Using IOTA terminology to evaluate fetal ovarian cysts: analysis of 51 cysts over 10‐year period

To describe ultrasound features of fetal ovarian cysts as reported by the original ultrasound examiner, to apply International Ovarian Tumor Analysis (IOTA) terminology after retrospective analysis of the images and to describe patient management and evolution of fetal cysts during pregnancy and after delivery.


with a fetal ovarian cyst at the Prenatal Diagnosis Division of the Bambino Gesù Children's Hospital, in Rome, between March 2011 and May 2020. Cysts were classified by the original ultrasound examiner as 'simple' (unilocular anechoic cyst) or 'complex' (cyst with other morphology).
In addition, three ultrasound examiners, experienced in gynecologic ultrasound, classified retrospectively the fetal ovarian cysts according to IOTA terminology, by reviewing stored ultrasound images. The evolution of these fetal ovarian cysts during pregnancy and after birth was recorded.

INTRODUCTION
Ovarian cysts are the most common abdominal cysts observed in the female fetus, with an incidence of one in 2600 pregnancies with a female fetus 1 . They are almost always described as benign 2 , but they can show a variable 'natural' course: they may undergo spontaneous resolution (50% of cases) 1,3 , or can be complicated with torsion, hemorrhage or rupture 2,3 . Ovarian torsion is the most frequent complication (22%) 1 . The prognosis of fetal cysts depends on their size and morphology 1 . Such knowledge is essential to schedule prenatal follow-up and select fetuses eligible for intrauterine aspiration (IUA), and to arrange postnatal care planning 4,5 .
Fetal ovarian cysts are classified on the basis of their ultrasound features as 'simple' (anechoic with thin walls) or 'complex' (isoechoic or hyperechoic content, with fluid, septa or calcifications) 1,6 . No consensus has been reached regarding the prenatal management of patients with complex cysts and this differs according to the referral center. Most newborns affected by complex ovarian cysts undergo surgery to prevent complications 1 ( Figure 1). However, complex cysts may include cases that regress spontaneously in the uterus or in postnatal life and, in some cases, histology findings following oophorectomy reveal the presence of residual normal ovarian parenchyma 7,8 . This suggests that classification into 'simple' and 'complex' categories is imprecise for predicting the evolution of fetal ovarian cysts and guiding clinical management.
In the year 2000, the International Ovarian Tumor Analysis (IOTA) group proposed terminology to be used in adult women for the assessment of adnexal lesions 9 . This lexicon helps examiners to classify cysts in a standardized manner, with detailed description of cyst morphology and content.
The aim of this study was to describe ultrasound features of fetal ovarian cysts, as reported by the original ultrasound examiner in a single referral center, and to apply IOTA terminology after retrospective view of the ultrasound images. Additionally, we aimed to describe patient management and evolution of the fetal cysts during pregnancy and after delivery.

METHODS
This was a retrospective observational study. All pregnant women with a diagnosis of fetal ovarian cyst on ultrasound examination between March 2011 and May 2020 were enrolled at the Prenatal Diagnosis Division of the Bambino Gesù Children's Hospital in Rome (Medical and Surgical Department of the Fetus-Newborn-Infant). Both patients diagnosed at the center during a routine obstetric ultrasound examination and those referred due to an ultrasound diagnosis of fetal ovarian cyst were included in the study. Other inclusion criteria were: at least two ultrasound examinations during pregnancy performed for the assessment of fetal ovarian cysts; and availability of data from postnatal follow-up. In cases which did not undergo surgery, follow-up data were retrieved from reports of ultrasound examinations performed by radiologists. All prenatal ultrasound examinations were performed by one of two examiners (A.R., L.C.), using a 4-8-MHz convex probe (Voluson E8, GE Healthcare, Zipf, Austria).
Cysts were classified by the original ultrasound examiner as 'simple' (unilocular anechoic cyst) or 'complex' (cyst with other morphology). Clinical information (age, gestational age at diagnosis and maternal or fetal comorbidity), type of management and evolution of the cyst, including spontaneous resolution, change of ultrasound pattern, IUA and postnatal surgery, were also recorded. According to our institutional policy, management of patients with fetal ovarian cysts was based on an arbitrarily defined cut-off point: cases with a cyst classified as simple that was smaller than 47 mm in maximum diameter underwent ultrasound follow-up once a week until delivery or until the cyst disappeared in utero, in order to identify any change in size or morphology; in cases with a cyst classified as simple that was 47 mm or larger, IUA was offered and, if it was declined, weekly follow-up examinations were performed; and in cases with a cyst classified as complex, ultrasound examination was performed once a month antenatally and, if the cyst persisted after delivery, surgical treatment was offered, regardless of cyst size. Surgery was also offered in cases with a simple cyst at first ultrasound examination during pregnancy, in which complex morphology was then observed during prenatal follow-up. When surgery was not indicated or was declined, postnatal follow-up was planned monthly until resolution, and any complication was noted (Table S1).
Three ultrasound examiners, experienced in gynecologic ultrasound, (A.C.T, with more than 20 years' experience, and C.C. and F.M., each with more than 10 years' experience) classified retrospectively the fetal ovarian cysts according to IOTA terminology, by reviewing the stored ultrasound images. The outcome of the fetal ovarian cysts was analyzed according to the original morphological classification of simple or complex. An additional analysis was also performed for the simple-cyst subgroup according to their maximum diameter, using a 47-mm cut-off. Finally, cyst outcome was analyzed according to the terminology proposed by the IOTA group 9 .
All clinical and ultrasound data were entered into a dedicated Excel file (Microsoft Office Corp., Redmond, WA, USA). Results are presented as absolute frequency (percentage) for categorical variables and as median (range) for continuous variables.

RESULTS
A total of 48 women with singleton pregnancy with a diagnosis of fetal ovarian cyst on ultrasound examination were included in the study. Three fetuses had bilateral masses; therefore, 51 ovarian cysts were included. Clinical data of the study population are shown in Table 1. The median age of the pregnant women at diagnosis was 32 (range, 25-41) years and the median gestational age at diagnosis was 32 (range, 29-38) weeks. The median number of follow-up examinations during pregnancy was three (range, two to five).
The management and outcome of cases with a simple fetal ovarian cyst, according to maximum cyst diameter (< 47 mm vs ≥ 47 mm), and of cases with a complex fetal ovarian cyst, is shown in Figure 2. Among the 29 fetal ovarian cysts classified originally as simple, 22 (75.9%) remained unchanged in morphology during pregnancy, two (6.9%), both measuring ≥ 47 mm, resolved after IUA, and five (17.2%) evolved complex morphology during pregnancy. Of the 22 cysts which remained unchanged during pregnancy, 19 resolved spontaneously within 12 months after delivery (17 measuring < 47 mm and two measuring ≥ 47 mm), two, both measuring < 47 mm, resolved after postnatal aspiration (one performed 15 days and one 180 days after delivery) and one (measuring ≥ 47 mm) was removed surgically after delivery (oophorectomy with benign ovarian cyst and normal ovarian parenchyma at histology). Among the five cysts which evolved complex morphology during pregnancy, one resolved spontaneously after delivery and four were removed surgically (one cystectomy with benign ovarian cyst at histology and three oophorectomies with benign ovarian cyst and normal ovarian parenchyma at histology. In the two cases managed with IUA, no signs of bleeding were observed for 2 h after aspiration. For all 22 fetal ovarian cysts classified originally as complex, neonatal surgery was planned, although surgery was declined in two of these cases and these resolved spontaneously. Of the 22, seven in total (31.8%) resolved spontaneously after delivery (six cysts resolved within 4 months after delivery and one within 9 months) and 15 (68.2%) underwent a surgical procedure. In 12 (80.0%) of these 15, necrosis was observed at histology, and in four of these 12 cases, ovarian amputation was observed, while a benign epithelial cyst with normal ovarian parenchyma was observed in the other 3/15 (20.0%). Bowel involvement (adhesion), requiring lysis of adhesions, was diagnosed in 4/22 (18.2%) cases with a complex cyst.
Three of the pregnant patients in our series presented bilateral fetal ovarian cysts. One patient (Case 25, Table S1) presented bilateral fetal ovarian cysts described as simple at first ultrasound examination. The right cyst was < 47 mm and the left cyst was ≥ 47 mm. The patient declined IUA and prenatal follow-up was proposed. Both cysts remained unchanged during pregnancy and spontaneous resolution was observed after delivery. Another patient (Case 26, Table S1) presented bilateral fetal ovarian cysts described as simple and both measuring < 47 mm at first ultrasound examination. Prenatal follow-up scans showed that the right cyst evolved into a complex cyst, and oophorectomy was performed after delivery, whereas the left cyst remained unchanged in size and morphology and was aspirated after delivery. The third patient (Case 46, Table S1) presented bilateral fetal ovarian cysts described as simple on first ultrasound examination (Videoclip S2). Both cysts were < 47 mm and remained unchanged during pregnancy. Spontaneous resolution was observed after delivery.
On reviewing the ultrasound images and applying IOTA terminology, neither papillary projections nor other solid components were observed, and all 51 (100%) fetal cysts were described as unilocular. Twenty-nine (56.9%) cysts had anechoic content (described as simple cysts by the original ultrasound examiner), and 10 (19.6%) had low-level, one (2.0%) had ground-glass, nine (17.6%) had hemorrhagic, one (2.0%) had mixed and one (2.0%) had undefined content (all described as complex by the original ultrasound examiner) ( Figure 3 and Table 2).
Classifying the cysts according to IOTA terminology, spontaneous resolution was observed in 20/29 (69.0%) and resolution was observed after IUA in 4/29 (13.8%) cysts with anechoic cyst content, in 7/10 (70.0%) cysts with low-level content, and in none of the other cysts, described as having ground-glass, hemorrhagic, mixed or undefined cyst content ( Table 2). Necrosis at histology following surgery was documented in none of the five   cysts with anechoic content, in one of the three cysts with low-level content and in 11 of the 12 (91.7%) cysts with ground-glass, hemorrhagic, mixed or undefined content.

DISCUSSION
In this retrospective study including 51 ovarian cysts in 48 fetuses, we found that most (69.0%) fetal ovarian cysts classified by the original ultrasound examiner as 'simple' resolved spontaneously after delivery, whereas most (68.2%) fetal ovarian cysts classified as complex were removed surgically, according to the institutional protocol, and necrosis was observed in most (80.0%) of these. We also observed that almost all (70.0%) cysts described as complex by the original ultrasound examiner, and classified as having low-level content according to IOTA terminology, resolved spontaneously. In contrast, no spontaneous resolution was observed in cysts with ground-glass, hemorrhagic, mixed or other cyst content, and there was necrosis at histology in almost all (91.7%) of them.
To the best of our knowledge, this is the first study applying IOTA terminology to fetal ovarian cysts, and analyzing patient management and outcome of the different categories defined according to cyst content.
The retrospective nature of this study and the relatively small number of cases are the main limitations, preventing conclusions from being drawn regarding management of patients with fetal ovarian cysts. Additionally, intra/interobserver reproducibility regarding application of the IOTA terminology and the diagnostic/predictive accuracy of morphological classification could not be assessed. Other limitations are that the study included patients examined at our institution over a 10-year period, so the ultrasound equipment used will have changed, improving in quality during this time, and that patients were examined by only two observers, so no data on heterogeneity in observers' skills can be assessed.
Previous studies on the outcome of fetal ovarian cysts, including a meta-analysis performed by Bascietto et al. 1 and more recently published studies 3,4,8 are summarized in Table 3. In particular, the meta-analysis 1 , including 954 fetuses, demonstrated that simple fetal ovarian cysts frequently resolve spontaneously, whereas patients with complex cysts are frequently managed with surgical procedures. Hara et al. 8 also examined the prognosis of cysts based on their content, in a retrospective series of 36 fetal ovarian cysts. They observed that necrosis was found in cysts appearing on ultrasound examination as 'complex with fluid-debris level'. This definition (complex with fluid-debris level) had been used previously by Ozcan et al. 10 as a sonographic hallmark of necrosis. We believe that the description of cyst content as 'low level' or 'ground glass', according to IOTA terminology, could provide additional information for patient management and prognosis of complex cysts. In our series, we observed that almost all cysts with anechoic and low-level content had a good prognosis (spontaneous resolution or presence of normal ovarian parenchyma at histology), whereas the case with anechoic and ground-glass cyst content was associated with necrosis.
Our study provides useful information on the outcome of fetal ovarian cysts and on the application of IOTA terminology in their description. We confirm that the description of morphology (as simple or complex) proposed by several authors is simple and effective in predicting the outcome of fetal ovarian cysts. We also believe that the introduction of IOTA terminology into clinical practice is important for the following reasons: (i) to improve characterization of the morphology of fetal cysts, with a standardized description of cyst content; (ii) to improve diagnostic test accuracy (e.g. description of cyst content could help in prediction of which cases are likely to resolve spontaneously); (iii) to aid in planning prospective multicenter studies to define the best management in cases with ovarian cyst prenatally. The role of surgery in newborns with a persistent ovarian cyst is controversial, considering the high percentage of cases with normal ovarian parenchyma at histology and the lack of complications in these cases when surgery is not performed. A randomized controlled trial should be performed to evaluate the effectiveness and safety of prenatal and postnatal intervention in cases with fetal ovarian cyst.
In conclusion, applying IOTA terminology provided a more detailed and accurate description of fetal ovarian cysts compared with the original classification into 'simple' and 'complex' categories. This study has demonstrated that anechoic fetal ovarian cysts (described as simple cysts by the original ultrasound examiner) and cysts with low-level content (described as complex by the original ultrasound examiner) frequently resolved spontaneously, whereas cysts with ground-glass, hemorrhagic, mixed or undefined cyst content (described
as complex by the original ultrasound examiner) were frequently associated with necrosis at histology following surgery.