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.

There Is Benefit to Increasing the Sulfonylurea Dose

Mayer B. Davidson, MD, Lauren C Ficks, MD, and Anne L. Peters, MI

The recommended maximum dose for (nonmicronized) glyburide is 20 mg/day.¹ However, few data support this recommendation. Indeed, acute studies in normal subjects strongly suggest that the maximal effects of the drug on insulin secretion and action would occur with doses <10 mg/day.² No further beneficial effect (with doses >10 mg/day) was noted in nine patients with type II diabetes in whom glyburide was increased progressively at weekly intervals.³

It has been stated that most patients require 5-10 mg of glyburide per day, with few benefiting from doses exceeding 15 mg/day.⁴ In a recent study, patients with type II diabetes were treated with placebo or 10, 20, or 40 mg of glipizide for 3 months.⁵ The average response to all three doses of glipizide was nearly identical (notwithstanding that 40 mg is considered the maximal dose in the United States)¹ and significantly better than placebo.⁵

Considering these results and the (admittedly scanty) available dose-response data for glyburide, we hypothesized that few patients would respond to doses of glyburide >10 mg/day. To evaluate that hypothesis, we retrospectively reviewed the responses to increasing doses of glyburide in patients with type II diabetes who are managed in our Diabetes Managed Care Program (DMCP).

METHODS
The DMCP at Cedars-Sinai Medical Center was established in 1987 to manage the diabetes of a large number of patients enrolled in pre-paid health plans affiliated with the hospital. The management of these patients is carried out by nurse clinicians who follow clinical protocols established and supervised by two diabetologists.

Sulfonylurea agents were increased progressively depending on the results of a fasting plasma glucose (FPG) concentration measured in a clinical laboratory 2 weeks after the dose was increased. FPG concentrations are also routinely measured every 2 months in every patient. If the FPG concentration was >140 mg/dl (7.8 mM) in patients receiving less than the maximal dose of a sulfonylurea agent, the present dose was increased, and the patient was notified by phone to go to the laboratory in 2 weeks for another FPG measurement.

No diet or exercise advice was given at that time. Nutritional counseling and exercise advice were given to patients when they enrolled in the DMCP. Therefore, subsequent changes in FPG concentrations were due to the dose increase rather than to any lifestyle alterations.

These protocols are based on the following reasoning. Since drug levels reach a plateau after five half-lives, the effect of a new dose will be evident by 2 weeks after the change. The FPG concentration is relatively stable in patients with type II diabetes.⁶ In our experience, the stability of that value is affected little by a fixed dose of a sulfonylurea agent. Therefore, the FPG concentration 2 weeks after a dose change will reflect the effect of that adjustment. (Instead of a laboratory measurement, some diabetologists use the results of fasting blood glucose [FBG] concentrations measured by patients at home if they believe those values to be accurate.)

The goal is to lower the FPG concentration to <140 mg/dl (or the FBG concentration to <120 mg/dl). At that point, the glycated hemoglobin level 2-3 months hence drives the therapeutic decision. The goal here is a value <1.5% above the upper limit of normal for the assay used. (In our laboratory, the normal range for glycated hemoglobin levels performed by affinity chromatography is 4.2-6.8%.) In this manner, patients do not remain on suboptimal doses for long periods during the lag phase before the glycated hemoglobin level plateaus at its new equilibrium value.

Study Design
All patients who were taking sulfonylurea agents and were eligible for this study A) had failed several months of treatment with diet and exercise alone; B) had a degree of symptoms of uncontrolled diabetes and a glucose concentration that in our clinical judgment required a low dose of drug initially; or C) were taking the drug when referred to the DMCP.

The charts of 870 patients were reviewed, 196 of whom were taking <20 mg of glyburide alone (i.e., they were not taking insulin, 20 mg of glyburide, or another sulfonylurea agent). Seventy-four patients were excluded from further analysis because the glyburide dose was decreased (FPG <100 mg/dl [5.5 mM]), not increased (FPG 100-140 mg/dl [5.5-7.8 mM]) or, if increased, the subsequent FPG concentration was not measured within 1 month. This left 122 patients with dose increases of glyburide and measurement of FPG concentrations within 1 month.

Dose Change Groups
The dose increases were divided into three groups as follows. Group A (not previously on a sulfonylurea agent) contained either asymptomatic patients who had failed diet or mildly symptomatic patients not given a trial of diet alone in whom initial doses of the drug ranged from 1.25-10 mg. Group B contained those dose increases from a smaller amount (2.5 mg was the lowest) to any higher dose up to and including 10 mg. Two patients in group B had more than two dose increases and, to simplify the analysis, were excluded. Group C contained all dose increases between 10 mg and 20 mg. Nineteen patients were in group A and there was obviously only one dose increase for each. Of the 58 patients in group B, 49 had one dose increase and 9 had two dose increases. Group C consisted of 73 patients, 46 with one dose increase and 27 with two dose increases. In addition to repeat dose increases within each group, some patients were also included in more than one group. Eight patients in group A were also in group B, and two in group A were also in group C. Eighteen patients in group B were also in group C, and one patient was in all three groups. Therefore, 120 individual patients were used in this evaluation.

Patients
The mean age (+SEM) of the 120 patients (the 2 with three or more changes in group B were excluded from the study) was 55.1 + 1.1 years (range 30-81 years). The mean body mass index (BMI) was 31.4 + 0.6 kg/m² (range 19.7 + 52.2 kg/m²). Twentyeight patients had newly diagnosed diabetes (< 1 month), some of whom had been started on small doses of glyburide before referral to the DMCP.

The mean duration of diabetes in the remaining 92 patients was 5.2 + 0.6 years (range <1-29).

Definition of Response
A response to a dose increase was defined as a >10% decrease in the FPG concentration (measured within 1 month). The effects of age (<65 vs. >65 years), duration of diabetes (<5 vs. >5 years), BMI (<27 vs. >27), and FPG concentration (<200 mg/dl [11.1 mM] vs. >200 mg/dl [11.1 mM]) on the responses were evaluated.

Statistical Analyses
All statistical analyses⁷ were performed by Fisher's exact test except for the differences in FPG concentrations before and after a dose change, which was analyzed by a paired t-test. A chi-square analysis was performed to evaluate whether the responses to dose increases of glyburide were a chance variation. Significance was accepted at the 0.05 level (two-tailed) for both tests. Data are expressed as means + SEM.

RESULTS
Fifteen of the 19 patients in group A responded. The average response in these 15 patients was 37.5 + 5.3%. The results in groups B and C are shown in Table 1. "All increases" in group B included 49 patients who had only one dose increase and 9 patients who had two dose changes. The "first increase" group includes the patients with only a single dose increase plus the first (i.e., smallest starting dose) increase in those patients who had two dose changes. The "second increase" group includes the patients with only a single dose increase plus the second dose increase in those patients who had two dose changes. (In these two cohorts of group B, the number of patients equals the number of dose increases.)

The second increase group was compared with the first increase group for the following reason. It seemed possible that if patients failed to respond to one dose increase, they would be less likely to respond to a second one. Since the second one would likely be at the higher doses, this enrichment of patients with two dose increases (27 of 73 in group C vs. 9 of 58 in group B) would lead to the conclusion that increases above 10 mg of glyburide are less effective.

In group B, the proportion of positive responses was similar for the all increases, first increase, and second increase groups. The results in group C are depicted in a similar manner. All of the second dose increases were from 15 mg to 20 mg. As in group B, the proportion of responses in group C was also similar among the three subgroups. Fisher's exact test revealed no differences between the response rates in groups B and C. The similar proportion of responders in the first and second increase groups indicates that a patient failing to respond to an increased amount of a small dose of glyburide may still respond to a dose increase at a higher dose of the drug.

Since ~50% of subjects responded in both groups B and C (Table 1), it might also be argued that changes in FPG concentrations were simply chance variation. If this were the case, one would expect that the number of patients in whom FPG levels fell would be similar to the number that rose. However, note that responders were defined as a >10% fall in FPG concentration, so that the nonresponders include some patients whose FPG concentrations fell by between 0 and 9%. In fact, the number of instances in which dose increases resulted in a fall in FPG concentration was significantly greater than the number of instances in which a dose increase resulted in an increase in FPG concentration in both groups B and C. Thus, in group B, 47 dose increases resulted in a fall in FPG with concentration, compared with 20 in which there was a rise (P < 0.025 compared with a 50% distribution). In group C, 67 dose increases led to a decrease in FPG concentration, compared with 37 in which the value increased (P < 0.05 compared with a 50% distribution). Therefore, the responses to dose increases of glyburide were not chance variations in FPG concentrations.

The average response in the 28 responders after the first and second dose increases in group B was -22.6 + 2.1% and -21.5 + 2.1%, respectively. Similarly, in group C, the average response in the 38 responders after the first dose increase and in the 35 responders after the second dose increase in group C was -26.6 + 2.5% and -23.6 + 2.4%, respectively. Thus, in the -50% of patients who responded to increasing doses of glyburide, the percentage fall of FPG concentrations was similar across all baseline doses of glyburide between 2.5 mg and 15.0 mg.

Apparent exceptions were diet failures and mildly symptomatic patients with type II diabetes who were started on glyburide for the first time. More of these patients in group A responded (78% [15/19] vs. ~50% in groups B and C), and their response was somewhat greater (-37.5 + 5.3% vs. -22 to 27%).

No patients in groups B and C complained of hypoglycemia after a dose increase. Likewise, no patient in group A experienced hypoglycemia, because their FPG concentrations were measured within 1 month of starting glyburide, and the dose was decreased if necessary (FPG <100 mg/dl [5.5 mM]).

The response to increasing doses of glyburide was unrelated in both groups B and C to age, BMI, duration of diabetes, or FPG concentration before the dose increase, whether evaluated for all increases or after first or second dose increases (data not shown). As pointed out above, we do not feel that changes in diet or exercise were important factors because patients were simply told via telephone to increase their doses and had an FPG concentration measured 2-4 weeks later. There were no instructions given for changes in diet or exercise patterns.

DISCUSSION
Increasing the dose of glyburide lowered the FPG concentration >10% ~50% of the time in patients already on the drug, with a greater response in those just starting the medication. It seems unlikely that dietary factors played much of a role in the better initial response to the drug (group A) compared with further increases (groups B and C), because FPG concentrations in group A were often measured before the patients could be scheduled to see our busy dietitian.

It should also be pointed out that although up to 50% of patients responded in groups B and C, our definition of response (>10%) does not necessarily mean that patients achieved good control. As mentioned above, our criterion for good control is a glycated hemoglobin value <1.5% above the upper limit of normal for our assay. Since responses to dose increases were evaluated 2-4 weeks later, changes in glycated hemoglobin levels could not be used in this analysis. However, 70% of our type II patients on sulfonylurea agents alone are able to achieve good control as defined above as indicated by glycated hemoglobin levels measured routinely every 2 months (data not shown).

A recent report evaluated the maximal dose of glipizide, another second-generation sulfonylurea agent.s In that double-blind, crossover study, diabetes medication in 23 patients with type II diabetes was stopped, and a placebo was given for 2 weeks. Following the placebo period, patients took 10, 20, or 40 mg/day of glipizide for 3 months at each dose. The glycemic end point evaluated was the mean of blood glucose measurements performed at home before breakfast, lunch, and supper and at bedtime 2 days/week. Values at each dose of glipizide were significantly less than placebo but similar to each other.⁵ The authors concluded that increasing the dose of glipizide above 10 mg daily produced little benefit.

We would argue that considering only the mean values does not identify the instances in which patients might respond to higher doses of sulfonylurea agents. The negative responses, some of which might include increases in glucose levels, would obscure the positive responses.

In our study, FPG concentrations for all dose increases in groups B and C changed minimally because the negative responses counterbalanced the positive responses (Figure 1). In fact, the absolute change of FPG concentrations in the nonresponders was a small increase. In spite of a much greater absolute fall in FPG concentrations in the responders, the changes in nonresponders were enough to cause either a minimal change or no change when all responses to a dose increase were considered together (Figure 1).

Figure 1. Change in FPG concentration within 1 month after increasing glyburide dose. Group B: 67 dose increases between 2.5 mg and 10 mg. Group C: 104 dose increases between 10 mg and 20 mg.

It might also be argued that using the same patients in groups B and C might select for those in group C more likely not to respond, because it is these patients in whom doses would be progressively increased. If this were true, it might sway the data in the direction of fewer positive responses in group C and favor the hypothesis that increases in doses of glyburide above 10 mg would be ineffectual. The results in Table 1 indicate that this was not the case, because the proportion of positive responses was similar in groups B and C.

In summary, up to 50% of patients with type II diabetes responded (defined as a >10% decrease in FPG concentrations) to dose increases of glyburide from 2.5 mg to 10 mg and from 10 mg to 20 mg. The proportion of patients responding at the lower and higher doses was similar. The average decrement in FPG levels in the groups who did respond was 20-25%. Patient age, BMI, duration of diabetes, or FPG concentrations before the dose increase did not predict a response.

Finally, it should be noted that the best response (~80% of the patients with an average decrease of FPG levels of nearly 40%) occurred in the small group of asymptomatic patients who failed diet or mildly symptomatic patients not given a trial of diet alone, all of whom were started on a low dos (<10 mg) of glyburide. This suggests that the initial response to glyburide may be better than responses following subsequent dose increases.

Our approach is to continue to increase the dose of glyburide up to 20 mg/day. If at any time the FPG concentration falls below 140 mg/dl, the dose is maintained, and an FPG concentration and a glycated hemoglobin level 2-3 months later are used for further dose adjustments, if necessary. The goal is to achieve a glycated hemoglobin level <1.5% above the upper limit of normal for the assay used. Whenever the maximal dose of 20 mg/day o glyburide is reached and the glycated hemoglobin level remains >1.5% abort the upper limit of normal for our assa' we strongly consider adding metformi: (if there are no contraindications) or bedtime NPH insulin.

Acknowledgments
We gratefully acknowledge the excellent clinical skills of Liz Perry, RN, CDE; Mary Rose Deraco, RN, CDE; and Cindy Dunlap, RN, CDE, and the secretarial skills of Beth Avecilla.

References

¹Physicians' Desk Reference, 48th edition',
Montvale, N.J., Medical Economics Data Production Company, 1994.

²Groop LC, Barzilai N. Ratheiser K, Luzi L, Wahlin-Boll E, Melander A, DeFronzo RA: Dosedependent effects of glyburide on insulin secretion and glucose uptake in humans. Diabetes Care 14: 724-27, 1991.

³Groop L, Groop P-H, Stenman S. Saloranta C, Totterman K-J, Fyhrquist F. Melander A:Comparison of pharmacokinetics, metabolic effects, and mechanisms of action of glyburide and glipizide during long-term treatment. Diabetes Care 10:671-78, 1987.

⁴Editorial. Glibenclamide: a review. Drugs 2:11640, 1971.

⁵Stenman S. Lemander A, Groop P-H, Groop LC: What is the benefit of increasing the sulfonylurea dose? Ann Int Med 118:169-72, 1993.

⁶Kosaka K, Mizono Y. Kuzuya T: Reproducibility of the oral glucose tolerance test and the rice-meal test in mild diabetics. Diabetes 15:901-904, 1966.

7Matthews DE, Farell VT: Using and Understanding Medical Statistics. Basel, Switzerland, S. Karger, 1985.

Mayer B. Davidson, MD, is the associate director of clinical diabetes at the City of Hope National Medical Center, in Duarte, Calif. Lauren G. Ficks, MD, is an internist/endocrinologist at HealthCare Partners Medical Group, in Duarte. Anne L. Peters, MD, is an associate professor of medicine at the UCLA School of Medicine, in Los Angeles.

Return To American Diabetes Association Home Page

Copyright © 1997 American Diabetes Association

Last updated: 6/15/97
For Technical Issues contact webmaster@diabetes.org