Dose Dependence of Growth Response to Human Growth Hormone in Growth Hormone Deficiency M. A. PREECE, J. M. TANNER, R. H. WHITEHOUSE, AND N. CAMERON Department of Growth and Development, Institute of Child Health, University of London, and Growth Disorder Clinic, the Hospital for Sick Children, London, England end of the third. Adjusting treatment increment by covariance for bone age at the beginning of treatment, pre-treatment velocity, and body surface area did not alter these mean differences. Bone age velocity during treatment was the same in both treatment groups (mean 1.09 'years'/year in first year); thus we anticipate a gain in final adult height of the order of 10 cm from employing the larger dose. The decrease in skin folds occurring on treatment, however, was no different with the larger than with the smaller dose. This reinforces previous observations that the short-term metabolic and longer-term auxologic effects of hGH are not necessarily related. (/ Clin Endocrinol Metab 42: 477, 1976)
ABSTRACT. A trial of the relative effect on growth of 20 IU/week and 10 IU/week of human growth hormone has been made in 38 patients with "isolated" growth hormone deficiency over 1 year of treatment, 18 patients over 2 years and 10 over 3 years, and in 17 patients with surgically treated craniopharyngiomata over 1 year. The velocity of height growth in the first year of treatment, compared with a full year of pre-treatment control, was 1.3 times as great in both groups of patients on the larger dose as it was in those on the smaller one. Second-degree equations fitted to the treatment catch-up curves gave estimates of 1.7 cm more height gained on the larger dose by the end of the first year, 2.7 cm by the end of the second, and 3.4 cm by the
F
ROM 1959 to 1970 all patients entering the Medical Research Council Clinical Trial of Human Growth Hormone were treated with a nominal dose of 20 international units of hGH per week, divided into 2 doses administered 3 or 4 days apart (1). This dose, which was not varied according to body size, was greater than that used by investigators in other countries. In Switzerland, for example, Prader's group used 5 mg/m2 surface area twice weekly, which amounts to about 6 IU/week in a GHdeficient eight-year-old and 10 IU/week in a similar 12-year-old (2). In Norway, Trygstad (3) used 4 IU/week or 8 IU/week according to body weight. In the United States 3 injections a week were usually given with doses totalling between 6 IU/week and 12 IU/ week (4-7). From 1970 to 1975, therefore, children receiving treatment were allotted at random to one of two courses of treatment, 20 IU/ week and 10 IU/week. We have compared the effects on height growth in the first year of treatment in 38 children with Received March 10, 1975.
"isolated" growth hormone deficiency only, in 17 children who had had craniopharyngiomata surgically removed, and in the second and third years of treatment in a smaller number of patients. Materials and Methods Thirty-eight pre-pubertal children (35 boys and 3 girls) with "isolated" growth hormone deficiency (that is, deficiency either of GH solely, or of GH plus gonadotrophins) have been followed, using the pre-treatment and treatment schedule and techniques previously described (1). All measurements on each child were made by a single anthropometrist. All had a full year of three-monthly measurements before treatment. All had a clear-cut diagnosis with no signs of ACTH or TSH deficiencies or of intra-cranial lesions, either clinically or on appropriate radiological and laboratory investigation, which included testing for TSH and T4 levels and cortisol response in the insulin tolerance test. All responded to hGH administration. None developed signs of puberty during the treatment years considered. Gonadotrophin and sex hormone levels in blood or urine were not systematically measured. The age of the patients at starting treatment ranged from 4.7 to 13.9 years,
477
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 January 2016. at 13:01 For personal use only. No other uses without permission. . All rights reserved.
PREECE, TANNER, WHITEHOUSE AND CAMERON
478
averaging 10.0, for the small dose, and 5.6 to 14.3 years, averaging 9.3, for the large dose. Height velocity (cm/yr) in the pre-treatment year was calculated by fitting a straight line to the 5 available measurements and taking its slope. Height velocity in the first year of treatment was similarly calculated. We have used covariance analysis to allow for possible differences in treatment velocity associated with pretreatment levels of velocity (related to age at beginning of treatment) and pretreatment bone age (representing maturity at beginning of treatment). Bone age was calculated by the revised Tanner-Whitehouse (TW2) method (8) using the radius, ulna, metacarpals, and phalanges only, that is, the section of the method known as the RUS score (8). Skin folds were measured with the Harpenden caliper, and standard deviation scores of log transforms were calculated (1). In children who had been treated for 2 or more years, heights (h) at all available ages (t) after the beginning of treatment have been fitted by the parabola h = a + bt + ct2, the fit being by linear least squares (9). The constants a, b, and c were obtained for each individual. The means of these constants were calculated for individuals in the 10 IU/week and in the 20 IU/week treatment groups, and these means were then compared. Mean-constant curves (10,11), i.e., the curves constructed from the means of the 3 constants, have been plotted for the 2 treatment groups (in Fig. 2, below). The growth hormone used was prepared by Hartree's modifications of either the Raben method (12) or the Wilhelmi method (13). Results for the two sorts of hormone are presented both separately and pooled. Five different batches of Raben and 7 of Wilhelmi were involved, but the potency of all, as estimated by the growth in weight of the hypophysectomised rat, was between 0.8 and 1.2 IU/mg. Thus 5 mg TABLE 1. First-year treatment velocity (cm/yr) in children with 'isolated' growth hormone deficiency, according to dose of hGH, mean ± SE N 15 23
Dose hGH, IU/week
Raben
Wilhelmi
Both preparations
10 20 Difference
7.57 ±0.84 8.93 ±0.55 1.36 ± 1.00
7.17 ± 0.47 9.55 ± 0.57 2.38 ± 0.74f
7.33 ± 0.42 9.28 + 0.40 1.95 ± 0.58J
t Significant at 1% level.
JCE & M , 1976 Vol 42 No 3
of the preparation was injected intramuscularly twice weekly in the 10 IU/week cases and 10 mg twice weekly in the 20 IU/week cases, both dissolved in Sorensen's glycine buffer. Patients were allotted to the 2 treatment schedules according to the last digit of their case notes. In the supplementary study 17 children with craniopharyngiomata surgically removed were treated in an exactly similar way. These patients were receiving substitutive treatments as necessary, most being on thyroxine 0.1 to 0.3 mg/day, cortisol 7.5 to 15.0 mg/day, and syntopressin. Their range of ages at initial treatment was 9.2 to 18.7 years.
Results 1. "Isolated" growth hormone deficiency First treatment year. Table 1 shows the mean velocities in the first treatment year. Greater treatment velocities occurred on the larger dose of both preparations, the difference reaching the 1% level of significance for the Wilhelmi preparation and for both preparations pooled. On the average, the velocity on 20 IU/week was 1.3 times the velocity of 10 IU/week. The regression of treatment velocity on pre-treatment velocity and bone age at the beginning of treatment (all 38 cases) was: Treatment velocity (cm/yr) = 9.68 + 0.22 (±0.27) pre-treatment velocity (cm/yr) -0.31 (±0.12) bone age ('years') The partial regression coefficient for pretreatment velocity was not significantly different from zero, but that for bone age was significant at 5%; the multiple correlation was R = -0.46. The regressions of treatment velocity on bone age are shown in Fig. 1; the slopes are almost identical (high dose, -0.25 ± 0.14, low dose, -0.28 ± 0.12). The mean responses (9.06 cm/yr and 7.45 cm/yr at mean bone age) are little different from those without allowance for bone age (i.e., those in Table 1), which indicates that our randomization procedure has been effective. Adding surface area into this multiple
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 January 2016. at 13:01 For personal use only. No other uses without permission. . All rights reserved.
DOSE-DEPENDENT GROWTH RESPONSE TO hGH
479
14
12 >»
k 10 20 lU/Wk
t
8
O
o UJ
6
10 lU/Wk
SI
4
UJ
oc
2
0
2
4
6
8
10
12
14
INITIAL BONE AGE (years ) FIG. 1. Regressions of first-year treatment velocity on initial bone age, in high (A) and low (•) dosage groups.
regression did not significantly increase difference became still smaller, and the the correlation. The partial correlation co- pooled mean increase was 1.09 ± 0.11 efficient of surface area with treatment 'years.' velocity when bone age and initial velocity were partialled out was insignificantly dif- Second and third treatment years. Thirteen ferent from zero. patients were followed for 2 years at the high The bone age advanced an average of dose and 5 at the low. The measured 1.14 ± 0.15 'years' in the large-dose group heights (h) of each individual from the beand 1.02 ± 0.16 'years' in the small-dose ginning of treatment werefittedby the curve group. These values do not differ signifi- h = a + bt + ct2 where t is age in years. In cantly. Bone age increase is related to initial every case the fit was excellent, the runs bone age delay, (i.e., chronological age less test (14) (for points being systematically bone age). The regressions were not signifi- above or below the curve according to the cantly different in the two groups and the value of t) negative and the residuals about pooled correlation coefficient was 0.43. the curve of the order of 2 to 3 mm. When the mean bone age increments in the Table 2 shows the means (±SE) of the two treatment groups were compared using parameters in each treatment group. The covariance analysis to adjust for this, the table also displays similar fits for the first
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 January 2016. at 13:01 For personal use only. No other uses without permission. . All rights reserved.
T A B L E 2.
JCE & M • 1976 Vol 42 • No 3
PREECE, TANNER, WHITEHOUSE AND CAMERON
480
Mean values (±SE) of parameters a, b, and c of the fitted parabola h = a + bt + ct2 according to length of treatment and dose of hGH (h is treatment height increment, t is age), mean ± SE hGH, 10 IU/week
Length of treatment (years)
hGH, 20 IU/week N
a
b
c
N
a
b
c
1 2 3
23 13 7
-0.09 ±0.05 -0.07 ± 0.06 0.12 ± 0.11
11.29 ±0.61* 11.20 ± 0.56f 9.96 ± 0.62*
-1.95 :tO.40 - 1 . 5 3 : t0.21 -0.88 :tO.07
15 5 3
0.02 ±:0.03 -0.01 ±:0.13 -0.22 ±:0.20
8.59 ± 0.83 8.56 ± 0.56 8.50 ± 0.34
-1.09 ± 0.59 -0.86 ± 0.33 -0.69 ± 0.18
Corresponding b's significantly different at 5% (*) and 1% (f) levels.
year only, and for individuals who were followed for 3 years. In none of the 6 groups (i.e., 1, 2, 3, years, and 10 IU/week, 20 IU/ week dosage) does the parameter a differ significantly from zero; while all b's and c's do so. The parameter b differs significantly between the two treatment groups at 1 year, 2 years, and 3 years. Though c is significantly different from zero it does not differ between treatment groups. Figure 2 illustrates the two-year and threeyear mean constant curves. The height increment achieved at the end of two years' treatment by the patients on 20 IU/week averaged 2.7 ± 0.8 cm more than the increment achieved by the patients on 10 IU/ week: at 3 years the difference averaged
3.4 ± 1.5 cm. (The differences of height increment at 2 years between all two-year patients and those two-year patients followed for 3 years was statistically insignificant.) The increments in bone age in the two treatment groups were practically equal over the second and third years, as they were over the first. 2. Patients with surgically treated craniopharyngiomata Table 3 shows the mean first-year treatment velocity for the craniopharyngiomata cases, with the Wilhelmi and Raben preparations pooled. The results were similar to
24 20 lU/Wk
16
FIG. 2. Mean-constant curves of height increment from the beginning of treatment to 2 years ( ) or 3 years ( ). Constants are calculated from 13 cases on 20 IU/week and 5 cases on 10 IU/week for 2-year followup, and from 7 cases on 20 IU/ week and 3 cases on 10 IU/week for 3-year follow-up.
1
2
YEARS OF HGH TREATMENT
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 January 2016. at 13:01 For personal use only. No other uses without permission. . All rights reserved.
D O S E - D E P E N D E N T GROWTH RESPONSE TO hGH
N
Dose IU/week
Both Preparations
10 20 Difference
5.45 ± 0.41 7.70 ± 0.60 2.25 ± 0.731
t Significant at 1% level.
those for "isolated" growth hormone deficiency patients, with the differences averaging 2.25 ± 0.73 cm/yr. The high dose produced 1.4 times the velocity of the low dose. In this instance, however, randomization worked less well, and the covariance analysis assumes some importance. The bone age in the low dosage group was by chance greater than in the high dosage group (averaging 10.6 against 9.0 'years'). When adjusted for bone age, the means of the two groups were 7.37 cm/yr and 5.83 cm/ yr, a difference of 1.54 ± 0.56 cm/yr (P = < 0.05). With this necessary allowance, then, the large dose produced an acceleration of 1.3 times the small dose, in agreement with the results for children with "isolated" growth hormone deficiency. 3. Skin folds One of the most regularly seen effects of the administration of hGH is a decrease in skin folds (15). Table 4 shows the mean decrease caused by each dose of hGH in the two sets of patients, at 6 months and 12 months. The average of the triceps and subscapular folds is given, as changes in the means of the standard deviation scores of the two log transforms. Though the larger dose did cause a slightly greater decrease than the smaller one, the difference was far from statistically significant. Discussion There are few reports of dosage comparisons to date. In the American collaborative study (6,7) 6 IU/week was compared with 30
IU/week in a small number of idiopathic growth hormone-deficient patients, all given 3 injections per week as contrasted to our 2 injections per week. Patients on the lower dose had an average increment of 9.3 cm in the first year and those on the higher dose (numbering only 9, an increment of 11.9 cm. Due to differences in the availability of hGH however, these patients were more heavily selected than ours, with bone ages and height much lower in relation to age. If our regressions of first-year response on bone age are used to adjust, at least approximately, for this, the USA values would be about 8.2 cm for 6 IU/week and 10.8 cm for 30 IU/week, compared with our values of 7.5 cm for 10 IU/week and 9.1 cm for 20 IU/week. The apparent advantage of 0.5 to 1 cm of the American collaborative project may be due to the greater frequency of injection, and we have now embarked on a trial of 5 IU 3 times per week. But the two trials may not be altogether comparable; the results for bone age increase are quite dissimilar. In our earlier trial, 6 months' treatment with a dose of 40 IU/week, administered when the growth velocity had dropped to within normal limits, did not result in re-acceleration of growth (1). Hall and Olin (16) in Sweden have also reported greater growth rates at treatment doses of 10-18 IU/week than at 8 IU/week. Their material and methods are too dissimilar to permit closer comparison however. TABLE 4. Decrease in triceps and subscapular skinfolds (mean of SD scores of log transforms) after 6 months and 12 months of hGH at 10 and 20 IU/week, mean ± SE N
Dose hGH, IU/week
6 months
12 months
0.69 ± 0.17 0.76 ± 0.13
0.67 ± 0.19 0.83 ± 0.14
0.57 ± 0.20 0.71 ± 0.13
0.74 ± 0.20 0.68 ± 0.09
'Isolated' GH Deficiency 15 23
10 20
Craniopharyngiomata CD 00
TABLE 3. First-year treatment velocity (cm/yr) in children with craniopharyngioma, according to dose of hGH, mean ± SE
481
10 20
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 January 2016. at 13:01 For personal use only. No other uses without permission. . All rights reserved.
482
PREECE, TANNER, WHITEHOUSE AND CAMERON
There is little basis in experimental work either for the present dose of hGH or for the frequency of its administration. Treatment schedules were set up before it was appreciated that growth hormone acted through the release of somatomedin and not directly. Hall (17) has shown that plasma somatomedin in growth hormone-deficient children rises some 3 hours after intravenous infusion of hGH and remains elevated for some period greater than 24 hours. Thus, a thrice-weekly frequency is probably near optimal, and a twice-weekly frequency may combine acceptability and therapeutic effect reasonably well. The dose schedule was originally based on the effects of metabolic indices in short-term studies, but these have been shown to be not at all predictive of growth response to a year's treatment in children with growth hormone deficiency (18). Our data on skin folds reinforce this conclusion, by failing to show a dose effect at either 6 or 12 months. Little is yet known about the average production rate of growth hormone per 24 hours in children. The integrated 24-hour serum levels are considerably higher than those of adults (19) and the daily production rate in adults has been estimated to be on the order of 1 mg/day (20). It is really not possible at present to derive a satisfactory treatment schedule from these considerations. We remain dependent on empirical trials such as those described here, and on this evidence a schedule of 20 IU/week produced a clearly greater response than a schedule of 10 IU/week at the same twiceweekly frequency. Since the larger dose gave no greater increase in bone age than did the smaller one, it seems likely that children who start treatment at around 6 years of age would have a total increment in height of between 5 and 10 cm greater on the larger dose than on the smaller one. Acknowledgments We wish to thank Dr. W. A. Marshall and Mr. M. J. R. Healy for helpful comments and the Medical Research
JCE & M • 1976 Vol 42 • No 3
Council for supplies of growth hormone and for financial support. We are especially grateful for the secretarial and administrative assistance of Miss Janet Baines.
References 1. Tanner, J. M., R. H. Whitehouse, P. C. R. Hughes, and F. P. Vince, The effect of human growth hormone treatment for 1 to 7 years on the growth of 100 children with growth hormone deficiency, low birthweight, inherited smallness, Turner's syndrome, and other complaints, Arch Dis Child 46: 745, 1971. 2. Prader, A., M. Zachmann, J. R. Poley, R. Illig, and J. Szeky, Long-term treatment with human growth hormone (Raben) in small doses: Evaluation of .18 hypopituitary patients, Helv Paediatr Ada 22: 423, 1967. 3. Trygstad, O., Human growth hormone and hypopituitary growth retardation. Ada Paediatr Scand 58: 407, 1969. 4. Goodman, H. G., M. M. Grumbach, and S. L. Kaplan, Growth and growth hormone. II. A comparison of isolated growth hormone deficiency and multiple pituitary-hormone deficiencies in 35 patients with idiopathic hypopituitary dwarfisrn, New Engl J Med 278: 57, 1968. 5. Soyka, L. F., H. H. Bode, J. D. Crawford, and F. J. Flynn, Jr., Effectiveness of long-term human growth hormone therapy for short stature in children with growth hormone deficiency, J Clin Endocrinol Metab 30: 1, 1970. 6. Aceto, T., S. D. Frasier, A. B. Hayles, H. F. L. Meyer-Bahlburg, M. L. Parker, R. Munschauer, and G. Di Chiro, Collaborative study of the effects of human growth hormone in growth hormone deficiency. I. First year of therapy, J Clin Endocrinol Metab 35: 483, 1972. 7. Aceto, T., S. D. Frasier, A. B. Hayles, H. F. L. Meyer-Bahlburg, M. L. Parker, R. Munschauer, and G. Di Chiro, Collaborative study of the effects of human growth hormone in growth hormone deficiency. 3. First eighteen months of therapy, In Advances in Human Growth Hormone Research: A Symposium 1973, U.S. Dept. of Health, Education, and Welfare, Publication No. NIH 74-612, p. 695. 8. Tanner, J. M., R. H. Whitehouse, W. A. Marshall, M. J. R. Healy, and H. Goldstein, Assessment of Skeletal Maturity and Prediction of Adult Height, Academic Press, London, 1975. 9. Snedecor, G. W., and W. G. Cochran, Statistical Methods, ed. 6, Iowa State University Press, Ames, 1967, p. 453. 10. Merrill, M., The relationship of individual growth to average growth, Hum Biol 3: 37, 1931. 11. Tanner, J. M., Growth at Adolescence, ed. 2,
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 January 2016. at 13:01 For personal use only. No other uses without permission. . All rights reserved.
DOSE-DEPENDENT GROWTH RESPONSE TO hGH
12. 13. 14. 15.
16.
Blackvvell Scientific Publications, Oxford, and Charles C Thomas, Springfield, 1962, p. 9. Hartree, A. S., Separation and partial purification of the protein hormones from human pituitary glands, Biochem J 100: 754, 1966. Hartree, A. S., Preparation and properties of human growth hormone, In Mason, S. (ed.), Human Growth Hormone, London, Butterworth, 1972, p. 1. Siegel, S., Nonparametric Statistics for the Behavioral Sciences, McGraw-Hill, New York, 1956, p. 52. Tanner, J. M., and R. H. Whitehouse, The effect of human growth hormone on subcutaneous fat thickness in hyposomatotrophic and panhypopituitary dwarfs, J Endocrinol 39: 263, 1967 Hall, Kerstin, and Patrick Olin, Sulphation factor activity and growth rate during long-term treatment of patients with pituitary dwarfism with human growth hormone, Acta Endocrinol (Kbh) 69: 417, 1972.
483
17. Hall, K., Effect of intravenous administration of human growth hormone on sulphation factor activity in serum in hypopituitary subjects, Acta Endocrinol (Kbh) 66: 491, 1971. 18. Clayton, B. E., J. M. Tanner, and F. P. Vince, Diagnostic and prognostic value of short term metabolic response to human growth hormone in short stature, Arch Dis Child 46: 405, 1971. 19. Blizzard, R. M., R. G. Thompson, A. Baghdassarian, A. Kowarski, C. J. Migeon, and A. Rodriguez, The interrelationship of steroids, growth hormone and other hormones on pubertal growth, In Grumbach, M. M., G. D. Grave, and F. E. Mayer (eds.) Control of the Onset of Puberty, Wiley, New York, 1974, p. 342. 20. Kowarski, A., R. G. Thompson, C. J. Migeon, and R. M. Blizzard, Determination of integrated plasma concentrations and true secretion rates of human growth hormone,7 Clin Endocrinol Metah 32: 356, 1971.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 January 2016. at 13:01 For personal use only. No other uses without permission. . All rights reserved.
ABSTRACT. A trial of the relative effect on growth of 20 IU/week and 10 IU/week of human growth hormone has been made in 38 patients with "isolated" growth hormone deficiency over 1 year of treatment, 18 patients over 2 years and 10 over 3 years, and in 17 patients with surgically treated craniopharyngiomata over 1 year. The velocity of height growth in the first year of treatment, compared with a full year of pre-treatment control, was 1.3 times as great in both groups of patients on the larger dose as it was in those on the smaller one. Second-degree equations fitted to the treatment catch-up curves gave estimates of 1.7 cm more height gained on the larger dose by the end of the first year, 2.7 cm by the end of the second, and 3.4 cm by the
F
ROM 1959 to 1970 all patients entering the Medical Research Council Clinical Trial of Human Growth Hormone were treated with a nominal dose of 20 international units of hGH per week, divided into 2 doses administered 3 or 4 days apart (1). This dose, which was not varied according to body size, was greater than that used by investigators in other countries. In Switzerland, for example, Prader's group used 5 mg/m2 surface area twice weekly, which amounts to about 6 IU/week in a GHdeficient eight-year-old and 10 IU/week in a similar 12-year-old (2). In Norway, Trygstad (3) used 4 IU/week or 8 IU/week according to body weight. In the United States 3 injections a week were usually given with doses totalling between 6 IU/week and 12 IU/ week (4-7). From 1970 to 1975, therefore, children receiving treatment were allotted at random to one of two courses of treatment, 20 IU/ week and 10 IU/week. We have compared the effects on height growth in the first year of treatment in 38 children with Received March 10, 1975.
"isolated" growth hormone deficiency only, in 17 children who had had craniopharyngiomata surgically removed, and in the second and third years of treatment in a smaller number of patients. Materials and Methods Thirty-eight pre-pubertal children (35 boys and 3 girls) with "isolated" growth hormone deficiency (that is, deficiency either of GH solely, or of GH plus gonadotrophins) have been followed, using the pre-treatment and treatment schedule and techniques previously described (1). All measurements on each child were made by a single anthropometrist. All had a full year of three-monthly measurements before treatment. All had a clear-cut diagnosis with no signs of ACTH or TSH deficiencies or of intra-cranial lesions, either clinically or on appropriate radiological and laboratory investigation, which included testing for TSH and T4 levels and cortisol response in the insulin tolerance test. All responded to hGH administration. None developed signs of puberty during the treatment years considered. Gonadotrophin and sex hormone levels in blood or urine were not systematically measured. The age of the patients at starting treatment ranged from 4.7 to 13.9 years,
477
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 January 2016. at 13:01 For personal use only. No other uses without permission. . All rights reserved.
PREECE, TANNER, WHITEHOUSE AND CAMERON
478
averaging 10.0, for the small dose, and 5.6 to 14.3 years, averaging 9.3, for the large dose. Height velocity (cm/yr) in the pre-treatment year was calculated by fitting a straight line to the 5 available measurements and taking its slope. Height velocity in the first year of treatment was similarly calculated. We have used covariance analysis to allow for possible differences in treatment velocity associated with pretreatment levels of velocity (related to age at beginning of treatment) and pretreatment bone age (representing maturity at beginning of treatment). Bone age was calculated by the revised Tanner-Whitehouse (TW2) method (8) using the radius, ulna, metacarpals, and phalanges only, that is, the section of the method known as the RUS score (8). Skin folds were measured with the Harpenden caliper, and standard deviation scores of log transforms were calculated (1). In children who had been treated for 2 or more years, heights (h) at all available ages (t) after the beginning of treatment have been fitted by the parabola h = a + bt + ct2, the fit being by linear least squares (9). The constants a, b, and c were obtained for each individual. The means of these constants were calculated for individuals in the 10 IU/week and in the 20 IU/week treatment groups, and these means were then compared. Mean-constant curves (10,11), i.e., the curves constructed from the means of the 3 constants, have been plotted for the 2 treatment groups (in Fig. 2, below). The growth hormone used was prepared by Hartree's modifications of either the Raben method (12) or the Wilhelmi method (13). Results for the two sorts of hormone are presented both separately and pooled. Five different batches of Raben and 7 of Wilhelmi were involved, but the potency of all, as estimated by the growth in weight of the hypophysectomised rat, was between 0.8 and 1.2 IU/mg. Thus 5 mg TABLE 1. First-year treatment velocity (cm/yr) in children with 'isolated' growth hormone deficiency, according to dose of hGH, mean ± SE N 15 23
Dose hGH, IU/week
Raben
Wilhelmi
Both preparations
10 20 Difference
7.57 ±0.84 8.93 ±0.55 1.36 ± 1.00
7.17 ± 0.47 9.55 ± 0.57 2.38 ± 0.74f
7.33 ± 0.42 9.28 + 0.40 1.95 ± 0.58J
t Significant at 1% level.
JCE & M , 1976 Vol 42 No 3
of the preparation was injected intramuscularly twice weekly in the 10 IU/week cases and 10 mg twice weekly in the 20 IU/week cases, both dissolved in Sorensen's glycine buffer. Patients were allotted to the 2 treatment schedules according to the last digit of their case notes. In the supplementary study 17 children with craniopharyngiomata surgically removed were treated in an exactly similar way. These patients were receiving substitutive treatments as necessary, most being on thyroxine 0.1 to 0.3 mg/day, cortisol 7.5 to 15.0 mg/day, and syntopressin. Their range of ages at initial treatment was 9.2 to 18.7 years.
Results 1. "Isolated" growth hormone deficiency First treatment year. Table 1 shows the mean velocities in the first treatment year. Greater treatment velocities occurred on the larger dose of both preparations, the difference reaching the 1% level of significance for the Wilhelmi preparation and for both preparations pooled. On the average, the velocity on 20 IU/week was 1.3 times the velocity of 10 IU/week. The regression of treatment velocity on pre-treatment velocity and bone age at the beginning of treatment (all 38 cases) was: Treatment velocity (cm/yr) = 9.68 + 0.22 (±0.27) pre-treatment velocity (cm/yr) -0.31 (±0.12) bone age ('years') The partial regression coefficient for pretreatment velocity was not significantly different from zero, but that for bone age was significant at 5%; the multiple correlation was R = -0.46. The regressions of treatment velocity on bone age are shown in Fig. 1; the slopes are almost identical (high dose, -0.25 ± 0.14, low dose, -0.28 ± 0.12). The mean responses (9.06 cm/yr and 7.45 cm/yr at mean bone age) are little different from those without allowance for bone age (i.e., those in Table 1), which indicates that our randomization procedure has been effective. Adding surface area into this multiple
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 January 2016. at 13:01 For personal use only. No other uses without permission. . All rights reserved.
DOSE-DEPENDENT GROWTH RESPONSE TO hGH
479
14
12 >»
k 10 20 lU/Wk
t
8
O
o UJ
6
10 lU/Wk
SI
4
UJ
oc
2
0
2
4
6
8
10
12
14
INITIAL BONE AGE (years ) FIG. 1. Regressions of first-year treatment velocity on initial bone age, in high (A) and low (•) dosage groups.
regression did not significantly increase difference became still smaller, and the the correlation. The partial correlation co- pooled mean increase was 1.09 ± 0.11 efficient of surface area with treatment 'years.' velocity when bone age and initial velocity were partialled out was insignificantly dif- Second and third treatment years. Thirteen ferent from zero. patients were followed for 2 years at the high The bone age advanced an average of dose and 5 at the low. The measured 1.14 ± 0.15 'years' in the large-dose group heights (h) of each individual from the beand 1.02 ± 0.16 'years' in the small-dose ginning of treatment werefittedby the curve group. These values do not differ signifi- h = a + bt + ct2 where t is age in years. In cantly. Bone age increase is related to initial every case the fit was excellent, the runs bone age delay, (i.e., chronological age less test (14) (for points being systematically bone age). The regressions were not signifi- above or below the curve according to the cantly different in the two groups and the value of t) negative and the residuals about pooled correlation coefficient was 0.43. the curve of the order of 2 to 3 mm. When the mean bone age increments in the Table 2 shows the means (±SE) of the two treatment groups were compared using parameters in each treatment group. The covariance analysis to adjust for this, the table also displays similar fits for the first
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 January 2016. at 13:01 For personal use only. No other uses without permission. . All rights reserved.
T A B L E 2.
JCE & M • 1976 Vol 42 • No 3
PREECE, TANNER, WHITEHOUSE AND CAMERON
480
Mean values (±SE) of parameters a, b, and c of the fitted parabola h = a + bt + ct2 according to length of treatment and dose of hGH (h is treatment height increment, t is age), mean ± SE hGH, 10 IU/week
Length of treatment (years)
hGH, 20 IU/week N
a
b
c
N
a
b
c
1 2 3
23 13 7
-0.09 ±0.05 -0.07 ± 0.06 0.12 ± 0.11
11.29 ±0.61* 11.20 ± 0.56f 9.96 ± 0.62*
-1.95 :tO.40 - 1 . 5 3 : t0.21 -0.88 :tO.07
15 5 3
0.02 ±:0.03 -0.01 ±:0.13 -0.22 ±:0.20
8.59 ± 0.83 8.56 ± 0.56 8.50 ± 0.34
-1.09 ± 0.59 -0.86 ± 0.33 -0.69 ± 0.18
Corresponding b's significantly different at 5% (*) and 1% (f) levels.
year only, and for individuals who were followed for 3 years. In none of the 6 groups (i.e., 1, 2, 3, years, and 10 IU/week, 20 IU/ week dosage) does the parameter a differ significantly from zero; while all b's and c's do so. The parameter b differs significantly between the two treatment groups at 1 year, 2 years, and 3 years. Though c is significantly different from zero it does not differ between treatment groups. Figure 2 illustrates the two-year and threeyear mean constant curves. The height increment achieved at the end of two years' treatment by the patients on 20 IU/week averaged 2.7 ± 0.8 cm more than the increment achieved by the patients on 10 IU/ week: at 3 years the difference averaged
3.4 ± 1.5 cm. (The differences of height increment at 2 years between all two-year patients and those two-year patients followed for 3 years was statistically insignificant.) The increments in bone age in the two treatment groups were practically equal over the second and third years, as they were over the first. 2. Patients with surgically treated craniopharyngiomata Table 3 shows the mean first-year treatment velocity for the craniopharyngiomata cases, with the Wilhelmi and Raben preparations pooled. The results were similar to
24 20 lU/Wk
16
FIG. 2. Mean-constant curves of height increment from the beginning of treatment to 2 years ( ) or 3 years ( ). Constants are calculated from 13 cases on 20 IU/week and 5 cases on 10 IU/week for 2-year followup, and from 7 cases on 20 IU/ week and 3 cases on 10 IU/week for 3-year follow-up.
1
2
YEARS OF HGH TREATMENT
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 January 2016. at 13:01 For personal use only. No other uses without permission. . All rights reserved.
D O S E - D E P E N D E N T GROWTH RESPONSE TO hGH
N
Dose IU/week
Both Preparations
10 20 Difference
5.45 ± 0.41 7.70 ± 0.60 2.25 ± 0.731
t Significant at 1% level.
those for "isolated" growth hormone deficiency patients, with the differences averaging 2.25 ± 0.73 cm/yr. The high dose produced 1.4 times the velocity of the low dose. In this instance, however, randomization worked less well, and the covariance analysis assumes some importance. The bone age in the low dosage group was by chance greater than in the high dosage group (averaging 10.6 against 9.0 'years'). When adjusted for bone age, the means of the two groups were 7.37 cm/yr and 5.83 cm/ yr, a difference of 1.54 ± 0.56 cm/yr (P = < 0.05). With this necessary allowance, then, the large dose produced an acceleration of 1.3 times the small dose, in agreement with the results for children with "isolated" growth hormone deficiency. 3. Skin folds One of the most regularly seen effects of the administration of hGH is a decrease in skin folds (15). Table 4 shows the mean decrease caused by each dose of hGH in the two sets of patients, at 6 months and 12 months. The average of the triceps and subscapular folds is given, as changes in the means of the standard deviation scores of the two log transforms. Though the larger dose did cause a slightly greater decrease than the smaller one, the difference was far from statistically significant. Discussion There are few reports of dosage comparisons to date. In the American collaborative study (6,7) 6 IU/week was compared with 30
IU/week in a small number of idiopathic growth hormone-deficient patients, all given 3 injections per week as contrasted to our 2 injections per week. Patients on the lower dose had an average increment of 9.3 cm in the first year and those on the higher dose (numbering only 9, an increment of 11.9 cm. Due to differences in the availability of hGH however, these patients were more heavily selected than ours, with bone ages and height much lower in relation to age. If our regressions of first-year response on bone age are used to adjust, at least approximately, for this, the USA values would be about 8.2 cm for 6 IU/week and 10.8 cm for 30 IU/week, compared with our values of 7.5 cm for 10 IU/week and 9.1 cm for 20 IU/week. The apparent advantage of 0.5 to 1 cm of the American collaborative project may be due to the greater frequency of injection, and we have now embarked on a trial of 5 IU 3 times per week. But the two trials may not be altogether comparable; the results for bone age increase are quite dissimilar. In our earlier trial, 6 months' treatment with a dose of 40 IU/week, administered when the growth velocity had dropped to within normal limits, did not result in re-acceleration of growth (1). Hall and Olin (16) in Sweden have also reported greater growth rates at treatment doses of 10-18 IU/week than at 8 IU/week. Their material and methods are too dissimilar to permit closer comparison however. TABLE 4. Decrease in triceps and subscapular skinfolds (mean of SD scores of log transforms) after 6 months and 12 months of hGH at 10 and 20 IU/week, mean ± SE N
Dose hGH, IU/week
6 months
12 months
0.69 ± 0.17 0.76 ± 0.13
0.67 ± 0.19 0.83 ± 0.14
0.57 ± 0.20 0.71 ± 0.13
0.74 ± 0.20 0.68 ± 0.09
'Isolated' GH Deficiency 15 23
10 20
Craniopharyngiomata CD 00
TABLE 3. First-year treatment velocity (cm/yr) in children with craniopharyngioma, according to dose of hGH, mean ± SE
481
10 20
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 January 2016. at 13:01 For personal use only. No other uses without permission. . All rights reserved.
482
PREECE, TANNER, WHITEHOUSE AND CAMERON
There is little basis in experimental work either for the present dose of hGH or for the frequency of its administration. Treatment schedules were set up before it was appreciated that growth hormone acted through the release of somatomedin and not directly. Hall (17) has shown that plasma somatomedin in growth hormone-deficient children rises some 3 hours after intravenous infusion of hGH and remains elevated for some period greater than 24 hours. Thus, a thrice-weekly frequency is probably near optimal, and a twice-weekly frequency may combine acceptability and therapeutic effect reasonably well. The dose schedule was originally based on the effects of metabolic indices in short-term studies, but these have been shown to be not at all predictive of growth response to a year's treatment in children with growth hormone deficiency (18). Our data on skin folds reinforce this conclusion, by failing to show a dose effect at either 6 or 12 months. Little is yet known about the average production rate of growth hormone per 24 hours in children. The integrated 24-hour serum levels are considerably higher than those of adults (19) and the daily production rate in adults has been estimated to be on the order of 1 mg/day (20). It is really not possible at present to derive a satisfactory treatment schedule from these considerations. We remain dependent on empirical trials such as those described here, and on this evidence a schedule of 20 IU/week produced a clearly greater response than a schedule of 10 IU/week at the same twiceweekly frequency. Since the larger dose gave no greater increase in bone age than did the smaller one, it seems likely that children who start treatment at around 6 years of age would have a total increment in height of between 5 and 10 cm greater on the larger dose than on the smaller one. Acknowledgments We wish to thank Dr. W. A. Marshall and Mr. M. J. R. Healy for helpful comments and the Medical Research
JCE & M • 1976 Vol 42 • No 3
Council for supplies of growth hormone and for financial support. We are especially grateful for the secretarial and administrative assistance of Miss Janet Baines.
References 1. Tanner, J. M., R. H. Whitehouse, P. C. R. Hughes, and F. P. Vince, The effect of human growth hormone treatment for 1 to 7 years on the growth of 100 children with growth hormone deficiency, low birthweight, inherited smallness, Turner's syndrome, and other complaints, Arch Dis Child 46: 745, 1971. 2. Prader, A., M. Zachmann, J. R. Poley, R. Illig, and J. Szeky, Long-term treatment with human growth hormone (Raben) in small doses: Evaluation of .18 hypopituitary patients, Helv Paediatr Ada 22: 423, 1967. 3. Trygstad, O., Human growth hormone and hypopituitary growth retardation. Ada Paediatr Scand 58: 407, 1969. 4. Goodman, H. G., M. M. Grumbach, and S. L. Kaplan, Growth and growth hormone. II. A comparison of isolated growth hormone deficiency and multiple pituitary-hormone deficiencies in 35 patients with idiopathic hypopituitary dwarfisrn, New Engl J Med 278: 57, 1968. 5. Soyka, L. F., H. H. Bode, J. D. Crawford, and F. J. Flynn, Jr., Effectiveness of long-term human growth hormone therapy for short stature in children with growth hormone deficiency, J Clin Endocrinol Metab 30: 1, 1970. 6. Aceto, T., S. D. Frasier, A. B. Hayles, H. F. L. Meyer-Bahlburg, M. L. Parker, R. Munschauer, and G. Di Chiro, Collaborative study of the effects of human growth hormone in growth hormone deficiency. I. First year of therapy, J Clin Endocrinol Metab 35: 483, 1972. 7. Aceto, T., S. D. Frasier, A. B. Hayles, H. F. L. Meyer-Bahlburg, M. L. Parker, R. Munschauer, and G. Di Chiro, Collaborative study of the effects of human growth hormone in growth hormone deficiency. 3. First eighteen months of therapy, In Advances in Human Growth Hormone Research: A Symposium 1973, U.S. Dept. of Health, Education, and Welfare, Publication No. NIH 74-612, p. 695. 8. Tanner, J. M., R. H. Whitehouse, W. A. Marshall, M. J. R. Healy, and H. Goldstein, Assessment of Skeletal Maturity and Prediction of Adult Height, Academic Press, London, 1975. 9. Snedecor, G. W., and W. G. Cochran, Statistical Methods, ed. 6, Iowa State University Press, Ames, 1967, p. 453. 10. Merrill, M., The relationship of individual growth to average growth, Hum Biol 3: 37, 1931. 11. Tanner, J. M., Growth at Adolescence, ed. 2,
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 January 2016. at 13:01 For personal use only. No other uses without permission. . All rights reserved.
DOSE-DEPENDENT GROWTH RESPONSE TO hGH
12. 13. 14. 15.
16.
Blackvvell Scientific Publications, Oxford, and Charles C Thomas, Springfield, 1962, p. 9. Hartree, A. S., Separation and partial purification of the protein hormones from human pituitary glands, Biochem J 100: 754, 1966. Hartree, A. S., Preparation and properties of human growth hormone, In Mason, S. (ed.), Human Growth Hormone, London, Butterworth, 1972, p. 1. Siegel, S., Nonparametric Statistics for the Behavioral Sciences, McGraw-Hill, New York, 1956, p. 52. Tanner, J. M., and R. H. Whitehouse, The effect of human growth hormone on subcutaneous fat thickness in hyposomatotrophic and panhypopituitary dwarfs, J Endocrinol 39: 263, 1967 Hall, Kerstin, and Patrick Olin, Sulphation factor activity and growth rate during long-term treatment of patients with pituitary dwarfism with human growth hormone, Acta Endocrinol (Kbh) 69: 417, 1972.
483
17. Hall, K., Effect of intravenous administration of human growth hormone on sulphation factor activity in serum in hypopituitary subjects, Acta Endocrinol (Kbh) 66: 491, 1971. 18. Clayton, B. E., J. M. Tanner, and F. P. Vince, Diagnostic and prognostic value of short term metabolic response to human growth hormone in short stature, Arch Dis Child 46: 405, 1971. 19. Blizzard, R. M., R. G. Thompson, A. Baghdassarian, A. Kowarski, C. J. Migeon, and A. Rodriguez, The interrelationship of steroids, growth hormone and other hormones on pubertal growth, In Grumbach, M. M., G. D. Grave, and F. E. Mayer (eds.) Control of the Onset of Puberty, Wiley, New York, 1974, p. 342. 20. Kowarski, A., R. G. Thompson, C. J. Migeon, and R. M. Blizzard, Determination of integrated plasma concentrations and true secretion rates of human growth hormone,7 Clin Endocrinol Metah 32: 356, 1971.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 January 2016. at 13:01 For personal use only. No other uses without permission. . All rights reserved.
