Foregut Erosion Related to Biomedical Implants: A Scoping Review

Introduction

To enable surgeons standardize surgical procedures and improve patient-reported outcomes for common medical conditions, biomedical devices implanted transabdominally have gained popularity over the past 50 years in the treatment of gastroesophageal reflux disease, hiatal hernia, and morbid obesity. The Angelchik antireflux prosthesis (AAP) was first introduced in the seventies for treating gastroesophageal reflux disease. ¹ Subsequent innovations were the adjustable gastric band (AGB) for morbid obesity, various types of mesh for crural reinforcement, and the magnetic sphincter augmentation device (LINX^®). Biomaterials used at/near the esophagogastric junction have included silicon, polypropylene (PP), polyester, polytetrafluoroethylene (PTFE), and titanium. Device-related foregut erosions (FEs) represent a challenging event that demands special attention owing to the potential of severe postoperative complications and even death. Full-thickness endoluminal erosions occur when externally implanted devices penetrate the mucosal lining of the upper gastrointestinal tract. The causes of erosion are multifactorial and may involve defective design/testing, manufacturing errors, or surgical implantation issues leading to pressure, ischemia, infection, or exaggerated foreign body reactions triggered by the device.^2–4 Understanding the complexities and the dynamics of device-related FE is crucial and clinical vigilance remains important even in patients with long-term implants since late complications can potentially be reported even after many years the device is dismissed. ⁵

We provide an overview of full-thickness foregut injury leading to erosion associated with four types of biomedical devices that have been used in the past or are still in use today for the treatment of gastroesophageal reflux disease, hiatal hernia, and severe obesity.

Materials and Methods

This scoping review followed the methodological framework proposed by Arksey and O’Malley, Levac and colleagues, and the Joanna Briggs Institute.^6–8 This framework includes the following six stages to guide scoping review processes: (1) specifying the research question, (2) identifying relevant literature, (3) selecting studies, (4) data mapping, (5) summarizing, synthesizing, and reporting the results, and (6) expert consultation (optional). Our review was reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) Checklist (Supplementary Table S1). ⁹ Since the study included a review of published articles and study-level results, institutional review board approval or exemption was not required.

Results

Literature search

The literature search in this review yielded 908 records from all databases (Fig. 1). After duplicate removal, 861 records were screened. After title and abstract assessment, 136 full-text articles were finally considered. Four articles were finally excluded after full-text assessment because of patient overlap. Overall, 132 articles were finally included for a total of 1292 patients suffering from device-related FE. All included studies were carried out between 1982 and 2023. The majority were case series and case reports, while the remaining were observational cohort studies (n = 27).

OPEN IN VIEWER

FIG. 1.

PRISMA diagram illustrating the article selection process for the scoping review.

Discussion

This study shows that device-related FE is relatively rare but it may occur even many years from the implantation. Erosion has rarely been associated with mortality; however, significant morbidity was reported, and esophagogastric resection was required in some patients especially after nonabsorbable mesh-reinforced cruroplasty (Table 5) (Supplementary Data S1).

The dynamics of soft-tissue pressure injuries associated with medical devices have long been investigated, and preventive strategies have been effective in mitigating the severity of these lesions. ¹⁴⁴ Regarding device-related FE, other factors besides mechanical pressure and ischemia can determine visceral injury. Similar to retained surgical gauzes that have the potential to erode and completely migrate into the bowel lumen without any apparent opening of the intestinal wall, ¹⁴⁵ biomedical implants stimulate a foreign body response by cytokine and chemokine trigger, neutrophil/macrophage activation, fibroblast migration, and collagen deposition with formation of a fibrotic capsule. This walls off the implant and prevents it from interacting with the surrounding tissue, but the effect of pressure and ischemia on the gut wall may still cause erosion. Factors such as biomaterial properties (hydrophobic versus hydrophilic), surface geometry, porosity, and mechanical mismatch are important features determining the foreign body response to the embedded device. Smooth/flat silicon devices seem associated with greater foreign body giant cell adhesion and a thicker fibrotic capsule.^146,147 Additionally, implant of a biomedical device can set the stage for potential microbial colonization and infection with biofilm formation and subsequent host inflammatory response.^148,149 Other well-known risk factors for erosion into the digestive tract are the inadvertent/unrecognized visceral injury or compromised blood supply, which is more likely to occur especially after previous esophagogastric surgery, or the implant of devices in close proximity of staple lines. ⁹⁵ Interestingly, animal experiments have shown that foreign biomaterials implanted into the peritoneal cavity under sterile conditions can become colonized by bacterial translocation from the gut lumen.^150,151 Woodcock showed that bacterial translocation may represent a rarely recognized source of contamination or infection of aortic prosthetic grafts, probably through an increase of intestinal permeability. ¹⁵² Gram-negative Escherichia coli bacteria are the most commonly isolated pathogens. ¹⁵³ Although bacterial clearance can be seen in mesenteric lymph nodes and other organs, inert materials such prosthetic grafts do not undergo bacterial clearance. ¹⁵⁴ Finally, immune-mediated mechanisms have been advocated in generating systemic side effects, but the role of sensitization or allergy in determining visceral erosion remains unclear. ¹⁵⁵

Transmural erosion is an insidious process that may be asymptomatic or manifest with variable symptoms and signs. ¹⁵⁶ Symptoms associated with device-related FE include abdominal pain, new-onset dysphagia, and gastrointestinal bleeding; low-grade fever and laboratory signs of systemic infection may also be present. Early recognition and accurate diagnosis of device-related FE involve a combination of radiological and endoscopic studies. Management strategies for device-related FE include device removal through a surgical, endoscopic, or combined approach. In most circumstances, there is no free communication with the peritoneal cavity, and partial or total endoscopic retrieval is feasible.

The AAP, first used in 1973 and approved by the Food and Drug Administration in 1979, became initially popular due to its simple design and ease of insertion. The AAP is a C-shaped elastomer shell ring filled with silicone gel, 46 g in weight, 7 × 3.3 × 2.8 cm in size, with a radiopaque circumferential tantalum strap, and Dacron retention ties to hold the two ends of the “C” together. The device was placed through laparotomy around the gastroesophageal junction, with the Dacron straps tied together. Despite the initial positive expectations, utilization of the AAP progressively decreased due to widespread reports of complications and had become unpopular in the early 1990s coinciding with the start of the laparoscopic era. ¹⁵⁷ The most commonly reported complications were dysphagia (40%), followed by device migration and device erosion. In one of the few long-term reports, device removal was required in 15% of the patients most frequently due to migration, slippage, persistent dysphagia, or erosion. ²² It has been hypothesized that leakage of low-molecular-weight silicon droplets through the elastomer shell was responsible for capsule formation since silicon concentration in the capsule increased over time. Generally, minimal inflammation and no granuloma was found. The long-term host response and the structural integrity of the AAP remain unknown.^158,159 Overall, it is estimated that more than 30,000 Angelchik procedures have been performed worldwide, and thus, it is still possible to encounter patients implanted with this device in current surgical practice.

The AGB was introduced in the 1980s. Between 2011 and 2018, the popularity of laparoscopic AGB declined over time from 35.4% to 1.1% of total bariatric procedures in the United States. A systematic review published in 2013 and including 25 studies and more than 15000 patients found a 1.4% erosion rate. ⁹⁶ There was large interstudy variability regarding the incidence of erosion, from .2% to 33%, with a median time to event of 33 months (11–170). The erosion rate was higher (6.7%) with perigastric implant and dropped to 1% with the pars flaccida approach. Overfilling, 10 cm band, and gastrogastric sutures were potential risk factors for erosion. The clinical course of FE after gastric banding is usually benign, and the first step should be surgical/endoscopic removal. Delayed replacement is associated with maintenance of weight loss and re-erosion has been uncommon. Delayed conversion to Roux-en-Y gastric bypass (RYGB) or sleeve gastrectomy is feasible.

Nonabsorbable mesh for hiatal reinforcement was introduced in the 1990s. Jansen et al. first demonstrated PP mesh erosion into the esophagus in an experimental model and found that the distance from the edge of the mesh and the esophageal wall should be greater than 3 mm. ¹⁶⁰ In 2009, Staldhuber et al. reported 28 cases of mesh-related complications after crural repair. ³⁶ Mesh erosion was a common complication. Cheng et al. reviewed 50 cases of hiatal mesh erosions reported between 1998 and 2019. The latency period ranged 1 month–20 years, with 79% of the events occurring within 2 years. The most common mesh was PTFE and PP. Thirteen (26%) patients necessitated open esophagogastric resection. In 2019, Sanchez-Pernaute et al. reported on 122 consecutive patients undergoing circular dual-mesh hiatal repair between 2005 and 2016. ⁵³ The absolute erosion rate was 4.9%. The erosion rate after recurrent hiatal hernia repair was greater (16%) compared with primary repair (4%), and 1 of the 6 patients needed esophageal resection. The new synthetic absorbable types of mesh are expected to protect from erosion over a medium-term follow-up period. Our review confirms that no erosions have been reported so far with synthetic absorbable meshes.^161–166

The magnetic sphincter augmentation (LINX ^® ) device was introduced in 2007. It is a nonbulky device (1.2 g) composed of a series of beads designed to be noncompressive by its roman-arch configuration. The device is implanted around the esophagogastric junction after tunnelization between the posterior vagus nerve and the esophagus. Proper sizing is critical to avoid compression on the esophageal wall. Current American College of Gastroenterology guidelines for gastroesophageal reflux disease (GERD) recommend the LINX ^® procedure as an alternative to laparoscopic fundoplication for patients with regurgitation who fail medical management (strong recommendation, moderate level of evidence). ¹⁶⁷ Cumulative published rates of revisional procedures reflect evolution in patient selection, sizing technique, type of mediastinal dissection, type of hiatal repair, and postoperative management. In one large series, a drop in the explant rate to 5% has been noted after 2014 due to changes in sizing technique and systematic crura repair. ¹⁴³ Erosion rates based on the MAUDE database range from .1% to 4.1%. The majority of erosions were associated with size 12. Erosion represents the third cause for device removal, with a mean time to explant of 25 months and a 7-year cumulative erosion risk of .28%. ¹⁶⁸ The eroded LINX ^® device can be safely removed either laparoscopically or endoscopically. It is well known that titanium implants stimulate a foreign body response, which walls off the titanium with a fibrous capsule. Exuberant scar tissue contraction may cause stricture and infection and may lead to erosion.^169–171 The most likely risk factor for erosion is undersizing the device that makes it compressive around the distal esophagus. Erosion with the magnetic augmentation device remains a rare event, has a benign clinical course, and may be prevented by proper surgical technique and sizing.

Sudden onset of dysphagia, severe heartburn, epigastric pain, or low-grade fever should alert proper investigation after implant of the LINX ^® device. Individual titanium beads can be visible endoluminally on upper endoscopy. No signs of perforation generally occur due to the protective effect of pericapsular fibrosis. Nothing per os, antibiotics, and planning elective endoscopic or laparoscopic removal are recommended once erosion has been endoscopically proven.

In the past, device-related FE was managed via urgent open surgery and device removal. In some patients, more extended surgical procedures, including visceral resection, were necessary because of the strong adhesions and perivisceral fibrosis, which made sometimes impossible safe and complete retrieval of the device. It appears from this study that resections were more common in the setting of eroded PTFE and PP mesh. Recent advances in minimally invasive techniques have allowed a progressive shift toward less invasive approaches leading to reduced postoperative complications and shorter length of hospital stay. Furthermore, a wait and see or a two-stage approach has been advocated in selected asymptomatic patients. ¹⁰² Upfront endoscopic removal may be feasible and successful when 50% or more of the eroded device is inside the visceral lumen. From a technical standpoint, a metal cutting wire may be appropriate to cut the eroded AAP or the AGB. Bleeding has been reported as a complication of endoscopic AGB removal. A safe option for endoscopic explant of the LINX ^® device is the Ovesco^® remove system. An independent, next to the scope intraluminal grasping device can be inserted to assist during division of the connecting wire links between beads. The device is then grasped with an endoscopic loop and removed transorally.

Historically, most implantable medical devices have been introduced into clinical practice despite the lack of robust evidence to support effectiveness as indicated by the IDEAL guidelines. ¹⁷² Hopefully, pilot trials, randomized clinical trials, and international registries to monitor surgical procedures will allow safer widespread adoption of biomedical devices in the future. Interestingly, the miniaturized design of new technologies such as the RefluxStop^173,174 allows the silicon device to fall apart into 4 small pieces and easily pass the pylorus in case erosion of the gastric pouch occurs, thus avoiding the need of surgical or endoscopic removal.

This study does have some limitations. First, underreporting cannot be ruled out, and additional information may have been missed also because of the exclusion of non-English language literature. Second, reporting of outcomes was inconsistent across the studies, while almost 51% of included reports were classified as having a moderate-high risk of bias. Third, undetected occult erosions may have occurred in asymptomatic patients not undergoing regular follow-up assessment; therefore, the determination of the global incidence for erosion related to each single device is arduous.

Conclusions

This study shows that device-related FE is rare, but it may occur even many years from the implantation of AAP, nonabsorbable mesh, adjustable gastric banding, and magnetic sphincter augmentation devices. Mortality-related FE is infrequent, however, increased postoperative morbidity and need for esophagogastric resection were observed for nonabsorbable mesh-reinforced cruroplasty. The search for novel minimally invasive therapies will continue in the future to improve standard of care, patient-reported outcomes, and possibly reduce complication rates. Lessons learned from the past indicate that safe implementation of new biomedical devices in clinical practice is crucial to protect patients, surgeons, and health care providers.