Chronic lymphocytic leukemia (CLL), a disease of clonal CD19+CD5+CD23+ B cells, remains incurable by conventional chemotherapy, despite the use of increasingly efficacious combination chemotherapy regimens. Despite its immunophenotypic homogeneity, this disease is marked by high heterogeneity in the clinical course of patients; patients that are chemorefractory have particularly poor prognosis and early mortality. One promising approach for CLL treatment is through immunotherapy since CLL is highly immune-responsive, as demonstrated by its susceptibility to the graft-versus-leukemia (GvL) effect after hematopoietic stem cell transplantation (HSCT). A key question is whether we can develop simpler and safer tumor-specific immunotherapy that can enhance GvL or even replace HSCT.
Our previous studies have focused on dissecting mechanisms of effective tumor immunity in patients who have demonstrated long-lasting clinically evident immunity following HSCT. Together, those studies indicate that productive anti-tumor immune responses are long-lived, directed against multiple epitopes, elicit both B and T cell responses, and in general, are specific for an individual patient. We are now completing an analysis of the activation state of leukemia-infiltrating T cells before and after allotransplantation for CML and CLL, which appear to be predictive of clinical response. Altogether, these insights have led me to undertake three general approaches to devise immunotherapy for CLL patients that is personalized and directed against antigens that are truly specific to that individual tumor. (1) Whole tumor vaccines to augment the GvL effect: I am leading a high-priority clinical trial at DFCI to examine the safety and immunologic effects of vaccination with whole tumor cells (using irradiated autologous leukemia cells) and the cytokine adjuvant GM-CSF in post-transplant patients. Vaccinated subjects have developed CD8+ T cell responses to CLL-specific antigens, independent of allo-immunity, implying that personalized whole tumor cell vaccination can enhance anti-tumor immunity with minimal toxicity. Our completed and ongoing studies of this vaccine are laying the groundwork for using the post-transplant setting as a platform for developing tumor vaccines. (2) Expansion of tumor-reactive donor T cells as adoptive T cell immunotherapy. (3) Personal neoantigens at tumor-specific immune targets for vaccination: A high priority in cancer immunotherapy is the identification of tumor-specific antigens that would allow effective tumor targeting without collateral toxicity. We are pioneering the use of exome and transcriptome sequencing technologies to identify unique mutated CLL antigens that arise from individual-specific genetic alterations within a tumor, thus paving the way for developing personalized tumor vaccines. We have therefore undertaken large-scale genome analysis of CLL to systematically identify tumor-specific genetic alterations. Through integration of sequencing data of 91 CLL samples with clinical annotation, defining 9 key mutated genes and 6 pathways involved in CLL pathogenesis, we have discovered novel associations with prognostic factors. Our ongoing studies in this area focus on: (a) systematic analyses to couple tumor genotype with phenotype; (b) understanding CLL tumor heterogeneity; (c) mechanistic studies of specific tumors drivers, such as SF3B1.