Immunological clustering from peripheral blood to define prognosis and T cell exhaustion profiles in renal cell carcinoma.

Abstract

536 Background: Peripheral blood markers such as neutrophil-to-lymphocyte ratio (NLR) and CRP are established prognostic factors in renal cell carcinoma (RCC). However, their association to the tumor immune microenvironment (TME) remains unclear. We developed a novel leukocyte-based clustering system combining NLR and minor leukocyte fractions, and investigated its prognostic and immunological significance. Methods: We retrospectively analyzed 115 treatment-naive RCC patients who underwent nephrectomy or partial nephrectomy between November 2018 and July 2023 at Tokyo Women’s Medical University and Adachi Medical Center. Overall Survival (OS) and RFS (Recurrence Free Survival) were evaluated using Kaplan–Meier analysis, and hazard ratios were estimated by the Cox proportional hazards model. Resected tumor specimens were subjected to flow cytometric analysis to assess tumor-infiltrating immune cells. Patients were classified into three clusters using a novel leukocyte-based stratification: Group1 (NLR<3 and monocyte + eosinophil + basophil <8%), Group2 (NLR<3 and monocyte + eosinophil + basophil ≥8%), and Group3 (NLR≥3).Furthermore, we applied this classification to evaluate its utility in predicting prognosis and guiding the indication of immune checkpoint inhibitors (ICI) in patients with metastatic renal cell carcinoma. Results: This classification clearly stratified outcomes. The median RFS was not reached in Groups 1 and 2, and no significant difference was observed between them, whereas Group 3 showed a significantly shorter RFS (p = 0.0009). Group 1 showed the longest OS, while Group 3 had the shortest (p = 0.0243).Furthermore, multivariate analysis suggested that the leukocyte-based cluster was an independent prognostic factor (p=0.0059).Expression of exhaustion markers was highest in Group3, including PD-1 (p=0.0280) and LAG3 (p=0.0227), suggesting an association with T cell dysfunction. Applying this system to 295 patients receiving first-line ICI-based regimens, OS and PFS (Progression Free Survival) were significantly shorter in Group3 treated with IO-IO therapy (OS: p=0.0011, PFS: p=0.0003). Conclusions: We propose a simple, non-invasive clustering system integrating NLR and minor leukocyte fractions, which reflects TME immune status, predicts prognosis in RCC, and may guide selection of ICI-based therapy. Multivariate analysis of RFS. HR (95% CI) p value Age 1.02 (0.97 – 1.08) 0.5067 Sex Male reference 0.4491 Female 0.59 (0.15 – 2.37) 0.4491 Histological Type Clear cell reference 0.3372 Others 0.34 (0.44 – 3.00) 0.3372 Pathological stage I/II reference 0.0009 III 8.78 (2.01 – 38.21) 0.0038 Peripheral blood cluster EMB low 8.28 (1.03 – 66.78) 0.0059 EMB high reference 0.0471 NLR high 7.78 (1.64 – 36.99) 0.0099