Mifepristone priming and subsequent misoprostol for second trimester medical abortion in women with previous caesarean delivery

To assess clinical outcomes and complications in women with ≥1 prior caesarean delivery (CS) during mid‐pregnancy medical abortion with misoprostol following mifepristone priming.


Caesarean birth is a central component of modern obstetrics.
There has been a progressive increase in caesarean section (CS) over the past three decades, rising in Australia from 19.5% in 1996 1 to 36% in 2019. 2 Globally, between 2000 and 2015 the average annual CS rate increased by 3.7% (from 12.1 to 21.1%). 3 While this increase may have led to improvements in obstetric and perinatal outcomes, it has also created many secondary issues such as placenta accreta spectrum disorders, caesarean scar pregnancy and challenges in pregnancy management following a prior CS. One such challenge is the management of second trimester abortion in women with at least one prior CS. While the sequential use of mifepristone and misoprostol has been demonstrated to be an effective method for medical abortion, [4][5][6] there are limited published data regarding its safety profile for women with prior CS.
Abortion in the second trimester of pregnancy accounts for the minority of abortion procedures, although reported to account for the majority of procedural complications. 4 Medical abortion does not require the specialist surgical skill of surgical abortion, makes assessment of the fetus for autopsy in cases of fetal malformation possible and may be less emotionally distressing for the healthcare providers. 4 The presence of prior uterine surgery, most typically CS, adds a layer of complexity and potential risk from uterine scar rupture in second trimester abortion. A systematic review of medical abortion with misoprostol between 16 and 28 weeks gestation in 461 women reported a uterine rupture rate of 0.43% (95%CI 0.08-1.67%) in those with one prior low transverse CS. 7 There were insufficient data on risk with >1 prior CS or prior classical CS.
It is now common practice to use a protocol of sequential mifepristone and misoprostol for medical abortion, as this technique has been demonstrated to result in improved procedural outcomes. 8,9 Whether the use of mifepristone prior to prostaglandin administration in second trimester abortion for women with prior CS alters procedural complication rates is unclear.
We conducted a retrospective review of all consecutive medical abortions at our hospital between 13 and 28 weeks gestation using sequential mifepristone and misoprostol over an 11-year period to assess the procedural outcomes and complication rates for women with and without prior CS.

MATERIALS AND METHODS
Women between 13 and 28 weeks gestation who underwent a medical abortion using sequential mifepristone and misoprostol for fetal anomalies or severe obstetric complications between January 2008 and December 2018 at King Edward Memorial Hospital, Perth, Western Australia, were identified from the institutional Fetal Medicine Service database. Gestational age was assigned based on certain menstrual dates with confirmatory ultrasound examination, or ultrasound dating if the menstrual dates were uncertain or varied significantly from the menstrual dates. All fetuses were alive at the commencement of the termination process, apart from those cases >22 weeks gestation where feticide was performed prior to the administration of misoprostol.

Mifepristone was introduced into our institutional medical
abortion regimen in 2008. The process of medical abortion is performed within a specific clinical pathway at King Edward Memorial Hospital. The regimen used in this series consisted of 200 mg of mifepristone administered orally as an outpatient, usually 24-48 h prior to the administration of misoprostol. The dosage regimen of misoprostol varied, principally depending upon gestation.
The route of administration was predominantly vaginal; however, a randomised clinical trial was completed during the early part of this study where the misoprostol was administered orally threehourly (100 women) or sublingually three-hourly (102 women). 10 For pregnancies <22 weeks gestation, women were administered a loading dose of 800 μg of misoprostol vaginally, followed by 400 Maternal demographic and clinical data were obtained from review of the medical record chart and entered into a computerised database. The primary outcome data for this study were the duration of abortion (defined as the time from commencement of misoprostol to fetal expulsion) and the procedural complications for women with at least one prior lower uterine segment CS, using the women with no prior CS as the reference group. We were specifically interested in the frequency of severe maternal complications, defined for the purposes of the study as maternal haemorrhage >1000 cc, requirement for blood transfusion, uterine rupture and hysterectomy.
Continuous data were summarised using medians and interquartile ranges (IQR). Categorical data were summarised using frequency distributions. Univariate outcome comparisons were conducted using Kruskal-Wallis and Mann-Whitney tests for continuous data, and Fisher exact or χ 2 tests for categorical data.
Procedural duration was examined using Kaplan-Meier survival probabilities and Cox proportional hazards regression where covariate adjustments for parity, number of prior CS, gestation at the procedure, first misoprostol dose and maternal demographic characteristics were also considered. Logistic regression analysis was used to assess procedural complications. Covariate effects were summarised using univariate and adjusted hazard ratios (HR, aHR) and odds ratios (OR, aOR) and their 95% confidence intervals (CIs).
The median gestation at abortion was 19 weeks (IQR 17-21), with the gestational age distribution demonstrated in Table 1.   (0.36%) delivered after the mifepristone alone, prior to any misoprostol (gestation range 14-18 weeks Table 3). This was also reflected in the greater total misoprostol dosage to achieve delivery in nulliparous women (1600 μg) compared with multiparas (1200 μg; Table 3). For women with a prior CS, the duration of termination was less than for those with no prior pregnancies (Table S1). Overall, increasing gestation was associated with an increased procedural duration (Table S1).    Uterine rupture occurred in four cases (Table S2)

DISCUSSION
In this single institutional series of 1399 consecutive second trimester medical abortions, the sequential use of mifepristone and misoprostol in general provided an efficient and safe method of termination.
For women without a prior CS, the median duration from first administration of misoprostol to fetal expulsion was 8.5 h, impacted significantly by parity and gestation. The incidence of severe complications such as post-partum haemorrhage >1000 cc (2.3%) and blood transfusion (1.2%) was consistent with other published data. 13 However, as these data demonstrate, significant morbidity may occur, and all units who provide a medical termination of pregnancy service must have the facilities to deal with these adverse events in a timely manner.
Women with at least one prior CS, had an increased complication rate, predominantly due to uterine rupture and its associated  0.28% 14 and 0.43%, 7 when using misoprostol for second trimester abortion in the presence of at least one CS. In the systematic review by Goyal, 14 a sub-analysis of with prior CS with mifepristone priming provided a uterine rupture rate of 1.15%. An additional systematic review assessed several forms of prostaglandins and mechanical methods, again without mifepristone priming, with a uterine rupture rate of 0.8%. 15 We report a uterine rupture rate of 1.3% for women with at least one prior CS (4/304) and 0.28% overall (4/1399). This rupture rate was higher than reported in the systematic reviews, although given the low denominator in all series, one or two cases may have a substantial impact on rates.
The first two uterine ruptures occurred early in our experience with mifepristone/misoprostol regimens and led us to reduce the dose of misoprostol, as this medication may be the key component in uterine rupture given its uterotonic effects. Mifepristone blocks progesterone receptors, producing cervical softening and increased uterine sensitivity to prostaglandins. 4 We hypothesise that a lower misoprostol dose should be used in women with a prior CS following mifepristone priming, and that this dosage modification may reduce uterine rupture; however, this is by no means certain and further research in this area is required.
Similarly, others have reduced the misoprostol dose for women with a prior CS, in regimens with and without mifepristone priming, to potentially reduce the occurrence of uterine rupture. 16 least one prior CS. 18 Mifepristone was used in only a quarter of cases and a variety of regimens, making comparisons with our data challenging; however, they reported a uterine rupture rate of 0.78% and an increased complication rate when oxytocin was used (OR 3.44; CI 1.12-10.52, P = 0.03). 18 Recently, Latta et al. 19 reported a study admixed with abortion and fetal demise cases, of 109 women with prior CS, 32 who had >1 CS. The uterine rupture rate was 6.25% with >1 prior CS, although only 26.5% of women received mifepristone priming.
Second trimester surgical abortion (D&E) techniques also have a recognised complication rate which is increased in women with prior CS. 20 In a series of 2973 surgical abortions, a major complication rate of 1.3% was reported. 20 The most frequent complications were cervical laceration (3.3%), atony (2.6%), and haemorrhage (1.0%). The authors reported an OR of 7.4, 95% CI 3.4-15.8 for a major complication with ≥2 prior CS. Prior CS was the strongest independent risk factor for a major complication.
Techniques for second trimester abortion continue to be refined to improve efficacy, safety and access for women. The presence of a prior CS increases the procedural risks regardless of the approach to abortion, and key messages are to ensure adequate training for medical staff in whichever technique they are using, early recognition of complications should they occur, and comprehensive pre-procedural counselling for women.

FUNDING INFORMATION
No funding was provided for this study.