Oncolytic viruses: A novel therapeutic approach for pancreatic cancer

Abstract

Pancreatic cancer, particularly the exocrine-type pancreatic ductal adenocarcinoma, has a dismal prognosis, with a 5-year survival rate of only 2%–9%, depending on the geographic region. The high aggressiveness of this malignancy is attributed to factors such as late diagnosis, an immunosuppressive and desmoplastic tumor microenvironment, early metastasis, and resistance to conventional therapies. Even novel immunotherapies such as immune checkpoint inhibitors have shown little efficacy in clinical trials. Surgical resection is the only potentially curative treatment; however, few patients are eligible for surgery at diagnosis, and the recurrence rates are quite high. Recently, oncolytic viruses have emerged as promising alternatives for treating this disease. Oncolytic viruses selectively infect, replicate, and lyse cancer cells, triggering immunogenic cell death and initiating antitumor immune responses, offering a potential strategy to overcome the immunosuppressive tumor microenvironment. Additionally, oncolytic viruses can be genetically engineered to enhance tumor selectivity and antitumor immunity, providing increased safety and efficacy. This review aims to characterize oncolytic viruses and pancreatic ductal adenocarcinoma, explore the current state-of-the-art research on oncolytic virotherapy for treating this disease, and provide a summary of ongoing and completed clinical trials. Furthermore, the challenges of oncolytic virotherapy in pancreatic cancer have been highlighted, along with future perspectives for advancing this field.

Graphical abstract

Pancreatic cancer (pancreatic ductal adenocarcinoma [PDAC]) has a grim prognosis with a 5-year survival rate of 2%–9% and limited treatment options. Surgery is often not feasible, and recurrence rates are high. Oncolytic virotherapy shows promise by selectively targeting cancer cells, enhancing the immune response. Genetic modifications improve treatment efficacy, offering hope for PDAC management.

Introduction

Presently, cancer is one of the leading causes of mortality worldwide. Although efforts in early diagnostics and the expanding field of oncologic research have led to increased survival rates, the discovery and development of new, more effective, and less-toxic cancer treatments remain a high priority.¹ When considering pancreatic cancer (PC), this is especially relevant, as it is one of the deadliest cancers, with a very high mortality rate. Pancreatic ductal adenocarcinoma (PDAC), an exocrine form that represents 90% of all PC diagnoses, is the most aggressive and frequent type of PC and the 12th most common malignant disease worldwide.^2,3 The 5-year survival rate of PC is about 10%, with only slight improvements in patient outcomes since the 1980s.² The prognosis of PDAC is particularly poor because the disease is frequently diagnosed in an advanced stage, with most patients already presenting metastasis and because PC is highly refractory to conventional cancer treatments, such as chemotherapy and radiotherapy, due to its immunosuppressive and desmoplastic nature.^3,4 Not even immune checkpoint inhibitors (ICIs), which revolutionized many tumor treatments, showed efficacy in treating PC.⁴ Currently, surgical resection is the only curative option, but most patients present with unresectable disease.³

Recently, oncolytic viruses (OVs) have emerged as a possible therapeutic for PC. Several ongoing clinical trials (CTs) are evaluating the use of OVs either as monotherapy or in combination with other treatments.⁴ OVs can be genetically engineered or can occur naturally, in which case they are often referred to as wild-type OVs.⁵ In the 1950s, naturally occurring OVs were experimentally used in cancer treatment for the first time, and the outcomes were documented in numerous case series. Although some positive results were observed, it became apparent that most wild-type viruses tested lacked safety and efficacy. These viruses were often not selective enough for cancer cells and could cause significant adverse events. Among these, adenoviruses showed the most promising results, indicating a natural tropism for tumor cells without considerable safety issues. Despite this, progress in virotherapy was limited and declined considerably in the 1970s, when CTs involving live pathogens adopted stricter regulations. Consequently, this form of immunotherapy was put mainly aside.⁵

In 1991, Martuza et al. demonstrated that a genetically engineered herpes simplex virus type I (HSV-1) could replicate selectively in cancer cells. Moreover, this mutant virus showed positive results in treating experimental gliomas in nude mice, inhibiting tumor growth and prolonging survival.^6,7 These findings re-ignited interest in OVs and marked the beginning of a new era in their development by demonstrating that leveraging advancements in genomic engineering to modify viral genomes is not only possible but also essential for optimizing the therapeutic potential of OVs.⁷

This review explores the concepts of OVs and PC individually, subsequently exploring the therapeutic potential of OVs as a therapeutic strategy for PC. The objective is to identify and characterize promising genetically engineered and naturally occurring OVs with clinical relevance, alongside their respective CTs. Our focus was directed explicitly toward OVs that have reached clinical evaluation in PDAC, particularly those with published or publicly available data as of July 2025, including both replication-competent OVs and non-replicating vectors. Among the various viral platforms, adenoviruses and HSVs are the most extensively studied and genetically modified for PDAC treatment; consequently, these two viral types are also prioritized.

Included in our analysis are ONYX-015, VCN-01, CAN-2409, LOAd703, and HF10. This selection aims to emphasize OVs that have potential near-term impact in PDAC therapy, as of present. To provide a more comprehensive overview, Table 1 summarizes other ongoing and unpublished CTs for various OVs in PDAC treatment. Additionally, the challenges of implementing oncolytic virotherapy in PC have been highlighted, as well as the strategies to overcome them. By selectively targeting and destroying cancer cells while promoting antitumor immunity, OVs can improve survival rates and provide new treatment avenues for this disease.

Table 1.

A representative sample of CTs evaluating oncolytic virotherapy in PDAC

Virus type	Phase	Title	Virus and administration route	Other therapies	Status and no. of subjects	Study results	Trial no./reference
Adenovirus	1/2a	Phase 1/2a trial evaluating safety of LOAd703, an armed oncolytic adenovirus for pancreatic cancer	LOAd703 Intratumoral	Gemcitabine Nab-paclitaxel Atezolizumab	Active, not recruiting (51)	–	NCT02705196
Adenovirus	1	AdV-tk + valacyclovir therapy in combination with surgery and chemoradiation for pancreatic cancer	CAN-2409 Intratumoral	Valacyclovir chemoradiation Surgery	Completed (27)	Arm A (resectable disease): 8/12 patients resected with R0 margins; mOS was 14 mo; one long-term survivor (>66 mo); increased cytotoxic T cell infiltration. Arm B (locally advanced disease): PR in 3 patients, SD in 7, and PD in 1; mPFS was 5.8 mo and mOS was 12 mo; 1-year survival rate of 50%. No DLTs occurred, and most AEs were mild. Treatment was safe and well tolerated. Benefit seen at all dose levels.	NCT00638612/Aguilar et al.8
Adenovirus	2	Neoadjuvant CAN-2409 + prodrug in combination with chemoradiation or stereotactic body radiation therapy for borderline resectable pancreatic adenocarcinoma	CAN-2409 Intratumoral	Valacyclovir chemoradiation SBRT Surgery	Active, not recruiting (54)	Subgroup of 13 patients with borderline resectable PDAC. Control group (n = 6): mOS was 12.5 mo; 1 long-term survivor; mPPS was 7.2 mo; no evidence of immune activation. Test group (n = 7): mOS was 31.4 mo; 3 long-term survivors; mPPS was 21.2 mo; evidence of robust systemic immune activation; no DLTs or pancreatitis observed. CAN-2409 + valacyclovir was well tolerated and showed a significant survival benefit.	NCT02446093/Candel Therapeutics9
Adenovirus	1	A phase 1, multicenter, open-label, dose-escalation study of intratumoral injections of VCN-01 oncolytic adenovirus with intravenous gemcitabine and Abraxane in advanced pancreatic cancer	VCN-01 Intratumoral	Gemcitabine Nab-paclitaxel	Completed (8)	Viral replication and transgene expression were confirmed. All injected lesions remained stable or decreased in size. Mean time to tumor progression was 8.4 mo. Treatment was well tolerated.	NCT02045589/Bazan-Peregrino et al.10
Adenovirus	1	A phase 1, multicenter, open-label, dose-escalation study of intravenous administration of VCN-01 oncolytic adenovirus with or without gemcitabine and Abraxane in patients with advanced solid tumors	VCN-01 Intravenous	Gemcitabine Nab-paclitaxel	Completed (42)	Viral replication and transgene expression were confirmed. Delayed chemotherapy had fewer AEs but a shorter time to tumor progression than chemotherapy and VCN-01 on the same day. Intravenous VCN-01 was well tolerated.	NCT02045602/Garcia-Carbonero et al.11
Adenovirus	2b	A phase 2b, open-label, randomized study of nab-paclitaxel and gemcitabine ± VCN-01 in patients with metastatic pancreatic cancer	VCN-01 Intravenous	Gemcitabine Nab-paclitaxel	Completed (96)	VCN-01 plus Gem-nabP improved mOS (10.8 vs. 8.6 mo), mPFS (7.0 vs. 4.6 mo), and mDoR (11.2 vs. 5.4 mo) vs. Gem-nabP alone. In patients receiving 2 VCN-01 doses and at least 4 chemotherapy cycles, mOS reached 14.8 vs. 11.6 mo. AEs were mostly mild, transient, and less frequent after the second VCN-01 dose.	NCT05673811/TherivaTM Biologics12
Adenovirus	1	Phase 1 trial of huCART-meso administered in combination with VCN-01 in patients with pancreatic and serous epithelial ovarian cancer	VCN-01 Intravenous	huCART-meso cells	Active, not recruiting (13)	–	NCT05057715
Adenovirus	1	A first-in-human phase 1 trial of binary oncolytic adenovirus in combination with HER2-specific autologous CAR T cells in patients with advanced HER2-positive solid tumors	CAdVEC Intratumoral	HER2-specific CAR T cells	Recruiting (45)	–	NCT03740256
Adenovirus	1	Phase 1 study evaluating MEM-288 oncolytic virus alone and in combination with standard-of-care therapy in advanced solid tumors	MEM-288 Intratumoral	Nivolumab Docetaxel	Recruiting (40)	–	NCT05076760
Adenovirus	2	Oncolytic virus + PD-1 inhibitor to patients with advanced pancreatic cancer	H101 Intratumoral	Camrelizumab	Not yet recruiting (30)	–	NCT06196671
HSV	1b/2	Phase 1b/2 study of OH2 injection, an oncolytic type 2 HSV expressing granulocyte-macrophage colony-stimulating factor, in pancreatic cancer	OH2 Intratumoral	–	Terminated (Low enrollment) (5)	–	NCT04637698
HSV	1a	A clinical safety and efficacy study on oncolytic virus injection (R130) for the treatment of relapsed/refractory advanced solid tumors	R130 Intratumoral/intraperitoneal	–	Recruiting (24)	–	NCT05886075
HSV	1a	A clinical safety and efficacy study on oncolytic virus injection (R130) for the treatment of advanced solid tumors	R130 Intratumoral/intraperitoneal	–	Recruiting (20)	–	NCT05860374
HSV	1	Phase 1 study of intratumoral administration of HF10, oncolytic HSV-1, in patients with advanced unresectable pancreatic cancer	HF10 Intratumoral	Gemcitabine Erlotinib	Completed (12)	Of the 9 patients who completed treatment, 3 had PR, 4 had SD, and 2 had PD; mPFS was 6.3 mo and mOS was 15.5 mo. Two patients exhibited tumor downstaging and underwent resectable surgery, with OS of 22 and 39.6 mo. In both cases, significant T cell infiltration was detected near residual cancer cells. No DLTs occurred, and HF10 showed a favorable safety profile.	UMIN000010150 /Hirooka et al.13
HSV	1	Phase 1 study of combination of TBI-1401(HF10), a replication-competent HSV-1 oncolytic virus, and chemotherapy in patients with stage III or IV unresectable pancreatic cancer	HF10 Intratumoral	Gemcitabine Nab-paclitaxel TS-1	Active, not recruiting (36)	–	NCT03252808
HSV	1b	A phase 1b, dose-escalation study of the safety and preliminary efficacy of STI-1386 oncolytic virus in patients with relapsed or refractory solid tumor conditions	STI-1386 Intratumoral	–	Not yet recruiting (36)	–	NCT05361954
Reovirus	1b	A phase 1b study of pembrolizumab (Keytruda) in combination with Reolysin (pelareorep) and chemotherapy in patients with advanced pancreatic adenocarcinoma	Reolysin Intravenous	Gemcitabine Irinotecan Folinic acid 5-Fluoroucil Pembrolizumab	Completed (11)	Viral replication was confirmed. Three patients achieved disease control: one had a PR lasting 17.4 mo, and two had SD lasting 9 and 4 mo. The treatment was well tolerated, with fever, chills, and fatigue being the most common AEs.	NCT02620423/Mahalingam et al.14
Vaccinia virus	1b	An open-label, non-randomized phase 1b study to investigate the safety and effect of the oncolytic virus GL-ONC1 administered intravenously prior to surgery to patients with solid-organ cancers undergoing surgery for curative intent or palliative resection	GL-ONC1 Intravenous	–	Terminated (Low enrollment) (5)	–	NCT02714374
Vaccinia virus	1	A phase 1 study evaluating the safety, tolerability, biodistribution, and shedding of the virus; pharmacodynamics; immunogenicity; and antitumor activity of GC001 oncolytic vaccinia virus injection in patients with advanced solid tumors	GC001 Intratumoral	–	Recruiting (21)	–	NCT06508307
Vaccinia virus	1	A first-in-human, phase 1, multi-center, open-label, dose-escalation study to evaluate the safety, tolerability, pharmacokinetics, pharmacodynamics, and preliminary evidence of antitumor activity of IDOV-Immune in adult participants with advanced solid tumors	IDOV-Immune Intravenous	–	Not yet recruiting (78)	–	NCT06910657
Vaccinia virus	2	A phase 2 trial to evaluate the efficacy of recombinant human IL-21-expressing oncolytic vaccinia virus injection (hV01) in patients with advanced pancreatic tumors	hV01 Intratumoral	–	Recruiting (12)	–	NCT07006077
Alphavirus	1/2	A single-arm, open-label clinical study to evaluate the efficacy and safety of VRT106 in combination with chemotherapy for resectable pancreatic cancer	VRT106 Intravenous	Chemotherapy	Recruiting (18)	–	NCT06758544
Alphavirus	1/2	A single-arm, open-label clinical study to evaluate the efficacy and safety of VRT106 in combination with chemotherapy for advanced pancreatic cancer	VRT106 Intravenous	Chemotherapy	Not yet recruiting (10)	–	NCT06866977
Unspecified	1	Oncolytic virus + anti-PD-1 and chemotherapy as preoperative therapy for patients with borderline resectable and locally advanced pancreatic cancer	Unspecified	Camrelizumab Gemcitabine Capecitabine	Not yet recruiting (20)	–	NCT06346808

AEs, adverse events; DLTs, dose-limiting toxicities; Gem-nabP, gemcitabine and nab-paclitaxel; huCART-meso, human chimeric antigen receptor-modified T cells; mDoR, median duration of response; mPPS, median post-progression survival; mPFS, median progression-free survival; mOS, median overall survival; mo, months; PR, partial response; PD, progressive disease; PD-1, programmed cell death 1; SD, stable disease; SBRT, stereotactic body radiation therapy.

Oncolytic viruses

The current panorama of oncolytic viruses in cancer therapy

OVs represent a novel form of cancer immunotherapy, where naturally occurring or genetically engineered viruses selectively infect and replicate in cancer cells, ultimately causing their lysis, while leaving normal cells unharmed.⁵ The majority of OVs being studied are genetically engineered to increase tumor selectivity and reduce virulence, minimizing the risk to healthy cells.¹⁵ In the last two decades, there was a significant expansion in oncolytic virotherapy research, supported by preclinical studies and CTs.¹ There are numerous registered CTs involving OVs in the treatment of various types of cancer, including melanoma, glioma, ovarian cancer, lung cancer, PC, head and neck cancer, bladder cancer, breast cancer, liver cancer, and colorectal cancer. Most of these CTs are in the early stages, with only a few reaching phase 3. The most significant research hotspots are adenovirus-based and HSV-based OVs.¹⁶ However, other viruses are also being studied, including reoviruses, parvoviruses, vaccinia virus, Newcastle disease virus, coxsackieviruses, measles virus, Seneca Valley virus, vesicular stomatitis virus, echoviruses, retroviruses, polioviruses, and rhabdoviruses.^1,17,18,19

At present, four OVs have been approved for use in cancer immunotherapy in various countries. The first-ever approved OV is a naturally occurring picornavirus named ECHO-7, also known as Rigvir. It was authorized for use exclusively in Latvia in 2004 for the treatment of melanoma.²⁰ In 2005, the second OV, named H101 or Oncorine, was approved in China. It is a genetically modified serotype 5 adenovirus, used for the treatment of nasopharyngeal carcinoma.²¹ In 2015, the Food and Drug Administration and the European Medicines Agency approved the first OV-based drug in the United States and Europe, respectively. This drug, a genetically modified HSV-1, is known as T-VEC or Imlygic and is used for the treatment of unresectable metastatic melanoma.²² As of now, T-VEC remains the only approved OV in Europe. Most recently, in 2021, Japan approved a fourth OV, named G47Δ or Delytact, a genetically modified HSV-1 for the treatment of malignant glioma.¹

Replication-competent OVs cause cancer cell death by a dual mechanism of action (Figure 1): selective replication within cancer cells followed by their lysis and the induction of systemic antitumor immune responses.²⁴ After infecting tumor cells, OVs replicate until the cell reaches viral overload and undergoes lysis, realizing viral progeny, which infects the surrounding cancer cells, continuing the process.¹⁷ This not only kills the infected cancer cells but also prepares the next phase of antitumor action: the initiation of a systemic immune response against the tumor.²⁵ The induction of systemic innate and systemic tumor-specific adaptive responses is the critical aspect of oncolytic viral action. During oncolytic cell death, cancer cells release tumor-associated antigens (TAAs), which are proteins derived from cancer cell mutated genes.¹⁷ These antigens are recognized as foreign by the immune system, and antigen-presenting cells (APCs) such as dendritic cells (DCs) will process and present them to T cells.²⁴

Figure 1.

The dual mechanism of action employed by replication-competent OVs starts with infection, replication, and oncolysis of cancer cells

This leads to the release of TAAs, which initiates an adaptive antitumor immune response, culminating in the destruction of the infected tumor and also of distant metastasis not directly infected by the virus. This particular virus was genetically engineered to include an immunostimulatory transgene in order to express proteins of interest, such as human GM-CSF, which further aid the antitumor immune response. Therefore, it is referred to as an armed OV. This image was created with illustrations from Servier Medical Art and drew inspiration from Timmer et al.²³

In addition to TAAs, lytic tumor cells release pathogen-associated molecular patterns (PAMPs) and danger-associated molecular patterns (DAMPs), which activate DCs by being recognized by their pattern recognition receptors.²³ TAAs bind to molecules of the major histocompatibility complex (MHC) on the surface of DCs. These DCs then migrate to the lymph nodes, where they present the antigen-MHC complex to T cells. This presentation activates T cells by engaging the T cell receptor (TCR) and the primary costimulatory molecule, CD28. The combined signals from the antigen stimulation (via TCR) and costimulation (via CD28) prime naive T cells to differentiate into effector T cells. These primed cytotoxic T cells can then recognize and target cancer cells, migrating to tumor sites to eliminate them. Consequently, an adaptive immune response is triggered, leading to tumor regression not only at the site of viral infection but also at distant metastases that have not been directly infected by the virus.²⁵

By killing cancer cells, OVs enhance tumor visibility to the immune system, attracting immune cells previously absent from the tumor core, such as cytotoxic T cells, natural killer (NK) cells, mature DCs, and macrophages. This process, known as immunogenic cell death, helps overcome the immune system’s difficulty in targeting tumors, particularly in cases like PC, where the tumor microenvironment (TME) is highly immunosuppressive. OVs can help convert “cold tumors” into “hot tumors” by inducing a pro-inflammatory environment and promoting a more effective antitumor immune response.¹

The mechanism of action described above relies on the ability of OVs to replicate selectively within tumor cells. However, non-replicating viral vectors have also emerged within the broader field of virotherapy. Non-replicating viruses are engineered to enter cancer cells and deliver therapeutic transgenes without undergoing further replication.^26,27 These vectors serve as delivery systems and can be genetically engineered to express immunostimulatory molecules or suicide genes that convert systemically administered prodrugs into cytotoxic metabolites, triggering immunogenic cell death.^26,28 Unlike replicating OVs, non-replicating viral vectors do not directly lyse tumor cells; nevertheless, they can induce antitumor immune responses and are being studied in CTs for several malignancies, including PC.⁸

Strategies to enhance the safety and efficacy of oncolytic viruses

The selectivity of OVs for cancer cells is partly associated with their natural tropism for specific gene and protein expression profiles, as many OVs naturally target cell surface proteins that are overexpressed in cancer cells.¹⁷ A particular example is the HSV entry mediator and nectin-1, which are cell surface receptors that HSV-1 uses to enter the cell. These receptors are significantly overexpressed in melanoma cells, enhancing their susceptibility to HSV-1 infection, compared to normal cells, which express lower levels of these receptors.²⁵ OVs selectively replicate in cancer cells by exploiting their abnormal pathways. In healthy host cells, antiviral pathways are responsible for detecting and eliminating viral infections. However, these pathways are often defective in tumor cells, making them more susceptible to viral replication and destruction compared to healthy cells.^7,17 Nevertheless, this natural selectivity has proven insufficient on its own, and to improve safety, current research focuses on engineering the viral genome with these defective pathways in mind by deleting or inserting specific genes to enhance tumor-selective targeting.¹⁷

A concrete example is H101 (Figure 2), an engineered oncolytic adenovirus (OAd) in which the gene encoding protein E1B55K, a key adenoviral early protein, was deleted. E1B55K binds to and inactivates the tumor suppressor protein p53, preventing apoptosis. Therefore, healthy cells infected with H101 will suffer apoptosis because E1B55K is absent and p53 remains active, making this OAd highly inefficient in normal cells and thus relatively safe.²⁵ However, cancer cells frequently have a defective p53 and do not initiate apoptosis upon infection, allowing viral replication. Initially, the p53 status was thought to be the sole determinant of whether viral infection would result in successful replication and subsequent lytic cell death. However, further research revealed that, although p53 dysfunctionality may contribute to H101’s selectivity, other factors, such as the infectivity of tumor cells and their ability to support the expression of early adenoviral proteins, play a more critical role in determining the replication efficiency of this OAd in cancer cells.²⁹

Figure 2.

The genetically engineered oncolytic adenovirus H101 selectively replicates in cancer cells by exploiting their high susceptibility to adenoviral infection, their ability to support the expression of early adenoviral proteins, and their frequently dysfunctional p53 pathway

These factors allow H101 to replicate efficiently within cancer cells, causing cell lysis and subsequent viral dissemination. In contrast, when H101 infects a normal cell with an intact p53 pathway, the cell undergoes programmed cell death, effectively halting further viral spread. This image was created with illustrations from Servier Medical Art.

Another emerging theme in OV research is the insertion or exchange of human genes within the viral genome to achieve higher immunogenicity and, consequently, greater efficacy in cancer cell destruction.³⁰ Viruses hijack the host cell’s protein translational machinery and use it to produce viral proteins, which can be manipulated to our advantage by engineering the viral genome to produce proteins of interest.⁵ When an OV has an engineered genome containing immunostimulatory transgenes, it is referred to as an armed OV (Figure 1).¹⁷ One such example is T-VEC, which has been genetically engineered to express human granulocyte-macrophage colony-stimulating factor (GM-CSF).¹ GM-CSF is a cytokine whose primary role is to promote the proliferation, differentiation, and migration of APCs, including DCs and macrophages, thereby enhancing their recruitment to the immunosuppressive TME.³¹

T-VEC is created by deleting the ICP34.5 and ICP47 genes from the HSV-1 genome, both of which are non-essential for replication. Deleting ICP34.5 reduces its neurovirulence, latency, and ability to reactivate. Furthermore, this deletion attenuates the virus’ activity, as this gene encodes the neurovirulence factor responsible for replication in neurons.¹⁷ In place of ICP34.5, two copies of the GM-CSF gene are inserted. ICP47 encodes a protein that inhibits antigen presentation; therefore, the deletion of this gene increases tumor destruction.³² Moreover, deleting ICP47 further increases the safety of T-VEC, as it decreases the virus’ ability to evade immune recognition in normal cells.¹⁹ These modifications enhance antitumor immunity and treatment safety, preventing neural damage. Besides GM-CSF, other immunomodulatory factors have been inserted into oncolytic viral vectors, such as interleukin (IL)-2 and type I interferon (IFN-1).³³

Genetic engineering of OVs has allowed for the creation of stronger and more durable antitumor immune responses, enhanced cancer cell selectivity, and reduced viral pathogenicity, which lead to safer and more effective oncolytic viral treatments.¹

Pancreatic cancer

PC has an abysmal prognosis, consistently falling behind all other cancers in terms of positive prognostic outcomes. Depending on the region and country, the 5-year survival rate is anywhere between 2% and 9%. Still, it never exceeds 10%, being much lower than the median 5-year survival rate of all cancer types, which was 69% in 2016.^3,34 There are various types of PCs, divided into two major categories: exocrine and endocrine, developing from exocrine and endocrine cells, respectively. More than 95% of all PCs are of the exocrine type, and less than 5% are endocrine cancers.³⁵ The most prevalent and deadliest PC is PDAC, an exocrine type, which represents about 90% of all diagnosed PCs.³⁴ Given that PDAC is, by far, both the most prevalent and lethal form of PC, this review focuses specifically on PDAC physiopathology and treatment.

The hallmarks of pancreatic cancer

Cancer hallmarks provide a comprehensive framework for understanding the fundamental mechanisms driving tumorigenesis. Below, specific aspects of selected hallmarks of cancer, particularly relevant to PDAC, will be detailed, as well as how OVs could help overcome or take advantage of each one.

Acinar-to-ductal-metaplasia

The pancreas has acinar cells (exocrine glandular region), epithelial cells (pancreatic ducts), and endocrine cells (endocrine glandular region).³⁶ Among these, acinar cells are known for their remarkable capacity for plasticity, which aids pancreatic regeneration and homeostasis. Under certain conditions, such as inflammation and tissue damage, acinar cells can dedifferentiate to a progenitor-like state with a phenotype similar to epithelial cells, resembling the cells of the pancreatic ducts. This process, known as acinar-to-ductal metaplasia (ADM), is the first step toward PDAC development. ADM makes cells more vulnerable to oncogenic mutations, such as the mutational activation of the proto-oncogene KRAS, which in turn promotes the formation of neoplastic precursor lesions.^36,37 Currently, there are three different precursor lesions identified: pancreatic intraepithelial neoplasia (PanINs), intraductal papillary mucinous neoplasms, and mucinous cystic neoplasms, with PanIN being the most common, present in about 90% of PDAC cases.^37,38 These lesions progress as oncogenic mutations continue to occur and ultimately will lead to PDAC.³⁹ OVs are not designed to inhibit the ADM process itself, since it consists of an early, pre-neoplastic change in pancreatic tissue. However, they can be used as therapeutic agents to target and eliminate early-stage cancer cells that have arisen from these lesions. If administered at an early stage, OVs could selectively infect and lyse these transformed cells, potentially halting tumor development and preventing progression to a more advanced, invasive state.

Driver genetic mutations

In genetic terms, PDAC is marked by mutations in oncogenes and tumor suppressor genes crucial to disease progression. The four most common genetic alterations are oncogenic KRAS mutation and loss of the tumor suppressors TP53, CDKN2A, and SMAD4.^36,40 KRAS belongs to the RAS superfamily and encodes a small GTPase that regulates a range of cellular activities, including cell proliferation, differentiation, survival, and migration. Mutant KRAS remains in a permanently active state, leading to continuous activation of effector pathways that drive growth, proliferation, survival, metabolism, motility, and gene transcription. KRAS mutation confers growth, survival, and metastatic advantages to cancer cells, making it the most frequently occurring mutation in PDAC, present in over 90% of cases.⁴⁰ The KRAS mutation is usually the earliest genetic alteration in the development of nearly all cases of PDAC, typically present in precursor lesions like PanINs. However, KRAS-activating mutations alone will not cause PDAC. What allows tumor progression is the combined inactivation of several tumor suppressor genes (TP53, CDKN2A, and SMAD4), which normally counteract oncogenic mutations through mechanisms like cell-cycle arrest, apoptosis, or senescence.^40,41 Therefore, their inactivation is a major catalyst for PDAC development.

CDKN2A encodes p16, a protein that regulates the cell’s growth and division cycle by blocking the transition of the G1 to S phase. This is the first tumor suppressor gene to be inactivated during tumorigenesis since its loss is detected in early PanINs.³⁹ TP53 encodes p53, a transcription factor that, when mutated, loses its DNA-binding ability and, therefore, its function as a transcription factor. This leads to multiple tumor-promoting effects, including genomic instability, altered cell metabolism, enhanced metastatic potential, and impaired cell-cycle control and apoptosis.^37,41 SMAD4 encodes a protein that acts as a key mediator in the signaling pathway of transforming growth factor β (TGF-β). As a transcriptional coactivator, SMAD4 mediates the role of TGF-β in regulating cell growth and apoptosis. The loss of SMAD4 impairs these antiproliferative effects, thus contributing to tumor development and progression.^41,42 The functional loss of both TP53 and SMAD4 usually occurs in PanINs of advanced stages, marking the final stages of tumor initiation.⁴¹

OVs are often genetically engineered to exploit these genetic mutations associated with malignancy. For instance, the inactivation of TP53, resulting in a defective p53 pathway, has been leveraged as a therapeutic vulnerability. Certain OAds, such as ONYX-015 and H101, were specifically designed to selectively replicate in tumor cells with a non-functional p53 pathway, thereby enabling OVs to target cancer cells while sparing normal cells.^25,43 Similarly, the Rb protein pathway is frequently dysfunctional in PDAC cells, and the OAds LOAd707 and VCN-01 were engineered to replicate in cancer cells with a defective Rb protein pathway.² These examples highlight the ability of OVs to exploit common genetic mutations in PDAC for selective and effective targeting.

Epigenetic reprogramming

The transcriptomic profile of PDAC is influenced not only by genetic mutations but also by significant epigenetic alterations. Epigenetic mechanisms such as histone modifications, DNA methylation, and regulatory non-coding RNAs are highly dysregulated in PDAC. Epigenetic reprogramming is closely related to the loss of tumor suppressor genes and the increased expression of oncogenes.³⁶ In 15% of cases, the loss of function of CDKN2A, one of the most relevant tumor suppressor genes in PDAC, happens due to hypermethylation of the promoter region of the gene, resulting in its inactivation.⁴¹ This illustrates how epigenetic reprogramming is an underlying aspect of PDAC development.

OVs can reshape the epigenetic landscape of tumor cells by inducing changes in DNA methylation, histone modifications, chromatin remodeling, and non-coding RNA expression. These modifications can reactivate silenced tumor suppressor genes and silenced cancer-promoting genes and enhance the expression of genes involved in antigen presentation, consequently increasing tumor immunogenicity and promoting antitumor immune responses. For instance, through DNA hypermethylation, OVs can silence genes responsible for inhibiting TAA and neoantigen presentation, thereby enhancing the recognition of these antigens by APCs. The epigenetic modifications triggered by OVs can also render tumor cells more permissive to viral replication and oncolysis, further amplifying their therapeutic effects.⁴⁴

Furthermore, preclinical studies have shown that combining OVs with epigenetic modulators, such as DNA methyltransferase inhibitors and histone deacetylase inhibitors, further enhances treatment efficacy. This strategy helps to reverse aberrant epigenetic silencing, overcoming resistance to virotherapy and boosting antitumor immune responses. Beyond co-administration, OVs can be genetically engineered to deliver epigenetic modulators directly to the TME.^44,45 By strategically targeting the epigenetic dysregulation of tumor cells, OVs can enhance their own replication and lytic activity and counteract the immunosuppressive and pro-tumorigenic effects of epigenetic dysregulation in PDAC, ultimately improving therapeutic outcomes and combating the resistance often seen in conventional PDAC treatments.⁴⁴

Tumor microenvironment

One of the most challenging aspects of PDAC is the TME, which is widely seen as a crucial enabler of disease progression. The TME of PDAC comprises multiple fibroblast populations, endothelial cells, neurons, a diverse and mostly suppressive population of infiltrating immune cells, and a dense extracellular matrix (ECM).⁴⁶ Often, cancer cells are a minority within the tumor, and the non-cancerous components contribute nearly as much as the tumor cells to the disease’s aggressive biology, therapy resistance, and heterogeneity.³⁷

Extensive desmoplastic nature of the TME

Desmoplasia is a fibrotic reaction observed in tumors, characterized by the pathological deposition of ECM components that can occupy the majority of the tumor mass.^36,37 The ECM consists of a dense network of structural proteins, adaptor proteins, proteoglycans, and enzymes. In normal conditions, ECM provides biochemical and structural support to maintain tissue homeostasis. However, in PDAC, there is excessive accumulation of dense ECM primarily originating from fibroblasts.⁴⁶ Cancer-associated fibroblasts (CAFs) are the predominant cells in the TME.³⁷ CAFs are stimulated by cancer cells to overexpress ECM proteins, including collagen (specifically type I, III and IV), hyaluronic acid, and fibronectin.⁴⁶ This dense deposition creates a physical barrier that protects the tumor against systemic chemotherapy delivery. Moreover, hyaluronic acid is a glycosaminoglycan polymer with great capacity for water retention.³⁷ This promotes a higher interstitial fluid pressure (IFP) within the TME, and it is estimated that tumors exhibit an IFP up to 10 times greater than that of healthy pancreas tissue, potentially leading to the collapse of intratumoral vessels and further reducing the permeability of the tumor to systemic chemotherapy.⁴⁶

Desmoplasia presents a significant physical barrier that hinders the access of systemic therapies to the tumor. However, OVs pose a promising approach to overcoming this obstacle.⁴⁷ The OAd VCN-01, for instance, is genetically engineered to express human hyaluronidase, an enzyme that degrades hyaluronic acid. By targeting this key glycosaminoglycan polymer in the ECM, VCN-01 reduces the density of the tumor stroma, decreases IFP, and consequently enhances viral dissemination throughout the tumor mass.^2,3 This mechanism not only facilitates OV spread within the tumor but also enables the infiltration of immune cells and chemotherapeutic agents, overcoming a key hallmark of PDAC resistance.

Immunosuppressive nature of the TME and immune evasion

The largest immune cell populations in the TME of PDAC are myeloid cell types, including myeloid-derived suppressor cells (MDSCs) and tumor-associated macrophages (TAMs). These comprise the two major immunosuppressive cell types in PDAC.⁴⁸ Oncogenic KRAS signaling drives cancer cells to express GM-CSF, which, in this context, promotes the recruitment of MDSCs that will inhibit T cell proliferation and activation.³⁷ TAMs can originate from circulating monocytes or tissue-resident macrophages. In their M2 phenotype, TAMs have tumor-promoting immunosuppressive characteristics, including the secretion of highly immunosuppressive factors such as IL-10, reduced expression of antigen-presenting MHC class II molecules, and the secretion of arginase 1 (therefore depleting local L-arginine levels, which is necessary for T cell proliferation and function). TAMs are recruited to the TME by factors such as colony-stimulating factor 1 and CC-chemokine ligand 2, which are secreted by cancer cells.⁴⁸ Both MDSCs and TAMs suppress T cell function and hinder antitumor immune responses, contributing to immune evasion.³⁶

In terms of adaptive immunity, the TME of PDAC is characterized by a lack of cytotoxic T cells, the presence of pro-tumorigenic T cell populations such as Th2 and T regulatory (Treg) cells, and lower levels of neoantigens compared to immune-sensitive cancers.^37,48,49 Th2 T cells produce IL-4 and IL-13, which can contribute to the suppression of antitumor immune responses. Treg cells modulate DCs to underexpress the costimulatory ligands necessary to activate cytotoxic T cells during antigen presentation.⁴⁸ PDAC is often considered poorly immunogenic; however, recent studies have shown that many patients do possess targetable neoantigens, particularly in the context of fully resected or borderline resectable disease.^8,50 The main barrier is the immunosuppressive TME, which, once established, severely limits effective antigen presentation and T cell activation and, therefore, significantly contributes to the immune evasion of PDAC.⁴⁹ This explains why immunotherapies achieve better outcomes in early disease stages, where the TME is less developed and antitumor immune responses can be more easily activated.

Because of their unique mechanism of action (Figure 1), OVs offer a promising approach to overcoming the deeply immunosuppressive TME of PDAC, which poses a significant obstacle to effective anticancer immunity. OVs selectively infect and lyse tumor cells, triggering immunogenic cell death and the release of TAAs, PAMPs, and DAMPs. These signals promote the recruitment and activation of APCs, such as DCs, and consequently stimulate cytotoxic T cell responses, which will migrate to tumor sites and eliminate cancer cells. This process remodels the TME and converts a “cold tumor” into an immunologically “hot tumor,” characterized by increased immune cell infiltration and activation.^1,23,25 Additionally, many OVs are genetically engineered to express immunostimulatory molecules. For instance, the OAd LOAd703 is armed with genes encoding the immune-activating molecules CD40 ligand and 4-1BB ligand.⁵¹ OVs restore immune surveillance and stimulate robust and sustained antitumor immune responses, which support their integration into combination immunotherapy strategies for PDAC.

Epithelial-to-mesenchymal transition

One of the factors significantly contributing to the poor prognosis of PDAC is the tendency for early metastasis. Dissemination can occur even before the primary tumor reaches a substantial size, and PDAC exhibits a remarkable propensity to metastasize to distant sites, with the liver, peritoneum, and lungs being the most common.^49,52 Genome sequencing studies reveal that cells in metastatic lesions and cells of the primary tumor share similar genome and driver gene mutations, which implies that transcriptional or post-transcriptional alterations are crucial for enabling cancer cells to invade and metastasize.⁵² A key process in metastasis is epithelial-to-mesenchymal transition (EMT) in which epithelial cells acquire a mesenchymal cell phenotype due to a complex system of transcription factors that induce mesenchymal gene expression while suppressing epithelial gene expression.^36,52 Through EMT, cancer cells lose epithelial characteristics such as polarity, tight cell-cell adhesion, and orderly tissue structure and adopt mesenchymal traits such as increased mobility, allowing the cells to migrate from the primary tumor. After extravasation into distant tissues, PDAC cells need to re-adopt an epithelial phenotype to colonize; therefore, they experience a reverse form of EMT.³⁶ The TME significantly influences invasion and metastasis, since CAFs produce soluble factors capable of inducing EMT. Examples include hepatocyte growth factor (HGF) and insulin-like growth factor 1, which activate signaling pathways that alter cell adhesion, morphology, and motility.⁵²

While OVs do not inhibit the EMT process itself, they could potentially help counteract it by targeting both tumor cells undergoing EMT and the TME that promotes it. Interestingly, in a preclinical study conducted using in vitro cell lines and in vivo mouse models, cancer cells undergoing EMT exhibited increased susceptibility to oncolytic herpesviruses, particularly HSV-1. During EMT, the surface expression of a key HSV-1 entry receptor, nectin-1, increases due to the loss of adherens junctions, where the receptor is usually sequestered. As a result, nectin-1 becomes more accessible, facilitating viral entry. A human follicular thyroid cancer cell line and a human PDAC cell line were used, and, in both models, EMT-induced cells exhibited higher HSV-1 binding, entry, gene expression, viral replication, and oncolysis compared to control cells. In the in vivo models, tumors derived from EMT-induced cells showed significantly enhanced regression in response to oncolytic HSV-1 compared to control tumors.⁵³ By exploiting EMT-associated upregulation and exposure of nectin-1, OVs could potentially transform a significant driver of metastasis into a therapeutic vulnerability.

Moreover, the ability of OVs to remodel the TME provides an indirect way to counter the factors responsible for inducing EMT, particularly through modulation of CAFs and stromal signaling. For instance, the OAd LOAd713 was engineered to express the IL-6 receptor and the immune-activating molecule CD40 ligand, to inhibit IL-6 signaling (a cytokine with a prominent role in cancer pathogenesis) and stimulate CD40. Preclinical testing in human cancer cell lines, including PDAC, showed that LOAd713 effectively induced oncolysis in PC cells.⁵⁴ In contrast, pancreatic stellate cells, one of the primary sources of CAFs in PDAC, were not lysed by the virus due to their intact Rb protein pathway, which prevents viral replication.^54,55 However, LOAd713 infection in these cells led to a significant decrease in the expression of EMT-promoting factors, such as HGF and TGF-β, demonstrating that OVs can disrupt key mediators of EMT within the TME.⁵⁴

Pancreatic cancer and oncolytic viruses

Pancreatic cancer and immunotherapy

The main principle of cancer immunotherapy is to stimulate the patient’s immune system, particularly T cells, to identify and destroy cancer cells.⁵⁶ Immunotherapy has shown remarkable efficacy in many cancers, even surpassing chemotherapy as the preferred first-line treatment in some instances.⁵⁷ However, in the particular case of PC, CTs of single-agent immunotherapies have had limited results. For example, ICIs, which have shown great success in the treatment of other cancers, seem to have minimal activity in PC.⁵⁸ Currently, immunotherapy is used in routine clinical practice only for a small subset of PCs that have specific genetic and molecular biomarkers that increase immune sensitivity. In these cases, patients may be treated with pembrolizumab, an anti-programmed cell death 1 monoclonal antibody.⁵⁹ It has become clear that achieving effective immunotherapy likely depends on exploring the synergistic effects of combination treatments.^23,60 Future strategies may commonly combine diverse and novel immunotherapies with chemotherapy. This approach would target a broad spectrum of immune evasion mechanisms, thereby increasing the tumor's sensitivity to immune attack. There are a wide array of immunotherapies currently in CTs for the treatment of PC, including immunomodulators (such as ICIs, immune stimulatory agonists and cytokines), adoptive cell therapies (involving T cells and NK cells), cancer vaccines, and OVs.²³

Oncolytic adenoviruses under clinical investigation in PDAC

A virus should have some desirable characteristics to be used as an OV, namely, it should infect tumor cells selectively; undergo rapid replication, to produce the desired effect before being neutralized by the host immune system; have a low prevalence of specific antibodies among the general population; and have a favorable safety profile, with limited adverse effects, risk of transmission, and mutation.^19,61

Adenoviruses are members of the Adenoviridae family, characterized by a double-stranded DNA genome of approximately 35 kb, enclosed in a non-enveloped icosahedral capsid.¹⁸ These viruses have been extensively studied for oncolytic virotherapy in PDAC and have been the subject of numerous preclinical trials and CTs.⁶² There is a wealth of knowledge regarding them, and they offer several clinically beneficial attributes for use as OVs, including high efficiency for in vivo gene delivery; a lytic life cycle that involves infection, replication, and destruction of host cancer cells (oncolysis); and a reduced risk of genotoxicity since the viral genome does not integrate into the host cell’s genome.⁶³ While numerous adenovirus serotypes have been identified, the majority of OAds are derived from serotype 5.^51,64 Several CTs are underway to assess the efficacy and safety of OAds for PDAC treatment.¹⁸ Most of the OAds being studied are genetically engineered to enhance their efficacy and selectivity for cancer cells, thus improving safety. Below, a selection of relevant genetically engineered OAds and their corresponding CTs is depicted.

ONYX-015

ONYX-015, similar to H101 discussed above, is a genetically modified adenovirus with a deletion of the E1B55K gene, which encodes an anti-apoptotic protein responsible for inactivating p53. ONYX-015 showed promise in preclinical studies and was the first OV to enter CTs for the treatment of PDAC. However, it failed in CTs due to a fundamental design flaw: the E1B55K deletion severely compromised the virus’ ability to replicate, even in tumor cells.⁴³ Consequently, ONYX-015 was unable to disseminate within the tumor mass or induce an objective clinical response. CT results published in 2001 confirmed this failure, showing no evidence of effective intratumoral viral replication or tumor reduction.^3,65 Although intratumoral injection of ONYX-015 exhibited no efficacy, its safety was well established, which incentivized further CTs involving genetically modified OVs.⁴³ ONYX-015 is no longer under investigation for PDAC, and newer OAds with improved genetic engineering are now being explored.⁵¹

LOAd703

LOAd703 is a mutant OAd with several genomic modifications. It has the deletion of the E1ACR2 region, which encodes a protein that inhibits the Rb protein to promote viral replication. This deletion allows selective replication in cancer cells with a dysfunctional Rb protein pathway. To further promote viral replication in tumor cells, E2F-binding sites were inserted upstream of the E1A gene, which is a critical gene to initiate viral replication. E2F is a transcription factor often overactive in cancer cells due to the dysfunctional Rb protein; therefore, this modification ensures that the virus only replicates in cells with overactive E2F, enhancing treatment safety. To increase the virus’ capacity to trigger an immune response (increase immunogenicity), the E3 gp19K and E3 6.7K genes were deleted, as these genes help the virus evade immune detection. However, the most significant and innovative genomic alteration of LOAd703 is the insertion of genes encoding the immune-activating molecules CD40 ligand and 4-1BB ligand to boost immune cell activation and further stimulate an antitumor immune response.^2,43,51 An ongoing phase 1/2a CT (NCT02705196) is evaluating the safety and efficacy of intratumoral LOAd703 treatment in PDAC patients, when administered with gemcitabine and nab-paclitaxel (Gem-nabP) and with or without atezolizumab, an anti-programmed death-ligand 1 monoclonal antibody (Table 1).⁶⁵

VCN-01

Similar to LOAd703, VCN-01 also has the deletion of the E1ACR2 region, allowing selective replication in cells with a dysfunctional Rb protein pathway, as well as insertion of E2F-binding sites. Additionally, VCN-01 features a mutation in the viral capsid to reduce liver tropism and increase its half-life in the bloodstream after intravenous administration. Furthermore, a gene encoding human hyaluronidase was inserted into the viral genome to degrade the dense ECM characteristic of PDAC, facilitating viral dissemination and tumor infiltration of immune cells, as well as chemotherapy agents.^2,3

Results from two completed phase 1 CTs investigating VCN-01 for PDAC treatment were published in 2021 and 2022.^3,10,11 In the 2021 CT (NCT02045589), VCN-01 was administered via intratumoral injection at doses of 10¹⁰ and 10¹¹ viral particles per patient, either with gemcitabine or Gem-nabP. Eight participants with advanced PDAC were divided into two groups: one received the lower dose of VCN-01 (10¹⁰ viral particles) with gemcitabine and the other received the higher dose (10¹¹ viral particles) with Gem-nabP. The most frequent adverse events were asthenia (six patients), pyrexia (four patients), and elevated liver transaminases (three patients). The pharmacokinetic profile of VCN-01 showed two peaks: one corresponding to the initial distribution phase and the other indicating successful intratumoral viral replication. Furthermore, after intratumoral injection, the serum concentration of recombinant human hyaluronidase increased, proving viral replication and expression of the transgene. Regarding clinical outcomes, all injected tumor lesions remained stable or regressed, with 12.5% of patients showing a minor response to treatment. The mean time to tumor progression was 8.4 months.^3,10

The 2022 CT (NCT02045602) enrolled 42 participants and was the first to administer an OAd intravenously for PDAC treatment. The doses administered were 3.3 × 10¹² and 3.3 × 10¹³ viral particles per patient. The participants received Gem-nabP concomitantly, and this study evaluated how the timing of chemotherapy and oncolytic virotherapy administration affected treatment efficacy and adverse events. One group received VCN-01 and chemotherapy on the same day, while the other group received chemotherapy 7 days after VCN-01 administration. It was observed that the group receiving delayed chemotherapy suffered fewer adverse events. However, this group had a median time to tumor progression of 6.7 months, compared to 9.9 months for the group given chemotherapy and VCN-01 on the same day. Compared to the 2021 CT, the response rates of both groups were lower, suggesting that systemic administration may be less effective than intratumoral injection. However, this CT reported fewer adverse events per patient, which is noteworthy given that systemic intravenous administration is typically associated with a higher potential for adverse effects. This finding indicates that systemic administration of OAds is feasible and well-tolerated, with adverse events comparable to those observed with intratumoral injection.^3,10,11 Overall, in these two CTs, VCN-01 demonstrated an encouraging safety profile, with the most common adverse events being asthenia, nausea, and pyrexia, often associated with chemotherapy. Additionally, most reported events were of mild severity.^3,10,11

Recently, the top-line data of a phase 2b CT (NCT05673811) completed in March 2025 was reported.¹² This study aimed to further evaluate VCN-01 in combination with Gem-nabP as first-line therapy for metastatic PDAC in a larger cohort. The trial enrolled 96 patients, randomized 1:1 to receive either Gem-nabP alone or Gem-nabP preceded by intravenous VCN-01 administration. In the VCN-01 arm, the OV was administered 7 days before the first and fourth cycles of chemotherapy. The median overall survival (OS) was 10.8 months in the group receiving VCN-01 plus chemotherapy and 8.6 months in the group receiving chemotherapy only. The median progression-free survival (PFS) was 7 months vs. 4.6 months, respectively. The median duration of response (DoR) was 11.2 months in the VCN-01 group compared to 5.4 months in the control group. Subgroup analysis revealed better therapeutic outcomes in patients who received two doses of VCN-01 and at least four chemotherapy cycles, with median OS reaching 14.8 months compared with 11.6 months in patients treated with chemotherapy only. This suggests that a second dose of VCN-01, administered approximately 3 months after the first, may provide additional therapeutic benefit in this subgroup.¹²

In this phase 2b CT, VCN-01 was well tolerated, with the most frequent adverse events being pyrexia, flu-like symptoms, nausea, vomiting, and elevated liver transaminases. These events were transient, reversible, and occurred with less frequency and lower severity following the second VCN-01 dose. Overall, the safety profile was consistent with previous CTs and considered acceptable for the target population.¹² A phase 3 confirmatory CT is being planned, positioning VCN-01 as one of the leading OVs with potential for approval as a therapeutic option for PDAC.

CAN-2409

Unlike the replicating OAds presented above, CAN-2409 is a non-replicating adenoviral vector, genetically engineered to encode the HSV thymidine kinase (HSV-tk) gene. CAN-2409 is delivered intratumorally as a localized oncolytic immunotherapy. Once administered, HSV-tk expression enables localized prodrug activation within tumor cells. Orally administered prodrugs such as valacyclovir or ganciclovir are transformed into phosphorylated toxic metabolites, which are incorporated into the genome of transduced and neighboring tumor cells, inducing immunogenic cell death.^8,28 In addition to its enzymatic role in prodrug conversion, HSV-tk also acts as a superantigen-like molecule, contributing to a potent immune activation.⁸

The first CT evaluating CAN-2409 for the treatment of PDAC was published in 2015. This phase 1, open-label, dose-escalation study (NCT00638612) primarily assessed the safety and feasibility of intratumoral CAN-2409 as an adjunct to standard-of-care treatment and secondarily evaluated its potential clinical benefit. A total of 27 patients were enrolled and divided into two cohorts. Arm A comprised 14 patients with potentially resectable disease who received CAN-2409 as neoadjuvant therapy before surgery, while arm B comprised 13 patients with locally advanced PC who were treated with CAN-2409 in combination with chemoradiotherapy. Patients received two doses of intratumoral CAN-2409 via either computed tomography or endoscopic ultrasound (EUS) guidance, along with oral valacyclovir for 14 days, beginning 1–3 days following CAN-2409 injection. Four dose levels (3 × 10¹⁰, 1 × 10¹¹, 3 × 10¹¹, and 1 × 10¹² viral particles per patient) were tested.⁸

In arm A, the second CAN-2409 dose was administered at the time of surgery, 2–3 weeks after the first, either into the resection bed or directly into the tumor if resection was not possible. In arm B, patients initiated 5-fluorouracil plus radiotherapy the week following the first injection, and the second CAN-2409 dose was delivered 2–4 weeks after the first, during chemoradiation. Of the 27 patients enrolled, 24 completed the study, 12 in each arm, with three patients treated at each dose level per arm. Three patients withdrew early: one due to unrelated death, one due to metastatic disease discovered during surgery, and one due to valacyclovir intolerance.⁸

All 12 patients in arm A underwent surgical exploration, and 8 had successful tumor resections with R0 margins. Four patients were deemed unresectable, one due to locally advanced disease and three due to unexpected metastases. Among the resected patients, one remained disease free over 66 months after treatment. The remaining patients had OS ranging from 9 to 30.4 months, with a median of 14 months. Immune cell infiltration in tumor tissue was evaluated in seven patients by comparing pre-treatment tumor biopsies with the resected tumors, and all showed increased cytotoxic T cell infiltration, while helper T cell levels remained mostly unchanged. This suggests that CAN-2409 administration was able to activate an effective antitumor immune response.⁸

In arm B, three patients experienced a partial response, seven achieved stable disease, and one had progressive disease based on metastatic spread identified 2.1 months post-treatment. One patient declined further treatment after chemoradiation but maintained quality of life until death at 12 months. Among those with partial response, two had a DoR of 7 and 7.9 months. The third patient was considered for resection based on tumor shrinkage, but surgery was not performed due to metastasis found during laparotomy. Median PFS and OS in arm B were 5.8 and 12 months, respectively, with a 1-year survival rate of 50%. These results compare favorably with historical outcomes reported for 5-fluorouracil plus radiotherapy alone, which typically show a median OS of approximately 9 months and a 1-year survival rate of 28%. The three patients with metastases discovered during surgery were subsequently treated with gemcitabine-based chemotherapy and survived between 9 and 12 months, exceeding the expected median survival for this subgroup, which is approximately 6 months.⁸

Regarding safety, CAN-2409 was well tolerated across all dose levels and no dose-limiting toxicities occurred. The most frequently reported adverse events included fatigue, anorexia, nausea, vomiting, and abdominal pain, which are typical in PC patients. Most clinical abnormalities were grade 1 or 2 and deemed unrelated or unlikely to be related to the treatment. Four patients experienced transient grade 3 clinical abnormalities that were considered possibly related to CAN-2409. Overall, CAN-2409 demonstrated a favorable safety profile and encouraging early signs of efficacy when combined with either chemoradiotherapy or surgical resection. The addition of this viral therapy did not appear to increase toxicity and produced promising clinical outcomes when compared with historical controls, where PDAC patients received standard of care alone. Interestingly, this study suggests that efficacy was not dose dependent since clinical activity and immune activation were observed at all dose levels.⁸ This trial demonstrated that intratumoral injection of CAN-2409 is feasible and safe, supporting it as a promising approach to improve the efficacy of PDAC treatment without increasing toxicity.

A phase 2 CT (NCT02446093) is currently ongoing to assess the safety, preliminary efficacy, and immunological effects of intratumoral CAN-2409 in combination with a prodrug (valacyclovir or acyclovir) in patients with borderline resectable PDAC undergoing neoadjuvant chemoradiation or stereotactic body radiation therapy before surgical resection. This trial enrolled a total of 54 patients with borderline resectable disease and remains active, although it is not currently enrolling new participants. In February 2025, a press release reported final survival data from a subgroup of 13 patients enrolled in this trial. These patients were randomized into a test group (n = 7) that received CAN-2409 plus valacyclovir in addition to standard-of-care chemoradiation followed by surgical resection and a control group (n = 6) that received standard chemoradiation alone prior to surgical resection. Final survival outcomes were reported after an additional 9 months of follow-up beyond the initial analysis. Patients treated with CAN-2409 showed a significant survival benefit. The estimated median OS was 31.4 months compared to 12.5 months in the control group. At the data cutoff, three of the seven patients in the experimental arm were still alive, with post-enrolment survival durations of 66, 63.6, and 35.8 months and survival from diagnosis of 73.5, 68.8, and 41.3 months, respectively. Among these, one patient had stage IV metastatic disease identified at surgery and another had residual tumor at the resection margin.⁹

In contrast, only one patient in the control group remained alive at data cutoff (61.2 months post-enrolment and 65.5 months from diagnosis). At the time of surgical resection, this patient had no histological evidence of residual invasive PDAC, which is associated with better outcomes. Although the long-term survivors in the experimental group had disease recurrence, they responded to salvage chemotherapy and exhibited prolonged post-progression survival. Median survival following progression was 21.2 months in the CAN-2409 group, compared to 7.2 months in the control group. Notably, resected tumor tissue from patients receiving the experimental treatment showed dense infiltration of CD8+ granzyme B+ cytotoxic T cells, DCs, and B cells. Additionally, increased levels of granzyme B, granzyme H, and IFN-γ were detected in peripheral blood, indicating robust systemic immune activation. These findings were not observed in the control group, which supports CAN-2409’s capacity to activate a potent anti-tumor immune response.⁹

In this subgroup of 13 patients, CAN-2409 was well tolerated, with no dose-limiting toxicities or cases of pancreatitis reported. The safety profile observed is consistent with prior CTs of CAN-2409 across multiple indications, including the previous trial in PDAC, reinforcing the safety of this investigational therapy when combined with standard-of-care treatment for borderline resectable PDAC.⁹ Given the durable survival benefit, favorable safety profile, and evidence of immune activation, CAN-2409 is one of the most promising viral therapies currently in development for the treatment of PDAC. A larger late-stage randomized controlled CT is being planned to further investigate its clinical benefit in this disease.

Oncolytic herpesviruses under clinical investigation in PDAC

HSVs are members of the Herpesviridae family, characterized by a double-stranded DNA genome of about 154 kb, enclosed in an enveloped icosahedral capsid.²⁵ The approval of T-VEC for the treatment of unresectable metastatic melanoma increased interest in the investigation of HSVs, particularly HSV-1, as platforms for oncolytic virotherapy. HSV-1 offers several advantageous characteristics, including its broad infection range, making it able to infect nearly all types of human cells; a large genome encoding over 74 distinct genes, which allows for significant capacity to incorporate transgenes; a greater infectivity compared to OAds; and availability of effective antiviral molecules, such as acyclovir and ganciclovir, able to control adverse events.^17,51,66

HF10

HF10, unlike T-VEC, is a spontaneously mutated HSV-1. Meaning that its genome was not engineered but rather selected after mutations occurred naturally and made it suitable for oncolytic virotherapy.⁶² HF10 has multiple natural gene deletions, with LAT and UL56 being particularly significant, as their absence reduces the virus’ neurovirulence, reducing the risk of damage to the nervous system. Although not fully characterized, the virus’ selectivity for cancer cells is associated with abnormalities in the IFN pathway, which facilitate viral replication.⁴³

A phase 1 CT (UMIN000010150) evaluated the safety and efficacy of HF10 administered intratumorally under EUS guidance in combination with gemcitabine and erlotinib in patients with unresectable locally advanced PDAC without distant metastasis. This dose-escalation, single-arm, open-label trial began with a cycle of erlotinib and gemcitabine therapy, and the first HF10 injection was performed on day 1 of the second chemotherapy cycle. HF10 administration continued every 2 weeks for up to four doses, provided no dose-limiting toxicities occurred. Initially, 12 patients were enrolled; however, two withdrew after the first chemotherapy cycle due to lymph node metastases and interstitial pneumonia. Ten patients initiated HF10 treatment, and nine completed all four injections. HF10 was administered at doses of 1 × 10⁶, 3 × 10⁶, or 1 × 10⁷ plaque-forming units (PFU) per injection, corresponding to three patient cohorts. Five patients developed myelosuppression, which was caused by chemotherapy. One patient discontinued treatment after developing severe hepatic dysfunction following the third injection, and another experienced perforative peritonitis due to pre-existing duodenal stenosis. These adverse events were considered to be unrelated to HF10 administration, and no dose-limiting toxicities were observed.¹³

Among the nine participants who completed the treatment, three exhibited a partial response, four had stable disease, and two presented progressive disease during surveillance time.^13,65 Regarding target lesion analysis, three patients had a partial response and six exhibited stable disease. Median PFS was 6.3 months, and median OS was 15.5 months. Two of the three patients who showed partial response experienced tumor downstaging and underwent resectable surgery. The first patient, who received 3 × 10⁶ PFU, underwent surgery 5 months after initiating treatment. Histology of the resected tumor showed 99% had been eliminated and substituted with fibrotic tissue, with significant cytotoxic T cell infiltration near the residual cancer cells. This patient developed peritoneal dissemination 6 months after surgery and had an OS of 22 months. The second patient received 1 × 10⁷ PFU, and surgery was performed 7 months after enrollment, with histology confirming 90% tumor elimination and cytotoxic T cell infiltration. Six months after surgery, disease recurrence was observed in the mesenteric lymph nodes, but the patient achieved a prolonged survival of 39.6 months. In both surgical cases, the detection of significant T cell infiltration close to the residual cancer cells supports that HF10 exerted a therapeutic effect not only by direct oncolysis but also through the stimulation of antitumor immune responses.¹³

In this phase 1 trial, HF10 demonstrated a favorable safety profile, with no complications related to the intratumoral administration of the OV. Although therapeutic outcomes remain preliminary, HF10 showed signs of efficacy, and combination therapies involving this OV warrant further investigation in a larger cohort of PDAC patients. Presently, another phase 1 CT (NCT03252808) is ongoing to assess the recommended dose of intratumoral HF10 administered concomitantly with chemotherapy for the treatment of unresectable PDAC. The OV is being injected every 2 weeks for up to 1 year, alongside either Gem-nabP or TS-1 (a combination of tegafur, gimeracil, and oteracil) in a larger cohort of 36 patients (Table 1).

Although adenoviruses and herpesviruses are the most extensively studied in CTs, other virus types, such as reoviruses, vaccinia viruses, and alphaviruses, are also being investigated for oncolytic virotherapy in the context of PDAC. Table 1 presents a representative sample of these CTs, including additional OAds and oncolytic HSVs not previously mentioned. It contains all CTs registered on ClinicalTrials.gov as of July 14, 2025, identified using the search terms “pancreatic cancer” and “oncolytic virus.”

Challenges of oncolytic virotherapy in pancreatic cancer

Limitations of preclinical models

In the preclinical development of OVs, in vitro models are typically the starting point for assessing viral infectivity, replication, and cytopathic effects. To date, the vast majority of these studies have been conducted using two-dimensional (2D) PDAC cell line cultures, in which cells grow as a monolayer attached to a flat surface. The widespread use of these models is mainly due to their accessibility, ease of manipulation, cost-effectiveness, and suitability for high-throughput screening. However, 2D cultures fail to replicate key features of pancreatic tumors, such as spatial organization, ECM composition, cellular heterogeneity, and the immunosuppressive TME.⁶⁷ While these studies remain valuable for initial screening and mechanistic insights, their inability to recapitulate the complexity of in vivo tumors significantly limits their predictive value for clinical efficacy, contributing to poor translational outcomes.⁶⁸

Recent research has highlighted the advantages of three-dimensional (3D) PDAC models, particularly spheroids and patient-derived organoids (PDOs), which better reflect the biology of human tumors. Research comparing 2D monolayer and 3D PDAC cell culture models has shown that OVs exhibit significant differences in their ability to infect, replicate, and induce cell death depending on the model, emphasizing how this choice strongly impacts study results.⁶⁸ Tumor spheroids are generated from conventional PDAC cell lines cultured under non-adherent conditions that promote spontaneous cell aggregation. These structures more accurately mimic tumor architecture, cell-cell interactions, hypoxic gradients, and limited nutrient diffusion, offering greater physiological relevance than 2D cultures.^67,68

Alternatively, PDOs are derived directly from patient tumor tissue obtained through surgical resection or image-guided biopsy. The tumor cells are embedded in a hydrogel matrix that supports 3D growth and spatial organization.⁶⁷ PDOs have been shown to preserve the histological architecture, genomic and transcriptomic profiles, and cellular morphology of the original tumor, even after long-term in vitro expansion. The ability of PDOs to retain these features makes them a valuable platform for developing personalized and precise oncolytic virotherapy strategies. Their patient-specific nature allows for the assessment of individual tumor responses to OVs under controlled conditions. PDOs are particularly well-suited for identifying genetic markers associated with sensitivity to virotherapy. These markers could then be used to screen patients and guide the selection of the most appropriate OV-based treatment. Achieving this, however, requires the analysis of larger and more diverse PDO panels to capture interpatient heterogeneity and enable robust biomarker discovery.⁶⁸

Despite their advantages, 3D models have limitations. Their generation and maintenance are costly and labor intensive. Standard PDO models lack critical components of the TME, including vasculature, CAFs, and immune cells, which limits their utility for studying key mechanisms of oncolytic virotherapy, such as immunogenic cell death. Moreover, the inherent heterogeneity of PDOs and the complexity of their manipulation constrain scalability for high-throughput screening.⁶⁷ Nevertheless, the increased adoption of 3D models represents progress toward more physiologically relevant and predictive in vitro platforms.

In vivo models remain indispensable for evaluating OV safety and efficacy. Among these, xenograft models, including cell line-derived xenografts (CDXs) and patient-derived tumor xenografts (PDTXs), are commonly used. In CDX models, established human PDAC cell lines are transplanted into immunocompromised mice, either subcutaneously or orthotopically. Subcutaneous implantation is inexpensive, enables rapid screening of OV efficacy and toxicity, and facilitates tumor size monitoring. In contrast, orthotopic implantation into the pancreas better depicts the TME but is more technically demanding, expensive, and requires animal sacrifice for response evaluation. Despite their practicality, CDX models fail to capture the cellular and stromal heterogeneity of PDAC, often producing uniform tumor masses with limited stromal infiltration. As a result, their predictive value is limited, with poor correlation between preclinical outcomes and clinical efficacy.⁶⁷

PDTX models involve the engraftment of tumor fragments obtained from patient biopsy or surgical resection into immunocompromised mice. These models more accurately preserve the architecture and metastatic potential of the tumor. Nonetheless, PDTXs are resource intensive, with long engraftment times and high maintenance costs. Furthermore, the progressive replacement of human stroma by murine stromal components after serial passages introduces a species mismatch that may affect study results.⁶⁷ However, the most significant limitation shared by both CDXs and PDTXs is that they require the use of immunodeficient hosts. This precludes the study of immune responses and renders these models suboptimal for evaluating oncolytic virotherapy, which heavily depends on interactions with the host immune system.^67,69

Another valuable and widely used in vivo model for studying novel PDAC treatments is the genetically engineered mouse model, which is developed by introducing specific mutations in oncogenes and tumor suppressor genes associated with the disease into the mouse genome. The most established genetically engineered mouse model in PDAC research is the KPC model, which carries mutations in KRAS and TP53, leading to the spontaneous development of pancreatic tumors that closely mimic the human TME, including its dense desmoplasia and poor vasculature.⁶⁷ Importantly, the animals in the KPC model retain a functional adaptive immune system, making it particularly relevant for evaluating immunotherapeutic strategies.^65,67 However, many clinically relevant OVs, including OAds, exhibit strict species specificity and do not replicate in murine cells. As a result, the KPC model is unsuitable for studying key aspects of replication-competent OVs, such as viral replication and spread within the tumor.^63,65

Alternative animal models, such as the Syrian hamster, have been explored. The Syrian hamster model for PDAC is a syngeneic model, meaning it is established by transplanting tumor tissue or cells between genetically identical animals of the same strain. This enables the use of fully immunocompetent hosts, as the immune system does not reject the transplanted tumor.⁶⁷ Consequently, these models are suitable for studying how host immune responses influence the efficacy of OVs. Among syngeneic models, the Syrian hamster is particularly relevant for studying OAds because, unlike murine models, it is permissive to adenovirus replication. Before the development of this model, testing replication-competent OAds was restricted mainly to human tumor xenografts in immunodeficient mice.⁶⁹ However, the Syrian hamster is only partially permissive to human adenoviruses. For instance, while human adenovirus serotype 5 is capable of replicating in various hamster cell lines, the viral burst size is about seven times lower than that observed in the highly permissive A549 human lung cancer cell line.⁷⁰ Despite this limitation, the Syrian hamster remains a valuable model, as it allows the evaluation of both viral oncolysis and immune responses in an immunocompetent setting. Still, it has substantial drawbacks: the tumor cells are of rodent origin and do not fully mimic human PDAC.⁶⁷ Moreover, the lack of widely available genetic tools and limited immune profiling in hamsters makes it challenging to investigate the interplay between virotherapy and host immune responses.⁷⁰

Quite recently, a novel murine PC cell line, KPC-1, was developed, which enables human adenovirus replication in immunocompetent mice. Unlike non-permissive murine models or partially permissive systems such as the Syrian hamster, KPC-1 enables adenoviral replication in both immunocompetent and immunodeficient hosts. This allows the assessment of viral replication, immune activation, and therapeutic efficacy in a more clinically relevant context. Importantly, it also allows the investigation of the complex relation between antitumor and antiviral immune responses, which was previously hindered by species-specific barriers. Thus, the KPC-1 model represents a significant advancement, providing a more clinically relevant platform for evaluating OAds in PDAC.⁷⁰

Despite the progress made, a fundamental limitation of existing preclinical models lies in their inability to fully recapitulate the complexity of the human TME and its interactions with an intact immune system. These limitations contribute to discrepancies between preclinical data and clinical outcomes. Although available models offer complementary strengths, no single one captures all relevant aspects of tumor-virus-host dynamics. Recent advances, such as PDOs and the KPC-1 model, represent necessary steps toward more predictive platforms. To improve the translational success of oncolytic virotherapy in PDAC, future efforts should focus on integrating diverse models to better replicate the disease’s defining characteristics, while also continuing the search for innovative, more clinically relevant and predictive preclinical models.

Route and timing of administration

Effective delivery of viral particles to the tumor site is crucial for the success of oncolytic virotherapy, and selecting the optimal route of administration is essential.^1,71 Intratumoral injection is commonly used in CTs as it maximizes drug concentration at the tumor site and is relatively simple to perform in anatomically accessible cancers, such as melanoma.^2,15 For tumors located in less-accessible areas, including PDAC, intratumoral injection can be more technically challenging.^1,2 However, with the use of imaging guidance, interventional radiology has made intratumoral delivery feasible for anatomical locations such as the pancreas and liver.^72,73 Other possible delivery methods include loco-regional administration, namely, hepatic arterial infusion using the TriSalus device and intraperitoneal administration for peritoneal metastases.^74,75,76

Oral administration is not a viable option due to the enzymes and pH of the gastrointestinal tract, which can inactivate many viruses.⁷¹ Ideally, systemic intravenous administration would be preferred as it is less invasive and painful for patients and simpler for healthcare workers. However, ensuring that viral particles reach the tumors in sufficient quantities to kill cancer cells effectively and activate an antitumor immune response is a significant challenge of systemic delivery.^2,21 Using OAds as an example, significant barriers to systemic delivery include pre-existing antibodies, non-tumor uptake, erythrocyte binding,² heterogeneous expression of the coxsackie and adenovirus receptor (CAR) on tumor cells, and variable tumor permissiveness.^77,78

The majority of the population possesses neutralizing antibodies (NAbs) against adenovirus serotype 5 from previous natural exposure. Since most OAds derive from serotype 5, this results in augmented immune-mediated viral clearance and reduced therapeutic efficacy.⁴³ Even individuals who have not been previously exposed and lack acquired immunity can develop it within 2 weeks of the first administration.² To overcome this obstacle, several strategies are being explored. One approach is switching to adenoviral serotypes with lower seroprevalence, e.g., serotype 6, 11, or 35, to reduce the likelihood of pre-existing immunity.^16,79 Additionally, engineering strategies, such as modifying capsid proteins or fiber domains, are being explored. For instance, VCN-11 is an OAds engineered with an albumin-binding domain in its capsid, enabling it to bind human albumin and evade NAbs.⁸⁰ Beyond evasion, studies have also explored strategies to harness pre-existing NAbs to improve delivery, instead of avoiding them. For instance, one method uses bifunctional adaptor molecules that link NAbs to tumor cells, consequently redirecting the immune system to attack the tumor and enhancing antitumor immune activation.⁸¹ Another strategy to use NAbs to our advantage is exploiting Fcγ receptor-mediated uptake of virus-antibody complexes by myeloid cells, which can transport the virus to tumor sites, bypassing antibody neutralization.⁸²

Another significant barrier to systemic delivery is non-tumor uptake by Kupffer cells, which are specialized resident macrophages of the hepatic reticuloendothelial system. These cells sequester a significant portion of adenoviral particles within minutes of systemic administration. Erythrocyte binding also hinders systemic delivery, as adenoviruses exhibit a strong tendency to bind with human erythrocytes, causing them to sequester viral particles and significantly reduce the amount of free virus in circulation and its overall bioavailability.²

CAR expression on the surface of cancer cells is another key factor, as most OAds, including those derived from serotype 5, must bind to this receptor to ensure attachment and successfully enter the cell. However, CAR expression is often dysregulated in cancer cells, leading to significant heterogeneity not only between cancer and normal tissues but also across tumor types and disease stages.⁷⁸ This variability can result in inefficient infection of tumor cells or even complete resistance to viral entry.

Variable tumor permissiveness is another critical limitation. For successful oncolysis, replication-competent OAds must enter tumor cells and replicate within them to produce new viral particles. However, many tumors lack the necessary intracellular environment for productive viral replication, rendering them non-permissive. In such cases, viral entry may occur, but replication is impaired or absent, resulting in minimal or no therapeutic benefit. Recent studies have shown that the TME plays a key role in modulating permissiveness to OVs, highlighting the complexity of host-virus interactions in solid tumors. To improve clinical outcomes and translation, testing patient-derived tumor biopsies for OV permissiveness before treatment and using ex vivo human tumor models may help identify responsive patients and prioritize tumor types for future CTs. These models also support the identification of factors driving permissiveness and contribute to personalized treatment decisions.⁷⁷

Another emerging delivery strategy is the use of nanoparticle-based and cell-based delivery systems, both designed to avoid rapid immune clearance.⁸³ Nanoparticle-based delivery systems consist of encapsulating the OV within nanomaterials such as polyethylene glycol. This process, known as PEGylation, entails covalently attaching polyethylene glycol to the viral capsid. By coating the virus with a polymer, the circulation time is increased by preventing NAbs from binding to the virus.^25,83 Liposomes are another type of nanomaterial widely tested for the encapsulation of OVs.⁸³

Cell-based delivery systems, particularly those using mesenchymal and neural stem cells, are promising vectors for OV delivery. These cells not only shield the virus from early immune detection and sequestration by the hepatic reticuloendothelial system but also facilitate delivery to tumor sites due to natural tumor tropism and can serve as efficient viral production factories.^83,84 Furthermore, chimeric antigen receptor T (CAR-T) cells, engineered ex vivo to have tumor-specific cytotoxic effector functions, are currently being explored in preclinical and clinical studies as carriers for OVs. While CAR T cells have shown substantial efficacy in hematologic malignancies, they have demonstrated limited success in solid tumors. To overcome this challenge, combining CAR T cells with OVs represents a promising strategy. CAR T cells could shield OVs from NAbs and enhance their delivery to the tumor, while simultaneously providing cytotoxic effects. The interaction between CAR T cells and viral oncolysis can result in a synergistic response, as viral oncolysis promotes local inflammation and antigen release, which attract additional CAR T cells and establish a positive feedback loop. Moreover, viral replication within the TME enhances T cell-mediated antitumor responses, further supporting this approach.^85,86

The timing of administration is also a critical factor. Given the potential of oncolytic virotherapy as part of combination treatments, it is crucial to assess whether it should be administered before, during, or after other therapies.⁶⁵ This aspect requires further evaluation in CTs.

Lack of predictive biomarkers

Currently, no predictive biomarkers are available to determine how a patient with PDAC will respond to oncolytic virotherapy. Given the significant intratumoral and intertumoral heterogeneity of PDAC, this aspect is of great relevance. Identifying predictive biomarkers would allow selecting patients who are most likely to benefit based on the specific characteristics of their tumor, leading to more personalized and effective treatment strategies. This approach could also maximize treatment efficacy by identifying the most suitable OV for each patient and avoid unnecessary treatments or trial-and-error approaches.^21,87 The lack of predictive biomarkers may be linked to low response rates observed in CTs, since there is no way of guaranteeing the patients selected have favorable tumor characteristics for that specific treatment. Addressing this issue is a key aspect of oncolytic virotherapy development, and efforts are underway to discover predictive biomarkers through studies investigating the sensitivity of patient-derived PDAC cell cultures to various OVs.⁸⁷

Future perspectives and conclusions

Despite great efforts, PC, particularly PDAC, remains a daunting malignant disease with an abysmal prognosis. This is mainly attributed to late diagnosis, early and aggressive local invasion and metastasis, and a characteristic immunosuppressive TME.⁶⁵ Available treatments for this disease are highly unsatisfactory. The tumor is often resistant to chemotherapy, and surgical resection, the only potentially curative treatment, is often unfeasible at the time of diagnosis. Immunotherapies, which have transformed treatment paradigms in several other cancers, have shown limited efficacy in PDAC, primarily due to the immunosuppressive nature of the TME.⁵⁹ Amid these disheartening facts, OVs have emerged as potential new allies because they not only selectively target and lyse cancer cells but also have the ability to elicit systemic antitumor immune responses.

Some CTs have been completed and multiple are ongoing, investigating a wide range of various virus types as platforms for oncolytic virotherapy, including adenoviruses, HSVs, reoviruses, vaccinia viruses, and alphaviruses. The key to implementing immunotherapy in PDAC seems to be combining different immunotherapies to target various immune evasion mechanisms and sensitize the tumor to the immune system. An example would be leveraging the synergistic effects of OVs and ICIs. ICIs in monotherapy have proven ineffective in PDAC, possibly due to a lack of cytotoxic T cells in the TME. Administered together, OVs would recruit cytotoxic T cells to the tumor and ICIs would ensure those cells remain active, thus having a synergistic effect.^15,88

To date, OVs have demonstrated a favorable safety profile, with most adverse events being mild and transient. This supports their integration into PDAC treatment regimens without compounding the substantial toxicity already associated with the current standard-of-care therapies. In terms of clinical efficacy, the available data are particularly encouraging for VCN-01 and CAN-2409, both of which have completed phase 2 CTs and demonstrated statistically significant improvements in survival outcomes. Phase 3 trials are currently being planned for both therapies, positioning them as leading candidates for future regulatory approval in PDAC treatment. Nonetheless, several challenges remain, and to optimize oncolytic virotherapy, it is essential to continue exploring different genetic modifications to produce OVs with enhanced tumor selectivity and improved immunogenicity.

Additionally, it is crucial to develop effective strategies for systemic delivery, to overcome biological barriers and minimize the loss of viral particles when using this administration route. Further research is also needed to determine the optimal timing and sequencing of OVs in combination with other therapies, including chemotherapy, radiotherapy, and immunotherapy, to maximizing therapeutic benefit while minimizing toxicity. Establishing optimal dosing regimens for each OV is equally important to improve treatment outcomes. Last, the identification of robust predictive biomarkers will be critical for guiding patient selection and enhancing clinical efficacy.

In conclusion, additional preclinical and clinical studies are essential to address these concerns and enhance the efficacy of oncolytic virotherapy, allowing its full therapeutic potential to be realized. If these challenges can be addressed, OVs may be integrated into the arsenal of treatment options for PDAC, a disease that is in urgent need of novel therapeutic alternatives.

Acknowledgments

This work was financially supported by the Portuguese Science and Technology Foundation (FCT) and the European Regional Development Fund (ERDF), through the Centro 2020 Regional Operational Programme under projects UIDB/04539/2020 and UIDP/04539/2020. This graphical abstract was created with illustrations from Servier Medical Art.

Author contributions

M.T.C. designed and structured the manuscript. M.E. performed the literature search, wrote the first draft of the manuscript, and prepared the figures. M.T.C. and A.M.M. revised and edited the final version of the manuscript. All authors read and approved the final manuscript.

Declaration of interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.