Alterations of Monocyte Function in Patients with Growth Hormone (GH) Deficiency: Effect of Substitutive GH Therapy¹

GH deficiency (GHD) is associated with increased prevalence of atherosclerosis and cardiovascular morbidity. Because monocytes play a crucial role in the development of atherosclerosis, we investigated in the present study the effect of GH deficiency and subsequent GH replacement on monocytic function in hypopituitary subjects. Twelve patients were randomized to receive GH replacement therapy (either 3 or 6 μg/kg·day, sc) for 3 months. Plasma levels and monocyte production of cytokines and monocyte adhesion to endothelium were determined in controls and patients with GHD before and after GH treatment. Before GH therapy, patients with GHD had increased basal plasma tumor necrosis factor-α (TNFα; 220% over control values; P = 0.004) and interleukin-6 (IL-6; 340% over control values; P = 0.0009) levels. Basal monocyte production of both cytokines was also significantly higher in patients with GHD [484% over control values for TNFα (P= 0.0007); 1479% over control values for IL-6 (P = 0.035)]. GH treatment for 3 months led to a reduction in plasma TNFα (135% over control values; P = 0.03, pre- vs. post-GH therapy), monocyte TNFα production (204% over control values; P = 0.01), plasma IL-6 (219% over control values; P = 0.07), and monocyte IL-6 production (448% over control values; P = 0.01). Plasma TNFα levels positively correlated with monocyte TNFα production in patients with GHD both before and after GH therapy (P = 0.003 and P = 0.049, respectively). A positive correlation (P = 0.0003) was also observed between monocyte TNFα production and monocyte IL-6 production. There were no correlations between these plasma cytokine levels or monocyte cytokine production and parameters of body composition, lipid profile, or IGF-I and IGF-binding protein-3 levels. Before GH treatment, adhesiveness of monocytes to cultured aortic endothelial cells was also enhanced. This alteration was not reversed by GH administration. In conclusion, our results demonstrate that markers of monocyte activation are increased in patients with GHD and that GH replacement partly reduces these abnormalities. Reduction of cellular activation of monocytes by GH therapy could potentially contribute to reduce the risk of cardiovascular events in patients with GHD.

HYPOPITUITARISM is associated with increased prevalence of atherosclerosis (1) and enhanced cardiovascular morbidity and mortality (2, 3) despite conventional replacement therapy with glucocorticoids and T₄, suggesting a role for GH deficiency (GHD) in the vascular disease of hypopituitarism. Although GHD is only one of the possible contributors to increased cardiovascular disease in adult hypopituitarism, the recent availability of GH replacement therapy allows evaluation of the role of GHD in the development of cardiovascular disease. Adults with GHD have been shown to have increased cardiovascular risk factors, including altered body composition with increased body fat and abnormal levels of serum lipids and lipoproteins (4), both of which were improved by GH replacement therapy (5). The mechanisms involved in the increased prevalence of atherosclerosis associated with GHD are not well known.

Atherosclerosis is characterized by a chronic and excessive inflammatory response resulting from the trapping of low density lipoprotein (LDL) in the arterial wall. Evidence has been provided that immune mechanisms play an important role in atherogenesis. Monocyte adhesion and migration into the arterial wall are among the earliest events in atherogenesis (6). Once in the intima, these cells are exposed to a milieu of modified lipoproteins, cytokines, chemoattractants, and growth factors, all of which can cause further activation and differentiation into tissue macrophages. These cells participate in the atherogenic process not only as scavenger cells, but also by their capacity to produce numerous proinflammatory cytokines and growth factors.

A growing body of evidence suggests that GH plays a role in the regulation of the immune system. Several alterations in the immune system of patients with GHD have been described. These changes include reduced activity of natural killer cells and antibody synthesis, thymic hypoplasia, significant delay in rejecting allogenic skin grafts, and defective antibody- and cell-mediated immunity (7–13). Despite the abundance of information on the immune system in GHD patients, monocytic function in these subjects has not been investigated. Because monocytes seem to contribute to the early development of atherosclerosis and produce proatherogenic cytokines (14), we sought to investigate the function of monocytes/macrophages in patients with GHD. We report here on the effect of GHD and low dose GH replacement therapy on plasma and monocyte/macrophage cytokine production and on monocyte adhesiveness to endothelium in a group of adult hypopituitary subjects.

Subjects and Methods

Patients

Twelve patients (11 men and 1 woman) with hypopituitarism (10 with the adult-onset form) were studied. Clinical, biochemical, and hormonal data are shown in Tables 1¹ and 2². GHD was diagnosed by an insulin tolerance test (0.075–0.1 U/kg BW, iv, regular insulin) in which the peak GH concentration was less than 3 μg/L. Two had isolated GHD, and 10 were affected with multiple hormone deficiency and treated with substitutive therapies, such as T₄, testosterone or estrogen/progestin, and cortisone, at standard doses. The duration of GHD in the study population varied from 2–38 yr. Primary pituitary or hypothalamic pathologies included 2 craniopharyngiomas, 1 reticulosarcoma, and 5 adenomas. Three patients had idiopathic GHD, and 1 had posttraumatic deficiency. Seven patients underwent transphenoidal surgery, and 1 was treated by pituitary irradiation and chemotherapy. Seven patients were hyperlipidemic; 1 was glucose intolerant.

Table 1.

Clinical characteristics of the patients

Patient no.	Age (yr)	Sex	Diagnosis	Therapy	T4	HC	Sex steroids
1	51	M	NFPA	TSS	+	+	+
2	62	M	CR	TSS	+	+	+
3	66	M	PRL	TSS	−	−	−
4	45	M	NFPA	TSS	+	+	+
5	49	F	NFPA	TSS	+	+	−
6	33	M	IGHD		+	+	+
7	18	M	IGHD		−	−	−
8	45	M	RS	RO	+	+	+
9	18	M	IGHD		+	+	+
10	53	M	NFPA	TSS	+	+	+
11	33	M	CR	TSS	+	+	+
12	40	M	PTD		+	+	+

NFPA, Nonfunctioning pituitary adenoma; PRL, prolactinoma; IGHD, idiopathic GH deficiency; CR, craniopharyngioma; RS, reticulosarcoma; PRD, posttraumatic deficiency; TSS, transsphenoidal surgery; RO, radiotherapy.

Table 2.

Characteristics of the study population before and after 3 months of GH therapy

	Patients		Control group
	Pre	Post-GH therapy
Age (yr)	43 ± 4		39 ± 5
Sex (M/F)	11/1		10/2
BW (kg)	85.7 ± 4.2	87.4 ± 4.2	80.3 ± 6.1
BMI (kg/m2)	29 ± 1.4	29 ± 1.4	26.8 ± 2.4
Fat mass (kg)	31.6 ± 2.7	31.2 ± 2.8	ND
Trunk fat mass (kg)	17.1 ± 1.6	16.6 ± 1.6	ND
Lean mass (kg)	55.8 ± 3.5	56.1 ± 3.4	ND

ND, Not determined.

Protocol

The protocol was approved by the Notre Dame Hospital ethics committee, and informed written consent was obtained from all subjects. This study was ancillary to a larger multinational, multicenter, randomized study (Eli Lilly & Co., Indianapolis, IN) comparing the effects of two different algorithms of GH replacement in hypopituitary adults (either 3 or 6 μg/kg·day, sc). In this study, seven patients were randomized to receive the 3 μg/kg daily dose, and five patients were randomized to receive the 6 μg/kg dose for a period of 3 months. GH was self-administered by the patients as a daily injection in the evening at bedtime. Because clinical, hormonal, and metabolic data were similar both before and 3 months after GH therapy in the two subsets of patients with GHD randomized to the 3 and 6 μg/kg doses of GH, the data from the two subgroups were pooled and presented as mean values. Blood samples for lipid and lipoprotein analyses were drawn in the morning after an overnight fast before drug administration. Control subjects, matched with patients for sex, age, and body mass index, were recruited from the hospital staff and relatives. Subjects with infectious or inflammatory conditions or treated by antiinflammatory or antioxidant drugs were excluded from the study.

Body composition

Body composition (total fat and central fat) was determined by dual energy x-ray absorptiometry, using Lunar DPX (Lunar Radiation Corp., Madison, WI) with an objective, highly reproducible measurement of central fat, using a computer-derived rectangle individually adjusted for each patient. The sd in similar patients was approximately 5%.

Assays

Plasma tumor necrosis factor-α (TNFα) and interleukin-6 (IL-6) were measured by highly sensitive, commercial, enzyme-linked immunosorbent assay kits (Quantikine HS, R & D Systems, Minneapolis, MN). The minimum detectable concentrations by these assays were 0.11 pg/mL for TNFα and 0.09 pg/mL for IL-6. The intra- and interassay coefficients of variation of the assays were 5.6% and 10% for TNFα and 3.3% for IL-6, respectively.

Cytokines were measured in the culture medium by enzyme-linked immunosorbent assay kits (Quantikine, R & D Systems). The minimum detectable concentrations by these assays were 4.4 pg/mL for TNFα and 0.7 pg/mL for IL-6. The intra- and interassay coefficients of variation of the assays were less than 7%.

Total serum antioxidant status was assessed using a commercial kit (Randox Laboratory, Mississauga, Ontario). Lipid peroxides were determined in the serum by measuring the thiobarbituric acid-reactive substances, expressed as malondialdehyde equivalents (nanomoles per 500μ L serum) (15, 16).

Serum insulin-like growth factor I (IGF-I) and IGF-binding protein-3 (IGFBP-3) were measured by RIA in the Lilly laboratory of Dr. W. Blum (University Children’s Hospital, Giessen, Germany). IGF-I was measured with a third generation IGFBP-blocked assay without extraction. The intra- and interassay coefficients of variation were 3.6% and 13.1%, respectively. IGFBP-3 was measured using authentic IGFBP-3 as previously described (17). The intra- and interassay coefficients of variation were 3.9% and 11%, respectively.

Human monocyte isolation

Fresh heparinized blood (100 mL) was obtained from GH-deficient patients and healthy nonsmokers donors in the morning between 0800–0900 h. Peripheral blood mononuclear cells were isolated by density centrifugation using Ficoll (Life Technologies, Grand Island, NY) (18), allowed to aggregate in presence of FCS, then further purified by the rosetting technique. After density centrifugation, the recovery of highly purified monocytes (85–90%) as assessed by FACS analysis was obtained. Monocytes were resuspended in serum-free RPMI 1640 medium (Life Technologies) with 2 mmol/L glutamine supplemented with 100 U/mL penicillin and 100 μg/mL streptomycin and used immediately for adhesion assay.

Adhesion assay

Bovine aortic endothelial cells (19th passage) were grown to confluence in DMEM (ICN Biochemicals, Inc., Costa Mesa, CA) supplemented with 10% FCS (HyClone Laboratories, Inc., Logan, UT), 2 mmol/L l-glutamine (ICN Biochemicals, Inc., Costa Mesa, CA), 100 U/mL penicillin, and 100 μg/mL streptomycin (DMEM-FCS; Flow Laboratories, McLean, VA) at 37 C in a 5% CO₂-95% air atmosphere for 6 days. The cells were then trypsinized and subcultured in 96-well culture plates (Costar) for 48 h, at which time cell confluence was reached. In all experiments, cells were used between the 20th and the 24th passage. The day of the experiment, the medium of confluent monolayers of bovine aortic endothelial cells was removed and replaced by fresh serum-free RPMI 1640 medium. One hundred microliters of a monocytic cell suspension (2.3 × 10⁶ cells/mL) was then added to each well. After a 2-h incubation period, nonadherent monocytes were removed by washing twice with phosphate-buffered saline without calcium or magnesium (PBS-A). Adherent cells were lysed in 50μ L hexadecyltrimethylamine ammonium bromide (Sigma Chemical Co., St. Louis, MO; 0.5%) in PBS-A at pH 6.0 for 30 min. Quantification of adherent monocytes was made by measuring monocyte myeloperoxidase activity (19). Briefly, myeloperoxidase activity was determined by the addition to each well of 250 μL dianisidine dihydrochloride (Sigma Chemical Co.; 0.2 mg/mL in PBS-A) warmed to 37 C and mixed with hydrogen peroxide (0.4 mmol/L, final concentration). After 2–5 min of incubation, the optical density of the plated wells was read at 450 nm using a Titer-Tek multiscan spectrophotometer (Flow Laboratories, Rockville, MD).

Culture of human monocytes and monocyte-derived macrophages (MDM)

For determination of cytokine production, freshly isolated human monocytes were cultured for 24 h in RPMI 1640 supplemented with 10% (vol/vol) autologous serum in the presence or absence of 1 μg/mL LPS. Differentiation of monocytes into MDM was obtained by culturing the cells in RPMI 1640 supplemented with 20% autologous serum. After 7 days of culture, MDM were treated with or without LPS for 24 h, and cytokines were measured in the cell supernatants.

Statistical analysis

Data are expressed as the mean ± sem. Statistical differences between the variables were determined using Student’s t test for paired data and nonparametric tests (Wilcoxon and rank sum tests) for unpaired data. Linear regression analysis was used to determine whether correlation existed between variables. P < 0.05 was considered statistically significant.

Results

Body composition in patients with GHD

Mean values for body weight, body mass index, total fat mass, trunk fat mass, and fat-free mass before and 3 months after GH replacement therapy are shown in Table 1. A trend for a reduction in total and central fat and an increase in fat-free mass was seen, but did not reach statistical significance.

Plasma IGF-I and IGFBP-3 in patients with GHD

Plasma IGF-I increased significantly (P < 0.005) in patients with GHD after GH replacement therapy. Plasma IGFBP-3 tended to increase after therapy, albeit not significantly (Table 2).

Lipid profile

Serum levels of total cholesterol, LDL cholesterol, and triglycerides were significantly greater (P < 0.005, P < 0.01, and P < 0.005, respectively) in patients with GHD than in normal controls (Table 3). The mean fasting serum glucose levels were similar in both groups. GH therapy did not significantly modify the lipid profile in patients with GHD (Table 2).

Table 3.

Lipid and hormonal profiles in the study population before and after 3 months of GH therapy

	Patients		Control group
	Pre	Post-GH therapy
Glucose (mmol/L)	4.9 ± 0.1	4.9 ± 0.2	4.9 ± 0.1
Cholesterol (mmol/L)	5.52 ± 0.29a	5.76 ± 0.25	4.34 ± 0.18
LDL (mmol/L)	3.39 ± 0.16a	3.63 ± 0.21	2.69 ± 0.21
TG (mmol/L)	2.05 ± 0.33a	2.57 ± 0.35	1.01 ± 0.11
HDL (mmol/L)	0.96 ± 0.10a	1.03 ± 0.10	1.44 ± 0.11
Cholesterol/HDL	6.4 ± 0.6a	6.3 ± 0.7	3.1 ± 0.2
IGF-I (μg/L)	79 ± 13	157 ± 18b	ND
IGFBP-3 (μg/L)	2.38 ± 0.29	2.88 ± 0.23	ND

ND, Not determined.

^a

P < 0.05, pre-GH therapy vs. controls.

^b

P < 0.005, pre vs. post-GH therapy.

Plasma lipid peroxides and total plasma antioxidant status

Plasma lipid peroxide levels did not differ between the control and GHD groups before GH administration (data not shown). Total plasma antioxidant status was also similar in the two groups (controls, 0.75 ± 0.03 mmol/L; GHD patients, 0.78 ± 0.06 mmol/L).

Plasma levels and monocyte production of cytokines

Before GH therapy, basal plasma TNFα levels were significantly greater in patients with GHD than in normal controls (P< 0.005; Fig. 1a). GH therapy reduced plasma TNFα in these patients to levels similar to those observed in the control group (Fig. 1a). The mean plasma IL-6 levels were also elevated in GHD patients before therapy compared to control values (P < 0.001). There was a trend for a lowering of plasma IL-6 levels by GH treatment, although the effect was not statistically significant (pre- vs. post-GH therapy, P = 0.07; Fig. 1b). Before GH treatment, basal monocyte TNFα and IL-6 productions were significantly higher (P < 0.001 and P < 0.05, respectively) in the GHD group than in the control group (Fig. 2, a and b). Administration of GH markedly decreased basal release of these two cytokines by monocytes (Fig. 2, a and b). A strong positive correlation was observed between plasma TNFα levels and monocyte TNFα production in patients with GHD both before (r = 0.87; P < 0.001; Fig. 3a) and after GH therapy (r = 0.63; P < 0.05). Monocyte TNFα production in GHD patients positively correlated with monocyte IL-6 production (r = 0.87; P < 0.001; Fig. 3b). In GHD patients, an increase in MDM basal TNFα production was also found (controls, 10.1 ± 1.6 pg/mL; GHD patients, 37.0 ± 7.1; P < 0.005). This alteration was reversed by GH therapy (17.3 ± 5.6; P < 0.05 vs. before GH). There was no significant difference in the lipopolysaccharide-stimulated release of monocyte and MDM TNFα between the GHD and control groups (data not shown).

Figure 1.

Plasma TNFα (a) and IL-6 (b) levels in control subjects and patients with GHD before and after GH replacement therapy. Data represent the mean ± sem. *, P < 0.05, pre- vs. post-GH therapy;*** , P < 0.005, pre-GH therapy vs. controls.

Figure 2.

Basal monocyte TNFα (a) and IL-6 (b) production in control subjects and patients with GHD before and after GH replacement therapy. Data represent the mean ± sem. ***, P < 0.005 vs. controls; **, P = 0.01 vs. before GH therapy; *, P < 0.05 vs. controls.

Figure 3.

a, Correlation between monocyte TNFα production and plasma TNFα levels in patients with GHD before GH replacement therapy. b, Correlation between monocyte TNFα production and IL-6 production in patients with GHD before GH replacement therapy.

Relations between cytokines and body composition, and hormonal and lipid profile

There were no correlations between plasma cytokine levels or monocyte cytokine production and parameters of body composition, lipid profile or IGF-I and IGFBP-3 levels.

Monocyte adhesion to endothelium

A significant increase in the adhesion of monocytes isolated from patients with GHD to endothelial cells was observed before GH therapy (Fig. 4). There was no correlation between the degree of monocyte binding in these subjects and any of the metabolic parameters measured. This alteration was not reversed by GH administration (Fig. 4).

Figure 4.

Monocyte adhesion to endothelial cells in control subjects and patients with GHD before and after GH replacement therapy. Results are expressed as the percent adhesion over control values. Data represent the mean ± sem. **, P < 0.01 vs. controls.

Discussion

This study showed that adults with GHD have elevated plasma concentrations of TNFα and IL-6 and increased adhesiveness of their blood monocytes to endothelial cells in vitro. The study also established that the increased plasma concentrations of TNFα and IL-6 are related to increased monocyte production of these cytokines. These findings may represent novel features of increased risk of atherosclerosis in patients already known to have central obesity and elevated serum concentrations of cholesterol, LDL cholesterol, and triglycerides. Furthermore, this study shows that GH replacement therapy for 3 months almost totally reversed the abnormally elevated production of TNFα and IL-6, apparently independently of any effect on body composition or serum lipid concentrations.

The clinical relevance of our findings is supported by several findings from clinical and experimental data that indicate a major role of the proinflammatory cytokines TNFα and IL-6 in the development of atherosclerosis. TNFα (20, 21) and IL-6 (22, 23) are present in the human arterial atherosclerotic wall. TNFα has been shown to promote the adhesion of leukocytes to endothelial cells (24, 25), to induce chemotaxis (26), and to increase the expression of several cell adhesion molecules (27). IL-6 also promotes lymphocyte adhesion to endothelium (28), increases endothelium permeability (29), stimulates monocyte transformation into macrophages (30), and induces vascular smooth muscle proliferation (31–33). Finally, TNFα and IL-6 have a major role in the regulation in the liver of acute phase proteins (34, 35), including fibrinogen, an important vascular risk factor (36). A relationship between levels of cytokines and acute phase proteins and risk factors for atherosclerosis or the degree of atherosclerotic disease has been shown (37, 38).

Our data suggest that the increased plasma cytokine levels are produced by blood monocytes and should be considered as plasma markers of monocyte activation. First, there is a correlation between plasma levels of TNFα and those of IL-6. Second, plasma levels of both cytokines correlate with basal concentrations obtained ex vivo from monocytes. Third, although theoretically circulating TNFα may originate in part from adipose tissue, plasma TNFα levels in our patients with GHD do not correlate with total or abdominal fat mass.

The data of the present study may appear discordant with the current view that GH is an activator of monocyte function. GH, IGF-I, and PRL have potential pleiotropic actions on the lymphohemopoietic system (39). In monocytes, GH stimulated chemotaxis, and both GH and PRL were able to prime macrophages for superoxide anion production (11, 40, 41). Also, GH participates, indirectly via IGF-I, in the control of TNFα expression by monocytes and macrophages (12, 42). In hypophysectomized rats or dwarf mice, there is a hypoplasia of the lymphoid system (43) and a reduced TNFα response to LPS by peripheral macrophages (42). There abnormalities are partly restored by GH therapy (42). These data suggest that GH acts physiologically as a hemopoietic growth factor, but do not imply that in human hypopituitarism monocyte functions are depressed. For instance, lack of GH can be compensated for by IGF-I (which is less dependent on GH in man than in rodents), IGF-2, and insulin and by the redundancy of the lymphohemopoietic growth factor network (which is greater in man than in rodents) (39). Simultaneously, GHD may be associated with an as yet unidentified metabolic, hormonal, or cytokine alteration that will activate basal monocyte functions.

Our results have also shown that monocyte adhesiveness to cultured endothelial cells is enhanced in patients with GHD. However, at variance with the increased production by monocytes of TNFα and IL-6, the enhanced adhesion of monocytes is not reversed by GH replacement therapy. This suggests that either different factors or pathways are responsible for the abnormalities of monocyte function in patients with GHD or that the enhanced monocyte adhesion is not GH related in hypopituitary cases.

In our study, GH treatment for 3 months at either 3 or 6 μg/kg daily had no significant effect on body composition or lipid profile, although it did increase plasma biochemical markers such as IGF-I. IGFBP-3 did not increase significantly, probably because it is a less sensitive marker of GH therapy than IGF-I (44).

Many studies have shown that GH replacement therapy for 6–12 months in adults with GHD reduces total and abdominal fat mass, increases fat-free mass, and reduces total cholesterol and LDL cholesterol (4, 5, 45–48). The fact that we did not observe these beneficial effects of GH therapy on body composition and lipid profile may be due to the shorter period of treatment and the lower doses of GH in our study. Nevertheless, our study shows that GH appears to play independent roles in the maintenance of adiposity, blood lipid profile, and blood monocyte functions. Monocyte function is altered by GHD in a way that predicts an increased vascular risk of atherosclerosis and vascular morbidity.

In conclusion, we have shown that in adult patients with GHD, markers of monocyte activation associated with atherosclerotic and cardiovascular risk factors are increased. We have also shown that GH replacement therapy reduces cellular activation of blood monocytes/macrophages. These changes could account for some potential beneficial effects in reducing the risk of atherosclerosis and cardiovascular disease in patients with GHD.

Acknowledgemens

The authors thank Mrs. M. Jetté for her technical assistance, Mrs. J. Auclair for secretarial assistance, and Mrs. L. Pedneault for her precious collaboration.

¹

This work was supported in part by Eli Lilly & Co..