50 integration site analysis in a clinical trial of lentiviral vector based HSC gene therapy for MLD shows high levels of polyclonal hematopoietic reconstitution and no signs of genotoxicity at one year after transplant

50 Integration Site Analysis in a Clinical Trial of Lentiviral Vector Based HSC Gene Therapy for MLD Shows High Levels of Polyclonal Hematopoietic Reconstitution and No Signs of Genotoxicity at One Ye[.]

Molecular Therapy Volume 20, Supplement 1, May 2012 Copyright © The American Society of Gene & Cell Therapy

S20

CLINICAL GENE & CELL THERAPY ORAL ABSTRACT SESSION

in vivo selection drug resistance transgene mutant methylguanine

methyltransferase P140K, anti-HIV, and fl uorescent reporters We

are now able and plan to test the engraftment of MniPSC derived

hematopoietic cells transplanted at different stages of differentiation

in an autologous transplantation setting for their ability to give us to

the elusive HSC as well as to use this system as a model to evaluate

the effi cacy of stem cell therapies for the treatment of HIV and

hematologic genetic diseases

48 Derivation of “Semi-Universal Donor Stem

Cells” That Express a Specifi c Class I Human

Leukocyte Antigen (HLA) A0201

German G Gornalusse,1 Roli K Hirata,1 Laura Riolobos,1 David

W Russell.1,3

1 Department of Medicine, University of Washington, Seattle, WA;

2 Department of Biochemistry, University of Washington, Seattle,

WA.

The clinical use of Human Embryonic Stem Cells (hESCs) is

restricted because of immunological rejection reactions, mostly due to

genetic disparities in the Human Leukocyte Antigen (HLA) locus One

way to circumvent this problem is to create HLA-defi cient, universal

donor stem cells by targeted disruption of the β2-Microglobulin

(B2M) gene, which encodes the common subunit of all HLA class I

heterodimers However, this strategy may result in immunological

rejection mediated by Natural Killer (NK) cells, which can eliminate

HLA class I-negative cells, and could also pose problems if the

transplanted cells harbor viral infections or transform into cancer cells

To overcome this problem, we demonstrate that it is feasible to express

a specifi c HLA class I allele (HLA-A0201) as a single chain dimer

on the surface of B2M knock-out hESCs by employing an integrating

foamy virus vector ΔΨ-PPE-HLA(bBA0201) This vector includes

a puromycin resistance gene driven by the ubiquitously expressed

PGK promoter, and a separate expression cassette with the EF1α

promoter driving the expression of the HLA single chain construct

This novel engineered protein is a fusion polypeptide in which the

β2-Microglobulin subunit is covalently attached to the coding sequence of

HLA-A0201 We transduced 2 independent β2M-/- hESC clones with

the ΔΨ-PPE-HLA(bBA0201) foamy vector and obtained puromycin

resistant colonies (at MOI=10, we obtained 40 puroR clones/1500

unselected CFU, which is ∼3% resistant CFU) Southern blot analysis

demonstrated that all the clones analyzed (n=21) carried the intact

provirus, as evidenced by the presence of a unique ∼3.9 kb band, and

the majority of them were single copy integrants By fl ow cytometry

analysis, we showed that the B2M-HLA-A0201 fusion protein was

robustly expressed on the cell surface of B2M-/- hESCs, and we are

currently analyzing the immunological reactions to these cells and

their differentiated derivatives In addition to these overexpression

experiments, in which we ectopically express the B2M-HLA-A0201

fusion protein from an EF1a promoter, we are also employing a

gene targeting protocol to knock-in an HLA-A0201 gene as a fusion

construct at the B2M locus and confer properly regulated gene

expression Both approaches ultimately generate hESC lines that

express a single, unique, HLA class I protein on the cell surface In the

case of HLA-A0201, this allele is particularly common in the United

States where it exists in 48% of Caucasians, 46% of Hispanics, and

24% of African-Americans Therefore this single cell line should be

compatible with a large percentage of the US population, and can

be considered a semi-universal donor cell A similar approach could

also be used to uniquely express other common HLA class I proteins

on B2M-/- hESCs.

Clinical Gene & Cell Therapy Oral Abstract Session

49 Transgene T Cell Response in a Phase 2 Clinical Trial of rAAV1.Alpha-1 Antitrypsin Gene Therapy Vector: CD8+ Mediated and HLA C*05 Restricted

Roberto Calcedo,1 Suryanarayan Somanathan,1 Qiuyue Qin,1

Terence Flotte,2 Jeffrey Chulay,3 James M Wilson.1

1 Gene Therapy Program, Department of Pathology and Laboratory Medicine, University of Pennsylvania, Philadelphia, PA; 2 University of Massachusetts Medical School, Worcester, MA;

3 Applied Genetic Technologies, Alachua, FL.

T cell responses to A1AT transgene product was analyzed in 9 subjects with A1AT PI*Z mutations enrolled in a Phase 2 clinical trial

of a recombinant adeno-associated virus (rAAV) alpha-1 antitrypsin (A1AT) gene therapy vector None of the subjects had detectable T cells to the A1AT transgene product before vector administration After vector administration, 2 subjects developed a T cell response to A1AT measured by IFN-γ ELISPOT The magnitude of the response

in one of the subjects was high enough for epitope mapping and intracellular cytokine staining (ICS) analyses The mapping analysis revealed an immuodominant epitope located at position 202-216 of the mature protein and distant from the site of the PI*Z mutation; which

is located at position 342 of the mature protein The ICS analysis revealed an IFN-γ and TNF-α CD8+ T cell mediated response with a CD45RO- CD27- effector phenotype We also evaluated differential antigen processing and presentation of the mutant PI*Z A1AT using an

ex vivo mouse model of antigen presentation where mouse fi broblasts expressing PI*Z- or PI*M-A1AT were co-cultured with A1AT-specifi c

T cells Both antigens stimulated a similar level of activated IFN-γ secreting cells Next, we HLA typed all subjects enrolled in the trial and found that HLA-C* alleles 04 and 05 were uniquely present in this subject In silico analysis using a HLA-peptide binding prediction algorithm showed that HLA-C*04/05 alleles bind 9-mer sequences

in the dominant epitope more strongly than other alleles To evaluate antigen presentation of the A1AT dominant epitope by HLA-C*04/05 alleles, an ex-vivo antigen presentation model was developed by transfecting individual HLA alleles into K562 cells, loading them with the immunodominant peptide and using them as antigen presenting cells Only cells transfected with HLA-C*05 and loaded with the A1AT dominant epitope activated T cells from the subject These data demonstrated that the A1AT-specifi c T cell response observed

in this subject was CD8+ mediated and restricted by HLA C*05 This importance of HLA haplotype in infl uencing the adaptive immune response to vector encoded transgene should be considered in the design of clinical trials

50 Integration Site Analysis in a Clinical Trial

of Lentiviral Vector-Based HSC Gene Therapy for MLD Shows High Levels of Polyclonal Hematopoietic Reconstitution and No Signs of Genotoxicity at One Year after Transplant

Eugenio Montini,1 Alessandra Biffi ,1 Andrea Calabria,1 Martina Cesani,1 Fabrizio Benedicenti,1 Laura Lorioli,1 Tiziana Plati,1

Simone Leo,2 Gianluigi Zanetti,2 Maria Sessa,1 Christof von Kalle,3

Manfred Schmidt,3 Luigi Naldini.1

1 San Raffaele Telethon Institute for Gene Therapy, Milan, Italy; 2 CRS4, Pula, Italy; 3 National Center for Tumor Diseases, Heidelberg, Germany.

A self-inactivating lentiviral vector (LV) has been used in an ongoing HSC-based clinical trial for metachromatic leukodystrophy (MLD) performed in our Institute in Milan High sustained levels of vector marking have been achieved in all 4 treated patients up to now,

Molecular Therapy Volume 20, Supplement 1, May 2012

CLINICAL GENE & CELL THERAPY ORAL ABSTRACT SESSION

reaching up to 80% of the reconstituted multi-lineage hematopoiesis

Vector integration site profi ling in vector-marked cells from patients

of HSC gene therapy enables to assess the safety and effi cacy of gene

transfer as well as the dynamics of hematopoietic reconstitution Here

we report the integration site analysis in different cell lineages and

time points after transplantation in the fi rst 3 MLD patients LAM

PCR was used to retrieve LV/genomic DNA junctions from patients’

transduced CD34+ cells, and from bulk or lineage marker-sorted cell

populations from bone marrow or peripheral blood (1, 3, 6, 9 and 12

months after transplant) A highly polyclonal pattern of bands was

observed in all cell compartments with exception of T-cells at the

earliest time point This is an expected fi nding because the patient

did not receive a T-cell ablative preconditioning treatment By

high-throughput next-generation 454-pyrosequencing of LAM PCR

products we obtained 6x10e6 sequencing reads and mapped >25000

unique integrations from the 3 analyzed patients The LV integration

profi le in this trial is highly reminiscent to those previously described

in the LV-based adrenoleukodystrophy clinical trial and our preclinical

xenotransplantation mouse models, including the occurrence and

identity of Common Insertion Sites (Biffi et al., Blood 2011) CIS

clustering at specifi c mega-base wide chromosomal regions suggests

that these CIS may be originating by an intrinsically benign viral

integration bias No skewing from in vitro to in vivo in the frequency

of CIS was found so far Gene Ontology analysis at different time

points shows constantly similar gene classes such as transcription

factors and chromatin remodeling genes Analysis of sequencing reads

as a surrogate of the abundance of specifi c cell clones shows long-term

variable contribution of multiple clones without evidence of clonal

dominance Importantly, sequencing of the same samples multiple

times provided important information on complexity and clonality

in a dynamic fashion By this approach we show that the number of

integrations shared between different sequencing runs of the same

LAM PCR products is relatively small (about 50%), indicating that

our retrieval and sequencing depth is far from saturation These

fi ndings further indicate that a highly polyclonal reconstitution has

been achieved in this clinical trial Overall, our data show that despite

the unprecedented high vector marking levels obtained in MLD

treated patients no overt evidence of genotoxicity has been observed

EM and AB contributed equally

51 Results of a Phase I Trial of SGT-53: A

Systemically Administered, Tumor-Targeting

Immunoliposome Nanocomplex Incorporating a

Plasmid Encoding wtp53

Neil Senzer,1,2,3 John Nemunaitis,1,2,3 Derek Nemunaitis,1 Cynthia

Bedell,1 Gerald Edelman,1,3 Minal Barve,1,3 Robert Nunan,1

Kathleen F Pirollo,4 Antonina Rait,4 Esther H Chang.4,5

1 Mary Crowley Cancer Research Centers, Dallas; 2 Texas

Oncology, P.A., Dallas; 3 Medical City Dallas Hospital, Dallas;

4 Georgetown University Medical Center, Washington, DC;

5 SynerGene Therapeutics, Inc., Potomac.

The effective oncologic application of gene therapy is predicated on

the effi cient and selective systemic delivery of therapeutic molecules

to both primary and metastatic cancer cells throughout the body

SGT-53 is a therapeutic complex comprised of a DOTAP:DOPE cationic

liposome encapsulating a plasmid encoding normal human wild

type (wt) p53 cDNA and decorated with the anti-transferrin receptor

single chain antibody fragment designed to target cancer cells by

binding to the transferrin receptor (TfR) Eleven patients with solid

tumors, having exhausted all standard and/or approved therapies,

were enrolled in this sequential dose-escalating study ranging from

0.6 mg DNA per infusion through 1.2, 2.4 and 3.6 mg Pertinent

inclusion criteria were ECOG 0-2, measurable biopsy-accessible

disease, willingness to allow skin biopsy, and standard adequate

organ function SGT-53 was administered twice weekly for fi ve

weeks for a total of ten infusions A 48-hour inpatient observation was required for the fi rst infusion and all patients received prophylactic dexamethasone 8 mg i.v 1 hour prior to dosing, H1 and H2 blockers i.v 30 minutes prior, indocin 25 mg, p.o approximately 30 minutes prior to receiving SGT-53, and acetaminophen 650 mg p.o prior and subsequent to SGT-53 for pyretic reactions In patients with biopsies

of tumor ± normal skin, p53 was detected using DNA PCR employing primers specifi c for the exogenous p53 gene Eight of 11 evaluable patients (73%) demonstrated SD at week 6 assessment and three had

PD The tumor in one patient (adenoid cystic carcinoma) underwent signifi cant necrosis and was reclassifi ed from inoperable to operable

In another patient (leiomyosarcoma with metastases in liver and lung), post-treatment CT assessment showed development of necrotic centers in all the liver metastases DNA PCR analysis for exogenous p53 DNA in the tumors (metastases) from three patients and normal skin biopsies showed a tumor specifi c and DNA dose-dependent presence of the exogenous p53 (5+ tumor vs 0 normal skin) One patient experienced a possibly-related grade 3 adverse event (AE), fatigue (at 0.6 mg DNA, concurrent with tumor necrosis) and a second patient possibly-related grade 2 chest pain and tachycardia (at 3.6 mg DNA) SGT-53-related grade 1, 2 AE occurring in ≥5%

of patients primarily consisted of transient fever starting 6-8 hours after infusion and lasting 12-16 hours (5 of 12; 42%) and transient hypotension over a similar time frame in 7 (7 of 12; 58%) This is the

fi rst report of tumor-specifi c delivery to metastatic lesions in patients after systemic administration Integrating the results of the current study with data derived from pre-clinical p53 restoration models, a clinical Phase Ib study of SGT 53 combined with docetaxel, has been activated and is accruing patients

52 Increases in CD4 Counts and Effects on HIV in Aviremic HIV-Infected Subjects Infused with Zinc Finger Nuclease (ZFN) CCR5 Modifi ed Autologous CD4 T-Cells (SB-728-T)

Winson Tang,1 Jay Lalezari,2 Carl June,3 Pablo Tebas,3 Gary Lee,1

Marty Giedlin,1 Shelley Wang,1 David Stein,5 Hiroyu Hatano,4

Elizabeth Sinclair,4 Joseph K Wong,4 Cindy Desmarais,6 Travis Wood,1 Steven G Deeks,4 Geoff M Nichol,1 Ronald Mitsuyasu,7

Dale Ando.1

1 Sangamo BioSciences, Inc, Richmond, CA; 2 Quest Clinical Research, San Francisco, CA; 3 University of Pennsylvania, Philadelphia, PA; 4 UCSF, San Francisco, CA; 5 Albert Einstein College of Medicine, Bronx, NY; 6 Adaptive Biotechnologies, Seattle, WA; 7 UCLA, Los Angeles, CA.

Background: ZFN modification of the CCR5 receptor in

autologous CD4 T-cells may render a survival advantage to these cells

in HIV-infected subjects We report data relating to safety, increases

in CD4 T-cell counts, persistence and traffi cking of SB-728-T, and effects on HIV-RNA (VL) and proviral DNA from two Phase 1

studies of SB-728-T Methods: In the CA study, 9 immunologic

nonresponders (INR) with CD4= 200-500 cells/uL were enrolled into

3 cohorts that received 1x, 2x, and 3x1010 total cells In the PENN study, 6 immunologic responders (IR) with CD4 ≥450 cells/uL and

6 INR with CD4 ≤500 cells/uL were infused with 1010 total cells At

Wk 4, IR underwent a 12-week HAART Treatment Interruption (TI) The median duration of follow-up was 246 days (range 42-561) for all subjects T-cell profi ling was performed with multiplex PCR on 4 subjects to sequence TCR CDR3 chains to determine the composition

of various T-cell clones in the peripheral blood, gut mucosa, and lymph nodes The method amplifi es rearranged TCR CDR3 sequences

to explore all Vbeta and Jbeta combinations Results: SB-728-T was

well tolerated with only one SAE attributed to a transfusion reaction-fever, chills, myalgia, and arthralgia All AEs were reversible and most were mild to moderate in severity The mean CD4 and SB-728-T count increased by 1533 cells/uL (range 216-3025) and 83 cells/uL,