Berberine Ursodeoxycholate for the Treatment of Type 2 Diabetes: A Randomized Clinical Trial.

Introduction

Type 2 diabetes (T2D) is one of the most common metabolic disorders worldwide, with its rapid increase in prevalence in recent decades closely linked to the obesity epidemic.1 Its prevalence has been estimated to be 10.5% globally and 11.2% in China, with China having the largest number of patients with T2D globally.2,3 T2D is primarily caused by defective insulin secretion by pancreatic β cells and the inability of insulin-sensitive tissues to respond appropriately, leading to a disruption in glucose homeostasis.4 In an excessive nutritional state, hyperglycemia and hyperlipidemia are often present in patients with T2D, further promoting insulin resistance and chronic inflammation. Furthermore, T2D is frequently comorbid with chronic metabolic abnormalities including hypertension or metabolic dysfunction–associated steatotic liver disease (MASLD), which can exacerbate T2D and lead to a worse prognosis with increased risk of mortality and cardiovascular outcomes.5,6

Although there are multiple classes of hypoglycemic drugs approved, China’s overall effectiveness in the control of T2D remains insufficient, with only 49.4% of individuals achieving a hemoglobin A1c (HbA1c) level less than 7.0%.3 Hence, there is a significant demand for antidiabetic medications that can comprehensively address metabolic risk factors beyond glycemic control.

Berberine ursodeoxycholate (HTD1801), a first-in-class orally administered gut-liver anti-inflammatory metabolic modulator, provides a unique dual mechanism of action (adenosine monophosphate [AMP] kinase activation and NLRP3 inflammasome inhibition) compared with the current therapeutic landscape. HTD1801 is an ionic salt of ursodeoxycholic acid and berberine, representing a new molecular entity that offers the possibility of combination therapy for chronic metabolic and nonviral liver diseases in a single treatment. Berberine primarily acts as an AMP kinase activator, which is expected to promote the utilization of glucose and free fatty acids, therefore decreasing inflammation and improving insulin resistance. Ursodeoxycholic acid, a key regulator of bile acid and inflammatory pathways, has been used for the treatment of hepatobiliary disorders and cholestatic conditions. Due to its unique structure and the interaction of these 2 moieties, HTD1801 exhibits enhanced and novel pharmacologic effects and physicochemical characteristics compared with the individual moieties.7

HTD1801 has been evaluated in multiple clinical studies across metabolic and nonviral liver diseases. In patients with presumed metabolic dysfunction–associated steatohepatitis (MASH) and T2D, HTD1801 treatment for 18 weeks showed improvements in liver fat, markers of liver fibrosis, liver injury, body weight, and lipids.8 Despite 93% of patients taking medications for the treatment of T2D (including glucagon-like peptide-1 receptor agonists), a significant additional reduction in the HbA1c level was observed with HTD1801.9 Based on these findings, a study evaluating the efficacy and safety of HTD1801 was conducted in patients with T2D, which was inadequately controlled with diet and exercise.

Methods

This phase 2 double-blind, placebo-controlled randomized clinical trial was conducted at 14 sites (universities and hospitals) in China between March 2022 and January 2023. The study was designed in collaboration with local partners, and the trial protocol (Supplement 1) was approved by the ethics committees of all participating centers. All participants provided written informed consent. This study was conducted in accordance with the Declaration of Helsinki10 and regulations consistent with the International Conference on Harmonization guidelines for Good Clinical Practice.11 This study followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline.

Key inclusion criteria included a diagnosis of T2D (World Health Organization criteria12), having undergone 8 or more weeks of diet and exercise consistent with standard of care,13 having a fasting plasma glucose (FPG) level of less than 250.5 mg/dL (to convert to millimoles per liter, multiply by 0.0555), and having an HbA1c level between 7.0% and 11.0% (to convert to proportion of total hemoglobin, multiply by 0.01). Among those patients who participated in the trial, self-reported ethnicity categories included Han Chinese and Hui Chinese. These ethnicities were included as part of baseline demographics to allow for the evaluation of subgroup analyses on response. To stabilize glucose, eligible patients participated in a 4-week, single-blind placebo run-in period and received guidance on lifestyle modification, medication, and procedures for self-monitoring of blood glucose (standard fingerstick blood glucose monitor and patient diary). Laboratory results were collected 5 to 7 days prior to randomization to confirm eligibility, for which patients had an HbA1c level of 7.0% to 10.5% and an FPG level of less than 250.5 mg/dL. One protocol revision occurred during the study; no changes were made to eligibility criteria following study initiation.

Assuming an HbA1c level change from baseline to week 12 of −0.4% with placebo, −1.1% with HTD1801 (Shenzhen HighTide Biopharmaceutical Ltd) 500 mg twice daily, and −1.4% with HTD1801 1000 mg twice daily, the pooled SD was 1.1%, and 2-sided α = .05.14 A sample size of 81 patients was estimated to provide 90% power to detect a statistical difference between HTD1801 1000 mg twice daily and placebo. Accounting for an approximate 20% dropout rate, the planned sample size was 33 patients per group, or 99 total. Sample size analyses were performed using SAS, version 9.4 (SAS Institute Inc).

Randomization was performed with an interactive web response system (Clinflash). A block randomization method was used, with liver fat content (assessed by controlled attenuation parameter ≥274 dB/m or <274 dB/m) and HbA1c level (<8.5% or ≥8.5%) as stratification factors. Patients were randomized 1:1:1 to receive HTD1801 500 mg twice daily (hereinafter referred to as the 500-mg group), HTD1801 1000 mg twice daily (hereinafter referred to as the 1000-mg group), or placebo. Treatments were provided as matching capsules. All patients and study and sponsor personnel were blinded to treatment assignment.

The primary end point was the HbA1c level change from baseline to week 12. Secondary end points included the change from baseline to week 12 in glycemic markers, liver fat content (using controlled attenuation parameter [FibroScan, Echosens]), liver biochemistry, body weight, lipids, and safety. Following baseline, visits occurred every 4 weeks. The HbA1c level was collected after baseline at weeks 8 and 12. Samples were collected in a fasted state with the exception of postprandial samples, which occurred after a standardized meal. Homeostatic model assessment for insulin resistance was calculated using standard formulas.15 Treatment-emergent adverse events were coded according to the Medical Dictionary for Regulatory Activities, version 25.1,16 and severity was graded using the Common Terminology Criteria for Adverse Events grading criteria.17 According to T2D treatment guidelines in China, hypoglycemia was defined as a blood glucose level of 70.3 mg/dL.13

Descriptive statistics were used to summarize continuous data. Counts and proportions were used for categorical data. The full analysis set included all randomized patients who received at least 1 dose of the study drug. The safety set included all randomized patients who received at least 1 dose of the study drug and underwent at least 1 safety assessment. The primary end point was analyzed using a mixed-effects model for repeated measures, with the HbA1c level change from baseline as the dependent variable. Treatment group, measurement time point, and interaction between treatment group and measurement time point were independent variables. Randomization stratification factors were covariates, and patients were included as a random effect. Between-group differences were judged based on model-adjusted least-squares mean. In a sensitivity analysis of the primary end point, missing data were handled using the last observation carried forward approach, and the analysis was conducted using an analysis of covariance (ANCOVA) model.

For secondary analyses, the primary estimation method was analyzed using a mixed-effects model for repeated measures. Only 1 postbaseline sample (week 12) was collected for controlled attenuation parameter and high-sensitivity C-reactive protein; therefore, an ANCOVA model was used for analysis. The ANCOVA model included treatment, randomization stratification factors, and the corresponding baseline measurement. For continuous variables, missing data were not routinely imputed, whereas categorical variables were imputed using the nonresponder imputation method.

Comparisons between active treatment groups and the placebo group were tested using an estimation based on the statistical model (mixed-effects model for repeated measures or ANCOVA) of the 95% CI of the difference in the least-squares mean between groups (treatment vs placebo). The least-squares mean and 95% CI in each group were estimated, and a 2-sided P value = .05 for between-group comparison was provided. Testing of secondary analyses and the associated P values reported are considered nominal. Analyses were performed using SAS, version 9.4.

Results

Of the 129 patients with T2D who entered the run-in period, 113 (mean [SD] age, 54.3 [10.6] years; 41 female [36.3%] and 72 male [63.7%]) were randomized. Among these patients, the mean (SD) HbA1c level was 8.2% (0.8%); body mass index, 25.5 (3.7), calculated as weight in kilograms divided by height in meters squared; and FPG level, 160.7 (38.3) mg/dL. Baseline demographic and clinical characteristics were balanced among treatment groups (Table 1). Patients had other features of metabolic dysfunction, including low-density lipoprotein cholesterol near the upper limit of normal and elevated triglycerides. A total of 113 patients were treated and contributed to both the efficacy and safety populations (placebo group, n = 38; 500-mg group, n = 37; and 1000-mg group, n = 38). Among these patients, 110 (97.3%) were Han Chinese, and 3 (2.7%) were Hui Chinese. Overall, 110 patients (97.3%) completed the study, and no patients discontinued due to an adverse event (Figure 1).

Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); CAP, controlled attenuation parameter; GGT, γ-glutamyltransferase; HbA1c, hemoglobin A1c; HDL-C, high-density lipoprotein cholesterol; hs-CRP, high-sensitivity C-reactive protein; HTD1801, berberine ursodeoxycholate; LDL-C, low-density lipoprotein cholesterol.

SI conversion factors: To convert HbA1c level to proportion of total hemoglobin, multiply by 0.01; fasting and postprandial plasma glucose levels to millimoles per liter, multiply by 0.0555; fasting and postprandial insulin levels to picomoles per liter, multiply by 6.945; fasting C-peptide and postprandial C-peptide levels to nanomole per liter, multiply by 0.331; total cholesterol, HDL-C, non–HDL-C, and LDL-C levels to millimoles per liter, multiply values by 0.0259; triglyceride level to millimoles per liter, multiply by 0.0113; total bilirubin level to micromoles per liter, multiply values by 17.104; ALT, AST, and GGT levels to microkatals per liter, multiply by 0.0167; and hs-CRP level to milligrams per liter, multiply by 10.

Data are presented as mean (SD) unless otherwise noted.

HTD1801 indicates berberine ursodeoxycholate.

The primary end point was achieved with significant dose-dependent reductions in the HbA1c level after 12 weeks of treatment with HTD1801 compared with placebo (Figure 2A). The least-squares mean (SE) change from baseline in the HbA1c level for the placebo group was −0.3% (0.1%); for the 500-mg group, it was −0.7% (0.1%), and for the 1000-mg group, it was −1.0% (0.1%). The least-squares mean difference was −0.4% (95% CI, −0.79% to −0.03%; P = .04) for the 500-mg group and −0.7% (95% CI, −1.10% to −0.35%; P < .001) for the 1000-mg group compared with the placebo group. Improvement in the HbA1c level was observed with both HTD1801 treatment groups by week 8 and continued to improve with no plateau evident at week 12. A sensitivity analysis performed to address missing data was consistent with the primary analysis (eTable in Supplement 2).

In this full analysis set (N = 113), least-squares means are derived from a mixed-effects model for repeated measures, with the measured value or change from baseline as a dependent variable; the treatment group, measurement time point, and interaction between treatment group and measurement time point as independent variables; and randomized stratified controlled attenuation parameter (≥274 dB/m or <274 dB/m), hemoglobin A1c (HbA1c) level (<8.5% or ≥8.5%), and patient baseline measurements as covariates. HTD1801 indicates berberine ursodeoxycholate. To convert HbA1c level to proportion of total hemoglobin, multiply by 0.01; to convert fasting plasma glucose level to millimoles per liter, multiply by 0.0555.

The number of patients achieving target HbA1c level thresholds (<7.0% and <6.5%) increased with dose and treatment duration, with twice as many patients in the 1000-mg group achieving an HbA1c level less than 6.5% at week 12 compared with week 8 (Figure 2B). At week 12, treatment with 1000 mg twice daily resulted in a greater proportion of the 34 patients achieving both HbA1c levels (<7.0%: 19 patients [55.9%] and <6.5%: 10 patients [29.4%]) compared with the 33 patients receiving placebo (<7.0%: 5 [15.2%] and <6.5%: 2 [6.1%]).

Consistent with reductions in the HbA1c level, HTD1801 was associated with improved markers of glucose metabolism and insulin resistance. Dose-dependent mean (SD) reductions in the FPG level were observed with both HTD1801 dosages (500-mg group, −13.0 [38.2] mg/dL; 1000-mg group, −18.4 [21.8] mg/dL) compared with placebo (0.3 [50.9]) (Table 2); reductions were observed by week 4 and sustained through 12 weeks of treatment (Figure 2C).

Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; CAP, controlled attenuation parameter; GGT, γ-glutamyltransferase; HbA1c, hemoglobin A1c; HDL-C, high-density lipoprotein cholesterol; HOMA-IR, homeostatic model assessment for insulin resistance; hs-CRP, high-sensitivity C-reactive protein; HTD1801, berberine ursodeoxycholate; LDL-C, low-density lipoprotein cholesterol; LS, least-squares; NA, not applicable.

SI conversion factors: To convert HbA1c level to proportion of total hemoglobin, multiply by 0.01; fasting and postprandial plasma glucose levels to millimoles per liter, multiply by 0.0555; fasting and postprandial insulin levels to picomoles per liter, multiply by 6.945; fasting C-peptide and postprandial C-peptide levels to nanomole per liter, multiply by 0.331; total cholesterol, LDL-C, HDL-C, and non–HDL-C levels to millimoles per liter, multiply values by 0.0259; triglyceride level to millimoles per liter, multiply by 0.0113; hs-CRP level to milligrams per liter, multiply by 10; ALT, AST, and GGT levels to microkatals per liter, multiply by 0.0167; and total bilirubin level to micromoles per liter, multiply values by 17.104.

Unless otherwise specified, the analysis was performed using a mixed-effects model for repeated measures, with the change from baseline as a dependent variable; the treatment group, measurement time point, and interaction between treatment group and measurement time point as independent variables; and a randomized stratified CAP (≥274 dB/m or <274 dB/m), HbA1c level (<8.5% or ≥8.5%), and patient baseline measurements as covariates. P values were nominal.

Dose-dependent reductions in 0.5-hour postprandial plasma glucose level and homeostatic model assessment for insulin resistance were also observed with HTD1801 compared with placebo (Table 2 and eFigure 1 in Supplement 2). HTD1801 resulted in dose-dependent reductions in fasting C-peptide level; however, no differences were observed between HTD1801 and placebo in fasting insulin, the change in 0.5-hour postprandial insulin, or 0.5-hour postprandial C-peptide levels (Table 2).

HTD1801 treatment resulted in dose-dependent reductions in low-density lipoprotein cholesterol and non–high-density lipoprotein cholesterol levels by week 4 through week 12 compared with no change or increases with placebo (eFigure 2 in Supplement 2). Dose-dependent reductions in triglyceride levels were also observed with HTD1801. The high-density lipoprotein cholesterol level remained unchanged from baseline across all treatment groups.

As shown in Table 2, the high-sensitivity C-reactive protein level, a marker of inflammation and cardiovascular risk, decreased in all treatment groups at week 12. Body weight remained stable across treatment groups throughout the study (Table 2).

Despite most patients having liver biochemistry within the normal range at baseline, HTD1801-treated patients had reductions in alanine aminotransferase, aspartate aminotransferase, and γ-glutamyltransferase levels by week 4 (eFigure 3 in Supplement 2). Improvements across liver biochemistries with 1000 mg twice daily maintained a downward trend through week 12. Reductions in total bilirubin level were observed in all groups by week 4 (Table 2). No meaningful changes in controlled attenuation parameter were observed in any group.

Treatment with HTD1801 was generally safe and well tolerated. Treatment-emergent adverse events, generally mild in severity, occurred in 15 patients (39.5%) in the placebo group, 17 (46.0%) in the 500-mg group, and 27 (71.1%) in the 1000-mg group, totaling 59 patients (52.2%) overall. The most common treatment-emergent adverse event was hyperlipidemia, which occurred in 3 patients each in the 500-mg and 1000-mg groups and in 2 patients who received placebo (Table 3). One patient randomized to the 1000-mg group experienced an event of hypoglycemia, which was mild in severity and considered unrelated to study drug. The occurrence of nausea (placebo) and diarrhea was rare, with only 1 instance of each reported during the study (both in the 1000-mg group).

Abbreviations: HTD1801, berberine ursodeoxycholate; TEAE, treatment-emergent adverse event.

Medical Dictionary for Regulatory Activities, version 25.1 was used for adverse event coding.16

Four patients experienced severe treatment-emergent adverse events, including hypertriglyceridemia (placebo group), retinal hemorrhage (500-mg group), hyperlipidemia (1000-mg group), and increased blood creatine phosphokinase level (1000-mg group). All other treatment-emergent adverse events were mild to moderate in severity. The incidence of treatment-related treatment-emergent adverse events was similar across groups, all mild to moderate in severity (placebo group, n = 5; 500-mg group, n = 2; and 1000-mg group: n = 6).

One placebo-treated patient required rescue therapy during the treatment period. A serious adverse event of retinal hemorrhage occurred in 1 patient in the 500-mg group. The event was considered unlikely related to treatment but rather associated with uncontrolled hypertension and T2D. No clinically significant abnormalities were observed in laboratory tests, vital signs, physical examinations, or 12-lead electrocardiograms in any group.

Discussion

In this phase 2 placebo-controlled randomized clinical trial, treatment with HTD1801 (500 mg twice daily and 1000 mg twice daily) resulted in significant reductions in HbA1c levels compared with placebo at 12 weeks. Additional improvements were observed in glycemic, cardiometabolic, and liver-related parameters, demonstrating the comprehensive benefit of HTD1801 for the treatment of T2D and its comorbidities. These findings support HTD1801 as a well-tolerated oral treatment option that could be used alone or in combination with other available therapies for T2D.

HTD1801 is a first-in-class, orally administered gut-liver anti-inflammatory metabolic modulator that, upon ingestion, dissociates into the 2 active moieties (berberine and ursodeoxycholic acid). Nonclinical data indicate improved berberine absorption in the liver when administered as HTD1801, suggesting its potential for improved efficacy. Furthermore, HTD1801 shows improved solubility compared with the parent moieties. These data are indicative of greater bioavailability, potentially resulting in improved efficacy compared with ursodeoxycholic acid or berberine alone.7

HTD1801 is active through multiple pathways addressing the core aspects of metabolic disease. Berberine activates AMP kinase, inhibits the NLRP3 inflammasome, stimulates the insulin receptor, and induces the low-density lipoprotein receptor.18,19 Specific to improved glucose metabolism, activation of the AMP kinase pathway modulates a wide variety of interrelated pathways that regulate energy utilization and may result in improved insulin sensitivity.18,20 These pathways include inhibition of gluconeogenesis, stimulation of hepatic fatty acid oxidation and glucose uptake by skeletal muscle, increased expression of the insulin receptor, and inhibition of cholesterol and triglyceride synthesis and de novo lipogenesis.21,22,23 Additionally, inhibition of the NLRP3 inflammasome helps regulate an obesity-induced inflammatory state, therefore improving glucose homeostasis and insulin resistance.24,25 Ursodeoxycholic acid helps to ameliorate insulin resistance, promotes insulin secretion, and exhibits anti-inflammatory and antioxidative stress effects.26,27,28,29 This unique combination of actions supports HTD1801 as a new treatment option for chronic metabolic and liver diseases.

T2D often coexists with chronic metabolic abnormalities that can exacerbate T2D and lead to a worse prognosis with increased risk of mortality and cardiovascular outcomes.5,6 Although there are available treatment options for T2D, many patients still have an incomplete response to treatment.3 Additionally, few of the available therapies comprehensively address disease burden beyond glycemic control. Injectable glucagon-like peptide-1 receptor agonists are increasingly recommended due to their cardiometabolic benefits. However, injectable treatments impose a significant burden on patients with poor long-term adherence. Oral glucagon-like peptide-1 receptor agonists are currently under development to circumvent the burden of injections but thus far have not demonstrated a cardiometabolic benefit beyond improved glycemic control.30,31

This study provides evidence that HTD1801 can treat the core aspects of metabolic syndrome. The mechanistic pathway has been associated with improvements in glucose metabolism, insulin resistance, lipid metabolism, and hepatic inflammation, providing a comprehensive treatment platform for the multifaceted nature of metabolic diseases. HTD1801 treatment for 12 weeks resulted in significant dose-dependent and clinically meaningful reductions in the HbA1c levels that were driven by improvements in both FPG and postprandial glucose levels. HTD1801 decreased homeostatic model assessment for insulin resistance in a dose-dependent manner with no impact on fasting or postprandial 0.5-hour insulin secretion, indicating that the glucose-lowering effect of HTD1801 might be attributed to improved insulin sensitivity rather than direct stimulation of insulin secretion. When these observations are combined with improvements in lipids (low-density lipoprotein cholesterol and non–high-density lipoprotein cholesterol levels) and inflammation (high-sensitivity C-reactive protein level), they suggest an overall improvement in the metabolic profile and a potential reduction in cardiovascular risk.

Body weight remained stable in all groups throughout the study. However, in a previous 18-week study on presumed MASH and T2D, in which the study population had a baseline body weight of 30 kg greater than the current study, a 3.5-kg reduction was observed after treatment with HTD1801 1000 mg twice daily.8

Liver diseases of metabolic origin (eg, MASLD) and T2D frequently coexist, leading to overall greater risk of disease progression and clinical outcomes.5,6 In patients with T2D, up to 56% will also have MASLD or simple steatosis.32 There is evidence that T2D is one of the strongest risk factors contributing to a faster progression of MASLD, with up to 30% of patients developing MASH, a chronic, life-threatening inflammatory liver disease.33,34 Despite the present study population having evidence of mild liver disease, both HTD1801 groups had improvements in markers of liver injury (alanine aminotransferase and aspartate aminotransferase levels) and γ-glutamyltransferase level, which, based on enzyme activity related to atherosclerosis and plaque formation, may be associated with cardiovascular risk.35,36 This supports the potential of HTD1801 to reduce underlying liver damage, potentially prevent development of MASLD or progression to MASH, and reduce cardiovascular risk. Although a reduction in controlled attenuation parameter was not observed, this could be attributed to the relatively mild liver steatosis in this study. In an 18-week, placebo-controlled study on presumed MASH and T2D and a liver fat content of 10% or more, a significant 4.8% reduction (magnetic resonance imaging–derived proton density fat fraction) was observed after treatment with HTD1801 1000 mg twice daily.8 Considering the pleiotropic effects of HTD1801 across hepatic and cardiometabolic parameters, HTD1801 holds promise for the treatment of T2D alongside other comorbid metabolic diseases.

HTD1801 was found to be safe and well tolerated with treatment-emergent adverse events that were generally mild in severity. There was excellent treatment adherence with no discontinuations due to adverse events. In previous studies of HTD1801, the most common treatment-emergent adverse events were mild to moderate gastrointestinal-related adverse events (primarily diarrhea and nausea), which increased with higher doses.8,37 In this study, the occurrence of nausea and diarrhea was infrequent. Hypoglycemia, a major limiting factor in intensive glycemic control, occurred in 1 patient in the 1000-mg group.

This study has some limitations. It was only 12 weeks in duration, and a longer treatment duration is needed to further assess the long-term benefit of HTD1801. Additionally, the study population was ethnically homogeneous (97.3% of patients self-reported as Han Chinese); however, previous studies that have included other ethnicities have demonstrated similar benefits with HTD1801 treatment.8,38

Conclusions

This placebo-controlled randomized clinical trial of HTD1801 found that improvements across key glycemic, cardiometabolic, and liver-related parameters support HTD1801 as a potential novel oral treatment option for T2D. HTD1801 treatment provides a spectrum of therapeutic effects that appear to address comorbid conditions that exacerbate disease and worsen prognosis of patients with T2D. These findings are being studied in ongoing trials on MASH39 and T2D.40,41