A modified γ-retrovirus vector for X-linked severe combined immunodeficiency

Abstract

Background: In previous clinical trials involving children with X-linked severe combined immunodeficiency (SCID-X1), a Moloney murine leukemia virus-based γ-retrovirus vector expressing interleukin-2 receptor γ-chain (γc) complementary DNA successfully restored immunity in most patients but resulted in vector-induced leukemia through enhancer-mediated mutagenesis in 25% of patients. We assessed the efficacy and safety of a self-inactivating retrovirus for the treatment of SCID-X1.

Methods: We enrolled nine boys with SCID-X1 in parallel trials in Europe and the United States to evaluate treatment with a self-inactivating (SIN) γ-retrovirus vector containing deletions in viral enhancer sequences expressing γc (SIN-γc).

Results: All patients received bone marrow-derived CD34+ cells transduced with the SIN-γc vector, without preparative conditioning. After 12.1 to 38.7 months of follow-up, eight of the nine children were still alive. One patient died from an overwhelming adenoviral infection before reconstitution with genetically modified T cells. Of the remaining eight patients, seven had recovery of peripheral-blood T cells that were functional and led to resolution of infections. The patients remained healthy thereafter. The kinetics of CD3+ T-cell recovery was not significantly different from that observed in previous trials. Assessment of insertion sites in peripheral blood from patients in the current trial as compared with those in previous trials revealed significantly less clustering of insertion sites within LMO2, MECOM, and other lymphoid proto-oncogenes in our patients.

Conclusions: This modified γ-retrovirus vector was found to retain efficacy in the treatment of SCID-X1. The long-term effect of this therapy on leukemogenesis remains unknown. (Funded by the National Institutes of Health and others; ClinicalTrials.gov numbers, NCT01410019, NCT01175239, and NCT01129544.).

Figure 1. Immune Reconstitution and Gene Marking after Gene Therapy

Panels A, B, and C show changes over time in the numbers of CD3+, CD4+, and CD8+ lymphocytes, respectively. Panel D shows CD3+ cell counts at the indicated times after gene therapy, compared between the 20 patients enrolled in previous trials using the MFG-γc vector (open blue circles) and the 8 patients enrolled in the current trials using self-inactivating γ-retrovirus (SIN-γc) (solid black circles) (P = 0.28). Patient 3 was excluded because of high-level maternal engraftment. Panel E shows changes over time in naive CD4+CD45RA+ lymphocyte numbers. Panel F shows T-cell receptor diversity in T cells of 7 patients after gene therapy. The length of complementarity-determining region 3 (CDR3) in each indicated T-cell receptor beta chain variable (TCRBV) gene family was measured after amplification with family-specific primers. The horizontal axis represents CDR3 length, and the vertical axis represents the frequency of sequences with a given CDR3 length; a Gaussian distribution of CDR3 lengths is indicative of normal diversity. Panel G shows that vector copy number (VCN) in peripheral-blood CD3+ lymphocytes was detectable in 7 patients and was sustained over time. Panel H shows T-cell proliferation in response to phytohemagglutinin (PHA), measured 6 months after gene therapy and at the last follow-up. Panels I and J show increases in CD3–CD56+ (natural killer) lymphocyte numbers in patients with CD34+ cells infused with a VCN of at least 0.7 copies per cell (Panel I) but not in those infused with a VCN of less than 0.7 copies per cell (Panel J).

Figure 2. Analysis of Integration-Site Distributions in Patients Treated with the SIN Vector

Panel A shows a heat map summarizing the placement of integration sites (data sets in columns) relative to mapped genomic features (in rows). Each column summarizes results for a pool of unique integration sites from all patients in each trial or the pretransplantation pools from patients in the current trial. Each row indicates a form of genomic annotation; relevant databases are given in parentheses. Some associations were measured with the use of a sliding window of defined length; because the most meaningful length for comparison was often not known in advance, comparisons over multiple different lengths are shown (indicated by numbers on the left). Colors indicate departures from random distributions; darker shades of red indicate more strongly positive associations, and darker shades of blue indicate more strongly negative associations, as compared with the random distribution. Associations were summarized with the use of the receiver-operating-characteristic (ROC) method. The integration frequency relative to gene activity (“Expressed genes”) was quantified as for gene density, but only genes in the 1/16 highest expression category or top half expression category in 1-Mb intervals were counted. Panel B shows the distribution of integration sites relative to published mapped sites of epigenetic modification in hematopoietic progenitor cells (CD34+CD133+ cells). Darker shades of blue indicate more strongly positive associations, and darker shades of yellow indicate more strongly negative associations, as compared with the random distribution. Panel C shows the three top-scoring clumps of integration sites, all of which were enriched in the trials with the MFG-γc vector relative to the current trial (a detailed description is provided in the report in the Supplementary Appendix). The numbers denote coordinates on the indicated chromosome; blue lines denote mapped transcripts, and vertical lines indicate the positions of integration sites from each trial (the MFG-γc trial in green and the SIN-γc trial in orange). Clumps were found at MECOM (top), CCND2 (middle), and LMO2 (bottom). Panel D shows the comparison of the frequencies of integration sites near the lymphoid proto-oncogenes. Each point represents an individual patient who has undergone gene therapy, and the P value of 0.003 is for the comparison between trials on a per-patient basis (i.e., each patient was analyzed as a single data point). The integration-site data sets are catalogued in the Methods section of the Supplementary Appendix.