Childhood-Onset Schizophrenia

We commend Kumra and associates¹ for their contribution to the scant literature on the treatment of childhood schizophrenia with atypical antipsychotics. Their controlled study found that clozapine (dosage range, 25-525 mg/d) is more effective than haloperidol in treatment-resistant childhood-onset schizophrenia. However, controlled studies may not detect unusual phenomena, such as individuals who require a very high clozapine dosage for adequate therapeutic response. We therefore report a case of childhood-onset schizophrenia in a youth who needed a clozapine dosage of 975 mg/d for optimal response. This case is also presented for heuristic purposes: Is it possible that in some cases inadequate response to clozapine may be due to inadequate dosage? We would be interested in the comments of Kumra et al.

A 15-year-old boy with paranoid schizophrenia (DSM-III-R criteria) was transferred to a residential care unit under our care from another such facility. He was receiving clozapine, 975 mg/d (325 mg, 3 times daily; serum level, 130 ng/mL) and phenobarbital, 100 mg twice daily. Weekly blood cell counts were normal and he had no significant adverse effects from the medication. His thought processes were well-organized, but he appeared to be suspicious.

Prior to clozapine treatment and since he was 8 years old, the patient had daily auditory command hallucinations. In response to these hallucinations, he was frequently physically aggressive or self-abusive. At age 9 years, he had attacked his mother with a knife, responding to command hallucinations. His symptoms did not respond to trials of several psychotropics including methylphenidate, imipramine, lithium, valproate, chlorpromazine, haloperidol, risperidone, and a risperidone-haloperidol combination. Clozapine therapy was started, and the dosage was rapidly increased to 100 mg, 3 times daily. He continued to hallucinate and was hospitalized for the second time. The clozapine dosage was increased to 300 mg, twice daily, which resulted in seizures. The seizures were controlled by phenobarbital, 100 mg twice daily. Medical examinations, including an electroencephalogram, revealed no abnormalities. An intellectual assessment revealed borderline IQ (full-scale IQ=78, performance IQ=70, verbal IQ=90). On discharge to the residential home, he still had occasional hallucinations that resulted in a dosage increase to 325 mg, 3 times daily, and his hallucinations remitted completely.

At his present residential facility, no medication changes were made. He continues to receive clozapine, 325 mg, 3 times daily, which is higher than the upper limit recommended in the package insert (900 mg/d for adults).² At 1-year follow-up, the patient continues to be free of hallucinations. There have been no suicide attempts or aggressive episodes.

We recognize the limitations of the case-study method, but case studies can be a source of ideas and hypotheses. Our patient did not have a typical case of childhood-onset schizophrenia and had a verbal performance IQ discrepancy. A unique feature of this case is the dramatic response to high-dose clozapine therapy. The treatment response in the sample of Kumra et al was apparently variable. Their patients had severe psychopathological symptoms as indicated by a mean (SD) baseline Brief Psychiatric Rating Scale (BPRS) score of 83.7 (14.0) in the clozapine-treated group and 84.7 (17.6) in the haloperidol-treated group, declining to 52.5 (12.6) and 64.7 (18.1), respectively, at week 6. We believe that the 6-week BPRS scores in both groups still suggest severe symptoms, despite the significant drop in BPRS score from baseline. We wonder if dosage increase would have further benefited some of these patients. Kumra et al have reported on long-term follow-up of the study patients and indicate a 38% dropout rate owing to various adverse effects. That still leaves some patients who could benefit from a higher dosage. We wonder if the dosage was increased in their patients after the study and if their experience was similar to ours.

In our patient, there was a decrease in serum clozapine level as phenobarbital was introduced and the dosage adjusted upward over time. Our patient's clozapine dosage was adjusted upwards as well. The serum clozapine level decreased from 483 ng/mL to the present level of 130 ng/mL. It would be interesting to know if the serum clozapine levels in the study by Kumra et al had a relationship to treatment response, as is known to occur in the literature on adult patients.³

While interindividual differences in absorption and metabolism might explain differences in dosage requirements, we raise the question if there are cases of treatment-resistant childhood-onset schizophrenia that require higher doses per se (perhaps because they differ from "garden-variety" cases with respect to brain pathophysiology). We agree with Kumra et al that childhood-onset schizophrenia remains poorly understood. Their study provides a much-needed basis for rational clinical practice.