Caroline Bouche1, Manuel Gomes David1, Julia Salleron2, Philippe Rauch1, Léa Leufflen1, Julie Buhler1, Frédéric Marchal1,3. 1. Surgery Department, Institut de Cancérologie de Lorraine, 6, Avenue de Bourgogne, 54519 Vandoeuvre-les-Nancy, France. 2. Biostatistics Department, Institut de Cancérologie de Lorraine, CEDEX 54519 Vandoeuvre-les-Nancy, France. 3. CRAN, UMR 7039, CNRS, Université de Lorraine, Boulevard des Aiguillettes, 54506 Vandoeuvre-les-Nancy, France.
Abstract
OBJECTIVE: The aim of this retrospective cohort study is to evaluate the concordance between the preoperative MRI and histology data with the final histopathological examination. METHOD: This is a retrospective observational study of 183 patients operated for endometrioid cancer between January 2009 and December 2019 in the surgical oncology department of the Lorraine Cancer Institute (ICL) in Vandœuvre-lès-Nancy. The patients included are all women operated on for endometrioid-type endometrial cancer over this period. The exclusion criteria are patients for whom the pre-therapy check-up does not include pelvic MRI and those who have not had first-line surgery. The final anatomopathological results were compared with preoperative imaging data and with endometrial biopsy data. RESULTS: For the myometrial infiltration, the sensitivity of MRI was of 37% and the specificity of 54%. To detect nodal metastases, the sensitivity of MRI was of 21% and the specificity of 93%. We observed an under estimation of the FIGO classification (p = 0.001) with the MRI in 42.7% of cases (n = 76) and an overestimation in 24.2% of cases (n = 43). There was a concordance in 33.1% of cases (n = 59). We had a poor agreement between the MRI and final histopathological examination with an adjusted kappa (κ) of 0.12 [95% IC (0.02; 0.24)]. There was a moderate concordance on the grade between the pretherapeutic biopsy and the final histopathological examination on excised tissue with an adjusted kappa of 0.52 [95% IC 0.42-0.62)]. Endometrial biopsy underestimated the tumor grade in 28.9% of cases (n = 50) (p < 0.001), overestimated the tumor grade in 6.9% of cases (n = 12) and we observed a concordance in 64.2% of cases (n = 111). CONCLUSION: The pre-operative assessment of endometrial cancer is inconsistent with the results obtained on final histopathological examination. A study with a systematic review should be done to assess the performance of MRI, only in expert centers, in order to consider a a specific care management for endometrial cancer patients: patients who have had an MRI in an outpatient center should have their imaging systematically reviewed, with the possibility of a new examination in case of incomplete sequences, by expert radiologists, and discussed in multidisciplinary concertation meeting in expert centers, before any therapeutic decision. The sentinel node biopsy must be used for low and intermediate risk endometrial cancer.
OBJECTIVE: The aim of this retrospective cohort study is to evaluate the concordance between the preoperative MRI and histology data with the final histopathological examination. METHOD: This is a retrospective observational study of 183 patients operated for endometrioid cancer between January 2009 and December 2019 in the surgical oncology department of the Lorraine Cancer Institute (ICL) in Vandœuvre-lès-Nancy. The patients included are all women operated on for endometrioid-type endometrial cancer over this period. The exclusion criteria are patients for whom the pre-therapy check-up does not include pelvic MRI and those who have not had first-line surgery. The final anatomopathological results were compared with preoperative imaging data and with endometrial biopsy data. RESULTS: For the myometrial infiltration, the sensitivity of MRI was of 37% and the specificity of 54%. To detect nodal metastases, the sensitivity of MRI was of 21% and the specificity of 93%. We observed an under estimation of the FIGO classification (p = 0.001) with the MRI in 42.7% of cases (n = 76) and an overestimation in 24.2% of cases (n = 43). There was a concordance in 33.1% of cases (n = 59). We had a poor agreement between the MRI and final histopathological examination with an adjusted kappa (κ) of 0.12 [95% IC (0.02; 0.24)]. There was a moderate concordance on the grade between the pretherapeutic biopsy and the final histopathological examination on excised tissue with an adjusted kappa of 0.52 [95% IC 0.42-0.62)]. Endometrial biopsy underestimated the tumor grade in 28.9% of cases (n = 50) (p < 0.001), overestimated the tumor grade in 6.9% of cases (n = 12) and we observed a concordance in 64.2% of cases (n = 111). CONCLUSION: The pre-operative assessment of endometrial cancer is inconsistent with the results obtained on final histopathological examination. A study with a systematic review should be done to assess the performance of MRI, only in expert centers, in order to consider a a specific care management for endometrial cancerpatients: patients who have had an MRI in an outpatient center should have their imaging systematically reviewed, with the possibility of a new examination in case of incomplete sequences, by expert radiologists, and discussed in multidisciplinary concertation meeting in expert centers, before any therapeutic decision. The sentinel node biopsy must be used for low and intermediate risk endometrial cancer.