Ectopic low-lying implantation pregnancy: analysis of outcomes depending on gestation age

Introduction

Rate of non-tubal ectopic pregnancies (EP) ranges from 5.0% to 10% from all EPs [1-3] though their overall incidence has been increasing in recent years [4]. Furthermore, non-tubal EPs contribute disproportionately to maternal morbidity and mortality in comparison to tubal EPs [2, 3].

There are no standards for diagnostics and treatment of this pathology [5-9]. Management of non-tubal EP may involve expectancy, medical or surgical treatment, or their combination, minimally invasive radiological approaches depending on EP location and the patient's clinical stability [2, 3, 9].

Some of these non-tubal pregnancies are characterized by trophoblast invasion in the lower segment of the uterus including previous cesarean scars, cervico-isthmus or cervix. They are similar in incidence rarity but cause life-threatening bleeding.

The number of reported cases of Cesarean scar pregnancy (CSP) has increased in the last 10 years because of the increase in Cesarean sections number and the improvement in ultrasound detection in the first trimester [3, 10-14]. Cervico-isthmic implantation is also viewed as a variation of EP [9, 15-17]. Some authors associate this localization with CSP as pregnancy located below the internal os [15, 17]. Due to their relative rarity CSPs have proved difficult to study and there is still lack of agreement regarding the most appropriate diagnostic criteria and optimal strategies to treat this condition [9, 10, 12, 18-20]. Some specialists affiliate CSP, cervico-isthmic pregnancy (CIP) and cervical pregnancy into “pregnancy in an abnormal location” or “low-lying-implantation ectopic pregnancy” because the low segment is less capable of fibromuscular contraction to control bleeding [17,21]. However, this term is not quite appropriate, because livebirth is considered impossible for cervical pregnancy, but is being discussed for CSP and CIP [13, 22].

In recent years the number of reports of livebirth delivery has increased although this situation is always connected with high risk of bleeding and hysterectomy [2, 13, 17, 22-25]. Unfortunately, in clinical practice the majority of clinicians have very limited experience and confidence in the differential diagnosis of this pathology and treatment is at the physicians’ discretion [16]. Literature about non-tubal EP has been restricted to small case reports and series, with limited data on optimal treatment protocols [1].

The objective is to define the optimal term of treatment of ectopic low-lying implantation pregnancy (LIP), including CSP and CIP.

Material and methods

A single-center retrospective study of 997 consecutive patients with EP of gynecological department of the City Hospital #1 of Krasnodar (2011-2016 years) found CSP and CIP before 14 gestation weeks in 37 cases (3.7%). Taking into consideration the similar clinical characteristics and management we unified patients with CSP and CIP into a category of LIP. The diagnosis was based on ultrasound examination (US) data, performed in the hospital and human chorionic gonadotropin (β-HCG) level in blood.

Each patient with confirmed LIP gave the informed consent to treatment options, including open surgery, systemic chemotherapy, mini-invasive radiological approaches procedures (uterine artery embolization under X-ray control (UAE), dilatation and curettage/vacuum aspiration under US control (D@C/VA,) local methotrexate (MTX) administration under US control) and their combinations.

US protocols, blood loss, time of β-HCG level decrease, spotting duration and outcomes were evaluated. We considered the outcomes as successful if no further interventions were required after the procedure, and complicated if any additional interventions were performed. Patients were divided into two groups taking into account gestation age (GA) according to the last menstrual period (LMP) (1^st group – GA less 56 days, n=22; 2^nd group – GA 56 and more days, n=15).

The statistical analysis was performed using statistical software (SPSS, Version 18.0). Examination of the form of distribution of signs was carried out using Shapiro–Wilk test. Median and interquartile range were used for description of numerical characteristics of statistical distribution of signs – Me (LQ, UQ). Mean values of quantitative attributes were estimated using the nonparametric Mann-Whitney U test. To determine the statistical significance of differences of observed and expected frequencies of qualitative attributes the chi-square test (χ²) was used.

The critical level of significance was defined with a two-tailed test р<0.05. Comparison of risk of repeated invasive procedures (RIP) between the groups was performed using the calculation of odds ratio and relative risk factor (RR). To visualize the nature of relation between the variables a scattering diagram was applied.

Results

According to the retrospective data ultrasound protocols the sensitivity of routine pelvic ultrasound examination for LIP diagnostics amounted to 54% (prior to hospital admission the diagnosis was confirmed by US in 20 cases: 10 in the 1^st group and 10 in the 2^nd group). US criteria for CSP (LIP) are well-known: location of “empty” uterus and “empty” cervical canal; visualization of gestation sac/trophoblast in scar/isthmus; myometrial thinning of anterior wall of uterus (1-3 mm); gestation sac located in the “niche” surrounded by myometrium/fibrose tissue separated from cavum uterus; hypervascular pattern with trophoblastic blood flow into scar with positive test for β-HCG; negative “sliding-sign”; placental bulge, placental lacunae [13, 15, 16, 26].

Most of these criteria were absent or misinterpreted in medical reports. Eight patients (21.6%) were admitted with bleeding after D@C/VA performed in other hospitals while LIP was overlooked. All of them needed RIP for the purpose of haemostasis. 2 among 8 these patients had GA less than 56 days, and 6 patients – 56 days and more. In all cases RIP were performed: with less 56 days GA – in 1 case D@C under US control with systemic МТХ administration; in 1 case – D@C+UAE+MTX; with 56 days and more GA – in 2 cases D@C under US-control with systemic МТХ administration, in 4 cases – D@C+UAE+МТХ were carried out.

Analyses of all examined LIP cases discovered that D@C/VA as a self-sustained treatment method was efficient only in cases where GA was less than 8 weeks (up to 55 days from LMP) (table 1). For patients with LIP, whose GA overpassed 55 days from LMP, all attempts of D@C/VA resulted in repeated interventions (table 1). Out of 6 cases, which required RIP in the 1^st group: in one case UAE was carried out two months later in connection with profuse bleeding due to arterio-venous malformation, despite of decrease in β-HCG up to the subliminal level. As a whole, without considering the methods of treatment, the difference in the necessity of RIP between the groups was significant (χ²=3.96, p=0.046) (Table 1).

Analysis of treatment and outcomes based on GA demonstrates significant difference in blood loss and numbers of RIP between the groups with GA less than 8 weeks and more than that (Table 2).

Table 2. Outcomes of LIP in relation to GA

Scattering diagram was used for graphical visualization of relation between the observation period before decrease in β-HCG and blood loss. Points’ position demonstrates that the group with gestational age of 8 weeks and more has a wider field of distribution than the other group (Figure 1).

Figure 1. Relation between the observation period before decrease in β-HCG and blood loss

For the group with GA>55 days Relative Risk (RR) of RIP was 2.2 times higher, than in the group with GA<55(6/22 vs 9/15; RR: 2.2, 95% CI 0.991-4.885) (Table 3). Odd ratio calculation showed that the factor of GA>55 increases the risk of RIP four times (9/6 vs 6/16, χ²=3.96, p=0.046; OR=4.00, 95% CI 0.99-16.15) (Table 3).

Table 3. Risk evaluation of RIP depends on GA

CI, confidence interval.

Discussion

Estimated incidence of CSP is 1.5 per 10,000 maternities [19]. Despite rarity of LIP diagnostics, difficulties and morbidity present the actuality of this problem. Ectopic cesarean scar pregnancy needs to be differentiated from normal but low intrauterine pregnancy, abortion in progress, true cervical pregnancy [7, 12, 14, 16]. There are lots of publications focused on early diagnostic criteria of cesarean scar pregnancy because treatment during first trimester in most cases allowed to preserve fertility [9, 11, 14, 16, 18-20, 27]. Unrecognized or misdiagnosed CSPs treated with inadequate procedures may lead to severe and sometime uncontrollable vaginal bleeding and hysterectomy.

The clinical diagnosis without the visualization methods is extremely difficult. Unfortunately, our study shows insufficient level of knowledge about diagnostic criteria of LIP among ultrasound specialists. Without standard compulsive protocol, we have a high number of false-negative conclusions, which caused prolongation of diagnostic time and treatment with repeated invasive procedure. In cases when LIP is missed in first trimester, the probability of fertility loss increases multi-fold. Four patients with LIP prolonged for more than 14 weeks not included in analysis in our study had profuse bleeding (600-5,000 ml) and fetal demise in all cases. Hysterectomy was performed in two cases.