Glycine transporter inhibitors
Medicines designed to strengthen the NMDA receptor by raising glycine. Large trials found no benefit, and their development for schizophrenia stopped.
At a glance
- Type
- Medicine: glycine transporter 1 (GlyT1) inhibitors
- Used for
- Tested for negative symptoms (bitopertin) and for problems with memory and attention (iclepertin)
- Status
- Development stopped, after phase 3 trials found no benefit
- Administration
- Tablets, once a day, added to an antipsychotic
Clinical summaryFor health professionals
- Bitopertin, SearchLyte (Bugarski-Kirola 2016): 3 phase 3 trials, 1,794 patients with suboptimally controlled symptoms, 12 weeks; 1 of 6 active arms beat placebo.
- Bitopertin, FlashLyte and DayLyte (Bugarski-Kirola 2017): about 1,200 patients with predominant negative symptoms, 24 weeks; no separation from placebo.
- Iclepertin, CONNEX (Keefe 2025): 3 phase 3 trials, 1,835 patients, 338 centres, 41 countries, 10 mg a day for 26 weeks; no difference in MCCB cognition score.
- Iclepertin lowered haemoglobin slightly and reversibly, a class effect (Nagy 2026).
What they are
The NMDA receptor, which is thought to work too weakly in schizophrenia, needs glycine bound alongside glutamate to work. Glycine transporter inhibitors block the pump that clears glycine away, so more of it stays near the receptor.1 Two of them, bitopertin and iclepertin, reached the largest trials ever run in schizophrenia, and both failed. The reasoning behind them is explained in Beyond dopamine (opens in a new tab).
How well they work
Bitopertin, from Roche, was promising in a first trial of 323 people with mainly negative symptoms.2 Three phase 3 trials then added it to antipsychotics in 1,794 people whose symptoms were not well controlled, and only one of six doses tested did better than placebo.3 Two further trials, in about 1,200 people with mainly negative symptoms, found no benefit at 24 weeks.4 Bitopertin is now being developed for erythropoietic protoporphyria, a rare inherited disorder.5
Iclepertin, from Boehringer Ingelheim, was tested for problems with memory and attention, for which no medicine is approved. Three phase 3 trials included 1,835 people at 338 hospitals in 41 countries. After 26 weeks, thinking skills did not differ from placebo.6 The company stopped the long-term extension trial in January 2025.7
Side effects
Both were generally well tolerated.4,6 Iclepertin lowered haemoglobin slightly, and the level returned to normal.8
Alternatives
Other approaches to the NMDA receptor include D-serine and sarcosine (opens in a new tab), sodium benzoate (opens in a new tab) and evenamide (opens in a new tab), which acts on glutamate release instead.
What we don’t know yet
- Why raising glycine did not help, when so much evidence points to the NMDA receptor.
