← Back to The Absolute Effect
Renal Cell Carcinoma · Post-ICI progression · Reviewed 14 October 2026

CONTACT-03

After progression on or after an immune checkpoint inhibitor (ICI), does continuing/rechallenging ICI by adding atezolizumab to cabozantinib improve outcomes compared with cabozantinib alone?

Population
522 patients with locally advanced or metastatic RCC progressing during or after prior ICI treatment; 53% had received ICI in first line.
Experimental arm
Atezolizumab 1,200 mg IV every 3 weeks + cabozantinib 60 mg orally daily.
Comparator
Cabozantinib 60 mg orally daily.
Design
Randomized, open-label phase III; co-primary endpoints PFS by blinded independent review and OS.

Key outcomes

Median PFS
10.6 vs 10.8 months

Atezolizumab + cabozantinib vs cabozantinib; HR 1.03 (95% CI 0.83–1.28), P=0.78.

Median OS
25.7 months vs not reached

HR 0.94 (95% CI 0.70–1.27), P=0.69.

Objective response rate
41% vs 41%

No response-rate gain from adding atezolizumab.

Serious adverse events
48% vs 33%

More serious toxicity with the combination.

Interpretation: previous first-line ICI matters

CONTACT-03 directly tested the clinically tempting strategy of continuing or rechallenging checkpoint blockade after ICI progression. Adding atezolizumab to cabozantinib did not improve PFS, OS or response and increased toxicity. These data argue against routine sequential ICI use outside a clinical trial.

Practice implication: If the patient has progressed after an ICI-containing first-line regimen, switching mechanism—often to an active VEGFR/MET-targeted agent such as cabozantinib when appropriate—is better supported than simply adding another ICI. This is a trial-level conclusion, not a rule for every individual clinical scenario.

Critical appraisal

The randomized design is a major strength and the control arm performed well. The study included patients with different prior ICI regimens and lines; it was not a head-to-head comparison of all possible post-ICI options. Treatment-related toxicity and adverse events leading to death were higher in the combination arm. The result should be considered alongside TiNivo-2, which similarly failed to show benefit from adding nivolumab to tivozanib after prior ICI exposure.

No NNT is calculated because the primary results did not demonstrate benefit and time-specific absolute event rates are not used here.

Current-practice context

As of 14 October 2026, CONTACT-03 is key negative randomized evidence against routine ICI rechallenge after progression on prior ICI in advanced RCC. Select subsequent therapy based on prior VEGFR TKI exposure, IMDC risk, tempo of progression, comorbidities, access and current guidelines.

Dated context added 14 October 2026; full guideline audit remains pending.

Sources