Long-term safety and effectiveness of systemic bevacizumab for chronic severe bleeding in idiopathic gastrointestinal angiodysplasia, esophageal varices, and angiodysplasia of von willebrand disease and acquired von willebrand syndrome Free

Background Angiodysplasia of the gastrointestinal (GI) tract may be idiopathic or occur secondary to other causes, such as von Willebrand disease. Bleeding from GI angiodysplasia may be chronic and severe, resulting in severe anemia and dependence on hematologic support (IV iron and RBC transfusion). Because GI vascular lesions are often numerous and recur despite procedural intervention, few durable therapeutic options are available for these patients. Given the angiogenic dysregulation associated with the genesis of these lesions, and the success of systemic bevacizumab in the management of hereditary hemorrhagic telangiectasia (HHT), bevacizumab may have therapeutic utility in patients with bleeding GI vascular lesions of other etiologies. Data describing use of systemic bevacizumab in this setting is very limited, primarily describing case reports and use for short durations.

Methods We performed an observational study of non-HHT patients with severe chronic bleeding from GI vascular lesions of any etiology refractory to standard therapies who were treated with off-label systemic bevacizumab. Data was extracted via manual chart review by physicians. The primary endpoint was hematologic support (RBC transfusions and IV iron infusions) requirements, measured in RBC unit equivalents (1 RBC unit equivalent = 1 RBC unit or 250 mg IV elemental iron). Secondary endpoints included hemoglobin, IV iron and RBC transfusions measured independently, and safety.

ResultsPatients & Dosing. 23 patients with chronic, severe bleeding from GI vascular lesions received bevacizumab. Median (range) age was 78 (42-90) years, all but 1 patient was age ≥60, and 52% were female. All patients initiated at 5 mg/kg daily (77%) or 2.5 mg/kg daily (23%) and were infused every 2 weeks for 4 treatments, followed by 2.5-5 mg/kg daily every 4 weeks. Patients were treated for up to 4.2 years.

Hematologic Outcomes. Hematologic support required significantly improved from 6 months pretreatment [median (IQR) 14.2 (10.0-23.9) RBC unit equivalents] during months 1-6 of bevacizumab treatment [4.1 (2.0-7.1) RBC unit equivalents, P=0.0003] and further improved during months 7-12 of bevacizumab treatment [median (IQR) 0.6 (0.0-3.5) RBC unit equivalents, P>0.001]. These hematologic support improvements coincided with improvements in median (IQR) hemoglobin, measured at 8.2 (7.7-9.4) g/dL at baseline, 10.1 (8.6-12.9) g/dL at month 3, 12.2 (9.6-13.0) g/dL at month 6, 10.6 (8.8-13.1) g/dL at month 9, and 10.4 (8.9-13.3) g/dL at month 12 of bevacizumab treatment (P=0.0039). Both individual components of required hematologic support (IV iron and RBC transfusion) improved from 6 months pretreatment to months 1-6 on treatment and months 7-12 on treatment: median (IQR) milligrams elemental iron infused 1530 (780-2030) vs 1020 (0-1530), P=0.053 and 0 (0-510), P<0.0001, respectively, and units RBCs transfused 8.0 (2.0-16.5) vs 0.0 (0.0-2.0), P=0.002 and 0.0 (0.0-1.5), P=0.0017, respectively. Twenty patients (87%) required RBC transfusion in the 6 months pretreatment vs 9 (37%) in the 12 months after initiating bevacizumab (P=0.0018).

Healthcare Utilization. Patients in the cohort required 60 emergency department visits/hospital admissions in the 12 months specifically for GI vascular lesion bleeding in the 12 months prior to bevacizumab initiation, versus 21 ED visits/admissions in the 12 months after initiation. The median (IQR) number of ED visits/admissions per patient improved from 3.5 (1-5) in the 12 months prior to treatment to 1 (0-1.25) in the 12 months after initiation (P=0.0049).

Safety and Discontinuation. Proteinuria was the most common treatment-emergent adverse event (TEAE), occurring in 7 patients (30%), followed by hypertension in 6 (26%), and fatigue in 3 (13%). No patient suffered from a venous or arterial thromboembolic event. There were no fatal TEAEs. Three patients (13%) discontinued for TEAEs (all grade 3 proteinuria) and 5 (22%) discontinued for inadequate treatment effect.

Conclusions In this study, systemic bevacizumab was safe and highly effective in treating severe chronic bleeding from GI vascular lesions across a wide variety of non-HHT etiologies, including idiopathic angiodysplasia, esophageal varices, von Willebrand disease and acquired von Willebrand syndrome in a highly comorbid mostly older adult/late elderly population. Systemic bevacizumab may be a novel therapeutic option for clinically challenging vascular bleeding outside of HHT.