Berberine derivative lowered HbA1c by 1% in 12-week diabetes trial

A 12-week randomized trial published in March 2025 in JAMA Network Open found that HTD1801, an oral compound pairing berberine with ursodeoxycholic acid (UDCA), reduced hemoglobin A1c by 1.0 percentage point compared to placebo in adults with type 2 diabetes. The higher of two tested doses also improved fasting glucose, insulin sensitivity, LDL cholesterol, and liver enzyme levels. Insulin secretion did not increase. The tolerability profile caught the attention of the trial’s independent commentators.

HTD1801 is a first-in-class gut-liver anti-inflammatory metabolic modulator developed by HighTide Therapeutics. It combines berberine, a plant alkaloid used for centuries in traditional Chinese and ayurvedic medicine, with UDCA, a bile acid already prescribed for gallstone dissolution and primary biliary cholangitis. Berberine shares several mechanisms with metformin: AMP-activated protein kinase (AMPK) activation, inhibition of gluconeogenesis, and modulation of gut microbiota. But berberine’s clinical use has been held back by poor oral bioavailability. Attaching UDCA improves absorption and adds anti-inflammatory effects through NLRP3 inflammasome inhibition, a pathway metformin does not touch.

How the study was designed

The phase 2 trial enrolled 113 adults with type 2 diabetes who were predominantly overweight. Ji and colleagues at Peking University People’s Hospital led the study across 14 other Chinese centers. Participants were randomized to one of three arms: HTD1801 500 mg twice daily, HTD1801 1000 mg twice daily, or placebo. The primary endpoint was change in HbA1c from baseline to week 12.

The study population was 97.3 percent Han Chinese. Both the authors and an accompanying invited commentary flagged this homogeneity as a limitation. Mean baseline HbA1c sat around 8.0 percent. Participants had mild liver disease at most, with low prevalence of hepatic steatosis at enrollment.

What the trial found

At the 1000 mg twice-daily dose, the least-squares mean reduction in HbA1c was 1.0 percentage point greater than placebo. Fasting plasma glucose dropped by 20.5 mg/dL. Postprandial glucose at 0.5 hours fell by 23.0 mg/dL. Insulin resistance, measured by HOMA-IR, improved. In the high-dose group, 55.9 percent of patients reached an HbA1c below 7.0 percent, against 15.2 percent on placebo. Of those on the higher dose, 29.4 percent got below 6.5 percent, compared to 6.1 percent with placebo.

The glucose-lowering did not come from pushing the pancreas harder. Fasting insulin and C-peptide levels stayed flat, which points to improved insulin sensitivity as the driver. This separates HTD1801 from sulfonylureas and from other agents that raise insulin output.

Beyond glucose: lipid and liver effects

The metabolic changes went past glycemic control. Total cholesterol fell by 15.1 mg/dL and LDL cholesterol by 11.2 mg/dL in the 1000 mg group, each compared to placebo. Liver transaminases (ALT, AST, and gamma-glutamyltransferase) all declined. Most participants had normal liver enzyme levels at baseline, so the drop is notable. It fits the drug’s design as a gut-liver axis modulator and suggests a possible role in the metabolic dysfunction that frequently comes with type 2 diabetes, including metabolic dysfunction-associated steatotic liver disease.

The trial did not detect weight change or improvements in liver steatosis on imaging. The commentary authors pointed out steatosis prevalence was low at baseline: not much room to show a reduction in 12 weeks.

Safety and tolerability

GI side effects are the historical knock on berberine, as they are for metformin. In this trial, the tolerability numbers were clean. One participant reported nausea. One reported diarrhea. Both were at the lower dose. A single retinal hemorrhage occurred in the treatment group, which investigators judged unrelated to the study drug. Hypoglycemia was rare and never severe.

Nestoras Mathioudakis, an endocrinologist at Johns Hopkins who wrote the invited commentary, said the low GI-effect rate is one reason HTD1801 could find a role. “For certain patient populations, berberine derivatives like HTD1801, if proven effective and safe in larger-scale trials, could offer advantages over metformin, such as fewer gastrointestinal adverse effects, safety for individuals with kidney impairment or liver conditions, and benefits beyond glycemic control.”

Where it fits clinically

The glucose-lowering potency of HTD1801 at 1000 mg twice daily lands in the range of DPP-4 inhibitors and SGLT2 inhibitors. It is a bit less than what metformin monotherapy achieves at equivalent doses. A 2010 meta-analysis by Sherifali and colleagues found that metformin 1000 mg twice daily typically cuts HbA1c by 1.5 to 2.0 percentage points over similar time frames.

What HTD1801 offers, Mathioudakis argued, is a broader set of metabolic changes from a single pill. Metformin does not do much for LDL cholesterol or liver enzymes in most patients. HTD1801 does both. Its twin mechanism, AMPK activation plus NLRP3 inhibition, is different from any diabetes drug currently on the market. “As an innovative drug candidate, HTD1801 exhibits a unique dual mechanism of action that is distinct from any existing T2DM drugs on the market,” said Dr. Linong Ji, the trial’s lead investigator and former Vice President of the International Diabetes Federation.

Phase 3 and what happens next

Two phase 3 trials have since read out. In April 2025, HighTide Therapeutics announced that SYMPHONY 1 and SYMPHONY 2 both met their primary endpoints. SYMPHONY 1, a monotherapy trial in 408 patients, found a 1.3 percentage point placebo-adjusted HbA1c reduction at 24 weeks. SYMPHONY 2, testing HTD1801 on top of metformin in 551 patients, reported 1.2 percentage points. Patients who started with more severe disease did better: those with baseline HbA1c at or above 8.5 percent saw drops of 1.5 and 1.6 percentage points.

Gated secondary endpoints were also met: more patients hit HbA1c below 7.0 percent, and LDL cholesterol, GGT, and high-sensitivity C-reactive protein all moved in the right direction. Fewer than 2 percent of patients stopped treatment because of side effects. HighTide plans to submit a New Drug Application to China’s National Medical Products Administration this year.

What to watch

For a drug derived from a natural product, the data are better than the category usually produces: a well-conducted phase 2 in a major journal, an independent commentary that endorsed further study, and two phase 3 trials that confirmed the phase 2 result. But the gaps are worth naming. The phase 2 population was almost entirely Han Chinese. Twelve weeks is short. Questions about durability beyond 24 weeks, weight effects over years, and cardiovascular outcomes are unanswered. The drug has not gone head-to-head against metformin, a GLP-1 receptor agonist, or an SGLT2 inhibitor.

Mathioudakis closed his commentary with a call for “larger, longer-term trials across more diverse patient populations” and for studies testing HTD1801 in combination with existing therapies. If China’s NMPA accepts the NDA and the drug reaches the market, post-marketing surveillance and real-world evidence will start filling those gaps. Right now, HTD1801 is the most clinically advanced berberine derivative in the diabetes pipeline. No berberine capsule from a supplement aisle will match its absorption profile. And no existing oral diabetes drug hits both AMPK and the NLRP3 inflammasome at the same time.

Consult your doctor before starting, stopping, or changing any diabetes medication.

References

1. Ji L, Ma J, Ma Y, et al. Berberine ursodeoxycholate for the treatment of type 2 diabetes: a randomized clinical trial. JAMA Netw Open. 2025;8(3):e2462185. https://doi.org/10.1001/jamanetworkopen.2024.62185
2. Ai X, Yu P, Peng L, et al. Berberine: a review of its pharmacokinetics properties and therapeutic potentials in diverse vascular diseases. Front Pharmacol. 2021;12:762654. https://doi.org/10.3389/fphar.2021.762654