Lamotrigine and lithium are both approved by the Food and Drug Administration (FDA) for the maintenance treatment of bipolar disorder. Findings from recent major randomized clinical trials indicated that they had similar efficacy overall, although secondary analyses suggested some possible complementarity, with lamotrigine more effective for depressive prophylaxis and lithium more effective for manic prophylaxis.1,2 It therefore appears possible that combination treatment with these agents might be more effective than either agent alone. In spite of the fact that almost all large randomized controlled trials involve monotherapy (to conform to perceived FDA preferences), the use of more than one mood stabilizer is quite common in the clinical treatment of bipolar disorder.3 Thus it is encouraging that some recent studies have begun to provide some evidence in support of combined treatment.3 Lithium plus lamotrigine is also a combination with no major contraindications.
While widely used, there are no published clinical studies concerning the use of this combination in bipolar disorder. In this article, we report our clinical experience with the combination of lithium plus lamotrigine in the long-term treatment of bipolar disorder in order to generate hypotheses and as pilot data for further research.
The charts of all patients with bipolar disorder who were treated between 2000 and 2004 in a private community psychiatric practice (JK and DAB) and an academic bipolar specialty clinic (SNG) were screened to identify all subjects who had been treated simultaneously with lamotrigine plus lithium. All charts were included that reported use of lamotrigine in combination with lithium. Study approval and waiver of consent for this chart review were obtained from the Cambridge Health Alliance Institutional Review Board. Data were retrieved by three researchers (JK, DAB, and ES) and reviewed by two researchers (ES and SNG). Data were harvested anonymously and entered into a computerized program which was double-checked by two researchers (ES and SNG). Because of the small sample size and the observational nature of the study, only descriptive statistics are provided.
Charts were reviewed for the following clinical and demographic variables: age, sex, current substance abuse, concurrent medications, mood stabilizer use and dosage, evidence regarding poor response to standard mood stabilizers in the past, indications for treatment with lamotrigine plus lithium, evidence of mania or hypomania during lamotrigine plus lithium treatment, adverse events, dose and duration of maintenance treatment with lamotrigine plus lithium, and whether and why lamotrigine plus lithium was discontinued. Diagnoses were based on the consensus application of DSM-IV criteria for mood disorders by three psychiatrists with expertise in mood disorders (JK, DAB, SNG). Indications for treatment (acute depression, mood elevation, or mood-cycling) were established based on a similar consensus review of the chart data by three researchers (JK, DAB, SNG). Treatment response was established retrospectively by consensus of four researchers (ES, JK, DAB, and SNG) applying the Clinical Global Impressions-Bipolar Disorder-Improvement scale (CGI-BP): (−3 = very much worse, −2 = much worse, −1 = slightly worse, 0 = unchanged, +1 = slightly improved, +2 = much improved, +3 = very much improved). This scale is divided into measures of acute depression, acute mania, and prophylaxis of mood episodes, based on the clinical impression of the investigators. Utilization and validation of this scale have been described elsewhere.4 The standard definition of treatment response on this scale is a CGI-BP score of 2 or 3. Acute efficacy refers to effects during treatment while a patient was experiencing a current major depressive, manic, hypomanic, or mixed episode. Since manic symptoms can co-occur with depressive symptoms in bipolar disorder in the narrow DSM-IV definition of the mixed episode as well as in broader definitions of agitated depression or dysphoric mania, the CGI-BP allows researchers to assess benefit for the two components of bipolar illness even in cases in which treatment is predominantly focused only on acute depression or only on acute mania. Prophylaxis refers to efficacy in preventing mood episode morbidity as evidenced by reduced frequency or severity of symptoms or episodes during treatment as compared to before or after treatment. Prophylaxis applies to the maintenance phase of treatment, defined as 6 months or longer after the end of the acute phase. Good reliability of treatment response ratings was established between the researchers (all kappa values > 0.70).
Adequacy of previous mood stabilizer treatment was defined as lithium at serum levels above 0.8 mmol/L, valproate at serum levels above 75 μg/mL, or carbamazepine at serum levels above 8 μg/mL. When serum levels were not available, adequacy of dosing was defined as lithium ≥ 900 mg/day, valproate ≥ 1,000 mg/day, or carbamazepine ≥ 600 mg/day. Efficacy is reported on the basis of intent-to-treat analysis.
Clinical and demographic characteristics of the sample are shown in Table 1. Figure 1 presents Clinical Global Improvement scores by patient (numbering corresponds to that shown in Table 1).
The sample consisted of 21 patients (11 females, 10 males, mean age 43.2 years, range 20-73 years, SD = 11.9). Of the 21 patients, 16 were diagnosed with bipolar I disorder and 5 with bipolar II disorder. Mean doses of lithium and lamotrigine were 963 mg/day (range 150-2000 mg/day) and 179 mg/day (range 25-500 mg/day), respectively, used for a mean of 55.7 weeks (range 4-228 weeks). Concomitant medications included anticonvulsants (n = 14, excluding clonazepam), antidepressants (n = 10), antipsychotics (n = 6), thyroid hormone (n = 6), and methylphenidate (n = 1). The main indications for treatment were depression 57% (n = 12), mixed states 29% (n = 6), and rapid cycling 14% (n = 3).
Combination therapy was initiated in 24% of patients by adding lithium to lamotrigine and in 76% by adding lamotrigine to lithium. The sequence of usage did not clearly affect outcomes, since overall prophylactic response was observed in 2/5 (40%) of those treated with lamotrigine followed by lithium compared with 4/16 (25%) of those treated with lithium followed by lamotrigine.
Of the 15 patients for whom such data were available, 12 (80%) had previously failed to respond to at least one mood stabilizer due to lack of efficacy despite adequate duration and dose of treatment (as defined in the methods section): 3 (20%) had failed to respond to combination therapy with valproic acid and carbamazepine, 7 (47%) had failed to respond to monotherapy with valproic acid, and 2 (13%) had failed to respond to carbamazepine monotherapy.
Twelve of the subjects (57%) had at least one other axis I diagnosis, including substance abuse (n = 10), generalized anxiety disorder (n = 2), attention-deficit/hyperactivity disorder (n = 1), obsessive-compulsive disorder (n = 1), posttraumatic stress disorder (n = 1), panic disorder (n = 1), and anorexia/bulimia nervosa (n = 1). Fourteen percent were currently abusing substances, with an additional 33% reporting past substance abuse.
As shown in Figure 2, in this sample of 21 patients, acute antidepressant benefit was observed in 10 patients (48%), acute anti-manic benefit in 3 patients (14%), and overall prophylactic benefit in 6 patients (29%).
Of the 21 patients, 13 (62%) had no side effects; in those who did experience side effects, the most common were cognitive difficulty (6/21, 29%), rash (2/21, 9%), sedation (2/21, 9%), and constipation (2/21, 5%). Ten patients (48%) discontinued combination therapy, mainly due to lack of efficacy (n = 4, 19%) or the appearance of manic-like symptoms (n = 4, 19%), with only two patients discontinuing due to side effects (rash).
Manic symptoms were recorded in 5 patients (mania in 1, hypomania in 3, and a mixed episode in 1), all of whom were treated with lamotrigine added to lithium. These manic symptoms led to discontinuation of lamotrigine in 2 patients and reduction of dose in 1 patient.
In the long-term follow-up at a mean of approximately 1 year, lamotrigine plus lithium appeared effective in some patients with refractory bipolar disorder. In the acute phase, 48% of the 21 patients showed an acute response in depressive symptoms, while only 14% of the patients showed an acute response in manic symptoms. Long-term prophylactic efficacy for prevention of future mood episodes appeared to be limited to 29% of this sample. The magnitude of benefit was not as robust as might be expected based on data from clinical trials. This observation may be a consequence of the treatment-refractory nature of this sample, since the patients who had not responded to lithium may have been more refractory to whatever new treatments were added. However, this sample size is too small to definitively address this issue.
Most patients tolerated the combination well, with only 2 patients discontinuing due to side effects (rash), while approximately half of the patients dropped out due to insufficient efficacy.
The occurrence of mania in 5 patients, in all of whom lamotrigine was added to lithium, raises the possibility that such cases of mania were not just spontaneous, but rather that they might have been induced by lamotrigine. The manufacturer-sponsored clinical trials have tended to report no risk of lamotrigine-induced mania compared with placebo, a view commonly endorsed by many researchers. However, this interpretation ignores the fact that the lamotrigine clinical trials were not powered to assess whether lamotrigine has a higher risk of inducing mania than placebo. In cases where observations are made but no specific hypothesis has been powered for testing, the appropriate statistical methods are not p-values, with the implication that a placebo comparison p-value greater than 0.05 means the event did not occur. Rather, as in this study, descriptive statistics should be used. As we have shown elsewhere,5 in such an appropriate statistical analysis, a fair reading of the available lamotrigine clinical trial data indicates a four-fold increased risk of acute mania in patients treated with lamotrigine compared with placebo. Real-world clinical data as presented here and in other reports6 find that mania does sometimes occur with lamotrigine, consistent with the above interpretation of the lamotrigine clinical trial data.
This study is generally consistent with the implications of large randomized controlled trials that have suggested that these agents might be complementary. However, larger observational studies, and preferably, randomized studies, are needed to further delineate the relative validity or invalidity of these preliminary results.
Limitations of the Current Study
The observational nature of this study is justified because, based on a MEDLINE search, this appears to be the first published clinical study concerning this combination treatment, despite the fact that use of the combination appears to be a common clinical practice. However, the results of this study should be interpreted with caution because of a number of limitations. First, these data are open, uncontrolled, and retrospective. In addition, the ratings of response partially involved the assessment of the treating clinician, so that these data are open to bias in testing hypotheses. Second, the observational circumstances of treatment, especially the frequent use of other concomitant medications, may have obscured the real treatment effect of lamotrigine plus lithium.
Nevertheless, with all their limitations, observational studies have value in terms of enhanced generalizability of findings compared with the special populations used in clinical trials. This study also provides pilot data on a combination of treatments that is widely used in practice but that has not yet been studied in controlled settings.
Although lamotrigine and lithium were well tolerated in this study and the data suggested effectiveness in about half of the patients in the sample who had bipolar depression, the anti-manic and long-term efficacy of these agents appeared less robust in this sample. This observation may reflect the treatment-refractory nature of illness in the sample. The findings in this small, non-randomized sample cannot be seen as proof of these results, but rather should be taken as pilot data to encourage investigators to undertake larger, preferably randomized, studies to further assess this commonly used combination of agents.
1. Bowden C, Calabrese J, Sachs G, et al. A placebo-controlled 18-month trial of lamotrigine
maintenance treatment in recently manic or hypomanic patients with bipolar I disorder. Arch Gen Psychiatry 2003;60:392-400.
2. Calabrese J, Bowden C, Sachs G, et al. A placebo-controlled 18-month trial of lamotrigine
maintenance treatment in recently depressed patients with bipolar I disorder. J Clin Psychiatry 2003;64:1013-24.
3. Ghaemi SN. Polypharmacy
in psychiatry. In: Ghaemi SN, ed. New York: Marcel Dekker; 2002.
4. Spearing MK, Post RM, Leverich GS, et al. Modification of the Clinical Global Impressions (CGI) scale for use in bipolar illness (BP): The CGI-BP. Psychiatry Res 1997;73:159-71.
5. Ghaemi SN, Soldani F, Hsu DJ. Commentary: Evidence-based pharmacotherapy of bipolar disorder
. Int J Neuropsychopharmacol 2003;6:303-8.
6. Raskin S, Teitelbaum A, Zislin J, et al. Adjunctive lamotrigine
as a possible mania
inducer in bipolar patients. Am J Psychiatry 2006;163:159-60.