Volume 106 · Issue 01 · 2026

With a focus on strengthening innovative drug development and promoting the clinical application of emerging medical technologies, we have summarized innovation-driven pharmaceutical development and clinical advancement, highlighting the key challenges, current hotspots, and potential breakthroughs, while outlining priority directions for fostering deeper integration between pharmaceutical development and medical practice. This paper provides a systematic analysis of the progress, challenges, and future directions of China′s innovative drug development. Currently, through biomedical technology innovation, comprehensive policy support, accelerated regulatory reviews and approval processes, and reforms in the payment system, China has risen to a leading global position in terms of the number of innovative drugs developed and license-out deals. However, challenges remain, including weak capacity for original innovation, insufficient alignment with clinical needs, and limited international engagement. To address these, the country has recently launched the national major science and technology special project on drug development in the new phase. This initiative aims to shift the focus toward capacity building, drug discovery and original innovation, while optimizing the project organization and management to foster a collaborative innovation ecosystem. The goal is to establish an independent, controllable, and world-leading drug innovation system by 2035, thereby providing strong support for the "Healthy China" initiative.

The deep integration of artificial intelligence (AI) is driving a profound transformation in the research paradigm of sports medicine. This article systematically reviewed the primary models of AI-sports medicine integration, including intelligent early warning for sports injury risks, intelligent diagnosis of musculoskeletal health issues, intelligent assistance of surgeries, and intelligent prediction of patient prognosis. It further analyzed the key challenges hindering deep integration, covering technical bottlenecks and data barriers, patient privacy and health equity risks, as well as the lack of regulatory and health insurance mechanisms. On this basis, systematic pathways for promoting high-quality development were proposed, including strengthening a clinical value-oriented technology research and development and transformation system, building a full-cycle data security and ethical governance framework for technologies, and improving value-based payment mechanisms and policy support environments. Through multi-dimensional and coordinated advancement, the potential of AI-sports medicine integration could be fully unleashed, achieving a leap from technological empowerment to health value creation, thereby providing robust support for the implementation of the"Healthy China 2030"strategy.

More and more studies have confirmed that robotic-assisted total knee arthroplasty (RATKA) can improve surgical accuracy and has a user-friendly learning curve. However, most current studies evaluating the learning curve use operative time as the standard of assessment, although proficiency does not equate to complete mastery of RATKA surgical techniques. This article will summarize experience from two perspectives: surgical technical pitfalls and non-surgical technical adverse events in RATKA. Regarding surgical techniques, it is essential to fully understand the differences in workflow and principles between RATKA and traditional total knee arthroplasty to avoid the occurrence of"unsatisfactory cases". At the non-surgical technical level, a comprehensive understanding of the various new complications that may arise during RATKA is necessary in order to consciously prevent adverse events. Ultimately, the robot will become a powerful tool for surgeons.

The rapid advancement of artificial intelligence (AI) technologies, particularly deep learning algorithms and hardware devices, has profoundly transformed diagnostic and therapeutic paradigms in sports medicine. This article reviews the applications and recent progress of AI across multiple domains in the field, including the use of deep learning and machine learning models for injury risk prediction and early screening; the application of computer vision techniques in imaging diagnosis; the integration of intelligent algorithms into surgical planning and intraoperative assistance; the dynamic monitoring of rehabilitation processes; and the deployment of large language models to support patient education and consultation. The deep integration of sports medicine and AI requires interdisciplinary collaboration across clinical medicine, sports science, materials science, and computer science. This article highlights the current applications, discusses existing challenges and potential risks, and provides an outlook on future development trends.

The incidence of osteoporosis increases significantly with age. When elderly patients require internal fixation due to degenerative changes or trauma in the thoracolumbar spine, concomitant osteoporosis considerably elevates surgical difficulty and risk. Common complications include screw loosening, internal fixation failure, and nonunion of bone grafts, posing substantial challenges to spine surgeons. Currently, there is a lack of relevant clinical standards or guidelines for the application of thoracolumbar internal fixation in elderly patients with osteoporosis. To address this gap, based on the principles of evidence-based medicine, integrating scientific rigor with practicality, conducting a systematic review and analysis of relevant literature, the Innovation and Transformation Group of the Chinese Orthopaedic Association and the Spine Trauma Group of the Chinese Association of Orthopaedic Surgeons and the Orthopedic Minimally Invasive Injury Professional Committee of Shaanxi Province Association of Orthopedic Surgeons jointly led an expert initiative leading to the development of this guideline. Recommendations are proposed primarily covering preoperative education, surgical strategies, intraoperative considerations, and postoperative rehabilitation, aiming to standardize the clinical application of this technique.

Neutropenia is one of the most common hematologic toxicities during the treatment of hematologic malignancies. It can lead to delays in subsequent treatments, dose reduction of chemotherapeutic agents, or even treatment discontinuation. Furthermore, neutropenia increases the risk of febrile neutropenia and severe infections, which adversely affect patient prognosis. Pegylated recombinant human granulocyte colony-stimulating factor (PEG-rhG-CSF) is effective in preventing and treating neutropenia, and is widely used in the treatment of hematologic malignancies and hematopoietic stem cell transplantation (HSCT). To further promote the standardized application of PEG-rhG-CSF in hematological malignancies, the Hematology Oncology Committee of China Anti-Cancer Association organized domestic experts in the related fields, based on current evidence-based medical data and clinical practice, to focus on the clinical application of PEG-rhG-CSF in hematologic malignancies, hematopoietic stem cell mobilization and hematopoietic function reconstitution after HSCT. Eight recommendation opinions were finally proposed to guide the standardized use of drugs in clinical practice.

With the rapid aging of the population, disability and dementia have become major public health challenges affecting the health and quality of life of the elderly in the community. The multidisciplinary collaborative intervention model for the risk of disability and dementia in the elderly, covering key areas such as clinical medicine, public health and social support systems, has become a current research focus. This expert consensus concentrates on eight core functional dimensions—mobility, cognition, emotion, vision, hearing, excretion, swallowing, and comorbidity management—and the corresponding modifiable risk factors. It systematically reviews the clinical evidence and latest advances in relevant interventions in China and abroad. By integrating opinions from multidisciplinary experts through the Delphi method, the consensus ultimately formulates 15 intervention strategies and recommendations across these eight dimensions for the elderly in the community, with the aim of establishing effective approaches to delay the progression of disability and dementia, providing evidence-based support for primary health-care services, and promoting healthy ageing in China.

To quantitatively evaluate the efficacy, precision, and reproducibility of the surgical navigation module of a novel robotic-assisted anterior cruciate ligament reconstruction (ACLR) system during bone tunnel drilling using in vitro models.

Thirty 3D-printed knee models were randomized into a robotic group (n=15) and a conventional group (n=15) using a random number table. Both groups utilized the same individualized surgical plans generated by the system′s preoperative planning module. In the robotic group, bone tunnel drilling was guided by the system′s mechanical arm; in the conventional group, drilling was performed manually by senior sports medicine surgeons using traditional instrumentation. Operative time was recorded. Postoperative bone tunnel lengths and tunnel exit positions in both groups were compared with the preoperative planning values among the three groups.

The total operative time was significantly longer in the robotic group [(25.83±2.20) vs (15.55±1.96) min, P<0.001], whereas the robotic group achieved a significantly shorter duration for the core bone tunnel drilling phase [6.08 (5.72, 6.75) vs 7.60 (6.80, 8.52) min, P=0.003]. There were no statistically significant differences among the three groups regarding the numerical values of the three length indicators (femoral and tibial tunnel lengths, intra-articular distance) or the four tunnel exit position dimensions (deep-shallow, high-low, anterior-posterior, medial-lateral) within the Bernard quadrant (all P>0.05). However, the variance in all positioning dimensions was significantly lower in the robotic group compared to the conventional group (all P<0.05), particularly in the femoral high-low and tibial anterior-posterior directions (both P<0.001), indicating superior consistency in robot-assisted operation.

Robot-assisted ACLR technology improves efficiency during the critical drilling phase and significantly minimizes manual operational variability. These findings provide experimental evidence supporting the application of high-precision surgical navigation in ACLR.

To determine the preventive effect of a WeChat platform-based rehabilitation exercise prescription on deep vein thrombosis (DVT) after total knee arthroplasty (TKA).

Prospective data were collected from patients who underwent TKA at the First Affiliated Hospital of Kunming Medical University between December 2023 and March 2025 (set as the experimental group). Retrospective clinical data were collected from patients who underwent TKA at the same hospital between January 2022 and October 2023 (set as the conventional group). All surgeries were performed by the same senior chief physician, and preoperative bilateral lower extremity vascular ultrasound showed no thrombosis in any patient. One day before surgery, the patients in the experimental group received a personalized Quick Response (QR) code for a WeChat public platform from a rehabilitation specialist nurse. After scanning the code, patients could access rehabilitation exercise videos. The rehabilitation nurse explained key points and precautions for the exercises. Patients began rehabilitation training as early as postoperative day 0. The conventional group followed a paper-based rehabilitation exercise prescription and received guidance from rehabilitation nurses on key points and precautions. The incidence of DVT within one week after surgery, postoperative length of stay, total inpatient medical costs (including material costs and non-material costs), and Knee Society Score (KSS) (including clinical and functional scores) were compared between the two groups.

A total of 118 patients were enrolled, with 59 in the conventional group and 59 in the experimental group. In the conventional group, there were 14 males and 45 females, aged (68.0±7.2) years. In the experimental group, there were 11 males and 48 females, aged (66.5±7.4) years. There was no statistically significant differences in gender, age, operation time, preoperative thrombosis risk score, or anticoagulant dosage between the two groups (all P>0.05). The incidence of DVT in the operated limb was lower in the experimental group than that in the conventional group [35.6% (21/59) vs 66.1% (39/59), P=0.001]. The postoperative hospital stay was shorter in the experimental group [(6.9±2.7) vs (8.9±2.5) d, P<0.001]. The total inpatient medical costs were lower in the experimental group [18 873 (17 561, 20 880) vs 23 492 (18 362, 45 027) yuan, P<0.001], with both material costs [2 214 (2 214, 5 379) vs 2 899 (2 214, 30 500) yuan, P=0.028] and non-material costs [15 344 (13 837, 16 157) vs 15 912 (14 563, 19 046) yuan, P=0.011] in the experimental group were both lower than those in the conventional group. The clinical score of KSS was higher in the experimental group [(76.5±5.7) vs (68.5±3.6) points, P<0.001], while no significant difference was found in the functional score of KSS between the two groups (P=1.692).

The WeChat public platform-based rehabilitation exercise prescription can reduce the incidence of DVT, shorten the hospital stay, and decrease total inpatient medical costs in patients after TKA.

To assess the bilateral differences in static postural stability among patients with chronic ankle instability(CAI) using inertial sensors, and to further analyze their dependence on visual input.

The CAI patients prospectively enrolled from the Sports Medicine Department outpatient clinic of Peking University Third Hospital between May 2024 and December 2024 as the CAI group, and concurrently enrolled healthy people from society as the healthy control group. An inertial sensor was firmly affixed to the lumbar region (L₃-L₅) of each participant. Under both eyes-open and eyes-closed conditions, participants performed three single-leg standing trials on each leg while data were simultaneously collected from a force plate and the inertial sensor. Static postural stability was assessed using two primary indicators: the root mean square of acceleration (RMS_acc), calculated from the resultant acceleration derived from tri-axial accelerometer signals; and the mean velocity of the center of pressure (COP_MV), derived from the anterior-posterior and medial-lateral displacement velocities of the center of pressure measured by the force plate. Differences in RMS_acc and COP_MV between the two groups under different conditions were compared, and the intraclass correlation coefficient (ICC) was calculated to assess agreement between the inertial sensor and force plate measurements.

A total of 56 participants were included, 28 in the CAI group and 28 in the healthy control group. The CAI group included 14 males and 14 females with a mean age of (28.5±6.1) years, while the healthy control group included 10 males and 18 females, aged (27.5±6.4) years; there was no significant differences between the two groups in terms of gender or age (both P>0.05). The CAI group had significantly lower CAIT score than the healthy controls [(15.2±3.0) vs (29.2±1.5) points, P<0.001] and reported a greater number of previous ankle sprains [(4.2±1.3) times vs 0 time, P<0.001]. Under eyes-open conditions, the affected side of the CAI group exhibited higher RMS_acc [M(Q₁, Q₃)][0.81 (0.68, 1.38) m/s²] and COP_MV [(2.45±0.36) mm/s] than the non-dominant limb of the healthy controls [0.31 (0.22, 0.41) m/s², 1.69 (1.54, 2.25) mm/s, respectively] (both P<0.05). Under eyes-closed conditions, both the affected [RMS_acc: 0.95 (0.62, 1.11) m/s²; COP_MV: (6.20±1.72) mm/s] and unaffected sides [RMS_acc: 0.51 (0.37, 1.03) m/s²; COP_MV: 4.17 (4.05, 5.36) mm/s] of the CAI group were significantly higher than those of the corresponding sides of the healthy controls [(0.38±0.28) m/s² and (3.74±1.61) mm/s, respectively] (all P<0.05). Consistency analysis between RMS_acc and COP_MV showed moderate agreement in the CAI group across different visual conditions and stance sides (ICC=0.51-0.72), whereas the healthy control group displayed lower and more variable consistency (ICC=0.39-0.65).

Lumbar-mounted inertial sensors can objectively capture postural instability during single-leg stance in individuals with CAI. Participants with unilateral CAI exhibited impaired postural stability both on the affected and the unaffected side, and show an increased dependence on visual input.

To compare the healing status of medial meniscus posterior root tear (MMPRT) in patients treated with either isolated high tibial osteotomy (HTO) or HTO combined with MMPRT repair, and to analyze its relationship with clinical outcomes.

A retrospective analysis was conducted on 38 patients who underwent isolated HTO and 26 patients who underwent HTO combined with MMPRT repair at the Hebei Medical University Third Hospital between June 2022 and June 2024. The MMPRT healing rates (complete/partial/failure) were compared between the two groups during second-look arthroscopy. Based on MMPRT healing status, the patients were divided into the Healed group (n=44) and the Non-healed group (n=20). The cartilage repair status of the medial femoral condyle (MFC) and medial tibial plateau (MTP) was compared between these two groups using the International Cartilage Repair Society (ICRS) grading system during second-look arthroscopy. The clinical outcomes were assessed using the visual analog scale (VAS) of pain, Lysholm score, and Tegner score prior to the second-look arthroscopy. The radiographic evaluations included the hip-knee-ankle angle (HKA), joint line convergence angle (JLCA), posterior tibial slope (PTS), medial proximal tibial angle (MPTA), and Kellgren-Lawrence grading.

A total of 64 cases were enrolled, there were 28 males and 36 females with a mean age of (50.9±6.6) years. Although the MMPRT healing rate was slightly higher in the HTO+MMPRT group, there was no statistically significant difference between the two groups [88.5%(23/26) vs 81.6%(31/38), P=0.456]. Cartilage regeneration in the Healed group was superior to that in the Non-healed group. For patients in the Healed group and the Non-healed group, the improvement rates of ICRS grade for the MFC was 59.1% (26/44) and 25.0% (5/20), respectively, and the improvement rates of ICRS grade for the MTP was 50.0% (22/44) and 15.0% (3/20), respectively, and there were statistically significant differences in both comparisons between the two groups (both P<0.001). At the final follow-up, the VAS, Lysholm, and Tegner scores of patients in both the Healed group and the Non-healed group were significantly improved when compared with the preoperative values (all P<0.05), but there was no statistically significant differences between the two groups[(1.3±1.1) vs (1.7±1.1) points, (87.9±5.3) vs (86.6±2.5) points, and (3.7±1.4) vs (3.5±1.1) points, respectively] (all P>0.05). At the final follow-up, the HKA and MPTA in the Healed group and the Non-healed group were all significantly increased when compared with those before the operation (all P<0.05), while the JLCA was significantly decreased (P<0.05); however, no statistically significant differences was found between the two groups (182.3°±1.9° vs 181.7°±1.4°, 92.7°±3.1° vs 93.1°±2.0°, and 2.2°±0.9° vs 2.1°±0.7°, respectively) (all P>0.05). At the final follow-up, neither group showed significant improvement in Kellgren-Lawrence grade compared to preoperative assessments (all P>0.05), with no statistically significant difference between the two groups (P=0.590).

During short-term follow-up, HTO combined with MMPRT repair does not significantly improve the MMPRT healing rate when compared to isolated HTO. Although patients with healed MMPRT exhibit better cartilage repair, they show no superior improvement in clinical and radiographic outcomes.

A retrospective analysis was conducted on the clinical data of patients with giant refractory gastric bezoars treated at Zhongda Hospital of Southeast University between January 2019 and May 2025. All patients underwent modified endoscopic holmium laser-yellow zebra guidewire loop combined lithotripsy. Follow-up gastroscopy was performed 4 weeks postoperatively to analyze the efficacy of this combined technique in treating giant refractory gastric bezoars. A total of 38 patients were included, including 14 males and 24 females, aged (66±12) years (44-88 years). Thirty-five patients successfully achieved bezoar removal after a single treatment session, with a single-session success rate of 92.1% (35/38). Postoperative complications, such as nausea, vomiting, gastric mucosal injury, or pain, occurred in 4 patients (10.5%). At the 4-week follow-up gastroscopy, the complete ulcer healing rate was 87.0% (20/23). The modified endoscopic holmium laser-yellow zebra guidewire loop combined lithotripsy is a safe and effective treatment for giant refractory gastric bezoars.

Artificial intelligence (AI) is changing the way healthcare is practiced, with an important impact on the diagnosis and treatment of sports medicine-related diseases. A branch of AI that is widely used is machine learning (ML), which in turn covers deep learning (DL) and convolutional neural networks (CNN). Another important branch of AI is computer vision (CV). Anterior cruciate ligament (ACL) injury is a common type of sports injury in clinic, and ACL reconstruction is the main treatment for this disease. This article reviews the application and evaluation of various AI tools during the perioperative period after ACL injury from the perspective of preoperative management, intraoperative application and postoperative rehabilitation. Preoperative management focuses on preoperative diagnosis and risk prediction; Intraoperative applications, particularly intraoperative navigation, to help identify important anatomical landmarks during surgery; Postoperative rehabilitation, assisting patients with functional exercise and preventing recurrence of injury.

Mantle cell lymphoma (MCL) has both aggressive and indolent characteristics, strong heterogeneity and poor prognosis. Regarding MCL, there are a number of new research advances in 2024. This article focuses on MCL prognostic model optimization and new targeted combination therapy strategies, focusing on the identification of high-risk characteristics such as TP53 mutations and blastotypes, as well as the combined application of Bruton's tyrosine kinase (BTK) inhibitors, B-cell lymphoma-2 protein (BCL-2) inhibitors, immunomodulators and other drugs in new and recurrent MCL, providing a reference for precise treatment.