Bone remodeling is a dynamic process that responds to external mechanical loading, involving bone resorption by osteoclasts and subsequent synthesis by osteoblasts. Although numerical models such as the finite element method (FEM) can simulate the remodeling process at the macroscale, there is a significant lack of predictive models for the microscale morphological changes in trabecular bone, which directly affect its mechanical properties. In this study, we propose a novel deep learning framework based on conditional generative adversarial networks using the Pix2Pix architecture to predict the morphological evolution of trabecular bone. To generate training data, a large virtual trabecular bone structure is synthesized using a deep convolutional generative adversarial network (DCGAN). This virtual bone is then subjected to bone remodeling simulations and segmented into multiple statistical volume elements (SVEs). Our results demonstrate that the trained Pix2Pix model can predict the morphological changes of trabecular bone SVEs with a mean error of less than 2

BACKGROUND/OBJECTIVES:Stereotactic body radiation therapy (SBRT) provides improved pain response and local control for spinal metastases. However, management of local failure after initial SBRT is challenging. We report institutional outcomes, dosimetry, and toxicity for reSBRT following SBRT. METHODS:We retrospectively reviewed 61 lesions (55 patients) treated with reSBRT after prior SBRT. Both SBRT courses delivered a median dose of 27 Gy. Patients underwent clinical and radiological evaluation every three months. Toxicity was graded using CTCAE v5.0. Dosimetric parameters for the spinal cord (SC), cauda equina (CE), planning organ-at-risk volumes (PRV), and thecal sac were converted to equivalent dose in 2 Gy fractions (EQD2) using the linear-quadratic model (α/β = 2). RESULTS:Median follow-up was 10.3 months. Forty lesions (65%) were cervicothoracic and 21 (35%) were lumbosacral. One- and two-year overall survival (OS) were 45% and 29%, respectively, and one- and two-year local control (LC) were 89% and 88%, respectively. Gastrointestinal primary tumors were associated with inferior LC (HR 2.41, 95% CI 1.11-5.23, p = 0.026). Fifteen patients (27%) reported myelitis/neuropathic symptoms during follow-up; four (7%) developed new post-radiation myelitis or neuropathy (RMN) without radiologic progression. Five patients (9%) developed vertebral compression fractures (VCF). Cumulative EQD2 was not significantly associated with RMN (p = 0.344); all affected patients had thecal sac EQD2 > 95.5 Gy and relevant nerve roots EQD2 > 108 Gy. CONCLUSIONS:ReSBRT provided a favorable LC with acceptable toxicity. High cumulative dose to the thecal sac and nerve roots may contribute to neurologic toxicity as peripheral nerve injury.

Bone remodeling models are typically phenomenological or mechano-biological but often lack mechanisms to incorporate patient-specific data, limiting clinical use. We present a patient-specific multiscale framework that couples finite element (FE)-based shape optimization at the microscale with a mechano-biological model at the macroscale. The model predicts % bone mineral density (BMD) changes at the macroscale, which in turn drive microscale trabecular adaptation via % bone volume fraction (BV/TV) changes. Micro-QCT imaging data are used to train a DCGAN-based ReconGAN for virtual reconstruction of trabecular microstructures, from which FE models are generated. Apparent BMD changes predicted by the macroscale model guide the microscale shape optimization to simulate adaptation. The framework reproduces BMD losses of 9.8% (trabecular) and 4.9% (whole vertebra) over a 215-day spaceflight scenario, consistent with results from prolonged bed rest and controlled experimental datasets. In vertebral compression fracture simulations, it captures trabecular bone degeneration by reducing peak load from 3.532 to 3.280 kN and energy absorption from 0.243 to 0.218 J, and recovery restores close agreement to the original microstructure. These results demonstrate a path toward patient-specific simulation of bone remodeling and its mechanical consequences, with strong potential for treatment planning and assessment of skeletal interventions.

BACKGROUND:Lumbar disc degeneration is a leading cause of disability worldwide and a major contributor to low back pain. Biomechanical models have emerged as powerful tools to investigate the structural and mechanical changes associated with disc degeneration, offering insights that are difficult to obtain in vivo. METHODS:This narrative review examines the progression of biomechanical modeling efforts in lumbar disc degeneration over the past decade. We reviewed studies from 2013 to 2024, focusing on both finite element and musculoskeletal modeling approaches. Key developments in model personalization, tissue degeneration simulation, and validation techniques were analyzed to assess their contribution to understanding spinal mechanics in degenerative states. FINDINGS:Recent advancements have enabled more accurate representations of intervertebral disc pathology, incorporating subject-specific imaging, detailed tissue behavior, and complex loading scenarios. While finite element models have improved in capturing disc-level mechanical changes, musculoskeletal models have advanced in simulating whole-body dynamics and compensatory mechanisms. However, gaps remain in integrating these approaches and validating models against in vivo measurements. Few studies have directly linked modeling outcomes with clinical decision-making or patient-specific interventions. INTERPRETATION:Biomechanical models have significantly improved our understanding of lumbar disc degeneration and its mechanical consequences. Future efforts should prioritize multimodal validation, integration of imaging and motion capture data, and development of clinically actionable models. Advancing these tools may enable more personalized and predictive approaches to spine care and support novel therapeutic strategies.

Accurate representation of spinal geometry is necessary in biomechanical modelling to properly understand the function of the spine. The objective of this study was to create a large database of image-derived digital spine surface models for geometric studies and computational biomechanics investigation. Computed tomography scan data was acquired from 60 asymptomatic subjects (30 males and 30 females). Subjects ranged in age from 20 to 68 with equal numbers selected in each 10-year age bracket. Three-dimensional geometric surface models were generated for each subject's spine. A series of distance measures were also computed for each vertebral body to assess variation in the population. Geometric measures were able to show the significant variation in vertebrae size and shape within the subject population and as a function of vertebral level, gender, and age. This database enables the biomechanical evaluation of the variation in spinal loads that occurs within the population due to geometry.

BACKGROUND:Spine metastases often cause significant pain, instability, and/or neurological morbidity. Local control (LC) of spine metastases has been augmented with advances in systemic therapies, radiation, and surgical technique. Prior reports suggest an association between preoperative arterial embolization and improved LC and palliative pain control.OBJECTIVE:To further elucidate the role of neoadjuvant embolization on LC of spine metastases and the potential for improved pain control in patients receiving surgery and stereotactic body radiotherapy (SBRT).METHOD:A retrospective single-center review between 2012 and 2020 identified 117 patients with spinal metastases from various solid tumor malignancies managed with surgery and adjuvant SBRT with or without preoperative spinal arterial embolization. Demographic information, radiographic studies, treatment characteristics, Karnofsky Performance Score, Defensive Veterans Pain Rating Scale, and mean daily doses of analgesic medications were reviewed. LC was assessed using magnetic resonance imaging obtained at a median 3-month interval and defined as progression at the surgically treated vertebral level.RESULTS:Of 117 patients, 47 (40.2%) underwent preoperative embolization, followed by surgery and SBRT and 70 (59.8%) underwent surgery and SBRT alone. Within the embolization cohort, the median LC was 14.2 months compared with 6.3 months among the nonembolization cohort ( P = .0434). Receiver operating characteristic analysis suggests ≥82.5% embolization predicted significantly improved LC (area under the curve = 0.808; P < .0001). Defensive Veterans Pain Rating Scale mean and maximum scores significantly decreased immediately after embolization ( P < .001).CONCLUSION:Preoperative embolization was associated with improved LC and pain control suggesting a novel role for its use. Additional prospective study is warranted.

NECK HYPEREXTENSION is used commonly during surgical positioning to provide optimal anatomic positioning and exposure. Iatrogenic tetraplegia has been reported in patients with asymptomatic cervical stenosis after prolonged neck extension during cardiac and noncardiac surgeries. Standardized protocols to screen patients for underlying degenerative cervical disease, such as cervical spondylotic myelopathy or cervical stenosis, have been shown to alter perioperative decision-making. 1 Diaz A Chin C Burks SS et al. A retrospective pilot study for preoperative screening to prevent iatrogenic cervical spinal cord injury. Cureus. 2021; 13: e12550 PubMed Google Scholar Here, the authors present the case of a patient with a significant risk of intraoperative spinal cord injury during combined cardiac and vascular surgery, which was managed successfully without complication. They discuss the pathophysiology, prevention, and treatment of this devastating complication.

We present the application of ReconGAN, introduced in a previous study, for simulating the vertebroplasty (VP) operation and its impact on the fracture response of a vertebral body. ReconGAN consists of a Deep Convolutional Generative Adversarial Network (DCGAN) and a finite element based shape optimization algorithm to virtually reconstruct the trabecular bone microstructure. The VP procedure involves injecting shear-thinning liquid bone cement through a needle in the trabecular region to reinforce a diseased or fractured vertebra. To simulate this treatment modality, computational fluid dynamics (CFD) is employed to predict the morphology of the injected cement within the bone microstructure. A power-law equation is utilized to characterize the non-Newtonian shear-thinning behavior of the polymethyl methacrylate (PMMA) bone cement during injection simulations. The CFD model is coupled with the level-set method to simulate the motion of the interface separating bone cement and bone marrow. After predicting the cement morphology, a data co-registration algorithm is employed to transform the CFD model to a high-fidelity continuum damage mechanics (CDM) finite element model of the augmented vertebra for predicting the fracture response. A feasibility study is presented to demonstrate the ability of this CFD-CDM framework to investigate the effect of VP on the mechanical integrity of the vertebral body in a cancer patient with a lytic metastatic tumor.

Background: Adjuvant regional nodal irradiation (RNI) after breast conserving surgery or mastectomy is supported by clinical trials for patients with node-positive breast cancer. RNI results in increased radiation dose to organs-at-risk (OARs) such as the heart and lungs. While regular acquisition of both free breathing (FB) and deep inspiration breath hold (DIBH) scans has been widely adopted for left-sided breast cancers (LBCs) as a cardiac-sparing technique, DIBH scans are not routinely acquired for right-sided breast cancers (RBCs). Therefore, when OAR constraints cannot be met with 3D conformal radiation therapy (3DCRT) planning on the FB scan, the only option is intensity modulated radiation therapy (IMRT), with its inherent increased cost, resource utilization, and requirement for insurance authorization. Given these challenges, we have regularly acquired FB and DIBH scans for right-sided RNI since 2018. We hypothesized that acquisition of DIBH scans would result in a reduced need for IMRT and reduced dose to OARs. Methods: We retrospectively identified patients who were treated with right-sided RNI who had both FB and DIBH scans. All patients had target volumes (breast or chest wall and regional lymph nodes [undissected axillary, supraclavicular, and internal mammary nodes]) prospectively contoured on the FBCT scan based upon the RTOG Breast Atlas. This initiated a treatment planning algorithm that began with creating a FB 3DCRT plan and changed to DIBH 3DCRT then FB IMRT when OAR constraints could not be met while maintaining acceptable planning target volume (PTV) coverage. For patients who did not have contours available on the DIBH scan, the treating physician retrospectively completed the PTV contours. For each patient, three total plans were created for comparison using our institutional target coverage and OAR metrics: FB 3DCRT, FB IMRT, and DIBH 3DCRT. We compared PTV coverage and doses to multiple OARs including the contralateral breast, esophagus, heart, lungs (left, right, total lung dose), and liver. PTV coverage and OAR doses were evaluated by a one-way ANOVA followed by Bonferroni comparison. A p < 0.05 was considered statistically significant.Results: We identified 38 patients in whom FB and DIBH scans were acquired. Only 32% (N=12) were treated with the standard FB 3DCRT. Of the remaining 26 patients 73% (N=19) were treated DIBH 3DCRT, and only 27% (N=7) were treated with FB IMRT, resulting in a FB IMRT rate of 18% overall. Without DIBH scans, 68% (N=19) would have advanced to FB IMRT. Dosimetric comparison across these 38 patients (N=114 plans) demonstrated that DIBH 3DCRT had at least equivalent OAR metrics as compared to FB 3DCRT, with significant improvement in max heart dose (9.6 Gy vs. 14.9 Gy; p = 0.034), right lung V20 (32.1% vs 37.8%; p < 0.01), mean total lung dose (8.9 Gy vs. 10.5 Gy; p < 0.01), and mean liver dose (1.8 Gy vs. 4.0 Gy; p < 0.01). FB IMRT plans resulting in significantly lower right lung V20 (26.3% FB IMRT vs. 37.8% FB 3DCRT vs. 32.1% DIBH 3DCRT), but resulted in higher dose to the heart and contralateral breast: mean heart dose (2.2 Gy FB IMRT vs. 1.0 Gy FB 3DCRT vs. 0.9 Gy DIBH 3DCRT; p < 0.01), maximum heart dose (16.4 Gy FB IMRT vs. 14.9 Gy FB 3DCRT vs. 9.6 Gy DIBH 3DCRT; p < 0.01) and contralateral breast D5% (5.0 Gy FB IMRT vs. 2.9 Gy FB 3DCRT vs. 3.0 Gy DIBH 3DCRT; p < 0.01).Conclusions: We found that acquiring DIBH scans for RBC patients receiving RNI reduced the need for FB IMRT from 68% to 18%. As compared to FB 3DCRT, DIBH 3DCRT resulted in in equivalent target coverage with significantly lower lung and liver doses. FB IMRT is useful to keep the right lung V20 within acceptable limits at the expense of higher dose to other OARs. Our data support the routine acquisition of DIBH scans in RBC patients undergoing RNI in order to decrease the proportion of patients that require FB IMRT. Citation Format: Sachin R Jhawar, Kylee Lindsey, Karla Kuhn, Kayla Tedrick, Ian Zoller, William Taylor, Eric Cochran, Erin Healy, Sasha Beyer, Julia White, Jose G Bazan. Should deep inspiration breath hold scans be standardly acquired for right-sided breast/chestwall and regional nodal irradiation? [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P3-19-02.

BackgroundWith survival improving in many metastatic malignancies, spine metastases have increasingly become a source of significant morbidity; achieving durable local control (LC) is critical. Stereotactic body radiotherapy (SBRT) may offer improved LC and/or symptom palliation. However, due to setup concerns, SBRT is infrequently offered to patients with ≥3 contiguous involved levels. Because data are limited, we sought to evaluate the feasibility, toxicity, and cancer control outcomes of spine SBRT delivered to ≥3 contiguous levels.MethodsWe retrospectively identified all SBRT courses delivered between 2013 and 2019 at a tertiary care institution for postoperative or intact spine metastases. Radiotherapy was delivered to 14–35 Gy in 1–5 fractions. Patients were stratified by whether they received SBRT to 1–2 or ≥3 contiguous levels. The primary endpoint was 1-year LC and was compared between groups. Factors associated with increased likelihood of local failure (LF) were explored. Acute and chronic toxicity was assessed. In-depth dosimetric data were collected.ResultsOverall, 165 patients with 194 SBRT courses were identified [54% were men, median age was 61 years, 93% had Karnofsky Performance Status (KPS) ≥70, and median follow-up was 15 months]. One hundred thirteen patients (68%) received treatment to 1–2 and 52 to 3–7 (32%) levels. The 1-year LC was 88% (89% for 1–2 levels vs. 84% for ≥3 levels, p = 0.747). On multivariate analysis, uncontrolled systemic disease was associated with inferior LC for patients with ≥3 treated levels. No other demographic, disease, treatment, or dosimetric variables achieved significance. Rates of new/progressive fracture were equivalent (8% vs. 9.5%, p = 0.839). There were no radiation-induced myelopathy or grade 3+ acute or late toxicities in either group. Coverage of ≥95% of the planning target volume with ≥95% prescription dose was similar between groups (96% 1–2 levels vs. 89% ≥3 levels, p = 0.078).ConclusionsFor patients with ≥3 contiguous involved levels, spine SBRT is feasible and may offer excellent LC without significant toxicity. Prospective evaluation is warranted.

Background With advances in systemic therapy translating to improved survival in metastatic malignancies, spine metastases have become an increasingly common source of morbidity. Achieving durable local control (LC) for patients with circumferential epidural disease can be particularly challenging. Circumferential stereotactic body radiotherapy (SBRT) may offer improved LC for circumferential vertebral and/or epidural metastatic spinal disease, but prospective (and retrospective) data are extremely limited. We sought to evaluate the feasibility, toxicity, and cancer control outcomes with this novel approach to circumferential spinal disease. Methods We retrospectively identified all circumferential SBRT courses delivered between 2013 and 2019 at a tertiary care institution for post-operative or intact spine metastases. Radiotherapy was delivered to 14–27.5 Gy in one to five fractions. Feasibility was assessed by determining the proportion of plans for which ≥95% planning target volume (PTV) was coverable by ≥95% prescription dose. The primary endpoint was 1-year LC. Factors associated with increased likelihood of local failure (LF) were explored. Acute and chronic toxicity were assessed. Detailed dosimetric data were collected. Results Fifty-eight patients receiving 64 circumferential SBRT courses were identified (median age 61, KPS ≥70, 57% men). With a median follow-up of 15 months, the 12-month local control was 85% (eight events). Five and three recurrences were in the epidural space and bone, respectively. On multivariate analysis, increased PTV and uncontrolled systemic disease were significantly associated with an increased likelihood of LF; ≥95% PTV was covered by ≥95% prescription dose in 94% of the cases. The rate of new or progressive vertebral compression fracture was 8%. There were no myelitis events or any grade 3+ acute or late toxicities. Conclusions For patients with circumferential disease, circumferential spine SBRT is feasible and may offer excellent LC without significant toxicity. A prospective evaluation of this approach is warranted.

Background: Computational spine models of various types have been employed to understand spine function, assess the risk that different activities pose to the spine, and evaluate techniques to prevent injury. The areas in which these models are applied has expanded greatly, potentially beyond the appropriate scope of each, given their capabilities. A comprehensive understanding of the components of these models provides insight into their current capabilities and limitations.Methods: The objective of this review was to provide a critical assessment of the different characteristics of model elements employed across the spectrum of lumbar spine modeling and in newer combined methodologies to help better evaluate existing studies and delineate areas for future research and refinement. Findings: A total of 155 studies met selection criteria and were included in this review. Most current studies use either highly detailed Finite Element models or simpler Musculoskeletal models driven with in vivo data. Many models feature significant geometric or loading simplifications that limit their realism and validity. Frequently, studies only create a single model and thus can't account for the impact of subject variability. The lack of model representation for certain subject cohorts leaves significant gaps in spine knowledge. Combining features from both types of modeling could result in more accurate and predictive models.Interpretation: Development of integrated models combining elements from different model types in a framework that enables the evaluation of larger populations of subjects could address existing voids and enable more realistic representation of the biomechanics of the lumbar spine.

AORTIC ARCH pathology, as a part of acute aortic syndrome, is a difficult clinical problem to address. Whether it presents as an intramural hematoma with impending rupture, isolated aneurysmal disease, or acute localized dissection, it creates a clinical challenge to the perioperative team. Some of the difficulty stems from determining the best operative strategy, either an open approach, total endovascular repair, or a hybrid of the two. Hybrid aortic arch repair involves 2 steps—open aortic branch revascularization and endovascular aortic repair. 1 Gregory AJ Prusinkiewicz CA Herget E et al. Subclavian graft thrombosis as an alternative cause for delayed spinal cord ischemia following hybrid aortic arch repair. J Cardiothorac Vasc Anesth. 2014; 28: 718-722 Google Scholar Variations of these 2 steps may be used. Each involves its own set of risks.

This article presents an effort toward building an artificial intelligence (AI) assisted framework, coined ReconGAN, for creating a realistic digital twin of the human vertebra and predicting the risk of vertebral fracture (VF). ReconGAN consists of a deep convolutional generative adversarial network (DCGAN), image-processing steps, and finite element (FE) based shape optimization to reconstruct the vertebra model. This DCGAN model is trained using a set of quantitative micro-computed tomography (micro-QCT) images of the trabecular bone obtained from cadaveric samples. The quality of synthetic trabecular models generated using DCGAN are verified by comparing a set of its statistical microstructural descriptors with those of the imaging data. The synthesized trabecular microstructure is then infused into the vertebra cortical shell extracted from the patient's diagnostic CT scans using an FE-based shape optimization approach to achieve a smooth transition between trabecular to cortical regions. The final geometrical model of the vertebra is converted into a high-fidelity FE model to simulate the VF response using a continuum damage model under compression and flexion loading conditions. A feasibility study is presented to demonstrate the applicability of digital twins generated using this AI-assisted framework to predict the risk of VF in a cancer patient with spinal metastasis.

BACKGROUND Spine surgery is indicated for select patients with mechanical instability, pain, and/or malignant epidural spinal cord compression, with or without neurological compromise. Stereotactic body radiotherapy (SBRT) is an option for durable local control (LC) for metastatic spine disease. OBJECTIVE To determine factors associated with LC and progression-free survival (PFS) for patients receiving postoperative stereotactic spine radiosurgery. METHODS We analyzed consecutive patients from 2013 to 2019 treated with surgical intervention followed by SBRT. Surgical interventions included laminectomy and vertebrectomy. SBRT included patients treated with 1 to 5 fractions of radiosurgery. We analyzed LC, PFS, overall survival (OS), and toxicity. Univariate and multivariate analyses were performed. RESULTS A total of 63 patients were treated with a median follow-up of 12.5 mo. Approximately 75% of patients underwent vertebrectomy and 25% underwent laminectomy. One-year cumulative incidence of local failure was 19%. LC was significantly improved for patients receiving radiosurgery ≤40 d from surgery compared to that for patients receiving radiosurgery ≥40 d from surgery, 94% vs 75%, respectively, at 1 yr (P = .03). Patients who received preoperative embolization had improved LC with 1-yr LC of 88% vs 76% for those who did not receive preoperative embolization (P = .037). Significant predictors for LC on multivariate analysis were time from surgery to radiosurgery, higher radiotherapy dose, and preoperative embolization. The 1-yr PFS and OS was 56% and 60%, respectively. CONCLUSION Postoperative radiosurgery has excellent and durable LC for spine metastasis. An important consideration when planning postoperative radiosurgery is minimizing delay from surgery to radiosurgery. Preoperative embolization and higher radiotherapy dose were associated with improved LC warranting further study.

Metastatic adamantinoma involving the spine is an extremely rare occurrence. In this case report, we present a patient with recurrent adamantinoma of the tibia which was found to have metastasized to the spine. The metastatic involvement was diagnosed pathologically, status post CT guided percutaneous core needle biopsy, performed after the patient returned with concerning symptoms and imaging findings suggestive of metastasis. The patient was ultimately treated surgically with vertebrectomy and reconstruction. A thoughtful review of this disease process is explored, emphasizing the pathology, imaging characteristics, and pertinent differential diagnostic considerations. While uncommon, knowledge of this rare disease process and its presentation can improve future patient diagnosis and outcomes.

K. Dibs: None. J.D. Palmer: Research Grant; Varian Medical Systems, The Kroger Company. Consultant; Huron Consulting. Speaker's Bureau; Varian Medical Systems, Depuy Synthes. Advisory Board; Novocure. Member of panel; NCCN.A. Olausson: None. E. Bourekas: None. D. Boulter: None. A.S. Ayan: None. E.R. Cochran: None. V. Yildiz: None. J.C. Grecula: None. A.L. Arnett: None. R. Raval: None. S. Beyer: None. T. Scharschmidt: None. J.B. Elder: None. A. Chakravarti: None. E. Mendel: None. D.M. Blakaj: None.

Both endovascular repair (EVR) and open repair (OR) surgery of thoraco-abdominal aortic aneurysms cause spinal cord (SC) injury that can lead to paraparesis or paraplegia. It has been assumed that mechanisms responsible for SC damage after EVR are similar to those after OR. This pilot study compared the pathophysiology of SC injury after EVR versus OR using a newly developed EVR dog model. An increasing number of stents similar to those used in patients were inserted in the aorta of three dogs to ensure thoracic or thoracic plus lumbar coverage. The aorta of OR dogs was cross-clamped for 45 min. Behavior assessment demonstrated unique patterns of proprioceptive ataxia and evolving paraparesis in EVR versus irreversible paraplegia in OR. MRI showed posterior signal in lumbar SC after EVR versus central cord edema after OR. Histopathology showed white matter edema in L3–L5 localized to the dorsal column medial lemniscus area associated with loss of myelin basic protein but not neurons after EVR, versus massive neuronal loss in the gray matter in L3–L5 after OR. Metabolome analysis demonstrates a distinctive chemical fingerprint of cellular processes in both interventions. Our results call for the development of new therapeutics tailored to these distinct pathophysiologic findings.