Bezafibrate

Bezafibrate is a decades-old fibrate approved in Europe and Japan for high triglycerides, now being tested at Massachusetts General Hospital in a Phase 2 proof-of-concept trial for bipolar depression [1]. The hypothesis: bipolar disorder involves mitochondrial dysfunction, and a pan-PPAR agonist that improves cellular energy metabolism might help depressive episodes that don't respond to lithium or atypicals [9]. The trial (NCT02481245) is small - 30 patients, investigator-initiated, no commercial sponsor - which tells you exactly what this is: a signal-finding study, not a path to FDA approval. Bezafibrate isn't even approved in the US for its original indication, so the commercial story is essentially nonexistent unless someone repositions the asset. Worth tracking as a window into the mitochondrial-psychiatry hypothesis, not as an investable program.

Approved drug in a new indication, not a novel compound. Bezafibrate (ATC C10AB02, RxNorm 1525) has been marketed since the 1970s as Bezalip and other brands across Europe, Japan, and emerging markets, but was never FDA-approved [2]. The bipolar depression trial is Phase 2, currently recruiting per ClinicalTrials.gov, with a target of 30 participants [1]. No FDA designations - breakthrough, fast track, orphan, RMAT - none apply because this is investigator-initiated work on a non-US-approved drug. Massachusetts General Hospital is the sole sponsor. No expected readout date has been disclosed. The trial was first registered in 2015 and is still actively recruiting in 2026 - an eleven-year window without a readout is itself a meaningful signal about enrollment pace and program priority. Separately, bezafibrate continues to accumulate Phase 2/3 data in primary biliary cholangitis (NCT06443606, n=108, AP-HP-sponsored) and biliary atresia [3,4] - those are the more credible non-cardiovascular paths for this molecule, not psychiatry. Bezafibrate has also been tested in primary mitochondrial disease (MELAS, fatty acid oxidation disorders), with short-term metabolic benefits but mixed long-term safety findings, including evidence of mitochondrial toxicity in some patients [10] - directly relevant to the bipolar mechanistic story.

PPARs (peroxisome proliferator-activated receptors) are nuclear receptors that work like genetic dimmer switches - they sit inside cells, bind fatty-acid-like molecules, and turn on genes that control how the cell burns fat and produces energy [5]. Bezafibrate activates all three subtypes - alpha, gamma, and delta - which is why it's called a pan-PPAR agonist. PPAR-alpha is the dominant driver of its lipid-lowering effect in liver. The bipolar depression rationale leans on a separate biology: people with bipolar disorder show altered brain energy metabolism on imaging, postmortem mitochondrial findings, and there's a recognized overlap between mitochondrial disease syndromes and bipolar-like presentations [9]. The bet is that boosting mitochondrial fatty-acid oxidation via PPAR activation will lift the energetic floor in depressive episodes. Honest read: the mitochondrial-bipolar hypothesis is plausible and has supporting correlative data, but it has never produced a drug. Creatine, CoQ10, and N-acetylcysteine have all been tried under similar mitochondrial/oxidative-stress logic and failed to deliver clean wins in bipolar depression [9]. PPAR-alpha is also primarily a peripheral metabolic regulator - central nervous system effects in humans are largely unproven. A critical mechanistic gap: bezafibrate's CNS penetration is limited. Preclinical work in mitochondrial encephalopathy models shows the drug crosses the blood-brain barrier poorly, with most of the dose staying in the periphery and only modest induction of PGC-1α in cortex and hippocampus [11]. The MGH trial is implicitly betting that either (a) limited CNS exposure is enough to engage neuronal PPAR-driven mitochondrial biogenesis, or (b) peripheral metabolic improvement and reduced systemic inflammation propagate centrally. Either way, this is a second layer of mechanistic uncertainty stacked on top of the unproven PPAR-bipolar hypothesis. The mechanism is biologically reasonable, not biologically validated for this indication.

NCT02481245 is a Phase 2 proof-of-concept trial, randomized placebo-controlled, n=30, with the primary endpoint being change in the Montgomery-Åsberg Depression Rating Scale (MADRS) from baseline to week 8 [1]. Patients have bipolar disorder I or II in a current depressive episode, on stable mood-stabilizer regimens (so bezafibrate is being added on, not tested as monotherapy). Sponsor is MGH; PI work links to the Nierenberg group's history in bipolar pharmacology. The design is reasonable for what it is - placebo-controlled with a validated depression scale is the right setup for a signal-finding study. Concerns: n=30 only detects effects with a Cohen's d around 0.8 or larger (a 'large' effect size - roughly the magnitude of a strong antidepressant response over placebo, far bigger than the d ≈ 0.3 typical of real-world antidepressant trials). Realistic antidepressant signals will be invisible at this sample size. Second concern: the trial has been recruiting since 2015 - eleven years to enroll thirty patients is slow even for academic psychiatry, suggesting funding gaps, recruitment difficulty, or de-prioritization. Third: no biomarker enrichment. Bipolar depression is a clinical syndrome with many biological subtypes, and if mitochondrial dysfunction applies to only a subset, an unselected sample dilutes any real signal.

Our model gives this drug a 9% chance of eventually being approved. It starts from the historical approval rate for Phase 2 drugs in this area-about 24%-then adjusts based on ten facts about the trial and the sponsor. The estimate is pulled up by a non-randomized design and light or open-label blinding, and pulled down by the sponsor's thin approval record and weak earlier-phase results. The remaining factors fell close to average and left the number roughly where it started.

Efficacy risk dominates. The mitochondrial-dysfunction hypothesis in bipolar disorder has correlative support but no successful drug - creatine, CoQ10, and NAC trials failed to produce labeled treatments [9], and PPAR agonists have not shown CNS efficacy in any human psychiatric study. Without a biomarker to enrich for patients whose depression actually involves mitochondrial pathology, a true effect in a 20% subset can disappear in a 30-patient unselected sample. Pharmacokinetic risk: limited blood-brain barrier penetration means even the right pharmacological hypothesis may fail at the exposure stage [11]. Safety risk is comparatively low because bezafibrate has a known profile from cardiovascular use: rare myopathy, transaminase elevation, and creatinine rise. The interaction worth flagging is with lithium - many bipolar patients are on lithium, which itself raises creatinine and affects renal handling; combining them needs careful monitoring. Long-term safety in mitochondrial disease populations has shown evidence of mitochondrial toxicity at therapeutic doses, which is worth tracking if any CNS program advances [10]. Execution risk is the real story: a trial that has been recruiting since 2015 with a 30-patient target is moving at a pace that suggests it may not complete on a useful timeline, or may finish with substantial missing data. Commercial risk is structural - bezafibrate is generic globally and not FDA-approved in the US. Even with a strong Phase 2 signal, no economic actor has clear incentive to fund a Phase 3 and pursue a 505(b)(2) pathway (a regulatory shortcut that lets a sponsor reference existing safety data to gain approval for a new indication) unless a repositioning company picks up the indication.

Noise, not signal - at this stage. The trial is too small, too slow, and too commercially orphaned to move on its own. The data point that would convert this to a signal: a published MADRS readout showing a clean separation from placebo, ideally with secondary biomarkers (lactate, NAA [N-acetylaspartate, a neuronal health marker visible on brain MRI spectroscopy], or peripheral mitochondrial function markers) supporting the mechanistic story. That would justify a properly powered Phase 2b and potentially attract a repositioning sponsor via a 505(b)(2) pathway (a regulatory shortcut that lets a sponsor reference existing safety data for a new indication). Check back when MGH publishes - if they publish - or if a development partner emerges. The more interesting bezafibrate stories right now are hepatology: the AP-HP Phase 3 in PBC (NCT06443606, ongoing) and the recent network meta-analysis ranking bezafibrate and seladelpar at the top of PPAR agonists as second-line options after ursodeoxycholic acid [8]. That's where the molecule's near-term value lives. Competitive context for bipolar depression matters too: lumateperone (Caplyta, approved 2021) and cariprazine (Vraylar, approved 2019) are the recent commercially-backed wins in this indication, and several Phase 2/3 programs in treatment-resistant bipolar depression (ketamine/esketamine derivatives, psilocybin assisted therapy, novel glutamate modulators) are advancing with biomarker-rich designs and real sponsors. For psychiatry watchers, the bigger metabolic-CNS bet to track is GLP-1 receptor agonists in depression and addiction trials, where the data signal is stronger and the commercial machinery is real. Bezafibrate-for-bipolar is a fascinating mechanistic experiment, not a developable asset on current trajectory.