De-escalation of Treatment in Metastatic Renal Cell Carcinoma: Why? How? Who? When?

The therapeutic landscape of metastatic renal cell carcinoma (mRCC) has evolved considerably with the introduction of immune checkpoint inhibitors (ICIs) and VEGFR-targeted therapies, leading to unprecedented improvements in survival. As durable responses and deep remissions become more frequent, the question of how to optimize treatment intensity has gained prominence. Therapeutic de-escalation, through reduced dose intensity, extended dosing intervals, treatment interruption, or integration of local therapies, has emerged as a promising approach to mitigate chronic toxicity, preserve quality of life, and reduce healthcare costs without compromising oncologic outcomes. Pharmacokinetic and pharmacodynamic data provide a strong biological rationale for immunotherapy de-escalation, with ICIs achieving sustained PD-1/PD-L1 receptor occupancy that persists for weeks after infusion. Historical observations, such as prolonged responses after nivolumab discontinuation, together with modern evidence from trials like OMNIVORE, TITAN-RCC, STAR, and ongoing initiatives (PDIGREE, MOIO), support the feasibility of treatment reduction in selected responders (Table 1). De-escalation may also be facilitated by consolidative local therapies, including metastasectomy, stereotactic radiotherapy, and delayed nephrectomy, that can increase complete response rates and prolong treatment-free survival. Identifying candidates for de-escalation requires integrating clinical response depth, metastatic patterns, and emerging biomarkers. Minimal residual disease assessment using circulating tumor DNA, as recently demonstrated in prospective studies, may refine patient selection and inform optimal timing for treatment interruption. Despite encouraging retrospective and early prospective data, robust randomized trials remain essential to define standardized strategies. Overall, treatment de-escalation represents a key step toward personalized, value-based care in mRCC.