Marla Bernbaum

Narrative Review: A Rational Approach to Starting Insulin Therapy

Tailoring insulin therapy to the blood glucose profiles of patients with diabetes is indisputably a rational strategy. Yet, with the advent of various insulin preparations, the pharmaceutical industry aggressively promotes products to fit the needs of most, if not all, patients with diabetes. Of note, the Diabetes Control and Complications Trial investigators used several different insulin regimens to achieve blood glucose control in patients with type 1 diabetes (1), a population generally more homogeneous in terms of underlying pathophysiologic characteristics than those with type 2 diabetes. Thus, the need for an individualized approach to insulin therapy in patients with diabetes cannot be overemphasized. Nevertheless, certain subgroups of patients with diabetes can be differentiated from each other according to the pattern of blood glucose changes during the day. Therefore, profiling these groups may help practitioners rationally choose a particular insulin regimen. The insulin regimen should subsequently be modified on the basis of the individuals response to therapy. We describe 3 common blood glucose profiles that represent typical patterns among patients with diabetes. On the basis of the blood glucose profile, we suggest an initial insulin regimen that would then be modified according to individual responsiveness. To prime the discussion, we briefly review the pharmacologic characteristics of currently available insulin products. This review is based on previously published manuscripts that were identified through a MEDLINE search (1996 to 25 February 2006) of English-language literature. The literature search was limited to core clinical journals that have accessible full texts, and the key phrase used was therapeutic use of insulin. A total of 420 manuscripts were reviewed, but only a select number are quoted to avoid repetitiveness and to adhere to space limitations. Individual references were chosen at the discretion of the authors. All authors independently reviewed the relevant available literature. This literature, along with our clinical experiences, was used to construct practical suggestions. Pharmacologic Characteristics of Currently Available Insulins There are several excellent reviews of the available insulin preparations (29). The pharmacokinetic variables of various insulins are summarized in Table 1. Regular human insulin peaks at 2 to 3 hours when given subcutaneously and has a variable duration of action that can range from 6 to 8 hours. Although regular insulin historically has been used to target postprandial hyperglycemia, its time-action profile is far from the physiologic pancreatic postprandial insulin burst. The advent of recombinant DNA technology has made it possible to improve the time-action profile of regular insulin. The structural modifications of these novel analogues are shown in Figure 1. Currently, there are 3 rapid-acting insulin analogues with similar pharmacokinetic profiles that are more suitable for targeting postprandial hyperglycemia: insulin lispro (Humalog, Eli Lilly, Indianapolis, Indiana), insulin aspart (Novolog, Novo Nordisk, Bagsvaerd, Denmark), and insulin glulisine (Apidra, Aventis Pharmaceuticals, Inc., Bridgewater, New Jersey). These insulin analogues are rapidly absorbed (< 30 minutes) after subcutaneous injection, peak at 1 hour, and have a shorter duration of action (3 to 4 hours) than regular insulin (1012). Furthermore, the intraindividual variability in time to maximum serum insulin concentration is clinically significantly less for rapid-acting insulin analogues than for regular human insulin preparations (1013). Table 1. The Pharmacokinetic Variables of Some of the Most Commonly Used Insulin Preparations* Figure 1. The structural modifications of insulin found in insulin analogues. = Insulin lispro differs from human insulin by the substitution of proline with lysine at position 28 and the substitution of lysine with proline at position 29 of the insulin chain. = Insulin aspart is designed with the single replacement of the amino acid proline by aspartic acid at position 28 of the human insulin chain. * = Insulin glulisine is designed with the substitution of the amino acid lysine with asparagine at position 3 of the human insulin chain and by substitution of the amino acid lysine at position 29 with glutamine. = Insulin glargine differs from human insulin in that the amino acid asparagine at position A21 is replaced by glycine and 2 arginines are added to the C-terminus of the chain. = Insulin detemir is designed to bind albumin in plasma after absorption. Threonine is omitted from position 30 of the insulin chain and replaced by myristic acid, a C14 fatty acid chain. Figure reprinted with permission from reference 2: Oiknine R, Bernbaum M, Mooradian AD. A critical appraisal of the role of insulin analogues in the management of diabetes mellitus. Drugs. 2005;65:325-40. [PMID: 15669878] The rapid-acting insulin analogues should be injected 5 to 15 minutes before a meal. However, in infants or in older adults with dementia who both have unpredictable eating patterns, rapid-acting analogues can be administered after the meal without excessive deterioration of glycemic control (14, 15). Intermediate-acting insulins, such as neutral protamine Hagedorn (NPH) or lente insulin, have a delayed onset of action ranging between 2 to 4 hours, can take approximately 6 to 7 hours to reach peak concentration, and can last up to 20 hours. This leads to a distinct peak-and-trough effect, and therefore, when used as basal insulin, 2 or more injections are often required each day to minimize the daily excursions of insulin levels. Recently, production of lente and ultralente insulins was discontinued because of their decreasing market shares. Long-acting insulin glargine has an onset of action of approximately 2 hours, reaches a plateau of biological action at 4 to 6 hours, and lasts up to 24 hours (9, 16). The timing of insulin glargine administration should be individualized, although it is minimally important as long as the insulin is administered at the same time every day (17). Compared with bedtime NPH insulin, insulin glargine is associated with less nocturnal hypoglycemia in patients with type 2 diabetes (28.8% vs. 12.6%, respectively; P= 0.011) (18, 19). If nocturnal hypoglycemia occurs after evening or bedtime administration of insulin glargine, the timing of the injection should be changed to the morning. Insulin detemir is another long-acting basal insulin analogue that is soluble at neutral pH. It has a unique mechanism of action (2023). After subcutaneous injection, insulin detemir binds to albumin through its fatty acid chain. At steady state, the variability in the concentration of free, unbound insulin is then greatly reduced, resulting in stable plasma glucose levels (2023). The duration of action of insulin detemir is approximately 20 hours at a dose of 0.4 U/kg of body weight (22). Insulin detemir is therefore administered twice daily in most patients. As with insulin glargine, insulin detemir has a more stable, less variable pharmacokinetic profile than does NPH. There are 2 premixed conventional insulins: Humulin 50/50 (Eli Lilly, Indianapolis, Indiana), which consists of 50% NPH and 50% regular insulin, and Humulin 70/30 (Eli Lilly) or Novolin (70/30) (Novo Nordisk, Bagsvaerd, Denmark), which consists of 70% NPH and 30% regular insulin. In some countries, additional premixed insulins with different proportions of NPH and regular insulin are available. Rapid-acting insulin analogues are also available in premixed preparations with rapid and intermediate insulin activity (Table 1). Self-mixing of insulin combinations in 1 syringe may be difficult for some patients and may result in errors of dosing. Fixed-mixed combinations may simplify the insulin regimen and reduce the number of daily injections. Three types of fixed-ratio insulin analogue mixes are currently available: Humalog mix 75/25 (Eli Lilly), a 75% insulin lispro protamine suspension with 25% insulin lispro; Humalog mix 50/50 (Eli Lilly), a 50% insulin lispro protamine suspension with 50% insulin lispro; and Novolog mix 70/30 (Novo Nordisk), a 70% insulin aspart protamine suspension with 30% insulin aspart. Conventional premixed human insulins have an onset of action of approximately 0.5 to 2 hours, usually plateau at 3 to 6 hours, and last up to 24 hours. Insulin analogue mixes have an onset of action of approximately 15 minutes, reach a peak biological action at 1 to 4 hours, and last up to 24 hours (24). Of note, there is clinically significant interindividual variability in the pharmacokinetics of insulin, and the profile often depends on the dose that is administered. However, the pharmacokinetics of insulin analogues have less intraindividual variability than human insulin. The availability of insulin analogues enhances flexibility and convenience for insulin-treated patients, and this increased flexibility in insulin timing is associated with an improvement in quality-of-life measures (25). The increasing variety of insulin preparations allows more opportunities to achieve control. The downside is the increased challenge facing health care providers in making the right choices in insulin therapy. Choosing the Insulin Preparation Protocols for insulin use depend on patient location, comorbid conditions, and available resources. All patients with type 1 diabetes require insulin therapy, whereas the indications for insulin therapy in patients with type 2 diabetes include symptomatic hyperglycemia, failure of oral therapy, pregnancy, acute illness necessitating surgery, acute myocardial infarction or cardiovascular surgery, and admission to an intensive care unit. The proper timing of initiation of insulin therapy, when oral agents are not achieving glycemic goals, depends on such variables as severity of hyperglycemia, economic and psychosocial circumstances of the patie

A Critical Appraisal of the Role of Insulin Analogues in the Management of Diabetes Mellitus

Insulin is one of the oldest and best studied treatments for diabetes mellitus. Despite many improvements in the management of diabetes, the nonphysiological time-action profiles of conventional insulins remain a significant obstacle. However, the advent of recombinant DNA technology made it possible to overcome these limitations in the time-action profiles of conventional insulins. Used as prandial (e.g. insulin lispro or insulin aspart) and basal (e.g. insulin glargine) insulin, the analogues simulate physiological insulin profiles more closely than the older conventional insulins. If rapid-acting insulin analogues are used in the hospital, healthcare providers will need a new mind-set. Any error in coordination between timing of rapid-acting insulin administration and meal ingestion may result in hypoglycaemia. However, guidelines regarding in-hospital use of insulin analogues are few. The safety profile of insulin analogues is still not completely established in long-term clinical studies. Several studies have shown conflicting results with respect to the tumourigenic potential of this new class of agents. The clinical implications of these findings are not clear. Although novel insulin analogues are promising ‘designer drugs’ in our armamentarium to overcome some of the limitations of conventional insulin therapy, cost may be a limiting factor for some patients.

Promoting diabetes self-management and independence in the visually impaired: a model clinical program.

Program Structure and Content A program has been developed at the St Louis University School of Medicine aimed at encouraging independence and improved self-esteem for the visually impaired individual with diabetes. This clinical program employs an integrated multidisciplinary approach to include ( 1 ) diabetes education focusing on the teaching of skills and devices needed by the visually impaired for diabetes selfmanagement ; (2) a monitored exercise program, individually tailored to accommodate visual, cardiovascular, and neurologic limitations; (3) group support with individual counseling as indicated; and (4) a prospective evaluation of measures of glucose control, exercise tolerance, and psychological adjustment. The program was initiated by several St Louis University diabetologists, and was designed through collaboration with consultants in the areas

Psychosocial Profiles in Patients With Visual Impairment Due to Diabetic Retinopathy

The psychological impact of vision loss due to diabetic retinopathy is compounded by the loss of diabetes self-management skills. The appropriate role and timing for rehabilitative intervention has not been determined. Twenty-nine individuals with diabetes mellitus, 16 with stable visual impairment and 13 with fluctuating and transitional visual impairment, underwent psychological assessment before and after entering into a specially designed rehabilitation program. Low levels of performance were demonstrated by the Rosenberg Self- Esteem Scale and the Diabetes Self-Reliance Test in both groups. The Minnesota Multiphasic Personality Inventory, the Zung Self-Rating Depression Scale, and the Rand Mental Health Index suggested that subjects with stable vision impairment were moderately compensated relative to the transitional group, although the former group may have been totally blind. Both groups demonstrated significant improvements in psychological profiles after the program. It is suggested that a rehabilitation program may be of clinical benefit early in the course of vision loss associated with diabetic retinopathy.

Personal and family stress in individuals with diabetes and vision loss

The intrapersonal distress and the impact of diabetes and vision impairment on marital functioning were assessed. Significant degrees of intrapersonal distress were demonstrated by the Beck Depression Inventory, Rosenberg Self-Esteem Scale, and Rand Mental Health Index. Family functioning as assessed by the Family Assessment Device was significantly compromised. Vision impairment was a major stressor in the spousal relationship. Of 18 subjects who had been involved in a committed relationship at the onset of vision impairment, 9 had separated. Separation occurred at a mean of 1.6 years after the vision impairment. Totally blind individuals were at greater risk for separation than those who were legally, but not totally, blind. Psychological intervention was a limited benefit. Studies are necessary to identify the appropriate timing for further interventions.

Cerebral and Ophthalmic Artery Hemodynamic Responses in Diabetes Mellitus

OBJECTIVE To address whether hemodynamic responses in the cerebral arteries and OAs may be altered in patients with diabetic retinopathy. We used TCD to evaluate the effects of changes in BP, posture, and exercise on MCA and OA blood flow velocities. RESEARCH DESIGN AND METHODS We evaluated 13 patients with BDR, 19 with PDR, and 11 control subjects. Each was tested while supine, breathing 100% oxygen, sitting, and during exercise. RESULTS Control subjects exhibited linear increases in velocity in the MCA and OA with increases in BP. Those with BDR had higher baseline sBP than control subjects. The MCA velocity response to BP in the BDR group was parallel to, but differed significantly from, the response in control subjects in compensation for the level of BP (ANCOVA F1,53 = 10.1, P = 0.003). The OA velocity response to BP was indistinguishable between the control subjects and the group with BDR. The group with PDR had more elevated BP than control subjects, and those with BDR had more advanced autonomic neuropathy. The PDR group had heterogenous velocity responses in the MCA and OA with respect to BP. None of the control subjects and 3 of 13 of the BDR group had abnormal autoregulatory velocity responses in the MCA or OA to 100% oxygen breathing, whereas 12 of 19 PDR patients were abnormal (P < 0.01). Of those with PDR, 4 had elevated MCA and 3 had elevated OA velocities while supine at rest. CONCLUSIONS Patients with PDR demonstrated abnormal hemodynamic responses of the cerebral and ophthalmic circulation both at rest and with exercise.

A Model Clinical Program for Patients With Diabetes and Vision Impairment

A program was developed to improve independence, self- esteem, and glycemic control in patients with diabetes and blindness. Twenty-nine individuals with both insulin- dependent and noninsulin dependent diabetes mellitus entered 12-week programs that included education focusing on diabetes self-management skills for the visually impaired, monitored exercise sessions, and group support. Glycated. hemoglobin valuesfellfrom 13.0% ± 0.6% (SEM) to 11.4% ± 0.5% (P=.001). Exercise tolerance in a 12-minute walk test improved from 0. 48 ± 0. 04 to 0. 64 ± 0.05 miles (P=.001). Marked improvements occurred in psychosocial indices, demonstrated through changes in the Rand Mental Health Index from 155 ± 6 to 174 ± 5 (P=. 0001), the Rosenberg Self- Esteem Scale from 22 ± 1 to 19 ± 1 (P =. 001), the Zung Depression Scale from 0.50 ± 0. 02 to 0. 44 ± 0. 02 (P =.001), and the Diabetes Self- Reliance Test from 60 ± 3 to 74 ± 2 (P=.0001).

The Effect of Glipizide Gastrointestinal Therapeutic System on Islet Cell Hormonal Responses to a Test Meal in NIDDM

OBJECTIVE To determine whether the abnormal glucagon and amylin secretions in NIDDM are secondary to hyperglycemia and relative hypoinsulinemia. RESEARCH DESIGN AND METHODS A total of 13 patients with NIDDM were studied before and after treatment with glipizide gastrointestinal therapeutic system (GITS) in a randomized double-blind placebo-controlled fashion. Of the 13 subjects, 9 were randomized to the glipizide GITS arm and 4 were randomized to the placebo arm of the study. Serum glucose, insulin, C-peptide, plasma glucagon, and plasma amylin concentrations were measured under fasting and postprandial (post-Sustacal ingestion) conditions. The Sustacal challenge was performed at baseline and after 12 weeks of treatment with either glipizide GITS or placebo. RESULTS Glipizide GITS treatment resulted in a significant reduction in hyperglycemia and increases in insulin and C-peptide secretion. Hyperglucagonemia was not ameliorated, and amylin secretion was not altered after glipizide GITS treatment. Placebo-treated patients did not show significant changes in any of the parameters measured. CONCLUSIONS Glipizide GITS treatment failed to ameliorate the hyperglucagonemia of NIDDM and did not alter amylin secretion even though it increased insulin secretion and significantly ameliorated the hyperglycemia. These observations suggest that NIDDM related abnormalities in some of the islet cell hormonal responses are the result of changes inherent in the islet cells and may be independent of hyperglycemia and relative hypoinsulinemia.

Exercise for Patients With Diabetic Retinopathy

E xercise is considered an integral component of therapy for diabetes. In addition to enhancing insulin sensitivity and glucose utilization, it promotes improved respiratory and cardiovascular function, flexibility, and general sense of well-being. However, the safety and benefits of exercise for patients with diabetic retinopathy have not been established. Diabetologists and ophthalmologists usually recommend that patients with proliferative diabetic retinopathy (PDR) refrain from exercise to avoid vitreoretinal hemorrhage. Often burdened with multiple advancing diabetic complications and fearful of further vision loss, patients with PDR withdraw from normal daily activities. This leads to further compromise of physical and psychosocial ability and cardiovascular deconditioning. It is important to develop exercise prescriptions for individuals who have PDR with other concurrent diabetic complications that will result in improved participation in normal daily activities and psychosocial well-being, while minimizing risk of further visual deterioration.

Effectiveness of Glucose Monitoring Systems Modified for the Visually Impaired

OBJECTIVE To compare three glucose meters modified for use by individuals with diabetes and visual impairment regarding accuracy, precision, and clinical reliability. RESEARCH DESIGN AND METHODS Ten subjects with diabetes and visual impairment performed self-monitoring of blood glucose using each of the three commercially available blood glucose meters modified for visually impaired users (the AccuChek Freedom [Boehringer Mannheim, Indianapolis, IN], the Diascan SVM [Home Diagnostics, Eatontown, NJ], and the One Touch [Lifescan, Milpitas, CA]). The meters were independently evaluated by a laboratory technologist for precision and accuracy determinations. RESULTS Only two meters were acceptable with regard to laboratory precision (coefficient of variation <10%)—the Accuchek and the One Touch. The Accuchek and the One Touch did not differ significantly with regard to laboratory estimates of accuracy. A great discrepancy of the clinical reliability results was observed between these two meters. The Accuchek maintained a high degree of reliability (y = 0.99X + 0.44, r = 0.97, P = 0.001). The visually impaired subjects were unable to perform reliable testing using the One Touch system because of a lack of appropriate tactile landmarks and auditory signals. CONCLUSIONS In addition to laboratory assessments of glucose meters, monitoring systems designed for the visually impaired must include adequate tactile and audible feedback features to allow for the acquisition and placement of appropriate blood samples.