Is Biopsy Sampling in Endometrial Cancer Accurate?

Peter Kovacs, MD, PhD


October 12, 2017

Accuracy of Endometrial Sampling in Endometrial Carcinoma: A Systematic Review and Meta-analysis

Visser NCM, Reijnen C, Massuger LFAG, et al
Obstet Gynecol. 2017;130:803-813


Endometrial cancer is the most common gynecologic cancer.[1] Most endometrial cancers are of endometrioid type and are associated with exposure to unopposed estrogen. Typically, these cancers are well differentiated and are diagnosed at an early stage when treatments can result in high survival rates.[2] There are known risk factors that are associated with endometrial cancer (ie, older age, infertility, irregular cycles, nulliparity, unopposed estrogen use, tamoxifen use, obesity, hypertension, diabetes).[3]

Endometrial cancer is staged surgically. In addition to the stage of the disease and histologic type, the degree of differentiation influences management.[3]

Treatment is primarily surgical, but in patients with lymph node metastasis, postoperative radiotherapy is recommended as well. Radiotherapy can be considered in cases of local disease that is poorly differentiated or for those who are poor candidates for surgery. Chemotherapy can be considered for advanced or recurrent disease.[3,4] Therefore, the stage of the disease and the histologic degree of differentiation do affect both initial and adjuvant therapies. This paper reviewed the accuracy of endometrial biopsy during the preoperative assessment of endometrial carcinoma.

Study Summary

This meta-analysis was based on the results of 45 studies involving 12,459 patients. Preoperative endometrial sampling was performed by hysteroscopic directed biopsy, aspiration biopsy, or dilatation and curettage (D&C). Biopsy results (histology and grading) were compared with the pathologic findings of the hysterectomy specimens.

Overall, the preoperative and pathologic findings agreed in 67% of the cases (95% CI, 60%-75%). When histologic grade was evaluated the lowest agreement was for the preoperative diagnosis of grade 2 (61%; 95% CI, 53%-69%). An agreement was found in 75% of preoperative grade 1 diagnosis (95% CI, 69%-81%) and in 75% (95% CI, 65%-86%) of preoperative grade 3 diagnosis. Hysteroscopic biopsy had the highest (89%; 95% CI, 80%-98%) and D&C had the lowest (70%; 95% CI, 60%-79%) degree of agreement.

Clinically relevant downgrading was found in 26% of the cases (from high grade to low grade), and clinically relevant upgrading was found in 8% of the cases (from low grade to high grade).

Preoperative biopsy correctly identified the endometrioid histologic subtype in 95% (95% CI, 94%-97%) of the cases and the nonendometrioid subtype in 81% (95% CI, 69%-92%).


Endometrial cancer typically presents with abnormal uterine bleeding. Ultrasound and endometrial sampling are used for the evaluation of abnormal uterine bleeding.[3] An endometrial sample can be obtained with office biopsy, hysteroscopy, or D&C. The initial histologic test can differentiate between benign and malignant causes and in case of malignancy may determine the surgery the patient has to undergo.

Most endometrial cancers are stage I or stage II and are treated/staged surgically; those with more advanced stage are candidates for chemotherapy or radiation therapy. Patients with endometrioid cancer and grade 1 or 2 histology localized to the uterus may undergo hysterectomy only and require no extensive dissection of lymph nodes. Therefore, the advanced knowledge of histology and grading could determine the extent of surgery. Performing a hysterectomy and salpingo-oophorectomy alone, however, may prove to be suboptimal if the final diagnosis indicates grade 3 disease. The contrary is also true if a patient undergoes lymph node dissection; if the final diagnosis is stage I, grade 1 or 2 disease, she may expose herself to unnecessary postoperative morbidity (lymphedema) due to the extensive surgery. Worldwide practices differ significantly when it comes to lymph node sampling/ dissection.[3]

This study found a moderate agreement on histologic grading between biopsy and final pathology results. In one of 12 cases the final pathology indicated significant upgrading where more extensive surgery would be indicated. In one of four cases there was clinically significant downgrading from preoperative histology. In these cases, less aggressive surgery can prove to be enough.

Overall, one can conclude that current preoperative grading is not too reliable. The authors suggested that either the low-, medium-, high-grade system should be changed to low and high grade only or further immunohistochemical markers should be evaluated to improve accuracy of preoperative grading.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: