[Drug-induced immune hemolytic anemia: a retrospective study of 10 cases]
A Bollotte,T Vial,P Bricca,C Bernard,C Broussolle,P Sève
DOI: https://doi.org/10.1016/j.revmed.2014.05.009
Abstract:Purpose: Drug-induced immune haemolytic anemia occurs in one case per million and can be fatal. Our aim was to describe the main characteristics and the type of drug involved. Methods: Cases were retrospectively identified using spontaneous notifications collected by our pharmacovigilance centre and the results of immuno-hematological investigations performed by the laboratory of French blood establishment of Lyon between 2000 and 2012. Inclusion criteria were: an immune (positive direct antiglobulin test), hemolytic, anemia (haemoglobin <100 g/L), with at least a plausible causal relationship with drug exposure according to the French method of imputability or the presence of drug-dependent antibodies, and exclusion of other causes of hemolysis. Results: Ten cases (5 men and 5 women, median age 54.4 years) were identified. Causal drugs were ambroxol, beta-interferon, cefotetan, ceftriaxone, loratadine, oxacillin, oxaliplatine, piperacillin-tazobactam, pristinamycin, and quinine. The median time to onset of anemia after starting the culprit drug was 6 days (2 hours to 16 days). The median nadir of hemoglobin was 57.9 g/L (range: 34-78). The direct antiglobulin test was positive in 8 patients: IgG only (n=4), IgG and complement (n=3), and IgA (n=1). Drug-induced immune haemolytic anemia was considered as definite in 5 cases with positive drug-induced antibodies, probable in 4 cases negative for the detection of drug-induced antibodies but with plausible or likely causal relationship with drug exposure, and probable with an autoimmune mechanism in 1 case. Conclusion: The diagnosis of DIIHA is often difficult because of the similarities with autoimmune haemolytic anemia and the inconstant sensitivity of immunologic tests that sometimes required repetitive assessment.