| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Pediatrics, Division of Endocrinology, Sophia Children’s Hospital, Erasmus University, 3000 CB Rotterdam, The Netherlands
 |
Abstract |
|---|
 Top Abstract I. IntroductionII. Normal vs. Extremes...III. Endocrinology of Bone...IV. Sex Steroid Action...V. Height PredictionVI. Treatment of CTS:...VII. Treatment of...VIII. Estrogen Treatment in...IX. Alternative Treatment...X. Conclusions and...References |
|---|
|
I. Introduction |
|---|
|
TopAbstractI. IntroductionII. Normal vs. Extremes...III. Endocrinology of Bone...IV. Sex Steroid Action...V. Height PredictionVI. Treatment of CTS:...VII. Treatment of...VIII. Estrogen Treatment in...IX. Alternative Treatment...X. Conclusions and...References |
|---|
Growth is a result of complex processes. Genetic constitution, nutrition, endocrine function, and psychosocial well being are all involved in the process of growth (1, 2). The genetic component of height has been estimated to be 0.5–0.9, i.e., 50–90% of the height variation is accounted for by genetic factors. Assessment of the parental height as an indicator of the genetic component of growth and development of the child is therefore of critical importance (3). In addition, socioeconomic factors such as social class, family size, birth rank, housing, and crowding are associated with growth. Improved socioeconomic conditions and better health have led to the manifestation of a positive secular trend in growth and development over the last centuries. In 1865 the mean adult height among Dutch army recruits was 165 cm. One century later, in 1965, the mean adult height in boys was 178 cm. Fifteen years later, in 1980, the mean adult height had increased by another 4 cm to 182 cm. In the middle of the 19th century, age of menarche of European girls was about 16–17 yr. Nowadays, the mean age of menarche is 13 yr or even less.
Phenomena responsible for both positive and negative secular trends have affected height throughout our history. Studies of fossil remains of our hominid ancestors demonstrate that the stature of individuals living during the last hundred-thousands of years reached the range of heights seen today: the mean stature of early anatomically modern Homo sapiens in Europe was 184 cm in males and 167 cm in females (4, 5). Thus, stature is based on many factors, including heredity and environment.
In recent years, information concerning auxology and (neuro)endocrinology of tall stature has expanded. In addition, long-term results of height-reducing treatment modalities have become available. In this review we will give an update of the (neuro)endocrinology, the auxology, the differential diagnosis, and the therapeutic modalities available in the management of constitutionally tall stature (CTS).
|
II. Normal vs. Extremes of Growth |
|---|
|
TopAbstractI. IntroductionII. Normal vs. Extremes...III. Endocrinology of Bone...IV. Sex Steroid Action...V. Height PredictionVI. Treatment of CTS:...VII. Treatment of...VIII. Estrogen Treatment in...IX. Alternative Treatment...X. Conclusions and...References |
|---|
CTS is defined as a condition in which the height of an individual is 2
SDs above the corresponding mean height for a given age,
sex, and population group. As shown in Table 1
, height p97 values vary
substantially among various populations. The Scandinavians and the
Dutch are among the tallest people in the world.
|
The diagnosis is generally made from the family history of growth and
from physical examination. No apparent abnormalities are present at
physical examination, which permits distinction from excessive growth
syndromes such as Marfan syndrome and Klinefelter syndrome (see Table 2).
|
At the time of puberty GH secretion increases. Sex-specific changes regarding the timing of the pubertal increase of GH secretion during puberty have been found by analyzing 24-h GH profiles in healthy boys and girls (25). This increase, occurring about 1 yr earlier in girls than in boys, is correlated to estradiol levels in both boys and girls (26, 27). Similarly, serum levels of IGF-I rise at puberty (23). In boys with delayed puberty, testosterone (T) treatment caused increased GH pulse amplitude, thereby increasing the mean serum GH concentration. T exerts its effect on GH via an estrogen-dependent mechanism by increasing hypothalamic GHRH release (28). Paradoxical GH responses to glucose loading and to administration of TRH or GHRH, similar to those seen in acromegaly, have been observed in some CTS children (29, 30, 31). However, these observations have not been substantiated in studies properly controlled for age and stage of puberty. Tauber et al. (32) showed a clear heterogeneity of GH secretion in tall children, with some of them even having low GH secretion (32). Therefore, abnormal responses may be related more to the stage of puberty of the child than to abnormalities of GH secretion (33).
In conclusion, constitutionally tall children are healthy children without hard evidence of pathology of the GH-IGF-I axis.
|
III. Endocrinology of Bone Growth and Maturation |
|---|
|
TopAbstractI. IntroductionII. Normal vs. Extremes...III. Endocrinology of Bone...IV. Sex Steroid Action...V. Height PredictionVI. Treatment of CTS:...VII. Treatment of...VIII. Estrogen Treatment in...IX. Alternative Treatment...X. Conclusions and...References |
|---|
Longitudinal bone growth is the result of expansion of the growth plate cartilage. As puberty proceeds, a progressive decrease in cartilage expansion occurs. Because the vascular invasion and resorption of calcified cartilage by chondroclasts exceeds cartilage expansion, there is a progressive thinning of the growth plate. Ultimately, the growth plate becomes perforated and longitudinal bone growth ceases (34, 35).
The regulation of longitudinal bone growth is very complex, and several factors, including nutritional, endocrine, paracrine, and autocrine, are necessary (36). For normal bone growth and maturation the gonadal steroids are essential in conjunction with several hormones and growth factors:
1. GH. GH has been shown to stimulate long bone growth in a dose-dependent fashion. At the cellular level GH interacts with the GH receptor. The expression of GH receptors is developmentally regulated in epiphyseal chondrocytes (37). There is ample evidence that GH interacts directly with cells of the growth plate and not only through IGF-I (vide infra). Green and associates (38) have proposed the dual effector theory based on in vitro observations (see also Ref. 39). This theory states that GH is a prerequisite of cartilage maturation. Priming of resting chondrocytes in the growthplate by GH is required for IGF-I to promote ’clonal expansion’ of growth plate chondrocytes and to stimulate skeletal growth. More recently it has been suggested that this theory may not apply to the in vivo situation. The presence of GH receptors is not limited to resting chondrocytes. It was shown that both IGF-I and GH exerted their effect at each stage of differentiation rather than acting specifically upon particular subpopulations of cells at certain phases of chondrocyte differentiation (40). The observation that GH-overproducing transgenic mice have a significantly larger size than their controls, while IGF-I transgenic mice have a normal size, would support the theory of a differential effect of GH and IGF.
However, transgenic models are inadequate for studying the independent actions of IGF-I and GH as in IGF-overproducing transgenic mice GH production is not completely suppressed (41).
2. Thyroid hormone. Thyroid hormones are crucial for bone growth because of their direct effects on bone. In addition, there are indirect effects by stimulating pituitary GH release (42), thereby increasing serum IGF-I levels. Lastly, effects of thyroid hormone on IGF-I generation by chondrocytes have been demonstrated (43). The obvious experiment of nature is represented by the syndrome of resistance to thyroid hormone hallmarked by short stature and marked delayed bone maturation (44).
3. Vitamin D. In addition to the crucial role of vitamin D in bone mineralization, several observations point to a possible role in chondrocyte proliferation and bone growth. In rats, proliferation of growth plate chondrocytes is decreased by high doses and increased by low doses of 1,25-dihydroxyvitamin D3 [ 1,25-(OH)2D3] (45). Moreover, short stature and delayed bone maturation may be present in the syndrome of vitamin D resistance (46).
4. Growth factors. As IGF-I and -II are among the most prevalent growth factors secreted by skeletal cells and have important actions on bone formation, it is reasonable to predict that they play a significant role in bone growth and maturation. In the circulation IGF-I and -II form a complex with IGF-binding proteins, and there is little, if any, free IGF-I or -II available in bone. Thus, locally secreted IGF-I and -II acting in a paracrine or autocrine manner might play an even more important role in the regulation of growth plate cell function. IGFs enhance the differentiated function of the osteoblast and increase collagen synthesis. Moreover, IGFs decrease collagen degradation and IGF-I (but not IGF-II) increases osteoclast recruitment (41).
The synthesis of IGF-I and -II takes place in cells of osteoblast lineage and is controlled by systemic hormones (stimulatory: PTH, GH, estrogens; inhibitory: cortisol, vitamin D) and local factors (stimulatory: bone morphogenetic proteins, PGE2; inhibitory: fibroblast growth factor, transforming growth factor, platelet derived growth factor).
In addition, skeletal cells synthesize a variety of IGF-binding proteins (IGFBP). They act to regulate and modulate the local actions of IGFs. Most IGFBPs have been shown to have inhibitory effects on either IGF-I or -II. However, IGFBP-5 has been shown to potentiate IGF-II action on osteoblast- derived cell lines (47).
|
IV. Sex Steroid Action on Bone Growth and Maturation |
|---|
|
TopAbstractI. IntroductionII. Normal vs. Extremes...III. Endocrinology of Bone...IV. Sex Steroid Action...V. Height PredictionVI. Treatment of CTS:...VII. Treatment of...VIII. Estrogen Treatment in...IX. Alternative Treatment...X. Conclusions and...References |
|---|
Girls with Turner syndrome show a growth spurt during estrogen replacement therapy (53, 54, 55). Similarly, girls with central or pseudo-precocious puberty with clearly elevated estrogen levels show increased height velocity and premature epiphyseal closure (56). Whereas in patients with androgen insensitivity syndrome the growth spurt resembled that of women both in magnitude and timing, mean final adult height was taller than in normal women but shorter than in normal men (22, 48, 57). This would suggest that Y chromosome-related factors contribute to stature. Patients with familial male precocious puberty treated with antiandrogens alone did not revert to a normal prepubertal growth rate but only when an aromatase inhibitor was added (58).
Studies of acute infusion of gonadal steroids in peripubertal children have illustrated the complexity of the relationship between gonadal steroids and the GH/IGF axis (49, 59, 60). These studies suggest sex differences in the control of GH secretion in the response to E2 during puberty. Moreover, acute and chronic E2 exposure may have different effects: acute infusions decrease whereas prolonged exposure may increase GH bioactivity (60).
There is also evidence for a dose dependency as high doses of estrogens result in decreased growth velocity in both males and females (58, 61, 62). High-dose estrogen treatment decreased IGF-I levels in acromegalic patients as well as in tall girls (63, 64). Recently, a man was described with estrogen resistance. He had a normal prepubertal growth and normal timing of onset of secondary sex characteristics. Despite full masculinization he continued to grow. At 18 yr his height was 204 cm, and the growth velocity was 1 cm/yr. The bone maturation and mineralization were both markedly retarded (65). Moreover, a phenotypic female with aromatase deficiency was described showing markedly delayed bone maturation (66). It is well known that in boys with pseudo-precocious puberty as a result of congenital adrenal hyperplasia, height velocity is increased and epiphyseal maturation is advanced, resulting in stunted adult height. There is ample evidence that androgen-stimulated growth is largely based on influencing GH release and augmentation of IGF-I. However, it is not excluded that these effects are estrogen mediated. On the other hand, nonaromatizable androgens have growth-promoting effects not mediated via GH-IGF-I axis (67). Keenan et al. (68) reported that nonaromatizable DHT induced and maintained accelerated growth rate in short boys with delayed puberty in spite of a 50% decline in integrated GH concentration and no change of IGF-I level, suggesting that strictly androgen-mediated skeletal growth might be exerted locally in growing cartilage.
In vitro studies using rat- and human-derived cells have shown that there might be a sex-specific and age-dependent responsiveness of cartilage and bone cells to sex steroids. Cells and tissues derived from males respond primarily to T, whereas cells and tissues derived from females respond primarily to estrogen. The best response was obtained in tissue from children in early puberty (69, 70, 71). The mechanism of action of the gonadal steroids on growth plate cartilage is poorly understood. The effect of estrogens on proliferation of human chondrocytes in vitro was shown to be biphasic: at low concentration a stimulatory effect was observed, while at supraphysiological doses inhibition was observed.
High doses of estrogens stimulate the maturation of cartilage without increasing the growth rate: cell division by cartilage cells is inhibited in the proliferative zone of the growth plate, and the age-related decrease in size of the hypertrophic chondrocytes is accentuated by estrogens (72, 73, 74, 75, 76). The latter effects were not overcome by the addition of GH or IGF, suggesting that estrogens may act directly on chondrocytes or may influence the release of factors that inhibit cell proliferation locally (77). Using fetal rat osteoblasts in culture, McCarthy et al. (78) established that estrogens do not alter IGF-I promotor activity but inhibit the biological effects of all hormones that act through cAMP to regulate skeletal IGF-I expression and activity.
Collectively, bone growth and bone maturation are the result of a complex interplay of various hormones in which GH and the gonadal steroids have a pivotal role. Moreover, there might be a sex-specific and age-dependent responsiveness of cartilage and bone cells to sex steroids. At the level of the growth plate, estrogen receptor-mediated processes appear essential in expressing the effects of sex steroids.
|
V. Height Prediction |
|---|
|
TopAbstractI. IntroductionII. Normal vs. Extremes...III. Endocrinology of Bone...IV. Sex Steroid Action...V. Height PredictionVI. Treatment of CTS:...VII. Treatment of...VIII. Estrogen Treatment in...IX. Alternative Treatment...X. Conclusions and...References |
|---|
A. Skeletal maturity or BA
A measure of skeletal maturity is generally obtained by assessing
the appearance and shape of the bones of the hand and wrist on an
x-ray. These appearances change with age, and their rate of change is a
direct measurement of the rate of maturation. Various methods of
evaluation of BA and maturity have been developed over the years. The
methods most commonly used are the Greulich and Pyle Atlas (83) and the
Tanner-Whitehouse (TW2) method (79, 80). In addition, other methods
such as the FELS-method are also available (85).
The method of Greulich and Pyle for BA estimation is presented as an atlas of examples of radiographs of the left hand and wrist of healthy children at various ages (83). The children who form the standard group were drawn from the Brush Foundation Study, which selected children from the better socioeconomic strata in Cleveland, Ohio, from 1930 onward. All the children were white, had been born in the United States, and were of North European ancestry. Each of the standards of the atlas was selected from a large number of children of the same sex and age. All films were arranged in order of increasing maturity, and the film chosen as the standard is the one most representative of the central tendency or anatomical mode. The BA is determined by comparison with the standards.
The Tanner-Whitehouse technique describes maturity indicators for each bone of the hand and wrist (79, 80). Each bone progresses through a series of specific stages with assigned weighted scores. These scores are added to form a maturity score, which in turn can be converted to a corresponding BA. The source data for this method were obtained from large groups of British children of an average socioeconomic level in the 1950s.
The major problem in both techniques is that subjective processes and discontinuous scales are used that result in considerable inter- and intrarater variability in BA (84, 86, 87, 88). In a direct comparison, the BA as determined by the Greulich and Pyle method is generally about 1 ‘year’ less than that as assessed by the Tanner-Whitehouse technique (86, 87, 89). One should realize, however, that BA determinations are estimates of maturity and that, in fact, there is no objective quantification available. Recently, various authors have discussed the main problems of skeletal maturation assessment and the sources of possible bias (90, 91, 92). Therefore, it is mandatory that one is acquainted with the specific qualities of the BA determination method used.
B. Computed assisted skeletal age-scoring systems
As stated above, estimates of BA do not advance smoothly as the
child matures but in a saltatory fashion. In the TW2 system, a
difference of one stage in the rating of a particular bone may result
in an increase of 0.3 ‘years’ in BA. In actuality, skeletal maturity
will advance gradually. To diminish the errors in the interpretation of
maturity and to improve BA ratings, the TW2 system has been transformed
recently by the original author into a computerized image analysis
system using a continuous scale Computer-Assisted Skeletal Age System:
CASAS (93, 94). So far, a limited number of studies have been performed
on the reliability and validity of CASAS in healthy children and in
children with tall stature (95, 96, 97). Results indicate that reliability
of the CASAS ratings is extremely high, both within and across
operators. Moreover, in longitudinal series, BA does advance far more
smoothly compared with manual scores. With regard to children with tall
stature, CASAS was found quite applicable. The CASAS method, however,
is not without drawbacks and is still, to some extent, user dependent.
Further developments are needed to improve these aspects.
C. Accuracy of height prediction
Prediction methods that have survived the tests of clinical
usefulness are those that incorporate a multitude of variables relating
to adult height and maturity and that are sensitive in their assessment
of childhood maturity (82). Most prediction methods are based on growth
data of unselected normally growing children. Therefore, when applied
to children with tall stature, critical appraisal of their qualities is
required. Knowledge concerning the specific advantages and
disadvantages of the various methods is of utmost importance
since it may influence possible therapeutic intervention.
Thus far, only a limited number of studies have been performed testing the reliability of height prediction methods in large groups of untreated children with tall stature. The accuracy of height prediction may be expressed mathematically as the difference between predicted adult height and actual adult height. In this way, positive values indicate overestimation, and negative values reflect underestimation of the final adult height. Absolute errors demonstrate the method’s overall predictive error and is not influenced by over- or underestimation.
Table 3 summarizes the accuracy of
various prediction methods in boys with tall stature
(98, 99, 100, 101). The systematic tendency of the prediction methods to over-
or underestimate final adult height are not consistent. Variation in
initial clinical data (CA and BA) and time definition of reaching adult
height may account for this inconsistency. In general, however, the
method of Bayley-Pinneau tends to overestimate final height, whereas
the method of Tanner-Whitehouse slightly under- or overestimates final
height. Joss et al. (102) described a systematic
overprediction of the Bayley-Pinneau technique in a study of 32 tall
boys. In addition, they reported a systematic overprediction using the
Tanner-Whitehouse method, which was even more pronounced at an older
BA. Some investigators have used repeated predictions in the same
subject for accuracy assessment. This may have induced bias in reported
means and/or standard deviations of the errors of prediction. When the
group of patients was divided into age-specific subgroups it appeared
that with increasing age both methods became more accurate in
predicting adult height (100, 101). In our study the Index of Potential
Height (IPH) based on the BA of Greulich and Pyle was found to be the
most reliable method as it showed the lowest mean error and mean
absolute error, -0.1 (2.9) cm and 2.3 (1.8) cm, respectively
(101). The IPH is based on the assumption that the height
SD scores for BA [rather than for chronological age (CA)]
remains constant up to final height.
|
(64, 101, 103, 104, 105, 106, 107, 108, 109). Again, inconsistency is present in reported errors of
prediction, probably due to differences in initial clinical data and
timing of adult height assessment. Nevertheless, the mean errors are
found to be rather small, indicating that height prediction in tall
girls is quite accurate regardless of which method is used. As in boys,
predicting adult height became more accurate with increasing age (101).
The mean absolute errors were also small, a confirmation of
their reliability. It seems that there is no ‘best’ prediction method
in tall girls; the choice of method for use in daily practice may
therefore depend on the preference and experience of the clinician.
|
To improve height prediction in children with constitutionally tall
stature, we developed regression equations based on growth data derived
from a sample of untreated tall children (55 boys/88 girls) (111).
Since the quality of a regression model is reflected in its residual
standard deviation (RSD), the smaller the RSD the better the model
predicts the dependent variable (final height) by the combination of
predictor variables. In our newly developed prediction equations the
RSD was 2.5 cm for both boys and girls (see Table 5). This implies that about 95% of the
predictions lie within approximately 5 cm of the real value (± 2 RSD).
From a clinical point of view, this inaccuracy is quite acceptable. In
comparison, Tanner et al. (79) reported RSD values up to 4.1
cm in boys and 3.6 cm in girls for the same age ranges. The prediction
equations were tested on a separate sample of 32 tall children (16
boys/16 girls) and compared with other prediction techniques including
the TW Mark 2 and Bayley-Pinneau method. The absolute errors of our
prediction equations were smaller (though not reaching statistical
significance) than the TW Mark 2 method and the Bayley-Pinneau
technique, indicating less variability. These results give support to
the idea that height prognosis in children with excessive growth is
more accurate when based on growth data derived from tall children.
|
|
VI. Treatment of CTS: General Concepts |
|---|
|
TopAbstractI. IntroductionII. Normal vs. Extremes...III. Endocrinology of Bone...IV. Sex Steroid Action...V. Height PredictionVI. Treatment of CTS:...VII. Treatment of...VIII. Estrogen Treatment in...IX. Alternative Treatment...X. Conclusions and...References |
|---|
Sex steroids have been used in the treatment of tall boys and girls since the late 1950s. The basis for the use of sex hormones to limit adult height came from observations in children with (pseudo-) precocious puberty. These children show early closure of the epiphyses due to premature production of gonadal steroids, which limits their eventual adult height (118, 119).
Studies mainly in children with gonadal dysgenesis have suggested a biphasic dose-dependent effect of estrogens on growth rate, low dose having a stimulatory and high dose an inhibitory effect (36, 120, 121). Furthermore, it has been demonstrated that the administration of high doses of gonadal steroids, specifically estrogens, accelerate bone maturation (98, 99, 103, 104, 112, 122, 123, 124, 125). Since the first study in 1956, many reports have appeared describing the height-reducing effect of administration of high doses of sex hormones in girls (64, 103, 104, 105, 106, 107, 108, 109, 112, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131) and in boys (98, 99, 132).
|
VII. Treatment of Constitutionally Tall Boys |
|---|
|
TopAbstractI. IntroductionII. Normal vs. Extremes...III. Endocrinology of Bone...IV. Sex Steroid Action...V. Height PredictionVI. Treatment of CTS:...VII. Treatment of...VIII. Estrogen Treatment in...IX. Alternative Treatment...X. Conclusions and...References |
|---|
It is generally agreed that the steroid hormone effects on bone maturation are due to an indirect action mediated by the GH/IGF-I axis combined with a direct effect at tissue level after metabolic conversion into estrogens (48).
The choice of the androgen preparation is unambiguous. ‘Natural’ compounds are preferred over 17-alkylated compounds as it has been observed that the latter may cause cholestasis and hepatic tumors and may negatively influence lipoprotein levels (115). The doses of the long-acting T esters (such as T propionate, enanthate, and decanuate) used in most studies are about 500 mg/m2/month, which correspond to roughly 4 times the normal T production rate of adult men or to about 8–10 times that of early adolescence. In clinical practice, two weekly im injections of 500 mg or 250 mg once a week are used. Whether such high doses are really necessary to obtain a maximum effect on bone maturation is not known. However, treatment with T at a mean dosage of 265 mg/m2/month resulted in a lower BA velocity and thereby less reduction of adult height in a small group of tall patients with hypogonadotropic hypogonadism (98). Theoretically, an alternative treatment modality is testosterone undecanoate. Testosterone undecanoate in oleic acid is administered orally and is absorbed preferentially through the lymphatics into the bloodstream bypassing first-phase degradation in the liver (133). However, dose frequency is 2–3 times a day, and circulating blood levels tend to vary substantially (134). Dose finding studies for CTS treatment have not been performed.
B. Height reduction
The ‘uncorrected’ effect of height reductive therapy,
i.e., height prediction minus achieved adult height, varies
with the prediction method applied. Since every single prediction
method has its own prediction error, the mean effect may be
‘corrected’ by subtraction of the corresponding mean prediction
errors (98, 99, 132) (see Table 6). In
our own studies the Bayley-Pinneau prediction showed the
greatest mean ‘corrected’ effect of 2.0 cm, while the IPH, being the
most accurate method, calculated a mean ‘corrected’ effect of only
0.6 cm (101). An important finding, however, was that at the time of
referral the control groups (tall boys and girls) were significantly
different from children who had received sex hormone therapy. The
proper ’net’ treatment effect on final height was calculated by
multiple linear regression analysis adjusting for differences in age,
BA, and height prediction between treated and untreated children. The
mean adjusted effect for the various prediction methods varied from
-1.7 to +0.7 cm in boys. There was, however, a marked variation in the
individual height-reducing effect, ranging from -2.6 to +15.8 cm.
Figure 1 shows the adjusted effect of
therapy according to the IPH-Greulich-Pyle prediction and its
95% confidence interval.
|
|
It has been clearly shown that height reduction was dependent on the BA at start of therapy: height reduction was more pronounced when treatment was started at a younger BA (100, 101, 132). However, an important issue that caused a significant reduction in the height-limiting effect was the observation of a marked additional posttreatment growth after cessation of therapy. This posttreatment growth might partly be explained by late-pubertal completion of spinal growth. On the other hand, the additional growth could result from the fact that treatment had been stopped before complete closure of the epiphyses. A significant negative relationship between posttreatment growth and BA at the time of stopping therapy (r = -0.53; P < 0.001) was observed (101). The latter contrasts with the opinion of Brämswig and co-workers (132), who advocated short-term therapy and reported significant height reduction (uncorrected: 7.6 cm) with a mean BA (SD) of 15.3(0.8) yr at the time of stopping therapy. However, these results seem too optimistic, since final height assessment was performed at a relatively young mean (bone) age. Moreover, others failed to show any height reducing effect using the same therapeutic strategy, but with assessment of final height at a definite later point in time (135).
As illustrated in Fig. 1
, when therapy was started at a BA of 14 yr or
older, adult final height significantly exceeded height prognosis at
the time of start of treatment. This suggests that treatment had
resulted in induction rather than reduction of growth.
C. Effects on gonadal function
High doses of T induce suppression of the
hypothalamo-pituitary-gonadal axis (113, 136). Contraceptive studies in
adult men have shown that androgen-induced suppression of gonadotropins
and of spermatogenesis is reversible (137, 138). However, extrapolation
of these data to the management of tall stature in pubertal boys must
be viewed with caution since factors that regulate spermatogenesis in
normally functioning adult testes may not be the same as during puberty
(139). Androgen therapy in tall boys is usually initiated at the first
signs of puberty, and it is in this peripubertal period that important
maturational changes take place in the testis (139, 140, 141, 142, 143, 144). Influenced by
complex hormonal actions, these maturational processes eventually lead
to initiation of spermatogenesis. Onset of spermatogenesis (spermarche)
as detected by urine analysis (spermuria) appears to be an early
pubertal event: the median age of spermarche has been estimated to be
13–14 yr (145, 146, 147, 148). In addition, it is noteworthy that administration
of T esters at high doses may cause morphological and cytological
changes, as shown in rat and human adult testes (149, 150, 151).
1. Plasma hormone levels. High T levels are obtained during treatment with supraphysiological doses of androgens suppressing the hypothalamo-pituitary-gonadal axis (152).
Zachmann, Prader, and co-workers (98, 113) reported a slow recovery of pituitary gonadotropins during LHRH-stimulation tests after discontinuation of T therapy. Brämswig et al. (136) demonstrated normalization of gonadotropin levels in 100 tall boys after discontinuation of treatment with follow-up periods up to 48 months, although transient hypergonadotropic LH- and FSH- secretory patterns were observed. In a recently published study by the same group (153), hormonal levels and testicular function were evaluated after a follow-up period of approximately 10 yr and compared with normal volunteers. Mean values of LH, FSH, PRL, T, estradiol, and sex hormone-binding globulin were in the normal range in both groups. T was lower and FSH was higher in treated tall men compared with volunteers, but the only statistically significant difference was for T. We observed different levels of gonadotropins in previously treated tall men compared with controls (tall and ‘normal’ men) (154). Androgen-treated tall men had significantly higher FSH levels compared with controls. Levels of plasma hormones were not significantly correlated with parameters of sperm quality; however, we observed significant negative correlations between plasma FSH levels and sperm concentration as well as the age at start of therapy in the androgen-treated men. We speculate that the higher levels of FSH may reflect intratesticular changes due to androgen treatment received during a period of testicular maturation especially during the earlier pubertal stages (155). These increased FSH levels may compensate for partially disturbed germinal function to maintain normal sperm quality (156). In a subgroup of previously treated and untreated men, we also measured inhibin B, which probably is a more direct marker of spermatogenesis than FSH (157). We found similar levels, well within the normal range (F. H. De Jong and W. J. De Waal, unpublished results). On the other hand, the difference in gonadotropin levels may also reflect a change in responsiveness at the hypothalamo-pituitary level (136).
2. Testicular volume. Treatment with high doses of androgens induces reduction in testicular volume in adult men (158) as well as in tall adolescent boys (98, 99, 113). This implies major intratesticular changes during therapy such as a decrease in seminiferous tubule size (149, 150). These processes are likely to be reversible since testicular volume normalizes after discontinuation of therapy as shown in several studies (98, 99, 113, 158). This is in contrast to the observations of Willig et al. (159, 160), who reported significantly smaller testicular sizes in previously treated men. In contrast, in our studies at a mean follow-up period of 8 yr after cessation of treatment, there was no difference in mean testicular volume between treated and untreated tall men (154).
3. Sperm quality. When sperm quality is evaluated, one must be aware of the normal distribution in the population as well as of confounding factors interfering with parameters of sperm quality. It is well established that varicocele (161, 162, 163, 164), smoking (165), sexually transmitted disease (166), and cryptorchidism (167) are likely to affect sperm quality and/or plasma hormone levels. Semen analysis in our study of previously androgen-treated men showed that sperm quality was comparable with a control group of untreated tall men, even after correction for the above mentioned possible interfering conditions, after a mean follow-up period of 8 yr. These findings are in agreement with the experiences reported by Zachmann and Prader and co-workers (98, 113). In contrast, Willig and co-workers (159, 160) found significantly reduced sperm concentrations in previously treated tall men compared with controls. Their control group, however, showed a relative high mean value of sperm concentration of 120.2 x 10'6/ml, almost twice as high as values found in the normal population at present (168, 169). Their treated group showed a mean sperm concentration of 63.4 x 10'6/ml, which is comparable with values found in our study (154). It is possible that differences in patient selection, semen analysis methodology, and treatment regimens may account for the observed differences. In addition, the extent to which interfering conditions are present may cause important bias as well. In a recent report Lemcke and co-workers (153) showed that 10 yr after T treatment, none of the tall men had azoospermia, and the mean ejaculate parameters were in the normal range or only slightly subnormal. Overall, seminal parameters of T-treated tall men were slightly, but not significantly, lower compared with normal statured volunteers. Interestingly, they found a significantly higher prevalence of varicocele and maldescended testes in the tall men compared with their control group of normal volunteers (153). They surmised that varicocele and maldescended testes, rather than T treatment, caused the somewhat lower semen quality in the tall men. In our studies in treated and untreated tall men, we observed an overall prevalence of varicocele of 42% (12% subclinical and 30% clinical) (154). This would suggest that varicocele occurs more often than reported in the normal population (12.4–25.8%) (170, 171, 172). A relationship to androgen treatment is unlikely since no difference in the prevalence of varicocele was observed between androgen-treated men and controls. One could speculate on the impact of stature on the pathogenesis of varicocele (153, 154).
4. Pregnancy/paternity. Thus far, only casuistic and exclusive female data have been available on successful pregnancies after height-reductive therapy. At the time of our follow-up studies five of the 43 androgen-treated men and six of the 30 untreated tall men had fathered one or more children (154). All 11 men reported that pregnancy had occurred in their partners after less than 1 yr of unprotected coitus. Two other pregnancies, fathered by a previously treated man and a control, respectively, ended in spontaneous abortion. These very limited numbers do not allow any further conclusions.
D. Other clinical effects
Many patients experience side effects during therapy (98, 99, 113, 132, 173, 174). Most of these, however, are mild and transient (see
Table 7). In some patients, slight to
moderate edema, notably in the pretibial or malleolar area, was
associated with marked weight gain during the first 6 months of
treatment. This indicates that the early gain in weight is not only due
to protein anabolism but also to water retention (98). Acne was by far
the most reported side effect (98, 99, 175). Occasionally acne
fulminans has been reported and necessitated discontinuation of therapy
(173, 176). A causal relationship with androgen therapy is likely as
shown by Fyrand et al. (177). Hinkel et al. (178)
investigated the effects of high doses of androgens on lipoproteins
during and after the cessation of therapy. Although during treatment a
significant fall of triglycerides and HDL was observed, all values
normalized after the end of treatment (178). In our studies,
gynecomastia occurred in 13% of the cases. Since gynecomastia is
rather prevalent in population studies in pubertal boys (179), it is
difficult to say whether the condition had increased. One would expect
that treatment would have effects on sexuality (sex interest,
masturbation). Although in one study a marked increase of sexuality in
younger, but not in older patients, was noted, it never exceeded the
normal range seen in adolescence (98). Treatment with
supraphysiological doses of T were not shown to provoke aggressive
behavior in adolescents or young adults (180, 181).
|
|
VIII. Estrogen Treatment in Tall Girls |
|---|
|
TopAbstractI. IntroductionII. Normal vs. Extremes...III. Endocrinology of Bone...IV. Sex Steroid Action...V. Height PredictionVI. Treatment of CTS:...VII. Treatment of...VIII. Estrogen Treatment in...IX. Alternative Treatment...X. Conclusions and...References |
|---|
B. Height reduction
Studies on the effect of height reduction in tall girls have shown
various results using different prediction models within the same study
population (104, 105, 107, 108, 109, 124, 127, 129). The mean calculated
effect of therapy is ‘corrected’ by subtracting the systematic
prediction error, as has been commonly reported in the literature. The
mean reported height reduction (‘corrected’ and uncorrected) in girls
with CTS ranged from 2.1 to 10 cm (64, 101, 103, 104, 105, 106, 107, 108, 109, 112, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131)
(see Table 8). A clear comparison,
however, is hampered by differences in initial clinical data
(especially CA and BA), duration of treatment, therapeutic regimen
(different doses and estrogen preparations), and the point in time of
final height assessment. Concerning the latter, De Waal et
al. (101) observed a mean (SD) additional growth of
2.7 (1.1) cm after cessation of therapy, which is of the same order of
magnitude as in boys (107, 123, 127). The cause of the observed
additional growth is not quite clear. It could be explained by
cessation of therapy before complete closure of the epiphyses. In
addition, it is plausible that part of the remaining posttreatment
growth reflects additional spinal growth.
|
) that the effect of treatment
was more pronounced when treatment had been started at a younger BA
(105, 108, 112, 123, 130, 131). However, others did not find such a
relationship (103).
|
C. Effects on gonadal function
1. The hypothalamo-pituitary-gonadal axis and menstrual
cycles. Suppression of the hypothalamo-pituitary-gonadal axis
induced by pharmacological doses of estrogens via a negative feedback
mechanism was found to be reversible (136, 192). Hanker et
al. (192) assessed the functional state of the
hypothalamo-pituitary axis by standardized LHRH testing in 16 tall
girls treated with 300 µg EE daily for 7–26 months. While absent LH
responses were observed in all girls immediately after therapy was
stopped, 4 to 8 weeks later the LH responses had normalized in 13 girls
and 12 weeks after therapy in 14 girls. All girls experienced
spontaneous menstrual bleeding within 3 to 22 weeks after termination
of therapy. The same was observed in most later follow-up studies in
which first menstruation was reported within 1–6 months after
cessation of treatment, in most cases even after the first month (103, 114, 127, 130). Amenorrhea of longer than 6 months after cessation of
height-reductive therapy was reported in about 5% of the cases (175).
The incidence of prolonged amenorrhea after cessation of oral
anticonceptive therapy is about 0.5% (193). In addition, the overall
prevalence of secondary amenorrhea of more than 6 months in women aged
15–34 yr is about 1.3% (194). This may suggest an increase of
amenorrhea after height-reductive therapy. It should be noted, however,
that there are no convincing data that use of oral contraceptives (OCs)
is causally related to amenorrhea and that other risk factors for
amenorrhea, such as smoking, nutrition, and exercise, were not
adequately investigated (193, 195). No differences were found in
menstrual cycle characteristics between previously treated and
untreated tall women after a mean follow-up period of almost 11 yr
(175).
2. Pregnancy. In girls, pregnancy, which is the ultimate ’proof’ of complete reversibility of hypothalamo-gonadal suppression, has been reported in various single cases (112, 113, 127, 128, 130, 131, 189). In our own study, information on a total of 63 pregnancies was obtained from 40 previously treated tall women. No distinct differences in details and outcome of pregnancies between treated and untreated tall women were found (175). These results indicate that long-term effects of high doses of estrogens on fertility are unlikely.
Although a mean follow-up period of 10 yr is still too short to draw definite conclusions, there is no clear evidence that treatment with high doses of sex steroids does induce harmful effects on reproductive function in tall girls.
D. Other adverse effects
In most studies, unwanted side effects have been reported only
during treatment or shortly after discontinuation of therapy (109, 112, 113, 127, 128, 130, 131, 183, 189, 196, 197). Most side effects were
found to be mild and reversible (see Table 9). In a large retrospective study
short-term and long-term effects of high doses of estrogens in the
management of CTS were evaluated at a mean follow-up period of 10 yr
after discontinuation of height-reductive therapy. OCs were used by a
high proportion of previously estrogen-treated girls as well as
controls. An impressive bulk of data on the association between
long-term OC use and possible health risks (reviewed in Refs. 198, 199) form an excellent reflection of the prospective risk in
estrogen-treated girls. The proportion of OC’s use and reported side
effects were not significantly different between estrogen-treated and
nontreated women. More than 70% of the estrogen-treated women had
experienced one or more side effects during therapy, such as weight
gain, headache, nausea, leg cramps at night, increased pigmentation of
areola and nipples, and vaginal discharge. Although most of them were
mild and transient, the adverse effects of estrogens occurred more
frequently during therapy than during OC use (175). This would suggest
a dose-dependent effect of estrogens on the incidence of adverse events
(107, 130).
|
The effect of high-dose estrogens on lipid metabolism was evaluated in several studies (178, 204). The changes of serum lipids and lipoproteins during estrogen therapy were reversible after cessation of treatment.
Malignancy was not reported in our follow-up study (175). Although there have been no reported cases of ovarian, uterine, vaginal, or breast malignancies in girls treated for tall stature, the risk of cancer in young women receiving estrogens remains uncertain. The possibility of a dose-dependent effect and a relationship with OC use at a young age and duration of OC use with increased risks of breast cancer (205) point to the need for long-term follow-up of patients treated with pharmacological doses of estrogens.
|
IX. Alternative Treatment Modalities and Future Research |
|---|
|
TopAbstractI. IntroductionII. Normal vs. Extremes...III. Endocrinology of Bone...IV. Sex Steroid Action...V. Height PredictionVI. Treatment of CTS:...VII. Treatment of...VIII. Estrogen Treatment in...IX. Alternative Treatment...X. Conclusions and...References |
|---|
)
represent the 95% confidence interval of the calculated adjusted
effect. [Reproduced with permission from W. J. de Waal et
al.: J Clin Endocrinol Metab 81:1206–1216,
1996 (