5 First in Class Therapies to Look Out For at ESMO 2026

Introduction: The next breakthrough may come from changing the therapeutic approach

The oncology pipeline is entering a phase in which innovation is increasingly defined by how a therapy engages tumor biology rather than simply which target it addresses. Antibody–drug conjugates (ADCs), immune checkpoint inhibitors and molecularly targeted therapies have expanded treatment options across multiple malignancies, but significant limitations persist. Resistance can undermine initially effective treatments, molecular eligibility can exclude patients whose tumors express a relevant target, and some cancer-associated signalling pathways remain difficult to inhibit pharmacologically. Even when immune checkpoints are blocked, the immune system may not generate a sufficiently effective or durable antitumor response.

These challenges are prompting a new generation of investigational therapies to explore different ways of delivering cytotoxic payloads, inhibiting transcriptional signalling and redirecting or activating immune cells. The European Society for Medical Oncology (ESMO) Congress 2026, taking place in Madrid from October 23–27, offers an opportunity to assess five programmes that illustrate this shift: JSKN033, ISM6331, BHV-1530, XmAb819 and invikafusp alfa. Their development strategies range from combining established therapeutic principles in a novel formulation to pursuing difficult biological targets and selectively activating T-cell populations.

The significance of these candidates lies not simply in their first-in-class positioning, but in the treatment limitations they seek to address. JSKN033 aims to integrate targeted tumor killing with checkpoint inhibition; ISM6331 tests whether AI-designed chemistry can translate inhibition of the Hippo–TEAD pathway into a clinical treatment; BHV-1530 explores whether an ADC can extend FGFR3 targeting beyond genomically defined populations; XmAb819 redirects T cells towards clear cell renal cell carcinoma (ccRCC); and invikafusp alfa seeks to strengthen selected T-cell populations in cancers resistant to PD-1-directed therapy.

These are distinct therapeutic hypotheses, and the evidence supporting them is not equivalent. Some programmes have generated preliminary clinical signals, while others are approaching important early clinical readouts. The central question for ESMO 2026 is therefore not which candidate sounds most novel, but which programme is beginning to establish a credible link between its mechanism, patient selection and clinically meaningful benefit.

JSKN033: Combining tumor-directed cytotoxicity with immune checkpoint blockade

The treatment barrier: Advanced cervical cancer remains difficult to manage after progression on standard systemic therapy. Although ADCs and checkpoint inhibitors have expanded treatment options, their mechanisms address different aspects of tumor biology and do not necessarily overcome resistance when used individually.

JSKN033 combines JSKN003, a biparatopic HER2-directed ADC, with envafolimab, a PD-L1 inhibitor, in a subcutaneous fixed-dose formulation. JSKN003 is designed to recognise two HER2 epitopes and deliver a topoisomerase I inhibitor payload to tumor cells, while envafolimab targets PD-L1-mediated immune suppression. The approach brings targeted cytotoxicity and checkpoint inhibition together within a single treatment formulation.

Its potential differentiation is twofold. First, the combination tests whether tumor-directed payload delivery and immune checkpoint blockade can produce complementary antitumor activity. Second, subcutaneous administration could simplify treatment delivery compared with regimens requiring separate intravenous infusions. This may reduce administration time and healthcare-resource requirements, although convenience alone cannot establish superior efficacy, improved adherence or better quality of life.

The programme has progressed beyond a purely exploratory concept. The ESMO 2026 presentation is expected to report efficacy and safety findings from the Chinese Phase I/II study JSKN033-102 in patients with cervical cancer whose disease has progressed after standard-of-care therapy. The company has also reported development of additional cervical cancer studies, including a Phase III monotherapy programme and a Phase II combination study in the first-line setting.

Analytical outlook: JSKN033 represents a broader industry trend: integrating therapeutic modalities while improving how treatment is administered. Its clinical proposition will be strongest if the combined mechanism delivers meaningful benefit and the subcutaneous formulation reduces treatment burden without compromising safety or efficacy. The important distinction is between a more convenient formulation and a genuinely differentiated treatment.

ISM6331: Bringing AI-designed pan-TEAD inhibition into clinical testing

The treatment barrier: Cancer cells can exploit signalling networks that regulate growth, survival and adaptation to treatment. The Hippo pathway and its downstream YAP/TAZ–TEAD transcriptional machinery are implicated in tumor biology, but transcriptional regulators have historically presented challenges for conventional small-molecule drug discovery. ISM6331 is an investigational, AI-designed small molecule intended to inhibit all four TEAD family transcription factors. Its first-in-human, multicentre Phase I study is evaluating the compound in patients with mesothelioma and other advanced solid tumors. Initial clinical findings are scheduled for presentation at ESMO 2026. The programme has also received FDA Fast Track and Orphan Drug designations, according to the developer. 

The scientific rationale extends beyond the use of artificial intelligence. AI-enabled molecular design may help identify and optimise compounds against challenging targets, but the resulting drug must still demonstrate adequate potency, exposure, selectivity and tolerability in humans. The first clinical readout will therefore be important in determining whether ISM6331 can translate its design rationale into a viable therapeutic candidate.

Pan-TEAD inhibition also raises a biological question. Inhibiting all four TEAD family members could provide broader pathway suppression than selectively targeting one member, but broader inhibition does not automatically translate into superior clinical outcomes. Tumors may vary in their dependence on TEAD signalling, and the therapeutic window will depend on whether sufficient pathway inhibition can be achieved without unacceptable effects on normal tissues.

Analytical outlook: ISM6331 is a test of both computational drug discovery and target biology. Its success would not mean that AI independently guarantees better medicines; it would demonstrate that an AI-designed molecule can achieve clinically relevant inhibition of a difficult pathway. The key milestone is the transition from a compelling design hypothesis to evidence of target engagement, tolerability and activity in patients.

BHV-1530: Expanding the reach of FGFR3-directed treatment

The treatment barrier: Precision oncology has established FGFR3 as a clinically relevant target, particularly in urothelial carcinoma. However, approved FGFR-directed kinase inhibitors are used in molecularly selected settings, leaving some patients with FGFR3-expressing tumors outside the eligible population. This creates an opportunity for a different therapeutic modality to exploit the target without relying exclusively on the presence of an eligible genomic alteration.

BHV-1530 is an investigational FGFR3-directed ADC incorporating a proprietary topoisomerase I inhibitor payload. Rather than primarily inhibiting receptor signalling, the ADC is designed to use FGFR3 recognition to deliver a cytotoxic agent to tumor cells. Its Phase I study includes patients with advanced urothelial cancer, non-small cell lung cancer and head and neck squamous cell carcinoma, including patients with FGFR3 overexpression without a detectable activating alteration.

The programme has already generated preliminary clinical signals. In July 2026, Biohaven reported early tumor reductions and confirmed partial responses in heavily pretreated patients, including patients with FGFR3-altered and wild-type FGFR3-overexpressing tumors. The company also reported no dose-limiting toxicities at the time of its update and no observed toxicities characteristic of FGFR kinase inhibitors, such as hyperphosphataemia and certain ocular or nail disorders. These remain sponsor-reported early findings and do not establish a superior safety profile or comparative efficacy.

The key opportunity is to extend the population potentially treatable through FGFR3 recognition. However, protein expression is not synonymous with treatment sensitivity. Successful ADC activity requires sufficient target availability, effective binding and internalisation, payload delivery and susceptibility of the tumor cells to the payload. The relationship between FGFR3 expression and clinical response will therefore be central to the programme’s development.

Analytical perspective: BHV-1530 is testing whether ADC technology can expand the therapeutic relevance of an established molecular target beyond alteration-defined populations. If expression-based selection proves predictive, the programme could help address an important limitation of precision oncology. However, a larger potential patient population is meaningful only if it translates into durable clinical benefit and an acceptable safety profile.

XmAb819: Redirecting T cells towards clear cell renal cell carcinoma

The treatment barrier: Advanced ccRCC has benefited from immune checkpoint inhibitors and VEGF-targeted therapies, but resistance to established treatments remains a major clinical challenge. A continuing development objective is to direct immune-cell activity towards tumor cells through mechanisms that do not depend solely on restoring pre-existing antitumor immunity.

XmAb819 is an investigational bispecific antibody designed to bind ENPP3 on tumor cells and CD3 on T cells. By engaging both targets, the molecule is intended to bring T cells into proximity with ENPP3-expressing cancer cells and trigger T-cell-mediated killing. Its development is based on ENPP3 expression in ccRCC and the possibility of using this target to direct immune activity towards malignant tissue. 

XmAb819 has already produced preliminary clinical findings. In an October 2025 company update, 20 efficacy-evaluable patients treated within the predicted target-dose range had an overall response rate of 25%, including four confirmed partial responses and one unconfirmed partial response. The reported disease control rate was 70%. These findings support further investigation, but the small, selected population and early-phase design limit the conclusions that can be drawn. They cannot establish superiority over approved therapies or reliably predict outcomes in a larger trial. The mechanism also creates a significant safety challenge. T-cell engagers can trigger cytokine release syndrome and other inflammatory toxicities, making dose optimisation essential. The ability to engage tumor cells and activate T cells must be balanced against the risk of excessive immune activation. The ESMO 2026 presentation is expected to provide updated Phase I expansion-cohort findings relevant to dose selection and the programme’s planned pivotal development. Xencor has indicated that a Phase III monotherapy study is planned to begin in 2027.

Analytical perspective: XmAb819 illustrates the promise and complexity of T-cell engagers. The early response signal is a reason to continue development, not proof of clinical superiority. Its value will depend on whether ENPP3-directed immune engagement can produce durable disease control while keeping immune-related toxicity manageable in a population with substantial unmet need.

Invikafusp alfa: Selective T-cell activation after checkpoint resistance

The treatment barrier: Immune checkpoint inhibitors have transformed oncology, yet many patients never respond and others develop resistance after an initial benefit. In some tumors, releasing inhibitory signals may be insufficient because the antitumor T-cell response itself needs to be strengthened or redirected.

Invikafusp alfa, also known as STAR0602, is an investigational bispecific dual T-cell agonist designed to selectively activate T cells expressing particular T-cell receptor β-chain variants, including Vβ6 and Vβ10. The molecule combines selective T-cell receptor engagement with an interleukin-2 signal, aiming to expand relevant T-cell populations while avoiding indiscriminate activation of the broader T-cell compartment. It is being evaluated in the START-001 Phase I/II study in patients with advanced solid tumors resistant or refractory to PD-1-directed treatment.

Earlier clinical findings provide a rationale for further investigation. Phase I data reported before ESMO 2026 identified a recommended Phase II dose of 0.08 mg/kg and showed selective expansion of predominantly CD8-positive Vβ6/Vβ10 T cells. Preliminary antitumor activity included confirmed partial responses in patients with TMB-H, microsatellite-stable colorectal cancer, with one reported response lasting approximately 12 months. The programme received FDA Fast Track designation for TMB-H colorectal cancer. These findings are encouraging but do not establish broad efficacy across checkpoint-resistant tumors. The ESMO 2026 update is particularly important because the programme has moved into Phase II dose expansion. The planned oral presentation will report updated findings from biomarker-enriched cohorts, including patients with TMB-H solid tumors resistant to immune checkpoint blockade. The study is intended to evaluate whether selective activation of Vβ6/Vβ10 T-cell subsets can produce single-agent activity after PD-1-directed treatment has failed. The central challenge is heterogeneity. TMB-H, mismatch-repair-deficient and virally associated tumors can differ substantially in antigen presentation, immune-cell composition and resistance mechanisms. A broad immunotherapy strategy will therefore need to demonstrate which tumor types and biomarker-defined populations are most likely to respond.

Analytical outlook: Invikafusp alfa tests whether selective immune activation can address a limitation of checkpoint blockade without broadly activating the entire T-cell compartment. If its clinical activity is reproducible, the programme could provide a differentiated immunotherapy approach for selected patients after PD-1 resistance. Its value will depend on durable responses and a manageable safety profile, not immune expansion alone.

What these programmes reveal about the next wave of oncology innovation

Three broader themes emerge from these programmes.

First, innovation is becoming more specific about the biological problem it seeks to solve. The next generation of oncology therapies is not simply adding new agents to existing classes. It is attempting to exploit tumor-specific expression, inhibit difficult signalling machinery or reshape immune activity. However, a compelling mechanism is only the starting point; the clinical development programme must establish where that mechanism works and where it does not.

Second, patient selection may determine the eventual reach of a therapy. BHV-1530 raises the possibility of extending treatment beyond mutation-defined FGFR3 populations, while invikafusp alfa uses biomarker-enriched development to investigate checkpoint-resistant disease. Both programmes highlight a central precision-oncology challenge: broadening eligibility without weakening the ability to predict benefit. A wider addressable population is meaningful only when a reliable treatment-response relationship is established.

Third, early activity must be separated from clinical validation. Phase I studies primarily establish safety, tolerability, pharmacokinetics and dose selection, while exploring preliminary activity. Small cohorts and uncontrolled designs cannot establish comparative superiority. For these candidates, response durability, consistency across patients, validated biomarkers and subsequent clinical trials will determine whether the initial signal translates into a treatment option.

This is why ESMO 2026 should be read as more than a showcase of new mechanisms. It is an opportunity to assess whether these mechanisms are generating the evidence needed for the next stages of development. The most consequential findings will not necessarily be those with the largest early response rates, but those that clarify the relationship between target biology, patient selection, treatment exposure and durable benefit.

Conclusion: First-in-class is a starting point, not the endpoint

JSKN033, ISM6331, BHV-1530, XmAb819 and invikafusp alfa represent five distinct attempts to address persistent barriers in oncology. They span combination therapy, difficult-to-drug pathway inhibition, ADC-based target expansion, tumor-directed T-cell engagement and selective immune activation.

Their significance will ultimately depend on whether they deliver more than mechanistic novelty. A clinically meaningful advance requires a demonstrable relationship between the proposed mechanism and patient outcomes, an acceptable safety profile and a clearly defined place in the treatment pathway. Some candidates may prove most useful in narrowly selected populations rather than across the broader groups initially considered. That could still represent important progress if the benefit is durable and clinically relevant.

The next wave of oncology innovation will be judged not by how many first-in-class therapies enter development, but by how convincingly they transform biological hypotheses into effective treatment strategies. ESMO 2026 offers an opportunity to evaluate that transition from the promise of a novel mechanism to the evidence required to change clinical practice.