Preoperative ultrasound assessment of myometrial invasion in endometrial cancer and prediction of surgical stage

Abstract The aim of the study was to determine the performance of two-dimensional transvaginal ultrasonography (TVUS) in the preoperative assessment of myometrial invasion and cervical stromal invasion (CSI) in women diagnosed with endometrial cancer. Two-dimensional transvaginal ultrasound was performed prior to the diagnostic procedure and before surgical treatment. Myometrial invasion was assessed by measuring the ratio between the endometrial thickness and the antero-posterior diameter of the uterus in the midsagittal plane, by measuring of the smallest tumor-free margin of myometrium and was modified by the subjective assessment of the investigator. Cervical stromal invasion of the tumor was assessed by measuring the distance between the lowest detectable point of the tumor and the internal cervical os. Patients were operated by open surgery, laparoscopic or robot-assisted methods. No statistical difference was found between the number of preoperatively assessed patients (n = 161/180, 89.4%) and those with confirmed diagnosis after surgery (n = 144/180, 80%), (p > 0.05). Before surgery, 55.9% (n = 90/161) of cases were assessed with invasion <50% by TVUS, and 56.3% (n = 81/144) were confirmed after surgery. Cases assessed with invasion of >50% before surgery were 44.1% (n = 71/161). Surgery confirmed 43.8% (n = 63/144) of cases. No myometrial invasion was found preoperatively in 19 cases, using TVUS as a diagnostic tool. After surgery, in 36 cases no myometrial invasion was found. We found more cases with Grade 1 tumor (n = 29/81, 35.8%) when myometrial invasion was less than 50% compared to predominance of Grade 2 (n = 45/63, 71.4%) and Grade 3 (n = 15/63, 23.8%) tumor when invasion was more than 50% (p < 0.01).


Introduction
Endometrial cancer (EC) is the most common gynecologic cancer in the developed countries, affecting 170,000 women annually. Globally there are 320,000 new cases per year [1]. North America has the highest incidence (19.1 per 100,000) and South-Central Asia the lowest (2.7 per 100,000), leaving Northern Europe with an intermediate-high incidence (14.1 per 100,000) [2]. This is the most common tumor of the female genital organs in Bulgaria, and it ranks first in the structure of oncogynecological morbidity with a relative share of 32.1% (2016). The registered cases for 2016 are 1177 [3].
The endometrioid-type EC is subdivided into three grades: well (grade 1), moderately (grade 2) and poorly (grade 3) differentiated carcinomas conforming to the extent of solid, non-gland forming clusters of neoplastic cells. Grades 1 and 2 together constitute up to 85% of endometrioid endometrial cancer (EEC) [4].
The division of EC into type I and type II subgroups is based on clinicopathological findings [5]. Type I accounts for about 80% of all ECs and includes estrogen-related well-differentiated endometrioid adenocarcinomas. It is commonly diagnosed at an early stage in younger obese women. Type II EC is predominantly poorly differentiated endometrioid carcinomas or non-endometrioid carcinomas that more frequently occur in the elderly. They are not estrogen dependent and are associated with a higher risk of metastases and poorer prognosis [5]. Type II ECs constitute around 15-20% of all ECs but cause 40% of deaths [6,7]. This classification is useful as it underlines differences in the pathogenesis of various types of endometrial carcinoma, but it is not applicable to clinical decision-making.
Exposure to endogenous or exogenous estrogen is the major generator of EC. Obesity, early menarche, late menopause, nulliparity, polycystic ovarian syndrome (PCOS) and estrogen-secreting tumors are risk factors for EC via an overload of estrogen [8,9]. Epidemiological studies show that a 5 kg/m 2 increase in body mass index (BMI) is associated with a three-fold higher risk of EC [10]. High BMI accounts for over one-third (34%) of the worldwide EC incidence and practically half (48%) of the cases in North America [8].
Abnormal uterine bleeding is the cardinal symptom (90%) of EC frequently presenting in an early phase of the disease [11]. Abdominal pain and enlargement occur in more advanced stages [7]. Depending on individual risk factors, EC is the cause of 1-14% of postmenopausal bleeding [12].
EC diagnosis is based on the histopathological evaluation of an endometrial sample. Preoperative sampling may be performed by office endometrial biopsy, hysteroscopic biopsy or dilation and curettage. Office endometrial biopsy is successful in obtaining adequate tissue sample in around 90% of cases [13].
Transvaginal ultrasonography is a suitable first-line examination method in postmenopausal bleeding given that endometrial thickness of 4 mm or less has a negative predictive value for EC of more than 99% [14,15]. Nevertheless, histological evaluation of the endometrium should be elicited for women with persistent or recurrent bleeding or individual risk factors for EC regardless of endometrial thickness [14,16]. Transvaginal ultrasound is advocated within the preoperative clinical examination of the EC patient, identifying the size of the tumor, possible myometrial invasion (MI) and cervical stromal invasion (CSI); these parameters influence the likelihood of high-risk disease and advanced stage [17]. Two-dimensional transvaginal ultrasonography (2D TVU) has a moderate diagnostic performance of MI with an overall pooled sensitivity of 82% and a specificity of 81% according to a meta-analysis [18]. So far, the three-dimensional transvaginal ultrasonography (3D-TVU) shows lower agreement, reliability, and accuracy in MI and CSI evaluation than 2D TVS [19]. Magnetic resonance imaging (MRI) is the most common imaging technique for preoperative assessment of endometrial cancer for deep MI and CSI, along with ultrasound, which is an acceptable alternative [20].
The goal of EC surgery is definitive treatment, or if surgery alone is not curative, it should abolish gross disease and stage the patient for a decision on postoperative adjuvant treatment. The mainstay surgery consists of total hysterec tomy, bilateral salpingo-oophorectomy and peritoneal washing. Additional omentectomy is performed in cases with serous EC (SEC). Sentinel lymph node mapping (SLN) or comprehensive lymphadenectomy is carried out based on tailored risk assessment [17,21]. Intraoperative frozen section analysis of the uterus is the mainstay for the decision of performing a staging lymphadenectomy. Preoperative estimation of myometrial invasion and cervical involvement may be useful for surgical planning [22,23].
The aim of the present study was to determine the performance of 2D TVU in the preoperative assessment of myometrial invasion and cervical stromal invasion (CSI) in women diagnosed with endometrial cancer. We also aimed to assess the predictive value of the method as to the type and grade of differentiation of the tumor.

Ethics statement
The research protocol was approved by the local review board and contained questions regarding each patient's medical and reproductive history, vaginal bleeding, demographic and biometric variables as well as sonographic variables.

Patients and design
It is a prospective study including 180 consecutive patients diagnosed with endometrial cancer by D&C or hysteroscopy at the University Hospital of Ob&Gyn Maichin Dom -Sofia, between October 2018 and August 2021 and subsequently subjected to surgical treatment. Ultrasound examinations were done using ultrasound scanner UGEO WS80A ELITE with an endocavitary convex array probe EA 2 − 11 MHz (Samsung, Korea). The patients who did not have endometrial carcinoma in the postoperative histological result were excluded from this study.
Depending on the preoperatively established degree of myometrial invasion with TVUS, patients fell into one of two groups: 90 cases with invasion less than 50% and 71 cases with established myometrial invasion greater than 50%. The group of patients with suspected cervical invasion of the tumor was also analyzed-38 cases.
The patients were examined in dorsal lithotomy position with an empty bladder by transvaginal 2D ultrasound prior to the diagnostic procedure and once again before surgical treatment. Myometrial invasion was assessed by measuring the ratio between the endometrial thickness and the antero-posterior diameter of the uterus in the midsagittal plane and by measuring the smallest tumor-free margin of myometrium. Endometrium:myometrium thickness ratio of more than 1:2 and tumor-free margin of less than 9 mm were accepted as criteria for deep myometrial invasion. The subjective assessment of the investigator apart from the measurements was also taken into consideration. Cervical stromal invasion of the tumor was assessed by identifying the lowest detectable point of the tumor and its relation to the internal cervical os. If the lowest point of the tumor was above the internal cervical os no CSI was presumed. All patients were operated on within 3 days after the ultrasound examination.
As previously described [24], following hysterectomy, a pathologist with substantial experience in gynecologic oncology examined the pathological specimens. The histopathological variables assessed were histological type, grade of differentiation, depth of myometrial invasion and presence of cervical stromal invasion. Only epithelial malignant tumors, i.e. endometrial cancers (endometrioid, mucinous, serous, clear cell, mixed and undifferentiated cancer) and mixed epithelial and mesenchymal malignant tumors, i.e. carcinosarcomas (malignant mixed Müllerian tumor), were included. Endometrioid and mucinous cancers were classified into three grades: Grade 1, well differentiated; Grade 2, moderately differentiated; and Grade 3, poorly differentiated.
According to the histological type and grading of the tumor and the preoperative assessment of its myometrial and cervical stromal invasion, surgical treatment was tailored ranging from simple total hysterectomy with bilateral salpingooophorectomy to hysterectomy with selective lymphonodectomy and hysterectomy with pelvic lymph node dissection. Patients were operated by open surgery, laparoscopic or robot-assisted methods. Histological evaluation of the surgical specimen was considered the reference standard. The International Federation of Gynecology and Obstetrics 2009 staging criteria were used, based on surgical findings at hysterectomy [25].

Data analysis
Statistical analysis was performed using the SPSS for Windows, v 26. The quantitative data were represented by n (%). Statistical methods used include chi-square test and statistical testing of hypotheses. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy were calculated. Statistical significance was established at p < 0.05.

Results
The study involved 180 patients with proven EC who underwent preoperative TVUS. There was a significantly higher number of menopausal women (91.1%, n = 164) than premenopausal women (8.9%, n = 16), as well as those with irregular uterine bleeding (75%, n = 135). Slightly more than half of the women (50.6%, n = 91) had endocrine-metabolic diseases. Women with one or more births were 85.6% (n = 154) compared to women without childbirth (14.4%, n = 26). The demographic characteristics of the studied patients are presented in Table 1.
More than 70% of cases were classified by FIGO in stages I a and b. Endometroid type endometrial carcinoma was prevalent compared to non-endometroid type. After performing a hysterectomy, carcinoma was found in situ in 12 of the cases, representing slightly more than 6.5% of the women surveyed. The number of endoscopic operations-laparoscopic and robot-assisted methods-represented about 66% of the operations performed (Table 2).
In 10.6% (n = 19) of the cases, TVUS invasion did not reveal myometrial invasion. In the postoperative results, the absence of myometrial invasion was found in 20% (n = 36) of all cases. These results do not show a statistically significant difference between the two groups-preoperatively diagnosed with TVUS and postoperatively confirmed by histopathological examination (Table 3). Table 4 presents the postoperative results regarding the established degree of tumor differentiation at different degrees of myometrial invasion. The predominance of cases with moderately or poorly differentiated tumor was impressive, when the myometrial invasion was greater than 50% (71.4% and 23.8%, respectively). When myometrial invasion was less than 50%, highly or moderately differentiated tumors predominated (35.8% and 49.4%, respectively).
Tumor extension to the cervix was correctly assessed in 25 of 38 women in which it was detected by the pathological investigation. This defines sensitivity of 66%, specificity of 97%, PPV of 82% and NPV of 94%.

Discussion
The correct preoperative diagnostics of endometrial cancer is of key importance for the accuracy of the surgery to be performed and for the prognosis of the treatment. Factors of great importance are the stage of the differentiation of cancer, the endometrial thickness invasion and the existence of cervical stromal invasion. Myometrial invasion higher than 50% is directly connected with high grade pelvic and para-aortic lymph metastasis. It is considered that the frequency of pelvic and para-aortic lymph metastasis is connected directly with the differentiation stage of the tumor-in Grade 1 tumor it is 3% and 2%, in Grade 2 tumor it is 9% and 5%, in Grade 3 tumor is 18% and 11% [26]. On the other hand, the cervical localization of the tumor requires more radical surgical treatment [23]. That is why earlier and precise diagnosis by clinicians is necessary. The diagnostic methods of EC used are MRI, CT and TVUS. It is considered that MRI is much more precise compared to CT for myometrial invasion [27]. The ultrasound method is much more used, cheaper and easy to practice. With the development of the ultrasound technology, the precision of this method is close to that of MRI [28][29][30]. Ultrasound diagnosed more cases of deep MI compared to MRI, however, the difference was not statistically significant. The sensitivities and specificities of ultrasound and MRI for deep MI were 86% vs. 77% and 66% vs. 76%, respectively. For CSI, ultrasound and MRI correctly diagnosed the same number of cases [31]. Furthermore, in regions of limited resources, ultrasound is recommended over MRI for the preoperative staging of endometrial cancer [32].
The ultrasound assessment was of very high specificity and sensitivity in our study. No statistical difference was found between the numbers of preoperative TVUS established patients with EC and   those with postoperative confirmed EC. The cases in I stage of disease prevailed. We found that when the myometrial invasion was assessed under 50%, we confirmed to a large extent that endometrial cancer with differentiation G1 or G2 (in 85% of the cases) was found. In contrast, when the preoperative invasion was above 50%, low differentiated cancer (G2 or G3) was found in the postoperative treatment in 95% of the cases. We found similar results from other studies that search the relationship between the preoperative ultrasound diagnoses and MI, CSI in endometrial cancer [33]. In another study with 80 patients the results are similar -preoperative TVUS had 76.5% sensitivity, 91.3% specificity, whereas frozen section analysis had 85.3% sensitivity and 93.5% specificity, and they correlate well with each other [22]. These data will be useful for the necessity and volume of surgical treatment.
We found that the sensitivity of TVUS when myometrial invasion is measured was 90%. It is higher than that calculated in another study, where the number of cases was higher [24], and similar as the percentage when the number of cases was lower [34]. These results show again that regardless of some differences in the reported data, the ultrasound study proves a reliable method for establishing a myometrial invasion in EC when high tech ultrasound apparatuses are used. It is very important to underline that the results of the study and the subjective assessment of the specialist are of great significance. Some authors think that the subjective assessment is more reliable [24]. The sensibility of TVUS in detection of CSI was 66% in our study. This result corresponds to other similar studies [24,34].
The object of future research and introduction into our practice could be the implementation and analysis of the results of included intraoperative frozen sections. This approach can improve the intraoperative staging of the process.

Conclusions
Our observations showed that 2D TVUS is a reliable and simple method for preoperative assessment of myometrial invasion and cervical stromal involvement in endometrial cancer with acceptable sensitivity and specificity. Myometrial invasion of more than 50% is seen exclusively in high-grade tumors or in non-endometriod (type II) cancers. Ultrasound assessment of myometrial invasion is a reliable and useful predictive marker that helps the decision making as to the type of surgical treatment. Preoperative ultrasound study should be used as an important element when the treatment protocol of women with EC is determined.