Clinical Descriptors Research Articles

Segmentation-based quantitative measurements in renal CT imaging using deep learning

BackgroundRenal quantitative measurements are important descriptors for assessing kidney function. We developed a deep learning-based method for automated kidney measurements from computed tomography (CT) images.MethodsThe study datasets comprised potential kidney donors (n = 88), both contrast-enhanced (Dataset 1 CE) and noncontrast (Dataset 1 NC) CT scans, and test sets of contrast-enhanced cases (Test set 2, n = 18), cases from a photon-counting (PC)CT scanner reconstructed at 60 and 190 keV (Test set 3 PCCT, n = 15), and low-dose cases (Test set 4, n = 8), which were retrospectively analyzed to train, validate, and test two networks for kidney segmentation and subsequent measurements. Segmentation performance was evaluated using the Dice similarity coefficient (DSC). The quantitative measurements’ effectiveness was compared to manual annotations using the intraclass correlation coefficient (ICC).ResultsThe contrast-enhanced and noncontrast models demonstrated excellent reliability in renal segmentation with DSC of 0.95 (Test set 1 CE), 0.94 (Test set 2), 0.92 (Test set 3 PCCT) and 0.94 (Test set 1 NC), 0.92 (Test set 3 PCCT), and 0.93 (Test set 4). Volume estimation was accurate with mean volume errors of 4%, 3%, 6% mL (contrast test sets) and 4%, 5%, 7% mL (noncontrast test sets). Renal axes measurements (length, width, and thickness) had ICC values greater than 0.90 (p < 0.001) for all test sets, supported by narrow 95% confidence intervals.ConclusionTwo deep learning networks were shown to derive quantitative measurements from contrast-enhanced and noncontrast renal CT imaging at the human performance level.Relevance statementDeep learning-based networks can automatically obtain renal clinical descriptors from both noncontrast and contrast-enhanced CT images. When healthy subjects comprise the training cohort, careful consideration is required during model adaptation, especially in scenarios involving unhealthy kidneys. This creates an opportunity for improved clinical decision-making without labor-intensive manual effort.Key PointsTrained 3D UNet models quantify renal measurements from contrast and noncontrast CT.The models performed interchangeably to the manual annotator and to each other.The models can provide expert-level, quantitative, accurate, and rapid renal measurements.Graphical

Normative Modeling of Thalamic Nuclear Volumes and Characterization of Lateralized Volume Alterations in Alzheimer’s Disease Versus Schizophrenia

BackgroundThalamic nuclei facilitate a wide range of complex behaviors, emotions, and cognition and have been implicated in neuropsychiatric disorders including Alzheimer’s disease (AD) and schizophrenia (SCZ). The aim of this work was to establish novel normative models of thalamic nuclear volumes and their laterality indices and investigate their changes in SCZ and AD. MethodsVolumes of bilateral whole thalami and 10 thalamic nuclei were generated from T1 magnetic resonance imaging data using a state-of-the-art novel segmentation method in healthy control participants (n = 2374) and participants with early mild cognitive impairment (n = 211), late mild cognitive impairment (n = 113), AD (n = 88), and SCZ (n = 168). Normative models for each nucleus were generated from healthy control participants while controlling for sex, intracranial volume, and site. Extreme z-score deviations (|z| > 1.96) and z-score distributions were compared across phenotypes. z Scores were associated with clinical descriptors. ResultsIncreased infranormal and decreased supranormal z scores were observed in SCZ and AD. z Score shifts representing reduced volumes were observed in most nuclei in SCZ and AD, with strong overlap in the bilateral pulvinar, medial dorsal, and centromedian nuclei. Shifts were larger in AD, with evidence of a left-sided preference in early mild cognitive impairment while a predilection for right thalamic nuclei was observed in SCZ. The right medial dorsal nucleus was associated with disorganized thought and daily auditory verbal hallucinations. ConclusionsIn AD, thalamic nuclei are more severely and symmetrically affected while in SCZ, the right thalamic nuclei are more affected. We highlight the right medial dorsal nucleus, which may mediate multiple symptoms of SCZ and is affected early in the disease course.

Using the Progression Independent of Relapse Activity Framework to Unveil the Pathobiological Foundations of Multiple Sclerosis.

Progression independent of relapse activity (PIRA), a recent concept to formalize disability accrual in multiple sclerosis (MS) independent of relapses, has gained popularity as a potential clinical trial outcome. We discuss its shortcomings and appraise the challenges of implementing it in clinical settings, experimental trials, and research. The current definition of PIRA assumes that acute inflammation, which can manifest as a relapse, and neurodegeneration, manifesting as progressive disability accrual, can be disentangled by introducing specific time windows between the onset of relapses and the observed increase in disability. The term PIRMA (progression independent of relapse and MRI activity) was recently introduced to indicate disability accrual in the absence of both clinical relapses and new brain and spinal cord MRI lesions. Assessing PIRMA in clinical practice is highly challenging because it necessitates frequent clinical assessments and brain and spinal cord MRI scans. PIRA is commonly assessed using Expanded Disability Status Scale, a scale heavily weighted toward motor disability, whereas a more granular assessment of disability deterioration, including cognitive decline, using composite measures or other tools, such as digital tools, would possess greater utility. Similarly, using PIRA as an outcome measure in randomized clinical trials is also challenging and requires methodological considerations. The underpinning pathobiology of disability accumulation, that is not associated with relapses, may encompass chronic active lesions (slowly expanding lesions and paramagnetic rim lesions), cortical lesions, brain and spinal cord atrophy, particularly in the gray matter, diffuse and focal microglial activation, persistent leptomeningeal enhancement, and white matter tract damage. We propose to use PIRA to understand the main determinant of disability accrual in observational, cohort studies, where regular MRI scans are not included, and introduce the term of "advanced-PIRMA" to investigate the contributions to disability accrual of the abovementioned processes, using conventional and advanced imaging. This is supported by the knowledge that MRI reflects the MS pathogenic mechanisms better than purely clinical descriptors. Any residual disability accrual, which remains unexplained after considering all these mechanisms with imaging, will highlight future research priorities to help complete our understanding of MS pathogenesis.

Portability rules detection by Epilepsy Tracking META-Set Analysis

Epilepsy is a severe and common neurological disease that causes sudden and irregular seizures, necessitating patient-specific detection models for effective management. The proposed methodology, Epilepsy Tracking META-Set Analysis, establishes portability rules that identify similar patients, enabling the transfer of these detection models from one patient to another. Main issue is to identify clusters of patients analyzing a set of meta-features of each patient in terms of clinical descriptors, performance metrics of a machine learning model for seizure detection, and data complexity measures. The investigation of complexity measures represents a novelty in such a medical field, allowing to compare patients and to support automated seizure detection methods. The proposed methodology is validated using the well-known Epileptic Seizure EEG Database from the Epilepsy Center of the University Hospital of Freiburg and demonstrates promising results in transferring detection models to new cases.

Outcome Prediction Using Multi-Modal Information: Integrating Large Language Model-Extracted Clinical Information and Image Analysis.

Survival prediction post-cystectomy is essential for the follow-up care of bladder cancer patients. This study aimed to evaluate artificial intelligence (AI)-large language models (LLMs) for extracting clinical information and improving image analysis, with an initial application involving predicting five-year survival rates of patients after radical cystectomy for bladder cancer. Data were retrospectively collected from medical records and CT urograms (CTUs) of bladder cancer patients between 2001 and 2020. Of 781 patients, 163 underwent chemotherapy, had pre- and post-chemotherapy CTUs, underwent radical cystectomy, and had an available post-surgery five-year survival follow-up. Five AI-LLMs (Dolly-v2, Vicuna-13b, Llama-2.0-13b, GPT-3.5, and GPT-4.0) were used to extract clinical descriptors from each patient's medical records. As a reference standard, clinical descriptors were also extracted manually. Radiomics and deep learning descriptors were extracted from CTU images. The developed multi-modal predictive model, CRD, was based on the clinical (C), radiomics (R), and deep learning (D) descriptors. The LLM retrieval accuracy was assessed. The performances of the survival predictive models were evaluated using AUC and Kaplan-Meier analysis. For the 163 patients (mean age 64 ± 9 years; M:F 131:32), the LLMs achieved extraction accuracies of 74%~87% (Dolly), 76%~83% (Vicuna), 82%~93% (Llama), 85%~91% (GPT-3.5), and 94%~97% (GPT-4.0). For a test dataset of 64 patients, the CRD model achieved AUCs of 0.89 ± 0.04 (manually extracted information), 0.87 ± 0.05 (Dolly), 0.83 ± 0.06~0.84 ± 0.05 (Vicuna), 0.81 ± 0.06~0.86 ± 0.05 (Llama), 0.85 ± 0.05~0.88 ± 0.05 (GPT-3.5), and 0.87 ± 0.05~0.88 ± 0.05 (GPT-4.0). This study demonstrates the use of LLM model-extracted clinical information, in conjunction with imaging analysis, to improve the prediction of clinical outcomes, with bladder cancer as an initial example.

Adjuvant mFOLFIRINOX (mFFX) for resected pancreatic cancer (PC): Long term clinical and genomic outcomes.

4170 Background: Adjuvant chemotherapy with mFFX is the standard for pts with ECOG 0-1 (1) after pancreatectomy for resectable PC. Herein, we report long-term survival data on an expanded cohort and present comprehensive genomic analyses. Methods: Pts with resected PC who received &gt;1 dose of adjuvant mFFX identified from institutional databases. Primary endpoints: Recurrence Free Survival (mRFS) and Overall Survival (mOS) calculated from start of treatment until recurrence or death. Key secondary endpoints: clinical factors, genomic descriptors ( KRAS status, whole genome doubling (WGD), and homologous recombination deficiency (HRD)) associated with outcome. Kaplan-Meier method used to estimate RFS and OS. Univariate Cox proportional hazards model correlated clinico-genomic characteristics with RFS and OS. Results: N=147 identified between 01/2015-01/2022. Median age: 67 years; N= 22 (15%) &gt; 75 years. Several prognostically unfavorable groups; N=52 (36%) stage III disease/ N2 nodal status; N=115 (78%) lymphovascular invasion (LVI); N=133 (90%) perineural invasion (PNI). Median time from surgery to start mFFX: 1.78m (IQR; 1.45, 2.12). Median CA 19-9 at start mFFX: 21 (range 0-1,124). N=124 (84%) pts completed 12 doses mFFX (+/- 12 doses oxaliplatin). N=98 pts (67%) stopped oxaliplatin early for neuropathy. With the median follow up 35.1m (range 5.1, 73.0), mRFS 27m (95% CI 20, 39) and OS not reached (NR). For &gt;75 cohort, mRFS 12m (95% CI 9.6, NR) and mOS 30m (95% CI 19, NR). mFFX started &lt;8 weeks from resection associated with improved RFS (HR 0.62; 95% CI 0.41, 0.96; p= 0.033) and OS (HR 0.53; 95% CI 0.3, 0.94; p= 0.030). T3 tumors associated with worse RFS (HR 2.82, 95% CI 1.28, 6.20; p=0.010) and OS (HR 5.10, 95% CI 1.45, 18; p=0.011). Genomic variables summarized (Table). KRASmutation (+), WGD (+), and higher KRAS CNV trended to shorter RFS and OS; statistical significance limited by small subgroups. HRD status did not confer differences in OS. Conclusions: Adjuvant mFFX is an effective therapy and results in long-term OS outcome in resected PC in a non-trial setting, including for pts &gt;75 years. Early initiation of adjuvant mFFX may optimize outcome. Trends observed for improved outcome in select genomic subsets, which require large-scale validation. 1. Conroy, JAMA Onc, 2022. [Table: see text]

Characterization of clinical data for patient stratification in moderate osteoarthritis with support vector machines, regulatory network models, and verification against osteoarthritis Initiative data

Knee osteoarthritis (OA) diagnosis is based on symptoms, assessed through questionnaires such as the WOMAC. However, the inconsistency of pain recording and the discrepancy between joint phenotype and symptoms highlight the need for objective biomarkers in knee OA diagnosis. To this end, we study relationships among clinical and molecular data in a cohort of women (n = 51) with Kellgren–Lawrence grade 2–3 knee OA through a Support Vector Machine (SVM) and a regulation network model. Clinical descriptors (i.e., pain catastrophism, depression, functionality, joint pain, rigidity, sensitization and synovitis) are used to classify patients. A Youden’s test is performed for each classifier to determine optimal binarization thresholds for the descriptors. Thresholds are tested against patient stratification according to baseline WOMAC data from the Osteoarthritis Initiative, and the mean accuracy is 0.97. For our cohort, the data used as SVM inputs are knee OA descriptors, synovial fluid proteomic measurements (n = 25), and transcription factor activation obtained from regulatory network model stimulated with the synovial fluid measurements. The relative weights after classification reflect input importance. The performance of each classifier is evaluated through ROC-AUC analysis. The best classifier with clinical data is pain catastrophism (AUC = 0.9), highly influenced by funcionality and pain sensetization, suggesting that kinesophobia is involved in pain perception. With synovial fluid proteins used as input, leptin strongly influences every classifier, suggesting the importance of low-grade inflammation. When transcription factors are used, the mean AUC is limited to 0.608, which can be related to the pleomorphic behaviour of osteoarthritic chondrocytes. Nevertheless, funcionality has an AUC of 0.7 with a decisive importance of FOXO downregulation. Though larger and longitudinal cohorts are needed, this unique combination of SVM and regulatory network model shall help to stratify knee OA patients more objectively.

Presentation of enthesitis-related arthritis and juvenile-onset spondyloarthritis: a cross-sectional study in a pediatric and adult clinic

BackgroundJuvenile idiopathic arthritis (JIA) comprises a whole spectrum of chronic arthritis starting before 16 years of age. The study aims to explore the clinical and demographic descriptors, treatment, and disease progression of enthesitis-related arthritis (ERA) in comparison with juvenile-onset spondyloarthritis (SpA).MethodsCross-sectional analysis of consecutive patients in two dedicated clinics, with a single visit and retrospective case-notes review. Arthritis, enthesitis and sacroiliitis were evaluated by scoring disease activity and damage. Continuous variables were reported by median, interquartile range; categorical variables were reported by the frequency comparison of the two groups.ResultsThirty-three cases were included, being 23 (69.7%) with ERA. The median age at diagnosis was 12.5 y (SpA) vs. 9 y (ERA) (p < 0.01); the time from symptom onset to diagnosis was 5.5 y (SpA) vs. 1.5 y (ERA) (p < 0.03). In both groups, the predominant presentation was a single joint or < 5 lower limb joints and asymmetric involvement, with a high frequency of enthesitis. There was a higher frequency of mid-tarsal and ankle synovitis in the ERA group and hip involvement in those with SpA. The comparison of the frequency of spine symptoms at presentation, 30% SpA vs. 21.7% ERA (p = 0.7), was not significant, and radiographic progression to spinal involvement occurred in 43.5% of ERA patients. The median time for spinal progression and age at onset was 2.2 and 12 y for ERA, and 4 and 16.5 y for SpA, respectively. Activity and damage scores were not significantly different between the groups. Treatment comparison resulted in 91.3% of ERA and 100% SpA being treated, predominantly with NSAIDs in both groups, followed by DMARDs and biologics, with a higher frequency of biologics in SpA.ConclusionThe main differences were the late diagnoses of SpA, and the hip and spine involvement, with higher frequency of biologic treatment in juvenile-onset SpA compared to ERA.

Factors Associated With Prolonged Mental Health Admissions at US Children's Hospitals.

Mental health (MH) hospitalizations at medical hospitals are associated with longer length of stay (LOS) compared with non-MH hospitalizations, but patient factors and costs associated with prolonged MH hospitalizations are unknown. The objective of this paper is to assess patient clinical and demographic factors associated with prolonged MH hospitalizations and describe variation in MH LOS across US children's hospitals. We studied children aged 5 to 20 years hospitalized with a primary MH diagnosis during 2021 and 2022 across 46 children's hospitals using the Pediatric Health Information System database. Generalized estimating equations, clustered on hospital, tested associations between patient characteristics with prolonged MH hospitalization, defined as those in the 95th percentile or above (>14 days). Among 42 654 primary MH hospitalizations, most were aged 14 to 18 (62.4%), female (68.5%), and non-Hispanic white (53.8%). The most common primary MH diagnoses were suicide/self-injury (37.4%), depressive disorders (16.6%), and eating disorders (10.9%). The median (interquartile range) LOS was 2 days (1-5), but 2169 (5.1%) experienced a hospitalization >14 days. In adjusted analyses, race and ethnicity, category of MH diagnosis, and increasing medical and MH complexity were associated with prolonged hospitalization. Our results emphasize several diagnoses and clinical descriptors for targeted interventions, such as behavioral and inpatient MH resources and discharge planning. Expanded investment in both community and inpatient MH supports have the potential to improve health equity and reduce prolonged MH hospitalizations.

Performance of Radiological and Biochemical Biomarkers in Predicting Radio-Symptomatic Knee Osteoarthritis Progression.

Imaging biomarkers permit improved approaches to identify the most at-risk patients encountering knee osteoarthritis (KOA) progression. This study aimed to investigate the utility of trabecular bone texture (TBT) extracted from plain radiographs, associated with a set of clinical, biochemical, and radiographic data, as a predictor of long-term radiographic KOA progression. We used data from the Foundation for the National Institutes of Health (FNIH) Biomarkers Consortium dataset. The reference model made use of baseline TBT parameters adjusted for clinical covariates and radiological scores. Several models based on a combination of baseline and 24-month TBT variations (TBT∆TBT) were developed using logistic regression and compared to those based on baseline-only TBT parameters. All models were adjusted for baseline clinical covariates, radiological scores, and biochemical descriptors. The best overall performances for the prediction of radio-symptomatic, radiographic, and symptomatic progression were achieved using TBT∆TBT parameters solely, with area under the ROC curve values of 0.658 (95% CI: 0.612-0.705), 0.752 (95% CI: 0.700-0.804), and 0.698 (95% CI: 0.641-0.756), respectively. Adding biochemical markers did not significantly improve the performance of the TBT∆TBT-based model. Additionally, when TBT values were taken from the entire subchondral bone rather than just the medial, lateral, or central compartments, better results were obtained.

MLe-KCNQ2: An Artificial Intelligence Model for the Prognosis of Missense KCNQ2 Gene Variants.

Despite the increasing availability of genomic data and enhanced data analysis procedures, predicting the severity of associated diseases remains elusive in the absence of clinical descriptors. To address this challenge, we have focused on the KV7.2 voltage-gated potassium channel gene (KCNQ2), known for its link to developmental delays and various epilepsies, including self-limited benign familial neonatal epilepsy and epileptic encephalopathy. Genome-wide tools often exhibit a tendency to overestimate deleterious mutations, frequently overlooking tolerated variants, and lack the capacity to discriminate variant severity. This study introduces a novel approach by evaluating multiple machine learning (ML) protocols and descriptors. The combination of genomic information with a novel Variant Frequency Index (VFI) builds a robust foundation for constructing reliable gene-specific ML models. The ensemble model, MLe-KCNQ2, formed through logistic regression, support vector machine, random forest and gradient boosting algorithms, achieves specificity and sensitivity values surpassing 0.95 (AUC-ROC > 0.98). The ensemble MLe-KCNQ2 model also categorizes pathogenic mutations as benign or severe, with an area under the receiver operating characteristic curve (AUC-ROC) above 0.67. This study not only presents a transferable methodology for accurately classifying KCNQ2 missense variants, but also provides valuable insights for clinical counseling and aids in the determination of variant severity. The research context emphasizes the necessity of precise variant classification, especially for genes like KCNQ2, contributing to the broader understanding of gene-specific challenges in the field of genomic research. The MLe-KCNQ2 model stands as a promising tool for enhancing clinical decision making and prognosis in the realm of KCNQ2-related pathologies.

Clinical observations in acute neurology: A historical explanation of the present.

The history of the development of acute neurologic disease as a biologic mechanism is of interest. Equally important is how it translated to the bedside and how the clinical examination differentiated itself. This paper reviews primary sources pertaining to acute neurologic conditions described mostly in the 19th and 20th century. A review of monographs, treatises, textbooks, and peer-reviewed articles was conducted. The evolution of clinical signs and syndromes associated with dynamic intracranial pathologies was predicated on the idea that animal studies informed clinicians, who then linked clinical signs to these observations. A dominant theme is that innovative technologies could trace acute processes through all their various stages, affording a complete picture of the disease process. Just as clinical descriptors of central nervous system processes evolved, the presentation of acute neuromuscular respiratory failure became better defined. Once practices incorporated these acute clinical signs, textbooks cemented their "gold standard" status with relative impunity. The practise of acute neurology and neurocritical care must find out what, historically, others were seeking but could not find. Patterns of clinical presentation in acute neurology are sufficiently recognizable to guide practise decisions. Although, the currently well-documented clinical syndromes of acute neurologic conditions (at presentation and during deterioration) have been taught for generations, practitioners have noted that they lack consistency and predictability. History also taught us that part of this improved knowledge came with designated units-a clear example of how protean systems (not always innovative neurologists or neuroscientists or technologies) shaped the history of neurology.

Nurses' experiences of using a computer-based triage decision support system in the emergency department.

Emergency department triage is a vital and complex decision-making process. There is limited knowledge about nurses' experiences with triage decision support systems in emergency departments. This study was conducted to examine nurses' experiences with a computer-based triage decision support system in the emergency department. This is a qualitative and phenomenological study. Data were collected through interviews from 14 triage nurses who used a triage decision support system in the emergency department of a university hospital. The nurses were recruited for semi-structured interviews. All interviews were recorded on a voice recorder and then transcribed. Data were analysed with the inductive content analysis method. The interviewer asked comprehensive questions about the nurses' experiences with the triage decision support system. Three main and 11 sub-themes were elicited as a result of the analysis of the in-depth interviews: (a) the facilitating the triage decision theme, which included help in case of a dilemma, team collaboration, monitoring/supervision, and error reduction sub-themes; (b) the contribution to professionalism theme, which included ease of learning and teaching triage, professional autonomy, creating a database, and evidence-based practice sub-themes; (c) the areas that need improvement theme, which included reducing screen clicks, the effect of the hospital automation system performance, and clinical descriptors not included in the algorithm sub-themes. Triage nurses stated that the decision support system was beneficial and facilitated decision-making. The decision support system enabled triage nurses to make their own decisions using their clinical knowledge and experience, without a restriction on their professional autonomy, and this was perceived positively. It was emphasized that this system could be a support tool in educating nurses new to triage. However, nurses stated that the pace of the decision support system was affected by the performance of the hospital automation system and that the large number of steps used for electing items caused a waste of time. They also added that these were technical areas that needed improvement. The study provides important data that will help healthcare organizations and professionals better understand the emergency department nurse triage decision support system and gain a versatile, comprehensive, and general understanding.

Exploring Diagnostic Precision and Triage Proficiency: A Comparative Study of GPT-4 and Bard in Addressing Common Ophthalmic Complaints.

In the modern era, patients often resort to the internet for answers to their health-related concerns, and clinics face challenges to providing timely response to patient concerns. This has led to a need to investigate the capabilities of AI chatbots for ophthalmic diagnosis and triage. In this in silico study, 80 simulated patient complaints in ophthalmology with varying urgency levels and clinical descriptors were entered into both ChatGPT and Bard in a systematic 3-step submission process asking chatbots to triage, diagnose, and evaluate urgency. Three ophthalmologists graded chatbot responses. Chatbots were significantly better at ophthalmic triage than diagnosis (90.0% appropriate triage vs. 48.8% correct leading diagnosis; p < 0.001), and GPT-4 performed better than Bard for appropriate triage recommendations (96.3% vs. 83.8%; p = 0.008), grader satisfaction for patient use (81.3% vs. 55.0%; p < 0.001), and lower potential harm rates (6.3% vs. 20.0%; p = 0.010). More descriptors improved the accuracy of diagnosis for both GPT-4 and Bard. These results indicate that chatbots may not need to recognize the correct diagnosis to provide appropriate ophthalmic triage, and there is a potential utility of these tools in aiding patients or triage staff; however, they are not a replacement for professional ophthalmic evaluation or advice.

Abstract C074: Clinico-genomic characterization of N=2,460 pancreatic adenocarcinoma identifies KRASMUT dosage as prognostic of overall survival across disease stages

Abstract Background: The majority (~80%) of patients (pts) with pancreatic adenocarcinomas (PDAC) present with advanced disease. Most sequencing efforts to date have focused on early stage disease. Herein, we leverage a large clinical sequencing cohort of N= 2460 pts with PDAC to characterize the clinico-genomic landscape across different disease stages. Methods: Tumors from 2460 pts were sequenced using MSK-IMPACT clinical sequencing assay and mutations called via CLIA-approved pipelines. Detailed clinical descriptors including lines of treatment and response were abstracted for N=1451 (59%) pts. Intratumor heterogeneity estimates including allele-specific copy number were inferred using FACETS. Results: Of N=2460, at diagnosis 34% of pts had resectable tumors, and 24% and 42% had locally advanced and metastatic disease, respectively. 94% of pts had KRASMUT tumors. Expectedly, MAPK pathway alterations were significantly enriched in KRASWT tumors (60% vs. 7% in KRASMUT, p &amp;lt; 0.001) with gain-of-function kinase oncogene fusions exclusively observed in these tumors (p&amp;lt;0.001). In addition, GNAS, PTEN, FOXP1, SMARCB1 and CREBBP were enriched in the KRASWT tumors (q &amp;lt; 0.05). Collectively, pts with KRASWT/MAPK MUT and KRASWT/MAPK WT tumors comprise 4% and 2% of all pts PDAC, respectively. Germline pathogenic variants in high/moderate penetrance genes were more common in pts with KRASWT tumors compared to pts with KRASMUT tumors (20% vs. 10%, p &amp;lt; 0.001). Pathogenic germline alterations in BRCA1/2 and PALB2 were observed in 6% (N=144) of cohort. Biallelic inactivation of BRCA2 specifically was associated with prolonged response of metastatic pts to PARP-inhibitor maintenance therapy with 10.3 months on treatment prior to progression compared to 3.4 months for those with biallelic inactivation of other HRD genes and 2.2 months for those with non-biallelic inactivation. Among KRASMUT tumors assessed for KRAS copy-number allelic imbalance (AI) (N=1097), 41% (N=452) had AI at the KRAS locus with mutant-allele selected for in 94% of tumors with imbalance. Among KRASMUT tumors with normal genome ploidy, overall, 20% had a mutant allele gain ranging from 10% in resectable to 29% of metastatic tumors. We observe significantly lower OS among pts with KRASMUT dosage gain compared to heterozygous KRASMUT across pts with resectable (HR: 0.66, 95% CI: 0.18-1.15, p&amp;lt;0.01) and metastatic (HR: 0.39, 95% CI: 0.16-0.63, p=0.001) diseases with statistical significance not reached among the locally advanced disease (HR: 0.4, 95% CI: -0.01-0.81, p=0.056). Notably, the OS of pts with loss of the wild-type allele and no mutant copy gain was indistinguishable from that of heterozygous KRASMUT tumors (HR: -0.11, 95% CI: -0.37-0.15, p=0.4). Conclusion: In this large cohort of 2460 pts with PDAC, we observed that KRASMUT dosage gain is a significant negative predictor of OS. These findings highlight the importance of incorporating allelic-dosage into comprehensive molecular testing in PDAC given prognostic importance and the emerging era of direct RAS targeting. Citation Format: Anna M. Varghese, Maria A. Perry, Daniel Muldoon, Subhiksha Nandakumar, Amanda Erakky, Bastien Nguyen, Amanda Zucker, Hulya Sahin Ozkan, Olca Basturk, Debyani Chakravarty, David P. Kelsen, Wungki Park, Kenneth H. Yu, Alice C. Wei, Nikolaus Schultz, Michael F. Berger, Christine Iacobuzio-Donahue, Chaitanya Bandlamudi, Eileen M. O'Reilly. Clinico-genomic characterization of N=2,460 pancreatic adenocarcinoma identifies KRASMUT dosage as prognostic of overall survival across disease stages [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Pancreatic Cancer; 2023 Sep 27-30; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(2 Suppl):Abstract nr C074.

Artificial intelligence–adjudicated spatiotemporal dispersion: A patient-unique fingerprint of persistent atrial fibrillation

BackgroundSpatiotemporal dispersion-guided ablation is a tailored approach for patients in persistent atrial fibrillation (PsAF). The characterization of dispersion extent and distribution and its association with common clinical descriptors of PsAF patients has not been studied. ObjectivesArtificial intelligence–adjudicated dispersion extent and distribution (AI-DED) was obtained with a machine/deep learning classifier (VX1 Software, Volta Medical) in PsAF patients undergoing ablation. The purpose of this study was to test the hypothesis that AI-DED is unique to each patient and independent of common procedural and clinical parameters. MethodsIn a subanalysis of the Ev-AIFib study (NCT03434964), spatiotemporal dispersion maps were built with VX1 software in 78 consecutive persistent and long-standing PsAF patients. AI-DED was quantified using 2 distinct approaches (visual regional characterization or automated global quantification of AI-DED). ResultsAI-DED paired-subregion Euclidean distance measurements between 78 patients (average distance 5.07 ± 0.60; min 2.23; max 9.75) demonstrate that AI-DED is a patient-unique characteristic of PsAF. Importantly, both AF type and AF history do not correlate with AI-DED levels (R2 = 0.006, P = .53; and R2 = 0.03, P = .25, respectively). The most extensive AI-DED levels are not associated with poorer procedural (83%, 81%, and 83% of AF termination in low, medium, and high dispersion groups, respectively; P = .954) and long-term (88%, 75%, and 91% of freedom from AF/atrial tachycardia after multiple procedures; P = .517) outcomes. ConclusionThe atrial distribution and extent of multipolar electrogram spatiotemporal dispersion follow a nonrandom, albeit patient-unique, distribution in PsAF patients. AI-DED may represent a procedure-implementable fingerprint of the PsAF substrate.

CMAT: ClinVar Mapping and Annotation Toolkit.

Semantic ontology mapping of clinical descriptors with disease outcome is essential. ClinVar is a key resource for human variation with known clinical significance. We present CMAT, a software toolkit and curation protocol for accurately enriching ClinVar releases with disease ontology associations and complex functional consequences. The software and ontology mappings can be obtained from: https://github.com/EBIvariation/CMAT.

CHARACTERIZATION OF CLINICAL DATA FOR IN MODERATE OSTEOARTHRITIS WITH SUPPORT VECTOR MACHINES AND REGULATORY NETWORK MODELS

Knee osteoarthritis (KOA) diagnosis is based on symptoms, assessed through questionnaires such as the WOMAC. However, the inconsistency of pain recording and the discrepancy between joint phenotype and symptoms highlight the need for objective biomarkers in KOA diagnosis. To this end, we study relationships among clinical and molecular data in a cohort of women (n=51) with Kellgren-Lawrence grade 2–3 KOA through Support Vector Machine (SVM) and a regulation network model (RNM). Clinical descriptors (i.e., pain catastrophism (CA); depression (DE); functionality (FU); joint pain (JP); rigidity (RI); sensitization (SE); synovitis (SY)) are used to classify patients. A Youden's test is performed for each classifier to determine optimal binarization thresholds for the descriptors. Thresholds are tested against patient stratification according to baseline WOMAC data from the Osteoarthritis Initiative, and the mean accuracy is 0.97. For our cohort, the data used as SVM inputs are KOA descriptors, synovial fluid (SL) proteomic measurements (n=25), and transcription factors (TF) activation obtained from RNM [2] stimulated with the SL measurements. The relative weights after classification reflect input importance. The performance of each classifier is evaluated through AUC-ROC analysis. The best classifier with clinical data is CA (AUC = 0.9), highly influenced by FU and SE, suggesting that kinesophobia is involved in pain perception. With SL input, leptin strongly influences every classifier, suggesting the importance of low-grade inflammation. When TF are used, the mean AUC is limited to 0.608, which can be related to the pleomorphic behaviour of osteoarthritic chondrocytes. Nevertheless, FU has an AUC of 0.7 with strong importance of FOXO downregulation. Though larger and longitudinal cohorts are needed, this unique combination of SVM and RNM shall help to map objectively KOA descriptors.Acknowledgements: Catalan &amp; Spanish governments 2020FI_b00680; STRATO-PID2021126469ob-C21-2, European Commission (MSCA-TN-ETN-2020-Disc4All-955735, ERC-2021-CoG-O-Health-101044828). ICREA Academia.

Clinical descriptors of disease trajectories in patients with traumatic brain injury in the intensive care unit (CENTER-TBI): a multicentre observational cohort study

Patients with traumatic brain injury are a heterogeneous population, and the most severely injured individuals are often treated in an intensive care unit (ICU). The primary injury at impact, and the harmful secondary events that can occur during the first week of the ICU stay, will affect outcome in this vulnerable group of patients. We aimed to identify clinical variables that might distinguish disease trajectories among patients with traumatic brain injury admitted to the ICU. We used data from the Collaborative European NeuroTrauma Effectiveness Research in Traumatic Brain Injury (CENTER-TBI) prospective observational cohort study. We included patients aged 18 years or older with traumatic brain injury who were admitted to the ICU at one of the 65 CENTER-TBI participating centres, which range from large academic hospitals to small rural hospitals. For every patient, we obtained pre-injury data and injury features, clinical characteristics on admission, demographics, physiological parameters, laboratory features, brain biomarkers (ubiquitin carboxy-terminal hydrolase L1 [UCH-L1], S100 calcium-binding protein B [S100B], tau, neurofilament light [NFL], glial fibrillary acidic protein [GFAP], and neuron-specific enolase [NSE]), and information about intracranial pressure lowering treatments during the first 7 days of ICU stay. To identify clinical variables that might distinguish disease trajectories, we applied a novel clustering method to these data, which was based on a mixture of probabilistic graph models with a Markov chain extension. The relation of clusters to the extended Glasgow Outcome Scale (GOS-E) was investigated. Between Dec 19, 2014, and Dec 17, 2017, 4509 patients with traumatic brain injury were recruited into the CENTER-TBI core dataset, of whom 1728 were eligible for this analysis. Glucose variation (defined as the difference between daily maximum and minimum glucose concentrations) and brain biomarkers (S100B, NSE, NFL, tau, UCH-L1, and GFAP) were consistently found to be the main clinical descriptors of disease trajectories (ie, the leading variables contributing to the distinguishing clusters) in patients with traumatic brain injury in the ICU. The disease trajectory cluster to which a patient was assigned in a model was analysed as a predictor together with variables from the IMPACT model, and prediction of both mortality and unfavourable outcome (dichotomised GOS-E ≤4) was improved. First-day ICU admission data are not the only clinical descriptors of disease trajectories in patients with traumatic brain injury. By analysing temporal variables in our study, variation of glucose was identified as the most important clinical descriptor that might distinguish disease trajectories in the ICU, which should direct further research. Biomarkers of brain injury (S100B, NSE, NFL, tau, UCH-L1, and GFAP) were also top clinical descriptors over time, suggesting they might be important in future clinical practice. European Union 7th Framework program, Hannelore Kohl Stiftung, OneMind, Integra LifeSciences Corporation, and NeuroTrauma Sciences.

How to recognise a Behçet's ulcer from other types of oral ulceration? Defining Behçet's ulceration by an International Delphi Consultation.

To define the clinical characteristics of oral ulceration (OU) in Behçet's disease (BD), to allow differentiation from other causes of OU, including aphthous ulcers, by an International Delphi consultation. To develop a clinical guideline on how to recognise BD ulcers. Round 1. 40 clinical images of OU in BD, recurrent aphthous stomatitis (RAS), inflammatory bowel disease (IBD) and mucous membrane pemphigoid (MMP) were shown. Participants answered, independently, which images would be consistent with a BD ulcer. Round 2. The results from marking independently were shown. The panel remarked the questions through iteration process. The images not agreed to be a possible BD ulcer were discarded. Round 3. 10 clinical descriptors that may define BD ulcers were suggested. Participants ranked the level of importance for each descriptor on each image presented. Round 4. Participants re-ranked their level of agreement for each descriptor through iteration process. Whether the clinical pictures would be different from RAS was also explored. A final agreement was reached. This study has shown clear differentiation between BD, IBD and MMP ulcers when defining them by phenotype through clinical images only. On the other hand, no differentiation between RAS and BD ulcers was found. The most important clinical descriptors that define BD ulcers have been agreed. New clinical guidance for Health Care Professionals (HCP) on how to recognise a BD ulcer has been proposed. This should elucidate an earlier diagnosis, quicker access to treatment and control of the disease enhancing patient's quality of life.