The levonorgestrel-releasing intrauterine system (LNG-IUS) is a long-acting, reversible contraceptive (LARC) that is inserted into the uterine cavity. It has been used in clinical practice for over 30 years and functions by locally and gradually releasing levonorgestrel (LNG) within the uterus, thereby exerting sustained progestogenic effects for 5 to 8 years. These effects include thickening of the cervical mucus and inhibition of endometrial maturation [1]. In addition to its high contraceptive efficacy, the LNG-IUS is also a first-line, non-surgical treatment for several benign gynecological conditions due to its prolonged progestogenic action on the endometrium. These conditions include adenomyosis [2], endometriosis [3], and endometrial hyperplasia (EH) [4]. It is also effective in treating menorrhagia, and prevents the recurrence of endometrial polyps (EPs) [5, 6].

However, the high rate of LNG-IUS expulsion is a concern. Women presenting with dysmenorrhea, menorrhagia, or adenomyosis experience a higher rate of LNG-IUS expulsion, ranging from 8.10% to 31.10%, compared to those using the LNG-IUS for contraception alone [7]. Patients with a history of LNG-IUS expulsion have a markedly higher risk of re-expulsion, ranging from 27.60% to 57.00% [8, 9]. Consequently, a substantial majority (58.60%–84.50%) of these women discontinue its use after expulsion [7, 8, 10].

Currently, there is no effective strategy to significantly reduce the recurrence of LNG-IUS expulsion in women with high-risk factors. Although pre-treatment with gonadotropin-releasing hormone analogs (GnRHa) can shrink the uterine volume, this does not substantially reduce the expulsion rate, which remains at 13.30%–14.00% [7, 11]. Some combination therapies have shown additional benefits for adenomyosis, including the relief of symptoms such as dysmenorrhea and menorrhagia, while also reducing the expulsion risk. For instance, LNG-IUS placement after GnRHa and high-intensity focused ultrasound (HIFU) reduced the 2-year expulsion rate to 6.38% [12], concurrent endometrial ablation reduced it to 1.90% [13], and laparoscopic adenomyomectomy combined with LNG-IUS showed an expulsion rate of 3.80% [14]. However, these approaches may not be suitable for women requiring LNG-IUS monotherapy.

Zhu et al. [15] first introduced hysteroscopic suturing of the LNG-IUS in a patient with adenomyosis in 2021, effectively preventing expulsion. Subsequent small-scale studies reported lower expulsion rates, ranging from 0.00% to 8.33% [15, 16, 17, 18, 19]. However, not all patients in these studies had a history of LNG-IUS expulsion.

The present study enrolled women with at least one prior LNG-IUS expulsion who wished to continue its treatment. The LNG-IUS was sutured to the uterine cavity under hysteroscopic guidance and using non-absorbable sutures to preserve therapeutic efficacy and reduce the risk of re-expulsion. The aim of this prospective observational study was to evaluate the feasibility, expulsion rate, and adverse effects of this fixation procedure.

This study included 73 women who had previously benefited from LNG-IUS treatment but had experienced at least one episode of LNG-IUS expulsion. A documented history of LNG-IUS expulsion may be a more appropriate indication for hysteroscopic LNG-IUS suture fixation than other reported high-risk factors, such as adenomyosis, uterine fibroids, dysmenorrhea, an enlarged uterine cavity (uterine volume $>$150 mL), being overweight (BMI $>$25 kg/m²), or heavy menstrual bleeding [7, 8, 9, 21, 22, 23]. Repeated insertions and expulsions of the LNG-IUS place a significant financial burden on both the patients and healthcare system, while also increasing the risks associated with related procedures. All patients in this study expressed a strong desire to undergo suture fixation of the LNG-IUS in order to prevent a recurrence of the expulsion. In addition to a history of expulsion, the patients also presented with other known high-risk factors for LNG-IUS expulsion, including dysmenorrhea (64.38%), menorrhagia (67.12%), adenomyosis (73.97%), a uterine volume $>$150 mL (52.24%), and being overweight (38.46%).

All patients successfully underwent hysteroscopic LNG-IUS suture fixation, with a mean operative time of 37.81  $\pm$  14.90 min. Additional hysteroscopic procedures were performed concurrently in 52.05% of cases. Analysis of the learning curves of the four primary surgeons indicated that proficiency in the suturing technique was generally achieved after five procedures (Fig. 4). No hysteroscopy-related complications occurred during any of the operations. The procedure was successfully completed as day surgery in all patients, with a mean hospital stay of 5.63  $\pm$  1.31 h. No late complications were observed at the one-week postoperative follow-up time. These findings confirm the hysteroscopic LNG-IUS suture fixation procedure is easy to master and well-suited for implementation in the setting of day surgery.

Fig. 4.

Learning curve for LNG-IUS suture fixation.

Patients with a history of LNG-IUS expulsion have a high rate of re-expulsion, ranging from approximately 27.60% to 57.00% [8, 9]. In our study, the expulsion rate was successfully reduced to 4.29% through hysteroscopic suture fixation. Similar studies have also reported low expulsion rates following hysteroscopic LNG-IUS suture fixation, ranging from 0.00% to 8.33% (Table 2) [15, 16, 17, 18, 19]. However, in the limited number of published clinical studies, not all patients who underwent the procedure had a prior history of LNG-IUS expulsion. In the study by Zhang et al. [17], only 25.00% (3/12) of patients had a history of expulsion, and a re-expulsion rate of 8.33% was observed 6 months postoperatively. In the study by Lv et al. [19], 49.40% (39/79) of patients had a history of expulsion, with a re-expulsion rate of 2.60%.

In the present study, three patients experienced a recurrence of LNG-IUS expulsion at 4–5 months postoperatively, corresponding with the high-risk expulsion period reported in a previous study [19]. In our three patients with recurrence of expulsion, the initial suture fixation procedure was performed by three different surgeons. Moreover, the suture thread loops and knots were intact, suggesting the expulsion may have been caused by tearing of the muscular layer. Upon reviewing the operation videos for these cases, we identified that the first two knots were over-pushed, therefore creating excessively tight sutures. Zhang et al. [17] have suggested that expulsion could be related to the absorption of absorbable sutures. In a subsequent study, these authors improved fixation by using non-absorbable sutures, resulting in a significantly reduced expulsion rate of 2.60%. However, the specific reasons behind this reduction were not explored in detail [15, 19].

Five patients in the present study experienced a downward shift of the LNG-IUS, all of whom had a diagnosis of adenomyosis. One patient had a distorted uterine cavity, with the uterine fundus protruding into the cavity as a pseudo-septate uterus (Fig. 5A). In this case, the LNG-IUS was fixed at a low position, as the tail reached the internal cervical orifice (Fig. 5B). During follow-up of this patient, transvaginal ultrasound consistently demonstrated that the LNG-IUS remained positioned at least 1 cm from the uterine fundus (Fig. 5C,D). The other cases of downward shift may have been caused by loose formation of the thread loop (Fig. 5E). We also noted the occurrence of an “LNG-IUS swing” phenomenon in two cases with loose thread loops, which has not been reported previously. Both cases exhibited a downward shift in the LNG-IUS during the 1-month follow-up period, although it remained in situ at the 2- and 3-month follow-up times (Fig. 5F,G). The loose thread loop could potentially cause the LNG-IUS to swing within the uterine cavity. It is important to note that even when the LNG-IUS shifts downward, it still provides contraceptive and therapeutic benefits [7]. Therefore, we closely monitored the five patients with a downward shift, in light of the continued benefits they were likely to receive from the device.

Fig. 5.

LNG-IUS downward shift and ‘swing’ phenomenon. (A) Distorted uterine cavity with adenomyosis. Protrusion of the uterine fundus, mimicking a ‘pseudo-septate uterus’ (5 August 2022). (B) Void arrow: Distal tip of the LNG-IUS reaching the internal cervical orifice (5 August 2022). (C) 1.35 cm downward shift of the LNG-IUS at 1-month follow-up (5 September 2022) (void arrows indicate position change). (D) 1.14 cm downward shift of the LNG-IUS at 8-month follow-up (7 April 2023) (void arrows indicate position change). (E) Void arrow: Single case exhibiting a loose thread loop (30 January 2023). (F) Void arrow: 1.36 cm downward shift observed at the 1-month follow-up (23 February 2023). (G) In situ position confirmed at the 2-month follow-up time in the same patient (22 March 2023).

Consistent with previous findings, our study demonstrated that LNG-IUS suture fixation can reduce the menstrual volume and alleviate dysmenorrhea symptoms (Table 2). A large majority of our patients (68/73, 93.15%) experienced either reduced menstrual bleeding or amenorrhea. Among the 47 patients with preoperative dysmenorrhea, 97.87% (46/47) achieved complete or partial symptom relief, comparable to the relief rate of 97.40% reported by Lv et al. [19] in 2024.

In addition to cases of adenomyosis, our study is the first to include patients with EH. Of the 7 patients with this condition, four required postoperative hysteroscopic evaluation. Follow-up multi-point endometrial biopsies revealed complete histopathological regression in two patients with complex hyperplasia, and in two with atypical hyperplasia. The LNG-IUS remained in place in all four cases. These findings suggest that LNG-IUS fixation may also be a viable treatment option for EH. Moreover, in three patients with recurrent EPs, no recurrence was observed after LNG-IUS fixation, indicating a potential preventive effect. However, due to the absence of a control group, the efficacy of LNG-IUS suture fixation for treating EH and recurrent polyps requires further confirmation through prospective, controlled clinical trials.

Nonetheless, suboptimal treatment outcomes were observed in three patients with adenomyosis. One case involved a unicornuate uterus with a rudimentary horn (type IIC). Despite successful LNG-IUS fixation and no postoperative expulsion, the patient experienced persistent dysmenorrhea and ultimately underwent hysterectomy 5 months later. The other two cases, with uterine volumes of 215.78 mL and 619.11 mL, experienced worsening menstrual bleeding despite proper LNG-IUS placement, requiring additional interventions. Similarly, Lv et al. [19] reported two cases in which patients opted for hysterectomy due to unsatisfactory symptom control following LNG-IUS fixation for adenomyosis. The fixed 52 mg dose of the LNG-IUS could limit its effectiveness in patients with a high adenomyosis burden. Therefore, careful patient selection and thorough preoperative evaluation are essential when considering LNG-IUS fixation for such cases.

In our study, irregular uterine bleeding was defined as spotting lasting more than 8 days within one month, or more than 24 days within 90 days. During the 6-month follow-up period, we observed a relatively high incidence (71.23%) of irregular uterine bleeding. A previous study on LNG-IUS suture fixation reported an irregular uterine bleeding rate of 44.30% at 12 months post-procedure [19]. This exceeds the 24.00% incidence of irregular bleeding observed at 12 months following standard LNG-IUS placement in patients with adenomyosis [10]. However, neither the earlier clinical studies on LNG-IUS suture fixation, nor those on standard placement, have clearly defined what constitutes LNG-IUS—related irregular bleeding. Therefore, the relatively high rate of irregular spotting observed in our study may partly reflect differences in the definition of irregular uterine bleeding. As one of the most common adverse effects associated with standard LNG-IUS placement, irregular uterine bleeding typically decreases over time and is negatively correlated with duration of use [24, 25]. The follow-up period in the present study was relatively short compared with the study by Lv et al. [19], which may partly explain the higher incidence of irregular uterine bleeding. Although suture fixation of the LNG-IUS may increase the risk of irregular bleeding, confirmation of such an association requires a control group undergoing standard LNG-IUS placement, a longer follow-up period, systematic data collection at predefined intervals, and quantitative evaluation using bleeding diaries. Despite the relatively high incidence of vaginal spotting in our study, no cases of pelvic inflammatory disease were identified during the 6-month follow-up. All patients tolerated the device well, and none required early removal of the levonorgestrel-releasing intrauterine system due to bleeding.

This study represents the largest clinical investigation to date on hysteroscopic suture fixation of the LNG-IUS in patients with a documented history of expulsion. At 6-month follow-up, the rates of expulsion (4.29%) and downward displacement (7.14%) were both relatively low. LNG-IUS suture fixation also improved dysmenorrhea (97.87%, 46/47) and reduced menstrual volume or the onset of amenorrhea (93.15%, 68/73) in the large majority of patients, while proving effective in the treatment of EH and recurrent EPs.

However, several limitations should be acknowledged. First, the absence of a control group prevents direct comparison with alternative treatment approaches, thereby weakening the strength of our conclusions. Unlike previous clinical studies on LNG-IUS suture fixation, all participants in our study had experienced at least one prior LNG-IUS expulsion. These patients opted either for suture fixation in the hope of continued therapeutic benefit, or abandoned LNG-IUS treatment for alternative therapies. Additionally, LNG-IUS is not covered by medical insurance in China, and hence the recruitment of a control group willing to undergo standard re-insertion posed significant challenges. Second, the study population was heterogeneous, with 73.97% (54/73) of patients diagnosed with adenomyosis. As a result, quantitative assessments such as the Pictorial Blood Loss Assessment Chart (PBAC) and Visual Analog Scale (VAS) for dysmenorrhea were not included. This weakened the reliability of our evaluation of symptom improvement. Third, due to the novelty of the surgical technique at the time this study was initiated, the increased incidence of irregular bleeding was not anticipated. Moreover, previous studies lacked clear definitions or evaluation criteria for irregular bleeding [17, 19]. Objective tools such as bleeding diaries were not employed to assess spotting frequency or volume, thus limiting the robustness of the safety analysis related to irregular bleeding.

Future studies should adopt more refined subgroup designs based on underlying conditions, and incorporate standardized quantitative measures to better assess the efficacy and safety of this technique.

Among patients with a history of LNG-IUS expulsion, hysteroscopic suture fixation demonstrates dual benefits: it reduces the re-expulsion rate to 4.90% while simultaneously alleviating symptoms of dysmenorrhea and menorrhagia.

The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.

YG and YW designed the research study. BL and YG conducted follow-up visits and collected the data. YG and GX analyzed the data and drafted the manuscript. GX and YW critically revised the manuscript. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

This study was conducted in accordance with the guidelines of the Declaration of Helsinki. The study was approved by the Women’s Hospital of Zhejiang University School of Medicine ethics committee (IRB-20220332-T), and all participants provided written informed consent.

The uterine image featured in the graphical abstract was sourced from http://www.album-online.com (alb11341740) and has been properly licensed.

This research received no external funding.

The authors declare no conflict of interest.