MedAvante Responds to Antidepressant Analysis Published in JAMA

1/18/2010 4:24:00 PM

MedAvante’s Earl Giller, Janet Williams and Michael Detke authored the following letter to the Journal of the American Medical Association (JAMA) in response to a study published in the January JAMA titled, “Antidepressant Drug Effects and Depression Severity: A Patient-Level Meta-analysis.”

The patient-level meta-analysis of major depressive disorder (MDD) by Fournier et al.¹ is consistent with the findings of previous studies that subjects with more severe symptoms show a greater response to antidepressant treatment compared with placebo. We strongly urge caution however, regarding the conclusion that patients with less-severe depression do not respond to antidepressant treatment.  The failure to see drug-placebo differences in these subgroups is fundamentally an inability to reject the null hypothesis, rather than a conclusive finding of a non-difference.  In other words, absence of evidence is not evidence of absence.

One methodological concern is the potential for inflation of baseline scores in mildly ill patients. Our experience with centralized raters who are blinded to entry severity criteria, compared with site-based raters, in placebo-controlled, randomized clinical trials reveals that inclusion of inappropriate subjects may be a major problem. When assessed by centralized raters, more than 1/3 of almost 3,000 subjects entered into 9 randomized controlled trials (Detke et al.²) should not have been enrolled.

Fournier et al. acknowledge the problem of baseline score inflation but state that it would work against their findings. Figure 2 is interpreted to show that patients on antidepressant treatment do not do significantly better than those on placebo treatment (HDRS point difference > 3) until subjects score 25 or higher at baseline. Yet the change from baseline scores is 4 or greater from baseline HDRS scores as low as 11, suggesting the problem may be more with placebo response than medication response.

Subjects with baseline scores close to the inclusion criterion are those most likely show a high placebo response (Williams et al.³) substantially reducing the possibility of detecting a drug effect, making accurate assessment at baseline paramount. In fact, when centralized raters are employed to select subjects, thereby reducing the risk of baseline inflation, the effect size almost doubled in one study (Coric, et al.⁴) largely due to a lower placebo response rate. Indeed, Shelton et al.⁵, in an analysis of four pooled randomized, placebo-controlled studies that used centralized raters showed nearly identical drug-placebo differences in patients with HAMD17 scores of <19, 20-24, and >25 at baseline. 

While greater severity of illness predicts greater response, we must be cautious about the cutoff for efficacy at lower severities, and newer methodological approaches may help clarify this further.

                Earl Giller, M.D., Ph.D.

                Janet B. W. Williams, D.S.W.

                Michael J. Detke, M.D., Ph.D.

                MedAvante Research Institute, Hamilton, NJ

¹Fournier JC, DeRubeis RJ, Hollon SD, Dimidjian S, Amsterdam JD, Shelton RC, Fawcett J.  Antidepressant drug effects and depression severity:  a patient-level meta-analysis.  JAMA.  2010; 303(1):47-53.

²Detke MJ, Williams JBW, Kobak KA, Ellis A, Giller E, Leon A, Reines S, Kane J.  The challenge of patient ascertainment in clinical trials – new data.  ACNP Annual Meeting.  2009. 

³Williams JBW, Dunn J, Kobak KA, Giller E, Wilson P, Detke MJ.  Placebo Response Assessed by Site and Blinded Centralized Raters in a GAD Trial.  ACNP Annual Meeting.  2009.

⁴Coric V, Goddard A, Oren D, Pultz J, Dockens R, Wu X, Gentile K, Thomas M, D’Souza B, Shekhar A, Stock E.  A Randomized, Double-Blind, Placebo-Controlled and Active Comparator Trial of Pexacerfont, a Corticotropin Releasing Factor Receptor-1 Antagonist, in the Treatment of Generalized Anxiety Disorder.  ACNP Annual Meeting.  2008.

⁵Shelton RC, Prakash A, Mallinckrodt CH, Wohlreich MM, Raskin J, Robinson MJ, Detke MJ.  Patterns of depressive symptom response in duloxetine-treated outpatients with mild, moderate or more severe depression.  Int J Clin Pract, 2007. 61(8), 1337–1348.