How Insulin Became a Treatment: A Breakthrough Revisited

Insulin became a medical breakthrough when a pancreatic extract could be purified, measured, given repeatedly to people with severe diabetes, and manufactured at useful scale; the Toronto work of 1921 and 1922 was decisive, but it was not a solitary moment. It rested on decades of research linking the pancreas, islet cells, glucose, and diabetes.

The history is more instructive once you set the familiar hero story aside. Ideas mattered. So did laboratory space, experimental design, and chemical purification. So did early patient care, assays, production, and correction after an imperfect first attempt. A treatment emerged from a system of contributions.

Before Toronto, the pancreas hypothesis had a history#

Nineteenth-century anatomy identified clusters of pancreatic cells later called the islets of Langerhans. In 1889, research showing that removal of the pancreas caused severe diabetes in dogs strengthened the causal connection, and by the early twentieth century, investigators had linked damage to the islets with diabetes and proposed an internal pancreatic secretion that regulated glucose.

Researchers in several countries tried pancreatic extracts. The central difficulty was not merely imagining a missing substance. The pancreas also contains digestive enzymes, and crude preparations were inconsistent, irritating, or ineffective. The active principle had to survive extraction, be separated from unwanted material, and retain measurable activity.

This prior record changes the meaning of “discovery.” The Toronto group did not invent the idea that the pancreas mattered; their breakthrough was to turn a plausible but technically blocked idea into a preparation that produced repeatable clinical improvement.

A proposal became an experiment#

In 1920, Frederick Banting developed a proposal to isolate the pancreatic internal secretion while reducing interference from digestive tissue. He brought the idea to J.J.R. Macleod, a physiologist with extensive work in carbohydrate metabolism. Macleod provided laboratory space, animals, equipment, guidance, and a student assistant. Charles Best worked with Banting during the summer of 1921.

The early experiments used dogs and measured blood and urine glucose alongside clinical status. The duct-ligation concept was not the only route that ultimately mattered; extracts from other pancreatic material also became part of the work. The researchers revised methods as results accumulated.

Animal glucose reduction established biological activity, not a finished medicine. Extract strength varied, preparation was laborious, and translation to repeated human dosing required cleaner material and reliable assays.

The lesson is methodological, and it outlives the story. A mechanism-inspired intervention still has to survive contact with measurement and manufacture, however convinced you are by the mechanism. A compelling hypothesis earns a test, not a conclusion.

Purification changed what could be attempted#

James Collip, a biochemist, joined the effort late in 1921 and worked on purification. His contribution addressed the bottleneck between an active crude extract and a preparation suitable for human use. Alcohol fractionation and other adjustments improved potency and reduced unwanted material.

Macleod contributed to experimental interpretation, resources, and presentation of the findings. Clinicians Walter Campbell and Almon Fletcher helped care for early patients and assess response. Laboratory technicians and hospital staff supported the work. These roles are not interchangeable, which is precisely why the team could do what one person could not.

The article often described as the first official clinical report listed Banting, Best, Collip, Campbell, and Fletcher. Authorship itself preserves a more distributed record than the shortest popular retelling.

Leonard Thompson's treatment included correction, not magic#

Leonard Thompson, a young patient with severe diabetes at Toronto General Hospital, received a pancreatic extract in January 1922. The first administration produced limited biochemical improvement and a local reaction. Treatment paused while purification continued.

A later, more purified preparation produced a much clearer fall in glucose and ketones together with clinical improvement. That sequence matters. The breakthrough did not arrive fully formed at the first injection. An adverse and insufficient result prompted revision, and the revised product performed better.

The March 1922 CMAJ report described seven patients treated with pancreatic extracts. The paper followed glucose in blood and urine, ketones, diet, dose, and observed clinical condition. The samples were tiny and uncontrolled by modern standards. The effects in people with profound insulin deficiency were nevertheless large, rapid, biologically coherent, and repeatedly linked to dosing.

This is an example of effect size changing how you read the evidence: a randomized mortality trial was neither available nor needed to decide whether a person with otherwise fatal insulin deficiency improved after the missing hormone was replaced. Randomization remains essential when effects are smaller, outcomes fluctuate, and alternative explanations are plausible.

A treatment required production and standardization#

Success created an immediate supply problem. Extracting insulin from animal pancreases at clinical scale required collection, purification, potency testing, sterility, and consistency, and the University of Toronto's Connaught laboratories and a commercial manufacturing collaboration expanded production while methods continued to evolve.

Standardization was clinical science. A “unit” needed reproducible biological meaning so that a dose from one batch did not behave unpredictably against another. Impurities caused reactions, and variable potency created danger. Quality control was not an administrative step after discovery; it was part of making the discovery real.

The archival patent record adds nuance to a common slogan. On December 19, 1922, Banting, Best, and Collip assigned the Canadian patent for the preparation of insulin to the governors of the University of Toronto for a token payment of one dollar. Holding the intellectual property through the university was intended to support controlled production and dissemination. The document does not mean every later insulin formulation, delivery device, or manufacturing process became patent-free.

Recognition compressed a complicated collaboration#

The 1923 Nobel Prize in Physiology or Medicine was awarded to Banting and Macleod for the discovery of insulin; popular accounts often use that award as the complete credit map, and most of the single-name answers you hear come from it. Read it instead as one historical act of recognition inside a larger record. That record includes Best, Collip, Campbell, and Fletcher. It includes prior investigators, early patients, and production teams.

Disputes over credit are not unusual when science is told through prizes. A prize has a fixed number of recipients; translation has many dependencies. The responsible history does not need to diminish one contribution to recover another. It specifies what each part added and where the evidence is uncertain.

Insulin was a treatment, not a cure#

Insulin replacement transformed the outlook for people with type 1 diabetes, but it did not restore a self-regulating pancreas, and early therapy required repeated injections, dietary calculation, glucose monitoring available at the time, and management of hypoglycemia. Long-term vascular complications did not disappear.

Later advances changed purity, duration, and species source. They changed concentration, delivery, and monitoring. Researchers determined insulin's amino-acid sequence and three-dimensional structure. Recombinant human insulin reduced dependence on animal extraction. Analogs altered timing. Glucose meters, continuous sensors, pumps, and automated delivery systems brought treatment closer to physiological regulation without making it effortless.

The site's articles on how insulin therapy works and insulin pumps and closed-loop systems follow that scientific continuation.

Access remains part of the history#

WHO's centenary report emphasizes a difficult fact: insulin is essential. Yet access, affordability, and supply remain unequal. So do diagnostics and trained care. The one-dollar patent assignment is morally resonant. But it did not by itself determine a century of global production, pricing, health-system capacity, or modern intellectual property.

That does not erase the original decision. It corrects the myth that one generous act permanently solves implementation. Access has to be built repeatedly through manufacturing capacity, procurement, and competition. It is built through regulation, coverage, and distribution. It is built through refrigeration where required, education, and clinical follow-up.

Five lessons from the breakthrough#

First, prepare the intellectual ground. Decades of anatomy and physiology made the Toronto hypothesis testable. Second, credit the bottleneck: purification was as necessary as biological activity. Third, treat an imperfect result as information; the first human attempt led to a better preparation. Fourth, measure outcomes close to the mechanism while tracking what patients actually experience. Fifth, count manufacturing and access as part of translation.

The insulin story is why history belongs beside evidence appraisal. It hands you something more useful than a polished legend: collaboration, revision, and unfinished work. That is the broader purpose of medical humanities in clinical practice and the site's research framework.