Abnormally invasive placenta: changing trends in diagnosis and management

Authors


  • The authors have stated explicitly that there are no conflicts of interest in connection with this article.

Correspondence

Bindiya Gupta, Department of Obstetrics and Gynecology, University College of Medical Science and Guru Teg Bahadur Hospital, Delhi, India. E-mail: dr_bindiya_gupta@yahoo.co.in

Abstract

Trends in patient profile, clinical presentation, diagnosis, management options and outcome of abnormally invasive placenta (AIP) were retrospectively evaluated at a tertiary care centre from 2001 to 2010. AIP was diagnosed when confirmed by ultrasound or MRI, when complete manual removal of placenta was not possible or when histological confirmation was achieved in a hysterectomy specimen. The first and second halves of the time period were compared. The total number of cases was 56 and the incidence increased significantly in the second half. Main risk factors were placenta previa and previous cesarean section. Antenatal diagnosis by ultrasound was made in 72% in the second half, compared with 35% in the first half (= 0.009). Maternal mortality was 29% in the first half and 21% in the second half. Hysterectomy was the mainstay in management; elective procedures being significantly higher in the second half.

Abbreviations
AIP

abnormally invasive placenta

Introduction

Abnormally invasive placenta (AIP) encompasses various types of abnormal placentation in which chorionic villi attach directly to or invade the myometrium, replacing the historical definitions, i.e. superficial invasions (placenta accreta), middle-layer invasions (placenta increta), and deep invasions (placenta percreta) [1]. It occurs in 1:2500 deliveries and the incidence has increased 10-fold in the last 50 years [2]. The underlying pathology is a defect in the deciduas basalis caused by a variety of insults such as previous cesarean section, previous uterine surgery, excessive curettage or infections [3]. Two-dimensional and three-dimensional color Doppler has 82–100% sensitivity and 92–97% specificity in antenatal diagnosis of AIP [4, 5]. A high index of suspicion is required for diagnosis. Ultrasound features suggestive of AIP include the presence of irregularly shaped placental lacunae within the placenta, thinning of myometrium overlying the placenta, loss of retroplacental “sonolucent line,” protrusion of the placenta into the bladder, increased vascularity of the uterine serosa–bladder interface and turbulent blood flow through the lacunae on Doppler ultrasonography [6]. It is advised to use MRI in cases with a non-conclusive ultrasound and doubtful parametrial invasion [6]. Optimal management involves early recognition, accurate preoperative diagnosis, maternal counseling and elective surgery. Hysterectomy has been the gold standard in management, but conservative management can be used in selected cases [7].

The present study evaluates the changing trends in patient profile, risk factors, clinical presentation, antenatal diagnosis, management and maternal fetal outcome of AIP in our institution in a middle- to low-resource country over the last 9 years.

Material and methods

Retrospective analysis of complete medical records of patients with a diagnosis of AIP, whether antenatal or postnatal, admitted to the obstetrics ward of the University College of Medical Sciences and Guru Teg Bahadur Hospital, Delhi, India, a tertiary care referral centre, from October 2001 to October 2010, was carried out. All women who delivered in the hospital between 22 and 41 weeks of gestation were included. Gestational age was calculated from first day of the last menstrual period or estimated by first obstetrical ultrasound. AIP was diagnosed by the presence of one of the following; ultrasound or MRI diagnosis, manual removal of placenta partially or totally impossible with no cleavage plane between part or whole of the placenta and uterus, or histological confirmation of morbidly adherent placenta on hysterectomy specimen.

To evaluate the changing trends we compared the first and second halves of the 9-year time period and divided it into Period A and Period B. The number of women was 17 and 39 in the two groups, respectively. Data regarding a woman's demographic profile and risk factors were noted. The woman was taken to be a booked case if she had at least three antenatal hospital visits during her pregnancy. Ultrasound or MRI reports wherever available were noted. Records of intraoperative and postoperative complications, maternal and fetal outcome, and histopathology reports, were also obtained.

Maternal morbidity was assessed using the following indicators: need for blood or blood product transfusion, hysterectomy, disseminated intravascular coagulation (defined by platelet count of less than 100 000/mm3, prolongation of prothrombin time and the activated partial-thromboplastin time and by the presence of fibrin-degradation products in plasma), postpartum endometritis (defined by fever of more than 38°C and pelvic pain) with or without sepsis (defined by positive blood culture), need for intensive care admission and duration of intensive care unit stay. The anesthetist and obstetrician jointly decided upon transfusion of blood and blood products. The transfusion policy did not change in the two periods.

Univariate analysis was conducted with Fisher's exact test for categorical variables and the Mann–Whitney U-test for continuous variables. A p-value <0.05 was taken as significant. Hospital ethics committee approval was obtained for the study.

Results

The number of complete hospital records retrieved was 56. In Period A, among 53 994 deliveries, 17 women met the diagnostic criteria of AIP. In Period B, 39 of 46 898 delivered women satisfied the inclusion criteria. The incidence of AIP increased from 0.3/1000 in Period A to 0.8/1000 in Period B.

Table 1 summarizes the demographic characteristics. There were significantly more booked cases (= 0.006) in Period B and 60% had an elective cesarean section in this group of women, in contrast to only 17% in group A (= 0.010). The risk factors during both periods were similar; the major ones being, placenta previa and history of previous cesarean. Other risk factors included previous myomectomy, history of curettage and endometritis.

Table 1. Patient characteristics and clinical presentation
 Period A (n = 17)%Period B (n = 39)%p-value
Age (years)27 ± 4.2 27 ± 2.83 0.82
Parity 2 ± 0.9 3 ± 1.20 0.02
Multiple pregnancy0 1  
Booking status
Booked318.723590.0006
Unbooked1482.31641 
Gestational age at delivery (weeks)35.2 ± 3.2 35.5 ± 3.5 0.72
Retained placenta2 3 0.32
Type of delivery
Vaginal317.538 
Elective cesarean section317.524600.01
Emergency cesarean section11651232 

In Period A, a provisional diagnosis of AIP was made in six women (35%) preoperatively on ultrasound with confirmation by MRI in one, whereas in Period B 28 women (72%) had preoperative diagnosis based on ultrasound and confirmed by MRI in eight (= 0.009). Histopathological confirmation was obtained in 76% (13/17) and 80% (31/39) in the two groups, respectively. Ultrasound was negative in 2/17 cases of AIP diagnosed during delivery in Period A and 3/39 cases in Period B. The false-positive rates of ultrasound were 1/6 (16.67%) and 3/28 (10.7%) in Periods A and B, respectively.

Of 17 cases of AIP diagnosed during Period A, two women presented in the postnatal period with retained placenta. Primary hysterectomy was performed in 76.5% (13/17) of women; 10 (77%) emergency and three (23%) elective. Conservative management was carried out in four women (23.5%) but was successful only in two. In one case where the woman had presented postnatally, Doppler ultrasound showed retention of the complete placenta with evidence of placenta increta and this was managed with 50 mg methotrexate (three injected doses) intramuscularly. The woman was closely followed up with serial ultrasound and human chorionic gonadotropin and discharged after 3 weeks with the residual placenta in situ. Three months later she resumed normal menses and transvaginal ultrasound showed an empty uterine cavity. Another woman was managed expectantly with the placenta in situ without methotrexate. The other two women were initially managed conservatively with partial manual removal of placenta and placental bed suturing. Later, both underwent hysterectomy for secondary hemorrhage after 10 and 24 h, respectively. The first woman also required internal iliac artery ligation and partial cystectomy because of bladder involvement.

In Period B, 31/39 (79.4%) women underwent hysterectomy; 11/31 (35%) elective and 20/31 (65%) emergency procedures. In one woman with an intraoperative percreta diagnosis, the placenta was left in situ. A second hysterectomy was carried out after 9 weeks when there was much reduced vascularity on Doppler ultrasound; the operation was technically easier, safe and cystectomy was avoided. Of the women who underwent hysterectomy, two had urinary bladder injury that required repair. Conservative management was used in eight women (20.5%), but only four were successful (50%). Two of these women had piecemeal removal of the placenta with placental bed suturing. In the two other partial AIP cases, the adherent part of the placenta was left in situ and managed with uterine packing in one case and methotrexate in the other. The follow-up was uneventful and showed complete resolution of placental tissue on ultrasound in both instances. There was no significant difference in the number of hysterectomies in the two groups (= 0.24). However, elective surgery was more frequent in group B (= 0.01).

Table 2 summarizes maternal mortality and morbidity, which was similar in the two groups. The average intraoperative blood loss was 2.7 L in Period A and 1.8 L in Period B. The average numbers of packed red blood cell units and fresh frozen plasma transfused were 4 and 3 per patient, respectively. The average duration of intensive care unit stay was 2.6 days (from 2 h to 13 days) in Period A and 1 day (from 4 h to 3 days) in Period B. Maternal mortality was 5/17 (29%) in group A and 8/39 (20.5%) in group B. Of 13 maternal deaths, 11 (90%) of the women had been admitted as an emergency and 10 (85%) did not have a previous diagnosis of AIP. The main cause of death was hemorrhage (9/13); other women died of septicemia and disseminated intravascular coagulation. The main newborn complication was prematurity.

Table 2. Maternal morbidity and mortality
 Period A (n = 17)Period B (n = 39)
Bladder injury33
Disseminated intravascular coagulation1214
Infection66
Burst abdomen1
Intensive care unit transfer1012
Maternal mortality, n (%)5 (29)8 (20.5)

Discussion

There has been an alarming rise in the incidence of abnormally invasive placentation. Miller et al. [8] reported an incidence of 1 in 2510 in 1994 which had increased to 1 in 533 in 2002 [9]. In the present study also, the incidence was higher in Period B. This may be attributed to increasing cesarean section rates [10]. The cesarean section scar is avascular, which promotes the proliferation of cytotrophoblasts and embryo implantation due to hypoxia. Lack of oxygen further delays the arrest of mitosis and initiation of differentiation, thereby enabling the trophoblast to invade more deeply into the scar area [11]. In the presence of placenta previa, the risk of having placenta accreta increases from 24% in women with one cesarean section to 67% for women with three previous cesarean sections [3]. In the present study the risk factors were the same in the two groups, the major ones being placenta previa and previous cesarean section(s).

As a consequence of better health services, the number of registered or booked patients and antenatal health check-ups and the accuracy of antenatal imaging were significantly higher in the second group. This resulted in more elective admissions and more elective surgery, which had an obvious impact on overall maternal morbidity and mortality in the second group. With improved imaging modalities, the antenatal diagnosis increased from 35 to 72% in the present study, in line with other studies [11].

With AIP an extirpative approach is most commonly used [1]. If risk factors and prenatal imaging together strongly suggest the diagnosis of AIP, generally a cesarean hysterectomy is planned with informed consent, especially for women who do not wish to continue their fertility. If diagnosis is made after delivery of the baby, in most cases, the forced delivery of the placenta induces massive hemorrhage and leads to emergency hysterectomy. In addition, surgery in cases of AIP can cause injury to contiguous organs and so increase maternal morbidity and mortality. A newer approach of second-stage hysterectomy was tried in one case in the present study resulting in favorable dissection planes and less blood loss. This warrants further research. The conservative options include leaving the placenta in situ, awaiting either spontaneous resorption or expulsion. It can also be managed with prophylactic or therapeutic uterine artery embolization, internal iliac artery ligation or methotrexate [12]. The latter was here used successfully in 27% of the women of whom two-thirds had a successful outcome. Success of conservative treatment is reported to be as high as 78% [13].

In their series Rosen et al. and O Brien et al. reported an average blood loss of 3 L and an average of 10 units (3–29 units) of packed red blood cells were transfused per patient [11, 14]. The incidence of intensive care unit admissions and the intraoperative and postpartum complications they reported were similar to what was seen in the present study [11, 14]. The maternal mortality was, however, 22% in the present study compared with 7% [14]. This can be attributed to a large number of women being diagnosed in an emergency situation, a delayed arrival for tertiary care, limited availability of blood and blood products and non-availability of intensive care unit facilities out of hours in our setting. In under-resourced nations undiagnosed AIP in women with a previous cesarean section may lead to serious consequences and even maternal death. With improved antenatal diagnosis, there is an increase in elective surgery and a corresponding decrease in maternal morbidity and mortality may follow.

Funding

No specific funding.

Ancillary