Monte Minute‎ > ‎Monte Minute‎ > ‎

Gastrointestinal Angioectasia - 3/21/2013

posted Mar 21, 2013, 11:20 AM by Rohit Das   [ updated Mar 22, 2013, 7:48 AM by Purnema Madahar ]

The venue for this daily – an older woman with a history of rheumatic heart disease à mitral valve stenosis à mitral valve replacement now on long-term anticoagulation presented with a history of recurrent episodes of melena and/or hematochezia. Ultimately, the identified culprit lesion was a cecal angioectasia (AE). Sooooo:

·         What would be the differential in this scenario?

·         What is the epidemiology and pathophysiology behind gastrointestinal angioectasias?

·         How do they present clinically and how can the diagnosis be made?

·         So what are the potential management options?

What would be the differential in this scenario?

·         First off, defining a couple of terms…

o   Melena – refers to black or “tarry” feces, with the black color being a result of oxidation of iron present in hemoglobin during its passage in the ileum and/or colon. Main concept here is this process TAKES TIME – so, though most cases of melena are due to sources from the upper part of the GI tract, it can also occur with slow bleeds from lesions distal to the ligament of Treitz, as far as the right colon.

o   Hematochezia – refers to the passage of bright red blood per rectum, which generally implies lower GI bleeding, as the blood’s transit time through the GI tract hasn’t occurred for a long enough period to allow hemoglobin to oxidize, thus not allowing the bleeding to manifest as melena. Very importantly, very brisk upper GI bleeding (proximal to the Ligament of Treitz) may manifest as hematochezia – this happens 10-15% of the time.

·         Onto the differential…

o   Upper GI Bleeding – in the context of hematochezia, always important to first think about upper GI sources, as it may signify a potentially brisk, severe bleed that requires intensive monitoring and sooner-rather than-later intervention. The most common cause, by far, is Peptic Ulcer Disease. In one series of ~1600 patients with acute hematochezia, 3-15% was due to an Upper GI source.

o   Diverticular Bleeding – diverticular bleeding is the most common cause of brisk hematochezia, representing 40-50% of cases of massive rectal bleeding. Diverticular bleeding is common because diverticula are common – prevalence of diverticula is as high as 60% in patients over 60 years old. Bleeding occurs in around 15% of patients with diverticula, and of those cases the majority (50-90%) is from right-sided diverticula. Since diverticular bleeding is arterial and colonic in origin, it almost NEVER presents with melena – i.e., arterial bleeding from the colon is too fast to allow for transition to melena, even if right-sided in origin.

o   Angioectasias – the true prevalence of GI AE disease is not well known, but some reports indicate that it may be most common cause of hematochezia in patients over 65-70 years old when the bleeding pattern is chronic and/or recurrent (as it was in this patient). Since bleeding is venous in origin and rates of bleeding can be variable, angioectasias – when located in the cecum or ascending colon (which is their usual home) – can present with hematochezia and/or melena. Will delve into much more detail soon…

o   Malignancy/Polyps – in large series, polyps and/or malignancy was the etiology for hematochezia in about 10% of cases. Bleeding from cancers results from erosion and mucosal ulceration, and the bleeding tendency tends to depend on their location – right sided lesions can present with melena (but usually presents with occult bleeding), whereas left-sided lesions may present with hematochezia (more common with rectal cancer than sigmoid colon cancer). Main thing though – bleeding rate typically does NOT vary, and the bleeding manifestation, whether its melena, hematochezia, or occult, is usually uniform throughout the patient’s history.

·         Obviously there are several more, albeit rare, causes of GI bleeding that I haven’t gotten into, but the above are the most common causes to think of when a patient presents with bright red blood per rectum. In THIS patient’s case, the alternating manifestations of hematochezia and melena suggest a right-sided lower GI source OR an upper GI lesion that bleed at a variable rate. Given her age, chronic and recurrent history of bleeding episodes, the most likely etiology would be a cecal angioectasia.


What is the epidemiology and pathophysiology around gastrointestinal angioectasias?

·         The true prevalence of AE is not known. In a pooled analysis of patients undergoing screening colonoscopy, AEs were detected in only 1% of that population. However, in a study that looked at injection studies of right colons removed for surgical reasons in patients over 60 years old, the prevalence was as high 30-50%. The theme here is that AE is difficult to diagnose, and as a result of that, the epidemiology is a bit vague.

·         AE is most frequently found in patients over 60 years old, and certain conditions predispose to AE-related bleeding – end stage renal disease and congenital or acquired von Willebrand disease.

o   One interesting etiology of the latter risk factor is aortic stenosis. The syndrome of AE bleeding in patients with AS is known as Heyde’s Syndrome. Though this association is controversial, the pathophysiology is quite interesting. It’s thought that shearing forces across a stenosed aortic valve leads to disruption of vWF multimers, consequent decreased platelet activation and therefore a tendency to bleed. Acquired vWF deficiency has been described in up to 70-90% of patients with AS, and the severity of this deficiency correlates directly with the severity of AS. Obviously this is a chicken and egg kind of deal – it’s not that AS directly leads to the formation of AE; rather, the bleeding tendency associated with AS makes AE (probably more common than what we think) more clinically apparent and more readily diagnosed, thus leading to studies finding a positive, but not necessarily causal, association...

·         The pathophysiology behind AE is not fully understood, but there a couple of contributing mechanisms – a staged process due to intermittent low-grade obstruction of submucosal veins, and/or increased expression of VEGF…

o   To elaborate on the first mechanism, it’s thought that long-term colonic muscular contraction eventually leads to dilatation and tortuosity of veins that go through the muscular layer of the colon. Such vein disruption occurs most commonly in the cecum, due to greater intramural tension as compared to other areas of the colon. Angiographically, this can be detected by a “late emptying vein” on the venous phase of an angiogram – this is an early hallmark of AE and pathophysiologically very specific.

o   Ultimately, the higher pressures within the veins are transmitted proximally to the capillaries and arterioles, leading to precapillary sphincter dysfunction, and eventual arteriovenous communication. At this stage, AEs are most susceptible to bleeding, and angiographically transition to “early emptying veins.”


How do they present clinically and how can the diagnosis be made?

·         In the context of overt bleeding symptoms (hematochezia and/or melena), AE is invariably located in the colon, and even more specifically, in the right side of the colon (80-90%, based on the pathophysiology mentioned above). Though multiple AEs are usually found in a given patient, only one is usually responsible for bleeding. Diagnosis is difficult as mentioned – typically appearing as a fan-shaped, vascular lesion, the main problem is that AE can look like an innumerous number of other lesions of the GI tract. Additionally, the fragile nature of these lesions makes them very difficult to see in patients who aren’t adequately resuscitated. Angiography is the most reliable way to diagnose AE, but is obviously invasive and has its associated risks, and practically not usually necessary…


So what are the potential management options?

·         Several – endopscopic and angiographic options are both possible. Endoscopically, GI doctors can ablate lesions with lasers and stuff, fry them with a heating probe or mechanically stop bleeding with hemostasis clips (which usually come off after a few weeks). Control of bleeding is achieved endoscopically in 50-80% of cases, and importantly, no option has been proven to be better than another. Angiography is probably more effective – control of bleeding is achieved by either injecting vasopressin into the feeding vessel or, more commonly, microembolization. This is effective in up 80-90% of cases, but can lead to ischemic complications in 5% of patients.

·         Unfortunately, there aren’t any good long-term therapies for AE, short of taking them out surgically. In aortic stenosis related AE, case series have shown that the acquired vWF does improve with aortic valve replacement.  Furthermore, based on the association between AE and VEGF, there have been case reports of Thalidomide (an indirect VEGF inhibitor) improving refractory AE bleeding. An interesting aspect of this case was the patient’s need for Coumadin…it’s difficult to say how much anticoagulation is contributing to the natural history of this patients cecal AE, but the team took an interesting approach of having her follow-up with CT Surgery to perhaps replace her mechanical valve with a porcine one, thus obviating the need for further anticoagulation…interesting…


Attached is an okay review on AE in the elderly, as well as a NEJM paper describing the interesting relationship between AS and acquired vWF deficiency. Though, I had a really hard time finding a good review article on this subject that wasn’t from the 1970s-1980s…one of those topics worth diving inta a textbook for…anyway, enjoy.


Angiodysplasia and Lower Gastrointestinal Tract Bleeding in Elderly Patients

Sharma et. al., Arch Int Med 1995, Volume 155: 807-812

Vincentelli et. al., NEJM 2003, Volume 349: 343-349

Courtesy of Dan Massera - some more recent nuts and bolts around Heyde's Syndrome from NEJM
Loscalzo, NEJM 2012, Volume 367 (20): 1954-56