Extragastric Dieulafoy lesions (DLs) are extremely rare; in particular, duodenal and jejunal DLs account for only 3.5% of gastrointestinal (GI) bleeding cases.¹ The challenges impeding their diagnosis and treatment include the difficulty in localizing lesions, a high rate of recurrent bleeding, and their rare occurrence.^1–3 We present a case of a young male with jejunal DL who had atypical manifestations. After conventional modalities failed to provide access to the DL, his polypoid lesion was successfully managed by a combination of arterial embolization, intraoperative enteroscopy (IOE), and fluoroscopy-assisted laparoscopy.

A 29-year-old male was transferred to our hospital with sudden life-threatening hematochezia and significant hypotension. He had experienced episodes of vomiting, lower abdominal discomfort, and watery diarrhea before admission. His medical history was unremarkable, and he reported normal bowel habits without any symptoms of GI dysfunction or bleeding. He stated that he did not smoke, drink alcohol, or take any medications. The abdominal examination findings were unremarkable, although a rectal examination revealed cherry red blood. Microscopic examination of his stool sample revealed the presence of erythrocytes and leukocytes, but no parasitic organisms. Initial flexible sigmoidoscopy showed no bleeding focus. The first abdominal computed tomographic angiography (CTA) revealed a large volume of hyperdense clots in the rectum across all three phases, but detected no active extravasation. Subsequent selective angiography and upper and lower GI endoscopy were performed; however, no source of bleeding was identified. As the bleeding was self-limited without any hemostatic intervention, the patient was discharged and closely followed up. Five days later, he was readmitted with recurrent hematochezia, but did not report any other GI symptoms. Repeated conventional endoscopic investigations revealed no evidence of the GI bleeding source. Emergent CTA demonstrated leakage of contrast medium into the jejunum in the right lower abdominal quadrant. This raised a suspicion that the active bleeding lesion arose from the mucosal surface of the jejunum. Soon afterward, arterial angiography was performed, revealing the contrast medium leaking out of a vessel branching from the right ileocolic artery, a branch of the superior mesenteric artery. Coil embolization was successfully performed using three Microflex 10 coils (Microvention Terumo, Tustin, CA, USA), including two Helical 10 Soft coils and one HyperSoft Helical coil. Unfortunately, another recurrent bleeding episode occurred one week later. The third CTA found two small highest-attenuation clots in the hypogastric area (Fig. 1C). To accurately localize the bleeding lesion, video capsule endoscopy (VCE) (MicroCam; Intromedic, Seoul, Korea) was employed, which subsequently detected an active bleeding jejunal polypoid lesion protruding at an approximate distance of 90 to 100 cm below the duodenum-jejunal junction (Fig. 1A). Push upper enteroscopy to the proximal jejunum and push lower enteroscopy to 40 cm from the ileocecal junction were performed, but no lesion was seen. Finally, the patient underwent fluoroscopy-guided laparoscopy with IOE. The radiopaque coils of the previous embolization were utilized to assist the surgeons in lesion localization (Fig. 1B). Through a minor jejunotomy at the site where the applied coil was identified under C-arm X-ray visualization, the endoscope was inserted toward the anal side. After identifying the bleeding lesion, which was located 10 cm below the jejunotomy site, bidirectional endoscopic screening was continued to eliminate the coexisting bleeding lesions (Fig. 2A). The bleeding lesion was then removed by segmental and minimal resection of the jejunum (Fig. 2B). A histological examination revealed a pseudopolyp DL with large submucosal artery ectasia (Fig. 2C). After 4 months of follow-up, the patient remained asymptomatic without any event of recurrent bleeding. These three admissions are summarized in Table 1.

Table 1 . Summary of the Clinical Case through Three Admissions.

Presentation and diagnostic facilities	First admission	Second admission	Third admission
Gastroenteritis symptoms	Yes, before hematochezia	No	No
CTA	Hyperdense material in the rectum	Active extravasation in the right lower quadrant	Hyperdense material in the hypogastric small intestine
VCE	-	-	An active bleeding polypoid lesion in the jejunum
Endoscopy	Bi-directional with no bleeding focus	Bi-directional with no bleeding focus	Push enteroscopy without identifying bleeding focus
Angiography	No active bleeding focus	Active bleeding from a branch of the SMA	Fluoroscopy-assisted with coils
Intervention	Self-limited	Coil embolization	Laparoscopic surgery and IOE

CTA, computed tomographic angiography; VCE, video capsule endoscopy; SMA, superior mesenteric artery; IOE, intraoperative enteroscopy; -, not performed..

Figure 1. Multiple imaging modalities used to localize the bleeding GI lesion. (A) Wireless capsule endoscopy showed an active bleeding polypoid lesion (arrow) in the jejunum. (B) Localization of the lesion (arrow) through the detection of the recent embolic material under fluoroscopy. (C) Computed tomographic angiography revealed an active bleeding lesion (arrow) arising from the mucosal surface of the jejunum.

Figure 2. Dieulafoy lesion detected by intraoperative enteroscopy and its macroscopic and histopathologic analysis. (A) Intraoperative enteroscopy revealed an actively bleeding, protruding, polypoid lesion with a pulsatile vessel (arrow) arising from the mucosal surface of the jejunum. (B) The surgical specimen showed a well-demarcated protruding mass (arrow) measuring 1.8 cm × 1.0 cm × 1.0 cm in the jejunum. The cut surface was grayish-brown, with hemorrhagic changes. (C) A histological examination showed the presence of blood clots within the large, tortuous, caliber-persistent artery through the overlying normal-appearing surrounding mucosa.

DLs are GI vascular abnormalities characterized by solitary mucosal defects eroded by an abnormally large, so-called caliber-persistent, tortuous, submucosal artery. The erosion may result from ischemic injuries related to aging and underlying comorbidities, such as cardiovascular and renal disease, diabetes, heavy alcohol intake, or tobacco use.^1,2,4 DLs can occur at any age, but most often in the fifth decade of life. In contrast to gastric DLs, no sex predominance was observed in jejunal DLs.¹ In agreement with several previous case reports of jejunal DLs, our 29-year-old patient did not present with any previous health problems.^1,3

GI bleeding can have diverse manifestations, and its progression can vary from trivial and self-limited bleeding to life-threatening blood loss.^1,3 Our patient presented with acute, intermittent, recurrent massive hematochezia. Two hypotheses can be proposed to explain his preceding epigastric pain and nausea in the first admission: (1) acute massive blood loss from the DL induced GI tract ischemia, which accounted for the patient’s abdominal pain and nausea; or (2) gastroenteritis per se was a trigger of the onset of the GI bleeding episode. Nevertheless, the absence of these symptoms in the second and third admissions makes the latter hypothesis more compelling.

No standard treatment protocol has been established for DLs.¹ The diagnosis and management strategies of DLs depend on their presentations, lesion sites, and available expertise.^1,3 Endoscopic therapy is often the first choice for a hemostatic intervention.^2,3 It has been demonstrated that for a rapid-onset episode of a self-limited but recurrent bleeding lesion with a normal appearance of the surrounding mucosa, it is difficult to provide an accurate diagnosis, thus, endoscopy should be repeated.^1,5 Although conventional endoscopy was performed repeatedly, we failed to control the active bleeding. Radiographic examinations including multiphase CTA, mesenteric angiography, and nuclear scintigraphic imaging are alternative methods to detect sources of bleeding.^6–8 CTA is the most common and convenient technique to detect bleeding sources. An active bleeding rate at 0.5 mL/min gives the best results, with a pooled sensitivity of 89% and specificity of 85%.^5,6 Finding the highest attenuation can reveal the closest location to the site of active bleeding.⁸ Although DLs have no specific angiographic pattern, several features such as visualization of a non-tapering (caliber-persisting) and tortuous artery at the bleeding site may indicate this entity.^5,6,8 The discrepancy of CTA results regarding the bleeding source could be associated with the presence of several high-density materials within or near the bowel lumen, including ingested materials and fecaliths, which could be mistaken for hemorrhage without evaluating the non-contrast-enhanced phase.⁸ Coil embolization can be indicated for patients who have contraindications for surgery, or who have lesions beyond the reach of therapeutic endoscopy.⁷ In our case, the second CTA showed active extravasation in the jejunum, and the following embolization was performed successfully. Although successful hemostasis was observed in 96% of embolization cases, approximately 20% of them experienced an early recurrence of lower GI bleeding.^1,3,8 To identify the source of bleeding, current guidelines recommend VCE, whose diagnostic yield is expected to be 30% to 73%.^9,10 However, some significant drawbacks of this method might be encountered. Firstly, lesions can be overlooked due to poor bowel preparation. Secondly, rapid or delayed small-bowel transit during the procedure itself could cause small-bowel obstruction. Thirdly and most importantly, therapeutic approaches are not always feasible.⁸ Finally, in many developing countries, VCE is still unrealistically expensive and uncommon. The VCE results of our patient revealed a pseudopolyp lesion above the junction between the jejunum and the ileum. If the patient’s health status improves, single- or double-balloon enteroscopy would be pursued to visualize the small bowel and apply therapeutic modalities as needed. Nonetheless, this strategy is highly dependent on the endoscopist’s skills and the availability of such devices.⁹ In low-resource settings where balloon-assisted enteroscopy is unavailable, push enteroscopy is a choice with a diagnostic yield ranging from 3% to 70%.^7,11 Our patient underwent push enteroscopy, but no bleeding source was reported. While balloon-assisted enteroscopy was unavailable at our facility, laparoscopic-assisted IOE should be considered as the last resort, since it has yielded excellent treatment outcomes and may help to avoid unnecessary bowel resection.^1,3 This modality is required in 3% to 16% of all DL hemorrhages.¹ In this patient, as the CTA and VCE findings were inconsistent and balloon-assisted enteroscopy was unavailable, we believed that IOE could play a vital role not only in localizing and treating the culprit lesion, but also in screening for coexisting bleeding lesions. In fact, IOE has been reported to provide consistently good results in terms of the cure rate, minimal invasiveness, and improved cosmetic effects in these challenging situations.³ Furthermore, the location of previously deployed metallic coils was used as an important radiopaque marker to guide IOE towards the bleeding location. Postoperative rebleeding has been reported to be rare in previous cases.^1,3 Table 2 summarizes the 12 previously published case studies regarding the management of jejunal DLs. In most of these cases, DLs were detected by multiple modalities, some of which were repeated in various attempts in some cases, and tailored combinations of measures were performed to control the bleeding.^12–23

Table 2 . Literature Review of Case Reports about Jejunal Dieulafoy Lesions.

Author	Patient’s age (yr) and sex	Pre-illness symptom	Diagnosis method for DLs	Therapeutic approach
				Endoscopy	A	IOE
Goins et al14	36, Male	Epigastric pain	Mesenteric angiography	Rigid proctosigmoidoscopy	√	√
Sai Prasad et al21	13, Male	Dizziness	VCE	Not performed		√
Holleran et al16	67–76, Female	Not described	VCE, CT, and mesenteric angiography	DBE	√
Seo et al22	25, Male	Not described	Coil and laparotomy	EGD, colonoscopy, SBE		√
Kozan et al18	21, Female	Not described	IOE	Not performed	√	√
Han et al15	54, Male	No abdominal pain	CT and mesenteric angiography	Not described	√	√
Lee et al19	20, Female	Not described	Radionuclide scanning and angiography	Not performed	√	√
Saada et al20	27, Male	Abdominal pain	Radionuclide scanning	EGD, colonoscopy	√	√
Jaiswal et al17	21, Male	Low abdominal discomfort	IOE	EGD, colonoscopy, push enteroscopy	√	√
Yehya et al23	19, Male	Nausea, abdominal cramping	Radionuclide scanning	EGD, colonoscopy	√	√
Chao et al12	15, Male	Epigastric pain, vomiting	CTA	EGD, colonoscopy	√	√
Dvorská et al13	26, Male	Not described	CTA	EGD, colonoscopy	√	√

DL, Dieulafoy lesion; A, angiography; IOE, intraoperative enteroscopy; VCE, video capsule endoscopy; CT, computed tomography scan; CTA, computed tomographic angiography; DBE, double-balloon enteroscopy; EGD, esophagogastroduodenoscopy; SBE, single balloon enteroscopy..

In conclusion, jejunal DLs presenting with untypical manifestations should not be ignored when they cause episodes of sudden, recurrent, and massive GI bleeding. Bleeding events may be triggered after gastroenteritis. A multidisciplinary approach could be employed for the accurate diagnosis and effective management of jejunal DLs. Metallic coils are not only useful for embolization, but also constitute promising radiopaque markers to localize a lesion during laparoscopy. Although minimally invasive techniques are currently gaining more emphasis, our case is an illustration of the hands-on strategy at centers where advanced endoscopic capabilities are unavailable.

None.

No potential conflict of interest relevant to this article was reported.