IV
As soon as it was realized that the pancreas controls diabetes, attempts began to treat the disease, literally, with pancreas – just as diseases of the thyroid were being treated with thyroid. Minkowski was the first of many researchers to try to restore the pancreatic function to diabetic animals (others experimented on human diabetics) by preparing and administering extracts of pancreas. The extracts could be made in a variety of ways; they could also be administered in a variety of ways, although the most obvious were orally and by injection. The important observation would be of sugar in the urine. If an extract reduced glycosuria it might be potent. It might contain the internal secretion; indeed, it might supply the proof that there actually was an internal secretion, for until its effect could be practically demonstrated, the internal secretion of the pancreas was merely a good-looking hypothesis.
The results of the early experiments with pancreatic extracts were mixed, tending towards the negative. Some extracts had no effect; some had decidedly harmful effects, throwing the animals into shock or worse. Others had temporary sugar-reducing effects that were more than cancelled out by harmful side-effects – so much so that it was impossible to tell whether it was the extract or its toxic effect on the system that was the true cause of the reduction in glycosuria. If an extract caused kidney failure, for example, it might be changing the contents and quantity of the urine without affecting the diabetic condition at all. A few researchers did report encouraging results with extracts, but others who tried to repeat their work got discouraging results. It will never be known precisely how many researchers tried giving pancreatic extracts to diabetic animals and humans. Estimates run to more than four hundred. It was an easy experiment for even a country doctor to try, but if the results were not encouraging many would decide there was no point publishing. As it was, there was no shortage of publication, on every conceivable aspect of the problem of diabetes and the pancreas, it seemed. In 1910 Opie complained that the literature on diabetes was voluminous. A few years earlier Lydia Dewitt estimated that more thought and investigation was going into the islets of Langerhans than any other organ or tissue of the body.11
Despite the discouragement, the search for a workable pancreatic extract continued. Perhaps the problem with extracts was that somehow the pancreas’s external secretion, or the tissues producing it, destroyed the internal secretion in the extirpated organ. Laguesse suggested using extracts made from foetal pancreases, because it seemed that the islet cells develop well before the acinar cells in gestation. If the experiment was tried, it failed. So did a number of other experiments involving fish. In certain species of fish the islet tissue had been found to be anatomically distinct from the acinar tissue, making it possible, it seemed, to get an extract which was more purely an extract of the islets of Langerhans. Between 1902 and 1904 two Scots researchers in Aberdeen, John Rennie and Thomas Fraser, fed an extract of boiled fish islets to four diabetic patients. After inconclusive results, including a toxic reaction when they tried to inject the extract into a fifth patient, they gave up.12
The most persistent and important of the early extractors was Georg Ludwig Zuelzer, a young internist in Berlin who in the early 1900s became interested in the theory that diabetes was actually caused by adrenalin. Experimental evidence that large doses of adrenalin could produce glycosuria convinced Zuelzer that the function of the internal secretion of the pancreas was simply to neutralize adrenalin in the system. He decided to try to prove this by injecting an extract of pancreas into rabbits along with adrenalin. When no glycosuria developed, Zuelzer was encouraged to go on and see if his extract could reduce diabetic symptoms in depancreatized dogs. When it appeared to reduce the sugar excreted in the urine of two diabetic dogs, Zuelzer was encouraged to go further.
Dying diabetics were hopeless cases, so it must have seemed that nothing could be lost in experimenting on them. On June 21, 1906, Zuelzer injected eight cubic centimetres of his pancreatic extract under the skin of a comatose fifty-year-old diabetic in a private clinic in Berlin. The next day he injected another ten cc. Whatever effect the extract was having on the patient’s glycosuria could not be measured, for the man had lost control of his bladder and was wetting his bed. What was clear was that the patient seemed to be coming back from the edge of the grave. His overall condition improved, his appetite returned, and his severe dizziness disappeared.
But there was no more extract. The patient sank into deep coma on June 30 and died on July 2. What Zuelzer had seen was tremendously encouraging, a diabetic momentarily pulled out of coma. “Whoever has seen how a patient lying in agony soon recovers from certain death and is restored to actual health will never forget it,” he wrote years later just after insulin had been discovered in Toronto. He was almost certainly referring to his first experience with his own pancreatic extract, which he named “acomatol.”13
Zuelzer had immense practical difficulties carrying out his experiments. It was hard to get a supply of pancreases, for example. Workers at local slaughterhouses thought the doctor who wanted them to give him fresh sweetbreads for medical research must be a little crazy. The extract was not at all easy to make, and had a frustrating tendency to lose its potency (Zuelzer tested his batches on rabbits, measuring the potency by the amount of extract needed to neutralize the sugar-creating effects of a unit of adrenalin). But there were those early results, and it was obvious that a workable pancreatic extract would be a wonderful thing. When Zuelzer approached the Schering drug company with his idea they offered him financial support and technical help and applied for patents on his methods. By the summer of 1907 he was ready to try again on humans.
The extract produced the amazing effect of completely suppressing for a few days glycosuria and acidosis in a twenty-seven-year-old man. Other diabetics – a six-year-old, a thirty-five-year-old, and two sixty-five-year-olds – had their symptoms dramatically relieved by acomatol. (Some others, it appears, did not; Zuelzer reported only the most interesting cases.) On the other hand, in every case after the first two there were serious reactions to the injections: vomiting, high fevers, sometimes convulsions. Knowing that his preparation was not yet a practical therapy, Zuelzer was still confident enough to publish his results in 1908. He came to the triumphant conclusion “that it is possible through the injection of a pancreatic extract to eliminate the excretion of sugar, acetone, and acetoacetic acid by a diabetic without making any changes in the patient’s diet.”14
These exciting findings caught the attention of a worker in the clinic directed by Minkowski in Breslau. J. Forschbach obtained samples of Zuelzer’s extract and tested it on three dogs and three humans. His verdict was negative. Yes, Zuelzer’s was the first pancreatic extract to suppress glycosuria in both the short and the long run. But it did so at the cost of severe toxic side-effects, especially fever, so severe that Forschbach stopped his human experiments for fear of doing permanent damage to achieve only temporary relief. “It will be difficult to convince a patient who has been made severely ill by a single injection,” he wrote, “that this result was connected to a significant beneficial effect upon his diabetes.” Forschbach was fairly convinced, especially after some impotent extract caused no ill effects in one case, that the cause of the potency and the cause of the side effects were the same. So there was no future in it. Forschbach’s 1909 paper on his tests of Zuelzer’s extract was decidedly discouraging, and must have been more so because of Forschbach’s association with the great Minkowski himself. The giants in the field had passed judgment.15
At about the same time the Schering company decided that the results did not justify the cost of the work and withdrew their support. Zuelzer’s application for a grant of 500 marks (about $125 at that time) to spend six weeks at a zoological station seeing if he could make an extract from those interesting fish pancreases was rejected by the University of Berlin. Zuelzer was evidently neither wealthy nor well-connected, an outsider in Berlin medical circles now left on his own with his e
rratic extract. He published nothing more about it.
In fact he carried on, a big, shambling doctor hawking his idea and his method from one drug company to another, bribing slaughterhouse workers to give him pancreases. In 1911 the big Hoffman-La Roche chemical firm put him back in business, funding a small lab and some co-workers. The next year Zuelzer took out an American patent on his “Pancreas Preparation Suitable for the Treatment of Diabetes.” The patent was wishful thinking, though, for there were still problems with the extract. When the first big batch was made in the new lab, from 100 kilograms of pancreas, the animals on which it was tested went into severe convulsions. Zuelzer had never seen anything like this before. He decided it was the old story of toxic side-effects, perhaps caused this time by the use of copper containers, and threw out the batch. After more problems with ineffective extracts from horse pancreas, Zuelzer was ready for another round of experiments with what looked like promising material in the summer of 1914. When the war began, the hospital he was working in was turned over to the military. Georg Zuelzer was called to the front.16
The main effect of Zuelzer’s work was probably to set back the search for an effective pancreatic extract. His published findings, plus Forschbach’s report, seem to have convinced researchers of the impossibility of the enterprise: even if you did get an extract with anti-diabetic effects, whatever good effects it might have would be more than cancelled out by its bad effects. Experienced scientists had learned to be cautious, and it became something of a mark of professional prudence to qualify any findings about pancreatic extracts.*
A classic example of this learned cautiousness was the treatment of a student’s work at the University of Chicago in 1911–12. The student, E.L. Scott, who had been deeply affected by the death of a friend from diabetes, took up the search for the internal secretion as the research project for his master’s degree. He reasoned that previous failures had been caused by the external secretion, the powerful proteolytic (protein-destroying) enzymes, destroying the internal secretion. Perhaps the answer lay in getting rid of all traces of these enzymes. One way might be to ligate the ducts of the pancreas; this apparently would cause the tissues producing the external secretions to atrophy; from the remaining tissue an extract could be prepared and then tested. (Lydia Dewitt had already tried this method in 1906, but had tested her extract only on test-tube solutions, not living animals.) Scott abandoned the idea as impractical when, working under primitive conditions in a very hot summer, he found that it was almost impossible to get a pancreas to atrophy after ligation. Instead he turned to alcohol, a fairly common solvent and one that Zuelzer had also used in the preparation of his extract, to do the same job. Using extracts which had gone through various stages of development through mixing the pancreas with alcohol, filtering it, treating the residue, and other chemical procedures, Scott found one formula that gave encouraging results on three of the four diabetic dogs he treated with it. Not only did their sugar excretion diminish, but “if one dared to say it,” Scott wrote, the dogs “seemed even brighter for a time after the injection than before it.”
Like Zuelzer before him and others afterwards who observed the subjective signs of improvement in diabetic animals and patients, Scott was convinced that he had been successful. The first two conclusions of his master’s degree thesis were:
1st. There is an internal secretion from the pancreas controlling the sugar metabolism.
2nd. By proper methods this secretion may be extracted and still retain its activity.
Scott’s thesis adviser, the noted physiologist Anton Carlson, did not share his student’s confidence. Having just read of recent work by Hédon questioning the effectiveness of pancreatic extracts, Carlson worried that Scott had not sufficiently controlled his experiments. He urged that the conclusions be rewritten, probably supplying the new wording himself:
It does not follow that these [good] effects are due to the internal secretion of the pancreas in the extract. The injections are usually followed by a slight temporary rise in the body temperature, and this may be a factor in the lowered sugar output. Physiologists are not agreed as to whether the internal secretion acts by diminishing or retarding the passage of sugar from the tissues into the blood, or by increasing the oxidation of the sugar in the tissues. The pancreas extract may decrease the output of sugar from the tissues by a toxic or depressor action, rather than by a specific regulatory action of the pancreas secretion….The work is being continued in the hope of clearing up these points.17
Despite his conservatism, Carlson urged Scott to continue the research and work out his “salvation or damnation along the pancreas extract line…. There is something ahead in that line – possibly both shoals and open water. Puzzle: find the channel.” Scott tried half-heartedly, attempting to buttress his urinary sugar results with studies of his extract’s effect on the blood sugar of cats. He reported the “very surprising” result that it caused an increase in their blood sugar.18 Having struck a shoal, Scott veered away from pancreatic extracts to study problems relating to blood sugar.
Before giving up the work, Scott chatted about it with some of the other experts in the field. One of these was a professor at Western Reserve University in Cleveland, Ohio, John James Rickard Macleod. Macleod was a Scotsman, trained in Aberdeen, Germany, and London, who had emigrated in 1903 to take his American appointment at the age of twenty-seven. He had been working for several years in the area of carbohydrate metabolism. A competent researcher and a prolific writer and synthesizer of current knowledge in physiology, Macleod was particularly knowledgeable about the literature in his field. The only knowledge we have of his discussion with Scott is that it began with a consideration of how to cure Scott’s child’s diarrhoea. When the talk turned to pancreatic extracts, Macleod may have discouraged the younger man, for about this time he was working on his own main research contribution to the search for the internal secretion of the pancreas. Macleod was able to show that the findings of two leading Britishers, Knowlton and Starling, who thought they had a pancreatic extract which assisted the heart of a diabetic dog to utilize sugar in the blood, were not repeatable.19
Knowlton and Starling’s joined Scott’s and Zuelzer’s in the list of apparently ineffective pancreatic extracts. Two young Americans, John R. Murlin and Benjamin Kramer, continued to fiddle with pancreatic extracts similar to Knowlton and Starling’s, but their work led them off into examinations of the influence of alkaline solutions on metabolism.20
Macleod summarized the state of the search for an internal secretion in his 1913 book, Diabetes: Its Pathological Physiology. After due deliberation he concluded that there was an internal secretion of the pancreas, but suggested several reasons why it might never be captured in a pancreatic extract. The powerful pancreatic juice might destroy it; there might be no reserves of it in the pancreas to be captured by extraction; or it might exist in the pancreas only in latent form and not be activated until secreted into the blood. Macleod’s own interest and his work tended to be on the behaviour of blood sugar rather than pancreatic extracts. He thought the most convincing proof of the existence of an internal secretion came in Hédon’s early work (now questioned by Hédon himself) using grafts of pancreatic remnants to show that a small, isolated portion of the pancreas could stave off diabetes.21
In 1913 Dr. Frederick Allen pronounced what seemed to be the epitaph of a generation’s attempts to treat diabetes with pancreatic extracts: “All authorities are agreed upon the failure of pancreatic opotherapy in diabetes….injections of pancreatic preparations have proved both useless and harmful. The failure began with Minkowski and has continued to the present without an interruption….The negative reports have been numerous and trustworthy.”22
V
Frederick Madison Allen wrote with particular authority. Born in Iowa in 1876, trained in medicine in California, Allen had come east to do medical research, drifted into a poorly paying fellowship at the Harvard Medical School, and found himself work
ing on problems of sugar consumption. The study turned into three years of intensive research concentrating on diabetes. Most research is reported upon in journal articles. Allen’s was not. His first publication, subsidized by his father, was a remarkable 1913 volume entitled Studies Concerning Glycosuria and Diabetes. “Its spirit is that of an enlarged journal article,” Allen wrote in the introduction, claiming it was a book in which he hoped to give “simplicity and order” to the study of diabetes.23 The reader who waded through the following 1,179 pages – in which Allen set his own research, involving experiments on more than two hundred dogs, the same number of cats, and assorted guinea-pigs, rabbits, and rats, in the context of everyone else’s research (his bibliography contained approximately twelve hundred listings) – knew that the subject was anything but simple and orderly, except possibly in the head of Dr. Allen. Both the research and the book were prodigious achievements in themselves, and even more significant for the revolution in diabetes therapy that flowed from them.
Most researchers created experimental diabetes in animals by taking out the whole pancreas. Allen’s approach was to remove a large part of the pancreas, about 90 per cent in dogs, but leave the rest. He thus created a state of mild diabetes in animals which was probably much closer to the diabetes most humans experienced than was the severe, quickly fatal diabetes arising from total pancreatectomy. Although he tried some experiments involving pancreatic extracts, Allen’s chief interest was in the effect of diet on the diabetic animals. What kinds of diet would enable an animal with a partial pancreas to keep metabolizing his food without becoming more diabetic? What kinds of diet were harmful to animals with these crippled pancreases, making the diabetes worse?
The Discovery of Insulin Page 4