0021-972X/90/7106-1496$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1990 by The Endocrine Society
Vol. 71, No. 6 Printed in U.S.A.
Urinary Human Growth Hormone Measurement Using a Highly Sensitive Sandwich Enzyme Immunoassay: Diagnostic and Therapeutic Uses in Patients with Growth Hormone Deficiency* HITOSHI KOHNO, YOSHIAKI MURAKAMI, AND TSUKASA KODAIRA Department of Endocrinology and Metabolism, Fukuoka Children's Hospital Medical Center (H.K), Fukuoka 810; Research Laboratories, Research and Development Division, Sumitomo Pharmaceuticals Co., Ltd. (Y.M.), Takarazuka, Hyogo 665; and Clinical Testing Servicies, Otsuka Assay Laboratories, Otsuka Pharmaceuticals Co., Ltd. (T.K), Tokushima 771-01, Japan
ABSTRACT. Several reports indicate that urinary hGH excretion is significantly lower in patients with either partial (PGHD) or complete GH deficiency (CGHD) than in normal but short children (NSC) or normal children (NC). However, there is an overlap between the NSC and NC groups and the PGHD group. Using a highly sensitive sandwich enzyme immunoassay, we investigated whether the measurement of urinary hGH can clearly separate the PGHD and CGHD groups from the NSC and NC groups. In addition, we measured the urinary excretion of synthetic methionyl-hGH (met-hGH) in PGHD and CGHD after sc injections of 2 and 4 IU and im injections of 4 IU in an attempt to determine the optimal replacement dose. Total 24-h urinary hGH excretion in each patient examined for 2 consecutive days varied from 1 day to the next. There were no differencies in urinary hGH excretions between the NSC group and the NC group. The lower values for daily urinary hGH excretion in the NSC group overlapped some of the higher values in the PGHD group. However, when the mean urinary hGH level of both days was used, the 24-h urinary hGH excretion clearly separated the PGHD (5.5 ± 2.3 ng/day; range, 1.3-9.2; n
R
ECENTLY, a highly sensitive sandwich enzyme immunoassay (EIA) for human GH (hGH) was developed that used monoclonal anti-hGH immunoglobulin G-coated polystyrene balls and affinity-purified rabbit anti-hGH Fab'-horseradish peroxidase conjugate (1, 2). This EIA method made it possible to measure very low levels of unconcentrated urinary hGH that previously could not be measured by RIA. Using this EIA technique, we evaluated total 24-h urinary hGH excretion in normal children (NC), in normal but short children (NSC), in patients with partial Received July 10,1989. Address requests for reprints to: Dr. Hitoshi Kohno, Department of Endocrinology and Metabolism, Fukuoka Children's Hospital Medical Center, 2-5-1, Tojin-machi, Chuo-ku, Fukuoka 810, Japan. * This work was supported by a grant from the Research Foundation, Fukuoka Children's Hospital Medical Center.
= 21) and CGHD (1.9 ± 0.9 ng/day; range, 0.6-3.6; n = 14) groups from the NSC (12.8 ± 3 . 1 ng/day; range, 9.3-17.5; n = 10) and NC (14.6 ± 3.1 ng/day; range, 10.6-19.0; n = 6) groups without any overlap. A mean urinary hGH value less than 9.0 ng/day during a 2-day collection strongly suggested GH deficiency. Ten of 16 patients with PGHD and CGHD who received 2 IU met-hGH, sc, had urinary hGH levels within the range of the mean ± SD in NSC. These patients received daily sc 0.097 ± 0.024 IU/kg hGH injections. These results suggest that the measurement of 24-h urinary hGH excretion is noninvasive, accurate, and useful for the screening of GH deficiency. The mean value on 2 days of 24-h urinary hGH excretion for the screening of GH deficiency is estimated to be less than 9.0 ng/day. The optimal dose of GH as therapy for GH deficiency is demonstrated as daily sc injection of 0.1 IU/kg hGH, 0.7 IU/kg/week. To convert international units of met-hGH to milligrams, divide by 2.4. (J Clin Endocrinol Metab 7 1 : 1496-1500, 1990)
GH deficiency (PGHD), and in patients with complete GH deficiency (CGHD). Some investigators have reported that urinary hGH excretion is significantly lower in patients with GH deficiency than in either NC or NSC, but an overlap between these groups can not be excluded (3-6). We investigated whether the measurement of 24-h urinary hGH excretion can clearly separate the PGHD and CGHD groups from the NC and NSC groups. Furthermore, in order to evaluate the optimal dose of GH for therapy, we measured the daily variations in urinary hGH excretion after im and sc administration of synthetic methionyl hGH (met-hGH) in GH-deficient children.
Subjects and Methods Subjects Six normal children (4 males and 2 females, aged 6.6-10.9 yr) served as the controls. Ten NSC (9 males and 1 female,
1496
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 14:04 For personal use only. No other uses without permission. . All rights reserved.
URINARY hGH MEASUREMENT IN GHD aged 6.8-14.3 yr; height SD score, -2.1 to -2.4), 21 patients with PGHD (15 males and 6 females, aged 5.6-12.9 yr; height SD score, -2.2 to -3.4), and 14 patients with CGHD (9 males and 5 females, aged 5.8-14.9 yr; height SD score, —2.5 to —3.9) were studied. All children were prepubertal according to Tanner's pubertal stage. The groups of NSC, PGHD, and CGHD were more than 2 SD below the mean height for their ages based upon the growth charts produced by the Japanese Ministry of Health and Welfare. None of the children had renal disease, and they were all euthyroid. The NSC group had a normal GH response to 2 or more GH provocative tests, with peak GH levels above 10 ng/mL. They had no clinical and labolatory abnormalities except for a short stature. The PGHD group had peak GH levels between 7 and 10 ng/mL, and the CGHD group had peak GH levels up to 7 ng/mL. Methods GH secretion was studied after overnight fasting in all patients by insulin hypoglycemia (0.1 U/kg), arginine infusion (0.5 g/kg), and/or L-dopa (10 mg/kg). To evaluate the optimal dose of met-hGH as a treatment for GH deficiency, the daily variations in urinary hGH excretion after administration of met-hGH were examined for 3 days after 2-day urine collection on a hGH-free day. Met-hGH (0.5 IU/kg/wk) was administered in variously divided doses as follows. Patients did not receive more than 0.5 IU/kg/wk methGH in an examination week. Met-hGH was injected in PGHD and CGHD either im or sc. In the group receiving im injections (n = 19; 16.1-26.0 kg), 4 IU hGH were administered every 2 days. In the group receiving sc injections (n = 16; 15.2-46.3 kg), daily administration of hGH was performed with a dosage of 4 IU on the first day and 2 IU on the following 2 days. To
convert international units of met-hGH to milligrams, divide by 2.4. All subjects collected urine for 24 h on admission, which was kept at room temperature with 0.1% NaN3 and 0.1% BSA. Neither diet nor activity was restricted during the urine collection period. Five milliliters of the 24-h urine samples were dialyzed against 0.01 mol/L sodium phosphate buffer, pH 7.0, containing 0.1 mol/L NaCl and 0.1% NaN3 at 4 C for 24 h. The dialyzed urine was kept at 4 C until assay, which was performed within a week after dialysis. The stability of hGH in urine before and after the dialysis was studied at various stock conditions, such as at room temperature, 4 C, and -20 C. Urinary GH concentrations were measured in duplicate by the highly sensitive EIA method described by Hashida et al. (1, 2). Dialyzed urine samples or GH standard were incubated with monoclonal anti-hGH IgG-coated polystyrene balls at 37 C for 6 h. The polystyrene balls were then incubated with affinitypurified rabbit anti-hGH Fab'-horseradish peroxidase conjugate at 4 C for 16 h and at 20 C for 6 h. Peroxidase activity bound to the balls was assayed by fluorimetry with 3-(4-hydroxyphenyl)propionic acid as the substrate. The intra- and interassay coefficients of variation were 5.8% and 8.9%, respectively. The detection limits of hGH in serum and urine were 1.5 pg/mL with 20 ^L serum and 0.2 pg/mL with 0.15 mL urine, respectively. Urinary hGH was expressed in terms of nanograms per day and nanograms per g creatinine. To convert
1497
urinary hGH values to nanograms per mmol creatinine, divide by 8.84. Serum GH was measured in duplicate by a RIA kit (Dinabot, Tokyo, Japan). Urinary creatinine was measured by a conventional kit (Wako Pure Chemical Industries, Ltd., Osaka, Japan). The results are given as the mean ± SD. Student's t test was used for statistical analyses. Results hGH in urine stored at - 2 0 C with 0.1% NaN3 and 0.1% BSA before dialysis was stable for at least 2 months. When the urine was dialyzed and stored at 4 C, the hGH in urine was stable for 7 months. The hGH levels in urine collected with 0.1% NaN3 and 0.1% BSA at 4 C and at room temperature for 24 h were similar. The 24-h urinary hGH excretion varied during 2 consecutive days in the four groups of children (Fig. 1). During the 2 days, the mean (±SD) lower values for 24h urinary hGH excretion were 11.9 ± 2.4 ng/day (range, 8.8-15.2; n = 6) in NC, 10.5 ± 2.6 ng/day (range, 7.214.6; n = 10) in NSC, 4.7 ± 2.0 ng/day (range, 1.1-8.1; n = 21) in PGHD, and 1.4 ± 0.8 ng/day (range, 0.5-3.0; n = 14) in CGHD. The mean higher values were 17.3 ± 4.0 ng/day (range, 12.2-22.8; n = 6) in NC, 15.5 ± 3.9 ng/ day (range, 9.3-18.0; n = 10) in NSC, 6.3 ± 2.7 ng/day (range, 1.5-11.6; n = 21) in PGHD, and 2.5 ± 1.2 ng/day (range, 0.7-4.8; n = 14) in CGHD. There was no significant difference in urinary hGH excretion between the NC group and the NSC group. Although there was a significant difference (P < 0.01) between the lower values in NSC and the higher values in PGHD, many of the higher PGHD samples overlapped with those of the lower NSC group. In terms of nanograms per g creatinine, the mean lower values in 24-h urinary hGH excretion were 19.4 ± 6.5 ng/g creatinine (range, 13.8-33.0; n = 6) in NC, 21.2 ± 6.9 ng/g creatinine (range, 11.8-33.9; n = 10) in NSC, 16.5 ± 6.1 ng/g creatinine (range, 6.0-32.8; n = 21) in PGHD, and 4.2 ± 2.1 ng/g creatinine (range, 1.7-9.1; n = 14) in CGHD. The mean higher values were 31.7 ± 9.5 ng/g creatinine (range, 20.2-51.0; n = 6) in NC, 29.3 ± 6.6 ng/g creatinine (range, 18.5-48.1; n = 10) in NSC, 20.5 ± 7.9 ng/g creatinine (range, 6.8-39.8; n = 21) in PGHD, and 6.3 ± 2.5 ng/g creatinine (range, 2.8-9.8; n = 14) in CGHD. There was no significant difference in urinary hGH excretion between NC and NSC, and no significant difference between the lower values in NSC and the higher values in PGHD. The mean values for 24-h urinary GH excretion for both days combined (Fig. 2) were 14.6 ± 3.1 ng/day (range, 10.6-19.0; n = 6) in NC, 12.8 ± 3.1 ng/day (range, 9.3-17.5; n = 10) in NSC, 5.5 ± 2.3 ng/day (range, 1.39.2; n = 21) in PGHD, and 1.9 ± 0.9 ng/day (range, 0.6-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 14:04 For personal use only. No other uses without permission. . All rights reserved.
KOHNO, MURAKAMI, AND KODAIRA
1498
JCE & M • 1990 Vol 71 • No 6
20 15.0
-P < 0.001-
•N.S.
•o
I — P < 0.001-| i — P < 0.001 ro
10.0
10
5.0 Low
High
Low
High
PGHD
CGHD
(n = 21)
(n = 14) NSC (n = 10)
NC (n = 6) 42.0 .
CGHD (n = 14)
40.1 • -N.S.
30
30
PGHD (n = 21)
1 ^
-P < 0.001I—P < 0.05 —11
P < 0.001-
o>
20 20 •e x
10
10 Low
High
NC (n = 6)
Low
High
NSC (n = 10)
Low
High
PGHD (n = 2!)
Low
High
CGHD (n = 14)
FlG. 1. Total 24-h urinary hGH excretions examined for 2 days in NC, NSC, PGHD, and CGHD. Twenty-four-hour urinary hGH excretion showed the daily variations in patients with GH deficiency as well as in NC and NSC. There was an overlap between lower levels in NSC and higher levels in PGHD. Large circles and bars show the mean ± SD. To convert urinary hGH values to nanograms per mmol creatinine, divide by 8.84.
3.6; n = 14) in CGHD. There was no significant difference in urinary hGH excretion between NC and NSC. The mean combined values of 24-h urinary hGH excretion in PGHD as well as CGHD were significantly lower than in NSC (P < 0.001), and there was no overlap between the NC and NSC groups and the GH deficiency group. A mean urinary value less than 9.0 ng/day over a 2-day period strongly suggested GH deficiency. When the data were expressed in terms of nanograms per g creatinine for both days combined, significantly higher excretion of urinary hGH was also observed in NSC compared to patients with GH deficiency (NSC vs. PGHD, P < 0.05; NSC vs. CGHD, P < 0.001). However, there was an overlap between NSC and PGHD (Fig. 2). As for the daily variations in urinary hGH excretion
X O
NC (n = 6)
NSC (n = 10)
PGHD (n = 21)
CGHD (n = 14)
FIG. 2. Mean values on 2 days of 24-h urinary hGH excretion in NC, NSC, PGHD, and CGHD. There was no overlap in 24-h urinary hGH excretion between the NC and NSC groups and the GH deficiency group when the data were expressed in terms of nanograms per day. Bars show the mean ± SD. To convert urinary hGH values to nanograms per mmol creatinine, divide by 8.84.
in the group with GH deficiency (PGHD and CGHD) after administration of met-hGH, there were dynamic changes in urinary GH excretion (Table 1). In the group receiving im injections, urinary hGH concentrations were 21.0 ± 11.8 ng/day (range, 9.2-56.6; n = 19) and 24.5 ± 15.7 ng/day (range, 5.7-53.2; n = 19) on the day of 4 IU met-hGH administration. On the hGH-free day, the mean level of urinary hGH was 4.0 ± 2.2 ng/day (range, 0.7-8.6; n = 19), which was comparable to that during the baseline periods. After a sc injection of 4 IU met-hGH, urinary hGH levels increased to 28.9 ±11.2 ng/day (range, 13.9-48.4; n = 16). After a sc injection of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 14:04 For personal use only. No other uses without permission. . All rights reserved.
URINARY hGH MEASUREMENT IN GHD
1499
TABLE 1. Daily variations in 24-h urinary hGH excretion in PGHD and CGHD after im or sc administration of met-hGH for 3 days Days of met-hGH treatment Groups
n
im° sc°
19 16
4.0 ± 3.4 4.6 ± 2.4
im 6 sc*
19 16
13.3 ± 10.4 13.3 ± 6.4
2
-1
1
2
3
3.7 ± 2.5 4.1 ± 2.3
21.0 ± 11.8 28.9 ± 11.2
4.0 ± 2.2 16.1 ± 3.7
24.5 ± 15.7 15.4 ± 8.2
13.5 ± 10.5 12.5 ± 6.3
60.6 ± 37.5 83.4 ± 37.9
14.0 ± 9.5 41.7 ± 16.1
67.7 ± 40.2 36.0 ± 18.5
In the group receiving im injections, 4 IU met-hGH were administered every 2 days. In the group receiving sc injections, daily administration of met-hGH was performed with a dosage of 4 IU on the first day and 2 IU on the following 2 days. The data are expressed as the mean ± SD. To convert urinary hGH values to nanograms per mmol creatinine, divide by 8.84, and to convert international units of met-hGH to mg, divide by 2.4. 0 Nanograms per day. 6 Nanograms per g creatinine.
hGH (IU) 4
I
30
±1 SD of values in the NSC group (Fig. 3). These 10 patients received 0.097 ± 0.024 IU/kg met-hGH.
I
Discussion
03
20
10
- 2 - 1
1
2
3 day
*Dose of GH 0.097 ±0.024 lU/kg/day 0.51-0.85 lU/kg/week FlG. 3. Daily variations in 24-h urinary hGH excretion after sc injections of synthetic met-hGH in 10 of 16 patients with GH deficiency. Their urinary hGH levels were between the range of mean ± SD in NSC on the day with 2 IU of met-hGH (*). They received daily sc injections of 0.097 ± 0.024 IU/kg met-hGH. A screen shows the range of mean ± SD in NSC, and bars represent the mean ± SD. To convert international units of met-hGH to milligrams, divide by 2.4.
2 IU, urinary hGH values decreased to 16.1 ± 3.7 ng/day (range, 8.9-22.8; n = 16) and 15.4 ± 8.2 ng/day (range, 6.0-35.8; n = 16). Urinary hGH levels in 10 of 16 patients on the day of 2 IU met-hGH administration were within
Recently, several investigators, using highly sensitive sandwich EIAs (1-5) and immunoradiometric assays (68), reported that measurements of urinary hGH were useful for the diagnosis of GH deficiency and GH excess. In this study we also documented that urinary hGH excretion was lower in patients with GH deficiency than in children with normal GH secretion (NC and NSC), without any overlap between the two groups in total 24h urinary hGH excretion. Thus, measurements of urinary hGH were useful for the screening of GH deficiency. Mean 24-h urinary hGH excretion over 2 days of less than 9.0 ng/day strongly suggested GH deficiency. Although total 24-h urinary hGH excretion examined for 2 days showed daily variations in both children with normal GH secretion (NC and NSC) and patients with GH deficiency, the mean values on the 2 days of 24-h urinary hGH excretion clearly separated the PGHD and CGHD groups from the NC and NSC groups. A period of 2 days or more for urine collection is necessary to accurately evaluate urinary hGH excretion. As for the daily variations in urinary hGH excretion after administration of met-hGH in the GH deficiency group, there were dynamic changes in urinary hGH excretion. In the group receiving im injections, urinary hGH levels were higher on the day of hGH administration and then returned to the levels during the baseline period on an hGH-free day. These results indicated that met-hGH administered im was excreted in urine within 24 h. A very small amount of the synthetic hGH administered was excreted in the urine. Similar results have been previously reported (3, 6, 9). In the group receiving sc injections, after subtracting basal hGH values, 4 IU met-hGH resulted in twice the urinary excretion of hGH as 2 IU. These results indicate
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 14:04 For personal use only. No other uses without permission. . All rights reserved.
KOHNO, MURAKAMI, AND KODAIRA
1500
that urinary hGH measurements using this EIA method accurately reflect the dose of administered met-hGH. Ten of 16 patients who received daily sc injections had urinary hGH values within the range of the mean ±SD in NSC on the day of administration of 2 IU met-hGH. These patients received 0.097 ± 0.024 IU/kg hGH daily during this period. It has already been established that more frequent injections by the sc route result in higher growth rates compared with the same dose given less frequently or hGH given im (10). Based on these results, daily sc injections of 0.1 IU/kg hGH (0.7 IU/kg-week) are recommended as therapy for GH deficiency. This dose is higher than that we usually use (0.5 IU/kg-week) as therapy for GH deficiency. It has also been reported that the minimum effective dose of hGH is 0.03 IU/kg, three times a week (11), and that marginally greater acceleration of bone age in patients on a higher dose regimen can be observed (12). Further investigations on the effects of different dose regimens on the final outcome of height are urgently needed. Acknowledgments We wish to thank Drs. S. Honda and J. H. Holcombe for helpful comments, and Mr. K. Iwasa for statistical analyses.
References 1. Hashida S, Ishikawa E, Kato Y, Imura H, Mohri Z, Murakami Y. Human growth hormone (hGH) in urine and its correlation to
2.
3.
4. 5. 6. 7. 8. 9. 10. 11.
12.
JCE & M • 1990 Vol 71 • No 6
serum hGH examined by a highly sensitive sandwich enzyme immunoassay. Clin Chim Acta. 1987;162:229-35. Hasida S, Ishikawa E, Mohri Z, Nakanishi T, Noguchi H, Murakami Y. Sensitive sandwich enzyme immunoassay of human growth hormone (hGH) in serum and urine using monoclonal antibodycoated polystyrene balls. Endocrinol Jpn. 1988;35:171-80. Okuno A, Yano K, Itoh Y, Hashida S, Ishikawa E, Mohri Z, Murakami Y. Urinary growth hormone determinations compared with other methods in the assessment of growth hormone secretion. Acta Paediat Scand. 1987;337(Suppl):74-81. Hattori N, Shimatsu A, Yamanaka C, Momomi T, Imura H. Nocturnal urinary growth hormone excretion in children with short stature. Acta Endocrinol (Copenh). 1988;119:113-17. Sukegawa I, Hizuka N, Takano K, et al. Urinary growth hormone (GH) measurements are useful for evaluating endogenous GH secretion. J Clin Endocrinol Metab. 1988;66:1119-23. Albini CH, Quattrin T, Vandllen RL, Macgillivray MH. Quantitation of urinary growth hormone in children with normal and abnormal growth. Pediatr Res. 1988;23:89-92. Girard J, Erb T, Pampalone A, Eberle AN, Baumann JB. Growth hormone in urine: development of an ultrasensitive assay applicable to plasma and urine. Horm Res. 1987;28:71-80. Erb T, Karolyi G, Pampalone A, et al. Human growth hormone in urine: development of an ultrasensitive radiometric assay. Experientia. 1988;44:155-7. Hanssen KF. Immunoreactive growth hormone in human urine. Acta Endocrinol (Copenh). 1972;71:665-76. Albertsson-Wikland K, Westphal O, Westgren U. Daily subcutaneous administration of human growth hormone in growth hormone deficient children. Acta Paediatr Scand. 1986;75:89-97. Frasier SD, Aceto T, Hayles AB, Mikity VG. Collaborative study of the effects of human growth hormone in growth hormone deficiency. IV. Treatment with low doses of human growth hormone based on body weight. J Clin Endocrinol Metab. 1977;44:22-31. Preece MA, Tanner JM, Whitehouse RH, Cameron N. Dose dependence of growth response to human growth hormone in growth hormone deficiency. J Clin Endocrinol Metab. 1976;42:477-83.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 14:04 For personal use only. No other uses without permission. . All rights reserved.
Vol. 71, No. 6 Printed in U.S.A.
Urinary Human Growth Hormone Measurement Using a Highly Sensitive Sandwich Enzyme Immunoassay: Diagnostic and Therapeutic Uses in Patients with Growth Hormone Deficiency* HITOSHI KOHNO, YOSHIAKI MURAKAMI, AND TSUKASA KODAIRA Department of Endocrinology and Metabolism, Fukuoka Children's Hospital Medical Center (H.K), Fukuoka 810; Research Laboratories, Research and Development Division, Sumitomo Pharmaceuticals Co., Ltd. (Y.M.), Takarazuka, Hyogo 665; and Clinical Testing Servicies, Otsuka Assay Laboratories, Otsuka Pharmaceuticals Co., Ltd. (T.K), Tokushima 771-01, Japan
ABSTRACT. Several reports indicate that urinary hGH excretion is significantly lower in patients with either partial (PGHD) or complete GH deficiency (CGHD) than in normal but short children (NSC) or normal children (NC). However, there is an overlap between the NSC and NC groups and the PGHD group. Using a highly sensitive sandwich enzyme immunoassay, we investigated whether the measurement of urinary hGH can clearly separate the PGHD and CGHD groups from the NSC and NC groups. In addition, we measured the urinary excretion of synthetic methionyl-hGH (met-hGH) in PGHD and CGHD after sc injections of 2 and 4 IU and im injections of 4 IU in an attempt to determine the optimal replacement dose. Total 24-h urinary hGH excretion in each patient examined for 2 consecutive days varied from 1 day to the next. There were no differencies in urinary hGH excretions between the NSC group and the NC group. The lower values for daily urinary hGH excretion in the NSC group overlapped some of the higher values in the PGHD group. However, when the mean urinary hGH level of both days was used, the 24-h urinary hGH excretion clearly separated the PGHD (5.5 ± 2.3 ng/day; range, 1.3-9.2; n
R
ECENTLY, a highly sensitive sandwich enzyme immunoassay (EIA) for human GH (hGH) was developed that used monoclonal anti-hGH immunoglobulin G-coated polystyrene balls and affinity-purified rabbit anti-hGH Fab'-horseradish peroxidase conjugate (1, 2). This EIA method made it possible to measure very low levels of unconcentrated urinary hGH that previously could not be measured by RIA. Using this EIA technique, we evaluated total 24-h urinary hGH excretion in normal children (NC), in normal but short children (NSC), in patients with partial Received July 10,1989. Address requests for reprints to: Dr. Hitoshi Kohno, Department of Endocrinology and Metabolism, Fukuoka Children's Hospital Medical Center, 2-5-1, Tojin-machi, Chuo-ku, Fukuoka 810, Japan. * This work was supported by a grant from the Research Foundation, Fukuoka Children's Hospital Medical Center.
= 21) and CGHD (1.9 ± 0.9 ng/day; range, 0.6-3.6; n = 14) groups from the NSC (12.8 ± 3 . 1 ng/day; range, 9.3-17.5; n = 10) and NC (14.6 ± 3.1 ng/day; range, 10.6-19.0; n = 6) groups without any overlap. A mean urinary hGH value less than 9.0 ng/day during a 2-day collection strongly suggested GH deficiency. Ten of 16 patients with PGHD and CGHD who received 2 IU met-hGH, sc, had urinary hGH levels within the range of the mean ± SD in NSC. These patients received daily sc 0.097 ± 0.024 IU/kg hGH injections. These results suggest that the measurement of 24-h urinary hGH excretion is noninvasive, accurate, and useful for the screening of GH deficiency. The mean value on 2 days of 24-h urinary hGH excretion for the screening of GH deficiency is estimated to be less than 9.0 ng/day. The optimal dose of GH as therapy for GH deficiency is demonstrated as daily sc injection of 0.1 IU/kg hGH, 0.7 IU/kg/week. To convert international units of met-hGH to milligrams, divide by 2.4. (J Clin Endocrinol Metab 7 1 : 1496-1500, 1990)
GH deficiency (PGHD), and in patients with complete GH deficiency (CGHD). Some investigators have reported that urinary hGH excretion is significantly lower in patients with GH deficiency than in either NC or NSC, but an overlap between these groups can not be excluded (3-6). We investigated whether the measurement of 24-h urinary hGH excretion can clearly separate the PGHD and CGHD groups from the NC and NSC groups. Furthermore, in order to evaluate the optimal dose of GH for therapy, we measured the daily variations in urinary hGH excretion after im and sc administration of synthetic methionyl hGH (met-hGH) in GH-deficient children.
Subjects and Methods Subjects Six normal children (4 males and 2 females, aged 6.6-10.9 yr) served as the controls. Ten NSC (9 males and 1 female,
1496
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 14:04 For personal use only. No other uses without permission. . All rights reserved.
URINARY hGH MEASUREMENT IN GHD aged 6.8-14.3 yr; height SD score, -2.1 to -2.4), 21 patients with PGHD (15 males and 6 females, aged 5.6-12.9 yr; height SD score, -2.2 to -3.4), and 14 patients with CGHD (9 males and 5 females, aged 5.8-14.9 yr; height SD score, —2.5 to —3.9) were studied. All children were prepubertal according to Tanner's pubertal stage. The groups of NSC, PGHD, and CGHD were more than 2 SD below the mean height for their ages based upon the growth charts produced by the Japanese Ministry of Health and Welfare. None of the children had renal disease, and they were all euthyroid. The NSC group had a normal GH response to 2 or more GH provocative tests, with peak GH levels above 10 ng/mL. They had no clinical and labolatory abnormalities except for a short stature. The PGHD group had peak GH levels between 7 and 10 ng/mL, and the CGHD group had peak GH levels up to 7 ng/mL. Methods GH secretion was studied after overnight fasting in all patients by insulin hypoglycemia (0.1 U/kg), arginine infusion (0.5 g/kg), and/or L-dopa (10 mg/kg). To evaluate the optimal dose of met-hGH as a treatment for GH deficiency, the daily variations in urinary hGH excretion after administration of met-hGH were examined for 3 days after 2-day urine collection on a hGH-free day. Met-hGH (0.5 IU/kg/wk) was administered in variously divided doses as follows. Patients did not receive more than 0.5 IU/kg/wk methGH in an examination week. Met-hGH was injected in PGHD and CGHD either im or sc. In the group receiving im injections (n = 19; 16.1-26.0 kg), 4 IU hGH were administered every 2 days. In the group receiving sc injections (n = 16; 15.2-46.3 kg), daily administration of hGH was performed with a dosage of 4 IU on the first day and 2 IU on the following 2 days. To
convert international units of met-hGH to milligrams, divide by 2.4. All subjects collected urine for 24 h on admission, which was kept at room temperature with 0.1% NaN3 and 0.1% BSA. Neither diet nor activity was restricted during the urine collection period. Five milliliters of the 24-h urine samples were dialyzed against 0.01 mol/L sodium phosphate buffer, pH 7.0, containing 0.1 mol/L NaCl and 0.1% NaN3 at 4 C for 24 h. The dialyzed urine was kept at 4 C until assay, which was performed within a week after dialysis. The stability of hGH in urine before and after the dialysis was studied at various stock conditions, such as at room temperature, 4 C, and -20 C. Urinary GH concentrations were measured in duplicate by the highly sensitive EIA method described by Hashida et al. (1, 2). Dialyzed urine samples or GH standard were incubated with monoclonal anti-hGH IgG-coated polystyrene balls at 37 C for 6 h. The polystyrene balls were then incubated with affinitypurified rabbit anti-hGH Fab'-horseradish peroxidase conjugate at 4 C for 16 h and at 20 C for 6 h. Peroxidase activity bound to the balls was assayed by fluorimetry with 3-(4-hydroxyphenyl)propionic acid as the substrate. The intra- and interassay coefficients of variation were 5.8% and 8.9%, respectively. The detection limits of hGH in serum and urine were 1.5 pg/mL with 20 ^L serum and 0.2 pg/mL with 0.15 mL urine, respectively. Urinary hGH was expressed in terms of nanograms per day and nanograms per g creatinine. To convert
1497
urinary hGH values to nanograms per mmol creatinine, divide by 8.84. Serum GH was measured in duplicate by a RIA kit (Dinabot, Tokyo, Japan). Urinary creatinine was measured by a conventional kit (Wako Pure Chemical Industries, Ltd., Osaka, Japan). The results are given as the mean ± SD. Student's t test was used for statistical analyses. Results hGH in urine stored at - 2 0 C with 0.1% NaN3 and 0.1% BSA before dialysis was stable for at least 2 months. When the urine was dialyzed and stored at 4 C, the hGH in urine was stable for 7 months. The hGH levels in urine collected with 0.1% NaN3 and 0.1% BSA at 4 C and at room temperature for 24 h were similar. The 24-h urinary hGH excretion varied during 2 consecutive days in the four groups of children (Fig. 1). During the 2 days, the mean (±SD) lower values for 24h urinary hGH excretion were 11.9 ± 2.4 ng/day (range, 8.8-15.2; n = 6) in NC, 10.5 ± 2.6 ng/day (range, 7.214.6; n = 10) in NSC, 4.7 ± 2.0 ng/day (range, 1.1-8.1; n = 21) in PGHD, and 1.4 ± 0.8 ng/day (range, 0.5-3.0; n = 14) in CGHD. The mean higher values were 17.3 ± 4.0 ng/day (range, 12.2-22.8; n = 6) in NC, 15.5 ± 3.9 ng/ day (range, 9.3-18.0; n = 10) in NSC, 6.3 ± 2.7 ng/day (range, 1.5-11.6; n = 21) in PGHD, and 2.5 ± 1.2 ng/day (range, 0.7-4.8; n = 14) in CGHD. There was no significant difference in urinary hGH excretion between the NC group and the NSC group. Although there was a significant difference (P < 0.01) between the lower values in NSC and the higher values in PGHD, many of the higher PGHD samples overlapped with those of the lower NSC group. In terms of nanograms per g creatinine, the mean lower values in 24-h urinary hGH excretion were 19.4 ± 6.5 ng/g creatinine (range, 13.8-33.0; n = 6) in NC, 21.2 ± 6.9 ng/g creatinine (range, 11.8-33.9; n = 10) in NSC, 16.5 ± 6.1 ng/g creatinine (range, 6.0-32.8; n = 21) in PGHD, and 4.2 ± 2.1 ng/g creatinine (range, 1.7-9.1; n = 14) in CGHD. The mean higher values were 31.7 ± 9.5 ng/g creatinine (range, 20.2-51.0; n = 6) in NC, 29.3 ± 6.6 ng/g creatinine (range, 18.5-48.1; n = 10) in NSC, 20.5 ± 7.9 ng/g creatinine (range, 6.8-39.8; n = 21) in PGHD, and 6.3 ± 2.5 ng/g creatinine (range, 2.8-9.8; n = 14) in CGHD. There was no significant difference in urinary hGH excretion between NC and NSC, and no significant difference between the lower values in NSC and the higher values in PGHD. The mean values for 24-h urinary GH excretion for both days combined (Fig. 2) were 14.6 ± 3.1 ng/day (range, 10.6-19.0; n = 6) in NC, 12.8 ± 3.1 ng/day (range, 9.3-17.5; n = 10) in NSC, 5.5 ± 2.3 ng/day (range, 1.39.2; n = 21) in PGHD, and 1.9 ± 0.9 ng/day (range, 0.6-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 14:04 For personal use only. No other uses without permission. . All rights reserved.
KOHNO, MURAKAMI, AND KODAIRA
1498
JCE & M • 1990 Vol 71 • No 6
20 15.0
-P < 0.001-
•N.S.
•o
I — P < 0.001-| i — P < 0.001 ro
10.0
10
5.0 Low
High
Low
High
PGHD
CGHD
(n = 21)
(n = 14) NSC (n = 10)
NC (n = 6) 42.0 .
CGHD (n = 14)
40.1 • -N.S.
30
30
PGHD (n = 21)
1 ^
-P < 0.001I—P < 0.05 —11
P < 0.001-
o>
20 20 •e x
10
10 Low
High
NC (n = 6)
Low
High
NSC (n = 10)
Low
High
PGHD (n = 2!)
Low
High
CGHD (n = 14)
FlG. 1. Total 24-h urinary hGH excretions examined for 2 days in NC, NSC, PGHD, and CGHD. Twenty-four-hour urinary hGH excretion showed the daily variations in patients with GH deficiency as well as in NC and NSC. There was an overlap between lower levels in NSC and higher levels in PGHD. Large circles and bars show the mean ± SD. To convert urinary hGH values to nanograms per mmol creatinine, divide by 8.84.
3.6; n = 14) in CGHD. There was no significant difference in urinary hGH excretion between NC and NSC. The mean combined values of 24-h urinary hGH excretion in PGHD as well as CGHD were significantly lower than in NSC (P < 0.001), and there was no overlap between the NC and NSC groups and the GH deficiency group. A mean urinary value less than 9.0 ng/day over a 2-day period strongly suggested GH deficiency. When the data were expressed in terms of nanograms per g creatinine for both days combined, significantly higher excretion of urinary hGH was also observed in NSC compared to patients with GH deficiency (NSC vs. PGHD, P < 0.05; NSC vs. CGHD, P < 0.001). However, there was an overlap between NSC and PGHD (Fig. 2). As for the daily variations in urinary hGH excretion
X O
NC (n = 6)
NSC (n = 10)
PGHD (n = 21)
CGHD (n = 14)
FIG. 2. Mean values on 2 days of 24-h urinary hGH excretion in NC, NSC, PGHD, and CGHD. There was no overlap in 24-h urinary hGH excretion between the NC and NSC groups and the GH deficiency group when the data were expressed in terms of nanograms per day. Bars show the mean ± SD. To convert urinary hGH values to nanograms per mmol creatinine, divide by 8.84.
in the group with GH deficiency (PGHD and CGHD) after administration of met-hGH, there were dynamic changes in urinary GH excretion (Table 1). In the group receiving im injections, urinary hGH concentrations were 21.0 ± 11.8 ng/day (range, 9.2-56.6; n = 19) and 24.5 ± 15.7 ng/day (range, 5.7-53.2; n = 19) on the day of 4 IU met-hGH administration. On the hGH-free day, the mean level of urinary hGH was 4.0 ± 2.2 ng/day (range, 0.7-8.6; n = 19), which was comparable to that during the baseline periods. After a sc injection of 4 IU met-hGH, urinary hGH levels increased to 28.9 ±11.2 ng/day (range, 13.9-48.4; n = 16). After a sc injection of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 14:04 For personal use only. No other uses without permission. . All rights reserved.
URINARY hGH MEASUREMENT IN GHD
1499
TABLE 1. Daily variations in 24-h urinary hGH excretion in PGHD and CGHD after im or sc administration of met-hGH for 3 days Days of met-hGH treatment Groups
n
im° sc°
19 16
4.0 ± 3.4 4.6 ± 2.4
im 6 sc*
19 16
13.3 ± 10.4 13.3 ± 6.4
2
-1
1
2
3
3.7 ± 2.5 4.1 ± 2.3
21.0 ± 11.8 28.9 ± 11.2
4.0 ± 2.2 16.1 ± 3.7
24.5 ± 15.7 15.4 ± 8.2
13.5 ± 10.5 12.5 ± 6.3
60.6 ± 37.5 83.4 ± 37.9
14.0 ± 9.5 41.7 ± 16.1
67.7 ± 40.2 36.0 ± 18.5
In the group receiving im injections, 4 IU met-hGH were administered every 2 days. In the group receiving sc injections, daily administration of met-hGH was performed with a dosage of 4 IU on the first day and 2 IU on the following 2 days. The data are expressed as the mean ± SD. To convert urinary hGH values to nanograms per mmol creatinine, divide by 8.84, and to convert international units of met-hGH to mg, divide by 2.4. 0 Nanograms per day. 6 Nanograms per g creatinine.
hGH (IU) 4
I
30
±1 SD of values in the NSC group (Fig. 3). These 10 patients received 0.097 ± 0.024 IU/kg met-hGH.
I
Discussion
03
20
10
- 2 - 1
1
2
3 day
*Dose of GH 0.097 ±0.024 lU/kg/day 0.51-0.85 lU/kg/week FlG. 3. Daily variations in 24-h urinary hGH excretion after sc injections of synthetic met-hGH in 10 of 16 patients with GH deficiency. Their urinary hGH levels were between the range of mean ± SD in NSC on the day with 2 IU of met-hGH (*). They received daily sc injections of 0.097 ± 0.024 IU/kg met-hGH. A screen shows the range of mean ± SD in NSC, and bars represent the mean ± SD. To convert international units of met-hGH to milligrams, divide by 2.4.
2 IU, urinary hGH values decreased to 16.1 ± 3.7 ng/day (range, 8.9-22.8; n = 16) and 15.4 ± 8.2 ng/day (range, 6.0-35.8; n = 16). Urinary hGH levels in 10 of 16 patients on the day of 2 IU met-hGH administration were within
Recently, several investigators, using highly sensitive sandwich EIAs (1-5) and immunoradiometric assays (68), reported that measurements of urinary hGH were useful for the diagnosis of GH deficiency and GH excess. In this study we also documented that urinary hGH excretion was lower in patients with GH deficiency than in children with normal GH secretion (NC and NSC), without any overlap between the two groups in total 24h urinary hGH excretion. Thus, measurements of urinary hGH were useful for the screening of GH deficiency. Mean 24-h urinary hGH excretion over 2 days of less than 9.0 ng/day strongly suggested GH deficiency. Although total 24-h urinary hGH excretion examined for 2 days showed daily variations in both children with normal GH secretion (NC and NSC) and patients with GH deficiency, the mean values on the 2 days of 24-h urinary hGH excretion clearly separated the PGHD and CGHD groups from the NC and NSC groups. A period of 2 days or more for urine collection is necessary to accurately evaluate urinary hGH excretion. As for the daily variations in urinary hGH excretion after administration of met-hGH in the GH deficiency group, there were dynamic changes in urinary hGH excretion. In the group receiving im injections, urinary hGH levels were higher on the day of hGH administration and then returned to the levels during the baseline period on an hGH-free day. These results indicated that met-hGH administered im was excreted in urine within 24 h. A very small amount of the synthetic hGH administered was excreted in the urine. Similar results have been previously reported (3, 6, 9). In the group receiving sc injections, after subtracting basal hGH values, 4 IU met-hGH resulted in twice the urinary excretion of hGH as 2 IU. These results indicate
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 14:04 For personal use only. No other uses without permission. . All rights reserved.
KOHNO, MURAKAMI, AND KODAIRA
1500
that urinary hGH measurements using this EIA method accurately reflect the dose of administered met-hGH. Ten of 16 patients who received daily sc injections had urinary hGH values within the range of the mean ±SD in NSC on the day of administration of 2 IU met-hGH. These patients received 0.097 ± 0.024 IU/kg hGH daily during this period. It has already been established that more frequent injections by the sc route result in higher growth rates compared with the same dose given less frequently or hGH given im (10). Based on these results, daily sc injections of 0.1 IU/kg hGH (0.7 IU/kg-week) are recommended as therapy for GH deficiency. This dose is higher than that we usually use (0.5 IU/kg-week) as therapy for GH deficiency. It has also been reported that the minimum effective dose of hGH is 0.03 IU/kg, three times a week (11), and that marginally greater acceleration of bone age in patients on a higher dose regimen can be observed (12). Further investigations on the effects of different dose regimens on the final outcome of height are urgently needed. Acknowledgments We wish to thank Drs. S. Honda and J. H. Holcombe for helpful comments, and Mr. K. Iwasa for statistical analyses.
References 1. Hashida S, Ishikawa E, Kato Y, Imura H, Mohri Z, Murakami Y. Human growth hormone (hGH) in urine and its correlation to
2.
3.
4. 5. 6. 7. 8. 9. 10. 11.
12.
JCE & M • 1990 Vol 71 • No 6
serum hGH examined by a highly sensitive sandwich enzyme immunoassay. Clin Chim Acta. 1987;162:229-35. Hasida S, Ishikawa E, Mohri Z, Nakanishi T, Noguchi H, Murakami Y. Sensitive sandwich enzyme immunoassay of human growth hormone (hGH) in serum and urine using monoclonal antibodycoated polystyrene balls. Endocrinol Jpn. 1988;35:171-80. Okuno A, Yano K, Itoh Y, Hashida S, Ishikawa E, Mohri Z, Murakami Y. Urinary growth hormone determinations compared with other methods in the assessment of growth hormone secretion. Acta Paediat Scand. 1987;337(Suppl):74-81. Hattori N, Shimatsu A, Yamanaka C, Momomi T, Imura H. Nocturnal urinary growth hormone excretion in children with short stature. Acta Endocrinol (Copenh). 1988;119:113-17. Sukegawa I, Hizuka N, Takano K, et al. Urinary growth hormone (GH) measurements are useful for evaluating endogenous GH secretion. J Clin Endocrinol Metab. 1988;66:1119-23. Albini CH, Quattrin T, Vandllen RL, Macgillivray MH. Quantitation of urinary growth hormone in children with normal and abnormal growth. Pediatr Res. 1988;23:89-92. Girard J, Erb T, Pampalone A, Eberle AN, Baumann JB. Growth hormone in urine: development of an ultrasensitive assay applicable to plasma and urine. Horm Res. 1987;28:71-80. Erb T, Karolyi G, Pampalone A, et al. Human growth hormone in urine: development of an ultrasensitive radiometric assay. Experientia. 1988;44:155-7. Hanssen KF. Immunoreactive growth hormone in human urine. Acta Endocrinol (Copenh). 1972;71:665-76. Albertsson-Wikland K, Westphal O, Westgren U. Daily subcutaneous administration of human growth hormone in growth hormone deficient children. Acta Paediatr Scand. 1986;75:89-97. Frasier SD, Aceto T, Hayles AB, Mikity VG. Collaborative study of the effects of human growth hormone in growth hormone deficiency. IV. Treatment with low doses of human growth hormone based on body weight. J Clin Endocrinol Metab. 1977;44:22-31. Preece MA, Tanner JM, Whitehouse RH, Cameron N. Dose dependence of growth response to human growth hormone in growth hormone deficiency. J Clin Endocrinol Metab. 1976;42:477-83.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 14:04 For personal use only. No other uses without permission. . All rights reserved.
