Safety and Efficacy of Treatment with Sitagliptin or Glipizide in Patients with Type 2 Diabetes Inadequately Controlled on Metformin: A 2-year Study

T. Seck; M. Nauck; D. Sheng; S. Sunga; M. J. Davies; P. P. Stein; K. D. Kaufman; J. M. Amatruda

Disclosures

Int J Clin Pract. 2010;65(5):562-576. 

In This Article

Results

Patient Disposition and Characteristics

In this study, 519 of the 1172 randomised patients (44%) completed the 2-year treatment period, of which 504 were included in the PP analysis (sitagliptin n = 248; glipizide n = 256) (Figure 1). The proportions of patients discontinuing treatment and the reasons for discontinuation were generally similar between groups, with patients discontinuing for lack of efficacy (i.e. patients not meeting the progressively stricter, protocol-specified glycaemic criteria and/or not meeting the investigator's expectations of glycaemic improvement) accounting for 52% of the discontinuations over the 2-year treatment period (Figure 1). For the randomised cohort, the baseline characteristics were generally similar between treatment groups, as previously published.[9] The baseline characteristics of the patients in the 2-year PP cohort were also similar between the sitagliptin and glipizide groups (Table 1). As a result of the progressively stricter glycaemic criteria requiring discontinuation from the study, patients in the 2-year PP cohort tended to have better glycaemic control at baseline, with lower mean HbA1c and FPG values and a shorter mean duration of type 2 diabetes compared with those who did not complete the 2-year treatment period (data not shown).

Figure 1.

Patient disposition for the 2-year study. *In the sitagliptin group, the number of patients who discontinued was previously reported as 202 (Nauck et al. (9)), but was corrected to 200 during the 2nd year of the study. Patients discontinued for lack of efficacy included patients not meeting the progressively stricter protocol-specified glycaemic criteria and/or not meeting the investigator's expectations of glycaemic improvement. aAll-patients-treated (APT) cohort includes randomised patients who received at least one dose of study treatment and who had both a baseline and at least one post-baseline measurement. bPer-protocol (PP) cohort includes randomised patients who completed 2 years of treatment and did not have any reasons for exclusion from this cohort, including no baseline data, no treatment data at week 104, or major protocol violations.

Over the 2-year treatment period, the mean daily dose of glipizide in the PP population was 9.2 mg per day, with approximately 66% reaching a dose of glipizide of at least 10 mg per day. At study end, 16% of the glipizide-treated patients were on a dose of 20 mg per day. Down-titration or interruption of glipizide was permitted as needed to prevent recurrent hypoglycaemia, and 10% of patients were not taking glipizide in the 2 weeks period prior to the end of the study. For the APT cohort, the mean daily dose of glipizide was 9.5 mg per day over the 2-year treatment period. For all randomised patients, the mean duration of exposure to study drug was modestly greater in the sitagliptin group [483.1 days (69.0 weeks)] compared with the glipizide group [467.0 days (66.7 weeks)]. The mean (median) compliance rates, defined as the proportion of study drug taken in relation to the prescribed study drug over 2 years, were 98.5% (99.8%) and 98.0% (99.7%) in the sitagliptin and glipizide groups respectively.

Efficacy

In the PP cohort, the addition of sitagliptin to ongoing metformin monotherapy led to similar reductions in HbA1c from baseline after 2 years of treatment compared with the addition of glipizide (Table 2). In the PP cohort, the proportion of randomised patients with an HbA1c < 7% at the end of the 2nd year was similar between the sitagliptin (63%; n/N = 157/248) and the glipizide (59%; 151/256) groups. Of those patients in the PP analysis who had an HbA1c < 7% at the end of the 1st year, completed 2 years of therapy and had a HbA1c measurement at year 2, 73% (n/N = 141/193) in the sitagliptin group and 69% (135/196) in the glipizide group had an HbA1c < 7% at the end of the 2nd year.

The rise in HbA1c from week 24 to the end of the 2nd year was less with sitagliptin treatment compared with glipizide [coefficient of durability (COD) (95% CI): 0.16%/year (0.10, 0.21) vs. 0.26%/year (0.21, 0.31) respectively; between-group difference in COD (95% CI) = −0.10%/year (−0.16, −0.05)] (Figure 2).

Figure 2.

For the per-protocol cohort, HbA1c change (LS mean ± SE) over 2 years in patients on ongoing metformin therapy treated with sitagliptin 100 mg q.d. or glipizide

Results based on the PP cohort were supported by the results based on the APT cohort. For the APT cohort, LS mean HbA1c change (95% CI) from a mean baseline of 7.69% was −0.33% [(−0.42, −0.25); n = 576] in the sitagliptin group and from a mean baseline of 7.65% was −0.35% [(−0.44, −0.26); n = 559] in the glipizide group [between-group difference (95% CI) = 0.01% (−0.08, 0.10)]. In the APT cohort at week 104, 42% and 39% of the patients had an HbA1c < 7% in the sitagliptin and glipizide groups respectively. Similar between-group differences for COD were noted for the PP and APT cohorts (data not shown).

In the PP cohort for both treatments, the FPG change from baseline at the end of the 2nd year was similar between groups (Table 2 and Figure 3). For the APT cohort, the LS mean change from baseline (95% CI) in FPG was −0.6 mmol/l (−0.8, −0.3) [−10.1 mg/dl (−14.4, −5.7); n = 581] in the sitagliptin group and −0.6 mmol/l (−0.8, −0.3) [−10.3 mg/dl (−14.7, −6.0); n = 566] in the glipizide group [between-group difference (95% CI) = 0.0 mmol/l (−0.2, 0.3)].

Figure 3.

For the per-protocol cohort, fasting plasma glucose (FPG) change (LS mean ± SE) over 2 years in patients on ongoing metformin therapy treated with sitagliptin 100 mg q.d. or glipizide

At the end of the 2nd year for the PP cohort, fasting insulin increased from baseline in the glipizide group with no change observed in the sitagliptin group, resulting in a modest difference between groups (Table 2). A decrease from baseline in fasting proinsulin and the proinsulin/insulin ratio at the end of the 2nd year was observed in the sitagliptin group relative to the glipizide group. An increase in HOMA-β from baseline was observed with glipizide. Improvements in HOMA-IR and QUICKI were observed with sitagliptin relative to glipizide at the end of the 2nd year (Table 2), suggesting a small decrease in insulin resistance associated with sitagliptin treatment.

In the subset of patients who volunteered to undergo the 9-point meal tolerance testing, the baseline demographic and disease characteristic profiles were generally similar to those of the overall study population, and were also similar between treatment groups within this subset (data not shown). After 2 years of treatment and following a 4- to 7-day blinded (the patients received placebo matching their study drug) treatment washout period, patients in the sitagliptin treatment group had lower postprandial plasma glucose excursion (AUC) after the meal relative to baseline, with no change from baseline in glucose excursion in patients in the glipizide treatment group (Table 3, Figure 4A). Serum C-peptide and insulin 3-h AUC were numerically increased with sitagliptin compared with numerical reductions with glipizide (Table 3, Figure 4B and C). The insulin AUC to glucose AUC ratio was improved from baseline in the sitagliptin treatment group [LS mean change from baseline in ratio (95% CI) = 4.21 pmol/mmol (0.11, 8.42)], whereas the ratio was minimally changed in the glipizide treatment group [0.11 pmol/mmol (−3.89, 4.21)].

Figure 4.

Plasma glucose (A), serum insulin (B), and serum C-peptide (C) profiles during the nine-point meal tolerance test at baseline and following a 4- to 7-day washoff of study drug following 2 years of treatment with sitagliptin or glipizide added to ongoing metformin therapy (mean ± SE)

Beta-cell responsiveness to the meal challenge (performed after the 4–7 day washout period) was assessed using a standard model-based approach. Measures of beta-cell responsiveness postmeal (Φs and Φd) remained stable relative to baseline in patients who had been treated with sitagliptin, while a reduction in responsiveness was observed in patients who had received glipizide (Table 3). The Φtotal, a composite of static and dynamic beta-cell responsiveness, improved from baseline with sitagliptin and remained unchanged with glipizide. The measure of basal beta-cell responsiveness (Φb) remained unchanged relative to baseline in both groups. DI, an assessment of beta-cell responsiveness in relation to insulin sensitivity, remained stable over 2 years with sitagliptin, but declined from baseline with glipizide (Table 3). Overall, these changes suggest that sitagliptin, compared with glipizide, led to better maintenance of beta-cell function after 2 years of treatment. Similar trends for within-group changes were observed after 1 year of treatment (data not shown).

For the PP cohort, LS mean percent changes from baseline in TC were 4.2% and 0.2% with sitagliptin and glipizide respectively [between-group difference (95% CI) = 4.0% (1.1, 6.9)] from a mean baseline of approximately 4.5 mmol/l. For LDL-C, LS mean percent changes from baseline were 7.4% and −1.0% with sitagliptin and glipizide respectively [between-group difference (95% CI) = 8.4% (2.5, 14.3)] from a mean baseline of approximately 2.5 mmol/l. For HDL-C, LS mean percent changes from baseline were 4.4% and 1.3% with sitagliptin and glipizide respectively [between-group difference (95% CI) = 3.1% (0.7, 5.4)] from a mean baseline of approximately 1.2 mmol/l. For TG, LS mean percent changes from baseline were 9.2% and 11.8% with sitagliptin and glipizide respectively [between-group difference (95% CI) = −2.7% (−11.2, 5.8)].

Safety and Tolerability

Over 2 years of treatment, the incidence of clinical adverse experiences overall was lower in the sitagliptin group compared with the glipizide group (Table 4). The proportion of patients experiencing drug-related clinical adverse experiences was also lower with sitagliptin compared with glipizide. These differences were related primarily to the higher incidence of hypoglycaemia in the glipizide group. Nine deaths occurred over the 2-year treatment period: eight in the glipizide group [sudden cardiac death, myocardial infarction (n = 2), cancer-related deaths (n = 3), sepsis and a suicide that occurred 41 days following discontinuation of study drug] and one in the sitagliptin group (trauma related to being struck by a motor vehicle) (Table 4). None of the deaths was considered by the investigator as related to study drug.

The incidences of serious clinical adverse experiences were similar between treatment groups. For serious adverse experiences that were considered by the investigator to be related to study drug, there were three in the glipizide group (myocardial infarction, spontaneous abortion, and hydronephrosis) and one in the sitagliptin group (thrombocytopenia). The patient with thrombocytopenia experienced this event on day 407 and discontinued treatment with sitagliptin on day 409. A couple of days prior to the event the patient took nimesulide for headache. The investigator believed that thrombocytopenia was probably related to nimesulide; however, the investigator could not exclude study medication as a cause for the event and reported the adverse experience as possibly related to study medication. The platelet count improved upon initiation of corticosteroid therapy, with resolution of the event by day 446. The proportions of patients who discontinued because of adverse experiences were similar between treatment groups (Table 4).

Over the 2 years, there was a higher incidence (defined as between-group difference in incidence ≥ 1%) for 13 specific clinical adverse experiences in the sitagliptin group and 10 specific clinical adverse experiences for the glipizide group (Table 5). Of the adverse experiences with a higher incidence in the sitagliptin group, the 95% CI around the between-group difference in incidence excluded zero for cystitis, urinary tract infection, weight decrease, pain in extremity and asthma. Overall, the pattern of the adverse experiences (i.e. intensity, onset/duration, duration, etc.) of urinary tract infection and cystitis were similar in both treatment groups and most cases were assessed as mild or moderate in intensity and did not result in discontinuation. The mean duration of urinary tract infection and cystitis were similar in both groups and recurrence of events was uncommon. The related adverse experience of pyelonephritis was uncommon over the 2-year treatment period, reported in three patients in the glipizide group and one patient in the sitagliptin group. The overall incidence of infection-related adverse experiences was similar in the two treatment groups.

Of the adverse experiences with a higher incidence in the glipizide group, the 95% CI around the between-group difference in incidence excluded zero for cataract, toothache, hypoglycaemia and hypoaesthesia (Table 5). With the exception of hypoglycaemia, these adverse experiences (occurring in either group) were generally rated as mild in intensity, not considered related to study drug, and resolved while patients continued on study drug.

The incidences of gastrointestinal events overall and of the prespecified gastrointestinal events, abdominal pain, diarrhoea, nausea, and vomiting, were similar in the sitagliptin and glipizide groups.

There were 199 (34.1%) glipizide-treated patients for whom 805 episodes of hypoglycaemia were reported compared with 31 (5.3%) sitagliptin-treated patients for whom 57 episodes of hypoglycaemia were reported over the 2-year treatment period (Table 5). In addition to a substantially greater proportion of patients with one or more events of hypoglycaemia in the glipizide group relative to the sitagliptin group, the frequency of events among patients who had hypoglycaemia was also higher with glipizide treatment (Figure 5). Of the 726 episodes with concurrent fingerstick glucose values in the glipizide group, 554 (76%) were < 3.9 mmol/l (70 mg/dl), 242 (44%) were < 3.3 mmol/l (60 mg/dl) and 61 (8%) were < 2.8 mmol/l (50 mg/dl). Of the 49 episodes with concurrent fingerstick glucose values in the sitagliptin group, 35 (71%) were < 3.9 mmol/l (70 mg/dl), 14 (29%) were < 3.3 mmol/l (60 mg/dl), and 4 (8%) were < 2.8 mmol/l (50 mg/dl). During the first year, 83% of the 805 episodes of hypoglycaemia in the glipizide group and 88% of the 57 episodes of hypoglycaemia in the sitagliptin group occurred. Of the patients who experienced hypoglycaemic episodes, 18 and 2 patients required assistance in the glipizide and sitagliptin groups, respectively. Nine patients (1.5%) in the glipizide group had an episode that required medical assistance or exhibited markedly depressed level of consciousness compared with one patient (0.2%) in the sitagliptin group. Furthermore, nine patients (1.5%) on glipizide had a hypoglycaemic episode that required non-medical assistance compared with one patient (0.2%) on sitagliptin.

Figure 5.

Proportion of patients with 0, 1, 2, 3, 4, 5, 6, or more than 6 hypoglycaemic episodes in the sitagliptin group (black bars) and glipizide group (grey bars) during the 2-year study

No meaningful between-group differences were observed for change from baseline in vital signs or ECG data. The addition of sitagliptin to metformin over the 2-year treatment period was associated with a reduction in body weight relative to baseline [LS mean change from baseline (95% CI) = −1.6 kg (−2.3, −1.0)], whereas the addition of glipizide increased body weight [0.7 kg (0.0, 1.3)]. The different pattern of body weight change led to a between-group difference of −2.3 kg (−3.0, −1.6) (Figure 6).

Figure 6.

For the all-patients-as-treated cohort, body weight change (LS mean change from baseline ± SE) over 2 years in patients on ongoing metformin therapy treated with sitagliptin 100 mg q.d. or glipizide

The proportions of patients for whom a laboratory adverse experience was reported were similar between groups (14.5% for sitagliptin and 12.8% for glipizide). There were no clinically meaningful differences between groups in the proportion of patients with values meeting predefined limits of change criteria for any specific chemistry or haematology analyte, including hepatic transaminases.

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as:

processing....