Nösslinger and coworkers1 are to be complimented on a carefully conducted retrospective survival study of 431 patients with primary myelodysplastic syndromes (MDSs) comparing the original French-American-British (FAB) proposals2 and the recently published World Health Organization (WHO) proposal classifications of MDS.3 However, we are concerned about the authors' interpretation of the WHO criteria and the resulting impact on their survival studies. The critical changes in the WHO classification from the FAB include the following: (1) lowering the blast percentage for the diagnosis of acute myeloid leukemia (AML) to 20% from 30%, thus eliminating refractory anemia with excess blasts in transformation (RAEB-T); (2) moving dysplastic chronic myelomonocytic leukemia (CMML) into a proposed new category of myeloid disease with features overlapping myelodysplastic syndromes and myeloproliferative disorders; (3) subdividing RAEB into 2 types: RAEB-1 (5%-9% marrow blasts) and RAEB-2 (10%-19% marrow blasts); and (4) separating refractory anemia (RA) and refractory anemia with ringed sideroblasts (RARS) into 2 broad categories based on the presence of multilineage (2 or 3 myeloid cell lines) or unilineage (mainly affecting the erythroid series) dysplasia.
Unfortunately the reference quoted3 for the WHO classification did not provide sufficient information on the precise criteria described in detail in the recently published WHO manual,4 and, in addition, the authors do not reference the Germing et al paper,5 which confirms these new proposals. An additional review of the new WHO classification was published before the final criteria of the precise percent of dysplastic cells and the consideration of merging the dysplastic and proliferative forms of CMML were agreed upon.6
The major difficulty we have with the Nösslinger et al study is the adoption of the “50%” criteria for dysplasia in 2 or more cell lines for refractory cytopenia with multilineage dysplasia (RCMD). In the WHO proposals the threshold of 50% dysplasia has been utilized only in identifying AML with multilineage dysplasia7 but not for the MDS category of RCMD. In fact, in the WHO proposals RCMD is defined as an MDS subgroup with fewer than 5% blasts in the bone marrow, and dysplasia in 10% or more of the cells of 2 or more myeloid lineages (erythroid, granulocytic, and/or megakaryocytic). These criteria were adapted in the study of 1600 patients with MDS by Germing et al,5 although they did elect to use a 40% threshold for megakaryocytes. Germing et al and others8,9 have confirmed the worse prognosis of RCMD compared to RA or RARS. To accurately evaluate the WHO proposals it will be necessary to reassess the “unclassified” group in the Nösslinger et al study utilizing these criteria. It is very likely that the “unclassified” category (MDS-U) would diminish considerably, impacting the survival results.
In addition others have demonstrated that the survival of CMML is dependent on the bone marrow blast percentage10 and that CMML is much more heterogeneous than other subtypes of MDS. In order to emphasize the prognostic importance of the blast percentage in CMML the WHO classification divides CMML into 2 categories, CMML-1 and CMML-2, depending on the blast count in the peripheral blood and the bone marrow. It does not subdivide CMML according to the white blood cell count. In Table 1 of the Nösslinger et al article,1 CMML resembles RAEB in the International Prognostic Scoring System11 (IPSS) distribution. The separation of RAEB into 2 types (5%-9% blasts and 10%-19% blasts) is of importance as the authors demonstrate with a significant difference in IPPS distribution. A similar analysis of their patients with CMML should be performed. Confirmation of the similarities in outcome for RAEB-T and AML in the Nösslinger et al study provides further evidence in support of allowing such patients (20%-30% marrow blasts) to enter AML trials where appropriate.
In summary it is our hope that Nösslinger and colleagues will consider reviewing their data using the recently published WHO criteria. Such an effort would be important, because, although the Nösslinger et al study is interesting, it does not justify any statement about the validity of the WHO classification. We anticipate that a new look at the data of Germing et al would confirm the conclusion that the WHO system does provide improved and relevant guidelines for the classification of patients with MDS.