Long-Term Real-World Effectiveness of Pharmacotherapies for Schizoaffective Disorder

Jonne Lintunen; Heidi Taipale; Antti Tanskanen; Ellenor Mittendorfer-Rutz; Jari Tiihonen; Markku Lähteenvuo

Disclosures

Schizophr Bull. 2021;47(4):1099-1107. 

In This Article

Results

In both cohorts, approximately 60% of the patients with SCHAFF were females (60.2% in the Finnish and 60.5% in the Swedish cohort) (supplementary table 3). Mean age was also of similar range, 46.7 years (SD 14.6) in the Finnish cohort and 45.1 years (SD 12.1) in the Swedish cohort. During the follow-up, 13.1% of the Finnish cohort and 22.4% of the Swedish cohort were censored due to change of diagnosis into schizophrenia.

Median follow-up time was longer in the Finnish cohort (11.2 years, IQR 5.6–11.5) than in the Swedish cohort (7.6 years, IQR 3.8–10.3). During the follow-up, 50.5% of the Finnish and 46.2% of the Swedish cohort had psychosis hospitalization, and corresponding figures for any psychiatric hospitalization were 57.9% and 54.7%, respectively. In both cohorts, 93%–94% used antipsychotics during the follow-up and FG-use was generally more common in Sweden, whereas clozapine and antipsychotics polytherapy were more common in Finland (supplementary table 3). Concurrent antidepressant and antipsychotic use was more frequent in the Swedish cohort (56% of the Swedish vs 49% of the Finnish cohort), whereas use of mood stabilizers was more common in the Finnish cohort (41% of the Swedish, 47% of the Finnish cohort). BZDRs were the most common adjunctive pharmacotherapy in both countries, used by 72% of Swedish and 61% of the Finnish cohort.

Clozapine, LAIs, and antipsychotic polytherapy were consistently associated with a decreased risk of psychosis hospitalization in both cohorts (Figure 1). Quetiapine was not associated with a decreased risk as compared with nonuse of antipsychotics in either country. Exposure to aripiprazole LAI or levomepromazine was associated with a decreased risk in the Finnish cohort, but not in the Swedish cohort.

Figure 1.

Relative hazard of psychosis hospitalization (hazard ratio [HR], with 95% confidence interval [CI]) associated with specific antipsychotics in schizoaffective disorder (SCHAFF) compared with no use of antipsychotics, within-individual comparisons. In bold are depicted agents that are significant after Benjamini-Hochberg correction for 5% false discovery rate.

When all persons who were diagnosed with schizophrenia during the follow-up (ie, diagnosis was changed) were removed from the analyses, the results remained similar (supplementary figures 1 and 2). Between-individual analyses had a rather similar rank order for antipsychotics as the main analyses (supplementary figures 3 and 4). Clozapine and LAIs were also associated with the lowest risk of treatment failure (Figure 2). Of all antipsychotics, only use of levomepromazine was associated with an increased risk of treatment failure in the Swedish cohort.

Figure 2.

Relative hazard of treatment failure (hazard ratio [HR], with 95% confidence interval [CI]) associated with specific antipsychotics in schizoaffective disorder (SCHAFF) compared with no use of antipsychotics, within-individual comparisons. In bold are depicted agents that are significant after Benjamini-Hochberg correction for 5% false discovery rate. Treatment failure consists of psychiatric hospitalizations, any changes in antipsychotic medication (switch, addition, discontinuation), and death due to any cause.

Use of mood stabilizers in combination with antipsychotics was associated with a 24% (HR 0.76, 95% CI 0.71–0.81, Finnish cohort) and 16% (HR 0.84, 0.78–0.90, Swedish cohort) decreased risk of psychosis hospitalization when compared with antipsychotic only use (P < .0001). When compared with the most common antipsychotic monotherapies, combining mood stabilizer was beneficial for clozapine (HR 0.73, 0.62–0.86), olanzapine (HR 0.83, CI 0.71–0.98), quetiapine (HR 0.71, 0.61–0.84) and SG-LAI (HR 0.69, 0.52–0.91) in the Finnish cohort (Figure 3). The results in the Swedish cohort were mainly in line with the Finnish results, but CIs were wider and the results more often nonsignificant.

Figure 3.

Relative hazard of psychosis hospitalization (hazard ratio [HR], with 95% confidence interval [CI]) associated with combinations of specific antipsychotics and any mood stabilizer (MS) compared with specific antipsychotic monotherapies, within-individual comparisons. The agents that are significant after Benjamini-Hochberg correction for 5% false discovery rate are indicated with an asterisk (*).

Antidepressant use was associated with a 10% decreased risk of psychosis hospitalization (HR 0.90, 0.83–0.97) in the Swedish cohort, but not in the Finnish cohort (1.00, 0.94–1.07). Of the specific commonly used antipsychotics, only quetiapine combined with an antidepressant was associated with a decreased risk in both cohorts compared with specific antipsychotic monotherapies (HR 0.75, 0.60–0.94 in the Swedish cohort, HR 0.75, 0.63–0.90 in the Finnish cohort) (Figure 4).

Figure 4.

Relative hazard of psychosis hospitalization (hazard ratio [HR], with 95% confidence interval [CI]) associated with combinations of specific antipsychotics and antidepressants compared with specific antipsychotic monotherapies, within-individual comparisons. The agents that are significant after Benjamini-Hochberg correction for 5% false discovery rate are indicated with an asterisk (*).

BZDR use was associated with a 7% (HR 1.07, 1.01–1.14, Finnish cohort) and 21% (1.21, 1.13–1.30, Swedish cohort) increased risk of psychosis hospitalization. The results did not change when the first 30 days of use were censored from the analyses (HR 1.07, 1.01–1.15 in the Finnish cohort and HR 1.21, 1.12–1.30 in the Swedish cohort). Regarding the most common antipsychotics used, SG-LAI (HR 1.59, 1.22–2.09) and clozapine (HR 1.19, 1.01–1.41) combined with BZDRs were associated with a higher risk of psychosis hospitalization in the Finnish cohort when compared with specific antipsychotics without concomitant BZDRs (Figure 5). For the Swedish cohort, no specific antipsychotic and BZDR combination was associated with an altered risk.

Figure 5.

Relative hazard of psychosis hospitalization (hazard ratio [HR], with 95% confidence interval [CI]) associated with combinations of specific antipsychotics and benzodiazepines and related drugs (BZDR) compared with specific antipsychotic monotherapies, within-individual comparisons. The agents that are significant after Benjamini-Hochberg correction for 5% false discovery rate are indicated with an asterisk (*).

The results remained similar when the outcome was defined as any psychiatric hospitalization and not just hospitalization due to psychosis (supplementary table 4). The results remained similar for adjunctive pharmacotherapies: combining mood stabilizers with antipsychotics was associated with a decreased risk (HR 0.90, 0.85–0.95 in the Finnish cohort and HR 0.78, 0.72–0.86 in the Swedish cohort), combining antidepressants had varying results (HR 1.03, 0.97–1.09 in the Finnish cohort and HR 0.91, 0.83–1.00 in the Swedish cohort) and use of BZDRs was associated with an increased risk (HR 1.15, 1.09–1.21 in the Finnish cohort and HR 1.21, 1.11–1.32 in the Swedish cohort).

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....