0021-972X/99/$03.00/0
The Journal of Clinical Endocrinology & Metabolism
Copyright © 1999 by The Endocrine Society
Vol. 84, No. 6
Printed in U.S.A.
Final Height after Long-Term Treatment with
Triptorelin Slow Release for Central Precocious Puberty:
Importance of Statural Growth after Interruption of
Treatment
JEAN-CLAUDE CAREL, MARC ROGER, SIMONA ISPAS, FRANÇOISE TONDU,
NAJIBA LAHLOU, JOELLE BLUMBERG, JEAN-LOUIS CHAUSSAIN, AND
THE FRENCH STUDY GROUP OF DECAPEPTYL IN PRECOCIOUS PUBERTY*
Department of Pediatric Endocrinology (J.-C.C., J.-L.C.) and Laboratoire de Biochimie Hormonale
(M.R., N.L.), and INSERM U-342 (J.-C.C., J.-L.C., M.R., N.L.), Hôpital Saint Vincent de Paul, 75014
Paris; and IPSEN-BIOTECH Laboratories (S.I., F.T., J.B.), 75014 Paris, France
ABSTRACT
The impact of treatment of central precocious puberty (CPP) with
GnRH agonists on final statural height (FH) remains controversial,
and guidelines on the optimal time point for interruption of these
treatments have not been established. We analyzed the long term
results of 58 girls and 8 boys uniformly treated with triptorelin slow
release formulation (Decapeptyl, triptorelin-SR) for CPP and compared their FH with predicted height before treatment and with the
FH of a historical group of patients not treated with GnRH agonist.
The FH SD score was close to 0 and was not different from the genetic
target height. In girls, FH was improved by 4.8 6 5.8 cm compared
with predicted height before treatment and by 8.3 cm by comparison
L
ONG TERM treatment of central precocious puberty
(CPP) with GnRH agonists efficiently blocks the somatic manifestations of premature sexual development and
their psychological consequences (1–5). One of the aims of
this treatment is to improve final height (FH), compromised
by the premature exposure of growth plates to sex steroids.
As controlled studies are not feasible in this situation, evaluation of the true benefit regarding FH has to rely on indirect
methods, namely comparisons of attained height with predicted height at the initiation of treatment or with untreated
historical patients.
Reported results of GnRH agonists on FH in CPP have
been very variable, ranging from a complete restoration of
growth prognosis to partial or absent benefit (6 –15). Several
reasons might explain these discrepancies, in particular the
heterogeneity of treated patients and the use of different
GnRH agonists.
Although the criteria for the diagnosis of CPP and for the
with a historical group. In boys, comparison with a historical group
revealed a 13.7-cm improvement, whereas predicted height before
treatment was similar to FH. Three variables were independently
associated with FH in girls: the bone age/statural age ratio at the
onset of treatment (negatively), the height SD score at the end of
treatment, and the posttreatment growth spurt (D FH 2 height at the
end of treatment). The influence of the posttreatment growth spurt,
itself dependent on age and bone age at the interruption of treatment,
suggests that continuing treatment beyond the age of 11 yr in girls
does not improve and could actually decrease FH. This point should
be evaluated in a formal controlled trial. (J Clin Endocrinol Metab 84:
1973–1978, 1999)
indication of treatment with GnRH agonists, have been
widely discussed, the optimal age for interruption of these
treatment has not been the focus of much attention. Recent
textbooks recommend interrupting treatment “at an age acceptable for the onset of puberty” (16) or acknowledge that
the “optimal time for discontinuation of therapy has yet to
be determined” (17). This issue somewhat overlaps with the
problem of the use of GnRH agonists in normal short children, as the decision to interrupt GnRH agonists in patients
with CPP is made around the normal age of puberty. Although blocking puberty in short normal children transiently
improves height prognosis, the results for FH are very disappointing (18), with the possible exception of patients
treated for extended periods (19).
We report here FH results for 58 girls and 8 boys uniformly
treated with triptorelin slow release formulation (Decapeptyl, triptorelin-SR) for CPP.
Subjects and Methods
Received September 1, 1998. Revision received January 8, 1999. Accepted March 15, 1999.
Address all correspondence and requests for reprints to: Dr. JeanClaude Carel, INSERM U-342, Hôpital Saint Vincent de Paul, 82 avenue
Denfert Rochereau, 75014 Paris, France. E-mail: [email protected].
* The French Study Group of Decapeptyl in Precocious Puberty is
made up of Pascale Berlier, Michel Bost, Pierre Chatelain, Michel Colle,
Paul Czernichow, Michel David, Patrick Garandeau, Marcel Lecornu,
Georges Malpuech, Roger Mariani, Charles Sultan, and Jean-Edmond
Toublanc.
Patients
All of the patients included in a French multicenter trial of triptorelin-SR in CPP that had been followed to FH (58 girls and 8 boys among
83 girls and 10 boys who participated in the study) were included in this
study. Twenty-seven patients who participated in the French multicenter trial were not included in this analysis for the following reasons:
onset of puberty after the age of 8 yr in girls or 10 yr in boys (n 5 4),
duration of treatment less than 2 yr (n 5 4), and loss of follow-up (n 5
17). Two girls with severe intrauterine growth retardation and FH of 148
1973
1974
JCE & M • 1999
Vol 84 • No 6
CAREL ET AL.
and 142 cm were also excluded because their short stature was not due
to CPP.
The principal characteristics of this trial were as follows.
TABLE 1. Initial characteristics of the patients
Inclusion criteria. The inclusion criteria were onset of symptoms of precocious sexual development before the age of 8 yr in girls and 10 yr in
boys, plasma testosterone levels greater than 0.5 ng/mL in boys, and
pubertal response of LH to GnRH in both sexes (20).
Age at clinical onset of symptoms (yr)
Treatment. Triptorelin-SR (Decapeptyl, Ipsen-Biotech, Paris, France) was
given im every 28 days at a dose of 3.75 mg in children weighing more
than 20 kg and 1.87 mg in children weighing less than 20 kg as previously
described (21). The efficacy of the treatment was assessed by the periodic
evaluation of clinical pubertal development, measurement of plasma sex
steroids, and LH response to GnRH. Treatment was given for at least 2
yr, and the decision to interrupt the treatment was left to the judgment
of individual physicians following the patients and was mainly based
on chronological age, bone age (BA), and height.
Follow-up. Patients were considered as having reached their FH if they
had reached a BA of 15 yr in girls and 17 yr in boys (99% of FH by the
Bayley-Pinneau method) or if their growth rate was inferior to 1 cm/yr,
documented on at least two measurements 6 months apart. No correction was made to predict FH from BA at this stage.
Methods
BA was determined by the method of Greulich and Pyle (22). Height
prognosis was calculated according to Bayley-Pinneau (23). French
growth standards were used to calculate height sd score (24), and target
height was calculated as the midparental height sd score. As hormonal
measurements were not centralized and were performed at various
centers with different methods, we cannot homogeneously present hormonal values, in particular gonadotropin levels (25). However, as each
center had at each time point validated its method of gonadotropin
measurement in prepubertal normal controls, every center could code
the result of peak plasma LH response to GnRH test in one of three
categories: type 1, peak LH value equal to or less than the median value
observed in prepubertal children; type 2, peak LH value comprised
between the median value of prepubertal children and the lower limit
in pubertal children; and type 3, peak LH value superior to the lower
limit in pubertal children.
Historical controls
Data on FH and age at onset of puberty of untreated children with
CPP (58 girls and 25 boys) were collected from published reports of 24
girls and 13 boys (26), 8 girls and 4 boys (11), and 26 girls and 8 boys
(27). Data from children treated with progestins for CPP at Hospital Saint
Vincent de Paul were obtained (28 girls and 1 boy). Data from untreated
and progestin-treated children were not statistically different and were
combined. We also age-matched patients and controls on the basis of age
at diagnosis of CPP. The pairs were selected to have an age difference
of 0.5 yr or less, and when more than 1 patient was available for a pair,
we randomly selected among them.
Statistical methods. Results are expressed as the mean 6 sd. Analyses
were performed with paired Student’s t test (significance level, a 5 0.05,
two-tailed), Wilcoxon test, and Pearson correlation analysis. To determine the determinants of FH, a stepwise multiple regression analysis
was performed. The predictors used in the final model were the parameters showing a significant correlation with FH in the univariate
analysis. To evaluate the effect of qualitative predictors and to compare
control and treated patients, a variance analysis was also performed.
Results
Initial characteristics of the patients (Table 1)
All patients had clinical criteria for true CPP, with an early
onset of clinical symptoms of puberty, increased growth
velocity, and advanced BA maturation. The peak LH value
was coded as described in Subjects and Methods. In girls, 69%
of the GnRH tests were considered indicative of an evolutive
Age at initiation of treatment (yr)
Ht at initiation of treatment (SD score)
BA at initiation of treatment (yr)
BA/statural age at initiation of
treatment
Growth velocity before treatment
(cm/yr)
Etiology of CPP
Idiopathic
Neurogenic
Peak LH value to GnRH test
Type 1 (prepubertal)
Type 2 (intermediate)
Type 3 (pubertal)
Girls (n 5 58)
Boys (n 5 8)
6.3 6 1.5
[0.9/8]
7.5 6 1.3
[3.8/9]
2.4 6 1.5
[21.5/5.5]
10.1 6 1.5
[6/12]
1.1 6 0.1
[0.9/1.3]
8.4 6 2.2
[5/13.5]
7.1 6 3
[1.3/9.7]
9.1 6 1.7
[6.6/10.9]
2.2 6 2.4
[21.6/5.5]
11.6 6 1.7
[9/13.5]
1.1 6 0.1
[0.9/1.2]
8.9 6 1.9
[6.4/11]
57/58 (98)
1/58 (2)
6/8 (75)
2/8 (25)
6/55 (11)
11/55 (20)
38/55 (69)
0/8
0/8
8/8
Mean 6 SD and ranges ([/]) are presented. Percentages are in
parentheses.
TABLE 2. Auxologic parameters during and after treatment
Girls
Duration of treatment (yr)
Age at the end of treatment (yr)
Ht at the end of treatment (cm)
Ht at the end of treatment (SD score)
BA at the end of treatment (yr)
Growth velocity during treatment
(cm/yr)
Variation in BA during treatment
(yr of BA/yr)
DFH 2 ht at the end of treatment
(cm)
Age at the end of follow-up (yr)
Ht at the end of follow-up (FH, cm)
Ht at the end of follow-up
(FH, SD score)a
Boys
3.7 6 1.5
4.7 6 1.8
[1.9/9]
[1.9/7.4]
11.2 6 1
13.8 6 1.5
[8.5/13.8]
[12.4/16.4]
150.1 6 5.3
163 6 6.3
[139.6 –162] [150.3–170]
1.4 6 1.5
0.8 6 2.0
[21.9/4.6]
[22.3/3.3]
12.2 6 0.8
14.3 6 0.7
[10.5/14]
[13.3/15.5]
5.0 6 1.1
4.6 6 0.8
[2.1/7.5]
[3.3/5.7]
0.5 6 0.2
0.6 6 0.2
[0.1/1.3]
[0.3/0.8]
11.1 6 4.7
9.9 6 3.3
[1.0/23.5]
[4.3/14.5]
15.3 6 1.4
17.2 6 1.0
[12.5/18]
[15.4/18]
161.1 6 5.9 172.8 6 6.4
[149.5/173.2] [159.9/179.5]
20.4 6 1.1
20.4 6 1.1
[22.5/1.7]
[22.5/0.7]
Mean 6 SD and ranges ([/]) are presented.
a
Using reference values for adults.
puberty, whereas 20% and 11% of the tests were intermediate
or clearly prepubertal, indicating that some of the girls had
been treated on the grounds of clinical symptoms rather than
on the basis of GnRH test. In girls, there was a 1.2 6 1-yr delay
(range, 0 –5) between the onset of clinical symptoms and the
initiation of treatment. Fourteen girls and five boys had been
treated with progestins before triptorelin-SR.
Auxological and hormonal parameters during and after
treatment (Tables 2 and 3)
As previously described (28), triptorelin-SR efficiently
suppressed the gonadal axis, as illustrated by a suppressed
LH response to GnRH. In girls, an average of 0.7 6 0.3 GnRH
FINAL HEIGHT AFTER TREATMENT OF CPP WITH GnRH AGONISTS
1975
TABLE 3. Final height in triptorelin-SR-treated and historical control patients
Girls
Age at clinical onset of
symptoms (yr)
Target ht (cm)
Predicted ht before
treatment (cm)
Predicted ht at the
end of treatment (cm)
FH (cm)
Boys
Triptorelin-SR (n 5 58)
Historical controls (n 5 86)
Triptorelin-SR (n 5 8)
6.3 6 1.6a
[0.9/8]
160.1 6 4.4
[150.5/171]
156.4 6 6.3
[143.2/169.9]
162.4 6 6.4
[147.2/176.4]
161.1 6 5.9c
[149.5/173.2]
5.3 6 1.9
[0.3/7.8]
7.1 6 3b
[1.3/9.7]
171.8 6 3.7
168/180
174.2 6 6.6
[163.5/184.7]
175.3 6 4.4
[168.8/182.7]
172.8 6 6.4d
[159.9/179.5]
152.3 6 7.6
[134/171.5]
Historical controls (n 5 26)
4.8 6 2.9
[0.7/9.0]
156.6 6 7.5
143/171
Mean 6 SD and ranges ([/]) are presented.
a
P 5 0.001 vs. historical control group (by variance analysis).
b
P 5 0.01 vs. historical control group (by Wilcoxon test).
c
P , 0.001 vs. predicted height before treatment and P 5 0.02 vs. predicted height at the end of treatment; P 5 NS vs. target height (by
Student’s t test); P , 0.001 vs. the historical control group (by variance analysis after adjustment for age at diagnosis).
d
P , 0.001 vs. the historical control group (by variance analysis after adjustment for age at diagnosis).
test/yr was performed during treatment; 98% of these tests
indicated gonadotropin suppression, whereas 1% were type
2 (intermediate) and 0.7% were type 3 (no suppression). In
boys, an average of 0.6 6 0.2 GnRH test/yr was performed
during treatment; 95% of these tests indicated gonadotropin
suppression, and 5% were type 2. Under further follow-up,
GnRH tests indicated adequate suppression in these patients,
suggesting that poor compliance was the explanation for
incomplete suppression.
During treatment in girls, growth velocity declined from
8.4 6 2.2 cm/yr to 5.9 6 1.1 cm/yr during the first year, 5.3 6
1.3 cm/yr during the second year of treatment, and 4.0 6 1.2
cm/yr during the rest of the treatment (time effect, P ,
0.001). BA maturation declined to 0.5 6 0.2 BA yr/yr during
the entire treatment period. This led to an improvement of
both BA/statural age and height prognosis at the end of the
treatment (Table 3).
Treatment was interrupted at 11.2 6 1.0 yr in girls with
very wide variations, as no precise guidelines had been given
to individual physicians participating in the study. After
interruption of the treatment, clinical pubertal development
resumed in all cases and was biologically confirmed by a
GnRH test in most patients as previously described (29). Girls
grew 11.1 6 4.7 cm after interruption of the treatment,
whereas boys grew 9.9 6 3.3 cm, resulting in FH of 161.1 6
5.9 and 172.8 6 6.4 cm, respectively.
Improvement of FH after triptorelin-SR treatment (Table 3)
The FH of treated patients was similar to the midparental
height sd score. In girls, FH exceeded height predicted before
treatment by 4.8 6 5.8 cm, with wide variations (27.7 to
119.1 cm; P , 0.001). There was a small decrease in FH
compared to predicted height at the end of treatment (21.2 6
3.8 cm; range, 212.5 to 16.0; P , 0.03). In contrast, in boys,
predicted height before treatment was similar to target
height and to the achieved FH.
As height benefit and more generally the use of GnRH
agonists are discussed in girls presenting with CPP after 6 or
7 yr of age, we analyzed FH gain (i.e. the difference between
predicted and actual FH) in girls with first signs of puberty
occurring before (n 5 16) or after (n 5 42) 6 yr of age. FH gain
FIG. 1. Variation in final height gain (difference between final height
and predicted height before treatment) with age at onset of pubertal
signs (r 5 0.12; P 5 0.38).
was 5.3 6 7.2 cm in younger and 4.5 6 5.3 cm in older girls
(P , 0.02 and P , 0.001, respectively, predicted vs. FH, by
Wilcoxon test; P 5 NS, younger vs. older girls). No correlation was found between FH gain and age at onset of puberty
(Fig. 1) or age at initiation of treatment.
As FH prediction using the Bayley-Pinneau method is not
always accurate, in a situation where BA maturation does not
follow the normal evolution (30) we also compared treated
patients to an historical group of patients, as described in
Subjects and Methods (Table 3). At the time of diagnosis,
patients in the treated group were older than those in the
control group. Variance analysis, after adjustment for the age
at diagnosis, indicated a positive effect of triptorelin-SR treatment on FH. The FH difference between treated patients and
controls, after adjustment for age at diagnosis, was 8.3 cm in
girls and 13.7 cm in boys. To further ascertain that the difference in FH between patients and controls was not due to
the age difference between the two groups, the comparison
was made between age-matched pairs of treated and untreated patients, as described in Subjects and Methods. In the
39 pairs of girls (age difference, 0.02 6 0.1 yr), the FH difference was 8.9 6 8.7 cm (P , 0.0001, by Wilcoxon test). The
FH difference between the treated and control girl was 5 cm
1976
JCE & M • 1999
Vol 84 • No 6
CAREL ET AL.
or more in 64% (25 of 39) of pairs. In the 7 pairs of boys, the
FH difference was 12.9 6 8.1 cm (P , 0.02, by Wilcoxon test),
with an age difference of 0.2 6 0.4 yr (P 5 NS). The FH
difference between the treated and control boys was 5 cm or
more in 86% (6 of 7) of pairs.
Factors predicting FH in triptorelin-SR-treated girls
Factors affecting FH were analyzed in treated girls only,
because the small number of boys did not allow such an
analysis. Predictors in the univariate analysis (Table 4) were
used to construct a multivariate model (Table 5). Three explanatory variables were found that could explain 66% of the
variance of FH, one of them descriptive of the initial characteristics of the patients and two related to interruption of
treatment. The BA/statural age ratio was a negative predictor, indicating that severe initial characteristics had a detrimental effect, not completely reversed by the treatment. The
height sd score at the end of treatment was positively associated with outcome, whereas the effect of target height or
height before treatment disappeared in the multivariate analysis. Unexpectedly, height gain after interruption of the treatment (averaging 11 cm) was a predictor of FH with a slope
of 0.54, indicating that a 1-cm difference in posttreatment
growth accounted for a difference of 0.54 cm in FH when the
other predictors were held constant.
This finding prompted us to analyze separately the factors
associated with height gain after interruption of treatment
(Table 6). Age and BA at the end of treatment were independently and negatively associated with height gain after
treatment.
Discussion
In this report, we have analyzed FH of 58 girls and 8 boys
treated for CPP with a single depot GnRH agonist, triptorelin-SR. The average FH was 161.1 6 5.9 cm in girls and
172.8 6 6.4 cm in boys, values close to the mean height of the
French population. Moreover, these values were identical to
the genetic target height of this group of children and indiTABLE 4. Factors associated with final height (centimeters) in
triptorelin-SR-treated girls: univariate analysis
Predictive factor
r
P
Target ht (cm)
Ht at onset of treatment (SD score)
D BA 2 statural age at onset of
treatment (yr)
BA/statural age at onset of treatment
Growth velocity before treatment (cm/yr)
Predicted FH at onset of treatment (cm)
Growth velocity during treatment (cm/yr)
Growth velocity at the end of treatment
(last value, cm/yr)
D Ht during treatment (SD score)
Ht at the end of treatment (SD score)
Predicted FH at the end of treatment (cm)
Age at the end of treatment (yr)
D FH 2 ht at the end of treatment (cm)
0.654
0.469
20.465
,0.001
,0.001
,0.001
20.470
0.402
0.545
0.508
0.485
,0.001
0.008
,0.001
,0.001
,0.001
0.392
0.656
0.815
20.316
0.522
0.002
,0.001
,0.001
0.016
,0.001
Factors not associated with FH included age at diagnosis and at
onset of treatment, treatment duration, difference between age at
diagnosis and at onset of treatment, BA progression during treatment, BA at the end of treatment, and coded peak LH before treatment.
TABLE 5. Multivariate analysis of predictive factors of final
height in triptorelin-SR-treated girls (in centimeters; r2 of the final
model 5 0.660)
Predictive factor
Slope (95% CI)a
BA/statural age at onset of
treatment
Ht at the end of treatment
(SD score)
D FH 2 ht at the end of
treatment (cm)
216.21 (228.23/24.19)
P
0.002
1.82 (1.02/2.61)
,0.001
0.54 (0.29/0.78)
,0.001
a
The regression coefficient represents the change in FH per unit
change in the predictor.
TABLE 6. Factors associated with D final height 2 height at the
end of treatment in triptorelin-SR-treated girls (centimeters; r2 of
the final model 5 0.539)
Predictive factor
Slope (95% CI)a
P
BA at the end of treatment (yr)
Age at the end of treatment (yr)
23.32 (24.61/22.02)
21.52 (22.48/20.56)
,0.001
,0.001
a
The regression coefficient represents the change in D FH 2 height
at the end of treatment per unit change in the predictor.
cated a 4.8 6 5.8 cm increase, compared to height predicted
before treatment, in girls.
The magnitude of height benefit induced by gonadotropin
suppression in CPP has been widely discussed, mainly due
to methodological problems. First, patients treated for CPP
are heterogeneous in terms of age of onset, progression of
pubertal development, and other factors influencing statural
growth. Second, indirect methods are used, by us and others,
to evaluate the height benefit: comparison with height predicted by the Bayley-Pinneau method or with historical patients. A few reports have evaluated the accuracy of prediction in untreated girls with CPP and indicate that the BayleyPinneau method overestimates FH by 4.2 cm (sd 5 4.4 cm)
(30), 3.7 cm (9), or 5.9 cm (14). Several studies have reported
an increased FH in GnRH agonist-treated girls, measured as
the difference between predicted and attained height (6 –15).
However, the outcomes were variable, with results ranging
from 1 cm (12) to 10 cm in younger children at diagnosis (11).
Our results in a large group of children (14.8 6 5.8 cm) fall
in the average of recently published series of patients (3– 6
cm). Several factors can explain the discrepancies between
these studies, including the use of daily forms of GnRH
agonists in some reports, treatment of patients with long
standing disease in older reports, and heterogeneity of
treated patients. One important and unsolved question is
whether treatment with GnRH agonists has beneficial effects
on height in girls with pubertal onset after the age of 6 or 7
yr, the age of pubertal development in a significant proportion of girls of the general population (31). The analysis of the
subgroup of 42 patients with onset of puberty between the
ages of 6 – 8 yr shows a significant 4.5 6 5.3 cm increase in
FH over predicted height. In addition, no statistical association was found between age at onset of puberty or at initiation of treatment and FH gain or FH itself, suggesting that
girls with late onset CPP benefited from GnRH agonist treatment similarly to girls with earlier pubertal onset.
The other method for estimating height benefit, i.e. the
comparison with an historical control group, revealed an
FINAL HEIGHT AFTER TREATMENT OF CPP WITH GnRH AGONISTS
8.3-cm increase in FH after adjustment for the age at diagnosis and a similar difference after age matching patients and
controls. This result is consistent with those of Paul et al. (11),
who found a difference between treated patients and historical controls of 14.1 and 4.2 cm in patients with onset of
puberty before and after the age of 5 yr, respectively. Altogether, if we take into account the probable overestimation
of height prognosis in the absence of treatment, the true
benefit in a group of patients such as ours is probably in the
range of 8 –10 cm.
The availability of FH in a large group of treated girls
allowed the analysis of factors associated with end results.
We elected to use as the end point variable FH itself rather
than the difference from predicted height because it avoids
the error associated with prediction methods and because
BA, a major determinant of height prognosis, could be analyzed as an independent variable. One variable descriptive
of the patients at the onset of treatment and two variables
relevant to the interruption of treatment explained 66% of the
variance in FH. The negative association of BA/statural age
at the onset of treatment emphasizes the fact that GnRH
agonist treatment is not capable of restoring a full FH potential if started after an irreversible advancement of BA.
Indeed, the slope of 216 indicated that a variation in this
ratio of 0.1 altered FH by 1.6 cm. The second variable, height
sd score at the end of treatment, is the most difficult to
integrate, because it results from genetic height potential, the
accelerating effect of CPP, and the decelerating effect of
GnRH agonist treatment. Its slope (21.82) indicates that a
difference of 1 sd at interruption of treatment (roughly 5 cm)
accounted for a 1.8-cm difference in FH. The last predictive
variable, the amplitude of the growth spurt after the end of
treatment, is of major interest because it is dependent on the
time point selected for interrupting the treatment. This lead
us to analyze separately the determinants of posttreatment
growth; BA and age at interruption of treatment were negative independent predictors, indicating that discontinuing
the treatment late had a negative effect directly and through
BA advancement. This finding is reminiscent of the model
proposed by Bourguignon in sex steroid-treated patients
with hypopituitarism (32). Similarly, we found that interruption of puberty with GnRH agonists in girls with short
stature and a normal timing of puberty has a limited effect,
indicating that increasing prepubertal growth induces an
equivalent decrease in pubertal growth (18). Although much
emphasis has been put on the criteria for initiation of GnRH
agonist treatment (3), determining the optimal time point for
interruption of treatment has not received much attention. In
a published series of patients, the mean age at interruption
of treatment ranged from 10.6 –11.3 yr (9, 10, 12–15). In our
patients, the age at interruption of treatment was similar
(11.2 6 1 yr), but the dispersion was somewhat wider. To put
things into perspective, it can be derived from our data that
an 11-yr-old girl, growing 4 cm and gaining 0.5 BA yr in a
year, will lose 2.5 cm of FH if treatment is discontinued 1 yr
later (combined effects of age, BA, and decreased height sd
score). Therefore, until this point is evaluated through formal
trials, we recommend stopping GnRH agonist treatment no
later than 11 yr of age in girls.
The small number of boys did not allow a thorough anal-
1977
ysis of the variables associated with the effect of treatment,
as in other studies (7, 11, 13, 15). Initial height prognosis was
not impaired compared with target height in our patients as
well as in those reported by Oostdijk et al. (13) due to a
moderate increment in BA. This should be regarded with
caution due to the lack of data on the accuracy of height
prediction in boys with CPP and to the markedly decreased
FH in untreated boys. In consequence, only the comparison
with historical controls indicated a FH benefit in boys.
In conclusion, our data confirm the restoration of genetic
target height or a 5- to 10-cm increase in FH in girls treated
with triptorelin-SR for CPP. We highlight a specific point of
important practical and economical consequences, i.e. the
optimal time for interrupting GnRH agonist treatments with
respect to the improvement of FH. Although this point
should certainly be better evaluated in a formal controlled
trial, our results suggest that continuing GnRH agonists beyond the age of 11 yr in girls does not improve and could
actually decrease FH by reducing the posttreatment growth
spurt.
Acknowledgments
We thank Drs. M. M. Grumbach and E. Thamdrup for providing us
with results in untreated children.
References
1. Kaplan SL, Grumbach MM. 1990 Clinical review 14: pathophysiology and
treatment of sexual precocity. J Clin Endocrinol Metab. 71:785–789.
2. Kletter GB, Kelch RP. 1994 Clinical review 60: effects of gonadotrophinreleasing hormone analog therapy on adult stature in precocious puberty.
J Clin Endocrinol Metab. 79:331–334.
3. Rosenfield RL. 1994 Selection of children with precocious puberty for treatment with gonadotropin releasing hormone analogs. J Pediatr. 124:989 –991.
4. Conn PM, Crowley WFJ. 1994 Gonadotropin-releasing hormone and its analogs. Annu Rev Med. 45:391– 405.
5. Kappy MS, Ganong CS. 1994 Advances in the treatment of precocious puberty. Adv Pediatr. 41:223–261.
6. Kauli R, Kornreich L, Laron Z. 1990 Pubertal development, growth and final
height in girls with sexual precocity after therapy with the Gn-RH analogue
d-TRP-6-LHRH. A report of 15 girls, followed after cessation of gonadotrophin
suppressive therapy. Horm Res. 33:11–17.
7. Oerter KE, Manasco PK, Barnes KM, Jones J, Hill S, Cutler Jr GB. 1991 Adult
height in precocious puberty after long-term treatment with deslorelin. J Clin
Endocrinol Metab. 73:1235–1240.
8. Partsch CJ, Hummelink R, Peter M, et al. 1993 Comparison of complete and
incomplete suppression of pituitary-gonadal activity in girls with central precocious puberty: influence on growth and predicted final height. The GermanDutch Precocious Puberty Study Group. Horm Res. 39:111–117.
9. Antoniazzi F, Cisternino M, Nizzoli G, et al. 1994 Final height in girls with
central precocious puberty: comparison of two different luteinizing hormonereleasing hormone agonist treatments. Acta Paediatr. 83:1052–1056.
10. Brauner R, Adan L, Malandry F, Zantleifer D. 1994 Adult height in girls with
idiopathic true precocious puberty. J Clin Endocrinol Metab. 79:415– 420.
11. Paul DL, Conte FA, Grumbach MM, Kaplan SL. 1995 Long term effect of
gonadotropin-releasing hormone agonist therapy in children with true precocious puberty treated at a median age of less than 5 years. J Clin Endocrinol
Metab. 80:546 –551.
12. Stasiowska B, Vannelli S, Benso L. 1994 Final height in sexually precocious
girls after therapy with an intranasal analogue of gonadotrophin-releasing
hormone (buserelin). Horm Res. 42:81– 85.
13. Oostdijk W, Rikken B, Schreuder S, et al. 1996 Final height in central precocious puberty after long term treatment with a slow release GnRH agonist.
Arch Dis Child. 75:292–297.
14. Kauli R, Galatzer A, Kornreich L, Lazar L, Pertzelan A, Laron Z. 1997 Final
height of girls with central precocious puberty, untreated vs. treated with
cyproterone acetate or GnRH analogue. A comparative study with re-evaluation of predictions by the Bayley-Pinneau method. Horm Res. 47:54 – 61.
15. Galuzzi F, Salti R, Bindi G, Pasquini E, La Cauza C. 1998 Adult height
comparison between boys and girls with precocious puberty after long-term
gonadotrophin-releasing hormone analogue therapy. Acta Paediatr.
87:521–527.
1978
CAREL ET AL.
16. Bridges NA, Brook CGD. 1995 Disorders of puberty. In: Brook CGD ed.
Clinical paediatric endocrinology, 3rd ed. Oxford: Blackwell; 253–273.
17. Blizzard RM, Rogol AD. 1994 Variations and disorders of pubertal development. In: Kappy MS, Blizzard RM, Migeon CJ, eds. The diagnosis and treatment of endocrine disorders in childhood and adolescence, 4th ed. Springfield:
Thomas; 857–918.
18. Carel JC, Hay F, Coutant R, Rodrigue D, Chaussain JL. 1996 Gonadotropin
releasing hormone agonist treatment of girls with constitutional short stature
and normal pubertal development. J Clin Endocrinol Metab. 81:3318 –3322.
19. Cutler GBJr, Yanovski JA, Rose SR, et al. 1997 Luteinizing hormone-releasing
hormone agonist (LHRHa)-induced delay of epiphyseal fusion increases adult
height of adolescents with short stature. Horm Res. 48S2:28.
20. Roger M, Lahlou N, Chaussain JL. 1996 Gonadotropin-releasing hormone
testing in pediatrics. In: Ranke MB ed. Diagnostics of endocrine function in
children and adolescents. Heidelberg: Verlag; 346 –369.
21. Boucekkine C, Blumberg Tick J, Roger M, Thomas F, Chaussain JL. 1994
Treatment of central precocious puberty with sustained-release triptorelin.
Arch Pediatr. 1:1127–1137.
22. Greulich WW, Pyle SI. 1959 Radiographic atlas of skeletal development of the
hand and wrist. Stanford: Stanford University Press;.
23. Bayley N, Pinneau S. 1952 Tables for predicting adult height from skeletal age.
J Pediatr. 14:432– 441.
24. Sempé M, Pédron G, Roy P. 1979 Auxologie, méhodes et séquences. Paris:
Théraplix.
JCE & M • 1999
Vol 84 • No 6
25. Chuang-Stein C. 1992 Summarizing laboratory data with different reference
ranges in multi-center clinical trials. Drug Information J. 26:77– 84.
26. Sigurjonsdottir TJ, Hayles AB. 1968 Precocious puberty. A report of 96 cases.
Am J Dis Child. 115:309 –321.
27. Thamdrup E. 1961 Precocious sexual development. A clinical study of 100
children. Copenhangen: Munksgaard.
28. Roger M, Chaussain JL, Berlier P, et al. 1986 Long term treatment of male and
female precocious puberty by periodic administration of long-acting preparation of [scap]d-Trp6-luteinizing hormone-releasing hormone microcapsules.
J Clin Endocrinol Metab. 62:670 – 677.
29. Chaussain JL, Swaenpoel C, Bost M, et al. 1993 Growth and ovarian function
in girls with central precocious puberty after interruption of LHRH agonist
therapu. In: Grave GD, Cutler GB, eds. Sexual precocity: etiology, diagnosis,
management. New York: Raven Press; 55–59.
30. Bar A, Linder B, Sobel EH, Saenger P, DiMartino-Nardi J. 1995 BayleyPinneau method of height prediction in girls with central precocious puberty:
correlation with adult height. J Pediatr. 126:955–958.
31. Herman-Giddens ME, Slora EJ, Wasserman RC, et al. 1997 Secondary
sexual characteristics and menses in young girls seen in office practice: a
study from the Pediatric Research in Office Settings network. Pediatrics.
99:505–512.
32. Bourguignon JP. 1988 Linear growth as a function of age at onset of puberty
and sex steroid dosage: therapeutic implications. Endocr Rev. 9:467– 488.
Erratum
In the article “Adult Height in Girls with Central Precocious Puberty Treated with Gonadotropin-Releasing
Hormone Analogues and Growth Hormone” by Anna Maria Pasquino, Ida Pucarelli, Maria Segni, Marco
Matrunola, and Fabio Cerroni (The Journal of Clinical Endocrinology & Metabolism, 1999, vol. 84: 449 – 452), one
of the authors’ names was misspelled.
Fabio Cerroni is the correct spelling for the last author.