Prognostic significance of pretransplant 18F-FDG PET/CT metrics in relapsed/refractory diffuse large B-cell lymphoma with partial response to salvage chemotherapy prior to autologous stem cell transplantation

Br J Haematol. 2025 Oct 1. doi: 10.1111/bjh.70186. Online ahead of print.

Abstract

Relapsed/refractory diffuse large B-cell lymphoma (R/R DLBCL) patients with partial response (PR) to salvage chemotherapy show heterogeneous outcomes after autologous stem cell transplantation (ASCT). Current visual positron emission tomography (PET) assessment inadequately identifies patients likely to benefit from transplantation. We retrospectively evaluated the prognostic value of semiquantitative metrics from interim 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) in 50 R/R DLBCL patients with PR to salvage chemotherapy who underwent ASCT. PET/CT scans were analysed visually using the Deauville 5-point scale and semiquantitatively using relative metrics measuring percentage maximum standardized uptake value (SUVmax) changes between baseline and interim scans (%ΔSUVmax-variants) and absolute metrics including tumour-to-liver SUV ratios (SUVTLR). Cut-off values were determined using quartile-based analysis. With a median follow-up of 107 months (range, 1.2-191.3), the median event-free survival (EFS) was 9.87 months. Interim SUVTLR ≥4.2 was the predictor of inferior EFS (hazard ratio [HR] 1.25, p = 0.002), while the percentage change in the hottest lesion's SUVmax (%ΔSUVmax-index) also showed prognostic significance (HR 0.91, p = 0.043). In contrast, visual Deauville scores at interim PET showed no prognostic difference between scores 4 and 5 (p = 0.094 for EFS). Therefore, semiquantitative PET metrics, particularly interim SUVTLR of 4.2, can provide prognostic information beyond visual assessment in PR patients prior to ASCT.

Keywords: 18F‐FDG PET/CT; Deauville 5‐point scale; diffuse large B‐cell lymphoma; prognostic biomarker; semiquantitative analysis; tumour‐to‐liver SUV ratio.