Characterization of Vaccine Site and Cellular Infiltrate in Subjects with Chronic Myeloid Leukemia (CML) Receiving K562/GM-CSF Vaccination in Combination with Imatinib Mesylate with or without Topical Imiquimod.

Increased understanding of the local injection site infiltrate in response to tumor vaccines may facilitate more effective anti-cancer vaccination strategies. A pilot vaccination strategy was developed to determine if K562/GM-CSF immunotherapy could enhance T cell reactivity and clinical responses in CML patients undergoing therapy with imatinib mesylate. K562/GM-CSF is a tumor vaccine derived from a CML cell line that expresses several defined CML associated antigens and has been genetically modified to secrete GM-CSF. We undertook a correlative project comparing the cellular infiltrates from pre- and post-vaccination skin biopsies in this immunotherapy trial.

GM-CSF producing tumor vaccines are effective in recruiting antigen presenting cells (APCs), however alone they are insufficient to initiate APC maturation. Because topical imiquimod (a Toll-like receptor 7 agonist) is known to enhance the in vivo maturation of the recruited APCs, the vaccines (1 x 10*8 cells distributed over 10 injection sites) were given with or without topical 5% imiquimod cream. Imiquimod was applied 4 hrs post-vaccination and then 3d and 5d post-vaccination to injection sites, with at least 1 site left without imiquimod treatment. A series of 4 vaccines were administered in 3 wk intervals. Six mm punch biopsies were taken at baseline, and 3d following the 1st and 4th vaccination. Biopsies were performed at the imiquimod site with the largest area of induration, as well as a site not exposed to imiquimod.

Immunohistochemistry of CD3, CD4, CD8, CD1a (Langerhans cell (LC)), factor XIIIa (dermal dendritic cell), and CD68 (monocyte/macrophage) and Geimsa staining was performed. Staining is reported as number of live cells per mm2 in the epidermis and dermis for CD1a+ cells and dermis for the remaining stains. Fifteen subjects agreed to the procedures as part of the clinical trial. Mean area of induration of imiquimod sites was increased significantly compared to the non-imiquimod sites after both the first (p=0.005) and fourth vaccinations (p=0.068). Geimsa staining revealed significant increases in proportion of neutrophils, eosinophils, and mononuclear cells to total number of staining cells after the 1st vaccination in the sites treated with imiquimod compared to the pre-vaccination biopsies while the increases at the sites without imiquimod treatment did not reach statistical significance. We observed increases in CD3+, CD4+, and CD8+ cells at post-vaccination sites. Interestingly, the total number of CD1a+ LCs in the area measured did not appear to be affected by the administration of imiquimod and was constant after vaccinations. However, distribution of CD1a+ LCs shifted from the epidermis to the dermis after vaccination. In addition we observed recruitment of factor XIIIa+ dermal dendritic cells and CD68+ macrophages to the vaccination site that was increased by imiquimod. Epidermal APCs known as LCs migrate to the dermis, yet maintain homeostasis of the total LCs following vaccination independent of Imiquimod application. A correlation across subjects between these histologic features and clinical response to vaccination is on-going.