Identification of leukemic cells in the cerebrospinal fluid from children with acute lymphoblastic leukemia: advances and dilemmas.

As more children survive acute lymphoblastic leukemia, new challenges face the pediatric oncologist. One is the need for early recognition of subclinical nonsymptomatic disease, or occult meningeal leukemia. By obtaining spinal fluid specimens, one can frequently monitor the central nervous system (CNS) for the presence of residual or recurrent leukemia. Methods for identifying small numbers of lymphoblasts in the spinal fluid have progressed from subjective descriptions of nucleated cells to assignment of objective methods for labeling the cells in question. These more sensitive methods of detection have challenged the "old" definitions of remission. The unequivocal identification of lymphoblasts in the spinal fluid of a child with less than 5 cells/mm3 of fluid is an example of such a dilemma. Does this finding herald overt CNS relapse or hematologic relapse? Does it warrant further, more aggressive treatment beyond giving standard prophylactic CNS therapy? It can be suggested that the finding of even a small number or proportion of leukemic cells in the spinal fluid in a child being given continuation/maintenance therapy has a high probability of being followed by an adverse event (relapse). If this is true, then the definition of CNS leukemia needs refinement and the methods for unequivocally identifying leukemic cells in a spinal fluid sample depend on techniques of cytologic examination beyond routine cytomorphology. Such laboratory methods are now available, but they require more clinical testing and standardization before their sensitivity and specificity in identifying leukemic cells is established.