7.2.10 Premature sexual maturation

Premature sexual maturation is a frequent cause for referral in paediatric endocrinology. Although clinical evaluation will suffice to reassure the patient and family in a majority of cases, premature sexual maturation can reveal severe conditions and need a thorough evaluation to identify its cause and potential for progression, in order to propose an appropriate treatment (1). Although the use of long-acting GnRH agonists has revolutionized the treatment of central precocious puberty, questions remain regarding their optimal use (2). One of the main ongoing controversial issues in the area is the definition of normal pubertal development and there is a need for longitudinal assessments of normally developing children in the various areas of the world and of a better understanding of the factors affecting normal pubertal development to improve the recognition and proper management of premature sexual maturation.

Normal pubertal development results from the activation of pulsatile GnRH secretion and of the activation of the hypothalamo-pituitary-gonadal axis. The onset of puberty is marked clinically by breast development in girls and testicular enlargement in boys. Tanner stages (Fig. 7.2.10.1) are used to evaluate pubertal development and the onset of puberty corresponds to Tanner 2 breast (B2) stage in girls (best assessed by both inspection and palpation) and Tanner 2 genitals (G2) stage in boys (testicular volume greater than 4 ml or testicular length greater than 25 mm).

Defining the normal limits of pubertal development is difficult, given the paucity of truly normative data and the number of components to consider including not only pubertal onset, but also progression of puberty and onset of menarche. Cross-sectional data obtained in the 1960s led to the designation of the normal age range of pubertal onset (the age at which 95% of children reach Tanner stage 2) as between 8 and 13 years in girls, and 9.5 and 13.5 years in boys. Cross-sectional data obtained in the USA have shown that pubertal milestones were being reached earlier than previously thought by African American and, to a lesser extent, by Mexican American or non-Hispanic white girls (3). A similar tendency has also been noted in Europe (4) and in Asia (5). In Copenhagen, the mean age at the B2 stage has decreased from 10.9 to 9.9 years between 1991 and 2006 (4). Although there have been discussions to decrease the cut-off defining early pubertal onset in girls (6), the traditional limits of 8 years in girls and 9.5 years in boys are still used by most paediatric endocrinologists (1, 2). Sexual hair development is a component of pubertal maturation that reflects the actions of androgen produced by the gonads or by the adrenals. Similarly, the traditional limits of pubic hair development have been set to 8 years in girls and 9.5 years in boys with wide ethnical variations.

There are several elements to remember when considering normal and abnormal pubertal development. First, the activation of the gonadotropic axis is not an all-or-nothing phenomenon, but evolves over several years, starting 2–3 years before the clinical onset of puberty. Second, the mean duration of the transition from one stage to the next is generally close to 6 months on average, but varies among individuals. In slowly progressive puberty, pubertal development can remain at the B2 stage or revert to the B1 stage before resuming later. It is noteworthy that, although the mean age at the B2 stage has decreased in the past decades, the age at menarche has been relatively stable, indicating a longer duration of puberty (4). Thirdly, the onset of puberty is affected by a number of factors in addition to ethnicity (7). Puberty occurs earlier in girls with early maternal menarche, low birthweight, excessive weight gain or obesity in infancy and early childhood, after international adoption (10–20 times increase in risk for unclear reasons (8)) and possibly after exposure to estrogenic endocrine-disrupting chemicals or if no father is present in the household (1, 7). These factors are generally not considered in definitions of normality in practice, but should be kept in mind. It is important to recognize that a ‘normal’ timing of onset of pubertal development does not rule out a pathological condition (9). The prevalence of precocious puberty is about 10 times higher in girls than in boys, and has been estimated at 0.2% of girls and less than 0.05% of boys in Denmark (10).

Figure 7.2.10.2 summarizes the mechanisms underlying premature sexual development. Premature sexual development results from the action of sex steroids or compounds with sex steroid activity on target organs. The most common mechanism of progressive precocious puberty is the early activation of pulsatile GnRH secretion, i.e. central or gonadotropin-dependent precocious puberty. Peripheral or gonadotropin-independent precious puberty is due to the production of sex steroids by gonadal or adrenal tissue, independently of gonadotropins, which are generally suppressed. Exposure to exogenous sex steroids or to compounds with steroidal activity can also result in premature sexual development. It is also important to recognize variants of pubertal development that can mimic precocious puberty, but do not lead to long-term consequences and are usually benign.

Central precocious puberty is due to the premature activation of GnRH secretion and results in an hormonal pattern that is similar to that of normal puberty, although early. Central precocious puberty can be due to hypothalamic tumours or lesions or be idiopathic in the majority of cases, in particular in girls (Table 7.2.10.1).

Peripheral precocious puberty can result from gonadal, adrenal, or hCG-producing tumours, activating mutations in the gonadotropic pathway and exposure to exogenous sex steroids (Table 7.2.10.2). Peripheral precocious puberty can rarely lead to activation of pulsatile GnRH secretion and to central precocious puberty (Table 7.2.10.1).

It is essential to recognize that most cases of premature sexual maturation correspond to benign variants of normal development that can occur throughout childhood (Table 7.2.10.3). This is particularly true in girls below the age of 2–3 years where the condition is known as premature thelarche. Similarly, in older girls, at least 50% of cases of premature sexual maturation will regress or stop progressing, and no treatment is necessary. Although the mechanism underlying these cases of non-progressive precocious puberty is unknown, the gonadotropic axis is not activated. Premature thelarche probably represents an exaggerated form of the physiological early gonadotropin surge that is delayed in girls relatively to boys.

Progressive premature sexual maturation can have consequences on growth and psychosocial development. Growth velocity is accelerated as compared with normal values for age and bone age is advanced in most cases. The acceleration of bone maturation can lead to premature fusion of the growth plates and short stature. Several studies have assessed adult height in individuals with a history of precocious puberty. In older published series of untreated patients, mean heights ranged from 151 to 156 cm in boys and 150 to 154 cm in girls, corresponding to a loss of about 20 cm in boys and 12 cm in girls relative to normal adult height (11). However, these numbers correspond to a historical series of patients with severe early onset precocious puberty, which are not representative of the majority of patients seen in the clinic today. Height loss due to precocious puberty is inversely correlated with the age at pubertal onset, and currently treated patients tend to have later onset of puberty than those in historical series (11).

Parents often seek treatment in girls because they fear early menarche (12). However, there is little data to predict the age of menarche following early onset of puberty. In the general population, the time from breast development to menarche is longer for children with an earlier onset of puberty, ranging from a mean of 2.8 years when breast development begins at age 9 to 1.4 years when breast development begins at age 12 (13).

Adverse psychosocial outcomes are also a concern, but the available data specific to patients with precocious puberty have serious limitations (2). In the general population, a higher proportion of early-maturing adolescents engage in exploratory behaviours (sexual intercourse, legal, and illegal substance use) and at an earlier age than adolescents maturing within the normal age range or later (14). In addition, the risk for sexual abuse seems to be higher in girls or women with early sexual maturation (15). However, the relevance of these findings to precocious puberty is unclear, and they should not be used to justify intervention.

The evaluation of patients with premature sexual development should address several questions: (1) Is sexual development really occurring outside the normal temporal range? (2) What is the underlying mechanism and is it associated with a risk of a serious condition, such as an intracranial lesion? (3) Is pubertal development likely to progress, and (4) would this impair the child’s normal physical and psychosocial development?

A complete family history (age at onset of puberty in parents and siblings) and personal history, including the age at onset and progression of sexual development, should be taken. Any evidence suggesting possible central nervous system disorder, such as headache, increased head circumference, visual impairment, or seizures should be collected. Growth should be evaluated by drawing a complete growth chart, because progressive precocious puberty is almost invariably associated with and sometimes preceded by an acceleration of growth velocity.

The stage of pubertal development should be classified according to Tanner (Fig. 7.2.10.1). Careful assessment is needed in obese girls to avoid overestimating breast development. The development of pubic hair results from the effects of androgens, which may be produced by testes or ovaries in central precocious puberty. Acne, oily skin, and hair may be present and result from the action of androgens. In girls, pubic hair in the absence of breast development is suggestive of adrenal disorders, premature pubarche, or exposure to androgens. In boys, measurement of testicular volume may suggest the cause of puberty, as testicular volume increases in central precocious puberty as in normal puberty and in cases of peripheral precocious puberty due to testicular disorders (although generally less so), whereas it remains prepubertal in adrenal disorders, premature pubarche, and other causes of peripheral precocious puberty. Physical examination should also assess for signs of specific causes of precocious puberty, such as hyperpigmented skin lesions suggesting neurofibromatosis or McCune–Albright syndrome. It is also important to recognize clinically the benign variants of precocious pubertal development with usually isolated and non progressive secondary sexual characteristic (breast or pubic hair), normal or slightly increased growth velocity and no or slight bone age advancement, if performed (Table 7.2.10.3).

Premature sexual development can be associated with high levels of anxiety in girls, and psychological evaluation of the child and of the familial environment is important.

Additional testing is generally recommended in all boys with precocious pubertal development, in girls who present with precocious Tanner 3 breast stage or higher, or in girls with precocious B2 stage, and additional criteria, such as increased growth velocity, advanced bone age, symptoms, or signs suggestive of central nervous system dysfunction or of peripheral precocious puberty.

Bone age measured using a reference atlas such as Greulich and Pyle evaluates the impact of sex steroids on epiphyseal maturation, and is usually advanced in progressive precocious puberty. Caution should be taken in over interpreting bone age, since there is a physiological scatter of approximately plus or minus 1 year of bone age versus chronological age in white people, and a systematic advance of bone age in Africans when using references obtained in white people,. Bone age can also be used to predict adult height, although with a low precision (95% confidence interval of about ± 6 cm) and a tendency to overestimate adult height in precocious puberty.

Hormonal measurements that can be useful for the evaluation of premature sexual maturation are summarized in Table 7.2.10.4.

In girls, pelvic ultrasonography can be used to detect ovarian cysts or tumours. Uterine changes due to oestrogen exposure can be used as an index of progressive puberty. A uterine volume greater than 2.0 ml and an uterine length of more than 34 mm have 89 and 80% sensitivity, and 89 and 58% specificity, respectively, for precocious puberty in one series (18). Testicular ultrasound scans should be performed if testicular volume is asymmetric or in peripheral precocious puberty, in order to detect Leydig cell tumours, which are generally not palpable.

Brain MRI is important to detect hypothalamic lesions in progressive central precocious puberty (19). The prevalence of such lesions is higher in boys (40–90% of cases) than in girls (8–33%) and is much lower when puberty starts after the age of 6 years in girls (about 2% in (20)). It has been suggested that an algorithm based on age and oestradiol levels may obviate the need for MRI in one third of girls, but this has not been extensively validated (19, 20).

Clinical evaluation, hormonal measurements and imaging usually identify one of the following situations (Fig. 7.2.10.1):

Table 7.2.10.5 summarizes features reflecting the intensity and duration of the gonadotropic axis activation that are useful in distinguishing between progressive central precocious puberty and non-progressive forms of precocious puberty. Although these criteria are not fully evidence-based, and reflect personal experience, as well as data obtained in cross-sectional and small-sized longitudinal studies, they can provide useful orientation. When discrepant results are obtained, it is recommended to wait a few months and reassess, to avoid unnecessary treatment (21).

GnRH agonists are generally indicated in progressive central precocious puberty. GnRH agonists continuously stimulate the pituitary gonadotrophs, leading to desensitization and decrease in luteinizing hormone release and, to a lesser extent, FSH release (22). Several GnRH agonists are available in various depot forms and their approval for use in precocious puberty varies with countries. Despite nearly 30 years of use of GnRH agonists in precocious puberty, there are still ongoing questions on their optimal use and an international consensus statement in 2007, has summarized the available information and the areas of uncertainty (2).

GnRH agonist treatments should be followed by experienced clinicians and result in the regression or stabilization of pubertal symptoms, decrease of growth velocity and bone age advancement (2). GnRHa-injection dates should be recorded and adherence with the dosing interval monitored. A suppressed luteinizing hormone response to the stimulation by GnRH, GnRH agonist, or after an injection of the depot preparation (which contains a fraction of free GnRH agonist) is indicative of biochemical efficacy of the treatment, but is not recommended routinely. Progression of breast or testicular development usually indicates poor compliance, treatment failure, or incorrect diagnosis, and requires further evaluation.

There are no randomized controlled trials assessing long-term outcomes of the treatment of central precocious puberty with GnRH agonists, but in most studies height outcomes have been evaluated. Among approximately 400 girls treated until a mean age of 11 years from several published series, the mean adult height was about 160 cm and mean gains over predicted height in the various series of patients varied from 3 to 10 cm (11). Individual height gains were very variable, but were calculated using predicted height, which is itself unreliable. Factors affecting height outcome include initial patient characteristics (lower height if bone age is markedly advanced and shorter predicted height at initiation of treatment) and, in some series, duration of treatment (higher height gains in patients starting treatment at a younger age and with longer durations of treatment).

Other outcomes to consider include bone mineral density, risk of obesity, and psychosocial outcomes. Bone mineral density may decrease during GnRH agonist therapy. However, subsequent bone mass accrual is preserved, and peak bone mass does not seem to be negatively affected by treatment (2). There have been concerns that GnRH agonist use may affect body mass index (BMI). However, childhood obesity is associated with earlier pubertal development in girls, and early sexual maturation is associated with increased prevalence of overweight and obesity. Altogether, the available data indicate that long-term GnRH agonist treatment does not seem to cause or aggravate obesity, as judged from BMI (2). However, the risk of obesity is a concern in girls with premature sexual maturation and BMI should be closely monitored. As discussed above, psychosocial evaluation data are scarce in patients with premature sexual maturation and there is little evidence to show whether treatment with GnRH agonists is associated with improved psychological outcome (2).

Although tolerance to GnRH agonist treatment is generally considered good, it may be associated with headaches and menopausal symptoms, such as hot flushes. Local complications (3–13%), such as sterile abscesses may result in a loss of efficacy and anaphylaxis has been described exceptionally (23).

The optimal time to stop treatment has not been established and factors that could influence the decision to stop GnRH agonists include aiming at maximizing height, synchronizing puberty with peers, ameliorating psychological distress, or facilitating care of the developmentally delayed child(Table 7.2.10.6). However, data only permit analysis of factors that affect adult height. Several variables can be used to decide on when to stop treatment including chronological age, duration of therapy, bone age, height, target height, and growth velocity. However, these variables are closely interrelated and cannot be considered independently. In addition, retrospective analyses suggest that continuing treatment beyond the age of 11 years is associated with no further gains (24). Therefore, it is reasonable to consider these parameters, and informed parent and patient preferences, with the goal of menarche occurring near the population norms (2). Pubertal manifestations generally reappear within months of GnRH agonist treatment being stopped, with a mean time to menarche of 16 months (25). Long-term fertility has not been fully evaluated, but preliminary observations are reassuring (25).

The addition of growth hormone (26) or oxandrolone (27) when growth velocity decreases or if height prognosis appears to be unsatisfactory has been proposed, but data are limited on the efficacy and safety of these drugs in children with precocious puberty.

When precocious puberty is caused by a hypothalamic lesion (e.g. mass or malformation), management of the causal lesion has generally no effect on the course of pubertal development. Hypothalamic hamartomas should not be treated by surgery for the management of precocious puberty. Precocious puberty associated with the presence of a hypothalamic lesion may progress to gonadotropin deficiency.

Surgery is indicated for gonadal tumours and postoperative chemo- or radiotherapy should be discussed as part of a multidisciplinary team including surgeons and oncologists.

Large ovarian cysts (greater than 20 ml or 3.4 cm in diameter and, typically, more than 75 ml or 5.2 cm) should be managed very carefully given the risk of adnexal torsion (28). In such cases, puncture (possibly ultrasound-guided) should be considered and allows molecular analysis of the cystic fluid for an activating GNAS mutation.

Removal of exogenous exposure to sex steroids is obvious, but the search for occupational exposure is often very difficult and requires careful investigation.

Benign variants of premature sexual maturation should be followed clinically with reassurance to the parents. There are limited data on long-term outcomes of individuals with these conditions and it has been suggested that premature pubarche is a risk factor for hyperandrogenism in adulthood.

There is no available treatment for peripheral causes of precocious puberty directed at the aetiology and the rarity of the diseases renders evaluation of therapeutic strategies very difficult. In McCune–Albright syndrome and recurrent ovarian cysts, aromatase inhibitors (29) and selective oestrogen receptor modulators (SERMs) (30) have been used to inhibit the production or action of oestrogens respectively. These approaches are partly effective, but no definitive strategy has emerged. In familial male precocious puberty due to luteinizing hormone receptor activating mutations, ketoconazole, an inhibitor of androgen biosynthesis, has been shown to be effective in the long term (31), and the combination of antiandrogens and aromatase inhibitors has been proposed. However, caution must be used with the use of ketoconazole given the risk of liver toxicity. Nonclassical and classical forms of congenital adrenal hyperplasia should be managed with glucocorticoids.

The main concern when examining a patient with premature sexual development should be the existence of a malignant or potentially-threatening lesion, either intracranial, in the gonads, the adrenals, or elsewhere. However, these lesions are exceedingly rare and, on a daily basis, the main difficulty is with the differentiation of progressive and non-progressive forms of precocious puberty, and with the decision to treat, particularly for girls with an onset of puberty between the ages of 6 and 8 years.