Portal de Periódicos da CAPES

They took a mulligan and mostly got it right … the issue of prophylactic platelet transfusion for patients receiving autologous stem cell transplantation

2014; Wiley; Volume: 54; Issue: 10 Linguagem: Inglês

10.1111/trf.12821

1537-2995

Charles A. Schiffer,

Blood groups and transfusion

A “mulligan” is a colloquial phrase for a “do over” in golf; it is not strictly legal, but tolerated by some amateurs who have illusions about their abilities. It doesn't always help, however, since the second attempt is almost invariably much worse than, or at best equivalent, to the first. In this issue of TRANSFUSION, Stanworth and colleagues1 provide a somewhat modified analysis of a large, relatively unique, well-conducted, randomized study comparing different approaches to platelet (PLT) transfusion in patients with severe thrombocytopenia. In a previously published version of this study, the authors concluded in the abstract of their article that prophylactic PLT transfusion administered at a PLT count of fewer than 10 × 109/L was of benefit to patients with “hematologic cancers” (the inference being all such patients) compared to an approach where transfusions were administered when bleeding was clinically apparent.2 As they seem to acknowledge in the current analysis, which focuses on the heterogeneity of results in different patient subgroups, the conclusions are much more nuanced than was originally presented. Of note, however, is that this information was readily apparent in the original publication. These issues were mentioned briefly in an accompanying editorial,3 correspondence to the original journal,4 in a follow-up commentary that I wrote5 and, indeed, in the discussion section of the primary article. Unfortunately, in a trend perhaps started and certainly abetted by USA Today, the public is now accustomed to receiving information in small bites that eschew details and subtleties. Physicians are not immune to this temptation; after all, there are hosts of great books published monthly that compete for our reading time (not to mention other diversions such as golf). Thus, we frequently and perhaps disproportionately rely on the headlines and abstracts and/or the brief abstractions of articles to be found on the burgeoning number of medical news Web sites, particularly if the subject is not near and dear to our heart or specialty. Nonetheless, the authors do not conclude that a “therapeutic” transfusion approach is acceptable for autologous transplant recipients but rather state in the abstract that “there is evidence that the effectiveness of prophylactic PltTx may differ between sub-groups, with chemo/alloHSCT patients receiving prophylactic PltTx appearing to show a greater reduction in bleeding outcomes compared to patients following a no-prophylaxis policy” (italics added). The focus in the abstract on the modest methodologic issue of subgroup analysis is disappointing, particularly given the supportive evidence provided by a large retrospective analysis and a subsequent randomized trial of very similar design in autologous transplant recipients done by Wandt and colleagues in Germany.7, 8 These investigators also noted almost no episodes of grade 4 hemorrhage using the therapeutic transfusion approach and have adopted this practice as their standard, a conclusion I would endorse. I would not advocate additional large trials as mentioned in the discussion section: “further research to address the role of prophylactic PLT transfusions in autoHSCT and to establish whether a no-prophylaxis approach is noninferior should include assessments of the impact of different grades of bleeding upon patient health-related quality of life, in addition to full cost-effectiveness analyses of different PLT transfusion policies.” That would represent another mulligan. Many other issues deserve the attention of these well-organized and productive investigators, with my recommendation being to focus on the persisting superstitions lacking substantial evidence surrounding specific PLT count “thresholds” needed to perform invasive procedures in patients with thrombocytopenia.9 It can be difficult to define and quantify the occurrence and magnitude of less severe episodes of hemorrhage, a methodologic issue in PLT transfusion studies of this sort.10 It is striking, however, that the rate of serious and even minor bleeding was so low in these patients with profound thrombocytopenia and other, often severe, complicating infections and medical problems. This observation is true in large cooperative group studies of older patients with acute leukemia, attesting to the skills of the participating clinicians and the maturation of the blood supply system so that PLTs are available rapidly when needed, but mostly to the fact that a “normal” PLT count represents a luxury that the remarkably resilient microvasculature does not require to maintain clinically acceptable hemostasis. The physiology of this well-documented phenomenon is of great interest and deserves further attention. Finally, as noted in these and other studies, when the occasional catastrophic central nervous system hemorrhage happens, it often occurs in patients with higher PLT counts and might not have been prevented by PLT transfusions. How about those other patients with hematologic cancers and in particular those with acute leukemia expected to have long durations of thrombocytopenia after induction chemotherapy? In future studies, acute leukemia patients should not be lumped with recipients of allogeneic transplants, because, like recipients of autologous transplants, the allogeneic recipients also have a predictably short duration of thrombocytopenia using peripheral blood stem cells. The threshold for prophylactic PLT transfusion for leukemia patients has decreased from the long-standing practice of fewer than 20 × 109/L (also never evidence based) to fewer than 10 × 109/L based on results from randomized trials,11 and results from both the Wandt and Stanworth trials suggest that prophylactic transfusions continue to be appropriate for such patients. Whether the prophylactic threshold can be lowered further could be addressed in future trials but we might be getting close to the land of angels on the head of a pin. The critical principle remains, however, that these studies simply provide general guidelines. Decisions must still be highly individualized with considerations for transfusion, often at higher counts, for critically ill or septic patients or those with hyperleukocytosis or other coagulation abnormalities that commonly occur in acute progranulocytic leukemia. Conversely, prophylactic transfusions are often not necessary, even at lower counts, in clinically stable, nonbleeding patients with long-standing thrombocytopenia, such as individuals with myelodysplasia or aplastic anemia. It is not simply a numbers game; there is still an art to medicine and a role for thoughtful clinicians. The author has disclosed no conflict of interest.