CLINICAL DIABETES
VOL. 15 NO. 2 May/June 1997

These pages are best viewed with Netscape version 3.0 or higher or Internet Explorer version 3.0 or higher. When viewed with other browsers, some characters or attributes may not be rendered correctly.

F E A T U R E   A R T I C L E

Combination Oral Agent/Insulin Therapy in Patients With Type II Diabetes Mellitus

John R. White, Jr., PharmD

Clinicians managing patients with type II diabetes mellitus today are faced with the sometimes bewildering question: What is the most appropriate regimen for this particular patient? The options for monotherapeutic oral agents now include a choice of seven sulfonylureas, a biguanide, a alpha-glucosidase inhibitor, and a thiazolidinedione. A second alpha-glucosidase inhibitor is anticipated to be available soon. The complexity of the regimen decision is enhanced when one considers dual-agent and triple-agent oral therapy options. Complicating matters further are the possible permutations of oral agent/insulin therapy and insulin monotherapy. Which insulin should be used in a given situation? When should it be administered? What role does the insulin analog lispro play in combination therapy?

This article focuses on the use of combination oral agent/insulin therapy in patients with type II diabetes and includes discussions of:

• effects on glycemic control
• compliance
• effects on lipid profile
• effects on endogenous insulin con- centration and exogenous insulin dose
• when to use combination oral agent/ insulin therapy
• cost of the various regimens
• pros and cons of combination therapy versus insulin monotherapy

Sulfonylurea/insulin combinations have been used since the late 1950s, when sulfonylureas were introduced to widespread clinical use.¹ Combination metformin/insulin therapy has been studied for more than a decade.² Similarly, combination acarbose/insulin therapy has been evaluated over the past few years. The thiazolidinedione troglitazone, which was approved by the U.S. Food and Drug Administration (FDA) in late 1996, has as its primary indication combination insulin therapy.

RATIONALE FOR THE USE OF
COMBINATION THERAPY
Type II diabetes mellitus is a heterogeneous, chronic, progressive disease. It is characterized by diminished tissue sensitivity (in both hepatic and muscle tissue) along with relatively impaired ß-cell function.³

While its course is variable and sometimes difficult to predict, type II diabetes more frequently than not follows a rather predictable sequence. Initially, a patient with impaired glucose tolerance (IGT) is insulin resistant and hyperinsulinemic.⁴ At this point, there are sufficient pancreatic reserves to override the insulin resistance, and the patient remains relatively euglycemic. The cause for hepatic and peripheral insulin resistance is not completely understood. However, the resistance is probably heterogeneous and may be secondary to alterations in insulin-binding capacity (<1% of cases), abnormalities in insulin-regulated glucose transporters (GLUT-4), or other causes. Eventually, the insulin resistance escalates to a point where the patient becomes overtly diabetic and in many instances remains hyperinsulinemic. Over time, the patient’s endogenous insulin reserves may be reduced, and often after 10–15 years of disease, insulinopenia may be present.

From a treatment standpoint, diet and exercise therapy may be a reasonable treatment option for patients with IGT or early type II diabetes. Patients with more severe or longstanding type II disease may be more appropriately managed by the addition of an oral agent. As insulin reserves are reduced, combination therapy with insulin and oral agents may be more effective. Finally, patients with loss of insulin reserves may require multiple injections of insulin.

Combination therapy is used today for several different reasons, which will be expanded on in the discussion of the specific agents. In general terms, reasons for combination therapy include the following.

First, since insulin resistance is present in essentially all patients with type II diabetes who have fasting plasma glucose levels of >140 mg/dl,³ it seems reasonable to incorporate a medication into the patients’ regimens that will ameliorate or at least mitigate this fundamental abnormality. Second, the use of combination therapy may reduce the intensity of some of the side effects of insulin.

Third, combination oral agent/insulin therapy usually allows for comparable, if not better, glycemic control with a lower exogenous insulin dose. This prospect is appealing to many practitioners secondary to their concern regarding epidemiological studies that have demonstrated a correlation between hyperinsulinemia (patients with and without diabetes, treated and not treated with exogenous insulin) and the risk of macrovascular disease.⁵ However, no prospective studies to date have demonstrated that exogenous insulin accelerates macrovascular disease, and a recent consensus statement from the American Diabetes Association (ADA) concluded that "(e)xogenous insulin administration does not have direct adverse effects on cardiovascular events and may even favorably affect the cardiovascular risk profile if improved glycemic control and lipid profile are sustained."⁶ Therefore, the validity of this line of reasoning has not been adequately demonstrated in the literature.

Fourth, combination oral agent/ insulin therapy serves as a palatable transition for many patients who are not responding adequately to oral agent therapy alone. It allows them time to adjust to the idea and mechanics of using a simple insulin regimen while at the same time continuing to receive secondary coverage from the oral agent. Most patients easily transition from combination therapy to insulin monotherapy.

SULFONYLUREAS
Sulfonylureas are widely used in the United States in the management of type II diabetes mellitus and have been used in combination with insulin since their inception. They are undoubtedly the most studied oral agents in the arena of combination therapy. Recently, the sulfonylureas were cited to account for approximately 1% of all prescriptions written in the United States.⁷ Presumably, many of these prescriptions were for combination insulin therapy.

A new sulfonylurea, glimepiride, was recently introduced to the market in the United States. Glimepiride is the only sulfonylurea that carries an FDA approval for combination sulfonylurea/ insulin use. However, its superiority to other sulfonylureas in this regard has not been demonstrated.

Mechanism of Action and Rationale for Combination Use
Sulfonylureas exert both pancreatic and extrapancreatic effects but are useful only in patients with viable ß-cells.^8,9 Sulfonylureas directly stimulate the release of insulin.⁷ In vivo, sulfonylureas sensitize ß-cells to glucose, increasing insulin secretion indirectly. Thus, under the influence of sulfonylureas, more insulin is secreted at all glucose levels than would be secreted in the absence of the sulfonylurea.⁸ Other potential pancreatic effects include inhibition of glucagon release.

Sulfonylureas may also affect glucose levels by several extrapancreatic mechanisms, such as increasing insulin receptor binding affinity, increasing insulin effect by a post-receptor action, and decreasing hepatic insulin extraction. The relative clinical importance of each of these mechanisms of action is still a subject of research and debate.⁸

Efficacy of Combination
Sulfonylurea/Insulin Therapy
Since the 1950s, dozens of studies evaluating combination sulfonylurea/insulin therapy have been published. In the past few years, two major meta-analyses of studies evaluating sulfonylurea/insulin therapy have been published.

The first of these meta-analyses evaluated studies published between 1966 and 1991.⁹ This included 17 randomized, controlled trials of at least 8 weeks in length and studies with at least one of the following outcome parameters: fasting glucose, HbA_lc, or C-peptide. The researchers calculated effect size and weighted mean changes of the three outcome parameters for control and treatment groups.

In the treatment group, the fasting plasma glucose concentration decreased from a mean of 206 mg/dl at baseline to a mean of 165 mg/dl post-treatment. The control group showed a reduction from 204 to 194 mg/dl. HbA_lc values in the treatment group dropped from 11.0 to 10.2%. Conversely, HbA_lc values in the control group increased from 11.0 to 11.2%. Fasting C-peptide concentrations in the treatment group increased from 1.45 to 1.75 ng/ml and were reduced in the control group from 1.42 to 1.30 ng/ml. This analysis concluded that "combined insulin-sulfonylurea therapy leads to modest improvement in glycemic control compared with insulin therapy alone. With combined therapy, lower insulin doses may be used to achieve similar control."

A more recent meta-analysis evaluated studies published from 1980 to 1992 using the National Library of Medicine’s Medline database.¹⁰ Sixteen of the forty-three cited studies were included in the analysis after strict inclusion criteria were applied (randomized, placebo-controlled, homogenous target population, uniform outcome measures). The outcome measures evaluated included body weight, serum glucose concentrations, glycohemoglobin, C-peptide concentrations, and daily insulin dose.

In the treatment group, the fasting plasma glucose concentration decreased from a mean of 210 mg/dl at baseline to a mean of 167 mg/dl post-treatment. A reduction from 209 to 202 mg/dl was demonstrated in the control group. The HbA_lc in the treatment group dropped 1.1%, from 11.2 to 10.1%. In the control group, HbA_lc values were only slightly reduced, from 11.1 to 10.9%. Fasting C-peptide concentrations in the treatment group increased from 0.75 to 0.91 nmol/L and were reduced in the control group from 0.68 to 0.67 nmol/L. Insulin dose was reduced from a mean of 50 to 38 U/day in the treatment group and was reduced by 1 U (46 to 45 U/day) in the control group. Patients in the treatment group experienced a mean weight change from 76 to 78 kg, while those in the control group had an increase in weight from 76 to 77 kg. This study concluded that "combination therapy with insulin and sulfonylureas may be a more appropriate and suitable option to insulin monotherapy in subjects with non-insulin-dependent diabetes mellitus in whom primary or secondary failure to sulfonylurea developed. It may be a more cost effective way of long-term management in this group of subjects, especially in the elderly."

Potential Problems
The primary side effects of sulfonylureas in these studies included hypoglycemia, weight gain, and an increase in endogenous insulin secretion. Unfortunately, most sulfonylurea/insulin trials have not reported the relative incidence of hypoglycemia in their populations. However, it can be expected that the rates of hypoglycemia would be at least as high and probably higher than rates experienced in the sulfonylurea monotherapy patients.The incidence of hypoglycemia in this population is variable, but one study revealed a 20% chance of hypoglycemia every 6 months in patients treated with sulfonylurea monotherapy.⁷

Weight gain is common with sulfonylurea therapy. One study evaluating the effects of glyburide monotherapy reported a mean weight increase of 2.8 (± 0.7) kg, while another study evaluating the effects of tolbutamide reported a mean weight increase of 1.8 kg.^11,12 However, weight gain observed in one of the above-mentioned meta-analyses of combination therapy was modest (2 kg) and was similar to weight change experienced in the control group (l kg).¹⁰

Studies of sulfonylurea/insulin combination therapy have shown consistently that exogenous insulin doses may be reduced in responding patients. Nevertheless, endogenous insulin secretion is significantly enhanced. While endogenous insulin enters the circulation directly via the portal vein and may have different metabolic effects than subcutaneous insulin, there are no data to support the hypothesis that insulin delivery into the portal vein has unique advantages.¹³ Thus, the state of hyperinsulinemia, although altered in its constitution, may still be present in a patient treated with combination sulfonylurea/insulin therapy. Also, the necessity of increasing insulin doses after 6 months of therapy has been noted by some.

Additional, less common side effects that may occur with the use of sulfonylureas include dermatological reactions, hematological reactions, and gastrointestinal disturbances.⁷ Disulfiram-like reactions and hyponatremia have been reported with chlorpropamide.⁷

ACARBOSE
Acarbose is an oral alpha-glucosidase inhibitor indicated for the management of hyperglycemia secondary to type II diabetes mellitus. It was released for use in the United States in early 1996, but has been used extensively in Europe and Canada for several years. Acarbose is a mild antihyperglycemic agent, which may be used a monotherapy in new- onset or mild type II diabetes mellitus and may also be used in combination with insulin or other agents in more severe type II disease.

Mechanism of Action and Rationale for Combination Use
Acarbose competitively inhibits alpha-glucosidase enzymes in the brush border of the small intestine.¹⁴ Additionally, acarbose inhibits pancreatic alpha-amylase.¹⁵ Alpha-amylase is responsible for the hydrolysis of complex starches to oligosaccharides in the lumen of the small intestine.

The alpha-glucosidase enzymes are responsible for the hydrolysis of oligosaccharides, trisaccharides, and disaccharides in the brush border of the small intestine and include maltase, isomaltase, glucoamylase, and sucrase. Inhibition of these enzyme systems effectively reduces the rate of complex- carbohydrate digestion and the subsequent absorption of glucose, which reduces postprandial glucose excursions in patients with diabetes.

Acarbose reduces postprandial glucose by mechanisms that are additive to the effects of insulin. Therefore, it might offer a rational alternative for combination oral agent/insulin therapy. Additional clinical features of acarbose that make it an attractive alternative for combination therapy include 1) no change in body weight, 2) decrease in postprandial hyperglycemia, 3) increase in postprandial glucose-like peptide 1, and 4) possible small reduction in postprandial triglyceride concentrations.¹⁶

Efficacy of Combination Acarbose/Insulin Therapy
Acarbose has also been shown to be effective when used in combination with metformin, sulfonylureas, and insulin.¹⁷ This study evaluated the use of acarbose in 354 patients with type II diabetes mellitus who were previously treated with diet and insulin (91 patients), diet and sulfonylureas (103 patients), diet and metformin (83 patients), or diet therapy only (77 patients).

The study began with a 6-week run-in period, during which patients were asked to follow a diet and take their previous medications. Patients then continued their previous therapy and additionally were randomized to either placebo or acarbose. Acarbose dose was titrated to therapeutic effect (1-hour postprandial glucose concentration of <180 mg/dl without intolerable side effects) in doses ranging from 50 to 200 mg three times daily.

HbA_lc, fasting and postprandial glucose concentrations, fasting and postprandial C-peptide concentrations, and fasting serum lipids were evaluated at baseline and at 3-month intervals. HbA_lc values at 12 months were lower in patients receiving acarbose than in those receiving placebo in all of the treatment groups. The difference was 0.9% in the diet-only group (P = 0.005), 0.8% for the metformin group (P = 0.011), 0.9% for the sulfonylurea group (P < 0.002), and only 0.4% for the insulin group (P = 0.077).

Fasting plasma glucose concentrations at 12 months were also lower when compared with baseline in patients receiving acarbose in all of the treatment groups. The difference in the diet-only group was -0.7 mmol/L with acarbose versus 1.4 mmol/L with placebo (P = 0.01). In the metformin group, the difference was -1.5 mmol/L with acarbose versus 0.0 mmol/L with placebo (P = 0.11). The difference in the sulfonylurea group was 0.2 mmol/L with acarbose versus 1.6 mmol/L with placebo (P = 0.01). In the insulin group, the difference was 0.1 mmol/L with acarbose versus 0.1 mmol/L with placebo (P = 0.97). Therefore, no significant change in fasting plasma glucose was observed when acarbose was added to insulin.

Acarbose caused a significant reduction when compared with placebo in postprandial glucose concentrations in each of the groups. In the diet-only group, concentrations were reduced from 7.7 mmol/L at baseline to 13.2 mmol/L at 12 months. In the metformin group, the reduction was from 19.3 mmol/L at baseline to 15.8 mmol/L at 12 months. For sulfonylureas, the reduction was from 20.7 mmol/L at baseline to 16.6 mmol/L at 12 months. In the insulin group, the reduction was from 18.4 mmol/L at baseline to 15.7 mmol/L at 12 months. No statistically significant changes in C-peptide or serum lipids were noted during the study.

In summary, fasting plasma glucose levels in patients treated with acarbose and insulin were not different from patients treated with insulin and placebo. HbA_lc concentrations were 0.4% lower in the acarbose/insulin group than in the placebo/insulin group. Based on this study, the addition of acarbose to insulin therapy seems to provide a modest improvement in glycemic control.

Potential Problems
The common side effects of acarbose that would be expected to occur with combination therapy are dose-related gastrointestinal complaints, which may be attenuated by continued administration of the medication. These side effects include flatulence, diarrhea, and abdominal pain, which occurred in the U.S. phase III trials with an incidence of 77%, 33%, and 21%, respectively.¹⁵

While elevated hepatic enzymes have been reported at higher doses, elevated serum transaminase levels were no more frequent than those observed with placebo when doses of 100 mg three times daily or less were used.

Acarbose should be avoided in patients with inflammatory bowel disease, colonic ulceration, or obstructive bowel disorders. Relative contraindications to the use of this drug include chronic intestinal disorders of digestion or absorption and medical conditions that might deteriorate with increased intestinal gas formation.¹⁵ Finally, acarbose therapy should be avoided in patients with serum creatinine levels of >2.0 mg/dl, since studies have suggested increases in acarbose plasma concentrations with renal dysfunction and long-term studies have not been carried out in this population.¹⁵

Another problem that might be encountered in patients treated with acarbose, and insulin is difficulty in treating hypoglycemic episodes. The absorption rates of sucrose and other complex carbohydrates are drastically reduced with the administration of acarbose, and this may result in prolonged hypoglycemia. Therefore, patients treated with combination therapy should always be counseled to have oral glucose tablets or gel on hand in the event of a hypoglycemic episode.

METFORMIN
The biguanide metformin was introduced in 1957 as an antihyperglycemic agent in patients with type II diabetes mellitus.¹⁸ Although the FDA only recently approved it for use in the United States, metformin has been used worldwide since the late 1950s and is a safe and effective medication for the management of hyperglycemia in patients with type II diabetes mellitus.

Metformin is also effective when used in combination with other oral agents and insulin. Interestingly, a recent study demonstrated that 76% of insulin-treated patients (42 of 55 patients) successfully discontinued insulin when treated with the combination of metformin and glyburide.¹⁹ HbA_lc was significantly reduced (by an average of 1.3%), and patients experienced continued weight loss (average 5 lbs) after 6 weeks. Most of the primary and secondary failures were successfully managed with combination metformin and insulin.

Mechanism of Action and Rationale for Combination Use
Metformin is an oral antihyperglycemic agent, not a hypoglycemic agent. Its mechanism of action differs from that of sulfonylureas in that it does not increase insulin secretion, is not associated with hypoglycemia, and does not produce weight gain.

Metformin has two primary sites of action: the liver and muscle tissue. It lowers blood glucose by enhancing insulin-stimulated glucose transport in skeletal muscle. The range of observed enhancement of glucose uptake ranges from 10–40%, depending on the population being evaluated.²⁰

Metformin has also been shown in numerous studies to reduce hepatic glucose production in patients with type II diabetes. Initially, this effect was thought to be mediated via a reduction in glycogenolysis.²¹ However, a more recent study has suggested that this effect may be secondary to a reduction in gluconeogenesis.²² Metformin, in effect, reduces blood glucose concentrations via amelioration of insulin resistance in the liver and in the periphery.

In addition to its effects on glucose, metformin therapy has been associated with a reduction in triglyceride concentrations (16%), LDL cholesterol (8%), and total cholesterol (5%) and was associated with an increase in HDL cholesterol (2%).²³ Also, as noted above, patients treated with metformin may lose significant amounts of weight. The mechanism of action, along with other metabolic effects, make metformin a reasonable candidate for use with combination oral agent/insulin therapy.

Efficacy of Combination Metformin/Insulin Therapy
A trial of 53 patients with type II diabetes mellitus who were not adequately controlled with either maximum doses of a sulfonylurea or metformin compared the effects of randomizing these patients to either insulin monotherapy or combination oral/insulin therapy.²⁴ The study concluded that both combination therapy and insulin monotherapy were equally effective and well tolerated. However, the average insulin dose at 6 months was 15 U for the combination group and 57.2 U for the insulin monotherapy group. Also, insulin monotherapy patients gained an average of 5.2 kg in 6 months, while the patients treated with combination therapy gained an average of 2.1 kg. Unfortunately, the researchers did not do a subanalysis of the patients receiving metformin versus those on sulfonylureas.

In another study, 50 obese patients with type II diabetes mellitus who were poorly controlled with insulin monotherapy were randomized to the addition of placebo or metformin.²⁵ This study was doubleblind, lasted 6 months, and included a 4-week run-in period in which all patients received placebo. Metformin was administered in doses of 850 mg twice daily.

At 6 months, serum glucose profile concentrations (mean of every 2-hour measurement from 8:00 a.m. to 8:00 p.m.) dropped from 268 to 176 mg/dl in the metformin/insulin group and from 261 to 253.1 mg/dl in the insulin monotherapy group. HbA_lc values were significantly (P < 0.05) lower in the combination therapy group than in the insulin monotherapy group at 6 months (9.8 vs. 11.3%). Insulin doses were significantly (P < 0.05) reduced by ~25% over 6 months in the combination therapy group (90 to 68.4 U) and were not significantly reduced in the monotherapy group (88 to 85.8 U). Total cholesterol and serum triglycerides were significantly (P < 0.05) reduced in the combination group at 6 months when compared to baseline and to 6-month measures in the insulin monotherapy group.

Potential Problems
In the study mentioned above, the only reported adverse effect of metformin was diarrhea, which occurred in 2 of the 27 patients and was self-limiting. However, patients in the U.S. pivotal trials treated with metformin had 30% more reports of abdominal bloating, nausea, cramping, a feeling of fullness, and diarrhea than did patients receiving placebo.²⁶ As in the Gugliano study mentioned above, these side effects were usually self-limiting and transient. In many instances, these effects can be mitigated by starting with a low dose and titrating up slowly and also by taking the medication with food.

Additional, less common side effects include metallic taste and a reduction in vitamin B-12 levels. Lactic acidosis can occur with the administration of metformin but is extremely rare (0.03 cases per 1,000 patient-years). It has occurred primarily in patients with significant renal dysfunction.²⁶

The use of metformin in either monotherapy or combination therapy is contraindicated in patients with renal dysfunction (serum creatinine >1.5 mg/dl in men or >1.4 mg/dl in women), since metformin is excreted renally and can accumulate in patients with renal dysfunction. Because acidosis is sometimes associated with hepatic dysfunction, metformin is contraindicated in patients with clinical or laboratory evidence of hepatic dysfunction. It is also contraindicated in patients with acute or chronic lactic acidosis or a history of alcoholism or binge drinking. Finally, metformin should be temporarily withheld in patients with acute conditions predisposing them to acute renal failure or acidosis, such as cardiovascular collapse, acute myocardial infarction, acute exacerbation of congestive heart failure, use of iodinated contrast media, or a major surgical procedure.²⁶

TROGLITAZONE
Troglitazone (CS-045), a thiazolidindione, was approved by the FDA for use in the United States in December 1996. Two other thiazolidinediones, pioglitazone and ciglitazone, which were also being evaluated, have been discarded due to their adverse effect profiles. Currently, troglitazone is approved for use in patients with type II diabetes mellitus whose glucose is not adequately controlled by >30 U of insulin per day.

Mechanism of Action and Rationale for Combination Use
Troglitazone appears to work by enhancing insulin action without affecting insulin secretion.²⁷ The thiazolidinediones have been referred to as "insulin sensitizers." Several animal and human studies have demonstrated that troglitazone enhances insulin action at the receptor and postreceptor levels in hepatic and peripheral tissues.

Studies have confirmed that troglitazone does not enhance endogenous insulin secretion. One study in type II patients concluded that troglitazone was associated with a reduction in plasma glucose levels and in endogenous insulin concentration.²⁸ The findings suggest that troglitazone is not only an antihyperglycemic compound but also a compound that reverses insulin resistance.

When used as monotherapy, response rates have been ~75%.^28,30,31 An analysis of individual data has revealed that all patients demonstrate an improvement in insulin resistance after troglitazone monotherapy.³² The patients who had a lower concomitant reduction in plasma glucose were noted to have the lowest levels of insulin secretion. It has been hypothesized that thiazolidinediones will improve insulin resistance in virtually all patients with type II diabetes mellitus but that improved insulin action will occur only in those with sufficient pancreatic response to glycemic challenge.

An association between insulin resistance and hypertension and other cardiac risk factors has been reported. Thiazolidinediones may have a positive impact on these risk factors secondary to its ability to reduce insulin resistance. A significant correlation was demonstrated between blood pressure reduction (164 to 146 mmHg diastolic) and a decrease in plasma insulin levels (9.1 to 6.3 µU/ml) in patients treated with troglitazone.³³

One study compared the effects of troglitazone to glyburide on the following cardiac risk factors: left ventricular mass index (LVMI), cardiac index (CI), stroke volume index (SVI), total cholesterol, HDL, LDL, and triglycerides.³⁴ Patients treated with troglitazone had no significant change in LVMI at 48 weeks. They had significant increases in CI, SVI, and HDL concentrations at 48 weeks, while no significant changes were observed in glyburide-treated patients. Additionally, statistically significant reductions in triglyceride concentrations were reported in the troglitazone group, while no changes were observed in the glyburide group. Mild increases in total cholesterol (troglitazone: 5.64 to 5.93 mmol/L; glyburide: 5.76 to 5.80 mmol/L) and LDL (troglitazone: 3.48 to 3.77 mmol/L; glyburide: 3.56 to 3.58 mmol/L) were observed in both groups.

Considering the effects of troglitazone on glucose, insulin, glucose disposal, insulin resistance, lipids, and other cardiac risk factors, the use of this agent in combination therapy seems rational.

Efficacy of Combination
Troglitazone /Insulin Therapy
One study of 222 insulin-treated patients with type II diabetes mellitus compared the effects of placebo to 200 mg/day and 400 mg/day of troglitazone.³⁵ This study was a randomized, double-blind, placebo-controlled, parallel group, multicenter trial. The patients were obese (average BMI 35 kg/m²), had a mean fasting serum glucose (FSG) of 224.5 mg/dl, had a mean HbA_lc of 9.2%, and at baseline were treated with a mean insulin dose of 72 U/day. After a 2- to 4-week single-blind, placebo baseline phase, patients were randomized to placebo or to 200 or 400 mg of troglitazone. Outcome measures included change from baseline in total daily insulin dose (TDID), percentage of those discontinuing insulin, percentage with a >50% reduction in TDID, change in number of insulin injections, HbA_lc, FSG, mean fasting blood glucose (MFBG), weight, BMI, blood pressure, and safety.

Mean reductions at 6 months in MTID for patients treated with placebo, 200 mg of troglitazone, and 400 mg troglitazone were 13 U (18%), 30 U (41%), and 41 U (57%), respectively. Seventy percent of patients treated with 400 mg/day had at least a 50% decrease in TDID. Patients who were able to discontinue insulin altogether included 1% of placebo patients, 7% of the 200 mg/day group, and 15% of the 400 mg/day group. Patients in the 400 mg/day group had a mean decrease of 2 injections per day at 6 months. Reductions in HbA_lc of 0.09, 0.13, and 0.41% (P = 0.04) in the placebo, 200 mg, 400 mg groups, respectively, were reported. MFBG and FSG were reduced significantly in the 400 mg/day group. No statistically significant changes or trends were observed in weight, BMI, or blood pressure.

Another clinical trial that compared placebo to 200 or 600 mg/day of troglitazone in insulin-treated patients with type II diabetes mellitus reported that 30% of the 200 mg/day group and 57% of the 600 mg/day group had reductions in HbA_lc to below 8% at 6 months.³⁶ Reductions in TDID for the placebo, 200 mg/day, and 600 mg/day group were 1%, 15%, and 42%, respectively.

Potential Problems
In the North American trials for troglitazone (n = 2,510 patients), 20 patients experienced liver function test abnormalities.³⁶ Two patients developed reversible jaundice with biopsies that were consistent with idiosyncratic drug reaction. Troglitazone may cause resumption of ovulation in premenopausal anovulatory patients with insulin resistance, which could be considered a positive or negative effect based on the perspective and wishes of the patient. In general, troglitazone is well tolerated and is virtually without side effects in the majority of patients.

INSULIN/ORAL AGENT
THERAPY
When Should Combination Insulin/Oral Agent Therapy Be Employed?
The initiation of combination therapy can, of course, take place in two scenarios: for patients unsuccessfully treated with insulin monotherapy and for those unsuccessfully treated with oral agents alone. Additionally, the advent of the thiazolidinediones has raised the question: "Should combination therapy be used early on, rather than as a second- or third-ditch effort?"

Insulin monotherapy should be considered inadequate in most cases when the recommended ADA goals (Table 1) are not met despite a reasonable dose of insulin.

What is a reasonable dose of insulin? In the recent past, when only sulfonylureas were available, doses of >70 U/day of insulin were considered excessive, and in these cases combination therapy was promoted.³⁷ However, given data in the past 10 years regarding insulin resistance, the possible deleterious role of hyperinsulinemia, and the release of oral agents that attenuate or at least mitigate insulin resistance, should earlier combination insulin oral therapy be routinely considered? Perhaps patients not well controlled on more than 30 U/day should be considered candidates for combination therapy.

Patients should be considered oral-agent failures if they do not show an adequate response to sulfonylurea/metformin combination therapy at therapeutic doses. If other oral regimens have been used unsuccessfully, this author believes that sulfonylurea/metformin should be attempted before the initiation of insulin if the patient is hesitant to use insulin despite adequate education. In that case, the clinician can at least honestly assure the patient that the most potent known combination of oral agents was used and, if unsuccessfully so, then insulin is really the only alternative. The use of triple oral therapies lies in uncharted territory, and while it may be effective, poor compliance and high cost might obviate its use.

Choice of Insulin and Insulin
Regimen
Insulin has been widely used since 1922 as monotherapy and since the late 1950s in combination therapy management of hyperglycemia. It has been estimated that approximately 29% of all patients diagnosed with diabetes mellitus are type II diabetes patients treated with insulin,³⁸ and presumably many of these patients are being treated with combination therapy.

While a myriad of articles has been published delineating the metabolic effects of insulin, very few studies have directly compared the efficacy of various insulin regimens in type II diabetes patients with insulin monotherapy or combination therapy. Most combination trials carried out to date have simply enrolled patients who were already treated with insulin therapy and added an oral agent or, conversely, compared placebo to a single-insulin regimen added to oral therapy.

Even given the vast array of insulins available today, the pharmacokinetics and pharmacodynamics of conventional exogenous insulin sometimes limit its ability to attenuate the metabolic effects of type II diabetes.³⁸ This is primarily due to the fact that no conventional preparation can provide the low insulin concentrations needed to suppress hepatic glucose production (i.e., a "basal" insulin) in the postabsorptive phase, and, until the release of lispro, no preparations were able to provide the high concentrations needed to stimulate post-prandial peripheral glucose disposal.

The ability of insulin therapy to lower blood glucose levels is governed by dose, regimen, level of insulin resistance, and other factors. Insulin is effective in reducing blood glucose levels in patients with type II diabetes when used as monotherapy^5,38 and as cited above, when used in combination with sulfonyulreas, metformin, troglitazone, and acarbose.

Probably the most widely accepted combination insulin/oral agent therapy regimen for previously insulin-naive patients who are currently treated with oral agents is single-injection intermediate-acting insulin administered at bedtime. In this scenario, a dose of 10 U of intermediate-acting insulin administered at 2200 or 2300 has been suggested.³⁷ This dose is titrated by 3–5 U every 3–4 days, based on fasting blood glucose concentrations.

As an alternative to the above-mentioned regimen, lispro has been shown to be effective when used in combination with sulfonylureas.³⁹ When compared with sulfonylurea therapy, combined therapy resulted in reductions in FPG (154 vs. 193 mg/dl, P < 0.0001), 2-hour postprandial plasma glucose (251 vs. 331 mg/dl, P < 0.0001), and HbA_lc (7.2 vs. 9%, P < 0.0001). Body weight was higher in the combination therapy group (209 vs. 202 lb, P < 0.0001). Another study evaluating the effects of lispro on glucose tolerance in patients with type II diabetes concluded that lispro restored early plasma insulin rise, improved glucose tolerance, induced a prompt and transient hepatic glucose production inhibition, and reduced late hyperinsulinization.⁴⁰

When combination therapy is initiated in patients taking insulin, the oral agent should be slowly titrated up while insulin doses are slowly titrated down. The rapidity with which this should occur should be based on the oral agent being used and on the clinical status of the patient. In many cases, insulin regimens may be simplified (reduction in the number of injections), and in some cases insulin may be discontinued.

Potential Problems
Before the institution of combination therapy, several possible side effects, including hypoglycemia, weight gain, and the potential for accelerated macrovascular disease, should be considered with insulin and lispro use in patients with type II diabetes.

Severe hypoglycemia, while a major concern in patients with type II diabetes, probably occurs with less frequency than that observed in patients with type I diabetes.⁵ Additionally, the counterregulatory response to hypoglycemia in patients with type II diabetes is blunted to a lesser degree than in patients with type I disease. The annual rates of hypoglycemia and severe hypoglycemia (requiring the assistance of another person) in insulin-taking patients in the 3-year follow-up of the UK prospective study (UKPDS) were 33.4 and 1.4%, respectively.⁴¹ The annual rates of hypoglycemia and severe hypoglycemia in glyburide-managed patients in this trial were 27.8 and 1.3%, respectively.

One could assume that the rates of hypoglycemia would be similar or higher (if the therapy is effective) for combination therapy. The incidence of any hypoglycemia in patients treated with sulfonylurea/metformin combination was reported to be 40%, combination ultralente/sulfonylurea to be 33%, and ultralente/regular insulin to be 47%.⁴² However, unlike the Diabetes Control and Complications Trial⁴³ and the UKPDS, none of the cases of hypoglycemia in this study were incapacitating.

Cost, Patient Adherence, and
Effectiveness
In addition to contraindications, clinicians must also consider the cost, level of patient adherence, and medication effectiveness when choosing the most appropriate regimen for a given patient. The following paragraphs and tables summarize these parameters.

Cost. A cursory consideration of the cost of therapy must include actual cost of the medication and cost of baseline and sequential lab work. Table 2 gives these parameters for the most commonly used medications.

Drug costs are based on average wholesale price (AWP) of the drug plus an average customary pharmacy charge (15% of AWP) for 12 monthly prescriptions. While markup varies considerably from one insurance plan to another and from one pharmacy to another, Table 2 offers a reasonable estimate and, applied across the board, should offer a good comparison. Insulin and syringe costs are based on patient charges applied in many chain pharmacies ($16.39/vial of human NPH or 70/30 and $20/100 U-100 syringes).

Laboratory costs are simply estimates of patient charges and vary from area to area. An estimated cost of $65 was applied for a standard electrolyte, liver function, lipid, and renal panel (Panel A), $80 for this panel plus a complete blood count (CBC) with differential (Panel B), and $20 for an HbA_lc measurement. The frequency of the lab tests will also vary widely, but in this instance it was based on ADA recommendations and strict monitoring criteria. For example, not every practitioner will feel the need for a quarterly electrolyte panel in order to assess renal function in every patient treated with metformin; however, in many cases this would be a prudent course of action.

Readers are encouraged to evaluate Table 2 in the context of their own practices. Costs are expressed in dollars per year of therapy.

Adherence. Another consideration in choice of a regimen should always be degree of patient adherence to the regimen. Under this rubric, two major factors must be considered: ease of use and discomforting side effects. Ease of use of oral tablets and capsules is mainly governed by the number of daily doses needed. Adherence drops dramatically when one goes from daily dosing to twice daily and from twice daily to three times daily. Discomforting side effects can also limit patient adherence to medication regimens. It was noted, for example, that a significant number of patients treated with chlorpropamide in the UKPDS refused the medication after they experienced ethanol-induced flushing.

Table 3 ranks sulfonylureas, acarbose, metformin, and troglitazone in terms of ease of use and discomforting side effects. The rankings are the author’s best estimate based on what is reported in the literature and on personal clinical experience. The rankings for "discomforting side effects" are, in a sense, subjective educated guesses and should be viewed as such. Under this heading, only side effects that the patient will commonly experience are considered. Ease of use is ranked from 1 (good) to 4 (difficult), while discomforting side effects are ranked from 1 (few) to 4 (many). Patient adherence is ranked based on the sum of the two listed factors, with low numerical scores suggesting greater patient adherence.

Effectiveness. Table 4 provides an overview of effectiveness by describing changes in fasting plasma glucose, postprandial glucose, HbA_lc, lipids, and insulin dose. This table does not include data on all parameters for each medication because, in some instances, data are not available.

SUMMARY
The recent introduction to the U.S. market of three new oral agents and one new insulin has dramatically increased the number of options available to practitioners managing patients with type II diabetes. No consensus exists regarding when and how combination oral agent/insulin therapy should be instituted. Clinicians must use their own judgment to determine when combination therapy is appropriate for a particular patient.

Once combination therapy is deemed appropriate, choice of the regimen should include reflection on patient-specific contraindications, costs, glycemic goal, medication effectiveness, and the likelihood of patient adherence. Hopefully, as more is learned about the pathogenesis of type II diabetes and the comparative effectiveness of various combination regimens on long-term outcomes, the most appropriate methods of management will become more evident.

REFERENCES

¹Bailey T, Mezitis N: Combination therapy with insulin and sulfonylureas for type II diabetes. Diabetes Care 13:687-95, 1990.

²Leblanc H, Marre M, Billault B, Passa PH: Combined continuous subcutaneous insulin influsion in 10 overweight insulin requiring diabetic patients. Diabete and Metab 13:613-17, 1987.

³Raskin P, Ed.: Medical Management of Non-Insulin Dependent Diabetes Mellitus. Third edition. Alexandria, Va., American Diabetes Association, p. 15-17.

⁴DeFronzo RA: Lilly Lecture 1987: The triumvirate: ß-cell, muscle, liver: a collusion responsible for NIDDM. Diabetes 37:667-87, 1987.

⁵Genuth S: Insulin use in NIDDM. Diabetes Care 13:1240-64, 1990.

⁶American Diabetes Association: Consensus statement: The pharmacologic treatment of hyperglycemia in NIDDM. Diabetes Care 19 (Suppl 1):S54-61, 1996.

⁷Gerich JE: Oral hypoglycemic agents. N Engl J Med 321:1232-45, 1989.

⁸Skillman, TG, Feldman JM: The pharmacology of the sufonylureas. Am J Med 70:361-69, 1981.

⁹Pugh JA, Wagner ML, Sawyer J, Ramirez G, Tuley M, Friedberg SJ: Is combination sulfonylurea and insulin therapy useful in NIDDM patients? A meta-analysis. Diabetes Care 15:953-59, 1992.

¹⁰Johnson JL, Wolf SL, Kabadi UM: Efficacy of insulin and sulfonylurea combination therapy in type II diabetes: a meta-analysis of randomized, placebo-controlled studies. Arch Int Med 156:259-64, 1996.

¹¹Hermann LS, Schersten B, Bitzen P, Kjellstrom T, Lindgarde F, Melander A: Therapeutic comparison of metformin and sulfonylurea, alone and in various combinations. Diabetes Care 17:1100-109, 1994.

¹²Coniff RF, Shapior JA, Seaton TB, Bray G: Multicenter, placebo-controlled trial comparing acarbose (BAY g5421) with placebo, tolbutamide, and tolbutamide-plus-acarbose in non-insulin-dependent diabetes mellitus. Am J Med 98:443-51, 1995.

¹³Lebovitz HE, Pasmantier R: Combination insulin-sulfonylurea therapy. Diabetes Care 13:667-75, 1990.

¹⁴Santeusanio F, Compagnucci P: A risk-benefit appraisal of acarbose in the management of non-insulin-dependent diabetes mellitus. Drug Safety 11:432-44, 1994.

¹⁵Precose™ package insert. West Haven, Conn., Bayer Pharmaceuticals.

¹⁶Lebovitz H: A new oral therapy for diabetes management: alpha glucosidase inhibition with acarbose. Clinical Diabetes 13:99-102, 1995.

¹⁷Chiasson J-L, Josse RG, Hunt JA, Palmason C, Rodger NW, Ross SA, Ryan EA, Tan MH, Wolever TMS: The efficacy of acarbose in the treatment of patients with non-insulin-dependent diabetes mellitus. Ann Int Med 121:928-35, 1994.

¹⁸Bailey C: Biguanides and NIDDM. Diabetes Care 15:755-772, 1992.

¹⁹Bell D, Mayo M: Outcome of metformin facilitated reinitiation of oral diabetic therapy in the insulin requiring non-insulin-dependent diabetic patient. Endocrine Pract 3:73-76, 1997.

²⁰Bailey CJ: Metformin: an update. Gen Pharmacol 24:1299-1309, 1993.

²¹DeFronzo RA, Barzilai N, Simonson DC: Mechanism of metformin action in obese and lean NIDDM subjects. J Clin Endocrinol Metab 73:1294-1301, 1993.

²²Stumvall M, Nurjhan N, Perriello G, Dailey G, Gerich J: Metabolic effects of metformin in non-insulin-dependent diabetes mellitus. N Engl J Med 333:550-54, 1995.

²³DeFronzo RA, Goodman AM, the Multicenter Metformin Study Group: Efficacy of metformin in patients with NIDDM. N Engl J Med 333:541-49, 1995.

²⁴Chow CC, Tsang LW, Sorensen JP, Cockram CS: Comparison of insulin with or without continuation of oral hypoglycemic agents in the treatment of secondary failure in NIDDM patients. Diabetes Care 18:307-14, 1995.

²⁵Gugliano D, Quatraro A, Consoli G, Minei A, Ceriello A, DeRosa N, Onofrio ED: Metformin for obese, insulin-treated diabetic patients: improvement in glycaemic control and reduction of metabolic risk factors. Eur J Clin Pharmacol 44:107-112, 1993.

²⁶Glucophage package insert. Wilton, Conn., Bristol-Meyers Squibb.

²⁷Whitcomb RW, Saltiel AR: Thiazolidinediones. Exp Opin Invest Drugs 4:1299-1309, 1995.

²⁸Sutter S, Nolan J, Wallace P, Gumbiner B, Olefsky J: Metabolic effects of new oral hypoglycemic agent CS-045 in NIDDM subjects. Diabetes Care 15:193-203, 1992.

²⁹Iwamoto Y, Kuzuya T, Matsuda A, Awata T, Kumakura S, Inooka G, Shiraishi I: Effect of new oral antidiabetic agent CS-045 on glucose tolerance and insulin secretion in patients with NIDDM (Abstract). Diabetes Care 14:1083-86, 1991.

³⁰Valiquett T, Balagtas C, Whitcomb R: Troglitazone dose-response study in patients with NIDDM. Diabetes 44 (Suppl 1):109A, 1996.

³¹Kumar S, Boulton AJM, Beck-Nielsen H, Berthezene F, Muggeo M, Persson B, Spinas GA, Donoghue S, Lettis S, Stewart-Long P: Troglitazone, an insulin action enhancer, improves metabolic control in NIDDM patients. Diabetologia 39:701-709, 1996.

³²Saltiel AR, Olefsky JM: Thiazolidinediones in the treatment of insulin resistance and type II diabetes. Diabetes 45:1661-69, 1996.

³³Ogihara T, Rakugi H, Ikegami H, Mikami H, Masuo K: Enhancement of insulin sensitivity by troglitazone lowers blood pressure in diabetic hypertensives. Am J Hypertension 8:316-20, 1995.

³⁴Ghazzi MN, Perez JE, Antonucci TK, Driscoll JH, Huang SM, Faja BW, Whitcomb R, the Troglitazone Study Group: Cardiac and glycemic benefits of troglitazone treatment in NIDDM. Diabetes 46:433-39, 1997.

³⁵Buse J, Gumber B, Mathias N, Nelson D, Whitcomb R, the Troglitazone Insulin Study Group: Personal correspondence.

³⁶Rezulan package insert. Morris Plains, NJ, Parke Davis.

³⁷Lebovitz HE: Combination therapy for hyperglycemia. In Therapy for Diabetes Mellitus and Related Disorders. Second edition. Lebovitz H, Ed. Alexandria, Va., American Diabetes Association, 1994, p. 147-51.

³⁸Galloway JA: Treatment of NIDDM with insulin agonists or substitutes. Diabetes Care 13:1209-39, 1990.

³⁹Feinglos MN, Lane JD, Thacker CH, English JS: Pre-prandial insulin lispro improves glucose control in patients with secondary sulfonylurea failure (Abstract). Diabetes 45 (Suppl 2): 286A, 1996.

⁴⁰Bruttomesso D, Pianta A, Valerio A, Tadiotto F, Lavagnini T, Del Prato S: Fast acting insulin lispro improves glucose tolerance of NIDDM by inhibiting hepatic glucose production. Diabetes 45 (Suppl 2):140A, 1996.

⁴¹United Kingdom Prospective Diabetes Study (UKPSD) 13: Relative efficacy of randomly allocated diet, sulfonylurea, insulin, or metformin in patients with newly diagnosed non-insulin-dependent diabetes followed for three years. British Med J 310:83-88, 1995.

⁴²Holman RR, Steemson J, Turner RC: Sulfonylurea failure in type 2 diabetes: treatment with basal insulin supplement. Diabetic Med 4:457-62, 1987.

⁴³The DCCT Research Group: The effect of intensive insulin treatment on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med 329:977-86, 1993.

John R. White, Jr., PharmD, is an associate professor of pharmacy practice and director of the Washington State University/Sacred Heart Medical Center Drug Studies Unit, in Spokane, Wash.

Note of disclosure: Dr. White has received speaking honoraria from Eli Lilly, Bristol-Myers Squibb, Parke Davis, and Bayer Pharmaceuticals and has worked as a paid consultant for Eli Lilly and Parke Davis.

Return To American Diabetes Association Home Page

Copyright © 1997 American Diabetes Association

Last updated: 7/3/97
For ADA Related Issues contact CustomerService@diabetes.org

For Technical Issues contact webmaster@diabetes.org