Radiofrequency vapor ablation for duodenal mucosal ablation in the treatment of type 2 diabetes: results from the first-in-human pilot study

Abstract

Background and Aims

Duodenal mucosal ablation (DMA) is an emerging endoscopic treatment modality for type 2 diabetes (T2D). The radiofrequency vapor ablation (RFVA) system (Aqua Medical Inc, Pleasanton, Calif, USA) is a single-use, through-the-scope, circumferential ablation catheter under investigation for DMA. In this pilot study, we assessed the safety, tolerability, procedural feasibility, and initial efficacy of RFVA.

Methods

We conducted a first-in-human, prospective, single-center trial enrolling 27 patients with poorly controlled T2D (glycated hemoglobin [HbA1c], 7.5%-10%) despite ≥1 oral antidiabetic drug. DMA was performed with the patient under anesthesia in several dose titrations as follows: safety cohort (180 J once, n = 2), first treatment cohort (180 J twice, n = 11), and second treatment cohort (200 J twice; n = 14). Primary outcomes were safety (number of serious adverse events [SAEs]), tolerability (based on a visual analog scale [VAS] pain score), feasibility (procedure time), and initial efficacy (change in HbA1c at 4, 12, and 24 weeks).

Results

Twenty-seven patients with a mean T2D duration of 6 years (standard deviation [SD], 2.6) underwent DMA with 100% technical success. Mean patient age was 54 years (SD, 6.6), 48% were men, and baseline HbA1c was 8.6% (SD, 0.6). Mean procedure time was 49.7 minutes (SD, 14.3) and catheter time 34.5 minutes (SD, 10.9). There were no SAEs, and the maximum mean VAS pain score was 1.4 (SD, 2.1) on day 2. We observed a change in HbA1c at 4 weeks (–0.9%), 12 weeks (–1.2%), and 24 weeks (–0.8%) after the procedure.

Conclusions

DMA using the RFVA system is a simple through-the-scope procedure that appears safe, well tolerated, and feasible for the treatment of T2D. (Clinical trial registration number: NCT05887635.)

Type 2 diabetes (T2D) is one of the most significant and fast-growing health challenges of the 21st century. In 2021, 529 million people were living with diabetes, with figures expected to rise to 1.31 billion by 2050.¹ The upsurge in T2D is primarily driven by the rising rates of obesity, with an estimated 90% of patients with T2D being overweight or obese.² This is concerning given that diabetes is a chronic, multisystem disease associated with significant long-term comorbidity.³ Furthermore, despite an increase in the number of available glucose-lowering agents, including glucagon-like peptide-1 receptor agonists (GLP-1RAs), many patients do not attain or maintain adequate glycemic control.

Bariatric surgery, originally proposed for the treatment of severe obesity, was found to be highly effective as a metabolic treatment for T2D. The reason for these profound metabolic benefits is still under scientific debate but appears to involve calorie restriction, alterations in satiety by postprandial neurohumoral signals, increase in beta cell mass, improvement in insulin production, and gut microbiome adaptation.^4,5 Through bariatric surgery, the proximal small bowel has been highlighted as a critical metabolic center because of the immediate and weight-independent effects on glycemic control. In addition, animal models and clinical human studies have demonstrated that the small-intestinal mucosa becomes abnormal and hypertrophied in response to chronic hyperglycemia.^6,7

The profound metabolic effects after bariatric surgery together with the endoscopic accessibility of the duodenum have enabled pursuit of nonpharmacologic, nonsurgical treatments for T2D. This new endoscopic paradigm is collectively known as “metabolic endoscopy,” which targets the duodenal mucosa to mimic the antidiabetic effect of bypass surgery. One of these techniques is duodenal mucosal ablation (DMA), which involves selective, reversible damage to the duodenal mucosa to improve glycemic control. To date, 2 technologies have been explored for this indication: the Revita system (Fractyl Laboratories Inc, Lexington, Mass, USA) and the ReCET system (Endogenex Inc, Plymouth, Minn, USA). These are over-the-guidewire devices that require fluoroscopy for duodenal catheter placement. Revita is a single-use, balloon-based system that requires submucosal injection followed by hydrothermal ablation in a process termed duodenal mucosal resurfacing.⁸ ReCET is a nonthermal system that unfolds in the duodenum to deliver electroporation, resulting in mucosal cell death.⁹ Given the safety and initial efficacy of these 2 systems, we hypothesized that the circumferential radiofrequency vapor ablation (RFVA) system (Aqua Medical Inc, Pleasanton, Calif, USA) could be used for DMA as a primary treatment for patients living with T2D.

The RFVA system is a catheter-based, through-the-scope (TTS) device that does not require fluoroscopy or submucosal lifting and delivers thermal energy to coagulate gastrointestinal (GI) tissue. Thermal energy is delivered to target tissue in the form of heated water vapor that is generated using radiofrequency (RF) energy passing through saline as it flows over a bipolar electrode positioned within the lumen and at the distal tip of the catheter. Vapor is a high-temperature (100^oC), high-energy (540 cal/g) medium resulting in short (<5 seconds) application and causing mucosal damage without significantly affecting the submucosa or muscularis propria.¹⁰ In this first-in-human pilot study, we investigated the use of RFVA for DMA in a cohort of patients living with T2D with inadequate control despite the use of oral antidiabetic drugs (OADs). Our aim was to assess a series of stepwise dose escalations to determine safety, tolerability, and procedural feasibility to help guide the design of future trials using RFVA. Additionally, we sought to observe early efficacy in terms of glycated hemoglobin (HbA1c) reduction to inform sample size calculations and power estimates for larger clinical trials.

Methods

Study design

This was a pilot, first-in-human, nonrandomized, prospective, open-label study to investigate the safety, tolerability, and procedural feasibility of RFVA for DMA among patients with uncontrolled T2D despite OADs (NCT05887635). The study was conducted at a single center in Huechuraba, Santiago, Region Metropolitana Chile (Clinica Colonial). An initial sample size of 30 patients, allowing for a 10% dropout rate, was chosen to establish the appropriate dose, safety, and tolerability through a series of dose escalations. The early efficacy assessment was preliminary and intended to inform the methodology and design of larger clinical trials given the small sample size.

Patient population

We included adult patients aged 18 to 65 years with a confirmed diagnosis of T2D (duration of 3-10 years) and body mass indices of ≥24 and ≤40 kg/m². Patients were required to be on stable doses of ≥1 OADs for the last 12 weeks and have an HbA1c of 7.5% to 10.0% (59-86 mmol/mol) at the end of a 4-week run-in period. The run-in period ensured that patients had stable glycemic control on their existing OAD regimen before the intervention. Patients without stable control at the end of the run-in period, as determined by changes in OADs, HbA1c, and/or capillary blood glucose measurements, were excluded.

Study flow

Patients were invited to a screening visit to undergo a physical examination including observations, biometric data, and laboratory assessments (Fig. 1). Eligible patients were advised to start a 4-week run-in period to ensure stable glycemic control. During this period, patients checked finger-stick capillary blood glucose twice daily and recorded their results alongside OAD use in a diary. At the end of the run-in period, fasting plasma blood glucose and HbA1c measurements were drawn to confirm trial eligibility. Patients then underwent DMA with RFVA as described.¹¹ After the procedure, patients monitored capillary blood glucose 4 times a day for the first 14 days and then twice daily for 6 months. On days 2, 7, and 14 after the procedure, patients received a safety phone call.

Figure 1.

Study flow and selection. CBG, Capillary blood glucose; BD, twice daily; D, days; HbA1c, glycated hemoglobin; !, excluded from study.

All patients underwent follow-up endoscopy at 28 days to assess the duodenal mucosa under high-definition white-light endoscopy and obtain duodenal biopsy samples from the ablated segment. Primary and secondary endpoints were determined by clinical evaluation and laboratory assessments during site visits at day 28, week 12, and week 24 after the procedure.

Interventions

Duodenal mucosal ablation

All eligible subjects underwent DMA with the RFVA system, which is cleared by the U.S. Food and Drug Administration for the coagulation of bleeding and nonbleeding sites in the GI tract (Fig. 2). Procedures were completed with patients under a general anesthetic between October 2023 and December 2023 by a surgeon trained in GI endoscopy. Patients were positioned in the left lateral or prone position at the endoscopist’s discretion. Each procedure was completed with a therapeutic dual-channel endoscope with a 3-mm distal transparent attachment, but a pediatric or adult colonoscope with a 3.7-mm working channel to accommodate the 3.4-mm catheter could also be used.

Figure 2.

Radiofrequency vapor ablation system with processor, foot pedal, and catheter (arrow points to catheter).

The complete procedural steps (Fig. 3) have previously been published.¹¹ In brief, the papilla is identified, and a TTS clip is placed on the contralateral wall to mark the proximal boundary. The catheter is inserted TTS and deployed distal to the papilla. This reveals two 30-mm catheter discs (Fig. 2) that border a 2.5-cm treatment zone where steam is released. The catheter is attached to a generator, which controls the RF vapor generation and total energy delivered per ablation. The first circumferential ablation is delivered by foot-pedal activation to the proximal treatment zone. This is followed by a series of sequential ablations aiming for ≥9 cm of ablated duodenum in a proximal-to-distal direction. A second overlapping series of ablations are delivered after 5 minutes of the first ablations that should aim to retrace the same path. During an ablation run, there should be limited overlap between sequential ablations, which is aided by the mucosal changes and direct visualization as the device passes TTS. A small degree of overlap is expected, but the nature of the vapor being short applications and the wide safety margin observed in preclinical studies¹⁰ means safety issues are unlikely.

Figure 3.

Procedural steps of radiofrequency vapor ablation. A, The duodenum is inspected and washed with 2% acetylcysteine. B, An endoscopic clip is placed on the duodenal wall contralateral to the ampulla, which marks the proximal boundary. C, Radiofrequency vapor ablation catheter is inserted into the duodenum and 2 discs are deployed (only 1 visible in image). D, Vapor ablation is delivered to the duodenal mucosa. E, Discs are withdrawn, and the ablation zone is assessed. Additional ablations are then delivered in a proximal-to-distal direction. F, Complete postampullary treatment zone is assessed endoscopically. G, Second run of radiofrequency vapor ablations are delivered to give a double application. H, Final review of the ablated treatment zone.

Dosimetry

An initial dose of 180 J with a single application was chosen based on a series of preclinical subacute (24-hour survival) and chronic (3-week survival) animal studies.¹⁰ These confirmed the RFVA system was effective in causing mucosal necrosis at a dose of 180 J with both single and double application and led to adverse events (AEs; eg, stricture, perforation) at 400 J. Stepwise dose escalations were completed as part of the planned dosimetry to ensure safety and tolerability before escalating both the dose and number of applications (Fig. 4).

Figure 4.

Study dosimetry. An initial safety dose of 180 J with single application was administered to 2 patients. After 3 days, a clinical assessment was conducted and deemed safe (no serious adverse events) to move to the next dose of 180 J with a double application to 11 patients. After 4 weeks, the postprocedure gastroscopy and duodenal biopsy samples were reviewed, and it was deemed safe (no serious adverse events) to move to the final dose of 200 J with a double application in 14 patients. Exclamation mark (!) refers to a temporary pause and review before proceeding to the next dose.

Periprocedural instructions

At screening, patients were given standard dietary advice in line with recommendations for patients living with T2D and advised not to make any significant lifestyle changes during the trial. Patients fasted for a minimum of 6 hours before the procedure and then underwent an initial screening gastroscopy before proceeding to DMA. After the procedure, patients were admitted overnight for observation and discharged the following day on a modified diet for 2 weeks (clear fluids on days 1-3, liquid diet on days 4-5, and soft diet on days 6-14). After the initial cases, the procedure was converted to a day case intervention once tolerability and safety signals demonstrated that an overnight stay was not clinically warranted. Patients were discharged on oral proton pump inhibitors twice daily and sucralfate 4 times a day for 30 days.

Primary study endpoints

The primary endpoints of the study were safety, tolerability, and feasibility. These were predominantly determined during the first 14 days postprocedure with additional safety assessments determined at the 3 follow-up time points (28 days, 12 weeks, and 24 weeks) to ensure no emerging safety concerns were found. The primary safety endpoints were the incidence rates of procedure- and device-related serious AEs (SAEs) and unanticipated AEs. In addition, the number and severity of AEs were recorded by the study investigators. All AEs were classified according to Common Terminology Criteria for Adverse Events,¹² and the casual relationship to both the device and procedure were determined independently by the principal investigators (L.A.R.G. and R.H.).

Procedure tolerability was assessed by a visual analog scale (VAS) score for pain over the first 14-days after the procedure and the use of any analgesic medications. Patients were asked to document their level of pain each day by grading on a scale of 0 to 10. Feasibility was based on technical success, procedural time, and number and length of ablations. Technical success was defined as an ablated immediate postampullary duodenal length of ≥9 cm with ≥75% circumferential coverage by endoscopic assessment. Procedural time (minutes) was defined as both the endoscopy time (time between insertion and final removal of the endoscope) and catheter time (time between insertion and final removal of the RFVA catheter). The number of ablations was documented throughout the procedure, and once complete, the ablated segment was assessed to determine the total length.

Exploratory study endpoints

We determined the change in HbA1c from screening to 4, 12, and 24 weeks after the procedure among patients undergoing a treatment dose of RFVA. Our intention-to-treat cohort represented all patients undergoing an attempted treatment dose (ie, double application) of RFVA (n = 25). Additional exploratory endpoints were determined because of their close relationship with HbA1c including changes in weight (weight, body mass index) and glycemic parameters (capillary blood glucose, fasting plasma glucose, diabetic medication changes) over the same follow-up time points.

Ethical consideration and consent

The study protocol and trial oversight were approved by an independent Ethics Committee Authority (Servicio de Salud Metropolitano Oriente Comité de Ética Científico) and complied with recommendations of the Declaration of Helsinki. All patients were provided with written information about the trial and had written informed consent obtained before enrollment.

Statistical analysis

Data analysis was conducted using R version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria). Categorical variables are summarized using frequencies and percentages, whereas continuous variables are summarized using mean (SD) or median (interquartile range [IQR]). The change in HbA1c during the run-in was determined using the dependent t test (paired, 2-sided) to confirm stable glycemic control. The change in HbA1c over follow-up was determined using a generalized linear model with degrees of freedom adjusted using the Kenward-Roger method to improve estimate accuracy and pooled standard error. Pairwise comparisons between each follow-up time point and the screening HbA1c levels were conducted with Dunnett’s adjustment for multiple comparisons. Nominal P values and corresponding 95% confidence intervals (CIs) were reported. A P < .05 was considered statistically significant.

Results

Patient characteristics

Eighty-one patients with T2D who met the initial inclusion criteria were screened and entered the run-in period. Of these, 46 (56.8%) failed screening because of fluctuating diabetic control and 8 (9.9%) because of poor study concordance or withdrawal. The remaining 27 patients all proceeded to DMA with the RFVA system. The baseline characteristics are shown in Table 1. Mean patient age was 54 years (SD, 6.6), 51.9% were women, average duration of T2D was 6 years (SD, 2.6) and the mean baseline HbA1c was 8.6% (SD, 0.6).

Table 1.

Baseline patient characteristics (n = 27)

Characteristics	Values
Mean age, y (SD)	54 (6.6)
Sex
Male	13 (48.1)
Female	14 (51.9)
Mean length of type 2 diabetes diagnosis, y (SD)	6 (2.6)
Mean glycated hemoglobin, % (SD)	8.6 (0.6)
Mean fasting plasma glucose, mmol/L (SD)	9.3 (2.5)
Mean body weight, kg (SD)	80.2 (11.0)
Mean body mass index, kg/m2 (SD)	30.3 (4.2)
No. of oral antidiabetic drugs
1	13 (48.1)
2	12 (44.4)
>2	2 (7.4)
Values are n (%) unless otherwise defined.

SD, Standard deviation.

Safety

There were no SAEs, no unanticipated device- or procedure-related AEs, and no episodes of hypoglycemia during the study period. Of 27 AEs occurring in 19 patients, 23 were mild and 4 were considered moderate because of as-needed medications (Table 2). The most frequently occurring AEs were GI (n = 10; 37.0%), with the most common being abdominal pain (n = 7; 25.9%). Among the 4 moderate AEs, 2 were abdominal pain (7.4%) requiring an as-needed oral antispasmodic and 2 were headache (7.4%) that required an analgesic agent. Six AEs (abdominal pain) were deemed possibly procedure or device related, and the remaining AEs were probably not related. There were no major differences in AEs between the first (180 J) and second (200 J) treatment cohorts.

Table 2.

Summary of adverse events

Summary	No. of cases (%)
Adverse event severity
Mild	23 (85.2)
Moderate	4 (14.8)
Severe	0 (0)
Adverse event etiology
Gastrointestinal
Abdominal pain	7 (25.9)
Diarrhea	2 (7.4)
Hemorrhoids	1 (3.7)
Constipation	1 (3.7)
Sore throat	2 (7.4)
Hypoglycemia	0 (0)
General disorders
Musculoskeletal pain	4 (14.8)
Fatigue	3 (11.1)
Headache	3 (11.1)
Other
Throat pain and general discomfort	4 (14.8)

All patients underwent a follow-up endoscopy and duodenal biopsy sampling from the ablated duodenal segment at 4 weeks after the procedure. The duodenal endoscopic appearance was normal in all patients. Duodenal biopsy samples underwent routine hematoxylin and eosin staining and were reviewed by 2 independent histopathologists. All biopsy samples were read as normal with complete mucosal healing (Fig. 5).

Figure 5.

Duodenal biopsy sample 4 weeks after radiofrequency vapor ablation. Normal mucosal architecture and villous maturation are seen. No chronic duodenitis and no acute inflammation are found. Hematoxylin and eosin (H&E) staining, 20× magnification.

Tolerability

Postprocedure, all patients tolerated the 2-week modified diet without significant discomfort. The highest mean reported VAS pain score was 1.4 (SD, 2.1) on day 2, which resolved to 0.6 (SD, 1.0) on day 3 (Fig. 6).

Figure 6.

Patient-reported mean VAS for pain after procedure. VAS, Visual analog scale.

Procedural feasibility

The overall technical success rate was 100%. Among the 25 patients receiving a treatment dose (double application), the mean procedure time was 49.7 minutes (SD, 14.3), catheter time 34.5 minutes (SD, 10.9), mean number of ablations 15.8 (SD, 3.2), and average treatment length 13.8 cm (SD, 2.0). There was a reduction in both the procedure time (56.3 vs 43.6 minutes) and catheter time (38.8 vs 30.6 minutes) between the first treatment cohort (180 J with a double application; n = 11) and second treatment cohort (200 J with a double application; n = 14) despite an increase in the total number ablations (13.5 [SD, 2.5] vs 17.6 [SD, 2.7]) and total length of ablated segments (13.5 [SD, 2.2] vs 14.1 [SD, 1.8]).

Preliminary exploratory efficacy

In the intension-to-treat cohort (n = 25), we observed a significant reduction in HbA1c of 0.9% (95% CI, 0.51-1.29), 1.2% (95% CI, 0.81-1.59), and 0.8% (95% CI, 0.41-1.19) at 4, 12, and 24 weeks, respectively, compared with baseline (Fig. 7). There was no significant change in HbA1c during the run-in period (–0.1%, P = .482). Over this time, patients also had a small reduction in body weight at 4 weeks (–3.6 kg [SD, 2.6]), 12 weeks (–2.4 kg [SD, 5.0]), and 24 weeks (–1.2 kg [SD, 5.2]). During follow-up, the median number of OADs did not change from screening (1 [IQR, 1-2]). Four patients had OAD alterations on follow-up. Three patients changed their metformin at the 12-week follow-up (stopped = 1, reduction = 1, increase = 1) and 1 patient stopped 3 medications between 12 and 24 weeks without informing investigators. Finally, we observed no significant difference in change in HbA1c between the 2 treatments cohorts (180 J vs 200 J [double application]), although the pilot was insufficiently powered to assess a dose response.

Figure 7.

Change in glycated hemoglobin after radiofrequency vapor ablation.

Discussion

We demonstrated in this first-in-human, prospective, pilot trial that the RFVA system is a safe, well-tolerated, and easy to perform procedure for DMA. The procedure was completed successfully in all cases, with a procedure time of 43.6 minutes and catheter time of only 30.6 minutes in the second treatment cohort. The procedure is easy to perform with its TTS design, whereby the endoscopist can deliver controlled ablations under direct visualization at an enhanced rate of 2.5 min/cm. By comparison for the Revita system (Fractyl Laboratories Inc), their catheter time was 56 minutes in their randomized, sham-controlled trial even in the hands of endoscopists skilled in advanced therapeutic procedures.¹³ Moreover, data from several initial open-label prospective trials on the ReCET system (Endogenex Inc) showed similar results with a catheter time of 58 minutes¹⁴ and 68 minutes.¹⁵ Both Revita and ReCET have added complexity because of their over-the-guidewire design requiring fluoroscopy for both catheter placement and advancement. Additionally, catheter manipulation cannot be controlled with the endoscope tip, which limits control and reduces scalability.

DMA with RFVA was exceptionally well tolerated with a maximum mean VAS pain score of 1.4 on day 2 postprocedure. In addition, there were no SAEs and only minimal AEs, with most being mild transient abdominal pain (25.9%). This is in line with other duodenal ablation devices, which commonly report transient abdominal pain, sore throat, and altered bowel habits.13, 14, 15 To date, there have been no SAEs with electroporation.^15,16 With hydrothermal ablation using the Revita system, SAEs occurred in 7.5% with the first-generation catheter.¹⁷ These SAEs were due to 3 cases of duodenal stenosis, which demonstrates the risk of excessive small intestinal ablation. Nevertheless, no major SAEs have been reported with the second-generation catheter, although there were 2 procedure-related SAEs (3.6%; jejunal perforation and hematochezia) in the REVITA-2 trial.^13,18 Safety is critical when working in the duodenum, which is aided by the design of the RFVA system. Passing the catheter TTS provides the endoscopist with more control and eliminates guidewire-assisted pushing through the tortuous duodenum. RF vapor is then delivered under direct visualization to avoid overlap between sequential applications. These ultrashort durations of application of RF vapor limit deep thermal injury to the submucosa and muscularis.¹⁰ This is exemplified by the excellent tolerability, low VAS pain scores, and complete endoscopic and histologic healing at 4 weeks after the procedure.

Our preliminary efficacy data are encouraging with a reduction in HbA1c at 12 and 24 weeks by 1.2% and 0.8%, respectively. However, we acknowledge this was a pilot study that was not adequately powered to determine primary efficacy. These early results appear to be consistent with other ablation devices, which have demonstrated an HbA1c reduction of 10.0 mmol/mol in the open-label Revita study among 36 patients at 24 weeks,¹⁸ a reduction of 1.5% in the U.S.-based open-label study evaluating the ReCET procedure at 12 weeks (n = 20),¹⁶ and a reduction of ∼0.8% with the first-generation ReCET catheter at 24 weeks (n = 30) within an Australian-based open-label study.¹⁵ The reasons why efficacy may vary between technologies are the degree of ablation (eg, extent and depth of circumferential ablation and/or ablation length), baseline patient population (eg, severity of diabetes, disease duration, local diet and lifestyle practices, and use of continuous glucose monitoring), mechanism of ablation (eg, thermal vs nonthermal), or study design (eg, lack of a control group, underpowered) rather than long-term structural differences. Regardless of these differences, there does appear to be a consistent, early efficacy signal across 3 independent DMA technologies.

Another key question is why these procedures appear to impact glycemic control, with exploratory data from duodenal mucosal resurfacing showing it can improve insulin sensitivity and beta cell function at 3 and 6 months after the procedure.¹⁹ We know from gastric bypass that improvements in glycemic control can occur almost immediately in a weight-independent manner as nutrients are diverted away from the proximal intestines without any structural injury.²⁰ Therefore, rather than long-term structural injury after RFVA, which we have shown does not occur, we hypothesize it is the alteration in the interaction between the proximal intestine and enteral glucose that may have downstream effects on hepatic gluconeogenesis and insulin sensitivity. These changes may occur by alteration in cellular glucose receptors, intestinal and hepatic gluconeogenesis, bile acid signaling, or other complex transcriptomic changes.

The change in HbA1c we observed after RFVA could be due to poor medication concordance, a significant change in lifestyle habits, or an inadequate procedure durability because of the chosen ablation dose. These factors are important when planning future studies to determine the efficacy of RFVA. We hypothesize that durability would be enhanced with an increased dose; however, further dose escalation could increase the risk of AEs from deeper thermal injury. Therefore, focusing on optimizing the extent of circumferential ablation is a key area. These suggestions mirror those of the second-generation ReCET catheter that has a bigger diameter, higher dose, and more electrodes to improve circumferential coverage. This second-generation catheter, by optimizing the extent of circumferential ablation, led to a higher mean reduction in HbA1c of 1.7% at 24 weeks among 21 patients in their U.S. trial.¹⁵ In addition, ablation length will also play a critical part of improving future efficacy. We know that by using the TTS RFVA system, we can get an adequate ablation length above and beyond over-the-guidewire devices at 13.8 cm that could be easily lengthened beyond the ligament of Treitz, particularly with a colonoscope. After establishing that RFVA is safe and well tolerated, we have moved to enhance procedural feasibility with the next-generation TTS basket-tip design that retains the 3.4-mm catheter, thus enabling compatibility with more routinely used endoscopes. This is to improve both the circumferential ablation uniformity and the length of ablation that can be safely delivered. These refinements will be critical in ultimately designing a well-powered clinical study to determine the procedural efficacy. Furthermore, we believe the need for a duodenoscope to aid intraprocedural papilla identification is extremely unlikely, which would enable RFVA in the ambulatory facility setting.

With the growing global use of GLP-1RAs, a key question will be how RFVA fits into the treatment algorithm. To date, it is observed that DMA may have a mechanism independent of GLP-1RAs that promotes the idea of a synergistic treatment effect.¹⁹ This hypothesis has been investigated in a cohort of 16 insulin-requiring patients with T2D who underwent duodenal mucosal resurfacing with Revita followed by the addition of liraglutude.¹⁸ At 12 and 18 months postprocedure, 56% and 53% remained off insulin, respectively. Furthermore, the EMINENT-1 study¹⁴ explored the combination of ReCET with semaglutide among a population of 14 insulin-requiring T2D patients. At 12 months of follow-up, 86% remained off exogenous insulin with associated improvements in HbA1c. Therefore, future studies will need to explore the efficacy of RFVA with and without GLP-1RA therapy as well as in different populations (eg, new-onset disease, advanced disease, insulin-requiring).

Limitations of the current study are its open-label design, meaning patients were not blinded to treatment, and inherent biases may have occurred given it was a single center. The small sample size means it was not adequately powered to determine primary efficacy, and we cannot completely exclude external factors such as lifestyle contributing to these results, particularly because of a lack of a control group. Because of study geography, it may be difficult to generalize the feasibility of the device in North American and European populations. Finally, because of the study nature being dosimetry finding, not all patients underwent the optimal RFVA dose. Additional studies have been planned with our new catheter design to further refine procedural feasibility and explore initial efficacy to enable development of larger, well-powered clinical trials.

In summary, we demonstrated that DMA using RFVA is a safe and well-tolerated procedure. Because of its TTS design, we have shown that it is feasible as an easy-to-use endoscopic procedure for patients with poorly controlled T2D. Preliminary efficacy data are encouraging and provide further support for the importance of the proximal intestines in the control of metabolic disease in the growing field of metabolic endoscopy.

Disclosure

The following authors received research support for this study from Aqua Medical Inc: B. C. Norton, A. Neiponice, P. B. Hoebel; P. Vignolo, L. A. R. Grunert, and R. Haidry. All other authors disclosed no financial relationships.