A clinical hold is the regulatory equivalent of having your driver's license suspended while you're on the highway. FDA tells you to stop dosing patients — or stop before you start — and you don't proceed until the issue is resolved. For a Series A biotech spending $300-500K per month on operations, a clinical hold isn't an inconvenience. It's a potential company-ending event.
I've talked to regulatory affairs leads at a dozen small biotech companies over the past year. The fear of a clinical hold comes up in every single conversation. Not because these people are paranoid — because they've seen it happen. A colleague's company. A competitor. Someone at a conference. The stories circulate.
And the frustrating part: most clinical holds are preventable. The deficiencies FDA flags are things that could have been caught before filing.
The legal framework
Clinical holds are governed by 21 CFR 312.42. FDA can impose a clinical hold at two points:
- Before the trial begins — a "hold on opening" when FDA reviews the IND and finds insufficient basis to allow the trial to proceed
- After the trial has started — a "hold on an ongoing study" when new information raises safety concerns
For Phase 1, FDA has 30 calendar days after receiving the IND to notify you if they're placing a clinical hold. If you don't hear anything by Day 30, you may proceed. This is the "30-day safety review" period, and the silence-means-assent mechanic is both efficient and terrifying. No news is good news, but you spend those 30 days refreshing your email like it's match day.
FDA must specify the deficiency in writing within 30 days and has a subsequent call with the sponsor within 30 days of issuing the hold. Under the FDARA amendments, they're required to include clear, specific recommendations for resolving the hold. In practice, the quality of these recommendations varies. Sometimes you get a precise list of what's needed. Sometimes you get language that requires three follow-up calls to interpret.
The 7 most common reasons
Based on FDA's own reports, enforcement data, and my conversations with regulatory affairs professionals:
1. Insufficient nonclinical safety data
This is the big one. FDA's stated criterion per 21 CFR 312.42(b)(1)(i): "Human subjects would be exposed to an unreasonable and significant risk of illness or injury."
What that looks like in practice: missing or inadequate repeat-dose toxicity studies, gaps in the safety pharmacology core battery, genotoxicity data that doesn't meet ICH S2(R1) requirements, or toxicology study duration that doesn't match the proposed clinical dosing period per ICH M3(R2) Table 1.
The most common specific gap I've seen discussed: the non-rodent species study. Teams run rat tox, get clean data, and file without dog or primate data. FDA flags it immediately. This alone can account for a 6-9 month delay because large-animal toxicology studies can't be scheduled on short notice — CRO capacity for beagle and NHP studies is booked months in advance.
2. Inadequate dose justification
Starting dose calculation is not optional formatting. FDA wants to see a NOAEL from your GLP toxicology studies converted to a human equivalent dose (HED) using body surface area scaling, with a safety factor applied (typically 1/10 of the HED). For biologics, the MABEL approach may be more appropriate.
What triggers a hold: the math doesn't add up, the NOAEL isn't from a GLP study, the species scaling is wrong, or the safety factor isn't justified. I've also seen holds where the dose justification was adequate but FDA couldn't find it — it was buried in an appendix instead of being in the Nonclinical Overview where reviewers expect it.
3. GLP compliance issues
21 CFR Part 58 is non-negotiable for pivotal nonclinical safety studies. A study that wasn't conducted under GLP is, from FDA's perspective, a study that might as well not exist.
The tricky part: partial GLP compliance is sometimes worse than no GLP. If a study claims GLP status but the QA audit reveals deviations — equipment calibration gaps, protocol deviations not documented, test article characterization incomplete — FDA may reject the study entirely. This is worse than submitting a non-GLP study and flagging it as exploratory, because it suggests carelessness with the studies that are supposed to be your most rigorous.
4. Inadequate manufacturing controls
Module 3 problems. The drug substance or drug product characterization is insufficient. Stability data doesn't cover the planned trial duration. The manufacturing process isn't adequately described. Impurity profiles aren't characterized.
For small molecules, the common gap is insufficient impurity qualification — you haven't identified or qualified impurities above the ICH Q3A/Q3B thresholds. For biologics, it's comparability. If your manufacturing process changed between the tox batch and the clinical batch, you need to demonstrate comparability, and the bar for that is higher than most first-time sponsors expect.
5. Protocol design deficiencies
FDA reviews your Phase 1 protocol as part of the IND. If the stopping rules are unclear, if there's no dose-limiting toxicity definition, if the dose escalation scheme doesn't have adequate safety monitoring between cohorts, or if the informed consent doesn't adequately describe known risks from preclinical data — any of these can trigger a hold.
The informed consent gap is particularly common. Your toxicology studies showed liver enzyme elevations at the high dose. That finding needs to be in the informed consent. Not as a footnote. As a clearly described risk. FDA checks.
6. Investigator qualification issues
Less common, but it happens. The principal investigator's Form FDA 1572 is incomplete or indicates the investigator lacks adequate facilities or training for the proposed trial. Financial disclosure issues. Debarment history.
7. Missing pharmacology data
A gap in primary pharmacodynamics data — you haven't demonstrated that your drug does what you claim it does in an animal model. For targeted therapies, this includes evidence of target engagement. FDA wants to see that the drug hits the target at the doses you plan to use in patients.
For secondary pharmacology, the concern is off-target effects. If your kinase inhibitor hits 15 other kinases at clinically relevant concentrations, FDA wants to know what those other kinases do. The secondary pharmacodynamics data should address this.
What actually happens when you get a hold
Day 1: You get a letter. It specifies the deficiency. Your stomach drops.
Day 2-7: Internal panic, followed by a structured response. You assemble the team — regulatory, nonclinical, CMC, clinical. You parse the hold letter sentence by sentence.
Within 30 days: FDA schedules a call to discuss the hold. This call is important. Ask clarifying questions. Make sure you understand exactly what they want. Don't argue. Don't try to convince them their concern is overblown. Listen. Take notes.
Then you fix it. Run the missing study. Generate the missing data. Revise the protocol. Whatever they asked for.
When you're ready, you submit a "complete response" to the clinical hold. FDA has 30 days to review it. If they're satisfied, they lift the hold. If not, you get another letter.
Total timeline for a typical clinical hold resolution: 3-12 months, depending on what's missing. If it's a documentation issue (buried dose justification, missing informed consent language), you might resolve it in 6-8 weeks. If it's a missing GLP toxicology study, you're looking at 6-12 months for study execution alone, plus CRO scheduling.
Prevention
Here's what I'd tell any team that's 12-18 months from an IND filing:
Take the pre-IND meeting. I wrote about this in detail in the pre-IND meeting guide. The meeting costs you nothing except time. FDA tells you, in writing, whether your nonclinical plan is adequate. If they flag gaps, you fix them before filing. If they agree with your plan, you have documentation that you followed FDA's advice. This alone prevents a large percentage of clinical holds.
Run a gap analysis against ICH M3(R2). Go through Table 1 of M3(R2) line by line. Map your completed and planned studies against the requirements for your clinical phase and proposed duration. Every cell in that table should have a study assigned to it. Empty cells are clinical hold risks.
The RegFo compliance checker does this automatically — you input your study package and get a gap analysis with specific ICH citations. But even doing it manually with a spreadsheet is better than not doing it at all.
Check GLP status early. Don't wait until you're assembling the IND to discover that your pivotal tox study has a GLP deviation. Get the QA audit report from your CRO early. Read it. If there are deviations, discuss them with the study director. Some deviations are minor and can be documented. Others require repeating the study.
Verify your dose justification math. Walk through the NOAEL-to-HED calculation yourself. FDA's dose guidance specifies the body surface area conversion factors by species. Make sure you're using the right one. Make sure the safety factor is appropriate. And put the entire calculation in the Nonclinical Overview, not in an appendix.
Cross-reference your Module 2 summaries with Module 4 data. The Nonclinical Overview should accurately summarize what's in Module 4. If there's a discrepancy — the summary says 28-day repeat-dose study but the actual report is 14 days — FDA will notice. This sounds obvious. It happens more than you'd think, especially when different people write different sections.
Review the informed consent against your tox findings. Every adverse finding from your nonclinical studies should be reflected in the informed consent document. Liver enzyme elevations, target organ toxicity, reproductive findings — all of it. FDA reviewers cross-reference these.
The cost of a hold
I've tried to quantify this from conversations with biotech companies:
- Direct costs: CRO fees for additional/repeated studies ($200K-$2M depending on what's needed), regulatory consultant fees for hold resolution ($50-100K), team time diverted from other programs
- Indirect costs: 6-12 month delay to clinical timeline, damaged credibility with investors, missed partnership timelines, competitor advantage
- Opportunity cost: burned runway during the hold period ($300-500K/month for a typical Series A-C biotech)
A 6-month clinical hold at a company burning $400K/month costs $2.4M in runway alone, before you factor in the cost of fixing the problem. For a Series A company with $15-20M raised, that's 12-16% of total funding gone to a preventable error.
This is why regulatory affairs people are cautious. They've done the math.
The mindset shift
The teams that avoid clinical holds think about FDA requirements at the beginning of their preclinical program, not at the end. They don't design studies and then check for compliance afterward. They start with the requirements and design studies to meet them.
It's the difference between building to code and getting a building inspection after you've already built the house. One approach involves a blueprint. The other involves a sledgehammer.
Related reading:
- FDA IND Submission Checklist 2026 — complete requirements by CTD module
- GLP Compliance Checklist for Preclinical Studies — 21 CFR Part 58 requirements
- How to Check Preclinical Studies Against ICH Guidelines — 8-step gap analysis
Want to check your preclinical package against clinical hold risk factors? RegFo's compliance checker maps your studies against ICH M3(R2) requirements and flags gaps with severity ratings — before FDA does.
FAQ
How long does a clinical hold last?
A clinical hold lasts until FDA lifts it — there's no automatic expiration. Once you submit a complete response addressing each point in the hold letter, FDA has 30 calendar days to review it and either lift the hold or issue a follow-up response with remaining concerns. In practice, resolution timelines break down by hold type: documentation gaps (buried dose justification, missing GLP statement, incomplete informed consent) typically resolve in 6-10 weeks. Missing toxicology studies require 5-9 months for CRO scheduling and study execution before you can even submit a response. Protocol design gaps fall in between — 8-16 weeks depending on complexity. The average across all hold types runs 3-12 months. For a company spending $400K/month, the math on a 6-month hold is straightforward: $2.4M in burn rate before you've dosed a single patient.
Can you appeal an FDA clinical hold?
There's no formal appeals process for a clinical hold under 21 CFR 312.42. Your options are to respond to the hold by addressing the gap FDA identified, or to request a meeting with FDA to discuss the basis for the hold and clarify what data would satisfy their concern. The meeting request is the practical substitute for an appeal — you can argue your case, present additional data, or negotiate what constitutes a complete response. FDA must schedule that meeting within 30 days of the hold being issued. If you believe the hold is scientifically unwarranted, you can pursue a formal dispute resolution through FDA's Center for Drug Evaluation and Research dispute resolution process, but this is slow and rarely changes outcomes. The more effective path is the clarification meeting, clear documentation of what FDA wants, and a targeted complete response.
What is the difference between a clinical hold and a partial clinical hold?
A full clinical hold under 21 CFR 312.42 stops all clinical investigations under the IND. No patients can be enrolled, and patients already in the trial must stop receiving the investigational drug unless FDA determines that discontinuation poses greater risk than continuation. A partial clinical hold restricts specific parts of the investigation — typically a specific dose level, patient subpopulation, or clinical site — while allowing other aspects to continue. For example, FDA might issue a partial hold stopping enrollment of women of childbearing potential pending embryo-fetal development data, while permitting continued enrollment of men. Partial holds are more common when the safety concern is specific rather than program-wide. Both types require a written complete response to lift, and both carry the same 30-day FDA review clock after submission.
Does a clinical hold mean my drug failed?
No. A clinical hold means FDA found a gap in your submission package — not that your drug is unsafe or ineffective. The majority of clinical holds stem from missing data, inadequate documentation, or study design issues that are fixable. Common triggers include missing non-rodent toxicology data (a scheduling problem, not a safety failure), non-GLP hERG assay submitted as pivotal (a compliance paperwork issue), or buried dose justification (a formatting problem). None of these mean your compound has a safety problem. Approximately 5-10% of original INDs receive clinical holds; most are resolved and proceed to trial. A clinical hold on an ongoing study is more serious because it involves new safety signals, but even then resolution is possible. The hold letter tells you exactly what FDA needs — treat it as a detailed checklist, not a rejection.
How do I respond to a clinical hold?
Start within 48 hours of receiving the hold letter: assemble your full team — regulatory, nonclinical, CMC, clinical. Parse every sentence of the hold letter; each stated gap requires a specific, addressable response. Request the clarification call FDA owes you within 30 days — use it to confirm exactly what data will satisfy each concern before you spend money on studies. Then prepare a complete response that addresses each hold issue sequentially. Don't argue that FDA's concern is overblown; give them the data they asked for. If a study is needed, get a CRO on the phone that week — slots for GLP toxicology, especially large-animal work, book out 3-6 months. Submit the complete response only when every point is addressed. A partial response resets the 30-day clock and signals disorganization. After submission, FDA has 30 calendar days to lift the hold or issue follow-up.
Can FDA place a clinical hold on a Phase 2 or Phase 3 trial?
Yes. Under 21 CFR 312.42, FDA can issue a clinical hold at any phase of development and at any point — before trials begin or after they're underway. For ongoing Phase 2 and Phase 3 trials, the most common triggers are unexpected serious adverse events that suggest an unreasonable risk to participants, new nonclinical findings that weren't present at IND filing (such as carcinogenicity signals), or manufacturing issues that raise concerns about product quality at scale. Late-phase holds are generally more consequential because more capital has been invested, patient populations are larger, and the reputational impact is wider. FDA also has the authority under 21 CFR 312.42(b)(2) to place a hold if an IND phase is complete and the data don't support proceeding to the next phase — though this is less commonly used than safety-based holds.
What percentage of INDs receive clinical holds?
FDA doesn't publish a precise annual clinical hold rate broken down by phase, but available data suggests roughly 5-10% of original IND submissions receive clinical holds before first patient dosing. FDA's own reports indicate they review several thousand original INDs annually and issue holds on a meaningful fraction. The rate varies significantly by sponsor experience and drug type: first-time sponsors, small biotechs without dedicated regulatory affairs teams, and sponsors in novel modality areas (gene therapy, cell therapy) face higher hold rates. For Phase 1 oncology programs operating under ICH S9, nonclinical requirements are less stringent, which reduces hold rates in that subset. The more practically useful number: among holds that do occur, over 80% trace to the same five categories — missing nonclinical data, inadequate dose justification, GLP gaps, CMC issues, and protocol design problems — all of which are preventable with early gap checking.