External Validation On Biological Basis of New Prognostic Index in Early Asymptomatic Chronic Lymphocytic Leukemia (CLL) Patients: The Gruppo Italiano Studio Linfomi (GISL) Experience.

Abstract 2375

Poster Board II-352

A prognostic index based on widely available clinical and laboratory features was recently proposed to predict survival in patients with previously untreated patients with chronic lymphocytic leukemia (CLL) by MD Anderson investigators. However, whether proposed clinical risk categories may surrogate new biological variables of prognostic relevance (i.e., mutational status of the IgVH gene regions, ZAP-70 or CD-38 expression, cytogenetic abnormalities) is unclear thus far.

In a series of 160 asymptomatic Binet stage A patients enrolled in a Gruppo Italiano Studio Linfomi (GISL) multicentre trial designed to validate prospectively biological parameters in early CLL as well as to assess the impact on clinical outcome of an early versus delayed policy of treatment with subcutaneous alemtuzumab in the high biological risk, we evaluated whether clinical categories derived from newly proposed prognostic index reflected biological risk.

Since the original prognostic index was derived from a database including cases with more advanced disease we used an optimal cutoff search to determine how to best split Binet stage A patients in different prognostic groups. To this purpose an independent patient cohort consisting of 310 Binet stage A patients included in a GIMEMA (Gruppo Italiano Malattie EMatologiche Maligne dell'Adulto) database was used. According to recursive partitioning (RPART) model, a classification tree was built that identified two subsets of patients who scored respectively: 0-3 (low risk) and 4-7 (high risk). Therefore, by prognostic index, 48.7% and 51.2% of 160 asymptomatic stage A patients, respectively, met criteria of low risk and high risk disease.

In our prospective series high- risk score was more frequently associated with both unmutated IgVH status (P=0.009) and higher CD38-expression (P=0.002); in contrast only a trend towards an increased ZAP-70 expression could be found (P=0.06). As far as cytogenetic abnormalities are concerned, we observed that 11q deletion occurred more frequently among patients belonging to high-risk score (P=0.005), while cases with 13q deletion or trisomy 12 were homogeneously distributed among low- and high-risk patient category(P=0.151 and P=0.452, respectively). We did not consider suitable for correlation analysis 17p deletion since observed only in 2 out of 160 Binet stage A patients.

In conclusion, our results demonstrate in a prospective cohort of patients with early CLL that clinical categories of a revised score index may surrogate biological parameters of prognostic relevance. The observation reinforces the revised IWCLL guidelines recommendations to assess the risk of CLL patients on clinical basis and to deserve biological studies to patients eligible for clinical trials.