Genetic Correction of Thalassemia: Development of a Transformative Is Delivered

In many ways the perfect suitors, thalassemia and gene therapy, seem to have come together at last.¹  Superb science and transformative clinical responses represent a landmark in medicine. Indeed, one might be (naively) tempted to ask, “why did it take so long?”

The β-globin gene represented the site of arguably the world’s first identified “genetic disease” in 1956 when Dr. Vernon Ingram identified the valine to glutamic acid substitution at position 6 in patients with sickle cell disease.²  Dr. David Weatherall and colleagues defined the biochemical basis of the thalassemias in the 1960s, and with the introduction of molecular biology, it was soon established that a broad range of mutations within the globin genes could lead to the clinical features of anemia, hemolysis, and dyserythropoiesis — all of which are typical of the syndrome. Indeed, the concept of thalassemia as a comprehensive model for genetic disorders has been established for more than 25 years.

Gene therapy has a more contemporary history. The first clinical application was undertaken in 1990 for the treatment of adenosine deaminase deficiency. However, as is often observed following the introduction of novel innovations, initial enthusiasm was followed by critical reappraisal, as well as a more gradual and measured appraisal of its enormous potential. Most of the striking recent reports have been focused in relatively uncommon clinical disorders. This landmark study offers an approach for applying gene therapy to one of the most common monogenic disorders in the world.

The article reports on 22 patients with β-thalassemia that were treated at two different centers. Nine had the β⁰/β⁰ genotype with very low globin gene transcription, and the other 13 had a β⁰ allele in combination with a less severe β⁺ allele (in most cases β^E). The patients were between 12 and 35 years of age and were considered inappropriate for sibling donor stem cell transplantation. The treatment was based on purification of autologous hemopoietic stem cells, virally mediated integration of a normal β-globin gene, and reinfusion into the patient after myeloablative chemotherapy. This may sound straightforward, but delivery of the reality has been somewhat more challenging.

Selection of CD34+ cells was achieved through the use of filgrastim and plerixafor. Most of these cells were then used for therapeutic genetic transduction, and a small portion (2×10⁶/kg) was retained for hemopoietic rescue in the potential event of failure of hemopoietic reconstitution. A lentiviral vector was selected for therapy and was optimised to ensure high levels of transcription and ability to generate high titres. The inserted β-globin gene had an extended structure and included elements of the locus control region as well as an allelic insertion that allowed selective monitoring of the “transduced” hemoglobin. Between 7 and 11 million transduced CD34⁺ cells/kg were infused following in vitro transduction.

A troublesome issue remains, the need for myeloablative conditioning prior to reinfusion of the transduced cells. Busulfan was used in all patients and drug concentrations were monitored regularly to ensure optimal dosage for the four days of therapy. Genetically-modified cells were then infused after 72 hours.

Clinical results were very strong with a very respectable median follow up of 26 months. Red cell transfusion independence was achieved in three of the nine patients with a severe β⁰/β⁰ genotype, and the rest demonstrated a 73 percent reduction in transfusion requirement. For those with a less severe non-β⁰/β⁰ genotype, the results were even more striking and all but one of the 13 patients became transfusion-independent. The concentration of “transduced” hemoglobin was measured at 34 to 100 g/L with total hemoglobin levels at 82 to 137 g/L. Moreover, additional hematological features of thalassemia, namely hemolysis and dyserythropoiesis, were also largely corrected.

A striking success of the treatment was the number of CD34⁺ cells that were transfused with the transgene. As such, the cells that expanded after infusion represented a genuine “polyclonal” population with between 202 and 5,501 unique integration sites. Adverse effects of therapy were modest and entirely consistent with the use of the conditioning therapy, with veno-occlusive disease being observed in two patients. Clinical factors such as age, previous splenectomy, and genotype were not predictive of response, and the primary determinant was vector copy number in the therapeutic product. Indeed, as the seven patients who achieved levels of transduced hemoglobin greater than 80 g/L all had a vector copy number higher than 0.6 in the infusion, the concept of a “gene dosage-hemoglobin response” relationship is now emerging.