| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Medical Research Laboratories (J.F.), Clinical Institute and Medical Department M, Aarhus University Hospital, DK-8000 Aarhus, Denmark; Departments of Medicine (R.A.R., J.M., W.S., M.A.N.) and Surgery (M.B., R.L., J.K.), Ruhr-University, Knappschafts-Krankenhaus, 44780 Bochum, Germany; Department of Internal Medicine I (R.A.R.), University of Heidelberg, 69120 Heidelberg, Germany; Department of Visceral and Transplant Surgery (R.L., J.K.), Medizinische Hochschule Hannover, 30625 Hannover, Germany; and Diabeteszentrum Bad Lauterberg (M.A.N.), 37431 Bad Lauterberg im Harz, Germany
Address all correspondence and requests for reprints to: Dr. Jan Frystyk, Medical Research Laboratories, Aarhus University Hospital, Nørrebrogade 44, DK-8000 Aarhus C, Denmark. E-mail: jan{at}frystyk.dk.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Objective: We hypothesized that normalization of portal insulin in patients with portal pancreas graft drainage stimulates the GH/IGF-I axis and thereby contributes to glucose control.
Methods: We compared patients after combined kidney and pancreas transplantation with portal drainage (n = 7) to patients with systemic drainage of the pancreas graft (n = 8) and nondiabetic controls (n = 8). Overnight fasting sera were analyzed for free and total IGF-I and IGF-binding proteins. Glucose regulatory hormones were examined after an oral glucose tolerance test and GH after stimulation with GHRH.
Results: Systemic drainage led to higher basal and stimulated insulin levels than portal drainage (P < 0.05), but increments in response to oral glucose were reduced in both transplanted groups (P < 0.05 vs. controls). However, glucose tolerance was similar in all groups. Circulating free and total IGF-I and IGF-binding protein-3 were similar to control levels in the systemic drainage group but elevated in the portal drainage group (P < 0.05). Consistently, the GH response was reduced in the portal drainage group (P < 0.05 vs. controls) and correlated inversely with free IGF-I (r = –0.63, P < 0.05).
Conclusion: Portal drainage of pancreatic endocrine secretion in pancreas graft recipients raises IGF-I and lowers GH secretion. These changes might explain that glucose regulation is maintained despite lower peripheral insulin levels, compared with patients with systemic graft drainage and nondiabetic control subjects.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Several different surgical techniques for pancreas transplantation have been described, including systemic-venous (SV) drainage and portal-venous (PV) drainage of the graft. The majority of centers prefer enteric drainage of exocrine pancreas graft secretion combined with SV drainage of the endocrine pancreas (5). SV drainage has the disadvantage that it circumvents the portal circulation and results in hyperinsulinemia, both under basal and stimulated conditions, when compared with patients with PV drainage of the pancreas graft, kidney transplant recipients, and nondiabetic control subjects (6, 7, 8, 9, 10). Hyperinsulinemia has caused concern because it may be associated with atherosclerotic complications, which are a frequent problem in patients with long-term type 1 diabetes, who are the typical candidates for pancreas transplantation (4, 11, 12). In patients with portal drainage of the pancreas graft, peripheral insulin concentrations are markedly lower with similar control of glucose metabolism (6). However, the mechanism underlying regulation of glucose metabolism in these patients is unknown.
Insulin is an important regulator of the GH/IGF-I axis (13). It has been suggested that insulin increases the hepatic GH sensitivity by up-regulation of the GH receptors and thereby indirectly stimulation of the hepatic IGF-I synthesis (13, 14). In addition, insulin down-regulates the hepatic secretion of IGF-binding protein (IGFBP)-1 and -2 (the insulin sensitive IGFBPs), which both are considered to inhibit IGF-I mediated actions in vivo (13). Therefore, we hypothesized that in pancreas transplanted patients the anatomic location of the venous drainage determines basal and stimulatory concentrations of glucose regulatory hormones and factors of the circulating GH/IGF-I axis.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The study protocol was approved by the ethics committee of the medical faculty of the Ruhr-University, Bochum, Germany, and written informed consent was obtained from all participants.
Subjects
Seven type 1 diabetic patients were studied after combined kidney and pancreas transplantation according to the PV drainage technique and compared with eight matched patients with SV pancreas drainage (into the iliac vein). Details of the surgical procedures have been described elsewhere (15). For comparison, eight nondiabetic volunteers of similar age, sex distribution, and body mass index were examined. Subject characteristics are shown in Table 1
. Of note, there were no differences between the two patient groups with respect to time interval after transplantation and immunosuppressive therapy (Table 1).
|
All participants were studied on two occasions in the morning after an overnight fast in random order: 1) during a 75-g oral glucose challenge over 240 min, samples were collected at baseline and in 30-min intervals for the measurement of insulin, proinsulin, C-peptide, free fatty acids (FFAs), GH, glucagon, gastric inhibitory peptide (GIP), and glucagon-like peptide (GLP)-1; and 2) during an iv GH stimulation test with 50 µg GHRH (Ferring, Kiel, Germany) over 60 min, samples were collected at 0, 30, and 60 min for the measurement of GH. In addition, overnight fasting levels of circulating free and total IGFs and IGFBPs were determined. On the test days, the usual morning medication of the patients was postponed until after the investigation.
Laboratory determinations
All measurements were performed in duplicates within the same assay run unless otherwise stated. Insulin, C-peptide, and proinsulin were measured in EDTA plasma using commercial immunoassay kits. Insulin was measured directly in serum using a monoclonal antibody-based insulin microparticle enzyme sandwich assay (Abbott Laboratories, Wiesbaden, Germany), which has a negligible cross-reactivity with proinsulin (16). C-peptide was measured using C-peptide antibody-coated microtiter wells (C-peptide ELISA; Dako Diagnostics Ltd., Cambridgeshire, UK), which cross-reacts with human proinsulin (63%) as well as proinsulin conversion intermediates (71–87%). Plasma proinsulin was assayed using an ELISA (DRG Instruments GmbH, Marburg, Germany). The proinsulin ELISA also detects des (65,66) split proinsulin (cross reactivity 55–64%) but cross-reacts with neither des (31, 32) split proinsulin nor insulin.
Pancreatic glucagon was assayed in ethanol-extracted plasma using antibody 4305 as previously described (17). GLP-1 was determined in ethanol-extracted plasma as previously described (18). GIP was determined without plasma extraction using antibody R 65, which does not detect the void volume component (molecular weight 8000 GIP). As previously described, this assay does not cross-react with other peptides from the glucagon-secretin-vasoactive intestinal polypeptide family of gastrointestinal peptide hormones (19).
IGF-I and -II were determined by time-resolved immunofluorometric assays as previously described (20). Serum total IGF-I and -II were determined in acid ethanol extracts (20), whereas serum-free IGF-I and -II were determined in triplicates using ultrafiltration by centrifugation at approached in vivo conditions (13). The lower detection limit of ultrafiltered free IGF-I and -II was 0.03 and 0.05 µg/liter, respectively. The dimeric complex of IGF-I and IGFBP-1 was determined by an in-house, highly specific time-resolved immunofluorometric assay, which uses a monoclonal IGFBP-1 antibody (MAB 6303) for capture and an europium-labeled monoclonal IGF-I antibody for detection (21). IGFBP-1, -2, and -3 were determined by commercial immunoassays from Diagnostic Systems Laboratories Inc. (Webster, TX), and GH was determined by a commercial enzyme immunoassay from DRG Instruments. For the commercial assays, the intra- and interassay coefficients of variation were all less than 10%. For in-house assays intra and interassay CVs were as follows: pancreatic glucagon less than 5 and 8% (17); GLP-1 less than 8 and 10% (18); GIP less than 10 and 12% (19); total IGF-I and -II less than 5 and 10% (20); free IGF-I and -II both less than 16%; and dimeric complex of IGFBP-1 and IGF-I less than 5 and 15% (21).
Plasma FFAs were assayed using reagents from Wako Chemicals (Neuss, Germany) on a Hitachi 709 autoanalyzer (Hitachi, Krefeld, Germany). Triglycerides and total cholesterol were measured using standard clinical chemistry.
Statistics
Results are reported as mean ± SEM. Integrated incremental (or decremental) values were calculated according to the trapezoidal rule. The statistical calculations involving all three study groups were carried out using repeated-measures ANOVA. If there was a significant difference among the three experimental groups or if a significant interaction of experimental group and time was documented (P < 0.05), values at single time points were compared by one-way ANOVA. If the ANOVA was significant, it was followed by Student’s unpaired t test to analyze for differences among the three study groups. Comparison of two study groups was performed using Student’s unpaired t test or the 
2 test when appropriate. P <0.05 was considered to denote statistical significance.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
and Table 2)
Overnight fasting plasma glucose levels (Fig. 1A) were less than 100 mg/dl in all participants, with no significant differences among the three groups. Within the first 180 min, glucose excursions after oral glucose were similar, but at 210 and 240 min, plasma glucose remained elevated in both transplanted groups (P < 0.05; Fig. 1A). According to World Health Organization criteria, glucose tolerance was diabetic in one patient operated with PV pancreas drainage, whereas it was impaired in two and diabetic in two patients operated with SV pancreas drainage. In all remaining subjects, glucose tolerance was normal (2 test, P = 0.26).
|
|
) were elevated in patients with SV vs. PV pancreas drainage (13.5 ± 2.7 vs. 5.4 ± 0.6 mU/liter, P = 0.02). However, in comparison with controls (7.4 ± 0.8 mU/liter), there were no statistically significant differences (P = NS). After oral glucose stimulation, plasma insulin increased more markedly in control subjects than in both transplanted groups (Fig. 1B) and returned to baseline values earlier. From 60 to 240 min. insulin levels were markedly higher in patients with SV than PV pancreas drainage (P < 0.05). Integrated insulin responses confirm these results (Table 2). Baseline C-peptide levels were similar: 0.30 ± 0.03 vs. 0.53 ± 0.24 vs. 0.90 ± 0.43 nmol/liter (controls vs. SV drainage vs. PV drainage; P = NS), whereas the C-peptide responses reached higher levels in controls, compared with transplanted patients. In contrast to insulin, the two transplanted groups showed similar C-peptide responses (Fig. 1C and Table 2). Proinsulin levels behaved similarly (Fig. 1D), but there was no time point revealing significant differences. However, in controls integrated proinsulin increments (Table 2) were higher than in patients with PV pancreas drainage (P = 0.04). After ingestion of oral glucose FFA levels changed significantly with time (Fig. 1E); however, there were no differences at individual time points and no differences of integrated decremental responses among the three study groups (Fig. 1E and Table 2). Baseline GH levels did not differ among the three study groups (P = 0.6; data not shown). During the oral glucose tolerance test, GH levels were similarly suppressed in the groups and neither differences in the time course nor in the reactive increment starting at 150 min were observed (Fig. 1F and Table 2). Serum IGF-I remained constant in all study groups (details not shown).
Plasma glucagon differed neither in the fasting state [10.6 ± 1.1 vs. 14.1 ± 2.4 vs. 9.6 ± 1.4 pmol/liter (controls vs. SV drainage vs. PV drainage; P = NS)] nor during the time course of the experiment (details not shown). However, integrated decremental glucagon concentrations (Table 2) were lower in patients with PV drainage than patients with SV pancreas drainage (P < 0.01) and control subjects (P < 0.01). Baseline GIP levels were elevated similarly in the two patients groups: 24 ± 6 vs. 27 ± 8 vs. 6 ± 1 pmol/liter (PV drainage vs. SV drainage vs. controls; P < 0.05). This was accompanied by lower integrated incremental responses after oral glucose for the SV drainage group (Table 2). Baseline GLP-1 was similar among the groups, and levels increased similarly after oral glucose in all groups (details not shown).
Fasting levels of the circulating IGF-axis (Figs. 2–4)
Patients with SV pancreas drainage had normal levels of free IGF-I when compared with controls (Fig. 2A). In contrast, there was an approximately 4-fold elevation of free IGF-I in patients with PV pancreas drainage (Fig. 2A). Mean levels of free IGF-II (Fig. 2B) tended to be higher in patients with PV pancreas drainage than in controls and patients with SV pancreas drainage (P = 0.05).
|
|
|
, C and D). The calculated ratio of free to total IGF-I (Fig. 2E) revealed that the increase in free IGF-I was relatively more pronounced than that of total IGF-I. In contrast, the ratio of free to total IGF-II was unaltered (data not shown).
The pattern of IGFBP-3 concentrations was similar to that of total IGF-I and -II, with elevated levels in patients with PV pancreas drainage (Fig. 2F). Accordingly, the molar ratios of IGF-I to IGFBP-3, IGF-II to IGFBP-3, and the sum of IGF-I plus IGF-II to IGFBP-3 all were identical when comparing the three study groups (P = NS) (data not shown). Levels of IGFBP-1 and -2 (Fig. 3, A and B) displayed the same pattern: both patient groups had elevated levels of IGFBP-1 and -2, and additionally, patients with SV pancreas drainage had higher levels than patients with PV pancreas drainage. In contrast to IGFBP-1, levels of the dimeric complex of IGF-I and IGFBP-1 (i.e. IGFBP-1 bound IGF-I, Fig. 3C) were elevated to a similar degree in both patient groups. However, the fraction of IGFBP-1 saturated with IGF-I (Fig. 3D) was significantly elevated in patients with PV pancreas drainage, confirming the overall increment of IGF-I in these patients.
Linear regression analysis was performed to examine the relationship between free IGF-I and the other members of the IGF-system in the two pancreas transplanted patient groups (n = 15). We decided to include the two patient groups only in this analysis because they were well matched in all aspects, whereas they both differed substantially from controls with respect to medical history and immunosuppressive medication. Linear regression analysis showed that levels of free IGF-I correlated positively with total IGF-I (r = 0.74, P < 0.01) and IGFBP-3 (r = 0.84, P < 0.0001) and inversely with IGFBP-1 (r = –0.80, P < 0.001) and IGFBP-2 (r = –0.66, P < 0.01). Multiple linear regression analysis revealed that changes in these four variables could explain as much as 89% of the variation in free IGF-I (P < 0.0001).
GHRH stimulation test (Fig. 4)
Integrated incremental responses of GH (Fig. 4A) were significantly lower in patients with PV pancreas drainage, compared with controls and patients with SV pancreas drainage. To study the possible association between baseline levels of free and total IGF-I, respectively, and the integrated GH secretion after GHRH stimulation, a linear regression analysis was performed. However, GH secretion showed a skewed distribution, and therefore data were log transformed. The regression analysis showed that levels of free IGF-I correlated inversely with the integrated GH secretion (Fig. 4B), whereas total IGF-I did not show any significant association (r = –0.22, P = 0.43).
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The present study confirms that SV pancreas drainage leads to significantly higher peripheral insulin levels than PV drainage (Fig. 1) (8, 9), whereas the first-phase insulin release seems more physiological in the PV drainage group. Still, both transplanted groups showed relatively lower and/or delayed insulin responses, compared with controls, and we believe this is caused primarily by perisurgical organ damage. This view is supported by the observation that some of the patients showed an impaired or diabetic oral glucose tolerance at the time of study (22, 23). Additionally, the blunted insulin response may be a consequence of slow gastric emptying, a view supported by the longer elevation of insulin and C-peptide concentrations during the later periods of the oral glucose challenge test and by the small but significant difference in plasma glucose (Fig. 1A) at 210 and 240 min. Alternatively, it could be a consequence of organ denervation after transplantation. However, the cephalic phase of insulin secretion lasts only a few minutes (24), which makes this explanation unlikely. Also, it has been reported that pancreas denervation does not interfere with pulsatile insulin secretion (25) or the overall incretin stimulation of insulin secretion after oral glucose (26), making it unlikely that these mechanisms contributed to the impaired insulin secretion observed in pancreas transplanted patients.
Glucose stimulated plasma levels of C-peptide and proinsulin did not differ between the two transplanted groups, whereas plasma insulin was higher in the SV group. These findings indicate that although the overall glucose stimulated insulin secretion was similar in the two groups, the relative distribution between the liver and peripheral tissues differed, with a relatively higher hepatic delivery of insulin after PV than SV pancreas drainage, which is in good agreement with the anatomical conditions.
The circulating IGF system originates primarily from the liver, and there is solid evidence that exposing the liver to insulin has important effects on plasma levels of IGF-I and the insulin-sensitive IGFBPs. Thus, in diabetic rats ip insulin was more potent than sc insulin in restoring hepatic IGF-I mRNA synthesis, despite unchanged blood glucose levels (27). In type 1 diabetic patients, continuous ip or intraportal insulin infusion increased circulating IGF-I and IGFBP-3 more efficiently than continuous sc insulin infusion, again without affecting the metabolic control (28, 29), and C-peptide-positive type 1 diabetic patients had higher levels of free and total IGF-I, and lower levels of IGFBP-1 and -2, compared with C-peptide-negative patients with a similar metabolic control (30). Therefore, a greater hepatic insulin exposure provides the most plausible explanation for the observation that patients with PV pancreas drainage had elevated levels of IGF-I, IGF-II, and IGFBP-3 and suppressed levels of IGFBP-1 and -2 when compared with patients with SV drainage (Figs. 3 and 4). Taken together, it seems reasonable to conclude that the portal delivery of insulin plays a crucial role for the synthesis of IGFs and IGFBPs. However, we acknowledge that although the two patient groups were matched with respect to diabetes duration, organ ischemia time, immunosuppressive medication, and time from transplantation to study, the duration of dialysis was longer for patients with SV than PV pancreas drainage. Hence, we cannot fully exclude that some of the observed differences between the two groups may be related to differences in the duration of dialysis.
Previous studies have shown that serum free IGF-I is associated positively with total IGF-I and IGFBP-3 and inversely with IGFBP-1 and -2 (13), and this relationship was maintained in the two transplanted groups. However, in the present study, both transplanted groups were able to maintain normal to elevated concentrations of free IGF-I and -II despite clearly elevated IGFBP-1 and -2 levels. Most likely, this observation is at least partly explained by the use of immunosuppressive drugs. Studies in normal rats (31) and kidney-transplanted children (32, 33) have shown that the combined use of prednisone and cyclosporine and/or azathioprine results in increased concentrations of IGFBP-1, -2, and -3. Similarly, administration of glucocorticoids to healthy subjects has been reported to increase serum total IGF-I (34, 35). Hence, the use of immunosuppressive drugs is likely to influence the circulating IGF-system, and this makes it difficult to compare directly the circulating IGF-system in transplanted patients and normal subjects.
Pancreas transplantation with PV drainage resulted in higher levels of free and total IGF-I than SV drainage, and accordingly the former group had reduced GH levels during the GHRH test. Human in vivo studies have clearly shown that IGF-I administration suppresses the endogenous GH secretion, and there is accumulating evidence supporting that it is circulating free rather than total IGF-I that is responsible for the pituitary feedback control (13, 36, 37). In this context it is noteworthy that free IGF-I correlated inversely with the integrated GH secretion (Fig. 4B), whereas no such correlation was observed for total IGF-I. Finally, measurement of the dimeric complex of IGFBP-1 and IGF-I supported the finding of increased levels of free IGF-I in patients with PV pancreas drainage. Thus, as expected, elevated levels of free IGF-I resulted in an increased saturation of IGFBP-1, which was about 2-fold increased in patients with PV drainage, compared with controls and patients with SV drainage.
Patients with type 1 diabetes on sc insulin therapy are characterized by an exaggerated GH secretion, markedly reduced levels of free IGF-I and -II, subnormal levels of total IGF-I, total IGF-II, and IGFBP-3 and markedly elevated levels of IGFBP-1 and -2 (13). Thus, with respect to the circulating GH/IGF-I axis the conventionally treated type 1 diabetic patient is clearly different from the pancreas transplanted patient, whether the graft is drained into the portal or iliac vein.
Despite an impaired insulin response, most transplanted patients were able to maintain euglycemia during an oral glucose challenge. Because we observed no differences in GH, FFAs, glucagon, or incretin hormones, our observation raises the question whether other glucose regulatory hormones such as IGF-I may be of importance. At present there is little if any support for a role of circulating IGF-I as an acute glucose regulator in vivo (13), but IGF-I appears to be of importance in regulating insulin sensitivity. The exact mechanism remains to be elucidated, but it may involve direct effects of IGF-I on insulin target tissues (38, 39, 40) as well as an IGF-I mediated feedback inhibition of GH secretion (41). Thus, we hypothesize that in patients with PV pancreas graft drainage, the elevated serum levels of free IGF-I are able to maintain an insulin sensitivity at a level sufficient to preserve euglycemia even in the setting of subnormal peripheral insulin levels.
In conclusion, peripheral insulin levels were lower after PV than SV pancreas graft drainage without affecting oral glucose tolerance. In patients with SV drainage, GH responses to GHRH and circulating levels of free and total IGF-I were normal despite an unphysiological pancreas graft drainage circumventing the hepatic circulation. However, reestablishment of the pancreatic-hepatic circuit by PV pancreas graft drainage resulted in a marked stimulation of the circulating IGF-system and accordingly a suppression of the stimulated GH secretion. Stimulation of the circulating IGF system might have contributed to regulation of glucose metabolism in patients with PV pancreas graft drainage.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure Statement: The authors have nothing to disclose.
First Published Online March 4, 2008
1 J.F. and R.A.R. contributed equally to this work. 
Abbreviations: FFA, Free fatty acid; GIP, gastric inhibitory peptide; GLP, glucagon-like peptide; IGFBP, IGF-binding protein; PV, portal venous; SV, systemic venous.
Received October 23, 2007.
Accepted February 26, 2008.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
) and transplanted patients with PV (
) and SV pancreas drainage (
). Data are mean ± SEM. Results of ANOVA are shown in each panel (A: differences between groups; B: changes with time; AB: the interaction of group and time). 
, P < 0.05 for comparison of patients with SV pancreas drainage vs. control subjects. 
, P < 0.05 for comparison of patients with PV pancreas drainage vs. control subjects; 
, P < 0.05 when comparing patients with PV and SV pancreas drainage. For graphs D–F, the ANOVA found significant changes with time or a significant interaction of time x group, but at no single time point a significant difference could be achieved.
