Introduction

Post surgical fistulas of the upper digestive tract(UDT) are associated with significant morbidity and mortality.

Aims & Methods

We aimed to evaluate the efficacy and safety of a primarily endoscopic management of post surgical fistulas of the UDT.
We performed a multicentre retrospective analysis in which 126 consecutive patients underwent primarily endoscopic management of post surgical fistula between March 2009 and January 2022. Factors associated with successful fistula closure and adverse events were evaluated. Qualitative variables were expressed as percentages and evaluated with Chi-square test. Quantitative variables were evaluated for normal and non-normal distribution. Non-normal quantitative variables were expressed as median (min-max) and evaluated with Mann-Whitney´s test. Logistic regression analysis was performed to determine factors associated with fistula closure and adverse events. P value <0,05 was considered statistically significant (IBM SPSS 28).

Results

The median age was 57(17-89) years and 73(60%) patients were female. Main Indications for surgery were obesity in 73(58%) and cancer in 46(36%) (of which 32(70%) were gastric cancer). The median time between surgery and fistula diagnosis was 7(1-387) days. Average size of fistula orifice was 7(1-30)mm.
Fistulas were located in the esophagus in 8 patients (6.3%), gastric cardia in 46 (36.5%), stomach non-cardia in 21 (16.7%), esophago-gastric or esophago-jejunal anastomosis in 9 (7.1%), gastro-jejunal anastomosis in 39 (31.0%), small intestine in 2 (1.6%) and unknown in 1 (0.8%).
Endoscopic interventions included placement of a covered metallic stent in 98(78%) patients (fully-covered n=46, partially covered n=37, both fully and partially covered n=15), clips in 58 (47%) patients (OTSC n=37, TTSC n=17, both OTSC and TTSC n=4) and argon plasma in 16(13%) patients.
Early adverse events (<1week) occurred in 22(17%) and included stent migration (n=10), sepsis (n=6), gastrointestinal bleeding (n=5) and death (n=1). Late adverse events occurred in 31(20%) patients, namely stricture (n=11), mucosal overgrowth (n=6), migration (n=4), vomiting (n=3), stent fracture (n=2), perforation and right hemicolectomy (n=1) and death (n=4). After a median follow up of 25 (1-147) months, 32(25%) died, but only 15(12%) due to fistula (3 due to aorto-esophageal fistula and the others due to sepsis), 12(10%) due to neoplasia and 5 due to other causes.
Need for surgical reinterventions (p=0.002) was associated with early adverse events. Late adverse events were less frequent in patients with clip placement (p=0.011).
Clinical success defined as definitive fistula closure was documented in 106(84%) patients after a median of 8(1-192) weeks and required a median of 3(1-13) endoscopies. Factors predictive of fistula closure were female gender (p=0.01), indication for surgery (p=0.02), lower C reactive protein(p=0.003), lower urea (p=0.006), lower AST(p=0.003) and lower ALT(p=0.002).

Conclusion

The primarily endoscopic approach to management of UDT post surgical fistulas was clinically successful in >80% of the patients with a good safety profile.

Introduction

Post surgical fistulas of the upper digestive tract(UDT) are associated with significant morbidity and mortality. The indication for surgery may influence outcomes.

Aims & Methods

We aimed to evaluate the clinical presentation,efficacy and safety of a primarily endoscopic approach to management of post surgical fistula of the UDT comparing oncologic versus non-oncologic indications for surgery.

Multicentre retrospective analysis involving 126 consecutive patients who underwent primarily endoscopic management of post surgical fistulas of the UDT between March2009-January2022. Clinical efficacy (fistula closure) and adverse events in patients with fistulas after oncologic and non-oncologic surgery were compared.Qualitative variables were evaluated with Chi-square test. Quantitative variables with non-normal distribution were expressed as median(min-max) and evaluated with Mann-Whitney´s test. We performed logistic regression analysis to determine factors associated with fistula closure and adverse events. A p value <0.05 was considered statistically significant (IBM SPSS28).

Results

Median age was 57(17-89) years and 73(60%) patients were female. Indications for surgery were non-oncologic in 80 patients(63.5%) ((mostly bariatric, n=73)) and cancer in 46(36.5%) patients.
Patients undergoing oncologic surgery were older and with male predominance. There were no differences regarding median time to detect fistula or fistula size between the groups. Fever, abdominal pain and abdominal collections were more frequent in non-oncologic surgery group. On laboratory analysis, only haemoglobin was significantly lower in patients undergoing oncologic surgery(Table1). We found no significant difference on leukocyte count, neutrophil percentage, C-reactive protein, AST, ALT, total bilirubin, INR or creatinine.

		Oncologic	Non-oncologic	p-value
Characterization	Median age	71(52-89),	45(17-79)	<0.001
	Female gender	21(26%)	35 (76%)	<0.001
	Median time to fistula detection(days)	8(1-83)	6(1-387)	0.08
	Fistula size(mm)	5.0(2.0-25.0)	5.0(1.0-30.0)	0.671
Clinical presentation	Fever	18(39%)	49(61%)	0.049
	Abdominal pain	13(28%)	46(58%)	0.006
	Abdominal collections	11(n=24%)	48(60%)	<0.001
	Sepsis	13(n=28)	21(26%)	0.309

Endoscopic interventions were similar in both groups and included placement of a covered metallic stent in 64(80.0%) non-oncologic patients and 34(73.9%) oncologic patients (p=0.506) and clips in 99(48.8%) oncologic patients and 19(41.3%) (p=0.461) non-oncologic patients. The frequency of early adverse events (<1 week) (8(17%) vs 14(18%), p=1.0),as well as late adverse events (10(22%)vs 21(26%), p=0.670) were similar in both groups
We documented clinical success (fistula closure) in 72(90%) of non-oncologic patients and 34(74%, p=0.023) of oncologic patients, but after a similar median duration of management (13 (1-423) weeks vs 6(1-55) weeks, p=0.11) and number of endoscopies (3(1-8) vs 2(1-13), p=0.706).
After a median follow up of 36(1-147) and 12.5(1-113) months, 6(7.5%) and 28(60.9%) patients died, respectively from the group of non-oncologic and oncologic surgery. In the oncologic surgery group, 12 patients (48%) died from progression of oncologic disease.

Conclusion

A primarily endoscopic approach is effective in >70% of patients with fistulas after UDT surgery but it´s efficacy was significantly lower in patients with fistulas after oncologic surgery compared to non-oncologic surgery and this may have been due to the higher age and male predominance.

Introduction

Managing leaks and defects of the upper gastrointestinal (UGI) tract remains challenging. Endoscopic Vacuum Therapy (EVT) is a promising option^1-3; However, data on its efficacy in large cohorts remain limited.

Aims & Methods

This retrospective study with prospectively collected data enrolled all consecutive patients with UGI defects treated by EVT at a single academic center (San Raffaele Hospital, Milan, Italy) from March 2015 to September 2024. The primary outcome was clinical success (CS). Secondary outcomes included median treatment duration/devices, technical success (TS) and adverse events (AEs). Clinical success (CS) was defined as complete endoscopic and/or radiologic defect healing, with the patient remaining asymptomatic. Adverse events (AEs) were classified according to the AGREE classification.

Results

Over a period of 114 months, 96 patients [72.9% male, median age 67 years (IQR 57-75), 49% ASA 3] were included. Among them, 66 patients (68.7%) had anastomotic leaks after Ivor-Lewis esophagectomy, with 57% undergoing neoadjuvant chemotherapy. The median leak size was 7 mm (IQR 4-10), with a median circumferential involvement of 20% (IQR 10-40). A fistula involving the airways was present in 28.1% of cases. EVT was the first-line treatment in 78.1% of patients, with a median time from surgery to EVT of 7 days (IQR 4-9).
The overall clinical success (CS) rate was 72.9% (70/96 patients). Of the remaining patients, 20 underwent redo-surgery. Technical success for all procedures was 100% (302/302 procedures). The median treatment duration was 11 days (IQR 5-18), with a median of 3 EVT sessions (IQR 2-4). Three adverse events (AEs) (3/302, 1%) were recorded, including one bleeding episode and two sponge dislodgments, all managed endoscopically (AGREE IIIa). During a median follow-up of 62 days (IQR 33-155), two cases (2.1%) of esophageal strictures were reported (Table 1)
In the subgroup analysis, the intraluminal (IL) group (n=70) had a lower mean number of EVT sessions (3±2 vs. 5±3, p<0.001) and a shorter mean treatment duration (11±10 vs. 21±16 days, p<0.001) compared to the intracavitary (IC) group (n=26).

Table 1: Outcomes

Patients n=96	Median (IQR) OR n (%)
Clinical success, n	70 (72.9%)
Technical success, n	302/302 (100.0%)
Adverse events, n	3/302 (1%)
Treatment duration (days)	11 (5-18)
Number of EVT sessions	3 (2-4)
Hospital stay (days)	28 (17-43)
ICU admission, n	14 (14.9%)
Stricture, n	2 (2.1%)
Follow-up (days)	62 (33-155)

Conclusion

EVT is a safe and effective treatment for UGI defects. The IL approach was associated with shorter treatment duration and fewer sessions compared to the IC approach. Further controlled studies are needed to optimize EVT application and outcomes.

References

1. Mandarino FV, Barchi A, Leone L, et al. Endoscopic vacuum therapy versus self-expandable metal stent for treatment of anastomotic leaks < 30 mm following oncologic Ivor-Lewis esophagectomy: a matched case-control study. Surg Endosc. 2023;37(9):7039-7050. doi:10.1007/s00464-023-10213-8
2. Mandarino FV, Barchi A, D'Amico F, et al. Endoscopic Vacuum Therapy (EVT) versus Self-Expandable Metal Stent (SEMS) for Anastomotic Leaks after Upper Gastrointestinal Surgery: Systematic Review and Meta-Analysis. Life (Basel). 2023;13(2):287. Published 2023 Jan 19. doi:10.3390/life13020287
3. Mandarino FV, Barchi A, Fanti L, et al. Endoscopic vacuum therapy for post-esophagectomy anastomotic dehiscence as rescue treatment: a single center case series. Esophagus. 2022;19(3):417-425. doi:10.1007/s10388-022-00912-x

Introduction

The Luso-Cor^​ esophageal stent was developed for management of fistulas and dehiscences after bariatric and gastric oncologic surgery.

Aims & Methods

The study aims to evaluate the efficacy and safety of Luso-Corin managing fistulas and strictures after gastric surgery. This is a retrospective and multicentric study involving 35 consecutive patients with adverse events after gastric surgery who were managed with the Luso-Cor^Ò esophageal stent, between July 2016 and February 2023. The adverse events were: fistulas after sleeve gastrectomy(n=28), anastomotic dehiscence(n=4) and gastric tube stricture after bariatric surgery(n=3). The technical success was defined as the effective exclusion of the fistula and the clinical success as fistula closure or effective stricture dilation after stent removal. Endoscopic reinterventions and adverse events were analyzed with descriptive statistics.

Results

Median age was 50(20-84) years, 69%(n=24) were women. Median fistula orifice size was 5.5(2–20)mm. The median duration of stent implantation was 6(1-18) weeks. An endoscopic reintervention was required in 31.43%(n=11) patients due to the following reasons: Inadequate coaptation of the proximal flare to esophageal wall(n=6), migration(n=3), stricture(n=1), stent fracture(n=1). Adverse events were observed in 14.23% (n=5) patients: GI bleeding(n=3), esophageal stricture requiring balloon dilation(n=1), recurrent vomiting(n=1).
The technical success in exclusion of fistulas / anastomotic dehiscences was 100%(32). There were 5 deaths: sepsis(n=3); aortoesophageal fistula 3 cm above the proximal edge of the stent (n=1) e probable pulmonary thromboembolism (n=1).
After a median stent implantation of 6 (1-18) weeks, a clinical success rate in fistula closure of 93%(26/28) was observed. In two patients with persistent fistulas after stent removal were successfully managed with 10 mm over-the-scope-clips. The clinical success of the stent in managing strictures with concomitant fistulas was 100%(13/13) and isolated stricture was 33%(1/3). One patient required endoscopic dilatation and the other underwent total gastrectomy.

Conclusion

The innovative design of the Luso-Cor. stent is very effective in managing fistulas and concomitant strictures after gastric bariatric and oncologic surgery. The esophageal anchoring reduces the migration risk and facilitates the stent removal.

Introduction

The Luso-Coresophageal stent was developed to manage fistulas and dehiscences after bariatric and gastric oncologic surgery.

Aims & Methods

The study aims to evaluate the effectiveness and safety of Luso-Cor in treating sleeve gastrectomy fistulas. Retrospective and multi-centric analysis of 28 consecutive patients with fistulas after sleeve gastrectomy who were managed with the Luso-Cor stent, between July 2016 and February 2023. Qualitative variables were analyzed as percentages. Quantitative variables were analyzed as median (min-max) and Mann-Whitney test (non-normal distribution). Predictive factors of fistula closure and adverse events were analyzed with logistic regression. A p-value <0.05 was considered statistically significative. (IBM SPSS 28).

Results

Median age was 45.5 (20-66) years, 78% (n=22) were women. Median fistula orifice size was 5(2 – 20)mm and the location of leaks was: Cardia–71.4%(n= 20); gastric tube–17.9%(n=5); other–10.7%(n=3). The median duration of stent implantation was 6(1–18) weeks. In 28.57%(n=8) patients, a re-intervention was necessary due to: inadequate coaptation of proximal flare of the stent(n=6), migration(n=1) and esophageal stricture(n=1). In 3 cases a new stent was required. Adverse events were observed in 17.9%(n=5) patients: GI bleeding (n=3), esophageal stricture (n=1) stricture requiring balloon dilation, and recurrent vomiting (n=1). Two patients died, one due to aorto-esophageal fistula and the other one due to pulmonary embolism.
The overall success rate of Luso-Corstent was 89.3% (25/28). One patient died before stent removal. In 2 of the remaining cases the persisting fistula orifice was closed with an over-the-scope clip (OTSC). The stent was used as first line in 50% (n=14) patients and as rescue therapy in the remaining 50% (n=14)second line, with a significant reduction in the number of endoscopies (3 vs 4) (p=0.031), without difference in efficacy (p=1.225) and adverse events (p=1.195) between groups.

Conclusion

The Luso-Cor^​ esophageal stent was effective in managing sleeve gastrectomy fistulas, with low stent migration rates. The use of the stent as first line therapy significantly o reduced the number of endoscopies until fistula closure.

Clinical Case Summary

Introduction: Fistulas after laparoscopic sleeve gastrectomy (LSG) occur in 1-4% of patients, with 90% located at the cardia.
Case report: A 66-year-old female with grade III obesity, obstructive sleep apnea, type II diabetes and hypertension underwent LSG. She developed hypovolemic shock, abdominal pain and severe anemia (Hb 6.5 mg/dL). A CT scan revealed haemoperitoneum without active bleeding, managed with transfusion of packed blood cells. A week later, a new CT scan performed for leucocytosis and abdominal pain revealed pneumoperitoneum. An esophagogastroduodenoscopy (EGD) revealed a 2 cm fistula orifice at the cardia. A novel esophageal covered metallic stent was placed for five weeks. The fistula orifice decreased to 4mm and communicated through a fistulous tract with a residual subphrenic abscess measuring 62x20mm. Pus was collected from the abscess and drained internally with a double pigtail plastic 10Fr stent through the fistula orifice. After initial improvement, there was clinical deterioration requiring percutaneous subphrenic abscess drainage. Five weeks after the first EGD, the stent was removed, the fistula orifice ablated with argon plasma 40W/1L and closed with an over the scope clip of 7 mm. Patient improved and was discharged 4 months after LVG.
Discussion:The Luso-Cor esophageal stent is a specifically-designed covered stent with a 5mm uncovered ring near the proximal edge reducing risk of migration. Two articulating zones in the middle portion allow better adaptation to altered anatomy after LVG and a distal flare prevents retrograde reflux of fluid. This stent overcomes strictures in the gastric tube, concomitantly present in nearly 50% of patients with fistulas after LSG and has a low radial force to reduce tissue trauma. The novel Luso-Cor esophageal stent was able to serve as bridge to clinical stability with a significant decrease in the size of the fistula orifice which was closed with complementary therapeutic endoscopic procedures.

References

1. Rodrigues-Pinto E, Pereira P, Sousa-Pinto B, et al. Retrospective multicenter study on endoscopic treatment of upper GI postsurgical leaks. Gastrointest Endosc. 2021;93(6):1283-1299.e2. doi:10.1016/j.gie.2020.10.015
2. Bège T, Emungania O, Vitton V, et al. An endoscopic strategy for management of anastomotic complications from bariatric surgery: A prospective study. Gastrointest Endosc. 2011;73(2):238-244. doi:10.1016/j.gie.2010.10.010
3. Bemelman WA, Baron TH. Endoscopic Management of Transmural Defects, Including Leaks, Perforations, and Fistulae. Gastroenterology. 2018;154(7):1938-1946.e1. doi:10.1053/j.gastro.2018.01.067
4. Noronha Ferreira C, Reis D, Santos P, et al. Clinical Efficacy and Safety of a Novel Stent for Fistula Management Following Sleeve Gastrectomy. Tech Innov Gastrointest Endosc. 2024;26(1):12-20. doi:10.1016/j.tige.2023.07.005