Background and Objectives: Venous thromboembolism (VTE) can be the first manifestation of an underlying cancer. This study aimed to develop a predictive model to assess the risk of occult cancer between 30 days and 24 months after a venous thrombotic event using machine learning (ML). Materials and Methods: We designed a case–control study nested in a cohort of patients with VTE included in a prospective registry from two Spanish hospitals between 2005 and 2021. Both clinically and ML-driven feature selection were performed to identify predictors for occult cancer. XGBoost, LightGBM, and CatBoost algorithms were used to train different prediction models, which were subsequently validated in a hold-out dataset. Results: A total of 815 patients with VTE were included (51.5% male and median age of 59). During follow-up, 56 patients (6.9%) were diagnosed with cancer. One hundred and twenty-one variables were explored for the predictive analysis. CatBoost obtained better performance metrics among the ML models analyzed. The final CatBoost model included, among the top 15 variables to predict hidden malignancy, age, gender, systolic blood pressure, heart rate, weight, chronic lung disease, D-dimer, alanine aminotransferase, hemoglobin, serum creatinine, cholesterol, platelets, triglycerides, leukocyte count and previous VTE. The model had an ROC-AUC of 0.86 (95% CI, 0.83–0.87) in the test set. Sensitivity, specificity, and negative and positive predictive values were 62%, 94%, 93% and 75%, respectively. Conclusions: This is the first risk score developed for identifying patients with VTE who are at increased risk of occult cancer using ML tools, obtaining a remarkably high diagnostic accuracy. This study’s limitations include potential information bias from electronic health records and a small cancer sample size. In addition, variability in detection protocols and evolving clinical practices may affect model accuracy. Our score needs external validation.

Background: Inherent to the challenge of acute pulmonary embolism (APE), the breadth of presentation ranges from asymptomatic pulmonary emboli to sudden death. Risk stratification of patients with APE is mandatory for determining the appropriate therapeutic management approach. However, the optimal clinically most relevant combination of predictors of death remains to be determined. Radiomics is an emerging discipline in medicine that extracts and analyzes quantitative data from medical images using mathematical algorithms. In APE, these data can reveal thrombus characteristics that are not visible to the naked eye, which may help to more accurately identify patients at higher risk of early clinical deterioration or mortality. We conducted a scoping review to explore the current evidence on the prognostic performance of radiomic models in patients with APE. Methods: PubMed, Web of Science, EMBASE, and Scopus were searched for studies published between January 2010 and April 2025. Eligible studies evaluated the use of radiomics to predict adverse outcomes in patients with APE. The PROSPERO registration number is CRD420251083318. Results: Nine studies were included in this review. There was significant heterogeneity in the methodology for feature selection and model development. Radiomic models demonstrated variable performance across studies. Models that combined radiomic features with clinical data tended to show better predictive accuracy. Conclusions: This scoping review underscores the potential of radiomic models, particularly when combined with clinical data, to improve risk stratification in patients with APE.

Background/Objectives: Unusual-site venous thrombosis (USVT) lacks robust evidence guiding anticoagulant selection between vitamin K antagonists (VKAs) and direct oral anticoagulants (DOACs). This study aimed to evaluate recanalization, recurrence, and major bleeding outcomes in real-world USVT patients and to replicate these findings through a validated digital twin model with Monte Carlo simulation. Methods: We conducted a retrospective study of 90 USVT patients (72% VKAs, 28% DOACs). A conditional generative adversarial network was used to generate digital twins matched on age, sex, thrombosis site, and malignancy. Logistic regression was applied to estimate treatment-specific outcome probabilities for recanalization, recurrence, and major bleeding. A nested stochastic simulation framework simulated 500 iterations across clinical scenarios, including increased DOAC use, cancer prevalence, cerebral vein thrombosis proportion, and optimized VKA control. Results: The mean age was 67.5 years, and 54.4% were female. 61.1% of splanchnic vein thrombosis, 36.7% of upper extremity deep vein thrombosis, and 2.2% of cerebral vein thrombosis were included. In the real cohort, complete recanalization occurred in 40.0% of patients with DOACs and 36.0% with VKAs. Recurrence was 8.0% with DOACs and 7.7% with VKAs, and major bleeding occurred in 8.0% and 10.8% of cases, respectively. All-cause mortality was 20% in DOAC-treated patients and 60% in those receiving VKAs. Digital Twin-based predictions replicated these results (recanalization 40.3% versus 38.0%; recurrence 10.9% versus 8.6%; bleeding 7.6% versus 9.1%). Simulated scenarios preserved the directionality effect, with the most significant differences observed in high-cerebral vein thrombosis and cancer-enriched patients. Conclusions: DOACs showed comparable efficacy and slightly lower bleeding risk than VKAs in USVT. Digital twin and Monte Carlo modeling provided robust, reproducible simulations of treatment effects under varying clinical conditions. Separating empirical and simulation-based findings, the digital twin supported the internal consistency of real-world observations and demonstrated the potential of in silico modeling as a complementary tool in rare thrombotic diseases.

Background/Objectives: The role of direct oral anticoagulants (DOACs) vs. vitamin K antagonists (VKAs) in preventing recurrent thrombosis in patients with antiphospholipid syndrome (APS) remains uncertain. Using real-world data, we aimed to evaluate the effectiveness and internal validity of a digital twin (DT) approach for modeling thrombotic recurrence risk in APS patients treated with DOACs or VKAs. Methods: We conducted a multicenter observational study that included thrombotic APS patients treated with DOACs or VKAs. Clinical data were used to generate DT via conditional generative adversarial networks (CGANs), incorporating a directed acyclic graph (DAG) to preserve causal relationships. Validation metrics included absolute standardized mean differences (ASMD), mean ASMD (MASMD), and Spearman correlation matrices to assess structural fidelity. Treatment effects were estimated in a CGAN-conditioned cohort matched on key covariates. Results: Eighty-nine thrombotic APS patients were included: 70 (78.7%) received VKAs and 19 (21.3%) received DOACs. Thrombotic recurrences occurred in 5 DOAC patients (26.3%) and 17 AVK patients (24.3%). The CGAN-generated synthetic cohort closely mirrored the original data (MASMD = 0.073 ± 0.041), with 85.4% of pairwise correlations differing by <0.1 in absolute value. In the conditioned DT cohort, predicted recurrence was 24.2% for DOACs and 19.9% for VKAs. Recurrence risk increased with antibody burden, reaching 41.3% in triple-positive patients and 46.8% in those with index arterial thrombosis treated with DOACs. Conclusions: DT technology accurately replicated the clinical structure of APS patients, supporting its application for simulating counterfactual scenarios and estimating individualized treatment effects.

Gut microbiota dysbiosis has been proposed as a potential contributor to venous thromboembolism (VTE), although its clinical relevance remains uncertain. This review was conducted following the PRISMA 2020 guidelines. We conducted a systematic review of studies evaluating the association between gut microbiota dysbiosis and VTE. MEDLINE/PubMed, Web of Science, EMBASE, and Scopus were searched from the inception to February 2026. Observational studies and Mendelian randomization (MR) analyses assessing microbiota composition, dysbiosis, or related metabolites in relation to VTE were included. Risk of bias was assessed using established tools. Thirteen studies were included (five observational and eight MR). Observational evidence showed heterogeneous findings. Microbial metabolites such as trimethylamine N-oxide and lipopolysaccharides were associated with prothrombotic profiles in some studies, but no consistent association with VTE risk or recurrence was observed. The differences in microbiota composition were reported, although based on small populations. MR analyses identified microbial taxa with potential protective or risk associations, but most findings were modest and not robust after multiple testing correction. Gut microbiota dysbiosis may contribute to VTE through inflammatory and metabolic pathways; however, current clinical evidence is limited and inconsistent. Further prospective studies are needed to clarify causality and clinical implications.