Ethosuximide Fell Behind Placebo in an IBS Pain Trial

Ethosuximide, an older drug used for absence seizures, did not beat placebo for irritable bowel syndrome (IBS) abdominal pain in the best available clinical test. At a planned interim analysis, 14% of participants receiving the drug met the study's response definition, compared with 36% receiving placebo—a 22-percentage-point deficit for ethosuximide. The estimate came from a small interim sample and does not settle whether the underlying gut target is useful. It does make ethosuximide a poor candidate for self-experimentation.

The trial matters because it tested a specific idea about IBS pain: some patients may have oversensitized nerves in the gut wall, and T-type calcium channels may help keep those nerves firing. That is a plausible biological hypothesis. A plausible hypothesis, however, still needs a drug that reaches the right tissue at a tolerable dose and improves outcomes patients can feel.

A peripheral target with an inconvenient drug

IBS affects roughly 11% of people worldwide and can cause recurrent abdominal pain alongside diarrhea, constipation, or both. Treatments often help some patients but leave plenty of room for better options. The rationale for ethosuximide focused on Cav3.2, a T-type calcium channel. The 2026 randomized trial by Kerckhove et al. describes Cav3.2 channels as overexpressed in colonic mucosa from people with IBS and notes that inhibiting them prevented visceral hypersensitivity in preclinical models.

Visceral hypersensitivity means the bowel's sensory nerves react too strongly to stretching or other ordinary signals. It is one reason IBS pain can feel severe even when routine testing finds no structural disease. The preclinical findings support Cav3.2 as a target worth studying, but animal experiments and expression findings do not prove that blocking this channel will reduce pain in people.

Ethosuximide can inhibit T-type calcium channels, which made it an available way to test the idea. It also crosses into the central nervous system and inhibits all three T-type channel subtypes—Cav3.1, Cav3.2, and Cav3.3. That broad activity is a problem if the desired effect depends on selectively quieting Cav3.2 in gut sensory nerves. A drug can be biologically active and still be the wrong instrument for the job.

What the trial actually tested

Kerckhove and colleagues ran a multicenter, double-blind, placebo-controlled trial across 10 French academic gastroenterology departments. Adults met Rome IV criteria for IBS. Participants took ethosuximide or placebo; the active drug was titrated over 28 days to a maximum of 500 mg per day, followed by a 12-week treatment period.

The primary endpoint was demanding and clinically relevant. A participant had to achieve both at least a 30% reduction in weekly abdominal-pain numerical-rating-scale score and a score of at least 4 out of 5 on a Subjective Global Assessment. In other words, lower pain alone was not enough; participants also had to report meaningful overall improvement.

A prespecified blinded interim analysis occurred after 50 participants, or about 25 per group. It happened earlier than the original plan for 100 participants because recruitment proved difficult. The responder rate was 14% with ethosuximide and 36% with placebo. The publication excerpt does not provide a confidence interval or P value for that interim comparison, so it cannot establish statistical significance. It does establish the observed direction: ethosuximide performed worse.

The trial's interim result does not show that Cav3.2 is irrelevant to IBS pain. It shows that this CNS-penetrant, nonselective T-type channel blocker did not produce a favorable signal in this population.

Why the result needs restraint, not rescue

Small interim analyses can swing wildly, especially in IBS. The placebo response here was 36%, and placebo responses in IBS trials often run high. With roughly 25 people in each arm, the observed 14% versus 36% split has limited precision. The protocol was amended after the interim analysis, with the total planned sample recalculated to 128 participants. The new assumptions—40% response with ethosuximide and 15% with placebo—nearly reversed the interim pattern, which shows how uncertain those early percentages were.

Yet uncertainty does not turn a negative signal into a positive one. The trial also reported substantial discontinuation related to central nervous system adverse effects, including headache, sleep disruption, nausea, and vomiting. Those events fit ethosuximide's established pharmacology. They also muddy any completers-only analysis: people able to stay on the drug may differ systematically from people who stopped because of side effects.

The accessible publication information does not allow independent assessment of final intention-to-treat efficacy, responder rates among completers, or IBS-subtype analyses for diarrhea-predominant, constipation-predominant, or mixed IBS. Those gaps limit conclusions about whether a narrower patient group responded differently. They do not create evidence that such a group exists.

The target remains unproven in humans

The clean experiment would use a compound that selectively blocks Cav3.2 and stays largely outside the brain. That design could separate two questions that ethosuximide bundles together: whether peripheral Cav3.2 blockade reduces visceral pain, and whether a centrally active antiseizure drug is tolerable enough to test it. No human trial of a peripherally restricted, Cav3.2-selective blocker is identified in the available clinical evidence.

That distinction is more than pharmacology trivia. If a treatment causes enough headache, nausea, or sleep disruption to prompt discontinuation, those effects can erase or distort any modest benefit for IBS pain. The outcome may reflect inadequate gut exposure, poor subtype selectivity, central adverse effects, or a target that does not translate from preclinical models. The ethosuximide trial cannot disentangle those possibilities.

IBS pain has more than one peripheral route

A separate pilot trial points toward another peripheral pathway. In Kanazawa et al.'s 2025 randomized placebo-controlled study, 16 people with IBS received camostat mesilate 200 mg three times daily and 16 received placebo for 14 days. Camostat inhibits serine proteases, and the investigators examined whether it changed fecal bacterial functions related to FKBP-type peptidylprolyl cis-trans isomerases. Increased fecal protease activity has been proposed as a contributor to visceral hypersensitivity.

This was a 32-person pilot, not evidence for routine camostat use in IBS. Its value is narrower: it supports the idea that gut pain may be influenced by several peripheral processes, rather than one master switch.

Patients do have other evidence-based avenues to discuss with a clinician, but none should be treated as universal. A 2025 meta-analysis of 10 randomized trials involving 590 participants found that a low-FODMAP diet changed gut-microbiota measures; dietary treatment should still be individualized, since microbiota changes are not the same as symptom relief. A 2023 probiotics meta-analysis found results remained controversial and strain-dependent. A melatonin meta-analysis reported reduced IBS severity across four randomized trials, while an updated fecal microbiota transplantation review found better quality-of-life scores but no difference in overall symptom improvement or IBS Severity Scoring System results.

The practical verdict

Do not use ethosuximide off-label for IBS pain. The only multicenter blinded trial described here produced an unfavorable interim responder pattern, and adverse effects caused substantial discontinuation. The dose was not a casual supplement experiment; it was a prescription-drug regimen monitored in a clinical trial.

Cav3.2 remains a coherent research target because of the reported mucosal expression and preclinical visceral-hypersensitivity findings. It remains unproven in humans. Until a selective, peripherally restricted compound is tested against placebo with complete outcome reporting, the sensible conclusion is simple: the mechanism deserves better tools, but ethosuximide has not earned a place in IBS self-treatment.