Adherence and mental side effects during hepatitis C treatment with interferon alfa and ribavirin in psychiatric risk groups
Editorial note from Jules Levin: although this study is small, it finds individuals on methadone maintenance and individuals with psychiatric problems respond as well as HCV-infected patients without these issues. Patients in this study who were formerly addicted to drugs had a greater dropout rate due to reasons discussed below but these patients still had a good viral response rate. This study underscores that these special population groups can respond well with adequate support services. The psychiatric patients in this study received psychiatric evaluation and support services by a multi-disciplinary support team, and medication. These patients did not only have depression but a variety of other psychiatric problems. With adequate funding these special population, who make a substantial proportion of those with HCV-infection, can receive the support services they need to respond well to therapy. It is cost-effective to provide these services, because if these patients discontinue therapy due to noncompliance and difficulty in tolerating the therapy they will likely receive therapy again. So, rather than treating these patients again it is cost-effective to treat them properly the first time.
Psychiatric disorders or drug addiction are often regarded as contraindications against the use of interferon alfa (IFN-) in patients with chronic hepatitis
C. Our aim was to obtain prospective data on adherence to as well as efficacy and mental side effects of treatment with IFN- in different psychiatric
risk groups compared with controls. In a prospective trial, 81 patients with chronic hepatitis C (positive hepatitis C virus [HCV] RNA and elevated alanine
aminotransferase[ALT] level) and psychiatric disorders (n = 16), methadone substitution (n = 21), former drug addiction (n = 21), or controls without a
psychiatric history or drug addiction (n = 23) were treated with a combination of IFN--2a 3 MU 3 times weekly and ribavirin (1,000-1,200 mg/d).
Sustained virologic response (overall, 37%) did not differ significantly between subgroups. No significant differences between groups were detected with
respect to IFN--related development of depressions during treatment. However, in the psychiatric group, significantly more patients received
anti-depressants before and during treatment with IFN- (P < .001). Most of those who dropped out of the study were patients with former drug addiction
(43%; P = .04) compared with 14% in the methadone group, 13% in the control group, and 18% in the psychiatric group. No patient in the psychiatric
group had to discontinue treatment because of psychiatric deterioration.
In conclusion, our data do not confirm the supposed increased risk for
IFN--induced mental side effects and dropouts in psychiatric patients if interdisciplinary care and antidepressant treatment are available. Preexisting
psychiatric disorders or present methadone substitution should no longer be regarded as contraindications to treatment of chronic hepatitis C with IFN-
and ribavirin in an interdisciplinary setting.
(HEPATOLOGY 2003;37:443-451. Department of Psychiatry, Charite, Humboldt University, Berlin; and Departments of 2Psychiatry, 3Gastroenterology, and 4Biostatistics, Ludwig Maximilians University, Munich, Germany)
Editorial note: a pilot study from Diana Sylvestri at the OASIS clinic in Oakland finds that with extensive support services patients on methadone maintenance can respond well to HCV therapy even if they relapse into drug use. The support services monitor patients closely and intervene if relapse occurs, and intervention can be successful in stopping the relapse.
Combination treatment with IFN- and ribavirin has not yet been evaluated in a controlled manner in psychiatric patients and those with a former drug
addiction. We studied compliance with treatment, side effects, response rates, and dropout rates during combination treatment of chronic hepatitis C in
patients with psychiatric disorders, former drug addiction, or ongoing methadone substitution compared with controls without a psychiatric history or drug addiction.
Six patients in the psychiatric group had major depression, one had a
general anxiety disorder, 2 had schizoaffective disorder, 4 had chronic schizophrenia, and 3 had severe borderline personality disorder combined with a major depression. Five patients in this group also had a history of former drug abuse, and 2 patients with schizophrenic disorders additionally received methadone.
In the intention-to-treat analysis, at the end of treatment, 47% of all patients were HCV RNA negative (virologic response): 34% with genotype 1, 71% with
genotype 2, 60% with genotype 3, and 33% with genotype 4.
During the 6 months after the end of therapy, 11% (only genotypes 1 and 3) of all patients experienced a relapse without significant differences between
treatment groups. Overall, 37% of the patients showed a sustained virologic response: 35% in the control group, 38% in the psychiatric group, 43% in the
methadone group, and 28% in the former addiction group. Group differences were not significant. A sustained response was shown in 22% of the patients with genotype 1, 71% with genotype 2, 47% with genotype 3, and 33% with genotype 4.
Therapy was discontinued early in 22% of the patients; reasons were somatic (5%) or psychiatric (2%) side effects, relapse in drug and alcohol abuse (2%), and noncompliance (13%). A total of 13% in the control group and 14% in the methadone group dropped out, mostly because of noncompliance. A total of 18% of the psychiatric group discontinued treatment because of somatic complications or noncompliance. No patients in the psychiatric group had to stop treatment because of psychiatric side effects. In 43% of patients in the former addiction group (P < .01 vs. patients in the control and psychiatric groups), treatment was terminated prematurely because of noncompliance (13%), depression (5%), suicidal thoughts (5%), relapse in alcohol or drug abuse (10%), or somatic side effects (10%). The dropout rate was highest during the first 2 months of therapy.
The mean age was 40 ± 12 years. Forty-nine men and 32 women were included. Patients in the former addiction and methadone groups were significantly
younger than those in the control group (P < .01 and P < .001, respectively), as expected from the epidemiology of addiction. There were more women in the
psychiatric group compared with the methadone and former addiction groups (P < .05), which might be explained by the higher incidence of major
depression and borderline personality in women. Pretreatment with IFN-, baseline ALT levels, virus load, the frequency of histologically proven cirrhosis,
body weight (mean, 73.3 ± 12.2 kg), and height (mean, 173.6 ± 8.1 cm) did not significantly differ between groups.
The most frequent genotypes were 1 and 3 compared with genotypes 2 and 4. Genotype 1 was more frequent in patients in the control and psychiatric groups, whereas genotype 3 (P = .01 vs. controls) was predominant in the drug addiction groups.
Depression and psychiatric side effects
More patients in the psychiatric group were in a depressed mood when entering the study (P < .001 vs. all other subgroups). A total of 16% of all patients
developed a new depression during treatment with IFN-. We found no significant differences in frequency and severity of new depressions during treatment between the groups. However, patients with drug addiction tended to have more often but milder depressions, whereas the severity of depressive episodes in patients in the control and psychiatric groups tended to be moderate or severe. Suicidal thoughts were reported in 4% to 6% of patients, without significant differences between the subgroups. Only 2 patients (3%) dropped out because of a worsening of preexisting depression or development of suicidal thoughts. However, none of these patients had to stop treatment from the psychiatric point of view and suicidal thoughts disappeared under psychiatric care in all cases. Depression before or during treatment had no statistically significant influence on therapeutic outcome (sustained response) or dropout rate. In addition, patients with depression had significantly less problems with alcohol consumption during treatment (P = .046).
As expected, craving for drugs or alcohol was reported less frequently in controls compared with patients in the psychiatric group (P = .02), methadone group (P < .001), and former addiction group (P < .001). 76% in methadone and 62% in the former addiction groups reported "craving" compared to 38% in the psychiatric group and 0 in the control group. The most frequent psychiatric side effect was irritability. The frequency of sleeping disturbances,
concentration difficulties, and irritability did not differ significantly between groups. Sleeping disturbances were treated with benzodiazepines
(zolpidem or zopiclone) in 28%. Admission to the psychiatric ward was necessary for 3 patients in the psychiatric group and 2 patients in the methadone and former addiction groups but for none of the controls. It could not be determined in any case that admission was caused by a direct association with psychiatric side effects of treatment with IFN-.
Reporting on new depressive episodes during treatment with interferon: 6/21 (29%) in the former addiction group, 3/21 in the methadone group, 1/16 (6%) in the psychiatric group, and 3/12 (12%) in the control group. The numbers of patients were small in the study which may be the reason why there were no significant diiferences between the groups although the former addition group had a higher percentage experiencing new depressive episodes.
Use of antidepressants
Overall, 4 of the 81 patients (5%) were on antidepressants before and 13 of 81 patients (16%) received antidepressants during treatment with IFN- for hepatitis C. For 6 of the patients in the psychiatric group, an additional antidepressant treatment was initiated during the first 2 weeks of treatment because of preexisting depression. The following antidepressants were used: citalopram (10 times), mirtazapine (3 times), nefazodone (one time), paroxetine (2 times), fluoxetine (one time), and amitriptyline (one time). We found significant group differences in use of antidepressants before (P = .0322) and highly significant differences (P < .0001) at the end of the treatment with IFN-, with the most frequent use in the psychiatric group. Regarding the influence on liver function, antidepressant treatment did not influence ALT levels during or after treatment compared with patients not taking antidepressants. Relapse after treatment was also not related to antidepressants.
Seven of 23 patients in the control group (30%), 4 of 16 patients in the psychiatric group (25%), 8 of 21 patients in the methadone group (38%), and 7 of 21 patients in the former addiction group (33%) reported occasional alcohol consumption before hepatitis treatment. During treatment, 3 patients in the control group, 3 patients in the psychiatric group, 6 patients in the methadone group, and 9 patients in the former addiction group reported alcohol intake. Two patients in the methadone group and 4 patients in the former addiction group had to be treated for alcohol abuse. However, only one patient in the former addiction group dropped out after 8 weeks; all other patients could be stabilized by inpatient treatment over 8 to 14 days.
Alcohol abuse before treatment was associated with increased ALT levels after 6 months (P = .038) and 12 months (P = .039) of treatment but not with the
mean ALT level 6 months after treatment (P = .302). No differences were found between treatment groups. Alcohol consumption during treatment had no
influence on the mean ALT levels at 6 months (P = .226) and 12 months
(P = .147) during and 6 months after treatment (P = .151), respectively. Furthermore, patients who reported occasional alcohol consumption before or during hepatitis treatment showed no significant differences with respect to a virologic relapse with positive polymerase chain reaction.
In summary, we could not find differences in alcohol intake between treatment groups. Furthermore, alcohol use before or during treatment did not influence
the therapeutic outcome and ALT levels after treatment.
History of addiction and methadone substitution
Patients in the former addiction group reported a duration of drug abuse between 11 months and 22 years (mean, 8.9 years) and patients in the methadone group reported a duration between 7 months and 24 years (mean, 11.6 years). Patients were in methadone substitution between 5 months and 13 years (mean, 3.4 years). The mean dosage of methadone before treatment was 57 mg/d (6-160 mg/d) and increased slightly to 59 mg/d (0-170 mg/d) at the end point of treatment with IFN-. Eight of 21 patients (38%) were kept on a stable dosage. The dosage was increased in 38% and decreased in 24%. Thus, we found no statistically significant differences between methadone dosage before and during treatment with IFN-.
The dosage of ribavirin had to be reduced in 38%. Diarrhea was significantly more frequent in the methadone group compared with the former
addiction group (P = .011) and control group (P < .001). In those with former addiction, itching was rare (P = .002 compared with controls) but coughing was
frequent (P = .002 vs. controls, P = .001 vs. patients in the psychiatric group). Coughing was reported in 53% in former addiction group, 24% in methadone group, compared to 0 in the psychiatric group and 9% in the control group. Fatigue differed significantly only between patients in the methadone and
former addiction groups (P = .004). 86% in the methadone group reported fatigue, 38% in the former addiction group, 69% in the psychiatric group, and 69% in the control group. Hyperthyroidism in 2 patients in the psychiatric group and pneumonia in one patient in the methadone group were treated successfully. Other frequent somatic side effects, such as headache, myalgia, pyrexia, alopecia, nausea, exanthema, and pharyngitis, did not differ between subgroups. Differences in cough, diarrhea, and itching might be explained by long-term effects of methadone and opiates on the nervous system of the skin and the abdominal and pulmonary organs.
Discussion by authors
IFN- has been therapeutically used for more than 15 years in patients with chronic hepatitis, but there only are few psychiatric investigations about the
incidence and severity of psychological side effects, especially in patients with predisposing conditions. In a prospective study using internationally
standardized psychiatric diagnostic criteria, depression during treatment of hepatitis with IFN- occurred in 38% of nonpsychiatric patients after 12 weeks. In 2 multicenter studies, in which DSM criteria were not used, the incidence of depression ranged between 11% and 37%. Several cases of IFN--related
suicide attempts or severe psychosis have been reported, but most of these patients had no psychiatric history. Furthermore, a prior individual or family history of psychiatric illness does not seem to predict depression or anxiety during IFN- therapy. Suicide attempts during IFN- therapy have been reported mainly in patients with organic personality changes or delirious syndromes. These patients and their families did not attribute thought and behavioral changes to IFN- because they had no experience with psychiatric symptoms and therefore did not seek psychiatric help. Taken together, the assumption of a generally increased risk for psychiatric side effects in psychiatric patients has been deduced from reports about complications in nonpsychiatric patients admitted to clinical trials.
We therefore studied psychiatric symptoms during IFN- therapy in patients believed to be at increased risk. In the present trial, patients in all subgroups
developed psychiatric side effects that were systematically evaluated by standardized diagnostic criteria (DSM-IV), but no significant differences in the
frequency and severity of psychiatric side effects between subgroups were noted. Psychiatric patients had more depressive symptoms before and during
treatment compared with the other groups, with an increased need for treatment with antidepressants. However, these differences also seem to reflect their psychiatric condition before IFN- therapy and cannot be completely ascribed to IFN--induced mood changes. This is supported by the observation that psychiatric patients had the best compliance and a low dropout rate. In our trial, depression and suicidal thoughts occurred in most cases independently of a preexisting psychiatric disorder. These results are in line with 2 recent studies reporting that psychiatric patients with chronic hepatitis C can be safely treated with IFN- monotherapy.
It has been suggested that schizophrenic and schizoaffective patients as well as patients with depression or anxiety disorders show no worsening during IFN-
therapy.24-26 In our study, even patients with a severe borderline personality disorder and some patients with drug addiction tended to be more stable and
therapeutically reliable during treatment. The awareness of the risk of liver damage and the opportunity to receive a potentially successful treatment seemed to support cooperative behavior.
In all patients who had depression during treatment with IFN-, improvement was possible with psychiatric and psychopharmacologic support. Novel
antidepressants with less sedation (selective serotonin reuptake inhibitors) and benzodiazepines were especially effective in treating sleeping disturbances,
irritability, and depression. The successful use of selective serotonin reuptake inhibitors for the management of IFN--related psychiatric effects has been
described in several case reports.27-30 Most patients in the psychiatric group continued treatment with antidepressants because of a preexisting depression.
Only one patient developed a new major depressive episode, although he received treatment with antidepressants. The earlier use of antidepressants in the psychiatric group may explain the low incidence of major depressive episodes and suicidal syndromes. Recently, Musselman et al. showed that pretreatment with paroxetine as an antidepressant was highly effective in preventing depression and reducing the dropout rate during adjuvant treatment of melanoma with IFN-. In addition, Kraus et al. reported high efficacy in treating acute IFN--associated depressive symptoms.28 Thus, our data extend the efficacy of antidepressants in preventing and treating IFN--associated depressive episodes, especially in psychiatric patients and patients with methadone substitution and chronic hepatitis C. Overall, the low rate of dropouts due to psychiatric side effects may be in part the result of the timely use of antidepressants in all groups. Moreover, the high dropout rate of patients with former drug addiction may also be explained by the significantly less frequent use of antidepressants in this group.
Interestingly, we could not find differences in ALT levels during and after treatment with IFN- in patients who received treatment with antidepressants
compared with others. Thus, antidepressants did not affect liver function during treatment of hepatitis C. However, controlled studies focusing on the efficacy of antidepressants in preventing psychiatric side effects of IFN- and interactions with response and liver enzymes during and after treatment of hepatitis C are needed.
We are aware that our results have to be interpreted cautiously because of the heterogeneity in psychiatric diagnoses, the lack of psychiatric controls without
IFN- therapy, and the limited number of patients. Final conclusions about the incidence of psychiatric side effects for special diagnostic subgroups during
IFN- therapy and a causal relationship between the necessity of hospitalization of psychiatric patients cannot be drawn. Hyperthyroidism led to hospitalization in at least one patient. Hyperthyroidism should be closely monitored, particularly in psychiatric patients, because of its relevance for the mental state.
Study of the virologic response was not the main purpose of this investigation, but some comments seem appropriate. The somewhat-low sustained response
may be explained by the inclusion of IFN- relapsers and nonresponders and the high number of patients with genotype 1. The better response in the
methadone group can be ascribed to the higher prevalence of genotypes 2 and 3, whereas the low rate of sustained response by intention-to-treat analysis in
the former addiction group was due to the high dropout rate in this group. With better adherence, higher response rates will be possible in both
methadone-substituted patients and in patients with former drug addiction.
Methadone-substituted patients are often classified as active drug users and excluded from treatment of hepatitis. Although a recent study on treatment of
hepatitis C in injection drug users showed encouraging sustained response rates, no controlled studies regarding methadone-substituted patients are available thus far. In our trial, methadone-substituted patients finished treatment without relevant differences in psychiatric side effects compared with controls. In contrast to patients with former drug abuse, no patient dropped out because of drug relapse during treatment, although most patients reported craving. Cannabis abuse or occasional intake of benzodiazepines and even a single intravenous drug abuse using sterile needles were not regarded as a reason for cessation of therapy because of the small risk of reinfection. This is in line with the current discussion about treatment of HCV in active drug users. Patients with former addiction had the highest dropout rate and mental instability, although severe psychiatric comorbidity was excluded. Time of abstinence, ranging from 3 months to several years, did not seem to influence the number of dropouts due to a drug relapse. However, more patients are necessary to answer this question. The increased dropout rate of patients with former addiction in our study is in line with the results of Renault and Hoofnagle. Therefore, treatment results, especially in those with a former drug addiction and methadone-substituted patients, depend on their motivation, compliance, and toleration of side effects. However, former drug users with known personality and behavioral problems are the main "difficult to treat" group and will need special interdisciplinary management concepts in future clinical hepatitis C treatment trials to minimize the dropout rate and increase response to treatment with IFN-.
In conclusion, our prospective controlled study evaluating IFN- therapy in various psychiatric risk groups with chronic hepatitis C should encourage
physicians to offer this effective therapy to these patients. Although psychiatric-risk patients need special psychiatric care, this can be achieved in an interdisciplinary setting. Because of the frequent psychiatric changes during treatment with IFN-, we recommend close cooperation with a psychiatrist
starting before treatment. Our results add objective psychiatric data to the present discussion concerning the tenability of psychiatric contraindications to IFN- and the practicability of IFN- during methadone substitution.