Gene Therapy for SCID-X1: Focus on Clinical Data

A recent report by Gaspar et al. summarizes the clinical long-term follow-up (LTFU) of 10 patients presenting with X-linked severe combined immunodeficiency (SCID-X1) who were treated with first-generation gamma-retroviral vectors to reconstitute the expression of the interleukin-2 receptor common γ-chain (IL2RG) in autologous hematopoietic cells.¹ As in a related study of nine patients conducted by Hacein-Bey-Abina et al.,² the children enrolled in this study had no human leukocyte antigen–matched donor, thus reducing the safety and effectiveness of an allogeneic stem cell transplant.^3,4 In both studies, gene-modified hematopoietic cells were infused without prior myelosuppressive conditioning. Despite some differences in the design of vector backbones, envelope proteins, and transduction conditions, these studies share five important major findings.^1,2

1. The overall survival of the recipients of gene-modified hematopoietic cells was 95% (18/19 patients; range of reported LTFU 54–144 months), which compares favorably to that of recipients of mismatched allogeneic cells.

2. Regeneration of gene-corrected T cells continues over the long term and is largely responsible for the clinical benefit of the patients; de novo T-cell regeneration was even observed after intense cytotoxic chemotherapy to treat patients who suffered the complication of T-lymphoblastic leukemia (T-ALL), indicating the long-term survival of a gene-corrected thymic progenitor population.

3. T-ALL, caused by insertional activation of proto-oncogenes (predominantly but not exclusively LMO2), responded to standard chemotherapy regimens by entering complete remission in four of the five affected patients.

4. Regeneration of gene-corrected B and natural killer (NK) cells occurred for only a few months after infusion of gene-modified cells; subsequent marking levels in these compartments dropped to the baseline, suggesting lack of persistence of gene-corrected multipotent hematopoietic stem cells (HSCs).

5. Despite marginal regeneration of gene-corrected NK and B cells, immune recovery was sufficient to enable a normal development of the patients, in some cases requiring supportive treatment with antibiotics and/or immunoglobulins.

Taking together these findings, the overall survival and benefit to the gene therapy patients compare favorably to those attained with the mismatched transplantation regimens available for SCID-X1 patients lacking a matched donor. It is prudent, however, to propose protocol modifications before concluding that gene therapy should be considered first-line treatment for SCID-X1 patients who lack a human leukocyte antigen–matched donor. Gaspar et al. consider modifications to enhance HSC engraftment,¹ which could be achieved via partially myeloablative conditioning regimens, improved transduction protocols, or lentiviral vectors that support transduction of quiescent HSCs. Although all these measures may improve the long-term correction of B- and NK-cell function, it is more difficult to predict the effects on the risk of integrant-associated adverse events in the T-cell lineage. One possibility is that more robust long-term engraftment of gene-modified HSCs will increase the numbers of gene-corrected thymus-invading cells and thus reduce the proliferation pressure on individual clones that undergo thymic maturation. The resulting smaller clone size may decrease the risk of acquiring secondary leukemogenic mutations, which have been reported in all cases of T-ALL after insertional mutagenesis.^5,6

Conversely, a larger pool of thymic immigrants may increase the probability of a rare insertional mutant to home to this unique environment with its specific stimulatory conditions and thus unexpectedly increase the risk of T-ALL. In the absence of experimental data addressing the validity of these possibilities, it is important to carefully evaluate vectors that reduce the risk of insertional mutagenesis before increasing engraftment levels. Therefore, a revised clinical protocol, conducted by a transatlantic consortium of academic researchers and clinicians, has started to test the impact of changing a single variable in gene therapy of SCID-X1, namely, a redesigned vector-expression cassette intended to reduce the risk of proto-oncogene activation.^7,8 Data obtained from this trial will be instrumental in designing modifications to engraftment conditions as a second step.

Another remarkable observation can be made when reading the article by Gaspar and colleagues.¹ This appears to be the first published paper to report the LTFU of patients enrolled in a gene therapy trial for an inherited immunodeficiency syndrome that focuses on clinical data, without presenting complex molecular integrome analysis to address the “subclinical noise” of clonal imbalance caused by insertional mutagenesis. Similarly, the benefit–risk evaluation of genotoxic chemotherapy in oncology is based entirely on clinical data, including the analysis of adverse reactions such as secondary malignancies, rather than on attempts to define the residual genotoxic damage in surviving cells. Nevertheless, it is also important to recognize that the increasing accuracy in deciphering the complex integration pattern of gene vectors, and in describing the various waves of clonal selection of gene-modified cells occurring in patients over time, makes a major contribution to our understanding of basic mechanisms of hematopoiesis and establishes indirect evidence for the risk prevention associated with new vector tools.^{9,10,11,12,13,14} However, to contribute to the clinical decision making, we need further data on the prospective power of integrome biomarkers. Therefore, the approach taken by Gaspar et al. reflects a pragmatic two-step analysis of benefit–risk assessment in gene therapy: clinical data first, molecular findings second.