This systematic review (SR) included 16 studies. GCF was most prevalent in LSG, accounting for 84.6% of the studies, one case after the gastric band was removed and converted to LSG, and one biliopancreatic diversion with duodenal switch. In the case reports, it was not diagnosed after RYGB and OAGB. In a cohort study, only one RYGB was described as a complication arising from a secondary complex GCF following stent placement for a primary cardia leak. The transverse colon was identified as the most common site of occurrence, and exploratory laparoscopy was utilized as the final treatment or rescue surgery.

One of the primary causes of GCF is the failure to maintain proper staple lines and anastomotic integrity during MBS. Early detection and management of leaks or infections are crucial in significantly reducing the risk of fistula formation. However, this systematic review revealed that early detection was infrequent, highlighting a critical area for improvement in clinical practice.

The time from the index surgery to the development of GCF varies widely. Acute studies (25.0%) can develop GCF as early as 10 days post-surgery (in one RYGB case of secondary complex GCF), while chronic studies can take up to 9 years, with 75% occurring more than 6 months post-surgery. The time from diagnosis to closure of the GCF also varies significantly. In acute studies, closure occurs between 14 and 61 days, while in chronic studies, it ranges from 1 to 12 months. Previous systematic reviews (SRs) conducted by Sakran et al. and Bawa et al. reported differing resolution times for fistula formation in the chest and gastrocutaneous fistulas, respectively. Bawa et al. documented a median time to closure for gastrocutaneous fistulas of 81 days. In contrast, Sakran et al. reported a wide range of closure times for chest fistulas, spanning from 2 days to 7 years, with 47.8% of cases remaining unresolved beyond six months [6, 26]. Thus, across all SRs, the resolution times for different types of fistulas were unpredictable, varying significantly between short and long periods.

The unpredictable evolution of leaks and fistulas makes management challenging. GCF, in particular, should be considered a late possible complication rather than an acute disease, given its potential to manifest long after the initial surgery.

Due to its anatomical position, leaks after LSG typically occur in the upper part of the staple line. The published incidence of leaks in LSG is between 0.1 and 7%, with a decrease in recent years, primarily due to experience and the standardization of the surgical technique [4, 5]. These leaks can result in the persistence of collections below the diaphragm, leading to pathological communication between the stomach and the transverse or left-angle colon, thereby causing GCF.

If these leaks are refractory or left untreated, they are prone to become persistent and evolve into chronic fistulas. This progression is often due to the increased intraluminal pressure within the stomach, a result of its tubular shape post-surgery, even in the absence of strictures at the gastric incisura [27].

It is suggested that proper surgical techniques are essential to prevent such complications, including high-quality staplers and thorough intraoperative leak testing. While our study did not have access to specific data regarding the types of staplers used in primary surgeries, we acknowledge that the quality of staplers can vary significantly between manufacturers, even though all must meet basic regulatory standards such as CE marks. This variability in quality could influence surgical outcomes, although this hypothesis was beyond the scope of our current study and requires further investigation. Nevertheless, ensuring reliable, high-quality staplers is critical in minimizing the risk of leaks. Additionally, gentle handling of tissues to reduce trauma, reinforcement of the staple line, and ensuring an adequate blood supply are crucial to maintaining anastomotic integrity and promoting healing during MBS. While these measures can significantly reduce the risk of complications, they cannot provide complete protection. The anatomical location of upper gastrointestinal tract leaks and their proximity to critical abdominal and thoracic structures accounts for the heightened risk of fistulization when these leaks are not promptly diagnosed and managed [15, 28].

Nevertheless, the relationship between the anatomical location of upper gastrointestinal tract leaks and the onset of GCF in this SR remains unclear. We can hypothesize that tissue healing was complete, as most studies occurred more than six months after the index surgery. Despite this, the formation of fistulas was consistently observed along the staple line to the transverse colon in 31.3% of the studies described in this SR. This pattern suggests that factors beyond initial tissue healing, such as the mechanical stress at the staple line or persistent subclinical inflammation, may play a role in the development of GCF, but it stays unclear. For other fistulas that develop after MBS, gastrocutaneous fistulas commonly form from the proximal third of the gastric remnant. This condition is associated with extended hospitalizations and significant morbidity and mortality, with an incidence reported to range between 1.7 and 4.0% [29, 30]. Gastric fistulas in the chest following LSG present unique challenges due to the distinct physiology of the sleeved stomach. The high-pressure system, resulting from an intact pylorus at the distal end and a lower esophageal sphincter at the proximal end, contributes to the increased risk of persistent leaks or fistulation into adjacent anatomical compartments. Additionally, several other factors can contribute to the formation of gastric fistulas, including iatrogenic injuries, improper vascularization, ischemia, hematoma formation, technical issues, and staple misfiring [6, 31, 32].

The management of GCFs typically necessitates a multimodal approach, incorporating both non-surgical and surgical interventions based on the severity and complexity of the condition. Non-surgical management is the preferred initial strategy for well-controlled leaks without hemodynamic instability. This approach includes drainage, broad-spectrum antibiotic therapy, and adequate nutritional support to control local sepsis and facilitate healing [33, 34]. This SR confirms the use of a nasojejunal tube for enteral nutrition. Nevertheless, this SR shows that exploratory laparoscopy was necessary in almost all studies (81.3%). And seven out of 16 studies (43.75%) conducted another intervention first (stent, dilation, over-the-scope clip) before converting to exploratory laparoscopy.

A systematic review by Hernández et al. reported an overall success rate of 85.89% (95% CI, 82.52–89.25%) for leak closure, with a median interval of 44 days between stent placement and removal. Stent migration occurred in 18.65% of cases (95% CI, 14.32–22.98%), while the proportion of re-operations was 13.54% (95% CI, 9.94–17.14%), typically involving rescue surgery for leaks that failed to resolve with stent management. The review highlighted that most leaks were at the angle of His In LSG and some RYGB procedures. The authors concluded that the endoscopic placement of self-expanding stents is an effective option for managing leaks after MBS in selected patients, demonstrating high efficacy and low associated mortality rates [35].

Another SR examined the effectiveness of Endoscopic Internal Drainage (EID) treatment in MBS. The overall success rate for EID treatment was 91.6% for leak closure, with a median treatment duration of 78.4 days (50.1–106.7 days). Complications associated with EID were reported in four studies and included stenoses (n = 8), perforation (n = 1), esophageal ulceration (n = 3), bleeding (n = 2), and splenic hematoma (n = 1). Across all studies, 232 patients were included, with 80% having undergone LSG and 20% having undergone RYGB [36]. Nevertheless, reducing stent migration and re-operation rates is a significant challenge in both SRs [35, 36].

This SR was shown in the study by Donatelli et al., where internal pigtail placement in RYGB was used and the one case of GCF. A complication led to the development of a complex GCF, which required endoscopic necrosectomy, pigtail resenting, and a nose-jejunal feeding tube for four more weeks before being put on a regular diet and definitive removal of the pigtail for an overall 99 days of treatment. The GCF healed but persisted in a sub-clinic GGF.

One other study by Trellis et al. used a nitinol silicone-covered stent, but this needed a laparoscopic repairment to treat and heal the GCF [24]. In the cohort study by D’Alessandro et al., 52.5% of cases experienced failure of one or more double pigtail catheter treatments after 12 months, necessitating laparoscopic intervention as rescue surgery in all such instances [23]. Although other SRs have demonstrated positive responses to stent and drain treatments for leaks and fistulas using various diagnostic methods, this SR can conclude, even with the limited amount of evidence and power of the studies and the late onset and challenging anatomical position of GCF, stent placement appears to be neither recommended nor sufficiently effective.

The OTSC system is an endoscopic technique involving a clip to close the fistula opening. Studies have demonstrated its effectiveness in studies where conventional stenting fails or is not feasible.

An SR by Shoar et al. tested the OTSC in managing leaks and fistulas in patients who underwent LSG. The time between LSG and leak/fistula ranged from 1 to 803 days. Most of the leaks/fistulas were located at the proximal staple line, with sizes ranging from 3 to 20 mm. The time between leak diagnosis and OTSC clipping ranged from 0 to 271 days. Thirty-three out of 53 patients (63.5%) required one clip to close the lesion.

Regarding OTSC-related complications, a leak occurred in five patients (9.3%), and OTSC migration, stenosis, and tear occurred in one patient each (1.8%). Of the 73 patients with post-LSG leaks treated with OTSC, 63 had an overall successful closure (86.3%). The OTSC system is a promising endoscopic approach for managing post-LSG leaks in appropriately selected patients. Unfortunately, most studies are series with small sample sizes, short-term follow-up, and mixed data of concomitant procedures with OTSC. Further studies should distinguish the net efficacy of the OTSC system from other concomitant procedures in treating post-LSG leaks [37].

This SR included one study by Delong et al. that used OTSC. In this case, Ovesco clips were applied to both ends of the fistula tract, and post-procedure fluoroscopy did not demonstrate a leak after 62 days [13]. A study by Garofalo et al. also used OTSC but failed due to the fibrotic nature of the fistula. Eventually, laparoscopic resection of the gastrocolic fistula with omental interposition and perioperative endoscopy was necessary [20].

Given that the systematic review by Shoar et al. was unable to draw definitive conclusions about the effectiveness of OTSC and excluded GCF from its analysis, coupled with findings from this review indicating its use in only 2 out of 100 patients with a success rate of 50%, OTSC does not appear to be the first-line treatment choice. Its application for GCF remains highly debatable.

When endoscopic methods are unsuccessful or contraindicated, alternative treatments must be considered. Endoscopic procedures may be contraindicated in studies of severe hemodynamic instability, extensive tissue necrosis, severe infection or sepsis, anatomical abnormalities that prevent safe endoscopic access, or when there is a risk of perforation that cannot be managed endoscopically; surgical repair remains the definitive treatment for GCF.

Surgical options may include resection of the fistula tract, re-anastomosis, and staple line reinforcement. In severe studies, a more extensive procedure, such as partial gastrectomy or colectomy, may be necessary. Surgical outcomes can be favorable, but the approach is associated with higher morbidity and longer recovery times. Predictors of success in surgical treatment include the patient’s overall health status, absence of severe associated medical problems, the timing of the intervention, and the surgeon’s expertise. Early identification and management of complications, as well as implementing a multidisciplinary care approach, also contribute to improved outcomes.

In terms of results, successful surgical interventions often lead to the resolution of the underlying issue, but extended hospital stays and increased morbidity rates typically accompany them. The duration of hospital stay can vary significantly, depending on the complexity of the surgery and the patient’s postoperative recovery. Morbidity rates can include infections, prolonged wound healing, and other surgery-related complications. Specific data on these outcomes were not clear or available in this SR; therefore, it is important to acknowledge these limitations and suggest areas for future research to better understand and optimize surgical treatment strategies.

This SR revealed that, in 81.3% of studies, surgical intervention was necessary as the primary choice for complication management or as a rescue operation. All included studies consistently described exploratory laparoscopy, which involved surgical treatment through various forms of fistula resection and suturing.

Given the variability in treatment approaches and the absence of a single best treatment strategy for this rare complication, it is essential to emphasize the fundamental surgical principles that should guide management. These principles include optimizing the patient’s nutritional status, resecting devitalized or infected tissue, ensuring effective drainage, and creating a tissue buffer between the affected organs. By adhering to these core principles, surgeons can tailor their approach to the specific clinical circumstances of each patient, potentially improving outcomes despite the challenges posed by the complexity of these studies.

Among the reviewed studies, Tan et al. attempted to develop an algorithm-based approach for diagnosing and managing leaks, including one GCF following LSG. Their management strategy incorporated laparotomy, laparoscopy, endoscopic covered stenting, percutaneous radiologically guided drainage, jejunal enteric feeding, and total parenteral nutrition. While this approach could inform future studies, the study needed more statistical power to be considered robust. The proposed algorithm was based on a small cohort of only 14 patients treated with varying methods, and it remains to be seen how the single GCF was explicitly managed. Considering each procedure’s unique benefits and limitations, only larger, well-matched cohorts with reliable follow-up can provide definitive guidance for such management strategies [25].

Another systematic review by Bawa et al. on gastrocutaneous fistulas detailed the steps involved in the investigation and management strategy. All included studies initially attempted medical management, comprising antibiotics, skin protection, and artificial nutrition [26]. In this SR, an explicit treatment strategy was outlined. Specifically, 75.0% of the studies employed a nasojejunal tube for enteral nutrition to optimize the patient’s nutritional status, and 37.5% reported using antibiotics as a pre-treatment measure. Evidence from other SRs supports these approaches, indicating they are recommended practices [26]. However, it is essential to note that these interventions with antibiotics were not consistently presented across all studies in this review or other SR [38]. Therefore, implementing nasojejunal tube feeding for nutritional support and using antibiotics should be considered as primary actions in managing gastrocutaneous fistulas.

Furthermore, this SR could not build any form of algorithm, as 13 out of 16 studies were single-case reports. One cohort had 33 patients, with only one GCF case as a secondary complication, and another cohort was unable to provide conclusive results, as over 52.5% of patients who failed one or more double pigtail treatments after 12 months required laparoscopy as rescue surgery.

The primary limitations of this review include the variability in surgical techniques and postoperative management protocols and significant missing data regarding patient signs and symptoms, antibiotic treatment, hospital stay, and other clinical parameters. These are crucial for developing any predictive model and complicate the generalization of results to perform a meta-analysis.

Additionally, the predominance of single case reports and small sample sizes introduce a high risk of bias and confounding factors. The retrospective nature of most studies and the lack of comprehensive clinical data, such as patient signs, symptoms, and antibiotic treatments, hinder the development of reliable predictive models and treatment algorithms. Furthermore, we excluded marginal ulcers, which could have required a different treatment approach in various other MBS procedures. Larger, well-matched cohorts with reliable follow-up are necessary for more definitive conclusions.