The antimyotonic effect of lamotrigine in non-dystrophic myotonias: a double-blind randomized study

Abstract

Mexiletine is the only drug with proven effect for treatment of non-dystrophic myotonia, but mexiletine is expensive, has limited availability and several side effects. There is therefore a need to identify other pharmacological compounds that can alleviate myotonia in non-dystrophic myotonias. Like mexiletine, lamotrigine is a sodium channel blocker, but unlike mexiletine, lamotrigine is available, inexpensive, and well tolerated. We investigated the potential of using lamotrigine for treatment of myotonia in patients with non-dystrophic myotonias. In this, randomized double-blind, placebo-controlled, two-period cross-over study, we included adult outpatients recruited from all of Denmark with clinical myotonia and genetically confirmed myotonia congenita and paramyotonia congenita for investigation at the Copenhagen Neuromuscular Center. A pharmacy produced the medication and placebo, and randomized patients in blocks of 10. Participants and investigators were all blinded to treatment until the end of the trial. In two 8-week periods, oral lamotrigine or placebo capsules were provided once daily, with increasing doses (from 25 mg, 50 mg, 150 mg to 300 mg) every second week. The primary outcome was a severity score of myotonia, the Myotonic Behaviour Scale ranging from asymptomatic (score 1) to invalidating myotonia (score 6), reported by the participants during Weeks 0 and 8 in each treatment period. Clinical myotonia was also measured and side effects were monitored. The study was registered at ClinicalTrials.gov (NCT02159963) and EudraCT (2013-003309-24). We included 26 patients (10 females, 16 males, age: 19–74 years) from 13 November 2013 to 6 July 2015. Twenty-two completed the entire study. One patient withdrew due to an allergic reaction to lamotrigine. Three patients withdrew for reasons not related to the trial intervention. The Myotonic Behaviour Scale at baseline was 3.2 ± 1.1, which changed after treatment with lamotrigine by 1.3 ± 0.2 scores (P < 0.001), but not with placebo (0.2 ± 0.1 scores, P = 0.4). The estimated effect size was 1.0 ± 0.2 (95% confidence interval = 0.5–1.5, P < 0.001, n = 22). The standardized effect size of lamotrigine was 1.5 (confidence interval: 1.2–1.8). Number needed to treat was 2.6 (P = 0.006, n = 26). No adverse or unsuspected event occurred. Common side effects occurred in both treatment groups; number needed to harm was 5.2 (P = 0.11, n = 26). Lamotrigine effectively reduced myotonia, emphasized by consistency between effects on patient-related outcomes and objective outcomes. The frequency of side effects was acceptable. Considering this and the high availability and low cost of the drug, we suggest that lamotrigine should be used as the first line of treatment for myotonia in treatment-naive patients with non-dystrophic myotonias.

myotonia, Becker, Thomsen, paramyotonia, lamotrigine

Introduction

In the autosomal inherited non-dystrophic channelopathies, myotonia congenita and paramyotonia congenita, myotonia causes disabling muscle stiffness and pain, which are the most troublesome symptoms that limit daily living in these patients (Sansone et al., 2012; Heatwole et al., 2013; Trivedi et al., 2013). Myotonia congenita exists in a dominantly (Thomsen myotonia) and a recessively inherited form (Becker myotonia), while all cases of paramyotonia congenita are dominantly inherited (Johnsen and Friis, 1980; Colding-Jørgensen, 2005). Myotonia can be described as involuntary delay of muscle relaxation following a contraction. In myotonia congenita, myotonia usually occurs after the first contraction, but improves with repeated contractions; the so-called warm-up phenomenon. In paramyotonia congenita, myotonia usually worsens after repeated contractions (paradoxical myotonia), and is relieved by rest (Trivedi et al., 2013). In both conditions, myotonia usually presents in childhood and may affect all skeletal muscles, but the pattern and severity of muscle affection varies among mutations, and individuals, even among family members (Colding-Jørgensen, 2005; Trivedi et al., 2013). Myotonia can result in falls during running and awkward situations when the patient cannot relax the hand during a handshake or close the mouth whilst eating. Dangerous situations can occur during falls and if eyes stay closed after sneezing. Therefore, patients with myotonia are limited in their physical as well as social activities. The limitation of myotonia is very dependent on the challenges of each patient’s life, and patient-related outcome measures are well suited to capture these limitations in non-dystrophic myotonia (Sansone et al., 2012; Trivedi et al., 2013).

The clinical presentation varies and includes phenotypic overlap between myotonia caused by sodium and chloride channel gene mutations. In this paper, we refer to myotonia congenita and paramyotonia congenita based on the genotype and not the clinical presentation. Mutations in the sarcolemmal chloride ion channel (CLCN1) reduce resting chloride conductance, which causes myotonia congenita. Mutations in the sodium ion channel gene (SCN4A) cause long-lasting depolarization due to impaired sodium conductance, which causes paramyotonia congenita. (Matthews et al., 2010). In both channelopathies, the changed conductance results in sarcolemmal hyperexcitability. Several anticonvulsant drugs, anti-arrhythmic drugs, and other channel blockers have been suggested as treatment for myotonia, however, mexiletine is the only drug that has shown evidence of alleviating myotonia in non-dystrophic myotonia patients (Statland et al., 2012). Mexiletine is a voltage-gated sodium channel blocker that reduces hyperexcitability. The use of mexiletine is challenged by its high price, its discontinuation in many countries due to unfavourable side effects, and its limited availability because it is only manufactured in one place worldwide. At least a third of our patient cohort had had bad experiences with mexiletine, and thus declined future treatment with the drug. There is therefore a need to identify alternative pharmacological compounds that can alleviate myotonia in non-dystrophic myotonias. Lamotrigine has a benign side effect profile, is easy to obtain and costs ∼10% of the price of mexiletine (Rambeck and Wolf, 1993). Lamotrigine is also a voltage-gated sodium channel blocker that blocks the channel’s subunit 1.4, which is mainly expressed in muscle cells (Nakatani et al., 2013). Lamotrigine prolongs the refractory period of the voltage-gated sodium channel and also affects the voltage-activated calcium channels (Rambeck and Wolf, 1993; Errington et al., 2008; Nakatani et al., 2013). Therefore, lamotrigine may reduce myotonia. Compelled by these circumstances, we treated a few patients off-label with lamotrigine, and observed promising results. Lamotrigine has never been investigated for treatment of myotonia. Therefore, we investigated benefits and side effects of lamotrigine by both patient-related and objective outcomes in patients affected by non-dystrophic myotonia.

Materials and methods

Study design

This phase II randomized, double-blind, placebo-controlled crossover-study was conducted after GCP (Good Clinical Practice) standards, and monitored by the GCP Unit of the Copenhagen University Hospital (2013-562). The study is in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of Copenhagen (H-2-2013-131) and Danish Health and Medicines Authority (2013083535). It was registered at ClinicalTrials.gov (NCT02159963) and EudraCT (2013-003309-24) before it was performed from 1 November 2013 to 30 November 2015. To boost recruitment of participants, patients were studied at two geographic locations, the Neuromuscular Centers in Copenhagen and Aarhus, Denmark. The same investigators collected data at both locations (G.A., G.H.).

Participants

Inclusion criteria were: adults (≥18 years), genetically confirmed myotonia congenita or paramyotonia congenita, and clinical obvious myotonia in eye, hand, or leg muscles, which affected daily living. Exclusion criteria were: contraindications to treatment with lamotrigine, e.g. known allergy, reduced kidney or liver function, other concomitant treatments or conditions that could potentially interact with lamotrigine (epilepsy, long QT-interval on ECG), participation in other treatment studies <30 days before enrolment, pregnancy or breastfeeding.

In Denmark, patients with non-dystrophic myotonia are genetically tested and registered at the Copenhagen University Hospital, Rigshospitalet. All adult myotonia congenita or paramyotonia congenita patients from this registry were invited by letter (Fig. 1), and all who responded positively were asked to invite family members, with the same disorder, to participate in the study. We screened patients for clinical myotonia and exclusion criteria. The screening included medical history, an interview about myotonic symptoms, blood samples (to exclude liver and kidney disorders and pregnancy in fertile women), ECG, and MRC (Medical Research Council scale for manual muscle strength testing). MRC was assessed for shoulder abduction and flexion, elbow extension and flexion, wrist extension and flexion, hip flexion, knee extension and flexion, and ankle dorsal and plantar flexion (Table 1).

Table 1

Baseline characteristics of the randomized patients

	Group 1 (n = 13) Lamotrigine then placebo		Group 2 (n = 13) Placebo then lamotrigine
Sex
Male, n (%)	6 (46)	–	10 (77)	–
Female, n (%)	7 (54)	–	3 (23)	–
Age, years	40 ± 16	19–74	49 ± 12	25–67
BMI, kg/m2	26 ± 2	22–31	29 ± 5	21–39
Disorder (n)
Thomsen	5 (39%)	3 families	3 (23%)	2 families
Becker	3 (23%)	3 families	3 (23%)	3 families
PMC	5 (39%)	2 families	7 (54%)	5 families
Normal tests (n)
Leg: TUG/stair	3/3	1 MC, 4 PMC	0/2	2 PMC
Hand/Eye	0/0	–	1/2	2 MC
MBS	3	1 MC, 2 PMC	1	1 PMC
Scores
MBS	3.4 ± 1.3	1.0–5.0	3.4 ± 0.8	2.1–5.1
MRC	4.8 ± 0.3	4.2–5	4.9 ± 0.2	4.3–5
SF36	60 ± 20	25–85	68 ± 16	39–89

Data are: n (%), mean ± SD; range: X–X; BMI = body mass index; PMC = paramyotonia congenita; Normal tests: 11 patients had normal values in 1–2 of the five tests. MBS = Myotonia Behaviour Scale, range 1–6 (1 = no myotonia), average of 1 week; MRC = Medical Research Council scale for manual muscle strength testing ranging from 0–5 (5 = normal), average of all measured muscle groups; SF-36 = the 36-item Short Form Health Survey of quality of life, range 0–100 (0 = poor, 50 = normal), overall health status.

Figure 1

Flow of participants. Diagnosed adult patients: >18 years and genetically confirmed diagnoses of MC (myotonia congenita) or PMC (paramyotonia congenita). No contact info = patients who were not followed at our clinic and without accessible address or telephone numbers. No medical need = patients with few symptoms of myotonia. Patients lost to follow-up were included in the NNT and NNH analyses, but not in other analyses.

^aExcluded after 13 days due to allergic reaction.

^bExcluded after 32 days due to hospitalization caused by bleeding ulcers.

^cNo follow-up at third evaluation due to elucidation of other disorders (attention deficit hyperactivity disorder and cerebral vascular disease).

^dNo follow-up at fourth evaluation due to food poisoning.

Excl. cr. = exclusion criteria.

Before enrolment, some patients were treated for their myotonia with lamotrigine, mexiletine, or acetazolamide (Table 2). Patients treated with lamotrigine were provided with mexiletine as escape medicine during the trial. Escape medication was not allowed five half-life times before and during evaluations (Fig. 2). Participants were covered by the national patient insurance and transportation costs were covered, but they received no remuneration. All patients gave informed oral and written consent to participate.

Table 2

Clinical characteristics of participants

Family	Mutation(s)	Symptom(s)	Debut	Other disorders and treatments
B-Ia	c.854 G > A, c.854 G > A	Leg	4
B-II	c.920 T > C, c.2518_2519del	Leg, hand	6
B-III	c.1886 T > C, c.1886 T > C	Leg, hand, jaw, neck	5	Myotoniac
B-IV	c.815 C > T, c.1478C > A	Leg, jaw	6	Migrainec myotoniac anti-hormonal
B-V	c.1437_1450del, c.2680C > T	Leg, hand, jaw	7	Ulcerc
B-VI	c.501C > G, c.1936A > G	Leg	17
T-I	c.2680C > T	Leg, hand	25	Angina pectoris, hypocholesterolaemia
T-IIa	c.1439C > T	Leg, hand, eyes	5	Hypertensionc
T-IIa	c.1439C > T	Leg, hand	6	Herniated discc
T-III	c.382A > G	Leg, hand, jaw, eyes	4	PTSc myotoniac
T-IVa	c.1439C > T	Leg, hand, jaw, eyes, tongue	1	Pollen allergyc
T-IV	c.1439C > T	Leg, hand, jaw	5	Ehlers-Danlos
T-IV	c.1439C > T	Leg	4	Ehlers-Danlos
T-V	c.920T > C, c.1238T > Gb	Leg	6	Rosen glaucomac
PM-I	c.3877G > A	Leg, hand, jaw, eyes	6	Anti-hormonal
PM-I	c.3877G > A	Leg, hand, jaw	14
PM-II	c.3938C > T	Hand, jaw	0
PM-II	c.3938C > T	Hand, jaw	6
PM-III	c.4765G > A	Leg, jaw	6	Hypertensionc
PM-III	c.4765G > A	Leg, hand, throat, eyes, jaw, tongue	14	Hypertensionc GI refluxc diabetesc asthmac
PM-IV	c.3917G > C	Leg, hand, jaw, eyes	7	Myotoniac
PM-V	c.3938C > T	Leg, hand, jaw	1	Anti-hormonal
PM-V	c.3938C > T	Hand, jaw	2	Hypertensionc stressc
PM-V	c.3938C > T	Hand, jaw	5
PM-V	c.3938C > T	Leg, hand, jaw	5
PM-V	c.3938C > T	Hand	2	Hypertensionc hypercholesterolaemiac

B = Becker; GI = gastrointestinal; PM = paramyotonia; PTS = post-traumatic syndrome; T = Thomsen. Reference sections: SCN4A (NM_000334.4) and CLCN1 (NM_000083.2). Symptoms, debut of first symptom, other disorders, and treatments are reported by the patients.

^aDropout.

^bDominant inheritance.

^cPatient in treatment of the disorder.

Figure 2

Design of the study. All participants were randomized to receive lamotrigine in either the first (Group 1) or second (Group 2) treatment period. In both periods, participants were evaluated while on the following doses: 0 mg, 150 mg, and 300 mg (Weeks 0, 6, and 8). CK = plasma creatine kinase levels.

Randomization and masking

Participants were randomized to two 8-week periods of treatment with first lamotrigine then placebo (Group 1) or vice versa (Group 2). An independent pharmacy (Glostrup Apotek), without any contact to the trial participants, generated the random allocation sequence, using computer-generated block randomization, with blocks of 10 and a 1:1 allocation rate. The pharmacy filled, labelled and delivered 45 sets of sequentially numbered containers of trial medicine. Participants were given consecutive numbers in the order the investigators included them. The allocation concealment was in sealed non-transparent envelopes. Serum lamotrigine was analysed during the trial, and therefore one research assistant, with no other relation to the study, was unmasked. The results of the blood samples were masked until unblinding of the trial. All contact with trial participants was kept to two investigators (G.A., G.H.). They screened and enrolled patients, assigned patients to interventions, collected all trial data, and typed all primary and secondary data twice in the database to verify correct typing of data. Masking was kept until all data were collected and all data, including adverse events, were verified in the database. Therefore, patients, care providers, data collectors, and outcome adjudicators were all blinded. The pharmacy produced the trial medication in accordance with the rules of GMP and GDP. Placebo and lamotrigine were covered by hard gelatin capsules (Coni-Snap no 1, Capsugel^®), consisting of lactose monohydrate (0.5 ml) and doses of 0 mg, 25 mg, or 75 mg of lamotrigine (Lamotrigin Actavis). After each period, participants were asked to guess which treatment they had received.

Procedures

Participants were treated for two 8-week periods. Washout time was set to 1 week (five half-life times) as a minimum and 3 weeks as maximum (Fig. 2). The half-life of lamotrigine is 33 h. Thus, 1 week of washout ensured that serum lamotrigine levels were well below therapeutic levels when starting the second period. For each treatment period, patients received two containers of trial capsules with different doses (25 mg and 75 mg). The treatment doses increased every second week (25 mg, 50 mg, 150 mg, 300 mg). Patients were told to take the capsules in the morning. If a dose was forgotten, it should be noted in the dairy. Double doses were not allowed. After each treatment period, patients returned the remaining capsules, which were counted to verify the compliance. Blood samples were collected and stored in a −20° freezer until analysis for serum lamotrigine.

Outcomes

Primary and secondary outcomes were evaluated after 0, 6 and 8 weeks of each treatment (Fig. 2). The primary outcome measure was the Myotonic Behaviour Scale (MBS) (Hammarén et al., 2005). This is a patient-related outcome that has a good validity and reliability in patients with non-dystrophic myotonia in long-term evaluations (Hammarén et al., 2005). The MBS scale consists of six statements and participants had to choose the one that described their condition best. MBS ranges from 1 (no stiffness) to 6 point (incapacitating stiffness exists). One week before each of the six MBS evaluations, participants received a reminder SMS or e-mail to enhance compliance. Patients filled a Danish translation of the MBS in the diary during 7 days prior to each evaluation day, which was then collected at each evaluation day.

Secondary outcome measures were four functional timed tests, which assessed clinical myotonia of eye and hand closure, and leg muscles using a stopwatch (Kwieciński et al., 1992; Hammarén et al., 2005; Statland et al., 2012; Trivedi et al., 2013). The eyelid muscle relaxation time was measured after 5 s of maximal contraction. After the command ‘open’ participants should open their eyelids as fast as possible. The time from closure to fully open eyes was determined five times. The relaxation time of hand muscles was performed the same way. After 5 s of tight closure of the dominant hand, the time from the command ‘open’ to fully stretching of all five fingers was determined five times. The five repetitions were performed to identify the warm-up phenomenon. The treatment effect was calculated from the highest relaxation time, if the warm-up phenomenon occurred, or otherwise as the average of the five relaxation times.

Leg myotonia was assessed by two tests. First, the modified Timed Up and Go (TUG) test (Hammarén et al., 2005; Trivedi et al., 2013) was performed after 10 min of rest in the chair, measuring the time it took participants to stand up, walk 3 m and then return to sitting in the chair at their usual pace. The other test was the 14 Step Stair Test (14SST), in which patients walked 14 steps up the stairs and returned to the base as fast as possible, as described previously (Andersen et al., 2015).

Normal values of the four timed tests were obtained in a linear-mixed cohort of non-myotonic subjects (age: 20–70, n = 11 males, n = 10 females) (Andersen et al., 2016).

Other outcomes were self-assessed health measured by the short form (SF-36) questionnaire (McHorney et al., 1994; Hammarén et al., 2005; Sansone et al., 2012; Trivedi et al., 2013), creatine kinase levels, and the use of escape medication. Escape medication was noted in the diary during the trial.

If a patient took escape medication five half-life times before the evaluation day, the serum level was analysed. Data from evaluation days were excluded if the serum level was in the therapeutic window. During treatment periods, MBS data were excluded if escape medication had been taken within the five half-life times of lamotrigine. During placebo periods, data were not excluded.

Adverse events

Lamotrigine has a long track record for treatment of epilepsy, mood disorders and pain, and generally side effects are mild and well described, with the most common being headache and skin rash. The product resume of Lamotrigin ‘Actavis’ (D.SP.NR 21862) was used as a reference of suspected adverse events, and intensity of an event was evaluated by WHO definitions of SAE (serious adverts event), SAR (serious adverse reaction), and SUSAR (suspected unexpected serious adverse reaction). Participants were informed about the risk of side effects before inclusion. During the trial, inconvenience experienced by the participants was registered in the diary and participants were interviewed about adverse events at evaluations. Trial continuation was stopped in individuals in whom a suspected serious adverse event of lamotrigine was encountered.

Statistical methods

The sample size was estimated to include 42 patients, based on a defined clinical meaningful change of one MBS score, a conservative standard deviation (SD) of two MBS scores, power at 0.9, and α at 0.05. After completion of half the sample size, an interim analysis was made, and according to protocol, the study should be stopped if the treatment effect or worsening of lamotrigine was highly significant (P < 0.01).

We tested the crossover design for period effects by Mann-Whitney tests (MBS) and paired t-test (timed tests). To identify a treatment effect of lamotrigine, a paired two-tailed t-test was used if the normality test passed (Shapiro-Wilk) and the Wilcoxon signed rank test was used if the normality test failed. Meaningful changes in the timed tests were defined as a 50% reduction of the abnormal time. We log(10)-transformed the data from the timed tests before statistical analysis. The transformed data are presented together with the raw data. The relaxation times of eyelid and hand muscles at baseline varied between periods, irrespective of treatment allocation. In general, values from the first period were reduced compared with the second period. For this reason, baseline values from the second period, irrespective of treatment allocation, were used to calculate the treatment effect in these two tests.

Changes in SF-36 scores and plasma levels of creatine kinase were tested by paired t-tests between treatment allocations. Use of escape medication and side effects are presented as descriptive data, and were not subjected to a statistical analysis.

Number Needed to Harm (NNH) and Number Needed to Treat (NNT) were calculated as an intention to treat analysis that included dropouts. The McNemar test was performed to clarify whether the proportion of observations were significantly different from random occurrence. Dropouts were only included in the NNT and NNH analyses. If MBS scores were noted less than 4 days of an evaluation week, we defined it as missing data. Missing data at baseline (MBS: n = 2, SF-36: n = 1, TUG: n = 1, 14SST: n = 1) in one period were replaced by the baseline value from the other period. Missing data at the 300 mg evaluation (MBS: n = 1) were replaced by the 150 mg evaluation from the same period.

We estimated the standardized effect size of MBS data (effect size/baseline SD) with confidence interval (CI), to compare the effect of mexiletine (Statland et al., 2012) and lamotrigine.

Descriptive data are expressed as mean ± SD unless otherwise stated. Treatment effects are expressed as mean with 95% CI. A P-value < 0.05 was considered significant. Analysis and graphs were performed using Sigmaplot 11.2 Systate Software inc. US.

Role of funding source

The funders of the study had no role in study design, data collection, data analyses, data interpretation, or writing of the manuscript. The corresponding author has full access to all data in the study and had final responsibility for decision to submit for publication.

Results

From 13 November 2013 to 6 July 2015, we included 26 patients (Table 1) who completed the study in 134 ± 12 days of the 133 scheduled days. Four patients dropped out due to illness (Fig. 1). The study was stopped after the interim analysis according to protocol, because the treatment effect was highly significant. The warm-up phenomenon was demonstrated in the hand test of 11 myotonia congenita and three paramyotonia congenita patients. Paradoxical myotonia, i.e. prolongation of relaxation time with repeated contractions, was demonstrated in five patients with paramyotonia congenita and no patients with myotonia congenita. Disease characteristics of each patient are shown in Table 2. In total, participants forgot 156 of 2772 capsules during the whole trial without differences between treatments (placebo/lamotrigine: 77/79 capsules). Five most non-compliant patients forgot 16–24 capsules (12–20%) each in one treatment period [placebo (n = 4)/lamotrigine (n = 3): 64/63 capsules]. After the first period, 11/23 patients guessed the correct treatment allocation, after the second period 19/22 patients guessed the correct treatment allocation.

The primary outcome decreased significantly after treatment with lamotrigine (Fig. 3A and F). The mean MBS score decreased by 29% and NNT was 2.6 (P = 0.006, n = 26). If drop-outs (n = 4) and patients with low capsule compliance (n = 3) are excluded, the NNT was 1.9 (P = 0.009, n = 19). Serum lamotrigine levels increased from dose 150 mg (8 ± 5 mmol/l range: 2–24, n = 23) to dose 300 mg (16 ± 6 mmol/l, range: 7–27, n = 22). The effect of lamotrigine was smaller on dose 150 mg lamotrigine (NNT = 6.5, P = 0.046, n = 26). There was no relation between treatment effect assessed by MBS and TUG, and serum levels of lamotrigine.

Figure 3

Effect size of placebo and lamotrigine in primary and secondary outcomes. (A–E) Outcome measures before and after treatment with placebo and lamotrigine. Box plot: box includes the 25th–75th percentiles, median line, bars include 5th–95th percentiles, and dots are outliers. The horizontal lines indicate the normal value of the outcome measure. (A) MBS; (B) TUG; (C) Hand = relaxation time of finger flexor muscles; (D) Eye = relaxation time of eye closure. (E) 14SST; and (F) treatment effect (n = 22). Columns are the treatment effect of placebo and lamotrigine for each outcome measure. Bars are standard error (SE). P-values and confidence intervals of differences in treatment effects. TUG, 14SST, HAND, and EYE are measured in seconds (s). HAND and EYE are subtracted from the second baseline value, irrespective of treatment allocation.

The standardized effect size of lamotrigine was 1.5 (CI: 1.2–1.8) and of mexiletine 1.4 (CI: 0.6–2.2) and 3.0 (CI: 0.1–3.1). Thus, the standardized effect size was in the range of the other treatment’s CI, indicating a similar treatment effect of mexiletine and lamotrigine.

Baseline and end-treatment results of the timed tests are present in Fig. 3B–E. There was no effect of placebo on timed tests, but lamotrigine treatment significantly reduced the time in all four timed tests (Fig. 3B–D). Only patients with effect in three to four other outcome measures improved in the 14SST. The effect size (effect on lamotrigine minus effect on placebo), was significant for TUG, eye, and hand tests (Fig. 3F). NNT for timed tests (n = 26) were: TUG: 2.9 (P = 0.01); 14SST: 8.7 (P = 0.15); Hand: 2.4 (P = 0.007); and Eyes: 3.7 (P = 0.07).

The log-transformed data emphasized the effect of lamotrigine treatment compared with placebo (Table 3).

Table 3

Lamotrigine treatment compared with placebo

	Baseline (s)	Effect size (s)	CI	P
TUG	5.1 ± 2.9	1.2	1.2–1.3	0.003
14SST	12.1 ± 2.3	1.1	1.1–1.2	0.015
Hand	4.3 ± 3.8	2.8	2.2–3.6	0.01
Eye	4.8 ± 4.1	2.3	1.8–2.8	0.02

Normal health, measured by SF-36, is defined as 50 (McHorney et al., 1994), the higher the better. At baseline, the overall health status in patients was 65 ± 18. The domain with lowest score was role limitations due to physical health (43 ± 37). The overall health status was improved after treatment with lamotrigine compared with placebo (5 ± 13, CI: 3–7, P = 0.002, n = 22). This improvement was driven by an increase in the domains of: physical functioning CI: 10–15%, P = 0.001, n = 22; role limitations due to physical health CI: 10–21%, P = 0.04, n = 21; social function CI: 3–9%, P = 0.02, n = 22.

Plasma creatine kinase levels (baseline: 427 ± 618 mmol/l, range: 57–3080, n = 25) were not influenced by lamotrigine treatment. No patients took escape medication when on the dose of 300 mg lamotrigine. Two patients used escape medication in the placebo period: Patient B-III used one dose of 100 mg of mexiletine and Patient PM-IV used three doses of 250 mg Diamox^®. A third patient (Patient B-IV) needed escape medication during the whole placebo period, except on evaluation days, and during the low doses of lamotrigine, in total 84 daily doses of 200 mg of mexiletine.

Two events occurred during the trial, which lead to exclusion (blinding was kept). One patient on lamotrigine experienced an allergic reaction. After 2 weeks of treatment, the patient experienced muscle and join pain and a skin rash, which was reversible after stopping the trial medication. Another patient on placebo experienced anaemic symptoms, was hospitalized, and diagnosed with bleeding bacterial ulcer. The most common side effects reported were headache, fatigue, and skin rash. Fifteen patients experienced side effects during the trial, including the two patients who were excluded. Six of 15 patients experienced side effects in both periods, seven only in the lamotrigine period, and two only in the placebo period. Eleven patients did not report any side effects at all (six in Group 1, five in Group 2). Some patients experienced more than one side effect, and on several days. The number of days with side effects was similar, with 8% of days on lamotrigine and 6% of days on placebo (Table 4). The NNH was estimated at 5.2 (P = 0.11, n = 26).

Table 4

Side effects during the trial

Common side effects	Lamotrigine (n = 25)		Placebo (n = 24)
	n	Days	n	Days
>10%
Headache	11	49 (15)	6	50
Skin rash/itch	3	23 (15)	1	12
Fatigue	6	24 (3)	5	18
Muscle/join pain/oedema	6	15 (8)	2	6
Sore throat	3	11(0)	1	4
Nausea	2	4 (3)	3	9
1–10%
Dizziness	6	29 (6)	2	2
Irritability	1	4 (2)	-	-
Palpitations	2	2 (0)	1	2
Flu, fever, unleashed	2	3 (0)	2	7
Total	13	118 (48)	8	75

The table describes the side effects that the 26 participants noted in their diaries during the trial. Eleven participants did not note any side effects and six patients noted side effects in both treatment periods. ‘Common’ refers to the frequency of side effects in the product leaflet of lamotrigine. n = number of patients who had the side effect during the treatment period; Days = the total of days where one patient had the side effect; (n) = days with side effect in the period of lamotrigine doses 150/300 mg.

The treatment effect as assessed by MBS tended to be better in Group 2 versus Group 1 (difference: 1.0 MBS point, CI:−0.01–2.00, P = 0.51), but treatment effects were similar for the timed tests in the two groups. The magnitude of the treatment effect was unrelated to genetic diagnosis. Thus, eight myotonia congenita and 10 paramyotonia congenita patients had effect, whereas four patients had no effect (Patients B-III, T-IV₃, PM-I₂ and PM-V₅, Table 2).

Discussion

In this double-blind randomized controlled study, we found that lamotrigine effectively reduced myotonia in patients with non-dystrophic myotonia. The results are emphasized by the consistency between effects on the patient-related outcome (MBS) and the objective indices of clinical myotonia (Fig. 3F). These improvements were accompanied by improved quality of life in the domains of physical functions, which was significantly improved after treatment with lamotrigine compared with placebo. As a further indication of effect of lamotrigine, 86% of the participants were able to guess the treatment allocation at the end of the study, and despite dropout or low compliance in seven patients, NNT was 2.6 (P < 0.01). Finally, no serious events or unknown adverse reactions to lamotrigine occurred, and the frequency of side effects was acceptable (NNH = 5.2, P = 0.11). These results suggest that lamotrigine should be an alternative to mexiletine treatment in non-dystrophic myotonias. Lamotrigine should be considered first-line therapy in treatment-naive patients with NDM, due to effect, price, and availability of the drug, and in patients who are intolerant to mexiletine.

Treatment of non-dystrophic myotonias was systematically reviewed in 2009, and the conclusion then was that evidence of efficacy for any drug did not exist because of lack of studies (Tripet al., 2006). Mexiletine has been used empirically to treat non-dystrophic myotonias, until 2012 where mexiletine, as the only treatment for non-dystrophic myotonias, was investigated systematically (Statland et al., 2012). Similar to our study, mexiletine was investigated in a blinded, cross-over randomized placebo-controlled trial, using a patient-related primary outcome and clinical myotonia tests as secondary outcomes. As with lamotrigine in our study, mexiletine reduced self-assessed myotonia severity score and myotonia in clinical tests and the SF-36 domain of physical function improved (Statland et al., 2012). Unfortunately, 25% of primary outcome data were missing and it was not reported how these missing data were interpreted. Furthermore, although similar, the outcome measures were different, and therefore a direct comparison of efficacy between the two studies is not possible. However, the estimated standardized effect size in the studies indicates similar efficacies. A head-to-head comparison between the two drugs could not be performed in our study, as about a third of our patients declined mexiletine treatment because of side effects from previous experience with the drug. The most common side effects of mexiletine were gastrointestinal, neurologic, and pain, which affected 18 during mexiletine treatment versus two participants during placebo. From the number of participants who experienced side effects during active and placebo treatments, we estimated the NNH to be 3.0 for mexiletine. This is in accordance with a retrospective study of experienced side effects during mexiletine treatment, in which 52% of patients reported one or more adverse events during 6 months to several years of treatment (Suetterlin et al., 2015). The frequency of side effects was also high in the mexiletine study of patients with myotonic dystrophy type 1 (Logigian et al., 2010), where all participants experienced at least one side effect on dose 200 mg three times daily (n = 18). Thus, based on a good effect of both drugs, the estimated NNH seems to favour lamotrigine, but this cannot be concluded with certainty, as the two drugs were not evaluated head-to-head.

The side effects of lamotrigine are well known after treating patients with epilepsy and mood disorders with lamotrigine for decades (Chadwick, 1997). In general, the frequency of side effects is low, which was also the case in our study, where common and well known side effects of lamotrigine were reported and no serious events or previously unknown adverse events of lamotrigine occurred. Thus, more events and more participants, in absolute terms, were affected in the lamotrigine-treated period compared with the placebo period (Table 4), and the NNH analysis was not statistically significant. One patient experienced reversible symptoms of an allergic reaction, which excluded this patient from further treatment. Most side effects of lamotrigine occur within the first 8 weeks of treatment (Rambeck and Wolf, 1993), which was the treatment period in our study. Still, long-term studies are warranted to verify the benign side effect profile of lamotrigine in non-dystrophic myotonias.

Combining drugs that act by different mechanisms may reduce myotonia at lower doses of each drug. In a preclinical study, a synergistic anti-myotonic effect of lamotrigine and rufinamide was observed in vitro (Skov et al., 2017). This may be a promising treatment strategy, if side effects or drug interactions are not potentiated. Therapeutic options that reduced the sarcolemmal excitability through other channels, than the sodium channel, could also be potential treatment options for non-dystrophic myotonias (Kwieciński et al., 1992). For instance, potassium channel openers (Su et al., 2012), slow inactivators of sodium channels (Novak et al., 2015), and new derivatives of known sodium blockers (De Bellis et al., 2017) have a better effect than mexiletine on myotonia in in vitro models of myotonic muscles. However, these drugs need to be tested in clinical trials such as the present one.

Lamotrigine acts on sodium channels via a different mechanism than mexiletine and the tolerability of lamotrigine may differ from mexiletine, as suggested by the higher NNH for lamotrigine.

The effect of lamotrigine in our trial occurred irrespective of the type of channel mutations carried by the patients. Sodium channel mutations cause myotonia by overactivation of the channels (Matthews et al., 2010), and lamotrigine inactivates the voltage-gated sodium channels through hyperpolarization (Nakatani et al., 2013), which explains its effect on myotonia in patients with paramyotonia congenita. In myotonia congenita, defective chloride channels enhance the resting membrane potential, and thereby increase excitability, which produces myotonia (Matthews et al., 2010). The warm-up phenomenon, which reduces myotonia after a few minutes of muscle contractions, is proposed to include slow activation of the sodium channels (Novak et al., 2015). Lamotrigine does not affect the chloride channels (Su et al., 2012), but suppresses the excitability by blocking voltage-gated sodium channels, and hence reduces myotonia. In line with this, lamotrigine was found to fully abolish myotonia within clinically relevant concentrations in vitro in isolated CLCN1 inhibited human skeletal muscle (Skov et al., 2017).

There was a clear effect on the primary outcome, MBS, but also secondary outcomes in our study. The MBS has a high sensitivity to detect treatment-induced changes in patients with non-dystrophic myotonia (Hammarén et al., 2005). Another strength of the study was the high capsule compliance (95%). The number of dropouts was low (15%) and equally distributed in the two groups. Missing data are often a problem when using patient-related outcomes, as was also seen in the mexiletine study of non-dystrophic myotonias (Statland et al., 2012). In our study, missing data were minimal (3% of MBS, 1% SF-36). The high compliance of patient-related outcome was probably associated with reminder messages and a high interest of patients to participate. The few missing data points for MBS, in addition to good reliability and validity (Hammarén et al., 2005), emphasizes MBS as a relevant patient-related outcome measure for non-dystrophic myotonias. Participants were included continuously during the inclusion period of 21 months, which reduced the potential bias between periods of the trial, e.g. winter and summer, which affects patients with paramyotonia congenita in particular (Nielsen et al., 1982; Colding-Jørgensen, 2005).

A cross-over study offers many advantages, because one doesn’t have to consider possible confounders between treatment groups, often found in a parallel design, because patients are their own control. The effect size of MBS data tended to be lower in Group 1 versus Group 2, but effect sizes were similar for tests of clinical myotonia, which suggests a random cause of the MBS variability. This is also supported by no differences in baseline MBS data between periods, and that the treatment effects were low on the 150 mg dose of lamotrigine, at which dose lamotrigine was many fold higher than after the washout period.

The study was stopped according to protocol after the interim analysis showed superior efficacy for lamotrigine. The interim analysis was carried out to terminate the trial if patients did statistically worse or better on lamotrigine. The treatment effect, although highly significant, was likely underestimated because one or two of the five outcome measures were within normal range in some patients before treatment, thus hindering any chance to see improvement (Table 1). A limitation of the study is the observer bias in measuring fully open eyes and fully stretching of the fingers, which was particularly difficult in the least affected patients.

In this study, one-third of all diagnosed adult patients with non-dystrophic myotonia in Denmark were included. The study included patients of both sexes, all ages (19–74 years), three diagnoses of myotonia, different comorbidities (Table 2), different disease severities, and thus included a representative cohort of non-dystrophic myotonia patients. Most patients, who did not participate, did so because they felt no medical need for treatment of their myotonia, or they wished to live drug-free lives. The results indicate that treatment-naive non-dystrophic myotonia patients with clinically troublesome myotonia should be treated with lamotrigine.

Funding

The Jascha Foundation, Aase and Einar Danielsens Foundation, Augustinus Foundation, and A. P. Moeller Foundation supported the study financially.

Conflicts of interest

G.A. declares no competing interest. G.H. declares no competing interest. N.W. received research support from the Danish Council for Independent Research in Medical Sciences. M.D. declares no competing interest. H.A. has received research, travel support and speaker honoraria from Octapharma, CSL Behring, Pfizer and Genzyme/Sanofi, and has served as consultant on advisory board of UCB Pharma within the last 3 years. J.V. has received research and travel support and speaker honoraria from Genzyme/Sanofi and Ultragenyx Pharmaceuticals, and has acted as consultant on advisory boards for Genzyme/Sanofi, Sarepta, Lundbeck, Ultragenyx Pharmaceuticals, NOVO Nordisk, aTyr Pharma and Alexion Pharmaceuticals within the last 3 years.

Abbreviations

Abbreviations
 
• 14SST

  14 Step Stair Test

 
• MBS

  Myotonic Behaviour Scale

 
• NNH

  number needed to harm

 
• NNT

  number needed to treat

 
• TUG

  Timed Up and Go