International Journal of Gynecology and Obstetrics 127 (2014) 206–210

Contents lists available at ScienceDirect

International Journal of Gynecology and Obstetrics

j ourna l homepage: www.e lsev ie r .com/ locate / i jgo

CLINICAL ARTICLE

Outcomes of subsequent pregnancies after conservative treatment forplacenta accreta☆

Doron Kabiri a,⁎,1, Yael Hants a,1, Neta Shanwetter a, Moshe Simons a, Carolyn F. Weiniger b,Yuval Gielchinsky a, Yossef Ezra a

a Department of Obstetrics and Gynecology, Hadassah-Hebrew University Medical Center, Jerusalem, Israelb Department of Anesthesiology and Critical Care Medicine, Hadassah-Hebrew University Medical Center, Jerusalem, Israel

☆ Presented in part at the 32nd Annual Meeting ofMedicine; February 6–11, 2012; Dallas, TX, USA.⁎ Corresponding author at: Department of Obstetric

Hebrew University Medical Center, Ein-Kerem, P.O. BoIsrael. Tel.: +972 50 8946898; fax: +972 2 6777541.

E-mail address: doronkabiri@gmail.com (D. Kabiri).1 These authors contributed equally to this work.

http://dx.doi.org/10.1016/j.ijgo.2014.05.0130020-7292/© 2014 International Federation of Gynecology

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 22 December 2013Received in revised form 20 May 2014Accepted 26 June 2014

Keywords:Adherent placentaConservative treatmentPlacenta accretaPostpartum hemorrhage

Objective: To estimate the association between conservative treatment for placenta accreta and subsequent preg-nancy outcomes. Methods: In a retrospective study, data were analyzed on women who received conservativetreatment for placenta accreta (removal of the placenta with uterine preservation) at a tertiary hospital inJerusalem, Israel, between 1990 and 2000. Data were collected on subsequent pregnancies and neonatal out-comes until 2010, and compared with those from amatched control group of womenwho did not have placentaaccreta. Results: A total of 134 women were included in both groups. Placenta accreta occurred in 62 (22.8%) of272 subsequent deliveries in the study group for which data were available and 5 (1.9%) of 266 in the controlgroup (relative risk [RR] 12.13; 95% confidence interval [CI] 4.95–29.69; P b 0.001). Early postpartum hemor-rhage occurred in 23 (8.6%) of 268 deliveries in the study group and 7 (2.6%) of 268 in the control group (RR

3.29; 95% CI 1.43–7.53; P b 0.001). The odds ratio for recurrent placenta accreta in subsequent deliveries in thestudy group was 15.41 (95% CI 6.09–39.03; P b 0.001). Conclusion: Although subsequent pregnancies after con-servative treatment for placenta accreta were mostly successful, the risk of recurrent placenta accreta and post-partum hemorrhage is high in future deliveries.© 2014 International Federation of Gynecology and Obstetrics. Published by Elsevier Ireland Ltd. All rights reserved.

1. Introduction

Placenta accreta is a severe obstetric complication that is character-ized by abnormal adherence of the placenta to the uterine wall. Histo-pathologically, it is characterized by partial or complete absence of thedecidua basalis or the Nitabuch layer of the decidua, which results inabnormal attachment of the chorionic villi to the myometrium [1].

Placenta accreta can be a life-threatening obstetric condition; it isassociated with high maternal morbidity and mortality due to uncon-trolled bleeding, unplanned cesarean hysterectomy, and other compli-cations resulting from abnormal invasion of the placenta into adjacentorgans [2]. The reported incidence of placenta accreta has been risingconstantly over the past few decades, and this disorder has become animportant obstetric problem in medical practice [2–11].

the Society for Maternal-Fetal

s and Gynecology, Hadassah-x 12000, Jerusalem 9112001,

and Obstetrics. Published by Elsevier I

To reduce maternal morbidity and mortality, elective cesarean hys-terectomy in a tertiary care hospital with a multidisciplinary careteam is considered to be the safest andmost common treatment for pla-centa accreta diagnosed before delivery [9,12–17]. However, placentaaccreta may be diagnosed after delivery, when the placenta or placentaltissue fails to separate from the uterine wall. In an attempt to obtain anempty uterine cavity, systematic manual separation of the placentafrom the uterine wall is usually performed. This forcible manual bluntdissection can inducemassive hemorrhage, whichmay result in hyster-ectomy [2,18–20].

The aim of the present study was to estimate the associationbetween conservative treatment for placenta accreta (removal of theplacenta with uterine preservation) and subsequent pregnancy out-comes under the hypothesis that conservative treatment might nega-tively affect the obstetric outcomes of subsequent pregnancies.

2. Materials and methods

The present retrospective cohort study analyzed data from womenconservatively treated for placenta accreta at the maternal–fetal unitof the tertiary Hadassah-Hebrew University Medical Center, Jerusalem,Israel, between January 1, 1990, and December 31, 2000. In addition tothe women who were treated for placenta accreta, the study enrolled

reland Ltd. All rights reserved.

207D. Kabiri et al. / International Journal of Gynecology and Obstetrics 127 (2014) 206–210

a control group of women who had a delivery in the study period andwho were matched for parity and age (stratified by age group:b25 years, 25–29 years, 30–34 years, 35–39 years, and ≥40 years).The present study was approved by the institutional review board ofthe hospital on June 1, 2011 (reference number 0157-11-HMO), andall patients provided informed consent for participation by telephone.The cohort of patients has been used in other analyses [21,22].

Placenta accreta was diagnosed on the basis of sonographic, clinical,or histopathologic findings. Patients had to meet one or more of the fol-lowing criteria [17,21,22]: (1) impossibility of, or incomplete, manualremoval of the placentawith evidence of placental retention, despite ac-tive management of the third stage of labor (transabdominal manualmassage of the uterus accompanied by controlled traction of the umbil-ical cord); (2) sonographic evidence of retained placental fragments re-quiring removal (curettage or hysteroscopy) after vaginal delivery;(3) heavy bleeding from the implantation site after placental removalduring cesarean delivery with excision of part of the uterine wall andthe attachedplacenta or oversewingof the bleeding defects; and (4) his-tologic confirmation of placenta accreta (invasion of chorionic villi tothe myometrium). Women in the control group were selected fromthe registry database of the hospital in consecutive order by date ofthe index delivery.

Clinical data for the present study—including maternal demo-graphics, obstetric parameters, and subsequent pregnancy outcomesuntil 2010—were retrieved from the Ministry of Health CentralBureau of Statistics report and from patients’ medical records. Tele-phone interviews were conducted to obtain additional data regardingsubsequent pregnancies, obstetric complications, and maternal andneonatal outcomes.

Conservative management of placenta accreta was defined as re-moval of the placentawith uterine preservation. Early postpartumhem-orrhage was defined as bleeding requiring medical or interventionaltreatment in the first 24 hours after delivery.

Deliveries 2000

Women lost to follow-up or declined to participate

(n=126)

Women with conservative treatment for placenta

accreta (n=260)

Women enrolled in the study group(n=134)

Fig. 1. Enrollment of the

Statistical analysis was performed with SPSS version 18 (SPSS Inc,Chicago, IL, USA).The t test was used to assess differences in continuousvariables between the two groups. Differences in categorical variablesand in each of the dependent variables were analyzed via the χ2 orFisher exact test. Odds ratios (ORs) were analyzed via a logistic regres-sion model. P b 0.05 was considered statistically significant.

3. Results

Between 1990 and 2000, there were 34 450 deliveries at the studyinstitution and 260 women received conservative management for pla-centa accreta. Of these women, 99 (38.1%) were lost to follow-up, 5(1.9%) lacked a matched control, and 22 (8.5%) declined to participate.As a result, subsequent pregnancy outcomes were compared between134 women affected by placenta accreta (study group) and 134womenmatched for age and obstetric history who did not have placen-ta accreta (control group) (Fig. 1, Table 1).

The two groups were similar with regard to demographic and ob-stetric parameters; the only significant difference was a higher inci-dence of previous placenta accreta among patients in the study group(Table 2). Of the 107women in the study groupwho attempted concep-tion after placenta accreta, 99 (92.5%) successfully delivered live new-borns, with 280 deliveries. 105 women in the control group attemptedpregnancy after the index delivery, and 94 (89.5%) women achieved271 deliveries (P = 0.205).

Data for complications and outcomes were available for 272 subse-quent deliveries in the study group and 266 deliveries in the controlgroup. Placenta accreta occurred in 62 (22.8%) subsequent deliveriesin the study group and 5 (1.9%) in the control group (relative risk [RR]12.13; 95% confidence interval [CI] 4.95–29.69; P b 0.001) (Table 3).Early postpartum hemorrhage occurred in 23 (8.5%) and 7 (2.6%) deliv-eries in the study and control groups, respectively (RR 3.29; 95% CI

between 1990 and (n=34 450)

Women enrolled in the matched control group (n=134)

study participants.

Table 1Matching criteria between the study and control groups.a

Criterion Study group(n = 134)

Control group(n = 134)

P value

Maternal age, y 0.9918–24 30 (22.4) 31 (23.1)25–29 34 (25.4) 35 (26.1)30–34 41 (30.6) 40 (29.9)35–39 24 (17.9) 23 (17.2)≥40 5 (3.7) 5 (3.7)

Mode of delivery N0.99Cesarean delivery 3 (2.2) 3 (2.2)Vaginal delivery 131 (97.8) 131 (97.8)

Previous live births 0.900 36 (26.9) 37 (27.6)1 41 (30.6) 41 (30.6)2 24 (17.9) 20 (14.9)3–5 18 (13.4) 23 (17.2)≥6 15 (11.2) 13 (9.7)

a Values are given as number (percentage) unless stated otherwise.

Table 3Obstetric complications during subsequent pregnancies.a

Complication Subsequent deliveries Relative risk(95% CI)

P value

Studygroup(n = 272)

Controlgroup(n = 266)

Placenta accreta 12.13 (4.95–29.69) b0.001Yes 62 (22.8) 5 (1.9)No 210 (77.2) 261 (98.1)

Early PPHb 3.29 (1.43–7.53) b0.001Yes 23 (8.6) 7 (2.6)No 245 (91.4) 261 (97.4)

Placenta previac 2.45 (0.48–12.54) 0.280Yes 5 (1.9) 2 (0.8)No 265 (98.1) 263 (99.2)

Preterm birth 13 (4.7) 19 (7.1) 0.6 (0.3–1.3) 0.235Twin pregnancy 6 (2.2) 5 (1.9) 1.16 (0.36–3.76) 0.802Gestational diabetes 8 (2.9) 15 (5.6) 0.51 (0.22–1.20) 0.123Hypertensive disorders 10 (3.7) 10 (3.8) 0.97 (0.41–2.29) 0.941Intra-uterine fetal death 0 1 (0.4) 0.32 (0.01–7.89) 0.488Placental abruption 11 (3.7) 5 (1.9) 2.13 (0.75–6.05) 0.156Cesarean delivery 35 (12.9) 35 (13.2) 0.97 (0.62–1.50) 0.884Postpartum infection 3 (1.1) 2 (0.8) 1.45 (0.24–8.62) 0.682

Abbreviations: CI, confidence interval; PPH postpartum hemorrhage.a Values are given as number (percentage) unless stated otherwise.b Data available for 268 deliveries in the study group and 268 in the control group.c Data available for 270 deliveries in the study group and 265 in the control group.

208 D. Kabiri et al. / International Journal of Gynecology and Obstetrics 127 (2014) 206–210

1.43–7.53; P b 0.001) (Table 3). The OR for recurrent placenta accretaduring subsequent delivery in the study group was 15.41 (95% CI6.09–39.03; P b 0.001). There were no significant differences in therates of pregnancy complications, mode of delivery, or neonatal out-comes (Tables 3 and 4).

To validate the results, three demographic and obstetric parametersknown to be related to the incidence of placenta accreta—maternal age,number of pregnancies, and number of deliveries before the firstextirpative treatment of the placenta accreta—were compared betweenwomen with conservatively managed placenta accreta who wereincluded and excluded from the study group. Information on theparam-eters was obtained from medical records. The comparison revealedno differences (Table 5), reducing the likelihood of selection bias inthe study.

4. Discussion

In the present study, pregnancies after conservative treatment ofplacenta accreta were mostly successful. However, among the studygroup of women, there was an increased risk of the recurrence ofplacenta accreta and postpartum hemorrhage.

Although conservative treatment is common when the diagnosis ofplacenta accreta is made after delivery, there have been few studieson the long-term effects of this treatment on subsequent pregnancies,deliveries, and neonatal outcomes [23]. The high recurrence rate ob-served in the present study might be associated with the factors thatled to the initial placenta accreta, such as uterine factors (congenitalanomalies and infection), placental factors (abnormal location and

Table 2Demographic and obstetric characteristics at the elected delivery by group.a

Characteristics Study group (n = 134)

DemographicMaternal age 30.58 ± 5.65 (21–44)Gravidity 2.71 ± 2.99 (0–13)Parity 1.92 ± 2.22 (0–10)Abortions 0.71 ± 1.27 (0–9)Ectopic pregnancy 0.04 ± 0.27 (0–2)Prior cesarean delivery 0.08 ± 0.31 (0–3)Living children 1.95 ± 2.25 (0–10)Previous placenta accreta 23 (17.2)Previous placenta previa 2 (1.5)Previous postpartum hemorrhage 11 (8.2)

ObstetricHypertension disorders 11 (8.2)Infections 7 (5.2)

a Values are given as mean ± SD (range) or number (percentage) unless stated otherwise.

developmental defects), and uterine scarring (previous curettage, ce-sarean delivery, and fibroid removal) [19,24].

A common problem in studies on placenta accreta is the definition ofthis condition: some studies are based on prenatal imaging criteria,some on clinical criteria, and others on histopathology. A literaturesearch was conducted via Medline, PubMed, Embase, and the CochraneLibrary for reports published between January 1, 2002, and December31, 2012, using the keywords “placenta accreta,” “maternal outcomes,”and “neonatal outcomes,” to identify studies with methodologies simi-lar to the present study. Because histologic examination is not alwaysavailable when hysterectomy does not occur, in most recent studies, di-agnosis of placenta accreta has been based on a combination of findings(Table 6). The present study used diagnostic criteria that were acceptedin previous studies andwere based on sonographic, clinical, or histolog-ic parameters, and almost all cases fulfilled more than one diagnosticcriterion. Selection bias due to the reliability of the inclusion criteria isunlikely given the differences between the groups in terms of recurrentplacenta accreta and postpartum hemorrhage.

The location of the placentamight influence the risk of recurrence ofplacenta accreta. The present study had a long follow-up period, whichmeant that it was not possible to retrieve placental location because thisinformation was not contained within medical records. The higher

Control group (n = 134) P value

30.4 ± 5.77 (20–40) 0.9252.64 ± 2.89 (0–15) 0.8792.07 ± 2.32 (0–10) 0.4560.56 ± 1.09 (0–7) 0.2630.01 ± 0.09 (0–1) 0.1740.11 ± 0.28 (0–1) 0.4101.98 ± 2.33 (0–12) 0.992

4 (3.0) b0.0013 (2.2) 0.8683 (2.2) 0.037

4 (3.0) 0.1044 (3.0) 0.372

Table 4Neonatal outcomes during subsequent pregnancies.a

Outcome Study group(n = 286)

Control group(n = 276)

P value

Birth weight, g 3397 ± 493 3294 ± 552 0.31Gestation, wk 39.9 ± 1.9 39.6 ± 1.9 0.44Sex 0.51

Male 153 (53.5) 149 (54.0)Female 133 (46.5) 127 (46.0)

No. with 5-min Apgar score b7 0 0 N0.999No. admitted to NICU 0 0 N0.999

Abbreviation: NICU, neonatal intensive care unit.a Values are given as mean ± SD or number (percentage), unless stated otherwise.

Table 5Comparison of demographic and obstetric characteristics betweenwomenwith conserva-tively managed placenta accreta who were and were not included in the study group.a

Characteristic Women instudy group(n = 134)

Women excludedfrom study group(n = 126)

P value

Current maternal age, y 46.03 ± 6.25 45.86 ± 6.19 0.836No. of pregnancies before firstextirpative treatment

2.71 ± 2.99 3.19 ± 2.43 0.176

No. of deliveries before firstextirpative treatment

1.91 ± 2.22 1.58 ± 1.88 0.226

a Values are given as mean ± SD unless stated otherwise.

209D. Kabiri et al. / International Journal of Gynecology and Obstetrics 127 (2014) 206–210

incidence of placenta accreta and postpartum hemorrhage in pregnan-cies before the index pregnancy among the study group was foundonly after data analysis, and might bias the results of the study. Addi-tional studies may be needed to investigate the effect of these parame-ters on the risk of recurrent placenta accreta.

Single-center studies havemethodological advantages, such as unityin the care protocol and a relatively homogeneous population. Never-theless, the external validity of these results might be limited. A studybased on a larger cohort across multicenter databases might be war-ranted to validate the results. Recall bias and incomplete data sets areanticipated in every retrospective study; however, comparison ofthe results to a matched control group strengthens the validity of thepresent study’s results.

Table 6Diagnostic criteria of placenta accreta in recent clinical studies.

Author Year Diagnos

Gielchinsky et al. 2002 HistologGielchinsky et al. 2004 HistologKayem et al. 2004 HistologBretelle et al. 2005 HistologWu et al. 2005 HistologEller et al. 2009 HistologDoumouchtsis et al. 2010 HistologProvansal et al. 2010 HistologSentilhes et al. 2010 HistologSentilhes et al. 2010 HistologEller et al. 2011 HistologEsh-Broder et al. 2011 HistologAmsalem et al. 2011 HistologMeyer et al. 2012 HistologAggarwal et al. 2012 HistologFitzpatrick et al. 2012 HistologEshkoli et al. 2013 HistologMorlando et al. 2013 HistologKamara et al. 2013 ImagingGuleria et al. 2013 ImagingKayem et al. 2013 HistologFitzpatrick et al. 2013 HistologHiggins et al. 2013 HistologWeiniger et al. 2013 Imaging

The CI for the odds of recurrent placenta accreta was wide, whichmeans that the estimation may be imprecise. A larger sample sizewould narrow the CI and strengthen the conclusions.

In summary, although subsequent pregnancies after conservativetreatment for placenta accreta were found to be mostly successful,appropriate preparations should be made to minimize morbidity andmortality resulting from the recurrence of placenta accreta and postpar-tumhemorrhage. Previous studies [3,7,25,26] have also shown that staffawareness and suitable preparations (including the preparation ofblood products) may result in better treatment and a reduction in relat-ed risks, such as massive blood loss, consumption coagulopathy,unplanned cesarean hysterectomy, organ failure, and even death.

Conflict of interest

The authors have no conflicts of interest.

References

[1] Benirschke K, Kaufmann P, Baergen RN. Pathology of the Human Placenta. 5th ed.London: Springer; 2006.

[2] Publications Committee, Society for Maternal–Fetal Medicine, Belfort MA. Placentaaccreta. Am J Obstet Gynecol 2010;203(5):430–9.

[3] Wortman AC, Alexander JM. Placenta accreta, increta, and percreta. Obstet GynecolClin North Am 2013;40(1):137–54.

[4] Miller DA, Chollet JA, Goodwin TM. Clinical risk factors for placenta previa–placentaaccreta. Am J Obstet Gynecol 1997;177(1):210–4.

[5] ACOG Committee on Obstetric Practice. Number 266, January 2002: placentaaccreta. Obstet Gynecol 2002;99(1):169–70.

[6] Wu S, Kocherginsky M, Hibbard JU. Abnormal placentation: twenty-year analysis.Am J Obstet Gynecol 2005;192(5):1458–61.

[7] Oyelese Y, Smulian JC. Placenta previa, placenta accreta, and vasa previa. ObstetGynecol 2006;107(4):927–41.

[8] Solheim KN, Esakoff TF, Little SE, Cheng YW, Sparks TN, Caughey AB. The effect ofcesarean delivery rates on the future incidence of placenta previa, placenta accreta,and maternal mortality. J Matern Fetal Neonatal Med 2011;24(11):1341–6.

[9] Committee on Obstetric Practice. Committee opinion no. 529: placenta accreta.Obstet Gynecol 2012;120(1):207–11.

[10] Aggarwal R, Suneja A, Vaid NB, Yadav P, Sharma A, Mishra K. Morbidly adherentplacenta: a critical review. J Obstet Gynaecol India 2012;62(1):57–61.

[11] Royal College of Obstetricians and Gynaecologists. Placenta praevia, placenta praeviaaccreta and vasa praevia: diagnosis and management. Green-top Guideline No. 27.http://www.rcog.org.uk/files/rcog-corp/GTG27PlacentaPraeviaJanuary2011.pdf .Published January 2011. Accessed June 24, 2014.

tic criteria of placenta accreta Journal

y or clinical Placentay or clinical Obstet Gynecoly or clinical Obstet Gynecoly or clinical Eur J Obstet Gynecoly or clinical Am J Obstet Gynecoly or clinical BJOGy or clinical Acta Obstet Gynecol Scandy or clinical Int J Obstet Gynecoly or clinical Obstet Gynecoly or clinical Hum Reprody or clinical Obstet Gynecoly BJOGy J Obstet Gynaecol Cany Ceylon Med Jy or clinical J Obstet Gynecol Indiay or clinical PLOS Oney AJOGy or clinical Acta Obstet Gynecol Scand, histology, or clinical BJOG, histology, or clinical Acta Obstet Gynecol Scandy or clinical Acta Obstet Gynecol Scandy or clinical BJOGy Eur J Obstet Reprod Bioland clinical Int J Obstet Anesth

210 D. Kabiri et al. / International Journal of Gynecology and Obstetrics 127 (2014) 206–210

[12] Jolley JA, Nageotte MP, Wing DA, Shrivastava VK. Management of placenta accreta: asurvey of Maternal–Fetal Medicine practitioners. J Matern Fetal Neonatal Med2012;25(6):756–60.

[13] Esakoff TF, Handler SJ, Granados JM, Caughey AB. PAMUS: placenta accreta manage-ment across the United States. J Matern Fetal Neonatal Med 2012;25(6):761–5.

[14] Fox H. Placenta accreta: 1945-1969. Obstet Gynecol Surv 1972;27(7):475–90.[15] Eller AG, Porter TF, Soisson P, Silver RM. Optimalmanagement strategies for placenta

accreta. BJOG 2009;116(5):648–54.[16] AmsalemH, Kingdom JC, Farine D, Allen L, Yinon Y, D’Souza DL, et al. Planned caesar-

ean hysterectomy versus “conserving” caesarean section in patients with placentaaccreta. J Obstet Gynaecol Can 2011;33(10):1005–10.

[17] Kayem G, Davy C, Goffinet F, Thomas C, Clément D, Cabrol D. Conservative versusextirpative management in cases of placenta accreta. Obstet Gynecol 2004;104(3):531–6.

[18] Bretelle F, Courbière B, Mazouni C, Agostini A, Cravello L, Boubli L, et al. Managementof placenta accreta: morbidity and outcome. Eur J Obstet Gynecol Reprod Biol2007;133(1):34–9.

[19] Jauniaux E, Jurkovic D. Placenta accreta: pathogenesis of a 20th century iatrogenicuterine disease. Placenta 2012;33(4):244–51.

[20] Sentilhes L, Goffinet F, Kayem G. Management of placenta accreta. Acta ObstetGynecol Scand 2013;92(10):1125–34.

[21] Gielchinsky Y, Rojansky N, Fasouliotis SJ, Ezra Y. Placenta accreta–summary of10 years: a survey of 310 cases. Placenta 2002;23(2–3):210–4.

[22] Gielchinsky Y, Mankuta D, Rojansky N, Laufer N, Gielchinsky I, Ezra Y. Perinatal out-come of pregnancies complicated by placenta accreta. Obstet Gynecol 2004;104(3):527–30.

[23] Tandberg A, Albrechtsen S, Iversen OE. Manual removal of the placenta. Incidenceand clinical significance. Acta Obstet Gynecol Scand 1999;78(1):33–6.

[24] Hung TH, Shau WY, Hsieh CC, Chiu TH, Hsu JJ, Hsieh TT. Risk factors for placentaaccreta. Obstet Gynecol 1999;93(4):545–50.

[25] O’Brien JM, Barton JR, Donaldson ES. The management of placenta percreta: conser-vative and operative strategies. Am J Obstet Gynecol 1996;175(6):1632–8.

[26] Hudon L, Belfort MA, Broome DR. Diagnosis andmanagement of placenta percreta: areview. Obstet Gynecol Surv 1998;53(8):509–17.