Inside Duke Medicine

Inquiry: Turning the tides

It’s always intriguing to hear about how Duke’s world-class physicians and researchers got their start. Occasionally, life can take one’s career down unexpected paths.

Take Mark W. Dewhirst, DVM, Ph.D., for example. Trained as a veterinarian, he is now the Gustavo S. Montana Professor of Radiation Oncology and Director of the Radiation Oncology Program at the Duke Comprehensive Cancer Center. Here, he describes how he made a career of going with the flow.

I often tell my students to pay attention to unexpected opportunities. Sometimes it’s better to take the road less traveled, because in doing so, one can discover unique ways to address problems.

There are many examples of such opportunities in my career, some were fruitful and others a dead-end.  By far the best, though, came during my first post-doctoral fellowship at the University of Arizona. My program director, Eugene Gerner, met with me shortly after arriving in Tucson and asked me the following question, “Since you are a veterinarian, you must know something about physiology, correct?” I said of course, I had taken a couple of courses on this in veterinary school. However, that combined with my B.S. in Chemistry, was the extent of my knowledge. Undaunted by my lack of expertise, he suggested that I meet with the chair of chemical engineering to discuss a novel “window chamber” method that he had established to study tumor physiology in mice. This “window” is literally a glass-covered chamber that is surgically implanted in the skin of mice that permits direct observation of tumor growth and vascular function. We have gone on to use this model for many different applications after my arrival at Duke in 1984, but the one that has been the greatest challenge has dealt with why tumors have inadequate oxygen delivery – the term for this is “hypoxia”. When I entered into this field, it had been known since the 1950s that tumors were hypoxic and that this lack of oxygen led to resistance to radiation treatment.  Literally hundreds of papers had been written on the subject, with many approaches having been used to try to overcome it – none had been particularly successful. We decided that the way to overcome hypoxia had to start with understanding why tumors are hypoxic to begin with. In going through this discussion, keep in mind the concept of how tides and waves in the ocean influence how far waves travel up the beach.  What we know now is that tumors are not in a stable state of oxygen deprivation – in fact they experience continuous fluctuations in oxygen in a manner very similar to ocean tides. In 1980, the belief was that hypoxia was caused by inadequate numbers of vessels in tumors, leading to what is still termed “diffusion limited” hypoxia.  That is, cells that reside far away from blood vessels would not get adequate oxygen because it is used up by the cells that are closer to the vessels.  An implicit assumption in this paradigm was that all vessels contained some oxygen. One of our major discoveries was that some tumor blood vessels are completely devoid of oxygen, in spite of the fact that they have blood flow. The real breakthrough on this came in 2005, when Brian Sorg, Ph.D., a postdoctoral fellow in my group, established a method to visualize hemoglobin saturation in tumor microvessels.  Armed with the ability to see entire vascular networks in one image, we discovered that tumor vessels that traverse through larger tumor regions are often stripped of all of their oxygen. We can think of such regions as “low tides”.  Alternatively, tumor vessels with adequate oxygen content can be thought of as “high tides”. Another concept that was just emerging at the time that I started in the field was that some blood vessels may temporarily experience a stop in blood flow and that when this occurs, it will cause the surrounding tumor cells to experience “acute hypoxia”.  Most investigators believed that acute hypoxia was relatively rare in tumors and there was considerable debate as to how important it was therapeutically. We published our first studies relating to this phenomenon in 1996, showing for the first time that the flux of red blood cells (number of red cells traversing a vessel per unit time) in tumor microvessels was unstable and that this flux instability led to fluctuations in oxygen concentration.

Think of such fluctuations as the ‘waves’ in the ocean analogy. We subsequently proved that tumor regions could experience injury as a result of these instabilities. Still, this concept was largely ignored by most investigators. Undaunted by lack of attention to our work, we perservered, using modern sophisticated optical methods to show that the oxygenation of tumor vascular networks is most often unstable. Thus, we changed the paradigm from one in which “acute” hypoxia was thought to be a rare event to a ubiquitous phenomenon that involves large tumor regions. Finally, it is important to think about why we would care about all of this in the first place. When we started in this work, the only scientists and clinicians who cared were those who worried about the tumor cells resistance to radiation. Today, we know that the oxygenation state of a tumor contributes to altered cellular functions that drive the tumor toward a more malignant state. Thus the problems extend far beyond the simple application for radiation treatment.