How to customize common data models for rare diseases: an OMOP-based implementation and lessons learned.

BACKGROUND: Given the geographical sparsity of Rare Diseases (RDs), assembling a cohort is often a challenging task. Common data models (CDM) can harmonize disparate sources of data that can be the basis of decision support systems and artificial intelligence-based studies, leading to new insights in the field. This work is sought to support the design of large-scale multi-center studies for rare diseases. METHODS: In an interdisciplinary group, we derived a list of elements of RDs in three medical domains (endocrinology, gastroenterology, and pneumonology) according to specialist knowledge and clinical guidelines in an iterative process. We then defined a RDs data structure that matched all our data elements and built Extract, Transform, Load (ETL) processes to transfer the structure to a joint CDM. To ensure interoperability of our developed CDM and its subsequent usage for further RDs domains, we ultimately mapped it to Observational Medical Outcomes Partnership (OMOP) CDM. We then included a fourth domain, hematology, as a proof-of-concept and mapped an acute myeloid leukemia (AML) dataset to the developed CDM. RESULTS: We have developed an OMOP-based rare diseases common data model (RD-CDM) using data elements from the three domains (endocrinology, gastroenterology, and pneumonology) and tested the CDM using data from the hematology domain. The total study cohort included 61,697 patients. After aligning our modules with those of Medical Informatics Initiative (MII) Core Dataset (CDS) modules, we leveraged its ETL process. This facilitated the seamless transfer of demographic information, diagnoses, procedures, laboratory results, and medication modules from our RD-CDM to the OMOP. For the phenotypes and genotypes, we developed a second ETL process. We finally derived lessons learned for customizing our RD-CDM for different RDs. DISCUSSION: This work can serve as a blueprint for other domains as its modularized structure could be extended towards novel data types. An interdisciplinary group of stakeholders that are actively supporting the project's progress is necessary to reach a comprehensive CDM. CONCLUSION: The customized data structure related to our RD-CDM can be used to perform multi-center studies to test data-driven hypotheses on a larger scale and take advantage of the analytical tools offered by the OHDSI community.

Applications of artificial intelligence to myeloproliferative neoplasms: a narrative review.

INTRODUCTION: Artificial intelligence (AI) is a rapidly growing field of computational research with the potential to extract nuanced biomarkers for the prediction of outcomes of interest. AI implementations for the prediction for clinical outcomes for myeloproliferative neoplasms (MPNs) are currently under investigation. AREAS COVERED: In this narrative review, we discuss AI investigations for the improvement of MPN clinical care utilizing either clinically available data or experimental laboratory findings. Abstracts and manuscripts were identified upon querying PubMed and the American Society of Hematology conference between 2000 and 2023. Overall, multidisciplinary researchers have developed AI methods in MPNs attempting to improve diagnostic accuracy, risk prediction, therapy selection, or pre-clinical investigations to identify candidate molecules as novel therapeutic agents. EXPERT OPINION: It is our expert opinion that AI methods in MPN care and hematology will continue to grow with increasing clinical utility. We believe that AI models will assist healthcare workers as clinical decision support tools if appropriately developed with AI-specific regulatory guidelines. Though the reported findings in this review are early investigations for AI in MPNs, the collective work developed by the research community provides a promising framework for improving decision-making in the future of MPN clinical care.

Oral Decitabine/Cedazuridine Is an Effective Ambulatory Therapy for Patients With Myelofibrosis Refractory to JAK2 Inhibitor Therapy.

BACKGROUND: Outcomes are dismal for patients with myelofibrosis (MF) who are no longer responsive to JAK2 inhibitors (JAKi) and/or have increasing blast cell numbers. Although prior reports have suggested the benefits of intravenous decitabine (DAC) combined with ruxolitinib for patients with Myeloproliferative Neoplasm (MPN) accelerated/blast phase (AP/BP), decitabine-cedazuridine (DEC-C), an oral fixed-dose combination providing equivalent pharmacokinetic exposure, has not been evaluated in MF. METHODS: We conducted a retrospective analysis of 14 patients with high-risk MF refractory to ruxolitinib or MPN-AP (10-19% blasts) treated with DEC-C +/- JAKi at Mount Sinai Hospital from 2021 to 2024. RESULTS: The cohort was elderly (median age,76 years) and almost uniformly possessed high risk mutations with 13 of the 14 patients progressing on JAKi therapy. With a median follow-up of 9.4 months, the median overall survival (OS) was 29 months for the entire cohort. Median OS was 10.8 months for MPN-AP and was not reached for ruxolitinib refractory MF patients. All patients (n = 9) receiving > 4 cycles of DEC-C had clinical benefit exemplified by a reduction in blast cell numbers, spleen size, and lack of progression to MPN-BP (78%). Furthermore, 3/14 patients proceeded to allogeneic stem cell transplant. Myelosuppression was a common adverse event which was managed by reducing the number of days of administration of DEC-C from 5 to 3 per cycle. CONCLUSIONS: This report demonstrates the feasibility, tolerability, and clinical benefit of an exclusively ambulatory regimen for high-risk, elderly patients with advanced MF which warrants further evaluation in a prospective clinical trial.

Equivalent thrombotic risk with Warfarin, Dabigatran, or Enoxaparin after failure of initial direct oral anticoagulation (DOAC) therapy.

BACKGROUND: The direct oral anticoagulants (DOACs) are now commonly regarded as first line anticoagulants in most cases of venous thromboembolism (VTE). However, the optimal choice of subsequent anticoagulant in instances of first line DOAC failure is unclear. OBJECTIVES: To describe and compare outcomes with second line anticoagulants used after DOAC failure. METHODS: Patients seen at an urban hospital system for an episode of acute VTE initially treated with either apixaban or rivaroxaban who experienced a subsequent recurrent thrombosis while on anticoagulation (1st recurrent thrombosis) were included. RESULTS: In total, 166 patients after apixaban or rivaroxaban failure were included. Following DOAC failure (1st recurrent thrombosis), the subsequent anticoagulant was warfarin in 60 patients (36%), dabigatran in 42 patients (25%), and enoxaparin in 64 patients (39%). Enoxaparin was preferentially prescribed in patients with a malignancy-associated etiology for 1st recurrent thrombosis (p < 0.01). The median follow-up time in our cohort was 16 months. There was no difference in 2nd recurrent thrombosis-free survival (p = 0.72) or risk for major bleeding event (p = 0.30) among patients treated with dabigatran, warfarin, or enoxaparin. CONCLUSIONS: In this retrospective analysis of patients failing first line DOAC therapy, rates of 2nd recurrent thrombosis and bleeding did not differ among subsequently chosen anticoagulants. Our study provides evidence that the optimal 2nd anticoagulant is not clear, and the choice of 2nd anticoagulant should continue to balance patient preference, cost, and provider experience.

Slideflow: deep learning for digital histopathology with real-time whole-slide visualization.

Deep learning methods have emerged as powerful tools for analyzing histopathological images, but current methods are often specialized for specific domains and software environments, and few open-source options exist for deploying models in an interactive interface. Experimenting with different deep learning approaches typically requires switching software libraries and reprocessing data, reducing the feasibility and practicality of experimenting with new architectures. We developed a flexible deep learning library for histopathology called Slideflow, a package which supports a broad array of deep learning methods for digital pathology and includes a fast whole-slide interface for deploying trained models. Slideflow includes unique tools for whole-slide image data processing, efficient stain normalization and augmentation, weakly-supervised whole-slide classification, uncertainty quantification, feature generation, feature space analysis, and explainability. Whole-slide image processing is highly optimized, enabling whole-slide tile extraction at 40x magnification in 2.5 s per slide. The framework-agnostic data processing pipeline enables rapid experimentation with new methods built with either Tensorflow or PyTorch, and the graphical user interface supports real-time visualization of slides, predictions, heatmaps, and feature space characteristics on a variety of hardware devices, including ARM-based devices such as the Raspberry Pi.

Deep learning generates synthetic cancer histology for explainability and education.

Artificial intelligence methods including deep neural networks (DNN) can provide rapid molecular classification of tumors from routine histology with accuracy that matches or exceeds human pathologists. Discerning how neural networks make their predictions remains a significant challenge, but explainability tools help provide insights into what models have learned when corresponding histologic features are poorly defined. Here, we present a method for improving explainability of DNN models using synthetic histology generated by a conditional generative adversarial network (cGAN). We show that cGANs generate high-quality synthetic histology images that can be leveraged for explaining DNN models trained to classify molecularly-subtyped tumors, exposing histologic features associated with molecular state. Fine-tuning synthetic histology through class and layer blending illustrates nuanced morphologic differences between tumor subtypes. Finally, we demonstrate the use of synthetic histology for augmenting pathologist-in-training education, showing that these intuitive visualizations can reinforce and improve understanding of histologic manifestations of tumor biology.

Integration of clinical features and deep learning on pathology for the prediction of breast cancer recurrence assays and risk of recurrence.

Gene expression-based recurrence assays are strongly recommended to guide the use of chemotherapy in hormone receptor-positive, HER2-negative breast cancer, but such testing is expensive, can contribute to delays in care, and may not be available in low-resource settings. Here, we describe the training and independent validation of a deep learning model that predicts recurrence assay result and risk of recurrence using both digital histology and clinical risk factors. We demonstrate that this approach outperforms an established clinical nomogram (area under the receiver operating characteristic curve of 0.83 versus 0.76 in an external validation cohort, p = 0.0005) and can identify a subset of patients with excellent prognoses who may not need further genomic testing.

Predictors and clinical complications associated with antiphospholipid antibodies in sickle cell disease.

Although a higher prevalence of antiphospholipid autoantibodies (aPL) has been observed in some cohorts of sickle cell disease (SCD) patients, the clinical risk factors for the development of aPL and its associated complications remain unclear. In a retrospective study of 63 SCD patients, a lower hemoglobin concentration and higher white blood cell count were independently associated with an elevated aPL. SCD patients with elevated aPL had increased pregnancy complications (≥3 miscarriages, preterm delivery, pre-eclampsia) and venous thrombotic events. Our findings suggest that SCD may predispose to the generation of aPL and that aPL itself may contribute to the vasculopathy of SCD. Prospective testing for aPL is warranted in patients with SCD.

Deep learning applications in visual data for benign and malignant hematologic conditions: a systematic review and visual glossary.

Deep learning (DL) is a subdomain of artificial intelligence algorithms capable of automatically evaluating subtle graphical features to make highly accurate predictions, which was recently popularized in multiple imaging-related tasks. Because of its capabilities to analyze medical imaging such as radiology scans and digitized pathology specimens, DL has significant clinical potential as a diagnostic or prognostic tool. Coupled with rapidly increasing quantities of digital medical data, numerous novel research questions and clinical applications of DL within medicine have already been explored. Similarly, DL research and applications within hematology are rapidly emerging, although these are still largely in their infancy. Given the exponential rise of DL research for hematologic conditions, it is essential for the practising hematologist to be familiar with the broad concepts and pitfalls related to these new computational techniques. This narrative review provides a visual glossary for key deep learning principles, as well as a systematic review of published investigations within malignant and non-malignant hematologic conditions, organized by the different phases of clinical care. In order to assist the unfamiliar reader, this review highlights key portions of current literature and summarizes important considerations for the critical understanding of deep learning development and implementations in clinical practice.

Uncertainty-informed deep learning models enable high-confidence predictions for digital histopathology.

A model's ability to express its own predictive uncertainty is an essential attribute for maintaining clinical user confidence as computational biomarkers are deployed into real-world medical settings. In the domain of cancer digital histopathology, we describe a clinically-oriented approach to uncertainty quantification for whole-slide images, estimating uncertainty using dropout and calculating thresholds on training data to establish cutoffs for low- and high-confidence predictions. We train models to identify lung adenocarcinoma vs. squamous cell carcinoma and show that high-confidence predictions outperform predictions without uncertainty, in both cross-validation and testing on two large external datasets spanning multiple institutions. Our testing strategy closely approximates real-world application, with predictions generated on unsupervised, unannotated slides using predetermined thresholds. Furthermore, we show that uncertainty thresholding remains reliable in the setting of domain shift, with accurate high-confidence predictions of adenocarcinoma vs. squamous cell carcinoma for out-of-distribution, non-lung cancer cohorts.

Evaluation of point-of-care International Normalized Ratio in sickle cell disease.

BACKGROUND: Point-of-care (POC) International Normalized Ratio (INR) measurement provides efficient monitoring of warfarin therapy; however, its reliability may be affected in patients with anemia, such as those with sickle cell disease (SCD). OBJECTIVES: To evaluate the correlation of POC-INR to clinical laboratory INR (CL-INR) in SCD and use of a correction factor. PATIENT/METHODS: In this retrospective study, the accuracy of POC-INR compared to CL-INR was evaluated in a cohort of patients with SCD and in a non-SCD Black cohort. RESULTS: Despite the difference in anemia, the SCD cohort showed a similar percentage of in-range POC-INR values as observed in the non-SCD cohort (37% vs 42%). The SCD cohort was randomly divided to form discovery and validation cohorts. In the discovery cohort, 86% of POC-INRs were in range when the POC-INRs were ˂4.0, but only 24% were in range if POC-INRs were ≥4.0. A linear regression of CL-INR versus POC-INR for POC-INR values ≥4.0 yielded a coefficient of 0.72 (95% confidence interval, 0.69-0.75); Multiplying POC-INR by this correction factor, rounded to 0.7 for ease of use in clinical practice, improved the proportion of in-range POC-INR values ≥4.0 from 24% to 100% in the SCD discovery cohort and from 19% to 95% in the SCD validation cohort. Similar findings applied to analyses of the non-SCD cohort. CONCLUSIONS: POC-INR and CL-INR in patients with SCD are similar when POC-INR is <4.0, and the accuracy of POC-INR values ≥4.0 can be improved by applying an institution-specific correction factor.

Antimicrobial resistance is a risk factor for mortality in adults with sickle cell disease.

Not available.

Clinical, laboratory, and genetic risk factors for thrombosis in sickle cell disease.

Sickle cell disease (SCD) patients are at a four- to 100-fold increased risk for thrombosis compared with the general population, although the mechanisms and risk factors are not clear. We investigated the incidence and predictors for thrombosis in a retrospective, longitudinal cohort of 1193 pediatric and adult SCD patients treated at our institution between January 2008 and December 2017. SCD diagnosis and thrombotic complications were identified using International Classification of Diseases coding and verified through medical chart review. Clinical and laboratory data were extracted from the medical records. With a median follow-up of 6.4 years, 208 (17.4%) SCD patients experienced 352 thrombotic events (64 strokes, 288 venous thromboembolisms [VTE]). Risk factors for stroke included older age and HbSS/Sß0-genotype and a lower hemoglobin (Hb) F% in the subset of HbSS/Sß0-genotype patients (P < .05). VTE risk was independently associated with lower estimated glomerular filtration rate, hydroxyurea (HU) use, HbSS/Sß0 genotype, and higher white blood cell (WBC) counts and Hb (P ≤ .03). Two thrombomodulin gene variants previously associated with thrombosis in the general African American population, THBD rs2567617 (minor allele frequency [MAF] 0.25; odds ratio [OR], 1.5; P = .049) and THBD rs1998081 (MAF, 0.24; OR, 1.5; P = .059), were associated with thrombosis in this cohort. In summary, thrombotic complications are common, and several traditional and SCD-specific risk factors are associated with thrombotic risk. Future studies integrating clinical, laboratory, and genetic risk factors may improve our understanding of thrombosis and guide intervention practices in SCD.

Pan-cancer image-based detection of clinically actionable genetic alterations.

Molecular alterations in cancer can cause phenotypic changes in tumor cells and their micro-environment. Routine histopathology tissue slides - which are ubiquitously available - can reflect such morphological changes. Here, we show that deep learning can consistently infer a wide range of genetic mutations, molecular tumor subtypes, gene expression signatures and standard pathology biomarkers directly from routine histology. We developed, optimized, validated and publicly released a one-stop-shop workflow and applied it to tissue slides of more than 5000 patients across multiple solid tumors. Our findings show that a single deep learning algorithm can be trained to predict a wide range of molecular alterations from routine, paraffin-embedded histology slides stained with hematoxylin and eosin. These predictions generalize to other populations and are spatially resolved. Our method can be implemented on mobile hardware, potentially enabling point-of-care diagnostics for personalized cancer treatment. More generally, this approach could elucidate and quantify genotype-phenotype links in cancer.