Background: ICI-based regimens are the standard of care for first-line (1L) treatment of metastatic clear cell (cc) RCC. Treatment options following disease progression during or after ICI therapy are limited but can include single-agent TKIs, such as cabozantinib (cabo). CONTACT-03 evaluated anti–PD-L1 atezolizumab (atezo) + cabo vs cabo alone in patients (pts) with metastatic RCC that progressed during or after prior ICI treatment and is the first phase 3 randomized trial to test the benefit of ICI rechallenge by direct addition to a control arm. Methods: CONTACT-03 enrolled pts with histologically confirmed, inoperable, locally advanced or metastatic cc or non-cc RCC, regardless of PD-L1 status, that progressed on or after ICI treatment. Randomization was 1:1 to atezo (1200 mg IV q3w) plus cabo (60 mg oral qd) or cabo alone. Stratification factors were IMDC risk factors (0 vs 1-2 vs ≥3); most recent line of prior ICI therapy (adjuvant vs 1L vs 2L); and histology (dominant cc without sarcomatoid vs dominant non-cc [papillary or unclassified] without sarcomatoid vs cc or non-cc with any sarcomatoid component). The multiple primary efficacy endpoints were centrally reviewed RECIST 1.1 PFS and OS. Key secondary endpoints were investigator (INV)-assessed PFS, centrally reviewed RECIST 1.1 ORR and DOR and safety. Results: Of 522 pts randomized to atezo + cabo (n=263) or cabo (n=259), 55% and 51% had most recent ICI in the 1L setting and 10% and 11% had sarcomatoid RCC, respectively. At the data cutoff (Jan 3, 2023), median follow-up was 15.2 mo. No PFS or OS benefit was observed with atezo + cabo vs cabo. ORR was 41% in both arms; DOR was 12.7 (95% CI: 10.5, 17.4) mo with atezo + cabo and 14.8 (95% CI: 11.3, 20.0) mo with cabo. All-cause Grade 3/4 adverse events (AEs) occurred in 68% (177/262) and 62% (158/256) of safety-evaluable pts receiving atezo + cabo or cabo, respectively; all-cause Grade 5 AEs occurred in 6% and 4%. AEs leading to treatment withdrawal occurred in 16% of pts on atezo + cabo and 4% on cabo. Conclusions: The addition of atezo to cabo did not improve clinical outcomes and led to increased toxicity in patients with RCC that progressed on or after prior ICI treatment. CONTACT-03 is the first randomized, phase III oncology trial to test the benefit of PD-(L)1 inhibitor continuation by direct addition to a standard control arm; the results prompt caution with this approach in other cancers. Clinical trial information: NCT04338269.

NE, not evaluable. ^a Interim analysis.