Endoscopic Therapy for Anastomotic Bleeding After Lower Gastrointestinal Surgery in Patients With Crohn’s Disease

Abstract

Background and Aims

Crohn's disease (CD) patients may experience anastomotic bleeding after lower gastrointestinal surgery, but its endoscopic management remains underexplored. We aimed to analyze the risk factors for anastomotic bleeding and the efficacy of endoscopic treatment.

Methods

We retrospectively analyzed clinical characteristics, endoscopic treatments, and outcomes of CD patients with postoperative anastomotic bleeding from January 2021 to May 2025 at multiple centers. Anastomotic bleeding was defined as (i) direct bleeding (melena/hematochezia) or (ii) hemoglobin decrease ≥2 g/dL. Patients were followed up for 1 year, with rebleeding requiring endoscopic reintervention.

Results

We included a total of 21 patients with 100% of patients having onsite bleeding controlled with initial endoscopic treatment and 9 patients (42.9%) having rebleeding. There was no statistical difference between the rebleeding group and nonrebleeding group in terms of age, gender, comorbidity, anticoagulation or antiplatelet therapy, clotting disorder, hemoglobin before or after procedure, bleeding type, the Rutgeert’s score, and endoscopic treatment. However, a greater number of patients in the rebleeding group required blood transfusion (88.9% vs 33.3%, P value = .011), and a greater volume of blood transfusion 2.0 (2.0–3.5 units) vs 0.0 (0.0–1.8 units) than those in nonrebleeding group (P value = .013). None of the 21 patients had no endoscopy-associated complications.

Conclusion

Endoscopic therapy appears to be safe and effective for the majority of CD patients with anastomotic bleeding following bowel surgery. The requirement for blood transfusion at the inception, in conjunction with a high volume of blood transfusion appear to be high risk factors for rebleeding.

Introduction

Crohn's disease (CD) is an idiopathic, incurable, chronic inflammatory disease of the gastrointestinal (GI) tract with transmural inflammation at the any part of the GI tract, leading to the development of strictures or obstructions, fistula, abscesses. The last 2.5 decades have witnessed a widespread use of biological agents in inflammatory bowel disease. Still, some patients with CD, eventually require surgical intervention for medically refractory disease, disease-associated complications, or colitis-associated neoplasia. At the time of diagnosis, approximately 20% of patients with CD have stenoses and fistulas, and this number may increase to 50%–70% within the first decade after diagnosis.^1,2 The surgical procedures most commonly employed in the treatment of CD include intestinal resection with anastomosis, strictureplasty, diverting ostomy, and bowel bypass.³ Surgical intervention for CD is not curative, and postoperative disease recurrence is common. Regular monitoring and treatment to prevent relapse is necessary. In addition, postoperative complications, such as anastomotic bleeding, anastomotic fistula, and anastomotic stenosis can occur.

Anastomotic bleeding is uncommon. The reported frequency of anastomotic bleeding after CD surgery is estimated about 3%, depending on surgical procedures.4, 5, 6, 7 While most anastomotic bleeding is self-limiting, some requires endoscopic, radiographic, or surgical intervention. Endoscopy is a main treatment modality for GI bleeding. Its role in the management of postoperative anastomotic bleeding in CD is not defined with scant literature. This study describes the experience of multicenter using endoscopic treatment for lower GI anastomotic bleeding in patients with CD.

Methods

Patients

This was a retrospective multicenter study involving 3 inflammatory bowel disease centers: the Center of Interventional Inflammatory Bowel Diseases at Columbia University Irving Medical Center/New York Presbyterian Hospital, New York, NY; Department of Medical Gastroenterology, Asian Institute of Gastroenterology, Hyderabad, Telangana, India; and Inflammatory Bowel Disease Unit, Department of Gastroenterology and Hepatology, Hospital Universitario Puerta de Hierro-Majadahonda, Madrid, Spain. Patients’ data were collected from January 2021 to May 2025. Each center identified eligible patients through electronic medical record review based on unified inclusion and exclusion criteria. The Columbia University Irving Medical Center's protocol number approved by the Institutional Review Board was AAAN3966. The Asian Institute of Gastroenterology of India's institutional review board number was ECR/346/Inst/AP/2013. The case from Hospital Universitario Puerta de Hierro-Majadahonda had an exempt from institutional review board as the patients’ information was deindentified. Given the retrospective nature of the study, informed consent was waived.

Definition

“Postoperative anastomotic bleeding” was defined as the presence of the following signs which could not be attributed to causes other than GI bleeding after bowel resection and anastomosis surgery with intestinal or colonic anastomosis: 1) direct bleeding such as melena or hematochezia or 2) indirect bleeding with hemoglobin decrease ≥ 2 g/dL; Furthermore, all instances of anastomotic bleeding were confirmed by endoscopy to be located at the site of the anastomosis, and were treated with endoscopic hemostasis. All enrolled patients were followed up for 1 year, rebleeding was defined as the occurrence of any of the above bleeding signs requiring endoscopic reintervention for bleeding control.

Clinical Variables

Data collection included: age, gender, smoking, associated comorbidities, previous anticoagulant or antiplatelet treatment, clotting disorder, indication for blood transfusion, volume of blood transfusion, hemodynamic instability, hemoglobin before and after procedure, bleeding type, endoscopic techniques, the Rutgeert’s score for postoperative recurrence of CD.⁸

Endoscopy Variables

The types of anastomotic bleeding can be classified as either active or nonactive. Active bleeding is characterized by arterial pulsatile bleeding, small venous bleeding, or oozing of blood. In contrast, nonactive bleeding refers to the presence of traces of bleeding, which can be considered the responsible lesion for bleeding. The hemostatic techniques utilized for endoscopic procedures comprise the use of hemostatic clips, thermal coagulation, local injection (1:10,000 epinephrine solution), topical spraying with 50% dextrose and hemostatic gel, and staple removal. We categorized endoscopic treatments as single or combinations treatments.

Outcome Measurement

We observed the rate of onsite control of bleeding and rebleeding with endoscopic treatment in all cases, and we also compared the clinical characteristics and endoscopic treatment in the group of rebleeding and nonrebleeding.

Statistical Analysis

Normally distributed continuous data were expressed as mean ± standard deviation and were compared between 2 groups using Student's t-test. Non-normally distributed continuous data were expressed as median (Q1, Q3) and were compared between 2 groups using the Mann–Whitney U test. Categorical data were expressed as number (percentage) (n [%]) and were compared between groups using the χ² test. A P value < .05 was considered statistically significant. Survival curves for the no rebleeding rate were constructed using the Kaplan–Meier method. SPSS version 27.0 was used to perform all analyses.

Results

A total of 21 CD patients with anastomotic bleeding after ileocolonic or colorectal anastomoses were treated endoscopically included in this study. Onsite control of bleeding was successfully achieved by endoscopic treatment in all 21 cases, with a 100% onsite hemostasis rate, but rebleeding occurred in 9 patients (42.9%), after rescue endoscopic treatment, 2 patients rebleeded again, but did not require surgery or embolization therapy. No patient experienced endoscopy-related complications, such as perforation or anastomotic separation, either during initial or repeat endoscopic treatment.

Of the 21 patients who underwent endoscopic hemostasis, 57.1% were male and 42.9% were female. The median age was 51.0 years (interquartile range [IQR]: 35.0–62.5 years). Comorbidities outside the GI system were present in 6 patients (28.6%). Two patients (9.5%) received anticoagulant therapy before surgery, while 4 patients (19.0%) were administered antiplatelet agents. Only 1 patient (4.8%) had clotting disorder. Twelve cases (57.1%) received a blood transfusion. The median blood transfusion was 2 units (IQR: 0.0–2.0 units). The median hemoglobin prior to endoscopic was 8.5 g/dl (IQR: 7.1–10.9 g/dl), and the median hemoglobin after surgery was 10.2 g/dl (IQR: 8.3–11.0 g/dl). Endoscopic examination revealed active anastomotic bleeding in 9 cases (42.9%) (Figure 1), while in the remaining 12 cases (57.1%), stigmata of blood were observed, which were classified as inactive. The median of Rutgeert’s score before endoscopic treatment was 2.0 (IQR: 0.0–1.0). Fourteen patients (66.7%) used single endoscopic treatment technique, with hemostatic clips being the most commonly applied (Figure 2, Video A1), the other 7 cases used a combined endoscopic technique.

Figure 1.

Active bleeding at the ileocolonic anastomosis.

Figure 2.

Hemoclipping treatment.

Based on the results of initial endoscopic treatment, all the patients were divided into 2 groups (rebleeding and nonrebleeding). No significant difference was found in age, gender, comorbidities, anticoagulation or antiplatelet therapy, clotting disorders, or hemoglobin before or after procedure between the 2 groups. The endoscopic findings, including bleeding type, endoscopic treatment modality type, and the Rutgeerts score did not differ between the 2 groups. However, the number of patients in the rebleeding group who had a transfusion indication and the volume of transfusions were higher than those in the nonrebleeding group, with the percentage of transfusion (N = 8; 88.9% vs N = 4; 33.3%, P value = .011), and the median volume of blood transfusion 2.0 (IQR: 2.0–3.5 units) vs 0.0 (IQR: 0.0–1.8 units), P value = .032 (Table).

Table.

Comparison of Clinical Data and Endoscopy-Related Characteristics in the Rebleeding Group and the Nonrebleeding Group

Variables	Total patients (n = 21)	Non rebleeding patients (n = 12)	Rebleeding patients (n = 9)	P
Age (y)	51.0 (35.0–62.5)	48.0 (41.8–59.5)	62.0 (26.0–66.5)	.831
Gender
Male	12 (57.1%)	7 (58.3%)	5 (55.6%)	.899
Female	9 (42.9%)	5 (41.7%)	4 (44.4%)
Smoking
Yes	5 (23.8%)	2 (16.7%)	3 (33.3%)	.375
No	16 (76.2%)	10 (83.3%)	6 (66.7%)
Comorbidity
Yes	6 (28.6%)	2 (16.7%)	4 (44.4%)	.163
No	15 (71.4%)	10 (83.3%)	5 (55.6%)
Anticoagulation therapy
Yes	2 (9.5%)	1 (8.3%)	1 (11.1%)	.830
No	19 (90.5%)	11 (91.7%)	8 (88.9%)
Antiplatelet treatment
Yes	4 (19.0%)	2 (16.7%)	2 (22.2%)	.748
No	17 (81.0%)	10 (83.3%)	7 (77.8%)
Clotting disorder
Yes	1 (4.8%)	1 (8.3%)	0 (0.0%)	.375
No	20 (92.5%)	11 (91.7%)	9 (100.0%)
Indications for blood transfusion
Yes	12 (57.1%)	4 (33.3%)	8 (88.9%)	.011
No	9 (42.9%)	8 (66.7%)	1 (11.1%)
Hemodynamic instability
Yes	16 (76.2%)	9 (75.0%)	7 (77.8%)	.882
No	5 (23.8%)	3 (25.0%)	2 (22.2%)
Bleeding type
Active	9 (42.9%)	4 (36.4%)	5 (55.6%)	.391
Nonactive	12 (57.1%)	7 (63.6%)	4 (44.4%)
Rutgeert’s score before endoscopic treatment (i0–i4)	2.0 (0.0–2.0)	2.0 (1.0–2.0)	2.0 (0.0–2.0)	.318
Endoscopic therapy
Single	14 (66.7%)	9 (75.0%)	5 (55.6%)	.350
Combined	7 (33.3%)	3 (25.0%)	4 (44.4%)
Time from anastomotic bleeding to endoscopy (d)	30.0 (7.0–101.3)	30.0 (7.0–105.0)	30.0 (4.0–105.0)	.789
Hemoglobin before procedure (g/dl)	8.5 (7.1–10.9)	8.9 (7.4–14.0)	8.4 (7.0–10.4)	.342
Hemoglobin after procedure (g/dl)	10.2 (8.3–11.0)	10.2 (8.9–11.7)	10.0 (7.8–10.5)	.311
Blood transfusion (unit)	2.0 (0.0–2.0)	0.0 (0.0–1.8)	2.0 (2.0–3.5)	.013

Bold values indicate statistical significance (P < .05).

At 1 year of follow-up, 9 patients (42.9%) had recurrent bleeding end received rescue endoscopy. The risk of no rebleeding rate at 12-month follow-up was (9 of 21) 57.1% (Figure 3).

Figure 3.

No rebleeding rate after 1 year of follow-up.

Discussion

The issue of postoperative lower GI anastomotic bleeding in CD patients is not frequently addressed in the medical literature. As previously indicated, the condition is rare and can be managed conservatively. However, endoscopic intervention is a crucial aspect of treatment in a subset of severe cases. In our retrospective observational study. All patients (100%) achieved onsite control of bleeding with initial endoscopic treatment. Although 9 patients (42.9%) rebleeded, they eventually succeeded in bleeding control without the need for more invasive treatments such as surgery or embolization following a rescue endoscopic treatment. The absence of endoscopic treatment-related complications regardless of initial or rescue endoscopic treatment. Of the endoscopic anastomotic bleeding types, 42.9% were active bleeding. In terms of the type of endoscopic treatment, the use of a single endoscopic technique accounts for 66.7% of the cases. There was no significant difference between rebleeding group and nonrebleeding group in terms of whether or not they used anticoagulant and antiplatelets before endoscopic treatment. Similarly, the Rutgeerts scores were not statistically significant between the 2 groups. However, a significantly higher volume of blood was transfused to the rebleeding group than to the nonrebleeding group, and the proportion of patients with an indication for transfusion was also higher in the rebleeding group.

We reviewed the literature, and there are very few studies or case reports on endoscopic therapy of anastomotic bleeding after lower GI surgery in CD patients. However, studies have been conducted in instances where the etiology is changed to colorectal cancer (CRC) or diverticular disease. The risk factors for anastomotic bleeding include surgical modality, primary disease, and the overall status of the patient (age, gender, smoking, and current medications). It had been revealed that the age greater than 80 years was an independent risk factor for anastomotic bleeding following right-sided colectomy.⁹ Furthermore, findings from the Wang’s study indicated that anastomotic bleeding is more likely to occur in male patients than in female patients following rectal cancer.¹⁰ The type of primary disease and stapled anastomosis may increase postoperative anastomotic bleeding.11, 12, 13 Nevertheless, it had also been documented that the surgical techniques do not affect postoperative anastomotic bleeding.¹⁴ Concerns have been raised regarding the impact of antithrombotic medications on anastomotic bleeding. The Noguchi’s study revealed no heightened risk of bleeding associated with either preoperative or postoperative antithrombotic drug utilization.^9,11 The conservative management of anastomotic bleeding typically involves the administration of a blood transfusion. The presence of blood transfusion indicates the potential for more aggressive bleeding, necessitating intervention. Hébert’s study showed that the lower GI anastomotic bleeding group was more likely to have transfusion and hemoglobin drop complications.¹³ In the event of postoperative anastomotic bleeding, endoscopy represents an effective means of identifying the site of bleeding, establishing the cause of bleeding, and performing endoscopic treatment. Furthermore, endoscopy can be employed as a secondary intervention to reduce the likelihood of surgical reoperation. Liu and colleagues¹⁵ observed an 81% (63 of 78) efficacy rate in the initial endoscopic hemostasis in the endoscopic treatment of postoperative anastomotic bleeding. among 78 cases, 67% of patients were treated with a single endoscopic therapy and 15% with a combination of treatments. Of the 37 anastomotic bleeding patients treated endoscopically by Besson's team, 21 (56.7%) were treated with a single endoscopic technique; 7 (18.9%) were given combined endoscopic therapy to achieve hemostasis.¹¹ However, it is important to note that the endoscopist cannot ignore the possibility of rebleeding and potential complications in addition to considering success rates. Current literature resulted on rebleeding rates show considerable variability in outcomes, ranging from 14.3% to 100%.15, 16, 17, 18 It is considered that this result is dependent on the size of study groups and the duration of follow-up. The Serrano's study showed that 1 of 7 (14.3%) anastomotic rebleeding was referred to surgical treatment.¹⁶ The most common endoscopic complication is leakage.

The discrepancies between our findings and those reported in the literature are not substantial; however, some differences persist. Our results revealed that the volume and indication of blood transfusions were higher in the rebleeding group than in the nonrebleeding group. In addition we also found that antithrombotic medications did not increase the risk of rebleeding, which was consistent with the findings in the research literature mentioned above. However, we did not find differences in age, gender, type of anastomotic bleeding, the Rutgeerts score, and endoscopic treatment modality between the 2 groups, which may be related to the small sample size that prevented a more detailed subgroup analysis. Furthermore, it also implied that the disease activity inherent in CD does not play a pivotal role in the occurrence of anastomotic bleeding. The rebleeding rate in our research was likely to be higher than that documented in the extant literature, which is thought to be due to the longer follow-up period employed in our study, as opposed to the 1-month follow-up reported in the literature.

There are limitations of the study. Due to its small sample, retrospective, observational design, the results of the study should be interpreted with caution. A large sample, prospective, randomized study would provide a more comprehensive and stratified analysis.

Our findings suggest that the volume and requirement of blood transfusion may be risk factors for rebleeding. Endoscopic treatment is an effective diagnostic and therapeutic tool for lower GI anastomotic bleeding, whether a single therapy or a combination technique, and appropriate endoscopic treatment of anastomotic bleeding can achieve bleeding control and avoid reoperation. The endoscopic procedures were found to be safe, with no complications observed.

Conclusion

In conclusion, lower GI anastomosis bleeding in CD patients usually requires interventional management in addition to conservative treatment. Endoscopic therapy is a safe and effective approach for the management of these patients. Appropriate techniques selection and cautious handling can reduce complications and decrease reoperation rates.