PFS and ORR fall short as OS surrogates in metastatic RCCAnalysis shows limitations in using certain metrics to predict cancer survival

A large meta-analysis of 24,518 patients with metastatic renal cell carcinoma (RCC) evaluated whether progression-free survival (PFS) and objective response rate (ORR) can serve as reliable surrogate endpoints for overall survival (OS). The analysis found only moderate correlations: PFS-OS r = 0.52 (95% CI, 0.38-0.66) and ORR-OS r = -0.59 (95% CI, -0.72, -0.45). These correlations suggest that PFS and ORR explain only a portion of the variability in OS, limiting their surrogacy.

The surrogate threshold effect (STE) for PFS was a hazard ratio of 0.92 (95% CI, 0.79-1.09), indicating that a PFS benefit must be at least this large to predict an OS benefit. For ORR, the STE was an odds ratio of 1.40 (95% CI, 0.99-1.89). Both confidence intervals cross the null value, meaning the thresholds are not statistically significant. This undermines the reliability of these endpoints as predictors of OS benefit.

Subgroup analyses were frequently imprecise, especially for ORR and in smaller treatment-class subgroups. The meta-analysis also had limited representation of non-clear cell RCC, preventing reliable histology-specific conclusions. Substantial heterogeneity across included trials further complicates interpretation. These limitations suggest that the findings should be applied cautiously to specific patient populations.

From a clinical perspective, these results challenge the assumption that improvements in PFS or ORR reliably translate to improved OS in metastatic RCC. While PFS and ORR remain valuable for assessing treatment activity and guiding early clinical decisions, they should not be considered definitive surrogates for OS. Clinicians should interpret PFS and ORR benefits with caution, especially when OS data are immature or unavailable.

The study did not report safety outcomes, funding sources, or conflicts of interest. The association between surrogate endpoints and OS was assessed via correlation coefficients, which cannot establish causality. The certainty of evidence was not reported. Given the moderate correlations and non-significant STEs, PFS and ORR are not confirmed as reliable surrogates for OS in this population. Until more robust validation is available, OS remains the gold standard for assessing treatment benefit in metastatic RCC.

In practice, these findings support the continued use of OS as the primary endpoint in pivotal trials, while PFS and ORR can serve as supportive endpoints. For individual patient care, decisions should integrate multiple factors, including toxicity, quality of life, and patient preferences, rather than relying solely on PFS or ORR improvements. Future research should focus on identifying more reliable surrogate endpoints, possibly incorporating molecular or imaging biomarkers, and ensuring adequate representation of diverse histologies.

In conclusion, this meta-analysis demonstrates that PFS and ORR are not sufficiently consistent surrogates for OS in metastatic RCC. The moderate correlations and non-significant STEs highlight the need for caution when using these endpoints to infer survival benefits. Clinicians should prioritize OS data when available and interpret PFS and ORR improvements in the context of overall clinical benefit.