Significance of initial, interim and end-of-therapy 18F-FDG PET/CT for predicting transformation risk in follicular lymphoma

Mixue Xie,Lulu Wang,Qi Jiang,Xuxia Luo,Xin Zhao,Xueying Li,Jie Jin,Xiujin Ye,Kui Zhao

DOI: https://doi.org/10.1186/s12935-021-02094-5

IF: 6.429

2021-07-26

Cancer Cell International

Abstract:Abstract Background Histological transformation (HT) of follicular lymphoma to a more aggressive lymphoma is a serious event affecting patients’ outcomes. To date, no strong clinical HT predictors present at diagnosis have yet been identified. The fluorodeoxyglucose (FDG)-positron emission tomography (PET)/computed tomography (CT) is highlighted as a non-invasive diagnostic tool for the detection of HT, but its ability to predict HT at early stage of disease has not been clear. Therefore, this study investigated the predictive values of the pre-transformation standardized uptake value (SUV max ) for the risk of transformation in FL. Methods This retrospective study involved 219 patients with FL between June 2008 and October 2019 who had undergone 18 F-FDG PET/CT scan. One hundred and thirty-two, 64, and 78 patients underwent PET at baseline (PET baseline ), interim (PET interim ) and end-of-induction therapy (PET end ), respectively. Qualitative assessment was performed using the 5-point Deauville scale. Statistical analysis was done using Cox regression models, receiver operating characteristic (ROC) analysis, and Kaplan–Meir survival curves. Results Of the 219 patients included, 128 had low-grade FL (grade 1–2) and 91 had high-grade FL (grade 3a). HT eventually occurred in 30 patients. The median time to HT was 13.6 months. Among clinical indicators, advance pathological grade was shown as the most significant predictor of HT (HR = 4.561, 95% CI 1.604–12.965). We further assessed the relationship between PET and HT risk in FL. Univariate Cox regression determined that SUV baseline and SUV end were significant predictors for HT, while neither SUV interim nor qualitative assessment of Deauville score has predictive value for HT. Due to the noticeable impact of high pathological grade on the HT risk, we conducted the subgroup analysis in patients with low/high pathological grade, and found SUV baseline could still predict HT risk in both low-grade and high-grade subgroups. Multivariate analysis adjusted by FLIPI2 score showed the SUV baseline (HR 1.065, 95% CI 1.020–1.111) and SUV end (HR 1.261, 95% CI 1.076–1.478) remained as significant predictors independently of the FLIPI2 score. According to the cut-off determined from the ROC analysis, increased SUV baseline with a cutoff value of 14.3 and higher SUV end with a cutoff value of 7.3 were highly predictive of a shorter time to HT. Conclusions In follicular lymphoma, quantitative assessment used SUV max at the pre-treatment and end-of-treatment PET/CT scan may be helpful for early screen out patients at high risk of transformation and guide treatment decisions.

oncology