Welcome to this week's digest. This edition covers the latest official announcements, regulatory actions, and expert perspectives from the frontlines of artificial intelligence, gene therapy, and radioligand theranostics. All links lead directly to source material from regulatory agencies, peer-reviewed journals, clinical trial registries, and official press releases.

🤖 Artificial Intelligence

Agentic AI in Oncology Clinical Trials — ESMO TAT 2026 Expert Address

At the ESMO Targeted Anticancer Therapies Congress 2026 (Paris, 16–18 March), Jorge Reis-Filho (Chief AI and Data Scientist, AstraZeneca) presented a keynote on redesigning oncology trials with agentic AI. He outlined how semi-autonomous foundation models trained on multimodal data (EHR, pathology, radiology, genomics) are being used to predict treatment response, de-risk Phase II-to-III transitions, optimize trial eligibility, and generate synthetic control arms. He cited the TROPION-Lung01 AI-based predictive biomarker for datopotamab deruxtecan as a validated example.

Why it matters: These tools aim to reduce late-stage failure rates in oncology drug development, potentially shortening timelines and cutting costs. For clinical development teams, the message is that AI-enabled patient selection and adaptive trial design are moving from exploratory to operational expectations in early-phase oncology protocols.

Source: ESMO Targeted Anticancer Therapies Congress 2026 →

🧬 Gene Therapy

Ultragenyx DTX301 (AAV8) Hits Phase 3 Co-Primary Endpoint in OTC Deficiency

Ultragenyx reported that the Phase 3 Enh3ance trial of DTX301, an investigational one-time AAV8 gene therapy for ornithine transcarbamylase (OTC) deficiency, met its Week 36 co-primary endpoint: a statistically significant 18% reduction in 24-hour plasma ammonia AUC versus placebo, with an acceptable safety profile and no dose-limiting toxicities. Eight of nine patients with abnormal ammonia at baseline achieved normal levels post-dose.

Why it matters: This is a meaningful efficacy signal for a disease with no approved curative option. The result validates ammonia AUC0-24 as a regulatory-grade surrogate co-primary endpoint for rare metabolic disease gene therapy strategies, providing a template for future IND/BLA submissions in similar urea cycle disorders.

Source: ClinicalTrials Arena →

☢️ Radioligand Theranostics

First Head-to-Head Observational Comparison of 225Ac-PSMA vs 177Lu-PSMA in mCRPC

A peer-reviewed observational study published in Frontiers in Oncology directly compared outcomes in 116 patients treated with 177Lu-PSMA and 43 patients treated with 225Ac-PSMA at a single centre. PSA50 response rate was 42.2% for 177Lu-PSMA and 40.5% for 225Ac-PSMA; median overall survival was 13.0 months versus 11.8 months, with no statistically significant difference between the groups. Receiving more than two RLT courses was associated with longer survival across both agents.

Why it matters: This single-centre, non-randomized observational study is the most granular direct comparison to date. Despite the theoretical advantage of alpha-emitters, it finds equivalence in PSA response and OS at current clinical dosing schedules, raising important questions about patient selection, dosing optimization, and sequencing for future randomized trials.

Source: Frontiers in Oncology →

📡 Expert Signals

Jorge Reis-Filho on AI Integration in Oncology R&D

At ESMO TAT 2026, Reis-Filho argued that the leading oncology R&D organizations will be those fully integrating multimodal AI foundation models into drug development workflows. He highlighted how these tools enable accurate generalization from relatively small datasets, reducing the workforce and cost of Phase II trials, and described AI-based biomarkers as an active, present capability rather than a future aspiration.

Insight: The framing is that agentic AI is not a speculative R&D project but an operational tool already being deployed in leading pharma and academic consortia to de-risk trial design and accelerate translational validation.

Source: ESMO Daily Reporter →

📰 In Case You Missed It

Novartis Pluvicto Receives MHRA Approval for Pre-Taxane mCRPC

On 3 February 2026, the UK MHRA authorized Pluvicto for PSMA-positive mCRPC patients who have progressed on ARPI therapy and are considered appropriate to delay taxane-based chemotherapy, based on PSMAfore phase III data showing a 59% reduction in risk of radiographic progression or death versus ARPI switch.

Why still relevant: This expands the eligible European patient population and creates a regulatory precedent that will inform similar EMA/other jurisdictional applications, directly affecting site initiation and patient eligibility planning in UK/Ireland-based radioligand therapy programmes.

Source: Novartis UK →

Novartis Real-World Data: Pluvicto Shows 13.5 Months Median PFS in Chemo-Naïve PSMA+ mCRPC

Real-world evidence presented at GU ASCO 2026 reported median PFS of 13.5 months in chemo-naïve patients with PSMA-positive mCRPC receiving Pluvicto after one ARPI, consistent with PSMAfore trial data.

Why still relevant: This provides external validity and reinforces the importance of ARPI treatment sequencing as a critical stratification variable and endpoint modifier for clinical development teams managing Pluvicto trials.

Source: Novartis →

Radiopharm Theranostics Doses First Patient in Phase 1/2a Trial of 177Lu-BetaBart (RV-01)

RV-01 is the first 177Lu-conjugated monoclonal antibody targeting the 4Ig isoform of B7-H3 — an immune checkpoint overexpressed across multiple solid tumours — to enter first-in-human testing, using hepatic clearance to minimize renal toxicity.

Why still relevant: This signals expansion of the RLT modality beyond PSMA targets, directly relevant to future theranostics trial design in non-PSMA tumour types.

Source: PR Newswire →

Ultragenyx Completes BLA Rolling Submission for DTX401 (AAV8, GSDIa); FDA Grants Priority Review

DTX401 (pariglasgene brecaparvovec) for glycogen storage disease Ia (GSDIa) has a PDUFA date set for 23 August 2026, supported by 96-week Phase 3 data showing sustained reduction in cornstarch dependence.

Why still relevant: This serves as a model regulatory submission for complex AAV gene therapies in rare metabolic disease, with manufacturing at a dedicated US gene therapy facility.

Source: SEC Filing →

FDA Qualifies AIM-MASH AI Assist as First AI-Based Drug Development Tool for MASH Trials

The FDA's qualification of AIM-MASH (PathAI) through the Biomarker Qualification Program means any sponsor can now incorporate this AI-driven liver biopsy scoring tool into Phase 2/3 MASH submissions without re-submitting validation data.

Why still relevant: This is the first time a machine learning pathology tool is embedded into the regulatory endpoint infrastructure—a template for AI-driven endpoint tools in other disease areas (e.g., digital pathology for oncology biopsy endpoints in RLT trials).

Source: BauPharma →

🎯 Practical Implications