International Society for Cell and Gene Therapy of Cancer 2009 Annual Meeting Held in Cork,Ireland

Introduction

N oriyuki Kasahara (David Geffen School of Medicine, University of California Los Angeles; ISCGT president) gave the introductory talk of the meeting, and opened by noting that this represents the 10th annual meeting of the society since its inaugural meeting held in London, in 1999. Professor Kasahara commented on the appropriateness of the return of the annual meeting to Europe, and welcomed the presence of the majority of the society's past presidents, as well as delegates from many countries where the society has held previous meetings, including Singapore (Farzaneh et al., 2004), China (Guinn et al., 2007), India (Guinn and Mulherkar, 2008), and Japan (http://www.bioscience.org/current/special/tagawa.htm). Professor Kasahara declared that gene therapy has now come of age, and provided a historic overview of the advances and remaining barriers in gene therapy technology, manufacturing, regulations, and clinical milestones during the 10 years of the existence of the society, which he described as “a roller-coaster ride.” Now, however, the first market-approved gene medicines for cancer treatment have reached clinical use in China in 2005 and include Ad-p53 (Gendicine) and H101 (DE1B-55kD conditionally replicating adenovirus [CRAd]), and in fact, China has now surpassed the United States and Europe in terms of numbers of patients treated with gene therapy protocols (Peng, 2005). Equally important, during this past decade, we now have developed a more realistic assessment of what gene therapy can accomplish, and it is important to note that the approved gene therapy medicines work best in conjunction with traditional cancer treatments. With market approval has come a renewed interest in gene therapy as a medicine, including from the commercial sector, and therapeutic efficacy has been observed in at least five clinical settings—adenosine deaminase (ADA) deficiency, X-linked severe combined immunodeficiency (SCID), lipoprotein lipase (LPL) deficiency, critical limb ischemia, and retinitis pigmentosa.

Work from the Kasahara laboratory includes gene transfer, functional genomics/proteomics, cell/bioengineering tools, and transplantation/regenerative medicine, in addition to their main focus on oncolytic virotherapy. The Kasahara group uses gene transfer technology to bioengineer cells, primarily for immune targeting of cancer cells, and Professor Kasahara also works in association with collaborators on the engineering of acute myeloid leukemia (AML) cells with costimulatory genes (B7-1) and cytokines (granulocyte-macrophage colony-stimulating factor [GM-CSF] or interleukin [IL]-2). This expertise on attenuation of host immune responses is also being applied in his laboratory to transplantation/regenerative medicine, by using the same mechanisms that tumors use to tolerize the immune system, to avoid immune rejection of allogeneic transplants. In this arena, their work focuses on “conditioning graft tissue” so that it is tolerated by the immune system and involves lentiviral (LV) delivery of short hairpin RNA (shRNA) against HLA to the tissue graft cells. A dose-dependent knockdown of HLA expression in human CD34⁺ cells was observed, with complete protection against HLA-directed cell killing in vitro at transduction efficiencies of 75%, indicating the potential for this strategy of graft treatment before transplantation into the host. This technology could be further refined by use of small interfering RNA (siRNA) to specifically target only mismatched allele sequences in parent-to-offspring transplants. Professor Kasahara cited potential targets for this technology in the near future in HLA-restricted cell transplants such as bone marrow (BM) transplant/hematopoietic stem cells (HSCs), pancreatic islet cells, and skin grafts, with future possibilities for solid organ transplants if transduction efficiencies can be improved, as well the possibility of generating “universal” SCs for regenerative medicine.

Professor Kasahara then described his laboratory's work on oncolytic virotherapy. One of the main problems with applying gene therapy in the clinic is the low efficiency of gene delivery to the majority of cells in solid tumors. Human clinical trials of glioblastoma show <1% (often <0.1%) transduction efficiency by nonreplicating viruses, despite prior debulking and multiple injections, with transduction limited to the area surrounding the needle track (Lang et al., 2003). Kasahara and collaborators have sought to overcome this drawback through the use of tumor-specific replication-competent viruses, so that each tumor cell becomes a source of more vector. Naturally cytolytic viruses have shown clinical efficacy; however, these large complex viruses are often difficult to manipulate; their basis for attenuation/selectivity is not well understood; a robust immune response often results in rapid clearance of virus, and the lack of persistence leads to recurrence of tumor. Murine leukemia virus (MLV) is a stable integrating retrovirus, a simple virus that is well understood. It fails to infect unless the cell is proliferating, providing a level of selectivity, and it is not intrinsically cytolytic. It shows stable integration, acting as a benign parasite, and it is less immunogenic than many of the other oncolytic viruses, and antiretroviral drugs are available. It can be modified to carry suicide genes, which convert prodrugs to cytotoxins and thereby induce death. Use of a green fluorescent protein (GFP)-labeled replication-competent retroviral (RCR) vector, ACE-GFP, showed spread of this virus over time in a multifocal CT26 tumor in liver after locoregional delivery, with a sharp demarcation between normal and tumor tissue. Effective tumor killing was achieved in murine in vivo models with RCR-mediated delivery of cytosine deaminase (CD) suicide gene/5-fluorocytosine (5-FC) prodrug therapy, with a single administration of vector followed by multiple injections of prodrug. Administration of multiple cycles of prodrug further improves survival and suggests persistence of RCR-CD in metastatic intracranial glioma cells (Tai et al., 2005). This approach is about to enter a phase I/II dose escalation clinical trial for glioblastoma multiforme.

Cancer Molecular Targets

Manfred Dietel (Institut für Pathologie, Universitätsklinikum Charité der Humboldt-Universität zu Berlin, Germany) described biomarkers as being diagnostic, prognostic, or predictive. This latter designation is the most important as it can predict how a tumor will behave in the body of the patient before treatment begins. Predictive biomarkers indicate whether a treatment will be effective and provides a new concept in oncology-targeted therapy. The patient used to be given a nonspecific treatment such as cisplatin for ovarian cancer, and although as many as 60–70% respond, many have side effects. Medics cannot predict response, which would have helped avoid treating the patients who develop side effects but receive no clinical benefit. The basic idea is to target treatments to patients who will respond, which is important for both patients and clinicians. In the last 10–15 years scientists have detected a number of clinically effective and relevant biomarkers for breast cancer, including Her2, which have indicated which patients should receive trastuzumab (Herceptin). In 70% of all malignancies we are now on the way to finding predictive biomarkers. Trastuzumab targets metastatic breast cancer. Gefitinib and erlotinib (anti-Her2 antibodies) target non-small cell lung cancer (NSCLC) with mutated epidermal growth factor receptor (EGFR). In future the U.S. Food and Drug Administration (FDA) is likely to approve only drugs that target known biomarkers with predictive indications.

Current predictions of tumor response are tissue based and this is one of the challenges to histopathologists. Anti-EGFR therapy involves the following:

Examination of tumor tissue for the expression of EGFR/Her2 by immunocytochemistry/immunohisto-chemistry

NSCLC patient survival is less than 1 year and the currently available treatment is not effective and always toxic. At present the FDA will approve a new drug only if it is more than 50–60% effective. Gefitinib (Iressa) is being developed by AstraZeneca (London, UK) and erlotinib (Tarceva) is being developed by Hoffmann-La Roche (Basel, Switzerland); both are in clinical trials. They are both EGFR-targeting kinase inhibitors. Again, an eligibility test is being used. If erlotinib is given to all patients the response rate is 10–15% in Europe whereas the response rate in Asia is ≤50%. If the patient has a mutation in exon 19 or 21 the efficacy of the drug is much higher. At present anti-EGFR treatment is approved only if patients have an activating mutation that must be shown in tests before treatment. Treatment is forbidden in the absence of either mutation. This has saved insurance companies considerable money and patients unnecessary treatment with a drug when it is predicted not to work for them. The basis of the effect is that the catalytic kinase domain is activated by several types of mutation. No mutation is ineffective. In terms of frequencies, 5–10% of mutations occur in exon 18, 45% in exon 19, and 40–45% in exon 21, and these are now evaluated before the treatment of every NSCLC patient. The test is ISH based and there is a notable difference between survival rates among patients with low and high copy numbers. Only tumor is tested for mutations; biopsies need to be microdissected to ensure that only tumor tissue is tested, otherwise this material can lead to false results because the tissue excised is mostly normal.

Patients can also be selected on the basis of the presence of KRAS mutations if they have invasive colon cancer; this exclusion of KRAS wild-type (WT) patients saves €325,000 per year in therapy costs. This again demonstrates how important it is to exclude patients from receiving drugs that will not be effective for them, but will give them side effects, and also shows the need to select tumor cells for pretesting that indicates which one of the two monoclonal antibodies (mAbs) should be given to patients. If the routine laboratory is to do all these tests, large clinical trials are required (Gutman and Kessler, 2006). Professor Dietel said that in future a typical diagnosis will include mutational analysis and sequencing to predict tumor response to provide a solid basis for targeted therapy.

Jim Norris (Medical University of South Carolina, Charleston, SC) explained that there have been reductions in the death rates per capita for heart disease, cerebrovascular disease, and infections such as pneumonia/influenza, from 1950 to 2001, but that sadly the same is not true for cancer rates, which have stayed almost the same. The frequencies have improved slightly since 2001, but the statistics suggest we (scientists/clinicians) need to do better. p53 is a tumor suppressor gene that regulates hundreds of genes in a number of pathways in the cell including senescence, cell cycle, angiogenesis, apoptosis, and tumor immunity. The group led by John Nemunaitis has worked to develop Advexin (Ad-p53), which is equivalent to the Ad-p53 used in clinical trials in China, with the aim that the WT p53 in the Ad-p53 vaccine would replace the mutated p53 in cancer cells. WT p53 is normally a tetrameric protein that binds DNA. Mutated p53 tetramers are inactive and typically do not bind DNA but block WT p53 DNA binding and thus prevent WT p53 from performing its normal function. Advexin has been shown to be safe in more than 600 treated patients, with no notable side effects or adverse events, suggesting this treatment would improve the quality of life for patients. Regarding p53 expression there are four categories of patients: those with (1) low WT p53 expression, (2) high WT p53 expression, (3) low mutated p53 expression, and (4) high mutated p53 expression. It is in this latter group that Advexin therapy exhibits an unfavorable outcome. Of the patients that fall into groups 1–3, 86% have a favorable response when treated with Advexin, whereas only 25% of the patients in the unfavorable group do. Patient with favorable p53 status had highly significant improvements in survival. Median survival over 48 months was 7.2 months in the 40 favorable patients whereas in the 13 unfavorable patients median survival was 2.8 months (log rank p = 0.0103). A clinical trial is now ongoing in China that will enlist more than 1000 patients.

All treatment approaches including chemotherapy, radiation, and gene therapy target sphingolipid metabolism. Ceramide is an important signaling molecule in the late stage of the execution process of cell death after treatment. Acid ceramidase (AC) is upregulated in head and neck cancer (70%) and prostate cancer (60%) (Norris et al., 2006). Overexpression of AC shifts the equilibrium between ceramide toward formation of sphingosine and sphingosine 1-phosphate, the latter of which prevents or decreases apoptosis. The effects of increasing AC can be reversed by an inhibitor called LCL204. Cells that overexpress AC are more resistant to chemotherapy drugs than those that do not, including the drugs doxorubicin, taxol, gemcitabine, and etoposide. AC silencing by siRNA sensitizes AC–EGFP cells to these drugs. Genes that are upregulated by AC expression include the eukaryotic translation initiation factor-4E (eIF4E) and proliferating cell nuclear antigen (PCNA). eIF4E is an interesting oncogene with two functions. One is to move a subset of RNAs from the nucleus to the ribosome; RNAs that are transported are associated with growth and include IL-2 and IL-8.

Barbara Guinn (University of Southampton School of Medicine, Southampton, UK) described their work on AML antigen identification. Survival rates have improved significantly over the last 30 years for AML patients less than 65 years of age, because of improvements in supportive care and the development of HSC transplants; but for patients more than 65 years of age, treatment options are limited and 5-year survival rates are about 11%. Many AML patients achieve a first remission, when tumor loads are low, and this may be a promising time point at which immunotherapy could be used to delay or even prevent relapse. When the Guinn group started looking for antigens in AML in 1999, few immunotherapy targets had been identified. They used the SEREX technique, an acronym for the serological analysis of recombinant cDNA expression libraries (Sahin et al., 1995), which benefits from having few limitations regarding starting material, one-third of all SEREX-identified genes are novel and SEREX identifies tumor-associated antigens (TAAs) that induce both humoral and T cell-mediated responses. To try to maximize their chances of finding a cancer-specific or cancer-testis (CT) antigen the group immunoscreened a normal donor testis cDNA library with pooled sera from five AML patients with the M4 and M5 subtypes of AML and no detectable cytogenetic abnormalities (Guinn et al., 2005). Through two rounds of immunoscreening, they identified 44 mostly known antigens including MYH11, RHAMM, MAPK10, BRAP, and TRIP11, which have previously been shown to be involved in AML, and NAB2, NY-BR-1, and ACVR2, which have been shown to be involved in solid tumors. To determine which antigens to investigate further the group immunoscreened the antigens with sera from 16 additional AML patients, 22 chronic myeloid leukemia (CML) patients, and 40 age- and sex-matched normal donors. They identified eight antigens that were recognized only by patient sera, including SPAG9 and NY-BR-1, and eight antigens that were preferentially recognized by patient sera, including BRD1 and NAB2. Reverse transcription-polymerase chain reaction (RT-PCR) analysis of the eight antigens recognized only by patient sera and two antigens preferentially recognized by patient sera (SSX2IP and BRD1) identified three antigens (GRINL1A, SSX2IP, and PASD1) that were expressed only in cell lines and patient samples and not normal donor samples. The Guinn group and others have now shown that PASD1 has restricted expression in normal tissues and characteristics typical of CT antigens. The Guinn group has identified a number of HLA-A2-binding epitopes within PASD1, using the BIMAS and SYPEITHI algorithms, but none of the wild-type epitopes bound with any notable frequency to HLA-A2 on T2 cells. Therefore the Guinn group modified one of the anchor residues and was able to show that these modified peptides could induce T cell responses (in terms of interferon [IFN]-γ production and pentamer staining) from normal donors and AML patients. They used the pDOM-epitope vaccine design (Rice et al., 2008) to prepare PASD1 epitope vaccines and showed that when HLA-A2 transgenic mice were immunized with the pDOM.PASD1 epitope vaccine, T cells could respond to both the modified and WT epitopes in enzyme-linked immunospot (ELISpot) and cytotoxic T lymphocyte (CTL) assays. Dr. Guinn also mentioned their work on the peptide–MHC (pMHC) tetramer array (Soen et al., 2003). One major advantage of the technique is the ability to analyze multiple T cell populations, using a relatively small number of CD8⁺ T cells (92 × 10⁶ cells per array). In addition, pMHC arrays use small amounts of pMHC for each spot (1 ng) and can display a large number of pMHC spots (91,000) within each array without haplotype restriction. The group has assessed a range of slide surfaces and fluorochromes, and has enhanced the sensitivity and specificity of the pMHC array, thereby increasing its use as a diagnostic tool for leukemias. CD8⁺ T cells were negatively purified from patients presenting with acute lymphocytic leukemia, chronic myeloid leukemia, and acute myeloid leukemia (n = 11) and incubated with hydrogel slides printed with commonly associated tumor antigen pMHCs, together with viral pMHCs (cytomegalovirus [CMV], influenza A, and Epstein–Barr virus [EBV]). They found in the peripheral blood of presentation AML patients T cells that recognized AML-associated tumor antigens despite not having undergone ex vivo expansion, suggesting the pMHC array may provide a useful end-point assay in future clinical trials.

Kam Man Hui (National Cancer Centre, Singapore) examined patients who had been treated by curative surgery, which accounts for 20% of hepatocellular carcinoma (HCC) patients, to determine whether the disease recurs. Global gene expression profiling of human cancer identified more than 100 genes that are important predictors of outcome and now these genes are simultaneously analyzed by quantitative real-time polymerase chain reaction (Q-PCR). The group is now increasing the number of genes analyzed to 10,000 with new funding. HCC is the third most lethal cancer worldwide and the fifth most common disease in the world. It is often diagnosed late, which is the reason it accounts for 10% of all cancer deaths worldwide. In this cancer, like most, it is important that patients be diagnosed early. When analyzing samples the group also receives adjacent matched normal tissue and uses normal tissue from normal donors as additional controls. Often adjacent normal tissue shows signs of having undergone cirrhosis and is not normal at all. The group collects details of additional clinicopathological features at diagnosis including age, sex, history of viral infections (hepatitis B virus [HBV], hepatitis C virus [HCV], and non-B, non-C), capsulation, tumor size, α-fetoprotein (AFP) level (median nanograms per milliliter), lesions, histological grading (G1, G2, G3, G2–G4), cirrhosis, and invasion. Most patients in Singapore stay with the same doctor throughout treatment, which aids in the collection of data. Using microarray analysis the Hui group identified 57 genes whose expression predicts relapse in recurrent HCC. These genes provide 100% prediction of recurrence and also many potential biomarkers. Two particularly interesting genes were RACGAP1 and LECT2, which perform different and opposing functions in the oncogenesis of HCC. RACGAP1 expression is significantly higher in patients with recurrent HCC; however, this is not 100% as some patients have high expression but do not undergo disease recurrence. RACGAP1 expression correlates with the migration property of HCC cells. The Hui group showed that the ability of cells to metastasize could be significantly inhibited with siRACGAP1. It also suppresses Cdc42, Racl, and Rho activation. Suppression of RACGAP1 with siRNA occurs because of the activation of caspases leading to apoptosis. LECT2 is significantly decreased in invasive HCC. Normally it is expressed in fetal liver and when the Hui group transfected cells with Ad-Lect-2 these cells take longer to form a monolayer. Overexpression of LECT2 in tumor cells reduces the migration of tumor cells and their ability to invade. The group has now developed huHCC xenograft models for gene expression studies so that they can test drugs for therapeutic use.

The laboratory of Ann Leen (Baylor College of Medicine, Houston, TX) is interested in using T cells to target EBV-negative Hodgkin's lymphoma. Adoptive T cell transfer has proven successful for the treatment of EBV-positive tumors; however, 80% of patients referred to their center with relapsed or refractory lymphoma have EBV-negative tumors. Thus, Dr. Leen's group has been investigating alternative T cell target antigens. To date they have examined more than 100 biopsies to map the TAA expression profile of EBV-negative Hodgkin's lymphoma. They found that 40% of the tumors were MAGE (melanoma-associated antigen) positive, and 73% were survivin positive. Further, they found that SSX2 expression was present at baseline but could be increased with demethylating agents. To generate CTL lines targeting these TAAs, the Leen group used monocyte-derived dendritic cells (DCs) as antigen-presenting cells (APCs) either pulsed with protein-spanning mixtures of overlapping peptides (pepmixes) or nucleofected with antigen-encoding DNA plasmids as a source of antigen, and the lines were initiated in the presence of a cytokine cocktail. The group was concerned about targeting only one antigen, as the remaining tumor cells can downregulate expression as a means of immune evasion, and therefore they used combinations of antigens to generate multi-TAA-targeting CTLs (Leen et al., 2006, 2009). When the group tested the functional capacity of the CTLs they demonstrated that the lines were able to produce IFN-γ after antigenic stimulation, and were able to recognize and kill antigen-expressing target cells in a cytotoxicity assay. Finally, the group showed that this technology was robust and could be extended to the generation of multi-TAA CTLs targeting leukemia-expressed antigens, such as Wilm's tumor-1 (WT1), PRAME (preferentially expressed antigen of melanoma), and survivin.

Brian Lichty (McMaster University, Hamilton, ON, Canada) talked about their work boosting tumor vaccine efficacy using oncolytic viruses. Viral oncolysis can be aided by antitumor immune responses. By combining the two, the patient receives the benefits of both approaches. Vesicular stomatitis virus (VSV) is sensitive to IFN and most tumors are hyporesponsive to IFN. The Lichty group has been working with murine intracranial B16F10 tumors, which are aggressive. They initially treated mice with VSV–GFP to debulk tumor cells, without success. They then started to vaccinate mice with the human transgene DCT, which is a melanoma-associated antigen. They showed that the mice respond to DCT after they have been treated with the vaccine even when they have established tumors. The group noted they get tumor debulking. However, when they delivered the transgene in an oncolytic virus, antitumor effects were transient with no improvement over oncolytic virus lacking the gene. They engraft 10³ B16 cells and on day 7 vaccinate the mice with Ad hDCT intramuscularly. On day 21 they give the mice 10⁹ VSV-hDCT intravenously and on day 27 perform immune analysis. Throughout this treatment the Lichty group showed they could shift immune dominance from the virus to the TAA transgene, and now the survival curve is at 20% in a mouse model with preexisting melanoma in the brain. The group has found that the anti-TAA response is largest in tumor-bearing mice as the boosting oncolytic vaccine vector amplifies in the tumor, where they found peak frequencies of DCT_180–186-specific CD8 tumor-infiltrating lymphocytes (TILs). Therefore they have turned a tumor antigen into a viral antigen, increasing the response in terms of TILs, and this translates into a therapeutic effect. They also found a spontaneous response to gp100 when they boost with an oncolytic vector and successful therapy leads to extensive vitiligo, hence the TAA should be carefully chosen as the response induced was significant as shown by vitiligo data.

Intersection of Cell and Molecular Therapy with Immunotherapy

Nagy Habib (Imperial College, London, UK) described their research with SCs, which are found in the brain, liver, and BM. The HSC is a single cell with CD34 expression. Only 2% of these cells stick to plastic and these are called omnicells or omnicytes, which are present in all of us. These cells can divide from day 10 to provide a 5-log increase in cell numbers but are self-limiting and express CD34, CD45, and various other markers including HoxB4, Sox2, Oct-4, and Nanog (the master regulator of all the SC markers; its expression indicates how differentiated the cell is). These cells can be manipulated to increase the “stemness” of the cells. When young and in the early stages of division omnicytes are highly stem cell-like in nature, but they lose stemness as they differentiate. The Habib group will start a clinical trial on October 1, 2009 for patients with myocardial infarction. Some SCs produce insulin if cultured with the right growth factors. In a SCID rat model of diabetes, the group injected SCs into the pancreas via the tail vein and observed decreases in blood sugar that were maintained for at least 6 months on average. The group has now taken cells from diabetic patients, induced them to produce insulin, and given them back to the patient. They observed a decrease in the amount of insulin required. The third patient started on the clinical trial on the day Professor Habib gave his talk at the ISCGT meeting. The SCs used could be of neuronal or microglial derivation and could protect the brain from cancer. In stroke patients a small area of the brain is dead but a much larger area is penumbral, that is, hovering between life and death. If the doctor does nothing then this area dies as well. Therefore patients who have had a massive stroke are given omnicytes to limit damage to their brain function. When patients have lost one-quarter of their brain, 95% will die or have a major disability. SC injections can be given into the carotid artery and these patients become part of the 5% of poststroke patients who are back to work within 6 months of a stroke. The results so far are encouraging but the Habib group now must determine what happens with the next patients. With rat liver cells the group adds toxin and the cells detach, round up, and die. If omnicytes and toxin are added, the cells stay alive. This suggests that there is something in the omnicytes that stops the insult from causing cell death. A factor secreted through a semipermeable membrane stops apoptosis, and cells can be induced to secrete various things including insulin, as already described for antifetoprotein. One patient on the liver transplant waiting list had biliary cirrhosis and after omnicyte transplantation the doctors could see the liver ascites disappear. In addition, the patient's albumin increased and bilirubin decreased; however, this treatment may not work with all patients. The Habib group treated a 6-year-old female with a metabolic gene defect by giving her cells from her father and could see correction of the glycogen storage disease, but the real clinical challenge will be how to increase the number of cells available for therapy without losing their healing properties, that is, the future aims will be to cause the proliferation and expansion of omnicytes while maintaining their stemness, and also how to do it quickly in acute cases.

Karin M.L. Gaensler (University of California San Francisco, San Francisco, CA) talked about establishing neonatal murine models as tools for exploring strategies for the induction of tolerance. Using adeno-associated virus (AAV)-mediated gene delivery in a neonatal model, they were able to show induction of tolerance to human factor IX (hFIX). Higher levels were detected after administration of AAV8-hFIX than AAV1-hFIX. The Gaensler group also combined transplantation during the neonatal period with a positive selection strategy for in vivo amplification of drug-resistant donor HSCs. BM-derived HSCs were transduced ex vivo by LV-mediated gene transfer of the DNA repair enzyme P140K-O ⁶-methylguanine-methyltransferase (P140-MGMT). The P140-MGMT DNA repair enzyme confers resistance to benzylguanine, an inhibitor of endogenous MGMT, and to chloroethylating agents such as BCNU. Enrichment of donor SC levels was achieved by successive cycles of low-dose, in vivo chemoselection. The group suggests that these approaches could enable further evaluation of mechanisms of tolerance induction during immune ontogeny.

Yasufumi Kaneda (Division of Gene Therapy Science, Graduate School of Medicine, Osaka University, Osaka, Japan) described the development of anticancer strategies using the hemagglutinating virus of Japan (HVJ) envelope. This virus was isolated in Japan in the 1950s. It is a mouse parainfluenza virus that binds to glycoproteins and glycolipids containing sialic acid and has membrane fusion activity, producing abundant viral proteins from the RNA genome. The researchers used inactive HVJ particles and by incorporating plasmid DNA they created the HVJ-E vector, allowing the direct entry of genes into the cell cytoplasm, and then to the nucleus. siRNA can also be delivered at high efficiency by HVJ-E to cells either in vitro or in vivo. HVJ-E/Eg5 siRNA affects only growth-promoting cells during the separation of centrioles and formation of mitotic spindles. They investigated the treatment of glioblastoma using HVJ-E/Eg5 siRNA and cell survival was greatly reduced. The centrioles did not separate and mitotic spindles did not form. They eradicated intracranial U-118 mouse glioblastoma tumors by a single stereotactic injection of HVJ-E/Eg5 siRNA. After 3 months, 80% of the mice had survived and no tumors were seen in the brain. They also injected tHVJ-E/Eg5 siRNA into normal brain tissue and no apoptosis or tissue damage was observed. They also found that HVJ-E itself has antitumor activity when they injected the vector into CT26 tumors. Sixty percent of the mice became tumor free after three injections of the empty vector. The surviving mice were rechallenged with the same CT26 tumors 10 months later and all mice rejected the tumors. Similar suppression was observed in renal cancer tumors. Their finding suggested that natural killer (NK) cells can be activated by HVJ-E, through the secretion of chemokines and cytokines by DCs.

The Kaneda group incubated cells treated with and without HVJ-E plus anti-IL-6 antibody and showed attenuation of HVJ-E activity by IL-6, the secretion of which from DCs has been shown to be able to overcome suppression by regulatory T cells (Tregs) (Kurooka and Kaneda, 2007). HVJ-E stimulates IL-6 production by BM-derived DCs. In tumor-bearing mice that have been treated with HVJ-E, there was increased infiltration of IL-6⁺ DCs in lymph nodes (LNs) and tumoral tissue, suggesting that in vivo suppression of Tregs is probably mediated by IL-6 production. Antitumor immunity by HVJ-E is seen to be related to NK and CTL activation, and Treg suppression. Suppression of tumor metastases occurred in CT26 tumor-bearing mice with bilateral flank tumors. When treating only right-side tumors, tumors on both flanks regressed in 80% of mice. Similar results were obtained in treating B16/BL6 subcutaneous melanoma tumors, inhibiting the development of lung metastases. HVJ-E was also found to induce direct killing in hormone-resistant prostate cancer, by preferentially fusing with hormone-resistant cells. One intratumoral injection of HVJ-E suppressed orthotopic human prostate cancer in SCID mice.

Immunotherapy of Cancer

Masatoshi Tagawa (Chiba Cancer Center, Chiba, Japan) described the biological activities of a new type of IFN, which belongs to a type III family of IFNs and has a subset of sensitive cells different from that of IFN-α/β, type I IFN. IFN-λ induced G₁ arrest in TE-11 human esophageal carcinoma cells with upregulated p21 and dephosphorylated retinoblastoma protein. Two other sensitive lines, YES-5 and T. Tn, became apoptotic with cleavage of caspase-3 and poly(ADP-ribose) polymerase (PARP). The IFN-λ receptor complexes, IL-28Rα and IL-10Rβ, are not ubiquitously expressed as are those for IFN-α/β. Expression is not observed in normal fibroblasts or other tumor cells but is expressed in immortalized esophageal cells (Het-1A cells), suggesting tissue-specific expression. In contrast, IFN-α and IFN-β are suppressive to normal cells. Cells treated with a combination of 5-fluorouracil (FU) and IFN-λ produced additive growth suppression effects on the esophageal carcinoma cells but not on normal cells including Het-1A cells. Ads expressing the IFN-λ gene (Ad/IFN-λ) induced growth inhibition of esophageal cells and led to increases in the sub-G₁ fraction after infection of IFN-λ-sensitive cells, which were accompanied by cleavage of caspase-3 and PARP. Regarding apoptosis, caspase-8 and −9, as well as Bcl-2, were involved, whereas Bax was less likely to be responsible. The forced expression of IFN-λ in tumors inhibited their tumorigenesis. Unique receptor expression showed a different utility of IFN-λ in other systems. P6X fibroblasts do not express the IFN-λ receptors and consequently are insensitive to IFN-λ. When P6X cells were infected with Ad/IFN-λ and cocultured with YES-2 tumor cells, YES-2 cells became apoptotic but P6X cells did not. The cell-mediated delivery of IFN-λ suppressed YES-2 tumor growth in vivo. Further analyses of IFN-λ-mediated antitumor effects suggested that NK activity and antiangiogenesis were not involved. In summary, IFN-λ produces additive inhibitory effects when combined with chemotherapeutic agents, and administration of fibroblasts expressing IFN-λ is a possible strategy for cancer treatment.

Bao-En Shan (Hebei Cancer Institute, Shijiazhuang, China) described the antitumor activity and immunoregulation mechanisms of IL-27 and IL-23 in MA891 cells. The IL-12 family includes IL-23, IL-27, and IL-35 molecules, which are made up of two subunits of p35, p19, and/or p28. These cytokines have similar functions—they induce innate and adaptive immune responses and are produced primarily by activated APCs, although they exert distinct biological effects on helper T cell type 1 (Th1) and Th17 memory T cells and Treg differentiation. IL-35 is produced by Tregs and as an antiinflammatory cytokine it can expand CD4⁺CD25⁺ Tregs. The Shan group placed IL-23 and IL-27 genes into retroviruses and inserted these genes into the mouse cancer cell line MA891. They found no difference in the morphology of transduced cells after IL-27 or IL-23 production, or changes in the expression of MHC class I, class II, CD70, or CD86. However, the cells were able to produce many picograms of IL-27 per milliliter compared with IL-23. Expression of either IL-23 or IL-27 induced the same amount of CTL activity while in vivo, the MA891/IL-27 cells had decreased tumor volume compared with vector control groups, while MA/891/LXSn tumors were obvious on day 40 but no differences in growth kinetics were apparent in the early days of the experiment. Mice immunized with MA891/IL-23 could reject subsequent MA891 challenge but not H22 challenge, whereas mice injected with vector control all died before the challenge time point. IL-27 and IL-23 can upregulate MHC class I and II and the Shan group observed an increase in the expression of TNF-β, IL-12Rβ2, TNF-α, IL-1β, and IL-6 in IL-27-transduced MA891 cells. IL-27 promotes NK, CD4⁺, and CD8⁺ cell responses, and the Shan group observed infiltration of these cells into the tumor tissue. IL-27 and IL-23 did not induce apoptosis but in vivo the group saw more death of MA891 cells. IL-23 induced more apoptosis than did IL-27. Survivin expression is lower in MA891/IL-23 cells compared with vector control and parental cells after subcutaneous inoculation. There is an observed reduction in Fas and survivin, and reduced vascular endothelial growth factor (VEGF) in the tumor tissue of mice immunized with MA891/IL-27 and IL-23 compared with vector controls and parental cells. Day 10 postinjection there is an increase in nitric oxide (NO) production compared with parental and vector controls. The group also observed antitumor activity, but this was lower when injection was done intraperitoneally rather than by other routes such as subcutaneously or via the tail vein. IL-23 expression by the tumor cells increased TNF-α and NO levels, increased the cytotoxicity of macrophages, and greatly increased survival. The secretion of IL-23 or IL-27 cytokine by genetically modified tumor cells induced apoptosis and decreased vessel density within the tumor tissue; however, IL-23 enables tumors to completely disappear from tumor-bearing mice and induces memory T cells.

Helen Tayton-Martin (Adaptimmune, Oxford, UK) explained that the company focuses on T cell adoptive therapy, originating from the early research of its chief scientific officer Bent Jakobsen, in the engineering of soluble T cell receptors (TCRs) (Reid et al., 1996). Avidex was the name of the company initially set up in 1999 as an Oxford spinoff whose aim was to test soluble TCRs as therapeutics in their own right. Conventional wisdom indicates that in any individual there are ∼25 × 10⁶ different circulating T cells guided by their different TCRs to kill other cells, with autoreactive T lymphocytes removed by thymic deletion in early development. This enables “self-antigen”-expressing tumor cells to avoid T cell-mediated killing by “tolerized” T cells expressing natural affinity TCRs to these self-antigens. The aim of the group has been to engineer enhanced affinity TCRs to overcome this tolerance, through one of two methods: (1) transfection of affinity-enhanced TCRs into T cells and adoptive transfer to patients or (2) use of the high-affinity soluble TCR protein as a targeting molecule bringing in a toxic payload to kill cancer cells. Both approaches rely on the same fundamental approach to generate the high-affinity TCR at the outset. The starting point is a T cell clone specific for the WT TCR, from which the TCR genes are isolated and used to generate soluble TCR proteins in Escherichia coli (Boulter et al., 2003). The group performs Biacore binding experiments to measure on and off rates of the WT TCR. TCRs bind MHC-peptide through CDR loops (Li et al., 2005) and a phage display is used to generate mutant TCR gene libraries that present the mutant TCRs on the surface of infected E. coli cells, allowing competitive ELISA binding through limiting dilution to the peptide MHC antigen to identify higher affinity binding TCR mutants with Biacore measured affinities from nanomolar to picomolar sensitivity, a process referred to as “molecular evolution.” Higher affinity mutant TCR chains can be combined to pull out the combination with the highest affinity modified TCR (mTCR). Examples of what has been developed so far include NY-ESO-1 with off rates extended from 7 sec to 19 hr; telomerase, for which the off rate has been increased from 3 sec to 39 hr; and melanA, for which the off rate is now 67 hr. TCR recognition of escape variants is also conserved for the group's enhanced affinity HIV TCR specific to the HIV-1 Gag epitope SL9 (SLNTVATL) now in clinical development for the adoptive T cell therapy of HIV. Initial in vitro work demonstrated that TCR-transduced cells clear HIV infection in vitro. The group initially compared micromolar with picomolar affinity TCRs, but picomolar affinity TCRs cross-react with both HLA-A2-positive and HLA-A2-negative antigen-positive cells. There is an obvious loss of specificity at this level of sensitivity. However, micromolar TCRs activate T cells specifically against HLA-A2 antigen-positive cells, indicating a loss of the autoreactivity with HLA-A2-negative cells. It appears, therefore, that TCRs with a small enhancement of specificity remain antigen specific. On the basis of these observations the group has aimed for moderate enhancement of TCR recognition in the 100–10 μM range for its cancer program. The HLA-A1-restricted MAGE-A3-specific TCR is now a lead candidate targeting a cancer testis antigen (CTAg) present in multiple cancer types and present only at low levels in normal testis tissue. The group has also analyzed a panel of different affinity mutations and found by tetramer staining that they show reasonable transduction and expression. There are a range of different tumor types (cell lines) that can be recognized by and blocked with soluble TCRs. As they increased affinity they started to see a rise in nonspecificity that could not be blocked with soluble TCRs. In vivo work is ongoing to establish efficacy data for the optimal TCR candidate for clinical development. Further TCRs in the pipeline recognize gp100, NY-ESO-1, telomerase, WT1 (HLA-24), MAGE-A3 (HLA-A2), prostate stem cell antigen (PSCA), survivin, and Her2/neu.

Yajun Guo (International Joint Cancer Institute, China and president-elect, ISCGT) talked about his group's work constructing chimeric TCR containing anti-Her2 single-chain variable fragment (scFv) in an LV. T cells are preactivated with anti-CD3, anti-CD28, and IL-2. The group used the mouse breast cancer models D2F2 and 4TI, which are both Her2^–; the human SK-Br-3 breast cancer cell line, which is Her2⁺; and the MCF-7 breast cancer cell line, which is Her 2^–. T cells that were transduced with an scFv-CD28-ζ construct could kill SK-BR-3 and D2F2/E2 (transduced with human Her2) but not the Her2^– MCF7 or D2F2 cell line. The group observed antigen-specific cytotoxicity and cytokine production was triggered by scFv-CD28. Inhibition of tumor growth was observed after transfer of scFv-CD28-ζ-modified T cells. The transduced T cells killed only Her2⁺ cells and therefore they killed D2F2/E2 but not parental D2F2 cells. Mice treated with scFv-CD28-ζ-modified T cells survive rechallenge with D2Fd or D2F2 cells with a frequency of 70 or 100% respectively, at 100 days. However, all controls that did not receive scFv-CD28-ζ T cells died. The group has shown that the memory response, that causes rejection of tumor cells at challenge, involves host T cells because no adoptively transferred T cells remain long term. The response has also been shown to involve both CD4⁺ and CD8⁺ cells. The group used T cells from WT1, pfp^–/–, or FasL^–/– mice but showed that only the WT1 TCR-modified T cells work. The group showed that the killing of tumor cells in vitro by modified T cells requires both IFN-γ and perforin and other cytokines such as TFN-α. Cytokines such as IFN-γ and TNF-α play key roles in the induction of protective and curative immunity.

The second strategy described by the Guo group used TAA-activated DC vaccines. The group constructed dimeric receptors containing anti-Her2 scFv. The constructs were either (1) scFv-V_H-link-V_L-Myc-CD8-CD40 or (2) scFv-CD40-V_H-link-V_L-myc-CD8-TM-cytoplasmic CD40, and these constructs were inserted into an AdEasy vector that was transfected into murine DCs. The scFv-CD40 chimeric receptor induced NF-κB-dependent DC activation after ErbB2 stimulation. Costimulatory molecule expression and cytokine release after stimulation with Her2-positive tumor cell stimulation was observed. The Guo group showed that the vaccine was functional with scFv-CD40 DCs, causing 60% cell death compared with 10–15% of DCs transfected with DF2 alone. The group could show scFv-CD40 DCs in the tumor and draining LNs, whereas scFv-CD40 was found only in the tumor and not in the draining LNs. They also found D2F2/E2 cells transduced with scFv-CD40 in the tumor but not D2F2 cells transduced with scFv-CD40. They also showed that the antitumor immunity induced by these DCs is mediated by CD4⁺ and CD8⁺ cells and that specific CTLs could be induced in vitro.

Farzin Farzaneh (King's College London, London, UK) described work from his group (Hardwick et al., 2009; Ingram et al., 2009) showing NK cell stimulation by IL-2- and CD80/B7.1-expressing AML cells. U937 AML cells were modified to express IL-2 and B7.1 or both and were shown to stimulate T cells, through an increase in total T cell numbers, expression of IFN-γ, and attraction of NK cells. After initial stimulation with U937/IL-2/B7-1, the T cells respond to both K562 and U937 myeloid cell lines, but not to the Raji B cell line. NK cells activated by IL-2/CD80-expressing AML cells acquire the ability to lyse NK cell-resistant cells, including autologous unmodified AML cells. Interestingly, there is no significant expansion of CD25⁺CD4⁺FoxP3⁺ cells despite the use of IL-2 and/or CD80. The Farzaneh group has opened a clinical trial for relapsed AML patients with poor prognosis. This trial is designed to give the immune system the greatest opportunity to respond. The patients receive chemotherapy to induce partial remission and then receive an allogeneic HSC transplant. When the immune system is reconstituting, and if there is no evidence of graft-versus-host disease, then the patients receive a donor leukocyte infusion and the autologous AML cell vaccine every 3 weeks. Having demonstrated the efficacy of this approach in syngeneic murine leukemia models and having shown the effective in vitro stimulation of autologous NK and T cells, the group is now seeking approval for a similar phase I study of IL-2/CD80 AML vaccination in the autologous setting in the absence of allogeneic HSC transplantation.

Professor Farzaneh then described the human telomerase reverse transcriptase (hTERT) peptide vaccination study in which he has been involved (Aloysius et al., 2009). In this study DCs were loaded with peptide P540 or P865 and stimulated with a cocktail of cytokines (GM-CSF and IL-4 and/or TNF-α). In their phase I clinical trials patients were inoculated with 5–10 × 10⁶ peptide-pulsed DCs every 2–3 weeks. The group investigated the effect of IFN-α on the ex vivo maturation of DCs. P865 is a better peptide for vaccination than P540 and there was no significant effect whether IFN-α was included or excluded in the ex vivo treatment of DCs. The hTERT DC trial results are now published on 10 patients with advanced malignancies of the prostate, colon, breast, and head and neck. These patients were vaccinated with autologous DCs pulsed with two class I hTERT peptides every 2–3 weeks. Eight of 10 patients generated hTERT-specific CTLs, and 2 of 10 had transient but significant regressions of disease during vaccination. The group realized there is a need for the induction of helper functions. In the next phase I trial the group included two promiscuous MHC class II helper peptides called P766 and P762, which they gave to HLA-A2 patients. They were then able to show that there was a greater response to hTERT by hTERT-specific CD8⁺ T cells in patients who had also received the class II peptides. They have now treated eight prostate cancer patients and showed a slightly higher frequency of clinical response in terms of disease stabilization, with two patients having achieved remission. All patients showed significantly lower numbers of Tregs and an increase in antigen-positive CTLs after vaccination. The treatments were well tolerated, and there have been no adverse events and no deterioration in the quality of life, but to date there have been only short-lived immune responses and no sustained clinical responses.

In the final part of Professor Farzaneh's talk, he described the CASAC adjuvant. CASAC, an acronym for combined adjuvant for synergistic activation of cellular immunity (Wells et al., 2008), uses a combination of two Toll-like receptor (TLR) agonists to produce sustained increases in T cell responses, to a greater extent than has been achieved with a single TLR. Many of the tested combinations of TLR work, such as TLR1 and −2 plus TLR3, −4, or −9. Cognate help is provided by anti-CD40, a class II peptide or protein. They combined two TLR agonists, a single MHC class I peptide (either ovalbumin [OVA] or tyrosinase-related protein [TRP]-2), IFN-γ (100 μg), and linked bystander helper in the form of an activating anti-CD40 antibody, MHC class II peptides (related or not, 100 μg), or a protein, all in a 100-μl volume as an emulsion. The best responses were achieved with TLR4 and TLR9, IFN-γ, and anti-CD40. In some of the mice that had been vaccinated with OVA and the CASAC vaccine as many as 57% of total CD8⁺ cells recognized the SIINFEKL epitope. Three months later they challenged the mice with peptide in saline and could show clear recall responses and tumor rejection by vaccinated mice. There was no evidence of autoimmunity in the mice and it is obvious that the CASAC combination adjuvant works better than many other adjuvants.

Albert Deisseroth (FDA, Silver Spring, MD) described our understanding of the response to seasonal flu vaccination, which decreases with age, with the survival advantage in people over 55 years of age being 0%. The reason for this is that the CD40 ligand (CD40L) signal is needed for the induction of a response to vaccination and CD40L is missing in older patients. In older patients, B cells and CD8⁺ T cells cannot multiply without CD40L. DCs need CD40L to become active and in older people helper T cell responses to CD40L decrease. On the basis of this the Deisseroth group developed a vaccine that targets antigen and provides a link to the missing CD40L in older patients. By replacing the absent signal in older people, which binds CD40 on APCs, the ligand complex internalizes CD40L and promotes the MHC class I presentation of peptide. The group then found they could administer the vaccine as a chimeric protein or could embed the chimeric cDNA in an adenovirus. This vaccine activates the innate immune system, leading to cytokine release and causing T cell activation, and if DCs are infected then the signals needed are provided at the site of activation by DCs. They developed a vaccine, Ad-sig-hMUC1-ecd-CD40L, and tested whether it could induce memory responses in mice. One year after vaccination the Deisseroth group took spleen cells and adoptively transferred them into nude mice. They showed that the T cells could suppress the growth of tumor antigen-expressing tumor cells. Through the injection of protein they could expand the magnitude of the immune response induced by the Ad-sig-TAA-ecd/CD40L vaccine but they could not boost responses with vector or obtain an antiadenovirus response. However, through the use of adenoviral vector followed by repeated protein boosts they could produce antibody responses. The Deisseroth group demonstrated that circulating levels of TAA-specific antibodies were induced by the Ad-sig vaccine. When they tried the vaccine in old mice (18 months of age) versus younger mice (2 months), Ad-sig-E7 vaccination was shown to lead to an expansion of antigen-specific T cells in old mice and to suppress the growth of tumor. Protein boosting can clear tumor progression in mice even when they are old. Mucin (MUC)-1 is overexpressed in epithelial cells but barely expressed in all other normal cells, but overexpression predicts chemotherapy resistance and the metastatic propensity of tumors. Through the use of various combinations of vector (V) and protein (P), they could show that hMUC-1-specific antibodies were generated by vector immunization and two protein boosts (VPP), and that vaccinations of hMUC-1 transgenic mice produce antibodies that can bind human breast cancer biopsies unlike serum from unvaccinated mice. The Deisseroth group also saw differences in lung metastases between vaccinated and unvaccinated mice. They are currently examining Treg numbers in cancer nodules and whether these vaccines can work in mice with low levels of lymphoid cells. They have developed a mouse model of extreme lymphopenia: on day –3 the mice are injected with 32D bcr-abl p210 cells, on day 0 the mice are administered total body irradiation, and 1 hr later T cells are depleted and the mice are given a BM transplantation. On day 3, 10⁶ spleen cells from vaccinated or unvaccinated donors are given to the recipient mice and then they are vaccinated with the VPP vaccine combination on days 10, 17, and 28. The group showed a long-term response and are now moving into a breast cancer clinical trial. They have also developed an HA/ecdCD40L vaccine for flu. Production of a swine flu vaccine in chicken eggs is difficult and therefore they are looking at vaccines for avian and swine flu, especially as the currently available vaccines are weak and only partially protective. To do this they have taken part of the stalklike hemagglutinin, where physical and clinical neutralizing antibodies are located. They attached three peptides to CD40L and tested them in young and old mice. They found that antibody levels increase (serum titer against hemagglutinin antigen) in young mice more than in old mice. The injection of vaccine leads to the production of neutralizing antibody to the virus.

The future plan is to:

Produce HA/ecdCD40L vaccine.

Gerald O'Sullivan (Cork Cancer Research Centre, Cork, Ireland) explained that most patients die of untreatable or minimal residual disease (MRD). The impact of this disease in communities remained significant in 2009. Despite improvements in our understanding of the biology, better diagnosis, and staging, there is limited screening and too few specialized centers. There is a need to increase radical treatment, to develop neoadjuvants with chemo- and radiotherapy, and to develop a preventive approach. Professor O'Sullivan's belief is that we should target MRD in the treatment of metastatic cancers. The MRC Adjuvant Gastric Infusional Chemotherapy (MAGIC) trial (Cunningham et al., 2006) showed that 70–80% patients do not benefit from chemotherapy in adenocarcinoma. The gain has been significant but modest, due mostly to MRD multicellular aggregates that lack an autonomous blood supply. Dissemination is the final step in the cancinogenesis process, with 88% patients with adenocarcinoma and squamous carcinoma having micrometastases in the rib marrow. Micrometastases are of prognostic significance; the presence of only 2–7000 per 10⁵ marrow cells equates to 150 × 10⁶ in a 70-kg human, and only one cell needs to go rogue for a patient to die (Kaplan et al., 2005). An increase in BM-derived DCs precedes malignant cells due to obstruction signals from the primary tumor or directly to BM cells by metastases. Variations in the susceptibility of esophageal cells to cytotoxic drugs is associated with autophagy: if cells are resistant to a drug they do not die but instead undergo temporary autophagy. When the drug is withdrawn the cells recover. Autophagy may be inhibited with siRNA targeting beclin I, resulting in reduction or recovery of cancer cells after drug treatment. Autophagy is likely to be a survival mechanism in response to stress and this is probably why so many patients with esophageal, stromal, and lung cancers are resistant to chemotherapy. O'Sullivan's group has already shown that the immune system controls micrometastases (Murphy et al., 2000) and through a study at Mercy University Hospital (Cork, Ireland) has shown that Duke's B colon cancer patients who display immune reactivity to BM micrometastases have improved survival. The O'Sullivan group has hypothesized that the immune therapy of primary tumors could be used to control MRD. To develop a therapy for this they electroporated a plasmid carrying GM-CSF/B7-1 into growing murine fibrosarcoma tumors (Collins et al., 2006). Through in vivo electroporation of the tumor they achieved 60% clearance of the primary tumor and induced a durable systemic response. They found that treatment of subcutaneous tumor prevented growth of distal tumors in a liver metastasis model and that through a combination of immune-gene therapy and depletion of Tregs there was up to 95% clearance of tumors in mice. The O'Sullivan group has also successfully used electroporation to deliver chemotherapeutic drugs (electrochemotherapy) to subcutaneous tumors in patients and canine rectal tumors, and will soon initiate clinical trials in endoscopic delivery to gastrointestinal tumors.

With the described immune-gene therapy, there is no requirement to identify target antigens within the tumor. An alternative strategy, also examined by this group, involves defined endogenous antigen vaccination, again using plasmid delivery by electroporation. The group showed that a murine PSCA vaccine strategy against the TRAMPC1 murine prostate xenograft model improves survival. While 20% of patients have radical prostatectomy relapse, electrogene therapy can help local control of tumor regrowth. It can break tolerance to endogenous antigen and be used to therapeutic advantage. In mice there is no evidence of immune attack of the normal prostate tissue and no indication of the induction of autoimmune disease. The vaccination strategy slowed the growth of primary tumors but did not eliminate them, further indicating the appropriateness of such immune-gene therapies as adjuvant or neoadjuvant treatments in combination with surgical resection of primary tumors.

Cell Therapies

Mark Tangney (Cork Cancer Research Centre, University College Cork), spoke about his group's work in the field of bacterial gene and cell therapy for cancer. Specific properties of various bacterial genera can be exploited in their use as gene therapy vectors. Two different strategies have been employed by this group. The first relates to bactofection, involving the development of a novel attenuated Listeria monocytogenes strain capable of high-efficiency lysis and release of plasmid DNA within tumor cells. Tangney's group has been the first to demonstrate, in vivo, successful gene delivery by Listeria, delivering firefly luciferase to growing tumors in murine models and to patient breast tumor samples ex vivo (van Pijkeren et al., 2009). This strategy was successfully employed in a therapeutic setting for the delivery of thymidine kinase (TK)–ganciclovir therapy of breast tumor xenografts in murine trials.

Another bacterial cell therapy strategy involves delivery of noninvasive bacteria to the tumor site with bacterial expression of genes locally (external to the tumor cell). Several bacterial genera have been demonstrated to localize to and replicate in tumor tissue when intravenously administered in rodent models. With anaerobic bacteria, targeting is thought to be achieved through specific localization of species such as members of the Bifidobacterium genus in hypoxic regions of tumors after intravenous application. By transfection with plasmids suitable for bacterial expression of heterologous genes, these bacteria can home to hypoxic tumors, replicate within them, and locally express therapeutic proteins. Bifidobacteria are native, harmless residents of the human gut, and certain bifidobacterial strains have been shown to have health-promoting or probiotic benefits. A number of bifidobacterial strains that harbor plasmids expressing therapeutic agents, such as endostatin or prodrug-activating enzymes, have been shown to induce regression in rodent tumor models when administered intravenously. However, intravenous injection of bacteria presents concerns about unnatural invasive administration of biological agents, coupled with the traditional association of blood-borne bacteria and bacterial sepsis conditions. Work from the Tangney laboratory has uncovered the novel finding that oral administration of bifidobacteria is as effective as intravenous administration in achieving specific targeting of distal tumors. Bacterial translocation is defined as the passage of viable bacteria from the gastrointestinal tract to extraintestinal sites and this phenomenon is well described in pathogenic bacterial sepsis and related diseases. This group has demonstrated translocation of a nonpathogenic species of bacteria in healthy mice, using a Lux luminescence-based tagging system in bifidobacteria developed by them. Oral administration of B. breve to mice resulted in their translocation from the gastrointestinal tract with subsequent homing to and replication specifically in tumors. Lux-labeled B. breve fed to mice bearing subcutaneous tumors was detected by whole body imaging specifically in tumors, at levels similar to those following intravenous administration. These bacteria also colonized and replicated in B16 melanoma pulmonary metastases as small as 2 mm³. These findings indicate potential for the treatment and/or detection of tumors via ingestion of nonpathogenic bacteria.

Kah-Whye Peng (Mayo Clinic, Rochester, MN) provided an update on their phase I clinical trial investigating the intraperitoneal administration of oncolytic measles virus expressing soluble human carcinoembryonic antigen (MV-CEA) in patients with recurrent ovarian cancer. Although preliminary data are promising, it is clear that viral delivery, especially in these measles-immune patients, can be improved. Dr. Peng and colleagues are developing adipose tissue-derived mesenchymal stromal cells as cell carriers for oncolytic measles virus and she showed that these virus-infected cell carriers can circumvent antimeasles immunity and significantly prolong survival of immunized mice bearing orthotopic human ovarian tumors.

Keiya Ozawa (Jichi Medical University, Tochigi, Japan) gave an interesting presentation on cancer-targeted gene therapy using mesenchymal stem cells (MSCs). On transfer to hosts, MSCs have the ability to accumulate not only at the sites of wounded tissues (as applied in regenerative medicine) and inflammatory tissues, but also tumors, when injected into tumor-bearing mice, and therefore MSCs may be used as a platform for targeted delivery of anticancer agents to tumors (including metastatic sites) in order to augment the therapeutic efficacy without causing systemic side effects. MSCs display low immunogenicity, with HLA matching not required for cell therapy. Although various growth factors and chemokines may be involved, the detailed molecular mechanisms of MSC accumulation at the site of tumors are poorly understood. The Ozawa group researches MSC–endothelial cell adhesion after TNF-α stimulation. Fiber-modified adenovirus was used to tag MSCs. To assess tumor tropism, GFP-expressing MSCs or fibroblasts were injected into the left ventricular cavity in tumor-bearing nude mice. Ex vivo histological analyses revealed that injected MSCs were detected at tumor lesions, unlike fibroblasts. Luciferase-expressing MSCs or fibroblasts were also injected into mice and cells were visualized as localizing to tumors over time, using the IVIS in vivo imaging system. When the migration capacity of MSCs and fibroblasts was examined in a Transwell migration assay, the factors that induced MSC migration in vitro also enhanced fibroblast migration, unlike the situation observed in vivo. It was hypothesized that this phenomenon was due to adherence of MSCs to endothelial cells in tumor stroma. Stimulation of MSCs with TNF-α was shown to enhance MSC adhesion to endothelial cells in vitro. Partial inhibition by blocking antibodies against vascular cell adhesion molecule (VCAM)-1 and very late antigen (VLA)-4 implicated the involvement of these molecules in adhesion, although a similar phenomenon was not observed for fibroblasts. It was also noted that the administration of parthenolide (which reduces TNF-α production) inhibited MSC accumulation in tumors.

The second part of Professor Ozawa's talk outlined the development of vector-producing tumor-tracking MSCs to augment suicide cancer gene therapy. This approach involves the production of herpes simplex virus thymidine kinase (HSV-TK)-expressing retrovirus from MSCs at tumor sites for ganciclovir prodrug therapy. Immunostaining for luciferase demonstrated significant reporter gene expression at tumor peripheries with this approach, while a PCR/restriction mapping method was developed to identify retrovirally transduced tumor cells. This treatment was shown to significantly improve survival in murine tumor models.

Irish Society for Gene and Cell Therapy

Alex McKee (Trinity College Dublin, Dublin, Ireland) delivered a presentation on AAV2/5-SNAP25, altered memory, and learning behaviors. Alex outlined the research goals of the Applied Neurotherapeutics Research Group (ANRG), to understand the nature of the memory/learning at a molecular level, and to identify targets for the development of novel drugs for the treatment of mental illness. Long-term memory is formed by alterations in glutamate-dependent excitatory synaptic transmission, which is in turn regulated by synaptosomal-associated protein (SNAP)-25, a key component of the soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) complex essential for exocytosis of neurotransmitter-filled synaptic vesicles. Both reduced and excessive SNAP-25 activities have been implicated in various disease states that involve cognitive dysfunctions such as attention-deficit hyperactivity disorder, schizophrenia, and Alzheimer's disease. The work presented focused on the overexpression of SNAP-25 in the adult rat dorsal hippocampus by infusion of a recombinant AAV2/5 vector, and evaluation of the consequence of adult dysfunction of the SNARE protein in the absence of developmental disruption. P49 rats received a stereotaxic injection into the hippocampus, followed 5 weeks later by behavioral analyses (water maze, passive avoidance, fear conditioning, pre-pulse inhibition [PPI]) and petrochemical and molecular analyses (microdialysis, paired-pulse facilitation, RT-PCR, immunohistochemistry, and Western blotting). Gene transfer and expression of GFP constructs were visualized microscopically ex vivo, and SNAP-25 overexpression was validated by fluorescence-activated cell sorting (FACS) and RT-PCR. A specific and significant increase in extracellular glutamate levels as well as a reduction in paired-pulse facilitation in the hippocampus were detectable by microdialysis. In addition, SNAP-25 overexpression produced cognitive deficits, delaying acquisition of a spatial map in the water maze and impairing contextual fear conditioning, both tasks known to be dependent on the dorsal hippocampus. The high background transmission state and presynaptic dysfunction likely result in interference with requisite synapse selection during spatial memory consolidation. Together these studies provide the first evidence that excess SNAP-25 activity, restricted to the adult period, is sufficient to mediate significant deficits in the memory formation process.

Roger Smith (Royal Veterinary College, London, UK) presented his current research into the area of bone marrow-derived mesenchymal progenitor cells (MPCs) for naturally occurring tendinopathy in horses. Professor Smith stated that the horse is an attractive translational model for exercise-induced tendinopathy and partial rupture, which is extremely similar in human and equine athletes. The superficial digital flexor tendon, like the Achilles tendon in humans, functions as a “spring” in locomotion, and exercise-related strain resulting in clinical injury is common. Healing results in poorly functional and disorganized scar tissue that reduces performance and leads to high reinjury rates. MPCs offer the potential for tendon regeneration and improved functional outcome associated with a regenerated matrix. There are several features of tendon disease in the horse that lend themselves to cell therapy; the injury provides a “receptacle” for cells, with ease of administration and retention, and it provides an appropriate mechanical environment with vascularized granulation tissue as a scaffold to support cells. The Smith group hypothesized that autologous MPC implantation with marrow supernatant would reduce the reinjury rate compared with conventional management in horses and would induce regeneration of normal tendon matrix as evidenced by the normalization of biomechanical, histological, and compositional parameters. Although it would have been logical to choose differentiated tenocyte cells from the tendons themselves or related tendons, problems such as age-related reduction in synthetic ability or differential phenotypic protein synthesis from tendon fibroblasts from different sources precluded this option. In looking at undifferentiated cells, the use of embryonic SCs was ruled out, as were MPCs from tendon, umbilical cord, and fat. Hence MPCs from BM were employed as they were easy to recover and have been extensively researched. BM was aspirated from the sternum of horses that had suffered naturally occurring superficial digital flexor tendinopathy within the previous month. MPCs were selected, enriched, and expanded by culture on tissue culture plastic. After approximately 3 weeks, 10 million MPCs were suspended in the citrated supernatant of the original BM aspirate and implanted into the central defect in the damaged tendon under ultrasonographic guidance. Engraftment and survival of labeled MPCs were demonstrated histologically up to 125 days after implantation in horses with natural tendon injury, but not injuries induced by surgery or enzymatic induction, where the border of the tendon was disrupted. Engraftment in ex vivo tendon sections was also shown. Several other groups, in addition to the Smith group, have also shown that BM-MSCs can induce the formation of improved-quality tendon matrix in a number of different animal models. Functional tenocytic differentiation of BM-MPCs was also assessed. In an on-going controlled experimental equine study, MPCs were administered to two horses whereas saline was injected into two control animals. Horses underwent a standardized controlled exercise program and were killed after 6 months. MPC-treated tendons had greater elasticity, and were more similar to uninjured tendons, compared with saline-treated controls. Histologically, MPC-treated tendons had lower (improved) scores for cellularity and organization at the injured site compared with the saline-treated group, and were comparable to the uninjured sites of the treated tendon. DNA content was statistically significantly reduced in SC-treated tendons, whereas tissue-linked fluorescence, collagen, and water content of the MPC-treated injured sites were not statistically different from uninjured sites and contralateral limbs. The MPC-treated injured sites had lower total glycosaminoglycan content compared with uninjured sites and the saline-treated group, providing support for improved scar-free healing.

Professor Smith also presented outcomes from a 5-year clinical study, in which BM-MPCs were prepared and implanted, as for the experimental study, into 25 National Hunt racehorses with naturally occurring superficial digital flexor tendon injury. Horses were monitored for 2 years after a return to full work and the number reinjured was recorded and compared with 17 control National Hunt racehorses that suffered the same injury and that were followed up in an identical fashion but treated conventionally (Crowe et al., 2004). Results indicated a significant reduction in reinjury rates in treated animals, with a good safety profile, and no histological evidence of abnormal tissue or neoplastic transformation in those horses in which histological evaluation was possible due to death by other causes. This procedure could potentially be applied to other equine tendon injuries or Achilles and rotator cuff injuries in humans.

Audrey Rayssac, from the Anne-Catherine Prats laboratory (INSERM, Université de Toulouse, Toulouse, France), described her work on an internal ribosome entry site (IRES)-based vector expressing both fibroblast growth factor-2 (FGF2) and Cyr61 for gene therapy of lower limb ischemia. In describing the pathogenesis and traditional therapies for peripheral arterial disease, and mechanisms of critical limb ischemia, she outlined the rationale for gene-based proangiogenic therapy. Because of the potential side effects associated with the production of a single angiogenic factor and given the natural cooperation between various angiogenic factors to form functional vessels, this group designed a system expressing synergistic molecules in low and controlled doses. They developed an IRES-based bicistronic plasmid expressing two angiogenic molecules, FGF2 and Cyr61. This construct was delivered by electroporation into ApoE^–/– mouse hindlimb ischemic muscles, and compared with monocistronic plasmids. This construct had more stable expression than plasmids expressing either gene alone. In vivo analyses using laser Doppler analysis, arteriography, and immunochemistry showed that the bicistronic vector generated more abundant and functional revascularization than the monocistronic vectors, despite producing 5–10 times lower amounts of each angiogenic molecule. As an in vivo model for demonstrating excessive, undesirable angiogenesis, the monocistronic Cyr61 vector was shown to accelerate B16 melanoma growth in mice, unlike the bicistronic vector.

Dr. Erin Vaughan (National University of Ireland, Galway, Ireland) described research from the Regenerative Medicine Institute (REMEDI) laboratory into endothelial progenitor cells (EPCs) for autologous cell transplantation in the treatment of type 1 diabetes mellitus, and specifically their finding that osteopontin (OPN) enhances the angiogenic response in EPCs. This group found differences between healthy and diabetic host EPCs from both humans and rabbits, with reduced ability in vitro in both tubule formation as evidenced in Matrigel assays, and endothelial binding. Numbers of EPCs were also reduced in diabetic patients. Work on OPN knockout (KO) mice showed an association between reduced OPN production and a decreased ability to recover from ischemic injury, using Doppler blood flow analysis, leading them to the hypothesis that OPN downregulation in EPCs contributes to a decreased angiogenic response. To determine whether a lack of OPN could contribute to a decreased ability of EPCs to adhere to the site of injury, labeled KO or WT EPCs were assessed in vitro for endothelial cell binding, and it was found that the ability of the OPN KO cells to adhere to TNF-α-activated endothelial cells was impaired, suggesting that OPN may play a role in cell adherence at sites of injury. Exogenous addition of OPN restored their ability to adhere. Similar results were found in Matrigel tubule formation assays, supporting a role for OPN production in the angiogenic response mediated specifically by EPCs.

Mary Murphy (REMEDI, National University of Ireland, Galway, Ireland) presented the results of research on DNA methylation as a regulatory mechanism in SC differentiation, which was done in collaboration with the Cancer Epigenetics and Biology Program in Barcelona, Spain. Because there is little information about methylation patterns or the role of histone modifications in regulating the expression of key genes involved in self-renewal or differentiation of adult SCs, they concentrated their research efforts on the mesengenic process of MSCs, particularly on those pathways related to proliferation and differentiation into chondrocytes, osteocytes and adipocytes. Their primary interest was in DNA methylation, based on previous studies showing that DNA methylation patterns are better correlated with histone methylation patterns than with the underlying genome sequence context and that CpG methylation imparts dynamic epigenetic marks that undergo extensive changes during cellular differentiation (Meissner et al., 2008). They performed a different type of global methylation array using a high-throughput epigenomic discovery platform with custom CpG island arrays (Infinium HumanMethylation27 BeadChip assay; Illumina, San Diego, CA) to test the hypothesis that DNA methylation is integral to the regulation of gene expression in self-renewal or differentiation of adult SCs. Genomic DNA was isolated from undifferentiated MSCs, all from healthy male donors, and cells were induced to undergo chondrogenic, osteogenic, and adipogenic differentiation. Analysis of 28,000 CpG probes indicated that promoters of undifferentiated MSCs were generally hypomethylated, in agreement with the comparatively unrestricted gene expression of these cells. Analysis of the methylation patterns indicated that chondro- and osteodifferentiated MSCs clustered closer to each other than to undifferentiated MSCs or to adipogenic progeny, with the latter two groups showing the least similarity. Lineage-specific, differentially methylated genes were selected to create a differentiation-specific methylation signature for validation by RT-PCR and bisulfite sequencing. Their analysis showed that DNA methylation contributed to the rapid downregulation of redundant genes on commitment to a specific lineage whereas demethylation was associated with upregulation of genes associated with the differentiation process. From this work, they hypothesized that differentiation of adult SCs involves an initial wave of hypermethylation to rapidly downregulate expression of redundant genes involved in self-renewal or alternative differentiation pathways; in addition, they postulated that epigenetic mechanisms represent an important aspect of control to ensure the sequential and temporal expression of critical genes throughout the differentiation process.

Roisin Dwyer (REMEDI) presented a lecture entitled “MSCs: Tumor-Targeted Delivery Vehicle or Tumor-Promoting Agents?” It was based on studies from their laboratory showing tumor-specific migration and engraftment of fluorescently labeled MSCs at the sites of multiple metastases of breast tumors, after tissue harvest. They found that after systemic injection in vivo, MSC engraftment was present throughout the tumor mass and appeared to be lining the blood vessels. MSC migration was mediated, at least in part, by chemokines including monocyte chemotactic protein (MCP)-1. They are now advancing these studies to attempt to track MSC migration and engraftment in real time in vivo. MSCs were infected with adenovirus expressing the sodium iodide symporter (NIS), to support concentration of technetium-99m (Tc-99m) by the MSCs, for imaging their migration and engraftment in vivo. MSC-NIS were injected intravenously into tumor-bearing mice, followed by imaging with a Bazooka SPECT γ-camera. The images revealed Tc-99m concentration in the stomach and thyroid/salivary glands, representing endogenous mouse NIS expression. After intravenous MSC-NIS administration, Tc-99m was also concentrated at the tumor site. Quantitative PCR analysis of tissues harvested after imaging supported these data, with robust human NIS expression detected at the tumor site. However, hNIS was not detected in the liver or spleen of these animals, with low levels of expression detected in the heart and lungs. When Ad5-CMV-NIS was administered alone, as expected, ectopic hNIS expression was detected in multiple tissues including liver, lungs, spleen, and kidney. These preliminary images were promising and their data supported the specificity of MSC migration toward breast tumors. However, despite their promise, the fate of MSCs after engraftment continues to be poorly understood. Interactions between MSCs and breast cancer cells at the tumor–stromal interface have been reported to promote the development of lung metastases. The researchers further investigated the effect of MSCs on breast cancer cell proliferation, migration, and gene expression profile in vitro to characterize these interactions. A low-density array was used to analyze changes in expression of genes associated with breast cancer progression after interaction with MSCs. MSC-secreted factors were found to stimulate breast cancer cell migration and suppress proliferation, exhibiting upregulation of oncogenes, and genes associated with invasion, migration, and antiapoptosis, and increased expression of epithelial mesenchymal transition genes with decreased E-cadherin protein staining, and changes from a clustered to dispersed growth pattern. The authors believe that understanding breast cancer cell–MSC interactions is fundamental to determining whether MSC homing should be harnessed for delivery of therapeutic agents or whether MSC–tumor interactions could be considered a target for intervention.

Robert Montgomery (Blood Research Institute, Milwaukee, WI) presented their research results on platelet-directed factor VIII (FVIII) and FIX for gene therapy of hemophilia A and B. On the basis of their previous studies looking into the relationship of FVIII and von Willebrand factor, which is the carrier of FVIII, and the knowledge that neither coagulation FVIII or FIX is produced in megakaryocytes, they were able to postulate that megakaryocytes could synthesize FVIII by means of gene transduction. The platelet-specific integrin α_IIbβ₃ promoter was used for this purpose and two simultaneous approaches were studied: the development of a transgenic mouse using their α-IIb FVIII (2bF8) plasmid (pCIneo-2b-hB-FVIII) and an LV construct that incorporated the α_IIβ promoter, B domain-deleted (BDD)-FVIII, and woodchuck hepatitis virus posttranscriptional response element (pWPT-2b-gB-FVIII). In their experiments, the platelets of transgenic mice were found to have expression of FVIII, with twice the amount in homozygous versus heterozygous mice, and to optimize expression the platelets must have the presence of von Willebrand factor.

They looked at the dose response to tail clipping in mice, and all FVIII null mice survived tail clipping if they were infused with 2bF8 platelets to achieve a final level of 30% of FVIII in the platelets. If modest levels of FVIII were present in the plasma of KO mice, anti-FVIII antibodies inactivated the FVIII present. However, in the transgenic mice with platelet FVIII, the same level of inhibitors did not abrogate the beneficial effect of 2bF8 in platelets. All FVIII null mice survived tail clipping if infused to a final rhFVIII concentration of 2 U/dl. None of the 12 mice with inhibitors at 25 or 250 Bethesda units (BU)/ml survived tail clipping although rhFVIII (2 U/dl) was infused. In contrast, in the 2bF8 mice, even with the whole blood concentration of FVIII being less than 2% FVIII, hemostasis was achieved even at >2000 BU/ml. To determine whether stored platelet FVIII was being protected from inhibitory antibodies, they used three strategies for inhibitor model studies: a passive model infusing inhibitory plasma from highly immunized FVIII null mice, a chronic model using adoptive transfer of spleen cells from immunized FVIII null mice into 2bF8 transgenic mice, and active immunization of 2bF8 transgenic mice with human recombinant BDD-FVIII with adjuvant. Five Bethesda units renders patients untreatable, but not if platelet-directed FVIII is present through LV-mediated gene therapy. BM was harvested from the femurs and tibias of mice, and transduced with 2bFVIII and transplanted into irradiated mice. The platelets of all the transplanted mice produced similar amounts of FVIII, which were maintained for a period of a year, indicating the presence of repopulated BM. Even if antibodies were present they had the same dose result if appropriate marrow was present with the phenotype that corrected the bleeding problem. A similar approach was used for FIX in hemophilia B mice, using a transgenic model or LV-mediated transduction of HSCs and BMT. FIX was (1) expressed, (2) appropriately γ-carboxylated, (3) stored, and (4) released from platelet α granules, where it abrogated hemorrhage in FIX^–/– mice. In the BMT model, tolerance to human FIX could be demonstrated even when mice were immunized with human FIX with adjuvant. They concluded that FVIII or FIX can be expressed in platelets as a means of gene therapy in hemophilia. 2b-FVIII is found only in platelets and is effective even in the presence of high-titer inhibitory antibodies. 2b-FIX is found in platelets and minimally in plasma and corrects the hemorrhagic phenotype of FIX KO mice, but does not appear to be as effective in the presence of inhibitory antibodies.

Marie C.M. Lin, on behalf of Hong Yao (Biomedical Engineering Research Centre, Kunming Medical University, Kunming, China), described the creation of a novel polymeric nanoparticle, based on the concept of delivering gene therapy via low molecular weight polyethylenimine (PEI) (600 Da), cyclodextrin (CyD), and folate (FA) (H₁: PEI-CyD-folate) as a safer, less toxic, and high-efficiency alternative to viral vectors, due to its low immunogenicity, low cost, adaptability to chemical modifications, and capability of transporting large DNA plasmids. One of their research objectives was to use this nanopolymer in the delivery of IL-2 for melanoma cancer gene therapy. They characterized the structure of the polymer and examined its in vitro transfection efficiency, cytotoxicity, as well as in vivo transfection. Nuclear magnetic resonance (NMR) spectroscopy, ultraviolet (UV), and X-ray diffraction demonstrated successful conjugations of PEI, CyD, and FA in H₁. Atomic force microscopy (AFM), ζ potential, and transmission electronic microscopy (TEM) measurements were done to determine the size and charge of the H₁/plasmid polyplex as a function of nitrogen-to-phosphorus (N/P) ratios. The in vitro transfection efficiency was more than 50%, which could be improved by the presence of fetal bovine serum or albumin. The cytotoxicity of H₁/plasmid polyplex was lower than that of high molecular weight PEI (PEI-25 kDa). In addition, H₁-pEGFP polyplex effectively transduced pEGFP into cancer cell lines and intratumoral injection of H₁/pLuc polyplex showed high transfection/expression similar to that of Ad-Luc. They studied the combination of adenovirus and H₁ in order to shadow the epitope of the antigen of adenovirus, reduce the immune response, and allow for changes in the surface charge of the particles from negative to positive and thus enhance contact ability with the target cells. They used an in vivo model of melanoma B16 cell-bearing C57 mice. The repeated injection of Ad+H₁ into living mice did not attenuate its transduction efficiency. Previous studies of Ad-mediated IL-2 gene therapy in human melanoma clinical trials (Trudel et al., 2001; Dummer et al., 2008) have shown regression in 20–24% of tumors. The subcutaneous injection of H₁/pLuciferase showed high transfection/expression efficiency. Mice were injected with H₁/pLuc polyplex around tumor masses by subcutaneous injection (50 μg; N/P ratio, 20:1) and comparable to that of adenovirus-mediated luciferase transduction (10⁰–10⁹ plaque-forming units [PFU]) in vivo in mice bearing melanoma tumor. For this experiment melanoma xenograft C57 mice were used and subcutaneous injection of mice with B16 melanoma cells was performed, followed by treatment injection 7 days after tumor inoculation. Treatment was repeated and tumor volume was measured every other day. The mice were killed 3 weeks later, immune and cytokine profiles were measured in blood and in CTLs and NK cells, and a proliferation assay was performed with spleen cells. The results suggested that the therapeutic effect of H₁/pIL-2 polyplex is dose dependent and comparable to that of Ad-IL-2. The number of CD4-, CD8-, and CD49-positive cells and the levels of IL-2, IL-6, IFN, TNF-α, GM-CSF, IL-4, and IL-17 were increased significantly.

Oncolytic Virotherapy

There has been significant progress in the field of virotherapy to increase the potency and accuracy of RCRs for cancer therapy. A number of oncolytic viruses are in phase I/II clinical testing and the field was represented at this conference by investigators who are working on CRAd, reovirus, measles virus, and vaccinia virus.

Len Seymour (University of Oxford, Oxford, UK) started the session with a thought-provoking talk on challenges facing the systemic delivery of viruses. Viruses can be sequestered by the innate immune system (e.g., complement and antibody) or by the reticuloendothelial system of the liver and spleen, or have poor extravasation from the lumen of blood vessels out into the tumor parenchyma. These barriers can potentially be minimized by generating “stealth” vector particles through polymer coating or by depletion of Kupffer cells to decrease phagocytic uptake. Virus extravasation into tumor deposits can also be enhanced by increasing the permeability of the tumor vasculature through coadministration of cytokines such as IL-1β, VEGF, IL-2, histamine, and bradykinin.

Kevin Harrington (Institute of Cancer Research, London, UK) discussed various clinical strategies to maximize the antitumor activity of oncolytic viruses in cancer patients. Dr. Harrington presented data from clinical trials using oncolytic reovirus and showed that virotherapy outcomes would be more successful if virotherapy were used in combination with standard frontline therapy such as radiation or chemotherapy. In particular, he showed that combination therapy consisting of reovirus and chemotherapy blunts the generation of neutralizing antiviral antibodies. In general, toxicity from virotherapy trials has been mild and dose-limiting toxicity has not been observed.

Stephen Russell (Mayo Clinic) talked about the importance of noninvasive monitoring of viral gene expression and presented data on high-resolution analysis of sites of viral replication via single-photon emission computed tomography combined with computed tomography (SPECT-CT) imaging. Gene expression data are often lacking in human gene and virotherapy studies as it is not feasible to obtain repeat tissue biopsies. To enable noninvasive monitoring, Dr. Russell and colleagues have inserted the human thyroidal NIS into various vector systems (Ad, AAV, VSV, and measles virus) and showed that sites and duration of viral gene expression can be monitored accurately and regularly, using high-resolution SPECT-CT and commonly available tracers such as Tc-99m and I-123.

John Bell (University of Ottawa, Ottawa, ON, Canada) shared with the audience data from a number of clinical trials testing the safety and efficacy of the oncolytic vaccinia virus JX-594. JX-594 is a targeted oncolytic poxvirus designed to selectively replicate in and destroy cancer cells with cell cycle abnormalities. Direct oncolysis plus GM-CSF expression also stimulate shutdown of tumor vasculature and antitumoral immunity. JX-594 is tumor selective (mechanisms include IFN sensitivity, high TK pool in tumor cells, and EGFR-activated pathways) and induces fibrin clots in tumor blood vessels, causing vascular shutdown in tumor nodules. Phase II clinical trial data in which 16 patients with HCC or colorectal cancer were treated with JX-594 were also presented.

David Curiel (University of Alabama, Birmingham, AL) reviewed the quest to generate tumor-targeted CRAds. Dr. Curiel explained the unique challenges associated with the display of single-chain antibodies on adenovirus and stressed that although there are abundant in vitro data on tumor specificity of retargeted adenoviruses, in vivo data demonstrating the feasibility and utility of retargeted viruses are still lacking. He presented the strategy of using adapter molecules, soluble receptors linked to targeting ligands, as well as the addition of genetic modifications of the fiber knob domain to allow the adenovirus to infect cells through alternative receptors.

John Dong (University of South Carolina, Columbia, SC) presented data on dual-locked, prostate cancer-specific adenovirus, whose gene expression and replication is driven by probasin promoter to achieve selective replication in tumor cells, more efficient viral spread, and the creation of an inflammatory environment in tumors to induce an immune response against tumor antigens. They demonstrated that this dual-locked virus, with the probasin promoter driving E1 and E4 expression, has higher levels of transgene expression with higher specificity. When injected into LNCaP xenografts, the dual-locked virus has superior viral replication and oncolysis, resulting in extensive necrosis of tumors and achieving superior antitumor activity.

Chae-Ok Yun (Institute for Cancer Research, Yonsei University, Seoul, South Korea) stressed the importance of extracellular matrix as a physical barrier to the in vivo spread of virus. Because relaxin upregulates matrix metalloproteinases (MMPs), procollagenase, and tissue inhibitor of MMPs (TIMPs), and downregulates the synthesis and secretion of interstitial collagens and degradation of the extracellular matrix, Professor Yun and colleagues have generated CRAds expressing relaxin. Using spheroids of tumor cells, they demonstrated that relaxin-expressing virus is able to spread farther into the spheroid than does WT virus. A phase I clinical trial is in progress using intratumoral injection of the relaxin-expressing virus (DWP418) in patients with advanced solid tumors. The trial was approved by the Korean FDA in January 2008 and the first patient was treated in June 2008. Dose levels include 10¹¹, 10¹², 2 × 10¹², 5 × 10¹², and 10¹³ vector particles per patient.

Nonviral Gene Therapy

Andrew Miller (Imperial College) opened the session on nonviral gene therapy with a presentation on the use of nanomaterials and nanotechnology as a vector for nucleic acid therapies. The development of nucleic acid therapeutics (coding or noncoding) currently represents an underused resource in the pharmaceutical industry, because of poor bioavailability and efficacy of the therapeutic constructs. Professor Miller has focused on the development of cationic liposomes and polymers with the aim of defining an ABCD nanoparticle (NP) paradigm that defines the critical concepts for vector-mediated nucleic acid delivery in vivo (Kostarelos and Miller, 2005). This comprises a high degree of structural regularity, with the nucleic acid (A-component) entrapped in a central lipid envelope layer (B-component); this in turn is complemented with a stealth biocompatibility polymer (C-component) and an optional ligand-targeting layer (D-component). The process of formulation is adaptable to the delivery objective, resulting in tailor-made delivery solutions, be it for liver, lung, tumor, or for vaccination, with a focus on siRNA delivery because of its increased functional delivery relative to DNA in the majority of tissues. The structural integrity of ABC or ABCD NPs can be influenced by the surrounding pH, resulting in a pH-triggerable phenomenon that serves to release the nucleic acid cargo into the endosomal environment of the cell. Results on the delivery of a nontoxic, nonstimulating IFN-linked, pure siRNA NP targeting HBV were presented, which although not currently at a level delivering therapeutic efficacy do provide a platform for functional delivery of siRNA to liver. Although the internal environment of hepatocytes is currently unreceptive to NP-mediated functional DNA delivery, the same is not the case for delivery to the lung. Half-life-triggered ABC NPs displayed delivery levels comparatively better than those achieved with the recognized standard, namely JetPEI (Poly-Plus, France). In a tumor environment long-term circulating Gd-ABC and Gd-ABCD NPs were employed as molecular imaging tools to help understand the rules for delivery to tumors. Distribution of Gd-ABC NPs was evident throughout tumors in comparison with folate-targeted Gd-ABCD NPs, which were located primarily in the perivascular area. It is hoped that these findings will provide critical information regarding the synthesis of siRNA NPs for the treatment of tumors. Indeed, preliminary results in the treatment of metastatic and subcutaneous tumor models with long-term circulation siRNA ABC NPs and siRNA ABCD NPs are encouraging.

Daniel Scherman (CNRS-INSERM, Paris, France) provided an overview on their work in relation to nonviral targeting for tumor gene therapy and work on the development of a novel imaging technology (SPECT). Nonviral approaches to gene delivery such as the physical/chemical delivery methods of electroporation and cationic lipids, respectively, can offer localized targeted delivery to muscle, tumor, and joints. Electroporation can deliver a significant log fold increase in gene expression to various tumor types; the effect can be variable and depends on the tumor type. Immunization remains the most active field for electroporation, through the active expression of immunogenic transgenes. The effect of electrotransfer can be double and synergic. On the one hand, antigen expression after electrotransfer raises a specific immune response. On the other hand, electric fields induce some cell death, which serves to generate local inflammation, which increases the immune response. This synergistic effect will continue to play an important role in electroporation in future gene therapy trials.

Chemical methods described involve the use of cationic lipids in a targeting strategy, which employs the ABCD concept, as already described in this conference by Professor Andrew Miller. The addition of an appropriate ligand (modified antibody) allows for specific targeted delivery, with proof of concept being provided by the antibody-targeted delivery of luciferase reporter genes to ErbB2- and Her2-positive cells, respectively. This approach can be expensive and complex to develop the desired particles, and therefore an alternative strategy targeting the folate receptor, which is overexpressed on tumor cells, was explored. Although in vitro results showed targeted folate-dependent transfection of folate receptor-bearing cells, the results were not replicated in the in vivo setting, suggesting the need to identify a more appropriate cell surface receptor. Finally, the development of a novel imaging technology, SPECT, was described, which allowed for the scintigraphic quantification of hepatic function in vivo. Targeting of a specific asialoglycoprotein receptor, which requires galactose for specific activity, on the sinusoidal membrane of hepatocytes can be used to reflect the number of active hepatocytes. The technology has a number of potential applications in assessing liver disease, functional liver imaging to determine the clinical feasibility of partial hepatectomy in HCC, and determination of morbidity in liver donor patients.

Mahvash Tavassoli (King's College London) presented a summary of their current understanding regarding the mechanism of action and delivery strategies being developed for a viral protein, apoptin, that has an inherent tumor cytotoxicity function. The therapeutic potential of apoptin is highlighted by its ability to induce apoptotic cell-mediated killing in human cancer cells whereas no cytotoxicity is observed in normal cells. Apoptin-mediated cell death was shown to be independent of the p53 pathway; however, the role of a p53 functional homolog, p73, was shown to be extremely important in the cell death strategy. The effect was specifically linked to a decrease in expression of the oncogenic ΔNp73 isoform of the p73 protein. Novel strategies for the efficient delivery of apoptin specifically to tumor cells have focused on the use of protein transduction domains such as the trans-acting activator of transcription (TAT). The generation of recombinant proteins in bacteria (TAT-Apoptin), although allowing for efficient delivery and translocation to the nucleus of tumor cells but not normal cells, did pose problems regarding solubility and sufficient levels of production. These problems were overcome through the development of a novel mammalian furin-modified TAT-based protein transduction strategy, which provided efficient production and secretion of a functionally active recombinant protein and demonstrated successful transduction of target cells. This system represents an important protein transduction strategy or cell-mediated cancer therapy.

Hot Topics

A number of short papers, which focused on improving the targeting potential and therapeutic efficacy of suicide gene therapy strategies, were presented. In the first of these papers, Ujjwala Warawdekar (Advanced Centre for Treatment, Research and Education in Cancer, Kharghar, India) presented work on the enhancement of the bystander effect in suicide gene therapy, namely the transfer of toxic metabolites to neighboring cells, through the improvement of gap junction intercellular communication. Laser confocal microscopy and flow cytometric analysis confirmed a mean increase in gap junction communication of between 30 and 54% after treatment with the drugs valproic acid, retinoic acid, and phenyl butyrate. A concurrent increase in connexin-43 levels was also observed.

Torben Gjetting (Copenhagen University Hospital, Copenhagen, Denmark) discussed a transcriptionally targeted suicide gene therapy strategy focusing on the treatment of small-cell lung cancer (SCLC). Selective expression in SCLC is achieved through the use of the human insulinoma-associated antigen-1 (INSM1) promoter, an early developmental promoter, which is silent in adults but enables specific expression in SCLC. When fused to a construct containing a yeast cytosine deaminase (CD) and uracil phosphoribosyltransferase, the newly termed super-CD (SCD) can induce selective killing through conversion of 5-FC to its toxic 5-FU derivative. Intratumoral delivery of the construct in cationic liposomes resulted in significant reduction in tumor growth. Modification of the liposomes complexes with between 0 to 10% PEGylated lipids resulted in significant changes to the bioavailability and transfection efficiencies. The lower percentage of PEGylated complexes favored lung localization and greater gene expression.

Masato Abei (University of Tsukuba Graduate School of Comprehensive Human Sciences, Tsukuba, Japan) presented work on suicide gene therapy through the use of a Z33 fiber-modified adenovirus, which incorporates an IgG Fc-binding motif (Z33) from the Staphylococcus protein A, for the targeted antibody-mediated treatment of biliary cancers (BCs). The use of tumor-specific antibodies (i.e., antiepithelial cell adhesion molecule [anti-EpCAM], and anti-EGFR) coupled to a Z33 adenovirus expressing the uracil phosphoribosyltransferase (UPRT) enzyme led to increased sensitivity of BC cell lines to 5-FU. This was correlated to a marked reduction in BC xenograft growth in vivo. The effect was specific to the BC cell lines and did not result in increased sensitivity of either HeLa cells or normal hepatocytes to 5-FU. This strategy is open to a variety of gene therapy and molecular targeted therapy combinations for the treatment of a variety of cancers.

Moving Science into Business

In closing the conference Dr. Kevin Scanlon (International BioScience, Altadena, CA) provided a brief and thorough history from the initial Cancer Gene Therapy conference in 1991 in San Diego, California, to the publication of the first Cancer Gene Therapy journal in 1993 (more than 2000 papers published to date), and the initiation of the society in 1999. The primary aims of the society, that is, to provide an environment of inclusivity and debate and to encourage collaboration, have continued to grow year by year, as evidenced by the large contingents of delegates from the United States, Europe, and Asia as well as 9 of 10 of the past presidents of the society at this year's conference in Cork, Ireland. In the second part of his presentation Dr. Scanlon outlined the essential requirements for the gene therapy community in the commercialization of their work from clinical trial to medical products in the current economic environment. Consolidations in the pharmaceutical industry together with the fact that a number of major drugs are going off patent, leading to significant economic losses in the near future, have heightened the search for new “blockbuster” drugs. The development of a gene therapy drug (treatment) for cancer to rival those currently available for the treatment of diseases of the brain, lung, and cardiovascular and digestive systems, would meet this blockbuster criterion. The success of a new start-up is dependent on the meeting of a number of essential criteria, the first and most important being the innovative nature and therapeutic impact of the product. Subsequent criteria include the ability to safeguard the intellectual property (patent), identify the value and customer base for the product, and establish the advantage of the product over existing treatments. Finally, an experienced management team that understands the pitfalls and has the ability to satisfy the return projections of investors needs to be assembled. The establishment of a collaborative environment between government, business, and academics, similar to that pioneered by Germany in the 1880s, providing initial funding, mentoring, and support services within an incubator facility, is a key requirement in ensuring the success of a product in reaching the market. This in turn can lead to the support of “angel investor” groups helping to bridge the gap to the pharmaceutical industry. The development of a blockbuster gene therapy drug for cancer has been hampered by the lack of efficacious results in the clinic, but significant developments have been made in the last 10 years, many of which were outlined during the course of this conference and helped in no small way by the collaborative links forged between America, Europe, and Asia through the ISCGT.

Summary

The 10th meeting of the ISCGT marks a decade of progress in the field of human gene therapy, seeing our studies move from bench to bedside, and from hopes to reality. This meeting focused less on how gene therapy could or would be administered to further results from phase I/II clinical trials in which clinical responses were apparent. In addition, we listened to presentations on how better targeting of treatment to patients who are predicted to respond at disease presentation (Manfred Dietel) saves money for the state and side effects for patients, and will soon become standard. We heard of advances that have led to better targeting of gene therapy to tumors (Len Seymour, David Curiel, John Dong, and Mahvash Tavassoli), the identification of new biomarkers (Kam Man Hui), modes of treatment (Andrew Miller), targets for gene therapy (Barbara Guinn and Alex McGee), cellular therapy (Ann Leen, Helen Tayton-Martin, and Yajun Guo), SC therapy (Nagy Habib, Karin Gaensler, Roisin Dwyer, and Mary Murphy) and small-molecule therapy (Jim Norris and Marie C.M. Linn), noninvasive (in vivo) monitoring of gene expression (Stephen Russell), and clinical trial data (Farzin Farzaneh, Albert Deisseroth, Kah-Whye Peng, Kevin Harrington, John Bell, and Chae-Ok Yun). Despite gene therapy clinical trials no longer being novel, new treatments, better optimization and new challenges keep us busy. Of particular note as we reflect on the landmark anniversary of our growing society, we see that gene therapy is no longer a pipe dream; our presentations describe how to improve gene therapy in clinical trials and not the first steps into the clinic. Our experience of gene therapy in patients continues to grow and our treatments are becoming more effective as new waves of therapy are administered and monitored. This can only lead to breakthroughs, and in real terms to the saving of lives, one at a time.