Abstract
Polycystic ovary syndrome (PCOS) is the most common endocrine disorder among women in the reproductive age group. The reported prevalence of PCOS ranges between 5 and 15 percent . This variation is largely dependent on the population studied and the diagnostic criteria used to establish the diagnosis . Obesity, infertility, menstrual disorders and signs of hyperandrogenism are common clinical presentations of this syndrome. Women with PCOS are also at increased risk of long-term health problems such as cardiovascular disease and type-2 diabetes. In addition, it is associated with increased risk of endometrial hyperplasia and endometrial cancer secondary to exposure to unopposed oestrogen.
Introduction
Polycystic ovary syndrome (PCOS) is the most common endocrine disorder among women in the reproductive age group. The reported prevalence of PCOS ranges between 5 and 15 percent [1]. This variation is largely dependent on the population studied and the diagnostic criteria used to establish the diagnosis [2]. Obesity, infertility, menstrual disorders and signs of hyperandrogenism are common clinical presentations of this syndrome. Women with PCOS are also at increased risk of long-term health problems such as cardiovascular disease and type-2 diabetes. In addition, it is associated with increased risk of endometrial hyperplasia and endometrial cancer secondary to exposure to unopposed oestrogen.
Polycystic ovary syndrome is a diagnosis of exclusion. However, several sets of diagnostic criteria were proposed by different expert groups to make a diagnosis, namely the National Institute of Health (NIH) criteria, the Rotterdam criteria and the Androgen Excess and Polycystic Ovary Syndrome (AE-PCOS) Society criteria [3] (Table 7.1). The NIH PCOS consensus compiled in 1990 was largely based on expert opinion rather than evidence from clinical trials. According to the NIH consensus, both chronic anovulation and signs of hyperandrogenism (clinical or biochemical) must be present to establish the diagnosis of PCOS. Other disorders, including non-classic congenital adrenal hyperplasia (NC-CAH), Cushing’s syndrome, androgen secreting tumours, hyperprolactinaemia, and thyroid dysfunction, need to be excluded [4].
Table 7.1 Phenotypes of PCOS according to the Androgen Excess and PCOS Society (AE-PCOS)
| Parameter | Phenotype A | Phenotype B | Phenotype C | Phenotype D |
|---|---|---|---|---|
| Hyperandrogenism | + | + | + | – |
| Ovulatory dysfunction | + | + | – | + |
| Polycystic ovarian morphology | + | – | + | + |
Thirteen years later, the Rotterdam ESHRE/ASRM-sponsored PCOS Consensus Workshop Group revised the NIH consensus and added the polycystic ovary (PCO) ultrasound appearances to the diagnosis of PCOS. The 2003 Rotterdam consensus includes three criteria: (1) oligo- or anovulation; (2) clinical or biochemical signs of hyperandrogenism; and (3) polycystic-appearing ovaries on imaging. In order to make the diagnosis of PCOS, two of the three criteria must be present. The diagnosis of PCOS is by exclusion and other disorders of hyperandrogenaemia and ovulatory dysfunction must be first excluded [5]. The Rotterdam criteria for polycystic ovaries on ultrasound are 12 or more follicles measuring 2–9 mm and/or ovarian volume more than 10 cm3 in one or both ovaries.
More recently, an international evidence-based guideline for the assessment and management of PCOS has identified the most effective ultrasound criteria to diagnose PCOS considering the recent advances with improved resolution ultrasound [6]. We discuss these criteria later in this chapter.
Polycystic Ovary Syndrome Phenotypes
Four clinical phenotypes of PCOS have been recognized according to the combination of PCOS manifestations. These phenotypes are illustrated in Table 7.1. The phenotypes are classified into: (1) classic PCOS which includes groups A and B, and (2) newer PCOS which comprises groups C and D; both include the PCO morphology as a feature in contrast to the classic PCOS group [7,8].
The addition of the morphological appearance of polycystic ovary to the Rotterdam diagnostic criteria resulted in identifying two additional PCOS phenotypes: (1) women with ovulatory dysfunction and polycystic ovaries but without hyperandrogenism; and (2) ovulatory women with hyperandrogenism and polycystic ovaries [10]. In 2009, the AE-PCOS society expert review reassessed the key features of PCOS and compiled a new consensus. The AE-PCOS diagnostic criteria include: (1) hyperandrogenism, including hirsutism and/or hyperandrogenaemia; (2) ovarian dysfunction, including oligo-anovulation and/or polycystic-appearing ovaries; and (3) exclusion of other related disorders [3].
According to the AE-PCOS consensus, hyperandrogenism is a necessary criterion for the diagnosis of PCOS. Therefore, the PCOS phenotype of ovulatory dysfunction and polycystic ovaries but without hyperandrogenism (previously acceptable by the Rotterdam criteria) does not qualify for the diagnosis of PCOS [3].
Polycystic Ovary Morphology on Ultrasound
Pelvic ultrasound is considered an essential tool in the evaluation of women with suspected PCOS. Since the first ultrasound study of the female pelvis in the 1970s [11], several attempts have been made to identify ultrasound criteria to define PCO. However, there has been no complete consensus on these ultrasound criteria to date.
One of the early and widely used definitions was proposed in 1985 and defined PCO as the presence of 10 or more cysts measuring 2–8 mm in diameter, arranged peripherally around a dense core of stroma or scattered through an increased amount of stroma [12]. Other ultrasound studies reported enlarged ovaries with the follicle arranged peripherally or scattered throughout hyperechogenic stroma in 70 per cent of symptomatic women with PCOS [13]. These criteria were based on transabdominal ultrasound. Further studies have shown that the ovaries cannot be adequately assessed in 42 per cent of the cases by using transabdominal ultrasound [14]. Several limiting factors have been identified, such as obesity, low resolution, the full bladder distorting the pelvic anatomy and loops of bowel masking the ovaries [15].
The introduction of transvaginal ultrasound, with its improved resolution, has led to better visualization of the pelvic structures and development of more precise ultrasound criteria for the diagnosis of PCO [16]. The 2003 ESHRE/ASRM meeting in Rotterdam compiled a consensus definition for PCO. According to the Rotterdam criteria, polycystic ovaries are present when (1) one or both ovaries demonstrate 12 or more follicles measuring 2–9 mm in diameter; or (2) the ovarian volume exceeds 10 cm3. Only one ovary meeting either of these criteria is sufficient to establish the presence of polycystic ovaries [17]. This definition recognized two important parameters: (1) ovarian volume and (2) number of follicles.
Ovarian Volume and Area
There are several reports in the literature comparing ovarian volume in women with PCOS to those of healthy women. A volume of 10 cm3 was set as the threshold volume for PCO [18,19]. However, some reports suggest that ovarian volume alone is not enough for the diagnosis of PCO due to the high degree of volume overlap between normal ovaries and PCO [15]. There are many formulas for calculation of the ovarian volume [20]. Several studies proposed that ovarian volume should be calculated on the basis of the simplified formula for an ellipsoid (0.5 × length × width × thickness of the ovary) [18,21] (Figure 7.1). It is worth mentioning that increased total ovarian area has been proposed as an ultrasound criterion for the diagnosis of PCO. In a study of 48 control cases, the ovarian stroma was quantified by subtracting the cyst area from the total ovarian area on a longitudinal plane of the ovary and the upper normal limit (95th percentile) of the stromal area was set at 380 mm2/ovary. This study also observed a correlation between the total ovarian area and the stromal area [22]. The advantages of such an approach include reliable acquisition from transabdominal or transvaginal ultrasound, and it does not require a computerized-assisted analysis as the modern ultrasound machine software can readily measure the area of any outlined structure. In a large observational study, the sum of the areas of both ovaries was less than 11 cm2 in normal women and an ovarian area above this cutoff was found exclusively in women with PCOS [22]. Other authors set this cutoff at 5.5 cm2 per ovary [23].
Number of Follicles
The Rotterdam consensus identified the threshold to diagnose PCO on ultrasound as the presence of 12 or more follicles measuring 2–9 mm in diameter per ovary. This cutoff was based on a study published by Jonard et al. in 2003, which reported that 12 or more follicles offered the best compromise between 99 per cent specificity and 75 per cent sensitivity in detection of PCOS [19]. However, more recent studies have challenged this cutoff and showed high prevalence of ovaries with more than 12 follicles in healthy women [24,25]. The AE-PCOS sponsored a task force to review results from more recent studies and proposed increasing the threshold to 25 or more follicles [26]. A recent international guideline has recommended [6] a threshold of ≥20 follicles, measuring 2–9 mm, per ovary, to diagnose PCO.
Follicle Distribution
Two types of PCO have been identified based on their follicle distribution on ultrasound: (1) peripheral cystic pattern (PCP), where the follicles are distributed in the subcapsular region (pearl necklace appearance) (Figure 7.2); and (2) general cystic pattern (GCP), where the follicles are scattered throughout the entire ovarian parenchyma [10] (Figure 7.3). This peripheral distribution is usually observed in younger patients, while in older women the follicle distribution is more generalized [27]. In addition, each appearance reflects a specific endocrine pattern [16].
Multifollicular Ovary
The term multifollicular ovary (MFO) describes normal-sized or slightly enlarged ovaries with multiple (12 or more) follicles (4–10 mm in diameter) and having a normal stromal size [12] (Figure 7.4). Multifollicular ovary is a common ultrasound finding and can be seen in normal pubertal girls, girls with central precocious puberty, hyperprolactinaemia and in women with hypothalamic amenorrhoea. Therefore, the whole clinical picture should be considered before rushing into a diagnosis of PCOS when MFO is recognized on ultrasound.
Stromal Echogenicity and Stromal Volume
Ovarian stromal hyperechogenicity (Figures 7.5 and 7.6) has been reported as one of the earliest ultrasound criteria for the diagnosis of PCO [12]. Another study has shown no difference in stromal echogenicity between PCO and normal ovaries using transvaginal ultrasound, and suggested that the subjective impression of hyperechogenic stroma may be due to an increased stromal volume [28]. The role of stromal hypertrophy and hyperechogenicity has been emphasized as a reliable ultrasound sign to differentiate PCO from other causes of MFO with a reported sensitivity of 94 per cent [18,29]. However, hyperechogenicity is highly subjective and is dependent on the settings of the ultrasound machine. Total ovarian volume has largely replaced stromal volume as a parameter to diagnose PCO by ultrasound as it is easily measured in clinical practice and correlates well with stromal volume [30].
Stromal Blood Flow
The use of Doppler ultrasound allows for the detection of the vascularity of the ovarian stroma. Polycystic ovaries are characterized by increased stromal vascularity which contributes to the hyperechogenic appearance of the ovarian stroma. These findings were supported by observations from histological studies of PCO, which showed a twofold increase in the density of cortical stromal blood vessels compared to normal ovaries [31]. Several ultrasound parameters of the ovarian stromal blood flow have been proposed for the diagnosis of PCO, including 2D ultrasound pulse wave Doppler indices of ovarian stromal vessels (Figure 7.7) and 3D ultrasound power Doppler vascular indices: vascularization index (VI), flow index (FI) and vascularization flow index (VFI) (Figure 7.8). Some studies reported significant increase in these 3D power Doppler indices in PCO compared to normal ovaries [32,33]. However, there are other reports that did not find any significant difference in these indices between the PCO and normal ovaries [34,35]. A major limiting factor of the 3D power Doppler indices is that they are significantly dependent on the ultrasound machine settings [36].
Figure 7.8 3D power Doppler measurement of the ovarian blood flow. It measures three vascular indices – vascularization index (total Doppler signal, i.e. total blood flow), flow index (Doppler signal intensity) and vascularization flow index.
Polycystic Ovaries in Women without PCOS
The presence of PCO on ultrasound scan does not automatically establish a diagnosis of PCOS. It is estimated that polycystic ovaries are present in 25 per cent of normal ovulating women [37] and in 27–39 per cent of adolescent girls [38]. It has also been postulated that girls with PCO may have a genetic predisposition to the syndrome [39]. Some authors associated the presence of PCO with menstrual irregularities [38]. Others suggested that the presence of PCO in ovulatory women is associated with increased risk of recurrent miscarriage [40] and subfertility [41]. Therefore, it remains unclear whether PCO appearance on ultrasound represents a normal variant or an unexpressed form of PCOS.
3D Ultrasound Application
The current ultrasound criteria for the diagnosis of PCO are based on 2D ultrasound. However, 2D ultrasound has some limitations. First, there is significant inter- and intra-observer variability when making the diagnosis of PCO on ultrasound [42]. Second, the inter-observer agreement for follicle count was reported to be poor [43].
The introduction of 3D ultrasound in reproductive medicine has provided new methods for the assessment of antral follicle count (AFC), ovarian volume, and stromal vascularity.
Antral Follicle Count
There are two methods for assessment of follicle count: (1) 3D multiplanar view and (2) SonoAVC (sonography-based automated volume count). In the 3D multiplanar view method, the three orthogonal planes (longitudinal or A, transverse or B and coronal or C) of a stored 3D ovarian volumetric dataset are displayed simultaneously (Figure 7.9). Then, the observer counts the number of the follicles using the three perpendicular planes shown in the multiplanar view. The inter-observer reliability was significantly improved by utilizing this method [44] (Figure 7.9).
In the SonoAVC method, the 3D ovarian volume dataset is obtained first. Then, it is processed by the SonoAVC software, which identifies every single follicle, codes it with a specific colour and automatically measures the follicle diameters and volumes (Figure 7.10). This method is not robust or accurate enough currently and requires further post-processing to select or delete some follicles or hypoechoic areas that are falsely missed or recognized by the software, respectively. This automated method is highly valid and provides more accurate values than those obtained from 2D measurements [45].
