PET metabolic response in Lugano: Comparison of qualitative and quantitative outcome by independent central review.

Abstract

7065 Background: The 2014 Lugano Criteria is the standard for radiographic assessment in lymphoma clinical trials, emphasizing qualitative evaluation of 18F-FDG PET-CT (PET) metabolic response (MR). In 2017, Van Heertum et al. introduced a quantitative approach defining strict percent change of standard uptake values (SUV) determination of MR, primarily for partial metabolic response (PMR) and progressive metabolic disease (PMD). Both Lugano and Van Heertum compared to baseline for determination of PMD, though industry practice frequently compare PMD to nadir. While both qualitative and quantitative approaches have been utilized in pivotal trials, few studies have compared qualitative and quantitative Lugano radiographic metabolic assessments to date. Thus, this study aims to investigate PET MR timepoint responses (TPRs) from blinded independent central review (BICR) radiologists using qualitative Lugano approach vs quantitatively derived TPRs. Methods: 3416 radiologist qualitatively assessed designated (desPET) responses consisting of PET TPR, 5PS, most-FDG avid lesion SUV, and liver SUV were analyzed. For desPET TPRs, radiologists compared MRs to baseline and nadir. We quantitatively derived PET (derPET) TPRs using the Van Heertum et al.,2017 %Δ utilizing SUV and 5PS, comparing to baseline and nadir, defining PMR as ≥25% decrease with a 5-point score (5PS) of 4 or 5; PMD as ≥50% increase with 5PS of 4 or 5; complete metabolic response (CMR) as 5PS of 1–3 without residual disease; and no metabolic response (NMR) as 5PS of 4 or 5, not meeting PMR or PMD criteria. To understand the merit of qualitative PET assessments, the desPET were compared to the derPET TPRs. Then, the desPET TPRs were compared among radiologists who reviewed the same case to determine concordance. Results: Results showed desPET differed from the derPET for 1198 TPRs (35.1%). When PMD was the derPET TPR, the desPET TPRs were: 270 PMD (53.9%), 134 NMR (26.8%) and 89 PMR (17.8%). When CMR was the derPET TPR, the desPET TPRs were: 830 CMR (51.7%), 634 PMR (39.5%), and 122 PMD (7.6%). When PMR was derPET TPR, desPET TPRs were: 177 PMR (61.9%), 55 NMR (19.2%), 51 PMD (17.8%). In the instances where desPET TPRs differed from the derPET, radiologists reviewing the same cases had concordance in desPET TPRs for 79.5% of cases. Conclusions: Our data show notable differences in PET MR when comparing derPET and desPET TPRs. Almost 20% BICR radiologists assessed a PMR, when our derived values were PMD and in over half the cases BICR radiologists assessed CMR when the derPET was PMR, thus suggesting that the quantitative approach results in more PD and fewer CR cases. When determining reader concordance, nearly 80% of readers aligned in desPET TPR when their TPRs differed from the derPET, suggesting that a qualitative assessment supports medical judgment in determining patient metabolic disease and consistency between readers assessments.