An apparent stem cell population can be identified in many normal and malignant tissues by their side-population (SP) phenotype, characterized by a high level of Hoechst dye efflux. Malignant SP cells have increased resistance to many cytotoxic drugs and may contribute to relapse and disease progression. We have looked for SP cells in patients with B cell chronic lymphocytic leukemia (B-CLL) and determined whether these cells could be targeted by an immune response after immunization with hCD40L/IL2 gene modified tumor vaccines. We also assessed the impact of any anti-SP immune response on overall disease. For SP analysis, PBMC containing >90% CD5+CD19+ cells were obtained pre-treatment. We incubated the cells with 5mg/ml Hoechst 33342 dye for 2 hours at 37°C. Cells were then labeled with CD5, CD19 and propidium iodide and sorted by flow cytometry into SP and non-SP populations. 11 of 12 patient tumor samples contained a distinct CD5+CD19+ SP population (0.04%–2%), whereas SP were undetectable in PBMC from healthy donors. SP cells had increased gene expression compared to non-SP for ABCG2 (4±2-fold) and the “stemness” genes p21cip1/waf1 (8±12-fold) and Bmi-1 (12±21-fold) supporting their function as a precursor population for the bulk of tumor cells. In 5 of the patients, SP analysis was done on samples taken before, during and after receiving tumor vaccines. We observed a significant decrease (p< 0.05) in B-CLL SP cells in all 5, although in 4/5, the decrease was transient, returning to pre-treatment levels by 4 months post-immunization. To determine if the observed reduction in tumor SP was associated with an immune response to these cells, B-CLL-specific cytotoxic T lymphocytes (CTL) were generated from T cells during immunization (2–4 weeks). These CTLs were specific for autologous bulk B-CLL cells (week 2 CTL, 40±4 IFN-gamma spot forming units (SFU) versus unstimulated controls 20±12 SFU, rising to 334±14 SFU versus control 20±5 SFU by week 4, P<0.01). Moreover, CTL lines obtained at 4 weeks after immunization were highly reactive against SP B-CLL cells. To determine the effects of the induced anti-B-CLL SP reactivity on the patients' disease, we followed the 5 patients for >1 year. In the 4 individuals in whom the anti-SP cell response was transient, no sustained effect on disease burden was evident. In one patient, however, the tumor SP cells remained at undetectable levels after immunization. This individual subsequently had a delayed (from 6 months onwards) but progressive disease response, and by 15 months had resolution of lymphadenopathy, a reduction in splenomegaly and a >90% decrease in the circulating B-CLL count. Hence, B-CLL cells contain a primitive side-population, which can be targeted by an immune response after immunization with autologous CD40L and IL2 expressing tumor cells. A sustained anti-B-CLL SP response was associated with a delayed but progressive response in the bulk tumor population. This effect is consistent with “stem cell” activity in the SP of B-CLL, and suggests that tumor SP cells may be an important target for therapy.