February 2010  << Back  

  Can't open the PDF? Click here for help.

 

 RESEARCH ARTICLE

A blinded, randomized comparison of immediate-release and extended-release carbamazepine capsules in manic and depressed bipolar subjects

Rif S. El-Mallakh, MD

Mood Disorders Research Program, Department of Psychiatry and Behavioral Sciences, University of Louisville School of Medicine, Louisville, KY, USA

Mary Ruth Salem, RN, MA

Mood Disorders Research Program, Department of Psychiatry and Behavioral Sciences, University of Louisville School of Medicine, Louisville, KY, USA

Amarjit Chopra, MD

Mood Disorders Research Program, Department of Psychiatry and Behavioral Sciences, University of Louisville School of Medicine, Louisville, KY, USA

Gregory J. Mickus, BS

Mood Disorders Research Program, Department of Psychiatry and Behavioral Sciences, University of Louisville School of Medicine, Louisville, KY, USA

Praveen Penagaluri, MD

Mood Disorders Research Program, Department of Psychiatry and Behavioral Sciences, University of Louisville School of Medicine, Louisville, KY, USA

Radhika Movva, MD

Mood Disorders Research Program, Department of Psychiatry and Behavioral Sciences, University of Louisville School of Medicine, Louisville, KY, USA

BACKGROUND: The anticonvulsant carbamazepine is approved by the FDA for treatment of acute mania. It is available in 2 formulations: immediate-release (IR) and extended-release carbamazepine capsules (ERCC). The relative efficacy of these formulations in acutely ill bipolar patients has not been previously investigated.

METHODS: This study is a subanalysis of a 3-month, blinded, equal, random-assignment comparison of adverse effect load of an IR carbamazepine formulation (Tegretol) and ERCC (Equetro) in type I or type II bipolar patients already receiving carbamazepine or clinically determined to benefit from carbamazepine treatment. Dosages were titrated to patients’ clinical needs. Subjects who scored >15 on the Montgomery Åsberg Depression Rating Scale (MADRS) or >14 on the Young Mania Rating Scale (YMRS) were included in this analysis. The primary outcome measures were the relative mood scores at the end of the study.

RESULTS: At the end of 3 months of treatment, all patients improved compared with their baseline, but there was no difference in mood ratings in subjects with an initial MADRS >15 (ERCC, 18.2 ± SD 11.9, vs IR, 12.0 ± 4.5; P = .3) or YMRS >15 (ERCC, 6.5 ± 6.4, vs IR, 4.7 ± 3.1; P = .7). When compared with their baseline, patients receiving IR improved earlier than patients receiving ERCC. There were no differences in overall adverse events in patients receiving IR or ERCC (23.1 ± 13.42 vs 22.3 ± 13.40; P = .9).

CONCLUSIONS: Carbamazepine is effective in treating symptoms of both mania and depression, and there are no significant differences in the relative efficacy of the IR or ERCC formulations.

KEYWORDS: carbamazepine, immediate release, extended release, acute mania, bipolar disorder

ANNALS OF CLINICAL PSYCHIATRY 2010;22(1):3–8

  INTRODUCTION

Carbamazepine is one of the earliest mood stabilizing agents available but has only recently been studied in large, multicenter, placebo-controlled trials, and was only recently approved by the FDA for the treatment of acute mania.1,2 Early carbamazepine studies were performed using an immediate-release (IR) formulation. These studies showed that carbamazepine has efficacy in treating mania3 and bipolar depression3,4 and in the prevention of relapse in patients with bipolar disorder.5,6

Multiple formulations for carbamazepine are available. These include generic IR formulations, a brand name IR formulation (Tegretol), and 3 types of extended-release (ER) formulations (Carbatrol ER, Equetro ER, and Tegretol XR). Carbatrol ER and Equetro ER carbamazepine capsules (ERCC) use the Microtrol delivery system, in which carbamazepine is packaged within a large number of miniscule beads within the capsule. Three types of beads are used: 25% are IR, 40% are slow release (SR) and dissolve over 8 to 12 hours, and 35% are enteric coated and will not dissolve in the stomach, thereby providing a delayed peak.7 Tegretol XR uses an oral osmotic (OROS) pump system. This formulation has a laser-drilled hole in a semipermeable capsule that allows water influx to dilute a high osmotic load inside. The increasing osmotic pressure within the capsule extrudes the carbamazepine through the microscopic hole. ERCC is the only ER formulation documented to be beneficial in 2 large placebo-controlled trials for acute mania.1,2

Since carbamazepine has a relatively brief half life (approximately 6 hours), SR formulations have the advantage of less-frequent daily doses. However, there is also naturalistic evidence that SR formulations are better tolerated than IR formulations in patients with epilepsy8 and bipolar disorder.9 The primary purpose of the current study was to examine the adverse effect (AE) load associated with ERCC vs IR. The study found that patients receiving ERCC had significantly fewer gastrointestinal AEs compared with patients receiving IR.10 The current subanalysis examines a subgroup of patients who entered the study with a significant level of affective morbidity to determine if there was a major difference in efficacy of ERCC compared with IR in bipolar patients. This is the only head-to-head comparison of these formulations in bipolar illness.

  METHODS

Study design

The parent study was a prospective double-blind, parallel-group comparison that examined the tolerability of ERCC compared to IR carbamazepine in bipolar I or II patients who were either receiving carbamazepine or were determined to potentially benefit from receiving carbamazepine. This effectiveness trial examined the efficacy and tolerability of carbamazepine in the setting of routine clinical care. Patients qualified for the study if they were currently prescribed carbamazepine or were determined to likely benefit from its addition to their regimen. Volunteer patients were randomly and blindly assigned to either ERCC or IR carbamazepine and followed prospectively for 3 months. The carbamazepine dose was clinically titrated to maximize benefit and minimize AEs, and all other medications were kept unchanged. The last previous noncarbamazepine medication change occurred a minimum of 1 month prior to study entry. Psychiatric symptoms and AEs were monitored monthly. All patients provided informed consent, and the study was approved by our Institutional Review Board.

Scales/primary and secondary outcome measures

The primary outcome measures for this subanalysis were psychiatric symptoms, which were assessed using the following scales:

  • Young Mania Rating Scale (YMRS),11 which measures manic symptoms

  • Montgomery-Åsberg Depression Rating Scale (MADRS),12 which measures depressive symptoms

  • Psychiatric Symptoms Assessment Scale (PSAS),13 which measures a wide array of psychiatric symptoms

AEs were quantified using the Udvalg for Kliniske Undersogelser (UKU) side effects scale.14

Inclusion and exclusion criteria

As an effectiveness trial, patients were included if they met DSM-IV15 criteria for either type I or II illness, as confirmed with a Mini International Neuropsychiatric Interview (MINI)16; were over age 18; were manageable as outpatients; and were either receiving carbamazepine or were thought to benefit from it. Patients were excluded if they had an active and unstable medical condition or active substance abuse/dependence (excluding caffeine and nicotine). Patients whose initial YMRS was ≥15 or whose initial MADRS was ≥15 and who were not on carbamazepine prior to study entry qualified for this subanalysis.

Statistical considerations

Continuous variables were assessed with analysis of variance (ANOVA) with post hoc Fisher’s predicted least-square difference (PLSD) with a correction for the number of comparisons made (StatView software program). Paired t tests were used to examine within-category change over time. Since these were limited, Bonferonni correction was not required. Categorical variables were examined using the chi square. In case of study dropouts, we used the last observation carried forward. A probability level of ≤.05 was considered significant.

  RESULTS

The original study recruited 44 patients, of whom 20 received the IR formulation and 21 received ERCC. Of all patients, 27 had an MADRS score ≥15 and 15 had a YMRS score ≥15. All 15 patients with manic symptoms also had MADRS scores >15, suggesting a current mixed state. All 27 patients completed the 3 months of the study. Data regarding the original study is presented elsewhere10; this report limits its analysis to the 27 patients with significant mood symptoms at study entry.

Of these 27 individuals, 4 were men (15%) and 23 were women (85%). Ages were 41.6 ± 10.8 years for those receiving IR and 42.2 ± 9.5 years for those receiving ERCC (t = 0.2; P = .9). Nineteen patients had type I illness and 8 had type II illness. Of those randomized to IR, 10 had type I and 4 had type II; of those randomized to ERCC, 10 had type I, and 3 had type II (χ2=4.84; P = .09). Comorbidity was common. Dysthymia occurred in 51.9%, social phobia in 33.3%, panic disorder in 18.5%, obsessive-compulsive disorder in 22.2%, generalized anxiety disorder in 14.8%, and posttraumatic stress disorder in 11.1%. There were no differences in randomization in any of these disorders. Three patients were receiving carbamazepine alone. Other medications included lamotrigine, an antipsychotic (quetiapine, aripiprazole, ziprasidone, risperidone, or loxapine), a benzodiazepine (clonazepam or alprazolam), gabapentin, topiramate, a stimulant (methylphenidate or amphetamine), lithium, and an antidepressant (bupropion, escitalopram, mirtazapine, fluoxetine, or duloxetine).

Among patients with either an MADRS or YMRS score >15, all patients improved compared with their baseline scores. The baseline MADRS score was 23.6 ± SD 5.5. This dropped significantly to 18.3 ± 8.2 by study end (paired t = 3.55; P = .0015) and was significant by month 2 (19.8 ± 8.3; paired t = 2.98; P = .006), but not by the end of the first month (20.3 ± 8.1; paired t = 1.9; P = .07). Similarly, the YMRS score dropped from 17.5 ± 2.4 to 9.2 ± 7.1 by the end of the study (paired t = 4.4; P = .0007). Manic symptoms were significantly improved at the end of the first month (12.3 ± 7.9; paired t = 2.65; P = .02), and continued to improve into the second month (10.1 ± 5.8; paired t = 4.6; P = .0005).

Among depressed patients receiving the IR formulation, MADRS scores dropped from 24.1 ± 5.3 at baseline to 17.9 ± 7.4 (paired t = 2.3; P = .04). This improvement was not significant after 1 month of treatment due to a high level of variance (19.3 ± 8.1; paired t = 1.7; P = .1), but did reach significance at 2 months as the variance decreased (19.5 ± 7; paired t = 2.7; P = .02) (FIGURE 1). Similar changes were seen in patients receiving ERCC. Baseline MADRS was high at 23.1 ± 5.9 and improved significantly by study end (18.4 ± 8.9; paired t = 2.9; P = .01), but not at 1 or 2 months (20.8 ± 8.2; paired t = 0.9; P = .4; and 19.9 ± 9.4; paired t = 1.6; P = 0.14, respectively) (FIGURE 1). Side effects did not differ (TABLE).

For mixed manic and hypomanic patients receiving the IR formulation, the YMRS score dropped from 17.1 ± SD 2.7 at baseline to 7.6 ± 5.03 at study end (paired t = 3.9; P = .008). This improvement was noted by the end of the first month of treatment (11.1 ± 5.6; paired t = 2.5; P = .048) and grew in the second month of treatment (9.7 ± 4.4; paired t = 3.3; P = .017) (FIGURE 2). Similarly, patients receiving ERCC also experienced a significant improvement in YMRS from baseline (17.9 ± 2.1) to study end (10.4 ± 8.4; paired t = 2.4; P = .05) (Figure 2). This was notable by month 2 of treatment (10.9 ± 8.0; paired t = 3.0; P = .02) but not at the end of the first month (11.8 ± 9.7; paired t = 1.3; P = .2) (FIGURE 2). Depressive symptoms and side effects did not differ (TABLE).

The PSAS was used as a measure of general psychopathology. There were no significant differences for most measures with the exception of greater improvement among depressed patients receiving IR compared with those receiving ERCC (TABLE). At the end of the study period, mental health quality of life as measured by the 36-item Short-Form Health Survey (SF-36) was poor overall (30.4 ± 13.2). Physical health quality of life as measured by SF-36 was comparable to the American average (46.7 ± 10.7). There were no differences between depressed or mixed manic patients overall or according to carbamazepine formulation. There were no differences in SF-36 scores in either manic or depressed patients (TABLE).

Overall AE load as measured by the UKU did not differ (TABLE). Evaluation of autonomic and gastrointestinal AEs, which were significantly less in patients receiving ERCC in the entire sample of 44,10 were not different within this sample due to inadequate power (autonomic AE: ERCC=3.9 ± SD 3.9 vs IR = 5.9 ± 5.6; t = 1.1; P = .3; gastrointestinal AE: ERCC=1.1 ± 1.2 vs IR = 1.8 ± 2.2; t = 1.02; P = .32).

Patients receiving ERCC achieved higher levels of carbamazepine despite similar carbamazepine doses. This was true for patients who were depressed at baseline (IR mean dose, 705.6 ± 235.2 mg/d, ERCC mean dose 680.0 ± 262.6; t = 0.2; P = .8; IR mean level, 6.8 ± 1.3 mcg/ mL, ERCC mean level, 9.7 ± 1.9 mcg/mL; t = 3.4, P = .005), and who were mixed at baseline (IR mean dose, 695.2 ± 227.2 mg/d; ERCC mean dose, 600.0 ± 202.2; paired t = 0.8; P = .4; IR mean level, 6.4 ± 1.5 mcg/mL; ERCC mean level, 10.3 ± 1.9 mcg/mL; paired t = 2.9; P = .03).

At study end, there were no differences in response between IR and ERCC for either depressive (FIGURE 1) or manic symptoms (FIGURE 2). However, the power of the study was modest at .1 for both manic and depressive symptoms at study end. The effect size of the difference between IR and ERCC for both manic and depressive symptoms was approximately .06. This is unlikely to be clinically significant and would require a study in which over 200 patients were available for analysis. Despite the low power, IR was associated with faster improvement in both depressive and manic symptoms than was ERCC (FIGURES 1 AND 2).


TABLE

Key variables in bipolar I and II patients with mixed mania or depression randomized to receive either IR carbamazepine or ERCC*

VARIABLE FORMULATION BASELINE
MEAN ± SD
STUDY END
MEAN ± SD
P VALUE (T) BASELINE VS END P VALUE END IR VS ERCC
Manic
YMRS IR 17.1 ± 2.7 7.6 ± 5.03 .008 (3.9) ns
ERCC 17.9 ± 2.1 10.4 ± 8.4 .05 (2.4)
MADRS IR 22.0 ± 4.9 19.0 ± 7.8 ns ns
ERCC 21.3 ± 6.1 16.0 ± 7.8 .02 (3.1)
PSAS IR 30.7 ± 6.6 27.3 ± 12.6 ns ns
ERCC 33.4 ± 10.3 25.2 ± 9.4 ns
UKU IR 27.3 ± 6.4 26.6 ± 14.6 ns ns
ERCC 25.8 ± 12.9 23.7 ± 16.0 ns
SF-36 mental IR   27.0 ± 11.1   ns
ERCC   33.2 ± 11.4  
SF-36 physical IR   43.6 ± 9.5   ns
ERCC   49.8 ± 6.7  
Depressed
YMRS IR 24.1 ± 5.3 17.9 ± 7.4 .04 (2.3) ns
ERCC 23.1 ± 5.9 18.4 ± 8.9 .01 (2.9)
MADRS IR 12.2 ± 6.2 6.5 ± 5.2 .045 (2.2) ns
ERCC 14.3 ± 3.9 10.4 ± 7.5 ns
PSAS IR 29.5 ± 2.3 16.5 ± 7.7 .06 (2.0) .05
ERCC 37.0 ± 8.2 28.5 ± 11.7 ns
UKU IR 28.8 ± 6.1 12.0 ± 7.1 ns  
ERCC 30.4 ± 11.8 21. 9 ± 14.1 ns
SF-36 mental IR   29.9 ± 13.7    
ERCC   30.9 ± 13.2  
SF-36 physical IR   45.7 ± 13.1   ns
ERCC   47.8 ± 7.9  
ERCC: extended-release carbamazepine capsules; IR: carbamazepine immediate release; MADRS: Montgomery Åsberg Depression Rating Scale; ns: not significant; PSAS: Psychiatric Symptoms Assessment Scale; SF-36: 36-item Short-Form Health Survey; UKU: Udvalg for Kliniske Undersogelser scale; YMRS: Young Mania Rating Scale.
*Paired t tests were used to compare baseline and study end. ANOVA was used to compare IR vs ERCC.
Both IR carbamazepine and ERCC were effective in significantly reducing symptom levels in bipolar I and II patients with mixed mania or depression.

FIGURE 1 Change in MADRS score in depressed or mixed manic patients receiving carbamazepine IR or ERCC

ERCC: extended-release carbamazepine capsules; IR: carbamazepine immediate release; MADRS: Montgomery Åsberg Depression Rating Scale.
*P ≤ .05 vs baseline.
Improvement in MADRS score in depressed or mixed manic patients receiving IR or ERCC was not different at any time point, but IR separated from baseline 1 month earlier than ERCC.

FIGURE 2 Change in YMRS score in mixed manic patients receiving carbamazepine IR or ERCC

ERCC: extended-release carbamazepine capsules; IR: carbamazepine immediate release; YMRS: Young Mania Rating Scale.
*P ≤ .05 vs baseline.
Improvement in YMRS score in mixed manic patients receiving IR or ERCC was not different at any time point, but IR separated from baseline 1 month earlier than ERCC.

  DISCUSSION

Carbamazepine is a proven agent for the management of bipolar mania, depression, and prophylaxis.1-6 Studies of safety and efficacy exist for both the IR and ERCC formulations but not for the OROS formulation (Tegretol XR). Only the ERCC formulation is approved by the FDA for use in bipolar mania under the trade name Equetro. (An identical ERCC formulation for epilepsy is marketed under the name Carbatrol.) The current study is the first double-blind, randomized comparative study of the ERCC formulation (Equetro) and the IR formulation (Tegretol). In outpatients with acute mixed mania or acute bipolar depression, both ERCC and IR carbamazepine were effective in significantly reducing symptom levels (TABLE), and there was no difference between the 2 formulations at study end (FIGURES 1 AND 2).

This is a subanalysis of a randomized, blinded effectiveness trial of AE load experienced by bipolar patients receiving different formulations of carbamazepine. The protocol did not control the carbamazepine dosage, and it was titrated to maximize clinical efficacy and minimize AEs. The study was not powered to determine an efficacy difference between the formulations. Our power analysis suggests that a much larger study would be required to capture a difference, if one existed. Despite this, improvement with the IR formulation compared with baseline occurred faster than with the ERCC formulation for both manic and depressive symptoms. Separation from baseline was seen 1 visit (1 month) earlier in patients receiving IR (FIGURES 1 AND 2), but comparing IR and ERCC directly did not reveal any differences.

In the larger sample of 44 patients, ERCC was better tolerated.10 This is consistent with the finding that SR formulations generally have fewer AEs.17,18 However, in this smaller sample, there were no differences in side effects. This may be an issue of inadequate power. Similarly, the illness burden perceived by the patients as measured by the SF-36 did not vary in either group (TABLE).

Patients receiving ERCC achieved higher trough levels than did patients receiving IR at similar doses. This may be related to the IR regimen, in which patients took their doses twice daily. This dosing frequency may be suboptimal for IR carbamazepine.

There are problems with the current study. Although the use of an effectiveness design increases the clinical applicability of the data, it reduces the certainty of the conclusions. Effectiveness designs reduce the exclusion criteria. In the current study, we did not limit concomitant medications, which could have reduced the potential difference between the 2 formulations. Additionally, we did not titrate the medication to a specific plasma level. The final level achieved was statistically higher for patients receiving the ERCC. This was not associated with a greater clinical response.

  CONCLUSION

Despite its shortcomings, the current study demonstrates that both depressive and manic symptoms improve with carbamazepine treatment in outpatients with mixed mania and bipolar depression. Both the ERCC and IR formulations are effective.

DISCLOSURE: The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

ACKNOWLEDGMENTS: This study was supported by a research grant from Shire Pharmaceutical Company, Wayne, PA. The investigators did not receive additional funds for patient recruitment nor for data analysis. The contract allows the investigators to publish the data independent of outcome. However, the senior author (R.S.E.) had received speaking honoraria from Shire prior to, but not subsequent to, the performance of this study.

    REFERENCES

  1. Weisler RH, Keck PE Jr, Swann AC, et al. Extended-release carbamazepine capsules as monotherapy for acute mania in bipolar disorder: a multicenter, randomized, double-blind, placebo-controlled trial. J Clin Psychiatry. 2005;66:323–330.
  2. Weisler RH, Hirschfeld R, Cutler AJ, et al, and the SPD417 Study Group. Extended-release carbamazepine capsules as monotherapy in bipolar disorder: pooled results from two randomised, double-blind, placebo-controlled trials. CNS Drugs. 2006;20:219–231.
  3. Ballenger JC, Post RM. Carbamazepine in manic-depressive illness: a new treatment. Am J Psychiatry. 1980;137:782–790.
  4. Post RM, Ketter TA, Uhde T, et al. Thirty years of clinical experience with carbamazepine in the treatment of bipolar illness: principles and practice. CNS Drugs. 2007;21:47–71.
  5. Greil W, Kleindienst N. The comparative prophylactic efficacy of lithium and carbamazepine in patients with bipolar I disorder. Int Clin Psychopharmacol. 1999;14:277–281.
  6. Greil W, Kleindienst N. Lithium versus carbamazepine in the maintenance treatment of bipolar II disorder and bipolar disorder not otherwise specified. Int Clin Psychopharmacol. 1999;14:283–285.
  7. Wheless JW, Venkataraman V. New formulations of drugs in epilepsy. Expert Opin Pharmacother. 1999;1:49–60.
  8. Miller AD, Krauss GL, Hamzeh FM. Improved CNS tolerability following conversion from immediate- to extended-release carbamazepine. Acta Neurol Scand. 2004;109:374–377.
  9. Ginsberg LD. A retrospective analysis of changing from alternative agents to carbamazepine extended- release capsules in bipolar disorder. Ann Clin Psychiatry. 2006;18(suppl 1):31–34.
  10. Salem MR, Chopra A, El-Mallakh RS. Adverse event load in bipolar subjects receiving either carbamazepine immediate-release or extended-release capsules: a blinded, randomized study. Presented at: 7th International Conference on Bipolar Disorder; June 7-9, 2007; Pittsburgh, PA.
  11. Young RC, Biggs JT, Ziegler VE, et al. A rating scale for mania: reliability, validity, and sensitivity. Br J Psychiatry. 1978;133:429–435.
  12. Montgomery SA, Åsberg M. A new depression scale designed to be sensitive to change. Br J Psychiatry. 1979;134:382–389.
  13. Bigelow LB, Berthot BD. The Psychiatric Symptom Assessment Scale (PSAS). Psychopharmacology Bull. 1989;25:168–173.
  14. Lingjaerde O, Ahlfors UG, Bech P, et al. The UKU side effect rating scale. A new comprehensive rating scale for psychotropic drugs and a cross-sectional study of side effects in neuroleptic-treated patients. Acta Psychiatr Scand Suppl 1987;334:1–100.
  15.  Diagnostic and statistical manual of mental disorders 4th ed. Washington, DC: American Psychiatric Association; 2000.
  16. Sheehan DV, Lecrubier Y, Sheehan KH, et al. The Mini-International Neuropsychiatric Interview (M.I.N.I.): the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. J Clin Psychiatry. 1998;59(suppl 20):22–33.
  17. Horne RL, Cunanan C. Safety and efficacy of switching psychiatric patients from a delayed-release to an extended-release formulation of divalproex sodium. J Clin Psychopharmacol. 2003;23:176–181.
  18. Bowden CL. Key treatment studies of lithium in manic depressive illness: efficacy and side effects. J Clin Psychiatry. 1998;59(suppl 6):13–19.

CORRESPONDENCE: Rif S. El-Mallakh, MD, Mood Disorders Research Program, Department of Psychiatry and Behavioral Sciences, University of Louisville School of Medicine, MedCenter One, 501 East Broadway Suite 340, Louisville, KY 40202 USA E-MAIL: rselma01@louisville.edu