Gabapentin treatment in clozapine-induced restless legs syndrome: two cases and a review of the literature

Abstract

Restless legs syndrome (RLS) is a neuro-sensorimotor disorder affecting 2–4% of adults. It is characterized by intense urges to move the legs, associated with unpleasant sensory disturbances in the legs occurring at rest and manifests mostly in the evening and night, relieved by movement. Diagnosis is primarily based on clinical presentation and the consensus criteria for the diagnosis have been established. Antipsychotics, the dopamine antagonists, have been reported to induce RLS. Dopamine agonists, the effective first-line treatment of RLS, carry the risk of inducing or worsening psychosis. Many nondopaminergic agents including antiepileptic medications have also been used in the treatment of primary RLS. In this report we describe clozapine-induced RLS in two patients with schizophrenia and its successful treatment with gabapentin, a nondopaminergic agent. In addition, we have reviewed the available literature on clozapine-induced RLS and its management.

Introduction

Restless legs syndrome (RLS) is a neuro-sensorimotor disorder affecting 2–4% of adults [Allen et al. 2010]. RLS occurs predominantly at rest and manifests mostly in the evening and night. It is characterized by intense urges to move the legs, associated with unpleasant sensory disturbances in the legs that get relieved by movement [Allen et al. 2003]. Diagnosis is primarily based on clinical presentation and the consensus criteria for the diagnosis have been established [Allen et al. 2003, 2014]. Antipsychotics, the dopamine receptor antagonists, have been reported to induce RLS [Allen, 2014]. Dopamine agonists, the effective first-line treatment of RLS, carry the risk of inducing or worsening psychosis [Garcia-Borreguero et al. 2013; Stoner et al. 2009]. In this report we describe clozapine-induced RLS in two patients with schizophrenia and its treatment with gabapentin, a nondopaminergic agent. In addition, we have reviewed the available literature on clozapine-induced RLS and its management. The informed written consent was taken from both the patients to report it in a medical journal.

Case reports

Case report 1

Ms G, a 32-year-old female with treatment-resistant schizophrenia had significant clinical improvement with clozapine 225 mg/d (25 mg in the morning and 200 mg at bedtime) and haloperidol to 2.5 mg/d at the time of presentation to our clinic. She had been referred to address a psychosocial issue and was admitted in our hospital. She had been started on clozapine about 2 years before while on haloperidol 20 mg/d and received clozapine 375 mg/d and haloperidol 10 mg/d for the initial 6 months. Her psychotic symptoms improved after adding clozapine and over the period of the next 1.5 years the doses of clozapine and haloperidol had been decreased without any relapse of psychotic symptoms.

Careful evaluation during the inpatient stay revealed intermittent sleep disturbances in the patient for last 2 years since starting clozapine. She did not have this symptom while she was treated with haloperidol up to 20 mg/d. On enquiry she described the urge to move her legs and dull pain along with an unpleasant sensation in both the calf muscles in the night lasting for few hours for 2 years. These symptoms used to appear by around 9 pm every night, after about half an hour of ingesting clozapine. Movements like walking used to relieve her urge and unpleasant sensation. She used to pace for about 40–45 min every night before going to sleep. While in bed she used to frequently flex, extend and move her legs sideways to get rid of the urge and pain. Characteristically these symptoms used to be absent in the morning hours even though she used to take clozapine in the morning as well. She had these symptoms almost every day interfering with initiation of sleep for at least 3–4 days in a week. She had no difficulty in maintenance of sleep. The severity of these symptoms had remained unchanged with the reduction of clozapine dose over the last 1.5 years. Restriction of symptoms only to the legs, presence of unpleasant sensation rather than inner restlessness and presence of diurnal variation made us think of RLS and ruled out akathisia. On structured evaluation she fulfilled the International Restless Legs Syndrome Study Group (IRLSSG) diagnostic criteria for RLS [Allen et al. 2003]. On rating with International Restless Legs Syndrome Rating Scale (IRLS) she scored 12, amounting to moderate intensity of RLS [Walters et al. 2003]. Naranjo adverse drug reaction probability scale score was 5 suggesting clozapine-induced RLS to be probable [Naranjo et al. 1981]. There was no past or family history suggestive of RLS.

Ms G was a known case of hypothyroidism for the last 3 years and was on regular levothyroxine supplementation. She had no other physical comorbidities. General physical examination and detailed systemic examination did not reveal any abnormality. Serum thyroid stimulating hormone (TSH) level was 2.64 µIU/ml. Full blood picture, peripheral smear, liver and renal functions, and blood glucose levels were within reference limits.

We educated the patient and her parents about the nature of her symptoms and the diagnosis of RLS. She was started on gabapentin 100 mg, orally at night and monitored for possible side effects. Significant improvement in her RLS symptoms was noted within 2–3 days of treatment. Her IRLS score decreased to 2 at the end of a week and she maintained improvement on gabapentin for 3 months. She was referred back to the psychiatrist at a secondary care hospital.

Case report 2

Mrs A, a 45-year-old married female, with a known case of treatment-resistant schizophrenia had exacerbation of her psychotic symptoms while being adherent with 100 mg/d of clozapine. She had been started on clozapine about 1.5 years previously because of tardive dyskinesia with risperidone. She had not developed features suggestive of RLS while on risperidone. She had tolerated clozapine well and had never received clozapine more than 100 mg/d. After educating the patient and her caregivers, the dose of clozapine was gradually increased to 300 mg/d (given in divided doses of 100 mg in the morning and 200 mg at bedtime) over a period of 4–6 weeks. Adverse effects of clozapine were regularly monitored and her psychotic symptoms started improving gradually. Once the dose of clozapine was increased from 250 mg/d to 300 mg/d, within 2–3 days the patient developed features suggestive of RLS. She reported an intense urge to move the legs in the evening and night associated with an uncomfortable sensation in the legs below the knee joint. These symptoms would start by around 7 pm, even before she would ingest the bedtime dose of clozapine, and increase in severity further till midnight. She would pace for few hours before getting sleep and frequently would get up from the lying position to walk, as that would relieve these symptoms. She had difficulty both in initiation and maintenance of sleep and was sleeping only 4–5 hours/day in comparison with 7–8 hours/day previously. There was no history suggestive of akathisia. On structured evaluation she fulfilled the IRLSSG diagnostic criteria for RLS [Allen et al. 2003]. On rating with IRLS, she scored 22 amounting to severe intensity of RLS. Naranjo adverse drug reaction probability scale score was 6 suggesting clozapine-induced RLS to be probable [Naranjo et al. 1981]. There was no past or family history suggestive of RLS.

Mrs A had no physical comorbidities. General physical examination and detailed systemic examination was unremarkable. Full blood picture, peripheral smear, liver and renal functions, and blood glucose levels were within reference limits.

After discussing with the patient and her caregivers, the dose of clozapine was not decreased in view of the possibility of worsening of psychotic symptoms. She was started on gabapentin 100 mg orally at night. A marked reduction in her RLS symptoms was noted within the first few days of treatment. Her IRLS score at the end of 1 month of gabapentin was 1. She has been visiting our clinic regularly and has been doing well on the combination of clozapine and gabapentin for the last 3 months.

Discussion and review of literature

Antipsychotics and RLS

RLS is etiologically classified as primary (idiopathic) and secondary RLS. Antipsychotic agents have been reported to exacerbate or cause RLS [Allen, 2014]. Currently, RLS is proposed to be a hyper-dopaminergic state with increased dopamine production and synaptic transmission [Allen, 2015]. Regional brain iron deficiency and a genetic predisposition are important factors associated with the hyper-dopaminergic state in RLS. The down-regulation of striatal dopamine receptors secondary to a hyper-dopaminergic state and the circadian activity pattern of dopaminergic neurons leads to a hypo-dopaminergic state in the brain during evening and night causing RLS symptoms [Allen, 2015]. The antipsychotic agents, dopamine blockers, have been reported to cause or worsen RLS [Allen, 2014: 201]. In a study by Kang and colleagues, prevalence of RLS was around 21% in 182 treated schizophrenia patients, significantly higher than the control group [Kang et al. 2007]. There are several published case series and case reports on antipsychotic-induced RLS [Rittmannsberger and Werl, 2013; Zhao et al. 2014]. Genetic studies have shown polymorphisms of BTBD9 and CLOCK genes to be associated with antipsychotic-induced RLS [Jung et al. 2014; Kang et al. 2013]. Contrary to these findings Jagota and colleagues did not find any significant increase in prevalence of RLS in schizophrenia patients treated with antipsychotics [Jagota et al. 2012].

Dopamine receptor occupancy of antipsychotics and induction of RLS

Both the patients reported here had not developed RLS with antipsychotics having higher dopamine receptor occupancy (haloperidol in the first case and risperidone in the second case) but developed RLS with clozapine, a weak dopamine-receptor blocker. On a literature search, we could come across only three reports of first-generation antipsychotics inducing RLS [Horiguchi et al. 1999; Montplaisir et al. 1991; Perez-Lloret et al. 2012] and second-generation antipsychotics are more often reported to induce RLS. Among the second-generation antipsychotics, clozapine, quetiapine and olanzapine, which have lower dopamine receptor occupancy, are often reported to be associated with RLS [Duggal and Mendhekar, 2007; Raveendranathan et al. 2013; Rittmannsberger and Werl, 2013; Zhao et al. 2014]. In addition, there are published reports of high potency antipsychotics not inducing RLS when replaced for RLS-inducing low potency antipsychotics [Zhao et al. 2014] and low potency antipsychotics inducing RLS in patients who had tolerated high potency antipsychotics without RLS [Duggal and Mendhekar, 2007; Raveendranathan et al. 2013]. Also clozapine, olanzapine and quetiapine have significant anti-histaminic properties, since anti-histaminics are also known to induce RLS [Allen, 2014]. These findings suggest that the extent of dopamine receptor blockade by antipsychotics may not have positive correlation with their RLS-inducing or worsening property. Dopamine receptor blockade may not be the only mechanism by which antipsychotics induce RLS and the hypo-dopaminergic state in antipsychotic-induced RLS may be the final result of many intermediate processes involving different neurochemicals.

Clozapine-induced RLS

In both the patients reported here, clozapine seems to be the agent associated with induction of RLS. In the first case report, temporal correlation of onset of RLS symptoms with initiation of clozapine suggested that it was clozapine, which induced RLS not haloperidol. She had received higher doses of haloperidol (up to 20 mg/d) before clozapine was initiated without having RLS symptoms. Although thyroid dysfunction has been hypothesized to be associated with RLS [Pereira et al. 2010], it was unlikely to be the culprit in this patient as she was diagnosed with and started on appropriate treatment for hypothyroidism long before the onset of RLS symptoms. Her serum TSH levels were within normal limits. In the second case report, onset of RLS symptoms temporally correlated with an increase in the dose of clozapine and she was on monotherapy with clozapine in the absence of any physical comorbidities. As clinical examination and peripheral smear did not indicate iron deficiency, blood iron parameters were not assessed because of affordability issues. Since iron deficiency is a proven risk factor for restless legs syndrome, unavailability of blood iron parameters in our patients is a limitation of this report.

We did a literature search in Medline, PsychINFO, using the key word ‘clozapine’, in conjunction with ‘restless legs syndrome’, and ‘Willis-Ekbom disease’. We came across four cases of clozapine-induced RLS [Chathanchirayil, 2011; Duggal and Mendhekar, 2007; John et al. 2014; Raveendranathan et al. 2013] and the salient features of these reports are shown in Table 1. Patient 5 and 6 in Table 1 refer to the patients in the current report.

Table 1.

Summary of clozapine-induced restless legs syndrome cases.

Reference	Sex/age (years)	Diagnosis	Clozapine dose*	Dose response	Co - medications	Management of RLS
Duggal and Mendhekar, [2007]	Male, 26	Bipolar disorder	50 mg/d	Not mentioned	Valproate (1400 mg/d)	Replaced clozapine with olanzapine
Chathanchirayil, [2011]	Female, 29	Schizophrenia	325 mg/d	Not mentioned	Amisulpride (200 mg/d)	Diazepam (5 mg/d) No sustained improvement
Raveendranathan et al. [2013]	Female, 34	Schizophrenia	100 mg/d	Not mentioned	None	Aripiprazole (10 mg/d)
John et al. [2014]	Female, 28	Recurrent depressive disorder	300 mg/d	225 mg/d No change	Citalopram (80 g/d)	Pramipexole (0.25 mg/d)
Report 1	Female, 32	Schizophrenia	225 mg/d	375 mg/d No change	Haloperidol (2.5 mg/d)	Gabapentin (100 mg/d)
Report 2	Female, 45	Schizophrenia	300 mg/d	100 mg/d No RLS symptoms	None	Gabapentin (100 mg/d)

^*Clozapine dose, dose of clozapine at which RLS noted.

RLS, restless legs syndrome.

Dose response relationship of clozapine with RLS symptoms

The literature on dose response relationship of antipsychotic agents and RLS severity is inconclusive. The review of olanzapine and quetiapine-induced RLS reports suggests it to be a dose-dependent phenomenon [Rittmannsberger and Werl, 2013; Zhao et al. 2014]. In our first patient no change in RLS symptom severity was noted with the reduction of clozapine dose from 375 mg/d to 225 mg/d. In our second case, the patient had tolerated clozapine 100 mg/d for a long time without RLS symptoms. She developed RLS symptoms only when the dose of clozapine was increased, indicating that induction of RLS is a dose-dependent phenomenon. We could neither increase nor decrease the dose of clozapine further in our patients to assess its impact on RLS severity. As summarized in Table 1, clozapine has been reported to induce RLS in the dose range between 50–325 mg/d, suggesting that the threshold for induction of RLS is different in different individuals. Only a single report among the four published reports mentions the dose response relationship and in that report clozapine dose reduction did not have any effect on RLS symptom severity [John et al. 2014]. Kang and colleagues in their study on 182 patients with schizophrenia did not find any relationship between chlorpromazine equivalents and RLS score [Kang et al. 2007]. Further research is required before any definitive conclusion can be drawn regarding dose response relationship of clozapine and RLS symptoms.

Treatment of clozapine-induced RLS

Dopamine agonists, antiepileptic medications, benzodiazepines, opioids, and iron therapy are used in the treatment of primary RLS [Allen, 2014]. No definitive guidelines are available for the treatment of drug/antipsychotic-induced RLS. Deak and Winkelman suggest replacement with a suitable alternative medication wherever possible [Deak and Winkelman, 2010]. Duggal and Mendhekar have reported successful management of clozapine-induced RLS by replacing it with olanzapine [Duggal and Mendhekar, 2007]. Dopamine agonists, the effective first-line treatment of RLS, carry the risk of inducing or worsening psychosis [Garcia-Borreguero et al. 2013; Stoner et al. 2009]. Deak and Winkelman also suggest that in patients with psychosis having antipsychotic-induced RLS, dopamine agonists should be tried only after other approaches have been used [Deak and Winkelman, 2010]. Both of our patients were on clozapine for treatment-resistant schizophrenia. Replacing the clozapine with other antipsychotic agent had the risk of worsening of psychotic symptoms. After discussing with the caregivers we initiated our patients on gabapentin. Gabapentin, an alpha-2 delta ligand, has been shown to improve RLS symptoms and unlike dopamine agonists, augmentation in RLS has not been reported with gabapentin [Trenkwalder et al. 2015]. The recommended dose range for the treatment of primary RLS is between 300–1800 mg/d [Trenkwalder et al. 2015]. In both of our patients, RLS symptoms improved significantly with low-dose gabapentin (100 mg/d). Also, in our earlier report, olanzapine-induced RLS was successfully treated with 100 mg/d of gabapentin [Kumar et al. 2014] so it may be that the dose requirement of gabapentin is much lower to treat antipsychotic-induced RLS. In view of the risk of worsening psychotic symptoms with dopamine agonists, gabapentin may be a safer option in the treatment of antipsychotic-induced RLS in patients with schizophrenia. However, such a conclusion is very much premature due to lack of research data; this underscores the need for systematic research studies in this area. The mechanism of action of gabapentin in the treatment of RLS is not known and the research on the same may provide a much better understanding of the pathophysiology of RLS. As summarized in Table 1, case reports mention successful management of clozapine-induced RLS with aripiprazole and low dose pramipexole [John et al. 2014; Raveendranathan et al. 2013].

Conclusion

The presented cases show that antipsychotics including clozapine may induce RLS. Our first case remained undiagnosed and untreated for 2.5 years despite frequent visits to primary care physicians and psychiatrists. Despite clear diagnostic criteria, RLS is markedly underdiagnosed in primary care practice [Allen et al. 2010]. Hence, primary care physicians should be sensitized about the diagnosis of RLS and psychiatrists should be aware of this side effect and know the treatment of it. Gabapentin, a nondopaminergic agent, is one of the available therapeutic options for the treatment of RLS. Moreover, in patients with psychosis who carry the risk of worsening of psychotic symptoms with dopamine agonists, low-dose gabapentin may be a safe and useful treatment option for clozapine-induced RLS.