Lutetium extends survival in neuroendocrine tumours; health systems get AI governance tools
A Phase 3 trial, new genomic tools for blood cancers, drug-safety AI, and health systems' new AI governance playbook.
By Dr. Asher Knippel
The 2026 American Society of Clinical Oncology Annual Meeting (ASCO) and the Society of Nuclear Medicine and Molecular Imaging (SNMMI) Annual Meeting both concluded on 2 June, bookending a week that advanced radiopharmaceutical therapy, whole-genome sequencing for blood cancers, and computational drug safety — while health systems received new practical tools for deploying AI responsibly.
Monday, 2 June: Lutetium-177 edotreotide outperforms everolimus in Phase 3 neuroendocrine tumour trial
ITM Isotope Technologies Munich SE presented final Phase 3 data from the COMPETE trial at the SNMMI 2026 Annual Meeting, demonstrating that ¹⁷⁷Lu-edotreotide (ITM-11) significantly extends progression-free survival (PFS) in patients with gastroenteropancreatic neuroendocrine tumours (GEP-NETs). Median PFS reached 23.9 months with the radiopharmaceutical versus 14.1 months with the comparator everolimus — a difference that achieved statistical significance (p = 0.022). The trial also validated a single-timepoint dosimetry approach, with mean absolute percentage errors of 3–20% for kidney dosimetry, replacing multi-day imaging schedules with a single scan without sacrificing precision. GEP-NETs are rare tumours arising in the gut, pancreas, and appendix; this Phase 3 result strengthens peptide receptor radionuclide therapy (PRRT) as a preferred option and is expected to support regulatory submissions to the EMA and FDA.
Monday, 2 June: Foundation Medicine and SWOG launch biomarker-driven trials collaboration
Foundation Medicine and SWOG Clinical Trials Partnerships announced a strategic research collaboration on 2 June, pairing Foundation Medicine's portfolio of more than 100 companion diagnostic (CDx) approvals with SWOG's academic clinical-trials network. The partnership aims to embed genomic testing directly into trial protocols, identifying patients most likely to respond to a given therapy at enrolment rather than in retrospect. This translational infrastructure — connecting real-world genomic data to prospective trial design — is increasingly required by regulators and payers before approving new precision-medicine therapies.
ASCO 2026: Whole-genome sequencing enters clinical practice for blood cancers
Tempus AI announced the upcoming clinical availability of xH, a whole-genome sequencing (WGS) assay designed for patients with acute myeloid leukaemia (AML) and myelodysplastic syndromes (MDS), with data presented at the ASCO 2026 Annual Meeting. The assay achieved over 99% sensitivity in an MDS cohort and uncovered 40% more clinically relevant genomic findings compared with standard targeted panels — findings that can change treatment decisions. Conventional next-generation sequencing panels interrogate known disease-associated genes; whole-genome sequencing analyses all three billion base pairs, capturing structural variants and novel mutations that targeted panels miss. For blood cancer patients where treatment selection hinges on subtle genomic differences, this additional information can materially alter clinical management.
Cancer Discovery: AI model decodes tumour mutation landscape to predict treatment response
Researchers at the University of California San Diego published MutationProjector in Cancer Discovery, describing an AI model trained on genomic data from more than 30,000 tumours across 10 solid cancer types. Rather than flagging individual mutations, MutationProjector analyses the full combination of genetic alterations in a tumour, generates a compact biological-state representation, and then identifies disrupted molecular pathways potentially targetable by existing drugs. In validation experiments the model predicted treatment response based on pathway-level biology. This is translational work — it has not yet been tested in a prospective randomised trial — but publication in Cancer Discovery, a peer-reviewed journal of the American Association for Cancer Research (AACR), marks a rigorous early step toward clinical application.
Scientific Reports: Hybrid AI framework improves prediction of adverse drug reactions
A research team published PromptSE in Scientific Reports, presenting a hybrid computational framework that combines large-language-model semantic reasoning with deep-learning prediction to anticipate adverse drug reactions (ADRs) before a compound enters human trials. Unlike models relying primarily on statistical associations between molecular structure and known side-effect profiles, PromptSE guides the model to reason through the biological chain of events that could produce a given symptom, achieving stronger benchmark performance. Adverse drug reactions are among the leading preventable causes of death worldwide; more reliable computational screening at the pre-clinical stage could reduce the number of unsafe compounds advancing into costly early-phase trials.
Coalition for Health AI releases practical governance playbooks for health systems
The Coalition for Health AI (CHAI) released a series of governance resources developed in consultation with more than 150 clinicians and health AI leaders. The playbooks provide practical guidance for health systems on establishing internal AI policies, managing third-party AI developers, assessing risks from AI-generated clinical recommendations, and training clinical staff. CHAI also published a standardised "model card" template documenting each AI tool's intended use, performance characteristics, and known limitations — giving procurement teams a common language for evaluating vendors. The framework is designed to scale for smaller hospitals and community health centres that lack dedicated AI governance staff but are increasingly offered AI-assisted clinical tools by commercial vendors. CHAI was founded in 2021 by a consortium including academic medical centres, health systems, and federal health agencies in the United States.
The information in this roundup is drawn from published research, conference presentations, and institutional announcements. It is intended as journalistic reporting and does not constitute medical advice. Readers should consult a qualified clinician before making any change to their treatment or care.