Different transfusion care strategies are considered for the treatment of acute thrombopenia due to low or absence of production of platelets in Acute Myeloid Leukemia (AML) patients with controversies about both threshold of platelets to consider prior to transfusion and the quality of the product to transfuse. In addition, despite two recent studies (Wandt et al, Lancet. 2012 Oct 13;380:1309; Stanworth and al. N Engl J Med. 2013 May 9;368:1771), the choice between preventive (based upon the number of platelets) and curative (based on clinical symptoms) care, as discussed in 2004 by Slichter and al. (Transfus Med Rev. 2004 Jul;18(3):153), still remains a concern although preventive care is commonly used. In this study, we compared the treatment of AML patients at the Institut Paoli-Calmettes (Marseille, France, that elicited curative transfusion care strategy) herein called IPC, and Centre Hospitalier Lyon Sud (Pierre Benite, France, that elicited preventive transfusion care strategy) herein called CHLS, between January 2001 and December 2010, considering the use of transfusion product during induction, and death due to hemorrhage.

At the time of study, the curative transfusion care strategy at the IPC relied on the analysis of three complete blood counts a week, clinical examination of the patients (Pt) at least twice a day, and platelet transfusion in case of mucosal hemorrhage syndrome, headaches, fever, and high blood pressure. At the CHLS, the preventive transfusion care strategy relied on blood count, daily clinical examination of the Pt, and platelet transfusion when platelet number was found below 20 G/L. Both IPC and CHLS used Platelet Concentrate (PC) obtained from apheresis from single donor (CPA) or multiple donors (MCPS) and a threshold of 80 g/l Hb was considered for red blood cell transfusion. AML patients were treated by cytarabine combined with anthracycline (induction regimen), with subtle differences between CHLS and IPC and according to recommendation at the time of treatment.

Our analysis covered a 45 day induction time, and the 6 first months (185 days) of treatment. Between January 2001 and December 2010, 884 patients (median age 59 with 80% < 70) and 524 patients (median age 52 with 94% <70) were treated at the IPC and at the CHLS, respectively. 74% vs 82% of the patients underwent complete remission (CR) following one or two induction regimens, and 39% vs 40% are still alive. During induction, patients at the IPC and the CHLS received 5,1 ± 3,9 (mean ± Standard Deviation) and 10,2 ± 7 PC, with 4,4 and 9,8 CPA and 1,3 and 1,3 MCPS, respectively. During this period, the number of packed red blood cell unit used was 5.7 ± 4,6 and 10.6 ± 5,8, respectively, and 21 Pts deceased from hemorrhage at the IPC versus 2 at the CHLS, with 9 Pts refractory to platelet transfusion versus 1. When not considering Pts refractory to transfusion and Pts with disseminated intravascular coagulation (DIC), 7 Pts died at the IPC (at a median of 12 days after diagnosis) versus 1 patient at the CHLS at day 16 after diagnosis. Of note, ≥ grade 3 sepsis supported in intensive care unit was observed in 5 over 7 patients at the IPC.

This retrospective study confirms the results suggested in the TOPPS and German studies that have compared preventive and curative platelet transfusion strategies for leukemia patients. Two fold less platelet products are used in a curative strategy, but more patients deceased from hemorrhage, most of them earlier during the treatment and with high grade sepsis. Unexpectedly, two fold more packed red blood cell units are used in the preventive strategy. This study also suggests that increase of the number of platelet concentrate infused without regard to ABO typing could decrease the therapeutical impact of red blood cell transfusion. At the IPC, high number of patients deceased from hemorrhage in context of DIC or refractoriness to transfusion may be relied to the platelet and plasma transfusion strategy elicited in these particular situations. Altogether, these results suggest that risk factors like sepsis, have to be considered to elicit a preventive versus curative platelet transfusion strategy in the treatment of AML Patient.

Six months results and further statistical analyses are in progress and will be presented.


Prebet:Celgene Corporation: Honoraria. Vey:BMS: Honoraria.

Author notes


Asterisk with author names denotes non-ASH members.

This icon denotes a clinically relevant abstract