The predictive power of 18F-FDG PET/CT two-lesions radiomics and conventional models in classical Hodgkin's Lymphoma: a comparative retrospectively-validated study

Abstract

In a previous preliminary study, radiomic features from the largest and the hottest lesion in baseline ¹⁸F-FDG PET/CT (bPET/CT) of classical Hodgkin's Lymphoma (cHL) predicted early response-to-treatment and prognosis. Aim of this large retrospectively-validated study is to evaluate the predictive role of two-lesions radiomics in comparison with other clinical and conventional PET/CT models. cHL patients with bPET/CT between 2010 and 2020 were retrospectively included and randomized into training-validation sets. Target lesions were: Lesion_A, with largest axial diameter (D_max); Lesion_B, with highest SUV_max. Total-metabolic-tumor-volume (TMTV) was calculated and 212 radiomic features were extracted. PET/CT features were harmonized using ComBat across two scanners. Outcomes were progression-free-survival (PFS) and Deauville Score at interim PET/CT (DS). For each outcome, three predictive models and their combinations were trained and validated: - radiomic model "R"; - conventional PET/CT model "P"; - clinical model "C". 197 patients were included (training = 118; validation = 79): 38/197 (19%) patients had adverse events and 42/193 (22%) had DS ≥ 4. In the training phase, only one radiomic feature was selected for PFS prediction in model "R" (Lesion_B F_cm.corr, C-index 66.9%). Best "C" model combined stage and IPS (C-index 74.8%), while optimal "P" model combined TMTV and D_max (C-index 63.3%). After internal validation, "C", "C + R", "R + P" and "C + R + P" significantly predicted PFS. The best validated model was "C + R" (C-index 66.3%). No model was validated for DS prediction. In this large retrospectively-validated study, a combination of baseline ¹⁸F-FDG PET/CT two-lesions radiomics and other conventional models showed an added prognostic power in patients with cHL. As single models, conventional clinical parameters maintain their prognostic power, while radiomics or conventional PET/CT alone seem to be sub-optimal to predict survival.

Keywords: 18F-FDG; Hodgkin; PET/CT; Prediction; Radiomics.

Fig. 1

Target lesion contouring with PET_Edge tool from MIM Encore Software (version 7.0.5 MIM Encore Software Inc., Cleveland, OH). (A) Maximum-intensity projection of an 18-year-old male patient with Stage IIA nodular sclerosing Hodgkin’s Lymphoma. Axial CT (B) and PET (C) views of Lesion_A (pink contoured VOI, D_max = 4.33 cm). Axial CT (D) and PET (E) views of Lesion_B (cyan contoured VOI, SUV_max = 11.81)

Fig. 2

Total Metabolic Tumor Volume (TMTV) contouring with LesionID tool from MIM Encore Software (version 7.0.5 MIM Encore Software Inc., Cleveland, OH). (A) Maximum-intensity projection of an 18-year-old male patient with Stage IIA nodular sclerosing Hodgkin’s Lymphoma. (B) Semiautomatic threshold-based segmentation of the whole body, comprising some physiological sites of uptake (e.g., arrows), later excluded from the computation. (C) Revised lesion contouring, then processed by the software to obtain TMTV volumetric parameter (149,08 ml); pink circle: region of interest on the right liver lobe for PERCIST-based cut-off criterion for volumes of interest determination

Fig. 3

Statistical workflow

Fig. 4

Flowchart of patients’ selection. NLP-HL = Nodular Lymphocyte Predominant Hodgkin’s Lymphoma

Fig. 5

Baseline and further PET exams of (A) Patient 1 (Stage 4, IPS 3, bPET in A), who had a low F_cm.corr and still experienced no event at 79.8 months of follow-up; (B) Patient 2 (Stage 3, IPS 4, bPET in B) had a high value of F_cm.corr and progressed after 5.1 months of follow-up