Immune Phenotyping and Naive T Cells as a Predictor of Response to Therapy In Chronic Lymphocytic Leukemia (CLL)

Abstract 1362

The role of anti-tumor immunity is not well understood in CLL but does have the potential to be manipulated to improve treatment responses. We studied the predictive values of 43 immune phenotypes in patients enrolled in a clinical trial of early treatment of high risk CLL. The study enrolled 33 patients with CLL diagnosed using standard NCI-WG96 criteria who had previously untreated early - intermediate stage disease and did not meet NCI-WG96 guidelines for conventional treatment. High risk disease was defined as at least one of the following: 17p13 deletion; 11q22 deletion or unmutated IGHV together with either expression of ZAP-70 (>20%) or CD38 (>30%). Patients were treated with GM-CSF 250 mcg sc 3 × week × 5 weeks, alemtuzumab sc with dose escalation over 3 days then 30 mg/d 3 × week × 4 weeks, and rituximab 375 mg/m² weekly × 4 weeks for a total duration of therapy of 5 weeks. We measured immune phenotypes (representing Tregs, naïve, effector memory, central memory T cells, B cells, NK cells, monocytes, among others) at the time of enrollment, through treatment and at 1 year after enrollment in 28 patients. We identified pre-treatment immune phenotypes that correlated with disease burden (as measured by the number of circulating B cells). Increased disease burden was correlated with increased absolute numbers of CD3+, CD8+, CD4+ T cells, NK (CD56+CD16+), Dr+CD8+ T cells, and Tregs (CD4+CD25+CD127+); p<0.001. The disease burden was inversely related to the number of circulating monocytes (p<0.01) and unrelated to the circulating number of naïve T cells. We repeated the pre-treatment phenotype correlation studies using the CD19+ cell counts 12 months after treatment for 13 patients who have completed the final analysis. Interestingly, the number of naïve CD4 and CD8 cells (CD4+RA+62L+CD27+ and CD8+RA+62L+CD27+) of the patient prior to treatment correlated with CD19+ cells 12 months after treatment (p=0.024 and p=0.002 respectively; n=13). Finally, we correlated pre-treatment immune phenotypes with time to progression. Again, an elevated T cell phenotype (CD4+RA+62L+CD27- as a percent of CD4+ cells) correlated to increased time to progression (p=0.04; n=28). It is noteworthy that the naïve T cell phenotype did not correlate with tumor burden prior to therapy. Thus, while pre-treatment lymphocyte expansion is commensurate with tumor burden in CLL, the frequency and numbers of naïve T cells prior to treatment has some predictive value to response to treatment. These data suggest that CLL induces the expansion of lymphocytes, suppresses monocytes with an induction in naïve T cells acting as an indicator of higher B cell counts at 12 months in patients with durable remissions.