Martin Schaefer, MD; Rahul Sarkar, MD; Viola Knop, MD; Susanne Effenberger, MSc; Astrid Friebe, MD; Loni Heinze, MD; Ulrich Spengler, MD; Thomas Schlaepfer, MD; Jens Reimer, MD; Peter Buggisch, MD; Johann Ockenga, MD; Ralph Link, MD; Michael Rentrop, MD; Hans Weidenbach, MD; Gwendolyn Fromm, MD; Klaus Lieb, MD; Thomas F. Baumert, MD; Andreas Heinz, MD; Thomas Discher, MD; Konrad Neumann, PhD; Stefan Zeuzem, MD; Thomas Berg, MD
Acknowledgment: The authors thank U. Meyer, C. Mayr, B. Möller, B. Hintsche, R. Heyne, C. John, A. Jessen (Berlin); H. Schmidt (Münster); B. Ruf (Leipzig); M. Leuschner, S. Planz-Kuhlendahl (Offenbach); H. Jablonowsky (Salzgitter); C. Weber, W. Tacke (Königsstein); C. Gelbmann (Regensburg); C. Folwaczny, R. Zachoval, E. Schulte-Frohlinde, F. Schmidt (München); and J. Pausch (Kassel) for their help with recruiting patients.
Grant Support: The trial was supported in part by Roche Pharma, Grenzach-Whylen, Germany; Lundbeck, Hamburg, Germany (placebo medication and financial support); and the German Network of Competence on Viral Hepatitis.
Potential Conflicts of Interest: Disclosures can be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M11-2214.
Reproducible Research Statement:Study protocol: Available from Dr. Schaefer (e-mail, email@example.com). Statistical code and data set: Not available.
Requests for Single Reprints: Martin Schaefer, MD, Department of Psychiatry, Psychotherapy and Addiction Medicine, Kliniken Essen-Mitte, Henricistr. 92, 45136 Essen, Germany; e-mail, firstname.lastname@example.org.
Current Author Addresses: Dr. Schaefer: Department of Psychiatry, Psychotherapy and Addiction Medicine, Kliniken Essen-Mitte, Henricistr. 92, 45136 Essen, Germany.
Dr. Sarkar: Department of Psychiatry and Psychotherapy, Asklepios Westklinikum Hamburg, Suurheid 20, 22559 Hamburg, Germany.
Drs. Knop and Zeuzem: Department of Internal Medicine I, Klinikum der J. W. Goethe-Universität, Theodor-Stern-Kai 7, 60590 Frankfurt am Main, Germany.
Ms. Effenberger and Drs. Heinze and Heinz: Department of Psychiatry and Psychotherapy, Charité–Universitätsmedizin Berlin, Campus Charité Mitte, 10117 Berlin, Germany.
Dr. Friebe: Department of Psychiatry and Psychotherapy, Universitätsklinik Bochum Alexandrinenstraße 1, 44791 Bochum, Germany.
Dr. Spengler: Department of Internal Medicine, Universitätsklinikum Bonn, Sigmund-Freud-Straße 25, 53127 Bonn, Germany.
Dr. Schlaepfer: Department of Psychiatry and Psychotherapy, Universitätsklinikum Bonn, Sigmund-Freud-Straße 25, 53127 Bonn, Germany.
Dr. Reimer: Department of Psychiatry and Psychotherapy, Interdisziplinäre Suchtforschung der Universität Hamburg, Martinistraße 52, 20246 Hamburg, Germany.
Dr. Buggisch: Institute of Interdisciplinary Medicine, Asklepiosklinikum St. Georg, Lohmühlenstraße 5, 20099 Hamburg, Germany.
Dr. Ockenga: Department of Internal Medicine, Klinikum Bremen-Mitte, St.-Jürgen-Straße 1, 28205 Bremen, Germany.
Dr. Link: Department of Internal Medicine, St. Josefsklinik, Weingartenstr. 70, 77654 Offenburg, Germany.
Dr. Rentrop: Department of Psychiatry and Psychotherapy, Klinikum rechts der Isar Ismaningerstraße 22, 81675 München, Germany.
Dr. Weidenbach: Department of Internal Medicine, Klinikum Mittelbaden, Balger Straße 50, 76532 Baden-Baden, Germany.
Dr. Fromm: Department of Psychiatry and Psychotherapy, Justus Liebig Universität Gießen, Am Steg 22, 35385 Giessen, Germany.
Dr. Lieb: Department of Psychiatry and Psychotherapy, Klinikum der Johannes Gutenberg-Universität Mainz, Untere Zahlbacher Straße 8, 55131 Mainz, Germany.
Dr. Baumert: Department of Internal Medicine, Hugstetterstrasse 55, 79106 Freiburg, Germany.
Dr. Discher: Department of Internal Medicine II, Universitätsklinikum Gießen Marburg, 35392 Gießen, Germany.
Dr. Neumann: Institute for Biostatistics and Clinical Epidemiology, Charité–Universitätsmedizin Berlin, Charite Campus Benjamin Franklin, Hindenburgdamm 30, 12203 Berlin, Germany.
Dr. Berg: Section of Hepatology, Clinic and Policlinic for Gastroenterology and Rheumatology, Universitätsklinikum Leipzig, Liebigstraße 20, 04103 Leipzig, Germany.
Author Contributions: Conception and design: M. Schaefer, T. Berg.
Analysis and interpretation of the data: M. Schaefer, R. Sarkar, V. Knop, S. Effenberger, A. Friebe, J. Ockenga, K. Lieb, A. Heinz, K. Neumann, S. Zeuzem, T. Berg.
Drafting of the article: M. Schaefer, V. Knop, S. Effenberger, A. Friebe, T.F. Baumert, A. Heinz, K. Neumann, T. Berg.
Critical revision of the article for important intellectual content: M. Schaefer, R. Sarkar, V. Knop, U. Spengler, T. Schlaepfer, J. Reimer, K. Lieb, T.F. Baumert, A. Heinz, S. Zeuzem, T. Berg.
Final approval of the article: M. Schaefer, V. Knop, U. Spengler, T. Schlaepfer, P. Buggisch, J. Ockenga, R. Link, M. Rentrop, G. Fromm, K. Lieb, T.F. Baumert, A. Heinz, S. Zeuzem, T. Berg.
Provision of study materials or patients: M. Schaefer, V. Knop, T. Schlaepfer, J. Reimer, P. Buggisch, J. Ockenga, R. Link, H. Weidenbach, G. Fromm, K. Lieb, T.F. Baumert, S. Zeuzem, T. Berg.
Statistical expertise: M. Schaefer, V. Knop, S. Effenberger, K. Neumann, T. Berg.
Obtaining of funding: M. Schaefer, T. Berg.
Administrative, technical, or logistic support: M. Schaefer, R. Sarkar, V. Knop, S. Effenberger, L. Heinze, T.F. Baumert, A. Heinz, T. Berg.
Collection and assembly of data: M. Schaefer, R. Sarkar, V. Knop, S. Effenberger, A. Friebe, L. Heinze, U. Spengler, T. Schlaepfer, J. Reimer, P. Buggisch, J. Ockenga, R. Link, M. Rentrop, H. Weidenbach, G. Fromm, K. Lieb, T.F. Baumert, S. Zeuzem, T. Berg.
Schaefer M, Sarkar R, Knop V, Effenberger S, Friebe A, Heinze L, et al. Escitalopram for the Prevention of Peginterferon-α2a–Associated Depression in Hepatitis C Virus–Infected Patients Without Previous Psychiatric Disease: A Randomized Trial. Ann Intern Med. 2012;157:94-103. doi: 10.7326/0003-4819-157-2-201207170-00006
Download citation file:
Published: Ann Intern Med. 2012;157(2):94-103.
Depression is a major complication during treatment of chronic hepatitis C virus (HCV) infection with interferon-α (IFN-α). It is unclear whether antidepressants can prevent IFN-induced depression in patients without psychiatric risk factors.
To examine whether preemptive antidepressant treatment with escitalopram can decrease the incidence or severity of depression associated with pegylated IFN-α in HCV-infected patients without a history of psychiatric disorders.
Randomized, multicenter, double-blind, prospective, placebo-controlled, parallel-group trial. (ClinicalTrials.gov registration number: NCT00136318)
10 university and 11 academic hospitals in Germany.
181 HCV-infected patients with no history of psychiatric disorders enrolled between August 2004 and December 2008.
Escitalopram, 10 mg/d (n = 90), or placebo (n = 91) administered 2 weeks before and for 24 to 48 weeks during antiviral therapy.
The primary end point was the incidence of depression, defined as a Montgomery–Asberg Depression Rating Scale (MADRS) score of 13 or higher. Secondary end points were time to depression, incidence of major depression according to the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, quality of life, sustained virologic response, tolerability, and safety.
32% (95% CI, 21% to 43%) of the patients in the escitalopram group developed a MADRS score of 13 or higher compared with 59% (CI, 48% to 69%) in the placebo group (absolute difference, 27 percentage points [CI, 12 to 42 percentage points]; P < 0.001). Major depression was diagnosed in 8% of the patients in the escitalopram group and 19% in the placebo group (absolute risk difference, 11 percentage points [CI, 5 to 15 percentage points]; P = 0.031). Tolerability and safety parameters did not differ between the groups. In the escitalopram group, 56% (CI, 46% to 66%) of patients achieved a sustained virologic response compared with 46% (CI, 37% to 57%) in the placebo group (P = 0.21).
Results might not be generalizable to patients with previous psychiatric disease. Some patients withdrew or developed temporary elevated MADRS scores after randomization but before the study medication was started.
Prophylactic antidepressant treatment with escitalopram was effective in reducing the incidence and severity of IFN-associated depression in HCV-infected patients without previous psychiatric disease.
Roche Pharma and Lundbeck.
Depression is a common adverse effect of interferon-α (IFN-α) treatment for chronic hepatitis C virus infection. Whether escitalopram can prevent depression among patients without a history of psychiatric disease who require IFN-α treatment is uncertain.
Compared with placebo, continuous treatment with escitalopram starting 2 weeks before and during IFN-α therapy decreased the number of patients who had depression.
Patients with a history of psychiatric disease were not studied.
Escitalopram may be useful for the prevention of depression when IFN-α treatment is required for chronic hepatitis C virus infection.
Hepatitis C virus (HCV) infection is the most frequent cause of chronic hepatitis and an important risk factor for cirrhosis and hepatocellular carcinoma. The standard treatment for chronic HCV infection is pegylated interferon-α (PEG-IFN-α) combined with antivirals (1). Interferon-α is used to treat several other chronic inflammatory or malignant disorders, such as hepatitis B virus infection, melanoma, hairy cell leukemia, and chronic myelogenous leukemia. However, psychiatric adverse effects, such as depression, fatigue, and insomnia, have been frequently observed during this therapy (2).
Up to 70% of HCV-infected patients treated with IFN-α have been reported to have mild to moderate depressive syndromes (3–5), and 20% to 40% have major depression (5). Interferon-associated depression has a strong negative effect on quality of life and treatment adherence and is a risk factor for treatment failure (5–7). Suicide attempts represent the worst possible complication of IFN-induced depression (8). Preventing onset of depressive symptoms in patients is therefore an important goal.
Recent trials addressing the prophylactic use of selective serotonin reuptake inhibitors (SSRIs) in HCV-infected patients have been limited by small sample sizes, and the contradictory results do not allow for substantial conclusions about daily clinical practice (4, 9–15). Although data support the view that prophylactic antidepressant treatment might decrease the incidence of depression in patients with comorbid psychiatric disorders, whether antidepressants should also be given to patients without previous psychiatric disease remains unclear. We therefore designed a multicenter study to evaluate whether the antidepressant escitalopram can decrease the incidence or severity of IFN-associated depression in treatment-naive HCV-infected patients without a history of psychiatric disorders.
Patients were screened in the departments of hepatology of 10 university and 11 academic hospitals in different areas of Germany. Treatment-naive patients older than 18 years with chronic HCV infection and serum HCV RNA levels of 1000 IU/mL or higher were eligible for the study. Psychiatric exclusion criteria were a lifetime diagnosis of an affective disorder; drug abuse in the past 12 months; treatment with antidepressants during the past 3 years; or a history of any other axis I disorder according to the criteria of the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) (16). Medical exclusion criteria were pretreatment with IFN or immunotherapy, other chronic infection, or an autoimmune or severe somatic comorbid condition.
The study protocol was approved by the ethics committee of Berlin and confirmed by local ethics committees. The study was conducted in accordance with the principles of the Declaration of Helsinki and local laws and regulations. All participants provided written informed consent.
This trial was a multicenter, double-blind, prospective, randomized, placebo-controlled, parallel-group, phase 3 study consisting of preobservation, treatment, and follow-up periods. An independent pharmacist at Charité Berlin was responsible for central block randomization by fax and dispensation of the study medication. Patients were stratified according to age, sex, and genotype. The escitalopram and placebo tablets (Lundbeck, Hamburg, Germany) were identical in appearance. Study personnel and participants were blinded to group assignment.
In the preobservation period, patients were monitored for spontaneously developing symptoms of depression not related to antiviral therapy for 12 weeks before starting antidepressant therapy. After the preobservation period, patients received escitalopram, 10 mg/d (n = 90), or placebo (n = 91) (Figure 1). After 2 weeks of antidepressant pretreatment, all patients began receiving antiviral therapy with PEG-IFN-α2a plus ribavirin (Roche, Grenzach-Wyhlen, Germany) with continuous concomitant administration of escitalopram or placebo.
Study flow diagram.
Most patients who discontinued treatment early withdrew for reasons associated with the antiviral therapy (e.g., nonresponse). HCV = hepatitis C virus; PEG-IFN-α2a = pegylated interferon-α2a.
* During the 2-wk pretreatment period, antiviral therapy was not yet started.
† This patient died of causes unrelated to the study.
Patients with HCV genotype 1 or 4 received treatment for 48 weeks with PEG-IFN-α2a, 180 mcg weekly, and ribavirin, 1000 mg/d (body weight, <75 kg) or 1200 mg/d (body weight, ≥75 kg). Patients with genotype 2 or 3 received PEG-IFN-α2a, 180 mcg weekly, and ribavirin, 800 mg/d, for 24 weeks. Antidepressant treatment was stopped at the end of antiviral therapy.
Psychiatric changes, adverse events, laboratory values, dose adjustments of the antiviral therapy, and possible pharmacologic adverse effects were systematically monitored throughout the study. Hepatitis C virus RNA levels were assessed at weeks 4, 12, and 24 and at week 48 in patients with HCV genotypes 1 and 4, as well as at the end of follow-up (that is, at week 48 in patients with HCV genotype 2 and 3 or at week 72 in patients with HCV genotype 1 and 4) to evaluate sustained virologic response. Serum HCV RNA levels were measured by using the High Pure System COBAS TaqMan HCV Monitor Test (Roche).
Doses of ribavirin and PEG-IFN-2α were adjusted as needed. Adherence was assessed at each clinic visit by pill count. In cases of major depression, an experienced psychiatrist decided whether patients should receive open-label mirtazapine, 30 to 60 mg/d, as rescue medication in addition to the study medication (escitalopram or placebo), to which the patients remained blinded. Benzodiazepines were not allowed during the trial.
Escitalopram or placebo pretreatment was started 2 weeks before antiviral therapy. Psychiatric assessments were performed 14, 8, and 2 weeks before antiviral therapy was started (that is, during the preobservation period), as well as after 2; 4; 12; 24; and, for patients with genotypes 1 and 4, 48 weeks of antiviral therapy (that is, during the treatment period) and 24 weeks after antiviral treatment (that is, during the follow-up period).
In each participating center, 1 experienced psychiatrist performed the screening examination and psychiatric study visits. The Mini-International Neuropsychiatric Interview was used as systematic psychiatric screening to detect preexisting psychiatric disorders (17). The Structured Clinical Interview for DSM-IV Disorders was used to verify major depression before and during treatment (18).
Although depression severity is an important characteristic of major depression, data are lacking about the validity of the DSM-IV definition of the severity of major depression (19). Thus, the severity of depressive symptoms was evaluated by using the interviewer-based Montgomery–Asberg Depression Rating Scale (MADRS), which consists of 10 items (score range per item, 0 to 6) (20). The primary response criterion of depression was having a MADRS score of 13 or higher, and severe depression was defined as having a MADRS score of 25 or higher (20, 21).
Health-related quality of life (HRQOL) was evaluated by using the Short Form-36. This short questionnaire consists of 36 items that measure 8 multi-item variables (physical functioning, social functioning, role limitations due to physical problems, role limitations due to emotional problems, mental health, vitality, pain, and general health perception) and 2 summary scores for mental and physical health (22). For all subscales except “health changes,” a high score indicates a better state of health (0 = worst possible health state; 100 = best possible health state). For health changes, lower scores indicate fewer negative health changes.
The primary end point was the incidence of depression, defined as a MADRS score of 13 or higher during antiviral therapy (up to 48 weeks, depending on genotype). Although no specific MADRS cutoff score exists to define major depression, evidence shows that patients with a MADRS score less than 12 did not have major depressive episodes and that a mean MADRS score of 13 (range, 10 to 16) was associated with mild to moderate depression according to the global clinical impression scale (21, 23). Thus, by choosing the MADRS cutoff of 13 or higher, we aimed to capture all patients with a clinically relevant depressive syndrome. Secondary end points were time to depression defined as a MADRS score of 13 or higher, incidence of major depression defined by DSM-IV criteria, severe depression defined as a MADRS score of 25 or higher, HRQOL, sustained virologic response, tolerability, and safety.
Different trials using psychiatric rating scales showed a prevalence of depression during IFN treatment between 30% and 70% (5). On the basis of the assumption that the incidence of depression defined as a MADRS score of 13 or higher in the placebo group is 40% or more, a sample of 182 patients was needed for a statistical power of 80% to determine whether there was at least an absolute 20% lower rate of depression in the escitalopram group than in the placebo group (2-sided α value, 0.05).
The final analysis included only patients who received at least 1 dose of escitalopram or placebo. Between-group differences for the secondary outcome parameters (major depression, severe depression, sustained virologic response, and adverse events) were calculated with a chi-square test. Kaplan–Meier survival analysis by using the Mantel–Cox log-rank test was performed for group comparison of time to depression. For the primary end point (maximal MADRS score ≥13), we treated missing MADRS assessments by multiple imputation.
We generated m = 10 imputed samples by a linear model using all MADRS assessments, age, and body mass index (BMI). For each patient and sample, the maximum MADRS score was subsequently computed and dichotomized at a MADRS score of 13. For all imputed samples, differences between treatment groups were tested by a logistic model with the sole factor group. Combined point estimates and the covariance matrix of the model parameters were subsequently calculated by the Rubin rule. Finally, the combined P value, depression rates in both groups, and the difference of depression rates with 95% CIs were calculated by the delta method.
The multiple imputation samples were generated by using SPSS, version 19.01 (IBM SPSS, Chicago, Illinois). Patients without HCV RNA measurements at follow-up were considered nonresponders to antiviral treatment. Patients with missing data on major depressive disorder were categorized as nondepressed. A mixed-effects repeated measurement model was used to analyze group differences according to repeated measurements of depression severity (MADRS scores during antiviral treatment) and HRQOL (Short Form-36 scores).
Mixed-effects repeated measurement is well-recognized in psychopharmacologic trials and has been shown to be more robust to biases from missing data compared with the last-observation-carried-forward method (24–27). Baseline covariates, such as sex, age and MADRS score (both dichotomized), HCV genotype (1 and 4 vs. 2 and 3), BMI, study site, and group-by-time interaction terms were included as fixed independent variables. The time variable was treated as a categorical variable with 3 (genotypes 2 and 3) and 4 (genotypes 1 and 4) levels.
To allow for correlations of measurements from the same patient over time, an unstructured covariance matrix was specified within the SAS procedure MIXED (SAS Institute, Cary, North Carolina). Group-by-time contrasts were tested within the mixed-effects repeated measurement models, and P values were subjected to a Holm–Bonferroni procedure for multiple testing.
We used multiple logistic regression analysis to calculate odds ratios of the incidence of MADRS scores of 13 or higher between the 2 groups adjusted for sex, age, BMI, genotype (1 and 4 vs. 2 and 3), baseline MADRS score (median split at 1.42), and study site (4 major sites; all other small sites were combined in 1 category). Mean MADRS scores of the groups 6 months after treatment were compared by t test. The number needed to treat for benefit was calculated as the reciprocal of the between-group risk difference, and the upper and lower 95% CIs were found by inverting the CIs for the risk difference.
All tests and CIs were 2-sided, and the level of significance was an α level of 0.05. Statistical analyses were performed by using SAS 9.2; R, version 2.13.0 (R Foundation for Statistical Computing, Vienna, Austria); and SPSS Statistics, version 19.0.
This work was supported in part by Roche Pharma and Lundbeck. Lundbeck offered medication (escitalopram) and identical placebo pills. Roche was involved in part in the development of the study design by giving statistical support, including the preobservation period, and also gave support in organizing prestudy meetings and data presentation as posters on international scientific meetings. The funding source had no influence on any other aspects of the trial, including data collection, manuscript preparation, and the decision to submit the manuscript for publication.
A total of 208 of the 300 patients screened were enrolled between August 2004 and September 2008. Eighty-two patients did not meet the inclusion criteria. Twenty-seven patients left the trial during the preobservation period after randomization but before starting antidepressant and antiviral therapy because of newly diagnosed somatic or psychiatric disorders or unexpected personal or job-related reasons (Figure 1). There were no apparent systematic differences in baseline characteristics between patients who withdrew before antiviral treatment and those who remained in the study. Baseline data of the remaining patients did not differ between the study groups (Table 1).
Pretreatment Demographic and Clinical Characteristics
During the preobservation period, 3 of the patients in the placebo group and 2 in the escitalopram group developed MADRS scores greater than 10. One patient in the placebo group had a transiently increased MADRS score of 16, but this had decreased to a score of 8 by the time antidepressant therapy was started. No patient had a MADRS score of 13 or higher.
On the basis of an analysis using multiple imputation to account for missing data, 32% (n = 25 [CI, 21% to 43%]) of the escitalopram group and 59% (n = 49 [CI, 48% to 69%]) of the placebo group developed MADRS scores of 13 or higher during antiviral therapy (absolute risk difference, 27 percentage points [CI, 12 to 42 percentage points]; number needed to treat, 3.7 [CI, 2.4 to 8.5]; P < 0.001). Kaplan–Meier curves for the cumulative incidence of depression defined as a MADRS score of 13 or higher differed by group (P < 0.001) (Figure 2).
Cumulative Kaplan–Meier estimates of the time to first episode of depression during antiviral treatment.
More patients in the escitalopram pretreatment group remained depression-free (defined as a MADRS score ≥13) than those in the placebo group (hazard ratio, 0.43 [95% CI, 0.26 to 0.70]; log-rank test P < 0.001). MADRS = Montgomery–Asberg Depression Rating Scale.
In multivariable logistic regression analysis that accounted for escitalopram treatment, IFN-associated depression (MADRS score ≥13) was associated with female sex (P = 0.027) and baseline MADRS score (P < 0.001), whereas no association was seen for BMI (P = 0.055), genotype (P = 0.24), age (P = 0.195), and study site (P = 0.28). Under a sensitivity analysis where patients with missing data were assumed not to have a MADRS score of 13 or higher, 28% of the escitalopram group versus 54% of the placebo group developed a MADRS score of 13 or higher (absolute risk difference, 26 percentage points [CI, 16 to 36 percentage points]; P < 0.001).
The incidence of major depression as defined by the DSM-IV and diagnosed by an experienced psychiatrist was 8% (n = 7 [CI, 4% to 15%]) in the escitalopram group and 19% (n = 17 [CI, 12% to 28%]) in the placebo group (absolute risk difference, 11 percentage points [CI, 5 to 15 percentage points]; P = 0.031). Severe depression defined as a MADRS score of 25 or higher was found in 1% of the escitalopram group (n = 1 [CI, 0.3% to 6%]) compared with 12% of the placebo group (n = 11 [CI, 7% to 21%]).
Montgomery–Asberg Depression Rating Scale scores increased during IFN treatment (weeks 4, 12, and 24) compared with the preobservation period in the escitalopram and placebo groups. Overall, patients in the escitalopram group had lower mean MADRS scores at week 12 (mean difference, 2.8 [CI, 0.9 to 4.7]; P = 0.004), week 24 (mean difference, 3.1 [CI, 1.1 to 5.1]; P = 0.002), and week 48 (mean difference, 5.2 [CI, 2.1 to 8.4]; P < 0.001) of antiviral treatment (Figure 3).
Montgomery–Asberg Depression Rating Scale scores during hepatitis C virus therapy.
Significant group differences were seen at weeks 12 (P = 0.004), 24 (P = 0.002), and 48 (P = 0.001, genotypes 1 and 4).
* Only genotypes 1 and 4.
The treatment effect did not vary by week of antiviral therapy (P = 0.154). Mean MADRS scores 6 months after treatment returned to normal and did not differ between the placebo group (MADRS score, 4.0 [CI, 2.8 to 5.2]) and the escitalopram group (MADRS score, 2.6 [CI, 1.6 to 3.6]; P = 0.081) and if compared with baseline scores.
During the first 24 weeks of the trial, escitalopram therapy resulted in overall better mental health summary scores, as well as scores within the domains of general health perception, social functioning, and mental health, compared with placebo (Figure 4). No overall differences were seen in the physical health summary score or in the other Short Form-36 HRQOL domains. In addition to the differences during the first 24 weeks, patients with genotype 1 or 4 who received antidepressant pretreatment had better scores for general health perception, mental health, role limitations due to emotional problems, social functioning, health change, and mental summary score after 48 weeks of antiviral therapy.
General health status assessed by using the SF-36.
For all subscales except for “health change,” a high score indicates a better state of health (0 = worst possible health state; 100 = best possible health state). For “health change,” lower scores indicate fewer negative health changes. Significant group differences during the first 24 wk of the trial were seen for the mental health summary score (P = 0.002) and within the domains of general health perception (P = 0.037), social functioning (P = 0.035), and mental health (P < 0.001). At 48 wk, additional group differences were found in patients with genotype 1 or 4 for general health perception (P = 0.047), mental health (P = 0.003), health change (P = 0.003), role limitations due to emotional problems (P = 0.003), social functioning (P = 0.025), and mental summary score (P = 0.003). SF-36 = Short Form-36.
Sustained virologic response was achieved in 42 patients (46% [CI, 37% to 57%]) in the placebo group (genotypes 1 or 4, 35% [CI, 25% to 48%]; genotypes 2 or 3, 86% [CI, 72% to 96%]) and in 50 patients (56% [CI, 46% to 66%]) in the escitalopram group (genotype 1 or 4, 42% [CI, 31% to 55%]; genotype 2 or 3, 73% [CI, 56% to 88%]; P = 0.21).
Three patients (3% [CI, 1.2% to 9.4%]) in the escitalopram group and 16 (18% [CI, 11% to 27%]) in the placebo group received mirtazapine as a rescue medication (P = 0.004). No suicidal syndromes or suicide attempts were reported. Seventy-eight out of 90 patients in the escitalopram group (87%) and 80 out of 91 patients in the placebo group (88%) finished antiviral therapy according to protocol.
The number of patients who discontinued treatment early did not differ between the groups. Ten patients (6%) discontinued treatment early because of severe adverse events: 5 (6%) in the placebo group and 5 (6%) in the escitalopram group (P = 0.99). In 5 out of 10 patients, psychiatric events were responsible for treatment failure. Three patients in the placebo group had severe depression, and 2 patients in the escitalopram group reported drug abuse or self-harming behavior. Five patients were withdrawn because of somatic complications: 2 in the placebo group because of retinopathy or glioblastoma and 3 in the escitalopram group because of renal failure, cerebral tumor, or jaundice. Two patients in the escitalopram group withdrew because of nonadherence.
The overall incidence of adverse effects, including possible adverse events of escitalopram or PEG-IFN-2α and ribavirin, was higher in the placebo group (87%) than in the escitalopram group (74%) (absolute risk difference, 13 percentage points [CI, 4 to 22 percentage points]; P = 0.022). No group differences were found regarding specific adverse effects of escitalopram (Table 2). However, significantly fewer patients in the escitalopram group had fatigue (P = 0.040) or insomnia (P = 0.015), which were both frequently observed during IFN treatment.
Safety Data for Escitalopram and Placebo
Laboratory test data, such as levels of alanine and aspartate aminotransferase, γ-glutamyltransferase (Table 2), serum creatinine, hemoglobin, neutrophil granulocytes, and thrombocytes (data not shown), did not differ between treatment groups before, during, or after treatment. Dose reduction of PEG-IFN-2α was necessary for 19% of patients in both groups. Ribavirin dose was decreased in 29% of the placebo group and 24% of the escitalopram group.
To our knowledge, this is the first trial that demonstrates that preemptive and concomitant therapy with escitalopram significantly decreases the incidence and severity of PEG-IFN-α–associated depression in patients without preexisting psychiatric risk factors. Escitalopram treatment was safe and did not negatively influence sustained virologic response to antiviral therapy, which corroborates the findings of trials showing good safety and tolerability of antidepressants in patients with HCV infection during antiviral therapy (4, 10, 11, 15, 24, 25).
We consequently aimed to exclude patients with any history of psychiatric disorders to avoid spontaneous depression independent of antiviral therapy and to assess the real capability of SSRIs to protect patients from IFN-associated depressive mood changes. Although patients with preexisting psychiatric conditions were excluded, 18% of the placebo group developed major depression and 58% developed IFN-α–induced clinically relevant depressive episodes with a MADRS score of 13 or higher.
Most trials reported a 10% to 40% incidence of major depression during antiviral therapy (5). Some trials that did not exclude patients with preexisting psychiatric conditions found up to a 70% incidence of depression (3, 4). Thus, our results corroborate these data and those of recent trials showing that patients without psychiatric risk factors have a similar risk for depression during antiviral therapy with IFN-α as patients with psychiatric disorders (24, 25).
Interferon-α–associated depression has been related to a serotonergic deficit, and changes in serotonin signaling prove the rationale for therapeutic use of SSRIs (2, 26, 27). Indeed, in a prospective, controlled trial, acute treatment with the SSRI citalopram was shown to significantly decrease IFN-associated depressive symptoms (28). However, whether a prophylactic antidepressant therapy offers an advantage over a “treat-as-needed” strategy is uncertain.
Considering our trial's high efficacy in preventing depressive symptoms, as well as the good tolerability of escitalopram, our findings show good reasons to favor a preventive strategy for IFN-associated depression. The diagnosis of major depressive episodes is often missed, because patients with HCV infection are not regularly seen by a psychiatric specialist. According to a recent study, only 32% of patients with a major depressive episode are correctly diagnosed during routine medical interviews (6). Thus, the treat-as-needed strategy might lead to a significant delay in antidepressant therapy with an increased risk for worsening depressive symptoms, nonresponse, treatment discontinuation, or suicide attempts (29).
Depression that is especially severe may be associated with major complications (for example, suicidal thoughts and treatment discontinuation). In our trial, antidepressant pretreatment significantly decreased the incidence of IFN-associated severe depressive episodes.
Psychiatric adverse events of IFN-α treatment significantly decrease patients' quality of life (30). We prospectively investigated HRQOL for the first time in patients receiving a prophylactic antidepressant during antiviral therapy. Escitalopram pretreatment was associated with a better mental health summary score during the course of the study. Other psychiatric adverse effects of IFN-α that are important risk factors for depression, such as sleep disturbance and fatigue (31), also occurred less frequently when escitalopram was given as a prophylactic treatment.
So far, a systematic English-language MEDLINE search to December 2011 yielded 5 randomized, controlled trials investigating prophylactic antidepressant therapy for HCV-infected patients receiving antiviral therapy with IFN-α and ribavirin (9, 10, 13–15). Four trials showed no significant effects of an antidepressant pretreatment on the incidence of depression or depression severity (10, 13–15). Two different small trials of the same group did not find significant effects of a prophylactic treatment with either paroxetine (n = 33) (13) or citalopram (n = 39) (14).
Another trial with a small sample size (n = 52) that showed no general significant benefit demonstrated at least a decrease of depressive symptoms in a subgroup of patients with higher MADRS scores at baseline (15). The fourth negative study included 133 patients who received antidepressant pretreatment with escitalopram (10). However, depression was monitored only in the first 12 weeks of antiviral therapy; furthermore, patients were allowed to continue benzodiazepines or sleep medication, which might explain the overall very low incidence of depression in this trial (31). Finally, a very recently published trial (n = 79) also using escitalopram for antidepressant pretreatment found a significant reduction of single-symptom items in the MADRS scale and the incidence of depression (9), whereas mean MADRS scores or the incidence of major depression according to the DSM-IV criteria was not specifically assessed.
Most trials available so far have limited power owing to small sample sizes, high withdrawal rates, or a low incidence of depression. On the basis of their data, Raison and colleagues (15) calculated that at least 150 patients would be required for sufficient power, a number that so far has been realized only in our trial. Furthermore, in contrast to the other trials, we excluded patients with a history of psychiatric disorders or substance abuse in the past 12 months. In addition, IFN-associated depression was assessed and verified with different instruments for up to 84 weeks during and after antiviral treatment.
Our trial has some limitations. Random assignment of patients before the preobservation period followed by patient withdrawal from the trial independent of antidepressant treatment weakens the strength of the randomization. However, the groups were balanced once antidepressant treatment was started.
The overall observation period of 24 to 48 weeks depending on genotype was associated with a decreased number of cases available at week 48. Thus, the trial was not powered to detect group differences for sustained virologic response. Data about the outcome of acute antidepressant rescue treatment are not available. Finally, our results do not allow conclusions about the preventive effects of antidepressants in patients with drug addiction or psychiatric comorbid conditions.
Future trials should evaluate different antidepressants and possible additional effects on other psychiatric symptoms (such as fatigue and sleep disturbances) and on sustained virologic response. Moreover, more information is needed about new antivirals used in combination with IFN-α and ribavirin and about possible drug–drug interactions between the new antivirals and antidepressants and benzodiazepines (32–35).
In conclusion, our study shows that preemptive treatment with escitalopram is effective in reducing the incidence and severity of depressive symptoms associated with antiviral therapy for chronic HCV infection in patients without psychiatric risk factors. Given the current and future treatment options for PEG-IFN-α in gastroenterology, dermatology, and oncology and the burden of treatment-associated depression, our study's findings are highly relevant for improving and optimizing the safety and tolerability of IFN-α treatment.
The In the Clinic® slide sets are owned and copyrighted by the American College of Physicians (ACP). All text, graphics, trademarks, and other intellectual property incorporated into the slide sets remain the sole and exclusive property of the ACP. The slide sets may be used only by the person who downloads or purchases them and only for the purpose of presenting them during not-for-profit educational activities. Users may incorporate the entire slide set or selected individual slides into their own teaching presentations but may not alter the content of the slides in any way or remove the ACP copyright notice. Users may make print copies for use as hand-outs for the audience the user is personally addressing but may not otherwise reproduce or distribute the slides by any means or media, including but not limited to sending them as e-mail attachments, posting them on Internet or Intranet sites, publishing them in meeting proceedings, or making them available for sale or distribution in any unauthorized form, without the express written permission of the ACP. Unauthorized use of the In the Clinic slide sets will constitute copyright infringement.
Gijs W. D. Landman, MD
Diabetes cCntre, Isala Clinics, Zwolle
August 8, 2012
Primary Outcome Measure Needed
Schaefer et al. performed an industry sponsored trial investigating whether preemptive antidepressive medication could potentially decrease the incidence of moderate to severe depressions in patients treated with a pegylated IFN-a containing regimen in HCV infected patients without a history of depressive symptoms . They found a significant reduction in both moderate and severe depression in the escitalopram group. The incidence of depressive disorders was measured using the MADRS score with a cutoff value of 13 for moderate depressions.
With regard to the definition of a depressive disorder I have 2 concerns. Firstly, the original study protocol (available at clinicaltrials.gov) clearly states that the primary end-point, the incidence of moderate to severe depressions, was originally planned to be evaluated with the Hamilton Depression Rating Scale at week 24 after starting interferon. Secondly, the authors used a cutoff value of 13 for diagnosing moderate depression and 25 for severe depression. However, the cutoff value of 13 is used for defining remission after depression . Cut-off values of 19 and 32 are considered standard for diagnosing moderate and severe depression [3,4]. Using these lower cutoff values is not in accordance with previous studies and could lead to considerable over diagnosing of depressions and could possibly lead to inflation of effect of escitalopram.
To avoid the suggestion of a potential conflict, could the authors provide the primary outcome measured according to the incidence of depressions after 24 weeks of treatment in both the treatment and placebo group according to the Hamilton Depression Scale and also using the MADRS cutoff values of 19 and 32.
1.Schaefer M, Sarkar R, Knop V, Effenberger S, Friebe A, Heinze L, at al. Escitalopram for the Prevention of Peginterferon-α2a-Associated Depression in Hepatitis C Virus-Infected Patients Without Previous Psychiatric Disease: A Randomized Trial. Ann Intern Med. 2012;157:94-103
2.Zimmerman M, Posternak MA, Chelminski I. Implications of using different cut-offs on symptom severity scales to define remission from depression. Int Clin Psychopharmacol. 2004;19:215-20
3.Müller MJ, Himmerich H, Kienzle B, Szegedi A Differentiating moderate and severe depression using the Montgomery-Asberg depression rating scale (MADRS). J Affect Disord. 2003;77:255-60
4.Einarson TR. Evidence based review of escitalopram in treating major depressive disorder in primary care. Int Clin Psychopharmacol. 2004;19:305-10
Martin Schaefer 1,2, Susanne Sarkar 2, Thomas Berg 3
1. Dept of Psychiatry, Psychotherapy and Addiction Medicine, Essen, Germany, Dept of Psychiatry and Psychotherapy. 2. Charité–Universitätsmedizin Berlin, Campus Charité Mitte, Berlin, G
September 21, 2012
In his letter Dr. Landmann raises two queries concerning our use of the MADRS depression scale with a cutoff value of 13 for a moderate depression and also asks why the Hamilton Depressing Rating Scale (HAMD) was not further evaluated by us as initially planned. He suggests to present results using the HAMD scale as well as MADRS cut-off values of 19 and 32.
Here we can show that in our study neither the used depression scale nor the definition of the cut-off score had any influence on group differences regarding the effect of antidepressant pre-treatment on the incidence of depression. Fewer patients suffered from depression in the escitalopram group compared to the placebo group with a MADRS cut-off of ≥19 [7.8% (N=7) vs. 24.2% (n=22), Chi square =9.043, df=1, p=0.003] or a MADRS cut-off of ≥32 [1.1% (N=1) vs. 5.5% (N=5), Chi square =2.0713, df=1, p=0.10]. By choosing the HAMD with a cut-off score of 18 (1), escitalopram pre-treatment significantly reduced the incidence of depression until week 24 of IFN-α therapy [9.2% (N=7) vs. 1.3% (n=1), Chi square =4.832, df=1, p=0.028]. A HAMD score ≥10 (a cut-off comparable to MADRS ≥13) was found in 27.8% (N=25) of patients treated with escitalopram and 48.4% (n=44) with placebo [Chi square =8.119, df=1, p=0.004]. Similar to other groups we finally decided to use the MADRS scale for primary evaluation of IFN-associated depressive mood changes (2;3), because the HAMD scale was considered to include too many items concerning sleep disturbance and somatic problems, which are common side effects of IFN-α treatment but not specific to depressive syndromes (protocol available from the authors). We did not choose a moderate to high MADRS score as primary endpoint because this score may underrepresent patients suffering from mild to moderate depressive mood changes. While only a MADRS scores ≤4 indicate a mood state without depressive symptoms, it has been shown that patients with a MADRS score ≥13 might fulfill the diagnostic criteria for a major depression, which was therefore defined as cut-off for a clinically relevant depression in our trial (4;5).
Finally we wish to point out that our trial was an investigator sponsored trial supported only in part by industry, which had no influence on our data management or publication strategy.
(1) Morasco BJ, Rifai MA, Loftis JM, Indest DW, Moles JK, Hauser P. A randomized trial of paroxetine to prevent interferon-alpha-induced depression in patients with hepatitis C. J Affect Disord. 2007;103:83-90.
(2) de Knegt RJ, Bezemer G, Van Gool AR, Drenth JP, Hansen BE, Droogleever Fortuyn HA et al. Randomised clinical trial: escitalopram for the prevention of psychiatric adverse events during treatment with peginterferon-alfa-2a and ribavirin for chronic hepatitis C. Aliment Pharmacol Ther. 2011;34:1306-17.
(3) Diez-Quevedo C, Masnou H, Planas R, Castellvi P, Gimenez D, Morillas RM et al. Prophylactic treatment with escitalopram of pegylated interferon alfa-2a-induced depression in hepatitis C: a 12-week, randomized, double-blind, placebo-controlled trial. J Clin Psychiatry. 2011;72:522-28.
(4) Zimmerman M, Chelminski I, Posternak M. A review of studies of the Montgomery-Asberg Depression Rating Scale in controls: implications for the definition of remission in treatment studies of depression. Int Clin Psychopharmacol. 2004;19:1-7.
(5) Riedel M, Moller HJ, Obermeier M, Schennach-Wolff R, Bauer M, Adli M et al. Response and remission criteria in major depression--a validation of current practice. J Psychiatr Res. 2010;44:1063-68.
Martin Schaefer, MD, 1,2, Susanne Sarkar, MD,2, Thomas Berg, MD3
1:Dept of Psych/Psychother & Addiction Med, Kliniken Essen-Mitte, Germany. 2: Dept of Psych/Psychother, Charité–Univ Berlin.3:Div of Hepat, Clinic/Policlinic for Gastro/Rheum, Leipzig
November 20, 2012
Based on the comments of Dr. Landman, we noticed, that our primary entry in the database clinicaltrial.gov was not updated with details of the final protocol and indeed included a mistake: The Hamilton depression scale (HAM-D) was listed as the instrument for the primary outcome measurement instead of the MADRS-scale. This has been communicated to the ClinicalTrials.gov registry so that a correction can be posted. Nevertheless, we have to emphasize that the MADRS-scale and not the HAM-D was defined as primary outcome before any patient was included in our final study protocol, which was approved by the responsible ethic committee of the Charité Berlin and by the local ethic committees of the participating centers.
In order to demonstrate that the results in our trial are robust and independent from the finally chosen depression scale, we calculated a group comparison as in our original paper with the HAM-D scale (cut-offs ≥10 for mild to moderate and ≥18 for moderate to severe depression) using multiple imputations for missing data. Overall, 33% (95% CI, 23% to 45%) of the patients in the escitalopram group developed a HAM-D score of ≥10 compared with 54% (95% CI, 43% to 64%) in the placebo group (absolute difference, 20 % points [95% CI, 5 to 35 % points]; p=0.012). 6% (95% CI, 2% to 13%) of the patients in the escitalopram group developed a HAM-D score of ≥ 18 compared with 24% (95% CI, 16% to 34%) in the placebo group (absolute difference, 18% points [95% CI, 7% to 28% points]; p=0.003). Longitudinal analysis of continuous HAM-D scores using a mixed-effects repeated measurement model confirmed lower mean HAM-D scores in the escitalopram group at week 4 (estimated means 4.7 [95% CI, 3.6 to 5.9] vs. 6.3 [95% CI, 5.2 to 7.5], P=0.010), week 12 (estimated means 5.6 [95% CI, 4.3 to 6.9] vs. 7.8 [95% CI, 6.5 to 9.1]; p < 0.001), week 24 (estimated means 5.3 [95% CI, 3.8 to 6.7] vs. 8.6 [95% CI, 7.2 to 10.1]; p = 0.032), and week 48 (estimated means 4.4 [95% CI, 2.7 to 6.2] vs. 7.8 [95% CI, 6.0 to 9.5]; p=0.004) of antiviral treatment if compared to controls (P=0.001 for group difference all weeks combined).
Finally in our trial group differences are also independent from the MARDS cut-off score. Fewer patients suffered from depression in the escitalopram group compared to the placebo group with a MADRS cut-off of ≥19 (estimated means 11% [95% CI, 6 to 21] vs. 28% [95% CI, 20 to 39]; p=0.018) or a MADRS cut-off of ≥32 (estimated means 1% [95% CI, 0 to 8] vs. 6% [95% CI, 3 to 14]; p=0.13).
In conclusion, escitalopram pre-treatment significantly reduced incidence and severity of depression evaluated with the HAM-D or MADRS scales independent from the chosen cut-off score. Both the MADRS and the HAM-D scale are equally effective to monitor depressive mood changes during interferon treatment in patients with hepatitis C.
1 Department of Psychiatry, Psychotherapy and Addiction Medicine, Kliniken Essen-Mitte, Essen, Germany
2 Department of Psychiatry and Psychotherapy, Charité–Universitätsmedizin Berlin, Campus Charité Mitte, Berlin, Germany
3 Division of Hepatology, Clinic and Policlinic for Gastroenterology and Rheumatology, Universitätsklinikum Leipzig, Germany
Gastroenterology/Hepatology, Infectious Disease, Viral Hepatitis.
Results provided by:
Copyright © 2016 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use
This PDF is available to Subscribers Only