Key points

Introduction

Gastrointestinal angiodysplasia (GIAD) is a common cause of gastrointestinal bleeding, and the most common cause of small bowel bleeding in those older than 40 years. Capsule endoscopy and deep enteroscopy have increased the ability to visualize the small bowel and diagnose GIAD at a higher rate than in previous years. This review describes the pathophysiology for the development of GIAD and the current roles of endoscopic, medical, and surgical therapy in its treatment.

Introduction

Gastrointestinal angiodysplasia (GIAD) is a common cause of gastrointestinal bleeding, and the most common cause of small bowel bleeding in those older than 40 years. Capsule endoscopy and deep enteroscopy have increased the ability to visualize the small bowel and diagnose GIAD at a higher rate than in previous years. This review describes the pathophysiology for the development of GIAD and the current roles of endoscopic, medical, and surgical therapy in its treatment.

Definition of gastrointestinal angiodysplasia

GIADs are pathologically dilated communications between veins and capillaries. Histologically, they consist of an accumulation of ectatic, thin-walled veins, venules, and capillaries lined by endothelium in the mucosa and submucosa. The term GIAD was first used to describe a single or multiple, flat, arborized, nonvariceal vascular abnormality in the gastrointestinal (GI) tract, not associated with similar angiomatous lesions on the skin or in organs ( Fig. 1 ). The diagnosis of GIAD is based on the ability to distinguish them from other vascular lesions in the GI tract. The treatment of these vascular lesions is similar for endoscopy and radiography, but differs in surgical and pharmacologic approaches because of the natural history of the lesion.

GIAD of the gastric fundus.

Location and diagnosis of gastrointestinal angiodysplasia

It was once believed that GIADs were mainly found in the right colon (specifically the ascending colon and cecum) and to a lesser degree in the small bowel and stomach. Recent evidence demonstrates the opposite is true, as DeBenedet and colleagues found that of 1125 patients undergoing esophagogastroduodenoscopy (EGD), capsule endoscopy (CE), and colonoscopy, 114 patients were diagnosed with small bowel GIADs. Of the 114 patients diagnosed with GIAD, 32% were diagnosed in the stomach, 50% diagnosed in the duodenum, 37% diagnosed in the jejunum, 15% diagnosed in the ileum, and 44% diagnosed in the colon. Likewise, Carey and colleagues demonstrated in a retrospective study evaluating obscure GI bleeding that small bowel GIADs constituted more than 60% of the clinically significant lesions in patients who underwent CE. Bollinger and colleagues found that most of these GIADs were in the proximal small bowel and throughout the GI tract. CE and deep enteroscopy has allowed for better visualization of GIADs and their presence throughout the small bowel.

Endoscopy is the primary instrument used to diagnose GIAD, followed by CE. The ability of CE for diagnosing small bowel diseases, and it superiority over small bowel radiography, push enteroscopy, computed tomography, and angiography has been accepted in clinical practice. CE can be used to target and to determine the route for other forms of deep enteroscopy before ablation therapy.

Clinical presentation of gastrointestinal bleeding

The clinical presentation of bleeding GIADs can range from occult to a life-threatening event, requiring emergent hemostatic intervention. It has been shown that at first presentation, 90% of GIADs spontaneously stop bleeding, but there is a propensity for rebleeding. Reasons for rebleeding include underlying cardiac valve abnormalities, arrhythmias, left ventricular assist devices, chronic kidney disease, use of anticoagulation, cirrhosis, multiple GIADs, and prior history of the patient with respect to GIAD bleeding. Rebleeding also may be influenced by the location of the GIAD as well. A recently published meta-analysis found the rate of rebleeding to be 34% in generalized GIAD and 45% of GIAD isolated to the small bowel.

Treatments

The modalities for treating bleeding GIADs include endoscopic therapies (argon plasma coagulation [APC], mechanical clip placement, multipolar electrocoagulation [MPEC], and laser photoablation), angiography with embolization, surgical resection, and pharmacologic therapy. Because patients with GIADs are elderly and often have comorbidities, endoscopic therapies may not be the best initial option or the best long-term treatment strategy. The presence of multiple lesions in those patients with aortic stenosis, may be recalcitrant to endoscopic therapy and need aortic valve replacement (AVR).

Endoscopic therapy

Endoscopic therapy is an effective initial therapy for GIADs, although rebleeding rates are high, especially for small bowel GIADs. A recent meta-analysis of 623 patients with GIAD followed for a mean of 22 ± 13 months (range, 6–55 months) found a pooled recurrence rate of 34% (95% confidence interval [CI] 27%–42%). A subanalysis performed for small bowel GIADs revealed that the rebleeding rate increased to 45% (95% CI 37%–52%, I ² = 41%, P = .09). There was no difference in the mean follow-up times for all patients with GIADs compared with those with only small bowel GIADs ( P = .5).

Initial endoscopic therapy seems to be successful, but given that most rebleeding events occur between 1 to 2 years after therapy, this might be related to the presence of observational bias. There are reasons for higher rebleeding rates for patients who have small bowel GIADs, which include incomplete deep enteroscopy, specifically the insertion depth of the enteroscope, as this metric determines the ability to complete GIAD eradication. Small bowel insertion depth plays a significant role in the management of GIAD, because most GIADs are multiple and hidden throughout the small bowel. The possibility that missed GIADs play a role in rebleeding rates is very likely. Patients with recurrent small bowel bleeding or iron deficiency anemia with an initial negative capsule endoscopy have been found to have GIADs in 29% at repeat CE. This observation will lead to changes in patient management. It could be that there was an interval development of new GIADs or more likely that these lesions were missed on the initial CE. Some recent reviews of endoscopic therapy for GIADs suggest that risk reduction for rebleeding rates may be little if any, fitting with our previous suggestions. Another challenge in treating GIAD is that the disease course has periods of exacerbation and remissions. Half of patients will stop bleeding with long intervals between the time to the next bleeding episode ; thus, rebleeding may not be the best marker for the effectiveness of therapy. Some have suggested that the number of blood transfusions and hospitalizations for bleeding from GIADs may be a better endpoint. Pinho and colleagues observed a decrease in the transfusion rate 1 year after endoscopic therapy, but given the natural history of GIADs, that may have been too short an observational period. The most effective therapy is still unclear, although APC is the most studied therapy ( Figs. 2 and 3 ), other cauterization techniques (such as MPEC) and mechanical therapies (hemoclips or ligation) might be as effective or more effective in certain situations. More controlled, prospective studies are needed and will likely be forthcoming as technologies advance. Better recognition of GIAD will be important in determining the effectiveness of treatment.

Small bowel GIAD before APC therapy.

Small bowel GIAD after APC therapy.

Virtual chromoendoscopy used in concert with CE might enable the identification of previously unrecognized lesions, most of which are GIADs. Real-time images using narrow band imaging (NBI) may provide the same effect as flexible spectral imaging color enhancement (FICE) for the precise recognition of GIAD ( Figs. 4 and 5 ). In theory, it should have a similar diagnostic effect as FICE by increasing the recognition of true vascular lesions and by excluding insignificant red spots. Another potential technological advancement is the use of the Endocuff device on the push enteroscope to better visualize more of the GIADs. In personal experience, this technique enhances GIAD detection.

Small bowel GIAD on white light.

Small bowel GIAD on NBI.

Superselective transcatheter embolization

Superselective transcatheter embolization of bleeding GIAD has been shown to be successful in 80% to 90% of patients. The most used agents are biodegradeable gelatin sponges and microcoils. An advantage of this technique is that it can be repeated if rebleeding occurs. Complications occur from this technique in 5% to 9% of patients, with serious complications in fewer than 2%. The complications include hematomas, bowel infarction, arterial dissection, thrombosis, and pseudo-aneurysms.

Surgical therapy

Surgical options for GIADs include intraoperative enteroscopy (IOE) to target areas of GIADs and partial surgical resection or guided endoscopic therapy, and AVR in patients diagnosed with aortic stenosis and extensive GIADs.

There is limited published information on the IOE and GIADs other than anecdotes. Douard and colleagues reported on 15 patients with GIADs post IOE and demonstrated a recurrence rate of 30% at 19 months post IOE.

An important potential factor for increased bleeding of patients with GIAD has come from the observation of the association of loss of von Willebrand factor in patients with aortic stenosis. It has been shown that the loss of the high-molecular-weight multimers of von Willebrand factor in aortic stenosis (AS) is a cause of bleeding for those patients with GIADs. King and colleagues observed that 91 patients with AS had chronic anemia due to overt or obscure bleeding. Of these, 40 patients who did not undergo AVR continued to bleed. Thirty-seven patients underwent abdominal exploration, and cessation of bleeding was reported in only the 2 patients who underwent bowel resection. Of the 16 patients undergoing AVR, only 1 (6%) patient rebled after a mean follow-up of 9 years. Thompson and colleagues studied 57 patients with AS and GIAD who underwent AVR. Over a median follow-up period of 4.4 years, 45 (79%) did not demonstrate rebleeding. One prospective study demonstrated that the abnormalities in von Willebrand factor multimers and platelet function revered back to normal after AVR in all patients postoperatively. It seems that once these patients receive AVR, the acquired von Willebrand deficiency may be reversed secondary to restoration of high-molecular-weight multimers of von Willebrand factor, but some patients continue to manifest ongoing bleeding, despite AVR.