Effect of prophylactic interventional therapy of internal iliac artery balloon occlusion in pregnancies complicated by placenta previa and accreta: a retrospective cohort study

Background: Placenta previa and accreta are serious obstetric conditions that are associated with a high risk of intraoperative massive hemorrhage, the prophylactic intravascular balloon occlusion technique is increasingly used in managing uncontrolled hemorrhage in cesarean section (CS). We aim to examine the clinical effectiveness of prophylactic interventional therapy of the internal iliac artery balloon occlusion (PBOIIA) during CS in improving maternal outcomes for patients with placenta previa and accreta. Methods: A total of 420 women with placenta previa and accreta who underwent CS from January 2014 to December 2018 were included retrospectively. Patients were divided into balloon group in which patients had PBOIIA (n=248) and the control group in which patients did not have PBOIIA (n=172). Meanwhile, we performed a subgroup analysis in whether taking PTUI surgery. Information on conditions of patients and newborns, perioperative blood indicators, surgical outcomes were collected. Results: Median estimated blood loss (mEBL) was 2200mL in the balloon group and 2150mL in the control group respectively, there was no signicant difference between two-groups comparison (P (cid:0) 0.05). The rate of patients with hysterectomy (36.3% verus 35.5%) and amount of PRBCs transfused [(0-31.5) verus 3 (0-39)] were not different between two groups (P (cid:0) 0.05). However, the total hospitalization costs (7044.4±1708.0 verus 5793.6±2263.8) and surgery costs (3074.2±404.9 verus 1829.7±485.8) in balloon group were signicantly higher than those in control group (P<0.05). Subgroup analysis showed PTUI surgery had no signicant differences in EBL ((P (cid:0) 0.05), but it could signicantly decrease hysterectomy rates (P (cid:0) 0.05). Conclusions: PBOIIA has no signicant effect on reducing intraoperative EBL and hysterectomy to evaluate the effectiveness and practicality of PBOIIA in improving maternal outcomes for patients with placenta previa and accreta.


Background
Placenta previa and accreta associated with previous cesarean section are related to severe adverse maternal-fetal pregnancy outcomes and are even accountable for a high risk of maternal death [1]. These placental conditions can cause disseminated intravascular coagulation, shock, and a high rate of hysterectomy [2]. Placenta previa and placenta accreta are the major causes of postpartum hemorrhage, and is currently a leading cause of maternal death worldwide [3]. Massive hemorrhage during cesarean section (CS), which is hard to predict and control, is the main threaten to the life of patients with placenta previa and accreta. Cesarean hysterectomy is an important treatment for placenta previa and accreta, but it should be performed with caution [4]. Recently, a lot of conservative management has been conducted to reduce intraoperative hemorrhage and the hysterectomy rate, and to ensure maternal and newborn safety during CS. The prophylactic intravascular balloon occlusion technique is increasingly used in managing uncontrolled hemorrhage in CS. This technique includes prophylactic intravascular balloon occlusion of the internal iliac arteries (PBOIIA) and abdominal aorta arteries (PBOAA) [5].
Owing to the low risk of vascular complications [6], PBOIIA has been used in our hospital for patients with placenta previa and accreta since 2014. The effectiveness of PBOIIA remains controversial because of the inconsistent results of different research.
Deng, Wang et al reported that PBOIIA was an effective method of hemostasis in CS. However, Huang, Salim, Liu et al showed that PBOIIA had no bene t in reducing estimated blood loss (EBL) and improving maternal outcomes for patients with placenta previa and accreta. There is still a lack of large-sample studies on the effectiveness of internal iliac artery balloons. Therefore, we performed a large-sample, retrospective cohort study and aimed to evaluate the effectiveness and practicality of PBOIIA in improving maternal outcomes for patients with placenta previa and accreta.

Study population
The retrospective, observational study was approved by the Ethics Review Committee of West China Second University Hospital of Sichuan University, and because of the nature of the retrospective, observational setting for our study, and the data are anonymous, an informed consent was waived for all included patients by the Ethics Review Committee of West China Second University Hospital. All study methods of the retrospective study were conducted following the relevant regulations of a protocol approved by the Institutional Review Board from the West China Second University Hospital of Sichuan University. This study was conducted from January 2014 to December 2017. Placenta previa occurred when the placenta was wholly or partially implanted in the lower uterine segment. Placenta accreta was de ned as the situation where the placental trophoblast invaded into the myometrium, according to the depth of villous tissue invasiveness. Placenta accreta has been subdivided by modern pathologists into "creta" or "adherenta". An adherent placenta is where the villi adhere super cially to the myometrium without interposing the decidua. Placenta increta is where the villi penetrate deeply into the uterine myometrium down to the serosa. Placenta percreta is where the villous tissue perforates through the entire uterine wall and may invade the surrounding pelvic organs, such as the bladder [7,8]. All included patients had at least one prior CS and were diagnosed with placenta previa or placenta accreta when the placenta covered a previous cesarean scar by color Doppler ultrasonography and pelvic magnetic resonance imaging (MRI) examinations before delivery. The diagnosis was con rmed by intraoperative ndings or histopathological examination after surgery. Patients with an adherent placenta were excluded in this study because balloon occlusion was not routinely used for these patients in our clinical work. Therefore, we only included patients with placenta increta or percreta. Patients with serious medical and surgical diseases, incomplete data, multiple pregnancies, or those who delivered before 28 weeks of gestation were excluded.
According to the above-mentioned inclusion and exclusion criteria, of 713 patients with placenta previa and accreta associated with previous CS who delivered in our hospital, 420 were included in this study ( Figure 1). The 420 patients were divided into two groups according to whether they had PBOIIA (n=248, 59.0%) (balloon group) and whether they did not have PBOIIA (n=172, 41.0%) (control group). In 2017, doctors in our hospital investigated a novel approach called parallel transverse uterine incision (PTUI) surgery. PTUI had a signi cant effect on reducing intraoperative blood loss and the hysterectomy rate for patients with placenta previa and accreta [9,10]. Among our included patients, we found that some patients underwent PTUI surgery simultaneously. To avoid the effect of PTUI surgery on the results, we conducted a subgroup analysis on the two groups of patients according to whether they had PTUI surgery during CS. Among the 420 patients, 86 had PTUI surgery and 334 did not have PTUI surgery. In the PTUI group, the 86 women were subdivided into the balloon group that had PBOIIA (group A1; n=58) and the control group that did not have PBOIIA (group B1; n=28). In the non-PTUI group, the 334 women were subdivided into the balloon group (group A2; n=190) and the control group (group B2; n=144).
Comprehensive management of patients with placenta previa and accreta required the cooperation of a multidisciplinary medical team. All patients underwent scheduled CS at 35-36+6 weeks of gestation [11]. Fluid transfusion, blood transfusion, strong uterine contraction drugs, and conservative surgical treatments were provided to patients during CS. If the conservative treatments were not effective, hysterectomy was performed to save the patient's life. Since the effectiveness of PBOIIA in patients with placenta previa and accreta remains controversial and it is not de nitively reliable, and there are complications associated with PBOIIA. The decision of preoperative prophylactic placement of balloon occlusion was made on a case by case basis through preoperative multi-department discussion before surgery. This decision was jointly made by the surgeon and radiologist after evaluating the condition of patients in our hospital. All patients who accepted PBOIIA provided written informed consent, and balloon occlusion of the right and left internal iliac arteries was in ated in all cases.

Placement of the occlusion balloon
In the balloon group, all included pregnant women were fully informed of the bene ts and complications of PBOIIA by their doctors before CS. For the surgical procedure, after routine disinfection and laying of towels, bilateral femoral arteries were punctured by the Seldinger technique and a 5-French vascular sheath was inserted. A 5-French Cobra and 0.035-inch guidewire were used to guide the balloon catheter into the bilateral internal iliac artery through the vascular sheath, with the balloon catheter tip slightly above the opening of the bilateral uterine artery (low-pro le PTA balloon dilatation catheter PTA5-35-80-8-6.0; Cook Medical Inc., Bloomington, IN, USA). An indwelling catheter was xed in both lower limbs and the patients were brought into the operating room.
During the operation, approximately 2 mL of eunepiac, which is a diluent contrast agent, was injected to temporarily in ate the internal iliac artery balloon after the fetus was delivered and the umbilical cord was cut. This diluent contrast agent can block blood ow of bilateral internal iliac arteries and reduce the amount of uterine bleeding [12] (Figure 2).

Clinical characteristics and outcomes
The clinical characteristics of the included patients were collected by reviewing the medical records. The clinical indicators were retrospectively collected by two obstetricians, and if any discrepancy existed, it was resolved by a third obstetrician. All reviewers were blinded to the data of patients. The clinical indicators included maternal and neonatal characteristics, perioperative blood indicators, and surgical outcomes during hospitalization. Maternal and neonatal characteristics included maternal age, gravidity, parity, number of previous CSs and abortions, gestational age at delivery, body mass index (BMI), birth weight of the newborn, 1min Apgar score, neonatal intensive care unit (ICU) admission, neonatal asphyxia, and neonatal death. Perioperative blood indicators included hemoglobin (HGB), hematocrit, platelets, prothrombin time, activated partial thromboplastin time, and brinogen. Surgical outcomes included intraoperative EBL, cesarean hysterectomy, packed red blood cell (PRBC) transfusion (in China, one unit of PRBC is approximately equal to 200 mL of whole blood), fresh frozen plasma transfusion, platelet transfusion, the volume of autologous blood transfusion, blood loss within 24 hours postoperatively, maternal ICU admission rate, postoperative pyrexia (≥38.5°C), anemia (HGB <100 g/L), total hospitalization costs, and surgery costs.

Statistical analysis
Categorical variables are presented as number/proportion (%) and were analyzed by the chi-square test. The Kolmogorov-Smirnov test was performed to determine normality of continuous variables. Data are shown as mean ± standard for normally distributed variables. If variables were normally distributed, the independent t-test was used for analysis. The Mann-Whitney U test was used for the analysis of non-normally distributed data and data are shown as the median (range). All statistical analyses and data processing were conducted by SPSS18.0 statistical software (IBM, Armonk, NY, USA). P values and 95% con dence intervals are shown for assessment of clinical indictors of patients with placenta previa and accreta. P<0.05 indicates that the difference was signi cant.

Comparison of baseline characteristics between the balloon and control groups
The mean maternal age in the balloon group and control group was 32.48±5.09 and 32.70±4.72 years, respectively, the mean values of previous CSs were 1 (1-3) and 1 (1-3) in the balloon group and control group, and the mean gestational age at delivery in the balloon and control group were 36+3(28+3-40+4) weeks and 36+2(28+1-39+3) weeks respectively. There were no signi cant differences in maternal basic characteristics, such as maternal age, gravidity, parity, number of previous CSs and abortions, gestational age at delivery, and body mass index, between the balloon and control groups, which indicated that the two groups were comparable. With regards to neonatal characteristics, the newborn weight of the balloon group was signi cantly higher and the rate of neonatal ICU admission was lower than those in the control group (both P<0.05). There were no differences in the neonatal asphyxia rate, neonatal death rate, and 1-min Apgar score between the balloon and control groups ( Table 1).

Comparison of perioperative blood indicators between the balloon and control groups
There were no signi cant differences in preoperative blood indicators and postoperative blood indicators, including HGB, hematocrit, platelets, prothrombin time, activated partial thromboplastin time, and brinogen, between the balloon and control groups (P 0.05) ( Table 2).

Comparison of surgical outcomes between the balloon and control groups
Median blood loss and the cesarean hysterectomy rate were not signi cantly different between the balloon and control groups, the median blood loss was 2200 mL (range, 500-12000) in the balloon group and 2150 mL (range, 500-15000) in the control group, and in balloon group, 36.3% patients with cesarean hysterectomy versus 35.5% in control group, which were no signi cant difference (P 0.05). The amount of PRBC transfusion and the rate of maternal ICU admission was signi cantly lower in the balloon group than in the control group (P<0.05). The total hospitalization and surgery costs in the balloon group were signi cantly higher than those in the control group (7044.4±1708.0 verus 5793.6±2263.8 USD and 3074.2±404.9 verus 1829.7±485.8, both P<0.05). However, there were no signi cant differences in fresh frozen plasma transfusion, volume of autologous blood transfusion, blood loss within 24 h postoperatively, postoperative pyrexia (≥38.5°C), and anemia (HGB <100 g/L) between the balloon and control groups (P 0.05) ( Table 3).

Discussion
Placenta previa and accreta after previous CS are extremely serious obstetric conditions associated with a high risk of intraoperative massive hemorrhage. These conditions dramatically increase the risk of blood transfusion and hysterectomy during CS [13,14]. Placenta previa and placenta accreta are the two major risk factors for postpartum hemorrhage [15]. The incidence of placenta previa after CS is 1.22% [16]. However, prior CS appears to be the major risk factor for placenta accreta. Previous studies reported that the incidence of placenta accreta was 3.3%, 11%, 40%, 61%, 67%, and 67% after rst, second, third, fourth, fth, and sixth or more cesarean deliveries, respectively [11,17]. Owing to the high morbidity associated with placenta previa and accreta, accurate preoperative diagnosis and a multidisciplinary medical team for management of these conditions play a vital role [4,18].
Prenatal ultrasound and MRI techniques have been used to diagnose and guide clinical management, and favorable outcomes have been achieved [19,20]. With the wide use of ultrasound and MRI in the medical eld, abnormal placenta accreta can be diagnosed in advance, and a plan can be provided for reducing intraoperative blood loss. Surgical hemostasis and uterine contractile agents are used to manage intraoperative hemorrhage and preserve the uterus during CS. If all methods of hemostasis fail to control the bleeding, hysterectomy is the ultimate solution for patients with placenta previa and accrete [21].
In recent years, an increasing amount of hemostasis methods have been applied in CS, such as the prophylactic intravascular balloon occlusion technique. As early as 1997, PBOIIA was rst reported as a hemostatic method in CS for patients with placenta percreta [22]. With the increasing use of internal iliac artery balloon occlusion in CS for treating placenta previa and accreta, the role of PBOIIA in managing uncontrolled hemorrhage has become important for obstetricians and anesthetists. However, the e cacy of PBOIIA reported in the literature is controversial. Some scholars reported that PBOIIA was an effective hemostasis method in CS. Deng et al conducted a prospective observational study, which included 163 patients with placenta previa and accreta [21]. These authors found that PBOIIA was an effective strategy for controlling severe hemorrhage, and it reduced the amount of intraoperative blood loss for patients with placenta previa-accreta (1236.0 vs 1694.0 mL). Picel et al reported that, in 151 patients with invasive placenta undergoing cesarean hysterectomy, there was a signi cant difference in blood loss (2000 vs 2500 mL) and PRBC transfusions (2 vs 5 U) in the balloon group compared with the control group [23]. These ndings indicated that PBOIIA could reduce intraoperative blood loss and transfusion requirements. Wang et al studied 83 patients with pernicious placenta previa coexisting with placenta accreta and found that PBOIIA was an effective method for managing postpartum hemorrhage [24]. However, some researchers failed to show that PBOIIA improves maternal outcomes. Previous randomized, controlled trials included 13 patients who were diagnosed with placenta accreta in the intervention group and 14 cases in the control group [25,26]. There were no signi cant differences in calculated blood loss (4950 vs 4709 mL) and PRBC units transfused (5.2 vs 4.1 U) between the intervention and control groups. These ndings suggested that PBOIIA had no bene t in patients with placenta accreta. Liu et al compared 83 patients with placenta previa and accreta who underwent cesarean hysterectomy in the balloon group and 31 patients in the control group [27]. They found that PBOIIA had no signi cant effects on reducing EBL (3000 vs 3700 mL) and improving maternal outcomes.
A great deal of studies have investigated the effect of PBOIIA, but the sample sizes were small and the power of the studies was limited. We conducted a large-sample study to evaluate the effectiveness and practicality of PBOIIA for patients with placenta previa and accreta. In our study, we included 420 patients with placenta previa or accreta, and divided them into the balloon (248 cases) and control groups (172 cases). The results of this study are consistent with the previous negative nding that PBOIIA had no signi cant effects on improving maternal outcomes. Among the included patients, some underwent PTUI surgery simultaneously. A previous study showed that PTUI had a signi cant effect on reducing intraoperative blood loss and the hysterectomy rate [10]. Therefore, we conducted subgroup analysis to examine the effectiveness of PBOIIA when eliminating the effect of PTUI surgery during CS. We found that there were no signi cant differences in intraoperative EBL, but the rate of hysterectomy was signi cant decreased when PTUI surgery was taken. Additionally, no signi cant differences in postoperative blood indicators were found between the two groups. This nding indicated that PBOIIA may have no bene cial effect on recovery of the blood indicators in patients after CS. However, the total hospitalization cost and surgery costs in the balloon group were signi cantly higher than those in the control group, which greatly increased the nancial burden of patients. Therefore, we should take a cautious attitude towards PBOIIA according to the individual situation of the patients. Abdominal aorta artery balloon occlusion is more effective than internal iliac artery occlusion in hemostasis in patients with placenta previa and accreta, but it is associated with a higher risk of vascular-related complications [28]. The indications of abdominal aorta artery balloon occlusion should be strictly controlled in clinical practice. Further studies are required to compare the advantages and disadvantages of the two types of intervention surgery in patients with placenta previa and accreta [5,29].
Balloon occlusion-related complications are rarely reported, with a rate of approximately 6%-15.8%, and these include ischemia, thrombosis, pain, fever, anemia, hematoma, and infection [30,31]. In our study, there were no signi cant differences in the rates of postoperative fever (≥38.5°C) and anemia (HGB <100 g/L). There were also no balloon occlusion-related complications in patients after major surgery in our study, such as ischemia and thrombosis.

Strengths and limitations
The main advantage of our study is that the number of included patients was larger than that in previous studies, and therefore, the results are more reliable. We also performed subgroup analysis to evaluate the effectiveness of PBOIIA after eliminating the effect of PTUI surgery on hemorrhage. This provided a more reliable analysis of the results. The main limitation of our study is that it was a retrospective, single-center study. Additionally, our study was observational, and the included subjects could not be randomly allocated. Therefore, selection bias may have been present. However, we calculated the basic characteristics of the two groups, and the two groups of patients were comparable.

Conclusion
Our study shows that PBOIIA has no bene t in improving maternal outcomes for patients with placenta previa and accreta, including reducing intraoperative blood loss and hysterectomy rates. Additionally, PBOIIA signi cantly increased the nancial burden of patients. When the effect of PTUI surgery on hemorrhage is eliminated, there is still no signi cant difference in intraoperative EBL and the hysterectomy rate. Therefore, PBOIIA should not be routinely recommended to patients with placenta previa and accreta.