Purpose The aim of the present study was to test whether the coefficient of variation (CoV) of18F-FDG PET/CT images of metastatic lymph nodes and primary tumors may predict clinical outcome in patients with advanced non-small cell lung cancer (NSCLC). Materials and Methods Fifty-eight NSCLC patients who had undergone18F-FDG PET/CT at diagnosis were evaluated. SUVmax, SUVmean, CoV, MTV and TLG were determined in targeted lymph nodes and corresponding primary tumors along with Total MTV (MTVTOT) and Whole-Body TLG (TLGWB) of all malignant lesions. Univariate analysis was performed using Cox proportional hazards regression whereas the Kaplan–Meier method and log-rank tests were used for survival analysis. Results Fifty-eight metastatic lymph nodes were analyzed and average values of SUVmax, SUVmean, CoV, MTV and TLG were 11.89 ± 8.54, 4.85 ± 1.90, 0.37 ± 0.16, 46.16 ± 99.59 mL and 256.84 ± 548.27 g, respectively, whereas in primary tumors they were 11.92 ± 6.21, 5.47 ± 2.34, 0.36 ± 0.14, 48.03 ± 64.45 mL and 285.21 ± 397.95 g, respectively. At univariate analysis, overall survival (OS) was predicted by SUVmax (p= 0.0363), SUVmean (p= 0.0200) and CoV (p= 0.0139) of targeted lymph nodes as well as by CoV of primary tumors (p= 0.0173), MTVTOT(p= 0.0007), TLGWB(p= 0.0129) and stage (p= 0.0122). Using Kaplan–Meier analysis, OS was significantly better in patients with CoV of targeted lymph nodes ≤ 0.29 than those with CoV > 0.29 (p= 0.0147), meanwhile patients with CoV of primary tumors > 0.38 had a better prognosis compared to those with CoV ≤ 0.38 (p= 0.0137). Finally, we combined the CoV values of targeted lymph nodes and primary tumors in all possible arrangements and a statistically significant difference was found among the four survival curves (p= 0.0133). In particular, patients with CoV of targeted lymph nodes ≤ 0.29 and CoV of primary tumors > 0.38 had the best prognosis. Conclusions The CoV of targeted lymph nodes combined with the CoV of primary tumors can predict prognosis of NSCLC patients.
目的 本研究旨在探讨¹⁸F-FDG PET/CT图像中转移性淋巴结与原发肿瘤的变异系数（CoV）能否预测晚期非小细胞肺癌（NSCLC）患者的临床结局。材料与方法 对58例确诊时接受过¹⁸F-FDG PET/CT检查的NSCLC患者进行评估。测定目标淋巴结及对应原发肿瘤的最大标准化摄取值（SUVmax）、平均标准化摄取值（SUVmean）、变异系数（CoV）、代谢肿瘤体积（MTV）和肿瘤糖酵解总量（TLG），同时计算所有恶性病灶的总代谢肿瘤体积（MTVTOT）和全身肿瘤糖酵解总量（TLGWB）。采用Cox比例风险回归模型进行单因素分析，并运用Kaplan-Meier法和log-rank检验进行生存分析。结果 共分析58个转移淋巴结，其SUVmax、SUVmean、CoV、MTV和TLG的平均值分别为11.89±8.54、4.85±1.90、0.37±0.16、46.16±99.59 mL和256.84±548.27 g；原发肿瘤的对应值分别为11.92±6.21、5.47±2.34、0.36±0.14、48.03±64.45 mL和285.21±397.95 g。单因素分析显示，总生存期（OS）与目标淋巴结的SUVmax（p=0.0363）、SUVmean（p=0.0200）和CoV（p=0.0139），原发肿瘤的CoV（p=0.0173），MTVTOT（p=0.0007），TLGWB（p=0.0129）及临床分期（p=0.0122）显著相关。Kaplan-Meier分析表明，目标淋巴结CoV≤0.29的患者OS显著优于CoV>0.29者（p=0.0147）；而原发肿瘤CoV>0.38的患者预后较CoV≤0.38者更好（p=0.0137）。最后，我们将目标淋巴结与原发肿瘤的CoV值进行全排列组合分析，发现四条生存曲线存在统计学显著差异（p=0.0133），其中目标淋巴结CoV≤0.29且原发肿瘤CoV>0.38的患者预后最佳。结论 目标淋巴结CoV与原发肿瘤CoV的联合检测可预测NSCLC患者的预后。
肿瘤(癌症)患者之家