ESMO 2025 in review: A new era in oncology, from AI to next-gen therapeutics

The European Society for Medical Oncology (ESMO) Annual Congress has long been a bellwether for global oncology trends, and the 2025 meeting in Berlin was no exception. This year’s event showcased a striking convergence of technology and biology, where artificial intelligence (AI) meets immunotherapy, and precision medicine intersects with novel drug modalities.

From antibody–drug conjugates (ADCs) moving into early-stage disease to arenavirus-based immunotherapies and ctDNA-guided treatment strategies, ESMO 2025 reflected a field in the midst of rapid evolution – one that is increasingly defined by personalisation, data-driven decision-making, and the pursuit of durable outcomes.

The overarching themes were clear: integration of AI into clinical workflows, expansion of targeted therapies into earlier lines, and innovation in immuno-oncology beyond checkpoint inhibition.

AI and digital oncology: From promise to practice

Artificial intelligence is no longer a futuristic concept in oncology. Rather, it is becoming a practical tool. At ESMO 2025, sessions chaired by global experts highlighted how AI-based imaging biomarkers are transforming radiology and pathology. Tools such as Google’s Med-Gemini and Alpaca demonstrated impressive accuracy in whole-slide imaging, while spatial awareness models like SMMILe are enabling nuanced interpretation of tumour microenvironments.

Yet, implementation hurdles remain. As University of Bern Professor of Digital Pathology Inti Zlobec noted, only 5% of pathology labs in Switzerland are fully digitised – a figure mirrored across Europe. Without digitisation, though, AI adoption risks logistical disruption. The cost of upgrading laboratory information systems is estimated at €5 million per site, underscoring the economic challenge.

Still, the trajectory is clear. AI agents for multi-omic modelling and computational companion diagnostics are emerging, promising to integrate imaging, genomics, and clinical data into a unified decision-making framework. The CREATE study, presented by Qure.ai and AstraZeneca, exemplified this potential: an AI-driven chest X-ray tool identified 96% of lung cancer cases, including early-stage disease in non-smokers – a population often missed by traditional screening.

The implication of this? AI is poised to democratise precision oncology, particularly in resource-limited settings, by leveraging widely available imaging modalities for early detection.

ADCs: A continuing revolution

If 2024 was the year ADCs cemented their role in metastatic disease, 2025 marked their advance into early-stage settings. AstraZeneca’s DESTINY-Breast05 and DESTINY-Breast11 trials demonstrated trastuzumab deruxtecan (T-DXd) delivering unprecedented efficacy in high-risk HER2-positive early breast cancer. Pathologic complete response rates approached 67%, and invasive disease-free survival trends favoured T-DXd over trastuzumab emtansine (T-DM1).

Safety remains a watchpoint. Interstitial lung disease (ILD) occurred in ~9.6% of patients, though mostly grade 1–2 and reversible. Cardiac toxicity rates were reassuringly low. As panellists noted, the next challenge lies in sequencing: should T-DXd be used before or after surgery? While consensus leans toward post-operative use following standard neoadjuvant therapy, ongoing trials will refine these strategies.

Beyond breast cancer, ADCs are reshaping bladder cancer care, also. The KEYNOTE-905 study reported transformative results for perioperative enfortumab vedotin plus pembrolizumab in cisplatin-ineligible muscle-invasive disease – a population historically underserved.

Indeed, ADCs are evolving from salvage therapy to curative intent, signalling a paradigm shift that could redefine treatment algorithms across tumour types.

Immunotherapy reinvented: Beyond PD-1

Checkpoint inhibitors revolutionised oncology, but resistance remains a formidable barrier. At ESMO 2025, several presentations showcased strategies to overcome this challenge:

- Arenavirus-based immunotherapy (Abalos Therapeutics): ABX-001 activates both innate and adaptive immunity systemically, achieving complete remissions in preclinical models with a favourable safety profile. A first-in-human trial is imminent.
- GDF-15 blockade (CatalYm): Visugromab restored sensitivity to PD-1 inhibitors in refractory tumours, delivering responses lasting beyond 32 months in NSCLC and urothelial cancer. Notably, it also mitigated cancer cachexia – a dual benefit.
- Decoy-resistant IL-18 (Simcha Therapeutics): ST-067 enhanced the efficacy of bispecific T-cell engagers in solid and haematologic malignancies, addressing intrinsic T-cell limitations.
- Oncolytic virus strategies (Transgene/BioInvent): BT-001 combined with pembrolizumab induced tumour shrinkage in injected and distant lesions, reinforcing the promise of multifunctional viral platforms.

Immunotherapy’s next chapter will be defined by combination strategies and novel immune modulators that tackle resistance and broaden benefit.

Radioligand and alpha therapies: Precision at the cellular level

Radiopharmaceutical innovation featured prominently at ESMO. AdvanCell’s Lead-212-based PSMA-targeted alpha therapy (ADVC001) delivered encouraging safety and efficacy in metastatic prostate cancer – the first clinical validation of this approach. Meanwhile, Novartis’ Pluvicto extended its reach into hormone-sensitive disease, reducing progression risk by 28% when added to standard care.

These advances underscore the growing role of targeted radiotherapy in earlier lines, offering potent, localised cytotoxicity with manageable toxicity profiles.

Metabolic targeting: A new therapeutic pillar

Long considered “undruggable”, cancer metabolism emerged as a viable target at ESMO 2025. Faeth Therapeutics’ Phase 2 DICE trial in platinum-resistant ovarian cancer demonstrated a 34% reduction in progression risk and a 45% extension in progression-free survival when sapanisertib was added to paclitaxel. By simultaneously inhibiting PI3K, mTORC1, and mTORC2, Faeth’s multi-node strategy addresses metabolic plasticity – a key driver of resistance.

Post-ESMO 2025, the feeling is that metabolic targeting may join immunotherapy and ADCs as a foundational pillar of oncology, particularly in tumours linked to obesity and metabolic disorders.

Precision oncology and ctDNA-guided strategies

Minimal residual disease (MRD) detection via ctDNA dominated several high-impact sessions. The DYNAMIC-III trial in stage III colon cancer and IMvigor011 in urothelial cancer validated ctDNA as a prognostic marker and a tool for treatment de-escalation. Patients with negative ctDNA post-surgery achieved excellent outcomes without intensive chemotherapy, reducing toxicity and cost.

As Prof Angela Lamarca, ESMO Press Officer, emphasised, these findings herald a future where adjuvant therapy is tailored not by stage alone, but by molecular evidence of disease. And that is truly a leap toward personalisation.

Emerging targets and modalities

Innovation also extended beyond established pathways:

- KRAS G12D inhibition (Incyte): INCB161734 showed ORRs up to 34% in heavily pretreated PDAC – a breakthrough in a mutation long deemed intractable.
- Bispecific antibodies (INCA33890): Targeting TGFβR2 and PD-1, this agent achieved meaningful responses in MSS colorectal cancer, a population historically resistant to immunotherapy.
- Next-generation ADCs (Tubulis): TUB-040 delivered a 59% ORR in platinum-resistant ovarian cancer without biomarker selection, validating NaPi2b as a novel target and Tubutecan technology as a differentiated platform.

Additionally, big pharma and biotech alike leveraged ESMO to showcase pipeline depth:

- AstraZeneca: Expanded its ADC leadership with TROPION-Breast02 in TNBC and reinforced HER2 strategies.
- Ipsen: Acquired ImCheck Therapeutics for €350M, betting on BTN3A-targeted immunotherapy in AML.
- Arcus: Reported 26.7-month median OS with domvanalimab plus zimberelimab in gastric cancer, supporting TIGIT as a next-generation checkpoint.
- Merck, Exelixis, and others: Advanced ADCs, kinase inhibitors, and IO combinations across high-need indications.

Toward a data-driven, modality-diverse future

ESMO 2025 was more than a showcase of incremental progress; it was a manifesto for the future of oncology. Indeed, the meeting crystalised three imperatives:

1. Integration of technology and biology: AI and multi-omics will underpin decision-making, from early detection to adaptive treatment.
2. Diversification of therapeutic modalities: ADCs, radioligands, metabolic inhibitors, and immune modulators are expanding the armamentarium beyond traditional chemotherapy and checkpoint blockade.
3. Personalisation at scale: ctDNA-guided strategies and biomarker-driven trials promise to align intensity of care with individual risk, improving outcomes while mitigating toxicity and cost.

As these innovations move from congress halls to clinics, the challenge will be implementation and ensuring access, affordability, and equity in a rapidly evolving landscape.

For oncologists, regulators, and industry leaders, the message from Berlin is clear: the future of cancer care is not just targeted; it is integrated, intelligent, and transformative.