Introduction
For a long time, the small intestine was considered the “black box” by gastroenterologists because of the impossibility of an endoscopic visualization by a non-surgical method; this was due to its length, anatomical position, and size1. However, in the 21st century, endoscopic techniques were developed in order to study the small intestine without the need for surgery2,3. Study methods include endoscopic video capsule (VCE), enteroscopy with different techniques, and imaging studies.
VCE is considered a first-line study with a high negative predictive value, but with the limitation of not performing therapeutic interventions4. An additional technique is an enteroscopy, which should be performed if the findings of the VCE or other imaging studies are positive since it has a therapeutic possibility (biopsy, coagulation, polypectomy, among others)5. However, enteroscopy is invasive, requires sedation, and is difficult in patients with adhesions6. Some studies show that VCE and double-balloon-enteroscopy (DBE) have comparable diagnostic yields of up to 60 %4.
Among the methods that have been developed, there is the assisted DBE introduced in 2001, single-balloon DBE in 2007, spiral DBE in 2008, and through-the-scope balloon-assisted enteroscopy (TTS-BAE®) in 2013, which uses a standard endoscope without the need for an overtube. In general, these techniques have comparable diagnostic yields 7.8, and their choice depends on local experience and availability. There have been no reported differences in the diagnostic yield, therapeutic performance, or complication rate between spiral enteroscopy and DBE9.10 and between single-balloon enteroscopy and DBE11-13.
There are two possibilities to perform it. The access route is chosen depending on the clinical presentation and review of previous studies. In up to 85 % of cases, a complete examination of the small intestine is achieved when both pathways are used14.15. Its complications are low and are more related to the performance of therapeutic procedures, including pancreatitis (2 %) and perforation (1 %)4. Intraoperative enteroscopy is a useful but invasive diagnostic and eventually a therapeutic procedure; in general, its use should be limited to settings with difficulty performing enteroscopy, either by availability, previous surgery, or severe adhesions14.
The main indication is potential bleeding from the small intestine, which accounts for 5 % of the causes of gastrointestinal bleeding4; it is called this way because by using all diagnostic tools, it is possible to establish the cause of the bleeding 75 % of the time, while obscure gastrointestinal bleeding is diagnosed in the patient with negative endoscopic and small intestine studies16-18.
The study of small bowel bleeding is difficult, can be recurrent, and is related to neoplasms, especially in patients younger than 40 years who are more likely to have small bowel tumors (lymphoma, carcinoid tumors, adenocarcinoma, and hereditary polyposis)14. Therefore, a diagnostic algorithm that rationally uses the locally available tools is required. Considering the above, it is essential to carry out studies in patients brought to DBE that allow us to know the indications, findings, complications, and relationship with other diagnostic methods and their impact on the outcomes.
Methods
Descriptive observational study type case series. The information was collected prospectively. Patients over 18 years old were included, whose clinical indication or diagnostic imaging suggesting small bowel injury, and thus, DBE was performed. It excluded intraoperative enteroscopies. DBE was performed with a Fujinon EN-450T5® enteroscope according to the usual technique and without fluoroscopy (which is not used in most studies).
Information about sociodemographic variables was obtained-sex and age. In regard to the procedure, the following data were collected: indication, duration, access route defined according to the patient’s clinical picture, imaging studies or VCE, distance explored-it was calculated both in centimeters via antegrade starting from the pylorus and retrograde, from the ileocecal valve, adding the advances and subtracting the setbacks approximately- findings, results of biopsies, correlation with other studies, and patient’s clinical status in routine follow-up. Qualitative variables were described by absolute and relative frequencies. The quantitative variables were described using central tendency and dispersion measures: medians and interquartile ranges IQR if the distribution was not normal suggested or means and standard deviations (SD) if otherwise. Diagnostic yield was considered the relative frequency of abnormal findings and the correlation of similar findings between DBE and VCE or imaging studies.
Ethical considerations
The Ethics Committee approved the protocol of the Hospital de San José in Bogotá and the Research Committee of the Fundación Universitaria de Ciencias de la Salud.
This work does not involve additional procedures, but it does use the information from them. Nor is any additional benefit expected since its scope is descriptive. In addition, it is classified as risk-free research according to resolution 8430 of 1993 of the Colombian Ministry of Health.
Results
In total, 49 procedures were performed, four were excluded-three intraoperative enteroscopies and one performed on a child under 18 years; therefore, 45 enteroscopies were included out of 44 patients. The median age was 58 years (IQR: 48.5 to 70.5 years), with a minimum of 18 years and a maximum of 83 years. The majority were women (n = 29, 65.9 %).
The main indication was potential gastrointestinal bleeding from the small intestine (n = 24, 53.3 %) followed by chronic diarrhea (n = 5, 11.1 %) and chronic abdominal pain (n = 4, 8.9 %) (Table 1). The average procedure time was 90.2 minutes (SD: 23.6 minutes), with a minimum of 30 minutes and a maximum of 150 minutes.
Variable | n (%) |
---|---|
Indication | |
Potential gastrointestinal bleeding from the small intestine | 24 (53,3) |
Chronic diarrhea | 5 (11,1) |
Chronic abdominal pain | 4 (8,9) |
Crohn's disease | 3 (6,7) |
Ileum thickening | 2 (4,4) |
Intestinal polyposis (PAF and Peutz-Jeghers syndrome) | 2 (4,4) |
Stenosis (jejunum and ileocecal valve) | 2 (4,4) |
Tumor in the small intestine | 2 (4,4) |
Jejunitis under investigation | 1 (2,2) |
Procedure time in minutes-average (SD) | 90,2 (23,6) |
FAP: Familial adenomatous polyposis.
28 procedures via anterograde and 17 via retrograde were performed. The median length achieved via antegrade was 310 cm (IQR: 222-400 cm), and via retrograde was 195 cm (IQR: 176-475 cm).
Most of the procedures were normal (n = 21, 46.7 %). Among the most frequent findings were tumors (n = 5, 11.1 %), Crohn’s disease (n = 4, 8.9 %), and ulcers or erosions (n = 3, 6.7 %) (Table 2). In five cases (11.1 %), enteroscopy diagnosed lesions within the scope of conventional endoscopic studies, either from intermittent bleeding or undiagnosed lesions. One patient had a giant hiatal hernia with Cameron ulcers that was taken to surgery 11 months after diagnosis. Two patients presented hemorrhage secondary to colon diverticulum managed endoscopically with follow-ups at 14 and 71 months without new bleeding episodes. One presented cecum angiectasis, which was managed with electrocoagulation and a follow-up at 56 months without new bleeding episodes. The fifth case was an immunosuppressed patient with lymphoproliferative syndrome who presented an inflammatory ulcer due to histology in the rectum; anemia was corrected by controlling the underlying pathology with a follow-up at 69 months.
Overall diagnostic yield of BDE was 53.3 % (n = 24 abnormal findings). For the most frequent case, potential gastrointestinal bleeding of the small intestine, this yield was 54.2 %. The correlation of the results with the CVE was 37.5 %, and imaging studies were 42.9 %, whereas either of the two was 40 % (Table 3). There were no complications.
Diagnostics | n (%) |
---|---|
Normal | 21 (46,7) |
Tumor | 5 (11,1) |
Crohn's disease | 4 (8,9) |
Ulcer/Erosion | 3 (6,7) |
Ulcerated diverticulum | 2 (4,4) |
Polyps | 2 (4,4) |
Intestinal polyposis | 2 (4,4) |
Angiectasis in the cecum | 1 (2,2) |
Ischemic enteritis | 1 (2,2) |
Stenosis of the jejunum secondary to bridles | 1 (2,2) |
Postsurgical inflammatory granuloma in the mid-jejunum | 1 (2,2) |
Giant hiatal hernia with Cameron ulcers | 1 (2,2) |
Lipoma in the second duodenal portion | 1 (2,2) |
Study | VCE (n = 16) | Imaging studies* (n = 14) | All studies (n = 30) |
---|---|---|---|
Double-balloon enteroscopy | 6 (37,5 %) | 6 (42,9 %) | 12 (40 %) |
CAT, CAT Enterography, MRI Enterography, intestinal transit. CAT: computerized axial tomography.
Follow-up was achieved in 91.1 % (n = 41), in a median time of 56 months, with a minimum of 4 and a maximum of 87 months. The majority of these patients were asymptomatic and did not require further studies (n = 32, 78.1 %); four of the patients died from causes unrelated to the procedure (Table 4). Of the 24 patients with potential gastrointestinal bleeding from the small intestine, abnormal findings were reported in 13, of whom 2 (15.4 %) had a new episode of bleeding, while of the 11 patients with normal results, only 1 (9 %) presented a new bleeding episode that was considered secondary to an anal fissure in the coloproctology assessment.
Follow-up | n (%) |
---|---|
Follow-up patients | 41 (91.1) |
Follow-up time in months | Median: 56 IQR: 15-65 Minimum: 4 Maximum: 87 |
Findings | |
Resolution of the clinical picture | 32 (78.1) |
Persistence of the clinical picture | 5 (12.2) |
Death | 4 (9.8) |
For patients who underwent pathology studies (n = 21), the main result was cancer (n = 5) and chronic non-specific inflammation (n = 5), followed by infectious enteritis (n = 3) (Table 5), only 15.6 % (n = 7) of the DBE were therapeutic. 84.4 % (n = 38) had a diagnostic indication.
Results | n = 21 (%) |
---|---|
Cancer | 5 (23,8) |
Chronic non-specific inflammation | 5 (23,8) |
Infectious enteritis | 3 (14,3) |
Hyperplastic polyps | 2 (9,5) |
Villous adenoma with low-grade dysplasia | 1 (4,8) |
Crohn's disease | 1 (4,8) |
Hyperplasia of Brunner's glands | 1 (4,8) |
Hamartomatous polyps | 1 (4,8) |
Celiac disease | 1 (4,8) |
Normal | 1 (4,8) |
Discussion
This study describes the indications, findings, complications, and relationship with other DBE diagnostic methods. It also describes patient follow-up and compares the results with those reported nationally and internationally. At the local level, a series of cases of patients undergoing single-balloon enteroscopy 1 and DBE enteroscopy 19 have been published to date; the specific experience in bleeding was also recently published20. The median age in this study was similar to the DBE reported series 5.21-24.
Potential gastrointestinal bleeding of the small intestine was also the most frequent indication of DBE in other case series, including those published in Latin America3.5,19.21,22.24,25 and in a review that included 66 studies with 12, 823 procedures published between 2001 and 201026. This provides ground to define the procedure as a method for diagnosing and treating this condition.
The route of insertion varies according to the most likely location of the lesion, determined mainly by the medical history associated with image studies such as VCE. The anterograde pathway was the most frequent, which is also the most used in the reported case series5.21,23.24. If the location of a lesion is unknown, the antegrade pathway is generally used since the retrograde technique is more complex, requires preparation, less intestine can be explored, and sometimes the ileocecal valve cannot be cannulated22.
The median distance reached via anterograde was similar to that reported worldwide13,23. The median distance reached via retrograde apparently was greater than reported in the literature10,24; it is important to note that these measurements in all series are subjective, non-homogeneous, and therefore, inaccurate, so comparisons are difficult. The rate and clinical impact of the complete visualization of the small intestine are controversial27-29. There seems to be a consensus that this does not guarantee a greater diagnostic or therapeutic yield30.
The average time of the procedure was 90.2 minutes, and these are very variable (40-180 minutes)3,24,31 due to different factors, such as training and experience of the gastroenterologist, history of previous surgery, intestinal adhesions, and obesity30.
Despite the frequency of normal findings26, the benefit of DBE should not be ruled out since the possibility of treatment offered reduces the need for more invasive interventions and surgical risks; moreover, it is a safe method with low complication rates22. The detection rate of abnormal findings can be increased with a rigorous patient selection and early procedure 32.
The VCE has a variable diagnostic yield (between 45 % and 81 %), with an accuracy rate of up to 30 %; it is accepted that the DBE is a complement to its limitations3. In this study, overall diagnostic yield, defined as the percentage of abnormal findings, was 53.3 %, lower than in other studies3,21. This result can be explained in the late performance of the DBE due to administrative issues related to our healthcare system or overdiagnosis of previous studies that increased the amount of DBE that would not have to be performed.
The diagnostic yield for potential gastrointestinal bleeding of the small intestine was 54.2 %, similar to other studies4; however, it was low compared with series with results of close to 80 %3,31. This difference can be explained due to the long time taken from the onset of symptoms to the completion of complementary studies for administrative reasons already noted4,16,33,34.
The most frequent diagnoses in patients with potential gastrointestinal bleeding of the small intestine were tumors (12.5 %: two gastrointestinal stromal tumors (GIST), and one small intestine adenocarcinoma), followed by ulcerated colon diverticula (8.3 %); these results are similar to those reported in the world literature4.
Among the study’s strengths is the follow-up of 91 % of patients for an extended period (median 56 months). 78 % of the patients were asymptomatic and did not require further studies. Among the explanations for this finding, it may be suggested that the pathologies were intermittent alterations, for which the enteroscopy may not have been necessary, or the patients may have received some medical intervention, such as acetylsalicylic acid (ASA) suspension, non-steroidal anti-inflammatory drugs (NSAIDs), antibiotics, anticoagulants, among others, which led to a resolution of the symptoms. Follow-up was investigated for symptom resolution and additional studies. The study was raised as a series of cases with reduced sample size, so cohort studies are required to evaluate hypotheses related to symptom resolution. When the DBE has an abnormal result, it allows targeted management to the pathology; if the result is normal, this allows to indicate a follow-up, considering that it is the most advanced small intestine study.
The new bleeding episodes occurred in 15.4 % (n = 2), one secondary to colonic diverticulum and the other due to anemia secondary to giant hiatal hernia. In 9 % (n = 1) with normal findings that persisted with hematochezia, its cause was reported as an anal fissure. If the DBE reports abnormal findings, there is a possibility of up to 50 % rebleeding; on the other hand, if normal findings are reported, it is only 5 %14. Some authors suggest that there are predictors of recurrent bleeding, such as frequent bleeding episodes and transfusional requirement14.
During the study of potential gastrointestinal bleeding of the small intestine, conventional endoscopic procedures were not useful in 14 % of cases due to either intermittent bleeding or undiagnosed lesions (giant hiatal hernia with Cameron ulcers, ulcerated diverticulum, angiectasis, and rectal ulcer); up to 25 % of hemorrhagic lesions not diagnosed by conventional high or low endoscopy are reported in the literature and can be explained in inappropriate or null intestinal preparations; however, it is also important to insist on the academic education of gastroenterologists that guarantees conventional quality procedures by avoiding the practice of advanced and unnecessarily expensive ones14.
There were no complications in the present study, compared with other publications that mention a low frequency of pancreatitis and perforation3,26,35.
DBE is a useful tool in evaluating the small intestine with therapeutic possibility, low complication rates, and a diagnostic yield that can reach up to 85 %, depending on a rigorous selection of patients(36.37) and a decrease in overdiagnosis in previous studies. We suggest performing it in university hospitals that guarantee an objective training and learning curve15.38,39. The diagnostic yield for bleeding depends on the time interval between the indication of the procedure and its performance; this is only achieved by the patient’s proper understanding of their problem, academic knowledge of medical staff, and administrative entities that contribute to improving this window of opportunity.
The clinical follow-up to patients taken to DBE may define the need for a second procedure, an indication of additional studies, or the resolution of the problem.