Gene therapy trials: meeting the challenges

Gene therapy is a relatively new paradigm in medicine with enormous therapeutic potential. However, a number of widely reported adverse events  have focused attention on associated risks ahead of the exciting therapeutic progress being made. In 2000, the optimism of the gene therapy research community was bolstered by the first report of successful treatment of a genetic disease by gene therapy. The condition, X-linked severe combined immunodeficiency (SCID-X1), commonly diagnosed in early infancy, is characterised by recurrent infection as a result of an absence of cell-mediated and humoral immunity. Importantly, the majority of treated infants underwent full immunological reconstitution with eradication of established infections. A total of 20 infants were treated in two initial trials, nine in France: one by a group in Australia in collaboration with the French, and ten by a team in the UK. Long-term follow-up of the nine boys treated in the French trial, now aged between 8 and 11 years, has been reported, with eight patients alive after a median period of 9 years.This has established gene therapy as a realistic therapeutic alternative for patients without a suitably matched sibling donor, which is associated with less favourable survival rates. Overshadowing this impressive success, 30 months out from treatment, one of the initial patients in the French trial developed a T cell leukaemia, which was the direct result of the gene transfer vector used.This was shown to be related to retrovirus vector integration near the LMO2 proto-oncogene promoter, a phenomenon known as insertional mutagenesis, leading to aberrant transcription and expression of LMO2 .The trial was placed on voluntary hold while the cause was investigated and led to a worldwide effort to improve vector safety and reduce the likelihood of similar events in future clinical trials for this and other diseases using integrating vectors. A further three patients in the French trial and one in the British trial for SCID-X1 went on to develop T cell leukaemia, with four of the five patients being successfully treated with chemotherapy and in complete remission. Now, almost a decade later, a multi-national trial utilising a vector with improved safety features is open, and has begun treating patients.

Adenosine deaminase-deficiency (ADA-SCID) is another primary immunodeficiency for which gene therapy is showing great promise in the clinic. Initial trials for this disease were unsuccessful for several reasons that included the maintenance of patients on pegylated ADA (PEG-ADA) during gene therapy and the targeting of gene transfer to T lymphocytes . Cessation of PEG-ADA treatment during gene therapy facilitated a selective advantage for the gene-corrected cells. In addition, the targeting of haematopoietic stem cells (HSCs) using an improved gene transfer protocol and a myeloablative conditioning regime underpin the more recent success .More than 30 patients with ADA-SCID have been treated worldwide with gene therapy since 2000 , with the majority of those treated at the San Raffaele Telethon Institute for Gene Therapy in Milan.] In addition, six patients have been treated at Great Ormond Street Hospital in London, six at the Children's Hospital Los Angeles  and a further three at the University of California Los Angeles. In the majority of cases, reconstitution of immune function has been achieved without the need for supporting enzyme replacement therapy with PEG-ADA. Importantly, adverse events related to the gene transfer technology have not been observed.

Much excitement was caused by the report of successful immunotherapy of two patients with metastatic melanoma in September 2006.The Rosenberg group engineered tumour recognition into autologous lymphocytes from peripheral blood using a retrovirus encoding a T cell receptor. High, sustained levels of circulating engineered cells were retained in two patients up to 1 year after infusion, resulting in regression of metastatic melanoma lesions; a dramatic improvement for patients who had only been expected to live for 3–6 months. Although stable engraftment of the transduced cells was seen for at least 2 months after infusion in 15 other patients, they did not respond to the treatment. It appears that it is critical to obtain an effective tumour infiltrating lymphocyte population for the treatment to be successful, and further work is underway aiming to improve response rates and refine the approach. Recently, in a similar clinical trial, this strategy has been extended to treat patients with metastatic synovial cell carcinoma, which is one of the most common soft tissue tumours in adolescents and young adults. Clinical responses were observed in four of six patients with synovial cell carcinoma and in five of 11 patients with melanoma. Despite achieving similar levels of transduction and administering similar levels of gene-modified T cells to patients, the clinical responses were highly variable and require further investigation. Importantly, two of the 11 patients with melanoma were in complete regression at 1 year post-treatment and a partial response in one patient with synovial cell carcinoma was observed at 18 months.

Chimeric antigen receptors (CAR) represent another autologous cell-based therapy targeting tumour-associated cell-surface antigens. This approach is receiving increasing attention and brings together the expansion potential and persistence of cytotoxic T cells with the specificity of monoclonal antibodies. Researchers from the Abramson Cancer Center in Philadelphia have described the treatment of three patients with chronic lymphocytic leukaemia (CLL) with autologous T cells that were genetically modified to express a CAR with specificity for the B-cell antigen CD19. This publication builds on an earlier report for one of these patients in which complete remission was achieved following an infusion of 1.42 × 10⁷ transduced T cells. For these three patients, all carrying a considerable CLL tumour burden, cells were administered following conditioning chemotherapy designed for depletion of lymphocytes. Post-infusion complications were limited to a transient and treatable tumour lysis syndrome, occurring between 7 and 21 days post-infusion. At the time of publication, two of the three patients were in complete remission, at 10 and 11 months post-therapy, with the third patient showing a partial response at 7 months post-therapy. The persistence of modified cells was also demonstrated, suggesting that this therapy may provide sustained tumour control in these patients.

Until more efficient gene delivery systems and/or in vivo selection strategies are developed, many human diseases, potentially amenable to gene therapy, will remain beyond reach. For diseases such as SCID-X1 and ADA-SCID, genetically modified cells undergo selective expansion as a consequence of in vivo selection conferred by the disease pathophysiology despite the correction of only a modest number of progenitors. For diseases lacking a naturally occurring selection pressure, one promising possibility is to confer an exogenous selection pressure on gene-modified cells using a combination of gene and drug therapy. The potential of this type of strategy has recently been highlighted in a trial seeking to confer chemoprotection on human HSCs during chemotherapy with alkylating agents for glioblastoma.The strategy involves delivery of a mutant form (P140K) of methylguanine methyltransferase (MGMT) to HSC followed by alkylating chemotherapy in concert with a small molecule inhibitor of native MGMT, O⁶-benzylguanine (O⁶BG). The O⁶BG suppresses tumour MGMT expression, thereby enhancing sensitivity to alkylating chemotherapy. Gene-modified HSC expressing MGMT P140K are concurrently protected and selectively expanded. In this initial study, an extended survival of patients was reported but, in future, this approach could be used to expand progenitor cells bearing a therapeutic gene and P140K to allow in vivo selection and expansion of gene-modified cell numbers to a clinically useful therapeutic threshold.

Another strategy that can be employed is targeting disease states in which gene transfer to a small number of cells at anatomically discrete sites has the potential to confer therapeutic benefit. This has been impressively demonstrated in trials for a form of congenital blindness, Leber congenital amaurosis (LCA), which encompasses a group of incurable autosomal recessive dystrophies affecting the retina. One molecular form, accounting for approximately 10% of affected individuals, is caused by mutations in the gene encoding retinal pigment epithelium-specific 65-kDa protein, RPE65. This condition is characterised by a progressive deterioration of vision, with complete loss of sight by early adulthood. Using sub-retinal administration of recombinant adeno-associated viral vectors (rAAV) expressing RPE65, three independent clinical trials for LCA have been initiated . The preliminary results indicate there has been no detectable systemic dissemination of the vector from the treated eye, no evidence of a significant humoral immune response to either the vector or encoded RPE65 and no serious adverse events. Importantly, improvements in both objective and subjective measures of vision have been reported and maintained for up to 2 years ,supporting further clinical trials for this and other forms of LCA. Although the vector was initially administered unilaterally, redelivery to the contralateral eye has been safely performed with demonstrated efficacy, even years after the initial treatment . Given the progressive degenerative nature of this disease, the use of gene therapy as a potential treatment option needs to be undertaken before the disease reaches an advanced stage associated with photoreceptor cell death.

In a clinical trial performed in Paris, an 18-year-old male with β-thalassaemia was transplanted with autologous CD34⁺ cells transducedex vivo with a lentiviral vector expressing a marked β-globin transgene. Consistent with therapeutic benefit, the patient subsequently had a reduced transfusion requirement. Of interest, however, approximately half of the therapeutic effect observed appeared to be mediated by a dominant clone containing an integration site within the HMGA2 gene . Haematopoietic homeostasis is currently being maintained; however, long-term follow up is required to establish whether the presence of this dominant clone might lead to any adverse clinical effects.

A final trial targeting the haematopoietic compartment showing promising results is being performed at Hannover Medical School in Germany. In this trial for Wiscott–Aldrich syndrome, the results for ten patients have been reported. Following treatment, the clinical condition of the patients improved and long-term engraftment of gene-corrected cells was observed in the bone marrow; however, the development of an acute T cell leukaemia in one of the ten patients treated has been reported. The preliminary data implicate an insertion upstream of LMO2 and additional chromosomal mutations reminiscent of the genotoxic events occurring in the SCID-X1 trials. Again, gene transfer was achieved using a γ-retroviral vector containing transcriptionally active long terminal repeats.

Thus, at present, the field is making good progress, particularly in diseases targeting the haematopoietic system. Insertional mutagenesis, however, is now recognised as a clinically established risk associated with the use of integrating vectors. As a result, the field is currently exploring alternative vector systems, lacking strong viral enhancer elements and with safer integration profiles, aiming to reduce the likelihood of such events occurring in the future.

Number of trials per year

The number of trials initiated each year has tended to drop in those years immediately following reports of adverse reactions, such as in 2003 and 2007; however, 2005, 2006 and 2008 were strong years for gene therapy trials (Figure 1). The most recent years (2011 and 2012 in this case) tend to be underrepresented in the database because it takes time for articles to be published, causing a lag in obtaining information about the most recent trials.

 

Figure 1. Number of gene therapy clinical trials approved worldwide 1989–2012

Countries participating in gene therapy trials

Gene therapy clinical trials have been performed in 31 countries, with representatives from all five continents (Figure 2). We have located data on trials from three new countries since our last review , with these being Ireland (with three trials), as well as Romania and Saudi Arabia (with one trial each). The continental distribution of trials has not changed greatly in the last few years, with 65.1% of trials taking place in the Americas (64.2% in 2007) and 28.3% in Europe (26.6% in 2007), with growth in Asia reaching 3.4% from 2.7% in 2007.

Figure 2. Geographical distribution of gene therapy clinical trials (by country)