Publications : 2019

Röth A, Fryzek J, Jiang X, Reichert H, Morales J, Broome CM. Seasonal patterns of anemia, hemolytic markers, healthcare resource utilization, and thromboembolic events in cold agglutinin disease. Presentation at German Society for Hematology and Medical Oncology, Berlin, Germany, October 11–14, 2019.


Introduction: Cold agglutinin disease (CAD) is a rare form of autoimmune hemolytic anemia in which circulating IgM autoantibodies preferentially bind to the I antigen on red blood cells at low temperatures, resulting in chronic complement-mediated hemolysis. Patients with CAD have an increased risk of thromboembolic events (TEs). Although it is established that cold weather can elicit some of the circulatory symptoms of CAD (eg, acrocyanosis), its association with other CAD manifestations is not well understood. We therefore compared hemoglobin (Hgb), markers of hemolysis (bilirubin and lactate dehydrogenase [LDH]), healthcare resource utilization (HRU), and TE rates between seasons for patients with CAD.

Methods: Patients with CAD were identified from the Optum Humedica database. Hgb, bilirubin, LDH levels, HRU measures (inpatient days, outpatient visits, and emergency room visits), and number of transfusion days were evaluated. TEs were identified using diagnostic codes. Data were compared between seasons using logistic regression adjusted for age, sex, race, region, year, Charlson Comorbidity Index, and clustering within patients.

Results: 808 patients with CAD were identified (63% female; 66% aged ≥65 years). The median minimum Hgb for winter as compared with summer was decreased by 0.54 g/dL (P<0.001). The median maximum bilirubin and LDH increased by 0.12 mg/dL (P=0.005) and 42.1 U/L (P<0.001), respectively, in winter vs summer.  No significant differences in HRU measures or transfusion days were observed when stratified by season. One or more TE (n=287) occurred in 204 CAD patients (25%). Of these, 56 (19.5%) were in summer, 57 (19.9%) in fall, 79 (27.5%) in winter, and 95 (33.1%) in spring. Compared to summer, the adjusted TE risk was higher in spring (odds ratio [95% confidence interval]: 1.60 [1.09–2.33]; P=0.016), but not fall (1.06 [0.70–1.61]; P=0.785) or winter (1.42 [0.96–2.12]; P=0.082).

Conclusions: Patients with CAD had evidence of persistent hemolysis across all seasons. Variations in median Hgb, bilirubin, and LDH between winter and summer were not associated with differences in clinical outcomes, as there were no significant changes in HRU or transfusion days. Additionally, there was no association between colder weather and TE risk. The lack of seasonal variability in this cohort suggests that treatment considerations and monitoring of complications such as TEs in patients with CAD should be season independent.