Machine learning algorithm as a prognostic tool for venous thromboembolism in allogeneic transplant patients.

As a serious complication after allogenic hematopoietic stem cell transplantation (allo-HSCT), venous thromboembolism (VTE) is significantly related to increased nonrelapse mortality. Therefore distinguishing patients at high risk of death who should receive specific therapeutic management is key to improving survival. This study aimed to establish a machine learning-based prognostic model for the identification of post-transplantation VTE patients who have a high risk of death. We retrospectively evaluated 256 consecutive VTE patients who underwent allo-HSCT at our center between 2008 and 2019. These patients were further randomly divided into (1) a derivation (80%) cohort of 205 patients and (2) a test (20%) cohort of 51 patients. The least absolute shrinkage and selection operator (LASSO) approach was used to choose the potential predictors from the primary dataset. Eight machine learning classifiers were used to produce 8 candidate models. A 10-fold cross-validation procedure was used to internally evaluate the models and to select the best-performing model for external assessment using the test cohort. In total, 256 of 7238 patients were diagnosed with VTE after transplantation. Among them, 118 patients (46.1%) had catheter-related venous thrombosis, 107 (41.8%) had isolated deep-vein thrombosis (DVT), 20 (7.8%) had isolated pulmonary embolism (PE), and 11 (4.3%) had concomitant DVT and PE. The 2-year overall survival (OS) rate of patients with VTE was 68.8%. Using LASSO regression, 8 potential features were selected from the 54 candidate variables. The best-performing algorithm based on the 10-fold cross-validation runs was a logistic regression classifier. Therefore a prognostic model named BRIDGE was then established to predict the 2-year OS rate. The areas under the curves of the BRIDGE model were 0.883, 0.871, and 0.858 for the training, validation, and test cohorts, respectively. The Hosmer-Lemeshow goodness-of-fit test showed a high agreement between the predicted and observed outcomes. Decision curve analysis indicated that VTE patients could benefit from the clinical application of the prognostic model. A BRIDGE risk score calculator for predicting the study result is available online (47.94.162.105:8080/bridge/). We established the BRIDGE model to precisely predict the risk for all-cause death in VTE patients after allo-HSCT. Identifying VTE patients who have a high risk of death can help physicians treat these patients in advance, which will improve patient survival.

A prognostic model (BATAP) with external validation for patients with transplant-associated thrombotic microangiopathy.

Transplant-associated thrombotic microangiopathy (TA-TMA) is a potentially life-threatening complication following allogeneic hematopoietic stem cell transplantation (allo-HSCT). Information on markers for early prognostication remains limited, and no predictive tools for TA-TMA are available. We attempted to develop and validate a prognostic model for TA-TMA. A total of 507 patients who developed TA-TMA following allo-HSCT were retrospectively identified and separated into a derivation cohort and a validation cohort, according to the time of transplantation, to perform external temporal validation. Patient age (odds ratio [OR], 2.371; 95% confidence interval [CI], 1.264-4.445), anemia (OR, 2.836; 95% CI, 1.566-5.138), severe thrombocytopenia (OR, 3.871; 95% CI, 2.156-6.950), elevated total bilirubin (OR, 2.716; 95% CI, 1.489-4.955), and proteinuria (OR, 2.289; 95% CI, 1.257-4.168) were identified as independent prognostic factors for the 6-month outcome of TA-TMA. A risk score model termed BATAP (Bilirubin, Age, Thrombocytopenia, Anemia, Proteinuria) was constructed according to the regression coefficients. The validated c-statistic was 0.816 (95%, CI, 0.766-0.867) and 0.756 (95% CI, 0.696-0.817) for the internal and external validation, respectively. Calibration plots indicated that the model-predicted probabilities correlated well with the actual observed frequencies. This predictive model may facilitate the prognostication of TA-TMA and contribute to the early identification of high-risk patients.

Clinical risk factors and prognostic model for idiopathic inflammatory demyelinating diseases after haploidentical hematopoietic stem cell transplantation in patients with hematological malignancies.

Idiopathic inflammatory demyelinating diseases (IIDDs) of the central nervous system (CNS) are rare but serious neurological complications of haploidentical hematopoietic stem cell transplantation (haplo-HSCT). However, the risk factors and a method to predict the prognosis of post-transplantation CNS IIDDs are not available. This retrospective study first reviewed data from 4532 patients who received haplo-HSCT during 2008-2019 in our center, and 184 patients (4.1%) with IIDDs after haplo-HSCT were identified. Grades II to IV acute graft-versus-host disease (aGVHD) (p < 0.001) and chronic GVHD (cGVHD) (p = 0.009) were identified as risk factors for developing IIDDs after haplo-HSCT. We then divided the 184 IIDD patients into a derivation cohort and validation cohort due to transplantation time to develop and validate a model for predicting the prognosis of IIDDs. In the multivariate analysis of the derivation cohort, four candidate predictors were entered into the final prognostic model: cytomegalovirus (CMV) infection, Epstein-Barr virus (EBV) infection, IgG synthesis (IgG-syn) and spinal cord lesions. The prognostic model had an area under the receiver operating characteristic curve of 0.864 (95% CI: 0.803-0.925) in the internal validation cohort and 0.871 (95% CI: 0.806-0.931) in the external validation cohort. The calibration plots showed a high agreement between the predicted and observed outcomes. Decision curve analysis indicated that IIDD patients could benefit from the clinical application of the prognostic model. The identification of IIDD patients after allo-HSCT who have a poor prognosis might allow timely treatment and improve patient survival and outcomes.