CAJ | 학술논문

目的探讨18F-FDG PET/CT与电子内镜测量食管鳞状细胞癌(ESCC)长度的差异性,并确定PET显示ESCC长度的最适SUV阈值.方法前瞻性选择24例经组织病理学证实的ESCC患者入组(男19例,女5例;中位年龄59岁;TNM分期Ⅱ期3例,Ⅲ期14例,Ⅳ期7例).治疗前行18F-FDG PET/CT显像,SUV分别取2.0、2.5、3.0、3.5及35％、40％、45％、50％、55％SUVmax为阈值扣除本底,测量食管肿瘤长度,并分别标记为L2.0、L2.5、L3.0、L3.5、L35、L40、L45、L50和L55.患者PET/CT显像1周内行电子内镜检查,确定食管病变在体长度(Lst),并与PET/CT显示长度比较(配对t检验).相关性分析采用Pearson相关分析.结果原发肿瘤SUVmax为14.51±5.72,Lst为(5.27±2.45)cm.所测病变长度从大到小依次为L2.0、L2.5、Lst、L3.0、L3.5、L35、L40、L45、L50、L55.仅L2.5[(5.65±2.59) cm]和L3.0[(5.11±2.51) cm]与Lst比较差异无统计学意义(分别为t=-1.74和0.76,均P＞0.05).Pearson相关分析发现,L3.5、L2.5、L3.0与Lst的r值较高,分别为0.935、0.920和0.919,均P＜0.01.原发灶SUVmax＜15时,L2.5、L3.0、L35与Lst差异均无统计学意义[(4.95±2.76)、(4.45±2.50)、(4.42±1.85)和(4.82±2.14) cm;t =-0.439、1.299和2.011,均P＞0.05],且L2.5与Lst的相关系数最高(r=0.953,P＜0.05);SUVmax≥15时,仅L3.0与Lst差异无统计学意义[(6.11±2.61)和(5.67±2.64) cm;t=-0.897,P＞0.05],两者r值为0.791.结论对于原发灶SUVmax＜15的ESCC,确定原发肿瘤长度的最适阈值为SUV 2.5;对于原发灶SUVmax≥15者,SUV 3.0为阈值所测量的长度可能更接近电子内镜所测的肿瘤在体长度,但尚需扩大样本进一步验证.
목적 탐토18F-FDG PET/CT여전자내경측량식관린상세포암(ESCC)장도적차이성,병학정PET현시ESCC장도적최괄SUV역치.방법 전첨성선택24례경조직병이학증실적ESCC환자입조(남19례,녀5례;중위년령59세;TNM분기Ⅱ기3례,Ⅲ기14례,Ⅳ기7례).치료전행18F-FDG PET/CT현상,SUV분별취2.0、2.5、3.0、3.5급35％、40％、45％、50％、55％SUVmax위역치구제본저,측량식관종류장도,병분별표기위L2.0、L2.5、L3.0、L3.5、L35、L40、L45、L50화L55.환자PET/CT현상1주내행전자내경검사,학정식관병변재체장도(Lst),병여PET/CT현시장도비교(배대t검험).상관성분석채용Pearson상관분석.결과 원발종류SUVmax위14.51±5.72,Lst위(5.27±2.45)cm.소측병변장도종대도소의차위L2.0、L2.5、Lst、L3.0、L3.5、L35、L40、L45、L50、L55.부L2.5[(5.65±2.59) cm]화L3.0[(5.11±2.51) cm]여Lst비교차이무통계학의의(분별위t=-1.74화0.76,균P＞0.05).Pearson상관분석발현,L3.5、L2.5、L3.0여Lst적r치교고,분별위0.935、0.920화0.919,균P＜0.01.원발조SUVmax＜15시,L2.5、L3.0、L35여Lst차이균무통계학의의[(4.95±2.76)、(4.45±2.50)、(4.42±1.85)화(4.82±2.14) cm;t =-0.439、1.299화2.011,균P＞0.05],차L2.5여Lst적상관계수최고(r=0.953,P＜0.05);SUVmax≥15시,부L3.0여Lst차이무통계학의의[(6.11±2.61)화(5.67±2.64) cm;t=-0.897,P＞0.05],량자r치위0.791.결론 대우원발조SUVmax＜15적ESCC,학정원발종류장도적최괄역치위SUV 2.5;대우원발조SUVmax≥15자,SUV 3.0위역치소측량적장도가능경접근전자내경소측적종류재체장도,단상수확대양본진일보험증.
Objective To compare 18 F-FDG PET/CT and electronic endoscopy for measuring the length of esophageal squamous cell carcinoma (ESCC) and to evaluate the optimal SUV threshold for contour determination of the size of the lesion.Methods Twenty-four patients (19 males and 5 females,median age:59 years) with histologically confirmed ESCC were enrolled.Three patients had stage Ⅱ,14 had stage Ⅲ and 7 had stage Ⅳ diseases.PET studies were performed before treatment.The length of ESCC was measured on FDG PET imaging using different SUV thresholds of 2.0 (L2.0),2.5 (L2.5),3.0 (L3.0),3.5 (L3.5),and 35％(L35),40％(L40),45％(L45),50％(L50),55％(L55) of SUVmax.The length of ESCC on PET imaging was compared with the length of gross tumor in vivo measured by electronic endoscopy (Lst) to determine the optimal threshold of SUV using paired t test.Pearson correlation analysis was used to assess the correlation.Results The SUVmax of primary tumor was 14.51±5.72 and the Lst was (5.27± 2.45) cm.The length was in a descending order of L2.0,L2.5,Lst,L3.0,L3.5,L35,L40,L45,L50 and L55 when using different criteria.There were significant differences between the Lst and the lengths measured on PET except those by L2.5 and L3.0((5.65±2.69) cm,(5.11±2.51) cm; t=-1.74 and 0.76,both P＞0.05).The lengths measured on PET by all criteria were significandy correlated with the Lst,respectively,with the better r values by L3.5(0.935),L2.5(0.920) and L3.0(0.919) (all P＜0.01).When SUVm~＜15,there were no significant differences between the Lst ((4.82±2.14) cm) and L2.5((4.95±2.76) cm),L3.0((4.45±2.50) cm) and L35((4.42±1.85) cm),respectively (t=-0.439,1.299,2.011,all P＞0.05).The best correlation (r=0.953,P＜0.05) was between Lst and Lz5.When SUVmax ≥ 15,there was no significant difference between Lst ((5.67±2.64) cm) and L3.0((6.11±2.61) cm; t=-0.897,P＞0.05; r=0.791,P＜ 0.05).Conclusions For better correlation of ESCC lesion size,it is suggested that the optimal threshold of SUV for contouring is 2.5 for tumor SUVmax＜15,and 3.0 for tumor SUVmax ≥ 15.A larger sampling size is needed for further confirmation or modification.