Sodium Glucose Co-transport 2 Inhibitors in the Treatment of Type 2 Diabetes Mellitus

A Meta-analysis of Randomized Double-Blind Controlled Trials

Asres Berhan; Alex Barker

Disclosures

BMC Endocr Disord. 2013;13(58) 

In This Article

Results

Using the Google scholar search engine about 6,390 literatures on SGLT2 inhibitors were identified. One hundred and four articles were retrieved. After reviewing the abstracts of all the retrieved articles 26 articles were selected for full document review. Then, following the full document review of the 26 articles, 17 fulfilled the predetermined inclusion criteria. Eleven of studies were on dapagliflozin,[15,18–27] 3 on canagliflozin,[16,28,29] 2 on ipragliflozin,[30,31] and the remaining one was on empagliflozin[32] (Figure 1). A total of 4,811 patients with type 2 diabetes received one of the SGLT2 inhibitors alone (2,686) or in combination with other antidiabetic drugs (2,125) and 1,921 received placebo (887) or placebo with other antidiabetic drugs (1,034) ( Table 1 ). The risk of bias assessment in the included individual studies did not demonstrate the presence of bias in randomization, blinding or reporting. Similarly, funnel plot did not also show the existence of publication bias.

Figure 1.

Study selection flow diagram.

As presented in Figure 2, the pooled analysis of the mean change in HbA1c from baseline established a significant reduction in patients who were treated with SGLT2 inhibitors than placebo treated patients (overall SMD = −0.78; 95%CI, -0.86 to −0.69). All the SGLT2 inhibitors included in the meta-analysis, canagliflozin (subtotal SMD = −0.97; 95%CI, -1.25 to −0.69) dapagliflozin (subtotal SMD = −0.73; 95%CI, -0.82 to −0.64), ipragliflozin subtotal SMD = −0.68; 95%CI, -0.861 to −0.490) and empagliflozin subtotal SMD = −0.78; 95%CI, -0.967 to −0.599), demonstrated the significant reduction in HbA1c. The reduction in HbA1c appears more prominent in canagliflozin treated patients. However, heterogeneity testing revealed the presence of a considerable heterogeneity among the studies on canagliflozin (I2 = 90%) and a moderate heterogeneity among studies on dapagliflozin (I2 = 57%) and ipragliflozin (I2 = 56%).

Figure 2.

Standardize mean difference of the change in HbA1c from baseline.

Subgroup analysis based on the doses of SGLT2 inhibitors and the type of regimen (SGLT2 inhibitors monotherapy vs SGLT2 inhibitors in combination with other antidiabetic drugs) and meta-regression using duration of therapy and the doses of SGLT2 inhibitors as a covariates did not show a significant difference in HbA1c change from baseline. On the other hand sensitivity analysis confirmed the stability of the overall SMD when any of the studies with a specific dose removed from the analysis. The overall SMD ranged within −0.75 to −0.79%.

In support of the above analysis, the odds of SGLT2 inhibitors treated patients who achieved HbA1c < 7.0% were more than two folds of placebo treated groups (overall OR = 2.09; 95% CI, 1.77 to 2.46). Similarly, the mean FPG levels (overall SMD = −0.70 mg/mL, 95% CI, -0.79 to −0.61) and mean body weight (overall SMD = −0.59 kg; 95% CI, −0.66 to −0.52) of patients who were treated with SGLT2 inhibitors were significantly decreased from baseline compared to placebo treated patients (Figure 3). Furthermore, treatment with SGLT2 inhibitors was significantly associated with a reduction in both systolic (overall SMD = −0.27 (mmHg; 95% CI, -0.34 to −0.20) and diastolic (overall SMD = −0.24, 95% CI, -0.30 to −0.17) blood pressure from baseline. Most of the individual studies did not show the significant association of SGLT2 inhibitors with an increase in HDL cholesterol level from baseline. However, the overall SMD demonstrated a significant increase in HDL cholesterol level in patients who were treated with SGLT2 inhibitors (overall SMD = 0.21 mg/dl; 95% CI, 0.09 to 0.33). The change in the level of LDL cholesterol from baseline in SGLT2 inhibitors treated groups was not different from placebo treated groups (overall SMD = 0.07 mg/l; 95% CI, -0.01 to 0.14).

Figure 3.

Standardize mean difference of the change in body weight from baseline.

Even though the SGLT2 inhibitors with all doses did not show association with adverse events, the overall OR revealed the significant association of SGLT2 inhibitors with adverse events (overall OR = 1.18; 95% CI, 1.08 to 1.29) (Figure 4). The subtotal ORs in the subgroups of canagliflozin (subtotal OR = 1.31; 95% CI, 1.08 to 1.59) and dapagliflozin (subtotal OR = 1.17; 95% CI, 1.05 to 1.31) showed significant association with adverse events. Whereas the subtotal ORs in the subgroups of ipragliflozin was not statistically significant (OR = 0.95; 95% CI, 0.677 to 1.325). Dapagliflozin (subtotal OR = 3.07; 95% CI, 2.32 to 4.05) and canagliflozin (subtotal OR = 3.42; 95% CI, 1.86 to 6.28) were associated with genital tract infections. Dapagliflozin was also associated with urinary tract infection (subtotal OR = 1.32; 95% CI, 1.06 to 1.63). Nevertheless the number of patients who were treated with SGLT2 inhibitors and experienced serious adverse events was not different from placebo treated groups (overall OR = 0.83; 95% CI, 0.65 to 1.05). Similarly the number of patients who experienced hypoglycemic events during the study periods was not different from placebo treated groups (overall OR = 1.16; 95% CI, 0.94 to 1.43). As presented in Figure 5, the odds of discontinuation of medication due to adverse events in the SGLT2 inhibitors treated patients was not significantly different from placebo treated patients (overall OR = 1.05; 95% CI, 0.81 to 1.36).

Figure 4.

Mantel-Haenszel odds ratio of patients who experienced any adverse events.

Figure 5.

Mantel-Haenszel odds ratio of patients who discontinued the treatment due to adverse events.

processing....