One day after Eli Lilly and Company researchers reported finding olanzapine(Drug information on olanzapine) (Zyprexa) superior to risperidone(Drug information on risperidone) (Risperdal) on some measures in a 28-week comparative study presented to the American College of Neuropsychopharmacology (ACNP) meeting in 1996, Janssen Pharmaceutica Inc. issued a press statement critical of the methodology and results of the study. Janssen then announced it would undertake its own comparative trial. The results of this trial's initial eight-week acute treatment phase were presented at the European College of Neuro-psychopharmacology meeting held Oct. 31, 1998, through Nov. 4, 1998, in Paris. They indicate risperidone has an advantage for positive symptoms such as hallucinations and delusions, and for anxiety/depression in patients with schizophrenia and schizo- affective disorder.
The Janssen study associated the drugs with comparable improvements in other symptoms measured by the Positive and Negative Symptom Scale (PANSS), including negative symptoms such as apathy and social withdrawal; and with comparably low rates of extrapyramidal symptoms (EPS). Risperidone was provided in flexible dosing of 2 mg/day to 6 mg/day and olanzapine 5 mg/day to 20 mg/day. In this short-term treatment phase of the study, which will be extended to 52 weeks, risperidone was associated with significantly less weight gain and dry mouth; and olanzapine with less insomnia and headache (Conley et al., 1998).
The Lilly study had evidenced a statistically significant advantage for olanzapine in PANSS total score improvement among patients achieving at least a 30% improved total score (Tran et al., 1997). This study had employed olanzapine in flexible dosing of 10 mg/day to 20 mg/day, and risperidone at 4 mg/day to 12 mg/day. At these doses, a significantly greater proportion of patients receiving olanzapine completed the 28-week study (57.6%) than did patients receiving risperidone (47.3%).
The Lilly study also determined more treatment-emergent events-including palpitations and blurred vision, as well as EPS and remedial medications for EPS-among patients receiving risperidone rather than olanzapine. Lilly researchers have conducted subsequent analyses of these findings of greater efficacy and fewer side effects with olanzapine to further distinguish the antipsychotics in terms of treatment cost-effectiveness (see accompanying story).
Janssen issued a statement to the press when the Lilly findings were first presented in 1996, pointing out that the risperidone dosages in the study exceeded the daily average dose then used in the United States (approximately 4.7 mg). Janssen claimed, therefore, that the study data "lack clinical relevance."
After the Lilly study was published, spokespersons for Janssen Research Foundation from New Jersey and Belgium criticized not only the high risperidone dosages, but the statistical methods employed in the study (Gheuens and Grebb, 1998). They claimed that the one-sided test of significance applied to mean changes in total PANSS scores "made it impossible" to establish any clinical characteristics for which risperidone might be proved superior to olanzapine.
To this criticism, Lilly researchers responded that the dosage ranges enabled clinicians to optimize patient outcomes by adjusting the amount of medication in a manner similar to that used in practice (Tollefson and Tran, 1998). They defended the risperidone range of 4 mg to 12 mg as consistent with U.S. Food and Drug Administration-approved Risperdal labeling. In addition, they pointed out that the resultant mean modal dose of 7.2 mg in the study population was comparable to that employed in several studies that were pivotal in demonstrating risperidone efficacy for the U.S. market.
The Lilly researchers acknowledged using the one-side noninferiority test in one test claiming, "this analysis permitted testing unequivocally whether olanzapine was equivalent or superior to risperidone based on the change in the PANSS total score" (Tollefson and Tran, 1998a). They noted that all other efficacy and safety measures were analyzed with a conventional analysis of variance using a two-sided level of 0.05.
The last voice in the debate, for the moment, is that of Janssen's researchers who have just presented their own comparative data. They describe their study as the first large prospective, comparative study "that used clinically relevant doses." They have concluded "risperidone was shown to be both more efficacious and safer than olanzapine on important measures."
In the scientific process of presenting findings for others to assess, however, it is unlikely that the results will stand undisputed. It is also unlikely that these first large-scale comparative trials with newer antipsychotics will be the last; since the new agents are generally considered more effective and better tolerated than conventional neuroleptics, and there is little gained by continuing to compare new antipsychotics to old neuroleptics.
Risperidone will also continue to be targeted for comparative-and competitive-assessments, since it was the first antipsychotic, after the more toxic clozapine(Drug information on clozapine) (Clozaril), to successfully depart from the pharmacology of conventional neuroleptics. In a report at the ACNP meeting in December (Mullen et al., 1998), for example, researchers at Zeneca Pharmaceuticals Group indicated they had shown quetiapine(Drug information on quetiapine) (Seroquel) "more effective in treating depressive symptoms and equally effective in treating the positive and negative symptoms of psychosis when compared with Janssen's Risperdal."
Although the methodologies and results of these comparative studies will require scrutiny, the data should be useful to practitioners. These data can be used, according to John Kane, M.D., Hillside Hospital, Glen Oaks, N.Y., addressing an ACNP symposium in 1996, "in deciding how to choose among alternative medications; for which patient, at what point in time."
