Avelumab/Axitinib Combo Biomarker Analysis May Lend Insight to Complex RCC Landscape

Toni K. Choueiri, MD

A subgroup analysis of the phase III JAVELIN Renal 101 study identified molecular features that may define therapy-specific outcomes in patients with advanced renal cell carcinoma (RCC), explained Toni Choueiri, MD.

In the analysis of JAVELIN Renal 101—a trial which led to the May 2019 FDA approval of avelumab (Bavencio) plus axitinib (Inlyta) for the frontline treatment of patients with advanced RCC— researchers found that PD-L1 expression ≥1% appeared to be the greatest predictor of benefit from the PD-L1 inhibitor/VEGF TKI combination.¹ This factor was associated with the longest progression-free survival (PFS) in the combination arm and the shortest PFS in patients who received sunitinib (Sutent; HR, 0.63; 95% CI, 0.49-0.81), according to findings presented at the 2019 ASCO Annual Meeting.

Investigators also looked at relevant gene expression signature (GES) in relation to clinical outcomes; these analyses included published and de novo signatures. They noted significant differences in PFS between the treatment arms relative to wild-type when mutations in genes such as CD1631L, PTEN, or DNMT1 were present.

High-angiogenesis GES was associated with significantly improved PFS in the sunitinib arm but did not differentiate PFS in the combination arm. In the low-angiogenesis subset, the combination was found to result in improved PFS compared with sunitinib. Patients with a high number of effector T cells and inflamed T cells had longer PFS with the combination compared with sunitinib. Furthermore, in the combination arm, the PFS benefit was enhanced in patients with immune GES—positive tumors compared with those in the negative group (HR, 0.63; 95% CI, 0.46-0.86; 2-sided P = .004).

In an interview with OncLive during the 2019 ASCO Annual Meeting, Choueiri, director of the Lank Center for Genitourinary Oncology, director of the Kidney Cancer Center at Dana-Farber Cancer Institute, and the Jerome and Nancy Kohlberg Associate Professor of Medicine at Harvard Medical School, discussed the biomarker analysis results from JAVELIN Renal 101 and provided insight on where the RCC field stands following pivotal data presented at the meeting.

OncLive: What were the key takeaways from the biomarker analysis of JAVELIN Renal 101?

Choueiri: JAVELIN Renal 101 is a study that I had the honor to lead, and it led to the FDA approval of the combination of the PD-L1 inhibitor avelumab and the VEGF TKI axitinib in advanced clear-cell RCC. This study, published in the New England Journal of Medicine, showed a significant PFS benefit with the combination over sunitinib and a more than doubled response rate. This benefit was seen both in patients who had an independent review of their imaging and by investigator's assessment.

What we did here, in this biomarker analysis, is obtain a tissue sample from the kidney or metastatic site within 1 year. We had asked the important questions—a lot of them were preconceived—such as, "What does PD-L1 status do in positive versus negative patients? Can we come up with a gene signature based on RNA sequencing? Is there any mutational profile to distinguish responders from nonresponders in the sunitinib and combination arms?”

For immunohistochemistry, looking at PD-L1 in the combination arm, PD-L1 status didn't matter. On the sunitinib arm, PD-L1—positive patients did worse; that's consistent with our prior experience. With CD8, it's different. The invasive margin was associated with better benefit in the combination arm; however, the opposite was true in the sunitinib arm. That was comforting because it was in line with what we thought would happen. We had more than 700 patients, so then we said, "Let us do our own gene expression analysis." We started with 4000 genes and went down to 306 genes that had an influence on immune response, and then down to 26 genes that were part of some T-cell or natural killer cell response.

We came up with this signature that we call the JAVELIN Renal 101 signature. That signature did distinguish, in patients with high expression, which patients would have higher PFS. They were enriched for better PFS in the combination arm, and this didn't mean anything in the sunitinib arm.

Then we had a data set of 55 patients from the phase Ib study [and we found the] same thing: if you look at this signature, it was enriched for the responders; so we have a validation set. Finally, we did what many people do, which is whole-genome sequencing. We came up with very interesting genes like PTEN and others that were associated with either better response on the combination arm or on the sunitinib arm.

We want to think about this not just as relevant to avelumab plus axitinib but for immunotherapy plus TKI combinations in general. We are trying to validate this preliminary finding, and we have blood at certain intervals and tumor at progression. Hopefully, this will provide significant insight into the biology [of this disease] that we can take into the clinic.

What were some other studies presented at the 2019 ASCO Annual Meeting that you were excited about?

This was an exciting 2019 ASCO Annual Meeting for kidney cancer. Many important studies—although some were negative—were presented. One study in particular was an ECOG study looking at the adjuvant VEGF TKI pazopanib (Votrient). At the 2018 ASCO Annual Meeting, we saw that sorafenib (Nexavar) didn't have an effect. We didn't see a major difference in sorafenib versus no sorafenib. [We saw] the same thing with pazopanib in a very well-designed study by Leonard J. Appleman, MD, of University of Pittsburgh Medical Center Hillman Cancer Center.

Another interesting story follows the plenary session from last year. There was a study on cytoreductive nephrectomy in RCC, and it showed that this approach is not for everyone. Researchers looked at the International Metastatic RCC Database Consortium risk group, and we know how this group changes the way we think of metastatic RCC. Actually, they found that there were differences, and that maybe cytoreductive nephrectomy could be attempted in certain patients, especially if they are in the low-risk group. This is important, and it will play a role in practice. This will hopefully lead to better and better multidisciplinary approaches in practice.

Another interesting thing I noticed was the focus this year on metastatic disease, on the difficult-to-treat histologies. [These patients are] more rare and difficult to treat, the non—clear cell and sarcomatoid patients. There is some beautiful work from KEYNOTE-426 and from some retrospective studies [that we heard about]. There was also a phase II study of atezolizumab (Tecentriq) and bevacizumab (Avastin) in non–clear cell disease, which showed that patients do respond to checkpoint blockers.

What is refreshing to know is that there was actually an abstract around sarcomatoid histology. These can happen with every RCC, but more and more data are coming showing that patients seem to have better outcomes with checkpoint inhibitors. Brian Rini, MD, of Cleveland Clinic, and David F. McDermott, MD, of Beth Israel Deaconess Medical Center, and others, presented these data. It is quite interesting because these patients do not respond well to immunotherapy nor to targeted therapy. There is some immune signature here and some biologic differences.

I look forward to continuing with this path because these sarcomatoid-containing tumors are quite aggressive. Therefore, having a consensus on what to treat them with is very important. Thus far, the checkpoint inhibitors are emerging as the winners.

Following the FDA approvals of 2 different TKI/checkpoint inhibitor combinations in advanced RCC, how will we go about deciding which combination to use?

This question comes up a lot of the time, and I would even add nivolumab (Opdivo) plus ipilimumab (Yervoy) to that mix. Cabozantinib (Cabometyx) plus nivolumab is another combination that is coming if we see a positive trial. In a few years, we may have more combinations. At this point, in comparing [these combinations], in the community, people are already familiar with pembrolizumab (Keytruda). The dosing schedule is every 3 weeks, so people are telling me pembrolizumab is easier and that they use it. Probably the most important thing is that in KEYNOTE-426, the study looking at pembrolizumab plus axitinib, there was an overall survival (OS) benefit. There are no data with avelumab and axitinib showing an OS benefit, despite the fact that the response rates and PFS are very similar to a certain degree.

The third layer of nivolumab plus ipilimumab is actually just [indicated for] the intermediate- and poor-risk patients, but it has the highest complete response rates [out of these combinations]. We know these are very durable responses and we think that many of these patients are cured; this is very attractive.

The next generation of trials should [set out to answer] whether we start with nivolumab and add ipilimumab, or do we start with pembrolizumab and add axitinib, or do we start with avelumab and add axitinib? We also need to move away from sunitinib being the control arm. This would shy away from controversy. I know there are studies in the works that are establishing nivolumab plus ipilimumab as the new standard. RCC continues to be extremely interesting and dynamic, and to me, this means just more options for patients. Median OS, when I joined the staff in 2007, was 11 months. Now we are seeing median OS not reached at 3 and 4 years. Patients are living longer with their disease.

Are we exploring the use of subsequent combination regimens?

This is a very important question. We know by [giving subsequent single-agent TKIs] we have activity. The question is, really, “What we do after PD-1/PD-L1 blockade?” If somebody starts on pembrolizumab plus axitinib, would you then do nivolumab plus ipilimumab? I'm not aware of any trials looking at this. These PD-1/L1 inhibitors are very similar, so do you continue giving them or not? Hopefully we can soon design a study. We don't know the contribution of each agent.

For example, with the avelumab plus axitinib study, it would be interesting from an academic perspective if we have this combination versus axitinib versus avelumab versus sunitinib. That's thousands of patients, though. It will be interesting to see, just in the untreated patients, the contribution of each of these regimens. These are very important questions, and hopefully they will be addressed soon.

Choueiri TK, Albiges L, Haanen JB, et al. Biomarker analyses from JAVELIN Renal 101: Avelumab + axitinib (A+Ax) versus sunitinib (S) in advanced renal cell carcinoma (aRCC). J Clin Oncol. 2019;37(suppl 15; abstr 101). doi: 10.1200/JCO.2019.37.15_suppl.101.