Agomelatine and Other Antidepressants and the Risk of Acute Liver Injury, a Post-Authorisation Safety Study in 4 European Countries M

Home , Agomelatine

Agomelatine and Other Antidepressants and the Risk of Acute Liver Injury, a Post-Authorisation Safety Study in 4 European Countries M. Pladevall,1 A. Pottegård,2 T. Schink,3 J. Reutfors,4 R. Morros,5 B. Poblador-Plou,6 A. Timmer,7 J. Forns,1 M. Hellfritzsch,2 T. Reinders,3 D. Hägg,4 M. Giner-Soriano,5 A. Prados-Torres,6 M. Cainzos-Achirica,1 J. Hallas,2 B. Kollhorst,3 L. Brandt,4 J. Cortés,5 J. Aguado,1 G. Perlemuter,8,9,10 B. Falissard,10 J. Castellsagué,1 E. Jacquot,11 N. Deltour,11 S. Perez-Gutthann1

1RTI Health Solutions, Spain; 2University of Southern Denmark, Denmark; 3BIPS, Germany; 4Karolinska Institutet, Sweden; 5Universitat Autonoma de Barcelona Bellaterra, Institut Universitari d’Investigació en Atenció Primària Jordi Gol (IDIAP JGol), Spain; 6IACS, Spain; 7Medical Faculty–Carl Von Ossietzky University, Oldenburg, Germany; 8AP-HP, Hôpital Antoine Béclère, France; 9Université Paris-Saclay–Université Paris-Sud, Paris; 10Inserm, France; 11SERVIER, France P860

BACKGROUND RESULTS • Agomelatine is a melatonergic agonist and 5-HT2C antagonist indicated for major depressive episodes Study Population in adults. • There were 3,238,495 new users of study antidepressants (74,440 new users of agomelatine). • Hepatotoxicity, including acute liver injury (ALI), is an identified risk in the risk management plan for • agomelatine, and hepatotoxic reactions have been observed with other antidepressants; however, The cohort attrition process in each data source is presented in Figure 3. population-based studies quantifying this risk are scarce. Figure 3. Cohort Attrition

All users of study EpiChron SIDIAP GePaRD Denmark Sweden antidepressants OBJECTIVE 319,708 361,821 2,326,803 1,127,244 4,222,582

• To evaluate the risk of ALI associated with the use of agomelatine and other selected antidepressant drugs. Eligibility criteria Main cause of ineligibility: prevalent use (patients did not meet the new-user definition)

Exclusion criteria Main exclusion conditions: malignancy and alcohol use disorders METHODS New users 185,628 203,101 817,072 664,205 1,368,489 Study Design and Data Sources included (58.1%) (56.1%) (35.1%) (58.9%) (32.4%) • Multinational, multiple–data source, nested case-control study of new users of agomelatine and other

selected antidepressants Citalopram 9,016 (46.7%) 41,295 (59.4%) 229,895 (36.6%) 199,887 (56.1%) 302,719 (28.9%) • Population-based data sources: EpiChron (Aragon, Spain), SIDIAP (Catalonia, Spain), GePaRD (Germany), and Agomelatine 8,826 (82.5%) 30,155 (43.0%) 18,032 (81.5%) 14,184 (56.7%) Danish and Swedish national registers 3,243 (69.2%) Study Population Main Results • The selection of cases and controls is described in Figure 1. • The main results for agomelatine and the other antidepressants for the primary endpoint are presented in Figure 1. Selection of Cases and Controls Figure 4. A total of 472 cases for the primary endpoint was identified. PPVs for the primary endpoint ranged from 60% to 84%.

New user cohorts for agomelatine, • A total of 178 cases (confirmed by validation) for the secondary endpoint and 17,118 cases for the tertiary Inclusion criteria Exclusion criteria citalopram, and 8 other antidepressants endpoint was identified. Agomelatine results for the secondary (OR, 0.40; CI, 0.05-3.11) and tertiary endpoints (OR, 0.79; CI, 0.50-1.25) were similar to the results of the primary endpoint. PPVs for the tertiary endpoint ranged from 8% to 47%. • For the other study antidepressants that were compared with citalopram, most OR point estimates were also below 1.00 except for the tertiary endpoint, for which paroxetine and venlafaxine showed an increased risk of ALI. No prescription History of liver Up to 20 disease during the prior 12 ALI cases matched • Results of the planned SAs for agomelatine and the other antidepressants were, in general, consistent with months (new users) controls Chronic biliary or the main analysis and produced combined OR point estimates for agomelatine below 1.00 for current use. pancreatic disease Figure 4. Current Use Combined Adjusted Estimates for All Antidepressants Compared With Citalopram ≥ 18 years old Malignancy Matched on index date, (Primary Endpoint Main and Two Sensitivity Post Hoc Analyses Removing Exclusion Conditions)a age, sex, and calendar Continuous Alcoholism, year of start date Drugs for Data sources registration or drug abuse depression included Cases Controls Odds ratio (95% CI) I2 enrolment for at least 111 1,790 Main analyses 12 months prior Heart failure Citalopram 418 7,096 (Reference) No exclusion to the start date 168 2,870 criteria applied HIV/AIDS Internal validation Exclusion Ep, G, Dk 2 112 0.48 (0.13-1.71) 0% (EpiChron, SIDIAP, criteria applied, Agomelatine Ep, Sd, G, Dk, Sw 7 518 0.37 (0.19-0.74) 0% except alcohol Organ transplant and Denmark) and drug abuse Validation Ep, Sd, G, Dk, Sw 5 202 0.47 (0.20-1.07) 0% activities Women not Ep, G, Dk, Sw 9 217 1.07 (0.52-2.21) 0% Ep = EpiChron Sd = SIDIAP being pregnant External validation Fluoxetine Ep, Sd, G, Dk, Sw 41 933 0.63 (0.32-1.25) 36% G = GePaRD (Germany) Ep, Sd, G, Dk, Sw 23 434 0.89 (0.43-1.82) 34% Dk = Denmark Ep, Sd, G, Dk 7 201 1.29 (0.65-2.57) 0% Sw = Sweden Paroxetine Ep, Sd, G, Dk, Sw 39 912 0.70 (0.44-1.12) 0% Deﬁnition of ALI Ep, Sd, G, Dk, Sw 12 370 0.88 (0.41-1.90) 31% Ep, Sd, G, Dk, Sw 41 805 0.88 (0.58-1.34) 0% • Three ALI endpoints were defined (see Figure 2). Sertraline Ep, Sd, G, Dk, Sw 137 3,375 0.65 (0.49-0.85) 0% Figure 2. Deﬁnition of the Three Study Endpoints Ep, Sd, G, Dk, Sw 69 1,361 0.82 (0.59-1.15) 0% Ep, Sd, G, Dk, Sw 25 490 0.80 (0.45-1.40) 0% Escitalopram Ep, Sd, G, Dk, Sw 110 2,180 0.80 (0.58-1.09) 0% Code Description Specific codes Ep, Sd, G, Dk, Sw 44 875 0.82 (0.53-1.26) 0% ICD-9-CM Ep, Sd, G, Dk, Sw 41 967 0.69 (0.45-1.05) 0% 570.x Acute and subacute necrosis of liver Mirtazapine Ep, Sd, G, Dk, Sw 233 4,158 0.85 (0.61-1.20) 32% 572.2 Hepatic coma 573.3 Hepatitis unspecified Ep, Sd, G, Dk, Sw 87 1,643 0.83 (0.60-1.14) 0% Specific Specific ICD-10-CM Ep, Sd, G, Dk, Sw 35 637 0.89 (0.55-1.44) 7% codes codes K71.0 Toxic liver disease with cholestasis Venlafaxine Ep, Sd, G, Dk, Sw 132 2,603 0.78 (0.59-1.03) 0% K71.1 Toxic liver disease with hepatic necrosis K71.2 Toxic liver disease with acute hepatitis Ep, Sd, G, Dk, Sw 66 1,121 0.81 (0.49-1.36) 28% K71.6 Toxic liver disease with hepatitis, not elsewhere classified Ep, G, Dk, Sw 11 322 0.60 (0.31-1.16) 0% K71.9 Toxic liver disease, unspecified Duloxetine Ep, Sd, G, Dk, Sw 62 1,433 0.71 (0.39-1.28) 42% Nonspecific Nonspecific K72.0 Acute and subacute hepatic failure Ep, G, Dk, Sw 31 563 0.80 (0.48-1.33) 2% codes codes K72.9 Hepatic failure, unspecified K75.9 Inflammatory liver disease, unspecified Ep, Sd, G, Dk, Sw 22 467 0.55 (0.22-1.41) 54% K76.2 Central hemorrhagic necrosis of liver Amitriptyline Ep, Sd, G, Dk, Sw 95 1,847 0.83 (0.50-1.38) 46% Nonspecific codes Ep, Sd, G, Dk, Sw 30 673 0.71 (0.28-1.76) 68% Hospital Outpatient ICD-9-CM 573.8 Other specified disorders of liver 0.0 0.5 1.0 1.5 2.02.5 3.0 573.9 Unspecified disorders of liver Primary endpoint Favours other study Favours comparator 782.4 Jaundice, unspecified, not of newborn drugs for depression (citalopram) Secondary endpoint Expected V42.7 Liver transplant Tertiary endpoint PPV 790.4 Nonspecific elevation of transaminase or lactic acid dehydrogenase a OR estimates were adjusted for confounding factors. The list of confounders differed by data source and type of analysis (main vs. sensitive 789.1 Hepatomegaly analyses). The following confounders were included in most analyses: obesity, hyperlipidaemia and hypertriglyceridaemia, diabetes, ICD-10-CM Secondary and tertiary endpoints hypertension, indication of treatment with drugs for depression for major depression, indication of treatment with drugs for depression for K76.8 Other specified diseases of liver were validated in Denmark and Spain. K76.9 Liver disease, unspecified anxiety disorders, indication of treatment with drugs for depression for other mental and behavioural disorders, Charlson Comorbidity Index, R17 Unspecified jaundice, excludes neonatal number of liver tests performed, concurrent use of hepatotoxic drugs, and concurrent use of other drugs for depression. In Germany, PPVs of the primary and R16.0 Hepatomegaly, not elsewhere classified tertiary endpoints were estimated R16.2 Hepatomegaly with splenomegaly, not elsewhere classified (external validation). R74.0 Nonspecific elevation of transaminase and lactic acid dehydrogenase Z94.4 Liver transplant DISCUSSION

PPV = positive predictive value. Strengths Limitations • Large, multinational, and multiple–data source study • Low ALI incidence and low precision in the main Statistical Analyses including nine different antidepressants compared analyses of risk estimates and PPVs • Crude and adjusted odds ratios (ORs) and 95% confidence intervals (CIs) of ALI for current use of each study with citalopram • Despite the pharmacoepidemiological methods antidepressant compared with current use of citalopram were estimated via conditional logistic regression • Validation activities used to minimize its presence, bias as a result of models that used a prespecified list of confounders, and other potential confounders were added after a • Three different endpoints with different incidences potential misclassification of exposures or backward selection process. of ALI and different PPVs endpoints, and of residual confounding, is still possible. • Meta-analytic methods were employed to obtain the pooled adjusted ORs (random models were used when heterogeneity was present, i.e., I2 ≥ 30%). • Several preplanned sensitivity analyses (SAs) were done to check the robustness of results, and 2 post hoc CONCLUSIONS SAs (requested by the European Medicines Agency), one applying no liver-related exclusion criteria and one • The results of this study do not suggest that the risk of ALI with the use of agomelatine constitutes a public with all exclusion criteria but alcohol and drug abuse, were implemented to check the robustness of the results. health problem, at least among patient populations in health care systems with prescription patterns and risk-minimisation measures similar to those in this study. • When compared with citalopram, most antidepressants evaluated had OR point estimates for ALI below 1.00. CONFLICTS OF INTEREST However, specific studies to investigate this potential association of citalopram with ALI would be needed. The study was funded by Servier under a contract with RTI Health Solutions that grants the research team independent publication rights. MP, JF, MCA, JA, JC, and SPG are employees of RTI Health Solutions. AP, TS, JR, RM, BPP, AT, MH, TR, DH, MGS, APT, JH, BK, LB, JC, BF, and GP worked on other projects funded by pharmaceutical companies in their CONTACT INFORMATION institutions that were not related to this study and were without personal profit. GP reports participation in research projects Emmanuelle Jacquot, MD 50, rue Carnot funded by pharmaceutical companies. GP also received royalties and reports travel and participation in meetings funded Department of Pharmacoepidemiology 92284 Suresnes Cedex by pharmaceutical companies. ND and EJ are employees of Servier. Institut de Recherches Internationales Servier Tel: +33 1 55 72 40 57 E-mail: [email protected]

The power of knowledge. The value of understanding. Presented at the 31st Congress of the European College of Neuropsychopharmacology; October 6-9, 2018; Barcelona, Spain