Case of the season: detection of duodenal varices by CT enteroclysis

Small Intestine Varices in Cirrhosis at a High-Volume Liver Transplant Center: A Retrospective Database Study and Literature Review

INTRODUCTION

Esophageal varices are a well-characterized sequela of portal hypertension; however, less is known about varices arising in ectopic locations. We aimed to describe bleeding small intestine varices (SIV) in patients with cirrhosis and compare characteristics and outcomes to published case reports.

METHODS

We performed an institutional chart review using billing codes and natural language processing between 2008 and 2019. Inclusion criteria were adult patients with cirrhosis and SIV verified by endoscopy, video capsule, or imaging. Patients with noncirrhotic portal hypertension and stomal varices were excluded. We examined demographic and clinical factors, characteristics of SIV, bleeding, intervention, and outcomes in our series and collated data from published cases identified during a literature review.

RESULTS

We identified 71 patients with cirrhosis and SIV (18 bled). The literature search yielded 76 cases with bleeding SIV. Our series and published cases were matched for age, sex, liver disease etiology, and SIV location. Length of stay and transfusion requirements were similar. Aggregate initial treatments (number, hemostasis rate) included medical (n = 7, 57%), endoscopic (n = 48, 56%), interventional radiology (n = 31, 77%), and surgery (n = 8, 87%). Hospital and overall mortality rates were higher in our series (22% and 38%) compared with the published cases (5.3% and 18.4%), respectively (P = 0.02 and P = 0.07).

DISCUSSION

A quarter of patients with cirrhosis and SIV experience bleeding, with high transfusion requirements, frequent need for secondary interventions, and high mortality. These findings highlight the need for a multidisciplinary approach and second-line therapeutic modalities in the timely management of bleeding SIV in cirrhosis.

Enteroscopic Management of Ectopic Varices in a Patient with Liver Cirrhosis and Portal Hypertension

Portal hypertension and liver cirrhosis may predispose patients to varices, which have a propensity to bleed and cause significant morbidity and mortality. These varices are most commonly located in the gastroesophageal area; however, rarely ectopic varices may develop in unusual locations outside of this region. Haemorrhage from these sites can be massive and difficult to control; thus early detection and management may be lifesaving. We present a case of occult gastrointestinal bleeding in a patient with underlying alcoholic liver disease where an ectopic varix was ultimately detected with push enteroscopy.

Updates in the pathogenesis, diagnosis and management of ectopic varices

Ectopic varices (EcV) comprise large portosystemic venous collaterals located anywhere other than the gastro-oesophageal region. No large series or randomized-controlled trials address this subject, and therefore its management is based on available expertise and facilities, and may require a multidisciplinary team approach. EcV are common findings during endoscopy in portal hypertensive patients and their bleeding accounts for only 1–5% of all variceal bleeding. EcV develop secondary to portal hypertension (PHT), surgical procedures, anomalies in venous outflow, or abdominal vascular thrombosis and may be familial in origin. Bleeding EcV may present with anaemia, shock, haematemesis, melaena or haematochezia and should be considered in patients with PHT and gastrointestinal bleeding or anaemia of obscure origin. EcV may be discovered during panendoscopy, enteroscopy, endoscopic ultrasound, wireless capsule endoscopy, diagnostic angiography, multislice helical computed tomography, magnetic resonance angiography, colour Doppler-flow imaging, laparotomy, laparoscopy and occasionally during autopsy. Patients with suspected EcV bleeding need immediate assessment, resuscitation, haemodynamic stabilization and referral to specialist centres. Management of EcV involves medical, endoscopic, interventional radiological and surgical modalities depending on patients’ condition, site of varices, available expertise and patients’ subsequent management plan.

CT enteroclysis in the diagnosis of obscure gastrointestinal bleeding: initial results

AIM

To evaluate the usefulness of computed tomography (CT) enteroclysis in patients with obscure gastrointestinal (GI) bleeding.

MATERIALS AND METHODS

In a prospective study, CT enteroclysis was performed in 21 patients (median age 50 years; range 13-71 years) with obscure GI bleeding in which the source of the bleeding could not be detected despite the patient having undergone both upper GI endoscopic and colonoscopic examinations. The entire abdomen and pelvis was examined in the arterial and venous phases using multisection CT after distending the small intestine with 2l of 0.5% methylcellulose as a neutral enteral contrast medium and the administration of 150ml intravenous contrast medium.

RESULTS

Adequate distension of the small intestine was achieved in 20 of the 21 (95.2%) patients. Potential causes of GI bleeding were identified in 10 of the 21 (47.6%) patients using CT enteroclysis. The cause of the bleeding could be detected nine of 14 (64.3%) patients with overt, obscure GI bleeding. However, for patients with occult, obscure GI bleeding, the cause of the bleeding was identified in only one of the seven (14.3%) patients. The lesions identified by CT enteroclysis included small bowel tumours (n=2), small bowel intussusceptions (n=2), intestinal tuberculosis (n=2), and vascular lesions (n=3). All vascular lesions were seen equally well in both the arterial and venous phases.

CONCLUSIONS

The success rate in detection of the cause of bleeding using CT enteroclysis was 47.6% in patients with obscure GI bleeding. The diagnostic yield was higher in patients with overt, obscure GI bleeding than in those with occult obscure GI bleeding.