Emergency single-balloon enteroscopy in overt obscure gastrointestinal bleeding: Efficacy and safety

Abstract

We aimed to evaluate the impact of emergency single-balloon enteroscopy (SBE) on the diagnosis and treatment for active overt obscure gastrointestinal bleeding (OGIB).

Methods

SBE procedures for OGIB were retrospectively reviewed and sub-divided according to the bleeding types: active-overt and inactive-overt bleeding. The patient’s history, laboratory results, endoscopic findings and therapeutic interventions were registered. Emergency SBE was defined as an endoscopy that was performed for active-overt OGIB, within 24 hours of clinical presentation.

Results

Between January 2010 and February 2013, 53 SBEs were performed in 43 patients with overt OGIB. Seventeen emergency SBEs were performed in 15 patients with active overt-OGIB procedures (group A), which diagnosed the bleeding source in 14: angiodysplasia (n = 5), ulcers/erosions (n = 3), bleeding tumors (gastrointestinal stromal tumor (GIST), n = 3; neuroendocrine tumor, n = 1), and erosioned polyps (n = 2). Endoscopic treatment was performed in nine patients, with one or multiple hemostatic therapies: argon plasma coagulation (n = 5), epinephrine submucosal injection (n = 5), hemostatic clips (n = 3), and polypectomy (n = 2). Twenty-eight patients with inactive bleeding (group B) were submitted to 36 elective SBEs, which successfully diagnosed 18 cases. The diagnostic yield in group A (93.3%) was significantly higher than in group B (64.3%)—Fisher’s exact test, p = 0.038.

Conclusion

This study revealed an important role of emergency SBE in the diagnosis of bleeding etiology in active overt OGIB.

Introduction

The new methods available to directly observe the entire small intestine are capsule endoscopy (CE) and balloon-assisted enteroscopy (BAE), with reported similar diagnostic yield, for patients with obscure bleeding.^1–5 There is currently no consensus on the timing of BAE in the evaluation of overt obscure gastrointestinal bleeding (OGIB). It is believed that, as with upper and lower gastrointestinal bleeding earlier timing of endoscopy could result in a higher diagnostic yield and also augment the possibility of a therapeutic intervention.⁶ However, there are very few studies about the concept of emergency BAE for suspected small-bowel bleeding.

We aimed to evaluate the usefulness of emergency single-balloon enteroscopy (SBE) in active-overt OGIB on the diagnosis and enteroscopic treatment of small-bowel bleeding sources. We further compared the diagnostic yield of emergency SBE performed for active overt OGIB versus SBE for inactive-overt OGIB.

Patients and methods

Between January 2010 and February 2013, a total of 128 SBE procedures were carried out in our department, 78.9% for small-bowel examination (n = 101). Other reasons were mainly endoscopic retrograde cholangiopancreatography (n = 10) and percutaneous endoscopic jejunostomy (n = 6).

A total of 62 consecutive SBE examinations were performed for OGIB, 53 procedures for overt OGIB and nine for occult OGIB. Patients with overt OGIB were enrolled in this retrospective observational cohort study and subsequently divided into two groups according to the bleeding types: active-overt bleeding and inactive-overt bleeding. OGIB was defined and classified according to the published position statement of the American Gastroenterological Association (AGA). Emergency SBE was defined as an endoscopy that was performed for active-overt OGIB within 24 hours of clinical presentation.⁶

SBE was performed using either oral (antegrade) or anal (retrograde) insertion, depending mostly on the CE findings, and in some cases depending on radiological findings. The route of insertion of the SBE was found by calculating Gay’s index; the oral route was chosen if the time to lesion/time to cecum was less than 0.75.⁷ For the oral route, the endoscopic procedure was performed in the left lateral position, and no bowel preparation was required. For the anal route, four liters of polyethylene glycol-electrolyte (PEG) solution was given to the patient before the procedure, as soon as SBE was indicated. In some patients both approaches were used. Complete SBE was confirmed by tattooing the small bowel. All of the SBE procedures were performed by two experienced endoscopists.

All procedures were carried out under deep sedation with propofol, which was administered and supervised by an anesthesiologist. We used an SIF-Q180 enteroscope (Olympus Medical Systems, Tokyo, Japan) in all procedures, according to the standard push-and-pull technique.

Each patient had previously undergone at least one upper gastrointestinal endoscopy and colonoscopy (during a previous episode or at the present admission) in our institution that were negative for a source of bleeding.

CE was previously performed in all except one patient, which had a lesion suspected in a computed tomography obviating the need for CE. CE was performed during the present admission, except if there was a previous examination that already identified a potential source bleeding. Capsule findings were read as soon as the CE was downloaded, with a preliminary search for active bleeding. Whenever possible, frequent real-time views were performed.

The patient’s history and laboratory results were collected. Endoscopic findings and therapeutic interventions were registered, as well as complications during the procedure.

For comparison, patients with active-overt OGIB in whom an emergency SBE was performed were included in group A. Patients with inactive bleeding submitted to elective SBE were included in group B.

Endoscopic treatment

Hemostasis

Enteroscopic hemostasis was performed with electrocoagulation by using argon plasma coagulation (APC) or clipping. The equipment for APC included an argon gas source (APC 300; Erbe Elektromedizin, Tuebingen, Germany) and a high-frequency generator (Erbotom ICC 200; Erbe). The argon gas flow rate was set to 1.2–1.6 l/min, and the current was set at 30–40 W. For clipping, Resolution Clip Device (Boston Scientific, Natick, MA, USA) and QuickClip2 (Olympus Medical Systems Co, Ltd, Tokyo, Japan) were used. When necessary, submucosal adrenaline injection was performed prior to electrocoagulation and clipping.

Polypectomy

A simple snare polypectomy was performed to remove pedunculated polyps. Enteroscopic resection of sessile polyps was performed with a polypectomy snare, after the lesion was lifted by a submucosal saline injection.

Statistical analysis

The data were analyzed statistically using descriptive statistics and Fisher's exact test for comparative analysis. The proportion of patients with positive findings at emergency SBE and the control group were compared. Differences were considered significant with p values of < 0.05.

Results

During the study period of three years, a total of 53 SBE procedures were performed in 43 patients with overt OGIB. In our study, the type of overt OGIB was active in 15 patients (group A) and inactive in 28 patients (group B). Patients’ characteristics are summarized in Table 1.

Table 1.

Comparison of patient characteristics in the two groups

	Group A	Group B	p value
Number of patients	15	28
Age
Years (mean)	65.4 ± 10.9	64.3 ± 16.2	ns
Range	45–82	26–86
Sex
Male/female	9/6	16/12	ns
Symptom
Melena/hematochezia	4/11	10/18	ns
Hb value (g/dl)	8.2 ± 3.1	8.4 ± 3.3	ns
Diagnostic rate (%)	93.3	64.3	0.038 a

Hb: hemoglobin. ^aFisher’s exact test, p = 0.038.

Analysis of group A

A total of 17 emergency SBE procedures (nine oral, four anal, two oral+ anal) for active-overt OGIB were performed in 15 patients (nine men, six women; mean age 65.4 years, range 45–83). All patients presented with hematochezia or melena. The mean hemoglobin was 8.2 g/dl (range 4.2–8.4).

The bleeding source was successfully identified in 14 of 15 patients (93.3%) (Table 2). The following diagnoses were made: five cases of angiodysplasia (Figure 1), three cases of small intestinal ulcers and/or erosions (Figure 2), four cases of bleeding tumors, which were diagnosed as gastrointestinal stromal tumor (GIST) (n = 3, Figure 3), and a neuroendocrine tumor (n = 1) and two cases of erosioned polyps (Figure 4). In the remaining patient, only a large diverticulum in the ileum was found (without ectopic gastric mucosa), even though complete enteroscopy was achieved.

Table 2.

Comparison of diagnostic yields between emergency SBE and SBE

		Bleeding identified
		Positive	Negative	Total
Overt OGIB	Active	14	1	15
	Inactive	18	10	28
	Total	32	11	43

SBE: single-balloon enteroscopy; OGIB: overt obscure gastrointestinal bleeding.

Figure 1.

Bleeding angiodysplasia in the jejunum.

Figure 2.

Ulcers in the distal ileum.

Figure 3.

Bleeding gastrointestinal stromal tumor (GIST) in the distal ileum.

Figure 4.

Erosioned congestive polyp in the jejunum.

Endoscopic treatment was performed in nine patients (Table 3), with one or multiple hemostatic therapies, which included argon plasma coagulation (n = 5), injection of epinephrine (n = 5), hemostatic clips (n = 3) and polypectomy (n = 2). Furthermore, biopsies were taken in seven patients and carbon spot tattooing in five. The four patients with bleeding tumors underwent elective laparotomy with partial small-bowel resection.

Table 3.

Comparison of treatment between emergency SBE and SBE

		Endoscopic treatment
		Performed	Not performed	Total
Overt OGIB	Active	9	6	15
	Inactive	15	13	28
	Total	24	19	43

SBE: single-balloon enteroscopy; OGIB: overt obscure gastrointestinal bleeding.

Analysis of group B

A total of 36 elective SBE procedures (28 oral, eight anal) for inactive-overt OGIB were performed in 28 patients (16 men, 12 women; mean age 64.3 years, range 35–83). All patients presented with hematochezia or melena. The mean hemoglobin was 8.4 g/dl (range 5.4–8.9).

The procedures successfully diagnosed 18 cases of 28 (64.3 %) (Table 2): 13 cases of angiodysplasia, three cases of small intestinal ulcers and/or erosions, and two cases of bleeding tumors (GIST, n = 1 and carcinoma, n = 1).

Enteroscopic treatment was carried out in 15 patients (Table 3), with one or multiple hemostatic therapies, which included argon plasma coagulation (n = 13), injection of epinephrine (n = 4) and hemostatic clips (n = 3). Also, biopsy specimens were taken (n = 6), and carbon spot tattooing (n = 2) of lesions was performed.

The overall diagnostic yield of SBE for overt OGIB was 32 of 43 (74.4%). The most common bleeding etiology was angiodysplasia (n = 18), followed by erosions/ulcerations (n = 6) and bleeding tumors (n = 6). There were no complications regarding the examination or endoscopic treatment.

Comparative analysis

In comparative analysis between active- vs inactive-overt OGIB, there were no significant differences between the two groups in terms of age, gender, blood hemoglobin value on admission, or symptoms (Table 1). The diagnostic yield in group A was significantly higher than in group B—93.3 % in group A vs 64.3% in group B (Fisher’s exact test, p = 0.038).

Discussion

We studied the usefulness of SBE in overt OGIB. In our analysis, the overall diagnostic rate of SBE in identifying the source of bleeding for overt OGIB was 74.4%, which is as high as previously reported. In a study by Ohmiya et al.⁸ that included 479 patients with OGIB from seven medical centers, an overall diagnostic yield of 77% was reported.

The most common bleeding etiology was angiodysplasia, followed by erosions and ulcerations, with similar results in both groups of patients. Other previous studies report similar results, with the majority of bleeding sources being angiodysplasia, followed by erosions and ulcerations of various etiologies, polyps, and tumors, including malignancy. May et al.⁹ reported that in 137 patients with suspected small-bowel diseases, 109 (80%) had positive findings, and the majority of cases involved angiodysplasia (37%). Heine et al.¹⁰ reported that in 168 patients with suspected small-bowel bleeding, 123 (73%) had positive findings and the majority of cases involved angiodysplasia (52%).

We aimed to evaluate the concept of emergency SBEs in active-overt OGIB. In our study, we performed emergency SBEs in 15 patients with active-overt OGIB, and we were able to identify the bleeding sources in 14 cases. There are very few studies about the concept of emergency BAE. In a retrospective analysis by Mönkemüller et al.,⁶ a total of 17 emergency double-balloon enteroscopy (DBE) procedures for overt OGIB were performed in 10 patients, with a diagnostic yield of nine in 10. This suggests that in patients investigated for overt OGIB with emergency BAE it is more common to identify bleeding lesions.

In subsequent comparative analysis between emergency SBE and the control group B (inactive-overt OGIB), we found that the overall diagnostic yield was significantly higher in the first group (Fisher’s exact test, p = 0.04). There were no significant differences in the baseline clinical characteristics of the two groups, suggesting that emergency SBE in patients with active-overt OGIB is associated with positive findings.

Other investigators have previously reported this concept in CE. Pennazio et al.¹¹ reported that a source of bleeding was identified in 92.3% of cases of overt bleeding, in contrast to 12.9% for patients presenting with previous overt OGIB.

Using DBE, Mönkemüller et al.⁶ found that in patients investigated for overt OGIB with emergency DBE it is common to find treatable lesions, introducing the concept of emergency DBE.

Although there is currently no definition of emergency BAE, Mönkemüller et al. have suggested that it applies to any procedure performed within 24 hours of clinical presentation, a term that was used in our study. Also, in a study by Ohmiya et al.⁸ that enrolled 479 patients with OGIB from seven medical centers, it was suggested that in emergency cases, DBE should be selected for the diagnosis and treatment of small-bowel bleeding lesions.

In fact, there is currently no consensus on the timing of BAE in the evaluation of overt OGIB. However, it appears to make sense to perform an endoscopy with therapeutic capabilities when evaluating patients with massive small-bowel bleeding. It has been proposed that flexible enteroscopy should be the first-line exploration in patients with active ongoing bleeding with a high probability of therapeutic interventions.¹² It is believed that, as with upper and lower gastrointestinal bleeding, it is possible that earlier timing of endoscopy could result in a higher diagnostic yield and also augment the possibility of a therapeutic intervention. Ultimately, even if no therapeutic intervention can be applied, endoscopy, and hence BAE, permits potentially better mucosal inspection by forward and backward movement of the device and by rinsing with water the blood from the mucosa.

Additionally, BAE permits endoscopic therapy, which appears to be of crucial importance when there is a high chance of finding a bleeding source. In our study, endoscopic treatment was carried out in nine out of 15 patients with active-overt OGIB, with one or multiple hemostatic therapies. A modeled cost-minimization analysis of the management of OGIB proposed BAE as the most cost-effective initial test after standard endoscopy if the goal is treatment or definitive diagnosis.¹³ Another model suggested that initial BAE was a cost-effective approach for patients with OGIB who likely have angiectasias in the small bowel accessible with a single antegrade approach.¹⁴

It has been suggested that in select circumstances, namely when there is a high level of suspicion of small-bowel angiectasias, deep enteroscopy may be considered as the initial small-bowel diagnostic procedure in patients with OGIB.¹⁵

Endoscopic therapy is currently impossible with CE. But, on the other hand, a CE study preceding BAE might be helpful in choosing the initial insertion route and decrease the number of both approaches in BAE investigation.¹⁶ On the other hand, CE is well known to miss significant lesions, such as small-bowel tumors.¹⁷ Because these tests are complementary, a combination has additional benefits.

Additionally, BAE permits retrieval of tissue for diagnosis, which may facilitate the diagnosis and, in patients with bleeding tumors, planning for surgery.¹⁸

In experienced hands, enteroscopy can help avoid the more traumatic open surgery or, at least, aid preparation for surgical intervention and lower peri-interventional risk, by aiding diagnosis and locating the small intestine lesions prior to surgery.¹⁹ In our study, the bleeding tumors diagnosed with enteroscopy were tattooed, and the four patients underwent elective surgery with partial small-bowel resection.

Finally, by stopping the bleeding or guiding an intervention, emergency BAE may reduce the amount of blood transfusion needed.^20,21 It may be an option for patients with dye allergy or renal insufficiency.

However, BAE use requires available experienced endoscopists and endoscopy nurses and complex logistics, since additional skills and time are required regarding preparation, scheduling and patient monitoring during the enteroscopy and subsequent interventions.²²

In summary, our study highlights the usefulness of SBE for the diagnosis and treatment of OGIB. We found that emergency SBE is technically feasible, safe, facilitates both diagnosis and therapy, and enables the clinical management of patients with overt OGIB. We suggest that flexible enteroscopy should be performed as soon as possible in patients with active ongoing bleeding with a high probability of therapeutic interventions.

This study represents a contribution to better understanding the concept of emergency BAE for patients with overt OGIB. We emphasize, however, the need for a more comprehensive prospective investigation aimed at evaluating the impact of early endoscopic treatment and its cost-effectiveness in this subgroup of patients.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Conflict of interest

None declared.