Prognostic factors of early death in a cohort of 162 adult haemophagocytic syndrome: impact of triggering disease and early treatment with etoposide

• Published in: British journal of haematology Vol. 168; no. 1; pp. 63 - 68
• Main Authors: Arca, Marc, Fardet, Laurence, Galicier, Lionel, Rivière, Sébastien, Marzac, Christophe, Aumont, Cédric, Lambotte, Olivier, Coppo, Paul
• Format: Journal Article
• Language: English
• Published: England 01-01-2015
• Subjects: Adult; Antineoplastic Agents, Phytogenic - therapeutic use; etoposide; Etoposide - therapeutic use; Female; haemophagocytic syndrome; human immunodeficiency virus infection; Humans; Lymphohistiocytosis, Hemophagocytic - diagnosis; Lymphohistiocytosis, Hemophagocytic - drug therapy; Lymphohistiocytosis, Hemophagocytic - etiology; Lymphohistiocytosis, Hemophagocytic - mortality; lymphoma; Male; Middle Aged; Odds Ratio; Outcome Assessment (Health Care); Prognosis; prognostic factor; Retrospective Studies; haemophagocytic syndrome; human immunodeficiency virus infection; etoposide; prognostic factor; lymphoma
• ISSN: 0007-1048; 1365-2141
• DOI: 10.1111/bjh.13102

Summary
Reactive haemophagocytic syndrome is a life‐threatening disease for which factors influencing the outcome remain unclear. We sought to identify determinants of early mortality in patients with reactive haemophagocytic syndrome by conducting a non‐interventional retrospective multicentre study in three tertiary care teaching hospitals over a 6‐year period. The medical files of 162 patients fulfilling our diagnostic criteria of haemophagocytic syndrome were reviewed. Patients were classified according to 30‐d outcome following diagnosis. Thirty‐three patients (20·4%) died within 30 d. Clinical features at diagnosis associated with 30‐d death in univariate analysis were older age (P = 0·004), underlying lymphoma (P = 0·04), lower platelet count (P = 0·001) and elevated aspartate aminotransferase and lactate dehydrogenase (P = 0·04 both). The use of etoposide as a first‐line treatment tended to be associated with a better outcome (P = 0·079). In multivariate analyses, increasing age, decreasing platelet count, underlying lymphoma and no etoposide in the management were associated with a poorer prognosis (P = 0·03, 0·01, 0·003 and 0·04, respectively). These prognostic factors could help to identify those patients more severely affected by reactive haemophagocytic syndrome, who should benefit from aggressive supportive care, combined with specific treatment of the precipitating factor.