175. A clinical trial in Japan of retroviral-mediated gene transfer to bone marrow CD34+ cells as a treatment of adenosine deaminase (ADA)-deficiency

INBORN ERRORS OF METABOLISM: HEMOPHILIA AND OTHER DISEASES 175. A Clinical Trial in Japan of RetroviralMediated Gene Transfer to Bone Marrow CD34+ Cells as a Treatment of Adenosine Deaminase (ADA)-Deficiency Makoto Otsu,1,4 Tadashi Ariga,1 Yoshihiro Maeyama,1 Jukei Yoshida,1 Satoru Nakajima,1 Miyuki Kida,1 Nariaki Toita,2 Norikazu Hatano,2 Nobuaki Kawamura,2 Motohiko Okano,2 Ryouji Kobayashi,2 Osamu Tatsuzawa,3 Masafumi Onodera,4 Fabio Candotti,5 Kunihiko Kobayashi,2 Yukio Sakiyama.1 1 Research Group of Human Gene Therapy, Hokkaido Univ. Grad. Sch. Med., Japan; 2Dept. of Pediatrics, Hokkaido Univ. Grad. Sch. Med., Sapporo, Japan; 3Div. of Infectious Dis. & Rheumatology, Natl. Center for Child Health and Development, Tokyo, Japan; 4Div. of Hematology, Univ. of Tsukuba, Tsukuba, Japan; 5GMBB, NIH, Bethesda, MD. For ADA-deficiency, a form of severe combined immunodeficiency, retroviral-mediated gene correction of peripheral T cells or hematopoietic stem/progenitor cells (HSCs) is a treatment of choice alternative to bone marrow (BM) transplantation. Recently, two new such clinical trials aiming for the correction of patients’ HSCs were conducted in US and Italy, resulting in remarkably different outcomes: Italian patients showed immune reconstitution whereas the US trial led to no clinical improvement. There are two major differences between their protocols, either or both of which may explain these consequences. 1) In Italy, patients were mildly myeloablated, while not in the US trial. 2) US patients were on the concomitant enzyme replacement therapy (PEG-ADA) that may lessen in vivo selective advantages for gene-corrected cells and thus the efficacy of gene therapy. Italian group has not given the enzyme throughout the studies. We thus have proposed a trial that differs from these two to see If only the absence of PEG-ADA is enough to benefit patients, or if any kind of myeloablation is absolutely necessary. Two ADA-deficiency patients (#1, a 4 yr-old girl; #2, a 13 yr-old boy) were treated under the approval by the Japanese authorities. First, PEG-ADA was discontinued prior to the treatment. Rapid decrease in peripheral counts of lymphocytes and of neutrophils, liver enzyme abnormalities, and mild anorexia became evident, suggesting that the generalized toxicity due to ADAdeficiency had deteriorated. Nevertheless, both patients well tolerated the first several weeks with no serious problems. Good purity CD34+ BM cells (> 80%) were transduced with the retroviral vector GCsapM-ADA/PG13. The use of serum-free medium, a cocktail of cytokines (SCF, TPO, flt3-L, IL-6, and soluble IL-6 receptor) and the fibronectin fragments resulted in ∼50-70% transduction efficiencies and normal levels of ADA activities in the infused cell populations. No immediate adverse reactions were noted upon cell infusion (1.3 and 0.9 x 106 CD34+ cells/kg for #1 and #2, respectively). Patients did not receive any myeloablative conditionings. Although still preliminary in the limited time frames after treatment (6 weeks for #1 and several days for #2), we have observed normalization of liver enzyme abnormalities in #1 with gradual recovering from anorexia. Moreover, her neutrophil counts, not lymphocytes’ though, have also returned to normal by week 6, suggesting that gene-corrected HSCs have engrafted and started expanding at least towards the myeloid lineage, while providing the patient with significant amount of ADA enough to detoxify the affected organs. In light of the serious adverse events in the French gene therapy trial for XSCID, careful follow-up of the patients is mandatory and thus underway, including monitoring a clonal expansion of any type of hematopoietic cells . More detailed outcomes will be presented.

Molecular Therapy Volume 9, Supplement 1, May 2004 Copyright  The American Society of Gene Therapy

176. In Vivo Selective Enhancement of Correction of Human X-Linked Chronic Granulomatous Disease Neutrophils in NOD/SCID Mice Transplanted with CD34+ Stem Cells Transduced with a Bicistronic gp91phox/Benzyl Guanine Resistant Methylguanine Methyltransferase Vector Uimook Choi,1 Kouhei Yamashita,1 Lanise Caldwell,1 Harry L. Malech.1 1 Laboratory of Host Defenses, National Inst. Allergy and Infectious Diseases, NIH, Bethesda, MD. X-linked chronic granulomatous disease (X-CGD) results from a deficiency of gp91phox, the NADPH/flavin/heme binding component of the phagocyte oxidase. Correction of gp91phox expression provides no selective growth or survival advantage to hematopoietic stem cells (HSC). Achievement of clinical benefit from CGD gene therapy may require co-expressing a gene conferring a selective growth advantage at the stem cell level together with therapeutic gp91 phox gene. Methylguanine methyltransferase (MGMT) can repair alkylation of DNA resulting from 1,3-bis (2chloroethyl)-1-nitrosourea (BCNU) treatment. Human P140K mutant MGMT (hMGMT*) is highly resistant to inactivation by the irreversible inhibitor substrate, O6-benzylguanine (BG), while retaining DNA repair capacity, thus protecting HSC from synergistic cytotoxicity of sequential treatment with BG and BCNU. The advantage of this selective system as shown by others is that it can be applied in vivo to enhance the percent of cells marked with hMGMT*. We used a bicistronic onco-retrovirus bicistronic vector, RD114 pseudotyped MFGS-gp91 phox -IRES-hMGMT*, to transduce human CD34+ cells from patients with X-CGD. After transplant into NOD/SCID mice, we demonstrated in vivo selective enhancement of gp91phox expression in human neutrophils arising in these mice following treatment with BG/BCNU 20/6 mg/kg at week 4 and week 5 post transplant. 90% of the ex vivo transduced human X-CGD CD34+ cells expressed gp91phox transgene. Neutrophils differentiating from these transduced cells in culture demonstrate a high level of correction of oxidase activity, and this correction can be further enhanced by ex vivo selection with BG/BCNU. Transplant of the transduced X-CGD CD34+ cells into NOD/SCID mice was followed by BG/BCNU treatment of some animals. NOD/SCID mice were sacrificed at 7 weeks to evaluate gp91phox transgene expression and oxidase activity without or with in vivo BG/BCNU treatment. Average human cell engraftment (human CD45+) in mice not treated with BG/BCNU was 18.5±13 % versus 29±13 % in mice treated with BG/BCNU, a trend that did not reach statistical significance. However, in these experiments only 0.8±0.6 % of human neutrophils in bone marrow of untreated mice (CD45+/high side scatter) expressed gp91 phox, but 11.4±4.6 % of human neutrophils in BG/BCNU mice expressed gp91phox transgene, a difference that was highly significant (p