Safety outcomes in 602 GERD patients treated by RefluxStop: a multi-center real-world study from 22 centers across six European countries

Abstract

RefluxStop surgery aims to restore a functional anti-reflux barrier in gastroesophageal reflux disease (GERD) without encircling the esophagus, purportedly limiting dysphagia and gas-bloating. This report presents safety outcomes from 602 patients at 22 centers across six European countries and up to 6.75 years follow-up (mean [SD] 2 [1.25] years). Serious safety outcomes and reoperation occurred in 1.99% (12/602), all satisfactorily resolved. Thereof, two-thirds were reherniations (1.33%), which manifested from total disruption of crural repair, resolved by straightforward fundic repositioning and redo hiatal repair with RefluxStop unaffected in its pouch. Non-operative safety events included 4/602 (0.66%) cases of asymptomatic erosion (without action) in the early postoperative phase (4 weeks) as the learning curve of a new procedure. Dilatation for new-onset dysphagia was performed in one patient. This report presents an analysis focused on safety data from the largest independent RefluxStop study to date (n = 602). Despite inter-surgeon variability, the learning curve for a new procedure, and sizable proportions of large hernia and/or dysmotility patients, results show low rates of serious safety outcomes with reoperations occurring in <2%. The overall experience of the RefluxStop Evaluation Group demonstrated a consistent and favorable mid-to-long-term safety profile for RefluxStop in real-world surgical treatment of GERD.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-47034-3.

Introduction

Gastroesophageal reflux disease (GERD) is a common condition that elicits substantial medical resource use for both primary and secondary care^1,2. This often chronic and relapsing condition may impair quality of life comparably to acute myocardial infarction and congestive heart failure³, however, treatment itself may also contribute to disease burden by increased morbidity from adverse safety outcomes. The two predominant treatment options for patients with GERD are long-term acid suppression therapy with proton pump inhibitors (PPIs) and laparoscopic anti-reflux surgery (LARS)⁴, both pathways associated with a proportion of side effects that result in otherwise unnecessary healthcare utilization and undue patient suffering.

Medical acid suppression therapy has generally been proven an effective first-line therapy in comparison with previous medications, yet a 2020 population-based survey reported persistent symptoms in >50% of those taking daily PPI therapy⁵. In terms of the safety of medical therapy, a large >150,000-person study followed United States (US) veterans over a 10-year period of PPI use and found 7000 excess deaths related to adverse drug effects⁶, incurred by drug-induced consequences like cardiovascular events, digestive tract malignancy, renal impairment and illness, and infectious/parasitic disease. Despite the relative ease of access, PPIs oftentimes obtained over the counter, the lifelong safety implications are surprisingly prevalent and insidious from the patient perspective.

Traditional LARS, by fundoplication, is generally regarded as safe and effective, currently representing the surgical standard of care worldwide, but nonetheless, has a well-recognized association with side effects. Postoperative dysphagia (difficulty swallowing) and gas-bloat syndrome are prominent examples⁷. Although many surgeons adopt a partial wrap technique attempting mitigation of these outcomes, such modifications are a balancing act of fewer side effects against potential loss of efficacy⁸. Moreover, significant rates of serious safety signals like reherniation⁹ and reoperation⁷ have been reported for primary LARS.

An alternative to fundoplication techniques is the magnetic sphincter augmentation (MSA) intervention, which according to the literature carries identified risks of device erosion and/or need for removal (mostly due to dysphagia¹⁰, dysphagia seemingly related to device sizing¹¹. There is reportedly no significant short-term difference between MSA and fundoplication techniques in terms of postoperative dysphagia¹², and post-MSA cases appear to be more severe¹³.

RefluxStop is a non-active (not powered) device that is laparoscopically implanted on the exterior of the gastric fundus, laying encapsulated in a pouch of stomach wall in the gastric lumen like a teardrop. It is designed to treat GERD by fully reconstructing the anti-reflux barrier (ARB) rather than encircling or compressing the lower esophageal sphincter (LES). The RefluxStop pivotal study, a long-term prospective multi-center study used for CE mark and PMA US Food and Drug Administration (FDA) submission, was published in a series of safety and effectiveness reports up to 5 years^14–18.

Since the device gained CE mark in 2018, the procedure was implemented at >50 European centers where >1500 cases have been performed to date. Several individual centers have published safety and effectiveness outcomes based on observational study^19–27, reporting promising results even with proportionally large subcohorts having difficult-to-treat disease features prone to adverse safety outcomes. Given the growing adoption of this innovative treatment at several centers throughout Europe, and as part of responsible clinical practice, it is necessary to collate and report the available safety data thus far.

Objective

The purpose of this report, compiled by the independent RefluxStop Safety Evaluation Group, is to present all safety signals reported in real-world practice (i.e., in actual day-to-day healthcare settings) of 602 patients from 22 centers across six European countries. The report focuses on safety and technical feasibility data from this study of RefluxStop surgery.

Methods

Study design

This was a multi-center retrospective single-arm observational study. Twenty-two centers were included from Germany, Switzerland, Spain, Italy, the United Kingdom (UK), and Austria, with two-thirds of the patients from the early adopters in the German-speaking part of Europe. Data were collected from a review of medical records, collated, and analyzed for rates of adverse events (AEs) and safety signals.

Inclusion/exclusion criteria

This retrospective analysis included all consecutive patients (i.e., mean follow-up of 2 years, ranging from 2 months to 6.75 years) that underwent the RefluxStop procedure from 22 European centers that have participated in the contribution of patient-level safety data. The reporting period ended on March 31, 2025. Centers not able to provide such data prior to the conclusion of the reporting period were excluded.

The RefluxStop procedure was indicated for adults (≥ 8 years of age) with a diagnosis of GERD defined by abnormal 24-h pH testing results (acid exposure time pH <4 of ≥4.5%). This threshold was used for consistency, opposed to the Lyon consensus 2.0 (2024), since many included patients from this long-term evaluation were diagnosed prior to the new definition. As per practice norms at each clinic, patients were offered regionally available treatment options and only those selecting RefluxStop surgery were included. Study participants included all categories of reflux sufferers, also comprised of patients with large hiatal hernia (HHࣧ) defined as ≥3 cm in axial length in this study, reasonably so given inconsistency in the surgical literature⁹, severe ineffective esophageal motility (IEM), and obesity. Importantly, these data also included the learning curve of a new procedure at all centers. A standardized technique has since been developed based on the experience gained during early implementation.

Besides pH testing and, in a few centers, endoscopic verification of esophagitis or large HH (a requirement for all patients prior to surgery), other baseline investigations were performed in line with the clinical practice at each institution.

Primary objective

The primary objective of the RefluxStop Safety Evaluation Group was to report device- and procedure-related, serious and non-serious adverse events (SAEs and AEs) following RefluxStop surgery for the treatment of GERD in a pooled cohort of 602 patients from 22 centers in Europe. As such, the focus is safety and technical feasibility. AEs and SAEs were defined in line with European Union (EU) Medical Device Regulation (MDR) terminology²⁸, shown in Table 1 and grouped according to severity, an appropriate practice for this European-based study. This evaluation, prompted by the RefluxStop Safety Evaluation Group, was incepted as part of real-world surveillance of safety signals after RefluxStop surgery using the largest collated sample to date, anticipating further clarity regarding this novel treatment option for guidance in GERD management decision-making. Following collection of all safety signals, they were grouped into the following clinical categories:

Table 1.

The EU MDR definitions of safety outcomes.

Terms	Definition
Adverse event (AE)	Any untoward medical occurrence, unintended disease or injury or any untoward clinical signs, including an abnormal laboratory finding, in subjects, users or other persons, in the context of a clinical investigation, whether or not related to the investigational device
Serious adverse event (SAE)	Any adverse event that led to any of the following:  (a) Death  (b) Serious deterioration in the health of the subject, that resulted in any of the following:      (i) Life-threatening illness or injury      (ii) Permanent impairment of a body structure or function      (iii) Hospitalization or prolongation of patient hospitalization      (iv) Medical or surgical intervention to prevent life-threatening illness or injury of permanent impairment to a body structure or a body function      (v) Chronic disease  (c) Fetal distress, fetal death, or a congenital physical or mental impairment or birth defect
Device deficiency	Any inadequacy in the identity, quality, durability, reliability, safety, or performance of an investigational device, including malfunction, use errors, or inadequacy in information supplied by the manufacturer

EU, European Union; MDR, Medical Device Regulation.

• Early device erosion (< 4 weeks).

• Device dislocation via improperly closed deployment channel.

• Esophageal dilatation.

• HH recurrence.

• Bleeding from short gastric vessels.

• Reoperation (including explantation/device removal).

• Clinically non-indicated reoperation, identifiably.

• Explantation/removal of device.

• Intraoperative events.

• Non-serious safety events.

• Incisional trocar hernia.

• Conversion to open procedure.

Thereafter, each AE was reviewed to identify possible contributing factors, such as patient selection, surgical technique, and relevant comorbidities.

RefluxStop device and procedure

Rationale of treatment principle

The concept of the procedure is the anatomical restoration of the failed ARB, as recently described by two American Foregut Society publications^29,30, to allow for its normal physiological function (via correct placement rather than direct surgical manipulation with encircling of the lower esophagus).

Device description and role of safety features

The central component of the procedure is the RefluxStop device, a small, inactive, rounded cube that is invaginated in an exterior pouch of fundus wall (Fig. 1). Here, the package of invaginated fundus (with the device enclosed) purportedly acts as a mechanical stopper to anchor the LES intra-abdominally and is thought to reduce HH recurrence³¹.

Fig. 1.

The complete invagination of the RefluxStop device in its appropriate position on the external stomach wall, protruding into the stomach cavity in a teardrop shape formed by the encapsulated device. This fundic package serves as a stabilizing apparatus for the newly reconstructed ARB, where narrow (90°–110°) esophagogastric plication recreates the acute angle of His/gastroesophageal flap valve and the fundic package ensures sufficient intra-abdominal position of the LES with prevention of reherniation (an important safety feature of the procedural design). The procedure was designed in such a way that esophageal encirclement or compression is completely avoided, thereby mitigating the risk of postoperative dysphagia and overall reducing side effects due to the device invagination outside the stomach wall with its strong serosa layer. Other safety signals like esophageal erosion that may occur with MSA, have a substantially reduced risk profile due to the device design in pieces, as supported by this study. Image used with permission from Implantica.

The device (2.5 cm/1 inch in size) is made of medical-grade silicone and consists of five pieces that fit together to form the cuboid implant (Fig. 2), held together by resorbable suture placed while outside of the abdomen. Were the device to erode in the early phase to the stomach cavity, it is deductively posited as an asymptomatic event where the device pieces pass through the gastrointestinal tract uneventfully, not requiring further intervention.

Fig. 2.

The RefluxStop device (~2.5 cm) assembled and disassembled, respectively. The implant was developed in five separate segments as an inherent safety feature, designed to disassociate and pass through the digestive tract uneventfully (not requiring intervention) in the reportedly rare case of early erosion, in which poor surgical technique with overly tight suturing of the invagination pouch is the probable cause. It is important that implant invagination is performed loosely to avoid the risk of tissue degeneration from poor tissue circulation (caused by pressure from the device on the stomach wall in a too tight pocket). Image used with permission from Implantica.

Standardization of surgical steps

The laparoscopic procedure was carried out only by surgeons who had successfully completed standardized training from a licensed proctor authorized by the device manufacturer (Implantica, Switzerland).

In essence, the surgical steps of the RefluxStop procedure sequentially reinforce the newly reconstructed ARB, via laparoscopic access (Fig. 3). First, the esophagus is dissected extensively and superiorly into the mediastinum, enough to ensure at least 5 cm of relaxed intra-abdominal esophageal length. Any HH, if present, is repaired with loose cruroplasty to ensure the risk of postoperative dysphagia from constriction is avoided. Then, the fundus is dissected superiorly and posteriorly, with division of short gastric vessels, to achieve a tension-free and “floppy” status. Narrow (90°–110°) esophagogastric plication is performed on the patient’s left side and between the vagal trunks, recommended with single interrupted sutures in at least two rows. The RefluxStop implant is introduced to the abdomen via a dedicated deployment tool (integrated with a trocar) intended to simplify and improve successful invagination on the outside of the stomach fundus wall, where the implant pouch is loosely and fully closed, including a top and bottom suture for fixation¹⁴. The device ends up invaginated and hanging like a teardrop into the stomach cavity, avoiding interference with other organs and thereby minimizing AEs.

Fig. 3.

Main and alternative trocar placement—the latter often used for the RefluxStop procedure. Image used with permission from Implantica.

Follow-up of patients

Patients were followed up for a mean of 2 years (ranging from 2 months to 6.75 years). The distribution of follow-up time is further delineated by a median (IQR) of 21.3 (12–34.3) months. Clinical follow-up was conducted in accordance with the standard practices of individual institutions, which required variable real-world follow-up timescales consistent with existing patient treatment pathways. However, all patients were required to have a minimum of 2 months of follow-up to be included in the study. Follow-up was generally conducted in-person at 3 months and 6 months, and annually thereafter. Among the centers that were the first to perform RefluxStop surgery, follow-up data was obtained over a period of 4–7 years and included patients with a high proportion of IEM and large HH, for which a few separate articles have been published on these topics^11,32.

Statistical analysis

Data was reported descriptively, using measures of central tendency and dispersion as mean and standard deviation (SD) as well as absolute values and percentage frequency, n (%), for categorical data.

Ethics approval

Ethics approval was performed locally in line with local rules and praxis. The data collection formed part of standard monitoring conducted at each center. The procedure itself was offered as a routine treatment option; that is, patient care did not deviate from routine clinical practice for the purposes of this study. Informed consent was obtained from all subjects, and the study was performed in line with the Declaration of Helsinki. Generally, patients provided written informed consent for the use of anonymized data in research, as part of standard consent to undergo surgery. Informed consent was acquired by individual centers, and each center also approved the data reported. All methods were carried out in accordance with relevant guidelines and regulations. The procedure was performed in line with standard of care, and all centers went through training and proctoring, however, the training program improved over time based on new learnings. Experimental protocols were avoided.

Results

Sample population

Out of >1500 performed RefluxStop procedures overall, between September 2018 and January 2025 a total of 602 patients underwent real-world RefluxStop surgery at 22 centers in six European countries. Table 2 shows the country-specific number of cases performed and Table 3 presents baseline patient characteristics. Mean (SD) age was 51.6 (14) years and the median (IQR) was 53 (41–61) years. Median (IQR) body mass index (BMI) was 27 (24–29.3) kg/m² and gender included 44.5% female. Large hernia, defined in this study by HH axial length of ≥3 cm, was present in 31% of the 75% of subjects that reported hernia size.

Table 2.

RefluxStop procedures performed by country.

Country	Number of patients
Switzerland	208
Germany	180
Italy	74
United Kingdom	62
Spain	54
Austria	24*

*Including only published data from consecutive patients due to local EC regulations.

Table 3.

Baseline clinical characteristics.

Total	N = 602
Age, mean (SD), years	51.6 (14)
Male sex, %	55.5%
BMI, median (IQR), kg/m2	27 (24–29.3)
Large hiatal hernia > 3 cm (reported in 75% of subjects)	31%
Mean large hernia size	5.1 cm

BMI, body mass index; IQR, interquartile range; SD, standard deviation.

Serious safety outcomes

Table 4 lists all serious events that occurred after surgery, and a summary is provided below. All patients were resolved satisfactorily, with a detailed presentation of clinical context, management, and status:

Table 4.

Serious safety outcomes with and without reoperations after RefluxStop surgery (N = 602) with mean (SD) follow-up of 24 (15) months (range 2 months – 6.5 years).

Type of serious safety outcomes	Total cases, n (%)	Related to device or procedure?		Reason and clinical review		Management and consequences	Subtotal, n	Status
Hiatal hernia (HH) recurrence	8/602 (1.33%)	Related to antireflux procedure in general. Most important complication in all anti-reflux surgery All cases caused by total collapse of hiatal repair Straightforward procedure with extended dissection and repositioning of fundus with the device intact in its pouch		3 cases—Recurrence with baseline large HH, average 7 cm in size due to total hiatal repair collapse		All three patients were successfully reoperated with revision of HH repair, including repositioning of fundus with RefluxStop (in all cases) intact in its pouch and not impacted by the revision surgery. Hiatus was further reinforced with mesh	3	Recovered/Resolved Inform anesthesiologist to avoid retching/vomiting postop, which could be detrimental for the hiatal repair
				3 cases—Intense retching/vomiting directly post-op (1) or from food poisoning (2) causing total hiatal repair collapse		Intensive retching vomiting directly post-op (1) or admitted due to food poisoning (2) with total collapse of HH repair. Straightforward extended dissection and stomach fundus repositioning with RefluxStop untouched and intact in its pouch. In one case of food poisoning during the holiday season, open surgery performed at a different hospital. In one obese patient, the device was removed and the RefluxStop plication extended to a Dor fundoplication (with current experience level of the involved surgeon the device would probably not have been explanted). No residual symptoms	3
				2 additional cases of total hiatal repair collapse for unknown reason, unrelated to the RefluxStop device. Fundus reherniated with implant intact in-pouch		Laparoscopic revision in which the stomach fundus with the device intact in its pouch were repositioned, dissection extended, and the ripped cruroplasty repaired	2
Postoperative bleeding	1/602 (0.17%)	Procedure		Postoperative bleeding from short gastric vessels (SGVs): One patient had direct postoperative bleeding from SGVs and needed laparoscopic reoperation same day		Well-known SAE of standard-of-care LARS; non-specific to RefluxStop device or surgery	1	Recovered/Resolved
Device dislocation	1/602 (0.17%)	Procedure		This case was attributed to insufficient surgical closure of the pouch (deployment tool channel): an error in surgical execution, where the device exited the open pouch		This patient fully recovered with components of the implant recovered laparoscopically without further intervention needed	1	Recovered/Resolved Learning curve related.
Stomach ulcer	1/602 (0.17%)	Procedure		Stomach ulcer at device position: Patient was admitted to a different hospital than where RefluxStop was implanted due to stomach ulcer (patient-reported black stool)		Endoscopy at the hospital showed a small ulcer and no active bleeding. MRI showed that the stomach ulcer was superficial and only involved the inner mucosal layer. The implant was precautionarily surgically removed confirming the mucosal character of the ulcer, where a contributing factor was the lack of knowledge and understanding of the device and its AEs at this second hospital	1	Recovered/Resolved PPI therapy with double dose PPI for 8 weeks is an alternative ulcer treatment, successful in a patient at another hospital.
Early erosion	4/602 (0.66%)	Procedure No reoperative measures required		Generally, an overly tight closure of fundus pouch around implant. Contributing factors were: in one case likely a bleed in the stomach wall of the pouch, increasing pouch tightness; and in one case an unusual weak circulation in fundus. Standardized surgical technique with loose invagination designed to avoid this event		Device passed asymptomatically through GI tract without further intervention, except one case where the surgeon decided to remove the implant pieces endoscopically. Since the fibrotic pouch structure were intact, patients were well-treated even after this AE, at least for some time. One patient later had recurrence of reflux and underwent a Toupet fundoplication after 9 months	4	Recovered/Resolved Mainly learning curve-related.
Total reoperations	12/602 (1.99%)							All successfully treated or recovered/resolved

AE, adverse event; GI, gastrointestinal; HH, hiatal hernia; LARS, laparoscopic anti-reflux surgery; MRI, magnetic resonance imaging; PPI, proton pump inhibitor; SAE, serious adverse event; SGV, short gastric vessel.

Bold is to highlight the key results.

1. Early asymptomatic erosion–no reoperation (< 4 weeks)—0.66% (n = 4/602)

Four (n = 4/602, 0.66%) instances of early device erosion occurred out of 602 cases. All four patients were asymptomatic or had mild and temporary abdominal pain, some noticing the device in fecal matter and in others device erosion was detected during routine x-ray following local strategy. These cases were complemented by video review that showed an overly tight pouch around the implant. One case had a bleeding from the stomach wall inside the pouch (as visible by video recording) adding force on the stomach wall. The devices passed through the digestive tract asymptomatically, due to the device’s design disintegrating into pieces. No reoperation was needed but endoscopic removal was opted for in one case. These cases took place early within 4 weeks after surgery and all patients fully recovered. Later recurrence of reflux symptoms was appreciated in one case at 9 months after surgery, which was converted to Toupet fundoplication. The 4/22 surgeons involved only had one such early erosion case each as early adopters due to too tight suturing of the pouch.

• 2.HH recurrence—1.33% (n = 8/602)

All eight cases of HH recurrence had total disruption of the crural repair. Three (n = 3/8) cases of reherniation occurred in patients with large HH preoperatively (i.e., mean 7 cm in size). All patients were successfully repaired whereof n = 7/8 were managed by a straightforward repositioning of the fundus in combination with extended dissection (the implant intact and untouched in its pouch), followed by hiatal repair. In one obese patient, the device was explanted, and the patient was converted to Dor fundoplication. All patients fully recovered after re-surgery.

• 3.Postoperative bleeding—0.17% (n = 1/602)

One (n = 1) case of postoperative bleeding was identified from the short gastric vessels, requiring laparoscopic revision on postoperative day 1 for hemostasis.

• 4.Device dislocation—0.17% (n = 1/602)

During re-surgery, it became obvious that this case could be attributed to insufficient surgical closure of the pouch (deployment tool channel), an error in surgical execution, where the device (in pieces) exited the open pouch. Device components were unproblematic, laparoscopically retrieved.

• 5.Precautionary explantation—0.17% (n = 1/602)

One (n = 1) case was admitted, as an emergency at over 3 months after surgery, to a different hospital without RefluxStop experience. The patient had episodes of melena, and fainted once, but presented with no other symptoms. Hemoglobin was <30 g/L and endoscopy confirmed a mucosal ulcer at the site of the implant. However, no active bleeding was noted, although the overall clinical picture was consistent with the presentation of bleeding. The patient was therefore re-admitted to a nearby hospital; a second endoscopy showed no active bleeding, and magnetic resonance imaging (MRI) confirmed the mucosal character of the ulcer. Following discussion with the implanting surgeon, the implant was removed laparoscopically without subsequent intervention necessary and uneventful recovery of the patient followed.

Non- or less serious safety outcomes

Three types of non- or less serious events occurred, listed in Table 5. All cases were resolved with full recovery. Hiatal stenosis—0.17% (n = 1/602): In one case (n = 1), the patient experienced dysphagia postoperatively. Imaging with computed tomography (CT) and barium swallow suggested severe hold-up at the hiatus and inflammation at the operation site. Following failure to respond to intravenous steroids, the patient underwent revisional surgery 1 week after the index operation where the implant was uneventfully removed for precautionary reasons (i.e., contrast swallow x-ray showed the device was not involved in the stenosis). Follow-up was uneventful and the patient’s dysphagia resolved. Incisional trocar hernia – 0.50% (n = 3/602): These patients are reported separately from more severe AEs and reoperations, in line with the literature³³, since none occurred at the implant-deployment trocar. Outpatient endoscopy follow-up combined with esophageal dilatation for new-onset dysphagia—0.17% (n = 1/602): One case of esophageal endoscopic balloon dilatation for new-onset dysphagia occurred. To be noted, one center only (1/22) which mainly focused on IEM patients (comprising 66% of subjects) performed a few dilatations postoperatively to alleviate preoperatively originating symptoms.

Table 5.

Non- or less serious safety outcomes (N = 602).

Type of safety outcome	Total cases, n (%)	Related to device or procedure?	Reason and clinical review	Management and consequences	Status
Esophageal dilatation for new-onset dysphagia	1/602 (0.17%)	Procedure	Endoscopic dilatation is an outpatient procedure	Successful dilatation performed	Recovered/Resolved No dilatations in 21/22 centers
Hiatal stenosis	1/602 (0.17%)	Procedure	Hiatal stenosis post-surgery verified by contrast swallow x-ray: Often occurring in standard-of-care procedures but only in 1/602 cases in this study due to the standardization of the RefluxStop procedure	The contrast swallow x-ray showed a clear hiatal stenosis, which was unsuccessfully treated with steroids. However, RefluxStop was not involved in this stenosis placed > 3 cm away and should not have been explanted as in this case, as per the expert group’s judgement. Esophageal ballon dilatation would be an alternative approach	Recovered/Resolved
Trocar hernia abdominal wall	3/602 (0.50%)	Procedure	Incisional hernia at trocar site: This is an AE occurring with low frequency in all laparoscopic surgery. Not related to RefluxStop deployment trocar	Oftentimes surgical repair of the abdominal wall is needed, although seen as a lighter procedure with less risk involved	Recovered/Resolved Reported separately to not mix with more severe events or reoperations, following literature experience

IEM, ineffective esophageal motility.

Bold is to highlight the key results.

Safety outcomes not seen as AEs

A total of three (n = 3/602) safety events were reported that were not considered AEs. Conversion to open surgery—0.17% (n = 1/602), where open surgery was necessary in one case due to substantial intra-abdominal adhesions. Patient-requested explantation—0.33% (n = 2/602), occurred in two cases, where patients no longer wanted the implant for undefined reasons.

Overall reoperations—1.99% (n = 12/602)

As summarized from above, 1.99% (n = 12/602) of patients were reoperated on for the described serious safety outcomes, whereof n = 8/12 cases were reoperated on for reherniation primarily caused by episodes of intensive retching/vomiting (resulting in total rupture of the hiatal repair) or very large hernia preoperatively. These cases generally underwent straightforward re-surgery with fundus repositioning (after extended dissection) having the RefluxStop device left intact in its enclosed pouch, followed by new hiatal repair. Out of these 12 (1.99%) reoperations, three (0.50%) devices were explanted, further discussed below.

Discussion

Safety evaluation of novel treatments is a prudent and necessary methodology aimed at prioritizing mitigation of incurred harm in clinical practice, where identification of safety signals is important to uncover/rank the risk of unwanted effects of individual treatments. Such evaluations may be used to guide treatment decision-making, with the goal of maximizing relief from disease while minimizing side effects and requiring the least amount of further intervention or healthcare utilization. As such, both standard-of-care medical therapy⁶ and LARS (i.e., Nissen fundoplication, first described in 1956³⁴, have a well-known association with unwanted effects that may substantially impact health and quality of life.

SAEs may manifest with considerable variability, but a striking item listed from the EU MDR definition is surgical intervention for the prevention of life-threatening insult, simply put, reoperation, a safety signal that is arguably too common with standard-of-care surgery. For instance, according to a recently published systematic literature review of 63 randomized clinical trials (RCTs) evaluating laparoscopic Nissen fundoplication (LNF)⁷, the rate of reoperation at 10 years was 16.3%.

Several novel therapeutic options, such as MSA, have been introduced attempting to improve upon the standard of care, however, MSA therapy balances two opposite effects (due to its mechanism of action), where smaller device sizes may increase the risk of dysphagia (i.e., potentially necessitating esophageal dilatation or explantation/reoperation) or erosion (i.e., necessitating reoperation) and larger sizes apparently engage in a trade-off of treatment effectiveness for improved safety^11,32. The literature reports a high rate of reoperation (including removals and other indications) following MSA when compared to the standard of care³⁵. Although newer treatment approaches have emerged over the years, most seem to be dominated by LES envelopment or direct augmentation by other means, and better management may require broader view of the entire ARB with less impact on the LES.

RefluxStop employs a completely novel treatment principle that aims to improve LES function by correcting abdominal position and all three parts of the ARB, rather than encirclement or augmentation. RefluxStop has shown promising safety outcomes throughout the entire 5-year study period of its pivotal (prospective) trial^14–18 used in PMA for the FDA filing with no cases of device dislocation, migration/erosion, and/or reherniation (confirmed by contrast swallow x-ray at 5-year follow-up) as well as no cases of esophageal dilatation and only two severe SAEs that were completely resolved.

Clinical contextualization of adverse outcomes

This prompted the RefluxStop Safety Evaluation Group to collate available clinical safety data of RefluxStop implantation in consecutive non-selected patients and its associated procedure, after treatment of GERD at 22 centers in Europe offering RefluxStop surgery. This is the largest sample population to date from our experience in real-world practice (i.e., day-to-day healthcare settings), to evaluate safety signals and better understand RefluxStop’s role in surgical management of GERD. Below we discuss the results of our experience, focusing on the safety outcomes and technical feasibility of RefluxStop surgery only. The analysis of 602 patients with a mean (SD) follow-up of 2 (1.25) years, the median (IQR) being 21.3 (12–34.3) months, provides a robust illustration of serious events and reoperations after the RefluxStop procedure. It was noted that many severe sufferers were included in this group, since surgeons initially began use of the device in advanced IEM and large hernia (≥3 cm) patients based on the design logic of the device and procedure.

Early device erosion—asymptomatic and no reoperation

Early erosion of the device to the stomach cavity (<4 weeks) occurred in 0.66% (n = 4/602). Complemented with video review, these events were most likely attributed to overly tight suturing of the fundic pouch. When the implant is enclosed in a too-tight pouch, stomach wall circulation is stifled and causes subsequent early tissue damage leading to early erosion of the device through the stomach wall, usually within 4 weeks. It seems that this was part of the learning curve effect³⁶ which is generally difficult to fully eradicate in the early days of a new procedure³⁷. Today, this has been accounted for, and the surgical technique has been further standardized by case review and proctoring, involving measures that avoid an overly tight pouch by both posterior dissection of the fundus and the first plication row of sutures placed posterior of the greater curvature of the stomach where the short gastric vessels have been divided. This event occurred in 4/22 centers, with one early case in each center, and can be attributed to the adoption of a new technique. In the pivotal CE/FDA study, no early erosion or later migration occurred (also objectively verified by 5-year contrast swallow x-ray)^14–18.

All cases of early erosion (n = 4) resulted in resolution and full recovery without any surgical intervention needed. The device, in case of erosion, has been specifically designed to dissociate into its five smaller parts and exit per rectum with the patient remaining asymptomatic and without clinical consequence, thereby enhancing patient safety. One patient experienced recurrence of reflux symptoms much later (9 months after surgery) and subsequently received Toupet fundoplication.

Hiatal hernia (HH) recurrence

HH recurrence occurred in 1.33% (n = 8/602) of cases, was the most common serious AE in our analysis, and this low frequency is probably a result of the intra-abdominal anchoring apparatus created by the procedure³¹. All patients underwent successful HH repair, 7/8 by straightforward repositioning of the fundus combined with extended dissection (with the implant intact and untouched in its pouch), followed by new hiatoplasty. All patients were found at surgery to have a total failure of the crural repair with three cases known to be caused by intense retching/vomiting in the postoperative period. It is essential to instruct the anesthesiology department to provide stringent postoperative care to avoid retching/vomiting right after surgery. Three cases had a large hernia (mean 7 cm) preoperatively. In one patient, the device was explanted, followed by conversion of the RefluxStop plication to a Dor fundoplication, due to an obesity-related weak crura. All patients recovered fully. Standardized radiologic assessment was performed right after surgery for all 602 patients and thereafter only in cases of symptom-suggestive treatment failure. Index surgeries generally did not use mesh, but reoperation cases typically involved mesh. The number of mesh cases overall were in any case too few to draw conclusions regarding HH recurrence and device-related complications. One learning point from these events is that in the postoperative period retching and vomiting can totally disrupt the hiatal repair acutely, and therefore, all precautions must be taken to avoid this.

A recently published report of RefluxStop has demonstrated favorable treatment effect regardless of HH size at baseline: small (< 3 cm) or large (4–10 cm)¹⁹. In the 5-year results of the RefluxStop pivotal trial it was objectively verified (via contrast swallow x-ray imaging in all subjects) that no case of reherniation had occurred at the 5-year follow-up, although in a group with hernia ≤3 cm¹⁷.

After fundoplication with sutured hernia repair, symptomatic wrap herniation (9.4%) results in 18% of reoperations at a follow-up of 6–36 months, according to an analysis of 9433 cases by Carlson et al.³⁸. HH recurrence rates in the range of 10–20% (or higher) have been reported after traditional LARS, dependent on the series and definition³⁹, even exceeding 50% for long-term radiological recurrence^40,41. The overall recurrence rate in Armijo and colleagues’ study with a mean follow-up of 19.9 ± 23.8 months was 15.5%⁹.

According to Linnaus et al.⁴², most HH recurrences manifest through the anterior/upper-left aspect of the hiatal repair. The RefluxStop procedure, on the other hand, deductively provides a platform for intra-abdominal anchoring of the ARB by diaphragmatic interaction at the posterior/lower-left aspect of the hiatus (presumably less impacted by hiatal collapse) based on the RefluxStop procedure’s surgical design, including providing a large fundic package (dorsally and left-laterally positioned) of the implant and invaginated gastric wall, where both size and position seem to be features that may reduce HH recurrence. This study had a low rate (1.33%) of HH recurrence, however, radiologic evaluation was only performed in symptomatic patients and is a necessity for a comprehensive illustration of this relationship. Logically, a total circumferential collapse of HH repair, as was the case in all the reported subjects (n = 8/602) in this study, may be required for such a sizeable hiatal opening to allow reherniation of RefluxStop, based on reasons outlined above. As per Linnaus et al.⁴², a larger hiatal opening often occurs with circumferential collapse, which independent of size occurs in 29% of cases but what proportion have larger circumferential infirmity is unclear.

No other serious event occurred with frequency > 1/602 patients

The safety profile of the RefluxStop procedure is underlined by the fact that most AEs related to the learning experience occurred in only 1/602 cases (0.17%).

Explantation/removal of device

Explant operations of the RefluxStop device occurred in 0.5% (n = 3/602) of cases and for three different reasons, with some learning experience to be made. First, it is important to ensure the channel after the deployment tool is properly closed before completing the invagination step. The stomach wall may expand substantially causing a small hole to become larger. Second, reherniation is not necessarily a reason for explant of the device; instead, repositioning the fundus with the device intact in its pouch and performing a new hiatal repair are feasible. As mentioned, it is critical to avoid postoperative retching/vomiting directly after operation. Third, we have seen a case of ulcer at the wall of the RefluxStop invagination, successfully treated with double-dose PPIs, although the ulcer was deeply penetrating to the device.

Removal of the MSA device was required in 6.7% of cases at a median follow-up of 113 days (4 months), according to a recent US study, largely due to postoperative AEs (i.e., esophageal perforation, dysphagia, or persistent symptoms)⁴³. Another recently published study shows an explantation rate of 12.6% for MSA at a median follow-up of 39.5 months, mainly due to dysphagia and migration¹¹.

Esophageal dilatation

There was only one case requiring endoscopic esophageal dilatation for new-onset dysphagia, due to cruroplasty technique (too tight), that resolved after dilatation without affecting the device or treatment to any degree. For remaining preoperative symptoms, a few dilatations occurred in 1/22 centers (with 66% IEM subjects), but in the other 21 centers no dilatation occurred. Esophageal dilatation after LARS is commonly performed for postoperative dysphagia and occurs in 12.4% of fundoplication cases⁴⁴. Other studies have reported endoscopic dilatation for dysphagia in 10% of LNF cases and 16% of MSA patients at 1 year⁴⁵. Post-MSA esophageal dilatation rates are high due to associated severe dysphagia, reported between 16 and 50%^13,45. To reduce the incidence of troublesome dysphagia with MSA (via larger device sizing), a presumable trade-off is made by less acid reflux treatment effect, while dysphagia is still frequent in larger sizes^11,32. RefluxStop employs a mechanism that does not involve esophageal encirclement for normal LES function aiming to minimize postoperative dysphagia, as supported by the pivotal study results reporting no esophageal dilatations^16,17.

Total reoperations for serious safety outcomes (1.99%)

Reoperation increases morbidity and risk from the patient perspective but also consumes healthcare resources. The number of reoperations for serious safety outcomes in this study, as shown in Table 4, demonstrates a low incidence of 12/602 cases (1.99%). Out of these 12 cases, eight were related to reherniation, wherein 7/8 had the fundus repositioned in a straightforward procedure with new hiatal repair as previously described. The mean follow-up of 2 years in this study included severe GERD sufferers with up to 6.75 years of clinic visits in a real-world setting.

The systematic literature review by Zehetner et al. reported high rates of reoperation after LNF: 1 year (6.7%) and 10 years (16.3%)⁷. This aligns well to a separate systematic review with meta-analysis by Broeders et al., that found 7.0% of LNF patients required reoperation at 1-year follow-up (range 12–60 months)⁴⁶. Magnetic augmentation may be associated with severe dysphagia⁴⁷ or erosion that necessitates reoperation via device removal/explantation¹¹. The literature shows an even higher rate (10%) of reoperation following MSA when compared to the standard of care, LNF (p = 0.022), at 1-year follow-up³⁵.

Impact on the healthcare system

There are several published health economics studies involving the comparison of RefluxStop to standard treatments for acid reflux/GERD in different European countries, showing the overall impact of various treatment options on the healthcare system including both surgical and medical treatment^48–51. Crucial to these investigations are the advanced analytical methods using clinical data on safety outcomes. Several studies modeled for the UK, Italy, Switzerland, Sweden, Norway, and Spain have shown that RefluxStop was more cost-effective against standard-of-care medical treatment with PPI, Nissen fundoplication, and MSA, providing clinical benefit by reduced healthcare use for side effects at a marginal budget impact^48–53.

Thus, from the European perspective, healthcare expenditure per annum for GERD and lost productivity due to illness are substantial. The favorable safety profile of RefluxStop surgery is well-aligned with findings in the research performed and supports that this procedure is likely to also provide substantial economic benefits to global healthcare systems through improvement in clinical outcomes, particularly in regions with high rates of complicated disease features such as IEM and obesity^48–51. For example, the US has a large burden of disease where approximately USD $15–20 billion is spent on annual expenditure for direct and indirect costs related to reflux disease⁵⁴.

Strengths and limitations

The present analysis, which pooled safety data from 22 multiple centers in Europe, presents the largest cohort to date of 602 patients that underwent the RefluxStop procedure. This sample size provides more clarity to the overall understanding of RefluxStop and is more representative and generalizable than single-center studies, as well as less influenced by the outcomes of outlier patients or individual surgical practices. Surgeons in general were experienced in upper gastrointestinal surgery, but new to this technique, so there was undeniably a degree of learning curve effect during early experience, even beyond initial training. To some extent, this may be offset by surgeons’ overall experience but was nevertheless an influence on results. Some minor variability in procedural details may have occurred, both between centers and over time based on clinical experiences, which was not possible to account for. To minimize variability in execution, a surgical case review and proctoring system had been implemented to ensure optimal technique, where all surgeons in this study participated in the training program. After the initial launch in Europe, surgeons began inclusion of patients with large hernias, high BMI, and IEM in their clinical practice with RefluxStop. This makes for a population that is somewhat heterogeneous but arguably more reflective of real-world practice and thus more generalizable to the overall GERD population.

Another limitation is that this report is focused on safety outcomes alone and excluded any non-safety outcomes (i.e., efficacy). However, dysphagia (which is the most important outcome for patients) is well-reflected in the focus on esophageal dilatation, which is arguably a more objective and relevant factor when determining safety outcomes that necessitate further healthcare utilization but is also nonetheless influenced by inter-institutional practice differences. Previous studies by individual centers included in this study have presented such patient-reported outcomes^{14–18,21,26}, but is beyond the scope of this investigation focused on safety and technical feasibility. Finally, since many of the included patients were diagnosed with GERD and underwent RefluxStop surgery years ago, the Lyon consensus 2.0 from 2024 was not used for consistency.

Although this study may include potential selection bias and heterogeneous follow-up inherent to real-world practice, inclusion of 22 independent centers with mid- to long-term safety follow-up somewhat mitigates these limitations via diversification of clinical data. Standardized functional outcome assessment was absent due to the focus on safety outcomes. Furthermore, radiological evaluation was performed directly after surgery and thereafter only for symptom recurrence.

Conclusion

The independent RefluxStop Safety Evaluation Group focused on safety data of 602 patients, from 22 centers across Europe, having mean 2 years (2 months-6.75 years) of follow-up and concluded that the safety profile of RefluxStop surgery in real-world settings is highly favorable in the surgical treatment of GERD, including considerable portions of IEM and large HH patients. The total reoperation rate for serious safety outcomes was 1.99% (n = 12/602), with two-thirds due to HH recurrence based on total disruption of the crural repair, thus unrelated to RefluxStop, with all cases resolved by a standard approach. All explants (0.5%) were considered part of the learning curve of a new procedure. In 21/22 centers, no esophageal dilatations were performed. Notably, all safety events were successfully resolved.

Prospective studies on mid- to long-term clinical effectiveness have already been published, including RefluxStop data from the 5-year study used in the PMA submission to the FDA. Avoidance of esophageal encirclement or pressure application, as well as procedural invagination of the device in a free-hanging pouch, are key features designed to limit adverse outcomes with this procedure, potentially signaling an edict shift in the surgical treatment of GERD in terms of its low-risk profile.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

All authors collected institution-specific clinical data, participated in analysis, and contributed to writing/editing/review of the manuscript.

Data availability

Anonymized center-specific incidence data are uploaded as a supplementary file. The corresponding author can be contacted for any specific patient-level questions or information.

Declarations

Competing interests

SFS declares receiving travel grants from Implantica (Zug, Switzerland). ME, TL, YB, SFS, JZ, and JCL receive travel and conference presentation honoraria for education activities from Implantica (Zug, Switzerland). AA, NFG, MEBC, AHM, JDSL, MCA, BGO, CMM, CMS, LV, and GP declare no conflicts of interest.