| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Internal Medicine, Division of Metabolism, Endocrinology and Diabetes (E.V.D., A.L.B., C.A.J.), Pituitary and Neuroendocrine Center and Department of Neurosurgery (W.F.C., A.L.B.), Department of Biostatistics (M.B.B.), and Department of Pediatrics (V.P.), University of Michigan, Ann Arbor, Michigan 48109; Department of Veterans Affairs Medical Center (A.L.B., C.A.J.), Ann Arbor, Michigan 48105; and Department of Pharmacology (S.Y.K., R.T.), University of Texas Southwestern Medical Center, Dallas, Texas 75235
Address all correspondence and requests for reprints to: Eleni V. Dimaraki, M.D., M.S., 2650 Ridge Avenue, Suite 5111, Evanston, Illinois 60201. E-mail: e-dimaraki{at}northwestern.edu.
| Abstract |
|---|
|
|
|---|
Objective: The objective of the study is to examine whether GH secretion in acromegaly is dependent on endogenous GHRH.
Patients and Study Design: We studied eight patients with untreated acromegaly due to a GH-producing pituitary tumor. All patients received an iv infusion of normal saline for 24 h and GHRH-antagonist (GHRH-ant) at 50 µg/kg·h for 7 d. GH was measured every 10 min for 24 h during the normal saline infusion and on the last day of the GHRH-ant infusion. A group of nine different patients with untreated acromegaly served as the control group and underwent blood sampling for GH every 10 min for two 24-h periods to assess the day-to-day variability of GH secretion.
Setting: The study was set in a university referral center.
Main Outcome Measure: Twenty-four-hour mean GH was the main outcome measured.
Results: In six of eight subjects treated with GHRH-ant, 24-h mean GH decreased by 5.830.0% during iv GHRH-ant and, in three subjects, the change in the 24-h mean GH was greater than the upper limit of the 95% confidence interval of the spontaneous day-to-day variability of the mean GH in patients with acromegaly. Based on the binomial distribution, the probability of this magnitude of change to occur in three of eight subjects by chance alone is 0.0008.
Conclusion: In some patients with acromegaly due to a pituitary adenoma, GH secretion is under partial control by endogenous GHRH.
| Introduction |
|---|
|
|
|---|
In humans, ectopic GHRH production from peripheral tumors results in somatotroph cell hyperplasia, but not in adenoma formation (11, 12, 13, 14, 15). However, pituitary histopathology in patients with gigantism (prepubertal onset of acromegaly) showed coexistence of adenoma and somatotroph hyperplasia in adjacent nontumoral pituitary in three of eight specimens studied (16). Additionally, there are reports of GHRH-secreting brain tumors with neuronal components associated with somatotroph adenomas (17, 18, 19). A somatotroph pituitary adenoma cosecreting GHRH and GH, with high circulating GHRH levels, has also been described (20). These observations raise the possibility that similar to transgenic animals, very high local concentrations of GHRH in humans can result in somatotroph adenoma formation.
However, the in vivo involvement of GHRH in the maintenance of GH hypersecretion in human acromegaly cannot be demonstrated in a direct fashion.
We used the competitive GHRH-antagonist (N-Ac-Tyr1,D-Arg2)GHRH(129)NH2 (GHRH-ant) in vivo in patients with acromegaly to examine its effect on GH secretion and IGF-I levels and determine the extent, if any, of GHRH involvement in the pathogenesis of acromegaly.
| Subjects and Methods |
|---|
|
|
|---|
Two groups of patients with acromegaly were studied, a control group and a group treated with GHRH-ant. All subjects had clinical symptoms and signs of acromegaly, elevated serum IGF-I levels, and pituitary adenomas demonstrated by a dedicated magnetic resonance imaging study. All patients were studied before having transsphenoidal surgery (TSS) or medical treatment for acromegaly. Patients with visual field defects or with a pituitary tumor in proximity to the optic chiasm were excluded. Hypopituitarism was treated with hormone replacement as appropriate. No changes in hormone replacement were allowed during the participation in the study. Subjects did not have any other uncontrolled medical problems. After their participation in the protocol, all patients underwent TSS, and the diagnosis of a somatotroph adenoma was confirmed by immunohistochemistry in all patients.
The control group consisted of nine patients, seven men and two women, age 1858 yr. These patients had blood sampling for GH every 10 min for two 24-h periods to assess the reproducibility of GH secretion and determine the expected day-to-day variability of GH secretion in patients with acromegaly.
The treatment group consisted of eight patients, five men and three women, age 2562 yr. These patients were admitted to the GCRC for 10 d and were treated with GHRH-ant. At the time of the admission, two iv catheters were inserted, one for blood drawing and one for iv infusion of normal saline or GHRH-ant. Subjects first received iv normal saline and underwent 24-h blood sampling every 10 min for GH, from 07000700 h. After the 24-h sampling was completed, seven of eight subjects had an oral glucose tolerance test (OGTT); oral glucose 100 g was given at 0700 h and blood sampling every 10 min continued until 0900 h. Subsequently, at 1000 h, iv GHRH 1 µg/kg (Bachem, Torrance, CA) was administered to assess the responsiveness of GH secretion to GHRH stimulation. After these tests were completed, GHRH-ant was administered as a continuous iv infusion at 50 µg/kg·h, for 7 d. We have shown previously that, in healthy young men, an iv bolus of GHRH-ant of 400 µg followed by continuous iv infusion at 50 µg/kg·h suppressed nocturnal pulsatile GH secretion by 90% (21). On the last 2 d of the infusion, the 24-h blood sampling for GH, the OGTT, and GH stimulation by iv GHRH were repeated. Serum IGF-I was measured on the first and last days of the GCRC admission at 0650 h.
In subject 8, the protocol procedures were terminated prematurely by 1 d because he developed cellulitis around the iv infusion site. On the last day of participation, he had blood sampling for GH only for 10 h. Therefore, in this subject, the 10-h data from the last day were compared with the corresponding 10 h during normal saline infusion. This patient also had comparable GH data for only 1 h after iv GHRH administration. The same subject did not have an OGTT because of the diagnosis of diabetes mellitus.
After the completion of the protocol, all patients underwent TSS as indicated for the treatment of acromegaly. Samples of the pituitary tumor tissue were obtained from all subjects in the treatment group.
Assays
Plasma GH was measured by a chemiluminometric assay with sensitivity 0.01 µg/liter and intrassay coefficient of variation less than 10% (Nichols Institute Diagnostics, San Clemente, CA). Plasma IGF-I was measured by a two-site immunoradiometric assay (Diagnostic Systems Laboratories, Webster, TX).
Detection of gsp mutations
RNA was extracted from frozen dissociated cells (510 x 105 cells per sample) in five cases or frozen tumor tissue in two cases, using a Totally RNA Kit (Ambion, Austin, TX) according to the manufacturers protocol. First-strand cDNA was synthesized using random hexamer primers and reverse transcriptase (Ambion). Genomic DNA was extracted from similarly plated patient samples using a Puregene genomic DNA purification kit (Gentra Systems, Minneapolis, MN) according to the manufacturers protocol. In one case, paraffin-embedded tissue was used as the source of genomic DNA; sections (5 µm) were sliced and deparaffinized with xylene, and DNA was extracted by overnight incubation in a buffer containing 50 mM Tris (pH 8.0), 1 mM EDTA, 0.5% Tween 20, and 200 µg/ml proteinase K.
To amplify the specific Gs
subunit cDNA (203 bp) or genomic DNA (537 bp) fragments containing codons 201 and 227, the following primers were used: sense primer, 5'-GTGATCAAGCAGGCTGACTATGTG-3'; and antisense primer, 5'-GCTGCTGGCCACCACGAAGATGAT-3'. Using these primers, the PCR was performed for 32 cycles (0.5 min at 94 C, 0.5 min at 59 C, 1.5 min at 72 C). The PCR products were purified by agarose gel electrophoresis and sequenced directly using an ABI model 3100 DNA sequencer (Foster City, CA). Sequencing of all PCR products was carried out using each of the primers described above.
Data analysis
Mean GH was defined as the average of all values in a 24-h period. The change of mean GH in each subject in the treatment group was compared with the expected day-to-day variability of mean GH in patients with acromegaly. For each subject in the control group, we calculated the difference in 24-h mean GH adjusted for the average mean GH as the change of 24-h mean GH between the second day of sampling divided by their average. The mean of the absolute values of these adjusted differences ± 2.3 SD was used to define the 95% confidence interval (CI) for the day-to-day differences in mean 24-h GH in patients with acromegaly; 2.3 SD was used because there were only nine control subjects. In planning the study, we predetermined that the suppression of mean GH by GHRH-ant would be considered significant if the magnitude of suppression was greater than the mean + 2.3 SD day-to-day variability of 24 h mean GH in the control patients with acromegaly. If a significant number of subjects in the treatment group had a decline in mean GH greater than the expected day-to-day variability, then we would conclude that GHRH-ant has an effect on GH secretion.
The 24-h mean GH data from the subjects in the control group were used to calculate the amount of day-to-day variability of 24-h mean GH in patients with acromegaly. Therefore, we would expect that, for each subject in the treatment group, if there was no effect of the GHRH-ant on the GH secretion, the difference between 24-h mean GH during normal saline and GHRH-ant infusion would be within this CI. We calculated the number (n) of subjects in the treatment group who had a decrease of mean GH, adjusted for their mean GH, greater than the mean + 2.3 SD in the control group. We then used the binomial distribution to calculate the probability of n subjects having such a change in mean GH.
The GH response to iv GHRH is expressed as the amplitude of the response, the difference between the highest plasma GH concentration in the 2 h after the injection and the baseline plasma GH concentration at time 0, and as the area under the curve of GH concentration over the 2-h time calculated by the trapezoidal rule. The GH responses to GHRH at baseline and during GHRH-ant treatment were compared by paired t test.
The GH response to OGTT was defined as the nadir plasma GH concentration over the 2 h after oral glucose administration. The nadir GH concentrations during normal saline and GHRH-ant infusions were compared by paired t test.
| Results |
|---|
|
|
|---|
The 24-h plasma GH profiles of patients in the control group are shown in Fig. 1
. The 24-h mean plasma GH concentration in patients in the control group was 31.5 ± 10.0 µg/liter during the first 24-h period and 35.0 ± 11.6 µg/liter during the second 24-h period. The mean of the adjusted absolute values of the differences in mean 24-h GH between the two sampling periods was 11.3 ± 5.6% (mean ± SD). The 95% CI for the absolute adjusted day-to-day differences in mean 24-h GH in patients with acromegaly would be 0, 24.2%. Therefore, according to our predetermined criteria, only a suppression of 24-h mean GH by the GHRH-ant infusion greater than 24.2% was considered to be significant.
|
The 24-h GH concentration profiles of the eight patients in the treatment group are shown in Fig. 2
. In six of eight subjects, mean GH decreased by 5.8 to 30.0% during the GHRH-ant infusion. In two patients, the mean GH increased by 4.2 and 58.9%. In three of eight subjects, the magnitude of the decline of the 24-h mean GH adjusted for the average 24-h mean GH was greater than the upper limit of the 95% CI calculated by the data in the control group (Fig. 3
). Based on the binomial distribution, the probability of this magnitude of change to occur in three of eight subjects by chance is 0.0008.
|
|
The amplitude of the GH response to exogenous GHRH varied, but all subjects in the treatment group responded to GHRH stimulation. The mean amplitude of the response was 59.7 ± 27.1 µg/liter, during the normal saline infusion. During the GHRH-ant infusion, the amplitude of the GH response to GHRH was suppressed by an average of 58.0 ± 15.2% (to 22.6 ± 11.8 µg/liter, P = 0.013). Similarly, the area under the curve of GH concentration over the 2 h after GHRH administration was suppressed by 35.9 ± 7.5% (4413 ± 1941 vs. 7331 ± 3351 µg x min/liter, P = 0.003).
|
OGTT
In all seven subjects, the nadir GH concentration after oral glucose was not affected by the GHRH-ant infusion (28.7 ± 12.2 vs. 23.5 ± 9.8 µg/liter, P = 0.32).
Serum IGF-I
Serum IGF-I values were available for seven of eight subjects in the treatment group. Serum IGF-I concentrations did not change during the GHRH-ant infusion (1297 ± 163 vs. 1238 ± 145 µg/liter, P = 0.52)
gsp mutations
gsp mutations were detected at site R201 in the tumors of subjects 3 and 8 and at site R227 in the tumor of subject 4. In subject 3, there was a decrease in the mean GH during the GHRH-ant infusion, but this was not more than the mean adjusted change + 2 SD in the control group. In subject 8, 24-h mean GH was significantly suppressed by the GHRH-ant infusion. In subject 4, there was a small increase of 4.2% in the 24-h mean plasma GH during the GHRH-ant infusion.
| Discussion |
|---|
|
|
|---|
Detection of oncogenes is generally rare in sporadic pituitary tumors (8, 22). However, in the case of somatotroph adenomas, 3040% of the tumors have activating mutations of the
subunit of the Gs, called gsp mutations, that result in constitutive activation of the adenylate cyclase (24, 25, 26). It appears that, in these tumors, GHRH-independent activation of the GHRH signaling pathway is responsible for tumorigenesis and GH hypersecretion. In addition, X-chromosome inactivation analysis suggests that pituitary somatotroph adenomas are monoclonal in origin, i.e. they are the result of proliferation of a single mutated cell (27, 28). The validity of the X-chromosome inactivation analysis is based on the assumption of polyclonality of pituitary cells at the single cell level. However, it has been shown that multiple tissues consist of discrete groups of monoclonal cells (29, 30, 31, 32). Therefore, if an external stimulus promotes hyperplasia of one or more clones, sampling from the area of one clone would give the impression of a monoclonal tumor. Indeed, the majority of patients with uremic hyperparathyroidism have monoclonal parathyroid tumors (33). Clayton et al. (34) studied recurrent pituitary tumors using allelotyping of polymorphisms to detect loss of heterozygosity and X-chromosome inactivation analysis, and showed that more than half of the recurring tumors were derived from a different clone than the original tumor. These observations suggest that a common initiating event, possibly altered hypothalamic stimulation, promotes the development and/or proliferation of one or more clones. Alternatively, it is possible that hypothalamic stimulation is required for the expansion of a mutated clone, or hypothalamic overstimulation renders pituitary cells susceptible to mutations.
Until now, there was only indirect evidence that, in patients with acromegaly due to pituitary somatotroph adenomas, GH secretion is under hypothalamic control. As previously reported and demonstrated by the 24-h GH profiles of our control subjects, in patients with acromegaly, the pattern of GH secretion is highly consistent from day to day (35, 36). In addition, in patients with typical acromegaly, the nocturnal augmentation of GH secretion is preserved (35, 36). Together, these features suggest a degree of central GH control. Furthermore, patients with "cured" acromegaly, as defined by a normal serum IGF-I level and GH suppression by a glucose load, maintain increased GH pulse frequency after successful TSS, suggesting a primary hypothalamic control disturbance (37, 38). Because GHRH has been considered as the predominant stimulus of GH pulses, GHRH hyperpulsatility has been suggested.
The potential role of GHRH in the pathogenesis of acromegaly is supported by in vitro and in vivo data. In vitro, GHRH stimulates GH synthesis, GH release, and somatotroph cell proliferation (1, 2, 3, 4, 39). Human GHRH transgenic mice develop marked pituitary hyperplasia and, at age 1014 months, also develop pituitary adenomas (8, 10). In humans, ectopic GHRH secretion results in pituitary somatotroph hyperplasia and acromegaly (11, 12). Secretion of GHRH by neuronal tumors (gangliocytomas, hamartomas, choristomas) in the hypothalamus or the pituitary is associated with pituitary somatotroph adenoma formation and acromegaly (15, 17, 18, 19). Junctions between the neuronal tumor cells and the pituitary adenoma cells have been described. More recently, a case of a pituitary adenoma cosecreting GHRH and GH and resulting in elevated plasma GHRH concentration was described (20). These paradigms suggest that exposure to high concentrations of GHRH for a sufficient length of time can result in the formation of a GH-producing pituitary adenoma.
Most importantly, GHRH mRNA and protein as well as GHRH-receptor mRNA are present in somatotroph adenomas (6, 7, 40, 41). The functional importance of the tumor GHRH was demonstrated by its selectivity (absence in all other non-GH secreting pituitary tumors), as well as by the remarkable correlation between the levels of GHRH mRNA at the tumor level and the circulating GH concentration, the invasiveness of the tumor, and the chance of postoperative remission (7). Therefore, although the absence of hyperplasia surrounding the adenomas is an argument against an important role of GHRH, endogenous GHRH may be a causative or a supportive factor in the development of GH-producing pituitary adenomas and/or the maintenance of GH hypersecretion in acromegaly. Moreover, in three of eight patients with gigantism, a pituitary adenoma and pituitary somatotroph hyperplasia coexisted (16).
The high frequency of gsp mutations also suggests that activation of the GHRH signaling pathway plays a role in tumorigenesis. Classic somatotroph adenomas respond to exogenous GHRH both in vitro and in vivo (3, 4, 5, 42, 43, 44, 45). Our results further show that, at least in a subset of somatotroph adenomas in humans, GH secretion is partially controlled by endogenous GHRH in vivo. However, it is not clear whether this GHRH is of hypothalamic origin or locally produced by the pituitary tumor, exerting paracrine or even autocrine effects. In one study, Adams et al. (46) found consistent stimulation of GH by GHRH 2 nM in cell cultures. They also treated pituitary somatotroph adenomas with GHRH-ant at a concentration of 60 nmol/liter. They observed no effects of GHRH-ant on spontaneous GH secretion, but GHRH-ant did inhibit the stimulatory effects of GHRH. These data suggest a more important role of hypothalamic GHRH in comparison to paracrine/autocrine GHRH.
In the case of our patients, it would be expected that tumors with gsp mutations would be less likely to respond to the GHRH-ant infusion. However, in this small group, we did not observe any relationship between the presence of a gsp mutation and the behavior of the tumor during GHRH-ant infusion.
In conclusion, our results show that, in some pituitary somatotroph adenomas, GH secretion is in part under endogenous GHRH stimulation, raising the possibility that GHRH hypersecretion is the initiating event or a promoting factor for a subset of GH-secreting pituitary adenomas. Studies with more potent GHRH-ant would be warranted to further explore this possibility. Moreover, in this subset of patients, potent GHRH-ant might be a therapeutic option.
| Footnotes |
|---|
Disclosure Summary: The authors have nothing to disclose.
First Published Online March 14, 2006
Abbreviations: CI, Confidence interval; GHRH-ant, GHRH antagonist; OGTT, oral glucose tolerance test; TSS, transsphenoidal surgery.
Received November 2, 2005.
Accepted March 2, 2006.
| References |
|---|
|
|
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |