Prognostic value of baseline metabolic tumor volume in early-stage Hodgkin lymphoma in the standard arm of the H10 trial.

We tested baseline positron emission tomography (PET)/computed tomography (CT) as a measure of total tumor burden to better identify high-risk patients with early-stage Hodgkin lymphoma (HL). Patients with stage I-II HL enrolled in the standard arm (combined modality treatment) of the H10 trial (NCT00433433) with available baseline PET and interim PET (iPET2) after 2 cycles of doxorubicin, bleomycin, vinblastine, and dacarbazine were included. Total metabolic tumor volume (TMTV) was measured on baseline PET. iPET2 findings were reported negative (DS1-3) or positive (DS4-5) with the Deauville scale (DS). The prognostic value of TMTV was evaluated and compared with baseline characteristics, staging classifications, and iPET2. A total of 258 patients were eligible: 101 favorable and 157 unfavorable. The median follow-up was 55 months, with 27 progression-free survival (PFS) and 12 overall survival (OS) events. TMTV was a prognosticator of PFS (P < .0001) and OS (P = .0001), with 86% and 84% specificity, respectively. Five-year PFS and OS were 71% and 83% in the high-TMTV (>147 cm3) group (n = 46), respectively, vs 92% and 98% in the low-TMTV group (≤147 cm3). In multivariable analysis including iPET2, TMTV was the only baseline prognosticator compared with the current staging systems proposed by the European Organization for Research and Treatment of Cancer/Groupe d'Etude des Lymphomes de l'Adulte, German Hodgkin Study Group, or National Comprehensive Cancer Network. TMTV and iPET2 were independently prognostic and, combined, identified 4 risk groups: low (TMTV≤147+DS1-3; 5-year PFS, 95%), low-intermediate (TMTV>147+DS1-3; 5-year PFS, 81.6%), high-intermediate (TMTV≤147+DS4-5; 5-year PFS, 50%), and high (TMTV>147+DS4-5; 5-year PFS, 25%). TMTV improves baseline risk stratification of patients with early-stage HL compared with current staging systems and the predictive value of early PET response as well.

Deep Learning Approach to Automatize TMTV Calculations Regardless of Segmentation Methodology for Major FDG-Avid Lymphomas.

The total metabolic tumor volume (TMTV) is a new prognostic factor in lymphomas that could benefit from automation with deep learning convolutional neural networks (CNN). Manual TMTV segmentations of 1218 baseline 18FDG-PET/CT have been used for training. A 3D V-NET model has been trained to generate segmentations with soft dice loss. Ground truth segmentation has been generated using a combination of different thresholds (TMTVprob), applied to the manual region of interest (Otsu, relative 41% and SUV 2.5 and 4 cutoffs). In total, 407 and 405 PET/CT were used for test and validation datasets, respectively. The training was completed in 93 h. In comparison with the TMTVprob, mean dice reached 0.84 in the training set, 0.84 in the validation set and 0.76 in the test set. The median dice scores for each TMTV methodology were 0.77, 0.70 and 0.90 for 41%, 2.5 and 4 cutoff, respectively. Differences in the median TMTV between manual and predicted TMTV were 32, 147 and 5 mL. Spearman's correlations between manual and predicted TMTV were 0.92, 0.95 and 0.98. This generic deep learning model to compute TMTV in lymphomas can drastically reduce computation time of TMTV.