Neuroleptic Malignant Syndrome with Low Dose Lithium,Without Concomitant Antipsychotics

To the Editor,

Neuroleptic malignant syndrome (NMS) is a rare, life-threatening condition characterized by fever, rigidity, increased creatine phosphokinase (CPK) level, tachycardia, diaphoresis, blood pressure changes, leukocytosis, and altered mental status. ¹ NMS is almost always associated with the use of antipsychotics, especially first-generation antipsychotics and high dose or parenteral administration, but there are also reports of NMS with atypical antipsychotics. ² Very rarely, NMS may be associated with drugs that do not act predominantly on dopamine; an example to this is lithium, especially in the context of toxicity ³ and rarely with lithium-induced nephrogenic diabetes insipidus (NDI). ⁴ We describe a case of NMS associated with lithium at serum levels well within the therapeutic range and without NDI.

Case Report

The patient was a 52-year-old male with a 16-year history of bipolar disorder and a 3-month history of systemic hypertension, dyslipidemia, and ischemic stroke. He had been receiving oral haloperidol (10 mg/day) and olanzapine (2.5 mg/day) from a mental hospital for at least two to three years, and possibly 10 years. While on the above medications, he had presented to the treating hospital with acute mania and was given parenteral haloperidol (5 mg), oral lithium (600 mg/day), olanzapine, and lorazepam. There was no history of recent treatment with a depot antipsychotic. Two days later, he developed altered consciousness. There was no improvement across a week after withholding lithium and antipsychotics, and he was referred to our tertiary hospital.

On examination, he was conscious but disoriented to time, place, and person. He was restless. He had tachypnea, tachycardia, fever, sweating, and fluctuating blood pressure but no rigidity. Laboratory investigations were generally within normal limits: the total leukocyte count was 9800/mm³ and serum lithium was 0.2 mEq/L. Importantly, serum CPK was 1228 (reference level, 25–192) IU/L. A diagnosis of NMS was made, based on the presence of two major and four minor Levenson’s criteria of NMS.

Lithium and the antipsychotic drugs were stopped, and he was treated with Tab. bromocriptine 15 mg/day, Tab. sodium valproate 400 mg/day, and Tab. lorazepam 4 mg/day along with supportive measures. Bromocriptine was tapered and stopped when the CPK level decreased to 357 IU/L. Twelve days after stopping lithium and antipsychotics, the CPK level was 249 IU/L; NMS had also resolved clinically within a week of stopping the drugs.

Sodium valproate was uptitrated to 1 g/day. Plasma ammonia was checked because of continued restlessness, and the level was found raised (133 µmol/L; lab range, 16–66 µmol/L). Valproate was therefore stopped, and lithium (450 mg/day) was reintroduced 17 days after its discontinuation. No antipsychotics were started.

Five days after lithium reinitiation, fever (100.4º F), sweating, restlessness, tachycardia, tachypnoea, and tremors recurred; however, there was again no rigidity. The total leukocyte count was 13,300/mm³, and serum CPK was 5073 IU/L. The serum lithium level was 0.3 mEq/L, and there were no clinical signs specific to lithium toxicity. Meningitis and encephalitis were ruled out clinically by the physician. It appeared that there was a recurrence of NMS. MRI of the brain revealed gliosis of left parietal and occipital lobes and diffuse parenchymal atrophy, consistent with old cerebrovascular accident. No fresh changes were noted. Chest X-ray, electrocardiogram, abdominal ultrasonography, and other investigations identified no relevant abnormalities, and a presumptive diagnosis of lithium-related NMS was made. Two major and four minor Levenson’s criteria were fulfilled.

Lithium was again stopped, and bromocriptine was resumed. The CPK level dropped to 2700 within a day. The NMS attenuated, and serum CPK progressively fell to 160 IU/L across 10 days.

He was discharged in a relatively stable condition on clonazepam 3 mg/day. However, three weeks later, he required readmission for relapse into mania. He was successfully managed with clozapine and valproate; there was no recurrence of NMS.

On a final note: the patient had been receiving aspirin (150 mg/day), atorvastatin (20 mg/day), and amlodipine (5 mg/day) as part of his medical and neurological management. These drugs were continued throughout the period of NMS and thereafter.

Discussion

There are isolated reports of NMS associated with lithium combined with antipsychotics^5,6 and NMS associated with lithium toxicity. ⁷ Reports of NMS with lithium monotherapy are very rare; in a case report of lithium-induced NMS, ³ the serum lithium level was 1.5 mEq/L, which is above the therapeutic range. Our patient first developed NMS while on haloperidol and lithium, and we considered it to be haloperidol-induced NMS because haloperidol is one of the commonest antipsychotics associated with NMS. However, 22 days later, when lithium was reintroduced without concomitant antipsychotics, our patient developed NMS again, and the NMS resolved after discontinuing lithium. The score on Naranjo Adverse Drug Reaction Probability Scale was 6, indicating probable causality with lithium. ⁸ Our high index of suspicion, early detection, and prompt withdrawal of lithium prevented further clinical deterioration.

Could the second instance of NMS have been a relapse of the first, and, if so, might both have been related to haloperidol, which may have a long half-life in some individuals, particularly during long-term treatment? This is possible but unlikely. There was an interval of at least 10 days when the patient did not have NMS and was not on treatment for NMS. If the second instance of NMS had been a prolongation of the first, it would have emerged much earlier, within a few days of stopping bromocriptine. Furthermore, the CPK response to lithium rechallenge was higher than during the first instance of NMS, and the CPK level dropped by half, one day after lithium was again discontinued.

As far as we could ascertain, the patient had not received lithium before nor had he experienced NMS earlier. Our report indicates that lithium monotherapy at even subtherapeutic serum levels may very rarely be associated with NMS. We speculate that the recent stroke was a predisposing factor, as preexisting brain damage may be a risk factor for lithium-related neurotoxic effects with serum lithium within the therapeutic range. ⁹

On a final note, we suggest that NMS be defined by the spectrum concept ¹⁰ or using a broad definition in order to ensure early detection, especially when it develops with drugs that are not typically associated with the disorder and when not all the classical symptoms (e.g., rigidity, in our patient) are present.

The only reported case of lithium monotherapy causing NMS was at a higher dose of 900 mg and at a supratherapeutic level of 1.5 mEq/L, in a 72-year-old female. ³ Our case report is unique in that this is perhaps the first published report of lithium without concomitant antipsychotics and at a subtherapeutic dose and level being associated with NMS.