With the introduction of second-generation antipsychotics, findings of lower EPS burden and TD risk were coupled with expectations of superior efficacy for positive, negative and cognitive symptoms. Initial efficacy studies seemed to confirm the superiority of second-generation antipsychotics, but the comparator consisted predominantly of haloperidol, used at moderate to high doses and often without anticholinergic cotreatment, which made early treatment discontinuation and secondary negative symptom presentations more likely in haloperidol treated patients. Since then, a series of acute phase and longer-term studies have been completed, including large efficacy-effectiveness hybrid trials (e.g., 31, 58-60) that compared first- and second-generation antipsychotics. These data have been evaluated and interpreted in a number of different ways. Interpretations include that there is no difference between first- and second-generation antipsychotics, that second-generation antipsychotics are superior to first-generation antipsychotics, that some second–generation antipsychotics are superior to either all or some first-generation antipsychotics, in general, or in certain efficacy and/or side effect domains, or in patient subgroups that are not yet easily identified prior to choosing a specific agent. Since such a number of divergent interpretations have been offered, this indicates that blanket statements do not do justice to the complex clinical situation and data base.
Taken together, the evidence seems to suggest that in refractory patients, clozapine is superior to first-generation antipsychotics (
61-
63) and second-generation antipsychotics (although the latter was hardly confirmed in a recent meta-analysis, which was attributed to inappropriately low clozapine doses) (
64). Compared to first-generation antipsychotics, only three second-generation antipsychotics (amisulpride, olanzapine and risperidone) were superior based on PANSS score change differences (
61), but these were also the medications studied at a time when first-generation antipsychotics predominated, whereas the newer second-generation antipsychotics were tested mostly at a time of predominant second-generation antipsychotic use. While this could have introduced a cohort sampling bias, the differences between non-clozapine antipsychotics were very modest, with effect sizes as low as 0.1-0.3. Similarly, differences between second-generation antipsychotics studied head-to-head were either non-existent or also marginal, favoring in some comparisons risperidone (versus quetiapine and ziprasidone) or olanzapine (versus aripiprazole, quetiapine, risperidone and ziprasidone), with the same low effect size difference of only 0.1-0.3 (
59-
61). Moreover, the differences between second-generation antipsychotics were even more restricted when not analyzing mean total PANSS score differences, but analyzing discontinuation of medication due to inefficacy (
61). Thus, differences in design, including active or placebo control, dosing, and sponsorship (
65-
67) may have a greater impact on efficacy outcomes than the actual choice of non-clozapine antipsychotics.
Although the CATIE and CUtLASS studies seemed to suggest that there are generally no differences between second-generation antipsychotics and first-generation antipsychotics in all cause discontinuation, especially when analyzing switchers only (
68) and quality of life (
60), these conclusions have been challenged based on insufficient sample sizes to make non-inferiority claims (
69). Moreover, in first episode samples, all-cause discontinuation rates and relapse rates were significantly higher at one and two years, respectively, with modestly dosed haloperidol compared to the respective second-generation antipsychotic comparators (
31). Moreover, even in chronic samples, relapse rates were also significantly higher in first-generation antipsychotics, although haloperidol doses were higher than currently recommended (
70).
Moreover, the clinical effectiveness of first-generation antipsychotics, a measure of objective and subjective outcomes encompassing symptom-based and functional effects, is challenged by increased acute (
71) and chronic (
72) extrapyramidal side effects and related symptoms of dysphoria, compared to second-generation (atypical) antipsychotics (SGAs). Even though at chlorpromazine equivalents below 600 mg/d, there was no increased EPS rate with typical vs atypical antipsychotics, at those doses, the efficacy of second-generation antipsychotics was superior (
27). Furthermore, while masking of EPS can be achieved with prophylactic anticholinergic treatment (
58), the risk of TD is not reduced, but rather potentially increased (
72), and recent data suggest that anticholinergic medication load is associated with decreased efficacy of cognitive remediation treatment (
73).
Ultimately, we feel that the controversy regarding the likely oversimplified dichotomy between first- and second-generation antipsychotics has resulted in progress, in that it stimulated the conduct of large trials and examination of effectiveness outcomes beyond symptom reduction. These trials have generated new and important data, but also highlighted methodological challenges. These challenges include the definition of clinically meaningful endpoints, the effect of baseline medication and past treatment history, the limitation of available treatments used at a more chronic illness phase, and the importance of differences in acute and long-term adverse effects. All of these data point to the need for new treatments with novel mechanisms, tailored approaches that map onto the pathophysiology of the disease process (that may vary between patients and between different illness stages), and to the need for the biological dissection of patients into meaningful subgroups that can inform a stratified or, even, individualized treatment selection.