Treatment of Cesarean Scar Ectopic Pregnancy in China with Uterine Artery Embolization—A Systematic Review and Meta-Analysis

Cesarean scar ectopic pregnancy (CSP) is a rare form of ectopic pregnancy, and treatment of CSP with uterine artery embolization (UAE) is a novel approach. With increasing numbers of cesarean sections being performed annually, the incidence of this condition is likely to increase. The authors became aware of an unusually high number of published studies originating in mainland China regarding this unusual treatment and sought to perform a meta-analysis to provide comprehensive evidence on this novel practice. Methods: We performed a thorough search and included all forms of quality studies on this topic that reported UAE as a part of first-line management of CSP. We included only studies originating in China. Ultimately, 37 studies were included for qualitative and quantitative synthesis of evidence. After screening retrieved records and extracting data from eligible studies, we pooled continuous data as a mean estimate and 95% confidence interval (CI), and dichotomous data as proportion and 95% CI. Results: CSP patients treated with protocols including UAE had a mean time of 30 days for serum β-hCG normalization, 95% CI [26.816, 33.881]. They had a mean estimated intraprocedural blood loss of 4.19 ± 3.76 mL, a mean hospital stay of nine days, 95%CI [7.914, 9.876], and a success rate of 93.4%, 95%CI [0.918, 0.951]. The severe complication rate was 1.2%, 95%CI [0.008, 0.017]. Conclusion: UAE, in combination with other procedures is being used effectively for the treatment of CSP in China. Protocols including UAE have a success rate of approximately 93.4%, and a severe complication rate of approximately 1.2%. This data’s utility is limited by vast differences in the studied protocols and questionable feasibility outside of China.


Introduction
Ectopic pregnancy describes pregnancies outside of normal positioning in the uterus, most frequently in the fallopian tube and less frequently in other sites such as the ovaries, abdomen, cesarean scar, and other sites [1]. The incidence of all ectopic pregnancies has increased in recent decades and complicates approximately 2% of all pregnancies, following the increase in the cesarean section rate [1,2].
Cesarean scar pregnancy (CSP) is an ectopic pregnancy located at a previous uterine scar [3]. Its incidence is increasing due to the increased frequency of cesarean sections worldwide [4,5]. It occurs in 1 in 500 pregnancies among women with a previous cesarean delivery and compromises 4% of all ectopic pregnancies [6]. Despite its rarity, CSP can constitute a life-threatening condition [7].
Largely used in the treatment of uterine fibroids, uterine artery embolization (UAE) is a widely used procedure generally performed by interventional radiologists under local anesthesia and carried out by catheterization of the uterine arteries through a transfemoral approach. The procedure involves injecting gelatin sponge particles to block the supplying arteries to the uterus, resulting in the cessation of blood supply to the CSP [11]. It may be combined with a dose of MTX given in the intraprocedural period [11,12]. Other authors have reported using polyvinyl alcohol instead of gelatin sponge particles with similar results [13].
UAE may be used alone or combined with local or systemic MTX for treatment of CSP. Moreover, it can be performed before uterine D&C, laparoscopy, hysteroscopy, or local resection [13][14][15][16].
The authors of this study noticed a tremendous increase in published trials that included the use of UAE as a treatment coming out of China. The authors hypothesize that this may likely be due to a decreased regulatory effect on medical care in this country, versus the majority of the rest of the world. As a result, a large body of research on the usage of UAE in CSP from China has surfaced over the last ten years. We aimed to present a global report on the usage of protocols including UAE for the treatment of CSP in China, by conducting this systematic review and meta-analysis.

Materials and Methods
We conducted this systematic review and meta-analysis guided by the Cochrane Handbook for Systematic Reviews of Interventions [17], then we reported it using the "preferred reporting items for systematic review and meta-analysis" (PRISMA statement) [18].

Eligibility Criteria and Study Selection
We included case series, observational studies, comparative studies, and randomized controlled trials (RCTs) that reported UAE as a part of first-line management of CSP, originating from anywhere within China. Exclusion criteria included: (1) Case reports or review articles, (2) case series describing less than five cases managed by UAE, (3) studies where treatment modality or outcomes were not sufficiently detailed, (4) non-English language studies (5), and in vitro or animal studies. After removing duplicates by Endnote, title, and abstract screening, the full-text screening ensured the studies' eligibility for inclusion. Moreover, we screened the bibliographies of the included studies manually for other relevant studies. Screening was performed independently by two separate authors, and agreement was reached by consensus between the authors. Per our institute standards, a third researcher was assigned to resolve any disputes but was ultimately never needed. Only three studies that met our criteria were excluded because they were located outside of China.

Data Extraction
Extracted data included the year of publication, study design, inclusion period, the mean age of participants, gestational age, primary treatment modality, number of cases per group, success rate, causes of treatment failure, rate of severe complications, time for serum β-hCG normalization, length of hospital stay, intraprocedural blood loss, number of cases undergoing hysterectomy or laparotomy, cases with bleeding more than 500 mL or received a blood transfusion, conclusion, and study country of origin. The management was considered successful if there was no major complication and the patient needed no additional treatments.
Severe complications included a UAE procedure that required hysterectomy, laparotomy, involved bleeding >500 mL, or necessitated an unexpected blood transfusion.

Quality Assessment
We used the national institute of health (NIH) tools to assess the quality of cohort and case series studies [19]. For RCTs, we used the Cochrane risk of bias tool described in the Cochrane Handbook for Systematic Reviews of Interventions [17].

Data Synthesis
Analysis was conducted using Open Meta-Analyst software. We reported dichotomous outcomes as a proportion and a 95% confidence interval (CI) and continuous outcomes as a mean estimate and a 95% CI. When heterogeneity was significant (Chi-square p < 0.1), we employed the random-effects model and then made a sensitivity analysis to solve the heterogeneity.

Literature Search Results
The literature search retrieved 433 records; of them, 109 duplicates were removed. We excluded 232 studies during the title and abstract screening and 57 during full-text screening. In addition to the remaining 35 studies, 2 studies were included through the manual search, and a total of 37 studies were included for qualitative and quantitative synthesis of evidence [11,[13][14][15][16] (Supplemental Figure S1). Interestingly, only three studies that would have otherwise met our screening criteria were excluded because they originated in countries other than China.

Characteristic of Included Studies
Included studies are variable in their design, including cohort studies, case series studies, and RCTs with a total of 2655 patients. The most frequent treatment modality in the included studies was UAE combined with D&C or UAE combined with MTX and D&C. Table 1 shows the summary of included studies and the characters of the included patients.

Quality Assessment
The quality of most included studies was fair. The final judgments of each study quality are shown in Table 1, and the details of each quality assessment domain are shown in Supplementary Tables S1 and S2. 3.4. Analysis of the Outcomes 3.4.1. Time for Serum β-Human Chorionic Gonadotropin (β-hCG) Normalization (Figure 1) The time for serum β-hCG normalization (defined as reaching a level of less than or equal to 5 mIU/mL,) was reported by 20 studies in 23 different study groups. The overall mean time for β-hCG resolution to normal level was 29.817 days; 95% CI [26.158, 33.476], and the analysis was heterogeneous (p < 0.001, I 2 = 99%).

Quality Assessment
The quality of most included studies was fair. The final judgments of each study quality are shown in Table 1, and the details of each quality assessment domain are shown in Supplementary Tables S1 and S2.

Analysis of the Outcomes
3.4.1. Time for Serum β-Human Chorionic Gonadotropin (β-hCG) Normalization (Figure 1) The time for serum β-hCG normalization (defined as reaching a level of less than or equal to 5 mIU/mL,) was reported by 20 studies in 23 different study groups. The overall mean time for β-hCG resolution to normal level was 29.817 days; 95% CI [26.158, 33.476], and the analysis was heterogeneous (p < 0.001, I 2 = 99%).

Hospital Stay (Figure 2)
Thirty-six groups in 30 studies reported about the duration of hospital stay following the UAE. The overall mean time of hospitalization was 9.044 days; 95% CI [8.028, 10.060], and the analysis was heterogeneous (p < 0.001, I 2 = 99%).

Quality Assessment
The quality of most included studies was fair. The final judgments of each study quality are shown in Table 1, and the details of each quality assessment domain are shown in Supplementary Tables S1 and S2.

Analysis of the Outcomes
3.4.1. Time for Serum β-Human Chorionic Gonadotropin (β-hCG) Normalization (Figure 1) The time for serum β-hCG normalization (defined as reaching a level of less than or equal to 5 mIU/mL,) was reported by 20 studies in 23 different study groups. The overall mean time for β-hCG resolution to normal level was 29.817 days; 95% CI [26.158, 33.476], and the analysis was heterogeneous (p < 0.001, I 2 = 99%).

Amount of Intraprocedural Blood Loss (Figure 3)
Twenty studies with 24 variable groups reported the amount of intraprocedural blood loss. The overall effect estimate of the intraprocedural amount of bleeding was 41.881 mL; 95% CI [34.102, 49.661], and the analysis was heterogeneous (p < 0.001, I 2 = 99%).

Amount of Intraprocedural Blood Loss (Figure 3)
Twenty studies with 24 variable groups reported the amount of intraprocedur blood loss. The overall effect estimate of the intraprocedural amount of bleeding wa 41.881 mL; 95% CI [34.102, 49.661], and the analysis was heterogeneous (p < 0.001, I 2 99%).

Amount of Intraprocedural Blood Loss (Figure 3)
Twenty studies with 24 variable groups reported the amount of intraprocedural blood loss. The overall effect estimate of the intraprocedural amount of bleeding was 41.881 mL; 95% CI [34.102, 49.661], and the analysis was heterogeneous (p < 0.001, I 2 = 99%).   All included studies reported a severe complication rate. The overall proportion of severe complication rate was 0.012; 95% CI [0.008, 0.017], and the analysis was homogenous (p = 0.127, I 2 = 19%).

Discussion
We analyzed data of 2655 CSP patients treated with UAE as part of first-line management. The results show that UAE was associated with a mean of 30.3 days for β-hCG normalization, a mean hospitalization time of 8.9 days, a mean intraprocedural blood loss of 41.9 mL, a success rate of 93.4%, and a severe complication rate of 1.2%.
Regarding the time for serum β-hCG normalization, it was found to be about 30.3 days. Qiao et al. [55] reported in their meta-analysis comparing adjuvant therapies to D&C, that UAE plus D&C had a shorter β-hCG normalization period than MTX plus D&C, which supports our results and shows that UAE may help decrease the β-hCG normalization time. Other studies have also shown a shorter normalization time after UAE compared with other treatments [28,30,36,51].
The mean amount of blood loss was 41.9 mL in our study. A recent systematic review and meta-analysis of MTX therapy for CSP reported that the mean blood loss was 76.3 mL [56], which strongly indicates that UAE probably helps decrease bleeding. Other studies also reported that the addition of UAE to the treatment protocol led to less bleeding [11,14,54,57,58].
In this study, UAE was associated with a mean post procedural hospital stay of 8.9 days. A recent meta-analysis of MTX for CSP showed an average stay of 11.7 days when MTX was used on an inpatient basis as a solo agent for CSP [56]. Another study found that UAE followed by D&C had significantly less hospital stay than MTX plus D&C [55].
The success rate of UAE in our study was 93.4%, which means that about 2432 patients managed with UAE as a part of their therapy needed no additional follow-up treatments. The success rate for MTX combined treatment was lower and equaling 90.7% [56].

Discussion
We analyzed data of 2655 CSP patients treated with UAE as part of first-line management. The results show that UAE was associated with a mean of 30.3 days for β-hCG normalization, a mean hospitalization time of 8.9 days, a mean intraprocedural blood loss of 41.9 mL, a success rate of 93.4%, and a severe complication rate of 1.2%.
Regarding the time for serum β-hCG normalization, it was found to be about 30.3 days. Qiao et al. [55] reported in their meta-analysis comparing adjuvant therapies to D&C, that UAE plus D&C had a shorter β-hCG normalization period than MTX plus D&C, which supports our results and shows that UAE may help decrease the β-hCG normalization time. Other studies have also shown a shorter normalization time after UAE compared with other treatments [28,30,36,51].
The mean amount of blood loss was 41.9 mL in our study. A recent systematic review and meta-analysis of MTX therapy for CSP reported that the mean blood loss was 76.3 mL [56], which strongly indicates that UAE probably helps decrease bleeding. Other studies also reported that the addition of UAE to the treatment protocol led to less bleeding [11,14,54,57,58].
In this study, UAE was associated with a mean post procedural hospital stay of 8.9 days. A recent meta-analysis of MTX for CSP showed an average stay of 11.7 days when MTX was used on an inpatient basis as a solo agent for CSP [56]. Another study found that UAE followed by D&C had significantly less hospital stay than MTX plus D&C [55].
The success rate of UAE in our study was 93.4%, which means that about 2432 patients managed with UAE as a part of their therapy needed no additional follow-up treatments. The success rate for MTX combined treatment was lower and equaling 90.7% [56].
Fifty-five patients (1.2%) managed by UAE reported events of severe complications. Forty-three of them suffered bleeding more than 500 mL or received a blood transfusion, fifteen underwent a hysterectomy, and fourteen had laparotomy. Petersen et al. reported a severe complication rate in different modalities, reaching 3.4% in UAE plus D&C, 1.2% in UAE plus D&C and hysteroscopy, and 2.8% in UAE plus D&C and MTX, and higher proportions for other modalities devoid of UAE [59].

Limitations
Despite being the first study to group all available research on UAE in China as part of CSP treatment, our study has many limitations. The wide variety of treatment options used in the included studies makes summarizing the results challenging and prohibits any meaningful combination of protocols to show that any one regiment is superior or even notably efficacious. Therefore, our recommendations are somewhat limited as we are combining many different protocols, all of which included UAE as part of the primary treatment of CSP. The small sample size for some studies and heterogeneity in the results are also major limitations. Most studies were of fair rating of quality, so future research is needed with more structured study designs and a larger scale of participants to ensure the effectiveness of UAE in CSP treatment. Moreover, as our analysis looked at this novel technique's usage only in China, some of our findings may not be applicable to other countries. Notable is the fact that many consider China's outpatient care system to still be developing, which likely accounts for the very long inpatient stays associated with these studies. This would not be expected in most developed countries with robust outpatient care systems, and further limits the versatility of this data.

Strengths
This was the first meta-analysis to look at all the available research on the use of UAE for CSP coming from China, and we were able to find a very wide breadth of studies to include. Therefore, although the usefulness of this data in the rest of the world may be limited, we provide researchers and physicians outside of China a first look at a novel practice that is likely very foreign to their own modes of practice. In addition, we used strict adherence to PRISMA guidelines, and were able to solve heterogeneity in most situations. Many gynecologists around the world may not previously have been aware that these novel techniques were being used in China.

Conclusions
China is producing a large amount of literature on the novel usage of UAE in the treatment of CSP. Although our study was limited by including many variations in protocols and modalities that were included with UAE in the treatment of CSP, we were able to calculate an overall success rate of approximately 93.4%, and a severe complication rate of approximately 1.2%. Because of conditions unique to the healthcare system in China, this data may have limited utility in application to patient care in other countries. More high-quality trials will be needed to further elucidate which exact treatment combinations and protocols yield the safest and most efficacious results for patients.

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/jcm11247393/s1, Figure S1: PRISMA flow chart summarizing the process of study selection; Table S1: The Risk of Bias for all included Cohort studies (by NIH tool); Table S2: Risk of Bias for all included Case series studies (by NIH tool).
Author Contributions: All authors attest to significant contributions to this work. At a minimum, G.J.M. and A.T.M. were responsible for conception and initial draft; C.C., H.U., A.M. and J.P. were responsible for data analysis, results, and discussion writing; A.A., S.G., C.M., A.M. and M.G. were responsible for data collection and the final draft. Many of the listed authors made other contributions to other parts of this work. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.

Institutional Review Board Statement:
This Manuscript has been reviewed by the institutional IRB board at Marchand Institute and was found to be exempt from IRB review. (May 2021). Data used was exempt from consent to participate or publish secondary to the nature of the study being a systematic review, retrospectively looking at previously published data.
Informed Consent Statement: Not applicable to systematic review. Data Availability Statement: All supporting data is included or referenced in this manuscript. The authors have no additional data used in this study.