Lay Summary: An ssessment of the role of granulocyte transfusions in 2008.
Granulocyte transfusions are requested by clinicians for use in patients with refractory infection or at high risk of developing severe infection (Strauss 2003). Most patients prescribed granulocyte transfusions are those with cancer related neutropenia, who are receiving myeloablative chemotherapy with or without haemopoietic stem cell rescue. Interest in the use of granulocytes remains high (Van Burik & Weisdorf, 2002; Price 2006), and requests for granulocyte components for transfusion have steadily increased in England and Wales during the last five years. This has been driven by publications describing transfusion in neutropenic patients both for therapeutic indications, when they have an infection refractory to antimicrobials (Hubel et al. 2002) and for secondary prophylaxis, in patients who have had severe bacterial or fungal infections previously but who require a further cycle of chemotherapy or haemopoietic stem cell rescue (Kerr et al. 2003, Oza et al., 2006). Recent studies with promising but overall inconclusive results have been reported both in adults (Oza et al., 2006) and children (Sachs et al., 2006).
The exact clinical role for granulocyte transfusions (whether derived from whole blood or collected by apheresis) therefore remains unclear. Potential efficacy including a dose dependent effect has been raised by systematic reviews/meta-analyses (Vamvakas et al. 1996; Vamvakas et al. 1997; Stanworth et al., 2004), and in animal studies. The existing literature is, perhaps not surprisingly, otherwise heavily dominated by case reports and small case series, with the significant attendant risk of publication bias. However, it should be acknowledged that anecdotal evidence of benefit in selected patients from physicians in the UK and abroad can be found, and that a number of very recent publications have again pointed to evidence of benefit, including one study based on biological randomisation – although this study was underpowered to detect an effect on mortality (Oza et al., 2006).