Pediatric Blood & Cancer | 2019
Staging of acute leukemia based on central nervous system involvement alone: Is it appropriate?
Abstract
To the Editor: Youlden et al have reported a large Australian Childhood Cancer Registry study, describing the stage of hematological malignancies in children.1 We congratulate them for the commendable analysis of population-wide distribution and survival outcome of pediatric lymphomas and leukemias in Australia. The Toronto Paediatric Cancer Stage Guidelines have been utilized by the authors to stage the leukemias and lymphomas.2 Staging defines the extent of cancer and facilitates decisions on treatment intensity and duration, response assessment, and prognostication. Involvement of distinct sites and organs can be assessed and substantiated using appropriate staging investigations in lymphomasand solidmalignancies.However, a similar, tangible concept of a stage does not apply to leukemias. In the absence of conventional staging systems, the Toronto Paediatric Cancer Stage Guidelines have adopted the central nervous system (CNS) status to stage acute leukemias.2 The approach conveys merely one aspect of disease burden. It is not a comprehensive methodology for describing the “disease burden” or prognosis of acute leukemia. For a solid tumor, on the other hand, a stage is typically more informative. Overt CNS involvement, although predictive of inferior outcome, is not common in acute lymphoblastic leukemia (ALL).3 In a large Children s Oncology Group study of over 8000 children with ALL, merely 1.5% had CNS3 status.4 CNS involvement is more common in acute myeloid leukemia, albeit with an uncertain impact on the outcome.4,5 The authors of the Australian study mention the paucity of previous literature about stage-specific distribution and survival in pediatric leukemias.1 Contemporary management of leukemia utilizes frontline stratification into “risk groups” rather than “stages.” Risk stratification is based on several factors, including age, white blood cell count, tumor bulk, extramedullary disease, immunophenotype, the presence of genetic aberrations. and early treatment response.3 Therefore, utilization of CNS status as a standalone parameter to classify acute leukemia is suboptimal. It is inadequately communicative of the essence of the stage of a disease. Additionally, the CNS status of leukemia as a single parameter does not have a unique value for cancer surveillance in registry-based studies. From an epidemiological or surveillance perspective, categorizing patients into acute lymphoblastic andmyeloid leukemia and reporting survival outcome is informative. An endeavor to “stage” acute leukemia, akin to lymphoma or solid tumors, seems ineffectual.