VOL. 18 NO. 4 Fall 2000


The New "Designer" Insulins

  Howard M. Lando, MD, FACE, FACP


In the past year, several new "designer" insulins have become available for the treatment of patients with diabetes. These include glargine (Lantus), a peakless insulin with a 24-h duration of action mimicking a basal rate; aspart (Novolog), a very-quick-acting insulin analog similar to lispro (Humalog); and a 75/25 lispro mixture (Humalog Mix 75/25), which is a stable premixed insulin using lispro. The use of these new insulins will add a further dimension to the care of patients with diabetes to improve glycemic control and, ultimately, outcomes related to complications.

Since its discovery in the early 1920s by Nobel laureates Banting and Best, insulin has been one of the mainstays of treatment for diabetes.1 It is a 51­amino acid protein composed of an alphasmll.gif (897 bytes)- and a beta.gif (968 bytes)-chain linked by two disulfide bonds and produced initially as a prohormone called proinsulin in the beta.gif (968 bytes)-cell of the pancreas. Proinsulin is then cleaved into the active hormone insulin, and, in equalmolar amounts, its connecting bridge, called C-peptide. (Figure 1).

Because insulin has a short half-life, it was initially used as subcutaneous injections given shortly before each meal. Because of this "disadvantage," attempts were made soon after its discovery to prolong its duration of action by adding other substances to its formulation to retard its absorption from injection sites. These substances initially included gum arabic, lecithin, and oil suspensions, and later, protamine (NPH) and zinc (lente, ultralente, semilente).2

Diabetes Cover logo.GIF (10208 bytes)
Figure 1: Insulin

Initially, insulin was purified by grinding the pancreases of pigs and cows. Human insulin did not become available until 1979,3 with the advent of genetic engineering. It was originally made by adding the human insulin gene to a special strain of nonpathogenic Escherichia coli and soon thereafter by the chemical substitution of a single amino acid threonine for alanine at the terminal 30 position in the beta.gif (968 bytes)-chain of the porcine molecule.

These two processes have led to the creation of newer insulin analogs genetically engineered with further chemical substitutions to give them different absorption characteristics. When combined, these newer insulin analogs are better able to match normal physiological insulin production from the pancreas.

The first of these insulin analogs, insulin lispro (Humalog, Eli Lilly), was made by the reversal of the amino acids at the 28 and 29 positions of the beta.gif (968 bytes)-chain.4 This reversal produces an insulin that is absorbed more quickly than the native molecule and thus lowers postmeal blood glucose more rapidly, which could lead to better overall control.

In the past 12 months, the Food and Drug Administration (FDA) has approved two more insulin analogs and one new analog combination. These new insulins have been added to the armamentarium of diabetes treatments to help patients achieve the tight blood glucose control that the Diabetes Control and Complications Trial and the United Kingdom Prospective Diabetes Study demonstrated to be essential in delaying the development and progression of diabetes complications.

The first of these, insulin glargine (Lantus [HOE 901]), is produced by the Aventis Company (a recent merger of Hoechst Marion Roussel and Rhone Poulenc Rorer). This drug has the greatest potential for changing the way diabetes is treated at present.

Glargine is a recombinant human insulin analog that acts as a basal insulin in that it is absorbed equally throughout a 24-h period. Like most other human insulins presently available, it is produced by recombinant DNA technology utilizing a nonpathogenic strain of E. coli. It differs from native human insulin in that there is an amino acid substitution of asparagine with glycine at the 21 position of the alphasmll.gif (897 bytes)-chain, and two arginines are added at the C-terminus of the beta.gif (968 bytes)-chain. Unlike other available insulins, it has an acidic pH of ~4. Also, unlike most longer-acting insulins, it is a clear, colorless solution—not cloudy.5 These changes alter the disassociation properties of glargine, making its hexamer structure more stable and thus more slowly absorbed. Also, small amounts of zinc have been added, which further prolongs its duration of action.2

Once injected subcutaneously, the acidic solution is neutralized, which forms microprecipitates that slowly release this insulin in a relatively constant and thus peakless profile over 24 h. A euglycemic glucose clamp study using healthy volunteers compared the peak of glargine to that of NPH. Glargine had a peakless and flatline (square-wave) absorption lasting 24 h, whereas NPH showed its usual peak absorption at 4 h.6 Other studies have shown a similar effect in diabetic subjects.

Because glargine is a clear solution, there is no need to thoroughly mix it. This differs from NPH, lente, and ultralente insulins, for which inadequate mixing may lead to sedimentation of the insulin crystals and inhomogeneities in the suspension. This, in turn, could result in inaccurate dosing and thus hyper- or hypoglycemia. Glargine's clarity also makes it easier to identify "frosting," the precipitation of insulin on the walls of the vial due to freezing or overheating.2

The results of three large, controlled trials comparing glargine to NPH in individuals with type 1 diabetes were submitted to the FDA during the approval process. In the first two studies, one involving 585 patients and the other 534 patients, subjects were randomized to basal-bolus treatment with either NPH and regular insulin or glargine and regular insulin. The NPH was given once or twice a day with regular insulin at each meal. The glargine was given at bedtime with regular insulin at each meal. Both studies lasted for 28 weeks. The third study was of similar design, except it enrolled 619 patients, lasted 16 weeks, and used insulin lispro rather than regular insulin. All three studies showed a similar effect on HbA1c and a similar overall rate of hypoglycemia. However, there appeared to be a slightly greater decrease in total insulin dose and fasting blood glucose at the end of each of these studies with the glargine groups.

There has also been one large, randomized, controlled study of glargine in pediatric patients (age 6–15 years) with type 1 diabetes. This 28-week study enrolled 349 patients using a basal-bolus regimen of glargine once at night or NPH once or twice daily. The results were similar to those in the three studies of adults with type 1 diabetes.5

The results of two large studies of glargine in adults with type 2 diabetes were also submitted to the FDA. The first involved 570 patients treated for 52 weeks with either glargine or NPH at bedtime plus oral agents (sulfonylurea, metformin, acarbose) singularly or in combination. Unlike the studies mentioned above, in this study, the basal insulin and the total insulin dose both increased, whereas fasting glucose and HbA1c decreased equally.

The second type 2 diabetes study was of 518 patients not on oral agents but instead on a basal-bolus insulin regimen of glargine at bedtime plus regular insulin or NPH once or twice a day plus regular insulin. Much like the type 1 studies, this study lasted 28 weeks and showed a decrease in the HbA1c and fasting blood glucose with a slight increase in the basal insulin dose in the NPH group and a minimal decrease in basal insulin using glargine.5

The six glargine studies are summarized in Table 1. As with all insulins, hypoglycemia is the most common adverse effect with glargine. In a large, multicenter, randomized parallel-group study of 534 patients with well-controlled type 1 diabetes using a basal-bolus regimen, hypoglycemia (both symptomatic and nocturnal) occurred less frequently in the glargine group than in the NPH group despite a lower fasting blood glucose in the glargine group.7

Table 1. Insulin Glargine Studies

Study No Diabetes type

Therapy time (weeks)

Combination Agent n


Basal Insulin

Total Insulin


End delta.gif (936 bytes)* End delta.gif (936 bytes)* End delta.gif (936 bytes)* End delta.gif (936 bytes)*
Type 1


Type 1
28 Regular

P = NS

P = 0.01

Type 1
16 Lispro 310
Type 1
28 Regular
Type 2 52 Oral Agents 289
Type 2 28 Regular

P = NS

P = NS

*From Baseline
FBG, fasting blood glucose

When data from as yet unpublished U.S. and European studies of glargine in type 1 diabetes were pooled, the incidence of nocturnal hypoglycemia was almost half (16 vs. 30%, P < 0.05) as was severe hypoglycemia (3 vs. 7%, P < 0.05).8 Thus, it appears that glargine with either regular or lispro insulin may be a safer combination for intensive insulin therapy than other insulin combinations presently available.

It should be noted however that hypoglycemia may occur at different times with different insulin preparations. When switching from one preparation to another, careful observation and frequent home blood glucose monitoring are essential. Glargine's manufacturer recommends that if a patient is changing from a twice-daily NPH schedule to glargine, a 20% reduction in the dosage should be considered. With a once-a-day schedule, no decrease in dose is advocated.5

Information is not yet available regarding glargine use in the presence of either renal or hepatic impairment. In the former, insulin metabolism is usually decreased, whereas in the latter, reduced gluconeogenesis and insulin metabolism reduce insulin requirements. Until such studies are conducted and reported, it would be prudent to consider reducing the dose of glargine as would be done with other insulins.

No studies of glargine use in human pregnancies have been conducted. Therefore, the FDA has assigned this drug to its Category C—agents to be used only if clearly needed during pregnancy. It is also not yet known whether glargine is excreted in significant amounts in human milk. However, other human insulins are, so like the others, glargine should be used with caution in nursing mothers.5

Few adverse reactions to glargine have been reported in the studies to date except for the hypoglycemia discussed previously. There was a higher incidence of injection-site pain in the glargine-treated patients (2.7%) versus those treated with NPH (0.7%). This pain was mild and did not result in cessation of treatment in any subjects.5

Glargine cannot be used intravenously because severe hypoglycemia could result. It also cannot be mixed with other insulins because the solution could become cloudy, and the pharmacokinetics of each insulin would probably be altered.

To aid in distinguishing glargine from other clear insulins (regular, lispro, insulin aspart [Novolog]), its manufacturer is placing this product in a different shaped vial with a lavender color code. It will also be available in cartridges for a pen delivery system.

Thus, physicians now have another and perhaps better insulin to use in both type 1 and type 2 diabetic patients that may make optimal control easier to achieve. Because it mimics a basal secretion of insulin, glargine with either a quick-acting insulin in type 1 or type 2 diabetes or perhaps repaglinide (Prandin) in type 2 diabetes, will give patients an easy-to-administer basal-bolus sequence—something that until now has only been available with insulin pumps. This sequence will most likely allow patients to follow a more normal mealtime pattern that is individualized to their own habits, likely resulting in better control, less hypoglycemia and nocturnal hypoglycemia, and in many cases, a lower insulin requirement.

The potential importance of this new therapy is clear. Among the impediments to tight blood glucose control are patients' inability to take insulin at their scheduled times, their perceived inability to eat meals when they want to rather than when their insulin schedule dictates, and thus the increased possibility of hypoglycemia. Because they offer freedom from such concerns, insulin pumps are becoming more popular. However, many patients refuse to use pumps because they fear being tied down to a mechanical device. Also, many physicians fear the extra work and staff necessary to have a pump program and will not even consider insulin pumps for their patients.

Glargine changes all of that. With its ability to mimic a basal rate and with other drugs' ability to mimic the bolus rates of a pump, patients can get a similar pattern without the cost of the pump and supplies and without much more difficulty than they already face in terms of home blood glucose monitoring or physician visits. For physicians, glargine is just another insulin and should fit into their practice patterns easily.

Glargine is not a panacea, however. It will probably not have a perfect basal absorption in all patients, and it will have to be taken separately from other insulins. Its clear formulation leaves it open to mistakes, such as taking the wrong type or amount of insulin at the wrong times. But hopefully the different color code, vial shape, and pen will allay this concern. Patients with diabetes will still have to perform home blood glucose monitoring, and, in fact, 2-h postprandial glucose readings become almost more important in a basal-bolus pattern.

As a basal insulin, it will probably be able to substitute for intermediate insulins used in combination with other oral agents in type 2 diabetes. However, no studies on such uses have been conducted.

As Dr. Robert Ratner, one of the investigators of this insulin, said, "When we finished the study and the patients had to go back to their old regimens, most asked if there was a way to get this insulin immediately and not wait for it to come out on the market [paraphrased]."

The most recent insulin analog to be approved by the FDA is insulin aspart (Novolog), a product of Novo Nordisk. It is a very-rapid-acting insulin analog with absorption characteristics similar to lispro. It is homologous to human insulin except for the substitution of aspartic acid for proline in the 28 position of the beta.gif (968 bytes)-chain. This insulin analog also uses recombinant DNA technology but uses saccharomyces cerevisiae (baker's yeast) as the producing organism9 rather than nonpathogenic E. coli.

The amino acid substitution reduces the ability of the molecule to form hexamers, which allows it to be absorbed faster than regular insulin after subcutaneous injection. Regular insulin has a strong affinity to form hexamers in solution. These must be broken down in the tissues into dimers and then into monomeric form to be absorbed.10

Aspart is a clear, colorless solution like all rapid-acting insulins and also like glargine. Its onset of action is within 15 min of injection so that it can be taken just before a meal rather than 30 min before, as is the case with regular insulin. In a study of 107 healthy volunteers and 40 subjects with type 1 diabetes, aspart reached maximum peak serum concentrations within 40–50 min as compared to 80–120 min for regular insulin. In the diabetic group, aspart's peak was twice as high with equal amounts of insulin injected per body weight.9

Aspart's duration of action is 4–6 h. Its rapidity of action is its true benefit for most type 1 and type 2 diabetic patients. Glucose concentrations rise rapidly after meals and have not been well controlled by previous nonanalog insulins because of their delayed onset and peaks of action. Aspart more closely mimics the pre-formed insulin in the beta.gif (968 bytes)-cell that is missing in diabetes. It therefore allows much more freedom in the timing of meals and the amount of food ingested for patients who are well trained in managing their blood glucose.

The pharmacodynamics and kinetics of aspart have not been found to differ between sexes or in studies involving children. Like all insulins, package information for this product carries warnings regarding renal and hepatic impairment (i.e., a reduction in insulin dosage may be needed), although no studies have been done to date.

Raskin et al.11 compared aspart to regular insulin in a study reported at the American Diabetes Association (ADA) Annual Meeting and Scientific Sessions in June 1999. They randomized 884 patients with type 1 diabetes to aspart or regular insulin for 6 months and compared HbA1c and blood glucose values postprandially and before meals, as well as the incidence of hypoglycemia. They found a statistically insignificant improvement in HbA1c at 6 months (7.8 vs. 7.9%—though the difference from baseline HbA1c was 0.12 lower with aspart, P < 0.02) with better blood glucose values at all times with the aspart (P < 0.0001).

Raskin et al.'s extension trial enrolled 714 patients. At 12 months, there was a continued improvement in HbA1c (7.8 vs. 8.0%), with continued improvement in blood glucose values. Of particular note, the incidence of hypoglycemia was similar overall, but from midnight to 6:00 a.m., it was significantly improved (16% for insulin aspart vs. 34% for regular insulin, P < 0.05) despite the better blood glucose control.11

Similar results were obtained in two other studies—one by Rosenfalck12 et al. and the other by Home et al.13 In the first, a test meal was administered and pre- and postprandial blood glucose levels were compared with those of patients using regular insulin. Whereas there was a definite decrease in the postprandial glucose excursion if the insulins were taken at the same time immediately before the meal, there was no difference if the regular insulin was taken 30 min before the meal.12

Home et al., like Raskin et al., compared blood glucose values, HbA1c levels, and incidence of hypoglycemia. In their 6-month study, 707 patients used aspart immediately before a meal, and 358 patients used regular insulin 30 min before a meal. At the end of the study, there were lower HbA1c levels (P < 0.002), fewer nocturnal hypoglycemic events (P=NS), and better patient satisfaction (P < 0.0001), with no change in the amount of aspart or regular insulin used but an 8% increase in the amount of NPH used for good diabetic control in the aspart group.13

In another trial of 90 type 1 diabetic patients by Home et al., there were a similar number of hypoglycemic events, but there were significantly fewer major hypoglycemic episodes (requiring assistance from another person) in the aspart group than in the group using regular insulin—20 events in 16 subjects vs. 44 events in 24 subjects, (P < 0.002).14

Other abstracts involving aspart from the 1999 ADA meeting included one from Halberg et al.15 and one from Jacobson et al.16 The study by Halberg et al. compared separate versus mixed injections of NPH with aspart, showing no difference between the regimens. This confirms the ability to mix aspart with NPH without changing any of the very-rapid-acting aspart's absorption characteristics. However, it should be noted that there have been no studies to date studying mixtures of aspart and lente or ultralente insulin. Therefore, the safety and absorption characteristics of these mixtures cannot be ascertained. At present, they should only be used with caution.

The study by Jacobsen et al. compared a 30/70 premix of aspart plus protamine-retarded aspart compared to a conventional 30/70 mixture of regular and NPH insulin. In this study, the time of maximal insulin concentration for the aspart mixture was 60 min versus 110 min for the conventional mixture.16 Thus, when a premixed vial becomes available in the future, it will probably have characteristics similar to the native aspart.

As with other insulin analogs, there are no data regarding aspart use in pregnant women. Therefore, the product package bears the usual warning to use it during pregnancy "only if the potential benefit justifies the potential risk to the fetus."9 There are also no data regarding aspart's excretion in human milk. But again, because human insulin is excreted in human milk, one may surmise that aspart will be as well.

Like other insulins, aspart should not be frozen. Vials not being used should be refrigerated between 36 and 46° Fahrenheit. Insulin being used may be kept at room temperature out of direct sunlight or heat at < 86° Fahrenheit for up to 28 days.9 Health care providers must remind patients about this because patients often forget that their insulin may not be as effective after 1 month of being opened and in ambient temperatures. This could cause blood glucose control to gradually decline as patients try to use up all the insulin in an open vial before starting a new one.

Aside from some slight structural differences in the substitutions on the beta.gif (968 bytes)-chain of the molecule, aspart and lispro are equally effective and work essentially in the same manner to lower postprandial blood glucose rapidly and more effectively than regular insulin. Most physicians do not like to mix insulins from different companies in the same syringe because there are no data regarding the safety of such mixtures. Thus, the availability of aspart will allow those who must or wish to use Novo Nordisk's NPH to continue to do so while enjoying the benefits of better postprandial and overall diabetic control afforded by a rapid-acting insulin analog.

Lispro Mixture
A 75/25 lispro mixture (Humalog Mix 75/25) was approved by the FDA last year but was not released until this year. Made by Eli Lilly, this is a fixed mixture of their previously released very-rapid-acting lispro and a novel human insulin analog called NPL, which is insulin lispro protamine suspension. Adding protamine to lispro delays its absorption and changes its activity profile to that of NPH insulin except in a monomeric form.

This mixture was found to be necessary because, over a prolonged period of time (such as for formulation in a pre-mixed vial), an exchange occurs between lispro and human NPH. This exchange is not a problem when patients mix the two for immediate injection.17 The ratio of 75/25 was chosen based on previous experience with the amounts of each insulin, usually given by patients in a self-mixed ratio of lispro and NPH.18

As mentioned earlier, lispro is made by reversing the amino acids in position 28 and 29, proline and lysine, of the beta.gif (968 bytes)-chain of native human insulin. It is manufactured by recombinant DNA technology using a nonpathogenic strain of E. coli that has been altered by the addition of the gene for lispro. Lispro acts very quickly with a peak insulin concentration within 30–90 min in both normal volunteers and type 1 diabetic patients.19 This is why it can be used just before meals rather than 30 min before eating, as is necessary for subcutaneous regular insulin. Lispro also has a more rapid onset of glucose-lowering activity and a shorter duration of activity.

Mixtures are attractive options to many physicians because twice-daily injections of short- and intermediate-acting insulins are common. Because the 75/25 ratio is the most frequently prescribed combination, some believe that having a fixed mixture of that ratio will improve patients' compliance with such a regimen by increasing convenience and improving the accuracy of drawing up doses. Also, such a mixture can be placed in a pen device for greater ease of administration.

In a study by Lalli et al.,20 injections of a mixture of NPH and lispro or regular insulin were given at each meal, and NPH was also given at bedtime (not just twice a day). Improved blood glucose control was noted in both groups. There were 56 patients with type 1 diabetes enrolled in this open-label study.

Whereas there were no differences in the total amount of insulin used in either group, the lispro group used 30% less short-acting insulin and 30% more intermediate-acting insulin. Hypoglycemia was less frequent and HbA1c levels were improved over the 12 months of the study. What is most interesting is that, by giving some NPH at each meal, overall glucose control improved. This is probably because of the better basal insulin afforded by adding NPH at lunchtime and dinnertime to the usual morning and bedtime doses of NPH. While illustrating the need for ~70% NPH/30% lispro at breakfast, 60/40 at lunch, and 80/20 at dinner, this study makes a case for having mixtures, including premixed insulins of various concentrations, available.

There is already on the market a 70/30 premixed form of human insulin that is 70% NPH and 30% regular insulin. Other concentrations are available in Europe and elsewhere.

Several articles have compared the efficacy of lispro 75/25 to human insulin 70/30. The first, by Roach et al.,21 compared 84 patients with type 2 diabetes during a 6-month, randomized, open-label, crossover study injecting twice daily before breakfast and dinner. What was found, as in most other studies, was that the lispro 75/25 mixture resulted in better postprandial glucose control after breakfast and dinner when the lispro component of the mixture was most active (2-h excursion post-breakfast, P = 0.002; 2-h excursion post-dinner, P < 0.001). However, overall glycemic control as shown by HbA1c and the incidence of hypoglycemia were much the same. The only advantages noted by the authors were the improved postprandial control and the greater ease of administration in terms of time before meals.

Strack et al. and the NPL Study Group22 compared self-prepared mixtures of lispro and NPL versus regular insulin and NPH in an open-label, randomized, 12-month parallel study of 147 patients with type 1 or type 2 diabetes. There was very little difference in the ratios of short- to intermediate-acting insulins in both groups. As expected, the postprandial breakfast and dinner blood glucose levels were signi- ficantly lower in patients using lispro. In this study, however, there was a significant decrease in HbA1c levels of ~0.5% (7.95 vs. 7.44%, P = 0.001) in the lispro/NPL group after 12 months.

It should be noted that in both groups, the post-lunch and pre-dinner blood glucose levels were much higher, suggesting that in both regimens there was not enough NPL or NPH given at breakfast. Also one should note that this study did not use a fixed mixture but rather self-prepared mixtures. Still, the study showed the improved effectiveness of this analog combination.22

Finally, in another article, Roach et al. and the LM Study Group23 compared the 75/25 lispro mixture to an 80/20 human insulin mixture (80% NPH and 20% regular) in 53 individuals with type 2 diabetes for 6 months in an open-label, randomized, crossover study. Much like the first Roach study, there was little difference between the groups as far as glucose control as shown by HbA1c and better postprandial glucose levels post-breakfast and post-dinner. However, in this study, there was an improved nocturnal hypoglycemia pattern in those on the lispro mixture compared to those using the conventional insulin mixture (P = 0.034). One wonders if part of this was because there was less intermediate-acting insulin in the lispro mixture (75 vs. 80%) and thus less likelihood of hypoglycemia 6–10 h after injection when the intermediate-acting insulin peaks.23

Thus, premixed insulins can be easier to administer because they can be put in a pen device with concomitant better accuracy and improved transportability, perhaps leading to better patient compliance. Also, as in the Lalli study,20 mixtures that add NPH with each meal may improve glucose control.

The downside of premixed insulins is that one loses the ability to more accurately adjust each of the insulins in the mixture in terms of quantity and timing. This may be especially important for nocturnal hypoglycemia, where moving the intermediate-acting insulin to bedtime can obviate the problem by adjusting the peak time to more closely match the "dawn phenomenon" (early morning hyperglycemia caused by the physiological increase in cortisol production).

The new lispro mixture may also be a problem in some patients who are allergic to protamine, because both NPH and NPL are crystallized with protamine to delay absorption. Such patients can still be taught to self-mix other insulins with lispro to achieve better postprandial glycemia.

In type 2 diabetes, where insulin resistance is paramount, the minimal differences in the ratios of short- or very-short-acting to intermediate-acting insulin may not be so important. Therefore, the ease of administration afforded by pre-mixed insulins may be a true advantage. Because controlling blood glucose is essential to avoiding complications, any product that makes patients' lives easier and their regimens less complicated should be considered.

Three new insulin preparations have recently been approved for use in the United States: glargine, aspart, and a 75/25 lispro mixture. Each is unique and will have its own uses in the armamentarium for treating diabetes. Each is a human insulin analog derived from recombinant DNA technology.

Glargine is a true basal insulin with a 24-h peakless activity pattern. One can add a very-rapid-acting insulin, such as lispro or aspart, or perhaps a fast-acting oral agent, such as repaglinide, to provide a basal-bolus regimen similar to that provided by an insulin pump. Aspart is another very-rapid-acting insulin that allows mixing within the Novo Nordisk product line and will probably be seen in pre-mixed insulin preparations in the not-too-distant future. The 75/25 mixture of lispro is the first of the analog mixtures on the market. It should be quite useful in select patients who need better postprandial control and want administration via an insulin pen.

Singularly or together with other glucose-lowering agents, these new insulins will hopefully improve the lives of diabetic patients by allowing better glucose control in simpler regimens, thus helping to avoid the dreadful complications of this disease.


1Bliss M: The Discovery of Insulin. Chicago, The University of Chicago Press, 1982.

2Rosskamp R, Park G: Long-acting insulin analogs. Diabetes Care 22 (Suppl 2): B109-13, 1999.

3Goeddel DV: Expression in Escherichia coli of chemically synthesized gene for human insulin. Proc Natl Acad Sci U.S.A. 76:106­10, 1979.

4Marble A: Insulin in the treatment of diabetes. In Joslin's Diabetes Mellitus. 12th ed. Lea & Febiger, Philadelphia, 1985, p. 380-405.

5Lantus prescribing information, April 2000.

6Heinemann L, Linkeschova R, Rave K, Hompesch B, Sedlak M, Heise T: Time-action profile of the long-acting insulin analog insulin glargine (HOE 901) in comparison with those of NPH insulin and placebo. Diabetes Care 23:644-49, 2000.

7Ratner R, Hirsch I, Neifing J, Garg S, Mecca T, Wilson C: Less hypoglycemia with insulin glargine in intensive insulin therapy for type 1 diabetes. Diabetes Care 23:639-43, 2000.

8Information on file, Aventis Company.

9Novolog prescribing information, June 2000.

10White J Jr, Campbell RK, Hirsch I: Insulin analogues. Postgrad Med 101:58-60, 63-65, 70, 1997.

11Raskin P, Jovanovic L, Guthrie R, Leiter L, Boss A: Insulin aspart (Iasp) improves glycemic control compared to human insulin (HI) for patients with type 1 diabetes [Abstract]. Diabetes 48 (Suppl 1): A115, 1999.

12Rosenfalck A, Thorsby P, Kjems L, Birkeland K, Hanssen K, Madsbad S: Effects of the rapid-acting insulin analogue insulin aspart on postprandial glycemic excursions compared to human soluble insulin actrapid given immediately or 30 minutes before a meal in insulin treated type 2 diabetes patients [Abstract]. Diabetes 48 (Suppl 1):A116, 1999.

13Home P, Lindholm A, Riis A: Improved long term control with insulin aspart versus human insulin in people with type 1 diabetes. Diabetes 48 (Suppl 1): A358, 1999.

14Home P, Lindholm A, Hylleberg B, Round P: Improved glycemic control with insulin aspart: a multicenter randomized double-blind crossover trial in type 1 diabetic patients. Diabetes Care 21:1904-1909, 1998.

15Halberg I, Jacobsen L, Dahl U: A study on self-mixing insulin aspart with NPH insulin in the syringe before injecting [Abstract]. Diabetes 48 (Suppl 1):A104, 1999.

16Jacobson L, Sogaaard B, Riis A: Comparison of the pharmacokinetics and pharmacodynamics of a premixed formulation of soluble and protamine-retarded insulin aspart and a similar premix of human insulin [Abstract]. Diabetes 48 (Suppl 1):A108, 1999.

17Roach P, Yue L, Arora V: Improved postprandial glycemic control during treatment with Humalog Mix 25, a novel protamine-based insulin lispro formulation. Diabetes Care 22:1258-61, 1999.

18Vignati L, Anderson JH Jr, Iversen PW: Efficacy of insulin lispro in combination with NPH human insulin twice per day in patients with insulin-dependent or non-insulin-dependent diabetes mellitus. Clin Ther 19:1408-21, 1997.

19Humalog Mix 75/25 prescribing information, December 1999.

20Lalli C, Ciofetta M, Del Sindaco P, Torlone E, Pampanelli S, Compagnucci P, Cartechini M, Bartocci L, Brunetti P, Bolli G: Long-term intensive treatment of type 1 diabetes with the short-acting insulin analog lispro in variable combination with NPH insulin at mealtime. Diabetes Care 22:468-77, 1999.

21Roach P, Trautmann M, Anderson J, the Lispro Mixture 25 Study Group: Improved postprandial glycemia during treatment with a lispro/intermediate-acting insulin mixture, lispro mixture 25.  

22Strack T, Roach P, Anderson J, the NPL Study Group: Improved glycemic: control with self-prepared mixtures of insulin lispro and a new sustained-release lispro formulation, NPL.  

23Roach P, Trautmann M, Anderson J, the LM Study Group: Lower incidence of nocturnal hypoglycemia during treatment with a novel protamine-based formulation of insulin lispro.   

Howard M. Lando, MD, FACE, FACP, is an associate clinical professor in the Department of Endocrinology at George Washington University in Washington, D.C.

Note of Disclosure: Dr. Lando has served on an advisory panel for Aventis Pharmaceuticals and has received honoraria for speaking engagements from Takeda/Lilly. Both companies manufacture insulin products.

Copyright © 2000American Diabetes Association
Updated 10/00
For ADA Related Issues contact

For Technical Issues contact