All cell lines were grown according to ATCC/ECACC recommendations and ICAM-1-expressing KG-1 cells (ICAM-1/KG-1) were generated using lentiviral transduction.

Peripheral blood mononuclear cells (PBMC), obtained from healthy volunteers or leukocyte apheresis cones (National Health Service Blood and Transplant), were isolated using Lymphoprep™ (Fresenius-Kabi AS, Halden, Norway) density-gradient centrifugation as previously described [7]. PBMC were used at 2 × 10⁶ cells/mL and PBMC-conditioned medium (CM), collected 48 h after CVA21 treatment, was stored at − 20 °C prior to use. For type I IFN neutralisation, PBMC were pre-incubated for 30 mins with polyclonal antibodies prior to addition of CVA21 for 24 h [10]. For ICAM-1 blockade, PBMCs were either treated with 10 μg/mL LEAF™ purified anti-human ICAM-1-blocking antibody, a mouse IgG1-isotype control antibody (both BioLegend), or left untreated for 30 min at 37 °C. Subsequently, the PBMC were exposed to CVA21 for 24 h, before evaluation of NK cell activation and cytokine secretion.

Blood samples were obtained from patients diagnosed with AML or patients taking part in the STORM Phase 1 clinical trial (NCT02043665/Keynote-200/VLA009A) following additional informed consent. Written, informed consent was obtained from all patients in accordance with local institutional ethics review and approval (06/Q1206/106). Peripheral blood samples collected from patients recruited into the STORM clinical trial were collected from Cohort 1 and 3 (Cycle 1), who received 1 × 10⁸ or 1 × 10⁹ TCID₅₀ clinical grade CVA21, respectively. Samples were collected prior to the first CVA21 infusion in Cycle 1, then at 1 h, 3 days and 22 days after the first infusion; samples collected on day 3 and 22 were taken before scheduled CVA21 treatments.

CD14⁺ monocytes and plasmacytoid dendritic cells (pDC, Lin⁻, BDCA-2⁺, BDCA-4⁺, CD123⁺, CD4⁺, CD45RA⁺, BDCA-3^dim, BDCA-1⁻, CD2⁻) were isolated or depleted from whole PBMC using magnetic cell sorting on MACS® columns (Miltenyi Biotec, Bergish Gladbach, Germany), according to the manufacturer’s instructions. PBMC (±CD14⁺ monocytes or pDC) were used for collection of CM, assessment of NK cell activation and for CTL priming assays (see below).

mRNA was isolated using the RNeasy Plus kit (Qiagen, Hilden, Germany) and converted to cDNA using the SuperScript™ III Reverse Transcriptase kit and oligoDT priming (Thermo Fisher Scientific, Waltham, MA). Gene expression was evaluated by qPCR using TaqMan™ reagents and the QuantStudio 5 Real-Time PCR System (Thermo Fisher Scientific, Waltham, MA).

Wild-type coxsackievirus A21, Kuykendall strain, was provided by Viralytics Ltd. (Sydney, Australia) or propagated in-house from wild-type CVA21 obtained from ATCC® (ATCC®VR-850™). For propagation, supernatants were harvested following CVA21 infection of Mel624 cells for 24 h. CVA21 was pelleted by centrifugation at 36000 rpm for 2 h (SW45 rotor, Optima™ L-80 ultra-centrifuge, Beckman Coulter) and harvested virus was purified using OptiPrep™ density gradient centrifugation, 35–15% gradient (36,000 rpm, 1.5 h, SW41 Ti rotor). Viral titer was determined using a standard plaque assay on Mel624 cells.

MTS assays were performed according to the manufacturer’s protocol (Abcam, Cambridge, UK). Optical density was measured at 450 nm using a Multiscan EX microplate reader (Thermo Fisher Scientific).

IFN-α secretion was detected using matched paired antibodies (MabTech AB) and standard ELISA techniques. PBMC-CM was also analyzed using multiplex bead arrays (Bio-Plex Pro™ Human Cytokine 27-plex and 23-plex Assay; Bio-Rad) as per the manufacturer’s instructions. Plates were analyzed using a Bio-Plex 100 reader with Bio-Plex Manager software.

Immature DC (iDC) generation and CTL priming assays were performed as described by Prestwich et al [8]. Briefly, tumor cells (±0.1 pfu/cell CVA21 for 24 h) were loaded onto CD14⁺ monocyte-derived iDC and co-cultured with autologous PBMC for 1 week. CTL were then re-stimulated with tumor-loaded iDC (±CVA21) and cultured for a further 7 days. Primed CTL were then harvested for ⁵¹Cr release assay, CD107 degranulation or peptide recall assays.

Where indicated, CTL generation was also performed in the absence of iDC, using only CVA21-treated tumor cells. To do this, CVA21-infected tumor cells, treated with 0.1 pfu/cell CVA21 for 24 h, were centrifuged to remove free virus, and incubated for a further 48 h prior to addition of PBMC. PBMC were cultured for 1 week and re-stimulated with tumor cells (±CVA21 treatment – as above).

For TAA peptide stimulation, autologous CD14⁺ cells were thawed from frozen, washed with medium, and allowed to rest for 60 min prior to incubation with 6 nmol/mL PepTivator® peptide pools (15-mer peptide sequences with 11 amino acids overlap, Miltenyi Biotec) for 60 min at 37 °C; PepTivator® Peptide Pools were reconstituted in dH₂O at 30 nmol/mL and aliquots were stored at − 80 °C. Autologous CD14⁺ cells with or without PepTivator® peptide loading were then co-cultured with CTL and intracellular IFN-γ production was determined by flow cytometry.

For all flow cytometry analysis, cells were stained with relevant antibodies (see Additional file 1: Table S1 and analysis was performed on an Attune® Acoustic Focusing Cytometer (Thermo Fisher Scientific) or CytoFLEX S (Beckman Coulter). Cell viability was evaluated using a LIVE/DEAD® Fixable Dead Cell Stain Kit (Thermo Fisher Scientific). For NK cell and CTL CD107 degranulation assays, PBMC were co-cultured with target cell lines (2:1 ratio) for 1 h. Following this, Brefeldin A (1:1000, BioLegend, San Diego, CA) and relevant antibodies (see Additional file 1: Table S1) were added to each sample for 4 h. For intracellular IFN-γ staining, CTLs were co-cultured with target cells (2:1 ratio) or autologous CD14⁺ cells (±PRAME, Mucin-1 and MAGE-A1 peptide pools) for 1 h at 37 °C before addition of Brefeldin A (1:1000, BioLegend) and relevant antibodies (Additional file 1: Table S1) for a further 4 h at 37 °C. CTLs were then fixed in 1% paraformaldehyde (PFA) overnight and permeabilized using 0.3% saponin (Sigma-Aldrich) prior to intra-cellular IFNγ staining and acquisition by flow cytometry.

Target cells were labelled with 100 μCi ⁵¹Cr (PerkinElmer, Waltham, MA) then co-cultured with effector cells as previously described [6]; unlabeled K562 and Daudi cells were included in the analysis of CTLs to reduce non-specific killing. For NK cell ⁵¹Cr release assays, the data shown represents the effector:target ratio of 50:1 for AML (0.1pfu/cell CVA21), and 25:1 for MM (1pfu/cell). ⁵¹Cr was measured using a Microbeta² scintillation counter (PerkinElmer) and the percentage cell lysis was calculated (cpm: counts per minute): %lysis=100×samplecpm−spontaneouscpmmaximumcpm−spontaneouscpm\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \% lysis=100\times \frac{sample\ cpm- spontaneous\ cpm}{maximum\ cpm- spontaneous\ cpm} $$\end{document}

Statistical analysis was performed using Graphpad Prism 7.0 software. p-values were calculated using either Student’s t-test, one-way analysis of variance (ANOVA) or two-way ANOVA. Results were considered significant if p < 0.05 (* = p < 0.05, ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001). Pearson’s r was calculated to evaluate correlations.

As CVA21 is dependent on human ICAM-1 for cell entry, only human model systems or immunocompromised human xenograft murine models allow in vivo testing of CVA21, with the latter preventing exploration of the role of adaptive anti-tumor immunity. Therefore, to initially confirm the immunotherapeutic potential of CVA21 in vivo, and in potentially immunosuppressed cancer patients [36–39], we took advantage of access to samples from STORM (VLA009A) clinical trial patients. This Phase I dose escalation study examined the safety of i.v. delivery of CVA21 to patients with mixed types of advanced cancers including melanoma, prostate and squamous cell carcinoma of the lung (Fig. 1a). STORM patients that consented to have additional research blood samples taken for analysis received 1 × 10⁸ (Cohort 1) or 1 × 10⁹ TCID₅₀ (Cohort 3) clinical grade CVA21 on days 1, 3 and 5 (Cycle) and blood samples were collected prior to the first CVA21 infusion, then at 1 h, 3 days and 22 days after the first infusion (Fig. 1b). No consistent increase in IFN-α secretion, a key mediator of host anti-viral and anti-tumor responses, was observed in the patient plasma samples at these time points (data not shown). However, using qPCR, we observed an increase in the expression of interferon-stimulated genes (ISG), IFIT1, IFI44L and OAS1, 3 days after the first CVA21 infusion, compared with the baseline pre-infusion samples (Fig. 1c); this was more pronounced in patients who received a higher dose of virus and suggested the onset of a type I IFN response in the peripheral blood of CVA21-treated patients.Fig1Fig. 1

Intravenous CVA21 induces a type I IFN response and activates immune effector cells in vivo. a. As part of the STORM clinical trial, patients (n = 5) with advanced malignancies were administered with 1 × 10⁸ (red symbols) or 1 × 10⁹ TCID₅₀ (black symbols) clinical grade CVA21 i.v. b. CVA21 was infused on day 1, 3 and 5 and blood samples were taken pre-infusion (a), 1 h (b), 3 days (c) and 22 days (d) after the first infusion. c. cDNA was made from PBMC collected pre-infusion (a) and on day 3 (c) and the expression of IFIT1, IFI44L and OAS1 was measured by qPCR. Results were normalized to 18S RNA expression and the fold increase in expression (calculated as ΔΔCt) compared to pre-infusion is presented. d-f CD69 expression on NK cells (d), CD4⁺ T cells (e) and CD8⁺ T cells (f) was analyzed at each time point. *denotes statistical significance

The induction of ISGs following i.v. infusion of CVA21 implies an interferon response, and IFN-α2 has known cytotoxic potential against both AML and MM, with multiple IFN-α-based clinical trials having been completed [40, 41]. To date, the inflammatory milieu induced by CVA21 has not been thoroughly explored; moreover, its potential role in CVA21 efficacy remains unknown. To initially investigate this, we used a multiplex assay to examine the range of cytokines and chemokines induced from HD-PBMC following CVA21 treatment. These data confirmed that CVA21 elicited a strong cytokine response (Fig. 2a), with a number of potentially cytotoxic cytokines, including IFN-α2, TRAIL and IFN-γ, being identified. Having previously identified KG-1, HL-60, kasumi-1 and OPM2 cells as resistant to CVA21-direct oncolysis (Additional file 3: Figure S2B and D) these cells were subsequently used to examine the cytotoxic potential of CVA21-induced inflammation. KG-1, HL-60, kasumi-1 and OPM2 cells were cultured in PBMC-CM (±CVA21 treatment) for 96 h and cell viability was examined; CVA21-treated PBMC-CM significantly reduced the viability of all CVA21-resistant cell lines (Fig. 2b), which, given resistance to CVA21-direct oncolysis, was highly suggestive of CVA21-induced bystander cytokine killing. In support of this, reconstitution of culture media with recombinant IFN-α or IFN-γ (cytokines secreted in response to CVA21 treatment) also demonstrated a small, but significant, increase in killing of CVA21-resistant KG-1 cells (Additional file 4: Figure S3A).Fig2Fig. 2

CVA21 treatment induces cytokine-mediated bystander killing of CVA21-resistant cells. a. Conditioned-media (CM) was collected from healthy donor PBMC after 48 h (±CVA21 treatment; 1 pfu/PBMC) and analyzed for cytokines using a 48-plex multiplex assay. The mean fold change (n = 3) compared to untreated PBMCs is shown. Sample readings outside of the detection range were estimated using the assay range limits and are indicated by x. b. CVA21-resistant cell lines were cultured for 96 h in PBMC-CM (0, 0.1 and 1 pfu/cell CVA21 treatment for 48 h; n = 6 for KG-1, HL-60 and OPM2, n = 5 for kasumi-1) and cell viability was assessed using an MTS assay. Error bars indicate SEM. *denotes statistical significance

We have previously shown that OV-induced type I IFN-α can increase the anti-tumor properties of NK cells [7], therefore given that CVA21 stimulated IFN-α in vitro (Fig. 2a), and activated NK cells in vivo (Fig. 1d), we examined the ability of OV-activated NK cells to eradicate CVA21-sensitive and CVA21-resistant cells. Initially we confirmed that CVA21 could activate NK cells, in vitro, and demonstrated that treatment of HD-PBMCs with CVA21 induced a significant increase in CD69 expression on NK cells, as expected (Fig. 3a). Pivotally, this heightened state of NK cell activation was associated with improved recognition and killing of both CVA21-sensitive (H929, U266B and JIM3) and CVA21-resistant (AML cell lines and OPM2) cell targets, as measured by CD107a/b expression (Fig. 3b) and chromium release (Fig. 3c), respectively. This confirmed the ability of CVA21 to induce NK cell mediated anti-tumor immunity.Fig3Fig. 3

CVA21 treatment enhances NK cell activation and function. a. CD69 expression on healthy donor NK cells (CD3⁻CD56⁺) following CVA21 treatment for 48 h (n = 4). b. Healthy donor PBMC (± 0.1 pfu/cell CVA21) were co-cultured at a 2:1 ratio with AML or MM target cells for 5 h and the percentage of NK cells expressing CD107a/b was determined (n = 4); data for CVA21-sensitive (right) and CVA21-resistant (left) cell lines are shown. c. CVA21-sensitive (right) and CVA21-resistant (left) cell lines were labelled with ⁵¹Cr and then co-cultured with healthy donor PBMC (±CVA21 treatment) for 4 h and the percentage lysis of target cells was determined (n = 4). Error bars indicate SEM. *denotes statistical significance

While innate immunity is rapid and instrumental for the eradication of tumor cells, generation of adaptive anti-tumor immunity is necessary for long-term immunological memory. The activation of both CD4⁺ and CD8⁺ T cells in cancer patients following CVA21 treatment (Fig. 1e and f) suggests the induction of a T cell immune response; however, its relationship to anti-viral and/or anti-tumor immunity is unclear. To evaluate the ability of CVA21 to stimulate the production of tumor-specific CTLs, we adapted our previously established protocol for OV CTL priming [8], to the hematological setting. This protocol involved long-term co-culture of CVA21-infected tumor cell targets, pre-loaded onto myeloid-derived dendritic cells (mDC), with PBMC autologous to the tumor-loaded mDC [8]. We initially examined the ability of CVA21 to stimulate CTL priming using CVA21-sensitive cells (U266B MM cells and ICAM-1/KG-1 AML cells) which, following direct oncolysis, should release damage-associated molecular patterns (DAMPs) and pathogen-associated molecular patterns (PAMPs) to activate mDC and facilitate CTL priming. Firstly, we identified that for efficient CTL priming, and lysis of relevant cell targets, the presence of CVA21 was required (Fig. 4Ai and Bi); moreover, primed-CTL were tumor specific, as only relevant, but not irrelevant, target cells were capable of stimulating intracellular IFN-γ production (Fig. 4Aii and Bii).Fig4Fig. 4

CVA21 can prime tumor-specific CTL. a. and b: CVA21-sensitive U266B MM (a) and ICAM-1/KG-1 (b) cell targets were used. Tumor cells were pre-treated with CVA21 (0.1 pfu/cell) for 24 h, then loaded onto mDC prior to being co-cultured with autologous PBMC and one round of re-stimulation. a. and bi. CTLs primed in the presence or absence of CVA21 were co-cultured with ⁵¹Cr-labelled relevant targets (U266B and ICAM-1/KG-1 cells, respectively) at different effector:target ratios for 4 h. The percent cell lysis was determined using ⁵¹Cr release (n = 6). a. and bii. CTL intracellular IFN-γ production following a 5 h co-culture with relevant (U266B or ICAM-1/KG-1, respectively) or irrelevant (ICAM-1/KG-1 or Raji, respectively) targets (n = 3). a. and biii. Intracellular IFN-γ production following a 5 h co-culture with autologous CD14⁺ cells loaded with appropriate peptide pools (Mucin-1 and MAGE-A1; U266B CTLs, and PRAME; ICAM-1/KG-1 primed CTLs). c. CTL priming with CVA21-resistant cells (parental-KG-1 (i) and THP-1 (ii)). The percentage of tumor specific CTLs (CD3⁺CD8⁺) was determined using CD107a/b degranulation after a 5 h co-culture with relevant (KG-1; n = 3 or THP-1; n = 2) or irrelevant (Raji) cell targets. Spontaneous CD107 expression was subtracted from the values shown. Error bars indicate SEM. *denotes statistical significance

Given the efficient CTL priming demonstrated in Fig. 4, it was postulated that CVA21 would induce DC maturation following co-culture with CVA21-infected cell targets, and thus provide the necessary antigen presentation and co-stimulation to support efficient CTL priming. Surprisingly, upon phenotyping of mDC following treatment with CVA21 alone, or CVA21-loaded targets, limited mDC maturation was observed; CVA21 was unable to directly stimulate mDC maturation and only modest maturation (a small but significant increase in CD86 expression) was observed upon co-culture of mDC with CVA21-treated ICAM-1/KG-1 (Fig. 5a). Therefore, the importance for mDC during the course of the CTL priming assay was examined. To investigate this, we compared the ability of CVA21 to prime tumor-specific CTLs in the presence or absence of ex vivo generated, tumor-loaded, mDC. Figure 5b shows that CTLs primed with or without mDC were comparable in their capacity to lyse relevant tumor cell targets, demonstrating that CVA21-treated tumor cells can support CTL priming, irrespective of the presence or absence of mDC. Importantly, the CTLs generated in the absence of mDC retained their tumor specificity as CTL degranulation was only observed upon recognition of relevant, but not irrelevant, cell targets (Fig. 5c). Collectively, these data confirmed that in vitro generated mDC were not required for successful CTL priming, and suggested that all cellular components required for adaptive CTL priming, by CVA21-treated tumor cells, were present in the peripheral blood. This finding is of particular importance in the context of hematological malignances as CVA21-loaded tumor cells may co-exist, in the blood, with immune cell components that are necessary for effective CTL priming.Fig5Fig. 5

mDC are not necessary for priming of AML-specific CTL. a. mDC were treated with CVA21 or CVA21-treated ICAM-1/KG-1 cells for 48 h and expression of the activation markers, CD86, CD80 or HLA-DR, was examined (n = 4). b. ICAM-1/KG1-specific CTLs were primed with or without autologous mDC (±CVA21) and CTL-mediated lysis of relevant ICAM-1/KG-1 targets was measured by ⁵¹Cr release assay. Solid lines indicate CTL primed in the presence of mDC, dashed lines indicate CTL primed in the absence of mDC (n = 3). c. ICAM-1/KG1 CTLs, primed in the absence of mDC, were co-cultured with relevant (ICAM-1/KG-1) or irrelevant (Raji) target cells (n = 3) for 5 h and tumor specificity was examined using CD107a/b degranulation. Error bars indicate SEM. *denotes statistical significance

Data presented in Figs. 1, 2, 3, 4 and 5 demonstrate that CVA21 can activate innate and adaptive anti-tumor immunity against tumor cells which are both sensitive and resistant to CVA21-direct oncolysis; moreover, it appeared that all the cellular components required for this to occur were present in the blood. Therefore, to identify potential biomarkers of CVA21 response we sought to further characterize the molecular and cellular determinants required for CVA21-mediated immune activation. Our preliminary studies suggested that CVA21 was unable to directly activate isolated NK cells (data not shown), therefore, in the context of PBMC, we initially examined the role of anti-viral type I IFN for the induction of CVA21-mediated NK cell anti-tumor immunity. Using monoclonal type I IFN-blocking antibodies, we confirmed that NK cell activation was mediated by type I IFNs, as NK cell CD69 upregulation (Fig. 6a) and increased NK cell degranulation (Fig. 6b), following CVA21 treatment of PBMC, was not observed when type I IFN signaling was inhibited. Moreover, following this the importance of ICAM-1 (required for CVA21 infection of tumor cells [24]) on immune cell components was investigated; blockade of ICAM-1 within PBMC, prior to and during CVA21 treatment, completely abrogated the secretion of IFN-α (Fig. 6c) and prevented NK cell activation (no increase in CD69 expression; Fig. 6d) demonstrating a significant role for ICAM-1 in mediating CVA21-induced immune activation.Fig6Fig. 6

Type I IFN and ICAM-1 are required for CVA21-induced anti-tumor immunity. a-d. HD-PBMC were treated with CVA21 for 24 h, with or without pre-treatment with type-1 IFN blockade or an ICAM-1-blocking antibody. NK cell CD69 expression (a) and NK cell CD107a/b degranulation (b) were measured in the presence of type I IFN blocking antibodies or corresponding isotype antibodies (n = 4). c. HD-PBMC were treated with CVA21, in the presence or absence of ICAM-1-blocking antibodies, and IFN-α secretion was examined by ELISA (n = 3). d. NK cell CD69 expression was determined following CVA21treatment, with or without pre-treatment with an ICAM-1-blocking antibody or isotype control (n = 3). e. ICAM-1 expression was measured on mature hematopoietic immune cells (CD45⁺) from primary AML patients (n = 14). f. Correlation of CVA21 response (death of AML blasts) with ICAM-1 expression on mature CD45+ hematopoietic cells from primary AML samples. Error bars indicate SEM. *denotes statistical significance

As ICAM-1 was identified as a key mediator of CVA21 anti-tumor immunity, we next sought to identify the immune cell component responsible for CVA21 recognition, and downstream immune activation. Initially, we identified both monocytes (CD14⁺) and pDC as cell populations which expressed significantly more ICAM-1 than NK cells, CD4⁺ and CD8⁺ T cells (Fig. 7A), therefore, we hypothesized that these two cell types may act as key regulators of CVA21-mediated anti-tumor immunity. To test this, we first examined IFN-α production (a key mediator of NK cell activation; Fig. 6) from PBMC depleted of CD14⁺ monocytes, pDC, or both (CD14⁺ cells and pDC), along with isolated CD14⁺ monocytes and pDC. Figure 7B demonstrates that pDC, in isolation, secreted large amounts of IFN-α following CVA21 treatment, and that IFN-α secretion from PBMC was abrogated following pDC depletion. By contrast, CD14⁺ depletion had no significant effect on IFN-α levels following CVA21 treatment, and isolated CD14⁺ cells did not secrete IFN-α in response to CVA21 treatment. To further examine the role of monocytes and pDC for anti-tumor immunity, we repeated PBMC-CM toxicity, NK cell activation and degranulation assays, as well as T cell priming experiments but depleted CD14⁺ monocytes, pDC, or both, from the whole PBMC population. Interestingly, CVA21-treated PBMC-CM remained toxic to kasumi-1 and HL-60 cells in the absence of CD14⁺ cells; however, when PBMC were depleted of pDC, the cytotoxicity of CVA21-treated PBMC-CM was reduced to levels comparable with untreated PBMC-CM (Fig. 7C). Moreover, the absence of IFN-α following pDC depletion (but not monocyte depletion), abrogated CVA21-induced NK cell activation with regards to both CD69 up-regulation, and enhanced NK cell degranulation (Fig. 7D). Furthermore, CTL priming assays (performed without the addition of autologous mDC) also revealed the importance of pDC, as the absence of CD14⁺ did not significantly decrease the production of tumor-specific CTLs; however, removal of pDC significantly reduced levels of tumor-specific CTL (Fig. 7E). Taken together, these results demonstrate, for the first time, the critical role of pDC in orchestrating CVA21-induced innate and adaptive anti-tumor immune responses and confirm the immunotherapeutic potential of this agent.Fig7Fig. 7

pDC orchestrate innate and adaptive CVA21 anti-tumor immunity. a. ICAM-1 expression on immune cell components from healthy donors (n = 3). b-e CD14⁺ monocytes and pDC (CD123⁺BDCA-2⁺) were depleted from whole PBMC (W.PBMC) prior to analysis. b. IFN-α secretion from whole or depleted PBMC, or isolated CD14⁺ cells and pDC, was measured by ELISA 48 h post-CVA21 treatment. c. CM was generated from whole or depleted PBMC, following treatment with 0.1 pfu/PBMC CVA21 for 48 h. The cytotoxicity of CM against kasumi-1 (i) and HL-60 (ii) cells after 96 h was evaluated by MTS assay. d. NK cell CD69 expression (i) and NK cell CD107a/b degranulation (ii) after treatment of whole of depleted PBMC with 0.1pfu/cell CVA21 for 48 h was determined. e. ICAM/KG-1 cells (±CVA21 and without addition of mDC) were used to prime CTL and PBMCs depleted of CD14⁺ monocytes or pDC were used as effector cells. Tumor-specific CTLs were detected using CD107a/b degranulation assays against cell targets. Error bars indicate SEM. *denotes statistical significance. n.s. = not significant