Integration-Mediated Activation of PRDM16 and HMGA2 in Multiple Clones without Adverse Hematopoietic Consequences Following Transplant of Autologous MGMTP140K Gene-Modified CD34⁺ Cells

Abstract 2053

Drug resistance gene therapy with mutant MGMTP140K gene-modified hematopoietic CD34⁺ cells has been proposed to circumvent myelosuppression associated with alkylating agent chemotherapy; however, the safety of this approach has not yet been demonstrated. We have achieved successful engraftment of MGMTP140K gene-modified hematopoietic progenitor cells in 3 glioblastoma patients following BCNU (600mg/m²) conditioning, as part of a Phase I/II clinical study which includes post-transplant combination O⁶-benzylguanine (O⁶BG) and temozolomide (TMZ) chemotherapy. Two of these patients, one of whom remains alive with no evidence for disease progression at more than 2 years since diagnosis, received a total of 9 and 4 cycles of O⁶BG/TMZ chemotherapy, respectively. We observed transient increases in the number of circulating gene-modified white blood cells (WBCs) and granulocytes in these patients following each cycle. Analysis of CD34⁺ colony forming cells (CFCs) in peripheral blood revealed increases in the number of gene-modified CFCs over time and with multiple cycles of chemotherapy. Given this, we have conducted a longitudinal retroviral integration site (RIS) analysis in an attempt to identify any selective advantage for gene-modified clones following repeated O⁶BG/TMZ regimens. In the first patient, at day 200 following transplant, 2 separate RISs were identified, each mapping to the intronic region between exons 1 and 2 of the human PRDM16 gene, previously identified as being associated with insertional activation and clonal expansion in a gene therapy trial for X-linked chronic granulomatous disease (Ott et al 2006). Semi-quantitative capture site frequency analysis revealed that these 2 clones with PRDM16 integrations constituted >20% of the gene-modified cell pool, which included hundreds of RISs at this time point in this patient. Overall gene marking at this same time point was 39.9% of peripheral blood WBCs, indicating that as many as 8% of circulating WBCs arose from these two clones. In the second patient, 3 distinct RISs were identified mapping to the 3'untranslated region (UTR) of the HMGA2 gene, within a cluster of let-7 microRNA binding domains known to be responsible for post-transcriptional regulation of this gene product. While HMGA2-associated RISs have been reported in 2 other gene therapy trials for β-thalassemia (Cavazzana-Calvo et al 2010) and X-linked severe combined immunodeficiency syndrome (Wang et al 2010), the genomic loci of these RISs mapped to the third intron of this genomic sequence. Capture frequency analysis indicated that overall contribution of the most abundant HMGA2-associated clone identified in this study reached 5.2% of gene-modified peripheral blood WBCs at day 300 after transplant. Analysis of PRDM16 and HMGA2 transcripts in each patient's cells by reverse-transcriptase real-time PCR indicated increased transcription of these genes in peripheral blood and bone marrow WBCs. Analysis of WBC subsets in the first patient found PRDM16 RISs present in granulocyte and CD3⁺ blood cell lineages, suggesting these clones originated from multipotential gene-modified hematopoietic stem cells capable of differentiating despite PRDM16 transactivation. Furthermore, since discontinuation of chemotherapy, we have observed a steady decline in levels of both PRDM16 clones, with complete disappearance of one clone. Extended analysis of HMGA2 expression and clone tracking in WBC subsets of the second patient is currently under way. Importantly, analysis of both patients bone marrow demonstrated normal cytogenetics, with no evidence for MDS or leukemia by hematopathology and flow cytometry. We believe that these data, in combination with other clinical gene therapy studies identifying similar clones, indicate the PRDM16 and HMGA2 gene sequences as “hot-spots” for either initial retrovirus integration in CD34⁺ hematopoietic progenitor and stem cells, or as genomic loci where aberrant gene expression, as a result of retrovirus integration, is associated with increased cell proliferation. At this time, these observations suggest that in the absence of underlying genomic disease, in vivo fluctuations in PRDM16- and HMGA2-associated clones following transplant may be a common, clinically inconsequential phenomenon in retrovirus-mediated gene therapy targeting CD34⁺ cell populations for modification.