V. CALCATERRA*, P. SAMPAOLO, C. KLERSY, D. LARIZZA*, A. ALFEI, V. BRIZZI*, F. BENEVENTI and M. CISTERNINO*

Transcription

1 Ultrasound Obstet Gynecol 2009; 33: Published online 11 December 2008 in Wiley InterScience ( DOI: /uog.6271 Utility of breast ultrasonography in the diagnostic work-up of precocious puberty and proposal of a prognostic index for identifying girls with rapidly progressive central precocious puberty V. CALCATERRA*, P. SAMPAOLO, C. KLERSY, D. LARIZZA*, A. ALFEI, V. BRIZZI*, F. BENEVENTI and M. CISTERNINO* *Department of Pediatrics and Department of Obstetrics and Gynecology, University of Pavia and IRCCS Policlinico San Matteo and Biometric Unit, Scientific Direction, IRCCS Policlinico S. Matteo, Pavia, Italy KEYWORDS: breast ultrasound; multivariate model; precocious puberty; premature thelarche risk score ABSTRACT Objective To determine the utility of breast ultrasonography in the diagnostic work-up of precocious puberty and to create a prognostic index for early differentiation between non/slowly progressive or transient forms of precocious puberty and rapidly progressive central precocious puberty. Methods We recruited consecutively 60 girls with precocious pubertal development. In all the girls we evaluated Tanner stage, basal and gonadotropin-releasing hormone (GnRH)-stimulated follicle stimulating hormone (FSH) and luteinizing hormone (LH) levels, estradiol (E2) levels, and bone age, and performed pelvis and breast ultrasound examinations. Logistic regression models were fitted to identify possible diagnostic factors for rapidly progressive central precocious puberty and non/slowly progressive or transient forms. Results Ultrasound breast volume 0.85 cm 3 was associated with rapidly progressive central precocious puberty (P = 0.01). Uterine volume 5cm 3,LHpeak 7 IU/L, presence of an endometrial echo, E2 levels 50 pmol/l and bone age > 2 SD above expected were significantly associated with rapidly progressive central precocious puberty. A multivariate model including uterine volume, E2 level, bone age, presence of an endometrial echo and ultrasound breast volume revealed a strong ability to classify rapidly progressive forms. From this multivariate analysis a prognostic index for rapidly progressive central precocious puberty was defined. Conclusions Ultrasound imaging allows better definition of the breast and the maturation stage than does use of Tanner s stages. Ultrasound breast volume 0.85 cm 3 is an independent predicting factor of rapidly progressive central precocious puberty. A prognostic index that was created from a multivariate model including uterine volume, E2 level, presence of an endometrial echo, bone age and ultrasonographically determined breast volume, may help in the early differentiation between rapidly progressive central precocious puberty and non/slowly progressive or transient forms. Copyright 2008 ISUOG. Published by John Wiley & Sons, Ltd. INTRODUCTION Precocious pubertal development is generally defined as the appearance of secondary sex characteristics before the age of 8 years in girls and before the age of 9 years in boys 1. In girls, precocious puberty consists of a spectrum of disorders, including the non/slowly progressive or transient forms of precocious puberty characterized by stabilization/slow progression or regression of pubertal signs and the rapidly progressive central precocious puberty, in which pubertal development is highly accelerated and proceeds much faster than in normal puberty, with the deterioration of final height potential. The non/slowly progressive or transient forms are considered to be a benign entity and they do not require any treatment, but rapidly progressive central precocious puberty may be of idiopathic origin or secondary to central nervous system (CNS) lesions 2 5. When Correspondence to: Dr V. Calcaterra, Department of Pediatrics, Fond. IRCCS Policlinico S. Matteo, P.le Golgi, 2, Pavia, Italy ( v.calcaterra@smatteo.pv.it) Accepted: 25 July 2008 Copyright 2008 ISUOG. Published by John Wiley & Sons, Ltd. ORIGINAL PAPER

2 86 Calcaterra et al. diagnosed, rapidly progressive forms require magnetic resonance imaging (MRI) to check for CNS lesions, and gonadotropin-releasing hormone analog (GnRH-a) suppressive therapy. It may be difficult to distinguish between non/slowly progressive or transient forms and rapidly progressive forms, but the early recognition of rapidly progressive central precocious puberty is essential. The evaluation of breast development in patients with precocious puberty is usually performed by means of clinical examination using Tanner s stages 6. It is often difficult to obtain a correct assessment of the glandular tissue using these stages because it is not possible to distinguish mammary tissue from the surrounding loose, fat and connective tissue. Breast ultrasonography has not yet been used as a diagnostic procedure in the investigation of precocious puberty in girls, but it allows us to obtain an accurate morphostructural and volumetric study of the developing breast. An ultrasound grading of mammary development was proposed by Bruni et al. 7, in which five stages that are closely related to increasing estrogen production were described. The aim of this study was to determine the utility of the ultrasonographic evaluation of breast tissue, using a multivariate model, in the diagnostic work-up of precocious puberty. This model was used to create a prognostic index in order to identify as early as possible patients with complete progressive forms of precocious puberty who need further investigation, including brain MRI and GnRH-a suppressive therapy. PATIENTS AND METHODS We recruited consecutively 60 girls referred to our institution by their general practitioner or primary care pediatric consultant between 2002 and 2003 for evaluation because of the appearance of premature sexual maturation between the ages of 3 and 8 years. None of the patients had neurological signs or symptoms, or known CNS pathologies or exposure to exogenous androgens or estrogens. In all the patients diagnostic evaluation included measurement of height and weight, body mass index (BMI), physical examination with puberty rating, bone age assessment, measurement of basal and GnRH-stimulated follicle stimulating hormone (FSH) and luteinizing hormone (LH) levels and of serum estradiol (E2) levels, uterine and ovarian measurements by pelvic ultrasound and ultrasound assessment of breast volume and morphology. Evaluation of the pubertal development progression rate was performed after at least 6 months of follow-up by an experienced clinician (M. C.). Twenty-five girls were classified as having rapidly progressive central precocious puberty (Group 1) and the remaining 35 as having non/slowly progressive or transient precocious puberty (Group 2). The diagnosis of rapidly progressive central precocious puberty was based on the appearance of breast buds before 8 years of age accompanied by the presence of one or more of the following findings: menses, pubic hair, accelerated growth velocity or bone age greater than 2 SD above chronological age. In all the patients in Group 1 a rapid progression of pubertal stage was noted during the follow-up. The 35 patients of Group 2 with the diagnosis of non/slowly progressive or transient precocious puberty included 14 girls with development of breast tissue before the age of 8 without any other sign of sexual maturation; five patients with breast buds and bone age advancement and/or elevated height velocity without progression of breast development during follow-up; 12 patients with breast buds with pubic or axillary hair in the absence of bone or growth spurt and without progression of pubertal signs during follow-up; and four patients with breast buds with pubic or axillary hair and bone age advancement and/or elevated height velocity at the first evaluation, but with regression of breast development and no bone age progression during follow-up. All the diagnoses were confirmed after at least 2 years of clinical and auxological follow-up and the clinical, hormonal and ultrasound data of the two groups at the time of first evaluation were statistically compared. All the girls with rapidly progressive central precocious puberty underwent MRI of the CNS including the hypothalamic pituitary region. No abnormalities of this region were detected. Height was measured using a Harpenden stadiometer. Height and growth velocity were expressed in standard deviation scores (SDS, i.e. Z-scores) for chronological age 8. Puberty was assessed according to Marshall and Tanner 6. In the event of asymmetric breast development, the patient was considered as belonging to the higher clinical stage. Target height, expressed in SDS, was calculated from midparental heights 8. Body mass index (BMI) was calculated as body weight in kilograms divided by body height (m) squared, and the BMI-SDS was calculated according to the method of Cacciari et al. 9. Bone age assessment was determined according to the method of Greulich and Pyle, and expressed in SD for chronological age 10. Serum FSH, LH and E2 levels were determined by chemiluminescence immunometric assay (Siemens Medical Solutions Diagnostics, Milan, Italy). Pelvic ultrasound examination was performed according to the conventional full-bladder technique using an Aloka Prosound SSD 5500 machine (Aloka, Tokyo, Japan) with a MHz convex transducer. We evaluated uterine and ovarian volume (calculated using the ellipsoid formula V = D1 D2 D , where D1 is the largest longitudinal diameter, D2 the largest anteroposterior diameter and D3 the largest transverse diameter), endometrial echo visualization and presence of measurable follicles. Breast ultrasound examination was performed using an Aloka Prosound SSD 5500 machine with a 10-MHz linear transducer. We evaluated breast volume using the ellipsoid formula V = D1 D2 D , where the maximum perpendicular anteroposterior dimension (D1) and mediolateral dimension (D2) were measured on a transverse image of the largest diameter and the

3 Breast ultrasound in precocious puberty 87 Figure 1 Ultrasound images showing the morphology of the posterior profile of the breast as classified by Bruni et al. 7, who described five morphological stages of glandular development: (a) Stage A, (b) Stage B, (c) Stage C, (d) Stage D and (e) Stage E. maximum craniocaudal diameter (D3) was measured on a longitudinal image. Evaluation of the diameters of the breast was performed without pressing the breast tissue with the transducer. The morphology of the posterior profile of the breast was assessed according to the method of Bruni et al. 7, who described five morphological stages of glandular development (Figure 1): Stage A, absence of glandular bud; Stage B, first appearance of the glandular bud; Stage C, growing glandular bud which displays a linear edge, roundish shape and appears quite distinct from the surrounding connective-adipose tissue; Stage D, branching glandular bud in which ramifications appear stemming from the periphery of the bud into the stromal matrix; and Stage E, a triangular phase whose glandular structure in its final phase of growth has a pyramidal shape, with an upward apex, which appears triangular on sonographic imaging. Since breast development can be asymmetric, each gland was evaluated separately. Pelvic and breast ultrasound examinations were performed by the same observer. Informed consent was obtained from each parent or guardian and the study protocol was approved by the ethical committee of our institution. Statistical analysis Continuous variables were reported as mean and SD or median and 25 th 75 th percentile if the distribution was skewed, and categorical variables as absolute and relative (%) frequencies. A general linear model (with robust standard errors calculation) was used to compare breast volume across the stages defined by their appearance on ultrasound. Logistic regression models (with robust standard errors calculation) were fitted to identify possible factors useful for predicting rapidly progressive central precocious puberty as compared to non/slowly progressive or transient forms. Odds ratios (OR) with their 95% CI were reported. All common risk factors with a P < 0.1 on univariate analysis were included, after checking for collinearity, in a multivariate model to assess their independent prognostic value. The same risk factors were used to control the role of mammary gland volume in a second and final model. Model-based sensitivity and specificity were computed, together with model discrimination ability and model calibration, as measured respectively by the c statistic and the shrinkage coefficient (the closer to 1, the better). For the purpose of the analysis, continuous variables were dichotomized according to cutoffs from the literature, for LH peak 11,12,E2 13,uterine volume, ovarian volume and largest follicle diameter or to the median value of their distribution (for FSH and mammary gland volume); for bone age the chosen cutoff corresponded to 2 SD above the expected values for age 9. The repeatability of ultrasound measurements was assessed by calculating the Lin s concordance correlation coefficient in 11 girls. Intraobserver (R = 0.993, P < ) and interobserver (R = 0.988, P < ) correlation coefficients were very good. Stata 8 (Stata Corp, College Station, TX, USA) was used for computation, and a two-sided P-value < 0.05 was considered statistically significant. Bonferroni correction was used for post-hoc comparisons. RESULTS Auxological parameters Clinical and auxological characteristics of the two groups of patients, at the time of first evaluation, are reported in Table 1. As for auxological data, we did not find any significant difference between Group 1 and Group 2. Patients of Group 1 showed a higher bone age-sds than those of Group 2. Group 2 had a lower degree of breast maturation than Group 1. Hormonal parameters Median basal LH value was 0.60 IU/L (25 th 75 th centile, IU/L). No statistical difference was found between the mean basal LH levels of Group 1 and Group 2.The median values for peak serum LH

4 88 Calcaterra et al. Table 1 Clinical and auxological characteristics of patients with rapidly progressive central precocious puberty (Group 1) and with non/slowly progressive or transient precocious puberty (Group 2), at time of first evaluation Parameter Group 1 (n = 25) Group 2 (n = 35) P Age at appearance of 7.14 ± ± secondary sex characteristics (years) Height-SDS 2.16 ± ± Target height (cm) ± ± Target height-sds 0.27 ± ± Growth velocity-sds 3.32 ± ± BMI-SDS 0.82 ± ± Bone age-sds 1.95 ± ± Tanner stage (n) B * B * B * Continuous variables are expressed as mean ± SD. *Fisher s exact test; all other comparisons were with Student s t-test. BMI, body mass index; SDS, standard deviation score. and FSH levels on GnRH-stimulation test were 5.63 IU/L (25 th 75 th centile, IU/L) and 9.93 IU/L (25 th 75 th centile, IU/L), respectively. Median E2 levels were pmol/l (25 th 75 th centile, pmol/l). As shown in Table 2, a peak serum level of LH 7 IU/L and an E2 level 50 pmol/l were significantly associated with rapidly progressive central precocious puberty (P < and P = 0.004, respectively) with a fair discrimination ability as measured by the c statistic (c = 0.94 and c = 0.68, respectively). The sensitivities of LH peak and E2 levels were 88 and 58.3%, with specificities of 100 and 79.4%, respectively. The peak serum FSH level was not related to rapidly progressive central precocious puberty. In three patients in Group 2, the GnRH test was interrupted for technical reasons. Pelvic ultrasound Overall, the enrolled patients had a median uterine volume of 3.60 cm 3 (25 th 75 th centile, cm 3 ) and endometrium was present in 25 girls. Ovarian median volume was 2.11 cm 3 (25 th 75 th centile, cm 3 ) and the median dominant follicle diameter was 6.80 mm (25 th 75 th centile, mm). As shown in Table 2, uterine volume 5cm 3 and presence of an endometrial echo were significantly associated with rapidly progressive central precocious puberty (P = and P = 0.004, respectively), with fair discrimination ability (c = 0.66 and c = 0.69, respectively). Ultrasound measurement of uterine volume had a sensitivity of 52.0% and a specificity of 80.80% (cut-off value, 5 cm 3 ), while the presence of an endometrial echo had a sensitivity of 64.0% and a specificity of 74.2%. Diameter of the dominant follicle 1 cm and ovarian volume 2cm 3 were not related to rapidly progressive central precocious puberty. Breast ultrasound The median value for breast volume was 0.85 cm 3 (25 th 75 th centile, cm 3 ). Breast volume 0.85 cm 3 was significantly associated with rapidly progressive central precocious puberty (P = 0.01) with fair discrimination ability (c = 0.64) (Table 2). The ultrasound measurement of breast volume had a sensitivity of 66.0% and a specificity of 61.7% (cutoff value, 0.85 cm 3 ). The increase in breast volume was significantly associated with a more progressive ultrasound stage (regression model P overall < 0.001; Stage B vs. Stage C, P < 0.001; Stage C vs. Stage D, P < 0.001; Stage D vs. Stage E, P < 0.001). Ultrasound breast stage was only marginally associated with rapidly progressive central precocious puberty (P = 0.04). Breast ultrasound morphology, in relation to Tanner stage in all patients, is reported in Figure 2. Table 2 Hormonal and ultrasound data in girls with rapidly progressive central precocious puberty (Group 1) and with non/slowly progressive or transient precocious puberty (Group 2) Variable Group 1 (n (%)) Group 2 (n (%)) OR (95% CI) P Discrimination (c statistic) Peak serum LH < 7 IU/L 3 (8.57) 32 (91.43) Peak serum LH 7 IU/L 22 (100) 0 (0.00) < E2 level < 50 pmol/l 10 (27.03) 27 (72.97) 1 E2 level 50 pmol/l 14 (66.67) 7 (33.33) 5.40 ( ) Uterine volume < 5 cm 3 12 (30.00) 28 (70.00) 1 Uterine volume 5 cm 3 13 (65.00) 7 (35.00) 4.33 ( ) Absence of an endometrial echo 9 (25.71) 26 (74.29) 1 Presence of an endometrial echo 16 (64.00) 9 (36.00) 5.13 ( ) Breast volume < 0.85 cm 3 17 (28.81) 42 (71.19) 1 Breast volume 0.85 cm 3 33 (55.93) 26 (44.07) 3.13 ( ) Bone age 2 SD above expected 7 (21.12) 26 (78.79) 1 Bone age > 2 SD above expected 18 (66.67) 9 (33.33) 7.42 ( ) Associations between patient characteristics and rapidly progressive central precocious puberty are shown as the results of univariate models. Some variables had missing data. Percentages represent girls with the characteristic as a proportion of the total number with that characteristic in both groups. E2, estradiol; LH, luteinizing hormone; OR, odds ratio.

5 Breast ultrasound in precocious puberty 89 (a) Percent (b) Percent A B C D E same model. Of these two parameters, E2 level was retained owing to the higher discriminative ability of the respective multivariate model, and because it is a hormonal parameter that can be obtained more easily in clinical practice than measures that require a GnRHstimulation test. E2 levels, bone age advancement and presence of an endometrial echo were independent predictors of rapidly progressive central precocious puberty (P = 0.001, P < and P = 0.034, respectively) (Table 3, Model 1). Model sensitivity and specificity were 75.0% and 85.3%, respectively. When adding breast volume to the model, this also proved to predict rapidly progressive central precocious puberty independently (P = 0.012), with a further increase in discriminating ability (as measured by the c statistic) from 0.84 to 0.86 (Table 3, Model 2); model sensitivity and specificity were 77.1% and 83.3%, respectively Prognostic index (c) Percent A B C D E Model 2 was used to design a prognostic index. Each coefficient estimated by the model (log OR) was rounded to the closest unit or half unit. Covariates were coded as 1 if present ( cut-off) and 0 if absent (< cut-off). Each covariate was multiplied by its new coefficient and the sum of these values was calculated, as illustrated by the following algorithm: RPCPP risk = [0.5 (1 when uterine volume 5cm or 0 when uterine volume < 5cm 3 )] 0 + [2.5 (1 whene2level 50 pmol/l or 0 when A B C D E Ultrasound stage E2 level < 50 pmol/l)] Figure 2 Percentage distribution of ultrasound breast stages within TannerstagesB2(a),B3(b)andB4(c)ingirlswithprecocious puberty. A weak, although statistically significant, association between breast ultrasound morphology and Tanner stage was observed (Spearman s rho = 0.54; P < 0.001). Bone age Bone age advancement was > 2 SD from the expected value in 27 of the 60 patients, and this parameter was significantly associated with rapidly progressive central precocious puberty (P = 0.001). Evaluation of bone age had a sensitivity of 72.0% and a specificity of 74.3% (cut-off value, 2 SD). Multivariate models Uterine volume, peak serum LH, bone age advancement, presence of an endometrial echo and E2 levels were chosen as requisites for inclusion in a multivariate model. However, owing to the presence of collinearity, peak serum LH and E2 levels could not be included in the + [2 (1 when bone age > 2SD or 0 when bone age 2SD above expected)] + [0.5 (1 when presence of endometrial thickness or 0 when absence of endometrial thickness)] + [1.5 (1 when breast volume 0.85 cm 3 or 0 when breast volume < 0.85cm 3 )] (where RPCPP = rapidly progressive central precocious puberty). The risk score ranges from 0 to 7, with higher values corresponding to a higher probability of rapidly progressive central precocious puberty. Three risk groups (low-, intermediate- and high-risk) for rapidly progressive central precocious puberty were defined based on the tertiles of the distribution of this index. The relationship between risk score and probability of rapidly progressive central precocious puberty is summarized in Table 4. Using this prognostic index the majority (62.5%) of patients with rapidly progressive central precocious puberty showed a high risk score, while most patients with non/slowly progressive or transient forms showed a low risk score (57.6%).

6 90 Calcaterra et al. Table 3 Multivariate models showing the association between patient characteristics and rapidly progressive central precocious puberty Model 1 (LR Chi , P < , c statistic 0.84, shrinkage 0.84) Model 2 (LR Chi , P < , c statistic 0.86, shrinkage 0.85) Parameter OR (95% CI) P OR (95% CI) P Uterine volume 5 cm ( ) ( ) 0.75 E2 level 50 pmol/l 5.15 ( ) ( ) < Bone age > 2 SD above expected 6.24 ( ) < ( ) < Presence of an endometrial echo 2.98 ( ) ( ) 0.34 Breast volume 0.85 cm ( ) E2, estradiol; LR, likelihood ratio; OR, odds ratio. Table 4 Calculated risk score and corresponding risk groups for rapidly progressive central precocious puberty (RPCPP), showing proportion of patients with RPCPP (Group 1) and with non/slowly progressive or transient precocious puberty (Group 2) categorized into each of the three risk groups Risk group RPCPP risk score (range) Probability of RPCPP (95% CI) (%) Group 1 (n (%)) Group 2 (n (%)) Low risk ( ) 3 (12.5) 19 (57.6) Intermediate risk ( ) 6 (25) 11 (33.3) High risk ( ) 15 (62.5) 3 (9.1) Groups 1 and 2 total 24 and 33, respectively, due to missing data. DISCUSSION Premature sexual maturation is characterized by a wide spectrum of disorders of gonadal maturation ranging from premature thelarche to rapidly progressive central precocious puberty 2,5. These disorders have received different names such as thelarche variant 3, exaggerated thelarche 17, unsustained 18 and slowly progressive precocious puberty 19. All of these variants may be called non/slowly progressive or transient forms to differentiate them from complete rapidly progressive central precocious puberty. The distinction between non/slowly progressive or transient forms and complete rapidly progressive central precocious puberty is difficult because initially the two forms are clinically alike. Moreover, premature thelarche has a high incidence of progression to complete precocious puberty 20,21. Differentiation between non/slowly progressive or transient forms and rapidly progressive central precocious puberty is important for prognostic and therapeutic purposes. Patients with rapidly progressive central precocious puberty may require neuroimaging to detect occult intracranial lesions 4,22 and also need to be treated with GnRH-analog suppressive therapy, while patients with non/slowly progressive or transient forms do not. Breast ultrasonography has not yet been used in the diagnostic work-up for precocious puberty. Until now, the assessment of breast development has usually been made using Tanner s staging alone. This is often unable to define the presence of mammary parenchyma, mainly in obese girls, and the degree of estrogen-induced breast maturation. An ultrasound grading of breast growth and maturation was proposed by Bruni et al. 7 ; five stages, from A to E, are recognized on ultrasound examination in normal girls. Progression throughout these stages is positively correlated with the rise in estrogen levels 7. Although ultrasound breast stages were significantly correlated with Tanner stages in this study, we found that Tanner stages showed sonographically different degrees of maturation. In fact, in Tanner stage B2, which corresponds with the initial pubertal maturation, an ultrasound grading of D or E may be found, indicating a high degree of estrogen exposure. Conversely, in Tanner stage B3 or B4, ultrasound stages corresponding to a lower degree of breast maturation may be observed. In our study morphological staging of breast development assessed on ultrasonography demonstrated only a fair ability to distinguish girls with rapidly progressive central precocious puberty from those with non/slowly progressive or transient forms. This is probably because of the fact that the staging was only performed once, at the point when the patient was enrolled into the study. A longitudinal study monitoring the progression rate through breast stages would be helpful in identifying rapidly progressive central precocious puberty, in which pubertal development is highly accelerated and proceeds much faster than in normal puberty. Ultrasound measurement of breast volume was, however, found to be a helpful tool in the diagnostic workup of precocious puberty, as breast volume 0.85 cm 3 was significantly associated with rapidly progressive central precocious puberty in our cases. In addition to the GnRH stimulation test, in our study the main predicting factors for rapidly progressive

7 Breast ultrasound in precocious puberty 91 central precocious puberty were bone age advancement > 2 SD above expected, E2 levels 50 pmol/l, uterine volume 5cm 3, the presence of an endometrial echo and ultrasound breast volume 0.85 cm 3. A multivariate model including all these parameters showed a high discriminating ability for detecting rapidly progressive central precocious puberty, with a sensitivity and a specificity of 77.1% and 83.3%, respectively. This multivariate model was used to define a risk score for rapidly progressive central precocious puberty, which is easy to compute and helpful in the early identification of girls with rapidly progressive puberty. However, a prospective study is needed to validate the utility of this prognostic index. In conclusion, breast morphology as assessed on ultrasound examination reflects breast maturation better than Tanner staging, but it was unable, at least in this study, to discriminate non- or slow-progressive from rapidly progressive central precocious puberty. Conversely, breast volume seems to be associated with rapidly progressive forms. We believe that breast ultrasonography could be included in the diagnostic work-up of precocious puberty as it is non-invasive and well tolerated. The main predicting factors for rapidly progressive central precocious puberty other than ultrasonographically determined breast volume in our cases were E2 levels, uterine volume, bone age and the dominant follicle diameter. A risk score that has been designed using all of these parameters could serve as a useful tool in making the detection of girls with rapidly progressive central precocious puberty easier. REFERENCES 1. Grumbach MM, Styne DM. Puberty: Ontogeny, neuroendocrinology, physiology, and disorders. In Williams Textbook of Endocrinology (9 th edn), Wilson JD, Foster DW, Kronenberg HM, Larsen PR (eds). W. B. Saunders: Philadelphia, PA, 1998; Pescovitz OH, Hench KD, Barnes KM, Loriaux DL, Cutler GB Jr. Premature thelarche and central precocious puberty: the relationship between clinical presentation and the gonadotropin response to luteinizing hormone-releasing hormone. J Clin Endocrinol Metab 1988; 67: Stanhope R, Brook CG. Thelarche variant: a new syndrome of precocious sexual maturation? Acta Endocrinol 1990; 123: Cisternino M, Arrigo T, Pasquino AM, Tinelli C, Antoniazzi F, Beduschi L, Bindi G, Borrelli P, De Sanctis V, Farello G, Galluzzi F, Gargantini L, Lo Presti D, Sposito M, Tato L. Etiology and age incidence of precocious puberty in girls: A multicentric study. J Pediatr Endocrinol Metab 2000; 13 (Suppl 1): Partsch CJ, Heger S, Sippel WG. Management and outcome of central precocious puberty. Clin Endocrinol 2002; 56: Marshall WA, Tanner JM. Variations in patterns of pubertal changes in girls. Arch Dis Child 1969; 44: Bruni V, Dei M, Deligeoroglou E, Innocenti P, Pandimiglio A, Magini A, Bassi F. Breast development in adolescent girls. Adolesc Pediatr Gynecol 1990; 3: Tanner JM, Whitehouse RH, Takaishi M. Standards from birth to maturity for height, weight, height velocity, and weight velocity: British children. Arch Dis Child 1965; 41: Cacciari E, Milani S, Balsamo A, Spada E, Bona G, Cavallo L, Cerutti F, Gargantini L, Greggio N, Tonini G, Cicognani A. Italian cross-sectional growth charts for height, weight and BMI(2to20yr). J Endocrinol Inv 2006; 29: Greulich WW, Pyle SI. Radiographic atlas of skeletal development of the hand and wrist (2 nd edn). Stanford University Press: Stanford, Iughetti L, Predieri B, Ferrari M, Gallo C, Livio L, Milioni S, Forese S, Bernasconi S. Diagnosis of central precocious puberty: endocrine assessment. J Pediatr Endocrinol Metab 2000; 13 (Suppl 1): Eckert KL, Wilson DM, Bachrach LK, Anhalt H, Habiby RL, Olney RC, Hintz RL, Neely EK. A single-sample, subcutaneous gonadotropin-releasing hormone test for central precocious puberty. Pediatrics 1996; 97: Chalumeau M, Hadjiathanasiou CG, Ng SM, Cassio A, Mul D, Cisternino M, Partsch CJ, Theodoridis C, Didi M, Cacciari E, Oostdijk W, Borghesi A, Sippell WG, Breart G, Brauner R. Selecting girls with precocious puberty for brain imaging: validation of European evidence-based diagnosis rule. JPediatr 2003; 143: Haber HP, Wollmann HA, Ranke MB. Pelvic ultrasonography: early differentiation between isolated premature thelarche and central precocious puberty. Eur J Pediatr 1995; 154: Jensen AM, Brocks V, Holm K, Laursen EM, Müller J. Central precocious puberty in girls: internal genitalia before, during, and after treatment with long-acting gonadotropin-releasing hormone analogues. JPediatr1998; 132: Cisternino M, Sampaolo P, Alfei A, Bassi L. Ecografia pelvica in età pediatrica e adolescenziale. In Quaderni di endocrinologia pediatrica, Tonini G (ed). Archimedica: Torino, 2000; 12: Garibaldi LR, Aceto T Jr, Weber C. The pattern of gonadotropin and estradiol secretion in exaggerated thelarche. Acta Endocrinol 1993; 128: Schwartz HP, Zuppinger K. Unsustained central sexual precocity in girls. Pediatr Res 1986; 20: Fontoura M, Brauner R, Prevot C, Rappaport R. Precocious puberty in girls: early diagnosis of a slowly progressing variant. Arch Dis Child 1989; 64: Pasquino AM, Pucarelli I, Passeri F, Segni M, Mancini MA, Municchi G. Progression of premature thelarche to central precocious puberty. JPediatr1995; 126: Stanhope R. Premature thelarche: clinical follow-up and indication for treatment. J Pediatr Endocrinol Metab 2000; 13 (Suppl 1): Stanhope R. Gonadotrophin-dependent [correction of dependant] precocious puberty and occult intracranial tumors: which girls should have neuro-imaging? JPediatr2003; 14: