¹⁸F-FDG PET/CT Parameters Combined With Clinical Characteristic To Predict Prognosis In Diffuse Large B Cell Lymphoma

Background and Objective:In rituximab era,the four-risk classification of the International Prognostic Index(IPI)and Deauville 5-point scale were not enough to predict outcome of the patients with diffuse large B cell lymphoma(DLBCL).With the application of 18F-FDG PET/CT,multiple parameters have been proved to play an important role in predicting prognosis for different cancer.Our study expects to find out some effective parameters for prognosis,and develop a novel prognostic model combined with clinical characters to improve discriminative and predictive accuracy for DLBCL in rituximab era.Methods:We retrospectively collected data about 262 DLBCL patients who received R-CHOP and R-CHOP-like chemotherapy from January 2012 to September 2021 in Southern medical university Nanfang hospital.SPSS package(version 26.0)and R package version(4.2.1)were used for statistical analyses.A p<0.05 was used to define statistical significance.A Kaplan-Meier survival curve was generated and Logrank test was used to compare PFS and OS between each of the two groups.Survival analysis was performed by COX proportional hazard regression analysis,which were used to develop the nomogram.Results:1.Case information:262 patients with confirmed DLBCL had sufficient follow-up data and were eligible for analysis.A total of 228 patients(87%)were evaluated for an interim PET/CT response after four cycles of R-CHOP-like.The median follow-up was 34 months for the entire patient cohort.2.Baseline PET/CT imaging parameters and clinical characteristic for prediction of treatment outcome in DLBCL:Differentiating patients into two distinct groups based on IPI was easier to identify higher risk population.The Kaplan-Meier survival analysis for MTV,TLG,extranodal involvement,the maximum diameter of lesions,and the cell of origin(COO)classification showed a significant value of PFS and OS(p<0.01).However,the SUVmax was not suitable for predicting PFS and OS.Multivariate analysis revealed that higher MTV,higher IPI,bulky,and the non-GGB were a significant poor prognostic factor for both PFS and OS.The MTV,bulky,COO classification,and IPI were used to construct the nomogram and the web-based survival rate calculator,and the AUC values of the nomogram for predicting the OS rates for DLBCL was 0.759.3.Interim imaging parameters for prediction of treatment outcome in DLBCL:Differentiating patients into two distinct groups based on Deauville 5-point was easier to identify higher risk populations.The Kaplan-Meier survival analysis for interim SUVmax,MTV,TLG,and LLR showed a significant better PFS and OS in patients presenting lesser values as compared to those having higher(p<0.001).Moreover,the interim MTV was better than others including the Deauville 5-piont.4.The rate of change in the baseline and interim PET/CT parameters in predicting prognosis in DLBCL:the rate of change in the pre-and interim-SUVmax,MTV,and TLG were proved to predict prognosis for DLBCL.By the multivariate analysis showed theΔSUVmax was a significant index of predicting the prognosis for DLBCL.Conclusions:1.The analysis of baseline PET/CT parameter proved that the MTV,TLG,extranodal involvement,the maximum diameter of lesions,and the cell of origin(COO)classification were associated with prognosis of DLBCL.And the MTV,IPI,bulky,and the COO classification were independent prognostic factors for DLBCL.Giving the results of COX multivariate,a nomogram and web based survival rates was developed to guide cases’ recurrence risk assessment prior to treatment.2.The interim PET/CT parameters and the rates of change in the baseline and interim PET/CT parameters were associated with prognosis of DLBCL.Moreover,the interim MTV and the ΔSUVmax were better than the Deauville 5-point in prognosis.