 Section 9 of the final report of the Advisory Committee on Human Radiation Experiments. This is a LibriVox recording. All LibriVox recordings are in the public domain. For more information or to volunteer, please visit LibriVox.org. Recording by Stephen Sutleff. Final report of the Advisory Committee on Human Radiation Experiments. Introduction Part 5 Radioisotopes, what are they and how are they made? What are radioisotopes? The isotopes of an element are all the atoms that have in their nucleus the number of protons, atomic number, corresponding to the chemical behavior of that element. However, the isotopes of a single element vary in the number of neutrons in their nuclei. Since they still have the same number of protons, all these isotopes of an element have identical chemical behavior. But since they have different numbers of neutrons, these isotopes of the same element may have different radioactivity. An isotope that is radioactive is called a radioisotope or radionuclide. Two examples may help clarify this. The most stable isotope of uranium, uranium 238, has an atomic number of 92 protons and an atomic weight of 238, 92 protons plus 146 neutrons. The isotope of uranium of greatest importance in atomic bombs, uranium 235, though, has three fewer neutrons. Thus, it also has an atomic number of 92, since the number of protons has not changed, but an atomic weight of 235, 92 protons plus only 143 neutrons. The chemical behavior of uranium 235 is identical to all other forms of uranium, but its nucleus is less stable, giving it higher radioactivity and greater susceptibility to chain reactions that power both atomic bombs and nuclear fission reactors. Another example is iodine, an element essential for health. Insufficient iodine in one's diet can lead to a goiter. Iodine also is one of the earliest elements whose radioisotopes were used in what is now called nuclear medicine. The most common, stable form of iodine has an atomic number of 53 protons and an atomic weight of 127, 53 protons plus 74 neutrons. Because its nucleus has the correct number of neutrons, it is stable and is not radioactive. A less stable form of iodine also has 53 protons. This is what makes it behave chemically as iodine, but four extra neutrons, for a total atomic weight of 131, 53 protons, and 78 neutrons. With too many neutrons in its nucleus, it is unstable and radioactive, with a half-life of eight days. Because it behaves chemically as iodine, it travels throughout the body and localizes in the thyroid gland, just like the stable form of iodine. But because it is radioactive, its presence can be detected. Iodine 131 thus became one of the earliest radioactive tracers. How can different isotopes of an element be produced? How can isotopes be produced, especially radioisotopes, which can serve many useful purposes? There are two basic methods, separation and synthesis. Some isotopes occur in nature. If radioactive, these usually are radioisotopes with very long half-lives. Uranium 235, for example, makes up about 0.7% of the naturally occurring uranium on the Earth. The challenge is to separate this very small amount from the much larger bulk of other forms of uranium. The difficulty is that all these forms of uranium, because they have the same number of electrons, will have identical chemical behavior. They will bind in identical fashion to other atoms. Chemical separation. Developing a chemical reaction that will bind only uranium atoms will separate out uranium atoms, but will not distinguish among different isotopes of uranium. The only difference among the uranium isotopes is their atomic weight. A method had to be developed that would sort atoms according to weight. One initial proposal was to use a centrifuge. The basic idea is simple. Spin the uranium atoms as if they were on a very fast merry-go-round. The heavier ones will drift toward the outside faster and can be drawn off. In practice, the technique was an enormous challenge. The goal was to draw off that very small portion of uranium atoms that were lighter than their brethren. The difficulties were so enormous that the plan was abandoned in 1942. Instead, the technique of gaseous diffusion was developed. Again, the basic idea was very simple. The rate at which gas passed, diffused, through a filter, depended on the weight of the gas molecules. Lighter molecules diffused more quickly. Gas molecules that contained uranium-235 would diffuse slightly faster than gas molecules containing the more common, but also heavier uranium-238. This method also presented formidable technical challenges, but was eventually implemented in the gigantic gas diffusion plant at Oak Ridge, Tennessee. In this process, the uranium was chemically combined with fluorine to form a hexafluoride gas prior to separation by diffusion. This is not a practical method for extracting radioisotopes for scientific and medical use. It was extremely expensive and could only supply naturally occurring isotopes. A more efficient approach is to artificially manufacture radioisotopes. This can be done by firing high-speed particles into the nucleus of an atom. When struck, the nucleus may absorb the particles or become unstable and emit a particle. In either case, the number of particles in the nucleus would be altered, creating an isotope. One source of high-speed particles could be a cyclotron. A cyclotron accelerates particles around a circular racetrack with periodic pushes of an electric field. The particles gather speed with each push, just as a child swings higher with each push on a swing. When traveling fast enough, the particles are directed off the racetrack and into the target. A cyclotron works only with charged particles, however. Another source of bullets are the neutrons already shooting about inside a nuclear reactor. The neutrons normally strike the nucleus of the fuel, making them unstable and causing the nuclei to split, fission, into two large fragments and two or three free neutrons. These free neutrons, in turn, make additional nuclei unstable, causing further fission. The result is a chain reaction. Too many neutrons can lead to an uncontrolled chain reaction, releasing too much heat and perhaps causing a meltdown. Therefore, surplus neutrons are usually absorbed by control rods. However, these surplus neutrons can also be absorbed by targets of carefully selected material placed in the reactor. In this way, the surplus neutrons are used to create radioactive isotopes of the materials placed in the targets. With practice, scientists using both cyclotrons and reactors have learned the proper mix of target atoms and shooting particles to cook up a wide variety of useful radioisotopes. How does radiation affect humans? Radiation may come from either an external source, such as an X-ray machine, or an internal source, such as an injected radioisotope. The impact of radiation on living tissues is complicated by the type of radiation and the variety of tissues. In addition, the effects of radiation are not always easy to separate from other factors, making it a challenge at times for scientists to isolate them. An overview may help explain not only the effects of radiation, but also the motivation for studying them, which led to much of the research examined by the advisory committee. What effect can ionizing radiation have on chemical bonds? The functions of living tissues are carried out by molecules, that is, combinations of different types of atoms united by chemical bonds. Some of these molecules can be quite large. The proper functioning of these molecules depends upon their composition and also their structure, shape. Altering chemical bonds may change composition or structure. Ionizing radiation is powerful enough to do this. For example, a typical ionization releases 6 to 7 times the energy needed to break the chemical bond between two carbon atoms. This ability to disrupt chemical bonds means that ionizing radiation focuses its impact in a very small but crucial area, a bit like a karate master, focusing energy to break a brick. The same amount of raw energy, distributed more broadly in non-ionizing form, would have much less effect. For example, the amount of energy in a lethal dose of ionizing radiation is roughly equal to the amount of thermal energy in a single sip of hot coffee. The crucial difference is that the coffee's energy is broadly distributed in the form of non-ionizing heat, while the radiation's energy is concentrated in a form that can ionize. What is DNA? Of all the molecules in the body, the most crucial is DNA, deoxyribose nucleic acid, the fundamental blueprint of all of the body's structures. The DNA blueprint is encoded in each cell as a long sequence of small molecules linked together into a chain, much like the letters in a telegram. DNA molecules are enormously long chains of atoms wound around proteins and packed into structures called chromosomes within the cell nucleus. When unwound, the DNA in a single human cell would be more than two meters long. It normally exists as 23 pairs of chromosomes packed within the cell nucleus, which itself has a diameter of only 10 micrometers. One ten-thousandth of a meter. Only a small part of this DNA needs to be read at any one time to build a specific molecule. Each cell is continually reading various parts of its own DNA as it constructs fresh molecules to perform a variety of tasks. It is worth remembering that the structure of DNA was not solved until 1953, nine years after the beginning of the period studied by the advisory committee. We now have a much clearer picture of what happens within a cell than did the scientists of 1944. What effect can ionizing radiation have on DNA? Ionizing radiation by definition ionizes, that is, it pushes an electron out of its orbit around an atomic nucleus, causing the formation of electrical charges on atoms or molecules. If this electron comes from the DNA itself or from a neighboring molecule and directly strikes and disrupts the DNA molecule, the effect is called direct action. This initial ionization takes place very quickly in about one trillionth of a second. However, today it is estimated that about two-thirds of the damage caused by x-rays is due to indirect action. This occurs when the liberated electron does not directly strike the DNA, but instead strikes an ordinary water molecule. This ionizes the water molecule, eventually producing what is known as a free radical. A free radical reacts very strongly with other molecules as it seeks to restore a stable configuration of electrons. A free radical may drift about up to ten billion times longer than the time needed for the initial ionization. This is still a very short time, about one-tenth-thousandth of a second, increasing the chance of it disrupting the crucial DNA molecule. This also increases the possibility that other substances could be introduced that would neutralize free radicals before they do damage. Neutrons act quite differently. A fast neutron will bypass orbiting electrons and occasionally crash directly into an atomic nucleus, knocking out large particles such as alpha particles, protons, or larger fragments of the nucleus. The most common collisions are with carbon or oxygen nuclei. The particles created will themselves then set about ionizing nearby electrons. A slow neutron will not have the energy to knock out large particles when it strikes a nucleus. Instead, the neutron and the nucleus will bounce off each other, like billiard balls. In so doing, the neutron will slow down and the nucleus will gain speed. The most common collision is with a hydrogen nucleus, a proton that can excite or ionize electrons in nearby atoms. What immediate effects can ionizing radiation have on living cells? All of these collisions and ionizations take place very quickly in less than a second. It takes much longer for the biological effects to become apparent. If the damage is sufficient to kill the cell, the effect may become noticeable in hours or days. Cell death can be of two types. First, the cell may no longer perform its function due to internal ionization. This requires a dose to the cell of about 100 gray, 10,000 rad. For a definition of gray and rad, see the section below titled, How do we measure the biological effects of radiation? Second, reproductive death, mitotic inhibition. May occur when a cell can no longer reproduce, but still performs its other functions. This requires a dose of 2 gray, 200 rad, which will cause reproductive death in half the cells irradiated. Hence, such a quantity is called a mean lethal dose. Today we still lack enough information to choose among the various models proposed to explain cell death in terms of what happens at the level of atoms and molecules inside a cell. If enough, crucial cells within the body totally cease to function. The effect is fatal. Death may also result if cell reproduction ceases in parts of the body where cells are continuously being replaced at a high rate, such as the blood cell forming tissues in the lining of the intestinal tract. A very high dose of 100 gray, 10,000 rad, to the entire body causes death within 24 to 48 hours. A whole body dose of 2.5 to 5 gray, 250 to 500 rad, may produce death within several weeks. At lower or more localized doses, the effect will not be death, but specific symptoms due to the loss of a large number of cells. These effects were once called non-stochastic. They are now called deterministic. A beta burn is an example of a deterministic effect. What long-term effects can radiation have? The effect of the radiation may not be to kill the cell, but to alter its DNA code in a way that leaves the cell alive but with an error in the DNA blueprint. The effect of this mutation will depend on the nature of the error and when it is read. Since this is a random process, such effects are now called stochastic. Two important stochastic effects of radiation are cancer, which results from mutations in non-Germ cells, termed somatic cells, and heritable changes, which result from mutations in germ cells, eggs, and sperm. How can ionizing radiation cause cancer? Cancer is produced if radiation does not kill the cell It creates an error in the DNA blueprint that contributes to eventual loss of control of cell division, and the cell begins dividing uncontrollably. This effect might not appear for many years. Cancers induced by radiation do not differ from cancers due to other causes, and so there is no simple way to measure the rate of cancer due to radiation. During the period studied by the advisory committee, great effort was devoted to studies of irradiated animals and exposed groups of people to develop better estimates of the risk of cancer due to radiation. This type of research is complicated by a variety of cancers which vary in radio sensitivity. For example, bone marrow is more sensitive than skin cells to radiation-induced cancer. Large doses of radiation to large numbers of people are needed in order to cause measurable increases in the number of cancers and thus determine the differences in the sensitivity of different organs to radiation. Because the cancers can occur any time in the exposed person's lifetime, these studies can take 70 years or more to complete. For example, the largest and scientifically most valuable epidemiologic study of radiation effects has been the ongoing study of the Japanese atomic bomb survivors. Other important studies include studies of large groups exposed to radiation as a consequence of their occupation, such as uranium miners, or as a consequence of medical treatment. These types of studies are discussed in greater detail in the section titled, How do scientists determine the long-term risks for radiation? How can ionizing radiation produce genetic mutations? Radiation may alter the DNA within any cell. Cell damage and death that result from mutations in somatic cells occur only in the organism in which the mutation occurred and are therefore termed somatic or non-heritable effects. Cancer is the most notable long-term somatic effect. In contrast, mutations that occur in germ cells, sperm and ova, can be transmitted to future generations and are therefore called genetic or heritable effects. Genetic effects may not appear until many generations later. The genetic effects of radiation were first demonstrated in fruit flies in the 1920s. Genetic mutation due to radiation does not produce the visible monstrosities of science fiction. It simply produces a greater frequency of the same mutations that occur continuously and spontaneously in nature. Like cancers, the genetic effects of radiation are impossible to distinguish from mutations due to other causes. Today, at least 1,300 diseases are known to be caused by a mutation. Some mutations may be beneficial. Random mutation is the driving force in evolution. During the period studied by the advisory committee, there was considerable debate among the scientific community over both the extent and the consequences of radiation-induced mutations. In contrast to estimates of cancer risk, which are based in part on studies of human populations, estimates of heritable risk are based, for the most part, upon animal studies, plus studies of Japanese survivors of the atomic bombs. The risk of genetic mutation is expressed in terms of the doubling dose, the amount of radiation that would cause additional mutations equal in number to those that already occur naturally from all causes, thereby doubling the naturally occurring rate of mutation. It is generally believed that mutation rates depend linearly on dose and that there is no threshold below which mutation rates would not be increased. Spontaneous mutation, unrelated to radiation, occurs naturally at a rate of approximately 1 in 10,000 to 1 in 1 million cell divisions per gene, with wide variation from one gene to another. Attempts have been made to estimate the contribution of ionizing radiation to human mutation rates by studying the offspring of both exposed and non-exposed Japanese atomic bomb survivors. These estimates are based on comparisons of the rate of various congenital defects and cancer between exposed and non-exposed survivors, as well as on direct counting of mutations at a small number of genes. For all these endpoints, no excess has been observed among descendants of the exposed survivors. Given this lack of direct evidence of any increase in human heritable genetic effects resulting from radiation exposure, the estimates of genetic risks in humans have been compared with experimental data obtained with laboratory animals. However, estimates of human genetic risks vary greatly from animal data. For example, fruit flies have very large chromosomes that appear to be uniquely susceptible to radiation. Humans may be less vulnerable than previously thought. Statistical lower limits on the doubling dose have been calculated that are compatible with the observed human data, based on our inability to demonstrate an effect in humans, the lower limit for the genetic doubling dose is thought to be less than 100 rem. How do we measure the biological effects of external radiation? The methods of measuring radiation and radioactivity, purely physical events, were discussed earlier. In studying the effect of radiation on living organisms, a biological event, the crucial data are the amount of energy absorbed by a specific amount and type of tissue. This requires first measuring the amount of energy left behind by the radiation in the tissue, and second, the amount and type of tissue. What is an absorbed dose of radiation? The risk posed to a human being by any radiation exposure depends partly upon the absorbed dose, the amount of energy absorbed per gram of tissue. Absorbed dose is expressed in rad. A rad is equal to 100 ergs of energy absorbed by one gram of tissue. A more modern, internationally adopted unit is the gray, named for the English medical physicist L.H. Gray. One gray equals 100 rad. Almost all of the documents from the time period studied by the advisory committee use the term rad rather than gray. It is important to realize that absorbed dose refers to energy per gram of absorbing tissue, not total energy. Someone absorbing one gray, 100 rad, in a small amount of tissue, such as a thyroid gland, will absorb much less total energy than someone absorbing one gray, 100 rad, throughout his or her entire body. Thus, when speaking of absorbed dose, it is crucial to know the amount of tissue being exposed, not simply the number of gray or rad. What is an equivalent dose of radiation? Even the rad or gray, though, are still units that measure a purely physical event, the amount of energy left behind in a gram of tissue. It does not directly measure the biological effect of that radiation. The biological effect of the same amount of absorbed energy may vary according to the type of radiation involved. This biological effect can be computed by multiplying the absorbed dose in rad or gray by a number indicating the quality factor of the particular type of radiation. For photons and electrons, the quality factor is defined to be 1. For neutrons, it ranges from 5 to 20 depending on the energy of the neutron. For alpha particles, it is 20. Thus one gray, 100 rad, of alpha particles is currently judged to have an effect on living tissue that is 20 times more than one gray, 100 rads, of x-rays. Multiplying the absorbed dose in rad or gray by the quality factor, also known as the radiation weighting factor, produces what is called the equivalent dose. For the period studied by the advisory committee, this was expressed in terms of a unit called the REM, an acronym for Rankin Equivalent Man. The term equivalent simply means that an absorbed dose expressed in REM would have equivalent biological effects regardless of the type of radiation. Thus 10 REM of x-rays should have the same biological effect as 10 REM of neutrons absorbed by the same part of the body. The modern unit is the sievert, abbreviated SV, and named for the prominent Swedish radiologist Rolf Sievert, which is equal to 100 REM. Thus an equivalent dose of 200 REM would today be expressed as 2 sievert. What is an effective dose of radiation? Finally, the biological effect of radiation depends on the type of tissue being irradiated. As with different types of radiation, a weighting or quality factor is introduced depending on the type of tissue. The more sensitive the tissue is to radiation, the higher the factor. The effective dose is the sum of the equivalent doses of the various types of irradiated tissue, each properly weighted for its sensitivity to radiation. Tissue weighting factors are determined from the relative incidence of cancers in different tissues in the Japanese survivors of the atomic bombs. Calculating the effective dose makes it possible to readily compare different exposures, as illustrated by the accompanying graphs. Title of the graphs, experimental and non-experimental doses. Two plots appear on page 58. They are both titled Thyroid Studies with iodine 131. Each of the two plots shows a stratification. In the upper, a division is made between those receiving largest dose and those receiving smallest dose. While in the lower plot, those receiving largest thyroid dose versus those receiving smallest thyroid dose. Although the two plots differ in that one refers to millirem with thyroid excluded and the other represents rads received by the thyroid, the two plots are qualitatively similar, demonstrating the claim that effective dose makes it possible to readily compare different exposures. Three charts appear on page 59. They are titled Fernald School Nutrition Study Calcium Tracer, Fernald School Nutrition Study Iron Tracer, and Common Medical Procedures. The upper pair concern the Fernald School Nutrition Study. The first about the calcium tracer, the second about the iron tracer. Both display effective dose equivalent in millirems for those receiving the smallest dose, those receiving the largest dose, and a representative Denver resident. For each of these cohorts, effective dose equivalent is shown both within the study group and the annual natural background radiation they were exposed to. The pair of plots demonstrates that in the case of the iron tracer, the effective dose equivalent in the study group is of the same order of magnitude as the annual natural background radiation. In contrast, for the calcium tracer, the effective dose equivalent received by the study group is relatively negligible compared to the annual natural background radiation received by that group. This comparison can be made because the effective dose makes it possible to readily compare different exposures. The third plot on page 59 is titled Common Medical Procedures and shows the whole body effective dose equivalent in millirems for four different medical procedures. Chest x-ray, back x-ray, colon x-ray, and brain scan, showing both the annual natural background and dose received from the procedure. In the case of chest x-ray, the additional dose from the procedure is negligible compared to annual natural background. In the case of back x-ray, it is about a third more. In the case of colon x-ray, it is greater than the amount received in annual natural background. And in the case of brain scan, it is more than twice higher than that received in annual natural background. Although these medical procedures are directed towards very different parts of the anatomy, the comparison can be made because effective dose makes it possible to readily compare different exposures. End of section 9. Section 10 of the final report of the Advisory Committee on Human Radiation Experiments. This is a LibriVox recording. All LibriVox recordings are in the public domain. For more information or to volunteer, please visit LibriVox.org. Recording by Stephen Sutliff. Final report of the Advisory Committee on Human Radiation Experiments, Introduction Part 6. How do we measure the biological effects of internal emitters? The general principles just described acquire further refinement when applied to doses from internal emitters. What information is needed to calculate absorbed dose of a radionuclide inside the body? Calculating the absorbed dose from a radionuclide inside the body is complex, since it involves both the physics of radioactive decay and the biology of the body's metabolism. Six important factors that must be considered are these. 1. The amount of the radionuclide administered. 2. The amount of radiation emitted during the decay process. 3. The physical half-life of the radionuclide. 4. The chemical form of the radionuclide. 5. The fraction of the radionuclide that accumulates in each organ. 6. The length of time that the radionuclide remains in the organ. The biological half-life. How varied are the types of radiation that different radionuclides emit? Radionuclides can emit several types of radiation. For example, gamma rays, beta, or alpha particles. Each radionuclide emits its own unique mixture of radiations. Indeed, scientists identify radioactive materials by using these unique mixtures as if they were fingerprints. The mix of radiations for a specific radionuclide is always the same, regardless of whether the radionuclide is located on a bench in the physicist's laboratory or inside the human body. This means that the type of radiation of each radionuclide can be measured outside the body with great precision by laboratory instruments. A quality factor, discussed earlier, is used to adjust for the difference in the biological effects of different types of radiation. What determines how long a radionuclide will irradiate the body? The combination of the physical and biological half-life, the effective half-life, determines how long a radionuclide will continue to pump out energy into surrounding tissue. If the physical and biological half-lides of a particular chemical form of a radionuclide are very long, the radionuclide will continue to expose an individual to radiation over his or her lifetime. The total lifetime radiation exposure expressed in REM is called the committed dose equivalent. The physical half-life is the length of time it will take for half of the atoms in a sample to decay to a more stable form. The physical half-life of each radionuclide can be measured precisely in the laboratory. A shorter half-life means that the miniature power source will run down sooner. Sometimes, however, a radionuclide will not decay immediately to a stable form but to a second, still unstable form. A full calculation, therefore, must also include the types of radiation and physical half-lives of any decayed products. The biological half-life does not depend on the radionuclide but rather on the chemical form of the radionuclide. One chemical form of the radionuclide might be rapidly eliminated from the body, or as other chemical forms may be slowly eliminated. To measure the biological half-life of a particular chemical form of a radionuclide, that chemical form needs to be studied in animals. Since the biological processes of different animals vary considerably, an accurate determination of the biological half-life requires that each chemical form of the radionuclide be studied in each animal of interest. Prior to studying a chemical form of a radionuclide in a human being, animal studies will perform to get some idea of what to expect. Once the results of animal studies are available, scientists are able to predict what amount of that chemical form of the radionuclide can be safely injected into humans. An accurate determination of what fraction of each chemical form of the radionuclide accumulates in each organ, and how long it stays in each organ in humans can only be determined by studying humans. These type of studies are called biodistribution studies. What is the tissue-weighting factor? Some chemical forms of radionuclides are highly concentrated in one small organ, for example iodine in the thyroid gland. When this happens, that organ will absorb most of the radiated energy, and little energy will be deposited in the remainder of the body. Thus, for each chemical form of a radionuclide, there is an organ that will receive the highest dose from that radionuclide. Since organs also vary greatly in their sensitivity to radiation, the biological consequences of the radiation dose differ depending on the organ. This difference in sensitivity to radiation is represented by what is called a tissue-weighting factor. What is the difference between committed equivalent dose and committed effective dose? An estimate of the risk posed by a radionuclide in the body depends on its chemical form, its biodistribution, its physical properties, how to case, and the sensitivity of the organs exposed. When all these factors are considered in the calculation of risk for a single radionuclide, the total lifetime exposure is called the committed equivalent dose. If more than one radioisotope is present, the sum of all the committed equivalent doses is called the committed effective dose. Both are expressed in REM or the more modern units, sieverts. These calculations provide a basis for comparing the risk posed by different isotopes. How do radiation risks compare with chemical risks? It should be noted that radiation is not the only possible hazard resulting from the medical use of radionuclides. Few radioisotopes, whether intentionally or accidentally introduced into the body, enter in a chemically pure form. The radioactive atoms are usually part of a larger chemical compound. The chemical form of the radioisotope may pose its own hazards of chemical toxicity. Chemical toxicity depends on the chemical effect of the compound on the body, quite independent of any effects of radiation. Determining chemical toxicity is an entire field of science of its own. How do scientists determine the long-term risks from radiation? Where did the risk estimates in this report come from? Throughout their support, the reader will find numerous statements estimating the risks of cancer and other outcomes to individuals exposed to various types of radiation. These estimates were obtained from various scientific advisory committees that have considered these questions in depth. Their estimates, in turn, are based on syntheses of the scientific data on observed effects in humans and animals. How are risk estimates expressed? Epidemiologists usually express the risk of disease in terms of the number of new cases, incident rate, or deaths, mortality rate, in a population in some period of time. For example, an incident rate might be 100 new cases per 100,000 people per year. A mortality rate might be 15 deaths per 100,000 people per year. These rates vary widely by age, conditions of exposure, and various other factors. To summarize this complex set of rates, government regulatory bodies often consider the lifetime risk of a particular outcome, like cancer. When relating a disease, such as cancer, to one of its several causes, a more usable concept is the excess lifetime risk expected from one particular pattern of exposure, such as continuous exposure to one rad per year. It is well established that cancer rates begin to rise above the normal background rate only some time after exposure, the latent period, which varies with the type of cancer and other factors, such as age. Even after the latent period has passed and radiation effects begin to appear, not all effects are due to radiation. The excess rate may still vary by age, latency, or other factors. But for many cancers, it tends to be roughly proportional to the rate in the general population. This is known as the constant relative risk model. And the ratio of rates at any given age between exposed and unexposed groups is called the relative risk. Many advisory committees have based their risk estimates on models of the relative risk as a function of dose and perhaps other factors. Other committees, however, have based their estimates on the difference in rates between exposed and unexposed groups, a quantity known as the absolute risk. This quantity also varies with dose and other factors. But when this variation is appropriately accounted for, either approach can be used to estimate lifetime risk. What are the types of data on which such estimates are based? Human data are one important source discussed below. Two other important sources of scientific data are experiments on animals and on cell cultures. Because both types of research are done in laboratories, scientists can carefully control the conditions and many of the variables. For the same reason, the experiment can be repeated to confirm the results. Such research has contributed in important ways to our understanding of basic radiological principles. It also has provided quantitative estimates of such parameters as the relative effectiveness of different types of radiation and the effects of dose and dose rate. In some circumstances where human data are limited or non-existent, such laboratory studies may provide the only basis on which risks can be estimated. Why are human data preferable to data on animals or tissue cultures for most purposes? Most scientists prefer to base risk estimates for humans on human data wherever possible. This is because in order to apply animal or tissue culture data to humans, scientists must extrapolate from one species to another, or from simple cellular systems to the complexities of human physiology. This requires adjusting the data for differences among species in lifespan, body size, metabolic rates, and other characteristics. Without actual human data, extrapolation provides no guarantee that there are no unknown factors also at work. It is not surprising that there is no clear consensus as to how to extrapolate risk estimates from one species to another. This problem is not unique to radiation effects. There are countless examples of chemicals having very different effects in different species, and humans can differ quite significantly from animals in their reaction to toxic agents. How have human data been obtained? There are serious ethical issues with conducting experiments on humans as discussed elsewhere in the report. However, most of the human data that are used to estimate risks, not just risks from radiation, come from epidemiologic studies on populations that already have been exposed in various ways. For radiation effects, the most important human data come from studies of the Japanese atomic bomb survivors carried out by the Radiation Effects Research Foundation, formerly the Atomic Bomb Casualty Commission in Hiroshima. Other valuable sources of data include various groups of medically exposed patients, such as radiotherapy patients, and occupational exposed workers, such as the uranium miners discussed in Chapter 12. Why is it necessary to compare exposed populations with unexposed populations? Unlike a disease caused by identifiable bacteria, no signature has yet been found in cancerous tissue that would link it definitively to prior radiation exposure. Radiogenic cancers are identical in properties such as appearance under a microscope, growth rate, and potential to metastasize to cancers occurring in the general population. Finding cancers in an exposed population is not enough to prove that they are due to radiation. The same number of cancers might have occurred due to the natural frequency of the disease. The challenge is to separate out the effects of radiation from what would otherwise have occurred. A major step in this direction is to develop follow-up or cohort studies in which an exposed group has followed over time to observe their disease rates, and these rates are then compared with the rates for the general population or an unexposed control group. Why is the analysis of epidemiologic data so complicated? Simply collecting data on disease rates in exposed and controlled populations is not enough. Indeed, causal analysis may lead to serious errors in understanding. Sophisticated data collection techniques and mathematical models are needed to develop useful risk estimates for several reasons. One, random variation due to sample size. Two, multiple variables. Three, limited lifespan of most studies. And four, problems of extrapolation. In addition, individual studies may also be biased in their design or implementation. What is random variation? The observed portion of subjects developing a disease in any randomly selected subgroup, sample, of individuals with similar exposures is subject to the vagaries of random variation. A simple-minded example of this is the classic puzzle of determining in a drawer of 100 socks how many are white and how many are black by pulling out one sock at a time. Obviously, if we pull out all of the socks, we know for certain. In most areas of study, though, pulling out all the socks is far too expensive and time-consuming. But if we pull only 10, with what degree of confidence can we predict the color of the others? If we pull 20, we will have more confidence. In other words, the larger the sample, the greater our confidence. Using statistical techniques, our degree of confidence can be calculated from the size of the entire population, in this case 100 socks, and the size of the actual sample. The result is popularly called the margin of error. The most common examples of this in everyday life are the public opinion polls continually quoted in the news media. As can be seen in the simple example of the drawer of socks, the highest degree of confidence can be achieved simply by pulling all the socks out of the drawer. For public opinion polls, this will be far too expensive. Instead, a small sample is selected at random from the population. Nowadays, it is common to report not only the actual results, but also the sample size and the margin of error. The margin of error depends not only on the sample size, but also on how high a degree of confidence we desire. The degree of confidence is the probability that our sample has provided a true picture of the entire population. For example, the margin of error will be smaller for 80% degree of confidence than for 95%. Even where a study covers an entire exposed population, such as the atomic bomb survivors, the issue of random variation remains when we wish to generalize the findings to other populations. What are multiple variables? The effects of radiation will depend upon or vary with the dose of radiation received. However, these effects also may vary with other factors, other variables that are not dependent upon the radiation dose itself. Examples of such variables are age, gender, latency, time-since exposure, and smoking. Data on these other variables must be collected as well as data on the basic elements of radiation dose and disease. The challenge is to then distinguish between disease rates due to radiation and those due to other factors. For example, if the populations studied were all heavy smokers, this might explain in part a higher rate of lung cancer. Much of the science of epidemiology is devoted to choosing what factors to collect data on and then developing the multivariate mathematical models needed to separate out the effect of each variable. Radiation effects vary considerably across subgroups and over time or age. Because of this, direct estimates of risk for particular subgroups would be very unstable. Mathematical models must be used. These models allow all the data to be used to develop risk estimates that, while based on sufficiently large estimates to be stable, will be applicable to particular subgroups. A more subtle problem is mis-specification of the model finally chosen to calculate risks. The model may weigh selected factors in a manner that best fits the data from a statistical viewpoint. This model, while fitting the data, may not actually be a correct view of nature. Another model that does not fit the data quite as well may actually better describe the, as yet unknown, underlying mechanisms. Why does a limited time span reduce the value of a study? The most pronounced effects of large exposures to radiation manifest themselves quickly in symptoms loosely termed radiation sickness. However, another concern is understanding the effects of much lower levels of radiation. Unlike the more acute effects of large exposures, these may not appear for some time. Some cancers, for example, do not appear until many years after the initial exposure. These latent effects may continue to appear in a population throughout their entire lifetimes. Calculating the lifetime risk of an exposure requires following the entire sample until all of its members have died. Thus far, none of the exposed populations have yet been followed to the ends of their lives. Although the Radium Dial Painter Study for the group painting before 1930 essentially has been completed and the follow-up has been closed out. Why does extrapolation among human populations pose problems? As discussed earlier, extrapolating results from one species to another is problematic due to differences in how species respond to radiation. Even though humans are all members of the same species, there are similar problems when extrapolating results from one group of humans to another group. Within the human species, different groups can have different rates of disease. For example, stomach cancer is much more common and breast cancer much rarer among Japanese than among U.S. residents. How then should estimates of the radiation-induced excess of cancer among the atomic bomb survivors be applied to the U.S. population? Assumptions are needed to transport risk estimates from one human population to another human population that may have very different normal risks. Why does extrapolation from high to low doses pose problems? Acquiring high quality human data on low dose exposure is difficult. Past studies indicate that the effects of low doses are small enough to be lost in the noise of random variation. In other words, the random variation due to sample size may be greater than the effects of the radiation. Thus, to estimate the risk of low doses, it is necessary to extrapolate from the effects of high doses down to the lower range of interest. As with extrapolation among species or among human populations, assumptions must be made. The basic assumption concerns the dose effect. Is the effect of a dose linear? This would mean that half of the dose would produce half of the effect. One tenth of the dose would produce one tenth of the effect, and so on. Nature is not so reliable, however. There are many instances in nature of non-linear relationships. A non-linear dose effect, for example, could mean that half of the dose would produce 75% of the effects, or going in the other direction. A non-linear dose effect could mean that half of the dose would produce only 10% of the effect. Reliable data are too sparse to settle the issue empirically. Much of the ongoing controversy over low dose effects concerns which dose effect relationship to assume. Concerning dose response, most radiation advisory committees assume that radiation risks are linear in doses at low levels. Although these risks may involve non-linear terms at higher doses. Another assumption concerns the effect of dose rate. It is generally agreed that the effect of high dose x-rays is reduced if the radiation is received over a period of time instead of all at once. This reduction in acute effects due to the cell's ability to repair itself between exposures is one of the reasons that modern protocols for radiotherapy use several fractionated doses. The degree to which this also happens at low doses is less clear. There are few human data on the effect of dose rate on cancer induction. Most estimates of the effect come from animal or cell culture experiments. There is also evidence of quite different dose rate effects for alpha radiation and neutrons. How can a specific study be biased? When applied to an epidemiologic study, the term bias does not refer to the personal beliefs of the investigators, but to aspects of the study design and implementation. There are several possible sources of bias in any study. What is called a confounding bias may result if factors other than radiation have affected disease rates. Such factors, as mentioned earlier, might be a rate of smoking higher than the general population. A selection bias may result if the sample was not truly a random selection from the population under study. For example, the results of a study that includes only employed subjects might not be applicable to the general population, since employed people, as a group, are healthier than the entire population. An information bias may result from unreliability in a source of basic data. For example, basing the amount of exposure on the memory of subjects may bias the study, since sick people may recall differently than healthy people. Dose, in particular, can be difficult to determine when studies are conducted on populations exposed prior to the study, since there usually was no accurate measurement at the time of exposure. Sometimes, when dose measurements were taken, as in the case of the atomic veterans, the data are not adequate by today's standards. Finally, any study is subject to the random variation discussed earlier, which depends on how large the sample is. This is far more important for low dose than high dose studies, since the low dose effects themselves are small enough to be lost amid random variations if the sample is too small. To summarize, multiple studies may produce somewhat different results because there is an actual difference in the response between populations, or because studies contain spurious results due to their own inadequacies. In addition, it must be recognized that the entire body of scientific literature is itself subject to a form of bias known as publication bias, meaning an over-reporting of findings of excess risk. This is because studies that demonstrate an excess risk may be more likely to be published than those that do not. In view of all these uncertainties, what risk estimates did the committee choose? Despite all these uncertainties, it must be pointed out that more is known about the effects of ionizing radiation than any other carcinogen. The Beer V committee of the National Academy of Sciences estimated in 1990 that the lifetime risk from a single exposure of 10 rem of whole body external radiation was about 8 excess cancers of any type per 1,000 people. This number is actually an average over all possible ages at which an individual might be exposed, rated by population and age distribution. For continuous exposure to 0.1 rem per year throughout a lifetime, the corresponding estimate was 5.6 excess cancers, that is, over and above the rate expected in a similar but unexposed population per 1,000 people. It is widely agreed that for x-rays and gamma rays this latter figure should be reduced by some factor to allow for a cell's ability to repair DNA, but there is considerable uncertainty as to what figure to use. A figure of about 2 or 3 is often suggested. The estimates of lifetime risk from the Beer V report have a range of uncertainty due to random variation of about 1.4 fold. The additional uncertainties due to the factors discussed earlier are likely to be larger than the random variation. In comparison, for most chemical carcinogens, the uncertainties are often a factor of 10 or more. This agreement among studies of radiation effects is quite remarkable and reflects the enormous amount of scientific research that has been devoted to the subject, as well as the large number of people who have been exposed to doses large enough to show effects. End of section 10. Section 11 of Final Report of the Advisory Committee on Human Radiation Experiments. This is a LibriVox recording. All LibriVox recordings are in the public domain. For more information or to volunteer, please visit LibriVox.org. Final Report of the Advisory Committee on Human Radiation Experiments. Ethics of Human Subjects Research. A Historical Perspective. Chapter 1. Part 1. Part 1. Ethics of Human Subjects Research. A Historical Perspective. Part 1. Overview. When the Advisory Committee began work in April 1994, we were charged with determining whether the radiation experiments designed in administration adequately met the ethical and scientific standards, including standards of informed consent that prevailed at the time of the experiments and that exist today, and also to determine the ethical and scientific standards and criteria by which it shall evaluate human radiation experiments. Although this charge seems straightforward, it is in fact difficult to determine what the appropriate standards should be for evaluating the conduct and policies of 30 or 50 years ago. First, we needed to determine the extent to which the standards of that time are similar to the standards of today. To the extent that there were differences, we needed to determine the relative roles of each in making moral evaluations. In Chapter 1, we report what we have been able to reconstruct about government rules and policies in the 1940s and 1950s regarding human experiments. We focus primarily on the Atomic Energy Commission and the Department of Defense, because their history, with respect to human subjects research policy, is less well known than that of the Department of Health, Education, and Welfare, now the Department of Health and Human Services. Drawing on records that were previously obscure or only recently declassified, we reveal the perhaps surprising finding that officials and experts in the highest reaches of the AEC and DOD discussed requirements for human experiments in the first years of the Cold War. We also briefly discussed the research policies of DHEW and the Veterans Administration during these years. In Chapter 2, we turn from a consideration of governmental standards to an exploration of the norms and practices of physicians and medical scientists who conducted research with human subjects during this period. We include here an analysis of the significance of the Nuremberg Code, which arose out of the International War Crimes Trial of German Physicians in 1947. Using the results of our Ethics Oral History Project and other sources, we also examine how scientists of the time viewed their moral responsibility to human subjects, as well as how this translated into the manner in which they conducted their research. Of particular interest are the differences in professional norms and practices between research in which patients are used as subjects and research involving so-called healthy volunteers. In Chapter 3, we return to the question of government standards, focusing now on the 1960s and 1970s. In the first part of this chapter, we review the well-documented developments that influenced and led up to two landmark events in the history of government policy on research involving human subjects. The promulgation by DHEW of comprehensive regulations for oversight of human subjects research and passage by Congress of the National Research Act. In the latter part of the chapter, we review developments and policies governing human research in agencies other than DHEW, a history that has received comparatively little scholarly attention. We also discuss scandals in human research conducted by the DOD and the CIA that came to light in the 1970s and that influenced subsequent agency policies. With the historical context established in chapters 1 through 3, we turn in chapter 4 to the core of our charge. Here we put forward and defend three kinds of ethical standards for evaluating human radiation experiments conducted from 1944 to 1974. We embed these standards in a moral framework intended to clarify and facilitate the difficult task of making judgments about the past. 1. Government Standards for Human Experiments, the 1940s and 1950s When the advisory committee began its work, a central task was the reconstruction of the federal government's rules and policies on human experiments from 1944 through 1974. The history of research rules at the Department of Health, Education and Welfare, DHEW, was well known, at least from 1953 on, when DHEW's National Institutes of Health, NIH, adopted a policy on human subjects research for its newly opened research hospital, the Clinical Center. In the 1960s, the DHEW and some other executive branch agencies undertook regulation of research involving human subjects. These were early steps of a process that culminated, in 1991, in the comprehensive federal policy known as the Common Rule. The historical background of this process, including a well publicized series of incidents and scandals that motivated it, was also widely known and much discussed. 3. By contrast to DHEW, much less was known about the history of research rules for other agencies also involved in research with human subjects during this period, including the Department of Defense, DOD, the Atomic Energy Commission, AEC, and the Veterans Administration, VA. From the perspective of the charge to the advisory committee, these agencies were at least as important as DHEW. It was known that in 1953, the Secretary of Defense issued, in top secret, a memorandum on human standards based on the Nuremberg Code. In 1947, an international tribunal had declared the Nuremberg Code the standard by which a group of doctors in Nazi Germany should be judged for their horrific wartime experiments on concentration camp inmates. However, the actual impact of the Nuremberg Code on the biomedical community in the United States, both inside and outside of government, is a matter of some disagreement. C. CHAPTER II The general view was that, despite some developments in the 1940s and 1950s, there was little activity within the federal government on issues of human subjects research before the 1960s. But while scholars have known of the 1953 Secretary of Defense memorandum, which was declassified in 1975, other relevant Department of Defense documents remained classified or had lain buried in archives. Moreover, relevant records of the Atomic Energy Commission were largely unexplored, and in some cases still classified. These records are important because, from its creation in 1947, the AEC distributed radioisotopes that would be used in thousands of human radiation experiments, and it was a funding source for many other experiments. C. INTERDUCTION Along with the DOD, also created in 1947, the AEC was searching for biomedical information needed to understand the effects of radiation as it prepared for the possibility of atomic warfare. Although the AEC was thus the catalyst for a considerable amount of human experimentation after World War II, there has been literally no scholarship on the AEC's position on the use of human beings in radiation-related research. Now that the previously obscure, even classified, records are being made public, it appears that in the first years of the Cold War, officials and experts in the AEC and DOD did discuss the requirements for human experiments. In this chapter, we tell what we have learned about those discussions. We begin by telling the story of the AEC General Manager's early declarations on human research, which included a requirement that consent be obtained from patient subjects. This story requires a careful look at a series of letters and memorandums exchanged in the late 1940s. Together, these documents paint a clearly important, but nonetheless confusing, picture of a new agency's attempts to come to grips with the complexities of human experimentation. We consider not only what these documents say, but what we can piece together about what they meant in the context of the times. Central questions include the precise scope of the activities covered by the requirements, and whether and how these 1947 statements were communicated and put into effect in the AEC's burgeoning contract research and radioisotope distribution programs. We next turn to the Department of Defense, where we trace the history of rules on the use of healthy, normal volunteer subjects in military research from the time of Walter Reed through the Secretary of Defense's 1953 memorandum and beyond. This memorandum is the earliest known instance in which a federal agency that sponsored human experiments adopted the Nuremberg Code. What is known about how the memorandum was interpreted and implemented by the military establishment takes up much of the rest of this chapter. Here, as in the case of the AEC, key questions concern the scope of the activities covered by requirements and the extent to which they were put into effect. Finally, we discuss how research involving human subjects was addressed at the National Institutes of Health and the Veterans Administration in the 1950s. The evolution of policies governing human research at DHEW has been well documented and is only summarized here. We now know that NIH's 1953 policy was not the earliest federal requirement that consent be obtained from patients as well as healthy subjects. However, in contrast with the 1940s declarations by the AEC, it was a far more visible statement issued by an agency that was emerging as the leading sponsor of human subjects research. In contrast with what is known about NIH, the extent to which there were research rules at the VA in the 1940s and 1950s remains unclear. A recurring theme in this chapter is the uncertainty about the significance within government agencies of many of the official statements that are discussed. While these statements emanated from high and responsible officials and committees, often they cannot be linked to fuller expressions of commitment by the agencies. Some of these statements were not widely disseminated, and there were no implementing guidelines or regulations and no sanctions for failures to abide by them. Thus, it is sometimes unclear what formal, legal significance these statements had. We are no less interested, however, in what these statements can tell us about how government officials and advisors saw human research at the time and how they understood the obligations surrounding it. The Atomic Energy Commission. A requirement for consent is declared at the creation. Even before the AEC came into existence on January 1, 1947, Manhattan Project researchers and officials had begun to lay the groundwork for the expansion of the government's support of biomedical radiation research conducted under federal contract. By the time the AEC began operations, the parallel program to distribute federally produced radioisotopes to research institutions throughout the country was already well underway. The planning for these undertakings required both reflection on high-level matters of policy and attention to matters of small but critical legal and bureaucratic detail. Both legal rules and administrative processes were uncharted. For example, who would be responsible if things went awry and subjects were injured? When could the government tell private doctors or researchers how to conduct treatment or research? The need for rules seemed obvious, but the particular rules that would be arrived at were not. In April 1947, and again in November, Carol Wilson, the general manager of the new agency, wrote letters first to Stafford Warren and then to Robert Stone, both of whom played prominent roles in Manhattan Project medical research. Warren is medical director, and Stone is a key member of the Chicago branch of the project. In these letters, Wilson maintained that clinical testing with patients could go forward only where there was a prospect that the patient could benefit medically, and only after that patient had been informed about the testing and there was documentation that the patient had consented. What was the origin of this position, and what was its reach? It appears that these letters were the products of an agency that was not only seeking to devise rules for new programs, but also was trying to glean lessons from the experience with the secret research that had been conducted during the Manhattan Project. In the course of setting rules for the future, the AEC in its research community had to confront whether and how to proceed with human experimentation in the face of human experiments, including plutonium injections conducted under the auspices of the Manhattan Project, experiments that were conducted in secret and that had the potential for both negative public reaction and litigation. The First Wilson Letter General Manager Wilson's first 1947 letter on human research dated April 30th was, at least in part, a straightforward effort to define the rules according to which the AEC would provide contractors with research funding. The need for such rules had been discussed by the AEC's Interim Medical Advisory Committee, chaired by Stafford Warren, in January 1947, when it met to consider whether clinical testing should be part of the AEC contract research program. The report of the meeting records projects involving human subjects at the University of Rochester and the University of California at Berkeley, and perhaps others. In a January 30 letter to General Manager Wilson, Stafford Warren reported the committee's conclusion that, in the study of health hazards in the use of fissionable and radioactive materials, final investigations by clinical testing of these materials would be needed. Warren therefore requested that the AEC legal department determine the financial and legal responsibility of the AEC when such clinical investigations were carried out under AEC approved and financed programs. The term experiment was not used, and the precise meaning of clinical testing is not clear. A month later, in early March, Warren met with Major Birchard M. Brundage, Chief of the AEC's Medical Division, and two AEC lawyers, to consider the terms for the resumption of clinical testing. In a memorandum for the record, the lawyers summarized the meeting. In the case of clinical testing, the lawyers expressed the view that it was most important that it be susceptible of proof that any individual patient, prior to treatment, was in an understanding state of mind, and that the nature of the treatment, and possible risk involved, be explained very clearly, and that the patient expresses willingness to receive the treatment. Initially, the lawyers had proposed that researchers obtain a written release from patients. However, on Dr. Warren's recommendation, the lawyers agreed that it would be sufficient if at least two doctors certifying writing to the patient's state of mind to the explanation furnished him, and to the acceptance of the treatment. In his April 30 letter to Stafford Warren, Wilson announced that the AEC had approved Warren's committee's recommendations for a program for obtaining medical data of interest to the commission in the course of treatment of patients, which may involve clinical testing. Wilson's letter spelled out ground rules that were agreed upon. The commission understood that treatment, which may involve clinical testing, will be administered to a patient only when there is expectation that it may have therapeutic effect. In addition, the commission adopted the requirement for documentation of consent agreed upon in Warren's meeting with the lawyers. It should be susceptible of proof from official records that, prior to treatment, each individual patient, being in an understanding state of mind, was clearly informed of the nature of the treatment and its possible effects, and expressed his willingness to receive the treatment. The commission deferred to Warren's request that written releases from the patient not be required. However, it does request that in every case at least two doctors should certify in writing made part of an official record to the patient's understanding state of mind to the explanation furnished him and to his willingness to accept the treatment. Carol Wilson's April letter was sent to Stafford Warren as head of the Interim Medical Advisory Committee, which was responsible for advising the AEC on its contract research program, and forwarded to Major Brundage at the Oak Ridge office. Stafford Warren was at this point Dean of the Medical School at the University of California at Los Angeles, one of the dozen research institutions involved in the AEC contract research program. With one exception, the advisory committee on human radiation experiments did not locate documentation that the letter or its contents were communicated to any other research institutions involved with the AEC's contract research program. The exception is the University of California at San Francisco, where there is indirect evidence that someone at that institution had been apprised of Wilson's April letter. Of the 18 plutonium injections only the last one, that involving Elmer Allen or Cal III, took place after the April letter. In Mr. Allen's medical chart there is a notation signed by two physicians indicating that the experimental nature of the procedure was explained and that the patient agreed. Although the noted Mr. Allen's chart suggests an effort on the part of the researchers to comply with Wilson's April letter, the researchers did not comply with the other provision of the Wilson letter, that treatment, which may involve clinical testing, will be administered to a patient only when there is expectation that it may have therapeutic effect. As is discussed in more detail in Chapter 5, there was no expectation at the time that Mr. Allen would benefit medically from an injection of plutonium. Final Report of the Advisory Committee on Human Radiation Experiments Ethics of Human Subjects Research A Historical Perspective Chapter 1 Part 2 The Second Wilson Letter The context of the second Wilson letter as well as its precise terms further indicates that the April 1947 letter was given little distribution and effect. In the fall of 1947 the AEC laboratory at Oak Ridge requested advice from Carol Wilson's office on the rules for experiments involving human subjects. Just as the AEC's Washington headquarters had embarked on the funding of a new research program, Oak Ridge was also in the midst of considering the rules governing the expansion of its own medical research program and the distribution of isotopes, which was then headquartered at Oak Ridge. In September 1947 the manager of Oak Ridge operations wrote to Wilson, asking, What responsibilities does the AEC bear for human administration of isotopes? A. By private physicians and medical institutions outside the project and B. By physicians within the project. What are the criteria for future human use? Two weeks later Oak Ridge sent a memorandum to the Advisory Committee for Biology and Medicine, ACBM. The ACBM had succeeded both Stafford Warren's Interim Medical Advisory Committee and the Medical Board of Review, a group appointed by AEC chairman David Lilienthal to review the AEC's medical program. The memorandum emphasized the need for medical legal criteria for future human tracer research because some of that research would be of no immediate therapeutic value to the patient. The memorandum outlined the pros and cons of tracer studies. Pro. One. Tracer research is fundamental to toxicity studies. Two. The adequacy of the health protection, which we afford our present employees, may in a large measure depend upon information obtained using tracer techniques. Three. New and improved medical applications can only be developed through careful experimentation and clinical trial. Four. Tracer techniques are inherent in the radioisotope distribution program. Con. One. Moral, ethical, and medical legal objections to the administration of radioactive material without the patient's knowledge or consent. Two. There is perhaps a greater responsibility if a federal agency condones human guinea pig experimentation. Three. Publication of such researches in some instances will compromise the best interests of the Atomic Energy Commission. Four. Publication of experiments done by Atomic Energy Commission contractors personnel may frequently be the source of litigation and be prejudicial to the proper functioning of the Atomic Energy Commission insurance branch. The questions raised by Oak Ridge were discussed by the ACBM at its October 11, 1947 meeting, which decided to give the matter more study. The minutes of the October 11 meeting record that human experimentation was then discussed in the context of a request by Dr. Robert Stone to release classified papers containing certain information on human experimentation with radioisotopes conducted within the AEC research program. The request was part of a continuing effort by Stone and other scientists to obtain permission to publish the research, including the plutonium experiments that they had conducted in secret during the Manhattan Project. Earlier in 1947 the AEC had reversed a decision to declassify a report on the plutonium injections, citing the potential for public embarrassment and legal liability. C. CHAPTER V. The question of what to do with these requests continued to fester. The minutes explained that the problem raised by Stone had been dealt with by Chairman Lelyenthal's Medical Board of Review in June. In a cryptic statement, the minutes record the ACBM's agreement that papers on human experiments should remain classified unless the stipulated conditions laid down by the Board of Review were complied with. The stipulated conditions referred to are contained in General Manager Wilson's November 5, 1947 letter to Stone. According to Wilson's letter, at a June meeting the Medical Board of Review concluded that the matter of human experimentation would remain classified where certain conditions were not satisfied. Wilson then quoted from the preliminary, unpublished and restricted draft of the Medical Board report read to the commissioners as follows. The atmosphere of secrecy and suppression makes one aspect of the medical work of the commission especially vulnerable to criticism. We therefore wish to record our approval of the position taken by the medical staff of the AEC in point of their studies of the substances dangerous to human life. We, the Medical Board of Review, believe that no substances known to be or suspected of being poisonous or harmful should be given to human beings unless all of the following conditions are fully met. A. That a reasonable hope exists that the administration of such a substance will improve the condition of the patient. B. That the patient give his complete and informed consent in writing. And C. That the responsible next of kin give in writing a similarly complete and informed consent revocable at any time during the course of such treatment. In other words, the opinion of the Medical Board of Review was presented by Wilson in his November letter as both a prescription for the future conduct of human experiments and a presentation of the criteria that must be met for the declassification of past research. Wilson again referenced these conditions in a letter to ACBM Chairman Alan Gregg, also on November 5. I am sure, Wilson wrote Gregg, that this information will assist Dr. Stone in evaluating the present problem and inform him as to the conditions that must be met in future experiments. Thus, as discussed in more detail in chapters 5 and 13, the requirement that research proceed only with consent appears to have been coupled with the decision to withhold from the public information about experiments that failed to meet that standard. Two points should be made about the term informed consent, which appears in the November letter from Wilson to Stone. First, it is not clear what meaning Wilson and the members of the Medical Board of Review attributed to the term. No further explanation was given. Second, it is nevertheless a matter of some historical interest that this term is used at all. Previous scholarship had attributed its first official usage to a landmark legal opinion in a medical malpractice case that was issued a decade later. The April and November 1947 Wilson letters have some common elements in spite of their differences in detail. They both provided that research with humans proceed, one, only where there is reasonable hope of therapeutic effect, and two, with documentary proof that the patient subject was informed of the treatment and its possible effects and had consented to its administration. But there are many remaining mysteries about the AEC's 1947 statements. In interviews with advisory committee staff, Joseph Volpe, who served as an AEC attorney in its early days and became general counsel in 1949, explained that a letter authored by General Manager Wilson could state AEC policy and confidently recollected that informed consent from research subjects would have been required by the first AEC general counsel. This requirement, Volpe maintained, should be reflected in the commission's minutes. However, Committee and DOE review of the commission's minutes did not reveal evidence that the consent policy was expressly addressed. Even more troubling is that both Wilson letters precluded research that did not offer patient subjects a prospect of direct medical benefit. In the context of the concern about the plutonium injections and other non-therapeutic research conducted during the Manhattan Project experiments, this provision readily makes sense. Yet, as Oak Ridge's inquiry to Washington noted, non-therapeutic research in the form of tracer studies had been, and would continue to be, a mainstay of AEC-sponsored isotope research. How could it be that the Wilson letters were intended to ban exactly the kind of research that, at the same time, the AEC was so actively promoting? It is conceivable that the requirement of the isotope distribution program for risk review prior to the human use of radioisotopes was a means of addressing this notion. However, if the equation between that risk review procedure and the provision in the November Wilson letter seems implicit, the documentary evidence does not provide an express link between the requirements stated in the Wilson letter and the rules of the isotope distribution program. From statements to policy. A failure of translation. Despite the fact that they were developed in response to a need for clarity in the way that human research should be conducted, we have found little evidence of efforts to communicate or implement the rules stated by Wilson in coordination with the AEC's biomedical advisory groups and other AEC officials. In some cases, the evidence described in the following paragraphs suggests that policies for consent from subjects were established and implemented, while in other cases it suggests that, if there were any such policies, they were unknown or lost. Taken together, however, this evidence further supports the view that the ideas present in General Manager Wilson's 1947 statements were available to those working in the field during this time, albeit perhaps in a primitive form. Consider, for example, a 1951 exchange between the AEC's Division of Biology and Medicine, DBM, which directed the AEC's medical research program, and the Commission's Los Alamos Laboratory, which was in routine contact with Washington. An information officer at Los Alamos, Leslie Redman, who was charged to review papers that involve human experimentation, asked the DBM for a definite AEC policy on human experimentation. In the course of his work, Redman wrote, he had been advised by various persons at Los Alamos that regulations or policies of the AEC on human experimentation were available, but he had been unable to locate more than general information about these regulations. According to his letter, his understanding was that these regulations are comparable to those of the American Medical Association, that an experiment be performed under the supervision of an MD, with the permission of the patient, and for the purpose of seeking a cure. Redman's characterization of the American Medical Association's guidelines, as we shall see in Chapter 2, is partly incorrect. The requirement of a therapeutic intent is absent from the AMA guidelines. The possibility of direct therapeutic benefit for the patient was, however, a condition of research, according to both of General Manager Wilson's 1947 letters. Shields Warren, the DBM chief, responded to Redman by citing Wilson's November 5, 1947 letter to Stone, and by excerpting the conditions quoted above. But Warren did not term these conditions standards or requirements. Rather, Warren's response to Los Alamos urges compliance with these guiding principles. Though Los Alamos was provided with the criteria stated by Wilson in November 1947, General Manager Wilson's statements were not routinely communicated in response to requests for guidance from non-AEC researchers. In an April 1948 letter to the DBM, a university researcher explained that the isotopes division had approved his request to use Phosphorus 32 for experimental procedures in the human, simply for investigational purposes and not for treatment of disease. What, the researcher wanted to know, should be done about medical legal aspects and permission forms. The request could have been answered by referring to Wilson's 1947 statements about consent. Instead, the DBM simply referred the researcher to the isotopes division at Oak Ridge. In its response, the isotopes division did not indicate that consent should be solicited, as Wilson had stipulated. The isotopes division, stating it could be of little assistance, declined to provide legal advice, save to note that we understand that most hospitals do require patients to sign general releases before entering into treatment. From 1947 onward, the AEC had ample opportunity to disseminate a research policy. The AEC routinely provided educational and administrative materials to applicants for AEC funding and to the far greater number of applicants for AEC-produced radioisotopes. The isotopes distribution program, in particular, included a sophisticated structure of regulation, replete with review committees, training courses, and informational brochures. At the federal level, this included the subcommittee on human applications of the Committee on Isotope Distribution, whose very purpose was to review all initial requests for radioisotopes to be used experimentally or otherwise in human beings. The AEC subcommittee on human applications was supplemented by similar committees at the research institutions where the work was conducted. In principle, there does not seem to be any reason these local committees could not have been instructed by the isotopes division on consent requirements. Some evidence suggests that in March 1948 the subcommittee on human applications discussed consent requirements for healthy subjects and patient subjects. In a document dated March 29, 1948, the subcommittee on human applications appeared to resolve that, 1. Radioactive materials should be used in experiments involving human subjects when information obtained will have diagnostic value, therapeutic significance, or will contribute to knowledge on radiation protection. 2. Radioactive materials may be used in normal human subjects provided. A. The subject has full knowledge of the act and has given his consent to the procedure. B. Animal studies have established the assimilation, distribution, selective localization, and excretion of the radioisotope or derivative in question. 3. Radioactive materials may be used in patients suffering from diseased conditions of such nature that there is no reasonable probability of the radioactivity employed producing manifest injury provided. A. Animal studies have established the assimilation, distribution, selective localization, and excretion of the radioisotope or derivative in question. B. The subject is of sound mind, has full knowledge of the act, and has given his consent to the procedure. 4. Investigations are approved, one, by medical director or his equivalent at the installation responsible for the investigation, 2. By the director, division of biology and medicine, and 3. Full written descriptions of experimental procedures and calculated estimates of radiation to be received by body structure and organs must be submitted. We were unable to locate any further references to this document and do not know whether it represented a policy that was adopted. Perhaps it represents the consensus of the subcommittee on human applications, as it had met shortly before that, or perhaps it is simply a draft document prepared by staff. Whatever the ultimate disposition of this document it provides some idea of the problems that were under consideration at the time and indicates that views on human use were unsettled. The first numbered item, for example, appears to recommend human radiation experiments when they will offer diagnostic value and therapeutic significance or knowledge about radiation protection. If the document had, in fact, been adopted, the recognition that isotope experimentation could be undertaken to contribute to knowledge, item 1, would appear to revise the Wilson letter's prohibition of non-therapeutic experimentation. The third item also addresses consent and risk of injury to patient subjects without indicating that there should be any potential benefit. Another peculiarity is found in the second item, which refers to consent from normal human subjects but does not rule out experiments that present risk to the subject. In any event, at a 1948 meeting the subcommittee on human applications articulated a consent requirement as part of a decision to permit patient suffering from serious diseases to receive larger doses for investigative purposes. This requirement was disseminated to all radioisotope purchasers in 1949. The subcommittee allowed investigators to administer larger doses to seriously ill patients but only with the patient's consent. While it is possible that the basis for permitting larger doses was an assumption that smaller ones would be of no potential benefit to subjects, item 3 of the just quoted March 1948 document suggests the assumption was rather that in seriously ill patients other disease processes would be more likely to take their course before radiation injury was manifested. There is evidence that at least one AEC-funded entity did routinely provide some form of disclosure and consent in the early 1950s. From its opening in 1950 the AEC-sponsored Oak Ridge Institute for Nuclear Studies, O-R-I-N-S, a research hospital, advised incoming patients that procedures were experimental. Additionally, patients were given written information that advised them that probable benefit, if any, cannot always be predicted in advance. Patients were also asked to sign a form that indicated that they were fully advised about the character and kind of treatment and care which would be for the most part experiments with no definite promise of improvement in my physical condition. Thus, at least in the case of O-R-I-N-S and perhaps other AEC facilities, a local process was instituted apart from any known communication of the statements by AEC officials. Nonetheless, there is other evidence that the AEC did not communicate the requirements detailed in General Manager Wilson's 1947 letters to its own contract research organizations, which, as in the cases of Argonne, Los Alamos, Brookhaven, and Oak Ridge, had significant biomedical programs and were engaged in human research. When the Division of Biological and Medical Research at Argonne National Laboratory met in January 1951 to discuss beginning a program of human experimentation in cancer research, one of its members asserted that the ACBM had not established a general policy concerning human experimentation. The minutes of the meeting at Argonne record that the ACBM has been approached several times in the past for a general policy and has refused to formulate one. In 1956, Los Alamos asked the DBM to restate its position on the experimental use of human volunteer subjects for tracer experiments. The DBM responded by stating that tracer doses might be administered under certain conditions, which included the provision that subjects be volunteers who were fully informed. The focus of this position seems to have been research with healthy people and not patients, and no reference was made to the provisions of the Wilson letters. The DBM's 1956 formulation was given staff distribution by Los Alamos and restated in 1962. Also in 1956, the isotopes division did stay to requirement for healthy subjects. All subjects were to be informed volunteers. As part of its recommendations and requirements guidebook for the medical uses of radioisotopes, which was distributed to all medical users of radioisotopes, the isotopes division stated, Users of radioisotopes in normal subjects for experimental purposes shall be limited to A. Tracer doses which do not exceed the permissible total body burden for the radioisotope in question. In all instances, the dose should be kept as low as possible. B. Volunteers to whom the intent of the study and the effects of radiation have been outlined. C. Volunteers who are unlikely to be exposed to significant additional amounts of radiation. These requirements apparently applied to all uses of AEC radioisotopes, whether government or private researchers were involved. The experimental or non-routine use of radioisotopes in any human subjects was limited to institutional programs where local review committees existed to oversee the risk to which subjects were exposed. In stating these requirements, the AEC reiterated that patients in whom there is no reasonable probability of producing manifest injury may be used in some experiments not normally permitted, but did not reiterate the requirement that consent should be obtained from these patients, as was stated in 1948. What then can be said about the rules and policies of the AEC in the 1940s and 1950s? General Manager Wilson's 1947 letters clearly stipulate a requirement of informed consent from patient subjects, at least where potentially poisonous or harmful substances are involved. But with the exception of O-R-I-N-S, there is little indication that this requirement was imposed as binding policy on any AEC facility, contractor, or recipient of radioisotopes. By contrast, later requirements that healthy subjects be informed volunteers and that seriously ill patients be permitted to receive higher doses, only with their consent, appears to have been more broadly communicated and enforced. The only evidence of general attention to matters of consent from patient subjects comes from O-R-I-N-S, whose policies and practices show a striking similarity to those that, as we shall see, were being contemporaneously employed at another facility, essentially devoted to experimental work, the NIH's Clinical Center. At the same time, there is evidence of considerable attention in both policy and practice to issues of safety and acceptable risk. C. CHAPTER VI Questions of subject selection, as in the case of seriously ill patients, emerge only in this context of safety. There is no evidence that issues of fairness or concerns about exploitation in the selection of subjects figured in AEC policies or rules of the...