Subscribe to RSS
DOI: 10.1055/s-2007-984400
© Georg Thieme Verlag KG Stuttgart · New York
Dose-response Relationship of Pregabalin in Patients with Generalized Anxiety Disorder. A Pooled Analysis of Four Placebo-controlled Trials
Correspondence
Prof. P. BechMD
Psychiatric Research Unit
Frederiksborg General Hospital
48 Dyrehavevej
3400 Hillerød
Denmark
Phone: +45/4829 32 53
Fax: +45/4826 33 77
Email: pebe@noh.regionh.dk
Publication History
received 10.09.2006
revised 02.04.2007
accepted 09.05.2007
Publication Date:
10 August 2007 (online)
Abstract
Background: Pregabalin has been evaluated in randomised clinical trials in patients with generalised anxiety disorder (GAD) in a fixed-dose design and with the Hamilton Anxiety Scale (HAM-A) as outcome measure. Four of the available six placebo-controlled trials were found acceptable for a pooled analysis of dose-response relationship.
Method: Both the full HAM-A14 and the six-item subscale covering the core items of GAD (HAM-A6) were analysed. The unbiased effect size statistic was used to evaluate the advantage of pregabalin over placebo. An effect size of 0.40 or higher was used to indicate a clinically significant effect.
Results: Four placebo-controlled trials running over four weeks and covering the dose range from 150 mg to 600 mg pregabalin were sufficiently homogeneous to be pooled for the analysis. Both HAM-A6 and HAM-A14 showed that for the dose of 150 mg pregabalin daily the effect size was clearly below 0.40. For the dose range of 200-450 mg daily, the effect sizes exceeded 0.40, with a plateau-like curve. The maximum dose of 600 mg daily did not increase effect size. On the HAM-A14 as well as the item of sleep, effect size was generally higher, but followed the same pattern as the HAM-A6.
Discussion: The dose of 150 mg pregabalin over the four weeks of the trials was found insufficient for the treatment of GAD. In the dose range of 200-450 mg daily, a clinically significant effect was obtained, although with a plateau-like curve which was not increased for the maximum dose of 600 mg daily.
#Introduction
The Hamilton Anxiety Scale (HAM-A) [22] [23] is the internationally most widely used clinician-rated scale for the evaluation of treatment outcome in patients with generalised anxiety disorder [7] [33]. However, from the moment of its first use in a clinical trial Hamilton [24] recognised the difficulty in relying on the HAM-A total score as the sole measure of the severity of the patient's condition. A drug might demonstrate superiority to placebo on the HAM-A total score while not achieving efficacy on individual items which are central to the diagnosis. Alternatively, patients might report superiority to placebo on individual key items, but this may not be reflected in the HAM-A total score.
In a subsequent factor-analytical study of the HAM-A14, Hamilton [23] confirmed the multi-dimensionality of his scale by identifying two factors: a psychic anxiety factor, and a somatic anxiety factor. The usefulness of these two factors for discriminating between differential treatment responses was shown by a double-blind, placebo-controlled trial comparing diazepam to imipramine in patients with a DSM-III diagnosis of generalised anxiety disorder [2] [41]. Imipramine but not diazepam was superior to placebo on the HAM-A psychic factor, while both imipramine and diazepam were superior to placebo on the HAM-A somatic factor.
A recent evidence-based review on the pharmacotherapy of generalised anxiety disorder based on the HAM-A14 total scores has concluded that the selective serotonin reuptake inhibitors (SSRIs) should be considered as first-line treatment, with the serotonin-noradrenaline reuptake inhibitor venlafaxine as a reasonable alternative [7]. This is also the most recent consensus statement in guidelines for the treatment of patients with GAD [8]. Dose-response trials with SSRIs in patients with major depression have shown that an effect size of approximately 0.40 or higher, based on the total score of the Hamilton Depression Scale, signifies a clinically significant antidepressant effect [14]. In a placebo-controlled trial with the SSRI sertraline, Fabre et al. [19] showed that 50 mg daily over six weeks was the lowest effective dose. The effect size was calculated to be 0.38 [10]. Allgulander et al. [1], using the total score of HAM-A14, recently showed that also after four weeks of therapy a fixed dose of 50 mg sertraline in patients with generalised anxiety disorder resulted in an effect size of 0.38.
Because the HAM-D, like the HAM-A, is a multi-dimensional scale [6], we have advocated the use of the HAM-D subscale covering the six core items of depression (HAM-D6) to calculate the effect size of antidepressants in patients with major depression [12].
Analogously, we have derived a 6-item HAM-A subscale (HAM-A6) comprising five psychic anxiety symptoms: anxious mood, psychic tension, fears, intellectual disturbances, and anxious behaviour observed at the interview. Only one somatic item is contained in the HAM-A6 subscale, namely muscular tension. This is in accordance with the study by Snaith et al. [46] when capturing the core items of the DSM-III diagnosis of generalised anxiety disorder in relation to the HAM-A. The HAM-A6 has previously been found to be uni-dimensional [9] [12].
In the study by Meoni et al. [27] it was shown that the HAM-A6 symptoms were among the DSM-IV criteria GAD symptoms that most significantly discriminated between venlafaxine and placebo. The venlafaxine versus placebo difference was only modest on the HAM-A sleep item. Since sleep is a DSM-IV criterion symptom which is not included in the HAM-A6 we have focussed on it as a separate item in our dose-response analysis of pregabalin.
The objective of the current analysis was to test whether the uni-dimensional HAM-A6 reveals a significant dose-response relationship for pregabalin in clinical trials in patients with a DSM-IV diagnosis of generalised anxiety disorder [3]. We used an effect size of 0.40 or more to indicate a clinically significant anxiolytic effect. We pooled data from randomised, placebo-controlled clinical trials of pregabalin in the treatment of adult patients with a DSM-IV diagnosis of generalised anxiety disorder. Pregabalin is a novel compound structurally similar to gabapentin. In animal studies pregabalin has shown a broad spectrum of anxiolytic properties [21] and has now been evaluated in randomised clinical trials in patients with generalised anxiety disorder [20] [31] [35] [37] [42]. The pregabalin database is especially suited to the current analysis because the pivotal trials used a fixed-dose design.
#Methods
The protocol for this pooled analysis of placebo-controlled trials with pregabalin focused on dose-response analysis. In general, the selection of trials was based on the principles recommended by Peto [36], Bulpitt [17], and Schmidt et al. [43]. Because the optimal dose for achieving an anti-anxiety effect of the SSRI remains a problem due to the relatively high placebo effect [34], only trials in which 70% or more of the patients have completed four weeks of therapy were included to ensure an acceptable quality of the results for the dose-response analysis [4] [28]. To ensure a homogeneous and unbiased effect size analysis, US trials and European trials have been analysed separately [5] [11] [32]. Only trials in which co-morbid major depressive episodes (DSM-IV) were excluded have been considered, and only patients within the age range from 18 to 65 years have been included.
Only one trial was a European trial [31] and was excluded from the dose-response analysis, because the transatlantic difference as identified for depression [11] [32] was operating in the pregabalin trials, in that the baseline HAM-A14 score in this European trial was significantly higher than in the US trial (27.4 versus 24.5, P≤0.01). Furthermore, the patient age in the European trial [31] [32] was significantly higher than in the US trials (44.0±12 versus 37.2±10; P≤0.01). However, this European trial was analysed individually to evaluate the pregabalin versus venlafaxine effect in patients with generalised anxiety disorder.
Among the five US trials, only one was excluded from the dose-response analysis due to the percentage of patients who had completed four weeks of therapy was clearly below 70%[4], both concerning pregabalin 600 mg (66%) and lorazepam 6 mg (46%). This study is still unpublished, but has been analysed individually to show both the pregabalin effect and the effect of lorazepam in generalised anxiety disorder.
[Table 1] shows the four pregabalin trials selected for the dose-response analysis. The study period was four weeks for all except one [38], which had a planned duration of six weeks. In the pooled analysis, this trial has therefore been analysed after four weeks of therapy. Thus, in the present analysis focussing on dose-response relationship, we used the results of the first four weeks of this study.
Treatment (Number of patients) |
Pande et al. [35] |
Feltner et al. [20] |
Rickels et al. [42] |
Pohl et al. [38] |
pregabalin |
150 mg: 70 |
150 mg: 70 |
300 mg: 91 |
200 mg: 78 |
600 mg: 70 |
600 mg: 66 |
450 mg: 91 |
400 mg: 89 | |
- |
- |
600 mg: 90 |
450 mg: 88 | |
placebo |
69 |
67 |
89 |
86 |
benzodiazepines |
lorazepam, 6 mg: 68 |
lorazepam, 6 mg: 68 |
alprazolam, 93 | |
Completers (% of patients) | ||||
pregabalin |
150 mg: 90% |
150 mg: 70% |
300 mg: 89% |
200 mg: 71% |
600 mg: 70% |
600 mg: 70% |
450 mg: 80% |
400 mg: 72% | |
- |
- |
600 mg: 74% |
450 mg: 75% | |
placebo |
70% |
70% |
71% |
71% |
benzodiazepines |
lorazepam, 59% |
lorazepam, 53% |
alprazolam, 73% | |
HAM-A14 at baseline Mean (s. d.) | ||||
pregabalin |
150 mg: 23.3 (2.7) |
150 mg: 24.9 (3.9) |
300 mg: 24.9 (3.7) |
200 mg: 26.4 (4.8) |
600 mg: 23.2 (2.8) |
600 mg: 25.4 (4.6) |
450 mg: 24.5 (3.4) |
400 mg: 25.9 (5.0) | |
- |
- |
600 mg: 25.0 (3.5) |
450 mg: 25.5 (4.5) | |
placebo |
22.9 (3.4) |
24.8 (4.1) |
24.5 (3.6) |
25.5 (4.0) |
benzodiazepines |
lorazepam, 23.9 (3.2) |
lorazepam, 24.7 (3.7) |
alprazolam, 25.0 (3.2) | |
HAM-A6 at baseline Mean (s. d.) | ||||
pregabalin |
150 mg:12.0 (1.6) |
150 mg:12.3 (2.2) |
300 mg: 12.8 (1.8) |
200 mg: 13.2 (2.2) |
600 mg: 12.0 (1.9) |
600 mg: 12.3 (2.3) |
450 mg: 12.2 (1.8) |
400 mg: 12.9 (2.3) | |
- |
- |
600 mg: 12.5 (1.9) |
450 mg: 13.0 (2.5) | |
placebo |
12.0 (1.7) |
12.1 (2.2) |
12.5 (1.6) |
13.0 (2.1) |
benzodiazepines |
lorazepam 12.4 (1.3) |
lorazepam 12.2 (2.0) |
alprazolam 12.5 (1.8) | |
HAM-A6 sleep (item 4) at baseline Mean (s. d.) | ||||
pregabalin |
150 mg: 2.1 (0.8) |
150 mg: 2.5 (0.8) |
300 mg: 2.1 (0.9) |
200 mg: 2.3 (0.9) |
600 mg: 2.0 (0.8) |
600 mg: 2.5 (0.8) |
450 mg: 2.2 (0.8) |
400 mg: 2.2 (1.0) | |
600 mg: 2.4 (0.7) |
450 mg: 2.3 (0.9) | |||
placebo |
2.0 (0.8) |
2.3 (0.8) |
2.3 (0.9) |
2.3 (0.9) |
benzodiazepines |
lorazepam 2.0 (0.8) |
lorazepam 2.5 (0.8) |
alprazolam 2.2 (0.8) |
At screening, all current psychotropic medications were withdrawn. Zolpidem was the only hypnotic allowed during the trial on an as-needed basis for extreme sleeplessness. However, less than 5% of the patients in the trials shown in [Table 1] intermittently received zolpidem 10 mg daily.
Eligible patients were randomised to treatment with either pregabalin or placebo (or, in some trials, to an active comparator). Lists for randomisation were centrally prepared, using random number tables and the randomisation was blinded for the investigator.
The dosage was reached after a one-week titration, and remained fixed until the end of the study. The study medication was then tapered down during a follow-up one week period.
#Clinical outcome scales
The Hamilton Anxiety Rating Scale (HAM-A) [22] was used in all trials as the primary efficacy parameter. The inclusion criterion was a score of 20 or more on the total HAM-A14 scale. In the present analysis on dose-response relationship both the HAM-A14, the HAM-A6 and the HAM-A sleep item have been considered. The HAM-A6 includes the following items: anxious mood, psychic tension, fears, intellectual disturbances, muscular tension, and behaviour at interview [9] [12]. The theoretical score range of HAM-A14 is 0-56, and the range of HAM-A6 is 0-24.
#Statistical analysis
#The pooled-analysis method
In all individual trials, an intent-to-treat, last observation carried forward analysis was performed, and four weeks of assessments were used. The effect size was defined in accordance with Hedges and Olkin [25] as the mean change in scores from baseline to endpoint for the placebo-treated patients divided by the pooled standard deviation of the two treatment groups. It was the unbiased effect size that was calculated (25)[1], i.e., including an estimator in cases of small samples. However, because the mean number of patients in the various treatment groups was approximately 80, the sample size was not small. A positive effect size signifies that these results favour pregabalin.
After 4 weeks of treatment an effect size of 0.40 is of clinical significance, i.e., it equals a pregabalin advantage over placebo of around 20% expressed in terms of response rate [1] [9] [14] [45].
#The Mokken scale analysis
The HAM-A14 as well as the HAM-A6 were psychometrically analysed for uni-dimensionality by means of the Mokken scale analysis which is based on the Loevinger coefficient of homogeneity [14] [29] [30]. This analysis was made on data after 4 weeks of therapy, last observation carried forward, because the dispersion of scores for a correlation effect is greatest at this time of treatment [26]. A Loevinger coefficient of 0.40 or higher is acceptable for showing that the total score is a sufficient statistic, i.e., the scale is unidimensional [29].
#Results
[Table 2A] shows the Mokken analysis with the Loevinger coefficients for both the HAM-A14 and HAM-A6 after 4 weeks of treatment for the four selected dose-response trials, while [Table 2B] shows the two excluded trials. The results confirmed that the HAM-A14 is not a uni-dimensional scale (Loevinger coefficients lower than 0.40). In contrast, the Mokken analysis of the HAM-A6 yielded Loevinger coefficients above 0.40 for the four dose-response trials, individually and combined ([Table 2A]). This pattern was also seen when all six trials were included in the analysis ([Table 2B]).
Dose-response trials |
Loevinger homogeneity coefficient | ||
HAM-A14 |
HAM-A6 | ||
1 |
Pande et al. [35] |
0.32 |
0.45 |
2 |
Feltner et al. [20] |
0.36 |
0.46 |
3 |
Rickels et al. [42] |
0.39 |
0.51 |
4 |
Pohl et al. [38] |
0.34 |
0.44 |
all four studies combined |
0.35 |
0.46 |
Non-dose-response trials |
Loevinger homogeneity coefficient | |
HAM-A14 |
HAM-A6 | |
5 Unpublished |
0.38 |
0.50 |
6 Montgomery et al. [31] |
0.42 |
0.49 |
all six studies |
0.38 |
0.49 |
[Table 3A] shows the effect sizes after four weeks of treatment (last observation carried forward) for the selected four trials, while [Table 3B] shows the effect sizes in the two excluded trials. On the HAM-A6, the effect size for the lowest dose of pregabalin (150 mg), based on the combined data from two studies [20] [35] was clearly below 0.40, namely only 0.20 ([Table 3A]). For the intermediate pregabalin daily dose range of 200-450 mg, the effect size on HAM-A6 was above 0.40 in all five individual treatment arms, with a mean combined effect size of 0.49 ([Table 3A]). For the maximum dose of 600 mg, the mean HAM-A6 effect size, based on the three trials was 0.41, with a range from 0.36 to 0.50 ([Table 3A]).
Pregabalin dose |
Authors |
Effect size |
Effect size |
Combined dose-levels | |||||
individual trials |
Combined trials | ||||||||
HAM-ASleep |
HAM-A6 |
HAM-A14 |
HAM-ASleep |
HAM-A6 |
HAM-A14 | ||||
150 mg |
Pande et al. [35] |
0.38 |
0.22 |
0.38 |
0.24 |
0.20 |
0.31 |
150 mg | |
150 mg |
Feltner et al. [20] |
0.09 |
0.17 |
0.24 |
- |
- |
- |
- | |
200 mg |
Pohl et al. [38] |
0.59 |
0.48 |
0.63 |
- |
- |
- |
- | |
300 mg |
Rickels et al. [42] |
0.52 |
0.46 |
0.52 |
- |
- |
- |
- | |
400 mg |
Pohl et al. [38] |
0.61 |
0.51 |
0.62 |
0.59 |
0.49 |
0.56 |
200 mg to 450 mg | |
450 mg |
Pohl et al. [38] |
0.74 |
0.55 |
0.68 |
- |
- |
- |
- | |
Rickels et al. [42] |
0.47 |
0.44 |
0.37 |
- |
- |
- |
- | ||
600 mg |
Pande et al. [35] |
0.54 |
0.36 |
0.54 |
- |
- |
- | ||
600 mg |
Feltner et al. [20] |
0.45 |
0.50 |
0.58 |
0.51 |
0.41 |
0.53 |
600 mg | |
600 mg |
Rickels et al. [42] |
0.54 |
0.36 |
0.48 |
- |
- |
- |
- |
Authors |
Treatment |
Effect size | ||
Sleep |
Ham-A6 |
HAM-A14 | ||
unpublished |
pregabalin 150 mg (N=71) |
0.37 |
-0.16 |
0.19 |
pregabalin 600 mg (N=71) |
0.35 |
0.21 |
0.19 | |
placebo (N=70) | ||||
Montgomery et al. [31] |
pregabalin 400 mg (N=97) |
0.65 |
0.30 |
0.38 |
pregabalin 600 mg (N=110) |
0.54 |
0.28 |
0.31 | |
venlafaxine 75 mg (N=113) |
0.33 |
0.40 |
0.31 | |
placebo (N=101) |
On the HAM-A14, the effect size demonstrated the same pattern, but was generally numerically higher than on the HAM-A6. Thus, the mean effect size for 150 mg pregabalin was 0.31, for the intermediate dose of 200-450 mg it was 0.56, and for the maximum dose of 600 mg it was 0.53, with a range from 0.48 to 0.58. For the intermediate daily dose of 200-450 mg pregabalin, the effect sizes for both the HAM-A6 and HAM-A14 scales were similar to those achieved by the 600 mg daily dose of pregabalin, i.e., 0.41 and 0.46, respectively ([Table 3A]).
[Table 3B] shows the effect sizes for pregabalin in the two excluded trials. In the unpublished study, no effect size reached the level of clinical significance of 0.40 on any of the outcome measures (HAM-A14, HAM-A6 or sleep). In the Montgomery et al. study [31], the HAM-A6 effect size was 0.40 for venlafaxine, but not for pregabalin. However the difference between venlafaxine and pregabalin was not statistically significant. On the HAM-A14, the effect size was below 0.40 for both pregabalin and venlafaxine. In contrast, the effect size for sleep was greater than 0.40 for both doses of pregabalin, but below 0.40 for venlafaxine.
[Table 4] shows the effect sizes for the benzodiazepines (lorazepam 6 mg daily or alprazolam 1.5 mg daily) after 4 weeks of therapy. When all three lorazepam trials were combined, the effect size pattern on the HAM-A14, HAM-A6 and sleep was rather similar to the one obtained for alprazolam, i.e., only the sleep effect size reached the 0.40 level ([Table 4]).
Authors |
Treatment |
Effect size | ||
Sleep |
HAM-A6 |
HAM-A14 | ||
Pande et al. [35] |
lorazepam 6 mg (N=68) |
0.54 |
0.65 |
0.81 |
Feltner et al. [20] |
lorazepam 6 mg (N=68) |
0.34 |
0.22 |
0.35 |
Unpublished |
lorazepam 6 mg (N=70) |
0.35 |
0.01 |
-0.03 |
Total |
lorazepam 6 mg (N=206) |
0.41 |
0.29 |
0.38 |
Rickels et al. [42] |
alprazolam 7.5 mg (N=93) |
0.46 |
0.36 |
0.35 |
Discussion
In this pooled analysis of four randomised, double-blind, placebo-controlled, fixed-dose trials, pregabalin was found to exhibit a dose-response relationship at the lower end of the therapeutic dose range. This dose-response effect was more clearly seen on the HAM-A6 than on the HAM-A14. At a daily dose of 150 mg, pregabalin had an effect size below 0.40 on both the HAM-A6 (0.20), the HAM-A14 (0.31) and on the sleep item (0.24). In contrast, the effect sizes for daily doses of pregabalin in the range of 200-600 mg were above 0.40 on the HAM-A6, the HAM-A14, and on the sleep item. No dose-response relationship was observable within the 200-600 mg dosing range. An effect size above 0.40 is considered to be clinically significant [10] [14].
In a meta-analytical review of the drug treatment in GAD [28], the effect size statistic was calculated for all the included trials to compare the anti-anxiety effect, predominantly on the HAM-A14. Although the effect sizes in the meta-analytical review [28] were based on the principle published by Hedges and Olkin [25], an estimator of r=0.8 was used when the change scores in some of the trials are insufficiently reported [28] with reference to Ray and Shadish [40]. Mitte et al. [28] identified only one trial in which a selective serotonin reuptake inhibitor (SSRI) was compared to placebo, namely the study by Pollack et al. [39]. Because of insufficient information about the pooled standard deviation, Mitte et al. [28] obtained an effect size of 0.20. However, the more detailed data from the Pollack study have recently been published [18], which after unbiased effect size calculation [25] showed an effect size of 0.26 for the HAM-A14. However, in the placebo-controlled sertraline study in GAD, an effect size of 0.38 was obtained on the HAM-A14 [1].
Taking this uncertainty in the meta-analytical review by Mitte et al. [28] into consideration, the mean effect size for diazepam in ten placebo-controlled GAD trials was found to be 0.36. For venlafaxine, the mean effect size in 5 placebo-controlled trials was found to be 0.39, which is rather similar to what is seen for the venlafaxine arm of the study on by Montgomery et al. [31] on the HAM-A6 after 4 weeks of therapy. However, we demonstrated that venlafaxine had an insufficient effect on sleep, implying that a small dose of a sedative drug such as trazodone might be needed for insomnia problems [16].
The generally higher effect size on the HAM-A14 compared to HAM-A6 in the four GAD dose-response trials we have selected appears to indicate that pregabalin also has a positive effect on the somatic anxiety symptoms in GAD. The benzodiazepine arms ([Table 4]) for lorazepam as well as for alprazolam showed that an effect size of 0.40 or higher was only seen for the sleep item. The three lorazepam trials showed a diazepam-like profile [41], with a higher HAM-A14 effect size than HAM-A6, in contrast to the venlafaxine outcome as shown in [Table 3B]. The profile of pregabalin in the dose range 200-600 mg reflects both a clinically effect on HAM-A6 and on the sleep item.
This dose-response analysis of pregabalin in GAD is solely based on the effect size statistics. The background for this outcome measure for response in placebo-controlled trials is to be found in our meta-analysis of randomised controlled trials of fluoxetine versus placebo and tricyclic antidepressants [11]. In this study we compared different statistics, and the results showed that effect size was the most appropriate outcome measure.
A fundamental requirement for a drug to be considered an anti-anxiety compound for GAD patients is that it has shown efficacy in terms of symptom reduction in the core symptoms as covered by HAM-A6, which is more in accordance with the DSM-IV criteria for GAD than the HAM-A14 [13,27]. A Mokken analysis confirmed that the HAM-A6 is a uni-dimensional scale, implying that the total score is a sufficient statistic. A re-analysis of our first study with HAM-A [13] in patients with generalised anxiety disorder [15] confirmed our psychometric results, as the Mokken analysis showed a coefficient of homogeneity of 0.46 for HAM-A6 (i.e., uni-dimensionality) and 0.34 for HAM-A14 (i.e., multi-dimensionality). Finally, the four trials covered by this pooled analysis included a mean number of patients of approximately 80 (with a range from 66 to 91) in each treatment arm. As shown in the meta-analysis by Mitte et al. [28], this is a sufficient number of patients and, in fact, a condition for pooling data from different trials to obtain adequate homogeneity [37] [44].
The side-effect profile of pregabalin has been found safe in the dose range of 200-600 mg daily. As indicated in [Table 1] the data of completers on 600 mg daily did not differ from those found for placebo in terms of drop-out rates. The pharmacokinetics of pregabalin are unproblematic, as this compound is not metabolised, does not affect hepatic metabolism and does not have any significant drug interactions. No tolerance or abuse properties seem likely to develop in the chronic use in patients with epilepsy and DSM-IV generalised anxiety disorder.
In conclusion, this pooled analysis of DSM-IV-diagnosed patients with generalised anxiety disorder found that an acute (four week) course of treatment with pregabalin yielded clinically meaningful effect sizes on both the total HAM-A14 and on the HAM-A6. A clear dose-response relationship was identified indicating that a pregabalin dose of 150 mg daily is clearly insufficient, while the dose range between 200 mg and 600 mg daily is sufficient, demonstrating a plateau-like response curve. This dose-response pattern was also observed on the HAM-A sleep item. The drop-out rate was low, with doses of pregabalin up to 600 mg daily not differing significantly from placebo.
#Acknowledgements
I gratefully acknowledge Pfizer for making available the pregabalin data. The analyses were performed by Kem Phillips. I also gratefully acknowledge editorial input by Edward Schweizer.
#References
- 1 Allgulander C, Dahl AA, Austin C, Morris PL, Sogaard JA, Fayyad R, Kutcher SP, Clary CM. Efficacy of sertraline in a 12-week trial for generalized anxiety disorder. Am J Psychiatry. 2004; 161 1642-1649
- 2 American Psychiatric Association .
Diagnostic and Statistical Manual of Mental Disorders . 3rd edn., (DSM-III) Washington DC: APA 1980 - 3 American Psychiatric Association .
Diagnostic and Statistical Manual of Mental Disorders . 4th edn., (DSM-IV) Washington DC: APA 1994 - 4 Angst J, Bech P, Boyer P, Bruinvels J, Engel R, Helmchen H. et al . Consensus Conference on the Methodology of Clinical Trials of Antidepressants, Zurich, March 1988. Report of the Consensus Committee. Pharmacopsychiatry. 1989; 22 3-7
- 5 Ansseau M. The Atlantic gap: clinical trials in Europe and the United States. Biol Psychiatry. 1992; 31 109-111
- 6 Bagby RM, Ryder AG, Schuller DR, Marshall MB. The Hamilton Depression Rating Scale: Has the gold standard become a lead weight?. Am J Psychiatry. 2004; 161 2163-2177
- 7 Baldwin DS, Polkinghorn C. Evidence-based pharmacotherapy of generalized anxiety disorder. Int J Neuropsychopharmacol. 2005; 8 293-302
- 8 Bandelow B, Zohar J, Hollander E. World Federation of Biological Psychiatry (WFSBP) guidelines for pharmacological treatment of anxiety, obsessive-compulsive and posttraumatic stress disorders. World J Psychiatry. 2002; 3 171-199
- 9 Bech P. Rating scales in depression: Limitations and pitfalls. Dialogues Clin Neurosci. 2006; 8 207-215
- 10 Bech P, Andersen HF, Wade A. Effective dose of escitalopram in moderate versus severe DSM-IV major depression. Pharmacopsychiatry. 2006; 39 128-134
- 11 Bech P, Cialdella P, Haugh M, Birkett MA, Hours A, Boissel JP, Tollefson GD. A meta-analysis of randomised controlled trials of fluoxetine versus placebo and tricyclic antidepressants in the short-term treatment of major depression. Br J Psychiatry. 2000; 176 421-428
- 12 Bech P, Kajdasz DK, Porsdal V. Dose-response relationship of duloxetine in placebo-controlled clinical trials in patients with major depressive disorder. Psychopharmacology. 2006; 188 273-280
- 13 Bech P, Lunde M, Undén M. An inventory for the measurement of generalised anxiety distress symptoms, the GAD-10 Inventory. Acta Psychiatr Belg. 2005; 105 111-118
- 14 Bech P, Tanghøj P, Andersen HF, Overø K. Citalopram dose-response revisited using an alternative psychometric approach to evaluate clinical effects of four fixed citalopram doses compared to placebo in patients with major depression. Psychopharmacology. 2002; 163 20-25
- 15 Bjerrum H, Allerup P, Thunedborg K, Jakobsen K, Bech P. Treatment of generalized anxiety disorder: Comparison of a new beta-blocking drug (CGP361A), low-dose neuroleptic (flupenthixol), and placebo. Pharmacopsychiatry. 1992; 5 229-232
- 16 Bon OL. Low-dose trazodone effective in insomnia. Pharmacopsychiatry. 2005; 38 226
- 17 Bulpitt CJ. Medical statistics: meta-analysis. The Lancet. 1988; 2 93-94
- 18 Carpenter DJ, Pitts CD, Ruggiero L. Generalized anxiety disorder (GAD): Can the Hamilton psychic anxiety subscale be employed to measure primary drug response?. Poster at the NCDEU Meeting. 2006;
- 19 Fabre LF, Abuzzahab FS, Amin M, Claghorn JL, Mendels J, Petrie WM. et al . Sertraline safety and efficacy in major depression: a double-blind fixed-dose comparison with placebo. Biol Psychiatry. 1995; 38 592-602
- 20 Feltner DE, Crockatt JG, Dubovsky SJ, Cohn CK, Shrivastava RK, Targum SD. et al . A randomized, double-blind, placebo-controlled, fixed-dose, multicenter study of pregabalin in patients with generalized anxiety disorder. J Clin Psychopharmacol. 2003; 23 240-249
- 21 Field MJ, Ryszard JO, Lakhbir S. Pregabalin may represent a novel class of anxiolytic agents with a broad spectrum of activity. Br J Pharmacol. 2001; 132 1-4
- 22 Hamilton M. Diagnosis and rating of anxiety. Br J Psychiatry. 1969; , (special publ) 76-79
- 23 Hamilton M. The assessment of anxiety states by rating. Br J Med Psychol. 1959; 32 50-55
- 24 Hamilton M. Treatment of anxiety states. Components of anxiety and their response to benactyzine. J Ment Sci. 1958; 104 1062-1068
- 25 Hedges LV, Olkin I.
Statistical methods for meta-analysis. New York, NY: Academic Press 1985: 78-85 - 26 Licht RW, Qvitzau S, Allerup P, Bech P. Validation of the Bech-Rafaelsen Melancholia Scale and the Hamilton Depression Scale in patients with major depression: is the total score a valid measure of illness severity?. Acta Psychiatr Scand. 2005; 111 144-149
- 27 Meoni P, Salinas E, Brault Y, Hackett D. Pattern of symptom improvement following treatment with venlafaxine XR in patients with generalized anxiety disorder. Clin Psychiatry. 2001; 62 888-893
- 28 Mitte K, Noack P, Steil R, Hautzinger M. A meta-analytic review of the efficacy of drug treatment in generalized anxiety disorder. J Clin Psychopharmacol. 2005; 25 141-150
- 29 Mokken RJ.
A theory and procedure of scale analysis. Paris: Mouton 1971 - 30 Molenaar IW, Debets P, Systma K, Hemker BT.
User's manual MSP: A program for the Mokken scale analysis for polytomous items (version 30) . Groningen: lec ProGamma 1994 - 31 Montgomery SA, Tobias K, Zornberg GL, Kasper S, Pande AC. Efficacy and safety of pregabalin in the treatment of generalized anxiety disorder: a 6-week, multicenter, randomized, double-blind, placebo-controlled comparison of pregabalin and venlafaxine. J Clin Psychiatry. 2006; 67 771-782
- 32 Niklson IA, Reimitz PE. Baseline characteristics of major depressive disorder patients in clinical trials in Europe and United States: is there a transatlantic difference?. J Psychiatr Res. 2001; 35 71-81
- 33 Nutt D, Rickels K, Stein DJ. editors
Generalized anxiety disorder Symptomatology, pathogenesis and management. London: Martin Dunitz 2002 - 34 Nutt DJ, Ballenger JC, Sheehan D, Wittchen HU. Generalized anxiety disorder: comorbidity, comparative biology and treatment. Int J Neuropsychopharmacol. 2002; 5 315-325
- 35 Pande AC, Crockatt JG, Feltner DE, Janney CA, Smith WT, Weisler R. et al . Pregabalin in generalized anxiety disorder: a placebo-controlled trial. Am J Psychiatry. 2003; 160 533-540
- 36 Peto R. Why do we need systematic overviews of randomized trials?. Statist Med. 1987; 6 233-240
- 37 Pogue J, Yusuf S. Overcoming the limitations of current meta-analysis of randomised controlled trials. Lancet. 1998; 351 47-52
- 38 Pohl RB, Feltner DE, Fieve RR, Pande AC. Efficacy of pregabalin in the treatment of generalized anxiety disorder: double-blind, placebo-controlled comparison of BID versus TID dosing. J Clin Psychopharmacol. 2005; 25 151-158
- 39 Pollack MH, Zaninelli R, Goddard A, MacCafferty JP, Bellew KM, Burnham DB. et al . Paroxetine in the treatment of generalized anxiety disorder: results of a placebo-controlled, flexible-dosage trial. J Clin Psychiatry. 2001; 62 350-357 , Erratum in: J Clin Psychiatry 2001; 62: 658
- 40 Ray JW, Shadish WR. How interchangeable are different estimators of effect size?. J Consult Clin Psychology. 1996; 64 1316-1325
- 41 Rickels K, Downing R, Schweizer E, Hassman H. Antidepressants for the treatment of generalized anxiety disorder. A placebo-controlled comparison of imipramine, trazodone, and diazepam. Arch Gen Psychiatry. 1993; 50 884-895
- 42 Rickels K, Pollack MH, Feltner DE, Lydiard RB, Zimbroff DL, Bielski RJ, Tobias K. et al . Pregabalin for treatment of generalized anxiety disorder: a 4-week, multicenter, double-blind, placebo-controlled trial of pregabalin and alprazolam. Arch Gen Psychiatry. 2005; 62 1022-1030
- 43 Schmidt JE, Koch G, Vange LM La. An overview of statistical issues and methods of meta-analysis. J Biopharmaceut Statistics. 1991; 1 103-120
- 44 Smith GD, Egger M. Meta-analysis: Unresolved issues and future developments. Brit Med J. 1998; 316 221-225
- 45 Smith ML, Glass GV, Miller TI.
The benefits of psychotherapy. Baltimore: Johns Hopkins University Press 1980 - 46 Snaith RP, Baugh SJ, Clayden AD, Husain A, Sipple MA. The Clinical Anxiety Scale: an instrument derived from the Hamilton Anxiety Scale. Br J Psychiatry. 1982; 141 518-523
1 EF=(1-3/[4*(N1+N2-2)-1)]*{(M1-M2)/SQRT[(N1-1)*S1**2+(N2-1)*S2**2/(N1+N2-2)]} where M1, M2=means in the two groups; S1, S2=SDs in the two groups, and N1, N2=number of patients in the two groups.
#Correspondence
Prof. P. BechMD
Psychiatric Research Unit
Frederiksborg General Hospital
48 Dyrehavevej
3400 Hillerød
Denmark
Phone: +45/4829 32 53
Fax: +45/4826 33 77
Email: pebe@noh.regionh.dk
References
- 1 Allgulander C, Dahl AA, Austin C, Morris PL, Sogaard JA, Fayyad R, Kutcher SP, Clary CM. Efficacy of sertraline in a 12-week trial for generalized anxiety disorder. Am J Psychiatry. 2004; 161 1642-1649
- 2 American Psychiatric Association .
Diagnostic and Statistical Manual of Mental Disorders . 3rd edn., (DSM-III) Washington DC: APA 1980 - 3 American Psychiatric Association .
Diagnostic and Statistical Manual of Mental Disorders . 4th edn., (DSM-IV) Washington DC: APA 1994 - 4 Angst J, Bech P, Boyer P, Bruinvels J, Engel R, Helmchen H. et al . Consensus Conference on the Methodology of Clinical Trials of Antidepressants, Zurich, March 1988. Report of the Consensus Committee. Pharmacopsychiatry. 1989; 22 3-7
- 5 Ansseau M. The Atlantic gap: clinical trials in Europe and the United States. Biol Psychiatry. 1992; 31 109-111
- 6 Bagby RM, Ryder AG, Schuller DR, Marshall MB. The Hamilton Depression Rating Scale: Has the gold standard become a lead weight?. Am J Psychiatry. 2004; 161 2163-2177
- 7 Baldwin DS, Polkinghorn C. Evidence-based pharmacotherapy of generalized anxiety disorder. Int J Neuropsychopharmacol. 2005; 8 293-302
- 8 Bandelow B, Zohar J, Hollander E. World Federation of Biological Psychiatry (WFSBP) guidelines for pharmacological treatment of anxiety, obsessive-compulsive and posttraumatic stress disorders. World J Psychiatry. 2002; 3 171-199
- 9 Bech P. Rating scales in depression: Limitations and pitfalls. Dialogues Clin Neurosci. 2006; 8 207-215
- 10 Bech P, Andersen HF, Wade A. Effective dose of escitalopram in moderate versus severe DSM-IV major depression. Pharmacopsychiatry. 2006; 39 128-134
- 11 Bech P, Cialdella P, Haugh M, Birkett MA, Hours A, Boissel JP, Tollefson GD. A meta-analysis of randomised controlled trials of fluoxetine versus placebo and tricyclic antidepressants in the short-term treatment of major depression. Br J Psychiatry. 2000; 176 421-428
- 12 Bech P, Kajdasz DK, Porsdal V. Dose-response relationship of duloxetine in placebo-controlled clinical trials in patients with major depressive disorder. Psychopharmacology. 2006; 188 273-280
- 13 Bech P, Lunde M, Undén M. An inventory for the measurement of generalised anxiety distress symptoms, the GAD-10 Inventory. Acta Psychiatr Belg. 2005; 105 111-118
- 14 Bech P, Tanghøj P, Andersen HF, Overø K. Citalopram dose-response revisited using an alternative psychometric approach to evaluate clinical effects of four fixed citalopram doses compared to placebo in patients with major depression. Psychopharmacology. 2002; 163 20-25
- 15 Bjerrum H, Allerup P, Thunedborg K, Jakobsen K, Bech P. Treatment of generalized anxiety disorder: Comparison of a new beta-blocking drug (CGP361A), low-dose neuroleptic (flupenthixol), and placebo. Pharmacopsychiatry. 1992; 5 229-232
- 16 Bon OL. Low-dose trazodone effective in insomnia. Pharmacopsychiatry. 2005; 38 226
- 17 Bulpitt CJ. Medical statistics: meta-analysis. The Lancet. 1988; 2 93-94
- 18 Carpenter DJ, Pitts CD, Ruggiero L. Generalized anxiety disorder (GAD): Can the Hamilton psychic anxiety subscale be employed to measure primary drug response?. Poster at the NCDEU Meeting. 2006;
- 19 Fabre LF, Abuzzahab FS, Amin M, Claghorn JL, Mendels J, Petrie WM. et al . Sertraline safety and efficacy in major depression: a double-blind fixed-dose comparison with placebo. Biol Psychiatry. 1995; 38 592-602
- 20 Feltner DE, Crockatt JG, Dubovsky SJ, Cohn CK, Shrivastava RK, Targum SD. et al . A randomized, double-blind, placebo-controlled, fixed-dose, multicenter study of pregabalin in patients with generalized anxiety disorder. J Clin Psychopharmacol. 2003; 23 240-249
- 21 Field MJ, Ryszard JO, Lakhbir S. Pregabalin may represent a novel class of anxiolytic agents with a broad spectrum of activity. Br J Pharmacol. 2001; 132 1-4
- 22 Hamilton M. Diagnosis and rating of anxiety. Br J Psychiatry. 1969; , (special publ) 76-79
- 23 Hamilton M. The assessment of anxiety states by rating. Br J Med Psychol. 1959; 32 50-55
- 24 Hamilton M. Treatment of anxiety states. Components of anxiety and their response to benactyzine. J Ment Sci. 1958; 104 1062-1068
- 25 Hedges LV, Olkin I.
Statistical methods for meta-analysis. New York, NY: Academic Press 1985: 78-85 - 26 Licht RW, Qvitzau S, Allerup P, Bech P. Validation of the Bech-Rafaelsen Melancholia Scale and the Hamilton Depression Scale in patients with major depression: is the total score a valid measure of illness severity?. Acta Psychiatr Scand. 2005; 111 144-149
- 27 Meoni P, Salinas E, Brault Y, Hackett D. Pattern of symptom improvement following treatment with venlafaxine XR in patients with generalized anxiety disorder. Clin Psychiatry. 2001; 62 888-893
- 28 Mitte K, Noack P, Steil R, Hautzinger M. A meta-analytic review of the efficacy of drug treatment in generalized anxiety disorder. J Clin Psychopharmacol. 2005; 25 141-150
- 29 Mokken RJ.
A theory and procedure of scale analysis. Paris: Mouton 1971 - 30 Molenaar IW, Debets P, Systma K, Hemker BT.
User's manual MSP: A program for the Mokken scale analysis for polytomous items (version 30) . Groningen: lec ProGamma 1994 - 31 Montgomery SA, Tobias K, Zornberg GL, Kasper S, Pande AC. Efficacy and safety of pregabalin in the treatment of generalized anxiety disorder: a 6-week, multicenter, randomized, double-blind, placebo-controlled comparison of pregabalin and venlafaxine. J Clin Psychiatry. 2006; 67 771-782
- 32 Niklson IA, Reimitz PE. Baseline characteristics of major depressive disorder patients in clinical trials in Europe and United States: is there a transatlantic difference?. J Psychiatr Res. 2001; 35 71-81
- 33 Nutt D, Rickels K, Stein DJ. editors
Generalized anxiety disorder Symptomatology, pathogenesis and management. London: Martin Dunitz 2002 - 34 Nutt DJ, Ballenger JC, Sheehan D, Wittchen HU. Generalized anxiety disorder: comorbidity, comparative biology and treatment. Int J Neuropsychopharmacol. 2002; 5 315-325
- 35 Pande AC, Crockatt JG, Feltner DE, Janney CA, Smith WT, Weisler R. et al . Pregabalin in generalized anxiety disorder: a placebo-controlled trial. Am J Psychiatry. 2003; 160 533-540
- 36 Peto R. Why do we need systematic overviews of randomized trials?. Statist Med. 1987; 6 233-240
- 37 Pogue J, Yusuf S. Overcoming the limitations of current meta-analysis of randomised controlled trials. Lancet. 1998; 351 47-52
- 38 Pohl RB, Feltner DE, Fieve RR, Pande AC. Efficacy of pregabalin in the treatment of generalized anxiety disorder: double-blind, placebo-controlled comparison of BID versus TID dosing. J Clin Psychopharmacol. 2005; 25 151-158
- 39 Pollack MH, Zaninelli R, Goddard A, MacCafferty JP, Bellew KM, Burnham DB. et al . Paroxetine in the treatment of generalized anxiety disorder: results of a placebo-controlled, flexible-dosage trial. J Clin Psychiatry. 2001; 62 350-357 , Erratum in: J Clin Psychiatry 2001; 62: 658
- 40 Ray JW, Shadish WR. How interchangeable are different estimators of effect size?. J Consult Clin Psychology. 1996; 64 1316-1325
- 41 Rickels K, Downing R, Schweizer E, Hassman H. Antidepressants for the treatment of generalized anxiety disorder. A placebo-controlled comparison of imipramine, trazodone, and diazepam. Arch Gen Psychiatry. 1993; 50 884-895
- 42 Rickels K, Pollack MH, Feltner DE, Lydiard RB, Zimbroff DL, Bielski RJ, Tobias K. et al . Pregabalin for treatment of generalized anxiety disorder: a 4-week, multicenter, double-blind, placebo-controlled trial of pregabalin and alprazolam. Arch Gen Psychiatry. 2005; 62 1022-1030
- 43 Schmidt JE, Koch G, Vange LM La. An overview of statistical issues and methods of meta-analysis. J Biopharmaceut Statistics. 1991; 1 103-120
- 44 Smith GD, Egger M. Meta-analysis: Unresolved issues and future developments. Brit Med J. 1998; 316 221-225
- 45 Smith ML, Glass GV, Miller TI.
The benefits of psychotherapy. Baltimore: Johns Hopkins University Press 1980 - 46 Snaith RP, Baugh SJ, Clayden AD, Husain A, Sipple MA. The Clinical Anxiety Scale: an instrument derived from the Hamilton Anxiety Scale. Br J Psychiatry. 1982; 141 518-523
1 EF=(1-3/[4*(N1+N2-2)-1)]*{(M1-M2)/SQRT[(N1-1)*S1**2+(N2-1)*S2**2/(N1+N2-2)]} where M1, M2=means in the two groups; S1, S2=SDs in the two groups, and N1, N2=number of patients in the two groups.
#Correspondence
Prof. P. BechMD
Psychiatric Research Unit
Frederiksborg General Hospital
48 Dyrehavevej
3400 Hillerød
Denmark
Phone: +45/4829 32 53
Fax: +45/4826 33 77
Email: pebe@noh.regionh.dk