Ivonescimab

The sponsor table retains randomized arm sizes of 219. The primary publication reports safety denominators of 218 treated participants per arm.

The primary PFS analysis included 345 participants, not all 438 randomized participants. The primary OS comparison did not meet its significance boundary. The later HR of 0.76 is a follow-up result and does not replace the primary OS test. Intracranial findings were exploratory and outside the multiplicity-controlled hierarchy. Because the analysis included third-generation TKI-treated participants from HARMONi-A, the two trials should not be pooled as independent populations. The comparator was chemotherapy alone, not another active post-TKI combination.

All 322 participants were enrolled at sites in China, which limits direct geographic generalization. The initial publication reported PFS; the subsequent final OS publication should be considered separately rather than treating early survival immaturity as a continuing limitation. The final OS comparison favored ivonescimab, but grade 3 or higher TEAEs were reported in 67.1% versus 54.7%. HARMONi incorporates an overlapping third-generation TKI-treated population, so its findings are not wholly independent replication.

The EGFR-mutated cohort contained 19 participants in an open-label study without a concurrent randomized comparator. Its ORR confidence interval was 43.4% to 87.4%, and the PFS upper confidence limit was not estimable. These results do not establish a comparative survival benefit. The study-wide safety percentage is not an EGFR-cohort-specific estimate.

This trial compares amivantamab-based combinations with chemotherapy, not with ivonescimab. Its PFS results therefore cannot establish the relative efficacy of amivantamab and ivonescimab. Toxicities span hematologic, EGFR-related, and MET-related categories; a PFS comparison alone does not summarize the full benefit-risk balance.

The PFS P value of 0.0122 did not meet the prespecified one-sided boundary of 0.0117. Neither primary efficacy endpoint was met. These results concern pembrolizumab plus chemotherapy, not ivonescimab, and cannot determine the effect of simultaneous PD-1 and VEGF inhibition.

The enrollment target was reduced from 500 to 270 participants, and the estimated power with observed PFS events was 76%. The open-label study did not meet its primary PFS endpoint. Grade 3 or 4 TRAEs were reported in 44.7% versus 29.4%. The trial evaluates nivolumab, not ivonescimab.

There were 49 participants per arm, and eligibility required non-CNS progression. The OS hazard-ratio confidence interval included 1, so the numerical median OS difference does not establish an OS benefit. This study does not directly compare osimertinib continuation with ivonescimab.

The 137-participant population included multiple genomic subgroups; 56.9% had EGFR mutations. The overall ORR and safety percentages should not be presented as EGFR-specific estimates. The EGFR-subgroup ORR of 43.6% is not a randomized comparison with ivonescimab.

This single-participant case cannot estimate a comparative treatment effect or establish that the reported resistance pattern predicts benefit. Disease control during monotherapy and subsequent combination treatment should not be conflated. The qualitative safety description does not provide an adverse-event rate.

The report describes one participant with adenosquamous histology and multiple previous therapies. A response sustained over six cycles cannot establish an OS benefit, comparative efficacy, or the frequency of uncommon toxicity. The findings should not be generalized to all EGFR-mutated NSCLC.

This phase 1b evidence concerns monotherapy, not the chemotherapy combination evaluated in the pivotal trials. Grade 3 or higher TRAEs occurred in 22.2%, and TRAEs leading to death were reported in 2.8%. The monotherapy response estimate should not be substituted for post-TKI combination-treatment efficacy.

The study enrolled 51 participants with advanced solid tumors, with 47 contributing a postbaseline response assessment. Descriptive dose-escalation findings do not establish efficacy specifically in EGFR-mutated NSCLC. The publication gives different enrollment end dates in its abstract and results section; neither date has been silently substituted for the other. The maximum tolerated dose is specific to this study and schedule.

Only 1 of 59 participants, 1.7%, had NSCLC. These descriptive safety and pharmacology findings cannot establish efficacy for EGFR-mutated NSCLC. Unlike the Australian phase 1a study, this study did not reach a maximum tolerated dose; the results should not be merged into a single universal maximum-dose conclusion.