Association between Progesterone-Only Emergency Contraceptive Pill and Ectopic Pregnancy

Introduction

Ectopic pregnancy (EP) occurs when the blastocyst implants outside the endometrium of the uterus. More than 95% of ectopic pregnancies occur in the fallopian tube, while the remaining 5 % include cervical, ovarian, and primary abdominal pregnancies [1]. Tubal pregnancies occur in different locations in the tube, with the ampulla region being the commonest site, at 55%, isthmus at 25%, fimbria at 17%, and interstitial at 2% [1]. EP is the leading cause of first-trimester pregnancy-related maternal mortality in the developed world accounting for 4-10% of all maternal deaths in the United States of America [2]. Further, EP complicates 1-2% of all pregnancies [2]. In the developing world and Kenya in particular, unsafe abortion due to unwanted pregnancy is the leading cause of first-trimester maternal mortality [3].

In Kenya, at the Kenyatta National Hospital (KNH) in the year 2002, 5 patients were admitted every week because of EP [4]. EP also has significant implications on future fertility with only a third of patients reported to conceiving after salpingectomy following tubal pregnancy [5, 6]. A study at KNH reported that 5.9% of the patients had suffered repeat ectopic pregnancy [4]. For example, Miyoro et al. demonstrated that 56.6 % of patients treated for EP between 1991- 2000 had features suggestive of pelvic inflammatory disease (PID) [4].

Previous tubal damage from surgery and adhesions secondary to abdominal/pelvic surgery are known risk factors for ectopic pregnancy [1, 7]. Alteration in tubal motility due to certain contraceptive methods predisposes one to EP [7]. The use of progesterone-only contraceptive pills has a slightly increased risk of ectopic pregnancy, with reports of up to 4-6% EP rates [1, 7]. A systematic review published by the American College of Obstetricians and Gynaecologists (ACOG) however did not demonstrate an increased risk among women who had used progesterone- only emergency contraceptive pills reporting an ectopic pregnancy rate of 0.6%- 1.1%, which falls within the reported ectopic pregnancy rates of 0.8%- 2% [8, 9, 10, 11].

Emergency contraceptive (EC), also known as postcoital contraception prevents pregnancy after an unprotected sexual intercourse. Women seeking EC are typically younger than 25 years, have never been pregnant, and have used a form of contraceptive in the past [12]. The methods of emergency contraception available include intrauterine Contraceptive Devices (IUCD), combination Estrogen-Progesterone pills, progesterone-only emergency contraceptive (POEC) pills, and antiprogestins-progesterone receptor antagonists, and selective progesterone receptor modulators. The combination estrogen- progesterone pill also referred to as the YUZPE method consists of two doses of 100mcg ethinyl estradiol and 500mcg of levonorgesterol taken at 12-hour intervals. The POEC pill, marketed as “PLAN B” in the United States of America consists of 750mcg of levonorgesterol taken 12 hours apart. However, the POEC pill taken at 1.5mg at once is as effective as the interval dose and tends to promote adherence [13]. Moreover, LNG-EC reduces the risk of pregnancy following a single act of coitus in mid- cycle from 8% to 1.1 % [8].

Several case reports have been published of women who took the POEC with seemingly no other risk factors and ended up with an ectopic pregnancy [14]. They all recommended that an EP should be considered if menses are delayed after the use of POEC. In addition, to maximize efficacy, EC pills should be started as soon as possible following sexual intercourse. Earlier studies demonstrated that it is effective if started within 72 hours after intercourse [15, 16]. Studies that are more recent have however shown that the pill is effectively taken up to 5 days (120 hours) after intercourse [17, 18, 19]. The Kenya National guidelines for family planning providers has been modified allowing emergency contraception to used up to 120 hours post coitus [20].

In Kenya, there is no recent data on ectopic pregnancy, the last study having been conducted more than 10 years ago. Ectopic pregnancy remains one of the leading causes of morbidity and mortality from attendant complications. Future fertility following an ectopic pregnancy remains low and the risk of a repeat ectopic pregnancy is significant, therefore it is crucial to educate patients on the risk factors and signs of ectopic pregnancy for early intervention. Most of the documented cases of ectopic pregnancy with EC pills are based on case reports. Thus, the overall objective of this study was to determine the association between the use of the hormonal emergency contraceptive pill and ectopic pregnancy.

Results

We recruited 120 women with an index ectopic pregnancy. Each of the 120 cases of ectopic pregnancies were matched for age to control with intrauterine pregnancy (n=120). Adequate age matching was achieved with an average age of 26.9 years (SD 6.2) among cases compared to an average age of 26.7 (SD 6.2) among controls. The most frequent age groups were 25-29 years.

Cases Control OR(95% CI) p-value

<19 years 5(4.2) 5(4.2) 1 20-24 years 30(25.0) 30(25.2) 1.00(0.26-3.81) 1 25-29 years 43(35.8) 42(35.3) 1.02(0.28-3.80) 0.972 30-34 years 31(25.8) 31(26.1) 1.00(0.26-3.80) 1 >=35 years 11(9.2) 11(9.2) 1.00(0.22-4.46) 1

Table 1 is the risk of ectopic pregnancy was higher in single women. Single women were approximately four times more likely to have an ectopic compared to married women (OR= 4.14, 95% CI 2.06-8.32, p < 0.001).

Discussion

The objective of this study was to find out if the use of emergency contraceptive pills increases the risk of ectopic pregnancy. A significant association was noted from the study, with 27.5% of patients with EP reporting to have taken LNG-EC in the index menstrual cycle compared to 2.5% of the controls, which gave a twelve-fold increase in the risk of EP if LNG-EC fails. This increase though higher is in keeping with a recently published multicenter case-control study done in China between 2011 and 2013 which reported a three-fold increase in EP after the failure of EC [14]. Farkas, et al. in their study of the effect of oral ovulation inhibition contraceptives and postinor on steroid hormones reported an ectopic pregnancy rate of 6.4% following the failure of LNG-EC [21]. A systematic review by Cleland et al. however showed no increase in the risk of EP after the use of LNG-EC, reporting that the rate of ectopic pregnancy was 0.8%, which falls within the general population rates of 1-2% [2]. Most of the studies used in the systematic review however did not use EP as their endpoint. Two epidemiological studies reported higher rates of EP following the failure of LNG-EC. In Hong Kong in 2008, a failure rate of 1.8% with an ectopic pregnancy rate of 2.8% was reported [22], and Gainer et al. reported an ectopic pregnancy rate of 4.1% after the failure of LNG- EC [23]. There was no statistical significance between ectopic pregnancy and the level of education in this study. A recent study done in China however found that EP occurred more often in women with lower level education compared to those with tertiary education [14]. A study done in Italy did not show any association between education level and the occurrence of EP [24]. In this study, single women were four times more likely to have an EP compared to married women. This is most likely due to the possibility of high-risk sexual behavior and multiple sexual partners among single women. Miyoro found no statistical significance with marital status [4]. A study in China similarly did not show any statistical significance between marital status and the risk of EP [14].

There was a significant association between parity and EP, with women para 1-2 being at a higher risk of EP compared to nulliparous women. From this study, 92.5% of all cases had parity equal to or less than 2, with 34% being nulliparous. This is in keeping with other studies which have reported that most patients with EP tend to be of low parity. Miyoro reported that 77.2% of the cases had parity equal to or less than 2 [4]. Parazzini et al. reported similar association in Italy [24]. The majority (42.4%) of the LNG-EC users was aged 20-24 years, nulliparous (51.5%), and had college- level education (45.5%). This is in keeping with the ACOG practice bulletin report of May 2010 [17], which described EC users as being below 25 years and having never been pregnant. A study was done in Hong Kong between 2006 and 2008 to describe the characteristics of EC users [22], however reported a mean age of 30 years, with 65.5% being nulliparous.

Only 18.2% of the cases who took the pill knew their fertile days. Seventy two percent (72.2%) of the cases took the pill within 24 hours of unprotected coitus and all the 3 controls took the pill within 24-48% hours. The majority of the cases (78.8%) and all the controls swallowed both pills at once and most women (n=24 cases and n=3 controls) swallowed the pill in the secretory phase. From this study therefore, the majority of the women knew the correct timing and dosing of use of LNG-EC and used it correctly compared to a study done among Kenyatta University students by Nyawande, et al. in 2005, where despite significant knowledge of the fertile days (81%), only 52% knew how the correct timing [25]. This increase in knowledge of the timing could be attributed to a robust media campaign carried out by population services international to educate women of reproductive age on the correct use of the emergency pill over the last few years.

The most effective period of action of POEC has been described as the pre-ovulatory follicular phase with reports of delay of ovulation by up to 5 days by 100% [25, 26, 27], majority of the cases who took the pill (n=24) in this study and all the controls (n=3) took the pill in the secretory phase. Pulkinen and Talo [28, 29] described the myoelectrical activity of the fallopian tube and reported that progesterone inhibits tubal motility post-fertilization and could cause a delay in ovum transport resulting in EP.

Other risk factors for EP were also assessed. From this study, early age at sexual debut increased the risk of EP. The risk increased sevenfold in women who report their first sexual contact before their 15th birthday compared to those who had their first contact at 25 years and above. Sixty two percent (62.5%) of the cases had their sexual debut before their 20th birthday compared to 51.2% of the controls. A recent study in China showed a fourteen-fold increase in the risk of EP in patients who had their sexual debut before 18 years [14]. This is explained by the immaturity of the vaginal epithelium in younger people making them more susceptible to ascending infections with long-term sequelae.

Patients who conceived after a period of infertility/ subfertility were five times more likely to have an EP from this study. Twenty one percent (21%) of the cases reported a period of inability to conceive for at least 6 months after stopping the use of the contraceptive method versus 5% of the controls. Parazzini et al., in an Italian case-control study, reported a three times higher risk of EP among the cases [24]. The cause of infertility would most likely be implicated in the etiology of the EP.

There was no association between the previous use of IUCD and other long-term methods of family planning. This is in keeping with the findings of Miyoro et al. in 2002 in the same facility where he reported no association between IUCD and ectopic pregnancy [4]. Only one (0.8%) patient recruited during the study period conceived with an IUCD in situ and the pregnancy was an EP. The study done in China reported that previous use of IUCD was not significantly associated with EP. However, in the case of contraceptive failure with an IUCD in situ, the risk of EP was 21.08 higher than with no contraceptive method at the time of conception [14]. One patient (0.8%) conceived while on the mini pill during the entire study period. The resulting pregnancy was an EP. Progesterone-only pills have been reported to slightly increase the risk of EP up to 4-6% [1, 7], however from this study there was no significant association.

Sexually transmitted infections have a well-documented role in the causation of EP. The risk in this study increased three-fold with 25% of cases reporting a history of treatment for an STI compared to 10.1% of the control group. Miyoro, et al. in their study reported that 56.6% of the cases had features suggestive of PID [4]. Parazzini, et al. in an Italian study reported a six-fold increase in the risk of EP in patients who had been treated for an STI/PID [24], double what was reported in our study. Nathalie et al. in a review of several African studies concluded that PID resulting from STIs was the single most important risk factor for the development of EP in African developing countries [30]. The widespread use of antibiotics in the treatment of STIs allows some degree of patency to be maintained in the fallopian tubes resulting in EP.

This study did not show any association between EP and a history of previous abdominal pelvic surgery as reported in other studies. 20% of the cases and 27 % of the controls had a previous abdominopelvic surgery with the majority having undergone a caesarean section.

An Italian study reported an 8-fold increase in ectopic pregnancy among patients with an EP compared to the controls [24]. Our study however had a highly selected group in terms of previous caesarean section among the control. This is because KNH does not attend to low-risk antenatal mothers, who tend to receive antenatal care in smaller facilities within the county. Women with previous caesarean sections are usually referred for antenatal care and delivery to KNH which could explain the higher number of women with previous surgery in the control group compared to those in the cases.

There was no association between the previous history of tuboplasty and the risk of EP. However, only 3.3% of the cases and 1.7% of the controls had undergone tuboplasty making it difficult to draw any statistical association. This could be due to the low success rates of tuboplasty. Other studies have shown significant association in patients who had previous tuboplasty or reversal of tubal ligation who eventually conceived [24].

Acknowledgments

• We would like to acknowledge the following people for:
• Concept and design: xxx
• Collection and assembly of data: xxx
• Data analysis and interpretation: xxx
• Manuscript writing: xxx
• Final approval of manuscript: xxx

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Competing and conflicting Interests

The authors do not disclose any conflict of interest.