Diabetes mellitus: a look at the past, a glimpse to the future

Medical Department M
Aarhus Sygehus and University of Aarhus

Between the time of Aretaeus of Cappadocia’s accurate clinical description of diabetes almost 2000 years ago and the introduction of insulin around 90 years ago, knowledge of the disease and its treatment had advanced little. Type 2 diabetes is a combination of insulin resistance, partly related to lifestyle and obesity, and progressive loss of β-cell function. It has reached almost epidemic proportions in the West and Asia. Treatment is often ineffective, eg, 15% of US diabetics have an HbA1c of 10% or more. In Danish diabetics, the mean HbA1c is 8.0%, well above the 7% target proposed by the American Diabetes Association. Diagnosis, although simple, tends to be delayed, compounded by an often long clinically silent phase. The bulk of the world’s insulin is now injected by type 2 diabetic patients, after failure of diet and oral antidiabetic drugs. New drugs, such as glitazones and incretin modulators, are increasingly being used in the US, but long-term end point trials are not yet available. Intensive multifactorial intervention focused on lowering blood pressure and lipid and glucose levels is the key strategy in microalbuminuric patients. Landmark studies include the United Kingdom Prospective Diabetes Study (UKPDS) and Action in Diabetes and Vascular disease: PreterAx and DiamicroN MR Controlled Evaluation (ADVANCE), together with A Diabetes Outcome Progression Trial (ADOPT) and the Steno 2 study on sulfonylureas. Education and new therapeutic strategies have greatly improved clinical management, but, interestingly, some of the great steps forward in diabetes care and treatment have been surprisingly serendipitous.

Medicographia. 2011;33:9-14 (see French abstract on page 14

We owe the first description of a disease resembling type 1 diabetes, in the second century AD, to Aretaeus of Cappadocia. His writings contain a description that is accurate and clinical: “diabetes is an awful disease melting the body and limbs of the patient into urine. Life is short and painful and sooner rather than later the patient will expire.” Needless to say, he had no way of treating the patients he saw. Hippocrates never mentioned diabetes.

Pre-Aretaeus there are hints of diabetes in certain hieroglyphs and in accounts from India, where ants were drawn to sweet urine (almost a biological test for diabetes). In the past, type 2 diabetes was a disease of the rich, who were often obese from overeating. Nowadays, it has become more common among the poor and less educated. It is highly prevalent in certain populations, such as the Nauru Islanders, many of whom became obese and diabetic on the income from their phosphate de- posits, which were exhausted by the 1980s. Another muchstudied example is that of the Pima nation in Arizona, especially since they began earning money by running their own casinos. Diabetes remains unusual among Pima Indians south of the border in Mexico, where they are lean and hardworking physically. A decade ago, there was a move among the Pima to break off cooperation with researchers who, they claimed, had studied rather than treated them: National Institutes of Health (NIH) figures showed an increase in diabetes rates in the over-55s, from 40% in 1965 to 80% in 1999. Lifestyle might thus be more important than genes.

Figure 1
Table I. Twelve key points about diabetes.

The idea that lifestyle contributes to the development of type 2 diabetes was proposed in France over a century ago by Étienne Lancereaux (1829-1910). He classified diabetes into diabète maigre (“lean diabetes”) and diabète gras (“fat diabetes”), equivalent to diabetes types 1 and 2. Increased affluence and decreased physical activity have since created a near-epidemic of type 2 diabetes, which began mainly in the postwar US, but spread worldwide to China, India, and Japan, and to many countries in the Middle and Far East (but not among their guest workers). Treatment often remains ineffective, eg, 15% of US diabetics have an HbA1c of 10% or more. In Danish diabetics, the mean HbA1c is 8.0%, well above the 7% target proposed by the American Diabetes Association.

Diabetes is normally classified using clinical criteria and blood glucose data, although some centers also like to measure C-peptide levels to see if patients produce enough insulin. In obese patients with or without type 2 diabetes, C-peptide levels are often high due to a compensatory increase in insulin production, which decreases over time. No strict guidelines are available, and C-peptide and insulin assays are difficult to use in diagnosis as the results are not well standardized. For an illustration of the main facts concerning diabetes and its treatment, see Table I.

The discovery of insulin and beyond

Since the 1920s, we have had insulin to treat diabetic ketoacidosis and type 1 diabetes, but there are still regions in the world where not only insulin, but even proper diagnosis remain a luxury. Patients are still dying undiagnosed.

The bulk of the world’s insulin is injected by type 2 diabetics whose endogenous insulin production we now recognize as deteriorating over time due to a natural history of β-cell failure, as in type 1 diabetes. The sheer number of diabetics partly accounts for the enormous and increasing use of insulin, but most are also highly insulin resistant, requiring much more than the 40 units of insulin produced per 24 hours by healthy individuals. A dose of 200 units is not uncommon. It is axiomatic that the discovery of insulin for human use in Toronto in the early 1920s is the greatest event in the history of diabetology. The first patient, 14-year-old Leonard Thompson (1908-1935), was a typical type 1 diabetic. His beforeand after-treatment pictures—from ketotic emaciation to outward normality—went around the world. In his case, there was no time for a proper clinical trial, but it was hardly needed. The treatment was a revolution, instantly transforming a death sentence into a manageable, if lifelong, disease.1

But how did insulin come to be produced worldwide, in particular in Europe, and especially Denmark? Shortly after receiving a Nobel Prize in 1920 for his studies of capillary systems, the Danish zoophysiologist August Krogh (1874-1949) embarked on an obligatory lecture tour of the world’s important university hospitals, which included Toronto’s. There he agreed with Frederick Banting’s departmental head, John Macleod (1876-1935), to take care of insulin production in Denmark (and possibly Europe) on a nonprofit basis. As far as we know, no contract or patent was ever involved: it was a true gentlemen’s agreement. Behind the agreement lay a personal motive: Krogh’s wife Marie suffered from diabetes (although not of an aggressive type; her death in 1943 was due to breast cancer).

On his return home in 1923, Krogh promptly founded a laboratory, the Nordisk Insulinlaboratorium, and also a foundation, the Nordisk Insulinfond, which became an active sponsor of research. From the start, he was partnered by Hans Christian Hagedorn (1888–1971), a cofounder of the Steno Diabetes Center. Porcine insulin is close in chemical structure to human insulin, and being a major pork producer Denmark was well positioned to manufacture insulin from porcine pancreas. Heparin production in Denmark today shares a similar link; it is produced from porcine intestine by Leo Pharma, a company that was also briefly involved in insulin production at an early stage.

After early disagreements with the reputedly fiery Hagedorn, two remarkable brothers, Thorvald and Harald Petersen, left the Nordisk Insulinlaboratorium in 1924 to set up a rival production company, Novo, a year later. The rivalry continued for 65 years until the two companies merged to form Novo Nordisk, launching the world’s first prefilled insulin syringe the same year, following the introduction 4 years earlier of the NovoPen.

Porcine insulin is now rarely used in Europe and insulin production is based on gene technology. Rapid-acting insulin soon became available along with long-acting insulins, such as insulin glargine (Lantus®) and insulin detemir (Levemir®). Despite the popularity of the latter, it is difficult to demonstrate that these provide better glycemic control than that obtained with neutral protamine Hagedorn (NPH) insulin (eg, Insulatard®).

Leif Sestoft developed the insulin pen idea while working at Hvidovre Hospital in the Copenhagen suburbs. A 2007 analysis found no major difference in renal outcome, measured by glomerular filtration rate, between pen and pump insulin delivery, although continuous infusion achieved superior glycemic control.2 The pen was a major breakthrough for treating both types of diabetes. Reliable and user-friendly, it is now used by most insulin-requiring diabetics in many countries. Although it is likely that improved long-term glycemic control using pen injection will result in fewer complications, this has so far been difficult to document. Perhaps the single most successful application has been in the facilitation and marked improvement of glycemic control in diabetic pregnancy. Despite the dramatic developments in insulin analogs and delivery systems, the basic principles of treatment remain the same, as do its problems: exogenous insulin does not travel via the portal vein to the liver as does pancreatic insulin, nor are the doses of exogenous insulin accurately titrated to blood glucose concentration. Complete normalization of glycemia, as measured by HbA1c, thus remains elusive. Even the status of HbA1c as a marker has been called into question, on the basis of its sometimes less-than-linear relationship with estimated average glucose and random glucose levels.

Pharmacological treatment with oral antidiabetic drugs

Until the 1950s, insulin was the only pharmacological treatment of diabetes types 1 and 2, and it was serendipity that was responsible for the next paradigm shift in antidiabetic therapy. Marcel Janbon, an infectious disease physician experimenting with a new sulfonamide to treat the numerous cases of typhoid fever in wartime Montpellier (this was 1942), spoke to physiologist colleague Auguste Loubatières (1912- 1977) about his findings. Janbon reported posttreatment convulsions, prolonged coma, and severe falls in blood glucose in some of his patients to Loubatières, who had been conducting diabetes research in dogs during the previous decade.3,4 Although in 1946, after several more years’ work, Loubatières concluded in his doctoral thesis that sulfonamide was an insulin secretagogue acting directly on the pancreas, it was not until a decade later in Germany that the first sulfonylureas were developed for use in diabetes. They have since been widely used in trials such as the United Kingdom Prospective Diabetes Study (UKPDS)—including in its 10-year followup, the Action in Diabetes and Vascular disease: PreterAx and DiamicroN MR Controlled Evaluation (ADVANCE), and the Steno 2 study. In the latter two trials, gliclazide proved effective and devoid of cardiovascular side effects. Once-daily dosing with the modified release preparation greatly improves compliance.

In Europe, and subsequently in the US after a lag of several decades, the biguanide metformin was introduced as an insulin sensitizer and became widely used in combination with a sulfonylurea. Sulfonylureas lower HbA1c, often by 1.5% to 2%, and even more so when combined with metformin. Newer drugs, such as glitazones, which are not extensively used in Europe due to doubts about late effects and weight gain, cause a fall of between 0.5% and 1%. One study of particular interest where sulfonylureas are concerned, A Diabetes Outcome Progression Trial (ADOPT), compared three oral therapies in newly diagnosed type 2 diabetes: rosiglitazone, metformin, and glibenclamide.5 Baseline characteristics were similar in the three groups, each comprising around 1450 patients with blinded follow-up over 4- 5 years.

Tables IIA and IIB
Tables IIA and IIB. Advances in clinical trials and concepts of diabetes.

Abbreviation: EASD, European Association for the Study of Diabetes.

All-cause mortality was similar with all three therapies. Cardiovascular risk was lowest with glibenclamide, but hypoglycemia was most commonly seen with this drug (0.6% of episodes were considered serious). β-Cell function remained the same after 5 years, which is remarkable. Hospitalization and fractures were rarest with glibenclamide, and, in addition, less anemia was observed. With rosiglitazone, the rate of monotherapy failure was lowest, but weight gain and edema were more frequent. Metformin was associated with more frequent gastrointestinal events.

Figure 1
Figure 1. Major trials and conceptual developments
in type 2 diabetes from the 1940s to 2010.

Abbreviations: ADVANCE, Action in Diabetes and Vascular disease: PreterAx and DiamicroN MR Controlled Evaluation; BG, blood glucose; BP, blood pressure; IDF, International Diabetes Federation; NEJM, New England Journal of Medicine; SU, sulfonylurea; UGDP, University Group Diabetes Program; UKPDS, United Kingdom Prospective Diabetes Study.

Cost was factored into choice of treatment by the authors. Despite their initial bias in favor of rosiglitazone (the study drug in this comparison), it is not difficult to see why their data confirm the European preference for the familiar treatment stalwarts of sulfonylurea, metformin, and insulin.

Table IIA lists the major clinical trials in recent decades along with their key messages, while Table IIB shows conceptual innovations that have reshaped our day-to-day management of diabetes. Many of these trials were conducted in response to, or as an extension of, a trial that had gone before (Figure 1), at the same time as they attempted to resolve one of the keenly debated issues of day, beginning with the debate between Edward Tolstoi and Elliott Joslin (1869- 1962) on whether lax or strict glycemic control is better. This debate, along with several others, can now be laid to rest (Table III).

Table III
Table III. Much debated issues, past and present, concerning
diabetes and the kidney.

Abbreviations: ACE, angiotensin-converting enzyme; ARB, angiotensin receptor blocker.

Biomarkers for predicting complications

In 1984, we showed that microalbuminuria predicted not only renal disease, but also early mortality in both type 1 and type 2 diabetes (Figure 2).6 Microalbuminuria, an excellent marker of complications, was the basic instrument used in the Steno 2 study launched in 1992. Renal biopsy is very rarely required for diagnosing diabetic nephropathy, the only indication being atypical onset of sudden proteinuria. Steno 2 allocated type 2 patients with microalbuminuria to intensified combined intervention (glycemic control using gliclazide, lipid lowering, and blood pressure lowering) or conventional multifactorial treatment. The effect on mortality and end-stage renal disease after 8 years was dramatic. As a result, this treatment has become standard for microalbuminuric patients and even for some normoalbuminuric patients (mainly those with poor control and other risk factors). The Anglo-Danish-Dutch study of Intensive Treatment In peOple with screeN detected diabetes in primary care (ADDITION) used the same multifactorial strategy, but in newly diagnosed patients. In contrast, the results presented at the European Association for the Study of Diabetes (EASD) meeting in Stockholm in September 2010 were negative.

Figure 3
Figure 3. Ten partly or totally serendipitous discoveries in diabetes (outer circle) that radically improved diabetes care (inner circle).

Abbreviations: ESRD, end-stage renal disease; GFR, glomerular filtration rate; HbA1c, glycated hemoglobin; RAS, renin-angiotensin system, SU, sulfonylurea.

Table IV
Table IV. Possible future developments in diabetes.

What new prospects are on the horizon?

Although difficult to document in properly designed trials, our daily experience tells us that self-monitoring and the introduction of classes in which patients learn from specialists, nurses, and dieticians have transformed the clinical management of diabetes. It is interesting to note that the advances described in the editorial and in this paper have often been serendipitous (Figure 3, page 13), although it is no accident that most have emerged from well-established centers populated by prepared minds (discoveries rarely come out of the blue). Some have even created paradigm shifts.7 There is no reason to believe that this pattern is likely to change. Of the prospects listed in Table IV, a number are of “Holy Grail” status, unabashedly so: it is only by keeping them in our minds that we will recognize them when they present themselves in the most unexpected of guises. _

1. Bliss M. The Discovery of Insulin. Chicago, Ill: Chicago University Press; 1982.
2. Schmitz A, Christiansen JS, Christensen CK, Hermansen K, Mogensen CE. Effect of pump versus pen treatment on glycaemic control and kidney function in long-term uncomplicated insulin-dependent diabetes mellitus (IDDM). Dan Med Bull. 1989;36:176-178.
3. Janbon M, Chaptal J, Vedel A, Schaap J. Accidents hypoglycémiques graves par un sulfamidothiodiazol (le VK 57 ou 2254 RP). Montpellier Med. 1942;441: 21-22.
4. Loubatières-Mariani MM. [The discovery of hypoglycemic sulfonamides]. J Soc Biol. 2007;201:121-125.
5. Kahn SE, Haffner SM, Heise MA, et al; ADOPT Study Group. Glycemic durability of rosiglitazone, metformin, or glyburide monotherapy. N Engl J Med. 2006; 355:2427-2443.
6. Mogensen CE, Christensen CK. Predicting diabetic nephropathy in insulin-dependent patients. N Engl J Med. 1984;311:89-93.
7. Mogensen CE. Blood pressure, blood glucose, and diabetic renal disease. Medicographia. 2009;31:299-306.

Keywords: glycated hemoglobin; paradigm shift; serendipity; microalbuminuria; multifactorial intervention; ACE inhibition; sulfonylurea; insulin; metformin