Clinical features and outcome in childhood T-cell leukemia-lymphoma according to stage of thymocyte differentiation: a Pediatric Oncology Group Study

The immunophenotypes of lymphoblasts from children with newly diagnosed T-cell acute lymphoid leukemia (T-ALL, n = 101) or T-cell non-Hodgkin lymphoma (T-NHL, n = 31) were analyzed to correlate stage of thymocyte differentiation with clinical features and outcome. The 67 boys and 34 girls with T-ALL were 1 month to 18 years old (median, 8 years) with leukocyte counts ranging from 2 to 810 x 10(9)/L (median, 55 x 10(9)/L). Eighteen of these patients were black, and 70 had a mediastinal mass. Twenty-six boys and five girls with a median age of 9 years (range, 1 to 20 years) had T-NHL. Seven of these patients were black, and 24 had a mediastinal mass. The distributions of thymocyte developmental stages (early [CD7+], intermediate [CD1+ and/or CD4+ and/or CD8+], and mature [CD3+]) in cases of T-ALL and T-NHL were significantly different: 34%, 43%, and 23% v 6%, 62%, and 32% (P = .02). A comparison of the patients' clinical features according to the maturational stage of thymocytes failed to disclose significant differences in the majority of characteristics studied. However, patients with mature-stage T-NHL, with or without the addition of subjects with mature-stage T-ALL, were less likely to have a mediastinal mass (P = .02 for both comparisons). Those with intermediate-stage T-cell malignancy (T-ALL and T-NHL combined) were the subgroup most likely to have a mediastinal mass (P = .01). Response to remission induction therapy was significantly worse in the T-ALL subgroup with an early-stage phenotype: a failure rate of 21% v 0% and 6% for the two more differentiated phenotypic subgroups (P = .007). Event-free survival was not affected by thymocyte maturational stage in cases of either T-ALL or T-NHL. Despite evidence of clinical heterogeneity among the maturational stages of T-cell malignancies in children, these developmental subdivisions do not appear to be critical determinants of outcome once remission is achieved. We conclude that such phenotypes need not be included in the stratification plans for clinical trials using common induction treatment.