San Francisco, CA (UroToday.com) There has been a marked advanced in the understanding of the molecular underpinnings of kidney cancer over the past few years. Dr. Martin Voss, Medical Oncologist at Memorial Sloan Kettering Cancer Center, discussed clinically useful biomarkers in the current era.
Compared to other solid tumors, the prevalence of actionable mutations in renal cell carcinoma (RCC) is relatively low. 
Nevertheless, a number of biomarkers are easily evaluable in the clinic and can assist in personalizing therapy for individual patients. In patient with intermediate and high risk metastatic RCC, both ipilimumab plus nivolumab (Ipi-Nivo) and axitinib plus pembrolizumab (Axi-Pembro) have demonstrated overall survival (OS) benefit compared to sunitinib in the first-line treatment setting. Notably median follow-up is significantly longer for Ipi-Nivo (32.4 months) than Axi-Pembro (16.6 months). Ipi-Nivo demonstrates a clear tail of the OS curve (below) around 30 months. Sufficient follow-up is not yet available to assess this for thus feature for Axi-Pembro. In contrast, Axi-Pembro has a clear benefit in the first 9 months over sunitinib, which is not observed for Ipi-Nivo.
Dr. Voss proposed that PD-L1 expression can help determine which first-line therapy a patient is likely to derive the most benefit from. Response rates to Ipi-Nivo, reported as a ratio of overall response to progressive disease rates, were significantly higher for patients with PD-L1 expression greater than or equal to 1% (4.14) compared to those with PD-L1 expression less than 1% (1.85). Dr. Voss synthesized this data to convlude that in patients who are PD-L1 positive, consider first-line Ipi-Nivo; in patients who are PD-L1 negative and have symptomatic disease, consider first-line Axi-Pembro.
Another readily-available biomarker in the clinic is the neutrophil to lymphocyte ratio (NLR). Low NLR (less than 4) is associated with significantly higher rates of metastasis-free survival and OS following nephrectomy. Low NLR was subsequently validated as a biomarker to predict better progression-free survival, OS, and overall response rate in patients with metastatic RCC. Further, this prognostic association was observed across various disease scenarios including non-metastatic patients as well as metastatic patients treated with a tyrosine kinase inhibitor (TKI) or immunotherapy. Dr. Voss concluded that low NLR can be utilized as a clinical biomarker for determining prognosis and factored into decision-making regarding a patient’s candidacy for active surveillance.
Biomarkers based on genomics, RNA signatures, and circulating analytes remain in the research phase and Dr. Voss cautioned about applying them to patients at this time. There are however, some promising molecular biomarkers to consider. Loss of function alterations in the gene PBRM1 is associated with improved OS in metastatic RCC for patients receiving VEGF TKI, mTOR inhibitors, and PD-1 inhibitors. Notably in the exploratory biomarker analysis of the IMmotion150 Trial, PBRM1 was validated as a positive predictive biomarker in patients receiving sunitinib, but not in patients receiving atezolizumab.
DNA damage repair (DDR) genes have also been evaluated as a predictive biomarker in metastatic RCC. A retrospective analysis of 107 patients with metastatic RCC treated with immunotherapy demonstrated that patients whose tumors carried a deleterious DDR alteration had significantly better OS (hazard ration = 0.29; 95% confidence interval 0.09-0.95; p = 0.04).
Gene expression signatures have been developed that predict response to VEGF TKI and/or immunotherapy in metastatic RCC. In the IMmotion150 Trial, a gene expression signature of angiogenesis was predictive of response to sunitinib, but not to atezolizumab plus bevacizumab or atezolizumab alone. Gene expression signatures reflecting the immune microenvironment were also found to be predictive. The “T-effector-High Myeloid-Low” signature was predictive of better response to an immunotherapy-based regimen than VEGF TKI alone. In contrast, “T-effector-High Myeloid-High” signature was associated with better response to combination VEGF inhibitor plus immunotherapy. 
Non-invasive biomarkers, such as circulating tumor DNA (ctDNA) and metabolites, are still in the early phases of investigation. Profiling of paired tumor tissue and cell-free DNA (cfDNA) from 110 patients demonstrated a low concordance rate. Focusing on mutations in VHL, the most commonly altered gene in RCC, mutations were detected in 93% of tissue samples and only 28% of cfDNA samples.
Kynurenine is a metabolite of the amino acid tryptophan. Kynurenine-to-tryptophan ratio is hypothesized to predict response to immunotherapy by altering the tumor microenvironment. Metabolite profiling of samples from the CheckMate 025 Trial demonstrated that a decreasing kynurenine to tryptophan ratio was associated with significantly improved OS among patients treated with nivolumab compared to those with increasing ratios.
Dr. Voss concluded that there are numerous promising biomarkers for the management of RCC. Readily-available clinical biomarkers (such as PD-L1 status and neutrophil to lymphocyte ratio) can be used at this time to help guide first-line treatment decisions. Many other genomic, gene-expression-based, non-invasive, and metabolite biomarkers demonstrate potential, but will require validation before incorporation into routine clinical practice.
Presented by: Martin Voss, MD, Medical Oncologist at Memorial Sloan Kettering Cancer Center
Written by: Jacob Berchuck, MD, Medical Oncology Fellow at the Dana-Farber Cancer Institute (Twitter: @jberchuck) at the 2020 Genitourinary Cancers Symposium, ASCO GU #GU20, February 13-15, 2020, San Francisco, California