1.3.1 Torsion

Data on the incidence of ovarian torsion in pregnancy are conflicting, and rates in the literature vary from 0.1 to 15% [Reference Schmeler, Mayo-Smith and Peipert3, Reference Condous, Khalid, Okaro and Bourne4]. While an ovarian torsion may occur even in the absence of a cyst in women with elongated ovarian ligaments, the condition is more common if the ovary is enlarged and ‘heavy’, such as in the presence of large ovarian cysts especially those with solid components (e.g. dermoid cysts) or in ovarian hyperstimulation syndrome. Condous et al. estimated the overall risk of torsion at 0.1%, rising to 5–15% if an ovarian cyst coexists [Reference Condous, Khalid, Okaro and Bourne4]. Approximately 10–20% among all cases of ovarian torsion take place during pregnancy, and 60% of them occur between 10–17 weeks gestation, likely due to cephalad displacement of the ovaries [Reference Huang, Hong and Ding7, Reference Yen, Lin and Murk8]. Large ovarian cysts are also likely to undergo torsion during the early post-partum period; as the uterus involutes, more intra-abdominal space is available for a displaced ovary to undergo torsion.

The symptoms a pregnant woman with ovarian torsion will present with are similar to those of a non-pregnant woman: sudden-onset sharp abdominal pain, constant or intermittent in nature, radiating to the back, groin or flank, nausea, vomiting and occasionally anorexia. On examination, patients may display signs of peritonism, adnexal tenderness and cervical excitation [Reference Bottomley and Bourne5].

Serum markers such as leucocytosis have a limited role in the diagnostic evaluation of pregnant women with a suspected ovarian torsion. On the contrary, transvaginal ultrasonography, with a reported positive predictive value of 87.5% and specificity of 93.3%, is the most useful diagnostic tool [Reference Graif and Itzchak9]. The most common sonographic features observed in ovarian torsion include enlargement of the ovary when compared to the contralateral one, oedema of the ovarian parenchyma and peripherally displaced follicles, perhaps along with transudation of fluid into the displaced follicles. The presence of a displaced ovary (e.g. at the uterine-vesico fold or even at the contralateral side of the pelvis) as well as the ‘whirlpool sign’ on colour Doppler examination (i.e. twisted vascular pedicle) are also useful ultrasound features raising the suspicion of ovarian torsion. Finally, free fluid in the pouch of Douglas can also be noticed in patients with ovarian torsion (as well as in women with ovarian cyst rupture; in the former event, the fluid is reactive in origin and usually its volume is significantly reduced compared to cases of a ruptured cyst).

If ultrasound examination is inconclusive, then MRI can be utilised to investigate a possible ovarian torsion in pregnant women. The use of gadolinium contrast medium should be avoided. On the contrary, CT has no place in the diagnostic work-up of pregnant patients with a suspected ovarian torsion [Reference Senarath, Ades and Nanayakkara10].

It should be flagged that ovarian torsion remains a clinical diagnosis; even though ultrasound imaging is routinely used to aid diagnosis, clinical examination is still the diagnostic cornerstone, and a strong consideration is made to proceed to surgical intervention should the clinical findings be highly suspicious of ovarian torsion.

1.3.2 Ovarian Cyst Rupture and Haemorrhage

Rupture of an ovarian cyst and subsequent haemorrhage, or haemorrhage within an intact ovarian cyst, are quite common cyst accidents that may occur both in pregnant and non-pregnant women. Patients usually present with non-specific symptoms such as sudden-onset lower pelvic pain, nausea and vomiting; of note, pain tends to be worse at the onset of symptoms and may subside by the time of presentation. In the vast majority of cases, the associated haemorrhage is not clinically significant; patients are usually haemodynamically stable; symptoms are manageable with analgesia and tend to resolve within a few days. In the unlikely event of significant intra-abdominal bleeding though, women may present with hypovolemic shock and necessitate an acute surgical intervention.

Similar to ovarian torsion, ultrasound is the first-line imaging modality in suspected ovarian cyst rupture. An experienced sonologist can usually easily identify the presence of haemoperitoneum in the pelvis or even in the upper abdomen and hence, can assess the severity of intra-abdominal bleeding. Furthermore, haemorrhage within an intact ovarian cyst has distinct ultrasound features: a fresh blood clot may display a lace-like or spider-web appearance; it is avascular on Doppler examination and can be seen to ‘wobble’ in a jelly-like fashion on palpation with a transvaginal probe [Reference Bottomley and Bourne5]. The ultrasound appearance of blood may be anechoic in early stages, whereas in later stages accumulated old blood may have a ‘ground-glass’ appearance as in an endometrioma.

1.3.3 Mass Effect: Obstruction of Labour

Like any large pelvic mass, should an ovarian cyst enlarge significantly, it may apply pressure to surrounding structures, including the urinary bladder, ureters, urethra and intestines. The associated symptoms vary and will be determined by which organ is affected and to what degree. The same mechanism is responsible for the obstruction of labour: if a large ovarian cyst is located adjacent to the lower uterine segment and below the presenting part, it may contribute to labour dystocia [Reference Senarath, Ades and Nanayakkara10].

1.4.1 Imaging

Ultrasonography is considered the first-line imaging modality in pregnant women with an ovarian cyst [Reference Eskander, Berman and Keder11]. It is safe both for the mother and the fetus, widely available and cheaper compared to alternative options; hence, it is an ideal primary evaluation tool.

Transvaginal ultrasound can be used as early as the fourth to fifth week of gestation to confirm the implantation site and exclude an extrauterine or heterotopic pregnancy. Assessment of the ovaries and the Fallopian tubes is becoming standard practice during an early pregnancy scan, allowing early identification of adnexal cysts, even if they are quite small and asymptomatic.

Accurate characterisation of the nature of ovarian cysts identified in pregnancy is crucial. Subjective pattern recognition (i.e. subjective impression of the sonologist) has a very good diagnostic performance for the assessment of adnexal cysts with a sensitivity and specificity equal to those of complex logistic regression models [Reference Sokalska, Timmerman and Testa12, Reference Kaloo, Louden, Khazali and Hoy13]. However, this requires an expert ultrasound operator, and it remains questionable whether it can be safely applied to a non-expert setting [Reference Yazbek, Raju and Ben-Nagi14]. As a result, many algorithms and models have been developed in an attempt to improve the diagnostic accuracy of non-expert examiners. Some of these models, such as the Risk of Malignancy Index (RMI), take into account other factors besides the ultrasound characteristics (e.g. cancer antigen-125 (CA-125)), some of which alter during pregnancy and hence, cannot be applied to pregnant women [Reference Kaloo, Louden, Khazali and Hoy13].

The International Ovarian Tumor Analysis (IOTA) group has published extensively on ultrasound-based models to differentiate between malignant and benign ovarian cysts. The group has proposed ‘simple descriptors’ and ‘simple rules’. These help non-expert operators classify most ovarian lesions as benign or malignant with a reported sensitivity of 95% and specificity of 91%, leaving only a small proportion of indeterminate cases to be assessed by expert sonologists [Reference Timmerman, Testa and Bourne15, Reference Ameye, Timmerman and Valentin16]. Table 1.2 presents the IOTA ‘simple rules’ and ‘simple descriptors’ for differentiation between malignant and benign adnexal lesions [Reference Timmerman, Testa and Bourne15].

The pregnant status of patients should always be kept in mind during an ultrasound evaluation of an ovarian cyst. Even benign cysts such as endometriomas may undergo major morphological changes during pregnancy, referred to as decidualisation, and mimic malignancy. The effect of pregnancy-related high progesterone levels on the ectopic endometrial tissue of an endometrioma results in the formation of ectopic decidua. Sonographically, a decidualised endometrioma is characterised by a thick, irregular inner wall and prominent intraluminal papillary projections with increased blood flow on Doppler examination, similar to malignant ovarian tumours. As shown by Pateman et al. [Reference Pateman, Moro and Mavrelos17], about 12% of endometriomas will undergo decidualisation during pregnancy. In such cases, ultrasound examiners should look for other features to confirm their benign nature, such as the presence of extraovarian nodules of deep infiltrating endometriosis (e.g. in the uterosacral ligaments, the rectovaginal space or the rectosigmoid colon), obliteration of the pouch of Douglas as well as the tendency of endometriomas to decrease in size and regress rapidly during pregnancy, contrary to borderline and malignant lesions [Reference Pateman, Moro and Mavrelos17].

A proportion of pregnant women with an ovarian cyst will require further imaging tests. In such cases, MRI is the test of choice. Contrary to ultrasonography, MRI is more operator-independent; moreover, it has excellent resolution for soft-tissue pathology, high sensitivity in identifying malignancy and negligible risks to the mother and fetus [Reference Adusumilli, Hussain and Caoili18]. The use of gadolinium contrast medium should be avoided due to its teratogenic effect; actually, it is usually not required for the assessment of ovarian pathology [Reference Ball, Waters and Cooper19]. Left lateral positioning should be considered to avoid caval compression by the gravid uterus in case of a prolonged stay in an MRI scan machine.

The use of CT for the assessment of ovarian cysts in pregnant women is not recommended; besides the well-established safety concerns about ionising radiation, CT actually performs poorly when compared to ultrasonography and therefore, it does not offer any clinical benefits [Reference Senarath, Ades and Nanayakkara10].

1.4.2 Tumour Markers

Contrary to non-pregnant women, the use of serum tumour markers in patients with ovarian cysts during pregnancy is limited. The physiology of pregnancy alters baseline levels of some of these markers; the interpretation of abnormal results should be made with caution and always in conjunction with the results of imaging tests.

CA-125 is raised in about 80% of women with epithelial ovarian cancer, and it is the most frequently used serum tumour marker even though its specificity is low [Reference Kaloo, Louden, Khazali and Hoy13]. Its baseline levels increase in pregnancy; they peak in the first trimester with the upper limit of the normal CA-125 range reaching 112 U/mL between 11 and 14 weeks gestation and then decreasing as gestation advances [Reference Aslam, Ong, Woelfer, Nicolaides and Jurkovic20]. Even though measuring CA-125 values per se will not discriminate between benign and malignant lesions in pregnancy, it may have some merits in suspicious or indeterminate ovarian cysts, as a significantly raised level may flag a potential malignancy and trigger further investigations. It may be measured as a reference point before and after treatment in women with known ovarian cancer, thus allowing physicians to monitor the response to treatment [Reference Alalade and Maraj6].

Other serum tumour markers used to monitor germ cell tumours include alpha-fetoprotein, human chorionic gonadotropin and lactate dehydrogenase. Unfortunately, the former two increase physiologically during pregnancy and therefore are of limited use. The latter is not affected by pregnancy, and its raised levels are a significant marker, commonly noted in dysgerminoma [Reference Alalade and Maraj6].

Finally, the glycoprotein HE4, which is a more sensitive and specific marker than CA-125 in differentiating between benign and malignant ovarian tumours, has lower baseline levels during pregnancy, hence limiting its potential use in pregnant women with ovarian lesions [Reference Alalade and Maraj6].

1.5.1 Expectant Management

This is a safe approach in asymptomatic women with benign looking ovarian cysts without any suspicious imaging features; it is also supported by a recent evidence-based guideline commissioned by the British Society for Gynaecological Endoscopy (BSGE) and endorsed by the Royal College of Obstetricians and Gynaecologists (RCOG) [Reference Ball, Waters and Cooper19]. The vast majority of cysts diagnosed in pregnancy are benign and actually, most of them are functional in nature. About 70% of them will resolve spontaneously by 16 weeks gestation and hence, no interventions are needed [Reference Condous, Khalid, Okaro and Bourne4].

It has been suggested that cysts with benign features measuring more than 6 cm in size or those with a complex, albeit reassuring, ultrasound appearance should be followed up after four to six weeks [Reference Alalade and Maraj6, Reference Senarath, Ades and Nanayakkara10]. This is a reasonable practice but nevertheless, an experienced sonologist with expertise in women’s imaging can, most of the time, offer reassurance about the scan findings and reduce the need for routine serial scans. Of course, certain types of benign ovarian cysts such as endometriomas should be followed up during pregnancy given the decidualisation-related diagnostic challenges.

When expectant management is favoured, women should be counselled about the risk of ovarian torsion and cyst rupture, and the plan should be reviewed if a patient becomes symptomatic and a cyst accident is suspected.

1.5.2 Ultrasound-Guided Fine Needle Aspiration

This option applies only to symptomatic women with simple benign ovarian cysts. It is a straight-forward outpatient procedure performed under local anaesthesia and it can provide immediate symptomatic relief, therefore reducing the need for admission and more invasive procedures. It is also included in the recent BSGE–RCOG guideline as an alternative option to surgery [Reference Ball, Waters and Cooper19]. However, some authors have expressed concerns about high recurrence rates (33–40%), as well as the potential risk of intraperitoneal spillage of cancerous cells in the case of an undetected malignancy [Reference Bottomley and Bourne5, Reference Guariglia, Conte, Are and Rosati21].

1.5.3 Surgical Management

Surgery is usually reserved for patients with an acute abdomen due to a cyst accident and women with imaging findings suspicious of malignancy. It should also be considered in asymptomatic patients with large ovarian cysts (usually more than 10 cm, although there is no consensus in the literature about a cut-off limit) to prevent complications such as torsion, rupture or obstruction of labour [Reference Alalade and Maraj6].

Historically, the surgical approach included a laparotomy, nevertheless advances in minimal access surgery have now established laparoscopy as a safe and feasible approach during pregnancy [Reference Ball, Waters and Cooper19]. A growing volume of literature evidence has demonstrated its superior outcomes in patients with benign disease and having addressed previous safety concerns, recent guidelines reflect the change in practice [Reference Ball, Waters and Cooper19]. Laparoscopic surgery offers the same advantages to pregnant as to non-pregnant women. High quality studies have confirmed that laparoscopy is associated with improved visualisation of pelvic organs, reduced blood loss, less pain, reduced length of hospitalisation, faster recovery and a lower risk of uterine irritability compared to laparotomy, without an increase in adverse obstetric outcomes (miscarriage, preterm delivery or fetal growth restriction) [Reference Chen, Ding and Hua22, Reference Liu, Zhang, Huang and Wang23]. Patients mobilise more quickly following a laparoscopy, which is important given the hypercoagulable pregnancy status.

Previous concerns with regard to the impact of pneumoperitoneum and carbon dioxide on uteroplacental flow have now been discarded [Reference Candiani, Maddalena and Barbieri24]. A study by Reedy et al. did not find any statistically significant differences in five fetal outcome variables (birth weight, gestational duration, growth restriction, infant survival and fetal malformations) in women with singleton pregnancies between 4–20 weeks of gestation undergoing laparoscopy versus laparotomy [Reference Reedy, Källén and Kuehl25]. Most literature data as well as the BSUG–RCOG guidelines recommend an intra-abdominal operating pressure of 12 mmHg to minimise the risk of alterations to the feto-maternal perfusion, although visualisation may become more challenging [Reference Ball, Waters and Cooper19, Reference Candiani, Maddalena and Barbieri24]. An operating pressure of 15 mmHg, however, has been used without adverse fetal or maternal outcomes [Reference Chen, Ding and Hua22, Reference Pearl, Price, Tonkin, Richardson and Stefanidis26].

Another potential concern is related to the risk of inadvertent damage to the pregnant uterus with the Veress needle or the primary port during entry into the abdominal cavity. No randomised controlled trials have compared the safety of various laparoscopic entry points and techniques during pregnancy, therefore operating surgeons may utilise alternative options (e.g. open Hasson technique or direct gasless entry with optical trocar, entry at Palmer’s point, supra-umbilical or sub-xiphoid point) according to their preference, fundal height and location of the ovarian cyst [Reference Ball, Waters and Cooper19].

Despite its advantages, laparoscopic surgery during pregnancy can be technically challenging especially after the first trimester due to the enlarged pregnant uterus, absence of intra-uterine manipulation, relatively low intra-abdominal pressures and reduced intra-abdominal space. Therefore, it is recommended that that such procedures should be performed by experienced laparoscopic surgeons with advanced skills [Reference Ball, Waters and Cooper19]. If malignancy is suspected pre- or intra-operatively, then a laparotomy should be considered to reduce the likelihood of cyst rupture and spillage of cancerous cells.