Peter Huttenlocher was unendingly interested. Interested in music. In baking. In medicine. In people and their ways of thinking. In his patients. In the science of the human brain. “Peter is a great dreamer,” said Hanne, Peter’s paternal grandmother, in 1948. Although he delved deeply into philosophy in college, Peter decided early after arriving in the United States that he wanted to become a physician. He wanted to take care of people. So, he immersed himself in human medicine. He memorized. He took part in the multi-year training process for doctors that “knocks every ounce of originality out of you” [Reference Harper1].
Medicine requires certainty. It values decisiveness and a type of conformity. A respect for hierarchy and established protocol. Even as Peter delved into medical physiology, pathology, anatomy (and the grueling task of memorizing every muscle of the human body – over 600!), he also tried out science for the first time as a medical student. Science, unlike medicine, is uncertain and curiosity-driven. It requires a kind of wonderment as well as a measured disregard for authority and procedure. As a physician speaking to a patient, one is not supposed to say “I don’t know.” But for a scientist, “I don’t know” is the start of the next question, the next experiment, the new frontier [Reference Lane2].
How does a physician scientist wed these conflicting styles? Why train in both science and medicine? Can’t this merging of ideas and established knowledge be done efficiently by having scientists and physicians talk to each other? Absolutely – such collaboration is the backbone of biomedical science. However, the physician scientist is a different beast. There is something unique about understanding both the science and the medicine. For example, new connections between ideas are made when you walk into a patient’s room. And if it is part of your job designation, the physician scientist (who is only a subset of the physicians even at a medical school) is enabled to “think outside the box” with a patient and take some risk. For example, develop a new ketogenic diet as a treatment for particularly problematic cases of epilepsy.
Physician scientists have often played a critical role in fundamental discoveries. An example is the discovery of statins – a medication that more than 35 million Americans in any given year take to lower their cholesterol and prevent heart disease. In the 1970s,Joseph Goldstein and Michael Brown, physician scientists at University of Texas Southwestern (UTSW), were intrigued by a group of patients at increased risk for cardiovascular disease who had an inherited form of high cholesterol, known as familial hypercholesterolemia. Brown and Goldstein discovered that patients with familial hypercholesterolemia lack a critical lipid receptor known as the low-density lipoprotein (LDL) receptor, resulting in high levels of the “bad” lipid, LDL, which is associated with heart disease. This work was recognized by the Nobel Prize in Physiology or Medicine in 1985. Brown and Goldstein and others went on to find that statins lower cholesterol in patients with high cholesterol, who are at risk for cardiovascular disease. This is a classic example of physician scientists going from the patient’s “bedside” to the laboratory bench and back to the bedside with new knowledge that helps to advance new treatments.
The rare patient with a puzzling constellation of symptoms often finds their way into the physician scientist’s clinic. Their disease may become the physician scientist’s next experiment.Dan Kastner, a physician scientist at the National Institutes of Health (NIH), was puzzled by a group of patients with intermittent fevers, pain and systemic inflammation. Many of these patients have bizarre episodic symptoms that are debilitating, like patients with a disease called familial Mediterranean fever (FMF). Kastner and his colleagues uncovered the molecular defect in a protein, known as pyrin, that mediates FMF, launching a new set of discoveries that identified mechanisms for a group of disorders known as autoinflammatory diseases. Patients with these disorders have aberrant activation of inflammatory mediators known as cytokines, including interleukin 1 beta (IL1β). Kastner and others by now have identified many different autoinflammatory diseases in which cytokines are key, have advanced the field of immunology and have identified a key protein cluster known as the inflammasome that regulates IL1β activity. This work led to targeted therapies for autoinflammatory diseases through the inhibition of IL1β signaling (drugs such as anakinra). Treatment can be highly effective for patients with autoinflammatory diseases and other forms of systemic inflammation, including even cardiovascular disease. This is another example where physician scientists went from a rare disease to identifying mechanism to treatment. This work takes the collaboration of many scientists, physicians and physician scientists to finally reach the clinic and influence patient care. However, a key step along the way was the role of a physician scientist, Dan Kastner, who saw the patients and made pathbreaking connections regarding the disease mechanism and possible treatments.
Not all discoveries are worked out so quickly. Sometimes a discovery is made multiple decades before the development of associated disease treatments. This was the case with synaptic pruning. However, the fact that Peter had training in both medicine and science expedited the discovery. A non-clinician may have been less likely to even be studying post-mortem human brain specimens. A non-scientist might not have systematically counted synapses in multiple specimens. Peter’s understanding of the human brain and the ways it goes wrong provided a perspective, helping him to know that what he was finding in his studies of those brain samples was startling. We noted above that, as a scientist, when one sees something potentially exciting, a standard reaction is to think “it cannot be true.” But Peter figured out that what he was seeing was something fundamental that also explained much about the patients he saw in his clinic. Rather than ignore the oddities and pursue a different angle, he instead delved deeper.
Given the key insights and investigations so often launched by people with dual training in science and medicine, why is there a shortage of physician scientists? There has been much written about the “vanishing physician scientist” [Reference Schafer3]. It is hard to do both science and medicine well, especially in the age of clinical productivity metrics where there is pressure to see more patients due to reimbursement schemes. It is hard to win competitive research grant funding when you are only doing science part-time. And it is hard, while hearing one’s college friends speak of their new high-paying jobs and house purchases, to delay by multiple years one’s rise out of lower-paying medical trainee positions and then earn less because of lower clinical throughput. It has pushed many physician scientists to make a choice between science and medicine, and to make the choice earlier in their career paths. This attrition of physician scientists is well documented and many solutions have been proposed. One key aspect of dual training – whether through a combined MD/PhD program or through later training of MDs after residency as research fellows – is to teach a willingness to, in essence, “dabble” and take some risks. We have seen that one driver of fundamental medical discoveries is a willingness to respond to the disparate patients that arrive through the door and dig deeply into what is going on. To think about the problem with a “curiosity.” And to capitalize on that immense personal knowledge physician scientists possess of the human organism, spanning medicine and science. This is what drove Bill Kaelin, a physician scientist and cancer biologist who discovered a critical cancer pathway involving how cells sense and adapt to oxygen. He and a group of physician scientists, Peter Ratcliffe and Gregg Semenza, won the 2019 Nobel Prize in Medicine or Physiology for work that was influenced by the clinic and led to groundbreaking basic science discovery in oxygen sensing.
In part, this kind of thinking can be enabled by changes in medical education that encourage thinking out of the box, for both physicians and scientists. Training in fundamental research – allowing young physicians-to-be to receive the equivalent of a PhD in basic research – is critical for a subset of trainees. An additional element is to provide explicit training – role modeling – in translating findings from basic research to the clinic. A course developed as part of the University of Wisconsin–Madison MD/PhD program addresses this gap (other medical schools are also moving in this direction). In the final year of clinical training all dual-degree students engage in a clinical and translational research rotation, where they work closely with a physician scientist and develop a new translational research project that addresses a gap in patient treatment [Reference Stefely, Theisen and Hanewall4]. The class provides modeling of a physician scientist’s career and also results in a translational research publication for many of the participants. Stated course goals are to train students to have a willingness to take risks, to immerse themselves in a new area inspired by patients and to integrate science and medicine.
It was this kind of focused “dabbling” that enabled Peter Huttenlocher’s discovery of synaptic pruning. Major discoveries often start with an unexpected result. It can be a challenge to recognize the importance of the unexpected, especially if it is met with skepticism. This is where persistence and resilience come into play, traits that Peter certainly had learned during his childhood in Germany and as a new immigrant in the United States. Rejection is part of being a scientist (although rare for a doctor!). Perhaps more transparent persistence training will also be built into future physician scientist training curricula.
A productive population of physician scientists does not endure if the members are not also supported deep into their careers. MD/PhD training programs and early-career funding mechanisms, such as NIH K08 awards, launch many careers. But as with many other important findings, synaptic pruning was first described by a physician scientist who was between his late 40s and early 60s; an individual with many other demands on his time, who sustained the work through gaps in research funding. The support given to physician scientists at medical schools typically includes reduced clinical loads, provision of expensive laboratory space and well-placed internal funding. The money and space are borrowed from – prioritized over – other possible uses. This model has been successful: beneficial to the medical school’s reputation and immensely beneficial to society at large. In the United States, financing for this system relies heavily on the overhead charged to research funding (for every $100,000 in funding that goes to a researcher, their university will typically collect an additional $50,000 to $70,000). But physician scientist support also draws from the larger pool of clinical revenue, and is constantly threatened by medical reimbursement schemes and accounting models that punish “inefficient” sectors. Mid- to late-career physician scientists are not likely to “vanish,” but support for their explorations only continues when it is explicitly prioritized and battled for.
Once new biomedical science connections are made and an unexpected result becomes accepted, it becomes the work of many types of scientists to take the findings to the next level. The community digs deeper to understand the molecular mechanisms and practical utilities behind a process like synaptic pruning. In some ways, decades after the discovery of synaptic pruning, work in this field is just beginning. Questions are being generated more rapidly than answers, across a wide swath of applications from childhood learning to autism to psychiatric diseases and neurodegenerative diseases. It is not a simple gene disorder that leads to these aberrant conditions. Even within synaptic elimination, multiple pathways and mechanisms influence the refined process that is so central to human health. The initial discovery is now woven into the work of an incredibly diverse community of experts. This work is being supported because of its societal benefit – but a key part of it came into being in a setting where a conducive career path, persistent determination and financial support were available to a physician scientist.
