Evaluating Cyclin A1 As an Immunotherapeutic Target in Acute Myeloid Leukemia (AML)

T cell based immunotherapy targeting leukemia associated antigens (LAAs) such as Wilm's tumor 1 antigen (WT1) has proven both safe and clinically effective, with the induction of durable responses in patients with relapsed AML. However, because of heterogeneous LAA expression among leukemic blasts, targeting a single LAA has led to tumor immune escape and relapse, highlighting the need to identify and simultaneously target an array of tumor-expressed LAAs. To expand the target antigen spectrum we sought to immunologically characterize cyclin A1 (CCNA1), a meiosis-specific cell cycle antigen that is selectively expressed in AML blasts with limited normal tissue expression and which has been implicated as playing a role in leukemogenesis based on preclinical animal models.

To assess the suitability of CCNA1 as an immunotherapeutic target we first sought to expand reactive populations ex vivo using an overlapping peptide library (15mers overlapping by 11 amino acids) spanning the entire antigen as a source of stimulation. In 12 donors tested we achieved a mean 5.2±0.65 fold expansion of T cells, of which 11 proved to recognize CCN1 as measured by an IFNg ELISpot [mean 377.3 spot forming cells (SFCs)/1x10⁵ T cells (range 38 to 1188 SFCs) ]. The antigen specific lines were polyclonal (23±4% CD4+ and 64±5% CD8+ T cells) that expressed markers for both central and effector memory (29±7% CD45RO+CD62L+ and 28±6% CD45RO+CD62L-), required for long term in vivo persistence. Importantly both helper (CD4+) and cytotoxic (CD8+) subsets contained CCNA1-directed T cell populations, which secreted multiple cytokines (1.47±0.07% CD4+IFNg+TNFa+ and 9.52±0.25% CD8+IFNg+TNFa+) as measured by intracellular cytokine staining, confirming their polyfunctionality. Moreover, CCNA1-specific T cells were able to selectively kill pepmix-loaded autologous target cells and a partially HLA-matched cell line that endogenously-expresses cyclin A1 (44±7% and 55±1%, respectively; E:T 20:1). without non-specific activity against control (unloaded) targets (9±2%; E:T 20:1). This killing was confirmed to be HLA-restricted since the addition of an MHC class I blocking antibody diminished specific lysis (decrease from 44% to 26%). Finally, using individual peptides organized into minipools, we were able to identify 4 novel HLA class I-restricted epitope peptides.

To assess the clinical relevance of CCNA1-directed T cells we assessed whether patients with AML had detectable circulating T cells and in 5 of 9 patients who had undergone an allogeneic hematopoietic stem cell transplant as disease treatment we could detect CCNA1-specific T cells (mean 117±52.8 SFC/2x10⁵ expanded PBMCs), and detection of reactive cells correlated with a lower risk for relapse.

In summary, we have identified CCNA1 as a suitable candidate antigen to target using immunotherapeutic approaches for future clinical trials of adoptively transferred T cells.