Cytoreductive nephrectomy improves survival in metastatic papillary renal cell carcinoma: a population-based study with machine learning validation and landmark analysis.

BACKGROUND: The role of cytoreductive nephrectomy (CN) in metastatic papillary renal cell carcinoma (pRCC) remains poorly defined, as most CN evidence derives from clear cell RCC. Following the CARMENA trial, CN utilization has declined broadly, yet the applicability to pRCC is uncertain. METHODS: We identified 770 patients with metastatic pRCC (2010-2022) from a US population-based registry. Multivariable Cox regression, propensity score matching (PSM), competing risk analysis, and subgroup analyses were performed. Model validation included temporal split validation (training: 2010-2017; validation: 2018-2022), landmark analysis, and machine learning comparison (Random Survival Forest, Gradient Boosting Survival Analysis). RESULTS: CN was associated with improved OS (median 17 vs. 7 months, P < 0.001) and CSS (19 vs. 8 months, P < 0.001). After multivariable adjustment, CN independently predicted improved OS (HR = 0.61, 95% CI: 0.49-0.77) and CSS (HR = 0.56, 95% CI: 0.44-0.71; both P < 0.001), persisting after PSM (269 pairs). CN benefit was significant in single-organ metastasis (HR = 0.52, P < 0.001) but not multi-organ metastasis (HR = 0.75, P = 0.124), and was consistent across treatment eras (P interaction = 0.432). Temporal validation confirmed model stability (validation C-index: 0.662). Landmark analysis at 6 months demonstrated persistent CN benefit (HR = 0.68, P = 0.001), ruling out immortal time bias. CONCLUSIONS: CN confers a validated survival benefit in metastatic pRCC, particularly in single-organ metastasis. The declining CN utilization, potentially driven by extrapolation of CARMENA results to non-ccRCC subtypes, warrants reconsideration. Metastatic burden-based patient selection may be useful for informing clinical decision-making in pRCC.