Potential of duodenal mucosal resurfacing in achieving glycemic control in Asians with type 2 diabetes

Type 2 diabetes pathophysiology and incretin secretion in Asians

Approximately 60% of patients living with diabetes reside in Asia, the main focus of the current worldwide diabetes mellitus epidemic. ¹ Type 2 diabetes in Asians is characterized as having more β-cell dysfunction, impairment of insulin secretion, lower acute phase insulin response but with better insulin sensitivity. Among East Asians, the levels of post-prandial incretin hormones, glucagon-like peptide 1 (GLP-1) and glucose-dependent insulinotropic peptide (GIP) are comparable between persons with type 2 diabetes compared to healthy individuals. ² This is in stark contrast to Caucasians where the incretin effect consistently decreases from 10% to 40% among persons with type 2 diabetes from 50% to 70% in healthy individuals with normal glucose tolerance. On one hand, the incretin effect addresses insulin secretion and potentiation following an oral glucose tolerance test. On the other hand, a new and emerging concept called the gastrointestinal-mediated glucose disposal (GIGD) pertains to the role of the gastrointestinal tract in regulating glucose homeostasis. Despite having no differences in terms of the incretin effect between persons with type 2 diabetes and healthy individuals, following an oral glucose load, the GIGD decreases by half to 30% among those with type 2 diabetes compared to 60% in persons with healthy individuals. Hence, among Asians diagnosed with type 2 diabetes, both the incretin secretion and incretin effect are not decreased but rather they present with a reduced GIGD. ³ This would then mean that the incretin effect is not exactly just a summation of GLP-1 and GIP effects, but rather, other factors such as pancreatic hormones, neuroendocrine signals, glucose utilization, and insulin sensitivity contribute to the overall GIGD. ³

Duodenal mucosal resurfacing

Duodenal mucosal resurfacing (DMR) works on the assumption that resurfacing the mucosal interface will restore the normal mucosal surface allowing for a reset and correction of any abnormal signals emanating from the region and thus, resultant improvements in β-cell function and glucose homeostasis. ⁴ In brief, this minimally invasive technique performed via upper endoscopy with direct visualization is as follows: while at the target duodenal segment, the submucosa is first lifted with saline injection to separate it from the underlying muscularis layer (circumferential mucosal lift). This provides a thermally protective layer for the underlying structures below. After successful homogeneous lifting, hydrothermal ablation lasting for approximately 10 s each with temperatures reaching approximately 90°C are performed on the mucosal layer. The goal is to be able to perform up to five longitudinal ablations over a duodenal segment length of 9–10 cm starting at the post-papillary region up to the proximal area of the ligament of Treitz. Re-epithelialization of the ablated mucosal surface is expected to occur within days following the procedure. This procedure is currently performed under general anesthesia with a duration of approximately an hour or less. Over the next 24 h following DMR, patients are able to resume an oral diet with careful diet progression over the next 10–14 days without any form of caloric restriction.

There are currently three published human studies utilizing the Revita DMR system in type 2 diabetes, two prospective (one single-center study by Rajagopalan et al., ⁴ one international multi-center study by Van Baar et al.,^5,6 and one randomized-controlled trial by Mingrone et al. ⁷ ). In addition, there is one completed but unpublished study as well as one ongoing international multicenter randomized trial (Table 1) [ClinicalTrials.gov identifiers: NCT03653091 and NCT04419779]. The first of these is a human proof of concept study that was conducted in 2013 in Santiago, Chile. ⁴ This 6-month safety and efficacy showed an HbA1c reduction of 1.2% at 6 months (p < 0.001). This was followed by the REVITA-1 study, a multicenter study from 2015 to 2017 in seven endoscopy centers in Europe and South America with a 2-year follow-up period.^5,6 HbA1c levels were reduced from 8.5 ± 0.7% to 7.5 ± 0.8% at 6 months (p < 0.001); HbA1c reduction was sustained throughout the 24 months post-procedure at 7.5 ± 1.1% (p < 0.001). Lastly, the REVITA-2 is a double-blind, superiority randomize-controlled trial conducted from 2017 to 2018 in Europe and Brazil. ⁷ In this trial, the Asian population is grossly underrepresented with only 2/108 (1.9%) compared to 71/108 (66%) of Caucasians. ⁷ Results of the 108 enrolled patients were stratified according to region because of heterogeneity. In the European subgroup, 24 weeks post DMR, median (IQR) HbA1c change was −2.8 (3.8%) versus −2.5 (3.1%) post-sham (p = 0.033). However, the Brazilian subgroup results only trended toward benefit but was not statistically significant. Furthermore, those with high baseline fasting glucose of more than 180 mg/dL had significantly greater reductions in HbA1c post-procedure (p = 0.002). In all three published studies, the procedure was deemed safe as adverse events were largely mild and transient in nature. More recent evidence from analyses of the REVITA-1 ⁶ and REVITA-2 ⁷ cohorts demonstrated that fasting insulin, glucagon, C-peptide decreased significantly while indices of insulin sensitivity and beta cell function all improved substantially. However, levels of both GLP-1 and GIP remained unchanged opening the possibility that incretin changes are unlikely responsible for the improved glycemic control following DMR. ⁸

Table 1.

Summary of published studies and ongoing clinical trials on DMR.

Year published	Country	Duration	Participants	Design	Outcomes
20164	Chile	August–December 2013	N = 44 Adults with type 2 diabetes aged 28–75 years, BMI 24–40 kg/m2, and HbA1c 7.5–12% on at least one oral antidiabetic agent	Prospective, single-arm, open-label, single center	6 months post-DMR Efficacy: HbA1c reduction at 6 months 1.2% in the full cohort (p < 0.001) More potent glycemic effects in long duodenal segment cohort, 2.5% reduction in mean HbA1c 3 months after versus 1.2% in the short duodenal segment group (p < 0.05) and a 1.4% reduction at 6 months versus 0.7% in the short duodenal segment group (p = 0.3) Safety: DMR well tolerated with minimal gastrointestinal symptoms following the procedure Three patients experienced duodenal stenosis treated successfully by balloon dilation
20195,6	Belgium Chile Italy Netherlands United Kingdom	March 2015–September 2019	Adult with type 2 diabetes within the last 10 years aged 28–75 years, BMI 24–40 kg/m2, HbA1c of 7.5–10.0%, and stable diabetes treatment with at least one oral glucose-lowering drug for more than 3 months at enrollment	Prospective, single-arm, open-label, multi-center	2 years post-DMR Efficacy: Mean ± SD HbA1c level reduction, 8.5 ± 0.7% at baseline (N = 34) to 7.5 ± 0.8% at 6 months (p < 0.001); reduction sustained through 24 months post-DMR (7.5 ± 1.1%, p < 0.001) Greater than 50% of patients, glucose-lowering therapy was reduced or unchanged ALT decreased from 38.1 ± 21.1 U/L at baseline to 32.5 ± 22.1 U/L at 24 months (p = 0.048) HDL-C and TG/HDL ratios improved during 24-months of follow-up Safety: No device- or procedure-related serious adverse events, unanticipated device effects, or hypoglycemic events were noted between 12 and 24 months
20227	Belgium Chile Italy Netherlands United Kingdom	March 2017–December 2019	N = 109 Adults with type 2 diabetes aged 28–75 years, HbA1c 7.5–10.0%, BMI ⩾ 24 and ⩽40 kg/m2 on one or more oral glucose lowering medications of which one must be metformin, with no changes in dose or medication in the previous 12 weeks prior to enrollment	Prospective, randomized, double-blind, sham-controlled, multi-center study Active comparator: DMR Sham comparator: DMR sham	24-weeks post-DMR Efficacy: median (IQR) HbA1c change −3.1 (3.9%) in DMR group versus −2.8 (3.7%) in sham group (p = 0.147) Patients with baseline liver MRI-PDFF >5% (DMR n = 48; sham n = 43), 12-week post-DMR liver-fat change was −5.4 (5.6%) in DMR group versus −2.9 (6.2%) in sham group (p = 0.096) Heterogeneity between European (DMR n = 39; sham n = 37) and Brazilian (DMR n = 17; sham n = 16) populations (p = 0.063) European subgroup: median (IQR) HbA1c change −2.8 (3.8%) versus −2.5 (3.1%) post-sham (p = 0.033); 12-week post-DMR liver-fat change −5.4 (6.1%) versus −2.2 (4.3%) post-sham (p = 0.035) Brazilian subgroup: only trended toward DMR benefit in HbA1c, but not liver fat, in context of a large sham effect Patients with high baseline fasting plasma glucose ⩾180 mg/dL had significantly greater reductions in HbA1c post-DMR versus sham treatment (p = 0.002) Safety: most adverse events were mild and transient
Completed [ClinicalTrials.gov identifier: NCT03653091]	United States	September 2018–August 2020	N = 9 Men and non-pregnant women with type 2 diabetes for at least 3 years, aged 28–65 years, HbA1c of 7.5–9.5%, BMI ⩾28 and ⩽40 kg/m2, maintained on two to three oral agents with no changes in medication in the 12 weeks prior	Prospective, randomized, double-blind, sham-controlled, multi-center study Active comparator: DMR Sham comparator: DMR sham	Unpublished
Ongoing [ClinicalTrials.gov identifier: NCT04419779]	United States Belgium France Ireland Italy Netherlands Spain Switzerland United Kingdom	March 2021–2025 (estimated)	N = 506 Male, and non-pregnant, non-lactating females with type 2 diabetes aged 21–70 years, HbA1c 7.5–9.5%, fasting plasma glucose ⩾180 to <270 mg/dL and body mass index ⩾24 to ⩽40 kg/m2 on stable dose of metformin and up to 2 oral agents requiring a minimum of 20 units up to a maximum of 60 units of basal insulin	Prospective, randomized, double-blind, sham-controlled, multi-center study Active comparator: DMR Sham comparator: DMR sham	Ongoing

ALT, alanine aminoTransferase; BMI, body mass index; DMR, duodenal mucosal resurfacing; HbA1c, glycosylated hemoglobin; HDL-C, high-density lipoprotein cholesterol; IQR, interquartile range; MRI-PDFF, Magnetic Resonance Imaging-Proton Density Fat Fraction. SD, standard deviation; TG, triglyceride.

Currently accepted clinical indications for DMR based on the latest published literature are adults 28–75 years of age, poorly controlled type 2 diabetes, HbA1c ranging from 7.5% to 10.0%, BMI range of 24–40 kg/m², on stable diabetes regimen with at least one oral agent for at least 3 months. However, these parameters and anthropometric measurements were largely based on European and South American population data. Current available evidence remains limited as all published studies are either from animal data or from human studies conducted in Europe and South America; none of these have focused on Asians nor conducted in Asian countries. The published randomized trials conducted only have limited number of patients enrolled limiting its statistical power in terms of efficacy and safety as well the its generalizability to patients with type 2 diabetes mellitus. Long-term, large volume, and additional randomized controlled trials are still needed to clearly understand the efficacy, safety, and role of DMR.