Has our focus on the cell and the molecule eclipsed our view of the whole human being suffering from a disease? Has progress in basic biomedical research outrun our capacity to translate its discoveries into the treatment and prevention of serious diseases? Immunologist Ralph Steinman and freelance science writer Maia Szalavitz explore the unappreciated virtues of the profession traditionally responsible for bringing knowledge from the laboratory to the patient’s bedside—the physician-scientist— and see an endangered species. If the decline of the physician-scientist is to be reversed, we will need to change the organization of medical schools, ﬁnancial incentives for careers in clinical research, attitudes toward human research, and regulatory oversight. The rewards for human health could be immeasurable.
In 1944, three researchers at Rockefeller University published a paper that laid the foundation for the modern revolution called molecular biology. In discovering that DNA is the material that transmits genetic information from one generation to the next, Oswald Avery, Colin MacLeod, and Maclyn McCarty paved the way for the work of Francis Crick, James Watson, and Maurice Wilkins in the early 1950s. That work elucidated DNA’s structure and earned Crick, Watson, and Wilkins the 1962 Nobel Prize in Physiology or Medicine.
The pivotal studies carried out by Avery and his colleagues were a quest to understand a speciﬁc disease. In the early decades of the 20th century, pneumonia was the leading cause of death in America, killing more people than cancer or heart disease. Although today it is seen as a disease of old age, pneumonia then killed young and old alike. The research that led to the discovery of DNA’s role was stimulated by the observations of physicians who treated thousands of pneumonia patients and obtained biological samples from them to study. Without this investigation of how certain types of pneumonia “transformed” into others, the discovery that DNA carried genetic information would not have come as early as it did, and Watson and Crick’s work—and much research that followed— might well have been delayed.
Ironically, the discoveries by Avery and his colleagues, so energized by their study of human disease, helped set research on a path that has led away from work with human subjects and toward work at a cellular and molecular level. Indeed, this trail has taken scientists so far away from treating patients with devastating diseases that public and private granting agencies have begun to set up programs to lure larger numbers of scientists to combine work at the laboratory bench with care at the patient’s bedside. Thus far, these programs have failed to have an impact.
One seasoned biomedical researcher, Gerald D. Fischbach, M.D., executive vice president for health and biomedical sciences at Columbia University and former director of the National Institute on Neurological Disorders and Stroke, says: “I think the shortage is quite extreme. My experience at NIH and at Columbia is that we are not attracting enough, and enough high quality, people to do this work. I can’t put an exact number on it, but I think we need at least twice as many as we have now. The number of really important ideas generated is far greater than the number who can perform the research to test them.”
Historically, medical research was conducted by physicians, but the molecular and cell biology revolution changed that dramatically by the early 1960s. Since then, even basic research on particular diseases has required specialized skills that most doctors never develop.
The severing of basic or laboratory research from clinical or patient-related research is a systemic problem. In my own ﬁeld of immunology, for example, discoveries keep accumulating but, as yet, have not resulted in a cure for any major human disease. A crucial part of the solution—which is to increase the number of physician-scientists and improve the conditions under which they work—runs into obstacles at every step: recruitment, education, early career development, the structure of medical schools, and regulatory oversight of human research.
THE MAN IN THE MIDDLE
Historically, medical research was conducted by physicians, but the molecular and cell biology revolution changed that dramatically by the early 1960s. Since then, even basic research on particular diseases has required specialized skills that most doctors never develop. Programs like Clinical Scholars (now run by the Robert Wood Johnson Foundation) and the NIH’s Medical Scientist Training Program sprang up in the 1960s in part to address this problem, but most graduates of these programs wound up doing basic or animal research, not studying patients and human disease. For example, a 1996 study of the Medical Scientist Training Program found that 83 percent of its graduates published their research in journals of basic science, not clinical medicine.
Today, in the United States, we have 700,000 physicians, but only 14,000 of them are working to apply what is learned in the lab to human disease. What is worse, their numbers have been declining, both in absolute terms and as a proportion of the total number of physicians. From 1980 to 1997, the number of physician-scientists declined by 9 percent.
We risk being able to treat models of diseases such as multiple sclerosis (MS), cancer, and depression in rats and mice, but not having enough scientists, expertise, or funding to test much of this critical work on humans in a timely fashion. The physician-scientist, in the age of cellular biology, has become an endangered species just when more work is urgently needed in ﬁelds like modern neuroscience so we can actually help people with brain and other diseases.
CREATING AND CONTROLLING IMMUNITY
I will illustrate in just one vital area of basic research—the fundamental understanding of immunity—how far we have come in our discoveries without yet generating any major applications to curing or preventing disease.
If we could target these cells properly in humans, we might be able to accomplish some amazing things, but without more human research we simply cannot take this new information from the lab to patients suffering terrible diseases.
I study a type of immune cell called a dendritic cell. In the early 1990s, my colleagues and I showed that these cells were “nature’s adjuvant,” meaning that they direct the immune system to strike at a particular enemy. Later, we found that dendritic cells can also turn off, or “tolerize,” dangerous immune cells that may otherwise attack the body’s own tissues. So these cells have a role both in creating and curtailing immunity.
The immune system, of course, plays a role in many diseases, and we had managed to ﬁnd one of its main switches. Other researchers had been investigating how the immune system worked once activated, but we looked for what turned it on and controlled it. It had been apparent by the early 1970s that antigens (foreign substances in the body) did not switch on the immune system by themselves; if they did, they would cause all kinds of unnecessary and potentially harmful immune reactions. If you have a harmful antigen, I wanted to know, how do you make the immune system respond as it should?
In 1989 and 1990, we published the “nature’s adjuvant” experiments in The Journal of Experimental Medicine. When we loaded dendritic cells with a foreign protein and injected them into animals, the animals became immune to that protein. Likewise, when we injected a foreign protein into an animal, the dendritic cell was the main cell type that captured it as a ﬁrst step in stimulating immunity.
More than 10 years later, we are still trying to understand more about how dendritic cells operate. If we could target these cells properly in humans, we might be able to accomplish some amazing things, but without more human research we simply cannot take this new information from the lab to patients suffering terrible diseases.
We might, for example, learn how to tolerize cells to stop them from attacking the nervous system and causing autoimmune diseases such as MS. I have been developing a theory that the reason certain tissues are particularly vulnerable to autoimmune attack is that dendritic cells do not present their proteins to the immune system frequently enough to tolerize it to them. This could be part of the problem in MS, and we might be able to intervene if we knew more about human neuroimmunology.
The ability to control immunity and its ﬂip side, tolerance, might enable investigators to learn how to control the immune system better in other diseases, as well. We might activate cells against proteins found on cancers and thereby dissolve tumors. We might tolerize cells to proteins on transplanted organs and thereby prevent rejection. We might even activate cells to ﬁght infections like HIV, which seem somehow to fool the immune system into neglecting them. Even conditions like Alzheimer’s disease, atherosclerosis, and stroke seem to involve some components of immune activation in their progression. If we could deactivate or tolerize cells, might we be able to reduce the damage caused by these conditions? We cannot take this research much further without studies in humans.
...journals are far more likely to publish a simple, “elegant” experiment that shows straightforward, positive results than one with all the confounding variables present in research on humans.
THE ENDANGERED PHYSICIAN-SCIENTIST
It can take 12 to 14 years, including medical school, to train a physician-scientist. That is much longer than the training required for either pure basic research or the practice of medicine, because the work of physician-scientists involves major aspects of both. Even beyond formal education lies the formidable process of becoming established.
Getting to the point of publishing research results, upon which an academic career depends, is faster and easier in pure basic research. The time required to design and conduct an experiment, the money and staff needed, the approvals required, and the number and complexity of variables in the research itself are all less daunting in animal than in human research. Furthermore, journals are far more likely to publish a simple, “elegant” experiment that shows straightforward, positive results than one with all the confounding variables present in research on humans. Worse, journals rarely publish studies that show no effect for a treatment or intervention. Since the complexity of studying humans greatly increases the probability of such outcomes, it is a real risk for a young academic to try a human study. Spending years without publishing, then coming up with a null result, is not the royal road to the recognition, grants, and academic appointments that lead to continued success and more work in science.
Then there is the debt. Most physicians leave medical school with more than $100,000 in student loans. If they plan to do patient-oriented research following their residency and internship, they then do a fellowship. Fischbach points out: “You have people in their mid-thirties earning about $40,000 a year, which is an average salary for fellows. With over $100,000 in debt, they can’t afford to own a home and raise a family.”
Not that there would be much time for that. Many physician-scientists ﬁnd themselves slowly dropping the laboratory part of their work because managed care has increased the number of patients they must treat to make a living. Managed care also reduces support for patient-oriented research in a less obvious way, by cutting the funding to teaching hospitals that covers patients in clinical trials. “Managed care is like a wet blanket over everything,” says Fischbach.
Husband and wife Mahdav Dhodapkar (shown with a patient) and Kavita Dhodapkar (in the research laboratory) are pursuing careers as physician-scientists. Mahdav is an assistant professor at Rockefeller University and an associate in hematology at Memorial Sloan Kettering Cancer Center, both in Manhattan; Kavita is a Clinical Scholar at Rockefeller, working to cure brain tumors in children.
© 2003 Jeanine Pohlhaus
TWO CAREERS IN PATIENT-ORIENTED RESEARCH
The day-to-day obstacles faced by people pursuing this work, and the work’s tremendous potential, can be seen in the careers of Mahdav and Kavita Dhodapkar, young colleagues of mine at Rockefeller University. Both do patient-oriented research. Mahdav is an assistant professor at Rockefeller University and an associate in hematology at Memorial Sloan Kettering Cancer Center. His wife, Kavita, is a clinical scholar at Rockefeller, working to cure brain tumors in children.
Mahdav studies a deadly form of blood cancer called multiple myeloma, the disease that recently killed the beloved advice columnist, Ann Landers. He says: “Multiple myeloma is essentially incurable. It kills 15,000 people each year, and we have not really made much progress in terms of preventing it or even curing a subset of patients.” The cause of the disease is not known, but it does its deadly work when cancerous blood cells proliferate and build up inside the bone marrow, damaging tissue, causing pain, and increasing the risk of infections.
The system for treating children’s cancers could be a great guide to expanding patient-oriented research in other areas.
By contrast, huge progress has been made with another blood cancer, leukemia, particularly in children. In 1962, fewer than 5 percent of children with leukemia were long-term survivors; today, there is a 75 to 80 percent cure rate. Kavita points out that, unlike in adult cancer, where only 1 percent of patients are enrolled in clinical trials, in childhood cancer some 60 percent of patients are in trials. Pediatric oncologists have placed special emphasis on ensuring that physicians learn as much as possible, as systematically as possible, about children’s cancer in each tragic case. By creating two consortia to encourage participation in studies, they have had an unprecedented level of success. This research directly on patients has saved thousands of lives. The system for treating children’s cancers could be a great guide to expanding patient-oriented research in other areas.
Mahdav became interested in multiple myeloma while serving as a clinical fellow at the Mayo Clinic. “The most fascinating thing was that some patients had a very indolent form of disease, which would take forever to develop and cause harm, while in others, there was a very fulminant course. I wanted to know what distinguishes the two.”
When a colon polyp is found that could lead to cancer, it would be unethical to leave it in place to study its progress; similarly, if suspicious cells are found on the cervix, they must be removed rather than studied. But because there is no way to remove the dangerous cells in multiple myeloma, says Mahdav, “we are forced to watch, and this provides an opportunity to learn the basic principles by which precancerous cells can transform themselves into cancer.”
He emphasizes that this process is complicated, going far beyond the changes that take place inside a cell to make it reproduce uncontrollably. “There’s clearly more to the development of cancer than what goes on in the cancer cell itself,” says Mahdav. “The interaction between the host and the tumor is equally important from our perspective. That’s what drives us to study patients. There’s no way on earth you can model this effect in nonhuman animals.”
Mahdav is now working to use dendritic cells to produce an immune response against the cancer cells in patients with multiple myeloma. “We have done a series of studies using dendritic cells. We had successfully immunized mice many times, but we could not predict based on the mouse model what we would ﬁnd in people.”
In fact, when the researchers ﬁrst tried dendritic cells in humans, the cells had the opposite effect of what was expected: That’s when it was discovered that dendritic cells could either produce tolerance to a substance or cause the immune system to attack. “We learned that the maturational state of a dendritic cell determines whether it will have a positive or negative response to the same antigen,” says Mahdav. “We could either immunize or tolerize and that opened up an entirely new area. In cancer, our primary goal is to boost immunity; we try to stay away from suppressing it. But people studying MS or lupus want to suppress it, and the way this worked was not apparent from mice.” The time scale in studying rodents is also a disadvantage here, because it takes at least ten years for a colon polyp to become cancerous; mice and rats live only two to three years.
“As a physician, you get into this business to help patients, but doing patient-oriented research is even better because you are trying to create the tools that will help not just your patient directly today, but the next 10 patients someone else will take care of.”
Mahdav explains that in his case two factors were crucial in enabling him to become established as a physician-scientist. First, he was able to take time to work just in the lab for several years, to hone his research skills. Special funding from Rockefeller University and the NIH made this possible. Second, he found mentorship critical. “Most people lack mentors who will guide them through the period of time needed to learn the basic principles of patient-oriented research,” he says. After focusing exclusively on work in our lab (he now has his own, here at Rockefeller), Mahdav took his position at Sloan Kettering to add direct patient care into his work.
“A typical day is a lot of juggling,” he says. “There is one half-day a week which is purely in the clinic, and I try to spend a large amount of time doing hands-on research in the lab with students and post-docs. Every day I create a speciﬁc block of time to deal with each issue. It does ask for some longer days, but it is overall very satisfying.” He adds: “As a physician, you get into this business to help patients, but doing patient-oriented research is even better because you are trying to create the tools that will help not just your patient directly today, but the next 10 patients someone else will take care of.”
Kavita says she felt that patient-oriented research was the best ﬁt for her as a physician. “I want to directly help patients and ask questions as they relate to patients,” she says. She did her residency at the Mayo Clinic, then a clinical fellowship at St. Jude’s Children’s Hospital in Memphis. She recently completed a fellowship in neurooncology at New York University and has been at Rockefeller for nearly two years.
“Leukemia is often curable now,” she says, “but not brain cancer.” Kavita faces an especially difﬁcult challenge in working with human subjects, because research on children is, for good reason, highly regulated and has difﬁcult issues related to informed consent. Neuroscience, too, faces tricky consent issues, since patients with brain diseases may understandably have difﬁculty comprehending the potential risks of treatment. Combine the two and it is even more complicated—although the seriousness of the illnesses, the lack of effective alternatives, and the support for research among pediatric oncologists mean that the case for taking some risks is easier to make.
There is nothing more challenging, nothing more stimulating, than tackling a complex human disease and learning to treat it, even cure it, in patients. We need to reclaim that sense of excitement so that the distractions of basic science in animals do not deter many of our best people from working with humans.
Kavita is now studying the immune responses of children with certain brain tumors, and trying to enhance these responses in order to ﬁght the cancer. She has just begun studying the immune cells in culture, to measure what the immune system from a patient is capable of doing, and to learn how to enhance the immune system’s tumor-ﬁghting potential. If she succeeds in improving anti-tumor response, she will then test the process in patients.
EXCITEMENT THAT OVERCOMES OBSTACLES
What will motivate more young people to look to careers as physician-scientists? One issue is just letting them know how exciting and intellectually stimulating human research can be. For the last 50 years, there has been tremendous excitement about cellular and molecular biology—about decoding the human genome and ﬁnding the genes and proteins and receptors involved in various illnesses and traits. The brightest young people are attracted by these hot areas and believe them to be the best ﬁelds for research. But this has not resulted in cures or even signiﬁcantly helped most patients. We urgently need to bring all the work that has been done in the last few decades back to the patients.
There is nothing more challenging, nothing more stimulating, than tackling a complex human disease and learning to treat it, even cure it, in patients. We need to reclaim that sense of excitement so that the distractions of basic science in animals do not deter many of our best people from working with humans. We need to show young scientists real reasons why they should ﬁght to overcome the obstacles and spend their time on patient-oriented research.
Medical schools might produce better physician-scientists if such single-minded focus on a particular illness or condition were encouraged early on. When someone becomes interested in studying a particular condition, like Madhav with myeloma or Kavita with children’s brain cancer, it helps them integrate and assimilate what they are learning, rather than being drowned by the vast general knowledge base that is the proviso of current medical education and that often is of little use beyond the next exam.
The regulatory obstacles to actually conducting human research can be the ﬁnal straw for some young scientists, who ﬁnd that their research ideas take months, even years, to go forward in humans.
REGULATING HUMAN RESEARCH— BUT NOT TO DEATH
The regulatory obstacles to actually conducting human research can be the ﬁnal straw for some young scientists, who ﬁnd that their research ideas take months, even years, to go forward in humans, while colleagues working with mice can take their ideas from their brains to the lab practically the next day.
Certainly, tough rules and strict oversight are needed to ensure informed consent by patients and enable them to understand the risks and beneﬁts of participating in research. This is particularly tricky in connection with neurological diseases, as mentioned, which can affect cognition. The Institutional Review Boards (IRBs) that currently oversee human research ethics are doing a great job with the resources available to them. But they need to be overhauled so that the process becomes more friendly to researchers, more rewarding to board members, and faster to generate decisions.
For example, right now a researcher has to deal with mountains of paperwork from the IRB and months of back-andforth changes on these documents. IRB members are not paid for this work and are usually busy professionals, so the procedure can seem interminable.
If IRB members were paid, and if research institutions hired people speciﬁcally to help scientists shepherd their projects through the oversight process, researchers could focus on their patients and their data, and delays could be cut. These changes would also make the process safer for patients, because it would reduce the burden on both researchers and IRB members—providing more time to focus properly on ethical questions. Disasters like the death of 18-year-old Jesse Gelsinger in a University of Pennsylvania gene therapy trial would be far less likely if both researchers and IRB members could devote themselves fully to the most important questions, with supporting staff to handle the details.
The National Institutes of Health and several private foundations have recently begun grant and fellowship programs aimed at attracting physicians to patient-oriented research and guiding them through the difﬁcult transitions and balancing acts. The NIH now offers some debt-relief programs, through which medical school debts are paid if a physician commits to patient-oriented research. About 1,000 such students are now supported by the NIH, with 250 new slots added in 2002.
Likewise, some medical schools are beginning to act. Case Western Reserve University School of Medicine and the Cleveland Clinic are launching the Lerner College of Medicine to produce new physician-scientists. The program will begin in July 2004.
Other medical schools are trying particular measures in their curricula to spur more interest in patient-oriented research, but Case Western will be the ﬁrst to weave them into a ﬁve-year program for medical students. The teaching style will be interactive and focus on solving speciﬁc medical problems, with the hope of reducing the “forget it following the test” syndrome that can occur with big lecture classes. Students will work with patients, learning about research from the start, instead of having didactic education ﬁrst, followed by clinical work. To graduate, these future doctors will have to do original research for a thesis, in which they will be guided by mentors and aim for publication. The program also hopes to provide debt relief.
Patients have been patient long enough with basic science on animals and in culture. These efforts have been spectacularly productive. But now we need to direct the beneﬁts to more patients, more diseases.
“We want to produce an individual with the passion, knowledge, and skills to be a physician-scientist,” says Lindsey Henson, M.D., Ph.D., vice dean for education and academic affairs for Case Western Reserve School of Medicine.
In another positive step, some medical schools are beginning programs to offer doctors-in-training master’s and doctoral degrees in areas like biostatistics, which can help prepare them to do human research and streamline their training for both the lab and the clinic. A good understanding of statistics is crucial to proper design of studies and to critically reading research by others—which are necessary for both lab-based and human studies. Finally, many schools are working to make their curricula more exciting and more focused on working with speciﬁc diseases and problems.
It will, however, take more than these promising starts. As with the big research initiatives against AIDS and breast cancer, the public will have to demand that this research be done. From the patient’s perspective, nothing is more urgent. Disease-focused activist groups must add to their agendas a demand for support of physician-scientists; foundations must continue to give priority to this area. The NIH budget will have doubled between 1998 and 2003, and enormous amounts of private money have poured into biomedical science, but as yet, this has not been translated into cures for major diseases such as Alzheimer’s, Parkinson’s, brain cancer, MS, and spinal cord injuries.
Patients have been patient long enough with basic science on animals and in culture. These efforts have been spectacularly productive. But now we need to direct the beneﬁts to more patients, more diseases. In so doing, we will not only learn much more as scientists, but we will all beneﬁt as individuals whose lives may one day depend upon the progress of medicine.
I would like to thank my skillful cowriter, the science journalist Maia Szalavitz, for her research and writing assistance with this article.