Domestic and International Cosmetic Tourism Complications Presenting to a U.S. Tertiary Hospital.

BACKGROUND: Cosmetic surgery tourism has become a significant global industry. Oftentimes, patients who develop postoperative complications present for care in their home U.S. state. OBJECTIVES: This study evaluated patients who either traveled abroad or to other states within the United States for cosmetic surgeries and returned with complications treated in the authors' center. We sought to compare rates of complications between patients that underwent cosmetic surgery internationally and domestically. METHODS: This retrospective cross-sectional study reviewed patients who presented from June 2014 to June 2022 with concerns related to cosmetic surgeries performed in another state or abroad. Binary logistic regressions were performed to assess differences in outcomes between domestic and international cases, including complications, interventions, and admissions. RESULTS: One-hundred twenty-three patients (97.6% female, me an age 34.0 ± 8.7 years, range 16-62 years) comprised 159 emergency department consultations. The most common procedures included abdominoplasty (n=72) and liposuction (n=56). Complications included wound dehiscence (n=39), infection (n=38), and seroma (n=34). Over one-half of patients required intervention. Twenty-nine patients (23.6%) required hospital admission. On multivariate regression analyses, incidence of seroma (p=0.025) and oral (p=0.036) and intravenous antibiotic prescriptions (p=0.045) were significantly greater among the international cohort compared to domestic, whereas all other complication variables were non-significant. There were no other significant differences in operative interventions or hospital admissions between international and domestic cohorts. CONCLUSIONS: Compared to domestic tourism cases, international tourism cases were associated with significantly higher rates of seroma formation and antibiotic use. There were no significant differences otherwise in overall complications including infections, operative interventions, or hospital admissions.

Super-resolution neural networks improve the spatiotemporal resolution of adaptive MRI-guided radiation therapy.

BACKGROUND: Magnetic resonance imaging (MRI) offers superb non-invasive, soft tissue imaging of the human body. However, extensive data sampling requirements severely restrict the spatiotemporal resolution achievable with MRI. This limits the modality's utility in real-time guidance applications, particularly for the rapidly growing MRI-guided radiation therapy approach to cancer treatment. Recent advances in artificial intelligence (AI) could reduce the trade-off between the spatial and the temporal resolution of MRI, thus increasing the clinical utility of the imaging modality. METHODS: We trained deep learning-based super-resolution neural networks to increase the spatial resolution of real-time MRI. We developed a framework to integrate neural networks directly onto a 1.0 T MRI-linac enabling real-time super-resolution imaging. We integrated this framework with the targeting system of the MRI-linac to demonstrate real-time beam adaptation with super-resolution-based imaging. We tested the integrated system using large publicly available datasets, healthy volunteer imaging, phantom imaging, and beam tracking experiments using bicubic interpolation as a baseline comparison. RESULTS: Deep learning-based super-resolution increases the spatial resolution of real-time MRI across a variety of experiments, offering measured performance benefits compared to bicubic interpolation. The temporal resolution is not compromised as measured by a real-time adaptation latency experiment. These two effects, an increase in the spatial resolution with a negligible decrease in the temporal resolution, leads to a net increase in the spatiotemporal resolution. CONCLUSIONS: Deployed super-resolution neural networks can increase the spatiotemporal resolution of real-time MRI. This has applications to domains such as MRI-guided radiation therapy and interventional procedures.

Magnetic resonance imaging (MRI) is a medical imaging modality that is used to image organs such as the brain, lungs, and liver as well as diseases such as cancer. MRI scans taken at high resolution are of overly long duration. This time constraint limits the accuracy of MRI-guided cancer radiation therapy, where imaging must be fast to adapt treatment to tumour motion. Here, we deployed artificial intelligence (AI) models to achieve fast and high detail MRI. We additionally validated our AI models across various scenarios. These AI-based models could potentially enable people with cancer to be treated with higher accuracy and precision.

The Impact of a Centralized Plastic Surgery Research Infrastructure on Scholarly Productivity and Output.

Although research and innovation is a key within the field of plastic and reconstructive surgery, the impact of team structure, interpersonal dynamics, and/or standardized infrastructure on scholarly output has been infrequently studied. In this work, we present the formation and implementation of a novel plastic surgery research program that aims to unite previously disparate clinical and translational research efforts at our institution to facilitate critical inquiry. From July 2022 to June 2023, our department launched a pilot research program based on three pillars: (1) formalization of a research curriculum (monthly research meetings for agenda setting and discussion for project honing, formal research leadership for meeting facilitation and workflow regulation), (2) development of a centralized database to compile ongoing research (Google Drive repository to house all ongoing research documents, facilitate real-time editing, and provide resources/templates for assisting in the research process), and (3) bolstering of a core research identity built on mentorship and collaboration (more frequent interactions to shift previously siloed faculty-student mentorship into a robust milieu of intercollaboration). During the first year, we saw an increased number of publications and presentations, as well as robust participation and contribution from faculty, residents, and medical students. Future directions will focus on addressing resource limitation, such as project idea availability and funding, to sustain the success and growth of this novel research infrastructure.

Alpha Defensin-1 Level Correlates with Periprosthetic Infection Severity following Implant-based Breast Reconstruction.

Background: Accurate diagnosis of periprosthetic infections following breast reconstructions is paramount to reduce morbidity. Alpha defensin-1 (AD-1) is an antimicrobial peptide released by neutrophils. This study evaluates the relationship between quantitative AD-1 levels and infection severity in patients with suspected periprosthetic infection. Methods: Retrospective review was conducted of patients with prior breast implant reconstruction undergoing surgery for either suspected infection or prosthesis exchange and revision. The AD-1 level in periprosthetic fluid was sent for quantitative analysis. Association between AD-1 levels with outcomes, management, systemic markers of infection, and overall infection severity was evaluated. Results: Thirty-eight breasts were included. Infected breasts had higher AD-1 levels (3.91 versus 0.14, P < 0.01), greater odds of erythema [odds ratio (OR) 2.98 (1.53-5.82), P = 0.01], purulence [OR 2.84 (1.51-5.35), P = 0.01], fever [OR 1.84 (1.15-2.93), P = 0.01], threatened implant exposure [OR 2.97 (1.48-5.95), P < 0.01], and true implant exposure [OR 1.79 (1.04-3.08), P = 0.04]. Increasing AD-1 was an independent risk factor for washout (P < 0.01), and explant [OR 2.48 (1.47-4.2), P < 0.01]. AD-1 positively correlated with white blood cell count (ß = 1.81 cells/µL, P < 0.01), and serum lactate (ß = 0.19 meq/L, P < 0.04). Increasing AD-1 level was an independent predictor of infection severity (χ² = 22.77, P < 0.01). Conclusions: AD-1 levels correlate with infection severity, highlighting its potential both when clinical examination is ambiguous and when treatment response is being monitored. Although further evaluation is warranted, AD-1 may demonstrate utility in novel breast implant salvage algorithms.

Systematic Review of Prophylactic Plastic Surgery Closure to Prevent Postoperative Wound Complications Following Spine Surgery.

Spinal surgeries are increasingly performed in the United States, but complication rates can be unacceptably high at up to 26%. Consequently, plastic surgeons (PS) are sometimes recruited by spine surgeons (SS) for intraoperative assistance with soft tissue closures. An electronic multidatabase literature search was systematically conducted to determine whether spinal wound closure performed by PS minimizes postoperative wound healing complications when compared to closure by SS (neurosurgical or orthopedic), with the hypothesis that closures by PS minimizes incidence of complications. All published studies involving patients who underwent posterior spinal surgery with closure by PS or SS at index spine surgery were identified. Filtering by exclusion criteria identified 10 studies, 4 of which were comparative in nature and included both closures by PS and SS. Of these 4, none reported significant differences in postoperative outcomes between the groups. Across all studies, PS were involved in cases with higher baseline risk for wound complications and greater comorbidity burden. Closures by PS were significantly more likely to have had prior chemotherapy in 2 of the 4 (50%) studies (P = 0.014, P < 0.001) and radiation in 3 of the 4 (75%) studies (P < 0.001, P < 0.01, P < 0.001). In conclusion, closures by PS are frequently performed in higher risk cases, and use of PS in these closures may normalize the risk of wound complications to that of the normal risk cohort, though the overall level of evidence of the published literature is low.

Genomic landscape of patients with germline RUNX1 variants and familial platelet disorder with myeloid malignancy.

ABSTRACT: Familial platelet disorder with associated myeloid malignancies (FPDMM) is caused by germline RUNX1 mutations and characterized by thrombocytopenia and increased risk of hematologic malignancies. We recently launched a longitudinal natural history study for patients with FPDMM. Among 27 families with research genomic data by the end of 2021, 26 different germline RUNX1 variants were detected. Besides missense mutations enriched in Runt homology domain and loss-of-function mutations distributed throughout the gene, splice-region mutations and large deletions were detected in 6 and 7 families, respectively. In 25 of 51 (49%) patients without hematologic malignancy, somatic mutations were detected in at least 1 of the clonal hematopoiesis of indeterminate potential (CHIP) genes or acute myeloid leukemia (AML) driver genes. BCOR was the most frequently mutated gene (in 9 patients), and multiple BCOR mutations were identified in 4 patients. Mutations in 6 other CHIP- or AML-driver genes (TET2, DNMT3A, KRAS, LRP1B, IDH1, and KMT2C) were also found in ≥2 patients without hematologic malignancy. Moreover, 3 unrelated patients (1 with myeloid malignancy) carried somatic mutations in NFE2, which regulates erythroid and megakaryocytic differentiation. Sequential sequencing data from 19 patients demonstrated dynamic changes of somatic mutations over time, and stable clones were more frequently found in older adult patients. In summary, there are diverse types of germline RUNX1 mutations and high frequency of somatic mutations related to clonal hematopoiesis in patients with FPDMM. Monitoring changes in somatic mutations and clinical manifestations prospectively may reveal mechanisms for malignant progression and inform clinical management. This trial was registered at www.clinicaltrials.gov as #NCT03854318.

Real-time motion management in MRI-guided radiotherapy: Current status and AI-enabled prospects.

MRI-guided radiotherapy (MRIgRT) is a highly complex treatment modality, allowing adaptation to anatomical changes occurring from one treatment day to the other (inter-fractional), but also to motion occurring during a treatment fraction (intra-fractional). In this vision paper, we describe the different steps of intra-fractional motion management during MRIgRT, from imaging to beam adaptation, and the solutions currently available both clinically and at a research level. Furthermore, considering the latest developments in the literature, a workflow is foreseen in which motion-induced over- and/or under-dosage is compensated in 3D, with minimal impact to the radiotherapy treatment time. Considering the time constraints of real-time adaptation, a particular focus is put on artificial intelligence (AI) solutions as a fast and accurate alternative to conventional algorithms.

Experimental comparison of linear regression and LSTM motion prediction models for MLC-tracking on an MRI-linac.

BACKGROUND: Magnetic resonance imaging (MRI)-guided radiotherapy with multileaf collimator (MLC)-tracking is a promising technique for intra-fractional motion management, achieving high dose conformality without prolonging treatment times. To improve beam-target alignment, the geometric error due to system latency should be reduced by using temporal prediction. PURPOSE: To experimentally compare linear regression (LR) and long-short-term memory (LSTM) motion prediction models for MLC-tracking on an MRI-linac using multiple patient-derived traces with different complexities. METHODS: Experiments were performed on a prototype 1.0 T MRI-linac capable of MLC-tracking. A motion phantom was programmed to move a target in superior-inferior (SI) direction according to eight lung cancer patient respiratory motion traces. Target centroid positions were localized from sagittal 2D cine MRIs acquired at 4 Hz using a template matching algorithm. The centroid positions were input to one of four motion prediction models. We used (1) a LSTM network which had been optimized in a previous study on patient data from another cohort (offline LSTM). We also used (2) the same LSTM model as a starting point for continuous re-optimization of its weights during the experiment based on recent motion (offline+online LSTM). Furthermore, we implemented (3) a continuously updated LR model, which was solely based on recent motion (online LR). Finally, we used (4) the last available target centroid without any changes as a baseline (no-predictor). The predictions of the models were used to shift the MLC aperture in real-time. An electronic portal imaging device (EPID) was used to visualize the target and MLC aperture during the experiments. Based on the EPID frames, the root-mean-square error (RMSE) between the target and the MLC aperture positions was used to assess the performance of the different motion predictors. Each combination of motion trace and prediction model was repeated twice to test stability, for a total of 64 experiments. RESULTS: The end-to-end latency of the system was measured to be (389 ± 15) ms and was successfully mitigated by both LR and LSTM models. The offline+online LSTM was found to outperform the other models for all investigated motion traces. It obtained a median RMSE over all traces of (2.8 ± 1.3) mm, compared to the (3.2 ± 1.9) mm of the offline LSTM, the (3.3 ± 1.4) mm of the online LR and the (4.4 ± 2.4) mm when using the no-predictor. According to statistical tests, differences were significant (p-value <0.05) among all models in a pair-wise comparison, but for the offline LSTM and online LR pair. The offline+online LSTM was found to be more reproducible than the offline LSTM and the online LR with a maximum deviation in RMSE between two measurements of 10%. CONCLUSIONS: This study represents the first experimental comparison of different prediction models for MRI-guided MLC-tracking using several patient-derived respiratory motion traces. We have shown that among the investigated models, continuously re-optimized LSTM networks are the most promising to account for the end-to-end system latency in MRI-guided radiotherapy with MLC-tracking.

Natural history study of patients with familial platelet disorder with associated myeloid malignancy.

ABSTRACT: Deleterious germ line RUNX1 variants cause the autosomal dominant familial platelet disorder with associated myeloid malignancy (FPDMM), characterized by thrombocytopenia, platelet dysfunction, and a predisposition to hematologic malignancies (HMs). We launched a FPDMM natural history study and, from January 2019 to December 2021, enrolled 214 participants, including 111 patients with 39 different RUNX1 variants from 45 unrelated families. Seventy of 77 patients had thrombocytopenia, 18 of 18 had abnormal platelet aggregometry, 16 of 35 had decreased platelet dense granules, and 28 of 55 had abnormal bleeding scores. Nonmalignant bone marrows showed increased numbers of megakaryocytes in 12 of 55 patients, dysmegakaryopoiesis in 42 of 55, and reduced cellularity for age in 30 of 55 adult and 17 of 21 pediatric cases. Of 111 patients, 19 were diagnosed with HMs, including myelodysplastic syndrome, acute myeloid leukemia, chronic myelomonocytic leukemia, acute lymphoblastic leukemia, and smoldering myeloma. Of those 19, 18 were relapsed or refractory to upfront therapy and referred for stem cell transplantation. In addition, 28 of 45 families had at least 1 member with HM. Moreover, 42 of 45 patients had allergic symptoms, and 24 of 30 had gastrointestinal (GI) symptoms. Our results highlight the importance of a multidisciplinary approach, early malignancy detection, and wider awareness of inherited disorders. This actively accruing, longitudinal study will genotype and phenotype more patients with FPDMM, which may lead to a better understanding of the disease pathogenesis and clinical course, which may then inform preventive and therapeutic interventions. This trial was registered at www.clinicaltrials.gov as #NCT03854318.

SOX9 is a key component of RUNX2-regulated transcriptional circuitry in osteosarcoma.

BACKGROUND: The absence of prominent, actionable genetic alternations in osteosarcomas (OS) implies that transcriptional and epigenetic mechanisms significantly contribute to the progression of this life-threatening form of cancer. Therefore, the identification of potential transcriptional events that promote the survival of OS cells could be key in devising targeted therapeutic approaches for OS. We have previously shown that RUNX2 is a transcription factor (TF) essential for OS cell survival. Unfortunately, the transcriptional network or circuitry regulated by RUNX2 in OS cells is still largely unknown. METHODS: The TFs that are in the RUNX2 transcriptional circuitry were identified by analyzing RNAseq and ChIPseq datasets of RUNX2. To evaluate the effect of SOX9 knockdown on the survival of osteosarcoma cells in vitro, we employed cleaved caspase-3 immunoblotting and propidium iodide staining techniques. The impact of SOX9 and JMJD1C depletion on OS tumor growth was examined in vivo using xenografts and immunohistochemistry. Downstream targets of SOX9 were identified and dissected using RNAseq, pathway analysis, and gene set enrichment analysis. Furthermore, the interactome of SOX9 was identified using BioID and validated by PLA. RESULT: Our findings demonstrate that SOX9 is a critical TF that is induced by RUNX2. Both in vitro and in vivo experiments revealed that SOX9 plays a pivotal role in the survival of OS. RNAseq analysis revealed that SOX9 activates the transcription of MYC, a downstream target of RUNX2. Mechanistically, our results suggest a transcriptional network involving SOX9, RUNX2, and MYC, with SOX9 binding to RUNX2. Moreover, we discovered that JMJD1C, a chromatin factor, is a novel binding partner of SOX9, and depletion of JMJD1C impairs OS tumor growth. CONCLUSION: The findings of this study represent a significant advancement in our understanding of the transcriptional network present in OS cells, providing valuable insights that may contribute to the development of targeted therapies for OS.

Nano X Image Guidance in radiation therapy: feasibility study protocol for cone beam computed tomography imaging with gravity-induced motion.

BACKGROUND: This paper describes the protocol for the Nano X Image Guidance (Nano X IG) trial, a single-institution, clinical imaging study. The Nano X is a prototype fixed-beam radiotherapy system developed to investigate the feasibility of a low-cost, compact radiotherapy system to increase global access to radiation therapy. This study aims to assess the feasibility of volumetric image guidance with cone beam computed tomography (CBCT) acquired during horizontal patient rotation on the Nano X radiotherapy system. METHODS: In the Nano X IG study, we will determine whether radiotherapy image guidance can be performed with the Nano X radiotherapy system where the patient is horizontally rotated while scan projections are acquired. We will acquire both conventional CBCT scans and Nano X CBCT scans for 30 patients aged 18 and above and receiving radiotherapy for head/neck or upper abdomen cancers. For each patient, a panel of experts will assess the image quality of Nano X CBCT scans against conventional CBCT scans. Each patient will receive two Nano X CBCT scans to determine the image quality reproducibility, the extent and reproducibility of patient motion and assess patient tolerance. DISCUSSION: Fixed-beam radiotherapy systems have the potential to help ease the current shortfall and increase global access to radiotherapy treatment. Advances in image guidance could facilitate fixed-beam radiotherapy using horizontal patient rotation. The efficacy of this radiotherapy approach is dependent on our ability to image and adapt to motion due to rotation and for patients to tolerate rotation during treatment. TRIAL REGISTRATION: ClinicalTrials.gov, NCT04488224. Registered on 27 July 2020.

Distortion-corrected image reconstruction with deep learning on an MRI-Linac.

PURPOSE: MRI is increasingly utilized for image-guided radiotherapy due to its outstanding soft-tissue contrast and lack of ionizing radiation. However, geometric distortions caused by gradient nonlinearities (GNLs) limit anatomical accuracy, potentially compromising the quality of tumor treatments. In addition, slow MR acquisition and reconstruction limit the potential for effective image guidance. Here, we demonstrate a deep learning-based method that rapidly reconstructs distortion-corrected images from raw k-space data for MR-guided radiotherapy applications. METHODS: We leverage recent advances in interpretable unrolling networks to develop a Distortion-Corrected Reconstruction Network (DCReconNet) that applies convolutional neural networks (CNNs) to learn effective regularizations and nonuniform fast Fourier transforms for GNL-encoding. DCReconNet was trained on a public MR brain dataset from 11 healthy volunteers for fully sampled and accelerated techniques, including parallel imaging (PI) and compressed sensing (CS). The performance of DCReconNet was tested on phantom, brain, pelvis, and lung images acquired on a 1.0T MRI-Linac. The DCReconNet, CS-, PI-and UNet-based reconstructed image quality was measured by structural similarity (SSIM) and RMS error (RMSE) for numerical comparisons. The computation time and residual distortion for each method were also reported. RESULTS: Imaging results demonstrated that DCReconNet better preserves image structures compared to CS- and PI-based reconstruction methods. DCReconNet resulted in the highest SSIM (0.95 median value) and lowest RMSE (<0.04) on simulated brain images with four times acceleration. DCReconNet is over 10-times faster than iterative, regularized reconstruction methods. CONCLUSIONS: DCReconNet provides fast and geometrically accurate image reconstruction and has the potential for MRI-guided radiotherapy applications.

Alpha Defensin-1 Biomarker Outperforms Culture in Diagnosing Breast Implant-Related Infection: Results from a Multicenter Prospective Study.

BACKGROUND: Prompt diagnosis of breast implant infection is critical to reducing morbidity. A high incidence of false-negative microbial culture mandates superior testing modalities. Alpha defensin-1 (AD-1), an infection biomarker, has outperformed culture in diagnosing periprosthetic joint infection with sensitivity/specificity of 97%. After previously demonstrating its feasibility in breast implant-related infection (BIRI), this case-control study compares the accuracy of AD-1 to microbial culture in suspected BIRI. METHODS: An institutional review board-approved, prospective, multicenter study was conducted of adults with prior breast implant reconstruction undergoing surgery for suspected infection (cases) or prosthetic exchange/revision (controls). Demographics, perioperative characteristics, antibiotic exposure, and implant pocket fluid were collected. Fluid samples underwent microbial culture, AD-1 assay, and adjunctive markers (C-reactive protein, lactate, cell differential); diagnostic performance was assessed by means of sensitivity, specificity, and accuracy from receiver operating characteristic curve analysis, with values of P < 0.05 considered significant. RESULTS: Fifty-three implant pocket samples were included (cases, n = 20; controls, n = 33). All 20 patients with suspected BIRI exhibited cellulitis, 65% had abnormal drainage, and 55% were febrile. All suspected BIRIs were AD-1 positive (sensitivity, 100%). Microbial culture failed to grow any microorganisms in four BIRIs (sensitivity, 80%; P = 0.046); Gram stain was least accurate (sensitivity, 25%; P < 0.001). All tests demonstrated 100% specificity. Receiver operating characteristic curve analyses yielded the following areas under the curve: AD-1, 1.0; microbial culture, 0.90 ( P = 0.029); and Gram stain, 0.62 ( P < 0.001). Adjunctive markers were significantly higher among infections versus controls ( P < 0.001). CONCLUSIONS: Study findings confirm the accuracy of AD-1 in diagnosing BIRI and indicate superiority to microbial culture. Although further study is warranted, AD-1 may facilitate perioperative decision-making in BIRI management in a resource-efficient manner. CLINICAL QUESTION/LEVEL OF EVIDENCE: Diagnostic, II.

Genomic Landscape of Patients with Germline RUNX1 Variants and Familial Platelet Disorder with Myeloid Malignancy.

Germline RUNX1 mutations lead to familial platelet disorder with associated myeloid malignancies (FPDMM), which is characterized by thrombocytopenia and a life-long risk (35-45%) of hematological malignancies. We recently launched a longitudinal natural history study for patients with FPDMM at the NIH Clinical Center. Among 29 families with research genomic data, 28 different germline RUNX1 variants were detected. Besides missense mutations enriched in Runt homology domain and loss-of-function mutations distributed throughout the gene, splice-region mutations and large deletions were detected in 6 and 7 families, respectively. In 24 of 54 (44.4%) non-malignant patients, somatic mutations were detected in at least one of the clonal hematopoiesis of indeterminate potential (CHIP) genes or acute myeloid leukemia (AML) driver genes. BCOR was the most frequently mutated gene (in 9 patients), and multiple BCOR mutations were identified in 4 patients. Mutations in 7 other CHIP or AML driver genes ( DNMT3A, TET2, NRAS, SETBP1, SF3B1, KMT2C , and LRP1B ) were also found in more than one non-malignant patient. Moreover, three unrelated patients (one with myeloid malignancy) carried somatic mutations in NFE2 , which regulates erythroid and megakaryocytic differentiation. Sequential sequencing data from 19 patients demonstrated dynamic changes of somatic mutations over time, and stable clones were more frequently found in elderly patients. In summary, there are diverse types of germline RUNX1 mutations and high frequency of somatic mutations related to clonal hematopoiesis in patients with FPDMM. Monitoring dynamic changes of somatic mutations prospectively will benefit patients’ clinical management and reveal mechanisms for progression to myeloid malignancies. Key Points: Comprehensive genomic profile of patients with FPDMM with germline RUNX1 mutations. Rising clonal hematopoiesis related secondary mutations that may lead to myeloid malignancies.

Real-time radial reconstruction with domain transform manifold learning for MRI-guided radiotherapy.

BACKGROUND: MRI-guidance techniques that dynamically adapt radiation beams to follow tumor motion in real time will lead to more accurate cancer treatments and reduced collateral healthy tissue damage. The gold-standard for reconstruction of undersampled MR data is compressed sensing (CS) which is computationally slow and limits the rate that images can be available for real-time adaptation. PURPOSE: Once trained, neural networks can be used to accurately reconstruct raw MRI data with minimal latency. Here, we test the suitability of deep-learning-based image reconstruction for real-time tracking applications on MRI-Linacs. METHODS: We use automated transform by manifold approximation (AUTOMAP), a generalized framework that maps raw MR signal to the target image domain, to rapidly reconstruct images from undersampled radial k-space data. The AUTOMAP neural network was trained to reconstruct images from a golden-angle radial acquisition, a benchmark for motion-sensitive imaging, on lung cancer patient data and generic images from ImageNet. Model training was subsequently augmented with motion-encoded k-space data derived from videos in the YouTube-8M dataset to encourage motion robust reconstruction. RESULTS: AUTOMAP models fine-tuned on retrospectively acquired lung cancer patient data reconstructed radial k-space with equivalent accuracy to CS but with much shorter processing times. Validation of motion-trained models with a virtual dynamic lung tumor phantom showed that the generalized motion properties learned from YouTube lead to improved target tracking accuracy. CONCLUSION: AUTOMAP can achieve real-time, accurate reconstruction of radial data. These findings imply that neural-network-based reconstruction is potentially superior to alternative approaches for real-time image guidance applications.

Rapid distortion correction enables accurate magnetic resonance imaging-guided real-time adaptive radiotherapy.

Background and purpose: Magnetic resonance imaging (MRI)-Linac systems combine simultaneous MRI with radiation delivery, allowing treatments to be guided by anatomically detailed, real-time images. However, MRI can be degraded by geometric distortions that cause uncertainty between imaged and actual anatomy. In this work, we develop and integrate a real-time distortion correction method that enables accurate real-time adaptive radiotherapy. Materials and methods: The method was based on the pre-treatment calculation of distortion and the rapid correction of intrafraction images. A motion phantom was set up in an MRI-Linac at isocentre (P0 ), the edge (P 1) and just outside (P 2) the imaging volume. The target was irradiated and tracked during real-time adaptive radiotherapy with and without the distortion correction. The geometric tracking error and latency were derived from the measurements of the beam and target positions in the EPID images. Results: Without distortion correction, the mean geometric tracking error was 1.3 mm at P 1 and 3.1 mm at P 2. When distortion correction was applied, the error was reduced to 1.0 mm at P 1 and 1.1 mm at P 2. The corrected error was similar to an error of 0.9 mm at P0 where the target was unaffected by distortion indicating that this method has accurately accounted for distortion during tracking. The latency was 319 ± 12 ms without distortion correction and 335 ± 34 ms with distortion correction. Conclusions: We have demonstrated a real-time distortion correction method that maintains accurate radiation delivery to the target, even at treatment locations with large distortion.

Differential diagnosis of bone marrow failure syndromes guided by machine learning.

The choice to postpone treatment while awaiting genetic testing can result in significant delay in definitive therapies in patients with severe pancytopenia. Conversely, the misdiagnosis of inherited bone marrow failure (BMF) can expose patients to ineffectual and expensive therapies, toxic transplant conditioning regimens, and inappropriate use of an affected family member as a stem cell donor. To predict the likelihood of patients having acquired or inherited BMF, we developed a 2-step data-driven machine-learning model using 25 clinical and laboratory variables typically recorded at the initial clinical encounter. For model development, patients were labeled as having acquired or inherited BMF depending on their genomic data. Data sets were unbiasedly clustered, and an ensemble model was trained with cases from the largest cluster of a training cohort (n = 359) and validated with an independent cohort (n = 127). Cluster A, the largest group, was mostly immune or inherited aplastic anemia, whereas cluster B comprised underrepresented BMF phenotypes and was not included in the next step of data modeling because of a small sample size. The ensemble cluster A-specific model was accurate (89%) to predict BMF etiology, correctly predicting inherited and likely immune BMF in 79% and 92% of cases, respectively. Our model represents a practical guide for BMF diagnosis and highlights the importance of clinical and laboratory variables in the initial evaluation, particularly telomere length. Our tool can be potentially used by general hematologists and health care providers not specialized in BMF, and in under-resourced centers, to prioritize patients for genetic testing or for expeditious treatment.

National Analysis of Sternal Wound Complications and Readmissions After Coronary Bypass Surgery.

BACKGROUND: Wound complications are a cause for readmission after cardiac surgery. Health insurance status has been associated with poor postoperative outcomes. We investigate the association between health insurance status and post-CABG wound dehiscence or infection along with 30-day wound-related readmission using a national database. METHODS: We queried the National Readmissions Database for the year 2018 for patients aged 18 years or more undergoing multivessel coronary artery bypass graft surgery (CABG). Patients were subcategorized by health insurance status (private, Medicaid, Medicare, uninsured). Our primary outcomes were wound dehiscence or infection during the index admission and 30-day readmission after discharge for wound-related complications. RESULTS: In all, 131,976 patients met inclusion criteria: 32.7% private, 7.6% Medicaid, 59.3% Medicare, and 0.4% uninsured. Compared with patients having private insurance, Medicaid patients had greater odds of readmission for superficial wound dehiscence (odds ratio [OR] 2.11; 1.11-4.00; P = .022) and deep wound dehiscence (OR 2.11; 95% CI, 1.09-4.10; P = .026), as did Medicare patients (OR 2.34; 95% CI, 1.29-3.88; P = .004; and OR 3.23; 95% CI, 1.76-5.90; P = .001, respectively). Medicaid patients additionally had higher odds of readmission for superficial wound infection (OR 1.59; 95% CI, 1.11-4.00; P = .014). Compared with patients with private insurance, Medicaid patients had higher odds of deep wound dehiscence on index admission (OR 1.97; 95% CI, 1.02-3.83; P = .044), and Medicare patients had higher odds of superficial wound dehiscence (OR 2.55; 95% CI, 1.28-5.06; P = .001). CONCLUSIONS: Patients with Medicaid and Medicare had greater odds of readmission for wound complications and higher rates of wound dehiscence in their index admission. Further research is warranted to characterize factors driving readmission due to postsurgical wound complications in low socioeconomic status populations.

A RUNX1-FPDMM rhesus macaque model reproduces the human phenotype and predicts challenges to curative gene therapies.

Germ line loss-of-function heterozygous mutations in the RUNX1 gene cause familial platelet disorder with associated myeloid malignancies (FPDMM) characterized by thrombocytopenia and a life-long risk of hematological malignancies. Although gene therapies are being considered as promising therapeutic options, current preclinical models do not recapitulate the human phenotype and are unable to elucidate the relative fitness of mutation-corrected and RUNX1-heterozygous mutant hematopoietic stem and progenitor cells (HSPCs) in vivo long term. We generated a rhesus macaque with an FPDMM competitive repopulation model using CRISPR/Cas9 nonhomologous end joining editing in the RUNX1 gene and the AAVS1 safe-harbor control locus. We transplanted mixed populations of edited autologous HSPCs and tracked mutated allele frequencies in blood cells. In both animals, RUNX1-edited cells expanded over time compared with AAVS1-edited cells. Platelet counts remained below the normal range in the long term. Bone marrows developed megakaryocytic dysplasia similar to human FPDMM, and CD34+ HSPCs showed impaired in vitro megakaryocytic differentiation, with a striking defect in polyploidization. In conclusion, the lack of a competitive advantage for wildtype or control-edited HSPCs over RUNX1 heterozygous-mutated HSPCs long term in our preclinical model suggests that gene correction approaches for FPDMM will be challenging, particularly to reverse myelodysplastic syndrome/ acute myeloid leukemia predisposition and thrombopoietic defects.