Gastric Intestinal Vascular Ectasia Syndrome: Findings on Capsule Endoscopy

• Introduction
• Case Reports
• Patient 1
• Patient 2
• Discussion
• References

Introduction

Gastric antral vascular ectasia (GAVE) syndrome has been recognized as an important cause of gastrointestinal hemorrhage [1] [2]. Histopathologically, GAVE is characterized by superficial fibromuscular hyperplasia of the gastric antral mucosa, with capillary ectasia and microvascular thrombosis in the lamina propria. Although it is helpful to make a pathological diagnosis, the most reliable means of diagnosing GAVE and other areas of angioectasia in the gastrointestinal tract is by direct endoscopic visualization [2]. Since GAVE was first described, there have been several reports of patients with similar endoscopic findings in other regions of the gastrointestinal tract, including the cardia [3], duodenum [4], colon [5] [6], and rectum [7]. Terms such as “watermelon cecum”, “watermelon colon”, and “watermelon rectum” have been coined to describe such lesions.

We present here two patients with GAVE and anemia who we investigated using the Given M2A wireless capsule video system (Given Imaging Ltd., Yoqneam, Israel). We have described the technique of capsule endoscopy performed in our center previously and also the classification criteria for the resulting findings [8].

Case Reports

Patient 1

A 57-year-old white man was referred for investigation of intermittent melena, which he had had for 2 months. Endoscopy at another center revealed that he had GAVE. His hemoglobin level at presentation was 50 g/L (normal range 120 - 155 g/L) and he subsequently received a total of 23 units of blood. There was no other significant past medical history, he was not taking any prescribed or over-the-counter medications, and he had no constitutional symptoms. He was a retired farmer and his family history was unremarkable.

On laboratory investigation, his liver function tests, including serum albumin, renal function tests, thyroid function tests, blood platelet count, and coagulation profile were all within normal limits. The erythrocyte sedimentation rate was 65 mm/hour (normal range 0 - 9 mm/hour), the antinuclear antibody titer was 1 : 320 (speckled pattern), and the rheumatoid factor was < 20 IU/mL (normal range 0 - 19 IU/mL). There was no blood eosinophilia.

No abnormality was found on colonoscopy. Gastroscopy revealed columns of red ectatic vessels in the antrum, or “watermelon stomach” (Figure [1]); the cardia was also involved, with a few punctate angioectases in a patchy distribution. In addition, there were extensive punctate angioectases in the duodenum, within the reach of the gastroscope (Figure [2]). The small-bowel mucosa appeared otherwise normal and there was no evidence to suggest portal hypertensive gastropathy.

Capsule endoscopy revealed antral lesions with oozing. In addition, punctate angioectases were seen throughout the entire small bowel, with extensive involvement of the distal small bowel and terminal ileum (similar to the appearances found in patient 2 shown in Figure [3]). There were an average of four to eight punctate angioectases per video frame in the distal small bowel. Some of the lesions in the terminal ileum appeared to ooze slightly. The small-bowel mucosa was otherwise normal.

The capsule gastric transit time was 14 minutes and the small-bowel transit time was 3 hours 50 minutes. There was a regional transit abnormality above the ileocecal valve, of about 70 minutes, though this hold-up may occur in normal subjects [9].

Pathological examination of the biopsy specimens from the duodenum, antrum, and the cardia revealed dilated blood vessels within otherwise normal mucosa (Figure [4]).

After capsule endoscopy, we treated the antral angioectases aggressively with argon plasma coagulation. In view of the diffuse distribution of the angioectases in his small bowel, we recommended conservative management with long-term oral iron supplementation for this patient.

Patient 2

A 54-year-old white woman with known watermelon stomach had been treated periodically with argon plasma coagulation at endoscopy. She had had chronic gastrointestinal bleeding for many years and described intermittent melena. In order to maintain her hemoglobin level above 100 g/L (normal range 120 - 155 g/L), she had required multiple blood transfusions (a total of 25 units of packed red blood cells) over the previous 24 months. She also had hypertension, diabetes mellitus, and end-stage renal disease requiring hemodialysis. She stated that she had not been taking nonsteroidal anti-inflammatory drugs, and that she had no prior history of abdominal surgery or liver disease. Laboratory blood tests, including the platelet count, albumin, international normalized ratio, and liver function tests, were within normal limits.

As in patient 1, colonoscopy showed no abnormalities and gastroscopy revealed columns of red ectatic vessels in the antrum (watermelon stomach), with a few punctate angioectases in a patchy distribution in the cardia. Numerous punctate angioectases were visualized in the duodenum and proximal jejunum on examination with a pediatric colonoscope (Olympus PCF-160L; Olympus, Tokyo, Japan), reaching approximately 40 cm beyond the ligament of Treitz. Again, there was no evidence to suggest portal hypertensive gastropathy, and the small-bowel mucosa appeared otherwise normal.

Capsule endoscopy showed antral lesions with oozing. There were also punctate angioectases throughout the entire small bowel, with extensive involvement of the distal small bowel and terminal ileum (Figure [3]). Some of the lesions in the terminal ileum appeared to ooze slightly. There was more marked involvement of the proximal and mid-small bowel in this patient than there was in patient 1. There was, however, a comparable density of angioectases in the distal small bowel, with an average of more than eight punctate angioectases per video frame. The small-bowel mucosa was otherwise normal. The capsule gastric transit time was 15 minutes and the small-bowel transit time was 2 hours 35 minutes and there was no hold-up at the ileocecal valve.

Pathological examination of the biopsy specimens from the duodenum, antrum, and the cardia revealed dilated blood vessels within otherwise normal mucosa, similar changes to those seen in patient 1 (see Figure [4]).

After capsule endoscopy, we treated her antral angioectases aggressively with argon plasma coagulation. As for patient 1, we then recommended conservative management with long-term oral iron supplementation. Her hemoglobin has been maintained above 100 g/L without further blood transfusion or endoscopic therapy, for 18 months.

Discussion

Since GAVE was first described in 1953 by Ryder et al. [10], it has been increasingly recognized as an important cause of gastrointestinal hemorrhage. More than 70 % of patients with GAVE syndrome have neither cirrhosis nor portal hypertension [1]. GAVE has been associated with various autoimmune diseases, such as systemic sclerosis and rheumatoid arthritis, as well as having weaker associations with chronic renal failure, diabetes mellitus, and bone marrow transplantation. Although its pathogenesis remains unknown [1] [2], a number of hypothetical mechanisms have been proposed, including: a) partial antral mucosal prolapse, causing mechanical stress which results in submucosal ischemia and vascular ectasia; b) a possible role for locally released vasoactive substances, such as vasoactive inhibitory peptide or 5-hydroxytryptamine, or even hypergastrinemia; and c) the involvement of unique epitopes of RNA helicase, with a possible link to the autoimmune diseases [2].

To date, several authors have reported cases in which similar endoscopic findings have been found in sites outwith the antrum, including the cardia [3], duodenum [4], colon [5] [6], and rectum [7]. Many of these patients did not have underlying cirrhosis or portal hypertension [3] [5] [6] [7]. We propose the term “gastric intestinal vascular ectasia” (“GIVE”) to describe patients with gastric vascular ectasia with small- or large-bowel involvement. It is probable that GIVE is an under-recognized but potentially significant clinical condition. With the exception of one patient with protein-losing enteropathy in whom the entire colon showed reddish longitudinal stripes [6], vascular ectasia appears to be located within columnar mucosa close to gastrointestinal sphincters (the cardia, antrum, duodenum, cecum, and rectum). In our two patients, although vascular ectasia was observed throughout the entire small bowel, it was more severe in the duodenum and terminal ileum. This leads us to hypothesize that mucosal prolapse close to these sphincters or imbalance of sphincter activity and blood flow may play a role in the pathogenesis of GIVE.

The differential diagnosis of GIVE includes portal or mesenteric enteropathy, radiation enteropathy, and hereditary hemorrhagic telangiectasia syndrome. These diagnoses were unlikely in our patients, who had no prior history of radiation, abdominal surgery, or liver disease. In addition, in portal enteropathy and in radiation enteropathy the small bowel mucosa usually appears abnormal, with mucosal edema and villous swelling seen on capsule endoscopy; in our two patients, the background small-bowel mucosa appeared completely normal on both conventional and capsule endoscopy.

Treatment options for GAVE include endoscopic, pharmacological, and surgical interventions [1] [2]. Endoscopic thermal ablation of these angioectatic lesions is the mainstay of conservative therapy. However, when endoscopic therapy fails and patients experience recurrent acute or chronic gastrointestinal bleeding, surgical resection may be the only curative option [11] [12]. In patients with GIVE, the clinical response to endoscopic therapy of antral disease or antrectomy may be suboptimal if they have extensive small-bowel involvement. If antral disease is well controlled with endoscopic treatment and a patient’s bleeding is thought to arise from ectasia in the distal small bowel, further treatment by segmental small-bowel resection may be considered.

Capsule endoscopy should be considered in the management of patients with GAVE either when duodenal vascular ectasia is noted at endoscopy or when gastrointestinal blood loss continues despite satisfactory endoscopic ablative therapy to the antrum. In these situations, GIVE may be detected by capsule endoscopy, a finding that might direct decisions on further therapy. Antral oozing was seen on capsule endoscopy in both of our patients. In patient 2, after aggressive treatment with argon plasma coagulation, the hemoglobin remained stable for 18 months without further blood transfusion or argon plasma coagulation.

In conclusion, we report two patients with GAVE associated with extensive ectasia of the small bowel, confirmed by capsule endoscopy. We propose the term “gastric intestinal vascular ectasia”, or “GIVE” to describe the co-existence of small- or large-bowel ectasia with the classic antral changes seen in GAVE. We hypothesize that GIVE is an under-recognized, more generalized form of GAVE, which is characterized by vascular ectasia involving any part of, or all of the gastrointestinal tract, with a predilection for regions close to gastrointestinal sphincters. In managing patients with GAVE, we would recommend capsule endoscopy for further evaluation of patients with evidence of duodenal ectasia or for patients who demonstrate an incomplete response to ablative therapy of the gastric antrum. In patients with GIVE, the antral disease should be treated aggressively with endoscopic ablative measures such as argon plasma coagulation; the small-bowel disease may be managed conservatively with long-term iron supplementation.

References

• 1 Burak K W, Lee S S, Beck P L. Portal hypertensive gastropathy and gastric antral vascular ectasia (GAVE) syndrome [review].  Gut. 2001;  49 866-872
  CrossrefPubMedSearch in Google Scholar
• 2 Sebastian S, O’Morain C A, Buckley M J. Review article: current therapeutic options for gastric antral vascular ectasia.  Aliment Pharmacol Ther. 2003;  18 157-165
  CrossrefPubMedSearch in Google Scholar
• 3 Stotzer P O, Willen R, Kilander A F. Watermelon stomach: not only an antral disease.  Gastrointest Endosc. 2002;  55 897-900
  CrossrefPubMedSearch in Google Scholar
• 4 Cales P, Voight J J, Payen J L. et al . Diffuse vascular ectasia of the antrum, duodenum, and jejunum in a patient with nodular regenerative hyperplasia. Lack of response to portosystemic shunt or gastrectomy.  Gut. 1993;  34 558-561
  CrossrefPubMedSearch in Google Scholar
• 5 Tan S L, Huang W S, Lin J W. Watermelon colon: case report and review.  Gastrointest Endosc. 2003;  58 160-162
  PubMedSearch in Google Scholar
• 6 Bak Y T, Kwon O S, Kim J Y. et al . Protein-losing enteropathy with an endoscopic feature of “the watermelon colon”.  Eur J Gastroenterol Hepatol. 1999;  11 565-567
  PubMedSearch in Google Scholar
• 7 Steinberg J, Brandt L J. The watermelon rectum.  Gastrointest Endosc. 1989;  35 340-341
  PubMedSearch in Google Scholar
• 8 Tang S J, Haber G B. Capsule endoscopy in obscure gastrointestinal bleeding [review].  Gastrointest Endosc Clin N Am. 2004;  14 87-100
  CrossrefPubMedSearch in Google Scholar
• 9 Tang S J, Zanati S, Dubcenco E. et al . Capsule endoscopy regional transit abnormality revisited.  Gastrointest Endosc. 2004;  60 1029-1032
  CrossrefPubMedSearch in Google Scholar
• 10 Ryder J A, Klotz A P, Kirsner J B. Gastritis with veno-capillary ectasia as a source of massive gastric haemorrhage.  Gastroenterology. 1953;  24 118-123
  CrossrefPubMedSearch in Google Scholar
• 11 Sherman V, Klassen D R, Feldman L S. et al . Laparoscopic antrectomy: a novel approach to treating watermelon stomach.  J Am Coll Surg. 2003;  197 864-867
  CrossrefPubMedSearch in Google Scholar
• 12 Novitsky Y W, Kercher K W, Czerniach D R, Litwin D E. Watermelon stomach: pathophysiology, diagnosis, and management [review].  J Gastrointest Surg. 2003;  7 652-661
  CrossrefPubMedSearch in Google Scholar

P. Kortan, M. D.

The Center for Therapeutic Endoscopy and Endoscopic Oncology · Victoria Wing 16 - 048 · St. Michael’s Hospital · Toronto

Ontario, M5B 1W8 · Canada

Fax: +1-416-864-5619·

Email: kortanp.po3.smh@smh.toronto.on.ca

References

• 1 Burak K W, Lee S S, Beck P L. Portal hypertensive gastropathy and gastric antral vascular ectasia (GAVE) syndrome [review].  Gut. 2001;  49 866-872
  CrossrefPubMedSearch in Google Scholar
• 2 Sebastian S, O’Morain C A, Buckley M J. Review article: current therapeutic options for gastric antral vascular ectasia.  Aliment Pharmacol Ther. 2003;  18 157-165
  CrossrefPubMedSearch in Google Scholar
• 3 Stotzer P O, Willen R, Kilander A F. Watermelon stomach: not only an antral disease.  Gastrointest Endosc. 2002;  55 897-900
  CrossrefPubMedSearch in Google Scholar
• 4 Cales P, Voight J J, Payen J L. et al . Diffuse vascular ectasia of the antrum, duodenum, and jejunum in a patient with nodular regenerative hyperplasia. Lack of response to portosystemic shunt or gastrectomy.  Gut. 1993;  34 558-561
  CrossrefPubMedSearch in Google Scholar
• 5 Tan S L, Huang W S, Lin J W. Watermelon colon: case report and review.  Gastrointest Endosc. 2003;  58 160-162
  PubMedSearch in Google Scholar
• 6 Bak Y T, Kwon O S, Kim J Y. et al . Protein-losing enteropathy with an endoscopic feature of “the watermelon colon”.  Eur J Gastroenterol Hepatol. 1999;  11 565-567
  PubMedSearch in Google Scholar
• 7 Steinberg J, Brandt L J. The watermelon rectum.  Gastrointest Endosc. 1989;  35 340-341
  PubMedSearch in Google Scholar
• 8 Tang S J, Haber G B. Capsule endoscopy in obscure gastrointestinal bleeding [review].  Gastrointest Endosc Clin N Am. 2004;  14 87-100
  CrossrefPubMedSearch in Google Scholar
• 9 Tang S J, Zanati S, Dubcenco E. et al . Capsule endoscopy regional transit abnormality revisited.  Gastrointest Endosc. 2004;  60 1029-1032
  CrossrefPubMedSearch in Google Scholar
• 10 Ryder J A, Klotz A P, Kirsner J B. Gastritis with veno-capillary ectasia as a source of massive gastric haemorrhage.  Gastroenterology. 1953;  24 118-123
  CrossrefPubMedSearch in Google Scholar
• 11 Sherman V, Klassen D R, Feldman L S. et al . Laparoscopic antrectomy: a novel approach to treating watermelon stomach.  J Am Coll Surg. 2003;  197 864-867
  CrossrefPubMedSearch in Google Scholar
• 12 Novitsky Y W, Kercher K W, Czerniach D R, Litwin D E. Watermelon stomach: pathophysiology, diagnosis, and management [review].  J Gastrointest Surg. 2003;  7 652-661
  CrossrefPubMedSearch in Google Scholar

P. Kortan, M. D.

The Center for Therapeutic Endoscopy and Endoscopic Oncology · Victoria Wing 16 - 048 · St. Michael’s Hospital · Toronto

Ontario, M5B 1W8 · Canada

Fax: +1-416-864-5619·

Email: kortanp.po3.smh@smh.toronto.on.ca