Multimarker classification of acute lymphoblastic leukemia: evidence for further T subgroups and evaluation of their clinical significance

Abstract

Leukemic blasts from 339 consecutive patients (226 children, 113 adults) with acute leukemia negative for peroxydase and unspecific esterase, were tested in each case for the following surface markers: sheep erythrocyte receptor (E-R), T-cell antigens (HuTLA), surface immunoglobulin (Slg), and “common” non-T, non-B ALL antigen (cALLA). A clear distinction of 6 subgroups of patients with acute lymphoblastic leukemia (ALLA) was possible: (1) common null-cell or cALL (cALLA+) in 119 patients; (2) intermediate-cell or c/T-ALL (cALLA+, HuTLA+) in 87 patients; (3) pre-T-cell or pre-T-ALL (HuTLA+) in 47 patients; (4) T-cell or T-ALL (HuTLA+, E-R+) in 48 patients; (5) B-cell or B-ALL (Slg+) in 6 patients (another phenotype recently defined by Vogler et al. was recognized in some patients of the c-ALL group, but was not analyzed consistently); and (6) pre-B-cell or pre-B-ALL (cALLA+, cytoplasmic lg+). In 32 leukemias no surface marker could be demonstrated. Particular emphasis was put on various T phenotypes. Consecutive typing of ALL thus helped to establish the frequency of two phenotypes which so far were thought to be rather rare: HuTLA occurring together with cALLA (group 2) and HuTLA occurring without E-R (group 3). The measurement of T-antigen amounts by quantitative immunoautoradiography disclosed the existence of phenotypes of non-E-rosetting T leukemias with more T antigen per cell than found on normal or leukemic T lymphocytes expressing the E-R. ALL subdivision was further substantiated by additional markers such as terminal deoxynucleotidyl transferase, Fc receptors for IgG or IgM, C3 receptors on T and/or other lymphoblasts. The disappearance of the stimulatory capacity of allogeneic lymphocytes on the more differentiated T phenotypes was measured in mixed lymphocyte culture (MLC). A strong acid phosphatase reaction and a marked reaction for acid esterase were mostly found in pre-T- and T-ALL, whereas periodic acid Shiff staining prevailed in blasts with cALLA. An analysis of clinical data at presentation revealed significant differences for cell counts, age distribution, and organomegaly for the four main subgroups. The highest remission and lowest relapse rates were observed in the c/T-ALL group by using life-table analysis, whereas the response to treatment of the pre-T- and T-ALL groups was significantly poorer. We conclude that a combined use of a number of tests is essential to reach a precise diagnosis of subclasses of ALL with different biologic and prognostic characteristics.