Tobacco Documents Online

. A Mstoy iiof Cancer Control in the U n ited States 1946-1971 Book One A History of Scientific and Technical Advances in Cancer Control Prepared by the History of Cancer Control Project, UCLA School oi Public Health pursuant to Contract No. NOI-CN-55172, Division of Cancer Control and Rehabilitation, National Cancer Institute; principal investigator, Lester Breslow, M.D., M.P.H. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE Public Health Service National Institutes of Health National Cancer Institute Division of Cancer Control and Rehabilitation DHEW Publication No. (NIH) 79-1517 1

BOOK ONE A HISTORY OF SCIENTIFIC AND TECHNICAL ADVANCES r IN CANCER CONTROL 125

I1CKBLOCI{'S CARTOON JAN "COULP YOU NURRY AAV FINP A CURE FOR CANCER? TNAY WtllD BE SO MUCH EASIER TNAN PRE1dEtAl'lON

TABLE OF CONTENTS (BOOK ONE) Chapter 1. Occupational Carcinogenesis ................ 13i Chapter 2. Carcinogenesis Bioassays ................... 157 , Chapter 3. Tobaccogenic Cancer ........................ 109 Chapter 4. Detection of Uterine Cervix Cancer ......... 197 Chapter S. The Role of Mammography in the Detection of Breast Cancer ........................... Chapter 6. The Detection and Diagnosis of Large Bowel Cancer ............................... Chapter 7. Cancer Treatment ........................... Chapter 8. The Rehabilitation and Continuing Care of the Cancer Patient ...................... r 273 319 3069 423 ' r, 129

I CHAPTER 3 TOBACCOGDNIC CANCER Introduction In 1919, when Alton Ochsner was a junior in medical school at Washington University in St. Louis, a patient with cancer of the lung was admitted to Barnes Hospital, the teaching hospital for Washington University. In a short time, the patient died. Ochsner recalled the incident: Dr. George Dock, who was ar} eminent clinician and pathologist, asked the two senior classes to witness the autopsy because, as he succinctly said, the condition was so rare he thought we might never see another case as long as we lived. Being very young at the time and enamored by the clinical knowledge and judgment of our eminent professor of medicine, I was greatly impressed by this extremely rare condition. (1) hfiien Ochsner recorded these recollections in 1973, he was 77 years of age. Lung cancer--a condition "so rare. ..we might never see another case as long as we lived"--had, in fact, become a modern epidemic of massive proportions. The disease currently kills upwards of 83,000 Americans each year; in 1976 an estimated 65,200 men and 18,600 women died from the disease. (2) As a result of perhaps the most important * Principal Researcher/hYiter: Larry Agran 169 i i O Cr W

twentieth century advance in carcinogenesis, it is now stated with con- fidence that approximately 80 percent of all lung cancer deaths in the United States are caused by cigarette smoking. (3) The Early Evidence After graduating from medical school, A1ton Ochsner went on to become a surgeon and an early leader in lung cancer surgery. He was also among the earliest scientists to explore the relationship between cancer of the lung and the use of tobacco. Seventeen years elapsed before I saw another case of lung cancer, at the Charity Hospital in New Orleans after having come to Tulane University as Professor of Surgery in 1927. There was nothing particularly unusual about seeing a rare case in 17 years, but eight other additional cases were seen in a period of six months which,was extremely unusual. Having been impressed with the extreme rarity of the condition 17 years previously, the sudden increase in incidence represented an epidemic, and there had to be some reason for it. All the patients involved were men; they all smoked cigarettes heavily and had begun smoking in the First World War. I then ascertained that very few cigarettes were consumed before the First World War but during the war and after- ward there had been a tremendous increase. Since there was a parallel in the rise in sale of cigarettes and the appearance of the new disease with a lag of approximately 20 years from 1914 to 1936, I considered that this might be the necessary length of time for a possible carcinogenic agent in tobacco smoke to become evident. The evidence was admittedly very nebulous, but it seemed as if this was the most likely cause. (4) European studies, most notably that of MUller in Germany in 1939, (5) began to show a strong statistical association between lung cancer and smoking. Two years later, in 1941, Ochsner joined with Michael DeBakey in publishing the first American study to stress the cigarette-lung cancer connection. (6) Based on clinical observations of autopsies performed in the United States and in other countries, these researchers found that the incidence of pullranary carcinoma had doubled over the 18-year period studied, 'Vhereas the increase in the incidence of all carcinoma in all U 170 0 I~ 0

autopsies.kas relatively slight." (7) Noting the parallel increase in cigarette sales over the same period, the study concluded, "It is our definite conviction that the increase of pulmonary carcinoma is due largely to the increase in smoking, particularly cigarette smoking, which is universally associated with inhalation." (8) Dr. Michael Shimkin has written of epidemiologists in these terms: Epidemiologists are a mixed lot and come from many walks of medical, sociological, and economic persuasions. They include statisticians who refuse to be browbeaten by clinicians, physi- cians who acquire a nodding acquaintance with statistics, and geographic pathologists who learn to distinguish pathology specimens from people. There are also macroepidemiologists, who consider it beneath their dignity to deal with populations of less than 100,000, and microepidemiologists, who look for intuitive insights in unusual small clinical experiences. (9) Intuition. Pathology. Statistics. E4ch was important as the cigarette controversy emerged as the leading field in cancer epidemiology shortly after World War II. Interestingly, amid an atmosphere con- ducive to scientific inquiry, Washington University contributed more than its fair share of prominent personalities to the fray. Of course, Alton Ochsner was himself a product of Washington University. But there were others. Dr. Evarts Graham, who in 1933 performed the first success- ful pneumonectomy for cancer of the lung, was Ochsner's professor of surgery in his senior year. Some years after his graduation, when Ochsner first postulated that the increase in lung cancer was due to cigarette smoking because of the parallel between the sale of cigarettes and the increasing incidence of the disease, he was chided by Graham. Graham, who was a very heavy cigarette smoker, said, "Xes, there is a parallel be- ' tween the sale of cigarettes and the incidence of cancer of the lung, but there is also a parallel between the sale of nylon stockings and the in- cidence of cancer of the lung." (10) Ochsner recalled further: , 171 O ~ ~ C ... I

A few years later Dr. Graham wz-ote to me and reminded me (of the incident] and said that he would have to "eat crow" because a young man, a sophomore student at Washington University, had taken his jDr. Graham's] cases of cancer of the lung and studied them and the results of this-study convinced Dr. Graham that there was a relationship between cigarette smoking and cancer of the lung. This young sophomore student was Ernst Wynder. . . . (11) In 1950, Graham and Wynder together published the results of their epidemiological study. (12) In their investigation, they employed a retrospective method of study: They interviewed patients already known to have lung cancer and, inquiring about their smoking habits, they then compared these responses to the responses of patients without.lun,g cancer. The results indicated that proportionately more heavy smokers were found among the lung cancer patients than the control group population; fewer light and non-smokers were found among the cancer patients than among the controls. 'Graham and Wynder concluded: ~ "Extensive and prolonged use of tobacco, especially cigarettes, seems to be an important factor in the inducement of bronchogenic carcinoma." (13), Persuaded by the evidence, Graham altered his personal'smoking habits, decreasing his cigarette consumption to six per day=-two-after each-meal: Then, in 1953, when Graham and Wynder were able to prove that the tar from cigarette smoke when applied to the surface of animals produced skin cancer, (14) Graham quit smoking altogether. But it was too late. A few years later lle wrote to Alton Ochsner, "Because of our long friendship, you will be interested in knowing-that they found that I have cancer in both my lungs. As you know, I stopped smoking several years ago but after having smoked as much as I did for so many years, too much damage had been done." (1S) Wynder recalled the tragic irony of Graham's death. 172

When he was dying I went to St. Louis. He was laying (~ic3 in an oxygen tent. I remember he pointed to a little sign on the oxygen tent where it said "No Smoking." lie said, "I should have listened." He wrote me a very moving letter stating that fate had really done him badly for all the work he had done on lung cancer. (16) On March 4, 1957, Dr. Evarts Graham, the first person to surgically remove a human lung, was himself dead from lung cancer. Important as it was as an epidemiological study, the Wynder-Graham investigation was but the beginning of a great mid-century scientific debate. Serious questions remained, not the least of them directed to the techniques employed in the Wynder-Graham study itself. The questions were raised not only by the tobacco interests that were ob- viously threatened by the study's conclusions, but also by scientists of considerable reputation. Dr. Lester Breslow, currently Dean of the School of Public Health at UCLA, was a state public health epidemiologist in California in the late 1940s when he first encountered Ernst Wynder at work on the Wynder-Graham study. About 1947-48--in that period--we were visited in Berkeley by a medical student named Ernst Wynder. He came in with an obviously very strong conviction that cigarette smoking was a factor in lung cancer. Wynder had undertaken what we later be- gan to call a retrospective or case control study of the matter. He came by to let us know that he was going to be visiting the hospitals in the Bay Area to interview patients and controls in regard to their smoking practices. - We thought he was a pretty brash young man. ..and asked whether a member of our staff, named Hoaglin, could accompany him around to the hospitals just to see what he was doing. Hoaglin came back with a horrible story of poor technique, a very sloppy approach to the interviewing.. And so we decided we ought to do a proper kind of study. We were quite astonished with the results, which were almost identical with those Wynder was obtaining by what had appeared to us as very biased and sloppy techniques. (17) It was this remarkable similarity of results that made early con- verts of Breslow and others. But still more was needed before a con- sensus among scientists would emerge on the matter. Breslow's early 173

work, (18) like Ochsner's and Grahan: and Wynder's, was a retrospective study with all its attendant problems. According to Breslow: It seemed to me that the retrospective approach to the matter--the case control studies--were vulnerable methodologically on the grounds of bias of samples. The people already had the disease; they were selected people who were being interviewed; aad it is very difficult, if not impossible, to get perfect controls. You depend upon what the patient recollects and is willing to tell you and the accuracy of what he says about his smoking habits. There were a lot of reasons why one could doubt the significance of these retrospective studies. The issue was only going to be resolved by what we later came to call prospective studies. (19) The Emerging Consensus r Soon after L'he Wynder-Graham study was published in 1950, a number of prospective studies were organized throughout the country. The prospective method examines the smoking habits of a sizeable population-- apparently healthy--and then follows that group over a period of years in which -":te rateG and causes* of mortality are recorded. In this way, the problerrs or retrospective falsification or the failure of memory or the selection of improper controls are avoided. The most inf-'lue.itial pros.pective study in the United States was that undertaken by Drs. E. Cuyler Hanmond and Daniel Horn. (20) With the assistance of American Cancer Society volunteers, Hamrrond and Norn tracked 187,783 American men to determine what effect, if any, smoking habits had upon mortality. The results, published in the Journal of the American Medical Association in March, 1958, (21) confirmed the findings of the retrospective studies. In fact, all of the prospective studies of the late 1950s--including the 1956 study of British doctors by Doll and Hill (22) and the 1959 Dorn study of 200,000 U.S. war veterans with government life insurance policies--(23) showed that the total death rate for cigarette smokers was 174