Androgen-Dependent Somatotroph Function in a Hypogonadal Adolescent Male: Evidence for Control of Exogenous Androgens on Growth Hormone Release Nelly Mauras,
Robert M. Blizzard, and Alan D. Rogol
A 14’0/rz-year-old white male with primary gonedal failure following testicular irradiation for acute lymphocytic leukemia was evaluated for poor growth. He had received 2400 rad of prophylactic cranial irradiation. The growth velocity had decelerated from 7 to 3.2 cm/yr over 3 years. His bone age was 120/(a years (by TWZ-FLUS). and his peak growth hormone (GH) response to provocative stimuli was 1.4 ng/mL. The 24hour GH secretion was studied by drawing blood every 20 minutes for 24 hours. The resulting GH profile was analyzed by a computerized pulse detection algorithm, CLUSTER. Timed serum GH samples were also obtained after a 1 #g/kg IV bolus injection of the GH releasing factor (GRH). The studies showed a flat 24hour profile and a peak GH response to GRH of 3.9 ng/ml. Testosterone enanthate treatment was started, 100 mg IM every 4 weeks. Ten months after the initiation of therapy the calculated growth rate was 8.8 cm/yr. The 24hour GH study and GRH responses were repeated at the time, showing a remarkably normal 24hour GH secretory pattern and a peak GH response to GRH of 14.4 ng/mL. Testosterone therapy was discontinued, and 4 months later similar studies were repeated. A marked decrease in the mean 24hour GH secretion and mean peak height occurred, but with maintenance of the GH pulse frequency. The GH response to GRH was intermediate, with a peak of 8 ng/mL. There was no further growth during those 4 months despite open epiphyses. In conclusion, these data indicate that androgens may affect the function of previously quiescent somatotrophs, and that such androgen action may be necessary for the maintenance of normal GH secretory function. This may be either a direct androgen effect on the pituitary, or an indirect effect mediated through a modulation on the patterns of somatostatin and GRH secretion by androgen action. o 1SS~ by Grune & Stratton, Inc.
URING pubertal development in humans the spontaneous growth spurt is accompanied by rising serum testosterone concentrations and may be due, at least in part, to enhanced growth hormone (GH) secretion. Recently, we reported that the augmentation in GH secretion that occurs during either spontaneous puberty or exogenous testosterone therapy in boys, is a pulse-amplitude-modulated phenomenon, relatively independent of changes in pulse frequency.’ Even though these data strongly support a positive interaction between the sex steroids and GH secretion in children with established GH secretory patterns, evidence for a direct effect of androgens on the function of quiescent somatotrophs is limited.* Here we report a unique patient in whom the effect of exogenous androgen therapy on previously quiescent somatotroph function was carefully studied.
D
His evaluation included the measurement of pituitary gonadotropin concentrations, which were greatly elevated with a mean 24-hour luteinizing hormone (LH) concentration of 21.6 mIU/mL, a mean follicle-stimulating hormone (FSH) concentration of 56 mIU/mL, and a serum testosterone level of 125 ng/dL, which is in the prepubertal range, indicative of primary gonadal failure. Other pituitary function studies included a TSH of 3.4 mIU/mL, and a prolactin of 7.8 ng/mL. The T, was 7.5 pg/dL, and the morning serum cortisol was 21 pg/dL. A pharmacological stimulation test of GH secretion was performed using an arginine infusion and insulin-induced hypoglycemia. The peak GH response to these stimuli was 1.4 ng/mL, compatible with GH deficiency. MATERIALS
CASE REPORT
PL was 14io/12years old when evaluated by us for short stature. His height was 158 cm and his weight 64.5 kg. He had received standard chemotherapy for acute lymphocytic leukemia (ALL) when he was 3 years old, as well as 2,600 rad of prophylactic cranial irradiation. At 7 years of age he had a testicular relapse of his leukemia and received 2,500 rad of gonadal irradiation. He had been in continuous bone marrow remission since. The growth velocity had decelerated markedly from 7 cm/yr to 3.2 cm/yr over the preceding 3 years. The bone age was 120/1*years at the time we saw him.
From the Departments of Pediatrics and Pharmacology at the University of Virginia Medical Center, Charlottesville. Supported in part by NII-I Clinical Research Grant No. RRO0847. Address reprint requests to Nelly Mauras. MD, Nemours Children’s Clinic, PO Box 5720. Jacksonville, FL 32247. o 1989 by Grune di Stratton, Inc. 0026-0495/89/3803-0017%03.00/0 266
AND METHODS
Study Design The patient was admitted three times to the University of Virginia Clinical Research Center, and a history and physical examination were performed. A radiograph of the left hand and wrist was obtained for bone age determination by the TW2 (RUS) method.’ An intravenous (IV) heparin-lock needle was placed in a forearm vein for venous sampling for serum GH determinations every 20 minutes. Sampling was continued for 24 hours with special efforts made not to disturb his sleep. After the 24-hour sampling period, 1 pg/kg of the growth hormone releasing hormone (GRH) was given as an IV bolus dose, and sampling continued at 15minute intervals for serum GH determination for an additional two hours. The patient was then started on testosterone enanthate therapy, 100 mg intramuscularly (im), which was continued every 4 weeks for 10 months. An identical study was performed after the 10 months of testosterone therapy within three days after the last injection, and 4 months after its discontinuation.
Assays Serum GH concentrations were measured in duplicate using the Nichols Institute Diagnostic human GH immunoradiometric assay kits (San Juan Capistrano, CA). The intrassay coefficient of variaMetabolism,Vol 38, No 3 (March), 1989: pp 286-289
287
EFFECT OF ANDROGENS ON GH RELEASE
Table 1. OH Pulse Analysis
Baseline
4MoAftu Discontinuation of TQ
10MoAfterTFj
Chronological age fy~)
14%
1%
16%2
Growth velocity km/yrt
3.2
8.6
Nogowth
Bone age (vr) Testosterone (ng/dL)
12% <25
14% 552
14% t25
Somatomedin C (U/mL)
0.45
Peak GH response to GRH (ng/mL)
3.9
Mean 24-h GH concentration (ng/mL) Mean pulse frequency (pulses/24 h) Mean pulse amplitude (na/mL)
0.92
1.0
14.4
8.0
0.97
3.2
1.4
2
7
2.2
7.0
10 3.4
Abbreviation: TF$, testosterone treatment. tion was below 101, with a sensitivity of 0.8 ng/mL. Samples from each admission were measured in duplicate in the same assay run. Plasma somatomedin C concentrations were measured by the Nichols Institute, using the modified method of Furlanetto et al.“.5 Serum testosterone concentrations were measured using solid-phase RIA Coat-a-Count kits obtained from the Diagnostic Products Corporation (Los Angeles).
Pulse Analysis To detect significant hormone excursions (pulses) in the serum GH concentrations series, we applied CLUSTER analysis.6 Using this program, a pulse is defined as a statistically significant increase in a cluster of hormone values, followed by a statistically significant decrease in a second cluster of values. In the analysis we specified the number of points in testing prepeak and postpeak nadirs (one point) Ypeaks (two points), the median intraseries coeflicient of variation, and a r statistic to identify a significant increase and decrease. These parameters had previously been shown to constrain the false-positive pulse detection rate to t5%.’ RESULTS
The data are summarized in Table 1. When the patient was first evaluated at 14r%r years, the peak GH response to the GRH infusion was only 3.9 ng/mL. The 24-hour GH secretory profile showed a very low GH output as indicated by a low 24-hour mean GH concentration (0.97 ng/mL) and a low pulse frequency (2 pulses/24 h) and pulse amplitude (2.2 ng/mL, Table 1, Fig 1A). These data, as well as the poor growth and delayed bone age, prompted the diagnosis of growth hormone deficiency. GH therapy was offered but declined by the patient, who began testosterone therapy as described. During the 10 months after the initiation of therapy the growth velocity had accelerated to 8.6 cm/yr. Because of the unexpected growth response in the presence of GH deficiency, we elected to repeat the 24-hour GH and GRH study. While receiving testosterone therapy, the patient’s peak serum GH response to the GRH was significantly greater than at baseline (14.4 ng/mL). There was a marked augmentation of the mean 24-hour GH concentration (3.2 ng/mL), pulse frequency (7 pulses/24 h), and pulse amplitude (7 ng/mL), comparable to normal boys his age’ (Table 1, Fig 1B). Due to the apparent normalization of the GH secretory pattern after testosterone therapy, we elected to stop the testosterone and repeat the studies 4 months later to assess if the improvement was merely initiated, or if it also had been
maintained by testosterone. There was no further growth during those 4 months without therapy, despite open epiphyses. This time the mean GH response to GRH was intermediate (8.0 ng/mL), with a marked decrease in the mean 24-hour GH concentration (1.4 ng/mL) and pulse amplitude (3.4 ng/mL), but with maintenance of the pulse frequency (10 pulses/24 h) (Table 1, Fig 1C). For comparison, the 24-hour GH profiles of two normal boys, one prepubertal and another pubertal are shown in Fig 2. DISCUSSION
Due to markedly improved survival of childhood ALL, many investigators have looked at the effects of treatment on the endocrine function of individuals who have survived the disease. Although when higher dosages of cranial irradiation (ie, ~2,900 rad) are used (as for posterior fossa tumors), GH deficiency is not uncommon; when lesser dosages are used, such as those used in CNS prophylaxis for ALL, the data are sometimes contradictory. Shalet et al8 reported decreased GH reserve in children treated with 2,400 rad, yet the same authors later reported normal growth despite the biochemical finding of GH deficiency.’ Using 24-hour GH secretory profiles we” and others” have found decreased GH output under physiological conditions in those children treated with prophylactic cranial irradiation for ALL. Because of that, the use of GH therapy has been advocated in children treated for ALL who are not growing well.‘* Cranial irradiation per se, however, has not been shown to adversely affect the function of gonadotropes, and in general, gonadal function is preserved after cranial irradiation.‘0*13 The remarkable changes in the GH secretory dynamics on this patient represent a unique biologic model of the interaction of sex steroid hormones and GH secretion. We have previously reported that the marked increase in GH secretion either during spontaneous puberty or after exogenous testosterone therapy in boys represents a change in the amplitude of the GH pulses relatively independent of any change in pulse frequency.’ Those data, though, represent changes in GH secretion in children who were already producing GH. Penny and Blizzard’ had previously reported that in three males with isolated GH deficiency, GH secretion (as measured by the response to insulin-induced hypoglycemia and arginine infusion) increased when those children entered puberty. Similarly, in the patient reported here we have data to suggest activation, by exogenous testosterone therapy, of
288
MAURAS, 8LIZTARD. AND RCGCL
SASELINE
I
7.5
?
i
z
I 5.0 (3 1.s
I
0 Ai
I
210
so0
IS0
1000
12so
258
1500
(mid
-TIME
see
750
1008
7%
1809
e
lP.S
1
g
7.s
\ z 5.0 g 2.S
0.0 0
B
PSO 4
I
sOOT,M~;min;OoO
258
coo
1350
588
1258
1568
BB
TIME (min) Berum OH concentrations were determined by drawing Fig 2. blood every 20 minutes for 24 hours in two subjects. IA) represents the OH profile on a ll%r-year-old prepubertal male of normal size: mean 24hour OH concentration was 1.8 ng/mL. mean pulw frequency 4 pulses/24 h, and mean pulse amplitude 8.4 ng/mL. (Bj represents comparable data on a 14%year-old pubertal male (Tanner stage IV-V); mean P&hour OH conoentration was 4.8 ng/mL, mean pulse frequency 8 pulses/24 h. and mean pulse amplitude 16.4 ng/mL. Upward deflectiona represent the pukes determined by CLUSTER analysis. Please note the dlffarences in the ordinates.
c
o4
260
S;*ME7~min;000
12so
;oo
OH secretory profile in the patient before any therapy Fig 1. (A), 10 months after testosterone therapy (6). and 4 months after discontinuation of androgen therapy (C). Blood was sampled at 20-minute intervals for 24 hours beginning at 8 AM. Upward deffections represent the pulses detected by CLUSTER analysis.
previously quiescent somatotrophs. We found increased mean 24-hour GH concentration and increased GH pulse frequency and pulse amplitude, as well as a markedly accelerated growth rate after 10 months of treatment. It appears that sex steroids were necessary not only to activate, but also to maintain a normal GH secretory pattern even
though the frequency of GH pulses was sustained after discontinuation of testosterone therapy. Measuring the GH response to GRH and using continuous perfusion techniques on anterior pituitary cells of intact, castrated, and castrated and testosterone-treated male rats, Evans et al” showed evidence to suggest that exposure of the anterior pituitary gland in situ to testosterone correlates with enhanced GRH-stimulated GH release in vitro. Recent data, using the reverse hemolytic plaque assay also indicated that testosterone enhances GH secretion by increasing the secretory capacity, but not the sensitivity of somatotrophs to GRH by recruiting the function of a subpopulation of somatotrophs.‘5 Our data could suggest that the activation of the GH-producing cells in our patient may be directly controlled at the pituitary level, since he had blunted GH responses to the GRH injection prior to testosterone treatment and normal responses afterwards. However, since the
299
EFFECT OF ANDROGENS ON GH RELEASE
net secretion of GH is dependent on the underlying &radian rhythms of GRH and somatostatin secretion,16 androgen therapy may have also enhanced the GRH secretion and/or decreased the influence of somatostatin on GH secretion in this patient. It is possible, that the remarkable activation of somatotroph activity that we saw in this patient may also be due to a maturational process of the GRH receptor in response to the action of sex steroids.
ACKNOWLEDGMENT
We are deeply grateful to Sandra Jackson and the excellent nursing staff of the Clinical Research Unit for their devoted care of our patients, to Catherine Cassada and Anthony Amos for running the RIAs, to Dr Jay McDonald for assistance on the care of this patient, to Dr Michael 0. Thomer for providing us with the growth hormone releasing hormone, and to Diane Goin for typing this
manuscript.
REFERENCES
1. Mauras N, Blizzard RM, Link K, et al: Augmentation of growth hormone secretion during puberty: Evidence for a pulse amplitude-modulated phenomenon. J Clin Endocrinol Metab 64:596-601,1987 2. Penny R, Blizzard RM: The possible influence of puberty on the release of growth hormone in three males with apparent isolated growth hormone deficiency. J Clin Endocrinol34:82-84,1972 3. Tanner TM, Whitehouse RH, Cameron N, et al: Assessment of skeletal maturity and prediction of adult height (TW2 method). San Diego, Academic, 1983 4. Furlanetto RW, Underwood LE, Van Wyk JJ, et al: Eatimation of somatomedin-C levels in normals and patients with pituitary disease by radio-immunoassay. J Clin Invest 60:648-657,1977 5. Furlanetto RW: Pitfalls in the somatomedin-C radio-immunoassay. J Clin Endocrinol Metab 54:1084-1086,1982 6. Veldhuis JD, Johnson ML: Cluster analysis: A simple versatile and robust algorithm for endocrine pulse detection. Am J Physiol 250:E485-E493,1986 7. Evans WS, Faria ACS, Christiansen E, et al: Impact of intensive venous sampling on characterization of pulsatile GH release in normal men. Am J Physiol 15:E549-556, 1987 8. Shalet SM, Beardwell CG, Towmey JA, et al: Endocrine function following the treatment of acute leukemia in childhood. J Pediatr 90:920-923, 1977 9. Shalet SM, Price DA, Beardwell CG, et al: Normal growth
despite abnormalities of GH secretion in children treated for acute leukemia. J Pediatr 94719-722, 1979 10. Mauras N, Sabio H, Rogol AD: Neuroendocrine function in survivors of childhood ALL and NHL: A study of pulsatile GH and gonadotropin secretions. Am J Pediatr Hematol Oncol l&9-17, 1988 11. Blatt J, Bercu BB, Gillin JC, et al: Reduced pulsatile GH secretion in children after therapy for ALL. J Pediatr 104:182- 186, 1984 12. Romshe CA, Zipf WB, Wiser A, et al: Evaluation of growth hormone release and human growth hormone treatment in children with cranial irradiation-associated short stature. J Pediatr 104:177198,1984 13. Siris Es, Leventhal BG, Vaitukaitis JL: Effects of childhood leukemia and chemotherapy on puberty and reproductive function in girls. N Engl J Med 294:1143-1146, 1976 14. Evans WS, Krieg RJ, Limber ER, et al: Effects of in vivo gonadal hormone environment on in vitro hGRF-40 stimulated GH release. Am J Physiol249:E276-E280, 1985 15. Ho Ky, Horner MO, Kneis RJ, et al: Effects of gonadal steroids on somatotroph function in the rat: Analysis by the reverse hemolytic plaque assay. (submitted) 16. Tannenbaum GS, Ling N: The interrelationship of growth hormone releasing factor and somatostatin in generation of the ultradian. Endocrinology, 115:1952-1957, 1984