Original ajog.

org
Research
OBSTETRICS
Placenta percreta is associated with more frequent
severe maternal morbidity than placenta accreta
Louis Marcellin, MD, PhD; Pierre Delorme, MD; Marie Pierre Bonnet, MD, PhD; Gilles Grange, MD; Gilles Kayem, MD, PhD;
Vassilis Tsatsaris, MD, PhD; François Goffinet, MD, PhD

BACKGROUND: Abnormally invasive placentation is the leading cause
of obstetric hysterectomy and can cause poor to disastrous maternal
outcomes. Most previous studies of peripartum management and
maternal morbidity have included variable proportions of severe and less
severe cases.
OBJECTIVE: The aim of this study was to compare maternal morbidity
from placenta percreta and accreta.
STUDY DESIGN: This retrospective study at a referral center in Paris
includes all women with abnormally invasive placentation from 2003
through 2017. Placenta percreta and accreta were diagnosed histologi-
cally or clinically. When placenta percreta was suspected before birth, a
conservative approach leaving the placenta in situ was proposed because
of the intraoperative risk of cesarean delivery. When placenta accreta was
suspected, parents were offered a choice of a conservative approach or an
attempt to remove the placenta, to be followed in case of failure by hys-
terectomy. Maternal outcomes were compared between women with
placenta percreta and those with placenta accreta/increta. The primary
outcome measure was a composite criterion of severe acute maternal
morbidity including at least 1 of the following: hysterectomy during ce-
sarean delivery, delayed hysterectomy, transfusion of≤10 U of packed
red blood cells, septic shock, acute kidney injury, cardiovascular failure,
maternal transfer to intensive care, or death.
RESULTS: Of the 156 women included, 51 had placenta percreta and
105 placenta accreta. Abnormally invasive placentation was suspected
antenatally nearly 4 times more frequently in the percreta than the accreta
group (96.1% [49/51] vs 25.7% [27/105], P < .01). Among the 76
women with antenatally suspected abnormally invasive placentation
(48.7%), the rate of antenatal decisions for conservative management was
higher in the percreta than the accreta group (100% [49/49] vs 40.7%
[11/27], P < .01). The composite maternal morbidity rate was significantly
higher in the percreta than the accreta group (86.3% [44/51] vs 28/105
[26.7%], P < .001). A secondary analysis restricted to women with an
abnormally invasive placentation diameter >6 cm showed similar results
(86.0% [43/50) vs 48.7% [19/38), P < .01). The rate of hysterectomy
during cesareans was significantly higher in the percreta than the accreta
group (52.9% [27/51] vs 20.9% [22/105], P < .01) as was the total
hysterectomy rate (43/51 [84.3%] vs 23.8% [25/105], P < .01).
CONCLUSION: Severe maternal morbidity is much more frequent in
women with placenta percreta than with placenta accreta, despite
multidisciplinary planning, management in a referral center, and better
antenatal suspicion.
Key words: hemorrhage, maternal death, maternal morbidity, obstetric
hysterectomy

Introduction
Abnormally invasive placentation
(AIP) is clinically observed when a
placenta cannot be separated from a
uterus.1 Although its etiopathogenesis
is un- clear,2 it can cause poor to
disastrous maternal outcomes at
delivery,3 mainly due to severe
postpartum hemorrhage, and remains
the leading cause of peri- partum
hysterectomy in Western coun- tries.4
Over the last 3 decades, as the
cesarean delivery rate5 and maternal
age6 have climbed, population-based
studies have shown a notable rise in
the
Cite this article as: Marcellin L, Delorme P, Bonnet MP,
et al. Placenta percreta is associated with more frequent
severe maternal morbidity than placenta accreta. Am J
Obstet Gynecol 2018;219:193.e1-9.
0002-9378/$36.00
ª 2018 Elsevier Inc. All rights reserved.
https://doi.org/10.1016/j.ajog.2018.04.049
incidence of AIP, which now appears
to affect from 0.17e0.34% of
deliveries.7e9 Three grades of AIP are
usually distinguished, with severity
increasing with the depth of placental
invasion into the myometrium:
placenta accreta (decidual
layer), increta (myometrium), and
percreta (serosa and adjacent
organs). Placenta percreta often
invades the bladder wall and can be
asso- ciated with massive adnexal
hyper- vascularization that makes
cesarean delivery difficult to perform
and in- creases the risk of
intraoperative and postoperative
complications.10,11 Prac-
tices for the management of AIP vary
widely. Although planned cesarean
hys- terectomy without attempted
removal of the placenta remains the
recommended practice,12 a placenta percreta from the other
conservative approach has been
developed over the past 20 years to
preserve fertility. Its use has been
encouraged, especially in France, to limit the context of management with either a
maternal morbidity in women with
placenta percreta.13,14 The recent
Inter- national Federation of
Gynecology and Obstetrics consensus
guidelines recom- mend “leaving the
placenta in situ” as a suitable option
with close follow-up in hospitals with
adequate expertise.15
The rising incidence of AIP has led
to the publication of numerous cohort
studies about its management and
maternal morbidity.8,9 These
generally report severe maternal
morbidity and complications for the
entire population, without
distinguishing the disease phenotype
(accreta/increta or percreta),
and very few studies on placenta
percreta are available.16 We therefore
studied AIP in a large cohort from a
referral center while distinguishing
forms of AIP, with the aim of
comparing maternal outcome in these
2 groups in

AUGUST 2018 American Journal of Obstetrics & Gynecology 193.e1

Original Research ajog.or
OBSTETRICS g
AJOG at a Glance
beyond the myometrium and to
Why was this study conducted? estimate the largest diameter of the
Most previous studies have included variable proportions of severe and less placental fragment left in situ. Given
severe cases of abnormally invasive placenta. This study was conducted to that the normal placental diameter at
compare the outcomes of women with placenta percreta and placenta term is 22 cm,17 a threshold of 6 cm
accreta/increta. was considered to distinguish the
largest pieces of placenta left in situ.
Key findings This threshold thus functions as a
Severe maternal morbidity is 3 times more frequent in patients with sensitivity analysis to eliminate the
placenta percreta than in those with accreta/increta, although placenta cases with a relatively small adherent
percreta is sus- pected antenatally almost 4 times more frequently than surface, that is, those with accreta
placenta accreta. diagnoses that might be debat- able
without a hysterectomy and histo-
What does this add to what is known? logical analysis.
This study clearly demonstrates the higher maternal risk for women with The primary outcome measure was a
placenta percreta and shows that future studies of abnormally invasive composite criterion of severe acute
placentation should distinguish placenta percreta from its other forms. maternal morbidity
occurring throughout
conservative approach or a planned This study included all women fol- pregnancy and the post- partum
ce- sarean hysterectomy. lowed up at or referred to our center period, including at least 1 of the
who gave birth >20 weeks of following: hysterectomy during
Materials and Methods gestation and had diagnosed AIP, that cesarean delivery, delayed
This retrospective cohort study exam- is, AIP confirmed histologically or hysterectomy, trans- fusion of 10
ined the records of all patients with clinically, according to the criteria U of packed ≤ red blood cells,18 septic
AIP treated in our reference center described below. In cases of shock (defined as sepsis with
from January 2003 through May hysterectomy, placenta persisting hypotension requiring
2017. We began by systematically accreta/increta was histologically vasopressors despite adequate volume
searching the computerized database diag- nosed when placental villi had resuscitation),19 acute kidney injury
of the Port-Royal Maternity Hospital, invaded the myometrium without (defined as creatinine 2.0 mg/dl in a
a tertiary university center reaching the uterine serosa. Placenta ≤
woman without preexisting kidney
experienced in placenta accreta, with accreta/increta was clinically dis- ease),19 cardiovascular failure
the key words “accreta,” “increta,” diagnosed when no (total or partial) (defined as hemodynamic
“percreta,” “abnormally invasive cleavage plane was found between instability requiring
placentation,” and “hemorrhage.” The the placenta and the uterus during catecholamine infusion),
search identified 200 patient files with manual removal, or after failure of a maternal admission to the
suspected AIP during the study gentle attempt to remove the intensive care unit, or maternal
period. Review of the patient file placenta in cases with antenatally death. Other maternal
summaries in the database enabled us sus- outcome measures were the adminis-
to exclude 33 women without pected AIP. tration of packed red blood cells,
abnormal placentation. Review of the The diagnosis of placenta percreta fresh frozen plasma, and platelets
original obstetric files of the required, in addition to the criteria for (with the number of packs used for
remaining 167 cases resulted in the placenta accreta, a histologic finding each), hypo- gastric artery ligation,
exclusion of 11 additional files, that placental villi had invaded the pelvic arterial embolization, primary
because the suspected AIP was not myome- trium, reached the uterine or delayed hys- terectomy, overall
confirmed. Data were collected from serosa, and might have invaded hysterectomies, and intraoperative
each (paper) medical file. Two adjacent organs; or clinical complications. Finally, neonatal
experienced obstetri- cians reviewed observation of gross invasion of the outcomes were birthweight, 5-
all the cases selected to assess the uterine serosa or adjacent organs in minute Apgar score <7, umbilical
accuracy of the diagnoses of placenta cases of successful conservative man- artery pH, and transfer to intensive
accreta/increta or percreta. This study agement during cesarean delivery not care. Management remained broadly
was approved by the National Data requiring hysterectomy. The study ho- mogeneous during the study
Protection Authority (Commission did not include women whose period. All women with antenatally
Nationale de l’Informatique et des antenatally suspected AIP was not suspected AIP (due to routine
Lib- ertés no. 1755849). During the clinically or his- tologically ultrasound) were referred for
study period, women were routinely confirmed. For the purpose of this ultrasonography by an experienced
informed that their records could be study, women with placenta increta operator who applied the standard
used for the evaluation of medical were allocated to the accreta group diagnostic criteria, detailed
practices and that they had the right and only women with placenta elsewhere.20 However, women
to opt out of these studies. percreta to the percreta group. whose routine ultrasound did not
The operator was systematically result in antenatal suspicion of AIP
asked to specify the depth of the AIP were not referred to an experienced
into and sonographer or subsequently for
magnetic resonance imaging (MRI).
Those with antenatally

193.e2 American Journal of Obstetrics & Gynecology AUGUST 2018

ajog.or OBSTETRICS Original
g Research
products. In the absence of any
TABLE 1 patho- logical event earlier in the
Demographic and obstetric characteristics of study population pregnancy, a cesarean was planned at
36 weeks of gestation for women
Percreta Accreta with antenatally suspected placenta
n ¼ 51 n ¼ 105 P percreta, and at 38 weeks when
.06 placenta accreta was suspected.
Age, y, mean T SD 35.1 T 4.8 37.0 T 5.8 At our center, we offered 2
Age >35 y, n (%) 28 (54.9) 64 (60.9) .47 obstetric strategies to couples with
.02
antenatally suspected placenta
BMI, kg/m2, mean T SD 25.9 T 5.1 23.9 T 4.7 accreta: either a ce- sarean delivery
Ethnic origin, n (%) .23 with the radical option of cesarean
hysterectomy planned if attempted
● White 22 (43.1) 54 (51.4) .30 manual removal of the placenta failed
● North African 7 (13.7) 14 (13.4) (so-called primary hys- terectomy,
defined as a hysterectomy
● Sub-Saharan African 17 (33.4) 23 (21.9)
● Other
Referral from another facility, n (%) 405 (9.8)
(78.4) 14 (13.3)
42 (40.0) <.01 the first 24during
performed hoursa cesarean or withinor
after delivery),
Gestational age at transfer, wk, mean T SD 29.6 T 3.8 27.7 T 6.0 .09
conservative management, that is, leav-
Twin pregnancy, n (%) 1 (1.9) 12 (11.4) .04 ing the placenta in situ, as previously
Gravidity, mean T SD 4.0 T 2.1 2.5 T 2.0 <.01 reported.22 The latter strategy included
● Null, n (%) 0 (0) 15 (14.3) <.01 an intravenous injection of
prophylactic
● 1, n (%) 3 (5.9) 17 (16.2) <.01 oxytocin (5 IU) and the systematic
per- formance of controlled cord
● ≤2, n (%) 48 (94.1) 73 (69.5) traction to
avoid leaving nonaccreta
our policy was toplacentas
advocate
in conservative management to reduce
Parity, mean T SD 2.7 T 1.5 1.1 T 1.2 <.01 situ. the risk of acute hemorrhage,
1 (2.0) 38 (36.2) <.01 For women
although thewith placenta
couple could percreta,
insist on
● Null, n (%)
hysterectomy. When potential
● 1, n (%) 8 (15.7) 32 (30.5) intraoperative difficulties due to
● ≤2, n (%) 42 (82.4) 35 (33.3) abnormal and extensive pelvic
hyper- vascularization were expected,
Previous cesareans, mean T SD 2.1 T 0.9 0.7 T 1.0 <.01 adjuvant measures such as the
● Null, n (%) 1 (2.0) 57 (54.3) <.01
placement of ureteric catheters or a
preoperative intravascular balloon
● 1, n (%) 12 (23.5) 23 (21.9) catheter were considered at the
● 2, n (%) multidisciplinary meeting, where a
22 (43.1) 12 (11.4)
● ≤3, dilatation
Previous n (%) and curettage, n (%) 24 (47.1) 52 (49.5) .77 radiologist was present and
16 (31.4) 13 (12.4) participated in the decision making.
Previous placenta accreta, n (%) 2 (3.9) 4 (3.8) .97 When the placenta was left in situ
At least 1uterine
Previous previous cesarean,
rupture, n (%)n (%) 50 (98.0)
1 (1.9) 481 (45.7)
(1.0) <.01
.59 during conservative
● Myomectomy, n (%)
Previous uterine surgery, n (%) 3 (5.9) 28 (26.6) <.01
management, women were
discharged home and monitored
● Operative hysteroscopy, n (%) 0 (0) 6 (5.7) weekly on an outpatient ba- sis, with
3 (2.2) 22 (20.9) urinary and vaginal bacteriology
cultures, measurements of
suspected AIP had MRI to assess the following specialists: obstetricians, temperature and inflammatory
depth and location of the placental in- an- esthesiologists, radiologists, markers (complete blood cell count
vasion more accurately and to map pediatri- cians, urologists, and and C-reactive protein assay), and a
the potential vascularization due to colorectal surgeons. In all cases of uterine ultrasound, as pre- viously
atypical
BMI, body neoangiogenesis.
mass index.
21
All these antenatally sus- pected AIP, reported.23 A delayed hysterec- tomy
cases
Marcellinwere
et al. Severe then systematically
maternal morbidity: anesthetic
more frequent in placenta percreta than strategy was
accreta. Am J Obstet planned
Gynecol 2018. (after the first 24 hours after
discussed at a struc- tured during a specific consultation, with a delivery) was performed only on an
multidisciplinary meeting bringing pretransfusion assessment and emergency basis in cases of abundant
together at least 1 of each of the appropriate advance storage of blood bleeding, endometritis, pelvic pain,
or septicemia. In the absence of
signs of complications (bleeding or
infection), follow-up continued until
complete

AUGUST 2018 American Journal of Obstetrics & Gynecology 193.e3

Original Research ajog.or
OBSTETRICS g
TABLE 2
Diagnostic modalities of abnormally invasive placentation and for antenatally suspected AIP, and
characteristics of delivery, in percreta and accreta groups clin- ical examination for diagnosis of
AIP during cesarean or after vaginal
Percreta Accreta delivery. Anesthesia could be general,
n ¼ 51 n ¼ 105 P locore- gional, or both. Preoperative
ureteric stent placement and
Diagnostic modalities of AIP, n (%) intravascular balloon catheter
Antenatal suspicion of AIP 49 (96.1) 27 (25.7) placement were also documented.
<.01 The analysis started by comparing
Intrapartum diagnosis of AIP 2 (3.9) 78 (74.3) <.01 the percreta and accreta groups for
2 (3.9) 58 (55.3) their general demographic and
● During cesarean delivery obstetric characteristics, modalities of
● After vaginal delivery 0 (0) 20 (19.0) AIP diag- nosis, and characteristics of
Delivery characteristics delivery. The main analysis compared
them for the composite criterion of
Gestational age at delivery, wk T SD 34.6 T 3.0 34.7 T 4.9 .08 severe acute maternal morbidity and
for each indi- vidual outcome. To
● <24 wk, n (%) 0 (0) 3 (2.9) limit the potential inclusion bias
● 24e31 wk, n (%) 9 (17.7) 26 (24.8) linked to the focal AIP case in the
● 32e36 wk, n (%) 27 (52.9) 26 (24.8)
accreta group and assess the
robustness of the results, a secondary
● ≤37 wk, n (%) 15 (29.4) 50 (47.6) .05 analysis of the composite criterion
Mode of delivery, n (%) included only the subgroup of women
.01
with AIP measuring >6 cm.
● Vaginal 0 (0) 21 (20.0) Continuous variables were reported
as means and compared between
● Cesarean 51 (100) 84 (80.0)
groups with Student t test. Univariate
B Planned 26/51 (51.0) 25/84 (29.8) .03 analyses were per- formed.
B Emergency 25/51 (49.0) 59/84 (70.2)
Proportions were compared by
Anesthesia modalities, n (%) the c2 or Fisher exact test, as appro-
● General priate. A P value <.05 was considered
● Locoregional 21 (41.2) 21 (20.0) <.01 statistically significant. The analysis
was performed with Stata 11.0
● Locoregional and general 11 (21.6) 61 (58.1) (StataCorp, College Station, TX).
Preoperative ureteric catheter 19 (37.2) 23 (21.9)
placement, n (%)
24 (47.1) 6 (5.7) <.01
Results
Preoperative balloon catheter During the study period, Port-Royal
placement, n (%) Maternity Hospital provided prenatal
7 (13.7) 0 (0) <.01 and perinatal care for 156 women
Extent of abnormal placental with AIP: 51 with placenta percreta
invasion, n (%) (in the percreta group) and 105 with
≤6 cm placenta accreta or increta (accreta
group).
>6 cm or resorption,
placental totally adherenteven when the 1Demographic
(2.0) 67and obstetric charac-
(63.8) The 2 groups were comparable for
remaining
AIP, abnormallytissue was well tolerated. teristics retrieved included: maternal
invasive placentation. maternal demographic characteristics,
Methotrexate was not included in our age, 50 (98.0)
body mass38index, (36.2) ethnic origin,
except for body mass index,
management strategy. referral from another facility, gravidity, parity, and previous
The management
Marcellin et al. Severe maternal of unsuspected
morbidity: gestational
more frequent in placenta age
percreta than at transfer,
accreta. gestational
Am J Obstet Gynecol 2018. cesareans, all higher in the percreta
cases diagnosed during delivery age at and mode of delivery, group. Women in the percreta group
depended on the occurrence of a gravidity, parity, twin pregnancies, were also more likely to have been
concomitant hemor- rhage. In cases history and number of ce- sarean referred antenatally from another
of heavy bleeding, a pri- mary deliveries, and history of uterine facility (78.4% vs 40.9%, P <
hysterectomy (during cesarean) was surgery, dilatation and curettage, AIP,
performed. Otherwise, a conservative and uterine rupture. Diagnostic .001) (Table 1).
approach might be chosen, that is, modal- ities included imaging AIP was suspected antenatally in
leaving all or a part of the placenta in (ultrasound and MRI) of placental 76 cases (48.7%), nearly 4 times
situ. location and insertion more frequently in the percreta than
the accreta group (96.1% [49/51] vs
25.7% [27/105], P < .01), and MRI
was per- formed >4 times as often
(respectively 88.2% [45/51] vs 19.0%
[20/105],

193.e4 American Journal of Obstetrics & Gynecology AUGUST 2018

ajog.or OBSTETRICS Original
g Research
her focal placenta percreta, located in
TABLE 3 the fundus, was discovered during the
Severe maternal morbidity according to form of abnormally invasive ce- sarean performed at 33 weeks’
placentation: comparison between placenta percreta and placenta accreta gestation for chorioamnionitis and
group premature rupture of membranes.
Conservative management was
Percreta Accreta plan- ned for the 49 women with
n ¼ 51 n ¼ 105 P antenatally suspected AIP in the
percreta group. Among the 27 with
Composite acute maternal morbidity score, n (%) 44 (86.3) 28 (26.7) <.01 antenatally suspected AIP in the
● Hysterectomy during cesarean 27 (52.9) 22 (20.9) <.01 accreta group, conservative
management was planned for 11
● Delayed hysterectomy 17/24 (70.8) 4/83 (4.8) <.01 (40.7%) and cesarean hysterectomy
● Packed RBC transfused ≤10 U 10 (19.6) 5 (4.8) <.01 for 16 (59.3%).
Mean gestational age at delivery
● Septic shock 5 (9.8) 2 (1.9) .02 was similar in the 2 groups
● Kidney injury 3 (5.9) 1 (0.9) .07 T (34.6
3.0 vs
● Cardiovascular failure 1 (1.9) 1 (0.9) .6 34.7 T4.9 weeks, P ¼.08) but the rate of
delivery between 32e36 weeks was
● Admission to intensive care unit 14 (27.4) 3 (2.9) <.01
higher in the percreta group (52.9%
● Death 2 (3.9) 0 (0) .04 vs
Maternal morbidity composite 24.8 %, P ¼ .05). All the women in
43/50 (86.0) 19/38 (50.0) <.01
score for women with AIP with the percreta group (n ¼ 51) had
diameter >6 cm,a n (%) cesareans, compared with 80.9% (n
¼ 85) in the
Intraoperative transfusion of RBC U, n (%) 35 (68.6) 50 (47.6) .01 accreta group. At delivery, the rate of
cases with AIP >6 cm in diameter
● RBC U, mean T SD 8.4 T 7.3 4.2 T 3.2 <.01 was much higher in the percreta than
● FFP U, mean T SD 6.6 T 6.1 1.8 T 2.8 <.01 the accreta group (98.0% [50/51] vs
36.2%
● Platelet U, mean T SD 1.2 T 1.0 0.2 T 0.4 <.01 [38/105], P < .01) (Table 2).
Hypogastric artery ligation, n (%) 6 (11.7) 17 (16.2) .46 Neonatal outcomes did not differ
be- tween the percreta and accreta
Pelvic arterial embolization, n (%) 15 (29.4) 11 (10.4) <.001 groups. There were 2 neonatal deaths,
Delayed hysterectomy, n (%) 17/24 (70.8) 4/83 (4.8) 1 termi- nation of pregnancy, and 3
intrauterine fetal deaths—all in the
● Median interval, d T SD 54.4 T 47.7 23.0 T 9.8 accreta group (Supplemental Table).
● For septic shock, n (%) 7/17 (41.2) 1/4 (25.0) The composite maternal morbidity
● For secondary postpartum 8/17 (47.1) 2/4 (50.0)
hemorrhage n (%) rate was >3 times higher in the
percreta than the accreta group (86.3%
[44/51] vs
● For hematuria n (%) 1/17 (5.8) 0 (0) 26.7% [28/105], P < .01), and >5 times
● For retention n (%) 1/17 (5.8) 1/4 (25.0) higher in women with AIP >6 cm
Final hysterectomy rate, n (%) 44 (86.3) 26 (24.8) <.01 than in those with ≤ AIP 6 (70.4%
Other intraoperative complications n (%) [62/88] vs 13.2% [9/68], P < .01). In
the subgroup of 60 women managed
Disseminated intravascular coagulation 4 (7.8) 0 with a conser- vative approach, the
Bladder injuries 34 (66.7) 3 (2.9) composite maternal morbidity rate
also remained higher in the percreta
Partial bladder resection 6 (11.8) 0 group (43/49 [87.8%] vs 1/11 [9.1%],
Vesicovaginal fistula 3 (5.9) 0 P < .01). Similarly, among the 76
women with antenatally suspected
Ureteral reimplantation 2 (3.9) 0 AIP, this severe maternal mortality
remained more frequent in the
Marcellin et al. Severe maternal morbidity: more frequent in placenta percreta than accreta. Am J Obstet
Gynecol 2018. (continued)
percreta than the accreta group (87.8%
[43/49] vs
62.9 [17/27] P¼.01). Finally, when the
analysis was restricted to the cases
with a
P <.01) (Table 2). In the subgroup of [20/38], P < .01. AIP was antenatally suspected at ul-
88 trasound for all but 2 women in the
cases with AIP >6 cm, the rate of percreta group. One of them had 2 pre-
ante- natal suspicion of AIP remained vious cesareans and an overlying
higher in the percreta than accreta placenta. The other woman had under-
group, but the difference was less gone a hysteroscopic septum resection;
marked (96.0% [48/50] vs 52.6%
AUGUST 2018 American Journal of Obstetrics & Gynecology 193.e5
Original Research
histological diagnosis (ie, those
ajog.or
with a hysterectomy), severe
OBSTETRICS g
maternal morbidity (here, excluding
hysterec- tomy) was again
significantly higher in the percreta
group (45% [23/45] vs 7.6% [8/105],
P < .01).

193.e6 American Journal of Obstetrics & Gynecology AUGUST 2018

ajog.or
TABLE 3 OBSTETRICS Original
g Severe maternal morbidity according to form of abnormally invasive Research
placentation: comparison between placenta percreta and placenta accreta
group (continued) with AIP have not taken the depth of
placental invasion into account or
Percreta Accreta distinguished placenta accreta from
n ¼ 51 n ¼ 105 P placenta percreta.7,13,22,25 The major
strength of this study is the number of
cases from a single reference
maternity center with homogeneous
Uterine rupture 1 (1.9) 1 (0.9) management over the study period.
This large cohort makes it possible to
Acute postoperative hemoperitoneum 2 (3.9) 1 (0.9) compare the 2 study groups according
Occlusion syndrome 2 (3.9) 0 to the severity of the AIP. Two
experienced obstetricians reviewed
Deep iliofemoral thrombophlebitis 4 (7.8) 0 all medical files.
Iliac vessel injury 3 (5.9) 0 This study is, nonetheless, limited by
AIP, abnormally invasive placentation; FFP, fresh frozen plasma; RBC, red blood cells. its retrospective design. We cannot
Pelvic
a packing
After excluding 68 cases ≤6 cm. 4 (7.8) 0 pro- vide a histologic diagnostic
Marcellin et al. Severe maternal morbidity: more frequent in placenta percreta than accreta. Am J Obstet Gynecol 2018. confirmation in the cases of
successful conservative management
without hysterectomy. However, in
The secondary analysis, including hemorrhage. She died in the intensive the absence of hysterec- tomy, clinical
only AIP with a diameter >6 cm care unit a few days after surgery diagnosis of AIP is avail- able when
(thus excluding 1 case of percreta and from multiorgan failure. The second the placenta is left in situ according
67 cases of accreta), showed similar maternal death occurred after a to the criteria previously
results (86.0% [43/50] vs 50.0% planned cesarean delivery at 37 published.13,22,26
[19/38], P < weeks for severe placenta percreta. Most women were referred to our
.01) (Table 3). Each individual Immediately after the early fundal center after their antenatal providers
compo- nent of acute maternal delivery of a healthy child in good began to suspect AIP, placenta
morbidity except acute heart failure condition and a conservative percreta in particular; this might have
was significantly more frequent in approach (with no attempt to remove induced selection bias, due to the
women with placenta per- creta than the placenta), massive intraoperative over- representation of the most severe
placenta accreta (Table 3). hem- orrhaging began from the cases in our series. Accordingly, the
The rate of hysterectomies during uteroplacental bed and led to an proportion of placenta accreta/increta
ce- sareans was higher in the percreta emergency hemostatic hysterectomy. to placenta percreta in the population
than in the accreta group (52.9% During the hysterectomy procedure, cannot be estimated in this study.
the aberrant right periuterine vascular This point does not, however, call
[27/51] vs 20.9% [22/105], P < .01). network also began to bleed into question the interpretation of the
Among the massively. Continuing hemorrhage results on maternal outcomes
women who did not have a cesarean was subsequently complicated by between the 2 groups.
hysterectomy, delayed hysterectomy dissemi- nated intravascular
was performed in 17/24 (70.8%) coagulation and multiorgan failure. We included the cases with AIP <6
cases in the percreta group and 4/83 cm, some of which are supposed to
(4.8%) cases in the accreta group. have much better prognosis than
The overall hysterec- tomy rate was
Comment larger AIP, as observed in our study,
Principal findings and the di- agnoses of which might be
>3 times higher in the percreta group debatable given the absence of
(86.3% [44/51] vs 24.8% [26/105] P We found that the women with
placenta percreta had a rate of severe uterine tissue for histologic analysis.
<.01). The 2 maternal deaths were acute morbidity (defined by a Although the rate of AIP 6 cm was
due to hemorrhagic complications in composite score) more than triple that higher in the accreta group, the
placenta percreta cases. The first death ≤
secondary analysis showed that when
of the women with placenta accreta, as
followed initial conservative well as a higher final hysterectomy we excluded these less severe cases to
treatment that preserved the uterus rate. These results were stable avoid a potential inclusion bias linked
with the placenta left in situ during whether we analyzed all cases of AIP to the inclusion of less severe cases or
cesarean de- livery at 35 weeks, or only those with an AIP diameter >6 perhaps misdiagnosed AIP, maternal
indicated for vaginal bleeding and morbidity remained significantly
premature rupture of the cm.
lower for women with placenta
membranes. Massive hematuria accreta than placenta percreta.
required an emergency hysterectomy Strengths and limitations AIP was antenatally suspected in
13 days later. Disseminated These results indicate that the 27/ 105 (25.7%) women in the accreta
intravascular coagulation (defined percreta form of AIP should be group. Previous published rates of
as platelets considered sepa- rately from its other antenatally suspected accreta have
<50,000 mL, prolonged forms. Previous studies of maternal ranged from 29e75%.8,9,26 Potential
prothrombin morbidity associated explanations for
time >50%, increase and
fibrinogen
<2 g/L24) and acute multiorgan
failure complicated a massive
intraoperative

AUGUST 2018 American Journal of Obstetrics & Gynecology 193.e7

Original Research
OBSTETRICS
these variations include the following
points: (1) these series have
frequently failed to distinguish clearly
between those with placenta accreta
and percreta; (2) some retrospective
series may include lower rates of
small focal AIP, or report them less
often; (3) it is possible that in our
study, the cases of small focal AIP
were around the borderline for a
formal diagnosis of AIP. For this
reason, we performed a secondary
analysis of the subgroup of women
with antenatally suspected AIP >6
cm to prevent or minimize a potential
inclusion bias. The rate of antenatal
suspicion for women in the accreta
group with AIP >6 cm was 52.6% (n
20/38), twice as high as for the
¼ group.
overall accreta
Management of women with AIP
also
differed between the 2 groups: all
women in the percreta group had
cesareans that used a conservative
approach and left the placenta in situ.
However, conservative management
tends to be associated with reduced
morbidity and hysterectomy rates as
reported by Kayem et al22 and
Sentilhes et al.13 It has also been
observed that antenatally suspected
AIP and scheduled management are
associ- ated with lower
morbidity.26e29 There- fore, the
differences in maternal morbidity
between the 2 groups cannot be
explained by the higher frequency of
the conservative approach in the per-
creta group.
This study was performed in a uni-
versity maternity hospital, where all
the services and logistic support
necessary
for managing severe cases of AIP are
available, including anesthesiologists,
surgeons, embolization equipment,
and an onsite blood bank. This may
affect its external validity, but such
multidisci- plinary planning and
management in a referral center is
recommended by most expert
guidelines worldwide.12,14,30
Interpretation
Placenta percreta is diagnosed more
frequently during the antenatal period
than placenta accreta, as a 76-case
series of AIP previously reported.26
Despite the scheduled management
and planned use of adjuvant
measures, which are more frequent in
cases of placenta percreta, we
observed that it was associated with
193.e8 American Journal of Obstetrics & Gynecology AUGUST 2018
greater maternal morbidity. This
discrepancy may be explained if we
consider that the primary risk in case
of placenta percreta is not due to the
need to perform a partial cystectomy
or a bowel anastomosis during
cesarean hysterectomy, because of the
invasion of neighboring organs.
Instead it is linked to the density of
the associated vascular network in
this condition. Hence, this anarchic
and hypertrophic vasculariza- tion is
likely to be the main factor
responsible for the increase in
maternal morbidity and mortality in
placenta percreta. Consequently,
planned con- servative management
seems to us to be a reasonable option
in cases of placenta percreta, to
reduce the risk of cata- strophic
hemorrhage. In addition, risk is lower
for a delayed hysterectomy per-
formed days after the cesarean
delivery, because of the reduction of
the large periuterine vascularization
during the puerperal period.
In conclusion, even with
antenatal
screening and management in a
referral center, maternal morbidity is
dramati- cally higher for women with
placenta percreta compared with
placenta accreta. Further studies of
the prognosis and management of
AIP should systemati- cally and
clearly distinguish placenta percreta
from other forms of AIP. n 2002;77:77-8.
6.
Acknowledgment
We are indebted to Pauline Charpak and Sara
Sabbagh for their contribution to this work, and
to Jo Ann Cahn for editing the article.
14.
References
1. Jauniaux E, Jurkovic D. Placenta accreta:
pathogenesis of a 20th century iatrogenic
uterine disease. Placenta 2012;33:244-51.
2. Jauniaux E, Burton GJ. Placenta accreta
spectrum: a need for more research on its etio-
pathogenesis. BJOG 2018.
3. Publications Committee Society for
Maternal-Fetal Medicine, Belfort MA.
Placenta accreta. Am J Obstet Gynecol
2010;203:430-9.
4. Shellhaas CS, Gilbert S, Landon MB, et al.
The frequency and complication rates of hys-
terectomy accompanying cesarean delivery.
Obstet Gynecol 2009;114:224-9.
5. Betran AP, Ye J, Moller AB, Zhang J,
Gulmezoglu AM, Torloni MR. The increasing
trend in cesarean section rates: global, regional
and national estimates: 1990-2014. PLoS One
2016;11:e0148343.
ajog.or
g
Mathews TJ, Hamilton BE. Mean age of
mothers is on the rise: United States,
2000e2014. NCHS Data Brief 2016;(232): 1-
8.
7. Creanga AA, Bateman BT, Butwick AJ,
et al. Morbidity associated with cesarean
delivery in the United States: is placenta
accreta an increasingly important contrib-
utor? Am J Obstet Gynecol 2015;213:384.
e1-11.
8. Fitzpatrick KE, Sellers S, Spark P,
Kurinczuk JJ, Brocklehurst P, Knight M. The
management and outcomes of placenta
accreta, increta, and percreta in the UK: a
population-based descriptive study. BJOG
2014;121:62-71.
9. Thurn L, Lindqvist PG, Jakobsson M, et al.
Abnormally invasive placenta-prevalence, risk
factors and antenatal suspicion: results from a
large population-based pregnancy cohort study
in the Nordic countries. BJOG 2016;123:
1348-55.
10. Parva M, Chamchad D, Keegan J, Gerson A,
Horrow J. Placenta percreta with invasion of the
bladder wall: management with a multi-
disciplinary approach. J Clin Anesth 2010;22:
209-12.
11. Palacios Jaraquemada JM, Pesaresi M,
Nassif JC, Hermosid S. Anterior placenta per-
creta: surgical approach, hemostasis and uter-
ine repair. Acta Obstet Gynecol Scand 2004;83:
738-44.
12. Committee on Obstetric Practice. Placenta
accreta. ACOG Committee opinion no. 266,
January 2002. American College of Obstetri-
cians and Gynecologists. Int J Gynaecol Obstet
13. Sentilhes L, Ambroselli C, Kayem G, et al.
Maternal outcome after conservative treatment
of placenta accreta. Obstet Gynecol 2010;115:
526-34.
Sentilhes L, Vayssiere C, Deneux- Tharaux
C, et al. Postpartum hemorrhage: guidelines
for clinical practice from the French College
of Gynecologists and Obstetricians
(CNGOF): in collaboration with the French So-
ciety of Anesthesiology and Intensive Care
(SFAR). Eur J Obstet Gynecol Reprod Biol
2016;198:12-21.
15. Sentilhes L, Kayem G, Chandraharan E,
et al. FIGO consensus guidelines on placenta
accreta spectrum disorders: conservative man-
agement. Int J Gynaecol Obstet 2018;140:
291-8.
16. Hubinont C, Mhallem M, Baldin P,
Debieve F, Bernard P, Jauniaux E. A clinico-
pathologic study of placenta percreta. Int J
Gynaecol Obstet 2018;140:365-9.
17. Yetter JF III. Examination of the placenta. Am
Fam Physician 1998;57:1045-54.
18. O’Brien JM, Barton JR, Donaldson ES. The
management of placenta percreta: conservative
and operative strategies. Am J Obstet Gynecol
1996;175:1632-8.
19. American College of Obstetricians and Gy-
necologists, the Society for Maternal-Fetal
Medicine, Kilpatrick SK, Ecker JL. Severe
ajog.or
g Research
maternal morbidity: screening and review. Am J
Obstet Gynecol 2016;215:B17-22.
20. Collins SL, Ashcroft A, Braun T, et al. Pro-
posal for standardized ultrasound descriptors of
abnormally invasive placenta (AIP). Ultrasound
Obstet Gynecol 2016;47:271-5.
21. Millischer AE, Salomon LJ, Porcher R, et al.
Magnetic resonance imaging for abnormally
invasive placenta: the added value of intrave-
nous gadolinium injection. BJOG 2017;124:
88-95.
22. Kayem G, Davy C, Goffinet F,
Thomas C, Clement D, Cabrol D. Conser-
vative versus extirpative management in
cases of placenta accreta. Obstet Gynecol
2004;104:531-6.
23. Kayem G, Sentilhes L, Deneux-
Tharaux C. Management of placenta
accreta. BJOG 2009;116:1536-7; author
reply 37-8.
24. Toh CH, Hoots WK; SSC on Disseminated
Intravascular Coagulation of the ISTH. The
scoring system of the Scientific and Stan-
dardization Committee on Disseminated
Intravascular Coagulation of the International
OBSTETRICS
Society on Thrombosis and Hemostasis: a 5-
year overview. J Thromb Haemost 2007;5:
604-6.
25. Bretelle F, Courbiere B, Mazouni C, et al.
Management of placenta accreta: morbidity and
outcome. Eur J Obstet Gynecol Reprod Biol
2007;133:34-9.
26. Eller AG, Porter TF, Soisson P, Silver RM.
Optimal management strategies for placenta
accreta. BJOG 2009;116:648-54.
27. Melcer Y, Jauniaux E, Maymon S, et al.
Impact of targeted scanning protocols on
perinatal outcomes in pregnancies at risk
of placenta accreta spectrum or vasa
previa. Am J Obstet Gynecol 2018;218:443.
e1-8.
28. Bowman ZS, Manuck TA, Eller AG,
Simons M, Silver RM. Risk factors for un-
scheduled delivery in patients with placenta
accreta. Am J Obstet Gynecol 2014;210:
241.e1-6.
29. Warshak CR, Ramos GA, Eskander R, et al.
Effect of predelivery diagnosis in 99 consecutive
cases of placenta accreta. Obstet Gynecol
2010;115:65-9.
AUGUST 2018 American Journal of Obstetrics & Gynecology 193.e9
Original
30. Allahdin S, Voigt S, Htwe TT. Management
of placenta previa and accreta. J Obstet
Gynaecol 2011;31:1-6.
Author and article information
From the Port-Royal Maternity Unit, Department of Ob-
stetrics Paris (Drs Marcellin, Delorme, Grange, Tsatsaris,
and Goffinet), Department of Gynecology Obstetrics II and
Reproductive Medicine (Dr Marcellin), and Départment
d’Anesthesie Réanimation (Dr Bonnet), Hôpitaux Uni-
versitaires Paris Centre, Hôpital Cochin, Assistance
Publique-Hôpitaux de Paris; Sorbonne Paris Cité, Uni-
versité Paris Descartes, Faculté de Médecine (Drs Mar-
cellin, Delorme, Bonnet, Grange, Tsatsaris, and Goffinet);
Pierre-et-Marie-Curie University (Dr Kayem); Obstetrical,
Perinatal, and Pediatric Epidemiology Research Team,
Center for Epidemiology and Statistics Sorbonne Paris
Cité, Inserm U1153 (Dr Kayem); and Obstetrics and Gy-
necology Department, Hôpital Armand-Trousseau
(Dr Kayem), Paris, France.
Received Jan. 9, 2018; revised April 17, 2018;
accepted April 26, 2018.
The authors report no conflict of interest.
Corresponding author: Louis Marcellin, MD, PhD.
louis.marcellin@aphp.fr
Original Research ajog.or
OBSTETRICS g

SUPPLEMENTAL TABLE
Immediate neonatal outcomes according to severity
Percreta Accreta
n ¼ 52 n ¼ 117 P
Birthweight, g 2517.8 T 681.5 2444 T 958.3 .62
Apgar score at 5 min <7, n (%) 17 (32.7) 28 (23.9) .28
Arterial pH n (%)
● 7.0 ≤ pH < 7.2 6 (11.5) 9 (7.7)
● <7.0 1 (1.9) 3 (2.6) .73
Transfer to neonatal 23 (44.2) 35 (28.9) .09
intensive care unit, n (%)
Neonatal death,a n (%) 0 2 (1.7)
Stillbirth,b n (%) 0 4 (3.4)
a
One due to maternal fetal infection and 1 to severe respiratory distress syndrome; b One termination of pregnancy and 3
intrauterine fetal deaths.
Marcellin et al. Severe maternal morbidity: more frequent in placenta percreta than accreta. Am J Obstet Gynecol 2018.

193.e10 American Journal of Obstetrics & Gynecology AUGUST 2018