Document e5vXxoJEXE0dZLRb682MZXapq

BY K A R E N WR I G H T -WHAT I WANT TO TALK TO YOU ABOUT today are trends in cancer in industrial coun tries." So begins the lecture that Devra Lee Davis has delivered in dozens of formal appearances and extemporaneous discussions this year, a monologue she recites for anyone who wants to listen and for a good many who don't. "Our data draws on mortality statistics for six countries, provided through the World Health Organization and checked for quality," she says. Translation: 1am not pulling numbers out of thin air. ' "This is truly an international collaborative effort," she adds, naming colleagues from three countries. Translation: 1am not the lone evangel ist 1appear to be. * Karen Wright is a contributing correspondent o f Science magazine and a former editor at Scientif ic American. "May 1 have the first slide, please." Up in the and big reputations at stake in cancer resea top tier of the auditorium at the University of and there are more than half a million liv California at San Francisco in April, Davis's 11- year to be saved in this country alone. Bl year-old daughter. Lea, flicks on the projector. assertions about lack of progress in the so-ca Lea has heard her mother's disquisition so many war against cancer are never greeted warm ! times that she once fell asleep at the table when the cancer establishment. And consensus am the subject came up over dinner. epidemiologists, the fractious corp of med Davis's message has the opposite effect on scientists who study the distribution of diseas just about everyone else. As a specialist in hope of discovering causes, is rare even when public health policy and scholar in residence at stakes are low. the National Research Councilof the National But Davis goes further, suggesting that m Academy of Sciences, a prestigious position she of those cancers may be caused by prevent! was appointed to in 1989, Davis has compiled factors like exposure to chemicals in the wt one of the few system atic comparisons of recent place and changes In exercise and dietary hal changes in deaths from cancer. Culling informa Her studies .aren't detailed enough to point tion from millions of death certificates, Davis finger at any one culprit, but Davis suspects has shown that deaths from breast' cancer, .causes are there because the increases h brain cancer, kidney cancer, myeloma, melano -occurred so suddenly and because cancer n m a and non-Hodgkin's lymphoma have climbed are significantly lower in some countries t since the late 1960's, particularly among the they are in others. elderly. And then she ups the ante by using h e r con That claim in itself is enough to make headlines sions to argue that the United States is -- and enemies. There are big money, big prizes putting enough money into research on can A resident scholar at the prestigious N ational A cadem y o f Sciences, Davis recites her statistics to anyone w ho w ants to listen (and man] w ho d o n 't). H e re , she advises tw o law y ers w h o d o --- th ey a re w o rk in g o n a case in v o lv in g p u b lic h e a lth issues. Left: Davis chats w ith M arvin A. Sch neidermair, a form er associate director o f the N ational Cancer In stitu te, w hile A lbert Einstein eavesdrops outside the N ational Academy o f Sciences offices in W ashington. Below: M o th er and daughter, Lea, aw ait a routine doctor's appointm ent. prevention. With death rates rising from uniden tified causes, she says, research on treatm ent -- currently the lion's share of the Government's (2 billion annual cancer program -- seems a feeble palliative. Care for cancer patients al ready costs nearly $30 billion a year. Ultimate ly, Davis maintains, Ui only way to get canej: under control is by naming and eliminating its causes. ^ In some ways, her timing couldn't be better. On the 20th anniversary of the signing of the National Cancer Act, the country's cancer effort lias been skewered between new developments and new disappointments. Though basic re search programs have made spectacular ad vances In the understanding of cancer biology, those advances have not revolutionized treat ments for the most common cancers. The' National Cancer Institute's 1982 goal of reducing cancer deaths 50 percent by the end of the century looks ludicrous today. The mortality rate Is actually higher than it was when the w ar on cancer began, even though death rates for some cancers, like those that afflict children, have gone way down and even though at least 30 percent of all cancer deaths are due to a fully preventable cause, smoking. But Davis's scrutiny is not welcomed Whether through manner or m aterial, she has managed to alienate two of the world's leading experts on cancer statistics: th redoubtable Oxford physi cian Sir Richard Doll, who with A. Bradford Hill helped establish the link between smoking and lung cancer in the 1950's, and his notoriously feisty protg, Richard Peto. Doll explained to Davis years ago that he thought her ideas were both wrong and unoriginal; he hasn't changed his mind since. And Peto's feelings, after much subterranean rumbling, finally erupted in print in November 1990 when he and Doll told the journal Science that Davis's work was "uninter esting," "uninformative," "boring" and "old junk." "I do not know what explains these trends that I've shown you today," Davis says a t the conclu sion of her San Francisco lecture. "But I think we need to find out. What we learn may allow us ultimately to come up with strategies to prevent thousands of millions of deaths.'* Your average' scientist might try to tone down her rhetoric. Devra Lee Davis is not your average scientist. ALBERT EINSTEIN SITS IN A GARDEN OUTside the entrance to the building that houses the august National Academy of Sciences in Wash ington. His granite dais has a view across.Consti tution Avenue to the Mail, the memorials, the cherry blossoms, the tourists. But Einstein -- in bronze, larger than life -- never lifts his eyes from the stainless-steel map of the constellations at his feet. "The right to search for truth implies also a duty," he says in a marble exhortation. "One must not conceal any part of what one has recognized to be true." In an office a hundred yards away sits Davis, the antidote to the statue's sobriety. She talks a lot, wears bright red and stands 5 feet 8 inches tall, not counting the corona of thick brown curls that envel op her head. Photos of tier husband and children, Lea and Aaron, fill the walls where she could have hung her master's degree (tn. public health from Johns Hopkins) and lifer Ph.D.' (in "science studies" from the University o f Chicago). Davis does aero bics, speaks Russian, studies Torah, plays the cello, coaches her children's soccer teams, is host for international meetings, rescues stray dogs and is evasive about her age (she's 45). With little more than a decade of work on ca leer trendsunder her belt, she is an upstart of t ie most fearful kind* earnest, outspoken, indefatigable^ And, perhaps sig nificantly, female: a bejewehd brunette amid gray heads and bow ties. "A tremendous decline in cancer deaths has oc curred in people under age 45 in 'the United States," Davis begins. "A tremendous di;eline. Due largely to improvements in treatment for childhood can cers..But in the. age, groups over 45; tiieire. is an increase." It is most (Continued on page 77) *4 ffitinuedfront page 60) jnounced in people more a n 64 years old, who acKint for about two-thirds of 1cancer mortalities. "Now, we're not talking jout small increases here," ic says. "Some of these cansrs have increased 25 per mit to more than 200 per mit" In the past two decades. VTien something increases i rapidly over such a short :riod of time, that strongly iggests there may be a preintable cause behind it." Last year, in a paper pubihed fn the British medical mrnal Lancet, Davis and her )LJeagues reported similar ses In England and Wales, est Germany, France, Italy k! Japan. "All forms of am ir are increasing fn persons rer age 54 except lung and omach," she wrote. "The ranges .. . are so great and ipid that it would be impru* mt not to investigate their juses aggressively." That conclusion put Davis i network television and in itional newspapers. It also at her fn the doghouse as far s some of her peers were jncerned. Critics saw the ancet report as the most msatlonal in a series Davis nd various co-authors had egun in the late 197G*s, when avis worked as a policy anarst and science director at w Environmental Law Insti- ite in Washington. Between rticles on risk assessment, cid rain and asbestos* .Davis egan looking at occupational xposure to cancer-causing hemicals. She pointed out lat chemical production had icreased dramatically in his country since the 1960's nd that because of the laency periods typical of such arcinogens the potential efect of the Increase on cart e r mortality would probaiiy not be evident for an ther decade or so -- not mtil, say, the 1990's. But, as Davis concedes, nth cancers associated with ndustrial exposures amount- ng to only about 6 percent of ill cases, increases (hie to greater chemical production vould be difficult to detect. So h e began to examine deaths rom particular types of canair in people 65 and over, peo ple who were likely to have longer histories of exposure. That's when she crossed >jU and Peto the first tim a [n their classic 1981 treatise sn the causes of cancer, the Oxford duo had maintained Jiat the data on cancer vic tims older than 64 was too unreliable to be used in evalu ations of cancer trends. Ap parent increases in mortality in the elderly, they said, may have resulted from greater access to medical care and improvements in diagnosis-- influences that statisticians call artifacts. So Doll and Peto limited their appraisal to those 64 and under, and found no cause for alarm. In a conference report from the early 1980's, Peto took a side long dig at Davis with his as sertion that "opposite conclu sions by other commentators appear to derive chiefly from methodological oversights." Davis was unpersuaded. Doll and Peto's calculus dis counted more than 60 per cent of all cancer deaths. She did agree that advances in diagnostic equipment and techniques and the start-up of Medicare In the mid1960's could have introduced artifacts In the data of the time period -- from 1933 to 1977 -- that Doll and Peto had chosen to study. Her so lution was to consider the data only from 1968 on. There hadn't been much change in access to medical care after that date. It is true that CAT scans and magnetic resonance imag ing machines brought m ajor improvements in diagnosing cancer. But in the mid-80's, after she'd moved to the academy, Davis did a num ber of studies to show that improved diagnosis alone couldn't explain the increase in cancer deaths in the eld erly. One was an analysis of deaths between 1968 and 1978 from senility, cancer of unspecified type and "un specified" causes. If the ap parent increase in cancer mortality was simply the re sult of better diagnosis, Da vis reasoned, then deaths in these "default" categories should drop, reflecting the detection of previously un recognized tumors. But they had risen as well. The more she looked at the statistics; the more alarmed Davis became. At first, she confined her inter ests to brain, lung and bone cancer (myeloma), because those were among the can cers more likely to be caused by occupational ex posure to carcinogens. But the larger question of can cer trends in older popula tions led her to include other kinds of cancer, and to ex clude lung and stomach can cer, whose causes are fairly well known. Her Lancet pa per drew attention to in creases in brain and breast cancers and in myeloma, melanoma and non-Hodg kin's lymphoma. It also showed that rates of those cancers are relatively low in some industrialized coun tries, most notably Japan. The fact that Japanese im m igrants to this country quickly develop cancer rates similar to those of the natives further indicates that there may be avoidable causes of those cancers. Davis's analyses do not at tempt to pame what those causes might be; it takes epi demiologists years to arrive a t a cause, even In obvious cases like the link between asbestos and cancer of the respiratory tract. She likes to call her work "hypothesis generating" --meaning it raises more questions than it answers. Why has brain cancer doubled in people over age 65 since 1968? Why do Japanese women have lower rates of breast cancer than women In the Western Industrialized nations? Why do black men have a 50 per cent higher death rate from lung cancer than white men, even though they sta rt smoking later in life? Why is the death rate from melano m a increasing almost 5 per cent a ' year among fair skinned older people? Some observers have ex plained trends In the elderly by evoking what's known as the "competing causes of death" argument: that fewer people are dying of heart dis ease and other maladies, so more die of cancer. But that fact shouldn't change the rate of cancer mortality in spe cific age groups, Davis points o u t If more people live to be 75, then the total number of 75-year-otds who die from cancer will go up, but the pro portion of 75-year-olds should stay the same. It may be that the miracles of modem medicine are saving people who are more likely than their naturally resilient con temporaries to develop can cer. "But even if that's part of the answer, it still means we have something to pre vent," she says. The dissonance between the Doll and Davis camps reached a crescendo of sorts at a September 1990 meeting on the origins of human can cer held at the Cold Spring Harbor Laboratory on Long Island. In an opening night address, Samuel Broder, di rector of the National Cancer Institute, mentioned that "those over 65 have dispro portionate high rates of can cer mortality." Davis look it as an implicit endorsement of her position and couldn't re sist asking a rival what he thought of Broder's remark. Being Davis, she asked a fair ly illustrious rival -- Prof. Bruce N. Ames of the Univer sity of California at Berkeley, who pioneered the testing of potential carcinogens. Being Davis, she confronted Ames while dozens of colleagues looked on. And being Davis, she wound up in an argument with Peto, when Ames de ferred the question to him. Scientific disputes are rarely so public. The enmity over Davis's position seems obsessive at times, degener ating into what she calls "ad ieminam" attacks. Ames, who says cancer statistics are not a cause for alarm, keeps a file on Davis be cause "she's always coming out with something 1 dis agree with." He says Davis's interpretations present am biguous trends in the worst possible light. "She's out for publicity," he says. "I think she's wrong on everything. Well, not every thing -- she's not that reli able." Anyway, says Ames, "the world's leading epidemi ologists don't agree with her." He means Doll and Peto. "Let me tell you some thing," says Davis. "There's nobody who'd be happier to learn that there's some great artifact here. But I don't think it's doing any body any good to assume that all of these increases, in men and women every where, are just some fluke that seems to be afflicting medical statistics world wide." NOT ANXIOUS I to be joined in any controversy with Dr. I Davis," says Sir I Richard Doll, who is I speaking -- ever so I cautiously -- on the record. His partner, Peto has decided to decline that privilege. "Nobody argues about the mortality in many devel oped countries in people over 65 going up," says Doll. "The question is, Is it of any particular importance with respect to causes?" Doll maintains that when you look a t the trends for each type of cancer, they seem much less formidable. In creases in lung cancer, for example, are clearly due to smoking. Deaths from stom ach cancer have plummet ed, probably because of changes' in d ie t' Doll be lieves the increases in brain cancer are almost certainly the result of better diagno sis. (Davis maintains, with some corroboration from re cent independent studies, that these alone cannot ac count for all the increases.) "I'm not saying there are no new hazards," Doll says. But if there were unidenti fied risk factors, he says, they would show up as in creasing cancers in younger people as well as the elderly. That's one of the reasons he and Peto felt safe in exclud ing cancer deaths in people over 64 from their 1981 monograph: "There's no ex ample of any cancer that re quires longer than 15 or 20 years to reach a fairly high Incidence as a result of expo sure to a carcinogen. You're not going to have an epidem ic occurring 30 years after first exposure." And even if some cancers are rising mysteriously, be maintains, Davis isn't the first person to notice iL The increases in melanoma, my eloma and non-Hodgkin's lymphoma are probably le gitimate, he says, and are still Inexplicable But "these have been a matter of scien tific knowledge for a long time before Devra Davis drew attention to them." So Is Davis wrong, or just unoriginal? "Both," says DoIL "Both." Given his tepid opinion of her work, Doll might be sur prised to learn that Davis ends h er lectures with an homage to his efforts. In fact, he's the figure she seem s to identify with most -- not the esteemed Sir Richard of the present but J the young, modest, methodi- v cal researcher who antago- f nized the medical establish- \ ment in the late 1940's andj early `50's with his insist ence that there was a sinis-^ ter correlation between the * prevalence of cigarette^ smoking and the sharp rise^ in lung cancer. "1don't think you can imagine the respect and awe with which most of us regard Richard Doll," she says. In an adroit display of tautology, pathologists had argued that smoking could n't cause cancer because they knew of no mechanism by which the chemicals in cigarette smoke could cause cancer. The increase might ju st reflect the fact that doc tors were getting better at distinguishing between lung cancer and tuberculosis, they said. It wasn't until 1964 that the United States Sur- on General formally ac- she's going about it the right after, Davis, who had been define risk facto lowledged the smoking way, by just gathering data hired at the National Acad td lung cancer connection. and presenting it. You can emy of Sciences the year be The moral of the story, ac- draw your own conclusions." fore, decided to devote her Her work is won't know how to advfte * individuals to d g e t-:^ cancer. "The real danger is irding to Davis, is that a new self to cancer research. She `h y p o th esis,ea in science- rarely gets a DEVRA LEE DAVIS GREW says her father's Illness had that w e're going to be able to tell people they're a t risk be arm reception. "At first, up In Donora, Pa^ a small no effect on her decision to sople say it's ridiculous," town south of Pittsburgh pursue her studies of cancer re says. "Then they say, that's famous for two things: trends: "I don't want my own i. F fore we know enough about generating,' '. , It to prevent IL" The problem is that die fell, it might be true, but It's the baseball Hall of Famer story to detract from the * important.* Then they say, Stan MusLai and the country's m ajor concern I have, which raising more", most obvious --and potent -- causes of cancer 'have al* fell, actually, I discovered worst air-pollution disaster. is that there are preventable - .ready been discovered. Car Is myself.' " Davis unabash- In October 1948, when Da causes of cancer out there questions'.''; ,Uy applies this moral to her vis was 2 years old, a high- and we don't know what they cinogens like radiation and asbestos, for example, were vn situation. pressure condition swept tox are." But It's not clear what ic fumes' from a nearby steel A few years later, she be Lapter of the story she's In mill into the horseshoe gan a 16-month sabbatical 'thafrit, relatively simple to Identify because they caused spikes in cancer rates in select popula *w. It seems, at times, that shaped valley the Mononga- from her staff position at the answers. ' . 'avis cannot count even on hela River cut around Don National Academy of Sci tions. All an epidemiologist had to do-to pinpoint a cause e support of her collaboru ora. The hills that surround ences to travel in Europe was examine how the habits ns. Alan D. Lopez, a stalls- the town trapped the killer and collect the data that of those afflicted differed :ion for the World Health smog for five days. Twenty eventually went into her institute in 1986, when he from the habits of the popu rgani ration in Geneva, people died of heart and lung Lancet paper. As head of the published a heretical cri lation as a whole. But the rw claims he was "sur- trauma; all told, almost half board on environmental tique of the American can new gene ration of cancer ised" to see himself listed of the town's residents were studies and toxicology at the cer effort. Most of the insti causes, which specialists i a co-author of the Lancet affected. A decade later, academy, she had had three tute's so-called prevention prefer to call risk factors, tid e that pushed Davis death rates among Donorians personal assistants and 45 research money, he says, is are less potent and more dif to the spotlight. His office who had been affected were staffers, coordinated hun going Into basic etiology -- fuse -- like fat In. the diet. xrvided the data for the pa still higher than average. dreds of experts In 30 cities the cellular mechanisms of The tobacco or asbestos ir and he sent Davis copl- Davis was "completely un and presided over consensus cancer development -- rath-' models do not apply. "That's ts. comments on It; he ad aware" of the smog. "The sky reports like the one that led e r than epidemiological not the way breast cancer its he noticed that his was always dark. It did not to the ban on smoking on studies and prevention tri develops" says Edward ime was on several drafts, strike me as unusual." For a most domestic airline als. "The prevention of can Sondlk, deputy director of it he says the paper .was child, chronic pollution af flights. But when the sabbat cer on a big scale is going to the cancer Institute's divi iblished before he had a forded novel opportunities for ical was over, in 1989, Davis require that we change our sion of cancer prevention mnee to tell Davis to take ft adventure, including sliding resigned from her director habits, change our life and control. "That's not the r. down hills conveniently de ship to pursue cancer re styles, clean up the work way colon cancer develops. Davis thinks Lopez's d e nuded of vegetation. search full time. Now, as a place, clean up the environ Do you think we're going to al is the result of pressure Davis says it's just coinci scholar in residence, she ment, change the consumer find that there's a particular om Doll and Peto. Lopez dence that she ended up in works alone and without the products that contain haz cause of melanoma at this :>es say he had to re-lngra- public health, doing the kind academy's imprimatur. ardous m aterials," says Bai point in time? Do you think ate himself with the can- of work that eventually In her argument that more lar. "It's going to mean a we're going to find that ;r establishm ent a fter the cleaned up her hometown. money and effort should go whole new approach to ev there's a particular cause of ancet article cam e out -- Her undergraduate career into research on prevention eryday living. colorectal cancer, when it's re establishm ent being at the University of Pitts as the rational weapon "If we had good treat as widespread as it Is?" oil and Peto, with whom burgh was an amorphous against cancer, she has the ments, they would probably John Calms maintains that : has been working since collage of science courses, weight of history behind her. be far less expensive and there are "huge financial in )89 and who, Lopez ad- civil rights activism and The scourges of the recent disruptive than prevention. centives" for governments to dts, had "a few sharp German philosophy; after past -- malaria, tuberculosis, The problem with that ap do nothing about cancer pre ords" for him. she got her Ph.D. from the typhus, cholera and others -- proach is that it hasn't vention. Sales and excise tax "That's a sad comment on University of Chicago, she have been controlled primari worked. Cancer death rates es on cigarettes, for example, ie politics of the issue," Da wrote scholarly articles an ly through prevention, not are higher now than they provide about 514 billion in is says. "For some reason, acupuncture, door-to-door treatm ent On economic were when we started 50 annual revenue. And smoking oil and Peto are exercised interviewing in urban neigh grounds alone. It's unreason years ago. At some point, I saves billions more every bout this far beyond what is borhoods and the psycholog able to expect any other stra t think somebody has to step year on Social Security pay lerited. And I really can't ical effects of overcrowding egy to work against cancer. back and ask whether we're ments because smokers on g'ure out why." in cities: It was her postdoc The direct and indirect cost of going about this the wrong average die sooner than non- Lopez offers one explana- toral adviser at Johns Hop cancer now amounts to near way." smokers. ior? when he says Davis's kins University, the late ly 2 percent of the United There's no shortage of de There might be disincen rork "is dangerous in the Abraham M. Lilienfeld, who States gross national product fenders of the way the Na tives for the private sector, ense th at it draws the prin- encouraged h er to read Wil "Even If we could cure every tional Cancer Institute as welL "When you treat ipal public health concern liam F a rr and John Snow, single case of cancer, we spends its appropriations. cancer, profits a re made way^from cigarettes." Or It the English patriarchs of can't afford to," says Davis. Though Sam Broder, the in through drugs or surgery," iiaybe that statistical inter- epidemiology. The story of John Cairns of the Harvard stitute's director, declined to Davis points out. "But when rotation is a kind of esthet- how Snow stemmed a Lon University School of Public discuss the country's cancer cancer is prevented, nobody 0 and that scientists be- don cholera epidemic by de Health made a similar argu effort with me, Vincent T. makes any money." She ome wed to their personal ducing its connection to the ment in a 1985 article in Sci DeVita of the Memorial claims the impetus from the isions. "Doll and Peto stuck water supply is heady read entific American: "To put Sloan-Kettering Cancer Cen public isn't there, either. heir necks out before," says ing. even 140 years after the most of the effort into treat ter, who headed the institute "Where's the constituency rhomas G Chalmers of the fact. "We need people like m ent" he wrote, "is to deny from 1980 to 1988, says the for cancer research?" she iarvard University School Snow -- people who think all precedent." basic research has "paid off says. " It comes often from >f Public Health, referring about the big picture," Li- About a third of the Na handsomely." The discovery the families of cancer vic o the conclusions of their lienfeld told her. tional Cancer Institute's $2 of specific genes for predis tims, who are deeply moved .981 report, which Davis's In 1979, a year after Devra billion annual budget goes to positions to cancer "will give and want to do something vork contests. "And they're Davis wrote her first paper prevention research. But us the tools, finally, to define about treatment. T here is al inwtUing to admit they were on cancer trends, her father those are "rubber num qt-risk populations," he most no constituency for vrong. was told he had myeloma. A bers," says McGill Universi says. "It's the first time that cancer prevention." "It would be hard to insure former machinist at U.S. ty's John C Baitar 3d; a 35- there's a clarity to the idea Perhaps Davis knows bet hat Devra, too, wasn't blind- Steel, he was 55 and a suc year veteran of cancer- of preventing cancer." ter than anyone how hard id by a desire to make a cessful businessman at the trend research. Bailar him But even DeVita adm its ` building that constituency contribution. But l think time. He died in 1984. Soon self antagonized the cancer that without epidemiology to will be.