Kava-Kava extract LI 150 is as effective as Opipramol and Buspirone in Generalised Anxiety Disorder - An 8-week randomized, double-blind multi-centre clinical trial in 129 out-patients.
Objective: An 8-week randomized, reference-controlled, double-blind, multi-centre clinical trial investigated Kava-Kava LI 150 in Generalized Anxiety Disorder (GAD; ICD-1O: F41.1). Method: 129 out-patients received either 400 mg Kava LI 150, 10 mg Buspirone or 100mg Opipramol daily for 8 weeks. At week 9, subjects were seen to check for symptoms of withdrawal or relapse. Primary outcome measures comprised the HAMA scale and the proportion of responders at week 8. Secondary measures were the Boerner Anxiety Scale (BOEAS), SAS, CGI, a selfrating scale for well-being (Bf-S), a sleep questionnaire (SF-B), a quality-of-life questionnaire (AL) and global judgements by investigator and patients.
Results: In 127 patients (ITT) no significant differences could be observed regarding all efficacy and safety measures. About 75% of patients were classified as responders (50% reduction of HAMA score) in each treatment group, about 60% achieved full remission.
Conclusion: Kava-Kava LI150 is well tolerated and as effective as Buspirone and Opipramol in the acute treatment of out-patients suffering from GAD.
Key words: Kava, Buspirone, Opipramol, Generalized Anxiety Disorder, out-patient treatment, anxiety outcome-measures, clinical trial
With an estimated lifetime prevalence of 5.1% Generalized Anxiety Disorder (GAD) is an important anxiety disorder (Kessler et al. 1994). Its course is often chronic with a mean duration of illness of 20 years ore more (Yonkers et al. 1996). The illness shows a low rate of spontaneous remission (Yonkers et al. 1996) and is related to high comorbidity, especially with other anxiety disorders and depression (Brawman-Mintzer and Lydiard, 1996). GAD is associated with psychosocial impairment, increased morbidity and mortality and therefore constitutes a serious public health problem (Kessler et al. 1994).
The diagnostic criteria for this disorder were changed considerably over time. There are some important differences between the present operationalized diagnostic systems of ICD-10 and DSM-IV. While somatic complaints are emphasized in ICD-10, the central features of DSM-IV are unrealistic or excessive anxiety and worry.
Although the criteria for diagnosis are clear now and patients suffering from this disorder are frequently seen in primary care settings, misdiagnosis and inadequate treatment still represent major problems. Among primary care patients who were identified as distressed high utilizers of medical care, most were diagnosed with either major depression or GAD (Katon et al. 1990).
The pharmacological treatment of patients with GAD has been established very well during the last years with several options (Boerner, 2002a). Beside the traditional use of benzodiazepines (Shader and Greenblatt, 1993) Buspirone had been the golden standard in the past, especially in the U.S.A. (Schweitzer and Rickels, 1997; Anonymous, 1998). Also the tricyclic imipramine was tested to be effective in some clinical studies (Rickels et al. 1993). Although classified and licensed as an antidepressant, in Germany Opipramol is very popular with general practitioners for the treatment of anxiety disorders (Lohse and Mtiller-Oerling-hausen, 2001; Volz and Moller, 1998).
Although these options still provide beneficial drug treatment for these patients, new possibilities are arising. Some methodologically sophisticated studies have demonstrated the efficacy of the SSRI Paroxetin (Pollack et al. 2001) and of the SNRI Venlafaxin (Sheehan, 1999; Davidson et al. 1999; Rickels et al. 2000; Allgulander et al. 2001) in short and long-term strategies.
But it is a fact, that currently very few patients are provided with these medications in clinical practice. Several reasons may contribute to this observation: misdiagnosis or insufficient knowledge of doctors about these strategies, high costs of the treatment or negative attitude of the patients toward any pharmacological approach.
Although most of the medications mentioned, especially Paroxetin and Venlafaxin, were tolerated very well in the clinical GAD trials, some patients in these studies were afraid of any possible side effect or in fact were very sensitive and could not be pharmacologically treated.
Therefore there is a rationale to look for possible alternative strategies. Nonsynthetic drugs claiming anxiolytic properties like Kava Kava preparations are available as over-the-counter medicines. For some patients these kind of drugs are very attractive with respect to their natural origin and to the expectation of effectiveness combined with excellent tolerability.
Europeans first came into contact with Kava Kava (Piper methysticum Forster) in the 18th century during expeditions discovering the island world of Polynesia, Melanesia and Micronesia. The relaxing and calming beverage prepared from Kava-Kava roots today still plays a central role in social life of the natives of these islands and may be related to the low incidence of alcohol misuse and cancer in the regarding populations (Singh, 1992; Steiner, 2000). The lipophilic kavapyrones could be identified as the major pharmacologically active principles, exerting central muscle-relaxant and anticonvulsive, analgesic, anaesthetic and sedating effects in animal models (Meyer, 1962; Meyer, 1967; Buckley et al. 1967). Today, several products containing a dry extract made from Kava rhizome, standardized to contain a fixed amount of kavapyrones, are marketed as remedies for nervousness, tension and anxiety states.
For the preparation of a Cochrane Review, Pittler and Ernst (2001) very recently have evaluated the scientific literature reporting clinical studies of Kava preparations in conditions related to anxiety. They have identified 7 randomized, controlled, double-blind trials meeting the strict inclusion criteria to enter a metaanalysis. All of the reviewed trials suggest superiority of Kava over placebo. The meta-analysis of three studies out of these seven using the Hamilton Anxiety Scale (HAMA) as outcome measure revealed a significant treatment effect in favour of Kava (weighted mean difference: 9.69; 95% confidence interval: 3.54-15.83). The authors concluded that Kava extract is an effective symptomatic treatment for anxiety (Pittler and Ernst, 2001). Another recent clinical study not included in the above-mentioned review proved efficacy of a Kava extract in the treatment of non-psychotic anxiety, following pre-treatment with benzodiazepines (Malsch and Keser, 2001).
The shortage of most of these studies is the lack of precise diagnostic criteria for inclusion of patients. Either patients with a broad spectrum of anxiety disorders from phobias to adjustment disorders with anxiety were recruited or diagnosis was not made according to LCD or DSM by means of standardized and operationalized procedures like structured or semi-structured interviews. Clinicians may argue that this situation reflects the clinical reality where Kava preparations actually are used for many states or syndromes associated with mild to moderate anxiety symptoms (Boemer, 2001).
Furthermore, all previous clinical studies used dosages of Kava above the maximum daily dose of 120 mg of kavapyrones recommended by the commission E monograph which is the basis of all Kava drug Licenses in Germany (Bundesinstitut fur Arzneimittel und Medizinprodukte, 1990).
The aim of the present study was to investigate the efficacy and safety of a new Kava formulation in the treatment of Generalized Anxiety Disorder according to ICD-10 diagnostic criteria at a daily dose of 120 mg of kavapyrones, compared to Buspirone and Opipramol.
To better characterise the broad spectrum of anxiety psychopathology in GAD patients and monitor regarding changes under treatment, a new investigator rating instrument, providing a more precise and complete collection of anxiety symptoms than the HAMA scale, was introduced in this trial.
* Materials and Methods
The study was outlined as a randomized, double-blind, parallel-group, 8 weeks active phase multi-centre outpatient trial to assess efficacy and safety of Kava Kava extract LI 150 in comparison with the reference treatments Buspirone and Opipramol in Generalized Anxiety Disorder with a 1 week follow-up for detection of withdrawal symptoms and early relapses.
Patients who remained eligible one to seven days after a successful screening visit (open run-in phase) were randomly assigned in a 1:1:1 ratio to receive one of the three treatments (week 0). Computer-aided randomisation (Rancode; Wiedey, Konstanz, Germany) was performed in blocks of six by the independent Quality Assurance Unit of the sponsor. No further stratification was carried out. Blistered study medication and medication containers were labelled sequentially according to a coding list by the manufacturing department of the sponsor and provided to the investigators in blocks of six via the Contract Research Organisation. All investigators, personnel of the CR0 and of the sponsor actively involved in the trial were blinded to group assignment until lock of database at the end of the trial.
Following randomisation, patients were seen at weeks 2, 4 and 8. After another week without treatment (week 9), a follow-up visit was carried out to unravel early relapses and symptoms of withdrawal. A drugscreening test for narcotics and other psychotropic substances (TRIAGETM) was performed at the screening visit. Efficacy rating scales, adverse events, concomitant therapies and vital signs were assessed at all visits from week 0 to week 9. At weeks 2, 4 and 8, the global judgement of efficacy and tolerability by investigator and patient, drug accountability (pill-counting) and qualitative compliance ratings by investigators ("good", "moderate" or "poor") were recorded. Physical exams, blood chemistry and hematology were done at screening and at week 8. Patients gave written informed consent, the study protocol and patient information materials were approved by federal ethical committees in charge of the participating investigators.
The study was performed in full accordance with Good Clinical Practice (GCP) Guidelines, the Declaration of Helsinki and its revisions and with the EC Guideline "Clinical Investigations of Medical Products in the Treatment of Generalized Anxiety Disorder, Panic Disorder and Obsessive-Compulsive Disorder" (May 1994; EudraLex 3CC28a).
Outpatients meeting the ICD-1 criteria for Generalized Anxiety Disorder (GAD; F 41.1) (WHO, 1992) were recruited from community doctors. GAD was diagnozed according to ICD- 10 by means of the semistructured International Diagnostic Checklists (IDCL) (Hiller et al. 1993).
To enhance the quality of diagnostic procedures and of the use of rating instruments, investigators were trained before start of the trial by an experienced psychiatrist in a 1-day session with the help of sample patient interviews recorded on video tapes.
Inclusion criteria were: age 25 to 65 years; current diagnosis of GAD; a minimum total score of 19 on the HAMA and a maximum total score of 12 on the 17-item Hamilton Depression Scale (HAMD); capacity and willingness to give informed consent and to follow study procedures.
Exclusion criteria were: the following concurrent diagnoses according to ICD-1 criteria: depressive episode, panic disorder, phobias, obsessive-compulsive disorder, hypochondria, substance-related disorders, mental disorders of organic origin, psychotic mental disorders; history of seizure disorder or use of anticonvulsants; serious unstable acute or chronic medical illness; seriously impaired hepatic function (transaminases more than double of upper limit) or renal failure (plasma creatinine >2 mg/dl); blood count deviations >10% outside limits; pregnancy, breast-feeding, or use of inadequate methods for birth control in fertile women; concurrent initiation of hormonal contraception or other treatments with sex hormones; allergy or hypersensitivity to study medications; positive urine drug screen, participation in another clinical trial during the preceding six weeks.
Psychotropic medications had to be withdrawn at least two weeks prior to screening; no psychotropic substances were allowed during the trial. Any psychotherapeutic intervention had to be stopped at least four weeks prior to screening.
All active medications and matching placebos were provided by Lichtwer Pharma AG (Berlin, Germany) according to GMP requirements.
At baseline, subjects were randomly assigned to receive one of the following medications: one coated tablet containing 400 mg Kava Kava extract LI150 (standardized to a content of 30% kavapyrones, extraction solvent 96% ethanol in water, drug-extract ratio 13-20:1) once a day in the morning; one coated tablet containing 5 mg Buspirone-hydrochloride b.i.d.; one coated tablet containing 50 mg Opipramol-dihydrochioride b.i.d.
To ensure the blind, to simplify the manufacture of matching placebos and to reduce the total number of tablets to be taken by the patients, Buspirone and Opipramol tablets were encapsulated in capsules of identical shape, size and colour. Thus, each participant received one coated tablet and one capsule b.i.d. containing active substance or placebo. Medication was dispensed in blister packets, in double-dummy fashion as described above, labelled according to the provisions of the German Medicines Act.
Primary efficacy end points were the change in HAMA total score from baseline to week 8 and the proportion of responders at week 8. Response was defined as a decrease of more than 50% of the HAMA total score at week 8 relative to baseline.
Secondary endpoints comprised the HAMA subscales "psychic anxiety" and "somatic anxiety", the Boerner Anxiety Scale (BOBAS) (Boerner, 2002b), Self-Rating Anxiety Scale (SAS), Clinical Global Impressions (CGI), a self-rating scale for well-being (BfS) (von Zerssen, 1973), a sleep questionnaire (SF-B) (Gortelmeyer, 1985) and a quality-of-life questionnaire (AL) (Bullinger et al. 1993) as well as the global judgement of efficacy by patient and investigator.
Further, the response rate on the basis of the BOEAS scale (decrease >50%), and the remission rates on the basis of both, the HAMA (total score week 8 <9) and the BOEAS scale (total score week 8 <5) were calculated.
The Boerner Anxiety Scale (BOEAS) is a new investigator rating scale which has been developed to outline more precisely the broad psychopathological aspects of anxiety in accordance with the present classification of anxiety disorders in ICD-10 and DSM-IV These dimensions of disease which are of particular interest in Generalized Anxiety Disorder are mainly disregarded in the HAMA scale, where somatic symptoms dominate. The BOEAS comprises 11 items: anxiety mood, worry, phobic symptoms, anticipatory anxiety, social anxious symptoms, agoraphobic behaviour, panic attacks, nervousness, irritability, impairment of concentration and catastrophic ideas.
Severity can be rated from 0 to 3 points, thus a maximum total score of 33 can be reached. The BOBAS scale has been tested for validity and reliability with good correlation especially with the HAMA in this trial (r = 0.89) and in another pooled patient population (r = 0.84) (Boerner, 2002b). It was utilized here for the first time in a larger randomised clinical trial in GAD. Meanwhile, the BOEAS has also been used in GAD and other anxiety disorders drug trials investigating Paroxetin and Hydroxycin.
Adverse events were recorded by the investigator at each visit, based on patient interview. Patient and investigator independently rated tolerability on a simple 4-levelled scale (excellent, good, satisfactory, poor) at weeks 2, 4 and 8. The efficacy index of the CGI and laboratory results provide further information about safety and tolerability of treatments.
As the trial was outlined as a proof-of-concept study to gain first information on efficacy and safety of the new Kava preparation, no placebo control was introduced and no inference statistics were carried out. Hence, the sample size was set to 40 patients per group (total n = 120) to be able to detect medium- to large-sized effects.
Descriptive statistics for all variables were provided, tests for treatment differences included chi-square and Fisher's exact tests for categorical variables and Kruskal-Wallis tests for ordinal and continuous measurements. Mann-Whitney-U tests for group wise comparisons were performed in case of p-values below 0.05 in Kruskal-Wallis tests. ANCOVA was used to test for centre effects.
In a blind review meeting, the final schedule of statistical analyses was detailed after data cleaning was finalized and data base was locked. The primary analysis included all randomized patients who at least had one follow-up visit under medication (Intention-to-treat population, ITT). Missing values under treatment due to drop-outs were handled conservatively according to the Last-observation-carried-forward (LOCF) principle for ITT analysis. All patients who had completed the 8 week treatment phase without major protocol violations were analysed separately (According-to-protocol population, ATP).
95% confidence intervals of differences of means between treatments are given as the most appropriate method to quantify differences and to demonstrate equivalence.
A total of 18 community doctors (general practitioners, neurologists, psychiatrists) throughout Germany recruited patients between August 1998 and July 1999. Of 138 subjects who underwent screening procedures, 129 met all inclusion criteria and were randomised. Only two randomized patients were not eligible for ITT analysis (n = 127). One subject, who had been randomized erroneously, did not meet inclusion criteria, another was lost to follow-up after randomization. Eight patients dropped out during the 8 week treatment phase (5 in the Kava group, 3 in the Buspirone group), therefore 119 patients formed the ATP population (Fig. 1).
The ITT population comprised 107 (84.3%) female and 20 (15.7%) male patients. 17 (13.4%) patients were 20-29 years, 35 (27.6%) 30-39 years, 43 (33.9%) 40-49 years, 21 (16.5%) 50-59 years and 11(8.6%) 60-71 years old.
The mean duration of the current Generalized Anxiety Disorder was 40.3 [+ or -] 72.1 (median 15.8) months. About one third of subjects (35.4%) had suffered from 3 to 12 months, another third (30.7%) from one to two years. 20.5% had a history of anxiety of more than 3 years.
No anti-anxiety treatment prior to participation in the trial was reported for 79 (62.2%) patients. 18 (14.2%) patients had received drug treatment, 11 (8.7%) psychotherapy, and 19 patients (15.0%) a combination of both.
At screening the following 6 GAD symptoms out of 22 from the International Diagnostic Checklists (IDCL) (Hiller et al. 1993) were noted in more than 80% of patients: restlessness (n = 125; 98.4%), palpitations (n = 121; 95.3%), nervousness (n = 116; 91.3%), irritability (n = 109; 85.8%), insomnia (n = 105; 82.7%), loss of concentration (n = 103; 81.1%).
No significant differences between treatment groups could be observed with respect to baseline demographics and clinical characteristics of patients (Table 1).
Means of HAMA total scores decreased from about 23 at baseline to ca. 8 at week 8 with no significant differences between treatments (Fig. 2). Already at week 2 a substantial improvement of the HAMA total score could be seen to the same degree for Kava LI 150 (-6.8; 27.9%), Buspirone (-7.1; 29.7%), and Opipramol (-6.7; 27.9%), respectively. Treatment effects documented at week 8 were stable without medication through week 9 (follow-up) in all three groups.
Independent from the drug administered, about three fourths of patients responded well to the 8 week treatment regarding to the HAMA criterion, the figures for the BOEAS based response determination were a bit lower (Fig. 3).
On the basis of CGI subscale "improvement", 55.8% of patients were much or very much improved in the Kava group, 66.7% in the Buspirone and 71.4% in the Opipramol group after 8 weeks of treatment. 37.2% of Kava patients were rated to be "not ill" or "borderline ill" by means of CGI at that time-point (Buspirone: 42.9%; Opipramol: 42.9%).
Two thirds of patients were rated to be fully remitted after 8 weeks for all three medications, displaying a HAMA score below 9. BOEAS based remission figures again tended to be somewhat lower with a non-significant trend in favour of Opipramol.
Self rating of anxiety by means of SAS index scores decreased by 25 to 30% over 8 weeks: from 52.7 [+ or -] 6.5 to 39.7 [+ or -] 10.7 (Kava), 53.0 [+ or -] 7.4 to 37.5 [+ or -] 10.9 (Buspirone) and 54.4 [+ or -] 7.8 to 37.8 [+ or -] 9.5 (Opipramol).
14.6% (Buspirone), 21.3% (Kava) and 22.0% (Opipramol) were the mean improvements on the AL subscale "social life" as reported by patients with no significant differences between groups.
Bf-S sum scores decreased from baseline to week 8 from 34.8 [+ or -] 9.6 to 19.9 [+ or -] 12.0 (Kava), from 34.0 [+ or -] 10.7 to 18.8 [+ or -] 14.3 (Buspirone) and from 34.6 [+ or -] 9.6 to 17.2 [+ or -] 12.5 (Opipramol).
Sleep was impaired severely in all treatment groups as reflected by baseline scores around 2.5 in the sleep questionnaire SF-B. Improvements of clinical relevance could be seen with all treatments in a comparable manner. SF-B factor SQ (sleep quality) scores increased from 2.65 [+ or -] 0.68 to 3.60 [+ or -] 0.84 (Kava), 2.53 [+ or -] 0.67 to 3.68 [+ or -] 0.80 (Buspirone) and 2.66 [+ or -] 0.74 to 3.79 [+ or -] 0.82 (Opipramol).
No significant differences between treatments could be observed with regard to all primary and secondary efficacy variables (Table 2).
Except for one patient in the Opipramol group displaying depressive mood, no other symptoms of withdrawal or relapse were reported at week 9, one week after discontinuation of study medication.
Results from efficacy analyses in the ATP population closely corresponded with ITT results.
ANCOVA for HAMA total score at week 8 showed no centre effects.
A total number of 57 treatment emergent adverse events have been documented during 8 week treatment. A trend towards more adverse events under Kava medication is mostly due to some patients reporting a series of events. No systematic differences between treatments could be elicited with regard to frequency of single types of adverse events (Table 3).
One serious adverse event occurred in the Kava group. Three days after the week 4 visit, a patient suffered from a panic attack that required stationary treatment. Symptoms improved rapidly in the hospital and study medication was not discontinued.
Investigators rated tolerability to be "very good" or "good" in 37 cases (86.1%) for Kava and in 41 cases (97.6%) each for Buspirone and for Opipramol, respectively. Nearly the same figures were obtained by the patients' ratings (37/86.1%; 40/95.2%; 41/97.6%).
Of all laboratory values outside the normal range at week 8, only a gamma-GT increase in an Opipramol patient was rated to be of clinical relevance and was reported as an adverse event. Slight increases of transaminases to values above the upper limit of the normal range were documented for two Kava patients, three Buspiron patients and two Opipramol patients. Of these cases, one subject in the Kava group (Buspiron: three; Opipramol: one) had displayed transaminase values slightly above normal range already at baseline.
As far as we know, the data presented here are from the first clinical trial testing a Kava-Kava preparation in Generalized Anxiety Disorder. Furthermore it is the first trial investigating Kava in the daily dosage of 120 mg of kavapyrones in accordance with the recommendations of the German commission E monograph for Piperis methystici rhizoma (Bundesinstitut fur Arzneimittel und Medizinprodukte, 1990).
Somewhat surprisingly, study results show the same high response rates for Kava LI150 and for the two reference drugs. However, observed response rates coincide with some data from other clinical drug trials in GAD. For Venlafaxin, Sheehan (1999) reported a response rate of 69% from a 6-week trial, Allgulander et al. (2001) documented a rate between 50 and 70% after 8 weeks treatment and Davidson et al. (1999) saw only 49% responders after 8 weeks (Buspirone: 45%; Placebo: 36%).
Pollack et al. (2001) found a response rate after 8 weeks of Paroxetin treatment of 72.4% and a remission rate of 42%.
In these studies the mean duration of GAD ranged from 87 to 567 weeks--a figure comparable to the history of the patients in our study (40 months).
Only Allgulander et al. (2001) give information about pre-treatment of their GAD patients, reporting 74% of patients that had been treated for GAD at any time prior to entering the trial. This is a much higher proportion than in our study, where 62% of patients had never been treated for GAD before.
Severity of illness was somewhat lower in our sample, but still in the range of the other GAD trials mentioned above, as represented by a mean HAMA score of 23 at baseline. The HAMA cut-off for inclusion in our trial was only one point below the cut-off in the other trials (19 vs. 20).
Therefore, moderate severity of illness, absence of an inactive placebo control and especially a high proportion of previously untreated patients may have contributed substantially to the high response rates observed in our study.
The trial had been planned as a proof-of-concept study with a new Kava formulation containing the Kava dry extract LI150. Therefore no sample size estimation on a solid basis could be applied and more than 40 subjects per group are a generous dimension for such a type of study. Post-hoc analysis revealed that taken the actual number of randomized patients and the variation of the primary efficacy variable (HAMA; difference week 8--baseline), a one-sided t-test would have detected a group difference of 5 score points with a a of 5% and a type-II error probability of 20%. The observed HAMA standard deviations around 7 after 8 weeks of treatment represent a rather low figure as well as the unusual drop-out rate of only 6.3% (n = 8). Probably, the rater-training and the frequent on-site monitoring activities throughout the trial have significantly contributed to this finding.
Because in the Kava group 5 drop-outs occurred whereas in the Opipramol group not one patient withdrew, the main analysis was performed with the ITT population applying the LOCF procedure to ensure a more conservative approach avoiding any bias in favour of Kava. Further analysis of the Per-protocol population revealed no differences compared to the ITT results.
The lack of any significant difference of efficacy measures in terms of confidence intervals and results of Kruskal-Wallis tests (Table 2) can be judged as a strong hint for equivalence between groups although formally the premises for inference statistics were not given.
Critics may point to the missing placebo control questioning the internal validity of the study. Of course it is imaginable that even two standard treatments used as references may fail like the test drug and that the overall response would be in the range of placebo outcome. But presuming a rather high placebo response rate of 50% (Montgomery, 1999; Moller and Volz, 2001), the effect size difference still would be another 25% for all treatments in this study. Active drug--placebo differences in response rates amounted to a low of 17% in the Paroxetin trial (Pollack et al. 2001) and to 13% for Venlafaxin and 11% for Buspirone in the trial of Davidson et al. (1999). These differences are considered to be of clinical relevance by the authors of these trials.
The two active comparators used in this trial are very different anxiolytics in many matters. The selective 5-HT-1A receptor agonist Buspirone had been chosen as a reference drug as it still represents a scientific landmark in the development of anxiolytic substances. It was the first drug introduced for the treatment of GAD in the USA in 1986 and is frequently prescribed especially in Anglo-American countries. Interestingly, Buspirone plays almost no role in German primary care settings. In eight clinical studies, Buspirone proved to be effective in anxiety/GAD where the majority of patients received doses between 15 and 25 rug per day. HAMA total score reductions from 37 to 60% were observed. However, none of these trials lasted longer than 6 weeks. Initiation dosage of Buspirone is recommended to be 10-15 mg daily, the target therapeutic dosage should be 15-30 mg (Fulton and Brogden, 1997). Therefore, the 74% response rate of Buspirone 10 mg daily after 8 weeks treatment in our trial may overestimate the r eal pharmacodynamic effect to some extent and may be in part the result of the patients' positive attitude towards the study knowing they definitely would be administered an active drug in absence of a placebo group.
In contrary to Buspirone, Opipramol is the number one psychopharmacon in terms of numbers of prescriptions in Germany and is very popular especially with general practitioners (Lohse and Muller-Oerlinghausen, 2001). Classified as an antidepressant, it is widely used to treat a spectrum of anxiety and psychosomatic disorders although not being licensed for these indications (Volz and Moller, 1998). Only one pivotal 4-weeks clinical trial recently published demonstrated the efficacy of Opipramol 200 mg per day in GAD (Moller et al. 2001). In two other controlled trials, Opipramol was superior to placebo in somatoform disorders at 200 mg daily and in climacteric syndrome at 50 mg daily (Volz et al. 2000; van Lith and Motke, 1983). As no dose-response studies for Opipramol exist, dosing recommendations of the manufacturer result from a small number of older studies and from decades of clinical experience with this drug. For mild to moderate disease, a target dose of 50-150 mg daily is sought, 300 mg is denominat ed to be the maximum dose for severe states of illness (CIBA-GEIGY GmbH, 1994). The results reported here are in accordance with these recommendations.
The overall safety record from this study is excellent for all three treatments as reflected by the number and nature of adverse events as well as the global ratings of patients and investigators. On the first sight, Kava was burdened with a worse tolerability compared to Buspirone and Opipramol. However, this difference was statistically not significant and a closer look to the data qualifies this impression. The number of patients reporting adverse events were nearly the same for all three drugs. In the Kava group, more patients reported a series of adverse events than in the other groups. A detailed evaluation of all adverse events with regard to the causal relationship to the medication leads to the assumption that Kava may cause gastro-intestinal incompatibility in some patients. Other systematic side effects could not be observed. Only one patient in the Kava group dropped out due to an adverse event that seemed to be related to the medication (nausea, emesis). The only serious adverse event reported f rom this trial was a short hospitalisation of a Kava patient experiencing a panic attack as a symptom of a so far undetected comorbid panic disorder.
The safety data for Kava LI150 from this trial confirm Kava data from other clinical trials (Pittler and Erust, 2001) and from many years of clinical experience and traditional use (Bundesinstitut fur Arzneimittel und Medizinprodukte, 1990; Schulz et al. 2001).
During the last years an increasing number of reports of liver damage possibly related to Kava intake gained interest of the scientific community and of drug authorities in European countries (Strahi et al. 1998; Escher et al. 2001; Russmann et al. 2001). First investigations in a possible mechanism suggest an idiosyncratic immune-mediated process with a CYP2D6 deficiency as an additional risk factor (Russmann et al. 2001). However, data are too preliminary to allow any significant conclusions so far. It is also discussed, whether the manufacturing process of the two mainly used Kava preparations (acetonic and ethanolic extracts) may play a role in this context, e.g. by specifically enriching a toxic constituent or trigger substance (Hellwig, 2000).
No significant hepatotoxic reactions are to be reported from this trial in about 330 treatment weeks with Kava LI150.
Thus, this study adds further valuable evidence that patients suffering from GAD may benefit from an acute treatment with Kava. The growing basis of evidence for its clinical efficacy in anxiety disorders and, not to forget, the positive attitude of the vast majority of patients towards the herbal drug should be considered carefully in future benefit/risk assessments.
[FIGURE 1 OMITTED]
Fig. 2 Hamilton Anxiety Scale total score by week and treatment (ITT population, LOCF). Bars represent means [+ or -] SD. HAMA score Kava LI150 Buspiron Opipramol Screening 23.05 23.45 23.40 Week 0 23.14 23.55 23.93 Week 2 16.37 16.45 17.19 Week 4 11.84 11.52 11.24 Week 8 8.37 8.00 7.74 Week 9 8.16 7.76 6.81 Note: Table made from bar graph Fig. 3 Proportions of treatment responders (ITT population, LOCF); response to treatment was defined as a decrease of more than 50% of the Hamilton Anxiety Scale total score or of more than 50% of the Boerner Anxiety Scale total score, relative to baseline. Bars show total numbers and group wise percentage of treatment responders. No. of Patients Response Remission (HAMA red. (BOEAS red. (HAMA score (BOEAS score >50%) >50%) < 9) < 5) Kava LI150 33 (76.7%) 26 (60.5%) 28 (65.1%) 19 (44.2%) Buspiron 31 (73.8%) 26 (61.9%) 28 (66.7%) 22 (52.4%) Opipramol 32 (76.2%) 29 (69.0%) 28 (66.7%) 26 (61.9%) Note: Table made from bar graph Table 1 Baseline demographics and clinical characteristics of patients--95% confidence intervals of differences of means between treatment groups. Measure * Kava vs. Buspiron Kava vs. Opipramol (Baseline) (n = 43) (n = 42) (n = 43) Sex - - - Age (years) -4.45 5.05 -8.27 Height (cm) -3.37 3.02 -2.68 Weight (kg) -5.35 5.97 -8.89 Concomitant Diseases yes/no - - - Concomitant Therapies yes/no - - - No. of typicalAnxiety Symptoms -2.56 0.55 -1.17 Duration of Illness (months) -40.5 21.5 -42.6 Pre-existing Anti-Anxiety - - - Treatment yes/no HAMA (total score) -2.10 1.28 -2.37 BOEAS (total score) -1.83 1.64 -1.07 SAS (Index) -3.26 2.75 -4.81 Bf-S (total score) -3.55 5.23 -3.91 AL (total score) -11.4 6.7 -11.0 CGI (severity) -0.22 0.40 -0.13 Measure * Kava vs. Buspiron vs. Opipramol Opipramol (Baseline) (n = 42) (n = 42) (n = 42) Sex - - - Age (years) 1.19 -8.54 0.88 Height (cm) 4.00 -2.93 4.06 Weight (kg) 3.63 -9.49 3.61 Concomitant Diseases yes/no - - - Concomitant Therapies yes/no - - - No. of typicalAnxiety Symptoms 1.64 -0.26 2.73 Duration of Illness (months) 13.3 -39.9 29.5 Pre-existing Anti-Anxiety - - - Treatment yes/no HAMA (total score) 0.79 -1.98 1.22 BOEAS (total score) 2.40 -0.98 2.51 SAS (Index) 1.40 -4.75 1.85 Bf-S (total score) 4.40 -5.02 3.82 AL (total score) 7.7 -9.8 10.9 CGI (severity) 0.45 -0.24 0.38 Measure * p-value ** (Baseline) Sex 0.82 Age (years) 0.26 Height (cm) 0.93 Weight (kg) 0.64 Concomitant Diseases yes/no 0.59 Concomitant Therapies yes/no 0.34 No. of typicalAnxiety Symptoms 0.37 Duration of Illness (months) 0.96 Pre-existing Anti-Anxiety 0.82 Treatment yes/no HAMA (total score) 0.30 BOEAS (total score) 0.43 SAS (Index) 0.58 Bf-S (total score) 0.98 AL (total score) 0.68 CGI (severity) 0.45 * HAMA - Hamilton Anxiety Scale; BOEAS - Boerner Anxiety scale; SAS - Self-Rating Anxiety Scale; Bf-S - Self-Rating Scale for Well-Being according to von Zerssen; CGI, Clinical Global Impressions ** P-values of Kruskal-Wallis-Test, except for catergorical measures (Chi-Square-Test) Table 2 Summary of major efficacy measures--means [+ or -] SD at baseline and after 8 weeks of treatment (ITT population). Measure * Baseline Kava Buspiron HAMA (total score) 23.14 [+ or -] 3.19 23.55 [+ or -] 3.93 BOEAS (total score) 13.00 [+ or -] 4.03 13.10 [+ or -] 4.02 SAS (Index) 52.70 [+ or -] 6.54 52.95 [+ or -] 7.38 Bf-S (total score) 34.79 [+ or -] 9.64 33.95 [+ or -] 10.71 AL (total score) 140.14 [+ or -] 18.96 142.45 [+ or -] 22.82 SF-B (factor SQ) 2.65 [+ or -] 0.68 2.53 [+ or -] 0.67 (factor GES) 2.52 [+ or -] 0.70 2.40 [+ or -] 0.58 (factor PSYA) 2.66 [+ or -] 0.54 2.64 [+ or -] 0.59 (factor PSYE) 3.74 [+ or -] 0.61 3.97 [+ or -] 0.51 (factor PSS) 1.87 [+ or -] 0.63 2.01 [+ or -] 0.61 Measure * Baseline Week 8 Opipramol Kava HAMA (total score) 23.93 [+ or -] 3.42 8.37 [+ or -] 7.44 BOEAS (total score) 12.33 [+ or -] 4.02 5.79 [+ or -] 5.27 SAS (Index) 54.40 [+ or -] 7.81 39.67 [+ or -] 10.72 Bf-S (total score) 34.55 [+ or -] 9.63 19.86 [+ or -] 11.97 AL (total score) 141.79 [+ or -] 24.31 168.09 [+ or -] 24.75 SF-B (factor SQ) 2.66 [+ or -] 0.74 3.60 [+ or -] 0.84 (factor GES) 2.49 [+ or -] 0.71 3.41 [+ or -] 0.94 (factor PSYA) 2.70 [+ or -] 0.58 3.49 [+ or -] 0.75 (factor PSYE) 3.84 [+ or -] 0.58 3.30 [+ or -] 0.66 (factor PSS) 2.00 [+ or -] 0.65 1.46 [+ or -] 0.53 Measure * Week 8 Buspiron Opipramol HAMA (total score) 8.00 [+ or -] 7.56 7.74 [+ or -] 7.67 BOEAS (total score) 5.62 [+ or -] 5.40 5.12 [+ or -] 5.57 SAS (Index) 37.48 [+ or -] 10.88 37.79 [+ or -] 9.45 Bf-S (total score) 18.76 [+ or -] 14.31 17.19 [+ or -] 12.52 AL (total score) 170.55 [+ or -] 27.56 173.98 [+ or -] 23.26 SF-B (factor SQ) 3.68 [+ or -] 0.80 3.79 [+ or -] 0.82 (factor GES) 3.50 [+ or -] 0.98 3.49 [+ or -] 0.95 (factor PSYA) 3.74 [+ or -] 0.81 3.69 [+ or -] 0.70 (factor PSYE) 3.38 [+ or -] 0.74 3.44 [+ or -] 0.65 (factor PSS) 1.48 [+ or -] 0.49 1.43 [+ or -] 0.48 * HAMA - Hamilton Anxiety Scale BOEAS - Boerner Anxiety Scale SAS - Self-Rating Anxiety Scale Bf-S - Self-Rating Scale for Well-Being according to von Zerssen SF-B - Self-Rating Sleep Questionnaire ** P-values of Kruskal-Wallis-Test, except for "Response" (Chi-Square-Test) Table 3 Summary of efficacy measures - 95% confidence intervals of differences of means between treatment groups (ITT population). Measure * Kava vs. Buspiron Kava vs. Opipramol (Baseline) (n = 43) (n = 42) (n = 43) HAMA (total score) -3.99 2.43 -4.43 BOEAS (total score) -2.46 1.93 -1.83 SAS (Index) -7.61 2.70 -8.21 Bf-S (total score) -7.37 6.85 -8.55 AL (total score) -12.5 12.3 -16.1 SF-B (factor SQ) -0.69 0.28 -0.64 (factor GES) -0.70 0.29 -0.56 (factor PSYA) -0.70 0.16 -0.54 (factor PSYE) -0.48 0.19 -0.28 (factor PSS) -0.32 0.18 -0.38 (Week 8) CGI (severity) -0.45 0.65 -0.23 CGI (improvement) -0.43 0.56 -0.14 Global judgment of efficacy -0.42 0.39 -0.12 (investigator) Global judgment of efficacy -0.52 0.31 -0.30 (patient) Response yes/no (HAMA reduction - - - > 50%) Remission yes/no (HAMA <9) - - - Measure * Kava vs. Buspiron vs. Opipramol Opipramol (Baseline) (n = 42) (n = 42) (n = 42) HAMA (total score) 1.58 -3.91 2.62 BOEAS (total score) 1.82 -1.96 2.48 SAS (Index) 1.02 -6.39 4.11 Bf-S (total score) 3.70 -8.94 4.61 AL (total score) 7.6 -17.8 9.6 SF-B (factor SQ) 0.27 -0.47 0.52 (factor GES) 0.34 -0.38 0.56 (factor PSYA) 0.24 -0.30 0.53 (factor PSYE) 0.36 -0.14 0.52 (factor PSS) 0.10 -0.31 0.17 (Week 8) CGI (severity) 0.86 -0.34 0.77 CGI (improvement) 0.80 -0.23 0.75 Global judgment of efficacy 0.62 -0.16 0.69 (investigator) Global judgment of efficacy 0.47 -0.23 0.61 (patient) Response yes/no (HAMA reduction - - - > 50%) Remission yes/no (HAMA <9) - - - Measure * p-value ** (Baseline) HAMA (total score) 0.49 BOEAS (total score) 0.98 SAS (Index) 0.29 Bf-S (total score) 0.71 AL (total score) 0.83 SF-B (factor SQ) 0.50 (factor GES) 0.68 (factor PSYA) 0.48 (factor PSYE) 0.74 (factor PSS) 0.43 (Week 8) CGI (severity) 0.57 CGI (improvement) 0.34 Global judgment of efficacy 0.26 (investigator) Global judgment of efficacy 0.64 (patient) Response yes/no (HAMA reduction 0.95 > 50%) Remission yes/no (HAMA <9) 0.98 * HAMA - Hamilton Anxiety Scale; BOEAS - Boerner Anxiety Scale; SAS - Self-Rating Anxiety Scale; Bf-S - Self-Rating Scale for Well-Being according to von Zerseen; SF-B - Self-Rating Sleep Questionnaire; CGI - Clinical Global Impressions ** P-values of Kruskal-Wallis-Test, except for "Response" (Chi-Square-Test) Table 4 Frequencies of treatment emergent adverse events * during 8-week treatment phase by treatment group. Kava Kava Buspiron Opipramol (n=43) (n=42) (n=42) No. of Adverse Events 27 16 14 No. of Patients reporting Adverse 14 (33%) 14 (24%) 11 (26%) Events (%) No. of Adverse Events rated to be 1 3 1 "probably" related to Medication Nos. of specific Adverse Events Common Cold, Pharyngitis, 5 3 3 Bronchitis Nausea/Emesis 3 1 - Diarrhea 1 1 1 Other Gastro-intenstinal disorders 1 2 2 Weight Changes 2 1 4 Skin Affections 2 3 - Tachycardia 2 - - Sedation - 1 - Others 10 4 4 * Adverse Events that have not been pre-existing at screening or baseline visit
This study was funded by a grant from Lichtwer Pharma AG, Berlin, Germany.
We acknowledge the commitment of the investigators and the staffs of the clinical study sites: Aulich H, Bringmann W, Eyber P, Franz P, Guenter A, Heise M, Kleiber A, Larisch S, Mueller E, Reiss-Schreiber T, Ribbschlager M (all Berlin); Deters M, Halama P, Reinhard-Feyerabend A (Hamburg); Staehr B (Falkensee); Koppl G (Flensburg); Sommer H (Kothen); Schnabel M (Leipzig); Rogge J (Perleberg).
And of course we are indebted to the 138 patients who participated in the study.
Allgulander C, Hackett D, Salinas E (2001) Venlafaxine extended release (ER) in the treatment of generalized anxiety disorder. Twenty-four-week placebo-controlled dose-ranging study. Br J Psychiatry 179: 15-22
Anonymous (1998) Buspirone HCI. In: Drug facts and comparisons, 1998 edition: 1593-1596. Facts and Comparisons, St. Louis, MO
Boerner RJ (2001) Kava kava in the treatment of generalized anxiety disorder, simple phobia and specific social phobia. Phytother Res 15: 646-647
Boerner RJ (2002a) Angststorungen. In: Neuro-Psychopharmaka, vol. 3, 2nd ed. (eds Riederer P et.al.). Springer, Wien (in press)
Boerner RJ (2002b) The Boerner Anxiety Scale (BOEAS), a new scale for measurement of anxiety symptoms (submitted)
Brawman-Mintzer O, Lydiard RB (1996) Generalized anxiety disorder: issues in epidemiology. J Clin Psychiatry 57(suppl 7): 3-8
Buckley JP, Furgiuele AR, O'Hara MJ (1967) Pharmacology of Kava -2. Psychopharmacol Bull 4:11-12
Bullinger M, Kirchberger I, von Steinbuchel N (1993) Der Fragebogen Alltagsleben--em Verfahren zur Erfassung der gesundheitsbezogenen Lebensqualitat. Zeitschrift fur Medizinische Psychologie 2:121-131
Bundesinstitut fur Arzneimittel und Medizinprodukte (1990) Monographie: Piperis methystici rhizoma (Kava-Kava Wurzelstock). [Federal Office for Medicines and Medicinal Products. Monograph: Kava-Kava]. BAnz 101 (01 Jun 1990)
CIBA-GEIGY GmbH (1994) Insidon[R]. Drug Information Sheet, October 1994
Davidson JR, DuPont RL, Hedges D, Haskins JT (1999) Efficacy, safety, and tolerability of venlafaxine extended release and buspirone in outpatients with generalized anxiety disorder. J Clin Psychiatry 60: 528-535
Escher M, Desmeules J, Giostra E, Mentha G (2001) Hepatitis associated with Kava, a herbal remedy for anxiety. BMJ 322:139
Fulton B, Brogden RN (1997) Buspirone: an updated review of its clinical pharmacology and therapeutic applications. CNS Drugs 7: 68-88
Gortelmeyer R (1985) On the Development of a Standardized Sleep Inventory for the Assessment of Sleep. In: Methods of Sleep Research (eds Kubicki S, Hermann WM): 93-98. Gustav Fischer, Stuttgart
Hellwig B (2000) In der Diskussion: Kava-Kava--Schadliche Stoffe im Aceton-Extrakt? Dtsch Apoth Ztg 29: 3361
Hiller W, Zaudig M, Mombour W, Bronisch T (1993) Routine psychiatric examinations guided by ICD-10 Diagnostic Checklists (International Diagnostic Checklists). Eur Arch Psychiatry Clin Neurosci 242: 218-223
Katon W, Von Korff M, Lin E, et al. (1990) Distressed high utilizers of medical care: DSM-III-R diagnoses and treatment needs Gen Hosp Psychiatry 12: 355-362
Kessler RC, McGonagle KA, Zhao S, Nelson CB, Hughes M, Eshleman S, Wittchen HU, Kendler KS (1994) Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States. Results from the National Gomorbidity Survey. Arch Gen Psychiatry 51: 8-19
Lohse MJ, Mueller-Oerlinghausen B (2001) Psychopharmaka. In: Arzneiverordnungs-Report 2001 (eds Schwabe U, Paifrath D): 597-629. Springer, Berlin, Heidelberg
Malsch U, Kieser M (2001) Efficacy of kava-kava in the treatment of non-psychotic anxiety, following pretreatment with benzodiazepines. Psychopharmacology (Berl) 157: 277-283
Meyer HJ (1962) Pharmakologie der wirksamen Prinzipien des Kawa-Rhizoms (Piper methysticum Forst.). Arch Int Pharmacodyn Ther 88: 505-536
Meyer HJ (1967) Pharmacology of Kava--1. Psychopharmacol Bull 4: 10-11
Moller HJ, Volz HP, Reimana IW, Stoll KD (2001) Opipramol for the treatment of generalized anxiety disorder: a placebo-controlled trial including an alprazolamtreated group. J Clin Psychopharmacol 21: 59-65
Montgomery SA (1999) Alternatives to placebo-controlled trials in psychiatry. ECNP Consensus Meeting, September 26, 1996, Amsterdam. European College of Neuropsychopharmacology. Eur Neuropsychopharmacol 9: 265-269
Pittler MH, Ernst E (2001) Kava extract for treating anxiety (Cochrane Review). Cochrane Database Syst Rev 4: CD003383
Pollack MH, Zaninelli R, Goddard A, McCafferty JP, Bellew KM, Burnham DB, Iyengar MK (2001) Paroxetine in the treatment of generalized anxiety disorder: Results of a placebo-controlled, flexible-dosage trial. J Clin Psychiatry 62: 350-357
Rickels K, Downing R, Schweizer E, et al. (1993) Antidepressants for the treatment of generalized anxiety disorder, A placebo-controlled comparison of imipramine, trazodone, and diazepam. Arch Gen Psychiatry 50: 884-895
Rickels K, Pollack MH, Sheehan DV, Haskins IT (2000) Efficacy of extended-release venlafaxine in nondepressed outpatients with generalized anxienty disorder. Am I Psychiatry 157: 968-974
Russmann S, Lauterburg BH, Helbling A (2001) Kava hepatotoxicity. Ann Intern Med 135: 68-69
Schulz V, Haensel R, Tyler VE (2001) Rational Phytotherapy. A Physicians' Guide to Herbal Medicine (4th edition): 78-86. Springer, Berlin, Heidelberg
Schweitzer E, Rickels K (1997) Strategies for treatment of generalized anxiety in the primary care setting. J Glin Psychiatry 58(suppl 3): 27-31
Shader RJ, Greenblatt DJ (1993) Use of benzodiazepines in anxiety disorders. N Engl J Med 328: 1398-1405
Sheehan DV (1999) Venlafaxine extended release (XR) in the treatment of generalized anxiety disorder. I Clin Psychiatry 60(suppl 22): 23-28
Singh YN (1992) Kava: An overview. I Ethnopharmacol 37: 13-45
Steiner GG (2000) The correlation between cancer incidence and kava consumption. Hawaii Med 159: 420-422
Strahl S, Ehret V, Dahm HH, Maier KP (1998) Nekrotisierende Hepatitis nach Einnahme pflanzlicher Heilmittel. Dtsch Med Wochenschr 123: 1410-1414 van Lith ND, Motke JC (1983) Opipramol in the climacteric syndrome. A double-blind, placebo-controlled trial. Maturitas 5:17-23
Volz HP, Moller HJ (1998) Opipramol bei Angst- und Somatisierungsstudien. Fortschr Neurol Psychiat 66 (suppi 1): S21-S24
Volz HP, Moller HJ, Reimann I, Stoll KD (2000) Opipramol for the treatment of somatoform disorders, results from a placebo-controlled trial. Eur Neuropsychopharmacol 10:211-217
von Zerssen D (1973) Beschwerdenskalen bei Depressionen. Therapiewoche 46: 4426-4440
World Health Organisation (1992) The ICD-10 Classification of Mental and Behavioural Disorders. Clinical Descriptions and Diagnostic Guidelines. WHO, Geneve
Yonkers KA, Warshaw MG, Massion AO, Keller MB (1996) Phenomenology and course of generalised anxiety disorder. BrJ Psychiatry 168: 308-313
R.J. Boemer (1)
H. Sommer (2)
W. Berger (3)
U. Kuhn (3)
U. Schmidt (3)
M. Mannel (4)
(1.) Department of Psychiatry, Ludwig-Maximilians-University, Munich, Germany
(2.) Practice for Psychiatry and Neurology, Kothen, Germany
(3.) IMF - Institute for Interdisciplinary Medical Research, Cologne, Germany
(4.) Lichtwer Pharma AG, Berlin, Germany
Reinhard Boerner, Department of Psychiatry, LudwigMaximilians-University, Nussbaumstrasse 7, D-80336 Munich, Germany Tel.: ++49-89-51603370; Fax: ++49-89-51605343; e-mail: firstname.lastname@example.org