Three Guidance Events in One Week Redraw the Statistical Rulebook

The week of May 29, 2026 produced three distinct regulatory events, each demanding immediate attention from a different segment of the biometrics community — and each carrying its own implementation uncertainty.

The BE statistics guidance: long overdue, now operative

FDA finalized “Statistical Approaches to Establishing Bioequivalence” on May 29, superseding the February 2001 original and the December 2022 draft — the first full revision of FDA’s core BE statistics guidance for IND, NDA, and ANDA submissions in over 25 years. The companion ANDA-PK guidance, finalized the same day, closes a nearly five-year draft-to-final gap. The Federal Register notice catalogs the deltas: clarifying treatment of estimands and intercurrent events, updated sample-size determination, reworked outlier handling, and the absorption of two methodology blocks that previously lived elsewhere — population and modified population BE (from product-specific guidances), and reference-scaled average BE for narrow-therapeutic-index and highly variable drugs (from the August 2021 ANDA-PK draft). The PDF carries the operative parameter values; the notice already tells teams which sections to read first.

M11 and the E9(R1) dividend

ICH M11 (CeSHarP), adopted at Step 4 in November 2025 and finalized by FDA on May 21, delivers a three-part package: guidance document, standardized protocol template, and a technical specification with harmonized, machine-readable data fields for cross-jurisdictional electronic exchange. As covered last issue, this is a structurally demanding standard for protocol authoring teams. What’s new here is the frame: M11 finalization lands alongside a regulatory-insider validation of the estimand dividend it depends on. A former FDA OB Division III primary reviewer — 32 years of industry experience, 8 years at CDER — reports directly that applications using ICH E9(R1)-aligned protocols and SAPs are observably more reviewable and generate fewer information requests. That’s the first on-record confirmation of review-efficiency ROI from someone who actually processed the submissions. M11’s structured statistical methods sections only amplify that effect — if teams implement them correctly.

The MHRA IMA: writing the rules from scratch

The MHRA’s proposed Rare Disease Therapies Framework, open for consultation until July 30, 2026, introduces an Investigational Marketing Authorisation that collapses clinical trial approval and marketing authorisation into a single progressive pathway for conditions affecting ≤1 in 50,000 people. The press release commits the framework to adaptive clinical trial approaches, real-world data, prior knowledge from related programs, and predictive modelling — flagged as potentially “essential” for very small or highly heterogeneous populations. The framework is co-developed with the Rare Disease Consortium — MHRA, HRA, NICE, DHSC, NHS England, patient groups, academia, and industry — with NICE retaining NHS access control. The biometrics implication is structural: iterative rolling submissions will require adaptive SAPs and estimand frameworks designed for ultra-rare populations from the outset, not retrofitted at submission. The consultation closes July 30; it is the window to influence quantitative evidentiary thresholds before they harden.

The simultaneity problem

The sharpest observation about this week is not any individual guidance — it’s that the statistical infrastructure governing the oldest part of drug development (generic BE) and the newest (single-patient-scale ultra-rare) is being rewritten at the same moment, by agencies operating under varying degrees of institutional pressure. FDA leadership instability and known OB vacancies mean that guidance-to-review alignment cannot be assumed. Teams implementing all three updates in parallel should build in that uncertainty, not assume it away.

Protocol read: The structural infrastructure of clinical-evidence regulation is being rewritten at both ends of the spectrum at once, by agencies with varying levels of institutional stability. The implementation work is consequential, the institutional risk is real, and “wait and see” is no longer a defensible posture on any of the three.

What to do now:

• Read the FDA BE statistics guidance PDF before the next BE SAP cycle; the Federal Register notice flags the changed sections (estimands, sample-size, outliers, population BE, NTI/HVD RSABE) but the parameter values are in the PDF.
• File MHRA IMA consultation comments before July 30 if any program touches ultra-rare populations — quantitative evidentiary thresholds are still movable, post-finalisation they will not be.
• Build M11 implementation into protocol-authoring workflows now; the estimand-dividend evidence is no longer speculative and the structural template is operative this week.