The Diagnostic Value of Pelvic Ultrasound in Girls with Central Precocious Puberty

Abdominal imaging in precocious puberty in girls: can imaging determine onset of puberty?

Precocious puberty in girls is defined by the appearance of secondary sex characteristics before the age of 8 years. Precocious puberty is either peripheral or central. In most cases, the cause of central precocious puberty is unknown. Rapidly progressive forms of central precocious puberty may benefit from puberty suppression to prevent compromised final height and psychosocial problems related to the physical changes. Idiopathic central precocious puberty in girls can be a challenging diagnosis. Clinical examination, evaluation of growth charts, bone age, and hormonal tests are part of the workup. The role of pelvic ultrasound to evaluate pubertal features of the internal genitalia in this context is controversial. In this paper, we have reviewed the literature to find the scientific bases for several proposed ultrasound parameters and cutoff values for the determination of onset of puberty in girls with suspected central precocious puberty. There are no reliable cutoffs for the measurements of uterus and ovaries for the diagnosis of central precocious puberty and there is overlap in the appearance of the organs in both prepubertal and pubertal girls. The interpretation of the findings on pelvic ultrasound must be done with caution and always in close communication with the referring clinicians.

Pelvic ultrasound and pubertal attainment in girls with sexual precocity: the pivotal role of uterine volume in predicting the timing of menarche

Introduction

Among girls assessed for pubertal precocity, pelvic ultrasound (pUS) may represent a pivotal tool to predict the time expected to elapse between sonographic assessment and the onset of menarche (TUS-M). Accordingly, the present analysis is meant to define the statistical relationship between sonographic parameters and TUS-M, in order to identify the most reliable predictor of the timing of menarche.

Methods

Retrospective, multicenter analysis. Girls assessed for sexual precocity and showing sonographic and clinical findings consistent with pubertal onset upon referral were considered eligible. Patients treated with GnRH analogues were excluded and only those who had subsequently achieved complete and spontaneous pubertal attainment and for whom the exact date of menarche was available were included. Overall, we enrolled 184 girls from five tertiary care Italian Centers.

Results

The time elapsed (months) between baseline endocrine assessment and spontaneous achievement of menarche showed a negative statistically significant correlation (p<0.0001) with LH (r:-0.61), FSH (r:-0.59), estradiol (r:-0.52) and stimulated LH values (r:-0.58). Among pUS parameters, ovarian volume (r:-0.17 left, -0.30 right) and uterine body-to-cervix ratio (r:-0.18) poorly correlated with TUS-M, while uterine diameters (r:-0.61 longitudinal, -0.64 anteroposterior) and volume (r:-0.70) achieved a highly statistical significance (p<0.0001). Uterine volume (UV) showed a negative logarithmic relationship with TUS-M and represented the most reliable predictor of the timing of menarche in uni- and multivariable analyses (p <0.001). ROC analyses identified the UV thresholds that best predict the onset of menarche within 18, 12 and 6 months, respectively: 3.76, 6.02 and 8.80 ml.

Conclusion

The logarithm of UV shows the best statistical performance in predicting the timing of menarche in girls assessed for pubertal precocity. Accordingly, we developed a user-friendly online application that provides clinicians with an estimation of the months expected to elapse before menarche, based on the UV recorded upon pUS.

Diagnosis of Central Precocious Puberty

A thorough history and physical examination including Tanner staging and growth assessments can guide differential diagnosis and aid in the evaluation of precocious puberty. Basal luteinizing hormone levels measured using a highly sensitive assay can be helpful in diagnosing central precocious puberty (CPP). Brain MRI is indicated with males diagnosed with CPP and females under the age of 6 with CPP. As more information becomes available regarding the genetic etiologies of CPP, genetic testing may preclude the need for imaging studies and other hormonal testing, especially in familial cases.

Diagnostic model based on multiple factors for girls with central precocious puberty

OBJECTIVES

The GnRH stimulation test has been used as the gold standard for the diagnosis of central precocious puberty (CPP), but it has some practical barriers. This study intends to build a diagnostic model of CPP in girls based on the population in northern China.

METHODS

A total of 163 girls with precocious puberty (PP) were included from December 2018 to December 2019. Multifactor logistic regression analysis was conducted. Based on the results of multivariate logistic regression analysis, a nomogram was established for clinical application.

RESULTS

A multi logistic regression model showed that LH (OR=1.238, 95 % CI: 1.067-1.436, p=0.005), inhibin B (OR=1.066, 95 % CI: 1.032-1.100, p<0.001), bone age (OR=1.563, 95 % CI: 1.037-2.358, p=0.033), and uterine length (OR=1.180, 95 % CI: 1.034-1.348, p=0.014) were diagnostic factors for CPP. The prediction model AUC was 0.906 (95 % CI: 0.852-0.959, p<0.001).

CONCLUSIONS

We successfully developed a nomogram model for CPP patients based on clinical data. The diagnostic prediction model included four indicators: basal LH, inhibin B, bone age, and uterine body length.

Basal gonadotropin levels combine with pelvic ultrasound and pituitary volume: a machine learning diagnostic model of idiopathic central precocious puberty

OBJECTIVE

The current diagnosis of central precocious puberty (CPP) relies on the gonadotropin-releasing hormone analogue (GnRHa) stimulation test, which requires multiple invasive blood sampling procedures. The aim of this study was to construct machine learning models incorporating basal pubertal hormone levels, pituitary magnetic resonance imaging (MRI), and pelvic ultrasound parameters to predict the response of precocious girls to GnRHa stimulation test.

METHODS

This retrospective study included 455 girls diagnosed with precocious puberty who underwent transabdominal pelvic ultrasound, brain MRI examinations and GnRHa stimulation testing were retrospectively reviewed. They were randomly assigned to the training or internal validation set in an 8:2 ratio. Four machine learning classifiers were developed to identify girls with CPP, including logistic regression, random forest, light gradient boosting (LightGBM), and eXtreme gradient boosting (XGBoost). The accuracy, sensitivity, specificity, positive predictive value, negative predictive value, area under receiver operating characteristic (AUC) and F1 score of the models were measured.

RESULTS

The participates were divided into an idiopathic CPP group (n = 263) and a non-CPP group (n = 192). All machine learning classifiers used achieved good performance in distinguishing CPP group and non-CPP group, with the area under the curve (AUC) ranging from 0.72 to 0.81 in validation set. XGBoost had the highest diagnostic efficacy, with sensitivity of 0.81, specificity of 0.72, and F1 score of 0.80. Basal pubertal hormone levels (including luteinizing hormone, follicle-stimulating hormone, and estradiol), averaged ovarian volume, and several uterine parameters were predictors in the model.

CONCLUSION

The machine learning prediction model we developed has good efficacy for predicting response to GnRHa stimulation tests which could help in the diagnosis of CPP.

Clinical characteristics and genetic expansion of 46,XY disorders of sex development children in a Chinese prospective study

Diagnosis and management strategy of disorders of sex development (DSD) are difficult and various due to heterogeneous phenotype and genotype. Under widespread use of genomic sequencing technologies, multiple genes and mechanisms have been identified and proposed as genetic causes of 46,XY DSD. In this study, 178 46,XY DSD patients were enrolled and underwent gene sequencing (either whole-exome sequencing or targeted panel gene sequencing). Detailed clinical phenotype and genotype information were summarized which showed that the most common clinical manifestations were micropenis (56.74%, 101/178), cryptorchidism (34.27%, 61/178), and hypospadias (17.42%, 31/178). Androgen synthesis/action disorders and idiopathic hypogonadotropic hypogonadism were the most frequent clinical diagnoses, accounting, respectively, for 40.90 and 21.59%. From all next-generation sequencing results, 103 candidate variants distributed across 32 genes were identified in 88 patients. The overall molecular detection rate was 49.44% (88/178), including 35.96% (64/178) pathogenic/likely pathogenic variants and 13.48% (24/178) variants of uncertain significance. Of all, 19.42% (20/103) variants were first reported in 46,XY DSD patients. Mutation c.680G>A (p.R227Q) on SRD5A2 (steroid 5-alpha-reductase 2) (36.67%, 11/30) was a hotspot mutation in the Chinese population. Novel candidate genes related to DSD (GHR (growth hormone receptor) and PHIP (pleckstrin homology domain-interacting protein)) were identified. Overall, this was a large cohort of 46,XY DSD patients with a common clinical classification and phenotype spectrum of Chinese patients. Targeted gene panel sequencing covered most of the genes contributing to DSD, whereas whole-exome sequencing detected more candidate genes.

Accuracy of doppler assessment of the uterine arteries for the diagnosis of pubertal onset in girls: a scoping review

The uterine artery pulsatility index (PI) assessed by Doppler ultrasound reflects the impedance to the blood flow in the vessel distal to the sampling point. We aimed to assess the accuracy of the uterine artery PI for the diagnosis of puberty in girls. A PRISMA-ScR-compliant scoping review was performed in the MEDLINE and Embase databases with the search terms "puberty" and "Doppler ultrasonography". Studies that included girls aged 0-18 years who underwent pelvic Doppler ultrasound with calculation of uterine artery PI were eligible. Ten studies comprising 1385 girls aged 1.2-18 years were included. The selected studies included participants from Italy, Brazil, Iran, Belgium and Denmark, and were published between 1996 and 2021. Six studies selected girls who were referred for evaluation of pubertal disorders, while four studies included only healthy girls. Nine studies found a significant difference in Doppler signal pattern and PI according to pubertal stage, with PI cutoff points ranging from 2.5 to 4.6 for the diagnosis of puberty, with a sensitivity of 77%-94%, specificity of 85%-100%, and accuracy of 79%-98%. Doppler assessment of the uterine arteries with PI calculation is a useful noninvasive tool in the diagnosis of pubertal onset in girls.

Difference of Precocious Puberty Between Before and During the COVID-19 Pandemic: A Cross-Sectional Study Among Shanghai School-Aged Girls

OBJECTIVE

To compared the incidence rates and clinical features of precocious girls before and during the COVID-19 pandemic among Shanghai school-aged girls, and explored the potential mechanisms.

METHODS

This cross-sectional study collected medical data about precocious girls between 2016 and 2020 from Shanghai Children's Medical Center. Data of inpatient precocious girls from March to August in 2016-2019 (n=246) and 2020 (n=237) were collected. Subjects with abnormal brain and pituitary gland MRI reports, other endocrine diseases or chronic diseases were excluded. Finally, 209 precocious girls were included in the 2016-2019 group and 191 precocious girls were include in the 2020 group. Monthly incidence rates and clinical features were compared between before and during the COVID-19 pandemic. Linear regression models were used to examine the associations between biomarkers to explore the potential mechanisms.

RESULTS

Monthly incidence rates of precocious puberty in outpatient girls from March to December 2020 (0.44-1.36%) and in inpatient girls from March to August 2020 (27.04-47.83%) were higher than those in 2016-2019 (0.30-0.52% and 10.53-18.42%, respectively). Serum concentrations of GnRH were higher in the 2020 group than in the 2016-2019 group (2.81 vs 1.99 mg/L). Serum concentrations of MKRN3 (1.02 vs 1.93 ng/ml) and ghrelin (0.38 vs 0.88 ng/ml) were lower in the 2020 group than in the 2016-2019 group. Moreover, the serum concentration of ghrelin was positively associated with the serum concentration of MKRN3 [β=0.891 (95% CI, 0.612, 1.171); p<0.001].

CONCLUSIONS

These findings suggest an increased incidence of precocious puberty during the COVID-19 pandemic among Shanghai school-aged girls, which may be associated with decreased serum concentrations of MKRN3 and ghrelin, and indicated ghrelin as a potential regulatory mechanism of puberty.

Diagnosis and management of precocious sexual maturation: an updated review

The classic definition of precocious sexual maturation is the development of secondary sexual characteristics before 8 years of age in girls and before 9 years of age in boys. It is classified as central precocious puberty when premature maturation of the hypothalamic-pituitary-gonadal axis occurs, and as peripheral precocious puberty when there is excessive secretion of sex hormones, independent of gonadotropin secretion. Precocious sexual maturation is more common in girls, generally central precocious puberty of idiopathic origin. In boys, it tends to be linked to central nervous system abnormalities. Clinical evaluation should include a detailed history and physical examination, including anthropometric measurements, calculation of growth velocity, and evaluation of secondary sexual characteristics. The main sign to suspect the onset of puberty is breast tissue development (thelarche) in girls and testicular enlargement (≥4 mL) in boys. Hormonal assessment and imaging are required for diagnosis and identification of the etiology. Genetic testing should be considered if there is a family history of precocious puberty or other clinical features suggestive of a genetic syndrome. Long-acting gonadotropin-releasing hormone analogs are the standard of care for central precocious puberty management, while peripheral precocious puberty management depends on the etiology.Conclusion: The aim of this review is to address the epidemiology, etiology, clinical assessment, and management of precocious sexual maturation. What is Known: • The main sign to suspect the onset of puberty is breast tissue development (thelarche) in girls and testicular enlargement (≥4 mL) in boys. The classic definition of precocious sexual maturation is the development of secondary sexual characteristics before 8 years of age in girls and before 9 years of age in boys. • Long-acting gonadotropin-releasing hormone agonist (GnRHa) is the standard of care for CPP management, and adequate hormone suppression results in the stabilization of pubertal progression, a decline in growth velocity, and a decrease in bone age advancement. What is New: • Most cases of precocious sexual maturation are gonadotropin-dependent and currently assumed to be idiopathic, but mutations in genes involved in pubertal development have been identified, such as MKRN3 and DLK1. • A different preparation of long-acting GnRHa is now available: 6-month subcutaneous injection.

Explainable Dynamic Multimodal Variational Autoencoder for the Prediction of Patients with Suspected Central Precocious Puberty

Central precocious puberty (CPP) is the most common type of precocious puberty and has a significant effect on children. A gonadotropin-releasing hormone (GnRH)-stimulation test is the gold standard for confirming CPP. This test, however, is costly and unpleasant for patients. Therefore, it is critical to developing alternative methods for CPP diagnosis in order to alleviate patient suffering. This study aims to develop an artificial intelligence (AI) diagnostic system for predicting response to the GnRH-stimulation test using data from laboratory tests, electronic health records (EHRs), and pelvic ultrasonography and left-hand radiography reports. The challenges are in integrating these mul-timodal features into a comprehensive deep learning model in order to achieve an accurate diagnosis while also accounting for the missing or incomplete modalities. To begin, we developed a dynamic multimodal variational autoencoder (DMVAE) that can exploit intrinsic correlations between different modalities to im-pute features for missing modalities. Next, we combined features from all modalities to predict the outcome of a CPP diagnosis. The experimental results (AUROC 0.9086) demonstrate that our DMVAE model is superior to standard methods. Additionally, we showed that by setting appropriate operating thresholds, clinicians could diagnose about two-thirds of patients with confidence (1.0 specificity). Only about one-third of patients require confirmation of their diagnoses using GnRH (or GnRH analog)-stimulation tests. To interpret the results, we implemented an explainer Shapley additive explanation (SHAP) to analyze the local and global feature attributions.

A diagnostic model of idiopathic central precocious puberty based on transrectal pelvic ultrasound and basal gonadotropin levels

OBJECTIVE

To establish a diagnostic model of idiopathic central precocious puberty on the basis of transrectal pelvic ultrasound and basal gonadotropin.

METHODS

A total of 669 girls with Tanner breast development stage II were enrolled in this study from January 2015 to December 2018. The participants were divided into the ICPP group and the premature thelarche group. We analyzed various variables, including age at initial diagnosis, basal luteinizing hormone levels, the long diameter of the uterus, the transverse diameter of the uterus, the anterior-posterior diameter of the uterus, the volume of the uterus, maximum ovarian diameter, average ovarian volume, maximum ovarian volume, number of follicles (≥4 mm), maximum follicular diameter, endometrial thickness, and vaginal wall thickness.

RESULTS

The following diagnostic model was established: Y=-14.123 + 0.630 × age at initial diagnosis + 1.119 × transverse diameter of the uterus + 1.278 ×  anterior-posterior diameter of the uterus + 0.637 × average ovarian volume + 1.316 × maximum ovarian diameter + 0.146 ×number of follicles ≥4 mm + 2.925 × endometrial thickness + 0.559 × basal luteinizing hormone value. The area under curve was 0.922, sensitivity was 84.9%, and specificity was 86.2%.

CONCLUSION

Basal LH levels and transrectal pelvic ultrasound should be applied together to improve the accuracy of diagnosis in ICPP.

Machine learning identifies girls with central precocious puberty based on multisource data

OBJECTIVE

The study aimed to develop simplified diagnostic models for identifying girls with central precocious puberty (CPP), without the expensive and cumbersome gonadotropin-releasing hormone (GnRH) stimulation test, which is the gold standard for CPP diagnosis.

MATERIALS AND METHODS

Female patients who had secondary sexual characteristics before 8 years old and had taken a GnRH analog (GnRHa) stimulation test at a medical center in Guangzhou, China were enrolled. Data from clinical visiting, laboratory tests, and medical image examinations were collected. We first extracted features from unstructured data such as clinical reports and medical images. Then, models based on each single-source data or multisource data were developed with Extreme Gradient Boosting (XGBoost) classifier to classify patients as CPP or non-CPP.

RESULTS

The best performance achieved an area under the curve (AUC) of 0.88 and Youden index of 0.64 in the model based on multisource data. The performance of single-source models based on data from basal laboratory tests and the feature importance of each variable showed that the basal hormone test had the highest diagnostic value for a CPP diagnosis.

CONCLUSION

We developed three simplified models that use easily accessed clinical data before the GnRH stimulation test to identify girls who are at high risk of CPP. These models are tailored to the needs of patients in different clinical settings. Machine learning technologies and multisource data fusion can help to make a better diagnosis than traditional methods.

Diagnostic value of pituitary volume in girls with precocious puberty

BACKGROUND

To date, the gonadotropin-releasing hormone (GnRH) stimulation test is still the gold standard for precocious puberty (PP) diagnosis. However, it has many disadvantages, including low sensitivity, high cost, and invasive operation. This study aims to evaluate whether magnetic resonance imaging (MRI)-derived variables, including pituitary volume (PV), could be used as diagnostic factors for PP in girls, providing a non-invasive diagnostic approach for PP.

METHODS

A total of 288 young female patients who presented to the Clinic of Pediatric Endocrinology for evaluation of PP from January 2015 to December 2017 were enrolled. The sample included 90 girls diagnosed with premature thelarche (PT), 133 girls determined as idiopathic central precocious puberty (ICPP), 35 early pubertal girls, and 30 normal girls. All patients received pituitary MRI examinations.

RESULTS

The largest PV and pituitary height were shown in the ICPP and pubertal groups, followed by the PT group. The receiver operating characteristic (ROC) curve analysis showed that PV is a predictive marker for ICPP, with a sensitivity of 54.10% and a specificity of 72.20% at the cutoff value of 196.01 mm³. By univariate analysis, PV was positively associated with peak luteinizing hormone (LH), LH/follicle-stimulating hormone (FSH), age, bone age, and body mass index (BMI) (all P < 0.05). However, bone age and peak LH were the only significant predictors of PV as demonstrated by the stepwise multivariate regression analysis (Model: PV = 9.431 * bone age + 1.230 * peak LH + 92.625 [P = 0.000, R² = 0.159]).

CONCLUSIONS

The PV in the ICPP group was significantly higher than in PT and control groups, but there was no reliable cutoff value to distinguish ICPP from PT. Pituitary MRI should be combined with clinical and laboratory tests to improve the diagnostic value of PV for PP.

Rapid progressive central precocious puberty: diagnostic and predictive value of basal sex hormone levels and pelvic ultrasound

Objectives Data on the predictive values of parameters included in the diagnostic work-up for precocious puberty (PP) remain limited. We detected the diagnostic value of basal sex hormone levels, pelvic ultrasound parameters and bone age assessment for activation of the hypothalamic-pituitary-gonadal axis in girls with PP, in order to help in the decision to perform GnRH testing. Patients and methods We retrospectively considered 177 girls with PP. According to puberty evolution, the girls were divided into two groups: rapid progressive central precocious puberty (RP-CPP) and non/slowly progressive/transient forms (SP-PP). In all patients we considered Tanner stage, basal luteinizing hormone (LH) and estradiol (E2) values, bone age, and pelvis examination. We assessed the diagnostic value of each variable and identified the number of pathological parameters that best identify patients with RP-CPP. Results Basal LH ≥ 0.2IU/L, E2 level ≥ 50 pmol/L, uterine longitudinal diameter ≥ 3.5 cm, transverse uterine diameter ≥ 1.5 cm, endometrial echo and ovarian volume ≥ 2 cm3 were significantly associated with RP-CPP (p ≤ 0.01). The ability to diagnose RP-CPP was enhanced with increasing number of pathological hormonal and instrumental parameters (p < 0.001). With more than three parameters detected, sensitivity and specificity reached 58% (95%CI 48-67) and 85% (95%CI 74-92), respectively, with a PPV = 86% (95%CI 76-93) and PPN = 54% (95%CI 43-54); the area under the ROC curve was 0.71 (95%CI 0.65-0.78). Conclusion Despite the availability of different tests, diagnosing RP-CPP remains difficult. A diagnosis model including at least three hormonal and/or ultrasound parameters may serve as a useful preliminary step in selecting patients who require GnRH testing for early detection of RC-PP.

Basal serum luteinizing hormone value as the screening biomarker in female central precocious puberty

PURPOSE

Precocious puberty refers to the development of secondary sex characteristics before ages 8 and 9 years in girls and boys, respectively. Central precocious puberty (CPP) is caused by premature activation of the hypothalamus-pituitary-gonadal (HPG) axis and causes thelarche in girls before the age of 8. A gonadotropin-releasing hormone (GnRH) stimulation test is the standard diagnostic modality for diagnosing CPP. However, the test cannot always be used for screening because it is expensive and time-consuming. This study aimed to find alternative reliable screening parameters to identify HPG axis activation in girls &lt;8 years old (CPP) and for girls 8-9 years old (early puberty, EP).

METHODS

From January 2013 to June 2015, medical records from 196 girls younger than 9 years old with onset of breast development were reviewed, including 126 girls who had a bone age (BA) 1 year above their chronological age. All patients underwent a GnRH stimulation test, and 117 underwent pelvic sonography. The girls were divided into 4 groups based on age and whether the GnRH stimulation test showed evidence of central puberty. Subanalyses were also conducted within each group based on peak luteinizing hormone (LH) level quartiles.

RESULTS

Basal serum LH level was the most sensitive marker for screening CPP and EP. The cutoff values were 0.245 IU/L for CPP under 8 years old (P=0.049, area under the curve [AUC]=0.764, 88% sensitivity, 48% specificity) and 0.275 IU/L for EP between 8-9 years old (P=0.005, AUC=0.813, 79% sensitivity, 77% specificity). Peak LH level decreased as BMI z-score among subgroups increased when there was no difference in BA; however, higher BA eliminated this effect.

CONCLUSION

Basal serum LH level is a useful screening parameter for diagnosing CPP and EP in girls. Peak LH levels were lower with increasing BMI z-score, although older BA eliminated this effect.

Normative values for ultrasound measurements of the female pelvic organs throughout childhood and adolescence

BACKGROUND

Normative data from pelvic ultrasonography (US) of the pediatric female reproductive organs are outdated and limited by sample size.

OBJECTIVE

The purpose of this study was to report normal uterine and ovarian volumes and endometrial stripe thickness in children and young adults and throughout the menstrual cycle in post-menarchal adolescents.

MATERIALS AND METHODS

Females ages 0 through 20 years who underwent pelvic US between January 2010 and May 2014 were identified. After excluding patients with pelvic and/or endocrine pathology, nomograms of uterine and ovarian volumes and endometrial thickness by age year were created. Data for patients ages 12 years and older with a recorded day of last menstrual period were used to create additional nomograms of volumes/thickness throughout the menstrual cycle. Student's t-tests and linear regression were performed to assess differences in measurements between groups and association of volumes/thickness with age.

RESULTS

During our study period, 5,647 patients underwent 6,953 pelvic US examinations. After further review, 907 examinations from 889 patients were included (mean age: 11.3±6.0 years). Mean pelvic US volumes (cm³) per organ were 25.5±27.0 (uterus), 4.5±4.7 (right ovary) and 4.0±4.1 (left ovary). Mean endometrial thickness was 4.5±3.7 mm. Right ovarian volume was significantly larger than the left (P=0.0126). Uterine volume, ovarian volume and endometrial thickness were significantly associated with age (P-values<0.0001). Plots of mean organ measurements with respect to week of menses are provided.

CONCLUSION

We report normal volumes of the uterus and ovaries and endometrial stripe thickness measured by pelvic US throughout childhood and adolescence with reference to the menstrual cycle. These values are significantly associated with age and vary visually by menstrual cycle week.

Central precocious puberty in girls: Diagnostic study and auxological response to triptorelin treatment

INTRODUCTION

There are several controversies regarding the diagnostic tests and management of central precocious puberty (CPP). The aim of this study is to present the experience acquired in a group of girls with CPP treated with triptorelin, and to analyze the auxological characteristics and diagnostic tests.

MATERIAL AND METHODS

An observational, retrospective study in a group of 60 girls with CPP was conducted between January 2010 and December 2017. Sociodemographic, auxological and hormonal data were recorded at diagnosis, and pelvic ultrasound and magnetic resonance imaging of the head were performed. Girls were treated with triptorelin and monitored after treatment discontinuation until menarche.

RESULTS

At treatment start, chronological age and bone age were 7.7±0.7 and 9.7±0.8 years respectively, and growth velocity was 8.3±1.6cm/year. Target height was 161.1±5.8cm. Peak LH level after stimulation was 16.6±12.1 IU/l. Ovarian volumes were greater than 3mL in 35% of cases. MRI of the head was pathological in seven girls (11.7%). At treatment completion, chronological age and bone age were 10.3±1.1 and 11.2±0.8 years respectively, and growth velocity was 4.7±1.4cm/year. At the age of menarche (11.9±0.9 years), height was 157.5±5.7cm.

CONCLUSIONS

Treatment of CPP with triptorelin appears to be beneficial. The possibility to block pubertal development and slow skeletal maturation allows patients to reach their target height. However, individualized auxological monitoring would be mandatory.

Development of Prediction Models Using Machine Learning Algorithms for Girls with Suspected Central Precocious Puberty: Retrospective Study

Background

Central precocious puberty (CPP) in girls seriously affects their physical and mental development in childhood. The method of diagnosis—gonadotropin-releasing hormone (GnRH)–stimulation test or GnRH analogue (GnRHa)–stimulation test—is expensive and makes patients uncomfortable due to the need for repeated blood sampling.

Objective

We aimed to combine multiple CPP–related features and construct machine learning models to predict response to the GnRHa-stimulation test.

Methods

In this retrospective study, we analyzed clinical and laboratory data of 1757 girls who underwent a GnRHa test in order to develop XGBoost and random forest classifiers for prediction of response to the GnRHa test. The local interpretable model-agnostic explanations (LIME) algorithm was used with the black-box classifiers to increase their interpretability. We measured sensitivity, specificity, and area under receiver operating characteristic (AUC) of the models.

Results

Both the XGBoost and random forest models achieved good performance in distinguishing between positive and negative responses, with the AUC ranging from 0.88 to 0.90, sensitivity ranging from 77.91% to 77.94%, and specificity ranging from 84.32% to 87.66%. Basal serum luteinizing hormone, follicle-stimulating hormone, and insulin-like growth factor-I levels were found to be the three most important factors. In the interpretable models of LIME, the abovementioned variables made high contributions to the prediction probability.

Conclusions

The prediction models we developed can help diagnose CPP and may be used as a prescreening tool before the GnRHa-stimulation test.

Association between bisphenol a exposure and idiopathic central precocious puberty (ICPP) among school-aged girls in Shanghai, China

BACKGROUND

Bisphenol A (BPA) is a well-known and widely used endocrine disrupter, but data on its association with childhood reproductive development are limited.

OBJECTIVES

We investigated the possible relationship between exposure to BPA and idiopathic central precocious puberty (ICPP) in school-aged girls.

METHODS

We conducted a 1:1 matched case-control study in Shanghai, China, between July 2011 and September 2012. This study included 136 school-aged (6 to 9 years old) girls diagnosed with ICPP and 136 controls matched for age and body mass index (BMI). We measured the urinary BPA concentrations of all the girls and examined the association with odds of having ICPP. Laboratory examinations including serum estradiol (E2) levels, basal and gonadotropin-releasing hormone (GnRH)-stimulated luteinizing hormone (LH), and follicle-stimulating hormone (FSH) levels, bone ages (BA), and uterine and ovarian sizes were conducted in the ICPP girls.

RESULTS

Median concentrations of urinary BPA in the ICPP and control groups were 6.35 and 1.17 μg/g creatinine (Cr), respectively (p < 0.001). After adjustment for confounders, compared to those with the lowest concentrations of BPA, the highest concentrations were associated with a 9.08-fold increased odds of having ICPP [odds ratio (OR) = 9.08, (95% confidence interval (CI): 2.83-29.15)]. In the ICPP group, modest negative correlation was present between urinary BPA concentrations (μg/g Cr) and peak FSH levels [β = -0.090 (95% CI: -0.178, -0.003), p = 0.044].

CONCLUSIONS

Our findings suggest that BPA exposure is associated with increased odds of having ICPP in school-aged girls, and the potential mechanism may be attributable to the relatively low FSH levels.