These data were the rationale for launching vaccination with DC-derived exosomes in metastatic tumor-bearing patients. A GMP has been set up to harvest large amounts of MD-DC-derived exosomes (ultrafiltration followed by ultracentrifugation on a high-density sucrose/D2O cushion) and load exosomal MHC class I molecules with synthetic tumor peptides. Phenotypical analysis of clinical-grade exosomes is allowed by FACS-beads-assay, i.e., DC-derived exosomes, are fixed onto 4.5-micron beads stained using surface antigen-recognizing fluorescent antibodies. An immunocapture assay determines the quantity of MHC class I and/or II molecules in each individual exosome batch. A feasibility and safety phase I study was undertaken in France (Institut Gustave Roussy and Institut Curie with Anosys Inc. biotechnological partnership) in HLA-A1/B35 and -DP04 patients bearing stage III/IV melanoma expressing MAGE-3. Dexosomes were purified from the culture supernatant of day 7 autologous MD-DCs. MAGE-3 peptides were loaded onto MD-DCs (in the first six patients), or directly onto dexosomes (in nine patients). Escalating doses of cryopreserved dexosomes were administrated by four weekly sc/id injections, then every 3 wk in patients who achieved stable disease or a tumor regression. Fifteen patients undergoing progressive disease have entered the study (metastatic sites including skin, nodes, lung, and liver). Feasibility of exosome harvesting was about 100% with a yield of exosome production allowing from 6 to 120 vaccine administrations. The vaccine therapy was well tolerated, without any evidence of Grade 2 toxicity. In indirect loading process, one stage III patient achieved stabilization and received eight additional vaccine injections with continual stable disease. Using direct loading but low exosome dosage, a second patient achieved a mixed response with regression of subcutaneous sites but lung progression. A third patient exhibited partial response at node sites. Briefly, three of six patients who received the highest dosage of directly loaded DEX exhibited objective responses (skin and lymph node lesions). It is noteworthy that two patients previously progressing with ALVAC-MAGE3A1 and MAGE3 protein were responding to exosomes. A case of antigen spreading and MHC class I loss variant was documented following exosome administration. The MAGE3-specific CTL responses detectable using specific fluorescent tetramers highlighted monoclonal responses in three patients.
Was this article helpful?