The Discovery of Insulin
Page 12
Partially depancreatized dogs a la Allan [sic] – control diet to see if the tolerance of a partially diabetic dog can be improved by extract – If this occurs a human might be improved by a “course of treatment”.8
On the back of the card are notes summarizing an Allen article. The importance of the idea, which was not tried, is its underlining of Banting’s interest in the possibility of using the extract on humans. He was never a disinterested physiologist looking for the internal secretion of the pancreas, but always a practical researcher looking for a cure for diabetes in humans.
Banting’s practical desire to get on with the work, combined with his relative disinterest in medical scholarship, his weak background knowledge, and his inexperience at research, all militated against a careful, thorough study of the literature, including the publications of the others who had gone after the internal secretion. If Banting and Best were aware of the work of Zuelzer and E.L. Scott, for example, they either did not bother to read their articles, which are not listed on their surviving index cards, or they decided there was no need to cite this work in their early publications. They did, however, come across some of the results just published by Nicolas Paulesco.
Sometime between October and December 1921, Best read Paulesco’s July 23 publication on the action of his pancreatic extracts. Best summarized its contents on one of the index cards.9 Paulesco’s extract, he noted, lowers the blood and urinary sugar of diabetic animals and definitely reduces the acetone bodies in the urine. Its effect varies in duration and magnitude in proportion to the amount injected. Paulesco also “proves” that his extract lowers the blood sugar of a normal animal. Best also thought it germane to note, however, that Paulesco reported normal blood sugars in his dogs as low as .044 per cent and obtained hyperglycemic readings in his diabetic dogs no higher than .20 per cent. Both figures, Banting and Best knew, were considerably out of line, and this may have cast doubt on the Romanian’s methods. As well, Paulesco’s animals had been under a volatile anesthetic, chloroform, during his experiments, with the extract injected just after the pancreatectomy; Best may well have realized that the anesthetic’s effect on blood sugar would throw all experiments with extracts into question. Moreover, Paulesco did not report either the volume of extract injected or the volume of urine excreted. The index card suggests that Best did not find Paulesco’s paper particularly impressive.
One of the least impressive aspects of Paulesco’s work, according to Best’s summary, was that Paulesco “states that injections into jugular, portal, or mesenteric veins works, but into peripheral veins ‘no bon’.” Like most English-speaking Canadians then, and now, Best had a very rudimentary reading knowledge of French. In making this note he had misunderstood a key sentence in Paulesco’s article. The sentence reads: “Les mêmes effets, c’est-à-dire une diminution ou même une suppression passagère de l’hyperglycémie et de la glycosurie, s’observent aussi lorsqu’on injecte l’extrait pancréatique, non plus dans une veine périphérique, mais dans une branche de la veine porte, par example: dans une veinule mésaraïque ou dans une veinule splénique.”10 Best almost certainly mistook the phrase, “non plus,” meaning “not only,” for, as he wrote it, “no bon,” or no good.
The misreading added to the coolness of the note’s summary of Paulesco’s work. Banting and Best do not seem to have given much further thought to Paulesco. They did not look up his other publications, and if they reread his work the only result was to compound their misunderstanding by somehow concluding, incorrectly, that Paulesco’s experiments showed a less marked effect from second injections.11 Consequently their brief reference to Paulesco in their first publication in 1922 grossly distorts his work.*
Apart from the mistranslation of French and the other problems Banting and Best may have perceived with Paulesco’s work, there are two other possible reasons for the neglect of Paulesco. The most probable is that Paulesco was neglected because, as Banting’s cursory reading and notes indicate, the pair neglected almost everyone who had worked on pancreatic extracts before them. Wasn’t it obvious that all the precursors had failed to find the internal secretion? If others were working right now on extracts – a probability that Banting and Best, like modern students infatuated with ideas they think no one else has ever had, may not have clearly realized -they surely were not doing as well as the Torontonians. For if they were doing as well, or just a bit better, they would be about to start experimenting on humans. In contrast to the history of much recent scientific discovery, such as the structure of DNA, there is no evidence that the workers in Toronto thought they were in any kind of race or competition with outsiders to be the first to get to the internal secretion. Their intellectual, emotional, and experiential bias was entirely towards the goodness of their own experiments. This impeded their paying attention to the goodness of anyone else’s, and encouraged misreadings and mistranslations.
(The result of the sloppiness was harmful not only in creating later misunderstandings and resentments by Romanians and others, but also in the missed opportunities it involved for Banting and Best to plan a rational course of experiments. Had they thought about Paulesco carefully, for example, they might have decided to try their extract on normal animals and to measure its impact on ketonuria in diabetic ones, as he had done. Had they studied earlier workers they might have developed experiments to check for toxic effects of their extract. Had they presented a clear, well-thought-out and productive experimental plan to Macleod in October or November 1921, much later confusion and bitterness about credit might have been avoided.)†
The second probability, partially but not fully contradicting the first, is that Banting and Best saw Paulesco as being unimportant in surmounting the problem at hand, which was to get beyond the stage they had all – Paulesco, Kleiner, Banting and Best – reached of having extracts that suppressed hyperglycemia and glycosuria. Impressive as the blood sugar evidence was in all three labs, it still did not prove that an internal secretion had been captured. Substances might lower blood sugars and reduce sugar excretion without necessarily permitting the diabetic’s system to metabolize its food. Such a possibility must have been on Banting’s mind when he made an extensive note on one of his index cards of an article listing a dozen conditions that could lower blood sugar. Among these conditions were several factors, such as shock, moribund states, and the injection of foreign proteins, which could have affected his and Best’s dogs.12 There had to be other experiments, Banting might have reflected, other approaches beyond simple blood sugars and urine tests, to nail down the internal secretion. Paulesco’s work, which in any case was less impressive than Kleiner’s, appeared to be of no help at this stage of the problem.
Banting kept flirting with the idea of pancreatic grafting. A card dated October 4 has a note on possible kinds of grafts. On Wednesday, October 19, Banting wrote C.L. Starr to say that he had permission from the university’s Surgical Research Committee (recently set up by Macleod to head off the anti-vivisectionists) to do “an original investigation of the viability of (1) autogenous, (2) homogenous, (3) heterogenous grafts of pancreatic tissue.” He and Best planned to begin work the next day, October 20. “We believe that such an investigation will be of great value,” he wrote, “in ascertaining the clinical uses of substances contained in such tissue in the treatment of diabetes.”13
Banting and Best had dinner together that night.14 Their notebooks do not show any work done on October 20 or any grafting experiments ever attempted. Instead, on October 24 they began yet another round of injections of their degenerated gland extract into diabetic dogs. The most likely explanation of this sudden change of plan is that Macleod strongly advised the pair to stick to their extract. This may have been the time – alluded to directly in oral sources and ambiguously in the documents15 – when Macleod told Banting and Best, in effect, that their results were just not good enough and they would have to repeat the experiments to get more and better ones. In view of the likelihood that Banting�
��s proposed grafting experiments would have gone nowhere, this was sound and useful advice.
So more extract was made from degenerated pancreas (not very degenerated, though, for the waiting period after ligation was getting shorter and shorter; in one case it was only eighteen days) and another total pancreatectomy was done to make a dog diabetic. Things did not go well. The dog, number 21, was given twenty cc. of extract at 2:00 p.m. on October 26. At 2:30 it vomited. Excretion of urine almost stopped. The dog’s rectal temperature at 3:00 p.m., the only temperature record in all of Banting and Best’s notebooks, was a feverish 40C. Neither the first nor a second twenty cc. of extract had any significant impact on its blood sugar. Banting and Best stopped experimenting. The dog died suddenly after a drink of water on the 30th. The autopsy is recorded as showing a ruptured duodenal ulcer. It may have been brought on as the result of a slip of Banting’s scalpel during the pancreatectomy.
After most of another week off, Banting and Best began again on November 4, running a control test on one of the ligated dogs to study how it responded to sugar injections. A sugar-plus-extract experiment on yet another depancreatized dog, 26, looked good, except that 26 was a very sick dog. It died on November 10, another victim of a duodenal ulcer and extensive internal bleeding. This is the last dog written up in Banting and Best’s first paper; they refer euphemistically to the “early termination of the experiment.” The last two experiments had been particularly unsatisfactory, due perhaps to poor surgical technique injuring the animals. Perhaps Banting’s heart was not in it.
III
Macleod asked Banting and Best to talk about their work to a gathering of university students and staff at the Physiological Journal Club on November 14. Banting may have been pleased at the thought of presenting his results, which were exciting regardless of the most recent experiments, on his thirtieth birthday. A notice of the meeting lists Banting and Best as speaking on the subject of “Pancreatic Diabetes.”16
The importance of this talk has been grossly exaggerated, to the extreme of the Historic Sites Board of Canada stating on its major commemorative plaque on the University of Toronto campus that the meeting marked Banting and Best’s “public announcement of a therapy for use in the treatment of diabetes mellitus,” hence the discovery of insulin. This is nonsense. The session was an informal presentation to a semi-private university group. Banting and Best had not yet finished writing their first paper describing their experiments. When it was finished, a week or more later, the authors concluded that it was “very obvious” that the results of the experiments through November 10 “do not at present justify the therapeutic administration of degenerated gland extracts to cases of diabetes mellitus in the clinic.”17 At the Journal Club meeting on November 14 Banting and Best gave a preliminary report to some interested colleagues and students, possibly a few outside visitors, on their work-in-progress.
Banting liked to think and write late at night. At 12:15 a.m. on November 14 he noted in a diary the coming of his thirtieth birthday. His ambition was to write one article per year for five years – “I have my first paper well under way.” The questions of the moment, he thought, were whether to study for his fellowship in the Royal College of Surgeons, whether to leave surgery for experimental physiology, and whether to marry. “Time alone I suppose will only solve these problems. – At the present it behooves me to study & work at the internal secretion of the pancreas. & if possible isolate it in a form that will be of use in treating Diabetes.”18
There are no records of the presentation or the discussion at the Journal Club meeting later that day, but Banting’s and Macleod’s 1922 statements disclose two important consequences of the meeting. The first arose from a misunderstanding. Banting apparently had asked Macleod to introduce them. Best would show charts of their dogs while Banting talked about their work and its relation to that of other investigators. To Banting’s dismay, Macleod in his introduction said all the things he, Banting, had planned to say about earlier research. Banting was inexperienced as a speaker, nervous and inarticulate, and could not have adjusted easily to the surprising introduction. His natural reaction to the misunderstanding would have been to become angry – and to notice, as he stressed in writing about the meeting a year later, how often Macleod was using the pronoun “we” in describing the work. His mood could not have been helped after the meeting when he learned that students were talking of the remarkable work of Professor Macleod.19
Banting chose to say nothing to Macleod about his feelings. It was two months and many events later before Macleod heard of Banting’s sensitivity about the Journal Club meeting. “Had I been told of this attitude of Banting at the time,” he lamented in September 1922, “it would have served to warn me of his peculiar temperament and of his entirely unwarranted suspicions….”20
The more constructive consequence of the meeting came as the result of a suggestion by Dr. N.B. Taylor in the discussion after the presentation. He thought that a convincing demonstration of the extract’s effect would be to show that regular administration of it could prolong the life of diabetic dogs. When Banting, Best, and Macleod discussed the future course of the work the next day, Macleod suggested they try this longevity experiment. They agreed.21
IV
But that agreement brought what had always been a bedevilling question to the fore. Where would the extract come from for such an experiment? The duct-ligation method of obtaining extract was slow and cumbersome and expensive at the best of times. It involved delicate operations, many dogs, and a four-to-seven-week waiting period. At this time, November 15, 1921, when there was at most only one duct-ligated dog on hand, Banting and Best faced the depressing prospect of being able to do next to nothing for a month or two while waiting to obtain extract. The one short-cut they had tried, injecting secretin to exhaust pancreas of its external secretion, had not worked at all satisfactorily. This problem with the supply of extract, then, was a crippling limitation on the work. Indeed, in a larger sense the supply problem threatened to be an over-riding barrier: there would probably never be a practical clinical use of the internal secretion of the pancreas if duct ligation and degeneration was the only way of capturing it. There had to be a better way of obtaining pancreatic extract.
Banting thought about these problems late into the night on the 15th. His reading had given him some clues. Laguesse had found that in the pancreas of new-born and foetal animals the islet cells are more plentiful in relation to the acini than in mature animals. This should mean, Banting reasoned, that their pancreases produced abundant quantities of internal secretion. Perhaps he and Best should try to obtain an extract from new-born animals. But then, Banting realized, there might be a possibility that in the foetal pancreas, as opposed to the new-born, the internal secretion might exist and the external secretion not be found. Other internal secretions, such as adrenalin, were present in early stages of foetal development. Since digestion does not begin until after birth, it was likely that the external secretion was not potent in the foetus. Further interesting evidence was an article about Carlson’s work in Chicago in 1911 in which he and Drennan found that a depancreatized pregnant dog did not become diabetic until after delivery. They postulated that the foetal pancreas must supply the necessary deficiency. If it did have that kind of potency, and did not contain destructive pancreatic juice, perhaps the foetal pancreas could be used to make an effective extract. This idea “presented itself” to Banting at about 2 a.m. in the morning of November 16.22
Banting first thought of obtaining foetal pancreases by producing abortion in dogs. Then, surely realizing how cumbersome that procedure would be, the farmer’s son remembered that growers often bred their cattle just before slaughter to make them better feeders and fatter. There would be plenty of calf foetuses available at the slaughter-houses. The next morning he and Best went to the William Davies Company’s abattoir in northwest Toronto, cut out the pancreases from nine calf foetuses, and brought them back to their lab.
V
For you, the reader, perhaps starting to tire of dogs and extracts, the first month of the “new era,” described in the rest of this chapter, is the most technical part of the history of the discovery of insulin. I cannot present an accurate record without the technical detail, and you would be unwise to skip this important material. Try not to worry about the individual dogs and their blood sugars, but instead notice the pattern of development of the research problem and the achievements and failures of Banting, Best, and the others.
The longevity experiment was begun on dog 27, which was given one or two injections of foetal calf extract daily. A second depancreatized dog, 33, was used to test various doses of the extract, which Banting and Best were now determined to make in its most potent and effective form. No name was being used for the extract, “Isletin” having not been mentioned since early August. The job at hand was to capture the “active principle” of pancreatic extracts in some form that could eventually be tested clinically.