From VEGF-targeted agents to immune-based combinations and novel HIF-2α inhibitors, the expanding therapeutic arsenal for renal cell carcinoma (RCC) has significantly prolonged patient survival. At the same time, it has posed a more demanding challenge for clinicians: how to tailor the optimal treatment for each individual patient and truly achieve precision medicine.For many years, RCC was considered a “biomarker desert,” lacking clinically actionable indicators to guide treatment decisions. However, with the convergence of large-scale clinical trials and multi-omics technologies, this long-standing dilemma is finally beginning to shift.
At the 15th Shanghai Academic Conference on Urologic Oncology, Oncology Frontier – UroStream invited Professor Brian Rini from Medicine at Vanderbilt University to provide an in-depth perspective on the evolution of biomarker research in RCC. He discussed how KIM-1 may help identify patients who benefit from adjuvant therapy and outlined a forward-looking vision for a multidimensional, dynamic precision oncology paradigm.
Q1
Oncology Frontier – UroStream: You have more than 20 years of experience in the field of renal cancer. Could you share which biomarkers in RCC are relatively mature or particularly promising for prognostic stratification and treatment decision-making?
Professor Brian Rini: This is a question worth exploring in depth. Despite more than two decades of intensive effort, kidney cancer remains one of the few solid tumors without truly actionable clinical biomarkers. This contrasts sharply with breast and lung cancers, where biomarker-guided treatment strategies are well established.
The evolution of biomarker research in RCC can be broadly divided into two phases:
1. The VEGF Era (2005–2015)
Progress during this period was relatively slow. Clinical trials were limited in scale, biobanking infrastructure was underdeveloped, and molecular detection technologies were still immature. Most research focused on genes related to the VEGF pathway, yet none ultimately proved to have sufficient predictive power for routine clinical use.
2. The Immunotherapy Era
A major turning point came with the CheckMate 025 trial, which demonstrated a clear survival benefit for nivolumab in advanced RCC. However, unlike in lung cancer or melanoma, PD-L1 expression in RCC has shown limited predictive value. While PD-L1 correlates with response rates to dual immunotherapy (ipilimumab plus nivolumab), it does not reliably distinguish which patients should receive dual-IO therapy versus IO-TKI combinations.
More recently, advances in transcriptomic technologies have brought renewed optimism. Molecular subtyping based on the IMmotion151 study represented a major milestone, identifying seven molecular clusters through large-scale gene expression analysis. Subsequent work has consistently validated major RCC molecular phenotypes, including angiogenic, immune-inflamed, and proliferative subtypes.
In our OPTIC study, we demonstrated that these transcriptomic subtypes could predict responses to cabozantinib plus nivolumab. In particular, patients with strong angiogenic signatures exhibited higher response rates, providing a biologic rationale for treatment selection. Our current priority is to move these findings into prospective clinical validation, to determine whether molecular subtypes can meaningfully guide therapy in real-world practice.
Q2
Oncology Frontier – UroStream: KIM-1 has emerged as a novel biomarker in RCC, and has been explored in trials such as ASSURE and IMmotion010. Could you share key research findings and the future potential of this biomarker?
Professor Brian Rini: KIM-1 is indeed one of the most promising biomarkers we currently have. It was originally identified in the context of acute kidney injury, where it is markedly upregulated following tubular damage. In RCC, we have found that KIM-1 is specifically elevated in patient plasma and closely associated with tumor burden and disease progression.
The retrospective biomarker analysis from IMmotion010 is particularly compelling. Although the phase III trial did not meet its primary endpoint of disease-free survival (DFS), it created a valuable platform for biomarker discovery. Among approximately 2,500 plasma proteins analyzed, KIM-1 showed the strongest association with recurrence.
Importantly, KIM-1 demonstrated dual clinical utility:
- Prognostic value: Patients with high baseline KIM-1 levels (above the median) after nephrectomy had a 2.1-fold higher risk of recurrence (HR = 2.1, P < 0.001).
- Predictive value: In patients with high KIM-1 levels, adjuvant atezolizumab significantly prolonged DFS compared with placebo (HR = 0.72). In contrast, no meaningful benefit was observed in patients with low KIM-1 levels (HR = 0.98).
This is clinically important because current FDA-approved adjuvant therapies, such as pembrolizumab, are applied broadly to high-risk populations—meaning that up to 40% of patients may be overtreated. If KIM-1 is prospectively validated, it could enable us to precisely identify patients who truly benefit from adjuvant immunotherapy, thereby sparing others unnecessary toxicity and cost.
Q3
Oncology Frontier – UroStream: Looking ahead, what key research directions do you foresee for molecular subtyping and biomarker-guided precision therapy in RCC?
Professor Brian Rini: With the introduction of HIF-2α inhibitors such as belzutifan, we are now directly targeting the core VHL–HIF oncogenic pathway in RCC. At the same time, the widespread adoption of IO-TKI combinations and ongoing exploration of triplet regimens have created what I call “happy troubles”—namely, increasing cumulative toxicity and financial burden.
The central mission has shifted from finding a single “stronger” regimen for everyone to identifying the “most appropriate” regimen for each individual patient.
The future of RCC precision medicine must focus on a transition from broad-spectrum therapy to refined, personalized management, in several key areas:
1. Integrating Multidimensional, Dynamic Assessment Systems
We need to combine:
- Circulating biomarkers (KIM-1, ctDNA),
- Tissue-based features (transcriptomics, immune microenvironment),
- Genomic alterations (e.g., PBRM1, BAP1),
- Radiomic data.
By leveraging artificial intelligence and machine learning, these data can be integrated into predictive models that generate a personalized “treatment navigation map” at diagnosis.
2. Addressing Longitudinal Clinical Decision-Making
Biomarkers should not only guide treatment initiation, but also treatment de-escalation. Key questions include:
- Can ctDNA clearance identify patients who can safely discontinue therapy?
- Can we identify patients with the potential for “functional cure” in advanced disease?
- Can genetic or molecular features predict severe immune-related adverse events (irAEs)?
3. Driving Translation through Prospective Trials
We are clearly at the threshold of a breakthrough, but many insights remain hypothesis-generating. Only prospective, biomarker-guided interventional trials, directly comparing biomarker-based strategies with standard care, can convert these signals into reliable clinical tools.
Professor Brian I. Rini
Vanderbilt University Medical Center
