18氟-氟代脱氧葡萄糖PET／CT检查判断结直肠癌术后预后的临床价值

Clinical value of Fluorine-18-fluorodeoxyglucose PET/CT examination to predict the prognosis of patients after colorectal cancer operation

目的：探讨 18氟氟代脱氧葡萄糖（ 18F-FDG）PET/CT检查判断结直肠癌术后预后的临床价值。 方法：采用回顾性横断面研究方法。收集2007年3月至2015年10月南京医科大学第一附属医院收治的80例结直肠癌患者的临床病理资料。80例患者术前均行 18F-FDG PET/CT检查。完善术前相关检查后，遵循患者及家属意愿行手术治疗，依据患者具体情况施行术后辅助化疗。观察指标：（1）术前影像学检查结果。（2）治疗和随访情况。（3）预后因素分析。采用门诊及电话方式进行随访，术后第1年每3个月1次，第2年每半年1次，以后每年1次。随访内容包括肿瘤复发、进展和患者生存情况。无瘤生存时间为术后第1天至首次发现肿瘤复发、进展，患者死亡或随访截止。总生存时间为术后第1天至患者死亡或随访截止。随访时间截至2016年5月。偏态分布的计量资料以M（Qn）和M（范围）表示。以无瘤生存作为阳性事件，采用受试者工作特征（ROC）曲线获得最大标准摄取值（SUV max）、平均标准摄取值（SUV mean）、 代谢体积（MTV）及病灶糖酵解总量（TLG）的最佳界值点，并计算曲线下面积（AUC）值。如果ROC曲线下面积较小，则以中位数为界值点。采用KaplanMeier法进行生存分析。单因素分析采用Logrank 检验，多因素分析采用COX比例风险模型。 结果：（1）术前影像学检查结果： 结直肠癌患者PET/CT检查均表现为肠壁局限性增厚，肠腔狭窄，部分病灶周围脂肪间隙模糊，可见肿大淋巴结， 18F-FDG代谢异常增高。 80例患者结直肠癌原发灶SUV max、SUV mean、MTV、TLG分别为11.83（4.26，35.42）、7.06（2.38，20.92）、20.47 cm3（1.29 cm3，161.50 cm3）、138.58（14.17，857.89）。ROC曲线分析显示：SUV max的AUC=0.453［95%可信区间（CI）：0.307~0.600，P〉0.05］；SUV mean 的AUC=0.448（95%CI：0.303~0.594，P〉0.05）；MTV的AUC=0.815（95%CI：0.717~0.913，P〈0.05）；TLG的AUC=0.749（95%CI：0.635~0.863，P〈0.05）。由于SUV max、SUV mean的AUC较小，不能通过ROC曲线分析获得界值点，因此，以其中位数SUV max =11.83、SUV mean=7.06作为分组的界值点。MTV界值点为18.79 cm3（灵敏度为86.2%，特异度为68.3%）；TLG界值点为142.05（灵敏度为75.9%，特异度为70.7%）。（2）治疗和随访情况：80例患者中，13例发生肝脏转移行结直肠癌根治联合肝转移癌切除术，67例行单纯结直肠癌根治术。62例患者行术后化疗，其中45例患者的化疗方案以卡培他滨为主，17例化疗方案以氟尿嘧啶为主。80例患者均获得随访，中位随访时间为41.8个月（6.5～109.1个月）。随访期间，29例患者肿瘤复发、进展，19例患者死亡。80例患者中位无瘤生存时间为19.5个月（2.0～109.1个月），1、3、5年无瘤生存率分别为73.7%、36.3%、18.8%；中位总生存时间为31.8个月（3.3～109.1个月），1、3、5年总生存率分别为76.3%、37.5%、20.0%。（3）预后因素分析：单因素分析结果显示：肿瘤部位、N分期、M分期、临床分期、术后化疗、MTV及TLG是结直肠癌患者术后无瘤生存率的影响因素（HR=3.469，5.325，5.295，8.605，2.630，7.388，5.155，95%CI：1.522~7.906，2.256~12.568，2.405~11.657，2.969~24.937，1.063~6.504，2.550~21.403，2.178~12.204，P〈0.05）。肿瘤部位、分化程度、N分期、M分期、临床分期、MTV及TLG是结直肠癌患者术后总生存率的影响因素（HR=2.697，2.814，3.083，2.916，4.193，5.450，4.876，95%CI：1.011~7.197，1.121~7.062，1.166~8.149，1.140~7.454，1.386~12.678，1.581~18.786，1.727~13.766，P〈0.05）。多因素分析结果显示：临床分期为Ⅲ~Ⅳ期及TLG≥142.05是影响结直肠癌患者术后无瘤生存率的独立危险因素（HR=9.879，3.569，95%CI：1.854~22.836，1.127~11.306，P〈0.05）。M1期、临床分期为Ⅲ~Ⅳ期及TLG≥142.05是影响结直肠癌患者术后总生存率的独立危险因素（HR=4.522，9.315，10.120，95%CI：1.223~16.717，1.338~24.864，2.385~12.947，P〈0.05）。 结论： 18F-FDG PET/CT检查指标TLG是影响结直肠癌患者术后无瘤生存率及总生存率的独立预后因子，对结直肠癌患者术后预后判断具有一定的参考价值。

Objective：To investigate the clinical value of Fluorine18fluorodeoxyglucose （ 18F-FDG） PET/CT examination to predict the prognosis of patients after colorectal cancer operation. Methods：The retrospective crosssectional study was adopted. The clinicopathological data of 80 patients with colorectal cancer who were admitted to the First Hospital of Nanjing Medical University from March 2007 to October 2015 were collected. Eighty patients received first preoperative 18F-FDG PET/CT examination and underwent operations under decisions of patients and their families, and then adjuvant chemotherapy were performed according to the patients′ condition. Observation indicators included： （1） preoperative imaging examination, （2） situations of treatment and followup, （3） analysis of prognostic factors. The patients were followed up by outpatient examination and telephone interview once every 3 months within postoperative 1 year, once every half a year within postoperative 2 years and then once a year up to May 2016. The followup included tumor recurrence or progression and survival of patients. Tumorfree survival time was from postoperative day 1 to tumor recurrence or progression and death or end of followup. Overall survival time was from postoperative day 1 to death or end of followup. Measurement data with skewed distribution were represented as M （Qn） and M （range）. The optimal cutoff point of tumorfree survival of maximum standardized uptake value （SUV max）, mean standardized uptake value （SUV mean）, metabolism of volume （MTV） and total lesion of glycolysis （TLG） were investigated using the ROC curve analysis, and calculating area under the curve （AUC）. The median was used as a cutoff point if there was smaller AUC. The KaplanMeier method and Logrank test were respectively used for survival analysis and univariate analysis, and COX proportional hazards model for multivariate analysis. Results：（1） Results of preoperative imaging examination： results of PET/CT in patients with colorectal cancer showed there were circumscribed thickening of bowel wall, intestinal cavity strictures, fuzzy fat space around the some lesions, enlarged lymph node and 18F-FDG uptake increased abnormally. The SUV max, SUV mean, MTV and TLG of 80 patients were 11.83（4.26, 35.42）, 7.06（2.38, 20.92）, 20.47 cm3（1.29 cm3, 161.50 cm3） and 138.58（14.17, 857.89）, respectively. ROC curve showed that the AUC of SUV max , SUV mean, MTV and TLG were 0.453, 0.448, 0.815 and 0.749 [95% confidence interval （CI）： 0.307-0.600, P〉0.05; 0.303-0.594, P〉0.05; 0.717- 0.913, P〈0.05; 0.635- 0.863, P〈0.05], respectively. The median SUV max （11.83） and SUV mean（7.06） were used as the cutoff points due to smaller AUC of SUV max and SUV mean. The cutoff point of MTV was 18.79 cm3 （sensitivity=86.2% and specificity=68.3%）, and the cutoff point of TLG was 142.05 （sensitivity=75.9% and specificity=70.7%）. （2） Situations of treatment and followup： among 80 patients, 13 underwent the radical resection of colorectal cancer and resection of liver metastasis and 67 underwent radical resection of colorectal cancer. Sixtytwo patients received postoperative chemotherapy, including 45 with chemotherapy regimens of capecitabine and 17 with fluorouracil. Eighty patients were followed up for 41.8 months （range, 6.5-109.1 months ）. During the followup, 29 patients had tumor recurrence or progression, and 19 patients were dead. The median tumorfree survival time, 1, 3 and 5year tumorfree survival rates in 80 patients were 19.5 months （range, 2.0-109.1 months）, 73.7%, 36.3% and 18.8%, respectively. The median overall survival time, 1, 3 and 5year overall survival rates were 31.8 months （range, 3.3-109.1 months）, 76.3%, 37.5% and 20.0%, respectively. （3） Analysis of prognostic factors： results of univariate analysis showed that tumor location, N staging, M staging, clinical staging, postoperative chemotherapy, MTV and TLG were relative factors affecting postoperative tumorfree survival rate （HR=3.469, 5.325, 5.295, 8.605, 2.630, 7.388, 5.155, 95% CI： 1.522-7.906, 2.256-12.568, 2.405-11.657, 2.969-24.937, 1.063-6.504, 2.550-21.403, 2.178-12.204, P〈0.05）. The tumor location, tumor differentiation, N staging, M staging, clinical staging, MTV and TLG were relative factors affecting postoperative overall survival rate （HR=2.697, 2.814, 3.083, 2.916, 4.193, 5.450, 4.876, 95% CI： 1.011-7.197, 1.121-7.062, 1.166-8.149, 1.140-7.454, 1.386-12.678, 1.581-18.786, 1.727-13.766, P〈0.05）. In multivariate analysis, stage Ⅲ-Ⅳ of clinical staging and TLG≥142.05 were independent risk factors affecting postoperative tumorfree survival rate （HR=9.879, 3.569, 95% CI： 1.854-22.836, 1.127-11.306, P〈0.05）. The stage M1, stage Ⅲ-Ⅳ of clinical staging and TLG≥142.05 were independent risk factors affecting postoperative overall survival rate （HR=4.522, 9.315, 10.120, 95% CI： 1.223-16.717, 1.338-24.864, 2.385-12.947, P〈0.05）. Conclusion：TLG through 18F-FDG PET/CT examination is an independent prognostic factor affecting postoperative tumorfree survival rate and overall survival rate in patients with colorectal cancer after curative resection, and it has certainly reference value for prognosis.