Hum. Reprod. Advance Access originally published online on July 17, 2007
Human Reproduction 2007 22(9):2501-2508; doi:10.1093/humrep/dem202
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Ultrasound in polycystic ovary syndrome—the measuring of ovarian stroma and relationship with circulating androgens: results of a multicentric study
1 Dipartimento Chirurgico Materno Infantile e di Scienze delle Immagini, Sezione Ostetrica e Ginecologica, Università degli Studi di Cagliari, Via Ospedale 46, 09124 Cagliari, Italy 2 Dipartimento di Ginecologia ed Ostetricia, Università Cattolica del Sacro Cuore, Largo Gemelli 2, 00168 Roma, Italy 3 Clinica Medica 3, Dipartimento di Scienze Mediche e Chirurgiche, Università degli Studi di Padova, Via Giustiniani 2, 35128 Padova, Italy 4 Dipartimento di Farmacologia, Ostetricia e Ginecologia, Università degli Studi di Sassari, Viale S. Pietro, 07100 Sassari, Italy 5 Dipartmento di Medicina della Procreazione e dell' Età Evolutiva, Università degli Studi di Pisa, Via Roma 67, 56126 Pisa, Italy 6 Dipartimento di Igiene e Sanità Pubblica, Università degli Studi di Cagliari, Via Porcell 4, 09124 Cagliari, Italy 7 OASI, Institute of Research, Troina (EN), Italy
8 Correspondence address. Viale Poetto 194, 09126 Cagliari, Italy. Tel: +39-335-6622192; Fax: +39-070-6092231; E-mail: fulgh{at}tiscali.it
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
BACKGROUND: The introduction of transvaginal approach in ultrasound (US) has enabled the accurate evaluation of the structure of the ovary and stroma. Stroma represents an acknowledged US marker for polycystic ovary syndrome (PCOS). The proportion revealed between the stroma and the ovary surface in the median section (S/A ratio) had been indicated as a reliable marker for hyperandrogenism. In order to verify the feasibility of this determination in routine use and to confirm the efficacy of S/A ratio in predicting hyperandrogenism in PCOS, a multicentric study was performed in association with five Italian research groups.
METHODS: A total of 418 subjects of fertile age presenting oligomenorrhoea or secondary amenorrhoea, enlarged ovaries measuring >10 cm3 and/or >12 follicles measuring 2–9 mm in diameter took part in the study. Clinical, US and hormonal evaluations were performed in the early follicular phase or on random days in amenorrhoeic subjects. US assessment included ovarian volume, follicle number, ovarian and stroma area in median section. The hormonal study included a baseline plasma determination of LH, FSH, estradiol (E2), androstenedione (A), testosterone (T), dehydroepiandrosteronesulphate, 17-hydroxy-progesterone, sex hormone-binding globulin and prolactin. Correlations and receiver operator curves were used in statistical analysis of data.
RESULTS: S/A was found to be the best significant predictor of elevated A and T levels. In order to ascertain significant cut-off values in relation to A and T levels Youden indexes were calculated and indicated 0.32 as the best cut-off for the S/A ratio.
CONCLUSIONS: This work underlines the importance of stroma measure in improving US diagnosis of PCOS and suggest that this parameter may be used in routine clinical practice. In fact, multicentre study demonstrated the easy feasibility of such procedure without need of sophisticated machines or intensive training for operators.
Key words: PCOS/ovarian stroma/S/A ratio/androgen
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In recent years, advances in imaging technology and the introduction of transvaginal approach have enhanced the classical picture of polycystic ovaries (PCO), providing a more accurate view of the internal structure of the ovary (Adams et al., 1985
; Pache et al., 1992).
Following the consensus on PCOS diagnosis held in 2003, the ultrasound (US) criteria have been finally included as factor for the presence of the syndrome (The Rotterdam ESHRE/ASRM, 2004). Criteria deemed as being sufficiently specific and sensitive for use in defining PCO are as follows: ovarian volume (OV) (increased >10 cm3) and/or number of follicles (12 or more follicles measuring 2–9 mm in diameter) (Balen et al., 2003).
Data from the literature indicated ovarian stroma as an important marker for the presence of PCOS (Pache et al., 1992; Dewailly et al., 1994). An increased echodensity of the stroma corresponds to histological findings of prominent theca and fibrotic thickening of the albuginea described by Stein and Leventhal (1935). Although an increase in stromal volume represents one of the most specific features of PCO (Ardaens et al., 1991; Pache et al., 1992; Dewailly et al., 1994a,b), this form of assessment has been deemed of scarce efficacy for application in routine general practice and OV is thus considered a good surrogate. Nevertheless, both qualitative and quantitative stromal evaluation had previously been viewed as merely subjective and thus excluded from the consensus definition of PCO (The Rotterdam ESHRE/ASRM, 2004).
In a previous paper, we have proposed a reproducible method for stromal measure: the ratio between stromal and total area of median ovarian section (S/A). We have compared 80 oligo-amenorrhoeic women with PCOS with a control group of 30 using a 6.5 MHz probe (Fulghesu et al., 2001). Based on data from the control group the cut-off values for OV, stromal area (SA) and S/A ratio were calculated. The sensitivity of these parameters in diagnosis of PCOS was equal to 21, 62 and 100%, respectively, suggesting that a S/A ratio >0.34 (cut-off value) is diagnostic of PCOS. In addition, this parameter was significantly correlated with androgen plasma levels. At the consensus, this measure was estimated to be difficult to apply in daily routine practice (Balen et al., 2003).
In order to verify the feasibility of this determination in routine use and to confirm the efficacy of S/A in predicting hyperandrogenism in PCOS, a multicentric study, supported by the Società Italiana della Riproduzione (SIdR) was performed in association with five different Italian PCOS study groups.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Five different Italian centres took part in the study. The study was approved by the respective IRBs from each centre and informed consent obtained. All subjects were recruited between January 2003 and January 2005.
Patients
The study population was defined according to specific inclusion criteria in order to provide a homogeneous group of patients. All subject were studied by US and hormonal levels were assessed.
Inclusion criteria
Subjects of fertile age, euthyroid and normoprolactinemic (median 23.9 years, range 16–34) presenting oligomenorrhoea or secondary amenorrhoea and ovarian US revealing >12 follicles measuring 2–9 mm.
Each centre was required to identify a minimum of 20 subjects prior to inclusion in the study. Patients who had undergone hormonal treatment over the previous six months or presenting at US a dominant follicle were excluded from the study.
To exclude, the presence of a late onset adrenal enzyme defect in all patients showing hirsutism an ACTH test was performed (250 µg IV Synacthen, Ciba-Geigy, Basel, Switzerland) according to the criteria described by New et al. (1983).
The setting-up of the study was based on the calculation of sample size based on S/A attended results (280 cases). If a standard deviation of 0.5 in the S/A mean is assumed, and using a two-sided significance level of 5%, then a sample size of 260 cases gives 90% power to detect a difference of 0.2 in mean between hyperandrogenized and not hyperandrogenized subjects. This number had been increased to a minimum of 280 cases to allow for a possible presence of not evaluable cases.
Clinical protocol
Studies were conducted in amenorrhoeic patients on a random day, and oligomenorrhoeic subjects during the early follicular phase (days 2–5) of the menstrual cycle. Furthermore, in order to ascertain that ovulation had not occurred recently, the progesterone plasma level was measured on the day of the study. The testosterone (T) and androstenedione (A) levels were considered elevated where a value >0.7 for T and 3 ng/ml for A occurred, respectively. These cut-off values were extrapolated out from normal ranges indicated from laboratories of different centres. In addition, the mean and confidence intervals (CI) for T and A calculated from a control group, made up by 50 eumenorrhoeic normal weight subjects, were for T: 0.49 (0.10–0.68) ng/ml, for A: 1.75 (0.8–2.5) ng/ml, respectively.
Hormonal study included a baseline plasma determination of LH, FSH, E2, A, T, dehydroepiandrosteronesulphate (DHEAS), 17-hydroxy-progesterone (17-OHP), sex hormone-binding globulin (SHBG) and prolactin. The free androgen index was estimated for T and SHBG through application of the formula: 100x[T]x(6.11/[SHBG]). Patients showing a BMI >25<30 were considered overweight; those with a value 
30 obese.
The presence of hirsutism was determined applying the Ferriman–Gallway score criteria (Ferriman and Gallway, 1961). A score value >8 was considered positive for the presence of hirsutism.
Moreover, patients of center 1 and 2 underwent also an oral glucose tolerance test (OGTT): blood samples were collected before and 30, 60, 90, 120 and 180 min after ingestion of 75 g glucose in 150 ml water. Insulin and glucose were assayed in all samples. OGTT data are analysed as the area under curve (AUC) after the glucose ingestion and calculated by the trapezoidal rule. Accordimg to previous studies (Ciampelli et al., 2005) AUC values has been used has indirect index of insulin resistance: these data are only used for statistical analysis of linear correlation.
US examinations were performed following (The Rotterdam ESHRE/ASRM Consensus, 2004). Transvaginal US (TVUS) was performed on each patient using a 6.5 MHz endovaginal probe. In a group of 96 young patients, transabdominal US (TAUS) was performed with 3.5–5 MHz convex probe.
At least two different operators performed the US in each centre. In order to verify the interobserver variation, we calculated the data upon the first 100 cases recruited: 20 US in each centre were made by both operators in the same day and data reported separately. The statistician analysed the interobserver CV: interobserver variations did not exceed 5% in all measurements. No operator received specific training in the measurement technique. The only instructions were those reported as follows: OV was calculated for each ovary using the formula for a prolate ellipsoid: 
/6 x (D1 x D2 x D3), where D represented the maximum diameter in transverse, antero-posterior and long section. The S/A ratio was obtained as follows: the SA by outlining the peripheral profile of the stroma with a calliper, and the total ovarian area (OA) by outlining the external limits of the ovary in the maximum plane section (see Fig. 1 as example).
|
In the TVUS data used for statistical analysis were the mean of observed values for the left and right ovaries. In the TA, the results were calculated and analysed separately for each ovary. Data from right ovary were utilized for statistical analysis. We chose to analyse data from the right ovary in view of the fact that in the TA the data from the left ovary results appeared less homogeneous, possibly in relation to the presence of the rectum ampulla.
Assays
Blood samples were chilled and centrifuged at 4°C immediately after collection and then stored at –20°C until time of assay. Samples obtained from each patient were assayed simultaneously and in duplicate in each centre. All hormone concentrations were determined using the same commercial RIA kits (Diagnostic Systems Laboratories, Inc., Webster, TX). Gonadotropins were assayed by a double-antibody technique and all steroids were assayed by means of the dextran-charcoal separation technique. The intra- and inter-assay coefficients of variation were 
8 and 15%, respectively, for all determinations.
Statistical analysis
All results are expressed as mean ± SEM. Data were stored and analysed with the use of Statistical Program for Social Sciences release 12.0 (SPSS Inc., Chicago, IL, USA) on an IBM-compatible computer from a statistic Professor (L. Minerba). Normal distribution of sonographic parameters was tested with the Kolmorov–Smirnov goodness of fit test. A P-value <0.05 was considered statistically significant. The correlation between quantitative variables was ascertained by means of the Pearson coefficient of correlation. Receiver operator curves (ROC) (Zweig and Campbell, 1993) were constructed to examine the diagnostic test performance, i.e. the ability to discriminate between subjects with normal or high androgen levels. A statistical comparison of the ROC AUC was made according to methods previously described by Hanley and McNeil (1983). Starting from ROC curves, Youden indexes were calculated to find the significant cut-off values in relation to A and T levels for all parameters.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A total of 418 subjects were recruited from all centres. In 96 subjects, TAUS alone was performed in view of the fact that the transvaginal route was not approachable. These cases were analysed separately. Amenorrhoea was present in 33% of subjects (n = 138). Table 1 reports the clinical and hormonal findings from each participating centre as well as the overall results obtained. Different clinical aspects and hormone levels reflected the different populations coming to the recruitment centres, but all subjects fulfilled the inclusion criteria. Moreover, no significant difference between amerrhoeic and oligoamenorrhoeic subjects has been found (data not shown).
|
The presence of hirsutism and acne was 48.9 and 48.7%, respectively. A prevalence rate of overweight subjects was 11%, while obesity was registered in 8% of cases. No significant distribution in relation to different menstrual irregularities was observed for these clinical parameters. The prevalence of elevated androgen levels was 32.2% for both androgens. Moreover, four subjects presented elevated T with normal A levels, and three subjects elevated A with normal T levels.
Fig. 2 illustrates the US characteristics of all subjects studied. OV, S/A ratio, SA, OA and the number of follicles are reported. TAUS performed throughout the various study centres were compared: all ovarian characteristics were similar in the five centres and no differences in SA values were reported by the different operators. The measurement of SA, and S/A, in TA scans resulted significantly lower than transvaginal reports, likely due to technical issues, therefore, we chose to exclude TAUS from further statistical analysis.
|
Table 2 reports the linear correlations between TVUS findings, gonadotropins and androgen plasma levels. All US parameters are significantly related with LH. SA, OV, OA and S/A are significantly related to A and T, as well as to SHBG levels, although the strongest correlations have been demonstrated for T and A and S/A ratio. No significant findings were observed with regard to DHEAS levels. The lowest significant coefficient values were displayed between OV and A and 17-OHP. Due likely to the fact that a number of follicle >12 represented one of the admission criteria all correlations studies resulted o be not signficant.
|
Moreover, only S/A ratio significantly correlted with insulin AUCs (r = 0.33; P > 0.001), whereas all other echographic parameters did not show any correlation with insulin secretion.
Fig. 3 illustrates ROC curves for TVUS findings and elevated values of A (top) and T (bottom). ROC curves indicate diagnostic performance of US parameters in subjects characterized by elevated A and T levels. The ideal screening test will approach or reach the upper left corner of the graph: S/A was revealed as being the most accurate predictor for both elevated A and T levels (AUC 0.73, P < 0.01).
|
SA also proved to be a significant indicator of hyperandrogenism, displaying an AUC of 0.67 for both A and T. OV reported AUC values of 0.62 and 0.60 (P < 0.05) for A and T, respectively, whereas the OA AUC value did not prove to be a significant diagnostic predictor for either A or T (AUC 0.57 and 0.572, respectively).
By means of Youden indexes, starting from ROC curves, we calculated the best cut-off for each parameter. Table 3 indicates the calculated cut-off points. This method indicated 0.32 as the best cut-off for the S/A ratio for both A and T, 10.3 cm3 for the OV, 1, 4 cm2 for the SA. It was not possible to identify the cut-off for the median OA in view of the lack of statistical significance of ROC curves. Furthermore, in the same table (Table 3) sensitivity and specificity, positive predictor and negative predictor values of all parameters are indicated. Moreover, all other statistical analyses performed to evaluate ROC curves for each index have been reported.
|
S/A ratio demonstrated the highest degree of sensitivity and specificity. The last column illustrates a statistic comparison between the best index (S/A ratio) and other indexes obtained: S/A ratio is confirmed as the best indicator of hyperandrogenism respect to other US indicators.
Fig. 4 shows a flow chart reassuming the principal study findings: patients showing increased S/A ratio had higher incidence of hyperandrogenism. Moreover, US parameters evaluated separately in hyperandrogenic (A and/or T levels exceeding normal values) and normoandrogenic subjects demonstrate that hyperandrogenic subjects displayed significantly higher values of SA (1.9 ± 0.07 and 1.5 ± 0.04 mm2, respectively, with P<0.0005) and S/A ratio (0.38 ± 0.0009 and 0.31 ± 0.00 004, respectively, with P < 1.9x10–6), while no differences were demonstrated for OV and OA.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In a previous paper, we demonstrated how evaluation of the ratio between stromal and total area of median ovarian section (S/A), may enhance selection of PCOS patients with a virtual absence of false-positive and negative cases, being indeed strictly related to androgen concentration (Fulghesu et al., 2001
). The bulk of these data allowed us to define a cut-off capable of discriminating hyperandrogenic subjects. While this result was considered of use in a research setting, it was viewed as being rather hard to achieve in routine practice. The present paper attempts to demonstrate how the method could also be applied in a large patient group and provide significant results from a large number of operators. For this purpose, we persuaded the Italian Society for Reproduction (SIdR) to propose a multicentric study aimed at evaluating S/A ratio in patients affected by oligomenorrhoea or secondary amenorrhoea with US finding of PCO, thus classifiable, in line with Rotterdam criteria, as affected by PCOS. We have recruited 418 valuable cases and attempts were made to confirm the results previously reported.
Several researchers took part in the study, but only centres with at least 20 valuable cases were screened, in order to obtain more homogeneous hormonal data. Data from this multicentric study represented a large population recruited in different part of Italy. The clinical and hormonal differences reported in the Table 1 probably reflected the big spectrum of women who complains PCOS: in fact centre 1 is specialized in adolescent gynaecological disease and recruited several young and lean subjects; centre 2, is an important sterility centre; centre 3 is located in a countryside, whereas centre 5 is located in a rich town, where media are more important in influencing the women lifestyle.
No differences in US results seemed to be related to the place of origin or to the operator: interobserver comparison performed for the first 100 cases confirmed the virtual absence of subjectivity in measurements.
In this new series of patients, we provided confirmation that US findings, S/A, as well as SA and OV are positively related to androgen levels and demonstrated significant AUCs in ROC analysis. However, S/A would seem to be characterized by the most efficient diagnostic performance for hyperandrogenism. Furthermore, these data indicated 0.32 as the upper limit value of S/A for both androgens confirming the cut-off indicated by our preliminary results.
The presence of SA occupying 35% of the total ovarian surface area has recently been documented by Yoo et al. (2005) in an obese adolescent PCOS population. In this paper, the authors were not able to determine the SA by means of US, due to the impossibility of performing TVUS. However, data provided by magnetic resonance imaging (MRI) are similar to those found by all operators of our groups in hyperandrogenized subjects, confirming that a 35% presence of stroma in the median ovarian section could be taken as indicative for PCOS.
On the other hand, our data indicated a cut-off of 10 cm3 for OV as diagnostic of PCOS, as suggested by several authors (Reyss et al., 2006) and by the ESHRE/ASRM consensus, confirming that studied subjects are representative of general PCOS population.
Recently, OV has been reaccredited among the diagnostic criteria for PCOs. Puzigaka et al. (1991) demonstrated a good correlation between OV and androgen levels. Indeed, a threshold of 10 cm3 for OV, chosen at PCOS consensus has not to date been based on appropriate studies such as ROC analysis (Reyss et al., 2006). In the latter study, realized in a group of 154 PCOS patients compared with 54 controls, an arbitrary threshold at 7 cm3 is proposed as the best compromise between specificity (91.2%) and sensibility (67.5%). On the other hand, in this paper statistical analysis was performed taking into account the presence or absence of a diagnosis of PCOS according to criteria of the National Institute of Health. It should be underlined that our study is not concerned with the diagnosis of PCOS, but rather faces the possibility of detecting elevated androgen levels. Indeed, in our study only subjects having >12 follicles and oligo- or amenorrhoea, thereby PCOS to all effects as per indications provided by the consensus, were investigated.
Indirectly, the conclusions of our study are further supported by the correlation found between S/A ratio and insulin AUC values obtained on a limited number of patients recruited; in any case these findings are in agreement with those recently obtained on a very large population investigated for this purpose (Belosi et al., 2006).
In the past, several authors have considered the presence of increased stroma as indicator of ovarian hyperandrogenism, capable of differentiating between PCO and multifollicular morphology and representing the most sensitive and specific US indication of PCOS (Venturoli et al., 1995). Various attempts at stroma determination have included the use of computerized ultrasonic technique 24, 3D US and colour and pulsed Doppler US (Kyei Mensah et al., 1996a,b; Ajossa et al., 2001) and MRI (Yoo et al., 2005). Nevertheless, these methods failed to be widely adopted due to the lack of simplicity or excessive cost. Data from this multicentric study demonstrated that the method proposed by our group is easy to use and reproducible. In fact, several operators from five different centres obtained similar results without any specific training. Nevertheless, we have proposed this method for use only in TVUS.
The predictive value of US characteristics for endocrine abnormalities frequently associated with PCOS (elevated LH, A and/or T levels) is hard to find in literature. A Santbrink et al. (1997) study on 217 infertile PCOS patients, demonstrated that increased follicle number is the more sensitive (70%) and less specific (47%) predictor for increased androgen levels, compared with increased OV (>10.8 ml) (sensitivity 57%, specificity 67%). In the paper, concerned stroma evaluation was only of a qualitative nature and was considered too subjective (sensitivity 52%, specificity 57%). In 1998, Keyi-mensah et al. (1998) ademonstrated a positive correlation between OS volume and A in PCOS. Using computerized reading of US images, in 1997 in a small group of PCOS, Devailly demonstrated a significant correlation between both OA and stroma and A and 17-OHP, whereas neither OA nor SA correlated with T and LH (Stein et al., 1935). More recently, Nardo et al. (2002) failed to find any correlation between T levels and US parameters in 23 PCOS subjects studied by 3D TVUS. In our paper, we clearly demonstrate the presence of significant correlations between US and hormonal findings: the number of studied subjects, the reproducibility of scanning technique and the statistical approach applied all add further support to our findings.
In conclusion, we propose the introduction of S/A ratio as marker of hyperandrogenism in patients affected by menstrual irregularities and multicystic or enlarged ovary at US.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Adams J, Franks S, Polson DW, Mason HD, Abdulwheid M, Tucker M, Morris DV, Price J, Jacobs HS. Multifollicular ovaries: clinical and endocrine features and response to pulsatile gonadotropin releasing hormone. Lancet (1985) 2:1375–1378.[Web of Science][Medline]
Ajossa S, Guerriero S, Paoletti AM, Orrú M, Floris S, Mannias M, Mellis GB. Uterine perfusion and hormonal pattern in patients with polycystic ovary syndrome. Journal of Assisted Reproduction Genetics. 436–440.
Ardaens Y, Robert Y, Lemaitre L, Fossati P, Dewailly D. Polycystic ovarian disease: contribution of vaginal endosonography and reassessment of ultrasonic diagnosis. Fertil Steril (1991) 55:1062–1068.[Web of Science][Medline]
Balen AH, Laven JSE, Tan SL, Dewailly D. Ultrasound assessment of the polycystic ovary: international consensus definitions. Hum Reprod Update (2003) 9:505–514.
Belosi C, Selvaggi L, Guido M, Romualdi D, Fulghesu AM, Lanzone A. Is the PCOS diagnosis solved by ESHRE/ASRM 2003 consensus or could it include ultrasound examination of the ovarian stroma? Hum Reprod (2006) 21:3108–3115.
Ciampelli M, Leoni F, Cucinelli F, Mancuso S, Panunzi S, De Gaetano A, Lanzone A. Assessment of insulin sensitivity from measurements in the fasting state and during oral glucose tolerance test in polycystic ovary syndrome and menopausal patients. J Clin Endocrinol Metab (2005) 90:1398–1406.
Dewailly D, Robert Y, Helin I, Ardaens Y, Thomas Desrousseaux P, Laemetre L, Fossati P. Ovarian stromal hypertrophy in hyperandrogenic women. J Clin Endocrinol Metab (1994) 41:557–562.
Ferriman D, Gallway J. Clinical assessment of body hair growth in woman. J Clin Endocrinol Metab (1961) 21:1440–1447.
Fulghesu AM, Ciampelli M, Belosi C, Apa R, Pavone V, Lanzone A. A new ultrasound criterion for the diagnosis of polycystic ovary sindrome: the ovarian stroma/total area ratio. Fertil Steril (2001) 76:326–331.[CrossRef][Web of Science][Medline]
Hanley JA, Mc Neil BJ. A method of comparing the areas under receiver operating characteristic curves derived from the same cases. Radiology (1983) 14:839–843.
Kyei Mensah A, Maconochie N, Zaidi J, Pittrof R, Campbell S, Tan SL. Trans-vaginal three dimension ultrasound: reproducibility of ovarian and endometrial volume measurements. Fertil Steril (1996) 66:718–722.[Web of Science][Medline]
Kyei-Mensah A, Zaidi J, Campbell S. Ultrasound diagnosis of polycystic ovary syndrome. Baillieres Clin Endocrinol Metab (1996) 10:249–262.[CrossRef][Web of Science][Medline]
Kyei-Mensah AA, Lin Tan S, Zaidi J, Jacobs HS. Relationship of ovarian stromal volume to serum androgen concentrations in patients with polycystic ovary syndrome. Hum Reprod (1998) 13:1437–1441.
Nardo LG, Buckett WM, White D, Digesu AG, Franks S, Khullar V. Three-dimensional assessment of ultrasound features in women with clomiphene citrate-resistant polycystic ovarian syndrome (PCOS): ovarian stromal volume does not correlate with biochemical. indices. Hum Reprod (2002) 17:1052–1055.
New MI, Lorentzen F, Lerner AJ, Kohn B, Oberfield SE, Pollack MS, Dupont B, Stoner E, Levy DJ, Pang S, Levine LS. Genotyping steroid 21-hydroxylase deficiency: hormonal reference data. J Clin Endocrinol Metab (1983) 57:320–326.
Pache T, Wladimiroff J, Hop WC, Fauser BC. How to dicriminate between normal and polycystic ovaries: trans-vaginal US study. Radiology (1992) 183:421–423.
Puzigaca Z, Prelevic GM, Stretenovic Z, Balint Peric M. Ovarian enlargement as a possible marker of androgen activity in polycystic ovary syndrome. Gynecol Endocrinol (1991) 5:167–174. Sep.[Web of Science][Medline]
Reyss AC, Proust Richard C, Catteu-Jonard S, Dewailly D. Rotterdam consensus in adolescent girls: which investigations and how to interpret them to make the diagnosis of PCOS? Gynecol Obstet Fertil (2006) 34:341–346.[CrossRef][Medline]
Stein IF, Leventhal ML. Amenorrhea associated with bilateral polycystic ovaries. Am J Obstet Gynecol (1935) 29:181–191.[Web of Science]
The Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group. Revised 2003 Consensus on diagnostic criteria and long-term health risks related to polycistic ovary syndrome. Fertil Steril (2004) 81:19–25.[Web of Science][Medline]
Van Santbrink EJ, Hop WC, Fauser BC. Classification of normogonadotropic infertility: polycystic ovaries diagnosed by ultrasound versus endocrine characteristics of polycystic ovary syndrome. Fertil Steril (1997) 67:452–458.[CrossRef][Web of Science][Medline]
Venturoli S, Porcu E, Fabbri R, Pluchinotta V, Ruggeri S, Macrelli S, Paradisi R, Flamigni C. Longitudinal change of sonographic ovarian aspects and endocrine parameters in irregular cycle of adolescents. Ped Research (1995) 38:974–980.[Web of Science][Medline]
Yoo RY, Sirlin CB, Gottschalk M, Chang RJ. Ovarian imaging by magnetic resonance in obese adolescent girls with polycystic ovary syndrome: a pilot study. Fertil Steril (2005) 84:985–995.[CrossRef][Web of Science][Medline]
Zweig MH, Campbell G. Receiver operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. Clin Chem (1993) 39:561–577.
Submitted on November 3, 2006; resubmitted on March 20, 2007; accepted on May 8, 2007.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  D. Shi, M. K. Dyck, R. R. E. Uwiera, J. C. Russell, S. D. Proctor, and D. F. Vine A Unique Rodent Model of Cardiometabolic Risk Associated with the Metabolic Syndrome and Polycystic Ovary Syndrome Endocrinology, September 1, 2009; 150(9): 4425 - 4436. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||