放大胃镜联合窄带成像技术对不同分化程度早期胃癌的诊断效能

Diagnostic value of magnifying endoscopy combined with narrow-band imaging for early gastric cancer with various degree of differentiation

  • 摘要: 目的 探索放大内镜(magnifying endoscopy,ME)联合窄带成像技术(narrow-band imaging,NBI)对分化型和未分化型胃癌的诊断效能。 方法 采用ME联合NBI观察沧州市人民医院2014年8月- 2016年5月收治的67例早期胃癌患者胃黏膜表面微腺体(Microsurface,MS)结构和表面微血管(microvascular,MV)结构,并将MS分为4类:Ⅰ型椭圆形/管状,Ⅱ型乳头状,Ⅲ型部分消失,Ⅳ型消失;将MV分为3类:Ⅰ型线圈形,Ⅱ型细网格形,Ⅲ型锥形。采用WHO病理分型将早期胃癌分为分化型和未分化型,分析每一内镜下形态类别对分化型和未分化型胃癌的诊断效能。 结果 椭圆形和(或)管状的MS形态,线圈形和细网格形的MV形态更常见于分化型早期胃癌(P< 0.05);消失MS形态和锥形MV形态常见于未分化型早期胃癌(P< 0.05)。对诊断分化型胃癌,MSⅠ型椭圆形和(或)管状具有较高的敏感度(71.2%),而MVⅠ型线圈形和Ⅱ型细网格形诊断分化型胃癌特异度较高(均为93.3%)。3种类型对诊断分化型腺癌均显示较高的阳性预测值(88.1%,95.7%和96.6%)。Ⅲ型锥形诊断未分化型胃癌具有较高的精确度(94.0%)、敏感度(86.7%)、特异度(96.2%)、阳性预测值(86.7%)和阴性预测值(96.2%)。 结论 ME联合NBI下早期胃癌的形态分类更有利于判断其病理类型,尤其是MVⅢ型锥形诊断未分化型胃癌具有较高的精确度、敏感度、特异度、阳性预测值和阴性预测值。

     

    Abstract: Objective To analyze the diagnostic value of magnifying endoscopy (ME) combined with narrow-band imaging (NBI) for early gastric cancer (EGC) of various differentiation degree. Methods Microsurface (MS) pattern and microvascular (MV) pattern of 67 EGC patients admitted to Cangzhou People's Hospital from August 2014 to May 2016 were observed by ME combined with NBI. MS patterns were classified into 4 categories (oval/tubular, papillary, destructive and absent) and MV into 3 categories (loop, fine network and corkscrew). According to WHO criteria, EGC patients were categorized into differentiated-type (DT) and undifferentiated-type (UDT), and the diagnostic value of endoscopic classification on predicting DT and UDT EGC was analyzed. Results MS pattern of oval/tubular and MV pattern of loop and fine network mainly showed in EGC patients with DT (P< 0.05), whereas absense of MS pattern and corkscrew pattern were more common in UDT EGC patients (P< 0.05). For the diagnosis of DT EGC, oval/tubular MS pattern had high sensitivity (71.2%), and loop and fine network in MV pattern had high specificity (both 93.3%). The three types mentioned above showed relatively high positive predictive value (88.1%, 95.7% and 96.6%). Besides, a corkscrew pattern for the diagnosis of UDT tumors had high accuracy (94.0%), sensitivity (86.7%), specificity (96.2%), positive predictive value (86.7%) and negative predictive value (96.2%). Conclusion A combinaton of ME and NBI shows advantages in predicting histological type of EGC, especially for the MV pattern of corkscrew, which has high accuracy, sensitivity, specificity, positive predictive value and negative predictive value.

     

/

返回文章
返回