The African green monkey kidney cell line Vero was obtained from the American Type Culture Collection (CCL-81). FC1245 cells, derived from the KPC model (14), were obtained from A. Nowrouzi (DKFZ Heidelberg). FC1245-CD46 expressing the MV entry receptor human CD46 were generated by lentiviral transduction as described previously (15). B16-CD46 and MC38-CD46 were generated by stable transfection. Vero and MC38-CD46 cells were cultivated in Dulbecco’s modified Eagle’s medium (DMEM, Life Technologies) supplemented with 10% fetal calf serum (FCS, Biosera). B16-CD46, FC1245, and FC1245-CD46 cells were cultivated in Roswell Park Memorial Institute (RPMI, Life Technologies) medium with 10% FCS. All cell lines were cultivated at 37^oC in a humidified atmosphere with 5% CO₂. Routine tests for mycoplasma contamination were performed.

All experiments with human material were performed in accordance with the Declaration of Helsinki and were approved by the ethics committee of the Medical Faculty of Heidelberg University (323/2004, Amendment 03). Informed consent was received from participants before study inclusion. Generation and cultivation of patient-derived PDAC cultures have been described previously (16). Cultures were subjected to SNP typing and Multiplex Cell Contamination Testing (Multiplexion). Cultures were grown in DMEM Advanced F12 medium supplemented with 0.6% (w/v) glucose, 2 mM L-glutamine, 2% B27 supplement (1×) (all Thermo Fisher Scientific), 12 μg/ml heparin and 5 mM HEPES buffer (both Sigma Aldrich), 10 ng/ml rhFGF basic, 20 ng/ml rhFGF-10, and 20 ng/ml rhNodal (all R&D Systems). Cytokines were renewed twice per week.

MV-NIS is derived from the Edmonston vaccine strain of measles virus and encodes the sodium iodide symporter (NIS) (17). High-titer purified MV-NIS was obtained from Imanis Life Sciences. Measles Schwarz vaccine strain (MV) and its derivate encoding eGFP were generated and propagated as described previously (8, 18). Procedures for titration and characterization of recombinant measles vectors were carried out as in (18). To ensure comparability, all titrations were performed with Vero cells.

For infection, patient-derived PDAC cultures were seeded at a density of 1×10³ cells per well in 12 well plates and inoculated with MV-NIS at a multiplicity of infection (MOI) of 0.03 and 3 in triplicates, respectively. The inoculum was replaced with fresh medium 2 h post infection. Cell lysis was determined by lactate dehydrogenase (LDH) release assay. For 100% LDH release samples, mock infected cells were harvested 24 h post inoculation, subjected to two freeze-thaw cycles for complete cell lysis, and stored at -80°C. Culture supernatants were collected at designated time points, centrifuged (380 × g, 5 min) and stored at -80°C. LDH release into the supernatants was quantified using the CytoTox 96 Non Radioactive Cytotoxicity Assay kit (Promega). Relative cytotoxicity was calculated by normalization to 100% LDH release controls.

Murine tumor cells were seeded at a density of 1×10⁵ cells per well in 12 well plates and inoculated with MV-NIS at an MOI of 3 or with a measles vaccine derivative encoding eGFP at an MOI of 1. The inoculum was replaced with fresh medium 2 h post infection.

For ex vivo infections of FC1245-CD46, tumors were explanted from euthanized mice. Tumors were minced with a scalpel, passed through a 100 µm filter and cultivated in RPMI + 10% FCS with 1% antibiotic-antimycotic solution (Sigma) until confluent. Ex vivo cultures were inoculated with the measles vaccine variant encoding eGFP at an MOI of 1 as described above.

All experimental procedures with animals were approved by the regional council (Regierungspräsidium Karlsruhe, protocols G-192/15, G-58/17, G-17/19) and were performed in compliance with the German Animal Protection Law and the institutional guidelines. Six- to eight-week old female C57BL/6J mice were acquired from Harlan Laboratories. Mice were housed in groups of five in individually ventilated cages under specific pathogen-free (SPF) conditions at the Center for Preclinical Research of the German Cancer Research Center. After one week of acclimatization, 1x10⁶ FC1245-CD46 cells in 100 µl PBS were injected subcutaneously into the flanks of mice. For efficacy experiments, mice were assigned to treatment groups (n = 10 per group) when tumors reached an average size of 75 mm³, ensuring equal distribution of tumor sizes. Group size was calculated using nquery advisor 6.01 to enable detection of an effect with P ≈ 75% on tumor growth with a power of 80%, where P is defined as the probability that one subject in one group has a higher or lower value than a subject in the control group. Treatment was initiated as depicted in the schematics. Investigators were blinded to group assignment. MV-NIS (1x10⁶ ciu in 100 µl PBS) or 100 µl PBS (controls) were injected intratumorally on four consecutive days. Anti-PD-1 (clone J43, eBioscience; 100 µg in 200 µl PBS) or 200 µl PBS (controls) was injected intraperitoneally on the third day of virus treatment and every third day thereafter for a total of four doses. Tumor volume for each individual was monitored three times per week and was calculated using the formula

Animals were sacrificed when one of the following pre-defined endpoint criteria was reached: tumor volume exceeded 1000 mm³, the largest tumor diameter exceeded 15 mm, or tumor ulceration occurred. To minimize potential confounders, the order of treatments and measurements was varied. No animals were excluded from the efficacy analysis.

Paraffin sections of tumor tissue were prepared for immunohistochemical staining of PD-L1 and CD3. Deparaffinization and tissue staining were performed using a Benchmark Ultra IHC Staining module according to standard protocols (Ventana PD-L1 assay, clone SP263 and CONFIRM anti-CD3 Primary Antibody, clone 2GV6; Roche). Hematoxylin/eosin was used for counterstaining. IHC stainings were evaluated by a specialist in pathology according to standardized criteria (19).

To assess surface expression of human CD46, murine tumor cells were stained with 1 μl anti-human CD46 APC (clone TRA-2-10, Biolegend) in 100 μl PBS for 30 min on ice in the dark. Viability was assessed by staining with 0.2 μg/ml DAPI. Samples were acquired on a CytoFLEX flow cytometer (Beckman Coulter) and data were analyzed with FlowJo software (version 10.8.1, Tree Star Inc.).

To assess PD-L1 expression after measles virus infection, cells were infected at MOI 3 in 12 well plates and harvested 24 h and 48 h post infection. Patient-derived cells were stained with 1 µl anti-human PD-L1 PE (clone 29E.2A3, Biolegend) and murine cells were stained with 1 µl anti-mouse PD-L1 PE (clone 10F.9G2, Biolegend) in 100 µl PBS for 30 min at room temperature in the dark and with 0.2 μg/ml DAPI.

Samples of tumors and tumor-draining lymph nodes were prepared for flow cytometry as described in (20). In brief, fresh samples were minced, treated with 200 U/ml Collagenase I (Thermo Fisher Scientific) and passed through 100 µm cell strainers. 2x10⁶ cells in 100 µl PBS were used for staining. Prior to staining, samples were incubated with anti-mouse CD16/CD32 (Mouse Fc Block, BD Biosciences).

Tumor samples were stained with the following antibodies: 1 µl anti-CD45.2 PE/Cy7 (clone 104, Biolegend), 1 µl anti-CD3 PerCP-Cy5.5 (clone 17A2, BD Biosciences), 1 µl anti-CD4 APC-Cy7 (clone GK1.5, BD Biosciences), 1 µl anti-CD8a APC (clone 53-6.7, BD Biosciences), 1 µl anti-CD335 (NKp46) FITC (clone 29A1.4, Biolegend), 1 µl anti-CD69 PE (clone H1.2F3, Biolegend). Tumor-draining lymph node samples were stained with the following antibodies: 1 µl anti-CD3 PerCP-Cy5.5 (clone 17A2), 1 µl anti-CD4 APC-Cy7 (clone GK1.5), 1 µl anti-CD8a APC (clone 53-6.7) (all from BD Biosciences), 1 µl anti-CD44 PE (clone IM7), 1 µl anti-CD62L FITC (clone MEL-14) (both from Biolegend). After antibody staining for 30 min in the dark, washing, and staining with 0.2 µg/ml DAPI, acquisition was performed on a BD FACS LSR II (BD Biosciences). Data were analyzed using FlowJo software (version 10.0.7r2, Tree Star Inc.).

DNA was extracted from tumor samples using the DNeasy Blood and Tissue Kit (Qiagen). TCRB survey sequencing was performed by Adaptive Biotechnologies. Data were analyzed using vdjtools (21). Gini index was calculated using only productive TCRB with ineq (v0.2-13) package in R and plotted with ggplot2 (v3.3.5). Morisita index and heatmap were calculated and visualized using immunarch (v0.6.7) package in R (v4.0.5).

Total RNA was extracted from tumor samples using the RNeasy Mini Kit (Qiagen). NanoString gene expression analysis was performed by the nCounter Core Facility Heidelberg. In short, 25 ng RNA from total RNA samples was subjected to quality control by Agilent Bioanalyzer 2100 and Qubit system, hybridized with the CodeSet Immunology Panel (Mouse; NanoString Technologies), and processed using nCounter SPRINT Profiler. Raw data were normalized to the set of internal reference genes included in the CodeSet panel and data were analyzed using nSolver 4.0 software including the Advanced Analysis package (NanoString Technologies) for cell profiling and pathway analysis. For pathway analysis, genes differentially expressed between mock and treatment groups (p < 0.05) were mapped onto pathways using nSolver and log2-fold change expression values were calculated for each gene. Differential gene expression analysis was performed in R version 3.6.0 with the Bioconductor package “NanoStringDiff” (22). Data were normalized based on positive controls and housekeeping genes and corrected for background. A generalized linear model (GLM) likelihood ratio test was used as statistical test. Details are described in (22).

One microgram of tumor RNA extracted for gene expression profiling was reverse transcribed using Maxima H Minus RT (Thermo Fisher Scientific) with Olidgo(dT) primers. Quantitative PCR was conducted on a CFX96 Real-Time System (BioRad) with 1 μl cDNA or standard, 0.13 μl forward primer at 33 μM, 0.13 μl reverse primer at 33 μM, and 10.5 μl Power SYBR Green PCR Master Mix (Thermo Fisher Scientific) in a total volume of 20 µl. Standards consisted of 10-fold serial dilutions of a plasmid encoding MV N, pCG N (23), starting with 1x10⁷ gene copies/μl. The following primers were used:

The data were analyzed using CFX Manager Software (version 3.1, BioRad).

FC1245-CD46 tumors were established in six- to eight-week old female C57BL/6J mice (Janvier) and treated as described above. However, since MV-NIS was no longer commercially available, unmodified measles Schwarz vaccine (8), 1x10⁶ ciu in 100 µl OptiMEM per dose) or 100 µl OptiMEM (mock and anti-PD-1 groups) were used for treatment. Anti-PD-1 (clone J43) was obtained from BioXCell. Mice were sacrificed for spleen extraction at t1 (day 7 after treatment initiation) and t2 (day 13 after treatment initiation). Spleens were passed through 100 μm cell strainers, subjected to erythrocyte lysis with ACK Lysing Buffer (Thermo Fisher Scientific) for 10 min at room temperature and cultured in RPMI + 10% FCS + 1% penicillin/streptomycin.

To assess anti-tumor immunity, 1x10⁶ splenocytes were cultured with 1x10⁵ FC1245-CD46 target cells at an effector to target ratio of 10:1. To analyze measles-specific immunity, 5x10⁵ splenocytes were cultured with 7 μg/ml measles virus premium bulk antigen (Serion Immunologics, t1) or incubated with MV at an MOI of 0.5 (t2). Incubations with 10 μg/ml Concanavalin A (ConA, Sigma-Aldrich) or medium served as positive and negative controls, respectively. ELISpot co-cultures were established in clear 96-well MultiScreen_HTS-IP filter plates with a 0.45 μm pore size hydrophobic PVDF membrane (Merck) in a total volume of 200 µl. ELISpot was conducted using the mouse IFN-γ ELISpot pair comprising capture and detection antibodies (RRID: AB_2868948, BD Biosciences). Plates were coated overnight at 4°C with mouse IFN-γ ELISpot capture antibody prior to co-culture set-up. After 40 h of co-culture, ELISpots were developed using mouse IFN-γ ELISpot detection antibody, Streptavidin-HRP (BD Biosciences), and TMB substrate (Mabtech) according to manufacturer’s instructions. Data were acquired on a Bioreader 7000-E instrument (Biosys) and analyzed using EazyReader® software. Saturated wells were set to 450 spots. Splenocyte samples with an average spot count below 250 in the positive control (ConA) were excluded from analysis. To test for CD46 dependency of anti-tumor immune responses, 1x10⁶ splenocytes were cultured with 1x10⁵ FC1245 or FC1245-CD46 target cells at an effector to target ratio of 10:1. ELISpots were developed as described above.

Statistical analyses and data visualization were performed using Graph Pad Prism software (Versions 8.4.3 and 9.2.0, GraphPad Software, LLC). Comparisons of multiple groups were performed using ANOVA with Tukey’s post-test as indicated. Survival data were analyzed by Mantel-Cox (log rank) test with Bonferroni correction for multiple comparisons.

Patient-derived PDAC cultures which recapitulate hallmarks of human PDAC (16) were treated with MV-NIS or left untreated. Morphological differences were observed between cultures, signifying intertumoral heterogeneity. For instance, PC10 grew in islets while PC25 grew as a monolayer. Treatment with MV-NIS led to cytopathic effects, which differed between cultures. Syncytia formation, a hallmark of measles virus infection, was visible in PC25. Disruption of the cell monolayer was observed in PC31 and PC6. In other cultures, such as PC10, no major differences in culture morphology were observed upon infection with MV-NIS compared to mock treatment ( Figure 1A ). Direct oncolytic effects as determined by LDH release assay were also variable between cultures from different individuals ( Figure 1B ), reflecting tumor heterogeneity. Expression of the MV receptor CD46 did not correlate with oncolytic efficacy (Schäfer et al., unpublished data). Moreover, MV replication as determined by titration of viral progeny was modest in patient-derived PDAC cultures ( Supplementary Figure S1 ) compared to most human pancreatic cancer cell lines (24, 25). Thus, direct oncolytic activity of MV is most likely limited in PDAC. However, PDAC may be amenable to MV-mediated immunotherapy. Treatment with MV-NIS led to upregulation of PD-L1 on some cultures ( Figure 1C ), which is in line with previous reports (6, 7, 10). However, baseline expression and degree of PD-L1 upregulation were also variable. There was no direct correlation between MV permissiveness and PD-L1 upregulation. Overall, PD-L1 upregulation upon MV-NIS treatment provided a rationale for testing the combination of MV-NIS and PD-1/PD-L1 checkpoint blockade in PDAC.

We intended to further study the combination of MV and PD-1 checkpoint blockade in an immunocompentent murine PDAC model. Therefore, we tested viral replication, direct oncolytic effects and PD-L1 expression after treatment of FC1245-CD46 cells with MV-NIS. These cells are derived from the KPC mouse model (14) and stably express the MV entry receptor CD46 at levels comparable to other engineered murine cell lines ( Figure S2A ). Similar to other murine cells, FC1245-CD46 shows limited MV replication and limited sensitivity to direct MV oncolysis in vitro and ex vivo ( Supplementary Figures S2B–F ). As patient-derived PDAC cultures also showed modest viral replication and oncolysis upon treatment with MV-NIS ( Figures S1 , 1B ), we deemed FC1245-CD46 an appropriate model to study MV-mediated immunotherapy. Murine tumor cells cultured in vitro showed low expression levels of PD-L1, which were not or only slightly upregulated after inoculation with MV-NIS. Interestingly, treatment with the measles virus Schwarz vaccine strain led to stronger induction of PD-L1 than MV-NIS ( Supplementary Figure S2G ).

To test efficacy of MV combined with PD-1 checkpoint blockade, C57BL/6J mice bearing subcutaneous FC1245-CD46 tumors were subjected to mock treatment, treatment with intratumoral (i.t.) injections of MV-NIS, intraperitoneal (i.p.) injections of anti-PD-1, or the combination of i.t. MV-NIS and i.p. anti-PD-1, as outlined in Figure 2A . In this very aggressive model, mock treated animals showed rapid tumor progression ( Figure 2B ), reaching endpoint criteria within 12 to 17 days (median 14 days) after tumor implantation ( Figure 2C ). Monotherapy with either MV-NIS or anti-PD-1 had only limited effects on tumor growth and outcome, with a median time to endpoint of 16 days ( Figures 2B, C ). In contrast, combination treatment with MV-NIS and anti-PD-1 delayed tumor progression and prolonged survival significantly ( Figures 2B, C ). Nevertheless, all tumors ultimately progressed, reaching endpoint criteria at a median of 21 days. The benefits of combined MV-NIS plus anti-PD-1 compared to monotherapies were reproducible in an independent experiment ( Supplementary Figure S3 ).

To gain insights into mechanisms underlying efficacy of MV-NIS and anti-PD-1 combination treatment, mice bearing subcutaneous FC1245-CD46 tumors were treated as described above and tumor samples were collected as depicted in Figure 3A , with timepoints at baseline (t0, before treatment), one week (t1) and two weeks (t2) after treatment initiation. In line with our in vitro data, qRT-PCR of tumor samples indicated limited viral replication, with Cq values just above the detection limit at t1 after MV-NIS or combination treatment ( Figure S4 ). This indicates that direct oncolysis does not contribute substantially to anti-tumor efficacy. Since intratumoral T cell infiltration has been identified as an indicator of response to immunotherapy (26) and prognosis (27, 28) in PDAC, the quantity and distribution of CD3+ cells in tumor sections were assessed ( Figures 3B, C and Supplementary Table 1 ). Histopathological evaluation indicated increased CD3+ T cell infiltration in individual tumors after MV-NIS and MV-NIS plus anti-PD-1 combination treatment, but not after anti-PD-1 monotherapy, compared to mock treatment at t1. Slight increases compared to mock treatment were also observed for MV-NIS plus anti-PD-1 combination treatment at t2. The location of the CD3+ T cell infiltrate was circumferential in most samples.

As another potential marker for response to immunotherapy (26), PD-L1 expression within the tumor and tumor stroma was also assessed ( Figure 3B and Supplementary Table 1 ). Over 90% of stromal cells stained strongly positive for PD-L1 in all tumor sections from all treatment groups. Interestingly, tumor cell PD-L1 expression decreased from over 90% at baseline and t1 to 70% and lower at t2 in all treatment groups.

To complement these data, tumor-infiltrating lymphocytes were assessed by flow cytometry at baseline, t1, and t2 ( Figure 4 ). There was a slight, statistically non-significant increase in abundance of CD45+ cells after combination treatment at t1 compared to the other treatments ( Figure 4A ). There were no statistically significant differences in the relative abundance of CD3+ cells among all CD45+ cells ( Figure 4B ). Interestingly, MV-NIS and combination treatment led to a significant decrease in the relative abundance of CD3+ cells expressing the early activation marker CD69 compared to anti-PD-1 at t1 and mock at t2 ( Figure 4C ). Since natural killer (NK) cells have also been implicated in response to PD-1/PD-L1 checkpoint blockade (29), the relative abundance of CD335+ and activated CD335+ CD69+ NK cells was also determined ( Figures 4D, E ). The abundance of NK cells varied considerably between tumors and was not significantly different between the treatment groups at t1 and t2. Interestingly, these analyses yielded a significantly decreased abundance of NK cells expressing CD69 in the combination therapy compared to the monotherapies and mock treatment at t2.

Further, we analyzed T cells in tumor-draining lymph nodes (TDLN) by flow cytometry ( Supplementary Figures S5 , S6 ). While there were no significant differences in total CD3+ T cells in TDLN ( Supplementary Figure S6A ), there was a significant increase in the effector memory CD4+ population after MV-NIS monotherapy or combination treatment compared to anti-PD-1 monotherapy at t1 ( Supplementary Figure S6E ). Moreover, the central memory CD4+ population was significantly more abundant in TDLNs from mice treated with MV-NIS as compared to anti-PD-1 or mock at t1 ( Supplementary Figure S6F ). At t2, central memory CD4+ and CD8+ T cells were significantly increased in the combination group compared to the MV-NIS group ( Supplementary Figures S6F, I ).

We reasoned that not the quantity, but rather the quality of tumor-infiltrating lymphocytes may determine response to immunovirotherapy. Therefore, we performed TCR sequencing with tumor samples collected at baseline, t1, and t2. There were no overt differences in V and J segment usage and quantile statistics between the treatment groups ( Supplementary Figure S7 ). To further analyze clonality, the Gini index was calculated ( Figure 5A ), where 0 corresponds to minimal clonality and maximal diversity (that is, several clonotypes all having the same frequency), and 1 corresponds to the highest clonality (that is, dominance of few clonotypes). The Gini index was higher in all groups at t1 and t2 as compared to baseline (t0). No overt differences we observed in clonality of the intratumoral T cell repertoire between the different groups. Morisita’s Horn overlap index ( Figure 5B ) indicated a higher overlap in the TCR repertoire in samples from the MV-NIS and combination treatment groups, with some variability between different samples. In line with this, clonal tracking ( Figure 5C ) showed an overlap between clones from t1 and t2 in the MV-NIS and MV-NIS plus anti-PD-1 combination treatment groups, which might represent virus-reactive or tumor-reactive T cells.

We performed targeted transcriptome analysis using the Nanostring nCounter mouse immunology panel to gain insights into alternative immune-mediated mechanisms underlying initial response and subsequent tumor progression upon MV-NIS and anti-PD-1 combination treatment. Nanostring gene expression profiles revealed a unique expression pattern in mice treated with MV-NIS plus anti-PD-1 at t1, which was lost at t2 ( Figure 6 ). Immune pathway analysis using nSolver software ( Supplementary Figure S8 ) pointed towards enhanced adaptive immunity, cytokine secretion, chemokine signaling, cell adhesion, MHC I antigen presentation, T cell receptor signaling, and lymphocyte activation at t1 after combination treatment with MV-NIS and anti-PD-1 compared to the other treatment groups. Of note, also innate immune activation, genes associated with host-pathogen interaction and innate immune signaling were moderately elevated after combination treatment at t1, but not after monotherapy. These differences in gene expression signatures were no longer observed at t2.

Immune cell deconvolution analysis with nSolver ( Supplementary Figure S9 ) indicated an increased abundance of total CD45+ cells at t1 after MV-NIS and anti-PD-1 combination treatment. T cells, cytotoxic cells, and dendritic cells (DCs) appeared to be increased, and also macrophages as well as neutrophils. Interestingly, the gene signature associated with exhausted CD8+ T cells was also increased. After treatment with anti-PD-1 monotherapy moderate, but non-significant increases in DC- and neutrophil-associated genes were detected. No enrichment of specific immune cells was detected for MV-NIS monotherapy at this timepoint. Remarkably, at t2, the abundance of total CD45+ cells and cytotoxic cells appeared to be increased in this group, whereas the effects seen at t1 in the combination group had been lost. At large, the immune cell deconvolution analysis was consistent with the flow cytometry data on tumor-infiltrating lymphocytes (compare Figure 4 ) with overall less variability in the former.

To dissect the unique immune activation profile after combination treatment at t1, we performed differential gene expression analysis with tumor samples from the different treatment groups ( Supplementary Table 2 ). Compared to anti-PD-1 monotherapy, combination treatment led to increased intratumoral expression of T cell-associated genes (Cd4, Cd8b), costimulatory molecules (Cd27, Cd40l) as well as genes linked to lymphocyte activation (Cd69, Cd5, Cd6, Slamf1, Tnfrsf8) and cytotoxic effector cell function and differentiation as well as Th1 polarization (Il12a, Il12rb2, Runx3 and Tbx21, i.e. Tbet).

Overall, these data indicate broad immune activation and especially activation of effector T cells at t1 after combination treatment with MV-NIS and anti-PD-1. This immune signature was unique to the combination and not observed after either monotherapy. However, effects were lost at t2 ( Supplementary Table 3 ), indicating that the effects on the tumor immune environment were transient.

Finally we assessed whether MV plus anti-PD-1 induces not only local immunomodulation, but also systemic anti-tumor immunity in PDAC. We performed IFN-γ ELISpot analysis with splenocytes from mice from all treatment groups at t1 and t2 ( Figure 7 ). While splenocytes from mice treated with anti-PD-1 alone did not show enhanced tumor-specific reactivity compared to mock, we found significantly increased tumor-specific IFN-γ responses after treatment with MV at t1 ( Figure 7 , top left panel, adj. p < 0.01 between MV and mock). Combination with anti-PD-1 yielded similar reactivity as MV alone, indicating that anti-tumor immunity is driven by virotherapy. Remarkably, despite loss of the immune activation signature in the tumor at t2, systemic anti-tumor immunity was maintained, although the differences were no longer statistically significant ( Figure 7 , top right panel). Moreover, virotherapy induced MV-specific immunity at t1, which was maintained at t2 ( Figure 7 , bottom panels, adj. p < 0.01 between MV and mock at t1, adj. p < 0.001 between MV and mock at t2). Combination with anti-PD-1 did not affect anti-MV immunity.

Further, we confirmed that anti-tumor immunity does not depend on tumor CD46 expression by stimulation of splenocytes with both parental and CD46-positive FC1245 cells ( Supplementary Figure S10 ).

Overall, our results indicate that MV virotherapy can induce durable systemic anti-tumor immunity in PDAC, whereas peripheral immune signatures are transiently modulated by combination with anti-PD-1.