Clinician hesitation

Ever since John Kane and colleagues reported that clozapine was superior to chlorpromazine in terms of efficacy for patients with treatment-resistant schizophrenia, ^{Reference Kane, Honigfeld, Singer and Meltzer1} clinicians have debated the relative merits of clozapine as the only licensed drug for use in such patients. Avoidance of – or hesitation before – clozapine initiation is rife, despite current clinical guidelines advocating otherwise. This is most recently evidenced by a naturalistic study for patients commenced on clozapine reported by Howes et al ^{Reference Howes, Vergunst, Gee, McGuire, Kapur and Taylor2} in this issue, which examined the time taken to initiate clozapine following two adequate treatment trials with two different antipsychotics. Treatment episodes for each antipsychotic were categorised as to whether or not each treatment episode was ‘adequate’ in terms of dose and duration of treatment trial, in keeping with guidelines, ³ but evaluation of adherence was not possible. As with any retrospective study design, missing and incomplete data are inevitable and this study excluded a third of eligible patients owing to missing clinical notes, and further disregarded 91/825 treatment episodes for the remaining patients owing to incomplete prescribing data. For patients who received two adequate treatment trials (129/149, 87%), the subsequent mean time to delay for clozapine initiation was 47.7 months (range 0–219), which constitutes a reduction of approximately 1 year in comparison with a similar earlier study. ^{Reference Taylor, Young and Paton4} Howes et al concluded that the delay to clozapine initiation highlights a lack of clinician adherence to national clinical guidelines. ³ However, the recorded clinical reasons for that theoretical delay are notably missing and these might highlight the challenge in achieving two adequate trials on treatment with an antipsychotic in terms of dose, duration and adherence; but one wonders how they might fully account for the upper range of delay of up to 18 years.

Are interim prescribing strategies justifiable?

Although clozapine initiation is avoided, for sound clinical reasons or otherwise, switching multiple times between antipsychotics is commonplace. Howes and colleagues ^{Reference Howes, Vergunst, Gee, McGuire, Kapur and Taylor2} noted that, on average, patients received more than five different antipsychotic treatment episodes before clozapine initiation and that the mean number of different adequate treatment episodes was 2.8, reduced from 4 in the earlier study. ^{Reference Taylor, Young and Paton4} Other interim strategies include use of two or more antipsychotics at the same time as well as above licensed dose prescribing, either for a single antipsychotic or for a combination of two or more regularly prescribed antipsychotics. If ‘as required’ antipsychotics are also considered, then rates of antipsychotic polypharmacy are likely to be even higher. This finding is in keeping with a Danish study, which reported that two-thirds of psychiatrists interviewed would rather combine two antipsychotics than use clozapine. ^{Reference Nielson, Dahm, Lublin and Taylor5} As Howes and colleagues ^{Reference Howes, Vergunst, Gee, McGuire, Kapur and Taylor2} rightly point out, there is little empirical evidence to support either regular polypharmacy or high dosing. At best, one could argue that individualised prescribing using antipsychotic polypharmacy or above licensed doses be viewed as an N of 1 study, taking place after options endorsed in guidelines have been exhausted (including clozapine), and then only with a clearly documented clinical and psychopharmacological rationale, systematic monitoring of symptoms and side-effects and with a view to discontinuation if no clinical benefit is seen after a predetermined period of time.

Fool's gold…

So why are clinicians delaying clozapine initiation? Farooq & Taylor ^{Reference Farooq and Taylor6} highlight a possible perception among clinicians that clozapine is a dangerous therapeutic agent. In the UK, it was noted that clozapine was associated with an increased risk of death and that pneumonia was the most common single cause, followed by lung cancer. ^{Reference Taylor, Douglas-Hall, Olofinjana, Whiskey and Thomas7} Alternatively, in a larger-scale study in Finland, risk for all-cause mortality for antipsychotics was found to be lowest for clozapine, and this was attributed in particular to a lower risk of death from suicide. ^{Reference Tiihonen, Lonnqvist, Wahlbeck, Klaukka, Niskanen and Tanskanen8} Clinicians have been found to rate weight gain, hypersalivation and blood monitoring as the most problematic issues for clozapine but that a quarter of psychiatrists overestimated the risk of agranulocytosis. ^{Reference Nielson, Dahm, Lublin and Taylor5} Assuming that there is a good, supportive service provision for phlebotomy, there may be further uncertainty regarding the appropriateness of forced blood tests on the ward in the context of a compulsory treatment order. In this situation it is critical that the multidisciplinary team is given the opportunity to understand the role that clozapine can play within the care plan and is in agreement with the strategy. For Black and ethnic minority patients with lower than average white cell counts at baseline, clozapine may not be initiated owing to the lack of consideration of benign ethnic neutropenia. ^{Reference Whiskey, Olofinjana and Taylor9} Nonadherence to taking clozapine is also a key concern, ^{Reference Pai and Vella10} not least of all because clozapine retitration can take longer than the period of non-adherence.

Managing the risk of diabetes and cardiovascular side-effects of clozapine requires routine physical health monitoring, including metabolic parameters and an electrocardiogram, but this is now true of all antipsychotics. Pharmacological strategies to assist with some other side-effects include: anti-epileptic prophylaxis for risk of seizures for clozapine concentrations >500 μg/l; hyoscine hydrobromide for hypersalivation; and regularly prescribed laxatives for constipation. ^{3,Reference Taylor, Paton and Kapur11} Aripiprazole 5–10 mg/daily is also sometimes used to counter the effect of the risk of weight gain or oversedation; however, this is not endorsed by current clinical guidelines.

…Or a gold standard?

There is further potential cause for concern that clozapine is not always found to be effective in terms of symptom reduction for patients with treatment-resistant schizophrenia, although plausible explanations include that some studies limited the upper dose of clozapine to 400 mg/day. ^{Reference Farooq and Taylor6,Reference Leucht, Corves, Arbter, Engel, Li and Davis12} In a recently reported study, clozapine was not found to be superior to chlorpromazine, but 30% of individuals randomised to chlorpromazine were also exposed to clozapine during the 9-year follow-up phase. ^{Reference Girgis, Phillips, Li, Li, Jiang and Wu13} On the other hand, in the CUtLASS study, the first of two key randomised controlled trials in the era of second-generation antipsychotics, clozapine was superior with regard to symptom reduction. ^{Reference Lewis, Barnes, Davies, Murray, Dunn and Hayhurst14} In phase II the CATIE study, time to antipsychotic discontinuation due to poor response was significantly longer for those on clozapine v. commonly prescribed second-generation antipsychotics. ^{Reference McEvoy, Liebermann, Stroup, Davis, Meltzer and Rosenheck15} In randomised controlled trials for clozapine, few trials titrate the dose according to plasma concentration levels (target range 350–500 μg/l) ^{Reference Taylor, Paton and Kapur11} which is required due to interindividual metabolic variability caused in part by factors such as smoking. ^{Reference Couchman, Morgan, Spencer and Flanagan16,Reference Couchman, Bowskill, Handley, Patel and Flanagan17} Thus, it is quite possible that suboptimal doses are being evaluated. In a more recent large observational study, clozapine was found to have a significantly lower risk of rehospitalisation than risperidone and other second-generation antipsychotics for patients after discharge from their first hospitalisation with a diagnosis of schizophrenia. ^{Reference Tiihonen, Haukka, Taylor, Haddad, Patel and Korhonen18} None the less, there will always be a subgroup of patients for whom clozapine does not have a therapeutic effect even with a dose that achieves therapeutic plasma concentration levels, and the need for augmentation strategies is warranted. It remains to be determined whether these predominantly include patients with primary treatment resistance or with treatment resistance that emerges over time, possibly as a consequence of inconsistent antipsychotic treatment. ^{Reference Taylor, Young and Paton4}

Targets for changing prescribing behaviour

Even with a gold standard drug, not all patients will be able to tolerate its side-effects or have a good clinical response to it and clozapine is no exception to this rule. That said, studies continue to show clozapine underutilisation and that this varies by region. ^{Reference Nielson, Dahm, Lublin and Taylor5,Reference Downs and Zinkler19} Although there will be good clinical reasons for clozapine not being initiated for some patients, it is hypothesised here that clinicians’ knowledge, attitudes and preferences are more likely predictive factors for explaining the variation in clozapine prescribing rates and that these will need to be addressed before adherence to clinical guidelines for clozapine is likely to improve. Drawing on the literature on potential underutilisation of antipsychotic long-acting injections ^{Reference Patel and David20} and therapeutic drug monitoring for antipsychotic plasma concentration levels, ^{Reference Best-Shaw, Nagar, David, Rose and Patel21} it is anticipated that for clozapine there will be some specific aspects. These are likely to include: clinicians’ perceptions of side-effects and knowledge of the management of side-effects; potentially paternalistic concern regarding patient attitudes to and adherence with clozapine and the associated regular blood tests; service-level logistical restraints, including access to a laboratory for drug plasma concentration levels; clinician ambivalence regarding superior efficacy of clozapine in treatment-resistant illness; and lack of sufficient knowledge and ongoing experiential knowledge of prescribing clozapine. Endorsement by members of the multidisciplinary team also would appear to be critical and this could be compromised by lack of training in psychopharmacology.

The National Audit for Schizophrenia (NAS) ²² for England and Wales is due to report its findings later this month (December 2012) and will shed light on clinician preference, by evaluating whether or not patients are even offered clozapine after two different adequate antipsychotic treatment trials. However, the results will be on a trust-by-trust basis and then aggregated for England and Wales, but it is also within each trust that clinicians’ attitudes to clozapine should be considered as to whether or not clinicians’ attitudes and preferences are the primary cause for hesitation before clozapine initiation. Consequently, methods to improve rates will need to consider aspects of knowledge, attitudes and preferences before a change in prescribing behaviour is likely to occur. This will require local champions with good knowledge in clinical psychopharmacology, ^{Reference Harrison, Baldwin, Barnes, Burns, Ebmeier and Ferrier23} and without hindrance by service-level factors, to address the key concerns within each trust.