Long-term 2-year safety and efficacy of vildagliptin compared with rosiglitazone in drug-naïve patients with type 2 diabetes mellitus

Diabetes Obes Metab. 2009 Jun;11(6):571-8. doi: 10.1111/j.1463-1326.2008.01021.x. Epub 2009 Apr 5.


Aims: To assess the long-term safety and the sustained glycaemic control of vildagliptin compared with rosiglitazone over 2-year treatment in drug-naïve type 2 diabetes mellitus patients.

Methods: This was an additional 80-week, multicentre, double-blind and active-controlled extension to a 24-week core study comparing the treatments of vildagliptin (50 mg b.i.d., n = 396) to rosiglitazone (8 mg q.d., n = 202). The primary efficacy variable was the mean change in haemoglobin A1c (HbA1c) from the core study baseline (day 1) to the end of 104 weeks (the extension endpoint).

Results: Vildagliptin and rosiglitazone showed statistically significant and sustained HbA1c reductions from a core mean baseline of 8.6 and 8.7% to 7.8 and 7.3% respectively (both significant, p < 0.001). However, rosiglitazone-treated patients showed significantly greater mean HbA1c reductions (mean difference 0.62%, s.e. 0.13, p < 0.001) compared with vildagliptin. The overall lipid profile significantly improved with vildagliptin compared to rosiglitazone treatment. Body weight remained unchanged in vildagliptin-treated patients despite improvements in glycaemic control but significantly increased (mean change from core study baseline 4.67 kg) in rosiglitazone-treated patients (p < 0.001). Notably, a lower incidence of peripheral oedema was seen with vildagliptin (4.6%) compared with rosiglitazone treatment (11.1%). More serious adverse events (SAEs) occurred in vildagliptin- than rosiglitazone-treated patients (12.5 and 9.1% respectively), but only one SAE each in both treatment group was suspected to be related to study drug. Three non-study drug-related deaths (vildagliptin: 2 and rosiglitazone: 1) were reported. Four mild hypoglycaemic events were observed with vildagliptin.

Conclusions: This study showed that the similar short-term HbA1c reductions seen with both vildagliptin and rosiglitazone treatments were more durable after 104 weeks of treatment with rosiglitazone than vildagliptin. However, this greater durability with rosiglitazone was at the expense of weight gain (almost 5 kg), higher incidences of peripheral oedema and a less favourable plasma lipid profile compared with vildagliptin.

Trial registration: ClinicalTrials.gov NCT00138619.

Publication types

  • Comparative Study
  • Multicenter Study

MeSH terms

  • Adamantane / analogs & derivatives*
  • Adamantane / therapeutic use
  • Adult
  • Aged
  • Blood Glucose
  • Body Weight / drug effects
  • Diabetes Mellitus, Type 2 / drug therapy*
  • Diabetes Mellitus, Type 2 / metabolism
  • Dipeptidyl-Peptidase IV Inhibitors / therapeutic use*
  • Double-Blind Method
  • Edema / chemically induced
  • Female
  • Glycated Hemoglobin / metabolism
  • Humans
  • Hypoglycemic Agents / therapeutic use*
  • Lipids / blood
  • Male
  • Middle Aged
  • Nitriles / therapeutic use*
  • Pyrrolidines / therapeutic use*
  • Rosiglitazone
  • Thiazolidinediones / therapeutic use*
  • Treatment Outcome
  • Vildagliptin
  • Weight Gain


  • Blood Glucose
  • Dipeptidyl-Peptidase IV Inhibitors
  • Glycated Hemoglobin A
  • Hypoglycemic Agents
  • Lipids
  • Nitriles
  • Pyrrolidines
  • Thiazolidinediones
  • Rosiglitazone
  • Vildagliptin
  • Adamantane

Associated data

  • ClinicalTrials.gov/NCT00138619