Discussion
Though metformin can effectively inhibit the secretion of glycogen in the liver and increase the sensitivity of peripheral tissue to insulin, about 30% of patients of T2DM are still troubled by poor glycemic control or adverse reactions. Here our analysis showed that (1) GKAs, alone or combined with metformin, could effectively reduce FPG and glycosylated hemoglobin compared with placebo; (2) GKAs, alone or combined with metformin, did not affect the incidence of AE but increased the incidence of hypoglycemic events compared with placebo. Thus, our results suggest GKAs may help metformin to better regulate blood glucose with good tolerance.
Some previous studies have shown that the effective duration of GKAs is relatively short14,24. In this study, we included short-term treatment (2 days) and mid-term treatment (4 months), and the data showed that GKAs was effective in the treatment of T2DM compared with the placebo-controlled group, and there was no significant heterogeneity. Therefore, GKAs may reduce FPG and HbA1c in patients with T2DM in short to mid-term. Moreover, our study shows that the combination of GKAs may help patients, who still have poor glycemic control after taking metformin, better control their blood glucose. Of note, in the study of Zhang et al 2013, the combination of GKAs and metformin could effectively reduce FPG and HbA1c in the patients with poorly controlled blood glucose by metformin20, suggesting that GKAs can effectively help metformin adjust blood glucose by regulating GK. At the same time, earlier use of combination therapy to prolong the duration of drug effect is also consistent with the direction of ADA diabetes treatment guidelines25.
Regarding the safety, the main side effect of metformin is the gastrointestinal reaction, which can cause nausea, vomiting, diarrhea, and so on, which is related to the main effect of metformin on the intestine. Our analysis showed that the combination of GKAs and metformin did not significantly increase the risk of gastrointestinal related AEs, such as nausea and diarrhea, but increased the risk of hypoglycemia. Our subgroup analysis showed that dual-acting allosteric glucokinase activator did increase the risk of hypoglycemia compared with the control group, however liver-selective glucokinase activator did not. A recent study in China, D.Zhu et al.2018, has shown that the new liver selection of GKA (Dorzagliatin), which avoids the reduction of GK targeting in the pancreas, has no risk of hypoglycemia23. Meanwhile, another study showed that the incidence of hypoglycemia in GKAs can be effectively reduced by adjusting the affinity between GK and substrate26. However, some studies have shown that GKAs also bring the risk of hypoglycemia and increasing triglycerides. Zhai at el. showed that piragliatin(GKA) combined with other diabetes drugs increases the risk of hypoglycemia27. And Wilding at el. showed that triglycerides increased by 18–22% in randomized patients receiving titrated AZD1656(GKA) 4 months in the main study population19. Therefore, the safety of GKAs may need further evaluation, especially for hypoglycemia and triglycerides.
Although we included high-quality trials, several limitations in our meta-analysis remain to be noted as follows: (1) The number of included studies is limited, and only groups with higher dose and frequency of administration are included in each study, which may lead to the risk of selection bias; (2) There are other interference factors because of the study design of each study is different, such as the type and dosage of GKAs (MK-0941/AZD1656, etc.) and the time point or follow-up of FPG and HbA1c evaluation; (3) The long-term efficacy and safety of GKAs are still unclear.