The Optimal Background Cut-Off Threshold for Calculating Volumetric PET Biomarkers as Prognostic Factors in Classical Hodgkin Lymphoma

AIM. To determine the optimal background cut-off threshold for calculating volumetric PET biomarkers as prognostic factors in classical Hodgkin lymphoma (cHL). MATERIALS & METHODS. The retrospective analysis was based on the data from 79 patients with various stages of newly diagnosed cHL, who underwent PET/CT at the AN Bakulev Center for Cardiovascular Surgery. The analysis was focused on 341 PET/CT scans (prior to treatment, after completing the first-line therapy, and at different stages of follow-up period). The median follow-up was 12 months. During this period, 75 % of patients showed sustaining remission, in 16 % of patients refractoriness and in 9 % of patients relapse were observed; no deaths were reported. The volumetric PET biomarkers, i.e., the total metabolic tumor volume (TMTV) and total lesion glycolysis (TLG), were calculated automatically using four background cut-off thresholds: two absolute (SUVmax ≥ 2.5 and SUVmax ≥ 4.0), one relative (41 % of SUVmax), and one background (SUVliver) thresholds. RESULTS. To determine the prognostic value of the parameters under consideration, two patients subgroups were analyzed: the 1st one with remission (n = 59) and the 2nd one with relapsed/refractory cHL (n = 20). These subgroups showed significant differences in TMTV and TLG values with three different background cut-off thresholds (SUVmax ≥ 2.5, 41 % of SUVmax, SUVliver). A significant correlation was established between poor outcomes and high levels of TLG2.5 (p = 0.038), TMTV41%, and TLG41% (p = 0.012 and p = 0.040, respectively) as well as TMTVliver and TLGliver (p = 0.023 and p = 0.009, respectively). By univariate and multivariate analyses, the baseline TMTV41% values were associated with the risk of poor outcomes in cHL irrespective of the GHSG prognostic score. CONCLUSION. High level of the baseline volumetric PET biomarkers is associated with poor prognosis in cHL. By comparing the values of volumetric PET biomarkers, the optimal background cut-off threshold appeared to be 41 % of SUVmax. The TMTV and TLG values calculated with this threshold showed the highest accuracy in assessing the risk of poor outcomes irrespective of other prognostic factors.