Multivariate Analysis of Recurrence after Hysteroscopic Diagnosis and Treatment of Endometrial Polyps following IVF-ET Failure.

Objective: To explore the risk factors affecting the recurrence of endometrial polyps (EPs) after hysteroscopic diagnosis and treatment of EPs following in vitro fertilization-embryo transfer (IVF-ET) failure by multivariate analysis.
Methods: The clinical data of 369 patients with EPs hysteroscopically treated in our department due to IVF-ET failure from January 2017 to January 2020 were retrospectively analyzed, including the number and size of polyps, postoperative treatment, endometriosis (EM), hydrosalpinx (HSP), and polycystic ovarian syndrome (PCOS), and the effects of these factors on EP recurrence were observed.
Results: Of the patients enrolled, 184 cases (49.9%) were treated by curettage, and 185 cases (50.1%) by electrotomy. A total of 72 cases (19.5%) of postoperative recurrence were determined, including 34 cases (9.2%) without postoperative medication, 31 cases (8.4%) with one month of postoperative Didroxyprogesterone (DG) administration, and 7 cases (1.9%) with three months of postoperative DG administration. Surgical methods, 3 months of postoperative medication, PCOS, and polyp number and size significantly influence the recurrence of EPs, which were all the influencing factors of polyp recurrence. After controlling for other factors, the risk of EP recurrence after electrotomy was found to be lower than that after curettage, with an odds ratio (OR) (95% confidence interval (CI)) of 0.354 (0.163-0.767); the risk of EP recurrence after 3 months of postoperative medication was lower than that without postoperative medication, with an OR (95% CI) of 0.024 (0.005-0.104); the risk of EP recurrence in patients with PCOS was higher than that without PCOS, and the OR (95% CI) was 2.505 (1.113-5.639); patients with multiple polyps (≥2) were at an increased risk of recurrence than those with a single polyp, with an OR (95% CI) of 66.552 (14.711-301.084); patients with polyp diameter ≥ 2 cm had a higher risk of recurrence than those with polyp diameter < 2 cm, and the OR (95% CI) was 1084.76 (148.743-7910.999). Conclusions: PCOS patients are at an elevated risk of EP recurrence than non-PCOS patients. In patients with multiple polyps, those with a diameter ≥ 2 cm have an increased risk of polyp recurrence compared with those with polyp diameter < 2 cm; electrotomy is associated with a lower recurrence risk of EPs than curettage. The risk of EP recurrence in patients treated with postoperative progesterone for 3 months is lower than that of patients without postoperative medication.

1. Introduction

Endometrial polyps (EPs) are benign pathological entities induced by localized overgrowth of basal endometrium [1]. As one of the most common gynecological diseases in women of childbearing age, it is also an important reason for the failure of assisted reproduction [2]. Currently, the pathogenic mechanism of EPs is not yet completely clear, but it is shown to be associated with excessive estrogen, hypertension, and obesity [3]. EPs are prevalent with an incidence of about 7.8%–34.9% and the clinical presentations of abnormalities, uterine bleeding, menorrhagia, infertility, etc. [4, 5]. Routine hysteroscopy of infertile women abroad revealed an incidence of EPs as high as 17.6% [6]. And the incidence of EPs in primary infertility is 3.8%–38.5%, higher than that in secondary infertility (1.8%–17%) [7]. In addition, EPs have also been found to be a risk factor for pregnancy loss, with a prevalence rate of 15–50% among patients with recurrent abortion [8]. Big data studies at home and abroad have also shown that EPs are the most common endometrial lesions in patients with infertility or even repeated implantation failure [9, 10]. Therefore, it is of great significance to optimize treatment options for EPs to ameliorate repeated implantation failure.

With the development of minimally invasive techniques, transcervical resection of polyp (TCRP) has become the preferred treatment for EPs, with many advantages such as high accuracy, low blood loss, and postoperative recurrence rate [11]. Among them, hysteroscopic curettage as a traditional surgical method can temporarily control polyp bleeding, but it may cause abnormal uterine bleeding because polyps are easy to remain in the uterine cavity [12]. In addition, this surgical approach results in an increased risk of postoperative recurrence due to the curettage of EPs from the pedicle [13]. Hysteroscopic electrotomy, on the other hand, is a surgical method for the treatment of EPs through the electroresection of the pedicled and basal parts [14]. This procedure has been widely applied in clinical gynecological diseases such as perimenopausal abnormal uterine bleeding, intrauterine fibroids, and benign uterine lesions and plays a positive role in improving clinical symptoms, sex hormone levels, and reducing complications [15-17]. However, this treatment also has a certain risk of recurrence, with a postoperative recurrence rate of 13.3% [18], greatly compromising the therapeutic effect. Therefore, it is the current research focus of gynecologists to explore the causes of postoperative recurrence in EPs patients and reduce the recurrence rate. In this study, the high-risk factors of postoperative EP recurrence in patients undergoing hysteroscopic TCRP after in vitro fertilization-embryo transfer (IVF-ET) failure were retrospectively analyzed to provide evidence for clinical diagnosis and treatment.

2. Materials and Methods

2.1. Data Source

A total of 369 patients with hysteroscopically diagnosed EPs who underwent hysteroscopic TCRP due to IVF-ET failure in the Department of Gynecological Endocrinology of the Chongqing Health Center for Women and Children were enrolled. Inclusion criteria were (1) age 20–40 years; (2) EPs were suggested by transvaginal ultrasound and confirmed by postoperative pathology; (3) complete case data [19]; (4) no use of hormone drugs within the last month; (5) no history of hypertension, cardio-cerebrovascular diseases, high-risk factors of thrombosis, migraine, liver and kidney dysfunction, gallbladder diseases, submucosal uterine fibroids, nor malignant tumors. Exclusion criteria were (1) acute reproductive system infection; (2) severe organ dysfunction; (3) uterine diseases such as hysteromyoma, adenomyosis, endometrial tumor, and uterine malformation; (4) history of treatment with sex hormone drugs in recent 3 months; (5) cervical hardness and inability to fully dilate due to cervical scar; (6) coagulation dysfunction and tuberculosis; (7) malignant tumors of the reproductive system; and (8) recent history of uterine surgery. The specific diagnostic criteria for the identification of EPs by vaginal B-ultrasound were endometrial thickening, inhomogeneity, intrauterine space-occupying lesions or strong echogenic masses, and the presence of dotted blood vessels. This study was approved by the Ethics Committee of our hospital, and the study subjects were informed and provided informed consent.

2.2. Methods

2.2.1. Surgical Methods

Preoperative examinations were rounded out 3–7 days after menstruation to eliminate the contraindications. Hysteroscopic TCRP was performed under total intravenous anesthesia, with anesthesia induced by 2.5 mg/kg propofol injection (0.5 mg/kg/h) followed by continuous infusion of propofol injection for maintenance of anesthesia. (1) All patients completed various preoperative examinations and were explained about the operation-related matters. Usually, the operation time was scheduled within 3–7 days after menstruation, with 8–10 h of preoperative fasting. (2) Among the enrolled patients, 185 patients were treated with hysteroscopic electrotomy, with the methods described as follows: Patients received general intravenous anesthesia in the lithotomy position, and 0.9% normal saline was used as the uterine distention solution administered at a pressure of 100 mmHg. The cervix was explored, and a resectoscope was placed after the cervix was dilated to determine the position, number, size, and pedicle site of the polyps. Then the root of polyps was completely removed through the ring electrode to reach the basal layer. Thereafter, a curette was inserted into the uterine cavity, and reasonable scratching and curettage were carried out for residual polyp tissue to ensure its complete removal. Hysteroscopy was used again to check for residual polyps. The rest 184 underwent hysteroscopic curettage, and the methods were as follows: EPs were hysteroscopically observed. Then, a curettage spoon was used to scratch and scrape intrauterine polyps, especially those parts that could not be found or were not easily removed by the ring electrode, and careful examination and treatment were given. The curette was then inserted into the uterine cavity, and reasonable scratching and scraping were carried out for residual polyp tissue to ensure its complete removal. The presence of any residual micropolyp was observed by the hysteroscope again, and a small scissor was used to cut off if any. (3) All tissues removed during the operation were sent to the Department of Pathology for examination. (4) Postoperative treatment: no medication or medication (Didroxyprogesterone (DG), SFDA Approval No. H20090470), per os, 20 mg/d, for 1 month or 3 months). The experimental flowchart is shown in Figure 1.

Figure 1

Experimental flowchart. Note. EPs, Endometrial polyps; TCRP, transcervical resection of the polyp; IVF-ET, in vitro fertilization-embryo transfer; DG, Didroxyprogesterone.

2.3. Outcome Measures

The postoperative EP recurrence rate of patients with two different surgical modalities and between patients with and without medication (1 or 3 months of medication) were compared. Factors influencing EP recurrence were determined using the multivariate analysis.

2.4. Follow-up

We followed up with the subjects by means of telephone visits, visits, reexaminations, and inquiring about the case system to record patients' recurrence. The enrolled patients underwent B-ultrasound every 1 and 3 months after surgery. Excluding 4 (1.07%) losses to follow-ups, 369 cases were eventually included. During the follow-up, recurrence was diagnosed if the B-ultrasound suggested EPs.

2.5. Statistical Analysis

SPSS 23.0 statistical software (IBM Corp., Armonk, NY, USA) was used for analysis. Quantitative data following normal distribution were represented by ( ± s), and those of non-normal distribution were recorded as median (M) (interquartile range (IQR)). In this paper, binary Logistic regression (LR) was used to evaluate the influence of surgical methods, polyp number, polyp size, postoperative treatment, and complications on polyp recurrence. Those with statistical significance were further analyzed by binary LR to calculate the odds ratio (OR) value. P values < 0.05 were considered statistically significant.

3. Results

3.1. Basic Information

The median age of 369 patients with EPs was 31 years old. Among them, most cases (316 (85.6%)) had a body mass index (BMI) between 18.5 and 24.9 kg/m2. There were 193 cases (52.3%) with endometriosis (EM), 121 cases (32.8%) with hydrosalpinx (HSP) and 118 cases (32%) with the polycystic ovarian syndrome (PCOS). Single polyp was found in 192 cases (52%) and multiple polyps in 177 cases (48%). The polyp diameter was less than 2 cm in 337 cases (91.3%); 184 cases (49.9%) were treated by curettage and 185 cases (50.1%) were treated by electrotomy; 72 cases (19.5%) of postoperative recurrence were identified, including 34 cases without postoperative medication (9.2%), 31 cases with one month of postoperative DG administration (8.4%), and 7 cases with three months of postoperative DG administration (1.9%), Table 1.

Table 1

Basic data of EPs patients (M (IQR), n (%)).

Parameters	No. of cases
Age	31
BMI (kg/m2)
< 18	18 (4.9)
18.5–24.9	316 (85.6)
25–29.9	34 (9.2)
> 30	1 (0.3)
Complicated with endometriosis	193 (52.3)
Complicated with hydrosalpinx	121 (32.8)
Complicated with PCOS	118 (32.0)
Single polyp	192 (52.0)
Multiple polyps	177 (48.0)
Polyp diameter < 2 cm	337 (91.3)
Polyp diameter ≥ 2 cm	32 (8.7)
Curettage	184 (49.9)
Electrotomy	185 (50.1)
Postoperative recurrence	72 (19.5)
Recurrence without postoperative medication	34 (9.2)
Postoperative recurrence after one month of DG administration	31 (8.4)
Postoperative recurrence after three months of DG administration	7 (1.9)

Note. BMI, body mass index; PCOS, polycystic ovarian syndrome; DG, Didroxyprogesterone; EPs, Endometrial polyps; M, median; IQR, interquartile range.

3.2. Comparison of Data between Recurrent and Nonrecurrent Patients after Surgery for EPs

Through analysis, we found no significant difference in age and BMI between patients with and without recurrence of EPs (P > 0.05), but the presence of statistical differences in EM, HSP, PCOS, polyp number, polyp size, surgical methods, and postoperative treatment (P < 0.05), Table 2.

Table 2

Comparison of data between recurrent and nonrecurrent patients after surgery for EPs (n (%), mean ± SD).

Factors	n	Recurrence group (n = 72)	Nonrecurrence group (n = 297)	χ 2/t	P
Age (years old)	369	30.94 ± 4.44	30.52 ± 4.73	0.684	0.495
BMI (kg/m2)				4.290	0.232
< 18	18	4 (5.56)	14 (4.71)
18.5–24.9	316	60 (83.33)	256 (86.20)
25–29.9	34	7 (9.72)	27 (9.09)
> 30	1	1 (1.39)	0 (0.00)
Complicated with endometriosis	193	47 (65.28)	146 (49.16)	6.036	0.014
Combined with hydrosalpinx	121	35 (48.61)	86 (28.96)	10.158	0.001
Complicated with PCOS	118	31 (43.06)	87 (29.29)	5.046	0.025
Number of polyps				6.192	0.013
Single polyp	192	28 (38.89)	164 (55.22)
Multiple polyps	177	44 (61.11)	133 (44.78)
Polyp size				7.219	0.007
< 2 cm in diameter	337	60 (83.33)	277 (93.27)
≥ 2 cm in diameter	32	12 (16.67)	20 (6.73)
Surgical methods				6.768	0.009
Curettage	184	26 (36.11)	158 (53.20)
Electrotomy	185	46 (63.89)	139 (46.80)
Postoperative treatment				12.710	0.002
No postoperative medication	62	34 (47.22)	28 (9.43)
1 month of DG administration	200	31 (43.06)	169 (56.90)
3 months of DG administration	107	7 (9.72)	100 (33.67)

Note. BMI, body mass index; PCOS, polycystic ovarian syndrome; DG, didroxyprogesterone; EPs, endometrial polyps.

3.3. Univariate and Multivariate Logistic Regression Analysis

We assigned the variables with significant differences between recurrent and nonrecurrent patients after surgery for EPs as follows (Table 3).

Table 3

Variable assignments of univariate and multivariate Logistic regression analysis of influencing factors of polyp recurrence.

Variables	Assignments
Surgical methods	Curettage = 0, electrotomy = 1
Number of polyps	Single = 0, ≥2 = 1
Size of polyps	<2 cm in diameter = 0, and ≥2 cm in diameter = 1
Postoperative treatment	No medication = 0, medication for 1 month = 1, medication for 3 months = 2
Complicated with endometriosis	Without = 0, with = 1
Complicated with hydrosalpinx	Without = 0, with = 1
Complicated with PCOS	Without = 0, with = 1

Note. PCOS, polycystic ovarian syndrome.

3.4. Univariate Analysis of EP Recurrence

The univariate analysis results of EP recurrence are shown in Table 4. With α = 0.05 as the test level, the univariate analysis indicated that the p values of the number of polyps, polyp size, postoperative treatment, EM, HSP, and PCOS were all under 0.05, indicating statistical significance. Patients with multiple polyps were more likely to experience postoperative recurrence than those with single polyp; and compared with patients with polyp diameter < 2 cm, those with polyp diameter ≥ 2 cm were at an increased risk of postoperative EP recurrence. Besides, complications such as EM, PCOS, and HSP increased the risk of postoperative relapse. Patients were less likely to suffer from relapse after 3 months of medication than those without medication. Although no statistical significance was observed in EP recurrence between the postoperative medication (with vs. without) and between surgical methods (curettage vs. electrotomy), the two may be the influencing factors of postoperative polyp reexamination in a professional sense, so they were still included in the multivariate analysis model as classified variables.

Table 4

Univariate analysis of endometrial polyp recurrence.

Related factors	β	S.E.	OR (95% CI)	Wald value	P value
Surgical methods (with curettage as reference)
Electrotomy	−0.076	0.263	0.927 (0.552–1.552)	0.083	0.773
Number of polyps (with single lesion as reference)
Multiple	1.445	0.297	4.244 (2.373–7.589)	23.757	<0.001
Polyp size (with < 2 cm as reference)
≥2 cm	3.556	0.513	35.040 (12.832–95.681)	23.757	<0.001
Postoperative treatment (with nonmedication as reference)
1 month after medication	−1.48	0.289	0.863 (0.489–1.521)	0.260	0.610
3 months after medication	−1.857	0.438	0.156 (0.066–0.369)	17.931	<0.001
Complicated with endometriosis	0.459	0.268	1.582 (0.935–2.676)	2.926	0.087
Complicated with hydrosalpinx	1.131	0.270	3.099 (1.826–5.258)	17.572	<0.001
Complicated with PCOS	0.602	0.270	1.825 (1.075–3.098)	4.965	0.026

Note. PCOS, polycystic ovarian syndrome; OR, odds ratio; CI, confidence interval.

3.5. Multivariate Analysis of EP Recurrence

Multifactor analysis variables were screened by the forward LR method. Although the univariate analysis of surgical methods and postoperative treatment methods were suggested P > 0.05, they were still included in the multivariate analysis from professional considerations. The final model results are shown in Table 5. The stepwise regression results indicated significance in surgical methods, postoperative medication for 3 months, PCOS, polyp number, and polyp size, suggesting their roles as influencing factors of polyp recurrence. While postoperative medication for 1 month, HSP and EM were excluded. After controlling for other factors, the risk of polyp recurrence after electrotomy was found to be lower than that after curettage, with an OR (95% confidence interval [CI]) of 0.354 (0.163–0.767); The risk of polyp recurrence after 3 months of postoperative medication was lower than that without postoperative medication, with an OR (95% CI) of 0.024 (0.005–0.104); The risk of polyp recurrence in patients with PCOS was higher than that of non-PCOS patients, and the OR (95% CI) was 2.505 (1.113–5.639); The polyp recurrence risk in patients with multiple polyps (number ≥ 2) was higher than that of patients with a single polyp, and the OR (95% CI) was 66.552 (14.711–301.084); patients with polyp diameter ≥ 2 cm had a higher risk of recurrence than those with polyp diameter < 2 cm, and the OR (95% CI) value was 1084.76 (148.743–7910.999).

Table 5

Multivariate analysis of endometrial polyp recurrence.

Related factors	β	S.E.	OR (95% CI)	Wald value	P value
Surgical methods	−1.038	0.394	0.354 (0.163–0.767)	6.929	0.008
Postoperative treatment
1 month after medication	−0.82	0.395	0.922 (0.425–1.998)	0.043	0.836
3 months after medication	−3.744	0.758	0.024 (0.005–0.104)	24.414	<0.001
Complicated with PCOS	0.918	0.414	2.505 (1.113–5.639)	4.917	0.027
Number of polyps	4.198	0.770	66.552 (14.711–301.084)	29.714	<0.001
Size of polyps	6.989	1.014	1084.76 (148.743–7910.999)	47.532	<0.001

Note. PCOS, polycystic ovarian syndrome; OR, odds ratio; CI, confidence interval.

4. Discussion

Hysteroscopic TCRP is an accurate, safe, and effective procedure operated under direct vision, which is therefore the preferred treatment for EPs [20]. The results of this study showed that the postoperative recurrence rate was 19.5%, which was similar to previous literature [21]. At present, the causes of recurrent EPs in women of childbearing age are still elusive, but it is considered to be related to age, obesity, PCOS, estrogen stimulation, inflammation, etc [22]. In this research, 193 cases (52.3%) of EM, 121 cases (32.8%) of HSP, and 118 cases (32%) of PCOS were identified; There were 192 cases (52%) of a single polyp and 177 cases (48%) of multiple polyps; polyps smaller than 2 cm in diameter were determined in 337 cases (91.3%) and those larger than 2 cm in diameter were identified in 32 cases (8.7%); 184 cases (49.9%) were treated by curettage, and 185 cases (50.1%) underwent electrotomy; 72 cases (19.5%) had a postoperative relapse, among which 34 cases (9.2%) recurred without medication, 31 cases (8.4%) recurred after one month of DG administration, and 7 cases (1.4%) relapsed after three months of DG administration. Univariate and multivariate analysis showed that surgical methods, 3 months of postoperative medication, PCOS, and polyp number and size were predictive factors for the recurrence of EPs.

EM was found in 193 cases (52.3%) in this study. Lin et al. [23] reported that the incidence of EPs in stages 1–4 EM was 42.44%, 40.69%, 55.89%, and 51.52%, respectively, which were similar to our cohort data. As both EM and EPs are estrogenic diseases, the correlation between their occurrence has always been a hotspot in clinical research. It has been reported that most infertile patients with EM are accompanied by EPs, suggesting that EM is related to the occurrence of EPs [24]. Evidence has shown that EM syndrome in infertile patients is positively correlated with the occurrence of EPs [25], indicating that there may be mutual promotion between the pathological changes of EM and EPs in such patients.

Hysteroscopic TCRP is recommended as the most ideal treatment for EPs, as well as the gold standard for surgical diagnosis and treatment of polyps, which can not only significantly reduce bleeding in patients but also eliminate malignant lesions through polyp biopsy. The methods of hysteroscopic TCRP include unipolar electrotomy, bipolar electrocoagulation system, endoscopic scissors or graspers, and uterine curettage devices. Hysteroscopic treatment of multiple EPs mainly includes hysteroscopic electrotomy and hysteroscopy positioning curettage of polyps. Compared with curettage, hysteroscopic electrotomy can effectively remove polyps in the base and surrounding tissues. In addition, the recurrence rate of EPs by curettage is high, mainly because simple curettage cannot remove the base of polyps [19]. In this study, the recurrence of EPs in patients undergoing hysteroscopic electrotomy was lower than that of patients with curettage, which is consistent with the literature [26]. However, the thermal damage of electrotomy cannot be ignored as electrosurgical resection of polyps may lead to endometrial damage and intrauterine adhesion, and there is currently no definite evidence for its applicability to patients with fertility requirements.

Obesity and PCOS are high-risk factors for recurrence [27]. Since EP is a hormone-dependent disease, its recurrence may be related to high estrogen levels in patients. The conversion of peripheral adipose to estrogen increases in obese patients, and adipose tissue can increase estrogen storage. Serhat et al. showed that obesity, an independent risk factor for EPs, was not associated with diabetes and hypertension [28]. Hyperandrogenism is present in PCOS patients. Androgens can be converted to estrogen through the peripheral adipose tissue, which can increase estrogen storage and lead to long-term high estrogen effect in the endometrium, resulting in the imbalance of estrogen receptor (ER) and progesterone receptor (PR) in the endometrium [29]. Due to the decrease of PR expression in EPs, the endometrium shows a reduced sensitivity to progesterone response and even no response to progesterone, which leads to excessive hyperplasia of local endometrium tissues. Under the continuous stimulation of estrogen, the local endometrium of PCOS patients experienced excessive hyperplasia due to long-term anovulation and lack of progesterone antagonism, resulting in the formation of EPs. In addition, there are not only ER and PR in the endometrium but also insulin receptors. The combination of insulin and receptors can promote endometrial hyperplasia, so the increase in insulin level plays an important part in endometrial dysplasia [30]. Obese and PCOS patients often have insulin resistance and chronic inflammation, thus increasing the risk of EP recurrence. The results of this study confirm once again that PCOS is a high-risk factor for EPs. On the other hand, Gu et al. [21] also pointed out in their study that the risk of EP recurrence in patients with multiple EPs was about three times higher than that in patients with a single EP, suggesting that the number of EPs was an independent risk factor for EP recurrence, which was consistent with our research results. The novelty of this study and the contribution to the subject area is that by analyzing the clinical data of the included EPs patients who failed IVF-ET treatment, it was confirmed that PCOS, multiple polyps, polyp diameter ≥ 2 cm, curettage, and no postoperative medication were all risk factors for the recurrence of EPs, providing new references and predictive factors for the postoperative management and recurrence prevention of EP patients with IVF-ET failure. This study still needs to be improved on the following aspects. First, since this is a single-center study, there may be inevitable information bias. Second, the number and size of polyps and other factors were not further subdivided to analyze the influence of these factors on postoperative recurrence of EPs after IVF treatment failure. Last, potential risk factors such as estrogen stimulation and inflammation were not included to verify their role in the postoperative recurrence of EPs after failed IVF treatment. In the future, this research will be gradually improved from the above perspectives.

5. Conclusion

EPs are a kind of most common endometrial lesions in infertile women. At present, most studies suggest that EPs have negative impacts on fertility through mechanical interference and inflammatory stress. For infertile women, hysteroscopic polypectomy can improve the pregnancy rate of both natural pregnancy and assisted reproduction. However, due to the lack of research in this area and low data quality, it is still controversial whether polypectomy should be regarded as a standard treatment before assisted reproduction. Postoperative recurrence of EPs is a difficulty in clinical management. However, as its etiology and mechanism are still unclear, the means to prevent recurrence and the treatment after recurrence need more evidence-based medical evidence. We will design a series of rigorous randomized controlled studies, as well as more translational and basic research, as soon as possible to provide a more convincing theoretical and practical basis for the clinical treatment of EPs.