Semiquantitative analysis of interim 18 F-FDG PET is superior in predicting outcome in Hodgkin lymphoma patients compared to visual analysis.

PURPOSE: To investigate the prognostic value of interim PET (PETi) in adult HL patients, comparing visual with semiquantitative analysis.

MATERIAL AND METHODS: Retrospective analysis of Hodgkin's Lymphoma (HL) patients diagnosed between 2012 and 2016 in the Onco-Hematology Department of Instituto Português de Oncologia - Porto (median follow-up: 46.5 months [2,6-66,4]). Fifty-eight patients with available PET at diagnosis (PET0) and PETi data were included. PETi scans were analysed according to Deauville 5-point scale (5-PS), and cut-off values for changes in maximum standardized uptake value [SUVmax], peak SUV [SUVpeak], metabolic tumour volume [MTV] and total lesion glycolysis index [TLG] between PETi and PET0 were computed using ROC analysis. Visual and semiquantitative data were compared with each other in the prediction of patient outcomes.

RESULTS: Semiquantitative analysis obtained a higher sensitivity for persistent/relapsed disease compared to the 5-PS (70% vs. 10%, respectively), but lower specificity. It also demonstrated better predictive performance for response to first-line therapy (negative predictive value > 92%). The positive predictive value was similar for all five measurements. At 60 months of follow-up, there was a significant difference between the progression free survival (PFS) curves of patients with positive and negative PETi according to ΔSUVmax (56.9% vs. 88.0%, p < 0.05), ΔSUVpeak (55.9% vs 88.1%, p< 0.05), ΔMTV (35.3% vs. 88.7%, p< 0.05), and ΔTLG (42.4% vs. 88.1%, p< 0.05). Statistical significance wasn't reached when considering 5-PS results.

DISCUSSION: PETi interpretation according to a semiquantitative approach appears to discriminate HL patients better than the visual 5-PS analysis. This could allow better detection of persistent or early relapsed disease, while a negative PETi result could support de-escalating therapy intensity.