Cadonilimab

Cadonilimab (AK104, marketed as Kaitanni in China) is Akeso's bispecific antibody that hits both PD-1 and CTLA-4 with a single molecule [1]. It is already approved in China for three indications: second-line cervical cancer (June 2022) [2], first-line HER2-negative gastric/GEJ adenocarcinoma (2024) [3], and first-line cervical cancer (2025 label expansion, all-comer population) [12]. Akeso is now running global Phase 3 programs in NSCLC consolidation, hepatocellular carcinoma, and other solid tumors to broaden the label outside China [11]. The strategic bet: deliver the efficacy of the ipilimumab+nivolumab combination with less of its punishing toxicity, in a single infusion at a cheaper price point.

Cadonilimab is approved for three indications in China but remains investigational in the US and EU, so for Western markets it is treated as a novel compound. It carries no FDA breakthrough, fast track, or orphan designation that I can verify, and Akeso has not publicly announced an FDA pre-IND or Type B meeting outcome for cadonilimab specifically (absence of disclosure ≠ no engagement). The Phase 3 NSCLC consolidation trial referenced in this node (NCT06617416) compares cadonilimab against sugemalimab in patients with unresectable Stage III disease who have completed concurrent chemoradiation - a setting globally dominated by AstraZeneca's durvalumab off the PACIFIC trial. The strongest pillar in the package is COMPASSION-16 in first-line cervical cancer: cadonilimab + platinum chemo ± bevacizumab vs placebo + chemo ± bevacizumab, n=445, median PFS 12.7 vs 8.1 months (HR 0.62, 95% CI 0.49-0.80, p<0.0001), median OS not reached vs 22.8 months (HR 0.64, 95% CI 0.48-0.86, p=0.0011) [4][13]. Those numbers track essentially identically to pembrolizumab + chemo in KEYNOTE-826 (PFS HR ~0.62, OS HR ~0.67) - equivalence, not superiority, but the all-comer label (no PD-L1 cutoff) is a meaningful differentiator. The gastric pillar is COMPASSION-15: cadonilimab + chemo vs chemo in HER2-negative gastric/GEJ adenocarcinoma, n=610, median PFS 7.0 vs 5.3 months (HR 0.53, 95% CI 0.44-0.65), median OS 15.0 vs 10.8 months in the ITT, with a deeper PD-L1 CPS≥5 OS benefit (15.3 vs 10.9 months, HR 0.58) [14]. CheckMate-649 with nivolumab + chemo delivered ITT OS HR ~0.80 - cadonilimab's HR is more favorable in this comparison, though cross-trial caveats apply. Phase 2 readouts in endometrial cancer (interim 2026 [5]) and protocols for HCC conversion therapy with lenvatinib and stereotactic radiation are ongoing [6][7]. Expected readout for the NSCLC consolidation Phase 3 is not publicly disclosed; based on a Phase 3 launching in 2024 with PFS as a typical primary endpoint, a 2027-2028 interim is realistic, not 2026.

Two brakes on the immune system, one antibody. PD-1 sits on the surface of T cells and acts as an exhaustion brake - tumors that pump out PD-L1 squeeze that brake to shut T cells down. CTLA-4 is a different brake, applied earlier when T cells are first being trained in lymph nodes; blocking it lets more aggressive T cells get into circulation. The ipilimumab+nivolumab combo (CTLA-4 + PD-1, two separate drugs) is approved in melanoma and several other cancers and works better than either alone, but it causes substantial immune toxicity - in CheckMate-067 the combination produced ~55% Grade 3/4 treatment-related adverse events of any type, with treatment-related discontinuations of 36.4% (vs 7.7% for nivolumab monotherapy) [15]. Cadonilimab is engineered to bind PD-1 tightly and CTLA-4 weakly, with a modified Fc region that reduces systemic CTLA-4 engagement. The pitch is that the molecule preferentially activates T cells that are already in tumor tissue (where PD-1 is highly expressed on exhausted infiltrating T cells), concentrating the CTLA-4 effect locally. The mechanism is well-validated in principle - checkpoint dual blockade works - but whether the bispecific format actually improves the therapeutic window versus the cheap generic ipi+nivo combination (both small-molecule-priced once biosimilars arrive) is the real question.

NCT06617416 is a Phase 3 head-to-head of cadonilimab versus sugemalimab as consolidation therapy in unresectable, locally advanced NSCLC patients who have not progressed after concurrent chemoradiation. The comparator choice is telling - sugemalimab (CStone PD-L1, approved in China for this setting via the GEMSTONE-301 trial) is the China standard, not the global standard durvalumab. That makes this a China-centric registration trial, not a global label-expansion play. Primary endpoint and exact enrollment for NCT06617416 are not detailed in the available node data and are not estimated here. Separately, Akeso has the COMPASSION program in cervical (COMPASSION-16 Phase 3 positive readout, supported the 2025 first-line label expansion [4][12]) and Phase 2/3 work in HCC combination regimens [6][7]. NCT07400536 is a Phase 3 in locally advanced cervical cancer with chemotherapy followed by CCRT, n=378, primary endpoint PFS [8] - note this enrollment number refers to NCT07400536, not the primary NSCLC trial NCT06617416. The cervical and gastric data are the strongest pillars; NSCLC and HCC are the expansion bets that determine global commercial relevance.

Our model gives this drug an 11% chance of eventually being approved. That figure starts from a historical base rate of about 48% for Phase 3 drugs in this area, then adjusts based on ten specific facts about the trial and its sponsor. The estimate is pulled down mainly by heavier-than-usual blinding, the sponsor's thin or weak approval record, weak earlier-phase results, and a randomized trial design. The remaining factors are close to average for this stage, so they don't shift the number much further.

Efficacy: cadonilimab needs to beat sugemalimab on PFS or OS to register in China and would still need to clear a durvalumab-equivalent bar to matter outside China. Sugemalimab itself was approved on a relatively modest PFS benefit; the room above it is narrow. Safety: even with the engineered low-affinity CTLA-4 arm, immune-related adverse events (irAE) - colitis, hepatitis, hypophysitis, pneumonitis - are mechanism-based and will not go to zero. The CheckMate-067 ipi+nivo Grade 3+ irAE rate specifically (~36% per published analyses) is the relevant safety bar cadonilimab needs to beat, not the higher all-cause Grade 3/4 TRAE figure [15]. In the cervical first-line setting, Grade 3+ irAE rates need to stay competitive with PD-1 monotherapy or the toxicity penalty kills uptake. Commercial: this is the harder problem. Akeso has no US commercial infrastructure and no announced US partner for cadonilimab specifically (the Summit Therapeutics deal covers ivonescimab, a different bispecific against PD-1/VEGF [10]). To monetize cadonilimab in the West, Akeso needs a partnership or has to build out. In China, both cadonilimab and ivonescimab were included in the National Reimbursement Drug List (NRDL - the central government formulary that sets the prices Chinese insurers pay) during the 2024 negotiations [16], which compresses per-dose pricing but unlocks volume. Akeso reported 2024 commercial sales of RMB 2,002 million (+24.88% YoY from RMB 1,604 million in 2023) [16]; the company does not break out Kaitanni revenue separately from ivonescimab at the product level. IP angle: ipilimumab loses key patent protection in the 2026-2028 window in major markets, and nivolumab around 2028-2030 - if biosimilar ipi+nivo becomes available cheaply, the cost-advantage half of cadonilimab's pitch erodes, leaving the safety/convenience argument carrying more weight.

Worth watching, but the watchable data point is not the NSCLC consolidation trial - it is whether Akeso announces a Western commercialization partner for cadonilimab and whether the COMPASSION-16 cervical Phase 3 data package gets filed with the FDA. The ivonescimab/Summit deal showed the market is hungry for differentiated Chinese bispecifics; cadonilimab is older, more clinically de-risked, and arguably underpriced in that frame. The Akeso commercial sales trajectory through 2026 is also informative - 2024 grew ~25% YoY across the whole franchise [16]; sustained or accelerating growth in 2026 with first-line gastric and first-line cervical now both in the label would signal real uptake, while deceleration would indicate NRDL pricing pressure dominating volume gains. Check back when (1) any FDA-related announcement appears (pre-IND meeting outcome, BLA filing acceptance, breakthrough designation), (2) the COMPASSION-16 paper reads out longer follow-up data, or (3) Akeso reports half-year 2026 results with product-level revenue breakout. Until then, this is a China story with global optionality, not a global story.