Clinical outcomes in chronic lymphocytic leukaemia associated with expression of CD5, a negative regulator of B-cell receptor signalling

Abstract

Chronic lymphocytic leukaemia (CLL) is characterized by expression of CD5 on clonal B cells, and is partly driven by activated B-cell receptor (BCR) signalling. While CD5 is known to be a negative regulator of BCR signalling, it is unknown if variability in CD5 expression exists among patients and whether CLL cell CD5 expression affects CLL clinical outcomes. We assessed the extent to which CD5 expression is correlated with clinical outcomes, and whether this information adds to currently used prognostic markers. We evaluated CD5 expression from 1275 blood samples, established prognostic markers and time to event data from 423 CLL patients followed at the Duke University and Durham VA Medical Centers. CD5 median fluorescence intensity (MFI) was largely stable over time in individual patients, but ranged between 0·5 and 760 in the entire cohort. Lower CD5 MFI was significantly associated with a shorter time to first therapy. CD5 MFI, combined with established clinical and molecular prognostic markers, significantly improved risk-stratification. CD5 may affect disease outcomes by suppressing signalling through the BCR. Thus, a strategy to modulate CLL cell CD5 expression or function could be a therapeutic approach in CLL.

Keywords: B-cell receptor signalling; CD5; biomarker; chronic lymphocytic leukaemia; prognosis.

Figure 1:

A. Range of CD5 expression (MFI) displayed in CLL patients from the Duke University and Durham VA Medical Center cohort, ordered from lowest CD5 MFI to highest.B. Representative flow cytometry plots demonstrating range of CD5 expression on CLL cells.CLL: chronic lymphocytic leukaemia; FITC: fluorescein isothiocyanate; MFI: median fluorescence intensity; PE: phycoerythrin.

Figure 2:

A. Longitudinal CD5 MFI values for each patient that had more than six samples evaluated (39 CLL patients, 314 total samples). Data displayed as fold-change from the initial CD5 MFI value, connected by lines for each patient.B: Longitudinal CD5 MFI values for each patient that had more than three samples evaluated, from before and after therapy (5 CLL patients, 31 total samples). Data displayed as time from therapy in days and fold-change from the initial CD5 MFI value, connected by lines for each patient. CLL: chronic lymphocytic leukaemia; MFI: median fluorescence intensity.

Figure 3:

A. Higher CD5 median fluorescence intensity (MFI), dichotomized at the median value, is significantly associated with longer time to therapy (p = 0.008). The high CD5 curve had 201 patients and 84 events, and the low CD5 curve had 202 patients and 101 events.B. Higher CD5 MFI, dichotomized at the median value, is associated with superior overall survival (p = 0.068). The high CD5 curve had 201 patients and 20 events, and the low CD5 curve had 202 patients and 29 events.

Figure 4:

Forest plot of hazard ratios (HR; black squares) and confidence intervals (horizontal lines) of the association between CD5 MFI (continuous variable) and time to first treatment (TTT) in molecularly defined prognostic groups. HR less than 1 indicates higher CD5 MFI is associated with longer TTT, whereas HR above 1 indicates higher CD5 MFI is associated with shorter TTT. CD5 and CD38 have a significant P_interaction of 0.021, but all other interactions were not significant. FISH: fluorescence in situ hybridization; M: mutated; MFI: median fluorescence intensity; UM: unmutated.