A predictive algorithm using clinical and laboratory parameters may assist in ruling out and in diagnosing MDS

Area of Science:

• Hematology
• Computational Biology
• Medical Informatics

Background:

• Myelodysplastic syndromes (MDS) are a group of bone marrow disorders characterized by ineffective hematopoiesis and an increased risk of transformation to acute myeloid leukemia.
• Diagnosis of MDS traditionally relies on invasive bone marrow examination, which carries risks and may not be accessible to all patients.
• There is a need for noninvasive methods to accurately predict or exclude MDS, particularly in patients presenting with unexplained cytopenias.

Purpose of the Study:

• To develop and validate a noninvasive, Web-based application for the prediction or exclusion of myelodysplastic syndromes (MDS).
• To assess the diagnostic accuracy of the application using demographic, clinical, and laboratory variables.
• To provide a tool that assists physicians in managing patients with suspected MDS, potentially avoiding invasive bone marrow examinations.

Main Methods:

• Gradient-boosted models (GBMs) were trained using data from 502 MDS patients and 502 controls from the European MDS (EUMDS) registry.
• Model performance was evaluated using area under the receiver operating characteristic curve (AUC), sensitivity, and specificity, with validation through 100 times fivefold cross-validation.
• Key discriminating variables including age, sex, hemoglobin, white blood cells, platelets, mean corpuscular volume, neutrophils, monocytes, glucose, and creatinine were identified.

Main Results:

• The developed GBM model achieved a high AUC of 0.96 (95% CI, 0.95-0.97), indicating excellent discriminatory power.
• The application accurately predicted or excluded MDS in 86% of patients with unexplained anemia.
• A GBM score threshold (< 0.68) achieved a negative predictive value of 0.94 for excluding MDS, while a threshold (≥ 0.82) yielded a positive predictive value of 0.88 for predicting MDS.

Conclusions:

• A noninvasive, Web-based application utilizing GBMs can accurately predict or exclude myelodysplastic syndromes in most patients, particularly those with unexplained anemia.
• The tool demonstrates high predictive values, offering a valuable noninvasive alternative or adjunct to bone marrow examination for MDS diagnosis.
• Future research will incorporate peripheral blood cytogenetics/genetics and prospective validation for enhanced prognostication and clinical utility.