PACCRETA: Clinical situations at high risk of Placenta ACCRETA/Percreta: impact of diagnostic methods and management on maternal morbidity
- The authors have stated explicitly that there are no conflicts of interest in connection with this article.
Gilles Kayem, Department of Obstetrics and Gynecology, CHU Louis Mourier, APHP, Université Paris Diderot, Paris, France. E-mail: email@example.com
No population-based study has assessed the prevalence of placenta accreta, the predictive value of prenatal diagnostic examinations, the maternal morbidity associated with its management, or its psychological effects. Moreover, the technical resources necessary for the safest delivery of women at risk have not been clearly identified.
The primary objectives are to identify individual risk factors, as well as factors associated with types of management and with the healthcare facilities for postpartum hemorrhage. Secondary objectives are to (i) determine the prevalence of placenta accreta among women at risk; (ii) assess the predictive value of ultrasound and MRI for the diagnosis of placenta accreta in this population; (iii) evaluate psychological impact and physical complications; and (iv) report maternal physical complications during the year after delivery.
Population-based prospective observational study of pregnant women with a placenta inserted at a uterine scar.
A total of 182 centers in eight French regions, with 270 000 deliveries annually.
We expect to include 620 women at risk of placenta accreta in two years, of whom approximately 120 (20%) will have placenta accreta.
Main outcome measures
The primary outcome measure is severe maternal hemorrhage; secondary outcome measures are maternal morbidity and deaths, predictive value of ultrasound and MRI, and psychological evaluations at 6 and 12 months.
This study will be the first prospective population-based study to include women at risk of placenta accreta and to investigate incidence, prenatal detection, type of management, morbidity and maternal psychological consequences.
magnetic resonance imaging
The occurrence of placenta accreta is linked to abnormal invasion at the placental implantation site due to a decidual defect . The term increta is used in cases of invasion of the myometrium, while the term “percreta” refers to invasion of the serosa or even adjacent organs, most frequently the bladder. Nonetheless, the term accreta is frequently used to cover all three definitions .
Placenta accreta is often diagnosed after the baby's birth, when the placenta fails to deliver. Trying to force this delivery can result in a serious postpartum hemorrhage (PPH), emergency hysterectomy, and even death. Accordingly, abnormalities of placental insertion are responsible for 35–38% of peripartum hysterectomies in recent population-based studies [3, 4]. Other reported complications include both the possible visceral complications of a severe hemorrhage, such as multiple organ failure, and damage to adjacent organs such as the bladder. These are most common for placenta percreta, for which per- and postsurgical morbidity is elevated; it is reported to be lethal in 7% of cases . Finally, more rarely, spontaneous uterine rupture in the second or third trimester of pregnancy has been described, combined with massive hemoperitoneum [6-8].
Management of placenta accreta involves two unresolved aspects: first, the methods for its prepartum identification (subgroups at risk, additional work-up and imaging) and the type of management that can reduce maternal complications to the minimum possible, and secondly, the technical and human resources necessary for its management have not been clearly codified.
Our purpose in this article is first to review the current medical literature regarding placenta accreta and secondly to build a research protocol designed to provide an answer for some of the unanswered questions.
Incidence and risk factors of placenta accreta
The incidence of placenta accreta appears to be rising and may have multiplied by 10 over the past 50 years. Miller et al.  reported that among 155 670 deliveries, placenta accreta complicated 1/2510 births between 1985 and 1994. In a more recent study covering the period 1982–2002, Wu et al. found a still higher incidence in their establishment: 1 per 533 pregnancies . It is nonetheless important to note that all data about the prevalence of placenta accreta come from single-center studies in reference centers or studies of hospital databases. The disadvantages of these databases are probably flaws in their exhaustiveness, especially for the less serious cases, and coding errors. These problems are particularly acute for placenta accreta, which can be difficult to diagnose and for which strict diagnostic criteria must be defined. Because a pathologic examination is difficult in the absence of hysterectomy, the diagnosis most often has to be based on a combination of findings, which must be defined in advance. Moreover, the International Classification of Diseases, which is used for coding most of these hospital databases, has no specific code for placenta accreta cases, which are then coded unspecifically as “retained placenta” (O72.0) or “placenta previa with hemorrhage” (O44.0).
The primary risk factor for placenta accreta is a uterine scar, associated with either a cesarean or a gynecologic procedure (such as curettage, myomectomy or hysteroplasty). Placental implantation on the scar area, like anterior placenta previa after previous cesareans, can induce placenta accreta, and the risk increases with the number of previous cesarean sections. Other reported risk factors are maternal age greater than 35 years and multiparity, but the risk of placenta accreta in a healthy uterus is nonetheless very low. Miller et al. showed that placenta accreta occurred in 55 of 590 (9.3%) women with placenta previa compared with 7 of 155 080 (1/22 154) without placenta previa (relative risk 20.7, 95% confidence interval: 9.4–45.2) . Moreover, among women with placenta previa, placenta accreta was diagnosed in 36 of the 124 (29%) women for whom placental implantation overlaid a cesarean scar and in four of 62 (6.5%) women for whom it did not (relative risk 4.5, 95% confidence interval). Among women with placenta previa, age greater than 35 years and a history of cesarean were also independent risk factors for placenta accreta. Finally, in the case of placenta previa, the risk of placenta accreta ranged from less than 2% for women younger than 35 years with no previous cesarean section to 39% for women with two or more previous cesareans .
The principal risk factor for placenta accreta is thus a previous cesarean delivery. The regular augmentation of the cesarean rate from 10% 30 years ago to 20–30% in developed countries today has induced a substantial increase in the risk of placenta accreta . In France, the cesarean rate has climbed from 16% in 1995 to 21% in 2010 (results of perinatal surveys) . The current debates in obstetrics, leading to cesarean deliveries in situations with a very low excess risk for the fetus associated with vaginal birth,or even none at all, as in the case of ‘convenience cesareans’, must incorporate an analysis of their potential consequences, in particular, increased maternal morbidity.
Prenatal screening for placenta accreta
Prenatal diagnosis might have a beneficial effect on management. Warshak et al.  compared 62 cases of cesarean-hysterectomies diagnosed during the prenatal period with 37 cases discovered per partum in a single-center study in California. They showed that the risk of hemorrhage is reduced when the diagnosis precedes the cesarean section. In this study, the patients diagnosed before delivery had placenta percreta in 52% of cases, compared with 6% when the diagnosis was made at delivery. While the clinical forms of placenta accreta were more severe in the group with prenatal diagnoses, the risk of hemorrhage and use of transfused blood products tended to be lower than among the group for whom placenta accreta was diagnosed at delivery. The objective of prenatal screening of placenta accreta is therefore to improve maternal prognosis by being prepared for an obstetric emergency and thus avoiding the complications of a massive hemorrhage and an emergency hysterectomy. It allows planned multidisciplinary management including obstetricians, anesthetists and radiologists specialized in interventional radiology. Moreover, in the case of placenta percreta, damage to neighboring organs can require the participation of urologists or visceral surgeons. Diagnosis is generally suspected due to ultrasound and magnetic resonance imaging (MRI) findings in patients with risk factors.
Numerous studies have assessed the performance of ultrasound for predicting placenta accreta. The ultrasound criteria classically described are the absence of a hypoechoic zone between the placenta and the myometrium, interruptions of the echogenic area at the interface of the serosa and the bladder, a pseudotumoral appearance of the placenta in/around the uterine serosa and the presence of intraplacental lacunae in the accreta zone, which give the placenta a ‘Swiss cheese’ or moth-eaten appearance [14, 15].
Yang et al.  studied the predictive value of placental lacunae for maternal morbidity in 51 patients who had placenta previa and a history of cesarean delivery . The intraplacental lacunae were classified according to a score based on Finberg's criteria : the absence of lacuna was classified as grade zero, one to three small lacunae as grade 1, four to six large, irregular lacunae as grade 2, and numerous lacunae including some that are large and irregular as grade 3. Grade 1 lacunae had the best predictive value, with a sensitivity of 86%, a specificity of 78%, a positive predictive value of 76% and a negative predictive value of 88%. No hysterectomy was performed in women without lacunae.
MRI has been reported to be useful at diagnosis [17, 18], especially when the placenta is posterior . The most interesting study, because it was closest to daily clinical practice, is that reported by Warshak et al. . All the patients with a history of uterine scar (or myomectomy) and placenta previa or a low-inserted placenta initially had a Doppler ultrasound. If placenta accreta was suspected (uterine scar and placental implantation on the scar), an MRI with gadolinium injection was performed. Placentation was considered abnormal if manual removal of the placenta was difficult (due to its adhesion to the myometrium) and resulted in a clinical diagnosis of PPH, or if the pathology examination found villi but not decidual cells in the myometrium (pathology diagnosis). Of the 453 patients studied, 9% had placenta accreta. MRI allowed the authors to rule out the diagnosis of placenta accreta in 14 of the 16 Doppler ultrasound false-positives. Therefore, when ultrasound findings suggest placenta accreta, MRI seems useful for improving on the performance of Doppler ultrasound.
These different studies all present methodological problems that make their interpretation a matter of delicacy. The samples are not representative of the general population of women at risk of placenta accreta, since they were generally small and the women were identified retrospectively and were not included on a continuous basis. Finally, they all come from reference centers. Moreover, the external validity of the results obtained is uncertain, as the radiologic examinations were often performed in a specialized or research setting.
No study has ever prospectively evaluated their predictive value in the general population when assessed at general hospitals or by physicians not studying this issue. There are no guidelines in France for the work-up to be performed when placenta accreta is suspected.
Management of placenta accreta suspected before delivery
The consensual attitude for placenta accreta is that an extirpative approach must not be attempted [21, 22]. Trying to detach it is likely to induce a massive hemorrhage. Two types of management are possible: a cesarean with a hysterectomy or conservative treatment consisting of leaving the placenta in the uterine cavity without a hysterectomy.
For women who do not want more children, a cesarean hysterectomy is appropriate if the risk factors and imaging strongly indicate the diagnosis. The American College of Obstetrics and Gynecology currently recommends a cesarean hysterectomy without an attempt at manual removal of the placenta in cases of strong prenatal suspicion of placenta accreta . Very few series have assessed maternal morbidity after a cesarean hysterectomy. In a series of 76 cases of cesarean hysterectomy for placenta accreta, Eller et al.  found the following rates of various outcomes: transfusion (≥4 units of packed red blood cells) 42%, cystotomy 29%, ureteral wounds 7%, infectious complications 33%, and major morbidity 59%. In a more recent series in California, Warshack et al.  found no significant difference in maternal morbidity for women undergoing a cesarean hysterectomy after prenatal diagnosis of placenta accreta. Among the complementary management strategies proposed in cesarean hysterectomies are intraoperative embolization and preoperative placement of intravascular balloons. Uterine artery embolization performed after fetal extraction and before hysterectomy has thus been proposed. In a series of 26 women who had a preventive arterial embolization before the hysterectomy, Angtsmann et al.  observed a significant reduction in blood loss, percentage of patients receiving transfusions, and number of units of packed red blood cells transfused. The results for placement of intravascular balloons varied [26-28]. Nonetheless, the studies were all retrospective assessments of small samples.
Maternal morbidity appears to be influenced by the medical and technical resources available. A retrospective study in Utah (USA) found a reduction in maternal morbidity after a cesarean hysterectomy for placenta accreta in cases managed in a center that had an obstetrician and a surgeon always available (24/7), a blood bank and an interventional radiology department .
Management of placenta accreta without hysterectomy: conservative treatment that leaves the placenta in the uterine cavity
Hysterectomy results in permanent sterility. Moreover, when placenta accreta/percreta is present, it can be accompanied by high morbidity and be life-threatening. To try to minimize these complications, particularly when the patient expresses a desire for more children, an alternative conservative treatment can be offered [21, 30]. A study compared an extirpative strategy (attempt to obtain an empty uterus by forcible manual removal of the placenta) and conservative treatment performed during two consecutive periods in the same center [21, 31]. Conservative management was associated with a lower risk of hemorrhage and a lower hysterectomy rate but a higher risk of maternal infection than extirpative management. This conservative strategy seems beneficial in patients who want another child because this type of management has been shown to enable subsequent pregnancies, although they are associated with a risk of recurrent placenta accreta [32, 33].
Numerous case reports also describe this conservative management; they underline in particular the risk of complications from hemorrhage and infection . Two other French series described cases of placenta accreta managed conservatively. Bretelle et al.  used conservative management in 26 women and had a final hysterectomy rate of 19%. Sentilhes et al.  reported 167 cases of placenta accreta treated conservatively at 40 university hospital centers throughout France. The hysterectomy rate was 22% and that of severe maternal morbidity 6%; one woman died.
As for the cesarean hysterectomy, the specific modalities of conservative treatment for optimal management remain to be determined . Should manual removal of the placenta be attempted routinely in cases of prenatal suspicion of placenta accreta, by moderate traction on the umbilical cord, to avoid leaving a non-accreta placenta in place in case of a false-positive diagnosis from prenatal imaging, but at the risk of inducing a severe hemorrhage? Is a preventive uterine devascularization (embolization and/or surgical vascular ligation) appropriate in the absence of a hemorrhage?
Finally, although conservative treatment seems preferable for placenta percreta with invasion of the adjacent organs [37, 38], the least morbid management for placenta accreta or percreta without such invasion remains open. Practices in France differ from those in the USA where a cesarean hysterectomy is the standard recommendation .
Management of placenta accreta diagnosed after the baby's birth
In numerous cases, the diagnosis is made only in the third stage of labor. This can occur during delivery, when the placenta is not delivered and no plane of cleavage is found between the uterus and the placenta. If the attempt was careful, without force or insistence, and the patient's hemodynamic status stable, conservative treatment can be attempted. In other cases, management is the same as for serious PPH, except that uterotonics used alone are less effective than for causes such as uterine atony.
No evaluation has ever assessed the effect of the management strategy used for placenta accreta in general population studies, nor are there any management guidelines for this pregnancy complication in France. The various methods, conservative or radical, have been studied only in retrospective series that can not ensure exhaustive data. For this disease, with such major implications for maternal health and the continuous rise in the prevalence of its risk factors, it seems urgent that the determinants of maternal complications be specified in more detail – both the individual factors and those related to management and healthcare organization. Moreover, the benefit of using interventional radiology techniques should be assessed.
Psychological repercussions of placenta accreta
Women's satisfaction about the treatment used and the subsequent psychological repercussions have not been evaluated. It was recently suggested that a severe PPH handled by embolization might have major psychological repercussions not only for the woman but also for her spouse, in both the short- and long-term . This finding must be considered in light of the insufficient prevention of this type of effect. Thus psychological management was proposed in only 66% of the cases, although negative repercussions were observed in 41% .
It is possible that the psychological repercussions differ substantially for different management strategies. Better understanding of the psychological impact of the treatment used would make it possible to consider this aspect in choosing a treatment strategy and to implement early management to improve the women's quality of life. Similarly, the occurrence of potentially fatal complications during pregnancy can also have a prolonged and major psychological impact that impairs quality of life. Assessing these consequences would make it possible to define early and appropriate management strategies.
Material and methods: the study protocol
As shown by our overview, many questions remain unresolved. Particularly, in a population setting, few data exist on the incidence of placenta accreta, the predictive value of ultrasonography or MRI used for the diagnosis, the best setting and therapeutic strategy used or the maternal psychological consequences of these dramatic situations. For these reasons, we built a research protocol with a population-based approach, including eight French regions, 182 centers and approximately 270 000 deliveries per year.
The study will include all pregnant women with a uterine scar and anterior placenta previa (defined as the “women at risk”) and giving birth in one of the maternity units of the 11 participating perinatal networks over a two-year period (the study is expected to begin in 2013). Women who do not speak French or refuse to participate will be excluded. All women meeting the inclusion criteria will be identified and flagged by local coordinators at each participating center. Moreover, the exhaustiveness of the recruitment will be verified by the clinical research technicians responsible for the study, by consulting the delivery logbook and the computerized databases.
The primary outcome measure is severe PPH, defined as a PPH involving transfusion ≥4 units of packed red blood cells.
The secondary outcome measures are as follows:
Incidence of placenta accreta or percreta in patients with a uterine scar and placenta previa. The diagnosis of placenta accreta will be made if at least one of the following clinical and histologic criteria are met :
- Manual delivery of the placenta impossible or incomplete with partial or total absence of a plane of cleavage between the placenta and the myometrium,
- Massive bleeding of the site of placental insertion after difficult manual delivery of the placenta,
- Histologic confirmation of the placentation anomaly from the hysterectomy (uterine) or placental specimens, and
- Prenatal diagnosis of placenta accreta confirmed at delivery.
Diagnostic value for placenta accreta or percreta of ultrasound and of MRI. The signs described above will be collected.
Hysterectomy rates. The need for hysterectomy will be recorded, whether it occurs initially, in the immediate postpartum, or secondarily, in the year following the delivery.
Psychological repercussions and physical complications at six and 12 months. All women determined to be at risk of placenta accreta will receive a validated questionnaire to assess the psychological impact and physical complications (such as diverse functional signs, pain, fever, vaginal bleeding or secondary surgery), regardless of the management and final diagnosis (no placenta accreta, placenta accreta and cesarean hysterectomy or conservative treatment) at six and 12 months after delivery. The self-administered questionnaire will evaluate the presence of postpartum depression, anxiety, and posttraumatic stress. These self-administered questionnaires will include: (i) the Edinburgh Postnatal Depression Scale (EPDS) [40, 41], which will allow us to assess postnatal depression syndrome; (ii) an event impact scale , which evaluates symptoms of posttraumatic stress [43, 44]; and (iii) the StateTrait Anxiety Inventory, which assesses characteristics associated with anxiety .
Physical complications in the year after delivery. Long-term complications are especially likely to occur when the placenta is left in place: secondary infections, vaginal bleeding, pain, various functional signs. These may also have a substantial psychological impact.
In addition, the information collected will specifically include:
- The standard risk factors for placenta accreta and maternal morbidity, which will be selected from among the social and demographic characteristics, and aspects of medical, obstetric and gynecological history.
- The preoperative phase, including the type of management and reasons for the decision (placenta accreta or percreta, placental site, number of children).
- Additional treatment modalities in cases of uterine conservation (interventional radiology (embolization, iliac balloons) and second line treatments for PPH, methotrexate, uterotonics).
The resources of each center will also be studied, i.e. presence of a general surgeon, vascular surgeon, urologist, and their availability (specific hours or 24/7), presence of a maternal intensive care unit and technical equipment and facilities for interventional radiology, and its availability (specific hours or 24/7).
The total number of deliveries expected in the study is 540 000 women. Of these, 11% will have had a previous cesarean section . The rates of placenta previa usually reported in the literature are in the order of 0.8–1.5% [9, 47, 48]. With a hypothesis of 1%, 620 women with a previous cesarean section and an anterior placenta previa are expected during the study. Of them, 10–40% will have a placenta accreta. This assessment corresponds to a placenta accreta rate of 1/4700, that is, among the lowest range of prevalence rates reported in the literature.
Descriptive analyses will include the incidence of women at risk in the background population and of cases of placenta accreta. We will also describe screening practices, type of management, available human and technical resources, and the incidence of maternal complications. The psychological repercussions will be described with scores based on the different questionnaires.
Analysis of the diagnostic value of MRI and ultrasound of the placenta in the population of women at risk – the sensitivity, specificity, positive and negative predictive values, positive and negative likelihood ratios – will be calculated for each sign and for some combinations of ultrasound or MRI signs. An etiological analysis within the population of women with placenta accreta will use univariate analysis to test the associations between individual characteristics and those associated with management and with different markers of maternal morbidity and then multivariate analysis to examine the independent role of each factor.
This study will be the first prospective population-based study to include women at risk of placenta accreta and to investigate incidence, prenatal detection, type of management, morbidity and maternal psychological consequences.
Funding will be provided by the french government department for medical research [regional hospital clinical research program (PHRC) 2012].