| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Endocrine Care |
Department of Endocrinology, Christie Hospital, Manchester, United Kingdom M20 4BX
Address all correspondence and requests for reprints to: Dr. Peter J. Trainer, Department of Endocrinology, Christie Hospital, Wilmslow Road, Manchester, United Kingdom M20 4BX.
Abstract
In patients with acromegaly, there is a linear association between log10 serum GH and IGF-I. Healthy females secrete three times more GH than males but have broadly similar serum IGF-I levels, and in adult GH deficiency, the dose of exogenous GH required to achieve a given serum IGF-I is significantly greater in females than males. We report the influence of gender on the relationship between serum GH and IGF-I in subjects with active acromegaly. A single, fasted, serum sample was obtained from 153 subjects with active disease (87 males; median age, 47.8 yr; range, 2082 yr) in whom serum IGF-I was at least 30% above the upper limit of an age-related reference range after washout from medical therapy. A linear correlation between serum IGF-I and log10 serum GH was observed (r = 0.53; P < 0.0001), but this relationship was significantly influenced by gender. For a given serum GH value, females were estimated to have serum IGF-I values 82 ng/ml less than males [P < 0.02; 95% confidence interval (CI), 15.2149]. In females receiving oral E, mean serum IGF-I for a given GH value was 130 ng/ml lower than in males (P = 0.01; 95% CI, 29.8230.2) but only 60 ng/ml less than the remaining 45 females (NS; P = 0.2). This study demonstrates a gender difference in the relationship between serum GH and IGF-I in patients with active acromegaly consistent with relative GH resistance observed in normal and GHD females, which may, in part, be mediated by E. This observation has important implications for the use of IGF-I as a measure of disease activity.
ACROMEGALY IS CAUSED by excess GH secretion almost always from a pituitary somatotroph tumor (1). In turn, elevated circulating GH increases the concentration of GH-dependent peptides such as IGF-I, IGF binding protein-3 (IGFBP-3), and acid labile subunit (ALS). Although certain metabolic functions including lipolysis are mediated directly by GH, many of the observed effects of GH are the result of IGF-I generation (2), and a correlation between serum IGF-I and clinical measures of disease activity is reported (3). The relationship between serum GH and IGF-I is linear below GH levels of approximately 20 ng/ml, but at higher levels of GH secretion IGF-I generation is maximally stimulated with no further increase in circulating IGF-I (4, 5, 6, 7). When serum GH data are log10 transformed and plotted against serum IGF-I, a log-dose correlation is observed (4, 5, 6, 7).
Currently both serum GH and IGF-I are used to define treatment goals for acromegaly. A consensus statement has defined biological cure as the return to normal of all attributes of disordered GH secretion, as evidenced by the suppression of serum GH below 1 µg/liter after a 75-g oral glucose load and the presence of a normal age-matched serum IGF-I value (8). Numerous studies involving more than 2000 patients with acromegaly have demonstrated the benefit of lowering serum GH and the need for tight control because reduction below 2.5 ng/ml (5 mU/liter) is associated with reduced mortality (9, 10, 11). In these studies, the ratio of observed to expected deaths was increased 2.4- to 4.8-fold in those patients with persistent disease activity [GH >2.5 ng/liter (5mU/liter)] (9, 10, 12, 13). Less information exists on the prognostic use of serum IGF-I, although Swearingen et al. (13) recently alluded to normalization of mortality associated with a reduction of serum IGF-I into the normal range after transsphenoidal pituitary surgery.
Although serum IGF-I is principally regulated by GH, other factors are believed to modulate the relationship between serum GH and IGF-I. In patients with acromegaly, discordant serum GH and IGF-I results (either an elevated serum GH and normal IGF-I, or more commonly, elevated serum IGF-I in the presence of acceptable GH control) are seen in approximately one third of patients. In a study including 36 patients with acromegaly, only 15 of 24 (62.5%) individuals with safe serum GH levels [GH <2.5 µg/liter (5 mU/liter)] had a normal serum IGF-I level (7). Two other patients with normal serum IGF-I were observed to have elevated serum GH (7). Similarly, in 73 patients with acromegaly studied off medical therapy, 8% had elevated serum IGF-I despite GH values below 2.5 µg/liter, and 19% had elevated serum GH despite normal serum IGF-I values (14).
Little is known of the factors responsible for the discrepancy between serum GH and IGF-I in patients with acromegaly, but evidence from normal subjects and patients with GH deficiency suggests that gender influences the relationship between these parameters (15). Despite greater GH secretion, final adult height is lower in women compared with men, implying relative GH resistance in women (16, 17, 18, 19, 20).
Here we report the influence of gender on the relationship of serum GH to IGF-I in patients with active acromegaly.
Materials and Methods
Patients
One hundred fifty-three patients (87 male; median age, 47.8 yr; range, 2082 yr) were studied before their involvement in ethics committee-approved multicenter trials of pegvisomant therapy in acromegaly (21, 22). The entry criteria for these studies included a serum IGF-I at least 30% above the upper limit of an age-related reference range. All patients provided written informed consent and were instructed to withdraw medical therapy for acromegaly. A washout period of 5 and 2 wk was used for dopamine agonists and sc octreotide therapy, respectively. No patients had received a long-acting depot somatostatin analog in the preceding 12 wk. Eligible patients were at least 18 yr of age and had a confirmed diagnosis of acromegaly, based on historical data, symptoms and signs of acromegaly, and an elevated serum IGF-I concentration. One hundred twenty-nine of the 153 patients had previously undergone pituitary surgery, and 89 had received conventional three-field pituitary irradiation a median of 5.4 yr before sampling (range, 0.341.6 yr). One hundred twenty-six were receiving medical therapy at the time of enrollment and four patients had never received any previous treatment. Twenty-one were receiving oral E therapy but none were on transdermal E. Forty-two males were receiving T replacement therapy.
Methods
A single, fasted, early morning serum sample was obtained after
washout, and serum GH, IGF-I, IGFBP-3 and ALS were measured. The
dependency of mean serum IGF-I on covariates such as
log10 serum GH, sex, age, study cohort, and prior
radiotherapy was assessed with regression techniques. Exploratory
analysis using generalized additive
models did not reveal any gross
deviations from linearity for the
continuous terms considered, and so
standard multiple regression was used in subsequent analyses. One
outlying patient was excluded from the regression analysis (Fig. 1
), although inclusion of this subject
increased the magnitude of the influence of gender on the relationship
between serum GH and IGF-I. Serum ALS and IGFBP-3 were measured on 109
patients, and these data were used to analyze the influence of gender
on the relationship between serum GH and these parameters
(21).
|
Serum IGF-I was measured by a competitive binding RIA (Nichols Institute Diagnostics, San Juan Capistrano, CA); intra-assay coefficient of variation (CV) was 2.43%, and interassay CV was 5.28.4%. Serum GH was measured by RIA (Endocrine Sciences, Inc.). Interassay CV was 12% at 3.4 ng/ml, sensitivity 0.03 ng/ml. Serum IGFBP-3 was also measured by RIA (Endocrine Sciences, Inc.), intra-assay CV 13% at 1 mg and 5.1% at 2.7 mg, interassay CV 17% at 0.8 mg and 5.5% at 2.9 mg. ALS was measured by sandwich enzyme-linked immunosorbent assay (Diagnostic Systems Laboratories, Inc., Webster, TX), intra-assay CV 57% and interassay CV 510%.
Results
The results of the multiple regression analysis are shown in Table 1
. A significant linear association
between serum IGF-I and log10 serum GH was
observed for the cohort (R = 0.53; P < 0.0001).
Further analysis demonstrated a statistically significant dichotomy in
the linear association between log10 serum GH and
IGF-I in men and women (Fig. 1
). For a given serum GH value, females
were estimated to have an average serum IGF-I level 82 ng/ml lower than
male counterparts (P < 0.02; 95% CI, 15.2149 ng/ml)
(Fig. 1
). Twenty-one females were receiving oral E therapy, and in this
group, the mean serum IGF-I for a given GH value was 130 ng/ml less
than in males (P = 0.01; 95% CI, 29.8230.2) and 60
ng/ml less than in the remaining 45 females (P = 0.2;
test for interaction). In contrast, there was no effect of T
replacement on the relationship between serum GH and IGF-I in this
cohort. For a given serum GH value, males receiving T had serum IGF-I
levels 14 ng/ml below those of similar males not receiving androgen
replacement (P = 0.7).
|
|
|
This is the first study to have addressed the influence of gender on the relationship between serum GH and IGF-I in patients with acromegaly. We have confirmed a significant relationship between log10 serum GH and IGF-I in patients with active disease, similar to that observed by other workers and for the first time have observed a sexual dichotomy in the relationship between log10 serum GH and IGF-I in patients with active acromegaly (5, 6, 7, 23).
Healthy women have higher mean 24-h GH levels compared with men, the result of greater GH secretory burst mass, relatively reduced GH suppressibility following oral glucose, and a higher degree of disordered GH release (15, 16, 17, 18, 19). Despite greater GH secretion in women, serum IGF-I values were not statistically different between sexes in these studies. For an equivalent serum IGF-I concentration, van den Berg et al. (16) observed a 3-fold higher mean serum GH concentration and a 2.5-fold lower IGF-I/GH ratio in females, despite similar GH elimination rates. However, 3-fold higher serum IGF-I levels are not observed in normal women, who have broadly similar serum IGF-I levels to those of men (24). These observations imply relative GH resistance in women who, despite higher levels of GH secretion, have lower final adult height than men (20).
Evidence from patients with GH deficiency (GHD) also suggests that gender influences the relationship between serum GH and IGF-I. Despite use of the same cut-off to provocative testing, GHD females have significantly lower pretreatment serum IGF-I levels and require significantly higher doses of GH replacement to maintain equivalent serum IGF-I values to males of similar GH status (16). The observation that, for a given serum GH level, women with active acromegaly have lower serum IGF-I values than men is entirely consistent with these reports (25).
In contrast to some previous studies, we relied upon a single GH measurement rather than multiple measurements during a single day. A correlation has been reported previously between single measures of GH and mean 24-h GH secretion in patients with acromegaly and the relationship between fasting serum GH and mean 24-h GH secretion is stronger in patients with acromegaly than in controls (r = 0.82 and 0.39, respectively) (4, 5, 23). The logarithm of nonpulsatile GH release is also a significant independent predictor of the serum IGF-I (4). Because nonpulsatile GH release accounts for the majority of GH secretion in patients with acromegaly, a single fasting sample is sufficient in assessing the effects of gender on the relationship between serum GH and IGF-I in this large cohort (26).
That oral E therapy exaggerates the gender-dependent difference in the relationship between serum GH and IGF-I in patients with active acromegaly implies that E is a major determinant of serum IGF-I levels (15, 16, 17, 18, 19, 27). The influence of E in modifying the relationship between serum GH and IGF-I is evidenced by a 2-fold increase in serum GH at the mid-portion of the menstrual cycle without a concomitant rise in serum IGF-I (28, 29). In contrast, no sexual dichotomy in daily GH secretion rates is observed between males and young females studied in the early follicular phase of the menstrual cycle (30, 31). In postmenopausal women, oral E therapy lowers basal serum IGF-I (32), and this fall is associated with a rise in circulating GH (31). Both peak and absolute increase in serum IGF-I after a single dose of GH (0.1 mg/kg) are reduced by pretreatment with oral E in such subjects (32).
Oral and transdermal E may have different effects on GH secretory dynamics. Whereas oral E administration to postmenopausal women produces a significant rise in 24-h GH secretion, this is not observed during transdermal E replacement (31). The mechanism(s) underlying divergent serum IGF-I responses to oral and transdermal E are unclear, but first-pass hepatic E metabolism may directly inhibit hepatic IGF-I production (31). A recent report has suggested that E is inhibitory to GH-induced phosphorylation of JAK2/STAT5 in human embryonic kidney and hepatoma cells and that this inhibitory effect is abolished in the presence of the anti-E ICI 182780 (33).
In breast cancer patients receiving tamoxifen, a nonsteroidal partial antagonist to the E receptor, serum IGF-I levels are consistently lower than in patients not receiving the drug (34, 35). Tamoxifen produces a significant fall in serum IGF-I when administered to postmenopausal women with metastatic breast carcinoma, but does not alter GH response to GHRH (36). When administered to patients with acromegaly, tamoxifen therapy induces a transient rise in serum GH and long-term fall in serum IGF-I levels (37). Clemmons et al. (38) assessed the effect of oral E on serum IGF-I in acromegaly in five patients and observed significant reductions in serum IGF-I after 3 d of ethinylestradiol (1 mg/day) therapy, and the potential therapeutic role of oral E therapy in acromegaly now warrants further investigation. Serum IGF-I correlates with measures of disease activity (39) and the further suppression of serum IGF-I by the coadministration of oral E may lead to improvements in patient well-being and signs of active disease. The ability of oral E to lower serum IGF-I in patients with active acromegaly and the observation that, unlike oral replacement, transdermal E may increase serum IGF-I in healthy postmenopausal women (31) suggests that, when required, E replacement in acromegaly should be oral. In contrast, in GHD women, E replacement should occur via the transdermal route.
In this cohort, the effect of age on the serum GH/IGF-I relationship was highly significant, such that, for a given serum GH value, older patients had lower serum IGF-I values. This observation may imply that older patients have relative GH resistance compared with young patients (40). Increasing GH resistance with age in the normal population has been suggested by Lieberman et al. (32), who documented a significantly lower peak and absolute rise in serum IGF-I in older males compared with younger males in response to a single (0.1 mg/kg) bolus of GH. We believe that the fall in serum IGF-I with age for a given GH observed in this cohort may represent a selection bias, although further study is required. In healthy subjects, serum IGF-I declines with age and thus, when serum IGF-I is employed as a measure of GH secretion for recruitment into a therapy study, as was the case here, for a given serum IGF-I value older individuals with acromegaly are more likely to meet the specified entry criteria. Similarly, the effect of study is believed to be secondary to the entry criteria for the two studies analyzed. The patients reported here were initially recruited into two trials of pegvisomant therapy for acromegaly (21). For the first of these studies, patients with more severe acromegaly were recruited (serum IGF-I >50% above the upper limit of normal after washout). When the influence of study on the relationship between serum GH and IGF-I was corrected for in the multiple regression analysis, the effect of gender on this relationship between serum GH and IGF-I, as reported, was reduced but remained significant.
IGF-I circulates as a ternary complex with IGFBP-3 and ALS, which are also synthesized by the liver under GH regulation. Unlike IGFBP-3, which circulates in equivalent molar amounts to IGF-I plus IGF-II, ALS exists in molar excess to both IGF-I and -II, and IGFBP-3; approximately 50% of ALS exists in free form (41). In normal volunteers, in which GH secretion is greater in women, significantly higher ALS values are observed in females compared with age-matched males, despite the presence of similar serum IGF-I and IGFBP-3 levels (20, 42). However, it is not clear whether serum ALS is elevated in normal females compared with males with similar GH levels. We have observed 1) a significant difference in the relationship between GH and ALS in males and females with acromegaly such that, for a given serum GH level, females have significantly higher serum ALS levels, and 2) that oral E exaggerates this gender difference. These data contradict those of Kam et al. (43) who recently reported a parallel dose-dependent fall in serum IGF-I, IGFBP-3, and ALS after administration of various oral E preparations to postmenopausal and GHD women. The reason(s) for this discrepancy remain unclear.
In conclusion, we have demonstrated a gender-based difference in the relationship between serum GH and IGF-I and between serum GH and ALS in individuals with active acromegaly. Females manifest lower serum IGF-I concentrations compared with males with the same serum GH level, and this dichotomy is most marked when women receiving oral E therapy are compared with males. These data have major implications for the use of serum IGF-I and GH as markers of disease activity in acromegaly. Consensus guidelines specify treatment targets for both serum GH and IGF-I (8), but discordant serum GH and IGF-I results are not infrequent and may partly be due to gender differences in the relationship between serum GH and IGF-I in patients with acromegaly and/or the concomitant use of oral E. Additionally, retrospective studies relating treatment outcomes to biochemical control have not addressed the possible influence of gender on outcome (9, 10, 11, 13). Prospective data relating serum IGF-I to both morbidity and mortality in acromegaly are lacking, and as this information is gathered, the effect of gender on measures of disease outcome will need to be considered. Much work is needed to establish the relative merits of serum GH and IGF-I as markers of disease activity in acromegaly.
Acknowledgments
The Sensus Acromegaly Study Group includes the following individuals, in alphabetical order: A. L. Barkan, W. F. Bennett, B. A. Bengtsson, G. M. Besser, M. Bidlingmaier, D. R. Clemmons, D. M. Cook, R. J. Davis, W. M. Drake, E. V. Dimaraki, P. U. Freda, K. E. Friend, S. Hackett, V. Herman-Bonert, G. Johannsson, L. Katznelson, D. L. Kleinberg, A. Klibanski, M. Maldonado, S. Melmed, L. S. Phillips, J. S. Powell, D. R. Rose, J. A. Scarlett, M. C. Sheppard, S. Stavrou, P. M. Stewart, C. J. Strasburger, M. O. Thorner, M. L. Vance, A. J. van der Lely, and K. Zib.
Affiliations: St. Bartholomews Hospital, London, United Kingdom (W.M.D., G.M.B.); Massachusetts General Hospital, Boston, Massachusetts (L.K., A.K.); Columbia College of Physicians and Surgeons, New York, New York (P.U.F., J.S.P.); Cedars-Sinai Medical Center, Los Angeles, California (S.M., V.H.-B.); Academic Hospital Dijkzigt, Rotterdam, The Netherlands (A.J.v.d.L.); the University of Michigan Medical Center, Ann Arbor, Michigan (E.V.D., A.L.B.); the University of Birmingham, Birmingham, United Kingdom (P.M.S., M.C.S.); the University of Texas, M. D. Anderson Cancer Center, Houston, Texas (K.E.F., M.M.); the University of Virginia Health Sciences Center, Charlottesville, Virginia (M.L.V., M.O.T.); Sensus Drug Development, Austin, Texas (J.A.S., W.F.B., R.J.D.); University of North Carolina School of Medicine, Chapel Hill, North Carolina (D.R.C., D.R.R.); Sahlgrenska University Hospital, Gotenborg, Sweden (G.J., B.A.B.); New York University Medical Center, New York, New York (D.L.K., S.S.); Oregon Health Sciences University, Portland, Oregon (D.M.C.); Emory University School of Medicine, Atlanta, Georgia (L.S.P.); Klinikam Innesnstadt, Ludwig-Maximilians-Universitat, Munich, Germany (M.B., C.J.S.); and StatWorks, Chapel Hill, North Carolina (S.H., K.Z.).
Footnotes
Abbreviations: ALS, Acid labile subunit; CI, confidence interval; CV, coefficient of variation; GHD, GH deficiency; IGFBF-3, IGF binding protein-3.
Received March 27, 2001.
Accepted July 24, 2001.
References
This article has been cited by other articles:
![]() |
O. Alexopoulou, M. Bex, R. Abs, G. T'Sjoen, B. Velkeniers, and D. Maiter Divergence between Growth Hormone and Insulin-Like Growth Factor-I Concentrations in the Follow-Up of Acromegaly J. Clin. Endocrinol. Metab., April 1, 2008; 93(4): 1324 - 1330. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Schmid, P.-A. Krayenbuehl, R.-L. Bernays, C. Zwimpfer, F. E. Maly, and P. Wiesli Growth Hormone (GH) Receptor Isoform in Acromegaly: Lower Concentrations of GH but Not Insulin-Like Growth Factor-1 in Patients with a Genomic Deletion of Exon 3 in the GH Receptor Gene Clin. Chem., August 1, 2007; 53(8): 1484 - 1488. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Parkinson, P. Burman, M. Messig, and P. J. Trainer Gender, Body Weight, Disease Activity, and Previous Radiotherapy Influence the Response to Pegvisomant J. Clin. Endocrinol. Metab., January 1, 2007; 92(1): 190 - 195. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Coutant, F. B. de Casson, S. Rouleau, O. Douay, E. Mathieu, F. Gatelais, N. Bouhours-Nouet, C. Voinot, M. Audran, and J. M. Limal Divergent Effect of Endogenous and Exogenous Sex Steroids on the Insulin-Like Growth Factor I Response to Growth Hormone in Short Normal Adolescents J. Clin. Endocrinol. Metab., December 1, 2004; 89(12): 6185 - 6192. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. R. Biermasz, A. M. Pereira, M. Frolich, J. A. Romijn, J. D. Veldhuis, and F. Roelfsema Octreotide represses secretory-burst mass and nonpulsatile secretion but does not restore event frequency or orderly GH secretion in acromegaly Am J Physiol Endocrinol Metab, January 1, 2004; 286(1): E25 - E30. [Abstract] [Full Text] |
||||
![]() |
A. Mukherjee, J. P. Monson, P. J. Jonsson, P. J. Trainer, and S. M. Shalet Seeking the Optimal Target Range for Insulin-Like Growth Factor I during the Treatment of Adult Growth Hormone Disorders J. Clin. Endocrinol. Metab., December 1, 2003; 88(12): 5865 - 5870. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. J. Kopchick, C. Parkinson, E. C. Stevens, and P. J. Trainer Growth Hormone Receptor Antagonists: Discovery, Development, and Use in Patients with Acromegaly Endocr. Rev., October 1, 2002; 23(5): 623 - 646. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. J. Trainer Acromegaly--Consensus, What Consensus? J. Clin. Endocrinol. Metab., August 1, 2002; 87(8): 3534 - 3536. [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |