Gene Therapy for the Hemophilias

Christopher E. Walsh MD, Ph,D, Mt Sinai School of Medicine, One Gustave Levy Pl, New York City, NY, USA

Recent gene transfer trials for hemophilia A and B, bleeding hemophilia gene transfer require the long-term therapeutic disorders lacking either functional factor VIII or IX, respec- production of the coagulant protein without stimulating an tively, have produced tantalizing results, suggesting that the immune response to the transgene product or the vector. potential to correct these bleeding disorders at a molecular Based on a scientiﬁc understanding of the molecular and level may be at hand. Genetic correction of the hemophilias cellular defects, leading to the bleeding phenotype, impress- represents a model system to develop a basic understanding ive strides have been made in the last 2 years. of how gene therapy will be achieved. The goals for Gene Therapy (2003) 10, 999–1003. doi:10.1038/sj.gt.3302024

Keywords: hemophilia, gene transfer, gene repair, stem cells, viral vectors

In brief

Progress Prospects New AAV serotypes and lentiviral vectors Development of improved lentiviral, gutless adeno- have recently been studied for the production of virus and alternate AAV serotype vectors is in factor VIII or IX. progress Organs other than the liver can be used as ‘factories’ Gene repair, RNA repair (Trans-splicing) offer oppor- for factor VIII and IX production. tunities for the treatment of hemophila. The immune response to endogenous factor synthesis Gene-modified circulating endothelial progenitor and viral vectors is a potential problem in hemophilia may prove useful in the future. gene transfer. Gene-modified stem cell therapy may become Hemophilia clinical trials using either retrovirus, available. AAV or ex vivo transfected fibroblasts have been carried out or are ongoing.

Introduction severely affected patient with frequent spontaneous bleeding episodes (patients with o1% factor level) to a Current treatment for hemophila-related bleeding epi- moderate or mildly affected level. Coagulation assays are sodes utilizes intravenous infusion of purified and standardized and animal models (knockout mice and recombinant factor protein, which is effective, yet hemophilic canines) that mimic the human phenotype transient because of the short half-life of the proteins. are available for testing. A variety of gene transfer This treatment is expensive, restricts the prophylactic use approaches are currently being tested both in the of factors and can lead to crippling joint disease and laboratory and in the clinic. Results of two clinical trials, susceptibility to infectious agents. Effective hemophilia using nonviral and viral-based gene transfer approaches, gene transfer requires that a sustained, long-term (years) suggest that despite low factor levels, patients required production of coagulation factor at therapeutic levels be less factor infusion and reported fewer bleeding epi- generated. Thus, the method of gene delivery must be sodes. Although neither trial included a placebo arm, safe, and the risk of immune response to potential these results enforce clinical observations that low levels neoantigens must be minimal. Given recent scientific/ of factor dramatically reduce spontaneous bleeding. technical developments, genetic correction of hemophilic Despite the current excitement, results point to the need patients is now viewed as an achievable goal. for improved vectors. Here, we will review the recent The factor VIII and IX gene and protein products have advances over the past 2 years in this field that mirrors been extensively studied.1,2 Many tissues and cell types the advances in the field of gene transfer in general. (skeletal muscle, liver, spleen and skin) are capable of producing and expressing fully modified and functional New adeno-associated virus (AAV) serotypes factor IX protein. A therapeutic factor level is considered to be Z2% of normal levels. This is sufficient to convert a and lentiviral vectors have recently been studied for the production of factor VIII or IX

Correspondence: Dr CE Walsh, Mt Sinai School of Medicine, One Gustave In general, viral vectors exhibit long-term gene expres- Levy Pl, Rm 24-42C Annenberg Bldg, New York City, NY 10029, USA sion (years), whereas nonviral methods produce transi- Gene therapy for the hemophilias CE Walsh 1000 ent (weeks to months) factor expression. Hemostatic AAV serotypes 7 and 8 recently isolated from primates levels of factors VIII and IX were reached with first- and infected with high-dose adenovirus 6 are not neutralized second-generation adenovirus vectors. Unfortunately, by heterologous antisera raised to the other serotypes. the exuberant cell-mediated immune response engen- Recombinant serotypes 7 and 8 vector particles carrying dered by this vector leads to inflammatory response the alpha-1-antitrypsin cDNA were compared for trans- directed at transduced cells with the attendant loss of ducing effectiveness in mice. AAV7 was equivalent to protein expression. Newer gutless adenovirus with a AAV1 in efficient expression in skeletal muscle, whereas minimum of endogenous adenoviral genes may limit the AAV8 expressed at a 10 to 100-fold greater rate in liver- immune response, but as a consequence gene expression directed expression than all other serotypes. These data is drastically reduced. AAV has recently come to the fore confirm that relatively small differences in the capsid based on its now relative ease in preparation, ability to structure produce striking differences in transgene infect both dividing and nondividing cells,3 and expression in a wide variety of tissues. although it engenders a humoral immune response, it Lentiviral vectors have the potential to play an does not stimulate a cytotoxic lymphocyte response. important role in hemophilia gene therapy. One study Using this vector, therapeutic levels of factors IX and VIII used human immunodeficiency virus (HIV)-based lenti- were demonstrated in knockout mouse and hemophilic viral vectors containing human factors VIII or IX cDNA canines. Eight serotypes of AAV, have been isolated and expression for portal vein injection into C57Bl/6 mice. cloned (AAV1-8). Of these AAV, type 2 was the first Increasing doses of hFIX-expressing lentivirus resulted cloned and most extensively studied. Surprisingly, other in a dose-dependent, sustained increase in serum hFIX serotypes yield factor IX at levels two logs greater than levels up to approximately 50–60 ng/ml. Partial hepa- AAV-2 following skeletal muscle injection into mice4 and tectomy resulted in a 4- to 6-fold increase in serum hFIX produce sustained supratherapeutic factor levels leading of up to 350 ng/ml compared with the nonhepatecto- to complete loss of the bleeding diathesis.5 Such levels mized animals. The expression of plasma hFVIII reached are achieved as a result of more efficient effective skeletal 30 ng/ml (15% of normal), but was transient as the muscle gene transfer. Here, a linear relation exists plasma levels fell concomitant with the formation of anti- between input vector and factor expression. A unique hFVIII antibodies.7 side benefit is lack of an immune inhibitor response Owing to the potential safety issues using an HIV- presumably because of continuous production of factor based vector, an alternate approach is to utilize plasmid- as the major determinant for inducing tolerance. This based approaches that carry genetic elements that result is reminiscent of immune tolerance strategies promote integration. Early attempts using such a system currently used in the clinic, performed by repeated have provided encouraging results.8 infusion of factor. In particular, AAV1 produces robust transgene expression in muscle, but the exact mechanism is unclear. Data Organs other than the liver can be used as using AAV1/EGFP suggested that the number of transduced muscle fibers infected increases significantly sources for factor VIII and IX production with AAV1 compared to AAV2 (see Figure 1). A linear The liver is the principal organ synthesizing the relation between circulating levels of canine factor IX coagulation factors. However, other organs can synthe- protein and AAV1 dose suggests that this result is most size factors VIII and factor IX. Factor IX can be expressed likely because of the number of myocytes productively from skeletal muscle, fibroblasts, kerintinocytes, intest- infected. One working hypothesis explaining the ser- inal mucosa, cells lining the amniotic cavity and marrow otype transduction differences lies in the level of virus stroma.9 Factor VIII transgene expressing circulating specificity for binding cell receptors. At present, the endothelial cells are capable of secreting high levels of receptor for AAV type 1 has not been identified but factor VIII for a sustained period in animal models.10 appears distinct from type 2, that is, not inhibited by Circulating endothelial cells obtained from peripheral heparin. blood are expanded ex vivo and then genetically modified to express a gene of interest. The biochemical elements necessary for high-level factor expression including the endogenous coexpression of factor VIII and vWF may explain the high factor levels observed in vivo. The regulation of infused endothelial precursor cell growth kinetics and half-life of fully differentiated endothelium remains to be determined. Ex vivo gene transfer of hematopoietic progenitor cells and marrow stroma are also capable of factor VIII secretion in vivo.11

The immune response to endogenous factor synthesis and viral vectors is a potential Figure 1 AAV serotype transduction of skeletal muscle. Fluorecence of problem in hemophilia gene transfer skeletal muscle samples following injection of equivalent doses of serotype AAV1 and two vectors carrying an EGFP expression cassette. A uniform The antibody responses to exogenous factor replacement, ﬂuorescence pattern is observed with AAV1 compared to the patchy termed inhibitors, effect nearly 20% of factor VIII patients appearance of AAV2. and 3% of factor IX patients. Inhibitory antibodies that

Gene Therapy Gene therapy for the hemophilias CE Walsh 1001 bind to the particular regions of the factor molecule but not in germ cells. All patients had low (1:100–1000) inactivate by changing factor protein conformation.2 In preinjection anti-AAV2 neutralizing titers that increased general, the immune response may in part be related to after vector administration. An increase in the number of the type of mutation. For example, a large deletion in the injection sites from 10 to 90 produced no significant factor VIII gene and complete loss of protein typically increase in factor level. High titer neutralizing antibodies leads to a greater incidence of inhibitor formation. to AAV developed in all patients at levels sufficient to Bleeding episodes of patients with inhibitors are difficult preclude readministration of vector. Muscle biopsy to manage, relying on activated bypass factors and confirmed previous observations in animals that slow recombinant factor VIIa.12 Will a constant source of factor twitch muscle fibers expressed factor IX. engender high titer inhibitory antibodies negating any A dose-escalation study based on AAV2 vectors positive benefit and with consequences of worsening carrying human factor IX cassettes delivered via the bleeding? The clinical trials described below carefully hepatic artery has begun. Two patients received the screen for noninhibitor patients or patients with frequent lowest dose of virus (2 Â 1011/kg) via intra-arterial infusions where the chances of inhibitor are reduced. delivery without adverse effects. However, the trial was CD4+ subset activation in humans and Th-1 and Th-2 temporarily halted because of detection of the transgene lymphocytes in mice suggests that both MHC class I and in seminal fluid. Trial resumption was based on data that II mechanisms are involved. Involvement of both central germ cells were not infected with the virus. No (marrow, thymus) and peripheral (lymph nodes, Peyers detectable levels of factor FIX were observed above patch) tolerance mechanisms to factor VIII is described background in the two patients receiving the lowest dose but poorly understood.13–15 In addition to the immune of vector. Two patients received moderate doses of vector response to the transgene factor proteins, immune (1 Â1012/kg). One patient developed FIX levels up to 10– response to viral vectors is well described for adenovirus 12% within 2–3 weeks after injection; however, factor and AAV and thus prevent readministration of vector. levels subsequently dropped coincident with an eleva- tion of the liver transaminases. No data were presented on the second patient. Whether this represents toxicity of Hemophilia clinical trials using either the vector at this dose remains to be determined. It is interesting to note that based on dose–response experi- retrovirus, AAV or ex vivo transfected ments in hemophilic dogs, this moderate dose of virus fibroblasts have been carried out or are produced 4–14% of canine FIX for 1–2 years without ongoing significant toxicity. A nonviral approach used a factor VIII plasmid Within the past 2 years, five gene transfer trials were electroporated into autologous skin fibroblasts: cells approved (three for hemophilia A, two for hemophilia B) were expanded in vitro and 100 or 400 million cells in the US. Biotech firms that developed vectors specifi- injected into the greater omentum.18 Levels of factor VIII cally for factors VIII and IX sponsor all five trials. In a above pretreatment levels were measured in four of six phase I dose escalation study, 13 subjects with hemophi- patients with either a concommitant reduction in the use lia A received by peripheral intravenous infusion an of recombinant factor VIII or decreased number of amphotropic retroviral vector carrying a B-domain spontaneous bleeding episodes. However, factor VIII deleted human factor VIII gene. Infusions were adminis- decreased to pretreatment levels in all the patients after tered to patients with HIV and HCV infections and were 12 months. The explanation for the decline in factor well-tolerated. Factor VIII was measured and no subject expression may have been because of gene silencing, had sustained repeated FVIII levels 41% of normal immunological clearance or senescence of the fibroblasts levels. Patients were treated with vector ranging from after reimplantation. 3 Â 107 to 9 Â 108 vp/kg. Overall, there was no significant What do these clinical results tell us? No significant change in bleeding frequency. And there was no toxicity in any of the patients was reported. Furthermore, correlation between vector dose and time to FVIII the factor levels predicted from animal models were not activity response. This clinical outcome is consistent observed in patients. Thus, although testing of new with the limited capability of retroviral integration into factor proteins in hemophilic animals is traditionally nondividing liver cells and the lack of a liver-specific used because of similar pharmacokinetic profiles seen in promoter in the retroviral vector used. man, such extrapolation using gene transfer vectors may A trial using an AAV2 vector carrying the human not be clear-cut. A review of the preclinical data also factor IX cDNA injected intramuscularly was carried out suggests that animal studies may not be predictive of the in eight patients in a dose escalation study.16,17 One clinical outcome. For example, vector dosing based on a patient receiving the lowest dose (2 Â 1011 vp/kg) was vector particle-to-weight ratio produced discrepant reported to maintain factor levels at 1–2% and reported a results when comparing equivalent AAV vector dosing 50% reduction in factor usage and bleeding episodes for in mice and hemophilic dogs. Whereas experiments in a period up to 40 months postinjection. No evidence of hemophilic mice are dose-dependent and can produce inhibitor was reported despite preclinical data in dogs of supraphysiologic levels of factor IX (300% of a transient inhibitory response. Virus dissemination was normal), equivalent doses in hemophilic dogs produce transiently detectable in all body fluids, excluding factor IX at levels around 5% of normal and do not semen. At higher doses, no significant plasma factor appear to be dose-dependent.19 Recent data testing levels were reported. However, reduced frequency of AAV2/human factor IX vectors in non-human primates factor usage was one end point signifying treatment produced 4–10% factor IX,20 similar to data generated effectiveness. Molecular analysis of virus dissemination in hemophilic dogs,21 for a period of 1 year. These was detectable in all body fluids (saliva, blood and urine) outcomes reflect species differences in terms of the rate

Gene Therapy Gene therapy for the hemophilias CE Walsh 1002 of cell infectivity, gene expression, protein modification Advantages include ex vivo expansion and gene mod- and processing. However, testing in different animal ification with selected clones producing high levels of models serves to confirm the validity of each new factor. Autologous stem cells derived from each patient approach. would avoid transplantation rejection and immunosup- pression. Current disadvantages include the long lead time (months) required to generate the number of cells for transplantation and the ability to control the Future prospects differentiated fates of the transplanted multipotential cells. RNA repair (Trans-splicing) offers opportunities for the treatment of hemophila A novel approach for genetic correction involves the use of premessenger RNA (pre-m-RNA) repair. RNA trans- Summary splicing utilizes endogenous splicing mechanisms to correct a portion of the defective RNA. A pre-mRNA is A body of data suggests that genetic correction of the designed to base pair with a pre-mRNA transcribed from hemophilias is feasible. The subjective reporting by the defective gene. The pre-mRNA also contains all the patients of decreased bleeding episodes and an apparent requisite splicing signals that allows two independent self-declared reduction in bleeding episodes at nominal mRNAs to splice together, resulting in a correct copy of levels of factor strongly hint that reasonable factor levels mRNA that is translated into a normal protein.22 The if reached will achieve a major breakthrough in the advantage of this system is that large genes, unable to be treatment of hemophilia. Hemophilia gene transfer packaged into viral vectors, or genes that contain large represents the combination of vector delivery systems, regulatory elements could be corrected by using the animal models and clinical studies designed to answer smaller spliced sequences. We developed such a system specific questions. Not only will these studies benefit for factor VIII correction . Using the factor VIII exon 16 hemophilic patients, but should also instruct others in knockout mice as a model to test trans-splicing, we the field as well. Hopefully, this work will represent a demonstrated that by injecting plasmid or AAV carrying milestone in the use of genetics for treatment of human pre-mRNA encoding for exons 16–26 around 2–6% of ailments. factor VIII, that prevented bleeding challenge, was generated for 3–5 days and 3–4 months, respectively (Chao et al, submitted). As trans-splicing efficiency improves, this may be useful for the treatment of References autosomal dominant disorders involving other coagulation and thrombotic defects. 1 Kaufman R, Antonarakis S. Structure, biology, and genetics of factor VIII, In: Hoffman R, Benz EJ, Shattil S, Furie B, Cohen H, Silberstein L, McGlave P (eds). Hematology: Basic Principles and Gene-modified circulating endothelial progenitor may Practice, Vol. VIII-108, 3rd edn. Church Livingstone: New York, 2000, pp 1850–1868. prove useful in the future 2 Lillicrap D. Hemophilia treatment. Gene therapy, factor VIII The use of blood outgrowth endothelial cells (BOEC) as a antibodies and immune tolerance: hopes and concerns. Haema- source of cells synthesizing factor VIII has been tologica 85 (Suppl 10): 2000. 10 described. These circulating endothelial progenitor 3 Monahan P, Samulski R. AAV vectors: is clinical success on the cells are isolated from peripheral blood, expanded in horizon? Gene Therapy 2000;7:24–30. culture, and modified genetically to carry the normal 4 Chao H et al. Several log increase in therapeutic transgene FVIII gene. A significant advantage includes the expan- delivery by distinct adeno-associated viral serotype vectors. Mol sion of BOEC clones that synthesize vWF, the carrier Ther 2000;2:619–623. protein necessary for FVIII stability in plasma. Major 5 Chao H et al. Sustained and complete phenotype correction of questions of BOEC use include the half-life of these cells hemophilia b mice following intramuscular injection of aav1 in vivo, and potential for uncontrolled growth following serotype vectors. Mol Ther 2001;4:217–222. transplantation. 6 Gao G et al. Novel adeno-associated viruses from rhesus monkeys as vectors for human gene therapy. Proc Natl Acad Sci USA 2002;99:11854–11859. Gene-modified stem cell therapy may become 7 Park F, Ohashi K, Kay M. Therapeutic levels of human factor available VIII and IX using HIV-1-based lentiviral vectors in mouse liver. Recent reports on the plasticity of stem cells derived Blood 2000;96:1173–1176. from adult tissue forming liver, brain, muscle, skin, fat 8 Yant S et al. Somatic integration and long-term transgene expression in normal and haemophilic mice using a DNA have generated enormous interest on their use for the transposon system. Nat Genet 2000;25:35–41. genetic correction of hemophilia. A multipotential subset 9 Krebsbach P, Zhang K, Malik A, Kurachi K. Bone marrow of mesenchymal stem (MAPC) cells derived from stromal cells as a genetic platformfor systemic delivery of marrow stroma can be induced to differentiate into cell therapeutic proteins in vivo: human factor IX model. J Gene Med types with neuroectoderm, endoderm and mesoderm 2003;5:11–17. 23 characteristics. When MAPCs are injected into irra- 10 Lin Y et al. Use of blood outgrowth endothelial cells for gene diated animals, they differentiate into hematopoietic therapy for hemophilia A. Blood 2002;99:457–462. lineages as well as epithelium of the liver, gut and lung. 11 Chuah M et al. Long-term persistence of human bone marrow Potentially MAPCs could be genetically modified to stromal cells transduced with factor VIII-retroviral vectors and synthesize coagulation factors before re-transplantation. transient production of therapeutic levels of human factor VIII in

Gene Therapy Gene therapy for the hemophilias CE Walsh 1003 nonmyeloablated immunodeﬁcient mice. Hum Gene Ther 18 Roth D et al. Nonviral transfer of the gene encoding coagulation 2000;11:729–738. factor VIII in patients with severe hemophilia A. N Engl J Med 12 Poon M. Use of recombinant factor VIIa in hereditary bleeding 2001;344:1735–1742. disorders. Curr Opin Hematol 2001;8:312–318. 19 Wang L et al. Sustained expression of therapeutic level of factor 13 Chao H, Walsh C. Induction of tolerance to human factor VIII in IX in hemophilia B dogs by AAV-mediated gene therapy in liver. mice. Blood 2001;97:3311–3312. Mol Ther 2000;1:154–158. 14 Qian J, Collins M, Sharpe A, Hoyer L. Prevention and treatment 20 Nathwani A et al. Sustained high-level expression of human of factor VIII inhibitors in murine hemophilia A. Blood factor IX (hFIX) after liver-targeted delivery of recombinant 2000;95:1324–1329. adeno-associated virus encoding the hFIX gene in rhesus 15 Brown B, Lillicrap D. Dangerous liaisons: the role of ‘danger’ macaques. Blood 2002;100:1662–1669. signals in the immune response to gene therapy. Blood 21 Mount J et al. Sustained phenotypic correction of hemophilia B 2002;100:1133–1140. dogs with a factor IX null mutation by liver-directed gene 16 Kay M et al. Evidence for gene transfer and expression of factor therapy. Blood 2002;99:2670–2676. IX in haemophilia B patients treated with an AAV vector. Nat 22 Puttaraju M et al. Messenger RNA repair and restoration of Genet 2000;24:257–261. protein function by spliceosome-mediated RNA trans-splicing. 17 Manno C et al. AAV-mediated factor IX gene transfer to Mol Ther 2001;4:105–114. skeletal muscle in patients with severe hemophilia B. Blood, 23 Jiang Y et al. Pluripotency of mesenchymal stem cells derived Prepublished online Dec. 19, 2002. from adult marrow. Nature 2002;418:41–49.

Gene Therapy