The recent paper by Passmore et al (2003) provides useful information on clinical features and outcome from a large, population-based collection of cases of paediatric myelodysplastic syndrome (MDS) and juvenile myelomonocytic leukaemia (JMML). Although we have seen a few further cases of MDS in the 10 years since our small demographic study was published (Jackson et al, 1993), its incidence remains < 1 per million children per year in our region. So, our continued experience confirms the lower incidence revealed by the national register compared with that found in some other studies. JMML remains almost as elusive a condition in our region as ever; we have seen only one child, with neurofibromatosis, since 1993 that might fit into that category. However, with a total national accrual of less than seven cases of putative JMML per year, it could still simply be by chance alone that we have seen only one (as opposed to two or three) over the last 20 years. It is clear that the other forms of MDS, particularly refractory anaemia, are important in the UK, although perhaps not quite as common as in Danish children.
The suggestion that chronic myelomonocytic leukaemia (CMML) ‘as seen in adults virtually never occurs in children’ could perhaps be tempered slightly. I do agree it is rare. Some of those cases that ‘may have fulfilled the JMML criteria’, had the appropriate tests been done, could well have had JMML. But then, again, they might not. We reported to the register a case of CMML associated with t(13;14)(q12;q32), the details of which have already been published (Bown et al, 1992). The clinical and haematological features were much more in keeping with CMML as it occurs in adults than JMML; in particular, the patient was 8 years old, had no skin rash or purpura, only mild thrombocytopenia and no elevation of fetal haemoglobin. He subsequently developed a granulocytic sarcoma in the neck (Jackson et al, 1993) that heralded generalized relapse.
Passmore et al (2003) suggest that it would be important to determine whether the t(13;14) in their series was an acquired or constitutional abnormality. In our case, the breakpoints were clearly different from those (q10;q10) reported by others as a constitutional abnormality, and we also reported that we had confirmed, from repeated bone marrow examinations and fibroblast culture, that it was an acquired abnormality. It seems possible that our case, reported to the register and already published, and that in the series published by Passmore et al (2003) are the same patient. If not, then this letter serves to highlight yet another case of CMML with t(13;14).
It remains possible that one is not simply splitting hairs in arguments about myelomonocytic leukaemias in children. Perhaps JMML is distinct, biologically at least, from CMML. CMML does indeed occur in some older children, but perhaps only exceptionally in infants. Perhaps JMML, as a biological entity, can even occur in adolescents and young adults, albeit extremely rarely? In comparison, distinguishing chronic granulocytic leukaemia (CGL), by far the most common myeloproliferative (as opposed to dysplastic) disease in children, from JMML has been relatively simple. Yet one can still remember the extraordinary reluctance of some paediatricians and haematologists to make a diagnosis of CGL in someone who was ‘juvenile’.
The next 10 years of the register should help to refine even further our views about this fascinating aspect of haematology. It would be interesting to see studies of granulocyte–macrophage colony-stimulating factor hypersensitivity carried out prospectively in both JMML and CMML, including patients outside the paediatric age group; such studies might help us to decide whether progenitor hypersensitivity was truly useful in distinguishing putative JMML from putative CMML, or whether it is more valuable in distinguishing reactive from leukaemic myelomonocytic proliferations.