Making Insulin (Part Two)

On July 23, 1922, Dr. Frederick Banting was desperate for insulin when he boarded a train in Toronto bound for Indianapolis. Several events had brought Banting to this point, including unexpected obstacles that sharply limited the supply of pancreas tissue, coupled with limitations in Connaught Labs’ production equipment and insulin quality that Eli Lilly’s Dr. George Clowes pointed out when he visited the Labs. Also, at the time, Charles Best and J.J.R. Macleod were both away from Toronto, leaving Banting to respond to a growing crisis, especially in meeting the insulin needs of the three young diabetic patients who had recently come to Toronto from the United States: Ruth Whitehall, Myra Blaustein, and Teddy Ryder.

The Toronto insulin crisis came to a head when 57-year-old Charlotte Clarke, suffering from severe diabetes complicated by a badly infected, gangrenous leg, faced death. No one had performed surgery on a diabetic before, especially such a severe case, for fear of anesthesia complications. But perhaps there was hope if insulin injections would stabilize her long enough to tolerate the anesthesia so surgeons could amputate the infected limb.

In early July, Clarke had been referred to Banting by a medical officer friend, former Captain L.C. Palmer, whom Banting met during the First World War in Cambrai, France. Banting couldn’t refuse seeing Clarke, who was close to a diabetic coma. She was admitted to Toronto General Hospital, but there was no insulin available for her. However, Banting felt there was a chance to save Clarke, and possibly other diabetics needing critical surgery, if she was given insulin. As the alternative was certain death, Banting felt there was little risk in trying. 

On July 10th, Banting decided to temporarily limit the insulin being given to five other diabetic patients at TGH in order to provide an injection for Clarke. At the time, the quality of the insulin was poor and there was very little of it, but what there was, Banting felt, could be put to more urgent use to potentially save Clarke’s life. While the Toronto insulin supply was limited, he had received some from Eli Lilly and didn’t expect Connaught’s production to soon grind to a halt. The next day, with Clarke stabilized, Palmer proceeded to amputate her right leg above the knee. The surgery proceeded quite smoothly and Palmer and Banting were amazed to see Clarke make a normal recovery. A week later, Palmer removed the stitches and her incision appeared to have healed normally. But in light of the worsening insulin supply situation, Clarke received no further insulin for the next seven days.

Dr. R. D. Defries. The Albert and Mary Lasker Foundation.

In fact, Banting was growing increasingly distressed at the rapidly worsening Toronto insulin supply situation. And after Clowes’ visit, he grew suspicious about why Eli Lilly had been able to more consistently produce insulin than Connaught. Were they were holding some vital information back? Thus, in consultation with Dr. R.D. Defries, Connaught’s assistant director, Banting and David Scott, the Labs’ leading chemist, decided to visit Indianapolis to see for themselves what was happening at Eli Lilly. 

When they arrived, Banting was prepared for some resistance to his demands for answers. But he was surprised to be met at the train station by J.K. Lilly and Clowes, who had a package containing 150 units of insulin for Banting to take back to Toronto, with another 150 units ready to be sent. Relieved, Banting put his head on Mr. Lilly’s shoulder and wept. 

The Eli Lilly team was surprised to hear about the dire insulin shortage in Toronto and offered to delay their own local clinical work in order to supply Toronto until Connaught’s production stabilized. Banting and Scott were also given a full tour of Lilly’s facilities. Overwhelmed with what he saw, Banting was especially impressed with the large vacuum still that looked far superior to the antiquated tunnel evaporation process Connaught used. Indeed, Banting became very determined to press the University of Toronto to expedite the installation of a similar vacuum still at Connaught.

By the time Banting returned to Toronto on July 27th, Charlotte Clarke’s condition had sharply declined. The stump of her leg had become seriously infected and Palmer doubted much could be done to save her, with or without insulin. Banting immediately went to Toronto General Hospital to see Clarke with some of the insulin he had brought back from Indianapolis. As soon as insulin injections resumed, Clarke’s wound began to heal. She soon made a full recovery and lived for many years using a prosthetic leg. 

Portrait of Sir Edmund Walker. A history of the Canadian Bank of Commerce, Toronto Oxford University Press.

Banting next went to the office of Sir Edmund Walker, chair of the University of Toronto Board of Governors, with a request for $10,000. The money was urgently needed to enable Connaught to acquire a vacuum still and other equipment required to raise the Labs’ insulin manufacturing standard to a level that better matched what Banting had seen at Eli Lilly. However, Walker did not give Banting the answer he wanted; the request would be raised at the next Board of Governors’ meeting in the fall. Frustrated by further delays and the lives it could cost, Banting asked Walker if the Board would accept the funds if he could find them elsewhere. Walker agreed and the next morning, Banting was on a train to New York City. 

There had been several offers of financial help sent to Toronto from wealthy American doctors and Banting planned to follow up on one from Dr. Rawle Geyelin of New York City. Geyelin, who had several severe diabetics under his care, had recently visited Toronto and impressed Banting. As soon as Banting arrived at his office, Geyelin picked up the phone to speak with Robert Bacon, father of one of his severely diabetic patients. After a short conversation, Geyelin turned to Banting and asked how the $10,000 cheque should be made out. Banting immediately sent a telegram to Defries to tell him to go ahead and order the new vacuum still. The $10,000 gift was formally accepted by U of T’s Board of Governors on August 10th and on the 17th, the university’s Insulin Committee, at its first meeting, directed Connaught to purchase the equipment.

Meanwhile, Eli Lilly’s insulin plant operated continuously during the summer of 1922, with more than 100 employees working in three shifts focused solely on the challenges of large-scale production. Although yields increased, not unlike at Connaught, Lilly encountered problems securing sufficient pancreas tissue to meet production demands. By early August, Connaught and Eli Lilly were both producing fairly reliable supplies of insulin and able to provide it to more diabetic specialists. J.K. Lilly was particularly struck by the progress. As he wrote in an August 4th letter to Clowes, “I am almost overwhelmed with this tremendous situation, and experience some difficulty in keeping my feet on the ground and my brain in normal operation.” In an August 8th letter to Clowes, he added, “You have certainly entered the holy of holies and are sitting on the throne with the elect. It is a marvelous development and I rejoice in it.” 

Such sentiments, however, were soon disrupted as several batches of Lilly’s insulin failed key potency tests. Variations in alcohol concentration prompted by a change in the kind of alcohol Lilly was allowed to use by the U.S. government, coupled with variations in temperature and the extraction process, led to insulin of lower potency and purity, and thus some reactions. To keep up with demand, this situation prompted efforts at Lilly to double production, with Clowes asking Banting to reduce the amount he asked to be sent to Toronto and the dosage he could administer. The stress on the Eli Lilly insulin team, especially Walden, soon became overwhelming. During August, the insulin situation became a frustrating mix of miraculous recoveries and baffling failures.

These frustrations continued into September, although by early September, Eli Lilly’s output levels – almost double that of June and July — had recovered from the August setbacks. There were also promising developments at Eli Lilly with a new large-scale production process that would significantly increase the yields and purity of its insulin.

Meanwhile at Connaught, during September, Charles Best was preoccupied with a variety of insulin production challenges, not least of which was heavy rain flooding the Labs’ production area. Some insulin had to be returned from Banting’s practice and the Christie Street diabetes clinic as unusable. September 11th, as Best wrote in a letter to his fiancée, Margaret Mahon, “was a blue Monday for sure.” “Macleod was raving for more dope and better etc.,” as Best referred to the extract. 

By September 19th, after a couple of weeks of Best working in the lab every night and all day on Sunday, the problems had been largely resolved and he could tell Margaret it had been a good day with the “dope.” In the midst of this intense insulin work, Best also started his medical studies, although he was so busy he had to ask classmates to take lecture notes for him. 

Dr. Peter J. Moloney.

Best’s efforts at improving insulin production and purification were supported by the chemistry expertise of Dr. Peter J. Moloney, the Lab’s first research chemist. Before pH meters were available (the first was constructed in 1934), Moloney developed a quick-acting hydrogen electrode (known as the “Moloney Electrode”) designed to more accurately and rapidly determine the acidity of bacterial culture broths used in antitoxin and vaccine production. The Moloney Electrode also proved useful for measuring the pH during insulin production.

Moloney’s initial work on insulin purification was based on his earlier research that involved the purification of proteins using hippuric acid, a compound found in urine. The expense and loss of potency inherent in the boiling down of such solutions suggested a better approach would be to try the adsorption of proteins from aqueous solutions using certain materials, one of which was benzoic acid, which could be subsequently dissolved. Benzoic acid was useful because it was relatively insoluble in water and because of “the fineness of division in which they come down when a dissolved salt is treated with an acid,” as Moloney noted in the introduction to his paper, “Concentration of Insulin by Adsorption on Benzoic Acid,” written with D.M. Findlay.

The benzoic acid method marked a distinct advance in large-scale insulin production. The substantial amounts of alcohol and/or acetone necessary for the fractional precipitation of the proteins and the final precipitation of insulin involved in the previous method were avoided. The final insulin product was much freer of protein material than had been previously possible, resulting in far fewer reactions. However, this method involved long and tedious filtration steps that limited its ultimate utility. Nevertheless, some 250,000 units of insulin using the benzoic acid method (referred to as “Moloney insulin” by some Toronto nurses) were produced at Connaught and used clinically in Toronto during the fall of 1922, with “very satisfactory results,” according to Best.

Meanwhile, in Indianapolis, Eli Lilly’s new large-scale production equipment was being installed. As Clowes told Macleod, “large amounts of Iletin should be available very soon.” (“Iletin” was the trade name Eli Lilly used for the insulin it produced.) Best made another trip to Indianapolis to see the new equipment and also learn about the new insulin production process Walden had developed. 

Walden’s approach was based on utilizing what seemed like the magic isoelectric point, or pH level, of the alcoholic extract solution, at which a much purer active principle precipitated out of the solution. Walden and Eda Bachman had been involved in a basic research program at Lilly, focused on determining the isoelectric points of the materials used in several other Lilly products. The isoelectric point was a well-known metric for chemists, defined by the pH of a medium at which a protein carries no net charge and thus will not migrate in an electric field. At this specific isoelectric point, proteins readily precipitate, or fall out of solution — a property useful for separating mixtures of proteins or amino acids. 

During the fall of 1922, the cause of Lilly’s earlier insulin stability issues was traced to varying pH levels among batches. Walden closely tracked these levels and discovered that the weakening of the insulin solution was due to the gradual formation of a precipitate in the solution that also contained the active principle, the effect of which was to reduce the activity of the remaining insulin solution. It appeared that at the “wrong” pH level, insulin was being precipitated out of the insulin solution. Walden realized that the precipitate was indeed far purer and potent than had previously been observed. Thus, Walden deliberately adjusted the pH of the pancreatic extract solution to the isoelectric point for insulin to cause the maximum precipitation. As historian Michael Bliss described it, “Collect the precipitate, fiddle with it a bit more, and you had by far the best insulin yet.” “[It was] a product having a stability many times as great and a purity ranging from ten to a hundred times as great as the best product hitherto obtainable,” Walden wrote.

By October 1st, Lilly’s new large-scale insulin production plant was in operation and employing the isoelectric point precipitation and purification process. The new facility quickly produced a surplus supply, which then prompted discussions between Clowes and the U. of T. Insulin Committee about how best to distribute it. While Clowes felt considerable pressures to expand Iletin distribution beyond a specific list of diabetic specialists in the U.S. and Canada, Macleod and the other members of the Committee were quick to resist such proposals until large-scale methods were better established. Potency, Macleod understood, still varied and the protein content in samples of Lilly’s extract was often too high and potentially dangerous. 

Of further concern, later in October, was the discovery of stability problems. Previously, there had been little delay between when a vial of Iletin was produced and when it was administered. Thus, the chemists involved in the production process paid less attention to measuring potency levels over time. In order to overcome the stability problem, in fact, Lilly was forced to return to small-scale production and seek advice from the Toronto team, which meant Best had to make three trips to Indianapolis in early November. 

On November 11th, soon after returning from Indianapolis, Best reported to Connaught’s director, Dr. J.G. FitzGerald, that the Labs were producing about one litre of insulin per week and also that Lilly’s plant was now turning out insulin of good potency. However, Lilly was still having trouble with deteriorating potency, i.e., that the blood-sugar control effect of each injection did not last as long as it should. Connaught wasn’t having the same stability issues as its insulin was used as quickly as it was being produced. The issue was resolved by a more careful and standardized testing process to monitor stability over time.

By the end of December, and once the stability issue was resolved and insulin was being produced consistently by Lilly, the Insulin Committee established a policy that required samples of every batch of Iletin intended for sale to be sent to Toronto for physiological and clinical testing. The results would be sent by telegram from the Committee to Lilly as soon as possible, with a statement of approval or rejection. For a limited period, the Committee also agreed that if its approval of release did not arrive in Indianapolis within one week of the date that the samples had been sent, Lilly, if necessary, could distribute the insulin at its own risk, but would need to recall materials if the Committee afterwards reported it was unsatisfactory in dosage or potency. 

By mid-February, 1923, Eli Lilly was again producing Iletin far in excess of demand from diabetes specialists, once again prompting discussions with the Insulin Committee about a distribution strategy beyond designated specialists. While the members of the Committee were now confident in the large-scale production methods that yielded a consistent and stable insulin product, more general distribution could not proceed until they were certain that non-specialist physicians understood how to prescribe and administer insulin, and how to manage diabetic patients. In the U.S., such an education initiative relied on specialized publications for physicians, while in Canada, the main approach was the in-person training of physicians, initially at the University of Toronto, beginning in April 1923.

Eli Lilly and Company’s McCarty Street production plant in Indianapolis, Indiana, c. 1886.

Defries had also been thinking about the Canadian insulin production situation. Clowes suggested that Eli Lilly could meet Canada’s insulin supply needs directly, or indirectly through Connaught, such as shipping insulin in bulk from Lilly to Connaught to be filled, packaged and distributed. He could not envision a sufficiently large insulin production plant in a university building. In a March 14th letter to U of T president Falconer, Defries outlined his concerns. There had been considerable press attention focused on the limited capacity Connaught had in its Medical Building facility for increasing and accelerating insulin production, and in terms of the equipment and space required to handle the large amounts of beef pancreas tissue needed at the start of the process. 

There were suggestions in the press and from politicians that the federal government should provide funds to expand Connaught’s insulin production capacity so as to expedite the drug’s availability across Canada. At the same time, there were similar calls for the Ontario government to provide funds for Banting’s research work. The Insulin Committee would certainly welcome and put to use any such funds, especially since, as was emphasized to the press on March 3rd, “the work had been carried on up to the present by means of money borrowed from the Connaught Laboratories.” The Committee was referring to a $5,000 advance provided by FitzGerald from Connaught’s reserves to Banting, Best, Collip and Macleod in late January, 1922.

“[It] is desirable,” Defries stressed to Falconer, “to incorporate in any statement to the Government a request for the continuation of this work in the Connaught Antitoxin Laboratories.” In just the previous two weeks, insulin distribution had grown very rapidly and there was sharply increasing demand. Connaught was already assuming the responsibility for distributing insulin throughout Canada until June, 1923, at least, and a decision was needed. “Since the material must be used by many persons daily,” Defries suggested, “and since the treatment of many patients and their maintenance will be a problem for the Provincial Governments and the Municipalities, a valuable service can be rendered by distribution at the lowest possible prices.” It would also be a mistake, he stressed to Falconer, “to assume that additional space as contemplated in our plans will not be necessary.” It thus seemed the university should approach the provincial government about Connaught’s needs, with an emphasis on asking for $20,000 for increased laboratory space to continue research and the production of insulin for clinical and experimental use. “I do not know how we could obtain funds later if they were required, when the Government is probably willing to consider the need at present,” Defries said. 

At the April 10th meeting of the Insulin Committee, Defries reported there was “at present a considerable surplus of insulin on hand,” and asked for advice for how it might be distributed in Canada. Such plans depended on physicians in other provinces learning the proper clinical use of insulin based on the experience in Ontario. Expanded general distribution beyond diabetes specialists and to physicians and their diabetic patients also required upgraded and larger production facilities at Connaught. The application of the isoelectric point step in the precipitation and purification process was one of several advances that had enabled more efficient production with Connaught’s makeshift facilities in the Medical Building basement. But a clear limit had been reached. For Defries, the main challenge was finding a suitable space and sufficient capital to pay for the needed equipment.

In early April, 1923, an opportunity opened up for a larger space for Connaught’s insulin facility, and the University Board of Governors allowed Defries to take advantage of it. A building on the campus that had accommodated the university’s Y.M.C.A. facility, located just north-west of College Street and University Avenue (later supplanted by the current FitzGerald Building), was slated to be torn down and Defries had to act quickly. But financing the necessary renovations and required equipment remained a challenge until Defries personally approached Ontario premier, Ernest C. Drury, who approved a provincial grant of $35,000. 

With Connaught’s reserves providing the balance, renovations were expedited during June and July at a total cost of $43,000. As Defries told a reporter, Connaught would now be able to produce enough insulin for all of Canada in the new facility, and at a steadily declining price. In April, 1923, “Insulin-Toronto,” as Connaught’s product was named, was priced at 5¢ per unit. But in mid-May, the price had fallen to 4¢ per unit. By July 1st, Connaught’s goal was to supply Insulin “practically at cost.”

An August 1st article in The Globe noted that work on the refurbished building had been partially completed, and Best had already moved into his office there. The story pointed out that not only would Insulin be produced commercially nowhere else in Canada, but Connaught’s facility would also “provide the curative fluid for all countries which do not manufacture their own supply.”

By the end of August, the mood among the insulin team in Toronto, and in the city, seemed quite upbeat. Banting and Best had formally opened the annual Canadian National Exhibition, and Connaught’s newly expanded insulin production capacity was a key feature in a popular exhibit in the CNE’s Government Building. As The Globe reported on August 27th, “A huge sign in the Government Building, bearing in large letters the legend ‘Insulin,’ and proclaiming that it was discovered in the Department of Physiology of the University of Toronto and is prepared in the Connaught Laboratories, is a beacon to all who may be interested.” The display featured a large vacuum still, “which permits of the evaporation of the alcohol from the insulin mash at a low temperature,” along with samples arranged on a board that illustrated the various stages in the production process. Large crowds flocked to the exhibit. Public interest in insulin was so great, in fact, that “already more than 3,000 of the bulletins prepared by Best on the subject have been disposed of.”

Newspaper clipping from the Toronto Star on November 29th, 1923 about the Connaught Lab’s production of insulin in their new facility. Insulin Patent Sold for $1, The Banting House Museum.

By November, as the Toronto Star reported, Connaught’s “Insulin factory” was “worth upwards of $35,000” and producing 250,000 units weekly; the average requirement for a diabetic varied from 15 to 20 units per day. Best was in charge of the facility, with the close assistance of David Scott, who led a staff of 26 that worked around the clock to keep up with the demand in Canada and beyond. Connaught’s insulin was being exported to Ireland, South Africa, New Zealand, Australia, as well as to several other countries without their own production facilities. The price Connaught charged, as with its vaccines and antitoxins, was governed entirely by the cost of materials and production, since the Labs was “not engaged in commercial business.” As of June, the price of Insulin in Canada was 3¢ per unit and by late November it had dropped to 2¢; distribution was directed to specific diabetic departments in hospitals or to physicians specially trained in its use.

Eli Lilly was permitted to begin normal U.S. distribution of Iletin through physician prescription in October 1923, initially with some 7,500 physicians treating 25,000 diabetic patients with the extract. By this time, Lilly was also developing its plans to export Iletin to several countries. Throughout 1923, in fact, the Insulin Committee had worked with several U.S. pharmaceutical firms interested in applying for a licence to produce insulin. There had originally been nine applications, and seven were ultimately considered: Parke Davis & Co. of Detroit; Frederick Stearns & Co., also of Detroit; Wilson Laboratories in Chicago; H.K. Mulford & Co. of Philadelphia; Lederle Antitoxin Laboratories, based in Pearl River, N.Y.; E.R. Squibb in New Brunswick, NJ; and Arlington Chemical Co. of Yonkers, N.Y. Inspections of all these firms had revealed various issues with the facilities at Armour & Co. in Chicago, and Berg Biological Laboratories in Brooklyn, N.Y., eliminating them from consideration. 

At the same time, the Committee’s secretary was managing voluminous correspondence related to negotiating international insulin rights arrangements, particularly in India, Romania, Hungary, Czechoslovakia, China, Greece and Australia. The development of insulin production in Denmark at Nordisk Insulinlaboratorium, to meet the needs of diabetics in Scandinavia, was well established by the fall of 1923. Nordisk’s efforts began almost a year earlier with the visit to Toronto by Professor August Krogh of the University of Copenhagen. He had won the Nobel Prize in physiology and medicine in 1920, and, during the fall of 1922, while on a lecture tour in the U.S., heard increasingly dramatic reports about people being successfully treated with insulin.