The Regulatory Leadership Briefing
Insight & News
Field notes on the rules that move markets — written by the people who do the work.
Connor Griggs · 8 min read
The Regulatory Leadership Briefing
Field notes on the rules that move markets — written by the people who do the work.
Connor Griggs · 8 min read
Free proposal · three minutes
Get your free regulatory proposal
No sales call. No account. No obligation.
What is the product?
The Ledger
Practical guidance on the regulations reshaping your work — written to be applied, not just read.
124 articles
S11 replaced routine juvenile animal testing with a weight-of-evidence case made fresh before every pediatric trial — weighted most by age and developing organ systems, and reopened whenever the trial parameters change.
Connor Griggs · 8 min read
Article 117 and the Annex IX ancillary-substance opinion get used interchangeably and shouldn't be — they trigger on opposite combination scenarios, run different clocks, and one of them can block your CE certificate outright.
Connor Griggs · 8 min read
Annex XIV, Part B asks for an analysis against the clinical evaluation report's existing conclusions, not a summary of what PMCF collected — and for Class III and implantable devices, it runs on an annual clock.
Connor Griggs · 8 min read
HUD designation and HDE approval run through separate offices on separate questions. Clear both, and the Annual Distribution Number still caps how many units you can sell for profit — per device, not per company.
Connor Griggs · 8 min read
FDA's 2024 draft guidance says most interchangeable biosimilars no longer need a switching study — but the statutory standard, and FDA's ability to ask for one anyway, didn't move.
Connor Griggs · 8 min read
ICH S9 lets an advanced-cancer program skip studies other oncology programs can't — until the population, the study list, or a rising cure rate says otherwise.
Connor Griggs · 8 min read
The individual case safety report standard that has governed post-approval safety data since 2003 just moved for the first time — and the real change is new reportable data sources.
Connor Griggs · 7 min read
No CE mark, no standard conformity assessment — but Class III custom-made implantables still face a Notified Body quality-system audit under the article's second subparagraph.
Connor Griggs · 7 min read
ICH E2F's DSUR runs on the Development International Birth Date, not the calendar year — and it answers a different question than a PBRER does.
Connor Griggs · 8 min read
The Ombudsman mediates a stalled FDA review; it does not reverse a decision. Here is the formal path under 21 CFR 10.75, and the record that actually wins it.
Connor Griggs · 8 min read
FDA finalized the PBRER format in 2016 to replace the old aggregate safety report with an integrated benefit-risk analysis. Most PBRERs still read like the report it was built to retire.
Connor Griggs · 7 min read
Scientific validity, analytical performance, and clinical performance each get their own evidenced report. Annex XIII still expects a fourth document that ties them together.
Connor Griggs · 8 min read
Both run on the same six-month clock, and the resemblance ends there. One speeds the application it is granted to. The other can be sold to someone else's.
Connor Griggs · 7 min read
Three annexes, not three independent choices. Annex X is a type-examination step that only becomes a route once paired with Annex XI — and most manufacturers never need it.
Connor Griggs · 8 min read
Most CMC teams assume an ICH-harmonized pharmacopoeial test travels across regions untouched. Q4B's actual, annex-by-annex outcome says otherwise more often than expected.
Connor Griggs · 8 min read
21 CFR 801.45's direct-marking exceptions are not a labeling-team call. FDA expects the engineering determination behind them documented, not just the conclusion.
Connor Griggs · 8 min read
Diagnostics companies entering genetic testing often benchmark against a comparable non-genetic assay's classification. Rule 3 removes that option by name.
Connor Griggs · 8 min read
Q5D's cell-bank testing and Q5E's comparability assessments both rest on one unexamined assumption: that the expression construct inside the cell is exactly the one you characterized.
Connor Griggs · 8 min read
Fast Track runs on the lowest evidence bar of FDA's expedited programs. What it doesn't do is hand you Rolling Review — that takes a separate submission schedule FDA has to accept.
Connor Griggs · 8 min read
A non-EU manufacturer needs a written mandate before it can sell into the EU. What Article 11 requires it to cover, and when the Authorised Representative becomes liable alongside the manufacturer.
Connor Griggs · 8 min read
M15 formalizes model-informed drug development evidence around six elements. Model risk — not model sophistication — sets how rigorously a model must be verified before regulators trust it.
Connor Griggs · 8 min read
A 513(g) gets FDA's written view on classification, not feedback on your testing plan. It's fee-gated, non-binding, and FDA met its 60-day goal on roughly one in four FY2023 requests.
Connor Griggs · 7 min read
The Abbreviated pathway swaps full test data for declarations of conformity and summary reports against recognized standards. It doesn't waive the predicate or reduce the testing required.
Connor Griggs · 7 min read
E2E sorts safety concerns into identified risk, potential risk, and missing information. The category decides the pharmacovigilance plan — not a generic monitoring template.
Connor Griggs · 7 min read
A standalone control classifies under Rule 7 only without a quantitative or qualitative assigned value. MDCG 2020-16 Rev.5 just drew the line manufacturers kept crossing without noticing.
Connor Griggs · 7 min read
A dedicated thorough QT study was once the default. Since 2015 it hasn't been — but the waiver is earned with a specific exposure margin and a nonclinical risk assessment most sponsors bolt on too late.
Connor Griggs · 8 min read
RMAT and Breakthrough Therapy share a statute section. RMAT's evidence standard asks about unmet need, not comparison to existing therapies — and a request written to the wrong bar gets denied for the wrong reason.
Connor Griggs · 8 min read
A documented reliance framework can carry most of a batch's paper trail. Annex 16 keeps the certification decision itself personal to the Qualified Person — and specific about what that person still has to verify.
Connor Griggs · 8 min read
Article 22 lets a producer combine CE-marked devices under a declaration, not a new CE mark — until one of three conditions turns that producer into the manufacturer of the whole assembly.
Connor Griggs · 8 min read
Small-molecule stability leans on accelerated data to predict shelf life. Q5C makes real-time data primary for biologics instead — because their degradation doesn't follow the same curve.
Connor Griggs · 8 min read
FDA retired QSIT on February 2, 2026. CP 7382.850 replaces it with six QMS areas and four other requirements, assessed across the product's full life cycle.
Connor Griggs · 7 min read
ICH M13A reached Step 4 in July 2024. A risk classification — not habit — decides whether your bioequivalence study needs one arm or two.
Connor Griggs · 7 min read
Least burdensome has been law since 1997. Since 2022, FDA's own guidance requires a stated basis for every deficiency — and that requirement is a lever sponsors can use.
Connor Griggs · 7 min read
The MDR SSCP is not a one-time CE-marking deliverable. Article 32 ties it to your PSUR cycle, and every update needs fresh notified body validation.
Connor Griggs · 7 min read
ICH M11 finalized a fixed protocol template and a machine-readable technical specification. Adoption is voluntary; the real work is mapping legacy data to it.
Connor Griggs · 7 min read
Not every MDR Notified Body can review every device. MDCG 2019-14 device codes set each one's designation scope — verify it before you apply.
Connor Griggs · 6 min read
IVDR performance evaluation runs on scientific validity, analytical, and clinical performance — evidence an MDR PMCF plan was never built to generate.
Connor Griggs · 8 min read
21 CFR Part 4 lets combination product manufacturers pick a CGMP base independent of the PMOA-assigned center. Here is how that choice actually works.
Connor Griggs · 7 min read
FDA expects chemical characterization under ISO 10993-18 before biological testing fills any remaining gap. The analytical evaluation threshold is the decision that gets made first.
Connor Griggs · 7 min read
E8(R1) asks sponsors to name critical-to-quality factors at protocol design, not discover them at audit — the input ICH E6(R3)'s quality system assumes already exists.
Connor Griggs · 7 min read
The six-month shortage notice gets the compliance attention. The redundancy risk management plan under §506C(j) is the half FDA expects to already exist.
Connor Griggs · 7 min read
Final at Step 4 in August 2024, FDA-final that December. Efficacy can extrapolate from adult data — safety runs on a separate track that rarely does.
Connor Griggs · 7 min read
Twenty-three requirements, three chapters, and a checklist format Annex II actually specifies — not a formality, and not the same question as Annex XVI.
Connor Griggs · 8 min read
One regulation, two obligations with almost nothing in common — a 3-working-day event report and a once-a-year filing. Conflating them is how teams miss both.
Connor Griggs · 8 min read
Every classification memo starts the same way: run the software through Annex VIII Rule 11 and land on a class. MDCG 2019-11 Rev.1 asks a question that has to come first — is this software a medical device at all?
Connor Griggs · 7 min read
Before a bracketing design or an extrapolation argument means anything, ICH Q1A(R2) has already decided how many batches, which storage conditions, and how often you test. Get the framework wrong and everything built on it is wrong too.
Connor Griggs · 7 min read
Rule 21 doesn't classify by device type or duration of contact — it classifies by route and systemic absorption. Get the absorption question wrong and the class swings by two tiers.
Connor Griggs · 7 min read
21 CFR 314.70 doesn't hand you a menu of four filing tracks. It hands you one legal test — potential effect on identity, strength, quality, purity, or potency — and the answer picks the track.
Connor Griggs · 7 min read
Q1D designs the study. Q1E decides how far past your actual data a shelf-life claim can extrapolate — twice the long-term period, capped at 12 months beyond it, and never without statistical backing.
Connor Griggs · 6 min read
Rule 19 doesn't add a branch to the classification tree — it removes an outcome. Any device incorporating nanomaterial is Class IIa at minimum, and the exposure tier decides how much higher.
Connor Griggs · 7 min read
Both designs cut a stability protocol down. They do it on different assumptions — and a design chosen for the wrong reason is the finding a reviewer catches, not the cost it was meant to save.
Connor Griggs · 8 min read
The company proposes a strategy. FDA runs its own health hazard evaluation and assigns the class — and that decision, not the firm's framing, sets the depth of the recall and the reporting clock.
Connor Griggs · 8 min read
Gene therapy sponsors used to design biodistribution studies ad hoc, differently in every region. ICH S12 gives the actual criteria for folding biodistribution into your GLP toxicology study — and for when that combination fails.
Connor Griggs · 7 min read
A De Novo grant doesn't just clear your device. It codifies the special controls you proposed as a new classification regulation — the entry rule every future 510(k) in that product code, including your competitors', now has to meet.
Connor Griggs · 7 min read
A single Cmax,u/Ki,u ratio now decides whether an in vitro result clears your drug or forces a clinical study — and it isn't the number FDA used before 2024.
Connor Griggs · 7 min read
A Clinical Evaluation Report can be exhaustive on the subject device and still fail — because the benchmark Annex I asks for is the competing treatments, not the device's own history.
Connor Griggs · 7 min read
Designation only gets you a hypothesis. A same-drug rival in the same rare disease can still get approved unless you prove clinical superiority at approval, not before.
Connor Griggs · 7 min read
Certain Class III implantable and IIb active devices route through an independent expert panel before CE marking — unless one of three narrow exemptions applies.
Connor Griggs · 8 min read
21 CFR 814.39 sorts every post-approval PMA change onto a track by one test — safety or effectiveness — not by how big the change feels. Guess wrong and FDA resets the clock.
Connor Griggs · 7 min read
A specification list is not a control strategy. ICH Q11 expects process understanding and risk assessment to justify where every control actually sits — and nowhere else.
Connor Griggs · 8 min read
Classification puts most companion diagnostics in Class C under Annex VIII, Rule 3(a). Article 48(3) still requires a scientific opinion from the EMA or a national competent authority before the notified body can certify.
Connor Griggs · 8 min read
Comparable does not mean identical. ICH Q5E's stepwise, risk-based standard starts with analytical data — and only escalates to nonclinical or clinical bridging when that data leaves a real gap.
Connor Griggs · 8 min read
FDORA's 2023 amendments to FDCA section 506(c) let FDA require the confirmatory trial to be underway before approval, and gave the agency a faster path to withdraw it if that trial fails.
Connor Griggs · 8 min read
Repackaging or relabelling a device can make a distributor or importer the legal manufacturer under MDR Article 16, unless the 16(2) exemption applies.
Connor Griggs · 8 min read
Notified bodies must audit MDR manufacturers unannounced at least once every five years — but a vigilance report can trigger one earlier, at your supplier too.
Connor Griggs · 8 min read
A clinical hold's 30-day clock is FDA's to run, not the sponsor's to meet — and it only starts once a 14-day completeness check clears the response.
Connor Griggs · 8 min read
Whether a product is an MDR device or a medicinal product comes down to one legal test — principal mode of action — and MDCG 2022-5 is the guidance manufacturers keep misapplying.
Connor Griggs · 9 min read
A specification that just cites the compendial monograph looks complete and isn’t. Q6A and Q6B build in decision trees for exactly this reason — and most CMC teams never walk through them.
Connor Griggs · 8 min read
The pathway promises a 30-day FDA review instead of 90 — but FDA converts a submission to Traditional mid-review the moment it fails one of three eligibility conditions.
Connor Griggs · 7 min read
A BCS Class I result is not itself a biowaiver. ICH M9 gates it separately on rapid dissolution and excipient risk — either can sink an otherwise clean classification.
Connor Griggs · 7 min read
An executed SPA agreement is written and binding — but FDA's own guidance names three conditions under which it can still revisit it.
Connor Griggs · 8 min read
A multi-regional trial isn't one protocol run in several places. ICH E17 expects the regional-consistency question planned before the trial starts.
Connor Griggs · 8 min read
A refuse-to-file letter is not a scientific verdict. It's FDA saying your NDA or BLA never cleared the completeness threshold that lets substantive review start.
Connor Griggs · 8 min read
Six product groups without a medical purpose — dermal fillers and colored contact lenses among them — now clear the same clinical and post-market bar as a Class III implant.
Connor Griggs · 8 min read
Regulation 2023/607 moved the legacy-device deadline to 2027 or 2028. The extension is conditional, and it can lapse quietly, long before the date arrives.
Connor Griggs · 8 min read
An estimand needs five attributes and a deliberate strategy for every intercurrent event. "Intent-to-treat" names neither — it is a default, not a specification.
Connor Griggs · 8 min read
Notifying a substantial modification takes a sponsor one week. Implementing it takes at least 38 days — and the two clocks get confused more often than the rule does.
Connor Griggs · 7 min read
A REMS can change outside the 18-month, 3-year, and 7-year assessment schedule. Treating it as static between checkpoints leaves sponsors exposed.
Connor Griggs · 7 min read
Under Annex VIII, Class I is a narrow leftover category. Most clinical software lands in IIa or higher the moment it informs a decision.
Connor Griggs · 7 min read
ICH Q5A(R2) governs viral safety. ICH Q5D governs the cell substrate underneath it — derivation, banking, and the tumorigenicity call teams skip.
Connor Griggs · 7 min read
ICH Q7 defines an API starting material as a significant structural fragment. ICH Q11's Q&A makes clear that test alone won't justify it to a reviewer.
Connor Griggs · 7 min read
FDA's March 2026 draft guidance defines what a Form 483 response must contain: one 15-business-day submission with risk assessment, root cause, and CAPA.
Connor Griggs · 7 min read
A design space under ICH Q8(R2) is a multidimensional, interaction-tested region — not a set of proven acceptable ranges filed one variable at a time.
Connor Griggs · 7 min read
Scientific validity is a distinct, evidenced pillar of IVDR performance evaluation — not a literature summary borrowed from MDR clinical evaluation habits.
Connor Griggs · 8 min read
A comparability protocol lets you write the CMC change plan before you need it, then execute the change at a lower reporting tier. The catch: the tier structure isn't the same for a BLA as it is for an NDA.
Connor Griggs · 7 min read
Option 1's residual-solvent table assumes every product is dosed at 10 grams a day. Off that assumption, the table under- or over-restricts the limit, and Option 2's dose-based calculation is the one the guideline actually expects.
Connor Griggs · 7 min read
FDA's May 2026 final guidance sorts every marketing submission into one of three Human Factors Submission Categories, and eSTAR has required the call since August 1. Category 2's rationale is where most teams fall short.
Connor Griggs · 8 min read
Article 86 gets described as an annual PSUR requirement. Class IIb and III update annually — Class IIa only when necessary, at least every two years — and the notified body review path splits the same way.
Connor Griggs · 8 min read
The sponsor makes the first significant-risk call under 21 CFR 812.3(m), but the IRB can overturn it under 812.66 — and FDA's own determination, wherever it comes from, is final and controls over both.
Connor Griggs · 7 min read
Articles 13 and 14 give importers and distributors their own verification and complaint duties. A manufacturer's CE mark doesn't discharge them, and a supply contract can't sign them away.
Connor Griggs · 8 min read
ICH Q5A(R2), finalized November 2023 and adopted by FDA in January 2024, extends viral safety evaluation to viral-vector gene therapy products and accepts next-generation sequencing for virus detection. The scope boundary that matters for CGT sponsors is narrower than the headline: replication-competent, self-replicating products stay out.
Connor Griggs · 7 min read
Finalized alongside ICH Q14, Q2(R2) revised the validation-characteristics framework analytical teams have run for decades. The acceptance criteria a protocol needs now trace to the procedure's stated purpose, not a fixed category checklist.
Connor Griggs · 7 min read
Most sponsors read a Complete Response Letter for the deficiencies. FDA reads the resubmission for scope — and reclassifying it from Class 1 to Class 2 doesn't shorten the clock, it resets it.
Connor Griggs · 7 min read
A folder of risk assessments filed under the name “CCS” looks compliant until an inspector asks who owns it. Annex 1 expects one governed strategy, not a compliance artifact assembled after the fact.
Connor Griggs · 7 min read
Most vigilance procedures default to a single 15-day timer. Article 87 actually sets three — and the fastest one starts on suspicion, not confirmation.
Connor Griggs · 7 min read
Most CMC teams can recite “0.1%” from memory. It isn't in the guideline's default table — the real thresholds are three tiers that move with the dose.
Connor Griggs · 7 min read
Most complaint-handling teams know the 30-day clock under 21 CFR 803.50. Fewer can name the two specific conditions in §803.53 that shorten it to 5 work days — and getting that call wrong runs in both directions.
Connor Griggs · 7 min read
ICH M7's TTC gets quoted like a universal ceiling. It's a starting point for one exposure duration — and the classification workflow that precedes it decides whether it even applies.
Connor Griggs · 7 min read
Designation buys faster FDA access, not a lower bar. Why Sprint discussions and a negotiated Data Development Plan decide what happens at pivotal review — and why waiting until the deficiency letter is using the program backwards.
Connor Griggs · 6 min read
Annex XIV, Part B has governed PMCF since MDR applied. What separates an objective a Notified Body accepts from one that just restates the regulation — and when “not applicable” actually holds up.
Connor Griggs · 7 min read
Full and partial validation carry fixed numeric acceptance criteria. Cross-validation deliberately doesn't — and importing one from elsewhere in the guideline is the mistake reviewers catch.
Connor Griggs · 7 min read
Annex 1 makes pre-use integrity testing the default on every sterilizing filter. What a compliant risk assessment must document to justify skipping it — and why it has to live inside the Contamination Control Strategy.
Connor Griggs · 7 min read
FDA's Dec 2024 final guidance sets three required PCCP elements for AI-enabled devices. What each has to document, and what changed from the 2023 draft.
Connor Griggs · 7 min read
A finished-product test against default limits is not a Q3D risk assessment. The source-by-source work the guideline actually requires, and what the 2022 revision changed in the PDE table.
Connor Griggs · 7 min read
FDA judges substantial equivalence on a two-prong legal test, not resemblance. Split predicates have been off-limits since 2014, and a 2023 draft guidance adds a documentation burden most 510(k) Summaries still skip.
Connor Griggs · 6 min read
The qualification pathways are the easy part to satisfy. The compliance exposure that survives an audit is a PRRC who holds the title without the organizational authority Article 15's duties actually require.
Connor Griggs · 6 min read
FDA has treated Q14 as guidance since March 2024. Filed under the minimal approach by default, an analytical procedure locks down every parameter as an established condition — and every future method tweak becomes a prior-approval change.
Connor Griggs · 7 min read
All six Class D EURL categories are now operational — four since October 2024, two more since May 2026. The distinction that trips manufacturers up: initial conformity verification versus the ongoing batch-testing obligation.
Connor Griggs · 8 min read
FDA's Q-Submission Program covers more than the Pre-Sub. File a Submission Issue Request within 60 days of a hold letter and FDA's stated aim is 21 days — a third of the clock teams get once that window closes.
Connor Griggs · 7 min read
Q13 has been FDA guidance since March 2023. Most control-strategy sections still read like an equipment description — not the traceability and diversion argument reviewers are actually looking for.
Connor Griggs · 8 min read
Basic UDI-DI and UDI-DI are not interchangeable. Here is which device changes force a new UDI-DI — and where the reconciliation across labeling, technical file, and EUDAMED actually breaks.
Connor Griggs · 7 min read
FDA finalized ICH E6(R3) in September 2025. The guideline replaces uniform, procedure-heavy monitoring with critical-to-quality factors identified during protocol design — not retrofitted onto a plan already written.
Connor Griggs · 7 min read
An FMEA ranks failure modes by severity and occurrence. ISO 14971:2019 asks for a full risk management system — a plan, acceptability criteria, and an evaluation of overall residual risk — and reviewers notice exactly which piece is missing.
Connor Griggs · 7 min read
Article 61 still lets a manufacturer rely on another device's clinical data instead of running its own investigation — but only where it can show a genuine, ongoing right to that device's technical documentation.
Connor Griggs · 6 min read
Section 524B requires a vulnerability-management plan and a secure-update process alongside the software bill of materials — and FDA's refuse-to-accept screen is catching submissions that skip them.
Connor Griggs · 6 min read
Regulation (EU) 2024/1860 gave legacy Class C diagnostics a path to 2028 — but only for manufacturers who convert a filed application into a signed Notified Body agreement by 26 September 2026.
Connor Griggs · 7 min read
The primary mode of action determination happens early, in a filing most sponsors treat as paperwork. It quietly sets your review pathway, your user fees, and which quality system you will run for the life of the product.
Connor Griggs · 7 min read
Three of ICH Q10's four pharmaceutical quality system elements produce data. The fourth is supposed to act on it. Most companies run a meeting that reports the data and calls that management review.
Connor Griggs · 7 min read
Two years of transition ended on February 2, 2026. The manufacturers still exposed are not the ones without a certificate — they are the ones who mistook the certificate for the regulation.
Connor Griggs · 8 min read
The 2023 revision named the quiet failure modes of pharmaceutical risk management — subjectivity, ritual formality, decisions made before the assessment starts — and told industry to fix them. Most programs haven't.
Connor Griggs · 8 min read
The European database has been voluntary for years. A 2025 Commission decision set the clock for mandatory use — and the work that decides whether you make it is the data you don't have yet.
Connor Griggs · 8 min read
Established conditions, PACMPs, and the PLCM document are levers, not paperwork. Why most manufacturers capture none of the benefit — and how to change that before your next submission.
Connor Griggs · 7 min read
PMS remains the most common source of notified body findings, and the pattern is consistent: paperwork where a system should be. The recurring gaps, the EUDAMED clock, and how to build PMS that survives audit.
Connor Griggs · 9 min read
The statutory 30-day clock hasn't moved — almost everything around it has. Where a leaner, reorganized FDA actually changes your IND risk profile, and how to plan the interactions that remain.
Connor Griggs · 8 min read
Nothing matches that search.
New analysis and the regulatory shifts that actually matter — what we're telling clients, in one monthly briefing. No sales sequence. Unsubscribe anytime.
From the newsroom
Regulatory News is our independent newsroom — every story sourced to the primary record, published the day it lands. These are the beats that touch this page. RegulatoryNews.com →
If an article surfaced a problem that looks like yours, bring it to us. The analysis is free — the judgment behind it is how we work. We respond within one business day, and all inquiries are strictly confidential.