To evaluate the role of ultrasound and color Doppler analyses in improving the differential diagnosis of pubertal precocities.
To evaluate the role of ultrasound and color Doppler analyses in improving the differential diagnosis of pubertal precocities.
Sixty-nine girls with premature (<8 years old) breast development and/or pubic hair growth underwent: auxological (height, weight, body mass index, skeletal maturation), hormonal (basal, gonadotropin releasing hormone (GnRH)-test, adrenocorticotropic hormone test), and sonographic (uterine and ovarian volume and endometrial echo) including color Doppler (uterine arteries) evaluations.
The uterine size was greater in girls with a pubertal response to the GnRH test (Group II, n = 16; 7.48 ± 4.18 mL) than in those with a prepubertal response to the GnRH test (Group I, n = 17; 3.02 ± 1.36 mL; P = 0.006), an isolated pubarche (Group III; n = 20; 2.58 ± 1.32 mL; P < 0.001) or an isolated thelarche (Group IV, n = 16; 1.82 ± 1.07 mL; P < 0.001). Endometrial echo was observed in 87.5%, 29.4% and 5% of girls, respectively, in Groups II, I and III. The Doppler analysis of the uterine arteries showed the lowest impedance to be in patients with a pubertal response to the GnRH test (Group II).
Sonographic and color Doppler parameters may improve the diagnosis of GnRH-dependent precocious puberty and may be useful to determine which girls need treatment. Copyright © 2003 ISUOG. Published by John Wiley & Sons, Ltd.
Female puberty is the physiological process whereby a girl changes into a woman as a consequence of the progressive maturation of the hypothalamic–pituitary–ovarian axis. In western countries, females generally enter puberty between 9 and 13 years of age1. Sexual precocities may be considered as the expression of secondary sexual characteristics prior to the pubertal age and include isosexual precocious puberty, isolated thelarche and isolated premature pubarche.
Isosexual precocious puberty in girls may be defined as the premature development of secondary sexual characteristics associated with uterine and ovarian maturation. Linear growth acceleration and bone maturation, leading to early epiphyseal fusion and short adult height, may also be present. Clinically, the diagnosis is considered in girls developing pubertal changes prior to 8 years of age2. Approximately 95% of isosexual precocity is gonadotropin releasing hormone (GnRH)-dependent and is due to idiopathic activation of the hypothalamic–pituitary–ovarian axis without evident underlying anatomical causes3. The causes of GnRH-independent puberty include gonadal, adrenal and iatrogenic sources of hormone production4, 5 and its treatment is related to the solution of the specific underlying factors. GnRH-dependent isosexual precocity requires the GnRH-agonist suppression of the hypothalamic–pituitary–ovarian axis. However, some studies have described girls with intermediate forms of precocious puberty6 who were considered to be unsuitable for treatment with GnRH-agonist. Thus, it is important to define the limits of an active hypothalamic–pituitary–ovarian axis in order to determine that true puberty has begun.
The GnRH stimulation test is considered as the gold standard to distinguish between the intermediate forms of precocious puberty6 that are not suitable for treatment with GnRH-agonist, and true precocious puberty. However, there is no concordance as to the criteria which should be used for the diagnosis6, 10, 11.
Pelvic sonography is an accurate and non-invasive method for evaluating the female pelvis in infancy and childhood. Several investigators have documented increases in uterine and ovarian volume in childhood, with an increase in the number and size of the developing follicles in the years leading up to puberty4, 12, 13. The changes observed with ultrasound agree with those observed at postmortem examination14. Many studies have described pelvic sonographic variations in pubertal abnormalities, some of them reporting that sonography is useful to differentiate central precocious puberty from isolated premature thelarche or isolated premature adrenarche, while others describe wide overlap in quantitative and qualitative findings in the above pathological conditions2, 12, 15, 16.
The use of color Doppler ultrasound facilitates the detection of small vessels in the uteroovarian circulation and the measurement of impedance to flow in this vascular tree. To the best of our knowledge, only two papers have analyzed the influence of hormonal changes and subsequent internal genitalia transformations during puberty on uterine vascular modifications17, 18. However, we recently demonstrated that uterine artery Doppler analysis may assist the diagnosis of GnRH-dependent precocious puberty, that it may be useful for the selection of those girls needing treatment and that it may simplify the follow-up of girls treated for precocities5. In addition, Doppler analysis seems to be helpful in the management of premature pubarche19.
The aim of this study was to evaluate the role of ultrasound and color Doppler analyses in improving the differential diagnosis of pubertal precocities.
Sixty-nine girls referred to Modena Hospital Auxological and Paediatric Endocrine Clinic for the evaluation of premature breast development and/or pubic hair growth between June 1998 and June 2000 were selected to participate in this study. Each symptom developed before the age of 8 years. The study protocol was in accordance with the Helsinky II declaration and was approved by the hospital research review committee. Girls participated in the study after both informed consent from parents and agreement from the minors had been obtained.
On the basis of history, physical examination, basal sonography and laboratory data, patients who had been excluded from the study included those with: chronic disease, Cushing's syndrome, hyperprolactinemia, ovarian cysts (>10 mm in maximum diameter) and polycystic ovaries (> five subcortical follicles 2–10 mm in maximum diameter, increased ovarian volume and increased ovarian stroma echogenicity)20. Cases of GnRH-independent puberty (i.e. hypo/hyperthyroidism, sex steroid-secreting tumors, congenital adrenal hyperplasia) were also excluded. No patient had received hormonal therapy before the study.
Patients with both pubic hair and breast stage II underwent a GnRH stimulation test and based on the results they were subdivided into two groups. Group I (n = 17) included those presenting serum luteinizing hormone (LH) and follicle stimulating hormone (FSH) prepubertal response and Group II (n = 16) included those giving a pubertal response. Seven patients (four in Group I and three in Group II) were also part of a parallel study5. The first group was submitted to an adequate follow-up program (careful examination of the rate of progression of physical changes, linear growth and bone maturation, and of hormonal variations), and the latter was submitted to hypothalamic–pituitary–ovarian axis suppression with GnRH-agonists and periodical auxological, clinical and hormonal evaluations. A 1-year follow-up was completed in all patients.
To exclude congenital adrenal hyperplasia and verify a possible ‘exaggerated adrenal response’ an adrenocorticotropic hormone (ACTH) stimulation test was performed in patients with isolated pubarche (Group III; n = 20). Four of these patients were also part of a parallel study19. Girls with a prepubertal response to the GnRH test and those with an isolated thelarche (Group IV; n = 16), a benign condition which does not need any treatment, were submitted to an adequate follow-up program.
The pubertal development in all girls was staged by a single examiner (L.I.) according to the classification of Tanner21, 22. Standing height was measured using a Harpenden stadiometer (Holtain Ltd., Crymych, UK) to the nearest 0.1 cm; weight was measured on a digital scale with a precision of 0.1 kg (SECA 707, HH, Modena, Italy). The body mass index (BMI = weight (kg)/height2 (m2)) was calculated in all patients. Skeletal maturation (bone age) was determined from X-rays of the wrist and hand and was staged according to Greulich and Pyle23.
Uterine and ovarian sonographic examinations were performed using a AU4 Idea ultrasound machine (Esaote, Milan, Italy) with a 3.5-MHz convex transducer. The ultrasound scans were performed transabdominally when the participants had a full bladder, obtained by voluntary urine retention and oral administration of fluids. Uterine and ovarian volume, endometrial halo, number, diameter and distribution of the follicles were recorded as previously reported5, 19.
Doppler flow measurements of the uterine arteries were performed transabdominally with a 5.0-MHz color Doppler system (AU4 Idea color Doppler; Esaote) as previously reported5, 19. The pulsatility index (PI), defined as the difference between the peak systolic and end-diastolic flow divided by the mean maximum flow velocity, was calculated. For each examination, the mean value of three consecutive waveforms was obtained. No significant differences between the PIs of the left and right uterine arteries were observed, and therefore the average value of both arteries was used. The correlation between PI and heart rate was not tested24. Ultrasound and color Doppler analyses were performed by a single examiner (C.B.) who was unaware of the response to the GnRH and ACTH tests.
Peripheral blood was obtained from all patients between 08.00 and 11.00 h after an overnight fast, on the same day that sonographic and Doppler examinations took place, and different hormonal parameters were analyzed. Plasma concentrations of LH, FSH, estradiol (E2), and testosterone (T) were assayed as previously described25. Dehydroepiandrosterone sulfate (DHEAS), and 17-hydroxy progesterone (17-OH-Pg) were determined by radioimmunoassay (RIA) using the Coat-A-Count kit (DPC; Los Angeles, CA, USA). Androstenedione (A) was measured by RIA with the Quantitative Measurement of Androstenedione in Serum and Plasma kit (DSL Inc.; Webster, TX, USA). All hormone analyses were performed in duplicate.
The GnRH stimulation test was performed using a standard dose of 100 µg GnRH administered as an intravenous (i.v.) bolus. Serum LH and FSH concentrations were measured at 0, 30, 60 and 90 min. For defining a pubertal GnRH test the following criteria had to be fulfilled: a baseline LH value > 0.3 IU/L, a peak LH level > 15 IU/L, a LH/FSH peak ratio > 0.66 and a LH Δ value (peak—basal value) > 7 IU/L14–16. All samples were stored at −20 °C until they were assayed.
The ACTH test was performed with a single i.v. injection of 0.25 mg ACTH (Synachten, Novartis Farma; Origgio, Italy)26. The patients did not receive overnight dexamethasone suppression27. The test was performed between 08.00 and 11.00 h. Serum samples were drawn at −30, baseline, +60, +120 min. All samples were stored at −20 °C until they were assayed for A and 17-OH-Pg.
Results of hormonal values were converted to SI units using the following conversion factors: LH (IU/L) = mIU/mL × 1.0; FSH (IU/L) = mIU/mL × 1.0; E2 (pmol/L) = pg/mL × 3.761; T (nmol/L) = ng/mL × 3.467; A (nmol/L) = ng/dL × 0.0349; 17-OH-Pg (nmol/L) = ng/dL × 0.03026; DHEAS (µmol/L) = µg/mL × 2.714.
Statistical analysis was performed using SPSS software (SPSS Inc., Chicago, IL, USA) and included the Mann–Whitney U-test and one-way analysis of variance. The relationship between the parameters analyzed was assessed using the stepwise linear regression method. A probability of 0.05 was considered as statistically significant. Data are presented as mean ± standard deviation, unless otherwise indicated.
All 69 patients completed the study. The basal plasma LH, FSH, E2, T, A, 17-OH-Pg and DHEAS concentrations are reported in Table 1. Among patients with pubic hair and breast stage II the GnRH stimulation test allowed us to distinguish a prepubertal (Group I) from a pubertal (Group II) response (Table 2). The ACTH stimulation test excluded congenital adrenal hyperplasia in all patients with isolated premature pubarche (Group III). The BMI, the chronological age (CA) and the bone age (BA) are reported in Table 3. The BA/CA ratio was not significantly different among groups. However, in girls with a pubertal response to the GnRH test (Group II), the BA was always >2 years above the CA.
|Group I (n = 17)||Group II (n = 16)||Group III (n = 20)||Group IV (n = 16)||Normal range*||Significance|
|LH (IU/L)||0.15 ± 0.11||2.0 ± 2.1||0.12 ± 0.11||0.12 ± 0.28||<1.5||†|
|FSH (IU/L)||2.32 ± 2.06||5.56 ± 5.51||1.14 ± 0.87||0.51 ± 0.66||<1.5||‡|
|E2 (pmol/L)||45.5 ± 8.2||133.9 ± 85||41.3 ± 11.6||45.5 ± 9.4||<50||§|
|T (nmol/L)||0.43 ± 0.10||1.34 ± 1.0||0.41 ± 0.25||0.84 ± 1.14||0.3–1.3||NS|
|A (nmol/L)||2.35 ± 0.08||2.43 ± 1.54||2.47 ± 1.11||1.95 ± 0.51||1.2–3.5||NS|
|17-OH-Pg (nmol/L)||2.9 ± 0.9||4.11 ± 0.45||3.51 ± 1.66||2.63 ± 1.48||0.75–4.5||NS|
|DHEAS (µmol/L)||2.74 ± 1.68||3.05 ± 0.57||2.82 ± 1.32||1.43 ± 0.56||0.8–3.25||NS|
|Group I (n = 17)||Group II (n = 16)||Significance (P)|
|LH peak (IU/L)||3.1 ± 2.4||18.9 ± 2.1||<0.001|
|FSH peak (IU/L)||7.6 ± 6.8||12.6 ± 4.9||NS|
|LH peak/FSH peak||0.37 ± 0.19||1.47 ± 1.01||<0.001|
|Group I (n = 17)||Group II (n = 16)||Group III (n = 20)||Group IV (n = 16)||Significance|
|Chronological age (years)||6.4 ± 1.7||7.2 ± 0.6||7.0 ± 1.0||6.3 ± 0.9||*|
|Bone age (years)||7.8 ± 1.9||9.7 ± 1.1||8.5 ± 1.6||6.6 ± 0.7||†|
|Body mass index (kg/m2)||16.9 ± 1.7||17.7 ± 2.1||16.7 ± 1.2||18.0 ± 2.8||NS|
None of the studied patients was hirsute. Axillary hair was present in 4/16 girls in Group II and 2/20 girls in Group III. The follow-up at 1 year confirmed the diagnosis in all patients.
The ultrasound assessment allowed measurements of the uterine volume in 100% of the cases. The uterine size was greater in girls who presented a pubertal response to the GnRH test (Group II, 7.48 ± 4.18 mL) than in those who presented a prepubertal response (Group I, 3.02 ± 1.36 mL; P = 0.006), an isolated pubarche (Group III, 2.58 ± 1.32 mL; P < 0.001) or an isolated thelarche (Group IV, 1.82 ± 1.07 mL; P < 0.001). Furthermore, the uterine size was greater in Group I than it was in Group IV (P = 0.043). In addition, the uterine volume exceeded our own normal reference range (1.5–4.0 mL) for the prepubertal stage in 87.5% of girls in Group II, 17.6% of girls in Group I and 10% of the cases in Group III. In girls with isolated thelarche (Group IV), the uterine size never exceeded 4.0 mL in volume.
Endometrial echo was observed in 87.5% of cases in Group II, 29.4% of cases in Group I and 5% of cases in Group III. In none of the patients with an isolated thelarche (Group IV) was an endometrial echo observed.
In 58/69 (84%) participants, both ovaries were visualized. Only one ovary could be visualized in six girls in Group IV and in four girls in Group III and no ovaries were visualized in one patient in Group IV. The ovarian volume was greater in Group II (2.10 ± 0.91 mL) than in Group III (1.15 ± 0.37 mL; P = 0.003) and in Group IV (0.82 ± 0.46 mL; P = 0.001). No significant differences were observed in comparison with patients in Group I (1.90 ± 1.21 mL).
Similarly, the number of small (<1 cm) follicles was higher in girls with a pubertal response to the GnRH test (Group II, 2.8 ± 2.6) than in those with a prepubertal response to the GnRH test (Group I, 1.7 ± 2.3; P = 0.047), with isolated pubarche (Group III, 1.5 ± 1.8; P = 0.036) or with isolated thelarche (Group IV, 0.8 ± 1.0; P = 0.008).
On Doppler analysis, the uterine arteries were adequately visualized in 100% of the cases and the lowest impedance was observed in patients with a pubertal response to the GnRH test (Group II, Figures 1 and 2). A low PI (≤2.5) was considered to be an expression of a rapidly growing uterus5. The presence of low impedance to flow at the level of the uterine arteries had a generally high diagnostic value for precocious puberty, being more accurate than were uterine volume (cut-off value: 4 mL) and endometrial echo (presence/absence) (Table 4).
|Doppler||Uterine volume||Endometrial echo|
|Positive predictive value (%)||88||73||70|
|Negative predictive value (%)||98||96||96|
|Concordance with GnRH test (%)||98.5||93||91|
In the total study population, the uterine volume was directly correlated with FSH (r = 0.475; P = 0.001), LH (r = 0.530; P < 0.001), E2 (r = 0.676; P < 0.001) plasma concentrations and bone age (r = 0.411; P = 0.002). In addition, uterine volume correlated inversely with uterine artery PI (r = −0.529; P < 0.001). Finally, uterine artery PI was inversely correlated with FSH (r = −0.429; P = 0.001), LH (r = −0.507; P = 0.001) and E2 (r = −0.366; P = 0.024) plasma concentrations.
Among sexual precocities only GnRH-dependent puberty requires the GnRH-agonist suppression of the hypothalamic–pituitary–ovarian axis. Thus, in order to avoid unnecessary treatment, it is important to correctly diagnose the activation of the hypothalamic–pituitary–ovarian axis considering that isolated premature thelarche and isolated premature adrenarche may mimic the early features of true precocious puberty and produce diagnostic difficulties.
In the present study, the GnRH stimulation test was administered to distinguish between the intermediate forms6, which are not suitable for treatment with GnRH-agonist, and true precocious puberty. Although the stimulation test is considered as the gold standard, there is no concordance as to the criteria which should be used for the diagnosis6, 10, 11. Even though the criteria used by Oerter-Klein et al.10 reached a good diagnostic value, there was no basal or stimulated single level of LH, FSH or E2 with elevated sensitivity and specificity for the diagnosis of precocious puberty. In our study, to improve the diagnostic value of hormonal evaluation and to allow us to consider the GnRH stimulation test as the gold standard, we adopted very strict cumulative criteria5. Anamnestic, auxological and basal hormonal evaluation allowed us to exclude activation of the hypothalamic–pituitary–ovarian axis in girls with isolated premature thelarche and isolated premature pubarche. In addition, the ACTH test excluded congenital adrenal hyperplasia in all patients with isolated premature pubarche.
Pelvic sonographic evaluation has proved to give accurate and important information about internal genitalia28. In girls being evaluated for any disturbance in sexual maturation, sonography is considered as an adjunctive method in establishing the exact diagnosis of pubertal precocity. In almost all studies it has been found that these girls have enlarged ovaries, an increased uterine volume and the development of a midline endometrial echo3, 12, 13, 15. However, by evaluating the diagnostic value of these sonographic findings it has been underlined that although associated with good specificity, their sensitivity is low due either to the transabdominal approach or to the overlap between normal and pathological values.
It has been shown that the anatomical structure of the pelvic organs could not be adequately assessed in as many as 42% of cases (due to obesity, limited resolution of low-frequency transducers, troublesome full bladder technique and dilated bowel loops)29, 30. However, in our patients we visualized the uterine structures in 100% of cases, both ovaries in 84% of cases and at least one ovary in 99% of cases. As previously stated5, we confirm that the low diagnostic value of sonography is dependent on the operator's skill and the quality of the ultrasound equipment.
In the present paper sonography proved to be a test of high value in the diagnosis of female precocious puberty. Increased uterine volume (>4 mL) and the presence of a midline endometrial echo demonstrated adequate specificity (87% and 86%, respectively), good sensitivity (87.5% for both parameters) and high concordance with the GnRH-stimulation test (93% and 91%, respectively). These data agree with those of Haber et al.15 and show that no significant overlap between true precocious puberty and other premature pubertal anomalies exists. Thus we speculate that normal endometrial echogenicity and normal uterine volume, reflecting an infantile morphology, are useful sonographic parameters to exclude the activation of the hypothalamic–pituitary–gonadal axis. On the other hand, increased uterine volumes associated with an echogenic endometrium indicated the activation of the hypothalamic–pituitary–gonadal axis as demonstrated by the positive correlation between uterine volume and FSH, LH and E2 plasma concentrations. In addition, uterine volume was correlated with bone age. This shows that increased circulating estrogen concentrations determine the growth spurt, skeletal maturation and internal genitalia transformation.
Our data further show that the activation of the hypothalamic–pituitary–gonadal axis is associated with an increased ovarian volume and an increased number of small (<1 cm) ovarian follicles. This agrees with the findings of Salardi et al.28 who stated that if ‘… ovarian size is within the normal range… it is highly likely that patients have adrenarche or premature thelarche’. However, a partial overlap exists between true precocious puberty and isolated premature pubarche or thelarche, making it impossible to define clear cut-off values. This limits the usefulness of measurements of ovarian volume and follicular number in the assessment of pubertal status. Many factors (such as a partial rise of gonadotropin secretion or pulsatility and IGF-1 paracrine action) may, in fact, stimulate ovarian development and follicular selection.
In the present study we introduced a new parameter that could be used to correctly diagnose precocious puberty, namely color Doppler study of uterine arteries. We found that color Doppler analysis of the uterine arteries agreed well with the findings of the GnRH-stimulated test and was, better than sonography without color Doppler, a very useful diagnostic tool to differentiate between true precocious puberty and its intermediate forms. Contrary to the findings of Mosfeldt Laursen et al.17, the uterine artery was visualized in 100% of our population and measurable diastolic flow was observed in 40% of patients with premature thelarche and in almost 90% of patients with other precocities. This disagrees with the findings of Ziereisen et al.18, who found a modification of the Doppler signal pattern during the progression of pubertal maturation: absent diastolic flow in prepubertal girls was progressively replaced by a complete systolic–diastolic flow during puberty.
Uterine artery PIs, which seem to reflect arterial tone or modification in resistance to flow in the vascular bed, were lower in girls with true precocious puberty. This, associated with increased uterine volumes, confirmed that the growing uterus, during the pubertal spurt, may be the object of angiogenesis17, 18. Our data showed an inverse correlation between uterine vascularization and plasma LH concentrations. Elevated LH plasma levels may be responsible for increased vascularization by different mechanisms that may act individually or in a cumulative way: neoangiogenesis, catecholaminergic stimulation and leukocyte and cytokine activation31–33. In addition, in girls with an activated hypothalamic–pituitary–ovarian axis, the effects on uterine vascularization may be enhanced by estrogen activity, as underlined in our study by an inverse correlation between uterine artery PIs and circulating E2 levels. Horowitz and Horowitz34 found receptors for estrogens in the great vessels and the data of other investigators suggest that estrogens affect vessel wall physiology35, 36. Estradiol has been shown to decrease vascular resistances either by exerting a direct action on the smooth muscle cells in the media of the uterine artery vessel wall or by indirectly decreasing the calcium-mediated vessel constriction and/or periarterial sympathetic vasoconstrictor nerve activity. Furthermore, estrogens seem to increase nitric oxide production and ameliorate the plasma viscosity and prostacyclin/thromboxane balance37, 38.
Further longitudinal and extensive studies are necessary for the better understanding of the relationship between the low vascular impedance of the uterine arteries and hormonal modifications due to an activated hypothalamic–pituitary–ovarian axis. However, we conclude that sonographic parameters (uterine volume and presence/absence of endometrial echoes) and color Doppler analysis of uterine arteries may improve the diagnosis of GnRH-dependent precocious puberty and may be useful to determine which girls need treatment.