Insilico Medicine, listed on the Hong Kong Stock Exchange under stock code 3696, will present first-in-human Phase 1 results for ISM6331 at the European Society for Medical Oncology Congress 2026. The investigational oral pan-TEAD inhibitor is being evaluated in patients with advanced or metastatic malignant mesothelioma and other solid tumors, with the data scheduled for a rapid oral presentation in Madrid on October 25, 2026.
The presentation will be closely watched because it represents the first opportunity to judge ISM6331 using human clinical evidence rather than medicinal chemistry, animal models or claims about the speed of artificial intelligence-assisted discovery. Insilico Medicine has disclosed that the presentation will cover safety, tolerability, pharmacokinetics and preliminary antitumor activity, but it has not yet released patient numbers, response rates, dose-level results or detailed adverse-event findings.
Acceptance for a rapid oral session indicates that the ESMO scientific programme considers the dataset worthy of prominent discussion. It does not, however, establish that ISM6331 has demonstrated meaningful efficacy, an acceptable therapeutic window or an advantage over competing TEAD inhibitors. Those questions will remain unanswered until the underlying data are disclosed and examined in the context of dose exposure, patient selection and duration of follow-up.
Why does the ISM6331 Phase 1 design make mesothelioma the central clinical test?
The ongoing study, registered as NCT06566079, is an open-label, multicentre Phase 1 trial with estimated enrolment of 100 participants. It began in December 2024 and includes a dose-escalation stage followed by a dose-selection optimisation stage designed to identify a recommended Phase 2 dose while examining pharmacokinetics, pharmacodynamics and preliminary efficacy. Recruitment is taking place across clinical centres in the United States and China.
Patients must have unresectable advanced or metastatic malignant mesothelioma or another solid tumor for which standard treatment has failed or no effective standard therapy remains. Mesothelioma participants are required to have received both immune checkpoint therapy and platinum-based chemotherapy, placing ISM6331 in a heavily pretreated population with substantial unmet need. For non-mesothelioma patients entering the second part of the trial, documented dysregulation of the Hippo signalling pathway is required.
This distinction is important. The dose-escalation component can enrol patients regardless of whether a Hippo pathway alteration has been identified, allowing investigators to assess safety across a broader population. The dose-optimisation stage adopts a more selective strategy that may help determine whether molecular evidence of pathway dysregulation enriches for patients more likely to respond.
The study uses standard Response Evaluation Criteria in Solid Tumors for most cancers and modified criteria for malignant pleural mesothelioma. That matters because pleural tumors frequently grow along the chest wall rather than forming easily measured spherical masses, making conventional response measurements less suitable.
At ESMO 2026, clinicians will want to know how many patients were evaluable, how many had mesothelioma, which dose levels were tested and whether responses appeared concentrated in a specific molecular subgroup. A small number of responses scattered across multiple tumor types would carry a different development implication from a reproducible signal in mesothelioma patients with identifiable Hippo pathway abnormalities.
How does ISM6331 attempt to interfere with the Hippo, YAP, TAZ and TEAD cancer pathway?
ISM6331 is an orally administered small molecule designed to inhibit the TEAD family of transcription factors. TEAD proteins work with the transcriptional coactivators YAP and TAZ, which sit downstream of the Hippo signalling pathway and influence genes involved in cell proliferation, survival, tissue growth and resistance mechanisms.
When normal Hippo pathway controls are disrupted, YAP and TAZ can accumulate in the cell nucleus and interact with TEAD proteins, supporting abnormal transcriptional programmes that promote tumor growth. Alterations involving pathway regulators such as NF2 and LATS1 or LATS2 have made mesothelioma one of the most biologically plausible diseases for testing TEAD-directed drugs. Peer-reviewed research has consequently identified the YAP, TAZ and TEAD axis as a potentially actionable vulnerability in Hippo-driven cancers.
Insilico Medicine said ISM6331 was generated and designed using Chemistry42, the company’s generative artificial intelligence platform for molecular design. The candidate is described as a pan-TEAD inhibitor, meaning it is intended to inhibit activity across the TEAD protein family rather than selectively targeting one isoform.
The pan-TEAD approach could offer broad pathway suppression, but greater target coverage does not automatically produce a better therapeutic index. The clinical challenge is to inhibit tumor-driving transcription strongly enough to deliver durable antitumor activity without causing toxicities that prevent sustained dosing.
Preclinical potency, oral bioavailability and selectivity are useful starting characteristics, but they cannot establish human efficacy. The ESMO presentation therefore represents the point where the Chemistry42 narrative begins to face the same clinical standards applied to any conventionally discovered oncology compound.
Why has clinical progress with VT3989 raised the competitive benchmark for ISM6331?
ISM6331 is not entering an untouched therapeutic field. Vivace Therapeutics has already reported human proof-of-concept data for VT3989, an oral TEAD palmitoylation inhibitor evaluated in refractory solid tumors, particularly mesothelioma.
A peer-reviewed Phase 1 and Phase 2 analysis published in Nature Medicine included 172 treated patients, of whom 135 had mesothelioma. Among 47 mesothelioma patients treated at clinically optimised doses, the reported objective response rate was 26%. In a smaller 22-patient analysis incorporating selected doses and urine albumin-to-creatinine thresholds, the objective response rate was 32%, the disease-control rate was 86% and median progression-free survival was 10 months.
These findings provided early clinical validation for inhibiting the Hippo, YAP and TEAD pathway. They also raised the bar for later entrants. ISM6331 will no longer be judged solely on whether it produces any observable tumor shrinkage. Investigators and investors will compare its response depth, durability, dose intensity, patient-selection strategy and tolerability with the emerging profile of VT3989.
Any such comparison must remain cautious because the trials are not head-to-head studies. Differences in eligibility criteria, prior therapies, molecular selection, dose optimisation, follow-up and response assessment can materially influence reported outcomes.
The competitive question is therefore not whether one isolated response percentage is numerically higher. It is whether ISM6331 can generate a sufficiently consistent benefit across a clearly defined population while preserving a safety profile that allows long-term oral administration.
A differentiated pharmacological profile could still create room for ISM6331. Differences in binding mode, potency, selectivity, drug exposure, dosing flexibility or combination compatibility may become meaningful. Insilico Medicine will need human pharmacokinetic and pharmacodynamic evidence to demonstrate that those theoretical distinctions translate into a clinically useful candidate.
Which safety findings could determine whether ISM6331 advances beyond Phase 1?
Safety will be at least as important as preliminary efficacy in the ESMO dataset. The primary purpose of a first-in-human dose-escalation trial is to identify dose-limiting toxicities, characterise exposure and select a dose suitable for subsequent development.
The VT3989 experience provides clues about the findings that oncologists may scrutinise. Reported toxicities were mostly grade 1 or grade 2 and included increased urine albumin-to-creatinine ratios, proteinuria, peripheral oedema and fatigue. Proteinuria was described as reversible following dose modification and was not associated with renal impairment in the published analysis, but the finding became relevant to dose and patient-management decisions.
For ISM6331, the crucial disclosures will include the frequency of grade 3 or higher treatment-related adverse events, dose interruptions, dose reductions, discontinuations and serious adverse events. Investigators will also examine whether any toxicities increase predictably with exposure and whether they can be managed without compromising sustained pathway inhibition.
The recommended Phase 2 dose cannot be selected simply by choosing the highest dose below a conventional toxicity threshold. Modern oncology development increasingly requires a broader assessment of biological activity, pharmacokinetics, tolerability over repeated cycles and the exposure needed to suppress the intended target.
The October presentation should therefore show whether Insilico Medicine has identified a dose range that balances target engagement with practical chronic administration. A response observed at an exposure that cannot be maintained would have limited development value. Conversely, modest early activity accompanied by consistent target engagement and manageable tolerability could justify expansion in a biomarker-enriched mesothelioma cohort.
Where could ISM6331 fit as first-line mesothelioma treatment continues to evolve?
The treatment environment for unresectable malignant pleural mesothelioma has expanded, although outcomes remain poor for many patients. The United States Food and Drug Administration approved nivolumab with ipilimumab as a first-line option in 2020. It subsequently approved pembrolizumab with pemetrexed and platinum chemotherapy in September 2024 for unresectable advanced or metastatic malignant pleural mesothelioma.
In the KEYNOTE-483 study supporting the pembrolizumab combination, median overall survival was 17.3 months compared with 16.1 months for chemotherapy alone. The confirmed objective response rate was 52% with pembrolizumab plus chemotherapy and 29% with chemotherapy, although median progression-free survival was 7.1 months in both groups.
ISM6331 is not currently positioned to compete directly with these first-line regimens. The Phase 1 trial enrols mesothelioma patients who have already received checkpoint inhibition and platinum chemotherapy, making the immediate opportunity a later-line setting.
A credible signal in this population could support a mesothelioma-focused expansion study. Development could subsequently move toward molecularly selected patients, rational combinations or earlier treatment lines, but each step would require additional evidence.
Combination development may be particularly important because Hippo pathway activation has been associated with tumor survival and therapeutic resistance. A tolerable oral TEAD inhibitor could eventually be studied alongside immunotherapy, chemotherapy or other targeted treatments. Such strategies remain hypothetical for ISM6331 until its single-agent dose, safety profile and antitumor activity are better understood.
Does the ESMO presentation validate Insilico Medicine’s artificial intelligence platform?
The presentation is strategically important for Insilico Medicine because the company’s valuation and industry profile depend partly on the proposition that generative artificial intelligence can produce development candidates faster and more efficiently. ISM6331 offers another opportunity to test whether platform-generated molecules can survive the transition from computational design into human trials.
However, acceptance at a major oncology congress does not validate artificial intelligence as a superior discovery method. Clinical development ultimately measures the properties of the molecule, not the sophistication of the software that proposed it.
The strongest platform validation would come from repeated clinical success across independently designed programmes. Insilico Medicine has broadened its pipeline and signed substantial collaborations with pharmaceutical companies, while its lead fibrosis candidate rentosertib has moved into late-stage development. Yet none of the company’s investigational medicines has been approved for commercial sale, leaving clinical execution as the decisive test of its model.
For ISM6331, the October dataset should reveal whether Chemistry42 produced a molecule with useful oral exposure, target engagement, tolerability and preliminary activity. Even encouraging Phase 1 results would support further evaluation rather than prove that artificial intelligence can reliably improve the probability of oncology drug approval.
How are Insilico Medicine shares reflecting clinical, licensing and financial momentum?
Insilico Medicine completed its Hong Kong listing on December 30, 2025, at an offer price of HK$24.05 per share. The company raised approximately HK$2.28 billion through the initial public offering, giving it additional capital for clinical development, artificial intelligence models, automated laboratories and earlier-stage research.
The shares closed at HK$44.20 on July 28, 2026, down 7.45% for the session. The stock was approximately unchanged over the five trading days beginning July 22, but remained around 6.8% above its June 29 close and nearly 84% above the initial public offering price. Since listing, the shares have traded between HK$29.98 and HK$80.90, illustrating both strong investor interest and substantial volatility.
Sentiment around Insilico Medicine has recently been supported by licensing activity, pipeline milestones and a positive profit warning. The company expects first-half 2026 revenue of between US$102.5 million and US$106.5 million, compared with US$27.5 million a year earlier. It also expects net profit of between US$33.5 million and US$39.5 million, against a net loss of US$19.2 million in the corresponding period of 2025, primarily reflecting increased revenue from licensing, co-development and research collaborations.
The ESMO announcement should therefore be viewed as a long-term pipeline signal rather than the sole explanation for current share-price movements. Insilico Medicine’s market value reflects multiple programmes and commercial partnerships, while ISM6331 remains an early-stage candidate with significant development risk.
What must the ESMO 2026 dataset show for ISM6331 to become a serious oncology asset?
The October 25 presentation needs to move ISM6331 beyond a compelling mechanism and into a clinically interpretable development position. The most important evidence will be the number of evaluable patients, tumor types, prior treatment exposure, doses administered, dose-limiting toxicities and pharmacokinetic behaviour.
Attention will then turn to the number and duration of objective responses, the proportion of patients achieving stable disease and whether activity correlates with NF2 loss, other Hippo pathway alterations or broader evidence of YAP and TEAD activation. Detailed biomarker findings could determine whether future studies adopt a narrow genetically selected approach or retain a wider mesothelioma strategy.
The data will also need to clarify whether the study has identified a practical recommended Phase 2 dose. A dose suitable for expansion should provide sustained exposure and biological activity without requiring frequent interruptions or creating cumulative toxicity that becomes difficult to manage.
An encouraging presentation could justify a dedicated mesothelioma expansion cohort and sharpen Insilico Medicine’s oncology strategy. A weak or poorly differentiated dataset would make it harder to compete against VT3989 and other emerging Hippo pathway programmes.
ISM6331 has already reached a clinically important stage that many computer-designed molecules never achieve. The harder test arrives at ESMO 2026, when human data must show whether Insilico Medicine’s pan-TEAD candidate possesses the efficacy signal, tolerability and dose discipline needed to progress from an artificial intelligence discovery story into a credible oncology programme.
