Durable Remission of Renal Cell Carcinoma After Donor Lymphocyte Infusion Is Unavoidably Linked with Graft Versus Host Disease As Illustrated by the Detection of Allo Reactive T Cells Recognizing a Novel Minor Histocompatibility Antigen Encoded by the FUCA2 Gene

Abstract 4467

Allogeneic stem cell transplantation (allo SCT) followed by donor lymphocyte infusion (DLI) can be applied as immunotherapeutic intervention to treat malignant diseases. In hematological malignancies, remissions can be observed both in the presence and absence of graft versus host disease (GvHD), illustrating that graft versus leukemia and GvHD can be clinically separated. For efficient induction of an allo immune response, professional (hematopoietic) antigen presenting cells (APC) are required, which may result in skewing towards recognition of minor histocompatibility antigens (MiHA) selectively expressed by hematopoietic cells. However, for treatment of solid tumors including renal cell carcinoma (RCC), recognition of MiHA on non hematopoietic cells is essential. Therefore in these malignancies APC have to present broadly expressed MiHA to be targeted on non hematopoietic cells, making it inevitable that the graft versus tumor (GvT) effect is accompanied by GvHD.

Here we describe the treatment of a patient who presented with progressive metastatic clear cell RCC with T cell depleted non myeloablative allo SCT. At 7 months after allo SCT the patient received DLI at a single dose of 0.5*10⁷ T cells/kg. Acute skin GvHD occurred within 30 days after DLI and developed soon thereafter into persistent extensive chronic GvHD for several years that eventually resolved after prolonged topical and systemic immune suppression. Whereas the patient was mixed chimeric after allo SCT (84%, 80% and 95% donor after 2, 3 and 5 months, respectively), treatment with DLI resulted in conversion to full donor chimerism, which remained complete during the following years. The strong immune response was accompanied by an approximately 50% reduction in size of the measurable lung metastasis (according to RECIST criteria) and slight regression of the smaller lung lesions. To characterize the specificity of the immune response, we measured known MiHA, and detected a disparity for the previously identified MiHA LRH-1. T cells specific for this MiHA were detectable in peripheral blood at the onset of GvHD (0.14% of CD8+ T cells). Isolated LRH-1 specific T cells strongly recognized patient derived EBV-LCL, but were irresponsive to monocyte derived dendritic cells (monoDC), fibroblasts, keratinocytes, and a panel of LRH-1⁺ RCC cell lines and are therefore considered irrelevant for the GvT effect. In addition, we directly cloned activated CD8+ T cells and isolated an HLA-B*07 restricted T cell clone with unknown specificity. Whole genome association scanning unraveled a patient specific single nucleotide polymorphism in the FUCA2 gene encoding a novel MiHA which we designated LB-FUCA2-1V. High frequencies of circulating LB-FUCA2-1V specific T cells were detected after DLI (1.9% of CD8+ cells). Moreover, in contrast to LRH-1 specific T cells, LB-FUCA2-1V specific T cells recognized monoDC, and skin derived keratinocytes and fibroblasts after IFN-γ pretreatment. Our data also show recognition of 4 RCC cell lines, and therefore support a dominant role for LB-FUCA2-1V specific T cells in mediating tumor regression. Unfortunately, coinciding with the gradual resolvement of chronic GvHD, the GvT effect declined and eventually the metastases became progressive almost 3 years after allo SCT. In an attempt to re-initiate the GvT effect, escalating DLI doses of 0.2*10⁷, 0.5*10⁷, 1*10⁷ and 5*10⁷ T cells/kg were given. No GvHD developed but also no GvT effect could be achieved and the patient died of progressive disease 8.5 years after allo SCT.

In conclusion, we observed a strong donor derived immune response targeting the broadly expressed LB-FUCA2-1V MiHA that durably controlled metastatic clear cell RCC. Initiation of this response likely depended on the presence of patient hematopoiesis, including APC, since all attempts to re-induce the MiHA specific response to treat disease progression with escalating doses of DLI under full donor chimerism were ineffective. Our data suggest that successful treatment of solid tumors may require induction of an effective immune response against broadly expressed MiHA, inevitably inducing damage to normal tissue cells resulting in GvHD.