Efficacy of Applying Hyaluronic Acid Gels in the Primary Prevention of Intrauterine Adhesion after Hysteroscopic Myomectomy: A Meta-Analysis of Randomized Controlled Trials

Intrauterine adhesion (IUA), which mainly occurs after intrauterine surgery or an inflammatory process, is an important but often neglected condition in women of reproductive age. The presentation of IUA varies greatly, ranging from symptom-free to severe, with amenorrhea or infertility. With much advanced development of intrauterine instruments, more intrauterine diseases can be successfully cured by hysteroscopic surgery. Among these, submucosal myoma is one of the best examples. Submucosal myomas are often related to abnormal bleeding, anemia, and possible infertility or miscarriage. However, submucosal myoma after hysteroscopic myomectomy may be complicated by IUA in various grades of severity, and its incidence and prevalence might be nearly one-quarter to one-third of patients, suggesting an urgent need for efforts to decrease the risk of developing IUA after hysteroscopic myomectomy. Many strategies have been reported to be useful for this purpose, and intrauterine application of anti-adhesive gels, such as polyethylene oxide–sodium carboxymethylcellulose (PEO-NaCMC) or auto-crosslinked hyaluronic acid (ACHA), has become increasingly popular in routine clinical practice. This meta-analysis is aimed at investigating the effect of ACHA on the primary prevention of IUA formation after hysteroscopic myomectomy. A pooled analysis of three studies (hysteroscopic surgeries for fibroids, polyps, and septum) including 242 women showed that using PEO-NaCMC or ACHA gel decreased the IUA rate with an odds ratio (OR) of 0.364 (95% confidence interval (CI) 0.189–0.703, p = 0.03). Pooled analysis of two studies that limited the use of ACHA in 119 women showed that the application of ACHA gel for the primary prevention of IUA in patients after hysteroscopic myomectomy led to a statistically significant reduction of the development of IUA postoperatively (OR 0.285, 95% CI 0.116–0.701, p = 0.006). All of this suggests that the use of ACHA gel in patients after hysteroscopic myomectomy could significantly reduce de novo IUA, although more evidence is needed.


Search Strategy and Study Selection
The meta-analysis was conducted based on the recommendation of Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) and was registered in PROSPERO (ID: CRD42020176878) on 28/04/2020. We searched the PubMed, Embase, and ClinicalTrials.gov databases for relevant randomized controlled trials (RCTs) published online from their inception to May 2020. The search was performed without restrictions regarding language and country. Combined search terms included "hyaluronic acid", "adhesion", "intrauterine adhesion", and "hysteroscopic surgery". RCTs were eligible according to the following inclusion criteria: women undergoing hysteroscopic surgery for benign gynecologic disease, adhesion barrier of HA gel applied primarily at the end of surgery, and second-look hysteroscopy performed to identify the incidence and severity of IUA. Endpoints were reported as relative risk (RR) or odds ratio (OR) with corresponding 95% confidence interval (CI). Studies were excluded according to the following conditions: (1) patients with IUA before receiving surgery; (2) case reports, observational studies, or conference abstracts without adequate information for data synthesis; and (3) animal testing. Two reviewers (M.C. and P.-H.W.) independently evaluated all relevant articles retrieved from the databases according to the inclusion and exclusion criteria. Disagreements were resolved by discussion with the third author (W.-L.L.).

Procedures
Two investigators (M.C. and P.-H.W.) independently extracted data from each article, including authors' names, publication year, study period, sample size, indication for surgery, type of hysteroscopic surgery, and incidence of primary IUA after surgery. Risk of bias was assessed by using the Cochrane Collaboration's risk of bias tool covering allocation concealment, sequence generation, blinding, detection bias, attrition bias, and reporting bias.

Statistical Analysis and Data Synthesis
Heterogeneity between studies was evaluated using Cochran's Q test and measured by I 2 statistics. Low, moderate, and high heterogeneity were defined as I 2 values of 25%, 50%, and 75%, respectively. A two-sided p-value of ≤0.05 was regarded as statistically significant. Comprehensive Meta-analysis Version 3.0 (Biostat Inc., Englewood, NJ, USA) was used for data synthesis [68]. A random effect model was used to calculate effect size in meta-analysis due to potential clinical heterogeneity from different surgical indications and investigated populations. Odds ratios (ORs) were calculated for dichotomous outcomes, with 95% CI measuring the effect of applying HA gels in hysteroscopic surgery versus no administration of anti-adhesion products according to the Cochrane Handbook for Systematic Reviews of Interventions Version 6, 2019.

Strategy to Include Studies in the Current Meta-Analysis
After removing duplications and articles with unrelated topics, a total of 34 studies were reviewed in detail for eligibility; 31 studies were excluded, including 16 articles in review form, 1 observational study, 9 studies with evaluations of secondary intrauterine adhesion, 3 animal studies, 1 study with only one arm, and 1 conference abstract. Figure 1 shows the flowchart for identifying studies that met the criteria for the current meta-analysis. In the end, three randomized controlled studies [67,69,70] were included for meta-analysis.

Characteristics of Included Studies
The indications for hysteroscopic surgery were not consistent among the three studies [67,69,70]. Table 1 shows the basic characteristics of the included studies.  Table 2 shows an assessment of risk of bias, which was composed of five domains according to RoB 2, a revised Cochrane risk of bias tool for RCTs [71]. One study may have been at risk of randomization bias since it failed to report on allocation concealment [69]. Two studies failed to keep the investigation blind, which may have led to a higher risk of deviation from the intended Life 2020, 10, 285 5 of 12 intervention [67,70]. One study did not report the dropout rate [69], and the other two had dropout rates of 1.4% and 4.3% [67,70]. All dropout cases resulted from failing to attend follow-up hysteroscopy and outcomes were not evaluated based on intention to treat. However, dropout rates of the two studies were low, which may offset the risk of bias on missing data.

Characteristics of Included Studies
The indications for hysteroscopic surgery were not consistent among the three studies [67,69,70]. Table 1 shows the basic characteristics of the included studies.  On the evaluation of primary IUA rates, two of the three studies demonstrated a significant reduction ( Table 3). The time of follow-up after operation ranged from 9 to 12 weeks. Guida et al.
Life 2020, 10, 285 6 of 12 included cases of hysteroscopic surgery for myomectomy, polypectomy, and intrauterine septum resection, revealing 10.4% of IUA in the treatment group compared with 26.2% of IUA in the control group [67]. Huang et al. limited the patients with submucosal myoma treated by hysteroscopic myomectomy, and the results showed that 12.8% of patients had postoperative IUA in the treatment group compared with 39.1% in the control group [70]. On the other hand, De Iaco et al. did not show a significant difference in IUA rates between intervention and control groups, and their study also included different indications for hysteroscopic surgery, including myomectomy, polypectomy, and intrauterine septum resection [69]. All three of the analyzed studies included information on the IUA rate. For analysis, all three categories were included and pooled into the meta-analysis [67,69,70]. For these 242 patients, there was a significantly reduced risk of developing IUA in the ACHA and HA/NaCMC groups based on a random effect model ( Figure 2). On the evaluation of primary IUA rates, two of the three studies demonstrated a significant reduction ( Table 3). The time of follow-up after operation ranged from 9 to 12 weeks. Guida et al. included cases of hysteroscopic surgery for myomectomy, polypectomy, and intrauterine septum resection, revealing 10.4% of IUA in the treatment group compared with 26.2% of IUA in the control group [67]. Huang et al. limited the patients with submucosal myoma treated by hysteroscopic myomectomy, and the results showed that 12.8% of patients had postoperative IUA in the treatment group compared with 39.1% in the control group [70]. On the other hand, De Iaco et al. did not show a significant difference in IUA rates between intervention and control groups, and their study also included different indications for hysteroscopic surgery, including myomectomy, polypectomy, and intrauterine septum resection [69]. All three of the analyzed studies included information on the IUA rate. For analysis, all three categories were included and pooled into the meta-analysis [67,69,70]. For these 242 patients, there was a significantly reduced risk of developing IUA in the ACHA and HA/NaCMC groups based on a random effect model ( Figure 2).

Significant Reduction of Intrauterine Adhesion Rates in Patients Undergoing Hysteroscopic Myomectomy
While we focused on evaluating the effectiveness of applying ACHA in the primary prevention of IUA in patients after hysteroscopic myomectomy, two of the studies were included and pooled into the meta-analysis (Table 4) [67,71]. In Guida's study, 49 patients were included in the analysis, contributing to an incidence of IUA of 16% in the ACHA treatment group and one-third in the group without ACHA [67]. Since all patients in Huang's study were undergoing hysteroscopic myomectomy, upon further examination of their report, we found that two concentrations of ACHA (3% and 4%) were applied in the intervention group [70]. There was no statistically significant difference in the development of IUA between 3% and 4% ACHA application, although the trend showed a higher effect of 4% ACHA not only on the reduction of IUA incidence (17.4% vs. 8.3%, p = 0.352), but also on decreased severity (all had a mild degree of IUA in the 4% ACHA group and one-quarter had a moderate degree of IUA in the 3% ACHA group) [70]. However, compared with

Significant Reduction of Intrauterine Adhesion Rates in Patients Undergoing Hysteroscopic Myomectomy
While we focused on evaluating the effectiveness of applying ACHA in the primary prevention of IUA in patients after hysteroscopic myomectomy, two of the studies were included and pooled into the meta-analysis (Table 4) [67,71]. In Guida's study, 49 patients were included in the analysis, contributing to an incidence of IUA of 16% in the ACHA treatment group and one-third in the group without ACHA [67]. Since all patients in Huang's study were undergoing hysteroscopic myomectomy, upon further examination of their report, we found that two concentrations of ACHA (3% and 4%) were applied in the intervention group [70]. There was no statistically significant difference in the development of IUA between 3% and 4% ACHA application, although the trend showed a higher effect of 4% ACHA not only on the reduction of IUA incidence (17.4% vs. 8.3%, p = 0.352), but also on decreased severity (all had a mild degree of IUA in the 4% ACHA group and one-quarter had a moderate degree of IUA in the 3% ACHA group) [70]. However, compared with no use of ACHA in patients after hysteroscopic myomectomy, application of ACHA successfully decreased the incidence of IUA with both concentrations of ACHA gel (12.8% vs. 39.1%, p = 0.012) [70]. For these 119 patients, there was a significantly reduced risk of developing IUA in the ACHA application groups based on the random effect model (Figure 3).
Life 2020, 10, x FOR PEER REVIEW 8 of 13 no use of ACHA in patients after hysteroscopic myomectomy, application of ACHA successfully decreased the incidence of IUA with both concentrations of ACHA gel (12.8% vs. 39.1%, p = 0.012) [70]. For these 119 patients, there was a significantly reduced risk of developing IUA in the ACHA application groups based on the random effect model (Figure 3).

Discussion
The incidence of IUA after hysteroscopic surgery varies greatly depending on surgical indications and time of postoperative evaluation [1][2][3][4][5][6][7][8][9]24,26,30,47]. Taskin et al. reported IUA following hysteroscopic removal of a single myoma in 31.3% of cases, hysteroscopic resection for multiple myomas in 45.5% of cases, and hysteroscopic resection of intrauterine septum in 6.7% of cases [72]. On the other hand, a study by Yang et al. showed an incidence of IUA of 88% in patients who had undergone hysteroscopic septum resection and 40% in patients after hysteroscopic myomectomy, suggesting a significant proportion of IUA development after hysteroscopic surgery for more complicated diseases, such as uterine septation or myoma [35,36]. Although we found that a number of randomized controlled trials were performed to evaluate the application of ACHA gels as a barrier for the prevention of postoperative IUA, most of the studies did not exclude patients with IUA, and some studies also allowed adhesiolysis as one of the indications for hysteroscopic surgery, which could potentially show a relatively higher incidence of IUA resulting from intrauterine surgeries and underestimate the efficacy of ACHA gels on the primary prevention of IUA after hysteroscopic myomectomy. This meta-analysis focused on studies that enrolled patients who did not have IUA, with the expectation of conducting a more precise evaluation of the effect of ACHA gels on the primary prevention of IUA after hysteroscopic myomectomy.
Under similar clinical circumstances, including time of follow-up and surgical instruments used, Guida et al. [67] and Huang et al. [71] presented relatively consistent results on the efficacy of ACHA gels in the primary prevention of de novo IUA. Both studies revealed a significant reduction of the rate of IUA with the use of ACHA gels. This may be the first meta-analysis focusing on an evaluation of the incidence rate of de novo IUA with ACHA gels in patients who have undergone hysteroscopic myomectomy, demonstrating the low heterogeneity of eligible studies and a more

Discussion
The incidence of IUA after hysteroscopic surgery varies greatly depending on surgical indications and time of postoperative evaluation [1][2][3][4][5][6][7][8][9]24,26,30,47]. Taskin et al. reported IUA following hysteroscopic removal of a single myoma in 31.3% of cases, hysteroscopic resection for multiple myomas in 45.5% of cases, and hysteroscopic resection of intrauterine septum in 6.7% of cases [72]. On the other hand, a study by Yang et al. showed an incidence of IUA of 88% in patients who had undergone hysteroscopic septum resection and 40% in patients after hysteroscopic myomectomy, suggesting a significant proportion of IUA development after hysteroscopic surgery for more complicated diseases, such as uterine septation or myoma [35,36]. Although we found that a number of randomized controlled trials were performed to evaluate the application of ACHA gels as a barrier for the prevention of postoperative IUA, most of the studies did not exclude patients with IUA, and some studies also allowed adhesiolysis as one of the indications for hysteroscopic surgery, which could potentially show a relatively higher incidence of IUA resulting from intrauterine surgeries and underestimate the efficacy of ACHA gels on the primary prevention of IUA after hysteroscopic myomectomy. This meta-analysis focused on studies that enrolled patients who did not have IUA, with the expectation of conducting a more precise evaluation of the effect of ACHA gels on the primary prevention of IUA after hysteroscopic myomectomy.
Under similar clinical circumstances, including time of follow-up and surgical instruments used, Guida et al. [67] and Huang et al. [71] presented relatively consistent results on the efficacy of ACHA gels in the primary prevention of de novo IUA. Both studies revealed a significant reduction of the rate of IUA with the use of ACHA gels. This may be the first meta-analysis focusing on an evaluation of the incidence rate of de novo IUA with ACHA gels in patients who have undergone hysteroscopic myomectomy, demonstrating the low heterogeneity of eligible studies and a more conclusive effect with the use of a single anti-adhesion agent. However, the small sample size and limited number of available studies meeting our inclusion criteria were the major limitations of this meta-analysis, indicating an urgent need to include more randomized controlled trials to clarify the effect of ACHA as a tool for the primary prevention of IUA in patients following hysteroscopic myomectomy.

Conclusions
Applying ACHA gels in patients after hysteroscopic myomectomy could significantly reduce de novo IUA, although more evidence is needed.

Conflicts of Interest:
The authors declare no conflict of interest.

Abbreviations
The following abbreviations are used in this manuscript: ACHA auto-crosslinked hyaluronic acid CI confidence interval D&C dilation and curettage IUA intrauterine adhesion or intrauterine adhesions NaCMC sodium carboxymethylcellulose OR odds ratio PEO polyethylene oxide PRISMA Preferred Reporting Items for Systematic reviews and Meta-analyses RCTs randomized controlled trials RR relative risk