Real-world effectiveness of clozapine in patients with bipolar disorder: results from a 2-year mirror-image study

Clozapine remains the drug of choice for treatment-resistant schizophrenia but the evidence for its use in severe bipolar disorder (BD) remains sparse.

A pharmaco-epidemiologic database study was carried out in Denmark, investigating the effectiveness of clozapine in BD patients (without a schizophrenia-spectrum disorder), between 1996 and 2007, using a two-year mirror-image design.

A total of 21473 patients with a lifetime diagnosis of International Classification of Diseases-10 (ICD-10) BD were identified, of which only 326 (1.5%) were treated with clozapine and were included in the mirror-image analysis. The mean follow-up time was 544 ± 280 days, the mean clozapine dose was 307.4 mg [95% confidence interval (CI): 287.9-328.2], and 39.3% were male. During clozapine treatment, the mean number of bed-days decreased from 177.8 (95% CI: 149.4-211.6) to 34.6 (95% CI: 24.8-48.2) (p < 0.001). The mean number of admissions was reduced from 3.2 (95% CI: 2.9-3.7) to 2.0 (95% CI: 1.6-2.4) (p < 0.001). Overall, 240 patients (73.6%) had reduced bed-days and 130 (39.9%) were not admitted while treated with clozapine. Moreover, the number of psychotropic co-medications was reduced from 4.5 defined daily doses (DDD) (25-75 percentiles: 2.4-8.2) to 3.9 DDD (25-75 percentiles: 2.4-6.1) (p = 0.045). Somatic hospital visits for intentional self-harm/overdose reduced significantly from 8.3% to 3.1% (p = 0.004). However, non-psychotropic co-medication use for medical conditions did not increase; 0.7 DDD (25-75 percentiles: 0.0-2.9) to 0.8 DDD (25-75 percentiles: 0.1-2.89) (p = 0.3).

Clozapine use for BD was associated with a significant and clinically relevant reduction in the number of bed-days, psychiatric admissions, psychotropic co-medications, and hospital contact for self-harm/overdose, without increased medical treatments. Clozapine seems to be an appropriate choice for treatment-resistant BD and should be investigated in randomized controlled trials.