Eli Lilly Prozac Documents: What do They Reveal?

Eli Lilly Prozac Documents: What do They Reveal?

Thu, 27 Jan 2005

A report in the BMJ (formerly the British Medical Journal) that relied on internal Eli Lilly documents to challenge Lilly’s claims about the safety of Prozac has stirred a prickly hornet’s nest. [1] The documents show, among other things, that a 1988 review of clinical trials found that 38% of patients taking Prozac compared to 19% of patients on placebo experienced “activation,” which is linked by the FDA in current labeling to violent and suicidal behavior. [2] [PZ-477] .

Lilly took out a full page advertisement (“An open letter to patients and doctors” signed by Lilly CEO, Sidney Taurel) in 15 newspapers nationwide stating: “it was simply wrong to suggest that information on Prozac was missing, or that important research data on the benefits and possible side effects of the drug were not available to doctors and regulators.” [3] The BMJ has just issued a correction and an apology to Eli Lilly for suggesting that documents went “missing” stating: “all the documents supplied to the BMJ were either Eli Lilly documents or were in the hands of Eli Lilly [and] had in fact been disclosed during the suit.” [4]

The Associated Press reports that BMJ editor, Kamran Abbasi, said the apology was limited to the issue of whether the documents were missing from the court case. “Whether or not Eli Lilly made all of the information available to the FDA at the appropriate times is a question for the FDA to answer…. we await their response.” Morry Smulevitz, a spokesman for Lilly, said the company also awaited the FDA’s finding. [5]

Inasmuch as the documents’ authenticity is not in dispute, and the information about the hazardous drug effects contained in the documents (stamped “confidential) is of vital importance to public health – we will focus on their content rather than be diverted by who said what to whom. The documents reveal that as early as 1984 – years before the FDA approved the drug – Lilly knew or should have known that Prozac frequently induced severe adverse side effects, including “activation” and suicidality [6]:

“…frequency of side effects was very high (partly more than 90%) and side effects resulted nearly in each study in drop outs. The frequency of side effects depended on the dose, the age, and the duration of treatment… the clinical significance of side effects is not only the frequency of their occurrence but also their severity.”

Lilly’s submission of clinical trial data to the German regulatory authority (BGA) from a pooled group of 1,427 patients who tested Prozac, showed that there were 16 suicide attempts – two of which were successful. Furthermore, “In 15-20% of cases side effects…involve the central nervous system. As most of them resemble the clinical picture of the underlying disease, even from theoretical reasons one has to expect intensification and not an improvement in symptoms…” [PZ 281, p. 1525] [6]

A June 1984 Lilly memo notes: “A critical issue for the BGA is safety in long-term treatment.” Among BGA’s concerns: Prozac was “a stimulating/activating drug (side effect profile, suicides, suicide attempts)” and “the increase in agitating effect occurs earlier than the mood elevating effect and therefore an increased risk of suicide exists.” [PZ 2469, p. 489] [6]

In light of the risks, the German regulatory agency rejected Prozac in 1984 stating: “Considering the benefit & risk we think this preparation totally unsuitable for the treatment of depression.” [6] In 1985, BGA once again rejected Prozac stating:

“The use of the preparations [fluoxetine] seems objectionable, as the increase in agitating effect occurs earlier than the mood elevating effect and therefore an increased risk of suicide exists.” [PZ 1124][7]

One of the BMJ documents was a review of spontaneous domestic reports of adverse events (1982 to 1991) including charts and graphs comparing adverse effect reports for Prozac (14,198) and four older comparator antidepressants (Trazodone, Amitriptyline, Desipramine, Maprotiline (total 6,119 reports). Spontaneous reports constitute the FDA’s alert system for identifying side-effects in marketed drugs. The number of reports involving Prozac and suicide attempts and hostile or injurious behavior is astronomical compared to the older tricyclics – even when adjusting for the number of prescriptions for each. Among Prozac users there were 519 Suicide attempts reported; 468 cases of overdose; 234 reports of hostility; 115 reports of intentional injury; and 321 reports of drug-linked psychotic depression. When compared to Trazadone, (at the time the next most prescribed antidepressant) the suicide attempt rate among the adverse event reports was 3.7% for Prozac and 0.2% for Trazadone (4 reports). [8]

Although Lilly vehemently denied the risk of suicidal acts, agressiveness, and self-mutilation, in 2000, Lilly purchased a patent for an improved version of Prozac – R-fluoxetine–prompting the Boston Globe to investigate. The Globe reports revealed that documents submitted to the US patent office for R-fluoxetine contradict Lilly’s public position, stating that R-fluoxetine will not produce several existing hazardous side effects including: “akathisia, suicidal thoughts, and self-mutilation,” which the patent calls “one of its more significant side effects.” Lilly paid the patent holder of R-fluoxetine $20 million but never developed the improved Prozac. [9]

The Globe published excerpts from several Lilly documents stamped “confidential”–including documents obtained by BMJ. [9]. For example, an e-mail dated Nov. 13, 1990, from Claude Bouchy (a Lilly employee in Germany) to three corporate executives at the company’s US headquarters, including Lilly’s President, Sidney Taurel, complains about questionable instructions he had been given for coding adverse events: [10]

“the directions our safety people are getting from the corporate group (Drug Epidemiology Unit) requesting that we change the identification of events as they are reported by the physicians. . .Our safety staff is requested to change the event term ‘suicide attempt’ [as reported by the physician] to ‘overdose.’ …. it is requested that we change . . . ‘suicidal ideation’ to ‘depression.‘” [PZ 4003]

Bouchy tried to appeal to his US colleagues at Lilly headquarters:

“I do not think I could explain to the BGA, to a judge, to a reporter or even to my family why we would do this especially on the sensitive issue of suicide and suicide ideation.” [PZ 4003]

On November 14, 1990, Bouchy tried once more to appeal to his colleagues at Lilly headquarters:

“I personally wonder whether we are really helping the credibility of an excellent ADE system by calling overdose what a physician reports as a suicide attempt and by calling depression what a physicians is reporting as suicide ideation…it could be argued by people who have little sympathy for the company or by regulatory authorities that it is not a responsible way to deal with an issue which is getting so much attention in the scientific and in the general press.” [bold added PZ 1573]

The documents undermine the company’s 16 years of denial: “there is absolutely no medically sound evidence of an association between any antidepressant medicine, including Prozac, and the induction of suicidal ideation or violence.” [11].

Those who have seen the documents, including Congressman Maurice Hinchey, say the documents “clearly show a link between Prozac and actions of violence perpetrated by people taking the drug against themselves and against others. The documents… show that the company was instructing its employees to hide this information. We’re seeing evidence here that it was a conscious act on the part of the company.” [12]

Lilly may have made the documents available to plaintiff lawyers, expert witnesses, and the FDA, at some time, in some form. Nevertheless, the fact remains that most patients psychiatrists, pediatricians, family physicians that Prozac poses serious risks of violence and suicidal behavior. And even the FDA reviewer says he did not know about this information Indeed, it is hard to believe that responsible doctors would widely prescribe a drug that they knew might induce agitation, hostility, anxiety and violent / suicidal behavior in as many as 1 out of every 5 patients! [2] Doctors don’t generally play Russian roulette with their patients’ lives.

Dr Richard Kapit, who was the FDA clinical reviewer who approved fluoxetine based on the data Lilly submitted to the FDA, said he was not given the data in these documents. [1] And Dr. David Graham, currently associate director in the FDA’s Office of Drug Safety, discovered that Lilly excluded 76 of 97 cases of reported suicidality from its post-marketing surveillance data submitted to the FDA. Dr Graham concluded in a memo dated September 11, 1990: “because of apparent large-scale underreporting, [Lilly’s] analysis cannot be considered as proving that fluoxetine and violent behavior are unrelated.” [1] Dr. Martin Teicher, a prominent Harvard researcher and clinician who wrote the first crucial scientific paper in 1990 linking violent /suicidal acts to Prozac, told USA Today that “lives were threatened and Americans treated like “guinea pigs” because Eli Lilly & Co. officials lied 15 years ago in denying there was any evidence the anti-depressant Prozac could cause suicidal behavior.” [12]

Notwithstanding Lilly’s recent claim about the company’s “transparency,” the “open letter to doctors and patients” once again fails to disclose to the public vital safety information. Despite an FDA requirement (since March 2004) for all antidepressants, including Prozac, to warn physicians and the public, about the risks of activation and increased suicidal behavior, nowhere in this public document does Lilly make that disclosure. Instead, Lilly states that depression can cause suicide, but Lilly avoids mentioning that Prozac can cause activation, hostility, and suicide.

Furthermore, the “open” letter misrepresents the evidence in Lilly’s own documents by claiming: “It can be dangerous to stop taking any prescription medicine without a doctor’s guidance – and Prozac is no different.” [3] No different? A drug that increases the risk of suicide – whose label the FDA requires to carry a Black Box warning – is not like most medicines. Lilly’s “promise to publish the results of all Lilly clinical trials….and to make this information available online” is contradicted by its actions. Lilly fails to post any of the disputed documents, disingenuously claiming: “Due to court orders requiring confidentiality of some of the documents contained within the 52 pages, Lilly is unable to release the 52 pages of documents.” [3]

Now, as in the past, Lilly has aggressively sought to discredit researchers and journalists who published data linking Prozac to suicide – even as its own documents confirmed the risk. One of Lilly’s early targets was Dr. Martin Teicher, whose published finding of a 3.5% suicide risk for patients put on Prozac. [14] Others include, Dr. Peter Breggin (Talking Back to Prozac, 1995), Dr. Joseph Glenmullen (Prozac Backlash, 2000), and Dr. David Healy (Let Them Eat Prozac, 2003). Each has been savagely pilloried for bringing to public attention the concealed hazards of SSRI antidepressants. These psychiatrists have examined the Lilly documents in their special capacity as expert witnesses, and they dispute Lilly’s claims, arguing that the documents to the best of their knowledge were sealed and out of public reach. Their separate written reviews of the documents will be posted on the AHRP website. See: http://www.ahrp.org/risks/SSRIbreggin0105.php

The marketing of Prozac is a microcosm of the pharmaceutical industry’s modus operandi for marketing drugs of dubious benefit but hazardous side effects to millions of unsuspecting people. The marketing of such drugs relies primarily on promotional campaigns and the seal of legitimacy from paid medical opinion leaders. Neither the drug’s hazards nor the consulting fees of the experts are disclosed to the public. Vioxx and the other COX-2 drugs were marketed like the SSRIs to millions of people who didn’t need them but who were harmed from their concealed hazardous effects. [15] New York State Attorney General, Eliot Spitzer, called the concealment of adverse drug effect data from physicians–fraud. The matter of who did what, when, and how it impacted on public safety is grist for Congressional oversight and investigations by state and US attorneys general.

Lilly’s documents cast doubt on the legitimacy of current psychiatric practice guidelines that have been formulated on the basis of incomplete, biased information about the safety and efficacy of widely prescribed antidepressants, including Prozac. The focus of the debate should be to determine whether doctors and patients were adequately informed about the risks associated with Prozac, risks which Lilly documents show the company knew for over 16 years. Congress needs to consider the impact of not making the public aware of potentially harmful effects of commonly prescribed drugs.

Questions for Lilly:

1. Has Eli Lilly ever acknowledged – and does the company acknowledge today–that antidepressants, including Prozac (fluoxetine) and Cymbalta (duloxetine), cause suicidal behavior?

2. Has the information about this clinically significant risk been disclosed to physicians in prescribing guidelines? If so, show documented evidence.

3. Is there any evidence that Lilly tried to make the suicide / violence risks known to the public?

4. When can we expect the Black Box to appear on the labels of Lilly products containing fluoxetine and duloxetine? (This is an FDA requirement)

Questions for the FDA:

1. If, as Lilly insists, the company made available all “important research data on the benefits and possible side effects of the drug” to the FDA, why was information not disclosed about clinically significant harmful side effects, including the risks of suicide and violence, in the Prozac label all these years?

2. If, as Lilly insists, the FDA knew that Prozac induced activation linked to suicide attempts in 38% of adult patients in clinical trials, what was the justification for approving the drug for use in depressed children?

Contact: Vera Hassner Sharav


1. Lenzer, J. FDA to Review “Missing” Drug Company Documents, BMJ (formerly British Medical Journal), January 1, 2005. http://bmj.bmjjournals.com/cgi/content/full/330/7481/7 ;

2. See: Lilly Memo. Activation and Sedation in Fluoxetine Clinical Trials, 1988 [PZ 477]

3. Lilly An open letter….: http://www.lilly.com/news/prozac_bmj.html

4. BMJ. Eli Lilly: Correction and apology http://bmj.bmjjournals.com/cgi/content/full/330/7481/7/DC1

5. Associated Press. U.K. Medical Journal Retracts Report on Missing Prozac Data, WSJ, January 27, 2005 11:20, http://online.wsj.com/article_email/0,,SB110684070739638014-INje4Nglah4nZ2sa4KIa6yHm4,00.html

6.. Lilly Memo, Keitz; Bad Homburg to Corporate Headquarters, May 25, 1984, with Comment on Clinical Documentation Fluoxetine from BGA (Germany) rejecting Prozac for marketing. [BZ 281]

7. Lilly Memo, Hans Weber, February 27, 1985 with BGA rejection [PZ 1124 p. 336]

8. Spontaneous Domestic Reports January 1982-July 1991. [PZ-1548, See pp. 5-14] online at: http://www.house.gov/apps/list/press/ny22_hinchey/morenews/bmj_docs.pdf

9. Garnett, L. Prozac Revisited: As Drug Gets Remade, Concerns About Suicides Surface, Boston Globe, May 7, 2000 p. A-1

10. Lilly Memo, Claude Bouchy to Lilly Executives at Headquarters, including Sidney Taurel. November 13 and 14, 1990. [Pz -4003] [PZ-1573]

11. The statement was made under oath by Dr. Gary Tollefson, a Lilly scientist. See: Zuckoff, M. Prozac data was kept from trial, Boston Globe, June 8, 2000. p.A-1.

12. Belson K. “Lilly Shares Fall on Report About Prozac Documents NY Times, January 1, 2005, C- p.; CNN at: http://edition.cnn.com/2005/HEALTH/01/03/prozac.documents

13. Elias, M. Psychiatrist: Company hid link between Prozac, suicide Documents raise questions; Eli Lilly denies accusations USA TODAY. January 6, 2005 Page 2A http://www.usatoday.com/printedition/news/20050106/a_prozac06.art.htm

14. Teicher MH, Glod C, Cole JO. Emergence of intense suicidal preoccupation during fluoxetine treatment . American Journal of Psychiatry,1990: 147: 207-210.

15. Vioxx Went Mostly To Patients Who Didn’t Need It, Stanford Researcher Says. Press Release, Stanford University Medical Center. January 24, 2005. http://www.sciencedaily.com/releases/2005/01/050123210312.htm