Endothelial injury is a central driver of systemic IFN-β toxicity and is reversible through Jak inhibition Authors: Lianwen Zhang, Linh Pham, Michael B Steele, Nathan Jenks, Rebecca Nace, Stephen J Russell, Kah-Whye Peng Published: 2026 ABSTRACT AbstractOncolytic viruses (OVs) encoding interferon beta (IFN-β) are under clinical evaluation. IFN-β is an attractive transgene due to its antiproliferative and immunomodulatory properties; however, the systemic effects of sustained IFN-β expression remain poorly defined. Here, we demonstrate that prolonged circulating IFN-β induces dose-dependent toxicity in murine models. Persistent IFN-β exposure led to elevated transaminases, thrombocytopenia, lymphopenia, and reduced hemoglobin. Histopathologic analysis revealed sinusoidal endothelial sloughing, microvascular coagulation with microthrombi, and hepatocellular vacuolation and degeneration, along with bone marrow hemorrhage and cell death. These toxicities were absent in type I interferon receptor knockout mice. Importantly, termination of IFN-β production via inducible caspase-9 in adeno-associated vector-transduced cells reversed toxicity and normalized serum transaminases. In Balb/c mice bearing MPC-11 myeloma tumors which are highly permissive to IFN-β-expressing OVs, treatment induced rapid tumor lysis, elevated plasma IFN-β, and mortality. Importantly, co-administration of ruxolitinib, a clinically approved JAK1/2 inhibitor, dampened IFN-β signaling and rescued mice from lethal toxicity. Collectively, these findings define the pathophysiological consequences of sustained systemic IFN-β exposure and identify ruxolitinib as a potential mitigation strategy to manage IFN-β-mediated toxicity during OV treatment.Keywords: interferon beta toxicity, endothelial damage, elevated transaminases, ruxolitinib, Jak/Stat inhibitor Introduction Interferons (IFNs) are key mediators of innate antiviral defense and immune regulation.1,2 Type I IFNs (IFN-α and IFN-β) bind to and signal through the heterodimeric interferon-α/β receptor (IFNAR), activating Janus kinase 1 (JAK1) and tyrosine kinase 2 (TYK2), which in turn phosphorylate signal transducers and activators of transcription (STAT) proteins.1 Activated STATs form the interferon-stimulated gene factor 3 (ISGF3) complex that translocates to the nucleus, where it binds interferon-stimulated response elements (ISREs) to initiate transcription of more than 500 IFN-inducible genes with antiviral, antiproliferative, and antiangiogenic functions.3,4,5,6 Through these pathways, type I IFNs not only serve as a first line of innate defense by restricting viral replication but also orchestrate adaptive immunity by promoting dendritic cell maturation and enhancing antigen presentation to T cells.2,7 Among type I interferons, IFN-β demonstrates superior tolerability and potency as an anticancer agent compared to IFN-α.8 However, in a study where IFN-β was given as a single intravenous bolus dose followed by daily subcutaneous injections for 28 days, the dose-limiting toxicities were gastrointestinal toxicity, elevations of aminotransferases and albuminuria.9 To overcome these systemic toxicities while maintaining antitumor efficacy, tumor-targeted delivery systems such as viral vectors encoding IFN-β have been developed to achieve high intratumoral concentrations with limited systemic exposure.10,11,12,13 We and others have previously demonstrated the antitumor activity of oncolytic vesicular stomatitis virus (VSV) engineered to express IFN-β and the sodium iodide symporter (NIS) reporter gene (VSV-IFNβ-NIS) in preclinical and clinical settings.14,15,16,17,18 Expression of IFN-β enhances tumor selectivity by inducing interferon-stimulated genes in normal cells, restricting VSV replication in healthy tissues while preserving oncolytic potency within tumors. In addition to VSV-induced direct oncolysis, expression of IFN-β contributes to antitumor immunity by promoting antigen cross-presentation and immune cell activation. Notably, VSV-IFNβ-NIS demonstrated durable clinical responses in patients with relapsed or refractory peripheral T cell lymphoma, including complete remissions lasting up to three years following a single intravenous infusion.19 Despite these encouraging results, our prior studies using the highly permissive MPC-11 murine plasmacytoma model revealed that robust intratumoral viral replication can trigger dose-limiting toxicities due to rapid tumor lysis and elevated circulating IFN-β levels.20 While tumor lysis syndrome (TLS) is clinically manageable, no effective therapeutic strategy currently exists to mitigate IFN-β-mediated systemic toxicity. In this study, we established murine models that recapitulate high, sustained IFN-β production to define the clinical features, pathophysiology, and mechanisms underlying IFN-β-associated toxicity. We further demonstrate that the toxicity is dependent on IFNAR signaling and is reversible upon termination of IFN-β expression. Building on these mechanistic insights, we evaluated the use of ruxolitinib, a clinically approved JAK1/2 inhibitor, as a pharmacologic intervention to attenuate IFN-β signaling and prevent lethal outcomes in mice treated with oncolytic VSV-IFNβ-NIS. Our findings reveal that JAK inhibition effectively rescues animals from systemic IFN-β toxicity without compromising antitumor efficacy, providing a translational framework for managing IFN-associated adverse effects in patients receiving IFN-armed oncolytic virus therapies. Results High systemic mIFNβ levels are lethal following hydrodynamic plasmid injection To directly assess the relationship between circulating IFN-β levels and toxicity, sixty-four Balb/c mice were given increasing doses of SB100 transposon plasmid DNA encoding mIFNβ under the control of the EF1α promoter by hydrodynamic tail vein injection (Figure 1A). Hydrodynamic injection efficiently transduced hepatocytes, resulting in intrahepatic gene expression as early as 3 h post-DNA injection (data not shown).21 All 64 mice were bled on day 5–6 post-hydrodynamic injection and plasma IFN-β was quantitated by ELISA. Increasing plasmid doses produced correspondingly higher circulating IFN-β concentrations in a clear dose-dependent manner (Figure 1B). Clinical chemistry and hematologic analyses revealed elevated aspartate aminotransferase (AST) and alanine aminotransferase (ALT) levels in animals with high systemic IFN-β (Figures 1C and 1D). Correlation analysis demonstrated strong, statistically significant associations between plasma IFN-β levels and hepatotoxicity markers: ALT (r = 0.7659, p < 0.0001) and AST (r = 0.741, p < 0.0001), indicating that increasing systemic IFN-β strongly correlates with liver injury (Figures 1C and 1D). Elevated transaminases and toxicity from high plasma mIFNβ levels after hydrodynamic injection of plasmid DNA encoding mIFNβ in Balb/c mice (A) Schematic representation of the sleeping beauty transposon system with a transposon plasmid pKT2/mIFNβ-CLP PGK-Luc encoding mIFNβ, and sleeping beauty SB100X transposase. (B) Plasma IFN-β levels from Balb/c mice [0.0037 μg DNA (n = 5), 0.011 μg DNA (n = 3), 0.033 μg DNA (n = 3), 0.05 μg DNA (n = 3), 0.1 μg DNA (n = 4), 0.2 μg DNA (n = 3), 0.4 μg DNA (n = 6), 0.8 μg DNA (n = 3), and 1.6 μg DNA (n = 1)] at 6–10 days after hydrodynamic injection of increasing amounts of plasmid DNA. Mean ± SEM. (C) AST and (D) ALT levels of individual mice with the corresponding plasma mIFNβ are plotted. The respective Pearson correlation factor and p values are shown. High systemic IFN-β is associated with elevated AST and ALT. IFN-β toxicity is associated with anemia, thrombocytopenia, and elevated transaminases To determine whether systemic IFN-β itself, rather than local hepatic expression, is responsible for toxicity, we established a second murine model in which mIFNβ is produced in nonhepatic tissue. Adenoassociated virus (AAV) serotype 1/2 (AAV1/2) vectors encoding mIFNβ under the cytomegalovirus (CMV) promoter were generated and administered intramuscularly into the quadriceps of Balb/c mice across a dose range (1010, 3 × 1010, 1011, 3 × 1011, and 1012 vector genomes [vg] per mouse). Mice were monitored for clinical signs and body weight changes for 60 days, and blood was collected at multiple time points for complete blood count (CBC), clinical chemistry, and plasma mIFNβ measurements. Mice treated with saline or low-to-mid doses of AAV1-mIFNβ (1010 or 3 × 1010 vg) showed no adverse clinical signs and gained weight over time (Figure 2A). In contrast, mice receiving ≥1011 vg exhibited rapid weight loss and clinical manifestations consistent with systemic toxicity, including inactivity, hypothermia, and moribund condition, necessitating humane euthanasia. Kaplan-Meier analysis revealed significantly reduced survival in the high-dose groups (3 × 1011 and 1012 vg) compared with saline controls (p = 0.0005 and p = 0.0027, respectively) and compared with the 1011 vg cohort (p = 0.0059 and p = 0.0351; Figure 2B). Adverse changes in liver enzymes and hematology parameters due to high mIFNβ levels in the blood in a AAV gene delivery model Balb/c mice were given one intramuscular injection of saline (n = 5) or AAV1-mIFNβ vector (1010 [n = 5], 3 × 1010 [n = 5], 1011 [n = 10], 3 × 1011 [n = 10], and 1012 [n = 5] vector genome [vg] per mouse). (A) Body weights and (B) Kaplan-Meier survival curves of mice over time. (C) Plasma mIFNβ levels, (D) ALT, (E) AST, (F) hemoglobin HGB, (G) platelet counts, and (H) lymphocyte counts of mice at day 7, 11, or 21 after vector administration. The groups are color coded: black, saline; red, 1010 vg/mouse; orange, 3 × 1010 vg/mouse; blue, 1011 vg/mouse; green, 3 × 1011 vg/mouse; purple, 1012 vg/mouse. The grading (Gr) reflects fold increase in transaminases per CTCAE v.4.03 criteria. ∗ = p < 0.05, ∗∗p < 0.01. Mean ± SEM. Plasma mIFNβ was undetectable in mice given saline or ≤3 × 1010 vg AAV1-mIFNβ (Figure 2C). In contrast, high circulating levels of mIFNβ were detected in plasma after intramuscular injection of AAV1-mIFNβ at 1011 vg or more per mouse (Figure 2C), which also correlated with adverse clinical signs. All mice injected with the top two doses of AAV1-mIFNβ vectors died or were euthanized by day 12 (Figure 2B). At necropsy, mice injected with 3 × 1011 vg showed median plasma mIFNβ concentrations of 8,626 pg/mL (Figure 2C). Consistent with systemic IFN-β toxicity, mice with elevated circulating mIFNβ demonstrated marked increases in AST and ALT (Figures 2D and 2E), along with hematologic abnormalities including anemia (Figure 2F), thrombocytopenia (Figure 2G), and lymphopenia (Figure 2H). Pathologies associated with high sustained levels of circulating mIFNβ To unravel the mechanism of IFN-β toxicity in these mice, we performed histopathologic evaluation of liver, spleen, kidney, lung, brain, and femur sections from mice receiving high-dose AAV1-mIFNβ. All slides were reviewed by a board-certified veterinary pathologist. At necropsy, livers from high-dose animals appeared notably pale, consistent with biochemical evidence of hepatocellular injury as indicated by the clinical chemistry value of elevated transaminases (Figure 2). Representative photographs of H&E stained liver sections showed significant hepatocyte vacuolation and degeneration (Figures 3B and 3C). A prominent and recurring finding was the presence of intravascular cellular aggregates within hepatic blood vessels. Importantly, the endothelial lining was disrupted, characterized by sloughing of thin, elongated endothelial cells from the underlying basement membrane (Figures 3B and 3C). Several sections exhibited signs of microvascular coagulation, with erythrocyte aggregation suggestive of developing microthrombi. These histologic patterns align with the severe thrombocytopenia, lymphopenia, and anemia observed biochemically. Representative H&E stained sections from tissues harvested from AAV-mIFNβ treated mice that were euthanized due to adverse clinical signs (A, D, and G) Saline, (B, C, E, F, and H) 3 × 1011 vg AAV1-mIFNβ, (I) 1012 vg AAV-mIFNβ. (A–C) Liver, (D–F) spleen, (G–I) bone marrow of femur. Arrows in (B) and (C) point to sloughed off endothelial cells in the liver sinusoids, noting the microclots and cellular aggregation. In the spleen, high-circulating IFN-β resulted in increased apoptosis within the marginal zones of lymphoid follicles (Figures 3D–3F). Examination of bone marrow—which contains extensive sinusoidal vasculature—revealed large regions of necrosis and hemorrhage in animals receiving the top two vector doses (Figures 3G–3I). The presence of extravasated erythrocytes indicates compromised sinusoidal endothelial integrity and vascular leakage. It is apparent from these histological and biochemical findings that sustained systemic IFN-β induces widespread vascular injury, endothelial disruption, and hematopoietic damage, contributing to the multisystem toxicity observed in high-dose AAV1-mIFNβ-treated mice. Toxicity associated with high systemic mIFNβ levels is reversible To confirm that the observed toxicities were mediated through type I IFN signaling, we administered AAV1-mIFNβ to IFN-β-nonresponsive transgenic mice lacking the type I interferon receptor (IFNARko)22 and in C57/B6 mice as controls. C57/B6 mice (n = 6) or IFNARko transgenic mice (n = 10) received intramuscular injections of AAV1-mIFNβ while saline controls (n = 7) received saline only. High plasma IFN-β (2,000–18,000 pg/mL) was detected in all AAV-mIFNβ treated C57/B6 and IFNARko mice, with no detectable IFN-β in saline controls (Figure 4A). All C57B/6 mice developed signs of IFN-β toxicity and were euthanized by day 29 (Figure 4B). In contrast, all IFNARko mice, despite equivalent systemic IFN-β levels, survived to the end of the study without clinical abnormalities and continued to gain weight (Figure 4B; data not shown). Despite the high IFN-β levels, there was no significant elevations in AST/ALT levels of IFNARko mice compared to saline control animals (Figures 4C and 4D), and hematological parameters remained normal (data not shown). Lack of toxicity in the absence of IFNR signaling and toxicity is reversible after IFN-β production is terminated Transgenic mice knockout for type I IFN receptor (IFNRko) and C57/B6 mice received one intramuscular injection of saline or 2 × 1011 vg AAV1-mIFNβ and were monitored over time. (A) Plasma mIFNβ, (B) Kaplan-Meier survival curves, (C) ALT, (D) AST from saline (black, n = 6), or 2 × 1011 vg AAV1-mIFNβ-treated IFNRko mice (red, n = 10), and C57B/6 mice (blue, n = 7) were plotted at regular intervals. Line denotes mean. Half of the mice were bled for each time point for blood analysis. C57/B6 mice (n = 25) were given one intramuscular injection of AAV-mIFNβ-iCas9 and 14 days later, given one intraperitoneal dose of AP1903 a prodrug to activate iCas9 to initiate apoptosis of AAV-transduced muscle cells expressing IFN-β. (E) Plasma mIFNβ, (F) ALT, (G) AST, and (H) platelet counts in C57B/6 mice before (day 0) and after termination of mIFNβ production using AP1903. Mean ± SEM. To determine if IFN-β-mediated toxicity is reversible, we generated a bicistronic AAV1-mIFNβ-iCas9 vector encoding both mIFNβ and an inducible caspase-9 protein as a safety switch which allows us to terminate mIFNβ expression using the prodrug AP1903. Two mice treated with this vector exhibited high-circulating IFN-β (1,604 and 5,524 pg/mL) on day 9, accompanied by elevated AST/ALT and thrombocytopenia (Figures 4E–4H). Mice were then administered the dimerizing prodrug AP1903 on day 14 to activate iCas9 and terminate mIFNβ production. Within 3 days of AP1903 administration, plasma mIFNβ levels fell below the limit of detection (<15.6 pg/mL). This reduction was accompanied by rapid normalization of AST and ALT (Figures 4F and 4G) and recovery of platelet counts (Figure 4H) over the subsequent 3–5 days. Overall, we show here that IFN-β-mediated toxicity is IFNAR-dependent and importantly, is reversible upon cessation of IFN-β production. In addition, we tested if addition of ruxolitinib, a clinically approved Jak1 and Jak2 inhibitor (IC50 values for Jak 1 and Jak 2 of 3.3 nM and 2.8 nM, respectively) used in the treatment of myelodysplastic syndrome could mitigate toxicity in this model. In a small proof of concept study, we demonstrated that administering ruxolitinib to AAV-mIFNβ-treated mice from days 4–14 post-injection (10 days total) significantly prolonged survival, accompanied by a corresponding trend for reduced (although not significant) in elevation of liver enzymes (Figure S1). Ruxolitinib rescues mice from lethal TLS and IFN-β toxicity We previously reported that the MPC-11 murine plasmacytoma model is highly permissive to VSV-mIFNβ-NIS therapy.20 Rapid tumor lysis in this model results in release of progeny virus into the blood with viremia detectable at day 2 after virus infusion. This is accompanied by thrombocytopenia, lymphopenia, and elevated serum transaminases, and a significant increase in circulating IFN-β, rising rapidly to 10,000 pg/mL at day 2 after VSV-mIFNβ-NIS treatment before declining to approximately 300 pg/mL as the virally infected tumor cells are eliminated.20 Using this Balb/c mice implanted with the highly VSV permissive MPC-11 tumors, we explored if ruxolitinib could mitigate the lethal toxicity associated with TLS and IFN-β toxicity by inhibiting Jak/Stat signaling.23,24 Mice bearing established MPC-11 tumors received a single intravenous dose of VSV-mIFNβ-NIS (107 TCID50), followed by daily oral ruxolitinib for 10 days. As expected, 80% of saline-treated or ruxolitinib-alone mice were euthanized by day 12 due to progressive tumor growth (Figure 5A), confirming that ruxolitinib has no intrinsic antitumor activity against MPC-11 tumors (Figures 5B and 5C). In contrast, VSV-treated mice exhibited robust tumor control (Figures 5B and 5C), but all animals receiving virus alone deteriorated clinically and were euthanized by day 8 (Figure 5A). Strikingly, the addition of ruxolitinib completely rescued VSV-treated mice from early mortality due to TLS and/or IFN-β, and mice survived until they were euthanized for progressive tumor burden. Concomitant treatment with ruxolitinib protects VSV-IFNβ-NIS treated MPC-11 tumor-bearing Balb/c mice from toxicity (A) Kaplan-Meier survival, (B) individual tumor volumes, (C) predicted tumor volumes using mixed effects modeling, (D) plasma IFNβ, (E) ALT, (F) AST, and (G) platelet count of mice at scheduled bleeds on day 3 and day 14, or at necropsy (at various time points). Treatment groups are color coded, saline (black, n = 5), VSV alone (red, n = 10), ruxolitinib (gray, n = 5), and VSV plus ruxolitinib (blue, n = 10). One intravenous dose of VSV at 107 TCID50 was given per mouse, and/or ruxolitinib at 2 mg twice a day by oral gavage for 10 days, starting on the same day as VSV therapy. A separate cohort of mice was euthanized on day 3 for correlative studies. Serum levels of IFN-β were high in the VSV treated groups, reaching peak levels of 63,778 pg/mL (Figure 5D). The levels of virally encoded IFN-β were comparable between VSV and VSV plus ruxolitinib treated mice (Figures 5B and 5C). Importantly, ruxolitinib did not negate antitumor effects of VSV therapy or increase virus induced toxicity and no signs of neurovirulence were detected. Although both VSV-treated groups exhibited elevated AST and ALT and reduced platelet counts at day 3 (Figures 5E–5G), these abnormalities resolved by day 14. Tumor relapse eventually occurred in most mice, with dissemination to the liver and spleen, except for one long-term survivor. It is important to note that IFN-β plays a critical role in this Balb/c model, enhancing the antitumor efficacy of VSV-IFNβ-NIS. Complete blockade of IFN-β signaling using a neutralizing anti–IFNAR antibody significantly diminished therapeutic efficacy. In contrast, treatment with ruxolitinib improved tolerability while preserving substantial antitumor activity of the virus (Figure S2). Discussion Here, we describe the first murine model of lethal IFN-β toxicity using AAV-mediated mIFNβ gene transfer. Treated mice developed hematologic and biochemical abnormalities that closely mirror toxicities reported in relapsing-remitting multiple sclerosis (MS) patients receiving IFN-β therapy, including thrombocytopenia, lymphopenia, anemia, and elevations in aminotransferases. Mild, transient ALT/AST increases are common in MS patients and generally resolve even with continued dosing, though rare cases of severe liver injury and liver failure have been documented.25,26 We demonstrate a direct relationship between circulating IFN-β concentrations and AST/ALT elevation, firmly linking systemic IFN-β exposure to hepatocellular injury. Notably, even supra-physiologic IFN-β levels (∼10,000 pg/mL) were well tolerated in IFNAR-knockout mice, indicating that toxicity is strictly dependent on type I IFN receptor signaling. Importantly, IFN-β toxicity is reversible: termination of IFN-β production using an inducible caspase-9 safety switch rapidly normalized liver enzymes and hematologic parameters. Together with histopathologic findings, these results support a model in which sustained systemic IFN-β induces endothelial cell injury within liver sinusoids, leading to endothelial detachment, microvascular coagulation, platelet consumption, and secondary hepatocellular damage. Clinical observations in MS patients provide context for these findings. IFN-β therapy is associated with predictable early symptoms, such as fever, chills, tachycardia, malaise, and fatigue, which can be mitigated with symptomatic treatment.27 Leukopenia commonly develops within hours due to leukocyte redistribution, and thrombocytopenia occurs in up to 30%–50% of myeloma patients receiving IFN-β.27 While most hepatic enzyme elevations are mild and self-limited, rare but severe adverse events—including renal thrombotic microangiopathy, thrombotic thrombocytopenic purpura-like syndromes and multiorgan failure—have been reported.28,29,30,31,32 These events typically resolve with cessation of IFN-β therapy and steroid administration.30,33 The endothelial injury and microvascular dysfunction observed in our murine model are highly consistent with these clinical manifestations. Indeed, a prior study using transgenic mice expressing type I IFN in astrocytes demonstrated a dose-dependent microvascular pathology resembling thrombotic microangiopathy seen in MS patients receiving IFN-β.31,34 Together, these clinical and preclinical observations suggest that sustained systemic IFN-β triggers endothelial activation and loss of vascular integrity, initiating intravascular coagulation and propagating downstream hepatic and hematopoietic injury. The rapid recovery of liver function and platelet counts following termination of IFN-β production further supports the reversibility of this process. These mechanistic insights have direct implications for IFN-β-based cancer therapies, including oncolytic viruses engineered to express IFN-β to enhance tumor selectivity and immune activation.30,35,36,37,38 In the highly permissive MPC-11 plasmacytoma model, VSV-IFNβ-NIS induced rapid tumor lysis and robust IFN-β production, but mice succumbed to TLS and systemic IFN-β toxicity despite excellent tumor control. Remarkably, co-administration of the JAK1/2 inhibitor ruxolitinib prevented early mortality, preserved viral replication and antitumor efficacy, and mitigated IFN-β-mediated toxicity. Ruxolitinib did not alter circulating IFN-β levels but blocked downstream IFNAR/JAK/STAT signaling, providing a pharmacologically tractable strategy to widen the therapeutic window of IFN-β-armed virotherapies. This is particularly relevant for clinical translation, as ruxolitinib is already approved for myelofibrosis and polycythemia vera and has an extensive safety record in over 600-treated patients.39,40 Taken together, our results establish a mechanistic framework linking sustained systemic IFN-β exposure to endothelial injury, microvascular coagulation, and hepatic and hematopoietic dysfunction. They also demonstrate that this toxicity is fully reversible and can be effectively prevented through JAK inhibition. These findings support a strategy of integrating ruxolitinib or other JAK inhibitors with IFN-β-expressing oncolytic viruses to preserve antitumor activity while mitigating systemic toxicity. This approach is currently being evaluated in two ongoing phase 1 clinical trials of intravenous VSV-IFNβ-NIS in patients with relapsed or refractory hematologic malignancies (NCT03017820) and endometrial cancer (NCT03120624). Materials and methods Cells, plasmids, viruses, and vectors MPC-11 murine plasmacytoma cells (CCL-167, American Type Culture Collection, Manassas, VA) were maintained in DMEM supplemented with 10% fetal bovine serum. Plasmid DNA encoding murine IFN-β (mIFNβ) under the control of the alpha elongation factor-1 (EF1α) promoter was generated (Figure 1A) and used in hydrodynamic injection into mice.41 AAV vectors of serotype 1 encoding mIFNβ under the CMV promoter were generated by triple plasmid transfection into HEK293 cells and purified on a OptiPrep Density Gradient (D1556, Sigma, St. Louis, MO) as previously described. AAV vectors were characterized and titers measured by quantitative PCR for viral genomes using primers for the human growth hormone (hGH) gene in the vector construct. The Indiana strain of lab adapted VSV was genetically engineered to encode the mIFNβ cDNA downstream of VSV matrix (M) gene and the human sodium iodide symporter reporter gene (NIS) upstream of the VSV large (L) polymerase.14 The virus was propagated on baby hamster kidney (BHK) cells and viral titers were determined by TCID50 titration on BHK cells as previously described.14,17 In vivo experiments All animal procedures were approved by the Mayo Clinic Institutional Animal Care and Use Committee (IACUC). Female Balb/c mice (6–8 weeks old) received plasmid DNA via hydrodynamic tail vein injection, delivered in 2 mL of saline. For virotherapy experiments, Balb/c mice were implanted subcutaneously with 5 × 106 MPC-11 syngeneic plasmacytoma cells in the right flank. When tumors reached approximately 0.5 cm in diameter, animals received one intravenous injection of VSV-mIFNβ-NIS (107 TCID50) via tail vein injection. Ruxolitinib (S1378, Selleck Chemicals, Houston, TX) treatment began on the same day as VSV, given twice daily at 2 mg by oral gavage per mouse for a total of 10 days. An anti-murine IFN-α receptor blocking antibody (I-401, Leinco Technologies, Fenton, MO) was given intraperitoneally in a cohort of animals at a dose of 0.5 mg on day 0 and 0.25 mg on day 5 for a total of 2 doses. Animals were humanely euthanized per IACUC guidelines if they lost more than 20% of body weight, if tumor volume exceeded more than 10% of body weight, or if they became moribund or unable to access water or food. Blood analysis Blood was collected for clinical chemistry (200 μL into lithium heparin tubes, BD, 365971), CBC (100 μL in EDTA tubes, BD, 365973) or serum for IFN-β assays (400 μL in serum separator tube, BD, 365967). Blood chemistry was analyzed using the ABAXIS Piccolo Xpress (Union City, CA), and CBC analysis was performed on the ABAXIS VetScan HM5 hematology machines. Murine IFN-β was analyzed using VeriKine Mouse IFNβ ELISA kit (42400, PBL Assay Science, Piscataway, NJ). Statistical analysis Data handling, analysis, and graphic representation were performed using the GraphPad Prism program. Kaplan-Meier survival curves were generated, and survival differences were evaluated using the log rank test. Student’s t tests were used for pairwise comparison where indicated. To model tumor growth, a linear mixed-effects model was fitted to the log-transformed tumor volume data [log(volume +1)], using treatment group, study day, and their interaction as fixed effects. Random intercept and study-day slope terms were included to account for within-mouse correlations. Predicted tumor volumes over time were generated from this model for each treatment group. Tumor growth modeling was conducted using JMP Pro version 13 software (SAS Institute, Cary, NC). Data and code availability All data generated or analyzed during this study are included in the main text or supplemental information. Acknowledgments This work was funded by the NIH/NCI (R01CA175795), Mayo Clinic SPORE in Myeloma (P50CA186781), the David F. and Margaret T. Grohne Foundation, and the Mayo Foundation. We thank Dr. Ronald Marler (Mayo Clinic Arizona) for providing histopathology services. Author contributions L.Z., L.P., M.B.S., N.J., conducted the experiments; L.Z., L.P., and K.-W.P. analyzed the results and prepared the manuscript; S.J.R. and K.-W.P. directed the research. All authors reviewed the manuscript. Declaration of interests S.J.R. and K.-W.P. and the Mayo Clinic have a financial conflict of interest related to this research. This conflict is being managed by the Mayo Clinic Conflict of Interest Review Board in compliance with Mayo Clinic Conflict of Interest policies. Footnotes Supplemental information can be found online at https://doi.org/10.1016/j.omton.2026.201226. Supplemental information Document S1. Figures S1 and S2 Document S2. Article plus supplemental information References Associated Data Supplementary Materials Document S1. Figures S1 and S2 Document S2. Article plus supplemental information Data Availability Statement All data generated or analyzed during this study are included in the main text or supplemental information.