Discussion
Ectopic pregnancy can be a life-threatening condition and still remains a cause of up to 4.9 % of all maternal deaths in developed countries with almost 80% of all maternal deaths occurring during the first trimester of pregnancy 6,7,8.
Ectopic pregnancies pose an increased diagnostic challenge and to an even greater extent. It is possible that due to the low index of suspicion in spontaneous HP with the false reassurance of the presence of an intrauterine pregnancy, the diagnosis can be delayed and a good number of women will thus present with serious clinical presentations as tubal rupture, acute abdomen, shock and hemoperitoneum5.
In HP, the clinical picture may not be significant in asymptomatic patients, and seeing an intrauterine pregnancy by ultrasound may lower the vigilance of looking into the adnexa for another gestational sac or ‘thinking ectopic’. A presentation of vaginal spotting or lower abdominal or adnexal pain and tenderness may be attributed erroneously to a threatened miscarriage especially with a low index of suspicion after seeing an intrauterine pregnancy 9.
In a review of the world literature on 589 combined intrauterine and extrauterine pregnancies, a combination of signs and symptoms including abdominal pain, peritoneal irritation and enlarged uterus were the most significant findings, with the pelvic inflammatory disease a significant risk factor 4. The early diagnosis of an ectopic pregnancy is possible due to a combination of ultrasound and serum measurements β-hCG. A doubling time of serum β-hCG of 66% was initially used in the early 80s 10. Following that a doubling time of 53% 11 and more recently 35% or more, over a 2 day period, was suggested 12. A doubling time of more than 35% over a period of 2 days suggests the diagnosis of a viable intrauterine pregnancy, thus excluding an ectopic pregnancy in many (not all) cases. Another concept in the early diagnosis of ectopic pregnancies is the discriminatory zone. Seeing an intrauterine pregnancy by transvaginal ultrasound should be possible with a serum β-hCG level of 1500-2000 iu/ml 13, or more recently a conservative level 3500 iu/ml was suggested 14. Above that level, nearly all viable intrauterine pregnancies should be visualized using transvaginal ultrasound. Failing to see an intrauterine pregnancy at such a level would raise the probability of an ectopic pregnancy.
Unfortunately, in HP pregnancies, both concepts; the doubling time and discriminatory zones, commonly used in early diagnosis of ectopic pregnancy, are unlikely to be helpful, thus posing an increased risk of misdiagnosis, with a third to half of HP cases thus presenting late and already ruptured before a diagnosis was made 5,15. The level of serum β-hCG in HP represents the combined contribution of both the intrauterine (mainly) and extrauterine pregnancy. These confusing levels are unlikely to be of clinical use for the diagnosis of a HP.
Visualizing both an intrauterine and extrauterine gestational sac, yolk sac, fetal pole, or fetal heart activity is understandably diagnostic of a HP. Unfortunately visualizing both intrauterine and extrauterine fetal heart activity is rare 9.
To add to the diagnostic challenge, an HP can be misdiagnosed as a corpus luteum cyst with the low vigilance of having already confirmed an intrauterine pregnancy. When seeing what could be a corpus luteum in the adnexa using greyscale ultrasound, a Doppler ultrasound can be very useful in such cases, and determining the location of blood flow and ‘ring of fire’ sign, can improve the sensitivity of diagnosing an ectopic pregnancy over a corpus luteum cyst 16. This is what we performed in our case and seeing a ‘ring of fire’ sign on Doppler ultrasound, raised our suspicion towards the presence of a HP.
Although ultrasound remains the main imaging modality in ectopic and HP pregnancies, a subset of patients may need further imaging to provide additional information. These patients are a minority who for example need a precise diagnosis where the required information could not be obtained by ultrasound. This further imaging can be undertaken using MRI, which can provide images with high soft-tissue contrast without the use of intravenous contrast agents.
In a review of 1737 patients exposed to first trimester MRI exposure, as compared with non-exposure, was not associated with increased risk of harm to the fetus or in early childhood of up to 4 years of age, this includes the risk of stillbirth or neonatal death within 28 days of birth and any congenital anomaly, neoplasm, and hearing or vision loss from birth to age 4 years. In contrast, Gadolinium MRI at any time during pregnancy was associated with an increased risk of a broad set of rheumatological, inflammatory, or infiltrative skin conditions and for stillbirth or neonatal death 17.
Ultrasound remains the imaging modality of choice in pregnancy. MRI – in selected patients where a diagnosis cannot be made by ultrasound – has another advantage due to its excellent soft-tissue contrast without the use of ionizing radiation. We found it useful to review the findings suggestive of ectopic pregnancy on MRI and we feel it is important we are familiar with them. Findings on MRI include tubal dilation and wall enhancement, tubal hematoma, adnexal hematoma, and a gestational sac-like structure18. In our patient, an MRI revealed an adnexal rounded mass lesion with a thick wall showing a high T2 signal.
It should be emphasized that despite the evidence of reported safety of MRI, and the ability to gain valuable information without the resort to intravenous contrast agents as Gadolinium, a conservative approach should still be adopted, and MRI should only be used in very selected cases and ultrasound remains the main imaging modality in pregnant women with suspected HP or ectopic pregnancy.
Any treatment for HP should aim to target the ectopic pregnancy, selectively, without harmful effects on the ongoing intrauterine pregnancy.
With this in mind, systemic methotrexate is contraindicated with a viable intrauterine pregnancy 19 and local treatment modalities have thus been suggested to avoid the use of systemic agents in HP, as local injection of potassium chloride 20 or hyperosmolar glucose in the ectopic pregnancy in the tube. Local injection of 50% glucose after aspiration of the tubal gestational sac fluid under transvaginal ultrasonographic guidance seemed effective in resolving the ectopic pregnancy without adversely affecting the concurrent intrauterine pregnancy 21.
Although local injections of these agents avoid surgery – at least initially – the risk of failure of such treatments and subsequent surgery and salpingectomy is high, reaching 55%, making them not an attractive or advisable modality in the context of HP with the other pregnancy in the tube 22. These modalities, however, may have a place in ectopic pregnancies with no concomitant intrauterine pregnancy, scar pregnancies, or in HP where the extrauterine sac is in an unusual location for example cervical or cornual23.
Realistic and practical approaches in HP with one of the pregnancies in the tube are performing a laparoscopy (preferred option) or laparotomy (depending on the clinical condition and expertise) and undertaking a salpingectomy (usually if the other tube is normal) or salpingotomy19. Another advantage of the surgical approach is that laparoscopy (or laparotomy) can confirm the diagnosis in addition to providing a definitive treatment.
We do however feel that although salphingotomy for a ‘sole’ ectopic pregnancy (without concurrent intrauterine pregnancy) is an established modality in selected cases especially with a damaged contralateral tube, a salphingotomy in cases of HP is not similar for two reasons.
Firstly, there is a risk of around 21% of a repeat operation via salpingectomy due to persistent tubal bleeding 24 and this risk should not be taken lightly in a woman with HP and ongoing remaining intrauterine pregnancy already subjected to a salphingotomy.
A second reason is that, as opposed to a radical surgery as a salpingectomy, a salpingotomy carries the additional risk of persistent trophoblasts of around 7% 25 which cannot be followed up by β-hCG due to the concurrent intrauterine pregnancy nor treated with systemic methotrexate for the same reason.
We thus feel it is thus appropriate to perform a salpingectomy rather than a salpingotomy in most HP cases when dealing with the tubal element of pregnancy as it minimizes the risks at, and after the procedure, that is inherently associated with salpingotomy.