Clinical Application of Oncolytic Viruses: A Systematic Review

Leveraging the immune system to thwart cancer is not a novel strategy and has been explored via cancer vaccines and use of immunomodulators like interferons. However, it was not until the introduction of immune checkpoint inhibitors that we realized the true potential of immunotherapy in combating cancer. Oncolytic viruses are one such immunotherapeutic tool that is currently being explored in cancer therapeutics. We present the most comprehensive systematic review of all oncolytic viruses in Phase 1, 2, and 3 clinical trials published to date. We performed a systematic review of all published clinical trials indexed in PubMed that utilized oncolytic viruses. Trials were reviewed for type of oncolytic virus used, method of administration, study design, disease type, primary outcome, and relevant adverse effects. A total of 120 trials were found; 86 trials were available for our review. Included were 60 phase I trials, five phase I/II combination trials, 19 phase II trials, and two phase III clinical trials. Oncolytic viruses are feverously being evaluated in oncology with over 30 different types of oncolytic viruses being explored either as a single agent or in combination with other antitumor agents. To date, only one oncolytic virus therapy has received an FDA approval but advances in bioengineering techniques and our understanding of immunomodulation to heighten oncolytic virus replication and improve tumor kill raises optimism for its future drug development.


Introduction
Enhancing the body's own response to malignant cells through immune stimulation has been a vigorous focus of recent cancer research. Oncolytic viruses (OVs) are one such tool. These viruses are naturally occurring or can be modified to selectively infect and destroy cancer cells. In addition, there is evidence OVs can stimulate the host's immune response to combat tumors [1]. Multiple viruses are currently under investigation including herpesvirus, adenovirus, poxvirus, picornavirus, and reovirus as possible oncolytic treatments. In 2015 talimogene herparepvec was the first OV approved by the Food and Drug Administration (FDA) for human use in the USA [2] and there are a number of other OVs currently in phase III testing. Here we discuss how OVs have been adapted to destroy cancer cells and summarize the clinical data on OVs currently under investigation.

Methods
In our systematic review we collected all published clinical trials that utilized oncolytic viruses. Using Pubmed, we first narrowed our search to include "Clinical Trials" and "Humans." Subsequent "oncolytic vaccine". Our initial search returned 120 articles. This preliminary search also captured reviews, proposed study protocols, and opinion pieces, which were excluded (26 in total). Papers not available in full text for review, or with no available English translation were excluded (4). Last, we found a limited number of articles that described different facets of the same trial, and these redundant articles were excluded (4). One unpublished study was identified in clinicaltrials.gov. In summary, we used 86 total trials, including 60 phase I, 5 phase I/II, 19 phase II, and 2 phase III trials for our review. See Figure 1 for a schematic representation of systematic review of OV clinical trial data and exclusion criteria.

Mechanism of Action
OV therapy relies on a two-part process of selectively infecting tumor cells, and then inducing antitumor activity through release of tumor antigens and immune stimulation ( Figure 2).

Mechanism of Action
OV therapy relies on a two-part process of selectively infecting tumor cells, and then inducing antitumor activity through release of tumor antigens and immune stimulation ( Figure 2).

Targeting Cancer Cells
OVs can target and destroy cancer cells by taking advantage of a tumor's unique cellular activity. Strategies for infection and destruction of the tumor cells vary among OVs. Cancer cells can have altered or entirely absent signaling pathways that normal tissues use to defend against viral infection.
As an example, protein kinase receptor (PKR) and its associated interferon (IFN) pathway is integral to viral clearance, but is often underexpressed in certain cancers; low IFN expression renders the cancer cells susceptible to viral attack [3]. Another mechanism centers on the increased metabolic activity of tumor cells and rapid replication; for virally infected tumor cells this leads to increased viral replication and rate of lysis compared to normal tissue [4]. This situation enhances viral activity to destroy tumor cells.
Naturally occurring viruses already contain many strategies for evading host immunity and can take advantage of decreased defenses to target cancer cells. These cells often overexpress receptors or other proteins that the viruses already use you gain entry to normal cells. HSV-1 variants use nectins and herpes virus entry mediator (HVEM), expressed on cancer cell membranes, to gain entry. Some melanomas and carcinomas have increased expression of nectins and HVEM [5]. Measles virus utilize the CD46 receptor, which is often overexpressed in certain cancers [6].
Other OV's are enhanced viruses that have been tailored to increase their affinity for malignant tissues. They are engineered to target proteins overexpressed in cancer cells. One such example is thymidine kinase (TK). A strain of HSV has been modified to favor tumor cells due to their high TK

Targeting Cancer Cells
OVs can target and destroy cancer cells by taking advantage of a tumor's unique cellular activity. Strategies for infection and destruction of the tumor cells vary among OVs. Cancer cells can have altered or entirely absent signaling pathways that normal tissues use to defend against viral infection.
As an example, protein kinase receptor (PKR) and its associated interferon (IFN) pathway is integral to viral clearance, but is often underexpressed in certain cancers; low IFN expression renders the cancer cells susceptible to viral attack [3]. Another mechanism centers on the increased metabolic activity of tumor cells and rapid replication; for virally infected tumor cells this leads to increased viral replication and rate of lysis compared to normal tissue [4]. This situation enhances viral activity to destroy tumor cells.
Naturally occurring viruses already contain many strategies for evading host immunity and can take advantage of decreased defenses to target cancer cells. These cells often overexpress receptors or other proteins that the viruses already use you gain entry to normal cells. HSV-1 variants use nectins and herpes virus entry mediator (HVEM), expressed on cancer cell membranes, to gain entry. Some melanomas and carcinomas have increased expression of nectins and HVEM [5]. Measles virus utilize the CD46 receptor, which is often overexpressed in certain cancers [6].
Other OV's are enhanced viruses that have been tailored to increase their affinity for malignant tissues. They are engineered to target proteins overexpressed in cancer cells. One such example is thymidine kinase (TK). A strain of HSV has been modified to favor tumor cells due to their high TK expression [7]. TK is an enzyme used for DNA synthesis and repair during cell replication, is therefore highly expressed on tumor cells. This engineered variant of HSV holds a TK-knock out. The knockout of TK results in preferential infection and destruction of cancerous tissues.

Direct Tumor Cell Lysis
After target infection of the tumor cell occurs, lysis of the tumor cell leads to the release of viral particles, cytokines, and other cellular contents, and a secondary response to this lysis ensues. Within this system, there is direct killing of surrounding cells by the release of these cytotoxic elements, which include granzymes and perforins. The cellular contents ATP, uric acid, and heat shock protein are known as danger-associated molecular patterns (DAMPs), which cause local inflammation. DAMPs and cytokines attract natural killer (NK) cells to the surrounding tissue. In addition, viral particles have protein sequences that can stimulate the immune system, known as pathogen associated molecular patterns (PAMPs).
OVs can also be engineered to deliver suicide genes to enhance direct tumor cell lysis. These genes encode enzymes that can convert prodrug into an active form to destroy tumor cells. ProstAtak (aglatimagene besadenovec) is an OV engineered to express HSV TK in infected cells; after injection, valcyclovir is administered that halts DNA replication in the tumor cell.

Antitumor Activity
OV antitumor activity depends on both direct malignant cell destruction as well as a stimulating systemic antitumor immunity. This immunity is induced when an infected cell is lysed. The viral antigens and cellular components are liberated into the cancer microenvironment. Release of DAMPs, PAMPs, and cytokines lead to maturation of antigen presenting cells (APCs), including the dendritic cells (DC). By these processes, tumor cells that had previously evaded the immune system can now be recognized and targeted by the immune system; destruction pattern recognition receptors (PRRs) on immune cells are stimulated to recognize these PAMPs and DAMPs. PRRs specific to dendritic cells (DC)s within tumors can be directly stimulated by OV infection [8]. The subsequent activation of APCs then recruit CD4+ and CD8+ cells to destroy cells expressing viral antigens on tumors. One such example, in the activation antiviral CD8 T cells by reovirus lead to tumor regression [9]. This recognition of tumor antigens by the immune system is key to tumor destruction at distant sites that were not infected with the OV [10].
HSV and reovirus can prime antitumor immunity by stimulating T cell activity to destroy neighboring tumor cells [11,12]. Another genetically modified variant of HSV, Talimogene herparepvec (T-vec), has the GM-CSF cistron inserted into its genome. This modification enables high production and release of GM-CSF when the tumor cells are destroyed, resulting in enhanced DC recruitment and antigen presentation [13].
In addition to targeting the malignant cells themselves, OVs can destroy tumor vasculature. Some OVs are engineered to target angiogenesis, and other viruses inherently target the destruction of tumor vasculature. HSV and vaccinia viruses are one such example. These viruses depend on high expression of fibroblast growth factor and vascular endothelial growth factor (VEGF) for replication to target new tumor vessels [14,15]. The vesicular stomatitis virus (VSV) has been shown to infect blood vessels surrounding a tumor and cause thrombosis of its vessels [16].

Novel Trends in OV Advancements
Much of the emerging research for OVs focuses on how to amplify their immunogenicity and tumor destruction through insertion or exchange of human genes. These genes can improve cancer cell destruction through improved entry into cells, direct destruction of cancer cells, or recruitment of the patient's immune system. These emerging viruses are particularly focused on recruitment of the host's immune system to destroy the tumor.
One such virus in an ongoing phase 1 clinical trial utilizes a strain of HSV-1 to target glioblastomas (NCT02062827) [17]. This OV was modified to express IL-12, which improves the OV's efficacy in destroying tumor cells and in the surrounding microenvironment. The release of IL-12 with tumor lysis leads to recruitment of dendritic cells, monocytes, and macrophages to the area for increased tumor destruction. The release IL-12 also exhibits an anti-angiogenic effect, which slows the growth of blood vessels for the tumor and decreases tumor growth.
TILT-123 is another exciting example of an OV equipped with proteins to enhance its anti-tumor activity. TILT-123 is an adenovirus engineered to express TNF-alpha and IL2 (NCT04217473) [18]. While infection with the OV can lead to tumor destruction as previously discussed, the addition of IL2 and TNF-alpha increases the immunogenicity of the therapy. Expression of TNF-alpha and IL-2 attracts T cells to the tumor and also enhances their infiltration into the cancer cells [19]. NCT04217473 is the first clinical trial to evaluate the safety of TILT-123 s in patients with advanced melanoma, enrolling patients in early 2020.
In addition to studies currently underway combining OVs with simultaneous checkpoint inhibitor therapy administration, viruses are being engineered to produce the antibodies themselves. An HSV-1 virus has recently been developed for use in glioblastoma therapy, which can express PD-1 antibodies [20]. While this construct has only been tested in murine models as of now, it is a promising example of enhancing OV tumor activity through gene insertion.

Augmentation of Immune Checkpoint Inhibition. "Making Cold Tumors Hot: Immune Checkpoint Inhibitor-OV Combination Therapy Trials"
Remarkable progress has been made in the development of immunotherapy to treat cancers in the last decade. One type of immunotherapy is immune checkpoint inhibitors (ICBs), which are monoclonal antibodies that block receptor-ligand interactions that negatively regulate both innate and adaptive immunity. Prominent examples of ICBs include those that target cytotoxic T lymphocyte antigen 4 (CTLA-4) and programmed cell death protein 1 (PD-1) [21]. To date, multiple ICBs have been approved by the FDA to treat cancers; for example, ipilimumab (blocks CTLA-4), nivolumab (blocks PD-1), and pembrolizumab (blocks PD-1).
Combination therapies that use ICBs and oncolytic viruses are attractive. The oncolytic virus can recruit tumor-infiltrating lymphocytes into immune-deficient tumors and trigger the release of soluble tumor antigens, danger signals, and proinflammatory cytokines, which further increase T cell recruitment and promote immune cell activation [10,22]. Viral infection also increases the expression of CTLA-4, PD-1, and other immune checkpoint molecules, which would usually block T cell activation (and therefore antitumor immunity) but also sensitizes tumors to ICBs [23,24]. Non-clinical studies using B16-F10 melanoma demonstrated that localized injection of tumors with oncolytic Newcastle disease virus induced infiltration of tumor-specific CD4+ T cells and CD8+ T cells into both the injected tumor and distant tumors; this action increased the sensitivity of the tumors to systemic CTLA-4 blockade [10]. In another non-clinical model of triple-negative breast cancer, an oncolytic Maraba virus had activity as a neoadjuvant therapeutic, and it sensitized otherwise refractory tumors to ICB [25]. Other oncolytic viruses, such as B18R-deficient vaccinia virus [26] and vesicular stomatitis virus, express a library of melanoma antigens [27] and also showed significant (p < 0.05) therapeutic benefit in combination with ICB.
There is a strong mechanistic rationale for using combination oncolytic virus and ICBs since immune activation in the tumor environment as well as neo antigen production from oncolytic virus can improve efficacy of ICBs.
Ribas et al. confirmed that OV can potentially improve efficacy of anti PD1 immune checkpoint inhibitors. They evaluated safety and efficacy of combination T-VEC (talimogene laherparepvec) and pembrolizumab in a Phase 1b melanoma clinical trial (n = 21) and confirmed 62% objective response rate. The investigators found that OV can favorably alter tumor microenvironment (elevated CD8+ T cells, upregulation of INF-gamma, and increased expression of PD-L1 protein) resulted in improved response to the immune checkpoint inhibitor [28].
Improved efficacy was shown in Phase 2 clinical trial of combination oncolytic virus and CTLA-4 targeting ICBs [29]. Chesney et al. evaluated the combination T-VEC and Ipilimumab in a randomized phase II melanoma study (N = 198). 39 % patients had objective responses in the combination T-VEC/Ipilimumab arm while only 18% objective response was observed in the single agent ipilimumab arm (95% CI, P = 0.002). Of note, the responses were seen in injected lesions but also reported in distant visceral lesions in 52 % patients in the combination arm vs. only 23 % in the single agent ipilimumab arm.
A novel oncolytic virus called Coxsackievirus A21 (CAVATAK) shows synergy when combined with immune checkpoint inhibitors [30]. The CAVATAK virus makes entry to cancer cells via ICAM1, which is often abundant on tumors. A clinical trial evaluating cavatak virus in combination with ipilimumab showed 50% objective responses in melanoma patients. Similarly, a trial evaluating combination cavatak virus with pembrolizumab in advance solid tumors reveled no dose limiting toxicities (DLT) [31]. A phase 1b Keynote 200 clinical trial of combination cavatak + pembrolizumab in advance non-small cell lung cancer and bladder cancer confirmed no DLT; 12.8% grade 3 treatment-related adverse events were noted, however no grade 4 or 5 toxicities were seen [32,33]. Table 1 summarizes combination immune checkpoint and oncolytic virus clinical trials.

Modes of OV Delivery
Depending on both the OV and cancer, virus administration can be delivered using a systemic or more targeted approach.

Intralesional Route
A locoregional approach has been used for melanoma, prostate, and gliomas. Intradermal injections of T-vec is employed for melanoma, intracavitary injection for gliomas, and intraperitoneal injection for treatment of ovarian cancer are a few examples. Depending on the location and accessibility of the tumors the virus can be delivered once (i.e., to a glioma cavity during surgery) or rely on repeated injections such as in melanoma [35]. The efficacy of intratumoral injection can extend beyond the lesions injected. Studies report evidence of tumor lysis in satellite lesions not injected with virus [36]. OV delivery can also employ intravenous (IV) release, through either peripheral IV injection, or more targeted, through hepatic artery infusion for liver metastasis [37].
Intralesional injection can be limited by the extracellular matrix that surrounds otherwise accessible tumors. In desmoplastic tissues, the spread of the virus can be limited by excessive fibrin and connective tissues [38]. Agents such as bacterial collagenase have been used to increase the spread OV injection for local tumors such as melanoma, which demonstrated some efficacy [39]. However, collagenase cannot be used systemically, which limits its application. One of many systemic agents under investigation includes the antihypertensive agent losartan, which reportedly decreased collagen deposition, increased blood flow as well as uptake of the cytotoxic agent in pancreatic cancer [40].
Vascular injection also has limitations. Tumor vasculature is markedly disorganized and more unstable than normal vessels [41]; this property makes vascular injections less reliable for downstream delivery of the virus to surrounding cancer cells. However, currently there are several avenues under investigation to stabilize these tissues and enhance delivery of cancer immunotherapy; some examples include the use of nitric oxide [42] and VEGF [43].

Intravenous Route
Intra lesional OV administration is currently by far the most commonly used delivery method. However many trials are also evaluating intravenous route of administration [44,45] due to its inherent advantages. Some of the advantages of IV route includes ease of administration, standardization of administered drug dose, potential for multiple and long term administration, and wider application of IV based treatment especially in smaller community clinics [46]. The key detriment to IV administration so far has been development of neutralizing antibodies and clearance of OV [47][48][49]. There are at least eight clinical studies that we can confirm which used IV route for OV administration.

Phase I Trials
In our systematic review we have summarized 59 phase I trials investigating 36 OVs. Phase I trials seek to demonstrate safety and dosing. Of those 36 viruses, many did not proceed past this initial phase of investigation, or were altered in some way to improve efficacy. It is not feasible to summarize all Phase I trials here, but a summary of all reviewed trials can be found in Table 2 Here we discuss a few OVs of interest, many of which show promise and have progressed to Phase II investigations. Study used intratumoral injection at different dose levels in 30 patients. AEs: fever, and injection site reaction. Three patients with SD. Some evidence of tumor necrosis was seen on biopsy which was strongly positive for HSV.
G207 (HSV-1) GBM 9 Intratumoral injection of G207 given 24 h prior to radiation. 67% with SD or PR. 3 patients with measurable response to radiation. No HSV encephalitis occurred.
A herpes oncolytic virus can be delivered via the vasculature to produce biologic changes in human colorectal cancer [54] (Fong 2009).

NV1020
Metastatic colon cancer in liver previously refractory to chemotherapy (5 FU and leucovorin)

12
Study used IV injection of NV1020 into the hepatic artery. CEA levels dropped in patients, and one patient saw 75% reduction of tumor volume. Study reports 7 with SD, 3 with PD, 2 patients with reduction in tumor size at 28 days. Subsequent chemotherapy was hepatic artery injection of floxuridine with dexamethasone. All patients had PR. Results confounded by varying systemic chemotherapy regimens (7 had irinotecan, 2 with oxaliplatin and 3 got both).
Intradermal injection of Newcastle disease virus-modified autologous melanoma cell lysate and interleukin-2 for adjuvant treatment of melanoma patients with resectable stage III disease [55] (Voit 2003

RT3D (Reovirus Type 3 Dearing) Advanced solid tumors 36
Study used combination cyclophosphamide and RT3D (intravenously) to decrease neutralizing antibodies to the virus.
While it was well tolerated, it did not reduce NARA (neutralizing antireovirus antibody) titer.
A phase I study of the combination of intravenous reovirus type 3 Dearing and gemcitabine in patients with advanced cancer [60] (Lolkema 2011).

RT3D (Reovirus Type 3 Dearing)
advanced solid cancers 16 First study to combine IV Reovirus with chemotherapy. AE were similar to previous studies with fever, nausea/vomiting, and chills. Protocol revised because there were grade 3 rises in LFTs (but these patients were also taking acetaminophen.) Reovirus may exacerbate gemcitabine-related liver toxicity. Clinical response was best in a patient with nasopharyngeal cancer, but OR minimal. Protocol used escalating doses of radiotherapy followed by intratumoral injections of RT3D. AE included flu-like symptoms (grade 2 or less) as well as neutropenia and lymphopenia. Low dose group had 2 with PR and 5 with SD, and in the high dose group 5/7 had PR and 2/7 with SD.
No viral shedding seen in bodily fluids.

Ad5/3-(delta)24
Advanced solid tumors resistant to chemotherapy 21 Study used intratumoral injection with adenovirus followed by cyclophosphamide treatment in different dosing groups. AE were mostly grade 1-2 constitutional symptoms. The one year PFS and OS was increased compared to traditional chemotherapy resistant cancers. Study reports 8/12 patients with RECIST response: 2 with MR, 6 with SD, and 4 with PD.
Study reported 3 patients with a decrease in CA-125 levels at 1 month.  A phase I trial of intravenous infusion of ONYX-015 and Enbrel in solid tumor patients [83] (Nemunaitis 2007).

ONYX-015
Advanced cancers 9 Nine patients divided in 3 groups received IV infusion of onyx-015 of varying doses, with a dose of Enbrel. Study reported 4/9 with SD, but no regression was seen. AE were mild. Circulating viral DNA was higher when virus infusion is given in combination with Enbrel A phase I study of Onyx-015, an E1B attenuated adenovirus, administered intratumorally to patients with recurrent head and neck cancer [84] (Ganly 2000).

ONYX-15
Recurrent head and neck cancer 22 Study used single intratumoral injection that was well tolerated. Grade 1-2 constitutional symptoms were most common. Study reported no OR by RECIST criteria, but evidence of tumor necrosis seen on MRI in 5 patients with questionable PR was noted.

H103 (Adenovirus expressing HSP70) Advanced solid tumors 27
Study used intratumoral injection of virus in a dose escalation phase I study. Rare grade III fever and grave IV thrombocytopenia at high doses was observed. Most AE were fever, and a local reaction at the injection site, as well as thrombocytopenia, and depressed leukocyte and lymphocyte counts. Study reported 48% of patients had at least SD or better, and 11% had CR or PR.

Ad5-CD/TKrep Prostate 15
Patients had intraprostatic injection of adenovirus with cytosine deaminase and HSV thymidine kinase genes. 2 days later the patients received 5-fluorocytosine and valganciclovir prodrug for up to 4 weeks along with radiation. PSA 1 /2 life was decreased in those with more than 1 week of prodrug therapy. 94% of AE were mild to moderate; severe reactions were similar to reactions obtained with standard radiation therapy.
Phase I trial of replication-competent adenovirus-mediated suicide gene therapy combined with IMRT for prostate cancer [90] (Freytag 2007).
Ad5-yCD/mutTKSR39rep-ADP Prostate 9 This study used a second generation virus with improved enzyme activity administered via intraprostatic injection. Reported AE include 13% with grade 3 lymphopenia; other AE were grade 1-2. Prostate biopsies at the end of the trial had fewer positives for residual adenocarcinoma than expected (22% rather than >40%) Use of a targeted oncolytic poxvirus, JX-594, in patients with refractory primary or metastatic liver cancer: a phase I trial [91] (Park 2008).

Pexa-Vec/JX-594 (pexastimogene devacirepvec)
Primary or metastatic liver cancer 14 This was a dose-escalation study of intratumoral injection of JX-594. Grade 3 hyperbilirubinemia occurred in patients with the highest dose. All experienced flu-like symptoms, ranging grade I-III and 4 had short-lived grade I-III dose dependent thrombocytopenia. Study reported response in the injected as well as non-injected tumors; 3 PR, 6 with SD and 1 with PD (and 4 could not be evaluated via imaging for different reasons).

Pexa-Vec
Pexa-Vec/JX-594, or pexastimogene devacirepvec, is an oncolytic pox virus that has been studied across a variety of malignancies, including melanoma, colorectal cancer, and liver metastases. This vaccinia virus was altered for oncolytic activity by deactivation of its thymidine kinase, with the addition of granulocyte-macrophage colony-stimulating factor (GM-CSF). The GM-CSF is added to increase oncolysis and induce adaptive immune response to the tumor cells [107]. A majority of the studies focused on liver cancers, either HCC or metastases to liver. In one study, 14 patients received intratumoral injections of the virus. Grade III hyperbilirubinemia was observed with the highest dosed group, and 4 patients experienced transient thrombocytopenia, as high as grade 3. Under RECIST criteria, 3 patients had a partial response (PR) and 6 had stable disease (SD). Interestingly, a response could be seen not only in the tumors that had been injected, but also in the satellite lesions [108]. Pexa-vec has also been studied using IV administration. In a study of 15 patients with colorectal cancer, IV infusion of Pexa-vec resulted in SD in 67% of patients [94]. Pexa-vec showed promise in phase II clinical trials (see Table 3), but the Phase III trial that combined Pexa-vec with sorafenib in patients with metastatic HCC (Clinical Trial: NCT02562755) was terminated in august 2019 due to lack of effectiveness. Study reported that addition of Reolysin to paclitaxel+carboplatin, did not improve PFS, but was well-tolerated. Presence of KRAS mutation also did not affect outcome.
A randomized phase II study of weekly paclitaxel with or without pelareorep in patients with metastatic breast cancer: final analysis of Canadian Cancer Trials Group IND.213 [114] (Bernstein 2018).

Reolysin Metastatic breast cancer 74
There was no statistical difference in RR or PFS but there was increased OS: 17·4 months vs. 10·4 months with paclitaxel alone.
Prospective randomized phase 2 trial of intensity modulated radiation therapy with or without oncolytic adenovirus-mediated cytotoxic gene therapy in intermediate-risk prostate cancer [115] (Freytag 2014).
Intraprostatic distribution and long-term follow-up after AdV-tk immunotherapy as neoadjuvant to surgery in patients with prostate cancer [116] (Rojas-Martinez, 2013).

AdV-tk (aka aglatimagene besadenovec or ProstAtak
Prostate 10 Injected AdV-tk with valcyclovir or ganciclovir in 9 patients scheduled for prostatectomy. 6 of these patients had high risk features. At 10 years did not have signs of recurrence/metastasis. Historic controls with high risk features typically have 15% PSA failure with 34% of those have metastasis [117] (Pound 1999)  AdV-tk (aka aglatimagene besadenovec or ProstAtak Glioma 48 Patients with malignant gliomas were injected AdV-tk during surgery. They then received valcyclovir along with standard therapy of radiation and temozolomide. Overall survival was extended for those who received viral therapy (especially in the group in which total gross resection was achieved at the initial surgery).
A controlled trial of intratumoral ONYX-015, a selectively-replicating adenovirus, in combination with cisplatin and 5-fluorouracil in patients with recurrent head and neck cancer [119] (Khuri 2000).

ONYX-15
Recurrent squamous cell head and neck cancer 37 Patients were given cisplatin/5 FU along with intratumoral injection of ONYX-015. Study reported 27% CR and 36% PR. There were flu-like symptoms and injection site pain most commonly, and rare grade 3-4 mucositis.

ONYX-15
Recurrent head and neck cancer 37 Intratumoral injection of virus led to marked tumor regression (as defined as >50% of lesion) in 21% of patients. Biopsy of surrounding normal tissues were negative for viral infection.

Onyx-015 (Adenovirus dl1520)
Gastrointestinal cancer with liver metastasis 27 Study used hepatic artery infusion of the virus in combination with traditional chemotherapy (5-FU and leucovorin). Study reported biochemical response with virotherapy (increase TNF, IFN-gamma, IL-6 and IL-10) in some tumors resistant to chemotherapy alone. Some grade 3-4 adverse events were seen with hyperbilirubinemia.

Onyx-015 (Adenovirus dl1520)
Colorectal cancer with metastasis to liver 24 Study used hepatic artery infusion of virus in subjects that failed 5-FU/leucovorin. Study reported mixed results: some evidence of tumor necrosis and regression; many patients removed early due to CT-demonstrated enlargement of tumors. However, CT information may reflect inflammation in response to tumor infection with virus prior to regression; PET rather than CT use suggested in the future.
An open label, single-arm, phase II multicenter study of the safety and efficacy of CG0070 oncolytic vector regimen in patients with BCG-unresponsive non-muscle-invasive bladder cancer: Interim results [122] (Packiam 2017).

Ad5/3 delta 24 (CRAd)
Ad5/3 delta 24 is an oncolytic adenovirus being studied for application across a number of malignancies, and prominently with gynecologic cancers. The virus gains entry into cells by binding to the Ad3 receptor, which has increased expression on tumor cells. Delta-24 signifies a 24 base pair deletion in the E1a gene, leading to preferential replication in cells with defective retinoblastoma (Rb) expression, a common occurrence in tumor cells. Twenty-one patients with gynecologic malignancies received intraperitoneal injection of the OV. The treatment was well tolerated with only grade 1-2 adverse events (AE) of fatigue, malaise, and abdominal pain. The study reported that 67% achieved SD, with the remainder progressing after 1 month [72].
Ad5/3 delta 24 has been further manipulated with the addition GM-CSF gene to stimulate immune activity, with strains such as CGTG-102 and CGTG-602. In a study of CGTG-102, 60 patients with advanced solid tumors received either single of multiple dose intratumoral injections of the OV. Stable disease or better was reached in 51% of patients who received serial injections (vs. 41% in single injection). AE were minimal and limited to grade 1-2 flu-like symptoms.

Seneca Valley Virus
Seneca Valley Virus (SVV) is a more recent addition into OV group therapy. Its application focuses on tumors with neuroendocrine features, and has also proven safe in treatment of pediatric tumors [102]. In a study of 22 pediatric patients with neuroblastoma, rhabdomyosarcoma, and rare tumors with NET features, NTX-010, a SVV strain, was studied alone and in combination with cyclophosphamide. Adverse events included leukopenia, neutropenia, and tumor pain, but only 1 grade 3 toxicity (tumor pain) was reported. In another trial of SVV, SVV-001, 30 patients with advanced solid tumors with neuroendocrine features received IV injections of the OV. One patient with SCLC was progression free for 10 months [103]. This is a promising finding that warrants future study with neuroendocrine malignancies. However, a phase 2 study evaluating the efficacy of SVV in extensive stage small cell carcinoma was unfortunately negative. Non-progressive patients post 4 cycles of platinum doublet were randomized 1:1 to receive single IV infusion of SVV vs placebo. This trial was terminated due to futility as no improvement in median PFS was seen at interim analysis [127].
Recently a novel biomarker called tumor endothelial marker 8 (TEM 8) has been confirmed to act as a receptor for SVV, facilitating in viral entry in tumor cells. TEM 8 pre-screening can potentially form an important strategy to select patients for SVV therapeutic clinical trials [128].

Polio Virus
PVSRIPO is an oncolytic polio virus that has demonstrated promising results in the treatment of glioblastoma. It is engineered to target CD155/Necl5, the onco-fetal adhesion protein found in many solid tumors. In a study of 61 patients, the virus was injected intratumorally (Desjardins 2018) [69]. AE were minimal, with grade 4 intracranial hemorrhage occurring at only the highest injection dose. Survival is higher as compared to historical controls. Studies are underway to determine the safety of the therapy in pediatric glioma (NCT03043391) as well as invasive breast cancer (NCT03564782). It is also being studied in combination with immunotherapy for melanoma, including nivolumab (NCT04125719) and Atezolizumab (NCT03973879)

MV-NIS
MV-NIS is a unique oncolytic virus, composed of an Edmonston strain of measles virus which has been altered to express a sodium iodide transporter. Measles virus can target tumor cells through the CD46 antigen. Additionally, infection of tumor cells by the virus can be monitored through administration of Iodine-123 via SPECT imaging.
MV-NIS has been studied in the treatment of myeloma, and has exhibited promising results. In a study of 29 patients 1 CR was reported (Dispenzieri 2017) [100]. There were also some cases of decreased circulating free light chains, but they did increase once the patient cleared the virus.

Phase II Trials
Those viruses that were safe and had early clinical evidence of success progressed to Phase II studies. Phase II investigations seek to demonstrate clinical efficacy. Thirteen viruses across 24 trials were available for review. The review of Phase II data also included 6 Phase I/II trials. Again, because of limitation of space, we will discuss a few OVs used across a variety of cancers. A complete table of Phase I/II and Phase II trials in our review can be found in Tables 3 and 4, respectively.

Reolysin
Reolysin (Pelareorep) is a strain of the reovirus type 3D that has been investigated in a variety of cancers, including ovarian, breast, melanoma, pancreatic, glioma, and multiple myeloma. Ten Phase I trials are in summarized in Table 2. This OV has provided particularly promising evidence, and use in Phase II trials have increased OS in multiple cancers. In one study of 74 patients with metastatic breast cancer, OS increased as compared to paclitaxel monotherapy (17·4 months vs. 10·4 months) [114]. In a smaller study of 14 patients with melanoma, the addition of Reolysin to paclitaxel and carboplatin saw a 21% ORR (3 patients) [111]. Reolysin in combination with gemcitabine was superior to gemcitabine monotherapy in a study of 34 patients with pancreatic cancer [113]; OS and 1-and 2-year survival was increased. Outcomes were similar to those undergoing FOLFIRINOX therapy, but the side effect profile was better tolerated with Reolysin+gemcitabine. The same success has yet to be demonstrated ovarian cancer. In a trial with 108 patients with ovarian, tubal or peritoneal cancer receiving paclitaxel, adding Reolysin did not improve outcomes [109].

ProstAtak
ProstAtak/aglatimagene besadenovec/Adv-tk is an adenovirus which has been modified to express HSV thymidine kinase. After infection of tumor cells with the virus, valcyclovir or ganciclovir is administered that subsequently induces cell death. As the name may suggest, this OV was engineered for application in prostate cancer, but has also been tested in other cancers including a Phase II trial with gliomas. In one study of 48 patients with malignant gliomas, AdV-tk was injected into the tumor cavity following resection, and subsequently received valcyclovir. Overall survival was extended in the treatment group, particularly in those patients that achieved total gross resection at time of surgery [118]. In studies that involve prostate cancer, one example included a study of 10 patients who received intraprostatic injection of Adv-tk as neoadjuvant therapy to surgery. PSA levels were monitored as a marker of tumor suppression. Eight of nine patients continued to have reduced PSA levels 10 years following injection.

ONYX-015
ONYX-015 (previously known as Ad2/5 dl1520) is an oncolytic adenovirus which was engineered to target p53-deficient cells. This adenovirus is deficient in a portion of the E1B gene that produces a 55 kDa protein. This protein typically binds to p53 to prevent apoptosis; in p53-deficient cells, the virus can replicate and lyse cancer cells. Phase II trials involved those for head and neck cancer as well as liver metastases of gastrointestinal cancers. Intratumoral injection of head and neck cancers demonstrated marked (> 50%) tumor regression in 14% of patients [135] It has also been studied in combination with 5-FU/cisplatin, for which there was 27% CR and 36% PR [119]. Hepatic artery infusion of the virus was used in 2 studies, either in combination with 5-FU/Leucovorin, or as a single agent in patients who had previously failed 5-FU/Leucovorin therapy. While there was evidence of biochemical response, increased interleukins and other immune stimulatory molecules, no clinical response was seen [37]. In the ONYX-15 single agent study [121] many patients were removed from the study due to concern about cancer progression seen on CT imaging. It was suggested alternate imaging, such as PET should be utilized in the future as this technology suggested that the increase in tumor size may have been inflammatory response from viral induced infection and inflammation.

Cavatak
Cavatak, otherwise known as CVA21, is a coxsackievirus which targets ICAM-1 to gain entry into tumor cells. Currently there is data available for studies investigating the virus' use in advanced melanoma and non-muscle invasive bladder cancer. In a phase I study of 15 patients with non-muscle-invasive bladder cancer, patients received intravesicular injection of the virus. No serious AE occurred, and 1 CR was reported. A phase II study examined 57 patients with intratumoral injection CVA21 with advanced melanoma. There was a 21% durable response rate (CR+PR).
There are a number of current trials investigating Cavatak, predominantly in combination with immunotherapy. Its safety for use in NSCLC, melanoma, bladder cancer, and prostate cancer is being studied in the STORM trial (NCT02043665), with an additional arm in the bladder cancer and NSCLC groups to be used in combination with pembrolizumab. The CAPRA study (NCT02565992) is also combining Cavatak with pembrolizumab in metastatic melanoma. The MITCI trial (NCT02307149) is combining intratumoral Cavatak with Ipilimumab in metastatic melanoma.

Phase 3 Trials
In our systematic review, available Phase III data were limited to 2 trials of talimogene herparepvec (Table 5). T-VEC was compared to subcutaneous GM-CSF. Higher DRR at 6 months and greater median OS with T-VEC particularly in untreated advanced melanoma. Lesions that were not directly injected showed response.
OS was not improved.

Talimogene Laherparepvec Melanoma 41
Safety profile was consistent with previous trials. As compared to OPTIM trial above, this trial also included ECOG of 2 (OPTIM was 0, 1 only). Efficacy was not assessed, as the primary outcome was to provide expand access to T-VEC until FDA approval. AE, adverse response; DRR, durable response rate; OS, overall survival.

Talimogene Herparepvec
Talimogene herparepvec (T-vec) is currently the only FDA approved OV [2] available in the US. T-vec is an HSV-1 virus, which expresses human GM-CSFT-vec, that can only replicate in cancer cells. This happens because of a dysfunctional PKR pathway in cancer cells. In normal cells, viral infection causes PKR activation, which in turn, leads to an inhibition of protein synthesis within the cell and ultimately decreased viral replication. In contrast, with a dysfunction RKP system in cancer cells, viral replication proceeds unhindered [138]. However, in cancerous cells allow replication by the virus.
T-vec is indicated for unresected advance stage melanoma, administered via intratumoral injection [2]. Phase 3 trials demonstrated mixed results [36,137]. The safety profile was consistent with previous trials, with flu-like symptoms and injection site pain reported; 7% of patients had grade 3 or higher AE including nausea, vomiting, fever and abdominal pain. In the OPTiM Phase III trial [36] tumors that were not injected also demonstrated regression, which indicated possible systemic efficacy. Despite this, OS was not increased. Other studies compared T-vec to simple GM-CSF injection; T-vec produced a higher DRR (durable response rate) and increased OS than did GM-CSF alone [136].
T-vec has also been studied in combination with immunotherapy in initial phase I studies [28,139]. T-vec was injected into lesions with systemic ipilimumab therapy. Combining these two agents is more effective that either treatment alone. In another investigation, combination with pembrolizumab demonstrated increased ORR as compared pembrolizumab alone [28].
There is a Phase III trial underway examining combination therapy of systemic pembrolizumab with intratumoral injection of T-vec (NCT02263508).

Limitations of OVs
Despite the exciting evidence presented in the previously mentioned trials, OV therapy has limitations. One challenge is delivering the virus to the targeted lesions. Intratumoral injection is limited to accessible tumors such as melanomas, or targeted hepatic artery injection for hepatic lesions. While IV administration would be ideal, it can lead to sequestration of virus in the liver, limiting whole body distribution [140] as well as development of neutralizing antibodies.
Neutralizing antibodies are one of the greatest barriers to OV efficacy. Viruses selected for oncolytic virotherapy are those that can infect human cells, which carries both benefits and drawbacks for treatment. Many humans have been previously exposed to or vaccinated against some of the naturally occurring viruses used in OV therapy, and subsequently possess neutralizing antibodies to the virus. Nearly 90% of humans have antibodies against reovirus [141]. The measles virus is in development as an OV, and its success is hindered by the patient's circulating antibodies [142].

Conclusions
The success of immune checkpoint inhibitors has propelled the field of oncology towards exploring other immunotherapeutic mechanisms to abrogate cancer growth. Oncolytic viruses are one such promising avenue with the potential to advance oncology therapeutics. While there has been only one FDA approved anticancer agent to date, a multitude of the oncolytic virus in various stages of clinical development makes for a promising story yet to unfold. The biggest advantage we have in harnessing oncolytic viral therapy lies in our capability to bioengineer these nanomicrobes. With improved understanding of molecular pathophysiology, we have the potential to manipulate oncolytic viruses to match evolving cancer challenges.
Author Contributions: M.C.: literature search, data interpretation, and writing, A.C.: study design, data analysis and interpretation, and writing and editing of manuscript. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.