Among patients with newly diagnosed multiple myeloma (NDMM), identifying sarcopenia using a machine learning-based convolutional neural network algorithm appears to predict for overall survival (OS), according to research published in Cancer.
Previous oncological studies suggest that sarcopenia is linked with worse clinical outcomes among patients with solid tumors. Given that MM is typically diagnosed at a median age of 70 years, sarcopenia is more likely to be found in this patient population. There are, however, limited data about whether sarcopenia affects clinical outcomes among patients with NDMM, and analysis using techniques like computerized tomography (CT) can be both labor intensive and time consuming.
Machine learning tools may aid in this process. For this study, researchers evaluated whether a deep learning tool that analyzes CT images of the abdomen can help to determine prognoses among patients with NDMM.
Overall, data from 322 patients were included, among whom 171 had sarcopenia and 151 had no sarcopenia. At baseline, the median age was 66 years (range, 37-95), 62% of patients were male gender, and 25% of the 200 patients with available data had high-risk cytogenetics.
Analysis showed that the median OS in sarcopenic vs non-sarcopenic patients was 44 vs 90 months, respectively (P <.0001); the 2-year mortality rates were 40% and 18% (P <.0001).
Further analysis suggested that these findings were independent of disease stage, patient age, and the presence of high-risk cytogenetic features.
“Further investigations into the pathophysiology of sarcopenia and its negative impact on the OS of MM patients are required and may pave the way to understanding the complex interactions between the tumor micro‐environment and skeletal muscle tissue compartments,” the authors wrote in their report.
Disclosure: Some study authors declared affiliations with biotech, pharmaceutical, or device companies. Please see the original reference for a full list of authors’ disclosures.
Nandakumar B, Baffour F, Abdallah NH, et al. Sarcopenia identified by computed tomography imaging using a deep learning-based segmentation approach impacts survival in patients with newly diagnosed multiple myeloma. Cancer. Published online November 22, 2022. doi:10.1002/cncr.34545