Work

P812: Predicting High-risk Disease Biology Using Artificial Intelligence Based Fdg Pet/ct Radiomics in Newly Diagnosed Multiple Myeloma

Year: 2023

Type: article

Abstract: Topic: 13. Myeloma and other monoclonal gammopathies - Biology & Translational Research Background: Survival outcomes in multiple myeloma have improved significantly due to novel therapeutic strategies. However, patients with high-risk disease biology have benefited less from these advances. Patients with high-risk disease can partly be identified by serum biomarkers and cytogenetics, but accurate risk stratification remains a challenge. Aims: To extract radiomics features from PET/CT scans of newly diagnosed patients using a deep learning model and to evaluate the suitability of these radiomics features to identify patients with high-risk disease. Methods: After pre-processing of the native DICOM images, we extracted discrete radiomics features from the diagnostic FDG PET/CT scans of patients with newly diagnosed multiple myeloma using the nnU-Net architecture (with default parameters, 41% adaptive SUVmax thresholding) with a model trained on external patients with hematologic malignancies and PyRadiomics. Limitations of this approach include heterogeneity introduced by varying imaging quality, technical differences related to PET reconstructions, and potential inclusion of non-neoplastic FDG activity in segmented regions. We evaluated the prognostic significance of 10 select tumor shape- and texture-related radiomics features. We integrated these radiomics features with demographic information, serum biomarkers, and bone marrow cytogenetics. We fit logistic regression models to evaluate the association between the radiomics features and (1) International Staging System (ISS) stage III, (2) high-risk cytogenetics, (3) failure to achieve 3-year overall survival, and (4) failure to achieve a complete response to first-line therapy. For modeling purposes, patients without measurable metabolic tumor volume were categorized as having undetectable radiomics features (fixed at value 0 for all features). All radiomics features were then divided into quintiles and dichotomized for analysis (top/ bottom quintile as cut point). Results: We analyzed the diagnostic FDG PET/CT scans of 748 patients with newly diagnosed multiple myeloma seen at Mayo Clinic between 2004 and 2019. Median age at diagnosis of these patients was 64 years (29-94) and 460 of them (62%) were men. After a median follow-up of 1.8 years (95% CI 0.1-7.4 years), median overall survival was 7.9 years (95% 6.3-NR). Most patients had measurable metabolic tumor volume (n=505, 68%) on their diagnostic FDG PET/CT. The mere presence (compared to the absence) of measurable metabolic tumor volume was not a prognostic factor (HR 0.99, 95% CI 0.71-1.38, p=0.938). Measures of tumor shape and texture were associated with stage, high-risk cytogenetics, and 3-year overall survival, but not response to therapy. ISS stage III (OR 2.83, 95% CI 1.86-4.33, p<0.001, n=748) and high-risk cytogenetics (OR 1.63, 95% CI 1.04-2.57, p=0.032, n=560) were most strongly associated with maximum 2D tumor diameter. There was also an association with failure to achieve a complete response that did not reach statistical significance (OR 1.61, 95% CI 0.95-2.75, p=0.075, n=540). Failure to achieve 3-year overall survival was strongly associated with maximum 2D tumor diameter (OR 2.26, 95%CI 1.45-3.54, p<0.001, n=748) and variance in gray level intensities, a measure of tumor homogeneity across volume (OR 1.84, 95% CI 1.01-3.33, p=0.043, n=748). Conclusion: Radiomics features obtained from diagnostic FDG PET/CT scans were strongly associated with established markers of high-risk disease. Standardized radiomics analysis represents a promising approach to real-time non-invasive risk stratification and support of clinical decision making that warrants further investigation and validation. Keywords: FDG-PET, Prognosis, Multiple myeloma, High risk

Source: HemaSphere

DOI: 10.1097/01.hs9.0000970152.05861.06