RT Journal Article
SR Electronic
T1 Quantification of intratumor heterogeneity derived from baseline FDG PET/CT in untreated mantle cell lymphoma patients enrolled in a prospective phase III trial of the LYSA group: preliminary results
JF Journal of Nuclear Medicine
JO J Nucl Med
FD Society of Nuclear Medicine
SP 429
OP 429
VO 56
IS supplement 3
A1 Carlier, Thomas
A1 Bailly, Clément
A1 Hatt, Mathieu
A1 Kraeber-Bodéré, Françoise
A1 Visvikis, Dimitris
A1 Le Gouill, Steven
A1 Bodet-Milin, Caroline
YR 2015
UL http://jnm.snmjournals.org/content/56/supplement_3/429.abstract
AB 429 Objectives In untreated mantle cell lymphoma (MCL) patients enrolled in the prospective phase III Lyma trial, it has recently been shown that few quantitative metrics in pre-treatment 18FDG-PET are strong prognostic factors of progression-free survival (PFS). Heterogeneity quantification through textural features (TF) can be an alternative to conventional metrics or may help to enhance the prognostic value of SUV-based metrics. The objective of this study was to fully assess the characteristics of TF in relation with other standard metrics in untreated MCL patients.Methods Pre-treatment FDG-PET of 33 patients were considered. The region with the highest uptake was extracted and several metrics were derived: SUVmax, SUVpeak, SUVmean, metabolic tumor volume (MTV), total lesion glycolysis (TLG) and 15 TF chosen among the most used. For each metric requiring a segmentation step, 3 delineation methods were compared including the Fuzzy Locally Adaptative Bayesian approach, 40% of SUVmax and SUV>2.5. The robustness of each metric vs. the segmentation approach and the non-linear correlation between each metric were assessed by the Friedman and Spearman rank test respectively.Results Clinical characteristics and PFS of the 33 patients were similar to these of the entire Lyma cohort (n=299). Only 6 out of 15 TF were found independent of the delineation method. No correlation was found between any TF and SUVs. Three TF offered non redundant information with other TFs or standard metrics.Conclusions Our preliminary results show that most of the TF are redundant and/or dependent on the delineation approach. Thus only 3 TF may provide reliable and complementary information with respect to SUV-based metrics in this histological subtype of hematologic malignancy within the present context of multi-centric trial. Results on additional patients in the Lyma trial will be presented.