General principles
While antipsychotic drugs remain the cornerstone of treatment in child and adolescent onset schizophrenia, all young patients with schizophrenia require a multimodal treatment package that includes pharmacotherapy,  family and individual counselling, education about the illness and provision to meet social and educational needs (Clark & Lewis 1998).

Primary prevention and early detection
In theory at least, the onset of schizophrenia could be prevented if an intervention reduced the premorbid ‘risk’ status. However, the difficulty with the premorbid phenotype as currently conceived (subtle social and developmental impairments)  is its extremely low specificity and positive predictive value for schizophrenia in the general population, assuming that these premorbid features are a causal risk factor (Erlenmeyer-Kimling et al. 2000). Future refinement of the premorbid phenotype is likely to include genetic and neurocognitive markers in order to achieve acceptable sensitivity and specificity. At present, primary prevention remains on the distant horizon.

In contrast to primary prevention, the aims of early detection are to identify the onset of deterioration in vulnerable individuals with a high predictive validity. Predictive power increases markedly in adolescence around the onset of the prodrome (Davidson et al. 1999).

Recent work has attempted to identify ‘high-risk’ or early prodromal states with the aim of intervening to prevent the active phase of schizophrenia (McGrorry & Sing 1995; Yung et al. 1996). However, only about one-fifth of these ‘high-risk’ cases go on to develop frank psychosis and it has proved impossible to distinguish these ‘high-risk’ cases from others who remain non-psychotic. Clearly, interventions directed at ‘high-risk’ or prodromal states need to benefit the whole population at risk,  the majority of whom will not develop schizophrenia. A pragmatic stance would be to monitor children and adolescents with a strong family history and/or suggestive prodromal symptoms to ensure prompt treatment of psychosis.

Strong claims have been made that early recognition and treatment of psychotic symptoms in schizophrenia improves outcome.

The association between a long duration of untreated psychosis (DUP) and poor long-term outcome in schizophrenia (Loebel et al. 1992; Wyatt 1995; Birchwood et al. 1997) supports this view.

A similar association has been found in child and adolescent onset psychoses (Hollis 1999). While the association between DUP and poor outcome seems secure,  the causal connection is far less certain.  DUP is also associated with insidious onset and negative symptoms which could confound links with poor outcome. While there are good a priori clinical reasons for the early treatment of symptoms to relieve distress and prevent secondary impairments, as yet it remains unproven whether early intervention actually alters the long-term course of schizophrenia.

Pharmacological treatments
Because of the very small number of trials of antipsychotics conducted with child and adolescent patients, it is necessary to extrapolate most evidence on drug efficacy from studies in adults.

This seems a reasonable approach given that schizophrenia is essentially the same disorder whether it has onset in childhood or adult life. However, age-specific factors such as the greater risk of extrapyramidal side-effects (EPSs) and treatment resistance to traditional antipsychotics in younger patients (Kumra et al. 1998b) should also influence drug choice.

The typical antipsychotic haloperidol has been shown to be superior to placebo in two double-blind controlled trials of children and adolescents with schizophrenia (Pool et al.  1976; Spencer & Campbell 1994). It is estimated that about 70% of patients show good or partial response to antipsychotic treatment, although this may take 6 - 8 weeks to be apparent (Clark & Lewis 1998). The main drawbacks concerning the use of high-potency typical antipsychotics such as haloperidol in children and adolescents is the high risk of EPSs (produced by D₂ blockade of the nigrostriatal pathway), tardive dyskinesia and the lack of effect against negative symptoms and cognitive impairment. Treatment with typical antipsychotics is also associated with enlargement of the caudate nucleus which can be reversed with clozapine (Frazier et al. 1996). Clozapine (the prototypic atypical) has been shown to be superior to haloperidol in a double-blind trial of 21 cases of childhood onset schizophrenia (Kumra et al. 1996). Large open clinical trials of clozapine confirm its effectiveness in child and adolescent onset schizophrenia (Siefen & Remschmidt 1986; Remschmidt et al. 1994). Similar, although less marked, benefits of olanzepine over typical antipsychotics in childhood onset schizophrenia have been reported (Kumra et al. 1998a).

Drawing this evidence together, a strong case can be made for the first-line use of atypicals in child and adolescent schizophrenia (clozapine is only licensed in the UK for treatment-resistant schizophrenia).  Treatment resistance in child and adolescent patients should be defined as follows:

1  non-response with at least two conventional antipsychotics (from different chemical classes)  each used for at least 4 - 6 weeks; and/or

2  significant adverse effects with conventional antipsychotics.

While atypicals reduce the risk of EPSs, they can produce other troublesome side-effects (usually dose-related) including weight gain (olanzapine), sedation, hypersalivation and seizures (clozapine). The risk of blood dyscrasias on clozapine is effectively managed by mandatory routine blood monitoring. However, knowledge about potential adverse reactions with the newest atypicals is very limited in child and adolescent patients.  A further consideration is the cost of newer atypicals compared with traditional antipsychotics. In the UK, a 1-month supply of haloperidol costs less than £2, compared with £100 - 120 for the newer atypicals and £200 for clozapine. Although economic studies of cost-effectiveness have suggested that the costs of the atypicals are recouped in reduced inpatient stays and indirect social costs (Aitchison & Kerwin 1997), the availability of these drugs, particularly in developing countries, may well be limited because of their high cost. In the late 1990s, the use of atypical antipsychotics by UK child psychiatrists was still low. Over a 2-year period, only 10% of child and adolescent psychiatrists who prescribed antipsychotics in the Trent Health Region had used an atypical drug (Slaveska et al. 1998).

Currently, there is no clear consensus about the choice of antipsychotics in children and adolescents with schizophrenia.

Some authorities suggest starting with a trial of a traditional antipsychotic (e.g. haloperidol) with substitution of an atypical if the traditional antipsychotic is either not tolerated or ineffective after 6 - 8 weeks (Clark & Lewis 1998). However, clinical trial evidence suggests that clozapine is the most effective antipsychotic in child and adolescent onset schizophrenia, although its use is restricted to treatment-resistant cases.

A very powerful case can be made for using atypicals such as olanzapine, quetiapine or risperidone as a first-line treatment, given that child and adolescent onset schizophrenia is characterized by negative symptoms, cognitive impairments, sensitivity to EPSs and relative resistance to traditional antipsychotics.