Post Hoc Analysis Reveals Potential Serum Glycoproteomic Biomarkers for Nivolumab/Cabozantinib in Advanced RCC

David A. Braun, MD, PhD, discusses the potential for protein glycosylation as a biomarker to predict nivolumab/cabozantinib responses in advanced RCC.

The amount of glycosylation—specifically sialylation of serum proteins—is important in patients with advanced renal cell carcinoma (RCC), as higher levels of glycosylation, sialylation, and fucosylation were associated with worse clinical outcomes for patients with advanced renal cell carcinoma (RCC) treated in the frontline with either the combination of nivolumab (Opdivo) and cabozantinib (Cabometyx) or sunitinib (Sutent) monotherapy, according to data from a post-hoc analysis of the phase 3 CheckMate-9ER study (NCT03141177).

Findings from the analysis, which were presented at the 2024 ESMO Congress, also showed that serum glycoproteins involved in complement cascade and lipid metabolism could potentially serve as predictors for response to nivolumab plus cabozantinib vs sunitinib in patients with advanced RCC.

In an interview with OncLive®, lead study author David A. Braun, MD, PhD, discussed previous findings from the CheckMate-9ER study, expanded on the evidence supporting glycosylation as a potential biomarker, and explained the next steps for this research.

Braun is an assistant professor of medicine (Medical Oncology) and Louis Goodman and Alfred Gilman Yale Scholar at Yale School of Medicine in New Haven, Connecticut.

OncLive: What were key findings from the CheckMate-9ER trial that were previously reported?

Braun: CheckMate-9ER was a pivotal phase 3 trial for advanced kidney cancer. It looked at the combination of nivolumab plus cabozantinib vs sunitinib, [meaning] an immuno-oncology (IO)/TKI [combination] vs a TKI alone.

This was 1 of the key trials that established nivolumab plus cabozantinib as a standard of care for the first-line setting [for advanced RCC]. The [combination] improved response, progression-free survival [PFS], and overall survival [OS] relative to sunitinib. Despite all this, which has been clinically important, we still don't have good tools to distinguish which patients are likely to respond and which patients, unfortunately, have resistant tumors. That's where the idea of effective biomarkers comes along.

What evidence supports the potential use of protein glycosylation as a biomarker of response?

A lot of our group’s work and [the work of] other groups has focused on the genomics of [kidney] cancer itself and transcriptomic approaches to understand gene expression signatures. We also know biologically that the post-translational modification of proteins impacts [tumor] function.

Altered protein glycosylation could be considered a hallmark of cancer. Many malignant processes are guided by altered protein glycosylation. Just as one example, we know that when a particular form of glycosylation—sialylation—occurs in cancer at high levels, which it does, it can be immunomodulatory. It can engage things like cyclic molecules and ultimately lead to immune suppression and potential resistance to PD-1 blockade. It's with that background that this is a critical process essential to oncogenesis and maintaining oncogenic function, and [it has the] potential to be immunomodulatory. [Based on that], we asked the question of whether this might be relevant for predicting [response to] nivolumab plus cabozantinib.

What were some key findings from the exploratory biomarker analysis?

There are essentially 3 key parts that we looked at. First, as this is a relatively new type of investigation, we asked whether the regulation of glycosylation of different proteins was highly regulated and coordinated, meaning we'd expect to see big groups of patients acting similarly with respect to their glycosylation, or whether it was very disordered, [meaning it] was unlikely to be a good predictive or prognostic tool. We saw that many of these features were highly correlated across patients. Essentially, multiple proteins were glycosylated at different sites, but that was highly coordinated across big blocks of patients.

Second, was the total amount of glycosylation important? We looked at 2 specific forms of glycosylation: fucosylation [and] sialylation. We saw that as these processes [increase, patients] got progressively worse with respect to PFS and OS. However, this wasn't unique to nivolumab plus cabozantinib. This was observed in the sunitinib arm, as well. The idea is that this might be prognostic, meaning that having lots of protein glycosylation might overall be an indicator of a bad-acting tumor.

The final question we asked was whether this could potentially be predictive in a way. Were there particular glycopeptides that might distinguish response or outcomes with nivolumab plus cabozantinib vs sunitinib? The short answer is yes. There was a handful related to lipid metabolism, but they were largely related to the complement cascade. We highlighted 2 examples of this: CO3 and CFAH, which had opposite patterns. However, in their own way, both [CO3 and CFAH] were able to distinguish—when they were at altered levels—response or PFS with nivolumab plus cabozantinib vs sunitinib.

This is early [research] and an exploratory study, but [these data] at least give us some confidence that this was a worthwhile line of investigation that we need to pursue further and externally validate.

What are the potential future implications of these findings or direct next steps?

We've done a lot of genomics as a field in transcriptomics, but not a lot with protein glycosylation. Is this something that's worth studying? I think the answer is yes, and this is certainly an area worthy of further investigation.

We have some interesting hypothesis-generating findings, but the next step is to validate. In other cohorts with similar types of treatment, [we need to be able to] say, ‘[These data] are not just a one-off. This is something where we can consistently see specific glycopeptides that are altered and prognostic or potentially predictive of response in IO-based therapy.’

If [these findings are] validated, then there's a whole other area of biological investigation. What is this altered protein glycosylation doing? Can we somehow modify and improve the impact of these therapies for our patients? That's much further down the line after we get to the most critical next step, which is validation.

Reference

Braun D, Wang Y, Yu A, et al. 1694MO Novel serum glycoproteomic biomarkers predict response to nivolumab plus cabozantinib (NIVO+CABO) versus sunitinib (SUN) in advanced RCC (aRCC): Analysis from CheckMate 9ER. Ann Oncol. 2024;35(suppl 2):S1013. doi: 10.1016/j.annonc.2024.08.1787