Cesarean scar pregnancy: issues in management
Abstract
Objective
To evaluate our experience with the diagnosis and treatment of Cesarean scar pregnancy.
Methods
During a 6-year period, 12 cases of Cesarean scar pregnancy were diagnosed using transvaginal color Doppler sonography and treated conservatively to preserve fertility. Incidence, gestational age, sonographic findings, β-human chorionic gonadotropin ( β-hCG) levels, flow profiles of transvaginal color Doppler ultrasound, and methods of treatment were recorded.
Results
The incidence of Cesarean scar pregnancy was 1:2216 and its rate was 6.1% in women with an ectopic pregnancy and at least one previous Cesarean section. Gestational age at diagnosis ranged from 5 + 0 to 12 + 4 weeks. The time interval from the last Cesarean section to the diagnosis of Cesarean scar pregnancy ranged from 6 months to 12 years. High-velocity and low-impedance subtrophoblastic flow (resistance index, 0.38) persisted until β-hCG declined to normal. Patients were treated as follows: transvaginal ultrasound-guided injection of methotrexate into the embryo or gestational sac (n = 3), transabdominal ultrasound-guided injection of methotrexate (n = 2), transabdominal ultrasound-guided injection of methotrexate followed by systemic methotrexate administration (n = 2), systemic methotrexate administration alone (n = 2), dilatation and curettage (n = 2), or local resection of the gestation mass (n = 1). Eleven of the 12 patients preserved their reproductive capacity; the remaining patient, treated by dilatation and curettage, underwent a hysterectomy because of profuse vaginal bleeding. The Cesarean scar mass regressed from 2 months to as long as 1 year after treatment. Uterine rupture occurred in one patient during the following pregnancy at 38 + 3 weeks' gestational age.
Conclusion
Ultrasound-guided methotrexate injection emerges as the treatment of choice to terminate Cesarean scar pregnancy. Surgical or invasive techniques, including dilatation and curettage are not recommended for Cesarean scar pregnancy due to high morbidity and poor prognosis. Copyright © 2004 ISUOG. Published by John Wiley & Sons, Ltd.
Introduction
Cesarean scar pregnancy is defined as an ectopic pregnancy embedded in the myometrium of a previous Cesarean scar1. The incidence of Cesarean scar pregnancy is extremely low. A review of the published literature from January 1966 to October 2001 was performed by means of a computerized database search of MEDLINE. Medical subject heading search words used were ‘Cesarean scar’ + ‘ectopic pregnancy’ and the search was limited to the English language. This was supplemented by a manual search of the reference lists of the articles for additional relevant case reports. Our search produced only 15 articles1-15.
Ultrasound permits earlier and accurate diagnosis of Cesarean scar pregnancy, allowing successful preservation of the uterus without causing maternal complications2, 8, 12, 13. Treatments include local resection of the ectopic gestational mass, dilatation and curettage (D&C), and systemic or local administration of methotrexate. We present our experience with 12 patients with Cesarean scar pregnancy diagnosed using transvaginal color Doppler ultrasound during a 6-year period and treated conservatively to preserve the uterus.
Methods
From January 1995 to December 2000, a total of 642 ectopic pregnancies was diagnosed in Shin Kong Wu Ho-Su Memorial Medical Referral Center. Among these, 198 patients had a previous Cesarean section before the ectopic pregnancy; 12 of these were diagnosed with Cesarean scar pregnancy using transvaginal ultrasound. During this 6-year period, a total of 26 596 newborns was delivered in our hospital; 7980 of these deliveries were via Cesarean section. The Cesarean section rate was 30%.
The diagnosis of Cesarean scar pregnancy was confirmed if all of the following sonographic criteria were met: (1) the uterus was empty, with clearly demonstrated endometrium; (2) the cervical canal was empty, without a gestational sac or ballooning at the early diagnosis; (3) the gestational sac showed the ‘double ring’ sign in the anterior part of the isthmic portion of the uterus; (4) the gestational sac, with or without cardiac activity, was embedded and surrounded by the myometrium, the fibrous tissue of the scar, and it was separated from the endometrial cavity or Fallopian tube (Figure 1). This is different from a normally eccentric lower segment implantation in which the main gestational sac is located in the endometrial canal.
Transvaginal ultrasound image of a Cesarean scar pregnancy at 6 + 1 weeks, with cardiac activity (Case 10; crown–rump length, 6.3 mm). A sac is visible within the myometrium of the retroverted uterus with the endometrial and cervical canals both empty (arrows). The bladder wall (arrowheads) is seen anterior to the gestational sac.
After counseling about the risk of hysterectomy, all patients underwent conservative treatment including methotrexate injection, D&C, or wedge resection of the ectopic pregnancy, because preservation of fertility was desired.
The method of treatment was according to the decision made by the individual physicians. This study was reviewed and approved by the institutional review board of Shin Kong Wu Ho-Su Memorial Medical Center. All patients gave informed consent before treatment.
Local methotrexate injection under transabdominal ultrasound guidance was performed using a 22-gauge long needle (Tokyo Co., Tokyo, Japan). Using a needle guide, 1 mg/kg methotrexate was injected into the chorionic sac or embryo. This procedure was performed without anesthesia. Just prior to methotrexate injection, the chorionic sac cavity, amniotic fluid and gestational tissues were aspirated. Penetration of the bladder by the needle was avoided during the procedure.
To reduce the patient's discomfort and anxiety, local methotrexate injection under transvaginal ultrasound guidance was performed under general anesthesia with the patient in the lithotomy position. The gestational sac was punctured and aspirated under transvaginal sonographic guidance using a double-lumen ovum aspiration needle (ECHO-Tip, Cook, Sydney, Australia) (Figure 2). Afterwards, 1 mg/kg methotrexate was injected slowly, resulting in a hyperechogenic area in the region of the former gestational sac.
Image of puncture line for transvaginal ultrasound-guided local methotrexate injection in Case 12.
β-hCG testing was performed daily during hospitalization and weekly thereafter, until the level returned to <5 mIU/mL.
Transvaginal sonography using a Toshiba SSA-270A scanner (Toshiba Co., Tokyo, Japan) equipped with a 6.25-MHz vaginal probe with color and pulsed wave Doppler was used to monitor subtrophoblastic blood flow velocity. Vascular pattern, vascular waveform, pulsatility index, resistance index and peak systolic velocities were checked weekly until no subtrophoblastic blood flow velocity could be detected. The diameter of the Cesarean scar ectopic pregnancy mass was measured in centimeters if it persisted.
Results
Data are presented in Table 1. The mean age of the patients was 32 (range, 27–41) years. The time interval between the current Cesarean scar pregnancy and the previous Cesarean section ranged from 6 months to 12 years. Gestational age at diagnosis ranged from 5 + 0 weeks to 12 + 4 weeks. Five gestational sacs showed cardiac activity. With the exception of Case 5, in which the pregnancy was the result of in-vitro fertilization–embryo transfer, the patients conceived naturally. All patients complained of vaginal bleeding, and six had accompanying abdominal pain at the time of diagnosis; this is the reason these patients were referred to our medical center.
| Case | Age (years) | Time interval from last C/S (years) | Gestational age at diagnosis (weeks) | Cardiac activity in gestational sac (Yes/No) | Initial β-hCG level (mIU/mL) | Type of treatment | Time after treatment for β-hCG to reach normal level (days) | Time after treatment for flow* to disappear (days) | Pertinent follow-up information |
|---|---|---|---|---|---|---|---|---|---|
| 1 | 29 | 3 | 7 + 6 | No | 35 927 | D&C | 36 | — | Hysterectomy due to massive bleeding during D&C. Microscopic findings: chorionic villi and trophoblastic tissue inside the dense and fibrotic myometrium. |
| 2 | 35 | 13 & 10 | 7 + 4 | Yes | 19 755 | Wedge resection | 21 | — | No residual tissue. Microscopic findings: the myometrium adjacent to the chorionic villi showed dense interstitial fibrosis. |
| 3 | 31 | 7 | 7 + 3 | No | 36 023 | TAS local MTX + systemic MTX | 188 | 117 | Disappearance of ectopic mass after 8 months. |
| 4 | 30 | 7 & 4 | 5 + 0 | No | 3 217 | Systemic MTX | 30 | 14 | Disappearance of ectopic mass after 2 months. |
| 5† | 41 | 12 | 6 + 1 | No | 23 328 | TAS local MTX + systemic MTX | 76 | 90 | IVF treatment. Disappearance of ectopic mass after 1 year. |
| 6 | 28 | 5 | 6 + 0 | Yes | 16 628 | TAS local MTX | 56 | 70 | Disappearance of ectopic mass after 6 months. |
| 7 | 35 | 5 & 3 | 6 + 4 | Yes | 19 086 | TAS local MTX | 75 | 85 | Disappearance of ectopic mass after 4 months |
| 8 | 31 | 1.5 | 12 + 4 | No | 58 400 | D&C + cervical balloon tamponade | 63 | 58 | Blood loss 1000 mL during D&C; 3 months later patient was pregnant again but died due to uterine rupture at 38 + 3 weeks. |
| 9 | 30 | 0.5 | 8 + 4 | Yes | 47 752 | TVS local MTX | 117 | 130 | Persistent ectopic mass for 5 months (4.5 × 3.82 cm). |
| 10 | 27 | 4 | 6 + 1 | Yes | 24 195 | TVS local MTX | 78 | 92 | Persistent ectopic mass for 4.5 months (2.24 × 1.75). |
| 11 | 36 | 2 | 5 + 2 | No | 4 280 | Systemic MTX | 36 | 45 | Persistent ectopic mass for 10 months (2.11 × 1.51 cm). |
| 12 | 32 | 8 | 6 + 2 | No | 20 122 | TVS local MTX | 76 | 85 | Resolution of ectopic mass after 1 year. |
- * Subtrophoblastic flow.
- † This pregnancy was induced by in-vitro fertilization and embryo transfer. C/S, Cesarean section; D&C, dilatation and curettage, IVF, in-vitro fertilization; MTX, methotrexate; TAS, transabdominal ultrasound guidance; TVS, transvaginal ultrasound guidance.
The patient in Case 1 underwent a D&C for suspected incomplete miscarriage, since the gestational sac was located in the lower segment of the uterus. Uncontrolled bleeding of about 1000 mL occurred during the procedure, and an emergency hysterectomy was performed. During the hysterectomy, inadvertent perforation occurred during bladder separation. The microscopic findings revealed many chorionic villi and trophoblastic tissue inside the dense, interstitial, fibrotic myometrium.
A D&C was also performed in Case 8. This woman was only diagnosed with Cesarean scar pregnancy at 12 + 4 weeks of gestation due to the misdiagnosis of inevitable miscarriage. The gestational sac measured 6.37 cm in diameter and was located within the isthmus area. Profuse bleeding estimated to be 1000 mL occurred within a 10-min period during the D&C. A Foley catheter balloon filled with 50 mL normal saline was used for tamponade, and the uterus was successfully preserved.
Transvaginal color Doppler imaging of the patient in Case 8 demonstrated a complex vascular mass with an area of neovascularization surrounding the placental area in the lower uterine segment. A high peak systolic velocity of 65.2 cm/s and low impedance (mean resistance index, 0.38) were observed at the time of diagnosis (Figure 3). This low impedance and high subtrophoblastic flow velocity was consistent with a normal intrauterine pregnancy and remained unchanged until the serum β-hCG declined to normal. Thereafter, this unique area of vascular flow decreased progressively and disappeared within 2 weeks.
Doppler velocimetry demonstrated (Case 8) that high-velocity (peak systolic velocity, 65.2 cm/s) and low-impedance (resistance index, 0.38) turbulent blood flow was prominent and surrounded the Cesarean scar gestational mass.
The residual ectopic Cesarean scar persisted in-situ for several months after conservative treatment even though β-hCG had declined to a normal limit (Figure 4). The mass regressed from 2 months to as long as 1 year after the treatment.
Ultrasound image showing a persistent Cesarean scar mass (Case 3) with a diameter of 1.79 × 2.13 cm found during follow-up 3 months after treatment.
The uterus was preserved in 11 of the 12 patients, and the patient in Case 8 had a subsequent intrauterine pregnancy 3 months later, with normal implantation of the gestational sac. She was followed up regularly during the pregnancy. She refused an early Cesarean section at 36 weeks because the fetus would have had a low birth weight. Unfortunately, this patient had a uterine rupture at 38 + 3 weeks of gestation. Despite resuscitation she died due to hypovolemic shock secondary to internal bleeding and the fetus was stillborn.
Discussion
Cesarean scar pregnancy is a rare form of ectopic pregnancy, and its incidence in our institution may have been increased because of the increased rate of Cesarean sections performed. The incidence of Cesarean scar pregnancy was 1 : 2216 (12/26 596) in this study and its rate was 6.1% (12/198) for women with an ectopic pregnancy and at least one previous Cesarean section; the incidence of Cesarean scar pregnancy in women with a previous Cesarean section was 0.15% (12/7980). There are two reasons why the incidence of Cesarean scar pregnancy is high in our hospital. First, our hospital is a tertiary referral medical center in a local district to where many women who have abnormal or ectopic pregnancies are referred for further diagnosis and treatment. Second, the high frequency with which transvaginal ultrasound is used early in pregnancy leads to early diagnosis of abnormal pregnancies. If we exclude two patients referred from another hospital, the incidence of Cesarean scar pregnancy was 1 : 2656 (10/26 596) and its rate was 5.05% (10/198) for women with an ectopic pregnancy and at least one previous Cesarean section; the incidence of Cesarean scar pregnancy in women with a previous Cesarean section was 0.13% (10/7980).
The exact cause of Cesarean scar pregnancy is still unknown. Nonetheless, its occurrence may be linked to an existing scar defect or microscopic dehiscent tract generated between the prior Cesarean scar and the endometrial canal4. Such tracts or wedge defects, likely resulting from previous Cesarean sections, can be detected with transvaginal sonography even several years following the Cesarean section16 and their depth or size can be measured with saline infusion sonohysterography17. The conceptus enters and implants in the myometrium through this ‘ectopic tract’. We are not sure whether the risk of Cesarean scar pregnancy is related to the number of previous Cesarean sections. Nine of the patients had only one prior Cesarean section and the other three (25%) had two prior Cesarean sections, and two of the women had a history of intrauterine device (IUD) use, whereas three others had a history of previous pelvic inflammatory disease. Nonetheless, we did not find a close relationship between these factors and the risk of Cesarean scar pregnancy due to the small number of patients in this study. Further study is required to assess the risk of IUD use and pelvic inflammatory disease in Cesarean section ectopic pregnancy.
Clinical history can aid in differentiating Cesarean scar pregnancy from a failed chorionic sac and non-live embryo passing through the lower uterine segment and the cervix. All our patients complained of vaginal bleeding but only half of the patients had accompanying mild to moderate abdominal pain at the time of diagnosis. Cesarean scar pregnancy differs from a spontaneous or inevitable miscarriage in which the volume of bleeding is often greater due to detachment of a failed chorionic sac in the lower uterine segment and the cervix. Moreover, most patients with inevitable abortion will have cramp-like lower abdominal pain and some have cervical motion or adnexal tenderness18.
Sonography is the first-line diagnostic tool for Cesarean scar pregnancy. Transvaginal ultrasound equipped with color Doppler imaging may serve as an additional tool to augment the diagnostic capabilities of transvaginal ultrasound; high-velocity, prominent, low-impedance blood flow can be detected surrounding an ectopic gestational sac, consistent with normal early pregnancy19. With pulsed Doppler functions, more physiological information such as flow pattern and resistance and pulsatility indices can be obtained on the peritrophoblastic vasculature. This information can help to confirm or exclude the diagnosis of an ectopic pregnancy.
The early diagnosis of Cesarean scar pregnancy led to a high success rate of conservative treatment with local methotrexate administration alone6, 7, 20, 21. Direct local injection of methotrexate into the amniotic cavity of a Cesarean scar pregnancy using either transabdominal or transvaginal ultrasound-guided injection showed excellent outcomes4, 13. In our experience, the transvaginal approach is preferable for treating Cesarean scar ectopic pregnancy because it allows for a shorter distance to the gestational sac, approaching from the vagina rather than through the abdomen. The transabdominal route was used only in the first four patients due to insufficient experience with transvaginal procedures. Moreover, the thin 22-gauge needle used in abdominal puncture made the process of aspiration of gestation tissue difficult. In contrast, the transvaginal approach with the thick double-lumen IVF needle made the process of aspiration easier. Local methotrexate administration increases the success rate due to the high concentration of methotrexate deposited in the lesion, avoiding the side effects produced with systemic administration22. The same treatment principles of local methotrexate injection that we used in our patients with Cesarean scar ectopic pregnancy have been used in other ectopic pregnancies, such as cornual or cervical pregnancies, with similar outcomes23, 24.
Side effects associated with methotrexate treatment, such as pneumonitis, alopecia, nausea or stomatitis, were not experienced by our patients because each patient received no more than two doses of methotrexate25. Systemic methotrexate injection seemed to be effective for patients with β-hCG levels less than 5000 IU/mL. Surgical or invasive techniques, including D&C, were not recommended for Cesarean scar pregnancy due to high morbidity and poor prognosis, except for cases of failure of conservative treatment1 or rupture of the uterus due to delayed diagnosis. The surgical methods were used only in Cases 1, 2 and 8, but the outcome was dismal. Although some authors26, 27 have advocated expectant management of selected ectopic pregnancies, we do not believe that this approach should be used for Cesarean scar pregnancies, which are prone to catastrophic rupture and high β-hCG levels (3217–58 400 mIU/mL in our study). Expectant management achieved a high success rate (60%) in ectopic pregnancies when the starting β-hCG level was < 2000 mIU/mL, whereas in 93.3% of patients with an initial β-hCG level > 2000 mIU/mL expectant management failed27. The ectopic mass grows progressively without spontaneous resolution if the ectopic scar pregnancy is left untreated or incompletely treated, whether or not there is cardiac activity3, 11, 13.
Serial transvaginal color Doppler ultrasound examination was useful in monitoring Cesarean scar pregnancies and appeared to correlate well with serum β-hCG levels. The high-velocity, low-impedance, turbulent flow remained prominent without much change during the course of the follow-up, until the β-hCG level returned to normal. Those patients with such flow characteristics should be advised of the risk of uterine rupture with internal bleeding1, 28 due to the high-velocity flow, even if β-hCG progressively declines during follow-up28. Furthermore, the high peak systolic velocity of the patient in Case 8 should have been a clear warning to the physician not to perform D&C to terminate the ectopic pregnancy due to the danger of profuse bleeding from such an area of high-flow velocity.
Spontaneous uterine rupture has been reported in patients with prior Cesarean section29. Women with prior Cesarean section were 17 times more likely to experience uterine rupture than were those with an unscarred uterus30. The risk of uterine rupture may be higher after the implantation of a pregnancy in the thin uterine scar5. It is also unpredictable as to which kind of uterine scar is prone to complications in such a situation and when they might occur. Uterine rupture still unfortunately occurred in Case 8 despite the fact that weekly ultrasound examination of the uterine scar thickness was performed on the patient during the third trimester, and the thickness of the uterine scar remained 1.5 cm in diameter 1 week before the occurrence of the rupture. Uterine scar thickness is not the only risk factor for uterine rupture in women with prior Cesarean sections5, 31. Fertile patients in whom Cesarean scar pregnancy has occurred should be advised of the risk of uterine rupture when there is a sudden onset of lower abdominal pain during the course of pregnancy, especially in the mid-third trimester. Some authors have recommended surgical repair of the scar before the next conception following methotrexate treatment5, 12. However, reports of pregnancy outcome after such an occurrence are rare. Furthermore, repair of the uterine laparotomy scar may not prevent the occurrence of uterine rupture in the subsequent pregnancy32. In fact, in the absence of further reports on similar complications, we do not advise laparotomy for suturing the scar or arrange an emergency hysterectomy to terminate an ectopic pregnancy in favor of conservative treatment. However, a subsequent pregnancy should be avoided for more than 3 months and probably 1 or 2 years. For the pregnancy following a Cesarean scar pregnancy, we prefer an early Cesarean section be done as soon as the fetal lungs become mature, to avoid the possibility of spontaneous uterine rupture.
We cannot predict when the Cesarean scar mass completely resolves after conservative treatment; we found that some masses persisted for several months. One possible explanation for this is that the scanty venous flow within the fibrous scar tissue made the reabsorption of residual gestational tissue difficult. A second mechanism might be related to the proliferation of collagen fiber or fibrous tissue in the isthmic portion of the uterus in response to myometrial injury induced by placental villi invasion.
In summary, Cesarean scar pregnancy is a very unusual and possibly life-threatening complication of pregnancy. We hope that we have demonstrated that transvaginal color flow Doppler ultrasound can contribute to a clear, early diagnosis and effective non-surgical management and follow-up of Cesarean scar pregnancy. This helps to preserve fertility and avoid catastrophic complications because of delayed treatment. Uterine rupture is a possible but rare complication in the subsequent pregnancy. Early Cesarean section is recommended as soon as maturity of the fetal lungs is achieved.
