Summit/Akeso start SMT112-216, a global first-line ccRCC Phase 2 of ivonescimab: an exploratory design that screens combinations by ORR

Data as of 2026-10-21 (Taipei time). [Fact] statements are sourced; [Inference] statements are this article's judgment.

1. One-line conclusion

[Fact] SMT112-216 is a Phase 2 of about 100 patients in five arms, open-label, with investigator-assessed ORR as its primary endpoint (CT.gov). [Inference] It can answer which ivonescimab combination deserves Phase 3, not whether it can beat nivo + ipi. In the near term, what will most shape the kidney cancer strategy is FDA's 2026-11-14 decision on HARMONi (EGFR-mutant NSCLC) (Summit).

2. What happened

[Fact] On 2026-10-19 Summit announced site activation for SMT112-216 (AK112-216 / NCT07877194). It is a randomized, open-label, multiregional Phase 2 in locally advanced or metastatic ccRCC with no prior systemic therapy for advanced disease, planning about 100 patients, with sites in the US and China (Summit). Akeso issued a joint release the same day, noting that nivo + ipi and pembrolizumab + axitinib are both international first-line standards (Akeso).

• Arms: the release lists four regimens: ivonescimab + cadonilimab (a PD-1×CTLA-4 bispecific), ivonescimab + ipilimumab, ivonescimab monotherapy, and nivolumab + ipilimumab as the standard-of-care control (Summit).
• Primary endpoints: investigator-assessed ORR per RECIST 1.1, and adverse events; secondary endpoints include DCR, DoR, PFS, OS, PK and ADA (CT.gov).
• Timeline: primary completion 2029-04-12; 2 sites currently registered (Los Angeles, Beijing); status not yet recruiting (CT.gov).
• A second track: Summit is separately partnering with Arcus to test ivonescimab + casdatifan (a HIF-2α inhibitor) in a similar ccRCC setting (Summit). That trial's design and size were not found.
• This is a newly opened trial; there is no HR, CI or p value. Per-arm doses and randomization ratio: not found.

3. Why it matters

[Inference] The design's strength is placing nivo + ipi in the same randomized framework, avoiding comparison of single-arm ORR with historical data. It has three weaknesses.

• Endpoint: [Inference] investigator-assessed ORR in an open-label trial (CT.gov) is prone to overestimation. [Fact] In first-line ccRCC, IO + VEGF-TKI ORRs are already high: 59.3% for pembrolizumab + axitinib in KEYNOTE-426 (NEJM) and 71.0% for lenvatinib + pembrolizumab in CLEAR (independent review committee, NEJM full text Table 2; NEJM). This trial has no TKI arm; these figures are the bar a Phase 3 control will set. [Inference] ORR differences do not necessarily translate into OS.
• Sample size: [Fact] about 100 patients across five arms (CT.gov). [Inference] No pairwise comparison is adequately powered; only direction can be read.
• Choice of comparator: [Inference] ivonescimab blocks both PD-1 and VEGF, so clinically it challenges IO + TKI combinations (e.g., pembrolizumab + axitinib or lenvatinib). Using only nivo + ipi as the control means Phase 3 will still face an IO + TKI comparison.
• Safety: [Inference] dual IO (adding CTLA-4) on top of VEGF blockade may compound immune-related and vascular toxicities; the cadonilimab combination is split into two ivonescimab doses (CT.gov), possibly to find a tolerable dose. There are no ccRCC safety data yet.
• Region: enrollment in both the US and China. [Inference] One of FDA's concerns with HARMONi is whether Western and Asian patients have consistent results (Summit WCLC); a kidney cancer Phase 3 dominated by Chinese patients would face the same issue.

4. Regulatory and commercial implications

• FDA: [Fact] HARMONi's prespecified primary OS analysis did not reach statistical significance (HR 0.79, 95% CI 0.62–1.01; p=0.057); an updated analysis gave HR 0.76 (nominal p=0.0151) (Summit WCLC). [Inference] How FDA handles these OS data will set its overall stance on ivonescimab. If approval comes with positive PFS and OS short of prespecified significance, development costs for the bispecific in other tumor types fall; if rejected, Summit's GU Phase 3 trials would likely need OS as a primary or key secondary endpoint; that depends on FDA's reasons, which only the full response letter would show.
• NMPA: ivonescimab was approved in China in May 2024 (Summit); plans for a kidney cancer indication in China: not found.
• Competition and SoC: [Inference] HIF-2α inhibitors moving into earlier lines (belzutifan and casdatifan) will reshape first-line ccRCC combinations; Summit betting on both IO combinations and a HIF-2α combination is a hedge.
• For the next trial: [Inference] sponsors planning ccRCC combination trials should not lock Phase 3 endpoints before 11/14; once FDA decides, choose whether OS is the primary or key secondary endpoint.
• GU footprint: Summit also has HARMONi-GU1, a Phase II/III of ivonescimab + EV versus pembrolizumab + EV in first-line urothelial carcinoma, with global site activation planned to start in Q4 2026 (Summit).

5. Lessons for the next trial design and OncoLattice recommendations

All recommendations below are [Inference]; cited facts carry their own sources.

1. A combination-selection Phase 2 should prespecify go/no-go rules: e.g., an ORR difference versus nivo + ipi and a ceiling for grade ≥3 irAEs, with BICR confirmation of response; otherwise Phase 2 results carry no weight in a Phase 3 filing.
2. Align the Phase 3 control with IO + TKI, not only nivo + ipi.
3. Prespecify stratification: IMDC risk, sarcomatoid differentiation, PD-L1. IMDC intermediate/poor risk is where nivo + ipi performs best (CheckMate 214 primary population ORR 42% vs 27% for sunitinib, NEJM); without stratification, differences cannot be interpreted.
4. Site selection: first-line kidney cancer patients are mostly seen in urologic oncology or medical oncology clinics; select sites with post-nephrectomy referral and a multidisciplinary team able to manage dual-IO toxicities (endocrine, hepatitis, colitis).
5. Regional balance: set a minimum proportion of Western patients upfront to avoid HARMONi's regional-consistency dispute.

6. Unknowns and what to watch

• The HARMONi PDUFA outcome on 11/14 and whether review is extended.
• SMT112-216 per-arm doses, randomization ratio and interim analysis rules.
• The design of the ivonescimab + casdatifan trial.
• The actual start of HARMONi-GU1.

Evidence limitations

SMT112-216 has no results; this article's strengths-and-weaknesses assessment is design-level inference. HARMONi figures come from company releases, not peer-reviewed full text. No official document was found on FDA's OS requirement.

Endnote: differences from the original forwarded account

1. ClinicalTrials.gov lists five arms: the cadonilimab combination is split into ivonescimab dose 1 and dose 2 (CT.gov). The release says four regimens; the two are not contradictory, but "4 arms" is imprecise. Each arm averages only about 20 patients.
2. This article could not find a direct FDA statement in the Summit release or Fierce coverage that "FDA requires statistically significant OS"; it can only be labeled unverified. What can be confirmed: the HARMONi PDUFA date is 2026-11-14 (Summit BLA-acceptance release of 2026-01-29; Summit); the prespecified primary OS analysis was HR 0.79 (95% CI 0.62–1.01; p=0.057), not statistically significant; the June 2026 updated analysis was HR 0.76 (95% CI 0.61–0.95; nominal p=0.0151) (Summit WCLC). Fierce speculated that the update could constitute a major amendment and extend review by three months; this is media inference (Fierce Pharma).