ABBV-932

ABBV-932 (also known as RGH-932) is an oral investigational small molecule originated by Gedeon Richter and partnered with AbbVie, now in Phase 2 for two psychiatric indications: depressive episodes in bipolar I or II disorder [1] and generalized anxiety disorder (GAD) as an add-on to standard antidepressants [2]. The mechanism is disclosed: ABBV-932 is a prodrug of desmethyl-cariprazine (DCAR), a dopamine D3-preferring D3/D2 partial agonist, essentially the same target class as approved cariprazine (Vraylar) [8][9]. A 161-patient Phase 2 in bipolar depression (NCT06605599) [3] MISSED its primary MADRS endpoint in the overall population but showed a positive pre-specified subgroup signal in bipolar I patients, per Gedeon Richter's April 2026 update; the safety profile was generally similar to placebo including extrapyramidal side effect (EPS) rates [10]. The larger 200-patient follow-on (NCT07220460) [1] is a 26-week trial that AbbVie initiated to pursue the bipolar I signal. ABBV-932 sits inside a neuroscience franchise anchored by cariprazine, which generated roughly $3.4B in 2024 for AbbVie [4]; the strategic bet is a next-generation cariprazine-class asset that could extend the franchise into cleaner bipolar I depression and GAD positioning as Humira biosimilar erosion continues.

Novel compound; no marketing approval in any jurisdiction. Active Phase 2 program spans two indications. In bipolar depression, NCT07220460 is recruiting 200 adults with a depressive episode in bipolar I or II disorder; treatment period is 26 weeks followed by a 30-day safety follow-up [1]. In GAD, NCT06846320 is recruiting 315 adults with GAD and inadequate response to prior antidepressant therapy, testing ABBV-932 for 6 weeks as add-on to their existing SSRI or SNRI, with a 4-week safety follow-up [2]. The prior bipolar Phase 2, NCT06605599 (n=161), is completed [3]; primary MADRS (Montgomery-Åsberg Depression Rating Scale, a clinician-rated 0-60 scale of depression severity) analysis was negative in the overall population, positive in the bipolar I subgroup, safety comparable to placebo per Gedeon Richter's disclosure on 2026-04-29 [10]. As of the date of this writeup, ClinicalTrials.gov has not posted a full results section for NCT06605599; the sole public efficacy disclosure remains the Richter partner update. Supporting Phase 1 work includes NCT06300580, a completed dopamine D2/D3 receptor occupancy PET study in healthy volunteers [8], a mass balance study of oral [14C]-labeled ABBV-932 (NCT06953934, n=8; a mass balance study administers a radiolabeled dose to track absorption, distribution, metabolism, and excretion (ADME) of the compound in the body) [5], and a completed relative bioavailability study (NCT06849791). Dosing regimen for the ongoing Phase 2 trials (once vs twice daily) has not been publicly disclosed in ClinicalTrials.gov records. No breakthrough therapy, fast track, orphan drug, or accelerated approval designations have been announced. AbbVie's 2025 10-K [4] and 2026 10-K [7] reference an active neuroscience pipeline but do not disclose quantitative ABBV-932 milestones. Expected pivotal readout: likely 2027 for NCT07220460 based on a 26-week treatment design, 200-patient enrollment, and standard follow-up; AbbVie has not publicly committed to a date.

ABBV-932 is a prodrug of desmethyl-cariprazine (DCAR), the des-methyl metabolite of cariprazine, and pharmacologically acts as a dopamine D3-preferring D3/D2 receptor partial agonist [8][9]. This is the same mechanistic class as approved cariprazine (Vraylar), which has six- to eightfold higher affinity for human D3 over D2 receptors. The completed Phase 1 PET occupancy study NCT06300580 in healthy volunteers directly measured brain D2 and D3 receptor occupancy across oral doses of ABBV-932, confirming the target engagement profile [8]. AdisInsight's public record identifies ABBV-932 explicitly as a 'desmethyl cariprazine prodrug' developed by AbbVie under a partnership with Gedeon Richter (originator code RGH-932) [9]. Mechanistically the rationale is that DCAR delivered via a prodrug could achieve a differentiated exposure/tolerability profile relative to cariprazine itself, potentially reducing peak-related side effects like akathisia while retaining the D3-preferring pharmacology thought to drive antidepressant and anti-anhedonic effects. The Richter April 2026 update supports the tolerability angle: EPS rates were generally similar to placebo, which if replicated would represent a favorable delta versus Vraylar's known EPS liability [10]. The relevance to GAD is less mechanistically clean; D3-preferring partial agonism is not a validated GAD target, so the GAD add-on trial is exploratory. Because the target class is validated by an approved product in the same company, this program should not be scored as first-in-class.

NCT07220460 enrolls 200 adults with a depressive episode in bipolar I or II disorder over a 26-week treatment period followed by 30-day safety follow-up; the primary endpoint listed on ClinicalTrials.gov is number of participants experiencing adverse events, with change in MADRS as a secondary endpoint [1]. This structure is reasonable in context: the prior NCT06605599 already provided the primary MADRS efficacy read (negative overall, positive bipolar I subgroup) [10], and a 26-week safety expansion supports both a chronic-dosing safety database and enables a longer-duration efficacy assessment as a key secondary. Read that way, the AE primary is a regulatory framing choice, not a signal that AbbVie is punting on efficacy; the durability question at 26 weeks matters for a chronic condition like bipolar depression where 8-week Phase 2s poorly predict Phase 3 outcomes. NCT06846320 is a 6-week add-on study (n=315) in GAD patients with inadequate response to protocol-specified antidepressants, plus a 4-week safety follow-up [2]. Add-on designs test whether ABBV-932 adds efficacy over placebo add-on, which is pragmatic for a treatment-resistant population but historically produces small effect sizes that struggle to differentiate commercially. Placebo comparator is standard in both trials. No adaptive design elements, biomarker enrichment (e.g., D3 occupancy stratification), or pharmacogenetic stratification have been publicly disclosed. HAM-A total ≥ 20 is used as GAD eligibility gating [2]. Dosing frequency and dose levels are not disclosed in the ClinicalTrials.gov summaries; the PET occupancy study NCT06300580 would have informed dose selection but has not been publicly reported in detail.

Our model estimates a 18% chance this drug is eventually approved. It starts from the historical base rate for Phase 2 drugs in this area (about 24%), then adjusts using ten facts about the trial and sponsor. What moves the number most: it is helped by a non-randomized design and its light or open-label blinding; it is held back by its few secondary endpoints and weak or limited earlier-phase results. The other facts land near average for this stage, so they leave the estimate roughly where the base rate put it.

Efficacy risk dominates for bipolar depression. The prior NCT06605599 missed the primary MADRS endpoint in the overall population; the positive bipolar I subgroup finding is a pre-specified analysis, not the primary result, and Phase 3 replication of subgroup signals in psychiatry is historically weak [10]. Bipolar depression trials show high placebo response (often over 40%), and 26-week durability adds an attrition/dropout risk not seen in shorter Phase 2 designs. For GAD, competitive risk is severe: approved GAD therapies include the SSRIs escitalopram and sertraline, the SNRIs duloxetine and venlafaxine, buspirone, and pregabalin, plus off-label hydroxyzine; add-on trials in patients with inadequate SSRI/SNRI response face a moving symptomatic baseline and often produce modest effect sizes. Safety risk is now largely characterized by class: D3/D2 partial agonists like cariprazine carry known liabilities including akathisia and other EPS, restlessness, insomnia, weight gain (generally modest for this class), and QT effects to monitor. Richter's April 2026 update reports EPS rates generally similar to placebo, which if replicated would be a genuine differentiator against Vraylar [10]. The mass balance study (NCT06953934) and bioavailability study are routine ADME work; no early-safety flags have been publicly reported from healthy-volunteer Phase 1 work. Commercial risk for bipolar depression is real even on positive efficacy: cariprazine (Vraylar, ~$3.4B for AbbVie in 2024) [4] already occupies AbbVie's premium bipolar depression slot, and quetiapine, lurasidone, and olanzapine-fluoxetine are cheap generics; lumateperone (Caplyta) is approved for bipolar I and II depression and growing. Cobenfy (xanomeline-trospium, formerly KarXT), acquired by Bristol Myers Squibb from Karuna in a ~$14B deal and FDA-approved September 2024 for schizophrenia, is a launched mechanistic comparator opening a muscarinic lane with expected psychiatric label expansion. For ABBV-932 to matter commercially, it needs a differentiated readout on efficacy magnitude in bipolar I, a clear tolerability delta versus cariprazine (EPS/akathisia in particular), or a defined subpopulation that fails current options.

Watch, moderate signal. The single most consequential new fact is the April 2026 Gedeon Richter update [10]: NCT06605599 missed the primary MADRS endpoint in the overall bipolar depression population but showed a positive pre-specified subgroup signal in bipolar I with placebo-like safety including EPS rates. That reframes ABBV-932 from a black-box mechanism play into a next-generation cariprazine-class asset targeting bipolar I depression with a possible tolerability edge. It also explains why NCT07220460 is a 26-week safety-primary expansion in bipolar I/II: AbbVie is de-risking the safety database for a chronic dosing label while running MADRS as a key secondary. For AbbVie as a business, ABBV-932 is a small line item against $56.3B in 2024 revenue [4], but strategically it slots into the neuroscience franchise anchored by Vraylar, extending the same D3-preferring mechanistic class rather than opening a new one. This is not a Humira replacement conversation; it is franchise depth. Practical check-backs going forward from August 2026: (1) ACNP annual meeting in December 2026 for potential detailed presentation of NCT06605599 efficacy/safety data with subgroup breakdowns; (2) posting of ClinicalTrials.gov results section for NCT06605599 (12-month post-completion window applies); (3) AbbVie Q3 and Q4 2026 earnings calls for any R&D update on Phase 3 initiation timing for bipolar I depression; (4) EPS-driven tolerability data from NCT07220460 as a differentiator versus cariprazine. If AbbVie stays silent through end of 2026 with no follow-up disclosure beyond Richter's April 2026 update, that suggests the bipolar I signal is not being aggressively pursued and probability of success should be revised downward.