QTc changes and early major adverse cardiovascular events associated with antidepressant treatment for major depressive disorder in adults: individual participant data network meta-regression of double blind randomised trials.

OBJECTIVE: To compare the effect of specific antidepressants with placebo on heart rate corrected QT interval (QTc) during the first eight weeks of treatment for major depressive disorder in adults, and to investigate whether antidepressant use is associated with early major cardiovascular adverse events (MACE). DESIGN: Individual participant data (IPD) network meta-regression and aggregate pairwise meta-analysis. DATA SOURCES: Individual and aggregate level data from randomised controlled trials identified through Cochrane Central Register of Controlled Trials, CINAHL, Embase, LILACS, Medline, Medline In-Process, PsycINFO, the websites of regulatory agencies, and international registers. METHODS: IPD were harmonised to enhance comparability and analysed using network meta-regressions to explore the interplay with baseline QTc, age, sex assigned at birth, body mass index, serum potassium level, and estimated glomerular filtration rate. The primary outcome was changes in QTc Fridericia (QTcF) associated with specific antidepressants and placebo after taking into account modifiable and non-modifiable risk factors. Aggregate data on early MACE were analysed using pairwise meta-analysis comparing antidepressants with placebo. ELIGIBILITY CRITERIA FOR SELECTING STUDIES: Double blind randomised trials of adults (≥18 years) with major depressive disorder diagnosed according to standardised, operationalised criteria. RESULTS: Of 130 unique randomised controlled trials identified, 35 (27%) included individual level data (8679 participants) on QTc with placebo and 10 antidepressants (amitriptyline, bupropion, duloxetine, escitalopram, fluoxetine, mirtazapine, paroxetine, trazodone, venlafaxine, vortioxetine). Compared with placebo, escitalopram and amitriptyline were associated with increased QTc intervals (8.7 milliseconds (ms), 95% credibility interval (CrI) 1.6 to 15.8, and 5.3 ms, 0.8 to 9.8, respectively). When risk factors were considered, amitriptyline and escitalopram were also associated with the highest QTc prolongations (58.8% and 21.3%, respectively), whereas venlafaxine and vortioxetine were often associated with the lowest QTc intervals (24.1% and 9.7%, respectively). Important variability of effects was observed, especially for escitalopram, fluoxetine, and mirtazapine. An aggregate analysis of 139 trials (52 398 participants) did not identify an association between antidepressant use and increased rates of early MACE and non-suicidal death compared with placebo (risk difference 0.01%, 95% CrI -0.01% to 0.02%). CONCLUSIONS: The effect of specific antidepressants on QTc interval differed between adults with depression. No evidence associated antidepressant use with early MACE and non-suicidal sudden death. To support shared decision making and taking into account the trade-off between benefits and harms, individual modifiable and non-modifiable risk factors should be considered when choosing antidepressants. SYSTEMATIC REVIEW REGISTRATION: Open Science Framework https://osf.io/c8k65/.