Relationship between the Morphological Evaluation of the Pituitary and the Growth Hormone (GH) Response to GH-Releasing Hormone Plus Arginine in Children and Adults with Congenital Hypopituitarism¹

The relationship between the hypothalamus-pituitary morphology and the somatotroph responsiveness to maximal provocative tests exploring the GH releasable pool is still unclear. We evaluated the GH-releasing effect of GHRH plus arginine (GHRH plus Arg) in 36 patients with congenital GH deficiency (GHD) according to their pituitary magnetic resonance imaging findings, consisting of anterior pituitary hypoplasia, stalk agenesis (neural and or vascular component), and posterior pituitary ectopia. Seventeen children (12 boys and 5 girls, aged 1–5.2 yr) were evaluated at the time of diagnosis of GHD (mean age, 3.6 ± 1.4 yr), and 19 adults (13 males and 6 females, aged 15.9–28.6 yr) with childhood-onset GHD were reevaluated after completion of GH treatment (at least 6 months of withdrawal) at a mean age of 20.5 ± 3.5 yr. Eleven children had isolated GHD, and 6 had multiple pituitary hormone deficiency (MPHD) whereas 7 adults had isolated GHD, and 12 had MPHD. A residual vascular component of the pituitary stalk was visualized in 7 children and 7 adults with isolated GHD, whereas magnetic resonance imaging showed complete pituitary stalk agenesis (both vascular and neural components) in 10 children and 10 adults, including 16 with MPHD (6 children) and 4 children with isolated GHD. In the children, the median peak GH response to GHRH plus Arg (7.6 μg/L; range, 2.4–40.2 μg/L) was significantly higher than that in the adults (1.8 μg/L; range, 0.8–37.4 μg/L; P = 0.0039); it was also significantly higher in the isolated GHD patients (18 μg/L; range, 3.3–40.2 μg/L) than in those with MPHD (1.9 μg/L; range, 0.8–7.6 μg/L; P = 0.00004). In the patients with residual vascular component of the pituitary stalk the median peak GH responses to GHRH plus Arg (19.1 μg/L; range, 1.6–40.2 μg/L) was significantly higher than that in patients with complete pituitary stalk agenesis (2.2 μg/L; range, 0.8–8.8 μg/L; P = 0.00005). There was a trend toward a decrease with age in peak GH response to GHRH plus Arg. Mean serum insulin-like growth factor I (IGF-I) levels were 36 ± 7.1 μg/L in the children and 63.5 ± 22.6 μg/L in the adults (P = 0.0001). The mean IGF-I level did not differ between the children with (35.7 ± 4.8 μg/L) and those without (36.3 ± 8.7 μg/L) the pituitary stalk; it was much higher in the adults with residual vascular pituitary stalk (81.1 ± 17.7μ g/L) than in those with complete pituitary stalk agenesis (47.7± 12.5 μg/L; P = 0.0002). The IGF-I level was 36.1 ± 6.7 μg/L in the isolated GHD children and 36 ± 8.6μ g/L in those with MPHD; levels were 82.1 ± 19.4 and 52.7± 16.8 μg/L respectively, in the adults (P = 0.003).

In this study we have confirmed that the partial integrity of the hypothalamic pituitary connections is essential for GHRH plus Arg to express its GH-releasing activity and have shown that this provocative test is able to stimulate GH secretion to a greater extent in those patients with GHD, but with a residual vascular component of the pituitary stalk. This test is reliable in the diagnosis of congenital hypopituitarism in both children and adults when associated with complete pituitary stalk agenesis and MPHD. In younger children with congenital GHD but less severe impairment of the pituitary stalk the GH response to GHRH plus Arg may be within the normal range; deterioration of pituitary GH reserve with a GH response of less than 10 μg/L after 20 yr of age makes this test very sensitive in the diagnosis of adult GHD.

THE RELIABILTY OF pharmacological provocative tests in the diagnosis of GH deficiency (GHD) in children is controversial for several reasons, and the high frequency of transitory GH defect during childhood has been attributed also to the limitations of GH testing in the absence of a diagnostic gold standard (1–3). Although patient selection is one of the most important criteria in the diagnosis of GHD (4), there are no data on the reliability and reproducibility of provocative GH tests in patients sharing the same etiology of pituitary dysfunction. Moreover, the majority of the data on GH tests refer to mid-childhood, whereas today the management of patients with suspected GHD younger than those in the past raises additional questions about the age dependency of GH responsiveness to standard tests. Children with growth delay are now referred to the pediatric endocrinologist at a younger age than previously, and the use of combined provocative tests that may obviate current diagnostic limitations is still awaited.

Adult GHD is diagnosed when the GH response is less than 3 μg/L after insulin-induced hypoglycemia (ITT), which appears to be the most reliable provocative test, whereas clonidine, arginine, and glucagon are less discriminatory (5–9). The ITT test, however, is associated with frequent symptomatic hypoglycemia, as is also shown in our recent study, in the great majority of patients with permanent GHD of childhood-onset (4). Thus, alternative provocative tests are needed for patients in whom ITT is contraindicated (10), including the very young at presentation. The diagnostic value of GHRH in combination with arginine (Arg) has been demonstrated in both children and adults with hypopituitarism (6, 11–13). Nevertheless the GH response and the maximal secretory capacity of residual somatotrope cells have not been tested after GHRH plus Arg in prepubertal children younger than 6 yr and in adults with childhood-onset GHD in whom the magnetic resonance imaging (MRI) features of the hypothalamic-pituitary region were compatible with a congenital origin. In fact, MRI has been of great value in the diagnosis of disorders of the hypothalamic pituitary area (14). More recently, the use of contrast medium has allowed identification of abnormalities of the hypothalamic-pituitary stalk in patients with GHD, making it possible to define patients with complete agenesis of the pituitary stalk (neural and vascular components) and those with residual vascular component of the pituitary stalk (15). Interestingly, the absence of the vascular pituitary stalk was closely associated with the presence of multiple pituitary hormone deficiencies (MPHD), whereas the presence of a residual vascular component was predictive of a less severe impairment of pituitary function (15).

The aim of this study was to investigate the diagnostic accuracy of GHRH plus Arg in young children (under 6 yr) and adults (childhood-onset) with GHD associated with congenital hypothalamic-pituitary malformations such as anterior pituitary hypoplasia, ectopic posterior pituitary, and partial or complete pituitary stalk agenesis. To this end, we evaluated the GH-releasing effect of GHRH plus Arg compared with the responses to classical pharmacological tests in children and adults with isolated GHD or MPHD who share identical hypothalamic-pituitary abnormalities at MRI.

Subjects and Methods

The study population comprised 17 children with newly diagnosed hypopituitarism (12 boys and 5 girls) and 19 adults (13 males and 6 females) with childhood-onset GHD diagnosed at a mean age of 10.2± 3.7 yr by failure to respond to at least two pharmacological provocative tests. The data of some of these latter patients had been partially reported (4).

All of the children underwent complete studies of pituitary function at a mean age of 3.6 ± 1.4 yr as previously described (14). The mean height sd score was −2.6± 0.7, bone age was 1.9 ± 0.5 yr, and body mass index (kilograms per meter squared) was 16.3 ± 2.9. In particular, GH function was studied after Arg (0.5 g/kg BW l-arginine monohydrochloride given iv over 30 min) and insulin (0.1 U/kg BW, iv bolus) or glucagon (1 mg, im). The blood samples for GH determinations were obtained at 0, 30, 60, 90, and 120 min after Arg and insulin and at 150 and 180 min after glucagon. Eleven children (7 boys and 4 girls) had isolated GHD, and 6 (5 boys and 1 girl) had MPHD. Partial GHD (GH peak after insulin and Arg tests, 5.0–9.9 μg/L) was found in 4 of them, whereas the remaining 13 patients had a GH peak below 5.0 μg/L.

In the adult patients GH secretion was reevaluated after ITT and Arg at least 6 months from GH discontinuation at a mean age of 20.5 ± 3.5 yr. Mean height sd score (−3.6 ± 1.1 yr at the time of presentation) was −1.7 ± 1.4 at the time of reevaluation, and body mass index was 22 ± 4 kg/m². Seven (4 males and 3 females) had isolated GHD and 12 (9 males and 3 females) had MPHD. Partial GHD (GH peak after ITT and Arg tests, 5.0–9.9 μg/L) was diagnosed in 3 of them at the time of presentation, whereas the remaining 16 patients had a GH peak below 5.0 μg/L. MPHD patients were receiving appropriate hormone replacement therapy.

The MRI technique with the use of contrast medium, as previously described (4), revealed an ectopic posterior pituitary at the infundibular recess in all subjects. A residual vascular component of the pituitary stalk was visualized in 7 children and 9 adults, whereas MRI showed complete agenesis of the pituitary stalk (vascular and neural components) in 10 children and 10 adults. Isolated GHD was found in 7 children with residual vascular component and in 4 with complete pituitary stalk agenesis; isolated GHD was found in 7 adults with residual vascular component and in none with pituitary stalk agenesis.

All subjects were tested with GHRH [GHRH-(1–29), GEREF, Serono, Italy; 1 μg/kg BW, iv, at time zero] plus Arg (0.5 g/kg BW l-arginine monohydrochloride given iv from 0–30 min). The study was approved by the Department of Pediatrics institutional review board, and written informed consent was obtained from the parents and adult patients. Blood samples were drawn from an indwelling catheter inserted in an antecubital vein at −15, 0, 15, 30, 45, 60, 75, and 90 min. All studies started between 0800–0900 h after the patients had fasted overnight. The serum insulin-like growth factor I (IGF-I) concentration was determined in all patients at the time of the GHRH plus Arg test. The main clinical findings are summarized in Tables 1¹ and 2².

Assay procedures

Serum GH levels were measured by fluoroimmunoassay using a commercial kit (Auto DELFIA hGH, E.G.&G. Wallac Oy, Finland). The intra- and interassay coefficients of variations were 5.1% and 2.5%, at 0.430 mU/L, 2.7% and 2.1% at 5.0 mU/L, and 2.2% and 1.4% at 21.1 mU/L. Cross-reactivity was less than 0.001% for PRL and human placental lactogen. Serum GH levels were measured by RIA using a commercial kit (Sorin, Italy) in the adults at childhood presentation.

IGF-I was measured by RIA using a commercial kit (SM-C-RIA-CT, Biosource Technologies, Inc., Belgium). The intra- and interassay coefficients of variations were 6.1%, 4.1%, and 4.7% at 54.2 ± 3.3, 194 ± 8, and 491 ± 2 μg/L; 9.9%, 9.6%, and 9.3% at 121 ± 11, 251 ± 24, and 494 ± 46μ g/L. The sensitivity of the assay was 0.25 ± 0.1 μg/L. Normal age-related values in our laboratory are: 1–3 yr, 30–120μ g/L; 3–6 yr, 35–145 μg/L; 6–9 yr, 75–195 μg/L; 9–12 yr, 95–320 μg/L; 12–15 yr, 160–340 μg/L; and 15–21 yr, 210–520μ g/L.

Statistical analysis

Comparisons between groups were performed using the Mann-Whitney U test (when comparing two groups) or Kruskal-Wallis ANOVA analysis (comparing more than two groups); comparisons within groups were performed using the Wilcoxon test (for comparing the same group twice) or Friedman ANOVA analysis (more than twice). The odds ratio was calculated by Woolf’s method, and the significance of any deviation from unity was estimated by Fisher’s exact test; 95% confidence intervals were calculated for all odds ratios. P < 0.05 was considered to indicate statistical significance. All tests were two-sided; data are given as the mean ± se or as medians and ranges as appropriate.

Results

GH responses to provocative tests

The mean GH response after ITT and Arg or glucagon was higher in the children (3.8 ± 1.5 and 3.8 ± 0.9 μg/L) than in the adults at the time of reevaluation (1.1 ± 0.8 and 1.1 ± 0.7μ g/L; P = 0.0001). The mean GH response after ITT (2.7 ± 1.4 μg/L) and Arg (2.4 ± 1.1 μg/L) was higher in the adult patients at the time of presentation (during childhood evaluation) than at the reevaluation. The GH response after ITT and Arg was higher in the children than in the adults independently of the presence of the vascular pituitary stalk at MRI (3.2 ± 2.0 and 3.8 ± 1.2 μg/L vs. 1.3 ± 0.7 and 1.3 ± 0.7 μg/L; P = 0.0001) or of its complete agenesis (4.2 ± 0.9 and 3.8 ± 0.8 μg/L vs. 0.8 ± 0.6 and 0.8 ± 0.5 μg/L; P = 0.0001). The children with isolated GHD or MPHD showed a higher GH response after ITT (3.4 ± 1.7 and 4.3 ± 0.9 μg/L; P = 0.0001) and Arg (3.8 ± 1.1 and 3.9 ± 0.9 μg/L; P = 0.0001) than the adults (1.3 ± 0.7 and 0.9 ± 0.7 μg/L vs. and 1.3 ± 0.7 0.9 ± 0.6 μg/L; P = 0.0001).

In the children the mean peak GH response to GHRH plus Arg (14.7± 13.4 μg/L) was significantly higher than that in the adults (6.2 ± 9.4 μg/L; P = 0.03). It was also significantly higher in the patients with residual vascular component of the pituitary stalk (19.0 ± 13.5 μg/L) than in those with complete pituitary stalk agenesis (3.1 ± 2.4 μg/L; P = 0.000001). Moreover, the response was significantly higher in the patients with isolated GHD (17.9 ± 13.1 μg/L) than in those with MPHD (2.5 ± 2.0 μg/L; P = 0.000002).

The peak and median GH responses to GHRH plus Arg in the individual patients are shown in Figs. 1–3¹²³ according to the clinical characteristics, hormonal findings, and MRI features. In particular, the median peak GH response after GHRH plus Arg in the children (7.6 μg/L; range, 2.4–40.2 μg/L) was significantly higher than that in the adults (1.8 μg/L; range, 0.8–37.4 μg/L; P = 0.003). The GH peak was greater than 20 μg/L (normal value in children and adolescents) in five children and greater than 16.5 μg/L (normal value in adults) in three adults; none of the patients more than 20 yr old had a peak GH response above 3 μg/L. In the isolated GHD patients the median peak GH response to GHRH plus Arg (18 μg/L; range, 3.3–40.2 μg/L) was significantly higher than that in patients with MPHD (1.9 μg/L; range, 0.8–7.6 μg/L; P = 0.00004); it was significantly higher in the patients with residual vascular component of the pituitary stalk (19.1 μg/L; range, 1.6–40.2 μg/L) than in those with complete pituitary stalk agenesis (2.2 μg/L; range, 0.8–8.8 μg/L; P = 0.00005). A negative, but not significant, correlation was found between age at the time of GHRH plus Arg evaluation and peak GH response in both the children and the adults (Fig. 4).

Mean serum IGF-I levels were 36 ± 7.1 μg/L in the children and 63.5 ± 22.6 μg/L in the adults (P = 0.0001). Although the mean IGF-I level did not differ between the children with (35.7 ± 4.8 μg/L) and those without (36.3 ± 8.7 μg/L) the pituitary stalk, it was much higher in the adults with residual vascular pituitary stalk (81.1 ± 17.7 μg/L) than in those with complete pituitary stalk agenesis (47.7 ± 12.5 μg/L; P = 0.0002). The IGF-I level was 36.1 ± 6.7μ g/L in the isolated GHD children and 36 ± 8.6 μg/L in those with MPHD, whereas levels were 82.1 ± 19.4 and 52.7 ± 16.8μ g/L, respectively, in the adults (P = 0.003).

Symptomatic hypoglycemia during ITT was documented in the great majority of patients with MPHD and in 65% of patients with IGHD.

Discussion

This study shows that there is a close correlation between morphological impairment of the hypothalamic-pituitary unit and the maximal GH releasable pool evaluated by provocative testing with GHRH plus Arg in children and adults with congenital GHD. In fact, the patients with GHD and MRI evidence of anterior pituitary hypoplasia, stalk agenesis (neural and or vascular component), and posterior pituitary ectopia had a negligible GH response to GHRH plus Arg as well as to ITT. On the other hand, in the patients with a residual vascular component of the pituitary stalk, GHRH plus Arg induced a considerable pituitary GH releasable pool. The marked differences in GH response to ITT, Arg, and GHRH plus Arg between children and adults probably sharing a common etiology for GHD suggest a progressive impairment of the residual GH releasable pool during childhood and early adulthood. Our present findings also show that the GHRH plus Arg test is reliable for the diagnosis of GHD in both children and adults presenting with complete pituitary stalk agenesis and/or multiple pituitary hormone deficits. Moreover, although pituitary responsiveness to GHRH plus Arg in very young children does not definitively rule out an impaired GH response to ITT, GHRH plus Arg appear to be very sensitive in the diagnosis of young adult GHD patients.

MRI has greatly improved the diagnosis of disorders affecting the hypothalamic-pituitary area in patients with idiopathic hypopituitarism, showing not only anterior pituitary hypoplasia, posterior pituitary ectopia, and stalk agenesis, but even partial pituitary stalk abnormality with persistence of a residual vascular component (15). This raises the question of whether the functional link between hypothalamic neurohormones, i.e. GHRH and somatostatin, and somatotroph cell function in term of GH synthesis and release is at least partially preserved (15, 16). Our present data agree with this assumption, demonstrating that a provocative test as potent as GHRH plus Arg shows a pituitary GH releasable pool inversely correlated to the degree of morphofunctional impairment assessed by MRI. In fact, in the present study with GHRH plus Arg as well as in previous studies with GHRH or GHRP alone or with hexarelin (16–20), a considerable GH response was found in GHD patients with a residual vascular hypothalamus-pituitary connection. In the present as well as in previous studies, the GH responses to these stimuli were, however, lower than those obtained in normal subjects, suggesting that partial impairment of the GH releasable pool reflects the morphological alteration in the hypothalamic-pituitary unit.

The GHRH plus Arg test is one of the most powerful provocative tests of GH secretion (6, 11, 12) and, differently from classical provocative tests, including ITT, shows good intraindividual reproducibility (12). This test, given appropriate cut-off limits, has been proposed as a reliable alternative to the classical provocative tests for the diagnosis of GHD (13). In fact, it has been shown that testing with GHRH plus Arg distinguishes normal subjects from those with MPHD and that it is as sensitive as ITT for the diagnosis of GHD in adults and older adolescents (12). These data are confirmed by the present findings showing for the first time that in children under the age of 6 yr as well as in adults with congenital GHD (particularly if they have multiple anterior pituitary deficits) and MRI evidence of severe pituitary and stalk abnormalities, the GH response to GHRH plus Arg is clearly impaired as well as that to ITT. Thus, this study further indicates that the GHRH plus Arg test is reliable for the diagnosis of adult GH deficiency and that it distinguishes MPHD children and adults from normal subjects.

On the other hand, the present findings show that in younger children (5 yr of age) with congenital GHD but less severe impairment of the pituitary stalk the GH response to GHRH plus Arg may be within the normal range, thus failing to distinguish these patients with IGHD from normal subjects. This is probably due to residual functional link between hypothalamic GHRH activity and somatotroph cells allowing (some) GH synthesis and release. Reportedly, serum IGF-I concentrations do not correlate with GH status, as determined by provocative tests (3). In our study the children, in fact, had IGF-I levels within the lower normal range for age. Moreover, in this age group there was no difference in IGF-I values between patients with IGHD or MPHD and normal subjects, confirming that single IGF-I determination has limited value in the diagnosis of even congenital GHD during the first 5 yr of age. The potential for assessing GH status with a single estimation of the circulating IGF-I level, however, appears to be extremely attractive in young adults with childhood-onset GHD provided that evaluation of the IGF-I concentration has been performed after 6-month withdrawal of GH treatment, as shown by the results of our previous study (4). In particular, the IGF-I results discriminated better our adult patients with MPHD than those with IGHD.

In conclusion, this study emphasizes the close correlation between morphological impairment of the hypothalamus-pituitary unit and the maximal GH releasable pool evaluated by provocative testing with GHRH plus Arg in young children and adults with congenital hypopituitarism. Our present findings also show that this test is reliable for the diagnosis of GHD in both children and adults presenting with complete pituitary stalk agenesis and multiple deficits of anterior pituitary function; it remains true that the GHRH plus Arg test may fail to recognize GHD in children with partial impairment of the pituitary stalk. GHRH plus Arg was not associated with any remarkable side-effect, suggesting that it is safe in the evaluation of GH status in both in children and, especially, young adults with childhood-onset GHD. In our study the peculiar MRI features and/or GH responses to GHRH plus Arg and/or IGF-I concentrations showed that dynamic GH provocation tests may become unnecessary in young adults with childhood-onset GHD.