Prediction of Myometrial Invasion in Stage I Endometrial Cancer by MRI: The Influence of Surgical Diagnostic Procedure

SIMPLE SUMMARY: Fertility sparing treatment can be considered for young women with clinical stage 1A endometrial cancer (EC) without myometrial invasion (MI). Surgical diagnostic procedures (SDP) were needed to make diagnosis of EC, but different extents of SDP including diagnostic hysteroscopic biopsy (DHB, group 1), operative hysteroscopic partial resection (OHPR, group 2), operative hysteroscopic complete resection (OHCR, group 3), and cervical dilatation and fractional curettage (D&C, group 4) may affect the accuracy of MI assessment by magnetic resonance imaging (MRI) after SDP. Here, we retrospectively review those initially diagnosed with stage 1A EC and compare MI status on MRI reports and final histopathology of hysterectomy. We found that the MRI accuracy of MI was better in patients with EC diagnosed with D&C. Three diagnostic procedures using hysteroscopy might interfere with the diagnostic power of MI on MRI. Thus, D&C for diagnosis of EC and further hysteroscopic complete resection with hormone as a fertility sparing treatment for those confirmed as stage 1A without MI from MRI may be a choice in the future. ABSTRACT: Young women with endometrial cancer (EC) can choose fertility-sparing treatment for stage 1A disease without myometrial invasion (MI). The surgical diagnostic procedure (SDP) may affect the accuracy of magnetic resonance imaging (MRI) to assess MI. Here, we evaluated different SDP and compared the MI on MRI results with further pathologic results after hysterectomy. We retrospectively collected data on 263 patients with clinical stage IA EC diagnosed between January 2013 and December 2015. Patients were classified into four groups based on SDP, including diagnostic hysteroscopic biopsy (DHB, group 1), operative hysteroscopic partial resection (OHPR, group 2), operative hysteroscopic complete resection (OHCR, group 3), and cervical dilatation and fractional curettage (D&C, group 4). The sensitivity, specificity, diagnostic accuracy, positive predictive value, and negative predictive value of MRI to assess MI were 73.1%, 46.7%, 63.9%, 71.8%, and 48.3%, respectively. Three hysteroscopic procedures (groups 1 to 3) had a trend with a higher odds ratio of discrepancy between MRI and histopathology (p = 0.068), especially in group 2 (odds ratio 2.268, p = 0.032). Here, we found MRI accuracy of MI was better in patients with EC diagnosed with D&C. Three diagnostic procedures using hysteroscopy might interfere with the diagnostic power of MI on MRI.