Predicting Invasive Aspergillosis in Hematology Patients by Combining Clinical and Genetic Risk Factors with Early Diagnostic Biomarkers

Personalized medicine provides a strategic approach to the management of IA. The incidence of IA in high-risk hematology populations is relatively low (Aspergillus exposure in patients with a potentially similar clinical risk. Nonclinical variables, including genetic mutations that increase susceptibility to IA, could explain why only certain patients develop the disease. This study screened for mutations in 322 hematology patients classified according to IA status and developed a predictive model based on genetic risk, established clinical risk factors, and diagnostic biomarkers. Genetic markers were determined by real-time PCR and, with clinical risk factors and Aspergillus PCR results, subjected to multilogistic regression analysis to identify a best-fit model for predicting IA. The probability of IA was calculated, and an optimal threshold was determined. Mutations in dectin-1 (rs7309123) and DC-SIGN (rs11465384 and rs7248637), allogeneic stem cell transplantation, respiratory virus infection, and Aspergillus PCR positivity were all significant risk factors for developing IA and were combined in a predictive model. An optimal threshold requiring three positive factors generated a mean sensitivity/specificity of 70.4%/89.2% and a probability of developing IA of 56.7%. In patients with no risk factors, the probability of developing IA was 2.4%, compared to >79.1% in patients with four or more factors. Using a risk threshold of 50%, preemptive therapy would have been prescribed for 8.4% of the population. This pilot study shows that patients can be stratified according to risk of IA, providing personalized medicine based on strategic evidence for the management of IA. Further studies are required to confirm this approach.