- Email: [email protected]
Somatostatin analogues inhibit angiogenesis in the chick chorioallantoic membrane
JOURNAL
OF SURGICAL
RESEARCH
60,245-251
(1991)
Somatostatin Analogues Inhibit Angiogenesis in the Chick Chorioallantoic Membrane EUGENE A. WOLTERING, M.D., ROSEMARY BARRIE, M.S., THOMAS M. O’DORISIO, M.D.,* DENNIS ARCE, M.D., TINA URE, M.D., ANDREW CRAMER, M.D., DOUGLAS HOLMES, M.D.,? JOSEPH ROBERTSON, M.D.,? AND JOHN FASSLER, M.D.* Oregon Health Sciences University, Departments of Surgery and tOphthalmology, Portland, Oregon 97201-3980; and *Department of Medicine, The Ohio State University, Columbus, Ohio 43210-1870 Submitted
for publication
March 23, 1990
titative bioassay of angiogenic activity [6]. This assay has proven to be quick and inexpensive and permits continuous, serial observations of changes in blood vessel growth. The rapid vascular growth of the CAM makes it an ideal model for observing modulation of angiogenesis, since sustained release implants containing specific test substances can be placed on this growing capillary network. The CAM model has recently been used to demonstrate the angiogenic properties of the AIDS-KS cells [7]. Crum et al., using the CAM model, reported on a class of steroids which may have angiostatic properties when coadministered with heparin [8]. Glaser et al. have found that cultured human retinal pigment epithelial cells release a substance which, when applied topically to the chick embryonic yolk sac, causes the regression of new blood vessels and also inhibits proliferation of bovine endothelial cells in vitro [9]. Fassler et al. have presented preliminary data demonstrating octreotide acetate (SMS 201-995, Sandostatin, Sandoz Corporation, East Hanover, NJ (SMS)) when combined with endothelial cell growth factor induced inhibition of blood vessel growth [lo]. Another somatostatin analogue, RC-160, developed by Schally et al. has also been associated with high levels of antitumor activity [ll]. Based upon these observations, we hypothesized that somatostatin analogues may inhibit angiogenesis.
The mechanism responsible for alterations in tumor growth following administration of somatostatin analogues is unknown. Somatostatin analogues, SMS 201995 and RC-160, have demonstrated the potential to inhibit both tumor growth and vascularity, in viva and in vitro. We hypothesized that SMS and RC- 160 inhibit angiogenesis and this inhibition may alter tumor growth. To test this hypothesis, 2 mm methylcellulose disks containing concentrations of SMS 201-995 and RC-160 at 0, 0.5, 2.5, or 50 pg per disk, were implanted on the chorioallantoic membrane (CAM) of 6to 7-day-old shell-less chick embryos. Inhibition of blood vessel growth in the region of the disk was visually assessed 24-36 hr following disk implantation and graded (O-4) based on the radius of the zone of inhibition from the center of the disk. The overall incidence of inhibition for the somatostatin analogues at concentrations of 0.5,2.5, and 50 bug per disk was 13,56, and 61% for SMS and 27,49, and 68% for RC-160, respectively. Overall incidence of inhibition for the positive (inhibitory) control was 70.5% and those for buffer (negative) controls were 3-14%. Somatostatin analogues were associated in a dose-related fashion with both a greater percentage of inhibition of blood vessel growth and an increased grade of inhibition. Inhibition of angiogenesis may be a mechanism responsible for the tumor regression observed in vivo following SMS or 0 1991 Academic Press, Inc. RC-160 therapy.
MATERIALS INTRODUCTION
AND METHODS
To test our hypothesis we examined the effect of two chemically similar somatostatin analogues on vessel growth using a modification of the technique described by Dunn et al. for shell-less chick embryo culture [12]. Fertilized chicken eggs were obtained from the Oregon State University, Department of Poultry Sciences. The eggs were scrubbed with a mild detergent, dipped in alcohol, air dried, and placed on an automatic turner in a humidified incubator at 375°C. On Day 2 or 3 of development, each chick embryo was removed from the shell,
Tumor-related angiogenesis is a well-recognized, although not well-understood phenomenon. In 1971, Folkman isolated a tumor extract which could induce vascular proliferation (tumor angiogenic factor, TAF) and, since that time, many angiogenic-inducing agents derived from malignant and nonmalignant tissues have been described [l-5]. The chorioallantoic membrane (CAM) of the chick embryo can be used as a semi-quan245
All
Copyright 0 1991 rights of reproduction
0022-4804/91$1.50 by Academic Press, Inc. in any form reserved.
246
JOURNAL
OF SURGICAL
RESEARCH:
placed in a plastic wrap hammock (Safeway Stores, Inc., Oakland, CA), and returned to the incubator. Octreotide acetate (SMS 201-995, Sandoz, East Hanover, NJ) and RC-160 were tested for their capacity to inhibit angiogenesis in three concentrations: 0.52.5, and 50 pg disks. Hydrocortisone 21-phosphate (Sigma) 45 pg/lO ~1 with heparin (Sigma, St. Louis, MO) 37.5 pg/lO ~1 as modified from Crum et al. [8] was positive (inhibitory) control. Negative (no inhibition) controls were based on the buffer in which the drug was dissolved. These included octreotide acetate buffer or lyophilized octreotide acetate buffer (lyophilization was used to concentrate SMS to 50 pg/disk). Equal volumes of the somatostatin analogue and methylcellulose (Fisher Scientific, Kent, WA) were combined and 10 ~1 of the resultant suspension was layered on a Teflon rod, 2 mm in diameter, and allowed to air dry according to procedure used by Crum et al. [8]. The final drug concentration for each 10 ~1 disk was 0.5% methylcellulose and 0.5, 2.5, or 50 pg of SMS 201995 or RC-160. SMS 201-995 concentrations were chosen to represent a range that would exceed the tumor drug levels previously achieved following 50 pg subcutaneous injections in humans [ 131. We and others have clinically used up to 2000 pg per day of SMS with few side effects. Plasma levels of SMS have been demonstrated as high as 1.5 X lo6 pg/ml. Octreotide acetate levels in human tumors have been shown as high as 173 nanograms per gram wet tumor weight several hours following the administration of a 50-pg dose of Sandostatin subcutaneously [ 131. Doses of 500 pg subcutaneously 3-4 times per day and intravenous infusions of 50 pg of SMS per hour are common doses of this drug. We would assume that these doses would be associated with tissue concentrations in the range chosen for SMS or RC-160 test concentrations. Each disk was placed on the outer third of the CAM of a 6- to 7-day-old developing shell-less chick embryo and returned to the incubator. The radius of the zone of inhibition of blood vessel growth was visually assessed from the center of each disk to the furthest contiguous area in which tertiary blood vessels were absent. In some cases, visualization and subsequent photographic documentation of the small vessels were enhanced by the use of intravascular India ink injected into a vitelline vein using a 30-gauge needle (Fig. 1). A minimum of 30 evaluable eggs per concentration was tested. Embryos that did not survive (for any reason) or embryos in which disks could not be located were classified as nonevaluable and not included in the results. This represented approximately 20% of each set of eggs implanted and the incidence was equally distributed between control and the test substance embryos. Data were analyzed using analysis of variance techniques (Table 1). Secondary statistical analysis included Bonferroni t tests and two-tailed t tests (Table 2).
VOL.
50, NO. 3, MARCH
1991
RESULTS Chick embryo survival until disk implantation was approximately 67% with a 14% loss during removal from the shell and a subsequent loss of 19% due to either infection or abnormal differentiation. At 24-36 hr the positive control (hydrocortisone al-phosphate at 45 pg/disk combined with heparin, 37.5 pg/disk) inhibited angiogenesis (loss of tertiary and quaternary vessels) in 70.5% of the evaluable embryos (Fig. 2). Preliminary studies carried out for up to 96 hr demonstrate stable zones of angiogenesis inhibition after 24 hr of incubation. Increases in the concentration of either of the two somatostatin analogues were associated with an increase in the overall percentage inhibition of angiogenesis and an increase in the mean grade of zone of inhibition (Fig. 2, Table 1). SMS at 0.5,2.5, and 50 pg per disk was associated with 13, 56, and 61% overall incidence of inhibition, respectively (Fig. 2). The mean grades of the zone of inhibition for SMS disk concentrations of 0.5, 2.5, and 50 pg were 0.19 f 0.54, 0.68 rt 0.68, and 1.16 4 1.07, respectively (Table 1). Comparisons of the SMS associated mean grades of the zone of inhibition showed statistically significant inhibition between each concentration, P < 0.05 (Table 2a). This was performed first using analysis of variance on the three drug concentrations (P < 0.001) and then using Bonferroni t tests on each pair (P -c 0.05) as shown (Table 2a). RC-160 at 0.5, 2.5, and 50 pg per disk was associated with 27, 49, and 68% inhibition, respectively (Fig. 2). The mean grades of the zones of inhibition for RC-160 at disk concentrations of 0.5, 2.5, and 50 pg were 0.33 f 0.60, 0.57 f 0.65, and 1.21 f 0.91, respectively (Table 1). Comparisons of the RC-160 mean grades of the zone of inhibition showed statistically significant inhibition between the concentrations P < 0.05 of 50 and 0.5 pg per disk as well as 50 and 2.5 gg per disk (Table 2b). An analysis of variance was first performed on the three drug concentrations (P < 0.001). Bonferroni t tests were then performed between each pair of concentrations, with significance and P values as stated in Table 2b. Overall incidence of inhibition for the positive (inhibitory) control was 70.5% (Fig. 2) and the mean grade of the zone of inhibition was 1.34 * 1.18 (Table 1). Each concentration of somatostatin analogue was compared to the positive controls (Table 2~). An analysis of variance on the different concentrations and positive control was performed (P < O.l), followed by Dunnett’s test to determine post hoc significance. Significance and P values are stated the Table 2c. There was no statistical difference in inhibition of angiogenesis between the highest concentration of somatostatin analogue (50 cLg/ disk) and positive control. Overall incidences of inhibition for the negative buffer controls were 3.3% for the octreotide buffer and 13.9%
WOLTERLNG
ET AL.: SOMATOSTATIN
ANALOGUES
INHIBIT
ANGIOGENESIS
247
248
JOURNAL
TABLE Mean Group Lyophilized buffer Buffer Control Positive control SMS 0.5 pg SMS 2.5 pg SMS 50 cLg RC-160 0.5 /lg RC-160 2.5 pg RC-160 50 /.lg
Grade
OF SURGICAL
RESEARCH:
VOL.
50, NO. 3, MARCH
1
1991
TABLE
of the Zone of Inhibition
N
Mean*
SD**
SEM***
Variance
36 30
0.14 0.03
0.35 0.18
0.06 0.03
0.120 0.032
44 32 34 31 33 37 34
1.34 0.19 0.68 1.16 0.33 0.57 1.21
1.18 0.54 0.68 1.07 0.60 0.65 0.91
0.18 0.09 0.12 0.19 0.10 0.11 0.16
1.36 0.28 0.45 1.10 0.343 0.408 0.811
Note. Comparison of the mean grades of inhibition analogues and their respective buffers. * Mean grade of inhibition. ** Standard deviation. *** Standard error of the mean.
induced by SRIF
for the lyophilized octreotide buffer (Fig. 2). The mean grade of the zone of inhibition for buffer control was 0.03 + 0.18 and 0.14 + 0.35 for the lyophilized buffer (Table 1). Negative controls used for comparison with each analogue concentration were chosen according to the buffer used for each analogue/disk combination. Both somatostatin analogues in the two higher concentrations, 50 and 2.5 pg/disk, were statistically different from their appropriate buffers when compared using the two-tailed t test with Bonferroni adjustment, P < 0.001 (Table 2d). Increasing concentrations of both somatostatin analogues were associated with a greater percentage of inhibition in blood vessel growth as well as an increased grade of inhibition in a dose-related fashion (Tables 1 and 2). DISCUSSION
Somatostatin (somatotropin-release inhibiting factor (SRIF)), initially isolated from bovine hypothalamus, is a 14 residue peptide, which inhibits growth hormone release and modulates a variety of gut peptide levels and physiologic fluid secretion [ 141. Octreotide acetate, a more stable, more potent, and longer acting somatostatin analogue, has been used clinically for symptomatic control of functional endocrine tumors [ 13-171. Occasionally, tumor regression has been observed during octreotide treatment. A 70-year-old woman, treated in our institution, with MEN-I syndrome and gastrinoma metastatic to the liver showed complete tumor regression 4 months after beginning SMS therapy for control of ulcer symptoms. This tumor regression persisted for 10 months before tumor progression resumed [15]. Currently, several groups are examining the effects of SMS on human tumor growth in animals. Recently, Al-
2
Statistical Relationships between Different Concentrations of SMS and between Somatostatin Analogues and the Appropriate Controls Using the Mean Zone of Inhibition Concentration
Differs
from
(a) SMS 50 rg SMS 2.5 pg SMS 0.5 pg
SMS 2.5 pg SMS 0.5 pg SMS 50 pg
Significance
(P value)
Yes; P < 0.05 Yes; P -c 0.05 Yes; P < 0.05
These data were analyzed first using an analysis of variance on the (P < 0.001). Next Bonferroni t tests were three SMS concentrations performed on each pair of concentrations, with significance and P values as listed above. (b) RC-160 50 pg RC-160 2.5 yg RC-160 0.5 pg
RC-160 2.5 /.qg RC-160 0.5 /lg RC-160 50 pg
Yes; P < 0.05 No
Yes;P < 0.05
These data were analyzed first using an analysis of variance on the (P i 0.001). Next Bonferroni t tests three RC-160 concentrations were performed on each pair of concentrations, with significance and P values as listed above. (c) SMS 50 /Ig SMS 2.5 pg SMS 0.5 pg RC-160 50 jig RC-160 2.5 pg RC-160 0.5 /.lg
Positive Positive Positive Positive Positive Positive
control control control control control control
No Yes; P < 0.01
Yes;P < 0.01 No
Yes;P i 0.01 Yes; P < 0.01
These data were analyzed first using an analysis of variance on the six concentrations of analogue and positive control (P -e 0.001). Dunnett’s test was then used to determine post hoc significance. Significance and P values as listed above. (d) SMS 50 pg SMS 2.5 pg SMS 0.5 pg RC-160 50 pg RC-160 2.5 pg RC-160 0.5 pg
Lyophilized buffer control Buffer control Buffer control Buffer control Buffer control Buffer control
Yes; P -c 0.001 Yes; P i 0.001 No; P < 0.128 Yes; P < 0.001 Yes; P -c 0.001 No; P < 0.011
These data were analyzed using two-tailed t tests, with the above P values obtained. The Bonferroni adjustment for multiple comparisons was then used to determine significance, which states the calculated P value needs to be to.009 to be significant, since we have performed six comparisons. Significance is accordingly, listed above. Yes, difference is statistically significant. No, difference is not statistically significant. Note. Statistical comparison of inhibition of angiogenesis induced by SMS 201-995 or RC-160 vs their respective buffers and the positive control. Varying concentrations of SMS 201-995 were compared to one another. (a) SMS 201-995 concentrations were internally compared to each other using ANOVA, with Bonferroni post hoc comparisons. (b) RC-160 concentrations were internally compared to each other using ANOVA, with Bonferroni post hoc comparison. (c) Positive controls (heparin-steroid combinations) were compared to SMS 201-995 and RC-160 using analysis of variance (ANOVA) and post hoc comparisons by Dunnett’s test. (d) SMS 201-995 or RC-160 were compared to control buffers using two-tailed t tests.
WOLTERING ANGIOGENESIS:
INCIDENCE
Test Substance
N
lnhiition %
%sitive
44
70.5
30
3.3
36
13.9
suffer
Control l
*
ET AL.:
SOMATOSTATIN
OF INHIBITION
96.7
+2 29.5 3.3
0
13 9
0
Lyophilized Buffer
l
**
SMS
5ug
86
1
32
12.5
6.3
34
55 9
12 5
5oAlg
31
61.3
35.5
RC160
.5ug
33
27.3
61
RC160
2 5ug
37
48.6
8.1
RC160
5O*rg
34
67.6
52.9
SMS
25wg
SMS
FIG.
2.
Data results expressed as percentages
inhibition.
len et al. showed that SMS combined with fluoromethylornithine inhibits the growth of human carcinoid tumor [17]; while Milhoan et al. implanted human small cell lung carcinoma cells and human carcinoid tumor cells subcutaneously into nude mice and observed that subsequent tumor growth was less in SMS-injected mice than in control (saline-injected) mice [18]. Weber et al. in a similar animal model, showed that the growth of MCF-7 hormone-responsive human breast carcinoma was inhibited by SMS treatment when compared to growth of tumors in the buffer-treated control mice [ 191. Additionally, these workers noted that tumors in SMStreated mice were grossly less vascular. In this study, we used the chick CAM assay system to study the inhibition of angiogenesis. The CAM functions as the major site for gas exchange during normal chick development. Using sustained release implants to study test compounds on this exposed growing capillary network has made the CAM model a primary system for detection of angiogenic modulators [l, 2, 4, 6-81. It is widely accepted that human and animal tumors can induce endothelial cell mitosis and angiogenesis, resulting in extensive neovascularization [20, 211. This angiogenic activity is critical for tumor implantation and subsequent growth. Using the rabbit cornea, it has been shown that in the absence of neovascularization, the size of tumor grafts becomes limited [21]. It has been proposed that neovascularization is essential for tumor growth, development, and metastasis; furthermore, when angiogenesis is absent, tumors tend to remain dormant [22]. Folkman’s group has described a class of steroids which inhibits angiogenesis in the presence of heparin or a fragment of heparin having no anticoagulant activity [8, 231. The only known biological function of this new class of steroids is the inhibition of angiogenesis, which
ANALOGUES
INHIBIT
ANGIOGENESIS
249
appears to be independent of the steroids’ glucocorticoid and mineralocorticoid activity [8,23]. In the presence of heparin, angiostatic steroids inhibit capillary growth by an unknown mechanism. A proposed mechanism is the loss of basement membrane scaffolding to which anchorage-dependent endothelial cells are normally attached. This mechanism may be related to the associated capillary regression when in the presence of angiostatic steroids [8]. Maione et al. have recently described angiogenesis inhibition with recombinant human platelet factor-4 and have demonstrated that heparin reverses this angiostatic activity [24]. In our experiments we were able to achieve inhibition of angiogenesis without a facilitator such as heparin, supporting the concept that SMS acts directly on the cell, possibly through a cell membrane receptor. The concept of receptor-specific SMS binding is supported by the work of Reubi et al. [25]. Both SMS 201-995 and RC-160 have been shown to inhibit tumor development [14-16, 261. Another somatostatin analogue, MDK-678, has been found to inhibit the growth of human small cell lung carcinoma in nude mice [ 181. Schally et al. have synthesized several somatostatin analogues that have high levels of antitumor activity [26]. These investigators have suggested that somatostatin analogues be considered as adjuncts for tumor therapy in selected cancer patients [ll]. Using Syrian golden hamsters, Schally et al. have shown that combinations of the somatostatin analogue RC-160 and luteinizing hormone-releasing hormone (LH-RH) will inhibit pancreatic exocrine cancer growth and result in prolonged animal survival [ 111. Combinations of somatostatin analogue and LH-RH appear to be more efficacious than either substance administered alone [ll]. We have shown that somatostatin analogues inhibit angiogenesis of normal vascular development in the chick chorioallantoic membrane in a dose-related fashion (Fig. 2). There was no significant difference in the incidence of inhibition of angiogenesis between the highest concentration of SMS (50 pg) or RC-160 (50 pg) used and the positive control, hydrocortisone-21-phosphate (Table 2). These results may be clinically significant, since Folkman et al. have suggested the use of the heparin antiangiogenic steroid combinations as future chemotherapeutic agents [8, 231 or, alternatively, biologic response modifiers. The SRIF-like compounds RC-160 and SMS 201-995 are comparable to the steroidheparin combinations of Folkman in their ability to decrease angiogenesis. The structural similarities of RC160 [D-Phe-Cys-Tyr-D-Trp-Lys-Val-Cys-Trp-NH,] and SMS 201-995 [D-Phe-Cys-Phe-D-Trp-LysThr-Cys-Thr-ol] suggest that inhibition of angiogenesis may be a characteristic of a class of somatostatin analogues. This concept requires, however, validation with larger numbers of related compounds. These investigations are currently underway in our laboratory.
250
JOURNAL
OF SURGICAL
RESEARCH:
It is felt that neovascularization plays a critical role in the progression of several important disease states including diabetic retinopathy, retinopathy of prematurity, choroidal neovascularization, sickle cell retinopathy, arthritis, and tumor growth [ 1,9]. Diffusible factors released from growing tumors have been isolated and characterized and are believed to play a significant role in neovascularization [22]. One such factor is TAF, which seems to be related to aggressive tumor growth in uiuo [l]. TAF from rat Walker tumors has been purified and is an easily dialyzable compound of 200-300 Da [l, 201. This molecule has been found to be a very potent stimulator of angiogenesis (even in trace amounts) in the CAM model [l, 21,271. TAF has also been shown to enhance endothelial cell proliferation in vitro [ 11. A similar capillary-related factor, endothelial cell-stimulating angiogenesis factor (ESAF) thought to be produced in rheumatoid arthritis and other arthritides with inflammation, has been isolated from synovial fluids of patients with joint diseases [l]. There is strong evidence that ESAF and TAF may have similar structures and, in fact, may be identical [l]. As noted by Brown et cd., vascular proliferation is a feature of the pathology of these disorders with occasional formation of a pannus, which can lead to joint erosions and effusions [ 11. They also noted that capillaries of the rheumatoid pannus penetrate the articular cartilage in the early stages of destruction [l]. ESAF has been isolated from both tumor and nontumor sources. The cellular source of angiogenesis from the joint is not known, although lymphocytes and macrophage have both been suspected to produce angiogenesis factors, and both cell types are isolated from diseased synovial membranes [ 11. It seems that a wide spectrum of normal cells or tissues can induce angiogenesis. When tested in the CAM assay, mouse lymph node and spleen cells induce angiogenesis; also, mouse lymph nodes and allogeneic rabbit lymph nodes, when grafted intracorneally into adult rabbits [22], cause a neovascularization reaction. Inhibitors of neovascularization have been isolated from avascular tissues such as cartilage, vitreous humor, and the lens; also an inhibitor has been isolated from cell cultures of human retinal pigment epithelial cells, as reported by Glaser et al. [9]. In our study, somatostatin analogues have been shown to inhibit neovascularization in the normally developing CAM. To our knowledge, this is the first description of somatostatin analogues directly and quantitatively inhibiting angiogenesis. This observation could be responsible for the alterations in tumor growth previously seen in vitro and in uiuo. Optimization of drug concentration and drug exposure time will be necessary to determine if our results are clinically relevant. ACKNOWLEDGMENTS We thank and appreciate the technical and scientific assistance of G. H. Davis. We acknowledge and thank G. Sexton for his help with
VOL.
50, NO. 3, MARCH
1991
the statistical analysis. We also acknowledge and thank A. Schally of the Veterans Administration Medical Center, New Orleans, Louisiana for the kind gift of the sandostatin analogue RC-160. This research was supported, in part, by Sandoz Inc., East Hanover, NJ.
REFERENCES 1.
2. 3.
4.
5.
6. 7.
8.
9.
10.
11.
12.
13.
Brown, R. A., Weiss, K. B., Tomlinson, I. W., Phillips, P., and Kumar, S. Angiogenic factor from synovial fluid resembling that from tumors. Lancet 1: 682, 1980. Auerbach, R. Angiogenesis-inducing factors: A review. Lymphokines 4: 69, 1981. Gospodarowiez, D., Bialecki, H., and Thakral, T. K. The angiogenie activity of the fibroblast and epidermal growth factor. Exp. Eye Res. 28: 501, 1979. Shing, Y., Folkman, J., Haudenschild, C., Lund, D., Crum, R., and Klagsburn, M. Angiogenesis is stimulated by a tumor derived endothelial cell growth factor. J. Cell Biochem. 29: 275, 1985. Phillips, P., and Kumar, S. Tumor angiogenesis factor and its neutralization by a xenogeneic antiserum. Znt. J. Cancer 23: 82, 1979. Folkman, J. Angiogenesis: Initiation and control. Ann. N.Y. Acad. Sci. 401: 212, 1982. Salahuddin, S. Z., Nakamura, S., Biberfield, P., Kaplan, M. H., Markham, P. D., Larsson, L., and Gallo, R. C. Angiogenic properties of Kaposi’s sarcoma-derived cells after long-term culture in uitro. Science 242: 430, 1988. Crum, R., Szabo, S., and Folkman, J. A new class of steroids inhibits angiogenesis in the presence of heparin or a heparin fragment. Science 230: 1375,1985. Glaser, B., Ckampochiaro, P., Davis, J., and Jerdan, J. Retinal pigment epithelial cells release inhibitors of neovascularization. Ophthalmology 94: 780,1983. Fassler, J. E., Hughes, J. H., Cataland, S., and O’Dorisio, T. M. Somatostatin analogue: An inhibitor of angiogenesis? Biomed. Res. (Suppl.) 1: 81, 1988. Zalatnai, A., and Schally, A. V. Treatment of N-nitrosobis (2-0xopropyl) amine-induced pancreatic cancer in Syrian golden hamsters with n-Trp-6-LH-RH and somatostatin analogue RC-160 microcapsules. Cancer Res. 49: 1810, 1989. Dunn, B. E., Fitzharris, T. P., and Barnett, B. D. Effects of varying chamber construction and embryo pre-incubation age on survival and growth of chick embryos in shell-less culture. Anat. Rec. 199: 33, 1981. Woltering, E. W., Ellison, E. C., O’Dorisio, T. M., Sparks, J., Howe, B., and Dyben, T. Somatostatin-like peptides alter calcium but not secretin sensitivity of gastrinoma cells. J. Surg. Res.
40:605,1986. 14. Mozell, E., Woltering,
E. A., O’Dorisio, T. M., Fletcher, W. S., Sinclair, A., and Hill, D. Effect of somatostatin analogue on peptide release and tumor growth in the Zollinger-Ellison syndrome. Surg. Gyn. Obstet. 170: 476, 1990. 15. Woltering, E. A., Mozell, E., O’Dorisio, T., Fletcher, W., and Howe, B. Suppression of primary and secondary peptides with somatostatin analogue in the therapy of functional endocrine tumors. Surg. Gyn. Obstet. 167: 453, 1988. 16. Kvols, L. K., Moertel, C. G., O’Connell, M. J., Schutt, A. J., Rubin, J., and Hahn, R. G. Treatment of the malignant carcinoid syndrome: Evaluation of a long-acting somatostatin analogue. N. Engl. J. Med. 315: 663, 1986. 17. Allen, E., Evers, M., Townsend, J. C., Uchida, T., and Thompson, J. Somatostatin Analogue (201-995) and a-difluoromethylornithine inhibit growth of human carcinoid tumor. Surg. Forum
40:415,1989. 18.
Milhoan,
R., Trudel,
J., Lawrence,
J., Townsend,
C., and
WOLTERING
ET AL.:
SOMATOSTATIN
Thompson, J. Somatostatin inhibits growth of human small cell lung carcinoma in vivo. Surg. Forum 39: 438, 1988. 19.
20.
21.
22.
Weber, C., Meram, L., Koschitzky, T., and Logerfo, P. Inhibition of growth of MCF-7 human breast carcinoma in uivo by somatostatin analogue SMS 201-995. Surg. Forum 39: 66, 1988. Weber, C., Meriam, L., Koschitzky, T., Karp, F., Benson, M., Kenneth, F., and Logerfo, P. Inhibition of growth of human breast carcinomas in viuo by somatostatin analogue SMS 201995: Treatment of nude mouse xenografts. Surgery 106: 416, 1989. Phillips, P., Steward, J., and Kumar, S. Tumor angiogenesis factor (TAF) in human and animal tumors. Znt. J. Cancer 17: 549, 1976. Folkman, J., and Cotran, R. Relation of vascular proliferation to tumor growth. Znt. Rev. Exp. Pathol. 16: 207, 1976.
ANALOGUES 23.
24.
25. 26.
27.
INHIBIT
ANGIOGENESIS
251
Ingber, D. E., Madri, J. A., and Folkman, J. A possible mechanism for inhibition of angiogenesis by angiostatic steroids: Induction of capillary basement membrane dissolution. Endocrinology 119:33,1981. Reubi, J., and Torhorst, J. The relationship between somatostatin, epidermal growth factor, and steroid hormone receptors in breast cancer. Cancer 64: 1254,1989. Schally, A. V. Oncological applicat,ions of somatostatin analogues. Cancer Res. 48: 6977, 1989. Schor, A., Schor, S., and Kumar, S. Importance of a collagen substratum for stimulation of capillary endothelial cell proliferation by tumor angiogenesis factor. Znt. J. Cancer 24: 225, 1979. Auerbach, R., Kubai, L., and Skidky, Y. Angiogenesis induction by tumors, embryonic tissues and lymphocytes. Cancer Res. 36:
3435,1976.
OF SURGICAL
RESEARCH
60,245-251
(1991)
Somatostatin Analogues Inhibit Angiogenesis in the Chick Chorioallantoic Membrane EUGENE A. WOLTERING, M.D., ROSEMARY BARRIE, M.S., THOMAS M. O’DORISIO, M.D.,* DENNIS ARCE, M.D., TINA URE, M.D., ANDREW CRAMER, M.D., DOUGLAS HOLMES, M.D.,? JOSEPH ROBERTSON, M.D.,? AND JOHN FASSLER, M.D.* Oregon Health Sciences University, Departments of Surgery and tOphthalmology, Portland, Oregon 97201-3980; and *Department of Medicine, The Ohio State University, Columbus, Ohio 43210-1870 Submitted
for publication
March 23, 1990
titative bioassay of angiogenic activity [6]. This assay has proven to be quick and inexpensive and permits continuous, serial observations of changes in blood vessel growth. The rapid vascular growth of the CAM makes it an ideal model for observing modulation of angiogenesis, since sustained release implants containing specific test substances can be placed on this growing capillary network. The CAM model has recently been used to demonstrate the angiogenic properties of the AIDS-KS cells [7]. Crum et al., using the CAM model, reported on a class of steroids which may have angiostatic properties when coadministered with heparin [8]. Glaser et al. have found that cultured human retinal pigment epithelial cells release a substance which, when applied topically to the chick embryonic yolk sac, causes the regression of new blood vessels and also inhibits proliferation of bovine endothelial cells in vitro [9]. Fassler et al. have presented preliminary data demonstrating octreotide acetate (SMS 201-995, Sandostatin, Sandoz Corporation, East Hanover, NJ (SMS)) when combined with endothelial cell growth factor induced inhibition of blood vessel growth [lo]. Another somatostatin analogue, RC-160, developed by Schally et al. has also been associated with high levels of antitumor activity [ll]. Based upon these observations, we hypothesized that somatostatin analogues may inhibit angiogenesis.
The mechanism responsible for alterations in tumor growth following administration of somatostatin analogues is unknown. Somatostatin analogues, SMS 201995 and RC-160, have demonstrated the potential to inhibit both tumor growth and vascularity, in viva and in vitro. We hypothesized that SMS and RC- 160 inhibit angiogenesis and this inhibition may alter tumor growth. To test this hypothesis, 2 mm methylcellulose disks containing concentrations of SMS 201-995 and RC-160 at 0, 0.5, 2.5, or 50 pg per disk, were implanted on the chorioallantoic membrane (CAM) of 6to 7-day-old shell-less chick embryos. Inhibition of blood vessel growth in the region of the disk was visually assessed 24-36 hr following disk implantation and graded (O-4) based on the radius of the zone of inhibition from the center of the disk. The overall incidence of inhibition for the somatostatin analogues at concentrations of 0.5,2.5, and 50 bug per disk was 13,56, and 61% for SMS and 27,49, and 68% for RC-160, respectively. Overall incidence of inhibition for the positive (inhibitory) control was 70.5% and those for buffer (negative) controls were 3-14%. Somatostatin analogues were associated in a dose-related fashion with both a greater percentage of inhibition of blood vessel growth and an increased grade of inhibition. Inhibition of angiogenesis may be a mechanism responsible for the tumor regression observed in vivo following SMS or 0 1991 Academic Press, Inc. RC-160 therapy.
MATERIALS INTRODUCTION
AND METHODS
To test our hypothesis we examined the effect of two chemically similar somatostatin analogues on vessel growth using a modification of the technique described by Dunn et al. for shell-less chick embryo culture [12]. Fertilized chicken eggs were obtained from the Oregon State University, Department of Poultry Sciences. The eggs were scrubbed with a mild detergent, dipped in alcohol, air dried, and placed on an automatic turner in a humidified incubator at 375°C. On Day 2 or 3 of development, each chick embryo was removed from the shell,
Tumor-related angiogenesis is a well-recognized, although not well-understood phenomenon. In 1971, Folkman isolated a tumor extract which could induce vascular proliferation (tumor angiogenic factor, TAF) and, since that time, many angiogenic-inducing agents derived from malignant and nonmalignant tissues have been described [l-5]. The chorioallantoic membrane (CAM) of the chick embryo can be used as a semi-quan245
All
Copyright 0 1991 rights of reproduction
0022-4804/91$1.50 by Academic Press, Inc. in any form reserved.
246
JOURNAL
OF SURGICAL
RESEARCH:
placed in a plastic wrap hammock (Safeway Stores, Inc., Oakland, CA), and returned to the incubator. Octreotide acetate (SMS 201-995, Sandoz, East Hanover, NJ) and RC-160 were tested for their capacity to inhibit angiogenesis in three concentrations: 0.52.5, and 50 pg disks. Hydrocortisone 21-phosphate (Sigma) 45 pg/lO ~1 with heparin (Sigma, St. Louis, MO) 37.5 pg/lO ~1 as modified from Crum et al. [8] was positive (inhibitory) control. Negative (no inhibition) controls were based on the buffer in which the drug was dissolved. These included octreotide acetate buffer or lyophilized octreotide acetate buffer (lyophilization was used to concentrate SMS to 50 pg/disk). Equal volumes of the somatostatin analogue and methylcellulose (Fisher Scientific, Kent, WA) were combined and 10 ~1 of the resultant suspension was layered on a Teflon rod, 2 mm in diameter, and allowed to air dry according to procedure used by Crum et al. [8]. The final drug concentration for each 10 ~1 disk was 0.5% methylcellulose and 0.5, 2.5, or 50 pg of SMS 201995 or RC-160. SMS 201-995 concentrations were chosen to represent a range that would exceed the tumor drug levels previously achieved following 50 pg subcutaneous injections in humans [ 131. We and others have clinically used up to 2000 pg per day of SMS with few side effects. Plasma levels of SMS have been demonstrated as high as 1.5 X lo6 pg/ml. Octreotide acetate levels in human tumors have been shown as high as 173 nanograms per gram wet tumor weight several hours following the administration of a 50-pg dose of Sandostatin subcutaneously [ 131. Doses of 500 pg subcutaneously 3-4 times per day and intravenous infusions of 50 pg of SMS per hour are common doses of this drug. We would assume that these doses would be associated with tissue concentrations in the range chosen for SMS or RC-160 test concentrations. Each disk was placed on the outer third of the CAM of a 6- to 7-day-old developing shell-less chick embryo and returned to the incubator. The radius of the zone of inhibition of blood vessel growth was visually assessed from the center of each disk to the furthest contiguous area in which tertiary blood vessels were absent. In some cases, visualization and subsequent photographic documentation of the small vessels were enhanced by the use of intravascular India ink injected into a vitelline vein using a 30-gauge needle (Fig. 1). A minimum of 30 evaluable eggs per concentration was tested. Embryos that did not survive (for any reason) or embryos in which disks could not be located were classified as nonevaluable and not included in the results. This represented approximately 20% of each set of eggs implanted and the incidence was equally distributed between control and the test substance embryos. Data were analyzed using analysis of variance techniques (Table 1). Secondary statistical analysis included Bonferroni t tests and two-tailed t tests (Table 2).
VOL.
50, NO. 3, MARCH
1991
RESULTS Chick embryo survival until disk implantation was approximately 67% with a 14% loss during removal from the shell and a subsequent loss of 19% due to either infection or abnormal differentiation. At 24-36 hr the positive control (hydrocortisone al-phosphate at 45 pg/disk combined with heparin, 37.5 pg/disk) inhibited angiogenesis (loss of tertiary and quaternary vessels) in 70.5% of the evaluable embryos (Fig. 2). Preliminary studies carried out for up to 96 hr demonstrate stable zones of angiogenesis inhibition after 24 hr of incubation. Increases in the concentration of either of the two somatostatin analogues were associated with an increase in the overall percentage inhibition of angiogenesis and an increase in the mean grade of zone of inhibition (Fig. 2, Table 1). SMS at 0.5,2.5, and 50 pg per disk was associated with 13, 56, and 61% overall incidence of inhibition, respectively (Fig. 2). The mean grades of the zone of inhibition for SMS disk concentrations of 0.5, 2.5, and 50 pg were 0.19 f 0.54, 0.68 rt 0.68, and 1.16 4 1.07, respectively (Table 1). Comparisons of the SMS associated mean grades of the zone of inhibition showed statistically significant inhibition between each concentration, P < 0.05 (Table 2a). This was performed first using analysis of variance on the three drug concentrations (P < 0.001) and then using Bonferroni t tests on each pair (P -c 0.05) as shown (Table 2a). RC-160 at 0.5, 2.5, and 50 pg per disk was associated with 27, 49, and 68% inhibition, respectively (Fig. 2). The mean grades of the zones of inhibition for RC-160 at disk concentrations of 0.5, 2.5, and 50 pg were 0.33 f 0.60, 0.57 f 0.65, and 1.21 f 0.91, respectively (Table 1). Comparisons of the RC-160 mean grades of the zone of inhibition showed statistically significant inhibition between the concentrations P < 0.05 of 50 and 0.5 pg per disk as well as 50 and 2.5 gg per disk (Table 2b). An analysis of variance was first performed on the three drug concentrations (P < 0.001). Bonferroni t tests were then performed between each pair of concentrations, with significance and P values as stated in Table 2b. Overall incidence of inhibition for the positive (inhibitory) control was 70.5% (Fig. 2) and the mean grade of the zone of inhibition was 1.34 * 1.18 (Table 1). Each concentration of somatostatin analogue was compared to the positive controls (Table 2~). An analysis of variance on the different concentrations and positive control was performed (P < O.l), followed by Dunnett’s test to determine post hoc significance. Significance and P values are stated the Table 2c. There was no statistical difference in inhibition of angiogenesis between the highest concentration of somatostatin analogue (50 cLg/ disk) and positive control. Overall incidences of inhibition for the negative buffer controls were 3.3% for the octreotide buffer and 13.9%
WOLTERLNG
ET AL.: SOMATOSTATIN
ANALOGUES
INHIBIT
ANGIOGENESIS
247
248
JOURNAL
TABLE Mean Group Lyophilized buffer Buffer Control Positive control SMS 0.5 pg SMS 2.5 pg SMS 50 cLg RC-160 0.5 /lg RC-160 2.5 pg RC-160 50 /.lg
Grade
OF SURGICAL
RESEARCH:
VOL.
50, NO. 3, MARCH
1
1991
TABLE
of the Zone of Inhibition
N
Mean*
SD**
SEM***
Variance
36 30
0.14 0.03
0.35 0.18
0.06 0.03
0.120 0.032
44 32 34 31 33 37 34
1.34 0.19 0.68 1.16 0.33 0.57 1.21
1.18 0.54 0.68 1.07 0.60 0.65 0.91
0.18 0.09 0.12 0.19 0.10 0.11 0.16
1.36 0.28 0.45 1.10 0.343 0.408 0.811
Note. Comparison of the mean grades of inhibition analogues and their respective buffers. * Mean grade of inhibition. ** Standard deviation. *** Standard error of the mean.
induced by SRIF
for the lyophilized octreotide buffer (Fig. 2). The mean grade of the zone of inhibition for buffer control was 0.03 + 0.18 and 0.14 + 0.35 for the lyophilized buffer (Table 1). Negative controls used for comparison with each analogue concentration were chosen according to the buffer used for each analogue/disk combination. Both somatostatin analogues in the two higher concentrations, 50 and 2.5 pg/disk, were statistically different from their appropriate buffers when compared using the two-tailed t test with Bonferroni adjustment, P < 0.001 (Table 2d). Increasing concentrations of both somatostatin analogues were associated with a greater percentage of inhibition in blood vessel growth as well as an increased grade of inhibition in a dose-related fashion (Tables 1 and 2). DISCUSSION
Somatostatin (somatotropin-release inhibiting factor (SRIF)), initially isolated from bovine hypothalamus, is a 14 residue peptide, which inhibits growth hormone release and modulates a variety of gut peptide levels and physiologic fluid secretion [ 141. Octreotide acetate, a more stable, more potent, and longer acting somatostatin analogue, has been used clinically for symptomatic control of functional endocrine tumors [ 13-171. Occasionally, tumor regression has been observed during octreotide treatment. A 70-year-old woman, treated in our institution, with MEN-I syndrome and gastrinoma metastatic to the liver showed complete tumor regression 4 months after beginning SMS therapy for control of ulcer symptoms. This tumor regression persisted for 10 months before tumor progression resumed [15]. Currently, several groups are examining the effects of SMS on human tumor growth in animals. Recently, Al-
2
Statistical Relationships between Different Concentrations of SMS and between Somatostatin Analogues and the Appropriate Controls Using the Mean Zone of Inhibition Concentration
Differs
from
(a) SMS 50 rg SMS 2.5 pg SMS 0.5 pg
SMS 2.5 pg SMS 0.5 pg SMS 50 pg
Significance
(P value)
Yes; P < 0.05 Yes; P -c 0.05 Yes; P < 0.05
These data were analyzed first using an analysis of variance on the (P < 0.001). Next Bonferroni t tests were three SMS concentrations performed on each pair of concentrations, with significance and P values as listed above. (b) RC-160 50 pg RC-160 2.5 yg RC-160 0.5 pg
RC-160 2.5 /.qg RC-160 0.5 /lg RC-160 50 pg
Yes; P < 0.05 No
Yes;P < 0.05
These data were analyzed first using an analysis of variance on the (P i 0.001). Next Bonferroni t tests three RC-160 concentrations were performed on each pair of concentrations, with significance and P values as listed above. (c) SMS 50 /Ig SMS 2.5 pg SMS 0.5 pg RC-160 50 jig RC-160 2.5 pg RC-160 0.5 /.lg
Positive Positive Positive Positive Positive Positive
control control control control control control
No Yes; P < 0.01
Yes;P < 0.01 No
Yes;P i 0.01 Yes; P < 0.01
These data were analyzed first using an analysis of variance on the six concentrations of analogue and positive control (P -e 0.001). Dunnett’s test was then used to determine post hoc significance. Significance and P values as listed above. (d) SMS 50 pg SMS 2.5 pg SMS 0.5 pg RC-160 50 pg RC-160 2.5 pg RC-160 0.5 pg
Lyophilized buffer control Buffer control Buffer control Buffer control Buffer control Buffer control
Yes; P -c 0.001 Yes; P i 0.001 No; P < 0.128 Yes; P < 0.001 Yes; P -c 0.001 No; P < 0.011
These data were analyzed using two-tailed t tests, with the above P values obtained. The Bonferroni adjustment for multiple comparisons was then used to determine significance, which states the calculated P value needs to be to.009 to be significant, since we have performed six comparisons. Significance is accordingly, listed above. Yes, difference is statistically significant. No, difference is not statistically significant. Note. Statistical comparison of inhibition of angiogenesis induced by SMS 201-995 or RC-160 vs their respective buffers and the positive control. Varying concentrations of SMS 201-995 were compared to one another. (a) SMS 201-995 concentrations were internally compared to each other using ANOVA, with Bonferroni post hoc comparisons. (b) RC-160 concentrations were internally compared to each other using ANOVA, with Bonferroni post hoc comparison. (c) Positive controls (heparin-steroid combinations) were compared to SMS 201-995 and RC-160 using analysis of variance (ANOVA) and post hoc comparisons by Dunnett’s test. (d) SMS 201-995 or RC-160 were compared to control buffers using two-tailed t tests.
WOLTERING ANGIOGENESIS:
INCIDENCE
Test Substance
N
lnhiition %
%sitive
44
70.5
30
3.3
36
13.9
suffer
Control l
*
ET AL.:
SOMATOSTATIN
OF INHIBITION
96.7
+2 29.5 3.3
0
13 9
0
Lyophilized Buffer
l
**
SMS
5ug
86
1
32
12.5
6.3
34
55 9
12 5
5oAlg
31
61.3
35.5
RC160
.5ug
33
27.3
61
RC160
2 5ug
37
48.6
8.1
RC160
5O*rg
34
67.6
52.9
SMS
25wg
SMS
FIG.
2.
Data results expressed as percentages
inhibition.
len et al. showed that SMS combined with fluoromethylornithine inhibits the growth of human carcinoid tumor [17]; while Milhoan et al. implanted human small cell lung carcinoma cells and human carcinoid tumor cells subcutaneously into nude mice and observed that subsequent tumor growth was less in SMS-injected mice than in control (saline-injected) mice [18]. Weber et al. in a similar animal model, showed that the growth of MCF-7 hormone-responsive human breast carcinoma was inhibited by SMS treatment when compared to growth of tumors in the buffer-treated control mice [ 191. Additionally, these workers noted that tumors in SMStreated mice were grossly less vascular. In this study, we used the chick CAM assay system to study the inhibition of angiogenesis. The CAM functions as the major site for gas exchange during normal chick development. Using sustained release implants to study test compounds on this exposed growing capillary network has made the CAM model a primary system for detection of angiogenic modulators [l, 2, 4, 6-81. It is widely accepted that human and animal tumors can induce endothelial cell mitosis and angiogenesis, resulting in extensive neovascularization [20, 211. This angiogenic activity is critical for tumor implantation and subsequent growth. Using the rabbit cornea, it has been shown that in the absence of neovascularization, the size of tumor grafts becomes limited [21]. It has been proposed that neovascularization is essential for tumor growth, development, and metastasis; furthermore, when angiogenesis is absent, tumors tend to remain dormant [22]. Folkman’s group has described a class of steroids which inhibits angiogenesis in the presence of heparin or a fragment of heparin having no anticoagulant activity [8, 231. The only known biological function of this new class of steroids is the inhibition of angiogenesis, which
ANALOGUES
INHIBIT
ANGIOGENESIS
249
appears to be independent of the steroids’ glucocorticoid and mineralocorticoid activity [8,23]. In the presence of heparin, angiostatic steroids inhibit capillary growth by an unknown mechanism. A proposed mechanism is the loss of basement membrane scaffolding to which anchorage-dependent endothelial cells are normally attached. This mechanism may be related to the associated capillary regression when in the presence of angiostatic steroids [8]. Maione et al. have recently described angiogenesis inhibition with recombinant human platelet factor-4 and have demonstrated that heparin reverses this angiostatic activity [24]. In our experiments we were able to achieve inhibition of angiogenesis without a facilitator such as heparin, supporting the concept that SMS acts directly on the cell, possibly through a cell membrane receptor. The concept of receptor-specific SMS binding is supported by the work of Reubi et al. [25]. Both SMS 201-995 and RC-160 have been shown to inhibit tumor development [14-16, 261. Another somatostatin analogue, MDK-678, has been found to inhibit the growth of human small cell lung carcinoma in nude mice [ 181. Schally et al. have synthesized several somatostatin analogues that have high levels of antitumor activity [26]. These investigators have suggested that somatostatin analogues be considered as adjuncts for tumor therapy in selected cancer patients [ll]. Using Syrian golden hamsters, Schally et al. have shown that combinations of the somatostatin analogue RC-160 and luteinizing hormone-releasing hormone (LH-RH) will inhibit pancreatic exocrine cancer growth and result in prolonged animal survival [ 111. Combinations of somatostatin analogue and LH-RH appear to be more efficacious than either substance administered alone [ll]. We have shown that somatostatin analogues inhibit angiogenesis of normal vascular development in the chick chorioallantoic membrane in a dose-related fashion (Fig. 2). There was no significant difference in the incidence of inhibition of angiogenesis between the highest concentration of SMS (50 pg) or RC-160 (50 pg) used and the positive control, hydrocortisone-21-phosphate (Table 2). These results may be clinically significant, since Folkman et al. have suggested the use of the heparin antiangiogenic steroid combinations as future chemotherapeutic agents [8, 231 or, alternatively, biologic response modifiers. The SRIF-like compounds RC-160 and SMS 201-995 are comparable to the steroidheparin combinations of Folkman in their ability to decrease angiogenesis. The structural similarities of RC160 [D-Phe-Cys-Tyr-D-Trp-Lys-Val-Cys-Trp-NH,] and SMS 201-995 [D-Phe-Cys-Phe-D-Trp-LysThr-Cys-Thr-ol] suggest that inhibition of angiogenesis may be a characteristic of a class of somatostatin analogues. This concept requires, however, validation with larger numbers of related compounds. These investigations are currently underway in our laboratory.
250
JOURNAL
OF SURGICAL
RESEARCH:
It is felt that neovascularization plays a critical role in the progression of several important disease states including diabetic retinopathy, retinopathy of prematurity, choroidal neovascularization, sickle cell retinopathy, arthritis, and tumor growth [ 1,9]. Diffusible factors released from growing tumors have been isolated and characterized and are believed to play a significant role in neovascularization [22]. One such factor is TAF, which seems to be related to aggressive tumor growth in uiuo [l]. TAF from rat Walker tumors has been purified and is an easily dialyzable compound of 200-300 Da [l, 201. This molecule has been found to be a very potent stimulator of angiogenesis (even in trace amounts) in the CAM model [l, 21,271. TAF has also been shown to enhance endothelial cell proliferation in vitro [ 11. A similar capillary-related factor, endothelial cell-stimulating angiogenesis factor (ESAF) thought to be produced in rheumatoid arthritis and other arthritides with inflammation, has been isolated from synovial fluids of patients with joint diseases [l]. There is strong evidence that ESAF and TAF may have similar structures and, in fact, may be identical [l]. As noted by Brown et cd., vascular proliferation is a feature of the pathology of these disorders with occasional formation of a pannus, which can lead to joint erosions and effusions [ 11. They also noted that capillaries of the rheumatoid pannus penetrate the articular cartilage in the early stages of destruction [l]. ESAF has been isolated from both tumor and nontumor sources. The cellular source of angiogenesis from the joint is not known, although lymphocytes and macrophage have both been suspected to produce angiogenesis factors, and both cell types are isolated from diseased synovial membranes [ 11. It seems that a wide spectrum of normal cells or tissues can induce angiogenesis. When tested in the CAM assay, mouse lymph node and spleen cells induce angiogenesis; also, mouse lymph nodes and allogeneic rabbit lymph nodes, when grafted intracorneally into adult rabbits [22], cause a neovascularization reaction. Inhibitors of neovascularization have been isolated from avascular tissues such as cartilage, vitreous humor, and the lens; also an inhibitor has been isolated from cell cultures of human retinal pigment epithelial cells, as reported by Glaser et al. [9]. In our study, somatostatin analogues have been shown to inhibit neovascularization in the normally developing CAM. To our knowledge, this is the first description of somatostatin analogues directly and quantitatively inhibiting angiogenesis. This observation could be responsible for the alterations in tumor growth previously seen in vitro and in uiuo. Optimization of drug concentration and drug exposure time will be necessary to determine if our results are clinically relevant. ACKNOWLEDGMENTS We thank and appreciate the technical and scientific assistance of G. H. Davis. We acknowledge and thank G. Sexton for his help with
VOL.
50, NO. 3, MARCH
1991
the statistical analysis. We also acknowledge and thank A. Schally of the Veterans Administration Medical Center, New Orleans, Louisiana for the kind gift of the sandostatin analogue RC-160. This research was supported, in part, by Sandoz Inc., East Hanover, NJ.
REFERENCES 1.
2. 3.
4.
5.
6. 7.
8.
9.
10.
11.
12.
13.
Brown, R. A., Weiss, K. B., Tomlinson, I. W., Phillips, P., and Kumar, S. Angiogenic factor from synovial fluid resembling that from tumors. Lancet 1: 682, 1980. Auerbach, R. Angiogenesis-inducing factors: A review. Lymphokines 4: 69, 1981. Gospodarowiez, D., Bialecki, H., and Thakral, T. K. The angiogenie activity of the fibroblast and epidermal growth factor. Exp. Eye Res. 28: 501, 1979. Shing, Y., Folkman, J., Haudenschild, C., Lund, D., Crum, R., and Klagsburn, M. Angiogenesis is stimulated by a tumor derived endothelial cell growth factor. J. Cell Biochem. 29: 275, 1985. Phillips, P., and Kumar, S. Tumor angiogenesis factor and its neutralization by a xenogeneic antiserum. Znt. J. Cancer 23: 82, 1979. Folkman, J. Angiogenesis: Initiation and control. Ann. N.Y. Acad. Sci. 401: 212, 1982. Salahuddin, S. Z., Nakamura, S., Biberfield, P., Kaplan, M. H., Markham, P. D., Larsson, L., and Gallo, R. C. Angiogenic properties of Kaposi’s sarcoma-derived cells after long-term culture in uitro. Science 242: 430, 1988. Crum, R., Szabo, S., and Folkman, J. A new class of steroids inhibits angiogenesis in the presence of heparin or a heparin fragment. Science 230: 1375,1985. Glaser, B., Ckampochiaro, P., Davis, J., and Jerdan, J. Retinal pigment epithelial cells release inhibitors of neovascularization. Ophthalmology 94: 780,1983. Fassler, J. E., Hughes, J. H., Cataland, S., and O’Dorisio, T. M. Somatostatin analogue: An inhibitor of angiogenesis? Biomed. Res. (Suppl.) 1: 81, 1988. Zalatnai, A., and Schally, A. V. Treatment of N-nitrosobis (2-0xopropyl) amine-induced pancreatic cancer in Syrian golden hamsters with n-Trp-6-LH-RH and somatostatin analogue RC-160 microcapsules. Cancer Res. 49: 1810, 1989. Dunn, B. E., Fitzharris, T. P., and Barnett, B. D. Effects of varying chamber construction and embryo pre-incubation age on survival and growth of chick embryos in shell-less culture. Anat. Rec. 199: 33, 1981. Woltering, E. W., Ellison, E. C., O’Dorisio, T. M., Sparks, J., Howe, B., and Dyben, T. Somatostatin-like peptides alter calcium but not secretin sensitivity of gastrinoma cells. J. Surg. Res.
40:605,1986. 14. Mozell, E., Woltering,
E. A., O’Dorisio, T. M., Fletcher, W. S., Sinclair, A., and Hill, D. Effect of somatostatin analogue on peptide release and tumor growth in the Zollinger-Ellison syndrome. Surg. Gyn. Obstet. 170: 476, 1990. 15. Woltering, E. A., Mozell, E., O’Dorisio, T., Fletcher, W., and Howe, B. Suppression of primary and secondary peptides with somatostatin analogue in the therapy of functional endocrine tumors. Surg. Gyn. Obstet. 167: 453, 1988. 16. Kvols, L. K., Moertel, C. G., O’Connell, M. J., Schutt, A. J., Rubin, J., and Hahn, R. G. Treatment of the malignant carcinoid syndrome: Evaluation of a long-acting somatostatin analogue. N. Engl. J. Med. 315: 663, 1986. 17. Allen, E., Evers, M., Townsend, J. C., Uchida, T., and Thompson, J. Somatostatin Analogue (201-995) and a-difluoromethylornithine inhibit growth of human carcinoid tumor. Surg. Forum
40:415,1989. 18.
Milhoan,
R., Trudel,
J., Lawrence,
J., Townsend,
C., and
WOLTERING
ET AL.:
SOMATOSTATIN
Thompson, J. Somatostatin inhibits growth of human small cell lung carcinoma in vivo. Surg. Forum 39: 438, 1988. 19.
20.
21.
22.
Weber, C., Meram, L., Koschitzky, T., and Logerfo, P. Inhibition of growth of MCF-7 human breast carcinoma in uivo by somatostatin analogue SMS 201-995. Surg. Forum 39: 66, 1988. Weber, C., Meriam, L., Koschitzky, T., Karp, F., Benson, M., Kenneth, F., and Logerfo, P. Inhibition of growth of human breast carcinomas in viuo by somatostatin analogue SMS 201995: Treatment of nude mouse xenografts. Surgery 106: 416, 1989. Phillips, P., Steward, J., and Kumar, S. Tumor angiogenesis factor (TAF) in human and animal tumors. Znt. J. Cancer 17: 549, 1976. Folkman, J., and Cotran, R. Relation of vascular proliferation to tumor growth. Znt. Rev. Exp. Pathol. 16: 207, 1976.
ANALOGUES 23.
24.
25. 26.
27.
INHIBIT
ANGIOGENESIS
251
Ingber, D. E., Madri, J. A., and Folkman, J. A possible mechanism for inhibition of angiogenesis by angiostatic steroids: Induction of capillary basement membrane dissolution. Endocrinology 119:33,1981. Reubi, J., and Torhorst, J. The relationship between somatostatin, epidermal growth factor, and steroid hormone receptors in breast cancer. Cancer 64: 1254,1989. Schally, A. V. Oncological applicat,ions of somatostatin analogues. Cancer Res. 48: 6977, 1989. Schor, A., Schor, S., and Kumar, S. Importance of a collagen substratum for stimulation of capillary endothelial cell proliferation by tumor angiogenesis factor. Znt. J. Cancer 24: 225, 1979. Auerbach, R., Kubai, L., and Skidky, Y. Angiogenesis induction by tumors, embryonic tissues and lymphocytes. Cancer Res. 36:
3435,1976.