A. Venditti et al., TRIPLE IMMUNOFLUORESCENCE EVALUATION OF CD15, CD34 AND CLASS-II EXPRESSION BY FLOW-CYTOMETRY IN NORMAL AND LEUKEMIC BONE MARROWS, Haematologica, 78(6), 1993, pp. 359-363
Background. The CD15/CD34 phenotype has been reported as aberrant and
exploited for monitoring minimal residual disease in acute myeIoid leu
kemia (AML) patients. Moreover, CD15/CD34 has been described as a rare
phenotype in normal bone marrow (NBM) (<0.10%). We used a triple immu
nofluorescence assay to investigate the expression of CD15, CD34 and c
lass II antigens in normal (NMB) and leukemic (LBM) bone marrows. Meth
ods. A FACScan for quantitative fluorescence and the PAINT-A-GATE prog
ram for multiparametric analysis were utilized. Fifteen normal bone ma
rrow and fifteen leukemic bone marrow samples were studied with a trip
le immunofluorescence assay. A FACSorter was used for sorting. Results
. Eleven out of 15 normal marrows contained less than 0.67% (range 0.0
1-0.66) cells which coexpressed CD15, CD34 and classe II antigens. Thr
ee normal marrows contained more than 0.67% but less than 1.84% (range
1.05-1.83) triple stained cells. The entire group of leukemic marrows
coexpressed triple stained cells with a frequency higher than 1% (ran
ge 1.1-48.6); furthermore, 10 samples contained the same population at
frequencies higher than 10%. The difference between normal and leukem
ic marrows was statistically significant (p = <0.001). Triple positive
cells (TPc) from NBM were sorted for morphology, which proved to be c
onsistent with myeloid progenitor features (early myeloblasts). This c
onfirms that CDl5(+) CD34(+) Class II+ precursors are commonly express
ed in NMB, although at low frequency. Interestingly, 12 (80%) AMLs out
of 15 were diagnosed as M1 (5) or M2 (7), while the remaining were M4
(2) and M5a (1). Additionally, all M2 cases were positive at percenta
ges higher than 10%. Apparently CD15, CD34, class II expression correl
ates mainly with granulocyte differentiation. Two complete remission (
CR) LBM, positive at onset for the triple combination, were regularly
monitored. In one case the TPc percentage always remained near the nor
mal values found in NBM (0.72%), and this patient is still in CR. In t
he other, overt relapse was preceded by a progressive increase in the
TPc percentage. Conclusions. Although the presence in NBM of CD15(+) C
D34(+) and class II+ precursors hinders minimal residual disease detec
tion, we conclude that this unusual combination may distinguish a leuk
emic population and may allow monitoring of <<early relapse>>.