Prognostic role of bcl-xL and p53 in childhood acute lymphoblastic leukemia (ALL)

Molecular parameters involved in the prediction of response of childhood acute lymphoblastic leukemia (ALL) are still unclear. We have evaluated the expression and mutational status of p53 and the expression of bcl-xL and bax in a series of 62 consecutive children (median age: 4 years; 38 males and 24 females) affected by de novo ALL. Alterations and overexpression of p53 were uncommon events (9/62, 14.5%) while bcl-xL and bax overexpression were frequent (about 70%). EFS was directly correlated to age < 6 years (p = 0.0178), nonT phenotype (p = 0.0470), WBC at diagnosis ≤ 20000/μl (p = 0.0093), response to induction therapy with prednisone (p = 0.0211) and inversely correlated to mutated p53 or overexpression of p53 (p = 0.0039) and high intensity of Bcl-xL expression (p = 0.0055). OS was directly correlated with age < 6 years (p = 0.0004), female gender (p = 0.0139), nonT phenotype (p = 0.0012), WBC at diagnosis ≤ 20000/μl (p = 0.0187), response to induction therapy with prednisone (p = 0.0211), wild type p53 or low expression of p53 (p = 0.035). When all factors were considered in a stepwise Cox regression analysis, only the good response to PDN (p = 0.013) and the low intensity of Bcl-xL expression (p = 0.001) were independent significant prognostic factors with regard to EFS. Moreover, only age (p = 0.022), gender (p = 0.036) and WBC at the diagnosis (p = 0.050) were independent prognostic factors with regard to OS. Moreover, the mutated status of p53 was statistically correlated to the resistance to the induction therapy with PDN (correlation coefficient: -0.349, p = 0.008). In conclusion, both bcl-x L and bax were frequently expressed at high intensity, but only bcl-xL was an independent predictor of EFS in our series. Moreover, p53 alterations were uncommon and alone not strong independent predictors of outcome, but they were correlated to poor response to therapy with PDN and inversely correlated to EFS and OS in univariate analysis. ©2005 Landes Bioscience.