Abstract
Background: Follicular lymphoma (FL) is the second most prevalent non-Hodgkin lymphoma, characterized by an indolent course with recurrent patterns of remission and relapse. Despite high rates of prolonged remission with anti-CD20 immunochemotherapy, 20–30% of patients experience progression of disease within 24 months of first-line therapy (POD24), which predicts transformation and early mortality. There is currently no reliable baseline characteristic that can predict the incidence of POD24. Quantitative fluorine-18 fluorodeoxyglucose positron emission tomography (PET/CT) in the form of total metabolic tumor volume (TMTV) is a promising prognostic tool in indolent and aggressive lymphomas. This retrospective study aims to determine whether baseline TMTV predicts POD24 in patients with FL in a cohort of patients from Brazil.
Methods: A retrospective analysis was performed including patients diagnosed with FL between January 2015 and March 2025 at two reference centers in Brazil. Recorded features included epidemiologic, clinical, and imaging (SUVmax and TMTV). All PET/CT images were analyzed by a trained physician. TMTV was accurately calculated for 81 patients who had baseline PET/CTs performed at our institution. The remaining 19 patients, with external PET/CTs, were assessed for all other variables available. A threshold of 41% SUVmax was used to delineate the Volume of interest (VOI) and total metabolic tumor volume (MTV). TMTV was defined as the sum of the MTV values of all individual lesions. Bone marrow and splenic involvement were considered in volume measurement only if there was focal lesions. Association between POD24 status and baseline variables were evaluated by Pearson's chi-square test with Yates' continuity correction. Accuracy and optimal cutoff value for TMTV was determined based on ROC curve analysis and supported by thresholds previously reported in the literature.
Results: Our cohort included 100 patients (median age: 58 years; 53% male), with 20% experiencing POD24. The ROC curve analysis for TMTV yielded an AUC of 0.62 (95% CI: 0.47–0.77), indicating modest discriminative power for predicting POD24. At a cutoff of 400 cm³, TMTV demonstrated 83.3% specificity and 42.1% sensitivity. While a high TMTV (≥ 400 cm³) was significantly associated with POD24 (χ2 = 4.78; p = 0.029), none of the other evaluated variables – including age > 60, advanced stage, bone marrow involvement, extranodal involvement, B symptoms, elevated LDH, low albumin, high β₂-microglobulin or high SUVmax – reached statistical significance. In the multivariable logistic regression model adjusted for variables with strong clinical relevance such as FLIPI2 score and SUVmax, high baseline TMTV remained the only independent predictor of POD24 (OR 3.92; 95% CI: 1.20–12.80; p = 0.0236).
Conclusions: POD24 is one of the most powerful predictors of adverse outcomes in FL, including transformation and decreased overall survival. Identifying patients at risk of POD24 remains a major unmet need in clinical practice. In a large cohort from the FOLL12 trial, high TMTV emerged as an independent predictor of POD24, regardless of FLIPI2 score and treatment regimen, underscoring its role in identifying patients at high risk of early progression. We were able to validade this findings in a Brazilian cohort of Follicular lymphoma, an ethinicity usually not represented in other studies (latinos). TMTV remains a strong predictor of early failure and future trials should incorporate TMTV as a risk factor for first line therapy.
