Near normalization of adolescent height with growth hormone therapy in very short children without growth hormone deficiency C. L e s a g e , MD, J. Walker, MD, F. Landier, MD, P. C h a t e l a i n , MD, J. L. Chaussain, MD, a n d P. F. Bougn~res, MD From the Service d'Endocrinologie and U 342 INSERM,H6pital St Vincent de Paul, Paris, and U 307 INSERM,Lyon, France Ten prepubertal children with stature at or b e l o w the 1st percentile for height and without growth hormone d e f i c i e n c y received 0.3 U recombinant growth hormone per kilogram daily for 2 years before puberty. Their growth velocity increased from 4 _ 0.3 cm/yr before treatment to 10.7 • 0.6 and 8.8 • 0.6 cm, respectively, during the first and s e c o n d years of treatment, and then remained at 5.7 _+ 0.7 cm the year after the end of growth hormone administration. This resulted in a near normalization of a d o l e s c e n t height. Bone maturation paralleled c h r o n o l o g i c age, and therefore the expected final height of the children increased by a p p r o x i m a t e l y 10 cm. Administration of growth hormone induced a reversible hyperinsulinemia, with moderate and transient changes in glucose metabolism. A prospective, randomized study, including an untreated cohort, will be n e e d e d to confirm the effects on final height and to determine the magnitude of the response in familial short stature. (J PEDIATR1991;119:29-34)
Several short-term (6 to 12 months) studies have demonstrated that growth hormone at a dose of 0.05 to 0.1 U / k g per day can accelerate the growth of constitutionally short children who have no demonstrable deficit of GH secretion. 1"8 However, the limited magnitude of the height gain, the high cost of the treatment, and the concerns expressed by some pediatric endocrinologists9"13 have excluded the vast majority of such children from receiving GH treatment, even if their final stature is expected to be severely impaired. There is debate, however, regarding the appropriate use of recombinant GH; some pediatricians argue that it should be offered to very short children who have no deficiency of growth hormone secretion. TM We attempted to evaluate the feasibility of normalizing children's height when it was below the 1st percentile by Submitted for publication Oct. 8, 1990; accepted Jan. 28, 1991. Reprint requests: P. F. Bougn6res, Service d'Endocrinologie et U342 INSERM, Hopital St-Vincent de Paul, 82 Avenue DenfertRochereau, 75014 Paris, France. 9/20/28306
administering larger doses of G H for a period of 2 years preceding puberty. Our aim was to allow these children to enter adolescence with nearly normal stature and avoid the potential psychologic problems related to severe shortness. 15 Extrapolating from studies in both hypopituitary 16-1s and non-GH-deficient children, 1"8 we calculated that, for the growth rate to reach the desired value of 10 GH IGF-I OGTT
Growth hormone Insulin-like growth factor I Oral glucose tolerance test
cm/yr, GH should be given at a dose of ~0.3 U / k g per day. The trial, because of its preliminary nature, was not randomized and included a relatively limited number of children. METHODS Subjects. Ten children (seven boys) with short stature were included in the trial after the principles of therapy had been carefully explained to both them and their parents, To
29
30
Lesage et al.
The Journal of Pediatrics July 1991
Table I. Clinical characteristics of the studied children During GH treatment Before GH treatment
Chronologic age (yr) Boys (n = 7) Girls (n = 3) Weight (kg) Boys Girls SDS for height Height (cm) Boys Girls SDS CA SDS BA Growth rate (cm/yr) Boys Girls SDS BA Bone age (yr, GP/TW) Boys Girls Pubertal stages (n, M/F) Tanner I Tanner I1 Tanner III Tanner IV
At 12 mo
At 24 mo
At 12 mo after GH treatment
11.0 • 0.9 9.8 • 0.4
12.0 • 0.9 10.8 _+ 0.4
13.0 • 0.9 i1.8 • 0.4
I4.0 • 0.9 12.8 • 0.4
25.2 • 1.9 20.7 • 1.2 +0.31
32 • 2.3 25 _+ 1.4 +0.73
36.5 • 3.5 30 _+ 1.4 +0.17
43.3 _+ 3.7 34.8 • 2.3 +0.76
3.7 1.9 0.3 0.4
133.5 • 3.7 127.1 • 1.5 -2.1 • 0.3 -0.61
142 • 4 134.5 _+ 1.9 -1.6 _+ 0.2 -0.14
146.5 • 4.5 141.5 _+ 1.4 -1.7 • 0.3 -0.34
4.2 • 0.4 3.7 • 0.2 -1.65
10.5 • 0.7 t 1.0 _+ 0.3 6.9
9.4 _+ 1.2 7.3 • 0.7 3.1
5.2 _+ 0.7 7.0 • 0.7 -0.7
9,5/9.3 7.9/7.7
10.5/10.3 9.4/8.9
11.6/11.4 10.1/9.9
12.6/12.5 11.1/11
6/1 1/2 ---
5/1 2/2 ---
3/l 2/1 2/1 --
2/0 3/2 1/1 1/0
123.7 116 -3.1 -1.6
• • • +
SDS, Standard deviation score. CA, chronologicage; BA, bone age; GP/TW,GreuIich-Pyle/Tanner-Whitehousemethods.
be eligible, the children had to be prepubertal and to have a height at least 2.5 SD below the mean for chronologic age.19 Plasma G H response to pharmacologic stimuli (ornithine, arginine, insulin, glucagon) had to be greater than 12 ~Lg/L on two tests. No other disease could to be present; patients with a personal or family history of glucose intolerance or diabetes mellitus were also excluded. The main characteristics of the children are shown in Table I. The children's mean birth length was 47 _+ 1 cm and weight was 3015 + 189 gm. Their parents' heights were 164 __+ 2 cm (fathers) and 155 _+ 2.5 cm (mothers). We determined bone age by the methods of Greulich and Pyle 2~ and of Tanner and Whitehouse (TW2). 21 Height prediction was calculated with the methods of Bayley and Pinneau 22 and of Tanner et al., 21 including correction for parental height. The mean normal mid-parental height in France is 166.5 cm. 19 Height was expressed as the standard deviation score both for chronologic age and for bone age. Height velocity was expressed as centimeters per year and as the standard deviation score for bone age. The children received 0.3 U / k g body weight per day of biosynthetic G H (Genotropin, from KabiVitrum AB, Stockholm, Sweden), subcutaneously each evening, including the day before each study, for 24 months. Their response was
evaluated before therapy, then after 12 and 24 months of G H administration, and finally 12 months after the discontinuation of the treatment. For at least 3 days before the metabolic tests, the children ate a normocaloric standard diet containing 50% carbohydrates. On day 1 of each study a 12 ml blood sample was drawn after an overnight 12-hour fast for the measurement of concentrations of plasma glucose, free fatty acids, cholesterol, triglycerides, insulin, and insulin growth factor I. An oral glucose tolerance test was performed on day 2 after an overnight fast, with the use of 18% glucose solution at a dose of 45 g m / m 2. Blood samples were collected every 30 minutes for plasma glucose and insulin measurements. Plasma glucose concentration was measured with a glucose oxidase method (Yellow Springs Instruments, Yellow Springs, Ohio). Plasma insulin concentration was measured with double-antibody radioimmunoassay kits purchased from the Commissariat fi l'Energie Atomique, Gif sur Yvette, France. 23 Insulin measurements were performed in triplicate. Plasma I G F - I concentrations were measured with a previously reported radioimmunoassay. 24, 25 Statistical methods. The results are expressed as the mean _+ SE. Comparisons before, during, and after treatment were made with a two-tailed paired t test.
Volume t 19 Number 1, Part 1
Normal adolescent height after growth hormone treatment
31
Table II. Predicted adult height at the different stages of the study During GH treatment Before GH treatment Patient No. Boys 1 2 3 4 5 6 7 Mean _+ SEM Girls 8 9 10 Mean • SEM
At 12mo
At 42 mo after GH treatment
At 24 mo
B-P (cm)
p" (cm)
TW2 (cm)
B-P (cm)
TW2 (cm)
B-P (cm)
TW2 (cm)
B-P (cm)
TW2 (cm)
158.3 145.6 162.0 154.6 154.3 162.5 156.5 156.2 • 2
+1.8 -3.2 -0.8 -4.2 -3.5 -2.5 -1.2 -1.95
161.9 142.8 159.6 156.4 153.5 159.9 156.9 155.9 • 2.2
168.3 154.0 172.3 158.0 161.6 173.3 166.3 164.8 _+ 2.5
170.5 152.9 178.9 162.6 168.2 168.5 162.6 166.3 _+ 2.8
180.1 156.1 179.6 162.6 168 173.3 165.6 169.3 _+ 3.1
170.5 152. 9 178.9 162.6 168.2 168.5 162.6 166.3 _+ 2.7
177 152.2 !71.4 163.3 i69.1 171.2 165 167.0 _+ 2.6
!67.1 154.5 175 161.5 165.5 168.3 163.7 164.9 _+ 2.3
151 154.5 152 152.5 _+ 0.6
-1.8 -2.5 -2.5 -2.3
150.5 152 151 151.2 • 0.4
159.1 160.7 164.1 161.3 • 0.7
154.7 156.2 153 154.7 • 0.5
160.8 157.8 163.5 160.7 -+ 0.9
156.1 157.8 154.1 156 -+ 0.6
160.5 160.5 166.5 162.5 _+ 1
157 157.9 159 158 _+ 0.3
B-P, Bayley-Pinneaumethod22; TW2, Tanner-Whitehouse method,21 including a factor (p) correcting for the measured mid-parental height. *p =
Mid-parental height - 166.5 cm 3
Table III. Plasma substrates, insulin, and I G F - I in the studied children During GH treatment At 42 mo
At entry Plasma glucose (mg/dl) Fasting Area OGTT Free fatty acids (mmol/L) Claolesterol (mmol/L) Triglycerides (retool/L) Plasma Insulin (~U/rnl) Fasting Peak OGGT Plasma IGF-I (U/ml)
85 265 0.72 4.8 0.60
• 6 _+ 33 _+ 0.10 _+ 0.3 +_ 0.04
85 230 0.55 4.6 0.70
6 _+ 0.7 42 _+ 5 0.9 ___0.2
At 24 mo
_+ 6 _+ 21" _ 0.12 +_ 0.4 _+ 0.04
85_+7 254 _+ 32 0.85 _+ 0.15 4.4 _+ 0.3 0.66 _+ 0.08
13 _+ 1.1" 65 _+ 10" 3.2 • 0.5*
16 _+ 2* 104 _+ 21" 2.9 _+ 0.5*
At 42 mo after end of GH treatment 81 251 0.68 5.0 0.74
_+2 _+ 31 _+ 0.10 _+ 0.3 _+ 0.07
5 _+ 0.6 62 + 8 2.6 _+ 0.6*
*p
Several short-term (6 to 12 months) studies have demonstrated that growth hormone at a dose of 0.05 to 0.1 U / k g per day can accelerate the growth of constitutionally short children who have no demonstrable deficit of GH secretion. 1"8 However, the limited magnitude of the height gain, the high cost of the treatment, and the concerns expressed by some pediatric endocrinologists9"13 have excluded the vast majority of such children from receiving GH treatment, even if their final stature is expected to be severely impaired. There is debate, however, regarding the appropriate use of recombinant GH; some pediatricians argue that it should be offered to very short children who have no deficiency of growth hormone secretion. TM We attempted to evaluate the feasibility of normalizing children's height when it was below the 1st percentile by Submitted for publication Oct. 8, 1990; accepted Jan. 28, 1991. Reprint requests: P. F. Bougn6res, Service d'Endocrinologie et U342 INSERM, Hopital St-Vincent de Paul, 82 Avenue DenfertRochereau, 75014 Paris, France. 9/20/28306
administering larger doses of G H for a period of 2 years preceding puberty. Our aim was to allow these children to enter adolescence with nearly normal stature and avoid the potential psychologic problems related to severe shortness. 15 Extrapolating from studies in both hypopituitary 16-1s and non-GH-deficient children, 1"8 we calculated that, for the growth rate to reach the desired value of 10 GH IGF-I OGTT
Growth hormone Insulin-like growth factor I Oral glucose tolerance test
cm/yr, GH should be given at a dose of ~0.3 U / k g per day. The trial, because of its preliminary nature, was not randomized and included a relatively limited number of children. METHODS Subjects. Ten children (seven boys) with short stature were included in the trial after the principles of therapy had been carefully explained to both them and their parents, To
29
30
Lesage et al.
The Journal of Pediatrics July 1991
Table I. Clinical characteristics of the studied children During GH treatment Before GH treatment
Chronologic age (yr) Boys (n = 7) Girls (n = 3) Weight (kg) Boys Girls SDS for height Height (cm) Boys Girls SDS CA SDS BA Growth rate (cm/yr) Boys Girls SDS BA Bone age (yr, GP/TW) Boys Girls Pubertal stages (n, M/F) Tanner I Tanner I1 Tanner III Tanner IV
At 12 mo
At 24 mo
At 12 mo after GH treatment
11.0 • 0.9 9.8 • 0.4
12.0 • 0.9 10.8 _+ 0.4
13.0 • 0.9 i1.8 • 0.4
I4.0 • 0.9 12.8 • 0.4
25.2 • 1.9 20.7 • 1.2 +0.31
32 • 2.3 25 _+ 1.4 +0.73
36.5 • 3.5 30 _+ 1.4 +0.17
43.3 _+ 3.7 34.8 • 2.3 +0.76
3.7 1.9 0.3 0.4
133.5 • 3.7 127.1 • 1.5 -2.1 • 0.3 -0.61
142 • 4 134.5 _+ 1.9 -1.6 _+ 0.2 -0.14
146.5 • 4.5 141.5 _+ 1.4 -1.7 • 0.3 -0.34
4.2 • 0.4 3.7 • 0.2 -1.65
10.5 • 0.7 t 1.0 _+ 0.3 6.9
9.4 _+ 1.2 7.3 • 0.7 3.1
5.2 _+ 0.7 7.0 • 0.7 -0.7
9,5/9.3 7.9/7.7
10.5/10.3 9.4/8.9
11.6/11.4 10.1/9.9
12.6/12.5 11.1/11
6/1 1/2 ---
5/1 2/2 ---
3/l 2/1 2/1 --
2/0 3/2 1/1 1/0
123.7 116 -3.1 -1.6
• • • +
SDS, Standard deviation score. CA, chronologicage; BA, bone age; GP/TW,GreuIich-Pyle/Tanner-Whitehousemethods.
be eligible, the children had to be prepubertal and to have a height at least 2.5 SD below the mean for chronologic age.19 Plasma G H response to pharmacologic stimuli (ornithine, arginine, insulin, glucagon) had to be greater than 12 ~Lg/L on two tests. No other disease could to be present; patients with a personal or family history of glucose intolerance or diabetes mellitus were also excluded. The main characteristics of the children are shown in Table I. The children's mean birth length was 47 _+ 1 cm and weight was 3015 + 189 gm. Their parents' heights were 164 __+ 2 cm (fathers) and 155 _+ 2.5 cm (mothers). We determined bone age by the methods of Greulich and Pyle 2~ and of Tanner and Whitehouse (TW2). 21 Height prediction was calculated with the methods of Bayley and Pinneau 22 and of Tanner et al., 21 including correction for parental height. The mean normal mid-parental height in France is 166.5 cm. 19 Height was expressed as the standard deviation score both for chronologic age and for bone age. Height velocity was expressed as centimeters per year and as the standard deviation score for bone age. The children received 0.3 U / k g body weight per day of biosynthetic G H (Genotropin, from KabiVitrum AB, Stockholm, Sweden), subcutaneously each evening, including the day before each study, for 24 months. Their response was
evaluated before therapy, then after 12 and 24 months of G H administration, and finally 12 months after the discontinuation of the treatment. For at least 3 days before the metabolic tests, the children ate a normocaloric standard diet containing 50% carbohydrates. On day 1 of each study a 12 ml blood sample was drawn after an overnight 12-hour fast for the measurement of concentrations of plasma glucose, free fatty acids, cholesterol, triglycerides, insulin, and insulin growth factor I. An oral glucose tolerance test was performed on day 2 after an overnight fast, with the use of 18% glucose solution at a dose of 45 g m / m 2. Blood samples were collected every 30 minutes for plasma glucose and insulin measurements. Plasma glucose concentration was measured with a glucose oxidase method (Yellow Springs Instruments, Yellow Springs, Ohio). Plasma insulin concentration was measured with double-antibody radioimmunoassay kits purchased from the Commissariat fi l'Energie Atomique, Gif sur Yvette, France. 23 Insulin measurements were performed in triplicate. Plasma I G F - I concentrations were measured with a previously reported radioimmunoassay. 24, 25 Statistical methods. The results are expressed as the mean _+ SE. Comparisons before, during, and after treatment were made with a two-tailed paired t test.
Volume t 19 Number 1, Part 1
Normal adolescent height after growth hormone treatment
31
Table II. Predicted adult height at the different stages of the study During GH treatment Before GH treatment Patient No. Boys 1 2 3 4 5 6 7 Mean _+ SEM Girls 8 9 10 Mean • SEM
At 12mo
At 42 mo after GH treatment
At 24 mo
B-P (cm)
p" (cm)
TW2 (cm)
B-P (cm)
TW2 (cm)
B-P (cm)
TW2 (cm)
B-P (cm)
TW2 (cm)
158.3 145.6 162.0 154.6 154.3 162.5 156.5 156.2 • 2
+1.8 -3.2 -0.8 -4.2 -3.5 -2.5 -1.2 -1.95
161.9 142.8 159.6 156.4 153.5 159.9 156.9 155.9 • 2.2
168.3 154.0 172.3 158.0 161.6 173.3 166.3 164.8 _+ 2.5
170.5 152.9 178.9 162.6 168.2 168.5 162.6 166.3 _+ 2.8
180.1 156.1 179.6 162.6 168 173.3 165.6 169.3 _+ 3.1
170.5 152. 9 178.9 162.6 168.2 168.5 162.6 166.3 _+ 2.7
177 152.2 !71.4 163.3 i69.1 171.2 165 167.0 _+ 2.6
!67.1 154.5 175 161.5 165.5 168.3 163.7 164.9 _+ 2.3
151 154.5 152 152.5 _+ 0.6
-1.8 -2.5 -2.5 -2.3
150.5 152 151 151.2 • 0.4
159.1 160.7 164.1 161.3 • 0.7
154.7 156.2 153 154.7 • 0.5
160.8 157.8 163.5 160.7 -+ 0.9
156.1 157.8 154.1 156 -+ 0.6
160.5 160.5 166.5 162.5 _+ 1
157 157.9 159 158 _+ 0.3
B-P, Bayley-Pinneaumethod22; TW2, Tanner-Whitehouse method,21 including a factor (p) correcting for the measured mid-parental height. *p =
Mid-parental height - 166.5 cm 3
Table III. Plasma substrates, insulin, and I G F - I in the studied children During GH treatment At 42 mo
At entry Plasma glucose (mg/dl) Fasting Area OGTT Free fatty acids (mmol/L) Claolesterol (mmol/L) Triglycerides (retool/L) Plasma Insulin (~U/rnl) Fasting Peak OGGT Plasma IGF-I (U/ml)
85 265 0.72 4.8 0.60
• 6 _+ 33 _+ 0.10 _+ 0.3 +_ 0.04
85 230 0.55 4.6 0.70
6 _+ 0.7 42 _+ 5 0.9 ___0.2
At 24 mo
_+ 6 _+ 21" _ 0.12 +_ 0.4 _+ 0.04
85_+7 254 _+ 32 0.85 _+ 0.15 4.4 _+ 0.3 0.66 _+ 0.08
13 _+ 1.1" 65 _+ 10" 3.2 • 0.5*
16 _+ 2* 104 _+ 21" 2.9 _+ 0.5*
At 42 mo after end of GH treatment 81 251 0.68 5.0 0.74
_+2 _+ 31 _+ 0.10 _+ 0.3 _+ 0.07
5 _+ 0.6 62 + 8 2.6 _+ 0.6*
*p
