Objective: To investigate the prognosis value of early and interim ¹⁸F-FDG-PET/CT scan in patients with diffuse large B- cell lymphoma （DLBCL） to establish the suitable criteria for evaluating posttherapeutic lesions in scans.
Methods: Fifty- six newly diagnosed DLBCL patients were enrolled in the study, and underwent baseline, early and interim ¹⁸F- FDG PET/CT scans. Five- point and % ΔSUVmax criteria were used separately to interpret ¹⁸F- FDG PET/CT images. Interobserver reproducibility was assessed with the kappa test（κ）, and thresholds of %ΔSUVmax were calculated via receiver operating characteristic curve（ROC）. Survival curves were obtained using Kaplan-Meier curves and log-rank test. Cox regression analysis was used for multi-factor analysis.
Results: Median follow-up was 24 months（6 to 42 months）. The kappa value of the five- point scale was above 0.600 with the reference background set in the liver（Score≥4）. The optimal threshold of %ΔSUVmax was 81% for early PET/CT and 74% for interim PET/CT. Survival analysis showed both early and interim PET/CT scans could predict the outcome of 56 patients with DLBCL, and 3-year PFS and OS of PET-negative patients were significantly higher than those of PET-positive ones（P<0.05）. Five-score criteria were more accurate in evaluating 3- year PFS of DLBCL patients in the interim PET/CT scan（76.79%）. %ΔSUVmax criteria were better for interpreting 3-year OS（76.79% and 83.93%）. Multi-factor analysis demonstrated that early and interim PET/CT were solid predicting factors for DLBCL patients.
Conclusion: Early and interim PET/CT scans could predict the outcome of patients with DLBCL, treated with R-CHOP/CHOP. Three-year OS was more accurate in early and interim PET/CT using 66 %ΔSUVmax criteria as an interpretation, while 3-year PFS was more accurate in interim PET/CT by five scores criteria.
目的: 探讨化疗早期及中期18F-FDG PET/CT显像判定弥漫大B细胞淋巴瘤(DLBCL)患者预后的价值，对比分析显像中残存病灶的判读标准。
方法: 56例初治DLBCL患者接受R-CHOP/CHOP方案治疗，于化疗前、化疗早期(2周期，PET-2)及化疗中期(4周期，PET-4)行18F-FDG PET/CT扫描。使用五分法和最大标准摄取值减少率(%ΔSUVmax)法判读PET-2和PET-4。应用kappa检验评估五分法中3位阅片者判读结果的一致性，利用受试者操作特性(ROC)曲线计算%ΔSUVmax最佳阈值，构建%ΔSUVmax法、五分法的Kaplan-Meier生存曲线，使用Log-rank检验评估无进展生存(PFS)时间及整体生存(OS)时间。使用Cox回归模型进行预后相关因子多因素分析。
结果: 全部56例DLBCL患者中男28例、女28例，中位年龄55(20~80)岁，中位随访时间24 (6~42)个月。五分法(4分为界值)评估PET/CT图像，阅片者间一致性较好(κ>0.600)。ROC曲线计算出PET-2、PET-4的%ΔSUVmax最佳界值分别为81%、74%。Kaplan-Meier生存分析结果显示，PET-2、PET-4均能预测56例DLBCL患者的预后，PET阴性患者的3年PFS和OS率明显高于PET阳性患者，差异有统计学意义(P<0.05)。应用五分法判读PET-4，预测患者PFS准确性最高(76.79%)；66%ΔSUVmax法判读PET-2、PET-4预测OS准确性最好(76.79%、83.93%)。多因素分析结果显示，PET-2、PET-4均为优于国际预后指数评分的独立预后因子。
结论: PET-2和PET-4均能预测DLBCL患者预后。PET-4预测患者预后更准确，其中使用66%ΔSUVmax法预测3年OS率、五分法预测3年PFS率更准确。