Relative value of ZAP-70, CD38, and immunoglobulin mutation status in predicting aggressive disease in chronic lymphocytic leukemia

Abstract

Leukemia-cell expression of ZAP-70, CD38, or unmutated immunoglobulin heavy chain variable region genes (U-IGHV) each is associated with aggressive disease in patients with chronic lymphocytic leukemia (CLL). To assess the relative strength of each marker, we defined thresholds for designating a case as positive for CD38 or ZAP-70 in a test cohort of 307 patients and used these data-defined criteria to stratify patients in an independent cohort of 705 patients. Multivariable analysis revealed that ZAP-70 was the strongest risk factor. Knowledge of the IGHV mutation status or CD38 did not improve our ability to predict the time to first treatment except for ZAP-70-negative cases, which could be segregated into 2 groups of intermediate-risk or low-risk disease based on whether they expressed unmutated or mutated IGHV. ZAP-70 maintained its high relative prognostic value for the subset of patients with early-stage, asymptomatic disease, including patients evaluated within 1 year of diagnosis. Although it is premature to recommend therapy based on these risk factors, patients with ZAP-70-positive CLL cells should be monitored closely for disease progression as they have a median time from diagnosis to requiring initial therapy by standard criteria of approximately 3 years.

Figure 1

Relationship between ZAP-70 and CD38 or IGHV mutation status in defining the time from diagnosis to initial therapy. Kaplan-Meier curves depict the proportion of untreated CLL patients in the validation cohort (N = 705) from diagnosis to initial therapy. Panels A through D depict the time from diagnosis to initial treatment of patients who have CLL cells that express ZAP-70 (ZAP-70⁺), whereas panels E through H show the time from diagnosis to first treatment of patients with CLL cells that were ZAP-70 negative (ZAP-70⁻). Panels A and E or B and F show how IGHV mutation status can segregate cases that are CD38⁺ or CD38⁻, respectively. Panels C and G or D and H show how CD38 can segregate cases that use U-IGHV or M-IGHV, respectively. The P values were determined using the log-rank test. The symbols represent the time at which patients were censored.

Figure 2

Relationship between IGHV mutation status and ZAP-70 or CD38 in defining the time from diagnosis to initial therapy. Kaplan-Meier curves depict the proportion of untreated CLL patients in the validation cohort (N = 705) from diagnosis to initial therapy. Panels A through D depict the time from diagnosis to initial treatment of patients who have CLL cells that express M-IGHV genes, whereas panels E through H show the time from diagnosis to first treatment of patients with CLL cells use U-IGHV. Panels A and E or B and F show how CD38 can segregate cases that are ZAP-70–negative (ZAP-70⁻) or ZAP-70–positive (ZAP-70⁺), respectively. Panels C and G or D and H show how ZAP-70 can segregate cases that are CD38⁺ or CD38⁻, respectively. The P values were determined using the log-rank test. The symbols represent the time at which patients were censored.

Figure 3

Relationship between ZAP-70 and IGHV mutation status in defining the time from diagnosis to initial therapy. Kaplan-Meier curves depict the proportion of untreated patients with CLL according to the time since diagnosis. The symbols represent the time at which patients were censored. The patients were divided into 4 subgroups according to their ZAP-70 expression and IGHV mutation status. The legend is provided in each panel for these 4 subgroups, namely cases that were ZAP-70 negative and that used M-IGHV (ZAP-70⁻/M-VH; ■), cases that were ZAP-70–negative but used U-IGHV (ZAP-70⁻/U-VH; ▴), cases that were ZAP-70–positive but used M-IGHV (ZAP-70⁺/M-VH; ▾), and cases that were ZAP-70–positive and used U-IGHV (ZAP-70⁺/U-VH; ♦). The median time from diagnosis to initial therapy for subgroup is listed in parentheses in the legend. The number of patients in each subgroup is provided in the legend. Panel A provides data for the entire validation cohort of 705 patients. Panel B provides data for the 521 patients in the validation cohort who contributed samples for analyses prior to receiving therapy for CLL. Panel C provides data for the 199 patients who presented with early-to-intermediate stage disease within 1 year of diagnosis prior to receiving therapy for CLL. Panel D provides data for the 323 patients who had cytogenetic analyses performed on their leukemia cells at the time of sample collection (Table 1).