121 



He and other 



[he 



arch 



headqua 

 rk. N.C., agri 



>al animal research, by ilself, should no 

 .nger be accepted as a reliable means of 

 jjging risks lor humans 



The problem n we donl know what ihe 

 ndmgs really mean." Dr Roben Maronpoc. 



hid of the institute's cxpcrtinciual-pathol- = 

 £v laboratory, said of the aminal studies 



\s illustration. Dr Allen J « ikov chief of 

 ie institute's epidemiology branch, cued a 

 eccnl institute study showing that rodents 

 mistiming coli beverages ■showed an asso- 

 lanon between the cola beverages and renal 

 ailurc." or loss of kidney function 



'But the results are murky/' he wenl on 

 \nd so the institute is choosing not to draw 

 onclusions until more research is done. 



Another study, completed about a year 

 igo. found that rats and mice develop cancer 

 vhen fed high doses of oxazepam, a direct 

 hemical relative of Valium Valium is 

 imong the nation's mosl-often prescribed 

 Irugs, and the rodents taking the maximum 

 olerated dose ol oxazepam "had a 100 per- 

 :cni incidence ol lumors. all over the body — 

 ery quick.'' Dr Maronpot said 



And so the institute began a rare, crash 

 study, devoting all available resources to see 

 .vhat more could be learned 



Examining frozen DNA sections from Ihe 

 affected animals, "we found this was an 

 VI.T.D. result," Dr Maronpol said "Oxaze- 

 pam would not be a problem even for a 

 mouse at normal human dosage levels." 



But this kind of research is costly and 

 time-consuming, and the technology has ex- 

 isted for only a few years. 



Dr. Maronpot swept his hand toward a long 

 row of blue books stretching more than 10 

 feel along an upper shelf, reports on all 450 

 animal studies the Government has conduct- 

 ed over the lasl 30 years. 



"Il'S an impressive product, not produced 

 by anyone else in the world." he said Still. 

 Dr. Maronpot acknowledged, neither he nor 

 anyone else at the institute knows how many 

 of the tested substances that produced tu- 

 mors or other harmful effects in animals — 

 about half the total — might now be shown to 

 be benign at normal levels. 



A More Vexing Question 

 Even more worrisome, perhaps, is the 

 opposite question: How many substances 

 that caused no harm to rodents might be 

 dangerous lo humans' One chance finding 

 demonstrated this problem. 



"Arsenic is not a carcinogen in animal 

 studies," said Dr Joseph F Fraumeni, direc- 

 tor ol epidemiology and biostatistics at the 

 National Cancer Institute. But several years 

 ago, he recalled, a study of smelter workers 

 exposed lo high levels of arsenic in the air 

 showed a high level of lung cancer. 



From that. Dr Olden's review committee 

 concluded that the Government should no 

 longer rely only on animal studies They 

 should be simply one part of a program of 

 research also involving studies of population 

 groups found to have been exposed to the 

 substances without knowing of the possible 

 risk, and laboratory analyses showing how 

 the chemicals interact with cells. 



That is easy to say, institute officials 

 agree, but difficult and costly lo do 



Progress 



Studies of Cells 

 At Cutting Edge 



nd mor 

 'M.i 



,.,,0, 



i has been dii 

 utlc shop, the 



CI CO I 



lhal 



pathology laboratory It 

 crash oxazepam study was earned out And 

 it is in laboratories like these lhal the future 

 of toxicology research is believed to lie. 



Mere, scientists try lo figure oul how sub- 

 stances interact with human cells, whether 

 they cause mutations lhal can lead lo cancer 

 And when research like this works, scientists 

 believe they have the most credible findings 

 now achievable 



"Thai's what we're focusing on now. un- 

 derstanding the responses at the most basic 

 level " Dr. Maronpot said "We're making 

 progress " But he is at the culling edge of 

 science, and he acknowledged lhat "there's 

 still so much we do not know," 



As H is, if a substance lhat produces a 

 carcinogenic effect in mice is referred lo Dr 

 Maronpoi's lab. "typically it can take two 

 three, maybe even five, seven, eight years" 

 lo carry out one of these studies. So Dr, 

 Maronpoi's laboratory can offer a second 

 opinion on fewer than 10 percent of Ihe sub- 

 stances subjected lo animal studies, i 



Nearby, Dr. Wilcox heads the department 

 thai offers the second-best hope for validal- | 

 ing findings from the animal studies He and i 

 seven' other epidemiologists Iry to find spe- 

 cific groups of people who have been exposed 

 to the substances lo see if they have suffered 

 ill effects. The smelter workers exposed lo 

 arsenic are an example. Bui unfortunately, 

 Dr Wilcox said, similar examples of epide- 

 miological studies are rare. 



"The whole area of environmental epide- 

 miology is a frustrating one." he said. The"; 

 principal problems are thai people are gener- 

 ally exposed to low levels of the suspect 

 substances And even if they suffer unusual 

 health problems, it is hard lo know whether 

 ,he illnesses were caused by the substance or 

 something else - smoking, poor diet, etc 



"Epidemiology is a real crude tool for 

 looking lor associations." Dr, Wilcox ac- 

 knowledged It is also time-consuming as a 

 msuli his department, like the pathology 

 laboratory. Is able to examine only a tiny. 

 percentage' of the substances subiected to 

 animal studies. ,„.„- 



Thai means Ihe institute and the rest of the . 

 Government can seldom offer much more 

 than the animal studies as warnings of a 

 substance's possible danger to humans. 



""We're looking for alternative approach- 

 es," Dr Gnesemer said. "But right now, 

 that's what we've got." 



Quite often, that means no one lakes the 

 institute's warnings 6eriously any longer 



Frustrations Grow 

 With Knowledge 



nost two years ago. the results came in 

 rat and mouse studies of 1.2.3-u ichloro- 

 .inc. an industrial solvent used as a paint 

 •Mrnilli remover or a degrcasmg agent 

 most every animal exposed to the sub- 

 re was riddled with tumors "in several 

 ns " said Dr Richard D liwin. the ins 



cpor 



Th- 



is the type ol chemical lhal shows the great- 

 est potential for human effect." 



"Our understanding is that workers wash 

 themselves in this." Dr. Gnesemer said. And 

 since ihe chemical is absorbed in the skin, he 

 and others said, ihe finding was particularly 

 troubling 



In Dr Irwin's view, "It would be real good 

 to get some human data because I'm sure 

 ihcrc were people who were exposed lo it in 

 ihe pasl, maybe even now " 



So did the epidemiologists look for people 

 who had been exposed to the substance? 



"This isn't one were looking at." Dr. Wil- 

 cox said. But maybe, he added, the National 

 Cancer Institute's epidemiologists did look at 

 it. The cancer institute has what is probably 

 the world's largest cancer epidemiology de- 

 partment — 100 scientists and support staff 

 — and they gel the animal-study reports 

 automatically. But they seldom choose to 

 begin a study based on Ihe animal research, 

 and they did not initiate one in this case. 



In 1990, when a rodeni study suggested lhat 

 fluoride might be a carcinogen, "we look that 

 one on." said Dr. Fraumeni. head of epidemi- 

 ology for the cancer institute. "We found 

 nothing, and that was the lasl time" 



As (or tnchloropropane. he said. "1 haven't 

 heard of it." 



Dr. Irwin wondered if the Occupational 

 Safety and .Health Administration might 

 have done a survey or found a way lo check 

 on workers exposed lo Ihe chemical. 



But Dr. Edward Stein, a health scientist for 

 O.S.H.A., said ihe agency had done no sur- 

 veys and had nol changed its standards for 

 trichloropropane since January 1989. when it 

 issued a regulation limiting airborne emis- 

 sions of the subslance. 



As for telling people ol Ihe dangers, Dr 

 Slein added, "The primary manufacturers of 

 the product would be responsible." 

 ' "1 presume when updating training pro- 

 grams at companies lhat use this, say annu- 

 ally, whoever is doing that would be aware of 

 the new information," Dr Slein said. "They 

 would makV the employees aware of it, but 

 I'm nol sur'&lf that is actually being done." 

 "We always have a battle on the issue of 

 what to do with Ihe animal data." Dr. Slein 

 added. "I'm nol trying lodownplay it. but I do 

 ; believe other things ought to have priority." 

 So back in North Carolina. Dr. Irwin said 

 "1 really haven't heard of anything happen 

 ing It's almost as if our work just goes into a 

 black box." 



Acknowledging lhal problem. Dr. Olden 

 said: "1 have to say we don't serve the 

 American people. very well right now. Bui 

 that's where we are " 



