| Objective:The purpose of this study was to evaluate the diagnostic and prognostic value of18F-FDG PET/CT for nasopharyngeal carcinoma (NPC) patients follow-up after comprehensive therapy.Methods:1. A total of89NPC patients after comprehensive therapy were included in this retrospective study, and all the patients underwent18F-FDG PET/CT scan and formally follow-up. The diagnostic results of F-FDG PET/CT and traditional imaging were analyzed and compared with histopathological diagnosis or clinical follow-up data. The diagnostic sensitivity, specificity, accuracy, positive predictive value (PPV) and negative predictive value (NPV) of residual, recurrent or metastasis were calculated, respectively. The quality diagnosis and quantity diagnosis using the maximal standard uptake value (SUVmax) were determined by two experienced nuclear physicians and radiologists. Receiver operator curves (ROC) were determined to assess the area under curve (AUC) and the optimal cutoff value for evaluating diagnostic value of PET/CT and predicting survival. Kaplan-Meier method and log-rank test were used to evaluate and compare survival rate. Multivariate analysis was performed to identify the prognostic factors which influencing Overall survival (OS) and Disease-free survival (DFS) using Cox proportional hazards. Relative risk (RR) was presented with their95%confidence intervals (Wald x2test). All statistical tests were conducted at a two-sided level of significance of0.05. 2. A total of221NPC patients after comprehensive therapy were included in this retrospective study, and all the patients underwent18F-FDG PET/CT scan and formally follow-up. The diagnostic results of F-FDG PET/CT was analyzed and compared with histopathological diagnosis or clinical follow-up data. For each patient, every lesion was segmented by four SUV thresholds with3D-area growing algorithm, SUV=2.5,20%,40%and SUV background method. The detection rates were compared. The optimal SUV threshold method was then used to calculate whole body metabolic tumor volume (WBMTV) and whole body total lesion glycolysis (WBTLG). Receiver operator curves (ROC) were determined to assess the area under curve (AUC) and the optimal cutoff value for WBMTV and WBTLG. Kaplan-Meier method and log-rank test were used to evaluate and compare survival rate. The prognostic significance of SUVmax, WBMTV, WBTLG and other pathological variables for overall survival and disease free survival were assessed by Cox proportional hazards regression analysis and Kaplan-Meier survival analysis.3. A total of48NPC patients were included in this retrospective study, and all the patients underwent18F-FDG PET/CT scan and formally follow-up. Each visible lesion delineated by SUV background method, using3D-area growing algorithms. Whole body metabolic tumor volume(WBMTV) and whole body total lesion glycolysis (WBTLG) were calculated. The differences of SUVmax, WBMTV and WBTLG among pathologic subtypes were tested using one-way analysis of variance (ANOVA) or Kruskal-Wallis test. Two-tailed Spearman’s correlationwas used to analyse the relationships between themetabolic parameters and the TNM staging system. Receiver operating characteristic curves were used to determine the optimal WBMTV and WBTLG threshold predictive of death. The prognostic significance of SUVmax, WBMTV, WBTLG and other pathological variables for overall survival and progression survival were assessed by Cox proportional hazards regression analysis and Kaplan-Meier survival analysis. Results:1. In all89patients, recurrence and metastasis were confirmed in36patients,41patients in DFS and12died at the last follow-up. Five-year overall survival rate and mean survival time of all patients were73.00%and69.2±4.5mo. The diagnostic sensitivity, specificity, accuracy, PPV and NPV of PET/CT and conventional imaging were100.0%,90.2%,95.5%,92.2%and100.0%VS75.9%,78.9%,77.3%,80.0%and72.7%, respectively. The sensitivity and accuracy had significant difference. From ROC, the cut off value of SUVmax was2.5, which may be the best diagnostic value in the follow-up of NPC. Patients with SUVmax below2.5had significantly better5-year OS (81.9%vs62.0%, P=0.036) and Mean survival time (75.8±5.8mo vs59.8±5.8mo) than patients with an SUVmax of2.5or greater. PET/CT negative patients had significantly better5-year OS (59.9%vs100.0%, P=0.006) than18F-FDG PET/CT positive ones. Cox proportional hazards regression model showed both SUVmax (RR,1.205; P=0.014) and number of lesions (RR,2.734; P=0.005) could be used to predict OS. For DFS, number of lesions (RR,2.105; P=0.008) may have predictive relevance.2. In all221patients, recurrence and metastasis were confirmed in28patients,156patients in DFS and37died at the last follow-up. Five-year overall survival rate and mean survival time of all patients were71.6%and90.9±4.6mo. The221patients had a total of576lesions, The detection rates of the threshold SUV=2.5,20%,40%and SUV background method were65.6%(378/576)、80.2%(462/576).71.5%(412/576)、and90.4%(521/576) respectively. There were significant differences between the four methods (P<0.005). From ROC, the cut off value of WBMTV and WBTLG was8.10,35.58. Patients with WBMTV below8.10had significantly better5-year OS (90.7%vs51.2%, P=0.000) than patients with a WBMTV of8.10or greater. Patients with WBTLG below35.58had significantly better5-year OS (91.7%vs50.4%, P=0.000) than patients with a WBTLG of35.58or greater. Multivariate Cox regression modeling showed both WBTLG (RR,1.002; P=0.004) and age (RR,1.046; P=0.006) could be used to predict OS. For DFS, WBTLG (RR,1.003; P=0.000) may have predictive relevance.3. In all48patients, recurrence and metastasis were confirmed in9patients,36patients in PFS and3died at the last follow-up. Mean survival time of all patients were66.9+2.3mo. The differences of SUVmax, WBMTV(F=2.355, P=0.137) and WBTLG (F=3.106, P=0.082) among pathologic subtypes are all not significantly in this study. Positive correlations were observed between SUVmax, WBMTV and WBTLG and T-stage (P<0.05). From ROC, the cut off value of WBMTV and WBTLG was28.0,155.5. Patients with WBMTV below28.0had significantly better5-year OS than patients with a WBMTV of28.0or greater (P=0.011). Patients with WBTLG below155.5had significantly better5-year OS than patients with a WBTLG of155.5or greater (P=0.002). Multivariate Cox regression modeling showed both WBMTV and WBTLG could be used to predict OS (RR:1.050VS2.734; P=0.008VS0.005) and PFS(RR:1.030VS2.234; P=0.005VS0.006).Conclusions:1. F-FDG PET/CT imaging had significantly higher sensitivity and accuracy than conventional imaging in detection of the residual, recurrent and metastasis of NPC after comprehensive therapy. SUVmax2.5may be a relatively good diagnostic value in the follow-up study, and it may also have great valuable for predicting long-term survival. SUVmax and number of lesions may be predictors for OS and DFS, and number of lesions may be an important factor for DFS.2. The detection rate of the SUV threshold of background method was significantly higher than other methods in detecting metastatic lesions. WBTLG could be used as an independent prognostic factor for with NPC after comprehensive therapy. 3. WBMTV and WBTLG are valuable for predicting long-term survival of NPC. High WBMTV and WBTLG may be useful for identifying patients requiring more aggressive treatment. |
PDF Full Text Request