Premature thelarche, premature adrenarche or pubarche, precocious puberty

1. What every clinician should know

To recognize the signs of premature sexual development, the clinician must be knowledgeable of the normal physical findings of puberty. Table I outlines the mean ages of the classic physical findings of puberty; thelarche, pubarche and menarche. Approximately 50% of girls enter puberty with synchronous thelarche and pubarche, usually within 3-6 months. In the other half, about 66% of girls initially present with thelarche and 34% with pubarche.


Ethnicity and Age of Pubertal Milestones.

Girls entering through the thelarche pathway tend to have earlier menarche and greater percentage of body fat mass. The ethnic differences in the onset of puberty are mostly accounted for by differences in body fat content further discussed below. Typically, the time from thelarche to menarche is 2-3 years. The younger the age of pubertal onset, the longer the time span to menarche. In addition, girls who have an earlier menarche have a shorter duration until ovulatory cycles.

Continue Reading

Menarche does not necessarily reflect ovulation. Up to 55% of menses within the first 2 years of menarche are associated with anovulatory cycles, with this number reducing to 20% within 5 years. As a result, menses during the first few years after menarche tend to be more irregular. Within 3-5 years of menarche, most women establish a menstrual pattern that will be characteristic of their reproductive years, barring changes in weight or medical status.

Physical manifestations of puberty, originally described by Marshall and Tanner in 1969, outline 5 stages in breast (B1-5) and pubic hair (PH1-5) development and are discussed in more detail below.

The age of sexual development differs with ethnicity as shown in Table I, with these differences correlating with body fat percentage. Classically, early development was considered thelarche prior to age 8 years or adrenarche prior to age 9 years. The term adrenarche is often used interchangeably with pubarche, as the main clinical sign of adrenarche is pubic hair growth. Technically, adrenarche refers to the age-dependent increase in adrenal androgen precursor production around age 6-7 years. In general adrenarche is independent of gonadarche.

Due to the earlier onset of puberty in recent times, some experts have recommended that these guidelines be changed to thelarche prior to age 7 in White girls and prior to age 6 in African-Americans and adrenarche prior to age 8 years. To decrease the chance of missing significant pathology at these borderline ages, the following three conditions also warrant evaluation:

  • Rapid progression of puberty (bone age greater than 2 years ahead of chronologic age and a predicted height of more than 150 cm or 2 standard deviations or more below the genetic target height).

  • Central nervous system (CNS) findings, such as headaches, seizures or other neurological symptoms.

  • Pubertal progression that affects the emotional health of the family or the girl.

To determine these conditions, a minimal evaluation including history, physical examination and bone age is necessary. As one would expect, the younger the child, the more likely there is a pathological cause for premature sexual development.

Physical stages of pubertal development

The original description of pubertal changes in breast and pubic hair development were determined by Drs. Marshall and Tanner based upon photographs taken every 3 months of British girls of lower socioeconomic status living in group homes. The stages were re-evaluated in 1997 in a U.S.-based study of 17,000 girls based on physical examinations. Breast (B) and pubic hair (PH) development are described in stages 1-5. In regards to breast development: B1 is no development; B2 is characterized by increased size of the areolae and presence of a small mound (breast bud) under the areolae; B3 includes enlargement of the breast above the chest wall; B4 shows further enlargement with a secondary areolar mound above the breast tissue (not seen in all girls); and B5 is the mature breast with recession of the areolae to the same level as the remaining breast tissue. PH1 represents no pigmented pubic hair; PH2 shows pigmented hair on the labia majora; PH3 is characterized by spread of hair onto the mons pubis; PH4 demonstrates a fully covered mons; while PH5 is characterized by spread onto the inner upper thigh.

Body fat and puberty

The timing of puberty substantially changed from the mid-1800s to the mid-1900s, with menarche occurring an average of 4 years earlier. This is related to improved nutrition with a resultant increase in body fat percentage. The decreasing age of puberty has mostly plateaued, with a decline of only 4 months in the last 60 years. There is a strong correlation between the onset of puberty and body fat content, with overweight girls tending to have earlier puberty and thin girls later puberty. The explanation for this relationship is found in the production of adipokines, in particular leptin.

In mice, lack of leptin is associated with lack of pubertal development. In girls, leptin levels correlate with body fat mass and gradually increase during puberty. Case reports of leptin deficiencies in girls demonstrate a lack of gonadotropin secretion, which may be reversed with leptin therapy. It is evident that leptin is necessary for puberty, but it remains to be determined whether leptin initiates puberty. The mechanism of leptin action on GnRH neurons appears to be indirect, possibly acting via the kisspeptin system. Hypothalamic neurons produce kisspeptin from expression of the Kiss1 gene, which activates a G-protein coupled receptor (GPR54 or Kiss1R) on GnRH neurons to stimulate GnRH production. Leptin increases kisspeptin expression. More thorough reviews of the neuroendocrinology of puberty are available.

2. Diagnosis and differential diagnosis

Three forms of isosexual precocious development include premature thelarche (PT), premature adrenarche (PA) and precocious puberty (PP) (See Figure 1).

Figure 1.

Figure 1. Examples of premature sexual development. (A) A completely developed 8 yo with Tanner B5 and PH 5. (B) A 5 yo with premature thelarche characterized by Tanner B3 and PH1. (C, D) A 6 yo with premature adrenarche characterized by Tanner B1, PH3 and axillary hair growth. (Photos provided by G. William Bates, MD, permission of the American College Obstetrics Gynecology).

PT involves breast development as the only physical sign. Evaluation includes history, physical examination, review of growth records, bone age, pelvic ultrasound and testing for FSH, LH, estradiol and TSH. The bone age may be slightly advanced but does not exceed 2 SD. Pelvic ultrasound shows prepubertal uterine and ovarian dimensions. Baseline hormone tests are characteristically normal. GnRH stimulation testing shows an FSH dominant response. The classic case of PT presents around 2 years of age and rarely progresses to PP. Older children presenting with PT may be showing the first signs of subsequent PP and need to be carefully observed.

PA presents with development of pubic hair, axillary hair and adult body odor due to apocrine gland activity. This appears due to early maturation of the androgen-producing enzymes 17α-hydroxylase and 17, 20 lyase with low 3β-hydroxysteroid dehydrogenase activity. Evaluation includes history, physical examination, review of growth records, bone age and testing for DHEAS, testosterone, am 17-hydroxyprogesterone and TSH. The bone age may be slightly advanced, but rarely exceeds 2 SD and the predicted adult height is usually normal. Laboratory testing shows an increased DHEAS for age, but usually normal for the stage of pubic hair development. PA may be predictive of adult PCOS.

Children with PA may have increased insulin resistance, dyslipidemia and increased vascular inflammatory markers compared to peers. This is especially seen in a subset of children born with growth restriction followed by childhood obesity, referred to as “fetal deprivation with childhood catch-up.” A trial in 2006 of empiric therapy with metformin in children with PA showed less adiposity, decreased IgF-1 levels and a slight delay in the clinical onset of puberty compared to non-treated controls. The differential diagnosis for PA includes congenital adrenal hyperplasia, adrenal or ovarian androgen-producing tumors and Cushing syndrome.

PP may present with combinations of breast development, pubic hair development and uterine bleeding. Breast development is essentially always present, being the most sensitive physical sign of estradiol production. Typically the normal sequence of puberty is preserved, but the time intervals are shortened. However, this may not always be the case. Pubic hair is often seen in older children, but can be scant or absent in very young children with PP. Uterine bleeding is commonly seen with PP, but does not necessarily imply ovulation, as bleeding can be due to estrogen-stimulated endometrial proliferation. The causes of PP are outlined in Table II and may be divided into GnRH-dependent (central) and GnRH-independent (peripheral) etiologies.

Table II.

Causes of Precocious Sexual Development

In females, the cause of PP is idiopathic in over 50% of cases. GnRH-dependent PP essentially involves early activation of the HPO axis typically due to an intracranial process such as tumor, inflammation or hydrocephalus. A hypothalamic hamartoma (Figure 2) is one of the most common tumors resulting in CPP. In this case, neurons within the tumor may secrete transforming growth factor, leading to stimulation of adjacent GnRH neurons as the etiology. Interestingly, CPP is increased in children from international adoptions. The rate is higher in children as opposed to infants and may be due to catch-up weight from improved nutrition.

Figure 2.

Figure 2. Brain MRI showing a hypothalamic hamartoma causing CPP. Provided by Dr. M. Tariq Bhatti, Departments of Ophthalmology and Medicine, Duke University Medical Center.

Peripheral PP may be due to an ovarian tumor secreting estradiol, with granulosa-theca cell being the most common (See Figure 3). Rarely, spontaneous development of a single functional ovarian cyst has been associated with PP. Spontaneous resolution of the cyst is common resulting in correction of the PP. Uncommonly, cyst aspiration maybe required. Primary hypothyroidism may be associated with PP, most likely due to cross-reactivity of TSH with the FSH receptor. Hypothyroidism is the only cause of PP associated with a delayed bone age. Multiple ovarian follicular cysts are also commonly seen with hypothyroidism.

Figure 3.

Figure 3. Granulosa cell tumor. Low magnification (A) and high magnification (B) showing multuiple Call-Exner bodies (eosinophilic secretions between granulosa cells). Photos contributed by Rex Bentley, MD, Dept. of Pathology, Duke University.

McCune-Albright syndrome is a rare cause of peripheral PP with the classic symptoms of polyostotic fibrous dysplasia (cystic bone lesions), large pigmented skin lesions (cafe au lait patches) and PP (See Figure 4). Bone lesions involve replacement of normal bone with masses of fibroblasts and can involve any bone, such as the extremities leading to fractures, or the skull causing disfigurement. McCune-Albright syndrome is caused by a post-zygotic mutation of the Guanine Nucleotide binding protein, Alpha Stimulating activity polypeptide 1 (GNAS1) gene leading to unregulated activation of cAMP. A post-zygotic mutation leads to a mosaicism of affected and unaffected cells, resulting in great variation of which organ systems are ultimately affected. Other disorders may include hyperthyroidism, acromegaly, hypercortisolism, hypophosphatemia, fatty liver and cardiomegaly. Overall, peripheral PP is rare, estimated at 1.1 in 100,000 girls. The most common causes are a functional ovarian cyst and McCune-Albright syndrome.

Figure 4.

Figure 4. Large cafe au lait patch and B4 development in a 10 y.o. with presumed McCune-Albright syndrome. Photos contributed by John Nichols, Jr., MD, Piedmont Reproductive Endocrinology Group, Greenville, S.C.

Initial evaluation of PP includes history, physical examination, review of growth records, radiographic bone age, pelvic ultrasound, and laboratory testing of estradiol, FSH, LH, TSH and inhibin B. Use of a laboratory capable of determining low levels of estradiol is necessary, as the usual adult lower limit of 20 pg/mL may not be helpful. In typical PP growth records will show an increase in height velocity by a change in percentile (See Figure 5) and the bone age will be advanced greater than 2 SD, the only exception being hypothyroidism.

Figure 5.

Figure 5. Growth chart showing increase in height percentile (upper graph).

Yet, the degree of bone age advancement depends on the chronicity of the process and may be normal or slightly advanced if the diagnosis is made early. A pelvic ultrasound is useful to further determine the degree of estrogen effect and to evaluate for ovarian abnormalities (See Figure 6).

Figure 6.

Figure 6. Pelvic ultrasound of a child with CPP showing developing follicles on the left ovary (upper left panel), a hemorrhagic cyst on the right ovary (upper right panel) and an increased uterine fundus to cervix ratio (lower left panel).

Uterine growth is extremely sensitive to estradiol, with changes in the cervix to fundus ratio and in the dimensions of the fundus ( Ovarian imaging commonly shows developing follicles and/or follicular cysts and increased ovarian volume. The normal childhood ovary is homogeneous up to age 7-8 at which time microcysts (less than 10 mm antral follicles) appear. Greater than 10 mm follicles normally develop around the time of normal menarche.

An elevated estradiol level is concerning for an ovarian tumor, especially if associated with an elevated inhibin B level. In contrast, a normal estradiol is only helpful in excluding an estrogen-producing tumor and in no way negates the diagnosis of PP. Currently, the better clinical laboratories have a lower limit of estradiol detection of 1.0 pg/mL. In a previous study utilizing an ultrasensitive bioassay, the normal mean estradiol level in a 7 y.o. female was 0.6 pg/mL.

Gonadotropin assays also vary among laboratories and the lower level of detection and the reference used for standardization must be known. As the normal prepubertal LH is less than 0.1 IU/L, an assay with appropriate lower limit of detection is needed. Similarly, the gonadotropin levels are only useful if in the adult range, generally greater than 2 IU/L, suggestive of central PP. TSH levels are similar in children and young adults and thus the same reference range can be used. A child with PP with normal estradiol and gonadotropins will require a GnRH stimulation test to further define the activity of the HPO axis.

GnRH or GnRH agonist stimulation of the pituitary results in a predictable pattern of LH and FSH release. A normal child essentially has no or minimum increase in gonadotropins. A child with PT will have an increase in FSH but no significant change in LH. A normal adult or child with CPP will have an increase in both gonadotropins, but LH greater than FSH (Figure 7). A diagnosis of CPP is made with a peak LH (WHO second international standard) of greater than 8 IU/L using a stimulation with 100 μg of IV or SQ GnRH. Due to the unavailability of GnRH in the USA, stimulation testing is performed with the GnRH agonist, leuprolide acetate. In patients with CPP, 500 μg of SQ leuprolide acetate results in a peak LH at 3-6 hrs post-injection of greater than 8 IU/L. FSH levels are not as useful as LH levels for the diagnosis of CPP.

Figure 7.

Figure 7. Examples of GnRH stimulation test results in a child with PT (top), CPP ( middle) and a child with CPP undergoing treatment with leuprolide acetate depot (bottom).

The decision to perform an MRI on a patient with PP is based upon the exclusion of peripheral causes of PP, adult levels of gonadotropins, a positive GnRH stimulation test or other symptoms suggesting an intracranial process. Approximately 8% of girls with CPP have an unrecognized intracranial process.

GnRH Stimulation Testing

The classic GnRH stimulation test was performed by administering 100 μg GnRH (Factrel) IV and measuring LH and FSH at 0, 20, 40 and 60 min. Typically, LH levels peak at 20 min while FSH levels peak at 20-40 min. Subsequently, an abbreviated test was found to be equally effective with a 100 μg SQ administration of Factrel followed by a single blood draw at 40 min. The unavailability of this product in the U.S. led to alternate GnRH stimulation tests with a GnRH analogue.

In the U.S., this is most commonly performed with 20 μg/kg SQ administration of leuprolide acetate (Lupron). The increase in gonadotropins remains relatively stable from 2-24 hours, thus allowing for a single determination when convenient. Baseline gonadotropin values should be obtained just prior to Lupron administration. The commonly accepted value for a positive test is LH greater than 8 IU/L.

Vaginal examination

The tendency for vaginal examination of a child with precocious puberty varies among practitioners of different specialties. Certainly any deviation from the typical pattern of pubertal bleeding needs evaluation. This would include:

  • Vaginal bleeding without any breast development, bringing into question the diagnosis of PP.

  • Other vaginal abnormalities such as non-physiologic discharge.

  • Irregular bleeding with no pattern.

The method of examination also varies depending upon the comfort and expertise of the physician and the desires of the parents. Many physicians prefer examination under anesthesia, which is certainly easiest for the examiner.

Another option is an in-office vaginoscopy with a gynecological flexible hysteroscope. We have found this quite easy to perform in most children without any sedation. Warm normal saline is used as a distension media and distension is enhanced by pinching the labia shut with the examiner’s fingers. The key to success is to avoid touching the vaginal walls with the scope as this results in discomfort. As the main goal is to exclude a foreign body or tumor, this procedure can usually be done quite quickly (Figure 8).

Figure 8.

Figure 8. Cervix visualized in a 2 y.o. undergoing vaginoscopy for vaginal bleeding.

3. Management

The management for PT is simple observation and reassurance to the parents. Observation is very important in that only time will tell whether this is isolated PT or the first clinical evidence of subsequent PP.

The management for PA is also observation, as long as other etiologies such as CAH have been excluded. Consideration should be given to early evaluation of insulin sensitivity and lipid levels, especially in obese children or children with acanthosis nigricans. Of course weight optimization is critical.

The management of PP is dependent on the etiology, such as thyroid replacement in hypothyroidism or oophorectomy with a granulosa cell tumor. In cases of CPP in which the cause cannot be reversed, treatment with a GnRH agonist is based upon the preservation of final height and the need to reverse pubertal symptoms for social and psychological benefit. For example, the need for treatment in a 2 y.o. child is unquestionable, but the benefits may need to be carefully considered in an 8 y.o. child.

Final height optimization is based on the remaining growth potential when therapy is initiated. Thus, maximum final height negatively correlates with initial bone age (BA), age at the start of therapy and a delay in therapy; and positively correlates with the Height SD/BA. The term “Height SD/BA” refers to the standard deviation of the current height versus the normal height for the given BA. Thus, a less negative number shows preserved height for BA and is a positive predictor for final height.

Generally, the earlier the presentation of CPP, the less the final predicted height; and the earlier the time of onset of treatment, the greater the final height. Final height is estimated by the Bayley-Pinneau method, which is based on the bone age and current height. Bone age is determined by radiograph examination of the hand and wrist originally described by Greulich and Pyle. The average final height of American girls is 168.3 (5″ 4.5″) and using the Bayley-Pinneau method, 99% of girls are within 6.3 cm of predicted final height.

One criticism has been that the Bayley-Pinneau method tends to over predict final height in children with PP having a greatly advanced BA for chronological age (CA). Although there is obviously a lot of variation, the typical height of an untreated girl with CPP is approximately 152 cm (5 feet). Treatment with a GnRH agonist improves final height, but the magnitude is dependent upon the height for bone age at the onset of therapy. For example, a child at 120 cm with a bone age of 8 has much more potential than a child at 130 cm with a bone age of 13. In general, girls with initiation of treatment prior to age 6 years gain 9-10 cm while those from 6 to 8 y.o. gain 4 to 7 cm.

Treatment in the U.S. is commonly with depot leuprolide acetate (Lupron-Depot). Unlike typical treatment for endometriosis or leiomyomata, the goal is for suppression of the HPO axis close to a pre-pubertal state. A typical starting dose is 0.2-0.3 mg/kg, usually between 7.5 and 15 mg of the monthly depot preparation. The fixed dose 3-month preparation (11.25 mg) does not suppress pituitary gonadotropin secretion as well as the 7.5 mg month preparation.

Histrelin acetate is available as a 50 mg implant placed into the subcutaneous tissue of the inner upper arm releasing 65 mg/day for 12 months. Anesthetic choice has to be individualized and may include general, conscious sedation or local with distraction depending on patient age and maturity. Common side-effects of placement include bruising and transient localized pain. Due to scar tissue encapsulation, the implants can be challenging to remove. Standardly, HPO suppression is obtained 1 month after placement of the implant. HPO axis suppression may be verified by a GnRHa stimulation test approximately 1 month after initiation of treatment and every 6 months. After cessation of therapy, HPO axis normalization is seen by 6 months.

The goal of therapy is to maintain a stimulated LH level of less than 2 IU/L. A simpler way to evaluate treatment is by obtaining an LH level 30-60 mins after the monthly dose of depot leuprolide acetate, which serves the same purpose as a GnRHa stimulation test. GnRHa therapy is discontinued around 11 years of age and/or a bone age of 12-13 years.

Other drugs have been considered in conjunction with leuprolide acetate to increase final height including growth hormone and oxandrolone, but are not currently considered standard of care.

Treatment of McCune-Albright syndrome is more difficult, requiring the use of an estrogen receptor blocker or aromatase inhibitor. Both tamoxifen and letrozole decrease bleeding and linear growth. A bisphosphonate is often used for treatment of fibrous dysplasia.

Treatment of peripheral PP associated with a single ovarian cyst is simple observation. Other therapy may rarely need to be considered in cases not resolving in 3 months. In some cases, recurrent single ovarian cysts have been seen leading to persistently elevated estradiol levels. This persistent elevation in estradiol may activate the HPO axis leading to a GnRH dependent PP referred to as “combined PP.” In this circumstance treatment with a GnRHa may be necessary.

4. Complications

Complications as a consequence of the condition

The main complications of CPP include decreased final height and the social and psychological implications of early puberty. Girls with CPP have normal intelligence and no increased incidence of major psychiatric illness. Girls may have concerns over physical changes not seen in their peers and commonly overestimate their final height. The concern that girls with CPP have an early proclivity to sexual activity does not appear to be correct. Timing of sexual activity tends to correlate with chronological age and not puberty stage. Childhood and adult obesity are more common with CPP. In addition, premature adrenarche may predict the development of PCOS, especially in children with IUGR followed by catch-up obesity. These children have increased risk of insulin resistance and dyslipidemia.

Complications as a consequence of management

GnRHa are generally well tolerated. Vasomotor symptoms and vaginal dryness commonly seen in adults are not typical in children. Local reactions to the injection are possible, along with rare cases of anaphylaxis. Treatment with a GnRHa does not affect the ultimate gonadal function of girls with CPP, nor the risk or extent of obesity. GnRHa therapy does adversely decrease bone density, but since it is discontinued at an early age, final bone density is not affected.

5. Prognosis and outcome

The prognosis for PP depends upon etiology. With respect to CPP, treatment with a GnRH agonist quickly results in cessation of uterine bleeding along with regression of breast development and pubic hair, although these may not totally disappear. Reactivation of the HPO axis occurs within a few months of discontinuation of the GnRHa. GnRHa treatment does increase final height as discussed above, but most individuals do not reach the same height that they would have normally. There is no evidence to suggest decreased fertility, diminished ovarian reserve or early menopause in women with a history of CPP, but these questions have not been well researched.

6.What is the evidence for specific management and treatment recommendations

Carel, J.-C., Eugster, E. A.. (Most current consensus of the treatment of CPP with GnRH agonists.)

Marshall, W, Tanner, J. “Variations in pattern of pubertal changes in girls”. Arch Dis Child. vol. 44. 1969. pp. 291-303. (Original article describing physical puberty stages.)

Styne, DM. “Puberty, obesity and ethnicity”. Trends in Endocrinology and Metabolism. vol. 15. 2004. pp. 472-8. (Review of the relationships of ethnicity and adiposity to pubertal onset.)

Greulich, WW, Pyle, SL. Radiographic atlas of skeletal development of the hand and wrist. 2nd Edition. 1959. pp. 61-183. (Classic textbook describing bone age measurements with Bailey-Pinneau tables in the appendix.)

Kaplowitz, PB, Oberfield, SE. “Re-examination of the Age Limit for Defining When Puberty Is Precocious in Girls in the United States; Implications for Evaluation and Treatment”. Pediatrics. vol. 104. 1999. pp. 936-41. (Consensus statement of the ages defining precocious puberty.)

Roa, J, Tena-Sempere, M. “Energy balance and puberty onset: emerging role of central mTOR signalling”. Trends in Endocrinology & Metabolism. vol. 21. 2010. pp. 519-28. (Excellent review of the neuroendocrinology of puberty.)

Herman-Giddens, ME, Slora, EJ, Wasserman, RC. “Secondary sexual characteristics and menses in young girls seen in office practice: A study from the Pediatric Research in Office Settings Network”. Pediatrics. vol. 99. 1997. pp. 99-505.

Klein, K, Baron, J. “Estrogen levels in childhood determined by an ultrasensitive recombinant cell bioassay”. J Clin Invest. vol. 94. 1994. pp. 2475-80. (Most accurate determination of prepubertal estradiol level.)

Ibanez, L, Ong, K, Valls, C, Marcos, MV, Dunger, DB, de Zegher, F. “Metformin treatment to prevent early puberty in girls with precocious pubarche”. J Clin Endocrinol Metab. vol. 91. 2006. pp. 2888-91. (Article discussing insulin resistance related to PA.)

Wu, T, Mendola, P, Buck, GM. “Ethnic differences in the presence of secondary sex characteristics and menarche among US girls: The third national health and nutritional health survey, 1988-1994”. Pediatrics. vol. 110. 2002. pp. 752-7. (Describes ethnic differences in pubertal development.)