Abstract
Aim: To study outcomes of patients diagnosed with endometrial carcinoma (EC) after histological analysis of endometrial resections retrieved during operative hysteroscopy performed for a presumed benign lesion. Patients and Methods: A retrospective study was conducted using medical records of patients who underwent operative hysteroscopy for a presumed benign lesion with a final diagnosis of EC between January 1994 and April 2014 in two tertiary academic centers. Results: A total of 29 patients were selected. International federation of gynecology and obstetrics (FIGO) classification was distributed as follows: 16 stages IA, 7 stages IB, 4 stages II and 2 stages III. Peritoneal cytology was positive in one case (stage IIIA). Median follow-up was 4.2 years (range=0.3-20.51). Two deaths were observed and were attributed to endometrial cancer. Conclusion: Operative hysteroscopy does not appear to influence stage of EC nor cause retrograde seeding of EC for 27/29 (93%) patients. For 2 patients, the impact of operative hysteroscopy remains uncertain.
- Surgical hysteroscopy
- endometrial cancer
- endometrial adenocarcinoma
- prognosis of endometrial cancer
- peritoneal cytology
Endometrial cancer is diagnosed on endometrial tissue obtained by blind sampling or by hysteroscopy (1). A number of studies report inconsistent results regarding the risk of positive peritoneal cytology and potential dissemination of the disease (2). Nevertheless, the impact of peritoneal cytology after diagnostic hysteroscopy on long-term survival and the risk of recurrence has yet to be evaluated (2-4). There are, however, very few studies evaluating the impact of operative hysteroscopy performed for a presumed benign lesion with a final diagnosis of endometrial adenocarcinoma (5). The main differences between operative and diagnostic hysteroscopy are, for operative hysteroscopy, longer operative times, use of higher intrauterine perfusion pressures and ability to perform surgical procedures. Although this situation is less frequent, it is important to evaluate the risk of retrograde seeding of endometrial cancer during operative hysteroscopy. The objective of this study was to evaluate outcomes of patients with endometrial cancer diagnosed on endometrial resections obtained during an operative hysteroscopy performed for a presumed benign lesion.
Patients and Methods
This retrospective bicentric study was conducted in the Obstetrics and Gynecology Department of two academic hospitals (Hôpital La Conception and Hôpital Nord) in Marseille, France, between January 1994 and December 2014. The inclusion criteria were as follows: a preoperative sonography showing no suspicion of malignancy, an operative hysteroscopy for a presumed benign lesion with a final pathological diagnosis of endometrial adenocarcinoma. Medical records were selected using the DIM (Département d'information médical). Institutional Review Board approval was obtained for this study (no.CEROG-2014-GYN-0905) and all patients gave their informed consent to participate.
During this period, the operative technique was identical in both hospitals. All patients underwent an operative hysteroscopy under general anesthesia. Operative hysteroscopy was performed after cervical dilation with Hank dilators, using an 8- or 9-mm hysteroscopic resectoscope and a 5-mm loop electrode. The uterine cavity was distended using a glycine solution with a continuous monitoring of intake and output. Monopolar energy was used. Resection was performed after having excluded a suspected malignant lesion. Suspicious features of the endometrium suggesting endometrial carcinoma were defined by at least one of the following criteria: irregular or papillary surface, evidence of necrosis, irregular vessel pattern (6, 7). All patients diagnosed with endometrial carcinoma after pathological analysis underwent preoperative imaging (magnetic resonance imaging (MRI) or computed tomography (CT) scan). This preoperative imaging guided the decision concerning the therapeutic strategy established during a multidisciplinary meeting. After histopathological analysis, the disease was staged using the 2009 international federation of gynecology and obstetrics (FIGO) classification (7-9). Staging for patients who were operated before 2009 was updated using the revised 2009 FIGO criteria (10).
Information regarding patients' characteristics, initial treatment, histological characteristics and subsequent follow-up were collected from medical records. Information concerning patients lost to follow-up or who were followed up outside the department was collected from medical records held by the patient's general practitioner. In case of death, the year and cause of death were collected. Results are presented as mean±standard deviation and median values (minimum-maximum).
Results
During the study period, a total of 6,104 operative hysteroscopies were performed and 29 patients met the inclusion criteria. Median age was 62 years (range=48-72). Twenty-seven (93.1%) patients had a menopausal status. A hormonal replacement treatment was reported in 8 (27.6%) cases. Mean follow-up was 6.44±6.18 years.
All patients consulted for pre- or postmenopausal bleeding. Median duration of operative hysteroscopy was 15 minutes (range=4-60) and median volume of glycine solution used was 2 liters (range=0.3-11). Hysteroscopic procedure performed was a resection of an intrauterine lesion for 5 patients (17.2%), a total endometrectomy for 8 patients (27.6%) and both for 16 patients (55.2%). The pathological results of the lesion resected or total endometrectomy comprised: 21 endometrioïd carcinomas (72.4%), 2 clear-cell adenocarcinomas (6.9%), 4 mucinous adenocarcinomas (13.8%) and 2 papillary serous carcinomas (6.9%). A total of 25/29 (86.2%) type 1 histotypes and 4/29 (13.8%) type 2 histotypes were observed.
Following cancer diagnosis, all patients underwent surgical treatment consisting of a total hysterectomy with a bilateral salpingo-oophorectomy after preoperative imaging (MRI or CT scan). Lymphadenectomy was performed depending on the stage, patient's age and operability (obesity, comorbidities). Mean time between operative hysteroscopy and surgery was 35.3±19.9 days.
FIGO classification status according to the final histopathology report was distributed as follows: 16 stage IA, 7 stage IB, 4 stage II and 2 stage III patients (7). There was only 1/29 positive peritoneal cytology (3.4%) and 1/29 positive pelvic lymph node (3.4%). Peritoneal cytologies and lymphadenectomies performed for stage I and II patients were all negative.
Outcomes of stage IA patients. Post-operative histopathological analysis revealed 16 patients presenting a FIGO stage IA disease: 8 of these tumors were grade 1 (including 7 endometrioid carcinomas and 1 mucinous adenocarcinoma), 5 were grade 2 (including 4 endometrioid carcinomas and 1 mucinous adenocarcinoma). There was one grade 3 (endometrioid carcinoma) and 2 clear-cell carcinomas.
All patients underwent surgery during the initial phase of treatment. A pelvic lymphadenectomy was performed in 8 cases. For both patients with a clear-cell carcinoma a surgical treatment equivalent to ovarian cancer was performed. Seven patients underwent adjuvant radiotherapy. No patient received adjuvant chemotherapy. All patients were alive during the follow-up period, except one patient who died 7 years after initial treatment of gastric cancer.
Outcomes of stage IB patients. All 7 patients presenting with stage 1B endometrial cancer received surgical treatment followed by radiotherapy and brachytherapy. Four of these patients had a grade 1 endometrioid carcinoma and all of them were alive during the follow-up period. One patient had a mucinous adenocarcinoma (grade 1). She died 3.7 years later from another cause (heart failure). One patient had a grade 2 endometrioid carcinoma. She was alive at 1.9 years follow-up. The last patient presented a papillary serous carcinoma (type 2). She died 12 years later of breast cancer.
Outcomes of stage II patients. Treatment and outcomes for stage II patients are detailed in Table I.
Outcomes of stage III patients. There was one stage IIIA and one stage IIIC1 patient. The stage IIIA patient was a non-menopausal patient aged 48 who underwent a 15-min endometrectomy that enabled the diagnosis of a grade 1 endometrioid carcinoma. The preoperative MRI classified her as a presumed stage FIGO IA. Fourteen days later, she underwent a total hysterectomy with bilateral salpingo-oophorectomy and pelvic lymphadenectomy. The final histopathological analysis revealed a grade 2 endometrioid carcinoma with myometrial invasion >50%, an invasion of the serosa and adnexas, a positive peritoneal cytology and no lymphovascular invasion or nodal involvement. She received chemotherapy and radiotherapy and died two years later from disease. The stage IIIC1 patient, aged 60, underwent a 20-min endometrectomy that enabled the diagnosis of a grade 1 endometrioid adenocarcinoma. The preoperative MRI classified her as a presumed stage FIGO IB. She underwent the same surgical treatment as described for the previous patient 31 days after endometrectomy. Pathological analysis revealed a grade 1 endometriod adenocarcinoma with a myometrial invasion >50%, lymphovascular invasion, 3 metastatic pelvic nodes and a negative peritoneal cytology. She received radiotherapy and brachytherapy and is still alive with no recurrence 19 years later.
Patients with international federation of gynecology and obstetrics (FIGO) stage II (FIGO 2009).
Two patients died from endometrial cancer, 1 was stage IIIA and 1 stage II. The stage II patient was a 70-year-old and underwent a polyp resection that lasted 10 minutes. Analysis of hysteroscopic resection products revealed a papillary serous carcinoma. She underwent, 23 days later, a total hysterectomy with bilateral salpingo-oophorectomy, pelvic and aortic lymphadenectomy and an omentectomy. The pathologic results revealed a grade 2 papillary serous carcinoma with myometrial invasion >50%, lymphovascular invasion, isthmus and endocervix invasion. There were 0/12 involved lymph nodes and peritoneal cytology was negative. She received radiotherapy but refused brachytherapy. A recurrence occurred 4 years later consisting of a 7-cm metastatic iliac lymph node. She died from endometrial cancer after a median follow-up of 5 years. Three other patients died of other causes during the follow-up period: stomach cancer, breast cancer and end-stage heart failure.
Discussion
This study reports a series of 29 patients for which the definitive pathological results was an endometrial carcinoma diagnosed on endometrial resection specimens retrieved during an operative hysteroscopy performed for a presumed benign lesion. In all cases, pre-operative clinical examination ultrasound and hysteroscopy had not showed any signs of malignancy. The final staging found 16 stage IA, 7 stage IB, 4 stage II, 1 stage IIIA and 1 stage IIIC endometrial cancers. Peritoneal cytology was negative in 28/29 patients (96.5%). There were 2 FIGO stage III patients. Both patients underwent MRI after the diagnosis that reported no signs of poor prognosis and no extra-uterine invasion. In one of these two cases peritoneal cytology was positive.
This is a small series but, to our knowledge, this is the largest series of patients with endometrial cancer diagnosed during an operative hysteroscopy performed for a presumed benign lesion. Vilos et al. published a series of 19 patients who underwent operative hysteroscopy between 1990 and 2005 (5). In 16 of these women, prehysteroscopy office endometrial biopsy was not done, was impossible to perform, provided inadequate sample or was reported as proliferative endometrium or endometrial hyperplasia. All patients underwent a partial or complete hysteroscopic endometrial resection. Endometrial cancer was diagnosed after pathological examination of endometrial resections. All 16 patients underwent the standard treatment for endometrial cancer. Follow-up ranged from 1 to 14 years and there were no recurrences. The rate of positive peritoneal washings was not specified in the study. Three patients died from causes that were not related to their endometrial disease, 13 patients were still alive with no recurrence. A pilot study on operative hysteroscopy in endometrial cancer was conducted in 2006, including 35 patients (10). The objective was to evaluate accuracy of endometrial biopsy (2 to 4) by means of hysteroscopic resectoscope for the evaluation of tumor differentiation (grade) in endometrial cancer. The aim was to optimize surgical treatment by improving characterization of the tumor (histology, grade). After definitive pathological examination of the hysterectomy specimen, tumor grade was concordant in 97.1% cases. None of the patients had positive peritoneal cytologies.
In our study, there was only one positive peritoneal cytology; a stage IIIA patient. This positive cytology could be subsequent to retrograde seeding but, considering the other locations of the disease, it is likely that it could result from spontaneous dissemination. The malignant potential of disseminated tumor cells during hysteroscopy is debated and a number of studies report conflicting results; however, these studies concern diagnostic hysteroscopy alone (11-13). Nevertheless, its implication in endometrial cancer remains controversial (3, 4).
Regarding the stage IIIC1 patient, a nodal involvement was found, although myometrial invasion was minimal. The implication of operative hysteroscopy in her nodal involvement is questionable. It is theoretically possible that tumor cells could disseminate though the vascular or lymphatic system. Inflammatory and cicatricial processes could potentially be responsible of a faster dissemination in the lymphatic system (14, 15).
There is also a potential risk of local dissemination after resections (16). The stage IIIA patient underwent pelvic ultrasound examination, that reported no abnormality, followed by a preoperative operative hysteroscopy and MRI classifying her FIGO stage IA. However, the definitive pathological examination of the hysterectomy specimen revealed an invasion of >50% of the myometrium and a tumor spread to the adnexa. For the stage IIIC patient, pelvic ultrasound examination was normal but the preoperative MRI classified her stage IB. The implication of intrauterine perfusion pressure has been evaluated. Baker and Adamson reported that there was no spill into the peritoneal cavity when perfusion pressure was maintained under 70 mmHg (17).
In our series, 2/29 patients (7%) were finally diagnosed as stage III, although there was no evidence of malignant lesions neither after ultrasound examination nor hysteroscopy. An explanation could be that both patients had an advanced stage of endometrial cancer and ultrasonography underestimated the malignant nature and particularly myometrial invasion. Indeed, ultrasound is effective for detection of adenocarcinomas but not for evaluating myometrial invasion (18-20).
In conclusion, operative hysteroscopy was not associated with a positive peritoneal cytology for 27/29 patients (93%). Influence of operative hysteroscopy for these patients remains uncertain. For both patients that were finally classified stage III, a correct staging, before operative hysteroscopy and implication of operative hysteroscopy in the dissemination of the disease, could be considered. Various mechanisms could explain this dissemination. One of them could be retrograde tubal dissemination as reported for diagnostic hysteroscopy for which the prognostic impact is debated. However, other mechanisms, such as hyperpressure or post-resection processes, could be responsible for myometrial and lymphatic dissemination.
- Received May 21, 2016.
- Revision received June 12, 2016.
- Accepted June 14, 2016.
- Copyright© 2016 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved





