The Discovery of Insulin

by Michael Bliss

  It particularly rankled that Macleod had stepped in and expressed himself so smoothly. More than smoothly – almost proprietarily. Instead of feeling grateful that Macleod had bailed him out, Banting decided that the professor had gone too far. The bugger had kept using that word “we” even though he had never done a single experiment, “nor had he contributed one idea of value except estimation of haemoglobin… I was the only one who gave a paper to the Physiological Section who was not asked to respond to his paper.” Who was the chairman who had not asked Banting to respond? Macleod. Come to think of it, whose name was first on the program? Macleod’s.12

  Then there were all the earlier events to consider: Macleod’s discouraging comments at the beginning of the work; that quarrel in his office early in the fall; his having said so much, using “we” all the time, at the Journal Club meeting; the interest he was taking in the work now; the good results Collip was getting (and apparently reporting to Macleod)13; and then everything he had said this afternoon. It seems to have been during his emotional turmoil after the New Haven meeting that Banting first decided Macleod was trying to take over the work, trying, in fact, to steal his results. His memory is not to be trusted for particulars, but Banting’s 1940 account of the train trip back to Toronto leaves no doubt that the meeting triggered a personal crisis:

  I did not sleep a wink on the train that night – I did not even go to my berth but sat up in the smoker condemning Macleod as an imposter and myself as a nincompoop. I decided that I must first learn to write clearly, precisely, legally, explicitly and then be able to talk convincingly, freely and unhesitatingly. I knew Macleod for what he was, a talker and a writer. Apart from his pen and his tongue he would not even be a lab. man for he had no original ideas, he had no skill with his hands in an experiment. He only knew what he read or was told and then he could rewrite or retell it as though he were a scientist and a discoverer. It was foolish to spend weeks and months working night and day at experiments and then have them told beautifully by someone else who had the art as though they were his ideas and works.14

  Macleod knew nothing of Banting’s feelings. In early January he wrote to a colleague that the New Haven meetings had been “in every way a great success, the discussions being particularly interesting.” About this time Banting began telling his friends that Professor Macleod was stealing his results.15

  II

  Back in Toronto in the New Year the work advanced about as quickly as relations among the workers deteriorated. Our documentary sources also deteriorate somewhat, for no Banting and Best notebook has been located for the period between December 22, 1921, and February 13, 1922. This may be because they did very little work together. The longevity experiment continued on dog 33, Marjorie, but relatively little is reported in any publication about her condition in January. The dog had apparently been getting a daily injection of six cc. of whole gland extract over Christmas.16 They discontinued this on January 4, measured the sugar in the dog’s urine (but not its blood sugar), and observed its general condition. When it seemed to get worse, they resumed injections on January 8.17 There are no records of Banting and Best doing any other experiments on dogs in January.

  It is not clear what role they were to have in the ongoing work now that Macleod had turned his whole lab over to the problem. A rough division of labour appears to have been worked out informally around Christmastime.18 Clark Noble was added to the group to help work with the rabbit testing and the glycogen experiments. Best and Dr. John Hepburn were to do the respiratory quotient tests (this would be Best’s M.A. thesis project) when the apparatus could be set up. Collip was to try to purify the extract to see if he could get it pure enough for clinical trials. While waiting for his respiratory quotient experiments to begin, Best seems to have done the preliminary work collecting pancreas and taking it through the first stages of extraction before passing it to Collip.19 Banting appears to have done whatever surgery the group required. He may have expected to play an important role in the impending clinical tests because he was the only practising – more or less – physician in the group.20 It was well understood that all results were being pooled in what had now become a team effort.*

  Collip was working hard and enthusiastically at his several problems. It was probably in the first week of the New Year that he made a series of vitally important observations. When he first began injecting extract into normal rabbits he had noted how very hungry they became as their blood sugars fell, some of them avidly eating paper or wood shavings. As he started using more potent batches of extract, the rabbits would occasionally go into convulsive seizures. Their heads snapped back, eyeballs protruding, limbs rigid, they would violently toss themselves from side to side, then collapse into a kind of coma, lying still on their sides and breathing rapidly. The slightest stimulation, such as a shaking of the floor, would set them off again. Sometimes lying on its side the animal’s limbs would move rapidly, as in running. The convulsions would recur every fifteen minutes or so until in most cases the rabbit died, rigor mortis setting in immediately.21

  A pre-doctoral fellow in the pathological chemistry department, O.H. Gaebler, remembered witnessing Collip’s reaction to the first appearance of the convulsions. Collip’s first thought, according to Gaebler, was that the extract must have toxic properties to cause the reaction. “On second thought, he took a blood sample and set it aside, emptied solid glucose into water, shook it about, and injected it. The rabbit recovered shortly. Subsequent analysis of the blood indicated virtual absence of glucose. It all looks simple now, but it was the most thinking per square meter per minute that I have seen.”22

  Collip had been dealing with the hypoglycemic reaction, now called “insulin shock,” which develops when blood sugar falls below certain levels. He had learned the remarkable way in which sugar clears up the condition, the symptoms quickly disappearing as the blood sugar rises again. Clark Noble learned of the same phenomenon about the same time when Joyce, the animal keeper, told him of coming in in the morning and finding rabbits dead or convulsive (in one version of this story a rabbit was stuck in the ventilating system). Noble was doing blood sugars on the animals, finding them very low, when Macleod came in, took the pipe out of his mouth, and said, “Ah, Noble, very interesting. Did you give them glucose?” He had not, did now, and half a century later thought he and Macleod had been the first to see its effects.23 Macleod had probably earlier learned what to do from Collip. Or he and Collip had both been alerted to the phenomenon and its antidote by a paper published earlier that year by F.C. Mann and T.B. Magath reporting their observations of hypoglycemic shock after hepatectomy (removal of the liver).24

  Collip was meeting Macleod almost every day for lunch now (neither of the two professors would have gone out of their way to socialize with the inarticulate and probably increasingly sullen Banting) and telling him of his very satisfying results.25 Just how satisfying they were to Collip is clear from one of the group’s few surviving letters of that winter, a January 8 report Collip made to the president of his University of Alberta, H.M. Tory, on the use of his sabbatical time:

  I will never regret having decided to spend a year near Professor Macleod. At the recent Conference at Yale he stood out most obviously as the leading man present. Last spring the old problem of diabetes was again taken up for re-investigation in his laboratory. During the summer such encouraging results were obtained by Dr. Banting and Mr. Best that in the fall the scope of the work was much enlarged. I was given the chemical side and a good part of the Physiological to push along with.

  I planned a series of experiments the results of which when obtained gave me a direct lead to the solution of the basic functional derangement in diabetes. The crucial experiment was tried out just before the Christmas break and the results were so striking that even the most skeptical I think would be convinced. I have never had such an absolutely satisfactory experience before, namely going in a logical way from point to point into an unexplored field building a
bsolutely solid structure all the way. However to make a long story short we have obtained from the pancreas of animals a mysterious something which when injected into totally diabetic dogs completely removes all the cardinal symptoms of the disease. Just at the moment it is my problem to isolate in a form suitable for human administration the principle which has such wonderous powers, the existence of which many have suspected but no one has hitherto proved. If the substance works on the human it will be a great boon to Medicine, but even if it does not work out a milestone has at least been added to the field of carbohydrate metabolism.

  Professor Graham was in my laboratory today discussing the whole matter and in the course of a few days time we hope to have had a clinical test made. If it works we will turn over in all probability the formula to the Connaught Anti-Toxin laboratories for manufacturing purposes.

  To be associated in an intimate way with the solution of a problem which for years has resisted all efforts was something I had never anticipated. I only wish that the various papers which will be published on this work were coming from Alberta rather than Toronto.26

  III

  Fred Banting’s dissatisfaction with the state of affairs in the lab had not eased. Macleod had become the quarterback of the team. Collip seemed to be doing all the running with the ball. Collip expected soon to have an extract ready to try on humans. Nothing is more evident from Banting’s notes and ideas, going right back to October 31, 1920, than his belief that the real test of his work would be the one done on a human diabetic. He was determined to participate in the first clinical trial.

  He might not. It was Collip, not Banting or Best, who had the job of preparing the extract which would be used in the clinic. Surely Dr. Banting would administer it, though. Not necessarily, for he had no standing at Toronto General Hospital, the university’s teaching hospital, where the trial would take place. In any case, he wanted the first test to be of extract he and Best had made, not Collip, and he began urging Macleod to let him try on a human the extract he and Best were using on dog 33. To clear himself to do this, or perhaps to be in on the testing of Collip’s extract, Banting apparently applied to Professor Graham at the Department of Medicine for a temporary appointment in that department to make possible his testing pancreatic extract on humans at the hospital.27

  Duncan Graham, an Ontario-born Scotsman, trained in Toronto, the United States, Britain, and Germany, had recently become the Eaton Professor of Medicine at Toronto, one of the first appointments made under the controversial “full-time” system. Graham was a tough cookie at all times, but particularly so when it came to protecting patients in the hospital wards under his control from premature experiments or investigations.28 He decided that Banting, a surgeon who was not currently in practice, had no qualifications to experiment on his patients. Banting remembered Graham saying to him, on either this or a later similar occasion, “What right have you to treat diabetics? How many of them have you ever treated?”29 Not easy questions for Banting to answer.

  Banting was nothing if not persistent, and by now must have been desperate to stop what could have only seemed more and more like some nightmarish conspiracy – Graham and Macleod were the best of friends – to push him out of the picture. We know nothing of the arguments Banting used in persuading Macleod to let him and Best try their extract on a human. Perhaps he claimed it would only be fair to give them the first chance. Perhaps he convinced the professor that the extract being used on dog 33 was not having toxic effects. Perhaps Macleod thought it wise to give Banting the reassurance he seemed to want (Macleod did not yet know of Banting’s belief that he, Macleod, was trying to push him aside, but he had noticed, after a meeting of the Journal Club at which ketonuria was discussed, that there was a “strain” between Banting and Best on the one hand and Collip on the other).30 Perhaps Banting just wore Macleod down. Whatever the reasons, Macleod relented and agreed to intercede with Graham to make possible a clinical trial of extract prepared by Banting and Best.31

  The patient chosen to receive the extract was a fourteen-year-old boy, Leonard Thompson. Thompson was a public ward patient (i.e., a charity case) in the diabetic clinic Dr. Walter “Dynamite” Campbell had founded a few years earlier at Toronto General Hospital under Duncan Graham’s supervision. Leonard’s diabetes had been diagnosed in 1919. Allen therapy was tried. By December 1921 the boy was reduced to skin and bones. As a favour to his doctor, Campbell agreed to pull strings to have him admitted to the General Hospital rather than the Hospital for Sick Children, since the latter had no diabetic clinic. To arrange this he had Thompson’s father take the boy to Duncan Graham’s office. When the father walked in, carrying the boy, Graham’s secretary, Stella Clutton, was horrified. “I’ve never seen a living creature as thin as he was,” she told me sixty years later, “except pictures of victims of famine or concentration camps.”32

  Leonard Thompson weighed 65 pounds on admission to hospital on December 2. He was pale, his hair falling out, abdomen distended, breath smelling of acetone. He was dull and listless, content to lie in his bed day after day. “All of us knew that he was doomed,” a senior medical student in the hospital recalled.33 Campbell tried various adjustments to his hospital diet, finally settling on a regimen totalling 450 calories daily. When the boy continued to worsen, Campbell told his father that unless Banting and Best’s new extract had some effect the result was inevitable. The father agreed to let them try the extract on Leonard.34

  Best made some extract by the process worked out in December. Whole beef pancreas was ground up in an equal volume of slightly acid alcohol. The solution was filtered, most of the alcohol was evaporated off in a vacuum still, the solution was washed twice with toluene, and the remaining watery solution was sterilized with a Berkefeld filter. Banting and Best tested the extract’s potency on a dog. They may have given each other injections to make sure it was safe for humans; if so, there was only a little redness in their arms. The next day they took the extracts across the street to Campbell’s clinic on Ward H of Toronto General Hospital.35

  Campbell remembered the extract as being “a thick brown muck” in appearance.36 The actual injection was made by a young house physician, Ed Jeffrey. In the afternoon of January 11 he injected fifteen cc. of the (presumably diluted) extract into Leonard Thompson, seven and a half cc. into each buttock. The quantity chosen was one-half the dose it was thought would have a definite result on a dog of equal weight.37 The only detailed description of the scene that day is in Banting’s 1940 memoir:

  We went to the hospital and remained in the corridor while a houseman injected it into the patient. We had advised Campbell concerning the time for taking samples of blood for blood sugar estimations and also concerning specimens of urine. We waited around for the first specimens and could hardly contain our suppressed excitement. This was in reality the first human diabetic to be treated. When the specimen of urine arrived we were told that it would be tested in due course. We asked for a small sample, a few drops, but we found that the whole sample was the property of the hospital, that all specimens would be done together along with samples of blood and that we would be given the results on the following day. There was a cool atmosphere about the place but there did not seem to be anything to do so we went back to the laboratory.38

  The result of the injection, as reported in a publication signed by Banting, Best, Collip, and Campbell, was as follows: Leonard Thompson’s blood sugar dropped from .440 to .320. The twenty-four-hour excretion of glucose fell from 91.5 grams in 3,625 cc. of urine to 84 grams in 4,060 cc. The Rothera test for ketones continued to be strongly positive. “No clinical benefit was evidenced.”39 A sterile abscess, caused by the impurities in the extract, developed at the site of one of the injections.40

  There is substantial evidence that one or two other patients also received injections of Banting and Best’s extract,41 but there are no records or detailed references in the publications. Some of Banting’s accounts in the 1920s suggest that Thompson wa
s the one of three patients on whom the extract had some noticeable effect. Many years later Walter Campbell told Robert Noble (Clark Noble’s brother) that Thompson was the only one on whom they had even bothered to do blood sugars.42

  Banting and Best’s extract had failed. Of course a good face could be put on the results: the 25 per cent decrease in the blood sugar, the reduction of glycosuria (and Banting put a better face on the results in his Nobel Prize lecture by talking of a “marked reduction” in blood sugar and saying that the urine had been rendered sugar-free).43 But there was the overwhelming fact that the extract’s actually very modest impact did not outweigh the reaction it caused. Even though Leonard Thompson was a very sick diabetic boy, the doctors decided not to give him further injections of Banting and Best’s extract. It was “absolutely useless for continued administration to the human subject,” Collip wrote in 1923, in a mood, which we will come to understand, of considerable bitterness. Banting himself accurately summed up the situation after January 11 when he wrote in his published “History of Insulin” that “These results were not as encouraging as those obtained by Zuelzer in 1908.”44

  Banting may not have known at the time that the records of Toronto General Hospital listed Thompson as having received “Macleod’s serum.” When he found out about it, he did not appreciate the irony.45

  IV

  Macleod had probably made a serious mistake in bowing to Banting’s pressure for a clinical test. It was a kind of crossing of the Rubicon (which in the geography of insulin was Toronto’s College Street, running between the hospital and the university), from the clinical side of which there could be no real withdrawal. That boy was in the hospital dying. The impure extract had been a little bit effective. The pressure must have increased on Collip to come up with better extract, fast. He was working very long hours.