Radiomics-Based Models Can Accurately Diagnose MM and Bone Metastases, Study Finds

Differentiation between multiple myeloma (MM) and bone metastases from other cancers can be difficult, but radiomics-based models have potential to improve diagnostic accuracy.

Differentiation between spine multiple myeloma (MM) and bone metastases (BM) related to other cancer types can be difficult given their similar sites of occurrence, clinical traits, and imaging features. To assist traditional methods of classification, a study published in Frontiers in Medicine evaluated novel radiomics models based on 18F-fluorodeoxyglucose positron emission tomography/CT (18F-FDG PET/CT) for the classification of spine MM and BM.

Misclassification of bone marrow lesions can affect patient survival and quality of life. Lesions classified incorrectly are also often misdiagnosed as other orthopedic diseases, and treatment options are significantly different for each condition. Reducing the risk of misdiagnosis in this setting is therefore a crucial endeavor.

Although serum markers such as creatinine, globulin, and alkaline phosphatase can help differentiate between BM and MM, patients with certain forms of myeloma tend to have normal or low numbers of these markers. With 18F-FDG PET/CT imaging—the International Myeloma Working Group’s recommended imaging method for MM—anatomical and metabolic information are both used to assess bone damage and lesions with high sensitivity and specificity. But some lesions, like osteolytic lesions, are still difficult for even experienced clinicians to identify.

Radiomics uses machine learning to convert imaging features into high-dimensional data, allowing noninvasive evaluation of a tumor’s spatial heterogeneity and assisting in personalizing treatment for individual patients. The current study explores the potential for using radiomics in combination with 18F-FDG PET/CT to identify MM vs BM, given previous research has predominantly evaluated radiomics based on CT and MRI images.

A total of 131 patients were included in the study, 86 with a BM diagnosis and 45 with confirmed MM. In all, 184 lesions were randomly divided into a training group and a validation group at a 7:3 ratio to develop the radiomics models. The training group included 80 BM lesions and 49 MM lesions, and the validation group contained 34 and 21 lesions, respectively.

Ten and 8 texture features were selected from CT and PET, respectively, to build the models after least absolute shrinkage and selection operator regression and 10-fold cross-validation. There were 3 radiomics models: 2 constructed with CT and PET and using multivariate logistic regression and a ComModel using PET plus the maximum standardized uptake value of each lesion. Two experienced physicians evaluated the images in a double-blind format to test accuracy against the radiomics systems.

In both the training and evaluation groups, the 3 radiomics models performed well. The area under the receiver operating characteristic curve (AUC) was 0.909 in the CT training group, 0.949 in the PET training group, and 0.973 in the ComModel training group, and the CT, PET, and ComModel validation group AUCs were 0.897, 0.929, and 0.948, respectively.

The PET and ComModel were significantly better at diagnosing BM and MM vs the expert clinicians, while there was no significant statistical difference between the CT model and physician evaluation.

The study was limited due to its single-center nature. Therefore, more research would determine how the findings’ generalizability. Some patients also did not have pathological findings and received their diagnosis based on combined pathological findings and follow-up results, the authors noted. But overall, the study results are promising.

The authors concluded, “Radiomics could transcend subjective visual assessment to provide an objective evaluation of lesion and tissue heterogeneity, which served as a new tool to provide valuable information about the microenvironment of lesions that cannot be observed by the human eyes.”

Reference

Jin Z, Wang Y, Wang Y, Mao Y, Zhang F, Yu J. Application of 18F-FDG PET-CT images based radiomics in identifying vertebral multiple myeloma and bone metastases. Front Med (Lausanne). Published online April 18, 2022. doi:10.3389/fmed.2022.874847