Key Points
• Source: American Journal of Cancer Research
• Conclusions/Relevance: “A combination of imaging features, clinical variables and genomic data was able to successfully predict complete response to first-line treatment in DLBCL patients, with the amplification of BCL6 as the genetic marker retaining the highest predictive value. Additionally, a panel of imaging features may provide important information when predicting treatment response, with lesion dissemination-related radiomic features deserving especial attention.”
• First-line (1L) R-CHOP is ineffective in up to one-third of DLBCL patients, therefore early identification is important in testing alternative treatment options. Spanish researchers assessed the ability of 18F-FDG PET/CT imaging features along with clinical data plus or minus genomic parameters to predict complete response (CR) to 1L treatment, with images extracted from prior treatment. Researchers segmented lesions as a whole to recapitulate tumor burden in 39 patients. Overall, 23 patients achieved long-term CR.
• “The best performance metrics were obtained with the combined model including genomic data and built applying the LDA method (AUC of , and 90% of balanced accuracy). The amplification of BCL6 was found to significantly contribute to explain response to first-line treatment in both manual and LDA models. Among imaging features, radiomic features reflecting lesion distribution heterogeneity (GLSZM_GrayLevelVariance, Sphericity and GLCM_Correlation) were predictors of response in manual models,” the authors wrote.
• Intriguingly, the researchers found that when they applied dimensionality reduction, the whole set of imaging features, which were mostly composed of radiomic features, contributed to elucidate 1L response. The researchers were able to model a nomogram predictive for response to 1L treatment.
• The authors wrote, “Our study stresses the relevance of genomic features when predicting treatment response. Thus, the highest AUC, balanced accuracy, sensitivity, and specificity ( , 90%, 100% and 80%, respectively) was achieved when these variables were combined together with clinical and imaging features in a LDA [linear discriminant analysis] model.”
• Limitations of the current study include its single-center nature, small sample size, and imbalance in MYC rearrangement, which did not permit analysis.