Somatic Gene Therapy for X-Linked Severe Combined Immunodeficiency Using a Self-Inactivating Modified Gammaretroviral Vector Results in An Improved Preclinical Safety Profile and Early Clinical Efficacy in a Human Patient

Abstract 164^FN2

Somatic gene therapy for X-linked severe combined immunodeficiency (X-SCID) using a MLV-based gammaretroviral vector expressing the IL-2 receptor gamma chain (MFG-γc) resulted in excellent immunologic reconstitution but also in insertional oncogenesis. In 5/20 treated children, T cell leukemia developed, with insertional activation of LMO2 proto-oncogene in 4 of the 5. We reasoned that replacing the virus-derived promoter and enhancer elements with a weaker cellular promoter would result in improved safety yet retain efficacy. We therefore generated the pSRS11.EFS.IL2RG.pre* self-inactivating (SIN) gammaretroviral vector in which expression of γc is controlled by an intronless EF1-α promoter, in a MLV vector devoid of the LTR U3 (enhancer/promoter) region. In preclinical studies we determined ‘relative safety' using several surrogate assays. In a reporter assay, the pSRS11.EFS.IL2RG.pre* vector when inserted into the oncogenic LMO2 locus induced LMO2 expression 6–90-fold less than the parent MFG vector (MFG-γc). pSRS11.EFS.IL2RG.pre* had lower activity in a murine in vitro immortalization assay (0.2 clones per 10e5 cells, fitness score 0.00007), compared to MFG-γc (0.54 clones per 10e5 cells, fitness score 0.00025). Peripheral blood and bone marrow from C57BL6 mice transplanted with murine bone marrow transduced with pSRS11.EFS.IL2RG.pre* and followed in primary recipients over 4 months and in secondary recipients over 1 year in vivo showed no evidence of vector associated leukemias. Deep sequencing demonstrated 8 of 3621 insertions into the MDS-associated gene Evi1 in mice transplanted with MFG-γc-transduced cells while there were none (0 of 2690 insertions into Evi1) in mice transplanted with pSRS11.EFS.IL2RG.pre* vector transduced cells (P=0.025). In preclinical efficacy studies, circulating T and B lymphocytes were detectable in the peripheral blood of 7/7 γc-deficient mice transplanted with pSRS11.EFS.IL2RG.pre* transduced cells while 4/4 mice repopulated with SFFV-eGFP transduced cells remained alymphoid. Experimental animals were sacrificed approximately five months post-transplant for analysis of immune reconstitution. Flow cytometric analysis of the spleens and bone marrow revealed restoration of mature B220⁺IgM⁺ B cells and NK cell populations in all mice transplanted with pSRS11.EFS.IL2RG.pre* transduced cells. CD4⁺ and CD8⁺ T cells were also detected in both tissues and in thymi recovered from transplanted animals. T cells in these mice proliferated in response to mitogenic stimuli. Immunoglobulin subclasses IgG1 and IgG2a detected in the plasma from pSRS11.EFS.IL2RG.pre* reconstituted mice also indicated restored B cell function in these animals. Human preclinical studies also supported the correction of XSCID cellular defects using this vector. Based on these data, a multi-institutional phase I/II trial was initiated with the pSRS11.EFS.IL2RG.pre* vector using an identical clinical protocol as in the previous X-SCID trials, to determine efficacy and safety compared with the MFG-γc vector. The first patient was treated in December 2010. Six months post-gene therapy, he has attained CD3 T cell count of >800, normal proliferation to mitogens, and normal NK cell numbers. He has cleared medically-resistant oral ulcers, and a rotavirus infection acquired post-gene therapy. Nearly all of the circulating T cells (86%) and 41% of his NK cells express γc, albeit at modestly lower density than normal, as expected. However, the early kinetics of T cell reconstitution was comparable to several subjects treated with MFG-γc. These data suggest that the improved safety profile demonstrated with numerous surrogate preclinical studies is associated with efficacious transgene expression and functional immune recovery in the initial human patient treated.