Saxagliptin added to a submaximal dose of sulphonylurea improves glycaemic control compared with uptitration of sulphonylurea in patients with type 2 diabetes: a randomised controlled trial

Aims: Assess the efficacy and safety of saxagliptin added to a submaximal sulphonylurea dose vs. uptitration of sulphonylurea monotherapy in patients with type 2 diabetes and inadequate glycaemic control with sulphonylurea monotherapy. Methods and patients: A total of 768 patients (18–77 years; HbA1c screening ≥ 7.5 to ≤ 10.0%) were randomised and treated with saxagliptin 2.5 or 5 mg in combination with glyburide 7.5 mg vs. glyburide 10 mg for 24 weeks. Blinded uptitration glyburide was allowed in the glyburide-only arm to a maximum total daily dose of 15 mg. Efficacy analyses were performed using ANCOVA and last-observation-carried-forward methodology. Results: At week 24, 92% of glyburide-only patients were uptitrated to a total glyburide dose of 15 mg/day. Saxagliptin 2.5 and 5 mg provided statistically significant adjusted mean decreases from baseline to week 24 vs. uptitrated glyburide, respectively, in HbA1c (−0.54%, −0.64% vs. +0.08%; both p < 0.0001) and fasting plasma glucose (−7, −10 vs. +1 mg/dl; p = 0.0218 and p = 0.002). The proportion of patients achieving an HbA1c < 7% was greater for saxagliptin 2.5 and 5 mg vs. uptitrated glyburide (22.4% and 22.8% vs. 9.1%; both p < 0.0001). Postprandial glucose area under the curve was reduced for saxagliptin 2.5 and 5 mg vs. uptitrated glyburide (−4296 and −5000 vs. +1196 mg·min/dl; both p < 0.0001). Adverse event occurrence was similar across all groups. Reported hypoglycaemic events were not statistically significantly different for saxagliptin 2.5 (13.3%) and 5 mg (14.6%) vs. uptitrated glyburide (10.1%). Conclusion: Saxagliptin added to submaximal glyburide therapy led to statistically significant improvements vs. uptitration of glyburide alone across key glycaemic parameters and was generally well tolerated.


Introduction
As a result of the progressive decline in functional insulin-producing beta cells, the majority of patients with type 2 diabetes eventually require combination drug therapy to achieve and maintain glycaemic targets (1)(2)(3). The availability of newer agents may allow for different therapeutic combinations to improve glycaemic control without increasing the risk of adverse effects (4,5).
Saxagliptin is a potent, selective dipeptidyl peptidase-4 (DPP-4) inhibitor, specifically designed for extended inhibition of the DPP-4 enzyme (6,7). DPP-4 inhibitors enhance the levels of the glucoregu-latory hormones glucagon-like peptide-1 (GLP-1) and glucose-dependent insulinotropic peptide (GIP), thereby acting to promote insulin synthesis and release, and suppress glucagon secretion, among other important glucoregulatory effects (8). DPP-4 inhibitors are associated with a favourable safety profile, including a low risk of hypoglycaemia because of the glucose-dependent nature of incretin hormone activity, a neutral effect on body weight and the potential for improved beta-cell function (8,9). Proof of concept for saxagliptin was previously established in a 12-week dose-ranging trial (dose range: 2.5-40 mg) (6,7). In a phase 3 clinical trial, saxagliptin administered as initial therapy with metformin S U M M A R Y Aims: Assess the efficacy and safety of saxagliptin added to a submaximal sulphonylurea dose vs. uptitration of sulphonylurea monotherapy in patients with type 2 diabetes and inadequate glycaemic control with sulphonylurea monotherapy. Methods and patients: A total of 768 patients (18-77 years; HbA 1c screening ‡ 7.5 to £ 10.0%) were randomised and treated with saxagliptin 2.5 or 5 mg in combination with glyburide 7.5 mg vs. glyburide 10 mg for 24 weeks. Blinded uptitration glyburide was allowed in the glyburide-only arm to a maximum total daily dose of 15 mg. Efficacy analyses were performed using ANCOVA and last-observation-carried-forward methodology. Results: At week 24, 92% of glyburide-only patients were uptitrated to a total glyburide dose of 15 mg ⁄ day. Saxagliptin 2.5 and 5 mg provided statistically significant adjusted mean decreases from baseline to week 24 vs. uptitrated glyburide, respectively, in HbA 1c ()0.54%, )0.64% vs. +0.08%; both p < 0.0001) and fasting plasma glucose ()7, )10 vs. +1 mg ⁄ dl; p = 0.0218 and p = 0.002). The proportion of patients achieving an HbA 1c < 7% was greater for saxagliptin 2.5 and 5 mg vs. uptitrated glyburide (22.4% and 22.8% vs. 9.1%; both p < 0.0001). Postprandial glucose area under the curve was reduced for saxagliptin 2.5 and 5 mg vs. uptitrated glyburide ()4296 and )5000 vs. +1196 mgAEmin ⁄ dl; both p < 0.0001). Adverse event occurrence was similar across all groups. Reported hypoglycaemic events were not statistically significantly different for saxagliptin 2.5 (13.3%) and 5 mg (14.6%) vs. uptitrated glyburide (10.1%). Conclusion: Saxagliptin added to submaximal glyburide therapy led to statistically significant improvements vs. uptitration of glyburide alone across key glycaemic parameters and was generally well tolerated.

What's known
• Initial antihyperglycaemic monotherapy is frequently insufficient to enable patients with type 2 diabetes to achieve or sustain glycaemic targets. • Sulphonylurea therapy has demonstrated efficacy in improving glycaemic control; however, it is also associated with limitations, including the potential for weight gain and an increased risk of hypoglycaemia with sulphonylurea-induced hyperinsulinaemia.

What's new
• The availability of newer agents such as dipeptidyl peptidase-4 inhibitors may allow for additional therapeutic combinations to improve glycaemic control without significantly increasing the risk of adverse effects. • This study evaluated the safety and efficacy of saxagliptin added to submaximal-dose sulphonylurea (glyburide) therapy, compared with uptitration of sulphonylurea (glyburide) monotherapy for 24 weeks in patients with inadequate glycaemic control with submaximaldose sulphonylurea therapy alone.
improved glycaemic control and was well tolerated in patients with type 2 diabetes (10). Sulphonylureas are among the most frequently prescribed and least costly oral antidiabetic drugs (OADs); their mechanism of action involves binding to the beta-cell sulphonylurea receptor 1, which ultimately stimulates insulin release (11,12). While specific guidelines vary, sulphonylureas are most often recommended as second-line therapy (after metformin) (13) and in some cases, as first-line OAD therapy in patients who are not overweight (12). Studies have demonstrated clinically significant improvements in glycaemic control with sulphonylurea therapy. However, as monotherapy, 5-year failure rates based on a fasting plasma glucose (FPG) > 10.0 mmol ⁄ l after at least 6 weeks of treatment at the maximum dose are approximately 34% (3,11). Sulphonylurea therapy is associated with limitations, including the potential for beta-cell toxicity, weight gain and an increased risk of hypoglycaemia with sulphonylurea-induced hyperinsulinaemia (5,11,(14)(15)(16). Combination therapy of a DPP-4 inhibitor with a submaximal dose of a sulphonylurea represents an alternative treatment approach that may provide improved glycaemic control earlier in the disease course and allow the use of lower doses of sulphonylurea to reduce the risk for dose-related adverse events (AEs) (17,18).
The current trial evaluated the safety and efficacy of saxagliptin added to submaximal-dose sulphonylurea (glyburide) therapy, compared with uptitration of sulphonylurea (glyburide) monotherapy, in patients with type 2 diabetes and inadequate glycaemic control with submaximal-dose sulphonylurea therapy alone.

Patients
Patients aged 18)77 years (inclusive) with type 2 diabetes and inadequate glycaemic control (HbA 1c screening value ‡ 7.5 to £ 10.0%) on a submaximal sulphonylurea dose [defined as less than the maximum approved dose for each sulphonylurea (see Table S1

Study design
This study (CV181-040) was a 24-week, phase 3, randomised, multicentre, 3-arm, double-blind, international trial. Patients were recruited from outpatient settings, advertisements, postings and referrals. Eligible patients entered a 4-week, single-blind, dietary and exercise placebo lead-in period during which they discontinued their current sulphonylurea therapy and received open-label glyburide 7.5 mg ⁄ day. Patients were instructed by a registered dietitian, registered nurse, physician, certified diabetes educator or nutritionist on diet and exercise in accordance with the American Diabetes Association (ADA) or similar local guidelines to be followed for the study duration. Good compliance ( ‡ 80 to £ 120%) with placebo was required to be eligible for randomisation before the short-term treatment period. Patients with an HbA 1c level ‡ 7.0% and mean FPG (MFPG) or FPG ‡ 140 mg ⁄ dl (7.8 mmol ⁄ l) or mean fasting whole blood glucose (MFWBG) ‡ 131 mg ⁄ dl (7.3 mmol ⁄ l) continued treatment with open-label glyburide 7.5 mg ⁄ day and were randomised (1 : 1 : 1) via Interactive Voice Response System to one of three treatment groups (block size 3) utilising a double-dummy design: saxagliptin 2.5 mg ⁄ day (saxagliptin 2.5 mg + glyburide), saxagliptin 5 mg ⁄ day (saxagliptin 5 mg + glyburide) or placebo + blinded glyburide 2.5 mg ⁄ day [uptitrated glyburide; initial total daily dose (TDD) of glyburide 10 mg]. A one-time decrease in open-label glyburide to 5 mg ⁄ day was permitted at the investigator's discretion for patients who developed hypoglycaemia. Uptitration of blinded glyburide was permitted at weeks 2 and 4 in the uptitrated glyburide treatment group for patients satisfying prespecified glycaemic criteria [MFPG ‡ 100 mg ⁄ dl (5.5 mmol ⁄ l) or MFWBG ‡ 95 mg ⁄ dl (5.3 mmol ⁄ l)] to a maximum TDD of 15 mg (7.5 mg open-label + 7.5 mg blinded glyburide), provided the open-label glyburide dose had not been previously decreased because of hypoglycaemia. Throughout the study, double-blind study medication was to be taken twice daily, before the morning and evening meals to allow the glyburide dose to be split between morning and evening. Saxagliptin was to be taken in the morning. Patients were eligible for rescue therapy based on progressively strict glycaemic control criteria over 24 weeks if MFPG levels were: > 240 mg ⁄ dl (13.3 mmol ⁄ l) (weeks 4 and 6); > 220 mg ⁄ dl (12.2 mmol ⁄ l) (week 8); and > 200 mg ⁄ dl (11.1 mmol ⁄ l) (weeks 12, 16, 20 and 24), or if MFWBG > 221 mg ⁄ dl (12.3 mmol ⁄ l) (weeks 4 and 6); > 203 mg ⁄ dl (11.3 mmol ⁄ l) (week 8); or > 185 mg ⁄ dl (10.3 mmol ⁄ l) (weeks 12, 16, 20 and 24). Patients meeting rescue criteria entered into the long-term extension period, during which they were administered open-label metformin and glyburide in addition to blinded study medication. Patients completing 24 weeks of treatment without rescue also entered the long-term extension period. Long-term extension results will be reported in a future communication.

Efficacy assessments
The primary efficacy end-point was HbA 1c change from baseline to week 24. Secondary efficacy endpoints assessed at week 24 and listed in the order tested were change from baseline in FPG, proportion of patients achieving HbA 1c < 7.0%, and change from baseline in postprandial glucose (PPG) area under the curve (AUC) from 0 to 180 min in response to a 75-g oral glucose tolerance test (OGTT) with samples drawn at time )30 min, immediately prior to time 0 min, and at +30, +60, +120 and +180 min after oral glucose ingestion. Other efficacy end-points included the proportion of patients achieving HbA 1c £ 6.5%; change from baseline to week 24 in fasting insulin, C-peptide and glucagon; postprandial insulin, Cpeptide and glucagon AUC; beta-cell function [measured by homeostatic model assessment (HOMA)-2b]; insulin resistance (measured using HOMA-2IR); the proportion of patients requiring rescue for failing to achieve prespecified glycaemic targets or discontinuing for lack of efficacy; the proportion of patients achieving a glycaemic response at week 24 based on prespecified criteria; and PPG-AUC at the 120-min time point. Changes from baseline to week 24 in body weight, lipid parameters, insulinogenic index, Matsuda index (19) and oral glucose insulin sensitivity (OGIS) (20) were also examined. Subgroup analyses for baseline HbA 1c were prespecified.

Safety assessments
Safety and tolerability assessments included incidence of AEs, serious AEs (SAEs), AE-related discontinuations as well as results for electrocardiograms, vital signs and clinical laboratory tests. AEs of hypoglycaemia and confirmed hypoglycaemia, defined as symptoms of hypoglycaemia with a fingerstick glucose £ 2.8 mmol ⁄ l, were also recorded.

Statistical analyses
Efficacy analyses were conducted utilising data collected at baseline and postbaseline in the Randomised Patients data set, which consisted of all randomised patients who took at least one dose of double-blind study medication. An ANCOVA was performed on continuous efficacy end-points using last-observation-carried-forward (LOCF) methodology with treatment group as an effect and baseline value as the covariate. Within the framework of the ANCOVA model, point estimates and 95% confidence intervals were calculated for mean changes between each of the saxagliptin treatment groups and the uptitrated glyburide group. For the primary endpoint, each comparison between a saxagliptin treatment group and the uptitrated glyburide group was performed at the 0.027 alpha level from Dunnett's adjustment so that the overall (family-wise) type I error rate was controlled at the 0.05 significance level. Sequential testing methodology was used for secondary efficacy end-points. At each step in the testing sequence, only the saxagliptin treatment groups that were significantly superior to uptitrated glyburide were tested at the subsequent step. Summaries of categorical end-points such as the percentage of patients achieving a therapeutic glycaemic response at week 24, the proportion of patients requiring rescue ⁄ discontinuation because of lack of glycaemic control and the proportion of patients with reported and confirmed hypoglycaemia included frequencies and percentages for each treatment group; treatment groups were compared using the two-sided Fisher exact test. Demographic and other baseline characteristics were summarised using descriptive statistics by treatment group. LOCF methodology was used to handle missing data. The adjusted mean changes from baseline within each treatment group as well as the difference in mean change from baseline between each treatment group and the placebo or active comparator treatment group for each subgroup were calculated, as well as the corresponding subgroup by treatment interaction p-value. Systeme International (SI) conversion from mg ⁄ dl to mmol ⁄ l of glucose was calculated with the equation: mg ⁄ dl · 0.0555. All other SI conversions pertaining to data presented for this study are noted in Table S2. Estimated average glucose (eAG) values were calculated post hoc based on HbA 1c values using the linear regression: eAG mg ⁄ dl = 28.7 · HbA 1c ) 46.7 (21).
Safety analyses were performed in the treated patient population, consisting of patients who received at least one dose of study medication. Events of hypoglycaemia and confirmed hypoglycaemia were recorded and analysed separately from other AEs. Hypoglycaemic event intensity was graded according to the investigator's discretion, as were all other AEs. Efficacy and safety measurements obtained after rescue were not included in analyses.
Based on the primary end-point, the sample size afforded at least 90% power to detect a difference in HbA 1c means of 0.4% between each saxagliptin treat-ment group and the uptitrated glyburide treatment group, assuming a standard deviation (SD) of 1.2%.

Disposition, baseline demographics and disease characteristics
A total of 768 patients were randomised and treated with double-blind therapy; 563 patients completed the 24-week treatment period (Figure 1). Demographic and baseline clinical characteristics were generally well balanced across all treatment groups ( Table 1). The mean (SD) duration of the previous sulphonylurea treatment was 2.5 (3.54), 2.3 (2.96) and 2.4 (3.60) years in the saxagliptin 2.5-and 5-mg, and uptitrated glyburide groups, respectively. All patients had been previously treated with a sulphonylurea before study entry; 64% of patients received prior glyburide treatment.
Changes in other efficacy assessments are listed in Table 2. At week 24, saxagliptin treatment increased postprandial insulin and C-peptide AUC to a greater degree than did uptitrated glyburide. Saxagliptin treatment did not have an effect on fasting insulin or C-peptide. Postprandial glucagon AUC was decreased to a greater degree with saxagliptin treatment vs. uptitrated glyburide; decreases in fasting glucagon were only noted in the saxagliptin 5-mg group [)0.8 pg ⁄ ml ()0.8 ng ⁄ l)] vs. uptitrated glyburide [+4.5 pg ⁄ ml (+4.5 ng ⁄ l)]. There was no change in beta-cell function (HOMA-2b assessment) at week 24 in the saxagliptin treatment groups vs. uptitrated glyburide. Similarly, saxagliptin treatment was not associated with changes in early insulin response (as measured using the 30-min insulinogenic index), insulin sensitivity (based on OGIS and Matsuda indices) or insulin resistance (HOMA-2IR assessment) vs. uptitrated glyburide. Mean body weight increased in all treatment groups; adjusted mean increases were statistically significantly greater in each saxagliptin

Safety and tolerability
Overall, saxagliptin added to submaximal glyburide therapy was generally well tolerated. The proportion of patients reporting any AE was similar across all treatment groups, with no evidence of a doseresponse relationship (Table 3). One death (sudden cardiac death) occurred in the uptitrated glyburide group. The majority of AEs were mild or moderate in intensity.
The proportion of patients with skin-related AEs was: 8.9% [22 ⁄ 248 (saxagliptin 2.5 mg)], 4.7% [12 ⁄ 253 (saxagliptin 5 mg)] and 4.9% [13 ⁄ 267 (uptitrated glyburide)], with no evidence of a dose-related effect. AEs related to localised oedema were reported in two saxagliptin-treated patients: one in the saxagliptin 2.5-mg group and one in the saxagliptin 5mg group; none were reported in the uptitrated glyburide treatment group. Both events were reported to be of mild intensity and not related to study drug, and neither led to discontinuation. No events of Stevens-Johnson syndrome or angiooedema were reported. Cardiac disorder events were: 2.0% (5 ⁄ 248) for saxagliptin 2.5 mg, 4.0% (10 ⁄ 253) for saxagliptin 5 mg and 3.7% (10 ⁄ 267) for uptitrated glyburide. AEs of hypertension were reported for 9 ⁄ 248 (3.6%, saxagliptin 2.5 mg), 16 ⁄ 253 (6.3%, saxagliptin 5 mg) and 6 ⁄ 267 (2.2%, uptitrated glyburide) patients; however, mean systolic (SBP) and diastolic blood pressure (DBP) decreased in all treatment groups. The mean change from baseline at week 24 for saxagliptin 2.5 and 5 mg, and uptitrated There was no statistically significant difference in the incidence of reported and confirmed hypoglycaemic events in the saxagliptin 2.5-and 5-mg treatment groups vs. the uptitrated glyburide treatment group (Table 3). Confirmed hypoglycaemia occurred in 6 ⁄ 248 (2.4%), 2 ⁄ 253 (0.8%) and 2 ⁄ 267 (0.7%) of patients in the saxagliptin 2.5-and 5-mg treatment groups vs. the uptitrated glyburide treatment group. Of the 10 confirmed hypoglycaemic events, most were of mild or moderate intensity; a single hypoglycaemia-related AE judged to be severe in intensity by the study investigator occurred in a patient receiving saxagliptin 2.5 mg, which was easily managed by the patient. No hypoglycaemic event was judged by the study investigator to be an SAE or led to discontinuation of study therapy.

Discussion
The therapeutic goal for the treatment of type 2 diabetes is to achieve and maintain glycaemic levels without compromising safety or tolerability (22,23). This study demonstrated that in patients with type 2 diabetes not achieving glycaemic control on glyburide monotherapy, the addition of saxagliptin once daily to submaximal doses of glyburide for 24 weeks provided statistically significant and clinically meaningful reductions in key parameters of glycaemic control vs. uptitrated glyburide, without statistically significantly increasing the frequency of hypoglycaemia. Reductions observed in HbA 1c with saxagliptin added to submaximal glyburide corresponded to substantial improvements in FPG, the proportion of patients reaching goal, and PPG-AUC, with maximal benefits observed in the saxagliptin 5-mg group. These 24-week results are in contrast to increases in HbA 1c , FPG and PPG-AUC in the uptitrated glyburide group, indicating that the addition of saxagliptin to submaximal glyburide therapy is preferable to continued monotherapy with higher doses of sulphonylurea. HbA 1c reductions achieved with saxagliptin added to submaximal glyburide therapy occurred early. Decreases relative to uptitrated glyburide were noted by week 4 and were maintained throughout the remainder of the treatment period. Importantly, the effect of saxagliptin on HbA 1c lowering was consistent across prespecified subgroups including age, gender, ethnicity, BMI, geographic distribution and duration of diabetes, suggesting its appropriateness for use in a variety of patients with type 2 diabetes.
More than twice as many patients treated with saxagliptin 2.5 and 5 mg achieved HbA 1c goals recommended by the ADA and European Association for the Study of Diabetes (< 7.0%) vs. patients treated with uptitrated glyburide (24). Similarly, more than twice as many patients treated with saxagliptin

mg achieved the American Association of Clinical
Endocrinologists' recommended goal of HbA 1c £ 6.5%, compared with patients treated with uptitrated glyburide (25). Patients treated with saxagliptin achieved HbA 1c targets of < 7% more rapidly than patients receiving uptitrated glyburide, potentially reducing the harmful impact of glucotoxicity on betacell function earlier in the disease course (26). Relative to uptitrated glyburide, saxagliptin treatment led to an increase in postprandial insulin. A possible explanation for this observation is that the maximal insulin-releasing effect of a sulphonylurea may be enhanced by the addition of another OAD with a complementary mechanism of action also known to increase insulin levels or beta-cell function. DPP-4 inhibitors enhance GLP-1 and GIP levels, thereby acting to promote insulin synthesis and release and improve beta-cell function, suggesting their utility in combination with sulphonylurea therapy. Saxagliptin 2.5 and 5 mg were generally well tolerated, with no clinically meaningful difference between treatment groups in the proportion of patients reporting AEs. There was no evidence of a dose-response relationship for AEs. Combination therapy is often associated with an increased risk for hypoglycaemia, particularly combinations that use sulphonylureas or insulin (27). Treatment with saxagliptin did not statistically significantly increase the frequency of hypoglycaemia, compared with glyburide uptitration, even with significantly greater glycaemic efficacy achieved with add-on therapy. Despite no observed statistically significant increase in hypoglycaemia in this study, these results must be viewed with appropriate caution given that hypoglycaemia is a concern in any sulphonylurea-based therapeutic regimen.
An increase from baseline in body weight, of small magnitude and unclear clinical relevance, occurred in all treatment groups; however, weight gain was higher in the saxagliptin treatment groups. Improved glycaemic control has been shown to promote weight gain in certain instances by decreasing glucosuria (13,28). A possible explanation is that the weight gain in the saxagliptin treatment groups was a result of decreased glucosuria, whereas sulphonylureainduced weight gain may have been mitigated by the glucosuria caused by inadequate glycaemic control in the uptitration arm.
Only data collected prior to rescue were used for efficacy and safety analyses, which could be a study limitation in that experience following rescue therapy is not included in the analyses. However, this approach was taken to minimise the potential confounding of rescue therapy on both safety and efficacy parameters.
In patients with type 2 diabetes and inadequate glycaemic control, initiating therapy with a sulphonylurea and uptitrating the dosage until maximum concentrations are reached is an approach commonly used in clinical practice in certain regions. The addition of saxagliptin to submaximal sulphonylurea treatment -as opposed to uptitration -offers a potentially improved strategy for achieving tighter glycaemic control and contrasts with existing 'treatto-failure' approaches, in which clinical inertia often leads to an unacceptable glycaemic burden (29). Other randomised controlled studies of oral antidiabetic agents in combination with submaximal sulphonylurea therapy have also demonstrated improved glycaemic control, compared with uptitration of sulphonylurea monotherapy (30,31). Further, the addition of saxagliptin to submaximal sulphonylurea therapy has the potential to yield additional benefits, such as preservation of beta-cell function and reduction of potentially toxic beta-cell effects associated with maximal sulphonylurea therapy (15,16).

Conclusions
The glycaemic benefits demonstrated in this study support the use of the DPP-4 inhibitor saxagliptin as add-on therapy to glyburide or other sulphonylurea agents in patients with type 2 diabetes and inadequate glycaemic control with sulphonylurea monotherapy. The favourable tolerability profile of combination therapy assessed in the current trial indicates that the addition of saxagliptin may be appropriate when a submaximal dose of glyburide or other sulphonylurea provides inadequate glycaemic control.

Funding and Acknowledgements
uted to the data analysis ⁄ interpretation, drafting and critical revision of the manuscript content and approval of the final manuscript.

Supporting Information
Additional Supporting Information may be found in the online version of this article: Table S1 Eligible sulphonylureas and doses.