The treatment of type 1 diabetes is simple in theory: replace the missing endogenous production of insulin with exogenous insulin in a manner that mimics normal physiology. The difficulties in doing this are myriad, however, in part stemming from the fact that endogenous insulin is delivered in intricate pulses to the portal circulation whereas exogenous insulin is provided subcutaneously through the peripheral circulation.
Additionally, the beta cells in the pancreas sense minute changes in blood sugar levels and secrete insulin, glucagon, amylin, incretins, and other hormones to balance blood sugar levels. The traditional monitoring of blood sugar levels four times per day with two to three injections of insulin does not come close even approximately to what should be happening with regards to insulin regulation. Newer insulin production technology that has led to the development of insulin analogues has made insulins easier to use, with less hypoglycemia. Insulin delivery devices, from pens to pumps, provide more options for patients. Monitoring technology, with easy to use glucose meters and continuing glucose sensing, makes it easier to follow blood sugar levels and react to trends. None of this approaches the functionality of the human beta cell, however, and it will be our ability to restore and maintain beta cell mass that will truly treat (and potentially cure) type 1 diabetes. This review will focus on the treatments that are currently available, the evolving area of continuous glucose monitoring and possible cures for type 1 diabetes.
Prior to the discovery of the therapeutic role for animal insulin in the treatment of human diabetes by Banting and Best in 1922 type 1 diabetes was a fatal disease. In parts of the world where access to insulin is limited, people with type 1 diabetes continue to suffer from poor outcomes. In most of the developed world, however, insulin is readily available.
Characteristics of Insulin Preparations
Increasing numbers of various insulin types are becoming available, ranging from the traditional insulins to insulin analogues. This diversity of choice in terms of onset and duration of action allows use of exogenous insulin to mimic normal physiology more closely, thereby allowing for improvements in glycemic control with less hypoglycemia. However, insulin is not a simple drug to prescribe, since inappropriate doses can result in severe hypoglycemia. In people with type 1 diabetes integrating the carbohydrate content of the meal and other factors such as exercise and illness is necessary to determine the required insulin dose. Patients need access to a health-care team, with education on diabetes self-management and nutrition.
Knowledge of the pharmacology of each of the various insulin preparations is required, coupled with observation of individual patient reactions. Historically four properties characterized insulin preparations used for injection: concentration, species source, purity, and type. Issues regarding species source and purity have become moot, since most insulin preparations are now based on highly purified human insulin. As for concentration, insulin is generally marketed in 10-ml vials at a concentration of 100 units/ml (units-100). Thus, an injection of 0.5 ml delivers 50 units of insulin. Fortunately, calculations by the patient are obviated by the use of syringes with the number of units marked directly on the barrel. In Europe, other concentrations of insulin can be obtained (such as units-40). Additionally, a more concentrated form of insulin known as units-500 can be purchased by special order in the United States for use in patients who require large amounts of insulin due to severe insulin resistance. Care must be taken when using insulins that are other than units-100 since errors in dosing can occur if insulin syringes meant for the units-100 concentration of insulin are used.
Types of Insulin
From a therapeutic point of view, three characteristics of the time course of action of the different types of insulin preparations are important: onset of action, time of peak activity, and duration of action. These depend on the rate of absorption after the subcutaneous injection. Table 43.1 summarizes the data on the insulin preparations currently on the market. These are general guidelines and may not pertain exactly to the clinical situation in which patients’ physical activity and eating patterns differ from conditions imposed by a research study setting. The ranges are also only approximations because of the great intrinsic variability among patients and because the response of an occasional patient may differ considerably from the values listed.
Table 43.1 Time course of action of insulin preparations (times are approximations and may vary in different individuals)
Type Onset (min) Peak (h) Duration (h)
Rapid acting 5–15 0.5–2.0 3.0–5.0
Short acting 0.50–1.0 2.0–4.0 6.0–8.0
Intermediate acting 2.0–4.0 h 4.0–10.0 10.0–16.0
Long acting 2–4 None
• Glargine None ∼24 6–23*
*Duration depends on dose given (modified from AACE Diabetes Mellitus Guidelines, Endocr Pract. 2007;13(Suppl 1):17)
Variability of Insulin
There are many reasons that insulin has a variable action in a given individual. Insulin analogues tend to be less variable (that is, they have the least intra-subject variability when injected in the same individual and their activity is measured on different days) than the older insulins, particularly those of the lente series (lente and ultralente) which are the most variable. The volume of a dose of insulin may alter its absorption, although this may be less true with the newer analogues. The site of injection can influence rate of absorption of the insulin as well as the depth of injection (intramuscular versus subcutaneous versus intradermal). Once again, the analogue insulins tend to be less impacted by the site of injection than older insulin preparations. Finally, regional blood flow can alter the absorption of insulin with factors such as exercise, skin temperature, and hydration status impacting absorption. Patients with type 1 diabetes are often able to recognize variability in insulin activity and use this knowledge to inject insulin at different sites for different purposes (e.g., inject in a site that yields faster activity when the blood glucose level is high).
The initial insulins were purified from actual animal pancreases. They were named for the species they came from pork, beef, and beef/pork. In 1986 the first recombinant human insulin was released onto the market. Because it is human insulin, produced in Escherichia coli, it is less immunogenic than the older animal insulin and has largely replaced use of the older animal insulins.
Comparing Regular Versus Analogue Insulins
The first type of insulin to be produced was regular insulin. It has no modifying agent and currently is the only one that should be administered intravenously. Figure 43.1 shows the structure of the insulin molecule, with its α and β chains, and region for self-aggregation and modification. When regular insulin is injected there is a delay in its absorption due to self-aggregation that occurs between insulin molecules. Regular insulin forms a hexamer in subcutaneous tissue and must dissociate into a monomeric form to be absorbed.
To overcome this problem insulin analogues have been created based on the knowledge that the related hormone insulin-like growth factor I (IGF-1) acts similarly to insulin but does not self-aggregate due to differences in the C-terminal portion of the β chain. An analogue is regular human insulin that has been altered through a modification in its structure (usually in its amino acid composition, but other modifications are possible) that changes its tendency for self-aggregation and thus its absorption, but not its binding to the insulin receptor. Table 43.2 compares all available insulin analogues and describes the modifications of each. In summary, rapid-acting insulin analogues tend to reduce postprandial hyperglycemia and reduce rates of hypoglycemia compared to regular insulin and long-acting analogues reduce nocturnal hypoglycemia and weight gain. These findings are most evident in individuals with type 1 diabetes when all-analogue regimens are compared with all-human insulin regimens (early studies tended to compare a hybrid of analogue rapid-acting insulin with NPH as the bolus insulin). Finally, analogue insulins reduce intra- and inter-subject variability in blood sugar response, which is related to the reduction in rates of hypoglycemia.
Rapid-Acting Insulin Analogues
Approved by the US FDA in 1996, insulin lispro (Humalog) was the first insulin analogue to enter the market. It is produced through recombinant DNA methods using E. coli. Lispro differs from regular insulin by inversion of the amino acids lysine and proline in the C-terminus of the β chain. This inversion reduces the formation of dimers and hexamers (which typically occurs with regular insulin) and thereby significantly facilitates the rate of absorption of lispro. This increases both the onset of action and the time to peak concentration and decreases the time of return to baseline, more closely mimicking normal physiology. Insulin aspart (Novolog) was the second rapid-acting insulin to be introduced. Aspart differs from regular insulin by replacement of proline at position 28 of the β chain with the negatively charged aspartic acid. It is produced by recombinant DNA technology using a modified strain of the yeast Saccharomyces cerevisiae (baker’s yeast) as the production organism. The newest rapid-acting analogue is insulin glulisine (Apidra), which is produced from nonpathogenic E. coli. Insulin glulisine differs from human insulin in that the amino acid asparagine at position B3 is replaced by lysine and the lysine in position B29 is replaced by glutamic acid. All three rapid-acting analogues are approved for treatment of type 1 and type 2 diabetes mellitus. Lispro and aspart have a Category B designation in pregnancy (presumed safety based on animal studies); glulisine is Category C (uncertain safety).
Fig. 43.1 Structure of the insulin molecule and alterations for various insulin analogues
Table 43.2 Differences in amino acid sequence of different insulin analogues
Type Chain Alteration
Lispro β chain Proline (b28)/lysine (b29) switched to (b28)/proline (b29)lysine
Aspart β chain Proline (b28) replaced by aspartic acid(b28)
Glulisine β chain Asparagine (b3) replaced by lysine (b3) lysine (b29) replaced by glutamic acid (b29)
Glargine β chain Two arginines added to end (b30) of β chain
α chain Asparagine (a21) replaced by glycine (a21)
Detemir β chain C-14 fatty acid side chain (myristic acid attached to lysine (b29).
The amino acid modifications in lispro, aspart, and glulisine result in subcutaneous absorption rates that are twice as fast and peak levels that are higher than those of regular insulin. More importantly, peak insulin action occurs approximately twice as fast with the rapid-acting analogues as compared to regular insulin and the levels return to baseline more rapidly than with regular insulin. These pharmacokinetic and pharmacodynamic properties more closely resemble physiologic meal-induced insulin secretion and provide greater flexibility and convenience to the patient since the analogues may be injected immediately before a meal or even after eating (as opposed to 30 min prior to meals for regular insulin). Reviews of published clinical studies that compare the rapid-acting insulin analogues to regular insulin reveal the following generalizations: (1) All three rapid-acting insulin analogues are superior to regular insulin in controlling postprandial hyperglycemia; (2) inter- and intra-patient variability tends to be reduced with the insulin analogues; (3) rapid-acting insulin analogues usually result in less hypoglycemia (a finding that is more pronounced in studies comparing all-analogue insulin to human insulin versus studies in which rapid-acting insulin is added to NPH as a basal insulin where rates of hypoglycemia may not be different9); and (4) rapid-acting insulin analogues are usually comparable to regular insulin at lowering HbA1c levels, although occasionally there is a greater improvement with the analogues.
Two long-acting insulin analogues are currently available. The first insulin glargine (Lantus) was approved by the US FDA in April 2000 and is produced by replacement of asparagine at position 21 of the α chain of regular insulin with glycine and addition of two arginine molecules to the C-terminus of the β chain. These modifications shift the isoelectric point leading to formation of microprecipitates in the subcutaneous tissue from which small amounts of insulin glargine are slowly released, resulting in a relatively constant concentration over 24 h.28,29 In pharmacodynamic studies, insulin glargine was found to have a mean duration of action of 22 ± 4 h, without a pronounced peak. This profile allows glargine to be dosed once daily as basal insulin. In contrast, NPH insulin reaches a peak between 4 and 8 h, with a duration of action between 12 and 16 h. The fluctuations in diurnal serum insulin levels are significantly less in patients treated with glargine, compared to NPH or ultralente. The vast majority of clinical studies involving glargine have compared its efficacy and tolerability to NPH insulin. From these studies one common theme emerges: once-daily glargine appears to be similar if not more efficacious than NPH in glycemic control and is associated with a significantly lower rate of hypoglycemia (particularly at night) as well as less glucose fluctuation. Approved in June 2005, insulin detemir (Levemir) constitutes the other long-acting insulin analogue.
Whereas all of the other insulin analogues are produced by either amino acid addition, inversion, or substitution of regular insulin, insulin detemir is produced by deletion of the final C-terminal amino acid molecule of regular insulin and addition of a 14-carbon fatty acid chain to lysine in position 29 of the β chain. These modifications allow insulin detemir to self-associate into hexamers and to also bind to albumin, both of which slowdown its systemic absorption. Pharmacodynamic studies indicate that insulin detemir has a relatively flat action profile with a duration of action that appears dose dependent (mean duration of action ranging from 6 to 23 h). At higher doses (>0.4 units/kg), the duration of action is approximately 20 h. At lower dose (<0.4 units/kg), the duration of action is shorter and twice-daily dosing may be necessary.35 Multiple studies suggest that intra-patient variability in insulin action is less with detemir compared with NPH and glargine. The rates of hypoglycemia and weight gain are lower when detemir is compared to NPH.Both glargine and detemir are approved for the treatment of adult and pediatric patients with type 1 diabetes mellitus or adult patients with type 2 diabetes mellitus who require basal (long-acting) insulin. Because of their chemical properties, glargine and detemir cannot be mixed in the same syringe with other insulin preparations. Both have a category C designation in pregnancy. Detemir, because of its somewhat shorter duration of action, is often used as a twice a day drug, compared to glargine which is often used once a day in patients with type 1 diabetes. However, individual patients may differ, with some needing only once-daily detemir and others needing twice a day glargine.
Premixed insulin preparations should not be routinely used for the treatment of most patients with type 1 diabetes, due to the lack of flexibility, fixed ratios of rapid acting to longer acting insulin, and lack of data on achieving and maintaining tight control. However, in patients where reaching and maintaining near euglycemia is not possible and/or where this is the only insulin available, premixed insulin can be used to avoid the acute complications of diabetes and maintain control that is as close to target as possible.
Clinical Use of Insulin Analogues
In the treatment of type 1 diabetes, use of insulin analogues has become increasingly common. This is in part due to the clearly recognized need for achieving near euglycemia in people with type 1 diabetes and the marked increase in risk for severe hypoglycemia seen in studies using non-analogue insulin. In patients followed at one DCCT center HbA1c levels fell with an increase in the rate of hypoglycemia on non-analogue insulin. Once lispro was introduced, HbA1c levels continued to fall without a further increase in hypoglycemia. Reducing the variability of insulin activity and lessening weight gain make insulin analogues easier use in type 1 diabetes. An important concept in using premeal insulin, and adequately lowering postprandial blood glucose levels, is the concept of “lag time.” This is the ideal period of time during which a short or rapid-acting insulin should be injected before a meal in order to optimally control the postprandial blood glucose level. The higher the blood
sugar, the longer the lag time between insulin injection and eating. Although it is often easiest to inject insulin immediately prior to eating, use of continuous glucose monitoring makes the lag in onset of rapid-acting insulin more apparent in some individuals.
Intensive Insulin Therapy
The goal of insulin therapy is to provide insulin replacement in as physiologic a fashion as possible. Figure 43.2 shows the time course of action for the available insulin preparation – rapid-, short-, intermediate-, and long-acting. Ideally, for patients with type 1 diabetes, the most physiologic regimen is the use of a basal insulin combined with premeal boluses (Figs. 43.3 and 43.4). This can be accomplished either with a long-acting insulin such as glargine or with a detemir and premeal rapid-acting insulin (called multiple daily insulin injection or MDI therapy), or by using continuous subcutaneous insulin infusion (CSII) therapy. These approaches offer the most flexibility in lifestyle. However, these types of regimens require that patients either give 4–6 injections of insulin per day or master the use of an insulin pump plus learn how to do carbohydrate counting and test blood sugar levels before each insulin injection. Generally this requires self-monitoring of blood glucose levels more than three times per day since an increased frequency of self-monitoring of blood glucose (SMBG) in people with type 1 diabetes on intensified regimens is associated with an improved HbA1c. On non-analogue insulins patients on intensive regimens gain more weight and experience more hypoglycemia, as shown in the Diabetes Control and Complications Trial (DCCT).
Fig. 43.2 Duration of action of various injected insulin preparations
Fig. 43.3 Idealized insulin secretion and insulin replacement
Fig. 43.4 Basal bolus insulin therapy
Less intensive regimens, such as twice a day NPH and regular insulin, may require less testing by the patient, but because the patient is taking an intermediate-acting insulin (NPH) they have much less flexibility in lifestyle. The insulin will peak 6–12 h after injection and the patient will need to eat at that time. To avoid hypoglycemia patients will often keep their blood sugar levels above target. No matter what the regimen, the goal for every patient is to keep their HbA1c level as close to normal as possible with a minimum number of hypoglycemic reactions, particularly avoiding severe reactions (that is, insulin reactions that require assistance of another person for treatment).
CSII Versus MDI
Continuous subcutaneous insulin infusion therapy was first used in the late 1970s. Its use has gradually increased since. Initial models were large and bulky, while current insulin pumps are available from a wide range of manufacturers with many features, including technology that helps calculate insulin doses. Early benefits were a reduction in episodes of hypoglycemia and a lowering of the fasting blood sugar level (by increasing insulin delivery to overcome the dawn phenomenon). Risks include diabetic ketoacidosis and infusion site infections. Benefits of CSII have been hard to quantitate, in part because studies are small and technology advances quickly. In earlier studies, comparing CSII to NPH-based MDI regimens, CSII was associated with improvements in outcomes, such as a reduction in rates of severe hypoglycemia. The five-Nations Study Group, assessing 272 patients using CSII versus MDI, revealed less hypoglycemia and lower HbA1c levels in patients in CSII therapy. However, MDI therapy was NPH based, rather than based on a long-acting insulin analogue.
More recent studies comparing CSII to MDI regimens using rapid and long-acting insulin analogues show fewer differences. In a pilot study involving 21 people with type 1 diabetes who had hypoglycemia unawareness and recurrent episodes of severe hypoglycemia, CSII, MDI, and intensive education alone were compared over 24 weeks.
In both the analogue and the CSII groups there was a restoration of hypoglycemia awareness and a reduction in severe hypoglycemia. In both groups there was a reduction in HbA1c, although it did not reach significance in the CSII group. A short-term (5 weeks in each arm) randomized cross-over study in 100 people with type 1 diabetes showed a reduction in glycemic exposure (as measured by fructosamine levels and continuous glucose monitoring) in patients on CSII compared to MDI with aspart and glargine insulin. The biggest improvement in glucose control was in the overnight/early morning blood sugar levels. A retrospective study from Sweden, comparing patients who chose CSII (n = 563) versus MDI therapy with glargine (n = 513), showed a greater reduction in HbA1c with CSII therapy (−0.59% versus −0.2%, P < 0.001), especially in those with a higher initial HbA1c level.53 A retrospective study of the use of CSII in Spain revealed that patients followed on CSII for 2 years had greater improvements in HbA1c and greater reduction in the number of hypoglycemic episodes (mild and severe) compared to baseline therapy with MDI. Rates of diabetic ketoacidosis were not increased. None of these studies are well randomized or include use of newer insulin pump features, such as bolus delivery calculators. There is no well-studied approach to determining insulin to carbohydrate ratios (I:C) or correction ratios, and differing recommendations exist as to how to calculate these factors. In many cases an initial dose is calculated based on what the patient appears to be doing on his/her initial insulin regimen or estimated based on the patient’s weight. The dose is then adjusted based on pre- and 2 h postprandial blood sugar levels.
In the American Diabetes Association guide for treating type 1 diabetes, the algorithm for determining the I:C is to multiply the body weight (lbs) by 2.8 and dividing this product by the total daily dose of insulin. Another recommendation suggests calculating the amount of carbohydrate for one unit of rapid-acting insulin by dividing 500 by the total daily insulin dose. Thus using the former method, in a 145-pound patient using 38 units of insulin per day, the I:C ratio is 1:11 and with the latter method it is 1:13. Use of pump bolus calculators can be effective but requires accurately setting the target blood sugar level, the I:C, the insulin sensitivity (IS), and insulin on board factor, as self as accurate carbohydrate counting and a premeal blood sugar measurement. In a study using a hyperinsulinemic euglycemic clamp technique, individual carbohydrate ratios were determined in patients with type 1 diabetes. Patients undergoing a glucose clamp were fed a meal and their exact insulin to carbohydrate ratio was determined. The average carbohydrate ratio was 1:9.3.58 Additionally the investigators found that the peak postprandial blood sugar level occurred approximately 70 min (68.5 ± 8.8 min) after a meal. This finding is quite similar to a recent study using continuous glucose monitoring in which the time to peak postprandial blood glucose was 72 ± 23 min.
Much of the literature involving CSII is in children, where management is somewhat different than in adults. In a study comparing MDI with glargine and aspart versus CSII with aspart in 32 adolescents (average age of 13 years) with type 1 diabetes over 16 weeks the HbA1c was unchanged in the glargine group (8.2% versus 8.1%) and improved in the CSII group 61 (8.1–7.2%, P < 0.05 versus glargine group). Another pilot study showed that early use of CSII (within 1 month of diagnosis) in children (n = 28, mean age 12.1 ± 6 years) with new-onset type 1 diabetes led to a reduction in HbA1c from 10.5 ± 2.4% to between 6.5 and 7.4% over the next 18 months, with stable C-peptide levels over the first year of treatment. CSII offers the most flexibility with regards to reducing and adjusting the basal insulin levels and can calculate and deliver doses of insulin with greater accuracy than giving insulin by injection (the smallest increment on a syringe is 0.5 units, whereas pumps can deliver much smaller doses). In addition, pumps can make an estimation of how much insulin is still active, reducing correction doses to avoid overcorrection. Whether or not to use CSII versus MDI is often a personal decision, and patients may change from one therapy to another and then back again. Quality of life is similar with either approach. The greatest risk of CSII is an increased risk for diabetic ketoacidosis (DKA) because of the lack of available long-acting depot insulin. Therefore, some patients benefit from a hybrid of a once-daily dose of long-acting insulin to prevent DKA and CSII dosing for premeal and correction doses. It is important to have a health-care team knowledgeable in the use of CSII and the necessary troubleshooting.
Related Ads By Google :