Background. Cytogenetic analysis is the most important diagnostic tool for determining prognosis in acute myeloid leukemia (AML) (1). However, chromosomal aberrations can be found in only 50% of patients (pts) and the need for additional apt predictors of prognosis is obvious. No single antigen expression of leukemic cells has been proven to reliably predict prognosis in AML. In this study pts with AML were grouped according to antigen expression and prognostic significance of this classification was investigated.

Material and methods. Using multiparameter flow cytometry, we determined the expression patterns of four antigens: CD15, CD33, CD34 and HLA-DR in leukemic blasts from 129 previously untreated consecutive pts with AML diagnosed 1994–2001. Patients receiving up-front palliative treatment and M3 were excluded. The median age was 64 yrs (range 19–85). Median follow-up in 33 surviving pts was 44 m (range 20–105). The Cox proportional hazard model was used in survival analysis.

Results. Based on the expression of CD15 and CD33 five distinct groups were identified: [I: CD33−/CD15− (n=18), II: CD33+/CD15− (n=43), III: CD33+/CD15 heterogeneous (n=10), IV: CD33+/CD15+ (n=50), V: CD33−/CD15+ (n=8)]. In groups II and IV subgroups A and B were formed according to the expression of CD34 and HLA-DR. CR was achieved in 88 pts (69%) and 51 pts (58%) eventually relapsed. The median overall survival (OS) for all pts and complete remitters was 15 and 32 months, respectively. Cytogenetic results were available in 112 pts. Only 3 pts (2%) had favorable cytogenetic changes (all in group IV). 17 pts (13%) had unfavorable chromosomal changes, they were more frequent in groups IIA and IVA.

Group I and III pts (median age 63 in both groups) had a median OS of 28 and 29 months, respectively. The shortest median OS (8 months) was observed in pts in group II (median age 73 years). Pts <60 yrs in this group also had a short OS (9 months). Group IV pts with only small subsets of blasts with CD34 and HLA-DR expression (group IVA, n=33, median age 64) had a short OS (16 m) probably due to a high proportion of unfavorable chromosome abnormalities. Patients with pattern V had relatively short survival (14 months) despite a median age of 50 yrs and high CR rate (table 1). In univariate survival analysis age (p<0.0001), cytogenetics (p=0.005), and our immunophenotypic classification (p<0.0001) were found to be significant for OS. In multivariate analysis these factors remained independent of each other. The prognostic significance of our immunophenotypic classification was retained (p=0.0235) when patients younger than 60 yrs were analyzed separately (table 2). Analysis of single antigen expression of CD2, CD4, CD7, and CD34 did not have any prognostic significance.

Conclusion. Immunophenotype patterns using CD15/CD33 and CD34/HLA-DR antigen expression define distinct clinical subgroups of AML and seem to add useful prognostic information.

Table 1. All pts

IIIIIIIVVall
18 43 10 50 129 
age(yrs) 63 73 63 62 50 64 
unfav chrom 10 17 
OS(months) 28 29 21 14 15 
IIIIIIIVVall
18 43 10 50 129 
age(yrs) 63 73 63 62 50 64 
unfav chrom 10 17 
OS(months) 28 29 21 14 15 

Table 2. Patients <60 yrs

IIIIIIIVVall
11 23 51 
age(yrs) 53 54 49 50 45 50 
unfav chrom 
 34 58 32 19 36 
IIIIIIIVVall
11 23 51 
age(yrs) 53 54 49 50 45 50 
unfav chrom 
 34 58 32 19 36 

1.
Grimwade et al. The importance of diagnostic cytogenetics on outcome in AML.
Blood
92
:
2322
–33,
1998

Author notes

Corresponding author

Sign in via your Institution