Document VJX4nKJdxK3kK2EG9XBp9JR9w

1 PAGES 1 - 78 2 3 UNITED STATES DISTRICT COURT 4 FOR THE DISTRICT OF MASSACHUSETTS 5 6 ************** Brian K. Milward and * 7 Linda J. Milward * * 8 v. * Civil Action * No. 1:07-cv-11944-GAO 9 Acuity Specialty Products * Group, Inc., et al. * 10 * ************** 11 12 13 14 15 Deposition of David H. Garabrant, M.D., M.P.H. 16 Wednesday, February 18, 2009 17 Nixon Peabody LLP 18 100 Summer Street 19 Boston, Massachusetts 02110 20 21 22 23 ----------- J. Edward Varallo, RMR, CRR --------- 24 Registered Professional Reporter HG LITIGATION SERVICES ~ DALLAS, TEXAS 25 1-888-656-3376 1 1 Present for Plaintiffs: Steve Baughman Jensen, Esq. 2 Allen Stewart, P.C. Republic Center 3 325 North St. Paul Street - Suite 2750 Dallas, Texas 75201 4 214.965.8700 ~ Fax 214.965.8701 sjensen@allenstewart.com 5 6 Present for Radiator Specialty Company: 7 Brian R. Birke, Esq. Adler Pollock & Sheehan, P.C. 8 175 Federal Street - 10th Floor Boston, Massachusetts 02110 9 617.482.0600 ~ Fax 617.482.0604 bbirke@apslaw.com 10 11 Present for The Clorox Company: 12 Charles K. Mone, Esq. Campbell Campbell Edwards & Conroy, P.C. 13 One Constitution Center - 3rd Floor Boston, Massachusetts 02129 14 617.241.3000 ~ Fax 617.241.5115 cmone@campbell-trial-lawyers.com 15 16 Present via telephone for Berryman Products, Inc.: 17 Brian P. Voke, Esq. Campbell Campbell Edwards & Conroy, P.C. 18 One Constitution Center - 3rd Floor Boston, Massachusetts 02129 19 617.241.3000 ~ Fax 617.241.5115 bvoke@campbell-trial-lawyers.com 20 21 Present for The Sherwin-Williams Company; Henkel Corporation: 22 Kevin R. Powers, Esq. Cetrulo & Capone LLP 23 World Trade Center East Two Seaport Lane - 10th Floor 24 Boston, Massachusetts 02210 617.217.5500 ~ Fax 617.217.5200 25 kpowers@cetcap.com 2 1 Present for WD-40 Corporation: Christopher G. Betke, Esq. 2 Coughlin Betke, LLP 175 Federal Street - 9th Floor 3 Boston, Massachusetts 02110 617.988.8050 ~ Fax 617.988.8005 4 cbetke@coughlinbetke.com 5 6 Present for The Sherwin-Williams Company: James R. Miller, Esq. 7 Dickie McCamey & Chilcote, P.C. Two PPG Place - Suite 400 8 Pittsburgh, Pennsylvania 15222-5402 412.281.7272 ~ Fax 412.392.5367 9 jmiller@dmclaw.com 10 11 Present for United States Steel Corporation; Aristech Chemical Corporation; USX Corporation: 12 Tim Gray, Esq. Forman Perry Watkins Krutz & Tardy, LLP 13 City Centre Building - Suite 100 200 South Lamar Street 14 Jackson, Mississippi 39201 601.960.8600 ~ Fax 601.960.8613 15 tgray@fpwk.com 16 17 Present via telephone for Sunnyside Corporation: Leslie A. Sheehan, Esq. 18 Gordon & Rees LLP 633 West Fifth Street - Suite 4900 19 Los Angeles, California 90071 213.576.5000 ~ Fax 213.680.4470 20 lsheehan@gordonrees.com 21 22 Present for Rust-Oleum Corporation: Matthew P. Zayotti, Esq. 23 Keegan Werlin LLP 265 Franklin Street - 6th Floor 24 Boston, Massachusetts 02110 617.951.1400 ~ Fax 617.951.1354 25 mzayotti@keeganwerlin.com 3 1 Present for Nu-Calgon Wholesaler, Inc. Eileen P. Kavanagh, Esq. 2 Litchfield Cavo, LLP 6 Kimball Lane - Suite 200 3 Lynnfield, Massachusetts 01940-2682 781.309.1500 ~ Fax 781.246.0167 4 kavanagh@litchfieldcavo.com 5 6 Present for Boyle-Midway, Inc. William R. Covino, Esq. 7 McDonough Hacking & Lavoie, LLC One Washington Mall - 9th Floor 8 Boston, Massachusetts 02108 617.367.0808 ~ Fax 617.367.8307 9 wcovino@mhlattys.com 10 11 Present for Acuity Specialty Products Group, Inc.; The Steco Corporation: 12 Curtis L.S. Carpenter, Esq. Morrison Mahoney LLP 13 250 Summer Street Boston, Massachusetts 02210 14 617.439.7500 ~ Fax 617.439.7590 ccarpenter@morrisonmahoney.com 15 16 Present for CRC Industries, Inc.: 17 Joseph J. Leghorn, Esq. Nixon Peabody LLP 18 100 Summer Street - 25th Floor Boston, Massachusetts 02110 19 617.345.1000 ~ Fax 617.345.1300 jleghorn@nixonpeabody.com 20 21 Present for The Steco Corporation: 22 Christopher Sullivan, Esq. Sullivan & Collins, LLC 23 500 W. Cummings Park - Suite 4000 Woburn, Massachusetts 01801 24 781.939.5840 25 4 1 Present via telephone for American Grease Stick Company: 2 Syd A. Saloman, Esq. Tucker Heifetz & Saltzman, LLP 3 100 Franklin Street Boston, Massachusetts 02110 4 617.557.9696 ~ Fax 617.227.9191 saloman@ths-law.com 5 6 Present for NCH Corporation (Chemsearch Division): 7 Wesley A. Weathers, Esq. Weathers Riley & Sheppeard, LLP 8 4848 S.W. 21st Street - Suite 202 Topeka, Kansas 66604 9 785.273.2020 ~ Fax 785.273.4662 wes@wrslaw.net 10 11 Also Present: 12 Vic Houston Henry, Esq. Henry Oddo Austin & Fletcher, P.C. 13 1700 Pacific Avenue - Suite 2700 Dallas, Texas 75201 14 214.658.1900 ~ Fax 214.658.1919 vhhenry@hoaf.com 15 16 17 18 19 20 21 22 23 24 25 5 1 INDEX 2 ---------------------------------------------------- 3 DEPONENT PAGE 4 ---------------------------------------------------- 5 David H. Garabrant, M.D., M.P.H. 6 by Mr. Jensen................................. 7 7 8 ---------------------------------------------------- 9 GARABRANT EXHIBITS FOR IDENTIFICATION PAGE 10 ---------------------------------------------------- 11 1 Document headed Legal Testimony by David 7 H. Garabrant, MD, MPH since November 2004 12 (2 pages) 13 2 Dr. Garabrant's report, dated November 19 19, 2008 (7 pages) 14 3 Dr. Martyn Smith's supplemental 21 15 declaration (SMITH000694 - 715) 16 4 Collection of journal articles supplied 70 by Dr. Garabrant 17 18 (ORIGINAL EXHIBITS DELIVERED TO ATTORNEY JENSEN) 19 20 21 22 23 24 25 6 1 ---------------------------------------------------2 MORNING SESSION 3 9:17 a.m. 4 ---------------------------------------------------5 DAVID H. GARABRANT, M.D., M.P.H., 6 having been first duly sworn on oath, 7 was examined and testified as follows: 8 EXAMINATION 9 BY MR. JENSEN: 10 Q. Good morning, Dr. Garabrant. 11 A. Good morning. 12 Q. We were introduced earlier this morning, 13 but I'm Steve Jensen and I'm here on behalf of the 14 plaintiffs today to ask you some questions. 15 And I apologize to everyone. I have not 16 brought extra copies of my materials this morning, 17 but I don't believe I'm going to mark anything that 18 isn't part of reliance materials or reports, so 19 everyone should have copies available to them. 20 Marking as Exhibit 1 two pages that I 21 believe is a list of your legal testimony since 22 November 2004. Take a look at that and see if 23 that's correct, Doctor. 24 A. I believe it is. 25 Q. And do you have your own copy of that or 7 1 do you need to look at it? 2 A. No, I do not. 3 Q. Okay. Taking a look at that list on 4 Exhibit 1, Dr. Garabrant, are there any of those 5 cases that are listed on Exhibit 1 in which you were 6 engaged to testify by plaintiff's counsel in a 7 personal-injury case? 8 A. I don't believe so. 9 Q. And approximately how many cases -- Strike 10 that. 11 Were there any of the cases that are 12 listed on Exhibit 1 which involved litigation other 13 than personal injury litigation, to your knowledge? 14 A. I'm not sure I know the definition of 15 personal injury litigation. Some of these are 16 workers' compensation cases, so I'm not sure how to 17 answer that. 18 Q. Fair enough. 19 In each of those cases, were you asked to 20 provide opinions as to the extent to which an 21 exposure to a particular toxic substance is capable 22 of causing a particular disease? 23 MR. LEGHORN: Object to the form. 24 A. I am not sure I understand that question. 25 As I hear it, the answer would have to be no. 8 1 Q. And why is it no? 2 A. Well, I'm not sure what you mean by an 3 exposure to a toxic substance is capable of causing 4 particular disease. That's the part that I'm not 5 sure what you mean. 6 Q. Okay. And how would you in your own words 7 describe in general terms the areas of testimony in 8 which you have provided opinions in the cases 9 described in Exhibit 1 or listed in Exhibit 1? 10 A. (Pause) Well, many of those cases have to 11 do with chemical exposures and whether the 12 plaintiffs' alleged exposures caused them to have 13 various illnesses. 14 Q. And were you -- Let me back up for a 15 second. If I use the term general causation, does 16 that have any significance to you? Do you 17 understand what that means? 18 A. I believe so. 19 Q. And if I use the term specific causation, 20 does that have some meaning for you? 21 A. I believe I understand that. 22 Q. Okay. In those cases listed in Exhibit 1, 23 were you always providing opinions going to general 24 causation? 25 A. I would say in most of these cases, yes. 9 1 I'm not sure that it was always true, but certainly 2 in most of them it was true. 3 Q. Were there also cases in which you 4 provided both a general causation opinion and a 5 specific causation opinion? 6 A. Yes. 7 Q. Can you recall looking at the list which 8 of those cases that was true for? 9 MR. LEGHORN: Object to form. 10 A. That I provided both general and specific 11 causation opinions? 12 Q. That's my question, yes. 13 A. Not from memory, no. 14 Q. From memory, is it possible for you to 15 look at this list and identify what toxin or toxins 16 or alleged toxin or toxins were at issue in each of 17 those cases? 18 A. I could do most of them from memory; not 19 all. 20 Q. In approximately how many of those cases 21 was the toxin at issue asbestos? 22 A. When you say the toxin at issue was 23 asbestos, are you referring to cases involving 24 friction products? 25 Q. I am referring to any case listed on 10 1 Exhibit 1 in which you provided opinions and the 2 allegation by the plaintiff was that they had become 3 ill with some disease as a result of asbestos 4 exposure. 5 A. Well, there are a substantial number of 6 cases here where the allegation was that exposure to 7 friction products has caused them to become ill, and 8 I would estimate that that's perhaps half of the 9 cases listed here. 10 Q. And am I correct that the allegation by 11 the plaintiff in those cases was that the friction 12 product contained some amount of asbestos to which 13 the plaintiff was exposed and which contributed to 14 their injury. Is that correct? 15 A. Those cases involved allegations of 16 chrysotile asbestos in friction products. 17 Q. Let me take a quick look at that list, 18 please. Thank you. 19 Is it true that some of the cases listed 20 on Exhibit 1 involved allegations of injury from 21 benzene exposure? 22 A. Yes. 23 Q. Can you identify those cases? 24 A. (Pause) I think the one listed as Garcia 25 v. Allied Diagnostic; Henricksen v. Chevron. And 11 1 the third one I'm not sure, Baker v. Chevron. I 2 can't recall whether the allegation included benzene 3 or not. I know it had to do with petroleum-based 4 fuels such as diesel fuel and gasoline, but I don't 5 remember whether there was an allegation of benzene 6 exposure. 7 Q. In the Garcia v. Allied Diagnostic case, 8 do you recall what the injury was of the plaintiff, 9 or injuries, that was at issue? 10 A. I believe Mr. Garcia had a non-Hodgkin's 11 lymphoma. 12 Q. And do you recall what the substance of 13 your opinions in the case were in Garcia? 14 A. Mr. Garcia was a printer and it was my 15 opinion that his work as a printer did not cause his 16 non-Hodgkin's lymphoma. 17 Q. And is it your opinion generally that 18 exposure to organic solvents is not capable of 19 causing non-Hodgkin's lymphoma? 20 MR. LEGHORN: Object to the form. 21 A. It is my opinion that there is not 22 sufficient scientific evidence to reach a conclusion 23 that exposure to solvents causes non-Hodgkin's 24 lymphoma. 25 PHONE VOICE: I'm having a little trouble 12 1 hearing Dr. Garabrant. 2 MR. LEGHORN: Try to keep your voice up. 3 MR. SULLIVAN: We're having difficulty 4 down here at the end of the table too, 5 Dr. Garabrant. 6 THE WITNESS: I will speak more loudly. 7 MR. SULLIVAN: Thank you. 8 BY MR. JENSEN: 9 Q. Is it also your opinion, Doctor, that the 10 scientific evidence is insufficient to conclude that 11 benzene specifically is capable of causing non12 Hodgkin's lymphoma? 13 MR. LEGHORN: Object to the form. 14 A. It's my opinion that the scientific 15 evidence is not adequate to support a conclusion 16 that benzene causes non-Hodgkin's lymphoma. 17 Q. And were both of those opinions part of 18 the opinions you delivered in the Garcia case? 19 A. I'm not sure I recall exactly what my 20 opinions were in Garcia, but I think that's roughly 21 correct. 22 Q. Turning to the Henricksen versus Chevron 23 case, do you recall what the injury alleged in 24 Henricksen was? 25 A. I am not sure, but I believe it was acute 13 1 myelogenous leukemia. I could be wrong, though. 2 From memory I'm not sure. 3 Q. All right. Do you recall what the 4 substance of your opinions in that case were? 5 A. Mr. Henricksen worked as a fuel truck 6 driver, to the best of my recollection, and it was 7 my opinion that his exposure to motor fuels, 8 including gasoline and diesel fuel, did not cause 9 his AML. 10 Q. And was your opinion in that regard based 11 on the lack of evidence of sufficient exposure? 12 A. It was based on the scientific evidence 13 that fails to show that either gasoline or diesel 14 fuel is causally associated with increased risk of 15 AML. 16 Q. So am I correct in understanding that you 17 were providing a general causation opinion in that 18 case? Is that correct? 19 A. Yes. And I believe I also provided a 20 specific causation opinion. 21 Q. And with respect to your general causation 22 opinion, your general causation opinion was tied to 23 the specific kinds of products and materials that 24 Mr. Henricksen alleged exposure to. Correct? 25 A. Yes, my opinion was tied to the alleged 14 1 exposures of the plaintiff. 2 Q. Did the plaintiff allege that the benzene 3 in the gasoline was the cause of his injury? 4 A. I don't recall offhand. 5 Q. Do you recall what type of AML 6 Mr. Henricksen had? 7 A. No. 8 Q. Do you recall what injuries were at issue 9 in Baker versus Chevron? 10 A. There were four plaintiffs in the Baker 11 versus Chevron case. They suffered from monoclonal 12 gammopathy of unknown significance, or MGUS; a 13 chronic lymphadenopathy that persisted for years but 14 in which no cancer diagnosis was reached; AML that 15 was successfully treated, I believe. And I take 16 that back. I don't recall if there was an AML. 17 There was a case of Hodgkin's disease that was 18 successfully treated with radiation therapy and the 19 plaintiff subsequently developed breast cancer in 20 the radiation field; and one of the plaintiffs also 21 had a monoclonal gammopathy. 22 Q. So there were two monoclonal gammopathies 23 in the group of four? 24 A. That's correct. So one of the plaintiffs 25 had two diagnoses. Actually, and another one of 15 1 them had a diagnosis of Hodgkin's disease and then 2 subsequent breast cancer. 3 Q. Do you recall what substances the 4 plaintiffs alleged were causing these various 5 diseases? 6 A. The allegations revolved around a refinery 7 in which alleged contamination of groundwater under 8 the neighborhood where these plaintiffs lived had 9 caused their disease. 10 Q. Where was that case? 11 A. In southern Ohio. I'm blocking out the 12 name of the town. Hooven, Ohio. 13 Q. And Exhibit 1 indicates that you testified 14 both in a deposition and in a Daubert hearing in 15 that case. Is that correct? 16 A. Yes. 17 Q. Do you know what the outcome of the 18 Daubert hearing was? 19 A. I do not. 20 Q. As far as you can recall, do any of the 21 cases listed in Exhibit 1 involve allegations of APL 22 from exposure to benzene or benzene-containing 23 materials? 24 A. Not to my recollection. 25 Q. You have also provided testimony in legal 16 1 cases before November of 2004. Is that correct? 2 A. Yes. 3 Q. As best you can recall, have any cases in 4 which you have ever testified involved allegations 5 of acute promyelocytic leukemia from exposure to 6 benzene or benzene-containing materials? 7 A. I don't believe so. 8 Q. Do you recall when was the first time you 9 ever provided legal testimony in litigation? 10 A. I believe it was in the mid 1980s. 11 Q. Have you ever provided testimony at the 12 request of a plaintiff's counsel in a personal13 injury case? 14 A. Yes. 15 Q. Do you know how many times? 16 A. Offhand I do not know. 17 Q. And what kinds of cases, if you can 18 recall? 19 A. Well, the first time I gave testimony in 20 the mid 1980s had to do with a plaintiff who had 21 colon cancer in which he alleged it was caused by 22 asbestos exposure and I testified on behalf of the 23 plaintiff. 24 Q. Any other examples of testifying on behalf 25 of the plaintiff in a personal-injury case that you 17 1 can recall? 2 A. There have been numerous cases where 3 people have alleged injury from various chemicals 4 causing asthma, chronic dermatitis and other 5 conditions where I've testified on their behalf. 6 Q. Aside from the example of a plaintiff 7 alleging that his asbestos exposure was related to 8 his colon cancer, can you recall other examples of 9 cases in which you've testified on behalf of a 10 plaintiff who alleged that a chemical caused him to 11 develop cancer? Or her. 12 A. (Pause) There may well have been. I 13 can't recall any from memory, though. 14 Q. All right. Is it fair to say that of the 15 cases listed on Exhibit 1, that the vast majority if 16 not all of them involve allegations by a plaintiff 17 that they have developed cancer as a result of one 18 or more chemical exposures? 19 A. No. 20 Q. That's not fair to say? 21 A. No. 22 Q. Please identify for me the cases listed on 23 Exhibit 1 which involved injuries other than cancer 24 and, to the best of your recollection, what those 25 allegations of injury were. 18 1 A. I'm not sure I can do that for each and 2 every case. I will do my best, however. 3 Essenmacher versus Orkin Exterminating 4 Company, I believe the plaintiff alleged multiple 5 chemical sensitivity from pesticides. 6 Jerry Ward versus Albion Personnel 7 Services was not a cancer case. I don't remember 8 what Mr. Ward had. 9 MDL Docket 1535, Welding Rod Products, had 10 to do with allegations that welding caused 11 Parkinson's disease. 12 Presler versus Lincoln Electric involved 13 allegations that welding caused Parkinson's disease. 14 Karen Brown versus Christus Spohn Health 15 System had to do with a woman who received a barium 16 enema and they perforated her colon and she got 17 barium in her peritoneum and alleged that she had 18 barium poisoning. 19 I could go on, but that's out of the first 20 maybe eight or nine, about half of them had nothing 21 to do with cancer, and I think that that pattern is 22 true probably throughout the entire exhibit. 23 Q. I have put an exhibit sticker marking as 24 Exhibit 2 what I believe is a copy of your report in 25 this case, Dr. Garabrant, if you would verify that 19 1 that's true. 2 A. Yes. 3 Q. Have you done additional work in this case 4 since you prepared the report marked as Exhibit 2? 5 A. Yes. 6 Q. What additional work have you done? 7 A. I have reviewed Dr. Martyn Smith's 8 supplemental declaration. I have reviewed 9 additional scientific literature. I have reviewed 10 the declarations prepared by Dr. Pyatt and 11 Dr. Bennett, and I have reviewed Dr. Martyn Smith's 12 deposition testimony. 13 Q. Have you formed additional opinions that 14 are not reflected in Exhibit 2? 15 MR. SULLIVAN: Excuse me, counsel. Could 16 you just identify Exhibit 2 by the date of the 17 report? 18 MR. JENSEN: It's dated November 19, 2008. 19 MR. SULLIVAN: Thank you. 20 A. I have formed opinions regarding 21 statements made by Dr. Smith in his supplemental 22 declaration and in his deposition testimony. 23 Q. What opinions are those? 24 A. We would have to go through his 25 declaration and his deposition in order to identify 20 1 all of them. 2 Q. Dr. Garabrant, have you brought with you 3 today copies of Dr. Smith's original declaration and 4 supplemental declaration? 5 A. Yes. 6 Q. At the time you prepared Exhibit 2, you 7 had already reviewed Dr. Smith's original 8 declaration. Is that correct? 9 A. I believe I had. 10 Q. Do you have any opinion today relating to 11 Dr. Smith's original declaration that you did not 12 express in your report marked as Exhibit 2? 13 A. I believe Exhibit 2 contains all of my 14 opinions regarding Dr. Smith's original declaration. 15 Q. All right. If you could put in front of 16 you Dr. Smith's supplemental declaration, please. 17 And for the record, I will mark a copy of that as 18 Exhibit 3. And let me verify that Exhibit 3 is 19 indeed the same thing as what we're talking about, 20 Dr. Smith's supplemental declaration. 21 A. Yes, I believe that to be the supplemental 22 declaration. 23 Q. Okay, thank you. 24 Can you tell me what opinions you have 25 developed relating to Dr. Smith's supplemental 21 1 declaration that are not reflected in your report 2 which is Exhibit 2? 3 A. Well, okay, I'll do my best. The first is 4 at Smith's paragraph 2 where he claims defense 5 experts make the error that because of this unique 6 clinical standing, it is also a unique etiological 7 or causal entity, and there he's referring to APL. 8 I don't think that I made any such error. There is 9 simply good scientific evidence that the etiology of 10 APL has nothing to do with benzene, which makes it 11 different than some of the other forms of AML. 12 At paragraph 6 Dr. Smith says "Thus, 13 common environmental causes are likely for these 14 leukemias and may include low-level benzene 15 exposure, quinone compounds in our diet, ionizing 16 radiation and chemicals producing oxidative stress." 17 I believe he is simply wrong on that with respect to 18 APL. There is not evidence that benzene is a risk 19 factor for APL. 20 Dr. Smith reiterates a similar view at the 21 top of the next page, and I would again respond with 22 the point I just made. 23 At paragraph 7 Dr. Smith comments "If APL 24 were really so different from AML, why is it not 25 listed as a separate cause of death on death 22 1 certificates and why are there no epidemiological 2 studies which focus only on APL and benzene 3 exposure?" He then goes on to say "The problem is 4 that death certificates and case referrals in 5 epidemiology studies consider AML as one disease or 6 in our analogy as a car, independent of the color." 7 Dr. Smith seems to have forgotten that he 8 has identified a substantial body of epidemiological 9 literature which looks specifically at APL and 10 benzene and which fails to show any association, so 11 it's odd that he would question the existence of 12 that literature when he has reviewed it. 13 Secondly, death certificates in 14 epidemiology do not always consider AML as one 15 disease. There are many studies which consider 16 different forms of AML and Dr. Smith has done some 17 work to identify those studies. 18 At paragraph 8 he comments that "Because 19 APL is seen in studies of workers exposed to benzene 20 where the subtypes of AML have been separately 21 analyzed," I would comment that he has just 22 contradicted his previous point that there are no 23 studies that have done that, and he goes on to say, 24 quote, "and has been found at higher levels than 25 expected with an odds ratio that exceeds 2." There 23 1 are no studies that show an odds ratio that exceeds 2 2 for APL and benzene, and Dr. Smith has failed to 3 show any. 4 At paragraph 12 Dr. Smith comments on the 5 study by Golomb, G-o-l-o-m-b, and having reviewed 6 that myself, Dr. Smith has not counted up the number 7 of cases properly and has not calculated the odds 8 ratios properly, and there is no incidence of APL 9 that is approximately two times higher in those 10 patients with probable benzene exposure than in 11 unexposed patients, as Dr. Smith claims. 12 At paragraph 14 Dr. Smith says he has used 13 this information to calculate an odds ratio for APL 14 in the Chinese cohort study that is in excess of 2. 15 And I believe he has made errors in the assumptions 16 behind his calculations as well as in his methods of 17 calculating. 18 At paragraph 16 Dr. Smith comments 19 regarding my critique, quote, "Overall, his critique 20 could have some validity if APL were a separate 21 disease, which it is not." I believe Dr. Smith in 22 paragraph 2 indicated that, quote, "I do not dispute 23 that APL is a unique therapeutic entity among the 24 myeloid leukemias," so I am perplexed by his 25 contradiction of his earlier paragraph. 24 1 He then goes on to say at paragraph 16, 2 quote, "My point is that we know benzene causes AML 3 and that the Travis study shows that a significant 4 proportion of these AMLs, four of nine cases 5 analyzed, see below, are APL cases." I disagree 6 that we know that four of the nine are APLs because 7 of deficiencies in the diagnostic information in the 8 Travis report. 9 At paragraph 19 Dr. Smith attributes a 10 statement to me at lines 3 through 7 that I don't 11 believe I ever made. He then goes on to make an 12 error in claiming that the proportion of acute 13 leukemias that were APL is 28.6 percent. That's not 14 correct and the literature that Dr. Smith claims to 15 have read shows it's not correct, so I don't know 16 how he arrived at that conclusion. 17 He then calculates an odds ratio -- I'm 18 still at paragraph 19, at the top of page Smith 19 000704 -- he claims he calculated an odds ratio. In 20 fact he did not calculate an odds ratio; he is 21 wrong. His calculation is not based on any reliable 22 methodology. Moreover, he did not calculate any 23 test of statistical significance or any confidence 24 interval regarding his measure of association, which 25 he incorrectly calls an odds ratio, and therefore he 25 1 has no basis for saying at the bottom of paragraph 2 19, quote, "Thus, one can say to a reasonable degree 3 of scientific probability that the odds ratio for 4 APL associated with benzene exposure in the Chinese 5 cohort is more likely than not greater than 2." 6 At paragraph 20 Dr. Smith makes a comment 7 that my calculations have no validity because the 8 proportion of APL among AML cases in China appears 9 to be much higher than among AML cases in the U.S.; 10 and two, the Chinese cohort is a group of healthy 11 workers that is not representative of the general 12 population, making direct comparisons to other 13 population incidence untenable. He is simply wrong. 14 There's no basis for those claims. 15 At paragraph 21 Dr. Smith claims that both 16 Dr. Bennett and I ignore the difficulties of 17 performing research in China in the time period 18 before 1987. He's wrong about that; I don't ignore 19 that. I simply pointed out that the diagnoses of 20 APL were not reliable and were not based on the 21 diagnostic materials that are essential in order to 22 reach those diagnoses. 23 Dr. Smith then goes on to address a number 24 of responses to the declaration by Dr. David Pyatt. 25 Although I do not agree with a number of things he 26 1 has said there, I will not address those since those 2 are really directed at Dr. Pyatt's opinions. 3 Q. Starting with your opinion in regard to 4 paragraph 2 of Exhibit 3, Dr. Smith's supplemental 5 declaration at paragraph 2, as I recall, you told me 6 that your opinion is that there is good evidence 7 that the etiology of APL has nothing to do with 8 benzene exposure. Is that correct? 9 A. I don't think that's what I've said. 10 Q. Okay. The record will reflect what you've 11 said. Here's my question about what you've said. 12 Are you basing your opinion in regard to paragraph 2 13 on any evidence other than epidemiological studies? 14 A. Yes. 15 Q. What evidence besides epidemiological 16 studies? 17 A. Well, first let me comment on the role of 18 the epidemiological studies. The epidemiological 19 studies provide a direct test of the hypothesis that 20 APL is causally associated with benzene and they 21 fail to show that association, and so they provide 22 the strongest evidence in response to Dr. Smith's 23 allegation that APL is not a unique etiological or 24 causal entity with respect to benzene. 25 Secondly, I rely on more general concepts 27 1 derived from my education, training and experience 2 in medicine that goes to the issue of how do we 3 actually differentiate one disease from another. In 4 other words, why is rheumatoid arthritis different 5 from systemic lupus erythematosus? Why is 6 rheumatoid arthritis different from osteoarthritis? 7 Why is tuberculosis different from other mycoplasmal 8 diseases? 9 The history of medicine provides a firm 10 foundation for the overall process of 11 differentiating diseases based on seeing that they 12 have different clinical presentations, different 13 clinical courses, and different treatments, as well 14 as different etiologies. And so over the history of 15 medicine we have split broad diagnoses into more 16 clearly defined and differentiated diseases. At 17 times we have taken diseases that appeared to be 18 different and then realized that they were different 19 manifestations of the same underlying disorder, and 20 so on. 21 What I believe is clear for APL, and 22 particularly since cytogenetics and 23 immunophenotyping has been brought into widespread 24 use in the diagnosis of leukemias, is that not only 25 is it a distinct clinical entity, but to the extent 28 1 that we have been able to examine the association of 2 that distinct entity with various putative causes, 3 it has different causes than some of the other forms 4 of AML. 5 Q. Will you be providing any opinions in this 6 case dealing with the cellular origins for APL or 7 any other form of leukemia? 8 MR. LEGHORN: Object to the form. 9 A. No. 10 Q. Will you be providing any opinions in this 11 case relating to the natural history of how APL 12 develops? 13 MR. LEGHORN: Object to the form. 14 A. I'm not sure I know exactly what you mean 15 by that question. If that question includes how it 16 develops with respect to etiology, yes. 17 Q. Will you be providing any testimony in 18 this case relating to what is going on at the 19 cellular, subcellular, genetic level that is causing 20 APL to develop? 21 MR. LEGHORN: Object to form. 22 A. I have not been asked to cover that area 23 and at the moment I have no plans to testify in that 24 area. 25 Q. Moving on to paragraph 6 of the 29 1 supplemental declaration of Dr. Smith, which is 2 Exhibit 3, the first sentence of paragraph 6 says 3 "Common genetic variants in a gene called NQ01 for 4 NAD(P)H:quinone oxidoreductase, one, confer 5 susceptibility to AML and two studies in England and 6 China show that this is strongest for de novo AML 7 cases harboring translocations and inversions, 8 including t(15;17) in APL." Do you have any 9 opinions as to whether that sentence is correct? 10 A. I do not have any opinions regarding 11 whether that sentence is correct. 12 Q. I won't read the second sentence but I'll 13 ask you the same question about the second sentence. 14 Do you have any opinions as to whether the second 15 sentence of paragraph 6 is correct? 16 A. I do not. 17 Q. Am I correct, Dr. Garabrant, that your 18 opinions regarding paragraph 6 of Dr. Smith's 19 opinion are really relating to the third sentence of 20 the paragraph, which says "Thus, common 21 environmental causes are likely for these leukemias 22 and may include low-level benzene exposure" and he 23 goes on from there. Is that right, your opinions 24 are addressed at that sentence? 25 A. Well, my opinions are addressed at the 30 1 hypothesis that Dr. Smith has that since he has a 2 possible mechanism by which environmental causes 3 might cause APL, he then makes a conclusion that 4 common environmental causes are likely. And I think 5 that his conclusion is not borne out by the 6 available evidence. He's got a hypothesis that does 7 not have proof. 8 Q. The fourth sentence of paragraph 6, 9 Dr. Garabrant, says "Further, in a recent 10 epidemiological study of childhood AML, we separated 11 the, quote, good prognosis, close quote, AMLs with 12 translocations and inversions from other forms of 13 AML and found a strong association between parental 14 exposure to solvents and increased risk of these 15 specific forms." Then he gives a citation. Up to 16 that point in the sentence, do you have any opinions 17 as to whether what I've read so far is correct? 18 A. I don't have an opinion regarding that 19 sentence. This litigation is not about childhood 20 AML and so I don't know the relevance of Dr. Smith's 21 reference to that. 22 Q. Well, the rest of the sentence after the 23 citation says "Again suggesting a common cause for 24 APL and other AMLs harboring translocations and 25 inversions." I assume you disagree with that 31 1 portion of the sentence. Am I correct? 2 A. Well, again Dr. Smith is talking about 3 childhood AML. This is not a case about childhood 4 AML. The childhood leukemias are different than the 5 adult leukemias in many ways, and I do not know the 6 literature on childhood AML and parental exposure to 7 solvents in depth and I didn't review it with 8 respect to this case, and so I can't comment on the 9 validity of his claim without having actually 10 explored that area in detail. All I can say is I 11 think he is invoking something of marginal or no 12 relevance to the case at hand. 13 Q. And your view that the evidence he is 14 invoking is of marginal or no relevance is based on 15 the fact that it in this particular instance comes 16 from a study relating to childhood AML which you 17 believe is of marginal or no relevance to an adult 18 AML. Is that correct? 19 A. Well, there is a large literature on 20 childhood leukemia looking at parental exposures to 21 solvents, benzene, pesticides, incense, parental 22 occupations, and Dr. Smith is referring to a single 23 study pulled out of that large body of literature 24 and reaching a conclusion based on that single 25 study. And my comment is if we're going to talk 32 1 about childhood AML, that is a very large area that 2 needs to be examined rigorously and systematically. 3 You cannot reach a valid conclusion by citing to a 4 single study that has a finding that you like, 5 without considering all of the rest of the body of 6 literature. 7 This case is not about childhood AML and 8 neither Dr. Smith nor I have done a rigorous 9 examination of that large body of literature, and so 10 I think it is irrelevant to simply pick a single 11 study and to say "See, this supports me." 12 Q. Do you have any knowledge as you sit here 13 today, Dr. Garabrant, as to how many other studies, 14 if any, have looked at comparing translocations and 15 inversions of AMLs that have the so-called good 16 prognosis with other forms of AML in childhood AMLs? 17 A. Again, as I have already stated, I have 18 not reviewed the literature on childhood AML in 19 preparation for this case. I think it is of 20 marginal or no relevance to the case, but if 21 Dr. Smith wants to bring that up, I'd be happy to 22 review it and provide opinions regarding childhood 23 AML and what we know about its etiology. Dr. Smith 24 has not done that. He has cited to a single study. 25 Q. Turning to paragraph 12 of Dr. Smith's 33 1 supplemental declaration, you indicated that it was 2 your opinion that Dr. Smith had not correctly 3 counted the cases from the Golomb study. Is that 4 right? 5 A. I believe so. 6 Q. How do you believe the Golomb cases should 7 be counted? 8 A. (Pause) In the 58 unexposed patients in 9 the Golomb study, there are five patients who have 10 APL. Dr. Smith counts four and he then uses that 11 erroneous count to calculate an odds ratio of 2.1. 12 So he's counted wrong. 13 Q. Other than the four should have been a 14 five, are there any other errors in the calculation 15 that Dr. Smith made that is reflected in paragraph 16 12, in your opinion? 17 A. From memory, I don't know exactly how he 18 made his calculation, so I would have to repeat that 19 calculation in order to verify whether he has done 20 his arithmetic properly. 21 Q. Well, without pulling out a calculator to 22 do the arithmetic, what I'm asking you is whether 23 you have any methodological criticism of the way 24 that that calculation was done as it is described in 25 paragraph 12? 34 1 A. Again, without working back through what 2 he has described, I can't comment. I'd be happy to 3 do that, but I can't do that without sitting down 4 and working back through his logic and doing the 5 arithmetic. 6 Q. So it is fair to say as you sit here today 7 you have not developed any opinions beyond what 8 you've already told me regarding any errors that 9 Dr. Smith may have made in paragraph 12. Is that 10 correct? 11 MR. LEGHORN: Object to the form. 12 A. Well, no. I've already told you that he's 13 counted wrong. There are five cases; he only 14 counted four of them, and therefore his calculation 15 is in error. 16 Q. My question intended to encompass that 17 criticism, Doctor. But besides the counting four 18 cases when he should have counted five, as you sit 19 here today you have identified no other 20 methodological errors that Dr. Smith made in his 21 calculations at paragraph 12. Is that correct? 22 MR. LEGHORN: Object to the form. 23 A. No. In fact, to the extent I understand 24 what he's done, he chose to exclude one of those 25 five cases because there were not confirmed or I 35 1 should say because cytogenetic testing was not done 2 or was not reported in the Golomb paper. He is 3 inconsistent in his methods when he comes to the 4 Travis paper where he specifically includes cases 5 that have no confirmed cytogenetic abnormalities. 6 So in Golomb he excludes the one where there's no 7 confirmed cytogenetic abnormality; in Travis he 8 specifically includes them. That is a methodologic 9 inconsistency that is inconsistent with rigorous and 10 carefully done science. 11 Q. Again, that criticism goes to how many APL 12 cases are reflected in Golomb that should have been 13 counted in his calculation. Is that right? 14 MR. LEGHORN: Object to the form. 15 A. No. It goes to his methods and it goes to 16 the reliability of his thinking. 17 Q. Besides the number of APLs that are used 18 in the calculation in paragraph 12, as you sit here 19 today you have no other criticisms related to that 20 calculation. Is that correct? 21 MR. LEGHORN: Object to the form. It has 22 been asked and answered. 23 A. As I have already answered, in order to 24 determine whether I have other criticisms I would 25 need to check his arithmetic. 36 1 Q. Which you have not done as we sit here. 2 Correct? 3 A. I have not done it as I sit here. I did 4 work through his arithmetic previously, but I don't 5 have it and I don't recall whether I found any other 6 errors. 7 I should also mention that Dr. Smith fails 8 to provide any test of statistical significance or 9 any confidence interval with his calculated odds 10 ratio, and as a consequence of that, it is unclear 11 whether that odds ratio is reliable or not. Let me 12 rephrase that: It is unclear whether that odds 13 ratio is reliably different than 1.0 or not. 14 Q. Is it your view, Dr. Garabrant, that an 15 epidemiological study that shows an odds ratio or a 16 relative risk or a standardized mortality ratio or 17 whatever the equivalent measure of association is 18 for that type of study which is not statistically 19 significant but is above the odds ratio of relative 20 risk, whatever is above 1, that such a study 21 provides no evidence of an association? 22 MR. LEGHORN: Object to the form. 23 A. Well, the evidence of association from an 24 epidemiological study is whatever is calculated by 25 the SMR or odds ratio or whatever other measure of 37 1 association one calculates. The issue is whether 2 that measure of association is meaningfully 3 different from the value 1.0, which means there is 4 no association. 5 Q. Right. And my question for you is: If 6 the association calculated is not statistically 7 significant, is it your opinion that such a study, 8 regardless of whatever else its findings or data 9 show, provides no meaningful evidence of an 10 association? 11 A. It is my opinion that when you have 12 measures of association that are not statistically 13 significant, the data are reasonably compatible with 14 there being no association whatsoever. 15 Q. And if the data are reasonably compatible 16 with there being no association whatsoever, is it 17 your opinion that any such data should not be 18 considered as evidence of an association? 19 A. It's my opinion that good scientists 20 consider all reliable and relevant data, which is 21 what I think Dr. Smith systematically does not do in 22 his supplemental declaration. 23 Q. Let me pose a hypothetical to you, Doctor. 24 If there were ten epidemiological studies all 25 looking at the issue of whether benzene exposure may 38 1 cause APL and each of those studies found an 2 elevated relative risk above 2.0 but the confidence 3 interval for each of those studies encompassed 1.0, 4 so all ten of them come up with similar results, 5 none of them are statistically significant but all 6 of them have a relative risk that is calculated at 7 above 2.0. Would it be your opinion that the data 8 is insufficient to conclude that more likely than 9 not benzene exposure causes APL in that 10 circumstance? 11 MR. LEGHORN: Object to the form. 12 A. Your hypothetical isn't adequate to reach 13 any such conclusion one way or the other. 14 Q. I would add to the hypothetical that's the 15 only epidemiological evidence that was available 16 that was specific to both APL and benzene exposure. 17 MR. LEGHORN: Object to the form. 18 BY MR. JENSEN: 19 Q. Does that make any difference to you? 20 MR. LEGHORN: Object to the form. 21 A. No. The hypothetical still is not 22 sufficient to reach any meaningful conclusion one 23 way or the other. 24 Q. In your opinion, is it ever appropriate to 25 reach an opinion that a particular chemical causes a 39 1 particular disease in the absence of statistically 2 significant epidemiological evidence that supports 3 that proposition? 4 MR. LEGHORN: Object to the form. 5 A. It is my opinion that when epidemiological 6 evidence exists, you must consider it. 7 Q. That wasn't my question. Do you need me 8 to repeat the question? 9 A. No. If your question is whether you can 10 reach a conclusion of causation in the absence of 11 epidemiological evidence, my answer is that's not 12 the instance we're in. We have a case here where 13 there is a substantial body of epidemiological 14 evidence which must be considered. There are many 15 settings where we can conclude causation in the 16 absence of epidemiology, such as gunshot wounds, 17 such as falling out of airplanes when your parachute 18 doesn't open. You don't need epidemiology. But 19 when you have situations such as low-level chemical 20 exposures and the development of diseases years or 21 decades later and there is epidemiologic evidence, 22 you have to consider it and you have to consider it 23 fully. 24 Q. Is it your view, Dr. Garabrant, that the 25 available epidemiological evidence that exists with 40 1 respect to acute promyelocytic leukemia and benzene 2 exposure establishes that benzene exposure cannot 3 cause APL? 4 MR. LEGHORN: Object to the form. 5 A. As Dr. Smith himself pointed out, you 6 cannot prove the null hypothesis. I agree with that 7 statement. He got that one right. The 8 epidemiologic evidence that we have regarding 9 benzene and APL fails to prove the alternative 10 hypothesis. 11 Q. Am I correct, Dr. Garabrant, that because 12 in your view the epidemiological evidence with 13 respect to APL and benzene fails to prove the 14 alternative hypothesis, it is your view that it is 15 unnecessary to look to other bodies of evidence such 16 as biological mechanism evidence to determine the 17 question of whether benzene is capable of causing 18 APL? 19 MR. LEGHORN: Object to the form. 20 A. I am a strong adherent of Austin Bradford21 Hill's comments on causation. And consistent with 22 Hill's comments, I believe that one has to look for 23 evidence of an association when the purpose is to 24 determine whether an environmental exposure is 25 causally associated with a disease such as APL. One 41 1 has to look for evidence of an association and that 2 if one does not find evidence of an association, 3 then most of the other considerations that Hill 4 discusses are rendered moot. In other words, when 5 there's no association, you can't look for dose 6 response because there's no dose response when 7 there's no association. You really can't look at 8 specificity when there's no association. You can't 9 really judge coherence of the literature -- Well, I 10 guess you can when there's no association and when 11 you see a coherent picture of lack of association, 12 you can say "You know? There's really no 13 association." 14 So your comments about biological 15 plausibility I think go to the issue of, well, 16 should you look for biological plausibility when you 17 can't find evidence of an association? And my 18 comment to that would be I've spent my career as a 19 scientist, and smart scientists can come up with a 20 biologically plausible explanation for almost any 21 hypothesis. That's not hard. The issue is whether 22 you have empirical evidence that supports that the 23 hypothesis is correct in the first place. 24 Q. And in your view the only type of evidence 25 that counts as empirical evidence to support the 42 1 hypothesis with respect to environmental exposures 2 and latent disease are human epidemiological 3 studies. Is that correct? 4 A. No. My opinion is that when human 5 epidemiological studies exist, you can't ignore what 6 they may say, and that's the instance we're in here. 7 We have a substantial body of literature which 8 Dr. Smith has largely identified which is directly 9 relevant to whether there is an association between 10 benzene exposure and APL. You can't ignore it. 11 MR. JENSEN: Let's take a quick break. 12 MR. LEGHORN: I was going to say, before 13 we go to another paragraph, it would be a good time 14 for a break. 15 (In recess 10:38 a.m. to 10:51 a.m.) 16 BY MR. JENSEN: 17 Q. Turning to paragraph 19 of Dr. Smith's 18 supplemental declaration, I believe you testified 19 that it was your opinion that Dr. Smith's statements 20 about the proportion or assumptions about the 21 proportion of APLs in the Chinese population is 22 wrong. Am I correct that that was your opinion or 23 that is your opinion? 24 A. Let's read back exactly what I said about 25 that, because I don't think that's what I said. 43 1 Q. Well, I don't want to quibble with you 2 about what your opinion is, so I'll ask you: What 3 is your opinion with regard to the proportion of 4 APLs and what Dr. Smith has to say about that in 5 paragraph 19? 6 A. Well, my first opinion is that I can't see 7 any reason in the world why you would calculate 8 something based on proportions when the data is 9 available to actually compare incidence rates. So 10 Dr. Smith chose to do a calculation that has the 11 potential to have many sources of bias instead of 12 doing a calculation that would be much more 13 reliable. So, again, his fundamental approach is 14 just wrong. 15 Q. Do you have an opinion about whether his 16 statement in paragraph 19 -- it's the third 17 sentence, I believe -- that begins "Thus, four cases 18 of APL in fourteen cases of acute leukemia would be 19 equal to the general representation of APL in the 20 Chinese population, i.e., 4/14 equals 0.286 or 28.6 21 percent." Do you have an opinion about whether the 22 statement in that sentence is correct? 23 A. Well, it's wrong. We know from the Yang 24 paper that the correct proportion is not 28.6 25 percent. Now, that's a paper that Dr. Smith read. 44 1 So why would he choose to make an assumption that's 2 wrong when the actual proportion is in the paper? 3 That's the sort of incorrect methodology that we see 4 recurring in Dr. Smith's opinions. 5 Q. What is the proportion reflected in the 6 Yang paper of APLs to the total ANLL incidence in 7 China? 8 A. Well, the more important question is why 9 would you bother to use proportions when you can 10 actually use the actual incidence rates? Here's the 11 problem with proportions. When one disease is 12 overrepresented, the proportions of all others are 13 underrepresented; and conversely when one disease is 14 underrepresented, the proportions of all other 15 diseases are overrepresented. So proportions by 16 their very nature are a less reliable measure than 17 are actual incidence rates. 18 So I cannot think of a reason why any 19 reliable scientist would use proportions when he has 20 access to the underlying rates. It's just the wrong 21 thing to do. 22 Q. What is the proportion of APL in ANLL as 23 reported by the Yang paper? 24 A. 18.7 percent. 25 Q. Besides what you've already testified to, 45 1 do you have any other specific criticisms of 2 paragraph 19 and the approach that Dr. Smith took in 3 his paragraph 19? 4 A. Yes. One of the principles of science is 5 that a good scientist considers all relevant and 6 reliable data. Dr. Smith chose not to do that in 7 paragraph 19. So by choosing to make an erroneous 8 calculation based on proportions and ignoring the 9 far more reliable and relevant incidence rates, he 10 has not followed good scientific procedure. He has 11 ignored good data and chosen to do something that is 12 highly likely to give an unreliable answer, and in 13 fact the proportion he chose was based on a poorly 14 documented statement at the very end of the Travis 15 paper where Dr. Travis says the proportion of APL in 16 de novo ANLL was similar to that in the general 17 population. 18 So he took that and said, well, it's 19 really the same, it's 28.6, when it wasn't the same; 20 it was actually, in my opinion, not even very 21 similar. And my underlying opinion is I wouldn't 22 even use that approach because it's not a good 23 approach. The Yang paper gives tell you incidence 24 rates. You can calculate an incidence ratio. 25 Q. What is the healthy worker effect? 46 1 A. The healthy worker effect is a term that 2 really is an umbrella term for the commonly observed 3 phenomenon in the occupational epidemiology 4 literature coming from the United States and Western 5 Europe that in many cohort studies of working 6 populations, the workers are in general healthier 7 than the general population. 8 Q. Does the healthy worker effect, in your 9 opinion, play any role with respect to analysis of 10 the Travis et al. papers and series of studies by 11 Travis, Yin, Hayes et al.? 12 A. I think it deserves careful consideration. 13 Q. How did you consider the healthy worker 14 effect and what role did it play in your opinions in 15 this case? 16 MR. LEGHORN: Object to the form. 17 A. The Travis paper reports on hematopoietic 18 malignancies among benzene-exposed workers in China. 19 If we go back to some of the earlier reports from 20 that population, we find that the Yin paper 21 published in 1989 gives us the ability to evaluate 22 whether there is a healthy worker effect among 23 benzene-exposed workers in China. And what we find 24 is that the all-cause-of-death mortality rate in the 25 benzene-exposed workers is roughly double that in 47 1 the nonexposed workers. 2 So not only is there not a healthy worker 3 effect in the population under study in the Travis 4 paper, it goes the other way. It's a twofold 5 increase in risk. So if Dr. Smith were really 6 interested in adjusting for the healthy worker 7 effect, he's got it backwards. In fact, that 8 working population has double the all-causes 9 mortality of the general population. And if he's 10 going to adjust for it, then he should have adjusted 11 the other way. That's point number one. 12 So he has invoked something in a manner 13 where he's actually done it backwards. 14 Number two, to my recollection Dr. Smith 15 objected to using the incidence rates in the general 16 population of China derived from the Yang paper as 17 we've discussed and he chose to use the proportion 18 of APLs among AMLs instead of using the incidence 19 rates. And in part he did that because he said 20 there was a healthy worker effect in using the 21 population reported in the Yang paper. 22 His comment suggests that although he's 23 read the Yang paper, he didn't read it very 24 carefully, because many of the populations studied 25 in the Yang paper were working populations. They 48 1 came from the Daqing Oil Field, the Changchun 2 Automobile Company, the Yanju Forestry Farm, the 3 Baotou Steel Company, the Pingdingshan Coal Mine, 4 the Kailuan Coal Mine, the Fengfeng Coal Mine, 5 et cetera. 6 So in fact those rates to which he 7 objected using were largely derived from working 8 populations and so it would have been perfectly 9 appropriate to have compared the rates derived from 10 the Yang study, the cancer incidence rates, to the 11 cancer incidence in the Chinese benzene cohort. 12 Perfectly appropriate to do. 13 Q. Does the Yang paper indicate what 14 proportion of the people from which the incidence 15 rate for APL is calculated are workers as opposed to 16 nonworkers? 17 A. They didn't tabulate it or they didn't 18 give the summaries, but that could easily be derived 19 from Table 1. 20 Q. With respect to your statements regarding 21 the Yin 1989 paper, did I hear you correctly that 22 Yin was calculating all cancer mortality of the 23 exposed worker population by comparing it with an 24 unexposed worker population? 25 MR. LEGHORN: Object to the form. 49 1 A. That's not what I said. I was talking 2 about all causes of death. You were talking about 3 cancer. 4 Q. I'm sorry. But Yin did calculate all 5 causes of death in the exposed population by 6 comparing it with an unexposed population of 7 workers. Is that right? 8 A. Yes. 9 Q. So both populations were workers with 10 respect to the calculations made by Yin. Right? 11 A. That's correct. 12 Q. Isn't comparing an exposed worker 13 population to an unexposed worker population one of 14 the methods that epidemiologists commonly use to 15 avoid any impact of the healthy worker effect? 16 A. Yes. And that's why Dr. Smith should have 17 carefully looked at the Yin paper. He would have 18 realized that there's no healthy worker effect in 19 the benzene-exposed cohort. In fact, their all20 cause-of-death mortality is double that in the 21 unexposed population. 22 Q. Well, in order to measure the healthy 23 worker effect, if any, don't you have to compare a 24 worker population to a nonworker population? 25 A. If you want to measure the healthy worker 50 1 effect, yes, you have to. 2 Q. So there could be a number of different 3 reasons besides any healthy worker effect or lack 4 thereof for the mortality in the exposed workers 5 described in Yin to be higher than the unexposed 6 workers. Right? 7 A. Yes. And the most likely reason is far 8 better disease surveillance in the benzene-exposed 9 than in the unexposed. In other words, what this 10 really suggests is that there is systematic 11 difference in the way they're doing their 12 surveillance in the benzene-exposed population 13 compared to the unexposed. I mean, a doubling of 14 mortality is extraordinary. We never see that. 15 Never. It's absolutely extraordinary. 16 Q. Another contributor could be that the 17 exposures themselves are causing mortality. 18 Correct? 19 A. I think that's extremely unlikely here. 20 What exposures could you hypothesize that would 21 double mortality across all causes of death? I 22 can't think of any. There are no chemicals that 23 cause increases in cardiovascular disease, 24 cerebrovascular disease, renal disease, pulmonary 25 disease, all causes of cancer, diabetes, infectious 51 1 diseases, accidents, suicides, psychiatric disease, 2 neurologic disease. There aren't any. 3 Q. Exhibit 2, which is your report of 4 November 19, at page 3 -- Are you with me, Doctor? 5 A. Yes. 6 Q. Okay. The first full paragraph there on 7 page 3 discusses the Travis study and your analysis 8 of Dr. Smith's opinion of the Travis study. Is that 9 right? 10 A. Well, the first paragraph is what it is. 11 Q. And in that paragraph it says "If we use 12 Dr. Smith's estimate of three to eight cases of APL 13 per one million person-years," that estimate that 14 you're describing there is an estimate based on 15 incidence in the United States. Is that correct? 16 A. I don't recall where Dr. Smith got that 17 estimate, but that's his estimate. 18 Q. And you were comparing in that paragraph 19 an incidence rate that derives from a general 20 population with the incidence in the benzene-exposed 21 group discussed in Travis. Is that correct? 22 A. Well, I'm using Dr. Smith's own estimate 23 of the incidence rate to make an estimate of the 24 expected number of cases you'd see in Travis. 25 Q. And I'm asking you whether you even know 52 1 if that incidence rate that is estimated, you say, 2 by Dr. Smith at about three to eight cases of APL 3 per year, is that an incidence rate for a worker 4 population or for a general population? 5 A. I don't recall where Dr. Smith got that 6 number. To the best of my recollection, he never 7 said. Or, I should say, the range. 8 Q. So I take it, then, that it doesn't make 9 any difference to you whether or not the incidence 10 rate that Dr. Smith is using of three to eight cases 11 of APL per one million person-years is an incidence 12 rate derived from a general population or a worker 13 population for purposes of your analysis? 14 A. Well, it does matter. And in fact it is 15 entirely consistent with the incidence rate of APL 16 in what I was referring to as the Yang study, it's 17 actually Yang Chongli, in the Chinese Medical 18 Sciences journal. In that study, which is largely 19 derived from working populations in China, the 20 incidence rate is 3.5 per million person-years, so 21 it is entirely consistent with Dr. Smith's 22 estimates. 23 Q. Dr. Garabrant, have you ever published any 24 article, textbook chapter or any other publication 25 in which you've expressly discussed the issue of 53 1 whether benzene exposure is capable of causing acute 2 promyelocytic leukemia? 3 A. I don't believe so. 4 Q. Have you ever published a study of 5 benzene-exposed workers and leukemia risk? 6 A. I'd have to look at my C.V. I don't 7 recall any. I know my colleagues and I published a 8 study of childhood leukemia looking at parental 9 exposures where we specifically looked at benzene 10 and other solvents. I would have to look back 11 through my C.V. to see if there's anything else. 12 I know I've done research that is not published that 13 has looked at that issue. 14 Q. Is the paper you describe on childhood 15 leukemia the Buckley et al. paper of 1989? 16 A. Yes. 17 Q. What unpublished work have you written 18 with respect to benzene and leukemia? 19 A. We did a large cohort mortality study of 20 auto industry workers where we looked at each and 21 every one of about ninety causes of death, including 22 leukemias, and over a hundred different exposures. 23 In that list of exposures we included many different 24 types of solvents, oils, lubricants, greases, 25 machining fluids, petroleum distillates, paint 54 1 thinners. So these were auto industry workers and 2 we looked at leukemia mortality. 3 Q. In that cohort of workers, were there any 4 measured air concentrations of benzene that were 5 available for purposes of exposure analysis? 6 A. We did not have any at the time we did 7 this study. I am aware that there is some 8 information regarding industrial hygiene data from 9 the plants we studied, but it was not incorporated 10 into our research. 11 Q. Was that study commissioned by one or more 12 of the auto companies? 13 A. No. It was commissioned by the United 14 Auto Workers-Ford Motor Company National Joint 15 Committee. 16 Q. What were your findings with respect to 17 benzene and -- Excuse me. What were your findings 18 with respect to solvent petroleum exposures and 19 leukemia in that study? 20 A. To the best of my recollection, I don't 21 think we found any significant associations. 22 Q. Are you familiar generally with 23 Dr. Smith's work in the field of benzene toxicology? 24 A. I have seen some of it. I wouldn't say 25 I'm generally familiar with it. 55 1 Q. Do you have an opinion as to whether 2 Dr. Smith is considered a highly respected benzene 3 toxicology researcher? 4 MR. LEGHORN: Object to the form. 5 A. I don't know whether he is a highly 6 respected benzene toxicology researcher. I do 7 believe he's a highly respected scientist. 8 Q. Did you read either or both of 9 Dr. Cranor's reports in this case? 10 A. I read Dr. Cranor's report dated October 11 9, 2008. 12 Q. Dr. Cranor issued a supplemental report on 13 December 9, 2008. Did you have the opportunity to 14 review that? 15 A. I don't believe so. 16 Q. In paragraph 6 of his supplemental report 17 Dr. Craner states, quote, "As a part of a review of 18 the science, whether for assessing claims in physics 19 or human health, scientific judgment is critically 20 involved not only for drawing the ultimate 21 conclusions but also for a number of steps along the 22 way," close quote. Do you agree with that 23 statement? 24 MR. LEGHORN: Object to the form. 25 A. I've never read the report. Could I see 56 1 it and read that section of it? (Pause) Okay, 2 what's your question? 3 Q. If I may have it back, I'll repeat the 4 question. 5 A. Well, I need it. Do you want me to read 6 that sentence back? 7 Q. Yes. The sentence that I read to you, and 8 my question was simply whether you agreed with that 9 sentence. 10 A. I would say yes, I agree with that 11 sentence, and I strongly agree with the preceding 12 paragraph in which Dr. Cranor says "A scientist in 13 reviewing and assessing all the scientific evidence 14 for a conclusion such as whether benzene exposure 15 can cause APL must consider and integrate all the 16 available relevant evidence." And that is of course 17 the critical issue that Dr. Smith has not done. He 18 has not properly looked at the incidence rates of 19 APL in China. He has relied on calculations that 20 are not reliable, choosing them over calculations 21 based on much more reliable evidence. He has 22 invoked a healthy worker effect where none exists as 23 a potential objection to comparing incidence rates 24 in the benzene-exposed cohort in the Travis paper 25 compared to incidence rates in other populations in 57 1 China. 2 If he were to integrate and evaluate all 3 the relevant evidence, I think he would find that 4 his conclusions are not well-supported by the data. 5 Q. And in your opinion, biological mechanism 6 evidence is not relevant in this instance. Correct? 7 A. That is not my opinion. 8 Q. It is your opinion that because the 9 epidemiology in your view fails to show an 10 association between benzene and APL, that 11 consideration of biological mechanism evidence is 12 unnecessary. Isn't that true? 13 MR. LEGHORN: Object to the form. 14 A. That is not my opinion. It is my opinion 15 that the epidemiology provides a critical test of 16 whether hypotheses derived from mechanistic evidence 17 are borne out by the experience of the human 18 population, and in this setting they appear not to 19 be borne out. 20 Q. Is it your opinion that biological 21 mechanism evidence with respect to environmental 22 exposures and latent disease is never sufficient of 23 itself to establish that exposure causes the 24 disease? 25 MR. LEGHORN: Object to the form. 58 1 A. That is not my opinion. It is my opinion 2 that when you have epidemiologic evidence as we have 3 in this case, you must consider it, and it provides 4 a test of whether your mechanistic hypothesis is or 5 is not correct. 6 Q. And it is your view that regardless of the 7 limitations on that epidemiology, to the extent that 8 epidemiology fails to itself show an association, 9 that it renders the mechanistic evidence incapable 10 of reaching the conclusion of causation. Isn't that 11 right? 12 MR. LEGHORN: Object to form. 13 A. That is not my opinion. My opinion is 14 that you have or I should say Dr. Smith has a 15 mechanistic hypothesis. The epidemiology provides a 16 real test of whether populations exposed to benzene 17 are at increased risk of APL; it does not support a 18 conclusion that they are. 19 Q. And the flip side is, Doctor, that you've 20 admitted that the epidemiological evidence also 21 cannot prove the null hypothesis. Correct? 22 A. That is correct, and that is a widely 23 accepted principle of scientific inference across 24 all branches of science. 25 Q. So the epidemiological evidence with 59 1 respect to benzene and APL in your opinion neither 2 proves that benzene does cause APL nor does it prove 3 that benzene does not cause APL. Correct? 4 MR. LEGHORN: Object to the form. 5 A. The issue is this. In science if you have 6 a hypothesis that some agent causes a disease, let 7 it be the hypothesis that benzene causes APL, the 8 burden of proof is on the person who forms the 9 hypothesis to show that their hypothesis is correct. 10 We largely endorse the approach written by Karl 11 Popper, the idea of refutation of hypothesis, that 12 what most scientists do is try to refute hypotheses 13 that have been proposed. 14 So we are in a setting where Dr. Smith has 15 hypothesized that low-level benzene exposure causes 16 APL, and it is appropriate that the epidemiology 17 evidence and other forms of scientific evidence be 18 examined carefully and systematically to determine 19 whether they provide proof of Dr. Smith's hypothesis 20 or whether they fail to provide it. In other words, 21 do they refute it or do they provide data that are 22 inconsistent with his hypothesis? The answer is 23 they provide data that fails to support Dr. Smith's 24 hypothesis. 25 What that means then is he has a 60 1 hypothesis that lacks proof. 2 Q. I am going to repeat my question which you 3 didn't answer. Isn't it your opinion that the 4 epidemiological evidence with respect to benzene 5 exposure and APL is neither sufficient to show that 6 benzene exposure can cause APL nor is it sufficient 7 to show that benzene exposure cannot cause APL? 8 MR. LEGHORN: Object to the form, asked 9 and answered. 10 A. As Dr. Smith testified, you cannot prove 11 the null hypothesis. That is widely agreed across 12 all areas of science. So your question has a 13 presumption in it that somehow I should be forced to 14 do something that we all agree can't be done. In 15 science we seek to prove alternative hypotheses such 16 as that benzene causes APL and, failing to do so, 17 which is the situation we're in, we are left with 18 the conclusion that there is not adequate evidence 19 to conclude that benzene causes APL, period. 20 Q. Based on the human epidemiology studies. 21 Right? 22 A. Based on a consideration of all the 23 evidence. 24 Q. Well, Doctor, if the human epidemiology 25 studies themselves can't prove the null hypothesis, 61 1 then how can they be affirmative evidence that 2 benzene exposure does not cause APL? 3 MR. LEGHORN: Object to form, 4 argumentative. 5 A. I've already answered that question. 6 Q. Is there any circumstance in which you 7 would find that biologic mechanism evidence itself 8 is sufficient to reach a conclusion scientifically 9 that more likely than not exposure to agent A causes 10 disease B? 11 MR. LEGHORN: Object to the form. 12 A. There are numerous instances where that 13 could be true. If you had a gunshot wound to the 14 head and you explored the anatomy of what had 15 happened as the bullet passed through the head, I 16 think you could easily invoke mechanistic evidence 17 that would say that was the cause of death; you 18 don't need epidemiology. 19 Q. I'll try it again. Is there any 20 circumstance in which you would find based on 21 mechanistic evidence in the absence of epidemiology 22 evidence that an exposure to an environmental agent 23 causes a particular disease following some latency 24 period? 25 MR. LEGHORN: Object to the form. 62 1 A. I would be reluctant to accept mechanistic 2 evidence as the sole basis for a conclusion that an 3 environmental agent caused a disease such as cancer 4 with an appreciable time lag between exposure and 5 disease and in the setting where there were other 6 known or unknown causes of that disease. That 7 approach, I believe, is entirely consistent with for 8 example the International Agency for Research on 9 Cancer, which uses a set of rules in which they are 10 almost invariably unwilling to reach a decision that 11 a chemical agent is known to cause human cancer in 12 the absence of sufficient epidemiologic evidence. 13 They consider mechanistic evidence as a far less 14 important form of support for reaching those 15 conclusions than do they consider the epidemiologic 16 evidence. 17 If I were to try to apply the logic and 18 rules that IARC uses to the setting of benzene and 19 APL, it would be quite clear that the epidemiologic 20 evidence is inadequate to support a conclusion that 21 benzene causes APL. Whether you have mechanistic 22 evidence or not, I do not think you could sustain a 23 conclusion that benzene is known to cause APL in 24 humans. 25 Q. In fact, IARC has a classification for 63 1 whether something is a known carcinogen and whether 2 something is a probable carcinogen as well. 3 Correct? 4 A. I believe they use the terms known, 5 probable, possible, and not classifiable. They also 6 have a fourth classification for known not to cause. 7 There's only one chemical that's ever been 8 classified in that last category. 9 Q. And IARC will classify agents as probably 10 carcinogenic to humans in circumstances when there 11 is limited evidence of carcinogenicity in humans and 12 sufficient evidence of carcinogenicity in 13 experimental animals. Correct? 14 A. I believe that's correct. And I am not 15 aware that we have sufficient evidence in animals 16 for benzene and APL and I believe we do not even 17 have limited evidence in humans. I doubt this would 18 be IARC class 2A. It would probably be IARC 19 category 3. 20 Q. IARC also classifies a carcinogen or 21 possible carcinogen as 2A as a probable carcinogen 22 when there is inadequate evidence of carcinogenicity 23 in humans, sufficient evidence of carcinogenicity in 24 animals, and strong evidence that carcinogenesis is 25 mediated by a mechanism that also operates in 64 1 humans. Is that right? 2 A. I believe that's one of the set of 3 criteria by which they will classify a chemical as a 4 2A or probable carcinogen. But I don't believe that 5 we've satisfied those conditions. 6 Q. Would you agree that the human 7 epidemiology studies that have examined benzene 8 exposure and acute nonlymphocytic leukemias as a 9 group have shown a strong association between 10 benzene and ANLL? 11 MR. LEGHORN: Object to the form. 12 A. The epidemiology studies that have 13 examined that have shown that there is a strong 14 association between benzene exposure at levels that 15 exceed 40 to 50 ppm-years in some studies, in some 16 more than 200 ppm-years and increased risk of ANLL. 17 Q. And that association in the studies 18 between benzene and ANLL has also been a consistent 19 association. Would you agree? 20 MR. LEGHORN: Object to the form. 21 A. I'm not sure what you mean by consistent 22 association. 23 Q. I mean it has been consistently found in a 24 number of different cohorts and a number of case25 control studies that there is an association between 65 1 ANLL as a group and benzene exposure. 2 MR. LEGHORN: Object to the form. 3 A. There is a body of well-done epidemiology 4 that I believe supports the conclusion that benzene 5 exposures greater than 40 to 50 ppm-years are 6 causally associated with increased risk of ANLL. 7 Q. And those studies show or many of them 8 show an increase in the risk with increased dose. 9 Correct? 10 MR. LEGHORN: Object to the form. 11 A. I would not characterize it as many. It's 12 actually a few have adequate exposure information to 13 be responsive to your question. 14 Q. And taken as a whole, the body of 15 literature, epidemiological literature looking at 16 benzene exposure and ANLL shows a coherent picture. 17 Would you agree? 18 MR. LEGHORN: Object to the form. 19 A. You would have to define what you mean by 20 coherent. The previous series of questions I think 21 were more consistent with the concept of replication 22 rather than coherence. 23 Q. How do you interpret coherence as it 24 relates to the set of considerations described by 25 Sir Austin Bradford-Hill? 66 1 A. Let's use Hill's own description of it. 2 He says "Coherence. On the other hand, the 3 cause-and-effect interpretation of our data should 4 not seriously conflict with the generally known 5 facts of the natural history and biology of the 6 disease." He goes on to say "Thus, in the 7 discussion of lung cancer, the committee finds its 8 association with cigarette smoking coherent with the 9 temporal rise that has taken place in the two 10 variables over the last generation and with the sex 11 difference in mortality, features that might well 12 apply in an occupational problem. The known 13 urban/rural ratio of lung cancer mortality does not 14 detract from coherence, nor the restriction of the 15 effect to the lung." 16 And so Hill is actually talking about the 17 overall pattern of the findings whereas I think your 18 previous questions were talking about replication of 19 findings. Clearly, with respect to benzene and APL, 20 there's no coherent picture. There's none 21 whatsoever. We don't have the characteristics such 22 as Hill just described for smoking and lung cancer 23 where you have time trends that lag the prevalence 24 of smoking, where you have a sex ratio that reflects 25 the difference in smoking patterns between men and 67 1 women, where you have an urban/rural difference. In 2 other words, that's a coherent picture. We don't 3 have that for benzene and APL. 4 Q. Do you have a coherent picture for benzene 5 and ANLL? 6 A. I think we do. I might mention on that, 7 one of the important elements in that coherent 8 picture is the evidence that only the exposures 9 within about ten to fifteen years of diagnosis 10 convey risk. So not only do we have evidence of 11 dose response where the risk is clearly increased at 12 exposure levels above 40 to 50 ppm-years, but we 13 also have evidence that distant past exposures don't 14 matter. And that is really a picture that has 15 become coherent in the past ten years. 16 MR. JENSEN: Counsel, I think that I have 17 less than thirty minutes of questioning left. I 18 don't know for sure, but that's what I think. 19 MR. LEGHORN: Do you want this on the 20 record? 21 MR. JENSEN: Either way. I don't care. 22 We're getting close to lunchtime. I think it should 23 be the doctor's decision primarily what we do in 24 terms of whether to take a lunch break or not. But 25 if we were to take a short bathroom break now, I 68 1 could organize my thoughts and probably finish 2 within thirty minutes. It's up to you. 3 THE WITNESS: I would prefer to finish 4 without taking a lunch break. 5 MR. JENSEN: All right. Let's take a very 6 short break now. 7 (In recess 11:43 a.m. to 11:53 a.m.) 8 BY MR. JENSEN: 9 Q. Dr. Garabrant, is it true that for 10 purposes of your opinion in this case, you have not 11 done any review of the literature relating to the 12 issue of the extent to which topoisomerase 13 inhibitors can and do cause APL? 14 A. No, actually I have looked at some of that 15 literature. 16 Q. But you are not offering any opinion on 17 that subject. Is that correct? 18 A. That's correct. It is my understanding 19 that Dr. Pyatt is addressing that issue. 20 Q. And you are not offering any opinion on 21 the subject of the extent to which benzene exposure 22 inhibits topoisomerase II. Is that correct? 23 MR. LEGHORN: Object to the form. 24 A. I have no intention of offering opinions 25 regarding the extent to which benzene inhibits 69 1 topoisomerase II. I have not been asked to do that. 2 Q. I have just put an exhibit sticker, 3 Exhibit 4, on a collection of binder-clipped 4 articles that were left for me at my hotel when I 5 arrived last night. And I would ask you to take a 6 quick look, Doctor, and see if you recognize those 7 articles. 8 A. (Pause) Yes. 9 Q. Did you recently identify those articles 10 as articles that you wished to rely on for purposes 11 of your opinions in this case? 12 A. I believe these are articles I assembled 13 since submitting my report on November 19, 2008, 14 although -- Let me restate that. These are articles 15 I have reviewed since submitting my report in 16 November of 2008, although some of them are already 17 cited in my report and I had reviewed them prior to 18 submitting my report in addition. I mean, I don't 19 know how to answer these things I recently 20 accumulated. Some of them I've had for years; some 21 of them I've cited in my report; some of them I 22 re-reviewed; some of them were new since submitting 23 my report. 24 Q. My understanding from counsel is that you 25 have recently been out of the country. Is that 70 1 correct? 2 A. Yes. 3 Q. And after you got back from your trip, you 4 identified the set of articles that have been marked 5 as Exhibit 4 as materials that you wished to rely on 6 to defense counsel in this case. Is that correct? 7 A. I'm not sure that's exactly correct. 8 These are articles for which I provided defense 9 counsel with copies just yesterday. Some of them 10 are cited in my report of November 19 and I assume 11 that everybody has had copies of them since November 12 or had copies prior to that. Some of them are 13 articles that I looked up and read in response to 14 Dr. Smith's supplemental declaration. So it's a 15 combination of factors. The common theme is that I 16 gave copies of these electronically to Mr. Leghorn 17 yesterday. So some of them are old, some are new, 18 some are in response to Dr. Smith's work, some are 19 ones that I had reread looking now at some new 20 aspect of them. 21 Q. Okay, thank you. 22 MR. JENSEN: And just for the record, to 23 the extent that the articles in Exhibit 4 include 24 materials that were previously referenced in 25 Dr. Garabrant's original report, we certainly don't 71 1 have any objection to his reliance on them. To the 2 extent that they're new, we do have an objection. 3 That can be resolved however it will be resolved. 4 Thank you. 5 BY MR. JENSEN: 6 Q. Dr. Garabrant, who compiles for you your 7 list of legal testimony as reflected by Exhibit 1? 8 A. I do. 9 Q. You do. How do you do it? 10 A. I try to keep track of all the times I 11 have been deposed or testified in court. 12 Q. And the reason I ask is because I'm aware 13 of at least one area or one instance of testimony 14 that doesn't seem to be reflected, which is trial 15 testimony you gave in the case of Turner versus 16 Chevron. You've got your deposition testimony 17 reflected but no trial testimony. 18 A. If I've omitted it, it isn't by intent. 19 That list, I believe, reflects that that was 20 prepared on November 18. I don't remember when I 21 testified in Turner, whether it was just before that 22 or just after that. It was sometime last fall, so 23 I may have missed it or it may have been brand-new; 24 I don't know. 25 Q. Does anyone besides yourself check the 72 1 accuracy of the list? 2 A. No. 3 Q. If you will turn with me to Dr. Cranor's 4 report of October 9 at page 10, there is a 5 subparagraph iii in the middle of page 10 and the 6 third sentence of that paragraph says "As both 7 medical and legal commentators put the point, 'In 8 the final analysis, assessment of evidence and 9 causal inferences depend on accumulating all 10 potentially relevant evidence and making a 11 subjective judgment about the strength of the 12 evidence'." Do you agree with that statement? 13 MR. LEGHORN: Object to the form. 14 A. I would have to go back and read reference 15 39, which is footnoted at the bottom Kassirer & 16 Cecil, Annals of Internal Medicine, before I would 17 comment on what Dr. Cranor is concluding there. 18 Q. Do you have an opinion as to whether the 19 following sentence after that sentence is true, 20 which says "When professional judgment is so central 21 to the drawing of inferences, professionals may 22 disagree at various stages of reasoning in an 23 inference to the best explanation, and thus may have 24 disagreements about their overall view of the 25 scientific conclusions"? 73 1 A. Well, I mean, I think what Dr. Cranor is 2 saying there is nothing more than people can 3 disagree on their view of what a body of evidence 4 says. So I don't think there's any content to that 5 statement. We know that people can disagree, and of 6 course what we try to do is to explore why they 7 disagree and try to uncover whether they have looked 8 at all of the evidence, whether they have considered 9 it properly, whether they have made their 10 calculations properly, whether they have followed 11 the rules of science that are widely accepted and so 12 on. 13 So I don't think there's much content to 14 Dr. Cranor's sentence. People can disagree. 15 Q. Outside the context of litigation, 16 scientists disagree with one another about 17 scientific propositions all the time. Is that true? 18 MR. LEGHORN: Object to the form. 19 A. The essence of the scientific method is 20 based on the idea that every scientist has a 21 responsibility to be skeptical of everyone else's 22 work and to try to refute it. That's what we do as 23 scientists, and that's not limited to medicine or 24 epidemiology or physics or biology. That's part of 25 the method. And when we can refute other people's 74 1 conclusions with evidence, then their conclusions 2 should not be trusted as reliable. 3 Q. Well, oftentimes scientists have competing 4 inferences and competing conclusions based on the 5 same evidence. Isn't that true? 6 MR. LEGHORN: Object to the form. 7 A. I would have to say no because of your use 8 of the word "often." Largely scientists live in a 9 world where we agree to the vast body of evidence 10 and disagree only around the fringes. Once in a 11 while that's not true. As Cranor points out, 12 Newton's laws of motion were widely accepted until 13 Einstein came along and showed that at very high 14 velocities, newtonian physics don't really -- aren't 15 supported by real evidence, by data or by theory. 16 But largely we live in a world where bridges don't 17 fall down; where when you turn your furnace on it 18 heats your house; where when you hit the brakes in 19 your car, it stops; where when you have a blood test 20 done, the results mean what we all agree they mean. 21 The methods of science have turned out to provide an 22 extraordinarily broad and reliable body of evidence 23 that has allowed mankind to move forward and to make 24 great progress in almost every realm of life. 25 And so when you say we often disagree, my 75 1 answer is no. Overwhelmingly we agree. It's only 2 around the fringes of discovery where the pattern of 3 evidence is unclear that we disagree. And that is 4 of course what is going on in this case. Is there 5 evidence that benzene causes APL? And the answer 6 is: not nearly enough to conclude that there is. 7 And as we've discussed, Dr. Smith has simply looked 8 at the epidemiology improperly and failed to 9 consider the epidemiology that is directly relevant 10 to his opinions and that is a direct test of his 11 hypothesis and doesn't support it. 12 MR. JENSEN: I will let you close with 13 your closing argument. That's all the questions 14 I have. 15 MR. LEGHORN: Thank you. 16 (Deposition concluded at 12:10 p.m.) 17 18 19 20 21 22 23 24 25 76 1 COURT REPORTER'S CERTIFICATE 2 I, J. Edward Varallo, RMR, CRR, Registered 3 Professional Reporter and Notary Public in the 4 Commonwealth of Massachusetts (my commission expires 5 12/24/2015), hereby certify that the deposition of 6 David H. Garabrant, M.D., M.P.H. taken on February 7 18, 2009, in the matter of Brian K. Milward and 8 Linda J. Milward v. Acuity Specialty Products Group, 9 Inc., et al. was recorded by me stenographically and 10 transcribed; that before being sworn by me, the 11 deponent provided satisfactory evidence of 12 identification as required by Executive Order 455 13 (03-13) of the Governor. 14 I certify that the deposition transcript 15 produced by me is true and accurate to the best of 16 my ability. 17 I certify further that I am not counsel, 18 attorney, or relative of any party litigant, and 19 have no interest, financial or otherwise, in the 20 outcome of this suit. 21 22 23 24 _____________________________ 25 DATED: 2/25/2009 J. Edward Varallo 77 1 WITNESS: David H. Garabrant, M.D., M.P.H. 2 DATE: February 18, 2009 3 CASE: Brian K. Milward and Linda J. Milward v. 4 Acuity Specialty Products Group, Inc., 5 et al. 6 7 8 DISTRIBUTION TO COUNSEL The original signature 9 page/errata sheet was sent to Joseph J. Leghorn, 10 Esq., to obtain signature from the deponent. When 11 signed, please send original to Steve Baughman 12 Jensen, Esq., who will supply a copy of the signed 13 errata sheet to other counsel present at the 14 deposition. 15 16 17 WITNESS INSTRUCTIONS After reading the transcript 18 of your deposition, please note any change or 19 correction and the reason for it on the errata 20 sheet. DO NOT make any notations on the transcript 21 itself. Use additional sheets if necessary. 22 23 24 SIGN AND DATE THE ERRATA SHEET and return it, along 25 with the transcript, to your counsel. 78