Acta PEdiatr Scand [Suppl] 366: 6-12, 1990
Effects of 3 Years of Growth Hormone Therapy in Short Normal Children P.C. HINDMARSH, P.J. PRINGLE, L. DI SILVIO and C.G.D. BROOK From the Endocrine Unit and the Kabi International Growth Research Centre, 7he Middlesex Hospital, Monimer Street, London. UK
ABSTRACT. Hindmarsh, P.C., Pringle, P.J., Di Silvio, L. and Brook, C.G.D. (Endocrine Unit, The Middlesex Hospital, Mortimer Street, London, UK). Effects of 3 years of growth hormone therapy in short normal children. Acta Paediatr Scand [Suppl] 366: 6, 1990. The effect of 3 years of growth hormone (GH) treatment on growth rate, predicted height, carbohydrate and metabolic status, and thyroid function was studied in 16 short prepubertal children growing with a normal pretreatment growth rate. The height velocity SDS increased from a pretreatment value of -0.44 f 0.33 (mean f SD) to a value of +2.20 f 1.03 during the first year of treatment. It was maintained at a value above zero over the subsequent 2 years. By the end of the third year of treatment, the predicted final height had increased by 6.8 cm in the boys and by 4.2 cm in the girls ( p < 0.001 and p < 0.01, respectively). Increasing the dose of GH on a body surface area basis reduced the deceleration of growth observed during the second year of treatment, leading to an improvement in height prognosis over that year. Glucose homoeostasis was achieved initially at the expense of an elevation in fasting serum insulin concentration, but this had returned to pretreatment values by the end of the second year of therapy. No effects on thyroid function were observed. Key words: Gmwth hormone, short stature, insulin.
The use of growth hormone (GH) to promote growth in children with severe GH insufficiency has become established clinical practice following the original observations of Raben (1). The original studies demonstrating the efficacy of GH were carried out in children who had an unambiguous diagnosis of severe GH insufficiency. Since then, several groups have reported the short-term effects of exogenous GH treatment on the growth rate of children with short stature, abnormal growth velocity and ‘normal’ GH responses to conventional provocative stimuli. Very few of these studies have included short children growing at a normal rate; the majority have included children with delayed puberty or other diagnoses, such as intrauterine growth retardation. In general, no explanation has been sought to explain the poor growth rates, and the definition of normality has depended solely on the results of GH secretion test(s) (2-9). The limitations of defining normality by pharmacological testing have been shown by Bercu et al. (10, 11). Several groups have now confirmed that a relationship exists between the height velocity of short prepubertal children and the amount of GH secreted (12-14). This observation led to the hypothesis that short children growing with a normal velocity would respond to exogenous GH therapy if sufficient GH were given, but that their response would be less than that of children who secreted less GH and grew poorly as a consequence. The results of treating 16 such children with recombinant human GH (rhGH) for 1 year have been reported previously (15). The data demonstrated an increase in height velocity standard deviation score (SDS) from a pretreatment value of -0.44 k 0.33 (mean f SD) to a value of +2.20 k 1.03. This increase in growth rate was accompanied by a significant increase in height SDS for bone age. The only metabolic consequence of the treatment was an increase in the fasting serum insulin concentration. This report describes the longer term follow-up of the same children in terms of growth rate, predicted height and metabolic parameters.
GH therapy in short children
Acta Paediatr Scand [Suppl] 366
SUBJECTS AND METHODS The clinical details of the 16 children (10 boys, 6 girls) and the endocrine criteria of normality have been reported previously (IS). Growrh ussessmenr. Standard anthropometry was performed at 3-monthly intervals (16). Bone age was assessed at yearly intervals by the same observer, who was unaware of the child's status, using the method of Tanner and Whitehouse (17). Height predictions were made using TW-2height prediction equations (18). At each clinic visit, pubeny was staged according to the method of Tanner (19). All anthropometric measures were recorded as SDS using the British standards (20). Treurmenr regimens. The children received somatrem (Somatonorml) as a subcutaneous injection of 2 IU on 6 out of 7 nights for the first year of therapy. All doses were recorded in terms of IU/m2 body surface area/week. At the end of the first year of treatment, the children were randomly allocated to one of two groups: group A continued to receive 2 IU of rhGH for 6 out of 7 nights; group B received a revised dose of rhGH based on body surface area, equivalent to the dose given at the commencement of treatment at the beginning of the first year in terms of body surface area. The dose was continuously revised throughout the second year. After completion of the second year of treatment, all the children received a standardized dose of rhGH 20 IUlm2/week. Curbohydrure and merabolic studies. Glucose regulation was assessed by an oral glucose tolerance test (1.75 g glucose/kg body weight) at 6-monthly intervals during the first treatment year and thereafter at yearly intervals. During the course of the test, serum insulin concentrations were measured at 30-minute intervals. To complement these studies. glycosylated haemoglobin concentrations were measured. Fasting triglyceride and cholesterol concentrations were measured at 6-monthly intervals throughout the 3-year treatment period. 7hyroidfuncrion. Serum thyroxine concentrations were measured every 6 months during the first year of treatment and thereafter at the end of each treatment year. Biochemical evuluurion. A blwd sample was taken at each clinic visit for the measurement of plasma urea and electrolyte levels, serum calcium concentrations, liver function tests, haemoglobin concentrations and white cell and platelet counts. Srarisrical unulysis. The changes in growth parameters were compared to pretreatment values using two-way analysis of variance (ANOVA) with a repeated-measures design. The Newman-Keuls test was used to assess the significance of differences between the means. A similar procedure was adopted to compare the carbohydrate and metabolic parameters, thyroid function and biochemical test results over the 3 years. Student's r-test was used to assess the significance of differences between the two treatment regimens following the randomization procedure at the end of the first year of treatment.
RESULTS Growth variables. A significant acceleration in height velocity was observed over the first year of rhGH treatment (Fig. 1) and this was maintained over the subsequent 2 years (two-way ANOVA, F = 19.9, p c 0.001). By the end of the third year of treatment, two
Treatment period (years) Fig. I. The effect of 3 years of treatment with hGH on height velocity SDS in 16 short normal children. Open circles: group A, dose not adjusted for body size during the second year. Closed circles: group B, dose adjusted for body size during the second year. In year 3 both groups received a slightly higher dose. Results are expressed as mean f SEM. lSomatonorm, trademark KabiVitrum AB, Sweden, for somatrem.
7
8
P.C. Hindmarsh et al.
Acta
Paedialr Scand [Suppl] 366
Table I . Characteristics of the two groups of children at the beginning of the second year of GH therapy and the effect of treatment on height velocity SDS during the 3-year treatment period. Group A received a dose of 2 IU on 6 out of 7 nights throughout the first 2 years of treatment, while group B received a revised dose during the second year based on body surface area. During the third year of treatment, both groups received a standardized dose of 20 IU/m2/week. Results are expressed as mean f SD. Treatment group
Age (years) Dose rhGH Wlrn*/week) Pretreatment height velocity SDS Year 1 height velocity SDS Year 2 height velocity SDS Year 3 height velocity SDS
P
Group A (n = 8)
Group B (n = 8)
8.3 f 1.7 16.6 5 2.8 -0.48 f 0.40
10.3 f 1.5 14.6 f 2 . 5 -0.42 f 0 . 2 6
ns ns ns
* 1.11
1.85 f 0.92
ns
0.52 f 1.16
0.95 f 1.31
ns
1.67 f 2.11
2.25 f 2.05
ns
2.48
ns = not significant
of the girls had achieved breast stage I1 development and were undergoing their pubertal growth spurt, while one of the boys had reached a testicular volume of 10 ml and his pubertal growth spurt was expected. The rise in height velocity over the third year of treatment was significant even when these patients were excluded ( p = 0.05). On random allocation to the two different treatment regimens at the end of the first year of therapy, there was no significant difference between the two groups of children in terms of age, dose of rhGH received during the first year of therapy. pretreatment height velocity SDS. or height velocity SDS over the first year of treatment (Table 1). The dose of rhGH received by group A decreased by 22 % over the 2 years in terms of body surface area. Height velocity decreased over the second year of treatment and the decrease was less in group B than in group A, but not significantly so. The predicted adult height of the children increased by 6.8 f 3.2 cm (mean f SD) in the boys (Student’s t-test, paired samplep < 0.001) and 4.2 f 2.0 cm in the girls ( p < 0.01) over the 3-year treatment period. The children in group A showed a significant increase in predicted height SDS over the first year of therapy and all subsequent values were higher than pretreatment values (two-way ANOVA. F = 17.7. p < 0.001). During the period of randomization (year 1-year 2). however, no significant change in predicted height SDS was observed (year 1 SDS, - 1.05 -t 0.64. mean f SD; year 2. - 1.13 f 0.52). Changing to a dose of 20 IU/m2/week at the end of the second year of treatment led to a further significant increase in predicted height SDS (year 3, -0.169 f 0.61; Newman-Keuls, year 2 versus year 3, p < 0.05). By contrast, the children in group B displayed a consistent increase in predicted height SDS over the treatment period, with each year’s value being significantly greater than the preceding one as well as significantly greater than the pretreatment value (two-way ANOVA, F = 13.1, p < 0.001; Newman-Keuls for all combinations p < 0.05). The incremental changes in predicted height are shown in Fig. 2, which also includes the change in value for the 10 control children from the previous report (15). Both treatment groups wed significantly larger increments in predicted height than the control group ( p < 0.01). The increment in predicted height at the end of the second year of treatment was significantly less in group A than in group B (Student’s t-test, p < 0.05).
GH therapy in short children
Acta Paediatr Scand [Suppl] 366
0Group A 0Group 6
T
Control 9 rou P
T
0 i
Year 1
-
Year 2
I
u
u
Year 3
Year 3
Fig. 2. The change in predicted final height SDS over 3 years of treameni wilh hGH in the 16 short normal children compared to a control group ( n = 10). Group A. dose not adjusted for body size during the second year. Group B. dose adjusted for body size during the second year. In year 3 both groups received a slightly higher dose. Results are expressed as mean f SEM.
Glucosc tolerance and lipid studies. Fasting glucose concentrations and the incremental areas under the glucose and insulin curves were unchanged during the course of the study. Over the 3-year period, the glycosylated haemoglobin concentration remained unchanged (pretreatment,6.8 f 0.8%;year1,6.7 f 0.7;year2,6.4 f 0.4;year3,6.4 f 0.5;normal range 5 4 % ) . Glucose homoeostasis was initially preserved at the expense of an increase in the fasting serum insulin concentration, from a pretreatment value of 5.4 f 3.5 mU/I to a value of 15.3 f 8.7 mU/l at the end of the first year. The fasting serum insulin concentration decreased during the second year of treatment to a value of 8.7 f 5.6 mU/I, which was not significantly different from the pretreatment value (two-way ANOVA, F = 13.1, p < 0.001; Newman-Keuls: pretreatment versus year 1, p < 0.05; year I versus year 2, p < 0.05; pretreatment versus year 2 , not significant). Interpretation of insulin data at the end of the third year of treatment was complicated by the onset of puberty in three of the children. There was no significant difference between the fasting serum insulin concentrations of group A and group B at the end of either the first or second year of treatment. Fasting serum cholesterol concentrations increased significantly during the first year of treatment from 4.0 f 0.5 mmolll to 4.5 f 0.8 mmol/l. and remained elevated compared to 0.7 mmol/l; year 3, pretreatment values throughout the subsequent 2 years (year 2. 4.5 0.7 mmol/l; two-way ANOVA, F = 10.9, p < 0.001). Although this rise was 4.4 statistically significant, the serum cholesterol concentrations remained within the normal range (2.3-6.9 mmol/l) and the rise may be explained by the increase in fasting serum cholesterol concentration that normally occurs during the second half of the first decade of life. Fasting serum triglyceride concentrations were unchanged during the course of therapy. Thyroid state. Serum thyroxine concentrations did not change during the treatment period. Biochemical studies. There were no changes in any of the measured biochemical or haematological parameters over the 3-year period.
*
*
DISCUSSION The results of this study demonstratethat children who are short but growing along or parallel to the 3rd height centile will respond to rhGH treatment with an increase in growth rate. This observation confirms the hypothesis that shori children growing with a normal velocity will
9
10
P.C. Hindmarsh et al.
Acta Paediatr
Scand
respond to exogenous GH therapy if sufficient GH is given (12). The increase in growth rate has been translated into an increase in final predicted adult height of 6.8 cm in boys and 4.2 cm in girls over the 3-year period of this study. At the end of the first year of treatment, the children were randomly allocated to either an unchanged treatment regimen, so that the dose of rhGH gradually decreased in terms of body surface area, or to a revised dose regimen based on the initial dose given to the child in terms of body surface area. Although the results of this random allocation did not lead to a statistically significant difference in growth rate between the two groups, the improvement in predicted height coupled with the reduced decrement in growth rate over this period in the dose-adjusted group confirms that GH dose may prove to be. an important factor in determining height prognosis in such children. At the end of the second year of treatment, all children were transferred to a regimen of 20 IU rhGH/mZ/week, which led to a further increase in growth rate over the third year of treatment. The change in dose at this point led to a significant increase in predicted height SDS for the children in group A and group B, although the magnitude was greater in group A. The increase in growth rate in the third year of the study compared to the second year was not simply due to the fact that three of the children had entered puberty, because the result was significant even if these individuals were excluded. Similar observations have been made by Gertner et al. (21). These observations suggest that optimal therapy with rhGH would require at least a yearly adjustment of the dose regimen on the basis of body size. The results could explain the rather poor responses to GH treatment observed in the past using fixed-dose regimens of GH in children with GH deficiency (22-24). For example, over the first 2 years of this study, the children in group A effectively experienced a 22% reduction in dose in terms of body surface area. These observations also complement physiological studies of 24-hour GH secretion, where an increase in secretion between the ages of 5 and 10 years has been demonstrated (14, 25). It is still too early to predict the final outcome in these children and to explain fully the importance of dose and growth response in such individuals (15, 26); further studies need to be performed to determine optimal dose regimens for promoting growth. GH secretion increases two-fold to threefold around the time of puberty (25, 27), and its contribution to the pubertal growth spurt is wellestablished (28-31). Optimal therapy in short normal children and children with GH insufficiency must take account of these observations, and the dose of GH used in puberty may therefore become an important factor in determining outcome. The pubertal growth spurt contributes approximately 20 cm to final height, some 50% of which is due to GH (28, 29). Thus, inadequate therapy at this stage could reverse the beneficial effect of GH therapy on height gain in the prepubertal years. If GH is to be used in higher doses than have previously been employed, the interaction of GH with other endocrine organ systems must be considered. Insulin-like growth factors are important in modulating gonadal activity (32) and rhGH has been shown to enhance the effects of gonadotrophin therapy on the ovaries in humans and primates (33, 34). It could be postulated, therefore, that GH treatment might accelerate the pubertal process without altering the timing of the onset of puberty, which is dependent on nocturnal gonadotrophin secretion. Data are available which suggest that GH treatment can shorten the duration of puberty by 6 months (35). For these reasons, GH doses should not be increased indiscriminately, and the long-term effects of GH treatment in short normal children need to be considered carefully. Concern has been expressed that short normal children treated with GH may develop metabolic sequelae similar to those seen in acromegaly. However, none of the children in this study have developed glucose intolerance as measured by oral glucose loading tests and determination of glycosylated haemoglobin concentrations. Glucose tolerance was obtained
[SuppI]
366
Acta Paediatr Scand [Suppl] 366
GH theraw in short children
at the expense of a rise in fasting insulin concentrations, but the insulin response to glucose loading was unchanged. GH is known to produce insulin resistance (36) and directly affects glucose oxidase lipid synthesis (37).However, there is accumulating evidence to suggest that synergism between GH and insulin is essential for the generation of insulin-like growth factor I and for the promotion of growth (38, 39). Both the growth velocity and fasting serum insulin concentration data suggest that tolerance to the effects of GH occurs. This is not simply due to an effective reduction in dose since the dose was maintained at a constant level in terms of body surface area in half of the children. There would appear to be down-regulation of the effects of GH on carbohydrate and lipid metabolism at the tissue level and, if these effects also occurred in cartilage, this would explain the reduction in growth velocity in the succeeding years of treatment. These observations might explain to some extent the growth deceleration seen in succeeding years of therapy in hypopituitary children (22-24). To circumvent this problem, increases in GH doses would be advisable over the years. This is supported by the observation of an increased growth rate over the third year of therapy once the GH dose had been adjusted to 20 IU/m2/week. This study has demonstrated that rhGH treatment in children growing along or parallel to the third height centile induced a sustained increase in growth velocity which led to an increase in height prediction over 3 years. Apart from an increase in fasting serum insulin concentrations, there was no untoward effect of treatment. The data suggest that optimal growth can be achieved in these children by calculating the GH dose on a body surface area basis and by using doses in the region of at least 20 IU/m2/week. ACKNOWLEDGEMENT The author would like to thank Kabi Peptide Hormones, Stockholm, for their support of this study.
REFERENCES 1. Raben MS. Treatment of a pituitary dwarf with human growth hormone. J Clin Endocrinol Metab 1958; 18: 901-3. 2. Frazer T, Gavin JR. Daughaday WH, Hillrnan RE, Weldon VV. Growth hormonedependent growth failure. J Pediatr 1982; 101: 12-15. 3. Rudman D, Kutner MH, Blackston RD, Cushrnan RA, Bain RP, Patterson JH. Children with normalvariant short stature: treatment with human growth hormone for six months. N Engl J Med 1981; 305: 123-31. 4. Plotnick LP, Van Meter QL, Kowarski AA. Human growth hormone treatment of children with growth
failure and normal growth hormone levels by immunoassay: lack of correlation with somatomedin generation. Pediatrics 1983; 71: 324-7. 5 . Van Vliet G, Styne DM, Kaplan SL, Grumbach MM. Growth hormone treatment for short stature. N Engl J Med 1983; 309: 1016-22. 6. Gertner JM, Gene1 M, Gianfredi SP er al. Prospective clinical trial of human growth hormone in short children without growth hormone deficiency. J Pediatr 1984; 104: 172-6. 7. Carrascosa A, Vicens-Calvet E, Audi L, Gusinye M, Albisu M, Potau N. Chronic growth retardation with normal growth hormone response to provocative stimuli and low somatomedin activity: long-term therapy with human growth hormone. Acta Paediatr Scand 1987; 76: 489-94. 8. Albertsson-Wikland K, Hall K. Growth hormone treatment in short children: relationship between growth and serum insulin-like growth factor I and I1 levels. J Clin Endocrinol Metab 1987; 65: 671-8. 9. Wit JM, Rietveld DHF, Drop SLS er al. A controlled trial of methionyl growth hormone therapy in prepubertal children with short stature, subnormal growth rate and normal growth hormone response to secretatogogues. Acta Paediatr Scand 1989; 78: 426-35. 10. Spiliotis BE, August GP, Hung W, Sonis W, Mendelson W, Bercu BB. Growth hormone neurosecretory dysfunction. JAMA 1984: 251: 2223-30. I 1 . Bercu BB, Shulman D, Root AW, Spiliotis BE. Growth hormone (GH) provocative testing frequently does not reflect endogenous GH secretion. J Clin Endocrinol Metab 1986; 63; 709-16. 12. Hindmarsh P, Smith PJ, Brook CGD, Matthews DR. The relationship between height velocity and growth hormone secretion in short prepubertal children. Clin Endocrinol 1987; 27: 581-91. 13. Spadoni GL, Cianfarani S , Bernardini S er al. Twelve-hour spontaneous nocturnal growth hormone secretion in growth retarded patients. Clin Pediatr 1988; 27: 473-8.
11
12
P.C. Hinclmclrsh et al.
Acta Paediatr Scand [Suppl] 366
14. Albertsson-Wikland K, Rosberg S. Analyses of 24-hour growth hormone profiles in children: relation to growth. J Clin Endocrinol Metab 1988; 67: 493-500. 15. Hindmarsh PC, Brook CGD. Effect of growth hormone on short normal children. Br Med J 1987; 295: 573-7. 16. Brook CGD. Growth assessment in childhood and adolescence. Oxford: Blackwell Scientific, 1982: 164. 17. Tanner JM, Whitehouse RH. Cameron N, Marshall WA, Healy MJR, Goldstein H. Assessment of skeletal maturity and prediction of adult height (TW2 method). London: Academic Press, 1983: 106. 18. Tanner JM, Landt KW, Cameron N, Carter BS, Patel J. Prediction of adult height from height and bone age in childhood. Arch Dis Child 1983; 58: 767-76. 19. Tanner JM. Growth at adolescence. Oxford: Blackwell Scientific, 1962: 325. 20. Tanner JM, Whitehouse RH, Takaishi M. Standards from birth to maturity for height, weight, height velocity, and weight velocity: British children, 1%5. Part II. Arch Dis Child 1966; 41: 613-35. 21. Gertner JM. Tamborlane WV, Gianfredi SP. Gene1 M. Renewed catch-up growth with increased replacement doses of human growth hormone. J Pediatr 1987; 110: 425-8. 22. Bums EC, Tanner JM, Preece MA. Cameron N. Final height and pubertal development in 55 children with iiliopathic growth hormone deficiency, treated for between 2 and I5 years with human growth hormone. Eur J Pediatr 1981; 137: 155-64. 23. Joss E. Zuppinger K, Schwarz HP, Roten H. Final height of patients with pituitary growth failure and changes in growth variables after long term hormonal therapy. Pediatr Res 1983; 17: 676-9. 24. Bundak R, Hindmarsh PC, Smith PJ, Brook CGD. Long-term auxologic effects of human growth hormone. J Pediatr 1988; 112: 875-9. 25. Hindmarsh PC. Matthews DR, Brook CGD. Growth hormone secretion in children determined by time .series analysis. Clin Endocrinol 1988; 29: 35-44. 26. Smith PJ, Hindmarsh PC, Brook CGD. Contribution of dose and frequency of administration to the therapeutic effect of growth hormone. Arch Dis Child 1988; 63: 491-4. 27. Miller JD, Tannenbaum GS, Colle F, Guyda HJ. Daytime pulsatile growth hormone secretion during childhood and adolescence. J Clin Endocrinol Metab 1982; 55: 989-94. 18. Aynsley-Green A. Zachmann M, Prdder A. Interrelation of the therapeutic effects of growth hormone and testosterone on growth in hypopituitarism. J Pediatr 1976; 89: 992-9. 29. Tanner JM, Whitehouse RH. Hughes PCR, Carter BS. Relative importance of growth-hormone and sex steroids for the growth at puberty of trunk length, limb length, and muscle width in growth hormonedeficient children. J Pediatr 1976; 89: 1000-8. 30. Mauras N. Blizzard RM, Link K, Johnson ML. Rogol AD, Veldhuis ID. Augmentation of growth hormone secretion during puberty: evidence for a pulse amplitude-modulated phenomenon. J Clin Endocrinol Metab 1987; 64: 596-6431. 31. Stanhope R, Pringle PJ, Brook CGD. The mechanism of the adolescent growth spurt induced by low dose pulsatile GnRH treatment. Clin Endocrinol 1988; 28: 83-91. 32. Adashi EY. Resnick CE, Hernandez ER ef al. Insulin-like growth factor I as an amplifier of folliclestimulating hormone action: studies on mechanism(s) and site(s) of action in cultured rat granulosa cells. Endocrinology 1988; 122: 1583-91. 33. Homburg R. Eshel A, Abdalla HI. Jacobs HS. Growth hormone facilitates ovulation induction by gonadotrophins. Clin Endocrinol 1988; 29: 113-7. 34. Wilson ME, Gordon TP, Rudman CG. Tanner JM. Effects of growth hormone on the tempo of sexual maturation in female rhesus monkeys. J Clin Endocrinol Metab 1989: 68: 29-38. 35. Darendeliler F, Hindmarsh PC, Preece MA, Cox L. Brook CGD. Growth hormone increases rate o f pubertal maturation. Acta Endocrinol 1990; in press. 36. Rosenfeld RG, Wilson DM. Dollar LA, Bennett A, Hintz RL. Both human pituitary growth hormone and recombinant DNAderived human growth hormone cause insulin resistance at a postreceptor site. J Clin Endocrinol Metab 1982; 54: 1033-8. 37. Foster CM, Hale PM. Jing H-W, Schwanz J. Effects of human growth hormone on insulin-stimulated glucose metabolism in 3T3-F442A adipocytes. Endocrinology 1988; 123: 1082-8. 38. Binoux M, Lassarre C. Hardouin N. Somatomedin production by rat liver in organ culture. 111. Studies on the release of insulin-like growth factor and its carrier protein measured by radioligand assays. Acta Endocrinol 1982; 99: 422-30. 39. Scheiwiller E, Guler H-P. Menyweather J et ul. Growth restoration of insulin-deficient diabetic rats by recombinant human insulin-like growth factor 1. Nature 1986; 323: 169-71. (C.G.D.B.) Endocrine Unit The Middlesex Hospital Mortimer Street London WIN 8AA
UK
Effects of 3 Years of Growth Hormone Therapy in Short Normal Children P.C. HINDMARSH, P.J. PRINGLE, L. DI SILVIO and C.G.D. BROOK From the Endocrine Unit and the Kabi International Growth Research Centre, 7he Middlesex Hospital, Monimer Street, London. UK
ABSTRACT. Hindmarsh, P.C., Pringle, P.J., Di Silvio, L. and Brook, C.G.D. (Endocrine Unit, The Middlesex Hospital, Mortimer Street, London, UK). Effects of 3 years of growth hormone therapy in short normal children. Acta Paediatr Scand [Suppl] 366: 6, 1990. The effect of 3 years of growth hormone (GH) treatment on growth rate, predicted height, carbohydrate and metabolic status, and thyroid function was studied in 16 short prepubertal children growing with a normal pretreatment growth rate. The height velocity SDS increased from a pretreatment value of -0.44 f 0.33 (mean f SD) to a value of +2.20 f 1.03 during the first year of treatment. It was maintained at a value above zero over the subsequent 2 years. By the end of the third year of treatment, the predicted final height had increased by 6.8 cm in the boys and by 4.2 cm in the girls ( p < 0.001 and p < 0.01, respectively). Increasing the dose of GH on a body surface area basis reduced the deceleration of growth observed during the second year of treatment, leading to an improvement in height prognosis over that year. Glucose homoeostasis was achieved initially at the expense of an elevation in fasting serum insulin concentration, but this had returned to pretreatment values by the end of the second year of therapy. No effects on thyroid function were observed. Key words: Gmwth hormone, short stature, insulin.
The use of growth hormone (GH) to promote growth in children with severe GH insufficiency has become established clinical practice following the original observations of Raben (1). The original studies demonstrating the efficacy of GH were carried out in children who had an unambiguous diagnosis of severe GH insufficiency. Since then, several groups have reported the short-term effects of exogenous GH treatment on the growth rate of children with short stature, abnormal growth velocity and ‘normal’ GH responses to conventional provocative stimuli. Very few of these studies have included short children growing at a normal rate; the majority have included children with delayed puberty or other diagnoses, such as intrauterine growth retardation. In general, no explanation has been sought to explain the poor growth rates, and the definition of normality has depended solely on the results of GH secretion test(s) (2-9). The limitations of defining normality by pharmacological testing have been shown by Bercu et al. (10, 11). Several groups have now confirmed that a relationship exists between the height velocity of short prepubertal children and the amount of GH secreted (12-14). This observation led to the hypothesis that short children growing with a normal velocity would respond to exogenous GH therapy if sufficient GH were given, but that their response would be less than that of children who secreted less GH and grew poorly as a consequence. The results of treating 16 such children with recombinant human GH (rhGH) for 1 year have been reported previously (15). The data demonstrated an increase in height velocity standard deviation score (SDS) from a pretreatment value of -0.44 k 0.33 (mean f SD) to a value of +2.20 k 1.03. This increase in growth rate was accompanied by a significant increase in height SDS for bone age. The only metabolic consequence of the treatment was an increase in the fasting serum insulin concentration. This report describes the longer term follow-up of the same children in terms of growth rate, predicted height and metabolic parameters.
GH therapy in short children
Acta Paediatr Scand [Suppl] 366
SUBJECTS AND METHODS The clinical details of the 16 children (10 boys, 6 girls) and the endocrine criteria of normality have been reported previously (IS). Growrh ussessmenr. Standard anthropometry was performed at 3-monthly intervals (16). Bone age was assessed at yearly intervals by the same observer, who was unaware of the child's status, using the method of Tanner and Whitehouse (17). Height predictions were made using TW-2height prediction equations (18). At each clinic visit, pubeny was staged according to the method of Tanner (19). All anthropometric measures were recorded as SDS using the British standards (20). Treurmenr regimens. The children received somatrem (Somatonorml) as a subcutaneous injection of 2 IU on 6 out of 7 nights for the first year of therapy. All doses were recorded in terms of IU/m2 body surface area/week. At the end of the first year of treatment, the children were randomly allocated to one of two groups: group A continued to receive 2 IU of rhGH for 6 out of 7 nights; group B received a revised dose of rhGH based on body surface area, equivalent to the dose given at the commencement of treatment at the beginning of the first year in terms of body surface area. The dose was continuously revised throughout the second year. After completion of the second year of treatment, all the children received a standardized dose of rhGH 20 IUlm2/week. Curbohydrure and merabolic studies. Glucose regulation was assessed by an oral glucose tolerance test (1.75 g glucose/kg body weight) at 6-monthly intervals during the first treatment year and thereafter at yearly intervals. During the course of the test, serum insulin concentrations were measured at 30-minute intervals. To complement these studies. glycosylated haemoglobin concentrations were measured. Fasting triglyceride and cholesterol concentrations were measured at 6-monthly intervals throughout the 3-year treatment period. 7hyroidfuncrion. Serum thyroxine concentrations were measured every 6 months during the first year of treatment and thereafter at the end of each treatment year. Biochemical evuluurion. A blwd sample was taken at each clinic visit for the measurement of plasma urea and electrolyte levels, serum calcium concentrations, liver function tests, haemoglobin concentrations and white cell and platelet counts. Srarisrical unulysis. The changes in growth parameters were compared to pretreatment values using two-way analysis of variance (ANOVA) with a repeated-measures design. The Newman-Keuls test was used to assess the significance of differences between the means. A similar procedure was adopted to compare the carbohydrate and metabolic parameters, thyroid function and biochemical test results over the 3 years. Student's r-test was used to assess the significance of differences between the two treatment regimens following the randomization procedure at the end of the first year of treatment.
RESULTS Growth variables. A significant acceleration in height velocity was observed over the first year of rhGH treatment (Fig. 1) and this was maintained over the subsequent 2 years (two-way ANOVA, F = 19.9, p c 0.001). By the end of the third year of treatment, two
Treatment period (years) Fig. I. The effect of 3 years of treatment with hGH on height velocity SDS in 16 short normal children. Open circles: group A, dose not adjusted for body size during the second year. Closed circles: group B, dose adjusted for body size during the second year. In year 3 both groups received a slightly higher dose. Results are expressed as mean f SEM. lSomatonorm, trademark KabiVitrum AB, Sweden, for somatrem.
7
8
P.C. Hindmarsh et al.
Acta
Paedialr Scand [Suppl] 366
Table I . Characteristics of the two groups of children at the beginning of the second year of GH therapy and the effect of treatment on height velocity SDS during the 3-year treatment period. Group A received a dose of 2 IU on 6 out of 7 nights throughout the first 2 years of treatment, while group B received a revised dose during the second year based on body surface area. During the third year of treatment, both groups received a standardized dose of 20 IU/m2/week. Results are expressed as mean f SD. Treatment group
Age (years) Dose rhGH Wlrn*/week) Pretreatment height velocity SDS Year 1 height velocity SDS Year 2 height velocity SDS Year 3 height velocity SDS
P
Group A (n = 8)
Group B (n = 8)
8.3 f 1.7 16.6 5 2.8 -0.48 f 0.40
10.3 f 1.5 14.6 f 2 . 5 -0.42 f 0 . 2 6
ns ns ns
* 1.11
1.85 f 0.92
ns
0.52 f 1.16
0.95 f 1.31
ns
1.67 f 2.11
2.25 f 2.05
ns
2.48
ns = not significant
of the girls had achieved breast stage I1 development and were undergoing their pubertal growth spurt, while one of the boys had reached a testicular volume of 10 ml and his pubertal growth spurt was expected. The rise in height velocity over the third year of treatment was significant even when these patients were excluded ( p = 0.05). On random allocation to the two different treatment regimens at the end of the first year of therapy, there was no significant difference between the two groups of children in terms of age, dose of rhGH received during the first year of therapy. pretreatment height velocity SDS. or height velocity SDS over the first year of treatment (Table 1). The dose of rhGH received by group A decreased by 22 % over the 2 years in terms of body surface area. Height velocity decreased over the second year of treatment and the decrease was less in group B than in group A, but not significantly so. The predicted adult height of the children increased by 6.8 f 3.2 cm (mean f SD) in the boys (Student’s t-test, paired samplep < 0.001) and 4.2 f 2.0 cm in the girls ( p < 0.01) over the 3-year treatment period. The children in group A showed a significant increase in predicted height SDS over the first year of therapy and all subsequent values were higher than pretreatment values (two-way ANOVA. F = 17.7. p < 0.001). During the period of randomization (year 1-year 2). however, no significant change in predicted height SDS was observed (year 1 SDS, - 1.05 -t 0.64. mean f SD; year 2. - 1.13 f 0.52). Changing to a dose of 20 IU/m2/week at the end of the second year of treatment led to a further significant increase in predicted height SDS (year 3, -0.169 f 0.61; Newman-Keuls, year 2 versus year 3, p < 0.05). By contrast, the children in group B displayed a consistent increase in predicted height SDS over the treatment period, with each year’s value being significantly greater than the preceding one as well as significantly greater than the pretreatment value (two-way ANOVA, F = 13.1, p < 0.001; Newman-Keuls for all combinations p < 0.05). The incremental changes in predicted height are shown in Fig. 2, which also includes the change in value for the 10 control children from the previous report (15). Both treatment groups wed significantly larger increments in predicted height than the control group ( p < 0.01). The increment in predicted height at the end of the second year of treatment was significantly less in group A than in group B (Student’s t-test, p < 0.05).
GH therapy in short children
Acta Paediatr Scand [Suppl] 366
0Group A 0Group 6
T
Control 9 rou P
T
0 i
Year 1
-
Year 2
I
u
u
Year 3
Year 3
Fig. 2. The change in predicted final height SDS over 3 years of treameni wilh hGH in the 16 short normal children compared to a control group ( n = 10). Group A. dose not adjusted for body size during the second year. Group B. dose adjusted for body size during the second year. In year 3 both groups received a slightly higher dose. Results are expressed as mean f SEM.
Glucosc tolerance and lipid studies. Fasting glucose concentrations and the incremental areas under the glucose and insulin curves were unchanged during the course of the study. Over the 3-year period, the glycosylated haemoglobin concentration remained unchanged (pretreatment,6.8 f 0.8%;year1,6.7 f 0.7;year2,6.4 f 0.4;year3,6.4 f 0.5;normal range 5 4 % ) . Glucose homoeostasis was initially preserved at the expense of an increase in the fasting serum insulin concentration, from a pretreatment value of 5.4 f 3.5 mU/I to a value of 15.3 f 8.7 mU/l at the end of the first year. The fasting serum insulin concentration decreased during the second year of treatment to a value of 8.7 f 5.6 mU/I, which was not significantly different from the pretreatment value (two-way ANOVA, F = 13.1, p < 0.001; Newman-Keuls: pretreatment versus year 1, p < 0.05; year I versus year 2, p < 0.05; pretreatment versus year 2 , not significant). Interpretation of insulin data at the end of the third year of treatment was complicated by the onset of puberty in three of the children. There was no significant difference between the fasting serum insulin concentrations of group A and group B at the end of either the first or second year of treatment. Fasting serum cholesterol concentrations increased significantly during the first year of treatment from 4.0 f 0.5 mmolll to 4.5 f 0.8 mmol/l. and remained elevated compared to 0.7 mmol/l; year 3, pretreatment values throughout the subsequent 2 years (year 2. 4.5 0.7 mmol/l; two-way ANOVA, F = 10.9, p < 0.001). Although this rise was 4.4 statistically significant, the serum cholesterol concentrations remained within the normal range (2.3-6.9 mmol/l) and the rise may be explained by the increase in fasting serum cholesterol concentration that normally occurs during the second half of the first decade of life. Fasting serum triglyceride concentrations were unchanged during the course of therapy. Thyroid state. Serum thyroxine concentrations did not change during the treatment period. Biochemical studies. There were no changes in any of the measured biochemical or haematological parameters over the 3-year period.
*
*
DISCUSSION The results of this study demonstratethat children who are short but growing along or parallel to the 3rd height centile will respond to rhGH treatment with an increase in growth rate. This observation confirms the hypothesis that shori children growing with a normal velocity will
9
10
P.C. Hindmarsh et al.
Acta Paediatr
Scand
respond to exogenous GH therapy if sufficient GH is given (12). The increase in growth rate has been translated into an increase in final predicted adult height of 6.8 cm in boys and 4.2 cm in girls over the 3-year period of this study. At the end of the first year of treatment, the children were randomly allocated to either an unchanged treatment regimen, so that the dose of rhGH gradually decreased in terms of body surface area, or to a revised dose regimen based on the initial dose given to the child in terms of body surface area. Although the results of this random allocation did not lead to a statistically significant difference in growth rate between the two groups, the improvement in predicted height coupled with the reduced decrement in growth rate over this period in the dose-adjusted group confirms that GH dose may prove to be. an important factor in determining height prognosis in such children. At the end of the second year of treatment, all children were transferred to a regimen of 20 IU rhGH/mZ/week, which led to a further increase in growth rate over the third year of treatment. The change in dose at this point led to a significant increase in predicted height SDS for the children in group A and group B, although the magnitude was greater in group A. The increase in growth rate in the third year of the study compared to the second year was not simply due to the fact that three of the children had entered puberty, because the result was significant even if these individuals were excluded. Similar observations have been made by Gertner et al. (21). These observations suggest that optimal therapy with rhGH would require at least a yearly adjustment of the dose regimen on the basis of body size. The results could explain the rather poor responses to GH treatment observed in the past using fixed-dose regimens of GH in children with GH deficiency (22-24). For example, over the first 2 years of this study, the children in group A effectively experienced a 22% reduction in dose in terms of body surface area. These observations also complement physiological studies of 24-hour GH secretion, where an increase in secretion between the ages of 5 and 10 years has been demonstrated (14, 25). It is still too early to predict the final outcome in these children and to explain fully the importance of dose and growth response in such individuals (15, 26); further studies need to be performed to determine optimal dose regimens for promoting growth. GH secretion increases two-fold to threefold around the time of puberty (25, 27), and its contribution to the pubertal growth spurt is wellestablished (28-31). Optimal therapy in short normal children and children with GH insufficiency must take account of these observations, and the dose of GH used in puberty may therefore become an important factor in determining outcome. The pubertal growth spurt contributes approximately 20 cm to final height, some 50% of which is due to GH (28, 29). Thus, inadequate therapy at this stage could reverse the beneficial effect of GH therapy on height gain in the prepubertal years. If GH is to be used in higher doses than have previously been employed, the interaction of GH with other endocrine organ systems must be considered. Insulin-like growth factors are important in modulating gonadal activity (32) and rhGH has been shown to enhance the effects of gonadotrophin therapy on the ovaries in humans and primates (33, 34). It could be postulated, therefore, that GH treatment might accelerate the pubertal process without altering the timing of the onset of puberty, which is dependent on nocturnal gonadotrophin secretion. Data are available which suggest that GH treatment can shorten the duration of puberty by 6 months (35). For these reasons, GH doses should not be increased indiscriminately, and the long-term effects of GH treatment in short normal children need to be considered carefully. Concern has been expressed that short normal children treated with GH may develop metabolic sequelae similar to those seen in acromegaly. However, none of the children in this study have developed glucose intolerance as measured by oral glucose loading tests and determination of glycosylated haemoglobin concentrations. Glucose tolerance was obtained
[SuppI]
366
Acta Paediatr Scand [Suppl] 366
GH theraw in short children
at the expense of a rise in fasting insulin concentrations, but the insulin response to glucose loading was unchanged. GH is known to produce insulin resistance (36) and directly affects glucose oxidase lipid synthesis (37).However, there is accumulating evidence to suggest that synergism between GH and insulin is essential for the generation of insulin-like growth factor I and for the promotion of growth (38, 39). Both the growth velocity and fasting serum insulin concentration data suggest that tolerance to the effects of GH occurs. This is not simply due to an effective reduction in dose since the dose was maintained at a constant level in terms of body surface area in half of the children. There would appear to be down-regulation of the effects of GH on carbohydrate and lipid metabolism at the tissue level and, if these effects also occurred in cartilage, this would explain the reduction in growth velocity in the succeeding years of treatment. These observations might explain to some extent the growth deceleration seen in succeeding years of therapy in hypopituitary children (22-24). To circumvent this problem, increases in GH doses would be advisable over the years. This is supported by the observation of an increased growth rate over the third year of therapy once the GH dose had been adjusted to 20 IU/m2/week. This study has demonstrated that rhGH treatment in children growing along or parallel to the third height centile induced a sustained increase in growth velocity which led to an increase in height prediction over 3 years. Apart from an increase in fasting serum insulin concentrations, there was no untoward effect of treatment. The data suggest that optimal growth can be achieved in these children by calculating the GH dose on a body surface area basis and by using doses in the region of at least 20 IU/m2/week. ACKNOWLEDGEMENT The author would like to thank Kabi Peptide Hormones, Stockholm, for their support of this study.
REFERENCES 1. Raben MS. Treatment of a pituitary dwarf with human growth hormone. J Clin Endocrinol Metab 1958; 18: 901-3. 2. Frazer T, Gavin JR. Daughaday WH, Hillrnan RE, Weldon VV. Growth hormonedependent growth failure. J Pediatr 1982; 101: 12-15. 3. Rudman D, Kutner MH, Blackston RD, Cushrnan RA, Bain RP, Patterson JH. Children with normalvariant short stature: treatment with human growth hormone for six months. N Engl J Med 1981; 305: 123-31. 4. Plotnick LP, Van Meter QL, Kowarski AA. Human growth hormone treatment of children with growth
failure and normal growth hormone levels by immunoassay: lack of correlation with somatomedin generation. Pediatrics 1983; 71: 324-7. 5 . Van Vliet G, Styne DM, Kaplan SL, Grumbach MM. Growth hormone treatment for short stature. N Engl J Med 1983; 309: 1016-22. 6. Gertner JM, Gene1 M, Gianfredi SP er al. Prospective clinical trial of human growth hormone in short children without growth hormone deficiency. J Pediatr 1984; 104: 172-6. 7. Carrascosa A, Vicens-Calvet E, Audi L, Gusinye M, Albisu M, Potau N. Chronic growth retardation with normal growth hormone response to provocative stimuli and low somatomedin activity: long-term therapy with human growth hormone. Acta Paediatr Scand 1987; 76: 489-94. 8. Albertsson-Wikland K, Hall K. Growth hormone treatment in short children: relationship between growth and serum insulin-like growth factor I and I1 levels. J Clin Endocrinol Metab 1987; 65: 671-8. 9. Wit JM, Rietveld DHF, Drop SLS er al. A controlled trial of methionyl growth hormone therapy in prepubertal children with short stature, subnormal growth rate and normal growth hormone response to secretatogogues. Acta Paediatr Scand 1989; 78: 426-35. 10. Spiliotis BE, August GP, Hung W, Sonis W, Mendelson W, Bercu BB. Growth hormone neurosecretory dysfunction. JAMA 1984: 251: 2223-30. I 1 . Bercu BB, Shulman D, Root AW, Spiliotis BE. Growth hormone (GH) provocative testing frequently does not reflect endogenous GH secretion. J Clin Endocrinol Metab 1986; 63; 709-16. 12. Hindmarsh P, Smith PJ, Brook CGD, Matthews DR. The relationship between height velocity and growth hormone secretion in short prepubertal children. Clin Endocrinol 1987; 27: 581-91. 13. Spadoni GL, Cianfarani S , Bernardini S er al. Twelve-hour spontaneous nocturnal growth hormone secretion in growth retarded patients. Clin Pediatr 1988; 27: 473-8.
11
12
P.C. Hinclmclrsh et al.
Acta Paediatr Scand [Suppl] 366
14. Albertsson-Wikland K, Rosberg S. Analyses of 24-hour growth hormone profiles in children: relation to growth. J Clin Endocrinol Metab 1988; 67: 493-500. 15. Hindmarsh PC, Brook CGD. Effect of growth hormone on short normal children. Br Med J 1987; 295: 573-7. 16. Brook CGD. Growth assessment in childhood and adolescence. Oxford: Blackwell Scientific, 1982: 164. 17. Tanner JM, Whitehouse RH. Cameron N, Marshall WA, Healy MJR, Goldstein H. Assessment of skeletal maturity and prediction of adult height (TW2 method). London: Academic Press, 1983: 106. 18. Tanner JM, Landt KW, Cameron N, Carter BS, Patel J. Prediction of adult height from height and bone age in childhood. Arch Dis Child 1983; 58: 767-76. 19. Tanner JM. Growth at adolescence. Oxford: Blackwell Scientific, 1962: 325. 20. Tanner JM, Whitehouse RH, Takaishi M. Standards from birth to maturity for height, weight, height velocity, and weight velocity: British children, 1%5. Part II. Arch Dis Child 1966; 41: 613-35. 21. Gertner JM. Tamborlane WV, Gianfredi SP. Gene1 M. Renewed catch-up growth with increased replacement doses of human growth hormone. J Pediatr 1987; 110: 425-8. 22. Bums EC, Tanner JM, Preece MA. Cameron N. Final height and pubertal development in 55 children with iiliopathic growth hormone deficiency, treated for between 2 and I5 years with human growth hormone. Eur J Pediatr 1981; 137: 155-64. 23. Joss E. Zuppinger K, Schwarz HP, Roten H. Final height of patients with pituitary growth failure and changes in growth variables after long term hormonal therapy. Pediatr Res 1983; 17: 676-9. 24. Bundak R, Hindmarsh PC, Smith PJ, Brook CGD. Long-term auxologic effects of human growth hormone. J Pediatr 1988; 112: 875-9. 25. Hindmarsh PC. Matthews DR, Brook CGD. Growth hormone secretion in children determined by time .series analysis. Clin Endocrinol 1988; 29: 35-44. 26. Smith PJ, Hindmarsh PC, Brook CGD. Contribution of dose and frequency of administration to the therapeutic effect of growth hormone. Arch Dis Child 1988; 63: 491-4. 27. Miller JD, Tannenbaum GS, Colle F, Guyda HJ. Daytime pulsatile growth hormone secretion during childhood and adolescence. J Clin Endocrinol Metab 1982; 55: 989-94. 18. Aynsley-Green A. Zachmann M, Prdder A. Interrelation of the therapeutic effects of growth hormone and testosterone on growth in hypopituitarism. J Pediatr 1976; 89: 992-9. 29. Tanner JM, Whitehouse RH. Hughes PCR, Carter BS. Relative importance of growth-hormone and sex steroids for the growth at puberty of trunk length, limb length, and muscle width in growth hormonedeficient children. J Pediatr 1976; 89: 1000-8. 30. Mauras N. Blizzard RM, Link K, Johnson ML. Rogol AD, Veldhuis ID. Augmentation of growth hormone secretion during puberty: evidence for a pulse amplitude-modulated phenomenon. J Clin Endocrinol Metab 1987; 64: 596-6431. 31. Stanhope R, Pringle PJ, Brook CGD. The mechanism of the adolescent growth spurt induced by low dose pulsatile GnRH treatment. Clin Endocrinol 1988; 28: 83-91. 32. Adashi EY. Resnick CE, Hernandez ER ef al. Insulin-like growth factor I as an amplifier of folliclestimulating hormone action: studies on mechanism(s) and site(s) of action in cultured rat granulosa cells. Endocrinology 1988; 122: 1583-91. 33. Homburg R. Eshel A, Abdalla HI. Jacobs HS. Growth hormone facilitates ovulation induction by gonadotrophins. Clin Endocrinol 1988; 29: 113-7. 34. Wilson ME, Gordon TP, Rudman CG. Tanner JM. Effects of growth hormone on the tempo of sexual maturation in female rhesus monkeys. J Clin Endocrinol Metab 1989: 68: 29-38. 35. Darendeliler F, Hindmarsh PC, Preece MA, Cox L. Brook CGD. Growth hormone increases rate o f pubertal maturation. Acta Endocrinol 1990; in press. 36. Rosenfeld RG, Wilson DM. Dollar LA, Bennett A, Hintz RL. Both human pituitary growth hormone and recombinant DNAderived human growth hormone cause insulin resistance at a postreceptor site. J Clin Endocrinol Metab 1982; 54: 1033-8. 37. Foster CM, Hale PM. Jing H-W, Schwanz J. Effects of human growth hormone on insulin-stimulated glucose metabolism in 3T3-F442A adipocytes. Endocrinology 1988; 123: 1082-8. 38. Binoux M, Lassarre C. Hardouin N. Somatomedin production by rat liver in organ culture. 111. Studies on the release of insulin-like growth factor and its carrier protein measured by radioligand assays. Acta Endocrinol 1982; 99: 422-30. 39. Scheiwiller E, Guler H-P. Menyweather J et ul. Growth restoration of insulin-deficient diabetic rats by recombinant human insulin-like growth factor 1. Nature 1986; 323: 169-71. (C.G.D.B.) Endocrine Unit The Middlesex Hospital Mortimer Street London WIN 8AA
UK
