Therapeutic drug monitoring of LAI antipsychotics as a predictor of clinical relapse: a one-year follow-up

Abstract

IntroductionClinical relapses in schizophrenia remain a frequent event. Long-acting injectable (LAI) antipsychotics enhance adherence, but low blood levels can sometimes be observed despite an adequate posology. Nonetheless, the evaluation of this parameter is uncommon in clinical practice.ObjectivesTo explore the potential advantages of therapeutic drug monitoring (TDM) of LAIs as a predictor of relapse in clinically stable outpatients with schizophrenia.Methods44 individuals who had reached the pharmacokinetic steady state of LAI treatment (paliperidone, olanzapine, aripiprazole) underwent an anamnestic and psychopathological assessment. LAI blood levels were measured using liquid chromatography-mass spectrometry and classified as “in range” or “under range” according to the Arbeitsgemeinschaft für Neuropsychopharmakologie und Pharmakopsychiatrie (AGNP) guideline values. Individuals who relapsed during the one-year follow-up were compared to non-relapsers (Fisher’s exact test, χ2 or Mann-Whitney U). An exploratory binary logistic regression tested the role of other possible relevant predictors of relapse.ResultsNo differences were observed in baseline use of mood stabilisers (p=0.211), antidepressants (p=0.530), or prescribed LAI (p=0.563). Other comparisons are presented in the table: among these variables, in-range LAI levels were the only significant predictor of relapse (F=5.95, p=0.015; OR 0.04, 95%CI 0.02-0.56).Relapse (n=6)No relapse (n=38)pAge (years)41.33±10.7843.95±12.980.667Male4 (66.7%)21 (55.3%)0.600Illness duration (years)21.83±2.6419.13±11.820.289Previous acute episodes3.50±1.053.29±1.470.652PANSS-total49.33±14.8342.74±14.140.231In-range LAI2 (33.3%)32 (84.2%)0.006ConclusionsTDM of LAIs may optimise the clinical management of schizophrenia by highlighting a suboptimal dosage and a consequent higher relapse risk. Large-scale, drug-specific assessments are needed to confirm these findings.DisclosureNo significant relationships.