It works! Now what?

October 1921 began with considerable excitement and confidence for Banting, Best and Macleod. The success of the pancreatic extract work so far led to lots of new questions, prompting Banting and Best to explore many different ideas. Macleod, however, tried to rein them in, advising they stay focused on solving the outstanding problems associated with confirming the effectiveness of the pancreatic extract in controlling diabetes. One of those outstanding problems was somehow overcoming the limits they had in the supply of extract prepared from dog pancreas tissue. No one wanted to have to rely on dogs for the work. 

Exploring a potential alternative to dog pancreas opened up through Banting’s connection with Dr. John G. FitzGerald, director of U of T’s Connaught Antitoxin Laboratories. During the First World War, Connaught had expanded to include a farm site at what today is the northern limit of Toronto. Smallpox vaccine was prepared at the Connaught “Farm” and calves were involved in its production. Perhaps the pancreatic duct of a calf could be ligated to isolate the internal secretion, the researchers reckoned. On October 4th, Banting and Best visited the farm to try a pancreatic duct ligation procedure on a calf. However, the operation was unsuccessful as the calf died from the anaesthetic. The experience, although frustrating, fuelled further ideas.

Photo of the official opening of the Connaught Antitoxin Laboratories and University Farm on October 25, 1917. Courtesy of the Sanofi Pasteur Canada Archives, and “Connaught Laboratories & The Crucible of World War I” by Christopher Rutty.

Banting and Best spent considerable time in October doing research in the medical library, reading journal articles about diabetes and the pancreas, and jotting down ideas on index cards. Unlike Best, Banting was not the typical academic medical researcher. He had a practical streak and was driven by a desire to test the extract on human diabetics.

The library research helped Banting and Best prepare their first presentation about their work and their initial publication. They gave a talk on November 14th (Banting’s 30th birthday) to the Journal Club of the Department of Physiology and a small audience of students, faculty and staff. The presentation seemed to go well, with an important suggestion made during the discussion by Dr. N.B. Taylor, who thought a longevity experiment with the extract given regularly to a diabetic dog over an extended period of time would be useful. But where would the extract come from? The duct-ligation process was time consuming and expensive on several levels, not least of which involved the reliance on dogs. There had to be a better way.

On November 16th, at about 2:00 am, Banting woke from a fitful sleep with an idea, sparked by a memory about the cattle he had helped to slaughter on his father’s farm in Alliston, Ontario. He had recently read that the pancreatic islets cells were more richly concentrated in newborn infants because digestive juices such as trypsin, an enzyme, had not yet poisoned the internal secretion. From his farm experience, Banting also knew it was best to breed cattle before slaughter, since pregnant cows were more voracious feeders and would gain more weight. Banting was thus inspired to collect the pancreas tissue from foetal calves retrieved from freshly slaughter cows, rather than trying to induce abortions in dogs. The next day, Banting and Best went to a local slaughterhouse and returned to the lab with nine pancreases from foetal calves and immediately prepared extract. They administered the extract to Dog #27 and its blood sugar fell from .30 to .20 within 20 minutes. Additional injections brought the level down to .08, and tests 24 hours later showed the dog’s urine was sugar-free. This new approach meant that no more duct ligations would be needed. What’s more, there would be no more shortages of extract. Abattoirs could supply all the foetal pancreases Banting and Best would ever need.

Dog #33 standing on the roof of the University of Toronto Medical Building, November 1921. University of Toronto, The Discovery and Early Development of Insulin.

With a steady supply of extract, Banting and Best next focused on a longevity experiment, the focus of which was Dog #33, otherwise known as “Marjorie.” They removed her pancreas on November 17th. She received her first injection of foetal calf pancreas extract on December 6th. In the meantime, Banting and Best worked on improving the purity of the extract by adding a final filtration step using unglazed porcelain (also known as a “Berkefeld filter”), which trapped bacteria and assured sterility, although it reduced potency. Feeling additional confidence, and a certain boldness, on November 23rd, Banting and Best took a daring step and injected each other with small 1.5 c.c. doses (about 1/3 teaspoon) of extract. This was not an uncommon practice among doctors and researchers who were unwilling to test a new drug to a patient unless willing to first test it on themselves. Such self-experimentation does continue, but more rigorous codes of conduct have evolved since the late 1940s. Banting and Best’s brief note only said, “No reaction,” which meant no harmful effects. They then focused on writing the manuscript for their first publication, “Internal Secretion of the Pancreas,” which would appear in the February 1922 issue of Journal of Laboratory and Clinical Medicine.

Just as they submitted their article for review, Macleod received a call for papers for an American Physiological Society meeting planned for the end of December in New Haven, Conn. Macleod suggested to Banting that a report on the work be presented. Word about the encouraging diabetes research in Toronto was starting to get out to diabetes specialists, likely through Dr. Lewellys Barker, a Canadian and U of T graduate working in a U.S. hospital who had visited Toronto and attended the Journal Club talk. Barker had mentioned Banting and Best’s work to Dr. Elliott Joslin, who then wrote to Macleod looking for more details and “any grain of hopefulness I can give to patients.” There were, he said, “so many pathetic cases.”

On December 6th, Banting and Best took another major step when they tried using alcohol, instead of a salt-based solution, in the preparation of the foetal calf extract. Alcohol evaporates at a much lower temperature than water, plus it could dissolve and remove some of the contaminating impurities in the extract solutions. On December 7th, extract prepared with the alcohol method worked well with both Dog #23 and Marjorie. From the use of alcohol, they concluded that the active agent in the foetal pancreas must be soluble in the alcohol, as, they surmised, it would also be in fresh adult cow pancreatic tissue. On December 8th, Banting and Best removed the pancreas from Dog #35 and proceeded to prepare an extract from it using the alcohol extraction method. On December 11th, Dog #35 was given an injection of extract prepared from its own whole pancreas. The animal’s blood sugar level dropped from .38 to .18 in four hours. It was clear that alcohol worked to extract the internal secretion from the whole pancreas. No more foetal pancreas. The research could push forward using easily available fresh whole pancreas from cattle.

James Collip as a graduate student, c. 1914. University of Toronto, The Discovery and Early Development of Insulin.

The recent breakthroughs prompted Macleod to formally invite Dr. James Bertram Collip to join the pancreatic extract research team so he could use his specialized biochemistry skills to further the development and purification of the extract. Collip was born in Belleville, studied physiology and biochemistry at the University of Toronto, and earned his PhD in 1916. He became a full professor of physiology at the University of Alberta.

In 1921, Collip, who was happily married with a young family, had come to the University of Toronto on a sabbatical leave. Based in a lab in the Pathology Building at Toronto General Hospital, Collip began working on the whole beef pancreas on December 12th . Instead of dogs, Collip experimented with rabbits rendered diabetic experimentally, although he quickly discovered that normal rabbits could be employed to test the potency of the extract. He relied on a diabetic Airdale terrier to test the anti-diabetic effectiveness of his extract. 

At the same time, Banting and Best were also attempting to purify the whole beef extract. They experimented with dialysis and the use of toluene, a clear hydrocarbon that smells like paint thinner, after alcoholic evaporation in order to remove fat-like impurities in the extract. Banting and Best prepared a large batch of extract, but when they tested it, they found it lacked potency, based on tests on diabetic dogs. However, they still had some extract they knew to be potent and on December 21st, decided to try it on a fellow physician, Dr. Joseph Gilchrist, who was suffering from diabetes. They administered the extract through a stomach tube rather than by injection. To Banting and Best, it was worth a try in light of Gilchrist’s declining health. However, the extract had no effect on Gilchrist, either positive or negative. 

Note “To prepare insulin” written by James B. Collip, December, 1921. University of Toronto, The Discovery and Early Development of Insulin.

Banting and Best’s frustrations continued so they took a break for Christmas. But Collip was making significant progress. He was able to use a vacuum still to evaporate the alcohol from the extract solution, although only partially, followed by filtration. This extract worked very well at reducing the Airdale’s blood sugar level. Collip also tested the dog’s urine and discovered the extract left it completely sugar and keytone free. This finding marked another significant step. At the same time, Collip showed that the extract restored a key aspect of the diabetic dog’s liver function: making glycogen from the carbohydrates that the dog had consumed. Normally, the pancreas will respond to higher levels of blood glucose after eating by releasing insulin, which lowers blood glucose levels by prompting the liver and muscles to absorb glucose from the blood and store it as glycogen. After Christmas, Collip shared his impressive results with Banting, Best and Macleod on the train trip to New Haven.

Banting hoped the New Haven meeting on December 30th would be a triumph, but, at least for him, it proved to be especially disappointing. His nerves got in the way of giving a convincing presentation and Macleod felt the need to step in and rescue him. There was a who’s who of diabetes specialists in attendance, including Dr. Joslin and Dr. Frederick Allen, and they posed some tough questions. Macleod faced considerable scepticism from the specialists in the audience, although one, Dr. George Clowes, director of research at Eli Lilly & Co. of Indianapolis, Ind., was impressed by what he heard. Afterwards, Clowes approached Macleod, offering the collaboration of Eli Lilly to help develop the extract. He discussed the idea with Banting and Macleod, but they politely declined the offer as the work was not sufficiently advanced to be considering commercial preparation. However, Clowes suggestion would be kept in mind. Banting’s halting performance in New Haven and Macleod’s rescue only added to Banting’s mounting frustrations and fed his suspicions that Macleod intended to take over the work. It would be an especially difficult train trip home for Banting.

For Banting, the uncertainties continued during January. Banting felt isolated as Collip’s successes with the extract mounted and he saw Collip and Macleod working more closely. Best was also busy conducting specialized respiratory quotient (RQ) tests. The RQ test measured the ratio of carbon dioxide breathed out to oxygen absorbed and was thought to provide a reliable guide to whether or not carbohydrates were being burned in the body. If the ratio was higher than normal, diabetes was indicated, but the RQ test could show the potency of insulin in reducing the CO2/O2 ratio. Best also collected beef pancreases for Collip, and also did the initial extraction steps. Banting was left to conduct the occasional surgery, but he was increasingly anxious to see the extract given its first official trial. 

Naturally, Banting thought he should administer it. However, he had no official standing as a physician at the Toronto General Hospital. Nevertheless, Banting pressed Macleod for a human trial, and he, in turn, pushed Dr. Duncan Graham, who oversaw the diabetes clinic, for permission from the hospital. But Graham refused Banting’s request for temporary standing, noting he had no experience treating diabetic patients. Banting had no reasonable response, yet he continued to advocate for a trial. Macleod finally relented when the father of a 13-year-old boy gave TGH permission to administer the extract in a desperate effort to save his son, Leonard Thompson. The longevity experiment with Marjorie had been going well and it made sense to Banting and Best, and reluctantly to Macleod, to try the extract on Thompson.

On January 10th, Best prepared some whole beef pancreas extract using the method developed in December, which integrated the use of the vacuum still as developed by Collip. They tested its potency on a dog and then Banting and Best gave each other small injections of the extract to make sure it was safe. The next day, they took a vial of the extract across the street to the TGH diabetic clinic. On the afternoon of January 11th, a young house physician, Ed Jeffrey, administered 15 c.c. of the extract into Thompson’s buttocks, 7.5 c.c. in each one. All Banting and Best could do was wait in the corridor outside the ward. They were not permitted to handle any of the specimens as conducting the tests was the hospital’s responsibility. Thompson’s blood sugar fell from .44 to .32, and the sugar levels in his urine fell moderately. The results of “Macleod’s serum”, as was noted on Thompson’s chart, were positive, but disappointing to Banting, Best and Macleod. There had been a 25% drop in blood sugar, a reduction in sugar in the urine, but the extract’s modest impact did not outweigh the reaction to it. There was a sterile abscess that developed at one of the injection sites, caused by impurities that remained in the extract, not because of any infectious agent.

Dr. J. B. Collip, c. 1920. University of Toronto, The Discovery and Early Development of Insulin.

When Collip heard that a human trial of Banting and Best’s extract had gone ahead, he was quite upset. Their move undercut the research team’s arrangement and transformed the extract purification problem into a competition. On January 14th, the Toronto Star ran an article that discussed the first trial. With tensions among the team escalating, the publicity further exacerbated the situation. A reporter, Roy Greenaway, had heard about the Thompson test and found his way to Best and then to Macleod. Macleod was very careful in discussing the extract with the reporter and advised against publishing anything. However, Greenaway published the article, which barely mentioned Banting and Best. The oversight further annoyed Banting. His mentor, Dr. Starr, was soon able to facilitate a meeting between Banting and Macleod in an attempt to resolve Banting’s fear that the esteemed physiologist would steal his work. Macleod initially didn’t take Banting’s concerns seriously, but when Duncan Graham impressed upon him that Banting was quite serious, Macleod immediately went to Starr, who Macleod knew Banting trusted. When Banting and Macleod finally met, they were able to clear the air of misunderstandings between them, at least temporarily.

Aside from the personal dynamics of the research team, the pressure was now squarely on Collip to prepare a more purified extract that would be potent, but not cause any reactions. Collip’s skill in extracting insulin and several other hormones would become a legend in Canadian medical research. He was a mixture of part chef, part brewer, and part wizard, hurriedly bouncing in his lab between mixing and filtering, distilling and evaporating, concentrating and diluting, centrifuging and blending samples of the extract.  He worked quickly, reflecting his restless temperament, and rarely made the same extract twice, as he often forgot what he had done. However, in January 1922, Collip’s “bathtub chemistry” was just what was needed to prepare a better version of the whole beef extract. While there were endless variables, Collip’s method involved gradually increasing the concentration of alcohol in the pancreatic mixtures. He soon discovered that the active agent stayed in solution at higher and higher concentrations, while most of the proteins precipitated out and the fats and salts could eventually be extracted by centrifugation (applying centrifugal force through rapid spinning to separate particles from a solution) and washing. On January 16th, Collip discovered that at a precise limit in the alcohol concentration (a little over 90%), he could actually trap the internal secretion in the purest form yet seen. After testing its potency on rabbits and waiting to check for abscesses in the rabbit’s skin where the extract was injected, Collip knew he had an extract that could be tested on patients in the TGH diabetic clinic. 

On January 23rd, treatment of Leonard Thompson resumed with Collip’s extract; he had only received the one insulin treatment of Banting and Best’s extract, but then returned to a diet-based treatment. This time, after Thompson received Collip’s extract, the effect was quite dramatic; sharp drops in blood sugar levels and glygosuria (sugar in the urine) almost disappearing. Leonard quickly became brighter, more active, looked better and was stronger. Moreover, there were no abscesses or other side effects. This time, the pancreatic extract clearly worked.

Around the time Thompson received this new version of the beef extract, Collip paid a visit to Banting and Best. 

“Well fellows,” he told the young researchers, “I’ve got it.” 

How did you do it? Banting and Best asked.

Collip replied: “I have decided not to tell you.” Macleod, he added, agreed. 

He went on to announce he was leaving the group and threatened to take out his own patent on the extract. Simmering as he listened, Banting finally lost his temper and took a swing at Collip. Fortunately, Best stepped in to restrain Banting before the altercation turned violent. This confrontation reflected a perfect storm of fatigue, frayed nerves, frustration, excitement and paranoia among Banting, Collip and Macleod, and, to a lesser extent, Best. The conflict threatened the future of their work. 

However, by January 25th, the storm had passed. Shortly after the blow-up between Banting and Collip, Dr. FitzGerald, of Connaught Labs, along with Velyien Henderson, facilitated meetings with Collip and Macleod to settle the whole mess, exacerbated by distrust, especially between Banting and Collip. FitzGerald, an ally of Banting and Best, had become quite concerned that Collip was well within his rights to return to Alberta and patent his method of beef extract purification separately from what anyone else had done in Toronto. It would clearly be a disaster if Collip left under such circumstances. To head off such an outcome, FitzGerald was able to formalize in writing a seminal research and development agreement between Connaught and Banting, Best, Collip and Macleod. 

Dated January 25, 1922, and signed by the four men, as well as the chairman of U of T’s Connaught Committee, Albert E. Gooderham, the agreement was based on two key conditions: One, that the collaborators agree not to take out a patent with a commercial pharmaceutical firm during an initial working period with Connaught; and two, that no step that would involve any modification in policy concerning the research work would be allowed unless first discussed among FitzGerald and the four collaborators. FitzGerald offered Connaught’s modest facilities and the Lab’s expertise with biologicals production, along with $5,000 from the Labs’ reserve funds, to help expedite the further development and production of the extract for expanded clinical trials. There was also an agreement for Connaught to “consider subsequently the question of commercial production of the extract or extracts if and when its (or their) merit has been established and its (or their) method of preparation has been given to the world.”

By February 1922, six more diabetic patients at TGH has been successfully treated with the extract, along with Dr. Gilchrist. Banting and Best delivered a paper on the research to a group of doctors at the Toronto Academy of Medicine, and also contributed to a seminal paper, “Pancreatic Extracts in the Treatment of Diabetic Mellitus,” which was submitted for rapid publication in the Canadian Medical Association Journal. Meanwhile, Roy Greenaway of the Toronto Star was prodding Macleod for the full story. This time, Macleod felt much more comfortable seeing that it was told.