MRI Radiomics Show Accuracy in Predicting High-Risk Cytogenetic Abnormalities Among Patients With Multiple Myeloma
Nomograms based on magnetic resonance imaging (MRI) may be useful in predicting high-risk cytogenetic abnormalities among patients with multiple myeloma, according to study results published in Clinical Radiology.
Researchers conducted a 2-center cohort study to evaluate the use of MRI radiomic features in predicting high-risk cytogenetic abnormalities among patients with MM.
Patients (n = 195) were stratified by HRCA status (HRCA vs non-HRCA), confirmed via fluorescence in situ hybridization. Between both HRCA status cohorts, patients were further stratified into training (n = 111) and validation (n = 48) cohorts.
Radiomics predictive models were derived from T1WI, T2WI, and FS-T2WI images and clinical factors with C-indexes used to determine the most efficient models. The researchers tested the most efficient nomogram performance on 36 patients who served as an external test cohort in a separate test center.
Analysis of the optimal model revealed among the training and validation cohorts C-indexes corrected via the 1000 bootstrap method at 0.79 and 0.80 in single-sequence, 0.83 and 0.84 double-sequence, and 0.88 and 0.84 multi-sequence images, respectively. When the optimal model, FT2+age, FT2+1+age, and FT2+2+1+age combined, was tested in the external test cohort, the C-index was 0.70 in single-sequence, 0.76 double-sequence, and 0.77 multi-sequence, respectively.
“MRI radiomics can be used to predict HRCAs in MM patients, which will be helpful for clinical decision-making and prognosis evaluation before treatment,” the researchers concluded.
Source:
Liu S, Liu C, Pan H, et al. Magnetic resonance imaging–based nomograms predict high-risk cytogenetic abnormalities in multiple myeloma: a two-centre study. Clinical Radiology. December 5, 2024. doi: 10.1016/j.crad.2024.106768
