Balancing efficacy and safety: glucokinase activators in glycemic and metabolic management of type 2 diabetes mellitus—a meta-analysis

Abstract

Type 2 diabetes mellitus (T2DM) is a persistent condition typically defined by prolonged hyperglycemia resulting from beta-cell impairment and insulin resistance. Glucokinase activators (GKAs) are new medications that target glucokinase (GK) to increase glucose utilization in both the pancreas and liver. Its effectiveness and safety are inconsistent across various doses and types. The meta-analysis assessed the effectiveness and safety of GKAs in T2DM on glycemic control (hemoglobin A1c [HbA1c], fasting plasma glucose [FPG], postprandial plasma glucose [PPG]) as well as metabolic parameters (lipids, body weight, safety outcomes) stratified by dosage, type, and intervention time. The study involved a comprehensive analysis of 3,854 participants drawn from 11 randomized controlled trials (RCTs). Standardized mean differences (SMDs) and risk ratios (RRs) were computed employing random-effects models, alongside conducting sensitivity and subgroup analyses. GKAs effectively lowered HbA1c and PPG, especially high-dose (HbA1c: SMD = –0.43, 95% CI: –0.62 to –0.24) and dual-acting GKAs. Medium and high-dose GKAs were associated with increased risk of hypoglycemia (RR = 1.50; 1.63). At 24 weeks, GKAs led to increases in triglycerides, body weight, and liver enzymes, with the majority of these effects subsiding by 52 weeks. GKAs provide favorable glycemic control but carry dose-dependent concerns. While dual-acting GKAs demonstrate impressive efficacy, hepatoselective GKAs reveal improved safety. There exists a pressing need for further longitudinal studies and customized treatment approaches concerning the utilization of GKAs.

PROSPERO registration: CRD42020188517

1. Introduction

Type 2 diabetes mellitus (T2DM) is a prevalent metabolic disease distinguished by inadequate secretion of insulin and the lack of response to insulin, resulting in persistent hyperglycemia [1]. This condition adversely affects the quality of life for patients and is closely associated with several complications, such as cardiovascular disease, nephropathy, retinopathy, and nervous system disorders [2]. Despite the considerable advancements in diabetes management through the utilization of existing glucose-lowering agents, such as metformin and more recent classes like glucagon-like peptide-1 (GLP-1) receptor agonists and sodium-glucose co-transporter 2 (SGLT-2) inhibitors, attaining optimal glycemic control continues to pose challenges for numerous patients, alongside ongoing concerns regarding long-term efficacy and safety [3]. Such limitations have propelled the search for new therapeutic modalities targeting more complete metabolic control with favorable safety profiles.

Glucokinase activators (GKAs) are a class of small-molecule drugs that allosterically activate glucokinase (GK) and have shown distinctive properties in the treatment of T2DM [4]. It possesses the ability to improve the insulin response and optimize glucose metabolism through its effects on the pancreas and liver, thereby demonstrating its potential to lower levels of haemoglobin A1c (HbA1c) and postprandial plasma glucose (PPG) [5]. However, clinical studies in recent years have shown that long-term use of GKAs may be associated with the risk of hypoglycemia, metabolic disorders, and liver function abnormalities, which vary depending on diverse types and doses [6, 7].

The existing body of literature regarding GKAs presents a discordant view on their efficacy and safety, primarily relying on isolated studies or specific dosages, lacking a comprehensive evaluation. Considering this, the present study performed a meta-analysis to derive a comprehensive estimate of the beneficial and adverse effects of GKAs in people with T2DM. In particular, it analyzed the effects of type, dose, and duration of intervention on glycemic control measures (HbA1c, fasting plasma glucose [FPG] and PPG) as well as metabolic parameters (e.g., body weight and lipid profile). This analysis aims to provide substantial data to support the therapeutic use of GKAs, the personalization of treatment strategies, and the identification of research opportunities.

2. Materials and Methods

2.1 Study registration

This systematic review and meta-analysis was conducted according to the PRISMA guidelines and registered in the PROSPERO database (Registration No: CRD42020188517).

2.2 Search strategy

A comprehensive search was conducted across the PubMed, Embase, and Cochrane Library databases, with the most recent update occurring on August 6, 2024. Our approach involved a search strategy that integrated controlled vocabulary, such as MeSH terms in PubMed and Emtree in Embase, alongside free-text keywords, utilizing Boolean operators AND, OR, and NOT. The search terms encompassed glucokinase activators, GKAs, diabetes mellitus, type 2 (MeSH), type 2 diabetes (Title/Abstract), and type 2 diabetic mellitus (Title/Abstract). All search queries were confined to investigations involving human subjects and publications in the English language. The filter “NOT (animals [MeSH] OR mice OR rats)” was meticulously employed to eliminate studies involving animals. Furthermore, a meticulous examination of the reference lists from pertinent articles was conducted to incorporate additional studies. The specifics of the search strategy can be found in Supplementary Table 1.

2.3 Inclusion and exclusion criteria

Inclusion and exclusion criteria were specified according to the PICOS criteria as follows:

Inclusion Criteria:

Participants: Adults diagnosed with T2DM, aged ≥18 years, recruited from Asia (China, Japan, South Korea, Taiwan), Europe (Germany, Hungary, Poland, Slovakia), North America (United States, Canada, Mexico), Latin America (Chile, Peru), and Oceania (New Zealand).

Intervention: GKAs were given for at least 12 weeks.

Comparison: Placebo or other antidiabetic drugs.

Outcomes: The primary outcomes included changes in glycemic control (e.g., HbA1c, FPG, PPG, pancreatic beta-cell function, lipid profile, serum uric acid, body weight, liver or kidney function and safety profiles (e.g., hypoglycemia events; adverse events).

How to choose appropriate study design: Randomized controlled trials (RCTs)

Exclusion Criteria:

(1) Non-RCTs.

(2) Clinical diagnosis of type 1 diabetes; history of any serious cardiovascular events during the six months preceding the study; any malignancy; poorly controlled hypertension

(3) Study duration less than 12 weeks.

(4) Data limited or lacking for safety and efficacy endpoints.

2.4 Data extraction

The process of data extraction was conducted in accordance with established inclusion and exclusion criteria. The extracted data encompassed various study characteristics, including type and design, as well as baseline characteristics of the population, such as sample size, sex, age, body mass index (BMI), and initial weight. Additionally, laboratory measurements were recorded both at baseline and post-intervention, including HbA1c, fasting plasma glucose (FPG), postprandial glucose (PPG), homeostasis model assessment of insulin resistance (HOMA-IR), homeostasis model assessment-β (HOMA-β), lipid profile, liver function tests, and renal function tests. Furthermore, the incidence of hypoglycemia and pertinent adverse events were also documented. In instances where a study presented several follow-up periods, the longest duration accompanied by comprehensive data was chosen for analysis. Discrepancies in data extraction were addressed through internal discussions or, when appropriate, by referencing the existing literature.

2.5 Quality assessment

The risk of bias for each study was assessed independently using Cochrane Handbook for systematic reviews of interventions criteria. We assessed bias on several domains, including random sequence generation, allocation concealment and blinding of participants and outcome assessors. Studies were categorized as “low risk,” “uncertain risk,” or “high risk.” The detailed risk of bias is displayed in Fig. 1, and the risk of bias summary is displayed in Supplementary Fig. 1.

Fig. 1  Risk of bias graph

Risk of bias assessment for 11 included studies. Colors indicate risk levels: green (low), yellow (unclear), red (high).

2.6 Statistical analysis

Stata version 14.0 (StataCorp, College Station, TX, USA) was used for all analyses. Standardized mean differences (SMDs) were calculated for continuous outcomes and risk ratios (RRs), both with 95% confidence intervals (CIs), for dichotomous outcomes. Thresholds for SMD were established, categorizing values greater than 0.2, 0.5, and 0.8 as indicative of small, moderate, and large effect sizes, respectively [8]. A risk ratio (RR) >1 indicates higher risk in the intervention group (e.g., RR = 1.5 represents a 50% increase in risk). A meta-analysis was performed as all trials assessed GKAs in T2DM with outcomes that could be combined (HbA1c, hypoglycemia), meeting Cochrane’s feasibility criteria, while allowing for the investigation of sources of heterogeneity (e.g., GKA subtypes, dosages) and improving precision across a limited number of trials. A random-effects model was chosen in advance to address the expected clinical heterogeneity (such as differences in GKA types and dosages) and methodological diversity among studies, as clearly advised by the Cochrane Handbook for Systematic Reviews of Interventions in situations where heterogeneity is anticipated. To validate the selection of the model, further analyses utilizing fixed-effect models were performed. Heterogeneity was evaluated with the Q test (significance threshold p < 0.10) and I² statistic, interpreted according to Cochrane guidelines: 0–40% (low), 30–60% (moderate), 50–90% (substantial), and 75–100% (considerable) [8]. We employed meta-regression analysis to examine the influence of GKA types and the duration of intervention on alterations in HbA1c levels. To examine the robustness of the findings, a sensitivity analysis was conducted by sequentially excluding each study. Funnel plots were utilized to assess publication bias, complemented by trim-and-fill analysis to account for possible missing studies. Additionally, Begg’s and Egger’s tests were employed, with a p value of less than 0.05 designated as the threshold for statistical significance.

3. Results

3.1 Search results and study characteristics

This comprehensive meta-analysis included 11 independent, double-blind, randomized controlled trials, with a total of 3,854 participants. The study selection process is illustrated in Fig. 2. The included studies involved the following GKAs: AZD1656 [9, 10], Dorzagliatin [11-13], PF-04937319 [14], MK-0941 [15, 16], TTP399 [17], and PF-04991532 [18, 19]. Baseline characteristics of each group are presented in Table 1.

Fig. 2  Flow diagram of the included studies

PRISMA flow diagram of identification (n = 465 records), screening (n = 283), full-text review (n = 27) and inclusion (n = 11 studies).

Table 1 Baseline characteristics of included studies

Author, year	Duration of follow-up	Treatment group	Number of participants	Age (years)	Male (%)	Baseline HbA1c (%)	Type of GKAs	Geographic Distribution
Kiyosue 2013	4 months	AZD1656 10–80 mg	56	55 ± 10	85.7	8.9 ± 1	Dual-acting GKAs	Japan
		AZD1656 20–140 mg	58	55 ± 9	86.2	8.6 ± 0.9
		AZD1656 40–200 mg	55	55 ± 9	87.3	8.7 ± 1
		PBO	55	57 ± 9	83.6	8.4 ± 0.7
Wilding 2013	4 months	AZD1656 20 mg + MET	40	57.4 ± 9.2	47.5	8.2 ± 0.9	Dual-acting GKAs	Europe (Germany, Hungary, Poland, etc.), Latin America (Chile, Mexico, Peru)
		AZD1656 40 mg + MET	52	54.4 ± 9.1	42.3	8.4 ± 0.7
		AZD1656 10–140 mg + MET	91	57.1 ± 9.3	50.5	8.4 ± 0.9
		AZD1656 20–200 mg + MET	93	57.1 ± 8.5	50.5	8.4 ± 0.8
		PBO + MET	88	56.9 ± 9.6	51.1	8.3 ± 0.9
Yang 2022	24 weeks	Dorzagliatin 75 mg bid + MET	382	54.6 ± 10	64	8.3 ± 0.6	Dual-acting GKAs	China (multicenter)
		PBO + MET	385	54.4 ± 9.2	60	8.3 ± 0.6
Zhu 2018	12 weeks	Dorzagliatin 75 mg qd	53	57.6 ± 9.2	51	8.44 ± 0.8	Dual-acting GKAs	China (multicenter)
		Dorzagliatin 100 mg qd	50	56.7 ± 7.7	56	8.27 ± 0.64
		Dorzagliatin 50 mg bid	50	54.9 ± 8.1	69	8.33 ± 0.65
		Dorzagliatin 75 mg bid	49	55.4 ± 7.7	63	8.46 ± 0.67
		PBO	53	54.7 ± 8.5	58	8.39 ± 0.78
Zhu 2022	24 weeks	Dorzagliatin 75 mg bid	310	53.2 ± 9.6	65	8.3 ± 0.7	Dual-acting GKAs	China (multicenter)
		PBO	153	53.5 ± 10	66	8.4 ± 0.7
Meininger 2011	54 weeks	MK0941 10 mg + INS	119	56.2 ± 8.4	44	9.1 ± 1.1	Dual-acting GKAs	North America (USA), Oceania (New Zealand)
		MK0941 20 mg + INS	117	55.7 ± 8.4	51	8.9 ± 0.9
		MK0941 30 mg + INS	117	56.4 ± 8.1	54	9.0 ± 0.9
		MK0941 40 mg + INS	119	56.1 ± 8.6	50	9.0 ± 1.1
		PBO + INS	115	56.4 ± 9.2	50	9 ± 0.9
N. B. Amin 2015	12 weeks	PF-04937319 3 mg	57	48 ± 5.3	49	8.0 ± 1.1	Dual-acting GKAs	Global multicenter (North America, Europe, Asia; 10 countries unspecified)
		PF-04937319 20 mg	54	48 ± 5.9	59	7.9 ± 1.1
		PF-04937319 50 mg	56	47 ± 7.3	59	8.2 ± 1
		PF-04937319 100 mg	56	48 ± 6.3	54	8.3 ± 1.1
		PBO	57	48 ± 6.3	63	8.1 ± 1.1
	12 weeks	PF-04937319 10 mg + MET	60	55 ± 8.0	57	8.0 ± 0.9
		PF-04937319 50 mg + MET	61	56 ± 9.6	61	8.0 ± 1.0
		PF-04937319 100 mg + MET	61	56 ± 9.8	48	7.9 ± 1.0
		PBO + MET	61	55 ± 8	56	7.9 ± 1
Vella 2019	6 months	TTP399 400 mg	50	52.86 ± 10.7	46	8.45 ± 0.75	Hepatoselective GKAs	Unspecified
		TTP399 800 mg	42	576 ± 8.5	52.4	8.34 ± 0.81
		PBO	48	55.7 ± 11.7	59.2	8.41 ± 0.78
NCT00824616	20 weeks	MK-0941 5 or 10 mg tid	34	54.1 ± 10.2	52.9	/	Dual-acting GKAs	Unspecified
		PBO	34	55.5 ± 8.8	55.9	/
NCT01338870	12 weeks	PF-04991532 5 mg bid	49	59.2 ± 6.8	38.8	7.9 ± 0.939	Hepatoselective GKAs	North America (USA, Canada), Europe (Hungary, Slovakia), Latin America (Mexico), East Asia (Taiwan, China)
		PF-04991532 75 mg bid	50	56.1 ± 8.6	58	7.86 ± 0.998
		PF-04991532 150 mg bid	50	57.2 ± 8.7	60	7.93 ± 1.018
		PF-04991532 300 mg bid	52	56 ± 7.7	42.3	8.01 ± 1.064
		PBO	50	55.7 ± 8.8	68	8.11 ± 1.255
NCT01336738	12 weeks	PF-04991532 150 mg	52	55.3 ± 9.9	69.2	8.34 ± 0.906	Hepatoselective GKAs	North America (USA, Canada, Mexico), Europe (Hungary, Slovakia), East Asia (South Korea, Taiwan, China)
		PF-04991532 450 mg	54	55.1 ± 9.3	55.6	8.19 ± 0.947
		PF-04991532 750 mg	53	55.5 ± 7.3	67.9	7.96 ± 1.03
		PBO	53	55.6 ± 8.5	73.6	8.55 ± 1.351

3.2 Efficacy outcomes on glycemic control

3.2.1 HbA1c

All studies included reported alterations in HbA1c levels. The studies were systematically classified into groups of low-dose, medium-dose, and high-dose for analytical purposes, contingent upon the treatment doses of GKAs that were administered. HbA1c meta-analysis results found no differences in the low-dose group (SMD = –0.05%; 95% CI –0.26 to 0.16, p = 0.638; I² = 67.4%). Moderate HbA1c reduction was found in the medium-dose group (SMD = –0.34%; 95% CI –0.50 to –0.17, p < 0.0001; I² = 38.7%), corresponding to a moderate effect size, which could signify a clinically significant enhancement in glycemic regulation. For high-dose group, the decrease of HbA1c was significantly greater (SMD = –2.268; 95% CI: –3.882 to –0.654, p = 0.006; I² = 99.50%). Of these, Zhu et al. [11] and Yang et al. [12] showed a significantly larger reduction in HbA1c than other studies. Upon removing these two outlier studies, the SMD changed to –0.43 (95% CI: –0.62 to –0.24), and heterogeneity significantly decreased to 59.3% (p < 0.0001). Following the removal of outliers, this indicates that the estimates of the intervention effect in both high-dose groups demonstrated enhanced stability and consistency (Fig. 3). The sensitivity analysis, which compared random and fixed-effect models, substantiated the reliability of the medium-dose outcomes (random SMD = –0.34 compared to fixed SMD = –0.35; Supplementary Table 2). In the analysis stratified by selectivity, dual-acting GKAs exhibited a different influence on HbA1c levels (SMD = –0.36%; 95% CI –0.47 to –0.25, p = 0.018; I² = 45.1%) than hepatoselective GKAs (SMD = –0.04%; 95% CI –0.31 to 0.24, p < 0.0001; I² = 74.3%) (Fig. 4). Meta-regression showed a significant association between type of GKAs and effect size (p = 0.039) and found that dual-acting GKAs decreased HbA1c to a greater extent than hepatoselective GKAs by an average 0.27 (95% CI: –0.53 to –0.01). These results illustrate the considerable variation in HbA1c lowering potential among different classes of GKAs. With longer follow-up times, there was a trend towards a decrease in HbA1c for all GKAs groups, though not statistically significant (p = 0.127). The initial funnel plot suggested potential asymmetry (Begg’s p = 0.795; Egger’s p = 0.167, Supplementary Fig. 2). After excluding two extreme outliers, trim-and-fill analysis revealed no missing studies (adjusted SMD = –0.41, 95% CI: –0.63 to –0.19), with improved symmetry (Begg’s p = 0.452; Egger’s p = 0.087) and tight clustering around the adjusted effect (Supplementary Fig. 3), supporting the stability of this meta-analysis.

Fig. 3  The effects of GKAs on HbA1c level

Pooled analysis of HbA1c reduction in patients treated with GKAs vs. placebo (11 studies, random-effects model).

Fig. 4  HbA1c changes from baseline with GKAs treatment versus placebo stratified by selectivity

Subgroup analysis: HbA1c lowering based on selectivity (dual- vs. hepatoselective GKAs).

3.2.2 FBG

An analogous outlier effect was observed with HbA1c in the high-dose group for FPG. Upon conducting the sensitivity analysis, subsequent to the exclusion of Zhu et al. [11] and Yang et al. [12], it was observed that there existed no notable difference between the GKAs and placebo groups (SMD = 0.07 mmol/L; 95% CI: –0.17 to 0.03, p = 0.177; I² = 46.1%) (Supplementary Fig. 4). The small effect size (SMD = 0.07) failed to meet the criterion for clinical significance (SMD > 0.2), indicating a limited effect on the regulation of fasting glucose control.

3.2.3 2-hPPG

Changes in 2-hour PPG levels from baseline to the end of the study for low, medium, and high-dose groups compared with placebo are reported in two studies. Pooling of results using a random-effect meta-analysis showed a significantly larger reduction in PPG levels (SMD = –0.62 mmol/L; 95% CI: –0.80 to –0.43, p < 0.0001; I² = 48.8%) (Supplementary Fig. 5). The sensitivity analysis demonstrated that the exclusion of a single study did not significantly alter the estimate.

3.3 Efficacy outcomes on metabolic parameters

3.3.1 C-peptide and HOMA-based assessments

Two studies reported alterations in C-peptide levels. No significant alterations were observed within the low-dose cohort (SMD = –0.15, 95% CI: –0.59 to 0.30, p = 0.52), in the medium-dose group (SMD = –0.14, 95% CI: –0.40 to 0.11, p = 0.27), or in the high-dose group (SMD = –0.01, 95% CI: –0.43 to 0.41, p = 0.95) according to the results of meta-analysis with low heterogeneity (I² = 29.6%, p = 0.213) (Supplementary Fig. 6).

HOMA-β and HOMA-IR changes were reported in the studies by Zhu et al. [11] and Yang et al. [12]. There was a significant increase in HOMA-β (SMD = 3.39, 95% CI: 3.21 to 3.57, I² = 0%, p < 0.001) and a significant lower HOMA-IR (SMD = –2.23, 95% CI: –3.09 to –1.36, p < 0.001), hinting that GKAs ameliorated insulin resistance. The heterogeneity was high (I² = 97.0%, p < 0.001). Further investigations are necessary to validate these findings and enhance study methodologies to reduce variability.

3.3.2 Lipid profile

Three studies reported varying lengths of intervention about the impact on lipid metabolism. At 24 weeks, total cholesterol (SMD = 0.146, 95% CI: 0.031 to 0.261, p = 0.013) and triglycerides (SMD = 0.352, 95% CI: 0.236 to 0.468, p < 0.001) had also significantly increased. However, changes in total cholesterol and triglycerides at 52 weeks did not differ statistically. At 54 weeks, triglycerides significantly increased again (SMD = 0.214, 95% CI: 0.063 to 0.364, p = 0.005). There were no significant changes in low-density lipoprotein cholesterol [LDL-C] or high-density lipoprotein cholesterol [HDL-C] at any time point (Table 2).

Table 2 Metabolic parameter variations in meta-analysis

Lipid Parameter	Timepoint	Included Studies	SMD	95% CI	p-value	Heterogeneity (I2)	p-value (Heterogeneity)
TC	24 weeks	Wenying Yang 2022, Dalong Zhu 2022	0.146	0.031, 0.261	0.013	0.00%	0.96
	52 weeks	Wenying Yang 2022, Dalong Zhu 2022	0.041	–0.075, 0.156	0.49	0.00%	0.536
	54 weeks	Gary E. Meininger 2011	0.103	–0.081, 0.287	0.271	0.00%	0.393
TG	16 weeks	J. P. H. Wilding 2013	–0.017	–0.226, 0.191	0.871	0.00%	0.388
	24 weeks	Wenying Yang 2022, Dalong Zhu 2022	0.352	0.236, 0.468	<0.001	0.00%	0.402
	52 weeks	Wenying Yang 2022, Dalong Zhu 2022	–0.077	–0.192, 0.038	0.189	0.00%	0.458
	54 weeks	Gary E. Meininger 2011	0.214	0.063, 0.364	0.005	0.00%	0.523
HDL-C	24 weeks	Wenying Yang 2022, Dalong Zhu 2022	–0.009	–0.165, 0.148	0.914	42.60%	0.187
	52 weeks	Wenying Yang 2022, Dalong Zhu 2022	–0.017	–0.301, 0.268	0.909	82.10%	0.018
	54 weeks	Gary E. Meininger 2011	0.122	–0.028, 0.272	0.111	0.00%	0.668
LDL-C	24 weeks	Wenying Yang 2022, Dalong Zhu 2022	–0.066	–0.230, 0.098	0.428	47.30%	0.168
	52 weeks	Wenying Yang 2022, Dalong Zhu 2022	0.018	–0.097, 0.133	0.757	0.00%	0.418
	54 weeks	Gary E. Meininger 2011	–0.12	–0.270, 0.030	0.118	0.00%	0.785
Uric Acid	24 weeks	Wenying Yang 2022, Dalong Zhu 2022	0.45	0.263, 0.637	<0.001	58.40%	0.121
	52 weeks	Wenying Yang 2022, Dalong Zhu 2022	0.061	–0.054, 0.176	0.302	0.00%	0.583
ALT	24 weeks	Wenying Yang 2022, Dalong Zhu 2022	0.316	0.171, 0.461	<0.001	0.331	0.221
	52 weeks	Wenying Yang 2022, Dalong Zhu 2022	–0.069	–0.184, 0.046	0.079	0.00%	0.634
AST	24 weeks	Wenying Yang 2022, Dalong Zhu 2022	0.535	0.418, 0.652	<0.001	0	0.595
	52 weeks	Wenying Yang 2022, Dalong Zhu 2022	–0.103	–0.218, 0.012	0.079	0	0.769
Creatinine	24 weeks	Wenying Yang 2022, Dalong Zhu 2022	–0.105	–0.330, 0.119	0.358	0.714	0.061
	52 weeks	Wenying Yang 2022, Dalong Zhu 2022	0.03	–0.085, 0.145	0.605	0.00%	0.358
eGFR	24 weeks	Wenying Yang 2022, Dalong Zhu 2022	0.107	–0.143, 0.357	0.403	0.768	0.038
	52 weeks	Wenying Yang 2022, Dalong Zhu 2022	0.003	–0.112, 0.118	0.965	0	0.399

3.3.3 Body weight and BMI

Two to four studies reported changes in body weight and BMI. In analyses of different doses, intervention durations, and types of GKAs, only body weight (SMD = 0.18, 95% CI: 0.03 to 0.33, p = 0.017, Supplementary Fig. 7) and BMI (SMD = 0.19, 95% CI: 0.04 to 0.34, p = 0.016, Supplementary Fig. 8) significantly increased in studies by Zhu et al. [11] and Yang et al. [12] using dorzagliatin. The findings indicate there were no significant changes observed in body weight or BMI across the various intervention duration groups or dosage groups. Although statistically significant, the effect sizes for weight gain (SMD < 0.2) and BMI growth (SMD < 0.2) lacked clinical significance according to Cohen’s criteria. Nevertheless, even minor fluctuations in weight may warrant clinical vigilance in individuals with obesity-related comorbidities.

3.3.4 Uric acid, liver, and kidney function

This meta-analysis analyzed the alterations in uric acid, liver function, and kidney function based on varied intervention durations. Uric acid was significantly elevated at 24 weeks (SMD = 0.450, 95% CI: 0.263 to 0.637, p < 0.001) while it had no significant differences at 52 weeks (SMD = 0.061, 95% CI: –0.054 to 0.176, p = 0.302). For liver function, both alanine aminotransferase (ALT) and aspartate aminotransferase (AST) significantly increased at 24 weeks (ALT: SMD = 0.316, 95% CI: 0.171 to 0.461, p < 0.001; AST: SMD = 0.535, 95% CI: 0.418 to 0.652, p < 0.001), but no significant difference was noted at 52 weeks (ALT: SMD = –0.069, 95% CI: –0.184 to 0.046, p = 0.634; AST: SMD = –0.103, 95% CI: –0.218 to 0.012, p = 0.079). No significant changes were observed in creatinine or eGFR at any time point (Table 2).

3.4 Safety outcomes

3.4.1 Hypoglycemia

This meta-analysis assessed the risk of hypoglycemia among low, medium, and high-dose groups. Results showed that the risk of hypoglycemia was not significant in the low-dose group (RR = 1.09 95% CI: 0.81 to 1.48, I² = 0%, p = 0.553). A significant increase in hypoglycemia risk was observed in the medium-dose group (RR = 1.50, 95% CI: 1.12 to 2.02, I² = 0%, p = 0.006) and high-dose group (RR = 1.63, 95% CI: 1.23 to 2.15, I² = 0%, p = 0.001, Fig. 5). Fixed-effect values corresponded with random-effects models (Supplementary Table 3), validating dose-dependent hypoglycemia risks under homogeneity assumptions (I² = 0%). In high-dose subgroups, AZD1656 exhibited a marked risk elevation (RR = 9.11, 95% CI: 1.11 to 74.96, Table 3), corresponding to 811 additional events per 1,000 patients (10% baseline risk). This significant risk is especially alarming in susceptible populations (e.g., the elderly, individuals with cardiovascular diseases), necessitating caution with high-dose regimens and the consideration of safer alternatives. MK-0941 posed a clinically significant (RR = 1.52, 95% CI: 1.13 to 2.06; 52 additional events/1,000 patients) risk and might warrant more intensive monitoring (e.g., increased frequency of glucose monitoring, education on recognizing symptoms, dose adjustments) in at-risk populations. Subgroup analysis based on intervention course indicated that, if intervention duration was longer than 12 weeks, the risk of hypoglycemia were significantly higher in the medium-dose group (RR = 1.52, 95% CI:1.12 to 2.05, p = 0.007) and the high-dose group (RR = 1.79, 95% CI:1.06 to 3.03, p = 0.029) (Fig. 6). Nonetheless, there was no significant hypoglycemia risk of Dorzagliatin, PF-04937319, TTP399 and PF-04991532. Funnel plot symmetry (Begg’s p = 0.452; Egger’s p = 0.295) and trim-and-fill adjustment (imputing 3 studies, adjusted RR = 1.37 [95% CI: 1.16 to 1.62] vs. original RR = 1.40 [1.18 to 1.65]) demonstrated robustness against publication bias (Supplementary Fig. 9). The preserved effect direction, overlapping confidence intervals, and homogeneity across studies (I² = 0.0%) provided triangulated evidence for dose-response associations.

Fig. 5  The incidence of hypoglycemia of difference groups

Comparative hypoglycemia rates in GKA-treated vs. placebo groups.

Table 3 The incidence of hypoglycemia in different groups under varying intervention drugs

Dose Group	Intervention Drug	RR	95% CI	p-value	I2 (Heterogeneity)
Low Dose	Dorzagliatin	7	0.37–132.29	0.194	0.00%
	MK-0941	1.1	0.80–1.50	0.568	0.00%
	PF-04937319	0.69	0.14–3.36	0.646	0.00%
	TTP399	0.19	0.01–3.90	0.283	0.00%
	PF-04991532	1.02	0.07–15.87	0.992	0.00%
	AZD1656	6.44	0.27–154.70	0.251	0.00%
Middle Dose	Dorzagliatin	7.41	0.39–139.97	0.182	0.00%
	MK-0941	1.5	1.11–2.03	0.009	0.00%
	PF-04937319	1.01	0.21–4.89	0.992	0.00%
	PF-04991532	1.03	0.16–6.91	0.973	0.00%
High Dose	Dorzagliatin	4.47	0.78–25.64	0.093	0.00%
	MK-0941	1.52	1.13–2.06	0.006	0.00%
	PF-04937319	2.35	0.62–8.88	0.207	0.00%
	TTP399	1.71	0.30–9.77	0.544	0.00%
	PF-04991532	0.6	0.08–4.83	0.635	0.00%
	AZD1656	9.11	1.11–74.96	0.04	0.00%

Fig. 6  The incidence of hypoglycemia in different groups under varying intervention durations

Hypoglycemia risk stratified by therapeutic dose and treatment duration.

3.4.2 Other adverse events

Dizziness and headache were adverse events observed at all dosage levels. (RR = 1,655, 95% CI 1.073 to 2.552, p = 0.023). There were no statistically significant differences in upper respiratory tract safety events or urinary tract infections between the treatment and control groups within any of the dose group.

4. Discussion

A meta-analysis was conducted to thoroughly assess the efficacy and safety of GKAs in treating T2DM, considering dose-response relationships, mechanistic classifications (dual-acting versus hepatoselective), and temporal metabolic adaptations. Our studies reveal three key conclusions: First, high-dose GKAs produced substantial reductions in HbA1c, characterized by considerable variability, necessitating the initiation of lower dosages and a gradual rise while assessing efficacy and safety. Second, dual-acting drugs (e.g., dorzagliatin) demonstrate a higher level of glycemic control compared to hepatoselective subtypes, likely through the synergistic activation of pancreatic β-cell glucose sensing and the suppression of hepatic glucagon-mediated glucose production. Finally, short-term metabolic disruption—24-week weight gain and triglyceride elevation—were stabilized by 52 weeks, challenging prior assumptions of irreversible GKA-induced dyslipidemia. Nonetheless, the risk of hypoglycemia escalated in a dose-dependent manner (high-dose RR = 1.63), particularly with AZD1656 and MK-0941, underscoring the necessity for personalized dosage thresholds in high-risk groups.

According to the 2025 American Diabetes Association (ADA) recommendations [20], dipeptidyl peptidase-4 (DPP-4) inhibitors, in conjunction with SGLT2 inhibitors, continue to be the second-line standard regimens for combination therapy with metformin, particularly in patients with associated cardiovascular or renal conditions. For populations with progressive beta-cell functional decline, the guidelines highlight the necessity for “pathophysiology-based treatment strategies.” This study demonstrates that dual-acting GKAs preserve beta-cell function, surpassing current therapies due to their capacity to rectify glucose-sensing deficiencies and enhance insulin secretion. Dual-target mechanisms of these agents surpass limitations labelled to classical single-target agents such as DPP-4 inhibitors and SGLT2 inhibitors.

This meta-analysis shows that different doses of GKAs can be used to reduce HbA1c in a dose-dependent manner and the significant reduction in high-dose group. This dose-dependent effect is strongly associated with the activation potency of glucokinase activators (GKAs) on glucokinase (GK) in pancreatic β-cells and the liver. As the central sensor of glucose metabolism, GK regulates glucose-stimulated insulin secretion (GSIS) in the pancreas, while hepatocyte GK governs glycogen synthesis and breakdown [21-23]. Patients with type 2 diabetes commonly exhibit impaired GK function, leading to insufficient insulin secretion and increased hepatic glucose output [24]. GKAs ameliorate these pathological defects by activating GK [25].The dose-dependent effect of HbA1c improvement has been supported by several studies. For instance, Wilding et al. showed that HbA1c was significantly reduced in a dose dependent manner with Dorzagliatin treatment for 12 weeks [10]. In a similar vein, Meininger et al. observed that incorporating MK-0941 into insulin therapy for patients with T2DM resulted in a dose-dependent decrease in HbA1c levels, with the most beneficial outcome noted at a dosage of 30 mg administered three times daily [15]. The findings of these studies indicate a significant correlation between the glucose-lowering effects of GKAs and their administered dosages. Nonetheless, some studies, including Zhu et al. [11] and Yang et al. [12], denoted distinct changes in HbA1c. However, these studies were excluded from the analysis following sensitivity assessments due to significant heterogeneity. This discrepancy could be due to the differences in sample size, baseline characteristics of patients, treatment dose and follow up duration.

In contrast, GKAs can be further divided into dual activators and hepatoselective activators [6]. Dual-acting GKAs work on both the pancreas and liver, activating GK in these organs to improve glucose sensing and stimulate insulin secretion as well as reduce hepatic glucose output for improved glycemic control [5]. In subgroup analysis, a better effect of dual-acting GKAs on HbA1c reduction was observed when compared to hepatoselective GKAs. In vivo, relevant studies performed in Wistar rats, mice, and Gottingen minipigs indicate that hepatoselective GKAs (e.g., TTP399) do not induce hypoglycemia [17]. Furthermore, the clinical study Simplici-T1 illustrated that TTP399, characterized as a hepatoselective GKA, presents a safe and effective therapeutic option for individuals with type 1 diabetes (T1DM), significantly reducing HbA1c levels without inducing hypoglycemia [26]. The results align with previous observations indicating that hepatoselective GKAs exhibit constraints in their efficacy for glucose reduction. Nevertheless, hepatoselective GKAs are poised to offer a more secure pharmacological alternative for individuals who are at heightened risk for hypoglycemia, such as older adults or those with renal impairment.

The findings further illustrate that GKAs across all dosage groups are advantageous in diminishing the degree of PPG, albeit with a modest effect on FPG levels. Research conducted by Yang et al. [12], Meininger [15], and Chen [27] has already established this point. The observed difference can be elucidated by the mechanism through which GKAs primarily reduce blood glucose levels during postprandial hyperglycemia via enhanced insulin secretion, while their impact on glycemic control during the fasting state is comparatively minimal [25].

Dual-acting GKAs, such as Dorzagliatin, show a distinctive advantage in protecting β-cell function. A high-dose intervention significantly enhances HOMA-β and reduces HOMA-IR, all while maintaining stable C-peptide levels. This is closely aligned with the “pathophysiology-based treatment strategy” that the 2025 ADA guidelines recommended. In mechanistic studies, Dorzagliatin was shown to rescue β-cell dysfunction through several mechanisms, including the restoration of GK activity, upregulation of insulin receptor substrate-2 (IRS-2) expression, and promotion of mitochondrial metabolism [28-31]. The findings align with the extensive longitudinal data from the DREAM study [32], which indicated that extended treatment with Dorzagliatin improves β-cell glucose sensitivity and postpones functional decline. Long-term validation of metabolic safety is still not established for GKAs, but dual-target effects indicate a novel pathway for treatment guided by pathophysiology.

GKAs have been shown to preserve β-cell function, but their metabolic effects require careful investigation. This study found considerable metabolic changes after 24 weeks of GKA intervention, including increased total cholesterol, triglyceride levels, and body weight. In animal models, hepatic glucokinase overexpression resulted in not only reduced blood glucose but also highly increased triglyceride and free fatty acid level [33]. In addition, prolonged GKAs use has been reported to be related to a higher risk of non-alcoholic fatty liver disease (NAFLD) [34]. Excessive hepatic glucokinase activation can cause lipid metabolism dysregulation through several mechanisms: (1) Accumulation of glucose-6-phosphate (G6P) activates glycolysis, leading to increased acetyl-CoA production and triglyceride formation through fatty acid synthase (FASN) [35]; (2) Increased insulin secretion inhibits adipose tissue lipolysis and activates adipocyte differentiation [36]. Nonetheless, the elevated dosage of Dorzagliatin at week 24 correlated with statistically significant elevations in uric acid, ALT, and AST, highlighting the potential augmented metabolic strain on renal and hepatic functions.

Significantly, the 52-week follow-up data revealed a stabilization of metabolic parameters, suggesting that extended treatment might enable the body to progressively attain a dynamic equilibrium through intrinsic metabolic adaptations. Conversely, it may also indicate a decline in the effectiveness of GKA over time following the operation [5, 6, 35]. While the 2025 ADA guidelines do not make specific recommendations regarding GKA-induced metabolic risks, their weight management principles and safety monitoring framework provide relevant clinically actionable guidance [20]: (1) Apply intensive lifestyle interventions involving a 500 kcal deficit/d and at least 150 min of moderate-intensity exercise/week; (2) For people with TG ≥5.6 mmol/L, adjunctive therapy with fenofibrate or statins is indicated; (3) When ALT/AST exceeds 3× the upper limit of normal, consider switching to agents with lower hepatotoxicity (e.g., SGLT2 inhibitors).

Hypoglycemia is a major safety concern of GKAs treatment, which can be attributed to the increased action of GK and the effect on the dynamic control of pancreatic β-cell insulin secretion. By augmenting the response of GK to glucose, GKAs also reduce the threshold for insulin secretion resulting in insulin release at lower glucose concentrations, thereby increasing the propensity for hypoglycemia [25, 37]. In instances of elevated doses or significant GK activation, the associated risk may become pronounced due to the upregulation of the regulatory mechanism governing insulin secretion, resulting in a dysregulation of glucose levels [38, 39]. In this meta-analysis, the risk of hypoglycemia was statistically higher in the medium-dose group and high-dose group. The risk increased after 12 weeks of intervention. In comparison to other drugs, the administration of high-dose MK-0941 and AZD1656 markedly increased the likelihood of hypoglycemia. By contrast, Dorzagliatin showed no increased risk of hypoglycemia. This divergence stems from its dual-action mechanism: stimulation of GK in hypothalamic glucose-sensing neurons and the adrenal medulla induces sympathetic activation and resultant secretion of glucagon/adrenaline, both blunt hepatic glucose output while at the same time curbing excessive insulin secretion during hypoglycemia [40]. Other GKAs (i.e., MK-0941/AZD1656) do not possess this central-peripheral regulatory synergy, resulting in unmitigated GK activation and increased hypoglycemia risk [41]. Clinicians, according to the 2025 ADA hypoglycemia management strategy [20], should prioritize risk assessment of GKA-associated hypoglycemia for elderly (≥65 years), renally impaired (eGFR <45 mL/min/1.73 m²), or cognitively compromised patients. For patients on moderate- to high-dose GKAs, continuous glucose monitoring (CGM) is recommended for real-time tracking of glycemia, focusing on nocturnal and preprandial hypoglycemic events to lower the risk of hypoglycemia.

The incidence of dizziness and headache was notably elevated across all dosage groups, while gastrointestinal adverse events were particularly heightened in the medium-dose group. Gastrointestinal adverse events, including nausea, vomiting, abdominal pain, and diarrhea, could be related to metabolic regulation by the agents, but the underlying mechanism has not been well defined. Dizziness and headache mechanisms might reflect transient adaptations of the central nervous system to changes in glucose metabolism [42-44].

This meta-analysis provides insights into the potential role of GKAs in the management of diabetes, but there are some limitations. First, several of the trials were characterized by limited sample sizes, rendering them susceptible to underpowered statistical evaluations. Furthermore, secondary endpoints, such as alterations in beta-cell function, were inadequately reported, thereby increasing the likelihood of selective reporting bias. Second, open-label trial designs (e.g., AZD1656 with metformin) can overestimate efficacy. There are significant geographic and ethnic biases: Dorzagliatin studies were almost exclusively in Chinese populations, MK-0941 and PF-04937319 studies were largely on European/North American cohorts with little data on African or Middle Eastern populations. The median length of study (≤24 weeks) prevents evaluation of long-term cardiovascular safety, long-term durability of beta-cell function, and risk of NAFLD. Moreover, losing effectiveness over time seen with agents such as AZD1656 and MK-0941 is still mechanistically unresolved. There are no relevant head-to-head comparisons between dual-acting and hepatoselective GKAs to guide clinical choice of the most effective agent.

5. Conclusion

The research illustrates the efficacy of dual-acting GKAs, such as Dorzagliatin, in managing hyperglycemia by enhancing beta-cell function and hepatic glucose metabolism, a benefit that seems particularly significant for individuals exhibiting marked beta-cell dysfunction. Conversely, hepatoselective GKAs (e.g., TTP399) are safer for elderly or high-risk populations given their low hypoglycemia risk and minimal metabolic impact. Although there were short-term metabolic changes (e.g., increased body weight, elevated patient triglycerides), data from longer-term follow-up indicate that compensatory mechanisms may limit negative consequences. The mechanisms of GKAs, grounded in pathophysiologic considerations, represent an innovative strategy in diabetes management that aligns with contemporary guidelines emphasizing personalized treatment and risk evaluation through dose-dependent efficacy-safety trade-offs (Graphical Abstract). Future efforts should: (1) prolong the duration of follow-up to ≥5 years to establish cardiovascular and hepatic metabolic safety; (2) expand ethnic diversity in trials to improve generalizability; (3) compare dual-acting versus hepatoselective GKAs in head-to-head trials to further elucidate subtype-specific advantages; (4) explore combinations with existing agents (e.g., SGLT2 inhibitors) to define metabolic trade-offs.

Graphical Abstract 

Ethics Approval and Consent to Participate

This is a systematic review and meta-analysis; ethics approval and consent to participate are not applicable.

Consent for Publication

Not applicable. This study does not involve human participants.

Availability of Data and Materials

All data generated or analyzed during this study are included in this published article and its supplementary information files.

Conflict of Interest

The authors declare that they have no conflict of interest.

Funding

No Funding.

References

• 1   American Diabetes  Association (2020) Addendum. 9. pharmacologic approaches to glycemic treatment: standards of medical care in diabetes-2020. Diabetes Care 43: 1979.
• 2   Teck  J (2022) Diabetes-associated comorbidities. Prim Care 49: 275–286.
• 3   Nauck  MA,  Wefers  J,  Meier  JJ (2021) Treatment of type 2 diabetes: challenges, hopes, and anticipated successes. Lancet Diabetes Endocrinol 9: 525–544.
• 4   Agius  L (2009) Targeting hepatic glucokinase in type 2 diabetes: weighing the benefits and risks. Diabetes 58: 18–20.
• 5   Li  W,  Zhang  X,  Sun  Y,  Liu  Z (2020) Recent clinical advances of glucokinase activators in the treatment of diabetes mellitus type 2. Pharmazie 75: 230–235.
• 6   Toulis  KA,  Nirantharakumar  K,  Pourzitaki  C,  Barnett  AH,  Tahrani  AA (2020) Glucokinase activators for type 2 diabetes: challenges and future developments. Drugs 80: 467–475.
• 7   Hodson  L,  Brouwers  M (2018) Non-alcoholic fatty liver disease concerns with glucokinase activators. Lancet Diabetes Endocrinol 6: 684–685.
• 8   Cohen  J (2013) Statistical power analysis for the behavioral sciences. Routledge, New York, USA.
• 9   Kiyosue  A,  Hayashi  N,  Komori  H,  Leonsson-Zachrisson  M,  Johnsson  E (2013) Dose-ranging study with the glucokinase activator AZD1656 as monotherapy in Japanese patients with type 2 diabetes mellitus. Diabetes Obes Metab 15: 923–930.
• 10   Wilding  JP,  Leonsson-Zachrisson  M,  Wessman  C,  Johnsson  E (2013) Dose-ranging study with the glucokinase activator AZD1656 in patients with type 2 diabetes mellitus on metformin. Diabetes Obes Metab 15: 750–759.
• 11   Zhu  D,  Li  X,  Ma  J,  Zeng  J,  Gan  S, et al. (2022) Dorzagliatin in drug-naïve patients with type 2 diabetes: a randomized, double-blind, placebo-controlled phase 3 trial. Nat Med 28: 965–973.
• 12   Yang  W,  Zhu  D,  Gan  S,  Dong  X,  Su  J, et al. (2022) Dorzagliatin add-on therapy to metformin in patients with type 2 diabetes: a randomized, double-blind, placebo-controlled phase 3 trial. Nat Med 28: 974–981.
• 13   Zhu  D,  Gan  S,  Liu  Y,  Ma  J,  Dong  X, et al. (2018) Dorzagliatin monotherapy in Chinese patients with type 2 diabetes: a dose-ranging, randomised, double-blind, placebo-controlled, phase 2 study. Lancet Diabetes Endocrinol 6: 627–636.
• 14   Amin  NB,  Aggarwal  N,  Pall  D,  Paragh  G,  Denney  WS, et al. (2015) Two dose-ranging studies with PF-04937319, a systemic partial activator of glucokinase, as add-on therapy to metformin in adults with type 2 diabetes. Diabetes Obes Metab 17: 751–759.
• 15   Meininger  GE,  Scott  R,  Alba  M,  Shentu  Y,  Luo  E, et al. (2011) Effects of MK-0941, a novel glucokinase activator, on glycemic control in insulin-treated patients with type 2 diabetes. Diabetes Care 34: 2560–2566.
• 16   Merck Sharp & Dohme  LLC (2015) A study to test MK-0941 in adults with type 2 diabetes mellitus with inadequate glycemic control on insulin (MK-0941-018). https://clinicaltrials.gov/study/NCT00824616 accessed on April 25, 2024.
• 17   Vella  A,  Freeman  JLR,  Dunn  I,  Keller  K,  Buse  JB, et al. (2019) Targeting hepatic glucokinase to treat diabetes with TTP399, a hepatoselective glucokinase activator. Sci Transl Med 11: eaau3441.
• 18  Pfizer (2013) Study of safety and efficacy of PF-04991532 in subjects with type 2 diabetes. https://clinicaltrials.gov/study/NCT01338870 accessed on April 25, 2024.
• 19  Pfizer (2013) Study of safety and efficacy of PF-04991532 in subjects with type 2 diabetes mellitus. https://clinicaltrials.gov/study/NCT01336738 accessed on April 25, 2024.
• 20   Shankar  P,  Jali  M (2025) Standards of care in diabetes-2025. RJMS 15: 4–7.
• 21   Sternisha  SM,  Miller  BG (2019) Molecular and cellular regulation of human glucokinase. Arch Biochem Biophys 663: 199–213.
• 22   Coghlan  M,  Leighton  B (2008) Glucokinase activators in diabetes management. Expert Opin Investig Drugs 17: 145–167.
• 23   Agius  L (2008) Glucokinase and molecular aspects of liver glycogen metabolism. Biochem J 414: 1–18.
• 24   Haeusler  RA,  Camastra  S,  Astiarraga  B,  Nannipieri  M,  Anselmino  M, et al. (2015) Decreased expression of hepatic glucokinase in type 2 diabetes. Mol Metab 4: 222–226.
• 25   Matschinsky  FM (2013) GKAs for diabetes therapy: why no clinically useful drug after two decades of trying? Trends Pharmacol Sci 34: 90–99.
• 26   Klein  KR,  Freeman  JLR,  Dunn  I,  Dvergsten  C,  Kirkman  MS, et al. (2021) The simpliciT1 study: a randomized, double-blind, placebo-controlled phase 1b/2 adaptive study of TTP399, a hepatoselective glucokinase activator, for adjunctive treatment of type 1 diabetes. Diabetes Care 44: 960–968.
• 27   Chen L  ZJ,  Yang  R,  Feng  L (2021) Glucokinase activator dorzagliatin (HMS5552) regulates GLP-1 release in T2D patients and is synergistic with sitagliptin and empagliflozin in optimizing beta-cell function. Diabetes 70 (Suppl 1): 117-LB.
• 28   Chow  E,  Wang  K,  Lim  CKP,  Tsoi  STF,  Fan  B, et al. (2023) Dorzagliatin, a dual-acting glucokinase activator, increases insulin secretion and glucose sensitivity in glucokinase maturity-onset diabetes of the young and recent-onset type 2 diabetes. Diabetes 72: 299–308.
• 29   Futamura  M,  Yao  J,  Li  X,  Bergeron  R,  Tran  JL, et al. (2012) Chronic treatment with a glucokinase activator delays the onset of hyperglycaemia and preserves beta cell mass in the Zucker diabetic fatty rat. Diabetologia 55: 1071–1080.
• 30   Oh  YS,  Lee  YJ,  Park  K,  Choi  HH,  Yoo  S, et al. (2014) Treatment with glucokinase activator, YH-GKA, increases cell proliferation and decreases glucotoxic apoptosis in INS-1 cells. Eur J Pharm Sci 51: 137–145.
• 31   Shirakawa  J,  Togashi  Y,  Sakamoto  E,  Kaji  M,  Tajima  K, et al. (2013) Glucokinase activation ameliorates ER stress-induced apoptosis in pancreatic β-cells. Diabetes 62: 3448–3458.
• 32   Zeng  J,  Gan  S,  Mi  N,  Liu  Y,  Su  X, et al. (2023) Diabetes remission in drug-naïve patients with type 2 diabetes after dorzagliatin treatment: a prospective cohort study. Diabetes Obes Metab 25: 2878–2887.
• 33   O’Doherty  RM,  Lehman  DL,  Télémaque-Potts  S,  Newgard  CB (1999) Metabolic impact of glucokinase overexpression in liver: lowering of blood glucose in fed rats is accompanied by hyperlipidemia. Diabetes 48: 2022–2027.
• 34   Zhang  Z,  Ji  G,  Li  M (2023) Glucokinase regulatory protein: a balancing act between glucose and lipid metabolism in NAFLD. Front Endocrinol (Lausanne) 14: 1247611.
• 35   Agius  L (2014) Lessons from glucokinase activators: the problem of declining efficacy. Expert Opin Ther Pat 24: 1155–1159.
• 36   Berger  E,  Géloën  A (2022) Insulin prevents fatty acid induced increase of adipocyte size. Adipocyte 11: 510–528.
• 37   Davis  EA,  Cuesta-Muñoz  A,  Raoul  M,  Buettger  C,  Sweet  I, et al. (1999) Mutants of glucokinase cause hypoglycaemia- and hyperglycaemia syndromes and their analysis illuminates fundamental quantitative concepts of glucose homeostasis. Diabetologia 42: 1175–1186.
• 38   Matschinsky  FM (2009) Assessing the potential of glucokinase activators in diabetes therapy. Nat Rev Drug Discov 8: 399–416.
• 39   Nakamura  A,  Terauchi  Y (2015) Present status of clinical deployment of glucokinase activators. J Diabetes Investig 6: 124–132.
• 40   Wu  Y,  Wang  K,  Su  J,  Liu  X (2024) Efficacy and safety of dorzagliatin, a novel glucokinase activators, in the treatment of T2DM: a meta-analysis of randomized controlled trials. Medicine (Baltimore) 103: e36916.
• 41   Ren  Y,  Li  L,  Wan  L,  Huang  Y,  Cao  S (2022) Glucokinase as an emerging anti-diabetes target and recent progress in the development of its agonists. J Enzyme Inhib Med Chem 37: 606–615.
• 42   Cavestro  C,  Rosatello  A,  Micca  G,  Ravotto  M,  Marino  MP, et al. (2007) Insulin metabolism is altered in migraineurs: a new pathogenic mechanism for migraine? Headache 47: 1436–1442.
• 43   Rainero  I,  Limone  P,  Ferrero  M,  Valfrè  W,  Pelissetto  C, et al. (2005) Insulin sensitivity is impaired in patients with migraine. Cephalalgia 25: 593–597.
• 44   Islam  MR,  Nyholt  DR (2022) Glucose-related traits and risk of migraine—a potential mechanism and treatment consideration. Genes (Basel) 13: 730.