EDITORIAL Insulin Glargine (HOE901) First responsibilities: understanding the data and ensuring safety n this issue, Heinemann et al. (1) and Rat- decline in activity through the duration of other journals. They leave the reader with ner et al. (2) provide the results of stud- the 30-h study. The authors conclude that a sense of promise, but without the firm Iies of insulin glargine (formerly known as insulin glargine provides a flatter metabolic conviction that insulin glargine is clearly HOE901), the most recent addition to the profile than NPH insulin. Theoretically, this superior to other long-acting forms of growing family of insulin analogs and for- could prove to be advantageous in accom- insulin. The pharmacodynamic study sug- mulations. The authors report insulin modating the basal insulin requirements of gests that the peakless profile of action glargine’s pharmacodynamic characteristics patients with diabetes. should provide an advantage that can be (1) and data concerning its safety and effi- Ratner et al. (2) report initial results for used to produce clinically meaningful cacy when administered to subjects with the U.S. Study Group of Insulin Glargine in improvements in glycemic control. The type 1 diabetes who were treated with reg- Type 1 Diabetes. In their article, they clinical trial’s demonstration of a reduction ular insulin (2). Insulin glargine is produced describe the results of a large randomized in hypoglycemia and in fasting plasma glu- by recombinant DNA technology with 2 prospective 28-week trial of insulin cose levels appears to provide an opportu- modifications of the native human insulin glargine versus NPH insulin. Patients with nity to advance further glucose lowering to structure: substitution of the amino acid type 1 diabetes who were previously achieve a reduction in GHb. However, in glycine for the native asparagine at position treated with regular and NPH insulin par- this study, no change in GHb was evident. A21 of the A-chain of human insulin and ticipated in the study. Prior reports of clin- This characteristic does not lessen glargine’s the addition of 2 arginine molecules to the ical efficacy with insulin glargine have potential significance. Reducing hypogly- NH2-terminal end of the B-chain of human involved short-term studies and few are cemia is a worthy end in itself. Several insulin. The resulting insulin glargine available except as abstracts or in review recent reports have attested to the impor- demonstrates an isoelectric point at pH 6.7, articles (3). This protocol could not be tance of reducing the frequency and sever- which is in contrast to the native molecule’s masked, because glargine is a clear solution ity of hypoglycemic reactions. isoelectric pH of 5.4. As a result, glargine is and NPH insulin is a milky suspension. In Currently, the available data concern- soluble at acidic pH and less soluble at the study, subjects who were randomized ing insulin glargine create a situation simi- physiological pH. Insulin glargine is sup- to insulin glargine administered it as a sin- lar to the circumstances in 1996, when plied as a clear colorless solution at acidic gle bedtime injection. Subjects random- lispro insulin was first marketed in the U.S. pH balances. Upon subcutaneous injection, ized to NPH insulin remained on the NPH In the case of lispro insulin, good pharma- the acid in the vehicle is neutralized and schedule that they were using before ran- codynamic data and promising studies sug- glargine precipitates, thereby delaying its domization (once daily at bedtime or twice gested that hypoglycemia reduction was absorption and prolonging its duration of daily before breakfast and at bedtime). possible. Studies demonstrating improve- action. At the time this editorial was written, Dose titration of both insulin glargine and ments in GHb have only recently been the Food and Drug Administration (FDA) NPH insulin was performed on the basis of published (4,5). In particular, the research was conducting its review of insulin capillary blood glucose monitoring results. of Lalli et al. (5) demonstrates the difficulty glargine. The goal was to achieve a fasting blood glu- of translating the theoretical benefits of new Heinemann et al. (1) document the cose level between 4.4 and 6.7 mmol/l therapy into clinically significant benefits. results of single-dose double-blind (80–120 mg/dl), as limited by symptomatic The reported regimen used combinations crossover euglycemic clamp studies in hypoglycemia. The protocol examined of lispro and NPH in various ratios at each healthy male volunteers and compare the glycemic control and hypoglycemia as end meal. Certainly, the initial patients of the time-action profiles of subcutaneously points. Insulin glargine was well tolerated study group were not treated with the reg- injected glargine, NPH insulin, and without evidence of antibody formation imen that they eventually studied. Adapta- placebo. Prior pharmacodynamic studies and with only minimal injection site reac- tions were required due to the context of using insulin glargine have evaluated for- tions. Insignificant reductions from base- the society in which the patients lived mulations with higher and lower zinc con- line in glycohemoglobin (GHb) were (Italy) and because of the content of the centrations than the current formulation demonstrated in both groups. Insulin patient education program that was avail- (zinc 30 µg/ml), and they have only been glargine demonstrated a significantly able to them. The researchers’ understand- reported in abstract form or in review arti- greater reduction in fasting plasma glucose ing of the pharmacodynamics of lispro and cles. The work of Heinemann et al. demon- levels. After the initial titration phase their extensive clinical experience resulted strates that glargine provides an essentially (2 months), there also was a significant in the development of a technique that peakless profile of action with its onset of reduction in hypoglycemia, including noc- worked well for their patients. Their results action at 2–4 h and with a duration Ͼ24 h. turnal and severe hypoglycemic episodes. are nothing short of spectacular, and they By comparison, NPH human insulin was These two articles are among many are clearly the product of painstaking clin- again documented to provide a substantial studies conducted with insulin glargine ical observation complemented by excel- peak in action at 4 h with a subsequent that will be published in Diabetes Care and lent clinical research. Clinicians have 576 DIABETES CARE, VOLUME 23, NUMBER 5, MAY 2000 Editorial worked very hard to develop techniques to insulin (instead of regular insulin) for the coincide with the launch of insulin use the insulin formulations that have been therapy of type 1 diabetes is rapidly glargine. By contrast, it is an outrage that available for decades. It will inevitably take increasing in specialty clinics and offices, the data for a number of other products time to learn how to actually exploit the the use of glargine (in all probability) will marketed for various indications in the theoretical advantages of insulin glargine. be rapidly adopted in those practices. It is therapy of diabetes are available only under At the University of North Carolina, an even within the realm of possibilities that Freedom of Information Act queries to the insulin glargine study site, we tried to iden- the “poor man’s pump” (multiple daily FDA. It is even more infuriating when tify subjects who had done poorly on mul- injection regimens using insulin pens) may claims are implied by data presented in tiple daily injection regimens, despite prove to be as effective, less expensive, and promotional material, although there are maximal interventions. This was driven by more convenient than continuous subcu- neither available raw data nor plans to con- our unwillingness to change the insulin taneous insulin infusion with the use of duct the appropriate trials to fully investi- regimens of our patients who were doing insulin glargine. Data that suggest the ben- gate those claims. well. The subjects we sought would have efit of glargine therapy in type 2 diabetes Second, as a community of diabetes been excellent candidates for insulin pump also exist in abstract form. Full publication professionals, we need to help the pharma- therapy. However, they had either refused of those results in combination with clini- ceutical industry understand the limita- pump therapy or faced financial barriers cal experience may reveal the potential of tions of their data, whether published or that prevented them from using an insulin insulin glargine as an insulin supplement in promotional. We need to suggest trials pump. Other patients enrolled in the study, patients with moderate insulin deficiency and/or gather clinical data in a systematic motivated by personal interest or a desire to and as a basal insulin replacement in fashion to highlight the need for further mitigate health care expenses. Because the patients with more complete insulin insuf- study or to generate hypotheses. As an study was not masked, both subjects and ficiency. Many other potential uses of example, the reviewers of this manuscript investigators were fully aware of what for- insulin glargine will also require specific have raised the issue of how the use of mulation of insulin each individual was study before they can be widely consid- insulin glargine would compare with the taking. As Ratner et al. (2) suggest, this may ered, such as its use in pregnancy and as a use of human ultralente insulin. Given that have subconsciously limited our enthusi- form of immunomodulatory therapy in the there are both strong proponents and asm to aggressively manipulate the dose of prevention of autoimmune diabetes. opponents of ultralente use, this topic an unfamiliar product in subjects who were The limitations of the insulin glargine appears to be deserving of exploration in, at often quite well controlled.