Welcome to this week’s digest. For March 30 to April 5, 2026, the signal is clear: AI governance is becoming explicitly regulatory, gene therapy continues to expand beyond classic rare disease lanes, and radioligand theranostics is accelerating with new isotopes, targets, and real-world evidence programs.
EMA and FDA released a harmonized “Guiding Principles of Good AI Practice in Drug Development,” outlining 10 shared expectations across preclinical, clinical, manufacturing, and post-market use. Core requirements include human-centric design, context-of-use clarity, risk-based validation, lifecycle monitoring, and GxP-aligned documentation.
Why it matters: This marks a practical shift from fragmented guidance to coordinated regulatory expectations. Sponsors using AI for recruitment, trial optimization, external control arms, or regulatory support now need auditable governance and validation from day one.
Source: FDA/EMA Joint Publication →CDER highlights a sharp rise in submissions that include AI components spanning nonclinical, clinical, manufacturing, and post-market contexts. Regulatory focus is tightening around context-of-use definition, model risk assessment, validation strategy, and traceable lifecycle oversight.
Why it matters: AI outputs are increasingly treated as regulated evidence artifacts, not informal analytics. This has immediate implications for sponsor governance models, CRO/CDMO oversight, and timing of FDA engagement on credibility plans.
Source: Pharmaceutical Outsourcing (FDA CDER context) →The EMA/HMA Network Data Steering Group adopted its 2026–2028 workplan, prioritizing interoperability, real-world evidence, and AI implementation across medicines regulation. The plan aligns with broader EU pharmaceutical reforms and anticipated AI-specific implementation pathways.
Why it matters: Global programs should expect convergence in principles but divergence in implementation detail. Teams running US/EU strategies in parallel will need region-specific validation packages and documentation mapping.
Source: EMA/HMA Network Data Steering Group →On March 27, 2026, FDA approved KRESLADI (marnetegragene autotemcel) for LAD-I in patients without suitable allogeneic transplant options. The approval includes long-term evidence expectations and is paired with a Rare Disease Priority Review Voucher.
Why it matters: The decision reinforces a viable pathway for accelerated access coupled to extended post-approval evidence generation, including registries and long-term benefit tracking in ultra-rare settings.
Source: Yahoo Finance (Rocket Pharmaceuticals) →AskBio announced completion of enrollment in the Phase 2 GenePHIT study of AB-1002 for non-ischemic cardiomyopathy/NYHA class III heart failure, with initial results expected in H1 2027.
Why it matters: This is an important scale signal for AAV programs in chronic cardiovascular disease, where durability, dosing, and safety expectations differ from classical orphan paradigms.
Source: AskBio press release coverage →At its March 23–26, 2026 meeting, CHMP recommended conditional marketing authorization for Adstiladrin (nadofaragene firadenovec) in BCG-unresponsive non-muscle invasive bladder cancer with CIS.
Why it matters: This extends EU gene therapy momentum into localized oncology contexts and strengthens the precedent that accelerated access can be paired with staged confirmatory evidence obligations.
Source: EMA CHMP meeting highlights →Huntsman Cancer Institute reported first-in-human treatment of the first three patients in the Phase 1 RYZ401 study (NCT07165132), an Actinium-225 somatostatin receptor radioligand sponsored by RazyeBio/Bristol Myers Squibb.
Why it matters: The study expands alpha-emitter development beyond prostate-focused PSMA programs into neuroendocrine and other SSTR-positive solid tumor biology, where early dosimetry and safety data will be highly decision-driving.
Source: Huntsman Cancer Institute →Radiopharm Theranostics announced first-patient dosing in NCT07259213 for RAD 402, a Terbium-161 monoclonal antibody radioligand targeting KLK3 in advanced prostate cancer, with initial data anticipated in H2 2026.
Why it matters: This introduces a non-PSMA biological axis and a distinct isotope profile, potentially relevant for PSMA-heterogeneous disease and future sequencing strategies.
Source: Yahoo Finance (Radiopharm Theranostics) →Novartis initiated a prospective non-interventional study (NCT07290270) in China enrolling around 170 patients with mCRPC and mHSPC treated per local Pluvicto labeling through 2029.
Why it matters: Structured RWE generation in a major market will influence future access discussions, label strategy, and regional confidence in Lu-177 pathway adoption.
Source: Novartis clinical trials registry →The joint principles emphasize risk-proportionate validation, lifecycle drift monitoring, multidisciplinary oversight, and transparent documentation of AI role/scope in development decisions.
Insight: Clinical development teams should assume AI governance expectations similar to other regulated quality systems, including traceability and reproducibility standards.
Source: FDA/EMA joint statement →China’s 2026 Drug Administration Law implementing regulations (effective May 15, 2026) formalize accelerated pathways, overseas data acceptance, data exclusivity extensions, and flexible manufacturing provisions.
Insight: For global developers, China is increasingly a strategic launch/approval jurisdiction rather than a late regional add-on.
Source: Morgan Lewis analysis →Industry policy commentary highlights upcoming UK/EU changes, CRO pressure on timelines/privacy, and support for FDA’s risk-based AI guidance as a practical template.
Insight: Sponsors should prepare for cross-region alignment at the principle level while maintaining dual-track operational validation plans.
Source: BioSpace policy report →Why still relevant: This earlier-line use case changed treatment sequencing and expanded eligible population, directly shaping current comparator strategy for next-wave radioligand programs.
Source: Novartis media release →Why still relevant: Early dose-escalation safety/response benchmarks from heavily pretreated mCRPC continue to anchor expectations for alpha-emitter development and sequencing design.
Source: Grand Rounds in Urology (ASCO GU coverage) →Why still relevant: Organ dose findings for Lu-177 zadavotide guraxetan informed protocol adaptation and set practical precedent for dosimetry-driven cycle decisions in pivotal radioligand programs.
Source: BioSpace / Curium release →Why still relevant: Hold resolution is a useful regulatory signal for cardiac gene therapy risk management, CMC discipline, and contingency planning in late-stage programs.
Source: AHLA Daily →Why still relevant: The breadth of approvals across modalities supports China-first strategy discussions for innovative portfolios, including gene therapy and theranostics programs.
Source: DIA Global Forum →Operationalize context-of-use definitions, risk-credibility plans, lifecycle monitoring, and GxP-compliant documentation for all AI-enabled development workflows, with separate FDA/EMA implementation maps.
Recent FDA/EMA decisions reinforce conditional or accelerated pathways that rely on post-marketing evidence; registry architecture should be pre-specified early, not retrofitted post-approval.
First-in-human Ac-225 and Tb-161 programs increase pressure for robust renal, marrow, and salivary safety characterization, plus target-expression stratification and long-term follow-up design.
With regulatory reforms and strong approval throughput, China should be evaluated as an integrated development jurisdiction (not just commercialization geography), including early NMPA interaction models.
RLT programs should pair efficacy endpoints with quality-of-life and utilization-relevant outcomes to support payer negotiations and post-approval uptake across regions.