---------------------------------------------------------------------------- ---------------------------------------------------------------------------- CHAPTER 38 Summary on Underestimation of Dose in Our Study ---------------------------------------------------------------------------- ---------------------------------------------------------------------------- Part 1. Why We Need to Know about the Exposures We Have NOT Measured =========================================== The entire principle of this study of causation of breast-cancer is based upon the identification and measurement of sources of annual breast-irradiation over the period from 1920 to 1960. If our goal is to identify how large a fraction of breast-cancer causation ionizing radiation is responsible for, we need to identify ALL the sources, with measurement of the dose received annually from each source. For every source we cannot quantify, we are ""losing"" some part of the causation. And we will, therefore, be underestimating the radiation causation of breast-cancer as the result. We realized at the outset of these investigations that there were undoubtedly going to be some major sources which we would not be able to quantify. But we also realized that for every source we could identify and quantify, we could be sure of at least the minimum extent to which ionizing radiation was a cause of breast-cancer. That ionizing radiation is a cause of breast-cancer was without any doubt at the outset, no matter how poorly that appears to have been recognized by the uninformed or mis-informed. The Effort Went Better Than We Expected --------------------------------------- We have been surprised and amazed at what it was possible to accomplish in this quest. The fact that we have identified ionizing radiation as the major causal factor in breast-cancer and have identified much about the time periods which are involved in breast-cancer development, spurs us on to try to evaluate QUANTITATIVELY any remaining radiation factors. The problem is NOT that we do not know the remaining radiation factors. Rather, the nature of the scientific literature is such that it is simply not easy to get reliable information to quantify them. Part 2 discusses three major areas of concern. Part 2. The Major Remaining Radiation Factors: How Large a Role? ================================================================== o - First: Fluoroscopy (or Roentgenoscopy) -------------------------------------------------- Fluoroscopy was very popular, and was used by even more non-radiologists than radiologists (Chapter 31, Part 5). There is simply no doubt that fluoroscopy was used in multitudinous ways which irradiated breast tissue, but which were not covered by our method in Chapter 23. A prime example, separately and independently revealed by Blatz, Pifer, and Buschke/Parker, was inclusion of pediatric fluoroscopy in well-baby and well-child examinations (Chapter 31, Parts 2 and 3). Yet our Master Table completely excludes such use. We have described the papers from Metropolitan Life Insurance Company, where nearly 5,000 consecutive applicants were fluoroscoped to screen out tuberculosis (Chapter 30). After the initial publications, how many additional applicants nationally were processed by this same company? How many additional companies and industries did this too, and did not write a paper on it? Yet our Master Table excludes such pre-employment uses of fluoroscopy. And these examples are not the only uses of fluoroscopy for which there is no permanent record of dose received by female breasts. This is one of the big problems about fluoroscopy; it was (and is) widely used, and there is almost never a record of WHAT dose was given to HOW MANY persons. Numerous pieces of evidence point to fluoroscopy as likely to be the dominant source of radiation breast-exposure. We contemplate with amazement Braestrup's comment that New York City passed an ordinance forbidding fluoroscopic exposures higher than 100 Roentgens per exam. By itself, the unevaluated use of fluoroscopy in the 1920-1960 period might account for an annual average breast-dose larger than the combined sources in the Master Table. o - Second: Dermatologic Use of Radium Gamma Rays and X-Rays in Therapy -------------------------------------------------------- There can simply be no doubt that dermatological use of radium gamma rays and x-rays up through 1960 can only be described as massive. The perusal of Chapters 33 and 34 makes that quite clear. We know that the dermatologists prided themselves on the therapy of some 80 different dermatological disorders, wholly aside from any therapy of superficial or deep cancers. We know that this continued unabated over many decades. We know that the radiation doses to tissues would undoubtedly have been high in those dermatological applications which were in the physical neighborhood of breast-tissue, or where breast- skin was in fact the organ being treated. But we do not generally have any way to measure just what fraction of the breast-tissue was irradiated in the dermatological applications. Skin disorders are common. Doses were high. It is possible that, by itself, the unevaluated dermatologic uses of radiation in the 1920-1960 period also might account for an annual average breast-dose larger than the combined sources in the Master Table. o - Third: The Array of Inflammatory Disorders Reviewed in Chapter 36 ---------------------------------------------------------- A vast number of inflammatory disorders --- from tuberculous lymphadenitis to sore shoulder --- used to be treated with radiation therapy. Chapter 36 lists many disorders for which radiation therapy must have caused breast-irradiation, sometimes or often. Even if therapy of one disorder by itself did NOT add a great deal to the annual average breast-dose, the COMBINED uses may have made an appreciable contribution to dose --- not evaluated in our Master Table. Part 3. Review of the Probable Underestimates from Chapters 8-25 ================================================================= The major omissions, discussed above, are not the only issue. Even the entries in the Master Table represent a credible LOWER limit of breast-dose from the sources to which they apply. We will just review the probable underestimates incorporated in the Master Table, by chapter. o - Chapter 8, Thymus Irradiation before Age 1. In Part 2, Item 10, we assumed that for every infant who received therapeutic doses, only 2.5 infants received diagnostic doses. And in Item 11, we assumed a very low dose from the use of fluoroscopy in such diagnosis. And in Part 3, Item 12, we assumed that no diagnosis or therapy at all occurred in 1957, 1958, or 1959. o - Chapter 9, Thymus Irradiation to Reduce Sudden Death, Ages 1-9. In Part 5, Item 2, we assumed that only ONE hospital in the entire city of Boston and the entire Suffolk County, did this. In Item 8, we assumed that the dosage was no higher than in Chapter 8, even though it was probably higher. In Item 10, we assumed there was never any use of fluoroscopy in pre-therapy screening, and we assigned no dose at all from diagnostic screening. o - Chapter 10, Pre-Surgical Thymus Irradiation. In Part 4, Item 12, we subtracted about 25 % from the breast-dose in Chapter 8. o - Chapter 12, Pre-Birth Irradiation. In Part 2, we reduced the frequency reported in our source by 25 %. In Part 4, Item 5, we used only the lowest dose in the range reported by our source --- a dose 2-fold lower than the highest level reported. o - Chapter 13, Treatment of Acute Postpartum Mastitis. In Part 3, Item 2, we assumed that one report covered ALL cases treated in Monroe County. o - Chapter 14, Chronic Mastitis. In Part 5, we assumed that there was no ""silent"" use of radiation therapy going on, and also that Dr. Pfahler was the only physician in all of Philadelphia who ever used radiation therapy for chronic mastitis. o - Chapter 15, Fluoroscopic Exposure in Management of Pneumothorax Treatment of Tuberculosis. In Part 3, we assumed 1.5 rads as breast-dose per fluoroscopy, using the Massachusetts studies. Yet we know that the Nova Scotia studies considered 7.5 rads as the proper dose, based upon the direction of the radiation introduction to patients. We chose the lower dose, conservatively, to avoid any possibility of overestimate of radiation dose. o - Chapter 16, Mass-Screening for Tuberculosis. There was some uncertainty of the years of duration of this program. We arbitrarily cut the duration to one-half of the period of concern, thus possibly underestimating the radiation dose. o - Chapter 18, Treatment of Pneumococcal Pneumonia with X-Rays. We assumed that no institution other than Bowman Gray School of Medicine did therapy of pneumonia cases in Forsyth County. This may have led us to a large underestimate of radiation dose. Also, we eliminated all cases other than those with pneumococcus as the etiologic agent, when it is highly likely that pneumonias caused by other organisms were also treated with x-rays. o - Chapter 19, Treatment of Whooping Cough with X-Rays. We assumed that only the Boston Floating Hospital treated pertussis with x-rays, but other hospitals in Boston may have done so. Our assumption may have led to an underestimate of radiation dose (person-rads) to breasts. o - Chapter 20, Treatment of Hyper-Thyroidism and Breast Irradiation. We made the assumption that when radium gamma-rays were used in therapy of this disease, there was no direct gamma-ray exposure to the breasts. The conditions of radium use did not permit any estimate of this source of breast-dose, so for our effort to provide the Credible Lower Limit, we assigned Zero as the dose to the breast from this source. o - Chapter 21, X-Rays in the Management of Adolescent Scoliosis. Even though we suspected a larger radiation dose in 1920-1930 than in the later periods for which records were available, we made the decision simply to assign Zero as the x-ray dose in 1920-1930 to avoid any possible overestimate of dose. o - Chapter 22, Chiropractic Examinations and Breast Irradiation. Some of our colleagues in disability evaluation have indicated to us that both chiropractors and physicians commonly take many, many films of the back in cases of disability claims and in evaluation of ameliorative therapy. We would judge that a share of those films, not taken in hospitals, may not be counted in any total assessment of chiropractic use of x-rays. o - Chapter 23, Major Diagnostic Radiological Contributions to Breast-Tissue Dose. For the fluoroscopic part of several diagnostic procedures, we have assumed an entrance dose of 3 Roentgens per exposure --- probably a great underestimate at a time when some machines were putting out 100 Roentgens per minute and when New York City had to post a limit of 100 Roentgens per fluoroscopic exam. Diagnostic doses from fluoroscopy in office practice are simply not trackable. o - Chapter 24, Occupational Sources of Breast Irradiation. There are reasons to doubt official evaluations of occupational radiation doses in the 1920-1960 period. First, the 1941 Cowie-Scheele Study indicates how the good low-dose behavior by most technicians can be undone by a few high-dose exceptions. Second, the way that radium was really handled in many hospitals was shocking (Chapter 35). Third, much x-ray work was done outside hospitals (Chapters 31-34, 36). o - Omission of person-rads received at ages above 64. This, too, is a source of dose underestimation. Part 4. Our View of the Evidence at Hand ========================================= The evidence already solid is far more than enough to point the path to a very great deal of breast-cancer prevention --- EFFECTIVE cancer prevention. It is not necessary to account for 100% of causation or to await impossible exactitude about the radiation-induced percentage, before initiating preventive ACTION based upon the finding from this study. The estimate that 75% of all breast-cancers have a radiation etiology is based on the Master Table and the numerous instances of our underestimation of doses in the effort to produce a Credible Lower Limit. In view of all the ancillary evidence from fluoroscopy and dermatology and treatment of inflammatory conditions, the issue is hardly one of defending a 75% etiology; rather the question may well be how much HIGHER than 75% this might be. There may be some valuable resources in the literature which have not yet come to our attention. We intend to continue upgrading our estimates in an on-going fashion, and to report new findings as appropriate. We hope and expect that others will inquire into this topic too. # # # # #