Rationale For Combination Therapy

In the context of two major physiologic defects, insulin resistance and insulin secretory failure, combination treatments with differing actions are entirely logical. Several different advantages of combining agents can be distinguished.


The first rationale for combination therapy (either with oral agents alone or with oral agents and insulin or oral agents with other injectable medicine such as incretin mimetics) is its superior efficacy. One principle that emerges from randomized controlled trials of antidiabetic therapies is that switching from one medication to another does not work as well as combination therapy. Figure 4 shows a classic study of combination oral agent therapy that illustrates this point (11). Patients with inadequate glycemic control on maximal doses of glyburide were randomized to continuation of that monotherapy, to metformin monotherapy gradually titrated to maximal doses (850 mg orally three times a day), or to a combination of glyburide and metformin. Neither monotherapy resulted in any significant improvement in fasting plasma glucose (FPG), but combination therapy with an insulin secretagogue and metformin, showed a dramatic improvement. Similarly, studies with other combination

FIGURE 4 As shown in the U.S. pivotal studies by DeFronzo et al. in patients failing sulfonylureas (maximal dose glyburide in the dotted line), continuation of "failing" secretagogue therapy resulted in gradual worsening of glycemia. Similarly, a switch from glyburide to metformin in maximally effective doses showed no significant improvement in glycemic control. In fact, the deterioration of glycemia shown by rise in fasting plasma glucose (FPG) was marked in the metformin group after glyburide had been stopped and before full dose titration of metformin had occurred. Switching makes little sense as the combination of therapy (solid line) showed a substantial improvement in FPG not observed with either drug alone.

FIGURE 4 As shown in the U.S. pivotal studies by DeFronzo et al. in patients failing sulfonylureas (maximal dose glyburide in the dotted line), continuation of "failing" secretagogue therapy resulted in gradual worsening of glycemia. Similarly, a switch from glyburide to metformin in maximally effective doses showed no significant improvement in glycemic control. In fact, the deterioration of glycemia shown by rise in fasting plasma glucose (FPG) was marked in the metformin group after glyburide had been stopped and before full dose titration of metformin had occurred. Switching makes little sense as the combination of therapy (solid line) showed a substantial improvement in FPG not observed with either drug alone.

therapies showed no benefit of switching to a new agent class, but greater glucose-lowering efficacy through combining it with an agent of a different mechanism.

Tolerability and Convenience

Many side effects of medications are dose related. For example, hypoglycemia is a side effect of insulin or insulin secretagogues, gastrointestinal side effects are common with metformin, and with alpha-glucosidase inhibitors, and fluid retention or weight gain may occur with the thiazolidinediones. These side effects are dose related, and using lower doses of medications and slow titration may minimize them. Combinations of oral agents, therefore, may minimize side effects while achieving equal or better glycemic control. This principle has been tested directly for the combination of glyburide and metformin (12). Combinations of oral agents may seem more complex than monotherapy, but in some cases their convenience can be enhanced. Combining a single dose of a long-acting sulfonylurea, such as glimepiride, with one or two tablets of metformin, may have equal or more benefit than three or four tablets of metformin alone. Metformin-secretagogue (metformin-glyburide, metformin-glipizide) combination pills (Glucovance, Metaglip), have been introduced. Similarly, metformin-thiazoli-dinedione (metformin-rosiglitazone, metformin-pioglitazone) combination pills (Avandamet, Actosplusmet) are available. Recently, thiazolidinediones-secretagogue (rosiglitazone-glime-piride, pioglitazone-glimepiride) combination pills have become available (Avandaryl, Duetect). This trend will likely increase in the future. Convenient formulations of two agents in a single pill with dual actions may appeal to many patients and practitioners. While separate titration of agents may be desirable for many patients, for others a case can be made for combination preparations. This tactic may prove especially attractive for patients who must take not only two or more agents for glycemic control, but also many other medications for blood pressure, lipid abnormalities, heart disease, and other problems.

Avoiding Insulin

A final aspect of convenience of combined oral agent therapy deserves comment, the convenience of avoiding insulin treatment. As mentioned above, many patients and physicians prefer not to use insulin if it can be avoided. Use of insulin may be frightening and may appear to be a punishment for poor lifestyle choices, and brings the risk of hypoglycemia. It remains to be seen whether combinations of three or even four oral agents will prove equally or better tolerated, and will lead to equal or better outcomes than earlier introduction of insulin. The availability of dry powder inhaled insulin may facilitate earlier insulin use in some patients for whom the barrier of injections is otherwise insurmountable.

Insulin and Oral Agents

The combination of insulin and oral agents can also offer convenience and therapeutic benefits for patients. The use of oral agents can reduce the dosage of insulin required to meet therapeutic goals. This may have the benefit of reducing the weight gain associated with insulin use, and patients taking insulin often perceive lower doses to be an advantage, as minimizing the number of injections and the size of an individual injection can reduce the discomfort and inconvenience of multiple or large injections. This advantage is most evident for patients who are very insulin resistant, requiring more than 100 units (lesser amounts for insulin pens) at a time, thus making a single injection impossible. For some patients who are on pure basal insulin treatment with little or no prandial regulation of glycemia, use of secretagogues or other agents with prandial control (e.g., incretin mimetics) also may reduce the number of insulin injections needed.

Non-Glycemic Effects of Combination Therapy

A final rationale for combination therapy relates to proven or potential benefits other than those resulting from better glycemic control. This concept has been most emphasized for metformin and the thiazolidinediones. Both of these classes of agents have potential non-glycemic effects that may reduce cardiovascular risk. In the UKPDS, metformin use by obese patients reduced cardiovascular events, such as myocardial infarction, relative to the rate seen with diet alone. Statistically significant benefits of this kind were lacking with insulin or sulfonylurea treatment (2,3). Similarly, the thiazolidinediones may have various non-glycemic effects, among them reducing markers for procoagulant and inflammatory states, and normalizing endothelium-dependent vasorelaxation and smooth muscle migration. Some studies have found lipid and blood pressure benefits as well. Ongoing randomized clinical trials which are testing whether thiazolidinedione therapy improves clinical outcomes are just beginning to report results and suggest that there may be some reduction in ischemic cardiovascular events, but raise concerns about congestive heart failure which may mitigate the overall cardiovascular benefit of this drug class (13).


Both insulin deficiency and insulin resistance are present in most patients with type 2 diabetes. As a result, most patients will need treatments that address both physiologic abnormalities and, therefore, combination pharmacotherapy. In some patients with marked insulin resistance, endogenous insulin production may be adequate once insulin resistance has been aggressively counteracted through both lifestyle change and dual pharmacotherapy using metformin and a thiazolidinedione for pre-prandial glycemic control, but post-prandial control may still require secretagogues.

Secretagogues and Insulin-Assisting Agents

Insulin secretagogues that are currently available include several sulfonylureas (first and second generation) and the fast acting, short-duration insulin secretagogues (repaglinide, nateglinide). Insulin-assisting agents that are available include the biguanide metformin and the thiazolidinediones, rosiglitazone and pioglitazone. Not included in either of these groups are the alpha-glucosidase inhibitors acarbose, miglitol and voglibose. Conceptually, they do not neatly address a known physiologic defect of diabetes as do the other classes of agents. By delaying carbohydrate absorption from the small intestine, through inhibition of the breakdown of disaccharides and polysaccharides, they reduce the amount of insulin required to combat meal-related hyperglycemia. Thus, they may be considered another type of insulin-assisting agent. The most widely used forms of oral combination therapy for diabetes pair an insulin secretagogue with an insulin-assisting agent or two insulin-assisting agents together especially in more obese insulin-resistant patients. Incretins are peptide hormones from the gut that enhance insulin secretion with food. Some have other therapeutic effects including slowing gastric motility, reducing glucagon levels and curbing appetite, thus acting as both secretagogue and insulin-assisting agent. These hormones include GLP-1 and gastrointestinal insulinotropic polypeptide (GIP). In type 2 diabetes GLP-1 is deficient, while GIP is present in normal levels, but defective in its tissue action. Incretin agonists are available as the GLP-1 receptor agonists, such as exenatide and liragutide, which deliver superphysiologic GLP-1 activity, and also GLP-1 enhancers, which include the DPP-IV enzyme inhibitors sitagliptin and vildagliptin, which raise the levels of endogenous GLP-1 and GIP, and thus restores physiologic activity by preventing their rapid proteolytic degradation. While both exenatide and the DPP-IV inhibitors have similar glycemic reduction effects, to date, exenatide is the only incretin agonist that has been associated with sustained weight loss, which likely reflects the superphysiologic activity of the drug. Incretin agonists have been shown to be safe in combination with metformin (14), sulfonylureas (15), both (16), and with thiazolidinediones.

Combining Insulin and Insulin-Assisting Agents

Later in the course of type 2 diabetes, when insulin deficiency is more marked, oral therapy alone fails to maintain control and insulin therapy is needed. Continuation of previously used oral agents while starting insulin is a form of combination therapy that has become increasingly common. This tactic allows insulin to be started with a simple regimen and titrated gradually, giving the patient time to learn the new procedures and gain confidence with insulin therapy. It also avoids the temporary loss of glycemic control that may occur when oral agents are discontinued and the dosage of insulin required is being established. Later, as a more complex insulin regimen combining basal and meal insulin becomes necessary, the benefit of ongoing oral-insulin combination therapy is less obvious, but use of insulin-assisting agents may continue to improve the results of treatment.

ORAL AGENT COMBINATION THERAPY Secretagogues with Biguanides

This combination has become very widely used in clinical practice and, for this reason, requires few comments. The first published data on oral agent combination in the United States were for glyburide and metformin (11), as illustrated in Figure 4. Drug dosing in this study (20 mg of glyburide and up to 2550 mg of metformin daily) probably exceeded clinically effective maximum doses for both the sulfonylurea and the metformin. Sulfonylureas have hyperbolic dose-response curves. Thus, doses for most patients need not exceed one-half of the approved maximal dose because this conveys most of the long-term glycemic benefit. Although metformin has been used for many decades, dose-response data have been published (17) only in the last decade. This study showed that maximal glucose lowering occurred at 2 g/day, suggesting the most appropriate full dosage regimen should be 1000 mg twice a day. Similar benefit can be gained with combined use of other sulfonylureas or insulin secretagogues with metformin (18). The author tends to favor use of once a day sulfonylureas, such as glimepiride and extended release glipizide, in combination with metformin, because of their convenient once daily dosing and reduced risk of hypoglycemia in comparison with glyburide.

Secretagogues with Thiazolidinediones

Combining insulin secretagogues with insulin-assisting agents such as thiazolidinediones also are effective. One large trial (19) found that adding troglitazone restored glycemic control in patients with secondary failure of glyburide, in a dose-dependent manner. Similarly, a more recent study (20) found that patients failing sulfonylurea therapy had improvement with the addition of 4 mg of rosiglitazone. In addition, this trial (20) showed clearly that titration up of more than half the maximally recommended dose of glipizide failed to improve hyperglycemia over the 6 months of this trial. It is perhaps important to note that the most positive response to the thiazolidinedione in such studies was in the subjects who were obese, with mean body mass index (BMI) of 30 kg/m2 or above. Subjects in combination trials appear to have a better absolute response when their baseline glycemic control is relatively poor (A1c > 9%). Thiazolidinediones have a relatively linear dose-response curve within the recommended dose range. This means that if > 2% reduction of HbA1c is needed, maximally approved doses will usually be required to approach the glycemic target. Side effects, such as edema or weight gain may limit their use in a few patients (20). Since it is hard to predict which patients will gain excess weight, tracking weight gain is important. Presumably, maximum doses of other thiazolidinediones will yield similar improvements to those demonstrated by troglitazone combined with sulfonylureas. In the case of rosiglitazone (21), evidence for this comes from a study in which 574 patients were randomized to continue sulfonylureas, or add submaximal doses of rosiglitazone (1 or 2 mg twice daily, compared to the maximal approved 4 mg twice daily dosage) for 26 weeks in a placebo-controlled trial. The higher dose of rosiglitazone reduced HbA1c by 1.0% and FPG by 44mg/dL (2.44mmol/L), while the lower dose reduced HbA1c by 0.6% and glucose by 24mg/dL (1.35mmol/L). Likewise, in a study of similar size (22), pioglitazone was given at less than the 45 mg maximal dosage(15 and 30 mg), and reduced HbA1c and FPG by 0.9% and 39 mg/dL (2.17mmol/L) and 1.3% and 58 mg/dL (3.2mmol/L) in a randomized comparison with placebo. Although studies directly comparing the effects of these agents are few, these findings suggest they have similar therapeutic power when combined at full dosage with sulfonylureas. Short-acting secretagogues, such as repaglinide, can also be used in combination with thiazolidinediones. In a 22-week randomized study (23) of troglitazone (up to 600 mg) and repaglinide (up to 4 mg pre-prandially), combination therapy had a synergistic affect, reducing HbA1c by 1.7% and fasting serum glucose by 80 mg/dL in comparison to monotherapy with repaglinide alone (0.8% and 43 mg/dL) or troglitazone (0.4% and 46 mg/dL). Similar benefits have been seen with repaglinide and pioglitazone or rosiglitazone combination trials.

Secretagogues with Alpha-Glucosidase Inhibitors

Although alpha-glucosidase inhibitors are commonly used in Europe and Japan, they are less often used in the United States. Addition of an alpha-glucosidase inhibitor to an insulin secretagogue may reduce HbA1c by 0.5% to 1%. For example, in a 28-week trial of acarbose added to nutrition therapy or sulfonylurea-treated subjects with inadequate control, the mean HbA1c reduction was 0.66% compared with placebo (24). Much of this effect was due to reduction of postprandial hyperglycemia. The mean 1 hr PPG level declined by 41 mg/dL (2.3mmol/L) when acarbose was added.

Combinations of Insulin-Assisting Agents

Insulin resistance may occur at multiple sites and several kinds of insulin-assisting agents exist to address these various defects, making several combinations of these agents a plausible therapeutic option. However, the glycemic effect is generally less robust than that seen in studies of combined secretagogue and sensitizer therapy. In one early study (25), 3 months of treatment with metformin reduced fasting and postprandial plasma glucose levels by 20% (58mg/dL or 3.2mmol/L) and 25% (87mg/dL or 4.8mmol/L). The same duration of troglitazone treatment exerted similar monotherapeutic benefit, with a reduction in fasting and postprandial plasma glucose of 20% (54 mg/dL or 2.9mmol/L) and 25% (83 mg/dL or 4.6 mmol/L). The combination of these therapies resulted in a further reduction of fasting and postprandial glucose of 18% (41 mg/dL or 2.3mmol/L) and 21% (54 mg/dL or 3.0mmol/L) and a reduction of mean HbAic of 1.2%. Another study showed that addition of 30 mg of pioglitazone to metformin reduced HbA1c by 0.83 over 16 weeks, with further improvement in an open label extension of the trial that permitted higher doses of pioglitazone (26). A third study showed that addition of full-dose (8mg) rosiglitazone to metformin led to 1.2% reduction of HbA1c over 26 weeks. These studies suggest that combining a thiazolidinedione with metformin is useful for some patients (27), especially those who are very obese and have marked insulin resistance with significant endogenous insulin remaining.

Combinations of Three Oral Agents

Relatively few reports of triple oral agent therapy exist. A retrospective study (28) examined the addition of troglitazone 600 mg daily for patients inadequately controlled on metformin and the sulfonylurea glimepiride. In this study, significant declines in HbA1c occurred at 2 and 6 months (1.6% and 2.5%). In another small, non-randomized, prospective study (29) of patients offered troglitazone 400 mg for 3 months in addition to metformin and a sulfonylurea, about 62% of the patients achieved > 1% decline and of these 68% reached the minimal HbA1c goal of 8%. Increasing the dose to 600 mg and extending the observation to 6 months did little to improve glycemia further. One randomized placebo controlled trial (30) of triple agent oral therapy that has been published found that adding 400 mg of troglitazone for 6 months to patients with poor glycemic control (HbA1c 9.7%), already on a sulfonylurea and metformin, resulted in a mean 1.4% reduction of HbA1c. While this was far superior to placebo, only 43% of patients in this trial reached the minimally acceptable glycemic target of < 8% HbA1c. The question should be raised whether use of triple oral agent therapy makes sense from a cost-effectiveness standpoint when compared with injected insulin, the main therapeutic alternative. One study (31) addressed the efficacy of added thiazolidinedione versus basal insulin. In a 24-week trial, 217 patients with HbA1c ranging from 7.5% to 11% already on effective doses of sulfonylurea and metformin received insulin glargine at 10units/day that was titrated to target an FPG of < 5.5 to 6.7mmol/L (< 100-120 mg/dL). In a parallel arm rosiglitazone was started at 4 mg once daily and then rosiglitazone was increased to 8 mg/day any time after 6 weeks if FPG was > 5.5mmol/L (100g/dL). In this study reduction in HbA1c was similar for the two arms (-1.7% vs. -1.5% for insulin glargine vs. rosiglitazone, respectively). Nonetheless, when baseline HbA1c was > 9.5%, the reduction of HbA1c with insulin glargine was greater than with rosiglitazone (P < 0.05). Additionally, insulin glargine was associated with slightly more hypoglycemia but less weight gain, no edema, and beneficial lipid changes at a lower cost of therapy. The studies described above provide support for trying a thiazolidinedione for a few months in patients who are not successful with metformin plus a secretagogue, and have HbA1c levels that are less than 9%, or those for whom using insulin is problematic.

Incretin Therapy Additions to Failing Oral Agents

The recent availability of incretin mimetics, either GLP-1 receptor agonists such as exenatide or DPP-IV inhibitors, has opened up new possibilities in combination therapy. Exenatide is an injectable synthetic analog of the Gila monster (Heloderma suspectum) salivary protein exendin-4. This compound has substantial homology with GLP-1 and tightly binds to GLP-1 receptors and thereby mimics the actions of native GLP-1 when given in doses of 5 or 10 mcg twice daily.

Three published trials show the use of exenatide in 30-week long studies with patients with oral agent failure with either sulfonylureas (15), metformin (14) or both (16). The design of the studies was similar. After a 4-week placebo injection run in phase, subjects were randomized to blinded administration of placebo versus exenatide 5 mcg twice daily for 1 month and then continued this dose or followed with 10 mcg twice daily. All subjects continued use of prior oral agents. When exenatide versus placebo was added to metformin, 272 patients completed the study (14). They were middle-aged (53 - 10 years), obese (34.2 - 5.9 BMI) and with inadequate glycemic control (HbA1c 8.2 -1.1%). At 30 weeks, the HbA1c change from baseline was -0.78-0.1% (10mcg), -0.4-0.11% (5mcg) and 0.08-0.1% for placebo; P < 0.002. Exenatide was associated with weight loss: -2.8 - 0.5 kg (10 mcg), -1.6 - 0.4 kg (5 mcg); P < 0.001 versus placebo. Gastrointestinal side effects including nausea, vomiting and diarrhea were more common with exenatide but lessened toward the end of the trial. In the 5 and 10 mcg groups, exenatide resulted in a placebo subtracted percentage for nausea of 11% and 22%, for vomiting 7% and 8%, and for diarrhea 4% and 8 % overall during the study. In the sulfonylurea failure study (15), the study population was similar with obese, middle-aged subjects with slightly higher baseline glycemia (HbA1c 8.6 — 1.2%). The change from baseline HbA1c at 30 weeks was -0.86 - 0.11 (10 mcg), -0.46 - 0.12? (5 mcg) and 0.12 -0.09% (placebo); P < 0.001. Weight loss was somewhat less with 10 mcg than in the metformin alone study (-1.6 kg). The third trial was for patients inadequately controlled on the combination of effective doses of sulfonylurea and metformin. Similar subjects were studied with middle-aged obese poorly controlled subjects (baseline HbAic 8.5 —1.0%). The change from baseline HbA1c occurred at 30 weeks was -0.8 + 0.1% (10 mcg), -0.6 + 0.1% (5 mcg) and + 0.2 + 0.1% (placebo); P < 0.0001. Weight loss in this study averaged 1.6 kg for the 10 mcg dose group and was like the sulfonylurea alone study. A similar study has been conducted but not yet published showing comparable glycemic benefit in patients in a thiazolidinedione alone to which 5 and 10 mcg doses of exenatide were added for about 11/2 year.

Is it reasonable to choose exenatide as an alternative to basal insulin therapy? Perhaps if patients are not very far from glycemic goal. Heine et al. (32) reported a study of 551 type 2 diabetes subjects who were inadequately controlled. They were randomized to either insulin glargine once a day at bedtime versus 5 mcg for 1 month then 10 mcg of exenatide for the duration of this 26-week long trial. The results showed that baseline HbA1c was 8.2% for patients receiving exenatide and 8.3% for those receiving insulin glargine. By study end exenatide and insulin glargine therapies resulted in identical reduction of HbA1c levels by 1.11%. Exenatide reduced postprandial plasma glucose levels more than insulin glargine, while insulin glargine reduced FPG levels more than exenatide. This is particularly well illustrated in the 7-point self-monitored glucose levels before and after meals and at 3 am performed at study beginning and end (Fig. 6). Body weight decreased 2.3 kg with exenatide and increased 1.8 kg with insulin glargine. Rates of symptomatic hypoglycemia were similar, but nocturnal hypoglycemia occurred less frequently with exenatide (0.9 event/patient-year versus 2.4 events/patient-year). Gastrointestinal symptoms were more common in the exenatide group than in the insulin glargine group, including nausea (57.1% vs. 8.6%), vomiting (17.4% vs. 3.7%) and diarrhea (8.5% vs. 3.0%). The nearly identical lowering of average glycemia is noteworthy in comparison to the marked difference in prandial versus pre-prandial control, suggesting these interventions had different patterns of benefit.

In all of the studies of exenatide in which sulfonylureas were used, an increased risk of hypoglycemia occurred that sometimes required a reduction in sulfonylurea dose to reduce the risk of hypoglycemia symptoms. In patients on sulfonylureas treated with exenatide (and perhaps also with DPP-IV inhibitors), it may be appropriate to preemptively reduce sulfonylurea doses substantially if patients' lowest blood sugars are less than 100 mg/dL since the glucose-dependent insulin secretion with this combination is lost as a result of the sulfonylurea. Taken together, these studies suggest that exenatide may represent a desirable alternative for overweight patients for whom lifestyle intervention alone is insufficient in improving weight and who also need improved glycemia control but are reluctant to use insulin.

DPP-IV Inhibitors

Incretin action can also be provided by inhibiting the rapid degradation of GLP-1 and GIP. Although the levels of GLP-1 probably do not rise to a degree similar to that seen with receptor agonists such as exenatide and liraglutide, nonetheless DPP-IV inhibitors such as sitagliptin and vildagliptin are close to the glycemic lowering efficacy seen with receptor agonists. Vildagliptin (33) 50 mg once daily added to patients inadequately controlled on metformin (baseline HbA1c 7.7%) resulted in a placebo subtracted difference in HbA1c at 52 weeks of -1.0 — 0.2%; P < 0.001. DPP-IV inhibitors do not cause nausea and vomiting and 1 year studies lack the weight reduction effect (they are weight neutral) seen with the GLP-1 agonists. This is likely due to lower levels of GLP-1 activity. Sitagliptin (34) has recently been FDA approved for use as monotherapy and as additional treatment for those not meeting glycemic goals on either metformin or a thiazolidinedione. In doses of 100 mg for patients with normal renal function in two monotherapy studies of 18 and 24 weeks duration, sitagliptin reduced HbA1c by 0.6% and 0.8%. Added to either metformin or pioglitazone, it further reduced HbA1c by 0.7% in 24-week duration studies. DPP-IV inhibitors are weight neutral, probably due to an appetite effect of raising endogenous GLP-1 levels. Their glycemic effects on peak prandial control appear superior to effects on pre-prandial control. This should complement the primarily pre-prandial effects of metformin or thiazolidinediones.

Incretin drugs appear especially favorable for prandial glycemic control and may be favored also because of positive effects of weight loss or minimal weight gain. Prandial control has been observed by Monnier et al. (35,36) to be important particularly as HbA1c nears goal (Fig. 7). Also if prandial euglycemia contributes to decreased cardiovascular risk through reducing oxidative stress then incretins could have an additional favorable action.


Insulin therapy is eventually needed for most patients with type 2 diabetes. An evening insulin strategy is a simple way to begin insulin therapy that will achieve glycemic goals and is easily understood by patients. The rationale for evening insulin has previously been reviewed (37-39). In brief, an evening injection of intermediate or long-acting insulin addresses a fundamental need in management of type 2 diabetes by suppressing overnight endogenous glucose production and thereby preventing hyperglycemia prior to the first meal of the day. This approach is useful for most patients with type 2 diabetes, with the notable exception of patients taking morning glucocorticoid therapy.

There are several versions of this strategy, including the use of intermediate-acting insulin at bedtime, intermediate and quick-acting insulin mixed in a single injection at dinnertime, and insulin glargine or insulin detemir at bedtime. For some patients an alternate timing of glargine may be used earlier in the day; for those using large doses (0.8 units/kg) of detemir this may also be possible. Most of the evidence for these regimens comes from trials of insulin combined with a sulfonylurea alone or with metformin, with the oral agents continued while the insulin dosage is gradually increased until control is re-established.

NPH at Bedtime

Addition of an injection of NPH insulin within an hour of bedtime is usually able to restore adequate glycemic control for patients who are no longer well-controlled with one or more oral agents alone. This tactic may be best employed in leaner patients (BMI < 29), who seem to need short-acting insulin at suppertime less often. A multicenter trial has shown bedtime NPH insulin plus daytime oral agents achieves glycemic control as good as insulin taken in the morning with oral agents, or mixed intermediate and regular insulin twice daily without oral agents. However, there is less weight gain with evening NPH (40). Other trials show better glycemic control with bedtime NPH plus a sulfonylurea than with a single injection of insulin alone (41,42). Evening NPH insulin also reduces free fatty acids to a greater degree than use of daytime insulin (43).

Patients can begin with a low dose of NPH insulin, usually about 10 units. They are instructed to self titrate the dose up by 2 to 4 units every 3 to 7 days based upon the stability of their fasting glycemic response. Stable patients may titrate more quickly, based upon the pattern of response, but there is little reason to hurry because glucose control will steadily improve at any rate of titration so long as the oral agents are continued. The insulin dosage required is frequently in the range of 30 to 50 units daily, or about 0.4 to 0.5 units/kg of body weight. The target for fasting glucose should be individualized and adjusted when hypoglycemia occurs, but often can be the ADA recommended 90-130 mg/dL (5-7.2 mmol/L) value in plasma-referenced home glucose-monitoring systems. Patients need to wake at a reasonably consistent time and eat breakfast consistently. Oral agents are continued, although sulfonylureas are usually given only with the first meal of the day. A randomized trial has examined the choice of daytime therapy accompanying bedtime NPH insulin (44). This study compared four different regimens in a prospective 1-year randomized controlled trial. The four regimens included bedtime insulin combined with morning insulin, glyburide alone, metformin alone, or glyburide combined with metformin. The least weight gain occurred when metformin was the only oral agent, and hypoglycemia was a limiting factor when glyburide was used.

Pre-Mixed Insulin with the Evening Meal

A second form of evening insulin that appears to work better for more obese patients (BMI > 30) is the combination of morning sulfonylureas and suppertime mixed insulin, the latter commonly offered as 70/30 (70% NPH with 30% Regular) insulin. In a multicenter study using the long-acting sulfonylurea glimepiride (45), illustrated in Figure 5, patients achieved a more rapid restoration of glycemic control with self-titration of 70/30 insulin while continuing the oral agent, rather than with insulin alone. Insulin was started at 10 units and titrated weekly, seeking FPG equivalent to 140mg/dL (7.8mmol/L, plasma-referenced). Nearly all subjects using the combination regimen reached the titration target rapidly, but 15% of the subjects in the placebo plus insulin group dropped out, mainly due to hyperglycemia during the transition to insulin. The mean HbAic declined from almost 10% to 7.6% for subjects completing the trial in both groups. The mean dose in the insulin alone group was 78 units and, for the glimepiride plus insulin combination, was 49. More subjects on insulin alone needed doses higher than 100 units daily, and so had to take more than one injection. A smaller study with a more aggressive titration scheme found better glycemic control using 70/30 insulin with the evening meal plus glyburide once daily, than with evening insulin alone (46). Premixed rapid analog mixes [e.g., lispro/neutral protamine lispro (25%/75%) and aspart/ neutral protamine aspart (30%/70%)] similarly may achieve control with somewhat more convenient meal timing of insulin. Use of pens with meals is often desirable especially for those who eat outside the home frequently. These are commonly given twice a day, sometimes more or less frequently. Garber et al. (47) have reported a small observational study using either once, twice or three times daily administration at meal time of the 70/30 aspart mixture in patients inadequately controlled or oral agents with or without basal insulin treatment. In this study, once daily administration at dinner of 70/30 reduced HbA1c by 1.4%, twice daily at breakfast and supper by 1.9% and thrice daily with an added lunch dose by 1.8%.

Insulin Glargine at Bedtime

A study using insulin glargine (48) suggests this agent may offer another option for starting insulin with an evening injection. In this 1-year European study, 426 subjects were randomly assigned to either insulin glargine or NPH insulin at bedtime, while continuing previous oral therapy. The therapeutic target was a fasting blood glucose < 120 mg/dL (6.7mmol/L), using a method that was probably not plasma-referenced. The insulin dosages used (23 units for

FIGURE 5 A transition strategy of adding insulin for obese patients failing oral agents using morning secretagogue plus a suppertime mixed insulin preparation (70/30 insulin combining NPH and Regular insulin). The titration of insulin dosing based on the fasting glucose achieves more rapid control with combination therapy and does so at lower insulin doses.

FIGURE 5 A transition strategy of adding insulin for obese patients failing oral agents using morning secretagogue plus a suppertime mixed insulin preparation (70/30 insulin combining NPH and Regular insulin). The titration of insulin dosing based on the fasting glucose achieves more rapid control with combination therapy and does so at lower insulin doses.

Exenatide Insu tin glargine

Ë |jg BID 1st 4 weeks, then 10 pg SID 10 U/d, titraled to target FPG <100 mgftlL

101 Everyday Tips for Losing 10 Pounds

101 Everyday Tips for Losing 10 Pounds

Studies show obesity may soon overtake tobacco as the leading cause of death in world. Who Else Could Use 101 'Everyday' Ways to Lose 10 Pounds or more and Keep it Off! You've been putting it off too long. Hey, everyone needs to lose weight from time to time. You're no different!

Get My Free Ebook

Post a comment