Charles-Edwards Elizabeth M, Messiou Christina, Morgan Veronica A, De Silva Sonali S, McWhinney Norman A, Katesmark Mike, Attygalle Ayoma D, DeSouza Nandita M
Cancer Research UK Clinical Magnetic Resonance Research Group, Institute of Cancer Research and Royal Marsden NHS Foundation Trust, Surrey, England.
Radiology. 2008 Nov;249(2):541-50. doi: 10.1148/radiol.2491072165.
To establish apparent diffusion coefficients (ADCs) of invasive cervical carcinoma compared with nontumor cervical epithelium and determine sensitivity and specificity of diffusion-weighted (DW) magnetic resonance (MR) imaging used in conjunction with T2-weighted MR imaging to help detect invasive cervical carcinoma in patients with stage Ia and Ib1 disease.
Local research ethics committee approval was obtained with written consent from each subject. Group 1 comprised patients (mean age, 38.7 years +/- 13.2 [standard deviation]) with histologically confirmed cervical intraepithelial neoplasia (CIN) found on smear (n = 20) or stage Ib1 cervical tumors (n = 18). Patients were imaged with endovaginal T2-weighted fast spin-echo and single-shot DW echo-planar MR imaging of the cervix. ADCs from invasive cervical carcinoma and nontumor regions were compared within (t test) and between (U test) patients. A derived threshold ADC level indicative of invasive cervical carcinoma was used with T2-weighted imaging by two independent observers to identify possible invasive cervical carcinoma in group 2, patients with suspected disease (n = 21; mean age, 42.0 years +/- 16.4). Surgical specimens were the reference standard. Interobserver agreement was assessed.
In group 1, ADCs from cervical carcinoma (757 x 10(-6) mm(2)/sec +/- 110) and adjacent epithelium (1331 x 10(-6) mm(2)/sec +/- 159) or CIN (1291 x 10(-6) mm(2)/sec +/- 156) were significantly different (P < .0001). In group 2, respective sensitivity and specificity to help detect invasive cervical carcinoma on T2-weighted images were 55.6% and 75% for observer 1 and 66.7% and 41.7% for observer 2, and 88.9% and 66.7% for observer 1 and 77.8% and 58.3% for observer 2 when ADC maps with a threshold level of 1100 x 10(-6) mm(2)/sec were added. Interobserver agreement was fair (kappa = 0.37) for T2-weighted images alone and good (kappa = 0.80) with ADC included.
ADCs from invasive cervical carcinoma are significantly lower than those from nontumor epithelium; good interobserver agreement by using T2-weighted and DW MR imaging makes this technique potentially useful to help detect early-stage disease.
建立浸润性宫颈癌与非肿瘤性宫颈上皮的表观扩散系数（ADC），并确定扩散加权（DW）磁共振（MR）成像与T2加权MR成像联合使用时，在检测Ia期和Ib1期疾病患者浸润性宫颈癌中的敏感性和特异性。
获得当地研究伦理委员会批准，并取得每位受试者的书面同意。第1组包括经组织学证实涂片发现宫颈上皮内瘤变（CIN）（n = 20）或Ib1期宫颈肿瘤（n = 18）的患者（平均年龄38.7岁±13.2[标准差]）。患者接受宫颈的阴道内T2加权快速自旋回波和单次激发DW回波平面MR成像。在患者内部（t检验）和患者之间（U检验）比较浸润性宫颈癌和非肿瘤区域的ADC。由两位独立观察者将一个表示浸润性宫颈癌的推导阈值ADC水平与T2加权成像一起用于识别第2组疑似疾病患者（n = 21；平均年龄42.0岁±16.4）中可能的浸润性宫颈癌。手术标本为参考标准。评估观察者间的一致性。
在第1组中，宫颈癌的ADC（757×10⁻⁶mm²/秒±110）与相邻上皮（1331×10⁻⁶mm²/秒±159）或CIN（1291×10⁻⁶mm²/秒±156）有显著差异（P <.0001）。在第2组中，观察者1在T2加权图像上检测浸润性宫颈癌的敏感性和特异性分别为55.6%和75%，观察者2分别为66.7%和41.7%；当加入阈值水平为1100×10⁻⁶mm²/秒的ADC图时，观察者1的敏感性和特异性分别为88.9%和66.7%，观察者2分别为77.8%和58.3%。仅T2加权图像时观察者间一致性一般（kappa = 0.37），加入ADC后良好（kappa = 0.80）。
浸润性宫颈癌的ADC显著低于非肿瘤上皮的ADC；使用T2加权和DW MR成像时观察者间一致性良好，使该技术可能有助于检测早期疾病。
