Real world evidence of treatment patterns in hereditary hemorrhagic telangiectasia

Abstract Hereditary hemorrhagic telangiectasia (HHT), is an inherited bleeding disorder characterized by recurrent epistaxis, mucocutaneous telangiectasia, gastrointestinal (GI) hemorrhage and arteriovenous malformations (AVMs). Currently there are no approved therapies for HHT, and treatment practices are variable. Real-world characterization of treatment strategies and related side effects will be instructive as new therapeutic strategies are developed. A retrospective, multicenter real-world data analysis was conducted using the TriNetX database on 3 July 2025. All patients with a diagnosis of HHT (ICD-10 code I78.0) were examined, excluding those with a history of neoplasm, aplastic anemia, and bone marrow failure to ensure therapy was related to HHT diagnosis. This analysis captured demographic, clinical and treatment data. Treatment examined included bevacizumab, pazopanib, pomalidomide, aminocaproate, tranexamic acid (TXA), octreotide, lanreotide, thalidomide, tacrolimus, sirolimus, tamoxifen and raloxifene. For each treatment, ICD-10 codes for known side effects were assessed. Given the high prevalence of iron deficiency anemia (IDA) in HHT, iron therapy practices were explored. 24,780 patients were identified with a diagnosis of HHT. Amongst these, 38% had epistaxis, 15% pulmonary AVM, 7% brain AVM, and 3% liver AVM. 29% had IDA. TXA was the most frequently prescribed treatment, followed by bevacizumab, tacrolimus, tamoxifen, thalidomide, octreotide, and raloxifene. Due to limited patient numbers (≤10 patients) for pazopanib, pomalidomide, lanreotide, sirolimus, and aminocaproate, demographic and clinical data is not available. Of the 490 patients treated with TXA, 65% had epistaxis, 6% melena and 6% GI hemorrhage. Median hemoglobin was 13.1 g/dL (5.2-18.8 g/dL) and ferritin 38 ng/mL (2.4-906 ng/mL). Among the 140 patients treated with bevacizumab, 85% had epistaxis, and 8% liver AVM. ≤10 patients had GI bleeding or melena. Median hemoglobin was 12.9 g/dL (6.7-17.8 g/dL) and median ferritin was 38 ng/mL (3-828 ng/mL). Hypertension, fatigue or proteinuria occurred in ≤10 patients. Tacrolimus was the third most common therapy (n=70). 29% of patients had epistaxis, and ≤10 patients had a diagnosis of melena, GI hemorrhage, brain AVM, and pulmonary AVM. Median hemoglobin was 13 g/dL (8-16.7 g/dL), and median ferritin was 37 ng/mL (7-536 ng/mL). Abdominal pain was reported in 20 patients (29%). Venous thromboembolism (VTE; comprising DVT and PE) were infrequently reported and occurred in ≤10 patients treated with tamoxifen, thalidomide and bevacizumab. 0 patients in the raloxifene group had VTE. In patients treated with TXA, 20 patients (4%) had PE and 20 patients (4%) DVT. Given the limitations of the TriNetX database, association with VTE and underlying HHT-related treatment was not elucidated. The majority of HHT patients were treated with ferrous sulfate (n=910), followed by iron sucrose, ferrous gluconate, ferric carboxymaltose, iron dextran, and ferric derisomaltose. Epistaxis was the most frequent HHT-related complication in patients treated with iron therapy. Median hemoglobin and ferritin levels were similar across all types of iron therapy. Given the high prevalence of mucocutaneous bleeding in HHT, it is not surprising that TXA was the most frequently prescribed treatment given its known efficacy for this indication. In this cohort, limited off-label use of pazopanib and pomalidomide is reported. Characterization of treatment related side effects posed challenges given the small group size of some treatments. Most treatment related side effects were uncommon and occurred in ≤10 patients, although abdominal pain was reported in almost 1/3 of patients on tacrolimus. Despite historical concerns about thromboembolic risk with the use of some HHT treatments, patients with a diagnosis of VTE were infrequent in those treated with bevacizumab, thalidomide or estrogen. As the most frequently prescribed treatment, TXA provided more robust data on VTE rates. While 8% of patients treated with TXA had VTE, the data does not support a definitive conclusion of increased VTE in patients treated with anti-fibrinolytics. To further elucidate treatment practices in HHT, a multi-center approach incorporating review of medical records would provide valuable insights into real-world practices.