HHT-related GI bleeding develops in approximately 30% of HHT patients, typically manifesting in the 5th-6th decades(30, 31, 33, 40, 117, 118). Though most symptomatic patients have GI telangiectases in the stomach (46-75%) and the small bowel (56-91%), up to 30% also have telangiectases in the colon(30-33, 119). The prevalence of GI telangiectases and HHT-related GI bleeding increases with age, varying by the population studied (unselected HHT vs. those with suspected GI bleeding(30-33, 119)), and by genotype(120).

 The cardinal manifestation of GI tract involvement is anemia from occult GI bleeding. Clinically overt bleeding (melena, hematemesis) is less common. Anemia occurs in approximately half of HHT patients(32, 35, 36), with epistaxis often a significant contributor, and this anemia is severe in up to 25% of patients(35). Severe anemia has a considerable effect on QOL(37-40) and cardiovascular morbidity and mortality. Bleeding related complications are also the most common cause for hospitalization amongst HHT patients(41). Given the clinical impact of anemia, and the otherwise occult nature of the GI bleeding, the clinical assessment of the severity of HHT-related GI bleeding is based primarily on anemia severity and hematologic support required to maintain the target hemoglobin. Though some patients are clinically identified as having a “heavy burden” of GI telangiectases, to date endoscopic findings (number, size, distribution of telangiectases) have not correlated well with severity of anemia. Future studies are needed to determine if an endoscopic classification could replace or complement a classification scheme based on anemia severity. A severity classification is needed for HHT-related GI bleeding, as new systemic therapies reach clinical trials and clinical care.

 Esophagogastroduodenoscopy (EGD) remains the diagnostic gold standard for upper GI telangiectases. Capsule endoscopy (CE) has an excellent safety profile but lacks the capability of assessing the stomach(33, 34). Limited data are available comparing CE to EGD in the setting of HHT(30-32) (Supplement Table 5), but suggest the diagnostic yield for the small bowel is similar to EGD. As such, the role of CE remains complementary to EGD when anemia remains unexplained by the severity of epistaxis and gastric involvement, or when the EGD is negative.  

 Though Argon Plasma Coagulation (APC) is the first line of therapy for acutely bleeding GI vascular lesions(42, 43) in non-HHT patients, there are insufficient data supporting its systematic and repeated use in HHT. The rate of recurring lesions is high in non-HHT lesions but has not been studied in HHT. Complications of repeated treatments have not been assessed, and there is considerable variability in expertise among endoscopists(42). Coagulation of bleeding lesions with APC at diagnostic endoscopy is appropriate but repeated sessions should be limited to severe patients who continue to bleed despite systemic therapy. Small series have reported reduction in RBC transfusion requirement and improvement of hemoglobin after planned (capsule endoscopy driven) eradication of telangiectases with APC during double balloon enteroscopy(33, 121). Clinical trials are needed to explore the efficacy of other endoscopic therapeutics, such as Hemoclips, band ligation, Hybrid APC, etc., which may be particularly relevant for larger lesions that are felt to be at higher risk for severe bleeding.

There are small case series and case reports regarding systemic therapies for HHT-related GI bleeding. Early studies and experience suggested benefit with hormonal therapy(122-125), though more recent studies suggest a better benefit-risk ratio for antifibrinolytics(44) and anti-angiogenic therapies including bevacizumab(19, 21, 23, 45)and thalidomide(73, 126), with the 4 studies meeting evidence criteria reported in Supplement Table 5. For mild to moderate GI bleeding, tranexamic acid may prove useful although its effect is probably weak, with studies showing improved nasal bleeding, but no significant improvement in anemia(44). For moderate to severe patients, who are transfusion or IV iron dependent, the use of IV bevacizumab (see also Epistaxis section for additional background details) has shown significant reduction of transfusion requirements in several uncontrolled case series, with a good safety profile(19, 21, 23). Recurrence of GI bleeding after initial response to IV bevacizumab “induction” therapy is common and there is experience with maintenance dosing; the potential long-term benefits as well as the optimal treatment regimen remain to be defined. Other anti-angiogenic drugs (pazopanib, pomalidomide, doxycycline), and specific estrogen receptor modulators (SERMs, such as tamoxifen, raloxifene, or bazedoxifene) may be useful agents(127-129) however evidence in HHT-related GI bleeding remains limited to small numbers of cases.

Approximately 3% of HHT patients have SMAD4 mutation and overlap syndrome with juvenile polyposis syndrome(130). These patients are at high risk of colorectal cancer(131-133) and should be screened aggressively starting from age 15 years. HHT patients without Juvenile Polyposis have colorectal cancer risks similar to the general population and should be screened accordingly. Patients with SMAD4 mutation are also at risk for aortopathy and hyperlaxity and require appropriate screening(134).