Kayem, Gilles MD; Davy, Céline MD; Goffinet, François MD, PhD; Thomas, Carole MD; Clément, Denis MD; Cabrol, Dominique PhD
Placenta accreta occurs when a defect of the decidua basalis results in abnormally invasive placental implantation.1 It is often diagnosed only after delivery when manual removal of the placenta has failed. Attempting forcible manual removal of a placenta accreta can easily lead to dramatic hemorrhaging that may result in hysterectomy. Thus, placenta accreta and especially placenta percreta have been reported to result in a mortality rate of 7% and cause intraoperative and postoperative morbidity associated with massive blood transfusions, infection, ureteral damage, and fistula formation.2 Its incidence, which is correlated with the cesarean rate, has increased 10-fold in the past 50 years.3 With a frequency of approximately 1 per 1,000 deliveries, this disorder has become more common in our medical practice.4
Case reports describe several methods for decreasing blood loss after delivery and preserving the uterus: uterine packing, oversewing the placental bed, leaving the placenta in situ, prostaglandin administration, direct aortic compression, uterine and hypogastric artery ligation, methotrexate injection, selective embolization procedures, and argon beam coagulation.5–7
Until mid-1997, the “gold standard” for the management of placenta accreta in our tertiary referral center was an extirpative approach, in accordance with most recommendations from Europe and North America.8 In July of that year, we changed this strategy and agreed unanimously to modify our hospital protocol. This decision was based on several published case reports and followed the successful conservative management of one case of placenta accreta, by leaving the placenta in place.9 We began to treat most cases of placenta accreta conservatively, leaving in situ each placenta that adhered either partially or totally to the myometrium. The purpose of this study was to evaluate the impact of this conservative management on hysterectomy and maternal morbidity rates.
MATERIALS AND METHODS
We retrospectively reviewed the medical records of all patients admitted to our tertiary center with a diagnosis of placenta accreta, increta, or percreta during the decade from January 1993 through December 2002. Gestational age was generally estimated by ultrasound dating and otherwise measured from the first day of the last menstrual period. We included all women who delivered at our hospital after 22 weeks of gestation.
To simplify the analysis, we treated all abnormal placentation including the variants of placenta increta and percreta as accreta. We diagnosed placenta accreta according to clinical or histologic criteria as follows10: 1) manual removal of the placenta partially or totally impossible and no cleavage plane between part or all of the placenta and the uterus, 2) heavy bleeding from the implantation site after forced placental removal during cesarean delivery, 3) histologic confirmation of accreta on a hysterectomy specimen, and 4) prenatal diagnosis of placenta accreta confirmed by the failure of its gentle attempted removal during the third stage of labor.
We compared 2 periods: January 1993 to June 1997 (period A) and July 1997 to December 2002 (period B). During period A, our written protocol called for the systematic manual removal of the placenta to leave the uterine cavity empty, even if force was required. In period B, we changed our policy and attempted to treat patients with placenta accreta conservatively.
We used 2 different types of conservative treatment, depending on how the placenta accreta was discovered. When it was discovered during delivery, removal of the placenta was not forced; the conservative treatment left the placenta, in part or entirely, in the uterus when the patient's hemodynamic status was stable and no septic risk was present. When the placenta accreta was strongly suspected before delivery (based on history and ultrasound or magnetic resonance imaging suggestive of the diagnosis), the case was discussed at the weekly obstetrical staff meeting and conservative treatment proposed to the patient.
In this case, management included the following steps. The precise position of the placenta was determined by ultrasound. A cesarean was planned, with the abdominal incision at the infraumbilical midline, enlarged above the umbilicus if necessary, and a vertical uterine incision at a distance from the placental insertion. After extraction of the infant, delivery of the placenta was attempted prudently, with the injection of 5 IU oxytocin and moderate cord traction. If this failed, the placenta was considered to be “accreta” and left in situ. The cord was cut at the placental insertion and the placenta left in the uterine cavity; the uterine incision was closed. Prophylactic antibiotic therapy (amoxicillin and clavulanic acid) was administered systematically for 10 days.
During the postpartum period, all patients were evaluated weekly for 6 months with ultrasonography, clinical examination, and blood counts. To improve clinical follow-up and to help choose antibiotic therapy in cases of endometritis with or without sepsis, C-reactive protein was assayed, and vaginal samples were taken for bacteriologic study.
Maternal morbidity was assessed with the following indicators: need for blood (or blood product) transfusion, hysterectomy, intensive care admission, duration of stay in intensive care, disseminated intravascular coagulation (defined by platelet count of less than 100,000 per cubic millimeter or a rapid decline in the platelet count; prolongation of clotting times, such as the prothrombin time and the activated partial-thromboplastin time; the presence of fibrin-degradation products in plasma; and low plasma levels of coagulation inhibitors, such as antithrombin III), and postpartum endometritis (defined by fever of more than 38.0°C and pelvic pain suggestive of the diagnosis) with or without sepsis (defined by positive blood culture).
The anesthetist and obstetrician jointly decided upon the transfusion of packed red blood cells or fresh frozen plasma, according to the mother's hemodynamic situation, the extent of blood loss, and the risk of hemorrhage. The transfusion policy did not change between the periods.
Univariate analysis was conducted with Fisher exact test for categorical variables and the Mann-Whitney U test for continuous variables.
From 1993 through 2002, 31,921 deliveries occurred at our center. Thirty-three women (1.03/1,000) satisfied our diagnostic criteria, and their charts were reviewed for their medical and pregnancy history and their exposure to potential risk factors. Table 1 summarizes the characteristics of this population. The groups did not differ significantly in age, parity, or type of delivery. In all, 17 patients either went into spontaneous preterm labor or had vaginal bleeding before term. The ensuing spontaneous or induced preterm births resulted in a mean estimated gestational age of 33.8 weeks at delivery for the 33 patients.
Table 2 summarizes the patients’ risk factors. Placenta previa was observed in 20 patients (60.6%). The groups did not differ significantly for these risk factors. All but one patient had a history of at least one risk factor (Table 2).
Figure 1 shows the 2 groups and the treatment received, including the number of hysterectomies. Table 3 describes outcomes for all patients. Thirteen cases (0.93/1,000 deliveries) of placenta accreta occurred during period A, only one of which was diagnosed prenatally. Only 2 patients did not have a hysterectomy during this period. For the first patient, the forcible removal of the entire placenta caused moderate hemorrhaging that was controlled by embolization of both uterine arteries. For the second patient, a small part of placenta remained accreta, despite the attempt to remove it forcibly, and was finally left in the uterus and treated conservatively.
Twenty cases (1.11/1,000 deliveries) of placenta accreta occurred during period B. Two women had a placenta percreta with localized bladder invasion; conservative management for them was successful. Fifteen cases were discovered at delivery and managed conservatively, with the placenta left partially or totally in place. Four cases of placenta accreta suspected prenatally were confirmed during the cesarean and managed conservatively, with the placenta left totally in place. Only one multiparous patient, with a diagnosis of placenta accreta suspected prenatally, preferred surgical management in case of intraoperative confirmation; she had a hysterectomy. Two other patients underwent hysterectomies during period B. One involved forcible removal, in a case where the surgeon did not suspect placenta accreta; massive hemorrhage led to a hysterectomy. The second patient had a hysterectomy on day 26 because of severe endometritis and a recurrence of hemorrhage, although conservative treatment was initially successful.
The retained placenta was measured along the horizontal axis of the uterus by ultrasonography within 2 days of delivery (median 40 mm, range 10–100). All the patients treated conservatively were carefully monitored on a weekly basis until complete ultrasonographic resorption of the placenta, at a mean of 6 months (3–12 months). One patient was admitted to our department for uterine bleeding and contractions 15 days after delivery. Clinical examination showed that the cervix was fully dilated and the placenta partially expelled into the vagina. The delivered part of the placenta was therefore removed after surgical excision. The other women in group B had spotting until complete resorption of the placenta. Six cases of postpartum endometritis, 3 involving blood-culture–confirmed sepsis, were treated successfully with hospitalization and appropriate intravenous antibiotics. Surgical evacuation was not performed in any of these cases.
The comparison of the 2 periods shows fewer cases of hysterectomy, transfusion, and disseminated intravascular coagulation during period B than during period A (Table 3). The hysterectomy rate decreased between 1993 and 2002, although the number of cases of placenta accreta was higher during period B. Sulprostone was administered to 11 patients (85%) during period A and 16 (80%) during period B (P = .99). We performed one uterine artery embolization and one hypogastric artery ligation during period A, and one hypogastric artery ligation and 5 uterine artery embolizations during period B (P = .43).
Seven patients in group B were contacted from 1 to 5 years afterward, whereas 10 were lost to long-term follow-up. Of these 7 patients, one had another successful pregnancy 2 years later and another had 2 consecutive successful pregnancies, both complicated by placenta accreta treated conservatively. The others chose for various personal reasons not to become pregnant again. None sought subsequent treatment for sterility.
This study is a historical consecutive study comparing conservative and extirpative management of placenta accreta in a single center with consistent treatment protocols.
The approach most often recommended in cases of placenta accreta is extirpative.4 If risk factors and prenatal imaging together strongly suggest this diagnosis, a cesarean hysterectomy is generally planned, especially for patients who do not wish continued fertility. If the placenta accreta is discovered after delivery, the placenta is removed as soon as possible to empty the uterine cavity. In most cases, however, this forced delivery induces massive hemorrhaging and leads to hysterectomy. Moreover, when the placenta is percreta, the surgery can cause ablation of contiguous organs and thus substantial maternal morbidity, in addition to a mortality rate of approximately 7%.2
An alternative therapeutic approach is conservative rather than extirpative. Some cases of successful conservative management of placenta accreta have previously been reported,6,11–13 although they were counterbalanced by the hemorrhagic and infectious complications that sometimes occurred.14,15 Moreover, case reports, although useful for suggesting new management approaches, cannot be used to assess them and do not allow the benefits and disadvantages of each therapeutic strategy to be evaluated comparatively. Similarly, the risk of complications cannot be estimated accurately from a few published case reports.
Our study compared 2 periods during which 2 different treatment strategies were applied. To minimize the inclusion bias, the study included all patients with placenta accreta over a 10-year period. The comparison was based solely on the period and not on the type of treatment strategy. Nonetheless, our results show that a policy of conservative management–even although extirpative treatment was performed in some cases–decreased the hysterectomy and maternal morbidity rates and preserved childbearing ability.
Nonetheless, although our results support the use of conservative management to improve the outcome in selected patients, several questions remain unresolved. These include the optimal adjuvant management, the best prenatal imaging methods for diagnosis, the long-term prognosis, and the risk of rare and grave maternal morbidity or mortality.
Methotrexate, uterine artery embolization, and sulprostone are 3 adjuvant treatments described in several case reports involving conservative treatment.6,14,16–18 The outcome when the placenta is left in place after methotrexate administration varies widely; it ranges from expulsion at 7 days to progressive resorption in roughly 6 months.6,16–18 Similarly, only a few reports describe the outcome after embolization and leaving the placenta in situ.19 We embolized the uterine arteries in some cases, to diminish or prevent a postpartum hemorrhage, and did not use methotrexate at all. Sulprostone is a well-known uterotonic agent used in case of postpartum hemorrhage. It was used in period A to attempt to avoid hysterectomy and in period B in most cases to prevent or to treat immediate abnormal postpartum bleeding, particularly in case of substantial placental retention. Published reports do not currently prove the benefit of adding these two therapies to conservative treatment, but we preferred to use them to try to prevent major postpartum bleeding in the 2 or 3 days after delivery. Prenatal identification of placenta accreta would make it much easier to take the appropriate precautions and to make choices about delivery in advance. However, the sensitivity and specificity of transvaginal or transabdominal ultrasound and magnetic resonance imaging vary from 33% to 95% in different studies; they depend especially on placenta location.8,20–22 For these reasons, imaging should be considered only when placenta accreta is suspected for clinical reasons. In our practice, we systematically attempt a careful and gentle intraoperative delivery of the placenta to confirm the diagnosis, even when placenta accreta is strongly suspected before labor.
Conservative management had some disadvantages, including postpartum infection, treatment failure, and restrictive follow-up. Ideally, these should be discussed prenatally with the patient to give her complete information about the different therapeutic strategies (extirpative or conservative). Given the difficulties mentioned above for prenatal diagnosis, however, this discussion is rarely possible. Accordingly, we have always attempted to preserve maternal fertility and to diminish the risk of hemorrhage when placenta accreta is discovered during delivery.
In our study, conservative management of placenta accreta was associated with a reduction of severe postpartum hemorrhaging and hysterectomies. These results are promising and indicate that this treatment can be considered in selected cases of women without massive postpartum hemorrhage and desirous of future fertility. Thus, at least 2 of the women followed up in group B had subsequent successful pregnancies.
This strategy must be applied with discrimination, however. The number of patients included in the second period is too low for an adequate evaluation of the risk of rare severe maternal morbidity or mortality. Accordingly, this type of management is presently appropriate only when rigorous monitoring will follow, in centers with adequate equipment and resources.8
1. Khong TY, Robertson WB. Placenta creta and placenta praevia creta. Placenta 1987;8:399–409.
2. O'Brien JM, Barton JR, Donaldson ES. The management of placenta percreta: conservative and operative strategies. Am J Obstet Gynecol 1996;175:1632–8.
3. Miller DA, Chollet JA, Goodwin TM. Clinical risk factors for placenta previa-placenta accreta. Am J Obstet Gynecol 1997;177:210–4.
4. Placenta accreta. ACOG Committee Opinion No. 266. American College of Obstetricians and Gynecologists. Int J Gynaecol Obstet 2002;77:77–8.
5. Scarantino SE, Reilly JG, Moretti ML, Pillari VT. Argon beam coagulation in the management of placenta accreta. Obstet Gynecol 1999;94:825–7.
6. Mussalli GM, Shah J, Berck DJ, Elimian A, Tejani N, Manning FA. Placenta accreta and methotrexate therapy: three case reports. J Perinatol 2000;20:331–4.
7. Johanson R, Kumar M, Obhrai M, Young P. Management of massive postpartum haemorrhage: use of a hydrostatic balloon catheter to avoid laparotomy. BJOG 2001;108:420–2.
8. American College of Obstetricians and Gynecologists. Postpartum hemorrhage. ACOG Educational Bulletin 243. Washington, DC: ACOG; 1998.
9. Kayem G, Pannier E, Goffinet F, Grange G, Cabrol D. Fertility after conservative treatment of placenta accreta. Fertil Steril 2002;78:637–8.
10. Gielchinsky Y, Rojansky N, Fasouliotis SJ, Ezra Y. Placenta accreta—summary of 10 years: a survey of 310 cases. Placenta 2002;23:210–4.
11. Legro RS, Price FV, Hill LM, Caritis SN. Nonsurgical management of placenta percreta: a case report. Obstet Gynecol 1994;83:847–9.
12. Hollander DI, Pupkin MJ, Crenshaw MC, Nagey DA. Conservative management of placenta accreta: a case report. J Reprod Med 1988;33:74–8.
13. Komulainen MH, Vayrynen MA, Kauko ML, Saarikoski S. Two cases of placenta accreta managed conservatively. Eur J Obstet Gynecol Reprod Biol 1995;62:135–7.
14. Jaffe R, Dubeshter B, Sherer DM, Thompson EA, Woods JR. Failure of methotrexate treatment for term placenta percreta. Am J Obstet Gynecol 1994;171:558–9.
15. Butt K, Gagnon A, Delisle MF. Failure of methotrexate and internal iliac balloon catheterization to manage placenta percreta. Obstet Gynecol 2002;99:981–2.
16. Arulkumaran S, Ng CS, Ingemarsson I, Ratnam SS. Medical treatment of placenta accreta with methotrexate. Acta Obstet Gynecol Scand 1986;65:285–6.
17. Buckshee K, Dadhwal V. Medical management of placenta accreta. Int J Gynaecol Obstet 1997;59:47–8.
18. Gupta D, Sinha R. Management of placenta accreta with oral methotrexate. Int J Gynaecol Obstet 1998;60:171–3.
19. Lemercier E, Genevois A, Descargue G, Clavier E, Benozio M. MRI evaluation of placenta accreta treated by embolization—apropos of a case: review of the literature [in French]. J Radiol 1999;80:383–7.
20. Lam G, Kuller J, Mcmahon M. Use of magnetic resonance imaging and ultrasound in the antenatal diagnosis of placenta accreta. J Soc Gynecol Investig 2002;9:37–40.
21. Finberg HJ, Williams JW. Placenta accreta: prospective sonographic diagnosis in patients with placenta previa and prior cesarean section. J Ultrasound Med 1992;11:333–43.
22. Levine D, Hulka CA, Ludmir J, Li W, Edelman RR. Placenta accreta: evaluation with color Doppler US, power Doppler US, and MR imaging. Radiology 1997;205:773–6.