Publications

Osteoporosis is the most common chronic metabolic bone disease worldwide. Vertebral compression fracture (VCF) is the most common type of osteoporotic fracture. Approximately 700,000 osteoporotic VCFs are diagnosed annually in the USA alone, resulting in an annual economic burden of  $13.8B. With an aging population, the rate of osteoporotic VCFs and their associated burdens are expected to rise. Those burdens include pain, functional impairment, and increased medical expenditure. Therefore, it is of utmost importance to develop an analytical tool to aid in the identification of VCFs. Computed Tomography (CT) imaging is commonly used to detect occult injuries. Unlike the existing VCF detection approaches based on CT, the standard clinical criteria for determining VCF relies on the shape of vertebrae, such as loss of vertebral body height. We developed a novel automated vertebrae localization, segmentation, and osteoporotic VCF detection pipeline for CT scans using state-of-the-art deep learning models to bridge this gap. To do so, we employed a publicly available dataset of spine CT scans with 325 scans annotated for segmentation, 126 of which also graded for VCF (81 with VCFs and 45 without VCFs). Our approach attained 96% sensitivity and 81% specificity in detecting VCF at the vertebral-level, and 100% accuracy at the subject-level, outperforming deep learning counterparts tested for VCF detection without segmentation. Crucially, we showed that adding predicted vertebrae segments as inputs significantly improved VCF detection at both vertebral and subject levels by up to 14% Sensitivity and 20% Specificity (p-value = 0.028).

Fibrotic changes in musculoskeletal diseases arise from the abnormal buildup of fibrotic tissue around the joints, leading to limited mobility, compromised joint function, and diminished quality of life. Relaxin (RLX) attenuates fibrosis by accelerating collagen degradation and inhibiting excessive extracellular matrix (ECM) production. Further, RLX disrupts myofibroblast activation by modulating the TGF-β/Smads signaling pathways, which reduces connective tissue fibrosis. However, the mechanisms and effects of RLX in musculoskeletal pathologies are emerging as increasing research focuses on relaxin's impact on skin, ligaments, tendons, cartilage, joint capsules, connective tissues, and muscles. This review delineates the actions of relaxin within the musculoskeletal system and the challenges to its clinical application. Relaxin shows significant potential in both in vivo and in vitro studies for broadly managing musculoskeletal fibrosis; however, challenges such as short biological half-life and sex-specific responses may pose hurdles for clinical use.

In the 1960s, the American Society for Surgery of the Hand embarked on an endeavor to improve and standardize the educational experience in hand surgery. By the 1980s, numerous programs existed across the country with the Accreditation Council for Graduate Medical Education formally recognizing orthopedic surgery-based fellowships in 1985 and plastic surgery-based fellowships in 1986. In order to sit for what was then termed the Certificate of Additional Qualification examination, applicants had to demonstrate performance of a specific number of procedures while in practice. Borrowing from this theme, the Accreditation Council for Graduate Medical Education began to analyze programs according to the relative proportion of cases done by fellows at individual institutions compared to national trends. Beginning in 2019 and working collaboratively with the Accreditation Council for Graduate Medical Education, the Hand Fellowship Director's Association has since modified the methods by which programs are evaluated, pivoting away from comparative percentages to the establishment of case minimums. The development of this process has been iterative with the resultant outcome being an evaluation system that focuses on educational quality and technical proficiency over sheer numerical volume.

BACKGROUND: Stem fixation in reconstruction after resection of femoral tumors is debated. Cemented stems offer immediate stability but risk aseptic loosening, while press-fit stems allow bone ingrowth but risk stress shielding and subsidence. Our retrospective review aimed to determine implant failure rates and their associated factors, as well as the rates of infection, debridement, and mortality for both fixation groups (cemented or press-fit stems) used in patients undergoing resection of femoral tumor disease and subsequent arthroplasty.

METHODS: We retrospectively studied 252 patients who underwent resection of femoral tumors and subsequent arthroplasty using cemented (n = 173; 69%) or press-fit (noncemented) (n = 79; 31%) stems between 1999 and 2020. Implant failure was the primary outcome, with secondary outcomes including rates of implant infection, debridement, and mortality. Multivariable regression was done to assess risk factors for implant failures.

RESULTS: The study found implant failure rates of 11% and 18% for cemented stems and press-fit stems, respectively. Lower stem to diaphyseal ratios ( P = 0.024) and younger patients ( P = 0.008) were associated with a higher risk of implant failure in cemented stems. The infection rates were 14% and 10% for cemented and press-fit stems, respectively. Debridement rates were 16% and 13% for cemented and press-fit stems, respectively, while the 1-year mortality rate was 16% for cemented stems and 1.5% for press-fit stems.

CONCLUSIONS: This study is the largest of its kind, providing patient characteristics and outcomes in both cemented and press-fit stems in the setting of reconstruction for femoral tumors. Both methods can be effective, with outcomes dependent on patient-specific factors, such as life expectancy, activity level, and body habitus, as well as proper implant fit. Additional studies of both implants and longer follow-up are required to elucidate the optimal fixation method for each individual patient.

LEVEL OF EVIDENCE: Level III, retrospective noncomparative study.

In hand trauma, the uninjured forearm has been touted as the ideal site for ectopic banking in digit/hand amputations. Here, we describe the temporary ectopic implantation and subsequent replantation of a partially amputated hand and highlight the "Three R's" - Recovery, Rehabilitation, and Revision over the first year of recovery.

Hip fractures exceed 250,000 cases annually in the United States, with the worldwide incidence projected to increase by 240-310% by 2050. Hip fractures are predominantly diagnosed by radiologist review of radiographs. In this study, we developed a deep learning model by extending the VarifocalNet Feature Pyramid Network (FPN) for detection and localization of proximal femur fractures from plain radiography with clinically relevant metrics. We used a dataset of 823 hip radiographs of 150 subjects with proximal femur fractures and 362 controls to develop and evaluate the deep learning model. Our model attained 0.94 specificity and 0.95 sensitivity in fracture detection over the diverse imaging dataset. We compared the performance of our model against five benchmark FPN models, demonstrating 6-14% sensitivity and 1-9% accuracy improvement. In addition, we demonstrated that our model outperforms a state-of-the-art transformer model based on DINO network by 17% sensitivity and 5% accuracy, while taking half the time on average to process a radiograph. The developed model can aid radiologists and support on-premise integration with hospital cloud services to enable automatic, opportunistic screening for hip fractures.

The supraspinatus tendon plays a crucial role in shoulder abduction, making it one of the common structures affected by injury. Clinically, crescent-shaped tears are the most commonly seen tear shape. By developing six specimen-specific, three-dimensional, supraspinatus-infraspinatus finite element model with heterogeneous material properties, this study aimed to examine the changes in tissue deformation (maximum principal strain) of the supraspinatus tendon due to specimen-specific material properties and rotator cuff tear size. FE models with small- and medium-sized full-thickness crescent-shaped tears were subjected to loads seen during activities of daily living and physiotherapy. Six fresh-frozen cadaveric shoulders were dissected to mechanically test the supraspinatus tendon and develop and validate FE models that can be used to assess changes in strain due to small (< 1 cm, equivalent to 20-30% of the tendon width) and medium-sized (1-3 cm, equivalent to 40-50% of the tendon width) tears that are located in the middle and posterior regions of the supraspinatus tendon. FE predictions of maximum principal strain at the tear tips were examined to determine whether failure strain was reached during activities of daily living (drinking and brushing teeth) and physiotherapy exercises (prone abduction and external rotation at 90° abduction). No significant differences were observed between the middle and posterior tear failure loads for small- and medium-sized tears. For prone abduction, there was a potential risk for tear progression (exceeded failure strain) for medium-sized tears in the supraspinatus tendon's middle and posterior regions. For external rotation at 90° abduction, one model with a middle tear and two with posterior tears experienced failure. For all daily activity loads, the strain never exceeded the failure strain. Our three-dimensional supraspinatus-infraspinatus FE model shows that small tears appear unlikely to progress based on the regional strain response; however, medium-sized tears are at higher risk during more strenuous physiotherapy exercises. Furthermore, differences in patient-specific tendon material properties are important in determining whether the tear will progress. Therefore, patient-specific management plans based on tear size may be beneficial to improve clinical outcomes.

Background and Objectives: Despite the promise of phage therapy (PT), its efficacy in prosthetic joint infection (PJI) management is unknown. Much of the current literature is largely limited to case reports and series. Materials and Methods: In order to help inform power calculations for future clinical trials and comparative analyses, we performed a systematic review and proportional meta-analysis of early PT outcomes to provide a preliminary assessment of early phage therapy treatment outcomes for cases of PJI. Results: In a search of available literature across MEDLINE (Ovid, Wolters Kluwer, Alphen aan den Rijn, The Netherlands), Embase (Elsevier, Amsterdam, The Netherlands), the Web of Science Core Collection (Clarivate, London, UK), and Cochrane Central (Wiley, Hoboken, NJ, USA) up to 23 September 2023, we identified 37 patients with PJIs receiving adjunctive PT. Patients most frequently reported Staphylococcal species infection (95%) and intraarticular phage delivery (73%). Phage cocktail (65%) and antibiotic co-administration (97%) were common. A random-effects proportional meta-analysis suggested infection remission in 78% of patients (95% CI: 39%, 95%) (I2 = 55%, p = 0.08) and 83% with a minimum 12-month follow-up (95% CI: 53%, 95%) (I2 = 26%, p = 0.26). Conclusions: Our study provides a preliminary estimate of PT's efficacy in PJIs and informs future comparative studies.

Spine disorders can cause severe functional limitations, including back pain, decreased pulmonary function, and increased mortality risk. Plain radiography is the first-line imaging modality to diagnose suspected spine disorders. Nevertheless, radiographical appearance is not always sufficient due to highly variable patient and imaging parameters, which can lead to misdiagnosis or delayed diagnosis. Employing an accurate automated detection model can alleviate the workload of clinical experts, thereby reducing human errors, facilitating earlier detection, and improving diagnostic accuracy. To this end, deep learning-based computer-aided diagnosis (CAD) tools have significantly outperformed the accuracy of traditional CAD software. Motivated by these observations, we proposed a deep learning-based approach for end-to-end detection and localization of spine disorders from plain radiographs. In doing so, we took the first steps in employing state-of-the-art transformer networks to differentiate images of multiple spine disorders from healthy counterparts and localize the identified disorders, focusing on vertebral compression fractures (VCF) and spondylolisthesis due to their high prevalence and potential severity. The VCF dataset comprised 337 images, with VCFs collected from 138 subjects and 624 normal images collected from 337 subjects. The spondylolisthesis dataset comprised 413 images, with spondylolisthesis collected from 336 subjects and 782 normal images collected from 413 subjects. Transformer-based models exhibited 0.97 Area Under the Receiver Operating Characteristic Curve (AUC) in VCF detection and 0.95 AUC in spondylolisthesis detection. Further, transformers demonstrated significant performance improvements against existing end-to-end approaches by 4-14% AUC (p-values < 10-13) for VCF detection and by 14-20% AUC (p-values < 10-9) for spondylolisthesis detection.

BACKGROUND: Treatment strategies for meniscal tears range from nonoperative management to surgical intervention. However, national trends in cost-related outcomes and patient factors related to the failure of nonoperative management remain poorly understood.

PURPOSE: To describe the costs associated with nonoperative versus operative management of meniscal tears in the 2 years after diagnosis and examine the relationship between patient characteristics and timing of surgery.

STUDY DESIGN: Cross-sectional study; Level of evidence, 3.

METHODS: This study was conducted using the MarketScan databases. Patients diagnosed with a meniscal tear without concomitant knee osteoarthritis between January 1 and December 31, 2017, were included. The primary outcome was the total cost of meniscal tear-related procedures-including insurance deductibles, coinsurance, and net insurance payments-in the 2 years after diagnosis. Procedures included were as follows: (1) surgery-including meniscectomy or meniscal repair; (2) physical therapy; (3) medication-including nonsteroidal anti-inflammatories, opioids, and acetaminophen; (4) intra-articular injections-including professional fee, hyaluronic acid, and corticosteroids; (5) imaging; and (6) clinic visits to orthopaedic specialists. Patients were grouped as having undergone early surgery (ES) (≤3 months of diagnosis), late surgery (LS) (>3 months after diagnosis), or no surgery (NS). Multivariate logistic regression was performed to determine the likelihood of undergoing surgery early and failing nonoperative treatment.

RESULTS: The study population included 29,924 patients with a mean age of 43.9 ± 12.9 years (ES: n = 9507 (31.8%); LS: n = 2021 (6.8%); NS: n = 18,396 (61.5%)). Complex (36.6%) and medial (58.8%) meniscal tears were the most common type and location of injuries, respectively. The mean cost of management per patient was $3835 ± $4795. Costs were lower in the NS group ($1905 ± $3175) compared with the ES group ($6759 ± $5155), while the highest costs were observed in the LS group ($7649 ± $5913) (P < .001). Patients who were men, >40 years, and with a bucket-handle or lateral meniscal tear were more likely to undergo surgery early. Patients who were men, <30 years, and with a complex tear or tear to the lateral meniscus were more likely to fail nonoperative management.

CONCLUSION: Nonoperative management had the lowest cost burden and should be recommended for patients with appropriate indications. However, if surgery is necessary, it should be performed earlier.