Risk factors

Importantly, only about 50% of women diagnosed with an EP have identifiable risk factors. Recognition of these risk factors can assist the clinician not only in the early diagnosis of EP, but also in reducing the risk of massive intra‐abdominal haemorrhage and its morbidity and mortality [11]. The incidence of EP increases among women who have a history of EP [12–14]. A woman who has had two prior EPs has a 10‐fold increased risk of future EP. This increased risk could be attributable to underlying tubal dysfunction or secondary to the treatment of EP: after surgical or non‐surgical management the EP recurrence rate is 8–15% and after conservative management 15% [13,14]. The risk of EP also increases in women who have a history of any type of pelvic surgery. For example, previous appendectomy has a twofold increased risk of EP [15,16]. Among the group of women in whom tubal sterilization has failed, pregnancy can result in an EP rate as high as 33% [13].

The incidence of EP increased by more than twofold during the period 1970–1985, from 7 to 16 per 1000, and then declined by 30% during the years 1985–1997. This was explained by the increase and decline of pelvic inflammatory disease (PID) within those periods [17]. It has also been shown that having multiple sexual partners is a strong risk factor for EP, with an odds ratio of 2.1 [18,19]. There is an association between previous exposure to Chlamydia trachomatis and subsequent EP. In a recent study, previous exposure to Chlamydia trachomatis, as indicated by serum antibodies, doubled the risk of EP and was highest among women 35 years or older [20,21]. Overall, a history of genital infections including sexually transmitted infection, PID and/or any tubal pathology or surgery is a significant risk factor for tubal EP.

Smoking is a major risk factor for EP [22]. The risk of EP increases by threefold to fourfold in women who smoke more than one packet of cigarettes per day. The level of risk has been related to the number of cigarettes smoked per day. After stratification by the amount of daily smoking during the peri‐conception period, the odds ratio rises from 1.6 for women who have smoked one to five cigarettes, to 1.7 for women who have smoked six to ten cigarettes, to 2.3 for women who have smoked 11–20 cigarettes, and to 3.5 for women who have smoked more than 20 cigarettes per day [23].

Historically, EP rates following assisted reproductive technologies are known to have increased over time. The rate of EP is 2–3% higher in women undergoing in vitro fertilization (IVF) compared with natural conception [24]. In addition, treatment with gonadotrophin and other drugs such as clomifene in IVF pregnancy increases the incidence of EP [23]. The rate of heterotopic pregnancy in the assisted reproductive population could be up to 1 in 100 to 1 in 45 [25,26]. More recently, however, a population‐based retrospective analysis was carried out on all pregnancies following assisted reproductive cycles carried out in the UK between 2000 and 2012, in the anonymized database of the Human Fertilisation and Embryology Authority (N = 161 967 treatment cycles). The authors concluded that EP rates in the UK following IVF/intracytoplasmic sperm injection have progressively decreased, and this appeared to be associated with a reduction in the incidence of tubal factor infertility and the increased use of both a lower number of embryos transferred and extended embryo culture [27].

Intrauterine contraceptive devices containing copper and the Mirena® intrauterine system (IUS) decrease the risk of an EP, but if pregnancy does occur with the device in situ, the risk of EP is higher [28–30]. Of the 0.5 per 100 Mirena IUS users who become pregnant in 5 years (cumulatively), half are EPs. Women aged 35–44 years have a three times higher risk of EP compared with younger women [31–33]. Diethylstilbestrol exposure in utero increases the relative risk of EP by 3.84 [34].

Types of ectopic pregnancy

Current evidence supports the hypothesis that tubal EP is caused by a combination of retention of the trophoblast/embryo within the fallopian tube due to impaired tubal transport of the embryo and alterations in the tubal environment allowing early implantation to occur [35]. Over 95% of EPs are tubal in origin, with 80% located in the ampullary portion of the fallopian tube [35]. Therefore approximately 5% of cases represent non‐tubal EPs.

Non‐tubal EPs can be located in the cervix, ovary, interstitial portion of the fallopian tube (i.e. that part of the fallopian tube which traverses the myometrium of a normally shaped uterus), caesarean scar or abdomen [7,8]. Note that cornual pregnancies are extremely rare (1 in 100 000 to 1 in 140 000 pregnancies) [36] and these occur in the non‐communicating horn of a unicornuate uterus [37]. The term ‘interstitial pregnancy’ should not be used interchangeably with ‘cornual pregnancy’ [38]. Caesarean scar pregnancies occur when there is implantation into a lower anterior uterine segment at the level of a previous caesarean scar [39]. This incidence is quoted to be between 1 in 1800 and 1 in 2226 of all pregnancies, with a rate of 0.15% in women with a previous caesarean scar and a rate of 6.1% of all EPs in women who have had at least one caesarean delivery [40–42].

Diagnosis

TVS is the single best diagnostic modality for evaluating women with suspected EP [43–45]. This is the case for both tubal and non‐tubal EPs. The presence of abdominal pain and/or vaginal bleeding in a premenopausal woman who is sexually active should prompt urinary pregnancy testing and in the event that this is positive should result in an assessment of the pelvis by an experienced operator using TVS. In a meta‐analysis of the use of emergency‐physician ultrasonography in the evaluation of patients at risk of EP, bedside ultrasonography provides excellent sensitivity and negative predictive value as a diagnostic test for EP [46]. Visualization of an intrauterine pregnancy by an emergency physician is generally sufficient to rule out EP [46].

In experienced hands, when there are no signs of intrauterine or extrauterine pregnancy on TVS, women should be classified as pregnancy of unknown location (PUL) [1,9,47,48]. Importantly, PUL is a descriptive term and should not be used interchangeably with EP [49,50]. The final outcome for a PUL includes a failed PUL, intrauterine pregnancy, EP or persistent PUL (PPUL), with EP and PPUL considered to be high‐risk PUL [45,51]. An EP seen at the first ultrasound scan appears to be more symptomatic and larger in diameter and volume compared with an EP which starts as a PUL [52].

Tubal ectopic pregnancy

In experienced hands, over 70% of EPs are seen on ultrasound on the first scan [52,53] and well over 90% before surgery [54]. The diagnosis of a tubal EP by ultrasonography is based on one of the following grey‐scale morphological appearances [54]:

1. an inhomogeneous mass or ‘blob’ sign adjacent to the ovary and moving separately to this (Fig. 43.1); or
   
2. a mass with a hyperechoic ring around the gestational sac or ‘bagel’ sign; or
   
3. a gestational sac with an embryonic pole with cardiac activity, i.e. a viable extrauterine pregnancy; or
   
4. a gestational sac with an embryonic pole without cardiac activity, i.e. a non‐viable extrauterine pregnancy.

These various morphological ultrasound forms of tubal EP are reported with the following frequencies: ‘blob’ sign, 57.9%; ‘bagel’ sign, 20.4%; or gestational sac with a measurable embryonic pole, with or without positive cardiac beat, 13.2% [54,55]. In a recent meta‐analysis assessing the accuracy of first‐trimester ultrasound in diagnosis of tubal EP, TVS was more useful for confirming a tubal pregnancy rather than excluding one [56].

Non‐tubal ectopic pregnancy

The differences between interstitial, angular and cornual pregnancies on two‐dimensional sonography are subtle. Three‐dimensional sonography has the advantage of providing coronal views of the uterus that cannot be obtained with conventional two‐dimensional sonography, resulting in distinctive differences that assist in differentiating between interstitial, angular and cornual pregnancies [57].

Interstitial pregnancy

The following historical diagnostic criteria have been proposed based on two‐dimensional sonography [58]:

1. an empty uterine cavity;
   
2. a chorionic sac separate and at least 1 cm from the lateral edge of the uterine cavity; and
   
3. myometrial thinning (a thin, <5 mm, myometrial layer surrounding the gestational sac).

However, these two‐dimensional sonographic criteria only detect 40% of interstitial pregnancies. Ackerman et al. [59] added the term ‘interstitial line’, an echogenic line that runs from the endometrial cavity to the cornual region, abutting the interstitial mass or gestational sac. This criterion in particular provides a better sensitivity (80%) and specificity (98%) compared with the previously described eccentric gestational sac location (sensitivity 40%, specificity 88%) [60] and myometrial thinning (sensitivity 40%, specificity 93%) [58].

Although there are limited data on the use of three‐dimensional sonography for the diagnosis of interstitial pregnancy, this modality could well be the likely successor to two‐dimensional evaluation of potential interstitial pregnancies [57] (Fig. 43.2).

Cornual pregnancy

Mavrelos et al. [37] proposed the following ultrasound diagnostic criteria for a cornual pregnancy:

1. there is a single interstitial portion of the fallopian tube in the main uterine cavity;
   
2. a gestational sac, mobile and separate from the uterus, is surrounded by the myometrium; and
   
3. a vascular pedicle joins the gestational sac to the unicornuate uterus.

Caesarean scar pregnancy

The following ultrasound criteria have been proposed for the diagnosis of caesarean scar pregnancy [61]:

1. empty uterine cavity and empty cervical canal;
   
2. presence of the gestation sac with or without a fetal pole with or without fetal cardiac activity (depending on gestational age) located anteriorly at the level of presumed site of the previous scar in the lower uterine segment;
   
3. Doppler examination showing evidence of functional trophoblastic/placental circulation; and
   
4. negative ‘sliding’ sign.

Godin et al. [62] further described an empty cervical canal and absence of healthy myometrium between bladder and gestational sac.