| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Endocrinology (K.H.D., S.M.S.), Christie Hospital, Manchester M20 4BX, United Kingdom; and the Department of Medicine (M.O.T., S.S.P.), University of Virginia Health Science Center, Charlottesville, Virginia 22908
Address all correspondence and requests for reprints to: Professor S. M. Shalet, Department of Endocrinology, Christie Hospital, Wilmslow Road, Manchester M20 4BX, United Kingdom. E-mail: stephen.m.shalet{at}man.ac.uk.
| Abstract |
|---|
|
|
|---|
Study Subjects: Twenty-four-hour spontaneous GH secretion was studied by 20-min sampling both in the fed state (n = 16; six women) and the last 24 h of 33-h fast (n = 10; three women) in adult cancer survivors of normal GH status defined by two GH provocative tests, 13.1 ± 1.6 (range, 328) yr after cranial irradiation (1840 Gy) for nonpituitary brain tumors (n = 12) or leukemia (n = 4) in comparison with 30 (nine women) age- and body mass index-matched normal controls (fasting, 11 men and three women).
Results: Using previously published diagnostic thresholds, all patients had stimulated peak GH responses in the normal range to both the insulin tolerance test and the combined GHRH plus arginine stimulation test, as well as normal individual mean profile GH levels during the fed and fasting states. However, gender-specific comparisons revealed marked reduction (by 40%) in the overall peak GH responses to both provocative tests but similar GH secretory profiles; no differences were seen in the pulsatile attributes of GH secretion (cluster analysis) or the profile absolute and mean GH levels in the fed state or when the hypothalamic-pituitary axis was stimulated by fasting.
Conclusions: Radiation-induced GH neurosecretory dysfunction either does not exist or is a very rare phenomenon in irradiated adult cancer survivors. The normality of physiological GH secretion in the context of reduced maximum somatotroph reserve suggests compensatory overdrive of the partially damaged somatotroph axis and constitutes a relative argument against somatotroph dysfunction being explained purely by hypothalamic damage with secondary atrophy due to GHRH deficiency. It is therefore possible that radiation in doses less than 40 Gy causes dual damage to both the pituitary and the hypothalamus.
| Introduction |
|---|
|
|
|---|
In the light of these uncertainties about the genuine existence of GHNSD after irradiation, we sought to examine spontaneous GH secretion in irradiated adult cancer survivors with normal stimulated GH responses; if radiation were to cause neuroregulatory disturbances in GH secretion due to hypothalamic damage with consequent GHNSD, one would expect persistence of these disturbances into adult life and more so if the adult h-p axis was subjected to physiological stimuli normally associated with increased GH secretion, such as fasting.
| Patients and Methods |
|---|
|
|
|---|
Sixteen adult survivors (six women) of nonpituitary brain tumors (n = 12) or leukemia (n = 4) were studied 13.1 ± 1.6 (range, 328) yr after cranial irradiation. Radiation doses ranged between 1840 Gy; the corresponding biological effective doses to the h-p axis were calculated as previously described (10) and ranged between 27.286.4 Gy. All patients had normal GH status as defined by their peak GH responses to both the insulin tolerance test (ITT) and the combined GHRH plus arginine stimulation test (GHRH+AST) (see Table 2
).
|
|
Study protocol
The study was approved by the South Manchester Local Research Ethics Committee, and informed consent was obtained from all subjects before testing.
GH profiling at 20-min intervals over 24 h was performed in the fed and the last 24 h of 33-h fast, as described previously (11, 12). Of the 16 patients and 30 normal controls, fasting profiles were undertaken in 10 patients (three women) and 14 normal controls (three women). Women were profiled in the first half of their menstrual cycle, and none had taken any oral contraception for at least 6 months before the study.
Assays
Serum samples from the fed and the fasting profiles for each subject were analyzed in duplicate in the same assay run using the modified Nichols Luma Tag hGH chemiluminescence immunometric assay (Nichols Institute Diagnostics, Bad Vilbel, Germany) (11). Serum IGF-I was determined by a two-site chemiluminescent immunometric assay and the free fatty acids (FFA) and 3-hydroxybutyrate (3-HOB) were analyzed as described previously (11, 12).
Analysis of GH concentration profiles
Cluster algorithm for pulsatility analysis (13), Cosinor analysis for appraisal of diurnal variation (14) and approximate entropy statistic to appraise secretory orderliness (15), and calculation of the pulsatile AUCGH/total AUCGH (area under the curve for GH) were performed as described previously (11).
Diagnostic thresholds
Definition of GH status using the pattern of responses to stimulation tests relied on the previously published diagnostic thresholds using noncompetitive immunoradiometric GH assay (10, 16). The definition of GH status on physiological testing (profiling) relied on the cutoff levels generated from our previous studies of GH secretion in the fed and fasting states (11, 12); cutoff levels for normal responses are shown in Table 2
.
Statistical analysis
The data were expressed as mean ± SEM if normally distributed, or as median and ranges if the data were skewed. Differences between groups were examined by the t test if the data were normally distributed or the Mann-Whitney rank-sum test if the data were skewed. Statistical significance was accepted at P < 0.05.
| Results |
|---|
|
|
|---|
|
Despite individual normality of the peak GH responses to both the ITT and GHRH+AST, the overall responses were markedly and significantly reduced by more than 40% (Tables 4
and 5
and Fig. 1
).
|
|
|
Gender-specific comparisons in the fed and fasting states revealed that patients and normal subjects have similar GH secretory profiles with no differences in the timing or amplitude of the diurnal variation (data of cosinor analysis not presented), pulsatile attributes of GH secretion (pulse frequency, pulse duration, and interpulse interval) or the profile mean and absolute GH levels (Tables 46![]()
![]()
and Fig. 1
). In the fed state, the pulsatile AUCGH/total AUCGH was significantly reduced (Tables 4
and 5
).
|
| Discussion |
|---|
|
|
|---|
The first suggestion of GHNSD after h-p axis irradiation arose from the study of Chrousos et al. (17), in which two monkeys irradiated with 40 Gy showed impaired spontaneous GH secretion compared with two normal controls. GHNSD was proposed in view of the "normal" peak GH responses to arginine infusion, though the responses to the ITT were diminished. In humans, the notion that radiation might cause GHNSD arose from a number of studies that showed reduced prepubertal and/or pubertal spontaneous GH secretion after cranial irradiation for leukemia (1824 Gy); these studies, however, failed to examine stimulated GH secretion to confirm the presence of a neurosecretory defect (6, 7, 9, 18). Whereas, in studies in which stimulated GH secretion was coexamined, the reduction in prepubertal and/or pubertal spontaneous GH secretion after cranial irradiation for leukemia (1824 Gy) was associated with significant reduction in overall peak GH responses to stimulation tests including the GHRH test despite normality of the peak responses in individual patients (5, 8, 19, 20). Similarly, reduction in spontaneous GH secretion in children irradiated for brain tumors (>40 Gy) was associated with blunting of stimulated peak GH responses to the ITT and/or arginine infusion (21, 22) and a 50% reduction in the overall peak GH responses to GHRH, despite normality of the individual responses in most patients (23).
The results of these studies seem to have led to a general acceptance that radiation-induced GHNSD is probably a real entity. However, these studies failed to demonstrate a pure neurosecretory defect; the problem leading to the belief that GHNSD existed in the irradiated patients had been created partly by the definition of normality to a pharmacological test being an all-or-none phenomenon, i.e. a level that passes a certain threshold rather than that which clearly exists, a continuum. Analysis of the group data illustrates this well with obvious reduction in overall peak GH responses in all studies. The reduction in peak GH responses indicates a degree of h-p axis damage that is not sufficient to reduce stimulated responses below diagnostic thresholds but severe enough to impair the h-p axis capacity to meet the demand for increased GH secretion during puberty when GH secretion would normally be amplified 2- to 3-fold (24, 25).
Under basal conditions, a reduction in maximal somatotroph secretory capacity should result in hyperstimulation or compensatory overdrive of the residual functioning part of the somatotropic axis through various mechanisms involved in the regulatory control of GH secretion (25) to restore normality of GH secretion. This is clearly supported by our findings, where, despite reduction in maximal somatotroph secretory reserve evidenced by reduced but "normal" responses to direct stimulation with GHRH+AST, spontaneous GH secretion remained fully preserved. Thus, partial damage of the axis might not affect GH levels in adults; however, in children, where GH secretion is more important than it is in adults, failure of the irradiated h-p axis to meet the demands for increased GH secretion during growth and puberty may be explained by decompensation of a partially damaged somatotropic axis, rather than GHNSD as had been previously suggested (4, 5, 6, 7, 8, 9). Failure of GH secretion to rise adequately in irradiated patients during puberty can be explained by two mechanisms. Firstly, reduced secretory capacity due to partial damage of the somatotropic axis, and secondly and perhaps more critical is the presence of "near maximal" activation of the h-p axis allowing for no further amplification during puberty. In support of this explanation, Lannering and Albertsson-Wikland (23) studied 19 children irradiated at age 116.9 yr with greater than 40 Gy. The overall peak GH responses to GHRH were reduced by 50%, and many children had "normal" responses. The peak GH response to AST+ITT (in 10 patients) was reduced, but still greater than 14 mU/liter in 30% of patients. Only one patient had a peak GH response above 20 mU/liter (10 µg/liter). The mean 24-h GH was reduced (by about 75%) in all prepubertal and pubertal children compared with control groups with constitutional short stature, and prepubertal and pubertal children had similar 24-h GH secretion, i.e. GH secretion failed to rise during puberty. Similarly, Moell et al. (5) studied 13 prepubertal/pubertal girls irradiated with 2024 Gy. The peak GH responses to GHRH were reduced by 60%, whereas spontaneous GH secretion was reduced by 85%, and GH secretion failed to rise during puberty.
Apart from studies in irradiated patients, there are a number of other studies that showed a profound reduction (3050%) in the overall peak GH responses (somatotroph reserve) compared with normals using various stimuli including GHRH (26, 27) and GHRH+AST (28) in children with GHNSD as well as in adults who had previously been diagnosed and treated for "classical idiopathic GHNSD" in childhood (29).
The preservation of spontaneous GH secretion in the fed state and also when the h-p axis is stressed by fasting in the context of substantially reduced somatotroph reserve argues against somatotroph dysfunction being solely explained by hypothalamic damage with secondary atrophy due to GHRH deficiency, as the combined effects of reduced GHRH secretion with consequent somatotroph atrophy would be expected to result in reduced spontaneous GH secretion. Therefore, our findings are most likely explained mechanistically by a combination of direct (primary) radiation-induced pituitary damage with hypothalamic compensation (overdrive of the residual somatotrophs). In addition, the reduction in pulsatile AUCGH/total AUCGH is suggestive of a reduction in peak levels with an equivalent increase in the interpeak levels. This is attributed to radiation-induced reduction in somatostatin tone that results in relatively higher nadir levels and consequently reduced peak levels. This is analogous to the fasting-induced decrease in somatostatin tone that also leads to increased nadir levels (11, 12).
Thus, collectively, previous studies that demonstrated established radiation-induced somatotroph dysfunction evidenced by an overall reduction in GH responses to GHRH by 5080% despite apparent normality of individual responses (30, 31, 32, 33, 34) and current data strongly suggest that radiation doses of less than 40 Gy produce direct damage to both the pituitary and the hypothalamus. These conclusions are in accord with earlier observations that showed marked sensitivity of the somatotrophs after in vitro irradiation of rat pituitary cell culture with single doses as low as 300 cGy (35) and the infrequent/rare occurrence of hyperprolactinemia (characteristic marker of hypothalamic damage) after low-dose (<40 Gy) irradiation as opposed to its high frequency after intensive (>40 Gy) irradiation (36, 37, 38). The extent of radiation damage at these two sites and the intensity of the compensatory mechanisms must vary between individuals and ultimately determine the pattern of GH responses to various stimuli and the nature of discordancy between stimulated and spontaneous GH secretion at different phases of life, i.e. adulthood vs. puberty.
In addition, one must not equate radiation-induced hypothalamic damage with GHRH deficiency, as there is normally a huge redundancy in the system, which means that overactivity of the residual functioning GHRH neurons may still result in supranormal GHRH release to compensate for reduced somatotroph reserve. Studies have shown that GHRH is essential for GH nocturnal increase (39, 40) and pulsatility (41). Therefore, the preservation of GH pulsatility and diurnal variation in the current cohort of irradiated patients and even in patients with severe radiation-induced GHD (11, 12) provides further evidence against the concept of established pituitary atrophy being purely secondary to significant hypothalamic damage and GHRH deficiency. Furthermore, the greater attenuation in GH responses to the ITT compared with GHRH in many studies may not necessarily reflect reduced hypothalamic GHRH but rather underlying hypothalamic "hyperactivity," which restrains further stimulation with the ITT, in exactly similar fashion to that which is seen in normal women who have increased spontaneous GH secretion (11, 25) mediated by increased hypothalamic agonist activity (increased GHRH release) (42, 43), yet show an almost 50% reduction in GH responses to an ITT as shown in this and previous studies (44, 45, 46, 47), with similar (this study) or even increased GH responses to the GHRH+AST test (43, 48) compared with normal men.
In conclusion, this study provides novel evidence that radiation-induced GHNSD in adult cancer survivors either does not exist or occurs very rarely and that radiation with doses less than 40 Gy possibly causes direct damage to both the pituitary and the hypothalamus with reduced overall peak GH responses to stimulation tests despite normality of individual responses. Compensatory overactivity of the partially damaged axis (overactivity of the residual functioning GHRH neurons and somatotrophs) operates to restore normality of spontaneous GH secretion. We speculate that failure of further activation of the already hyperstimulated partially damaged h-p axis during puberty when there is substantially increased demand for GH secretion results in the entity previously attributed to GHNSD.
| Acknowledgments |
|---|
| Footnotes |
|---|
Disclosure Statement: K.H.D. has nothing to declare and was previously employed by the Christie Hospital NHS Trust. S.S.P. has nothing to declare and is currently employed by the University of Virginia. M.O.T. has received consulting fees and lecture fees from Novo Nordisk and Merck amounting to less than $10,000 per year and is currently employed by the University of Virginia. S.M.S. has received consulting fees from Novo Nordisk and Pfizer, and lecture fees from Eli Lilly, Ipsen, Novo Nordisk, and Pfizer and was previously employed by the Christie Hospital NHS Trust.
First Published Online February 6, 2007
Abbreviations: AUCGH/total AUCGH, Area under the curve for GH; FFA, free fatty acid; GHD, GH deficiency; GHNSD, GH neurosecretory dysfunction; GHRH+AST, combined GHRH plus arginine stimulation test; 3-HOB, 3-hydroxybutyrate; h-p, hypothalamic-pituitary; ITT, insulin tolerance test; SDS, SD scores.
Received November 27, 2006.
Accepted January 30, 2007.
| 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 |