Chongqi Tu

Background: Stage IIB conventional osteosarcoma exhibits highly heterogeneous responses to neoadjuvant chemotherapy, necessitating reliable, noninvasive pre-treatment evaluation systems. Methods: This retrospective study of 198 patients utilized pre-treatment T2-weighted magnetic resonance imaging (MRI) and baseline clinical data. Twelve classical convolutional neural network architectures were benchmarked to identify the optimal imaging backbone. This selected radiomics score was then integrated with core clinical variables using logistic regression to construct a multimodal model. Gradient-weighted Class Activation Mapping++ (Grad-CAM++) and SHapley Additive exPlanations (SHAP) algorithms provided image- and feature-level interpretability. Results: DenseNet121 achieved the highest mean cross-validated area under the curve (AUC; 0.775) and was selected, with AUCs of 0.797 (95% CI: 0.716–0.877) and 0.701 (95% CI: 0.513–0.877) in the training and test cohorts. Adding the clinical variables improved discrimination, yielding a test-cohort AUC of 0.827 (95% CI: 0.661–0.992) with significant reclassification gain over the radiomics-only model (NRI = 1.28, IDI = 0.177; DeLong p = 0.011). Grad-CAM++ activation concentrated on the solid medullary and extraosseous soft-tissue components, and SHAP ranked the radiomics score as the strongest contributor. Conclusions: Integrating deep learning radiomics with key clinical features provides a noninvasive tool for the individualized prediction of neoadjuvant chemosensitivity in osteosarcoma.

Introduction: Pulmonary metastasis in high‐grade conventional osteosarcoma remains a leading cause of treatment failure and mortality. Traditional monitoring methods are insufficient for early detection. This study aims to develop and validate an interpretable machine learning (ML) model using multi‐dimensional data from electronic medical records (EMR) to predict the early risk of pulmonary metastasis in high‐grade conventional osteosarcoma patients. Methods: This retrospective study included data from 522 high‐grade conventional osteosarcoma patients. After rigorous feature selection, 12 independent predictive factors were identified: prothrombin time (PT), international normalized ratio (INR), fibrinogen, globulin, prognostic nutritional index (PNI), eosinophil percentage, monocyte percentage, neutrophil percentage, monocyte count, maximum tumor diameter, gender, and amputation status. Eleven ML models were constructed and compared, with Shapley Additive Explanations (SHAP) analysis applied to enhance model interpretability and clinical transparency. Results: The gradient boosting model exhibited superior performance, achieving an area under the curve (AUC) of 0.891 in the training set and 0.742 in the independent test set. SHAP analysis revealed that tumor maximum diameter was the most influential predictor of pulmonary metastasis risk, while inflammation and coagulation‐related indicators also demonstrated significant contributions. Discussion: The gradient boosting model effectively predicts the early risk of pulmonary metastasis in high‐grade conventional osteosarcoma and provides interpretable risk factors. This model shows potential as a clinical decision support tool, facilitating personalized risk management and precision medicine, aiming to improve patient outcomes.

Background: Although the tumor coagulome interacts with the tumor immune microenvironment (TME) in solid tumors, its role in osteosarcoma (OS) remains uncharacterized. This study aimed to delineate this transcriptomic interplay and identify potential prognostic targets. Methods: This study was performed with bulk RNA sequencing (RNA-seq), single-cell RNA sequencing (scRNA-seq), and clinical phenotype data. Bioinformatic approaches were employed at the transcriptomic level to investigate the impact of the tumor coagulome on the TME and prognosis in OS. We validated the above findings using immunohistochemistry and immunofluorescence. Results: The activity of a coagulation-related transcriptional signature was found to correlate with the degree of malignancy in OS. Its activity score demonstrated predictive value for OS prognosis, with a maximum area under the curve (AUC) of 0.802. scRNA-seq analysis indicated that inflammatory cancer-associated fibroblasts (iCAFs) and APOE⁺ macrophages were predominantly enriched in the high coagulation score subgroup. Our data further suggest that iCAFs may facilitate the M2 polarization of APOE⁺ macrophages via the C3–C3AR1 axis, potentially contributing to poorer clinical outcomes in patients with OS. Conclusions: These findings imply that within a high coagulation-related transcriptional score group, the interaction between iCAFs and APOE⁺ macrophages, likely mediated by the C3–C3AR1 axis, could facilitate OS progression. Consequently, the C3–C3AR1 signaling pathway might represent a promising target for future therapeutic strategies and coagulome monitoring in OS.

Background: Parosteal osteosarcoma (PAOS) is a rare and low‐grade primary bone malignant tumor that typically occurs in the distal femur. Traditional reconstruction options for distal femur PAOS include distal femoral replacement (DFR), autografts, and allografts, but issues such as functional sacrifice, insufficient mechanical strength, and low bone union rate limit their further clinical application. 3D‐printed prosthesis, due to the matching shape of various bone defects, could be a novel strategy to reconstruct the hemicortical defect after distal femur PAOS tumor resection. In this study, we designed a 3D‐printed strut‐type prosthesis (3DPSP) to repair the hemicortical defect and evaluated the mid‐term effect in distal femur PAOS. Methods: A retrospective analysis was conducted on patients diagnosed with PAOS at the distal femur between April 2018 and January 2021. Six patients received DFR and six patients received customized 3DPSP reconstruction. Clinical outcomes including oncology status, defect lengths, and ranges of motion of the knee, Musculoskeletal Tumor Society (MSTS) scores, and complications were recorded. X‐ray was performed to evaluate the position and stability of the prosthesis and T‐SMART was applied to assess the interface osteointegration between prosthesis and bone. Results: A total of 12 patients were enrolled in this study with an average follow‐up of 42.1 ± 11.8 months, ranging from 22 to 59 months. In the 3DPSP group, the average operation time was 2.3 ± 0.4 h; the intraoperative blood loss was 175.0 ± 82.2 mL. The average range of motion of the knee was 0°–123° and the mean MSTS score was 28.3 ± 1.6. Compared to the DFR group, the 3DPSP group exhibited significantly less intraoperative blood loss and superior functional recovery. Conclusions: In this study, we designed a novel 3DPSP to reconstruct the hemicortical defect after distal femur PAOS resection. The mid‐term follow‐up result demonstrated that this prosthesis could be an effective reconstruction option for distal femur PAOS patients because of the restored limb function and favorable bone‐prosthesis integration. Further observation of this novel prosthesis is needed.

Uncemented endoprosthetic reconstruction is commonly used in pediatric patients after distal femoral tumour resection. However, conventional short stems may be associated with aseptic loosening, bone resorption, and end-of-stem pain, which may compromise long-term outcomes. This retrospective cohort study aimed to compare prosthetic loosening, bone resorption, and clinical outcomes between slotted stems and conventional short stems in pediatric patients undergoing uncemented distal femoral endoprosthetic reconstruction. We performed a retrospective cohort study of pediatric patients who underwent implantation of an uncemented endoprosthesis with either a slotted stem or a conventional short stem between 2021 and 2023. Aseptic loosening and bone resorption were evaluated radiographically using radiolucent lines, stem axis displacement, and translucent bands. Clinical outcomes were assessed using the 1993 Musculoskeletal Tumor Society (MSTS-93) score and the rate of end-of-stem pain. Complications and endoprosthetic failures were also recorded. A total of 47 patients were included, including 24 in the slotted stem group and 23 in the control group. The median age was 9 years (range, 8-14) in the slotted stem group and 10 years (range, 7-14) in the control group. The median follow-up was 35 months (range, 24-49) in the slotted stem group and 36 months (range, 25-48) in the control group. No significant between-group difference was observed in the rate of radiographic aseptic loosening (4.17% vs. 8.70%, p = 0.609) or in bone resorption in Zones 1-4. However, the slotted stem group had less bone resorption in Zone 5 (p = 0.022). MSTS-93 scores were similar between groups (p = 0.574), whereas the slotted stem group had a lower rate of end-of-stem pain (p = 0.048). In this retrospective cohort, the slotted stem showed comparable short-term outcomes to the conventional short stem, with less Zone 5 bone resorption and a lower rate of end-of-stem pain. Slotted stems are viable alternatives to short stems for distal femoral endoprosthetic reconstruction following tumour resection in pediatric patients.

Background: Osteosarcoma is the most common primary malignant bone tumor. However, research on predicting the prognosis of patients with osteosarcoma after chemotherapy (POC) remains limited. Notably, the Lung Immune Prognostic Index (LIPI) has emerged as a novel and effective prognostic factor in lung cancer. Therefore, this study aims to explore the prognostic significance of LIPI in POC for the first time, providing new insights and a foundation for evaluating the prognosis of these patients.Methods: This retrospective study analyzed patients with POC who were admitted to our center between January 2012 and January 2022. Hematological and clinical characteristics were collected and systematically evaluated. Kaplan–Meier survival analysis and Cox regression models were employed to assess the associations between various prognostic factors and overall survival (OS). Independent risk factors influencing OS were identified through both univariate and multivariate analyses. Based on these findings, a LIPI nomogram model was developed to predict OS in patients with POC.Results: This study included 150 patients who underwent chemotherapy, with 41 (27%), 80 (53%), and 29 (19.3%) classified into poor, moderate, and good prognostic categories, respectively, based on the LIPI classification (P < 0.0001). Time-dependent receiver operating characteristic (ROC) curve analysis demonstrated that LIPI exhibited superior prognostic predictive capability compared to other hematological and clinical parameters. Univariate and multivariate analyses identified LIPI as an independent prognostic factor. A nomogram was subsequently developed by integrating significant prognostic variables. Calibration curves confirmed the nomogram’s accuracy in predicting three- and five-year overall survival (OS) post-chemotherapy. Furthermore, decision curve analysis indicated that the LIPI-based nomogram would provide substantial clinical benefits for chemotherapy patients.Conclusion: This study assessed the prognostic efficacy of LIPI in patients with POC and developed a LIPI-based nomogram to assist clinicians in predicting three- and five-year overall survival (OS). The proposed model has the potential to facilitate timely interventions and guide personalized management strategies, thereby improving patient outcomes.

Objective: Durable and biologically integrated fixation is critical for long‐term implant survival in patients with primary bone tumors. However, limited evidence exists regarding the long‐term outcomes of uncemented stem designs in this population. Specifically, we investigated: (1) the long‐term patient and implant survivorship rates; (2) the influence of factors such as resection length and patient age on implant survival; and (3) the incidence and types of complications, particularly those requiring implant removal or revision.Methods: We retrospectively reviewed 76 patients (49 males, 27 females; mean age 41 years, range 14–78 years) who underwent proximal femoral replacement with a modular uncemented endoprosthesis between 2015 and 2022. The mean follow‐up was 63.4 months (median: 60.5 months; range: 12–104 months). Functional outcomes were assessed using the Musculoskeletal Tumor Society (MSTS) score, while complications were classified based on the Henderson et al. system. Survivorship analyses were conducted using Kaplan–Meier methods.Results: The 5‐year patient survival rate was 88.0%, and the 5‐year implant survival rate was 90.4%. The mean MSTS score at final follow‐up was 25.6 (range 16–30), with 87.7% of patients achieving good to excellent functional outcomes. Younger patients (< 30 years) exhibited poorer implant survival, while resection length did not significantly impact outcomes. Complications occurred in 18.4% of cases, categorized into type 1 (soft tissue‐related, n = 3), type 2 (aseptic loosening, n = 2), type 3 (structural failure, n = 0), type 4 (infection, n = 3), type 5 (tumor recurrence, n = 3), and three cases of acetabular failure.Conclusions: Modular uncemented endoprostheses for proximal femoral replacement demonstrated promising survivorship and functional outcomes in patients with primary bone tumors. The low rate of aseptic loosening highlights the benefits of uncemented stem designs. However, younger age remains a risk factor for reduced implant longevity.

To explore the feasibility of using biomechanical indicators as supplementary evaluation to the Musculoskeletal Tumor Society Scoring System (MSTS) for amputee patients. Twenty-four patients who underwent hemipelvectomy between September 2018 and January 2025 were enrolled. There were 15 males and 9 females with an average age of 61.4 years (range, 45-76 years). Participants performed gait tests at self-selected speeds using three assistive devices (prosthesis, single crutch, and double crutches). Motion data were analyzed using a customized OpenSim model. Biomechanical indicators of the intact limb exhibiting common characteristics were screened through correlation and sensitivity analyses. Test-retest reliability [interclass correlation coefficient (ICC)] of selected parameters was assessed to evaluate their potential as MSTS score supplements. All biomechanical indicators showed significant positive correlations with MSTS scores across assistive devices ( P0.8, including walking speed, maximum hip angle, maximum hip moment, peak hip flexion moment, peak hip extension moment, hip flexion impulse, and hip extension impulse. Among these, maximum hip moment, hip flexion impulse, and hip extension impulse exhibited significant between-group differences in adjacent MSTS levels ( P0.74, P<0.01). Biomechanical indicators statistically qualify as potential supplements to MSTS scoring. Maximum hip moment, hip flexion impulse, and hip extension impulse demonstrate particularly high sensitivity to MSTS score variations.

Integrating 4D printing technology in medical implants offers promising advancements for minimally invasive delivery (MID) and personalized orthopedic solutions. This study presents a 4D-printed shape memory nickel-titanium (NiTi) mesh implant for cavitary bone defect reconstruction, enabling a time-dependent shape transformation. Fabricated through selective laser melting (80 W laser power, 600 mm/s scanning speed, 70 μm hatch spacing, 25 μm layer thickness), the implant can be compressed during implantation and recover its original shape. Micro-computed tomography analysis confirmed high geometric fidelity (D50 = 58 μm), while scanning electron microscopy-energy dispersive spectroscopy analysis revealed a uniform microstructure and confirmed the homogeneous distribution of Ni/Ti across the mesh implant. Phase transformation testing showed that the austenite finish temperatures (austenite finish) of the as-built sample and the acid-washed sample were below the 37°C physiological threshold. Compression testing indicated that a force of 156 N was required for 30% deformation, with complete recovery to its pre-defined shape. Clinically, the implant reduced cortical bone fenestration by 20%. Post-operative imaging at 6 and 12 months showed excellent osseointegration and minimal residual cavities. Functional assessments at 12 months indicated excellent recovery, with a Musculoskeletal Tumor Society score of 29. In the present study, the clinical use of the 4D-printed mesh implant demonstrated not only satisfactory osteointegration but also a practical advantage in surgical handling. The shape recovery of the implant from a compressed state to its pre-designed shape allowed for MID and precise fit to the defect contour.

Objectives: Pelvic bone tumor resection and reconstruction present significant challenges due to complex anatomy and weight‐bearing demands. 3D‐printed hemipelvic prostheses, incorporating customized osteotomy guides and porous structures, offer a promising solution for enhancing osseointegration. This study evaluates the long‐term outcomes of 3D‐printed custom hemipelvic reconstruction with a focus on the integration of auxetic biomaterials with a negative Poisson's ratio to optimize mechanical properties.Methods: A retrospective analysis was conducted on 12 patients with primary pelvic malignancies who underwent reconstruction using 3D‐printed hemipelvic prostheses between January 2018 and May 2023. Follow‐up duration was 48 months (range, 29–64 months) Oncological, functional, surgical, pain control, and radiographic outcomes were assessed.Results: At the latest follow‐up, 8 patients (66.7%) were disease‐free, 3 (25%) had disease progression, and 1 (8.3%) died from metastatic complications. Functional outcomes improved significantly, with the MSTS‐93 score increasing from 15 (range, 12–17) to 26 (range, 21–29). Pain scores decreased from 5 (range, 4–7) to 1 (range, 0–2). The median surgical duration was 270 min (range, 150–560 min), with intraoperative blood loss averaging 3200 mL (range, 1900–6300 mL). Complications included poor wound healing in 2 patients (16.7%), managed with VAC drainage. No mechanical failures, loosening, or fractures occurred. Accurate osteotomy, prosthesis implantation, and screw fixation were achieved. Successful osseointegration was observed in all cases, with no signs of bone absorption or osteolysis.Conclusions: 3D‐printed custom hemipelvic prostheses with auxetic biomaterials offer an effective solution for pelvic reconstruction, providing promising oncological, functional, and radiographic outcomes. These findings support the use of 3D printing in complex pelvic defect reconstruction, optimizing both osteointegration and mechanical strength.