The lack of safe, minimally invasive strategies for targeted drug delivery to the spinal cord remains a major barrier to treating neurodegenerative disease and spinal injury. Intrathecally administered macromolecules, including gene therapy vectors, rarely penetrate spinal tissue at therapeutically relevant concentrations due to cellular barriers at the cerebrospinal fluid-spinal cord interface. Here, we demonstrate the application of laser-activated perfluorocarbon nanodroplets (PFCnDs) for nonsurgical, spatially controlled delivery from the subarachnoid space into spinal cord parenchyma. Following intrathecal injection in rats, transdermal laser irradiation produced a 7-fold increase in photoacoustic signal within the spinal cord parenchyma compared to nonirradiated controls, confirming spatially selective intraparenchymal nanodroplet delivery. Codelivery of 500 kDa FITC-dextran demonstrated approximately 1 mm penetration into the dorsal horn at irradiated sites, while nonirradiated regions showed only superficial perivascular accumulation. Neither nanodroplets alone nor laser irradiation alone produced intraparenchymal delivery, indicating that vaporization-induced cavitation is necessary for penetration. The PFCnDs (∼300 nm diameter) were engineered with a lipid shell, perfluorohexane core, and near-infrared absorbing dye for transdermal activation at 1064 nm. These results demonstrate that laser-activated nanodroplets can deliver macromolecular cargo into the spinal cord via lumbar puncture and transdermal irradiation, without surgical exposure.

Glioblastomas are aggressive, heterogeneous tumors that present significant challenges in both diagnosis and treatment. Despite advances in surgical resection, radiotherapy, and chemotherapy with temozolomide (TMZ), the prognosis for glioblastoma patients remains poor, largely due to tumor heterogeneity and resistance mechanisms, such as genetic mutations in DNA repair pathways. To address these specific heterogenous qualities of glioblastoma, single-cell RNA sequencing (scRNA-seq) has emerged as a powerful tool for characterizing glioblastoma tumors, enabling the identification of subpopulations that respond differently to treatment. However, utilizing the vast amount of data generated by scRNA-seq poses challenges in clinical applications. To overcome this challenge, computational models have been introduced to more effectively process patient scRNA-seq data into more digestible information for clinicians. More specifically, advanced deep learning approaches show promise for processing and analyzing patient scRNA-seq data, enhancing informed treatment approaches for highly heterogenous glioblastoma. This review aims to explain how scRNA-seq can be used to identify important areas of glioblastoma treatment resistance, evaluate current glioblastoma scRNA-seq-based deep learning models, and outline relevant training datasets to overcome patient scRNA-seq data availability limitations. Ultimately, these deep learning models can be utilized by researchers and clinicians to provide more informed and precise treatment to glioblastoma patients.

Schwannomatosis (SWN) is a rare genetic condition characterized by the development of benign schwannomas along peripheral and spinal nerves, and this tumor development often results in chronic peripheral pain and significant neurological deficits. Despite surgical resection being the current standard of care for SWN-related pain, scientists are searching for accurate preclinical models that can be used for drug efficacy testing, but existing models often fail to fully recapitulate the complex genotypic diversity and tumor-nerve interface pathology seen in patients. This review first summarizes the genotypic and phenotypic characteristics of SWN subtypes resulting from key mutations in NF2, SMARCB1, LZTR1, and other chromosome 22q-related genes. Then, it provides a comprehensive overview of current preclinical frameworks for schwannomatosis including cell lines, mouse xenografts, and genetically engineered mouse models. We critically evaluate the respective strengths and limitations of each framework in modeling disease pathogenesis. Specifically, this review examines how the conventional two-dimensional culture systems inadequately represent the three-dimensional tumor architecture and interactions found between schwannoma cells, nerve fibers, and surrounding structural components. Next, this review highlights emerging preclinical platforms such as patient-derived organoids, SWN-on-chip microfluidic systems, and more advanced microphysiological systems that integrate cellular compartments to better simulate the tumor microenvironment. Moreover, we analyze the inclusion of large animal models to improve the fidelity and clinical relevance of SWN investigation, particularly in their potential to recapitulate human-scale anatomy and immune responses, for future therapeutic testing and discovery. Current preclinical approaches for schwannomatosis rely on commercial and patient-derived cells and small-animal models, which we discuss in this review. The emergence of advanced platforms such as SWN-on-chip bioengineered systems and large animal models represents a critical evolution in preclinical modeling capabilities, offering more physiologically relevant frameworks for mechanistic investigation and translational therapeutic evaluations. Bridging the gap between these evolving preclinical systems and clinical outcomes will require continued refinement of models that faithfully reproduce the tumor-nerve interface pathology and genotype-phenotype relationships observed in patients.

Abstract Gliomas represent a diverse spectrum of central nervous system tumors, ranging from aggressive high-grade gliomas (HGG) to less aggressive low-grade gliomas (LGG). Our group previously developed a high-grade glioma model in the minipig spinal cord using PDGFB, HRAS, and TP53. We now report the development of a minipig low-grade glioma model. Lentiviral vectors gene expression of PDGFB, BRAF V600E, and TP53 were used to generate the low-grade glioma model. Disease progression was monitored through behavioral assessment, MRI imaging for lesion detection, and histopathological analysis. Unlike the previously characterized high-grade phenotype, these models demonstrated a distinct, less aggressive low-grade phenotype. Both groups developed spinal cord lesions with divergent histological features: Group 1 (PDGFB + shTP53) displayed diffusely infiltrative gliomas with uniform tumor cells, edema, and minimal necrosis, while Group 2 (PDGFB + BRAF V600E + shTP53) exhibited heterogeneous lesions combining diffuse gliomatous and spindle cell components with increased atypia, apoptosis, and greater axial involvement. These spindle cells appear to originate from myeloid rather than glial lineage. This work establishes the first minipig low-grade glioma model, providing a clinically relevant large animal system for investigating low-grade glioma biology and developing novel therapeutic interventions for spinal cord tumors. Citation Format: Kecheng Lei, Angeliki Mela, Thais Federici, Muhibullah S. Tora, Marybeth Yonk, Yuliya Lakhina, Brett Henshey, Melissa Danielle Babbitt, Roy Raheb Khelo, Jeffrey N. Bruce, Peter Canoll, Nicholas M. Boulis. Lentiviral vector induced modeling of low grade glioma in the minipig spinal cord [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 6075.

Recapitulation of the complex nerve-tumor-immune microenvironment is critical for understanding Peripheral Nerve Sheath Tumors (PNSTs). Conventional in vitro models and animal systems often struggle to model crucial intercellular and inter-organ communication with full physiological relevance, creating a bottleneck in discovering effective therapeutics. Organ-on-Chip (OoC) technologies offer a paradigm-shifting solution. These microengineered platforms precisely integrate human cell biology with controllable fluidic and mechanical cues to replicate tissue- and organ-level physiology, enabling real-time, quantitative monitoring of tissue dynamics. This review posits that OoC technology is poised to revolutionize PNST research. In this review, we summarize the development of advanced Tumor-Nerve-Immune-on-Chip systems that integrate patient-derived Schwann cells, immune components, and functional microvasculature. These systems will accurately model PNSTs. Adoption of OoC will enable a predictive and individualized experimental framework, accelerating therapeutic discovery and allowing for personalized drug prediction for these challenging diseases.

Schwannomatosis is a non-cancerous disorder causing peripheral nerve sheath tumors (schwannomas), often leading to chronic pain. It is linked to loss of SWI/SNF related, matrix associated, actin dependent regulator of chromatin subfamily B member 1 (SMARCB1) or leucine zipper-transcription regulator 1 (LTZR1) gene function, though some patients may initially show minor mutations or no clinical signs, resulting in misdiagnosis, missed assessments, increased recurrence risk, unawareness of malignancy and overlooked genetic counseling during pregnancy. The present study reports a patient with a sporadic synonymous mutation in the SMARCB1 gene [SNP c.1032 C>T (p.Gly344Gly {GGC>GGT}) in exon 8]. This patient, a 53-year-old female with an 8-year history of schwannomatosis, presented to the neurosurgical department for recurrent tumor removal. Tumor tissue was analyzed using immunohistochemistry, hematoxylin and eosin staining, and enzyme-linked immunosorbent assay. Initially, the clinical impact of the SMARCB1 mutation on schwannomatosis was unclear. However, resected schwannomas showed 10-60% mosaic loss of nuclear SMARCB1 protein, with protein assays confirming low SMARCB1 levels, particularly in the distal thigh schwannoma. This case highlights the tumorigenic potential of SMARCB1 single nucleotide polymorphisms, emphasizing the need for multimodal diagnosis, long-term follow-up, awareness of recurrence and malignancy, and timely surgical planning in schwannomatosis patients.

INTRODUCTION: Pediatric brainstem glioma is a common tumor with a poor prognosis. To develop a rat brainstem glioma model, we employed lentiviral gene transfer to deliver oncogene and tumor suppressor impacting the RTK/RAS/PI3K and p53 pathways. METHODS: Twenty-four 6-8 week-old Sprague Dawley rats received stereotactic injections of a combination of lentiviral vectors (PDGF-B, HRAS, shRNA-p53) and control virus in the pontine reticular nucleus. Animals were divided into groups as follows: the control group, the group exhibiting symptoms before neurologic deficit development, the group with a prominent neurological deficit, and the endpoint group. To confirm the tumor growth in each group, animals underwent endpoint MRI, H&E, and immunohistochemical analysis. Glutathione assay and protein carbonyl content assay determined the level of oxidative stress in the groups. RESULTS: All animals exhibited clinical symptoms of brainstem tumors within 21-36 days, including circling and gait instability. MRI scans at the endpoint revealed contrast-enhancing lesions in the pontine region. Immunohistochemistry analysis confirmed the absence of H3K27M and IDH mutations. There was a notable disparity in oxidative stress levels among the groups, correlating directly with the severity of tumor presentation. CONCLUSIONS: Our research has established a rat model of high-grade brainstem glioma induced by lentiviral vectors. Evaluation of various oxidative stress parameters validated the disease's advancement. This model shows promising potential as a platform for pediatric brainstem glioma antisense therapy.

Persons with neurofibromatosis type 1 (NF1) exhibit enhanced glucose metabolism, which is replicated in Nf1-mutant mice. Inflammatory macrophages invest NF1-associated tumors, and targeting macrophages appears efficacious in NF1 models. Inflammatory macrophages rely on glycolysis to generate ATP; thus, identifying whether neurofibromin, the protein encoded by NF1, controls glucose metabolism in macrophages is therapeutically compelling. Using neurofibromin-deficient macrophages and macrophage-specific Nf1-knockout mice, we demonstrate that neurofibromin complexes with glucose transporter-1 (GLUT1) to restrain its activity and that loss of neurofibromin permits Akt2 to facilitate GLUT1 translocation to the membrane. In turn, glucose internalization and glycolysis are upregulated and provoke reparative (MIL4) macrophages to undergo an inflammatory phenotypic switch. Inflammatory MLPSIFNγ macrophages and inflammatory-like MIL4 macrophages invest the perivascular stroma of tumors and induce pathologic angiogenesis in macrophage-specific Nf1-knockout mice. These studies identify a mechanism for the enhanced glycolysis associated with NF1 and provide a novel therapeutic target for NF1.

Atypical facial pain remain challenging to manage primarily due to the extensive history of previous surgeries in affected patients. Traditional methods often lead to prolonged ataxia and undesirable side effects. Neuromodulatory techniques offer promising alternatives, emphasizing reversibility and the ability to conduct trial lead placements before permanent implantation. Among these, spinal cord stimulation (SCS) has gained attention. We explore its application specifically to the trigeminal tract and nucleus caudalis. We reviewed 22 patients with refractory facial pain who underwent nucleus caudalis stimulation. Outcomes following the trial period (mean 23.5 days) and permanent stimulation were analyzed. 81% of patients had successful trial stimulation and received permanent implantation of the stimulation system in ante- or retrograde fashion. In the anterograde method, electrodes traverse under the C1 arch and through laminotomies in C2, terminating proximally near the foramen magnum. In the retrograde approach, they pass from the occpiput under both vertebral arches, terminating caudal to C2. The mean follow-up was 38.4 months (range: 6-60 months). Permanent implantation demonstrated a significant reduction in facial pain (mean VAS change from 8.8 to 3.5). Four (23.5%) patients demanded successful system revision, 75% of which were due to technical issues and one due to electrode displacement. The long-term treatment failure rate was 23.5%. One patient developed an epidural hematoma. Fisher’s exact test (p = 0.93) showed no significant association between the direction of lead placement and pain outcomes. Burst SCS effectively reduces atypical trigeminal neuropathic pain, thereby contributing to the decision-making process to alleviate neurological deficits compared to conventional procedures.

The blood-brain barrier (BBB), crucial for central nervous system (CNS) homeostasis, poses challenges for drug delivery in CNS diseases due to selective permeability. Because of this difficulty, there are limited treatments developed for CNS diseases. As a solution, computational models can be implemented in treatment development to enable rapid screening of drug permeability, saving time and resources. This study explores machine learning, deep learning, and transfer learning models to predict the BBB permeability of drug molecules, validated through an in vitro assay known as Parallel Artificial Membrane Permeability Assay-BBB (PAMPA-BBB). Using the Blood-Brain Barrier Database (B3DB) of ∼ 8,000 compounds of known BBB permeability, classification models including support vector machines (SVMs), deep neural networks (DNNs), direct message passing neural networks (D-MPNNs), and transfer learning with quantum chemical properties were developed. Experimental validation with 18 compounds from the Emory Enriched Bioactive Library (EEBL), a library containing 1,018 FDA-approved pharmacologically active compounds of known activity, highlighted PAMPA-BBB as a robust validation method. The SVM model with combined 2D RDKit and Morgan fingerprint molecular representation achieved high performance (accuracy: 89.08

BACKGROUND AND OBJECTIVES:Riluzole is the only treatment known to improve survival in amyotrophic lateral sclerosis (ALS) patients. However, its efficacy and dosing are limited by hepatic toxicity and interindividual pharmacokinetic variability. Recent experimental studies in hounds have shown that continuous intrathecal (IT) administration of riluzole is well tolerated and achieves significantly higher spinal cord tissue levels. We report the first 2 human ALS cases treated with IT riluzole.METHODS:A catheter was inserted into the lumbar cistern and advanced to the midcervical region under fluoroscopic guidance and connected to a subcutaneous pump. Therapy was initiated at 0.1 mg/h of riluzole. The infusion rate was gradually increased until it reached a maximum of 4.8 mg/d.RESULTS:The 2 patients tolerated dose escalation and treatment for over 2 years without apparent motor or sensory complications. Patients reported no asthenia, a central side effect often reported as a reason to abandon oral therapy.CONCLUSION:This is the first report of chronic IT riluzole infusion in humans at a dose found to be safe in canines. A phase 1 study is planned to establish the maximum tolerated human dose, followed by a randomized placebo-controlled trial to determine the safety and tolerability of IT riluzole in patients with ALS.

The management of spasticity poses a significant challenge for both physicians and patients. This condition, often characterized by increased muscle tone, clonus, and muscle spasms, can be painful, disrupt daily activities, and increases the risk of injuries. First-line treatment for spasticity includes oral medications, such as baclofen, but these drugs can have significant side effects due to inhibitory effects on neural circuits, such as drowsiness and dizziness. Surgical approaches can be applied to patients for whom oral medications fail. However, these treatments can have life-threatening side effects or are irreversible. An emerging technology, chemogenetics, has the potential to provide targeted relief of spasticity, combining a non-destructive surgical approach with the convenience of an oral medication. This Current Opinion describes the current treatment approaches for spasticity, the pathophysiology of this condition, the current state of chemogenetics, and how this technology may be applied to patients with poorly controlled spasticity. Although the use of this approach is at an early developmental stage, we believe that it shows great promise.

Spinal cord gliomas (SCG) are rare cancers, and relevant pharmaceutical research and treatments are limited. While drug repurposing of FDA-approved compounds has shown promise for orphan diseases, its potential for SCG remains uncertain. Our research team has recently developed an SCG model in minipigs, offering anatomical, immunological, developmental, and genetic similarities to humans, which makes it an ideal model for translational precision medicine. In this study, we used a novel drug repurposing strategy using 3D spheroids from our recently established large animal SCG model as a case study. First, we established a miniaturized 3D cell culture platform for ultra-high-throughput screening (uHTS). After screening the FDA-approved and Bioactive compound library with minipig SCG 3D spheroids, Clofoctol, showed potential as a glioma treatment by targeting genes in the p53 pathway. Evaluation in a xenograft model demonstrated Clofoctol's ability to suppress tumor growth over a 10-day treatment period, followed by immunohistochemistry and RNA-seq. These findings suggested Clofoctol is a promising therapeutic for glioma treatment. Thus, our study shows the feasibility of 3D spheroids uHTS to expedite orphan disease research, providing a strategy for future drug repurposing for rare cancer patients. Marybeth G. Yonk, Muhibullah S. Tora, Megan A. Lim, Thais Federici, Haian Fu, Yuhong Du, Bing Yao, Nicholas Boulis, Kecheng Lei. Drug repurposing via ultra-high throughput 3D spheroids screening in rare cancer treatment [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 521.

Neurofibromatosis type 1 (NF1) is an inherited genetic disease resulting from pathogenic mutations in NF1 that drive tumor formation along peripheral nerves, leading to many functional consequences. Tumor removal or treatment often results in regrowth and/or nerve damage. Addressing NF1 pathogenic variations at the cellular level through gene therapy holds great potential for long-term treatment of patients with NF1. Adeno-associated viruses (AAVs) are broadly used gene delivery vehicles for gene therapies because of their low pathogenicity, ability to transduce nondividing cells, and potential for long-term gene expression. This article explores the landscape of AAV-mediated gene delivery strategies for NF1, discusses the challenges of efficient delivery to relevant cell types, and highlights the progress in vector design strategies.

Abstract High-grade gliomas are the most common type of malignant intracranial tumor with dismal five-year prognostic outcomes despite developments in chemoradiotherapy and surgical intervention. Distinct oncogenes associated with tumorigenesis and progression have been identified, providing targets for potential therapeutic interventions and modeling efforts. In this study, we evaluated the efficacy of vectors targeting EGFR and PTEN, individually and in combination to produce high-grade brain gliomas. Four groups of rats underwent intracranial stereotaxic injection of lentiviral vector into the forceps minor of the corpus callosum. All animals were weighed and monitored for behavioral deficits three times per week for 90 days. Upon signs of clinical morbidity or at 90 days post-inoculation, tissues were harvested and analyzed using hematoxylin and eosin and immunohistochemical staining. Rats injected with the combination of the EGFR and the sh-PTEN vectors exhibited declines in weight, survival, and motor function compared to the control. The combination of the EGFR and sh-PTEN vectors was the only condition to induce tumorigenesis with histopathological confirmation of high-grade identity as confirmed by immunohistochemistry showing the most positivity for GFAP and SOX2 compared to the control. Significant increases in Ki-67 characterized combination tumors as highly proliferative. Increased oxidative stress and angiogenesis were supported by significant increases in 4HNE, SMA, and HIF-1α in the tumors compared to other treatment groups. Thus, the present model of high-grade rat brain glioma provides a potential model to assess pathogenic pathways and evaluate therapeutic strategies.

Peripheral nerve injuries (PNIs) are common and devastating. The current standard of care relies on the slow and inefficient process of nerve regeneration after surgical intervention. Electrical stimulation (ES) has been shown to both experimentally and clinically result in improved regeneration and functional recovery after PNI for motor and sensory neurons; however, its effects on sympathetic regeneration have never been studied. Sympathetic neurons are responsible for a myriad of homeostatic processes that include, but are not limited to, blood pressure, immune response, sweating, and the structural integrity of the neuromuscular junction. Almost one quarter of the axons in the sciatic nerve are from sympathetic neurons, and their importance in bodily homeostasis and the pathogenesis of neuropathic pain should not be underestimated. Therefore, as ES continues to make its way into patient care, it is not only important to understand its impact on all neuron subtypes, but also to ensure that potential adverse effects are minimized. This piece gives an overview of the effects of ES in animals models and in humans while offering a perspective on the potential effects of ES on sympathetic axon regeneration.