直肠癌术后病理标本与术前MRI/CT定义肿瘤范围的对比研究

摘要

Objective:This study aimed to compare rectal cancer tumor volume parameters measured by MRI sequences (T1WI, T2WI, and DWI) and/or CT with those by pathological specimen. Methods:Twenty-two patients with rectal cancer were prospectively enrolled. MRI sequences including T1WI, T2WI, and DWI, and/or CT of the pelvis were performed before operation. Volume parameters, such as tumor length along the rectal axis, maximum tumor width perpendicular to rectal axis, and tumor actual area in that perpendicular plane, were measured on T1WI, T2WI, DWI, and CT, respectively, for each patient. The respective pathological parameters were further measured in surgical specimen after total mesorectal excision. The two kinds of parameter values measured in imaging and pathology were statistically compared and accuracy appraisal was performed. Results:The mean Lpath-L was 4.06±1.14 cm. The mean LT1-L, LT2-L, LDWI-L, and LCT-L were 3.91± 1.51, 4.62±1.41, 3.39±1.05, and 3.94±1.23 cm, respectively. Correlation coefficients were 0.688, 0.635, 0.688, and 0.720 (P<0.05). An average 6 mm overestimation was found in T2WI, and 1 to 6 mm underestimation in T1WI, DWI, and CT in length values compared with those measured in surgical specimen. The mean Lpath-W was 2.56 ±0.94 cm. The mean LT1-W, LT2-W, LDWI-W, and LCT-W were 3.62±0.99, 3.66±0.76, 3.23±0.58, and 3.64±1.04 cm, respectively. The magnitude of mean overestimation ranged from 5.1 to 11.1 mm. The Apath was 4.30 ±2.83 cm2. AT1, AT2, ADWI, and ACT were 8.98±3.90, 8.99±3.43, 8.41±3.09, and 9.63±4.40 cm2, respectively, which double overestimated the tumor area in the perpendicular rectal plane. Conclusion:The difference in longitudinal length between MRI sequences/CT and pathological specimen was in the range of?6 mm to 6 mm. The mean maximum tumor width and areas in the maximum tumor perpendicular plane were overestimated. This study indicated that gross tumor volume delineation based on CT or MRI for rectal cancer irradiation should be conservative in the axial images of rectum, and meticulous consideration is required along the rectum.%目的:比较直肠癌术后病理标本与术前MRI(T1WI、T2WI、DWI)和/或CT显示的肿瘤范围的差异.方法:选取2016年3月至2016年5月间于湖南省肿瘤医院拟行全直肠系膜切除术(TME)的直肠癌患者22例,术前行盆腔MRI(T1WI、T2WI、DWI)和/或增强CT检查.分别测量肿瘤沿肠管纵轴长度、垂直肠管横轴位肿瘤最宽径和横轴位肿瘤实际面积,并与术后病理标本测量对应参数比较,评估各影像测量的精确性.结果:病理长度(Lpath-L)为(4.06±1.14)cm,LT1-L、LT2-L、LDWI-L、LCT-L分别为(3.91±1.51)、(4.62±1.41)、(3.39±1.05)、(3.94±1.23)cm,与Lpath-L的相关系数分别为0.688、0.635、0.688、0.720(P<0.05);T2WI测量结果存在平均6 mm高估,T1WI、DWI、CT测量结果存在1~6 mm不同程度的低估.病理横截面肿瘤最宽径(Lpath-W)为(2.56±0.94)cm,LT1-W、LT2-W、LDWI-W、LCT-W分别为(3.62±0.99)、(3.66±0.76)、(3.23±0.58)、(3.64±1.04)cm,测量结果存在平均5.1~11.1 mm的高估.肿瘤病理实际面积(Apath)为(4.30±2.83)cm2,AT1、AT2、ADWI、ACT分别为(8.98±3.90)、(8.99±3.43)、(8.41±3.09)、(9.63±4.40)cm2,各影像测量实际面积存在约2倍程度高估.结论:在病变长度方面,各影像存在-6~6 mm差异;最大横截面方面均存在不同程度高估.因此在直肠癌放疗GTV勾画时,断面侧方应适当保守内收,而上下端也不应过多延伸.