Exploiting Preexisting Immunity to Enhance Oncolytic Cancer Immunotherapy

Prophylactic vaccinations are among the most effective forms of medical interventions with direct clinical and health economic benefits, with the eradication of common deadly infectious diseases being the most obvious example (1). Most vaccines rely on the use of attenuated pathogens or parts of them and evoke a robust T-cell memory repertoire directed against the pathogen, with the CD4⁺ T cells dominating the memory response (2, 3). Following immunization, naïve T cells differentiate into T effector memory (T_EM) cells that are rapidly re-called by encountering the antigen, and into T central memory (T_CM) cells that are found mainly in lymphoid organs and are not immediately triggered in response to pathogens (4). A strong repertoire of memory T cells against pathogens included in the national vaccine programs exists in the worldwide population (5).

The efficacy of cancer immunotherapy relies on the generation of specific antitumor CD8⁺ T cells that recognize peptides presented on the MHC-I (6). Effective antitumor activity requires fast T-cell mediated responses, which is highlighted for example by clinical success with the chimeric antigen receptor (CAR) T cells targeting CD19 in B-cell malignancies (7). Importantly, it has been shown that the cooperation of CD4⁺ and CD8⁺ T cells is required for efficient antitumor immunity to occur (8). Indeed, CD4⁺ T cells provide signals that improve the functionality of CD8⁺ T cells within the tumor microenvironment (TME; ref. 9) and their depletion prior to tumor challenge results in complete loss of tumor rejection in murine tumor models (10).

Although the central role of CD4⁺ T cells in T-cell mediated immunity is well recognized, it is still unclear how to optimally utilize the interplay between CD4⁺ and CD8⁺ T-cell populations in cancer treatment strategies (8). To this end, our aim was to investigate how to exploit the pathogen-specific T-cell memory reservoir, mainly CD4⁺ T cells, to strengthen the antitumor CD8⁺ CTL response.

We utilized our PeptiCRAd platform that is based on oncolytic adenovirus coated with MHC-specific peptides (11) to evaluate the effect of reengagement of pathogen-specific CD4⁺ memory T cells on antitumor CD8⁺ T-cell responses in mice preimmunized with vaccines specific for human pathogens. Our hypothesis was that antigen-presenting cells (APC) would process the virus and tumor- and pathogen-specific peptides linked to its surface and present the tumor-specific epitopes to CD8⁺ T cells and the pathogen-specific epitopes to memory CD4⁺ T cells that would then sustain the CD8⁺ T cell–mediated immune response as a bystander effect (12).

We investigated the feasibility of our approach in naïve or tetanus-preimmunized immunocompetent mice engrafted with B16.OVA tumors. Mice were treated with intratumoral injections of PeptiCRAd coated with SIINFEKL (CD8⁺ T cell epitope of chicken ovalbumin) and tetanus toxoid (CD4⁺ T-cell epitope) peptides. As hypothesized, we observed a superior antitumor response in mice preimmunized with the tetanus vaccine and treated with TT-OVA-PeptiCRAd. Interestingly, in naïve mice, the superiority of TT-OVA-PeptiCRAd over control treatments was lost, highlighting the prerequisite of the preexisting immunity to exploit the CD4⁺ T memory. We validated this strategy by targeting different pathogens (Diphtheria and Pertussis), by targeting different tumor antigens (both tumor-associated antigens and tumor neoantigens) and by combining them with a checkpoint inhibitor treatment (anti–PD-1 antibody). Consistent with our previous results, engagement of CD4⁺ T cells by Diphtheria-Pertussis–specific MHC-II–restricted peptides resulted in a slower tumor growth in preimmunized mice. In addition, a more robust effector memory CD4⁺ T-cell infiltration was observed in the TME of treated animals when compared with control animals. This response correlated with the level of CD8⁺ antigen-specific tumor-infiltrating lymphocytes (TIL) and the level of tumor growth control. These results indicated that the proposed mechanism of action is not restricted to tetanus, but the principle could be applied to other vaccine formulations.

Thus, our data suggest that the preexisting CD4⁺ memory T-cell repertoire can be exploited to support the antitumor CTL response; moreover, our findings contribute to the knowledge on how to generate an optimal T-cell response against tumors, which is the key for the next major improvement in cancer immunotherapy.

The main goal of this study is revoking the CD4⁺ T-cell antipathogen memory repertoire to boost the antitumor response. First, our hypothesis was verified in tetanus immunized B16.OVA bearing mice compared with the naïve. To demonstrate that the use of the memory repertoire gave an advantage over the naïve, the mice immunologic background was examined. Subsequently, we validated our hypothesis using a clinically relevant tumor peptide in combination with immune checkpoint inhibitor. Finally, the experiment was repeated with a different type of vaccine, offering a conceptual framework. The control and treatments groups are specified in the figure legends. Animal number for each study type was determined by the investigators (each treatment group had not less of n = 8 mice). Animals were randomly allocated to the control and the treatment groups.

The cell line B16.OVA, a mouse melanoma cell line expressing chicken ovalbumin (OVA), was kindly provided by Professor Richard Vile (Mayo Clinic, Rochester, MN) and cultured according to ATCC recommendations. The cells were cultured in RPMI1640 with low glucose and supplemented with 10% FBS, 1% antibiotics, 1% l-glutamine, and 10% geneticin (G418). The cells were cultivated in 37°C, 5% CO₂ in a humidified atmosphere. The cell line JAWS II were purchased from ATCC and cultured in α-MEM supplemented with 20% FBS, 1% antibiotics, 1% l-glutamine in presence of 5 ng/mL of GM-CSF.

All cell lines were cultured under appropriate conditions and were routinely tested for Mycoplasma contamination.

The following peptides purchased from Ontores Biotechnologies Co. Ltd. were used throughout the study:

KKKKKKSIINFEKL (OVA), KKKKKKSVYDFFVWL (TRP2), KKKKKQYIKANSKFIGITEL (Tetanus toxin), KKKKARYVSQQTRANPNPY (Pertussis), KKKKIQSKRFAPLYAVEAK (Polio Mahoney), KKKKKKSPVYVGNGVHANLHV (Diphtheria), KKKKKKPVFAGANYAAWAVNVAQVI (Diphtheria), ARYVSQQTRANPNPY (Pertussis), IQSKRFAPLYAVEAK (Polio Mahoney), SPVYVGNGVHANLHV (Diphtheria), KKKKKKLCPGNKYEM (B16M27.2).

The peptide B16M27.2 was generated from the MHC-I–restricted tumor neoantigen B16M27 (REGVELCPGNKYEMRRHGTTHSLVIHD; ref. 13); the IEDB algorithm tool was used to predict the exact epitope that binds H2Db and contains the single point mutation.

For tetanus and diphtheria-tetanus-polio-pertussis vaccination, 4- to 6-week-old female C57BL/6 mice received a primary intramuscular vaccination of Anatetall (GlaxoSmithKline: 8 IU in 100 μL) or PolioBoostrix (GlaxoSmithKline: Diphtheria Toxoid 0.4 IU in 100 μL, Tetanus Toxoid 4 IU in 100 μL, Bordetella pertussis antigens: Pertussis Toxoid 1.6 mg in 100 μL, Hemagglutinin 1.6 mg in 100 μL and Pertactin 0.5 mg in 100 μL), respectively, administered bilaterally into the quadricep muscle (50 μL per leg). An intramuscular booster vaccination (50 μL) was administered twice at an interval of two weeks. Mouse IgG antibody responses to tetanus toxoid and diphtheria were measured by ELISA (Xpress Bio). Serum from immunized mice was harvested 5 days after the last immunization and prior to the animal experiment.

Oncolytic adenovirus and each epitope with a polyK tail (Ontores) were mixed to prepare the PeptiCRAd complex. We mixed polyK-extended epitopes with Ad-5-D24-CpG for 15 minutes at room temperature prior to treatments with the PeptiCRAd complexes. More details about the stability and formation of the complex can be found in our previous study (11).

All animal experiments were reviewed and approved by the Experimental Animal Committee of the University of Helsinki and the Provincial Government of Southern Finland (license number ESAVI/9817/04.10.07/2016).

Four- to 6-week-old female C57BL/6JOlaHsd mice were obtained from Envigo Laboratory. A total of 3 × 10⁵ B16.OVA cells were injected subcutaneously into the right flank. Details about the schedule of the treatment can be found in the figure legends. Viral dose was 1 × 10⁹ vp/tumor complexed with 20 μg of a single peptide or with 10 μg + 10 μg mixture of two peptides. Intratumorally administrated Anatetall vaccine was dosed at 2 IU per mouse. Checkpoint inhibitors were given intraperitoneally at a dose of 100 μg/mouse.

Splenocytes from tumor-bearing mice treated with different treatments regiments were harvested at the end of the experiment and processed into single-cell suspensions, after which, the splenocytes were cocultured with TT peptide (QYIKANSKFIGITEL)-pulsed JAWSII cells for 6 hours in the presence of Brefeldin A (eBioscience). After restimulation, the cells were washed, fixed, and stained and the data were acquired using BDLSRFORTESSA flow cytometer.

The antibodies used are the following: TruStain Fcblock and anti-CD8-FITC (eBioscience), Affymetrix (Thermo Fisher Scientific), Foxp3-PE (eBioscience), CD4-PeCy7 (eBioscience), CD3-PerCPCy5.5 (eBioscience), IFNg-APC (eBioscience), CD40L-BV650 (BD Biosciences Bel Art Scienceware (Thermo Fisher Scientific), IFNg-FITC (BD), IL17A-PE (BD), CD4-PerCPCy5.5 (BD), IL4-APC (BD), CD44-V450 (BD), CD44-PE (eBioscience), CD4-PeCy7 (eBioscience), CD3-PerCPCy5.5 (eBioscience), CD62L-APC (eBioscience), CCR7-V450 (BD), CD11c-FITC (BD), B220-PE (eBioscience), MHC-II(A-I/E-I)-PeCy7 (eBioscience), CD86-V450 (BD), CD40-APC (eBioscience), CD11b-PerCP-Cy5.5F4/80BV650 (BD), H-2Kb SVYDFFVWL-APC (ProImmune), CD8a-FITC (ProImmune). The data were acquired using BDLSRFORTESSA flow cytometer.

The data were analyzed calculating the ratio between the percentage of cells and the tumor volume.

Data were analyzed using FlowJo software v9.

IFNγ ELISPOT assays were performed using a commercially available mouse ELISPOT reagent set (ImmunoSpot) and 20 ng/μL of each peptide was tested in in vitro stimulations of splenocytes at 37°C for 72 hours. Spots were counted using an ELISPOT reader system (ImmunoSpot).

All data associated with this study are present in the article or in Supplementary Materials. These data are available by request from Drug Research Program ImmunoViroTherapy lab, University of Helsinki (Helsinki, Finland).

Statistical analysis was performed using Graphpad Prism 6.0 software (Graphpad Software Inc.). For animal experiment, two-way ANOVA with Tukey multiple comparisons test was used and P < 0.05 was considered statistically significant. All results are expressed as the mean ± SEM. Details about the statistical tests for each experiment can be found in the corresponding figure legend.

We assessed the potential of engaging the CD4⁺ T-cell memory to the concept of the PeptiCRAd vaccine platform (11, 14, 15) where we coated an oncolytic adenovirus with both MHC-I–restricted tumor-specific peptides and MHC-II–restricted pathogen-specific peptides, and studied the effect in tumor-bearing mice preimmunized for the pathogen (Fig. 1A). Our hypothesis was that by adding the MHC-II–restricted pathogen-specific peptides to the PeptiCRAd platform, we would provide a swifter and stronger T helper response, enhancing the tumor-specific CTL response.

First, we investigated the antitumor effect of PeptiCRAd in mice preimmunized with tetanus vaccine intramuscularly and bearing B16.OVA tumors, a melanoma model expressing chicken OVA as a model antigen (16). The OVA-epitope was selected because it has a high immunogenicity and hence provides a suitable model to analyze the generation of T-cell response (17). C57BL/6 mice were immunized with tetanus vaccine three times at 2-week intervals (Fig. 1B). Five weeks after the priming, serum samples were collected from mice and anti-tetanus antibody titer was measured to confirm the success of the vaccination (Fig. 1B; Supplementary Fig. S1A).

After tumor engraftment, mice were randomized and treated with PeptiCRAd coated with tumor-specific peptides (OVA-PeptiCRAd), tetanus-specific peptides (TT-PeptiCRAd), or both tetanus and OVA peptides (TT-OVA-PeptiCRAd). In addition, tetanus vaccine alone or in combination with OVA-PeptiCRAd was used to assess whether intratumorally administrated commercial vaccine can affect tumor growth. All treatments were all delivered by intratumoral administration according to the regimen depicted in the Fig. 1B.

Following therapy, TT-OVA-PeptiCRAd was superior to either one of the single coated viruses in controlling the tumor growth in mice preimmunized with tetanus toxoid vaccine (Fig. 1C), suggesting that the anti-tetanus memory response indeed enhances the primary immune response elicited against the OVA antigen. The ability of TT-coated PeptiCRAd to elicit mainly Th1-polarized CD4⁺ T-cell responses was further corroborated by intracellular staining (Supplementary Fig. S1B). Less surprisingly, the approach worked also when the tetanus vaccine was reintroduced as a combination with OVA-PeptiCRAd (vaccine + OVA-PeptiCRAd), whereas tetanus vaccine alone had no therapeutic efficacy (Fig. 1D). Notably, when comparing vaccine + OVA-PeptiCRAd to OVA-PeptiCRAd, the latter showed a significantly higher antitumor efficacy (P = 0.05). This suggests that the effect was not caused by the adjuvant contained in the vaccine itself but rather by the presentation of tetanus-specific peptides on MHC-II, engaging CD4⁺ T cells to help the cytotoxic CD8⁺ T-cell response.

Interestingly, when the same experiment was performed in naïve mice (mice that had not been preimmunized with tetanus vaccine), no statistically significant differences were observed between OVA-PeptiCRAd and TT-OVA-PeptiCRAd (Fig. 1E).

These results demonstrate that the antitumor efficacy of our virus-based PeptiCRAd cancer nanovaccine is significantly enhanced if it is simultaneously coated also with peptides that are specific for a pathogen for which a preexisting immunity exists.

To gain a deeper understanding of the mode of action of the double-coated PeptiCRAd, we wanted to investigate the quality of the immune response elicited by the different treatments. To this end, we analyzed the frequency of different cell populations in the tumor by flow cytometry, most importantly the activated dendritic cells (DC), CD4⁺ and CD8⁺ T cells with effector and memory phenotype and experienced and exhausted CD8⁺ effector T cells.

Interestingly, we found increased frequency of both total and activated intratumoral DCs in all of the groups that had been treated with PeptiCRAd in the context of tetanus antigens (either coated with the TT peptide or coinjected with the whole vaccine; Fig. 2A; Supplementary Fig. S2). In contrast to these combination treatments, the use of vaccine alone led to poor induction of DC maturation in the TME, suggesting that inclusion of an adenoviral adjuvant may be critical for a proper DC activation in this setting. Moreover, we saw increased levels of CD4⁺ and CD8⁺ T cells in the tumors in all groups of mice treated with PeptiCRAd (Fig. 2B and C), which is well in line with what has previously been observed following treatments with virus-based drugs (11, 14, 15, 18). Finally, we wanted to analyze the phenotype of these T cells. Majority of the TILs showed a T effector memory cell phenotype, with an increase in the frequency of CD8⁺ and CD4⁺ T_EMs in groups treated with TT-PeptiCRAd and OVA-TT-PeptiCRAd (Fig. 2D and E). Moreover, the expression level of T-cell immunoglobulin and mucin-domain containing-3 (TIM3) on PD-1⁺ TILs were assessed to study T-cell exhaustion (19–21). Interestingly, we observed a significantly lower frequency of exhausted CD8⁺ T cells in the group of mice treated with TT-OVA-PeptiCRAd compared with the other groups, indicating that CD4⁺ T-cell help is required for optimal CD8⁺ T-cell activity (Fig. 2F). We concluded that the tetanus preexisting immunity improved the overall efficacy of the treatment substantially by modifying the immune environment at the tumor site, especially when the treatment was virus based and contained the tetanus vaccine or the tetanus peptides. Of note, the serotype 5 human adenovirus used in these experiments is non-oncolytic in murine tumors, and therefore the effect on tumor control is solely based on antitumor immune response. To better elucidate this phenomenon, we reanalyzed all the datasets by stratifying the mice between responders and non-responders and assessed again their immunologic responses. As expected, we observed a significant difference between the two groups. Irrespective of the type of therapy, all responders had an ongoing measurable immune response, highlighting the importance of the immune system in controlling the tumor growth, regardless of what kind of treatment they had received. Importantly, the majority of these responders were found in the group of mice treated with TT-OVA-PeptiCRAd (Figs. 2A–F and 3A–E).

To dissect the possible mechanism of the observed therapeutic efficacy, we assessed levels and phenotype of immune cells in secondary lymphoid organs of preimmunized mice. As expected, PeptiCRAd-treated mice showed expansion of CD4⁺ T-cell compartment both in the spleen and in the draining lymph nodes (Fig. 4A; Supplementary Fig. S3A). More importantly, a significant increase of TT-specific CD4⁺ T cells expressing CD40 ligand (CD40L) was observed in TT-OVA–treated mice (Fig. 4B; Supplementary Fig. S3B). Majority of these CD40L⁺ cells were polarized toward Th1 phenotype, albeit some TT-specific Foxp3⁺ T regulatory cells (Treg) were also detected (Supplementary Fig. S3C and S3D). Analysis of dLNs revealed that the intratumoral vaccination with TT-OVA-PeptiCRAd induced mainly IFNγ producing Th1 memory cells in the expense of IL4 secreting Th2 cells, whereas no differences was observed in IL17A producing Th17 cells (Supplementary Fig. S3E–S3H). Because CD4⁺ T cell–associated CD40L has been shown to be important in stimulating cytotoxic CD8⁺ T-cell responses (22, 23), we wanted to study whether we can see CD40⁺ antigen-presenting cells. Indeed, when preimmunized mice were intratumorally treated with TT-OVA-PeptiCRAd, a significantly higher frequency of CD40⁺ expressing APCs was detected (Fig. 4C), further suggesting that double-coated PeptiCRAd stimulates TT-specific CD4⁺ memory T cells, that in turn could license professional APCs via CD40–CD40L interaction.

We have previously shown that combination of tumor-targeted PeptiCRAd with immune checkpoint inhibitors is synergistic in terms of improved antitumor efficacy (8). Thus, we wanted to assess whether the vaccine-induced preexisting immunity would further enhance this synergy, particularly by increasing the frequency of mice responding to the therapy.

To test this hypothesis, we coated the virus with TT and tyrosinase related protein 2 (TRP2) peptides (TRP2180-188; ref. 24), which is naturally occurring melanoma-associated antigen and hence more clinically relevant epitope than OVA. Tetanus toxoid preimmunized mice were implanted with subcutaneous tumors and treated intratumorally with a PeptiCRAd coated with TRP2 peptides only (TRP2-PeptiCRAd) or with a PeptiCRAd coated with both TRP2 and TT peptides (TT-TRP2-PeptiCRAd; Fig. 5A). Similarly, as in Fig. 1, we observed a significant inhibition of tumor growth in mice treated with the double-coated virus compared either to mock or to TRP2-PeptiCRAd groups (Fig. 5B).

Interestingly, when we combined the PeptiCRAd treatments with a PD-1 blocking mAb, we observed a significant increase in efficacy of both TRP2-PeptiCRAd and TT-TRP2-PeptiCRAd treatments (Fig. 5C and D). However, the double-coated PeptiCRAd was still more effective than the virus coated with a single peptide in terms of tumor growth control (Supplementary Fig. S4A).

More importantly, inclusion of TT-specific peptides in the cancer nanovaccine resulted in 75% response rate to anti–PD-1, whereas only 28% of mice treated with TRP2-PeptiCRAd and PD-1 blockade experienced a complete tumor eradication (Fig. 5E; Supplementary Fig. S4B). One of the biggest advantages of combining oncolytic viruses with checkpoint inhibitors is that the viruses in the tumor facilitate and increase the T lymphocyte recruitment, thereby unleashing an unprecedented activity of the monoclonal antibodies (25). Along this line, we observed a significant increase in total and TRP-2–specific CD8⁺ TILs in mice treated with TT-TRP2-PeptiCRAd, when compared with the control treatments (Fig. 5F and G).

To further validate the model in a more therapeutically relevant setting, we used a previously published, MHC-I restricted neoantigen B16-M27 as the tumor target (13) and double-coating PeptiCRAd with the predicted B16-M27 short peptide together with the tetanus toxoid peptide. Following intratumoral treatment, significant tumor growth suppression was observed in TT-B16M27 PeptiCRAd treated mice compared with B16M27 monocoated PeptiCRAd mice (Fig. 5H), demonstrating that the platform can also be applied to potentiate therapies based on induction of neoantigen-specific T cells.

Because we observed that preexisting immunity to tetanus toxoid potentiates the antitumor response of a double-coated PeptiCRAd alone and in combination with PD-1 blockade, we sought to further investigate whether our approach is valid also in the context of tetravalent vaccine. Polioboostrix is a tetravalent vaccine with a high coverage of 85% of infants immunized, making it an attractive study model (26). C57BL/6 mice were preimmunized with Polioboostrix vaccine with the same immunization regime as before (Fig. 6A). Serum samples and splenocytes were collected and analyzed in order to confirm the effectiveness in the immunization protocol. Tetravalent vaccine was found to efficiently generate both antibodies and CD4⁺ T-cell specific for pertussis and diphtheria (Supplementary Fig. S5A–S5C). For the tumor growth analysis, B16.OVA tumors in naïve or preimmunized mice were treated with anti–PD-1 antibody and PeptiCRAd coated with MHC-II–restricted Diphtheria–Pertussis peptides and MHC-I–restricted TRP2 peptides (DP-TRP2-PeptiCRAd). Consistent with our previous results, a superior antitumor response was detected in preimmunized treated with DP-TRP2-peptiCRAd and anti-PD1, whereas treatment efficacy was lost in naïve mice (Fig. 6B).

These results confirm that the pathogen-specific preexisting immunity can enhance the antitumor response and that the mechanism of action is dependent on the memory T cells. Moreover, this effect is not restricted to tetanus but is adaptable to other pathogens as well. To further verify that the mechanism of action behind the enhanced treatment efficacy using diphtheria and pertussis as the preimmunizing vaccine, we analyzed the T-cell repertoire of the tumor draining lymph nodes, TME, and spleen. The frequency of naïve CD8⁺ T and CD4⁺ T cells was lower in the draining lymph nodes of the preimmunized, DP-TRP2 PeptiCRAd-treated mice compared with the control groups (Fig. 6C). Concomitantly, increased levels of CD4⁺ T_EM cells were observed in the draining lymph nodes and in the TME of preimmunized mice compared with the naïve and mock treated mice (Fig. 6D). In addition, a significant higher infiltration of TRP2-specific CD8⁺ T cells was seen in the tumor tissue of the immunized mice when compared with the naïve mice (Supplementary Fig. S5D), and the level of CD4⁺ T_EM cells in the tumor and draining lymph nodes strongly correlated with the intensity of the TRP2-specific TIL response. Taken together, the double-coated PeptiCRAd vaccine platform can be used to stimulate preacquired, pathogen-specific CD4⁺ T-cell immunity to help the generation of effective antitumor CD8⁺ T-cell responses.

Vaccines lead to the formation of an immunologic memory that is able to deploy a much faster and more effective immune response when reencountering the pathogen; in fact, the primary immune response is rather weak and slow while the secondary immune response is faster and more effective (27). When cancer vaccines (e.g., peptides, nanoparticles, virus-like particles) are used as therapeutic vaccines and not as prophylactic vaccines, the immune response observed after treatment is usually more similar to a primary than a secondary one. Among cancer vaccine strategies, oncolytic viruses are strongly reemerging as leading biological drugs used to synergize with checkpoint inhibitors given their ability to trigger tumor-specific T-cell responses (18, 28, 29). The idea of utilizing mainly MHC-I–restricted peptides as vaccine regimens to induce antitumor CD8⁺ T-cell response has dominated the field of cancer immunotherapy. Only until recently, more attention has been directed toward exploiting the immunologic “help” provided by CD4⁺ T-cell population (12, 30). Indeed, the CD4⁺ T cells are required to shape the response and the memory of cytotoxic CD8⁺ T cells (12, 31), “highlighting the potential of CD4⁺ T cells as a tool for cancer immunotherapy” (32). Nevertheless, only few studies until now have actively exploited the promising interplay between CD4⁺ and CD8⁺ cells to elicit a strong and more effective antitumor response (8).

Here, we investigated the cross-talk between CD8⁺ and CD4⁺ T cells within the TME and evaluated the strategy of exploiting the memory repertoire of CD4⁺ T cells to mount a fast and reliable antitumor response. To this end, we describe a new cancer immunotherapy approach that takes advantage of the preexisting pathogen-specific immunologic memory present in the worldwide population of vaccinated individuals. To increase the antitumor response, we modulated the interplay between CD4⁺ and CD8⁺ T cells within the TME by using our PeptiCRAd platform, which is based on oncolytic adenovirus coated with MHC-restricted peptides. PeptiCRAd platform was further developed here into a hybrid treatment engaging not only the antitumor CTL but also the CD4⁺ help for an optimally shaped and robust enough antitumor immune response to occur. The feasibility of our strategy was demonstrated and validated in melanoma using tetanus and polioboostrix vaccines available for humans, highlighting the universal nature of the CD4⁺ memory in boosting cancer-specific CTL responses.

One of the most important aspects of the hybrid PeptiCRAd system is that when DCs process the virus and the peptides attached onto its surface, the DCs present not only tumor-specific peptides to CD8⁺ T cells to trigger the antitumor immune response, but importantly they also present pathogen-specific peptides to CD4⁺ T helper cells that potentiate and sustain the cytotoxic immune response. Following intratumoral therapy with double-coated PeptiCRAd, we were able to detect increased levels of TT-specific Th1 CD4⁺ cells expressing CD40L, which has been previously shown to engage its receptor CD40 on APCs (23), leading to licensing and maturation of these cells. Indeed, we saw an increase in infiltration and activation of DC within the TME in the mice that had the best control of the tumor growth, which is in line with the notion that DCs play a major role in linking the innate and the adaptive immune responses and they are capable of stimulating the T cells within TME (33). These results are in line with a recent clinical study (34) showing that preconditioning with tetanus–diphtheria toxoid is sufficient to enhance the DC migration and improve antitumor response in patients with glioblastoma by activating pathogen-specific CD4 T cells. Moreover, consistently with the high DC activation, a significantly higher frequency of effector memory CD4⁺ and CD8⁺ TILs as well as activated CD8⁺ T cells with lower expression levels of exhaustion marker TIM3 were seen in responding mice compared with the nonresponders. This highlights the importance of the interplay between the innate and adaptive arm of the immune system as well as the key role of effector memory CD4⁺ T cells in supporting the ongoing antitumor response.

The use of immune checkpoint inhibitors (ICI) to block the interaction between the immune checkpoint receptors and their ligands is worldwide accepted as the breakthrough of the cancer immunotherapy field with anti–CTLA-4 (ipilimumab) anti–PD-1 (nivolumab and pembrolizumab), and anti–PD-L1 (atezolizumab) agents approved for the treatment of several indications, including melanoma (35, 36). The concept has revolutionized the way cancer is treated, and in some cases, has completely changed the life expectancy of cancer patients. It has become very clear that the presence of CD8⁺ T lymphocytes within the tumor is a critical parameter, necessary but not sufficient, to predict patient's positive overall response even with ICI therapies (36, 37). However, many cancer types are resistant to ICI treatments or develop resistance over time (38). To this end, the use of oncolytic viruses in combination with ICIs has been proposed as an effective strategy, as oncolytic viruses possess a natural ability to induce beneficial changes in the TME (39). Indeed, T-VEC, the first oncolytic therapy approved for metastatic melanoma, has been tested in combination with ICI (18, 40). In these studies, T-VEC modulated the T-cell infiltration, promoting increased frequency of CD8⁺ T cells in patients that responded to the combination therapy (28). Oncolytic vaccines that actively promote T-cell activation and tumor infiltration have regained an enormous momentum, leading to the influx of public and private investments (18, 41). However, simply inducing a T-cell response without having control over the quality of this response might not be sufficient in many cases.

Our hybrid PeptiCRAd treatment is based on an oncolytic adenovirus and thus induces the infiltration of TILs to tumors, an ability common to several oncolytic viruses, which have been shown to mount an pro-inflammatory response that can result in the generation and infiltration of effector T cells in both virus-injected and distant tumors (28, 42). However, in contrast to these more traditional oncolytic virus strategies, our technology also effectively exploits the interplay between CD4⁺ and CD8⁺ T cells and engages the pathogen-specific CD4⁺ memory effector T cells to the fight against cancer. As a clear increase of CD8⁺ TIL infiltration was observed when hybrid PeptiCRAd was used as a single treatment, a combination treatment with ICI was a logical next step in this research. Indeed, irrespective of the identity of the pathogen peptides or the tumor peptides loaded onto the hybrid PeptiCRAd, the combination therapy with ICI clearly showed an advantage over the single treatments, resulting in the complete response rate of 75% in the group treated with anti–PD-1 and TT-TRP2-PeptiCRAd. This is a notable improvement in treatment efficacy, especially considering that PeptiCRAd is based on human adenovirus 5, which does not replicate in or lyse murine tumor cells, unlike other oncolytic viruses that have been reported to increase response rates to ICI in mouse models (28, 43–45).

The localization of the antitumor T cells to the tumor border has been recognized as a central aspect in successful immunotherapy (38). Further, CD4⁺ T cells have a role in licensing DCs via CD40-CD40L pathway to prime tumor-specific CTLs in lymphoid organs and in establishing a durable antitumor immunity (31, 46). Indeed, upon treatment with hybrid PeptiCRAd, we found a high level of CD4⁺ T_EM cells in the TME and in the draining lymph nodes that positively correlated with the frequency of TRP2-specific CD8⁺ TILs in preimmunized mice. These results further corroborate that the immunologic profile after preimmunization and hybrid PeptiCRAd treatment resembles a secondary response rather than a primary one that usually is seen with oncolytic vaccines.

In summary, we described for the first time how to modulate the interplay between CD4⁺ and CD8⁺ T cells by using our PeptiCRAd platform and demonstrated the importance of the CD4⁺ T-cell memory repertoire specific for a pathogen antigen in enhancing the CD8⁺ T-cell antitumor response in cancer immunotherapy. Finally, our technology significantly increased the antitumor efficacy of anti–PD-1, conferring a rationale for a combination therapy of the hybrid PeptiCRAd with checkpoint inhibitors.

S. Tähtinen is a postdoctoral research fellow at Genentech. C. Capasso is a consultant at Valo Therapeutics. V. Cerullo is the co-funder and shareholder at and has ownership interest (including patents) in Valo Therapeutics. No potential conflicts of interest were disclosed by the other authors.

Conception and design: S. Tähtinen, S. Feola, C. Capasso, L. Buonaguro, V. Cerullo

Development of methodology: S. Tähtinen, S. Feola

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): S. Tähtinen, S. Feola, N. Laustio, M. Fusciello, M. Medeot

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): S. Tähtinen, S. Feola, C. Capasso, C. Groeneveldt, E.O. Ylösmäki, L. Ylösmäki, B. Martins, M. Tagliamonte, K. Peltonen, V. Cerullo

Writing, review, and/or revision of the manuscript: S. Tähtinen, S. Feola, E.O. Ylösmäki, L. Ylösmäki, J. Chiaro, F. Hamdan, K. Peltonen, T. Ranki, V. Cerullo

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): S. Feola, B. Martins, M. Fusciello, M. Medeot, J. Chiaro, F. Hamdan

Study supervision: S. Tähtinen, S. Feola, V. Cerullo

We thank all the participants for their support and advice. Moreover, the flow cytometry analysis was performed at the HiLife Flow Cytometry Unit, University of Helsinki, and the animal experiment carried out at the Laboratory Animal Center (LAC) of the University of Helsinki. This work has been supported by European Research Council under the European Union's Horizon 2020 Framework programme (H2020)/ERC-CoG-2015 Grant Agreement n. 681219, Helsinki Institute of Life Science (HiLIFE), Jane and Aatos Erkko Foundation (decision 19072019), Orion Research Foundation, Jalmari and Rauha Ahokas Foundation, Maud Kuistila Foundation, and Cancer Society of Finland (Syöpäjärjestöt).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.