1 JULY 21, 1994 2 * * * * * 3 DIRECT EXAMINATION (CONTINUED) 4 5 BY MR. SMITH: 6 Q Doctor Thompson, is it 7 correct that Lilly never conducted any clinical 8 trial specifically designed to look at the 9 question of whether or not Prozac caused or 10 induced violent aggressive behavior? 11 A There is no protocol that 12 was carried out in which that was the primary 13 purpose rather than a secondary purpose. 14 Q Was there any protocol 15 that even mentioned an examination of violent 16 aggressive behavior with respect to employing any 17 specific scale designed to examine violent 18 aggressive behavior in individuals? 19 A Not as a specific 20 objective, but of course all of our protocols 21 examine all the safety and efficacy features of 22 the drug, even though their primary intent may be 23 something else. 24 Q No, that's not what I was 316 1 asking. I was asking whether or not there was any 2 protocols that specifically mentioned as even a 3 secondary intent any scales to be employed to look 4 at the question of violent aggressive behavior. 5 A By scales, you need to 6 include the comprehensive mechanism by which we 7 assess all safety things. If you mean by a scale 8 a specific point score question by question -- 9 Q Yes, that's what I 10 meant. 11 A The answer is no, but 12 that's only one of the kinds of scales that would 13 be used to assess a safety finding. 14 MR. SMITH: We would 15 object to the answer as being nonresponsive to the 16 question asked, after no. 17 Q You came to Lilly in 18 1982, is that correct? 19 A Yes, sir. 20 Q And at that time Prozac 21 was being investigated employing clinical trials 22 both in the United States and in foreign 23 countries? 24 A I am not sure about the 317 1 foreign countries. I am pretty sure that at that 2 time the trials were in North America, so I can't 3 speak to Canada, but I do not believe that we had 4 European trials going on at that time. 5 Q Do you know whether or 6 not, at the time you began with Lilly in 1982, 7 that an application had been made to market Prozac 8 in any countries outside the United States? 9 A I'm pretty sure the 10 answer is no, as of April 1 of 1982. 11 Q Do you recall when the 12 first application to market Prozac in a foreign 13 country occurred or when it happened? 14 A I think it came after our 15 submission of the New Drug Application to the FDA, 16 and I know when the first approval occurred, but 17 I'm not sure of the exact date of the submission 18 in Belgium, for example. 19 Q Is Belgium the first 20 country in which the product was approved? 21 A I think that's correct. 22 Q And the NDA was submitted 23 to the United States Food and Drug Administration 24 in September 1983, correct? 318 1 A Yes, sir. 2 Q Doctor Weinstein has 3 testified earlier that basically the Western 4 European countries and the United States and 5 Australia have similar regulatory requirements in 6 connection with registering a product within a 7 particular jurisdiction; is that correct? 8 A Well, they are similar, 9 but there are also very many differences as well. 10 Q Doctor Weinstein also 11 testified that the countries in Western Europe and 12 the United States and Australia generally look to 13 trials conducted within those countries in support 14 of applications. Do you follow what I'm saying? 15 A Yes, sir. 16 Q Is that your 17 understanding, that for instance in Belgium, they 18 will look at United States trials, United States 19 will look at trials conducted in Belgium? 20 A That's changed over 21 time. In the early 1980's it was possible for a 22 drug to be approved in a country with no studies 23 within that country, but most major countries were 24 eager to see some of the studies done within their 319 1 own borders and their own citizens. 2 Q But the countries would 3 look at clinical trials conducted in other 4 countries? 5 A Yes, sir, that's exactly 6 correct. 7 Q And consider those trials 8 in support of efficacy and in analyzing the 9 question of the safety of the product? 10 A Yes, sir. 11 Q And the United States 12 Food and Drug Administration does that also, do 13 they not? 14 A Yes, sir. 15 Q As I understand it from 16 talking with Doctor Weinstein, the regulatory 17 process in some of the Southeast Asian countries 18 and Middle East may be a little different; is that 19 correct? 20 A I'm not really an 21 expert. Doctor Weinstein is far closer to that 22 than I am. 23 Q International was under 24 your purview at one time, was it not? 320 1 A At one time, that's 2 correct. 3 Q When was that? 4 A That was primarily when I 5 was the executive vice president, so it's been 6 fairly recent. 7 Q So that would have been 8 what years, so I don't have to look back in my 9 notes? 10 A Probably '90, '91, could 11 have been '89. I would have to go back and try to 12 reconstruct that. 13 Q Were you aware that there 14 were any foreign countries that had raised the 15 issue of whether or not Prozac could, as a side 16 effect, induce or cause suicidality or violent 17 aggressive behavior? 18 A Yes, sir. 19 Q When did you first become 20 aware of that, sir? 21 A I can't remember the 22 exact date, but the function of the US group that 23 I was primarily responsible for at that time was 24 in fact assist the affiliates in answering 321 1 questions from all of their regulatory 2 authorities. So as the submissions occurred in 3 country after country, we would get sets of 4 questions, which amazed me because they differed 5 from country to country, and I think it was 6 Germany that first raised that specific issue, and 7 so part of my responsibility and my team was in 8 fact to help the affiliate answer that question. 9 Q When was it that you 10 first recall that Germany raised that question? 11 A I can't be specific on 12 the date, because there was a whole -- I mean over 13 a period of years we got questions in from all the 14 European regulators. 15 MR. MYERS: The question, 16 as raised, raises two questions, one on 17 suicidality and one on violent aggressive 18 behavior. So in answering the question, keep 19 those two elements in mind. 20 A But I can't be specific 21 about exactly when we got the questions on either 22 one. 23 Q The German equivalent to 24 the United States Food and Drug Administration is 322 1 an entity known as the BGA, correct? 2 A Yes, sir, the 3 Bundesgesundheit Amt. 4 Q Beg your pardon? 5 A Bundesgesundheit Amt. 6 Q What does that mean in 7 English? 8 A That means the health 9 authority. 10 Q The health -- 11 A The state health 12 authority. 13 Q The state health 14 authority, and to your understanding they're the 15 Germany equivalent to the United States Food and 16 Drug Administration? 17 A Yes, but their law is a 18 little bit different because there is actually 19 both a central and decentralized process in terms 20 of drug approval and regulations, and we have no 21 comparable state regulation, by and large, of 22 drugs. So, they're the central authority, and 23 therefore the most equivalent to the FDA that we 24 have. 323 1 Q But a drug can be 2 approved for marketing in Germany without approval 3 of the BGA, is that right? 4 A That's the best of my 5 understanding. 6 Q As a product can't be 7 approved for marketing in the United States 8 without approval of the United States Food and 9 Drug Administration, correct? 10 A As you know, you have to 11 define drug very precisely for that to be true, 12 but generally that's correct. 13 Q How precisely would you 14 like for me to define it, sir? 15 A Well, there are many 16 substances which in fact have pharmacological 17 action in the body which fall outside the legal 18 purview of the FDA. 19 Q This litigation involves 20 Prozac, fluoxetine hydrochloride, correct? 21 A Yes, sir. 22 Q Is that a drug, sir? 23 A Yes, sir. 24 Q Is that a drug that must 324 1 be approved by the United States Food and Drug 2 Administration? 3 A For interstate commerce, 4 yes, sir. 5 Q And is the drug Fluctin, 6 fluoxetine hydrochloride, the similar drug in 7 Germany? 8 A Yes, sir. 9 Q In fact, it's the exact 10 same drug, it's just got a different name, doesn't 11 it? 12 A I wouldn't want to 13 testify into the exact excipients and other inert 14 ingredients in the formulation, but the active 15 ingredient is the same. 16 Q The active ingredient is 17 fluoxetine hydrochloride? 18 A Yes, sir. 19 Q That's the kind of drug 20 I'm talking about, sir, when I say drug. 21 A Then you are exactly 22 correct. 23 Q All right? 24 A Yes, sir. 325 1 Q We've used the word drug 2 many times yesterday, and a few times today, and I 3 had always used that word drug as meaning a drug 4 like Prozac. Have you, in any of your answers, 5 construed me to have a different meaning toward 6 the word drug or have you given me an answer -- 7 given me an answer with a different meaning 8 concerning drug? 9 A I don't believe so, but 10 your question was very specific in terms of 11 regulatory authority, and as you know, there's a 12 huge debate in Congress today about the status of 13 things that I would call drugs, including vitamins 14 and things called by other people food. So, I 15 mean, this is not a cut and dried issue and I was 16 trying to be absolutely correct in answering your 17 question. 18 Q It appears to me that you 19 are just trying to argue with me, Doctor Thompson, 20 and I'm just trying to simply get some information 21 from you, and when I use the term drug, I'm going 22 to use that term as most ordinary human beings, 23 most individuals on a jury would use that, like 24 Prozac. Can we have that understanding now, sir? 326 1 A Lewis Carroll said that a 2 word means exactly what I mean and nothing more or 3 less, and if you'll define things specifically, 4 I'll be glad to use whatever word you like in 5 exactly your way. 6 Q I will now define drug as 7 to mean a drug like Prozac since that's what this 8 lawsuit involves and that's what Lilly is 9 marketing, all right? 10 A We market a number of 11 other -- 12 Q Can we use that -- 13 A No, sir, we market a 14 number of others things as well. 15 Q Can we use that 16 definition, sir? 17 A We will use your 18 definition. 19 Q We'll use the definition 20 of the term drug like Prozac, all right? 21 A Yes, sir. 22 Q Now, I believe we were 23 talking about whether or not the BGA is the 24 Germany equivalent of the United States Food and 327 1 Drug Administration, for most practical purposes? 2 A It is, except that there 3 are also regional regulatory oversight of drug 4 marketing and use, and that's not the same as in 5 the United States. 6 Q Do those regional 7 entities that have oversight over the drug in 8 Germany require approval independently of a drug 9 other than that approval that's given by the BGA? 10 A It's my understanding 11 that in some circumstances that's correct. 12 Q For Prozac? 13 A I don't believe so -- 14 well, I don't believe that that was ever raised as 15 an issue by any of those states within Germany. 16 Q Do you know any instance, 17 Doctor Thompson, where some subgovernmental 18 authority of some local province or state require 19 that you submit to them some application for 20 approval of Fluctin, Prozac, in that particular 21 instance over and above that approval that was 22 sent to the BGA in Germany? 23 A I think that occurred in 24 Canada in some of the provinces, which also have 328 1 both a central and a decentralized -- 2 Q No, I'm talking about 3 Germany, I said BGA. 4 A I'm sorry, I wasn't 5 responsible for it at that time and I don't have 6 my own knowledge of that. 7 Q Then why would you bring 8 it up? 9 A Because you asked me a 10 very specific question about the similarity 11 between the BGA and the FDA, and I wanted to give 12 you a full and complete answer to that. 13 Q Do you know of any 14 instance, sir, where any other entity in Germany 15 required that you submit an application to approve 16 Prozac in Germany other than the BGA? 17 A I do not. 18 Q Was all of the assistance 19 that your team gave to the German Lilly affiliate 20 assistance in dealing with the BGA in Germany? 21 A And -- no, that's not 22 correct, because we also provided scientific 23 analysis of the data for their dealing with 24 opinion leaders, professors, medical groups, as 329 1 well. 2 Q Well, but as far as any 3 regulatory dealings that that German affiliate was 4 doing, did you deal with -- do you know of any 5 instance where they were dealing with any entity 6 other than the BGA with respect to the 7 registration process of the drug? 8 A No, I do not. 9 Q How did you first learn 10 that the BGA in Germany had raised the issue of 11 Prozac and suicide? 12 A I don't remember 13 specifically, but when the questions were asked of 14 our affiliates by regulatory authorities, the 15 affiliates would generally communicate that here, 16 usually through Doctor Weinstein to one of his 17 staff, and those questions would then go to parts 18 of the organization I was responsible for, and 19 therefore in various meetings or by copies of 20 documents I would be shown what the questions were 21 and what we were doing to answer them. 22 (THOMPSON EXHIBIT NO. 12 MARKED FOR 23 IDENTIFICATION.) 24 THE WITNESS: I can't 330 1 really read the top couple of lines on Page 2, 2 but -- 3 MR. SMITH: You know, I 4 have a better copy. 5 THE WITNESS: I also 6 can't read the bottom line and I can't read the 7 top line on Page 3. In fact, it's hard to read 8 the whole first paragraph on Pages 2 and 3. I've 9 done the best I can to read it. 10 MR. SMITH: Yes, we're 11 trying to get you a better copy, Doctor. 12 MS. ZETTLER: Do you guys 13 have any objection with replacing this one for 14 that? It's a little bit different format, but 15 it's the same substance. 16 THE WITNESS: This is 17 much better, thank you. 18 MR. SMITH: What I would 19 like to do is mark the new one as Exhibit 2 or 20 whatever. 21 THE WITNESS: Do you want 22 this back? 23 MS. ZETTLER: Yes, let me 24 mark it real quick. 331 1 THE WITNESS: Yes, sir, 2 I've reviewed it. 3 (THOMPSON EXHIBIT NO. 13 MARKED FOR 4 IDENTIFICATION.) 5 Q (BY MR. SMITH) All 6 right, Doctor Thompson, Exhibit 12 and Exhibit 13 7 appear to be, in substance -- that is the words 8 written the same. Can you look and see if you can 9 tell if they seem to be the same? 10 A Well, the document 11 Exhibit 12 seems to be missing at least one line, 12 which in Document 13 is 3.B. That must be cut off 13 at the top of the page. 14 Q What page would that be? 15 A I'm sorry, it's between 16 the first and second page of Exhibit 12, there's 17 no line which corresponds to the line labeled 3.B, 18 which is on the second page of Exhibit 13. Other 19 than that, they're essentially the same. 20 MR. MYERS: Well, I do 21 notice some written words -- 22 A I'm sorry, you are 23 correct. There are some things written on Exhibit 24 12 that aren't written on Exhibit 13, but the 332 1 printed text is the same except again there's an 2 underline on 13 that isn't underlined on 12. 3 Q My inquiry was with 4 respect to the printing there, the documents may 5 have been produced from different files from 6 differently Lilly employees who may have made 7 those writings on the document. 8 If you look up in the top 9 left-hand page -- top left-hand of the first page 10 of Exhibit 12, you'll see a date 3/23/84? 11 A Yes, sir, I missed that. 12 Q And then if you'd look on 13 the top left of the first page of Exhibit 13, 14 you'll see a date 6/26/84? 15 A Yes, sir, you're correct, 16 those dates are different. 17 Q Can you explain that? 18 I'm not familiar with the Lilly computer system or 19 telex system, if that is any explanation as to why 20 there would be two different dates on these 21 documents. 22 A I'm sorry, I have no 23 idea. I've not seen that before and I missed it 24 this time. 333 1 Q Do you know what those 2 numbers and letters mean up in the top left-hand 3 side of the page of each of those documents? 4 A No. I presume the 5 fourteen, colon, forty GMT is the Greenwich mean 6 time. 7 Q Yes? 8 A And the date obviously of 9 6/26 or 3/23, but I have no -- well, INO4 is 10 probably an Indianapolis 04 location, but that's a 11 guess because I don't know for certain. 12 Q Is that the same on -- 13 it's not the same on both documents, is it? 14 A No, I think that's GG, 15 INGG, or maybe it's six six. 16 Q Have you ever seen the 17 term INGG or six six before? 18 A I don't think so. 19 Q Under the IN66, there is 20 a -- on Exhibit 12, a number 0012, it appears? 21 A Yes, sir. 22 Q And then it's -- on 23 Exhibit 13 it's 0068, correct? 24 A Yes, sir. 334 1 Q And then it appears that 2 it gives two different Greenwich mean times 3 underneath that, does it not? 4 A Yes, sir. 5 Q Could these be computer 6 recordings of times when a particular computer 7 printed this document? 8 A Well, I don't know for 9 sure. I would assume that that header that you're 10 talking about was in fact computer generated, but 11 I really don't know how the system works. 12 Q Have you seen numbers 13 such as this in communications from your German 14 affiliate or within the Lilly corporate system? 15 A I honestly don't pay a 16 lot of attention to that type header, and 17 therefore I can't really say that I've got much 18 knowledge about it. 19 Q Do you know who would 20 know within the Lilly organization or have any 21 explanation for the difference in those two 22 headers? 23 A Well, somebody in the 24 systems area, but I'm not sure who manages -- Don 335 1 Neal is the chief person in Lilly's systems, but I 2 don't know who specifically manages the E-mail 3 system. 4 Q In any event, it appears 5 that the text is the same, does it not? 6 A Yes, sir, the printed 7 text starting with the names seems to be the same, 8 except for that one line that is cut off. 9 Q All right. But we 10 just -- and the date, where it says Bad Homburg 11 under the names -- no, it's on the first page. 12 A June 20. 13 Q Six? 14 A Well, one says June 29, 15 it looks to me, on Exhibit 12. 16 Q Yes. 17 A And it looks to me like 18 June 26 on Exhibit 13, so I missed another 19 difference. 20 Q But under it -- under 21 each of them it says Telex No. 005, correct? 22 A You're correct. 23 Q Any explanation for that? 24 A No, I don't really 336 1 recall. I mean, telex is different, obviously, 2 than E-mail, and I don't remember what systems 3 were being used in 1984. 4 Q Do these both appear to 5 be telexes to you? 6 A I honestly can't tell, 7 I'm not an expert in that area. 8 Q Let's talk about the 9 printed words there, and feel free, since they're 10 essentially the same, to use whatever copy is most 11 legible to you, Doctor Thompson. That was our 12 original intent in giving you another copy. 13 A Thank you. 14 Q But then we saw the 15 differences in the dates up on the upper left-hand 16 corner. It appears that, whether this is a March 17 or June document, that it is being sent from your 18 German affiliate, Doctor Schenk and Weber, to 19 Indianapolis? 20 A Yes, sir. 21 Q It's specifically 22 addressed to Ashbrook, Hardison and Stark, 23 correct? 24 A Yes, sir. 337 1 Q And then a number of 2 individuals are copied, including yourself? 3 A Yes, sir. 4 Q Do you recall receiving 5 this, sir? 6 A I can't recall the date 7 when I received it back in, I presume, '84; I've 8 seen it more recently, however. 9 Q Is this one of the 10 documents that you were shown in preparation for 11 your deposition? 12 A Yes, sir. 13 Q And a document that you 14 reviewed in preparation for your deposition? 15 A Yes, sir. 16 Q The subject of the 17 document is Fluoxetine - Registration Germany, 18 correct? 19 A Yes, sir. 20 Q And it indicates that 21 there is data that's been identified as essential 22 during discussions with the BGA of June 15th, 23 1984, correct? 24 A Yes, sir. That does make 338 1 the date of March 23 hard to explain, doesn't it? 2 Q Kind of curious, doesn't 3 it? Apparently there was a discussion with 4 marketing individuals at the Fluoxetine Symposium 5 at the 14th C.I.N.P. Congress, is that right? 6 A Yes, sir. 7 Q Can you tell me what that 8 is, sir? 9 A That's a very prestigious 10 academic group, and -- the N and P are 11 neuropsychiatry, I think I is international, I 12 think C is -- 13 Q Congress? 14 A -- something, congress, 15 council, committee, something. 16 Q And they were meeting 17 somewhere, correct? 18 A That seems to be what it 19 says. 20 Q And is it your impression 21 that there were individuals who were members of 22 the BGA and Lilly people at this meeting? 23 A I'm sorry, I've never 24 attended one of these meetings, so I don't have 339 1 any idea. 2 Q Is that your impression 3 or from reading the first paragraph of the 4 exhibit? 5 A Oh, I'm sorry. Well, I 6 don't necessarily conclude from the first 7 paragraph that BGA personnel were at the C.I.N.P., 8 it seems to me what it says is that there were 9 medical, slash, marketing persons that had 10 discussions at the C.I.N.P., and then they had 11 discussions at the BGA on June the 15th. I'm not 12 sure that, of my knowledge, I could tell you 13 whether BGA people go to C.I.N.P. or not. 14 Q Well, it says that the 15 discussions were on the occasion of the Fluoxetine 16 Symposium, 14th C.I.N.P. Congress, correct? 17 A That's what it says. 18 Q Do you recall that there 19 was a fluoxetine symposium at the 14th C.I.N.P. 20 Congress? 21 A No, sir. 22 Q All right. In any event, 23 Doctors Schenk and Weber are reporting to you, or 24 to Indianapolis, the results of this meeting with 340 1 the BGA, are they not? 2 A Yes, sir. 3 Q And raising issues that 4 were raised to them by the BGA? 5 A I think that's what this 6 means. 7 Q Item 2 says: The BGA 8 stated that there is a disagreement between 9 patient's and doctor's judgment of efficacy. 10 Since in their opinion the patient's impression is 11 more important, and we have to demonstrate 12 correlation between SCL 58 and HAMD and CGI and 13 PGI -- R-E-S-P, I assume that's respectively? 14 A Yes, sir. 15 Q Paren, perhaps by graphs, 16 close paren, period; correct? 17 A Yes, sir. 18 Q Apparently it was the 19 BGA's position that what the patient was reporting 20 in those portions of the clinical form that 21 recorded the patient's impressions was more 22 important concerning efficacy than what the scales 23 were rating, is that correct? 24 A Well, the -- one of the 341 1 global impression things is not -- well, that's 2 one question scale, I think you're correct. They 3 said the patients' rather than the psychiatrists' 4 opinions counted more for that. 5 Q In other words, what a 6 patient was reporting to a psychiatrist concerning 7 whether or not Prozac was working as an 8 antidepressant was felt by the BGA to be more 9 important than what the psychiatrist's opinion 10 was, is that right, as reflected by the rating 11 scales? 12 A Well, the answer is a 13 little bit complicated because in fact the SCL 58 14 could be called a rating scale, HAMD is certainly 15 a rating scale, but CGI and the PGI are the 16 clinician's and the patient's global impression of 17 efficacy and safety rolled together, and I guess 18 you could call it a scale even if it just has one 19 question. 20 Q All right. 21 A In the US -- I mean, the 22 reason for the paragraph is that the FDA pays far 23 more attention than to the HAMD, which, as you 24 know, is elicited from the patient but in fact 342 1 actually scored by the psychiatrist -- 2 Q And you have to be 3 trained to score the HAMD, do you not? 4 A Well, not really. There 5 are people in medicine that have had patients 6 administer it to themselves using computers. But 7 the way we did the studies, we trained the 8 psychiatrists to make sure that they would be 9 consistent in how they scored it. 10 Q That's what I was going 11 to say, Lilly felt that it was important that 12 Lilly provide their investigators with training on 13 how to administer the HAMD, correct? 14 A Yes, sir, that's correct. 15 Q And so the HAMD is 16 something that the investigator is taught how to 17 score, is that right? 18 A That's correct. 19 Q Whereas the patient's 20 impression is something that is just recorded by 21 the physician in responding to the patient with 22 respect to -- or in asking the patient how the 23 product was working or how they were feeling? 24 A Well, again, your -- it's 343 1 difficult to answer that because the Hamilton 2 depression questions are in fact elicited from the 3 patient. The psychiatrists, by looking at the 4 patient or examining the patient without talking 5 to them, would have no way to complete the 6 Hamilton depression scale. The clinical global 7 impression or the patient global impression is 8 when you ask the patient one straightforward 9 question about how you are. The Hamilton 10 depression scale, as you know, has twenty-one 11 specific questions that you ask the patient, how 12 did you sleep, what are you thinking about and so 13 forth. 14 Q Do you think that the 15 Hamilton depression scale is more sophisticated 16 because it asks more questions? 17 A I think you have to put 18 all the data together. It certainly provides 19 more, it's a richer domain in terms of the 20 information you can get, but I -- and I'm not 21 really qualified to compare one of these versus 22 another because I'm not a psychiatrist. 23 Q But those psychiatrists 24 at the BGA apparently felt that the patients' 344 1 impressions were more important, did they not? 2 A It's very unlikely that 3 there's a psychiatrist working for the BGA. 4 Q Oh, really? 5 A Yes, sir. 6 Q Why do you say that, sir? 7 A Because the BGA does not 8 have very many physicians in their employ. 9 Q Well, the testimony up to 10 date, sir, is that the BGA employs a commission 11 system, and that they employ a specific commission 12 that reviews applications, and that commission is 13 comprised of experts from varying fields, and if 14 the drug is a psychiatric medication, the 15 commission will have psychiatrists on it. 16 MR. FREEMAN: Objection, 17 I believe that slightly misstates the testimony in 18 that the people on the commission are independent 19 of the BGA, they are outside, not employed by the 20 BGA, and I think that's why you and the witness 21 are not communicating. 22 A It's my understanding 23 that the European regulators do use experts, and 24 I'm sure they pay them, but they are not full-time 345 1 employees of the commission, so I don't know 2 exactly -- I've not heard it called a commission 3 before, but I'm sure the BGA uses academic 4 experts; but my knowledge of the people that I've 5 met with at BGA would suggest that there are very 6 few physicians who are full-time employees of the 7 BGA. 8 Q Well, obviously, the BGA 9 can only make opinions such as this through 10 physicians and psychiatrists who are competent, 11 can they not? 12 A I wouldn't want to 13 denigrate the competence of psychologists and 14 other scientists in this area. Again, I'm not a 15 neuroscientist. 16 Q Well, the document says 17 that the BGA stated that there is disagreement -- 18 that there is disagreement between patient's and 19 doctor's judgment of efficacy, since in their 20 opinion the patient's impression is more 21 important, correct? 22 A That's what it says. 23 Q So the BGA, whether they 24 were speaking through psychiatrists or through 346 1 bureaucrats, were saying we think the patient's 2 impressions are more important? 3 A That's what they say; 4 that's what this memo purports for them to have 5 said. 6 Q Well, do you think that 7 this memo is inaccurate? 8 A I have no basis for 9 knowing that; I don't think it's inaccurate. 10 Q Item 7 says the BGA 11 explained their reservations regarding CNS side 12 effects, does it not? 13 A Yes, sir. 14 Q Were you aware of CNS 15 side effects that had been seen during the 16 clinical trials on Prozac? 17 A Yes, sir. 18 Q Item 10 states: 19 Comparative use of concomitantly taken hypnotics 20 and benzodiazepines in agitated/retarded 21 fluoxetine patients versus agitated/regarded 22 patients on comparators. Reason: The BGA 23 suspects fluoxetine to be a stimulating/activating 24 drug, paren, side effect profile, suicides, 347 1 suicide attempts, close paren. 2 A That's what it says. 3 Q Correct? 4 A Yes, sir. 5 Q Were you aware that there 6 were concomitantly taken hypnotics and 7 benzodiazepines employed in the Prozac clinical 8 trials in the United States? 9 A At least in some of them, 10 yes, sir. 11 Q Were you aware that there 12 was some reports in the Prozac clinical trials in 13 the US that Prozac was a stimulating/activating 14 drug? 15 A I don't recall those 16 words specifically. 17 Q Generally? 18 A There were reports of a 19 whole variety of what we would classify as CNS 20 events that were seen in the people who were -- by 21 intent to treat were given Prozac, also the 22 comparators. But I don't recall, in the NDA for 23 example, the phrase stimulating/activating events 24 or side effects or anything like that. 348 1 Q When we use the term CNS, 2 we're talking about the central nervous system, 3 are we not? 4 A Yes, sir. 5 Q And stimulating and 6 activating are central nervous system type 7 responses, are they not? 8 A Yes, sir. 9 Q Item 12 says that the BGA 10 stated that due to the accumulation of fluoxetine 11 we should consider to recommended a lower 12 maintenance dose after having achieved a certain 13 relief of acute systems, correct? 14 A Yes, sir. 15 Q Were you aware of the 16 clinical trial data on fluoxetine that established 17 that there was an accumulation of fluoxetine and 18 its metabolites -- 19 A Yes, sir. 20 Q -- in individuals 21 following initial administration of the dosage? 22 A Yes, sir. 23 Q Had you ever seen or had 24 any physicians in the medical division at Lilly 349 1 ever recommended to you, up to this time, a lower 2 maintenance dose with respect to Prozac? 3 A Maintenance -- well, it 4 was, I think, at this time or even before this 5 time that Joe Wernicke, Bob Zerbe and I designed 6 dose response studies, which we've been told were 7 the first ever done with an antidepressant, or for 8 that matter any major CNS drugs, so we were 9 vitally interested in defining the dose response 10 relationship. Those studies were initially 11 designed to look at a constant dose throughout 12 therapy rather than distinguishing between an 13 initial and a maintenance dose, although having 14 the results from the first studies, we then went 15 on to design a second study which would look at 16 whether or not a higher dose would work in people 17 who had failed to respond adequately to a lower 18 initial dose. Now, in this era, I'm not aware 19 that we did any studies that looked at giving a -- 20 in essence, a larger dose to begin with, and then 21 going to a lower dose for maintenance. There were 22 observations on long term therapy, but I don't 23 think the dose was controlled. 24 Q But there has never been 350 1 a study looking at a lower maintenance dose than 2 had initially been given? 3 A Not in a controlled 4 sense, I don't believe. 5 Q Item 14 indicates, quote, 6 as we already explained by our telex to Doctor 7 Zerbe of June 8th, 1984, we need a careful 8 analysis of suicide and suicide attempts: Patient 9 by patient, symptomatology/severity upon entry 10 into the study and week by week until the event 11 occurred, dose of fluoxetine, side effects, 12 et cetera. This is a very serious issue in the 13 opinion of the BGA. It might well be that we have 14 to recommend concomitant tranquilizer intake for 15 the first two or three weeks in the package 16 literature, end quote. Correct? 17 A Essentially, yes, sir. 18 Q Did you count any 19 mistakes? 20 A Yes, sir, one, but that's 21 minor. 22 Q As I said, I'm getting 23 better. Had you seen the telex to Doctor Zerbe of 24 June 8th? 351 1 A I don't think so. 2 Q Have you seen the telex 3 of June 8th to Doctor Zerbe? 4 A I can't remember that 5 specifically. 6 Q Is that a document you 7 reviewed in preparation for your deposition here, 8 sir? 9 A Not that I can recall, 10 but again, as I said yesterday, I reviewed at 11 least four dozen documents probably. 12 Q Had anybody at Lilly, any 13 of the physicians at Lilly argued or suggested at 14 the time that Prozac be used in connection with a 15 tranquilizer? 16 MR. FREEMAN: Are you 17 speaking with every patient or not? 18 MR. SMITH: In any 19 situation. 20 A Well, we had had a lot of 21 discussion in planning new studies, because 22 remember, as I said yesterday, the studies that 23 were in the NDA had already been put in place when 24 I came to Lilly, but in designing new studies we 352 1 had had a lot of internal debate about what kinds 2 of sedative or tranquilizing drugs could be 3 allowed in those studies. But in terms of actual 4 use by patients outside of the clinical trial 5 paradigm, I don't recall any discussion up to this 6 time at all. The doses that had been employed 7 were, in my opinion, very, very small and unlikely 8 to have much of a measurable effect. But since 9 patients can have multiple diseases, the debate 10 largely was what do you do with a patient that has 11 depression, but also has anxiety or some other 12 reason to have a CNS active drug. And the purpose 13 of the debate is obviously you don't want to 14 confound the results of whatever it is you are 15 primarily trying to treat and measure. 16 Q I'm talking about 17 treatment as opposed to whether or not you are 18 going to get an accurate picture of your clinical 19 trials if you employ concomitant hypnotics. 20 A No, up until this -- 21 Q I'm talking about you've 22 got a patient who begins treatment on Prozac, 23 becomes anxious or agitated or develops insomnia, 24 the investigator employs benzodiazepines to 353 1 relieve the anxiety or prescribes chloral hydrate 2 for sleep to relieve the insomnia. Did anybody 3 ever suggest that that would be an appropriate use 4 of -- or an appropriate treatment for patients 5 with these type of side effects of Prozac? 6 A Well, certainly not up 7 until this time. I think that was your original 8 question. 9 Q Yes. 10 A No, I don't recall any 11 such discussions. 12 Q Up until that time, the 13 only discussions that had been had at Lilly were 14 in connection with whether or not using the 15 benzodiazepines and the hypnotics like chloral 16 hydrate in the clinical trials would confound the 17 results of the clinical trials, correct? 18 A Yes, sir. 19 Q Not whether or not it 20 would be a better treatment option for use with 21 certain patients with Prozac, correct? 22 A I don't recall any such 23 conversations before this time addressing that 24 issue, right. 354 1 Q Item 15 concludes that 2 this is not a very promising situation in Germany, 3 correct? 4 A Well, it says -- yes, 5 which could help in a not very promising 6 situation. 7 Q Do you recall -- 8 specifically you did, Doctor Thompson, 9 specifically in connection with this request by 10 Doctor Schenk and Weber? 11 A I'm sorry, you said do I 12 recall anything specifically -- 13 Q Anything, yes, sir. 14 A -- that I specifically 15 did? 16 Q Yes. 17 A No. 18 Q Or do you recall any 19 specific directions you gave, sir, in response to 20 this specific telex or E-mail, whichever it is? 21 A No, sir. 22 Q Were Mrs. Ashbrook -- is 23 it Doctor Hardison? 24 A Yes. 355 1 Q And Doctor Stark 2 individuals under your indirect supervision at 3 this time, in June 1984? 4 A Ms. Ashbrook certainly 5 was; Doctor Stark certainly was; I don't think 6 Doctor Hardison was at that time, although the 7 statistics group I became responsible for at some 8 time. 9 Q Is Doctor Hardison a 10 statistician? 11 A Yes, sir. 12 Q Do you remember anything 13 that was generally done in connection with that 14 telex, specifically that telex or E-mail? 15 A I remember that the 16 various people that I was responsible for began 17 working on trying to answer these questions and do 18 the analyses that had been requested. 19 (THOMPSON EXHIBIT NO. 14 MARKED FOR 20 IDENTIFICATION.) 21 THE WITNESS: Yes, sir. 22 Q (BY MR. SMITH) Have you 23 received -- or have you seen Exhibit 14 before? 24 A Well, it was addressed to 356 1 me in part, but I can't recall it. 2 Q Is Exhibit 14 one of the 3 documents that you reviewed in preparation for 4 your deposition today? 5 A I don't believe so. 6 Q And Exhibit 14 is a -- 7 includes a cover letter including a comment from 8 the medical in Germany in connection with Prozac, 9 correct? 10 A I'm not sure what 11 von Keitz's role was in Germany. 12 Q Do you know B. von Keitz? 13 A I've heard the name 14 before, but I don't honestly know what that 15 person's role was. 16 Q Well, the subject of his 17 letter is fluoxetine, and he says yesterday we 18 unofficially received a copy of the medical 19 comment on our fluoxetine application, a 20 translation is attached, doesn't he? 21 A Yes, sir. 22 Q So I take it from that, 23 that this is a comment from the BGA, the medical 24 comment from the BGA which was unofficially 357 1 secured by your German affiliate in connection 2 with the Prozac application in Germany. Now, do 3 you think that's an unreasonable conclusion? 4 A No, I think that's very 5 reasonable. 6 Q And in connection with 7 whether or not there's any doctors at the BGA, 8 apparently this is a medical comment authored by 9 physicians at the BGA or at the commission, is it 10 not? 11 A I have no idea who 12 authored it. 13 Q Well, I'm assuming that 14 somebody with some medical background has authored 15 it, otherwise it wouldn't be entitled by Mister or 16 Doctor von Keitz as a medical comment? 17 A I don't have any idea who 18 authored this document. 19 Q Well, neither do I, other 20 than what is transmitted and what is stated by a 21 Lilly employee? 22 A But you're making 23 assertions about whether it was psychiatrists or 24 not. There are lots of other people who are 358 1 expert in the CNS area. 2 Q I didn't use the word 3 once in connection with this exhibit, I used the 4 term medical every time. 5 A Medical doctor, yes, 6 sir. I don't know whether that's true or not. 7 Q Do you think this was 8 written by a lay person, or someone without any 9 medical training, this comment on the clinical 10 documentation, Doctor Thompson? 11 A Well, I certainly think 12 it's possible that it was written by a 13 statistician or by a doctor of pharmacy or by a 14 doctor of nursing or somebody without a doctoral 15 degree who is expert in this area. I don't know. 16 Q Your employees are 17 calling it a medical comment, are they not? 18 A We receive medical 19 comments from people who do not have an MD degree 20 all the time. 21 Q Whatever, it's called a 22 medical comment by von Keitz, isn't it? 23 A It says the medical 24 comment, yes, sir, that's what it says. 359 1 Q Do you recall seeing this 2 document at or about the time it was received by 3 you? 4 A Not specifically, no. 5 Q Do you recall receiving 6 it generally in May of 1984? 7 A No, not specifically. 8 Q The last -- well, is it 9 your understanding that the documentation that was 10 sent to the BGA was the same documentation that 11 was sent to the United States Food and Drug 12 Administration? 13 A I have no idea. 14 Q Well, is that a 15 reasonable assumption on your part? 16 A I wouldn't have any idea 17 what to assume. 18 Q Well, do you know of 19 any -- do you know of a situation that has existed 20 where the German government has gotten medical 21 information concerning the Prozac clinical trials 22 that the United States Food and Drug 23 Administration has not received, Doctor Thompson? 24 A I don't have any idea, 360 1 sir. I have no way of making any sort of judgment 2 on that. 3 Q So is it possible then, 4 as far as you're concerned, that the BGA received 5 different or additional documentation or had data 6 before them that the United States Food and Drug 7 Administration did not have, sir? 8 A I'm sure that we complied 9 with the regulations which later said that the NDA 10 in the United States needed to contain all 11 information that was pertinent, but at this 12 specific time, since I've not seen the application 13 in Germany, and since I don't read German very 14 well, I have absolutely no idea what's in it. 15 Q So it may be that the 16 German government is looking at documents and data 17 that the United States Food and Drug 18 Administration didn't have when they're making 19 these medical comments, is that correct? 20 A That's a possibility. 21 Q Now, the application to 22 approve Prozac in the United States was in fact 23 pending at this very time in May 1984, was it not? 24 A Yes, sir, it had been 361 1 submitted. 2 Q The NDA had been 3 submitted and it was under consideration by the 4 United States Food and Drug Administration, was it 5 not? 6 A Yes, sir. 7 Q All right. The last 8 paragraph of the first page of the medical comment 9 states: Of the forty-six attached study 10 protocols, in twenty-five the note is to be found 11 that these studies are not completed. Each 12 double-blind study is preceded by a one week 13 placebo washout period. As statements on the 14 medications, which were used in the pretreatment 15 of the depression, are missing, the question 16 occurs if firstly this period was not too short 17 and on the other hand, if a carry-over-effect of 18 these drugs had not influenced the result of the 19 first examination before the beginning of the 20 studies, end quote. Did I read that correctly? 21 A Yes, sir. 22 Q Did I make any mistakes 23 in reading that? 24 A No, sir. 362 1 Q At this time, in May 2 1984, were there study protocols or studies that 3 had not been completed? 4 A I don't know for 5 certain. As I indicated earlier, three of us 6 had -- with others, had been designing dose 7 response protocols. I think they had begun by 8 this time, but I'm not certain. 9 Q Well, do you take it from 10 reading this paragraph that I've just read that 11 Eli Lilly and Company had submitted protocols and 12 results from forty-six studies, but that there 13 were twenty-five of them that had not been 14 completed? 15 A That's exactly what the 16 paragraph says. 17 Q All right. Were you 18 aware that that situation existed at that time, 19 Doctor Thompson? 20 A No, sir. 21 Q Is it correct at the time 22 that each of the double-blind studies had a one 23 week placebo washout period? 24 A Some of them had a one 363 1 week placebo period; I wouldn't want to testify 2 that all of them did because I'm not certain on 3 that. 4 Q Do you have any expertise 5 or do you have any knowledge, sir, concerning what 6 a proper period of placebo washout is necessary 7 for studying a medication such as Prozac? 8 MR. FREEMAN: On what 9 drug, because I think it might vary? 10 A Well, as a clinical 11 pharmacologist, I guess I have some general 12 expertise in the area of design of clinical 13 trials. 14 Q All right. So do you 15 have an opinion on what the proper washout period 16 would be on an antidepressant drug such as Prozac? 17 A I think your question was 18 in the study of an antidepressant drug such as 19 Prozac, what would the washout period be, and the 20 washout, I think, as Mr. Freeman indicated, would 21 really refer to all the other drugs the patient 22 might have been taking at the time. 23 Q So? 24 A I can conceive of 364 1 situations where you would want to have a longer 2 washout period with certain drugs that aren't 3 ordinarily taken, but that's why the investigator 4 uses the investigator's judgment in enrolling 5 patients in a trial. 6 Q And this is something 7 that is being questioned by the medical comment at 8 the BGA, correct? 9 A Yes, sir, that's the way 10 it reads. 11 Q Additionally there, they 12 indicate there is a possibility that if you just 13 have a one week washout period that you might see 14 some carry over of the other drugs that might have 15 been taken prior to the patient beginning the 16 trial being -- influencing the first treatment or 17 the first assessment of the patient, is that 18 correct? 19 A That's what it says, but 20 it says that they didn't have any data at all 21 about what those medicines were, so they didn't 22 have any scientific basis of knowing whether that 23 was a legitimate concern or not. 24 Q Do you know why that data 365 1 wasn't submitted by Lilly to the BGA? 2 A I have no idea. 3 Q Turn to Page 3 of the 4 document. The medical comment of the BGA at that 5 time was, in the first paragraph, quote: The 6 frequency of side effects was very high, partly 7 more than ninety percent, and the side effects 8 resulted narrowly in each study in dropouts. The 9 frequency of side effects depended on the dose, 10 the age and the duration of therapy. 11 I can't read that next 12 word, can you, sir? 13 A I think it's a misspelled 14 deciding, but I don't know what that means. 15 Q Deciding for the clinical 16 significance of side effects is not only the 17 frequency of their occurrence, but also their 18 severity, end quote. Correct? 19 A Yes, sir. 20 Q Do you know whether or 21 not Lilly's analysis of the data that was 22 submitted to the United States Food and Drug 23 Administration or the BGA indicated that there was 24 a frequency of up to ninety percent side effects? 366 1 A I don't remember the 2 number being quite that high, but that in fact is 3 not an unusual frequency in a six-week trial. 4 Q Do you recall that the 5 frequency of side effects with respect to Prozac 6 was, as the medical comment in Germany states, 7 quote, very high, end quote? 8 A No, I don't believe that 9 the analysis that I'm aware of, of the data in the 10 NDA, I would have used the word very high. 11 Q Do you recall these side 12 effects resulting in study dropouts of patients 13 who had experienced these side effects? 14 A Oh, yes, sir. 15 Q Do you recall that the 16 frequency of the side effects depended on the 17 dose, the age and the duration of therapy? 18 A Well, that is very 19 difficult from the original NDA data, to make that 20 kind of assertion, because most of the patients 21 from most of their duration of therapy were 22 treated with sixty or eighty milligrams. So the 23 differentiation between those two doses is 24 difficult. And in terms of the age effect, later 367 1 data would suggest to me that in fact that's 2 correct. But I'm not sure those data are 3 available from the data in the original NDA. 4 Now, the duration of 5 therapy is also clearly a factor in terms of the 6 side effects in terms of when they occur and how 7 long they last. 8 Q So age and duration of 9 therapy are clearly related to the side effects, 10 is that right? 11 A And doses -- I mean, from 12 other data, dose clearly is related to the 13 incidence of the side effects. I'm just saying I 14 think it would be difficult to do that analysis on 15 the data in the original US NDA. 16 Q All right. Well, at 17 least the medical comment makes that comment 18 concerning the data that was before them at that 19 time, did it not? 20 A Well, it doesn't say the 21 dose of fluoxetine, and if you remember, the 22 tricyclic comparators clearly have dose dependent 23 adverse event incidents; so, it said in the 24 trials, the trials included placebo, three 368 1 tricyclics and fluoxetine, and I don't know what 2 this refers to. 3 Q Why would they be 4 commenting on the side effects critically that 5 occurred with the comparator drug or with placebo 6 when they have under application Prozac, 7 fluoxetine? 8 A Well, in fact I think the 9 incidence of side effects was at least as high on 10 the tricyclic comparators as on fluoxetine. 11 Q Is that the way you read 12 this, Doctor Thompson, that they're talking about 13 the comparators or the placebo instead of Prozac 14 when they talk about the frequency of the side 15 effects and the severity of the side effects and 16 the fact that there is dropouts in connection with 17 side effects, you think that medical comment is 18 talking about something other than Prozac? 19 A Yes, sir, absolutely. 20 Q All right. 21 A I think the dropouts are 22 very relevant to the ability to analyze the data, 23 and that would be dropouts on any form of therapy. 24 Q All right. In the fifth 369 1 paragraph of that same page, the medical comment 2 in Germany is: In fifteen -- is that "or" or 3 "to"? 4 A I think it's a hyphen, 5 but that's -- 6 Q Does that make it 7 difficult for you to read, the stamp across this, 8 Doctor Thompson? 9 A At that particular place, 10 yes, sir. 11 Q In fifteen to twenty 12 percent of cases side effects occur which involve 13 the central nervous system. As most of them 14 resemble the clinical picture of the underlying 15 disease, even from theoretical reasons -- even 16 from theoretical reasons one has to expect an 17 intensification and not an improvement of 18 symptoms, end quote. Correct? 19 A Yes, sir. 20 Q Did your knowledge of the 21 data in the clinical trials at that time indicate 22 that fifty -- indicate that in fifteen to twenty 23 percent of the cases of side effects, this 24 involved central nervous system side effects? 370 1 A I thought it was actually 2 a little higher than that, but I think that's 3 about right. 4 Q The next paragraph says, 5 during the treatment with the preparation -- can 6 we say there, Doctor Thompson, clearly that that 7 has to do with Prozac? 8 A Probably. 9 Q All right. It says, 10 quote: During the treatment with the preparation, 11 sixteen suicide attempts were made, two of these 12 were a success. As patients with a risk of 13 suicide were exclude from the studies, it is 14 probable that this high proportion can be 15 attributed to an action of the preparation in the 16 sense that -- of a deterioration of the clinical 17 condition which reached its lowest point, end 18 quote. Correct? 19 A Yes, sir. 20 Q Were you aware that in 21 May 1985 the data indicated that there were 22 sixteen suicides attempts in the clinical trials 23 on Prozac? 24 A I'm not sure I knew that 371 1 specific number, but in that ball park, yes. 2 Q Were you aware at that 3 time that many of the clinical trials excluded 4 individuals who had a serious risk of suicide? 5 A Yes, sir, a serious risk 6 of suicide, although those same trials were 7 admitting people who in fact had suicidal ideation 8 at the time of admission. 9 Q I said serious risk of 10 suicide because that's specifically what's written 11 in the protocols, isn't it? 12 A Yes, sir. 13 Q So was there any need to 14 expand on my question in your answer? 15 A I just want to tell the 16 whole truth, sir. 17 Q Are you trying to tell 18 the truth or trying to argue with me? 19 A I never try to argue with 20 you, sir. 21 Q Had you ever, up to May 22 1985, Doctor Thompson, examined these sixteen 23 suicide attempts to determine whether or not this 24 could be an action of Prozac? 372 1 A I think you mean 1984, 2 and the answer is yes, I had examined them, at 3 least many of them. 4 Q You had personally 5 examined them yourself? 6 A Yes, sir, at least some 7 of them. 8 Q Had you visited with any 9 of the investigators that had reported these 10 suicides to you at that time? 11 A No, sir. 12 Q Had you visited with any 13 of the patients who had attempted suicide at that 14 time, sir? 15 A No, sir. 16 Q Had you visited with any 17 of the treating physicians that might have also 18 had knowledge concerning the patients; in other 19 words, an internist, an OB/GYN, that was also 20 treating the patient -- 21 A No, sir. 22 Q -- for their general 23 health needs -- 24 A No, sir. 373 1 Q -- to determine whether 2 or not that suicide was related to the use of the 3 drug? 4 A No, sir, I hadn't visited 5 any physicians outside of Lilly on this issue. 6 Q Is it your testimony you 7 visited with Lilly -- with physicians at Lilly? 8 A Oh, yes, sir. 9 Q Do you know whether or 10 not they had spoken with any of the investigators? 11 A They told me that they 12 had talked with the investigators. Now, by 13 physician I'm including Doctor Stark, as well as 14 Doctors Bennett and Dobbs and the other 15 psychiatrists that were involved in this, and I 16 also talked to CRA's who were part -- they weren't 17 called CRA's at that time, who were part of the 18 team in terms of working up and getting more data 19 on adverse events. 20 Q Did you ever see any -- 21 well, were there 1639's filed in connection with 22 all of those sixteen attempts? 23 A I can't testify that they 24 were filed with all of them because in a clinical 374 1 trial, 1639 is it not required under the 2 regulations, even today, and so I'm not sure what 3 the mechanism of safety reporting was, whether 4 that was just included in the NDA or whether there 5 were IND safety reports made or not. 6 Q Is it your testimony that 7 if somebody attempted suicide and was hospitalized 8 while in a Prozac clinical trial, that a 1639 9 would not be required by the United States Food 10 and Drug Administration? 11 A The FD-1639 form was 12 specifically referenced in the NDA regulations 13 rather than the IND regulations. The IND 14 regulations actually specified different means of 15 reporting, different timing of reporting, 16 different criteria for reporting, and as a matter 17 of fact somewhere around here we were having a big 18 debate with the FDA as to whether they'd even 19 allow us to use the FD-1639 for reporting events 20 within a clinical trial. 21 Q Did the Food and Drug 22 Administration, Doctor Thompson, know in May 1984 23 that there had been sixteen suicide attempts in 24 the Prozac clinical trial? 375 1 A They knew -- I can't 2 testify that they knew of the number sixteen. 3 They knew everything that we had a regulatory 4 obligation to report, and they certainly, in the 5 New Drug Application, had all of the data that was 6 pertinent from those clinical trials, but I don't 7 know whether the number was exactly sixteen or 8 not. 9 Q Well, would that data be 10 pertinent -- 11 A Yes, sir. 12 Q -- to the safety of the 13 product, sixteen suicide attempts? 14 A Yes, sir, if that's the 15 right number. 16 Q Well, do you have any 17 reason to conclude that the sixteen suicide 18 attempts is an incorrect number? 19 A Well, you earlier asked 20 me questions about whether or not there were 21 ongoing trials and so forth, so I don't know what 22 was submitted to the BGA and what time frame, and 23 whether there were additional cases that were 24 included or some cases omitted in that 376 1 application. I don't have any knowledge of what 2 was submitted to the BGA, and this number seems to 3 come from the BGA submission rather than from the 4 NDA submission to the FDA. 5 Q Well, the sixteen suicide 6 attempts would have been data that would have been 7 obtained by the BGA based on data supplied by 8 Lilly, would it not? 9 A Yes, sir. 10 Q You don't know that the 11 BGA was making any outside inquiry concerning 12 other suicide attempts that might have been made 13 by individuals on Prozac in counting these sixteen 14 at that time, do you? 15 A I don't think they would 16 have done that. 17 Q The fourth page has a 18 statement -- has a section where it says, quote, 19 summarizing opinion, does it not? 20 A Yes, sir. 21 Q The second point there 22 is, quote, considering the benefit and the risk, 23 we think this preparation totally unsuitable for 24 the treatment of depression, end quote, correct? 377 1 A Yes, sir. 2 Q Do you recall any 3 specific action you took, Doctor Thompson, in 4 connection with this specific comment on the 5 clinical documentation when you received it? 6 A No, sir, I don't recall 7 any specific action. 8 MR. FREEMAN: When you 9 get to a good place, let's take a little break. 10 MR. SMITH: Okay, we may 11 as well now. 12 (SHORT BREAK TAKEN.) 13 MR. BOUR: The time is 14 10:45. 15 Q (BY MR. SMITH) Doctor 16 Thompson, there may be some confusion here on my 17 part. Was it your testimony earlier that in May 18 of 1985 -- '84; May 1984, that there may have been 19 some suicide attempts that occurred on the Prozac 20 clinical trials that were -- that had not been 21 reported to the United States Food and Drug 22 Administration? 23 A I don't know that of 24 certainty because the reporting after the New Drug 378 1 Application was filed would be under the IND 2 rules, and depending upon the circumstances of the 3 suicide attempt, they may or may not have fallen 4 in a criteria to require rapid reporting. If they 5 didn't require expedited reporting under the old 6 IND rules at that time, the next time they would 7 be reported would be in an annual report, and I 8 have no idea what the annual report time was under 9 the IND. 10 Q But they should have been 11 included in an annual report at least? 12 A Depending upon the 13 character of the event, yes. 14 Q It's either going to be a 15 spontaneous or an annual report? 16 A Well, the annual report I 17 was referring to was the annual report under the 18 IND regulations, and spontaneous I think you use 19 to mean a postmarketing event. There are annual 20 reports to the NDA as well as annual reports to 21 the IND. 22 Q Well, is there a -- would 23 a suicide attempt be required to be reported to 24 the United States Food and Drug Administration, 379 1 however it's reported? 2 MR. FREEMAN: Wherever it 3 occurs? 4 MR. SMITH: Yes. 5 A Not necessarily; it 6 depends on other characteristics of the event. 7 Q All right. How about in 8 May of 1984, would it have been required to have 9 been reported in some form at that time? 10 A No, it would depend upon 11 other characteristics of the event other than just 12 saying it's a suicide attempt. 13 Q All right, so there might 14 have been suicide attempts that occurred prior to 15 May 25th, 1984 that were not reported to the 16 United States Food and Drug Administration? 17 A I'm not aware of any, but 18 that's a possibility for sure. 19 Q Because of the 20 differences in reporting requirements that you've 21 spoken of? 22 A And the characteristic of 23 each individual event. 24 Q Go back to Exhibit 14, 380 1 Item 7. 2 A I don't see an Item 7, 3 I'm sorry. 4 Q Maybe I've got the wrong 5 exhibit; it's the one dated June 26th? 6 A Oh, I'm sorry, I have 7 that as Exhibit 13. 8 Q All right. Do you see 9 Item 7 there? 10 A Yes, sir. 11 Q It says the BGA explained 12 their reservations regarding CNS side effects, 13 correct? 14 A Yes, sir. 15 Q Under that it says: 16 There have been a few patients complaining of 17 psychosis and hallucinations. Please provide us 18 with detailed report, whether those patients 19 suffered from psychotic depression, whether the 20 hallucinations developed during treatment or have 21 perhaps been present already at start of 22 treatment, or whether those events may indeed be 23 interpreted by -- quote within a quote -- 24 aggravation of disease, end quote within the 381 1 quote, end quote. Correct? 2 A Yes, sir. 3 Q Were you aware of 4 patients complaining of psychosis during treatment 5 with Prozac at that time? 6 A I think so, but I can't 7 be a hundred percent sure, but I think so. 8 Q Were you aware of 9 patients who had reported hallucinations during 10 Prozac treatment at that time? 11 A Yes, sir, I think that 12 was in the original NDA. 13 Q Are you familiar with the 14 term "psychotic depression"? 15 A I'm familiar with the 16 term, but I'm not sure I could precisely define it 17 for you. 18 Q Had you ever seen -- have 19 you ever seen the term "psychotic depression"? 20 A Yes, sir. 21 Q Would psychotic 22 depression be a psychiatric term that would be 23 used by psychiatrists? 24 A At least by some 382 1 psychiatrists. They have now codified both in the 2 United States and internationally the meaning of 3 certain terms, and what I learned as psychotic 4 depression I'm sure is not exactly the same 5 definition that would be used today. 6 Q But the term "psychotic 7 depression" might be a term that would have been 8 used by one of your investigators or psychiatrists 9 at the time? 10 A Yes, sir, it could have 11 been. 12 Q Did you look to see, in 13 connection with this specific telex, whether or 14 not this -- these hallucinations developed during 15 treatment or were already present at the start of 16 treatment? 17 A No, sir, I don't remember 18 doing any specific search. 19 (THOMPSON EXHIBIT NO. 15 MARKED FOR 20 IDENTIFICATION.) 21 THE WITNESS: Yes, sir. 22 Q (BY MR. SMITH) Exhibit 23 15 is a telex or a document received from Doctors 24 Weber and either Doctor or Mr. Chandler and Mayr, 383 1 correct? 2 A Yes, sir. 3 Q Is Chandler or Mayr a 4 physician, do you know? 5 A Mr. Mayr is not, and I'm 6 not sure about Chandler, I don't believe so. 7 Q It's dated January 29th, 8 1985 and concerns fluoxetine registration in 9 Germany, does it not? 10 A I can't confirm the '85; 11 what I've got looks like twenty-five, but if you 12 say so. 13 Q If you'll look down under 14 the names -- 15 A Oh, there, I'm sorry, 16 there is -- I'm sorry, you're correct. 17 Q So we're talking about 18 January 1985, approximately seven months after the 19 May medical opinion submitted to you from the BGA, 20 correct? 21 A Yes, sir. 22 Q It says: We unofficially 23 received confirmation that fluoxetine was 24 discussed by the Commission A at the BGA on 384 1 January 21st. Two major concerns seem to be the 2 reason that the registration was not accepted. 3 Correct? 4 A Yes, sir. 5 Q Now, here it clearly -- 6 wouldn't you agree that there was discussion by 7 this specific panel of physicians known as 8 Commission A -- 9 A I don't know who makes up 10 Commission A. 11 Q -- employed by the BGA? 12 A I have heard of a 13 Commission A that does something for the BGA, but 14 I have no idea who sits on it. 15 Q Well, Doctor Weinstein 16 has testified that it's a commission of 17 physicians. 18 A Doctor Weinstein is an 19 expert in this area and I'm not. 20 Q Would you accept my word 21 that that's what Doctor Weinstein said? 22 A I will always accept your 23 word, sir. 24 Q All right. The 385 1 Commission A then is, would you agree with me, a 2 commission of physicians? 3 A Yes, sir. 4 Q And there they apparently 5 had reviewed the application of Prozac, 6 fluoxetine, and expressed two major concerns which 7 resulted in the registration not being accepted, 8 correct? 9 A Yes, sir. 10 Q One was efficacy was 11 questioned, correct? 12 A Yes, sir. 13 Q By efficacy, we mean the 14 product was not felt by them to be suitable to 15 show any improvement in the disease being treated? 16 A That's generally correct. 17 Q All right. The second 18 reason that the registration of fluoxetine was not 19 accepted by Commission A, as stated here, was 20 suicidal risk, correct? 21 A Yes, sir. 22 Q Had any other commission 23 that had been established by regulatory bodies up 24 to this time found that the data with respect to 386 1 the Prozac clinical trials suggested a suicidal 2 risk in connection with Prozac? 3 A Not to my knowledge. 4 Q Doctor Weber, et al., 5 suggests an action plan, do they not? 6 A Yes, sir. 7 Q They suggest a meeting 8 with -- they say our clinical expert to discuss 9 the possible ramifications of the BGA's position 10 toward the product, correct? 11 A Yes, sir. 12 Q Do you know who your 13 clinical expert was at that time? 14 A No, sir. 15 Q They also suggest 16 immediate follow-ups on all key opinion leaders on 17 the BGA commission for selected visitation next 18 week, do they not? 19 A Yes, sir. 20 Q So do I take it by that 21 that the local individuals there at Lilly were 22 going to talk to members of this commission 23 concerning this situation? 24 A Well, it might be that 387 1 investigators or other experts in Germany were 2 going to be asked by Lilly employees to visit with 3 the commission members; I wouldn't necessarily 4 interpret this as meaning Lilly employees were 5 going to visit with them. 6 Q But they were going to 7 have selected visitation with those opinion 8 leaders on the commission? 9 A That's what it says, yes, 10 sir. 11 Q Concerning, I assume, the 12 fact that fluoxetine had been not accepted in 13 Germany because of the efficacy problems and the 14 suicidal risk issue, correct? 15 A I think that's correct. 16 Q They conclude by saying: 17 It is our intention to review all appropriate 18 communication channels prior to the official 19 response by the BGA. All pertinent information 20 would be communicated to keep everyone informed on 21 this important issue. Correct? 22 A Yes, sir. 23 Q It also indicates that 24 there should be an official letter within four 388 1 weeks that would give a clear understanding of all 2 issues, correct? 3 A Yes, sir. 4 Q Did you do anything 5 specifically in connection with this telex of 6 January 29th, 1985? 7 A Well, I'm not an 8 addressee and I'm not sure I ever saw it until 9 recently. 10 Q So your answer would be 11 no? 12 A Yes, sir, the answer is 13 no. 14 Q And when you say I saw it 15 recently, I'm assuming that this is one of the 16 documents that you've seen in connection with 17 preparation for your deposition here? 18 A I think that's correct, 19 sir. 20 (THOMPSON EXHIBIT NO. 16 MARKED FOR 21 IDENTIFICATION.) 22 THE WITNESS: Yes, sir. 23 Q (BY MR. SMITH) Exhibit 24 16 is a telecopy message from Doctor Weber to 389 1 Doctor Weinstein that encloses the intent to 2 reject letter from the German government, does it 3 not? 4 A Yes, sir. 5 Q You are not an addressee 6 of the telex. Did you see this document at or 7 about the time it was received in Indianapolis? 8 A I think so, but there are 9 so many documents that were related that I can't 10 be sure that this specific document was one that I 11 saw, but I think so. 12 Q Well, did you see the 13 intent to reject letter? 14 A I think so. 15 Q Is this a document that 16 you've reviewed recently in preparation for your 17 deposition? 18 A No, sir, I don't think 19 so. 20 Q Neither one, neither the 21 transmittal letter or the intent to reject? 22 A I don't think so. 23 Q All right. Doctor Weber 24 says that the attached BGA response letter is 390 1 indeed an intention of rejection and not a 2 rejection itself, correct? 3 A Yes, sir. 4 Q And that Lilly has three 5 months in which to respond and that the contents 6 of the letter mainly consist of -- which are 7 objections that had been voiced previously, 8 correct? 9 A Yes, sir. 10 Q Objections voiced by 11 members of the BGA previously or members of the 12 commission previously? 13 A Yes, sir. 14 Q Had you heard these 15 objections concerning Prozac from anybody else 16 prior to this, Doctor Thompson? 17 A You mean other than from 18 Germany? 19 Q Yes. 20 A I think the 21 phospholipidosis issue in fact was one that had 22 been raised by other people, and we certainly had 23 been talking about it internally. 24 Q All right. 391 1 A In terms of the others, 2 again, I'm not a hundred percent sure what came 3 from what affiliate because I wasn't 4 distinguishing at the time, we were just trying to 5 answer a whole bunch of questions. 6 Q All right, the last 7 paragraph of the first page says the letter from 8 the BGA is nearly identical with the opinions of 9 Professor Herrmann, who -- whom Doctor H.J. Weber 10 and S. Heymanns visited a day before, correct? 11 A Yes, sir. 12 Q Do you know who Professor 13 Herrmann is? 14 A I think Professor 15 Herrmann may be one of our experts, but I don't 16 think I know any more about him or her. 17 Q When you say one of our 18 experts, you're speaking of a Lilly expert? 19 A Yes, sir. 20 Q And do you recall what he 21 was an expert in -- 22 A No, sir. 23 Q -- or she? 24 A No, sir. 392 1 Q Doctor Herrmann, in his 2 opinion, according to Doctor Weber, had suggested 3 that suicide was a problem, had he not? 4 A Yes, sir, he says 5 suicides because the suspicion is sufficient for 6 refusal, and then it goes on. 7 Q And that that was a 8 problem that exists? 9 A Yes, sir. 10 Q The document itself is 11 dated February 26th, 1985, is it not? 12 A 26th or 27th? 13 Q The actual intent to 14 reject is dated -- 15 A Oh, I'm sorry. 16 Q -- February 26th, 1985, 17 is it not? 18 A Yes, sir. 19 Q And is signed by Doctor 20 U. Schmidt, correct? 21 A Yes, sir. 22 Q Do you know Doctor 23 Schmidt? 24 A Not that I'm aware of. 393 1 Q Do you know who he is? 2 A No, sir. U is more 3 commonly a female first name in Germany. 4 Q All right. Do you know 5 who she is? 6 A No, sir. 7 Q Have you made any -- 8 attempted to make any contact with Doctor U. 9 Schmidt at any time? 10 A No, sir. 11 Q Have you attempted to 12 find out in fact who Doctor U. Schmidt is, what 13 his or her position is with the German government? 14 A No, sir. 15 Q The document is 16 concerning the application for registration of 17 Fluctin, formerly fluoxetine, Lilly, which is, we 18 know, Prozac? 19 A Yes, sir. 20 Q It says we intend to 21 review the registration of the above-mentioned 22 drug for the following reasons, correct? 23 A Yes, sir. 24 Q The BGA states, Point 1, 394 1 the drugs concerned are not sufficiently tested 2 according to the secured state of scientific 3 knowledge and the therapeutic efficacy, which is 4 claimed for them, is insufficiently substantiated, 5 end quote. And then it gives a -- apparently a 6 reference to a particular regulation or something 7 of that nature, correct? 8 A Yes, sir. 9 Q Up until this time, 10 Doctor Thompson, had any governmental agency 11 advised Eli Lilly and Company that Prozac had not 12 been sufficiently tested according to the secured 13 state of scientific knowledge? 14 A I don't think so. 15 Q Up until this time, had 16 any regulatory body advised anyone from Lilly that 17 the data indicated that the therapeutic efficacy 18 which was claimed was insufficiently 19 substantiated? 20 A I don't think so, but I'm 21 not a hundred percent sure. 22 Q Item 1.1 under the 23 scientific insufficiencies indicates that there 24 was too short a washout period, concomitant 395 1 treatment with other psychotropic drugs, and the 2 choice of control drug, correct? 3 A Yes, sir. 4 Q Point 2 of the intent to 5 reject letter states that for the drugs concerned 6 there is according to their specific profile of 7 adverse events the justified suspicion that there 8 have been unacceptable damaging effects, end 9 quote, correct? 10 A Close. 11 Q And it quotes a specific 12 regulation or something of that sort, does it not? 13 A Yes, sir. 14 Q Had anybody, any 15 scientist at Lilly, up to this time, ever 16 suggested that the adverse effects of Prozac have 17 an unacceptable damaging effects? 18 A No, sir. 19 Q Had Lilly done anything 20 to determine whether or not the side effects, 21 adverse effects of Prozac had an unacceptable 22 damaging effect? 23 A Oh, absolutely. 24 Q Item 2.1 says the use of 396 1 the preparation seems objectionable as the 2 increase in agitating effect occurs earlier than 3 the mood elevating effect and therefore an 4 increased risk of suicide exists, end quote, 5 correct? 6 A Yes, sir. 7 Q Did I read that 8 correctly? 9 A Yes, sir. 10 Q Were any of the 11 scientists at Eli Lilly and Company aware in 12 February 1985 that Prozac possessed an agitating 13 effect? 14 A No, we didn't believe 15 that that statement is scientifically correct. 16 Q Did you have any data 17 from which scientists could conclude that Prozac 18 had an agitating effect? 19 A Oh, yes, sir, we had 20 observations on a large number of patients week by 21 week as they were treated. 22 Q All right. Did you have 23 data that indicated that the agitating effect with 24 Prozac occurred earlier than the elevated effect 397 1 of Prozac? 2 A I'm not sure that's a 3 correct scientific assertion, so when you ask me 4 whether we had data that would affirm that, I 5 would have to answer that I don't think that's a 6 correct scientific interpretation of our data. As 7 I've already testified, we had a lot of data on 8 the specific measurements and the general 9 observation of patients on a week-by-week basis as 10 they were undergoing therapy, and as you know, the 11 drug has a beneficial effect which can be measured 12 at one week, but to distinguish the drug from -- 13 Q I don't know that, sir. 14 A Well, patients in 15 general, on average, get better in the first week 16 of participation in the protocols, but to 17 distinguish the drug in a statistically 18 significant fashion from placebo requires about 19 three weeks of therapy, and we've done intense 20 studies of exactly what the time course was of 21 both adverse events and efficacy. 22 Q So you just disagree with 23 the conclusion that Prozac's agitating effect 24 occurs earlier that the mood elevating effect? 398 1 A No, I disagree with the 2 conclusion that Prozac has an agitating effect, 3 period. 4 Q Okay. Therefore you're 5 certainly going to disagree of any conclusion that 6 the agitating effect occurs earlier than the mood 7 elevating effect? 8 A Yes, sir. 9 Q Now, did the German 10 government have some data about Prozac that Lilly 11 didn't have at the time they authored this intent 12 to reject? 13 A I'm sure they didn't have 14 any data on Prozac other than the data that Lilly 15 had presented, but of course they had seen data on 16 other serotonin selective antidepressants as well 17 as other antidepressants. 18 Q Well, do you see any 19 indication here that the data that the BGA is 20 basing their intent to reject on was any data in 21 connection with studies done on other specific 22 serotonin reuptake inhibitors? 23 A Well, there is, in the 24 first page, the statement at the bottom two lines 399 1 in parentheses that says suicides, because a 2 suspicion is sufficient for refusal, and then in 3 parentheses, his comment, which I think is 4 Professor Herrmann's comment, Duphar, which is a 5 European drug company, could show that there were 6 the same or even more attempted suicides in the 7 control group. Now, I don't know exactly what 8 that statement means, but it certainly indicates 9 that the people in Germany were not looking just 10 at data on Prozac, but whatever Duphar was doing. 11 It's my understanding that Duphar was studying a 12 different drug at that time, not Prozac. 13 Q Well, but what you're 14 reading now is the subject authored by Doctor 15 Weber, correct? 16 A Yes, sir. 17 Q The cover letter that 18 accompanies the intent to reject letter? 19 A Yes, sir. 20 Q But the intent to reject 21 letter doesn't say anything about reviewing data 22 other than data supplied to Lilly on Prozac? 23 A You're correct. 24 Q To the FDA -- I mean to 400 1 the BGA by Lilly concerning Prozac? 2 A You are correct in the 3 content of this two-page statement, but I also 4 have knowledge that regulators generally use data 5 that come from other trials of related drugs. 6 Q Well, did Lilly submit to 7 the BGA data concerning other drugs? 8 A Yes, sir. 9 Q That were being 10 investigated? 11 A Well, we used at least 12 three tricyclics as comparators, for example. 13 Q I thought you weren't 14 familiar with the data that the BGA was reviewing; 15 I thought you -- didn't you say that earlier? 16 A I do not have specific 17 knowledge of exactly what was submitted. However, 18 earlier it talked about the fact that there were 19 twenty-five completed trials out of forty-seven 20 trials, and you asserted that in fact at least 21 some of those were in the NDA, and I do know 22 what's in the US NDA. 23 Q Well, now can we say that 24 the BGA and the United States Food and Drug 401 1 Administration are looking at the same data? 2 A I have no idea, sir. 3 Q Well, did you recall 4 whether or not anybody at Lilly asked the BGA 5 whether or not they were basing their intent to 6 reject on other data that they -- on other 7 specific serotonin reuptake inhibitors that were 8 being investigated at the time? 9 A No. 10 Q Wouldn't it be reasonable 11 to assume, Doctor Thompson, that the BGA is 12 looking at the Prozac clinical trial data that was 13 submitted to them by Eli Lilly and Company? 14 A Well, certainly they were 15 doing that, sir, but your question encompassed the 16 possibility of them using data that had not been 17 submitted by Lilly, and I was just clarifying that 18 my answer took into account that as a possibility. 19 Q But it's not a 20 probability, is it? 21 A Oh, I am not sure. I 22 think it is probable that regulators like the BGA 23 are sophisticated, in fact use a totality of data 24 that they're seeing across whole therapeutic 402 1 domains. 2 Q Okay, so you think that 3 the BGA regulators are indeed sophisticated? 4 A Yes, sir. 5 Q And they were 6 sophisticated in making this intent to reject of 7 Prozac? 8 A The BGA and the 9 Commission A, I think they are bright people. 10 Q And you don't have -- 11 they did say that the use of the preparation seems 12 objectionable as the increase in agitating effect 13 occurs earlier than the mood elevating effect and 14 therefore an increased risk of suicide exists, did 15 they not? 16 A Yes, sir, that's what it 17 says. 18 Q And you've now 19 characterized them as sophisticated? 20 A Yes, sir. 21 Q The intent to reject 22 letter in Point 2.2 of Page 2 also says during 23 treatment with the drug, some symptoms of the 24 underlying disease, paren, anxiety, insomnia, 403 1 agitation, close paren, increase, which as adverse 2 effects exceed those which are considered 3 acceptable by medical standards, period, end 4 quote, correct? 5 A That's what it says. 6 Q Did I read that 7 accurately? 8 A Yes, sir. 9 Q Did you do anything, 10 Doctor Thompson, in connection with the specific 11 intent to reject letter? 12 A I can't recall any 13 specific action that I took. 14 Q It appears, though, that 15 several things were going to be done by the Lilly 16 affiliates, does it not? 17 A Yes, sir, and by the team 18 that I was a part of; I mean, it was our job to do 19 the primary scientific analyses on the data that 20 we had. So, I'm sure the team was working, but I 21 don't recall that I specifically had an action 22 relative to this document. 23 MR. SMITH: All right. 24 (SHORT BREAK TAKEN.) 404 1 (THOMPSON EXHIBIT NO. 17 MARKED FOR 2 IDENTIFICATION.) 3 Q (BY MR. SMITH) I'll hand 4 you Exhibit 17 and ask you to review that briefly 5 or at length. 6 MR. BOUR: Start of Tape 7 5, the time is 11:25. 8 A Yes, sir. 9 Q Exhibit 17 is a document 10 that is authored by Doctor Johanna Schenk, is it 11 not? 12 A Yes, sir. 13 Q And it concerns Prozac, 14 fluoxetine, does it not? 15 A Yes, sir. 16 Q It appears it bears three 17 dates. On the top of the page it says April 3rd, 18 1985, Lilly Bad Homburg, does it not? 19 A Yes, sir. 20 Q Then there's a date in 21 the middle of the page that says April 2nd, 1985, 22 correct? 23 A Yes, sir. 24 Q And below that there's a 405 1 statement, Report on Fluoxetine Working Session of 2 April 29th and 30th, 1985, correct? 3 A Yes, sir, and my copy 4 also has a date stamp across it. 5 Q Beg pardon? 6 A My copy has a date stamp 7 across it that says April 05 of 1985. 8 MR. FREEMAN: Yours 9 doesn't have that, Paul, apparently. 10 Q The document that you're 11 looking at is a document that was supplied to us 12 within the last five days by Lilly, and now this 13 document we see a stamp on it April 5th, 1985, 14 correct? 15 A Yes, sir. 16 Q Your copy has a date of 17 April 5th, 1985? 18 A Yes, sir. 19 Q And this document has 20 nothing redacted from it, correct? 21 A Redacted from it? 22 Q Redacted? 23 MS. ZETTLER: Blacked 24 out. 406 1 Q Blacked out? 2 A Oh. 3 Q We've seen some documents 4 that are blacked out? 5 A No, sir, there is nothing 6 blacked out. 7 Q Do you know who the Wally 8 is that's mentioned on this document? 9 A It's probably Mr. Lange. 10 Q Who is Wally Lange? 11 A At that time, I think he 12 was responsible for either US marketing or 13 worldwide marketing, I'm not sure. 14 Q Is he still with Lilly? 15 A No, sir. 16 Q Where is he now? 17 A He retired. 18 Q Do you know why this 19 document would have four different dates on it? 20 A No, sir. 21 Q Is that usual or unusual 22 to see in connection with documents that you see 23 at Lilly? 24 A In all honesty, I don't 407 1 pay a lot of attention to the dates, I look at the 2 content, and so I wouldn't really want to comment 3 on that. 4 Q It's curious to me how 5 you can have a report on a working -- a report on 6 fluoxetine working session of April 19th and 30th, 7 1985 -- 8 A 29th and 30th. 9 Q 29th and 30th, and then 10 have three other preceding dates stamped on the 11 document itself. 12 A I thought that was 13 interesting, too. 14 Q All right. But you have 15 no explanation for it? 16 A No, sir, I'm afraid not. 17 Q Apparently the objective 18 of this meeting, reflected by the document, is to 19 make Professor Herrmann and his co-worker familiar 20 with fluoxetine data so that he is in a better 21 position to give best advice as a consultant to 22 the company in the registration process of 23 fluoxetine, correct? 24 A Yes, sir. 408 1 Q And it appears that he or 2 they have reviewed, one, the original 3 documentation submitted March 1st, 1984, correct? 4 A Yes, sir. 5 Q Do you know what that 6 was? 7 A No, sir. 8 Q Do you know who would 9 know what that was? 10 A Well, I think the staff 11 of Lilly in Germany would probably know who that 12 was, or what that was; maybe Doctor Weinstein, I 13 don't know. 14 Q Lilly Indianapolis and 15 Lilly's affiliate in Germany were working hand in 16 hand on this issue, were they not? 17 A Yes, sir. 18 Q When you say Lilly's 19 affiliates in Germany, such as Doctor Weber -- 20 A Yes, sir. 21 Q Weber, Doctor Schenk -- 22 A Yes, sir. 23 Q -- those were employees 24 of Lilly, were they not? 409 1 A Yes, sir. 2 Q And they were -- their 3 conduct and actions were directed by supervisors 4 in Germany? 5 A Yes, sir. 6 Q That was Doctor Weber, 7 correct? 8 A Yes, sir. 9 Q And then Doctor Weber 10 reported to individuals in the United States and 11 Indianapolis? 12 A No, sir. 13 Q Who did he report to, 14 Perelman? 15 A No, sir, he reported to 16 the general manager of the German affiliate. 17 Q All right, and who was 18 that at the time? 19 A I have no idea. 20 Q And at that time, the 21 general manager of the German affiliate -- that 22 was Claude Bouchy, wasn't it? 23 A I'm sorry, I wouldn't 24 know. 410 1 Q At that time, the general 2 manager of the German affiliate reported to who in 3 Indianapolis? 4 A I'm sorry, I really don't 5 know, the marketing and sales organization, 6 international; eventually at that time, he would 7 have reported up through Doctor Perelman, who I 8 believe at that time was president of 9 international, but I'm not sure of the chain of 10 command. 11 Q But Lilly in Indianapolis 12 was working with Lilly in Germany? 13 A Yes, sir. 14 Q To accomplish Lilly's 15 goals? 16 A Yes, sir. 17 Q Throughout the United 18 States? 19 A Throughout the world. 20 Q I mean the world. It 21 says additionally that the data reviewed was 22 analysis of pooled studies, fluoxetine versus 23 imipramine versus placebo, Protocol No. 27, 24 submitted October 26th, 1984, correct? 411 1 A Yes, sir. 2 Q Are you familiar with 3 Protocol No. 27? 4 A Reasonably, yes, sir. 5 Q Are you familiar that 6 that is a comparator study done at multiple sites 7 comparing Prozac, imipramine and placebo? 8 A Yes, sir. 9 Q Are you aware that that 10 was a study used by the United States Food and 11 Drug Administration to base, in part, their 12 approval of Prozac in the United States? 13 A Yes, sir. 14 Q Otherwise known as a, 15 quote, pivotal study, end quote? 16 A Yes, sir. 17 Q It says with respect to 18 outcome of that meeting, Professor Herrmann left 19 an opinion of twenty-one typewritten pages, the 20 essential points are summarized as follows, 21 correct? 22 A Yes, sir. 23 Q And then the next two and 24 a half pages, down to our Bad Homburg 412 1 recommendation would be the following now, appear 2 to be a summary of that twenty-one-page opinion, 3 do they not? 4 A You mean the next page 5 and a half, because I only have two pages after 6 that first page. 7 Q Yes, I do. 8 A Yes, sir. 9 Q The point is that what we 10 see up to the last page, under our Bad Homburg 11 recommendation, is a summary of Doctor Herrmann's 12 report? 13 A That's the way I 14 interpret it. 15 Q He speaks concerning 16 safety, does he not? 17 A Yes, sir. 18 Q It says -- well, the 19 summary says that he says: Still not resolved is 20 the fact that suicide attempts have been observed 21 more frequently on fluoxetine as compared to 22 imipramine, paren, only epidemiologic data or 23 literature on other antidepressants may help to 24 identifying whether it happened by chance that 413 1 incidence of suicide attempts was abnormally high 2 on fluoxetine or abnormally low under 3 comparator -- under comparators, close paren, 4 period, end quote. Correct? 5 A Yes, sir. 6 Q Did I read that 7 correctly? 8 A Yes, sir. 9 Q Had you seen any data up 10 to that time, Doctor Thompson, that indicated a 11 larger number of suicide attempts on fluoxetine, 12 Prozac, as opposed to comparator drugs? 13 A In terms of a total 14 numerical count, yes; in terms of a true 15 epidemiologic incidence rate, no. 16 Q All right. Doctor 17 Herrmann apparently -- or Doctor Schenk also 18 characterizes Doctor Herrmann's report as saying 19 except nausea, fluoxetine side-effect spectrum is 20 unequivocally more favorable than that of 21 imipramine, but according to today's knowledge, 22 this is negatively affected by the increased 23 suicidal risk, correct? 24 A Yes, sir, that's what she 414 1 says. 2 Q Then the summary goes on 3 to say that the benefit/risk ratio, according to 4 Doctor Herrmann, was not unequivocally positive, 5 correct? 6 A Yes, sir. 7 Q And that the probable of 8 suggest -- the probability of success of getting 9 Prozac registered in Germany was not high? 10 A Yes, sir. 11 Q Have you ever seen this 12 document before? 13 A Yes, sir, I think so. 14 Q Did you see it at or 15 about the time it was authored? 16 A I think so. 17 Q You recall -- you think 18 you recall receiving it? 19 A Well, I'm not an 20 addressee on it, but I think I saw it or at least 21 heard a summary of it at around that time. 22 Q Is this a document you 23 reviewed in preparation for your deposition? 24 A I can't remember 415 1 specifically; it may have been. 2 Q It probably was, wasn't 3 it? 4 A Well, it's certainly 5 possible, and I would even say probable. 6 Q Have you seen the 7 twenty-one-page typewritten report of Doctor 8 Herrmann? 9 A I don't think so. 10 Q Do you know of anybody 11 who has seen that report? 12 A I'm sorry, I don't know. 13 I assume it's in German, and I told you I don't 14 read German very well at all. 15 Q Well, we have other 16 evidence that there were expert opinions being 17 rendered in Germany that were translated into 18 English and just flying to Indianapolis; will you 19 take my word for that? 20 A Yes, sir. 21 Q So I thought maybe this 22 was an opinion that might have been translated and 23 submitted to you scientists in Indianapolis. Have 24 you ever seen such a translation of Doctor 416 1 Herrmann's opinion? 2 A I don't think so. 3 Q Did you ask for a 4 translation of Doctor Herrmann's opinion when you 5 saw this report -- when you saw this memo? 6 A I don't recall doing so. 7 Q In looking at this memo 8 now, wouldn't you think it would be significant to 9 see the twenty-one-page typewritten report? 10 A You mean for me 11 personally to see it? 12 Q Yes. 13 A No. I mean, there were 14 very excellent experts, both in Germany and in the 15 United States, working on this problem. My job 16 really was to make sure that the issues were being 17 addressed, not to do an individual assessment of 18 an expert's report. 19 Q You don't have any reason 20 to believe that Doctor Herrmann wasn't an 21 excellent expert, do you? 22 A No, sir. 23 Q I mean, he was a 24 consultant hired by Eli Lilly and Company, was he 417 1 not? 2 A I think so. 3 Q And he's indicated that 4 there's a high number of Prozac individuals 5 attempting suicide versus a lower number of 6 individuals on the comparator drug attempting 7 suicide, was he not? 8 A That's what it says he 9 asserted, so he made a mistake; we all make 10 mistakes. 11 Q Who made a mistake? 12 A Well, if Doctor Herrmann 13 asserted that there was a epidemiologically 14 significant difference in the incidence rates, he 15 made a mistake; if he asserted that there was a 16 numerical difference in the number of reports, he 17 was correct. 18 Q You're going to make a 19 conclusion without seeing his report whether or 20 not he was correct or mistaken, is that correct? 21 A I can only work from what 22 you've given me, which says -- 23 Q That's all I can work on, 24 too, Doctor. That's all Lilly has given me; did 418 1 you know that? 2 A No, sir, I have no idea 3 what Lilly has given you except they collected an 4 awfully lot of my documents. 5 Q They've submitted an 6 awfully lot of documents, but for some reason they 7 have not been able to locate this twenty-one 8 typewritten page report, to date. 9 MR. MYERS: That's not 10 correct. 11 MS. ZETTLER: Are you 12 saying that you have? 13 MR. MYERS: Look in the 14 documents that you were sent in response to your 15 latest set of request for production of documents, 16 and there are numerous expert reports in there. 17 MR. SMITH: Not Doctor 18 Herrmann's. 19 MS. ZETTLER: Not Doctor 20 Herrmann's twenty-one-page report, Larry. 21 MR. MYERS: There is a 22 translation of a report by Doctor Herrmann in that 23 package of documents, if you'll simply examine 24 them. 419 1 MS. ZETTLER: I've looked 2 through them, I've not found anything, translated 3 or not translated. 4 MR. MYERS: Look a little 5 harder. 6 MS. ZETTLER: No, you're 7 wrong, Larry, there is nothing in that -- 8 MR. BRENNAN: Doctor 9 Herrmann's report is there, Ms. Zettler. 10 MS. ZETTLER: Trust me, 11 if it had been there, I'd have it with me; it's 12 not there. In fact, your own response to our 13 request for production states that as of the date 14 that the response was filed you had not found it. 15 So are you changing your response on this now? 16 MR. MYERS: No, ma'am, 17 I'm telling you to look at the documents that were 18 sent directly to your office -- 19 MS. ZETTLER: When? 20 MR. MYERS: On the same 21 day you received the responses to the request for 22 production. 23 MS. ZETTLER: It's not in 24 there. 420 1 MR. MYERS: I'll just ask 2 you to go back and look again. There were 3 numerous reports from experts in those documents. 4 MS. ZETTLER: I've got 5 them here. 6 MR. SMITH: Is it your 7 testimony -- is it your statement here, Mr. Myers, 8 that the twenty-one-page typewritten opinion that 9 is referred to in Thompson Exhibit 17 has indeed 10 been produced to the plaintiffs in this case? 11 MR. MYERS: My 12 understanding is that a translation of that report 13 has been submitted; it may or may not be 14 twenty-one pages in length. 15 MS. ZETTLER: A 16 translation. 17 MR. SMITH: When was it 18 submitted? 19 MR. MYERS: It was 20 submitted in response to your June request for 21 production of documents, the responses to which 22 were served earlier this week. 23 MR. SMITH: Is it the 24 same opinion that is referenced? 421 1 MR. MYERS: I believe 2 that to be the case. 3 MR. SMITH: Does it 4 reflect by its -- 5 MR. MYERS: I'm not under 6 examination. 7 MR. SMITH: I know that. 8 MR. MYERS: Let's go off 9 the record. 10 (OFF-THE-RECORD DISCUSSION HELD.) 11 12 Q (BY MR. SMITH) As I 13 understand it, you never asked to see the 14 twenty-one-page typewritten report? 15 A I don't think so, sir. 16 Q Or have it translated? 17 A I don't think so. 18 Q Do you recall discussing 19 that report with any of the scientists at Lilly in 20 Indianapolis? 21 A I think so, yes. 22 Q Who would you have 23 discussed it with? 24 A Well, I would have 422 1 discussed it probably with Doctor Weinstein, as 2 well as the specific scientists working on 3 fluoxetine at that time. 4 Q Why would you have 5 discussed it with Doctor Weinstein? 6 A Because he was 7 responsible for international and he would have 8 been the primary person who had all the contact 9 with the German affiliate. 10 Q Well, would -- would you 11 have discussed specifically the opinion itself? 12 A Well, again, I don't have 13 a recollection of the content of that document, 14 but as I indicated, I think I have seen this, and 15 certainly I was aware that there were a continuing 16 series of questions back and forth from the German 17 affiliate and the BGA that we were trying to 18 respond to by doing scientific analyses. 19 Q E under Prerequisites for 20 Successful Outcome, do you see that? 21 A Yes, sir. 22 Q It says that limited 23 indication, i.e. mild to moderate endogenous 24 depression plus precautionary statement concerning 423 1 suicidal risk, correct? 2 A Yes, sir. 3 Q Then No. 1, plausible 4 explanation for incidence of suicides attempts, 5 correct? 6 A Yes, sir. 7 Q Do you take it that 8 Professor Herrmann is expressing the opinion that 9 if you have a plausible explanation for incidence 10 of suicides and attempts that there's a 11 possibility of a precautionary statement 12 concerning suicidal risk to be contained in the 13 either German package insert or prescribing 14 information for physicians there? 15 A I wouldn't necessary 16 connect those two logically. I think what this 17 says is that he's asserting that we may end up 18 with a precautionary statement, and three points 19 above that, one of them is it asks for a plausible 20 explanation incidence of suicides attempts, but 21 I'm not sure the two necessarily are connected. 22 Q All right. Of course, 23 there was a precautionary statement concerning 24 suicidal risk that did -- in fact had to be placed 424 1 in the German package insert, did there not? 2 A Is it called a 3 precautions in Germany like it is in the FDA 4 label? 5 Q Well, there are 6 statements in Germany concerning suicidal risk 7 that are not present in the United States label, 8 correct? 9 A At all times? I mean, 10 I'm sorry, I don't -- 11 Q Now. 12 A I haven't looked at 13 either label, either Germany or the US, recently. 14 Q Why? 15 A Why have I not? 16 Q Yes. 17 A I haven't had any 18 occasion to. 19 Q Well, aren't you 20 interested in the instructions that physicians are 21 being given concerning how to use the product -- 22 A Yes, sir. 23 Q -- which you're 24 manufacturing? 425 1 A Yes, sir. 2 Q When did you last look 3 at -- or have you ever looked at the package 4 insert or prescribing information for Fluctin, 5 Prozac in Germany? 6 A I don't recall ever 7 seeing the actual German label, and it's my 8 understanding many European countries there 9 actually is both a lay person and a professional 10 label, although I'm not sure whether that's true 11 in Germany or not. 12 Q I think that is true in 13 Germany. 14 A But I've seen -- I've 15 seen translations of it, but the exact timing of 16 when those were either proposed or approved or 17 used I wouldn't want to testify to. 18 Q You can say that at 19 sometimes, at least, or the last time you compared 20 the two, the statements concerning suicidal risk 21 were different in the German information as 22 opposed to the United States information? 23 A I think that's true. 24 Q All right. 426 1 MR. SMITH: Let's stop 2 and eat. 3 (LUNCH BREAK TAKEN.) 4 MR. BOUR: The time is 5 1:13. 6 Q (BY MR. SMITH) Doctor 7 Thompson, did you ever go to Germany yourself in 8 connection with these problems raised by the BGA 9 with respect to registering Prozac in Germany? 10 A No, sir. 11 Q Did you ever direct that 12 the individuals there at the Indianapolis office 13 go to Germany for that purpose? 14 A I'm not sure about that, 15 because people that reported to me certainly were 16 involved in the analysis, and may well have gone 17 to Germany to help the Germany affiliate, but I 18 can't remember specific details. 19 Q Do you know whether or 20 not the United States Food and Drug Administration 21 was made aware of Professor Herrmann's opinion 22 concerning the high incidence of suicide attempts 23 on Prozac as compared to imipramine? 24 A I'm sorry, I don't have 427 1 any knowledge of that specifically. 2 Q Would this be something 3 that you'd think the United States Food and Drug 4 Administration should have? 5 A Well, if Professor 6 Herrmann was in fact a Lilly consultant, and if in 7 fact we scientifically disagreed with that 8 conclusion, I don't think that there's a 9 regulatory requirement for providing his 10 conclusion. I think there's a regulatory 11 requirement to provide facts, but not necessarily 12 conclusions that we disagree with. 13 Q Do you know Doctor 14 Dorothy Dobbs? 15 A Oh, yes, sir. 16 Q Did you know that she has 17 testified that in her opinion such an opinion 18 should be forwarded to the United States Food and 19 Drug Administration? 20 A No, sir, I didn't know 21 that. 22 Q And she was a regulatory 23 scientist at Lilly? 24 A Yes, sir, an excellent 428 1 psychiatrist, very good. 2 Q Did you ever make any 3 effort to withhold any information concerning 4 Prozac or what was happening in Prozac from Doctor 5 Dobbs? 6 A Absolutely not. 7 Q Did you know that Doctor 8 Dobbs has testified under oath in these cases that 9 she had a feeling that she was taken out of the 10 loop in connection with information on Prozac? 11 A No, sir, I have no idea 12 what she testified to. 13 Q Did she ever make any 14 complaint to you of such a feeling? 15 A No, sir. 16 Q Did you -- were you aware 17 when Doctor Dobbs left Eli Lilly's employment? 18 A I couldn't give you the 19 exact date, but surely -- I think at that time I 20 may have been responsible for her; I wasn't when I 21 first came to Lilly. 22 Q Did you ask -- did Doctor 23 Dobbs discuss with you her resignation -- 24 A I can't remember any 429 1 details of that. 2 Q -- from Lilly? 3 A I was very sorry to see 4 her go. 5 Q Did she tell you that she 6 had been advised by Lilly that she was not going 7 to be promoted in the manner that she wanted to be 8 promoted? 9 A I don't remember that 10 specifically. 11 Q Did you ever have any 12 criticism of Doctor Dobbs' work? 13 A You know, certainly not 14 any of her scientific work; I thought she was a 15 very bright psychiatrist and excellent with 16 regulatory. She was primarily responsible for 17 putting the NDA together. I think that she had 18 some difficulties dealing with other people in 19 terms of her personality, but absolutely no 20 criticism of her scientific acumen or integrity or 21 thoroughness at all. 22 Q Did you or she have any 23 difficulty with your personalities at any time? 24 A I don't think so. 430 1 Q Did you ever discuss with 2 any individual at the Food and Drug Administration 3 the fact that Germany was viewing suicide risk and 4 activation as a problem with Prozac? 5 A I don't recall any time 6 that I had that conversation with anybody at the 7 FDA. 8 Q So as far as you know, 9 you don't have any information concerning what 10 information might have been given to the United 11 States Food and Drug Administration concerning 12 Germany's position in connection with problems 13 that Prozac might be causing individuals with 14 respect to suicide and violent aggressive 15 behavior? 16 A No, I don't think that 17 statement is correct. 18 Q All right. Then what did 19 you do? 20 A Well, specifically in 21 regard to the safety update to the New Drug 22 Application of fluoxetine, I was part of a team 23 working very hard to perform extensive analyses; 24 in fact, as I recall, that document is bigger than 431 1 the original NDA, across all of our data, across 2 all of the safety issues, because I recall we 3 analyzed each safety issue by, as we were talking 4 about earlier, age, dose, duration, when it 5 occurred, when it went away, et cetera, and that 6 was one of the issues that I recall we 7 specifically worked on. 8 Q In connection with the 9 safety update, you say that was a pretty 10 comprehensive document? 11 A Yes, sir. 12 Q Who is Doctor Marlene L. 13 Cohen? 14 A Doctor Cohen is one of 15 our senior scientists in the discovery program at 16 Lilly. 17 Q Do you respect her 18 judgment as a scientist? 19 A Absolutely. 20 Q Did you ask her to do 21 some work in connection with the safety update? 22 A I think we called on 23 her -- I don't think I personally did, but again 24 people I was responsible for certainly would have 432 1 asked her to -- and a whole variety of other 2 experts within Lilly, to look at it. 3 (THOMPSON EXHIBIT NO. 18 MARKED FOR 4 IDENTIFICATION.) 5 THE WITNESS: Yes, sir. 6 Q (BY MR. SMITH) Look at 7 Point 8 -- well, let's identify the document 8 first. It appears that Exhibit 18 is a review of 9 this fluoxetine safety update that you've spoken 10 of? 11 A Yes, sir. 12 Q Done by Doctor Cohen? 13 A Yes, sir. 14 Q And apparently she's 15 taking the update itself, looking at it, and has 16 made notes concerning her review of the safety 17 update? 18 A That's probably one of 19 the drafts of the update rather than the final 20 version, because we'd normally circulate that for 21 scientific critique. 22 Q And she made general 23 comments on the first two pages? 24 A Yes, sir. 433 1 Q And specific comments 2 concerning the document on the last three pages, 3 correct? 4 A Yes, sir. 5 Q In Point 8 of the general 6 comments on Page 2, she states, quote: CNS 7 stimulation, agitation, and insomnia. From the 8 data taken in concert it might appear that there 9 is a somewhat greater incidence of CNS stimulation 10 with fluoxetine as compared to placebo. For 11 example, a greater incidence of insomnia and 12 anxiety was reported, paren, 9-95, 9-113, 9-370, 13 agitation seems pronounced, 9-160, and a decrease 14 in REM sleep was reported, and there may be a 15 suggestion of manic psychosis, 9-107, 9-111. As 16 with other antidepressant agents used in depressed 17 population, mania or CNS stimulation may be a 18 minor component of the actions of fluoxetine, end 19 quote. Correct? 20 A Yes, sir. 21 Q Had you seen this 22 document before? 23 A Yes, sir. 24 Q Is this one of the 434 1 documents that you reviewed in connection with 2 your giving your deposition here today? 3 A Yes, sir. 4 Q Do you know what the 5 9-95, 9-113 and 9-370 mean here? 6 A I think it's Volume 9, 7 and those, I think, are page numbers. 8 Q Volume 9 of the safety 9 update? 10 A Yes, sir. 11 Q And so she's making a 12 reference to specific pages where she's noting 13 these -- this data, is that correct? 14 A I'm assuming that that's 15 what those references are to. 16 Q Have you discussed this 17 with Doctor Cohen? 18 A No, sir. 19 Q Did you see this document 20 before it was presented to you for review in 21 connection with your deposition here? 22 A I think so, because I 23 went over the safety update and our critique of it 24 several times, so I probably saw this, but I don't 435 1 have a specific recollection of it. 2 Q Did the safety update 3 conclude that CNS stimulation, agitation, and 4 insomnia from the data taken in concert it might 5 appear that there is a somewhat greater incidence 6 of CNS stimulation with fluoxetine as compared to 7 placebo? 8 A I don't remember for 9 sure, but I don't think so. 10 Q Did the safety data -- I 11 mean did the safety update that was submitted to 12 the United States Food and Drug Administration 13 indicate that from the data there was a greater 14 incidence of insomnia and anxiety? 15 A I don't remember; it had 16 enormous data tables in there about all of the 17 safety elements, but I don't recall the specifics 18 of it. It was a huge document and it was a long 19 time ago. 20 Q Did that document -- was 21 there a conclusion or a statement made in the 22 safety update in any manner that there was a 23 greater incidence of insomnia and anxiety reported 24 with Prozac? 436 1 A I don't know. 2 Q Did the safety update 3 that was submitted to the United States Food and 4 Drug Administration reflect the fact that 5 agitation was pronounced on patients with 6 fluoxetine? 7 A I don't know. 8 Q Did the safety update 9 submitted to the Food and Drug Administration 10 indicate that there was reports of decrease in REM 11 sleep? 12 A I don't know. 13 Q Did the safety update 14 indicate that there was a suggestion of manic 15 psychosis in connection with use of Prozac? 16 A I don't know. 17 Q Did the safety indicate 18 that CNS stimulation may be a minor component of 19 the action of fluoxetine? 20 A I don't know. 21 Q Point 7 indicates that 22 she was reviewing the issue of temperature 23 changes, that is sweating, correct? 24 A Yes, sir. 437 1 Q It says from the sum 2 total of the data presented it might appear that 3 fluoxetine exerts some effect on the hypothalamic 4 temperature regulating mechanisms, correct? 5 A Yes, sir. 6 Q Do you have an opinion 7 concerning whether or not Prozac does exert an 8 effect on the hypothalamic temperature regulating 9 mechanism of humans? 10 A As best I recall the 11 data, there were no significant changes in body 12 temperature recorded in the clinical trials, and 13 so I don't believe that there's any evidence that 14 it has any effect on the temperature regulating 15 mechanisms, per se, at least as manifested by body 16 temperature. It is correct that sweating was 17 reported as an adverse event. 18 Q Well, Doctor Cohen 19 indicates at Point 7, quote, excessive sweating 20 was the most frequently reported adverse event on 21 skin and appendages, does she not? 22 A I think that's what it 23 says. 24 Q And she provides a 438 1 citation to the volume and page where she found 2 that? 3 A Yes, sir. 4 Q And in fact she has a 5 volume and page to indicate that the -- quote, the 6 change in temperature was reported to be greater 7 than the change occurring with placebo, does she 8 not? 9 A That's what she says. 10 Q Do you just disagree with 11 her, Doctor Thompson? 12 A My recollection of the 13 data was that there was not a clinically and 14 statistically significant effect -- 15 Q Did you discuss -- I'm 16 sorry, I didn't mean to cut you off. 17 A -- on body temperature 18 with the placebo versus Prozac versus the 19 comparators. 20 Q Do you disagree that 21 sweating was reported more frequently as -- the 22 most frequently as an adverse event on the skin 23 and appendage section? 24 A No, sir, I don't disagree 439 1 with that at all. 2 Q The -- were the 3 investigators taking the patients' temperatures 4 that participated in the clinical trials? 5 A Yes, sir, I believe they 6 were. 7 Q Do you disagree with 8 Doctor Cohen where she says the change in 9 temperature was reported to be greater than the 10 change occurring with placebo? 11 A Well, I've told you what 12 my memory is of the data as a whole over a 13 twelve-year period. Doctor Cohen was obviously 14 much closer to the data that was written in 15 whatever draft it was she reviewed of the safety 16 update, and I have a great deal of respect of her 17 scientific acumen, but I've told you what my 18 memory is of the body temperature data as a whole. 19 Q So you could be wrong and 20 she could be right? 21 A Absolutely that's 22 possible. 23 Q Because obviously you 24 don't have that data sitting right in front of you 440 1 now? 2 A No, sir, I don't. 3 Q But she did then when she 4 wrote these -- made these observations? 5 A I think so. 6 Q And the same would be 7 true with Point 8, she had data before her then 8 when she made those notations and you don't have 9 that data before you now as you're giving your 10 testimony, correct? 11 A That's correct. 12 Q Do you believe that 13 Doctor Cohen was erroneous when she says from the 14 data taken in concert it might appear there is a 15 somewhat greater incidence of CNS stimulation with 16 fluoxetine as compared to placebo? 17 A That was not my 18 conclusion at the time, and it is not my 19 conclusion now. 20 Q But again, she had the 21 data before her then when she made these 22 notations, did she not? 23 A And I certainly reviewed 24 those same data at that time, but I don't have 441 1 them in front of me now. 2 Q Doctor Cohen is a 3 scientist in your discovery sciences department? 4 A Yes, sir. 5 Q What does that mean? 6 A Well, there are about a 7 thousand scientists at Lilly that work in basic 8 drug discovery, and she's a world class expert in 9 several areas in terms of CNS research, especially 10 in regard to serotonin receptors. 11 Q All right. So she's no 12 stranger to CNS stimulation? 13 A Not at all. 14 Q And she's generally 15 knowledgeable concerning these areas? 16 A Yes, sir. 17 Q As a matter of fact, it 18 sounds like she's probably dedicated a great deal 19 of time and effort in examining these issues? 20 A Well, serotonin receptors 21 for sure. 22 Q Do you disagree with her 23 where she says CNS stimulation may be a minor 24 component of the actions of fluoxetine? 442 1 A Yes, sir, I disagree with 2 her. 3 Q She could be right and 4 you could be wrong, though, correct? 5 A That's possible. 6 Q You're just stating a 7 disagreement? 8 A Yes, sir. 9 Q You're not rendering a 10 conclusion on who is right and wrong? 11 A I told you what my 12 opinion is. 13 Q In Point 6 of Doctor 14 Cohen's review of the safety update that was 15 submitted to the Food and Drug Administration, it 16 says: Drug interaction studies were performed 17 primarily after acute single dose administration 18 of fluoxetine. However, since steady-state levels 19 of fluoxetine do not occur following a single dose 20 and since plasma levels were shown to increase 21 with multiple dosing and steady-state achievement 22 after fluoxetine administration, drug interaction 23 studies after chronic administration of fluoxetine 24 when plasma levels are highest should perhaps be 443 1 considered. Correct? 2 A Yes, sir. 3 Q Were there drug 4 interaction studies after chronic administration 5 of fluoxetine done? 6 A Yes, some of them were 7 done, I think, after this time, but yes. 8 Q Had there been any before 9 Doctor Cohen recommended that? 10 A I can't recall any 11 specific ones. 12 Q Was it based on Doctor 13 Cohen's observation in Point 6 that these specific 14 studies were undertaken? 15 A Not to my best 16 recollection in the sense of this directly leading 17 to those studies, because we had been concerned 18 about how to conduct such studies for a long time 19 because it's very difficult to do that with a drug 20 with a long half time. 21 Q So -- but at least Doctor 22 Cohen and you agree on that point, that there was 23 a necessity to do studies on drug interaction 24 after chronic administration of fluoxetine? 444 1 A I think you're misstating 2 what Doctor Cohen said; I think she said should 3 perhaps be considered, and I agree with that 4 assertion, which is subjunctive, not -- I think 5 you used the word necessary. 6 Q She said you should 7 consider it, correct? 8 A Should perhaps be 9 considered. 10 Q And in fact you -- I 11 thought you testified that it had been considered? 12 A Yes, sir. 13 Q And that it was indeed 14 done, correct? 15 A Yes, sir. 16 Q Okay, so her statement 17 that it should be considered was accurate, was it 18 not? 19 A I think her statement was 20 accurate, but you didn't reflect it accurately in 21 your question. 22 Q Do you know if the BGA 23 was ever supplied with Doctor Cohen's observations 24 concerning fluoxetine as expressed in Exhibit 18? 445 1 A I'm sorry, I have no 2 idea. 3 (THOMPSON EXHIBIT NO. 19 MARKED FOR 4 IDENTIFICATION.) 5 THE WITNESS: Yes, sir. 6 Q (BY MR. SMITH) Doctor, 7 Exhibit 19 is a document that you authored dated 8 June 13th, 1988, correct? 9 A Yes, sir. 10 Q And in June 1988, Prozac 11 was manufactured in the pulvule form in only 12 twenty milligram dosage, is that correct? 13 A I think that's correct. 14 Q In other words, if my 15 physician was to write me a prescription for 16 Prozac pulvules, the lowest dosage per pulvule 17 that that prescription could be dispensed by, by a 18 pharmacist, would be twenty milligrams, is that 19 correct? 20 A Unless they repackaged 21 it. 22 Q Repackaged what? 23 A The pharmacist can 24 compound, so what many of the physicians, 446 1 patients, and I don't know from my own knowledge, 2 but probably pharmacists were taking the pulvules 3 apart and separating the granules within. 4 Q Separating the what? 5 A The material within the 6 pulvule. 7 Q Well, do you know that 8 that in fact occurred with the twenty milligram 9 pulvules? 10 A Yes, sir. I mean, that 11 was one of the points of this -- from the meeting 12 that I had gone to. 13 Q So, let me see if I 14 understand. In June 1988, the only pulvules that 15 were leaving Lilly manufacturing facilities were 16 twenty milligram pulvules, correct? 17 A Yes, sir. 18 Q In other words, Lilly was 19 not manufacturing pulvules in any lower dosage 20 than twenty milligrams, is that correct? 21 A Not for sale. 22 Q For sale to the public? 23 A Yes, sir, that's correct. 24 Q And is what you're saying 447 1 that there was -- it was known in June of 1988 2 that physicians were requesting that pharmacists 3 take those twenty milligram pulvules apart and 4 repackage them? 5 A No, I said explicitly I 6 didn't know of my own knowledge of pharmacists 7 doing that, but I knew that physicians and 8 patients were in fact using fractions of the 9 contents of pulvules on a daily basis or using 10 less than one pulvule a day. 11 Q Well, I think you 12 originally said that it could be repackaged -- 13 A Yes, sir. 14 Q -- by the pharmacist? 15 A That's correct. 16 Q And how would a 17 pharmacist repackage a twenty milligram pulvule? 18 A Simply use an empty 19 capsule and take the pulvule and empty half of it 20 into one capsule and half into another; it's done 21 all the time. 22 Q Did you know that in June 23 of 1988? 24 A No, I said specifically I 448 1 did not know of my own knowledge of pharmacists 2 doing that as opposed to physicians and patients 3 using fractional amounts of the content of the 4 pulvule. 5 Q Do you know whether or 6 not a pharmacist could do that? 7 A Yes, sir, it's my 8 understanding that a pharmacist certainly could do 9 that. 10 Q And do you know of any 11 instances in which pharmacists -- where 12 pharmacists did do that? 13 A No, sir, I said 14 explicitly with my own knowledge, I do not. 15 Q Has somebody told you 16 that that was occurring? 17 A I know that patients were 18 using smaller doses by using fractional quantities 19 of what was in the pulvule, but I don't know 20 whether the patient did that him or herself or 21 whether they had a pharmacist do it or whether a 22 physician did it. 23 Q Those little granules 24 that are in a pulvule -- 449 1 A Yes, sir. 2 Q -- are they multicolored 3 granules? 4 A I don't have any idea. 5 Q But it doesn't make any 6 difference, all you have to do is just by quantity 7 take one apart and divide it in half and you could 8 theoretically have one ten-milligram -- you would 9 have ten milligrams of fluoxetine hydrochloride on 10 one side and ten milligrams on the other side? 11 A You said it doesn't make 12 any difference, there's a real art to compounding, 13 and I'm not a pharmacist, so I wouldn't want to 14 address that. I know that patients in fact were 15 using fractional amounts of the contents of 16 pulvules either as dry powder or dissolving it in 17 a liquid like orange juice. 18 Q I had heard some of the 19 Lilly investigators in public speeches make the 20 statement that they were having patients dissolve 21 a twenty milligram pulvule by emptying its 22 contents into a glass of orange juice, stirring it 23 up, dividing that orange juice into four separate 24 sections and having therefore four five-milligram 450 1 dosage -- doses of Prozac. Specifically that was 2 Doctor Feighner -- 3 A Yes, sir. 4 Q -- that has indicated 5 he's directed his patients to do that. 6 A I don't know whether 7 Doctor Feighner -- I mean, I don't doubt your 8 assertion, but again, the purpose of my memo here 9 was to point out that other people at the 10 meeting -- I don't remember Doctor Feighner being 11 at this meeting -- were also telling me that they 12 were having patients use doses of five or ten 13 milligram a day. 14 Q And the point is, is that 15 Lilly was only manufacturing Prozac in 16 twenty-milligram pulvules? 17 A Yes, sir. 18 Q And the question was -- 19 well, the question is: In June of 1988, had the 20 lowest efficacious dose of fluoxetine been 21 established? 22 A I think we had the data 23 from the -- what we call the low dose study by 24 then. 451 1 Q All right. And what was 2 the lowest efficacious dose? 3 A Twenty milligrams. 4 Q Ten milligrams wasn't 5 efficacious? 6 A In the obesity studies, 7 the weight loss on ten milligrams was 8 statistically significantly more than on placebo, 9 but also statistically significantly less than on 10 higher doses, and clinically certainly less than 11 on higher doses. 12 Q I'm not talking about 13 obesity, I'm talking about efficacy for 14 depression. 15 A I'm not aware that we had 16 any data on ten milligrams as a dose to treat 17 depression in any control study anywhere. 18 Q So is it your testimony 19 that in June of 1988, the lowest efficacious dose 20 of Prozac for depression was twenty milligrams? 21 A Yes, sir. 22 Q All right. You say in 23 your trip report, Exhibit 19 dated June 13, 1988, 24 beginning with the second paragraph: There is 452 1 tremendous pressure from gurus and practicing 2 psychiatrists to make ten milligrams or even five 3 milligrams available. They are wrong that twenty 4 milligrams is excessive from the safety data we 5 have, but they would not lose appreciable efficacy 6 at ten milligrams. As they will titrate, and are, 7 between five, ten, to forty milligrams daily, no 8 matter what we do, we should make it easier for 9 them. I agree with the proposal to make available 10 a ten milligram dose without changing the label 11 dose guidelines as soon as possible, end quote. 12 Is that correct? 13 A Yes, sir. 14 Q Who were the gurus at 15 that meeting, or gurus? 16 A Well, there were some 17 faculty members who were speakers, and I don't 18 really remember who they were except one down 19 here, Doctor Ries or Ries, and I can't remember 20 who he was exactly, and then there were somewhere 21 on the order of -- as I recall that meeting, a 22 hundred or a hundred and fifty people that I think 23 were mostly practicing psychiatrists. 24 Q And practicing 453 1 psychiatrists were putting on you tremendous 2 pressure to make ten milligrams available? 3 A Yes, sir. 4 Q And apparently -- you say 5 they are wrong that twenty milligrams is excessive 6 from the safety data we have. Apparently they are 7 advising you that they feel that twenty milligrams 8 was excessive? 9 A Well, I'm not sure 10 whether they advised me that it was excessive, but 11 the psychiatrists' practice, in America at least, 12 is to titrate the dose of virtually all the drugs 13 that they use, and this was the first drug that 14 they had encountered that really didn't require 15 such titration. So that their practice and their 16 experience from all the other drugs in their 17 armamentarium, or almost all of them, were in fact 18 to use their clinical judgment on adjusting the 19 dose up and down, and that was not necessary with 20 this drug, but they didn't know all of the data 21 that I knew, and they therefore were not in the 22 same position that I was in terms of knowing what 23 was the most efficacious and safe dose. 24 Q You're saying you were 454 1 more knowledgeable concerning the safety of Prozac 2 at that time than the practicing psychiatrists and 3 gurus in San Diego -- in Long Beach, California? 4 A Yes, sir. 5 Q Who were treating 6 patients with Prozac? 7 A Yes, sir. 8 Q Apparently they were 9 expressing to you safety concerns because you say 10 they are wrong that twenty milligrams is excessive 11 from the safety data we have, correct? 12 A I think that's a 13 reasonable conclusion. 14 Q What type of safety 15 concerns did they express to you, Doctor Thompson? 16 A As I can best recall, 17 there were things like nausea and sweating and 18 tremulousness. 19 Q Nausea, sweating and 20 tremulousness? 21 A Tremulousness. 22 Q Jitteriness? 23 A I think tremulousness is 24 a more specific medical term. I don't know what 455 1 jittery means. 2 Q All right, we'll talk 3 about that later. Do you know Doctor Jan Fawcett? 4 A I'm not sure I know him 5 personally, but I certainly know of him by 6 reputation. 7 Q He's used the term 8 jitteriness, that's why I was asking you about 9 it. He's on the Lilly psychiatrist advisory 10 board, you do know that, don't you? 11 A Yes, sir. 12 Q Were there any 13 individuals that were there at that trip that were 14 on Lilly's psychiatric advisory board? 15 A I don't really know the 16 whole composition of that board and I can't 17 remember who the faculty were that were speaking, 18 so I'm not sure. As a matter of fact, I think -- 19 isn't Doctor Dunner in Washington state? 20 Q Yes. 21 A I think he was one of the 22 speakers at this meeting, and I think he was 23 also -- he was certainly one of our consultants 24 and I think he might have been on the advisory 456 1 board. 2 Q Was he reporting to you 3 the fact that he felt that twenty milligrams was 4 excessive? 5 A I don't remember his 6 saying anything one way or the other. 7 Q Did he advise you at that 8 meeting, as he's testified in his deposition, that 9 he prescribes concomitant sedating 10 tranquilizing -- concomitant sedating 11 antidepressants to his patients taking Prozac? 12 A I don't remember that 13 coming from Doctor Dunner. I did note in here 14 Doctor -- it's either Ries or Ries had noted his 15 use of trazodone. 16 Q All right. And the 17 reason they were doing that was to counteract the 18 nervousness, anxiety, agitation, jitteriness, 19 tremulousness that was being reported with Prozac 20 by them and their patients? 21 A Reported in association 22 of the treatment of underlying disorders with 23 Prozac, yes. 24 Q You say here that they're 457 1 wrong that twenty milligrams is excessive from the 2 safety data we have, but you say that we would 3 not -- they would not lose appreciable efficacy at 4 ten milligrams, correct? 5 A Yes, sir. 6 Q So you made a 7 recommendation that there be a ten milligram dose? 8 A No, sir, I made the 9 recommendation there be a ten milligram dosage 10 form without changing the labeled indicated dose 11 of twenty to eighty milligrams a day. 12 Q I understand that. But 13 you say that you would not lose appreciable 14 efficacy at ten milligrams? 15 A Yes, sir, but in your use 16 of the words dose, you included two different ways 17 in which scientifically we use the word dose, and 18 I wanted to make sure that we had a distinction 19 between those. 20 Q All right, I'll try to be 21 careful with that. Ten milligrams, you're saying 22 here, is just as efficacious as twenty milligrams? 23 A No, sir, I didn't say 24 that; that's not what it reads. 458 1 Q All right, what it says 2 is they would not lose appreciable efficacy at ten 3 milligrams? 4 A That's what my assumption 5 was, yes. 6 Q Is that accurate? 7 A Well, it depends on how 8 you define the word appreciable. There are 9 certainly other scientists at Lilly that disagree 10 with my assumption, and it was an assumption 11 because we had no data at ten milligrams. But I 12 was really addressing making a dosage form 13 available, not a change in the daily dose; so if 14 one gave ten milligrams in the morning and ten 15 milligrams in the evening, from the data we have, 16 I would be pretty comfortable with saying we 17 wouldn't lose appreciable efficacy. 18 Q But that's not the way -- 19 the context within which it's written, is it, 20 Doctor? 21 A Yes, sir, it is. It very 22 explicitly says, if I can read it to you again, 23 that in fact it says will not lose appreciable 24 efficacy at ten milligrams, that I think we ought 459 1 to make available a ten milligram dose, which 2 clearly means dosage form is my use of the word 3 dose -- 4 Q But you didn't say dosage 5 form, did you? 6 A No, sir, I used just the 7 word dose. 8 Q Like we lay people use 9 the word dose. 10 A It's clear from the 11 context, sir, that it says specifically not 12 changing the indications. 13 Q I understand that, I 14 understand, but you say they are wrong that twenty 15 milligrams is excessive from the safety data we 16 have, but they would not lose -- it's got to be 17 from twenty milligrams, doesn't it -- appreciable 18 efficacy at ten milligrams. You use twenty 19 milligrams in the same sentence, don't you? 20 A Yes, sir. 21 Q And you're referring to 22 twenty milligrams as being -- the complaint being 23 that it was excessive, aren't you? 24 A And the twenty milligrams 460 1 refers both to the dosage form and to the total 2 daily dose. 3 Q Well, it doesn't say 4 that, does it? 5 A It does scientifically. 6 MS. ZETTLER: Statisti- 7 cally significantly. 8 Q Do you think that when 9 the gurus and practicing psychiatrists were saying 10 twenty milligrams is excessive, they were talking 11 about dose, weren't they, dosage form? 12 A Dosage form. 13 Q All right. 14 A Yes, and I think many of 15 them were also talking about the total daily dose. 16 Q All right. And you say 17 they would not lose appreciable efficacy at ten 18 milligrams, is that right? 19 A Yes, sir. 20 Q Or is it your testimony 21 under oath here today, Doctor Thompson, that what 22 you meant was that you intended that those 23 psychiatrists prescribe ten milligrams in the 24 morning and ten milligrams at night and they 461 1 wouldn't lose considerable efficacy at that? 2 A That was one of the 3 meanings of my statement; the other meaning was 4 that probably if they used ten milligrams in 5 selected patients that they wouldn't lose 6 appreciable efficacy as well. 7 Q All right. So do we have 8 it now that you meant in one sense that it would 9 be all right to give ten milligrams per day per 10 patient and not lose appreciable efficacy in that 11 particular patient? 12 A I think that's possible 13 in some patients. I think the original package 14 insert suggested reducing the dose in certain 15 patients. Now, of course, there are other 16 strategies for reducing the dose besides using a 17 smaller pulvule. 18 Q That's correct, you could 19 prescribe it every other day, couldn't you? 20 A That would be certainly 21 one strategy. 22 Q Is that called alternate 23 day dosing? 24 A It sounds like the right 462 1 word to me. 2 Q Is that a scientific word 3 or scientific term that you would use? 4 A Yes, sir. 5 Q But it's your testimony 6 as we sit here right now that in June 1988, the 7 lowest efficacious dose of Prozac had been 8 established at twenty milligrams? 9 A Yes, sir. 10 (THOMPSON EXHIBIT NO. 20 MARKED FOR 11 IDENTIFICATION.) 12 THE WITNESS: Yes, sir. 13 Q (BY MR. SMITH) Exhibit 14 20 is a document dated July 7th, 1988, and it 15 appears to be authored by Doctor Gary L. Tauscher, 16 is that correct, or he's not a doctor, is he? 17 A I don't think so. I 18 think he may be a Pharm.D., I'm not sure. 19 Q He's a -- you think he 20 probably is a Pharm.D.? 21 A I think he may be. 22 Q You are not copied in on 23 this document, are you? 24 A No, sir. 463 1 Q Have you ever seen this 2 document before? 3 A I can't remember it 4 specifically, but I certainly remember a lot of 5 discussions about the dosage issue. 6 Q No, I'm talking about 7 this document, whether you've seen it before. 8 A I can't recall it 9 specifically. 10 Q The subject of the 11 document is Prozac Dosage Forms, Interim Update, 12 correct? 13 A Yes, sir. 14 Q It says the following 15 inputs from our CNS medical team will be of 16 interest to you, correct? 17 A Yes, sir. 18 Q And the CNS medical team 19 would be who? 20 A Well, I think at this 21 time -- I think Doctor Zerbe had already left for 22 the UK, so I would say it would be headed by 23 Doctor Masica and would include people like Doctor 24 Heiligenstein, Doctor Wheadon, Doctor Beasley and 464 1 others. 2 Q It says number of 3 requests, all three physicians have received 4 requests for a lower dose from a broad range of 5 psychiatrists, correct? 6 A Yes, sir. 7 Q It says the number of 8 requests are sufficient to predict that a survey 9 will demonstrate a general consensus regarding the 10 need for a lower dosage form, correct? 11 A Yes, sir. 12 Q Rationale. There are 13 three patient groups reflected in the requests: 14 A, side effects, patient tolerance. Correct? 15 A Yes, sir. 16 Q And there it mentions a 17 dosage related side effect, correct? 18 A Yes, sir. 19 Q Special patient 20 populations, correct? 21 A Yes, sir. 22 Q Speaking of primarily 23 elderly with other physical and mental problems, 24 correct? 465 1 A It says medical illness. 2 Q All right. And then 3 patients on long-term therapy -- 4 A Yes, sir. 5 Q -- may benefit from a 6 lower dosage form, correct? 7 A Well, what it talks about 8 is symptoms of fatigue after one to three months 9 of therapy, and that's news to me because I wasn't 10 aware of that. 11 Q You're not aware of any 12 fatigue reported? 13 A No, I'm not aware of that 14 being a -- 15 Q It says there the CNS 16 medical team apparently, the psychiatrists on 17 Lilly's staff have indicated a formula 18 presentation based on the requests to date and the 19 patients populations involved, a five milligram 20 pulvule appears to be the optimum preparation, end 21 quote, does it not? 22 A Yes, sir. 23 Q Now, was Lilly making a 24 five milligram pulvule then -- 466 1 A Not -- 2 Q -- in July of 1988? 3 A Not for sale. 4 Q Was Lilly making for sale 5 a ten milligram pulvule? 6 A No, sir. 7 Q You recommended in June, 8 a month before, that ten milligram be made 9 available -- 10 A Yes, sir. 11 Q -- as a dosage form? 12 A That's correct. 13 Q The CNS physicians were 14 recommending five milligram be made available as a 15 dosage form? 16 A Yes, sir. 17 Q In 1988 -- 18 A Yes, sir. 19 Q -- correct? Ten 20 milligram finally became available manufactured in 21 a pulvule by Lilly for public sale, did it not? 22 A Yes, sir, as well as a 23 liquid form, which of course permits virtually any 24 dosage. 467 1 Q I'm not talking about -- 2 my question didn't say anything about liquid form, 3 did it? 4 A It said what forms we had 5 made available. 6 Q I said in a pulvule 7 form. 8 A We made a ten milligram 9 available, yes. 10 Q In a pulvule form? 11 A I think that's the form. 12 Q When? 13 A I don't really remember 14 the date. 15 Q Approximately? 16 A 1990 or '91. 17 Q How about 1992? 18 A That could be the date. 19 Q How about February 1993? 20 A If you say so. 21 (THOMPSON EXHIBIT NO. 21 MARKED FOR 22 IDENTIFICATION.) 23 Q Look at Exhibit No. 21, 24 Doctor Thompson, and see if that will help you in 468 1 when the FDA approved Prozac in a ten-milligram 2 pulvule form for manufacture by Lilly for sale to 3 the public. 4 A It looks like it says the 5 FDA approved that dosage form December -- before 6 December 25th of 1992. 7 Q All right, so will you 8 take my word for it, it first came out to the 9 public in February 1993, sixty days later, after 10 approval? 11 A Yes, sir, I'll take your 12 word. 13 Q Some five years after -- 14 strike that -- four and a half years after you 15 originally recommended it be made available in ten 16 milligram form, correct? 17 A Almost the same interval 18 that the FDA took to review the original NDA. 19 MS. ZETTLER: Move to 20 strike as nonresponsive. 21 Q Can't that be answered 22 with a "yes" or "no", Doctor Thompson? 23 A The answer is yes. 24 Q And what is the answer? 469 1 A Yes. 2 (THOMPSON EXHIBIT NO. 22 MARKED FOR 3 IDENTIFICATION.) 4 THE WITNESS: Yes, sir. 5 Q (BY MR. SMITH) Doctor, 6 Exhibit No. 22 is a document dated November 7th, 7 1990, authored by Doctor John Heiligenstein, is it 8 not? 9 A Yes, sir. 10 Q And Doctor Heiligenstein 11 is a psychiatrist employed by Eli Lilly and 12 Company, is he not? 13 A Yes, sir. 14 Q He is an individual who 15 was on the Eli Lilly CNS medical team, was he not? 16 A Yes, sir. 17 Q And he was a psychiatrist 18 that was under your direction -- 19 A Yes, sir. 20 Q -- in November 1990, 21 correct? 22 A Yes, sir. 23 Q Doctor Heiligenstein says 24 in Point 1 of that document: Clearly, very 470 1 clearly there is a need for a lower dose in 2 pulvule form. Five milligrams would be most 3 appropriate, although at the child psychiatry 4 meeting in Chicago two point five milligrams was 5 suggested. The liquid formulation will only 6 partially address this need. Correct? 7 A Yes, sir. 8 Q Did you agree then, in 9 November 1990 that clearly, very clearly there was 10 a need for a lower dose in pulvule form? 11 A Yes. 12 Q Did you agree that five 13 milligram would be most appropriate? 14 A I don't think I agreed 15 with that. 16 Q All right. Doctor 17 Heiligenstein goes on to say in Point 2: We have 18 not established the lowest effective dose and if 19 we listen to the prescribers some effort should be 20 made to do this. If not MD's will continue to 21 dissolve pulvules and will anxiously await new 22 medications with greater flexibility in dosing, 23 end quote. Correct? 24 A Yes, sir. 471 1 Q Now, you said earlier 2 that the lowest efficacious dosage of Prozac was 3 twenty milligrams, and it had been established at 4 twenty milligrams. Now this document authored by 5 Doctor Heiligenstein says we have not established 6 the lowest effective dose, correct? 7 A That's what it says. 8 Q Do you disagree then with 9 Doctor Heiligenstein? 10 A Yes, sir. 11 Q He's wrong and you're 12 right? 13 A We disagree. 14 Q Who is right? 15 A Well, I have my opinion 16 and Doctor Heiligenstein obviously has his 17 opinion. 18 Q Based on what? 19 A Based on the control 20 prospective double blind study, zero versus five 21 versus twenty versus forty milligrams in 22 depression, and the dose response studies that we 23 had been doing at various doses in bulimia, 24 obsessive compulsive disease, obesity, smoking 472 1 cessation and other indications. 2 Q And you say you're better 3 equipped to interpret that data than Doctor 4 Heiligenstein? 5 A No, sir, I didn't say 6 that. 7 Q What makes you think 8 you're right and he's wrong? 9 A I didn't say that either. 10 Q He could be right and you 11 could be wrong? 12 A That's possible. 13 (THOMPSON EXHIBIT NO. 23 MARKED FOR 14 IDENTIFICATION.) 15 THE WITNESS: Yes, sir. 16 Q (BY MR. SMITH) Doctor 17 Thompson, Exhibit No. 23 is a document dated June 18 11th, 1991 and it appears to be authored by you? 19 A Yes, sir. 20 Q Do you recall authoring 21 this document? 22 A Yes, sir. 23 Q Is this a document that 24 you reviewed in preparation for your deposition? 473 1 A Yes, sir. 2 Q When did you last review 3 this document in preparation for your deposition? 4 A Roughly two, three weeks 5 ago. 6 Q Was this a document that 7 was shown to you by your lawyers, or is this a 8 document -- beg your pardon -- or is this a 9 document that you yourself retrieved from your 10 files? 11 A No, sir, this was 12 provided to me by my attorneys. 13 Q It says: Charlie, let me 14 ask you for help in putting together two slides 15 for the board. And I assume when you say Charlie, 16 you're talking about Charles Sampson? 17 A Yes, sir. 18 Q Who is Charles Sampson, a 19 statistician? 20 A Yes, sir, he was director 21 of the whole statistical group at that time. 22 Q He's the head honcho 23 statistician? 24 A One of the best I've ever 474 1 met anywhere. 2 Q The big cheese, as we 3 say, statistician, in Texas? 4 A Yes, sir. 5 Q You say -- you continue 6 to say: I have got to say something about ten 7 milligram both in regard to attributes and to the 8 logistics of when we filed in the US. 9 International filing is a big, big problem, end 10 quote. Correct? 11 A Yes, sir. 12 Q All right, so you're 13 making a presentation to the board of directors of 14 Eli Lilly and Company? 15 A Yes, sir. 16 Q And you're talking to 17 them about ten milligrams, correct? 18 A Yes, sir. 19 Q Apparently there is going 20 to be filed in the US -- you mean with the US FDA? 21 A Yes, sir. 22 Q You say international 23 filing is a big, big problem? 24 A Yes, sir. 475 1 Q What was the big, big 2 problem in international filing? 3 A Different countries 4 required different kind of testing of each new 5 formulation, and the same formulation can't 6 necessarily be used in each country, and it was a 7 very big piece of work to do the testing of a new 8 dosage form just for the United States much less 9 having to repeat that work for all the various 10 countries in Europe and the other places we were 11 marketing the drug. 12 Q Well, in fact no new 13 clinical trials were done to secure approval of 14 ten milligram by the FDA, was there? 15 A That statement is 16 correct, but not directly responsive to my 17 previous answer. 18 Q Well, whether it is 19 responsive or not, it is correct, okay? 20 MR. FREEMAN: We object 21 to -- 22 A That's correct. 23 Q You go on to say: I 24 don't think we have any ten milligram efficacy 476 1 data. Correct? 2 A That's what I said, but 3 it has to be understood that I meant in depression 4 because, of course, we did have ten milligram 5 efficacy data in obesity. 6 Q All right. So when 7 you're talking here in Exhibit 23, you're talking 8 about ten milligrams for the indication of 9 depression? 10 A Yes, sir. 11 Q It says, we do have -- 12 you continue to say: We do have the Wernicke 13 study of five milligram versus twenty and forty. 14 Some people have massaged those data to make five 15 milligram look good -- look not quite as good as 16 twenty milligram. Correct? 17 A That's correct. 18 Q Who has massaged the data 19 to make five milligram look not quite as good as 20 twenty milligram? 21 A I don't remember who the 22 individuals were. 23 Q When was the data 24 massaged to make this lower dosage form look not 477 1 as good as twenty milligram? 2 A Sometime after we had 3 locked the data on the study and this date. 4 Q When did you first learn, 5 Doctor Thompson, that anybody at Eli Lilly was 6 massaging data in any form, under any 7 circumstances? 8 A My goodness, I started 9 massaging data I think the first day I arrived at 10 Lilly, and I hope everybody else was doing the 11 same. 12 Q All right, to secure a 13 particular result? 14 A No, sir, to understand 15 the data in its entirety because there are very 16 many different statistical tests or other analyses 17 that can be done. When one looks -- 18 Q One might -- I'm sorry, 19 go ahead. 20 A One looks at the 21 robustness of a scientific conclusion by looking 22 at how different tests of the same data would help 23 you either substantiate or refute the same 24 conclusion. 478 1 Q Well, it looks like that 2 somebody is massaging or altering or interpreting 3 data to make a lower dosage look not quite as good 4 as a higher dosage? 5 A I don't think the words 6 that you've used are correct at all, and that is 7 certainly not what my meaning is. 8 Q You deny any situation of 9 any improper motive when you say some people have 10 massaged those data to make five milligram look 11 not quite as good as twenty milligram? 12 A I, one, dispute any 13 improper motive, but two, you used the word alter, 14 and I will absolutely affirm that after the data 15 lock there is no alteration of the data at all. 16 Q Then how did the massage 17 occur? 18 A Massage means applying 19 different tests to different elements of the 20 data. We had collected a whole variety of scales 21 of efficacy, we had a whole variety of data on 22 safety, we had these week by week in a group of 23 patients. Massaging meaning looking at old people 24 versus young people, fat people versus thin 479 1 people, one dose versus another, one week versus 2 another, one element of one scale versus one 3 element of another scale. It's a matter of doing 4 a thorough analysis, and each one of those 5 analyses can be done by multiple statistical 6 tests, many of which are thought to be equally 7 valid, and in other cases statisticians would 8 argue as to which the most valid was. 9 Q You go on to say: Some 10 people have massaged those data -- no, I'm sorry. 11 You go on to say: I'm not sure I know how that 12 massage was done, but I really -- I'd probably 13 need a slide to show whatever it is that makes it 14 look less good. Correct? 15 A Yes, sir. 16 Q I'd also like a slide 17 that shows a more global view of how it works so 18 we have a display of -- by week of the HAMD total 19 changes. Right? 20 A Yes, sir. 21 Q Then you go on to say: 22 As I recall, five milligram actually went down a 23 little faster than the others and ended up about 24 the same. Help me if you will on what to show on 480 1 ten milligram to the board. 2 A Yes, sir. 3 Q Correct? So do I 4 interpret that, Doctor Thompson, to indicate that 5 on the HAMD scores, actually five milligram caused 6 the HAMD scores to decrease a little faster 7 initially? 8 A The total scores. 9 Q The total scores, and 10 that the HAMD scores of five versus twenty ended 11 up about the same? 12 A Yes, sir, the total 13 score, I think that's what I said. 14 Q All right. I take that 15 to mean, Doctor Thompson, that five milligram ends 16 up about being the same as twenty milligram as far 17 as efficacy. 18 A No, sir. 19 Q Am I wrong, sir? 20 A Yes, sir. 21 MR. SMITH: Let's take a 22 break. 23 (SHORT BREAK TAKEN.) 24 MR. BOUR: The time is 481 1 2:47. 2 (THOMPSON EXHIBIT NO. 24 MARKED FOR 3 IDENTIFICATION.) 4 THE WITNESS: Yes, sir. 5 Q (BY MR. SMITH) Doctor 6 Thompson, Exhibit 24 is a document dated December 7 17th, 1990 authored by Lorenzo -- what is the last 8 name, sir? 9 A Tallarigo. 10 Q And who is Mr. Tallarigo? 11 A He's a physician, I think 12 gastroenterologist, but I'm not sure of his 13 subspecialty, from Italy. 14 Q And he is a Lilly 15 employee? 16 A Yes, sir. 17 Q And he's directed this 18 document to Alan J. Webber, correct? 19 A Yes, sir. 20 Q Alan Webber is a 21 regulatory scientist, is he not? 22 A That's correct. 23 Q Or was at the time? 24 A Yes, sir, is still today. 482 1 Q He is second in the 2 regulatory scheme behind Doctor Talbott? 3 A That's correct. 4 Q Who is Sigrid Heymanns? 5 A I think she's in the 6 German affiliate, but I'm not sure. 7 Q And William Oldfield? 8 A Bill Oldfield was at that 9 time in the regulatory group at Erl Wood. 10 Q And Francoise Rossignol? 11 A I think it's Rossignol, 12 and I think she was -- I'm not sure exact 13 position, but she was in the French affiliate. 14 Q We knew -- know who 15 Doctor Talbott is, he's the director of regulatory 16 affairs at Eli Lilly and Company, is he not? 17 A Yes, sir. 18 Q And Steven Whittaker? 19 A I think Steve is a CRA, 20 and I'm not too sure exactly what his assignment 21 was at this time. 22 Q The document says re your 23 request -- have you seen this document before, I'm 24 sorry? 483 1 A I don't think so. 2 Q It says: Re: Your 3 requests for Prozac. 1) We need to produce 4 evidence, paren, clinical trials, close paren, for 5 the benefit of these new dosages. We should be 6 very carefully not to suggest that these are the 7 more appropriate dosages if we do not want to run 8 of risk to have the higher dosages withdrawn from 9 the market. Correct? 10 A Yes, sir. 11 Q Number two says: 12 Difficult to answer if we do not know which data 13 we are talking about. Correct? 14 A Yes, sir. 15 Q Would that be data that's 16 massaged or unmassaged data? 17 A I have no idea; hopefully 18 massaged. 19 Q Three says it depends 20 which evidences we have of the benefits, doesn't 21 it? 22 A Yes, sir. 23 Q Do you know to what he's 24 referring there, sir? 484 1 A No, I don't. 2 Q Do you suppose that has 3 to do with massaging data on benefits and evidence 4 of benefits? 5 A Well, if it's looking at 6 analysis of data for evidence, I do hope it was 7 massaged very thoroughly. 8 Q It says here: PS. As 9 you may imagine, our price will be half of the 10 actual for the ten milligram and one fourth for 11 the five milligram. And I can't read that next 12 word. 13 A I think it's W-A-T-H. 14 Q Wath our objective is. 15 Do you know that -- what he could be referring to 16 there in the word W-A-T-H? 17 A He probably means what. 18 Q Beg your pardon? 19 A He probably means what 20 our objective is, question mark. 21 Q Probably if you or I were 22 writing it, we would say what is our objective? 23 A Yes, sir. 24 Q But he is Italian and 485 1 that's probably some type of translation? 2 A I guess. 3 Q Is what he's saying there 4 that the price that Lilly would get for ten 5 milligrams would be half the price that they would 6 get for twenty milligrams? 7 A I think that's what he's 8 saying, but, you know, I honestly don't know the 9 pricing structure in Italy. It wouldn't be true 10 in other countries. 11 Q Do you know what the 12 difference in cost is between twenty milligram and 13 ten milligram in the United States? 14 A Cost to manufacture or 15 the price sold at a drug store? 16 Q The price sold by Lilly 17 to their distributors. Doesn't Lilly sell to 18 their distributors? 19 A Yes, sir. I don't know 20 exactly what the price is, but I don't think it's 21 half. 22 Q Is it less, sir, for ten 23 milligrams -- 24 A Yes, sir, I think -- 486 1 Q -- than twenty 2 milligrams? 3 A Yes, sir. 4 Q Is it less expensive to 5 manufacture than twenty milligrams? 6 A I doubt that there is 7 much -- well, I don't know what the difference is; 8 I don't know what the manufacturing cost is. 9 Q But you probably think it 10 really isn't much -- there isn't much difference 11 in manufacturing cost between ten milligrams and 12 twenty? 13 A Well, there are two 14 aspects of it, one is to manufacture the bulk 15 material, so obviously there is less bulk, and 16 then the other is fill finish, but for many 17 products the fill finish, labeling and so forth 18 and distribution is a big fraction of the cost. 19 Q Right. You're going to 20 have that cost whether it be ten milligram, five 21 milligram or twenty milligram, correct? 22 A Yes, sir. 23 Q But certainly the cost of 24 the materials is going to be the major difference 487 1 in the cost, or will it? 2 A I think if -- the 3 difference in cost to manufacture would largely be 4 due to the difference in the materials, and I 5 don't know how the formulation differs between ten 6 and twenty. 7 Q Is it true, sir, that ten 8 milligram of Prozac is half the strength of twenty 9 milligram? 10 A Yes, sir. 11 Q Is it true, sir, that 12 twenty milligrams is half the strength of forty 13 milligrams? 14 A Yes, sir. 15 Q I had read something -- 16 and I may be wrong -- concerning an unusual aspect 17 of fluoxetine is that effect in platelets maybe or 18 presence in blood didn't necessarily go up on a 19 linear fashion; am I stating that correctly? 20 A I'm not aware of a 21 concentration dependency on platelet effect 22 because if doesn't have much platelet effect 23 anyway. And the blood concentration effects in 24 usual people is we would say normally reasonably 488 1 linear with the dose, so that if you give twice as 2 much within the usual range, the blood 3 concentrations of both the parent, both in 4 anthemas -- and both in anthemas of the daughter 5 molecules are roughly in the same ball park. 6 Q So you're going to get -- 7 if you take twice as much Prozac, you're going to 8 get twice as much in your blood? 9 A Roughly. 10 Q Does that mean there's 11 going to be twice as much antidepressant effect? 12 A No, sir. 13 Q Does it mean there's 14 going to be twice as much side effects? 15 A No, sir. 16 Q That's a more complicated 17 issue, isn't it? 18 A Yes, sir. 19 (THOMPSON EXHIBIT NO. 25 MARKED FOR 20 IDENTIFICATION.) 21 Q While he's looking at 22 that document, Doctor Thompson, do you know Doctor 23 Irwin Slater, have you ever met him? 24 A No, sir, I haven't. 489 1 Q You know who he is 2 though, don't you? 3 A Yes, sir. 4 THE WITNESS: Yes, sir. 5 Q (BY MR. SMITH) Exhibit 6 25 is a document authored by you, is it not? 7 A Yes, sir. 8 Q And it's dated when, sir? 9 A July 18, 1990. 10 Q And what time was it 11 written by you, sir? 12 A 6:48 in the morning. 13 Q When do you normally 14 begin work each day? 15 A About 6:15, something 16 like that. 17 Q And does that mean you're 18 at the office by 6:15? 19 A Yes, sir. Sometimes I 20 make calls from home, but usually I go into the 21 office and make calls and pick up the computer 22 messages there. 23 Q What? 24 A I'm sorry, usually I go 490 1 into the office and make calls or pick up computer 2 messages in my office rather than doing it from 3 home; occasionally I do it from home early in the 4 morning. 5 Q But you're generally at 6 the office working by 6:15? 7 A 6:15, 6:30, 6:00 o'clock, 8 something like that. 9 Q And how long do you 10 usually work each day? 11 A Until I get through. I 12 work at the office until anywhere from, say, 5:00 13 to 6:00, and then I go home and work until 9:00 or 14 10:00. 15 Q All right. Mr. Wood has 16 indicated in his deposition that you rarely 17 sleep. Has that -- have others made that comment 18 about your work habits? 19 A No, not really. I mean, 20 I'm an old ICU doctor, I get sleep when I need it 21 and where I can. 22 Q You indicate in Exhibit 23 25 that Paul Leber had called you the previous 24 day? 491 1 A Yes, sir. 2 Q And that you contacted 3 him at 6:15 AM that morning and had with him a 4 half hour conversation, very, very pleasant, with 5 Paul and Tom Laughren, correct? 6 A Yes, sir. 7 Q You made this call from 8 your office? 9 A Yes, sir. 10 Q And this document that 11 has been marked as Exhibit 26 -- 12 A 25? 13 Q I'm sorry, is it 25? 14 A I have 25. 15 Q Exhibit 25 is dated 6:48, 16 is that right? 17 A Yes, sir. 18 Q Is that the time that you 19 began printing or began your entry into the 20 computer or is that the time that the entry is 21 complete? 22 A I don't know for sure. 23 Probably at the end, I would guess, because I 24 think the way Messenger works, I think it accepts 492 1 text and then tags it at the end, but I'm not a 2 hundred percent sure of that. But I've won a lot 3 of typing contests, so I type fast. 4 Q All right. Now the Paul 5 Leber that you mentioned in this document is who? 6 A Doctor Leber who heads 7 the Division of Neuropsychiatric Drug Products at 8 the Food and Drug Administration. 9 Q Now, at that time -- at 10 the time that you made this call to Doctor Leber 11 at the FDA, had you been offered the job as deputy 12 chairman or deputy commissioner of the FDA? 13 A No, sir. 14 Q And the Tom Laughren that 15 you mentioned, what was his position? 16 A He's a psychiatrist who I 17 think had the title of group leader reporting to 18 Paul in that division at that time. 19 Q Was Paul Leber a 20 psychiatrist or is he a psychiatrist? 21 A Yes, sir. 22 Q Were they at the Food and 23 Drug Administration when you contacted them? 24 A Yes, sir; I wouldn't know 493 1 how to get a hold of them otherwise. 2 Q Well, it's 6:15 in the 3 morning when you called them -- 4 A Yes, sir. 5 Q -- in Indianapolis, this 6 is in July, so there would be a one-hour 7 difference? 8 A Yes, sir. 9 Q It would have been 10 seventeen -- it will have been 7:15 in Washington? 11 A Yes, sir, but I know 12 Doctor Leber sometimes gets in before 7:00. 13 Q How did you know that? 14 A Because he's asked me to 15 call him early before, because he, like I, gets 16 very busy starting at 8:00 o'clock or so and the 17 best way for us to avoid telephone tag is to talk 18 early. 19 Q You say the call was 20 about suicide? 21 A Yes, sir. 22 Q They said this morning's 23 Wall Street Journal article in their mind was 24 trivial and the reporter had called them, correct? 494 1 A Yes, sir. 2 Q What article in the Wall 3 Street Journal are you and they speaking of? 4 A I don't remember. 5 Q Do you remember what the 6 subject of the article was? 7 A I presume it had to do 8 with suicide, but I'm sorry, I don't remember the 9 article. 10 Q It had to do with suicide 11 and Prozac? 12 A Yes, sir, I assume. 13 MR. SMITH: Do you want 14 to change the tape? 15 (SHORT BREAK TAKEN.) 16 MR. BOUR: This is the 17 start of Tape No. 6, it's 3:07. 18 Q (BY MR. SMITH) In 19 connection with the Wall Street Journal article, 20 had you read that article, Doctor Thompson? 21 A I can't remember, 22 probably. Well, let me go back for a second, if 23 it was 6:15 in the morning, I normally don't get 24 the Wall Street Journal that early, so I'm not for 495 1 sure. 2 Q Well, you said he had -- 3 earlier you said he had called you the day 4 before. Could it have been that the article came 5 out the day previously? 6 A I could have misstated it 7 in the memo, I don't know. 8 Q That's right, you say 9 this morning's Wall Street Journal, correct? 10 A That's what I say, yes, 11 sir. 12 Q But they characterized 13 the article on Prozac and suicide as -- 14 characterized it as trivial to you -- 15 A Yes, sir. 16 Q -- in their discussion 17 with you? 18 A Yes, sir. 19 Q Did you characterize the 20 article in any way, sir? 21 A I don't remember doing 22 so. 23 Q Does this -- you've 24 reviewed this article -- I mean this document, 496 1 have you not? 2 A Yes, sir. 3 Q And you've reviewed it in 4 preparation for your deposition, have you not? 5 A Yes, sir. 6 Q Does this document 7 accurately or completely reflect the essence of 8 what was said that morning between you, Doctor 9 Leber and Doctor Laughren? 10 A I think so. 11 Q You go on to state, 12 quote: The issue is building and will not go 13 away. They feel some data are required and wanted 14 a study or studies of the issue. Paul suggested 15 several designs. Correct? 16 A Yes, sir. 17 Q The first is, and I 18 quote: One, a case control retrospective study, 19 before Teicher, seeing what the frequency is of 20 patterns developing the Teicher -- of patients 21 developing the Teicher syndrome. Paul felt that 22 the instrument should be developed in concert with 23 Teicher because only Teicher knows what it was 24 that he is seeing, end quote. Correct? 497 1 A Yes, sir. 2 Q Was that retrospective 3 done -- study done in concert with Doctor Teicher? 4 A No, sir. 5 Q Doctor Leber -- or you 6 report that Doctor Leber also requested, quote: 7 Two, cohort study. Paul feels the power is low 8 but if the incidence is three point five percent 9 even a small study should be able to establish 10 relative rates between fluox and other 11 treatments. Also it should set an upper limit on 12 the incidence of the emergence of suicidal 13 ideation, end quote. Correct? 14 A Yes, sir. 15 Q Was a cohort study such 16 as you and Doctor Leber discussed done by Lilly? 17 A No, sir. 18 Q You go on to quote Doctor 19 Leber: Three, best would be a large -- larger 20 blind prospective study designed with the help of 21 Teicher to detect his phenomenon, end quote. 22 Correct? 23 A Yes, sir. 24 Q Then you put the word his 498 1 in quotation marks? 2 A Yes, sir. 3 Q Was that study done? 4 A No, sir. 5 Q Of these three studies 6 that were suggested in this -- what is written 7 here concerning these three studies, is this 8 pretty much a quote from what Doctor Leber was 9 saying, our is this an interpretation of what he 10 was saying? 11 A Well, I can't remember 12 his exact words, but I can tell you that my usual 13 practice is as much as possible to try to both 14 quote people directly and to as accurately as 15 possible reflect what they said. So I think this 16 is an accurate reflection of what he said. 17 Q And obviously you said 18 that the phone conversation was initiated at 6:15 19 AM, it was a half hour conversation, and the 20 computer printout shows 6:48 PM, so this was 21 done -- 22 A AM. 23 Q AM. 24 A Yes, sir. 499 1 Q So this was done 2 immediately after you hung up from talking with 3 them, is that correct? 4 A Yes, sir. 5 Q What did you do, just 6 swivel around at your desk and start with the 7 memo? 8 A I don't have to swivel, 9 but I doubt that I started typing this until the 10 call was over. 11 Q No, I mean after you hung 12 up you immediately started typing? 13 A I think so, probably; 14 that's my usual practice. 15 Q And your attempt here was 16 to accurately reflect the substance of the 17 conversation as closely as you could? 18 A Absolutely. 19 Q You go on to say: Paul 20 said he wasn't commanding any of these particular 21 designs but wanted to express strongly his feeling 22 that some data are needed. He was very, very -- 23 he was very pleasant but obviously concerned, end 24 quote. Correct? 500 1 A Yes, sir. 2 Q You say here that he 3 wasn't commanding any of these particular designs, 4 but he was expressing strongly that some data was 5 needed. Did you have the impression at the time 6 you talked with Doctor Leber that he felt like 7 Lilly should do some study, whatever it was? 8 A No, I thought that he 9 felt that we ought to get some data, which doesn't 10 necessarily require a new study. 11 Q All right. But the three 12 suggestions he made that you outlined that he 13 specifically mentioned would have required 14 specific trial studies, would they not? 15 A Yes. The case control 16 study retrospectively wouldn't have required 17 enrolling new patients, but it would have required 18 a new design, a new way of capturing that data. 19 Q And it would also require 20 Doctor Teicher looking at that data also? 21 A If it were done in 22 concert with him, of course. 23 Q And that's what Doctor 24 Leber had suggested as an alternative? 501 1 A Yes, sir. 2 Q Beginning on the second 3 page, the first full paragraph, you say: I then 4 informed him of the suicide expert meeting next 5 Tuesday and told him we would furnish him with 6 detailed reports, et cetera. He was very pleased 7 but he said I agree that the experts on suicide 8 can give you some good testimony on 20/20, 9 Nightline and 60 Minutes, but what we really need 10 are good data, end quote. Correct? 11 A You're very close. 12 Q You put that portion of 13 what Doctor Leber said specifically in quotes, do 14 you not? 15 A Yes, sir. 16 Q Does that mean that as 17 far as what you're saying there is that that is an 18 exact quote of what he said? 19 A Yes, sir. 20 Q You go on to say: I 21 agree and commented that Lilly would perform a new 22 study of this issue to be designed in concert with 23 him with no commitment to design, scope or 24 timing. Correct? 502 1 A Yes, sir. 2 Q Now, here you say you 3 committed that Lilly would perform a new study? 4 A Yes, sir. 5 Q By that, I would assume 6 you were talking about a new clinical trial? 7 A Yes, sir. 8 Q Is that what you meant 9 then? 10 A Yes, sir. 11 Q Was a new clinical trial 12 such as you meant here ever done? 13 A No, sir. 14 Q You go on to say that you 15 did not make any commitment to him with respect to 16 design, scope or timing? 17 A Yes, sir. 18 Q Why is that? 19 A Because we hadn't figured 20 out how to do the studies and we didn't know the 21 logistics of doing them, et cetera. All I said 22 was that I thought it was appropriate for Lilly to 23 do another study. 24 Q Have you ever changed 503 1 your mind about that, Doctor Thompson, that it's 2 appropriate -- that you think it's appropriate 3 that Lilly do another study? 4 A No, I -- well, if I can 5 answer that fully, I think it would have been very 6 helpful at that time to have another study, and in 7 fact we worked very hard on doing that. So I 8 don't think it was -- I would have said my mind 9 was changed that it was appropriate to do it; 10 however, we found that it was impossible to do it, 11 and therefore, whether it was appropriate or not, 12 it couldn't be done. 13 Q By virtue of the problems 14 that you suggested yesterday of not having the 15 ability to enroll enough patients and enough 16 investigators to do the study? 17 A Yes, sir. 18 Q You mention a suicide 19 expert meeting that was going to occur on Tuesday? 20 A Yes, sir. 21 Q Who were your suicide 22 experts that you were meeting with at that time? 23 A I think that was a 24 meeting that included Gary Tollefson, I think 504 1 Doctor Winokur was there, and I'm not sure who 2 else. If that's the meeting that I'm trying to 3 recall at which I was present for at least a hour 4 in the afternoon, I think we had about four 5 experts. 6 Q That would have been the 7 summer of 1990? 8 A Yes, sir. 9 Q Tollefson, Winokur -- 10 A Those are two that I 11 recall at some suicide expert meetings. Again, I 12 think that was the one that was referenced here, 13 but I can't really specifically distinguish one 14 from another. 15 Q You go on to say in the 16 second paragraph: He then said that Temple had 17 asked that we not effect label changes on the 18 suicide issue without clearing them with the FDA 19 first. I think Paul would like this in general, 20 but emphasized Bob's interest specifically on the 21 suicide issue. There wasn't anything else that 22 hinted to more need for label change. I promised 23 we would on suicide not use the provision for 24 effecting the change. Correct? 505 1 A Yes, sir. 2 Q What label change are you 3 speaking of there, sir? 4 A Well, the specific one 5 was a label change related to the suicide issue. 6 Q All right. Okay, then 7 are you saying there that Paul, meaning Doctor 8 Leber, would like a label change concerning 9 suicide? 10 A You know, in reading 11 that, I'm not sure exactly what that phrase means, 12 because the other thing is Paul would like that in 13 general. In other words, we don't change the 14 label without talking to the FDA before we effect 15 the label change. 16 Q All right. 17 A I know that to be his 18 position. I do not believe that it's correct to 19 assert that at this time Paul wanted to have a 20 change in the label, because of course he could 21 have commanded that at any time. 22 Q Well, that raises another 23 point: Could Lilly have requested to the FDA that 24 they change the label on Prozac to include a 506 1 warning concerning suicidality and the use of 2 Prozac? 3 A Yes, sir. 4 Q And that would not have 5 had to have been approved by the FDA? 6 A Depending upon the exact 7 wording, you're correct. 8 Q So Lilly could have done 9 that voluntarily? 10 A Yes, sir. Virtually all 11 of our label changes are voluntary. 12 Q And you could have 13 petitioned to the FDA or suggested to the FDA that 14 your label be changed, right? 15 A Yes, sir. 16 Q To include statements 17 concerning Prozac and suicidality? 18 A Yes, sir. 19 Q And there wouldn't have 20 been a problem getting the FDA to approve -- or 21 the FDA would not have had to approve a change of 22 that nature in order to effectuate the label 23 change? 24 A Depending upon the exact 507 1 wording, because you have to go back to the 2 regulation and look at the exact wording that 3 allows a company to make changes, increases the 4 level of safety. 5 Q But you can't think of 6 any wording in connection with suicide that would 7 have required FDA approval? 8 A Yes, I can. 9 Q What would that have 10 been? 11 A That the drug is never 12 associated with suicide or suicide thinking in any 13 patient. 14 Q If you had a warning that 15 said in a small subset of individuals, Prozac may 16 be related to increased suicidal ideation in some 17 patients, would that have been a change that would 18 have required FDA approval? 19 A I don't think so. 20 Q If you had a warning or a 21 statement in the Prozac labeling to the effect 22 that Prozac may in some individuals induce violent 23 aggressive behavior, would that be a warning that 24 would require FDA approval? 508 1 A I don't think so. 2 Q All right, then, what is 3 it that you are speaking about when you're talking 4 about suicide issue without clearing them with the 5 FDA first? 6 A Well, I think it's 7 exactly what I typed, and that was that Bob Temple 8 asked that we not effect label changes, in other 9 words use the provision he was speaking about just 10 to change the label on the suicide issue, so I 11 would assume that means any statement about 12 suicide, without clearing them with the FDA first. 13 Q All right. They wanted 14 to be aware that you were going to make that 15 change and asked you to do that as a matter of 16 courtesy? 17 A Yes, sir. 18 Q But they weren't 19 requiring that as a matter of regulation? 20 A No, sir, the regulation, 21 I think you and I agreed on. 22 Q You say I think Paul 23 would like this in general, but emphasized Bob's 24 interest specifically on the suicide issue. 509 1 A Yes, sir. 2 Q Are you saying there that 3 Bob was interested in the suicide issue or Bob was 4 interested in a label change on the suicide issue? 5 A Neither one. I think 6 what that means is that Paul in general would like 7 to have a consultation with the FDA before we made 8 a label change, in general, but that Bob Temple 9 was specifically making a request in regard to any 10 change related to the suicide issue. 11 Q Did he say why Bob was 12 making that request? 13 A No, I think I would have 14 put down what it was if there had been any 15 specific indication about that. 16 Q Can you give us any input 17 into any other discussion that you all had that 18 was appropriate to this paragraph? 19 A Not contemporaneously at 20 that time. Remember, we were making a large 21 number of changes in the Prozac label to add 22 various postmarketing adverse events as they 23 occurred, hyponatremia, pancreatitis and so forth, 24 and in one case the FDA had in fact mandated a 510 1 change which we felt was wrong and in fact they 2 backed off of that request. So in general, Lilly 3 had initiated those requests with the exception, 4 to my knowledge, of that one request where they 5 wanted us to put a couple of things in the label. 6 There are other instances -- I'm sorry, there are 7 other instances with other drugs, however, where 8 again a similar kind of request has been made that 9 Lilly not just change the label under that 10 provision, but rather have a discussion with the 11 FDA first. 12 Q But this was a matter of 13 courtesy as opposed to a matter of it's a 14 regulatory requirement that you do it? 15 A You and I are in absolute 16 agreement, that we had the regulatory right to 17 increase the safety statements in the label 18 without prior approval or even notification of the 19 FDA until we actually had the final printed label. 20 Q All right. And you 21 advised them that you wouldn't change the label on 22 suicide without running it by the FDA? 23 A Yes, sir, that's what I 24 say. 511 1 Q There is more discussion 2 about label changes and sending a chronology of 3 changes to Washington with Max; I assume you mean 4 Max Talbott, is that correct? 5 A Yes, sir. 6 Q I don't understand why 7 they were asking for a chronology of label 8 changes. 9 A Well, the FDA doesn't 10 have the easiest access to the large volume of 11 documents that we submit to them; their record 12 rooms are very difficult for them to make sure 13 that they get everything in the right chronology. 14 It's must easier to ask us, since we know exactly 15 our printed label, to provide them with that kind 16 of a list, so that's not an unusual request. 17 Q Why were they wanting 18 that? 19 A I don't know. 20 Q Did it have anything to 21 do with the suicide issue? 22 A I don't think so. 23 Q This was just an 24 unrelated discussion you were having? 512 1 A I think so. As I said, 2 we'd made a whole number of changes over time in 3 that label. 4 Q I understand that, but 5 that paragraph there where you're talking about a 6 chronology of label changes on Prozac and having 7 Max go to Washington and get it in his hands this 8 afternoon, is it your testimony that that didn't 9 have anything to do with the suicide issue? 10 A It is, but I thought I 11 was the one that offered to carry it down that 12 afternoon, I think I was asking Max to prepare it. 13 Q Then you say in the 14 middle paragraph there: He asked that we fax 15 nothing to him unless he has agreed to it 16 beforehand. He said that someone has found that 17 mailbox and it makes a second route of 18 information, end quote. Correct? 19 A Yes, sir. 20 Q I don't understand the 21 context of that sentence. 22 A Well, I think I explained 23 it better later because I think it had been public 24 information in pink sheets or one of those 513 1 newsletters that someone had broken into Doctor 2 Leber's office and had broken into his computer, 3 and in fact I recall when I visited him, I forgot 4 whether it was before or after this, that he'd had 5 some fancy new combination lock installed on his 6 door, and what he was indicating was he was 7 concerned that the fax machine, wherever it was 8 near his office, was not a secure route of 9 communication. 10 Q Well, was Lilly and the 11 FDA having some sensitive communications at this 12 time? 13 A Well, only sensitive in 14 the sense that if we faxed him something and 15 expected that he got it, but that somebody had 16 tampered with it and taken it away and he didn't 17 get it, that would have certainly been a 18 miscommunication problem. So what he was asking 19 is that when we fax him things, that we get -- we 20 talk to him specifically and say we're about to 21 fax this to this fax number. 22 Q Is it -- was he 23 suspecting that somebody was taking faxes that 24 were sent to him and he wasn't getting those 514 1 faxes? 2 A That's what he indicated 3 to me. 4 Q You go on to say: So he 5 said he was warning people like us not to use fax 6 unless we specifically agreed with him to do so. 7 Then you say: I think we better start using the 8 Washington office more aggressively to transmit 9 stuff as this links to the break-in into Paul's 10 computer, et cetera. I also want to reenergize 11 our periodic discussions about opening an office 12 adjacent to the Parklawn Building specifically for 13 the purposes of -- purpose of expediting 14 submissions, dialogs and meetings. 15 A Yes, sir. 16 Q End quote, correct? 17 A Yes, sir. 18 Q Now, the Parklawn 19 Building is where Doctor Leber officed? 20 A At that time. 21 Q Right. Is that correct? 22 A Yes, sir. 23 Q So Lilly had an office 24 next to the FDA at that time? 515 1 A No, sir, but I was asking 2 that we open an office next to the FDA at that 3 time. 4 Q All right. For the 5 purpose that you stated, expediting submissions, 6 dialog and meetings? 7 A Yes, we have a lot of 8 problem because we send such a large volume of 9 material to the FDA, a lot of that is sent through 10 the mail or some expedited route. Often we will 11 get back some receipt, postcard; we don't know 12 that it was received in the right room, we don't 13 know exactly -- we don't have a good record of 14 exactly who received it, and so one day as I was 15 leaving the FDA, I noticed that there was an 16 office for rent on the left-hand side of the 17 street right across from the US Pharmacopeia and I 18 came back and prepared a proposal that we have a 19 small office there where we could actually print 20 from our computers at that site and have a Lilly 21 employee physically walk the material down the 22 street and physically hand it to somebody and 23 record who that was. 24 Q Was that done -- 516 1 A No, sir. 2 Q -- did you say? 3 A No, sir, I'm sorry, it 4 wasn't done. 5 Q Never was done, still 6 hasn't been done to this day? 7 A Still has not been done 8 to this day. 9 Q Is that still a 10 suggestion that you think is good? 11 A Well, it has changed 12 recently because the FDA has now moved into a 13 whole bunch of different buildings that are 14 several miles apart, so at this time most of the 15 drugs part of the FDA was in that one location, 16 and so a single office would have been within 17 walking distance. Now, unfortunately, they're 18 in -- or whatever, they're in several different 19 buildings. 20 Q Would this have required 21 hiring an individual to staff that office full 22 time? 23 A Not necessarily. We 24 talked about that a great deal, because we have 517 1 several full-time staff within Washington, and 2 it's not clear that it would have required an 3 additional person. 4 Q Where was your offices in 5 Washington at that time? 6 A They were downtown, I've 7 forgotten exactly which address they were. 8 Q What would have been the 9 problem of submitting your material to the 10 downtown office and having it sent out to the FDA? 11 A Well, we sometimes did 12 that, but it was a long drive from downtown 13 through traffic out to Parklawn, parking, carrying 14 it in, et cetera. It would have been more 15 convenient -- there was another part of my 16 suggestion, and that was since many of us went to 17 the Parklawn building, many people were there 18 twice a week or more, we didn't have any place to 19 work when we were there, so you would fly down in 20 the morning, you may have a one hour or thirty 21 minute meeting, the plane wouldn't get you home 22 until evening, and you would have six or seven 23 hours where you could work and there was 24 absolutely no place to go work. So part of my 518 1 suggestion was to have a computer and a fax 2 machine and a phone line there so that Lilly 3 employees could productively work when they 4 weren't actually in meetings down the street. 5 Q But that was never done? 6 A No, sir. We tried to use 7 a room in a hotel not far away, a couple of blocks 8 down the hill, but that really wasn't very 9 popular, so it never was done. 10 Q You go ahead and say: 11 Paul said that this reminded him of the Bendectin 12 situation. You could just keep it in your 13 medicine chest and then if anything bad happened, 14 you had someone -- something to blame, end quote. 15 Correct? 16 A Yes, sir. 17 Q And you go on to 18 comment: This is worrisome considering the 19 terrible pain Bendectin has gone through. 20 Correct? 21 A Yes, sir. 22 Q What pain has Bendectin 23 gone through that you're aware of? 24 A There was a lot of public 519 1 debate about whether Bendectin produced birth 2 defects in women who took it during pregnancy to 3 ameliorate the nausea and vomiting of pregnancy, 4 and in my opinion that was a terrible pain because 5 there was a lot of scientific evidence that bore 6 on this that really wasn't reflected in the lay 7 press coverage and I think it unnecessarily 8 worried an awful lot of pregnant women. 9 Q Were you aware of a 10 situation in Philadelphia that occurred just 11 recently where a woman who gave birth to a child 12 with club feet alleged Bendectin was responsible 13 for this deformity and she and her child were 14 awarded nineteen point five million dollars? 15 A No, sir, I wasn't aware 16 of that. 17 Q And that a jury there 18 found that Bendectin was responsible for this 19 child's problems? 20 A No, I probably would have 21 said very terrible pain if I had known that at the 22 time. 23 Q You were speaking of pain 24 to the drug as opposed to the patient there? 520 1 A No, sir, I think probably 2 both; both in terms of unwarranted attacks on a 3 good drug as well as on the fact that this was 4 about the only effective therapy for women and 5 hyperemesis gravidarum can be a very -- virtually 6 a life-threatening illness in some women. 7 Q The words you used there 8 say, quote, this is worrisome considering the 9 terrible pain Bendectin has gone through -- 10 A Yes, sir. 11 Q -- correct? 12 A Yes, sir. 13 Q You don't say this is 14 worrisome considering the terrible pain Bendectin 15 and those individuals consuming Bendectin have 16 gone through, do you? 17 A No, sir, I don't say 18 that. 19 Q You go on to say: Paul 20 is taking a position in talking with outside folks 21 today that Lilly and FDA and working together on 22 this suicide issue and following closely the 23 postmarketing events, but that there are no 24 denominators and the best that can be done is put 521 1 a cap on the number of events. Correct? 2 A Yes, sir. 3 Q Is that what Doctor Leber 4 said? 5 A As best I could recall 6 it. 7 Q When you use the term 8 cap, in quotes, is that because that was a term 9 that was specifically used by Doctor Leber? 10 A That I can't be sure of 11 because it's also a slang term for the upper 12 confidence limit, and my guess is that I put those 13 quotes there not because he used that word 14 necessarily but to indicate it was a slang term. 15 Q You've been using quotes 16 to him -- attributable to him throughout this 17 document, have you not? 18 A Yes, sir. 19 Q And you've told me 20 already that you specifically used those quotes to 21 denote those as words uttered by Doctor Leber, is 22 that correct? 23 A Yes, sir. 24 Q Now are you telling me in 522 1 this sentence, when you're using the term cap, 2 that you're not attributing that as language he 3 used, sir? 4 A I'm saying that I can't 5 remember which usage was -- I was referring to, so 6 I don't remember whether the word cap was his 7 exact word or whether it was my slang for the 8 upper confidence limit. 9 Q Did Doctor Leber say "the 10 best that can be done is to put a cap on the 11 number of events"? 12 A He said something that 13 meant the same, but I can't be sure that he used 14 the word cap other than an equivalent like upper 15 ninety-five percent confidence limit or highest 16 number that makes sense. 17 Q All right. Now, what is 18 your explanation here, Doctor Thompson, of what 19 cap means in terms used here? 20 A It means like the highest 21 incidence value that would make sense from the 22 data that you've seen. That's what the 23 denominator here means. In other words, you can 24 count -- 523 1 Q Wait a second, I'm 2 writing this down. 3 A I'm sorry. 4 Q Highest incidence value 5 that would make sense -- 6 A From the data that you've 7 seen. 8 Q From the data that you've 9 seen, all right? 10 A Yes, sir. 11 Q So instead of saying cap, 12 you could have said highest incidence value that 13 would make sense from the data you have -- you've 14 seen, correct? 15 A Yes, sir. 16 Q So the sentence could 17 read: Paul is taking a position in talking with 18 outside folks that Lilly and FDA and working 19 together on the suicide issue and following 20 closely to postmarketing events, but that there 21 are no denominators and the best that can be done 22 is to put a highest incidence value that would 23 make sense from the data that you've seen on the 24 number of events? 524 1 A Yes, sir. 2 Q Is it your sworn 3 testimony here, Doctor Thompson, that when you use 4 the word cap, you're not using the word limit; you 5 don't mean to imply the word limit? 6 A I said upper confidence 7 limit. I mean, formally in statistics, and Paul 8 is an excellent statistician -- 9 Q Are you telling -- I'm 10 sorry, go ahead, complete your answer. 11 A I mean, the formal 12 definition of that would be probably be phrased in 13 statistical lingo as the upper confidence limit on 14 the estimate of incidence. 15 Q Are you saying cap as a 16 slang word used by statisticians? 17 A Yes, sir. 18 Q And it's defined by 19 statisticians, the word -- the slang word cap is 20 defined as highest incidence value that would make 21 sense from the data you've seen? 22 A Yes, sir. 23 Q All right. When you're 24 referring to cap, are you talking about cap on 525 1 number of events? 2 A Yes, sir; I think this 3 refers specifically to the suicide issue, not just 4 the total number of adverse events. 5 Q Right. It appears to me, 6 in reading this as a lay person, that what you're 7 saying -- that what Doctor Leber was saying to you 8 was that you should limit the number of adverse 9 events? 10 A Well -- 11 MR. FREEMAN: Just a 12 minute. I object to the form of the question in 13 that it's not clear as to what he means. 14 MR. SMITH: Reporting on 15 Prozac. 16 MR. FREEMAN: The intent 17 of the question seems to me to imply that you were 18 in some way to not report certain adverse events 19 relating to suicide, but limit them in number; I 20 think that's what the question is getting at and I 21 want to be sure you're understanding it before you 22 answer it. 23 Q (BY MR. SMITH) Yes, and 24 that is indeed what my question is getting at. 526 1 A Well, there's absolutely 2 no possibility that either Doctor Leber or I had 3 any -- any flavor to this conversation of not 4 reporting all the adverse events that we knew 5 about and were required to report. That's not at 6 all the content of this statement at all. Now, be 7 very careful, however, in using the word limit 8 because I've repeatedly said that in fact the 9 formal interpretation of what I was saying is that 10 from the numerator, which is the reports that 11 we're seeing, we don't have a solid denominator, 12 but the best we can do is saying from the data 13 that we're looking at, we can put an upper limit 14 on what we think the maximum value would likely be 15 of the incidence of those reports. That's the 16 correct interpretation of what I've seen, and 17 absolutely no way was there a suggestion -- I'd go 18 to jail for not reporting an adverse event, and 19 Doctor Leber may be the most strict person at the 20 FDA in terms of following the regulations 21 precisely, so if that's what you mean, that's 22 silly. 23 Q Well, did he use the word 24 cap? 527 1 A I don't remember, sir. 2 Q Was he suggesting that 3 the best that could be done is to put a cap on the 4 number of events? 5 A Yes, sir -- 6 Q I mean, that was a 7 suggestion by him? 8 A Yes, sir. Remember 9 earlier we talked about the three and a half 10 percent, and you know that the three and a half 11 percent comes from the Teicher article, and the 12 content of the discussion was with that Doctor 13 Leber and I did not believe that that was a 14 correct reflection on the incidence of this 15 phenomenon, and so what we're saying is we can get 16 a numerator on how many events we're seeing 17 because we're reporting all that through DEN to 18 the FDA, but we don't have a denominator because 19 we don't know exactly the number of people 20 exposed. So the best we can do is to make a guess 21 at the denominator, count exactly the numerator, 22 and as best we can statistically estimate what the 23 upper limit is on what the incidence rate is; 24 neither one of us thought it would be anywhere 528 1 close to three and a half percent, but we can do 2 that. 3 Q Why didn't you say that 4 here? 5 A Well, I do type a lot, I 6 must admit, but cap seems to me to be the way in 7 which I express that phenomenon here and 8 elsewhere. I'm sorry you misunderstood it, but 9 that's a perfectly legitimate use of the word. 10 Q The first part of that 11 sentence says Paul is taking a position in talking 12 with outside folks that Lilly and FDA and working 13 together -- 14 A I meant are, I mistyped. 15 Q I wanted to quote you 16 accurately with respect to what was said. 17 A You did it very well, 18 sir. 19 Q Lilly and FDA are working 20 together on the suicide issue? 21 A Yes, sir. 22 Q You say Paul is taking 23 that position, wasn't in fact that occurring? 24 A Absolutely. 529 1 Q All right. 2 A As in all other issues. 3 Q Why did you say there 4 Paul is taking a position in talking with outside 5 folks? 6 A Because he wanted to 7 reassure the outside folk that we were cooperating 8 and we were working together to understand this 9 phenomena. 10 Q You say you had a number 11 of individuals who had -- who had been the subject 12 of reports of suicide -- 13 A Yes, sir. 14 Q -- while on Prozac, 15 correct? 16 A Yes, sir. 17 Q You had that by virtue of 18 reports that were submitted to Lilly? 19 A Yes, sir. 20 Q That was a number that 21 was known? 22 A Yes, sir, that's the 23 numerator. 24 Q What was the number at 530 1 that time? 2 A Oh, I'm sorry, I can't 3 reconstruct that. 4 Q Approximately? 5 A Well, I don't know in 6 1990. We can go look up what the number is, but I 7 think there were more than a hundred. You mean -- 8 you said suicide, so you're not including suicidal 9 ideation or acts; and in my reply to you I meant 10 from all sources, including all worldwide clinical 11 trials, spontaneous reporting, et cetera. 12 Q Are you talking about 13 suicides or suicidal acts and ideation? 14 A Suicide -- 15 Q Or just suicide? 16 A My reference to the 17 number of more than a hundred was suicides, people 18 who died. 19 Q So you knew that number? 20 A Yes, sir. 21 Q And the way you know that 22 number is by virtue of reports to you? 23 A Yes, sir. 24 Q And you're saying there 531 1 wasn't any way to know -- when you say there are 2 no denominators, you're talking about there is not 3 any knowledge of how many people had been exposed 4 to Prozac? 5 A That's correct. We knew 6 from the clinical trials reasonably accurately, 7 but we didn't really have a solid number from the 8 spontaneous experience. 9 Q Well, didn't you have 10 market data available to know how many people were 11 taking Prozac? 12 A Yes, sir, we were 13 submitting probably quarterly at this time our 14 reports that included what's called probable 15 increased frequency records which have a very 16 explicit way of trying to calculate what in fact 17 these incidence rates are across different 18 populations, trials, spontaneous, US, 19 international, country by country. 20 Q Couldn't you use that as 21 your denominator? 22 A Well, we were, but it's 23 not an exact number and that's why you need a 24 confidence interval about the accuracy of the 532 1 ratio that you calculate. 2 Q Okay, but wouldn't that 3 have been something that Lilly could have done 4 with some reasonable degree of certainty? 5 A That's what a confidence 6 limit is, it's the expression of the certainty of 7 a specific point estimate; that's exactly what 8 we're talking about here. 9 Q You know how much with 10 certainty, how many Prozac pills are being 11 manufactured? 12 A Yes, sir. 13 Q You could probably call 14 up somebody at Eli Lilly and Company and get a 15 count today concerning how much total Prozac had 16 been manufactured, could you not? 17 A With reasonable accuracy. 18 Q Lilly keeps -- that's 19 because Lilly keeps careful records of their 20 manufacturing process, do they not? 21 A Very careful records. 22 Q You can also look at 23 Lilly's records and determine how much pills of 24 Prozac are sold to their distributors, can't you? 533 1 A Yes, sir. 2 Q Because Lilly keeps 3 careful records of the amount of pills sold -- 4 A Yes, sir. 5 Q -- right? 6 A Yes, sir. 7 Q Now, once that medication 8 gets to your distributor, you lose control of the 9 medication as far as specifically knowing where a 10 particular pill went, correct? 11 A Yes, that's correct. 12 Q But you also require your 13 distributors to give you certain information about 14 what they're doing with your -- with their 15 product, don't you? 16 A I presume so. 17 Q All right. In other 18 words, the -- you know the geographical area that 19 the distributor is covering? 20 A Yes, sir. 21 Q You know the pharmacies 22 that the distributor is supplying? 23 A I'm not sure we know 24 that, but if you say so. 534 1 Q Well, you know the volume 2 of sales that the distributor is making of Prozac? 3 A Of my own knowledge, I 4 don't know what we know from distributors. 5 Q Is there any way you know 6 how many prescriptions are written concerning 7 Prozac? 8 A Not specifically, and 9 especially not country by country because of the 10 parallel importing problem that occurs with places 11 like Europe. 12 Q Let's talk about the 13 United States only then. 14 A You can buy data from 15 sources like IMS that do audits that try to 16 measure the total number of prescriptions, but the 17 problem of course in estimating the number of 18 patients is that you've got to look both at dose 19 and duration, so obviously you can count how many 20 pills you think have in fact been sold by the end 21 distributor, like the pharmacy, but you've got to 22 factor in whether the patient is taking eighty 23 milligrams a day, whether they took it for one day 24 or whether they took it for a year and so forth, 535 1 so that's what makes that denominator particularly 2 uncertain. 3 Q But you can get it with 4 some degree of confidence, can't you? 5 A Yes, sir, that's why you 6 would put together a confidence level. 7 Q All right, is that the 8 cap that you're talking about? 9 A It's the upper bound of 10 that confidence limit, yes, sir. 11 Q Was that done? 12 A Yes, sir. 13 Q What is it? 14 A I don't know. I mean, it 15 was done very frequently to look at the upper 16 bound, and it would be in our quarterly reports if 17 in fact there had been any indication of an 18 increase in frequency. 19 Q Was that ever reported to 20 the FDA in terms of a count, here is the cap on 21 the number of events? 22 A Yes, if it was in fact 23 tripped on a quarterly report because there was 24 any indication of increased frequency, all of that 536 1 testing would have been there for sure. I don't 2 remember whether that in fact led to one of those 3 reports or not. We were making quarterly reports 4 on all the adverse events. 5 Q Well, can -- did you ever 6 report to the United States Food and Drug 7 Administration that on a particular date it 8 appears that a particular number of individuals 9 have consumed Prozac? 10 A Yes, sir. 11 Q All right. And do you 12 recall what that number has been at any time? 13 A I think now it's ten or 14 eleven million, something in that order. 15 Q Now, are you saying that 16 there are ten or eleven people that are currently 17 taking Prozac? 18 A No, sir. 19 Q Or are you saying that 20 there are ten or eleven million people who have at 21 one time taken at least one Prozac pill? 22 A Yes, sir, I thought 23 that's what your question was and that's what my 24 answer means. 537 1 Q I think we were on the 2 same wavelength, I just wanted to make sure I was 3 clear. But can't you tell me, or can Eli Lilly 4 give me a reasonable estimate of the number of 5 individuals that are currently taking Prozac as 6 part of the depression therapy currently? 7 A I'm not sure how 8 accurately we have that number. 9 Q Who would be able to 10 supply us with that data at Lilly? 11 A Someone in marketing 12 probably. 13 MR. SMITH: Let's take a 14 little break. 15 (SHORT BREAK TAKEN.) 16 Q (BY MR. SMITH) Have you 17 heard of Doctor Seymour Fisher, Doctor Thompson? 18 A I don't think so. 19 Q A Ph.D. psychologist at 20 the University of Texas, Galveston Medical Branch? 21 A I don't think so. 22 Q Who had done -- has done 23 some postmarketing surveys of psychotropic 24 medications? 538 1 A Again, I don't think I've 2 heard of him. 3 Q Have you heard of 4 postmarketing surveys where an entity can go 5 through a drug chain and furnish questionnaires to 6 individuals who obtain drugs from their pharmacy? 7 A Yes, sir, I think Upjohn 8 pioneered that kind of study. 9 Q All right, and 10 specifically can you give me some idea of how 11 those kind of studies work? 12 A I don't know Doctor 13 Fisher's study, but the way Upjohn does it is to 14 have the pharmacist ask the patient about adverse 15 events or other events that have happened on a 16 drug that they're taking, and then they collect 17 that data by phone. 18 Q All right. Have you 19 heard of a situation where the patient is given a 20 postcard or something to list adverse events and 21 then he calls a particular number that conducts a 22 survey of those events? 23 A I've certainly heard of 24 that in Europe, but I'm sure that's also done in 539 1 the US. 2 Q Doctor Stuart Montgomery 3 has done studies of that nature; are you aware of 4 that? 5 A No, sir. 6 (THOMPSON EXHIBIT NO. 26 MARKED FOR 7 IDENTIFICATION.) 8 THE WITNESS: Yes, sir. 9 Q (BY MR. SMITH) Is that 10 exhibit marked twenty -- 11 A Six. 12 Q Exhibit 26 appears to be 13 a document authored by you dated November 14th, 14 1990 -- '83, correct? 15 A Yes, sir. 16 Q And it speaks of 17 fluoxetine surveillance? 18 A Yes, sir. 19 Q And the subject of the 20 document is a surveillance program to detect the 21 ximelidine-like neuropathy that was thought to be 22 at that time associated with Prozac, is that 23 right? 24 A I don't think we ever 540 1 associated the ximelidine-like neuropathy with 2 Prozac, and in fact I think that sentence really 3 says that I don't believe -- I do not believe the 4 purpose of the surveillance program is to detect 5 ximelidine-like neuropathy. 6 Q That's why I said it was 7 thought at one time that -- 8 A We were concerned that it 9 might be related to serotonin and Prozac, but we 10 never had any evidence of it. 11 Q All right. What was -- 12 well, you go ahead and list the purpose of a 13 surveillance study, correct? 14 A Yes, sir. 15 Q And you say, one, it will 16 define much better the adverse experiences with a 17 population incidence between one and five thousand 18 and one in two hundred that we have -- that we may 19 have missed in the fourteen hundred souls to date 20 and those of which we have seen only a single 21 example, such as a severe rash or the -- 22 A Erythema multiforme. 23 Q Erythema multiforme. 24 Correct? 541 1 A Yes, sir. 2 Q So are you saying there 3 that a surveillance trial could detect adverse 4 experiences in a more sophisticated manner -- I 5 don't know if that's the word -- a more -- it's 6 more likely to pick up an adverse experience that 7 wouldn't be seen in a smaller clinical trial? 8 A Yes, sir, I think wrong 9 in being more sophisticated because the quality of 10 the data usually is much less than in a tightly 11 controlled trial, but the problem is with a rare 12 adverse event, you'd have to study a lot of people 13 before you'd know what its real incidence is. 14 Q All right. And number 15 two, if a surveillance study was done, you could 16 get data on individuals with less severe 17 depression? 18 A Yes, sir. 19 Q Three, you could get data 20 on varying dosages of Prozac? 21 A Yes, sir. 22 Q Four, it could provide 23 data on a -- on starting with a full dose if we 24 structure it that way, correct? 542 1 A Yes, sir. 2 Q What did you mean by 3 that, sir? 4 A Most of the pivotal 5 trials in a new drug application had started with 6 an initial dose of twenty milligrams and then gone 7 to forty milligrams, and allowed the investigator 8 to go then to sixty and eighty, and the average 9 dose used in those NDA trials, at least the 10 pivotal ones, I think was about sixty-four 11 milligrams. So most of the investigators had 12 started with twenty and then fairly quickly 13 progressed up to doses of sixty or eighty, 14 averaging sixty-four. The question is what 15 happens when you start with a sixty or eighty 16 milligram dose, and to my knowledge, at this time, 17 we didn't really have very much data on the first 18 dose being that great. 19 Q What is the full dose for 20 Prozac now? 21 A Twenty milligrams. 22 Q All right. 23 A For depression. 24 Q Yes, that's what I'm 543 1 speaking of. You say in number five it can 2 provide long term safety data, correct? 3 A Yes, sir. 4 Q Six, it can identify rare 5 drug-drug interactions, perhaps in unusual 6 populations, elderly, et cetera, that can only be 7 identified in such a trial, correct? 8 A Yes, sir. 9 Q When you were speaking of 10 drug-drug interactions in connection with Prozac 11 and depression, did you have anything specific in 12 mind? 13 A Well, we looked at 14 certain interactions like Prozac plus warfarin, 15 Prozac -- I think we tested digoxin, Prozac plus 16 Valium as I recall, I think we had some ethanol 17 data, I've forgotten whether it was here or later, 18 but you test a panel of drugs that differ in their 19 own metabolism to see if your drug will change 20 their kinetics and vice versa. 21 What I was saying in here 22 was it's always possible that of all the panoplia 23 of drugs somebody might take along with 24 fluoxetine, there could be a drug-drug interaction 544 1 that you wouldn't see in a designed trial because 2 you hadn't picked that drug, and it could be that 3 you only see that interaction in some special 4 population of patients. 5 Q All right. Could it be 6 said, Doctor Thompson, that clinical trial data is 7 limited in the number of patients that are 8 examined? 9 A Yes, sir. 10 Q And that clinical trial 11 data is limited by virtue of the selection of 12 patients that are examined? 13 A Yes, sir. 14 Q And that the 15 postmarketing experience in connection with a drug 16 will give you an experience with a greater number 17 of patients? 18 A If you design it that 19 way, sure. 20 Q And if you examine the 21 adverse reaction reports on Prozac that occurred 22 in the postmarketing phase, you'll see adverse 23 experiences that are experienced by a larger 24 number of people? 545 1 A Yes, sir. 2 Q And it will be adverse 3 experiences that are experienced by people who are 4 subject to a greater degree of varying illnesses? 5 A Depending on how the 6 physicians prescribe it, sure. 7 Q Well, you know that 8 Prozac is being prescribed by general 9 practitioners? 10 A I hope so. 11 Q It's designed to be 12 prescribed by general practitioners, or it's 13 intended by Lilly that general practitioners 14 prescribe Prozac? 15 A Sure, if they're treating 16 depression. 17 Q Yes. 18 A Yes, sir. 19 Q I mean, there's no 20 limitation suggested by Eli Lilly and Company, 21 that Prozac only be prescribed by psychiatrists? 22 A No, we actually launched 23 it, as you know, first to psychiatrists before 24 opening it up to internists like me, but there's 546 1 no limitation implied. It's for a disease, not 2 for the practitioner who treats the disease. 3 Q In fact Lilly encourages 4 that general practitioners prescribe Prozac for 5 depressive illness? 6 A I hope so. 7 Q Well, you know so -- 8 A Yes, sir. 9 Q -- that Lilly encourages 10 that Prozac be prescribed by general practitioners 11 for treatment of depression? 12 A Yes, sir. 13 Q Internists? 14 A Yes, sir. 15 Q Obstetricians? 16 A Yes, sir. 17 Q Orthopedic surgeons? 18 A Yes, sir. 19 Q Neurologists? 20 A Yes, sir. 21 Q The whole spectrum of 22 specialties? 23 A Yes, sir. 24 Q As well as just general 547 1 practitioners who see a spectrum of illnesses? 2 A Yes, sir. 3 Q So if you do a 4 surveillance study or if you examine postmarketing 5 reports of adverse experiences, you're going to 6 see the experience that your drug has had in a 7 greater number of people? 8 A Yes, sir. 9 Q With a greater variety of 10 problems? 11 A Usually. It depends on 12 how you design the clinical trials and how it's 13 used in practice. 14 Q But for Prozac, you know 15 so? 16 A That's a reasonable 17 expectation. 18 Q And those individuals in 19 the clinical trials, there were some parameters on 20 other medications that they could be taking? 21 A Yes, sir, that's correct. 22 Q But in the general 23 public, there is not necessarily the parameters 24 that are as strict as were employed in the 548 1 clinical trials? 2 A That's correct, sir. 3 Q As far as other drugs 4 that these patients can be taking? 5 A Yes, sir, that's correct. 6 Q So you're going to see a 7 lot more drug-drug interaction, as you mentioned, 8 in postmarketing evaluation? 9 A Well, there might not be 10 any drug-drug interactions, but you'd have a 11 greater chance of seeing them if in fact they did 12 in fact occur. 13 Q So do you think that the 14 idea of postmarketing surveillance studies is a 15 good idea in connection with following drug 16 safety? 17 A That's difficult to 18 answer yes or no. Again, the quality of the data 19 is usually not very good, and if you look at the 20 real experts in this field, the real 21 pharmacoepidemiologists, most of them now do not 22 believe that you get a lot out of postmarketing 23 surveillance, people like you tell us and Gerry 24 Faich in fact have reviewed a fairly large number 549 1 of postmarketing surveillance trials of the size 2 of ten to fourteen, twenty thousand patients, and 3 have found that very rarely do you find anything 4 that actually is new or that leads to a label 5 change. So, you absolutely are correct that you 6 have the potential to find a lot of new stuff in 7 terms of what actually has happened in doing 8 postmarketing surveillance with a bunch of drugs 9 is not very much has been found. 10 Q Well, isn't the FDA 11 spontaneous reporting system set up to pick up 12 adverse experiences that occur after a drug has 13 been marketed? 14 A Yes, sir. I think you 15 before talked about a study, which I 16 distinguished -- and maybe I shouldn't have -- 17 between that and spontaneous reporting. 18 Q Well, should you -- is 19 there a distinction? 20 A Absolutely. 21 Q What is that distinction, 22 sir? 23 A Well, in doing a study, 24 one would in fact define the population to be 550 1 enrolled and the data instrument to collect the 2 data. You suggested earlier mailing a postcard or 3 giving a postcard to the patient or calling them 4 or having them call or whatever, that's a 5 mechanism to try to increase the reporting of 6 adverse events and to define both the numerator 7 and the denominator so that you can make a more 8 accurate estimate of incidents. Whereas in 9 spontaneous reporting, you just record everything 10 that you're told, and as we talked about before, 11 make the best estimate you can of the number of 12 people exposed. 13 Q Well, are you saying that 14 the FDA postmarketing adverse event reporting 15 system is more likely to pick up problems with 16 drugs than a surveillance study? 17 A Well, there -- 18 Q Or neither is likely? 19 A No, but I would like to 20 try to be precise in answering that. The more 21 patients that are exposed to the drug, the more 22 likely you are to get a report of an adverse 23 event. But we believe that the spontaneous 24 reporting rate, just from marketing experience, is 551 1 relatively low, especially for less serious 2 events, and so the reason for doing a 3 postmarketing study as opposed to just marketing 4 the drug and seeing what happens to be reported to 5 you is an attempt to try to increase both the 6 number of reports you get per patient, but also 7 perhaps to get more information on more validity 8 on those reports. So, the bigger the marketing, 9 the more people that take the drug, the more 10 likely you are to get some kind of reports, but 11 you're probably going to get better quality data 12 from doing a specific study. 13 Q All right. Well, has a 14 study -- postmarketing surveillance study been 15 done on Prozac? 16 A I don't think so, nothing 17 that I would call that way, because we've done, I 18 don't know, thirty or thirty-five thousand 19 patients in controlled trials, which is bigger 20 than, I think, every postmarketing surveillance 21 study that I know of. 22 Q But the answer to my 23 question is no, there has not been any 24 postmarketing study done on Prozac of which you 552 1 are aware? 2 A Yes, sir, that's correct. 3 Q Well, have there been 4 thirty-five thousand people studied in the 5 depression trials? 6 A No, sir. 7 Q How many individuals have 8 been studied on the depression trials? 9 A I don't know for sure, 10 but it must be in the range of ten thousand. 11 Q And of those ten thousand 12 individuals, how many took Prozac? 13 A Probably roughly fifty 14 percent. 15 Q Well, wouldn't some of 16 those studies use comparators -- 17 A Yes, sir. 18 Q -- and some of those 19 studies use placebos? 20 A And some of the studies 21 had arms that had different doses of Prozac, so 22 some would have less than fifty percent and some 23 would have more than fifty percent, so I was 24 making a guess as to what they would be overall. 553 1 Q So somewhere around five 2 thousand people have consumed Prozac in the 3 clinical trials? 4 A In controlled prospective 5 trials of depression with Prozac. 6 Q Approximately five 7 thousand? 8 A That's my guess. 9 Q All right. How many of 10 those individuals finished the clinical trials? 11 A I don't have any idea on 12 that. 13 Q Would it be less than 14 five thousand? 15 A Yes, sir, it would be 16 less than a hundred percent of whatever the number 17 is. 18 Q What has been the dropout 19 rate generally of depressed individuals taking 20 Prozac during the clinical trials? 21 A I don't know 22 specifically. As I recall from the early trials, 23 it was in the range of twenty, twenty-five 24 percent, and I think it's been less in some later 554 1 trials. 2 Q So could we say that of 3 the five thousand people who got Prozac in the 4 depression trials, only four thousand completed 5 the study -- 6 A Well -- 7 Q -- in which they were 8 enrolled? 9 A We're getting to four 10 thousand because I made a guess that there had 11 been thirty-five thousand in worldwide controlled 12 trials overall, prospective trials, and then you 13 asked how many of those involved depression and I 14 made a guess that it might be ten thousand total, 15 and then I made a guess that maybe half of them 16 were on Prozac. 17 Q Yes, sir. 18 A And now I made a guess 19 about how many of them might have completed. 20 That's getting to be a fairly wild guess, but I 21 don't have a better number than four thousand. 22 Q All right. I'm just 23 trying to get some idea of how many individuals 24 completed the depression clinical trials on 555 1 Prozac. 2 A Yes, sir, I understand. 3 Q And if you -- are you 4 comfortable with approximately four thousand? 5 A Well, I think the number 6 is bigger than that, but that's the best number I 7 could come up with off the top of my head. 8 Q Now, the FDA was 9 suggesting in connection with studying the suicide 10 issue, studying ten thousand individuals, is that 11 right? 12 A Yes, but the ten 13 thousand -- my understanding of what they were 14 suggesting, and that was actually a number that 15 was just batted around between Lilly and the FDA, 16 was total patients in the trials for all diagnoses 17 and all doses, including placebo, comparators, as 18 well as Prozac. 19 MR. SMITH: Okay. 20 (THOMPSON EXHIBIT NO. 27 MARKED FOR 21 IDENTIFICATION.) 22 THE WITNESS: Yes, sir. 23 Q (BY MR. SMITH) Exhibit 24 27 is a document dated January 27th, 1984 that is 556 1 unsigned, is it not? 2 A No, it seems to be 3 authored by six statisticians. 4 Q All right. Enos, 5 Hardison, et al.? 6 A Yes, sir. 7 Q And they are offering 8 their comments on the fluoxetine surveillance 9 plan, correct? 10 A Yes, sir. 11 Q And you are listed as an 12 addressee? 13 A Yes, sir. 14 Q Or recipient of that 15 document, correct? 16 A Yes, sir. 17 Q Do you recall receiving 18 the document? 19 A Vaguely. 20 Q Is this a document you 21 reviewed in preparation for your deposition? 22 A No, sir. 23 Q Of course, in 1984, 24 Prozac had not been marketed? 557 1 A That's correct. 2 Q So the comments on the 3 first page appear to be discussing premarketing 4 surveillance studies, would that be right? 5 A Well, the first sentence 6 says a premarketing study, we believe that a 7 control is mandatory for any premarketing study, 8 but in general, I wouldn't have used -- but it was 9 not my document, but I wouldn't have used a 10 surveillance plan to mean a premarketing study. 11 Q All right. If you turn 12 to Page 2, they do start talking about 13 postmarketing surveillance, do they not? 14 A Yes, sir. 15 Q They say, quote: We 16 support postmarketing surveillance in general 17 since it allows us to collect comparative 18 information on other antidepressants and it 19 provides insight into the usefulness of our 20 premarketing experience for predicting efficacy 21 and safety in the general population. Our 22 preferred plan is the postlaunch surveillance via 23 pharmacists which is conducted by a contracted 24 firm. The use of another company should not 558 1 directly jeopardize any relationship between Lilly 2 and physicians since this other company is just 3 doing, quote, market research, end quote within a 4 quote. Also the possibility of bias of 5 suspicion -- or suspicion of bias is reduced since 6 Lilly's only involvement is to commission the 7 study and receive the results, end quote. 8 Correct? 9 A Yes, sir. 10 Q So they're saying there 11 that they would support a postmarketing 12 surveillance study employing pharmacists and an 13 independent company -- 14 A Yes, sir. 15 Q -- to conduct that, and 16 they talk about the benefits of that would be that 17 you reduce the suspicion of bias in the study, 18 correct? 19 A Yes, sir. 20 Q Was any such study done 21 on Prozac? 22 A I don't think so. 23 (THOMPSON EXHIBIT NO. 28 MARKED FOR 24 IDENTIFICATION.) 559 1 THE WITNESS: Yes, sir. 2 Q (BY MR. SMITH) Exhibit 3 28 is apparently a document dated in January 1985 4 authored by Steve Bandak at Lilly House, London, 5 correct? 6 A Yes, sir. 7 Q And what was -- is it 8 Doctor Bandak? 9 A Yes, sir. 10 Q Position? 11 A What's his position? 12 Well, he's a physician full time at Lilly. At 13 that time -- I think he was based at that time in 14 the UK, but at one time he was in the United 15 States briefly. Now he's -- excuse me, now he's 16 medical director in Canada. 17 Q All right. There is some 18 handwriting on the front page of this document. 19 Do you recognize that handwriting? 20 A No, sir, I don't. 21 Q Is it yours? 22 A No, sir, it's not. 23 Q Okay. Have you ever seen 24 this document before? 560 1 A I don't think so. 2 Q Do you know whether or 3 not the data analysis as proposed by Doctor Bandak 4 in connection with the side effects was done as 5 suggested? 6 A Well, some of them 7 certainly were done because I know, for example, 8 when Doctor Leber told us exactly how to do the 9 safety update, many of the kinds of analyses and 10 graphs and tables that he wanted are in fact 11 these, but -- so many of these in fact I know have 12 been done, but I can't tell you exactly what 13 sequence or what dates. 14 Q But -- and you can't tell 15 me for sure that all of them have been done 16 either, can you? 17 A No, I can't. 18 Q And those that were done 19 would have been done at the request of Doctor Paul 20 Leber with the Food and Drug Administration? 21 A Well, I said I know 22 explicitly, because I remember his instructions on 23 how to do many of the adverse event analyses. 24 That's why I say I know that that encompassed 561 1 this. What was done between the date of this memo 2 and that safety update I'm not a hundred percent 3 sure on. 4 Q When were these 5 instructions given to you by Doctor Leber? 6 A I don't remember when we 7 filed that first safety update, but it took us 8 about -- roughly about six months after the 9 instructions before we were able to get through 10 the analyses. 11 Q All right. You mentioned 12 earlier a -- a confrontation you had with Doctor 13 Leber at the 1985 advisory committee meeting. 14 A Yes, sir. 15 Q Tell me about that. 16 A Well, I was sort of the 17 leader of the Lilly delegation to that meeting, 18 and Doctor Leber was sort of the leader of the FDA 19 delegation. The advisory committee, as you know, 20 is a group of outside experts who meet 21 periodically to advise the FDA, but in fact those 22 meetings, at least the neuropsychiatric division 23 meetings, are pretty much run by FDA staff, and 24 Doctor Leber repeatedly attacked me on various 562 1 aspects of the Prozac data. It was an all-day 2 meeting. 3 Q All right. Was the issue 4 of suicide ever mentioned at that meeting? 5 A I don't think so. 6 Q Was the issue of violent 7 aggressive behavior mentioned at that September 8 1985 advisory committee meeting? 9 A I don't think so. But 10 again, I'm sure there's a transcript, and nowadays 11 they're videotaped. 12 Q Was the issue of mania or 13 psychosis addressed in that meeting? 14 A That I can't remember. 15 It was an all-day meeting that went into all the 16 efficacy and safety data, so the specific things, 17 I have difficulty remembering what was discussed 18 there. 19 (THOMPSON EXHIBIT NO. 29 MARKED FOR 20 IDENTIFICATION.) 21 THE WITNESS: Yes, sir. 22 Q (BY MR. SMITH) Exhibit 23 29 is a copy of the information concerning use or 24 German package insert for Prozac in Germany. The 563 1 first page of this document is a certification 2 from the translating service that it's an accurate 3 translation, all right? 4 A Yes, sir. 5 Q And I couldn't swear to 6 you that this is an accurate translation, but they 7 have sworn so, okay? 8 A Yes, sir. 9 Q Have you ever seen this 10 document before? 11 A I don't think this 12 specific document, no. I have seen translations 13 of Germany labels, but I don't think I've seen 14 this one before. 15 Q All right. Do you -- in 16 looking at this one, do you recall anything 17 different about this one than others that you've 18 seen? 19 A Frankly, this one had 20 some of the -- what I would call later 21 postmarketing events included in it, and I don't 22 recall seeing a German label that had pancreatitis 23 or hyponatremia or the five week monamine oxidase 24 inhibitor warning in it before. That's why I was 564 1 searching for the date, so it must have been a -- 2 it must have been a more recent one than the 3 introductory one. 4 Q This is just -- this was 5 just one that a couple of -- a year ago -- six 6 months ago we had sent to us and had translated. 7 But this is in effect, as far as your 8 understanding, where it says information 9 concerning use, this is something that goes to the 10 patients? 11 A I think that's right, but 12 I don't really know the German market that well. 13 Q All right. At least it 14 appears to be -- a lot of this language directed 15 to a patient as opposed to a physician? 16 A Yes, sir. 17 Q Turn to Page 2 of the 18 translation of the information going to patients 19 in Germany. You see the phrase "risk of suicide"? 20 A Yes, sir. 21 Q Do you have a problem as 22 the chief scientific officer with the phrase "risk 23 of suicide" in the Prozac package insert? 24 A With those three words? 565 1 Q Yes. 2 A No, sir. 3 Q Do you have a problem as 4 the chief scientific officer of Eli Lilly if those 5 three words are used as an implication that Prozac 6 is related to a risk of suicide in individuals? 7 MR. FREEMAN: That's not 8 an accurate statement, object to it on this 9 ground: It's under risk patients, not risk of 10 Prozac. 11 MS. ZETTLER: He didn't 12 ask him about that, Joe. 13 MR. SMITH: That was the 14 question back, so he can relax. 15 MR. FREEMAN: You talked 16 about it being related to risk of suicide is what 17 you said. 18 A If it says a decreased 19 risk of suicide, I have no problem at all. 20 Q Would you have any 21 objection if the implication was that it was an 22 increased -- that individuals taking Prozac run an 23 increased risk of suicide? 24 A Yes, I have an objection 566 1 to that. 2 Q The insert says Fluctin, 3 which is Prozac in Germany, correct? 4 A Yes, sir. 5 Q Fluctin does not have a 6 general sedative effect on the central nervous 7 system, correct? 8 A Yes, sir. 9 Q Do you have any objection 10 to that language in the German package insert as 11 the chief scientific officer of Eli Lilly and 12 Company? 13 A No, I think that's 14 generally correct. 15 Q The next sentence there 16 says therefore, for his or her own safety, the 17 patient must be sufficiently observed until the 18 antidepressive effect of Fluctin sets in. 19 A Yes, sir. 20 Q Do you see that? Do you 21 have any objection to that language being 22 contained in the German package insert as the 23 chief scientific officer of Eli Lilly and Company? 24 A No, sir. We're 567 1 specifically talking about depression here? 2 Q Yes. 3 A Yes, sir, I agree. 4 Q The next sentence says: 5 Taking an additional sedative may be necessary. 6 This also applies in cases of extreme sleep 7 disturbances or excitability. Actually, I read 8 you the next two sentences. I think they go 9 together, don't you? 10 A Yes, sir. 11 Q Do you have any objection 12 to that language being contained in the German 13 package insert as the chief scientific officer of 14 Eli Lilly and Company? 15 A Yes, sir. 16 Q What is that objection, 17 sir? 18 A That I don't believe we 19 have any controlled clinical trial data that would 20 suggest that an additional sedative may be 21 necessary in the general patient, and I don't know 22 that I would recommend giving a general sedative 23 for extreme sleep disturbances or excitability. 24 Again, I don't think we have data and I'm not 568 1 aware of any evidence that in fact that's an 2 appropriate medical statement. 3 Q Well, Lilly was using 4 sedatives in their clinical trials for individuals 5 that had sleep disturbances, were they not? 6 A We allowed, in controlled 7 clinical trials, doses of drugs -- the chloral 8 hydrate dose I regard to be essentially 9 ineffective in the doses used in both trials, and 10 very low doses of benzodiazepines, and I would not 11 regard anybody who was reading this to say taking 12 an additional sedative, I don't think they would 13 have either used chloral hydrate in those doses, 14 nor would they have used such a small dose of a 15 benzodiazepine, so I don't think that the data 16 from the clinical trials relates to this 17 specifically. 18 Q Do you know specifically 19 what dosage chloral hydrate and benzodiazepines 20 were administered in? 21 A I think that in many of 22 the protocols the dose of chloral hydrate was 23 point five to one gram. There was at least one 24 protocol where I think it went up to two grams. 569 1 Q So are you saying that 2 that's ineffective? 3 A Yes, sir. 4 Q Then why would it be 5 allowed by the protocol? 6 A Well, in the case that 7 the investigator was for some reason using that in 8 the patient, it didn't disallow the patient from 9 being included in the protocol. I'm just saying 10 as a clinical pharmacologist and a person who has 11 used a lot of chloral hydrate in my patients, I 12 don't think you see an effective dose in an adult 13 that small. 14 Q You don't think it would 15 make you sleepy? 16 A Not one gram, not unless 17 you take alcohol or another sedative on top of it. 18 Q What was the dosage of 19 benzodiazepines being administered? 20 A I don't remember exactly, 21 but it was a pretty small dose. 22 Q But it was being 23 administered for relief of anxiety, wasn't it? 24 A I don't recall that. I 570 1 recall that -- I recall the statement in those 2 protocols as saying that the patient could take 3 those sedatives in up to those doses and be 4 continued in the protocol, but I don't recall that 5 there was an indication for why they would be 6 used. 7 Q Well, but there was 8 statements in the protocols that allowed the use 9 of that if these situations occurred during the 10 trials, insomnia and agitation or anxiety or -- 11 A If you say so; I can't 12 remember the specific wording. 13 (THOMPSON EXHIBIT NO. 30 MARKED FOR 14 IDENTIFICATION.) 15 THE WITNESS: Yes, sir. 16 Q (BY MR. SMITH) Have you 17 seen Exhibit 28 -- is that 28? 18 A Thirty. 19 Q Exhibit 30 before? 20 A Yes, sir. 21 Q All right, and I assume 22 that was shown to you in preparation for your 23 deposition? 24 A Yes, sir. 571 1 Q That is a document dated 2 March 17th, 1993 authored by Tina for Doctor Atul 3 Pande, correct? 4 A It says thanks, Tina -- 5 it must be Kristine Schlanser, yes, sir. 6 Q Is she an assistant or 7 secretary for Doctor Pande, do you know her? 8 A I have no idea who she 9 is. 10 Q Who is Doctor -- identify 11 Doctor Pande for us? 12 A He's a psychiatrist who 13 at this time was a full-time employee of Lilly. 14 Q What's his position now? 15 A He's left the company 16 now. 17 Q Oh, what was his position 18 when he left the company? 19 A I think he was a clinical 20 research physician; I think he had been promoted 21 from being an associate clinical research 22 physician. 23 Q When did Doctor Pande 24 leave? 572 1 A Roughly two or three 2 months ago. 3 Q Do you know why he left? 4 A Yes, sir, I think he's 5 now department chairman of psychiatry somewhere. 6 Q All right. So apparently 7 he had requested a list of studies with 8 psychoactive medications, had he not? 9 A Yes, sir. 10 Q And Tina has prepared for 11 him a list of studies with psychoactive 12 medications? 13 A Yes, sir. 14 Q And there it says chloral 15 hydrate was allowed, correct? 16 A Yes, sir. 17 Q Chloral hydrate was 18 allowed for sleep, wasn't it? 19 A You're not in the first 20 section, let me catch up with where you are. 21 Q I don't know if it ever 22 talks about insomnia in this document, but isn't 23 it your understanding that chloral hydrate is a 24 hypnotic -- 573 1 A That's what I would call 2 it. 3 Q -- that's administered 4 for sleep, it's a sleeping pill? 5 A Yes, sir. 6 Q Does it have to be 7 prescribed by a physician? 8 A Yes, sir. 9 Q What's the -- does it 10 have a brand name or a trade name? 11 A Not that I'm -- not that 12 I can recall. 13 Q It also says also if a 14 patient complains of agitation, the dose study 15 should be reduced and the patient may receive a 16 benzodiazepine at the investigator's discretion, 17 does it not? 18 A Yes, sir, that's what it 19 says. 20 Q So the patients were 21 getting tranquilizers -- benzodiazepines are 22 tranquilizers, are they not? 23 A No, sir, I wouldn't call 24 them that. 574 1 Q Would you call them 2 sedatives? 3 A A little -- probably 4 closer to sedative than tranquilizer, but how 5 about antianxiety drug, I'd accept that one. 6 Anxiolytic would be an exact term. 7 Q Well -- or do many 8 sedatives have anxiolytic properties? 9 A I would doubt that 10 chloral hydrate does, but I'm not aware of any 11 study addressing that question. 12 Q What about 13 benzodiazepine? 14 A Yes, they were -- it was 15 very difficult for Upjohn to do this, but Upjohn 16 and Roche I think did in fact distinguish between 17 the sedating effect and the anxiolytic effect. 18 Q All right. And so I 19 simply point this -- give you this document to 20 point out that a sleeping pill and anxi -- 21 A Anxiolytic. 22 Q Anxiolytic was 23 administered in many of the clinical trials, was 24 it not? 575 1 A Yes, sir, at least those 2 listed here. 3 Q And that's what is 4 suggested in the German package insert, isn't it? 5 A Well, the same class of 6 drug is suggested, yes. 7 Q All right. But -- so do 8 you still maintain an objection to this language 9 being in the German package insert? 10 A I would not have put it 11 in the insert myself, and I'll be glad to explain 12 that more fully if you'd like me to. 13 Q Well, would you have 14 given -- would you -- do you have a problem with 15 the individuals in Germany getting sedatives when 16 they have problems on Prozac with extreme sleep 17 disturbances or excitability? 18 A Since I don't know the -- 19 all the diagnoses in those patients, and I'm 20 certainly not an expert in the practice of 21 medicine in Germany, I have difficulty saying that 22 I've got a problem with it. Your original 23 question asked do I have a problem with it being 24 in the label and what I've said is I wouldn't have 576 1 put it in the label. 2 Q Would you have put it 3 somewhere else or would you advise physicians of 4 this practice in some other manner? 5 A No. 6 Q Well, this package insert 7 is for depression? 8 A Yes, sir. 9 Q Prozac was approved 10 only -- has been approved only in Germany for 11 treatment of depression, isn't it? 12 A I don't know if they have 13 approved the other indications now or not. 14 Q Page 4 of that document 15 indicates that in addition to other adverse 16 experiences or side effects, although Fluctin may 17 not have been the cause, there have been reports 18 of suicidal thoughts and aggressive behavior, 19 correct? 20 A Let me catch up with 21 where you are because I haven't found that exact 22 statement. It sounded to me like you were 23 beginning to read from the paragraph that begins 24 in addition, is that right? 577 1 Q Yes. 2 A Well, then you jumped to 3 suicidal thoughts -- oh, I'm sorry, that's at the 4 very bottom of the next paragraph. 5 Q I said in addition, the 6 package insert states in addition there have been 7 reports, although Fluctin may not have been the 8 cause of suicidal thoughts and aggressive 9 behavior, that is contained within this package 10 insert? 11 A Yes, sir. 12 Q Do you have any problem 13 with that language being in the German package 14 insert? 15 A No, sir. 16 MR. SMITH: I better 17 quit, or Joe's not going to come back, for today. 18 (DEPOSITION ADJOURNED.) 19 * * * * * 20 21 22 23 24 578