Treatments for cesarean scar pregnancy: 11-year experience at a medical center

Abstract Background Cesarean scar pregnancy (CSP) is a long-term complication after cesarean section that can cause severe maternal morbidity and mortality. Although a variety of treatments have been described, there is no consensus as to the optimal management approach. Many grading systems for CSP have been proposed, among which the classification made by the consensus of Chinese experts in 2016 was shown to provide improved treatment guidance for clinical practice. The purpose of the present study was to analyze the success rate of different treatments for each type of CSP as classified according to the Chinese Expert’s Consensus (2016), and to develop a management strategy for CSP. Methods A retrospective study was performed among patients diagnosed with CSP at Shandong Provincial Hospital between January 2009 and December 2019. We reviewed clinical characteristics, treatment methods, and subsequent outcomes; and analyzed these endpoints using the statistical software package SPSS 22.0 (SPSS, Inc., Chicago, IL). Results For type I CSP, systemic methotrexate (MTX) administration exhibited a success rate of 79.2% for type Ia and 14.3% for type Ib. Local and systemic MTX administration success rates were 88.9% for type Ia and 66.7% for type Ib. Dilation and curettage (D&C), curettage after uterine artery embolization (UAE + C), and hysteroscopic curettage (H + C) were 100% successful. For type II, UAE + C, H + C, and laparoscopy combined with hysteroscopic curettage (L + H+C) were 100% successful. D&C had a success rate of 97.0% for type IIa and 88.9% for type IIb. The success rate of systemic MTX administration was 52.0% for type IIa and 62.5% for type IIb. Both UAE + C and L + H+C had 100% success rates for type IIIa CSPs, while for type IIIb, the success rate was 87.9% for UAE + C vs. 96.6% for L + H+C. Conclusions For type I CSPs, D&C was quick, easy, and safe; for type II, H + C was more suitable. For type III and some type II patients who wished to undergo simultaneous repair of the cesarean defect, L + H+C was the optimal method. UAE can be used as a complementary option instead of a prophylactic measure, and when difficulties with endoscopic surgeries were encountered, conversion to laparotomy was the ultimate treatment.


Introduction
A cesarean scar pregnancy (CSP) is an ectopic pregnancy in which an embryo implants in a previous cesarean scar. CSP is a long-term complication after cesarean sections and can cause severe maternal morbidity and mortality [1,2]. CSP was once considered a rare ectopic pregnancy occurring in 0.15% of women with previous cesarean deliveries [3]. However, during the past decade, the incidence of CSP has increased dramatically, especially in China, which might be attributable to the rising rate of cesarean section [4].
Although a variety of treatments have been described, there is no consensus on the optimal management approach [5][6][7][8][9][10]. Many grading systems for CSP have been proposed [11][12][13][14]. Viel et al. classified CSP into endogenic and exogenic types. The endogenic type occurs when the CSP progresses into the cervico-isthmic space and the uterine cavity. The exogenic type denotes a deep implantation of CSP with progression toward the overlying myometrium. This classification provides clues regarding the prognosis of CSP. The consensus of Chinese experts categorized CSP into three types based on the gestational sac (GS) location and remaining myometrial thickness [15]. This classification was shown to provide a better treatment option for different types of CSP.
In this study, we categorized our patients into three types and further categorized each type according to the Chinese Expert's Consensus (2016) [15] into a and b subtypes based on the size of the GS. We assessed the success rate of different treatments for each type and tried to develop a management strategy for CSP. The study flow is shown in Supplementary Figure 1.

Patients
We performed a retrospective study and reviewed the patients who visited Shandong Provincial Hospital in China between January 2009 and December 2019. In obtaining informed consent from individual or guardian participants, the maternal age, cesarean frequency, gestational age, the diameter of the GS, myometrial thickness, time of interval between last pregnancy and CSP and b-HCG level were collected by reviewing the medical records. More details are provided in Supplementary Table 1. This study was IRB approved (study number SWYX2020-100) by the review board of Shandong Provincial Hospital.

Diagnosis
The diagnosis of CSP mainly depends on ultrasonographic and clinical presentations. The condition of CSP was confirmed when ultrasonographic images met the following criteria: (1) no GS present in the uterine cavity and cervical canal; (2) GS located in the anterior isthmus of the uterus with or without cardiac activity; (3) myometrial defect or absence between the bladder and the GS [16,17]. All ultrasound examinations were performed by sonographers of the gynecologic ultrasound group and audited by an expert sonographer, which excludes potential interobserver bias.
Classification CSP was classified into three types according to the Chinese Expert's Consensus (2016) [15] and the sonographic findings of patients (Supplementary Figure 2). Type I is the endogenous type, in which the GS progresses into the cervico-isthmic space or uterine cavity, and the thickness of the myometrial wall between the sac and the bladder is >3 mm. Type I is further divided into two subtypes by the size of the GS. Type Ia refers to those with a GS 3 cm. Type Ib refers to those with a GS >3 cm.
Type II is also the endogenous type, in which the thickness of the myometrial wall between the sac and the bladder is 3 mm. Type IIa refers to those with a GS 3 cm. Type IIb refers to those with a GS >3 cm.
Type III is the exogenous type, which denotes deep invasion of the scar defect with a progression toward the bladder and abdominal cavity, in which the myometrium between the bladder and sac is <3 mm or is missing entirely. Abundant vascularization is present within the implantation site. Type IIIa refers to cases with a GS 3 cm. Type IIIb refers to those with a GS >3 cm.

Treatments
The treatment approach of CSP was determined based on a patient's condition, physician expertise, and the timely availability of the technique, including administration of methotrexate (MTX), dilation and curettage (D&C), curettage after uterine artery embolization (UAE þ C), hysteroscopic curettage (H þ C), or laparoscopy combined with hysteroscopic curettage (L þ HþC). In addition, the treatment of CSP also takes the fertility needs of patients into consideration and maintains the patient's fertility as much as possible.
MTX was administered to patients systemically or combined with local administration. Systemic MTX was given intramuscularly as a single dose of 50 mg/m 2 . Local injection of MTX was administered into the GS under ultrasonographic guidance at a dose of 1 mg/kg. D&C was performed by an experienced surgeon with ultrasonographic guidance. A Foley catheter was placed into the uterine cavity for compression hemostasis when necessary.
Patients who underwent UAE were transferred to the interventional radiology center for embolization. Super-elective UAE was performed by two experienced radiologists using gelatin sponge powder. Uterine curettage was carried out 24-48 h later.
The procedure encompassing H þ C was as follows: first, hysteroscopy was performed to explore the uterine cavity and determine the exact location of the conceptus. The GS was then removed by vacuum aspiration. Subsequently, the uterine diverticulum was examined carefully. The implanted chorionic villi were hysteroscopically resected. The blood vessels in the niche of the scar were electrocoagulated.
L þ HþC was an option when the pregnancy protruded into the abdominal cavity or bladder. In this approach, laparoscopy was arranged to incise the uterovesical flexion and expose the CSP mass. Hysteroscopic curettage was done under the laparoscopic monitoring. The lesion, as well as the cesarean scar defect, was resected, and then the lower uterine segment was reconstructed laparoscopically.
On the 7th day after the first MTX injection, we evaluated serum b-hCG concentrations [18]. For surgical procedures, serum b-hCG levels were measured, and ultrasonography was repeated on the third day after surgery.

Success rate of different treatments for each CSP type
Success of the primary treatment was defined as the return of b-hCG to normal levels without further surgical treatment [19]. The treatment modality, as well as the success rate of each subtype, was listed and analyzed.

Statistical analysis
We used the statistical software package SPSS 22.0 (SPSS, Inc., Chicago, IL) for data analysis. Data were presented as means ± standard deviation (SD), median, frequencies (n), or compositional ratios (%) as appropriate. Data were analyzed as to whether they followed a normal distribution by the Shapiro-Wilk test. If variables met the assumptions of normality, Student's t-test was used; otherwise, the nonparametric Mann-Whitney U-test was applied. Constituent ratio data were evaluated by Fisher's exact-probability test. We considered a p value <.05 to be statistically significant.

Results
A total of 725 patients with CSP were included during the study period. As shown in Supplementary Figure  3A, only two patients were diagnosed in 2009. By 2015, 79 patients were confirmed. In 2016, the number of CSP patients was 124, showing a significant surge in this condition. The incidence of CSP has continuously increased since 2016. Among them, 162 cases (22.4%) were diagnosed with type I, 465 (64.1%) cases were diagnosed with type, and 98 (13.5%) were diagnosed with type III (Supplementary Figure 3B).
In the subgroup analysis, which compared five different methods of CSP. Administration of MTX was initially practiced in our hospital, serving as one of the main approaches for approximately 11 years. However, two years after the initiation of the medical treatment, an increasing number of patients received surgical management. As such, after 2011, more patients received surgical management compared to MTX administration. From 2011 to 2016, the main surgeries were D&C and UAE þ C. The later had the most cases and reached its peak in 2016. With increasing utilization of H þ C, the application of D&C and UAE þ C declined. The number of cases with the option of H þ C had been on the rise since the time it was introduced in practice. L þ HþC was not as popular as H þ C, however, its application increased greatly in the last year (shown in Figure 1).
Success rates of the treatments are shown in Table  1 Furthermore, we compared UAE þ C and H þ C in the treatment of type II CSPs. The clinical information and outcomes are listed in Table 2. Patients who were treated with H þ C had less blood loss and shorter hospitalization periods (p< .05). We also compared UAE þ C and L þ HþC in the treatment of type IIIb CSPs. The post-treatment reduction of b-hCG was more pronounced and blood loss was higher in the L þ HþC group (p< .05, Table 3).

Discussion
Due to recent increases in the cesarean section rate worldwide, there has been a commensurate increase in the prevalence of CSPs. Our data showed that there was a surge of CSP in the year 2016, which might be association with the implementation of the 2-child policy in China. During the past 11 years, 725 cases were included in our study.
Of all the medically treated cases, including systemic-only and local combined with systemic administration of MTX, 40.7% required a subsequent surgical intervention, which was similar to data reviewed by Maymon et al. [20]. In practice, we found that patients pretreated with MTX were more likely to develop persistent scar pregnancy and had a higher risk of bleeding during the following surgery. Therefore, medical    treatment is not presently used as the primary treatment method in our department. Reports involving curettage in the treatment of CSP have shown varied results, with some acceptable and successful [21,22], and others with ineffective and insufficient outcomes [23]. There are also reports of massive hemorrhage during or after the curettage [24]. In our study, D&C was safe with type I CSP regardless of the size of the GS. However, D&C is insufficient for type II patients, especially type IIb, and it is dangerous if used in type III CSP. Both type IIIb patients who underwent D&C as the primary treatment developed excessive bleeding and had an emergency hysterectomy.
Concomitant use of UAE associated with a reduced risk of subsequent treatments and augments the success rate [25]. The success rates of this method were 100% in type I, type II, and type IIIa patients. Only four patients of type IIIb received subsequent surgical intervention due to persistent CSP after initial treatment. However, the cost of UAE was high. The occasional discomfort and pelvic pain caused by the ischemia after UAE, concerns over the possibility of endometrial atrophy, and impairment of ovarian function have made both patients and physicians hesitant to choose this option [26,27]. UAE has gradually become a complementary measure in patients with heavy bleeding that has occurred after initial treatment, and when a laparotomy was not indicated.
Hysteroscopy can be used to examine the uterine cavity and uncover intrauterine lesions, as well as to assess bleeding under direct visualization [28]. However, we found that bleeding during the procedure could impair vision and interfere with the manipulation. Subsequently, we modified our procedures to include both hysteroscopy and curettage. Hysteroscopy was performed before curettage to assess intrauterine conditions and to clear up the residual gestational tissue after curettage. In addition, electroexcision was used in patients with implanted chorionic villi to clear the abnormal vessels in the niche of the scar. In our study, the success rate for this treatment was 100% in type I and type II patients. Three out of four patients of type IIIa had successful procedures. Only one patient of type IIIb used this method and failed. In further comparison with UAE þ C for the treatment of type II patients, we found that blood loss was lessened with the same success rate, and that hospitalization was shorter in the H þ C group. Pre-UAE was thus deemed unnecessary.
The above treatments for type III CSP were associated with a high risk of massive hemorrhage, uterine perforation, bladder injury, and persistence of trophoblasts [29]. In our study, UAE þ C and L þ HþC had much higher success rates for type III patients. Additionally, in the L þ HþC group, the level of b-hCG after treatment was even lower. Laparoscopic surgery has the advantages of removing the protruding pregnancy tissue completely while simultaneously reconstructing the lower uterine segment [30]. We performed laparoscopic surgery assisted by hysteroscopy and coupled with curettage for type III CSP and some type II patients who required repair of the cesarean section defect. Combined visualizations by hysteroscopy and laparoscopy have thus provided an excellent view of the extant pathology and facilitated an exact diagnosis. Except for one patient who exhibited extensive pelvic adhesions accompanied by serious anemia and whose procedure was converted to laparotomy, no other patients required further intervention. The blood loss of the L þ HþC group was higher than in the UAE þ C group, which may be related to the procedure of laparoscopic resection of gestational tissue with abundant blood supply. In the process of L þ HþC surgery, we choose to inject dilute vasopressin solution (1 unit/mL, 5-10 mL) into the cervix in advance, then perform H þ C under laparoscopic monitoring, and continuous uterine Foley balloon compression in the process of laparoscopic uterine repair to reduce bleeding. In the future, we should focus more on the strategy of preventing intraoperative bleeding. The strength of this study was the relatively large sample size of 725 cases. Through analysis of the data, we offer a clear algorithm for which treatment to use based on the type of CSP. The approaches in our strategy should be feasible at most medical centers and easily carried out.
A limitation to the current study was its retrospective nature, and there may have been a possibility of selection bias. As this was an experience at a single center, the strategy did not include all the effective treatments reported in the literature. Thus, a prospective, multicenter, randomized controlled trial still needs to be performed to reach more definitive conclusions.

Conclusions
We recommend that patient management be individualized according to differences in patient characteristics. However, timely detection and appropriate treatment are critical in achieving the most successful outcome. In our experience, administration of MTX is not recommended as the primary treatment or pretreatment. For type I CSPs, D&C is quick, easy, and safe, whereas for type II, H þ C is more suitable. For those who want to simultaneously repair the defect of the cesarean scar, laparoscopy should be performed at the same time. As for type III, although the sample was small, it showed an association of poor prognosis with MTX treatment for type III CSP, and L þ HþC was the optimal method. UAE can be used as a complementary option in the case of massive bleeding, rather than as a simple prophylactic measure. When complex adhesions or uncontrollable bleeding were encountered during endoscopic surgery, conversion to laparotomy was the ultimate treatment ( Figure 2).