Atypicals No Better Than Typicals in Depression with Schizophrenia
Second generation antipsychotics have developed a reputation for being more effective for treating a number of the symptoms of schizophrenia than their first generation counterparts, even if research doesn’t always back up this claim. (For example, see TCPR November 2009 or Lieberman JA et al, N Engl J Med 2005;353(12):1209–1223.) Furthermore, American Psychiatric Association Clinical Practice Guidelines specifically recommend atypical over first generation agents to treat depression in schizophrenia (American Psychiatry Association. Practice Guidelines for Treatment of Patients with Schizophrenia, Second Edition. Am J Psychiatry 2004:161(suppl):1–55).
CATIE, or Clinical Antipsychotic Trials of Intervention Effectiveness, was the largest set of trials ever to compare the major second generation antipsychotics. The main results of Phase 1 were released in 2005 (New Engl J Med 2005;353(12):1209–1223), and since then, numerous studies have been done using that data (including this one). For a review of CATIE, see TCPR, March 2006.
Using data from the CATIE trial, researchers followed 1,460 patients with schizophrenia who were assigned treatment with the first generation antipsychotic perphenazine (Trilafon) (n=256) or one of four second generation antipsychotics: olanzapine (Zyprexa) (n=328), quetiapine (Seroquel) (n=328), risperidone (Risperdal) (n=332), or ziprasidone (Geodon) (n=182). Patients were assessed for depression using the Calgary Depression Scale for Schizophrenia (CDSS). Average medication dosages were as follows: Zyprexa—20.1 mg; Seroquel—543.4 mg; Risperdal—3.9 mg; Trilafon—20.8 mg; and Geodon—112.8 mg.
Patients in each medication group were followed for up to 18 months; median treatment times were as follows: Zyprexa—9.2 months; Seroquel—4.6 months; Risperdal—4.8 months; Trilafon—5.6 months; Geodon—3.5 months.
Depression scores improved for all medication groups, and there were no differences among the first and second generation antipsychotics. Among the second generation antipsychotics, however, there was a small but statistically significant greater improvement in depression scores for the group taking Seroquel when compared to the group taking Risperdal, but only among a subset of patients who met criteria for major depressive episode (only about 9 patients in each group) (Addington DE et al, J Clin Psychiatry 2010; Sept 21, online ahead of print).
TCPR’s Take: TCPR has disputed the purported advantage of Zyprexa for core psychotic symptoms in the CATIE study because the drug was dosed more robustly than its competing agents. In this analysis, however, the fact that high_dose Zyprexa (as well as the other atypicals) yielded no benefit over Trilafon for depression is an important finding. This may be specific to Trilafon, which is a medium potency agent that is less likely than other first generation agents to cause symptoms that may appear to be depression, such as parkinsonism.
Should We Give Patients the Treatment They Request?
Since antidepressants and psychotherapy are about equally effective for mild to moderate depression, how do we decide which to use for a given patient? Common sense would dictate that we simply ask patients which they would prefer. Presumably, if a patient has faith in one versus the other treatment, the placebo effect will augment whatever specific effect the treatment may have. But has research borne out this assumption?
As it turns out, there has been little research on this issue, and its quality has been mixed (Swift JK et al, J Clin Psychol 2009;65(4):368–381). A recent trial used an interesting methodology, both randomly assigning some patients to their preferred or non-preferred treatment, and also allowing some participants to select their treatment.
A total of 145 mildly to moderately depressed patients were randomly assigned to one of three groups: 1) 10 weeks of sertraline (Zoloft) 50 mg with possible dose escalation to 200 mg; 2) 10 sessions of group cognitive behavioral treatment (CBT); or 3) patient choice of either Zoloft or CBT. But before they were randomized, all patients were asked whether they preferred to receive a drug or psychotherapy for their depression.
The researchers then compared HAM-D scores and remission rates of patients who wanted or did not want the treatment to which they were assigned. And here’s where things get interesting. In the Zoloft arm, the remission rate (RR) of those preferring a drug was 46%; the RR of those preferring therapy was 43%. In the CBT group, the RR of those preferring CBT was 43.9%; the RR of those preferring Zoloft was 0%. Similar results were seen on HAM-D scores. Clearly, patients who did not prefer psychotherapy had poor outcomes when assigned to receive CBT (Mergl R et al, Psychother Psychosom 2011;80(1):39–47).
TCPR’s Take: The results only partially confirm the common sense view that when all other factors are equal, patients will do better when offered the treatment they prefer. Just about all patients improved equivalently in this trial— the glaring exception being those who wanted meds but were assigned to therapy. On the other hand, those wanting therapy did just as well whether assigned to therapy or meds—which is something of a surprise. A potential explanation is that, according to the authors, patients were provided with “extensive education” about Zoloft after they had stated their treatment preference. This may have influenced some patients who felt negatively about Zoloft to feel more positively about the drug—which could have inflated the apparent Zoloft efficacy. Overall, the study’s results suggest that we should provide treatment in accordance with our patient’s preferences when feasible—though it may be especially important to respect the wishes of those patients who prefer medication treatment.