Tardive dyskinesia in a pregnant woman with low dose, short-duration risperidone: possible role of estrogen-induced dopaminergic hypersensitivity

“Ms. S,” a 35-year-old married woman, had a known case of bipolar disorder-I without any comorbid medical or psychiatric disorder, substance use, or family history of psychiatric illness. The illness started in 1997 with a depressive episode of 7 months’ duration, characterized by depressive mood with anhedonia; ideas of hopelessness; persecutory delusion; decreased sleep, appetite, and personal care; irritability; occasional anger outbursts; and psychomotor agitation. She was treated with imipramine and chlorpromazine up to 200 mg each. Chlorpromazine was stopped after resolution of psychotic symptoms, while imipramine was continued.

In 1999, while the patient was on imipramine, she had a manic episode with psychotic features of 3 months’ duration, for which she required inpatient treatment for 50 days. Imipramine was stopped, and she was treated with lithium and haloperidol up to 1,500 mg and 10 mg, respectively. Haloperidol was stopped after resolution of psychotic symptoms and lithium after 18 months. Ms. S remained asymptomatic for the next 30 months.

Subsequently, she had three depressive episodes, in 2002, 2004, and 2007, lasting for 2 months, 3 months, and 1 month, respectively. Although the episodes in 2002 and 2007 occurred after lithium discontinuation, the episode in 2004 occurred despite an adequate dose of lithium (900 mg–1,050 mg per day). In each of these episodes, antidepressants were added during the symptomatic period and later tapered and stopped after the symptoms resolved. In 2004, the patient also had psychotic symptoms, for which risperidone 1 mg was added, but it was withdrawn after 1 month because of emergence of acute extrapyramidal symptoms (EPS). The patient maintained adequate compliance and never reported any significant side effects with lithium or antidepressants.

In 2011, lithium was tapered and stopped as the patient planned a pregnancy. One month after conception, she had a manic episode with psychotic symptoms. Lithium was not started because of the possibility of teratogenic effects in the first trimester. Quetiapine was initiated and increased up to 600 mg, but was not found effective. Next, olanzapine up to 20 mg was tried, with adequate clinical response, but had to be stopped because of progressive blood sugar derangement. Finally, risperidone was restarted and increased up to 2 mg. However, although her symptoms improved, she started having dyskinetic movements in the orofacial region. TD was diagnosed and objectively evaluated with the Abnormal Involuntary Movement Scale (AIMS),⁸ and showed a total score of 10, with moderate severity in perioral and facial areas. Risperidone was stopped within 1 week, with a cumulative total dose of only 20 mg. Lithium was restarted and increased up to 900 mg as the patient entered the second trimester of pregnancy. The manic symptoms resolved within the next 3 months. Ultrasonography at 16 weeks revealed no fetal anomaly. However, the symptoms of TD did not resolve, even after 3 months of having stopped risperidone. Neurological opinion did not reveal any abnormality. Clonazepam was started and increased up to 4 mg but was discontinued by the patient because of excessive sedation. Tetrabenazine was not started because it is contraindicated during pregnancy and lactation and because of its potential to cause/precipitate depressive symptoms. The dyskinetic movement continued to increase in severity, and also involved muscles of the neck and shoulder, causing moderate-to-severe dysfunction, and the AIMS score increased from 16 to 20. Later, Vitamin E 400 mg–600 mg/day was started and was tolerated by the patient, but the dyskinetic movements showed no significant response. Two months later, in July 2012, the patient delivered her child without any major complication and started breast-feeding. In the subsequent follow-up about 1 month after delivery, the dyskinetic movements did not show any further progression and had decreased in severity, and AIMS score decreased from 20 to 15. Over the next 2 months, the dyskinetic movement decreased further, and AIMS score decreased to 8. Later Vitamin E was increased to 1,200 mg, along with lithium 900 mg. The patient continued to have mild orofacial dyskinesia at the time of follow-up 6 months after childbirth. The dyskinesia worsens intermittently during situations characterized by emotional excitement and sleep deprivation.

In this case, the dyskinetic movements do not fulfill the essential minimum drug duration criterion of DSM-IV-TR or Schooler and Kane criteria.⁷ However the nonrhythmic, repetitive, choreo-athetoid nature of movements increasing with emotional arousal and distraction and diminishing with relaxation, sleep, or volitional effort justifies our diagnosis. Some of the risk factors for TD, such as female gender, bipolar disorder as primary psychiatric illness, and past history of sensitivity to risperidone were also present in this case. Spontaneous dyskinesia is unlikely, as the patient was not drug-naïve when the dyskinesia started. The role of lithium is also unlikely because the patient was off this medication for more than 2 months before TD started, and, also, uncertainty exists in the literature about its exact role in TD.⁶ Withdrawal TD caused by quetiapine and olanzapine administered before risperidone cannot be ruled out. However, to our knowledge, withdrawal TD has not been reported with these antipsychotics. Last, a clear temporal association between risperidone initiation and dyskinetic symptoms exists in this case. However, even with risperidone, there are only three case reports of TD in young subjects at a dose as low as in our case.^9–11 All these cases were of non-affective psychosis. In all of these cases, TD started at 2 mg–4 mg/day of risperidone, and dyskinesia resolved completely within 2–4 months of medication withdrawal in two cases and in 24 months in one case. Partial remission of TD was reported only in cases with concomitant neurological lesion.^10,11 Thus, these cases differ from our case with respect to the psychiatric diagnosis and the effect of withdrawing risperidone on course of TD.

Second, in our case, a temporal association between pregnancy and the onset and course of TD was also noted, which, according to our knowledge, has not been reported earlier. This aggravating effect of pregnancy on TD can be explained on the basis of the hyper-estrogenic state in pregnancy, which affects the central dopaminergic system.¹² The role of estrogen is proposed in the mechanism for chorea gravidarum,¹³ but never been mentioned for TD. However, as the two conditions are similar with regard to neurobiological mechanism, hyper-estrogenic state may also be the reason for heightened sensitivity to risperidone-induced TD and the progressive course observed in this case. There are limited studies on role of estrogen in TD, although it has been implicated in multiple neuropsychiatric disorders via dopamine modulation.

Women in the reproductive age-group are always considered to be at low risk group for TD. However, there is a need to look into the biological basis for the risk factors of TD. Furthermore, it is essential to study the neurobiological implications of the hormonal changes that occur in pregnancy and their significance in psychopharmacology in women.