T cells recognizing truly leukemia-specific antigens like bcr/abl or other leukemia-associated antigens such as proteinase3 and WT1 have been demonstrated by several groups. In chronic myeloid leukemia (CML), DC and leukemic cells share a common progeny, leading to constitutive expression of putative tumor antigens such as bcr/abl in DC.

Aim of our clinical GMP study was the use of autologous DC as a vaccine in patients with chronic phase bcr/abl+ CML, who had not achieved a major cytogenetic response after treatment with imatinib or a-interferon. Primary endpoint of the study were feasibility and toxicity of DC vaccination, secondary endpoints were reduction of bcr/abl+ mononuclear cells in peripheral blood (measured by FISH and quantitative PCR) and induction of an immune response against leukemia-associated antigens. 10 patients were enrolled, 9 patients did receive the vaccine. DC were generated by leukapheresis from peripheral blood monocytes. DC were cultured in RPMI 1640 supplemented with clinical grade FCS in the presence of GM-CSF (100 ng/ml) and IL-4 (1000U/ml). Final maturation was achieved by addition of TNF-a (50 ng/ml) for 3 days. Flowcytometric analysis on day 8/9 of the cell culture showed that DC had a mature phenotype with strong expression of CD80, CD86, CD83 and HLA-DR. In all patients sufficient numbers of functional DC could be generated. By FISH analysis it could be demonstrated that the mean value of bcr/abl-positive cells in the DC cultures was 33% (range: 3–70%). DC were pulsed with KLH for 3–4 hours prior to clinical use. Vaccination was performed by subcutaneous injection in the inguinal region on days 1, 2, 8 and 21 using increasing numbers of DC per injection (1 to 50 x 10e6 cells per injection). 36 vaccinations were performed in nine patients. The follow-up period was three months.

There was no severe toxicity, all vaccinated patients developed a strong DTH reaction, mostly after the third vaccination. In 5/9 patients bcr/abl+ PBMC decreased over the vaccination period, in 2/4 patients quantitative PCR also showed a decrease of bcr/abl in peripheral blood. Only in one of these patients this decrease could clearly be attributed to changes in concomitant therapy during the post-vaccination period. In two patients low frequency T cells recognizing the bcr/abl fusion region were detected by ELISpot or tetramer staining. In all patients PBMC showed an increase in the proliferative response against autologous DC over the post-vaccination period, this increase was partially due to immunization against FCS.

We conclude that vaccination of CML patients after large scale GMP generation of bcr/abl+ DC is feasible and safe. Leukemic DC are able to induce T cell activation. A decrease in tumor cell burden/circulating bcr/abl+ cells in some patients suggests that vaccination with DC might possibly be useful as post remission therapy in CML patients after imatinib.

Author notes

Corresponding author