Frequently Asked Questions

Yes. In this post hoc analysis, lurasidone was associated with significant improvement in both psychic and somatic anxiety symptoms over 6 weeks in adults with bipolar I depression. For psychic anxiety, least squares mean change was -4.82 b1 0.37 with lurasidone versus -2.97 b1 0.28 with placebo in monotherapy (P < .001; 95% CI for between-group difference, -2.69 to -1.00) and -5.56 b1 0.41 versus -4.26 b1 0.42 in adjunctive therapy with lithium or valproate (P = .009; 95% CI, -2.27 to -0.32). For somatic anxiety, change was -1.89 b1 0.17 versus -1.37 b1 0.23 in monotherapy (P = .048; 95% CI, -1.04 to -0.01) and -2.22 b1 0.22 versus -1.47 b1 0.23 in adjunctive therapy (P = .006; 95% CI, -1.29 to -0.23).

Lurasidone was associated with improvement in both major anxiety domains measured on the Hamilton Anxiety Scale: psychic anxiety and somatic anxiety. In this analysis, psychic anxiety referred to mental agitation and psychological distress measured by HAM-A items 136 and 14, while somatic anxiety referred to physical anxiety symptoms measured by HAM-A items 7313. The findings suggest benefit for both psychological and physical anxiety symptoms during treatment of bipolar I depression.

Yes, for lurasidone monotherapy. Baseline "decrease in sleep" on YMRS item 4 significantly moderated anxiety improvement with lurasidone monotherapy versus placebo. For overall anxiety, the treatment-by-sleep interaction was significant (P = .010, F2, 428 = 4.69), and when decreased sleep was present at baseline, effect sizes were Cohen d = 0.47 for pooled monotherapy doses, 0.55 for 20360 mg/d, and 0.38 for 803120 mg/d.

The same pattern was seen for psychic anxiety (interaction P = .018, F2, 428 = 4.03; Cohen d = 0.50 pooled, 0.58 for 20360 mg/d, 0.42 for 803120 mg/d) and somatic anxiety (interaction P = .032, F2, 428 = 3.47; effect size 0.30 pooled, 0.35 for 20360 mg/d, 0.24 for 803120 mg/d). This moderating effect was not significant in adjunctive therapy with lithium or valproate.

Anxiety symptoms and sleep disturbance were very common at study entry. All patients with available baseline data (100%, n = 824) had at least 1 psychic anxiety symptom, and 88.5% (n = 729) had at least 1 somatic anxiety symptom. The prevalence of baseline decrease in sleep on YMRS item 4 was 72.5% (597/824), and 72.1% of the pooled sample had sleep disturbance across HAM-A, MADRS, and YMRS sleep items, while only 5.7% had none of these sleep-related symptoms.

Yes. In the monotherapy study, lurasidone significantly improved MADRS total score versus placebo in both lower- and higher-anxiety groups defined by baseline HAM-A score. In patients with lower baseline anxiety (HAM-A total < 14, n = 223), the effect size for MADRS improvement was 0.53 (P = .001; 95% CI for between-group difference, -8.15 to -2.07; least squares mean change -15.19 with lurasidone vs -10.08 with placebo). In patients with higher baseline anxiety (n = 262), the effect size was 0.39 (P = .005; 95% CI, -6.39 to -1.16; least squares mean change -15.28 vs -11.51).

Yes. Changes from baseline in HAM-A total score, as well as psychic and somatic anxiety subcomponents, were significantly associated with changes in Sheehan Disability Scale scores during lurasidone monotherapy and adjunctive treatment (all P < .001). In monotherapy, the mediation association between change in HAM-A and change in functioning remained significant after accounting for improvement in depressive symptoms (P = .044, F1,395 = 4.08), whereas this was not significant for adjunctive therapy (P = .519).

Yes. Lurasidone monotherapy or adjunctive therapy with lithium or valproate was associated with significantly greater improvement than placebo in multiple sleep-related measures. This included reduced sleep on MADRS item 4 (P = .005 for pooled monotherapy dose groups vs placebo; P = .015 for adjunctive therapy), insomnia on HAM-A item 4 (P = .006 for pooled monotherapy dose groups; P = .016 for adjunctive therapy), and decrease in sleep on YMRS item 4 (P = .002 for lurasidone 20360 mg/d vs placebo; P = .043 for adjunctive therapy vs placebo), although the 803120 mg/d monotherapy group did not significantly separate from placebo on YMRS item 4 (P = .214).

This was a post hoc analysis of 2 randomized, placebo-controlled, 6-week trials in adults with bipolar I depression. One trial tested lurasidone monotherapy at 20360 mg/d and 803120 mg/d versus placebo, and the other tested flexibly dosed lurasidone 203120 mg/d added to lithium or valproate versus placebo added to lithium or valproate.

Because the analysis was post hoc, the parent trials were not originally designed specifically to test how anxiety symptoms or sleep disturbance moderated or mediated treatment response. The results are therefore exploratory and useful for hypothesis generation, but they should be interpreted more cautiously than prespecified primary outcomes.

• This was a post hoc analysis of trials not specifically designed to study anxiety symptoms or sleep disturbance.
• Monotherapy and adjunctive therapy may differ in response, and sleep disturbance moderated response in monotherapy but not adjunctive treatment.
• The adjunctive study may have lacked power to detect moderation by sleep disturbance because placebo-corrected anxiety effect sizes were small.
• Sleep disturbance was not measured with a dedicated sleep questionnaire or polysomnography.
• Psychic and somatic anxiety were both assessed using the HAM-A rather than separate domain-specific instruments.
• No multiplicity adjustments were performed for these exploratory analyses.
• The authors note that improvement in anxiety symptoms may partly reflect improvement in the underlying depressive syndrome.