs 39%;P= 5.25e-07).


Rini BI, Huseni M, Atkins MB, et al. Molecular correlates differentiate response to atezolizumab + bevacizumab vs sunitinib: results from a phase III study (IMmotion151) in untreated metastatic renal cell carcinoma. In: Proceedings from the 2018 ESMO Congress; October 19-23, 2018; Munich, Germany. Abstract LBA31.