Incidence of Thromboembolic Events Is Increased in a Retrospective Analysis of a Large Cold Agglutinin Disease (CAD) Cohort

Abstract

Abstract Introduction: Cold agglutinin disease (CAD) is a rare form of autoimmune hemolytic anemia (AIHA) with a prevalence of approximately 16 patients per million. CAD accounts for less than 20% of all AIHA and is characterized by immunoglobulin M (IgM) mediated agglutination of erythrocytes and subsequent activation of the classical complement pathway. In addition to hemolysis, which can vary from mild to severe, clinical manifestations of CAD have been reported to include livedo reticularis, acrocyanosis, circulatory symptoms and cutaneous necrosis (Swiecicki, P Et al Blood 2013). The rarity of this disease has resulted in a lack of data regarding its natural history. Thromboembolic events (TEs) are a known complication of other hemolytic disorders, however, little is known about its incidence and severity in CAD. We evaluated the largest cohort of CAD patients to date and compared them to a matched non-CAD cohort to evaluate clinical characteristics and occurrence rate of TEs. Methods: Patients were identified from the Optum-Humedica database, containing de-identified information on claims, medications, lab results, diagnoses, procedures and clinical notes for individuals from all 50 states. The presence of "cold agglutinin disease" must have appeared in their medical record on at least 3 separate dates from 2006-2016. Patients with "cold agglutinin disease" documented on 1 or 2 dates were included only after an independent review and agreement by 2 hematologists. A comparison cohort was identified from the database and matched on gender, race, region, overall follow-up time, age, and entry date into the health plan. Patient characteristics, TEs, treatment with rituximab, and lab parameters [hemoglobin (Hgb), bilirubin, lactate dehydrogenase (LDH)] were evaluated. LDH and bilirubin were included as they represent hemolysis and indirect measures of complement activity in CAD. TEs in both cohorts were identified from claims submitted. Results: 814 patients with CAD and 7,960 comparisons were identified. Median follow-up time for CAD patients was 75.6 months (0-125 months). Seventy percent of CAD patients were over 65 years of age, 62% were women and 84% were Caucasian. Forty-four percent of CAD patients reside in the midwest, 13% in the northeast, 29% in the south, 12% in the west, with 2% unknown. Thirty-one percent of CAD patients had a medical claim for any form of TE compared to 20% in the matched comparisons (p < 0.0001). Eighteen percent of CAD patients had 1 TE compared to 14% of the matched comparisons (p < 0.001). Thirteen percent of CAD patients had 2 or more types of TEs compared to 6% of matched comparisons (p < 0.001). Nearly 10% of all CAD patients had a venous TE compared to 3% of matched comparisons (p < 0.0001). Twenty five percent of all CAD patients had a cerebral TE compared to 16% of matched comparisons (p < 0.0001). Eight percent of all CAD patients had an arterial TE compared to 5% of matched comparisons (p < 0.0001). Of the 94 CAD patients that had available Hgb within a month prior to a medical claim for a TE, 47% had mild anemia (Hgb > 10 g/dL), 29% had moderate anemia (Hgb 8.1-10 g/dL) and 24% had severe anemia (Hgb < 8 g/dL). Forty-one of the 94 patients had LDH and bilirubin values documented within one month of a medical claim for TE. Ninety percent of these CAD patients had evidence of ongoing hemolysis, represented by an abnormal bilirubin or LDH. Seventy-nine percent of CAD patients treated with a rituximab-containing regimen had an abnormal bilirubin or LDH within 12 months of completion of therapy. Conclusion: Thromboembolic events have been an under-appreciated complication of CAD. Patients with CAD had a significantly higher overall incidence of TEs compared to matched comparisons including venous, arterial, and cerebral events. Furthermore, patients with CAD also had a higher incidence of more than one type of TE. The severity of anemia does not appear to predict TE risk, whereas markers of hemolysis (LDH/bilirubin) do. Standard therapy with rituximab did not result in sustained control of hemolysis, as demonstrated by abnormal bilirubin and LDH values in a majority of patients within 12 months of last dose. Therapeutic modalities that can quickly and completely inhibit complement activity in CAD may offer clinical benefits beyond correcting hemolysis and anemia, such as decreasing TE risk. Disclosures: This study was funded by True North Therapeutics, Inc., a Bioverativ Inc. company Disclosures Broome: TrueNorth Therapuetics: Honoraria; Alexion Pharmaceuticals: Honoraria. Cunningham: TrueNorth therapeutics: Honoraria. Rosenthal: Bioverativ Inc: Employment.