(Circulation. 1996;94:2465-2471.)
© 1996 American Heart Association, Inc.
Articles |
the Institute of Experimental Clinical Research, Aarhus (Denmark) University Hospital, Kommunehospitalet (J.F., C.S., H.Ø.); IPSEN ApS, Copenhagen, Denmark (N.A.); the Department of Cardiology, Sahlgrenska Sjukhuset, Goteborg, Sweden (H.E.); the Department of Cardiology, Hospital de Weezenlanden, Zwolle, the Netherlands (H.S.); the Department of Cardiology, Waldkrankenhaus St. Marien, Erlangen, Germany (H.B.); and Georgetown University Medical Center and Henri Beaufour Institute-USA (M.F.).
Correspondence to Jan Frystyk, MD, Institute of Experimental Clinical Research, Aarhus Kommunehospital, Nørrebrogade 44, DK-8000 Aarhus C, Denmark.
| Abstract |
|---|
|
|
|---|
Methods and Results We measured fasting serum total (extractable) IGF-I in 553 patients in a controlled clinical trial. Half of the patients received continuous subcutaneous infusion of the somatostatin analogue lanreotide from the day before (baseline) and for 4 days after PTCA. We also measured ultrafiltrated serum free IGF-I and IGF-II, total IGF-II, IGF-binding protein-1 (IGFBP-1), IGFBP-3, and insulin in a subgroup of 18 placebo-treated and 20 lanreotide-treated patients. Total IGF-I had decreased by 7% (P<.0001) 1 day after initiation of lanreotide infusion and stayed reduced, whereas no early changes occurred in placebo-treated patients. The same pattern was observed in the subgroup. Free IGF-I decreased significantly from baseline by 22% to 27% (P<.05) in lanreotide-treated patients and increased insignificantly by 10% to 30% (P=.054) in placebo-treated patients. IGFBP-1 increased (P<.05) in both groups postoperatively, but levels in lanreotide-treated patients exceeded (P<.05) those of placebo-treated patients. Lanreotide treatment resulted in minor reductions (P<.05) in free and total IGF-II and IGFBP-3, whereas insulin was unaltered.
Conclusions Lanreotide administration acutely decreases circulating total and free IGF-I, the latter relatively more, and increases IGFBP-1. These alterations in the IGF system may participate in the improvement of the long-term outcome after PTCA noted with lanreotide treatment.
Key Words: growth substances proteins angiography muscle, smooth
| Introduction |
|---|
|
|
|---|
Animal studies have shown that administration of the somatostatin analogue lanreotide inhibits SMC proliferation and thymidine incorporation in vivo and in vitro.9 10 11 Lanreotide also prevents the vascular IGF-I accumulation transiently occurring in the first days after endothelial denudation.12
The present study included 553 patients and was part of a multicenter, double-blind, controlled clinical trial that compared the effects of lanreotide versus placebo on 1-year clinical outcome after PTCA.3 Lanreotide treatment (n=278) significantly (P<.05) reduced the frequency of clinical events after 12 months and significantly improved event-free survival compared with placebo treatment (n=275).
The aim was to compare serum total (extractable) IGF-I in patients treated with lanreotide versus placebo. Serum samples were drawn for analysis of total IGF-I before, during, and after angioplasty. In a subgroup of 38 patients from two centers, the measurements were extended to include circulating levels of ultrafiltrated free IGF-I, free and total IGF-II, insulin, IGFBP-1, and IGFBP-3.
| Methods |
|---|
|
|
|---|
Fasting blood samples were obtained the day before PTCA (baseline), just before PTCA (day 0), the following 2 days (days 1 and 2), and at a postdrug follow-up interval on days 6 through 11. Serum was stored at -80°C until analysis. As will be seen, only a few centers were able to obtain all specimens. In particular, the day 2 samples were often lacking because patients had been discharged. In a small subgroup of patients from two centers comprising 18 placebo-treated and 20 lanreotide-treated subjects, practically all samples were successfully obtained. In these specimens, ultrafiltrated serum free IGF-I and IGF-II, IGFBP-1, and insulin were determined at baseline and days 1 and 2; serum total IGF-II and IGFBP-3 were determined during all days.
IGF-I and IGF-II were measured by use of two monoclonal antibody noncompetitive time-resolved immunofluorometric assays.13 Both assays are characterized by a high sensitivity (detection limits, 2.5 [IGF-I] and 10 [IGF-II] ng/L), a three-decade linear calibration curve (from 5 ng/L to 2.5 µg/L [IGF-I] and from 20 ng/L to 10 µg/L [IGF-II]), and a cross-reactivity in heterologous assays <0.0002%.
Serum total IGF-I and IGF-II were determined after acid ethanol extraction14 ; serum extracts were assayed directly in duplicate after dilution in assay buffer (final dilution, 1:1066 [IGF-I] and 1:2132 [IGF-II]) as previously described.13 Intra-assay and interassay coefficients of variation were <5% and <10%, respectively.
Serum free IGF-I and IGF-II were determined in triplicate by use of ultrafiltration by centrifugation.15 In brief, Amicon YMT 30 membranes and MPS-1 supporting devices were used (Amicon Division, WR Grace and Co). The membrane has a nominal cutoff value of
25 kD and was selected from several candidates because of its superior recovery of free IGF and its IGFBP-free filtrate.15 Before centrifugation, serum samples were diluted (1:11) in Krebs'-Ringer buffer containing 5% (wt/vol) human serum albumin (Behring AG) that had been adjusted to pH 7.4 by airing with CO2. From each dilution, an aliquot of 600 µL was applied to the membranes, incubated (30 minutes at 37°C), and centrifuged (1500 rpm at 37°C; model Rotixa/RP [swinging bucket rotor], Hettich Zentrifugen). Ultrafiltrates were collected in 5-mL polyethylene tubes that before centrifugation were coated with human serum albumin. In the ultrafiltrates, the detection limit was 52.5 (IGF-I) and 210 (IGF-II) ng/L. Intra-assay and interassay coefficients of variation after ultrafiltration and analysis were determined in two control sera and averaged 27% and 29%, respectively.
IGFBP-3 was measured by radioimmunoassay (Diagnostic System Laboratories Inc), IGFBP-1 by ELISA (Medix Biochemica), and insulin by a polyclonal radioimmunoassay using rh-insulin and I125-rh-insulin as calibrator and tracer (Novo Nordisk A/S).
Statistical Analysis
To compare clinical information on the small subpopulation of patients obtained from the two centers with information from the full sample population, we tested the placebo-treated and the lanreotide-treated subpopulations versus the full sample population minus the subpopulation using Fisher's exact test or an unpaired t test.
Data are described by mean±SD, although the statistical analysis is based on rank sum tests that compared medians because of skewed distributions (heteroscedastic data). Data within the groups were analyzed by use of Friedman's repeated measures ANOVA, followed by Wilcoxon's paired test for comparison of changes from baseline (
values) and for comparison of absolute levels. Data between the groups were analyzed by use of Mann-Whitney's unpaired test. All correlations were performed on log-transformed data. A value of P<.05 was considered significant.
| Results |
|---|
|
|
|---|
|
As expected, a weak age-dependent reduction in baseline levels of serum total IGF-I was observed in the entire population (data not shown), and significant (P<.05) differences between centers were observed (data not shown), probably because of enrollment of different age groups.
Serum Free and Total IGF and IGFBPs in a Subgroup of 38 Patients From Two Centers
Although the subgroup included patients from only two centers, it was representative of the total population, and only minor differences were detected (Table 2
). In the lanreotide-treated subpopulation, the fraction of individuals with a high risk for restenosis tended to exceed the expected value (11.5% versus 7.2%; P=.057), but we did not observe any significant difference in the severity of coronary heart disease or in well-established disposing risk factors.
|
Serum total IGF-I was similar in the two groups at baseline and remained unchanged after placebo treatment (Table 3
). In contrast, significant reductions were observed with lanreotide treatment at days 0, 1, and 6 through 11 (P<.05), and a clear trend was seen at day 2 (P=.073). Comparison of changes from baseline showed that lanreotide caused significant differences between the two groups at days 0, 1, and 2 (P<.05), whereas no difference was observed at days 6 through 11.
|
Serum total IGF-II was similar in the two groups at baseline and remained unchanged after placebo treatment: 754±193, 784±194, 728±189, 763±192, and 756±177 µg/L (baseline and days 0, 1, 2, and 6 through 11, respectively). Lanreotide treatment reduced levels at days 6 through 11 compared with baseline: 784±176, 750±173, 766±228, 772±196, and 728±117 µg/L (P<.05). Comparing changes from baseline, we observed differences at days 0 and 6 through 11 (P<.05; data not shown).
Serum IGFBP-3 was similar in the two groups at baseline. In placebo-treated patients, levels were decreased at day 0 but returned to baseline at day 1 and remained unchanged: 4.22±1.08, 3.90±1.05 (P<.05), 4.01±1.07, 4.33±1.06, and 4.21±1.24 mg/L (baseline and days 0, 1, 2, and 6 through 11, respectively). IGFBP-3 was also significantly decreased at day 0 in lanreotide-treated patients and at days 6 through 11: 3.90±0.74, 3.14±0.77 (P<.05), 3.87±0.93, 3.75±0.93, and 3.46±0.59 (P<.05) mg/L, respectively. This late reduction in IGFBP-3 differed from that observed in placebo-treated patients (P<.05; data not shown).
Serum free IGF-I averaged 730±240 ng/L (baseline) and increased, albeit insignificantly, after PTCA to 800±250 ng/L (10%; P=NS) at day 1 and to 950±330 ng/L (+30%; P=.054) at day 2 in placebo-treated patients (see the Figure
, part A). In contrast, lanreotide treatment significantly decreased serum free IGF-I from 680±400 (baseline) to 530±380 ng/L (-22%; P<.05) at day 1 and to 500±370 ng/L (-27%; P<.05) at day 2. Comparison of the two groups showed that changes from baseline were significantly different at days 1 and 2 (P<.05; Table 4
). These changes were so pronounced that absolute levels of serum free IGF-I also differed between placebo- and lanreotide-treated patients at days 1 and 2 (P<.05; the Figure
).
|
|
Serum free IGF-II was unchanged in both groups and averaged 1500±920, 1310±770, and 1630±880 ng/L (P=NS) in placebo-treated patients and 1300±670, 1060±800, and 1100±700 ng/L (baseline and days 1 and 2, respectively; P=.058) in lanreotide-treated patients (see the Figure
, part B). No difference in changes from baseline was observed; Table 4
. However, the absolute levels differed at day 2 (P<.05). At all 3 days, free IGF-I and total IGF-I correlated in lanreotide-treated patients (all P<.05; all r>.53), whereas no association was observed in placebo-treated patients. No correlation between free and total IGF-II was observed.
Serum IGFBP-1 increased significantly in both groups after PTCA, averaging 2.4±3.4, 3.9±3.3 (+63%; P<.05), and 3.8±3.2 (58%; P<.05) µg/L in placebo-treated patients and 2.5±2.7, 6.7±4.2 (+168%; P<.05), and 6.3±3.6 (152%; P<.05) µg/L in lanreotide-treated patients (baseline and days 1 and 2, respectively; Table 4
). Although the changes from baseline did not differ significantly between the groups, the absolute levels of IGFBP-1 in lanreotide-treated patients exceeded that of placebo-treated patients (P<.05; see the Figure
, part C). At days 1 and 2, free IGF-I and free IGF-II were inversely correlated to IGFBP-1 (all P<.05; all r<-.46) in lanreotide-treated patients, whereas no correlations were observed in placebo-treated patients.
Serum insulin was similar in placebo- and lanreotide-treated patients at baseline and increased insignificantly in both groups postoperatively, averaging 37±7, 63±8, and 74±15 mU/L (P=.12) in placebo-treated patients and 29±4, 40±6, and 51±8 mU/L (P=.32) in lanreotide-treated patients (baseline and days 1 and 2, respectively). Comparing changes from baseline, we observed no differences between the two groups. There was no significant correlation between serum insulin and serum IGFBP-1.
| Discussion |
|---|
|
|
|---|
Unexpectedly, in the lanreotide-treated group, serum total IGF-I was still highly significantly reduced by 15% at the follow-up interval at days 6 through 11, ie, 2 to 7 days after withdrawal of lanreotide. A 4% yet significant reduction was also observed in the placebo-treated patients. We cannot explain this late decrease in serum total IGF-I; still, it was significantly greater in lanreotide- than in placebo-treated patients. It may in part be associated with altered lifestyles after PTCA.
Serum free and total IGF-II were both relatively unaffected by lanreotide treatment. This is in accordance with previous observations of unaltered levels of serum free and total IGF-II in acromegalic patients treated with octreotide.16
The cellular events occurring after endothelial denudation begin with local proliferation of SMCs migrating to the intima within a week. Soon, the vessel is reendothelialized and SMC proliferation diminishes.1 Normally, the vascular expression of IGF-I is very low, but a transient increase is observed after balloon injury.17 In rats, the early vascular changes during the first week after injury include increased expression of IGF-I peptide, IGF-I mRNA, and thymidine incorporation into DNA, whereas results on IGF-I receptor mRNA have been contradictory.18 19 20 21 Recently, SMCs from de novo and restenotic human coronary atherectomy plaques were found to intensively accumulate IGF-I to a significantly higher degree than that of normal coronary SMCs.22 In summary, the events leading to restenosis after endothelial denudation appear to be initiated at the time of manipulation and to be intimately related to the IGF system.
The relative importance of circulating versus local paracrine or autocrine IGF-I for tissue growth in general has been a matter of dispute for about 2 decades. In intact rats, endothelial denudation increases the vascular SMC accumulation of IGF-I without concomitant changes in hepatic IGF-I mRNA expression and serum total IGF-I.21 In hypophysectomized rats, the expression of IGF-I in the vessel wall was found to be independent of pituitary growth hormone secretion.23 These findings could indicate that locally produced IGF-I is more important than circulating IGF-I. However, others have reported that SMCs from vessels subjected to endothelial denudation respond to the addition of exogenous IGF-I by further increased DNA synthesis.19 In humans with insulin-dependent diabetes mellitus, growth hormone deficiency reduced the incidence of macroangiopathy and microangiopathy compared with growth hormoneintact diabetics.24 25 We believe that the present sum of evidence26 indicates that IGF-I actions are mediated through endocrine as well as paracrine and autocrine mechanisms. In the present study, lanreotide significantly improved the clinical outcome after PTCA,3 but this was not correlated to the levels of circulating IGF-I. It may be that lanreotide exercises important biological effects locally by inhibiting the accumulation of IGF-I in the coronary vascular SMCs, as has been shown in rabbits in vivo.12 Such vascular changes may not be visible in the circulation. However, in human in vivo studies, the only way to monitor a possible effect of lanreotide on the IGF system is by measuring the circulating IGF levels.
To avoid restenosis and improve the clinical outcome of PTCA, interest has focused on the exploitation of somatostatin analogues, which are capable of suppressing IGF-I locally and systemically.27 Lanreotide has proved effective in reducing vascular stenosis after experimental endothelial denudation.9 10 11 We have recently shown that in placebo-treated rabbits, arterial IGF-I immunoreactivity increased about fourfold 4 days after endothelial denudation, whereas no increase was observed in rabbits treated with lanreotide.12 In vitro cultured human coronary SMCs are stimulated by IGF-I, whereas coincubation with lanreotide diminishes the growth response.22
Two other lanreotide PTCA trials have been performed.28 29 In the first study,28 lanreotide was administered in a different mode (two daily injections subcutaneously). This study failed to show statistically significant effects on clinical outcome and angiographic end points. However, as Kent et al28 pointed out, a schedule of two injections subcutaneously per 24 hours is insufficient to maintain stable, high plasma levels of lanreotide, considering that this peptide has a plasma half-life of about 90 minutes. In this context, we have recently compared the effects of three daily injections of octreotide versus a long-acting formulation given as one monthly injection on the IGF system in acromegalic patients.16 With respect to suppression of serum total IGF-I, no difference was observed. However, serum free IGF-I was significantly more reduced, with the latter therapy inducing a high, stable plasma somatostatin analogue concentration and a high, persistent elevation of IGFBP-1.16 In a Scandinavian study comprising 112 patients, lanreotide was infused on the same time schedule as in the present study but in a lower dosage.29 The angiographically documented restenosis was significantly reduced, and although the clinical event rate was not altered significantly, it showed a reduction of 26%. In the present study, a statistically significant effect on clinical events was observed despite similar angiographic results.3 However, the angiographic examination may be insufficiently sensitive to detect small differences in minimal lumen diameters.29 Lanreotide administration appears to be efficient in experimental in vitro and in vivo studies as well as in human trials, probably because of an effect on coronary SMCs. We cannot, however, exclude the possibility that lanreotide has other beneficial effects. An earlier study in insulin-dependent diabetes mellitus patients treated with octreotide for 1 year reported an improved lipid profile and altered coagulation parameters, both of which may be important in patients with ischemic heart disease,30 but it is unknown whether these effects are induced in only 5 days' treatment and persist months later.
The studies of a biological role of IGFBP-1 in the regulation of the proliferative effects of IGF-I have been conflicting. Most in vitro studies report that IGFBP-1 inhibits IGF-I action, whereas others observe an increased response to IGF-I after IGFBP-1 administration (for review, see References 5, 31, and 32). The reason for this controversy remains unclear but may be due to the application of different cell lines and preparations of IGFBP-1, which occurs in several phosphorylated forms.33 34 Interestingly, a recent study found that the administration of nonphosphorylated rh-IGFBP-1 to hypophysectomized rats inhibited the growth-promoting effects of IGF-I and growth hormone in vivo.35 It is possible that the inhibitory effects of IGFBP-1 observed in vivo and in vitro are caused by sequestration of IGF-I.32 36 In the present study, serum IGFBP-1 increased significantly from baseline in both placebo- and lanreotide-treated patients, and although IGFBP-1 clearly tended to increase more in the latter group postoperatively, this increase did not achieve significance (.07<P<.10). Comparison of absolute levels, however, showed that serum IGFBP-1 was identical in placebo- and lanreotide-treated patients at baseline but significantly higher at days 1 and 2 in lanreotide-treated patients. This is in agreement with observations that octreotide stimulates IGFBP-1 release in vivo in acromegalic patients16 37 and normal subjects.38 We propose that the reduction in serum free IGF-I after lanreotide treatment may be due to an inhibitory effect on IGF-I synthesis and to the stimulation of serum IGFBP-1. This is supported by the inverse association between free IGF-I and IGFBP-1 in the lanreotide-treated patients.
In conclusion, continuous infusion of lanreotide significantly decreased serum total IGF-I after PTCA. In the subgroup, the most marked inhibitory effect of lanreotide, however, was observed in levels of free IGF-I. In addition, lanreotide increased IGFBP-1, which is considered to be an inhibitor of IGF bioactivity. Obviously, we could not in this study evaluate local effects of lanreotide on IGF-I synthesis but suggest that the early decrease in circulating free and total IGF-I may participate in the improved clinical outcome.
| Selected Abbreviations and Acronyms |
|---|
|
| Acknowledgments |
|---|
Received January 23, 1996; revision received June 18, 1996; accepted July 8, 1996.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
R. J. Irving, M. N. Carson, D. J. Webb, and B. R. Walker Peripheral Vascular Structure and Function in Men with Contrasting GH Levels J. Clin. Endocrinol. Metab., July 1, 2002; 87(7): 3309 - 3314. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Ruotolo, P. Bavenholm, K. Brismar, S. Efendic, C.-G.o. Ericsson, U. de Faire, J. Nilsson, and A. Hamsten Serum insulin-like growth factor-I level is independently associated with coronary artery disease progression in young male survivors of myocardial infarction: beneficial effects of bezafibrate treatment J. Am. Coll. Cardiol., March 1, 2000; 35(3): 647 - 654. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Frystyk, T. Grofte, C. Skjarbak, and H. Orskov The Effect of Oral Glucose on Serum Free Insulin-Like Growth Factor-I and -II in Healthy Adults J. Clin. Endocrinol. Metab., September 1, 1997; 82(9): 3124 - 3127. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Circulation Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 1996 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |