AIM: To evaluate the efficacy of endoscopic diagnosis and therapy for jejunal diverticular bleeding.

METHODS: From January 2004 to September 2009, 154 patients underwent double-balloon enteroscopy (DBE) for obscure gastrointestinal bleeding. Ten consecutive patients with jejunal diverticula (5 males and 5 females) at the age of 68.7 ± 2.1 years (range 19-95 years) at Chang Gung Memorial Hospital, Academic Tertiary Referral Center, were enrolled in this study.

RESULTS: Of the 10 patients, 5 had melena, 2 had hematochezia, 2 had both melena and hematochezia, 1 had anemia and dizziness. DBE revealed ulcers with stigmata of recent hemorrhage in 6 patients treated by injection of epinephrine diluted at 1:10 000, Dieulafoy-like lesions in 4 patients treated by deploying hemoclips on the vessels, colonic diverticula in 2 patients, and duodenal diverticula in 3 patients, respectively. Of the 2 patients who underwent surgical intervention, 1 had a large diverticulum and was referred by the surgeon for DBE, 1 received endoscopic therapy but failed due to massive bleeding. One patient had a second DBE for recurrent hemorrhage 7 mo later, which was successfully treated with a repeat endoscopy. The mean follow-up time of patients was 14.7 ± 7.8 mo.

CONCLUSION: DBE is a safe and effective treatment modality for jejunal diverticular bleeding.

Jejunal diverticulum, first described by Sir Astley Cooper in 1807, is a rare lesion of the small intestine seen in 2%-2.3% of small-bowel contrast studies[1,2] and 1.3%-4.6% of autopsy studies[3,4]. Most jejunal diverticula are asymptomatic, but they are usually complicated by diverticulitis with or without bowel perforation, intestinal obstruction, and bleeding, once they become symptomatic[5,6]. Massive gastrointestinal bleeding secondary to jejunal diverticulosis is extremely rare, and only less than 50 cases have been reported[7]. Most of them present with obscure gastrointestinal bleeding.

In the past, jejunal diverticular bleeding was surgically treated[8], with associated operative morbidity and mortality, as well as postoperative ileus[9]. Double-balloon enteroscopy (DBE), a new technique developed by Yamamoto et al[10], using the push-and-pull method, enables optimal to-and-fro observation as well as therapeutic intervention[11]. Herein, we describe our clinical experience with this new technique in the diagnosis and treatment of jejunal diverticular bleeding at one academic tertiary-referral center in Taiwan.

Three patients were treated with non-steroidal anti-inflammatory drugs, one was treated with coumadin, and one was treated with clopidogrel. Esophagogastroduodenoscopy, colonoscopy or DBE showed colonic diverticula in two patients and duodenal diverticula in three patients, respectively. Other imaging studies including computed tomography (CT), barium study, angiography, and capsule enteroscopy (CE) were performed before DBE (Table 1). Five patients underwent CT scan with no positive CT finding found in any patient. Four patients underwent CE with diverticular bleeding detected in one patient. Eight patients underwent barium study and the results are shown in Table 1. Three patients underwent angiography with no positive finding observed in any patient.

All the 10 patients underwent peroral DBE, which was chosen according to the presenting clinical symptoms and signs, estimation of bleeding sites, and the results of previous studies.

Six patients (Table 2) had ulcers with SRH on the margin of diverticulum, which were treated by injection of 1-3 mL epinephrine diluted at 1:10 000 (Figure 1). Four patients had Dieulafoy-like lesions on the fold of diverticulum, which were treated by deploying hemoclips (Figure 2) on the vessels. The lesions were roughly estimated in the proximal to middle jejunum.

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Jejunal diverticular bleeding treated by double-balloon enteroscopy local injection. A: A shallow ulcer with stigmata of recent hemorrhage on the diverticular margin; B, C: Local injection with epinephrine diluted at 1:10 000; D: Tattooing with pure carbon on the distal side of the diverticulum for marking.

Open in New Tab   Full Size Figure   Download Figure

Figure 2 A Dieulafoy-like lesion observed in diverticular fold (A), which was managed by deploying two hemoclips on the vessel (B).

Of the two patients who underwent further surgical intervention, one had a large diverticulum (2.3 cm × 1.5 cm × 1.2 cm in size) and was referred by the surgeon for DBE, one experienced failure of endoscopic therapy due to massive bleeding. One patient underwent a second DBE for recurrent hemorrhage 7 mo later, and was successfully treated with a repeat endoscopy. The mean follow-up time of patients was 14.7 ± 7.8 mo. Occult blood testing was negative in stools of these patients, and no significant decrease was observed in hemoglobin during the follow-up period.

Gastrointestinal bleeding is frequently encountered by the endoscopist in his/her daily medical practice. Ninety-five percent of the bleeders can be identified by conventional upper endoscopy or colonoscopy. However, 5% of bleeders cannot be found even on repeat conventional upper endoscopy or colonoscopy. This phenomenon is known as obscure gastrointestinal bleeding, and can be further divided into frank bleeding (overt bleeding) and iron-deficiency anemia with guaiac-positive stools (occult bleeding)[12,13]. The bleeding site of most obscure gastrointestinal bleeders is in the small intestine. Lesions in potential bleeders can be further classified into 3 types: vascular lesions, neoplastic lesions, or inflammatory lesions, including diverticula[14].

It has been reported that most patients with jejunal diverticular hemorrhage are treated with surgical intervention[7,15]. Jejunal diverticula are frequently seen in the sixth and seventh decades of life, as an acquired lesion with no muscularis layer on the mesenteric side. Patients undergoing surgery for duodenal diverticula are at risk for operative mortality and morbidity, as well as postoperative ileus and complications of general anesthesia[9].

Many modalities have been introduced to evaluate obscure gastrointestinal bleeding, including barium contrast study, CT scan, angiography, and CE. However, they have some limitations. The yield of CT scan for jejunal diverticula is only 2%[1]. Although the yield of CE for obscure gastrointestinal bleeding is 30%-80%[14], CE is not a therapeutic modality for obscure gastrointestinal bleeding and may result in retention of the capsule in diverticulum[16], as occurred in two of our patients. Barium contrast study has a high yield for detection of jejunal diverticula, but it is not a therapeutic modality for gastrointestinal bleeding. It has been reported that angiography plays a role in the diagnosis and treatment of small-bowel bleeding[17,18]. However, this procedure has limitations, including the risk of ischemic damage to the small-bowel wall, and failure to detect all bleeders.

It has been recently reported that DBE, a novel and unique technique, can successfully treat jejunal diverticular bleeding[19,20], and allow to-and-fro observation of the small intestine, and more importantly, permits tissue diagnosis and therapeutic intervention[11,21,22]. Thus, the aim of our study was to evaluate the efficacy of DBE therapy for jejunal diverticular bleeding.

Since endoscopic injection in treatment of nonvariceal hemorrhage was first described by Soehendra et al[23] in 1985, many endoscopic therapeutic modalities including contact or noncontact thermal therapy, and mechanical therapy by deploying a hemoclip on the feeding vessel have been developed for upper and lower gastrointestinal hemorrhage[24].

To our knowledge, this is the first case series report of patients with jejunal diverticular bleeding treated via endoscopy with diluted epinephrine local injection or hemoclips. The complication of injection therapy for small intestine lesions has been reported by Yen et al[25] who reported a case of intestinal wall necrosis after injection therapy. In addition, he reminds us that advanced age, atherosclerotic vascular disease, anemia, hypoxemia, shock and thinner intestinal wall are the risk factors. The amount of diluted epinephrine used in this study ranged 1-3 mL and no intestinal necrosis was found in patients. Nevertheless, further study is needed to determine the ideal amount of diluted epinephrine.

In this study, the mean follow-up time of patients was 14.7 ± 7.8 mo, during which only two patients had surgical intervention due to the large diverticulum and massive bleeding. One patient had recurrent ulcer bleeding after 7 mo, and was successfully treated with a repeat DBE by local injection therapy. No additional bleeding occurred in the other seven patients during the most recent follow-up.

In conclusion, DBE is a safe and effective treatment modality for jejunal diverticular bleeding. If rebleeding occurs, a repeat endoscopic therapeutic procedure should be attempted in addition to surgical intervention.