Abstract
Background:
This study aimed to identify the predictive value of basal sex hormone levels for activation of the hypothalamic–pituitary–gonadal (HPG) axis in girls.
Methods:
Gonadotropin-releasing hormone (GnRH) stimulation tests were performed and evaluated in a total of 1750 girls with development of secondary sex characteristics. Correlation analyses were conducted between basal sex hormones and peak luteinizing hormone (LH) levels ≥5 IU/L during the GnRH stimulation test. Receiver operating characteristic (ROC) curves for basal levels of LH, follicle-stimulating hormone (FSH), LH/FSH, and estradiol (E2) before the GnRH stimulation test were plotted. The area under the curve (AUC) and 95% confidence intervals (CIs) were measured for each curve.
对总共 1750 名出现第二性征的女孩进行促性腺激素释放激素 （GnRH） 刺激试验和评估。在 GnRH 刺激试验期间，对基础性激素与黄体生成素 （LH） 峰值 （LH） 水平 ≥5 IU/L 进行相关性分析。绘制 GnRH 刺激试验前 LH 、促卵泡激素 （FSH） 、 LH/FSH 和雌二醇 （E2） 基础水平的受试者工作特征 （ROC） 曲线。测量每条曲线的曲线下面积 （AUC） 和 95% 置信区间 （CIs）。
Results:
The maximum AUC value was observed for basal LH levels (0.77, 95% CI: 0.74–0.79), followed by basal FSH levels (0.73, 95% CI: 0.70–0.75), the basal LH/FSH ratio (0.68, 95% CI: 0.65–0.71), and basal E2 levels (0.61, 95% CI: 0.59–0.64). The appropriate cutoff value of basal LH levels associated with a positive response of the GnRH stimulation test was 0.35 IU/L, with a sensitivity of 63.96% and specificity of 76.3% from the ROC curves when Youden’s index showed the maximum value. When 100% of patients had peak LH levels ≥5 IU/L, basal LH values were >2.72 IU/L, but the specificity was only 5.45%.
Conclusions:
Increased basal LH levels are a significant predictor of a positive response during the GnRH stimulation test for assessing activation of the HPG axis in most girls with early pubertal signs.
Introduction
Precocious puberty is a common disease in the field of pediatric endocrinology [1]. Most patients with precocious puberty suffer from inappropriate activation of the hypothalamic–pituitary–gonadal axis (HPG), which results in idiopathic central precocious puberty (CPP). Activation of the HPG is important in the diagnosis of CPP and is based on progressive sexual development, accelerated growth rate, and advanced bone maturation. In cases where the gonadal axis is not activated, peripheral precocious puberty (PPP) is considered. PPP can show similar clinical manifestations to CPP, although the pathogenesis, clinical outcomes, and treatment methods for PPP differ from those of CPP.
Measurements of peak luteinizing hormone (LH) following the gonadotropin-releasing hormone (GnRH) stimulation test is the gold standard for assessing early activation of the HPG axis in cases with clinical symptoms and signs of puberty [2], [3]. However, the GnRH stimulation test requires several blood samples over long time periods, with relevant technicians and facilities. This highlights the need for a simple measure that can be used as a screening test for assessing early activation of the HPG axis. Early in the activation of the HPG axis, amplitude and pulse frequency of serum LH and follicle-stimulating hormone (FSH) secretion are significantly increased. However, prepubertal and pubertal gonadotropin baseline values have some overlap [4]. With use of a third-generation gonadal hormone detection method, which uses an immunochemiluminometric assay, detection sensitivity has significantly improved compared with traditional detection methods. This has helped to differentiate between prepubertal and pubertal hormone levels [5]. Children with CPP were proposed to have higher basal FSH and LH levels than children with PPP in a cross-sectional study [6]. However, further studies are required to determine a more sensitive index for detecting gonadal axis activation and how to select the appropriate cutoff value.
Therefore, the present study aimed to identify the predictive value for activation of the HPG axis in girls. We analyzed basal LH and FSH levels, the ratio between LH and FSH (LH/FSH), and estradiol (E2) levels prior to the GnRH stimulation test.
Subjects and methods
Subjects
We studied a total of 1750 girls with breast enlargement before 8 years of age, who had a physical examination and breast ultrasound indicating breast development, with a breast of Tanner stage 2 or higher. These girls were diagnosed and treated in the Endocrinology Department of Shanghai Children’s Medical Center from January 2010 to June 2015. Patients with precocious puberty as a result of another etiology, such as a central tumor, infection, or cranial irradiation, were excluded from the study. Measurements included height, weight, bone age by X-ray photography, and ultrasonography of the uterus and ovaries. Bone age was measured by the Greulich-Pyle method [7]. The volumes of the uterus and ovary were calculated by multiplying the length by the width, thickness, and 3.14, and then dividing by six according to ultrasonography. GnRH stimulation tests were performed and evaluated. The girls were divided into two groups according to GnRH stimulation test results. Girls with peak LH values ≥5 IU/L were considered to have pubertal activation of the HPG axis. These girls were categorized into the positive GnRH stimulation test group. Girls with peak LH values <5 IU/L were considered to have inactivation of the HPG axis and were categorized into the negative GnRH stimulation test group. Informed consent was obtained from the participating children and their parents, and the study was approved by the Institutional Review Board of the Shanghai Children’s Medical Center. This study was in accordance with the tenets of the Helsinki Declaration.
Methods
The GnRH stimulation test was performed in the early morning after fasting for 10 h. Gonadorelin (AnhuiFengyuan Pharmaceutical Co., Anhui, China) was injected at a dose of 2.5 μg/kg, with a maximum dosage of 100 μg. Blood samples were drawn from an inserted intravenous cannula before and 30 and 60 min after GnRH injection. Previous studies have suggested that LH levels between 30 and 60 min are sufficient for diagnosis of activation of the HPG axis [8], [9], [10]. All samples were analyzed for LH and FSH levels. The maximal LH and FSH levels achieved at any time point of testing were considered to be the peak levels. Additionally, E2 levels were determined prior to GnRH administration. An electrochemiluminescence immunoassay (DxI800 automated chemiluminescence assay and commercial kit; Beckman Coulter, Inc., CA, USA) was used to determine hormone levels. The intra-assay coefficient of variation for LH was 3.6%–5.4%, with an inter-assay imprecision of 4.3%–6.4% and sensitivity of 0.2 IU/L. The calibration range of the assay was up to 250 IU/L. The intra-assay coefficient of variation for FSH was 3.1%–4.3% and inter-assay imprecision was 4.3%–5.6%. The sensitivity was 0.2 IU/L and the calibration range of the assay was up to 200 IU/L. E2 assay sensitivity was 20 pg/mL. The calibration range of the assay was up to 4800 pg/mL. The intra-assay coefficient of variation was 12%–21%.
Statistical analysis
The Student’s t-test was performed to compare the mean of subjects’ characteristics between groups, and normal distribution transformation was conducted on a 1/square. Correlation analyses were conducted between sex hormones and puberty status. The odds ratio was calculated according to basal test value between the positive GnRH test group and negative GnRH test group. Because the cutoff value was LHmax=5.0, the continuous variable LHmax was converted into a binary variable. The LHmax values were categorized as 0 or 1 if LHmax values were <5.0 or ≥5.0. Receiver operating characteristic (ROC) curves for basal levels of LH, FSH, LH/FSH, and E2 were plotted. The area under the curve (AUC) and 95% confidence intervals (CIs) were measured for each curve. Youden’s index (sensitivity+specificity−1) was used to determine the optimal gonadotropin cutoff point from the ROC. The test of equality of ROC areas was performed to compare the AUC between groups. ROC analysis for multiple comparisons was performed between different AUCs using the DeLong method [11]. If multiple comparisons were significant, every two AUCs were further compared. A p-value <0.05 was considered statistically significant. All statistical analyses were performed using Stata 13.0 (Stata Corporation, College Station, TX, USA).
Results
Clinical and hormonal characteristics in the patients
There were 1138 patients in the positive GnRH test group (mean age±standard deviation: 7.95±1.25 years) and 612 patients (7.16±1.59 years) in the negative GnRH test group. The difference between chronological age and bone age was 1.38 years in the negative GnRH test group and 1.43 years in the positive GnRH test group (p=0.671). There was no significant difference in body mass index between the two groups. Uterine volume and mean ovarian volume were larger in the positive GnRH test group than in the negative GnRH test group (both p<0.05). Basal LH levels, FSH levels, the LH/FSH ratio, and E2 levels were significantly lower in the negative GnRH test group than in the positive GnRH test group (all p<0.05). Peak LH levels, FSH levels, and the LH/FSH ratio were significantly higher in the positive GnRH test group than in the negative GnRH test group (all p<0.05, Table 1). The correlation coefficient of peak LH levels and bone age was 0.21, that of peak LH levels and the size of the uterus was 0.42.
GnRH 试验阳性组有 1138 例患者 （平均年龄±标准差：7.95±1.25 岁），GnRH 试验阴性组有 612 例患者 （7.16±1.59 岁）。GnRH 试验阴性组实际年龄和骨龄的差异为 1.38 岁，GnRH 试验阳性组为 1.43 岁 （p=0.671）。两组体重指数差异无统计学意义。GnRH 试验阳性组子宫体积和平均卵巢体积大于 GnRH 阴性组 （均 p<0.05）。GnRH 阴性组基础 LH 水平、FSH 水平、LH/FSH 比值和 E2 水平显著低于 GnRH 阳性组 （均 p<0.05）。阳性 GnRH 试验组的峰值 LH 水平、FSH 水平和 LH/FSH 比值显著高于阴性 GnRH 试验组（均 p<0.05，表 1）。LH 峰值水平与骨龄的相关系数为 0.21，LH 峰值水平与子宫大小的相关系数为 0.42。
Hormonal and clinical characteristics of participants.
| Variables | GnRH test (−), n=612 | GnRH test (+), n=1138 | t-Testa | p-Valuea | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Mean±SD | Median | Range | Distribution | Mean±SD | Median | Range | Distribution | |||
| Age, year | 7.16±1.59 | 7.46 | 4.91–10.08 | Normal | 7.95±1.25 | 8.17 | 5.12–10.33 | Normal | −5.99 | <0.001 |
| Bone age–chronological age, year | 1.38±1.000 | 1.38 | −2.08 to 3.63 | Normal | 1.43±1.11 | 1.50 | −3.03 to 4.47 | Normal | −0.43 | 0.671 |
| Height | 126.91±11.49 | 129.00 | 108.00–158.40 | Normal | 131.15±9.85 | 132.60 | 109.00–163.50 | Normal | −2.39 | <0.001 |
| Weight | 27.72±6.79 | 27.70 | 18.50–48.00 | Normal | 29.48±6.14 | 29.00 | 18.50–50.50 | Normal | −3.07 | 0.001 |
| BMI, kg/m2 | 16.98±2.43 | 16.47 | 12.49–27.51 | Normal | 16.99±2.31 | 16.74 | 10.92–32.53 | Normal | −0.06 | 0.475 |
| Mean ovarian volume, mL3b | 1.88±0.93 | 1.75 | 0.10–6.16 | Normal | 2.05±0.94 | 1.90 | 0.08–7.04 | Normal | 2.41 | 0.016 |
| Uterine size, mL3 | 2.61±1.83 | 2.23 | 0.22–21.30 | Normal | 3.75±2.21 | 3.22 | 0.15–16.12 | Normal | −7.24 | <0.001 |
| Basal LH, IU/L | 0.27±0.32 | 0.19 | 0.01–16.69 | Normal | 0.87±1.20 | 0. 52 | 0.01–16.31 | Normal | 4.74 | <0.001 |
| Basal FSH, IU/L | 2.51±1.33 | 2.31 | 0.03–8.10 | Normal | 3.93±2.03 | 3.46 | 0.57–17.75 | Normal | 2.34 | 0.010 |
| Basal LH/FSH ratio | 0.17±0.56 | 0.08 | 0.01–12.0 | Normal | 0.22±0.28 | 0.15 | 0.01–4.35 | Normal | 3.96 | <0.001 |
| Basal E2, pg/mL | 15.56±21.81 | 11.00 | 1.0–258.0 | Normal | 21.51±24.72 | 15.00 | 1.00–447.00 | Normal | 4.55 | <0.001 |
| Peak LH, IU/L | 3.18±1.15 | 3.32 | 0.12–4.98 | Normal | 14.95±14.04 | 9.78 | 5.00–158.12 | Normal | 4.22 | <0.001 |
| Peak FSH, IU/L | 13.73±5.26 | 13.48 | 0.14–33.71 | Normal | 15.98±5.96 | 14.98 | 1.56–46.37 | Normal | 2.39 | 0.017 |
| Peak LH/FSH ratio | 0.27±0.18 | 0.23 | 0.04–2.93 | Normal | 0.98±0.75 | 0.73 | 0.19–9.02 | Normal | 21.81 | <0.001 |
GnRH test (−): peak LH values <5 IU/L during GnRH stimulation test. GnRH test (+): peak LH values ≥5 IU/L during GnRH stimulation test. at-Test based on 1/square transformation. bMean ovarian volume was calculated by dividing the sum of right and left ovarian volumes by two. SE, standard error.
Logistic regression analysis
The biochemical parameters that were considered to be related to the GnRH stimulation test results were adjusted using binary logistic regression analysis (Table 2). After regression analysis, basal LH levels were the most significantly and positively related to a positive response in the GnRH stimulation test (p<0.05).
Binary logistic regression analysis of hormones compared between GnRH test (+) group and GnRH test (−) group.
| Variable | Odds ratio | 95% CI | z-Value | p-Value |
|---|---|---|---|---|
| Basal LH | 12.11 | 8.29–17.71 | 12.88 | <0.001 |
| Basal FSH | 1.73 | 1.60–1.87 | 13.69 | <0.001 |
| Basal LH/FSH ratio | 1.90 | 1.19–3.04 | 2.68 | 0.007 |
| Basal E2 | 1.02 | 1.01–1.02 | 5.10 | <0.001 |
GnRH test (−) group: peak LH values <5 IU/L during GnRH stimulation test. GnRH test (+) group: peak LH values ≥5 IU/L during GnRH stimulation test.
ROC analysis
ROC curves were plotted for basal LH levels, FSH levels, the LH/FSH ratio, and E2 levels (Figure 1). The AUC was measured for each curve (Table 3). A larger AUC represented a more positive rate of excitation. The maximum AUC was observed for basal LH levels, followed by basal FSH levels, the basal LH/FSH ratio, and basal E2 levels. This finding suggested that the basal LH value was best for predicting activation of the gonadal axis. The p-values of AUC comparisons was <0.05 between each AUC of basal hormone (Table 3). The appropriate cutoff value of basal LH levels associated with a positive response was 0.35 IU/L when Youden’s J index reached the maximum value. The sensitivity was 63.96% and specificity was 76.35% from the ROC curves. Therefore, a basal LH value that reached 0.35 IU/L suggested that gonadal axis activation was relatively high and further GnRH stimulation testing was required. A basal LH value that reached 2.72 IU/L (specificity was 100%, but sensitivity decreased to 5.45%) suggested a 100% peak LH level of ≥5 IU/L (Table 4).
ROC curves of basal LH and FSH, basal LH/FSH ratio, and E2 value for predicting positive results following GnRH stimulation testing.
Comparison of AUCs for each basal hormone.
| Variable | AUC | 95% CI | p-Valuesa | |||
|---|---|---|---|---|---|---|
| AUC(LH) | AUC(FSH) | AUC(LH/FSH) | AUC(E2) | |||
| AUC(LH) | 0.77 | 0.74–0.79 | – | 0.002b | <0.001 | <0.001 |
| AUC(FSH) | 0.73 | 0.70–0.75 | 0.002 | – | 0.021 | <0.001 |
| AUC(LH/FSH) | 0.68 | 0.65–0.71 | <0.001 | 0.021 | – | <0.001 |
| AUC(E2) | 0.61 | 0.59–0.64 | <0.001 | <0.001 | <0.001 | – |
AUC, area under curve, –, Data is not available. ap-Values of AUC comparisons between each AUC of basal hormone. bp-values of AUC(LH) vs. AUC(FSH).
Sensitivity and specificity of deferent basal hormone levels for predicting positive results on GnRH stimulation test.
| Hormone | Cutoff, AUC | Cutoff | True+ | False+ | False− | True− | Sensitivity, % | Specificity, % | Youden’s indexa | PPV, % | NPV, % |
|---|---|---|---|---|---|---|---|---|---|---|---|
| LH | 0.73 | 0.35 IU/L | 270 | 149 | 112 | 357 | 63.96 | 76.35 | 0.40 | 100.00 | 37.32 |
| 0.86 | 0.61 IU/L | 282 | 368 | 51 | 505 | 43.47 | 90.37 | 0.34 | |||
| 0.89 | 0.78 IU/L | 258 | 494 | 29 | 555 | 34.23 | 95.10 | 0.30 | |||
| 0.99 | 2.72 IU/L | 59 | 1018 | 0 | 612 | 5.45 | 100.00 | 0.05 | |||
| FSH | 0.73 | 3.03 IU/L | 274 | 179 | 117 | 339 | 57.99 | 76.69 | 0.35 | 100.00 | 37.30 |
| 0.87 | 4.22 IU/L | 267 | 455 | 55 | 496 | 34.75 | 90.03 | 0.25 | |||
| 0.93 | 5.24 IU/L | 204 | 671 | 29 | 553 | 22.99 | 95.10 | 0.18 | |||
| 0.998 | 8.12 IU/L | 38 | 1062 | 0 | 612 | 3.43 | 100.00 | 0.03 | |||
| LH/FSH | 0.65 | 0.11 | – | – | – | – | 63.00 | 65.70 | 0.29 | 97.50 | 36.91 |
| 0.93 | 0.28 | – | – | – | – | 23.46 | 90.07 | 13.53 | |||
| 0.97 | 0.44 | – | – | – | – | 10.28 | 95.03 | 5.31 | |||
| – | 12.00 | – | – | – | – | 0.00 | 100.00 | 0.00 | |||
| E2 | 0.66 | 18.00 pg/mL | 168 | 206 | 61 | 166 | 44.51 | 72.59 | 0.18 | 100.00 | 37.20 |
| 0.78 | 23.00 pg/mL | 164 | 351 | 53 | 223 | 31.46 | 80.54 | 0.12 | |||
| 0.90 | 39.00 pg/mL | 148 | 824 | 27 | 556 | 14.95 | 95.26 | 0.10 | |||
| 0.99 | 288.00 pg/mL | 2 | 1135 | 0 | 611 | 0.18 | 100.00 | 0.02 |
aThe maximum value of Youden’s J index. –, data is not available; NPV, negative predictive value; PPV, positive predictive value.
Discussion
Our study showed that the basal LH level was useful for predicting gonadal axis activation. Activation of the gonadal axis is important for diagnosing CPP, which can accelerate bone maturation, result in impaired adult height and early menstruation, and can greatly affect the patient’s psychological health [12]. Previous studies have shown that early menstruation is related to adverse health outcomes in later life [13], [14], [15]. Therefore, timely diagnosis and appropriate treatment can help to improve the prognosis of these patients. Diagnosing CPP is difficult and should include clinical manifestations and correct and timely assessment of activation of the HPG axis. However, clinically diagnosing activation of the HPG axis by a simple examination and clinical data is challenging. Currently, the biochemical criteria for diagnostic confirmation of gonadal axis activation are primarily based on the LH response during a standard GnRH stimulation test. A stimulated LH value ≥5 IU/L and/or a stimulated peak LH/FSH ratio >0.6 are considered pubertal responses during GnRH testing [2], [16], [17]. Pubertal LH secretion is characterized by high levels of peak LH secretion, which leads to higher levels of sex hormones in pubertal compared with prepubertal subjects. This eventually leads to the appearance of pubertal physical signs and accelerated growth [18]. With the development of newer and more sensitive immunoassays that measure serum gonadotropins, measurement of basal gonadotropins is hypothesized to allow discrimination between activated and inactivated values in HPG axis maturity.
我们的研究表明，基础 LH 水平可用于预测性腺轴激活。性腺电轴的激活对于诊断 CPP 很重要，CPP 可以加速骨成熟，导致成年身高受损和月经提前，并会极大地影响患者的心理健康 [12]。先前的研究表明，月经早发与晚年的不良健康结果有关 [13]、[14]、[15]。因此，及时诊断和适当治疗有助于改善这些患者的预后。诊断 CPP 很困难，应包括临床表现和正确及时评估 HPG 轴的激活。然而，通过简单的检查和临床数据在临床上诊断 HPG 轴的激活是具有挑战性的。目前，诊断性腺电轴激活的生化标准主要基于标准 GnRH 刺激试验期间的 LH 反应。在 GnRH 测试期间，刺激的 LH 值 ≥5 IU/L 和/或刺激的峰值 LH/FSH 比值 >0.6 被认为是青春期反应 [2]、[16]、[17]。青春期 LH 分泌的特征是 LH 峰值分泌水平高，与青春期前受试者相比，这导致青春期性激素水平更高。这最终导致青春期体征的出现和生长加速 [18]。随着测量血清促性腺激素的更新、更灵敏的免疫测定的发展，假设测量基础促性腺激素可以区分 HPG 轴成熟度的活化值和失活值。
In our single-center study, we investigated 1750 girls with early breast development. We found that basal LH values that were obtained during the GnRH stimulation test were significantly correlated with stimulated LH values. Additionally, LH values were useful as a screening test for predicting a positive response during the GnRH test. The highest Youden’s J index (0.40) was used to determine the appropriate cutoff LH value for diagnosing activation of the HPG axis. The basal LH cutoff point was 0.35 IU/L, with a sensitivity of 63.96% and specificity of 76.35%. When the basal LH value was 2.72 IU/L, the specificity reached 100%, although sensitivity decreased to 5.45%, which is higher than the value reported by Houk et al. [19]. They evaluated basal LH levels for discriminating activation of the HPG axis using two different chemiluminescent third-generation immunoassays (Wallac DELFIA and Architect) in 55 girls. Using the Architect assay, the LH cutoff point was 0.83 U/L, with a sensitivity of 93% and a specificity of 100%. Using the Delphi assay, the LH cutoff point was 1.05 U/L, with a sensitivity of 100% and a specificity of 100%. However, in the current study the basal LH cutoff level for evaluating activation of the HPG axis is different from previous studies. Pasternak et al. [20] measured basal serum LH and FSH levels using a chemiluminescent immunometric assay. They showed that low basal serum LH levels (≤0.1 IU/L) were sufficient for ruling out a positive response in the GnRH test in 94.7% of 38 prepubertal girls, with a sensitivity of 64%. Additionally, Suh et al. [21] reported cutoff values of basal LH (0.22 IU/L) that were measured using the sequential two-step immunoenzymatic assay (Access hLH, FSH Reagent Pack; Beckman Coulter, Inc., Brea, CA, USA). They detected a positive response of the GnRH stimulation test with 87.8% sensitivity and 20.9% specificity in 540 girls with clinical signs. They also demonstrated that basal FSH levels, basal E2 levels, and the basal LH/FSH ratio did not have predictive values for the diagnosis of CPP [21]. Mogensen et al. [22] showed that basal LH levels were superior in predicting the maximal LH level during GnRH testing compared with FSH, E2, and inhibin B levels. In another study, a total of 803 girls were included, and serum LH and FSH levels were measured using the immunoradiometric assay [23]. Based on the ROC curve, the optimal cutoff point for the basal LH level that was related to a pubertal response was 1.1 IU/L, which was associated with a sensitivity of 69.1% and specificity of 50.5%. Because of these different results among studies, clinicians must first determine the local cutoff before GnRH stimulation in patients with precocious puberty when applying this method.
在我们的单中心研究中，我们调查了 1750 名乳房发育早期的女孩。我们发现在 GnRH 刺激试验期间获得的基础 LH 值与刺激的 LH 值显著相关。此外，LH 值可用作预测 GnRH 测试期间阳性反应的筛选测试。最高的 Youden J 指数 （0.40） 用于确定诊断 HPG 轴激活的适当临界 LH 值。基础 LH 截断点为 0.35 IU/L，敏感性为 63.96%，特异性为 76.35%。当基础 LH 值为 2.72 IU/L 时，特异性达到 100%，但敏感性降至 5.45%，高于 Houk 等人报告的值 [19]。他们使用两种不同的化学发光第三代免疫测定法 （Wallac DELFIA 和 Architect） 评估了 55 名女孩的基础 LH 水平以区分 HPG 轴的激活。使用 Architect 分析，LH 截断点为 0.83 U/L，灵敏度为 93%，特异性为 100%。使用 Delphi 测定，LH 截断点为 1.05 U/L，灵敏度为 100%，特异性为 100%。然而，在目前的研究中，用于评估 HPG 轴激活的基础 LH 临界水平与以前的研究不同。Pasternak等[20]使用化学发光免疫测定法测量基础血清LH和FSH水平。他们表明，在 38 名青春期前女孩中，94.7% 的基础血清 LH 水平 （≤0.1 IU/L） 足以排除 GnRH 检测的阳性反应，敏感性为 64%。此外，Suh等[21]报道了基础LH的临界值（0.22 IU/L），使用序贯两步法免疫酶测定法（获取 hLH，FSH 试剂包;Beckman Coulter， Inc.， Brea， CA， USA）。他们在 540 名有临床症状的女孩中检测到 GnRH 刺激试验的阳性反应，灵敏度为 87.8%，特异性为 20.9%。他们还证明，基础 FSH 水平、基础 E2 水平和基础 LH/FSH 比值对 CPP 的诊断没有预测价值 [21]。Mogensen 等 [22] 表明，与 FSH 、 E2 和抑制素 B 水平相比，基础 LH 水平在预测 GnRH 测试期间的最大 LH 水平方面更胜一筹。在另一项研究中，共纳入 803 例女孩，使用免疫放射测定法测量血清 LH 和 FSH 水平 [23]。根据 ROC 曲线，与青春期反应相关的基础 LH 水平的最佳临界点为 1.1 IU/L，敏感性为 69.1%，特异性为 50.5%。由于研究之间的结果不同，临床医生在应用这种方法时必须首先确定性早熟患者 GnRH 刺激前的局部临界值。
In the present study, the AUC for LH was greater than that for FSH, the LH/FSH ratio, and E2. This finding indicated that basal LH levels were superior to FSH, the LH/FSH ratio, and E2 levels for determining activation of the HPG axis. Moreover, some researchers believe that determination of the LH/FSH ratio is helpful for improving the diagnostic accuracy of CPP [24]. However, FSH levels overlap between prepubertal and pubertal girls, which can affect the LH/FSH ratio and limit its application in evaluating activation of the gonadal axis. Our study showed that, when the cutoff value of basal LH levels was 0.35 IU/L, the sensitivity and specificity were 63.96% and 76.35%, respectively, which were relatively low. A basal LH value that reached 2.72 IU/L showed a specificity of 100%, but sensitivity decreased to 5.45%. These data indicated that increased basal LH levels were associated with a positive response to the GnRH test. Therefore, physicians should pay attention to basal LH testing in patients with early breast development. However, the appropriate cutoff value depends on sensitive measurement of basal gonadotropins and clinical manifestations. Therefore, conducting further GnRH stimulation tests might be necessary. Notably, evaluation of HPG axis activation based on LH cutoff values is not consistent between research centers. This could be due to hormone testing methods, apparatus, and GnRH stimulation test methods.
In conclusion, measurement of basal LH levels could be better than FSH levels, the LH/FSH ratio, and E2 levels for initial evaluation of HPG axis activation with clinically suspected early puberty. Increased basal LH values are a significant predictor of a positive response during the GnRH stimulation test for assessing activation of the HPG axis.
Acknowledgments
We thank Ellen Knapp from Liwenbianji (www.liwenbianji.cn) for linguistic assistance during preparation of this manuscript.
我们感谢来自 Liwenbianji （www.liwenbianji.cn） 的 Ellen Knapp 在准备本手稿期间提供的语言帮助。
Author contributions: H.X., L.S., and S.Y. designed the study; D.Y., L.J., Y.Y., Y P., and L.H. performed the study; D.Y. and L.S. drafted the manuscript and performed statistical analyses; L.S. and H. X. contributed to interpretation of the results and critically reviewed the manuscript; H.X. had primary responsibility for final content. All the authors have accepted responsibility for the entire content of this submitted manuscript and approved submission.
Research funding: This research was supported by the Shanghai Municipal Science and Technology Commission (Grant No. 12411950402), The Project of Shanghai Children’s Health Service Capacity Construction (GDEK201708), National Human Genetic Resources Sharing Service Platform (2005DKA21300), Science and Technology Development Program of Pudong Shanghai New District (PKJ2017-Y01), and Science Innovation Funding of Shanghai Jiaotong University School of Medicine (Z2016-02).
Employment or leadership: None declared.
Honorarium: None declared.
酬金：没有人宣布。Competing interests: The funding organization(s) played no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the report for publication.
利益争夺：资助组织在研究设计中没有发挥任何作用;在数据的收集、分析和解释方面;在撰写报告时;或决定提交报告以供发布。
References
1. Chen C, Zhang Y, Sun W, Chen Y, Jiang Y, et al. Investigating the relationship between precocious puberty and obesity: a cross-sectional study in Shanghai, China. BMJ Open 2017;7:e014004.10.1136/bmjopen-2016-014004Search in Google Scholar PubMed PubMed Central
2. Resende EA, Lara BH, Reis JD, Ferreira BP, Pereira GA, et al. Assessment of basal and gonadotropin-releasing hormone-stimulated gonadotropins by immunochemiluminometric and immunofluorometric assays in normal children. J Clin Endocrinol Metab 2007;92:1424–9.10.1210/jc.2006-1569Search in Google Scholar PubMed
3. Brito VN, Batista MC, Borges MF, Latronico AC, Kohek MB, et al. Diagnostic value of fluorometric assays in the evaluation of precocious puberty. J Clin Endocrinol Metab 1999;84:3539–44.Search in Google Scholar
4. Borges MF, Pacheco KD, Oliveira AA, Rita CV, Pacheco KD, et al. Premature thelarche: clinical and laboratorial assessment by immunochemiluminescent assay. Arq Bras Endocrinol Metabol 2008;52:93–100.10.1590/S0004-27302008000100013Search in Google Scholar
5. Harrington J, Palmert MR, Hamilton J. Use of local data to enhance uptake of published recommendations: an example from the diagnostic evaluation of precocious puberty. Arch Dis Child 2014;99:15–20.10.1136/archdischild-2013-304414Search in Google Scholar PubMed
6. Atta I, Laghari TM, Khan YN, Lone SW, Ibrahim M, et al. Precocious puberty in children. J Coll Physicians Surg Pak 2015;25:124–8.10.1186/1687-9856-2013-S1-P65Search in Google Scholar
7. Frisch H, Riedl S, Waldhor T. Computer-aided estimation of skeletal age and comparison with bone age evaluations by the method of Greulich-Pyle and Tanner-Whitehouse. Pediatr Radiol 1996;26:226–31.10.1007/BF01405306Search in Google Scholar PubMed
8. Eckert KL, Wilson DM, Bachrach LK, Anhalt H, Habiby RL, et al. A single-sample, subcutaneous gonadotropin-releasing hormone test for central precocious puberty. Pediatrics 1996;97:517–9.10.1542/peds.97.4.517Search in Google Scholar
9. Brito VN, Latronico AC, Arnhold IJ, Mendonca BB. A single luteinizing hormone determination 2 hours after depot leuprolide is useful for therapy monitoring of gonadotropin-dependent precocious puberty in girls. J Clin Endocrinol Metab 2004;89:4338–42.10.1210/jc.2003-031537Search in Google Scholar PubMed
10. Kim HK, Kee SJ, Seo JY, Yang EM, Chae HJ, et al. Gonadotropin-releasing hormone stimulation test for precocious puberty. Korean J Lab Med 2011;31:244–9.10.3343/kjlm.2011.31.4.244Search in Google Scholar PubMed PubMed Central
11. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics 1988;44:837–45.10.2307/2531595Search in Google Scholar
12. Carel JC, Leger J. Clinical practice. Precocious puberty. N Engl J Med 2008;358:2366–77.10.1056/NEJMcp0800459Search in Google Scholar
13. Dreyfus J, Jacobs DR Jr, Mueller N, Schreiner PJ, Moran A, et al. Age at menarche and cardiometabolic risk in adulthood: the Coronary Artery Risk Development in Young Adults Study. J Pediatr 2015;167:344–52.e1.10.1016/j.jpeds.2015.04.032Search in Google Scholar
14. Mueller NT, Duncan BB, Barreto SM, Chor D, Bessel M, et al. Earlier age at menarche is associated with higher diabetes risk and cardiometabolic disease risk factors in Brazilian adults: Brazilian Longitudinal Study of Adult Health (ELSA-Brasil). Cardiovasc Diabetol 2014;13:22.10.1186/1475-2840-13-22Search in Google Scholar
15. Lim SW, Ahn JH, Lee JA, Kim DH, Seo JH, et al. Early menarche is associated with metabolic syndrome and insulin resistance in premenopausal Korean women. Eur J Pediatr 2016;175:97–104.10.1007/s00431-015-2604-7Search in Google Scholar
16. Neely EK, Hintz RL, Wilson DM, Lee PA, Gautier T, et al. Normal ranges for immunochemiluminometric gonadotropin assays. J Pediatr 1995;127:40–6.10.1016/S0022-3476(95)70254-7Search in Google Scholar
17. Subspecialty Group of Endocrinologic, Hereditary and Metabolic Diseases, the Society of Pediatrics, Chinese Medical Association; Editorial Board, Chinese Journal of Pediatrics. [Consensus statement for the diagnosis and treatment of central precocious puberty (2015)]. Zhonghua Er Ke Za Zhi 2015;53:412–8.Search in Google Scholar
18. Iughetti L, Predieri B, Ferrari M, Gallo C, Livio L, et al. Diagnosis of central precocious puberty: endocrine assessment. J Pediatr Endocrinol Metab 2000;13(Suppl 1):709–15.10.1515/JPEM.2000.13.S1.709Search in Google Scholar
19. Houk CP, Kunselman AR, Lee PA. Adequacy of a single unstimulated luteinizing hormone level to diagnose central precocious puberty in girls. Pediatrics 2009;123:e1059–63.10.1542/peds.2008-1180Search in Google Scholar PubMed
20. Pasternak Y, Friger M, Loewenthal N, Haim A, Hershkovitz E. The utility of basal serum LH in prediction of central precocious puberty in girls. Eur J Endocrinol 2012;166:295–9.10.1530/EJE-11-0720Search in Google Scholar PubMed
21. Suh J, Choi MH, Kwon AR, Kim YJ, Jeong JW, et al. Factors that predict a positive response on gonadotropin-releasing hormone stimulation test for diagnosing central precocious puberty in girls. Ann Pediatr Endocrinol Metab 2013;18:202–7.10.6065/apem.2013.18.4.202Search in Google Scholar PubMed PubMed Central
22. Mogensen SS, Aksglaede L, Mouritsen A, Sorensen K, Main KM, et al. Diagnostic work-up of 449 consecutive girls who were referred to be evaluated for precocious puberty. J Clin Endocrinol Metab 2011;96:1393–401.10.1210/jc.2010-2745Search in Google Scholar PubMed
23. Lee HS, Park HK, Ko JH, Kim YJ, Hwang JS. Utility of basal luteinizing hormone levels for detecting central precocious puberty in girls. Horm Metab Res 2012;44:851–4.10.1055/s-0032-1321905Search in Google Scholar PubMed
24. Supornsilchai V, Hiranrat P, Wacharasindhu S, Srivuthana S, Aroonparkmongkol S. Basal luteinizing hormone/follicle-stimulating hormone ratio in diagnosis of central precocious puberty. J Med Assoc Thai 2003;86(Suppl 2):S145–51.Search in Google Scholar
©2018 Walter de Gruyter GmbH, Berlin/Boston
