Prognostic factors in childhood acute lymphoblastic leukemia.

Abstract

Treatment efficacy alters the impact of most prognostic factors. Among clinical features, only age and leukocyte count remain prognostically important. Immunophenotyping is useful for ALL classification and for assignment to specific therapy regimens, but, with the possible exception of CD10 expression, has little prognostic importance in the context of contemporary phenotype- and risk-directed therapy. Cytogenetic features are useful for risk assignment. Hyperdiploidy > 50 chromosomes is associated with a favorable prognosis, whereas Ph+ chromosome and t(4;11) confer an adverse prognosis. Pre-B cases with the t(1;19) do not fare well with antimetabolite-based therapy and should be treated with additional classes of chemotherapeutic agents. Finally, certain specific rearrangements such as dic(9;12) may in fact be associated with favorable prognosis. With the exception of treatment for B-cell ALL (and perhaps transitional pre-B ALL), phenotype- or genotype-directed therapies have not been successfully devised. Selection of treatment for individual patients, therefore, should be based on their estimated risk of failure. For patients with very high-risk leukemia (i.e., those with > 70% likelihood of treatment failure), the use of experimental therapeutic strategies, despite the potential for acute and long-term disabilities, may be justified. For the subset of children in the lower-risk category (< 20% probability of failure), antimetabolite-based therapy should be employed to minimize long-term sequelae.