In this interview, Steven Monda, MD, University of Michigan, shares insights from his recent study on biomarker performance in metastatic clear cell renal cell carcinoma (ccRCC). His findings highlight how the origin of tumor biopsies—primary vs metastatic—can significantly influence the predictive value of biomarkers used to guide targeted therapy decisions for patients with ccRCC.

Please introduce yourself by stating your name, title, organization, and relevant professional experience.

Steven Monda, MD: Hi, I am Dr Steven Monda. I am a urologic oncology fellow and the Clark Family Fellow for Kidney Cancer Research at the University of Michigan. I completed my urology residency at the University of California, Davis in Sacramento. My primary interests as a urologist are kidney cancer surgery and treatment, and kidney cancer biology. I appreciate the opportunity to talk about my work.

Can you briefly summarize your study and its key takeaways?

Dr Monda: This study investigated whether tissue-based biomarkers in metastatic clear cell renall cell carcinoma (ccRCC) perform differently depending on whether the tumor sequencing was derived from a primary or metastatic site.

By reanalyzing data from the IMmotion151 trial and incorporating matched primary-metastasis samples from an internal University of Michigan cohort and the TRACERx study, we found that angiogenic biomarkers had stronger predictive value for sunitinib when derived from metastatic tissue, while immunogenic biomarkers seemed to retain predictive utility even when sourced from primary tumors. These findings suggest that the origin of the tumor sample may impact the reliability of certain biomarkers, especially in the context of targeted therapy selection.

What prompted your team to revisit the IMmotion151 biomarker data with a focus on the origin of the tumor sequencing—primary vs metastatic sites?

Biomarker work in RCC has yielded mixed results, with many of the current trials, such as CLEAR, COSMIC-313, CheckMate 9ER, and others not showing clear patterns of who responds to which therapy. We hypothesized that a major limitation of this biomarker work is the DNA and RNA sequencing are largely derived from primary tumors—tumors we know are large and heterogeneous and not necessarily concordant with their metastatic sites. This motivated us to see how biomarkers perform depending on their site of origin.

This is important because medical oncologists are currently deciding between a pure immunotherapy regimen (ipilimumab/nivolumab) and a tyrosine kinase inhibitor (TKI)–containing regimen (lenvatinib/pembrolizumab or cabozantinib/nivolumab) largely based on intuition. Although there is some discussion of durable response with immunotherapy vs immediate control with a TKI, that decision-making is not very sophisticated.  There is biology driving who responds to immunotherapy vs vascular endothelial growth factor (VEGF)-targeted therapy.

Can you elaborate on the significance of the divergent responses seen between angiogenic and immunogenic groupings, particularly why immunogenic signatures still held predictive value when sourced from primary tumors?

My intuition—and we are looking on increasing our sample size to show this for sure—is that primary RCCs are large vascular tumors and globally enriched for an angiogenic signature, but that angiogenic status is not always enriched in the matched metastasis (met). We see this in our matched dataset where primary tumors are often more angiogenic than their matched met. However, our immunogenic signatures tend to be more concordant, meaning if this is detected in the primary tumor, it is also detected in the met. This is still a hypothesis at this point, and we are looking at validating this with further trial and matched primary-metastasis datasets.

Your findings suggest a stronger predictive value for angiogenic biomarkers when derived from metastases. How might this influence future biopsy strategies or clinical decision-making in advanced ccRCC?

One immediate implication of our work is that current and planned analyses of existing trials should stratify their analysis to tissue site of origin. There are differences between metastatic site and primary site, and by not taking that into account, we may be missing important biologic signals. Our findings are not ready to be applied to current clinical practice, especially if biopsy is just being done to yield a histologic diagnosis, but it should inform how we design future trials and scientific endeavors.

Looking ahead, how do you envision these results shaping the design of future trials or biomarker development for therapy selection in metastatic ccRCC?

As I mentioned earlier, perhaps the immediate implication of this work is how we interpret existing sequencing data. Site of origin matters, but for future trials, this work begins to suggest that metastatic site may offer a better representation of disease biology. I would also add that the one ongoing biomarker-driven trial in RCC, called OPTIC—led by Brian Rini, Scott Haake, and Katie Beckerman—has already moved from primary to metastatic site biopsy for their treatment stratification, so the field is perhaps already starting to move in that direction.