Lowering the Threshold for Prophylactic Platelet Transfusion: Is the Risk of Bleeding Increased?
ABSTRACT & COMMENTARY
The relationship between platelet count and bleeding risk was established in the 1960s.1 The data establishing a platelet count of 20,000/mm3 as the threshold for platelet transfusion did not emerge from prospective randomized trials that were designed to optimize the risk/benefit ratio. The threshold was established arbitrarily. However, since that time, it has become clear that a variety of factors other than platelet count contribute to bleeding risk, including fever, compromised platelet function, clotting factor deficiencies, and local factors such as vascular lesions. Furthermore, the increased demand for platelet products has placed serious stresses on blood banks to keep an adequate supply of platelets on hand.
The threshold for platelet transfusion has been examined in several recent studies2,3 and, most recently, by the German Suddeutsche Hamoblastosegruppe. Wandt and colleagues mounted a prospective study comparing the use of a 10,000 vs. a 20,000 platelet count trigger. The 17 participating centers were divided in their belief about the safety of a 10,000 platelet/mm3 trigger, and, so, the study was not a randomized comparison but a nonrandomized comparison between centers that decided to use the 10,000 trigger (8 centers) and those that insisted on keeping the 20,000 trigger (9 centers). The study involved 105 consecutive patients with de novo acute myeloid leukemia (excluding acute promyelocytic leukemia). Two hundred sixteen cycles of induction and consolidation chemotherapy were delivered, and bleeding complications were compared.
In group A, platelets were not given unless the morning platelet count was less than 10,000/mm3 of blood. The threshold for platelet transfusion was 15,000 in the face of fever, clotting factor abnormalities, or hyperleukocytosis. In group B, platelets were given for any morning platelet count less than 20,000/mm3 of blood. The groups were comparable in pretreatment characteristics, and all patients received the same chemotherapy regimen. Bleeding complications occurred in 20 of 110 cycles (18%) in group A and in 18 of 106 cycles (17%) in group B. No bleeding deaths were recorded, and the number of red blood cell transfusions was similar in the two groups. Group A did receive significantly fewer platelet transfusions. Overall, the cost of platelet support was about one-third lower in group A.
Serious bleeding complications (grade 3-4) were encountered, but all seven patients with serious bleeding were in group B and their bleeding was unrelated to their platelet counts. Four of these patients developed serious bleeding in parallel with systemic infections and sepsis. Five of the seven had local lesions that were the site of bleeding, and, in four of seven, bleeding occurred despite platelet counts higher than 20,000/mm3 of blood. Thus, not a single patient experienced a serious bleeding complication as a consequence of withholding platelet transfusions until the platelet count fell below 10,000/mm3 of blood. (Wandt H, et al. Blood 1998;91:3601-3606.)
COMMENTARY
As difficult as it is to accept a change in a tried and true clinical algorithm, the data from this study and two others2,3 agree that patients with acute leukemia are not placed at a significantly increased risk of serious bleeding by reducing the platelet count trigger to 10,000 platelets/mm3 of blood. In addition to these published studies, a recent paper reporting interim results from an Italian multicenter study also supports the 10,000 trigger.4 Although not specifically studied, the reduction in use of platelet products probably also reduces the risk of transmitting a blood borne viral or bacterial infection to the patient. Platelets are often pooled from 4-8 donors, and the contamination of platelets with donor leukocytes is known to be a source of cytomegalovirus and hepatitis viruses.
Of course, the next question that will likely be addressed by the cost-cutters in medicine is whether prophylactic platelet transfusions should be abandoned altogether. Most patients who bleed as a consequence of low platelet counts stop bleeding when platelets are transfused. This question would be a difficult one for me to deal with. Having lost patients from bleeding into the brain and having seen patients who lost vision from retinal hemorrhages, I am not enthusiastic about waiting for bleeding to appear in a life-threatening situation before intervening. Of course, these issues are the most difficult ones for physicians. We all face situations where what is best for the individual patient is not supported by the insurance carrier because of the costs of providing that service to all its clients. In such instances, the patient bears the risk while the company reaps the benefit.
References
1. Gaydos LA, et al. N Engl J Med 1962;266:905-910.
2. Gmur J, et al. Lancet 1991;338:1223-1226.
3. Heckman KD, et al. J Clin Oncol 1997;15:1143-1151.
4. Rebulla P, et al. N Engl J Med 1997;337:1870-1875.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content