Biohaven Advanced BHV-7000 to Pivotal Trials Without Answering a Question the FDA Requires Every Sponsor to Answer

On September 4, 2026, the FDA sent Biohaven a letter that stopped new patient enrollment in every BHV-7000 epilepsy study. The agency’s stated reason: rodent studies had flagged a specific metabolite, and the FDA did not have enough preclinical data to assess what that metabolite does in humans. Biohaven disclosed the action in an SEC filing the same week. The drug, opakalim, an ion channel modulator targeting Kv7, had been advancing toward what Biohaven positioned as a pivotal program in epilepsy. The partial hold froze that program mid-stride, not because of a clinical signal, not because of a site failure, but because a fundamental pharmacology question about the molecule’s own metabolic profile remained open.

That question should have been closed in the lab, not answered by the FDA in a hold letter.

What the Rodents Told Them

The sequence here matters. The FDA’s partial hold was precipitated by findings from rodent studies, studies that, by definition, were already completed and in Biohaven’s hands. The agency reviewed preclinical data, identified a metabolite of concern, and concluded the existing nonclinical package was insufficient to characterize the risk that metabolite poses in human subjects. Biohaven’s response was to pause new enrollment and commit to generating the additional nonclinical studies the FDA requested.

The uncomfortable arithmetic: Biohaven recorded a $93.7 million R&D expense in Q2 2022 tied to the Kv7 platform acquisition that brought BHV-7000 into the portfolio, plus a $25 million milestone payment that became payable that same June. More than $118 million committed to a platform before the company had resolved whether its lead asset’s metabolite profile was characterizable under FDA standards.

The FDA’s Safety Testing of Drug Metabolites guidance for industry, last revised in March 2021 and aligned with ICH M3(R2), is unambiguous on timing. If a metabolite is disproportionate, meaning it is identified only in humans or appears at higher plasma concentrations in humans than in any of the animal species used in standard toxicology testing, sponsors must complete the necessary toxicity studies and provide results before advancing. The guidance defines exactly when that work needs to be done relative to clinical progression. It is not an aspirational standard. It is a precondition for continued enrollment.

Biohaven either identified this metabolite late, failed to recognize its regulatory significance when it appeared in rodent data, or made a judgment call that the characterization could happen in parallel with clinical advancement. Whatever the internal logic, the FDA’s September 4 letter says that logic was wrong.

The Sponsor Defense Versus the Regulatory Record

Biohaven’s public posture has been measured: the hold is limited to new enrollment, patients already in the study can continue under a protocol that allows it, and the company intends to generate the additional data and seek lifting of the hold. That is a reasonable operational response. It is also precisely the kind of response that obscures how avoidable this was.

Partial clinical holds of this type, triggered by metabolite characterization gaps rather than acute clinical signals, follow a recognizable pattern in FDA enforcement. The agency does not place a hold because it believes the drug is dangerous. It places a hold because the sponsor cannot yet demonstrate the drug is safe enough to keep exposing new subjects to. The distinction sounds subtle. For a Phase III epilepsy program, it carries real operational cost: enrollment frozen, site contracts strained, timelines displaced, competitive positioning in an active therapeutic space eroded.

The FDA’s own September 4 hold letter framed the issue plainly: insufficient information to assess the risk of a specific metabolite in human subjects, pending results from additional nonclinical studies. That language maps directly onto the 2021 metabolite safety guidance’s requirement that disproportionate metabolite characterization precede, not accompany, the clinical development stage at which human exposure becomes substantial. Biohaven crossed that threshold without closing the loop.

The deeper problem is structural. When a company acquires a platform for $93.7 million and simultaneously triggers a $25 million milestone, the internal pressure to advance that asset clinically is intense. Timelines get compressed. Regulatory readiness reviews get optimistic. The metabolite characterization work that belongs in nonclinical development gets scheduled for “as needed” rather than “before enrollment opens.” That is the decision sequence the FDA’s hold just made visible.

What This Reveals About Preclinical Due Diligence

Biohaven is not operating outside the industry norm in the way it acquired and advanced BHV-7000. The Kv7 channel is a credible epilepsy target, opakalim has a mechanism with scientific rationale, and building toward a pivotal program in refractory epilepsy reflects reasonable commercial ambition. What the September 4 hold exposes is a failure mode that is common precisely because it is invisible until the FDA makes it visible: sponsors treating metabolite safety work as a regulatory checkbox rather than as a scientific prerequisite.

The ICH M3(R2) guideline, which the FDA’s 2021 metabolite guidance explicitly cross-references, requires that nonclinical safety packages be adequate to support the stage of clinical development being conducted. “Adequate” has a specific meaning in that context: it includes characterization of metabolites that appear in animal studies at concentrations that may be relevant to human exposure. A rodent study that surfaces a metabolite of potential concern is not a finding to note and move past. It is a trigger for a defined regulatory workflow that must be completed before the agency will consider new enrollment justified.

The September 4 letter arrived more than four years after Biohaven paid $118 million-plus to own this program. If the metabolite appeared in rodent studies conducted during that period, the gap between when the finding was available and when the FDA formalized its concern represents the window in which this hold could have been prevented. Sponsors reviewing their own ion channel, CNS, and epilepsy programs should be asking that question about their own nonclinical packages today, not after the hold letter arrives.

For sites running BHV-7000 studies: patients currently enrolled can continue under the partial hold’s parameters, but no new subjects may enter. Site coordinators managing screening pipelines need to communicate that boundary clearly, document the regulatory basis for the enrollment pause, and avoid the common operational error of continuing to screen candidates against a protocol whose enrollment status has changed. The 8-K was filed; the sites need the operational memo to follow it the same week, not the same quarter.

For FDA-watchers: the agency’s willingness to impose a partial hold on a pivotal-stage program over a nonclinical gap signals continued enforcement rigor on metabolite characterization, a posture consistent with the 2021 guidance revision and with ICH harmonization priorities. Watch Biohaven’s IND response timeline. If the additional nonclinical studies require new animal cohorts rather than reanalysis of existing samples, the hold could run six to twelve months before the FDA has sufficient data to lift it, which would push any pivotal readout well into 2028.

References

1. Endpoints News, “FDA places partial hold on Biohaven’s ion channel drug”
2. Investing.com / SEC Filing, “Biohaven pauses new enrollment in epilepsy trials after FDA partial clinical hold” (September 4, 2026)
3. Clinical Trials Arena, “Biohaven opakalim pivotal programme: FDA partial clinical hold”
4. SEC / Biohaven Ltd., Form 10-K (2022): BHV-7000 Kv7 Platform Acquisition R&D expense and milestone disclosure
5. FDA, “Safety Testing of Drug Metabolites: Guidance for Industry” (March 2021)