| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Articles |
Departments of Endocrinology (P.F.C.), Hospital Montecelo, Pontevedra 36.005-E, Spain; and Section of Endocrinology (M.A.A., R.V.G.-M.), Hospital Xeral-Cíes, Vigo 36.201-E, Spain; and Department of Functional Biology and Health Sciences (F.M.F.), University of Vigo, Vigo 36200-E, Spain
Address all correspondence and requests for reprints to: Federico Mallo, Ph.D., M.D., Laboratory of Endocrine Physiology, Department of Functional Biology and Health Sciences, Faculty of Sciences, Campus of Vigo, University of Vigo, 36200-E Vigo, Spain. E-mail: . fmallo{at}uvigo.es
Abstract
Type 1 diabetes mellitus (DM 1) is associated with elevated circulating GH concentrations. Because these high GH levels could be explained either by an augmented pituitary secretion and/or delayed elimination clearance or distribution, we sought to evaluate GH pharmacokinetics to propose a model that better explains the elimination kinetics in patients with DM 1 and assess possible differences with normal volunteers that could justify elevation in GH circulating levels in these patients. A multicompartmental analysis was applied to serum GH concentrations measured at different times for 150 min in six patients with DM 1 and six age-, sex-, and body mass index-matched normal subjects after the administration of an iv bolus of recombinant human GH (200 µg), previous suppression of endogenous GH release with octreotide.
The best fitting to the GH disappearance profiles was obtained with the biexponential equation in both groups. From it, we propose a bicompartmental model to explain GH kinetics in normal and diabetic patients. The mean transit time in both compartments and the mean residence time in patients with DM 1 were more than twice the values from control group. So in DM 1 elevated circulating GH concentrations are, at least partially, caused by a delayed GH plasmatic clearance. The DM 1 patients included in this study had a normal renal function; thus, our results agree with the hypothesis that DM 1 constitutes a GH-insensitivity state because a reduced GH clearance by its receptor-mediated mechanism might explain the delayed GH elimination kinetics shown in patients with DM 1. However, the possibility of additional factors contributing to the slowed GH removal from circulation is not completely excluded.
IN PATIENTS WITH type 1 diabetes mellitus (DM 1), pulsatility and circulating GH serum levels are augmented. There have been also reported exaggerated GH responses to a wide range of provocative stimuli in DM 1 patients (1, 2, 3, 4, 5, 6, 7). However, the mechanisms responsible for such altered GH concentration are not yet fully understood.
It is recognized that plasma GH levels at any particular time are determined not only by the pituitary somatotroph secretion rate, but also by the hormonal distribution and elimination phenomena, in general named elimination kinetics. Both phenomena, secretion and elimination, are equally relevant to determine GH levels; however, relatively little is known about the latter (elimination kinetics) and its role in regulating GH concentration in the plasma. Thus, an alteration in GH distribution and/or plasmatic elimination might also significantly contribute to the elevation of GH levels observed in DM 1. So, high circulating GH levels could be explained either by an augmented pituitary secretion and/or a delayed elimination clearance or distribution. For this reason, an accurate evaluation of GH elimination kinetics is crucial to obtain an exact knowledge of the pathophysiological mechanisms underlying the circulating high GH levels observed in DM 1.
Previous studies about GH elimination kinetics in DM 1 patients are scarce and have produced widely varying results (8, 9, 10, 11, 12, 13). Many of these studies have important methodological limitations, such as the use of pituitary-derived GH containing more than one GH molecular isoforms, and, in most instances, radiolabeled preparations, which can provide pitfalls on estimation of GH kinetics. Historical studies on this topic have considered a noncompartmental approach to GH elimination kinetics (8, 9, 11, 12). In all of these cases, no suppression of endogenous GH secretion might also distort the GH plasma profiles. In addition, proper evaluation of GH kinetics is highly dependent on the model assumed, either compartmental or not.
The present study was designed to assess accurately GH kinetic parameters in patients with DM 1, using an experimental approach that obviates some of the limitations of previous studies. Endogenous GH release was suppressed with octreotide, monomeric 22-kDa recombinant human GH (rhGH) was administered, GH serum concentrations were measured by a highly sensitive immunoradiometric assay, and kinetical profiles were analyzed by a multicompartmental approach.
Subjects and Methods
Subjects
Six male patients with DM 1 were studied. The mean age was 25.8 ± 3.1 yr. (mean ± SE), and the mean body mass index (BMI) was 23.0 ± 0.7 kg/m2. The duration of diabetes was 152.5 ± 53.3 months. The clinical characteristics of the patients are reported in Table 1
. Mean levels of hemoglobin A1c at the time of the study were 10.8 ± 0.6% (normal, <5.8%). All of the patients had normal thyroid function, and none was taking any medication other than insulin, before and during the period of study. DM 1 subjects followed an appropriate diet and were all taking a mixture of short- and intermediate-acting insulin by multiple sc injections. Background retinopathy was present in two patients. In the same two patients, urinary albumin excretion rate was in the range of 30–300 mg/d, with no alterations in plasma creatinine levels, neither other laboratory evidence of renal dysfunction. None of the patients had other associated diseases.
|
Study protocol
After an overnight fast, tests were begun at 0900 h, with the subjects recumbent. DM 1 patients did not receive their morning insulin doses until the end of each test to obviate the possible effects of exogenous insulin on GH binding protein (GHBP) serum levels and GH receptors (14). So, an ambient of hyperglycaemia was intentionally maintained in the diabetic group during the study, being basal venous glycaemia at the beginning of test: 257.3 ± 73.4 mg/dl in diabetics vs. 95.4 ± 6.6 mg/dl in controls.
Two indwelling catheters were placed in forearm veins, one in each arm, and kept patent with a slow infusion of 0.9% NaCl. One catheter was used for drug administration and the other for blood sampling. Thirty minutes after admission (-60 min) the first blood sample was obtained and, to suppress endogenous GH secretion, 100 µg of octreotide (Sandostatin SMS 201–995, Novartis Farmaceut\|[iacute]\|ca SA, Barcelona, Spain) was administered as an iv bolus. Sixty minutes later (0 min), a blood sample was obtained followed by the administration of 200 µg of rhGH (Saizen, Serono, Spain) as an iv bolus. Subsequent blood samples were obtained at 1, 2, 4, 6, 8, 10, 15, 20, 30, 45, 60, 90, 120, and 150 min, being immediately centrifuged and kept at -20 C until assay.
Assays
Serum GH levels were measured in duplicate by a highly sensitive two-site monoclonal antibody immunoradiometric assay (HGH Allegro, Nichols Institute Diagnostics, San Juan Capistrano, CA). The sensitivity of the method is 0.02 µg/liter. The mean intrassay coefficients of variation were 2.5% (2.6 µg/liter), 3.8% (6.9 µg/liter), and 2.6% (13 µg/liter). The interassay coefficients of variation were 2.9% (2.6 µg/liter), 3.4% (6.9 µg/liter), and 2.5% (13 µg/liter). All samples from each subject were assayed in the same run. HbA1c levels were analyzed by HPLC (Menarini, Firenze, Italy; normal range < 5.8%).
Data analysis
To establish the model for the GH disappearance curves in both groups, the experimental data of each participant subject were fitted to a multiexponential model using DIMSUM, which is software for multiexponential model discrimination (15), according to the following relationship:
 |
The model was estimated by nonlinear, weighted, least-squares regression, following Marquart’s iterative algorithm (16), and tested by the following criteria: 1) visual inspection of graph plots from the original data and the estimated model; 2) ANOVA between original data and the estimated model; 3) when several models were significantly fitted to the data, F test was performed for the sum of squares of the residuals, for the three models studied (mono-, bi- and triexponential), to establish significance; 4) the error of each estimated parameter as coefficient of variation was always lower than 100% (17); 5) the squared correlation coefficient (r2) was always higher than 0.98; and 6) we consider the Akaike’s and Schwarz’s informative criteria for further discrimination between models. With such procedure, we fitted the data to mono-, bi-, and tri-exponential forms of this model. The best-fitted model to the GH disappearance profiles was obtained with the biexponential one in all cases. This model can be mathematically expressed according to the following equation:
 |
1 and 2 are the decay constants. From the fitted biexponential model, the following pharmacokinetic parameters were calculated:
GH concentration at 0 min (C0 = A1 + A2);
Area under curve (AUC = A1/1 + A2/2);
Estimated plasmatic volume (Vp = Dose/C0);
Apparent distribution volume (Vda = Dose/(AUC · 2);
Estimated distribution volume in the steady-state (Vss = Vp + (Vp · k12/k21));
Metabolic clearance rate (MCR = dose/AUC);
Area under the moment of the curve (AUMC = A1/12 + A2/22);
hormone half-lives (t(1/2)i = ln 2/i);
the elimination rate constant (k10 = C0/AUC);
the transfer rate constants between compartments (k21 = 1 · 2/k10, k12 = 1+2-k10-k21);
the mean transit time in each compartments (MTT1 = 1/k10, MTT2 = k12/(k21 · k10));
and mean residence time in the body (MRT = MTT1+MTT2).
The data are presented as mean ± SEM. Statistical analysis was made by nonparametric tests for independent samples (Mann-Whitney test). Significance level was established at P < 0.05.
Results
The iv administration of the somatostatin analog octreotide blunted endogenous GH secretion in both groups. The infusion of a bolus of 200 µg of rhGH induced an immediate increased in its serum level followed by a time-dependent decline toward basal levels (Fig. 1). The plasma GH disappearance curves of all subjects studied, either diabetics or controls, were best characterized by a biexponential equation, with a first phase of distribution followed by a second phase of GH disappearance.
|
). GH concentration at 0 min (Co), the estimated plasma volumes (Vp), the apparent distribution volumes (Vda), and the volumes of distribution in the steady-state (Vss) were similar in both groups. But a marked reduction on clearance, in type 1 diabetic patients was found (MCR: 402.0 ± 76.8 vs. 205.9 ± 56.0 ml/min/m2; P = 0.01).
|
). Discussion
In patients with DM 1, basal GH serum levels are commonly elevated, and exaggerated GH responses to multiple provocative stimuli had been described (1, 2, 3, 4, 5, 6, 7). Such hormonal disturbances have usually been attributed to an augmented pituitary GH secretion (6, 14), undervaluing the relevance of GH elimination clearance from plasma. However, it is well known that the GH plasmatic level at any particular time is the result not only of concurrent somatotroph secretion, but also of other factors as tisular distribution and elimination from plasma. Therefore, no accurate inferences about instantaneous GH secretion can be made when information on distribution volumes and plasma clearance is lacking (18).
In our study, we investigated the impact of DM 1 on the pharmacokinetics of an exogenous rhGH bolus. The parameters of GH distribution and elimination have been assessed in both groups applying a four-step approach: 1) suppression of endogenous GH secretion by the somatostatin analog octreotide; 2) administration of a single iv bolus of rhGH-22K; 3) measurement of GH plasmatic concentration by a highly sensitive immunorradiometric assay; and 4) use of multiexponential approach to analyze the data.
The somatostatinergic analog octreotide used to suppress endogenous GH has been shown to be able to decreases glomerular filtration rate in acromegalic patients (19). However, unmodified creatinine clearance values have been observed during octreotide infusion, suggesting negligible effects of the drug on the glomerular filtration rate in normal subjects (20). Up to the date, there are no data suggesting a possible influence of octreotide on the assessment of GH kinetics parameters in normal and diabetic patients. The metabolic effects on carbohydrate metabolism of a single dose of octreotide may be considered negligible in normal subjects. Moreover, it has been demonstrated that the half-life of exogenous GH in diabetic patients is not affected by short-term metabolic changes (13).
The use of rhGH, which contains only the 22K variant, obviates the possible differences on clearance rates of the various molecular GH isoforms as reported by Baumann et al. (21). Opposite to pituitary-derived GH, rhGH is composed only by monomeric 22K form, which represents most of circulating GH. According to Vahl et al. (22), the magnitude of our GH bolus can be considered physiological compared with previous studies employing prolonged infusions or supraphysiological doses of GH. Several reports have estimated different GH kinetics using the constant infusion technique to reach a kinetical steady-state (10, 11, 12). However, long exposure to GH or high amounts of this hormone might more extensively saturate the GHBPs and GH receptors, then blocking receptor-mediated GH internalization, which seems to be an important route for elimination of this hormone (21). This fact would explain the marked delay in GH elimination kinetics shown by those studies.
The exact role played by the elimination phenomena in the GH disturbances of type 1 diabetes mellitus is far from been settled. Previous studies to assess GH kinetic parameters produced widely varying results in diabetic patients. Most of them were performed several years ago, when the tools now employed, such as highly purified hGH and highly sensitive immunoassays were not available and modern analytical methods were not employed. These methodological limitations may have contributed to the discrepancies in the results reported by different studies. In this sense, Owens et al. reported a mean half-life of 21.8 min in diabetic subjects that was not significantly different from the 19 min found in normal volunteers (10). On the other hand, Boucher et al. (8) found significantly increased half-lives of radiolabeled GH in diabetic patients compared with normal subjects. On measuring the MCR of GH in patients with DM 1, previous studies have also produced widely varying results (9, 11).
More recently, Mullis et al. (13) using a bolus injection of rhGH following suppression of endogenous GH secretion with somatostatin, obtained a GH half-life of 13.6 min, in a group of six type 1 diabetic patients. They could not concluded if this GH half-life was different from the normal subjects half-life because they did not study a control group. With the above-mentioned approach, we demonstrated an elimination half-life for GH of 15.8 min in normal subjects, which is very similar to those previously reported (22, 23).
Although there are also discrepancies in normal subjects, recently it has been demonstrated that the MCR and Vd of GH increases with age and correlate positively with fat mass (22). In our study, we obviated the impact of age, sex, and body composition on GH pharmacokinetics by using age-, sex-, and BMI-matched controls. In type 1 diabetic patients, we demonstrated a clearly delayed GH plasmatic clearance compared with normal subjects.
It is well established that drug kinetics are highly dependent on the compartmental model assumed (16). In our study, the data from both groups were analyzed using mono- and multiexponential models, and we have proven that in DM 1 patients, as occurs in normal subjects, GH elimination kinetics are better fitted to a biexponential model. Such a model led us to assume the distribution of GH in two intercommunicated compartments: a central compartment, which is represented by plasmatic water, where unbound GH circulates freely; and a peripheral compartment, which could be integrated by GHBPs and bound GH.
Although GH plasma disappearance curve is a mathematical function with a multiexponential structure, most of the authors had not tried to analyze it in this way. As mentioned, our data for GH kinetics are better explained on the basis of a biexponential equation (Fig. 2). This implies a bicompartmental model to explain GH kinetics, including the circulation of GH molecules through the organism. Thus, pituitary secretion is the unique source of the circulating GH molecules, introducing them into the main compartment, the intravascular space. From this central compartment, where GH circulates as a free form, GH molecules are distributed to the peripheral compartment, constituted by GHBPs. The exchanges between both compartments are bidirectional, as GH binding to GHBPs is reversible. The exchanging rate will be dependent on affinity of GHBPs for GH, and represented by the constants K21 and K12. Finally, GH elimination by receptor-mediated internalization, kidney, liver, and other minor metabolic pathways can only work over the free unbound fraction of circulating GH in the central compartment, and is represented by K10. This approach is more complex but allows a better understanding of the process.
|
MRT = 
MTTi
To study GH kinetics using these parameters have two main advantages: 1) MTTs are additive between compartments. That is, the time that the ideal molecule of GH expends in the transit of any single compartment (MTT), can be added to the time expended in any other. 2) MTTs and MRT have a proper physiological meaning. So, we could study how long GH stays in any compartment and in the whole organism (24). These are remarkable properties that cannot be applied to half-life. Moreover, half-life has proper sense only in mono-compartment models, but not in more complex situations like those discussed here (24, 25). Finally MTTs are so easy to calculate as half-lives (24).
In the present study the GH MRT of diabetic patients was more than twice of control subjects. The MTTs were increased in both compartments of the diabetic patients. Thus an ideal molecule of GH, once secreted to the circulation of a diabetic, stays a longer time before being eliminated, in both central and peripheral compartment.
There are many possible explanations for the increased MRT of GH found in DM 1. Decreased renal GH clearance might be a possible mechanism for high plasma GH concentrations in diabetic patients. But this is not likely to be the explanation for the delayed elimination kinetics in our patients. Four of six diabetics had not detectable alterations in renal function and none had clinical or laboratory evidence of nephropathy. The other two presented microalbuminuria, with no other biochemical alterations, nor clinical data of established renal dysfunction. Although microalbuminuria predicts the development of renal disease in DM1 patients, at this stage it is accompanied by a normal or even elevated glomerular filtration rate in DM1 (26). So we did not expect to find any relevant modifications in GH kinetics of these two patients with microalbuminuria.
GH exchanges between central and peripheral compartments are also delayed, as revealed by the marked reduction in the K12 and K21 constants, having a half of the values of control subjects. This suggests an alteration in the binding affinity of GHBPs for GH. As GHBPs correspond to the extracellular domain of the GH receptor, a GHBP reduced affinity might well represent a defective GH receptor binding (27). On the other hand low levels of GHBP have being reported in DM-1 patients (28, 29, 30). This reflects a reduced GH receptor density and a concomitant GH insensitivity. And it might explain a reduced GH clearance by the mechanism of receptor-mediated internalization and the delayed GH elimination kinetics in DM 1 (31). The last seems to be the more likely explanation for our data, although the possibility of additional mechanisms contributing to the delayed GH removal from plasma is not excluded.
The main mechanism for GH elimination is the receptor-mediated internalization in the target cells. But to reach its receptor, free GH must get out of circulation to the interstitial space through the capillary membrane. This process has a favoring concentration gradient: higher levels in the vascular space, meanwhile, target tissues act as a shrink for GH. So, there is a main resultant flux from plasma to target tissue. The amount of this flux will be ruled by the Fick’s Law, which states that it will be directly proportional to the magnitude of the gradient and the surface of the membrane, and inversely proportional to the thickness of the membrane. It has been describe that in diabetes mellitus the hyperglycemia could induce the thickening of basement membrane of the capillaries throughout the body (26). Then in diabetics the exchange of GH molecules from vascular to interstitial space might be lower, partially explaining the reduction of K10. But, also, the increase in GH circulating levels, can in fact augment the magnitude of the gradient, compensating, also in part, the increase on the membrane thickness.
In summary, we report GH kinetic parameters in DM 1 of rhGH previous suppression of endogenous GH with octreotide. Both in normal and diabetic patients GH elimination kinetics are better explained applying a biexponential equation, and from it we propose an explicative bicompartmental model. The MRT of rhGH is prolonged in patients with DM 1, and MCR and K10 markedly reduced, so abnormal GH concentrations in this disease may be attributed, at least partially, to a delayed GH elimination from plasma. These data should be taken into consideration when deconvolution analysis is used for evaluating GH secretory responses.
Acknowledgments
We are grateful to Dr. Carlos Diéguez for critical review of the manuscript and his encouraging ideas.
Footnotes
This study was supported by grants from Fondo de Investigación Sanitaria (FIS 96/1739), Spanish Ministry of Health and from Xunta de Galicia (PGIDT00-PXI30109PN).
Abbreviations: BMI, Body mass index; DM 1, type 1 diabetes mellitus; GHBP, GH binding protein; MTT, mean transit time; MRT, mean resident time; rhGH, recombinant human GH.
Received April 27, 2001.
Accepted January 11, 2002.
References
This article has been cited by other articles:
 |
|
 |
|
  J. D. Veldhuis, M. Bidlingmaier, S. M. Anderson, W. S. Evans, Z. Wu, and C. J. Strasburger Impact of Experimental Blockade of Peripheral Growth Hormone (GH) Receptors on the Kinetics of Endogenous and Exogenous GH Removal in Healthy Women and Men J. Clin. Endocrinol. Metab., December 1, 2002; 87(12): 5737 - 5745. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
