Relapse with oral antipsychotics versus long-acting injectable antipsychotics: new paradoxical findings

What is already known about this topic?

Earlier randomised controlled trials (RCTs) for first-generation antipsychotics (FGA) comparing long-acting injectable antipsychotics (LAIs) to oral antipsychotics (OAPs) suggested FGA-LAIs were superior at preventing relapse. By delivering a full-recommended dose on a continuous basis, LAIs would seem to overcome adherence issues associated with OAPs.¹

What does this paper add?

• Despite non-adherence in some OAP patients, OAP and LAI patients did not differ in relapse rates. This updated review completely overturns the authors' previous meta-analysis.²

• By improving methodology and including four recent RCTs that used second-generation antipsychotics (SGAs), this updated review completely overturns the authors’ previous meta-analysis, published in 2011, which seemed to support the older belief that LAIs were associated with significantly lower relapse rates than OAPs.²

• The current 2014 paper added a control for publication bias. In RCTs, conducted prior to 1991, unpublished negative findings were not registered. The positive difference in relapse rates between LAIs and OAPs in these studies may have been an artefact of this bias.

• Smaller than recommended doses of OAPs may be equally effective.

Limitations

• There was much heterogeneity among the relapse criteria used by the different RCTs. Combining RCTs with differing severity of relapse made it impossible to distinguish whether LAIs prevented more severe relapses than OAPs.

• Also, naturalistic studies, where the clinician can document very poor compliance before starting patients on LAIs, consistently show decreased relapse rates for LAIs compared with OAPs.

What next in research?

• To resolve the paradox, research needs to be directed towards the ‘less is more’ hypothesis of antipsychotic medication efficacy. In this hypothesis, McGorry, Alvarez-Jimenez and others argue that after remission, no or low-dose antipsychotics may lead to better clinical outcomes, fitting with the biological research of Seeman and others on dopamine dynamics.

• This research must control for the ‘medication discontinuation syndrome’ among placebo controls. The time gap between when a patient goes off an antipsychotic and prior to entering a placebo control group must be long enough to eliminate medication discontinuation induced rebound psychosis from occurring during the drug trial period.^{3 4}

• Account should be taken of the findings of various studies which suggest that dose reduction and antipsychotic discontinuation schemes may be beneficial, and that long-term antipsychotic prescribing may be less effective, ineffective or even harmful for some patients, compared to lower doses or no doses.^{3 4}

Could these results change your practice?

Yes, because patients and clinicians are reluctant to use LAIs because of the stigma, the needle pain, the side-effect concerns and the cost. If so, LAIs should be used in cases where adherence to OAPs is proven to be very low or negligible.