Lithium as a rescue therapy for regression and catatonia features in two SHANK3 patients with autism spectrum disorder: case reports

Patient 1 is a 21-year-old man with ASD diagnosis based on the DSM-5 criteria as well as on the Autism Diagnostic Interview-Revised (ADI-R) [10] and the Autism Diagnostic Observation Schedule (ADOS) [11]. He is the second child of non-consanguineous parents and had no perinatal history. He was first evaluated at age 6 (see Table 1). At the verbal level, his language was limited to simple vocabulary of everyday life, with literal comprehension. He used words and short sentences to make a request and presented echolalia and stereotypic language. At the cognitive level, he has severe intellectual disability. At the motor level, he had no early psychomotor delay and liked to do sports and practiced biking for several hours a week. At age 13, as he moved from usual autism day-care centre to a non-specific autism unit, major clinical changes were observed.

In the following, patient 1 progressively reduced motor and verbal initiative, stopped his favourite activities, presented a language regression and lost autonomy skills. Additionally, behavioural disorders (impulsive acts, unpredictable gesture and opposition) alternated with apathy. At last, sleep disturbances and insomnia appeared.

Patient 1 received different psychotropic medications which were partially effective but accompanied by multiple side effects. (1) Antipsychotics were able to control aggressive behaviours and insomnia, but induced catatonia with positive evaluation by Bush Francis Catatonia Rating Scale (BFCRS) [12] and elevated muscle enzymes (creatine phosphokinase = 1286 UI/l, normal <175, and lactate dehydrogenase = 735 UI/l, normal: 200–480). (2) Benzodiazepines reduced catatonia symptoms but didn’t control impulsive actions, and induced verbal and motor excitement, confusion and insomnia. (3) Mood stabilizer drugs had no clinical efficacy and induced somatic problems such as cytolysis with valproic acid, and DRESS syndrome (Drug Reaction with Eosinophilia and Systemic Symptoms) with carbamazepine. (4) Antidepressants didn’t show any efficacy. (5) Among multiple drug associations only the co-therapy of aripiprazole and clonazepam showed partial efficacy improving regression, autonomy, language and motor skills for about 12 months, but without behaviour stabilization. (5) Methylphenidate was not effective and induced insomnia, violence and shouting. (6) Lithium was initially introduced in association with other psychotropic drugs (e.g. methylphenidate) and for a short period (<2 months), without clear clinical effects.

As there was no symptoms improvement, the diagnosis of atypical bipolar disorders was suspected. Therefore, it was decided to re-introduce lithium (1500 mg/day), only associated with melatonin (4 mg/day). In the following, symptoms progressively stabilized during the next 3 months. Agitation, hetero-aggressiveness and impulsivity disappeared. The patient regained autonomy in everyday life, and urinary and faecal incontinence stopped. In addition, the patient did no longer show opposition and participated again in different activities with other patients. Finally, he began riding his bike again. As the patient did no longer present sleep disturbances, melatonin could be stopped after 3 months of lithium therapy. With a follow-up of 2 years, lithium treatment permitted to reverse regression and to stabilize behavioural disorders and is well tolerated (1500 mg/day, blood lithium level =0.8 mEq/l, normal: 0.6-1). One year after the beginning of lithium therapy, the patient integrated a unit of life for adults. He recovered a level of functioning similar to the level he had before regression (see Figure 1 and Figure 2).

Patient 2 is a 17-year-old female adolescent with ASD diagnosis based on the DSM-5 criteria as well as on ADI-R and ADOS. She has first evaluation at age 9 (see Table 1). She is the only child of non-consanguineous parents without perinatal history. At the verbal level, she presented echolalia and stereotypic language with words and short sentences used in everyday life and in context. At the cognitive level, she has severe intellectual disability. At the motor level, she had no early psychomotor delay and she liked sports, especially rock climbing and gym. At age 12, major clinical changes were observed at the same time as she changed the day-care centre for a specific autism unit.

The clinical presentation suggested the diagnosis of catatonia-like symptoms without the full picture of catatonia (negative evaluation by BFCRS and normal blood analyses including muscle enzymes). In line with this hypothesis, we decided to introduce benzodiazepine treatment with clonazepam (0.9 mg/day). The patient rapidly improved and regained motor abilities, normal body posture, language, participation, and autonomy in a few months. Sleep disturbances and faecal incontinence stopped but night urinary incontinence persisted. These improvements continued for about 8 months. However, verbal and motor excitement progressively increased. Surprisingly, the patient progressively lost weight whereas she continued to ate normally (loss of 8 kg in 8 months) and without other clinical symptoms such as vomiting. Her initial body mass index (BMI) of 19.7 decreased to 16.4. We therefore decided to progressively withdrawal clonazepam. During the 4 following months, the patient regained weight (BMI =18.1), but then regressed again by losing motor and verbal skills. At last, she presented once more catatonia-like symptoms including behavioural disorders (hetero-aggressiveness) and insomnia. Clonazepam (0.9 mg/day) was re-introduced. Two days after, motor and verbal excitement, sudden impulsive actions, and total insomnia were observed requiring hospitalization in the inpatient psychiatric unit. Clonazepam was switched to lorazepam (3 mg/day), complicated by confusion, faecal incontinence, agitation and insomnia.

Based on the clinical similarities between patient 1 and 2, our clinical experience with patient 1 and the hypothesis of atypical bipolar disorder, we decided to introduce lithium, associated with very low dose clonazepam (0.2 mg/day). Patient 2 stabilized behavioural symptoms and progressively improved verbal and motor abilities, participation, sleep and autonomy. However, night urinary incontinence persisted. Clonazepam was discontinued after 4 months. After one year with lithium treatment (1000 mg/day, lithium blood level 0.7 mEq/l, normal 0.6-1), the patient recovered a similar level of functioning as before regression (see Figure 3 and Figure 4). Lithium reversed regression and stabilized behavioural disorders without side effects.