The morphological features of the ovaries in women with polycystic ovary syndrome (PCOS) have been well described by ultrasound imaging technology. These include enlarged ovary size, multiple small follicles of similar size, increased ovarian stromal volume and echogenicity, peripheral distribution of the follicles, and higher stromal blood flow. Ultrasound identification of the presence of polycystic ovarian morphology (PCOM) has been recognized as a component of PCOS diagnosis. With the advance of ultrasound technology, new definition has been proposed recently. There is, however, a paucity of data for the ovarian morphology in normal and PCOS adolescents. Magnetic resonance imaging has the potential to be an alternative imaging modality for diagnosing PCOM in adolescence.
Highlights
- •
The diagnostic criteria for polycystic ovarian morphology are described.
- •
The ultrasound features of the polycystic ovary in adults are presented.
- •
The ovarian morphology in adolescence is elucidated.
Introduction
Polycystic ovary syndrome (PCOS) is a common endocrine disorder of unknown etiology, affecting 6–8% of reproductive-aged women . PCOS is defined as a clinical syndrome characterized by oligoamenorrhea, obesity, infertility, and signs of excess androgen. It is also associated with long-term complications including endometrial carcinoma, metabolic syndrome, and cardiovascular disease. There is no single diagnostic test for this complex disorder. The identification of PCOS is based on clinical findings that are heterogeneous and highly variable, which makes it challenging to establish the diagnosis.
In 1935, Stein and Leventhal reported a case series of seven women characterized by oligomenorrhea or amenorrhea, hirsutism, and enlarged bilateral cystic ovaries found at laparotomy . It was the first report to link ovarian pathology with clinical hyperandrogenism and oligoamenorrhea. The clinical triad they described later became the basis for PCOS diagnosis . The histopathological characteristics of Stein–Leventhal ovaries are as follows :
- 1)
Bilaterally enlarged ovaries (two to five times the normal size);
- 2)
Multiple small follicles typically of similar size (<1 cm in diameter) densely packed and lined within the superficial cortex;
- 3)
Increased stroma, occasionally with luteinized cells (hyperthecosis);
- 4)
Morphological signs of an absence of ovulation (thick smooth capsule and absence of corpora lutea and corpora albicans);
- 5)
Hyperplasia and luteinization of the inner theca cell layer.
Ultrasound by the transvaginal route has provided a noninvasive technique for assessing the ovarian morphology and is the most commonly used method for the identification of the polycystic ovary. The 2003 Rotterdam consensus ultrasound criteria of polycystic ovarian morphology (PCOM) proposed the presence of ≥12 follicles measuring 2–9 mm in diameter and/or increased ovarian volume (>10 cm 3 ) in a single ovary or both ovaries .
However, the role of ultrasound finding of polycystic ovary in the diagnosis of PCOS is under considerable debate. Although the polycystic appearance of the ovaries was part of the original disease description, it is not considered as a specific pathological entity, which may also be seen in other endocrine disorders . In addition, polycystic ovaries are common in young healthy women with a prevalence of 20–30% in women younger than 36 years . The high prevalence of the polycystic ovary has further reduced the importance of the ultrasound criteria and raised doubts on its precision.
With the advance of imaging technology, numerous efforts have been made to define the ovarian appearance in women with PCOS and the ultrasound criteria of PCOM have been refined over time. The aim of this article is to review the updated guidelines and current opinions on the sonographic features of ovarian morphology for clinical practice and further research.
Ultrasound features of the polycystic ovary
The features of a typical polycystic ovary appearance that can be identified by ultrasound show a high concordance with these histopathological characteristics :
- 1)
enlarged ovaries that are usually more spherical in shape;
- 2)
multiple small follicles of similar size arranged around the periphery, giving the appearance of a “string-of-pearls”;
- 3)
the increased and hyperechoic stroma occupying the center of the ovaries; and
- 4)
higher intraovarian stromal blood flow.
Among these features, follicle number and ovarian volume are the sonographic parameters chosen to establish the diagnostic criteria for polycystic ovary. In 2014, the Androgen Excess Society and Polycystic Ovary Syndrome Society (AEPS) guidelines recommended using FNPO (follicle number per ovary) ≥25 for the definition of PCOM when using the newer technology that affords maximal resolution of ovarian follicles (i.e., transducer frequency >8 MHz). If such technology is not available, the ovarian volume is recommended for the diagnosis of PCOM.
Ultrasound features of the polycystic ovary
The features of a typical polycystic ovary appearance that can be identified by ultrasound show a high concordance with these histopathological characteristics :
- 1)
enlarged ovaries that are usually more spherical in shape;
- 2)
multiple small follicles of similar size arranged around the periphery, giving the appearance of a “string-of-pearls”;
- 3)
the increased and hyperechoic stroma occupying the center of the ovaries; and
- 4)
higher intraovarian stromal blood flow.
Among these features, follicle number and ovarian volume are the sonographic parameters chosen to establish the diagnostic criteria for polycystic ovary. In 2014, the Androgen Excess Society and Polycystic Ovary Syndrome Society (AEPS) guidelines recommended using FNPO (follicle number per ovary) ≥25 for the definition of PCOM when using the newer technology that affords maximal resolution of ovarian follicles (i.e., transducer frequency >8 MHz). If such technology is not available, the ovarian volume is recommended for the diagnosis of PCOM.
Antral follicle number
Antral follicles measure 2–9 mm in average diameter and increased antral follicle number per ovary (FNPO) is a key and consistent morphological feature of polycystic ovary. In the physiological state, antral follicles are recruited during each menstrual cycle, and growth is followed by the selection of the dominant follicle; this follicle ovulates following the mid-cycle luteinizing hormone (LH) surge. In PCOS, follicular growth is arrested at the antral follicle stage, thereby resulting in excessive antral follicle counts (AFCs). The underlying mechanisms for excessive antral follicle formation remain unclear . A two- to threefold increase in the average counts of all forms of ripening follicles (from the stage of primary follicles to tertiary follicles) was found in the Stein–Leventhal ovaries compared to the control ovaries by histological observations . The ultrasound imaging allows the identification of antral follicles, which are characterized by anechoic cystic structures in the ovary. With the improvement in resolving power, antral follicles of diameter <2 mm, which have escaped detection previously, can be visualized by modern ultrasound equipment ( Fig. 1 A). This results in the counting of more antral follicles and thus a major but artificial increase in the prevalence of PCOM in normal populations. The criterion of >12 follicles per ovary for diagnosis becomes obsolete.
The 2014 AEPS guidelines recommend that the threshold of follicle number per ovary (FNPO) be increased to 25 with the technology available for optimal resolution (mainly transducer frequency ≥8 MHz) . The new cutoff value was based largely on two studies using the receiver operating characteristic (ROC) curve analysis. Lujuan et al. reported that the FNPO threshold of 26 follicles was the best compromise between sensitivity (85%) and specificity (94%) to distinguish women with PCOS from the normal controls . It is similar to the results obtained by Dewailly et al. In the latter study, the cutoff value is 25 follicles if women with isolated PCOM were included in the control group.
As expected, the percentage of healthy women with isolated PCOM was greatly reduced (only 7%) by applying the revised threshold of antral follicles . In a study of normogonadotropic anovulatory infertile women, an increase of antral follicle threshold from 12 to 25 lowers the prevalence of PCOS from 93.3% to 54.7%. The non-PCOM women with oligoamenorrhea had signs of less severe endocrine disturbance and 33% of them had normal LH, normal LH/FSH (follicle-stimulating hormone), and androgen levels, thereby suggesting a potential hypothalamic cause of amenorrhea . The shift of diagnosis from non-hyperandrogenic PCOS to hypothalamic anovulation challenged the inclusive Rotterdam definition of PCOS and manifested the inadequacy of diagnostic strategies. The diagnostic dilemma would remain unsolved until we gain a better understanding of the pathophysiological changes in the entity of patients with PCOM and anovulation.
The criterion of FNPO is crucial in guiding clinical diagnosis and further research work. While the accuracy of the updated threshold awaits further validation, it provides a new starting point for future PCOS studies. However, it needs to be highlighted that ultrasonography (US) findings of the ovary in the patients may exhibit a wide spectrum of the morphological patterns corresponding to the heterogeneity of this disorder . It is not possible to define the ovary as “normal” or “polycystic” distinctly by a single cutoff value which itself will certainly be renewed as spatial resolution of ultrasound continues to improve. Other morphological features, although not included in the diagnostic criteria, should also be considered when determining the presence of a polycystic ovary.
Ovarian volume
Ovarian volume is one of the diagnostic criteria for PCOS and the consensus definition of PCOM includes an ovarian volume >10 cm 3 . The ovarian volume is calculated by 2D (two-dimensional) ultrasound using the formula for a prolate ellipsoid: volume = π/6 × length × width × thickness of the ovary ( Fig. 2 A). Although the 3D (three-dimensional) method of measuring the ovarian volume avoids the use of geometric assumption, it is not recommended due to the technical and interobserver variability . In a recent diagnostic test study of Lujan et al., an ovarian volume of 10 cm 3 had 81% sensitivity and 84% specificity in diagnosing PCOS. Compared to FNPO, the threshold of 26 follicles had the best compromise between sensitivity (85%) and specificity (94%). The lower discriminatory power of the ovarian volume may reflect the greater likelihood of overlap among controls and PCOS patients . It is recognized that not all polycystic ovaries will be enlarged to this size or greater . Notably, the threshold of 10 cm 3 proposed by Rotterdam consensus was chosen empirically based on opinions of the expert panel. Recently, lower cutoff values have been proposed by other researchers ranging from 6.4 to 7.0 mL . Given that the ovarian volume threshold values might depend on the clinical and metabolic characteristics of the population studied, the AEPS guidelines highly recommended the use of in-house reference normal values. If the local norms are not available or the image quality does not allow a reliable estimate of FNPO, the existing volume threshold (>10 cm 3 ) can be used conservatively .
In addition, the changes in ovarian volume with age need to be considered when defining PCOM. The ovarian volume increases through childhood, achieves its maximum volume shortly after puberty, and declines significantly with each decade of life from age 30 to age 70 . According to the normative model developed by Kelsey et al., in the average case, the ovarian volume rises from 0.7 mL (95% confidence interval (CI) 0.4–1.1 mL) at 2 years of age to a peak of 7.7 mL (95% CI 6.5–9.2 mL) at 20 years of age with a subsequent decline to about 2.8 mL (95% CI 2.7–2.9 mL) at menopause and smaller volumes thereafter . The model also showed that 69% of the variation in ovarian volume is due to age alone. A longitudinal study of the ovarian volume in PCOS patients showed that the decrease in ovarian volume with age was less pronounced in premenopausal women with PCOS. However, the average ovarian volume was similar in postmenopausal subjects with PCOS and controls, thereby suggesting a greater decrement in the volume change in women with PCOS in transition to menopause . According to the pattern of change in ovarian volume with age found in PCOS patients and normal population, careful consideration would be given when using the threshold (>10 cm 3 ) to diagnose PCOS in adolescence or in women over the age of 40 years . In adult women, a linear pattern of decline was also observed in follicular number and anti-Müllerian hormone (AMH), both remaining higher in subjects of PCOS compared to the adult controls at all ages . The women with PCOS were found to reach menopause 2 years later than the normoovulatory women . Whether a larger follicle pool in PCOS patients observed in these findings is due to the prolonged survival of PCOS follicle remains unclear .
Follicle size distribution
It is hypothesized that there are intrinsic differences in folliculogenesis between PCOS ovaries and normal ovaries . The various sizes of the antral follicles observed in the ovary represent different stages of folliculogenesis. In an attempt to assess the follicle size distribution in PCOS ovaries, Jonard et al. compared the number of follicles categorized by different size ranges (2–5 and 6–9 mm) in the ovaries of PCOS patients and normal controls. The mean FNPO of follicles 2–5 mm in size was shown to be significantly higher in polycystic ovaries than in controls, while it was similar within the 6–9-mm range between the two groups. The results may be explained by folliculogenesis disorders: excessive early follicular growth and/or follicle arrest . The finding of small antral follicle excess was consistent with the histological observation that the contrast of follicle number between PCOS ovaries and control ovaries was best defined in the smaller tertiary follicles of diameter <4 mm . Furthermore, Webber et al. found a sixfold increase in the number of primary follicles in cortical biopsies from polycystic ovaries in anovulatory women than in normal ovaries . These data confirm the presence of an initial excessive early follicular growth and suggest that the abnormalities of folliculogenesis occur at early stages of follicle development. The subsequent follicular arrest is the failure of follicle development to proceed beyond the mid-antral stage, giving the characteristic appearance of multifollicular ovaries on ultrasound . In addition, the number of follicles of size 2–5 mm was found to be positively correlated with the serum testosterone and androstenedione levels in PCOS patients, thereby supporting the hypothesis that the increased number of smaller follicles is associated with the trophic effects of androgens .
Further ultrasound studies of different follicle cohorts, which may be facilitated by 3D technique, are warranted to substantiate the hypothesis of intrinsic aberrant folliculogenesis in PCOS patients.
Follicular distribution pattern
In 1985, PCOM was classified into two types based on the distribution of follicles in the ovary: peripheral cystic pattern (PCP) and general cystic pattern (GCP) . In the PCP, “microcysts” were aligned in the subcapsular region of the ovary ( Fig. 3 A). In the GCP, “microcysts” occupied the entire parenchyma of the ovary ( Fig. 3 B). Takahashi et al. reported that androstenedione was significantly higher in the GCP than the PCP and the LH/FSH ratio was significantly higher in the PCP than the GCP. It suggested that PCP and GCP appear to differ endocrinologically . The different patterns of follicle distribution with the ovary may reflect different pathophysiological process of disturbed folliculogenesis . However, no standardized method to assess this morphological feature exists. Recently, Christ et al. described a scoring method to determine the distribution patterns by evaluating the largest cross-sectional plane (contains ≥9 follicles) of each ovary: 1 = clear follicle aggregation around the periphery with ≤1 central follicle, 2 = follicle aggregation around the periphery with >1 central follicle, and 3 = follicle scattered throughout the ovarian stroma . Efforts in establishing an objective evaluation index of the follicle distribution pattern may help provide insights into the pathophysiology underlying the classic “string-of-pearls” appearance of a polycystic ovary.
Stromal measurements
The diagnostic values of ovarian stromal size have been evaluated by a number of studies using 2D/3D techniques. The 2D stromal area was evaluated by tracing with the caliper the peripheral profile of the stroma in the maximum plane section of the ovary ( Fig. 4 A). The 3D stromal volume was obtained by subtracting the total follicular volume from the total ovarian volume. The stromal area, stromal/total area ratio (S/A), stromal volume, and stroma/total ovarian volume ratio were found to be significantly higher in PCOS patients than controls . The S/A ratio was positively correlated with the testosterone and androstenedione levels . Similarly, the stroma/total ovarian volume ratio was reported as the most accurate predictor of hyperandrogenemia and hirsutism . However, a recent study recommended against the use of ovarian stromal measurements in the diagnosis of PCOS, as substantially lower diagnostic potential was found compared to FNPO and ovarian volume . Although ovarian stromal hypertrophy is related to the ovarian androgenic dysfunction, the question remains whether the parameters of ovarian stroma volume yield additional information for PCOS diagnosis.
Stromal echogenicity
Increased stromal echogenicity is a characteristic feature of the polycystic ovaries. But ultrasound assessment of the stromal echogenicity is a subjective assessment that may vary depending on the settings of the ultrasound machine and the patient’s body habitus. Although the 3D technique allowed quantifying the echogenicity of the ovarian stoma by calculating the mean pixel intensity, its validity has not been proved.
Stromal blood flow
High stromal vascularity is one of the characteristics observed in the polycystic ovary which may contribute to the hyperechoic appearance of the stroma ( Fig. 4 B). The sonographic finding was supported by the evidence from an experimental study of histological ovarian sections that ovaries from PCOS showed a twofold increase in blood vessel density in both superficial cortical stroma and deep cortical stroma with respect to the age-matched controls . It is hypothesized that high vascularization may lead to an abnormal growth of the theca interna (which is the site for androgen steroidogenesis) with subsequent hyperandrogenemia . Elevated concentrations of the angiogenic factors including vascular endothelial growth factor, angiopoietins, and basic fibroblast growth factors were also observed in the serum and/or follicular fluid of women with PCOS during controlled ovarian hyperstimulation and may play an important role in the increased risk of ovarian hyperstimulation syndrome in PCOS .
The increased ovarian expression of angiogenic factors and the associated increased ovarian stromal blood flow may be the underlying cause of dysregulated folliculogenesis, thereby resulting in the failure of diversion of blood flow from cohort follicles to leading follicles and an uninhibited growth of multiple follicles in women with PCO undergoing ovarian stimulation with gonadotrophins . Several studies were carried out to test the validity of ultrasound parameters of the intraovarian blood flow in diagnosing PCOS ovary. The vascular indices generated by 3D power Doppler technology have been used to quantify the blood flow in the ovarian tissue. Recently, the technique of spatiotemporal image correlation (STIC) and high-definition flow (HDF) were also introduced . A number of studies have shown that the vascularization index (VI), flow index (FI), and vascularization flow index (VFI) were significantly higher in the women with PCOS compared with the women with normal ovaries . However, there are some conflicting reports that did not find any significant difference in these indices between the PCOS subjects and controls .
The major technical limitation is that 3D power Doppler indices are highly dependent on the machine settings. Subtle changes in the power Doppler settings including gain, power, pulse repetition frequency, wall motion filter, signal rise and persistence, and speed of acquisition were found to significantly increase or decrease the final results of 3D vascular indices . This highlights the importance of strict maintenance of the machine settings if any two subjects are to be compared. Comparison of the numeric variables between different studies would be possible until a standard machine setting for 3D Doppler measurement is adopted universally . Despite the challenge for standardization, the studies of ovarian stromal blood flow are warranted to investigate and clarify the role of vascular changes in the pathogenesis of PCOS and OHSS.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
