Comparison of preclinical efficacy of immunotherapies against HPV-induced cancers

ABSTRACT

Introduction

Persistent infections with the human papilloma viruses, HPV16 and HPV18, are associated with multiple cancers. Although prophylactic vaccines that induce HPV-neutralizing antibodies are effective against primary infections, they have no effect on HPV-mediated malignancies against which there is no approved immuno-therapy. Active research is ongoing in the immunotherapy of these cancers.

Areas covered

In this review, we compared the preclinical efficacy of vaccine platforms used to treat HPV-induced tumors in the standard model of mice grafted with TC-1 cells, which express the HPV16 E6 and E7 oncoproteins. We searched for the key words, ‘HPV,’ ‘vaccine,’ ‘therapy,’ ‘E7,’ ‘tumor,’ ‘T cells’, and ‘mice’ for the period from 2005 to 2023 in PubMed and found 330 publications. Among them, we selected the most relevant to extract preclinical antitumor results to enable cross-sectional comparison of their efficacy.

Expert opinion section

We compared these studies for HPV antigen design, immunization regimen, immunogenicity, and antitumor effect, considering their drawbacks and advantages. Among all strategies used in murine models, certain adjuvanted proteins and viral vectors showed the strongest antitumor effects, with the use of lentiviral vectors being the only approach to result in complete tumor eradication in 100% of experimental individuals while providing the longest-lasting memory.

KEYWORDS:

• Antitumor immunity
• cervix cancers
• early E6 and E7 oncoproteins
• HPV-induced cancers
• immuno-oncotherapy
• oropharyngeal cancers
• preclinical antitumor efficacy
• T cell vaccines

1. Introduction

Human papillomaviruses (HPVs) are small, non-enveloped double-stranded DNA viruses of the Papillomaviridae family. HPVs are the most frequently sexually transmitted pathogenic agents in the world [1]. Persistent infections with ≈14 of more than 200 known related HPV genotypes are associated with epithelial tumors and cancers of the cervix, head and neck, vulva, vagina, penis, and anus [2–7]. Cervical cancer is the fourth most frequent cancer of women worldwide after breast, colorectal, and lung cancer. Almost all cases of cervical cancer result from chronic infections caused by high-risk HPV subtypes. Among them, 70% to 75% result from infection with the most oncogenic and prevalent genotypes, HPV16 and HPV18 [3]. The vast majority of cervical cancer cases occurs in developing countries and are typically diagnosed in 30- to 50-year-old women [8,9]. HPVs are also responsible for a rapidly increasing number of head and neck squamous-cell carcinomas, which affect the nasopharynx, paranasal sinuses, oral cavity, larynx, hypopharynx, and oropharynx [10]. There are 650,000-to-890,000 new cases of head and neck cancer worldwide, leading to 350,000–450,000 deaths per year [11,12]. Most head and neck cancers are diagnosed in 50- to 70-year-old patients, with a higher incidence in men than in women [12]. The global incidence of head and neck cancer includes 98,000 oropharynx cancers per year [11]. HPV infections cause 71% and 52% of all oropharyngeal cancers in the U.S.A. and U.K., respectively. Among them, at least 85% to 96% are caused by HPV16. Oropharyngeal cancers mainly concern the tonsils, base of the tongue, soft palate, and uvula [13].

The ≈8-kb circular DNA genome of HPV encodes: (i) two Late structural subunits of the icosahedral viral capsid, i.e. the major capsid L1 and minor capsid L2 proteins, and (ii) six Early proteins, i.e. E1-E2, E4-E7, main drivers of viral replication [14]. The mono-layered basal epithelial cells at the junction between the endocervix columnar epithelium and the ectocervix squamous epithelium, as well as the mono-layered reticulated epithelium of tonsillar crypts, are particularly susceptible to HPV infections and further carcinogenesis [10,15]. In the early stages of infection, the HPV genome is maintained in the episomal form in the nuclei of the host cells. At this stage, the E2 protein is predominantly expressed in squamous epithelial cells. During subsequent stages, expression of the E6 and E7 zinc-finger oncoproteins increases following integration of the HPV genome into the host genome, which disrupts the E2 open reading frame, leading to de-repression of the genes encoding E6 and E7. This process causes an exit from the productive viral life cycle and initiates the development of neoplasias with unfavorable prognoses [13,14]. However, the absence of viral genome integration and the presence of only the episomal form of the HPV genome have been recorded in 26% of both cervical and oropharyngeal cancer cases [16].

The main mechanism of carcinogenesis relies on the action of E7. This oncoprotein inactivates the tumor suppressor retinoblastoma (Rb) protein by altering its phosphorylation state, resulting in abnormal mitosis, aneuploidy, and genomic instability [14]. In parallel, E6 induces ubiquitination and proteasomal degradation of p53, the key regulator of cell proliferation, resulting in resistance to apoptosis. These processes disrupt the differentiation and maturation of metaplastic squamous epithelium and lead to dysplasia [14,17]. E6 and E7 also promote angiogenesis, metastasis, and epigenetic changes, which can be associated with somatic alterations in the host cell genome [18].

Precancerous cervical lesions are classified by the cervical intraepithelial neoplasia (CIN) stages. In >90% of the cases, the infection is cleared at CIN1 and CIN2 due to natural host immunity. The risk of CIN2 recurrence increases with HPV persistence, particularly during the first-year post-infection [19,20]. However, up to 70% of malignancies can still regress at CIN3. In 3% to 5% of the cases, low-grade lesions progress to malignant transformation after 10–20 years of chronic infection, persistent viral DNA synthesis, and the failure of anti-HPV T-cells to infiltrate pre-malignant lesions. The cancer cells may cross the basement membrane to invade the surrounding tissues. Later-stage disease is more often killed by local invasion than by metastasis [21–23].

The currently approved prophylactic anti-HPV vaccines are very effective against primary infections [24–28] but show no efficacy against persistent HPV infections or already malignant tissues. Despite active preclinical research, no immunotherapy has been thus far approved for use against human HPV-induced cancers. A cross-sectional comparison of the efficacy of various preclinical immuno-therapeutic strategies will highlight those with the greatest potential to produce successful clinical results in the future. Here, with this perspective, we overview the immunotherapy candidates which, for the most part, have been evaluated in the standard model of TC-1 tumors in mice [9]. In our research, we looked for the key words ‘HPV,’ ‘vaccine,’ ‘therapy,’ ‘E7,’ ‘tumor,’ ‘T cells’, and ‘mice’ in PubMed for the period from 2005 to 2023. Among the 330 found publications, a selection of 13 relevant articles was made. The latter include the most efficient preclinical results in the TC-1 tumor model while covering all relevant vaccination strategies, i.e. protein, bacterial or viral vectors, and nucleic acid-based approaches. The selected articles were analyzed in detail to enable an in-depth comparison of their preclinical efficacy, which should help focus attention on the most promising strategies for implementing future clinical trials.

2. Current prophylactic vaccines and possible future immunotherapies

The currently available preventive HPV vaccines are all based on the major capsomer L1 protein, packaged into virus-like particles and formulated in alum and/or mono phosphoryl lipid A, to induce neutralizing antibodies that prevent infection [29,30]. These prophylactic vaccines are widely used in 88% of high-income countries. Worldwide coverage with the first dose of HPV vaccine is currently estimated at 21%. These preventive vaccines have shown a compelling potential to reduce the incidence of HPV infections, anogenital warts, and subsequent CIN2, CIN3, and cervical cancers, especially for individuals vaccinated at the age of 12–13 years [24–28], with evidence of herd protection [31]. Nevertheless, prophylactic HPV vaccination has no impact on the course of persistent HPV infections or malignant tissues. In addition, a large proportion of individuals remain unvaccinated, particularly in developing countries, which will result in high numbers of new HPV-caused cancers in future decades. There is therefore an urgent need for new therapeutic approaches based on T cell-mediated immunity to eradicate the already transformed cancer cells [32].

The most appropriate antigens as targets of therapeutic T-cell vaccination are the E oncoproteins. Among them, E6 and E7 are expressed in all malignant cells, making them a major target for the immunotherapeutic strategies developed over the last decades [1,13]. The oncogenicity of E6 and E7 results from interactions between each of their multiple small protein-binding sites with several host proteins. Importantly, these individual-binding domains per se can preserve their oncogenic properties [33,34]. Therefore, for safety reasons, careful sequence modification is necessary to inactivate all known protein-binding sites in E6/E7-based antigen designs.

The ability of HPVs to interfere with immune mechanisms must be accounted for designing immuno-oncotheraies. In this regard, several mechanisms allow HPVs to evade the host immune system. HPVs mainly encode non-secreted nucleoproteins, which are thus not efficiently cross-presented by major histocompatibility complex (MHC) molecules of the host antigen-presenting cells. In addition, most non-structural HPV proteins are expressed at low levels relative to other viral immunogenic proteins [1]. HPV infections have no viremic phase because the infected cells do not undergo cytolysis. Consequently, no abundant systemic antigen expression or presentation occurs. HPV infections induce cell proliferation rather than cytolysis, which does not favor the expression of danger signals able to trigger innate immunity [35]. In addition, HPVs only infect stratified epithelium, which is an anti-inflammatory environment, characterized by predominant expression of the immunosuppressive cytokines IL-10 and TGF-β [36]. The E5 protein can induce alkalization of late endosomes, which inhibits MHC molecule migration to the cell surface. E6 can decrease type I IFN expression. E7 may cause the upregulation of programmed cell death-1 ligand (PD-L1) by the tumor cells and also inhibits the cyclic GMP-AMP synthase (cGAS) – stimulator of interferon genes (STING) signaling pathway, a main inducer of type I IFN expression [18].

However, despite the above-mentioned mechanisms developed by HPVs to evade the host immune system, there is strong evidence in support of active T cell-mediated immunosurveillance of HPV-induced malignancies. In fact, genetic predisposition, linked to human leukocyte antigen (HLA) alleles, i.e. HLA-DQB1^*0602 and HLA-DRB1^*1501, has been shown to be associated with the occurrence of HPV-induced cancers [37]. Increased incidence of HPV infections and HPV-mediated cancers has also been recorded in immunocompromised patients [38]. Furthermore, progression to the invasive phase is often accompanied by failure in the induction of anti-HPV T cells capable of infiltrating pre-malignant lesions.

Therefore, the success of prospective immuno-oncotherapies may result from their capacity to induce robust and long-lasting memory T cell responses against pertinent HPV-derived antigens along with the ability to infiltrate precancerous lesions and kill dysplastic precancerous cells by cytolysis. Several HPV-specific therapeutic vaccine candidates based on various platforms have been evaluated in preclinical and clinical trials by numerous laboratories. None of these strategies have yet been approved for human use, even though encouraging clinical results have been reported in anogenital cancer immunotherapy [39–41] (https://www.clinicaltrials.gov/study/NCT03260023).

Preclinical research on immuno-therapy of HPV-induced cancers has been carried out mainly using the TC-1 model in almost all expert laboratories [9]. This makes it easier to cross-compare the effectiveness of the strategies used. The TC-1 fibrosarcoma cell line, of pulmonary origin, has been immortalized by stable co-transfection with E6 and E7 from HPV16 (E6_HPV16 and E7_HPV16) and the Harvey rat sarcoma (HRas) viral oncogene [42]. Like HPV-induced human tumors, TC-1 cells have the advantage of constitutively expressing the E6_HPV16 and E7_HPV16 protooncogenes [9]. They readily form tumors after subcutaneous or orthotopic transplantation and are cytolyzable by CD8⁺ T cells. TC-1 cells have a basal level of PDL1 expression that is up-regulated when the tumor microenvironment becomes inflammatory, making them suitable for studying the effect of therapeutic combinations based on anti-PD1 treatments. However, it should be mentioned that it is challenging to transpose the promising results obtained with the TC-1 preclinical model to the clinical treatment of human HPV-induced cancers. These results should indeed be viewed in the light that TC-1 cells, like the vast majority of preclinical tumors, only partially recapitulate the characteristics of human tumors and have a relatively limited predictive value [43]. In fact, unlike HPV-induced human tumors, TC-1 tumors do not down-regulate the expression of their surface MHC-I in vivo and have a microenvironment well permeable to immune cell infiltration and relatively easy to switch to inflammatory with effective immunotherapies. In addition, the immune system of young, healthy C57BL/6 mice transplanted with TC-1 tumors is certainly more functional than that of cancer patients. Despite these differences between the TC-1 and human HPV-induced cancers, numerous therapeutic HPV vaccine candidates have reached clinical trials based on the proof of concept established in the TC-1 model [43].

In the following paragraphs, we outline selected experimental results generated using the most relevant immunotherapy strategies developed since 2005 and evaluated in the TC-1 preclinical model. An overview of the clinical results in the field is beyond the scope of this review, although rare examples are cited.

3. HPV immunotherapies based on protein vectors

3.1. Fusion of extra domain a from fibronectin with E7_HPV16/18

The 390 amino-acid protein extra domain A from fibronectin (EDA) is notably produced in response to tissue injury in inflammatory conditions [44,45]. EDA is a toll-like receptor (TLR)4 agonist and triggers dendritic-cell maturation. A fusion protein of human EDA (hEDA) and full-length E7_HPV16/18 was engineered to produce a bivalent therapeutic vaccine candidate (hEDA-E7_HPV16/18) [44]. C57BL/6 mice were immunized by two subcutaneous (s.c.) injections, 7 days apart, with 3 nmoles hEDA-E7_HPV16/18 adjuvanted with 50 µg of the TLR3 agonist poly ICLC (a stabilized polyinosinic-polycytidylic acid) [46]. Strong T-cell responses against both E7_HPV16 and E7_HPV18 were detected by ELISPOT [44].

The therapeutic antitumor effect of hEDA-E7_HPV16/18 + poly ICLC was evaluated in two HPV-induced tumor models: (i) flank s.c. engraftment of 1 × 10⁵ TC-1 P3 (A15) tumor cells, a variant of TC-1 cells with down-regulated MHC-I expression [47], a feature associated with ≈1/3 of HPV-induced cervical cancers [48], and (ii) orthotopic genital tumors established by intravaginal tumor challenge with 1.2–2.0 × 10⁴ TC-1 cells after pretreatment with 4% nonoxynol-9 [44]. In both tumor models, mice received two s.c. doses of the vaccine candidate at days 7 and 14 post-tumor transplantations. In the s.c. model, 100% of TC-1 P3 (A15) tumor-bearing mice showed complete tumor eradication (n = 8) and were still alive at day 100 post-tumor transplantation (Table 1). The same vaccination regimen resulted in a complete tumor regression in 100% of mice bearing orthotopic vaginal tumors (n = 5). These results show the strong potential of this adjuvanted protein approach in HPV immuno-oncotherapy, including against tumors with MHC-I down-regulation [44].

Table 1. Recapitulative results of antitumor efficacy of selected protein or live-attenuated bacterial vectors in the standard therapeutic TC-1 murine model.

Vector category	Vaccine platform	HPV antigen	Adjuvant/ Formulation	TC-1 cell number engrafted	Immunization regimen days (d) post tumor inoculation	Durable anti-tumor efficacy in treated mice	Ref
Protein vectors	Fibronectin extra domain A (EDA)	Full length E7HPV16/18	Poly ICLC	1 × 10^5 (TC-1 P3) s.c.	prime d7 s.c. boost d14 s.c.	100%	[44]
				1–2 × 10^4 intravaginal		100%
	B. pertussis Adenylate cyclase	Full length E7HPV16	Cyclophosphamide CpG-B + DOTAP	5 × 10^4 s.c.	single injection d25 i.v. single injection d30 i.v. single injection d40 i.v.	87% 42% 17%	[51]
	Shiga toxin A subunit (STxA)	E7HPV16: 43–57	α-galactosyl ceramide	5 × 10^4 tongue	prime d5 i.m. bootst d10 i.m. prime d5 i.n. bootst d10 i.n.	≈25% 100%	[59]
				1 × 10^5 lungs	prime d5 i.m. bootst d10 i.m. prime d5 i.n. bootst d10 i.n.	0% 40%
Live attenuated bacterial vector	Attenuated recombinant L. monocytogenes	E7HPV16	—	7 × 10^5 s.c.	prime d12 i.p. boost d19 i.p.	20%	[66,67]
			Anti-GITR mAb		prime d12 i.p. boost d19 i.p.	60%

3.2. Adenylate cyclase toxoid of Bordetella pertussis harboring E7_HPV16

Adenylate cyclase CyaA of Bordetella pertussis is a 1706 amino acid protein that can be engineered for use as a vaccine protein [49]. Detoxified CyaA binds to CD11b thus targeting a large variety of antigen-presenting cells [50]. CyaA delivers its N‐terminal catalytic domain to the cytosol of host antigen-presenting cells. Therefore, immunogenic regions of vaccinal interest inserted into the permissive sites of the CyaA catalytic domain readily gain access to the MHC-I presentation pathway of these host antigen-presenting cells [49]. Recombinant CyaA harboring intact full-length E7_HPV16 or the segments E7_HPV16:43–77 or E7_HPV16:49–57, encompassing an immunodominant H-2D^b-restricted T-cell epitope, were generated to evaluate the therapeutic potential of this vaccine platform in the TC-1 model [51]. These vectors triggered strong antigen-specific CD8⁺ T-cell responses after a single intravenous (i.v.) injection of 50 µg/mouse without adjuvant.

The antitumor efficacy was evaluated in C57BL/6 mice engrafted s.c. in the flank, with 5 × 10⁵ TC-1 cells following administration of a single dose of each CyaA vaccine on days 4, 7, or 11 post-tumor engraftments. Under these conditions, 94%, 80%, and 75% (n = 15–16/group) of the immunized groups showed complete TC-1 tumor eradication, respectively. However, the vaccine was not effective when applied later, i.e. on days 18, 25, or 30 post-tumor inoculation, as the tumors grew [51]. One of the obstacles decreasing the efficacy of immunotherapy in the treatment of late-stage tumors is the intratumoral presence of immunosuppressive cells, such as myeloid-derived suppressor cells (MDSC), regulatory T cells (Tregs), and fibroblasts [52–54]. Thus, there have been multiple attempts to combine vaccines with agents that counteract these cells. Combinations of CyaA:E7_HPV16 with (i) cyclophosphamide, which efficiently suppresses Tregs, or (ii) the TLR-9 agonist CpG-B formulated in the transfecting agent DOTAP (1,2-dioleoyl-3-trimethylammonium-propane) improved antitumor efficacy. Indeed, complete tumor eradication was achieved when such tritherapy was administered at day 25 or 30 post-tumor transplantation in 87% (14/16) and 42% (5/12) of mice, respectively, as recorded on day 50 post-tumor transplantation, and in a stable manner until day 150. However, if the tri-therapy was given on day 40 post-tumor transplantation, full tumor eradication occurred in only 17% (2/12) of the animals, followed by relapse [51] (Table 1).

In a phase II clinical trial, in women infected with HPV16 or HPV18 but with normal cytology, the use of a bivalent vaccine composed of CyaA carrying E7_HPV16 and E7_HPV18, administered via the intradermal (i.d.) route, was combined with topical application of imiquimod, a TLR7 and TLR8 agonist. This trial showed no efficacy in viral clearance over a two-year follow-up [49].

3.3. Shiga toxin conjugated to an E7_HPV16 T cell epitope

Shiga toxin (STx) is secreted by Shigella dysenteriae and enterohemorrhagic Escherichia coli [55]. Its B subunit (STxB) is a nontoxic 89 amino acid polypeptide that pentamerizes to interact with the glycolipid globotriaosyl ceramide of the host cell surface before internalization [56]. Tumor antigens fused to STxB can efficiently target the endogenous MHC-I presentation pathway in dendritic cells, resulting in efficient triggering of specific CD8⁺ T cells [57–59].

The E7_HPV16: 43–57 synthetic peptide was chemically coupled to STxB (STxB-E7) and administered to C57BL/6 mice intramuscularly (i.m.) or intranasally (i.n.) at a dose of 20 µg according to a prime-boost regimen [59]. For the prime, STxB-E7 was adjuvanted with 2 µg α-galactosylceramide. The latter is an immunostimulant that targets invariant NKT and TCR_γδ⁺ T cells, which, through their cytokine production, prime dendritic cells for T-cell activation [60]. E7_HPV16:49–57-specific responses, detected by an ex vivo tetramer assay and ELISPOT, were similarly induced in the spleens of mice immunized i.m. or i.n. By contrast, only i.n. administration of STxB-E7 induced mucosal E7_HPV16:49–57-specific CD8⁺ T cells in the mediastinal and cervical lymph nodes draining the pulmonary and head and neck tissues [59].

The antitumor effect of STxB-E7 was evaluated in mice after i.m. or i.n. immunization in three different tumor models: (i) s.c. engraftment of 1 × 10⁵ TC-1 cells in the flank, (ii) orthotopic implantation of 5 × 10⁴ TC-1 cells in the submucosal lining of the tongue, as a head and neck cancer model, and (iii) orthotopic engraftment of 1 × 10⁵ TC-1 cells in the lungs by intercostal injection at the median axillary line [59].

In the prophylactic setting, the antitumor effect was evaluated in the s.c. and orthotopic head and neck cancer models. In both models, mice were primed with 20 µg STxB-E7 i.m. or i.n. on day −21 and boosted via the homologous route on day −7 before tumor inoculation. Regardless of the route of immunization, no s.c. tumor was detected at day 40 and 100% of the mice were still alive on day 70 post-tumor inoculation. However, the route of immunization had an impact on STxB-E7 antitumor efficacy in the orthotopic head and neck cancer model. Indeed, 90% of the mice treated i.n. survived until day 40 post-tumor transplantation, whereas only 54% of the mice treated i.m. were still alive at this time point (n = 7/group) [59].

In the therapeutic setting, in the head and neck and lung cancer models, mice were primed i.m. or i.n. on day 5 and boosted via the homologous route i.m. or i.n. on day 10 post-tumor transplantation with 20 µg STxB-E7. In the head and neck cancer model, only i.n. STxB-E7 administration provided strong antitumor activity, with 100% of the mice alive until day 30 post-tumor transplantation vs only ≈25% treated via the i.m. route. In the lung tumor model, i.m. immunization did not provide protection, whereas 40% of the mice treated i.n. were still alive on day 30 post-tumor transplantation [59] (n = 7/group) (Table 1).

Therefore, STxB-E7 shows prophylactic antitumor efficacy when administered i.m. or i.n. against s.c. TC-1 tumors. In this study, no results were shown with the s.c. tumor model in the therapeutic setting, yet STxB-E7 showed remarkable therapeutic potential to induce mucosal antitumor activity when administered i.n. against established orthotopic head and neck or lung TC-1 tumors [59].

4. Attenuated recombinant Listeria monocytogenes expressing E7_HPV16

Live attenuated recombinant Listeria monocytogenes is a popular vector for immunotherapy [61]. The intracellular bacillus, L. monocytogenes enters mammalian cells either via its internalin A and B proteins or by phagocytosis. Once inside the phagosomes, L. monocytogenes can be killed by phagosome-lysosome fusion, generating a source of antigens for presentation by the MHC-II pathway. L. monocytogenes propagation is linked to the expression of the 529 amino-acid, nonhemolytic toxin listeriolysin O (LLO). The latter perforates phagosomal membranes to allow the bacilli to escape to the cytosol, while its antigens gain access to the MHC-I machinery [62]. There are several ways to attenuate L. monocytogenes strains, including deletion of internalin-encoding or metabolic genes, enabling its use as a vaccination platform [63–65].

Recombinant live attenuated L. monocytogenes secreting a fusion of LLO and E7_HPV16 (Lm-LLO-E7_HPV16) was generated [66,67]. C57BL/6 mice were engrafted s.c. with 7 × 10⁵ TC-1 cells on day 0. Lm-LLO-E7_HPV16 immunotherapy with 5 × 10⁶ colony forming units (CFU) of vaccine candidates was performed by an intraperitoneal (i.p.) prime on day 12, followed by an i.p. boost on day 19. Lm-LLO-E7_HPV16 treatment significantly inhibited tumor growth and resulted in complete tumor regression in 20% of the treated animals (n = 10) [67]. More recently, to improve the antitumor potential of Lm-LLO-E7_HPV16, the TC-1 tumor microenvironment was concomitantly modulated by co-administration of the agonist monoclonal antibody (mAb) DTA-1 on day 12 post-tumor inoculation. This antibody is specific to glucocorticoid-induced tumor necrosis factor receptor-related protein (GITR). Then, four more doses of this mAb were administered at 3- to 4-day intervals. After binding with its ligand on effector T cells, GITR activates cytokine production and cytotoxicity and enhances cell memory. GITR is also constitutively expressed by Tregs [68], and its stimulation can dampen Treg function, either by protecting effector T cells from the action of Tregs and/or by a direct negative impact on Treg activity [69]. Combined Lm-LLO-E7_HPV16 and anti-GITR mAb treatments successfully controlled tumor growth and led to complete tumor regression in 6 of 10 mice bearing s.c. TC-1 tumors [67] (Table 1).

5. HPV immunotherapies based on viral vectors

5.1. Adenoviral vectors encoding E6_HPV16/18 and/or E7_HPV16/18 antigens

Adenoviruses are non-enveloped icosahedral viruses with a linear, double-stranded DNA genome of 35 to 37 kb. Adenoviral vaccine vectors are replication-defective, and their genome can accommodate DNA fragments from 5 kb in first-generation to 36 kb in third-generation vectors [70,71]. The adenoviral genome remains as a non-integrated episome inside the nuclei of the host cells. Adenoviral vaccine vectors mainly transduce epithelial and endothelial cells [70]. Thus, the presentation of transgenic antigens delivered by adenoviral vectors to naïve T cells is indirect. This antigen cross presentation requires the uptake of primary infected cells or their debris by bystander dendritic cells. The main drawback of these vectors is their high seroprevalence, resulting in preexisting anti-vector immunity in human populations, as well as their pro-inflammatory features [72–76]. During the coronavirus disease 2019 (COVID-19) pandemic, the use of this vector also revealed side effects that necessitated temporary or definitive discontinuation of vaccination based on these vectors [77].

The low-prevalent adenovirus type 26 (Ad26) and Ad35 have been used in preclinical immunotherapy of HPV-induced cancer. These vectors were generated to encode fusions of E2_HPV16 or E2_HPV18 with detoxified shuffled fragments of either E6_HPV16 and E7_HPV16 or E6_HPV18 and E7_HPV18, under the control of the cytomegalovirus (CMV) immediate-early enhancer and promoter [78]. Immunization of H-2^b/d F1 mice by an i.m. prime with 5–10 × 10⁹ viral particles (vp) of Ad26:E2_HPV16-E6_HPV16-E7_HPV16 or Ad26:E2_HPV18-E6_HPV18-E7_HPV18, followed by a boost 4 weeks later with the same dose of Ad35 encoding the same antigens established the marked T-cell immunogenicity of the vectors at the systemic level [78] (Table 2).

Table 2. Recapitulative results of antitumor efficacy of selected viral vectors or nucleic acid vaccine candidates in the standard therapeutic TC-1 murine model.

Vector category	Vaccine platform	HPV antigen	Adjuvant/ Formulation	TC-1 cell number engrafted	Immunization regimen days (d) post tumor inoculation	Durable anti-tumor efficacy in treated mice	Ref
Viral vectors	Adenoviral vectors Ad26-Ad35	Detoxified, shuffled E2-E6-E7HPV16+ E2-E6-E7HPV18	—	5 × 10^4 s.c.	prime d4 i.m. boost d20 i.m. prime d6 i.m. boost d20 i.m.	30% 65%	[78]
	Adenoviral vector Ad5		—	1 × 10^6 s.c.	single injection d8 i.p.	10%	[79]
			Anti-GITR mAb	1 × 10^6 s.c.	single injection d8 i.p.	100%
	Poxviral vectors PCPV-MVA	Detoxified E7HPV16	—	5 × 10^5 s.c.	prime MVA d14 i.t. boost MVA d21 i.v. prime PCPV d14 i.t. boost MVA d21 i.v.	35% 58–80%	[82]
	Arenaviral vectors LCMV-PICV	Detoxified E6-E7HPV16	—	1 × 10^5 s.c.	prime LCMV d8 i.v. boost PICV d18 i.v.	37%	[84]
	Lentiviral vectors	E7HPV16	—	2 × 10^5 s.c.	single injection d i.m.	80%	[94]
		Detoxified E6-E7HPV16 E6-E7HPV18	—	1 × 10^6 s.c.	single injection d10 i.m.	100% 100% 100%	[96]
					single injection d13 i.m.
					single injection d19 i.m.
Nucleic acids	Naked DNA	Detoxified ubiquitin-E6-E7HPV16 and IgK-E6-E7HPV16	—	5 × 10^4 s.c.	prime d3 or d7 i.d. 1^st boost d10 or d14 i.d. 2^nd boost d17 or d21 i.d.	75%	[100]
	mRNA	E7HPV16	Lipid nanoparticle containing PEG	1 × 10^5 s.c.	conventional nucleosides modified nucleosides self-amplifying	42% 42% 42%	[104]

Antitumor efficacy was evaluated on the TC-1 murine model in H-2^b mice. Five × 10⁴ TC-1 cells were used for s.c. tumor engraftment. Mice were primed with Ad26:E2_HPV16-E6_HPV16-E7_HPV16 alone or with Ad26:E2_HPV18-E6_HPV18-E7_HPV18 as early as 6-day post-tumor transplantation, when the tumors were not yet palpable. Mice were then boosted using an Ad35 vector encoding the homologous antigens on day 20. The immunization regimen with Ad:E2_HPV18-E6_HPV18-E7_HPV18 alone did not lead to full tumor eradication and only delayed death from day 35 to day 55 post-tumor transplantation [78]. When the same immunization strategy with the mixture of Ad26 vectors was performed, only 30% survival was recorded on day 80 post-tumor transplantation (n = 12) [78]. Giving the prime early, on day 4 instead of day 6 post-tumor transplantation, significantly improved the antitumor effect with 65% survival recorded on day 80 post-tumor transplanatation (n = 11). It is plausible that the protection conferred by the Ad vectors encoding E6_HPV18-E7_HPV18 against TC-1 cells that only express E6_HPV16 and E7_HPV16 resulted from immune cross-recognition of the orthologous E6 and E7 antigens. The authors hypothesized that the effector cytokines produced by CD4⁺ T cells specific to E6_HPV18-E7_HPV18 could have had bystander helper activity for E6_HPV16- and E7_HPV16-specific cytotoxic CD8⁺ T cells [78] (Table 2).

In another adenoviral vector-based preclinical immunotherapy strategy, an Ad5 vector encoding detoxified and shuffled fragments of E6_HPV16, E7_HPV16, E6_HPV18, and E7_HPV18 was generated [79]. In a prophylactic setting, vaccination of C57BL/6 mice engrafted with 1 × 10⁶ TC-1 cells with a single i.p. injection of 1 × 10¹⁰ vp showed 100% antitumor efficacy in 10/10 mice. In a therapeutic setting, the vaccination was performed on day 8 post-tumor transplantation, when the tumor volume was 30–60 mm³. After the initial phase of TC-1 tumor regression, tumor recurrence occurred in 9/10 mice as soon as 20 days post vaccination [79]. Full eradication of the TC-1 tumors, without recurrence, in 100% of the animals required combining this adenoviral vector-based immunotherapy with systemic application of an anti-GITR agonist mAb, which resulted in higher CD8⁺ T-cell activation [79] (Table 2).

5.2. Poxvirus-based vaccination vectors encoding E6_HPV16 and/or E7_HPV16

Attenuated recombinant poxvirus vectors of the Orthopoxvirus genus are the most frequently used viral vectors in vaccine development [80]. They have a linear double-stranded DNA genome of 170 to 250 kb. Among these vectors, the replication-defective Modified vaccinia virus Ankara (MVA) has been deleted from six genomic regions, enabling their genome to accommodate large gene payloads of at least 10 kb that encode antigens of vaccine interest [81]. Although a single immunization with recombinant MVA vectors is immunogenic, most studies showed better immunogenicity when MVA was used as a booster [80]. Indeed, MVA primarily induces host cell apoptosis, a cell death mechanism considered to be less immunogenic than inflammation-associated programmed cell death, such as necroptosis. The efficacy of an MVA-based vaccine candidate encoding E6_HPV16, E7_HPV16, and IL-2 (TG4001) was evaluated, also in combination with anti-PDL1 mAb treatment, in HPV-induced cancer neoplasia. The clinical trials successfully demonstrated measurable effect on viral clearance and histological resolution, notably in patients without liver metastases [39–41].

More recently, a new-generation bovine pseudocowpox virus (PCPV) that better induces IFN-α secretion by host cells was tested to compensate for the low immunogenicity of MVA in prime vaccination [82]. This vector activates human monocytes and dendritic cells, degranulation by NK cells, and counteracts the effects of MDSC against T cells. The strong T-cell immunogenicity of a PCPV-based vector encoding detoxified E7_HPV16 has been established. In a therapeutic setting, C57BL/6 mice were engrafted s.c. with 5 × 10⁵ TC-1 cells and primed intratumorally (i.t.) with 5 × 10⁶ PFU of PCPV-E7_HPV16 on day 14 post-tumor transplantation, when the tumor diameter was ≈8 mm. Mice were then heterologously boosted i.v. on day 21 with 1 × 10⁶ PFU of MVA-E7_HPV16. Under these conditions, 58% to 80% of the treated mice showed complete tumor regression versus 35% for mice homologously primed and boosted with MVA-E7_HPV16 (n = 12/group). The antitumor protection was CD8⁺ T-cell dependent [82] (Table 2).

5.3. Arenaviral vector platform to induce immunity against E6_HPV16/18 and/or E7_HPV16/18

Arenaviruses are enveloped viruses of the Arenaviridae family and have two distinct, i.e. large and short, single-stranded RNA segments [83]. Their short RNA segment encodes the envelope glycoprotein and nucleoprotein. It is possible to develop replication-attenuated arenaviral vectors by duplicating the short RNA segment and segregating the genes encoding the glycoprotein and nucleoprotein. On each of the short RNA segments, one or the other viral gene can then be replaced by inserts of ≤2 kb encoding antigens of vaccine interest [84]. Arenaviral vectors have the advantages of globally low seroprevalence in human populations and of rarely inducing anti-vector neutralizing antibodies [84]. However, these vectors induce significant anti-vector T-cell immunity, which markedly dampens the boost effect in a homologous prime-boost regimen [84]. To avoid this problem and generate an optimal combination of arenaviral vectors for use in prime-boost immunization, a pair of two distantly related arenaviruses, the prototypic Old World virus, lymphocytic choriomeningitis virus (LCMV), and the well-documented New World clade A virus Pichinde virus (PICV), has been used [84]. Although LCMV is non-cytopathic, it induces robust innate and adaptive immune responses and is widely used in pre-clinical and clinical development as an anti-cancer agent or vaccinal vector [83].

The immunotherapeutic potential of anti-HPV arenaviral vectors has been evaluated in the TC-1 murine model using LCMV- and PICV-based vectors encoding detoxified E6_HPV16 and E7_HPV16 [84]. C57BL/6 mice were inoculated s.c. with 1 × 10⁵ TC-1 cells. On day 8 post-transplantation, when the tumor volume reached ≈100 mm³, mice were primed i.v. with 1 × 10⁵ plaque-forming units (PFU) of PICV-E6_HPV16-E7_HPV16, followed by an i.v. boost with the same amount of LCMV-E6_HPV16-E7_HPV16 on d18. The potential of arenaviral vectors to trigger T-cell activation/expansion relies on their ability to induce the secretion of IL-33 by splenic stromal cells then recognized by its receptor on CD8⁺ T cells [85]. Thus, the i.v. route was used in these experiments to favor direct vaccine delivery to the secondary lymphoid organs to optimize access to splenic stromal cells. Arenavirus vectors are also immunogenic when administered i.d., s.c, or i.m. In contrast to the control individuals, all eight primed and boosted mice showed TC-1 tumor regression until day 21 post-tumor transplantation. Full tumor eradication was observed in three of the mice, in which tumor cells did not proliferate following a tumor rechallenge on day 118, a sign of long-term memory. However, the other five mice experienced a relapse of the tumor from day 32 post-transplantation [84] (Table 2).

A first-in-human, phase I/II trial is in progress with a replication-competent live-attenuated LCMV-based vector encoding a fusion of detoxified E6_HPV16 and E7_HPV16. The vector has been administered combined with an anti-PD1 treatment, in HPV16⁺ patients with (i) recurrent/metastatic head and neck squamous cell carcinomas via i.v. or (ii) safely accessible tumor site for an i.t. prime followed by an i.v. boost with the same vector backbone [86].

5.4. Lentiviral vectors encoding E6_HPV16/18 and/or E7_HPV16/18 antigens

Lentiviral vectors are enveloped in single-stranded RNA viral particles derived from HIV-1 [87,88]. Their genomes are permissive to the insertion of ≈5-kb transgenes. These vectors show great potential for gene transfer to host cell nuclei. The absence of genotoxicity has been established for the integrative version of these vectors [89]. However, to improve safety, only the non-integrative version is used in vaccination and immuno-oncotherapy, which simply requires adjustment of the administered dose [87]. The catalytic site of the integrase of non-integrative lentiviral vectors is mutated. As a result, the viral retrotranscribed DNA does not interact with the host chromosome. The viral DNA remains in the host cell nucleus as non-integrating episomes [87,88]. Lentiviral vectors are non-replicative, self-inactivating, and non-cytopathic. Presumably, their lack of cytopathicity also explains their weak inflammatory capacity, but this does not prevent robust T-cell immunogenicity [90,91]. One of the main advantages of these vectors is the absence of preexisting anti-vector immunity in the human population, as they are predominantly pseudotyped with the envelope glycoprotein G of vesicular stomatitis virus (VSV), to which the human population has rarely been exposed [87,88]. Lentiviral vectors show tropism for immune cells, in particular dendritic cells in vivo and therefore deliver the transgene encoding the antigen of interest directly to these unique antigen-presenting cells able to trigger naïve T cells [92,93].

A non-integrative lentiviral vector encoding a fusion of non-oncogenic E7_HPV16 and calreticulin under the control of the CMV promoter was generated [94]. The non-oncogenic E7_HPV16 harbors the D21G, C24G, and E26G amino-acid substitutions at the Rb binding site, and calreticulin is known to enhance MHC-I antigen presentation [47,95]. Low levels of serum anti-E7 antibodies were detected in C57BL/6 mice immunized once with this lentiviral vector until 4 months post-immunization. Strongly reacting and long-lasting IFNγ⁺ T cells were detectable by ELISPOT against the immunodominant H2-D^b-restricted E7_HPV16:49–57 epitope until days 106 and 316 post-immunization. Intracellular cytokine staining detected E7_HPV16-specific IFNγ⁺/TNFα⁺ cells among CD8⁺ T splenocytes. To evaluate the antitumor effect, C57BL/6 mice were engrafted s.c. with 2 × 10⁵ TC-1 cells. A single i.m. injection of 1 × 10⁷ reverse transcriptase activity units (RTs) resulted in the absence of tumor proliferation in 4/5 (80%) mice with early-stage tumors when no measurable tumor mass was detectable, i.e. 2 weeks post-tumor transplantation. These mice remained tumor free, up to 11 months post-tumor transplantation [94]. Remarkably, in mice with established tumors 3–4 mm in diameter, a single i.m. injection of 1 × 10⁷ RT units resulted in full tumor eradication in 7/8 (87.5%) mice. Among the seven cured mice, four remained tumor free until day 40 post-vaccination and were sacrificed to study the T cells. All three of the cured mice that were conserved remained tumor-free for up to 90 days post-vaccination [94] (Table 2).

In a more recent study, a non-integrative lentiviral vector encoding detoxified forms of E6_HPV16, E7_HPV16, E6_HPV18, and E7_HPV18 was developed [96]. After a single i.m. injection of C57BL/6 mice with 1 × 10⁹ transduction units (TUs)/mouse of this vector, strong T-cell immunogenicity was observed in the spleens of the immunized mice against the immunodominant H2-D^b-restricted E7_HPV16:49–57 epitope, as well as against the subdominant regions of E6_HPV16 and E6_HPV18. A single i.m. administration of this vector into mice bearing established TC-1 tumors with a mean volume of 80–120 mm³ (small), 200 mm³ (medium), or 450 mm³ (large) resulted in complete tumor eradication in 100% of the mice (n = 7–8/group) [96]. A single i.m. administration of this lentiviral vector was also effective against TC-1 lung metastatic foci. The induced CD8⁺ T-cell immunity was long-lasting and prevented tumor relapse for up to 200 days post-vaccination. The memory precursor cells were still detectable in mice beyond 270 days post-treatment. The antitumor effect of this vector was accompanied by extensive remodeling of the tumor microenvironment, characterized by a marked reduction in the percentage of intratumoral Tregs, infiltrating activated effector CD8⁺ T cells and resident memory T cells (Trm), and transcription factor T-cell factor-1 (TCF-1)⁺ stem-like CD8⁺ T cells. TCF-1 contributes to reinforce responses to checkpoint blockade immunotherapy and is now considered as a potential biomarker for the clinical prognosis of immunotherapy [97,98]. A suboptimal dose of 1 × 10⁸ TUs/mouse of this lentiviral vector was shown to synergize with an anti-PD1 mAb in TC-1 cell immunotherapy [96] (Table 2).

6. Nucleic acid-based vaccines

6.1. Naked DNA encoding the E6-E7_HPV16 antigen

DNA vaccines involve genetically engineered plasmids that contain DNA sequences encoding antigens of interest. The introduction of DNA vaccines either by i.m. or i.d. delivery results in antigen production by the cells and induces consistent immunity in the host [99]. To investigate the immunotherapeutic potential of DNA vaccination against HPV-induced tumors, a plasmid vector was developed to express a detoxified E6_HPV16-E7_HPV16 fusion protein [100]. The human ubiquitin gene was inserted into the frame at the 5’ end of the E6_HPV16-E7_HPV16 fusion to enhance the immunogenicity of the plasmid vector, giving rise to a ubiquitin-E6 _HPV16-E7_HPV16 fusion protein [101]. Another version of the plasmid vector able to induce secretion of the E6_HPV16-E7_HPV16 fusion antigen was also constructed by incorporating the signal peptide of the murine Igϰ secretory sequence at the 5’ end. The DNA vaccine consisted of a 1:1 mixture of secreted and ubiquitinated constructs encoding the fusion protein.

Mice were immunized i.d. in the pinna of each ear with 30 µg of the vaccine twice, 3 weeks apart. Five weeks post-immunization, E6_HPV16- and E7_HPV16-specific T-cell responses were readily detected, which could be recalled after 5 months. In addition, an E7_HPV16-specific antibody response was observed 5 weeks post-immunization. To assess the potential of the DNA vaccine in tumor prevention, mice were immunized twice at a three-week interval and engrafted with 5 × 10⁵ TC-1 cells 1 week after the second vaccination. Such prophylactic immunization fully prevented tumor growth. In the therapeutic setting, the antitumor effect of the vaccine was evaluated in mice inoculated with 5 × 10⁴ TC-1 cells followed by administration of the DNA vaccine three times at weekly intervals either 3 or 7 days following TC-1 cell inoculation. Immunization significantly suppressed the progression of established TC-1 tumors, demonstrating the therapeutic potential of the DNA vaccine (Table 2).

In combination therapy, mice were first inoculated with 1 × 10⁵ TC-1 cells, followed 7 days later by administration of the DNA vaccine three times at weekly intervals and 200 µg of an anti-PDL-1 mAb by i.p. injection twice a week for 3 weeks. Although targeting PDL-1 alone had no effect on tumor growth and animal survival, the combination therapy enhanced antitumor immunity, which resulted in the survival of 6/8 mice up to 60 days and the complete elimination of tumors in 3/8 mice, i.e. 37% of the treated mice (Table 2).

6.2. mRNA vaccine candidates encoding the E7_HPV16 antigen

Despite their short-term effectiveness, mRNA-based COVID-19 vaccines have been administered to provide wide vaccination coverage, mainly due to their ease of production. Since then, this technology has been further developed in anticipation of generating a large number of prophylactic and therapeutic vaccines [102]. In this approach, mRNA encoding the antigen of vaccine interest is encapsulated in lipid nanoparticles conjugated with the excipient polyethylene glycol (PEG). PEG generates a repulsive force, which is essential for avoiding the aggregation of nanoparticles, but it can also cause hypersensitivity and allergic reactions [103]. The immuno-therapeutic potential of this technology was recently evaluated for the treatment of HPV-induced tumors in the murine TC-1 model [104]. The antigen encoded by the mRNA was a chimeric protein composed of the type 1 herpes simplex virus (HSV-1) glycoprotein D (gD) fused to E7_HPV16.

Three mRNA vaccine platforms were tested. The first consisted of non-replicating mRNA, composed of conventional nucleosides, which are sensed by various receptors of the innate immune system, i.e. toll-like receptor (TLR)3, TLR7, and TLR8, retinoic-acid-inducible gene I (RIG-I), and melanoma differentiation-associated gene 5 (MDA5) [105]. This leads to the expression of proinflammatory cytokines, which may help T-cell activation but also significantly increases the risk of serious adverse events. The second consisted of non-replicating mRNA, containing N1-methyl-pseudo-uridine-5’-triphosphate instead of uridine-5′-triphosphate to generate the nucleoside, which avoids recognition by innate immune sensors [106]. The third consisted of self-amplifying mRNA, also made of modified nucleosides, using alphavirus-derived sequences that encode proteins that enable mRNA replication to maximize and prolong antigen availability [107,108]. The self-amplifying mRNA modality prolonged the expression of a reporter antigen in vivo up to 23 days vs 6 days for the non-replicating mRNA vaccine versions. The T-cell immunogenicity of the vaccine candidates based on nucleoside-modified or self-amplifying mRNAs was significant and comparable [104].

Antitumor efficacy has been tested using C57BL/6 mice engrafted with 1 × 10⁵ TC-1 cells vs 1 × 10⁶ cells, generally used in the standard model. The i.m. injection of each mRNA vaccine modality was carried out on day 3 post-tumor inoculation when no tumor was yet palpable. The tumors become palpable only from day 12 to day 15 post-transplantation. Under these conditions, particularly favorable for demonstrating anti-tumor efficacy, CD8⁺ T cell-mediated immunological control of tumor growth was observed for the three mRNA vaccine versions, but to a lesser extent for the non-replicative version containing conventional nucleosides [104]. In mice injected intravaginally or grafted in the lining of the tongue with TC-1 cells, i.m. immunotherapy with the non-replicating conventional mRNA or self-amplifying nucleoside-modified mRNA allowed full tumor eradication until the end of the experiment, i.e. day 30 post-transplantation [104].

In another experiment, mice bearing palpable s.c. tumors were vaccinated with escalating doses of 0.1, 1, and 5 µg/mouse of the three mRNA vaccine modalities on day 12 post-tumor transplantation, when the average tumor area was ≈12 mm². In this setting, complete tumor eradication was observed for 7/7 mice on day 20 post-tumor transplantation for the three mRNA vaccine modalities. However, the tumor relapsed for three-sevenths individuals from day 30 post-transplantation, regardless of the mRNA modality, even at the highest dose tested. Notably, no improved antitumor efficacy was observed with the self-amplifying version of the mRNA vaccine [104] (Table 2). The partial success against established tumors in the s.c. TC-1 model shows that the T-cell immunity induced by mRNA vaccines was not sufficiently robust or durable to prevent the rapid resumption of solid tumor growth.

7. Conclusion

Here, we overview the most significant approaches and studies in preclinical immunotherapies of HPV-induced cancers published over the last decade by comparing the: (i) HPV antigen design, (ii) immunization platform and regimen, (iii) efficacy of T-cell immunogenicity, and (iv) magnitude and duration of the antitumor effect (Tables 1 and 2). We intentionally omitted the adjuvanted peptide vaccination strategy because of the short in-vivo lifespan of small synthetic peptides and because many more powerful technologies, some of them discussed here, have the ability to induce longer-lasting, higher-quality cellular immune responses.

The therapeutic strategies discussed in this review were based on adjuvanted protein vectors, a live attenuated bacterial vector, several viral vectors, and DNA and mRNA platforms. To enable a direct comparison of the results from various expert laboratories, we only considered the HPV-induced TC-1 cell murine model, which is the most highly standardized and most widely used. These therapeutic preclinical vaccination strategies trigger notable T-cell responses. Among the strategies studied, except for the adjuvanted extra domain A of human fibronectin (hEDA) (Table 1), viral vectors showed the strongest antitumoral effects, accompanied by their ability to induce strong CD8⁺ T cytotoxic responses. Among the viral vectors, lentiviral vectors were the only ones to lead to total eradication of solid s.c. TC-1 tumors in almost all experimental individuals, even those with large tumors, accompanied by a long-lasting memory effect, as established by two independent laboratories (Table 2).

8. Expert opinion

As mentioned in the introduction, even though HPVs have developed multiple distinct mechanisms to evade the host immune system and render HPV-induced malignant tissues invisible to the immune system, the results of several recent clinical trials using an MVA-based immunotherapy showed measurable and encouraging antitumor efficacy against HPV-induced anogenital cancers [39–41] (https://www.clinicaltrials.gov/study/NCT03260023). It is, therefore, important to persevere in the development of new immunotherapeutic strategies, in particular to treat head and neck as well as oropharyngeal cancers, which are undergoing a steady epidemiological rise [10].

In view of the results summarized here, although several protein vaccine candidates appear to be good options for immunotherapy against HPV-induced tumors and despite the reasonable safety of recombinant protein antigens, they require the use of adjuvants to improve the magnitude of adaptive immune response and provide maximum anti-tumor and protective effects. In this context, the evidence of adjuvant-induced autoimmune/inflammatory syndrome (ASIA), observed in both preclinical animal models and humans, should be taken into account [109,110]. Viral vectors, for which the inherent constituents are sufficient to stimulate the innate immunity required to induce appropriate adaptive cellular immunity, represent a more amenable alternative for immunotherapeutic applications. However, viral vectors that are highly pro-inflammatory or targets of preexisting immunity in humans, such as adenoviral vectors, should be avoided, even though they offer certain advantages in terms of production. Although low-incidence subtypes of adenoviral vectors have been used, for example in COVID-19 vaccination, it became clear during the pandemic that they show serious safety limitations in humans. In parallel, replication competent viral vectors, including arenaviral vectors and those based on measles virus, are often highly immunogenic, but may carry the unavoidable risk of recombination and reversion.

This review provides evidence that lentiviral vectors show superior preclinical antitumor efficacy in the standard TC-1 murine model than adenoviral, poxviral, and arenaviruax vectors. Importantly, lentiviral vectors are less inflammatory than other viral vectors, as they are not cytopathic. Moreover, they are not targets of preexisting immunity as a result of their pseudotyping by the envelope glycoprotein of VSV, to which human populations have had very little exposure. It should be noted that large-scale techniques for the production of lentiviral vectors have evolved considerably, making it now possible to produce them in quantities sufficient for immuno-oncotherapy, even if they are not yet easily producible in amounts necessary for prophylactic mass vaccination against infectious diseases. The safety of lentiviral vectors in the immunotherapy of sarcomas and other solid tumors expressing New York esophageal squamous cell carcinoma-1 (NY-ESO-1), administered i.d., has been established [111]. In a first-in-class, first-in-human study, the safety profile, immunogenicity, and potential clinical activity in patients have clearly recorded. Additional clinical trials are likely to reinforce the safety and non-genotoxicity of these vectors and render them realistically implementable in clinical practice in the coming decade. However, due to the relative ease of their industrial production, mRNA-based vaccination/immunotherapies are becoming increasingly popular. The comparative study of antitumor efficacy carried out in this review highlights the low duration of immunity conferred by the mRNA vaccination strategy, including the mRNA modality that uses a self-amplifying genetic message, which significantly extends the duration of antigen expression in the host organism.

Article highlights

• Persistent infections with the human papilloma virus HPV16 and HPV18 genotypes can cause multiple cancers.

• Prophylactic anti-HPV vaccines show no efficacy against persistent HPV infections or already malignant tissues.

• No immunotherapy against HPV-induced cancers has been thus far approved for use in humans.

• Active research is ongoing on immunotherapy of HPV-induced malignancies.

• We compared the efficacy of the immunotherapy strategies developed against HPV-induced cancers in the standard murine TC-1 tumor model since 2005.

• Certain adjuvanted proteins and viral vectors induce the strongest effects against HPV-induced tumors.

• Lentiviral vectors, able to induce the longest-lasting T-cell immune memory, give rise to full eradication of large solid tumors in 100% of mice.

Declaration of interest

P. Charneau is the founder and CSO of TheraVectys. A. Demidova, L. Douguet, and I. Fert are employees of TheraVectys. L. Majlessi has a consultancy activity for TheraVectys. L.D., I.F., P.C., and L.M. are inventors of a pending patent directed to the potential of a lentiviral vaccination vector against HPV-induced cancers. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Acknowledgments

The authors would like to thank Pierre Authié for his critical reading of the manuscript.

Additional information

Funding

This paper received funding from Institut Pasteur and TheraVectys.