March, 2000 —Article: "Emergence of antidepressant induced suicidality" by Dr David Healy finds that out of 20 healthy volunteers with no history of depression or suicidal ideation, two became suicidal on sertraline (Zoloft), an SSRI. This is important refutation to the contention that suicidal acts of people on SSRI medications could always be attributed to underlying depression.

April, 2000 — Study 701 on the effectiveness of paroxetine (Seroxat, Paxil) in treating children and adolescents with major depressive disorder (MDD) is completed. Study 701 involves 206 pediatric patients, 149 who complete the study. Nine Paxil patients (8.9 percent) withdraw from the study due to adverse events vs two (two percent) on the placebo, mostly due to depression, "emotional lability" (suicidality), agitation, hostility, and nervousness. In addition, three Paxil patients withdraw from the study due to "emotional lability" soon after the last dose of study medication. One Paxil patient overdoses and cuts herself, another experiences suicidal ideation, and three others experience exacerbation of their depressive symptoms. All these events are considered unrelated to Paxil by the study's investigators. Another overdose is regarded as related to the Paxil. Overall, the patients on Paxil fare worse than those on placebo on the primary measures. The study is never published.

April 26, 2000 —Letter: In a missive entitled "Revision of Paroxetine Adolescent Depression Manuscript", Sally Laden writes to Jim McCafferty noting the revisions she is making to the Study 329 manuscript, and outlining the plan for resubmission of the article to the Journal of American Psychiatry. All the revisions are "cosmetic" they relate to wording and presentation:

…I attempted to make as many of the JAMA reviewer comments (sic) as practical. A summary of the more substantive points is listed below:

1. A more thorough discussion of the high placebo response rate in the Conclusions section.
2. A "study limitations" paragraph has been added to the Conclusion that addresses the fact that the study was not designed to directly compare paroxetine and imipramine; that the HAM-D score at baseline was low relative to adult studies; and that patients with externalizing disorders were eligible for enrollment. (italics added)
3. The mean imipramine plasma concentrations at weeks 4 and 8 will be stated (SB: please provide this data), and it is stated that a more thorough discussion of this data will appear in a separate publication.
4. A new figure (Fig. 3) showing the HAM-D total score response vs time.
5. As noted by several reviewers, the definitions of remission and response overlap. Boris Mirmaher suggested that the definitions be revised to delete mention of remission and simply describe response. This suggestion is reflected in the enclosed manuscript.
6. As suggested by Reviewer #2, the title is revised to delete the phrase: but not imipramine."

This letter is interesting because it clearly illustrates who is making the decisions about the content of the final manuscript, a document that will impact the lives of thousands of young people. Sally laden decided that it was not "practical" to deal with the safety issues. The significant incidence of adverse events (11 out of 93) in paroxetine group, 7 of whom had to be hospitalized, is not addressed. Subjects experiencing adverse events on paroxetine continue to be omitted from the considered data.

May, 2000 —According to the April 26, 2000 letter from Sally Laden to Jim McCafferty, the intention is to submit the "revised" Study 329 manuscript to the Journal of American Psychiatry. Apparently the decision changed at some point in May, 2000 to submit it to the Journal of the American Academy of Child and Adolescent Psychiatry (JAACAP) instead. The reason for this change is not stated. At the time, Graham Emslie, one of the Study 329 authors, is on the Board of JAACAP.

July 27, 2000 —Letter: Mina Dulcan, M.D., Editor of the Journal of the American Academy of Child and Adolescent Psychiatry (JAACAP) writes to Dr Martin Keller acknowledging submission of Study 329:

"Dear Marty,

1. The rate of serious adverse events in the paroxetine arm is somewhat high (11 subjects, presumably out of 93). This is included in the results but not discussed at all.
2. Similarly, each group had a fairly high rate of not completing the 8-week trial that is somewhat glossed over.
3. Given the high placebo response rate, what algorithm should clinicians follow when treating a depressed teenager? Are SSRIs an acceptable first-line treatment if approximately one half of youth get better with only supportive interventions.
4. Although it is implied, a stronger statement could be made regarding the lack of indications for tricyclic antidepressants given the lack of efficacy and side effect profile.
5. In the discussion section, there is a statement suggesting that a traditional three arm comparative trial was not done due to the risk of exposing additional subjects to clinical research. This seems rather self-serving, since I suspect the power issues and sample size limitations prevented this from being done, not human subjects concerns.
6. In the discussion section, there is a statement that the entry HAM-D score required was lowered to greater than or equal to 12 "to reflect the severity of their disorder"in a pediatric population. What does this mean? Is the scale not valid for youth? Are their scores somehow different than adults'?"

September 14, 2000 — Medical Review: Lilly submits application to FDA to gain an additional 6-month patent exclusivity for Prozac. The application later updated to include Lilly's integrated safety data summaries (ISS) from 3 pooled Prozac pediatric trials: the NIMH-funded trial (Emslie, 1997)*, a Lilly-funded trial with depressed children (Emslie, 2002), and a trial with children diagnosed with obsessive-compulsive disorder (Geller, 2001).

In 2002, Dr Andrew Mosholder does the medical review on the efficacy of fluoxetine (Prozac), compared to placebo, for children and adolescents with major depression. He finds:

• "7.2.7 Conclusions: This study…failed to show statistically significant effect of fluoxetine according to the protocol specified primary outcome measure"
• "8.2 Deaths: There were no deaths in these pediatric studies."
• "The supplement is approvable, in my opinion."

*see November, 1997 and June 22, 2004

2000 — Marketing Facts: Between 1995 and 2000 the number of marketing staff working for US pharmaceutical companies increases by 60% while research staff declines by 29% (see page 15). GlaxoSmithKline spends $91.8 million US advertising the SSRI Paxil in the US, almost $15 million more than Nike spends advertising its top brands of running shoes. (See Page 17) This level of spending is associated with a $355.6 million increase in sales between 1999 and 2000.