Placental Pathologic Associations With Morbidly Adherent Placenta

Abstract

Background

The pathology that underlies morbidly adherent placenta (MAP) is poorly understood. The objective of this study was to describe the placental pathology, especially implantation site pathology, associated with MAP.

Methods

This was a single institution, retrospective case-control study design examining placentas of patients who delivered between January 2008 and September 2013. MAP cases were defined by the need for clinical intervention at delivery beyond spontaneous placental delivery or simple manual extraction of the placenta. Controls consisted of patients with placentas sent for examination due to a history of maternal malignancy with no clinical suspicion of accreta. Placental pathologic findings of maternal vascular underperfusion (MVU), acute inflammation, chronic inflammation, fetal vascular obstruction, and hemorrhage were recorded and compared using bivariable and multivariable analyses.

Results

Three categories of pathologic changes were seen more commonly in MAP placentas (N = 101) than control placentas (N = 110): chronic basal inflammation, villous changes of MVU, and retromembranous and subchorionic/intervillous hemorrhage. In multivariable analyses adjusted for confounders, basal chronic villitis (aOR 5.6, 1.73–18.18), plasma cell deciduitis (aOR 2.63, 1.08–6.39), increased syncytial knots (aOR 3.92, 1.57–9.75), villous agglutination (aOR 24.85, 2.78–221.75), increased perivillous fibrin (aOR 5.08, 1.49–17.34), and the presence of subchorionic/intervillous thrombi (aOR 4.01, 1.63–9.86) remained associated with MAP.

Conclusions

MAP is highly associated with evidence of intraparenchymal placental hemorrhage, villous changes of MVU, and a lymphoplasmacytic infiltrate at the implantation site. The contribution of this basal chronic inflammatory infiltrate to MAP requires further investigation.

Keywords

placenta accreta chronic deciduitis plasma cell deciduitis basal chronic villitis maternal malperfusion

Introduction

Morbidly adherent placenta (MAP) describes abnormal placental implantation¹ with adherence of the placenta to the myometrium leading to the inability to separate the placenta from the uterine wall at the time of delivery. MAP encompasses the pathologic entities of placenta accreta (75%), placenta increta (18%), and placenta percreta (7%).¹ In addition, subclinical forms of adherent myometrium in the delivered placenta may represent risk factors for the subsequent development of placenta accreta.² MAP occurs in approximately 3 in 1000 deliveries, and the incidence has increased nearly 10-fold over the last 50 years.^1,3–5 MAP frequently requires operative delivery of the placenta to remove the adherent placental mass from the uterus, and in severe cases, hysterectomy may be required. Consequently, MAP is now the most common indication for cesarean hysterectomy.^3,6,7

Interactions at the basal plate between maternal decidual tissue, maternal inflammatory cells, and fetal extravillous trophoblast (EVT) cells likely contribute to MAP. However, the exact pathogenic mechanisms that underlie MAP are poorly understood.¹ Although defective decidualization at the site of implantation is frequently theorized as one of the important underlying etiologic factors for MAP, few studies have reviewed pathologic findings at the basal plate of the placenta in cases with MAP. Existing small studies have described the associated placenta findings in cases of accreta or adherent basal plate myometrium to include uteroplacental vascular abnormalities,⁸ decidual hemosiderosis and infarction,⁹ and acute and chronic inflammation.¹⁰ The objective of this study was to thoroughly review the placental pathologic findings, especially implantation site pathology, in a large, well-defined MAP cohort to identify potential pathogenic mechanisms for the development of MAP.

Methods

Study Design

This was a retrospective case-control study design utilizing patient data between January 2008 and September 2013 at Northwestern Memorial Hospital/Prentice Women’s Hospital. A cohort of patients with a clinical suspicion and/or pathologic diagnosis of MAP whose placentas were submitted to pathology as part of their clinical care were further refined by the need for clinical intervention at delivery beyond either spontaneous placental delivery or simple manual extraction of the placenta. Patients were included in the MAP case group if, upon review of the description of the third stage by a Maternal Fetal Medicine subspecialist (ESM), the placental removal was considered complicated (ie, requiring oversewing of the placental bed, manual extraction in multiple pieces, a D&C or a hysterectomy). Controls were selected from consecutive patients with placentas sent for pathologic examination over the same time period due to a history of maternal malignancy, as indicated on the pathology report. Inclusion as a control required that there was no clinical suspicion of MAP, and delivery was achieved with no more than simple manual extraction.

Clinical Data Collection

Clinical demographic data were extracted from medical records using a standardized abstraction data sheet and included data such as maternal age, gravity, parity, history of previous cesarean sections, and any prior uterine instrumentation. Specific details about the pregnancy included the estimated gestational age, mode of delivery, estimated blood loss at delivery, and management of the third stage of labor. In addition, for controls, the history of malignancy was confirmed on review of the medical records.

Placental Pathologic Data Collection

Placentas were examined grossly using a systematic protocol, which included recording of the trimmed placental weight, membrane insertion and gross appearance, dimensions of the placental disc, and the insertion, length, and diameter of the umbilical cord. Histologic samples included sections of membranes, umbilical cord, and 2 full thickness sections of the placental parenchyma. In addition, maternal surface biopsies, consisting of central, rectangular sections of the maternal surface of the placenta with basal plate along the long axis were submitted. All cases were originally reviewed and verified by a perinatal pathologist. Retrospective data collection was performed by the authors (LME, RLL) who were masked to case/control status. All data were recorded in a Microsoft Access (Redmond, WA) database.

Placental histologic data recorded included evidence of 3 well-characterized placental pathologic conditions: (1) maternal vascular underperfusion (MVU) (decidual vessels with mural fibrinoid necrosis/acute atherosis, muscularized basal plate arteries, mural hypertrophy of membrane arteries, decidual vascular thrombosis, infarcts, increased syncytial knots, villous agglutination, increased perivillous fibrin, distal villous hypoplasia),¹¹ (2) evidence of amnionic fluid infection/acute inflammatory pathology (acute chorioamnionitis, acute chorionic vasculitis/umbilical vasculitis, funisitis),¹² and (3) fetal obstructive vascular pathology (fetal vascular thrombosis, avascular villi).¹³ In addition, chronic inflammation at the basal plate was characterized, including basal chronic villitis characterized by lymphocytes, histiocytes, or plasma cells infiltrating the anchoring chorionic villi. In the absence of basal chronic villitis, a basal decidual chronic inflammatory infiltrate with plasma cells was considered diagnostic for plasma cell deciduitis. Finally, evidence of membranous, intervillous, or retroplacental hemorrhage was recorded.

Hematoxylin and eosin-stained slides from the placentas and, when present, hysterectomy specimen were then also reviewed to determine the pathologic stage of adherent myometrial fibers at the basal plate. The stages were defined as previously described²: Stage 0: no evidence of myometrial fibers attached to the basal plate; Stage 1: basal plate with adherent myometrial fibers, but with intervening decidua present; Stage 2: basal plate with adherent myometrial fibers, ≤ 2 layers of decidual cells separating myometrium from anchoring villi and/or Rohr’s fibrin; Stage 3: changes consistent with focal accreta—myometrium directly upon anchoring villi and/or Rohr’s fibrin with no intervening decidual cells are present; Stage 4: placenta increta with invasion into the myometrium 4A = < 25%, 4B = 25%–50%, 4C = 50%–75%, and 4D = 75%–100%; Stage 5: placenta percreta but no involvement of adjacent organs; Stage 6: invasion/attachment of adjacent organs identified.

Statistical Analysis

Chi-square or Fisher exact tests, as appropriate, were used to compare differences in the categorical variables. Student’s t-test (two tailed) was employed to determine the difference in means between continuous variables. Demographic and clinical characteristics significantly associated with MAP (P < .05) were included in multivariable logistic regressions to examine the independent pathologic associations with MAP. Analyses were performed using IBM SPSS Statistics 22 (Armonk, New York), and P < .05 was used to define statistical significance.

Results

Patient Characteristics

A total of 211 patients were included in the study, 101 MAP cases and 110 controls. The demographic characteristics of the 2 populations are shown in Table 1. Women in the MAP group were older, less likely to be non-Hispanic white, and had significantly higher gravity and parity than controls. Regarding obstetrical history, the patients in the MAP group had significantly more previous cesarean sections, they were more likely to have conceived via in vitro fertilization, and they more often had a history of uterine instrumentation than control patients. MAP patients were more likely to have placenta previa, deliver by cesarean section, and had significantly greater blood loss at delivery than controls. Finally, MAP patients delivered at a slightly earlier gestational age than control patients and had neonates with lower birth weights.

Table 1.

Demographic and Clinical Characteristics MAP Versus Control.

	MAP N = 101	Control N = 110	P Value
Maternal age, years (mean ± SD)	35.0 ± 5.3	33.6 ± 4.3	.042
Advanced maternal age	53 (52.5%)	38 (34.6%)	.009
Race, N (%)			.017
Non-Hispanic white	64 (63.3%)	91 (83.7%)
Non-Hispanic black	14 (13.9%)	5 (4.6%)
Hispanic	11 (10.9%)	4 (3.6%)
Asian	2 (2.0%)	1 (1.8%)
Other/unknown	10 (9.9%)	8 (7.3%)
Gravida (mean ± SD)	3.4 ± 1.9	2.6 ± 1.5	.001
Parity (mean ± SD)	2.3 ± 1.4	1.8 ± 0.9	.002
Range	1–9	1–4
Prior cesarean	40 (39.6%)	17 (15.5%)	<.001
Number of prior cesareans (mean ± SD)	0.69 ± 1.1	0.18 ± 0.5	<.001
In vitro fertilization, N (%)	28 (28%)	13 (12%)	.004
Hx of uterine instrumentation, N (%)	45 (44.6%)	34 (30.9%)	.041
Dilation & evacuation (D&E)	2 (2%)	5 (4.5%)
Dilation & curettage (D&C)	41 (41%)	30 (27%)
Other instrumentation	7 (7%)	3 (3%)
Previa, N (%)	34 (34%)	0	<.001
Delivery mode, N (%)			<.001
Vaginal	32 (32%)	75(68%)
C-section	66 (65%)	35(32%)
D&E	3 (3%)	0(0%)
Blood loss, mL	3413 ± 3783	522 ± 291	<.001
Gestational age, weeks (mean ± SD)	36.0 ± 4.7	39.2 ± 1.4	<.001
Term	57 (56.4%)	105 (95.5%)	<.001
Birth weight, grams (mean ± SD)	2758.6 ± 834.8	3454.5 ± 423.8	<.001
Management of third stage of labor, N (%)			<.001
Spontaneous placental delivery	0	100 (91%)
Uncomplicated manual removal	0	10 (9%)
Complicated manual removal	39 (39%)	0
D&C	19 (49%)	0
Removal in pieces	14 (36%)	0
Hysterectomy	62 (61%)	0

MAP, morbidly adherent placenta; SD, standard deviation;

Placental Pathology

Placental weight was significantly lower in the MAP group (MAP 442.6 g ± 194.1 g vs control 497.5 ± 116.7 g, P = .02), likely due to the lower gestational age at delivery in the MAP group. The maternal surface of the placenta was grossly complete in only 18% of the MAP cases while 82% of the control placentas were complete grossly (P < .001). Not surprisingly, basal adherent myometrial fibers were seen in 96% of the MAP cases (Stage 1 = 9.9%, Stage 2 = 12.9%, Stage 3 = 33.7%, Stage 4 = 28.7%, Stage 5 = 10.9%). Forty percent of control placentas also had basal adherent myometrial fibers (Stage 1 = 31.8%, Stage 2 = 8.2%), but none of the control placentas showed stages of adherence that would be diagnostic for accreta (ie, Stage 3 or above). Furthermore, none of the control placentas showed any evidence of malignancy.

Markers of chronic inflammation (Figure 1(A) and (B)) were seen more commonly in MAP cases compared to controls, specifically chronic basal villitis (17/101 cases vs 5/110 controls, P = .004), and basal chronic plasma cell deciduitis (30/101 cases vs 13/110 controls, P = .001). Central chronic villitis was not present more frequently in MAP placentas than control placentas (9/101 cases vs 13/110 controls, P = .49). While there was no increased frequency of decidual vascular pathology in MAP cases compared to controls, villous lesions of MVU were more often identified in MAP cases including: multiple infarcts (5/101 cases vs 0/110 controls, P = .024), increased syncytial knots (27/101 cases vs 14/110 controls, P = .01), and villous agglutination (11/101 cases vs 1/110 controls, P = .002). Finally, evidence of hemorrhage (remote retromembranous hemorrhage with hemosiderin deposition (28/101 cases vs 9/110 controls, P < .001) [Figure 1(C)] and subchorionic/intervillous thrombi [34/101 cases vs 12/110 controls, P < .001]) were much more commonly identified in MAP cases compared to controls (see Table 2).

Table 2.

Pathologic Characteristics of MAP and Control Placentas.

	MAP N = 101	Control N = 110	P value
Chronic inflammation	36 (36%)	25 (23%)	.039
Chronic basal villitis	17 (17%)	5 (5%)	.004
Plasma cell deciduitis	30 (30%)	13 (12%)	.001
Villous changes of MVU	42 (42%)	30 (27%)	.029
Multiple infarcts	5 (5%)	0 (0%)	.024
Increased syncytial knots	27 (27%)	14 (13%)	.010
Villous agglutination	11 (11%)	1 (1%)	.002
Increased perivillous fibrin	15 (15%)	6 (6%)	.023
Hemorrhage
Remote retromembranous hemorrhage with hemosiderin deposition	28 (28%)	9 (8%)	<.001
Subchorionic/intervillous thrombi	34 (34%)	12 (11%)	<.001

MAP, morbidly adherent placenta; MVU, maternal vascular underperfusion. Data shown as N (%)

Multivariable analyses were performed, adjusting for potential demographic and clinical confounders. Estimated blood loss and management of the third stage were excluded as they were felt to potentially be on the biologic causal pathway. Birth weight was excluded due to potential collinearity with gestational age. Markers of chronic inflammation, basal chronic villitis and plasma cell deciduitis, remained significantly associated with MAP. Similarly, villous changes of MVU, increased syncytial knots, villous agglutination, and increased perivillous fibrin, remained associated with MAP. Finally, while remote retromembranous hemorrhage with hemosiderin deposition was not independently associated with MAP, the association between presence of subchorionic/intervillous thrombi and MAP remained (see Table 3).

Table 3.

Pathologic Correlates of MAP in Multivariable Regression.

	aOR^a	95% CI
Chronic inflammation	1.28	0.60–2.74
Chronic basal villitis	5.60	1.73–18.18
Plasma cell deciduitis	2.63	1.08–6.39
Villous changes of MVU	2.16	1.02–4.59
Multiple infarcts	N/A	N/A
Increased syncytial knots	3.92	1.57–9.75
Villous agglutination	24.85	2.78–221.75
Increased perivillous fibrin	5.08	1.49–17.34
Hemorrhage
Remote retromembranous hemorrhage with hemosiderin deposition	3.16	0.22–48.82
Subchorionic/intervillous thrombi	4.01	1.63–9.86

MAP, morbidly adherent placenta; MVU, maternal vascular underperfusion.

Adjusted for maternal age, race/ethnicity, parity, prior cesarean, conception via invitro fertilization, prior uterine instrumentation, route of delivery, and gestational age.

Discussion

Our data demonstrate that, in addition to adherent myometrium, MAP is associated with other significant placental pathology related to chronic inflammation, maternal vascular underperfusion, and hemorrhage. Previous smaller studies of the placental pathology associated with adherent myometrial fibers and placenta accreta have identified similar findings including uteroplacental vascular abnormalities,⁸ placental infarction, and hemosiderosis,⁹ and both acute and chronic inflammation in the myometrium adjacent to areas of placenta accrete.¹⁰ However, this is the largest and most systematic study, to our knowledge, to examine the placental pathologic features associated with MAP.

Chronic inflammation in the placenta is relatively poorly understood. Chronic inflammatory placental lesions can be caused by infectious agents such as cytomegalovirus, Toxoplasma gondii, and Treponema pallidum, but most chronic inflammatory lesions do not have demonstrable infectious agents.¹⁴ When infection can be excluded, the cause is typically unknown, but mounting evidence suggests that maternal anti-fetal rejection plays a role.¹⁴ Chronic deciduitis defined by the presence of lymphocytes and plasma cells in the basal plate has an incidence of 1%–4%.^15,16 While the exact etiology of chronic deciduitis is not known, both infectious and immune-mediated etiologies have been implicated.¹⁴ The presence of a chronic inflammatory infiltrate in MAP cases, as shown in our study, may represent an abnormal immune response at the maternal fetal interface. This chronic inflammation may be present as a reaction to the advancing trophoblastic tissue and/or may play a role in controlling its invasiveness.

A role has been described for certain subsets of lymphocytes, such as natural killer cells and regulatory T-cells, in placental implantation and maternal vascular remodeling.^17,18 One aspect of placental implantation, proper formation of the decidua, is critical to establish the normal separation of the placenta as well as to control the “invasiveness” of EVT cells.¹⁹ The normal cellular changes in the endometrium associated with decidualization of endometrial stromal cells appear to be accompanied by expression of different proinflammatory cytokines, chemokines and growth factors at implantation, and successful implantation requires a highly regulated interaction between inflammatory cells and decidual cells.²⁰ Inflammatory cells like decidual natural killer (dNK) cells have been shown to play a role in controlling this aspect of placental implantation and vascular remodeling.^17,18,21

Recent studies have examined the role of chronic inflammation at the basal plate in placenta accreta. Laban et al.²¹ explored the role of dNK cells in the pathogenesis of placenta accreta and found that a decreased population of dNK cells was closely associated with cases of MAP. They suggested that increased dNK cells may actually be protective against deep EVT invasion or MAP. Schwede et al.²² observed an increased number of regulatory T-cells and decreased number of CD4 + T-cells in MAP cases also suggesting a role in regulating trophoblast invasion. Although further characterization of the specific cell types in the inflammatory infiltrate at the basal plate in our MAP cases was not performed, the presence of plasma cells suggests that this infiltrate is abnormal since plasma cells are not a normal component of the endometrial stroma.²³ Further investigation of the specific cell types seen in MAP cases may advance our understanding of the abnormal implantation in accreta cases.

Changes of maternal vascular underperfusion related to poor placental implantation such as placental infarcts, increased syncytial knots, and villous agglutination can be explained by implantation in an inhospitable site, such as the lower uterine segment or over a cesarean section scar, as is common in MAP. Implantation in regions of the uterus with inadequate decidualization could lead to abnormal or incomplete establishment of the vasculature needed to support the developing placenta.²⁴ Abnormal implantation and placenta previa are well-known antecedents of intrauterine bleeding,²⁴ and therefore it is not surprising that evidence of intervillous and retromembranous placental hemorrhage were associated with MAP in this study.

The increasing incidence of accreta has been linked to the parallel increasing rate of cesarean section; however, there are many cases of placenta accreta that occur without this or other risk factors.³ Therefore, it is important to discover other possible risk factors in the development of accreta not only to identify potential women that could be affected but also to further elucidate the pathogenesis of this lesion. This is the largest descriptive study to date of the placental features of MAP, and its strengths include that placental pathology was well characterized by a perinatal pathologist. In addition, cases were defined based on a priori dichotomized clinical descriptions of the third stage. This allows for a more specific clinical definition of abnormal placental adherence. However, this study does have limitations. The most important is that it is retrospective and biased by examination of only those placentas sent to pathology at the time of delivery. Furthermore, the control group may be biased by the selection of placentas submitted to pathology for a history of maternal malignancy. Given the retrospective study design, our goal was to acquire the most nearly normal population of placentas for the control group. Many of the placentas sent to pathology for maternal melanoma were sent only solely because of the history of maternal melanoma, and the pregnancy was otherwise uncomplicated. It is possible the presence of maternal melanoma introduces a heretofore unrecognized placental pathology related to abnormal placental adherence, but then we would expect the bias would make it more difficult to find distinctive placental pathologic differences between the two groups, which was not the case in our study. Finally, at our institution, placentas with a suspicion of abnormal adherence are routinely sent to pathology, but there is no way to be certain that all cases of clinical MAP were incorporated. Given these limitations, our study represents a step forward in the understanding of the placental pathology associated with MAP.

In summary, work from our group has shown that MAP can be recognized at the basal plate of the delivered placenta² in patients with clinical diagnosis of MAP, and that accompanying placental features include chronic basal inflammation, evidence of uteroplacental underperfusion, and intervillous hemorrhage. While these associated placental findings should not be considered specific for placental accreta or MAP, they may have potential important implications for understanding the pathogenesis of placenta accreta/MAP.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Abuhamad

. Morbidly adherent placenta. Semin Perinatol 2013; 37(5): 359–364.

Linn

Miller

Lim

Ernst

. Adherent basal plate myometrial fibers in the delivered placenta as a risk factor for development of subsequent placenta accreta. Placenta 2015; 36(12): 1419–1424.

Committee on Obstetric Practice. Committee opinion no. 529: placenta accreta. Obstet Gynecol 2012; 120(1): 207–211.

Ballas

Pretorius

Hull

Resnik

Ramos

. Identifying sonographic markers for placenta accreta in the first trimester. J Ultrasound Med 2012; 31(11): 1835–1841.

Lorenz

. What is new in placenta accreta? Best articles from the past year. Obstet Gynecol 2013; 121(2 Pt 1): 375–376.

Amsalem

Kingdom

Farine

et al.

Planned caesarean hysterectomy versus “conserving” caesarean section in patients with placenta accreta. J Obstet Gynaecol Canada 2011; 33(10): 1005–1010.

Angstmann

Gard

Harrington

Ward

Thomson

Giles

. Surgical management of placenta accreta: a cohort series and suggested approach. Am J Obstet Gynecol 2010; 202(1): e31–39.

Sherer

Salafia

Minior

Sanders

Ernst

Vintzileos

. Placental basal plate myometrial fibers: clinical correlations of abnormally deep trophoblast invasion. Obstet Gynecol 1996; 87(3): 444–449.

Stanek

Drummond

. Occult placenta accreta: the missing link in the diagnosis of abnormal placentation. Pediatr Dev Pathol 2007; 10(4): 266–273.

10.

Fox

Sebire

. Pathology of the Placenta, Oxford: Elsevier Health Sciences, 2007.

11.

Redline

Boyd

Campbell

et al.

Maternal vascular underperfusion: nosology and reproducibility of placental reaction patterns. Pediatr Dev Pathol 2004; 7(3): 237–249.

12.

Redline

Faye-Petersen

Heller

Qureshi

Savell

Vogler

. Amniotic infection syndrome: nosology and reproducibility of placental reaction patterns. Pediatr Dev Pathol 2003; 6(5): 435–448.

13.

Redline

Ariel

Baergen

et al.

Fetal vascular obstructive lesions: nosology and reproducibility of placental reaction patterns. Pediatr Dev Pathol 2004; 7(5): 443–452.

14.

Kim

Romero

Chaemsaithong

Kim

. Chronic inflammation of the placenta: definition, classification, pathogenesis, and clinical significance. Am J Obstet Gynecol 2015; 213(4 Suppl): S53–S69.

15.

Bendon

Miller

. Routine pathological examination of placentae from abnormal pregnancies. Placenta 1990; 11(4): 369–370.

16.

Naeye

. Functionally important disorders of the placenta, umbilical cord, and fetal membranes. Hum Pathol 1987; 18(7): 680–691.

17.

Eastabrook

von Dadelszen

. The role of decidual natural killer cells in normal placentation and in the pathogenesis of preeclampsia. J Obstet Gynaecol Canada 2008; 30(6): 467–476.

18.

Wallace

Fraser

Cartwright

. Extravillous trophoblast and decidual natural killer cells: a remodelling partnership. Hum Reprod Update 2012; 18(4): 458–471.

19.

Tantbirojn

Crum

Parast

. Pathophysiology of placenta creta: the role of decidua and extravillous trophoblast. Placenta 2008; 29(7): 639–645.

20.

Mor

Cardenas

Abrahams

Guller

. Inflammation and pregnancy: the role of the immune system at the implantation site. Ann N Y Acad Sci 2011. 1221:80–87.

21.

Laban

Ibrahim

Elsafty

Hassanin

. Placenta accreta is associated with decreased decidual natural killer (dNK) cells population: a comparative pilot study. Eur J Obstet Gynecol Reprod Biol 2014; 181: 284–288.

22.

Schwede

Alfer

von Rango

. Differences in regulatory T-cell and dendritic cell pattern in decidual tissue of placenta accreta/increta cases. Placenta 2014; 35(6): 378–385.

23.

Katzman

. Chronic inflammatory lesions of the placenta. Semin Perinatol 2015; 39(1): 20–26.

24.

Heller

. Placenta accreta and percreta. Surg Pathol Clin 2013; 6(1): 181–197.