Abstract 918

Introduction:

Various factors are used to predict the clinical course of chronic lymphocytic leukemia (CLL). Here we evaluated the importance of the serum markers soluble CD23 (sCD23), β2-microglobulin (s-β2m) and thymidine kinase (s-TK) compared to other prognostic markers in patients (pts) receiving a first-line therapy with fludarabine and cyclophosphamide (FC) or FC plus rituximab (FCR) in the CLL8 trial of the GCLLSG.

Methods:

The analysis of the serum factors was centrally performed prior to treatment. Kaplan-Meier method and the log-rank test were used to compare progression-free survival (PFS) and overall survival (OS). A Cox regression model was applied for the multivariate analysis and included the following parameters: age, sex, disease stage, time from first diagnosis, ECOG performance status, presence of B-symptoms, white blood cell count, sCD23, s-TK and s-β2m genomic aberrations, IGHV mutational status, and type of therapy (FC versus FCR). Cut-off values for s-β2m and for s-TK were 3.5 mg/l and 10 U/l, respectively.

Results:

Since exploratory analyses suggested a particular role for sCD23, the contribution of sCD23 was examined in more detail. Data for sCD23 was available for 616 pts. This cohort was representative of the full trial population (n=817) with respect to baseline characteristics. Results for response rates, PFS, OS and sCD23 were available for 574 pts. Based on their sCD23 levels, pts were categorized by quartiles: 139 pts were assigned to group A (≤2807.5 U/mL), 145 to group B (>2807.5 U/mL -≤5507 U/mL), 146 to group C (>5507 U/mL-≤10920.5 U/mL) and 144 to group D (>10920.5 U/mL). Significant differences were observed with respect to complete remissions (CR): Pts of group A achieved a double fold CR rate (48.2%) compared to pts of group D (23.6%; p<0.001). Correspondingly, group D had the highest rate of non-responses (11.8%). PFS and OS were longest for pts with low serum levels for sCD23: The median PFS in group A was 51.9 months (mo) compared to 29.8 mo in group D (p<0.001). The median OS for group D was 62.5 mo, but was not reached in the other groups.

Next multivariate analyses (Table1) were performed and showed that s-TK, 17p-deletion (del(17p)) and the type of therapy were independent predictors of PFS and OS. del(17p) was the strongest adverse factor for both PFS and OS. sCD23 and IGHV were independent prognostic factors for PFS but not for OS. s-ß2m, age and ECOG status independently predicted OS.

Finally, we investigated whether combinations of different serum markers would define distinct risk groups. Using the three serum markers, four risk categories were defined: Low risk (LR) = no serum factor elevation, intermediate risk (IR) = either s-ß2m or s-TK high, but sCD23 low, high risk (HR) = high s-ß2m and s-TK, but low sCD23, and very high risk (VHR) = high sCD23, regardless of s-ß2m and s-TK. Median PFS was significantly different for the four groups, with 59.7, 49.3, 35 and 29.8 months for LR, IR, HR, and VHR patients respectively (p<0.001). However, the difference between HR and VHR was not statistically significant (p=0.2).

Conclusion:

Elevated serum levels of s-ß2m, s-TK, and sCD23 predict poor outcome after first-line therapy with FC or FCR in pts with CLL, independently of other features such as IGHV status or del(17p). Measurement of sCD23 in addition to s-TK and s-ß2m seems to allow a more precise prediction of PFS.

Table 1

Multivariate analysis of the effects of various prognostic parameters on PFS and OS

PFS n=504OS n=515
Parameter Hazard ratio 95% CI p-value Hazard ratio 95% CI p value 
sCD23 (>10920.5 U/mL) 1.340 1.020–1.761 0.036 – – – 
s-TK ≥10.0 U/L 1.454 1.052–2.009 0.023 2.007 1.085–3,711 0.026 
s-β2m ≥ 3.5mg/L – – – 1.726 1.123–2.652 0.013 
del(17p) 6.430 4.154–9.952 <0.001 8.660 4.841–15.490 <0.001 
IGHV unmutated 1.411 1.035–1.922 0.029 – – – 
Age – – – 0.646 0.419–0.995 0.047 
ECOG pre ≥ 1 – – – 1.868 1.239–2.816 0.003 
FCR treatment 0.506 0.393–0.651 <0.001 0.671 0.447–1.008 0.05 
PFS n=504OS n=515
Parameter Hazard ratio 95% CI p-value Hazard ratio 95% CI p value 
sCD23 (>10920.5 U/mL) 1.340 1.020–1.761 0.036 – – – 
s-TK ≥10.0 U/L 1.454 1.052–2.009 0.023 2.007 1.085–3,711 0.026 
s-β2m ≥ 3.5mg/L – – – 1.726 1.123–2.652 0.013 
del(17p) 6.430 4.154–9.952 <0.001 8.660 4.841–15.490 <0.001 
IGHV unmutated 1.411 1.035–1.922 0.029 – – – 
Age – – – 0.646 0.419–0.995 0.047 
ECOG pre ≥ 1 – – – 1.868 1.239–2.816 0.003 
FCR treatment 0.506 0.393–0.651 <0.001 0.671 0.447–1.008 0.05 
View Large
Disclosures:

Fink:Roche: . Pflug:Roche: Travel grants. Eichhorst:Roche: Research Funding, Travel grants. Wendtner:Roche: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding, Travel grants. Mendila:Roche: Employment. Wenger:Roche: Employment. Stilgenbauer:Amgen: Research Funding; Bayer: Consultancy, Honoraria, Research Funding; Boehringer-Ingelheim: Consultancy, Honoraria, Research Funding; Celgene: Consultancy, Honoraria, Research Funding; Genzyme: Consultancy, Honoraria, Research Funding; GSK: Consultancy, Honoraria, Research Funding; Mundipharma: Consultancy, Honoraria, Research Funding; Roche: Consultancy, Honoraria, Research Funding; Sanofi Aventis: Research Funding. Hallek:Roche: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding. Fischer:Roche: Research Funding, Travel grants.

Topics:
chronic b-cell leukemias, chronic lymphocytic leukemia, treatment outcome, brachial plexus neuritis, electrocorticogram, serum markers, cd23 antigen, cyclophosphamide, disease remission, fludarabine

Author notes

*

Asterisk with author names denotes non-ASH members.

© 2010 by The American Society of Hematology
2010
Sign in via your Institution