Olanzapine dosing above the licensed range is more efficacious than lower doses: fact or fiction?

A substantial number of patients with schizophrenia or bipolar disorder receive olanzapine in amounts that are greater than what is recommended in the product labeling approved by drug regulatory agencies. The purpose of this review is to describe the evidence supporting the use of olanzapine in excess of 20 mg/day. PubMed was queried using the keywords 'olanzapine' and 'dose' or 'dosing' for all English-language articles published between 1990 and December 2008, inclusive of articles that describe utilization of olanzapine at doses in excess of 20 mg/day. Also queried was clinicaltrials.gov for studies involving 'olanzapine'. Efficacy and safety data were extracted from case reports, case series, observational studies and double-blind, randomized clinical trials. We found that the use of olanzapine at doses greater than 20 mg/day appears to be increasing. Among patients hospitalized for intermediate and long-term care in New York State psychiatric centers in the period from 1997 to 2003, the average dose of olanzapine increased from 17.4 to 22.5 mg/day. The percentage of patients receiving olanzapine at a dose exceeding 20 mg/day increased from 16.2 to over 50% from 1997 to 2006. Case reports of patients receiving doses up to 60 mg/day describe a favorable benefit-risk ratio. Double-blind clinical trials that have examined doses of olanzapine greater than 20 mg/day are limited in number, but suggest that these doses may be helpful in selected patients who are treatment resistant, have high levels of psychopathology or who are acutely agitated. This must be balanced by an increased risk of weight gain and elevated prolactin that was observed among those receiving 40 mg/day in a large randomized clinical trial comparing doses of 40 versus 20 versus 10 mg/day. In conclusion, dosing of olanzapine in clinical practice is higher than what has been established in the registration program for schizophrenia or bipolar disorder. This is somewhat supported by double-blind, controlled clinical trial evidence, but only for selected patients with severe and/or persistent symptoms.