Akeso, Inc. (Hong Kong Stock Exchange: 9926) has announced the publication of Phase 2 clinical results showing that its PD-1 and VEGF bispecific antibody ivonescimab, combined with chemotherapy, achieved a 70.1% major pathological response rate among patients undergoing surgery for resectable non-small cell lung cancer. Published in Nature Medicine on 4 November 2026 and highlighted by Akeso on 5 November, the study also reported a 50.7% pathological complete response rate in the combination-treatment surgical population.
The findings provide peer-reviewed evidence of antitumour activity for ivonescimab in an earlier lung cancer treatment setting. However, the study was non-randomised and non-comparative, and the reported pathological response percentages were calculated among patients who completed surgery rather than every patient initially enrolled. The trial therefore provides a promising clinical signal without establishing that ivonescimab improves survival or is superior to existing perioperative immunotherapy regimens.
For Akeso and its international development partner Summit Therapeutics, the publication extends the scientific rationale for investigating ivonescimab beyond advanced lung cancer. The central development question is whether combining immune checkpoint inhibition with vascular endothelial growth factor blockade can generate durable benefits when treatment begins before potentially curative surgery.
What do the published Phase 2 ivonescimab results establish in resectable non-small cell lung cancer?
The multicentre Phase 2 study, registered as NCT05247684, enrolled 90 patients with previously untreated, resectable Stage II to Stage IIIB non-small cell lung cancer at two Chinese hospitals. The trial evaluated ivonescimab alone or in combination with platinum-based chemotherapy before surgery, followed by postoperative ivonescimab treatment.
The monotherapy cohort enrolled 11 patients, while the combination cohort included 79 patients. Participants received three or four cycles of neoadjuvant treatment before planned surgery, followed by additional ivonescimab after resection where appropriate.
In the combination cohort, 67 patients underwent surgery and were included in the completed-surgery analysis. Major pathological response, defined as 10% or less residual viable tumour in the resected specimen, was achieved in 47 of these patients, representing 70.1%.
Pathological complete response, reflecting the absence of residual viable tumour in the assessed surgical specimen, occurred in 34 patients, representing 50.7% of the surgical population. The study also reported that 46 of 67 patients, or 68.7%, had 5% or less residual viable tumour.
These results are noteworthy because pathological response measures provide an indication of how extensively tumour tissue has responded to treatment before surgery. Achieving a major pathological response or complete pathological response can be associated with favourable longer-term outcomes in lung cancer, although the relationship does not mean that every patient achieving such a response will remain disease-free.
The primary endpoint was met in the combination cohort according to the study’s prespecified statistical criterion. Investigators compared the observed major pathological response rate against a historical benchmark of 41%, with the lower bound of the specified 80% confidence interval required to exceed that threshold.
This is an important distinction from a randomised trial showing superiority over a contemporaneous control group. The findings support further evaluation of ivonescimab but do not establish that it produces better outcomes than currently authorised perioperative treatment combinations.
Why does the difference between surgical response rates and all enrolled patients matter?
The denominator used to calculate pathological response is particularly important in neoadjuvant oncology research. A patient who does not proceed to surgery may not have a resected tumour specimen available for the pathological assessment that determines whether the endpoint has been achieved.
In Akeso’s combination cohort, 79 patients began treatment, but only 67 underwent surgery. The published 70.1% major pathological response rate therefore reflects 47 responders among the 67 surgical patients rather than 47 responders among all 79 enrolled participants.
Expressed against the entire combination cohort, the 47 patients achieving major pathological response represent approximately 59.5% of those initially enrolled. Similarly, the 34 patients achieving pathological complete response represent approximately 43.0% of the enrolled combination population.
These alternative calculations do not invalidate the study’s prespecified endpoint analysis. They illustrate why results expressed in different analysis populations should not be compared without examining the definitions and treatment pathways involved.
Across both cohorts, 77 of the 90 enrolled patients proceeded to surgery. Reasons for non-completion included patient decisions, disease progression, deaths considered unrelated to study treatment and other medical or surgical factors.
The study reported that one participant experienced a treatment-related surgical delay. No operations were reported as cancelled because of treatment-related adverse events.
Among patients who underwent resection, all achieved an R0 resection, meaning that no residual tumour was identified at the surgical margins. This is an encouraging surgical finding, but it applies specifically to the patients who reached surgery.
From a clinical development perspective, the ability to maintain surgical eligibility while achieving tumour regression is important. A perioperative treatment must not only produce antitumour activity but also allow patients to undergo the planned operation without unacceptable toxicity or delay.
How does ivonescimab’s dual PD-1 and VEGF mechanism differ from existing immunotherapy?
Ivonescimab is a bispecific antibody designed to inhibit programmed cell death protein 1, or PD-1, and vascular endothelial growth factor, or VEGF, through a single therapeutic molecule. PD-1 inhibition is intended to restore immune activity against tumour cells, while VEGF blockade targets signalling associated with tumour blood vessel formation.
The biological rationale rests partly on the relationship between abnormal tumour vasculature and an immunosuppressive tumour microenvironment. VEGF activity can contribute to dysfunctional blood vessels and affect the movement and function of immune cells within tumours.
Blocking VEGF alongside PD-1 may potentially improve conditions for antitumour immune activity. However, mechanistic synergy observed in experimental systems does not automatically establish greater clinical efficacy than a conventional immune checkpoint inhibitor.
This distinction is particularly relevant in resectable lung cancer because perioperative immunotherapy has already become an established treatment strategy for selected patients. Regimens involving immune checkpoint inhibitors combined with chemotherapy before surgery, followed by postoperative immunotherapy, have demonstrated benefits in randomised Phase 3 trials.
The competitive question for ivonescimab is therefore more demanding than demonstrating activity against untreated disease. It must eventually show that its efficacy, safety or treatment profile provides a clinically meaningful advantage over existing approaches.
The Phase 2 results support that investigation, but their non-comparative design prevents reliable superiority claims. Differences in patient characteristics, tumour stage, chemotherapy regimens and follow-up can substantially affect apparent treatment outcomes across separate studies.
The addition of VEGF inhibition also introduces safety considerations relevant to surgery. Agents affecting angiogenic pathways may create concerns involving bleeding, wound healing and other complications, depending on their specific pharmacology and administration.
Consequently, perioperative development requires a particularly careful balance between antitumour activity and the safety of planned surgical treatment.
What do the squamous histology and PD-L1 subgroup findings suggest about patient selection?
The study reported substantial differences in pathological response between patients with squamous and nonsquamous non-small cell lung cancer. In the combination cohort’s surgical population, the major pathological response rate reached 81.5% among patients with squamous disease, compared with 23.1% among those with nonsquamous histology.
Pathological complete response was achieved in 61.1% of patients with squamous cancer, compared with 7.7% of those with nonsquamous disease. These findings raise the possibility that tumour histology may influence responsiveness to the ivonescimab-based combination.
However, the subgroup sizes were unequal. The surgical analysis included 54 patients with squamous cancer and only 13 with nonsquamous disease, creating substantial uncertainty around the smaller subgroup estimate.
The study was also not designed to establish a definitive interaction between histology and treatment benefit. Such a conclusion would require appropriately powered research and formal evaluation of whether the treatment effect genuinely differs between patient groups.
The PD-L1 findings present another potentially important research question. Major pathological response was reported in 69.2% of patients with PD-L1 expression below 1%, the same percentage observed among patients with expression of at least 1%.
This suggests that antitumour activity was observed even among patients traditionally considered less likely to benefit from some immune checkpoint strategies. Nevertheless, pathological complete response differed between the PD-L1 subgroups, and the exploratory findings cannot establish that PD-L1 status is irrelevant to treatment selection.
The population also included a substantial proportion of patients with Stage III disease. Among 55 surgical patients with Stage III cancer in the combination cohort, major pathological response was 65.5% and pathological complete response was 47.3%.
These results may help guide future trial design, particularly around histology and disease stage. They do not yet support a validated predictive biomarker strategy or a change in routine clinical treatment recommendations.
Does the 83.3% one-year event-free survival rate establish a durable benefit?
The investigators reported estimated 12-month event-free survival of 83.3% in the combination cohort and 81.8% in the monotherapy cohort. These estimates provide preliminary information about disease control after treatment and surgery.
However, median event-free survival had not been reached at the August 2025 data cutoff. Median follow-up was approximately 24.15 months in the combination cohort and 36.70 months in the smaller monotherapy cohort.
The difference in follow-up duration complicates comparisons between cohorts. More importantly, the study did not include a randomised standard-of-care control arm that could establish the relative effect of ivonescimab on recurrence or survival.
Event-free survival in perioperative lung cancer research may include disease progression that prevents surgery, recurrence after treatment or death, depending on the protocol definition. These outcomes are distinct from pathological response, which is assessed using tumour tissue at surgery.
A high pathological complete response rate can strengthen confidence in antitumour activity, but it is not proof of an overall survival advantage. Longer follow-up is necessary to determine whether tumour regression translates into sustained reductions in recurrence and other clinically important events.
The publication also includes circulating tumour DNA analyses investigating changes in molecular evidence of disease during treatment. Among 43 evaluable patients who were circulating tumour DNA-positive at baseline, 28 had cleared detectable circulating tumour DNA before surgery.
This represents a reported clearance rate of 65.1% in that evaluable subgroup. The findings support further research into molecular response monitoring, although they do not establish that circulating tumour DNA clearance can independently guide perioperative treatment decisions.
What does the safety evidence reveal about combining ivonescimab with chemotherapy before surgery?
Safety is a central consideration in any treatment administered before potentially curative surgery. The Nature Medicine publication reported Grade 3 or higher treatment-related adverse events in 24 of 79 patients, or 30.4%, receiving ivonescimab with chemotherapy.
In the smaller monotherapy cohort, four of 11 patients experienced Grade 3 or higher treatment-related adverse events. The investigators reported no Grade 3 or higher bleeding events and no treatment-related deaths across the study.
Surgery-related adverse events considered associated with ivonescimab occurred in seven of 67 surgical patients in the combination cohort and one of 10 patients in the monotherapy surgical cohort. No Grade 3 or higher events were reported within those surgery-related categories.
These findings provide reassurance concerning the absence of particular severe complications in the observed population. However, a study involving 90 patients cannot reliably exclude uncommon adverse events or establish how risks may differ across a broader patient population.
Treatment-related surgical delay occurred in one patient, representing 1.3% of the combination cohort. The study also recorded two deaths unrelated to the investigational treatment among patients who did not complete surgery.
Longer-term tolerability will matter because the regimen includes postoperative ivonescimab treatment. Any future programme must assess treatment discontinuations, immune-related complications, vascular events and the potential cumulative consequences of extended exposure.
The safety findings therefore support further development but should not be treated as proof that perioperative ivonescimab is safer than established regimens.
How does the Nature Medicine publication strengthen Akeso and Summit Therapeutics’ development strategy?
Ivonescimab is already approved in China for specified advanced non-small cell lung cancer indications. In August 2026, the Chinese National Medical Products Administration granted a third marketing authorisation for an ivonescimab-containing regimen, covering first-line treatment of advanced squamous non-small cell lung cancer in combination with chemotherapy.
Its development in resectable disease remains investigational. Approval for advanced cancer cannot be extended automatically to treatment before and after surgery, where patient eligibility, treatment objectives and clinical endpoints differ.
Summit Therapeutics holds exclusive development and commercialisation rights to ivonescimab across several major international markets, including the United States and Europe. The agreement has a total potential value of up to US$5 billion, although that headline amount includes contingent economics rather than consideration already received.
The newly published perioperative study adds scientific depth to a molecule being investigated across multiple lung cancer and other solid tumour settings. However, the study does not establish a new approval or independently quantify additional licensing revenue.
For investors, its significance lies primarily in strengthening the broader development rationale. The publication increases the amount of peer-reviewed evidence available for evaluating ivonescimab’s biological activity, but the transition from Phase 2 findings to routine perioperative use remains dependent on confirmatory research.
The next important clinical objective is a sufficiently powered randomised trial establishing whether the regimen improves event-free survival, overall survival or other relevant outcomes against an appropriate comparator.
Such a study would also need to characterise the relationship between pathological response and longer-term benefit while maintaining acceptable surgical and postoperative safety.
The Nature Medicine publication represents meaningful progress for Akeso’s ivonescimab programme, particularly because of the substantial pathological responses observed in the combination cohort. Its 70.1% major pathological response and 50.7% complete response rates provide a strong justification for further investigation, but do not settle the question of comparative clinical benefit. Establishing whether PD-1 and VEGF dual targeting can improve long-term outcomes beyond existing perioperative treatment standards remains the decisive test.
