Authors: Suleyman Dolu, Mehmet Emin Arayici, Soner Onem, Ilker Buyuktorun, Huseyin Dongelli, Goksel Bengi, Mesut Akarsu
Categories: Research, Double-balloon enteroscopy, Small bowel vascular lesion, Angioectasia
Source: BMC Gastroenterology
Authors: Suleyman Dolu, Mehmet Emin Arayici, Soner Onem, Ilker Buyuktorun, Huseyin Dongelli, Goksel Bengi, Mesut Akarsu
Small-bowel angioectasia is commonly diagnosed and managed using double-balloon enteroscopy; however, rebleeding rates can vary significantly. This study aimed to identify and evaluate the clinical predictors of rebleeding in patients with small-bowel angioectasia.
This retrospective study focused on adult patients who underwent endoscopic management for small bowel vascular lesions (SBVLs). A total of 67 patients were included in the study, all of whom were retrospectively analyzed. The SBVLs were classified using the Yano et al. classification system. Among these, 62 patients with angioectasia who received endoscopic treatment were specifically analyzed. To further investigate the clinical outcomes, the angioectasia group that required endoscopic treatment was divided into two subgroups based on the number of double-balloon enteroscopy (DBE) procedures performed. Univariate and multivariate binary logistic regression analyses were used to establish which predictor variables were significantly related to the recurrence.
A total of 67 patients (mean age 68.1 ± 8.9 years; 44 males) with SBVL, angioectasia was diagnosed in 62 (92.5%) patients, dieulafoy lesion in 1 (1.5%) patient, and arteriovenous malformation in 4 (6%) patients. Similarly, the prevalence of chronic renal failure and diabetes mellitus were significantly higher in the multiple DBE endotherapy group compared to the single DBE endotherapy group (p < 0.001; p = 0.032, respectively). In multivariate logistic regression analysis, anticoagulant use (OR = 9.648, 95% CI: 1.729–53.817, p = 0.010), chronic renal failure (OR = 15.683, 95% CI: 2.727–90.203, p = 0.002), localization of duodenum (OR = 13.509, 95% CI: 1.598–114.168, p = 0.017), and localization of the ileum (OR = 17.100, 95% CI: 1.477–197.905, p = 0.023) were all independently associated with a higher risk of angioectasia recurrence.
The findings of this study demonstrates a high success rate of endoscopic treatment, with a rebleeding rate of 27%, highlighting the significance of this approach. Chronic renal failure, use of anticoagulant, and lesion localization in the duodenum or ileum were identified as independent risk factors for rebleeding, underscoring the need for careful patient monitoring and targeted intervention in these cases.
Small bowel bleeding accounts for 5% of all gastrointestinal bleeding cases [1]. Small bowel vascular lesions (SBVLs) are the most prevalent cause of small bowel bleeding [2], with angioectasias being the most common observation. They can be multiple, have varying bleeding potentials, and are often seen in the proximal small bowel [3]. Double-balloon enteroscopy (DBE) enabled not only an endoscopic diagnosis but also treatment of SBVLs. However, endoscopically detected SBVLs vary in appearance, and the cause and best therapy for these lesions are unknown [4]. Endoscopic clipping, injection therapy and argon plasma coagulation (APC) are endoscopic procedures that have been used for the vascular lesions treatment [5, 6].
The nomenclature used in the literature to describe gastrointestinal vascular lesions is notably confusing, with various terms often being used interchangeably. This inconsistent terminology has led to significant challenges in accurately identifying and understanding these lesions. The categorization of gastrointestinal vascular lesions lacks standardization, further contributing to the ambiguity and uncertainty surrounding their classification and interpretation [7]. For example, terms such as angioectasia and arteriovenous malformation (AVM) have been applied to vascular abnormalities in the gastrointestinal tract without consistent or precise definitions, leading to potential miscommunication in both clinical and research settings. Recognizing the need for clearer classification, Yano et al. have made efforts to address these concerns by endoscopically classifying small bowel vascular lesions (SBVLs) into four distinct subgroups [4]. This classification aims to bring more clarity and uniformity to the understanding and diagnosis of these lesions, thereby improving both clinical outcomes and research consistency.
In this study, we investigated patient characteristics and endoscopic treatments for SBVLs diagnosed by DBE. Therefore, the purpose of this study was to evaluate the predictors and management of SBVLs diagnosed by DBE.
Seven hundred and forty patients who underwent DBE between 2006 and August 2021 at Dokuz Eylül University Hospital were screened retrospectively. DBE was performed in a cohort of 314 patients with gastrointestinal bleeding etiology, and 67 of these patients were diagnosed with SBVL through DBE (Fig. 1). Patient characteristics, including age, sex, bleeding type (overt or occult), drug usage (antiplatelet and anticoagulant) and underlying disease (cerebrovascular disease, cardiovascular disease, chronic renal failure, hypertension, diabetes mellitus, and portal hypertension) were recorded. DBE procedural details (insertion depth, insertion route, procedure time) were noted. Lesion characteristics for SBVL, including type, number of lesions (single or multiple), location in the small bowel, and rebleeding rate were evaluated. The localization of SBVLs was determined by measuring the small bowel depth of insertion, the size of the enteric cavity, and the shape of the mucosal folds and villi [8, 9].
Fig. 1Flowchart of the study. DBE, double balloon enteroscopy; SBVL, small bowel vascular lesion
Yano et al. created a simple categorization system for vascular lesions in the small bowel. In our study, we used this classification to type SBVLs: Type 1 is a punctuate (< 1 mm) or patchy (a few mm) erythema and considered to be an angioectasia. Type 2 is an artery that typically protrude through a small mucosal defect and considered to a dieulafoy lesion. Type 3 is pulsatile red protrusion with surrounding venous dilatation and considered to be an arteriovenous malformation. Type 4 lesions are not classified into any of the above categories [4]. According to the Yano endoscopic classification of SBVL, type 1, type 2, and type 3 lesions was included and type 4 lesions were not included in our study.
Small bowel bleeding was defined as bleeding between the papilla and the ileocecal valve that persists or recurs after a negative upper and lower endoscopy. Small bowel bleeding was classified as overt bleeding (melena or hematochezia) and occult bleeding (iron-deficiency anemia, with or without a positive fecal occult blood test) [10]. Rebleeding was defined as the conditions in which upper and lower gastrointestinal endoscopies were negative in patients presenting with bleeding (melena or hematochezia) or a subsequent decrease in the hemoglobin level by > 2 g/dL from baseline. DBE was performed again in patients who were followed longtime follow-up (> 6 months) and were considered for rebleeding. It was thought that DBE treatment was successful in patients who were not considered to have rebleeding during their outpatient clinic visits. Patient follow-up was performed through clinic visits. In our center, routine follow-up visits are normally scheduled every 3 months, along with clinical examination, laboratory test data, and data related to rebleeding. Three of the 72 patients with angioectasia died (unknown cause) in the first six month, and seven were lost to follow-up. Patients who did not die within the first six month after initial hemostasis were included in the study. The interval between the first DBE hemostasis and the last clinic visit or death time was the follow-up period. A retrospective analysis was performed on the data of the remaining 62 patients. If patients experienced rebleeding, our routine procedure consisted of repeated esophagogastroduodenoscopy and colonoscopy before undergoing another DBE to locate the bleeding site. Patients were considered to be lost to follow-up if they did not undergo examination at regular intervals. We divided the angioectasia group requiring endoscopic treatment into two groups according to DBE endotherapy numbers (1 and > 1).
In this study, clinical predictors of rebleeding following initial endotherapy were evaluated, and patients who did not undergo any treatment were excluded. Therefore, the inclusion criteria were patients presenting with gastrointestinal bleeding and SBVLs (types 1, 2, or 3) who underwent DBE combined with endotherapy. The exclusion criteria were being under 18 years of age or pregnancy.
All procedures were performed by the same endoscopist with more than 15 years of experience (more than 1000 procedures). All DBE procedures were performed with a Fujinon EN-450T5 enteroscope (Fujinon Corp, Saitama, Japan). If the enteroscopy was to be performed via the oral route, at least 6 h of fasting was required; if the procedure was to be performed via the anal route, normal colonoscopy preparation solutions were used. All patients’ written informed consents were taken before the DBE examination. Before the examination, each patient had a hemogram, coagulation test, liver, and kidney function tests, and an anesthesiologist assessed each patient’s eligibility for the procedure. All patients were accompanied by an anesthesiologist during the procedure, and deep or conscious sedation with simultaneous monitoring was employed. As DBE has been performed by the same team in our clinic for a long time, similar procedures were applied to all patients (preprocedural fasting time, laboratory, etc.). The description of the DBE procedure reflects the standardized approach routinely employed at our center during the study period, as documented in the medical records.
APC, injection therapy and clip application were used alone or in combination for hemostatic endoscopic treatment. Type 1 and type 3 lesions were generally treated with APC and/or clip placement, whereas clip placement was used for treatment of type 2. The endoscopist’s opinion at the time of DBE was used to determine whether or not endotherapy should be done. Lesions with a high risk of rebleeding or requiring surgical intervention during enotherapy were tattooed on the lesion margin.
The normality assumptions of the data were controlled by the Kolmogorov Smirnov and Shapiro-Wilk test. Continuous data were presented with mean and standard deviation (SD) or median (interquartile range (IQR)). Categorical variables were presented with frequency (n) and percentage (%) and analyzed with Pearson chi-square and Fisher’s Exact test. Independent sample t-test was used for comparison of age according to recurrence. Univariate and multivariate binary logistic regression analyses were used to establish which predictor variables were significantly related to the recurrence. Variables that could be associated with recurrence were included in the univariate analyses. Multivariate analysis included variables with p < 0.10 in univariate analysis, and calculated by using the backward stepwise conditional method. Odds ratio (OR) with their corresponding 95% confidence intervals (CIs) was documented. Statistical analyses were utilized using the IBM SPSS Statistics for Mac, Version 29.0 (IBM Corp., Armonk, NY) and STATA (v18, College Station, TX, USA) package programs. Two-tailed p-value less than 0.05 was considered statistically significant.
During the study period, 740 patients underwent a total of 1028 DBE procedures for various indications. Among patients presenting with gastrointestinal bleeding, we could identify the potential source of OGIB in the small intestine in 254 of 314 patients. The most common sources of bleeding were ulcers/erosions in 90 patients, tumor lesions in 82 patients, SBVLs in 67 patients, and other lesions (varix and diverticulum) in 15 patients. According to the Yano-Yamamoto endoscopic classification of SBVL, type 1, type 2, and type 3 lesions was included in our study and Type 4 lesions were not included in our study. Among 314 patients presenting with gastrointestinal bleeding, 67 (21%) had one or more SBVL (type 1, type 2, and type 3 lesions) detected at DBE (Fig. 2). 67 patients underwent a total of 122 therapeutic DBE procedures (117 via oral route and 5 via anal route).
Fig. 2Jejunal angioectasia (A), argon plasma coagulation of jejunal angioectasia (B), active bleeding a ileal angioectasia (C), argon plasma coagulation of multiple duodenal angioectasia (D)
A total of 67 (mean age 68.1 ± 8.9 years; 44 males) patients with SBVL were enrolled in this study. Angioectasia was diagnosed in 62 (92.5%) patients, dieulafoy lesion in 1 (1.5%) patient, and arteriovenous malformation in 4 (6%) patients. The median insertion depth was 320 cm (range 60–500) by the oral route and 150 cm (range 60–160) by the per anal approach. The median procedure durations were 60 min (range 25–120). Patients and DBE characteristics are shown in Table 1.
Table 1Baseline descriptive and clinicopathological characteristics of the study groupVariablesTotal (n = 67)Age, mean ± SD, years68.1 ± 8.9Sex, n (%) Male44 (65.7) Female23 (34.3)GIB, n (%) Occult44 (65.7) Overt23 (34.3)Drugs, n (%) Antiplatelet18 (26.9) Anticoagulant16 (23.9) Dual therapy5 (7.5)Comorbidities Portal Hypertension1 (1.5) Cardiovascular disease25 (37.3) Chronic renal failure15 (22.4) Cerebrovascular disease1 (1.5) Diabetes mellitus21 (31.3) Hypertension32 (47.8)Number, n (%) Single10 (14.9) Multiple57 (85.1)Localization, n (%) Duodenum29 (43.3) Jejenum54 (80.6) Ileum15 (22.4)Diagnosis, n (%) Angiodysplasia62 (92.5) Arteriovenous malformation4 (6.0) Dieulafoy lesion1 (1.5)DBE characteristics (n = 122)DBE type, n (%) Oral117 (95.9) Anal5 (4.1)Insertion depth per oral, median (min-max), cm320 (60–500)Insertion depth per anal, median (min-max), cm150 (60–160)Duration of examination, median (min-max)60(25–120)GIB: gastrointestinal bleeding, DBE: double-balloon enteroscopy, SD: standart deviation
Actively bleeding angioectasia was observed in five of the patients at the time of DBE. Fiftyseven patients with angioectasia were treated with APC alone, and 5 patients were treated in combination (one patient APC + sclerotherapy and four patients APC + clipping). The only patient diagnosed with a dieulafoy’s lesion could only be treated with sclerotherapy because clipping treatment could not be performed because of technical failures. Four patients diagnosed with AVM were treated with APC alone. The edge of 2 treated angioectasia lesions and the edge of 1 dieulafoy’s lesion were tattooed. No adverse event associated with DBE procedures and endoscopic therapy (such as perforation or pancreatitis) was observed in this study. No patients required angiographic embolization, hormonal therapy, or surgical resection during the follow-up period.
The characteristics of patients with angioectasia according to the number of DBE endotherapy were examined and summarized in Table 2. The median follow-up period was 25 (IQR: 8–46) months. The median time interval between DBEs was 4 (IQR: 3–10) months. The median number of additional DBE’s was 3 (IQR: 2–5). Rebleeding was observed in 17 of the 62 patients (27.4%) in this cohort. Seven of the 17 patients experienced rebleeding during the first year. The mean age of patients who underwent one DBE endotherapy was 69.7 ± 8.1 years, while those who underwent more than one procedure had a mean age of 67.8 ± 8.5 years (p = 0.443). Among the patients with one DBE endotherapy, 62.2% were male and 37.8% were female. In the group with more than one DBE endotherapy, 76.5% were male and 23.5% were female, and no significant difference was observed between groups (p = 0.290). A significantly higher proportion of patients in the multiple DBE endotherapy group (47.1%) were using anticoagulants compared to those in the single DBE endotherapy group (17.8%) (p = 0.026). The prevalence of cardiovascular disease was significantly higher in the multiple DBE endotherapy group (67.4%) compared to the single DBE endotherapy group (31.1%) (p < 0.001). Similarly, the prevalence of chronic renal failure and diabetes mellitus were significantly higher in the multiple DBE endotherapy group compared to the single DBE endotherapy group (p < 0.001; p = 0.032, respectively). In the multiple DBE endotherapy group, all patients (100%) had multiple lesions, while in the single DBE endotherapy group, 77.8% of the patients had multiple lesions (p = 0.049) (Table 2). A significantly higher proportion of patients in the multiple DBE endotherapy group (64.7%) had duodenal localization compared to those in the single DBE endotherapy group (33.3%) (p = 0.026).
Table 2Characteristics of patients with angioectasia for number of double-balloon enteroscopy endotherapyVariablesDBE endotherapy number > 1(n = 45)DBE endotherapy number > 1(n = 17)p valueAge, mean ± SD, years69.7 ± 8.167.8 ± 8.50.443Sex, n (%) Male28 (62.2)13 (76.5)0.290** Female17 (37.8)4 (23.5)GIB, n (%) Occult bleeding29 (64.4)10 (58.8)0.683** Overt bleeding16 (35.6)7 (41.2)Drugs, n (%)Antiplatelet Yes13 (28.9)5 (29.4)0.999 No32 (71.5)12 (70.6)Anticoagulant Yes8 (17.8)8 (47.1) 0.026** No37 (82.2)9 (52.9)Dual therapy Yes4 (8.9)1 (5.9) No41 (91.1)16 (94.1)0.999Comorbidities, n (%)Portal HT Yes0 (0)1 (5.9)0.274 No45 (100)16 (94.1)Cardiovascular disease Yes14 (31.1)11 (67.4) No31 (68.9)6 (35.3) 0.016** Chronic renal failure Yes5 (11.1)10 (58.8) No40 (88.9)7 (41.2) < 0.001*** Cerebrovascular disease Yes0 (0)1 (5.9)0.274*** No45 (100)16 (94.1)DM Yes11 (24.4)9 (52.9) 0.032** No34 (75.6)8 (47.1)HT Yes24 (53.3)8 (47.1)0.659** No21 (46.7)9 (52.9)Number, n (%) Single10 (22.2)0 (0) 0.049*** Multiple35 (77.8)17 (100)Localization, n (%)Duodenum Yes15 (33.3)11 (64.7) 0.026** No30 (66.7)6 (35.3)Jejenum Yes36 (80)14 (82.4)0.999*** No14 (20)3 (27.4)Ileum Yes5 (11.1)6 (35.3)0.056*** No40 (88.9)11 (64.7)*Independent sample t-test, **chi-square (χ2) test, ***Fisher’s Exact test, GIB: gastrointestinal bleeding, DBE: double-balloon enteroscopy, HT: hypertension, DM: diabetes mellitus. P values significant at the 0.05 level were indicated in bold
The univariate logistic regression analysis presented in Table 3 demonstrates that several clinical parameters are associated with the risk of angioectasia recurrence. Notably, the use of anticoagulants (OR = 4.111, 95% CI: 1.212–13.940, p = 0.023), the presence of cardiovascular disease (OR = 4.060, 95% CI: 1.250–13.185, p = 0.020), chronic renal failure (OR = 11.429, 95% CI: 2.991–43.673, p < 0.001), diabetes mellitus (DM) (OR = 3.477, 95% CI: 1.079–11.205, p = 0.037), duodenal lesions (OR = 3.667, 95% CI: 1.136–11.838, p = 0.030), and ileal lesions (OR = 4.364, 95% CI: 1.118–17.028, p = 0.034) were all significantly associated with an increased risk of angioectasia recurrence.
Table 3Univariate and multivariate logistic regression analyses of the association between clinical parameters and risk of angioectasia recurrenceVariablesOR [95% CI] (univariate)P valueOR [95% CI] (multivariate)P valueGender Female (reference) Male1.973 [0.553–7.043]0.295Antiplatelet No (reference) Yes1.026 [0.301–3.496]0.968Anticoagulant No (reference) Yes4.111 [1.212–13.940] 0.023 9.648 [1.729–53.817] 0.010** Dual therapy Yes (reference) No1.561 [0.162–15.052]0.700Cardiovascular disease No (reference) Yes4.060 [1.250–13.185] 0.020* Chronic renal failure No (reference) Yes11.429 [2.991–43.673] < 0.001* 15.683 [2.727–90.203] 0.002*** DM No (reference) Yes3.477 [1.079–11.205] 0.037* HT No (reference) Yes1.286 [0.420–3.933]0.660Duodenum No (reference) Yes3.667 [1.136–11.838] 0.030* 13.509 [1.598–114.168] 0.017*** Jejenum No (reference) Yes1.167 [0.275–4.949]0.834Ileum No (reference) Yes4.364 [1.118–17.028] 0.034* 17.100 [1.477–197.905] 0.023*** The model was adjusted by age. OR: odds ratio, CI: confidence interval. * Variables with p < 0.10 in univariate analysis. Multivariate analysis included variables with p < 0.10 in univariate analysis, and calculated by using the backward stepwise conditional method (four steps).* Statistically significant at p < 0.05 level. HT: hypertension DM: diabetes mellitus
In the multivariate logistic regression analysis, which adjusted for age and included variables with p < 0.10 from the univariate analysis, the associations remained significant for several factors. Anticoagulant use (OR = 9.648, 95% CI: 1.729–53.817, p = 0.010), chronic renal failure (OR = 15.683, 95% CI: 2.727–90.203, p = 0.002), localization of duodenum (OR = 13.509, 95% CI: 1.598–114.168, p = 0.017), and localization of the ileum (OR = 17.100, 95% CI: 1.477–197.905, p = 0.023) were all independently associated with a higher risk of angioectasia recurrence (Table 3).
Gastrointestinal bleeding from the small intestine is rare and SBVLs are the most common lesions for small bowel bleeding [11]. Except for the duodenum and terminal ileum, the small bowel is unreachable by flexible endoscopy [2]. Due to the difficulty in identifying and treating vascular lesions in the small bowel, patients may need numerous procedures to locate and address the source of bleeding in these cases [12]. DBE is a new method used since 2001 for visualize the entire small intestine directly. DBE is very useful not only for diagnosis but also in allowing endoscopic therapy. DBE has an established clinical role in investigating gastrointestinal bleeding [13]. In this study, we investigated SBVL diagnosed with DBE, and analyzed the risk factors for recurrent DBE in angioectasia. We excluded patients with type 4 vascular lesions, such as varices, diverticula, and malignancy, to obtain important information on the optimal management of bleeding from small-bowel angioectasia. In addition, we followed the patients only undergone endotherapy because several reports followed regardless of whether or not they undergone endotherapy [14].
The terms used for vascular lesions in the gastrointestinal tract have not been standardized, and these lesions are sometimes used interchangeably. In this study, the SBVL was subclassified into type 1, 2 and 3 according to the Yano-Yamamoto classification [4]. Similar to previous study, we found the most common type 1 lesions in our study [15]. Elderly patients are at an increased risk of developing vascular lesions such angioectasia because these conditions are more prevalent as people age. Since the patient population in our study was 68.1 ± 8.96 years of age, SBVLs are common. One patient (1.5%) in the current study had a dieulafoy lesion. Recent studies show that dieulafoy lesion was responsible for 1.6–3.5% of bleeding cases undergoing enteroscopy for suspected small bowel bleeding [16, 17].
In our cohort, 740 patients had various indications for double-balloon enteroscopy (bleeding, abnormal radiological findings, polyposis syndromes, abdominal pain, diarrhea, Crohn’s disease, etc.). The fact that bleeding was less of an indication was inconsistent with the results of many previous studies [18, 19]. The most obvious reason for this may be that patients with anemia underwent radiological imaging and had abnormal radiological imaging findings, which resulted in their evaluation in this group. Lahat et al. categorized DBE indications as anemia and bleeding separately, and the bleeding rates were found to be lower in this study [20]. Smilarly, in a study conducted in China, the rate of patients who underwent double balloon enteroscopy due to bleeding was found to be lower than in our study [9].
The current study additionally evaluated risk factors for rebleeding following endoscopic treatment of SBVLs. The risk factors for recurrence in patients with small bowel bleeding have been addressed by several studies. Niikura et al. reported that female gender, liver cirrhosis, warfarin use, overt bleeding and positive capsule endoscopy findings were predictors of rebleeding [21]. In a study of 64 patients with small-bowel angioectasia reported that small bowel angioectasia correlated with age, heart disease, and liver cirrhosis. Small-bowel angioectasia was found to be independently predicted by cardiovascular disease and liver cirrhosis using multivariate logistic regression analysis [22]. According to a prospective multicenter study that enrolled 183 patients, cardiac disease and the presence of overt bleeding was an independent risk factor for rebleeding in patients with gastrointestinal bleeding with small-bowel angioectasia treated by DBE [16]. According to studies, patients with SBVLs are more at risk than those with other lesions to rebleeding episodes [23, 24]. In the literature, scores have been reported to evaluate the rebleeding risk in patients with small bowel bleeding. Ohmiya et al. developed the Ohmiya index that identified patients who developed small bowel vascular diseases and recurrent bleeding [25]. The RHEMITT score developed by Magalhães et al. is a scoring system that predicts the individual risk of small bowel rebleeding [26]. Chronic kidney disease and heart disease are risk factors in these scoring systems, as in our study. Factors that were significantly associated with the risk of rebleeding were cardiovascular disease and chronic renal failure comorbidities, using anticoagulants, and having SBVL in the duodenum. Multivariate analysis showed that the strongest predictive factors for rebleeding were chronic renal failure and localization of duodenum or ileum. Starting from the distal ileum or duodenum, which is the starting point of DBE, may have led to more careful examination of these areas and more detection of lesions. The proximal small bowel is where angioectasias are frequently encountered [27, 28]. They may be dispersed along the length of the small bowel.
Recurrence rates after endoscopic therapy of SBVL remain unclear. The rebleeding rate after DBE has been reported to be as 35–45% [16, 29–32]. However, there are also different reports in various studies. In a multicenter study conducted in Korea, the authors divided patients with angioectasia into two groups according to whether they received endoscopic treatment. The authors found that endotherapy and non-endotherapy groups had rebleeding rates of 15.6% and 38.1%, respectively (the overall rebleeding rate was 22.7%) [33]. According to a recent Japanese study, the rebleeding rate in patients who underwent endoscopic treatment (22.7%) was slightly lower than that in patients who did not undergo endoscopic treatment (39.1%). The overall rebleeding rate was 33.8% [14]. Similarly, studies in Japan and China, the authors reported a rebleeding rates of 10.9% and 25% [22, 34]. This wide range was probably due to the heterogeneity in the definition, population characteristics, follow-up time, and different complete enteroscopy and treatment modalities. The recurrence of bleeding in a significant number of patients with SBVL following endoscopic therapy could be attributed to either the development of new lesions or the absence of untreated lesions that were not found during endoscopic examination. In addition, the follow-up time in previous studies may explain the difference the rebleeding rate. In a French study, the rebleeding rate nearly doubled between the first and third year of follow-up [35]. In our study, the rebleeding rate after one session of endoscopic treatment of SBVL was 27,4%. We believe that the relatively low rebleeding rates observed in our study may be due to the rebleeding criteria. In our study, rebleeding was stringently defined as a documented episode of melena, hematochezia, or a decrease in hemoglobin level > 2 g/dL following negative esophagogastroduodenoscopy and colonoscopy, confirmed by DBE. Shinozaki et al. defined rebleeding only clinically (the rebleeding rate was 37%). They did not perform repeat DBE in some patients [30].
SBVLs are treated using endoscopic endotherapy methods, such as APC or endoscopic contact-probe methods, hemoclipping, and band ligation. No established gold standard treatment is defined. According to Igawa et al., a combination of APC and endoscopic injection sclerotherapy with polidocanol was useful for achieving the successful hemostasis of angioectasia in the small bowel [22]. In current study, depending on the shape and location of the SBVLS and endoscopist experience, endoscopic endotherapy techniques were used. In patients with small bowel bleeding, studies revealed that endotherapy had no discernible impact on the rebleeding rates [14, 33, 36–38]. It is not feasible to identify every lesion or accomplish enteroscopic hemostasis, and there is a chance that the rebleeding rate will be misinterpreted. Therefore, careful observation and long-term follow-up of high-risk patients may be required, independent of hemostatic treatment.
The study demonstrates several strengths that contribute to its impact and clinical relevance. First, by employing Yano’s classification system for SBVLs, the study addresses inconsistencies in gastrointestinal vascular lesion terminology, providing a more structured framework that aids in diagnosis and has the potential to standardize treatment approaches. Additionally, identifying independent predictors of rebleeding, such as anticoagulant use, chronic renal failure, and lesion localization in the duodenum or ileum, offers valuable insights for clinicians, facilitating patient risk stratification and informing treatment decisions and follow-up protocols. Furthermore, the study’s focus on patients who received endoscopic treatment provides a real-world perspective, underscoring the practical value of DBE in diagnosing and managing SBVLs and reaffirming its applicability in clinical practice.
The study has several limitations that may impact the generalizability and robustness of its findings. As a retrospective study conducted at a single center, its results may not be fully applicable to broader populations; including multiple centers and prospective designs could enhance the validation of rebleeding predictive factors. Additionally, the small sample size and lack of randomization may limit the strength of the conclusions, highlighting the need for randomized controlled trials to compare endoscopic treatments and assess long-term outcomes comprehensively. The lack of a standardized treatment protocol, with procedures like argon plasma coagulation and clipping chosen based on the endoscopist’s discretion, also introduces variability, suggesting that a more uniform approach could improve outcome consistency and comparability across studies. Since our study was a 15-year retrospective study, we could not access enteroscopic images of some patients, and type 1 SBVLs were not evaluated in two separate groups as type1a/1b in some patients’ records. Finally, variability in follow-up duration may influence observed rebleeding rates; a standardized follow-up period could better reflect the long-term efficacy of DBE and provide a clearer assessment of post-treatment rebleeding. Additionally, while we included only patients who underwent endotherapy, which means the findings may not be representative of all SBVLs, this focus on patients who received endoscopic treatment is also one of the study’s strengths, as it allows for a more detailed analysis of the outcomes associated with this specific therapeutic approach.
In conclusion, our study demonstrated that endoscopic therapy during DBE in patients with SBVLs is beneficial over the long term. The observed rebleeding rate of 27% underscores the importance of endoscopic intervention. Chronic renal failure, anticoagulant use, and lesion localization in the duodenum or ileum were identified as independent risk factors for rebleeding. These clinical parameters are strong predictors of recurrence and suggest the need for close monitoring and targeted interventions in affected patients. However, accurately assessing rebleeding specifically from the small bowel can be challenging, and theoretical assumptions may not always reflect clinical reality. Despite this, DBE remains a safe and effective procedure for patients with a history of gastrointestinal bleeding. We strongly recommend further research, particularly prospective and multicenter studies, to better understand the long-term outcomes and optimize the management of small bowel vascular lesions.