Abstract Outcomes for high-grade pediatric brain tumors are poor,butthere is optimism that chimeric antigen receptor (CAR) T cell therapycan improve prognosis. We present results from the firsttwo cohorts of a phase I clinical trial of IL13BBζ-CAR T cellsinfused weekly into the cerebral ventricles for children and young adults with recurrent or refractory high-grade neuromalignancies. Results Among the 18 patients (ependymoma =5, DIPG/DMG =9, pHGG=4) treated on trial, patients in cohort 2 (n = 15) received systemic lymphodepletion prior to first infusion; cohort 1 (n = 3) patients did not. The trial met its primary objectives of establishing feasibility, safety, and tolerability. There was one dose-limiting toxicity (Gr3 hypoxia, cohort 2). Secondary objectives included CAR T cell distribution and persistence in CSF and peripheral blood, response rates by RAPNO criteria, and overall survival. Patients received a median of 8 (range: 2-19) infusions. Common adverse events included headache, fever, and fatigue. Patients receiving lymphodepletion also experienced cytopenias. Two patients met protocol criteria for radiographic response, and half experienced radiographic decreases consistent with an anti-tumor response. Median overall survival from diagnosis for patients receiving lymphodepletion was 187m for patients with ependymoma, 20.5m for patients with midline glioma, and 30m for other patients. Median overall survival was 36m from diagnosis for patients not receiving lymphodepletion. Importantly, patients who did not receive lymphodepletion developed anti-CAR humoral and cellular immune responses detectable in the CSF and peripheral blood, and patients receiving lymphodepletion had higher numbers of CAR+T cells detected in CSF over the course of therapy. Conclusions This study demonstrates the safety, tolerability, and biological activity of locoregionally-delivered IL13BBζ-CAR T cells for children and young adults with neuromalignancies. Moreover, we show anti-CAR immune responses in patients not receiving lymphodepletion, but not in patients receiving systemic lymphodepletion. ClinicalTrials.gov:NCT04510051.

Chimeric antigen receptor CAR T cell therapy faces notable limitations in treatment of solid tumors. The suppressive tumor microenvironment TME, characterized by complex interactions among immune and stromal cells, is gaining recognition in conferring resistance to CAR T cell therapy. Despite the abundance and diversity of macrophages in the TME, their intricate involvement in modulating responses to CAR T cell therapies remains poorly understood. Here, we conducted single-cell RNA sequencing scRNA seq on tumors from 41 glioma patients undergoing IL13Ra2-targeted CAR T cell therapy, identifying elevated suppressive SPP1 signatures predominantly in macrophages from patients who were resistant to treatment. Further integrative scRNA seq analysis of high-grade gliomas as well as an interferon-signaling deficient syngeneic mouse model both resistant to CAR T therapy demonstrated the role of congruent suppressive pathways in mediating resistance to CAR T cells and a dominant role for SPP1+ macrophages. SPP1 blockade with an anti-SPP1 antibody abrogates the suppressive TME effects and substantially prolongs survival in IFN signaling-deficient and glioma syngeneic mouse models resistant to CAR T cell therapy. These findings illuminate the role of SPP1+ macrophages in fueling a suppressive TME and driving solid tumor resistance to CAR cell therapies. Targeting SPP1 may serve as a universal strategy to reprogram immune dynamics in solid tumors mitigating resistance to CAR T therapies.

Pediatric and adult brain tumors, including glioblastoma, astrocytoma, ependymoma, and medulloblastoma, are associated with poor prognosis and limited treatment options. Standard of care (SOC) therapies such as surgery, radiation, and chemotherapy often fail to control tumor progression and are associated with severe toxicity. The lack of personalized therapeutic approaches underscores the urgent need for clinically relevant human models that reflect tumor heterogeneity and enable testing of novel therapies, including immunotherapies. We analyzed a panel of ten low-passage patient-derived brain tumor (PBT) lines and matched patient-derived xenograft (PDX) models from surgical specimens. Tumors were characterized using single-cell and bulk RNA sequencing, DNA exome analysis, multiparameter flow cytometry, and immunohistochemistry. Freshly dispersed tumor (FDT) samples and derived PBT lines were compared to assess fidelity in maintaining key molecular and cellular features. FDT-matched PBT lines were successfully established in approximately 70% of cases and preserved patient-specific mutations (IDH1, MGMT, p53, PTEN) and expression profiles of key therapeutic targets including IL13Rα2, EGFR, HER2, WNT1, JAK1/2, and NOTCH1-4. Correlation analysis demonstrated significant similarity between FDT and PBT profiles (R = 0.31, p = 0.018). Notably, PBTs and PDX models retained heterogeneous and non-clonal populations, enabling the identification of therapy-resistant subclones under in vitro selection pressures. These features support their use in evaluating used for the optimization of immunotherapy and other therapies for patients with brain tumors. We report the generation of clinically relevant PBT and PDX models that recapitulate the molecular and histopathological features of primary brain tumors. These models provide a powerful platform for dissecting therapeutic response and resistance, and for advancing precision medicine approaches. Our work lays the foundation for optimizing immunotherapy and next-generation treatments for both pediatric and adult brain tumor patients.

Tumor heterogeneity in glioblastoma (GBM) remains a great challenge for chimeric antigen receptor (CAR) T cell immunotherapy, as antigen-negative tumor populations can evade targeted destruction and drive recurrence. Epidermal growth factor receptor (EGFR) and interleukin-13 receptor alpha 2 (IL13Rα2) are promising targets in GBM. While murine mAb 806-derived CAR T cells effectively target overexpressed wild-type EGFR (EGFRwt) and the mutant EGFRvIII variant-expressing cells, repeated administration of this murine antibody-derived therapy risks immunogenicity. To address tumor escape and immunogenicity concerns in the setting of monovalent CAR T cell therapy, we have developed and optimized a bispecific CAR T cell incorporating both 806-derived humanized anti-EGFR single-chain variable fragment (scFv) and human IL13 targeting IL13Rα2. In this study, humanized anti-EGFR Fabs with binding affinities comparable to murine 806 were generated. Anti-EGFR CARs were constructed from these humanized scFvs and CAR T cells function were analyzed. The resulting humanized CAR T (6T) showed reduced cytotoxic activity against EGFRwt-low-expressing cells and eliminated patient-derived GBM xenografts in immunodeficient mice as efficient as 806-CAR T cells. Several bispecific CAR T cells incorporating humanized 6T scFv and human IL13 in one CAR construct were then engineered and tested. To model GBM heterogeneity, a patient-derived GBM cell line PBT206 was transduced to stably express either IL13Rα2 or EGFRvIII and mixed equally. Notably, in vitro cytotoxicity assays using this tumor mixture showed that a bispecific CAR construct (T4) efficiently eliminated both tumor cells, with comparable efficacy to the combination therapy comprised of both monovalent CAR T cells specific for either EGFR or IL13Rα2. Ongoing studies will compare the in vivo antitumor efficacy of the T4 bispecific CAR T cells with monovalent and combination CAR T treatments across multiple mouse models. Overall, this bispecific CAR T cells offer a potential therapeutic strategy to overcome antigen heterogeneity and immunogenicity in GBM.

Chronic graft-versus-host disease (cGVHD) remains the leading cause of late nonrelapse mortality and morbidities affecting quality of life (QoL) in patients who received allogeneic hematopoietic cell transplantation (HCT). While currently available drugs for cGVHD treatment achieve high overall response rates (ORR) and can improve QoL, failure-free survival (FFS) remains suboptimal, with high risk of infections. To address this unmet need, we developed a novel cellular therapy based on engineering regulatory T cells, isolated from HCT donors, expressing a CD6-targeting chimeric antigen receptor (CAR) with a novel CTLA-4 intracellular domain (CD6-CAR Tregs). In vitro, CD6-CAR Tregs immunomodulate T-cell effector function, and in vivo, prevented GVHD onset in a NSG mice xenograft model adoptively transferred with human PBMCs. A GMP-compliant manufacturing platform was developed and initiated a phase I clinical trial to evaluate the safety and feasibility of donor-derived CD6-CAR Tregs in patients with steroid-refractory or dependent cGVHD (NCT05993611). This trial employed a standard 3+3 dose-escalation design including 0.1, 0.3, 1.0, and 3.0 x106 CAR-Tregs/kg. To our knowledge, this represents the first clinical investigation of a genetically engineered human CAR-Treg product. Sixpatients have been treated in the study, assessable for the primary endpoint of safety based on dose limiting toxicity (DLT) and adverse event (AE) evaluation within the first 28 days. Three patients are evaluable for ORR at 9 months, and 4 patients at 6 months. The median duration of follow-up for patients is 6 months (range: 1-11). The median age of patients is 59 years (range 18-64). All received peripheral blood stem cell graft (PBSC) graft from a matched sibling donor, 4 female, 2 male, median age 56.2 years (range 25-67). HCT followed the treatment for ALL (n=3), MPN (n=2), or AML (n=1) using myeloablative (n=4) or reduced intensity regimen (n=2), and all patients received calcineurin inhibitor-based prophylaxis. The median time to cGVHD onset was 11.5 months (range 6-36) and time from cGVHD and CD6-CAR Tregs was 7.1 years (range 1.4-10.8). Median prior lines of therapy was 3.5 (range 2-7), including belumosudil (n=5) and ruxolitinib (n=4). Scleroderma was the most common organ involvement (n=5), followed by oral (n=4), ocular (n=3), joint fascia (n=2), and lung (n=1), with most patients (n=4) having multiorgan (≥ 3 organ) involvement. CD6-CAR Tregs were successfully manufactured, cryopreserved, and infused without prior lymphodepleting chemotherapy. Three patients at DL1 and DL2 received 6.4 (+/-0.2) and 24.1 (+/-7.1) x106 CAR+ cells, respectively. One patient in DL1 received a second infusion at DL2, 5 months after DL1, based on the initial safety/efficacy, seeking to consolidate a clinical response. No CRS, ICANS or infections appeared during the DLT period. The most common AEs during the DLT period were: cytopenias (n=9; neutropenia=4, lymphopenia=1, low platelets=3, anemia=1), sinus tachycardia (n=5), dyselectrolytemia (n=3), hypertension (HTN, n=1), and fatigue (n=2). Grade 3 AEs were noted in two patients (lymphopenia, HTN) without DLT or Grade 4/5 AEs. No growth factor or RBC transfusion was required, and one patient received a platelet transfusion for a count of 43,000/mm3. All patients with a 3-month or longer follow-up (n=5) experienced partial response (PR) based on NIH consensus criteria with continued FFS, including 3 patients with a 9-month follow-up associated with an improvement in patient-reported outcomes. CD6-CAR Tregs were at the peak levels within one month after infusion (DL1, 1856 (+/-1511) vector copy#/ml blood). Changes in serum cytokines were variable between patients (DL1), and IL-4 elevation ≥10-fold above baseline was the only finding commonly detected in two participants. One patient showed a ≥10-fold decrease from baseline in IL-1RA, MIG, and IL-8. T-cell depletion in PBMC was not detected at DL1. Evidence suggests that CD6-CAR Tregs persist up to two months after infusion, as observed in a skin biopsy from one patient. In summary, CD6-CAR Tregs can be successfully manufactured and administered to patients without significant toxicity, including CRS or ICANS. No hematologic malignancy was observed, and salvage therapy for GVHD has not been required. These early clinical findings support CD6-CAR Tregs as a promising cellular therapy with the potential to re-educate an imbalanced immune system in cGVHD.

BACKGROUND:Therapeutic resistance in glioblastoma (GBM) is multifactorial and results from genetic heterogeneity, the immunoprivileged localization, and the potently tolerogenic microenvironment. Signal transducer and activator of transcription 3 (STAT3) plays a key role in both glioma cell survival and immune evasion, reinforcing GBM resistance. METHODS:Here, we describe a new cell-selective and double-stranded STAT3 antisense oligonucleotide (CpG-STAT3dsASO) for targeting human/mouse glioma cells and GAMs but not T cells. The oligonucleotide safety and efficacy against orthotopic GBM was assessed in immunocompetent or immunodeficient mice. RESULTS:CpG-STAT3dsASO injected intracranially/intratumorally was well-tolerated and reduced progression of human U251 GBM xenotransplants and mouse GL261 or neural cell-derived QPP8 gliomas. Unlike the single-stranded oligonucleotide, local CpG-STAT3dsASO administration did not trigger type-I IFN-dependent neurotoxicities in immunocompetent mice within the therapeutic dose range. CpG-STAT3dsASO activated intratumoral GAMs, such as dendritic cells, macrophages and microglia, thereby expanding CD4+ Th1 cells while reducing TREG numbers. CpG-STAT3dsASO monotherapy did not have curative effects as it led to recruitment of only limited numbers of mostly exhausted effector CD8+ T cells. However, when combined with systemic PD1 inhibition, CpG-STAT3dsASO/anti-PD1 treatments caused regression of GL261 as well as immunotherapy-resistant QPP8 gliomas and resulted in long-term survival of the majority of mice. The combination treatment boosted CD8+ effector T-cell activity, while promoting their intratumoral interaction with activated CD4+ Th1 cells and activated macrophages as indicated by spatial transcriptomics. CONCLUSIONS:Our results suggest rationale for GBM immunotherapy using CpG-STAT3dsASO to disrupt GAMs-dependent immune evasion, thereby restoring sensitivity to PD1 blockade and facilitating T-cell-mediated antitumor immune responses.

We aim to develop clinically relevant human tumor organoid models of glioblastoma for optimizing chimeric antigen receptor (CAR) T cell therapy. We therefore have developed three-dimensional (3D) organoid cultures derived from patient tumors to preserve spatial cell-cell and cell-matrix interactions, offering advantages over traditional two-dimensional (2D) culture models. Although Matrigel is commonly used in 3D cultures, its batch variation, xenogeneic origin and non-brain relevant components raise concerns for modeling brain tumors and evaluating CAR T therapies. To address this, we evaluated brain-relevant hydrogels that contain hyaluronic acid, a major component of brain tumor extracellular matrix. We tested a panel of conditions and found one promising candidate, a mixture of PhoGel and PhotoHA (termed PGPH). PGPH allowed reliable tumor sphere migration, stable droplet formation, efficient sample recovery, and cost-effectiveness. Comparable to Matrigel, PGPH supports migration and growth of several patient-derived brain tumor spheroid cell lines. Transcriptomic analyses revealed similarities and differences between Matrigel and PGPH when culturing tumor cells or T cells. Notably, we demonstrated that T cells could migrate and kill organoid tumors in PGPH, and by incorporating soluble immune regulators, we were able to assess how these reconstructed tumor microenvironment (TME) compartments interact with CAR Ts and influence their killing potency. Moreover, PGPH enabled efficient and rapid tumor-derived explant culture (PDE) that closely resembled the native tumor ecosystem by preserving parental tumor heterogeneity and TME features, allowing for drug testing only 7 days after surgery. Lastly, we were able to evaluate CAR T cytotoxicity in PDE and are investigating the correlation of ex vivo CAR T responses with clinical responses in ongoing trials. In summary, this chemically defined, brain-relevant hydrogel offers opportunities to preserve the native or reconstructed TME, which could facilitate the translational success of CAR T treatments for brain tumors from bench to bedside.

IL13Rα2-targeted CAR-T cells are a promising therapy for aggressive primary brain tumors, but it remains unknown why survival benefit between patients is heterogenous. MR imaging of targeted tumors reveals changes in tumor physiology, but it remains unclear whether these changes are progression related or immune-modulated. To investigate, we performed image analysis on perfusion and diffusion MRI. A Phase 1 study of IL13Rα2-targeted CAR-T cells was performed at City of Hope National Medical center. Patients were imaged using diffusion-weighted MRI, and with DCE-MRI at two timepoints: 1) post-resection and prior to CAR-infusion, and 2) after receiving 3 infusions of CARs. The patient cohort consisted of 39 patients, suffering from primary recurrent GBM, astrocytoma, oligodendroma, or ependymoma, with 16 female and 28 male individuals. The apparent diffusion coefficient of water (ADC) was measured with diffusion-weighted MRI, and the interstitial fluid flow velocity (IFF) was measured using DCE-MRI. Increasing tumor density, and decreasing tumor IFF heterogeneity were associated with decreased progression-free survival. Patients with less-dense tumors after 3 CAR-T infusions had increased median progression-free survival of 5 months (25th percentile ADC > 535 E-6 mm2/sec, log-rank p < 1E-4), and patients with higher IFF heterogeneity had increased progression-free survival of 6 months (std(IFF) > 0.06 mm2/s, log-rank p <1E-4). IFF and ADC 25th percentile after CAR-T therapy were not significantly correlated (Pearson r = 0.30, p > 0.05). CAR-T therapy success may be determined as early as 1 month after beginning CAR-T therapy in patients with aggressive primary brain tumors. Standard medical imaging, such as diffusion-weighted and DCE-MRI may provide evidence that therapy is successful. Conversely, these same biomarkers can be an early warning that treatment is ineffective, and may allow for better decision making for future participation in the trial, or if modifications need to be made to better administer CAR-T cells to these patients.

Leptomeningeal disease (LD) from IL13Rα2-positive brain cancer is associated with poor prognosis, with current therapies offering limited efficacy and control. IL13Rα2-targeted CAR-T cells represent a promising therapeutic strategy, potentially overcoming the barriers of the central nervous system. All participants in this phase 1 study (NCT04661384) had IL13Rα2+ CNS tumors with leptomeningeal dissemination and received four serial intrathecal infusions of IL13Rα2-directed CAR T cells. One participant received a reduced dose (10M, 50M, 50M, 50M) due to due to low Tn/mem yield from the apheresis, and thus was not included for response assessment. Arm 1 (glioblastoma, n=4) had a median survival of 8.7 months, with one subject achieving stable disease. Arm 2 (ependymoma/ medulloblastoma, n=4) had a median survival of 6.5 months with two subjects still surviving after after 18 months, and two subjects achieving stable disease. Dynamic contrast-enhanced MRI was used to monitor tumor perfusion and vascularity, with all patients receiving imaging ~1 month after beginning treatment. In non-survivors, venous blood return rate (kep) was positively associated with survival time (Pearson r = 0.85, p<0.05, n=6), and tumor blood volume fraction (vp)) was negatively associated with survival (Pearson r = -0.94, p < 0.01, n=6).Imaging demonstrates surviving patients in Arm 2 had lower blood volume in tracked brain metastases (vp,surviving = 0.26+/-0.04 n = 2, vp,non-surviving = 0.44+/-0.09 n=2, mean +/- std) than non-survivors in Arm 2. Initial results from this Phase I trial show promise for patients with metastatic ependymoma. Treatments were well-tolerated with expected SAEs. DCE-MRI analysis may prove as a useful tool for prognosis for patients with LD while receiving IL13Rα2-targted CARs, indicating that blood plasma fraction and venous return rate of leptomeningeal lesions may be biomarkers of extended survival.

Background Glioblastoma (GBM) is the most aggressive form of glioma with a median survival rate of less than two years. Despite aggressive standard treatments, GBM remains uniformly fatal with a poor prognosis. Achieving responsiveness of GBMs to chimeric antigen receptor (CAR) T cell therapy has been a significant challenge due to the heterogeneity and evasive mechanisms employed by solid tumors, particularly GBM, to resist therapy. Recent studies have highlighted the substantial role of cancer-associated fibroblasts (CAFs) in GBM invasion by depositing various collagen subunits, including COL1A1, COL1A2, COL5A1, COL5A2, and COL8A1, within the extracellular matrix (ECM). This collagen deposition leads to increased ECM stiffness, modulating the tumor microenvironment, inducing immune suppression, and hindering T cell trafficking, ultimately worsening clinical outcomes and patient survival. Additionally, COL1A1 has been implicated in promoting tumor aggressiveness, particularly in GBM cases with wildtype isocitrate dehydrogenase (IDH-wt) status and a poor prognosis. Based on these findings, we investigated whether the expression level of COL1A1 could influence the response to CAR T cell therapy. Methods We performed immunofluorescent staining on tissue biopsies obtained from GBM patients to identify a subpopulation of CAFs expressing ACTA2 (alpha smooth muscle actin), PDGFRβ, and COL1A1. Confocal microscopy was used to visualize the stained CAFs. ACTA2 and PDGFRβ serve as markers for a specific subset of CAFs involved in the epithelial-to-mesenchymal transition (EMT), a process associated with immune suppression and tumor growth. The presence of these CAF subpopulations was previously confirmed in GBM patients enrolled in an IL13Ra2 targeted CAR T cell trial through single-cell RNA sequencing. Subsequently, we compared our immunostaining images with flow cytometry data obtained from cerebrospinal fluid (CSF) or tumor fragment (TF) biopsies collected from the same patients after CAR T cell therapy. Results Our preliminary analysis revealed an increase in the number of infiltrating CD3+ cells from baseline, particularly within the effector CD8+ subpopulation (characterized by CD27+CD28+ expression), during CAR T cell therapy in patients with low COL1A1 expression. Furthermore, FACS analysis demonstrated a significant increase in the proportion of CD8+ T cells relative to CD4+ T cells in these patients. Conclusions Ongoing experiments are currently investigating the distribution of CAF subpopulations producing COL1A1 and their correlation with the response to CAR T cell therapy.

Figure S1. Increased TLR9 expression in induced glioma spheres. Figure S2. TLR9 and STAT3 form a feed-forward loop in GSCs. Figure S3. Stat3 silencing by local CpG-Stat3siRNA delivery inhibits tumor growth. Figure S4. Stat3 silencing by local CpG-Stat3siRNA delivery inhibits GSCs. Figure S5. Targeting Brain tumors systemically with CpG-siRNA reaches the tumor site.

Introduction: In patients with severe hemophilia A, adeno-associated virus (AAV)-mediated factor VIII (FVIII) gene therapy is associated with reduced bleeding compared with FVIII prophylaxis. AAV5-FVIII gene therapy is approved in the U.S. and conditionally approved in Europe for the treatment of adults with severe hemophilia A. Gene therapy during childhood could potentially prevent the development of arthropathy and improve quality of life. However, there are unknowns regarding long-term safety and whether the predominant episomal nature of AAV persistence would lead to loss of transgene expression with hepatocyte divisions and liver expansion during childhood. The inflammatory response to AAV gene therapy in children is also largely uncharacterized. We have previously reported the treatment of five neonatal or infant dogs with a single vector infusion of a codon-optimized AAV5-B-domain deleted canine FVIII (cFVIII) construct with a hybrid liver promoter (AAV5-cFVIII). Dogs treated at 2 weeks of age demonstrated improved whole blood clot time (WBCT) over a 6-month period despite minimal cFVIII expression (<3%). Dogs treated at 2 months of age demonstrated stable cFVIII expression measured over a 6-month period and decreased WBCT. Aims: To describe the early inflammatory response to AAV5-cFVIII in infant and neonatal dogs, and to provide an update on the safety and efficacy of AAV5-cFVIII in these animals 12-16 months post-treatment. Methods: Hemophilia A dogs were treated with a single vector infusion (dose=2.0e14 vg/kg) of AAV5-cFVIII at 2 weeks (n=2) or 2 months (n=3) of age. Epigenomic data suggests that 2 months of age in dogs is equivalent to 9 months in humans. Liver volume was measured by magnetic resonance imaging at baseline, 3, 6, and 12 months post-AAV5-cFVIII treatment. Plasma cytokine levels were measured using a cytokine array pre-treatment (day 0) and post-treatment (days 2-28). Percutaneous liver biopsy samples were obtained at 2 weeks, 6 months, and 12 months post-treatment. Results: Cytokine array analyses of samples demonstrated that one dog who received AAV5-cFVIII at 2 weeks experienced transient post-treatment plasma elevation of IL-6, KC-like, MCP-1, TGFβ-1, and TGFβ-2 that returned to baseline levels by day 18. Two dogs treated at 2 months experienced a transient increase in GMC-SF, IL-2, IL-6, IL-7, IL-15, and IL-18 that returned to baseline levels by day 21. No changes in levels of IFNγ, IL-8, IL-10, IP-10, TNFα, or TGFβ-3 were observed. Changes in plasma cytokine levels were not associated with an elevation in plasma alanine transaminase levels. Hemophilia A dogs treated with AAV5-cFVIII at 2 weeks of age maintained an improved WBCT 12 to 16 months post-treatment, although cFVIII expression as measured by one-stage (OSA) and chromogenic substrate assays (CSA) remained minimal (<3%). Body weight and liver volume increased by 10.6-fold (measured at 16 months) and 12-fold (measured at 12 months) respectively. Dogs treated with AAV5-cFVIII at 2 months of age demonstrated sustained cFVIII expression by OSA and CSA after 12 to 16 months (8.9-13.39% measured at 16 months by OSA), despite a 3.2-fold increase in body weight (measured at 16 months), and a 3.9-fold increase in liver volume (measured at 12 months). This was associated with improved WBCT observed over the 12 to 16-month period for all three dogs. Overall, a reduction in bleeding events (BEs) was observed in all dogs. No BEs were recorded prior to AAV5-cFVIII treatment for dogs treated at 2 weeks of age. During the 16 months post-treatment, one dog experienced two spontaneous BEs, while no BEs were observed for the second dog. The dogs treated at 2 months experienced 4 (3 spontaneous and 1 traumatic) BEs prior to AAV5-cFVIII treatment, and 1 traumatic BE 48 hours post-treatment prior to the detection of FVIII expression. For the remaining 16 months post-treatment, no BEs were observed in this group. Conclusions: Treatment using AAV5-cFVIII at 2 months of age in a hemophilia A dog model resulted in stable FVIII expression into adult life despite significant liver expansion. For some dogs, AAV-cFVIII treatment resulted in an early transient increase in plasma proinflammatory cytokine levels with no evidence of transaminitis that resolved within three weeks.Further studies on liver biopsy samples are ongoing to evaluate cellular implications, vector genome distribution, and mechanisms of AAV persistence.