Long-Term Vector Integration Site Analysis Following Retroviral Mediated Gene Transfer to Hematopoietic Stem Cells for the Treatment of HIV Infection.

We previously reported the efficacy of nonmyeloablative allogeneic transplantation in 2 HIV positive recipients, one of whom received retrovirus transduced hematopoietic stem cells to confer resistance to HIV (Blood. 2002; 99:698–701). Half of the donor cells were genetically modified with a Moloney murine leukemia virus (MoMLV) based HIV resistance vector containing a transdominant negative mutant Rev (TdRev) (2.58×10e8 cells) or a control vector MoMLV based vector encoding GP91phox (4.04×10e8 cells). Here we report an assessment of retroviral integration sites recovered out to 3 years post-transplantation. We identified 213 unique retroviral integration sites (RISs) from the patient’s peripheral blood samples myeloid and lymphoid cells from 1 to 36 months after reinfusion of genetically modified CD34+ cells by linear amplification-mediated PCR (LAM-PCR). While overall vector integration patterns were similar to that previously reported, only 3.75% of RISs were common among early (up to 3 months) and late samples (beyond 1 year). This low percentage of overlap offers further evidence that the early phase of hematopoiesis after transplantation derives primarily from short-term repopulating cells. Additionally, we identified 14 common integration sites (CISs). Interestingly, common integration sites were enriched among late samples; 14.9% of early RISs were CISs vs. 36.8% late. A total of 3 RISs were found near or within known oncogenes, but 2 (Integrin alpha 9 [ITGA9] and ADP-ribosylation factor-like 11 [ARL11]) were limited to early time points. An integration site near the MDS1 gene was detected in a late follow-up sample by LAM-PCR. We confirmed the integration site near the MDS1 gene by PCR with integration site-specific primers amplifying the region between the 3’-LTR of the provirus and the MDS1 locus. The MDS1 integration was not detected in early, but became detectable at all time points from 6 months to 3 years post transplant from both lymphoid and myeloid populations. Q-PCR using an integration specific Taqman probe was utilized to assess the level of clonal contribution to hematopoiesis from the clone containing the MDS1 RIS. The overall contribution of the MDS1 integrated clone remained stable during followup. Given an overall gene marking level of 0.001-0.01% with an MDS1 marking level estimated at 0.00001% in the follow up samples, the frequency of the MDS1 integrated clone is predicted to be 1/1000 marked LT-HSCs. We infused an estimated 1324 transduced LT-HSCs based upon cell dose, transduction efficiency and an estimated LT-HSC frequency of 5 per 10e3 CD34+ cells. The single integration in MDS1 in the context of non-LT-HSC limited hematopoiesis may thus account for the stability observed over time. In summary, the pattern of contribution by genetically modified cells is distinct between the early and late phase post transplantation and emphasizes the importance of long-term studies to assess the risk of integrating vectors. Additionally, the enrichment for CISs in the late phase supports the concept that integrations in the LT-HSCs favors genes that may be involved in “stemness”. Furthermore, integrations in or near putative oncogenes are likely insufficient alone as a cause of oncogenesis. Finally, LT-HSC dose may be an important determinant of the risk of integrating vectors in the context of HSC gene transfer.