Interview with E. Donnall Thomas

Ostrom: Dr. Thomas, tell me about growing up in a little town in Texas.

Thomas: My father was a general practitioner, the only doctor in Prairie Hill, a town of about 800 people. He was born in 1870, and with his family migrated from Tennessee to Texas in a covered wagon in 1874. He went to medical school in Louisville, with virtually no training before that. He had three children by his first wife. She died of tuberculosis in 1917, and he married my mother. I was born in 1920. From the very earliest time I can remember, I had planned to be a doctor. I guess because my father was, and because I admired him so much. I never gave any thought to anything else. Except that things intervened. This was the Depression; there was no money. My father was killed in an automobile accident when I was a student at the University of Texas, and I did not see how I could have money to go to medical school. So I switched to chemical engineering.

Ostrom: It seems that something helped create a strong work ethic in you; you probably know that some people call you a workaholic.

Thomas: My father was. He was always on call. When I was growing up, we'd plan to go fishing or something, and some woman would go into labor, or someone would get a leg broken—he couldn't do major surgery, he couldn't in that environment—but if it were a trauma or obstetrics case, he'd cancel our trip. In the middle of the night, he'd get called. As I remember it, every night. I'm sure it wasn't every night, but that's the way I remember it. But later on, having grown up in this environment, and being in love with medicine, I decided that with my chemistry background I'd much prefer to be in the scientific end of medicine. I didn't want to be the only doctor in a small town.

Ostrom: You met your wife, Dottie, when she hit you with a snowball. How did you get from there to here, being not only husband and wife, but partners in research?

Thomas: I was a senior at the University of Texas when she was a freshman. I was waiting tables at the girl's dormitory, which is how I got my food. It snowed in Texas, which is very unusual—January 20, 1940. And I came out of the dormitory after we'd finished serving breakfast, and there was about six inches of snow. This girl whacked me in the face with a snowball. She still claims she was throwing it at another fellow and hit me by mistake. One thing led to another, and we seemed to hit it off. She's a workaholic, too, and was then. We were married in December of 1942, and I had mentioned, I always wanted to be a doctor. After finishing my master's degree in chemistry, I got a job at the medical school in Galveston as an instructor in pharmacology. I didn't know anything about pharmacology, but I spent three weeks reading a pharmacology book, and I knew enough to be a lab assistant. I went to Galveston and did my first semester as a medical student there. I had a half-time job so I could go to medical school. And in January of 1943, when the war was really getting going, it was announced that the Army and Navy were taking over the medical schools to accelerate the training of doctors for wartime purposes. Since I already had a reserve commission in the Army, I decided that as long as the Army was going to be paying my way to medical school, I might as well apply to some of the famous medical schools. And so I applied to Harvard and Johns Hopkins and Columbia, I think it was. On February 20, 1943, I got a telegram from Harvard, saying if I would get my credentials in for the class starting in March, they would consider me along with the other 1,200 applicants, because there had been one vacancy that appeared at the last minute. And about the first of March, I got a telegram from Harvard saying I'd been admitted.

Ostrom: Tell me about how you became interested in bone marrow. Was it something about bone marrow in particular, leukemia, or the challenge?

Thomas: It was all of those. As a medical student, I had some very stimulating teachers, and a couple of them were hematologists. Because Dottie was a hematology technician, we used to look at smears and bone marrow together when we were students. I found the bone marrow to be a fascinating organ. I can't think of any particular time when I decided to make that my specialty, but by the time I was a senior in medical school, I knew that's what I would do.

Ostrom: Were there a lot of unknowns about bone marrow then?

Thomas: There had been a lot of studies, but in retrospect, it seems we didn't know much at that time. There were people who had been studying bone marrow for 50 years, but a lot of its functions were still a mystery, and its diseases were poorly characterized. It used to be thought that pernicious anemia was a form of leukemia. Going back to my father, I can remember as a kid, his being so excited when [George] Minot and [William] Murphy got the Nobel Prize [in physiology or medicine] in 1934 for their earlier work on pernicious anemia. His enthusiasm was catching. Little did either one of us know at that time that I would later be personally acquainted with both Minot and Murphy.

Ostrom: You were inspired by some studies involving mice and radiation. Would you explain why these studies were so important to you?

Thomas: By the time I had graduated medical school and was a fellow, I spent my first year in hematology with Dr. Clement Finch. He was then in Boston, but he moved to Seattle in 1949 to establish a division of hematology at the then-newborn School of Medicine at the University of Washington. His interest was in iron metabolism. And of course, red blood cells are made in the bone marrow. Also about that time, it was realized that radiation kills animals and people primarily by damage to the bone marrow. It's the most sensitive organ in the body as far as radiation damage is concerned. And of course, in the late '40s, after the atomic bomb explosions, everybody was interested in this. And I became very interested in what governed the bone marrow's production of white cells and red cells and its other functions. At that time, there were some early experiments that suggested there were some growth hormones for bone marrow. Specifically, a little was beginning to be known about erythropoietin, which stimulates bone marrow production of red cells. I took a year off from my clinical work and went to the Massachusetts Institute of Technology [MIT], and worked in the biology department there, with Dr. John Loofbourow whose interest was in wound-healing substances that stimulate cells to proliferate. I worked on substances that are released from irradiated yeast that stimulate yeast cell growth. And my real interest was in transferring this to bone marrow.

Ostrom: But that did not prove to be something you would stick with.

Thomas: No, but I worked on that topic from 1950 to 1955. And it proved to be a very difficult area of endeavor. I set up cultures of bone marrow, and tried to study factors that would stimulate in vitro growth. The follow-up on that story is that other people tried to do this, and no real progress was made for 20 years, until recombinant technology came along, and then it became possible to manufacture erythropoietin, to have it in sufficient quantities.

In 1949 Leon Jacobson of the University of Chicago showed that if you put a lead foil around a mouse's spleen, so it's not exposed to radiation, he could give the mouse lethal radiation, but the mouse would recover. He thought this was due to a hormone released from the spleen. So of course, this is what I was working on. I met Dr. Jacobson in 1950, to talk about these experiments. He also had great trouble trying to isolate any "hormone." In 1955, I was moving from Boston to Cooperstown, New York. A paper appeared in the Journal of the National Cancer Institute. What the authors had done was to give the mouse lethal irradiation, give it an infusion of bone marrow—or of spleen cells, since in the mouse, the spleen is a bone marrow organ. They gave lethal irradiation, they gave an infusion of marrow cells, and the mouse recovered. The marrow cells came from a different strain of mouse. They followed this up by doing a skin graft from the donor mouse to the recipient mouse. Now ordinarily, this skin graft would be rejected. What they showed was that in fact, the skin graft was accepted, as though it was the mouse's own skin. These are different strains of inbred mouse. By that time, it was well known that if you gave bone marrow cells from the same inbred mouse, the mouse would recover from irradiation. Here, they took marrow from a white mouse after this gray mouse had been irradiated lethally, but given marrow from the white mouse. When the [gray] mouse had recovered, they put on the skin graft, and it was accepted. Later on, another investigator showed that a skin graft from a third strain would be rejected. Only the marrow donor's skin would be kept.

Ostrom: And you said.…

Thomas: Bingo! And so did Dr. Joseph Ferrebee, who was also in Cooperstown. I was going there; we were going to work together. He'd seen the same paper.

Ostrom: What did this paper say to you?

Thomas: You couldn't explain what happened on the basis of a hormone. A graft rejection is a cellular phenomenon. We looked at this, and we said: If we can do this in a mouse, we ought to be able to do it in human beings. That was 1955, I was 35 years old. I went to Cooperstown as physician-in-chief of the hospital there.

Ostrom: So you thought you could do it in humans. And you did. How did the first ones go?

Thomas: Terribly. By then, of course, it was well known that leukemic cells were very sensitive to irradiation, just like normal bone marrow. And we had patients dying of leukemia or lymphoma. It was logical to say: We could give lethal irradiation to all the leukemic cells, also killing the normal bone marrow cells, but prevent the death of the patient by an infusion of bone marrow. By then, through the work of others, it was known that you could give bone marrow intravenously in mice, and it would grow in the bone marrow. So we actually started to do this, to give irradiation to these patients and to give bone marrow. At the same time, we started studies in dogs. Other people were working with inbred mice, as were we. But we felt we needed an outbred species where clinical procedures could be done, so we could transfer information from inbred mice to outbred species to human beings. [Inbred mice are the offspring of successive brother-sister matings, which become essentially like identical twins. Outbred refers to normal matings between unrelated individuals.]

Like humans, dogs come in families. We started the dog as a counterpart of our clinical studies. Now, the clinical studies at that time, I guess you could say, were disasters. We didn't get any grafts. We did find that some patients actually had appreciable remissions after 300 or 400 rad [a unit of absorbed radiation dose], getting close to the lethal dose. But we didn't have any success with grafts. The marrow didn't engraft [grow]. But at that time, the lethal dose for human beings was thought to be about 400 rad, which is now called centi-Gray—they changed the terminology. We started with about 200. We worked our way up to 400, and didn't get grafts. In the dog model, we also didn't get grafts. We decided that since we weren't getting grafts, maybe the grafts were getting rejected by the recipient. So we doubled the irradiation. We went up to 800 rad in the dog, which is twice the lethal dose. Lo and behold, we began to get grafts. We went up to 1,000 rad, and we got pretty consistent engraftment, especially with littermate donors.

We decided that we ought to give 1,000 rad to human patients. We didn't even know that intravenous bone marrow in human beings would ever work. We knew it would work in mice, and in dogs. One of the first problems was that we couldn't give that much irradiation to human patients with the then-existing equipment. The most high-energy X-ray machines then were about 300 kilovolts. The thickness of the human body prevents getting homogenous irradiation. We set up a special radiation unit involving opposing cobalt-60 sources. This followed consultation with radiation physicists at MIT and at Columbia. Cobalt-60 has two gamma rays, on the order of 1.2, 1.3 million volts. It's a much higher energy, so we could get enough penetration to do this.

We continued these studies in dogs while we were setting up. We published those first papers [on the irradiation of human patients] in the New England Journal of Medicine in 1957. They were basically all failures. We had one transient graft. But after about six weeks, the graft ceased to function. This is what we now call graft rejection. But by late 1957, we had set up our cobalt-60 sources. And also, through various professional colleagues, we knew of some sets of twins, one of whom had leukemia. I think it was in late 1957, we treated the first of these sets of twins, who had all the then-available treatment for acute lymphoblastic leukemia. At that time, there were only steroids, methotrexate and 6MP [steroids are adrenal hormones that have an antileukemic effect; methotrexate and 6-mercapto-purine, or 6MP, were the first antileukemic drugs that also had immunosuppresive properties; immunosuppression is the use of chemicals or biological agents that interfere with the normal ability of animals or humans to mount an immune reaction to foreign substances]. We didn't have any of the other antileukemic agents that we have now. This girl was in the final stages of advanced leukemia. We decided to give her 800 rad, and an intravenous infusion of marrow from her twin sister. In those days, we didn't have these long, complicated informed consents. All we did was sit down with the patient and family, and explain everything, and explain what we didn't know. Of course, the alternative was that she was going to die within a few days or weeks.

Ostrom: How did you talk to the families about what you wanted to do? You knew the outcome was very likely to be grim.

Thomas: Yes, we did. We'd have to sit down, put all the cards on the table. Say what we know, point out that one option was dying quietly in bed. It's tough, but that's the way it is. With one twin, the parents hesitated. We hesitated. But the alternative was to simply watch her die. We went ahead and gave her 800 rad. The night before we were going to do this, one of the pediatricians called Dottie, and said, "Are you really going to put that child in there and cook her?" Dottie cried all night. The girl was 7. To make a long story short, she tolerated what was then horrendous irradiation with almost no ill effects. And after the intravenous infusion from her sister, she recovered her blood counts very quickly, within a couple of weeks. She was discharged from the hospital. She was an outpatient 30 days later, with no sign of leukemia. Unfortunately, six months later, her leukemia came back, and she died. We owe her for being able to show that an intravenous infusion of bone marrow would protect against twice the lethal dose of irradiation.

Ostrom: After those first twin experiments, then you had some dismal failures.

Thomas: We were all trying to do transplants in human patients. I think basically we stopped in 1959. By that time, a number of other places had tried to do this, and all without success. We concentrated on the dog model, and worked out a lot of the problems: how to get grafts, how to handle the bone marrow, how to freeze bone marrow. We can freeze bone marrow for years if we want to now. And we worked out the radiation requirements. All told, it was 15 years of work. Let me go back a little way again. When I started as an attending physician at the Brigham Hospital in Boston, when I finished my chief residency, that first year, which was 1953, we had a patient with renal [kidney] failure who had an identical twin. After a lot of soul searching, it was decided to carry out a kidney transplant from the normal twin to the one with kidney failure. I was the medical attending; I helped take care of this patient on the medical ward before the transplant. The transplant team was headed by Dr. Joseph Murray [co-winner of the 1990 Nobel Prize with Thomas]. It was decided to do this transplant, and that transplant was successful. That was the first one.

Ostrom: What was the message of that transplant?

Thomas: That told me, and everybody, that the tissues of identical twins were entirely compatible. We didn't know how to do tissue typing then. [Tissue typing ensures that there is a reasonable compatibility between the donor and the recipient.] And then, actually, at that time, it was thought that if one could do a marrow transplant, one could follow it with a kidney transplant, like the skin graft. Which was one of the reasons I was very interested in this, even before this paper appeared. We did that in our dog model. And Dr. John Mannick took a year off from his surgical training and worked with us in Cooperstown, and we did marrow transplants and kidney transplants in dogs that were successful. But about that time, some immunosuppressive drugs had been developed. And again, Dr. Murray and the Boston team were able to show that they could get kidney grafts to survive even from individuals who were not identical twins if they gave enough immunosuppressive drugs. So this irradiation approach was abandoned for kidney transplants.

Ostrom: Despite some successes, though, you stopped working on human patients during the 1960s.

Thomas: Except for identical twins, we did all the laboratory work with dogs. We were working on dog tissue typing. Now, at that time, a lot of people were working on human tissue typing. The human tissue-typing people were way ahead of us. We were simply trying to transfer what they had learned to the dog model, because in the dog model, we could test it. In the human model it was difficult. In the 1970s, when we started in again with people, we were very heavily criticized. There were very many responsible people who said this shouldn't be allowed to go on because our success rates, in the past, had been so poor. In the dog, we were finally able to develop antibodies that recognized dog leukocyte antigens, what we call the DLA system. [Dog leukocyte antigens and human leukocyte antigens are part of a system of mutual tolerance in each species that allows some tissues to be grafted effectively to others. The degree to which the antigens from two individuals are compatible determines the likelihood of a successful graft.] The human leukocyte antigen system (HLA) had been developed to a greater degree than we were able to do in the dog. In fact, Jean Dausset got the Nobel Prize in 1980 for developing the HLA system. For marrow, that turned out to be crucial. In the dog, we studied brothers and sisters, so we didn't accidentally have identical twins. What we found, once we'd developed these typing sera [a typing serum contains antibodies that recognize one of the many antigens on the surface of white blood cells], was that if we irradiated one littermate and gave marrow from a mismatched donor, either the grafts were rejected, or the dog died of graft-versus-host disease [the immunological reaction of the grafted marrow against the tissue of the recipient; the reaction can cause damage to the skin, liver, or gastrointestinal system]. But if the donor were matched, about half of those dogs became long-term survivors. And if we gave a few weeks of immunosuppression, to sort of stop the graft-versus-host reaction, more than 90 percent of those dogs became long-term survivors. Now at that point, we were ready to go back to doing human transplants between brothers and sisters.

Ostrom: In the 1970s, one of your early papers showed something like 12 out of 100 patients surviving. Some people thought you should stop. You kept going. Why?

Thomas: This is what we call "the 100-patient paper." This is when things really got hot. We had 100 patients; 46 with AML, acute myeloid leukemia, and the others with ALL, acute lymphoblastic leukemia. But they were all in the end stages of leukemia. To have 12 of them become long-term survivors was, to me, very impressive. But other people were saying, "Well, you know, only 12 out of 100 isn't worth all this effort and expense." To me, it meant if we can do this with patients with far-advanced disease, even if there's no other progress, we should be able to do much better if we do it much earlier, before they get to the end stage of the disease. Of those 12, 8 of them are living now. And they are all beyond 22 years.

Ostrom: In your work, your curiosity led you in a number of directions. Did you realize early on that solving the mysteries of the marrow transplant was going to involve tissue typing, immune-suppression, and a whole host of different areas?

Thomas: We were learning things as we went along. Again, the dog system was very informative. For example, we were learning about the fact that in the first two or three months after a marrow transplant, dogs and humans are very immunodeficient. They're susceptible to all sorts of opportunistic infections. Some of those things we learned the hard way. Some of our early human patients were dying of Pneumocystis carinii infections [a form of pneumonia]. It was a surprise at autopsy to find that. We learned about all the infections that AIDS [acquired immune deficiency syndrome] patients get long before AIDS patients came along.

Ostrom: In the early 1970s a lot of physicians had very negative attitudes about bone marrow transplants, and they would not refer patients until they were very, very sick.

Thomas: In the first part of the 1970s, we only took patients who had failed everything else, for ethical reasons. When we began to get some of those patients to be long-term survivors, without their disease, it then became possible to do the transplants earlier, while the patients were in good clinical condition. Back in the mid-'60s, we could get 90 percent of the dogs to be good long-term survivors. We could get 5 percent of the humans to be long-term survivors, in part because the human patients were so sick, and came in with advanced infections and everything else.

Ostrom: Did you have trouble funding your efforts?

Thomas: We were lucky that in the '60s and '70s, funding was fairly easy to get—if you had a good project, you could get it funded. As you know, in recent times, you can have a good project, and not get it funded. In about 1967 we were convinced we could do tissue matching, leukocyte matching, in dogs. And it would work. By that time, we also knew that our human patients, even twins, had to have excellent clinical care, and that probably this needed to be by a devoted team. In 1967 we wrote a grant application that included nursing support. In 1968 that was funded. We were then at the old Public Health Hospital here in Seattle. We were able to hire a team of nurses who really wanted to do this sort of thing, and some support personnel at various levels, and even some money for patient care. That grant was funded in July of 1968, and we had our team set up and ready to work in early 1969; we did our first patient in March of 1969. It was the first in what I call the modern era of transplantation, based on knowledge of tissue typing.

Ostrom: Who was that?

Thomas: A man with blast crisis of chronic myeloid leukemia [chronic myeloid leukemia evolves into a form of acute leukemia, called blast crisis, which is usually fatal in a matter of weeks], notoriously untreatable. We gave irradiation only, 1,000 rad. And marrow from his sister. And we got a graft. Which was important to us. After about three months, he began to get progressively sick, and died. But he didn't have leukemia. He was the first of our patients to die of cytomegalovirus (CMV) pneumonia, this opportunistic infection. We did 10 patients with irradiation only. Only one became a long-term survivor. Several of these patients had had recurrences of leukemia, as our first twins had had. So we decided to give some chemotherapy before the irradiation. We decided to give what was considered a terribly high dose of cyclophosphamide. And then we gave the irradiation. Part of the reason for doing this in these patients with advanced leukemia was to prevent the sudden destruction of a huge number of leukemic cells, because a couple of our patients died of plugged up kidneys from uric acid [dead cells can lead to excess uric acid, which can plug the kidneys]. We hadn't seen that in our dogs, because the dogs didn't have leukemia. We wanted to destroy the leukemic cells more slowly, and then give irradiation.

Ostrom: So this was all a series of small steps?

Thomas: Yes, exactly. My good friend and colleague at Johns Hopkins, George Santos, had been working with cyclophosphamide in the rat population, which is why we knew about cyclophosphamide. All these little pieces fall together; we learn from other people, they learn from us. People keep asking me what was the breakthrough for this or that. I always say there aren't any breakthroughs. It's always a step-by-step process.

Ostrom: I'll ask it this way, then: What were the areas in which those small advances turned out to be crucial to success?

Thomas: Later on, additional chemotherapeutic agents became available. Control of infections was one of the major problems. We began to have better antibiotics for bacteria. In those early days, the majority of our patients died of bacterial infections. Through the work of others, agents became available for preventing Pneumocystis infection. One of the big advances in the early '80s was the development of acyclovir for the control of herpes infections. And in the late '80s, the development of ganciclovir for the control of CMV, cytomegalovirus infections. Better immunosuppressive agents for control of graft-versus-host disease. Increasing knowledge of the nature of the gastrointestinal problems, and how to manage them. Esophageal ulcers, veno-occlusive disease of the liver, graft-versus-host disease of the gut are all major problems in marrow transplant patients. First you have to recognize the problems, then you have to figure out what you're going to do about them. It's just a series of steps. It's interesting looking back that some of the problems we spent several years on could now probably be solved in a month or two—the technology is amazing. I used to go in about 5:30 in the morning to set up these tissue typing trays; because the anti-sera we used were so imprecise, I was afraid that the technician might make a mistake, so I'd do it myself. But now, with the advent of the actual molecular nature of these antigens, and doing the tissue typing by molecular techniques, it's rapidly replacing all these imprecise serological things. It's an entirely different ballgame. By the way, some of those failures, when we thought we were transplanting matched brother or sister, it turns out they weren't matched. Now, we can go back retrospectively and repeat the typing. So some of our inexplicable failures are now explicable.

Ostrom: What are some of the big problems left to solve with bone marrow transplants?

Thomas: We still have patients dying of graft-versus-host disease. We are doing more and more transplants using unrelated donors. And with the modern molecular typing, this is getting to be much more feasible, but there are still many problems. The problem of recurrence of leukemia, which still happens despite our best efforts. There is a long list of problems, which is why the survival rate is not 100 percent.

Ostrom: Is there a magic bullet?

Thomas: I don't think so. For example, one of the annoying problems is this: We've controlled bacterial infections; we've now pretty much controlled viral infections and parasitic infections. We still have fungal infections. The candidal-type infections [yeast infections] we can control pretty well. But there is still a class of fungal infections like Aspergillus, which we don't have a good way of controlling. Somewhere between 5 to 10 percent of our patients die of Aspergillus infections. We thought back in the '80s that this could be solved if we set up laminar air-flow rooms [sterile rooms in which air is supplied through filters that remove bacteria]. We did that. Half of our rooms—30 rooms—were laminar air-flow rooms. We spent 10 years studying it. The bottom line was we decided we didn't need them. We don't need them for bacterial infections—we can control those. They don't prevent fungal infections, because the patients come in with fungal infections. And particularly the patients with leukemia—this is not so much true with inherited diseases—but patients with leukemia have all had chemotherapy for their initial treatment, and many of them get fungal infections. The reason the laminar air-flow rooms didn't control the Aspergillus infections was that the patients already had them when they came in; they were already on board, so to speak.

Another problem is the fact that some patients, even with molecular matches, still get graft-versus-host disease. This is due to antigen systems that are outside the HLA system. We call them minor transplantation systems, to distinguish them from the major. We've known for a long time from inbred mice that they exist, but it's been very difficult to study them, to define their nature, in human beings. One of the major laboratories here is directed at that problem—looking at ways to type for these other systems, if you will. When Dausset won the Nobel Prize for the HLA system, one of the co-winners was George Snell of Bar Harbor [Maine], who had worked out these systems in mice. He had shown very clearly that there is a major system in the mouse, as there is in man, but also because he could do inbreeding and backcrossing [backcrossing occurs when offspring are mated to a parent], he defined many of these other systems in the mouse. Well, we can't do controlled mating and backcrossing in human beings, so we have to have another approach.

Ostrom: Which brings up a question: Why did you pick dogs in the beginning?

Thomas: There's a history there. Surgeons have used dogs for a long time for experimental procedures. Alexis Carrel, who got the Nobel Prize about 1912, had shown that he could transplant kidneys in dogs. The kidney would function, but after about ten days, it would stop functioning, because it was being rejected. When I was a medical student, everybody knew that you could never transplant tissues from one individual to another because they would be rejected. You can see why I got excited about that picture [of a mouse with a skin graft]. Dogs had been studied, and they're relatively inexpensive. Some people—Dr. D. W. van Bekkum back in the '50s—had been trying to do these studies on nonhuman primates, but they were all infected, they didn't come in families, so he had big problems. Dogs were much less expensive than primates, pleasant to work with, and they come in families. And one can get a daily blood count, do blood transfusions.

Ostrom: Some people have ethical problems with research on animals. What's your view on that?

Thomas: Without animal research, life-saving transplants of human patients would not exist today. Every time I talk to a reporter, I worry about this. We've had remarkably little difficulty with the animal rights people. I like to think that one of the reasons was that early on, and even now, we do some of these things to treat dogs. About 15 years ago, one of the Seattle papers had a major article with pictures of dogs with leukemia or sarcoma that we were returning to their owners. What we've been learning is just as applicable to dogs as it is to human beings. Maybe that influences them. There are rational people in the animal rights movement and there are irrational people. The rational ones, we've always been willing to have them come and see what we're doing, and see how we're taking care of the dogs, and they see how human beings are benefiting. It would have been impossible to do this without mice or dogs. You can do a lot of things in test tubes and tissue, but ultimately, let's face it, the living patient or living animal is so much more complex than we can comprehend, that we have to try these things, as we say, in vivo. When we were first doing these studies in dogs, in Cooperstown, Dottie's good friend, the librarian—she was always getting books out of the library—was saying she hoped we stopped using these dogs, it was terrible what we were doing. Dottie said, "Don't you understand? This is going to save the lives of people!" And she said, "People aren't worth it."

Ostrom: What would you tell young people who might be interested in medical research?

Thomas: I've derived tremendous satisfaction from this life. I can't visualize any other, really. Where you can have important problems to work on, that involved lots of scientific disciplines. Where you can have bright, stimulating colleagues to work with. Where it's not only challenging and interesting, but fun. It's satisfying. Here I am, about to be 79, and I'm still coming to work every day, because I can't think of anything I'd rather do.