DECISION-TREE APPROACH TO THE IMMUNOPHENOTYPE-BASED PROGNOSIS OF THE B-CELL CHRONIC LYMPHOCYTIC-LEUKEMIA

Use of a nonlinear prediction method, such as machine learning, is a v aluable choice in predicting progression rate of disease when applied to the highly variable and correlated biological data such as those in patients with chronic lymphocytic leukemia (CL). In this work, decisi on-tree approach to cell phenotype-based prognosis of CLL was adopted. The panel of 33 (32 different phenotypic features and serum concentra tion of sCD23) parameters was simultaneously presented to the C4.5 dec ision tree which extracted the most informative of them and subsequent ly performed classification of CLL patients against the modified Rai s taging system. It has been shown that substantial correlation between the percentage of expression of the CD23 molecule on CD19(+) B-cells, the level of sCD23, the percentage of CD45RA(+), and the absolute numb er of CD4CD45RA(+)RO(+) T-cells and the clinical stages, exists. The p rediction vector, composed of their concatenated values, was able to c orrectly associate 83% of the cases in the low-risk group (Rai stage 0 ), 100% of the cases in the intermediate-risk group (Rai stage I and I I), and 89% of the cases in the high-risk group (Rai stage IB and IV) of CLL patients, Predictivity of this vector was 100%, 95%, and 89%, r espectively, In conclusion, from the described analysis, it may be inf erred that two processes play important roles in the progression rate of CLL: 1, deregulated function of the CD23 gene in a-cells accompanie d by the appearance of its cleaved product sCD23 in the sera; and 2, f unctionally impaired and imbalanced CD4 T-cell subpopulations found in the peripheral blood of CLL patients, Am. J. Hematol. 59:143-148, 199 8, (C) 1998 Wiley-Liss, Inc.