This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
Tumor regression by viral infection has been observed since the nineteenth (Dock 563-92). At that time, cancer therapy was limited to tumor excision by surgery under anaesthesia. Quickly, different cancer treatments had emerged. It mainly includes radiotherapy, chemotherapy and recently immunotherapy. Unfortunately, these treatments remain too long and/or not successful in many malignant cases and metastases. Meanwhile, numerous studies have shown that some viruses were able to kill cancer cells. But it was only in early 21st century that the first oncolytic virus was approved (Garber 298-300). Thus, virotherapy came out with a new hope to cure cancer.
The delay of virotherapy development was due to serious barriers (Kelly and Russell 651-59). In addition to the low efficiency of some natural oncolytic viruses, many clinical experiments were faced to important problems of pathogenicity and mortality (Kelly and Russell 651-59). Many were abandoned considering ethical rules. In order to overcome these obstacles, reprogrammed viruses were engineered in the goal to make them cancer cell specific and to improve their oncolytic properties (Cattaneo et al. 529-40). Oncolytic virotherapy owes its expansion these last decades to the better understanding of viruses and host immunity, molecular biology tools, the availability of cancer diagnostic markers, the advent of cell culture technics and sophisticated models of animals.
A number of oncolytic viruses are currently in different phases of clinical trials (Ferguson, Lemoine, and Wang 805629;Ferguson, Lemoine, and Wang 805629;Cattaneo et al. 529-40) such as (i) DNA viruses: the adenovirus (ad), vaccinia virus and herpes simplex virus (HSV) and (ii) RNA virus: measles virus (MeV), Newcastle Disease virus (NDV) and Reovirus. In this report, development of non-modified and reprogrammed paramyxoviruses as oncolytic vectors will be reviewed. By giving examples, this review will summarize the adopted approaches to study and investigate oncolytic viruses and current state of clinical trials for cancer therapy.
II. Paramyxoviruses and oncolytic viruses
Paramyxoviruses are enveloped RNA viruses. For many reasons, they are considered as attractive candidates for medical applications like gene therapy (Kobayashi et al. 2607-14), vaccine (Bukreyev et al. 10293-306) (Suryanarayana, Chattopadhyay, and Shaila) and oncotherapy (Cattaneo e1000973).
In one hand, these non-segmented negative-strand RNA viruses contain simple genomes mainly including six genes, which encode for six major proteins. Host attachment/fusion cell entry system is, as well, less complicated in Paramyxoviruses than in other viruses. It is indeed performed by two independent glycoproteins contrarily to retroviruses and DNA viruses, in which these functions are done by respectively single and multiple proteins. The lack of host genome integration and recombination with other viruses makes them more interesting than retroviruses and DNA viruses. Moreover, a number of Paramyxoviruses, like Sendai virus (SeV) and NDV, is naturally attenuated for human since it is only pathogen for animals such as mice and avian.
In another hand, viruses from Paramyxoviridae family replicate well in a variety of cell lines derived from various animal species. Their biology and replication system are well characterized and manipulation of their genome is possible thanks to the master of reverse genetic technique (Calain and Roux 4822-30;Garcin et al. 6087-94;Conzelmann 123-62;Radecke et al. 5773-84).
Together, these advantages may explain the considerable development and advance with oncotherapeutic studies using Paramyxoviruses. These studies concern natural as well as reprogrammed viruses.
From identification of oncolytic viruses to their clinical application in cancer therapy, studies covering different aspects of the oncolytic agent action and effect can take decades. Researches include virus preparation and characterization, demonstration and optimization of its oncolytic potency and finally clinical tests for side effect and toxicity. Examples cited in this paragraph will be detailed in paragraph (II.2).
In some cases, animal viruses represent good choice to circumvent pathogenicity in human. This is even better when the chosen virus has naturally oncolytic potency like Newcastle disease virus. In this case, most of investigated strains in oncotherapy field are wild type and naturally attenuated. Isolates of attenuated lineage were obtained after several passages of animal cells infection with low multiplicity (10 to 20 PFU/cell). For instance, by adopting these conditions, Hewlett G. et al have succeeded to produce, from MDBK cells, high titers of NDV virus with law HN expression level at the surface (Alexander, Reeve, and Poste 369-73). Many naturally attenuated NDV strains were purified using this system like PV701 (Pecora et al. 2251-66), MTH-68/H (Alexander, Reeve, and Poste 369-73), 73-T (Zorn et al. 225-35) and HUJ (Freeman et al. 221-28). Similarly, Edmonston attenuated MeV strains like-Zagreb were prepared from passages of the wild type MeV through human cells or embryonated chicken eggs (ENDERS and PEEBLES 277-86;Matumoto 152-76)
Many studies have sought to optimize oncolytic virus candidates by genetic modification. The aim of virus reprogramming is to enhance selectivity and/or efficiency of its action against cancer. Modification can be done in different levels of the virus infection cycle (Cattaneo e1000973;Cattaneo et al. 529-40).
? Retargeting allows virus tropism restriction to cancer cells. Retargeting approaches include: (1) mutating surface viral proteins, Hemagglutinin (H) or fusion (F) protein, responsible for virus entry to abolish either its binding to the natural cells receptor or virus fusion to host-cell plasma membrane (Ungerechts et al. 1506-16) (Kinoh et al. 392-403), (2) Inserting either antibody against tumor receptor like CD20 (Ungerechts et al. 1506-16), a B lymphocyte marker, or linker sequence sensitive to tumor markers like MMS or uPAR (Springfeld et al. 7694-700) (Muhlebach et al. 7620-29) (Jing et al. 1459-68) (Kinoh et al. 1137-45;Kinoh et al. 392-403), thus the paramoyxovirus particles, like MeV and SeV, enter cells through cancer specific receptors, and (3) Inserting of target sequences, such as miRNA target, limiting virus replication to cancer cells environment (Leber et al. 1097-106).
? Shielding approach consists in producing chimera viruses in order to escape pre-existing viral immune response or to pseudotype viral vectors expressing foreign envelop in order to insure gene transfer without genome vector integration (Funke et al. 1427-36) (Baum 1349-50). For instance, H protein of measles virus was engineered to specifically target CD20 receptor, a cancer marker and then was inserted into lentiviral vector.
? One of the purposes of arming viruses is to enhance oncolytic activity via insertion of therapeutic genes like PNP in order to improve the efficiency of FCR regimens (cancer therapy)(Ungerechts G;Bossow S), or IFN-ï¿½ gene either to modify host immune response in favour of cancer cells or in order to establish the apoptosis state in cancer cells. This method was used to produce an armed Sendai virus: ï¿½BioKnife-IFN-ß”(Hasegawa Y). Arming method is also used to track cells during virus infection using soluble marker peptides like CEA (Peng KW) and NIS(Ahn).
Oncolytic effect is usually first confirmed in vitro using models of cancer cells (Dalerba, Cho, and Clarke 267-84) like fibrosarcoma cell line (HT1080) and malignant pleural mesothelioma epithelial and biphasic subtypes, respectively H226 and MSTO-211H (Shetty et al. L972-L982) (Morodomi et al. 769-77). In vivo, it is frequent to test oncolytic viruses on animal models with spontaneously or artificially transplanted tumor. Numerous trials involve rodent models having usually orthotopic human xenografts (Sugiyama et al.) (Leber et al. 1097-106) (Ungerechts et al. 1506-16) (Springfeld et al. 7694-700) (Phuong et al. 2462-69) (Morodomi et al. 769-77). To be transplantable, mice should be immune-deficient. Formerly, transplanted mice were previously irradiated to avoid xenografts rejection. Since the end of the 60s different lineages of immune-deficient mice are available like severe combined immunodeficiency (SCID) mice (Bosma and Carroll 323-50) and nude/athymic mice (Animal Models in Developmental Therapeutics - Holland-Frei Cancer Medicine - NCBI Bookshelf) (Belizï¿½rio 79-85). Transgenic mice were also generated for preclinical trial and preliminary toxicity studies. Genetic modification allows mimicking human body conditions like the transgenic mouse strain ï¿½(IFNARKO) CD46 Geï¿½ used for preclinical MeV oncotherapeutic studies (Mrkic et al. 7420-27). For the same purpose, non-human primates were also used. They present the advantage to be genetically close to human considering evolution. Effect similarity between human and non-human patients is thus expected (Xia and Chen 70-80). Among primates used for oncolytic studies of paramyxoviruses, Squirrel monkey and Rhesus macaques were used for preclinical trials to evaluate toxicity of engineered measles virus strains: MeV-NIS (Myers et al. 700-10) and MeV-CEA (Myers et al. 690-98).
Among the abundant amount of oncolytic candidates, few were recently undertaken though clinical trials. This step involves real patients and can take place only after authoritiesï¿½ permission. Patient illegibility defines several criteria: the age, cancer type and/or phase, immune systemï¿½ etc., and is fixed before cohortï¿½s recruitment. Clinical trials intend to seek virotherapy effect and to assess disease treatment regarding following parameters: (i) the therapeutic index which is a ratio between tumor and healthy cells (Therasse et al. 205-16), (ii) TCID50 or 50% tissue culture infective dose and (iii) the maximum tolerated dose (MTD). Trials mainly include different phases from I to IV. Phase I is always planned to determine the MTD using dose escalation of the oncolytic virus. It generally includes, as controls, placebo-treated or non-treated patients. Other controls can be used during different phases like combination or not with others therapeutic agents. From the first to the fourth phase, patient number size increases from tens to hundreds/thousands in order to confirm observation from the previous one. The goal of phase I and II is also to evaluate side effect and body reaction related to the oncolytic virus and to determine the best administration way. Phase III trials include hundreds to thousands of patients usually from different regions (Cancer Clinical Trials Education Series). They are generally randomized in groups treated differently (varied way of injection and/or drug type). Finally, phase IV trials are in-depth studies of treatment efficiency, risks and long-term effect years after the treatment. As of today, main clinical trials of oncolytic paramxoviruses are in phase I and II (table3). Clinical trials remain recent and needs years to be completed.
2. Examples of oncolytic paramyxoviruses
a. Newcastle disease virus (NDV)
NDV virus (also called avian Paramyxovirus type 1: APMV-1) is an avulavirus. It is fatal for birds but causes slight illness when it infects human including soft Flu-like symptoms and conjunctivitis. Depending on their related illness, there are three NDV pathotypes: lentogenic, mesogenic and velogenic. NDV selectively replicates in tumor cells and kills them taking advantages of mainly two properties of tumorigenicty: (1) the lack of effective interferon system (Krishnamurthy et al. 5145-55) and immune response and (2) the overactivation of Ras signalling pathway. It has been demonstrated that NDV cytotoxic effect on human fibroblast lines increases when Ras protein is expressed (Lorence et al. 6017-21). Same oncolytic basis were also shown with reovirus (Strong et al. 3351-62). Characteristics of different NDV pathotypes and strains were investigated for anti-cancer purposes.
Newcastle disease virus strains are classified into lytic and non-lytic, considering the mechanism of oncolysis. Both groups have been undergone through anti-cancer studies as they efficiently and selectively replicate in neoplastic cells.
Lytic strains, such us MTH-68/H, PV 701 and 73-T, lyse tumor cells by directly attacking the plasma membrane and forming syncytia. They were tested in mouse and human models with subcutaneous, intradermal and intravenous (i.v) injections. Clinical trials showed their ability to kill tumor xenografts (Phuangsab et al. 27-36) with no severe side effect (Ravindra et al. 507-13).
? MTH-68/H (from ï¿½More Than Hope 1968ï¿½) is an attenuated NDV virus derived from a mesogenic strain (Csatary 825). It was reported that MTH-68/H was able to stimulate macrophages production in infected rats (Hrabak et al. 279-89) and to activate NF-?B pathway in carcinoma cells (Ten et al. 496-501). The oncolytic activity of MTH-68/H is due to the activation of apoptosis machinery via nitric oxide synthase (iNOS) induction (Hrabak et al. 279-89). This apoptosis activation appears to be p53-independent (Donehower et al. 215-21;Fabian et al. 2817-30). An in vitro comparative analysis of the anti-tumor effect demonstrated that MTH-68/H was the most potent IFN-a inducer in peripheral blood mononuclear cell (PBMC) (Apostolidis, Schirrmacher, and Fournier 1009-19). In the same study, in order to overcome side virus effect (important loss of weight), UV light inactivated MTH-68/H was injected via locoregional i.v. route for in vivo mice trial investigation. This inactivated strain exhibited a delay in tumor growth involving anti-cancer immune responses. MTH-68/H was also clinically tested to treat 4 patients with advanced high-grade glioblastoma multiforme (GBM) (Csatary et al. 83-93). Whereas conventional treatment failed, patients survived thanks to the oncotherapy (survival rate: 5 to 9 years) and, regularly receiving MTH-68/H injections, restarted to live normally. In another study, MTH-68/H was administered to 12-year old patient with high-grade glioma after inefficient post-operative radio- and chemotherapy. The oncotherapy was combined to an antiepileptic treatment (which also has antineoplastic properties) and resulted in 95% of tumor regression. Electron microscopy and immunohistochemistry confirmed NDV oncolytic activity in the cancer cells inducing apoptosis, as it has previously shown in vitro (Wagner et al. 731-43). However, the success of this trial was limited by three unexplained tumor apparitions, which required surgical intervention
? PV701 is a naturally attenuated strain purified from a mesogenic NDV virus. It has shown that when using this virus, anti-cancer response is induced by the stimulation of T-cells, macrophages and cytokine (like IFN) expression (Termeer et al. 316-23;Schirrmacher et al. 63-73). In phase I clinical studies, the injection of PV701 via intravenous route caused remarkable regression of different kinds of tumor (Lorence et al. 618-24). Using repeated desensitization, the tolerability to the increasing doses of the virus was improved. Total regression and/or cancer stabilization were observed in cohorts having respectively stable cancer following radiotherapy and cohorts with progressing cancer (Laurie et al. 2555-62) (Lorence et al. 157-67).
? 73-T strain (obtained after 73 passages through carcinoma cells) was used to produce oncolysates: in vitro preparations of plasma membrane fragments from NDV infected tumor cells. Based on vaccination principle, 73-T in oncolysates can play the role of adjuvant: infected tumor cells express at their external surface NDV F and HN proteins, this enhances their antigenicity. Using this system, Zorn, U., et al have demonstrated that, in vitro, the presence of 73-T is able to increase cellular cytotoxicity of PBMC by the induction of interferon- (IFN-a) and tumor necrosis factor-a (TNF-a) (Zorn et al. 225-35). Oncolysate immunotherapies have been undergone through clinical trial phase I/II. These studies were targeting melanoma (Cassel et al. 672-79) (Plager C), renal cell carcinoma (Kirchner, Anton, and Atzpodien 171-73) and breast or ovarian cancer (Mallmann, Eis, and Krebs 490-96). Unfortunately, results from these trials were not relevant either because of the lack of healthy subjects as control or the absence of any conclusive results regarding cancer treatment (Schirrmacher et al. 63-73). A phase II clinical study of NDV oncolysate (Cassel and Murray 169-71) has been also the subject of a long-term immunological effect analysis. This study involved postsurgical patients suffering from stage III malignant melanoma. It showed that adjuvant vaccination increased 15-years survival rate up to 55%. Analysis of the peripheral blood T cell population in these patients suggested that the efficient immune response against tumor was due the important presence of CD8+CD57+ subset (Batliwalla et al. 783-94).
Non-lytic strains are characterized by inactive glycoproteins, fusion (F) and hemagglutinin-neuraminidase (HN). However, they are able to cause an anti-tumor response by disturbing host-cell metabolism and stimulating the expression of cytokine and chemokine proteins (Schirrmacher et al. 63-73). Among these non-lytic strains, Schirrmacher, V. et al have suggested NDV-Ulster strain as a ï¿½safeï¿½ vaccine candidate against cancer. Like oncolysates, NDV-Ulster infected cancer cells were also used for the same vaccination purpose to enhance immune response against tumor. It is named autologous tumor-cell vaccine (ATV). ATV was clinically tested in colorectal cancer patients. Vaccination has improved DTH (delayed-type hypersensitivity) response especially by increasing tumor reactive T lyomphocytes (Lehner et al. 173-78;Liebrich et al. 703-10). Furthermore, a prospective randomized phase III trial including patients with hepatic metastases of colorectal cancer showed extended overall and metastases-free survival, but this was selectively in the case of colon cancer (Schulze et al. 61-69). Postoperative ATV-vaccination of 23 patients with GBM in a pilot study, showed encouraging results regarding the overall survival comparing to the control group and the improvement of DTH reactivity (Steiner et al. 4272-81).
Among the three pathotypes, lentogenic strains cause only a mild respiratory sickness and produce non-infectious virus particles. NDV-Ulster strain, described above, is one of the lentogenic strains used essentially for immunotherapy assays but not for virotherapy tests. LaSota strain was described as oncolytic agent against chemoresistant tumors via Type I TNF system induction (Mansour, Palese, and Zamarin 6015-23). It was also suggested as a good vaccination vector (Liang et al. 495-98). It has recently demonstrated that NDV-HUJ cleaves livin protein, a caspase inhibitor, in chemoresistant melanoma primary cultures. Cleaved livin becomes a truncated protein which contradictorily has a proapoptotic activity (Lazar et al. 639-46). This study suggested also that the interferon system does not play any role in NDV-induced oncolysis. The same findings were also described in the context of lung tumor cells treatment (Yaacov et al. 795-807). Clinical trials Phase I/II findings demonstrated a good tolerability to i.v injection of NDV-HUJ in patients with GBM (Freeman et al. 221-28).
A part of non-modified strains, some NDV strains were genetically modified in order to enhance, in different ways, their selectivity (Bian et al. 1359-69) and their efficiency especially to induce a stronger immune response (Fournier, Aigner, and Schirrmacher 1719-29). (1) A German group has shown that the tumor infection with a recombinant virus producing IL-2, rec-(IL2), activated T-cells and IFN-? production (Janke et al. 823-32). (2) Similarly, the same group generated another recombinant virus, rec(GM-CSF), expressing the granulocyte/ macrophage colony-stimulating factor. The activation of IFN-a was higher in the case of GM-CSF expression in PBMC and involved monocytes and plasmacytoid dendritic cells (Janke et al. 1639-49). (3) Furthermore, a recombinant fusogenic NDV expressing influenza NS1 protein, described as having IFN-antagonist and anti-apoptotic properties, was able to lyse human and mouse cancer cells. The expression of NSI protein increased IFN activation and as consequence prolonged overall long-term survival without any severe side effect (Zamarin et al. 697-706). (4) Using 3 Type I interferon-sensitive recombinant viruses (rNDV), Elankumaran, S. Et al have demonstrated that the selective replication of rNDV has allowed a differential regulation of IFN-a system and IRF-7 (interferon-regulatory factor 7) pathway. In a mouse model of human fibrosarcoma, important tumor shrinkage was observed without any side effect in mice (Elankumaran et al. 3835-44).
Findings of clinical trials using NDV virus with or without genetic modification are globally encouraging. Nevertheless, some hurdles are still limiting NDV oncotherapy development. These include safety concerns and the requirement of more investigation of long-term immune effect on treated patient.
b. Measles virus (MeV)
MeV belongs to the Morbillivirus genius. Wild type MeV strain is pathogen for human and can even be fatal for children especially in developing countries. Its incidence was decreased in industrialized world thanks to the administration of vaccine against measles disease. The Edmonston vaccine strain (MeV-Ed) is safe and has no cytopathic effect on normal cells. However, many studies have shown that it has a significant oncolytic effect on haematological (Bais et al. 186512) and solid malignancies in xenografts model including GBM (Phuong et al. 2462-69), breast and ovarian cancer (Peng et al. 732-38) (McDonald et al. 177-84), prostate cancer (Msaouel et al. 82-91). The molecular mechanism leading to cancer cells apoptosis by the attenuated MeV is poorly understood. However recent study suggested the involvement the reactive oxygen species (ROS) in this process. ROS is highly produced in MeV-infected ovarian cancer cells. This affects cell-cell contact by E-cadherin silencing (Zhou et al. 14-25). Furthermore, neutrophils are also considered to play an important role in vaccine MeV, but not the wild type, oncolytic activity. Indeed, when activated by the presence of this virus, neutrophils induced in vivo anti-tumor cytokine secretion (Zhang et al. 1002-10). A part of its oncolytic property, MeV was proposed as a vector for vaccine therapy. In this case, MeV vaccine can be a chimera virus which expresses envelop foreign protein (Spielhofer et al. 2150-59).
Binding of the hemagglutinin (H) to specific receptors insures MeV-host cells attachment. Two main cellular receptors are involved in MeV entry into host cell (Dhiman, Jacobson, and Poland 217-29): (1) CD46, a ubiquitous receptor in nucleated cells (but not erythrocytes) overexpressed on tumor cells (Dorig et al. 295-305) (Liszewski, Post, and Atkinson 431-55), and (2) the signaling lymphocyte activation molecule (SLAM or CDw150) which is exclusively expressed in cells from the immune system and predominantly on B and T cells (Veillette and Latour 277-85) (Hsu et al. 9-21). The differential expression of these two receptors in cells can explain the distinctive tropism between wild type and vaccine measles virus strains. Wild type strain can only enter into the host cells through the SLAM receptor, while the Edmonston strain recognizes both receptors but preferentially CD46. That is why MeV-Ed selectively replicates in tumor cells (Dorig et al. 295-305) (Yanagi, Takeda, and Ohno 2767-79). CD46 is ubiquitously expressed; nonetheless, the oncolytic effect is only triggered when a threshold of surface CD46 receptor density is reached in tumor cells (Anderson et al. 4919-26). As consequence, infected cells fuse forming syncytia and finally leading to cell death via apoptosis. In addition to these two receptors, a poliovirus receptor-like protein called PVRL4 (Nectin 4) belonging to the immunoglobulin superfamily was recently described as a MeV epithelial receptor (Muhlebach et al. 530-33) (Noyce et al. e1002240).
The attenuated phenotype seems to be due to the V protein (Nakatsu et al. 11996-2001). Wild type Measles virus is able to counteract the innate immune response thanks to the C and V proteins. These two non-structural proteins contain tyrosine-rich conserved regions, which are able to bind to STAT1 and 2 proteins respectively. This binding interferes with STAT (Signal Transducers and Activators of Transcription) phosphorylation and then antagonizes IFN signaling (Caignard et al. 351-62) (Devaux et al. 72-83). Mutations in the domain make V protein ï¿½blindï¿½ to STAT1 (Devaux et al. 348-56). V-mutated viruses indeed become attenuated but remain immunogenic in rhesus monkeys.
Despite virus proved oncolytic effect in human xenografts, non-modified MeV-Ed potency remains variable and sometimes not efficient in in vivo infection assays. For this reason, many engineered MeV-Ed strains were investigated in one hand to improve their oncolytic potency and their specificity against tumor and in another hand to be able to track MeV action during infection. For this purpose, multiple strategies were used to retarget and/or arm measles virus based on specific properties of tumor. The most common strategy involves the expression, in the context of MeV-Ed infection, of soluble marker peptides. The latters have the advantage of being stable, inert and not immunogenic. Genes encoding for these peptides were inserted into MeV genome.
CEA, the extracellular domain of carcinoembryonic antigen (Peng et al. 527-31;Peng et al. 4656-62), is one of those peptides expressed in the context of engineered MeV infection. The replication of MeV-CEA and viral gene expression can be followed by the measurement of CEA production in patient serum. The sodium/iode symporter (NIS), a membrane ion channel naturally concentrating the iode in theroid glands, is also used for oncolytic therapy. The expression of this protein by engineered MeV has allowed, in addition to the virus gene expression, the evaluation of MeV-NIS spread and localization (Dingli et al. 157-66;Dingli, Russell, and Morris, III 1079-86;Dingli et al. 1641-46) by following and dosing the administered radioisotopes 123I, 124I or Tc99m. These tracers are detected by a ?-camera based on radioiodide (Ahn 392-402). Combining radiotherapy and virotherapy is possible using 131I tracer which has by itself the capacity to kill tumor and thus to increase the oncolytic efficiency of MeV-NIS (Dingli et al. 1641-46).
? Preclinical trials were first made to assess the safety of engineered MeV treatment. Two models were used: a transgenic mouse and primateï¿½s model. (1) Rodents lack CD46 and SLAM receptors and thus were not suitable for this kind of MeV toxicity study. In order to mimic human clinical trial, transgenic mouse strain was generated in 1998 (Mrkic et al. 7420-27) expressing human CD46 receptor with the same human tissue distribution(Kemper C). Moreover, this strain does not express IFN type I, hence the name ï¿½(IFNARKO) CD46 Geï¿½ (Ge for genomic). MeV-CEA toxicology studies using (IFNARKO) CD46 Ge mice convincingly showed that, injected in peritoneal cavity (Peng et al. 1565-77) or central nervous system (CNS) (Allen et al. 213-20), MeV-CEA biodistribution remained limited to the part of the body where the virus was injected. No toxicity was observed in these trials. (2) Rhesus macaques (Macaca mulatta) were also used as a model to estimate neurotoxicity of MeV-CEA when injected in CNS. Macaques tolerated repeated intracerebral administration of a MeV with no evidence of toxicology, encephalitis, or other abnormalities (Myers et al. 690-98).
Similarly, MeV-NIS toxicity was tested using (IFNARKO) CD46 Ge model via intravenous route. In cyclophosphamide-pretreated mice (cyclophosphamide or CPA is a chemotherapy agent), MeV-NIS propagation was increased and prolonged first before its clearance by the immune system with no sign of toxicity. Clinical complications caused by chemotherapy were not enhanced (Myers et al. 700-10). Squirrel monkey was also used for the same purpose. Contrarily to the rhesus macaques, this primate species express truncated CD46 on their erythrocytes that are, like human ones, unable to agglutinate in the presence of MeV (Hsu et al. 6144-54). Like in transgenic mice model, MeV-NIS infection in CPA-pretreated squirrel monkey was prolonged then safely eliminated by the immune system without any extra clinical signs (Myers et al. 700-10).
? Clinical trials are currently in progress using these engineered viruses (Msaouel, Dispenzieri, and Galanis 43-53) (Galanis et al. 875-82). The most advanced assay is a Phase I trial testing MeV-CEA to treat ovarian cancer. 21 Adult patients are currently involved (18 years and older) (Cancer Clinical Trials Education Series) (Galanis;Galanis et al. 875-82;Phuong et al. 2462-69) with peritoneal persistent, recurrent or progressive ovarian cancer. They were treated with seven increasing intraperitonealy-injected dose levels: 103 to 109 TCID50. Viral replication is determined by the CEA titer in sera or peritoneal fluid. Interesting results showed no dose-limiting toxicity or immune response linked to MeV-CEA injection. Moreover, the oncolytic virus was not eliminated in urine and saliva and no increase of anti-CEA antibody was detected. Some clinical signs were observed including grade 1 and 2 fatigue, fever, anorexia and nausea (Galanis et al. 875-82). Two-thirds of patients showed disease stabilization during 55 to 277 days (median duration: 92.5 days). Last data collection of this 6 years trial was scheduled for July 2012.
MeV-CEA is also a subject of a phase clinical trial in MeV-immunized patients with recurrent GBM (Galanis) (Allen et al.). Increasing doses from 105 to 107 were administered to 40 patients divided into 2 groups. The first group underwent resection before MeV-CEA administration, while the second group was treated with this engineered virus via intratumoral way (catheter) before resection. Both groups are currently under study, which would be completed on June 2013. The goal of this study is to determine the maximum tolerated dose (MTD), to evaluate the oncolytic agent efficiency and safety and its associated symptoms. After this trial, patients will be regularly followed during 15 years.
Concerning MeV-NIS, recent phase I clinical trial was assessed involving patients suffering from multiple myeloma. MeV-Nis was co-administered with 123I (Galanis). It was initiated in 2007 and lasted 5 years. Two study parts were applied to two groups of patients depending on the treatment combination or not with CPA. During the first part, which was closed on December 2009, dose escalation MeV-NIS injection from to 106 to 109 TCID50 allowed MTD determination. In the second part, Patients were pre-treated CPA IV one day before MeV-NIS administration. The virus was injected in increasing doses until the MTD fixed in the first step. The goal of this part is to evaluate the efficiency of the oncolytic virus by following 123I isotope (level, distribution and oncolytic effect). No dose-limiting toxicity was observed. MeV-NIS replicated specifically in the tumor without anti-measles antibodies generation (Msaouel et al. 82-91).
In addition to MeV-CEA and MeV-NIS, many other manipulated measles viruses have been engineered during the last decades.
? In order to prove the possibility to modify MeV target, two engineered strains were generated in which H protein was replaced by one of the two hybrid proteins: H/EGF or H/IGF1. These hybrid proteins respectively contain H fused to the epidermal growth factor (EGF) or the insulin-like growth factor 1 (IGF1) (Schneider et al. 9928-36). These recombinant viruses were able to infect rodent cells lacking CD46 receptor but expressing EGF and IGF1 ones. Once proved, the same principle was adopted to retarget measles viruses in some studies. For instance, to enhance MeV selectivity to cancer cells, its fusion to host cells was made dependant on matrix metalloproteinases (MMP) (Springfeld et al. 7694-700), a cancer cell hallmarker (Hidalgo and Eckhardt 178-93). This was possible by the introduction in the F protein of a hexameric sequence, PQGLYQ, specifically recognized by MMP (Hartl et al. 918-26). An MMP inhibitor confirmed that MeV replication happened in a MMP-dependent manner. This recombinant virus was able to form syncytia in vitro in human fibrosarcoma cells (HT1080) and to delay in vivo xenografts growth in nude mice (Springfeld et al. 7694-700). For the same purpose, PQGLYA hexamer sequence was added to the fusion protein in an engineered MeV-MMPA1 virus to restrict its activation in liver tumor tissue (Muhlebach et al. 7620-29). This retargeted virus presents a potential safer MeV with high oncolytic potency for liver cancer therapy. In the same way, another tumor cells marker was investigated to increase MeV selectivity to cancer cells. It is the urokinase-type plasminogen activator receptor (uPAR) (Nakanishi et al. 724-32). Recombinant human and murine MeV viruses were engineered expressing uPA domain that has a high affinity to uPAR. Reprogrammed viruses were called respectively MV-h-uPA and MV-m-uPA (Jing et al. 1459-68). Their specific oncolytic activity was demonstrated in vitro and in vivo.
? In addition to retargeting viruses by foreign gene introduction in MeV genome, some recombinant viruses were armed to express proteins in order to improve the efficiency of anti-cancer therapy. This is, among others, the case of MV-PNP HblindantiCD20 virus (Ungerechts et al. 1506-16). A chimera protein containing a mutated H, which is unable to recognize CD46 and SLAM receptor, has replaced H protein of this virus. Mutated H was fused to two proteins: the first protein is added to retarget the recombinant virus. It is an antibody specifically recognizing CD20 receptors present in MCL (Mantel cells lymphoma). The goal of fusing the second protein to H is to arm MeV using PNP (the prodrug convertase E. coli purine nucleoside phosphorylase). The role of PNP is to improve FCR regimen anti-cancer efficiency by activating the prodrug fludarabine (Ungerechts et al. 10939-47). FCR regimens are triple cancer treatment including fludarabine, CPA and CD20 antibody Rituximab. While chemotherapy agents, failed to kill MCL xenografts treatment, a combined triple treatment adding MV-PNP-HblindantiCD20 virus with fludarabine-CPA enhanced the oncolysis efficiency (Ungerechts et al. 1506-16). Likewise, an armed measles virus was generated expressing PNP and prostate stem cell antigen (PSCA): MV-PNP-anti-PSCA (Bossow et al. 598-608). This virus efficiently improved pancreatic cancer xenografts treatment in the presence of fludarabine.
? Tropism of MeV was also adapted by miRNA7-mediated fusion protein suppression. MiRNA7expression is upregulated in normal brain cells but downregulated in GBM. The latter was targeted by a recombinant MeV whose genome harbors miRNA-7 target sequence at the 3ï¿½ non-coding extremity of F gene (Leber et al. 1097-106). This virus had an efficient oncolytic effect against gliobastoma xenografts and was safe for MeV-sensitive mice. Its spread was strictly limited in primary human brain resection expressing miRNA-7 (Leber et al. 1097-106).
A non-modified MeV-Edmonston Zagreb strain (MeV-EZ) was also investigated in phase I clinical trial (Heinzerling et al. 2287-94) involving patients with cutaneous T-cell lymphoma (CTCL). In the latter, CD46 receptor is expressed and T-helper2 cytokines are inhibited by the overexpression of IFN-? (Willers et al. 874-79). Therefore, it is considered as a good target of MeV oncolytic viruses. In this study MeV-EZ was injected via intratumoral route to IFN-a pretreated patients (CTCL 5 stage IIb or higher). Total of 16 injections were performed by one dose escalation. MeV doses were well tolerated. Among 5 patients, only one showed a completed CTCL lesion regression. Partial regression was also observed in 3 patients and finally no response was detected in one of these patients. Anti-MeV antibody titer increased in all MeV immunized patients (Heinzerling et al. 2287-94).
MeV vaccine lineage is not the only strain used for genetic manipulation. The HL strain (Kobune et al. 315-20) of wild type measles virus showed oncolytic effect on breast cancer cells which lacks SLAM receptors but present PVRL4 ones (Fabre-Lafay et al. 73). As this strain has naturally targets SLAM receptors, a genetically modified virus rMV-SLAMblind was recently generated to be unable to recognize them (Sugiyama et al.). Contrarily to the vaccine strain MeV-ED, it specifically binds to breast cancer cells via PVRL4 receptor rather than CD46. Its oncolytic property to kill breast cancer xenografts was proved in immunodeficient mice. Assays using MeV-seronegative monkeys showed tumor regression without any clinical signs (Sugiyama et al.).
In summary, measles virus strains and especially vaccine lineage derivatives are among the most advanced paramyxoviruses as oncolytic agents. A number of clinical trials are in progress. Further investigations will depend on results obtained from these studies. The most actual challenging hurdle remains to circumvent host-immune response, which can eliminate the administered virus. The use of non-human paramyxoviruses expressing MeV glycoproteins (Springfeld et al. 10155-63) is one of the approaches to address this challenge. For the same purpose, insertion of MeV retargeted glycoprotein into lentiviral vectors allowed chimera shielded measles viruses (Baum 1349-50) (Funke et al. 1427-36). In some clinical trials, either local treatment instead of intratumoral treatment or immunosuppression using CPA before MeV injection can increase the efficiency of virotherapy.
Actual clinical trials are limited to few cancer types. Findings demonstrating the MeV attenuation in the absence of V-STAT1 recognition (Devaux et al. 348-56) can be investigated to extend clinical trials to cover STAT-deficient tumor like lymphoma and myeloma (Sun et al. 570-76) (Wong et al. 28779-85).
c. Mumps virus (MuV)
Mumps virus is a non-fatal rubulavirus. It nevertheless causes an acute illness varying from low pathogenic symptoms like earaches, myalgia and slight fever, to more severe signs such as parotitis, meningitis, encephalitis and orchitis. Thanks to the V protein and like all rubulaviruses, MuV is capable to escape host cell immune response by proteasome-mediated STAT1degradation (Ulane and Horvath 160-66) (Parisien et al. 4190-98). Via an independent mechanism, MuV-V protein causes ubiquitylation and degradation of STAT3 which is responsible for cytokine activation and notably the interleukine-6 (Il-6). This degradation inhibits cytokine and oncogene signaling and would induce apoptosis in cancer cells (Ulane et al. 6385-93).
Mumps virus was one of the most ï¿½fashionableï¿½ potential oncolytic viruses from the 50s to the beginning of the 70s. A Japanese group demonstrated evidence of its oncolytic properties in clinical trials involving patients with various malignancies (Asada 1907-28;Okuno et al. 37-49). Asada et al. first showed that among 90 patients in terminal phase cancer, MuV administration in small amount was responsible at least clinical improvement for 79 almost without side effects (Asada 1907-28). Likewise, similar findings were observed in a second study testing Urabe MuV strain involving 200 patients with various cancers. However these researches in this field showed not sufficient effect to kill tumor comparing to chemotherapy or radiotherapy and were interrupted by the Japanese authorities (Asada et al communication 1994).
MuV was investigated in immunotherapy against gynecologic malignancies(Shimizu Y). In this study, 22 patients were initially pre-immunized via subcutaneous route by MuV in order to make their T helper cells reactive to the virus. Then, 10-fold higher MuV doses were injected locally or systemically to the vaccinated patients. Five of seven patients with local injections showed ascites or pleural fluid disappearance with no important clinical sign except fever. Nonetheless, MuV immunotherapy was not efficient in the case of patients with larger tumor.
Further studies have demonstrated the efficiency of MuV immunotherapy to cutaneous warts (caused by HPV, human papillomavirus) via intralesional (Horn et al. 589-94;Johnson, Roberson, and Horn 451-55) or intradermal (Dasher, Burkhart, and Morrell 373-79) injections (Mulhem and Pinelis 288-93) and to genital warts (King, Johnson, and Horn 1606-07). This field remains the most important application of mumps virus as a therapy vector. It is considered as a new immunotherapy approach to treat a wide range of warts in immune person and as an alternative to cryotherapy which efficiency was limited (Kwok et al. CD001781) (Gibbs and Harvey CD001781).
d. Sendai virus (SeV)
Sendai virus is a murine pathogen belonging to respirovirus genius. Its tropism is limited to the respiratory tract by the cleavage of its fusion protein by rodent lung cell proteases called ï¿½tryptase-like Claraï¿½ (Kido et al. 13573-79;Nagai 81-87). Since it is not pathogen for human, it was considered as a good vector for vaccine (Moriya et al. 8557-63) (Takimoto et al. 255-66) (Jones et al. 959-68) (Moriya C;Takimoto T;Jones BG) and gene therapy (Nakanishi and Otsu). With the emergence of oncolytic viruses, this good vector was suggested as a potential candidate for virotherapy against cancer.
For this purpose, retargeted SeV virus was engineered to have selective oncolytic activity on cancer cells (Kinoh and Inoue 2327-34). Like it was previously described for measles virus (paragraph II-2-b), SeV was also retargeted expressing a linker sequence sensitive to matrix metalloproteinase (MMP) in F protein. This sequence replaced F cleavage site, needed for the virus fusion and spread. In this virus matrix protein was deleted to impair its capacity to bud (Kinoh et al. 1137-45). In vivo, SeV-MMP infection was efficient and restricted to cells expressing MMP2 or MMP9. In vivo, this recombinant virus was able to selectively replicate and kill human transplanted tumor in nude mice. In both cases: in vitro and in vivo, infected tumor cells formed syncytia and died. However, this oncolytic effect remains limited due to the small amount of tumor cells expressing MMP or the low level of MMP expression. Hence it was important to identify species of proteases specific to the different cancer types and to optimize anti-tumor efficiency.
In this context, 14 amino acids residue in the cytoplasm domain of F protein was first truncated in order to enhance its fusogenic activity and then its ability to form syncytia (Kinoh et al. 392-403). Second, the cleavage site of F was replaced by another linker sequences sensitive to urokinase-type (uPA) and tissue type (tPA) plasminogen activators. These proteases, and especially uPA, are expressed in human cancer cells. Syncytia formation by the recombinant Sendai viruses expressing F(uPA) and F(tPA) was enhanced in vitro in the presence of recombinant human uPA or tPA respectively. uPA-targeted SeV vectors were able to selectively and efficiently spread in various tumor cells including prostate, renal, pancreatic and lung cancers with a high oncolytic activity. This study suggested SeV-F/uPA as a vector for oncolytic cancer therapy (Kinoh et al. 392-403).
A subsequent study has armed this recombinant virus (expressing 14 amino acids truncated cytodomain F with linker sequence sensitive to uPA) named ï¿½BioKnifeï¿½ in the goal to improve its oncolytic activity against Glioblastoma multiforme (Hasegawa et al. 1778-86). Arming BioKnife virus consisted in the insertion in its genome of IFN-ï¿½ gene (BioKnife-IFNï¿½). The study made evident uPA-dependant oncolytic effect of the engineered virus, which was synergic to the expression of IFN-ï¿½ gene in rat GM cells expressing uPA and uPAR (9 L-L/R cell lines). The latter showed high sensitivity to BioKnife virus translated by important cell-cell fusion. This increased IFN-ï¿½ expression, which in turn, enhanced virus fusion property. Likewise, significant synergy was also demonstrated between Bioknife and IFN-ï¿½ transgene in vivo using rat orthotopic brain GM model with 9L-L/R.
An extended investigation of BioKnife virus has focused on the treatment of malignant pleural mesothelioma (MPM), tumor highly expressing uPAR receptor (Shetty et al. L972-L982). Two orthotopic xenograft models of human MPM were recently used: H226 (epithelial subtype) and MSTO-211H (biphasic subtype) (Morodomi et al. 769-77). In these experiments, repeated administrations into thoracic cavity of BioKnife prolonged mice survival and caused the tumor death in vivo. BioKnife oncolytic effect was also observed in vitro using the same MPM cell lines. Interesting findings have figured out details about the mechanism of action of Bioknife against tumor: (1) even in the absence of uPA expression by MPM cells, the recombinant virus was able to activate uPA production in (RIG-I/NF-?B)-dependant manner and (2) this subsequently causes tumor cell death via apoptosis.
In a completely different purpose but still in the context of oncotherapy, SeV virus was also investigated in nanotechnology-based bioassays (Dudu, Rotari, and Vazquez 9). These assays are used in cancer diagnosis by identifying tumor cell specific markers like EGFR (epidermal growth factor receptor). Classical methods to deliver specific antibodies against EGFR marker using nanocarriers presented numerous disadvantages such as endocytosis disruption and subcellular damage of endosome structure (Torchilin 2333-34). SeV has the advantage to fuse to the membrane releasing particles in the cytoplasm. Thus, to overcome delivery problems during the Q-dots step, a chimeric Sendai virus was generated fused to liposomes which are previously suggested as nanocarriers (Dudu et al. 2293-300). In this recent study, targeted Qdots were efficiently delivered into Human glioblastoma and medulloblastoma samples via SeV virus-based liposomes (VBL) in vitro. Results have been given by fluorescence microscopy and Transmission Electron Microscopy and have shown an increase of specific VBL intracellular labelling of EGFR. These findings pointed out the utility of a chimera SeV to characterize tumor in order to develop appropriate treatment regimens.
Examples of Sendai virus applications in oncotherapy fields underlined its property as a good vector to express or deliver molecules in cancer cells for therapy or diagnosis. This is in the context of engineered SeV. Its safety makes it an attractive virus to treat tumor in clinical cancer settings.
III. Summary- Challenging ï¿½paramyxo-virotherapyï¿½ against cancer
In the last decades, we have assisted to the boom of oncolytic researches using natural (table1) and engineered (table2) paramyxoviruses. However, only some of them are today investigated through clinical assays (table3). Most advanced studies to treat human malignancies concern reprogrammed measles virus and naturally attenuated Newcastle disease virus. Clinical virotherapy, often combined to other cancer therapy regimens, appears to enhance oncolytic efficiency. Oncolytic paramyxoviruses were also investigated in veterinary medicine (Patil et al. 3) such as canine distemper virus (CDV), a morbillivirus similar to MeV explored in therapy against lymphoma (Suter et al. 1579-87).
Current studies are faced to a number of obstacles and try to design new strategies to overcome them. Different host-cell barriers can block delivery of the virus during oncotherapy (Ferguson, Lemoine, and Wang 805629) including neutralizing antibodies, agglutinating complement and antiviral cytokines. Virus oncolytic efficiency can also be impaired by its elimination or its filtration in non-targeted tissue like liver and spleen. One of the disadvantages of oncotherapy studies is the divergence between observations from experimental and clinical trials. Since animal models are usually used, difference of body reactions can be observed comparing to human. Besides, in experimental models, it is frequent to use immune-deficient animals like SCID, nude or athymic mice without consideration to the acquired humoral immunity. Also, the use of laboratory tumor cells can also mislead researches that would further fail to be validated in clinical trials using human patients. Finally, virus toxicity and related side effects remain one of the most challenging issues in oncolytic studies using viruses.
Despite hurdles, outcomes from oncolytic virotherapy are globally promising. They anyway need to be validated by long-term effect evaluation and larger patient population studies targeting a broader spectrum of tumor. In different words, virotherapy against cancer is still a young field that requires time to grow before joining existing old cancer therapies, for hopefully better treatment efficiency.
Virus species Strain Properties and action mechanisms
NDV Lytic strains : syncytia formation ? cancer cell lysis MTH-68/H activates NF-?B and apoptosis in carcinoma cells
induces IFN-a in PBMC
PV701 Induces anti-cancer response by stimulating of T-cells, macrophages and cytokine expression
73-T Oncolysates induce IFN-a and TNF-a
Non-lytic strains: cytokine expression and interference with host-cell metabolism NDV-Ulster ATV same principle as oncolysates
LaSota induces of TNF-a signaling
NDV-HUJ Apoptosis via livin protein activation in melanoma primary culture
MeV MeV-Edmonston Zagreb (EZ) Attenuated strain targeting CTCL cancer as it lacks IFN system
Table1: Non-modified oncolytic paramyxoviruses
Virus species strain Modification ? advantage
NDV rec-(IL2) Expression of IL2 ? activation of T-cells and IFN-? production
Rec(GM-CSF) Expression of GM-CSF ? IFN-a induction
rNDV-NS1 Expression of influenza NS1 ?increase of IFN activation
rNDV-IRF7 Expression of IRF7 ? IRF7 pathway activation and IFN-a regulation
MeV Vaccine strain MeV-CEA Expression of CEA ? evaluation infection level
MeV-NIS Expression of NIS ? evaluation of infection level and distribution
MeV-IGF1 and MeV-EGF H is fused to IGF1 and EGF respectively ? enhancement of selectivity to cancer cells
MeV-MMP and MeV-uPA Chimera H protein having recognition sequences to MMP and uPA receptors respectively ? targeting cancer receptors
MV-PNP-HblindantiCD20 Chimera H protein : (1) unable to recognize SLAM receptors (2) fused to CD20 antibody, a cancer marker (3) expressing PNP activating a chemotherapy agent ? selective action on cancer cells + enhancement of chemotherapy efficiency
MV-PNP-anti-PSCA Fully retargeted H protein expressing PNP and exclusively recognizing PSCA marker ? selective action on cancer cells + enhancement of chemotherapy efficiency
MeV-miRNA7 Fusion protein harbours a miRNA7 target ? virus entry impaired in normal cells overexpressing miRNA7 / selective infectivity
Wild type strain MeV-HL/SLAMblind Mutated H unable to recognize SLAM receptors ?Binding to cancer cells via PVRL4 receptor
SeV SeV-MMP, SeV-tPA and SeV-uPA Chimera H protein having recognition sequences to MMP, tPA and uPA receptors respectively ? targeting cancer receptors
BioKnife-IFNï¿½ Chimera H protein having binding sequence to uPA + expression of IFNï¿½? enhancement of selectivity and efficiency of SeV oncolytic effect
Table2: Engineered oncolytic paramyxoviruses
Virus species Strain Cancer type Way of administration Clinical trial phase
NDV MTH-68/H GBM Locoregional i.v Phase I
PV701 Different solid cancers (colorectal, renal, breast and nonï¿½small-cell lung carcinoma) i.v
previous chemotherapy treatment in different regimens Phase I
73-T Stage III malignant melanoma oncolysates s.c. injection Phase II
NDV-Ulster Colorectal cancer with liver metastases i.d Prospective randomized phase III
GBM i.d injection a week after radiotherapy Pilot study
NDV-HUJ GBM i.v Phase I/II
MeV MeV-CEA ovarian cancer i.v Combined to CPA phase I
recurrent GBM i.p phase I/II
MeV-NIS multiple myeloma i.v /123I coadministration
CPA patient pretreatment phase I
MeV-EZ CTCL stage IIb or higher i.t
patients pretreated with IFN-a Phase I
Table3: most relevant oncolytic paramyxoviruses in clinical trials
CDV: canine distemper virus
HPV: human papillomavirus
MeV: Measles virus
MeV-ED: Edmonston strain of measles virus
MuV: Mumps virus
NDV: Newcastle Desease
SeV: Sendai virus
TPMV: Tupia paramyxovirus
Proteins and receptors:
EGF: epidermal growth factor
EGFR: epidermal growth factor receptor
F: fusion protein
H: hemagglutinin protein
HN: hemagglutinin-neuraminidase protein
GM-CSF: granulocyte/macrophage colony-stimulating factor
IRF: interferon-regulatory factor 7
IGF1: insulin-like growth factor 1
iNOS: nitric oxide synthase
MMP: matrix metalloproteinases
PSCA: prostate stem cell antigen
PVRL4: poliovirus receptor-like protein
Ras: rat sarcoma
SLAM: signaling lymphocyte activation molecule
STAT: Signal Transducers and Activators of Transcription
TNF: tumor necrosis factor
tPA: tissue type plasminogen activator
uPA: urokinase type plasminogen activator
uPAR: urokinase type plasminogen activator receptor
Immune cells and Cancer types :
CTCL: cutaneous T-cell lymphoma
GBM: glioblastoma multiforme
MPM: malignant pleural mesothelioma
PBMC: peripheral blood mononuclear cell
DTH: delayed-type hypersensitivity
MOI: multiplicity of infection
MTD: maximum tolerated dose
PFU: plaque-forming unit
TCID50:50% tissue culture infective dose
Way of injection:
ATV: autologous tumor-cell vaccine
CEA: carcinoembryonic antigen
FCR regimens: treatment combining treatment CPA and CD20 antibody Rituximab
NIS: sodium/iode symporter
PNP: purine nucleoside phosphorylase
SCID: severe combined immunodeficiency
VBL: virus-based liposomes