From the Departments of *Surgery, and †Diagnostic Radiology, Keio University School of Medicine, Tokyo, Japan.
Clin Nucl Med. 2016 Oct;41(10):761-5. doi: 10.1097/RLU.0000000000001332.
PET/CT plays an important role in cancer diagnosis. Recently, novel metabolic parameters in PET/CT such as metabolic tumor volume (MTV) and total lesion glycolysis (TLG) have been reported to be diagnostic and prognostic biomarkers of various cancers. We evaluated the diagnostic value of these metabolic parameters in colorectal cancer (CRC).
The study included 138 patients who underwent surgical resection of CRCs between August 2012 and March 2014. The MTVs and TLGs of tumors were measured using various SUV thresholds. The diagnostic abilities of the metabolic parameters were analyzed using ROC curves and classification and regression trees.
The AUCs of the MTVs and TLGs for predicting T stage (0.881-0.892) were significantly higher than the AUC of the SUVmax (0.824). In the M stage, the AUCs of MTVs and TLGs (0.688-0.723) were significantly higher than that of the SUVmax (0.606). Recursive partitioning applying classification and regression trees demonstrated that the optimal cutoff values of the most important variables for discriminating T, N, and M stages are MTV2.5 = 9.35 and 63.33 mL, TLG50% = 328.1, and TLG50% = 94.81, respectively.
Metabolic tumor volumes and TLGs in PET/CT are reliable diagnostic biomarkers. Using these parameters, more accurate preoperative diagnoses for CRC can be made.
PET/CT 在癌症诊断中发挥着重要作用。最近，新型代谢参数如代谢肿瘤体积（MTV）和总病变糖酵解（TLG）已被报道为各种癌症的诊断和预后生物标志物。我们评估了这些代谢参数在结直肠癌（CRC）中的诊断价值。
该研究纳入了 138 例于 2012 年 8 月至 2014 年 3 月期间接受 CRC 手术切除的患者。使用不同的 SUV 阈值测量肿瘤的 MTVs 和 TLGs。使用 ROC 曲线和分类回归树分析代谢参数的诊断能力。
MTVs 和 TLGs 预测 T 分期的 AUC（0.881-0.892）明显高于 SUVmax 的 AUC（0.824）。在 M 分期中，MTVs 和 TLGs 的 AUC（0.688-0.723）明显高于 SUVmax 的 AUC（0.606）。应用分类回归树的递归分割显示，区分 T、N 和 M 分期的最重要变量的最佳截断值分别为 MTV2.5=9.35 和 63.33mL、TLG50%=328.1 和 TLG50%=94.81。
PET/CT 中的代谢肿瘤体积和 TLG 是可靠的诊断生物标志物。使用这些参数，可以对 CRC 进行更准确的术前诊断。
