Relapsed and refractory classical Hodgkin lymphoma: could virotherapy help solve the equation?

ABSTRACT

Classical Hodgkin lymphoma is a neoplastic hematological disease. Standard first-line therapy, including chemotherapy and radiotherapy, is curative in >85% of early-stage patients, with a 5-year survival rate of >95%. However, approximately 15% of patients have hard-to-treat lymphoma with poor outcomes, and new treatment strategies are needed for these young adults. There are several well-documented cases in the medical literature on hematologic cancer remission following natural human viral infections. Therefore, hoping to reproduce these spontaneous tumor regressions, researchers have been investigating various viruses with oncolytic properties. There is a high rationale for using virotherapy in the treatment of Hodgkin lymphoma, in which tumor cells are often infected with the Epstein-Barr virus. Modern viral technologies and current knowledge about the relationship between viruses and cancer could accelerate the discovery of effective viral oncolytic therapies. This article reviews the use of oncolytic viruses as innovative therapies for treating Hodgkin lymphoma.

KEYWORDS:

• Hodgkin lymphoma
• virotherapy
• oncolytic viruses
• Epstein-Barr virus

Introduction

Classical Hodgkin lymphoma (cHL) is a B-cell lymphoid neoplasm that comprises a rich microenvironment with a minority of tumor cells, called Hodgkin Reed-Sternberg cells.^1–3

On their surface, Hodgkin Reed-Sternberg cells express CD30, a cell membrane protein and part of the tumor necrosis factor receptor superfamily, and CD15 in approximately 70% of cases.^4,5 Programmed death-ligand 1, which blocks cytotoxic T lymphocytes, is overexpressed in virtually all cases of cHL.^6,7

Hodgkin Reed-Sternberg cells are infected by Epstein-Barr virus (EBV) in approximately 40% of cases.⁸ The EBV plays an important role in the pathogenesis of cHL (mainly the mixed cellularity subtype) and exhibits a type II form of virus latency with the Epstein-Barr nuclear antigen-1, latent membrane protein (LMP) 1, and LMP 2 A.⁹ Epstein-Barr nuclear antigen-1 is a transcription factor that regulates both viral and cellular gene expression and recruits regulatory T-cells. LMP 1 acts as a mimic of CD40 and may constitutively activate the NF-κB and JAK/STAT pathways. LMP 2A involves B-cell receptor (BCR) signaling, leading to B-cell activation in the absence of normal BCR signaling.¹⁰

From a therapeutic perspective, polychemotherapy combined with radiotherapy in early stage cHL is a common first-line therapy for cHL, with 85% curability.¹¹ The 15% of patients who do not achieve complete response after first-line therapy represent a hard-to-treat population.^12,13

In patients with cHL relapsing or refractory (R/R cHL) to first-line therapy, second-line polychemotherapy followed by consolidation with autologous stem cell transplantation (ASCT) is curative in approximately 50% of cases.^14,15 Of note, ASCT may not be an option for some elderly patients with comorbidities or in cases of chemorefractory disease at the time of salvage.¹⁶

As a new, modern, and vectorized chemotherapy, brentuximab vedotin is a targeted antibody-drug conjugate that is active against CD30-positive cancer cells. The drug was approved by the Food and Drug Administration (FDA) in 2011 for R/R cHL patients who were post-ASCT, or for those patients not eligible for ASCT after at least two lines of therapies.^17,18

More recently, monoclonal antibodies targeting the programmed death receptor 1 (PD-1) have dramatically improved the prognosis of patients with R/R cHL by attaining up to 68% antitumoral response.¹⁹ Two anti-PD1 therapies (nivolumab and pembrolizumab) were approved in the last 5 years by the FDA²⁰ for R/R cHL patients, showing impressive efficacy, even in heavily pre-treated patients.^21,22 The median progression free survival with anti-PD1 was approximately 12 months, and most patients achieved a partial response,^19,23–26 indicating that anti-PD1 does not cure most patients. To overcome resistance to anti-PD1 and improve response quality, combinations with other immunotherapies are in various stages of development;^27,28 however, to date, no anti-PD1 combination has demonstrated significant efficacy in overcoming anti-PD1 resistance in Hodgkin lymphoma.

In fact, new treatment methods that show accurate tumor targeting, powerful tumor-killing properties, and low toxic side effects need to be proposed.²⁹ Virotherapy could be a good candidate, especially in the treatment of Hodgkin lymphomas, in which tumor cells are often infected with the EBV.

Oncolytic viruses (OVs) are natural viruses or genetically modified viruses that exhibit antitumor and antiviral activity with tumor tropism that target and kill cancer cells without harming normal cells.³⁰ Oncolytic viruses have the ability to induce immunogenic cell death as part of its replication cycle, and they also have powerful immunomodulatory effects that induce both innate and adaptive tumor responses.³¹ In this review, we discuss approaches with OVs to treat patients with R/R cHL, and their ability to eventually enhance the antitumor effect of other therapies such as anti-PD1 immunotherapies.

General properties of oncolytic viruses

Oncolytic viruses consist of an envelope containing the viral capsid, which in turn surrounds the viral genome, a single- or double-stranded RNA or DNA.³² The selection of oncolytic viruses for therapeutic drug development is based on numerous virus criteria, such as the viral genome, size, pathogenicity, immunogenicity, transgene capacity, blood-brain penetration, mechanism of cell entry, and stability of the virus (Table 1).

Table 1. Overview of oncolytic viruses characteristics in ongoing drug development in oncology.³¹

Oncolytic virus	Génome	Size (Kb)	Transgene capacity	Pathogenicity	Immunogenicity	BBB Penetration
Adenovirus	DsDNA	32	Moderate	Gastroenteritis Pneumonia Meningitis Pharyngitis Follicular conjunctivis	Low	Very limited
VSV	Ss (-) RNA	11	Low	Fever Headache Myalgia weakness Occasionally vesicular lesions	High	High
MV	Ss (-) RNA	16	Low	Measles pneumonia	Low	High
HSV1	DsDNA	152	High	Herpes vesicle Possible CNS latent infection	Low	Limited
Poliovirus	Ss (+) RNA	7.5	Low	Poliomyelitis Paralysis	Moderate	High
Myxoma virus	DsDNA	130–375	High	Fever Myalgia Encephalitis Vaccinia Eczema Lymphadenopathy	High	Very limited
Newcastle Desease virus	Ss (-) RNA	15	Low	Transitory conjunctivitis	Low	Very limited

Ss (-), single-stranded negative sense; BBB, blood-brain barrier; CNS, central nervous system; DsDNA, double-stranded DNA; VSV, vesicular stomatitis virus; MV, measles virus; MYXV, myxoma virus; HSV1, herpes simplex virus type 1.

Clinical experience with DNA viruses such as herpes simplex virus (HSV) and adenovirus (Ad) is more extensive, and they are generally easier to engineer genetically than RNA viruses because they can encode large transgenes for increasing therapeutic activity or immune modulation. They induce a strong immune response with strong anti-tumor immunity, but they can also be cleared more rapidly by the immune system.³² Therefore, DNA viruses may be more suitable for intratumoral administration. In contrast, RNA viruses such as the measles virus (MV; paramyxovirus) and vesicular stomatitis virus (VSV; rhabdovirus) are not as conducive to genetic manipulation because of their small size, but can penetrate tumors with efficacy and are more likely to cross the blood-brain barrier. The reduced likelihood of preexisting immunity to RNA viruses could make intravenous administration more feasible.³³

There are two types of OV. On one hand, there are naturally occurring viruses that are nonpathogenic to humans, such as the myxoma virus (MYXV; poxvirus) and Newcastle disease virus (paramyxovirus). On the other hand, there are viruses that are genetically manipulated for use as vaccine vectors, including MV, which has an excellent safety record after decades of use in humans,³⁴ poliovirus (picornavirus), and vaccinia virus (VV; poxvirus), and those viruses genetically engineered with mutations or deletions in genes such as Ad. Although Ad encodes protein E1A, to promote safety and prevent replication in normal cells, the E1A gene was deleted.³⁵ The HSV has strong lytic properties, and many variants have been engineered via deletion of ICP34.5 neurovirulence and ICP6. VSV is an engineered variant that overexpresses interferon β to protect normal cells from infection and causes tumor cell–specific destruction and antitumor immune responses.³⁶

Insertion of foreign sequences can provide selectivity for cancer cells and safety, as well as alter virus tropism by targeting translation, transcription, or transduction.

Oncolytic viral activity toward cHL

Oncolytic viruses can target and kill cHL cells via the natural interactions between viruses, tumor cells, and the immune system³⁷ (Figure 1). They can be engineered to selectively target Hodgkin Reed-Sternberg cells without harming normal cells, as described in a study by Hanauer et al. study³⁸ where MV-CD30 and VSV-CD30 were generated by replacing envelope genes in the genome of MV and VSV with those encoding CD30-targeting glycoproteins.

Figure 1. Mechanisms of action of oncolytic viruses as indicators of anti-tumoral activity in Hodgkin lymphoma

(a) The oncolytic virus (OV) infects and replicates in the classical Hodgkin lymphoma (cHL) cell because of an abnormal response to stress and dysfunctional virus detection and clearance by protein kinase R (PKR), retinoic acid inducible gene 1 (RIG-1), and interferon type 1 (INF). (b) Direct cHL cell-killing by OV results in the release of tumor-associated antigen (TAA), viral pathogen-associated molecular patterns (PAMPs), damage-associated molecular patterns (DAMPs), and cytokines (Interleukin (IL)12, IL6, IL1, INF, etc). (c) TAA, PAMPs, DAMPs, and cytokines activate antigen-presenting cells (APC) resulting in recruitment of other APCs such as dendritic cells (DC) and natural-killer cells (NK), promoting innate immunity activation. (d) Immunologic death of cHL cells by NK in response to INF and IL12. (e) DC, in response to IL12, migrates to lymph-node areas to activate CD4+ and CD8 + T-cells through major histocompatibility complexes (MHC) I and II. CD4 + T-cells release IL2, which helps lymphocyte B (LB) cells produce cell-destructive anti-tumor antibodies. TCR, T-cell receptor. (f) Activated CD8 + T cells initiate antiviral and anti-tumoral activity. CD8 + T cells migrate, in response to chemokine, to cHL cells that express MHC I.

Importantly, in the context of EBV-positive Hodgkin lymphoma,^39,40 an adenoviral vector expressing a transgene (either LMP2 alone or both LMP2 and LMP1) can act as a vaccine to treat R/R cHL, as shown by Bollard et al.⁴¹

Oncolytic adenovirus can also accommodate genetic insertion of therapeutic and immunomodulatory transgenes, such as the immune target anti-PD1, leading to enhanced efficacy in R/R cHL.^42–44

Some researchers suggest that intratumoral or intravenous OV administration can selectively access tumor cells by fusion with the plasma membrane, endocytosis, or by binding to tumor cell surface receptors. The oncolytic virus then replicates in the cHL cells and induces direct cell death^32,45 through an abnormal response to stress and dysfunctional virus detection and clearance by protein kinase R, retinoic acid inducible gene 1, and interferon type 1 responses. The tumor tropism of the OV is used as a key property to maximize direct killing effects on the tumor, while minimizing damage to surrounding normal tissues.⁴⁶

The direct killing of cHL cells by OVs result in the release of danger signals such as damage-associated molecular patterns, pathogen-associated molecular patterns, and tumor-associated antigens, thus inducing innate immunity.⁴⁷ These danger signals are picked up by antigen-presenting cells (APCs), such as macrophages and dendritic cells, which express antigens coupled with the major histocompatibility complex classes II and I. In addition, APCs promote the release of cytokines (such as tumor necrosis factor-α, interferon, interleukin [IL] 12, and IL-6), resulting in the recruitment and activation of innate immune cells.⁴⁶ The innate cells recruited are natural killer cells, natural killer T-cells, and γδ T-cells.^32,48 Natural killer cells have the potential to induce cHL death through a cytotoxic immune response boosted by cytokine secretion. This cell death mechanism is also known as immunogenic cell death.⁴⁹

Mature dendritic cells, in the presence of IL-12, migrate to regional lymph nodes where they present T-cells with danger signals that are coupled to major histocompatibility complex classes I and II. This mechanism could lead to tumor regression at distant tumor sites not exposed to the OV via increased T-cell recruitment and immune cell activation.⁵⁰ Local administration of OV is globally well tolerated. The most common adverse effects of OV are local reactions at the injection site and flu-like symptoms, which may be more severe after systemic administration and which do not allow the inclusion of immunocompromised patients or those with active viral infections.⁵¹

Many viruses are also pathogens and cause human disease, which could limit their clinical use, although deletion of virulence genes may render these viruses safe for human administration. In addition, some viruses remain sensitive to oral antiviral drugs.^{52, 53}

Oncolytic viruses as an innovative approach to target tumor cells

Oncolytic viruses and solid tumors

In 2015, the FDA approved the first OV for melanoma, talimogene laherparepvec,^53,54 an oncolytic herpes virus (HSV1) engineered to produce granulocyte macrophage colony-stimulating factor (GM-CSF). Talimogene laherparepvec demonstrated a superior and durable response rate and overall survival in patients.⁵⁴ Since then, many trials have been conducted to assess the safety and efficacy of multiple OVs in patients with various types of cancers.⁵⁵

Oncolytic viruses and the general approach to treat Hodgkin lymphoma

The oncolytic potential of viruses in Hodgkin lymphoma was first recognized in the 1970s, with a case series showing regression of Hodgkin’s disease in patients with concurrent measles infections.^56,57 In the context of R/R cHL, some recent clinical studies have engineered OVs to target LMPI, LMP2A, and dominant negative receptor II TGF β (DNRII), and demonstrated cancer remission without significant toxicity.^41,58 Recent studies have demonstrated the safety and activity of in vivo expansion of OVs to treat patients with R/R cHL or to eventually prevent relapse (Table 2).

Table 2. Summary of safety and efficacy results for oncolytic viral approaches to treat relapsed or refractory Hodgkin lymphoma in clinical trials

Author, year	Type of OV	Expressed Transgene	No of included HL patients	Pré – OV		Post-OV
Bollard et al, 2014^Citation41	Adenovirus	LMP1-LMP2	25	17 CR	8R/R	17 + 4 CR	4 SD
Bollard et al, 2018^Citation58	Adenovirus	LMP1 – LMP2	8	1 CR	7R/R	1 + 2 CR 1 PR	4 SD

OV, oncolytic virus; HL, Hodgkin lymphoma; CR, complete response; R/R, relapsed/refractory disease; PR, partial response; SD, stable disease; LMP, latent membrane protein.

Oncolytic viruses in Hodgkin lymphoma, targeting LMP 1 and LMP 2

Bollard et al. engineered genetically modified APCs transduced with an adenoviral vector expressing either LMP2 alone or associated with LMP1, for the purpose of expanding autologous cytotoxic T-lymphocytes (CTLs) to target EBV-associated lymphomas, including R/R cHL. Twenty-eight of 29 (96%) high-risk or multiple-relapse patients receiving LMP-CTLs as adjuvant therapy remained in remission at a median of 3.1 years after CTL infusion. Sustained remission without toxicity related to virotherapy were observed.⁴¹ In comparison, in a study evaluating the efficacy of anti-PD1 re-treatment in R/R cHL after anti-PD1 discontinuation in seven patients, the disease was in complete remission in six patients and partial remission in one patient. The median time to relapse after anti-PD1 was 12.1 months (range: 5.3–26.7 months).⁵⁹

Oncolytic viruses in Hodgkin lymphoma, targeting TGF β

Another study was conducted in patients with R/R cHL by Bollard et al. and consisted of CTLs that expressed DNRII transduced by an Ad expressing LMP1 and LMP2 transgenes. In this study, DNRII-T cells expanded and persisted for up to 4 years after infusion in patients with EBV-positive R/R cHL. Four out of seven patients with active disease were clinically responsive, and of these four, two patients had ongoing clinical response at 4 years.⁵⁸

Oncolytic viruses in Hodgkin lymphoma, targeting LAG-3

Lymphocyte activation gene-3 (LAG-3) is a marker of regulatory T-cells and is strongly expressed on infiltrating lymphocytes present in proximity to Hodgkin Reed-Sternberg cells.⁶⁰ EBV-positive cHL contains regulatory T-cells in its microenvironment, which generally highly expresses LAG-3. Lymphocyte activation gene-3 expression on tumor-infiltrating lymphocytes increases impairment of T-cell activity against LMP1 and LMP2.⁶⁰ These data suggest the potential capacity of engineered OVs like Ad to block LAG-3, a critical immune checkpoint in EBV-positive Hodgkin tumors.

Oncolytic viruses in Hodgkin lymphoma, targeting CD30

Hanauer et al. generated CD30-targeted MV (MV-CD30) and VSV (VSV-CD30) in order to selectively target CD30-positive cHL-cells in a xenograft mouse model.³⁸ CD30 targeting relies on engineered MV envelope glycoproteins created by replacing envelope genes in the genomes of MV and VSV with single-chain variable fragments encoding the specific stability-engineered anti-CD30. Each virus exhibited limited infection and significant spreading within the CD30-positive cHL cells. On comparing their oncolytic activities, VSV-CD30 turned out to be more efficient than MV-CD30 and resulted in a more rapid and efficient killing of cultivated cHL-derived cell lines. In the clinical study, the OVs induced no side effects (including neurotoxic effects), regardless of intratumoral or systemic administration, and it was shown that the CD30-targeted viruses selectively infected CD30-positive cells. Furthermore, CD30-targeted viruses can be transported via CD30-positive extracellular vesicles in the vicinity of cHL tumor cells,⁶¹ leading to infection and killing of bystander cells in the tumor microenvironment.³⁸

Rationale for oncolytic viruses in combination with traditional therapies for Hodgkin lymphoma

Oncolytic viruses in combination with chemotherapy and radiotherapy

Some studies suggest that the combination of chemotherapy and radiotherapy with OV treatment could have synergic antitumor activity.⁶² Chemotherapy induces tumor apoptosis through endoplasmic reticulum stress, and combination with ONYX-015, an Ad with a deleted E1B gene, could enhance direct cell death.⁶³ In addition, some OVs like HSV1 sequester DNA damage-response proteins and enhance radiation sensitizers.⁶⁴ In addition, chemotherapy and radiotherapy could increase OV antitumor activity by releasing danger signals and cytokines.^31,65

Oncolytic viruses combined with autologous stem cell transplantation

Some researchers suggest that oncolytic viruses, such as adenoviral vectors expressing the thymidine kinase gene and HSV, have the potential to prevent relapse after autologous stem cell transplantation by treating residual tumor cells before the transplant procedure.⁶⁶ Despite improvements in outcomes after ASCT, nearly 50% of patients relapse after transplant. Some studies have proposed that graft contamination by tumor cells can be the origin of relapses.⁶⁷ CHL cells also release extracellular vesicles containing CD30 cleaved by metalloproteinases and detected in the plasma of cHL patients.^61(p30) Uccini et al. suggested that EBV-infected cHL cells may release LMP1 through exosomes and contribute to the generation of the immunosuppressive microenvironment of cHL by producing immunomodulatory cytokines such as IL10.⁶⁸ Ex vivo manipulations of the ASCT to purge cancer cells using chemotherapies and toxins have been attempted, but these strategies often impair normal hematopoietic stem cell survival in the graft.⁶⁷

Oncolytic viruses such as MYXV may be used as purging agents that selectively target cancer cells and exosomes while sparing normal stem cells within the collection product; this approach can be applied without any additional toxicity for the graft recipient.^69,70 Preclinical in vitro and in vivo studies have proposed the use of different OVs, such as MV, VV, parvovirus, VSV, and coxsackie virus, to purge cancer cells.^71–73 These data suggest that OVs targeting CD30 and/or LMP1-purging strategies could be explored in further studies in the context of R/R cHL patients undergoing ASCT.

Oncolytic viruses combined with immune checkpoint inhibitors (ICIs)

Oncolytic viruses can eventually enhance the antitumor effect of ICIs in R/R cHL, as demonstrated in multiple ongoing clinical trials to treat patients with solid tumors, and have shown encouraging preliminary results.^74,75 Oncolytic HSV1 may improve clinical benefits of ICIs without overlapping toxicity profiles.⁷⁶ It can modulate tumor-specific CD8-positive T-cell responses and release TAA, danger signals, and pro-inflammatory cytokines. When administered intratumorally, this heightened inflammatory response can induce a systemic or abscopal effect and render distant tumors increasingly susceptible to ICI therapy through increased T-cell recruitment and immune cell activation.^50,77,78

A phase II trial combining intratumoral oncolytic HSV1 engineered to produce GM-CSF and intravenous ipilimumab (talimogene laherparepvec plus anti-CTLA-4 antibody) in patients with advanced melanoma demonstrated abscopal responses at tumor sites distant to the injected lesions, and decreases in visceral lesions were observed in 52% of patients in the combination arm versus 23% of patients in the anti-CTLA-4 antibody alone arm.^79,80

Oncolytic adenoviruses also increase the expression of PD1 and other checkpoint molecules in immune cells. Approximately 30% of infused DNRII-T cells express PD1.⁵⁸ Some studies, such as the ongoing Masterkey-265 phase Ib/III trial, have proposed assessment of intralesional talimogene laherparepvec in combination with intravenous pembrolizumab.⁸⁰

Moreover, OVs can be engineered to express ICI genes to increase their antitumor activity without additional toxicity. Dias et al. reported an oncolytic Ad that included a transgene expressing a specific CTLA-4 antibody. In peripheral blood mononuclear cells of patients with advanced tumors, OVs can demonstrate prolonged and localized delivery of CTLA-4 antibodies at a high level.⁸¹

Oncolytic viruses in the context of allogenic stem cell transplant: A rationale for the prevention and treatment of graft-versus-host disease

In patients with cHL who relapse after anti-PD1 treatment, allogenic stem cell transplant (Allo SCT) can be a therapeutic option for those who achieve a complete response with salvage therapies and have a donor available. However, acute graft-versus-host disease (GvHD) and transplant-related mortality remain important issues with Allo SCT.⁸² The onset of GvHD is mediated by allogenic donor-derived T lymphocytes, which upon transplantation, are activated by mismatched major and/or minor histocompatibility complex tissue-specific antigens of the recipient. As a result, the pro-inflammatory cytokines released by these activated T-cells ultimately target and damage healthy organs and tissues such as the digestive tract, liver, and skin.⁸³ A pre-clinical study reported that MYXV can prevent GvHD in xenografted mice without impairing graft-versus-tumor (GvT) effects against residual cancer cells.⁸⁴ Several preclinical data have suggested that ex vivo virotherapy of human HSCT samples with MYXV can delay the onset of GvHD by interfering with donor T-cells. Furthermore, histological analyses of the recipient organs revealed very low infiltration by donor lymphocytes.⁸⁵

In a subsequent study,⁸⁶ MYXV confirmed its potential to inhibit the development of GvHD after xeno-SCT. The virus efficiently binds to resting human CD3-positive T lymphocytes, becoming “carrier cells,” following T-cell activation. The virus replication cycle simultaneously inhibits T-cell proliferation and decreases the expression of GvHD-promoting cytokines. Furthermore, MYXV-infected activated T-cells were able to deliver viruses to cancer cell lines, resulting in their infection and further killing of tumor cells, as well as an enhanced GvT effect. Therefore, in this study, MYXV demonstrated an interesting dual role in not only inhibiting GvHD, but also enhancing GvT effects against residual cancer cells.

Conclusion

Among the various virotherapeutic perspectives for treating patients with R/R Hodgkin lymphoma, the main therapeutic approaches are targeted toward the viral proteins LMP1 and LMP2. Virotherapeutic approaches can also target the tumor itself, such as via CD30, and the surrounding suppressive tumor microenvironment via TGF-β and LAG3. Virotherapy could thus have multiple applications in therapeutic sequences by providing consolidation treatment after ASCT relapse and boosting the efficacy of anti-PD1 (Figure 2). Further studies are warranted to assess the potential role of virotherapy in the treatment of Hodgkin’s lymphoma.

Figure 2. Choosing virotherapy for the treatment of relapsed and refractory classical Hodgkin lymphoma

(a) First-line therapy(b) Second-line therapy with OV as consolidation treatment after ASCT to prevent relapse.(c) OV monotherapy or in association with third-line therapy, such as anti-PD1, to treat R/R CHL.OV, Oncolytic virus; R/R CHL, relapsed and refractory classical Hodgkin lymphoma; ASCT, autologous stem cell transplantation.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.