Background
Placenta previa remains one of the major causes of massive postpartum hemorrhage and maternal mortality worldwide.
Objective
To determine whether internal iliac artery balloon occlusion during cesarean delivery for placenta previa could reduce postpartum hemorrhage and other maternal complications.
Study Design
This was a prospective randomized controlled trial conducted at a tertiary university obstetric unit in Hong Kong. Pregnant women who were diagnosed to have placenta previa at 34 weeks (defined as lower placenta edge within 2 cm from the internal os) and required cesarean delivery were invited to participate. Eligible pregnant women were randomized into internal iliac artery balloon occlusion (Occlusion) group or standard management (Control) group. Those randomized to the Occlusion group had internal iliac artery balloon catheter placement performed before cesarean delivery and then balloon inflation after delivery of the baby. The primary outcome was the reduction of postpartum hemorrhage in those with internal iliac artery balloon occlusion. Secondary outcome measures included hemoglobin drop after delivery; amount of blood product transfusion; incidence of hysterectomy; maternal complications including renal failure, ischemic liver, disseminated intravascular coagulation, and adult respiratory distress syndrome; length of stay in hospital; admission to intensive care unit; and maternal death.
Results
Between May 2016 and September 2018, 40 women were randomized (20 in each group). Demographic and obstetric characteristics were similar between the 2 groups. In the Occlusion group, 3 women did not receive the scheduled procedure, as it was preceded by antepartum hemorrhage that required emergency cesarean delivery, and 1 woman had repeated scan at 36 weeks showing the placental edge was slightly more than 2 cm from the internal os. Intention-to-treat analysis found no significant differences between the Occlusion and the Control groups regarding to the median intraoperative blood loss (1451 [1024–2388] mL vs 1454 [888–2300] mL; P = .945), the median length of surgery (49 [30–62] min vs 37 [30–51] min; P = .204), or the need for blood transfusion during operation (57.9% vs 50.0%; P = .621). None of the patients had rebleeding after operation, complication related to internal iliac artery procedure, or any other maternal complications. Reanalyzing the data using on-treatment approach showed the same results.
Conclusion
The use of prophylactic internal iliac artery balloon occlusion in placenta previa patients undergoing cesarean delivery did not reduce postpartum hemorrhage or have any effect on maternal or neonatal morbidity.
Placenta previa remains one of the major causes of massive postpartum hemorrhage and maternal mortality worldwide. The risk of life-saving hysterectomy after cesarean delivery for placenta previa is 5.3%, which is increased by 30 times when compared to those without previa. Since the prevalence of placenta previa and its related complications, including placenta accreta spectrum, has been climbing up because of the increasing cesarean delivery rate, it has significant impact on the cost of obstetric care.
Why was the study conducted?
It is unknown if perioperative prophylactic internal iliac artery balloon occlusion can reduce postpartum hemorrhage in placenta previa. No randomized controlled trial has been conducted on this topic.
Key findings
Perioperative prophylactic internal iliac artery balloon occlusion is safe but does not reduce postpartum hemorrhage in placenta previa.
What does this add to what is known?
Perioperative prophylactic internal iliac artery balloon occlusion does not reduce postpartum hemorrhage in placenta previa.
Postpartum hemorrhage from placenta previa is particularly severe because the hypervascular placental bed is located at the poorly contractile lower segment, where the uterine muscles are sparse. Even strong uterotonic agents may not be able to augment the uterine tone at the lower segment, owing to the nature of its anatomy and musculature. , , , Consequently, hemostasis in placenta previa relies heavily on surgical techniques. These methods include multiple placement of hemostatic sutures at the placental bed, uterine compression sutures with or without the use of intrauterine balloon tamponade, the application of FloSeal to the placental bed, and ligation of the uterine artery. If these conservative surgical approaches fail to control bleeding, the last resort is hysterectomy for life-saving.
However, massive bleeding often leads to poor visualization of the operation field and coagulopathy, making surgical treatment more difficult. Therefore, preventive measures to reduce primary blood loss could be a better approach. Recently, preoperative placement of internal iliac artery catheters with intraoperative arterial balloon occlusion has become a popular prophylactic method. After cesarean delivery of the baby and clamping of the umbilical cord, the intra-arterial balloons are inflated to occlude the arteries. They do not arrest blood flow to the uterus completely but decrease the pulse pressure distal to the occlusion site. By reducing the rate of blood loss, the hemostatic procedures could be easier and shorter. The intra-arterial catheters also allow therapeutic embolization to be performed if bleeding persists after the operation. Although there were a number of case series reporting the use of prophylactic balloon catheter insertion in placenta previa and accreta spectrum, there is only a single randomized controlled trial on this approach in placenta accreta spectrum, and its result shows no beneficial effect but 15.4% procedure-related adverse effect. Evidence on placenta previa is not yet available. Hence the objectives of the current study are to determine whether prophylactic internal iliac artery balloon occlusion during cesarean delivery for placenta previa could reduce postpartum hemorrhage, and its benefits may outweigh its potential adverse effects to justify the implementation of this preventive measure.
Materials and Methods
Study design
This was an open-label randomized controlled trial conducted at a university hospital in Hong Kong. This trial was approved by the Joint Chinese University of Hong Kong-New Territories East Cluster Clinical Research Ethics Committee (CRE ref. no.: 2015.260-T) and monitored by the Clinical Research Management Office, The Chinese University of Hong Kong. The trial was registered in the Centre for Clinical Research and Biostatistics (CCRB) Clinical Trials Registry (CUHK_CCRB00507) and Chinese Clinical Trial Registry (ChiCTR-IOR-16008281).
Participants
Pregnant women who were found to have low-lying placenta at routine 18- to 22-week ultrasound scan were scheduled to have another ultrasonic assessment for placental site at 32–34 weeks. Placenta previa was then diagnosed and graded by transvaginal ultrasonogram according to the distance between the lower placental edge and the internal os of the uterine cervix: Grade I previa was defined as a lower edge within 5 cm from the internal os, grade II as a lower edge reaching the internal os, grade III when the placenta partially covers the cervix, and grade IV or complete previa when the placenta completely covers the cervix. Cesarean delivery was arranged for these patients between 37 and 38 weeks of gestation, except for those patients with grade I placenta previa with placental edge more than 2 cm away from the internal os. Patients within this subgroup of grade I placenta previa were counseled for the possibility of successful vaginal delivery. , , Those placenta previa patients who were planned for cesarean delivery were invited to participate in the study. Exclusion criteria included (1) age of less than 18 years; (2) mentally handicapped or severely ill; (3) fetus diagnosed with in utero death; (4) had contraindication for internal iliac artery balloon occlusion; (5) unstable clinical condition in which patients could not be sent to interventional radiology suite for preoperative internal iliac artery catheter placement; (6) unable to understand English or Chinese to give consent; and (7) ultrasonic features suggestive of placenta accreta spectrum. The recruited subjects were reassessed within 1 week before the operation, when a transvaginal ultrasound scan was repeated to double check the distance between the placental edge and the internal os.
Randomization
After a written informed consent was obtained, the subjects were randomly assigned to either internal iliac artery balloon occlusion (Occlusion group) or standard management (Control group) at a 1:1 ratio. The randomization sequence was generated by a web-based program that was set up by the research unit. The allocation sequences were sequentially numbered and sealed into opaque envelopes and were kept by the trial coordinator, who was not involved in subject recruitment. Study subjects and investigators were aware of the allocation, as blinding in this study was impossible owing to the nature of the intervention.
Interventions and treatment procedure
On the day of the planned cesarean delivery, those who were randomized to the Occlusion group were brought to the interventional radiology suite before the surgery. Under local anesthesia, an arterial sheath was placed in both common femoral arteries using the Seldinger technique after arterial puncture. Under fluoroscopic guidance, the bilateral internal iliac arteries were catheterized using 6 French balloon catheters. The balloon of the catheter was positioned in the proximal portion of each of the 2 internal iliac arteries. Afterwards, test inflation of the balloons using diluted water-soluble contrast was performed under fluoroscopy and road-mapping guidance to find out the volume of contrast required to optimally inflate the balloon for complete occlusion of the internal iliac artery. The catheters were then securely taped to the skin. The whole procedure was performed by a qualified interventional radiologist in all the cases.
The subject was then sent to the operating theater, where cesarean delivery was performed. After the baby was delivered, the occlusion balloons were inflated immediately. The placenta was then delivered and cesarean delivery was continued in the usual manner. If bleeding was not controlled after the operation, uterine artery embolization was indicated in both groups. In the Occlusion group, if hemostasis was achieved after the cesarean delivery, the balloon was deflated; if bleeding was not controlled and uterine artery embolization was indicated, the balloons were left inflated until embolization was performed. Otherwise, the balloon catheters and vascular sheaths were left in situ for 12–24 hours after surgery in case emergency embolization might be required. Slow and continuous infusion of normal saline at a rate of 10 mL/hour was given through the arterial sheaths to maintain their patency. After removal of the arterial sheaths, hemostasis was achieved by external compression.
All cesarean deliveries in both groups were performed by a qualified obstetrician with sufficient experience and expertise in operating on placenta previa.
Measurement of intrapartum and postpartum blood loss
During the cesarean delivery, a disposable delivery drape with a blood collection system was applied around the abdominal wound for accurate collection and measurement of blood loss from the abdominal cavity. A gynecological drape with pouch was also used to collect blood from the vagina. Blood volume in the suction bottle was measured. All gauzes or pads with blood were collected and weighed and an equivalent volume was calculated. The total intraoperative blood loss was estimated by the on-site anesthetist who was not part of the research team, based on the blood volume in the delivery drape and gynecological drape collection systems, the suction bottle, and weighed gauzes. Any post–cesarean delivery blood loss was also quantified.
All subjects were monitored after the operation. Hemoglobin levels were checked within 1 day pre–cesarean delivery, immediately after operation, and on day 1 and day 3 of post–cesarean delivery to assess any drop in the hemoglobin level. All patients were followed 6–8 weeks postnatally for any complications.
Outcome measures
The primary outcome was reduction in intrapartum blood loss. The secondary outcomes included drop in hemoglobin level after the operation; the amount of blood products transfused; the incidence of hysterectomy; maternal complications including renal failure, ischemic liver, disseminated intravascular coagulation, and adult respiratory distress syndrome; incidence of uterine artery embolization; days of stay in the hospital after the operation; admission to intensive care unit; and maternal death.
Sample size calculation
The sample size calculation was based on a reduction of blood loss by half in the internal iliac artery balloon Occlusion group as compared to the Control group. From a previous report and our observation, the average blood loss in patients undergoing cesarean delivery for placenta previa was 800–1000 mL, and the standard deviation was 300 mL. To detect a difference of 400 mL blood loss, with a power of 80% and a significance level of 5%, 10 subjects were required in each arm. A total of 20% (2 more subjects) was added to the sample size because of the skewed distribution of blood loss. To allow possibility of abandoning the procedure because of unforeseen emergency, the sample size was increased to 16 in each arm.
Statistical analysis
Statistical analysis was performed using an intention-to-treat method and was repeated using on-treatment approach. Descriptive data were presented as the median (interquartile range [IQR]). Dichotomous and categorical data were reported as numbers and percentages. Comparison of variables between the 2 groups was analyzed using Mann-Whitney U test for continuous variables and by χ 2 test or Fisher exact test for categorical variables. A mixed ANOVA was used to assess the level of hematologic markers at different time points between the 2 groups. Statistical significance was reached when 2-tailed P < .05. All statistical analyses were performed using SPSS statistics software (version 22, IBM SPSS Statistics for Windows; IBM Corp, Armonk, New York).
Results
Between May 2016 and September 2018, a total of 50 women with an antenatal diagnosis of placenta previa were invited to participate in the study. Forty women consented and were randomly allocated to Occlusion group (n = 20) or Control group (n = 20) ( Figure 1 ). All women were ethnically Chinese. Twenty of them (50%) were nulliparous. The median age was 36.0 years (range, 23.8–44.1 years). The median body mass index at delivery was 26.9 kg/m 2 (IQR, 24.8–29.4 kg/m 2 ). The median gestational age at delivery was 37.4 weeks (IQR, 37.0–38.1 weeks). The demographic and obstetric characteristics were similar between the 2 groups, except that the prevalence of previous cesarean delivery was higher in the Occlusion group (60% vs 10%; P = .019; Table 1 ).
| Occlusion group (n = 20) | Control group (n = 20) | P value | |
|---|---|---|---|
| Maternal age (years) | 35.3 [31.5–37.6] | 36.6 [32.7–39.1] | .289 |
| Gestational age (weeks) | 36.6 [35.2–37.2] | 36.1 [34.8–37.7] | .968 |
| Nulliparous | 10 (50.0%) | 10 (50.0%) | .999 |
| Previous cesarean delivery | 6 (60.0%) | 1 (10.0%) | .019 |
| Body mass index at cesarean delivery (kg/m 2 ) | 27.0 [24.9–29.4] | 26.9 [24.5–29.7] | .862 |
| Assisted reproduction pregnancy | 1 (5.0%) | 2 (10.0%) | .999 |
| Active smoker | 4 (20.0%) | 1 (5.0%) | .342 |
| Gestational diabetes | 3 (15.0%) | 4 (20.0%) | .999 |
| Pregnancy-induced hypertension | 2 (10.0%) | 0 (0.0%) | .487 |
| Placenta previa graded at recruitment | .999 | ||
| I & II | 5 (25.0%) | 6 (30.0%) | |
| III & IV | 15 (75.0%) | 14 (70.0%) | |
| Placental site | |||
| Anterior | 7 (35%) | 4 (20%) | .479 |
| Nonanterior | 13 (65%) | 16 (80%) | |
| Antepartum hemorrhage before procedure | 11 (55.0%) | 10 (50.0%) | .752 |
Four women in the Occlusion group did not receive the planned internal iliac artery procedure. Three of them required emergency cesarean delivery because of significant antepartum hemorrhage before the scheduled date of the planned internal iliac artery balloon catheter insertion. They were included in the intention-to-treat analysis but were excluded from the analysis by on-treatment method. The remaining woman was reassessed at 36 weeks with ultrasound scan, which showed that the placental edge was 2 cm from the internal os. Therefore she opted for a trial of vaginal delivery and was excluded from the study.
Intention-to-treat analysis
The intraoperative and postoperative outcomes of the 2 groups are shown in Table 2 . The neonatal and maternal outcomes are shown in Table 3 . There was no significant difference between the Occlusion and Control groups with regard to the median intraoperative blood loss (respectively, 1451 [1024–2388] mL vs 1454 [888–2300] mL; P = .945), the incidence of massive postpartum hemorrhage ≥1500 mL (47.4% vs 50.0%; P = .869), and the median length of surgery (49 [30–62] min vs 37 [30–51] min; P = .204). The need for blood transfusion during operation (57.9% vs 50.0%; P = .621) and the transfusion of packed red cells during operation (57.9% vs 50.0%; P = .621) were not significantly different between the groups. The use of carboprost during operation was significantly more common in the Occlusion group than in the Control group (36.8% vs 5.0%; P = .020).
| Occlusion group (n = 19) a | Control group (n = 20) | P value | |
|---|---|---|---|
| Intraoperative blood loss (mL) | 1451 [1024–2388] | 1454 [888–2300] | .945 |
| Postpartum hemorrhage ≥1500 mL | 9 (47.4%) | 10 (50.0%) | .869 |
| Postpartum hemorrhage ≥1000 mL | 15 (78.9%) | 15 (75.0%) | .999 |
| Length of surgery (min) | 49 [30–62] | 37 [30–51] | .204 |
| Any transfusion at operation | 11 (57.9%) | 10 (50.0%) | .621 |
| Any transfusion of packed RBC | 11 (57.9%) | 10 (50.0%) | .621 |
| Units of packed RBC transfused | 2 [0–4] | 0.5 [0–2.75] | .550 |
| Any transfusion of platelet | 4 (21.1%) | 5 (25.0%) | .999 |
| Units of platelet transfused | 0 [0–4] | 0 [0–3] | .945 |
| Any transfusion of fresh frozen plasma | 2 (10.5%) | 4 (20.0%) | .661 |
| Units of fresh frozen plasma transfused | 0 [0–0] | 0 [0–0] | .411 |
| Any transfusion of cryoprecipitate | 1 (5.3%) | 1 (5.0%) | .999 |
| Units of cryoprecipitate transfused | 0 [0–0] | 0 [0–0] | .999 |
| Syntocinon infusion during operation | 17 (89.5%) | 20 (100.0%) | .231 |
| Misoprostol during operation | 7 (36.8%) | 5 (25.0%) | .501 |
| Carboprost during operation | 7 (36.8%) | 1 (5.0%) | .020 |
| Intrauterine balloon during operation | 0 (0.0%) | 1 (5.0%) | .999 |
| Compression suture during operation | 6 (31.6%) | 4 (20.0%) | .480 |
| Uterine artery or internal iliac artery ligation during operation | 2 (10.5%) | 0 (0.0%) | .231 |
| Uterine artery embolization after operation | 2 (10.5%) | 0 (0.0%) | .231 |
| Hysterectomy | 0 (0.0%) | 0 (0.0%) | – |
a One patient from the Occlusion group was reassessed with scan at 36 weeks that showed that the placental edge was 2 cm from the internal os; she was allowed to undergo vaginal delivery.
| Occlusion group (n = 19) a | Control group (n = 20) | P value | |
|---|---|---|---|
| Maternal outcomes | |||
| Postoperative hospital stay (days) | 4.0 [3.0–6.0] | 3.0 [3.0–4.0] | .070 |
| Intensive care unit admission | 0 (0.0%) | 1 (5.0%) | .999 |
| Rebleeding after operation | 0 (0.0%) | 0 (0.0%) | – |
| Relaparotomy | 0 (0.0%) | 0 (0.0%) | – |
| Maternal death | 0 (0.0%) | 0 (0.0%) | – |
| Internal iliac artery balloon occlusion–related complications | 0 (0.0%) | 0 (0.0%) | – |
| Other maternal complications b | 0 (0.0%) | 0 (0.0%) | – |
| Neonatal outcomes | |||
| Birthweight (grams) | 2690 [2530–2980] | 2895 [2673–3190] | .149 |
| Male | 10 (52.6%) | 11 (55.0%) | .882 |
| Apgar score <7 at 5 minutes | 0 (0.0%) | 1 (5.0%) | .999 |
| Umbilical cord arterial pH | 7.29 [7.26–7.30] | 7.29 [7.27–7.32] | .459 |
| Umbilical cord arterial base excess | -4.50 [-8.93 to -2.3] | -2.90 [-5.0 to -1.0] | .109 |
| Neonatal intensive care unit admission | 1 (5.3%) | 1 (5.0%) | .999 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
