Haematological effects of oral administration of bitopertin, a glycine transport inhibitor, in patients with non‐transfusion‐dependent β‐thalassaemia

Bitopertin is a small molecule selective inhibitor of glycine transporter 1 (GlyT1), initially developed to increase brain extracellular levels of glycine in the vicinity of neuronal N-methyl-D-aspartate receptors for the treatment of schizophrenia. GlyT1, the pharmacological target of bitopertin, is also present as a transmembrane transporter in erythroid cells and accounts for 50–55% of glycine uptake in human red blood cells (RBCs). Erythroid GlyT1 inhibition by bitopertin leads to reduced intracellular glycine availability, interfering with the first step of haem synthesis, in which 5-aminolevulinate synthase catalyses the condensation reaction between glycine and succinyl-coenzyme A, forming 5aminolevulinic acid. b-Thalassaemia is a haemolytic anaemia arising from different genetic alterations of the b-globin gene (haemoglobin subunit beta), resulting in abolished or significantly reduced polypeptide b-globin chain synthesis and a-globin chain accumulation. Membrane damage due to unpaired a-globin chain deposition with an excess of free haem and iron contributes to ineffective erythropoiesis and decreased erythrocyte survival in the circulation. A reduced availability of haem and iron in b-thalassaemia may have beneficial effects in reducing ineffective erythropoiesis and improving RBC survival. In a b-thalassaemia mouse model, oral administration of bitopertin resulted in reduced anaemia and haemolysis, enhanced in vivo survival of erythrocytes, and diminished ineffective erythropoiesis. Markers of cellular damage induced by reactive oxygen species (ROS) were also substantially improved. This report examines the haematological effects of oral bitopertin in patients with non-transfusion-dependent (NTD) b-thalassaemia. A summary of the clinical study protocol and data-sharing statement are provided in the supplementary online material (see Supplementary Material and Methods). Twelve patients were enrolled, five male (41 7%) and seven female (58 3%), with a mean ( standard deviation [SD]) age of 32 8 years ( 9 7). Five patients (41 7%) were Asian and seven (58 3%) were Caucasian. Relevant clinical characteristics for all patients are presented in Table SI. Patients received oral bitopertin 30, 60 and then 90 mg/day during a six-week intrapatient dose escalation phase, followed by ≤10 weeks at the target dose of 90 mg. Figure S1 displays the length of treatment and bitopertin dosing. Six patients (50%) completed the initial 16-week treatment. Reasons for treatment discontinuation were: lack of efficacy (one patient, 8 3%), physician decision (one patient, 8 3%), and study termination by the sponsor (four patients, 33 3%). Individual values for haemoglobin (Hb) during the trial are presented in Fig 1. The first eight patients assessed at an eight-week preliminary efficacy analysis showed a mean ( SD) total Hb reduction of 4 42 g/l ( 6 85 g/l; Table I). Other selected biomarkers of disease activity, presented in Table I, did not show clinically meaningful improvement. The lack of improvement, or the actual decrease, in Hb values at the target dose of 90 mg was attributed to excessive inhibition of haem biosynthesis and/or globin chain production. Consequently, the target dose of bitopertin was decreased from 90 mg to 30 mg daily. Overall, the preliminary efficacy analysis led to treatment discontinuation in three patients and a two-week treatment interruption in four additional patients. Treatment was re-initiated at 30 mg for patients who interrupted treatment, and all newly enrolled patients remained on the starting dose of 30 mg/day throughout treatment (Figure S1). A subsequent analysis from 11 patients at Week 9 showed a total Hb value reduction of 3 4 g/l ( 7 6) compared with baseline (Fig 1). Therefore, the study was terminated and the extension period cancelled. Eight patients (66 7%) had adverse events (AEs) considered related to the study drug, and one patient experienced three serious AEs of Grade 3. The most frequent AEs were dizziness (41 7%), headache (25%) and decreased Hb (25%; Table SII). Despite strong preclinical evidence for a potential therapeutic role of glycine restriction in a mouse model of b-thalassaemia, the present study of oral bitopertin in patients with NTD b-thalassaemia failed to show clinically significant improvements in haematological and chemical biomarkers of disease activity. A therapeutic effect for bitopertin would have required a significant increase in RBC count in conjunction with either no change or a comparatively smaller reduction in mean corpuscular haemoglobin (MCH), as seen in the mouse b-thalassaemia model. However, in most patients, the decrease in MCH was greater in relative magnitude than the increase in RBC count, resulting in diminished total Hb. The lack of therapeutic efficacy for bitopertin in human NTD b-thalassaemia challenges some of the assumptions correspondence