Memantine Augmentation in Treatment-Resistant OCD
Antipsychotic augmentation already failed her — memantine's distinct glutamatergic mechanism has real trial support in treatment-resistant OCD, but sits far behind antipsychotic augmentation in how often it's actually reached for.
H.M., a 44-year-old woman, has owned and run a small neighborhood restaurant for the past nine years, a business she describes as "the one thing I built entirely myself" after an earlier career change. She has well-controlled gastroesophageal reflux on omeprazole and no other significant medical history.
Her OCD, centered on symmetry and a need for exact repetition in routine tasks, has required her to hire additional staff specifically to handle closing procedures she can no longer complete herself within a reasonable time. She has completed two adequate SSRI trials at maximal doses, followed six months ago by a twelve-week risperidone augmentation trial added to her current sertraline, which produced no meaningful change in her Y-BOCS, currently 29. She tolerated the risperidone without significant side effects but saw no benefit from it.
With antipsychotic augmentation — the strategy with the most trial support of any augmentation option — already tried and genuinely failed, the group is looking at options that see meaningfully less use in practice despite having real evidence behind them. Memantine, an NMDA receptor antagonist better known for its role in Alzheimer's disease, works through glutamatergic modulation rather than anything serotonergic or dopaminergic, giving it a mechanism entirely distinct from both her SSRI and her failed antipsychotic trial. Several small randomized trials have shown a genuine augmentation benefit when added to an existing SSRI in treatment-resistant OCD, with a favorable side-effect profile relative to antipsychotic augmentation — no metabolic monitoring burden, no extrapyramidal risk. It remains underrepresented in real-world OCD treatment algorithms relative to its actual evidence, largely because most clinicians' working mental model of augmentation defaults to antipsychotics first, without memantine reliably appearing as the next name considered.
An underused option with real evidence behind it
She's an ideal candidate to look past the default augmentation pathway. Memantine's evidence in treatment-resistant OCD is real, from randomized trials, not just case reports, and its mechanism through NMDA receptor modulation is genuinely distinct from both her SSRI and the antipsychotic that already failed her. I'd add it to her current sertraline rather than switching anything.
I agree, and I want to name honestly why this isn't where most clinicians would land first — it's underused relative to its evidence largely because most working algorithms simply don't list it as a next name after antipsychotic augmentation fails, not because of some genuine safety concern that argues against it. Her tolerability history with both prior medications is reassuring for starting a new agent generally.
Practically, memantine also avoids everything that made the antipsychotic trial a real burden to monitor — no metabolic labs, no extrapyramidal symptom checks. I'd start at a standard titrated dose and plan a twelve-week reassessment, the same trial length we've used at every step so far, so the comparison across her treatment history stays apples-to-apples.
Agreed: memantine started today as an augmentation agent added to her unchanged sertraline dose, with a twelve-week Y-BOCS reassessment matching the trial length used at every prior step of her treatment history.
All three voices reached this conclusion together without meaningful disagreement — the discussion was less about weighing competing options than about surfacing an evidence-backed choice that isn't reliably reached for by default.