Document 9QqrLeRwqyVN5aVMEOwQ6rw6

1 IN THE UNITED STATES DISTRICT COURT 2 FOR THE NORTHERN DISTRICT OF OHIO EASTERN DIVISION 3 4 5 EDDIE A. FANNIN, 6 Plaintiff, 7 vs No. 5:93-CV-2594 8 NORFOLK & WESTERN RAILWAY COMPANY, 9 Defendant. 10 Deposition Of 11 OTTO WONG 12 Tuesday, December 16, 1997 13 14 15 16 17 18 19 REPORTED BY: DONNA LEE BECKETT, CSR #3450 20 21 22 NOGARA REPORTING SERVICE 23 130 Battery Street, Suits 580 San Francisco, California 94111 24 (415) 398-1889 25 1 1 BE IT REMEMBERED that, pursuant to Notice 2 of Taking Deposition and on Tuesday, the 16th day of 3 December, 1997, commencing at the hour of 2:05 o'clock 4 p.m. thereof, at the Law Offices of Burns, White & 5 Hickton, 50 California Street, Suite 2310, San 6 Francisco, California, before me, DONNA LEE BECKETT, a 7 Certified Shorthand Reporter in the State of 8 California, personally appeared 9 OTTO WONG, Sc.D. F.A.C.E., 10 herein, called as a witness by the Plaintiff, having 11 been by me first duly sworn, was examined and testified 12 as hereinafter set forth. 13 14 APPEARANCES OF COUNSEL 15 For the Plaintiff (Appearing telephonically) HARTLEY & O'BRIEN 16 827 Main Street Wheeling, West Virginia 26003 17 By: B. DEAN HARTLEY, Attorney at Law 18 For the Defendant BURNS, WHITE & HICKTON 19 2400 Fifth Avenue Place 120 Fifth Avenue 20 Pittsburgh, Pennsylvania 15222-3001 By: QAVID A. DAMICO, Attorney at Law 21 22 23 24 25 2 1 EXAMINATION BY MR. HARTLEY 2 MR. HARTLEY: Q. Dr. Wong, as an epidemiologist, 3 do you attempt to determine whether causality can be 4 determined between Substance X and Disease Y? 5 A. That's true. 6 Q. Could you explain to me how you do that? 7 A. Basically we have what we call the. Hill 8 criteria for causation. We look at the data on that 9 relationship, exposure to Substance X and Disease Y in 10 terms of what we call the strength of association, 11 consistency of association, specificity of association, 12 dose response relationship and so on. Those are the 13 major criteria that we use in determining whether there 14 is an association between exposure and the disease. 15 Q. Well, let's talk for a second about 16 specificity. When you utilize the term specificity, 17 what do you mean? 18 A. Well, in terms of specificity, we are talking 19 about both components of the relationship. First we've 20 got to be specific in terms of exposure. And we've got 21 to be specific in terns of the disease outcome as well. 22 For example, if 1 make the statement, chemicals cause 23 canter, that is not very specific, but certainly that 24 is true. But that's not very informative. We need to 25 specify what kind of chemical and we also need to 3 < < NOGARA REPORTING SERVICE 1 specify what kind of cancer. 2 Q. Does specificity require you, when you're 3 utilizing the Bradford Hill criteria, to exclude other 4 types of cancer and only center on one specific type? 5 A. No, it does not because a chemical may be able 6 to cause more than one type of cancer. But whatever 7 cancer we are talking about, we need to have specific 8 data on the relationship between that chemical and the 9 specific kind of cancer that we're talking about. 10 Q. Are you talking about specific data meaning 11 biological plausibility or something else? 12 A. I am talking about epidemiologic data. 13 Q. Would you agree with me that when you are 14 doing your epidemiologic studies that if you look at 15 particular types of cancers that are similar in nature 16 they can be lumped together to determine statistical 17 significance? 18 A. Well, when you say "similar," that's a big 19 word. To some you may be similar. To others it may 20 not be that similar. If I am reading your question 21 right, I don't think multiple myeloma is similar to 22 leukemia or Hodgkin's disease. That's what we're 23 talking about. 24 Q. If we look at the various studies you would 25 agree, would you not, that multiple myeloma is a 4 < < NOGARA REPORTING SERVICE 1 lymphatic disease? 2 A. It's one of the specific types of lymphocytic 3 disease, yes. 4 Q. And it is classified on ICD9 as 203, correct? 5 A. Yes, sir. 6 Q. And in your paper in 1989, which you did with 7 Dr. Raabe entitled -- R-a-a-b-a -- A Critical Review of 8 Cancer Epidemiology in Studies of Petroleum Industry 9 Employees with a Quantitative Meta-analysis by Cancer 10 Site, didn't you suggest that you could combine cancer 11 of other lymphatic tissues into one category for 12 statistical analysis? 13 A. In 1989 we looked at some old papers when we 14 do the review and the meta-analysis. At that time we 15 said that some of the groups overlap in the old studies 16 and some of them may be combined into different groups. 17 I did say that. But that doesn't mean 18 Q. Well, if 19 A. Can I finish? 20 Q. Yes, sir. 1 thought you were. I'm sorry. 21 A I guess that's the problem of not being able 22 to see face-to-face. I'm not quite done with the 23 answer. 24 But when you look at more recent studies 25 you've got to look at the biology of the disease, what 5 < < NOGARA REPORTING SERVICE 1 we understand at this point. Certainly, for multiple 2 myeloma it's very distinct. Now we do understand that 3 it's very distinct from leukemia and it's very distinct 4 from non-Hodgkin's lymphoma. 5 Q. Well, what you are saying is that it has 6 changed since 1989, in your opinion, whether you should 7 compact the various diseases into one category? 8 A. In 1989 I didn't say that. You have to look 9 at them as a combined group. I'm saying that it's open 10 to question. We should look at that as a total group. 11 We should also look at the specific groups as well. 12 But now the hematologists tell us that multiple myeloma 13 is a very distinct disease, very different from other 14 categories in the lymphopoietic cancer broad category. 15 Q. How long, in your opinion, has multiple 16 myeloma been considered distinct from lymphomas? 17 A. At least within the last, I would say, my 18 understanding is seven or eight years. Maybe longer 19 than that. 20 Q. And upon what do you base that opinion? 21 A. Based on -- well, I'm sure you have a copy of 22 my report. The report dated November 25th, 1997. 23 Q. November 28th, Doctor? 24 A. No, 25. 1 have two reports. The first one 25 was dated November 25th, 1997. 6 < < NOGARA REPORTING SERVICE > >' 1 Q. I don't have that report. Was that to 2 Mr. Donahoe? 3 A. Yes. 4 Q. I don't have that report. I never received 5 that report. 6 MR. DAMICO: You never received the November 25th, 7 '97? 8 MR. HARTLEY: No, sir, I did not. I've got 9 November 28th of '97. 10 Q. Does November 25th, 1997 discuss paints and 11 solvents and everything else? 12 A. The report dated November 25th, 1997 talks 13 about benzene exposure and multiple myeloma. 14 Q. I don't have that report. 15 A. Well, that's kind of unfortunate, because it's 16 going to be difficult for me to refer you to specific. 17 references and so on. 18 MR. DAMICO: We'll put it in the fax machine to 19 you right now. What's your fax number? Off the 20 record. 21 1fliscussion off the record.) 22 MR. HARTLEY: Q. Doctor, the last question to you 23 was; I believe, what do you base your opinion on that 24 the lymphoma should be considered separate and distinct 25 from multiple myeloma? 7 < < NOGARA REPORTING SERVICE 1 A. Right. And I was going to refer you to my 2 report which you would be receiving shortly. Anyway, 3 let me point out where I talk about that in my report. 4 That's on Page 3, the first full paragraph. And I do 5 have two references there. One is dated 1981. The 6 other one is dated 1985. Although those were published 7 in 1981 and 1985, 1 was not aware of those references 8 until much later. So as far as I'm concerned, 1 didn't 9 know of that until much later. 10 Q. Could you tell me the title of the 1981 11 article? 12 A. The 1981 is actually a textbook on hematology. 13 And the lead author of the textbook is Wintrobe. 14 W-i-n-t-r-o-b-e. 15 Q. And the other one? 16 A. And the other one is an article actually 17 written by someone at OSHA. Her name is Tsongas. 18 T-s-o-n-g-a-s. 1985. 19 Q. And what does Ms. Tsongas say in her report in 20 1985? 21 A 'Basically Dr. Tsongas said that we should 22 when we analyze the data we should separate the disease 23 entity out. 24 Q. Did she give a reason for that? 25 A. Basically, she cited some, I guess, hematology 8 < < NOGARA REPORTING SERVICE 1 papers, some epidemiologic papers and so on, to support 2 that. 3 Q. Is Dr. Tsongas a pathologist? 4 A. I don't know what she is. 5 . Q. So you just relied on her paper of 1995 that 6 said that you should separate those out? 7 A. Well, those are the two samples that indicate 8 that non-Hodgkin's leukemia and multiple myeloma are 9 different diseases. 10 Q. And you think that she is a doctor? 11 A. I think she has a Ph.D. 12 Q. And Ph.D. in what, Dr. Wong? 13 A. I don't know. 14 Q. Other than the Wintrobe book and the Tsongas 15 report of '85, was there any other pathological 16 evaluation that you reviewed that suggested that you 17 should not classify the lymphopoietic malignancies for 18 statistical purposes when you're doing an analysis? 19 A. There are quite a few. Those are just two 20 analysis. 21 Q 'When you testified before OSHA on the benzene 22 standards did you have an opportunity to talk to Dr. 23 Tsoingas? 24 A. No. 25 Q. She was not in the office at that time, the 9 < < NOGARA REPORTING SERVICE 1 OSHA office? 2 A. I didn't go to OSHA and testify. I testified 3 in front of an administrative judge in a hotel in Los 4 Angeles. 5 Q. Were there any OSHA representatives there? 6 A. As far as I know I recognized one person, Dr. 7 Infante. I-n-f-a-n-t-e. 8 Q. And were you hired by OSHA to review and 9 testify during the hearings on the benzene standard? 10 A. I was a consultant to OSHA for the hearing, 11 yes. 12 Q. That means you were paid by them, right? 13 A. Not a whole lot of money. Yes, if you call 14 that pay. 15 Q. You got some remuneration? 16 A. Right. 17 Q. Have you reviewed your '86 testimony before 18 OSHA in the recent past? 19 A. Not really, but I remember roughly what I 20 said. 21 Q: "Isn't it true that in response to the comments 22 by Dr. MacMahon -- M-a-c, capital M, a-h-o-n -- wherein 23 he recommended that maybe it was not appropriate to 24 analyze the lymphomas and the leukemias together, you 25 then broke down non-Hodgkin's lymphoma and included 10 < < NOGARA REPORTING SERVICE 1 multiple myeloma with non-lymphoma in your report to 2 OSHA? 3 A. That's true. And that's because the software 4 package that we used for the analysis at that time, we 5 were not able to break it up into non-Hodgkin's 6 lymphoma and multiple myeloma. We were limited by the 7 software that we used at that time. 8 Q. How did the software prevent you from doing 9 that? 10 A. Because basically the software grouped part of 11 the non-Hodgkin's lymphoma and multiple myeloma as one 12 group. And that's the way how we calculated the SMR at 13 that time. 14 Q. In your testimony before OSHA, you never 15 indicated in response to Dr. MacMahon's remarks that 16 the computer program was the cause of the problem but 17 rather you said that the difficulty in separating 18 certain poietic cancers from each other is that a 19 certain amount of overlap appeared to occur. And 20 there's no comment about a computer program in that 21 testimony, is that correct? 22 A. I don't think I talked about the computer 23 program at that time. I think right now you're 24 focusing on multiple myeloma. But at the OSHA hearing 25 we were talking about different diseases within that 11 < < NOGARA REPORTING SERVICE 1 broad category in general. We were not focusing on 2 multiple myeloma alone. And 1 can tell you that for 3 some disease I still think it's appropriate to combine 4 certain ICD groupings together. In fact, for 5 non-Hodgkin's lymphoma we have to combine at least two 6 ICD codes in order to get multiple myeloma. So up to 7 today I still think that certain combinations is 8 appropriate, but not across the board. 9 Q. Would it be appropriate to combine 10 non-Hodgkin's lymphoma with multiple myeloma in your 11 analysis of a cohort? 12 A. As -- well, it depends on what the question 13 is. If the question is on multiple myeloma, 14 specifically on multiple myeloma, no, it would not be 15 appropriate to utilize the results combining multiple 16 myeloma and non-Hodgkin's lymphoma together. 17 Q. But you combined those in your testimony 18 before OSHA, didn't you? You combined ICD 8, 200, 202, 19 and 203, which is multiple myeloma. 20 A. That's right, because, as I said, the software 21 program would not allow us to do a separate analysis. 22 Q. Well, you also testified in 1986 in front of 23 OSHA that the medical literature is replete with 24 reports documenting the transition from certain 25 lymphomas and multiple myelomas to leukemia. And 12 < < NOGARA REPORTING SERVICE 1 leukemia phase has long been recognized in some 2 non-Hodgkin's lymphoma cases. And leukemic 3 transformation of lymphoid cells is considered a part 4 of the natural course of some non-Hodgkin's lymphoma 5 and lymphoma in general. 6 A. When you say testified to that effect, I think 7 it's slightly misleading. Because if you read. the 8 statement, or if you read my report, my publication 9 which appeared in 1987, basically I quoted a number of 10 people saying what you just said, okay, those -- at 11 least some thoughts in the community, in the scientific 12 community at that time. And being an epidemiologist, I 13 simply say that there are people who think this way. 14 Then they think that we should combine some certain 15 categories for analysis. But at the same time for 16 certain categories we should also separate them. 17 Q. Now, then, was your computer program able to 18 differentiate between leukemia and lymphoma? 19 A. Leukemia and lymphoma, yes. 20 Q. Why did you combine lymphoma with leukemia in 21 your statistical analysis in '86 before OSHA? 22 A. Well, if you look at any epidemiologic 23 studies, we have a hierarchy of causes of death. We 24 started with all causes of death. And then we break it 25 down into cancer deaths and non-cancer deaths. Among 13 < < NOGARA REPORTING SERVICE > >- 1 cancer deaths we break that down by different body 2 systems, respiratory system, gastrointestinal system 3 and so on. So I don't see anything that's different 4 from we combine colon cancer and stomach cancer in some 5 analysis, but at the same time separate that in some 6 other analysis. 7 Q. Didn't you do that in your publication of what 8 is commonly referred to as Wong 1 and Wong 2? 9 A. I don't know what's commonly referred to. I 10 published two papers in 1987. Part 1 and Part 2. 11 Q. Part 1 and Part 2. 12 A. I commonly refer to them as Wong 1 and Wong 2. 13 Q. You did that in those publications as well, 14 did you not, combine the lymphatic malignancies along 15 with the leukemias? 16 A. That's right. But, as I said, it's not any 17 different from the situation of cancer of the digestive 18 system. In the cancer of the digestive system we 19 combine stomach cancer and colon cancer together. But 20 that doesn't mean that the stomach is the same as 21 colon. 22 Q. Well, even after Dr. MacMahon made the 23 recommendations that you split it out, you in your 24 testimony to OSHA justified the combination and then 25 published articles combining the various categories of 14 < < NOGARA REPORTING SERVICE > > 1 disease that came from the cohort, didn't you? 2 A. That's not correct at all. In 1987, very 3 specifically I have published results specific to 4 leukemia by itself. 5 Q. Did you combine the lymphatic leukemias with 6 lymphomas in your analysis? 7 A. In some analysis I did, but not in all. 1 8 present both sides at that time. 9 Q. And in '87 you found a statistically 10 significant increase in the cohort you were studying 11 for lymphatic leukemias, did you not? 12 A. Yes, I did. And I said that the increase was 13 driven by the increase in leukemia. 14 Q. And what cell type of leukemia did you find in 15 the cohort that you were reporting on in 1987? 16 A. In the 1987 paper I point out that I do not 17 have enough cases to do any cell type specific 18 analysis. 19 Q. What does your paper report as the cell type 20 you found? 21 A: I find all kind of cell types. 22 Q. Well, let's look at them. Do you have your 23 paper? 24 A. Yes, I do. 25 Q. In which one do we want to look at them, dose 15 < < NOGARA REPORTING SERVICE > > 1 response analysis or general results? 2 A. Hold on for just a second. I think we want to 3 look at the second paper and Table 16. 4 Q. What paper is that, 1 or 2, Doctor? I'm 5 sorry. I didn't hear you. 6 A. 2. 7 Q. Table 16? 8 A. Right. 9 Q. What is the title of Table 16? 10 A. Characteristics of 22 Deaths from Lymphatic 11 and Hemopoietic Cancer. 12 Q. What cell types did you find in this cohort? 13 A. I have CML. I have CLL. I have some 14 non-specified leukemia. So we don't know exactly what 15 cell type we're talking about. I think we also have 16 ALL. 17 Q. Were there any AMLs? 18 A. Not specifically, but we do have non-specified 19 acute leukemia, I think. So that could have been AML. 20 Q. How many of those did you have non-specific 21 acute leukemias? 22 A. I don't have the ICD code in front of me. 23 b. I do. What number do you think it would be? 24 What is AML? 25 A. AML? 16 < < NOGARA REPORTING SERVICE 1 Q. Yes. 2 A. I think AML is 205.0. 3 Q. Acute is 205.0. How many 205.Os do you 4 report? 5 A. As I told you, we don't have any. 6 Q. And how many non-specific acutes do you have? 7 Do you need that number as well? 8 A. Yes. I think that's 207.0, is that right? 9 Q. Leukemia of unspecified type, 208. 10 A. 208? 11 Q. Yes. 12 A. What about 207? 13 Q. 207 is acute arrhythmia and aleukro leukemia. 14 That's 207. 15 A. Yes. What revision are you looking at? 16 Q. I'm looking at the ninth. 17 A. The ninth. The one that I used, I think, is 18 the eighth in the study. So there may be a slight 19 difference there. 1 think 207.0, Case 12, Case No. 12. 20 Q. Was the non-specific acute? 21 A: Yes, the non-specific acute. 22 Q. How many multiple myeloma did you find? 23 A. I found three. 24 Q. Did you find an elevation of multiple myeloma 25 within the cohort? 17 < < NOGARA REPORTING SERVICE > >- 1 A. Based on my calculation I should expect 2 2 point -- it's in the report. You have a copy of the 3 report now? 4 Q. Not yet. I'm looking at Page 393 of the 5 second part of your 1987 paper. 6 A. As I said, in that paper I did not report the 7 expected because the calculation wasn't done at that 8 time based on that program. And also, the rates of 9 multiple myeloma was not widely available in the '80s. 10 But subsequently I have done a calculation on how many 11 multiple myelomas I should expect. And for that study 12 1 should expect -- let me find a number here. I should 13 expect 2.58. Okay. Roughly two and a half cases of 14 multiple myeloma. That's on Page 3 of my report. When 15 you get it you can take a look. 16 Q. What did you base your expected rate on? 17 A. The expected rate was based on the rates 18 published by the National Cancer Institute. 19 Q. Is that the U.S. general population? 20 A. Yes, sir. 21 Q: The rate that you just gave me, what was it? 22 A. No, that was not the rate. That was the 23 expected. 24 Q. What was your expected? 25 A. To 2.58. 18 < < NOGARA REPORTING SERVICE > > 1 Q. In your paper on Page 393, which is Wong 2, 2 you indicated that you're going to utilize 0.56 as your 3 expected. 4 A. Where are you? 5 Q. Under your chart you go eight lines down on 6 the left-hand column. 7 A. Okay. The difference is when I said. 2.58, 8 that was for the entire cohort, for the entire study. 9 The .56 refers to a much smaller group, what we call 10 the intermittent exposure. 11 Q. I understand that. 12 A. So that's the difference. 13 Q. So the expected for the intermittent exposure 14 group was still 0.56, correct? 15 A. It should be. 16 Q. And that means then that the SMR for the 17 intermittent multiple myeloma group was still 357.1, 18 correct? 19 A. Yes. 20 Q. And was that statistically significant? 21 A: 'No. 22 Q. What was the confidence interval? 46.2 to 23 1298? 24 A. Right. So that interval includes the norm of 25 100 percent. So, therefore, it's not statistically 19 < < NOGARA REPORTING SERVICE > > 1 significant. 2 Q. I understand that it's not statistically 3 significant, but it shows at least an elevated risk, 4 which could be by chance. But it does show an elevated 5 risk, correct? 6 A. It's based on two cases, remember. 7 Q. I thought you had three. 8 A. 1 think you're confusing the two analyses. 9 Q. There were not three multiple myelomas in the 10 intermittent( exposure group? 11 A. No, there were two in the intermittent 12 exposure group. For the entire study we had three. So 13 when you use the three as the observed, you have to use 14 the expected of 2.58. 15 Q. I understand what you did. 16 A. Do you understand what I'm saying? 17 Q. Yes. Why didn't you use internal controls for 18 your expected for the intermittent exposed multiple 19 myeloma cases? 20 A. Because that group is far too small to be 21 stable enough for internal comparison. I said very 22 clearly in my OSHA testimony as well as in my papers. 23 We started -- when we designed the study we have a much 24 larger internal control group planned, but we end up 25 with a very, very small group. And that is just not 20 1 stable enough at all. 2 Q. Could you use the overall industry as an 3 expected or as a group from which to derive your 4 expected multiple myeloma cases? 5 A. When you say "overall industry," what are you 6 referring to? 7 Q. The chemical industry? 8 A. There is no such data. You can't just go to 9 one place and say, okay. Here are the rates of 10 multiple myeloma or here are the rates for leukemia for 11 chemical workers. There's no such thing. No such data 12 base. 13 Q. Explain to me why you didn't use your own 14 industry controls. You had 6- or 7,000 people, didn't 15 you, in this study? 16 A. Yes. 17 Q. What was the amount of your control group? 18 How many were in your control group? 19 A. Out of the 7,000 or so I think 3,000 or so 20 were not exposed. 21 Q:-'So you couldn't use the 3,000 people to 22 determine your expected for multiple myeloma for 23 intermittently exposed? 24 A. No, it's too small. Multiple myeloma is very 25 rare. For comparison to be meaningful you have to have 21 < < NOGARA REPORTING SERVICE > >. 1 a large group that when you use the age specific rates 2 you should have at least a few deaths in each age 3 category. 4 Q. Well, you didn't have any leukemia deaths, did 5 you, in your controls? 6 A. No. 7 Q. So wouldn't that same analysis apply to your 8 leukemia calculations? 9 A. I think you just misspoke. Can you repeat the 10 question? 11 Q. Sure. If you assert that your control group 12 was too small to determine what the expected would be 13 for multiple myeloma, why wouldn't that also apply to 14 your leukemia expected rates since you found zero 15 expected in that group? 16 A. I said the same thing for leukemia, that the 17 rates, based on internal comparison, was not stable. 18 Q. You didn't say that to OSHA. You told OSHA in 19 response to Dr. MacMahon that you felt it was better to 20 use internal controls from this group then to go out 21 and use the general population? 22 A. No. 23 Q. Because -- let me finish -- because there was, 24 according to you, deficits reported by various 25 investigators of leukemia in various industrial 22 < < NOGARA REPORTING SERVICE > > 1 settings. 2 A. 1 said that. But I also said that using 3 internal control would have certain advantage, okay, 4 and that's what you just said. But at the same time I 5 also said that at the hearing, as well as in my papers, 6 that using internal comparison, especially when the 7 disease is rare and the comparison group is small, then 8 the rates would be very unstable. And if we don't see 9 any deaths in a comparison group, then the comparison 10 would not be appropriate. 11 What we are saying is if someone is not 12 exposed to benzene there would be no leukemia at all, 13 which is, of course, not true. 14 Q. But you still utilized that data. You had no 15 leukemias in your control group. 16 A. I utilized that data because in our contract 17 to the Chemical Manufacturers Association, we specified 18 that we have an internal control group. So, therefore, 19 1 analyzed the data both ways. But I spelled out both 20 the advantages and disadvantages of using internal 21 comparison. 22 Q. If you felt that the internal comparisons were 23 ineffective, for lack of a better term, for leukemia, 24 why did you. publish your data? 25 A. As I said, it has both advantages and 23 < < NOGARA REPORTING SERVICE 1 disadvantages. And we also have in our contract with 2 CMA that we would use an internal group as well as an 3 external group. And, therefore, I published both. And 4 I spell out the advantages and disadvantages of both 5 methods. I think it's very clear. 6 Q. I don't think it is, to be honest with you. 7 A. What is not clear? 8 Q. Because you testified to OSHA there's no 9 obvious reason to suspect that the comparison group was 10 any different from the exposed group. So why not use 11 internal controls for multiple myeloma as compared to 12 the general population? 13 A. Because we did not see any leukemia in the 14 comparison group, the internal comparison group. That 15 means zero. So anything compared to zero is an 16 elevation, is an excess, okay? That doesn't make 17 sense. Because what we are saying is in a general 18 population we should not expect any leukemia at all. 19 That's not true. 20 Q. That would be the same for multiple myeloma, 21 right? 'Would you not expect any multiple myeloma in 22 the general population? I don't understand the 23 distinction that you make. 24 It was permissible in your mind or acceptable 25 in your mind to utilize internal controls with no 24 < < NOGARA REPORTING SERVICE > 1 leukemias for your comparison with leukemias, but it 2 was not acceptable to use internal controls for 3 multiple myeloma because it was a smaller group. 4 A. Okay. 1 think you misunderstood what I did in 5 the paper. In my paper, back in 1987, the focus was on 6 leukemia. Okay. There was some interest on multiple 7 myeloma, but not the main focus. The main focus was 8 leukemia, okay? And for leukemia, I analyzed the data 9 using both the internal comparison as well as the 10 external comparison. 11 And the reason I used the internal comparison 12 was, when we design a study we were hoping that we have 13 a much larger group. And hopefully, that would give us 14 some stable rate for us to do some comparison 15 internally. Because internal comparison has certain 16 advantages. But it turns out, by the time we finished 17 the study, the comparison group was very small. And we 18 end up with no leukemia death in the internal 19 comparison group at all. 20 And 1 point out both advantages as well as 21 disadvantages of using internal comparison group. But 22 since 1 specified that analysis in my contract, in my 23 proposal to the sponsor, I went ahead and do that. 24 Besides, I also think it's a very interesting question, 25 whether to use internal comparison or not. When and 25 < < NOGARA REPORTING SERVICE > >. 1 how to use internal comparison. And this case 2 demonstrates -- my study demonstrates that 3 theoretically internal comparison groups are good. But 4 in reality it's very difficult to get a group large 5 enough to provide you with stable rates. And that's 6 the reason for doing that. 7 Q. Where in Part 1 or Part 2 of your 1987 8 publications did you explain that you were going to use 9 both internal controls and U.S. general population and 10 the explanation for why? Could you find that passage 11 for me? 12 A. Yes. Hold on for one second. Okay. On the 13 page in the first paper, on Page 378 14 Q. 368 or 378? 15 A. I said 378. 16 Q. The 1987 publication? . 17 A. Yes. Part 1. 18 Q. General results? 19 A. Yes. 20 Q. Okay. I don't have that page. But go ahead. 21 Let me see if it's in backwards or something. Yes, I 22 have it. I'm ready. It was in the wrong place. 23 A. On the right-hand side column the paragraph 24 that starts with "The SMR procedure." Do you find 25 that? 26 < < NOGARA REPORTING SERVICE > >. 1 Q. Yes. "The SMR procedure is usually 2 satisfactory"? 3 A. Yes. There I talk about the use of both 4 internal and external comparison groups and also talk 5 about the advantages and disadvantages of both. 6 Q. Did CMA suggest to you not to use internal 7 controls? 8 A. No, the proposal was prepared by us and we 9 basically said we would use both internal and external. 10 Q. Did you have any disagreements with CMA before 11 you published your findings in '87 about publishing the 12 findings on internal controls as opposed to utilizing 13 the general population as your control group? 14 A. Well, when you say "CMA," it's kind of like 15 CMA is one entity. Okay. But actually, it's not. We 16 have a committee from the industry that I contact in 17 terms of the study. And, of course, some people agree 18 with the way I analyze the data and some people may not 19 agree. 20 Q. And what particular individuals did not agree 21 with your use of internal controls? 22 A. I don't even remember the names of the 23 individuals. But, you know, that's what I have done. 24 And that's how I would like to publish. Nobody would 25 stop me on that. 27 < < NOGARA REPORTING SERVICE > 1 Q. Why couldn't you have used your entire group 2 as opposed to just intermittently exposed? Why 3 couldn't you have used all of your exposed for your 4 estimated for the multiple myeloma as opposed to the 5 general population? 6 A. That doesn't make sense to me, your question. 7 Q. What did you base your controls on for the 8 intermittent exposure group for multiple myeloma? 9 Maybe my question is poor. I assume that you -- did 10 you use your entire controls? You had 3,000 controls, 11 didn't you? 12 A. Yes. 13 Q. Did you utilize that or attempt to utilize 14 that to calculate your expected for multiple myeloma? 15 A. No, I used the U.S. general population. 16 Q. And the reason that you say you couldn't use 17 all of your control group was because it was too small? 18 A. It's too small, right. 19 Q. How many people were in the intermittent 20 exposed group? 21 A. Over a thousand, 1 think. 22 Q. And you could not have used 3,000 controls as 23 a comparison for your 1,000 intermittent control group? 24 A. I think you're getting confused. In 25 epidemiology, usually we have two groups. One is what 28 < < NOGARA REPORTING SERVICE > >. 1 we call the exposed group. That's the group where 2 we're interested in finding out if there is an 3 increased risk in certain diseases. The way to find 4 that out is to compare the rate of disease or mortality 5 in the exposed group to rates of the same disease in a 6 comparison group. Okay? Do you follow me so far? 7 Q. Yes, I do. 8 A. So I think you're mixing up the intermittent 9 group and the comparison group. Because the 10 intermittent-group is one of the exposed groups. In my 11 study 1 have two groups of people exposed to benzene. 12 One is what we call continuous exposed group. The 13 other one is intermittent exposed group. So those are 14 the two groups combined we call the exposed group. 15 And then I have the third group called the 16 comparison group. That group would not be exposed to 17 benzene indirectly, but may be exposed to only a low 18 level background. And I call that the comparison 19 group. That is our internal comparison group. But in 20 my study I also have the external comparison group, 21 which is the U.S. general population. 22 Q. Why couldn't you calculate then your expected 23 for your entire exposed population the three multiple 24 myelomas versus the control group? 25 A. Because the control group is too small. 29 < < NOGARA REPORTING SERVICE > >. 1 Q. Can you explain to me, Dr. Wong, how your '87 2 study failed to find any AMLs? 3 A. Well, there are number of reasons. No. 1, in 4 that study we have only seven leukemia deaths. So it's 5 not that many. There were a couple of them 6 unspecified. So it could have been AML. And we don't 7 know for sure. The other thing is the exposure may not 8 be high enough to really give us more than one or two 9 cases of benzene related AML. 10 In another report, which 1 published in 1995 11 using the NIOSH data, I find that in order to have an 12 increased risk of AML you have to have at least 200 PPM 13 a year, if not higher. That is pretty high exposure. 14 We just don't see that in my study. 15 Q. I understand what you published in '95. But 16 aren't AMLs the majority of leukemias in the adult 17 population? 18 A. The majority, did you say? 19 Q. Yes, sir. They compose the majority of 20 leukemias in the adult population? 21 A. It's not the majority. When you say the 22 majority, you mean more than 50 percent, right? I 23 don't think it's more than 50 percent. 24 Q. Aren't they the largest group of leukemias 25 found in adults, AMLs? 30 < < NOGARA REPORTING SERVICE 1 A. The largest type, right, yes. 2 Q. And would you agree with me that your study of 3 the various chemical plants found a statistically 4 significant relationship between other cell types of 5 leukemia and benzene exposure in '87? 6 A. There seems to be some indication, but the 7 increase was not comparing to the general population. 8 The increase was not statistically significant. 9 Q. What about comparing to your internal 10 controls? 11 A. Well, comparing to the internal controls since 12 we have zero, anything is significant compared to the 13 internal group. 14 Q. What was your SMR when you compared it to the 15 internal group? 16 A. We shouldn't use the word "SMR" if we use the 17 internal group. 18 Q. It should be the RR? 19 A. Yes, relative risk. 20 MR. DAMICO: What disease process are you talking 21 about? 22 MR. HARTLEY: Leukemia. 23 THE WITNESS: Hold on for one second. Well, for 24 leukemia it's infinite because you divide something by 25 zero it's infinite. It's not well defined. 31 < < NOGARA REPORTING SERVICE 1 MR. HARTLEY: Q. Did you find a dose response 2 relationship for leukemia in your study of the chemical 3 plants? 4 A. For leukemia, let me see. Yes. If you look 5 at Table 14 on Page 391. 6 Q. Of Wong 2? 7 A. Of 2, yes. 8 Q. Table 14, Wong 2. 39 -- all right. 9 A. You will see that under the column of relative 10 risk, it's undefined. Because again we divide 11 something by zero. But we were able to calculate a 12 chi-square, meaning that -- well, the chi-square is a 13 statistic that we used to determine to see if there is 14 a dose response relationship or not. And that turns 15 out to be not significant because the P value -- do you 16 see that? 17 Q. Yes. 18 A. The P value was 0.088. In order for it to be 19 significant it has to be less than 0.05. 20 Q. I must be missing something. On Table 14, 21 Page 3'91, entitled "Mantel-Haenzel Relative Risk and 22 Extension Chi-square for Lymplates and Hematopoietic 23 Cancer, Leukemia, non-Hodgkin's lymphoma, non-Hodgkin's 24 Lymphopoietic Cancer by Maximum Peak Occupational 25 Exposure to Benzene Adjusted for Age and Race." 32 < < NOGARA REPORTING SERVICE > 1 That's the table you're talking about? 2 A. Yes. 3 Q. And the leukemia undefined chi-square is 6.46, 4 right? 5 A. Are you looking at Table 14? 6 Q. That's what's published in the literature, 14, 7 yes. 8 A. The chi-square is .91. 9 Q. No, different document then, Doctor. I have 10 Page 391 of Wong 2. 11 A. That's interesting. Hold on for one second. 12 I know what the problem is. Table 14, the one you 13 looked at, must be a copy of the journal paper when 14 it's published. For some reason they duplicate Table 15 12. Go back to the previous page and look at Table 12. 16 Q. Yes. Table 12 is the same as Table 14. 17 A. Right. The one I have is a reprint which they 18 corrected Table 14. I'm sorry for the inconvenience. 19 But your Table 14 is actually the same as Table 12. 20 Q. What does Table 12 tell me? 21 A: 'Table 12 tells me that the chi-square is 2.91. 22 It's not quite significant. Looking at maximum P 23 exposure. And you are correct that when we look at 24 cumulative exposure it is significant. And that's 25 going to Table 12. 33 < < NOGARA REPORTING SERVICE > >. 1 Q. So when we look at cumulative exposure in 2 Table 12 there is a significance to the P .01 4evel 3 value? 4 A. Right. That is based on internal comparison. 5 And as you can see next to non-exposed, under the 6 column of observed deaths is zero. Do you see that? 7 Q. Yes, I do. 8 A. So that means we did not see any leukemia in 9 the small group of non-exposed people. 10 Q. Did you see it for non-Hodgkin's lymphopoietic 11 cancer? 12 A. Non-Hodgkin's lymphopoietic cancer, we have 13 two cases. 14 Q. And what was the chi-square for that, 364? 15 A. 364, which was not quite significant. 16 Q. It would be that there would be a 94 percent 17 chance that it would be -- within a six percent chance 18 that it would be within chance? 19 A. Right. 20 Q. Six out of a hundred it could be by chance? 21 A. Yes. 22 Q. So it's one-tenth of a point off in being 23 significant? 24 A. Yes. In the paper I also add that even for 25 that non significant trend, upward trend, that was 34 < < NOGARA REPORTING SERVICE 1 driven by the trend in the leukemia. Because leukemia, 2 if you see the ICD code after leukemia is 204 to 207. 3 Q. Yes, I do. 4 A. And that is part of the non-Hodgkin's lymphoma 5. cancer as well. Basically, I have included all the 6 lymphopoietic cancer in that group, minus 201. 201 7 being Hodgkin's disease. 8 A. Now, what was the cumulative exposure in parts 9 per million? Do you see that, 180 to 719 cumulative 10 exposure? 11 A. Yes. 12 Q. That would be under ICD 8, 200, 202 and 203, 13 which would include multiple myeloma, correct? 14 A. That would include -- let's see. 15 Q. 203 is multiple myeloma, correct? 16 A. 200 is lymphosarcoma which is part of the 17 non-Hodgkin's lymphoma. 202 is, again, non-Hodgkin's 18 lymphoma. And 203 is multiple myeloma. So it's a 19 combination of both non-Hodgkin's lymphoma, as well as 20 multiple myeloma. 21 Q. So you combined them for statistical purposes 22 in this analysis based on cumulative exposure, correct? 23 A. Yes, because 1 could not separate 202 from 203 24 in the analysis. 25 Q. What tables, Doctor, have your breakdowns of 35 < < NOGARA REPORTING SERVICE > >. 1 diseases? 2 A. Well 3 Q. Is that Page 358 or 368? Or where did you 4 identify the cell types? 5 A. The cell types? 6 Q. Yes, sir. 7 A. That would be the last table, Table 16 on Page 8 393. 9 Q. For the cumulative exposure classed, group of 10 200, 202 and 203 of 180 to 719 parts per million months 11 you had a relative risk of 2.23, correct? 12 A. What table are you looking at? 13 Q. I'm back on No. 12. 14 A. Okay. Can you repeat the question? 15 Q. Sure. For the class categories of 200, 202 16 and 203 with 180 to 719 parts per million months 17 cumulative exposure you had a relative risk of 2.23? 18 A. Yes, sir. 19 Q. And you also had a chi-square for that trend 20 of 14, correct? 21 A. 14 applies to the trend. The trend meaning 22 that comparing all four categories of exposure, 23 non=exposed, less than 180, 180 to 719 and 220 and 24 over. That 0.14 applies to all four. 25 Q. All four groups? 36 < < NOGARA REPORTING SERVICE > >. 1 A. Right. That's what we mean by trend, 2 including all four categories. And that clearly says 3 it's not significant at all. 4 Q. It was at .71. Where did you break out 5 what table do I need to look at for your SMR for 200, 6 202 and 203? Not the trend of exposure, but for the 7 cause of death. 8 A. For the overall exposed group you may want to 9 go to the first paper, Table 9. 10 Q. What page is that on? 11 A. On Page 374. If you look at the table, 12 probably in the middle of the table, you will see the 13 lymphatic and hemopoietic cancers. 14 Q. 200, 209? 15 A. Right. And underneath that the standard 16 grouping are the following, the following four groups, 17 lymphosarcoma, which is ICD Code 200. Do you see that? 18 Q. Yes. 19 A. And the next one is Hodgkin's disease, 201. 20 And then the third group is leukemia from 204 to 207. 21 Do you see that? 22 Q. Yes, sir. 23 A. And the last group is the residual, which is 24 other lymphatic tissue cancer which would include ICD 25 Code 202. By the way, that is a typo there. It should 37 < < NOGARA REPORTING SERVICE > >. 1 be 202 and not 22. And then 203. And then 208 is 2 really a very, very small group. For all practical 3 purposes the fourth category, other lymphatic tissue 4 cancer I would say would be 95 percent, would be either 5 202 or 203. And, as we said, 202 is non-Hodgkin's 6 lymphoma. And 203 is multiple myeloma. 7 In the old terminology they break 8 non-Hodgkin's lymphoma into two groups, lymphosarcoma 9 being ICD Code 200. And then what they called "other 10 lymphomas, which would be 202. And that's why I said 11 even back in the '80s I said certain groups we should 12 combine. And this is a very good example where we 13 should combine. Because in looking at 200 or 202 14 separately you don't get a good picture. You should 15 combine 200 and 202 in order to get a full picture of 16 non-Hodgkin's lymphoma. By the same time, if you want 17 to look at multiple myeloma, you've got to look at 203. 18 But in the statistical tabulation back then is 19 combined, 203 is combined with 202 in this group. So 20 what I'm trying to say is that that's why in my 21 analysis I was able to combine 200, 202 and 203 22 together, adding the first and the fourth category 23 together. 24 Q. Out of 202, 203 and 208, what percentage is 25 203, in your opinion? 38 < < NOGARA REPORTING SERVICE > >. 1 A. Forgetting about 208, I think 202 and 203 may 2 be roughly about the same, percentage wise. 3 Q. 50-50? 4 A. Roughly, yes. Well, it depends on age group 5 and so on. But roughly I would say about 50 and 50. 6 Q. What about in a younger population? 7 A. I really should look at the statistic before I 8 answer the question. 9 Q. Isn't multiple myeloma generally considered a 10 disease of older age? 11 A. Yes, older age. People in the older age group 12 would have much higher rates. 13 Q. Statistically it occurs in the sixth decade of 14 life? 15 A. Start ticking off in the 50s. 16 Q. That would be the sixth. decade of life. Zero 17 to 9 would be the first decade. So if you had a 18 population that was under 50, what would the percentage 19 breakdown be, in your opinion, between 202 and 203? 20 A. I don't know. 1'd have to look at it. 21 Q. Doctor, you've done some studies on gasoline? 22 A. Yes. 23 Q. And did you present at the Gasoline Symposium 24 in Canada? 25 A. I did not present any paper in Canada on 39 < < NOGARA REPORTING SERVICE > 1 gasoline. 2 Q. Were you present during the presentations? 3 A. Did you say Canada? 4 , Q. Yes, sir. 5 A. No, I did not. 6 Q. Were you present? 7 A. In Canada? 8 Q. In Miami. 9 A. I just found my note. It's in Miami. Miami 10 is not in Canada, for your information. 11 Q. Miami may be Canada if you live pretty south. 12 But anyway. 13 A. 1 think what you're referring to is 1990. 14 Early 1990s there was a symposium on gasoline. Yes, I 15 did present some results of my gasoline study at that 16 meeting, yes. 17 Q. Was there also a presentation about the 18 development of multiple myeloma from gasoline at that 19 symposium? 20 A. By whom? 21 Q I'm sorry? 22 A. By whom? By me? 23 Q. I can't remember what the guy's name was. 1 24 didn't bring my environmental health perspective with 25 me. 40 < < NOGARA REPORTING SERVICE > > 1 A. You have to tell me the name because I didn't 2 go to everybody's presentation, especially when the 3 meeting was held in Miami. 4 Q. It's warm down there. How about a guy named 5 Snadder? 6 A. Ralph Snadder from Exxon? 7 Q. Yes. Did he present some study on gasoline 8 and multiple myeloma? 9 A. He present -- well, okay. I think why you get 10 confused with Canada. He present the results of a 11 Canadian study that he did back then. He present some 12 preliminary results. And to be quite honest with you, 13 I don't remember -- I do not remember exactly what he 14 said at the meeting. 15 Q. Did you present at Benzene '95? 16 A. Oh, yes. We have -- actually, we have two 17 presentations at Benzene '95. You are referring to the 18 conference at the Rutger's University, right? 19 Q. Yes. Did you speak? Because I was there and 20 1 don't remember you speaking. 21 A. -As I said, we have two presentations. One 22 presentation was on cell type specific leukemia in the 23 - petroleum industry. My co-author actually made the 24 oral presentation. His name is Jerry Raabe. And at 25 the same time I also have a poster presentation on 41 < < NOGARA REPORTING SERVICE > 1 multiple myeloma. 2 Q. Did you do that with Dr. Lamm or did you have 3 your own poster? 4 A. I did not do any work with Dr. Lamm. 5 Q. Doctor, you presented in Miami on gasoline 6 workers? 7 A. 1 present some preliminary analysis, yes. 8 Q. And what was the topic of your presentation? 9 What were you looking for? 10 A. Well, I was looking for -- basically we 11 present mortality analysis by different causes of 12 death. Of course, the interest was leukemia. One of 13 the main interest was leukemia, because in gasoline we 14 have 2 to 3 percent of benzene. So the focus was on 15 leukemia. But at the same time kidney cancer was 16 another interest because of some animal studies. 17 Q. What population did you study, terminal 18 workers or distribution workers or what? 19 A. I think they mean the same thing, terminal 20 workers and distribution workers. 21 Q. As that who you studied? 22 A. Yes. 1 have two groups of people in the 23 gasoline study. One is what we call land based 24 terminal workers. The other one would be what we call 25 marine workers. These are the people who are 42 < < NOGARA REPORTING SERVICE > >. 1 responsible for the transportation of gasoline and 2 other petroleum products on barges or on ocean tankers. 3 Q. Were there criticisms of your control group 4 for having a small SMR for all causes in your 5 distribution workers? 6 A. In the land base group we found a low SMR for 7 all causes when we compared to the general population. 8 Q. What about your control group, did your 9 control group have low SMR also for all causes? 10 A. That was our control group. The control group 11 was the general population. 12 Q. The control group was from the United States? 13 A. Yes. 14 Q. Were there criticisms for utilizing that as a 15 control group by Dr. Enterline? 16 A. I think Dr. Enterline raised a question. 17 Actually, Dr. Enterline was one of the reviewers for 18 our study all along. So he was looking at the same set 19 of data over my shoulder for three, four years while we 20 were doing the study. And he raised the question, you 21 know. why we see such a deficit in our study. And we 22 tried to explain that. 23 One of the reasons may be the people who work 24 with gasoline may smoke less. Maybe that's one of the 25 reasons. In fact, there was another person who was an 43 < < NOGARA REPORTING SERVICE > > 1 outside reviewer for the conference. I think that was 2 Dr. McLaughlin at the National Cancer Institute at that 3 time. And he gave us some very interesting insight. 4 He said that the general population, sometimes 5 you have people who have very high rate of heart 6 disease, as well as high rate of non-malignant 7 respiratory diseases and so on. And maybe that 8 accounts for the deficit we saw in our study. 9 But regardless of what the explanation is we 10 also have the second group. The second group is the 11 marine distribution workers. And certainly for that 12 group we don't have that problem. We don't see an 13 unusually low SMR. 14 Q. What did Dr. Infante think of your control 15 group? 16 A. The general population?, 17 Q. Yes. 18 A. I don't know what he thinks of that. 19 Q. Did he explain that to you when you were doing 20 the study, what his thoughts on your control group 21 were? 22 A. He was not a reviewer of our study. I did not 23 ask him for consultation or advice. 24 Q. How about at the meeting on the gasoline 25 symposium, did he comment on that meeting on the 44 < < NOGARA REPORTING SERVICE > >. 1 control group? 2 A. He commented on a whole bunch of papers. And 3 I don't remember exactly what he said on our paper. 4 Q. Now, we got into all this because of 5 specificity and Bradford Hill. Would you agree that 6 specificity must not be overemphasized in determining 7 the causation of a particular substance to a particular 8 disease? 9 A. I don't quite understand exactly what you mean 10 by "cannot be overemphasized." 11 Q. Bradford Hill says in his 1965 article that 12 established the criteria, "We must not, however, 13 overemphasize the importance of the characteristic." 14 That being specificity. Do you agree that some 15 epidemiologists overemphasize the role of specificity 16 in determining a causal relationship? 17 A. I don't think so. I think when Bradford Hill 18 used the term "specificity" in that content he meant 19 that a chemical may be able to cause more than one 20 disease. We cannot say that simply because Chemical X 21 cause& Disease Y that would imply that Chemical X does 22 not or is not capable of causing another disease. I 23 think that's what he meant. Okay? 24 And I totally agree with that. Certainly 25 smoking can cause more than lung cancer. It can cause 45 < < NOGARA REPORTING SERVICE > 1 other disease as well. But the evidence has to come 2 from specific data, specifically related to death 3 disease. Okay? If you show me that benzene can cause 4 multiple myeloma specifically, you have data on that, 5 then I will buy it. 6 Q. That also goes right to Bradford Hill's other 7 point. And that is that you should not be totally 8 wedded to statistical significance. If there's an 9 elevated risk that is not statistically significant on 10 a recurring basis then there should be some 11 consideration of an association and causality. Would 12 you agree with that? 13 A. Absolutely. And that's why I've been doing 14 this meta-analysis. And that is when you have a small 15 study there may be an elevation. But the excess or the 16 elevation may not be statistically significant. And 17 that can be due to two things. One is the sample size 18 is not enough, not big enough to pick up that 19 difference. Or it can be that indeed there is no 20 relationship. 21 And I totally agree with Bradford Hill that if 22 you do see an elevation from study to study, then there 23 is something. Now, meta-analysis would be able to 24 handle that because if you see some consistent 25 elevation when you combine all the studies into the 46 < < NOGARA REPORTING SERVICE > >. 1 meta-analysis, you will certainly see a statistically 2 significant increase because you increase the sample 3 size by doing a meta-analysis. 4 Q. Do you also agree with Bradford Hill's 5 statement that, "No formal test of significance can 6 answer those questions of causation. Such tests can 7 and should remind us of the effects that the play of 8 chance can create and they will instruct us in the 9 lengthy magnitude of those effects. Beyond that they 10 contribute nothing to the proof of our hypothesis." Do 11 you agree with that statement? 12 A. I agree that when you do a statistical test by 13 and in itself is not the whole thing. You've got to 14 look at the whole picture. And that's the reason why 15 in 19- --I believe 1965, Bradford Hill formalized the 16 criteria. So you should not just look at one 17 statistical study. You really have to look at what 18 kind of chemicals are we talking about: how specific 19 exposure was in the study; how long the observation 20 period is, and how many studies reporting an increase 21 or a lalc of an increase. You have to take all that 22 into consideration. 23. a. Which goes back to the idea that if you have 24 elevated non-statistical risk for a substance on a 25 recurring basis and you have biological plausibility as 47 < < NOGARA REPORTING SERVICE > 1 well as a temporal relationship and you have strength 2 from a few studies, or a great deal of studies, that 3 you can make a determination even though there is no 4 statistical significance of any epidemiologic study 5 that Substance X caused Disease Y, correct? 6 A. I agree completely. It's unfortunate that you 7 did not get my November 25th report earlier because 8 that's exactly what 1 did. I analyzed the literature, 9 including some of my own studies on tissue of benzene 10 exposure and multiple myeloma and basically applied the 11 criteria. And the bottom line is when you look at all 12 the studies taking into strength of association, 13 consistency of association and so on, those response, 14 et cetera, the conclusion is there is no connection 15 between benzene exposure and multiple myeloma. 16 Q. I understand your conclusion. 17 A. I just want to explain to you how you get 18 there. 19 Q. We'll get to that in a little bit. How many 20 of the Bradford Hill criteria do you feel as an 21 epidemiologist are needed to evaluate and establish the 22 likelihood that Substance X caused Disease Y? Do we 23- need all of them? 24 A. I don't think we can specify we need four out 25 of six or five out of six or all six. I don't think we 48 < < NOGARA REPORTING SERVICE > >. 1 can do that. I think we really have to use our 2 professional judgment to see whether there is a 3 consistent picture, whether the risk is high enough for 4 us to rule out confounding, whether there is any dose 5 relationship and so on. I don't think we can quantify 6 that. 7 Q. Is a dose response relationship enough to 8 establish causality? 9 A. It's one of the criteria that we should use. 10 Q. Is that in and of itself enough? If we have a 11 temporal relationship is a dose response relationship 12 enough to establish causality? 13 A. I think you also need consistency from 14 different studies. 15 Q. How many studies do we need to be consistent 16 on a dose response basis with a temporal relationship 17 before we can say there's causality? 18 A. Again, 1 don't think we can quantify that. 19 And it's somewhat meaningless to ask how many studies. 20 How big are the studies? Are we talking about small 21 studies? If we're talking about small studies, we need 22 a lot of them. If they are good, well-designed large 23 studies, maybe we don't need as many. I think it all 24 comes down to when you combine the studies together 25 what does the mete-analysis tell you. How big is the 49 < < NOGARA REPORTING SERVICE 1 95 percent competency. 2 Q. Can one study negate another study by itself? 3 If we have one positive study and we have one negative 4 study, does the negative study tell us that the 5 positive study was just by chance so there's no 6 relationship? 7 A. If the two studies are about the same size, 8 the same quality, then I would say we have inconsistent 9 findings and we don't know what's what. 10 Q. Which would require the person doing the 11 evaluation to utilize their own professional judgment 12 and come to a conclusion as to whether Substance X 13 caused Disease Y, correct? 14 A. I'm not so sure. If you have two studies and 15 assuming that the two studies are the same in terms of 16 quality and sample size and so on, if you have 17 conflicting information, 1 don't see how one can come 18 to a conclusion either way. 19 Q. But that would be the professional judgment of 20 the evaluator, would it not? You may say you cannot 21 come to a conclusion. But someone else may utilize 22 their own professional judgment and say, "I think one 23 - study is either flawed. I think one study is better 24 than another one." Et cetera. For whatever reason 25 they may utilize that same data and their own 50 < < NOGARA REPORTING SERVICE 1 professional judgment and come to a different 2 conclusion than you would, correct? 3 A. I think we're changing the question slightly 4 now. Because when I said we cannot come to a 5 conclusion, that's based on the assumption that the two 6 studies are similar in terms of quality and sample 7 size. That's what I said. That's why I gave you that 8 preamble. Now, if you say that one study is clearly 9 better than the other, of course, we would put more 10 weight on the better study. That's just logical. 11 Q. Were you criticized by Dr. Savidz from North 12 Carolina as a result of your 1995 publication in 13 Occupational and Environmental Medicine entitled "Risk 14 of Acute Myeloid Leukemia, Multiple Myeloma and Workers 15 Exposed to Benzene in Regard to Specificity"? 16 A. I do not consider that as criticism. 17 Basically he said that I only looked at AML in my 18 analysis. And he said that if I add up all the other 19 leukemia cell types then I would also find a 20 statistically significant increase in cell types other 21 than AML. 22 And 1 have responded to his inquiry, I think, 23 quite adequately in my response published alongside 24 with his letter. And that is it's meaningless to 25 combine CML, ALL, and -- what was the other one -- the 51 < < NOGARA REPORTING SERVICE > >. 1 other three kinds. ALL, CLL and CML together. Okay? 2 His whole approach is if we add all those numbers 3 together we would have enough cases to justify the 4 analysis from the statistical point of view. 5 And my response to him was a medical question, 6 a health related question is more than just statistics. 7 It would be meaningless to combine them because CML is 8 not the same as ALL. ALL is basically a childhood 9 leukemia. CLL is not. And they have different risk 10 factors and so on. Okay? And I point out to him that 11 1 did not say that my analysis proved that there is no 12 relationship between benzene exposure and CML. I 13 simply say that the data in that study did not provide 14 us with enough data, sufficient data to show either 15 way. 16 In order to look at the relationship between 17 benzene exposure and CML or other cell types we have to 18 go to other studies. But I disagree with him that an 19 analysis is based on statistics only. An analysis must 20 be able to make biological sense. And I think he 21 failed. in that. 22 Q. Now, when he wanted to combine all the 23 leukemia, as you did in 1987 and as you did before OSHA 24 from the Pliofilm cohort, all those leukemias 25 originated from the same stem line, didn't they, or 52 < < NOGARA REPORTING SERVICE > >. 1 don't they? 2 A. They don't originate from the same, what we 3 call stem cells. That is one of the confusions. And 1 4 sympathize with you because I had a tough time trying 5. to understand that in the early days. 6 In the early 1980's, when people referred to 7 stem cell, most people thought that there is only one 8 kind of stem cell. But it turns out that later on, at 9 least I find out later, that there are different kinds 10 of stem cells. And some are more committed to a 11 specific cell line than the other. So when you say all 12 the leukemia cells coming from the same stem cell, it's 13 not quite true. They are coming from different cell 14 lines. And Professor Irons, at the University of 15 Colorado, certainly has written a lot about that. He 16 would be the expert to talk about that. 17 . Q. Would you agree with me that CML and AML come 18 from the same fluid-potent stem cells? 19 A. I'm not so sure I would agree with you. I 20 think specifically Richard Irons wrote a paper, or more 21 than the paper, saying that the metabolites of benzene 22 would only have an effect on AML, but not on CML. 23. Q. What does the metabolite of benzene, the 24 hydroquinone, have to do with CML? Don't other 25 investigators believe that there are other avenues that 53 < < NOGARA REPORTING SERVICE > >. 1 benzene utilizes to attack the stem cells in creating 2 the malignancy? 3 A. I don't know about that, whether there are 4 some people who believe in that or not. Certainly I 5 think Richard Irons is the authority on stem cell. And 6 he doesn't think that benzene can cause other cell 7 types besides AML. And from the epidemiologic point of 8 view, if indeed benzene can cause CML, then we should 9 be able to see that in epidemiologic studies. I have 10 yet to see a study linking benzene directly to CML. 11 Q. Didn't you have some CMLs in your study, 12 Doctor? 13 A. I think you misunderstand or you may not be 14 able to interpret epidemiologic study. 15 Q. That may be true, because I'm just a country 16 lawyer, Doctor. I have no idea what I'm talking about, 17 to be honest with you. I'm just wondering, didn't you 18 have some CMLs in your study? 19 A. Clearly if you're implying that if I have CMLs 20 in my study then, therefore, benzene must be able to 21 cause CML. 22 Q. I'm not asking you that. First of all, did 23 you have CML in your study? 24 A. Yes, I have CML in my study. And that is 25 quite to be expected because whenever you look at a 54 < < NOGARA REPORTING SERVICE > 1 group of 3,000, 4,000 individuals, let's say, lawyers. 2 If I have 3,000 lawyers and I observe them over 40 3 years I can tell you for sure I would expect to see 4 some CMLs. 5 Q. How many CMI-s did you have? 6 A. I don't remember. Maybe I have two. Not that 7 many cases. 8 Q. What number is CMI, what 1CD 8 number? 9 A. I think it's 205.1. 10 Q. If it's 205.1 it looks like you had two of 11 them. 12 A. That maybe right. 13 Q. Now, then, weren't CMLs found in the Chinese 14 cohort? 15 A. It depends on which study, which paper you're 16 talking about. 17 Q. What about the most recent update with the 18 lead author being Hayes? 19 A. I don't have that paper in front of me, so I 20 won't be able to tell you the numbers. 21 Q. Do you have the Quinn or the Travis update in 22 front of you? 23 A. No, I don't. 24 Q. Do you have any of the Chinese literature in 25 front of you? 55 < < NOGARA REPORTING SERVICE > 1 A. I have one, which is Yin, 1996. By the way, 2 in that paper, Yin certainly analyzed multiple myeloma 3 as a separate group. And he didn't find a single group 4 of multiple myeloma in a cohort of 70,000 workers 5 exposed to benzene. 6 Q. Multiple myeloma is a very, very, very rare 7 disease in China, is it not, Doctor? 8 A. Yes, but at the same time if benzene can cause 9 multiple myeloma certainly you would see at least one 10 or two cases. 11 Q. What if the Chinese do not develop multiple 12 myeloma generally because of genetic differences, you 13 would not see a rise regardless of exposure, would you? 14 A. Are you saying that we don't see multiple 15 myeloma at all in Chinese? 16 Q. No, I don't think you can say that. I think 17 you can say, though, that it is very, very rare. 18 A. Rare to the extent that we don't see a single 19 case in 70,000 people over many years? 20 Q. Doctor, how many multiple myeloma are in the 21 controls? 22 A. In what study? 23 'Q. In any of the Chinese study? 24 A. I'm not so sure they list that. 25 Q. He had to list it if he made a comment that he 56 < < NOGARA REPORTING SERVICE 1 didn't have any multiple myeloma in his exposed, didn't 2 he? 3 A. No, he did not list that. 4 Q. Did he just make the comment he didn't find 5 any multiple myeloma in his control group? 6 A. Yes, in the table he simply said it is zero. 7 Q. Did he list the expected? 8 A. He didn't list the expected. 9 Q. In the body of the report? 10 A. 1 don't see that. 11 Q. At least not in the '96 body? 12 A. Not in the paper that I have in front of me. 13 MR. DAMICO: The court reporter needs like a 14 two-minute break. 15 MR. HARTLEY: Okay. 16 (Brief recess.) 17 MR. HARTLEY: Q. Doctor, in your abstract to Part 18 2 of the 1987 publication you indicate there is a dose 19 response relationship between cumulative exposure and 20 non-Hodgkin's lymphopoietic cancer was a borderline 21 significant P equaling to point zero P6. 22 A. We looked at that before. 23 Q. We talked about that before? 24 A. Yes. 25 Q. You also mentioned -- I want to get one thing 57 < < NOGARA REPORTING SERVICE 1 clear. Are you saying that AML and CML do not 2 originate from the same stem cells, as far as you 3 understand? 4 A. I better not answer that question because 1 5 really should. go back and look at some of the Richard 6 Irons papers. What I said before was that the 7 metabolite of benzene only picked on, according to 8 Richard Irons, only picked on AML, not CML. 9 Q. I'm not concerned with what the metabolite 10 does. I'm concerned about what the stem cell AML and 11 CML originate from. 12 A. I don't know if they originate from the same 13 committed stem cell or not. That's what I don't know. 14 Q. How about the primitive stem cell, are you 15 familiar with whether AML and CML derive from the 16 primitive uncommitted stem cell, the same one? 17 A. Well, when you use the term "uncommitted 18 primitive stem cell" then everything has to come from 19 that. 20 Q. So everything comes from the primitive 21 uncommitted stem cell, does it not? 22 A. Right. But that theory is really out of date 23 now. 24 Q. It is out of what? 25 A. Out of date. That we should look at committed 58 < < NOGARA REPORTING SERVICE > > . 1 Q. Are you going to testify that benzene does not 2 cause non-Hodgkin's lymphoma? 3 A. In this case? 4 Q. Yes, in this case. If that question is put to 5' you by me during cross-examination, would you say that 6 benzene does not cause non-Hodgkin's lymphoma? 7 MR. DAMICO: Well, I'll object on relevance. 8 MR. HARTLEY: Q. That means you can answer the 9 question during deposition. 10 A. I have not done a complete search on 11 non-Hodgkin's lymphoma and benzene exposure. My 12 assignment coming here today was to look at benzene 13 exposure and multiple myeloma and that's what I have 14 done. If you can convince the lawyers to authorize me 15 to do work on non-Hodgkin's lymphoma and benzene 16 exposure, if I have time I'll look into that. 17 Q. What would you charge Mr. Damico by the hour 18 to do that, Doctor? 19 A. I charge $380 an hour. 20 Q. I kind of doubt he's going to want you to do 21 that. 22 MR. HARTLEY: But, Dave, if you do, let me know. 23 MR. DAMICO: I'll send you half the bill. 24 THE WITNESS: Maybe if you are interested in 25 finding out the truth, maybe you can pay me to do that. 60 < < NOGARA REPORTING SERVICE > >. 1 Just kidding you. You can ask the next question. 2 MR. HARTLEY: a. Thank you. I was going to 3 respond with a smart remark but I elected not to. 4 Does not multiple myeloma come from the same 5. cell line as non-Hodgkin's lymphoma? That was the 6 basis for the first question. 7 A. I don't know whether they're coming from the 8 cell line, the same, and what stem. I think we should 9 really go back and look at that. And again, I'm not an 10 expert in the area. So I would not be the one who 11 would even do that research. I think, again, someone 12 in the area of hematology or toxicology would be able 13 to answer that question and not me. 14 Q. How do you know what disease category each of 15 the various malignancies should be grouped in when you 16 do your analysis? Do you just go through and say this 17 is 2 and this is 203 so we can put them together or do 18 you have some understanding of that process? 19 A. I have some understanding based on the review 20 papers done by people like Richard Irons and some other 21 hematologists. As I point out to you in my report, 22 there's some textbooks on hematology, as well as review 23- articles, indicating that multiple myeloma, 24 non-Hodgkin's lymphoma and leukemia are different 25 diseases. And when 1 look at the publications by the 61 < < NOGARA REPORTING SERVICE > 1 National Cancer Institute, multiple myeloma is a 2 distinct entity. They don't put multiple myeloma and 3 non-Hodgkin's lymphoma together as one analysis. 4 Q. One final question about the Chinese study and 5 then we'll move on. When you are doing a study of a 6 population to determine whether there's an elevated 7 risk of a particular disease you certainly want to look 8 at a population that develops that particular disease 9 entity before determining whether or not that would be 10 the logical cohort to study, don't you? 11 A. I don't follow your question. It's kind of 12 long, Can you repeat that? 13 Q. Sure. When the study of the Chinese cohort 14 was made going in, if we know that multiple myeloma is 15 a very rare disease among the Chinese, it would not 16 make one surprised to learn that there is either a 17 deficit or no increase in multiple myeloma for the 18 benzene exposed population, would it? 19 A. When you say "rare," I don't know how rare 20 that is. Because they did not report -- at least based 21 on the article I have in front of me -- we don't know 22 what the expected would be. 1 guess the proper 23 question to ask is in the Chinese cohorts of 70-plus 24 thousand workers exposed to benzene, that's what they 25 claim, I would like to know how many multiple myeloma 62 < < NOGARA REPORTING SERVICE > 1 they expect to see in that cohort. Okay? If they only 2 expect one case or two cases and they don't see any, 3 then I would say that's reasonable. But on the other 4 hand, if they expect to see five or six or even more 5, cases and they don't see any, then I would say why is 6 that? I want to look into that. 7 Q. Is it better to study populations with a 8 greater disease frequency or one that has less of a 9 frequency of the disease you're studying? 10 MR. DAMICO: That's a difficult question. Did you 11 say "better"? 12 MR. HARTLEY: Q. Yes. Would you study a group of 13 30-year-old males for prostate cancer? 14 A. I think I understand what you're asking. If 15 you study a population in which the rate of disease is 16 rare, then you need a large population in order to come 17 up with something statistically meaningful. Is that 18 what you're asking? 19 Q. Yes. 20 A. By the same time -- well, I shouldn't say 21 that. On the other hand, if you study something that's 22 very common in the population, most likely that implies 23. theft disease in that population would have many, many 24 different risk factors. And that would create a 25 different set of problems. 63 < < NOGARA REPORTING SERVICE > 1 Q. What about smoking, is smoking an example of 2 that? 3 A. Well, it depends on what disease you're 4 talking about. Certainly for tobacco-related diseases 5 that would be a problem. 6 Q. For lung cancer. There are many risk factors 7 for lung cancer other than smoking? 8 A. Was that a question or did you just comment on 9 that? 10 Q. That's true, right? 11 A. What was true? 12 Q. That there are other risk factors for lung 13 cancer other than smoking? 14 A. Oh, absolutely. 15 Q. And that would be an example of what you're 16 talking about? 17 A. Yes. 18 Q. Are there risk factors for multiple myeloma? 19 A. Yes, I think so. 20 Q. Could you delineate those for me? 21 A.' I think I might have listed a number of them 22 as an example in my report, dated 23 Q. I didn't get that report yet. I'll have to 24 disconnect and call down. 25 A. On Page 3. 64 < < NOGARA REPORTING SERVICE > 1 Q. Of which report, Doctor? 2 A. The report dated November 25th, 1997. 3 Q. Why don't you list the risk factors for me. 4 A. Based on a number of studies, at least some 5 risk factors has been identified in those studies. And 6 that would be virus, use of laxatives, a history of 7 diabetes, and, well, talking about cigarette smoking, 8 including cigarette smoking. Those would be examples 9 of risk factors of multiple myelorna. 10 Q. How does cigarette smoking relate to multiple 11 myeloma? I believe you testified back in '95 that 12 based upon your review of the literature at that point 13 in time you did not think that there was a sufficient 14 connection between smoking and multiple myeloma for 15 someone to opine that there was a relationship or a 16 risk factor. 17 A. I don't know where -- I just list them as risk 18 factors that have been identified by some people in 19 their study. I did not say that the relationship is 20 definitive. Again, the lawyers have not asked me to do 21 a complete research on those issues. And 1 didn't 22 spend much time on it. In fact, those are the studies 23 that I come across when I look at benzene and multiple 24 myeloma. They also talk about other risk factors. I 25 just list them as an example that multiple myeloma may 65 < < NOGARA REPORTING SERVICE > >. 1 A. When you talk about studies, the only thing I 2 can think of would be the study published by Rinsky 3 R-i-n-s-k-y. Rinsky in 1987. But, as you know, 4 subsequently, I have updated the study and published in 5 1985. And basically we found that there was no 6 relationship. Even in the 1987 Rinsky paper he did not 7 find a dose response relationship between benzene 8 exposure and multiple myeloma. 9 Q. But he nevertheless found an elevated 10 didn't he find as SMR of three point something or four 11 point something that was statistically significant? 12 A. Close to four. 13 Q. And that cohort indicated or that study 14 demonstrated that there was an elevated risk of 15 multiple myeloma in this cohort, did it not, in '87? 16 A. On the surface it does. But at the same time, 17 if you go back to the Hill criteria, then the Rinsky 18 1987 paper only meets the first criteria. And that is 19 a significance of association. It does not meet 20 consistency. Because when you look at other studies 21 you don't find that. It does not meet dose response, 22 because even Rinsky himself admits that there was no 23 dose response relationship. 24 Q. Well,, you can't make the dose response 25 relationship on four cases, can you? 67 < < NOGARA REPORTING SERVICE > > 1 A. Well, but you based the findings on four 2 cases, the same four cases. You said that was a 3 significant increase. 4 Q. Is there enough statistical, Doctor, for four 5 cases to provide you with a dose response relationship 6 if you're breaking up the doses as Rinsky did? 7 A. Well, actually there are two questions. One 8 is indeed you see an upward trend, but the trend may 9 not be statistically significant. And that may be 10 because your study is too small. It does not provide 11 you with enough cases. So that you can say that there 12 is a significant trend. But there was no trend at all. 13 I'm not talking about whether it's significant or not. 14 Three out of the four cases were short-term workers. 15 One guy worked there for like four days. 16 Q. 1 understand that. But when you made your 17 comments on the study to Gulf Canada in 1983, didn't 18 you indicate that the cohort definition appeared 19 adequate? 20 A. The cohort definition of what study? 21 Q Of the Pliofilm study. 22 A. Yes. 23 Q. Didn't you indicate that the cohort definition 24 appeared adequate and the cohort definition included 25 anyone who worked on the job site for one day? That's 68 < < NOGARA REPORTING SERVICE 1 what you told Gulf Canada, that that definition was 2 adequate. 3 A. Was inadequate, right? 4 Q. No, adequate. A-d without an I-n. Adequate. 5 Wasn't that your statement to Gulf Canada in 1983? 6 A. You have to show me the document in 1983. I 7 want to read the whole paragraph. I don't want you to 8 pick out one sentence and read that to me. I don't 9 know what content that would be. 10 Q. Do you remember making comments on the NIOSH 11 study of leukemia and benzene studies to Gulf Limited 12 in August of 1983? 13 A. I remember making some comments. I don't 14 remember the date. It was sometime ago. But 15 certainly, if you want me to explain to you, you've got 16 to show me the document. 17 Q. It's hard when I'm on the East Coast and 18 you're on the West Coast. But the first sentence of 19 your second paragraph says, "The cohort definition 20 appeared adequate." At that point in time you did not 21 make a comment or contest the adequacy of the cohort 22 definitions. Now, you did go on to point out though, 23 and I will be fair to you, that, yes, there were some 24 people who were only on the job site for a very, very 25 short period of time. But nevertheless, you indicated 69 < < NOGARA REPORTING SERVICE > 1 that the definition of the cohort appeared adequate. 2 Do you recall that? 3 A. Yes, but that has nothing to do with the dose 4 response relationship. What I'm trying to point out to 5 you in terms of the multiple myeloma result is that 6 three out of the four cases were short-term workers. 7 Certainly that doesn't give us any indication of those 8 responses at all, significant or otherwise. 9 Q. If three out of the four were short-term 10 workers, when you reviewed the results for Gulf Canada 11 why didn't you say that the cohort definition is 12 inadequate because it over includes and has the 13 possibility of including people who were only on the 14 job site for one day? 15 A. No. No. No. I think you're mixing up two 16 things. One is the cohort of definition. The cohort 17 of definition, you can define it any way you want. 18 Most people would define it as anyone who worked at the 19 site for at least three months, six months or a year. 20 But NIOSH decided to use one day. Okay? And assuming 21 that indeed, you know, those people worked there for at 22 least one day, that is a correct definition. That's 23 okay. But that means they have to include a whole 24 bunch of short-term workers. 25 I would not design a study as such because you 70 < < NOGARA REPORTING SERVICE > 1 are wasting your money and effort on a lot of people 2 with short-term exposure at that particular site. But 3 those people may have a lot of other exposures from 4 elsewhere that we don't know, exposure to other 5 chemicals. What I'm saying is in terms of the multiple 6 myeloma finding there are only four cases. But three 7 of them work at that site for a very short period of 8 time. How can we say that those three multiple myeloma 9 deaths were related to the exposure at the site when 10 they were short-term workers, when they spent most of 11 their time, their career elsewhere? That's a 12 distinction between a cohort definition and a causation 13 analysis. 14 Q. I understand that. But since that was the 15 definition of the cohort we're stuck with those one- or 16 two-day workers and their development of multiple 17 myeloma, are we not? 18 A. Yes. But the interpretation has to be 19 appropriate. If in the Rinsky study they see all four 20 cases among people who worked there for 20-some years 21 with a lot of exposure, then I say maybe there is 22 something between benzene exposure and leukemia. But 23 that's not the case. It's the opposite. 24 Q. Do you think that no one considers benzene to 25 be a risk factor for multiple myeloma? 71 < < NOGARA REPORTING SERVICE > >. 1 A. Certainly you do. Certainly the experts you 2 hired do. I can't say that statement. That doesn't 3 make sense, that question. 4 Q. Well, do you believe that it is recognized 5 within the scientific community that benzene is not a 6 risk factor for multiple myeloma? 7 A. Certainly not based on properly done 8 epidemiologic studies. 9 Q. Do you know who Bernie Goldstein is? 10 A. Yes, .I know who he is. 11 Q. Are you familiar with his work? 12 A. Basically he's a toxicologist. I don't think 13 he's an epidemiologist. I really don't know his work 14 that well. 15 Q. Would you agree that he is well-known in the 16 area of benzene related diseases? 17 A. He is well-known in terms of toxicology 18 related to benzene, not on benzene epidemiology. 19 Q. And he has expressed his opinion, has he not, 20 in, I believe, the early 1990s, that benzene more 21 probably than not causes multiple myeloma? Has he or 22 has he not done that first? And then you can explain. 23 A. I have seen some of his editorial comments. 24 And basically I think he relies on a couple of case 25 reports to support his opinion. And he may even rely 72 < < NOGARA REPORTING SERVICE > >. 1 on the Rinsky 1987 paper. But that is incomplete 2 literature review. And most of that is based on case 3 reports. And I don't think case reports are 4 appropriate for causation analysis. 5 Q. He looked at the biological possibility, did 6 he not, as a toxicologist? 7 A. Yes, it's kind of interesting. When you look 8 at the biological possibility certainly if you find 9 something, then you look at whether it is possible or 10 not. But saying something that's possible is no proof 11 of causation. The bottom line is if indeed you do have 12 a relationship between benzene exposure and multiple 13 myeloma you should be able to see that in epidemiologic 14 studies. 15 Q. But isn't it more rare than leukemia? 16 A. It may be rarer than leukemia, but, still, we 17 have so many studies. When you have some time, read my 18 report. You will see that the studies indicate that 19 there is no increased risk. It is not just based on 20 one or two studies. 21 Q. "We were talking about the fact of the 22 scientific community, would there be anyone who would 23 think it would be related. And naturally you pointed 24 to me, as the plaintiff's counsel, and the experts that 25 I have hired. 73 < < NOGARA REPORTING SERVICE > 1 Goldstein is not hired by either side and was 2 in fact a consultant to API at one time. And he has 3 concluded that multiple myeloma is probably caused or 4 is more likely than not to be caused by benzene 5 exposure. 6 A. I don't think that statement is completely 7 true. Dr. Goldstein certainly has some project from 8 API in the past. He might have been a consultant to 9 API. But at the same time, you know, a lot of us have 10 been consultants to OSHA as well. I don't think how 11 much that can buy us. The part that you said Dr. 12 Goldstein was not hired by 13 Q. Either side. 14 A. -- either side, I don't think that's true. I 15 think at least in one case I ran into him. He was 16 hired by the plaintiff attorneys. 17 Q. Have you ever testified for the plaintiff in a 18 benzene case, Dr. Wong? 19 A. No. 20 Q. Have you ever been asked to? 21 A. Some people asked for my opinion and I 22 indicated to them. And I think that that was the end 23 of the conversation. 24 Q. Have you ever advised plaintiffs counsel that 25 have asked you that there would be a conflict of 74 < < NOGARA REPORTING SERVICE > > 1 interest in light of all the work you do for industry? 2 A. I don't see that as a conflict of interest. 3 Basically I just report on the findings that I have 4 done, as well as interpret similar studies done by 5 other people. 6 MR. HARTLEY: I'm going to stop for five or ten 7 minutes here. I'm going to stop and see where the 8 report is because they haven't brought it up. 9 MR. DAMICO: Let's go off the record. 10 (Discussion off the record.) 11 (Brief recess.) 12 MR. HARTLEY: Q. When you do your meta-analysis 13 and you combine various epidemiologic studies, do you 14 review those studies to note whether they are of equal 15 value? 16 A. When I combine those studies, for example, 17 refinery studies, the cohort definition would be quite 18 similar. And the exposure would be quite similar, as 19 well. I don't combine refinery studies with paper mill 20 studies, for example. I don't combine refinery studies 21 with truck drivers, so to speak. So at least there is 22 some common exposure across the studies. 23 Q. What about is latency important in those 24 various studies if you're looking at them when you have 25 a younger cohort as opposed to an older cohort in the 75 < < NOGARA REPORTING SERVICE > >. 1 various studies you're going to attempt to combine? 2 A. Certainly that creates somewhat of a problem 3 when you combine the studies because different studies 4 use different cut points for latency or for exposure, 5 length of employment, length of exposure and so on. 6 And that being the case, it would be impossible to 7 combine different studies by length of exposure and 8 length of latency or by any other exposure 9 measurements. That is somewhat of a problem. 10 But to the extent that when you do a 11 meta-analysis you also need to look at individual 12 studies. And we did that. And we do some, what we 13 call stratified meta-analysis. In other words, we look 14 at the refinery workers as one group. We look at the 15 distribution workers as another group and so on and see 16 where there are any major differences between different 17 groups. 18 Q. How do you account for short follow-up time in 19 the various cohorts? 20 A. We tried to do that. In fact, I don't know 21 whether - 1 assume you may not have a copy of our 22 multiple myeloma meta-analysis paper which was 23 published last month in the Journal of Regulatory 24 Toxicology and Pharmacology. In that paper, actually, 25 we tried to group the studies of short-term follow up, 76 < < NOGARA REPORTING SERVICE > > 1 well, shorter follow up. Let's see. That's Table 6 in 2 that paper. 3 We grouped the studies with length of follow 4 up less than 20 years as one group. And we grouped 5 studies whose length of follow up is 20 years or more 6 as another group. And in that table I present the SMR 7 for both groups of studies and they are about the same. 8 In one group the SMR was 0.99. In the other group the 9 SMR was 0.92. So they are very comparable. 10 Q. When you are looking for confounders in a 11 particular study; suppose we're doing a study for the 12 relationship, for simple terms, of cigarette smoking 13 and lung cancer. What types of confounders would you 14 be looking for when you are actually doing the study? 15 A. Well, if the interest is lung cancer, then you 16 have to find out what else do we know about lung 17 cancer. What other risk factors we have for lung 18 cancer and whether those risk factors would be more or 19 less the same, similar, between the exposed group, the 20 group who smoked, and the group who do not smoke. 21 Q. Now, this may be a stupid question, but what's 22 the risk factor? 23 A. The risk factor for lung cancer? 24 Q. No, a risk factor in general. 25 A. Risk factor meaning that if you do have that 77 < < NOGARA REPORTING SERVICE > >. 1 risk that risk will be increased to a certain disease 2 as a result of that. We would say that smoking would 3 be a very strong risk factor for lung cancer. 4 Q. To be a risk factor does there have to be a 5 determination of causality? 6 A. In a way it does because without that you 7 don't know that that is a risk factor. 8 Q. Who determines risk factors, the author or the 9 scientific literature? 10 A. I don't think it's black and white. The 11 authors present--- usually the authors present the 12 data, present the analysis and make some kind of 13 interpretation of the data and may or may not come to a 14 definitive conclusion in terms of causation. But other 15 epidemiologists looking at the same data can either 16 agree or disagree with the author. 17 Q. As to the confounders or as to the causation 18 analysis that he's trying to either prove or disprove? 19 A. All of the above. I mean, you know, other 20 people looking at a specific study may disagree with 21 how the data should have been collected, how the data 22 should have been analyzed or the interpretation was not 23 supported by the analysis. The criticism, the 24 agreement or disagreement can be throughout the entire 25 research process. 78 < < NOGARA REPORTING SERVICE > >. 1 Q. When you're looking at risk factors for a 2 particular disease process you only utilize the risk 3 factors that have been determined to be causally 4 related to the particular substance under 5 investigation? 6 A. I'm not so sure I understand your question. 7 But when you do a study you want to include some of the 8 confounders. You want the control for some of the 9 confounders. You would like to have information on the 10 confounders so that you can adjust for that in the 11 analysis or control for that in the design. 12 One simple confounder that everyone of us 13 would agree would be age on the disease process. For 14 example, if I have a group of workers exposed to 15 benzene and I can calculate the rates of leukemia in 16 that group. I want to compare that to the general 17 population. But, of course, 1 need to adjust for age 18 because we know that leukemia rarely changes as we get 19 older. So age is a confounder to that extent. And we 20 need to adjust for that in the analysis. And that's 21 why we have the procedure called standardized mortality 22 ratio, SMR. In the SMR the expected age is adjusted. 23 That means that we have adjusted for that. 24 Q. Are confounders and risk factors synonymous? 25 A. In most situations I would say. yes because 79 < < NOGARA REPORTING SERVICE > > 1 basically a risk factor is more general than 2 confounder. Confounder is a risk factor whose effect 3 we don't want to study, per se. But we need to be 4 aware of that so we can adjust for that. For example, 5 if I look at a study of, again, between benzene and 6 leukemia, age is a risk factor for leukemia. But in 7 this study, in this study that I conducted, age is also 8 a confounder because I'm not interested in the age 9 effect on leukemia. I'm only interested in adjusting 10 for or controlling the confounding effect into age. 11 Q. Is age a risk factor for multiple myeloma? 12 A. Oh, certainly because depending on what your 13 age is your risk of multiple myeloma would be 14 different. 15 Q. Did you look at any statistics to see what the 16 risk of developing multiple myeloma at the age of 39 17 would be? 18 A. I have not -- I can't give you a number. But 19 certainly that's something very easy to look up. 20 Q. Do you think it would be higher or lower than 21 the age at 60? 22 A. I think it would be lower. 23 Q. Would you agree that developing multiple 24 myeloma at the age of 39 is certainly statistically 25 improbable? 80 < < NOGARA REPORTING SERVICE > > 1 A. No. 2 Q. Would you agree that it would be unlikely to 3 occur absent some environmental agent? 4 A. No. 5 Q. Is it more likely than not to occur at 39 6 without some environmental agent being the causative 7 agent? 8 A. No. Regardless of age most multiple myeloma 9 cases that we report in this country are not related to 10 any specific exposure. 11 Q. Even at the age of 39? 12 A. That's true. 13 Q. And we talked a little bit earlier about 14 statistically multiple myelomas normally occur in the 15 sixth decade of life, correct? 16 A. No. We say that that is the age group we 17 start seeing more multiple myelomas than the other age 18 groups. 19 Q. Did you review Mr. Fannin's background, if you 20 will, one, his social background, to determine if there 21 were any risk factors in his background for the 22 development of multiple myeloma? 23 I don't know what you mean by any -- I guess 24 documents related to his social background. The 25 documents that I have reviewed specific to the 81 < < NOGARA REPORTING SERVICE > >. 1 plaintiff in this case are all listed on my November 2 25th report on the first page. Basically they include 3 answers to interrogatories, as well as deposition 4 transcripts and also medical records. That would be 5 the extent of my information. 6 Q. In your opinion, Doctor, is there any known 7 cause of multiple myeloma? 8 A. I don't know whether ionizing radiation would 9 be one. I don't know whether ionizing radiation is a 10 known -- known, I assume you mean well-accepted by 11 everybody. I don't know whether that's well-accepted 12 or not. 13 Q. Did you find any risk factors in your review 14 of this case for Mr. Fannin's development of multiple 15 myeloma? 16 A. I have not done that. 1 have not been asked 17 by the attorneys to look into other risk factors. 18 Q. Do you know whether there's an association 19 between metals and multiple myeloma? 20 A. Metals? 21 Q Metals generally. 22 A. I'm not aware of any. 23 Q. How about stainless steel and multiple 24 myeloma? 25 A. I'm not so sure 1 understand. 82 < < NOGARA REPORTING SERVICE > > 1 Q. Working with stainless steel. 2 A. Working with stainless steel itself? 3 Q. Yes, sir. 4 A. It's kind of difficult to answer that 5 question. Because when somebody works with stainless 6 steel I assume he or she is also exposed to other 7 things in relation to working with stainless steel. 8 And I don't know what that is. That person may be a 9 welder or that person may be -- I don't know. I don't 10 know what other exposure would be. But I don't know of 11 any direct relationship between stainless steel itself 12 and multiple myeloma. 13 Q. What about working with aluminum? 14 A. 1 have not done anything on -- 1 have not 15 looked into that. I can't answer the question. 16 Q. Propane, working with propane or working 17 around propane? 18 A. I'm not aware of any. 19 Q. What about working around coal or coal dust? 20 A. That I don't know. 21 Q is rheumatic fever related in any way to 22 multiple myeloma? 23 A. I think there may be one or two studies or may 24 be even two or three studies reporting a 25 non-significant increase for that. _But, again, I don't 83 < < NOGARA REPORTING SERVICE > >. 1 think I would be in a position today to make a 2 definitive statement on that, because I have not looked 3 at that issue specifically. 4 Q. Well, you had to rule out other potential risk 5 factors, didn't you, for Mr. Fannin, or did you just 6 assume that since benzene doesn't cause multiple 7 myeloma I don't need to look at anything else that he's 8 been exposed to? 9 A. I'm not sure I understand the question. But 10 my assignment in this case was to look at whether there 11 is any relationship, causal relationship between 12 benzene exposure and multiple myeloma. I have not been 13 asked to look at stainless steel or aluminum and so on. 14 Q. So when it comes time for trial you won't come 15 into the courthouse and tell the jury that Mr. Fannin's 16 exposure to metals or stainless steel or aluminum or 17 propane or coal or coal dust was the cause of his 18 multiple myeloma, will you? 19 A. At this point I don't have non-Hodgkin's 20 leukemia to say that. But on the other hand, if the 21 attorneys decide to ask me to look into that, then I 22 may be able to offer an opinion when I testify. But 23, I'm sure if they asked me to do that we would tell you 24 what the opinion is on that. 25 Q. What articles do you recall that indicate that 84 < < NOGARA REPORTING SERVICE > >. 1 rheumatic fever is a risk factor for multiple myeloma? 2 A. I think it's one of the papers that I have 3 reference to in my first report, the report dated 4 November 25th. I won't be able to find the paper for 5 you now, but one or two papers may have talked about 6 that. 7 Q. Would you be kind enough to advise Mr. Damico 8 sometime about what papers it is? 9 A. Okay. 10 Q. Has it been proven causally that rheumatic 11 fever is associated with multiple myeloma? 12 A. I cannot answer the question. As I said, I 13 have not done a thorough research on that that 1 feel 14 comfortable to answer your question today. And if the 15 lawyers ask me to do that, I will be happy to do it. 16 Q. So as we sit here today under your own 17 definition of a risk factor, that it needs to be 18 causally established as a substance that would cause 19 multiple myeloma, you can't say that rheumatic fever is 20 a risk factor? 21 A. I cannot say that it is or is not today. 22 Q. I understand that. Now, then, are there 23 certain occupations that the literature indicates have 24 a higher risk of developing multiple myeloms based upon 25 your review "of the literature? 85 < < NOGARA REPORTING SERVICE > >. 1 A. Well, again, 1 said that my literature review 2 was focused on the benzene as a potential risk factor. 3 1 did not look at occupations, per se. 4 Q. Well, you looked at diesel exposed workers, 5 - diesel exposed workers and creosote workers. And in 6 those articles they talk about various risk factors for 7 various occupations, don't they? 8 A. Yes, sometimes they talk about occupations. 9 Yes. 10 Q. And isn't it true that a few of the articles 11 that you yourself cited indicate there's an elevated 12 risk for railroad workers for the development of 13 multiple myeloma? 14 A. Well, the problem is when you talk about a 15 railroad worker we don't know exactly what they mean in 16 terms of exposure. Some studies indicate that railroad 17 workers might have an increased risk of multiple 18 myeloma, but other studies indicate that there is no 19 increased risk. So I don't think you would be, at 20 least as far as my assignment goes, talking about 21 benzene, creosote and diesel emission, I don't think 1 22 should rely on the job title, the occupation railroad 23 workers. I should rely on the specific chemicals. 24 Q. Has it been reported in the literature that 25 being a railroad worker is a risk factor for the 86 < < NOGARA REPORTING SERVICE > >. 1 development of multiple myeloma? 2 A. I told you that a couple of studies indicate 3 that there is an increased risk. But at the same time 4 there are also studies indicating there are no 5' increased risk. 6 Q. Can you tell me a study that indicates that 7 there is not an increased risk for being a railroad 8 worker for the development of multiple myeloma? 9 A. If you're patient, I can see if I can find the 10 paper. Okay. If you turn to the references I have. 11 Q. In your 25th report or the 28th report? 12 A. It may appear in both reports. Yes, in fact, 13 it appears in both reports. It's the study by 14 Erickson. Do you see that? 15 Q. Wait a minute. Erickson and Carlson? 16 A. Yes. Do you have a copy of that paper? 17 Q. I do by happenstance. 18 A. If you turn to Page 98. 19 Q. I'm there. Table 2. 20 A. Wonderful. 21 Q. Railroad workers, the relative risk was 1.14. 22 Wait a second. 90 percent. 23 A. Wait. Are we looking at the same paper? 24 Q. Maybe. Erickson. 25 A. Erickson and Carlson, Table 2. 87 < < NOGARA REPORTING SERVICE > > 1 Q. Yes. Erickson and Carlson. 2 A. The relative risk for a railroad worker is 3 0.5. The very first -- the one on top. 4 Q. Table 1 ? 5 A. Table 2. 6 Q. Railroad worker. 0.5. 7 A. Yes. 8 Q. All right. I looked at road workers and 9 misconstrued that. 10 A. Road workers are not the same as railroad 11 workers, I would assume. 12 Q. I would assume they're not either. Do you 13 have -- I can never pronounce his name. Bopheta. Do 14 you have Sopheta '88 in front of you? 15 A. Absolutely. 16 Q. Look at Bopheta '88. 17 A. Okay. 18 Q. Didn't they determine that there was a 19 relative risk or an odds ratio for the development of 20 multiple myeloma? 21 A. What table are you on? 22 Q. I'm on Table 4. 23 A. Table 4. 24 Q. Page 556. 25 A. Yes, I think you gave me the wrong reference. 88 < < NOGARA REPORTING SERVICE > >. 1 That's Bopheta 1989. Did you say '88? 2 Q. '89. I'm sorry. I meant '89. I said '88. 3 Didn't they determine that the odds ratio for railroad 4 workers and the instance of multiple myeloma was 6.0 5 , statistically significant to the .05 level? 6 A. Yes, I'm looking at that. That's why 1 say 7 you cannot rely on the occupation, because railroad 8 workers can mean different things to different people. 9 And certainly you have a big discrepancy between the 10 two studies. One thing you do want to look at is when 11 the two studies give you different results you may want 12 to look at the sample size. If you look at the Bopheta 13 paper, you will see that there are three cases. Do you 14 see the three slash two? That means there's three 15 cases of multiple myeloma in railroad workers and there 16 were two railroad workers in the control. 17 So the important thing is to look at those 18 three cases, the cases of multiple myeloma in railroad 19 workers. Whereas, if you look at the Erickson paper 20 let me find my Erickson paper again. If you look at 21 the Erickson paper, 1 look at Table 2 on Page 98. You 22 see that you have seven cases of multiple myeloma. 23. Railroad workers. 24 So everything being equal, then the Erickson 25 study should be more accurate and more reliable than 89 < < NOGARA REPORTING SERVICE > > 1 the Bopheta study because the sample size is much 2 bigger. 3 Q. Well, just because a study is smaller, Doctor, 4 doesn't necessarily mean that you can't find a 5 relationship, does it? I mean, really, what you're 6 concerned with is the Beta error, aren't you, not being 7 able to find a relationship with a smaller study? 8 A. Well, we are concerned with both. If, again, 9 if you think that they define railroad worker in an 10 accurate way in both studies, and everything being 11 equal in those two studies, then sample size would be 12 something to be considered. 13 Q. Why would sample size be something to be 14 considered when you found a relationship in the smaller 15 study? 16 A. Well, because a misclassification of one 17 single case would change the result completely. After 18 all, it's -- the whole thing depends on three cases. 19 Q. That happens all the time, doesn't it? 20 A. If the definition of railroad worker is not 21 accurate, if somebody answered the question railroad 22 worker incorrectly, if it happened in only one case, 23 the whole thing would go away. 24 Q. Isn't that accounted for in the statistical 25 tests, though? 90 < < NOGARA REPORTING SERVICE > 1 A. Yes. If you take out one case, the 95 percent 2 confidence level will be different. 3 Q. But it's accounted for when you do the 4 statistical analysis, isn't it? 5 A. Assuming all three cases were indeed railroad 6 workers. The statistical test does not account for 7 misclassification. . 8 Q. Do you have Heinman? 9 A. I do. 10 Q. Would you find that for me? It's Table 1. 11 Did you find it, Doctor? 12 A. I'm still looking. What page are you on? 13 Q. It looks like it's Page 558. 14 A. Okay. 15 Q. At the bottom, transportation and material 16 handling, Table 1, the continuation of Table 1 from the 17 previous page. 18 A. Right. 19 Q. Road and railroad workers. The odds ratio is 20 Although not statistically significant because it 21 includes 1, it nevertheless is elevated? 22 A. Yes. But it also includes road workers as 23 well. So it's not just railroad workers, but also road 24 workers. 25 Q. Turn to Table 2. 91 < < NOGARA REPORTING SERVICE > >. 1 A. And if you go back to Erickson, because you 2 make a mistake of looking at the road workers, the road 3 workers have a risk twice as high as railroad workers 4 in the Erickson study. So when you combine the two, 5 maybe the increase comes from the road workers and not 6 the railroad workers. 7 Q. Would you turn to Table 2 for me?, 8 A. Yes. 9 Q. At the bottom, railroad, that would be 10 transportation. Underneath that, railroads? 11 A. Yes. 12 Q. The OR is 1.4, although not statistically 13 significant, because the lower end of the range is 14 below one and still is an elevated risk for railroad 15 workers to develop multiple myeloma, correct? 16 A. Correct. 17 Q. Going back for a second to -- let me find my 18 note. Bopheta, '89. You indicated that if any one of 19 these three cases were mischaracterized or their 20 employment was mischaracterized in any way, that that 21 would.6hange the odds ratio. 22 A. Yes. 23 Q. Do you have any evidence that any of those 24 three cases were mischaracterized by anyone? 25 A. I'm not saying there is evidence. All I'm 92 < < NOGARA REPORTING SERVICE > 1 saying is that being a very small study, and the result 2 depends on only three cases, that we have to interpret 3 that very carefully. That's all I'm saying. 4 Q. I just want to make sure you didn't have 5 something that I was unaware of. Let's skip for a 6 second to engine exhaust, Dr. Wong. In your opinion, 7 is engine exhaust a risk factor for multiple myeloma7 8 A. Well, again, that's not very specific. 1 9 tried to include that because sometimes people include 10 engine exhaust as both gasoline exhaust as well as 11 diesel exhaust. And let me look at my report, what I 12 said. Well, there are two studies from Sweden. The 13 first study was done by, I think it's pronounced 14 Flodin. 15 Q. I think you're correct. 16 A. 1987. And Flodin found a marginally 17 significant increase relative risk of 2.1 for engine 18 exhaust. And the 95 percent confidence interval was 19 from 1.2 to 3.9. So that was very significant. But 20 the authors themselves point out that there may be some 21 confounding in their study. And certainly they would 22 like to see some additional studies, preferably with 23 some adjustment or control over potential confounding. 24 And the second study was done a few years 25 later published by Erickson and Carlson in 1992 from 93 < < NOGARA REPORTING SERVICE > >. 1 Sweden, as well, the same country. And that study was 2 twice as large as the first study. And they did not 3 find a significant increase for engine exhaust. 4 Q. But they nevertheless found an elevated 5 relative risk of 1.38. It might not have been 6 significantly significant, but it was elevated. 7 A. Well, the 1.38 was without any adjustment for 8 potential confounding. They presented two analyses, 9 one without adjusting for confounding. The other one, 10 adjusting for confounding, the relative risk was 11 reduced to 1.1. 12 Q. Going back to the Flodin study, you indicated 13 in your report on Page 3, the third line down the 14 sentence that begins, The authors further remarked. 15 A. Right. 16 Q. That no study on multiple myeloma as related 17 to engine exhaust has, to our knowledge, been 18 published. 19 Then you indicated up above that there was the 20 confounding factors that may have been involved that 21 the authors pointed out. 22 A. Right. 23 Q. Do you accept Flodin's confounding factors as 24 potential risk factors for the development of multiple 25 myeloma? 94 < < NOGARA REPORTING SERVICE > > 1 A. Based on the Erickson study, certainly that 2 seems to be the case. Because without adjustment for 3 confounding the risk ratio was 1.38. After adjustment 4 the risk ratio was reduced to 1.1. So there is some 5. indication of that. 6 Q. Let me find this article for a second, Doctor. 7 Doctor, look at Flodin for a second. Are you with me? 8 A. I have to find the paper. 9 Q. Okay. I thought maybe I lost you. 10 A. Okay. I have the paper. 11 Q. What are the confounders that the author in 12 the Flodin study were concerned about? 13 A. I'm trying to find the paragraph. Well, in 14 their study they find an increase related to engine 15 exhaust, creosote and fresh wood. And some of the 16 occupations would have exposure to all three. I think 17 that may be one of the confusing parts. 18 Q. In your opinion, then, would creosote, engine 19 exhaust and fresh wood be rich factors for the 20 development of multiple myeloma? 21 A. Well, in my report I talk about two of the 22 three things you mentioned. I think, looking at all 23 the studies we have today on diesel exhaust and 24 multiple myeloma, the literature is not completely 25 consistent. But certainly I would think that the 95 < < NOGARA REPORTING SERVICE > >. 1 weight of evidence would say that there is no 2 relationship between diesel exhaust and multiple 3 myeloma. And that's for diesel exhaust. 4 In terms of creosote, there are not a whole 5 lot of studies on the subject. There are only two or 6 three studies. And the findings from these two or 7 three studies are not consistent. So I would say at 8 this point we would not be able to say one way or 9 another in terms of creosote. 10 Q. If they're not risk factors then, Doctor, why 11 would you be concerned with the confounding effects of 12 those in any study? 13 A. I did not say that creosote is not. I just 14 don't know if it is or is not at this point. 15 Q. What about diesel exhaust, are you saying that 16 it is not a risk factor for multiple myeloma? 17 A. Based on other studies, 1 don't think it is. 18 Q. Now, when you made your analysis of whether 19 diesel exhaust was or was not a risk factor for 20 multiple myeloma, did you go through the Bradford Hill 21 exercise? 22 A. Yes. One thing is we have quite a few studies 23 saying there is no connection and studies that say yes. 24 Q. Well, the studies that say there are no 25 connections are simply not statistically significant, 96 < < NOGARA REPORTING SERVICE > >. 1 isn't that true? 2 A. No, some actually report lower than unity, you 3 know, the risk ratio. 4 Q. So is it based on the lack of consistency 5 among the studies that you are unable to conclude a 6 relationship between diesel exhaust and multiple 7 myeloma? 8 A. More than that. We are also looking at the 9 quality of the study and how big the studies are. The 10 studies that are reporting an increase are based on 11 only a few cases. 12 For example, if you look at one of the larger 13 studies on people exposed to diesel emissions, which 14 would be the study done by NIOSH, okay, the first 15 author of that study is by the name Stern. S-t-e-r-n. 16 And I have put the citation in the report. They looked 17 at close to 5,000 cancer deaths among construction 18 operating engineers exposed to diesel. And their 19 result was they observed 73 cases of multiple myeloma 20 and expect 81 cases. And the risk ratio was 0.90. 21 That was a very large study. So I have to put more 22 weight on that study. 23 4 Q. But didn't the authors themselves say, Doctor, 24 that it's possible that malignant neoplasms were 25 underestimated because workers that left before 97 < < NOGARA REPORTING SERVICE > >. 1 retirement or death were not included in the analysis? 2 A. That may have an effect for either direction 3 of multiple myeloma, depending on if indeed there was 4 some cancer deaths missing, depending on what it is. 5 Q. Didn't the multiple myeloma have a longer 6 latency period than the normal neoplastic disease? 7 A. I'm not so sure that's true for all other 8 cancers. 9 Q. Would you agree with me that a longer latency 10 period regardless of whether it's longer than the other 11 neoplastics or not, if someone left the union and their 12 vital status was not accounted for in the cohort, that 13 you could be missing, as the authors suggest, neoplasms 14 that have a long latency period? 15 A. Yes. But at the same time if they also 16 missed, for example, prostate cancer, which also has a 17 long latent survival, then you would actually make the 18 statistic go the other direction. So, 1 mean, we can 19 criticize the study for missing -- potentially missing 20 some deaths. But we don't know the direction of bias, 21 if indeed, that's the case. 22 Q. Well, the authors criticized it themselves 23 saying that the malignant neoplasm -- it's possible 24 that malignant neoplasms were underestimated in this 25 study since workers who left the union before 98 < < NOGARA REPORTING SERVICE > > 1 retirement or death were not included in the analysis. 2 So they're basically saying themselves when they wrote 3 this that there's a possibility of underestimation of 4 malignant neoplasms, correct? 5 A. Well, right now the risk ratio is 0.9. Even 6 if they underestimate by some extent it's going to be a 7 long way before it would be elevated. 8 Q. Did you notice in the written materials that 9 Mr. Fannin cleaned his paintbrushes with gasoline and 10 he cleaned his hands with gasoline? 11 A. I read the deposition, yes. But what is your 12 question? 13 Q. That was my question. Did you notice that in 14 the deposition? That was my first question. 15 A. Yes. 16 Q. And did you review the literature on gasoline 17 and multiple myeloma? 18 A. I did. I thought that was the focus of my 19 November 25th report. 20 Q. What gasoline studies did you cite? 21 A. The gasoline study that I have done. 22 Q. Which is petroleum workers, terminal workers 23 exposed to gasoline, the statistical reports submitted 24 to API? 25 A. One paper was published. One was reported to 99 < < NOGARA REPORTING SERVICE > 1 API. Yes. There are two reports. 2 Q. Did you rely on the technical report that you 3 did for API 4 A. Both of them, yes. 5 Q. in forming your opinions? 6 A. Yes. 7 Q. I'd like to have a copy of the technical 8 report that you submitted to API on gasoline, since 9 you're relying on that to state that gasoline does not 10 cause multiple myeloma. 11 A. I have to ask the lawyer to get it from API. 12 I'm not sure I can Xerox a copy of that and send it to 13 you. 14 MR. DAMICO: We'll check into it and see what we 15 can do. 16 MR. HARTLEY: Under the rules you know if he 17 relied on it I'm entitled to see it. 18 MR. DAMICO: I understand. He's telling me now 19 that he may not be able to give it to me without 20 approval. 21 THE WITNESS: I cannot make a copy. I think that 22 the report is copyright. So I don't want to just make 23 a copy. 24 MR. HARTLEY: a. Do you normally give out your 25 articles that are in press to other people to review, 100 < < NOGARA REPORTING SERVICE > > 1 Doctor? 2 A. Yes, I do. 3 Q. And did you give out your reports in press to 4 Dr. Bergosagel? 5 A. Yes. Some things are accepted. I don't mind 6 giving it out. Before acceptance by the Journal, then 7 I would not. 8 Q. When was your study on multiple myeloma and 9 benzene exposure accepted by the Journal of Regulatory 10 Toxicology and Pharmacology? 11 A. I think it was quite a while ago. 12 Q. How long is quite a while ago? 13 A. It must be accepted around June or July. 14 Q. Did you give a copy of that to Dr. Bergosagel? 15 A. I might have given a copy to him when I went 16 up to Canada to see him. 17 Q. You met with him? 18 A. Yes. 19 Q. What did you and he discuss? 20 A. Basically we talked about the potential of 21 writing paper, a position paper together, because he 22 is an oncologist and I'm an epidemiologist. We were 23 thinking about how to combine the two disciplines and 24 write a review article. And I told him about my paper 25 on the multiple myeloma paper based on epidemiologic 101 < < NOGARA REPORTING SERVICE > > 1 studies. And I told him to come up with some 2 oncological aspects of multiple myeloma so that we can 3 write a paper together. 4 Q. You're not telling me that out of the clear 5 blue sky you went to Ontario to meet with Dr. 6 Bergosagel to discuss whether the two of you were going 7 to write an article together? 8 A. No. My mother lives in Toronto and I was up 9 there to see my mother. And it just happens that he 10 10lives in Toronto. In fact, I made the plan of going up 11 to see my mother a long time ago. 12 Q. How did you know of Dr. Bergosagel? 13 A. I have seen some of his writings on multiple 14 myeloma. 15 Q. Did you see him on the street corner and say, 16 "Hey, can we talk about writing an article together?" 17 A. No. No. No. I have seen his writings. He 18 has published on multiple myeloma. 19 Q. Did you discuss this case with Dr. Bergosagel? 20 A. No, I did not at that time. In fact, when I 21 went up to see him I don't think I was hired at all in 22 this case. That was before I got involved in this 23 case. 24 Q. When were you hired in this case? 25 A. That goes back to the first time was sometime 102 < < NOGARA REPORTING SERVICE > 1 in September, I think. 2 Q. Doctor, do you have your CV in front of you? 3 A. Yes. 4 Q. I want to look at a few things with you on 5 that. 6 A. Okay. 7 Q. On your CV you list 121 articles. 8 A. Yes. 9 Q. At least the one I have. Those articles, and 10 I'm concerned with the more recent ones, which ones 11 were rejected by the various journals? 12 A. Which ones were rejected? 13 Q. Yes. Which ones did you have to submit to 14 more than one journal to get published? Did you have 15 to submit your study on Risk of Acute Myeloid Leukemia 16 and Multiple Myeloma in Workers Exposed to Benzene to 17 more than the Occupational and Environmental Journal? 18 A. Well, that paper was stimulated by a paper 19 written by Paxton. P-a-x-t-o-n. Do you know who I'm 20 talking about? 21 Q. She works for API, doesn't she? 22 A. She used to work for API. She doesn't work 23 for API anymore. 24 Q. She works for petroleum companies now. 25 A. I don't know about that. I don't know about 103 < < NOGARA REPORTING SERVICE > >. 1 that. You can make any comments you want. She wrote a 2 paper on the updated analysis of the Pliofilm study and 3 published that in a journal called Risk Analysis. 4 That's really a journal for risk assessment. 5 I wrote a paper -- basically I pointed out 6 that we should separate AML from other leukemias for 7 analysis. And I believe they sent the paper out to 8 different people to review. And 1 remember 1 received 9 comments from three reviewers, one reviewer saying that 10 the data is too limited to talk about a dose response 11 analysis. And one paper said they had done analysis 12 similar to that. I think that person, based on what he 13 said, I think that was Kenny Crumb. Because he said he 14 had already done some similar analysis and reached 15 similar conclusions using a different approach. And 16 one analysis said, yes, that's the right way to look at 17 it. 18 But the editor came back and said that my 19 reanalysis separating the leukemia into different 20 groups would not offer sufficient new information and 21 suggested that I may want to reduce it to a letter or 22 something as comments. Which I don't want to do. So I 23 resubmit to the Occupational and Environmental 24 Medicine. 25 Q. So it was rejected by at least one journal? 104 < < NOGARA REPORTING SERVICE > 1 A. Well, I didn't go by what they suggest. 2 Q. That's not my question. Can you answer my 3 question? Was it rejected by a journal when you 4 submitted it for publication? 5 A. It was not accepted at the current form. They 6 wanted me to reduce that and I don't want to. 7 Q. How many journals rejected it or refused to 8 accept it in its current form? 9 A. I didn't send it to any other journal. The 10 second journal I sent it to was Occupational and 11 Environmental Medicine and it was accepted without any 12 changes. 13 Q. Were any of your other more recent articles 14 not accepted as written by other journals? Was the 15 reference No. 112, Cell-type Specific Leukemia Analyses 16 in a Cohort of 208,000 Petroleum Workers not accepted 17 in its original form by any journal? 18 A. No, we submitted it to the Regulatory 19 Toxicology. We didn't send it to anyplace else. 20 Q. It was published in Epidemiology? 21 A. What? 22 Q. It was published in Epidemiology, No. 112? 23 A. Oh, 112. Yes, 112 is an abstract of a 24 presentation we make. It's not a full paper. It is an 25 abstract. It is the same paper that we published 105. 105 < < NOGARA REPORTING SERVICE > 1 Do you see 105? 2 Q. Yes. 3 A. That is the full paper. That is the full 4 paper. 5 Q. Was that accepted originally by Regulatory 6 Toxicology and Pharmacology? 7 A. Yes 8 Q. Did you submit that to any other journal? 9 A. No. 10 Q. What about An Updated Cohort Mortality Study 11 of Workers at Northeastern United States Petroleum 12 Refinery, was that originally rejected or not accepted 13 by any journal? 14 A. I believe that was the original place that we 15 submit to. 16 Q. What about The Application of Meta-analysis in 17 Reviewing Occupational Epidemiologic Studies? 18 A. Oh, that one is a special article. That was 19 what they call metalogical papers. That was actually 20 at the invitation of the editor. So that being the 21 case,. i don't think they would invite us to write an 22 article and then reject it. 23 Q. What about 104, Risk of Acute Myeloid Leukemia 24 and Multiple Myelome in Workers Exposed to Benzene, 25 Occupational and Environmental Medicine. Was that 106 < < NOGARA REPORTING SERVICE > >r 1 originally accepted or rejected by any reviewer? 2 A. 1 think that was the one that we talked about 3 before, that I told you that paper was stimulated by 4 the paper written by Paxton. Her paper was published 5 in Risk Analysis and I submit to that and I got three 6 reviewers' comments. 7 One was, you know, critical in terms of the 8 sample size, which I cannot do anything about because 9 that was based on the NIOSH cohort. And one paper says 10 that, well, that analysis had been looked at 11 differently and we come to the same conclusion. 12 Therefore, that reviewer doesn't think the information 13 is that new. And the third one thinks it was quite 14 favorable. And the editor said, "if you shorten that 15 maybe we'll consider it." But I didn't want to do 16 that. 17 Q. In your review of the materials in this case, 18 did you determine whether Mr. Fannin was exposed to 19 petroleum products? Do you understand that term? 20 A. I understand that term. But that is not my 21 area of expertise. I am not in a position to determine 22 what he was or was not exposed to. Basically, I was 23 asked to look at epidemiologic data between certain 24 exposure and multiple myeloma. And that is my focus. 25 Q. Well, then, did you look at Linet's study of 107 < < NOGARA REPORTING SERVICE > > 1 the hundred patients in the Baltimore hospital which 2 found a relative risk of 3.3 for those exposed to 3 petroleum products? 4 A. Yes. Petroleum product can mean anything. 1 5. was not asked to look at petroleum products in general. 6 1 was asked to look at benzene in specific. 7 Q. Did you look at the Linet study? 8 A. Yes, I did. 9 Q. Did you include it in your report? 10 A. Yes, I did. 11 Q. Where? 12 A. Look at the reference. 13 Q. Which one? 14 MR. DAMICO: It's in the November 25th report, the 15 first page of the bibliography, the third from the last 16 article. 17 MR. HARTLEY: Q. Did you comment on Linet in your 18 paper that I just received today? That's why I'm. 19 asking, because I haven't had a chance to look at it 20 completely. 21 A. Hold on for one second. 22 MR. DAMICO: I know you're in a hurry. I'll shut 23 up if you want me to. But Page 3 he mentions it, the 24 second full paragraph. I don't know if it's mentioned 25 anywhere else. 108 < < NOGARA REPORTING SERVICE > 1 THE WITNESS: Yes. 1 mentioned Linet in the 1987 2 paper as one of the papers that suggests a number of 3 risk factors for multiple myeloma. 4 MR. HARTLEY: Q. Didn't Linet report a 3.3 5 . relative risk for petroleum products? And you did not 6 mention petroleum products in your potential risk 7 factors for multiple myeloma. My question is, why not? 8 A. Why not? Because those are lifestyles. The 9 things that I put down there are all lifestyle risk 10 factors. If you look at Page 3. None of that is 11 related to occupational exposure. 12 Q. You didn't put down any occupational risk 13 factors, did you, in any report? 14 A. Yes, on Page 6. 15 Q. Which one? 16 A. Under the subheading of the first report. 17 November 26th on Page 6 under the subheading on case 18 control studies of multiple myeloma, the very first 19 study is Linet 1987. She reports a risk ratio of 1.1 20 for benzene exposure. 21 Q. But she also reported a 3.3 relative risk for 22 petroleum products and you didn't mention that anywhere 23 in here? 24 A. Sir, 1 was asked to look at benzene exposure, 25 not petroleum products in general. If you want to make 109 < < NOGARA REPORTING SERVICE 1 a comment, make a comment. 2 Q. Which comment would that be? I didn't know I 3 made a comment. 4 A. No. If you want to make a comment on 5. petroleum products, you go ahead. That's not my focus. 6 Q. 1 understand. You didn't focus on paints and 7 solvents either, did you? 8 A. I have not been asked to do that. 9 Q. Did you do any calculations of Fannin's 10 cumulative exposure to anything that he was exposed to 11 that you have reviewed or that you had gleaned from 12 your review of the materials? 13 A. I have not. 14 Q. Are you going to testify at trial, Dr. Wong, 15 about solvents or paints and whether there's a 16 relationship between those two substances and multiple 17 myeloma? 18 A. No, I have not done any work on that for this 19 case. And I'm not in a position to testify on that. 20 Q. What about polycyclic aromatic hydrocarbons, 21 are your in a position to testify regarding whether they 22 are associated with multiple myeloma? 23 A. Well, let me specify or emphasize again. The 24 three things I looked at in this case are benzene, 25 creosote and diesel exhaust. Those are the three 110 < < NOGARA REPORTING SERVICE 1 things I will talk about, unless the lawyers ask me to 2 look into something else. If that's the case, I'm sure 3 they will let you know. 4 Q. Did you forward any written correspondence to 5 Dr. Bergosagel concerning this case? 6 A. No, I have never talked to him about this case 7 at all. 8 Q. Doctor, do you have an understanding of 9 whether there is benzene in solvents? 10 A. It depends on what kind of solvents you're 11 talking about. 12 Q. How about mineral spirits? 13 A. If there is, it may be a very small percentage 14 as contaminants. You have to look at the specific 15 product. 16 Q. Your study on -- back on your CV, No. 58, Wong 17 and Ragland. 18 A. Yes. 19 Q. I'm looking at publication No. 88. Was it 20 ever published? 21 A. That is so old I lost track of that. I guess 22 we never changed the word processor. We didn't change 23 that. 24 Q. Cooper and Wong, No. 52, A Study of Mortality 25 in a Population of Nickel Miners and Refinery Workers 111 < < NOGARA REPORTING SERVICE > >. 1 submitted in '88. Was that ever published? 2 A. To be honest with you, I don't know what Dr. 3 Cooper did with that paper. He's retired now. So I 4 don't know exactly what he did with that paper. 5. Q. No. 62, Epidemiology, Toxicology and 6 Quantitative Risk Assessment, An Integrated Approach, 7 submitted for publication. Was it ever published? 8 A. We submitted it and then things changed so 9 rapidly in terms of risk assessment that I decided to 10 kind of take it back and maybe revise that. 11 Q. Did you ever revise it? 12 A. No, because risk assessment is really within 13 the last five, ten years, become very, very broad. 14 Q. Did the journal reject it as it was written? 15 A. No, basically I just retracted it. There's no 16 point in publishing a paper that's not complete. 17 Q. 63, Wong, Whorton, Morgan and Gordon, Cause 18 Specific Mortality and Morbidity Among Paper Mill 19 Employees, submitted for publication. Was it ever 20 published? 21 A That may be a preliminary analysis of a paper 22 that we subsequently published later. Let me see. We 23 did publish a paper on paper mills. I think -- let me 24 try to -- do you have Publication 116? 25 Q. Yes 112 < < NOGARA REPORTING SERVICE > > 1 A. That may be the same paper, except with more 2 data, with more paper mills added to the project. 3 Q. Do you know whether it is or not? 4 A. I'm not definitely sure because the first one 5 was sometime ago. 6 Q. What about No. 88, Wong and Foliart, 7 Assessment of Lung Cancer Risk by Histologic Category 8 in Workers Exposed to Chlorinated Chemicals, submitted 9 for publication. Was that ever published? 10 A. That was too short. And I decided - we 11 decided to just leave it as a technical paper. I think 12 we submitted a technical paper to the Chemical 13 Manufacturers Association. 14 Q. 119, that was published, right, A Nested Case 15 Control Study of Leukemia. 16 A. 119? Okay. I have a more updated CV, so the 17 numbers may be slightly different. What was the title 18 of the paper? 19 Q. A Nested Case Control Study of Leukemia and 20 Multiple Myeloma in a Cohort of Land-based Terminal 21 Workers Exposed to Gasoline in the Petroleum Industry. 22 A. No, actually we were going to submit it, but 23 we needed to do a little more work. 24 Q. It said it was submitted for publication. Did 25 you withdraw it? 113 < < NOGARA REPORTING SERVICE > > 1 A. No. No. No. We thought we were going to 2 submit it, but we have to do more work. So we have not 3 submitted it at this point. 4 Q. Doctor, who pays for your time to write all 5 these papers? 6 A. Writing the papers themselves? 7 Q. Yes, sir. 8 A. No one. 9 Q. You just go out and write them yourself? 10 They're not funded by any agency or any association? 11 A. Sometimes if the paper is based on a project, 12 of course, the project most likely would be funded by 13 some group. And we would, you know, as a result of the 14 project we would have a technical report submitted to 15 the sponsor. And a lot of times we would, for lack of 16 a better term, cut and paste the longer technical 17 -report into a sort of manuscript and submit that for 18 publication. But the sponsors would not pay for 19 writing of the manuscript itself. 20 Q. Are you ever funded by a university or a 21 college? Have you ever been funded by a university or 22 college? 23 A. I don't think universities or colleges are in 24 the business of funding studies. 25 Q. Have you ever been, absent your time at 114 < < NOGARA REPORTING SERVICE > 1 Georgetown, after you left Georgetown, have you ever 2 been asked to be on the faculty of any university and 3 obtain grants to provide for your remuneration as a 4 professor or a visiting professor? 5 A. Well, I'm on the adjunct faculty of Tulane 6 University and I'm also a visiting professor at one of 7 the medical schools in Taiwan. But not being there 8 full-time it would be very difficult for me to apply 9 for research -- apply for research funding as a 10 university executive member. 11 Q. Does the adjunct professorship bestow any form 12 of money to you? Do you get paid for doing that? 13 A. They pay for my traveling expenses and living 14 expenses and so on. But basically 1 contribute my time 15 to the university. 16 Q. Where does your general income come from then 17 as Applied Health Sciences? 18 A. Well, the income, the majority would be coming 19 from doing research, doing projects. 20 Q. Well, for whom? 21 A For mainly the industry. 22 Q. Can you give me some examples? 23 A. Well, certainly. 1 have done work for API. I 24 have done work for paper product companies. I'm doing 25 a study for the National Stone Association. I'm doing 115 < < NOGARA REPORTING SERVICE > > 1 work for the International Lead and Zinc Research 2 Organization. 1 have done work for Mobil Oil Company. 3 I have done work for Chevron. 4 Q. Would you agree with me, Doctor, that if the 5 industry is not happy with the results of your reports 6 that they wouldn't ask you to come back again and do 7 another report for them? 8 A. You have to ask the industry. 9 Q. How much time have you spent in this case? 10 A. As I said, I was contacted the first time in 11 September -- 1 think in September they sent me a 12 package of materials to review. And 1 have identified 13 the material in my November 25th letter. You see that 14 on the first page. Basically I spent about a couple of 15 days looking through, reviewing the package. And then 16 in November they asked that I should write a report for 17 this case. And I think I spent maybe two to three days 18 doing some review and also writing the report. 19 Q. So how many hours do you think you have in 20 this, 40? 21 A. Well, roughly two days in October. Probably 22 about two to three days in November. 23 Q. About eight-hour days? 24 A. Yes, I'm talking about eight-hour days. 25 Q. Have you billed them for your time yet? 116 < < NOGARA REPORTING SERVICE > > 1 A. Yes, I have. 2 Q. Have you been paid? 3 A. I don't want to embarrass the attorney sitting 4 across from me at the table. They have not paid me. 5 MR. DAMICO: I was unaware of this glaring error 6 until early this morning or late this morning, 1 should 7 say. 8 MR. HARTLEY: Doctor, I don't believe I have 9 anything further. 10 MR. DAMICO: We'll read and sign. 11 MR.-HARTLEY: I'd like to get a copy of his 12 technical report to API. 13 MR. DAMICO: We'll get it. 14 MR. HARTLEY: Thank you. 15 (Whereupon, the deposition concluded at 5:45 16 o'clock p.m.) 17 18 OTTO WONG, Sc.D. F.A.C.E. 19 20 21 ' r, 22 23 24 25 117 < < NOGARA REPORTING SERVICE > 1 STATE OF CALIFORNIA ) ) 2 COUNTY OF SAN FRANCISCO) 3 I, DONNA LEE BECKETT, a Certified Shorthand 4 Reporter of the State of California, duly authorized to 5 administer oaths pursuant to Section 2025 of the 6 California Code of Civil Procedure, do hereby certify 7 that 8 OTTO WONG, Sc.D. F.A.C.E., 9 the witness in the foregoing deposition, was by me duly 10 sworn to testify the truth, the whole truth and nothing 11 but the truth in the within-entitled cause; that said 12 testimony of said witness was reported by me, a 13 disinterested person, and was thereafter transcribed _ 14 under my direction into typewriting and is a true and t - 15 correct transcription of said proceedings. 16 I further certify that I am not of counsel or 17 attorney for either or any of the parties in the 18 foregoing deposition and caption named, nor in any way 19 interested in the outcome of the cause named in said 20 caption. 21 Dated the 30th day of December, 1997. 22 234 E.I:.V 24 DONNA LEE BECKETT CSR No. 3450 (California) 25 L 118 c < NOGARA REPORTING SERVICE > >, 1 Otto Wong, Sc.D. F.A.C.E. 2 181 Second Avenue, Ste. 628 San Mateo, CA 94401 3 - Date: Tuesday, December 30, 1997 4 Re: Fannin vs. Norfolk & Western Railway Deposition Date: Tuesday, December 16, 1997 5 Dear Dr. Wong, 6 Please be advised the original transcript of your 7 deposition is ready for your review. 8 You have 35 days from the date of this letter to read, correct if necessary, and sign your transcript. It 9 will then be sealed and sent to the examining attorney pursuant to the applicable law. You are not required ' 10 by law to read and sign your deposition. 11 You may either come to our office to read and sign the . original transcript, or you may contact your attorney 12 or the attorney who arranged for you to be present at your deposition. If they have ordered a copy of the 13 transcript, you may review their copy and make corrections by submitting, signing and returning the 14 attached form. If you choose to review your transcript at our office, please call first to make an 15 appointment. Should you have any question regarding these instructions, please call. 16 Sincerely, 17 18 NOGARA REPORTING SERVICE 130 Battery Street, Suite 580 19 San Francisco, California 94111 (415) 398-1889 20 21 cc: Aft counsel Original deposition 22 23. 24 25 . 119 < < NOGARA REPORTING SERVICE >