In kidney cancer, progression-free survival is a weak stand-in for overall survival

Design

Systematic review and trial-level meta-analysis with surrogate-threshold analysis

Population

62 randomised trials, 24,518 patients with metastatic renal cell carcinoma

Primary outcome

Correlation of progression-free survival and response with overall survival

Effect

PFS–OS r = 0.52 (0.38–0.66); ORR–OS r = −0.59; surrogacy not reliably met

Many metastatic renal cell carcinoma drugs are approved on progression-free survival or response rate because overall survival takes longer and is confounded by later therapies. This meta-analysis of 62 trials and 24,518 patients tested how well those endpoints actually predict survival.

Both were only moderately correlated with overall survival — progression-free survival r = 0.52, objective response r = −0.59 — and the surrogate-threshold analyses did not reach a level that would reliably predict a survival benefit. Associations varied by treatment class and line of therapy and were often imprecise, with too few non-clear-cell cases to judge that histology.

The practical consequence is interpretive. When a new kidney-cancer drug reports a progression-free survival gain without mature overall-survival data, treat the survival benefit as unproven rather than assumed. It argues for keeping overall survival as the anchor in counselling and for cautious reading of accelerated approvals.

• Progression-free survival correlated only moderately with overall survival (r = 0.52, 95% CI 0.38–0.66).
• Objective response correlated moderately too (r = −0.59), and neither met a reliable surrogate threshold.
• Surrogate validity varied by treatment class and line of therapy and was frequently imprecise.
• Read a progression-free-survival-only kidney-cancer result as an unproven survival benefit until overall-survival data mature.

Why it matters

It cautions against equating the progression-free survival gains that drive many approvals with living longer.

Don't overread it

Trial-level surrogacy with substantial heterogeneity — it questions reliability across the field, not any single drug's individual survival evidence.

The statistics, in plain English

A correlation of 0.52 is moderate, not strong — meaning a drug can improve progression-free survival without a matching survival gain. The surrogate-threshold test asks how large a progression-free effect must be to predict survival benefit, and here that bar was not reliably cleared.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.