Reply to Matsushita et al.

I did appreciate the reply letter from Dr Matsui et al.,1 which I believe clarifies most of the issues I had raised in reference to their recent paper.2 However, their reply, which refers to 'a potential confounding factor in their (our) analysis3 may have come from reliance on the FAB guidelines to accurately diagnose patients', is still somewhat misleading. Although the presence of a cytogenetic abnormality is useful in confirming a suspected diagnosis of myelodysplastic syndrome (MDS), its present is not mandated by either FAB or WHO 2001 guidelines. The diagnostic criteria for MDS in both classifications largely overlap. Therefore, the assumption that we might have 'misclassified' cases because of the use of FAB (the only system available in 1996) is incorrect. I completely agree with Dr Matsui et al.1 that the enumeration of bone marrow CD34 positive cells should be further tested, possibly in a randomized prospective way, in order to conclusively confirm its validity for separating the two disorders. The fact that the 'CD34 approach' has not become standard of practice up to now, I suspect, it is also due to the usual interlaboratory variability which plagues immunohistochemistry laboratories in general. As this is not likely to change in the near future, the best approach to this difficult differential diagnosis remains, in my judgment, a comprehensive one inclusive of morphology, immunohistology, flow cytometry and cytogenetics. Marrow cytogenetics still maintains its major prognostic role in this group of hypoplastic marrow disorders.