Nuclear Medicine Center/Semmelweis University, Budapest
Introduction: Interim 18F- FDG PET/CT scans are routinely performed for the evaluation of early therapeutic response in high-grade lymphoma patients. This assessment is most commonly based on the visual Deauville five-point scale (DS). However, recently, semi-quantitative methods based on the standardized FDG-uptake value (SUV) have been introduced.
Aim: To compare the visual and semiquantitative assessment methods in interim PET/CT scans of diffuse large B-cell lymphoma (DLBCL) patients.
Methods: Imaging and clinical data of 90 Chilean and Hungarian de novo DLBCL patients who participated in a clinical study organized by the International Atomic Energy Agency were retrospectively analysed. We measured SUV parameters in the most active residual lymphoma lesion and in an intrahepatic reference volume of interest with a diamater of 3 cm. For semiquantitative evaluation, the modified qPET (m-qPET) was calculated as SUVpeak(lesion)/ SUVmean(reference), while the ratio-PET (rPET) was given as SUVmax(lesion)/ SUVmax(reference).
Results: Receiver Operating Characteristic (ROC)-analysis yielded a similar area under the curve (AUC) with each three methods (AUC DS: 0.7; AUC m-qPET: 0.73; AUC rPET: 0.72). Patients were divided into two groups according by each optimal cut-off value (m-qPET: 1.53; rPET: 1.54) obtained from the ROC-analysis (and by the conventionally used grouping of DS1-3 vs. DS4-5) and significantly different progression-free survival could be observed between each of the two groups (p<0,001). Positive predictive values were higher with both semiquantitative methods (m-qPET: 61% and rPET: 69% vs. DS: 52%), while negative predictive values remained similarly high with the visual and the semiquantitative approaches (m-qPET: 86%, rPET: 86% and DS: 88%).
Conclusion: Semiquantitative evaluation of interim PET/CT scans has a better prognostic value than the visual score-based assessment. This – in conjunction with other biomarkers – could enable the early and precise identification of patients who need their treatment to be escalated or possibly de-escalated.
Doctoral School: Clinical Medicine
Program: Clinical Haematology
Supervisor: Tamás Györke
E-mail address: email@example.com