Unexpected consequences of a cesarean delivery

The classical long-term effects of a CD such as uterine rupture, placenta previa, pathologically adherent placenta in a subsequent pregnancy, ectopic pregnancy, infertility, and intraabdominal adhesions among others were duly described in the literature and in many textbooks.

Pathologically adherent placenta at the site of a cesarean section scar in a subsequent pregnancy

The incidence of pathologically adherent placentae (accreta-increta and percreta) increases with the increasing number of CDs.

Placenta accreta (PA) (we will use this term to refer to all 3 forms of pathologically adherent placentae [eg, such as accreta, increta and percreta]) was rare between 1930 and 1950, with a frequency of 1 in 30,000 deliveries. However, this frequency rose to 1/19,000 between 1950 and 1960. By 1980, the rate increased to 1/7,000. A 4-fold increase in frequency was noted between 1994 and 2002. Several investigators cite the frequency of 1/2000 to 1/2500 deliveries. Others estimate the frequency to be as high as 3/1000 deliveries occurring in the last decade.

This drastic change in the number of patients with a potential PA fundamentally changed the diagnostic approach as well as the obstetric (surgical) management of this pregnancy complication. Early and reliable diagnosis of PA, knowledge of its natural history, and the realization of the fact that most are implanted in the scar of the previous CD are slowly focusing attention to its first-trimester complications.

A detailed discussion of sonographic and clinical aspects surrounding placenta accreta in the second half of pregnancy is not the focus of this review because it is available in textbooks as well as review articles. The first purpose of this article is to review the consequences of placenta accreta in the first and early second trimester of pregnancy with an emphasis on a possible spontaneous rupture of the uterus and/or profuse bleeding because of implantation in the pregnancy in the hysterotomy scar of a previous cesarean delivery. Even though in this review we discuss placenta accreta occurring in the first and early second trimester (inclusive of 14-16 weeks) to underline the importance of recognizing the threat of a severe complication of placenta accreta before the third trimester and/or delivery, we included a handful of articles describing disastrous outcomes in midpregnancy. In these cases, the clinical picture and outcomes matched or were very similar to those occurring in the first and early second trimester.

Pathologically adherent placenta at the site of a cesarean section scar in a subsequent pregnancy

The incidence of pathologically adherent placentae (accreta-increta and percreta) increases with the increasing number of CDs.

Placenta accreta (PA) (we will use this term to refer to all 3 forms of pathologically adherent placentae [eg, such as accreta, increta and percreta]) was rare between 1930 and 1950, with a frequency of 1 in 30,000 deliveries. However, this frequency rose to 1/19,000 between 1950 and 1960. By 1980, the rate increased to 1/7,000. A 4-fold increase in frequency was noted between 1994 and 2002. Several investigators cite the frequency of 1/2000 to 1/2500 deliveries. Others estimate the frequency to be as high as 3/1000 deliveries occurring in the last decade.

This drastic change in the number of patients with a potential PA fundamentally changed the diagnostic approach as well as the obstetric (surgical) management of this pregnancy complication. Early and reliable diagnosis of PA, knowledge of its natural history, and the realization of the fact that most are implanted in the scar of the previous CD are slowly focusing attention to its first-trimester complications.

A detailed discussion of sonographic and clinical aspects surrounding placenta accreta in the second half of pregnancy is not the focus of this review because it is available in textbooks as well as review articles. The first purpose of this article is to review the consequences of placenta accreta in the first and early second trimester of pregnancy with an emphasis on a possible spontaneous rupture of the uterus and/or profuse bleeding because of implantation in the pregnancy in the hysterotomy scar of a previous cesarean delivery. Even though in this review we discuss placenta accreta occurring in the first and early second trimester (inclusive of 14-16 weeks) to underline the importance of recognizing the threat of a severe complication of placenta accreta before the third trimester and/or delivery, we included a handful of articles describing disastrous outcomes in midpregnancy. In these cases, the clinical picture and outcomes matched or were very similar to those occurring in the first and early second trimester.

Cesarean section scar pregnancy

Another less well-known and less studied consequence of CD is the cesarean section scar pregnancy (CSP), which, by definition, is not a classical ectopic pregnancy. The term CSP, scar pregnancy, or even the sporadically used term, isthmic pregnancy, refers to a gestational sac that has implanted in the scar or the niche of a previous cesarean delivery ( Figure 1 ) .

Number of early placenta accreta and cesarean scar pregnancy articles based on their year of publication

Timor-Tritsch. Early placenta accreta and cesarean section scar pregnancy: a review. Am J Obstet Gynecol 2012.

This serious consequence of the increasing rate of cesarean deliveries has consistently appeared in the literature in the last 10 years. CSP is fundamentally different from a cornual and tubal as well as a cervical pregnancy. Our second purpose of this article was to describe varying approaches in the diagnosis and management of the CSP along with its complications. We also provide support to demonstrate a possible link between the pathogenesis of PA and CSP.

Because the condition is relatively rare and has an unusually high complication rate, CSPs are likely to be published as single case reports or a case series containing few patients.

Methodology of the review

To be able to draw practical clinical conclusions regarding the early placenta accreta (EPA) and CSP, an in-depth literature search was performed. We searched PubMed/Medline electronic databases for key words such as “pathologically adherent placenta,” “placenta accreta,” “placenta increta,” “placenta percreta,” “spontaneous uterine rupture,” “first trimester,” “second trimester,” “cesarean scar pregnancy,” “cesarean scar,” “cesarean scar ectopic,” “previous cesarean section scar,” “first-trimester accreta,” “scar implantation,” “scar pregnancy,” and “isthmic pregnancy.” The articles obtained were cross-referenced by the “related articles” and “link” enabled by PubMed.

We also searched Embase, Web of Science, and BIOSIS and Scholar Google for additional search of the Internet. Additional references were also looked for in the respective reference lists of the citations cited in previous text. We reviewed original articles in English, French, German, Spanish, and Portuguese languages. In the case of other foreign languages (eg, Chinese, etc), we relied on the English abstract.

This review was focused not only on finding the possible causative and pathophysiologic link(s) between early placenta accreta and cesarean section scar pregnancy but also primarily on their clinical presentations, management approaches, and complications to provide the most focused answer.

The literature search yielded a total of 204 publications between 1972 and 2011. There were 49 articles related to early complications of PA and 176 articles dealing with CSP. These will be mentioned in detail in the following text in the article. Eleven review articles about PA provided information about the pathogenesis, incidence, diagnosis, and clinical behavior of first- and second-trimester PA. These were evaluated and referenced. There were 6 review articles related to CSP without case descriptions. There were numerous case reports with the subject of which were both PA and CSP at the end of which there was usually review of the literature. Most case series or case presentations expressed personal views based on either a few cases or a single case.

Results of the literature review on early placenta accreta

Forty-two articles containing descriptions of 47 cases of PA detected and treated in the first and second trimesters were found. We included 2 cases at 27 and 28 weeks because their clinical courses and outcomes were identical to those in the first trimester. Table 1 contains detailed information about their history, previous CDs’ gestational age at presentation, clinical presentation, treatment, and outcome. Nine patients did not have a history of CD (19%), 31 patients had 1 CD (66%), 6 patients had 2 (12.7%), and 1 patient had 3 previous cesarean sections.

In analyzing the published material, several observations can be made based on the 42 articles: 29 were published in or after the year 2000. This increasing number of publications reflects the increasing rate of CDs. Nine of the silent uterine ruptures occurred in patients without a history of previous CDs. Eleven cases had an elective dilation and curettage (D&C) followed by bleeding and hysterectomy. This fact raises questions about the difficult aspects of sonographic diagnosis of early PA and CSP.

Spontaneous rupture of the uterus was reported in 15 of the 47 cases (32%), most of them being silent ruptures and were followed by bleeding, shock, and hemoperitoneum leading to laparotomy, uterine artery embolization, or hysterectomy. Mean gestational age at rupture was 18.1 weeks (range, 6–28 weeks). Based on the case histories, it is reasonable to surmise that many other patients, whose reports did not specifically contain the words “rupture,” “uterine rupture,” or “silent rupture,” may have had a ruptured uterus because “severe bleeding,” “acute abdomen,” “profuse bleeding,” “shock,” or “hemoperitoneum” were mentioned. Five patients (10.6%) had uterine artery embolizations. Thirty-seven patients (78.7%) had laparotomies and 35 (74.4%) had hysterectomies (78.7%). One patient had an arteriovenous malformation after the D&C with subsequent embolization requiring a hysterectomy.

In 10 articles discussing patients with uterine ruptures in the first trimester who had previous cesarean sections, the clinical descriptions are identical for both the early PA and that of a CSP. These cases were counted in both series. Reading the case description, it is obvious that the rupture occurred due to and at the site of the CD scar. Striking similarities make these cases hard to classify as pure PA or, for that matter, pure cases of CSP. This may be another compelling reason to believe in the hypothesis of their common pathogenesis, at least in those with a history of previous CD.

Results of the literature search on CSP

One question is whether the diagnostic entity of CSP is indeed new. Between 1990 and 1999, we found only 19 articles regarding the subject. Between 2000 and 2005, 48 articles were found, and 104 articles were published from 2006 to 2011 ( Figure 1 ). This increase in publications is testament to the mounting awareness and attention now focused on the diagnostic challenges of this new disease.

We identified 751 CSP cases with sufficient clinical information to be evaluated for this review. We will review the diagnosis and treatment of this condition.

Undoubtedly and clearly, as a result of the mounting risk of the cesarean section delivery, the obstetrical and gynecologic community started to become aware of its unpredicted consequences. This increased awareness is reflected in the steadily increasing number of articles dealing with the problems. As mentioned before, Figure 1 shows the number of publications containing cases, case series, and reviews for the last 20 years. Their numbers steadily increased toward the year 2010. This increasing number of articles supports the fact that CSP is becoming another diagnostic and clinical entity in urgent need of a well-thought-out and applied standardized treatment protocol.

The diagnosis of CSP

Our own experience with scanning 26 cases with CSP made us aware of the difficulty in the diagnosis. Indeed, in at least 107 of the 751 cases (13.6%), the diagnosis of CSP was missed. Reading carefully the entire text of case presentation in all reviewed articles, there were at least as many cases, if not more, in which the authors did not clearly report that the diagnosis was missed. However, indirectly, based on the description of the events (“heavy bleeding,” “shock,” “hemoperitoneum,” etc) after D&C for “termination of an early pregnancy” or D&C for missed abortion, we suspected that in these cases there were some problems establishing the primary diagnosis. Because of the uncertainty of the diagnosis of CSP, these cases were not included in the 107 described in the previous text.

Findings of arterio venous malformations (A-VM)

There were 8 cases with descriptions of A-VM detected in the placentae in the scar. Some of these were diagnosed prior to interventions; others were detected as a result of treatment (current study). The primary management of 3 of these was by uterine artery embolization ; the other 5 were managed with hysterectomies. In our own series of 26 CSPs, we had 2 cases with A-VM. The first of these 2 cases presented as a possible complication of a D&C procedure, whereas the second one presented to the emergency department with heavy vaginal bleeding and was promptly treated with uterine artery embolization.

Heterotopic locations, twin/triplet, and recurrent CSP

Heterotopic pregnancies involving a CSP with singleton or twin intrauterine gestations were reported in 11 articles. The management of these cases will be discussed in the following text. Four twin cesarean section scar pregnancies were reported in 3 articles : 1 single case of triplet CSP was reported. Six cases with 1 single recurrent CSP; ; 2 unusual cases of 3 consecutive CSPs ; 1 molar pregnancy, and 1 case of an endometrioma in the scar were also reported. Sepsis secondary to cesarean section scar diverticulum was also seen.

Future obstetric performance after CSP

Our ability to properly counsel women who did not have their uteri removed and desired to become pregnant again has to rely on the reported literature. We therefore reviewed the literature reporting on pregnancies after a successful treatment of a CSP. Sixty-four intrauterine pregnancies after CSP were reported. We speculate that these cases (ie, pregnancies and deliveries after CSP) are underreported. We can only hypothesize that even though many patients who after treatment of CSP retain their ability to have another pregnancy may not want to do so. Another reason would be that practitioners do not see a successful pregnancy after a previous CSP as worthy of publication, just as few will report pregnancies after ectopic gestations. This issue requires a more in-depth study.

The different treatment modalities used in managing CSP

Analyzing the different primary management protocols and approaches led us to our greatest surprise. There were no less than 31 primary approaches to treat the 751 CSPs. These are summarized in Table 2 . The sporadic, mostly individual cases and their results were insufficient to enable to draw a clear conclusion as to which was the most effective management protocol leading to the least or no complications. The result of such a large variety of treatment approaches and their different combinations reported in relatively small case series left the community of obstetricians and gynecologists relatively unsure as to the best approach for treatment.

Importantly, without a wide review, it is almost impossible to assess which treatment had the most complications and therefore should be avoided. Based on reviewing the literature and observing treatment choice of most authors, we realized that gynecologic surgeons tended to do D&C, laparoscopy, and hysteroscopy as their preferred first-line approach. Obstetricians preferred injecting the chorionic sac and relied on the help of interventional radiology. Many times the choice of the patient to have a hysterectomy performed was granted. Few countries have a set of guidelines at hand when a patient has an early PA or a CSP. An overview of the individual treatment modalities including their references are detailed in Table 3 .

Treatment complications of CSP

Complications were defined as the immediate or delayed need for a secondary treatment. The most frequent reasons for the additional treatment involved blood loss defined by blood loss requiring transfusion; greater than 200 mL; heavy bleeding; shock; hemoperitoneum or surgical interventions such as laparoscopy, hysteroscopy, laparotomy, hysterectomy, or procedures reported as requiring general anesthesia; or embolization of the uterine arteries. If patients required more than 2 different, additional consecutive treatments short of major surgical intervention or embolization, the case was counted as a complication.

We compiled the number of times each treatment approach was used and listed the complications in each of the 31 treatment modalities ( Table 3 ). There were a total of 331 cases with complications in the 751 CSPs (44.1%). Table 4 demonstrates the 3 first-line treatment approaches that, if used alone or in any combination, resulted in the largest number of complications. Treatments that featured uterine artery embolization or D&C alone, or in any “combination treatment,” had complication rates (mean, 46.9; range, 0–80% and mean, 61.9%; range, 29.5–100, respectively). The reason is probably the fact that unlike the multilayered myometrium in the uterine body, which are able to contain bleeding at the placental site after its separation, the vessels exposed by curettage bleed because the scar tissue is unable to contract and contain the profuse bleeding.

Intramuscular methotrexate without any other treatment in combination had a 62.1% complication rate. Intramuscular methotrexate and D&C in combination had a complication rate of 86%. The only speculation we may offer to explain the failure of this primary noninvasive treatment approach is due to the slow action of the drug and its questionable ability to stop the cardiac activity and placental expansion. The expected result may take days. While waiting for the drug effect, the gestational sac, the embryo/fetus, and its vascularity are growing. In this case, a secondary treatment has to be directed toward a larger gestation with a richer vascularization; thus, the risk of complications may increase.

Table 4 contains the first-line treatment choices that, if used alone or in any combination, resulted in the lowest complication rate. Any treatment involving hysteroscopy had a relatively low complication rate (18.4%). The lowest complication rates were those using local, intragestational sac injection of methotrexate or KCl (KCl was used only for heterotopic CSP) under transabdominal and/or transvaginal real-time ultrasound guidance (8 of 81 cases: 9.6%). Our previous experience managing multifetal pregnancy reductions and various types of ectopic pregnancies using transvaginal ultrasound guided local intragestational injections leads us to the belief that locally injecting methotrexate in cases of CSP is minimally invasive and leads to a rather quick resolution with fewer complications.

There were several treatments resulting in a very large or very small complication rate; however, these were single cases or case series with a very low number of patients. Although included in the tables, these treatments should not be considered as a guide for their use or to be considered as preferential first-line treatments until they are proven by evaluating their efficacy in a larger number of patients.

The management approach suggested by Jurkovic et al deserves some discussion. This author used a Shirodkar cervical suture in arresting hemorrhage following D&C of 28 scar pregnancies. The suture was left untied until the end of the surgical procedure. In 22 women, the suture was tied to achieve hemostasis because they “bleed significantly during the evaluation of scar pregnancies.” For the purpose of this review, these cases were counted as complications.

Outcomes after treatments of CSP

Excluding 49 cases of uterine artery embolizations, 14 cases of laparotomies, and 3 elective hysterectomies, which were rendered as planned primary treatment approaches, there were emergency surgeries performed to treat complications. Thirty-six hysterectomies, 40 laparotomies, and 22 uterine artery embolizations were performed, mostly to rescue patients from life-threatening hemorrhages. We were not able to assess the number of patients receiving general anesthesia, and because most therapeutic approaches (hysteroscopy, D&C, laparoscopy, embolization, etc) are usually performed under general anesthesia, one should also keep in perspective the risks general or regional anesthesia present.