THE BIOCHEMICAL DIAGNOSIS of GH deficiency

Transcription

1 X/00/$03.00/0 Vol. 85, No. 11 The Journal of Clinical Endocrinology & Metabolism Printed in U.S.A. Copyright 2000 by The Endocrine Society Estrogen Priming Effect on Growth Hormone (GH) Provocative Test: A Useful Tool for the Diagnosis of GH Deficiency* ALICIA S. MARTÍNEZ, HORACIO M. DOMENÉ, M. GABRIELA ROPELATO, HÉCTOR G. JASPER, PATRICIA A. PENNISI, MARÍA E. ESCOBAR, AND JUAN J. HEINRICH División de Endocrinología, Centro de Investigaciones Endocrinológicas, Hospital de Niños R. Gutiérrez, Buenos Aires, Argentina ABSTRACT We have studied the effect of estradiol (E 2 )on the GH-insulin-like growth factor (GH-IGF) axis in 15 prepubertal GH deficiency (GHD) children and 44 prepubertal or early pubertal children with idiopathic short stature (SS). All of them received a daily dose of micronized E 2 (1 or 2 mg) or placebo, for 3 days, before a sequential arginineclonidine test. In SS children, GH maximal responses were on placebo and g/l on estrogen (P ). The lower 95% confidence limits for GH maximal response changed from 3.7 g/l (without E 2 )to8.3 g/l (on E 2 ). In GHD children, no significant stimulatory effect of estrogen on GH levels was observed. After placebo, a cut-off limit of 3.7 g/l (the lower 95% confidence interval limit) resulted in 73% sensitivity, 95% specificity, and an overall 90% diagnostic efficiency. After E 2, a cut-off limit of 8.3 g/l resulted in a sensitivity of 87%, a specificity of 98%, and a diagnostic efficiency of 95%. After placebo, 68% of SS showed normal IGF-I levels, and the mean Received March 1, Revision received May 10, Accepted July 20, Address all correspondence and requests for reprints to: Dra. Alicia Martínez, División de Endocrinología, Hospital de Niños R. Gutiérrez, Gallo 1360 (1425), Buenos Aires, Argentina. cedie@cedie. guti.gov.ar. * Supported by Grant BID/OC-AR PICT/ , Consejo Nacional de Investigaciones Científicas y Técnicas, CONICET. did not change on E 2 ( vs nmol/l, not significant). In 93% of SS, IGF binding protein (IGFBP)-3 levels were normal during placebo. On E 2, mean IGFBP-3 did not change ( vs mg/l, not significant). In 14 of 15 GHD patients, IGF-I values were below normal on placebo, and the mean of the group did not change after E 2. During placebo, 13 of 15 GHD children presented low IGFBP-3 values. During E 2, there was a small significant increase in IGFBP-3 values ( vs mg/l, P 0.02). The highest diagnostic efficiencies for IGF-I and IGFBP-3 were observed during placebo (75% and 91%, respectively). We conclude that GH stimulation tests after E 2 priming had the highest diagnostic efficiency. Our findings suggest that the effect of estrogen priming on GH stimulated levels, by reducing the number of false nonresponders, might be useful to better discriminate between normal and abnormal GH status in SS children. (J Clin Endocrinol Metab 85: , 2000) THE BIOCHEMICAL DIAGNOSIS of GH deficiency (GHD) has traditionally been based on provocative tests using a variety of GH stimulation agents. However, the criteria to define a response level to exclude the GHD diagnosis have relied more on arbitrarily defined GH levels than on normative data (1). Estrogen administration is known to increase GH responsiveness to provocative stimuli (2, 3). Although estrogen priming has been proposed to improve the GH response, there are no established criteria for either the estrogen administration schedule or the GH cut-off level. It has also been observed (4 7) that some children, initially diagnosed as having GHD when prepubertal, had normal GH responses when retested in adult life. This could be attributable, at least in part, to the lack of sex steroids at the time of initial testing. Thus, estrogen priming might reduce the percentage of falsepositive GHD diagnosis in prepubertal and in early pubertal subjects. Although several reports have described the normalization of the GH response to provocative tests after the administration of estrogen or during spontaneous puberty (8, 9), to our knowledge, there are few published data on the effect of estrogen priming in GH-deficient children. The aim of this double-blind placebo-controlled study was to evaluate the effect of estrogen administration on GH stimulation tests in both short normal and GHD patients and to compare the diagnostic efficiency of this approach with that of serum levels of insulin-like growth factor (IGF)-I and IGF binding protein (IGFBP)-3. Subjects and Methods The same pediatric endocrinologist evaluated all patients. Because clinical evaluation is an important aspect in the diagnosis of GHD, established clinical and laboratory criteria were used to assign short children to diagnosis of either GHD- or non-gh-deficient short stature (SS) (10, 11). This assignment was made according to the clinical and auxological evaluation, previous to test GH reserve. This schedule allows us to evaluate the diagnostic efficiency of GH provocative test under placebo or estradiol (E 2 ), independently of the results of GH response. We studied 15 prepubertal short children highly suspicious of GHD based on: 1) severe growth retardation; 2) the presence of typical phenotypic features (frontal bossing, immature face, midfacial hypoplasia, truncal adiposity, fat dimpling, male hypogenitalism, and high-pitched voice); 3) the exclusion of other clinical or endocrinological causes of growth retardation; 4) radiological findings in the magnetic resonance imaging; and 5) association of other hormonal deficiencies. This group consisted of 9 girls and 6 boys, yr old (8 with isolated GH 4168

2 ESTRADIOL PRIMING IN GHD DIAGNOSIS 4169 deficiency and 7 with multiple pituitary hormone deficiencies), under adequate replacement therapy when indicated (Table 1). We also studied 44 healthy short normal children (10 girls and 34 boys), between yr old, in whom an evaluation to rule out clinical or endocrinological causes of SS was performed, including blood count and thyroid, liver, and kidney function tests. Celiac disease was also excluded. All of them were short (height below 2.0 sd), with normal weight for height, and prepubertal or in early puberty (Tanner stage 2) (Table 1). Heights were expressed in sds, according to age and sex, compared with Argentinean references values (12). Pubertal stages were classified according to Tanner (13, 14). Bone age was assessed by the method of Greulich and Pyle (15). A single daily dose of micronized E 2 valerate, po (1 mg for children weighing up to 20 kg and 2 mg for those weighing more than 20 kg), or placebo was administered, at night, to each subject, for 3 days, preceding the GH stimulation test. Four weeks later, they were crossed over to the opposite treatment and retested. Because the sequence of administration was random and double-blind, some children received placebo initially, whereas others received E 2 first; and both patients and staff members were blind, regarding the order of E 2 or placebo administration. A sequential arginine (0.5 g/kg)-clonidine (100 g/m 2 ) test was performed on each occasion after basal blood samples had been obtained for 17 -E 2, IGF-I, and IGFBP-3. GH was measured in plasma samples obtained at 30, 0, 30, 45, 60, and 90 min after arginine infusion and at 30, 45, 60, 90, and 120 min after clonidine administration, by a polyclonal double-antibody RIA (anti-gh antibody, kindly supplied by National Hormone and Pituitary Program, NIH, Bethesda, MD). The World Health Organization International Reference preparation of hgh was used as standard. Sensitivity of the GH assay was 1 g/l. The intra- and interassay coefficients of variation were lower than 5% and 8.7%, respectively. E 2 was measured by RIA using a commercial kit (Estradiolo-6 DPC, Los Angeles, CA) with a sensitivity of 37 pmol/l. To assess compliance, basal E 2 levels were measured after both placebo and E 2 administration. Only the children who reached E 2 levels above 110 pmol/l were included in the study. Total IGF-I levels were determined in the samples, by RIA, after serum extraction by the acid-ethanol method (16). For assay purposes, human recombinant IGF-I (Bachem California, Inc., Torrance, CA) was used as standard and tracer. The hormone was iodinated with 125 Iby means of the chloramine-t method and purified on a Sephadex G-50 column. The rabbit polyclonal anti-igf-i antibody was a generous gift from Dr. L. E. Underwood (University of North Carolina at Chapel Hill). The intra- and interassay coefficients of variation were 4.7% and 14.2%, respectively. Under our working conditions, serial dilutions of rhigf-i standard and extracted serum gave parallel displacement curves. IGFBP-3 was measured by immunoradiometric assay using a commercial kit (DSL, Webster, TX). Intra- and interassay coefficients of variation were 5.4% and 7.2%, respectively. All samples from each child were analyzed in the same assay. Statistical Analysis The GH peak (GHmax) was defined as the highest value during the test. Results are expressed as the mean sd. The 95% confidence limits for the mean peak GH levels were estimated after logarithmic transformation of the data. Area under the curve was obtained by using the Prism Version 3.0 program (GraphPad Software, Inc., San Diego, CA); the 95% confidence limits were estimated after logarithmic transformation of the data. Statistical analyses were performed using a two-tailed paired or unpaired Student s t test, where appropriate. Diagnostic sensitivity was calculated as the number of true-positive results in GHD patients divided by the total number of patients of this group. Diagnostic specificity was calculated as the number of true negative results in SS children divided by the total number of SS subjects. Diagnostic efficiency was calculated as the number of correct results divided by the total number of children of both groups. The Scientific Committee of the Ricardo Gutiérrez Children s Hospital approved the study, and informed written parental consent was obtained. Results E 2 levels After placebo administration, levels of E 2 were similar in both diagnostic groups. After estrogen priming, E 2 levels increased in short normal children from to pmol/l (P ) and in GHD patients from 37 4to pmol/l (P ). The magnitude of the increment was not statistically different between the two groups. Only four subjects, who did not reach E 2 levels above 110 pmol/l, were excluded. All of them had GHmax response greater than 9.0 g/l under placebo, showing no further increase under E 2. GH response Mean basal GH levels rose from (placebo) to g/l (E 2 )(P 0.002) in SS children but remained the same [ and g/l, not significant (NS)] in the GHD subjects. In short normal children, mean GH maximal responses were g/l after placebo and g/l after administration of estrogen (P ). The lower 95% confidence limit for GHmax increased from 3.7 g/l (after placebo) to 8.3 g/l (on E 2 ), Fig. 1. In GHD children, no statistically significant stimulatory effect of estrogen pretreatment on GH levels was observed ( g/l without and g/l on E 2, NS). With placebo administration, 4 of 15 GHD patients overlapped the SS 95% confidence limits; and after E 2 priming, only 2 of 15 showed a GHmax slightly above the lower 95% confidence limit. In the GHD group, the two patients with GHmax response under E 2 slightly above the lower 95% confident limit were two children presenting isolated GHD with severe growth retardation (height at 3.4 and 3.8 SDS, respectively). Under placebo conditions, a cut-off limit of 3.7 g/l, the lower 95% confidence interval limit, resulted in 73% sensitivity, 95% specificity, and an overall 90% diagnostic efficiency. An arbitrary discriminator of 5.0 g/l slightly increased the sensitivity to 87%, leaving unaffected, at 90%, the diagnostic efficiency. Under E 2 conditions, a cut-off limit of 8.3 g/l resulted in a sensitivity of 87%, a specificity of 98%, and a diagnostic efficiency of 95%. Using 9.0 g/l as an arbitrary cut-off, sensitivity and specificity were 100 and TABLE 1. Clinical data of the patients Patients n Sex CA (yr) BA (yr) Height SDS Pubertal stage GHD 15 9 F, 6 M ( ) All Prepubertal ( ) ( ) ( 0.4 to 4.8) SS F, 34 M ( ) 37 Prepubertal (5 17.3) ( ) ( 2.0 to 5.4) 6 G2, 1 M3 Mean SD; (), Range; CA, ; BA, ; G2, ; MA,.

3 4170 MARTÍNEZ ET AL. JCE&M 2000 Vol. 85 No %, respectively, with a diagnostic efficiency of 98% (Table 2). By applying this arbitrary cut-off limit of 9.0 g/l, close to the most widely accepted limit of 10 g/l used to exclude GHD, only 1 out of 10 SS children under 8 yr of age failed to respond in the test without priming. Of the remaining subjects, 5 of 15 between 8 12 yr of age and 3 of 19 older than 12 yr also showed no adequate response. Under priming conditions, only 1 SS child (the oldest patient, a boy of 17.3 yr) had a GH response below 9 g/l. The area under the curve (AUC) increased from to g/l min in SS (P ) and from to g/l min in GHD (P 0.05). The 95% confidence limits of the AUC changed, after E 2, from to g/l min in SS. The power of AUC to discriminate both groups improved under priming; however, it was still less effective than GHmax values (Table 2). Serum IGF-I and IGFBP-3 Under placebo, 30 out of 44 (68%) SS children showed normal serum IGF-I levels. No child showed IGF-I levels above normal limits. Although after E 2 administration, 5 previously normal IGF-I levels became low, and three previously low became normal, mean IGF-I levels did not change significantly ( vs nmol/l, NS, Fig. 2A). In 38 out of 41 (93%) SS patients, IGFBP-3 levels were normal under placebo. No patient presented high IGFBP-3 levels. Although under E 2, two previously normal IGFBP-3 levels became low, and 1 previously low became normal, mean IGFBP3 levels did not change significantly ( vs mg/l, NS, Fig. 2B). In GHD patients on placebo, 14 out of 15 (93%) showed serum IGF-I values below normal levels. Neither the number of patients with low values nor the mean IGF-1 level of the group changed under E 2 ( vs nmol/l, NS, Fig. 2A). Under placebo, 13 out of 15 (87%) GHD children presented low IGFBP-3 values. Under E 2, 2 previously low IGFBP-3 values became normal; besides, since 10 out of 15 GHD patients increased their levels, there was a significant [albeit small (13%)] increase in IGFBP-3 values ( vs mg/l, P 0.02, Fig. 2B). Thus, IGF-I and IGFBP-3, under placebo, had diagnostic efficiencies of 75% and 91% (specificities of 68% and 93%, sensitivities of 93% and 87%), respectively. On the other hand, under E 2, diagnostic efficiencies were 71% for IGF-1 and 86% for IGFBP-3 (Table 2). Side effects In 12 out of 59 children (4 girls and 8 boys), a moderate and transient breast enlargement was noticed after estrogen administration. It should be mentioned that breast enlargement, an undesirable effect of estrogen pretreatment, in all cases went unnoticed by either the children or their parents and was only detected during the physical examination performed after the test. FIG. 1. GH maximal response, under placebo (P) or E 2 administration in SS and GHD children. Each dot represents the result for an individual child. Boxes enclose the 95% confidence limits; the solid line in the middle of each box indicates the mean. The dashed horizontal bars represent arbitrary cut-off of 5 g/l (on placebo) and 9 g/l (on E 2 ). Discussion To confirm the clinical suspicion of GHD, a variety of diagnostic procedures had been suggested, among them physiological and pharmacological GH tests (17 22) and IGF-I and IGFBP-3 measurements (22 28), but an absolute standard could not be established because all available methodologies have some degree of uncertainty. Despite the known difficulties of pharmacological provocative tests (nonphysiological, arbitrary cut-off limits; different GH assays; day-to-day variability; and others), they TABLE 2. Sensitivity, specificity, and efficiency of GH axis parameters at different cut-off levels, under placebo and estradiol conditions Cut-off Sensitivity (%) Specificity (%) Efficiency (%) Placebo AUC g/l min / / / GHmax g/l / / / / / / IGF-I 14/ / / IGFBP-3 13/ / / Estradiol AUC g/l min / / / GHmax g/l / / / / / / IGF-I 14/ / / IGFBP-3 11/ / /

4 ESTRADIOL PRIMING IN GHD DIAGNOSIS 4171 FIG. 2. Levels of IGF-I (A) and IGFBP-3 (B), under placebo or E 2 administration in SS and GHD children. In both panels, the solid lines represent the lower limit of local normal values. The 95% confidence limits of controls are, for IGF-I (nmol/l): 0 3 yr, (n 27); 3 7 yr, (n 17); 7 11 yr, (n 19); and yr, (n 16). For IGFBP-3 (mg/l), mean 2 SD values of local controls are: 0 3 yr, (n 27); 3 7 yr, (n 17); 7 11 yr, (n 21); and yr, (n 32). remain the most accepted tool for evaluation of GH secretory capacity. Although spontaneous GH secretion had a better reproducibility (29, 30), it has a lower sensitivity, at least in prepubertal short children, and it does not represent an advantage over provocative tests (19). Priming with sex steroids remains a controversial issue. In a survey of 251 USA pediatric endocrinologists, Wyatt et al. (31) found that two-thirds of them do not prime any children. In a placebo-controlled study, Marín et al. (8) reported that 61% of normal Tanner stage I and 44% Tanner stage II children failed to attain a GH peak greater than 7 g/l in response to three provocative tests when they were not primed with sex steroids before testing, and that estrogen pretreatment was effective in improving GH secretion after pharmacological stimuli in normal children. However, controversial data exist in short children (32, 33). As far as we know, no previous placebo-controlled study on the effect of estrogen priming on the stimulated levels of GH in short children has been performed. Moreover, using micronized oral E 2 as an estrogen-priming agent, we could assess, by measuring serum E 2 levels, both the compliance and the level reached in each patient. In the present study, the administration of E 2, for 3 days, before GH stimulation, in prepubertal or early pubertal children, resulted in an improvement of the discrimination power of GH provocative tests between short non-gh deficient and GHD children. The best discrimination was achieved by using a cut-off limit of 9.0 g/l for maximal GH-stimulated value under E 2 ; a diagnostic efficiency of 98% was attained, above the 90% efficiency under placebo, using a maximal GH cut-off limit of 5.0 g/l. It is remarkable that the higher number of nonresponder SS children without priming were in the 8- to 12-yr-old subgroup (33%) and not, as it could be expected, in older subjects (16%) (3). There is a spectrum of GHD; thus, it may be possible that the use of priming before provocative testing may prevent the identification of a child with inadequate GH secretion. In our study, the overlap of GH responses between short and GHD children was reduced on primed conditions. However, a modest (albeit significant) increase in GH AUC, after E 2 priming, could be seen in the group of hypopituitary patients; and two of them reached a GHmax above the lower level of the 95% confident limits of SS children. The significant increase in IGFBP-3 levels, found after estrogen priming in GHD patients, might be related to the aforementioned increased in GH response. These results support the concept that GH secretion is a continuum, and except in patients with GH gene deletions, some GHD children could have some GH secretion in response to strong stimuli. As in previous studies, IGF-I and IGFBP-3 measurements do not correlate perfectly with the GH response to the stimulation tests (34). One-third of our healthy SS children under placebo showed IGF-I levels below normal, as has been reported in patients with a bone age below 12 yr (34). The IGFBP-3 values were discordant from the GH provocative response in only 3 of 41 of our SS children. On the other hand, GH under E2 priming, IGF-I, and IGFBP-3 under placebo were in near-complete agreement with the clinical diagnosis in the selected GHD patients. Considering the 68% specificity of IGF-I vs. the 93% of IGFBP3, and the similar sensitivities (93% vs. 87%), IGFBP-3 with a diagnostic efficiency of 91% seems more suitable for the diagnosis of GHD than IGF-I (diagnostic efficiency of 75%). Regarding the effects of E 2 on IGF-I and IGFBP-3 levels, it did not improve their diagnostic efficiencies. Controversial data exist regarding the value of IGF-I measurement as a screening test for GHD. Whereas some authors (26, 27) found similar high sensitivity and specificity, especially in older children, others (11) have reported that IGF-I levels have lower specificity and similar sensitivity. This was also observed in the present study, although the sensitivity of 93% and the specificity of 68% are somehow better than in previous reports. This could be related to the strict clinical criteria used for the definition of GHD. On the other hand, measurement of serum IGFBP-3 levels has shown excellent specificity but with a lower degree of sensitivity (11, 35, 36). In our patients, IGFBP-3 measurements had both high sensitivity and specificity. The reason for these findings could be related to the type of immunoassay used (immunoradiometric assay vs. RIA), the etiology of GHD (idiopathic or acquired), and other factors. In conclusion, we found that GH stimulation tests under E 2 priming had the highest diagnostic efficiency, allowing a more reliable discrimination of patients with GHD. Our find-

5 4172 MARTÍNEZ ET AL. JCE&M 2000 Vol. 85 No. 11 ings suggest that the effect of estrogen priming on GHstimulated levels, by reducing the number of false nonresponders, might be useful in discriminating between normal and abnormal GH status in SS children. Acknowledgments We gratefully acknowledge the children and their parents who agreed to participate in this study. We extend our thanks to the medical staff of the Division of Endocrinology for its collaboration in the recruitment of patients and to Ms. Margarita Jaimes and Mrs. Graciela Negrete for taking care of the children. We also acknowledge the technical expertise of Mrs. Cora Quiroga, Mrs. Ana Montese, Mr. Martín Peña, Ms. Silvina González, and Mr. Daniel De Maio in hormonal measurements. Also, we are indebted to Gordon B. Cutler, Jr. for critical review of the manuscript. References 1. Reiter EO, Martha Jr PM Pharmacological testing of growth hormone secretion. Horm Res. 33: Moll G, Rosenfield R, Fan V Administration of low-dose estrogen rapidly and directly stimulates growth hormone production. Am J Dis Child. 140: Gourmelen M, Pham-Trung M, Girard F Transient partial hgh deficiency in prepubertal children with delay of growth. Pediatr Res. 13: Cacciari E, Tassoni P, Parisi G, et al Pitfalls in diagnosing impaired growth hormone (GH) secretion; retesting after replacement therapy of 63 patients defined as GH deficient. J Clin Endocrinol Metab. 74: Tauber M, Moulin P, Pienkowski C, Jouret B, Rochiccioli P Growth hormone (GH) retesting and auxological data in 131 GH-deficient patients after completion of treatment. J Clin Endocrinol Metab. 82: Toogood AA, Shalet SM Diagnosis of severe growth hormone (GH) deficiency in young adults who receive GH replacement therapy during childhood. Acta Paediatr. [Suppl 423]: Juul A, Kastrup KW, Pedersen SA, Skakkebæck EN Growth hormone (GH) provocative retesting of 108 young adults with childhood-onset GH deficiency and the diagnostic value of insulin-like growth factor I (IGF-I) and IGF binding protein-3. J Clin Endocrinol Metab. 82: Marín G, Domené HM, Barnes KM, Blackwell BJ, Cassorla F, Cutler GB The effects of estrogen priming and puberty on the growth hormone response to standardized treadmill exercise and arginine-insulin in normal girls and boys. J Clin Endocrinol Metab. 79: Veldhuis JD, Metzger DL, Martha PM, et al Estrogen and testosterone, but not a nonaromatizable androgen, direct network integration of the hypothalamo-somatotrope (growth hormone)-insulin-like growth factor I axis in the human: evidence from pubertal pathophysiology and sex hormone replacement. J Clin Endocrinol Metab. 82: Rosenfeld RG Editorial: is growth hormone deficiency a viable diagnosis? J Clin Endocrinol Metab. 82: Tillmann V, Buckler JMH, Kibirige MS, et al Biochemical test in the diagnosis of childhood growth hormone deficiency. J Clin Endocrinol Metab. 82: Lejarraga H, Orfila G Estandares de peso y estatura para niñas y niños argentinos desde el nacimiento hasta la madurez. Arch Arg Pediatr. 85: Marshall W, Tanner J Variation in pattern of pubertal changes in girls. Arch Dis Child. 44: Marshall W, Tanner J Variation in pattern of pubertal changes in boys. Arch Dis Child. 45: Greulich W, Pyle S Radiographic atlas of skeletal development of the hand and wrist. ed 2. Stanford: Stanford University Press. 16. Daughaday WL, Mariz IK, Blethen SL Inhibition of access of bound somatomedin to membrane receptor and immunobinding sites: a comparison of radioreceptor and radioimmunoassay of somatomedin in native and acidethanol-extracted serum. J Clin Endocrinol Metab. 51: Underwood LE, Azumi K, Voina SJ, Van Wyk JJ Growth hormone levels during sleep in normal and growth hormone deficient children. Pediatrics. 48: Frasier SD A review of growth hormone provocative tests in children. Pediatrics. 53: Rose SR, Ross JL, Uriarte M, Barnes KV, Cassorla FG, Cuttler Jr GB The advantage of measuring stimulated as compared with spontaneous growth hormone levels in the diagnosis of growth hormone deficiency. N Engl J Med. 319: Ghigo E, Bellome J, Aimaretti G, et al Reliability of provocative tests to assess growth hormone secretory status. Study in 472 normally growing children. J Clin Endocrinol Metab. 81: Shalet SM, Toogood A, Rahim A, Brennan BMD The diagnosis of growth hormone deficiency in children and adults. Endocr Rev. 19: Rosenfeld RG, Albertsson-Wikland K, Cassorla F, et al Diagnostic controversy: the diagnosis of childhood growth hormone deficiency revisited. J Clin Endocrinol Metab. 801: Underwood LE, D Ercole AJ, Van Wyk JJ Somatomedin-C and the assessment of growth. Pediatr Clin North Am. 27: Cacciari E, Cicognani A, Pirazzoli P, et al Differences in somatomedin-c between short-normal subjects and those of normal height. J Pediatr. 106: Rosenfeld RG, Wison DM, Lee PDK, Hintz RL Insulin-like growth factors I and II in evaluation of growth retardation. J Pediatr. 109: Clemmons DR, Underwood LE Clinical review 59. Uses of human insulin-like growth factor-i in clinical conditions. J Clin Endocrinol Metab. 79: Blum WF, Ranke MB, Kietzmann K, et al A specific radioimmunoassay for the growth hormone (GH) dependent somatomedin binding protein: its use for the diagnosis of GH deficiency. J Clin Endocrinol Metab. 70: Blum WF, Albertsson-Wikland K, Rosberg S, Ranke MB Serum levels of insulin-like growth factor I (IGF-I) and IGF binding protein 3 reflect spontaneous growth hormone secretion. J Clin Endocrinol Metab. 76: Donaldson DL, Pan F, Hollowell JG, Stevenson JL, Gifford AR, Moore WV Reliability of stimulated and spontaneous growth hormone (GH) levels for identifying the child with low GH secretion. J Clin Endocrinol Metab. 72: Ropelato MG, Martínez A, Heinrich JJ, Bergadá C.1996 Reproducibility and comparison of growth hormone secretion test. J Pediatr Endocrinol Metab. 9: Wyatt DT, Mark D, Slyper A Survey of growth hormone treatment practices by 251 pediatric endocrinologist. J Clin Endocrinol Metab. 80: Lanes R, Lifshitz F, Sekaran C, Fort P, Recker B Premarin priming does not alter growth hormone release following exercise. J Endocrinol Invest. 9: Wilson DM, Dotson RJN, Neely EK, Cohen P, Hintz RL, Rosenfeld RG Effects of estrogen on growth hormone following clonidine stimulation. Am J Dis Child. 147: Moore DC, Ruvalcaba RHA, Smith EK, Kelley VC Plasma somatomedin-c as a screening test for growth hormone deficiency in children and adolescents. Horm Res. 16: Cianfaroni S, Boemi S, Spagnoli A, et al Is IGF binding protein-3 assessment helpful for the diagnosis of GH deficiency? Clin Endocrinol (Oxf). 43: Nuñez SB, Municchi G, Barnes KM, Rose SR Insulin-like growth factor I (IGF-I) and IGF-binding protein-3 concentrations compared to stimulated and night growth hormone in the evaluation of short children a clinical research center study. J Clin Endocrinol Metab. 81: