Between-Drug Differences

There are two central aspects of psychotropic drug response: the proportion of patients in whom a therapeutic response is induced ("incidence"), and the time to onset of improvement ("latency"). These two aspects can be detailed through 2-dimensional cure models and particularly enable statistical tests [Kolmogorov-Smirnov, Cramer-von Mises] to explicitly compare the speed of improvement between treatments. It turns out that almost all (>90%) of the observed differences in efficacy between drugs, which are beyond clinical relevance among effective compounds, can be explained by incidence, while less than 10% of the differences can be attributed to latency. In fact, cumulative rates of improvers/responders are virtually identical under antidepressants of large pharmacological and biochemical differences (e.g., imipramine [TCA], moclobemide [MAOI], fluoxetine [SSRI]) at standard drug trial days 3, 7, 10, 14, 21, 28, 35 and 42. These striking between-drug similarity of these cumulative rates are independent of improvement criteria.

Between-Drug Differences in Efficacy Beyond Clinical Relevance

Differences between drugs relate exclusively to the percentage of improvers (78%-83% under active compounds; 58% under placebo) or responders (48% under fluoxetine and moclobemide). If one drug displays better efficacy than another, it apparently converts throughout the entire observation period a higher proportion of patients to responders who otherwise would remain nonresponders. A finding which is compatible with results in the literature on early drug-placebo separation [Posternak and Zimmerman, 2005], and with results derived through the "average patient approach", where differences among drugs are found to be marginal at best, rarely reaching statistical significanc even with sample sizes of >1000 patients.

Highly Significant Differences for Side Effects

By contrast, active compounds display significant differences with respect to their unwanted side effects. Newer treatments have generally better tolerability and safety profiles than the older substances. This is demonstrated in Figure 1, where the time characteristics of premature withdrawal were detailed for the older TCA amitriptyline, the investigational compound oxaprotiline, and placebo. Premature withdrawals occur early under placebo — patients implicitly associate "no side effect" with "no drug effect" — while drop-outs under amitriptyline typically occur after one week of treatment when dosages reach therapeutical levels. Under the newer compound oxaprotiline, on the other hand, drop-outs occur at similar rates over the entire observation period.

Ethnic Variation in Dose, Blood-Level, Weight Gain

It is worth noting that the inter-individual variation in dose, blood-level, weight gain and other side effects is ethnicity-specific, whereas response rates are very similar across ethnicities, suggesting that drug-induced therapeutic effects are ethnicity-independent. Here, the term "ethnicity" is meant in the biological sense of a distinct individuality in certain genetic patterns as reflected, for example, by the inter-individual variation of the cytochrome P450 system.

References