The introduction of gene therapy into the clinic ignited a firestorm of interest by many stakeholders, including the scientific community. The application of gene therapy was envisioned for a broad range of disorders. The limiting step was, and still remains, technologies capable of safely and effectively transferring genes into a wider array of target cells beyond those of bone marrow.
Gene transfer in an ex vivo setting was demonstrated using murine retroviruses in a number of relevant target cells of epithelial and endothelial lineages. Quick translation into the clinic was promised with hopes of early success. Desperation often evoked expectations since many diseases being considered for early clinical trials of gene therapy were disabling and lethal with no available treatments.
It became clear, however, that the only viable way of approaching many of these diseases was to directly deliver the gene to the patient rather than via a transplanted cell in an approach called in vivo gene therapy. The disease that surfaced as an early candidate for in vivo gene therapy was cystic fibrosis (CF), which is caused by a defect in a chloride channel expressed in a variety of epithelial cells, the most important being those of the lung.
Gene therapy for CF would require direct delivery of the normal version of the gene into the lung airways. Unfortunately, vectors based on murine retroviruses were not up to the task for in vivo gene therapy. Vectors were made based on a virus, called the adenovirus, that is normally tropic for the lung.
Direct delivery of this vector into the airway of various animal models did efficiently transduce lung epithelial cells, setting the stage for several clinical trials of in vivo gene therapy in patients with CF.2,3,4
Similarly high levels of transduction were observed with adenoviral vectors in other target organs such as liver, heart and brain to name a few. These preclinical successes with the initiation of the CF trials fueled enthusiasm and expectation far in excess of the reality of the science.
The incredible efficiency with which adenoviral vectors transduce genes in vivo was corroborated in many laboratories. However, a disturbing trend of associated inflammatory type toxicities was being observed. The basis for these toxicities was the activation of adaptive and innate immune responses to the vector and the transduced cells.5 Attempts by us and others to circumvent these immune responses by vector engineering or co–administration of immune modulatory drugs were not successful.
Efforts were directed to develop vector systems that may be less immunogenic with the prime contender being those based on adeno-associated viruses (AAV). This obscure family of viruses was discovered in the 1960–1970s as contaminants in laboratory stocks of adenoviruses.6
Of the six different serotypes discovered, AAV serotype 2 was developed as a vector. It has limited cargo capacity (i.e., <4.8 KB) although it has the remarkable ability of transferring genes in vivo with minimal inflammation or activation of T and B cells against the transgene product. The vector genome persists in a nonintegrated form and continues to express the transgene for the life of the target cell.
However, the genome is destabilized if the target cell divides indicating that the optimal targets are those that are post-mitotic and long lived such as neurons, muscle fibers, photoreceptors, and hepatocytes. Despite these favorable properties, vectors based on AAV serotype 2 have had limited success in the clinic due to low transduction efficiencies and the presence of antibodies in humans as a result of natural AAV infections that inhibit the vector.7,8
Problems with the first-generation vector technologies described in this article led to delays and repeated failures. Stakeholders grew impatient and support for the field waned. A number of research programs endured despite the erosion of support for gene therapy, several of which eventually were tested in the clinic with remarkable success.
Studies of gene therapy for SCID that launched the field in the mid 1980s continued for several decades using the ex vivo bone marrow approach with murine retroviruses as the vectors. The key was to find a disease in which the transduced HSC and/or its progeny had a selective advantage over nontransduced cells.
Incremental advances in the biology of HSCs and techniques of retroviral transduction also helped progress these programs. If successful in these kinds of diseases, the therapy would allow complete repopulation of the treated patient with genetically corrected progeny cells without the need to ablate the endogenous marrow with radiation and chemotherapy, which carries substantial risks of mortality.
One form of SCID caused by a defect in a cytokine signaling receptor was the test case. The results were spectacular in that virtually all who were treated regained substantial immune function.9 Unfortunately, many of these patients eventually developed T-cell leukemias due to integration of the vector genome into a site that activated a tumor-promoting gene.10 Most of these leukemias have been treated and the cohort of patients treated with gene therapy clearly is much better off than they would have been without gene therapy.
The development of cancer is a consequence of the biology of murine retroviral vectors and may not be easily overcome, indicating the need for second-generation technology.
The other widely celebrated clinical success using first-generation technology was in the treatment of children with an inherited form of blindness called Leber congenital amaurosis (LCA). The form of this disorder targeted for gene therapy is defective in an enzyme involved in sensing light and is found in a cell located in the back of the retina.
Studies in dogs with this same disease showed reconstitution of sight following injection of an AAV2 vector into the retina.11 Three independent groups demonstrated reconstitution of visual function in LCA patients treated with AAV2-based gene therapy.12,13,14 The remarkable success of this program was due to unwavering persistence of the scientists and a careful selection of the target disease in an application that was able to succeed despite limitations of the technology.
The ability of gene correction to reconstitute function that can be quantitatively measured noninvasively was critical to the early assessment of efficacy in LCA. The localized delivery of a vector at high concentrations and long dwell times in the back of the retina helped overcome the inherent inefficiency of AAV2.
The clinical successes described here set the stage for a remarkable resurgence in support for gene therapy. The question is how can we leverage these successes into a broader array of applications with the goal of delivering commercial products? These seminal clinical trials established the feasibility of gene therapy in humans and helped overcome a variety of political and regulatory barriers that had surfaced.