The epidemiology, clinical burden, and prevention of intrauterine adhesions (IUAs) related to surgically induced endometrial trauma: a systematic literature review and selective meta-analyses

Abstract

Background: Reproductive-age women with intrauterine adhesions (IUAs) following uterine surgery may be asymptomatic or may experience light or absent menstruation, infertility, preterm delivery, and/or peripartum hemorrhage. Understanding procedure- and technique-specific risks and the available evidence on the impact of surgical adjuvants is essential to the design of future research.

Objective and rationale: While many systematic reviews have been published, most deal with singular aspects of the problem. Consequently, a broadly scoped systematic review and selective meta-analyses identifying evidence strengths and gaps are necessary to inform future research and treatment strategies.

Search methods: A systematic literature review was performed seeking evidence on IUA incidence following selected uterine procedures and the effectiveness of hysteroscopic adhesiolysis on menstrual, endometrial, fertility, and pregnancy-related outcomes. An evaluation of the impact of surgical adjuvants designed to facilitate adhesion-free endometrial repair was included. Searches were conducted in the PubMed, Embase, and Cochrane databases following PRISMA guidelines and included English-language publications from inception to 8 November 2024. Inclusion criteria restricted articles to those reporting IUA epidemiology or related clinical outcomes. Risk of bias assessment used the US NIH tools for interventional and observational studies. Meta-analyses were conducted and reported only for outcomes where there were sufficient data. Per analysis, we report on proportions (with 95% CI), heterogeneity (I2), and the risk of bias for each study included.

Outcomes: The review identified 249 appropriate publications. The risks of new-onset IUAs following the removal of products of conception after early pregnancy loss, hysteroscopic myomectomy, and hysteroscopic metroplasty for septum correction were 17% (95% CI: 11-25%; 13 studies, I2 = 87%, poor to good evidence quality), 16% (95% CI: 6-28%; 8 studies, I2 = 93%, fair to good evidence quality), and 28% (95% CI: 13-46%; 8 studies, I2 = 91%, fair to good evidence quality), respectively. For primary IUA prevention with adjuvant intrauterine gel barriers, the relative risks were 0.45 (95% CI: 0.30-0.68; three studies, I2 = 0%, poor to good evidence quality), 0.38 (95% CI: 0.20-0.73; three studies, I2 = 0%, fair evidence quality), and 0.29 (95% CI: 0.12-0.69; three studies, I2 = 0%, fair to good evidence quality), respectively, following the above potentially adhesiogenic procedures. Following adhesiolysis without adjuvants, the IUA recurrence rate was 35% (95% CI: 24-46%; 13 studies, I2 = 95%, poor to good evidence quality), similar to the rate of 43% for both those treated adjuvantly with an intrauterine balloon (95% CI: 35-51%; 14 studies, I2 = 85%, poor to good evidence quality), or an IUD (95% CI: 27-59%; four studies, I2 = 85%, fair to good evidence quality). The recurrence rate for secondary prevention with gel barriers was 28% (95% CI: 4-62%; three studies, I2 = 94%, good evidence quality). Notably, there was an excess rate of associated adverse obstetrical outcomes, including preterm delivery, placenta accreta spectrum, placenta previa, peripartum hemorrhage, and hysterectomy, with evidence demonstrating the beneficial impact of adjuvant therapies on these outcomes.

Wider implications: This systematic review comprehensively analyzes IUA formation following uterine surgical procedures and adjuvant therapy effectiveness. Even following adhesiolysis, it is apparent that the basilar endometrial trauma thought to facilitate the formation of IUAs may persist and contribute to adverse reproductive outcomes. Many critical gaps remain in our knowledge of the pathogenesis, prevention, and management of endometrial trauma and IUAs.

Pregistration number: PROSPERO (ID: CRD42023366218).

Keywords: endometrial trauma; hysteroscopy; infertility; intrauterine adhesions; metroplasty; myomectomy; placenta accreta spectrum; postpartum hemorrhage; retained products of conception; septum removal.

Systematic review and selected meta-analyses of intrauterine adhesions related to endometrial trauma. IUA, intrauterine adhesions.

Figure 1.

The PRISMA flow diagram of studies identified in the systematic review.

Figure 2.

Pooled results of meta-analysis for the incidence of IUAs following hysteroscopic metroplasty for septa. IUAs, intrauterine adhesions; RCT, randomized controlled trial; τ, tau. Events are cases where IUAs were found at second-look hysteroscopy. Proportion refers to the risk of finding IUAs. All cases could have received antibiotics and/or estrogen-based hormone therapy.

Figure 3.

Pooled results of the meta-analysis for the incidence of IUAs following hysteroscopic myomectomy. IUAs, intrauterine adhesions; RCT, randomized controlled trial; τ, tau. Events include cases where IUAs were found at second-look hysteroscopy. Proportion refers to the risk of finding IUAs. All cases could have received antibiotics and/or estrogen-based hormone therapy.

Figure 4.

Pooled results of the meta-analysis for the severity of IUAs following hysteroscopic myomectomy, all study types. IUAs, intrauterine adhesions; τ, tau. Events are cases with the corresponding degree of IUA severity identified at second-look hysteroscopy. Proportion refers to the risk of finding IUAs by category of adhesion severity using the AFS classification system. All cases could have received antibiotics and/or estrogen-based hormone therapy.

Figure 5.

Pooled results of the meta-analysis for the incidence of IUAs following abdominal myomectomy (laparoscopic and laparotomic), all study types. IUA, intrauterine adhesions; τ, tau. Events are cases where IUAs were found at second-look hysteroscopy. Proportion refers to the risk of finding IUAs. Bortoletto et al. (2022) retrospectively evaluated a cohort of patients post laparoscopic myomectomy, whereas Gupta et al. (2013) retrospectively evaluated subjects post laparotomic myomectomy. Kubinova et al. (2012) was a prospective study that evaluated patients after either myomectomy approach. All cases could have received antibiotics and/or estrogen-based hormone therapy.

Figure 6.

Pooled results of the meta-analysis for the incidence of IUAs following removal of retained products of conception post-delivery (postpartum), all study types. IUAs, intrauterine adhesions; τ, tau. Barel et al. (2015) retrospectively evaluated subjects who exclusively underwent a hysteroscopic approach to removal of RPOC. Westendorp et al. (1998) retrospectively evaluated patients who underwent dilation with sharp curettage. Events include cases where IUAs were found at second-look hysteroscopy. Proportion refers to the risk of finding IUAs. All cases could have received antibiotics and/or estrogen-based hormone therapy.

Figure 7.

Pooled results of the meta-analysis for the incidence of IUAs following removal of products of conception in the first trimester. IUAs, intrauterine adhesions; RCT, randomized controlled trial; τ, tau. Events include cases where IUAs were found at second-look hysteroscopy. Proportion refers to the risk of finding IUAs. All patients experienced spontaneous pregnancy loss. All cases could have received antibiotics and/or estrogen-based hormone therapy.

Figure 8.

Pooled results of the meta-analysis for the severity of IUAs following removal of products of conception in the first trimester, all study types. IUAs, intrauterine adhesions; τ, tau. Events are cases with the corresponding degree of IUA severity identified at second-look hysteroscopy. Proportion refers to the risk of finding IUAs by category of adhesion severity using the AFS classification system. All cases could have received antibiotics and/or estrogen-based hormone therapy.

Figure 9.

Pooled results of the meta-analysis for the recurrence of IUAs following hysteroscopic adhesiolysis. IUAs, intrauterine adhesions; RCT, randomized controlled trial; τ, tau. Events are cases where recurrent IUAs were identified at second-look hysteroscopy. Proportion refers to the risk of finding IUAs. All cases could have received antibiotics and/or estrogen-based hormone therapy.

Figure 10.

Pooled results of the meta-analysis for the severity of IUA following hysteroscopic adhesiolysis, all study types. (A) Pooled results of the meta-analysis for the mean IUA severity following hysteroscopic adhesiolysis utilizing the AFS classification system, all study types. IUAs, intrauterine adhesions; AFS, American Fertility Society; SD, standard deviation; MRAW, raw mean. (B) Pooled results of the meta-analysis for categorical IUA severity following hysteroscopic adhesiolysis, all study types. τ, tau. Each study utilized a different classification system: Acunzo et al. (2003) utilized AFS, Robinson et al. (2008) utilized March, and Bhandari et al. (2015) utilized ESGE. To allow for meta-analyses of studies using different classification systems, the severities were re-categorized as mild, moderate, or severe following the protocol described by Hooker et al. (2014). Events are cases with the corresponding degree of IUA severity identified at second-look hysteroscopy. Proportion refers to the risk of finding IUAs by category of adhesion severity. All cases could have received antibiotics and/or estrogen-based hormone therapy.

Figure 11.

Pooled results of the meta-analysis for the comparative effectiveness of gel adjuvant therapies versus no adjuvant for primary prevention of adhesions following potentially adhesiogenic procedures (RCT study design only). IUAs, intrauterine adhesions; RCT, randomized controlled trial; RR, risk ratio; τ, tau. Events are cases where IUAs were identified at second look hysteroscopy. Adjuvant indicates cases receiving gel adjuvant. Control refers to cases receiving no adjuvant. While various biodegradable gel prophylactic therapies were utilized in these studies, the addition of hyaluronic acid as a component of the gel barrier was noted in all studies except for Di Spiezio Sardo et al. (2011). All cases could have received antibiotics and/or estrogen-based hormone therapy.

Figure 12.

Pooled results of the meta-analysis for the comparative effectiveness of biodegradable barriers versus no adjuvant for secondary prevention of adhesions following hysteroscopic adhesiolysis (RCT study design only). IUAs, intrauterine adhesions; RCT, randomized controlled trial; RR, risk ratio; τ, tau. Adjuvant indicates cases receiving a biodegradable barrier. Acunzo et al. (2003) utilized a traditional gel adjuvant, while Fernandez et al. (2024) positioned a hydrophilic polymer film within the endometrial cavity that expands in situ to form a biodegradable barrier. Control refers to cases receiving no adjuvant. All cases could have received antibiotics and/or estrogen-based hormone therapies.

Figure 13.

Pooled results of the meta-analysis for the recurrence of IUAs following hysteroscopic adhesiolysis with adjuvant therapy for secondary prevention of adhesions (RCT study design only). (A) Pooled recurrence of IUAs when using adjuvant therapies for secondary prevention of adhesions (RCT study design only). IUAs, intrauterine adhesions; RCT, randomized controlled trial; τ, tau. Events are cases where recurrent IUAs were identified at second-look hysteroscopy for each specified adjuvant category. Proportion refers to the risk of finding IUAs at second-look hysteroscopy by adjuvant type. Within the biodegradable barriers, Wang Y.Q. et al. (2020) and Acunzo et al. (2003) utilized a traditional gel adjuvant, while Fernandez et al. (2024) positioned a hydrophilic polymer film within the endometrial cavity that expands in situ to form a biodegradable barrier. All cases could have received antibiotics and/or estrogen-based hormone therapy. (B) Pooled recurrence of IUAs when using adjuvant therapies for secondary prevention with and without the concomitant use of systemic estrogens and progestins (RCT study design only). RR, risk ratio. Events on the left are cases where recurrent IUAs were identified at second-look hysteroscopy with the specified adjuvant along with systemic estrogen/progestin. Events on the right are cases where recurrent IUAs were identified at second-look hysteroscopy when the specified adjuvant was used without a systemic estrogen/progestin. The adjuvant used by Hanstede et al. (2023) was an inert (copper removed) ‘T-shaped’ intrauterine contraceptive device, while Yang et al. (2022b) used an intrauterine Foley balloon. (C) Pooled mean difference in the severity of recurrent IUAs when using adjuvant therapies with and without the concomitant use of systemic estrogens and progestins using the AFS classification system (RCT study design only). A, adjuvant; H, hormones—systemic estrogen/progestin; MD, mean difference; RCT, randomized controlled trial; τ, tau. The adjuvant used by Hanstede et al. (2023) was an inert ‘T-shaped’ intrauterine contraceptive device while Yang et al. (2022b) used an intrauterine Foley balloon.

Figure 14.

Pooled results of the meta-analysis for the comparative effectiveness of different adjuvant therapies for secondary prevention of adhesions following hysteroscopic adhesiolysis (RCT study design only). IUAs, intrauterine adhesions; RCT, randomized controlled trial; RR, risk ratio; τ, tau. Group A represents the first listed adjuvant while Group B represents the second listed adjuvant, e.g. in ‘Gel & Balloon vs Balloon’, Group A = Gel & Balloon and Group B = Balloon. Events are cases where recurrent IUAs were identified at second-look hysteroscopy for each specified category of adjuvant versus adjuvant. All cases could have received antibiotics and/or estrogen-based hormone therapy.

Figure 15.

Pooled results of the meta-analysis for the obstetrical outcomes in patients following hysteroscopic adhesiolysis with use of adjuvant therapies, all study types. IUAs, intrauterine adhesions; PAS, placenta accreta spectrum; τ, tau. Events are cases with each specified obstetrical outcome. Proportion refers to the risk of the specified outcome. Each outcome was evaluated as a proportion of all births. Studies utilizied various adjuvant therapies, and those that included multiple treatment arms were pooled within each study. All cases could have received antibiotics and/or estrogen-based hormone therapy.

Figure 16.

Pooled results of the meta-analysis for the obstetrical outcomes in patients following hysteroscopic adhesiolysis with no adjuvants, all study types. IUAs, intrauterine adhesions; PAS, placenta accreta spectrum; τ, tau. Events are cases with each specified obstetrical outcome. Proportion refers to the risk of the specified outcome. Each outcome was evaluated as a proportion of all births. Studies that included multiple treatment arms were pooled within each study. All cases could have received antibiotics and/or estrogen-based hormone therapy.

Figure 17.

Pooled results of the mean change in the proportion of patients attaining normal menstruation following hysteroscopic adhesiolysis (RCT study design only). IUAs, intrauterine adhesions; MRAW, raw mean; τ, tau.

Figure 18.

Pooled results of the mean change in the proportion of patients with recovery of any menstruation from a baseline of amenorrhea following hysteroscopic adhesiolysis (RCT study design only). IUAs, intrauterine adhesions; MRAW, raw mean; τ, tau.