Research Papers:

Preoperative diffusion-weighted magnetic resonance imaging and intraoperative frozen sections for predicting the tumor grade in endometrioid endometrial cancer

Tomohito Tanaka _, Yoshito Terai, Satoe Fujiwara, Yoshimichi Tanaka, Hiroshi Sasaki, Satoshi Tsunetoh, Kazuhiro Yamamoto, Takashi Yamada, Yoshifumi Narumi and Masahide Ohmichi

PDF  |  HTML  |  Supplementary Files  |  How to cite

Oncotarget. 2018; 9:36575-36584. https://doi.org/10.18632/oncotarget.26366

Metrics: PDF 938 views  |   HTML 1301 views  |   ?  


Tomohito Tanaka1,4, Yoshito Terai1,4, Satoe Fujiwara1, Yoshimichi Tanaka1, Hiroshi Sasaki1, Satoshi Tsunetoh1, Kazuhiro Yamamoto2,5, Takashi Yamada3, Yoshifumi Narumi2 and Masahide Ohmichi1

1Department of Obstetrics and Gynecology, Osaka Medical College, Takatsuki, Japan

2Department of Obstetrics and Radiology, Osaka Medical College, Takatsuki, Japan

3Department of Obstetrics and Pathology, Osaka Medical College, Takatsuki, Japan

4Department of Obstetrics and Gynecology, First Towakai Hospital, Takatsuki, Japan

5Department of Obstetrics and Radiology, First Towakai Hospital, Takatsuki, Japan

Correspondence to:

Tomohito Tanaka, email: [email protected]

Keywords: endometrial cancer; diffusion weighted image; apparent diffusion coefficient; frozen section; grade

Received: February 28, 2018     Accepted: November 03, 2018     Published: November 27, 2018


Objective: The histological tumor grade is a strong predictor of nodal metastasis in endometrial cancer; as such, an accurate pre- or intraoperative diagnosis is important for performing lymphadenectomy.

Methods: Ninety-one patients with endometrioid endometrial cancer were imaged on DW-MRI with the apparent diffusion coefficient (ADC) calculated and a frozen section (FS) diagnosis made before and at hysterectomy. The diagnostic accuracy for predicting the tumor grade for diffusion weighted magnetic resonance inaging (DW-MRI) and the FS diagnosis compared to the ultimate histologic status was analyzed.

Results: Among 91 patients with endometrioid endometrial cancer, high-grade (endometrioid G3) tumors had lower ADC values than low-grade (endometrioid G1/2) tumors. The cut-off of the mean ADCmean values for predicting high-grade tumors resulted in 743×10-6 mm2/sec according to the receiver operating characteristic curve. The true positive rates of ADC values and FSs for the prediction of high-grade tumors did not differ to a statistically significant extent (73.3% vs. 66.7%, p=0.7), however, the true negative rate of ADC values for the prediction of low-grade tumors was significantly lower than that of the FSs (64.5% vs. 98.7%, p=0.01). The kappa statistics of ADC values and FSs were 0.23 and 0.73, respectively. Of note, all five patients with high-grade tumors for whom intraoperative FSs indicated low-grade tumors were predicted to have high-grade tumors on preoperative DW-MRI.

Conclusion: A FS diagnosis is more suitable for predicting high-grade tumors than DW-MRI; however, physicians should pay close attention to tumors with low ADC values on preoperative DW-MRI.

Creative Commons License All site content, except where otherwise noted, is licensed under a Creative Commons Attribution 3.0 License.
PII: 26366