Accuracy of Imaging in Preoperative Local Staging of Endometrial Cancer: Could Imaging Predict Low Risk Patients?

Endometrial cancer is the most common gynecological cancer in women and is considered the sixth most common malignancy in women (1,2). The incidence of endometrial carcinoma in Iran accounts for 17.6% of female genital tract carcinomas based on officially published female cancer statistics (3). Surgical staging is the current way of endometrial cancer treatment; it includes total abdominal hysterectomy and bilateral salpingo-oophorectomy. Nowadays, there are controversies about routine lymphadenectomy. Some studies suggest less extensive surgery for lowrisk patients, reserving para aortic lymphadenectomy for high-risk patients (4-6). Several studies showed only 5% positive lymph node metastasis with superficial myometrial invasion (1,7-9). Extra-uterine disease and lymph node metastasis cannot be evaluated via presurgical clinical examination; on the other hand, these are related to the depth of myometrial invasion, the most important and independent prognostic factor, cervical involvement and histopathological type and grade (1,10,11). Preoperative imaging assessment of myometrial invasion may help predict probable lymphadenopathies, especially in low grade tumors, and provide guidance for a better surgical approach in avoiding extensive surgery in low-risk patients with low-grade histopathology and myometrial invasion of less than 50%. Magnetic resonance imaging (MRI) and transvaginal sonography (TVS) are widely accepted as imaging choices in preoperative local staging of endometrial cancer (12) and accuracy of TVS in some recent studies was comparable with that of MRI, though there are still challenges (8,13,14). The use of Doppler sonography in the evaluation of endometrial cancer is a controversial issue (15). In this prospective study, we evaluated the diagnostic accuracy of MRI, TVS and Doppler for detecting the Abstract Objectives: In this study we aimed to evaluate diagnostic accuracy of magnetic resonance imaging (MRI) and transvaginal sonography (TVS) in identifying the depth of myometrial invasion and cervical involvement and also their relationship with the uterine arteries resistance index (RI) and pulsatility index (PI) in endometrial carcinoma. Materials and Methods: We performed a prospective study on 45 women with histologically confirmed diagnosis of endometrial carcinoma. The study was performed from October 2009 to December 2012. All the patients were evaluated by 3T MRI and TVS and transvaginal color Doppler sonography of uterine arteries. All the patients underwent hysterectomy and the result of imaging and pathologic studies were compared. Results: Mean age was 54.5 ± 10.8 years (33 to 77 years), mean gravid was 3.93 ± 2.8 and the mean parity was 3.9 ± 2.5. The pathology results indicated 29 patients (64.4%) in stage IA and 16 patients (35.6%) in stage IB. Mean endometrial thickness in stage IA patients was 18.4±14.4 mm and in stage IB patients was 38.5±11.5 mm. TVS also showed positive predictive value (PPV) of 76.5%, negative predictive value (NPV) of 88.9%, sensitivity of 81.3%, specificity of 85.7%, and accuracy of 84.1% for assessment of the depth of myometrial invasion in endometrial carcinoma. We found PPV of 86.7%, NPV of 92.3%, sensitivity of 86.7%, specificity of 92.3%, and accuracy of 90.2% for MRI study. Conclusions: TVS can evaluate the depth of myometrial invasion with an acceptable accuracy when MRI is not available or costeffective, or when MRI is contraindicated. Both preoperative MRI and TVS can predict low risk patients (less than 50% of myometrial invasion) accurately; thereby avoiding lymphadenectomy in these patients.