Clinical Usefulness of Endometrial Cytology in Determining the Therapeutic Effect of Fertility Preserving Therapy

Surgery, including total hysterectomy, has been the standard treatment for endometrial cancer; however, fertility preservation therapy using medroxyprogesterone acetate (MPA) is a treatment option for patients with well-differentiated and early-stage endometrial cancer [1]. Although MPA therapies for endometrioid carcinoma G1 (EM G1) and atypical endometrial hyperplasia (AEH) have high complete response rates of 90.7% and 98.5%, respectively, it has previously been reported that their recurrence rates are also high [2], and pathological confirmation of disease remission before termination of treatment is critical. In the National Comprehensive Cancer Network (NCCN) guidelines [3], histological evaluation of the effect of progestin therapy by endometrial sampling is recommended. However, it does not mention the use of endometrial cytology (hereafter referred to as “cytology”). The significance of cytology in evaluating the effect of MPA therapy includes the following: (1) the addition of cytology to dilatation and curettage (D&C) will screen diseases that were overlooked by D&C alone; and (2) cytology can be an alternative to D&C. However, the available evidence is insufficient to support these findings. For example, from the standpoint of the latter, cytology is expected to reduce the burden on patients and decrease the possibility of complications caused by D&C. Therefore, this study aimed to clarify the usefulness and significance of cytology in evaluating the effects of MPA therapy.

In our hospital, MPA therapy is administered to patients who meet the following conditions: (1) EM G1 or AEH that is localized to the uterine body and have not invaded the myometrium based on imaging studies; (2) non-smoking patients of ≤42 years of age; (3) body mass index (BMI) of <35 kg/m²; (4) no liver function disorder, abnormal coagulability, and history of thrombosis based on the laboratory results; and (5) patients who desire to preserve their fertility. MPA can be administered not only to patients with first-onset EM G1 or AEH but also to those with intrauterine recurrence while being under strict monitoring. Generally, the administration of MPA is started at a dose of 600 mg/day in patients diagnosed with EM G1 or AEH by D&C. In addition, MPA was also administered to patients in whom the presence of EM G1 or AEH was confirmed using endometrial biopsy. This included patients examined by a previous doctor, despite the absence of any disease on D&C before the commencement of treatment. Its therapeutic effect is evaluated by D&C after 4 months, which is the minimum period of treatment, but we also perform cytology concurrently in all cases. In addition, the intrauterine lesions are evaluated using hysteroscopy prior to the evaluation of the effect of MPA therapy. MPA administration is terminated when remission is pathologically confirmed. In patients with residual disease, D&C is performed every 2 months while continuing treatment, which is terminated once remission is confirmed.

In performing cytology, specimens are collected from the patients under intravenous anesthesia after cervical dilation using Hegar forceps and immediately before endometrial curettage using a curette. The specimens for cytology are collected by scraping the endometrium using the Scree Brush (Soft Medical, Bunkyo-ku, Japan), a cell-sampling brush for uterine cancer and prepared using the conventional method, and performing a Papanicolaou staining prior to their evaluation by a cytopathologist and a cytotechnologist. In principle, hysteroscopy is performed by at least 2 gynecologists using the Olympus HYF Type XP (Olympus, Shinjuku-ku, Japan) to evaluate the presence or absence of residual neoplastic lesions and intrauterine adhesions.

Clinicopathological information, such as the patient’s characteristics, histological diagnosis by D&C, and results of the cytological evaluation, were collected from the medical records of patients who received MPA therapy and in whom the therapeutic efficacy of MPA was evaluated between January 2018 and December 2019 at our hospital. Patients who underwent MPA therapy as the initial treatment as well as those who underwent therapy for recurrences were included in the study. After excluding those who did not undergo cytology for the evaluation of the therapeutic effect of MPA, those whose pathological diagnosis was not sufficiently evaluated, and those whose medical record information was incomplete, 77 patients were identified and included in the study. This study complies with the principles of the Declaration of Helsinki and has been approved by the Ethics Committee of Keio University School of Medicine (Ethics Committee Approval Number: 20120243).

In the evaluation of the therapeutic effect of MPA, patients with endometrial hyperplasia without atypia (EH) as well as those with AEH or worse on D&C were defined as cases “with residual disease.” Regardless of whether the histological diagnosis was AEH or EM G1, a cytological diagnosis of “suspicious or positive” was defined as positive for residual disease (note that positive here does not mean positive as a cytological diagnosis but as a statistic term). Furthermore, to calculate the sensitivity and specificity of cytology, the patients with residual disease on D&C despite their negative or inappropriate cytological evaluation were defined as “cytology-only negative” cases, while those with no residual disease on D&C despite their suspicious or positive cytological evaluation were defined as “cytology-only positive” cases. To compare the accuracy of the evaluation of the therapeutic effect of MPA, the sensitivity and specificity of cytology were calculated from the results of 63 patients that underwent cytology during D&C before the commencement of treatment.

Prior to the evaluation of the therapeutic effect of MPA, hysteroscopy was used to identify the patients with adhesions in the uterine cavity, who were then defined as cases “with intrauterine adhesions,” and those with lesions that were not obviously raised were defined as cases “with a small tumor mass.” Furthermore, cases with disease recurrence of AEH or worse within 6 months after the termination of MPA therapy were defined as “early recurrence.” These pieces of clinical information were used to analyze the accuracy of cytology.

We analyzed the patients’ characteristics using t test to compare age and BMI, whereas Wilcoxon test was used to compare the interval (days) from the commencement of MPA therapy to the time when D&C was performed. χ² test (Fisher’s exact test with expected number <5) was used to compare other factors. All statistical tests were performed using JMP® 15 (SAS Institute Inc., Cary, NC, USA) software.

A total of 142 evaluations on the therapeutic effect of MPA using D&C and cytology were performed in 77 patients. The median age at the commencement of treatment was 36 years (range, 24–44 years), and the median BMI at the commencement of treatment was 25 kg/m² (range, 18–42 kg/m²). Of the 142 cases, 132 (93%) were nulliparous. Obese patients were given weight loss guidance before the commencement of MPA therapy. The histopathological diagnoses before beginning MPA therapy were 72 cases (51%) and 70 cases (49%) of EM G1 and AEH, respectively. Of the 77 patients, 39 (51%) received MPA therapy for the treatment of disease recurrence (second and subsequent MPA therapies).

The cytology reports of 63 cases were obtained during D&C prior to MPA administration. In the cytological evaluation of the 57 cases with EH or worse on D&C, 12 had positive cytology, 42 had suspicious cytology, and 3 had negative cytology. On the other hand, of the 6 cases with no disease on D&C, 1 had suspicious cytology, while 5 had negative cytology. Based on these results, the sensitivity and specificity of cytology before treatment were calculated to be 95% and 83%, respectively.

The histological diagnosis using D&C and cytology in all 142 evaluations of the therapeutic effect of MPA are shown in Figure 1. Residual disease (EH or worse) was found on D&C in 76 cases (54%), suspicious cytology or worse in 51 cases (36%), and negative/inappropriate cytology in 91 cases (64%). The relationship between the cytological evaluation and the histological diagnosis using D&C is shown in Table 1. With D&C being the gold standard for diagnosis, the sensitivity and specificity of cytological evaluation were 61% and 92%, respectively. In addition, the cytological evaluation before the start of treatment revealed that the sensitivity and specificity of cytology in patients with EM G1 and AEH were 71% and 86%, and 47% and 97%, respectively, showing that sensitivity was significantly lower in cases with AEH (p = 0.038) (Table 2).

To analyze the characteristics of the patients with “cytology-only negative” and “cytology-only positive,” we compared them with the patients who received concordant evaluations using D&C and cytology (“concordant” cases). Examples of the cytology and histopathological findings of patients with “cytology-only negative” and “cytology-only positive” are shown in Figures 2 and 3, respectively. Table 3 shows the clinicopathological characteristics of patients with “cytology-only negative.” The number of cases of intrauterine adhesions was significantly higher in patients with “cytology-only negative” than in patients classified as “concordant” cases (48% vs. 22%, p = 0.0061), although there was no significant difference in the factors such as age, BMI, parity, and diseases before the commencement of treatment. There was no significant difference between the 2 groups in terms of the interval from the beginning of MPA therapy to the time when D&C was performed (the period during which the effect of hormone therapy was added). The results of cases with “cytology-only positive” are shown in Table 4. The age of patients with “cytology-only positive” was significantly higher than that of patients classified as “concordant” cases (41 years vs. 35 years, p = 0.024), and the interval between the commencement of MPA therapy and the time when D&C was performed was significantly longer in patients with “cytology-only positive” compared to those in patients classified as “concordant” cases (250 days vs. 187 days, p = 0.020). In addition, patients with “cytology-only positive” tended to have a higher number of cases of intrauterine adhesions than patients classified as “concordant” cases, although the difference was not statistically significant (60% vs. 21%, p = 0.073).

Of the 61 “concordant” cases that had no residual disease on D&C and with negative cytology, MPA therapy was terminated in 59 cases based on the evaluation results. Treatment in the remaining 2 cases was continued due to hysteroscopic findings that could not rule out residual diseases, as well as severe uterine cavity adhesions and concerns about their underestimation. Therefore, including the 3 cases with “cytology-only positive,” MPA therapy was terminated in a total of 62 cases, and the presence or absence of early recurrence was examined in those cases. Early recurrence was observed in 9 (15%) out of 59 “concordant” cases and in 1 (33%) out of 3 cases with “cytology-only positive”; however, there was no significant difference between the 2 groups (p = 0.42).

This study evaluated the usefulness of cytology, together with the results of D&C performed concurrently, in determining the effect of MPA therapy. Although the sensitivity and specificity of cytology in the diagnosis of general endometrial cancer were reported to be 89% and 99%, respectively [4], there are few reports on the accuracy of cytology during fertility preservation therapy to the best of our knowledge. Furthermore, the European Society of Gynaecological Oncology/the European Society for Radiotherapy & Oncology/the European Society of Pathology guidelines [5] recommend endometrial biopsy or hysteroscopic guided biopsy to determine the effect of treatment, but there is no mention of endometrial cytology.

The sensitivity of cytology in all cases was 61%, and it was particularly low at 48% in patients with AEH, indicating that diseases are often overlooked when the therapeutic effect is evaluated using cytology alone. In examining patients with “cytology-only negative,” this tendency was evident especially in those with intrauterine adhesions. In addition, even though 95% of the cytological examinations performed before the commencement of MPA therapy in this study could detect EH or worse, its sensitivity was found to be low in the evaluation of the effect of MPA therapy, which shows that it is inappropriate to evaluate the therapeutic effect of MPA with cytology alone and omitting D&C. Moreover, considering the fact that the incidence of placenta accreta is reported to be higher in patients with a history of D&C [6], the use of cytology as an alternative to D&C would not only reduce the burden on patients but also lead to the prevention of perinatal complications in future. However, D&C is still deemed essential due to the low sensitivity of cytology. In addition, unlike our study, a previous report [4] targeted the general population and treated patients with AEH or worse as “cases with disease.” Therefore, it should be noted that the numerical values of sensitivity and specificity cannot be simply compared between those in that study and those of ours.

We then examined the usefulness of adding cytological evaluation to D&C. The addition of cytology was thought to capture diseases overlooked during D&C, but this has not been previously investigated. Focusing on cases of “cytology-only positive,” the percentage of patients with such cases was low (6%), and there was only 1 case of early recurrence, which was revealed in a subsequent cytological evaluation. In addition, the number of early recurrences was not particularly high in patients with “cytology-only positive” in which MPA therapy was terminated when compared to that in patients who were classified as “concordant” cases with no residual disease on D&C and had negative cytology. The number of patients evaluated as suspicious/positive only by cytology was low, and the number of early recurrences in those patients was not high. In addition, negative cytology is not necessarily a basis for safely terminating MPA therapy. Therefore, we did not find any significance in adding cytology to D&C in all cases. However, longitudinal study and further examinations with a larger number of cases may be necessary, as our study included only 3 cases with “suspicious” cytology in which early recurrence was investigated.

A comparative examination of patients with “cytology-only positive” and those classified as “concordant” cases showed that the interval between the commencement of MPA therapy to when D&C was performed was significantly longer in patients with “cytology-only positive,” and they tended to have a higher number of cases with intrauterine adhesions compared to those in “concordant” cases. These suggest that the factors that cause “cytology-only positive” (i.e., more information can be obtained by cytology than by D&C) may be the following: (1) less tissue mass can be collected with a longer administration period of MPA; (2) histological diagnosis becomes difficult due to the modification of histological images by hormone therapy; and (3) tissue collection in D&C using a curette is difficult in patients with intrauterine adhesions. These findings suggest that cytology with D&C may be considered an option in patients with a prolonged administration period of MPA or patients with intrauterine adhesions. Nevertheless, it is difficult to conclude whether cytology is beneficial in decision-making regarding treatment plans and prediction of prognosis, as mentioned above.

One limitation of this study is that the histological diagnosis using D&C was the gold standard for evaluating the sensitivity and specificity of cytology. Residual diseases are often found in the resected uterus, even in cases that show therapeutic effects in endometrial tissues during MPA therapy, as reported previously [7]. Therefore, it is normally desirable to confirm the presence or absence of residual disease in hysterectomy specimens; however, hysterectomy is not performed in most patients who desire fertility preservation due to the nature of its therapy. In this study, we first evaluated the usefulness of cytology by comparing it with the results of D&C and further examined the presence or absence of early recurrence within 6 months after the termination of MPA therapy to substitute for the lack of hysterectomy specimens. In addition, because this was a retrospective study based on medical records, a limitation of this study is that data on factors that are likely to affect the diagnostic performance of cytology and biopsy, such as intrauterine adhesions and tumor mass, were not quantitative. It is considered that tumor mass, for example, can be analyzed by recording the weight of the D&C specimens. However, given the low frequency of cases with “cytology-only positive” in this study and the low rate of early recurrences in those cases, collecting additional information is considered to be of less significance.

In conclusion, we confirmed that the histological evaluation using D&C is necessary to evaluate the effect of MPA therapy and that it is not desirable to evaluate the effect of hormone therapy using only cytology without D&C. In addition, there is less significance in adding cytology to D&C, and the usefulness of cytology in terms of decision-making regarding the treatment plans and prediction of recurrence was shown to be low.

We are indebted to Ms. Tomomi Noda, Ms. Hitomi Nishino, and Ms. Keiko Abe of Keio University School of Medicine for their secretarial help.

Study approval statement: This study complies with the principles of the Declaration of Helsinki and has been approved by the Ethics Committee of Keio University School of Medicine (Ethics Committee Approval Number: 20120243). Consent to participate statement: Informed consent was obtained in the form of an opt-out.

The authors have no conflicts of interest to declare.

Funding was received only from institutional sources.

Takuma Yoshimura contributed to conceptualization, data curation, formal analysis, investigation, visualization, and writing. Wataru Yamagami contributed to conceptualization, supervision, review, and editing. Mio Takahashi, Takuro Hirano, Kensuke Sakai, Takeshi Makabe, Tatsuyuki Chiyoda, and Kouji Banno supervised the study. Daisuke Aoki supervised the study and reviewed and edited the paper.

The data that support the findings of this study are available from the corresponding author upon reasonable request.