Insulin-like growth factors (IGF)-I and -II and IGF binding protein-1, -2, and -3 in patients with acromegaly before and after adenomectomy

Insulin-like Growth Factors (IGF)-I and -11 and IGF Binding Protein-l, and -3 in Patients With Acromegaly Before and After Adenomectomy Jens O.L. Jorgensen,

Niels Moller, Jens Mailer,

-2,

Jorgen Weeke, and Werner F. Blum

The interrelationship between insulin-like growth factors (IGFs) and their major binding proteins (IGFBPs) as a function of disease activity in acromegaly has not previously been prospectively evaluated. We studied basal and insulin-stimulated serum levels of IGF-I and -II and IGFBP-1, -2, and -3 in six acromegalic patients before and 2 months after successful adenomectomy compared with a group of sex- and age-matched healthy, untreated subjects. All were studied postabsorptively (11 AM) and at the end of a 2-hour euglycemic glucose clamp (0.4 mU insulin/kg x min). Serum IGF-I levels (mean 2 SE) were elevated in acromegaly but were normalized following therapy (basal state IGF-I [pg/L], 857 + 119 [active] Y 255 f 65 [postoperative] v 190 f 20 [control]). Serum IGF-II levels did not change following therapy and were similar to those of the control group. IGF levels did not change during the clamp. Serum IGFBP-3 levels were elevated in active acromegaly, but were normalized after therapy (basal state IGFBPS [pg/L] 6,983 2 612 [active] v 3,939 2 504 [postop] v 3,358 + 125 [control]). The molar ratio of (IGF-I + IGF-II):IGFBP-3 was similar in all studies. Serum IGFBP-1 interacted significantly with time in all studies, exhibiting a gradual decrease in the basal state and ensued by further suppression during the clamp. Insulin and IGFBP-1 correlated inversely in the pooled data and in the acromegalic patients. Serum IGFBP-2 was significantly decreased in active acromegaly, but became normalized following therapy (basal state IGFBP-2 [pg/L], 275 +- 11 [active] v 396 + 16 [postop] v 345 f 19 [control]). Both IGFBP-2 and IGFBP-3 were unchanged during the clamp. Multiple stepwise regression analysis showed IGFBP-2 to correlate inversely with IGF-I, whereas IGF-I and IGFBP-3 intercorrelated positively. In summary, (1) IGF-I but not IGF-II is elevated in active acromegaly, (2) the molar ratios of (IGF-I + IGF-II):IGFBP-3 were similar in all studies, (3) IGFBP-2 levels are decreased in active acromegaly and do not seem to be regulated by insulin, and (4) Successful adenomectomy results in normalization of circulating IGF-I and IGFBPs. We speculate that the increased IGF-I activity in active acromegaly may be accentuated by the decreased IGFBP-2 and -1 levels. Copyright C 1994 by W. B. Saunders Company

S

INCE THE OBSERVATION by Salmon and Daughaday.’ it has become well established that some but not all of the metabolic actions of growth hormone (GH) are mediated through hepatic or peripheral synthesis of insulinlike growth factors (IGFs), in particular IGF-I.? Unlike insulin, IGF-I and -II are associated with several specific binding proteins (IGFBPs).“.~ Although the regulation and functional role of these IGFBPs are far from clarified, there is little doubt that they significantly modulate the actions of IGF-I ,and -11. Consequently, measurements of levels of IGF-I and -II and their binding proteins have attracted attention in studies of growth disorders including acromegaly. Active acromegaly is associated with elevated levels of IGF-1,’ whereas IGF-II remains within the normal range.h It has also been shown that acromegalic patients exhibit elevated levels of IGFBP-3,7,8 which binds 80% to 90% of circulating IGFs. IGFBP-1 levels, on the other hand, are generally low in these patients,x-l” which is thought to be caused by the concomitant hyperinsulinemia. Little information is available about IGFBP-2 levels and its regulation in acromegaly, and in healthy subjects, for that matter. Clemmons et al” reported fasting serum IGFBP-2 levels in untreated acromegalic patients to be similar to those of normal adults, whereas Hardouin et aLx using Western ligand blotting, observed decreased IGFBP-2 signals in untreated acromegaly, which became normalized following treatment. In the present study, we prospectively evaluated serum levels of IGF-I, IGF-II, and IGBP-1, -2, and -3 in acromegalic patients, in the basal state and during euglycemic glucose clamp, before and after successful adenomectomy as compared with a group of healthy volunteers.

Metabolrsm, Vol 43, No 5 (May), 1994: pp 579-583

SUBJECTS AND METHODS Subjects Six patients 41.2 2 4.2 years; diagnosed women,

(four women and two men: [mean ? SE] aged body mass index, 24.3 ? 1.5 kg/m?)

active acromegaly

and 10 healthy control

with newly

subjects

two men; aged 38.6 2 4.6 years) were studied.

(eight

All patients

were studied twice, approximately I month preoperatively and 2 months after transphenoidal pituitary surgery. No evidence of coexisting disease was evident. The diagnosis was ultimately based on persistent elevations of serum GH greater than IO pg/L during 4 hours of sampling (mean k SE GH. 27.8 2 3.5 FgiL) and the presence Pituitary

of a pituitary macroadenomas

adenoma by computed tomography scan. were evident in all patients. One patient

had elevated serum prolactin concentrations preoperatively (79.8 kg/L). which decreased to unmeasurable levels following adenomectomy.

Apart

from slightly subnormal

serum testosterone

levels

in the two male patients, residual pituitary function was normal preoperatively as judged by routine screening. Postoperatively, all patients exhibited a significant decrease in serum IGF-I, and serum GH levels less than 4 pg/L during a 4-hour period (mean 2 SE. 2.2 ? 0.X kg/L). Two months following tients exhibited normal residual anterior

adenomectomy. and posterior

all papituitary

function and received no medications. The study was approved by the local ethical committee.

From Medical Department M (Endocrinology and Diabetes). Aarhus Kommunehospital, Aarhus. Denmark; and Uni~~e~~itrttskinderklinik. Tiibingen. Germany. Submitted Februaty 24, 1993; accepted Ju& 7. 1493. Address reprint requests to Copyright 0 1994 by W.B. Saunders Compu?l\ 0026-049519414305-0010$03.0010

579

J0RGENSEN

580

Design All subjects

were studied in the basal postabsorptive state following a 12-hour fast. At 7:30 AM, intravenous catheters were placed for blood sampling and infusions. At 11:00 AM. a ?-hour constant infusion of regular insulin (Actrapid, Novo Nordisk, Copenhagen. Denmark; 0.4 mu/kg x min) was started. Concomitantly, plasma glucose was clamped at 5 mmol/L as corroborated by adjusting the infusion of a 20% glucose solution to the plasma glucose concentration measured every 5 minutes. Data on intermediary metabolism and insulin sensitivity from the study have been published previously.‘?

Serum IGFBP-1

Active

&----A

After

Adenomectomy

Mean + SE

40

(pg/l)

e--+

ET AL

1

zoSerum GH and insulin levels were measured by radioimmunoassay (RIA) as previously described.i3 Likewise, IGF-I and -II and IGFBP-3 levels were measured by RIAs according to previous descriptions.14-16 IGFBP-I and polyclonal IGFBP-1 antiserum were kindly provided by Dr Hans Bohn (Behringwerke. Marburg, Germany). The tracer was prepared by radioiodination of pure IGFBP-1, and the standard dilutions were made from the pure peptide. Intrassay and interassay coefficients of variation were 3.4% and 8.1%. respectively. Performance of the assay has been described previously.” The IGFBP-2 level was measured by a novel specific RIA. Briefly. a polyclonal antibody against a synthetic partial sequence [hIGFBP-2 (176.190)] was produced in rabbits, No cross-reactivity was detectable by RIA with pure IGFBP-1 or IGFBP-3. The tracer was prepared by radioiodination of (Tyr)-hIGFBP-2 (176-190). For preparation of standards, hIGFBP-2 (176-190) was calibrated with recombinant hIGFBP-2 (kindly supplied by Dr Schwander. Base]. Switzerland). The assay was performed as a conventional RIA using a double-antibody technique for separation of bound and unbound tracer. lntraassay and interassay coefficients of variation were 3.7% and 9.6%%. respectively. Further details of the IGFBP-2 assay have recently been published.‘s All three IGFBP RIAs showed no cross-reaction with the nonmeasured IGFBPs up to 1 FgimL. By Western blotting, no cross-reaction of the three antisera was obtained with other IGFBPs besides the one that was used for antiserum production. Serum IGFBP-1 concentrations were measured in 12 samples during the entire study period as presented in Fig 1. Serum concentrations of IGF-I and -11 and IGFBP-2 and -3 were measured once before the euglycemic clamp (II:00AM) and twice at the end of the clamp (12:40 to I:00 PM).

Statistics Results are expressed as the mean ? SE. ANOVA (general linear model [GLM]) for two repeated measures (time and treatment) was used to compare serum IGFBP-I profiles among patients and control subjects. The relationship between serum insulin and IGFBP-1 was evaluated by conventional correlation analysis. One-way ANOVA for an unbalanced design was used to test for differences in additional parameters between groups. Multiple stepwise regression analysis of the mean levels of IGFBP-2 (dependent variable) and IGFBP-3, IGF-I, and IGF-II (independent variables) was performed with SPSS/PC+ (SPSS, Inc, Chicago, IL). A P value (two-tailed) less than .05 was considered significant for all statistical procedures. RESULTS

IGFBP-1 A significant change with time of serum IGFBP-1 was observed in the patients both before and after adenomectomy, as well as in the control group (P = .03). This was

10

-I

CLAMP 1

8

I

I

11

13 Cloktime

Fig 1. Mean + SE serum IGFBP-1 levels before and after adenomectomy in six patients with acromegaly. The shaded area represents the healthy control group.

characterized by a gradual decrease during the basal period, ensued by a further decline at the end of the clamp. This pattern was similar in the three studies, as evidenced by the lack of a significant time-treatment effect (P = .I). Furthermore, the absolute levels also failed to exhibit significant differences. but serum IGFBP-I tcndcd to be suppressed during active acromegaly and elevated after adenomectomy (P = .08). A significant inverse correlation was observed between serum insulin and IGFBP-1 levels when pooling the data from all three studies and also when comparing the two acromegalic studies separately. An even closer correlation was observed following log 10 transformation of serum insulin concentrations (all three studies, r = A, P < .OOl; active acromegaly, r = .71. P < ,001; following adenomectomy, r = 36, P < ,001). In contrast, when comparing the two parameters in the control group, no significant correlation was observed (r = .40. P = .16: Fig 1). IGF-I and -I1 Serum IGF-I concentrations were significantly clcvated in active acromegaly. Following adenomectomy. serum IGF-I decreased to a level not significantly diffcrcnt from that of the control group (basal state IGF-I [kg/L], 857 + 119 [active] v 255 ‘_ 65 [adenomectomy] v 190 ? 20 [control]). Serum IGF-II levels, on the other hand, were similar in all three studies (basal state IGF-II [pg/L], 421 2 19 [active] 11426 + 17 [adenomectomy] 11441 * 21 [control]). IGF-I and -II levels during the basal state and cuglycemic clamp did not differ significantly. On a molar basis, serum IGF-I levels (mean of basal and clamp values)

IGFs AND

IGFBPs

581

IN ACROMEGALY

BASAL STATE --P~oo1---P < 001.

500

CLAMP --P~OO,---P c 001.

x ,1

0 Serum IGF-II

500

j:

@?3N

OJ

Before

After

Fig 2. Mean + SE serum IGF-I and IGF-II levels in patients before and after adenomectomy, and in the healthy control group: Measurements were performed both in the basal state and during a euglycemic glucose clamp.

decreased from 109.5 following adenomectomy

2 15.5 to 32.5 (Fig 2).

k

8.1

mmol/L

IGFBP-3 mid -2 Serum IGFBP3 levels were increased in patients with active acromegaly, whereas adenomectomy was associated with a complete normalization of this parameter (basal state IGFBP3 [pg/L]. 6,983 2 612 [active] v 3,939 + 504 [adenomectomy] v 3,358 2 125 [control]). Similar levels were recorded during the basal state and euglycemic clamp. On a molar basis, serum IGFBP-3 levels decreased from 228.0 :k 18.8 to 130.1 +- 16.6 mmol/L following adenomectomy. The molar ratio of (IGF-I + IGF-II):IGFBP-3 did not change significantly following adenomectomy and was comparable to that of the control group (0.73 k 0.03 [active] 1’ 0.69 2 0.02 [adenomectomy] v 0.76 k 06 [control]). Serum IGFBP-2 levels exhibited changes opposite those of IGFBP-3, with significantly decreased levels during active acromegaly and increases following adenomectomy to a level comparable to that of the control group. However, again, the euglycemic clamp failed to impose any separate effect (Fig 3). When performing multiple stepwise regression analysis of the levels of IGF-I, IGF-II, IGFBP-3, and IGFBP-2 (overall mean values of basal and clamp levels in the three studies pooled together) with IGFBP-2 as the dependent variable, only IGF-I was found to correlate significantly, in an inv’erse manner, with IGFBP-2 (partial regression coefficient := -.58. P < .Ol). In addition, IGF-I and IGFBP-3 were found to intercorrelate positively (partial regression

coefficient = .9, P < .OOl). When analyzing the three studies separately, the inverse correlation between IGFBP-2 and IGF-I only reached statistical significance in the acromegalic patients following adenomectomy (partial regression coefficient = -.9, P < .OOl). DISCUSSION

This longitudinal study is the first to scrutinize concomitant changes in IGF-I and -II and their major binding proteins in acromegalic patients before and after successful surgical therapy during well-defined conditions. Distinct abnormalities of all parameters except IGF-II were recorded in untreated acromegaly, but an almost complete normalization was observed following adenomectomy. It was furthermore observed that only serum IGFBP-1 was affected by short-term changes in serum insulin levels. The study comprised a low number of patients, but on the other hand, they were studied in a prospective and strictly defined design. Still, caution should always be exercised when extrapolating data from a limited group to a larger population. The elevation of serum IGF-I in active acromegaly is well described, and there is evidence to indicate that serum IGF-I rather than GH measurements may be more sensitive to recognize persistent disease activity following therapy.?J,h.‘y Ou r observations concerning serum IGF-II levels extend the original finding by Zapf et al,h who reported that IGF-II was not elevated in acromegalic patients. The fact that serum IGF-II was virtually similar in our three studies underlines the contention that IGF-II production is not regulated by GH. The inverse correlation between circulating IGFBP-1 and insulin reported in our study seems to be a ubiquitous finding and is presumed to represent a direct inhibitory effect of insulin on IGFBP-1 production.?“.?’ The finding Serum IGFBP-3

BASAL STATE 10000

--Pcoo1---

(KIN

-P 40,1

Serum IGFBP-2 WI/l)

0

1

Before

After

Fig 3. Mean -t SE serum IGFBP-3 and IGFBP-2 levels in patients before and after adenomectomy, and in the healthy control group: Measurements were performed both in the basal state and during a euglycemic glucose clamp.

582

JBRGENSEN

that this inverse correlation did not reach statistical significance in all studies when evaluated separately is most likely due to the limited number of observations. Suppressed IGFBP-1 levels have previously been reported in active acromegaly,x-iO but the observation of a preserved decrease in the morning ensued by a further suppression during insulin infusion, as also seen in normal subjects, is new. We furthermore found serum IGFBP-1 levels to be slightly supranormal following adenomectomy, which contrasts with the previous report of Holly et al,l” in which serum IGFBP-1 (at 9:00 AM) was still suppressed in a heterogenous group of acromegalic patients who had been operated on, irradiated, or treated with somatostatin analogs or bromocriptine. This discrepancy can probably be explained by the fact that their patients still had elevated insulin levels despite therapy.“’ Information on IGFBP-2 levels in acromegaly is sparse. Clemmons et aIri did not detect any differences when comparing serum IGFBP-2 concentrations (at 8:OO AM) in normal subjects and a group of 31 acromegalic patients, although the mean value in the latter group was lower. Our finding of decreased IGFBP-2 levels in active acromegaly is compatible with two animal experiments demonstrating a suppressive effect of GH on IGFBP-2.“,” However, other animal studies have failed to show such an effect of GHZ.i? it has also been reported that GH administration to obese, calorically restricted subjects had no impact on IGFBP-2.” and we have previously been unable to detect any effects of GH (infused for 14 hours) on IGFBP-2 in GH-deficient subjects (unpublished data). Furthermore, fasting, which is associated with elevated GH levels, has been reported to cause a substantial increase in IGFBP-2.” It therefore seems unlikely that GH per se regulates IGFBP-2 metabolism. Insulin has been suggested to inhibit IGFBP-2 inductionL5 However, both the present study and that of Clemmons et ali’ did not record any changes in IGFBP-2 levels during insulin infusion. To reconcile the different findings on IGFBP-2 regulation, it could be speculated that IGFBP-3, which is increased in acromegaly and during GH administration to nonfasting subjects, and furthermore is decreased during fasting, may act to suppress IGFBP-2. This hypothesis is also compatible with the observation that IGFBP-2 is predominant in early fetal circulation, whereafter the levels decrease concomitantly with an increase in IGFBP-3.zh Our

ETAL

finding of a negative correlation between IGF-I and IGFBP-2 should be interpreted with great caution due to the limited number of observations, which furthermore were pooled from all three studies. With these reservations in mind, we suggest that the strong positive intercorrelation between IGF-I and IGFBP-3 is supportive of a regulatory role for IGFBP-3 in IGFBP-2 production. The notions that IGF-II as compared with IGF-I has a higher affinity for IGFBP-2 and that IGFBP-2 is elevated in patients with IGF-II-producing tumors have led to the suggestion that IGF-II regulates IGFBP-2 production.‘” The present study does not add direct support to that hypothesis, but it is noteworthy that IGFBP-3 decreased more than IGF-I in absolute molar terms following adenomectomy. Assuming unaltered IGF-II production, this would transiently result in more free IGF-II following adenomectomy, which then in turn could stimulate IGFBP-2 production. Still. such considerations remain speculative, and the physiological role of IGFBP-2 is still unclear. Like IGFBP-I, it has been shown to cross the capillary membrane” and may thus serve as a transporter of IGFs to the tissues. On the other hand. the kinetics of IGFBP-2 are clearly distinct from IGFBP-I, with almost constant circadian serum IGFBP-2 levels.” The increased levels of serum IGFBP-3 in active acromegaly have previously been reported.‘,” In adult circulation, IGFBP-3 is by far the major BP and is speculated to function as a reservoir and buffer for IGFs. The molar ratio of (IGF-I + IGF-II):IGFBP-3 was similar when comparing active and treated acromegaly and control subjects. Since active acromegaly is associated with increased IGF action, it could be speculated that this increased activity may be accentuated by the concomitantly decreased levels of IGFBP-1 and -2. In general, high IGFBP-1 lcvcls are associated with conditions of low IGF-I activity and vice versa.z8~2y However, it is unclear whether the decreased IGFBP-1 levels are associated with elevated levels of free IGF-I in the circulation, or whether IGF-I bound to IGFBP-1 is transported to the target cells. Clearly, further and more sophisticated studies are needed to outline the complex interaction between IGFs and their binding proteins. ACKNOWLEDGMENT

Anette Mengel is thanked

for superb

technical

assistance.

REFERENCES 1. Salmon WD, Daughaday WH: A hormonally controlled serum factor which stimulates sulfate incorporation by cartilage in vitro. J Lab Clin Med 49:825-836. 1957 2. Daughaday WH, Rotwein P: Insulin-like growth factors I and II. Peptide, messenger ribonucleic acid and gene structures. serum. and tissue concentrations. Endocr Rev 10:68-91, 1989 3. Reports on the nomenclature Clin Endocrinol Metab 70:817-818.

of the IGF binding 1990

proteins.

J

4. Report on the nomenclature of the IGF binding Clin Endocrinol Metab 130:1736-1737, 1992

proteins.

J

5. Clemmons DR, Van Wyk JJ, Ridgway EC, et al: Evaluation of acromegaly by radioimmunoassay of somatomedin-C. N Engl J Med 301:1138-l 142, 1979

6. Zapf J, Walter H. Froesch ER: Radioimmunological determination of insulin-like growth factors I and II in normal subjects and in patients with growth disorders and extrapancreatic tumor hypoglycemia. J Clin Invest 68:1321-1330, 1981 7. Baxter RC. Martin JL: Radioimmunoassay of growth hormone dependent insulin-like growth factor binding protein in human plasma. J Clin Invest 78:1504~1512, 1986 8. Hardouin S. Gourmelen M, Noguiez P, et al: Molecular forms of serum insulin-like growth factor (IGF)-binding proteins in man: Relationships with growth hormone and IGFs and physiological significance. J Clin Endocrinol Metab 69:1291-1301. 1989 9. Hall K. Lundin G. Povoa G: Serum levels of the low molecular weight form of insulin-like growth factor binding protein

IGFs AND IGFBPs IN ACROMEGALY

583

in healthy subjects and patients with growth hormone deficiency, acromeg,aly and anorexia netvosa. Acta Endocrinol 1 l&321-326, 1988 IO. Holly JMP, Cotterill AM. Jemmott RC, et al: Inter-relations between growth hormone, insulin, insulin-like growth factor-l (IGF-I). IGF-binding protein-l (IGFBP-I) and sex hormonebinding globulin in acromegaly. Clin Endocrinol34:275-280. 1991 11. Clemmons DR, Snyder DK, Busby WH: Variables controlling the secretion of insulin-like growth factor binding protein-2 in normal human subjects. J Clin Endocrinol Metab 73:727-733, 1991 12. Moller N, Schmitz 0. Jorgensen JOL, et al: Basal and insulin-stimulated substrate metabolism in patients with active acromegaly before and after adenomectomy. J Clin Endocrinol Metab 74:1012-1019, 1992 13. (Zlrskov H, Thomsen HG, Yde H: Wick chromatography for rapid and reliable immunoassay of insulin. glucagon and growth hormone. Nature 219:193-195, 196X 14. Blum WF. Ranke MB. Bierich JR: Isolation and partial characterization of six somatomedin-like peptides from human plasma Cohn fraction IV. Acta Endocrinol (Copenh) I 11:271-284. 19X6 15. Blum WF, Ranke MB. Bierich JR: A specific radioimmunoassay for insulin-like growth factor II: The interference of IGF binding proteins can be blocked by excess IGF-I. Acta Endocrinol I l&374-380, 1988 16. Blum WF. Ranke MB. Kietzmann K: A specific radioimmunoassay for the growth hormone (GH)-dependent somatomedinbinding protein: Its use for diagnosis of GH deficiency. J Clin Endocrinol Metab 70:1292-1298, 1990 17. B,lum WF. Ranke MB. Kietzman K: Growth hormone resistance and inhibition of somatomedin activity by excess of insulin-like growth factor binding protein in uremia. Pediatr Nephrol S:539-544. 1991 18. Elum WF, Horn N. Kratzsch J. et al: Clinical studies IGFBP-2 by radioimmunoassay. Growth Regu 3:100-104, 1993

of

19. Furlanetto RW. Underwood LE. Van Wyk JJ: Estimation of somatomedin-C levels in normals and patients with pituitary disease by radioimmunoassay. J Clin Invest 60:648-657. 1977 70. Conover

CA, Lee PDK:

Insulin

regulation

of insulin-like

growth factor binding protein production in cultured hepG2 cells. J Clin Endocrinol Metab 70:1062-1067. 1990 31. Zapf J, Hauri C, Waldvogel M. et al: Recombinant human insulin-like growth factor I induces its own specific carrier protein in hypophysectomized and diabetic rats. Proc Natl Acad Sci USA 863813-3817, 1989 22. Cohick WS, McCusker RH, Clemmons DR: Insulin-like growth factor binding proteins (IGFBPs) modulate IGF-I binding and cell growth responses of MDBK cells. Proceedings of the 72nd Annual Meeting of The Endocrine Society. Atlanta, GA. June 1990 23. Ooi CT. Orlowski CC, Brown AL, et al: Different tissue distribution and hormonal regulation of messenger RNAs encoding rat insulin-like growth factor-binding proteins-l and -2. Mol Endocrinol4:321-328, 1990 24. Margot JB, Binkert C. Mary J-L. et al: A low molecular weight insulin-like growth factor binding protein from rat: cDNA cloning and tissue distribution of its messenger RNA. Mol Endocrinol3:1053-1060, 1989 25. Zapf J. Schmid C, Guler HP, et al: Regulation of binding proteins for insulin-like growth factors (IGF) in humans. Increased expression of IGF binding protein 2 during IGF I treatment of healthy adults and in patients with extrapancreatic tumor hypoglycemia. J Ciin Invest 86:952-961. 1990 26. McCusker RH, Campion DR. Clemmons DR: The ontogeny and regulation of a 31.000 Mr insulin-like growth factor/ somatomedin (IGF) binding protein in fetal porcine plasma and sera. Endocrinology 122:2071-2079. 1988 27. Bar RS. Clemmons DR, Busby WH, et al: Transcapillary permeability and subendothelial distribution of endothelial and amniotic fluid (IGFBP-1) IGF-binding proteins tn the rat heart. Endocrinology 127:107X-1086, 1990 28. Hall K. Brismar K, Grissom F, et al: IGFBP-1. Production and control mechanisms. Acta Endocrinol (Copenh) 124:48-54. 1991 29. Taylor AM, Dunger DB, Preece MA. et al: The growth hormone independent insulin-like growth factor-l binding protein BP-28 is associated with serum insulin-like growth factor-l inhibitory bioactivity in adolescent insulin-dependent diabetics. Clin Endocrinol 32:229-239. 1990