Reader Comments

Post a new comment on this article

Referee comments: Referee 2

Posted by PLOS_ONE_Group on 04 Apr 2008 at 16:10 GMT

Referee 2's review:

Peters et al. investigated the impact of a comprehensive selection of variants in five CYP genes on treatment response and tolerability under antidepressant pharmacological treatment with selective serotonin reuptake inhibitors (SSRIs) in a large pharmacogenetic sample (N=1953). The authors failed to detect any association, which does not encourage routine screening for CYP variants in an attempt to individualize antidepressant treatment. Given the present discussion revolving around the potential importance of CYP genes in psychopharmacotherapy and the concomitant economic considerations, this study for the first time investigating a comprehensive selection of CYP gene variants in a sufficiently large sample will have considerable impact on the field.

Patient ascertainment, phenomenological description, molecular genetic and statistical methods as well as deductions and discussion embedded in the literature available to date are well done.

However, the present manuscript might profit from the suggestions:

1)
In the introductory section, authors might include and discuss two previous studies on CYP gene variation and antidepressant treatment:

Grasmäder et al: Impact of polymorphisms of cytochrome-P450 isoenzymes 2C9, 2C19 and 2D6 on plasma concentrations and clinical effects of antidepressants in a naturalistic clinical setting. Eur J Clin Pharmacol. 2004 Jul;60(5):329-36

Rau et al: CYP2D6 genotype: impact on adverse effects and nonresponse during treatment with antidepressants-a pilot study. Clin Pharmacol Ther. 2004 May;75(5):386-93

2)
Definition of response as a QIDS score reduction of >50% "on the final clinical visit" is not clear. Citalopram was administered for 12 weeks, so one would assume the final clinical visit to be scheduled after 12 weeks. Not being aware of previous STAR*D publications in detail, it would be more informative, if authors included the time point of the final clinical visit (in all cases 12 weeks or 42 days?). This point is also crucial for the discussion section with respect to fast/slow response and comparability of response across subjects.

3)
Also, the fact that no ratings before week two were available deserves more consideration in the discussion (advantage, e.g. exclusion of non-pharmacological effects; disadvantage, e.g. no comparability with initial score before onset of treatment).

4)
In order to avoid ethnic stratification, for statistical analysis authors split their sample into two subsamples (Caucasian, African-American), which methodologically is absolutely correct and necessary. However, this - in addition to the two-stage design - decreases statistical power, particularly in the African-American sample, which might be mentioned more explicitly in the discussion section as increasing type II as well as type I error.

5)
Authors should mention whether - apart from HWE analysis - multiple testing was corrected for in the pharmacogenetic analysis. If not so, this would further support an overall negative result.

6)
Introduction, line 8, Statistical Methods, line 1, and in some sentences in the discussion section: References in brackets should precede the full stop.

**********
N.B. These are the comments made by the referee when reviewing an earlier version of this paper. Prior to publication the manuscript has been revised in light of these comments and to address other editorial requirements.

RE: Referee comments: Referee 2

Sphamilton replied to PLOS_ONE_Group on 09 Apr 2008 at 21:21 GMT

We are greatly appreciative for the reviewer's comments, and have addressed each in the final published version of the manuscript.