Abstract
BACKGROUND: MAGE-A3 is an immunogenic tumor-associated antigen detected in 1/3 of newly diagnosed MM patients, and confers a poor prognosis, making it a rational target for immunotherapy. We previously reported (Cohen et al, ASH 2013, #154) that pre- and post-ASCT administration of recMAGE-A3 + AS15 adjuvant (containing MPL, QS21, and CpG7909) and infusion of vaccine-primed autologous peripheral blood lymphocytes (PBL) in the early post-ASCT period had an acceptable safety profile and induced robust antibody responses against MAGE-A3. We now report our initial cellular immune response data, and update the humoral response and clinical outcome data.
METHODS: The composition of recMAGE-A3 +AS15 and the immunization schedule (Fig. 1) for this pilot study have been described (ASH 2013, #154). Antibody responses were assessed by ELISA. CD4 and CD8 T cell responses were assessed by ELISpot and intracellular cytokine release assays after in vitro re-stimulation with MAGE-A3 overlapping peptide pools or controls and autologous antigen-presenting cells. Clinical responses were determined by IMWG criteria. M3H67 mAb (specific for MAGE-A3 and homologous MAGE-A family members) was used for assessing expression in MM cells by immunohistochemistry (IHC).
RESULTS: Thirteen patients enrolled (med. age 56; 45% high-risk cytogenetics; 42% ISS II/III). All had MAGE-A+ myeloma cells and had achieved at least VGPR following induction. Twelve of 12 (100%) subjects tested to date developed high-titer (1:104-106) antibodies against MAGE-A3 that persisted to at least 1 year post-SCT. These titers were 10-100-fold higher than those seen in a prior study in lung cancer patients with recMAGE-A3 + AS02b, an older adjuvant lacking CpG7909 (PNAS 2008; 105:1650). Epitope mapping identified at least 7 distinct MAGE-A3 epitopes clustering in the hydrophobic regions from aa. 1-100 and 220-300. Isotyping and IgG subclass analysis demonstrated IgG class switching in all patients, with IgG1 and IgG3 subclasses most prevalent. Peripheral blood T cell responses have been evaluated in 3 subjects to date. All had MAGE-A3-specific CD4 responses by IFNγ ELISpot starting as early as d+31 after ASCT, with significant expansion after booster vaccinations and persistence through 1 year post-ASCT. Intracellular cytokine staining confirmed a polyfunctional, Th1-biased CD4 T cell response (IFNγ+, TNFα+, IL5-) in all 3 patients. No CD8 responses against MAGE-A3 have been detected to date. Clinical response assessments were as follows: there were 12 VGPR and 1 CR at enrollment, 7 VGPR and 6 CR (3 stringent CR) at 3 months (mos.) post-ASCT, and 3 VGPR and 5 CR (4 sCR) at 1 year post-ASCT, with 4 patients relapsing at or before 1 year, and 1 not yet evaluable. With a median follow-up of 19 mos. (range 6-32), 6 patients have relapsed (estimated median PFS is 24 mos.) and 1 died of progressive MM. There was no difference between progressors and non-progressors with regard to cytogenetics, baseline MAGE-A expression, antibody titers, hematologic response, or use of lenalidomide maintenance (n=4). MAGE-A expression was assessed by IHC in 3 relapse bone marrow biopsies, and all were negative.
CONCLUSIONS: RecMAGE-A3 immunotherapy and PBL reconstitution is well-tolerated, feasible, and induces antibody and Th1-biased CD4 T cell responses, but not CD8 responses, in the setting of ASCT for MM. Cellular immune assessments are ongoing. The magnitudes of antibody and CD4 responses appear greater than those seen historically with older formulations of recMAGE-A3 in other cancers, despite significant immune compromise after ASCT, suggesting a benefit from the new AS15 adjuvant formulation, or from immunization and autologous PBL transfer in the peri-ASCT setting, or both. The loss of MAGE-A3 expression in relapsing patients implies antigen-specific immune selective pressure even in the absence of CD8 T cell responses, and also suggests that combination strategies aimed at limiting immune escape (eg multi-antigen vaccines) should be investigated. Clinical outcomes are promising for this high-risk patient population. These results support advanced phase clinical trials to investigate clinical efficacy of recMAGE-A3 vaccine immunotherapy in MM.
Figure 1 Figure 1.
Disclosures
Cohen: Onyx Pharmaceuticals: Advisory Board, Advisory Board Other; Bristol-Myers Squibb: Advisory Board, Advisory Board Other, Research Funding; Janssen: Advisory Board, Advisory Board Other; Celgene: Member, Independent Response Adjudication Committee Other. Bertolini:Ludwig Institute for Cancer Research: Employment. Pan:Ludwig Institute for Cancer Research: Employment. Venhaus:Ludwig Institute for Cancer Research: Employment. Fellague-Chebra:GlaxoSmithKline: Employment. Gruselle:GlaxoSmithKline: Employment.
Marburg virus survivor immune responses are Th1 skewed with limited neutralizing antibody responses