Recently, the team led by Professor Yu-jing Hu from Hebei Provincial People’s Hospital published a study exploring the clinical value of 18F-FDG PET/CT metabolic parameters in predicting molecular subtypes of diffuse large B-cell lymphoma. The study also used visualization methods to analyze the importance of each parameter in this predictive process. 
 

This study retrospectively analyzed 208 patients diagnosed with diffuse large B-cell lymphoma via pathology from January 2020 to February 2024. The patients were divided into two groups based on cellular origin: the germinal center group and the non-germinal center group. The study compared clinical data and PET/CT metabolic parameters between the two groups, including maximum standardized uptake value, minimum standardized uptake value, mean standardized uptake value, and standard deviation of standardized uptake value.

The results showed that the non-germinal center group had significantly higher rates of B symptoms and several metabolic parameters mentioned above compared to the germinal center group. By constructing clinical, PET, and combined models for predictive performance analysis, the combined model exhibited the highest area under the curve (AUC), reaching 0.759—a significant improvement over either the clinical or PET model alone. The study further employed the Shapley additive explanation method to analyze the importance of predictor variables, revealing that the maximum standardized uptake value contributed the most, followed by the average standardized uptake value, B symptoms, the minimum standardized uptake value, and the standard deviation of standardized uptake values. 

The study indicates that a combined model integrating B symptoms with PET/CT metabolic parameters can effectively distinguish between different molecular subtypes of diffuse large B-cell lymphoma (DLBCL). This provides new insights for early, non-invasive clinical assessment of patient prognosis and the formulation of personalized treatment strategies. The findings support the potential value of PET/CT in the precise subtyping of lymphoma and the prediction of treatment efficacy.

Source: China Nuclear Technology Network