Reproductive outcomes of different management strategies after hysteroscopic resection of the uterine septum with endometrial polyps: a retrospective study

Background

The uterine septum, a common congenital anomaly with an incidence of 2% to 3% in women of childbearing age (Rikken et al., 2019), is associated with recurrent pregnancy loss and obstetric complications (Prior et al., 2018; Practice Committee of the American Society for Reproductive Medicine, 2024; Turocy & Rackow, 2019). The guidelines from the Practice Committee of the American Society for Reproductive Medicine (ASRM) recommend surgical intervention for patients with a uterine septum and a history of recurrent pregnancy loss, following thorough consultation (Practice Committee of the American Society for Reproductive Medicine, 2024). The development of intrauterine adhesions is a major complication following septum resection, which can affect subsequent pregnancy outcomes and increase the risk of placental abnormalities (Practice Committee of the American Society for Reproductive Medicine, 2024; Shen et al., 2022b). The reported incidence of intrauterine adhesions varies greatly among different management protocols after septum resection, ranging from 0 to 37.5% (Tonguc et al., 2010; Shen et al., 2022a). Estrogen therapy is thought to promote endometrial repair and thereby prevent intrauterine adhesion (Ma et al., 2021). However, some early studies have suggested that endogenous estrogen is sufficient for endometrial regeneration and repair, making supplementation with exogenous estrogen unnecessary (Yu et al., 2016; Roy et al., 2014). The efficacy of such methods remains debated.

Patients with a uterine septum are sometimes accompanied by endometrial polyps. One study reported that the concurrent incidence of endometrial polyps in women with a uterine septum was approximately 25 to 33% (Xiong et al., 2024). Endometrial polyps can further impede fertility through mechanisms such as physical obstruction, the inflammatory response, and decreased endometrial receptivity (Vitale et al., 2021). Although approximately 25% of endometrial polyps, particularly those less than 1.0 cm in diameter, may resolve spontaneously (Bougie et al., 2024), performing endometrial polypectomy is recommended for women with fertility needs (Gu et al., 2018). Hysteroscopic surgery can ensure complete removal of the polyp and its base under direct visualization (Sheng & Lyons, 2020). The recurrence rates of endometrial polyps vary widely from 2.5% to 46% (Liu et al., 2021; Ciscato, Zare & Fadare, 2020). Estrogen receptors are highly expressed in the glandular tissue of endometrial polyps, leading to an excessive response of the endometrium to estrogen, which results in the formation of polyps. Consequently, progestins are frequently administered to inhibit the recurrence of polyps following surgical intervention (Raz et al., 2021; Tanos et al., 2017; Gündüz et al., 2022).

The artificial hormone cycle, consisting of sequential administration of estrogen and progestin, aims to facilitate the rapid regeneration of the endometrium at sites of injury within the uterine cavity, thereby preventing the formation of adhesions (Healy et al., 2016). Short-acting oral contraceptives contain low doses of estrogen and progestin, wherein the potent progestin component exerts a significant inhibitory effect on endometrial proliferation. As a result, these contraceptives are commonly utilized as a strategy to prevent the recurrence of polyps in women with recent fertility considerations (Raz et al., 2021). For patients with fertility needs after resection of the uterine septum combined with endometrial polyps, the optimal approach to restoring the uterine cavity while simultaneously preventing polyp recurrence remains undefined. This study aims to compare the outcomes of artificial hormone cycles, short-acting contraceptives, or no medication to establish evidence-based recommendations for postoperative management in this patient population.

Results

A total of 189 cases were included and divided into three groups: Group A (artificial cycle, n = 92), Group B (short-acting contraceptive, n = 52), and Group C (no treatment, n = 45). The baseline characteristics were similar across the three groups (Table 1). As shown in the table, there were no significant differences in age, BMI, previous number of induced abortions, types of fertility problems, and types of uterine septum.

Reproductive outcomes

Table 2 presents a comparative analysis of pregnancy outcomes across the three treatment groups. In Group A, 37 out of 92 women (40.2%) achieved a live birth, compared to 18 out of 52 women (34.6%) in Group B and 14 out of 45 women (31.1%) in Group C (χ² = 1.192, P = 0.547). There was also no evidence of a difference in clinical pregnancy, pregnancy loss, or preterm birth rates.

Table 2:

Comparison of different pregnancy outcomes across the three groups.

This table compares the rates of primary and secondary pregnancy outcomes among the three groups (A, B, and C). The primary outcome is live birth, while secondary outcomes include clinical pregnancy, pregnancy loss, and preterm birth. Data are presented as the number of cases followed by the percentage within each group. Group comparisons were performed using the chi-square test.

Pregnancy outcomes	Group A (n = 92)	Group B (n = 52)	Group C (n = 45)	χ2	p-value
Primary outcome
Live birth	37 (40.2%)	18 (34.6%)	14 (31.1%)	1.192	0.547
Secondary outcomes
Clinical pregnancy	44 (47.8%)	22 (42.3%)	18 (40.0%)	0.882	0.66
Pregnancy loss	7 (7.6%)	5 (9.6%)	4 (8.9%)	0.336	0.891
Preterm birth	2 (2.2%)	2 (3.8%)	1 (2.2%)	0.668	0.846

DOI: 10.7717/peerj.20669/table-2

Notes:

Data are n(%).

Multivariate Logistic regression analysis was performed after including confounding factors such as age, BMI, types of fertility problems, and types of uterine septum. The results showed that only age (adjusted OR = 0.892, 95%CI [0.822−0.968], P = 0.006) was significantly associated with live birth after surgery (Table 3).

Table 3:

Multivariate logistic regression analysis for live birth after surgery.

This table presents the results of a multivariate logistic regression analysis performed to identify factors independently associated with the likelihood of achieving a live birth following surgery.

Variable	P-value	Adjusted OR	95%CI
Age	0.006	0.892	0.822 to 0.968
BMI	0.777	0.987	0.904 to 1.078
Types of fertility problems (vs Infertility)	0.876	1.055	0.541 to 2.057
Types of uterine septum (vs Complete)	0.311	0.696	0.345 to 1.404
Postoperative medication
A vs C	0.195	1.689	0.765 to 3.730
B vs C	0.417	1.447	0.593 to 3.530

DOI: 10.7717/peerj.20669/table-3

The mean time to pregnancy resulting in a live birth after surgery was 9.6 months (95%CI [8.9–10.3]) for Group A, 10.2 months (95%CI [9.4–11.0])) for Group B, and 10.4 months (95%CI [9.5–11.3]) for Group C (log-rank P = 0.468) (Fig. 1).

Discussion

Our study showed that there were no significant differences in pregnancy outcomes and postoperative complications among the three groups of artificial cycle, short-acting contraceptives, and no medication. Despite expectations that hormonal therapy might reduce the risk of these complications, our results did not show any significant protective effect of either the artificial hormone cycle or short-acting contraceptives compared to no treatment.

It should be noted first that there is no consensus regarding the efficacy of uterine septum surgery in improving fertility outcomes (Rikken et al., 2020), despite the patient’s decision to proceed with the procedure for fertility purposes. According to the ASRM Practice Committee, surgical intervention is advised for patients presenting with a uterine septum accompanied by recurrent miscarriage. However, the current evidence substantiating a correlation between septum resection and improved infertility outcomes remains inadequate (Practice Committee of the American Society for Reproductive Medicine, 2024). Some studies showed that the excision of the uterine septum tissue, executed under hysteroscopic observation, endeavors to revitalize uterine cavity morphology and enhance pregnancy likelihood (Carrera et al., 2022; Tang et al., 2022). Noventa et al. (2022) concluded that uterine septum resection can reduce the rates of miscarriage and preterm birth. A randomized controlled trial conducted in 2021 compared hysteroscopic resection with observation management in patients with a septate uterus and found no significant difference in time to pregnancy or live birth rate (Rikken et al., 2021). Whether uterine septum surgery is necessary and whether surgery can improve pregnancy outcomes needs further study.

Previous studies have suggested that hysteroscopic procedures can increase the risk of placental abnormalities (Hajšek et al., 2022). Baldwin et al. (2018) noted that hysteroscopic interventions may lead to a higher likelihood of placenta accreta. The health of a pregnancy and proper placentation are influenced by the condition of the endometrium (Ng et al., 2020). A postoperative medication regimen for patients with a uterine septum combined with endometrial polyps aims to better restore the uterine environment, thereby enhancing the likelihood of successful pregnancy and reducing the rates of placental abnormalities. However, our findings suggested that the addition of hormonal therapy may not provide a clear advantage over expectant management without hormonal intervention. These results are in accordance with a previous prospective study on the effectiveness of estrogen following hysteroscopic septum resection, which found that estrogen therapy did not improve reproductive outcomes after the procedure (Roy et al., 2014). For patients with endometrial polyps, the use of progestins after surgery is to prevent the recurrence of polyps, and the guidelines suggest that patients with immediate fertility needs can attempt pregnancy without the need for postoperative medication (Bougie et al., 2024).

In our study, the incidence of intrauterine adhesions and polyps was low and comparable across all treatment groups, with no significant reduction in adhesion rates and polyps in the groups receiving hormonal therapy compared to those who did not. The rate of intrauterine adhesion in the three groups after the operation was 3.1−6.5%, which is consistent with the previously reported incidence of intrauterine adhesions after uterine septum surgery (Healy et al., 2016). Yu et al. (2016) conducted a cohort study on the incidence of intrauterine adhesions among patients who received estrogen, an intrauterine device, an intrauterine balloon, or no adjunctive treatment after uterine septum resection. Their findings indicated no significant difference in adhesion rates across these groups, implying that the uterine cavity wound was capable of complete spontaneous healing without the administration of supplementary estrogen therapy or other interventions (Yu et al., 2016). Short-acting contraceptives can promote endometrial growth through estrogen, helping to reduce the risk of intrauterine adhesion, while the strong and effective drospirenone can resist endometrial hyperplasia caused by the estrogen effect, resulting in its limited role in repairing wound endometrium (Healy et al., 2016). Regarding the risk of endometrial polyp recurrence, there is limited research on whether artificial cycles after polypectomy increase this risk (Raz et al., 2021; Tanos et al., 2017). The addition of a timely and appropriate dose of progesterone to the estrogen treatment regimen can inhibit the recurrence of endometrial polyps to a certain extent (Kossaï & Penault-Llorca, 2020). Our findings suggest that short-term, low-dose artificial cycle therapy does not increase the risk of postoperative polyps.

All patients included in this study were preoperatively diagnosed with a uterine septum by transvaginal three-dimensional ultrasound imaging, and the presence of endometrial polyps was subsequently confirmed through hysteroscopic surgery. While this approach ensured diagnostic accuracy for our cohort, we should acknowledge the inherent challenges in hysteroscopic diagnosis within clinical practice. A randomized controlled trial revealed that hysteroscopy, when used in isolation, is insufficient as a diagnostic tool for detecting a uterine septum. This limitation arises from its inability to adequately visualize the external contour of the uterus, thereby preventing definitive differentiation between a partial septate uterus and an arcuate uterus (Smit et al., 2015). This underscores the critical role of preoperative ultrasound in our diagnostic protocol (Practice Committee of the American Society for Reproductive Medicine, 2024). In addition, although hysteroscopic examination and hysteroscopic resection of endometrial polyps followed by pathological examination are the gold standards for diagnosing endometrial polyps (Sheng & Lyons, 2020), several studies have indicated that small endometrial polyps can be overlooked during hysteroscopic examination (Saponara et al., 2025; Campo et al., 2014), particularly in low-resource settings where hysteroscopists may have limited training or experience. Failure to detect them may lead to inadequate management of subfertility. This potential for oversight highlights the necessity of proper training and adherence to standardized diagnostic protocols to minimize missed diagnoses and ensure appropriate patient management.

This study has several limitations. As this was a retrospective study, a formal prospective sample size calculation was not performed. Given the sample size of 189 women, statistical analysis indicates that the study was sufficiently powered to detect differences exceeding 20% in live birth rates among the treatment strategies. However, the difference in live birth rates between the highest (Group A, 40.2%) and lowest (Group C, 31.1%) groups in our study was only 9.1%. Although no statistically significant difference in live birth rates was observed across the three groups, a post-hoc power analysis revealed that the sample size may have been insufficient to exclude a clinically meaningful difference. This limitation of statistical power is particularly evident in the case of our secondary outcomes, which exhibited very low event rates. The interpretation of our outcomes is limited by the risk of a Type II error. Our negative findings indicate that we did not find evidence for a large difference in postoperative therapy, rather than being able to draw equivalence conclusions in the different strategies. Moreover, the non-randomized assignment of treatments and the potential for unmeasured confounders (such as ovarian reserve and lifestyle) are inherent limitations of our retrospective design that may affect the robustness of the data. Furthermore, most patients began attempting pregnancy after a hysteroscopic follow-up at two months post-surgery if no abnormalities were found in the study, and did not undergo regular ultrasound examinations. As a result, it was not possible to assess the recurrence of endometrial polyps beyond two months postoperatively.

Conclusions

The results of this study found no significant differences in reproductive outcomes and complication rates between the use of artificial cycles, short-acting contraceptives, or no hormonal therapy after uterine septum resection with concurrent polypectomy. However, due to the limited sample size and statistical power, these findings should be interpreted with caution. Therefore, clinicians can flexibly choose postoperative management plans based on individual conditions of patients (such as fertility plans, comorbidities, etc.), as there is no clear evidence from this study to support the recommendation of routine hormone therapy. This provides a preliminary basis for simplifying the postoperative management process and reducing medical costs. Future prospective studies with larger sample sizes are needed to confirm these findings and provide clearer guidance on the postoperative management strategy for this patient population.

Supplemental Information