A C TA Obstetricia et Gynecologica
AOGS RE V I EW AR TIC LE
Management of placenta percreta: a review of published cases 1,2 € CAROLINE CLAUSEN1,*, LARS LONN & JENS LANGHOFF-ROOS3 1
Department of Cardiovascular Radiology, 2Department of Vascular Surgery, and 3Department of Obstetrics and Gynecology, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark
Key words Abnormally invasive placenta, placenta percreta, placenta accreta, cesarean hysterectomy, conservative treatment, local resection Correspondence Caroline Clausen, Department of Cardiovascular Radiology, Rigshospitalet, Copenhagen University Hospital, DK-2100 Copenhagen, Denmark. E-mail:
[email protected] Conflict of interest The authors have stated explicitly that there are no conflicts of interest in connection with this article. €nn L, Please cite this article as: Clausen C, Lo Langhoff-Roos J. Management of placenta percreta: a review of published cases. Acta Obstet Gynecol Scand 2014; 93:138–143. Received: 8 July 2013 Accepted: 23 October 2013
Abstract Publications on abnormally invasive placenta in general report what can be considered a mixture of the conditions true accreta, increta and percreta varieties. The aim of this review was to identify all published cases of the most severe condition, placenta percreta in order to describe complications associated with the three commonly used surgical strategies: local resection, hysterectomy or leaving the placenta in situ, and to describe the outcome, with respect to blood loss and transfusion requirements, with the different endovascular interventions that may be used as adjuncts in the management of the conditions. A PubMed search was performed in April 2013 and the final review included 119 published placenta percreta cases. Conservative management, where the placenta is left in situ for resorption, seems to be associated with severe long-term complications of hemorrhage and infections, including a 58% risk that a hysterectomy will eventually be needed up till nine months after the delivery. Local resection seems to be associated with fewer complications within 24 h postoperatively compared with hysterectomy or leaving the placenta in situ. A selection bias in the direction of less severe cases for the local resection technique might in part explain the lower complication rates with that approach. Future prospective data collection activities should include intended as well as actual management, and long-term follow-up of all cases is of vital importance.
DOI: 10.1111/aogs.12295
Introduction Placenta accreta, or abnormally invasive placenta, is a broad term that covers conditions where the placenta in histopathological terms is a true accreta, increta or percreta (1). To distinguish between these clinical entities in clinical situations, where the invasiveness of the placenta is not always known beforehand, is not easy. Even after delivery the diagnosis is a challenge, except when the uterus is removed and is subjected to pathological examination. Publications on placenta accreta in general report what can be considered a mixture of the conditions of true
138
accreta, increta and percreta varieties, which makes it difficult to assess the results of management for each individual pathological entity. Placenta percreta is the most serious of the abnormally invasive placentas and is associated with a significantly higher maternal morbidity
Key Message Conservative management in the sense that the placenta is left in situ for resorption seems to be associated with severe long-term complications of hemorrhage and infections, including a 58% risk of hysterectomy up till nine months after delivery.
ª 2013 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 93 (2014) 138–143
Management of placenta percreta
C. Clausen et al.
than are the other varieties (2). Placenta percreta is associated with a considerably more demanding set up before and at delivery, including requirements for possible endovascular intervention such as balloon occlusion and/or embolization of the vessels supplying the placenta and the uterus (3). Furthermore, different attitudes to the surgical strategy exist, such as whether to aim for local resection, hysterectomy or leave the placenta in situ to be resorbed, as done in abdominal pregnancies (4). Since there are no prospective controlled studies on management and outcome of placenta percreta, the only option is to rely on and learn from the published individual cases or case series. We have assessed and synthesized information on the outcome of placenta percreta from available publications on varieties of abnormally invasive placenta. The aim was to identify all published cases of placenta percreta in order to describe complications associated with the three commonly used surgical strategies: local resection, hysterectomy or leaving the placenta in situ. Furthermore, we aimed to describe the blood loss and transfusion requirements, considering the different endovascular interventions which are used as a management adjunct.
Material and methods We performed a review to identify publications with individual information on cases of placenta percreta. A PubMed search was performed in April 2013 using the terms: (“Placenta Accreta”[Mesh] OR placenta increta OR placenta percreta) AND (“Embolization, Therapeutic”[Mesh] OR “Uterine Artery Embolization”[Mesh] OR “Balloon Occlusion”[Mesh] OR intravascular balloon catheter OR embolization OR balloon occlusion OR lynch suture OR cho procedure OR one-step surgery OR devascularization OR hemostatic technique). The search was restricted to publications in English.
The initial search was constructed by the three authors together, and the recovered studies were individually reviewed by one of the authors (C.C.; clarification issues handled jointly by C.C. and J.L.-R.). Studies were only included if they encompassed placenta percreta cases confirmed histologically or during surgery. Furthermore, inclusion required individual and sufficiently accurate information on the initial surgical procedure, mode of interventional radiology, if applied, postoperative complications and blood loss or transfusion requirements. From the initial search, 246 studies were identified. Fifty-two met the inclusion criteria (Fig. 1) and encompassed 119 individual cases of placenta percreta (Supporting Information Appendix S1). It was not possible to find studies of evidence levels higher than III, graded according to Oxford Centre for Evidence-Based Medicine (www.cebm. net/index.aspx?o=5653). For each placenta percreta case the following information was registered: gestational age at delivery, method of prenatal diagnosis, existence of a preoperative plan for surgery and radiological intervention, whether it was a elective or acute cesarean section, indication for delivery in cases of emergency cesarean section, initial surgical procedure, mode of interventional radiology – if applied, postoperative complications and blood loss in mL or transfusion requirements in units. Postoperative complications were divided into two groups. Immediate postoperative complications were defined as complications occurring during or up to 24 h after the initial surgery. Late postoperative complications were defined as complications occurring later than 24 h postoperatively. We compared complications by the initial surgical procedures (local resection, hysterectomy or leaving the placenta in situ) and compared blood loss by mode of interventional radiology (no use of interventional radiology, embolization, embolization and balloon occlusion, balloon occlusion of the internal iliac arteries or balloon occlusion of the common iliac arteries or the
246 studies Sufficiently accurate information on: • percreta confirmed histologically or during surgery • the initial surgical procedure • mode of interventional radiology, if applied • postoperative complications • blood loss or transfusion requirements
Yes
52 studies
Number of individual cases extracted from the included studies
No
194 studies
119 cases
Figure 1. Flow diagram of the 246 studies recovered in the initial search.
ª 2013 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 93 (2014) 138–143
139
Management of placenta percreta
C. Clausen et al.
aorta). Pearson’s chi-squared tests and Fisher’s exact tests were used.
Results Complications occurring before and after 24 h postoperatively are summarized in Table 1 by mode of initial surgical procedure. Perioperative blood loss, transfusion requirements and types of surgical approach presented by mode of endovascular intervention are summarized in Table 2. Of the 119 cases in all, 66 were managed by hysterectomy as the initial procedure. In 36 cases it was initially chosen to leave the placenta in situ and in 17 cases local
resection was performed. The hysterectomy cases were provided by 26 different centers, the in situ cases by 36 centers and the local resection data by four centers. All deliveries were cesarean sections and the majority were planned in advance (92%). However, 24 of the 119 deliveries (20%) were emergency cesarean sections. The most frequent indications were severe hemorrhage (n = 9; 8%), bleeding and preterm labor, including preterm premature rupture of membranes (n = 4; 3%), and preterm labor without bleeding (n = 5; 4%). Uterine rupture occurred in four cases (3%). Complications appearing within 24 h after the cesarean section did not differ by mode of the initial surgical
Table 1. Surgical procedures and complications in 119 placenta percreta cases
Initial surgical procedure
Hysterectomy
Local resection
Placenta left in situ
Sum
No. of cases Mode of cesarean section Elective Emergency Unspecific Complications Bladder injury/resection Salpingo-oophorectomy Post-operative hemorrhage Post-operative infection Fistula Pulmonary embolism/cardiopulmonary arrest Femoral pseudoaneurysm or distal thrombus Other re-operation Sum Secondary hysterectomy Emergency Planned Cases with one or more complications 0–24 h Cases with one or more complications >24 h
66 (56%)
17 (14%)
36 (30%)
119
41 (62%) 8 (12%) 17
16 (94%) 1 (6%) 0
20 (56%) 15 (42%) 1
77 (65%) 24 (20%) 18
11 1 5 1 2 1 4 1 26
0 0 2 0 0 0 2 1 5
6 1 16 9 1 3 0 3 39
17 2 23 10 3 4 6 5 70
– – 19 (30%) 8 (12%)
0 0 2 (12%) 2 (12%)
18 3 9 (25%) 22 (61%)
18 3 30 32
Table 2. Surgery and blood loss by mode of radiological intervention in 119 placenta percreta cases. Initial surgical procedure
Interventional radiology None Embolization Embolization and balloon occlusion Balloon occlusion, IIA Balloon occlusion, aorta or CIA Sum Data missing
Intraoperative blood loss and transfusion
No. of cases (n)
Primary hysterectomy
Placenta left in situ
Local resection
Termination of pregnancy
Intraoperative blood loss in mL (SD)
rbc units (mean)
ffp units (mean)
plt units (mean)
35 30 14
25 7 10
7 23 4
3 0 0
2 5 0
4800 (9950) 2350 (5250) 2950 (4150)
9.5 6.0 4.4
4.6 1.7 1.7
2.9 0.7 0.8
36 4
21 3
1 1
14 0
0 0
3550 (3550) 1650 (1100)
4.4 1.8
2.0 0
0.8 0
119 0
66 0
36 4
17 27
7 28
3500 (6450)
6.2
2.6
1.4
rbc, packed red blood cells; ffp, fresh frozen plasma; plt, patelets; IIA, internal iliac arteries; CIA, common iliac arteries.
140
ª 2013 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 93 (2014) 138–143
C. Clausen et al.
procedure (Table 1). For women who had placenta left in situ, 61% had complications occurring later than 24 h postoperatively, compared with 12% of those who had a hysterectomy initially or a local resection. The most frequent complications when the placenta was left in situ were secondary hysterectomy and post-operative hemorrhage. Of 36 women, 21 (58%) had a secondary hysterectomy. Three women were assigned to planned elective procedures, whereas 18 were emergency cases. Post-operative hemorrhage occurred in 44% of the women who had the placenta left in situ and 25% of suffered infections. Both post-operative hemorrhage and infections did not exceed 12% in those who had a hysterectomy initially or who had local resection. Bladder injury or resection occurred in 11 of the 66 women in the hysterectomy group and in six of the 36 women who had the placenta left in situ, accounting for 17% in both groups. Neither bladder injuries nor resections were registered in the local resection group. Where the placenta was left in situ, 42% of the deliveries were emergency cesarean sections as opposed to 6% for the local resections and 12% when hysterectomy was performed electively. The estimated average blood loss of 1650 mL (SD 1100 mL) included cases where endovascular balloon occlusion of the common iliac arteries or the aorta was used. When the balloon occlusion was placed more distally, in the internal iliac arteries, the estimated blood loss was 3550 mL (SD 3550 mL). The women having embolization or the combination of embolization and balloon occlusion had estimated blood losses of 2350 mL (SD 5250 mL) and 2950 mL (SD 4150 mL), respectively, compared with 4800 mL (SD 9950 mL) when adjunctive interventional radiology was not used. The group managed with balloon occlusion of the common iliac arteries or the aorta required 1.8 units of packed red blood cells. The groups who had the more distal balloon occlusion of the internal iliac arteries or the combination of embolization and balloon occlusion received 4.4 units of packed red blood cells, whereas embolization alone required 6.0 units. In the group managed without endovascular procedures, 9.5 units of packed red blood cells were given.
Discussion Conservative management in the sense that the placenta is left in situ for resorption seems to be associated with severe long-term complications. Of the women who had the placenta left in situ, 61% suffered at least one late postoperative complication, compared with 12% of the women with local resection or initial management by hysterectomy. The complications associated with leaving the placenta in situ were primarily hemorrhage and infections,
Management of placenta percreta
including a 58% risk of hysterectomy up till nine months after delivery. Following local resection, there were no secondary hysterectomies. However, one has to consider that the choice for local resection is probably often first made when the placenta is diagnosed to be percreta and found to be accessible for primary resection as judged at the time from an assessment of the exact localization and extension into the uterine wall and adjacent tissues. The appearance at that time may discourage the surgeon from trying a local resection and, if blood loss is limited, the findings may tempt the operating team to leave the placenta in situ. We did not find information on the intended mode of surgery in the cases reported. This would have been of interest especially regarding hysterectomy, which is sometimes the preferred mode for mothers of an advanced age who do not want more children or for women known to have fears relating to the risk of recurrence of an abnormally adherent placenta. The decision on whether to choose a hysterectomy electively, leave the placenta in situ or perform local resection is also based on and influenced by the available expertise in each center. Due to the infrequency of placenta percreta it is often not possible to acquire proficiency in the local resection technique. Not all centers have an interventional radiological department, and to perform a hysterectomy when a placenta percreta site is bleeding heavily requires significant experience. A hysterectomy was chosen in the majority of the cases included in this review. However, the 66 hysterectomy cases were from 26 centers and the 17 local resection from only four centers. In comparison, the 36 cases where placenta was left in situ were from 25 centers. Leaving the placenta in situ was associated with multiple complications. Of the women with the placenta left in situ, 42% required emergency or sub-acute surgery at some point compared with 6% in the local resection group and 12% in the hysterectomy group. This high rate of emergency surgery may have contributed to an increased risk of complications when the placenta was left in situ. Regarding immediate postoperative complications, only the local resection group stood out, with 12% of the women suffering one or more complications, compared with 25–30% in both the hysterectomy group and the group with placenta left in situ. Since the excess morbidity in cases where the placenta was left in situ occurred mainly after 24 h, we suggest that the obstetricians who accidentally find a placenta percreta or who are not prepared for a primary repair, consider leaving the placenta in situ for subsequent referral of the patient to a tertiary center within 24 h for a secondary local resection and repair. The necessary expert knowledge and experience must be available at the tertiary center.
ª 2013 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 93 (2014) 138–143
141
Management of placenta percreta
Bladder injuries/resections were registered in 17% of cases when hysterectomy was the initial procedure or the placenta was left in situ. This may indicate an over-representation of more severe cases compared with when local resection alone was chosen. The frequency of bladder injuries/resections when the placenta is left in situ by choice beforehand appeared to be the same as when a hysterectomy is performed directly after the delivery. This was possibly due to the high frequency of secondary hysterectomy (58%) when the placenta was left in situ. In previous publications it has been found that a prenatal diagnosis of placenta percreta is associated with a lower complication rate than when it is detected at operation (5). Only 8% of the women in this material did not have a prenatal diagnosis, which could either reflect a publication bias (which we could not assess) or placenta percreta cases being most often diagnosed before delivery, in contrast to the abnormally invasive placentas that do not go through the uterine wall. Although the majority of the women had a prenatal diagnosis of suspected placenta percreta, about one-fourth of the cesarean deliveries were performed in some context of an emergency procedure. This could be due to a tendency to plan the elective cesarean sections at too advanced a gestational age. However, the optimal timing of delivery for these patients remains controversial (6). The estimated average blood loss of 1650 mL included cases where endovascular balloon occlusion of the common iliac arteries or the aorta was used. This was less than reported in the other treatment groups, but applied to only four cases. The women having embolization, balloon occlusion of the internal iliac arteries or a combination of the two techniques had estimated blood losses of 2350–3550 mL, compared with 4800 mL when adjunctive interventional radiology was not used. The estimated average blood loss in the embolization group was slightly less than in the group of balloon occlusion of the internal iliac arteries (2359 mL vs. 3550 mL). However, the amount of red blood cells transfused in the embolization group was 6.0 units and exceeded the amount transfused in the group of women with balloon occlusion of the internal iliac arteries, which was 4.4 units. This illustrates a substantial discrepancy between reported transfusion requirements and estimated blood loss and emphasizes the difficulty in comparing these outcome variables between the different techniques. The discrepancy has also been noted in other studies on transfusion practices (7). Main outcome variables vary among the included studies and transfusion protocols and the way intra-operative blood loss is estimated also differ substantially. The reporting of complications varied from only the most severe being reported, to also including mild complications with
142
C. Clausen et al.
only minor clinical impact. Therefore a comparison of results from different studies is subject to inaccuracies. This study is in essence an overview and cannot approach the reliability offered by a proper systematic review, since there are no randomized controlled studies or controlled observational studies available. However, a systematic search and approach to what is available in terms of uncontrolled and purely descriptive studies was required, in order to assess this material and draw upon it for collective information. A potential publication bias can, therefore, neither be avoided nor quantified. We presume that the bias originates from a tendency to publish very complicated but successful cases. However, there is no reason to believe that there is a large difference in publication bias by mode of management. Still, the methodological shortcomings in the included studies and this attempt to synthesize information from them imply that the results should be interpreted with caution. The general problem with evaluation of management, i.e. that acquired skills and experience are associated with the choice and outcome of a specific mode of management, should be kept in mind. Most publications on the abnormally invasive placenta, including reviews, have included a highly variable mix of women with placenta accreta, increta and percreta. The process used in the present study, where we aimed to include only placenta percreta cases, was complicated due to this lack of systematic and uniform reporting. When lack of transparency did not provide grounds for extraction of case-based information, the reported cases were excluded. Considering that placenta percreta is an uncommon condition, a relatively large material was obtained by compiling all the published cases according to predefined inclusion criteria. This clearly indicates the need for a standardized protocol when reporting future results. The low incidence also calls for an international collaborative effort such as the International Network of Obstetrics Survey Systems (INOSS) (www.npeu.ox.ac.uk/inoss) and the European Working group on Abnormally Invasive Placenta (www. EW-AIP.org) (8–10). Meanwhile, it seems that conservative management of placenta percreta, in the sense that the placenta is left in situ for resorption, is associated with severe long-term complications compared with local resection or hysterectomy as the initial choice. The complications associated with leaving the placenta in situ are primarily hemorrhage and infections, including a very high risk of hysterectomy up to at least nine months after delivery.
Funding The authors declare no source of funding.
ª 2013 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 93 (2014) 138–143
C. Clausen et al.
References 1. Chantraine F, Langhoff-Roos J. Abnormally invasive placenta – AIP. Awareness and pro-active management is necessary. Acta Obstet Gynecol Scand. 2013;92:369–71. 2. Palacios-Jaraquemada JM. Caesarean section in cases of placenta praevia and accreta. Best Pract Res Clin Obstet Gynaecol. 2013;27:221–32. 3. Palacios Jaraquemada JM, Pesaresi M, Nassif JC, Hermosid S. Anterior placenta percreta: surgical approach, hemostasis and uterine repair. Acta Obstet Gynecol Scand. 2004;83:738–44. 4. Clausen C, Stensballe J, Albrechtsen CK, Hansen MA, Lonn L, Langhoff-Roos J. Balloon occlusion of the internal iliac arteries in the multidisciplinary management of placenta percreta. Acta Obstet Gynecol Scand. 2013;92:386–91. 5. Kayem G, Davy C, Goffinet F, Thomas C, Clement D, Cabrol D. Conservative versus extirpative management in cases of placenta accreta. Obstet Gynecol. 2004;104:531–6. 6. Sentilhes L, Goffinet F, Kayem G. Management of placenta accreta. Acta Obstet Gynecol Scand. 2013;92:1125–34. 7. Chantraine F, Braun T, Gonser M, Henrich W, Tutschek B. Prenatal diagnosis of abnormally invasive placenta
Management of placenta percreta
reduces maternal peripartum hemorrhage and morbidity. Acta Obstet Gynecol Scand. 2013;92:439–44. 8. Langhoff-Roos J, Chantraine F, Geirsson RT. AIP (abnormally invasive placenta) – from a retained placenta to destruction of the uterine wall. Acta Obstet Gynecol Scand. 2013;92:367–8. 9. Bonnet MP, eux-Tharaux C, Dupont C, Rudigoz RC, Bouvier-Colle MH. Transfusion practices in postpartum hemorrhage: a population-based study. Acta Obstet Gynecol Scand. 2013;92:404–13. 10. Kayem G, Deneux-Tharaux C, Sentilhes L. PACCRETA: clinical situations at high risk of placenta ACCRETA/ percreta: impact of diagnostic methods and management on maternal morbidity. Acta Obstet Gynecol Scand. 2013;92:476–82.
Supporting information Additional Supporting Information may be found in the online version of this article: Appendix S1. Reference list of included case reports and series (1–52).
ª 2013 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 93 (2014) 138–143
143
