Energy Expenditure of the Diaphragm or "He Thinks the Diaphragm is the Heart"*

DUDLEY F. ROCHESTER

Professor Emeritus, University of Virginia School of Medicine

	ARTICLE

TOP ARTICLE

In 1974, I had a paper about blood flow to the diaphragm published in the Journal of Clinical Investigation (J Clin Invest 1974;53:1216-1225). I never thought I would become a physician, much less a clinical investigator. I was scientifically inclined but unfocussed in preparatory school. My father wanted me to become an engineer, so I went to the Massachusetts Institute of Technology but failed to thrive in that environment; indeed I flunked out.

I subsequently transferred to Columbia College of Columbia University and profited from the teachings of Professors Jacques Barzun and Mark van Doren, among others. At Columbia I considered pursuing a non-scientific track, but when I found I was good at helping fellow students with physics and chemistry homework, I again opted for a science-oriented career.

I thought about chemistry, biochemistry, and, ultimately, medicine; albeit with trepidation. At this point, Dr. William Sharpe, a neurosurgeon and my stepfather, invited me to meet him at the New York City Medical Examiner's office, followed by lunch. Learning that I could still eat lunch after having viewed cadavers in various stages of dissection, I decided that I could hack medical school. I was accepted at Columbia University College of Physicians and Surgeons in 1950, but had to drop out for a year with a mild case of pulmonary tuberculosis. Six months in The House of Rest in Yonkers, NY stimulated my interest in diseases of the chest, but I had not yet found my niche.

I enjoyed the first two years of medical school, but still had no real idea of what I wanted to do. Another critical event occurred in second year Surgical Pathology class. My classmate Henry Rogers asked a question about our teaching model of appendicitis in rabbits. Our instructor, Dr. Virginia Kneeland Frantz, suggested answering the question with an experiment to be conducted by the inquisitive Mr. Rogers. This excited and inspired me, and I begged to participate as an onlooker and helper. For the first time I saw an experiment as a useful tool to learn, not just a laboratory exercise. Indeed, I began to think that an original experiment was something that I might actually be able to do.

Shortly afterwards I began my third year clinical clerkship in internal medicine. My preceptor was Dr. René Wegria, a cardiologist and physiologist. Dr. Wegria offered me a chance to take a research elective in his laboratory. There I learned techniques to study coronary artery blood flow, and obtained insights into determinants of myocardial blood flow and oxygen consumption. After graduation from P&S, I had my internship and assistant residency in medicine at Presbyterian Hospital in New York, under Dr. Robert F. Loeb. I spent another three months in Dr. Wegria's laboratory when I was a second year assistant resident in medicine, furthering my knowledge of circulatory physiology.

Dr. Loeb suggested that I apply for a research fellowship in Cardio-Pulmonary Physiology in Dr. Andre Cournand's laboratory at Bellevue Hospital. My mentors in the famous C-6 Lab from 1958 to 1960 were Drs. Harry Fritts and William Briscoe. Harry Fritts taught me how to use the then new electronic equipment in physiology, and I participated in his study of intrapulmonary arteriovenous shunting using the radioactive isotope of a poorly soluble inert gas, Kr⁸⁵. Bill Briscoe taught me about the distribution of ventilation and perfusion in normal and abnormal lungs, and I expanded my knowledge of indicator dilution techniques. Expanding on a technique described by Drs. Chidsey, Fritts, Hardewig and colleagues, I used a constant infusion of Kr⁸⁵ dissolved in saline to measure cardiac output by the Fick principle, the indicator was the poorly soluble radioisotope rather than oxygen. With the help of Drs. Rejane Harvey, Jacques Durand, and Jack Parker, I measured the rapid changes in cardiac output in man at the onset of exercise. A side benefit of that study was that I learned how to perform right heart catheterization.

In 1960 my career was interrupted by two years of military service at a United States Army hospital in France. It was a fantastic cultural and medical experience. I had to deal with several difficult situations pretty much on my own, and I think this helped me later to become an independent investigator. Then in 1962 I joined the faculty on the First (Columbia) Division at Bellevue Hospital.

Some of my clinical work was on the Bellevue Chest Service, looking after patients with COPD who were in ventilatory failure. Contemporary explanations for ventilatory failure included acute bronchitis and/or pneumonia, ventilation- perfusion imbalance, and, perhaps, inadequate central neural drive to breathe. However, I was quite unorthodox in thinking that some patients retained CO₂ because they were worn out with the effort of breathing. That is, I thought there might be such a thing as respiratory muscle failure.

This clinical hunch and two events in the laboratory stimulated me to pursue the role of respiratory muscles in the pathophysiology of ventilatory failure, with particular reference to respiratory muscle energy expenditure. Harry Fritts showed me how he recorded electromyograms from intercostal muscles on his sabbatical with Curt von Euler in Stockholm. Bill Briscoe wanted to look at the distribution of ventilation and perfusion under controlled conditions, and to this end studied COPD patients in the Drinker tank respirator. The point of interest to me was that the patients often breathed spontaneously in the tank, and I was able to study this later on.

From these points of departure, I began to record electrical activity of the diaphragm, first in anesthetized dogs and subsequently in awake humans. I chose this because it was easy to do, and because published reports indicated that there was a relationship between the size of the electromyogram signal and the force of contraction, and between the electromyogram and muscle oxygen consumption. That is, it appeared that the magnitude of the electromyographic signal was related to muscle energy expenditure. To better understand this I looked into ways to study blood flow to the diaphragm.

In the first such experiment, Dr. Mariette Pradel and I exposed the diaphragm of anesthetized dogs through an abdominal incision to gain access for the recording electrodes. We used a tiny needle and catheter to inject a small bolus of radioactive xenon (Xe¹³³) dissolved in saline into the diaphragm muscle, then recorded its disappearance using an external counter. The disappearance rate and the solubility of xenon in muscle yielded an estimate of blood flow per 100 grams of diaphragm muscle. In this study, doubling of minute ventilation was associated with doubling of diaphragmatic electrical activity and a one-third increase in diaphragmatic blood flow.

To estimate total blood flow to the diaphragm, I had to excise the diaphragm and weigh the muscular part. In so doing, I observed a large vein on the abdominal surface of the left hemidiaphragm that drained directly into the inferior vena cava. Since the animal laboratory was equipped with a fluoroscope, I realized that the inferior phrenic vein could be catheterized percutaneously. This would eliminate the need to expose the diaphragm and diaphragmatic blood flow could then be measured using the Kety-Schmidt technique. Moreover, one could sample blood from this vein to obtain an arterial- venous difference of oxygen across the diaphragm, and thus estimate diaphragmatic oxygen consumption, a measure of diaphragmatic energy expenditure, from the product of blood flow and the arteriovenous oxygen content difference.

I wanted to find out how much energy the diaphragm used at rest, during quiet breathing, and during unobstructed hyperventilation. To these ends, I showed (Journal of Clinical Investigation 1974;53:1216-1225) that diaphragmatic blood flow and oxygen consumption during quiet breathing were approximately 1% of cardiac output and whole body oxygen consumption, respectively. Oxygen consumption of the paralyzed diaphragm was about 60% of that during quiet breathing. When ventilation was stimulated by inhalation of CO₂, diaphragmatic blood flow and oxygen consumption increased linearly with minute ventilation, but tripling minute ventilation only doubled diaphragmatic oxygen consumption, even in those dogs that had pneumonia. I was interested to observe that during quiet breathing, diaphragmatic blood flow and oxygen extraction varied inversely, whereas diaphragmatic oxygen consumption was quite constant.

Two additional studies were an immediate outgrowth of the preceding work. Dr. Guglielmina Bettini and I showed in anesthetized dogs that inspiring through an inspiratory resistance led to much bigger increases in diaphragmatic oxygen consumption than occurred with unobstructed hyperventilation, and that diaphragmatic oxygen consumption was linearly related to the diaphragmatic inspiratory pressure-time index and to diaphragmatic electrical activity. At the highest levels of effort, diaphragmatic oxygen requirements were met by increasing diaphragmatic blood flow, but diaphragmatic oxygen extraction plateaued. In all, the determinants of diaphragmatic blood flow and oxygen consumption appeared to be quite similar to those for myocardial blood flow and oxygen consumption.

Drs. Norma Braun, Saidel Laine and I studied diaphragmatic electrical activity in patients with COPD as well as a few others with kyphoscoliosis or morbid obesity. We recorded diaphragmatic electrical activity using esophageal electrodes fashioned from a cardiac pacemaker catheter, and based on the previous animal experiments, considered diaphragmatic electrical activity to be a surrogate for diaphragm energy expenditure. We chose the subjects because they were known to have an elevated work of breathing, and they also were in chronic ventilatory failure. We studied the patients during spontaneous quiet breathing, and again in the Drinker tank respirator. We used this mode of ventilation because I remembered Briscoe's experiment, and because that was how we treated ventilatory failure at Bellevue Hospital in those days. For most patients, there was a prompt and nearly complete cessation of diaphragmatic electrical activity; this was associated with a concomitant relief of dyspnea that was unrelated to any change in Pa_CO2. This was the first demonstration of two points: a ventilator could capture breathing, and cessation of diaphragmatic contraction was associated with relief of dyspnea.

To recapitulate, having tuberculosis got me interested in lung diseases, and caring for patients with COPD got me interested in respiratory muscles. Working with Rene Wegria taught me about myocardial blood flow and energy expenditure, and I transferred these ideas from the blood pump to the air pump. Harry Fritts opened my eyes to a new way to study respiratory muscles, electromyography. My isotope work with Harry Fritts and Bill Briscoe led me to become competent with various indicator dilution techniques, so when the anatomical opportunity presented itself, I knew what to do.

There are, it seems to me, three morals to this story. First, exciting stimuli that later lead to visions and insights appear randomly and unexpectedly. Second, one can develop a basic set of skills and knowledge while training on other investigators' projects. Third, with the help of a vision and insights, one can build on the basics to develop new techniques that lead to new knowledge that may have clinical and physiological significance.

	Footnotes

Correspondence and requests for reprints should be addressed to Dudley F. Rochester, M.D., 103 Shawnee Court, Charlottesville, VA 22901.