Angiogenesis and its control in the female
reproductive system
Hamish M Fraser and Stephen F Lunn
MRC Human Reproductive Sciences Unit, Centre for Reproductive Biology, Edinburgh, UK
The rapid, controlled and cyclical nature of angiogenesis in the female
reproductive tract suggests that interference with this process should provide a
novel approach to manipulation of reproductive function. Many factors involved
in the regulation of angiogenesis have been identified, and the possibility of
stimulating or inhibiting these paracrine control mechanisms is being addressed
using current advances in the development of angiogenic and anti-angiogenic
compounds. Studies with animal models indicate that the normal processes of
folliculogenesis, ovulation and corpus luteum function in the ovary, and the
control of menstruation and implantation in the endometrium could be
profoundly influenced by manipulation of angiogenesis. Novel therapeutic
agents targeted to the angiogenic pathway may also have a wide range of
applications in pathological processes in the reproductive tract such as cancer,
endometriosis, fibroid growth, and ovarian hyperstimulation syndrome.
Correspondence to
Dr Hamish M Fraser, MRC
Human Reproductive
Sciences Unit, Centre for
Reproductive Biology, 37
Chalmers Street
Edinburgh EH3 9ET, UK
Angiogenesis is the development of new blood vessels by endothelial cell
proliferation and outgrowth from pre-existing vessels. It is regulated by a
complex series of growth factor interactions, including stimulatory,
modulatory and inhibitory regulators and is associated with changes in
the extracellular matrix required to allow migration of the newly-forming
vessels. Apart from during tumour growth and wound healing, the adult
vascular endothelium is generally quiescent. The notable exceptions are
within the female reproductive system; the ovaries, uterus and placenta1"6.
These tissues undergo cyclic changes in angiogenesis necessary to supply
the nutrients and hormone precursors essential for the establishment and
maintenance of pregnancy. During the menstrual cycle, this physiological
angiogenesis in the corpus luteum and endometrium is distinctive in that,
in the absence of conception, it is followed by tissue regression and reinitiation of cyclicity. This tight physiological control contrasts to the
situation in tumours, and raises the possibility that endogenous inhibitors
of angiogenesis play an integral role in the reproductive tract. Thus, these
tissues provide an outstanding system in which to study the normal
physiology of angiogenesis. Understanding the mechanisms of control of
angiogenesis in the reproductive system and the harnessing of major
British Medical Bulletin 2000, 56 (No 3) 787-797
O The British Council 2000
Human reproduction: pharmaceutical and technical advances
advances in the development of angiogenesis inhibitors and stimulators
may reveal new approaches to regulation of reproductive function and
treatment of pathological conditions of the female reproductive tract. In
addition, insight should be gained into the changes which precipitate the
relatively uncontrolled angiogenesis responsible for the growth of solid
tumours and how this may be inhibited.
This chapter describes briefly the potential angiogenic control mechanisms within the female reproductive tract and how these might be targeted,
reviews recent data investigating the consequences of angiogenesis
inhibition in experimental models, and examines the projected clinical
applications. Since mechanisms involved in regulation of cyclic ovarian
and uterine angiogenesis have much in common, progress with respect to
the ovary is highlighted and the reader referred to recent reviews for the
uterus5"7.
Angiogenic control mechanisms
Folhcular angiogenesis commences at, or shortly before, the development
of a recognisable thecal layer and increases in follicles undergoing maturation to the antral stage. Blood vessels increase in number and size as the
follicle develops, but do not penetrate the granulosa cell layer while the
basement membrane surrounding it remains intact3-8. The extent to which
angiogenesis governs follicular atresia on the one hand, and selection of the
dominant follicle on the other, is of great interest; in some species, a
relationship has been demonstrated3, but differences are not obvious in the
human ovary8.
During the process of ovulation, the basement membrane loses its integrity
and the thecal blood vessels access the granulosa cell layer. Subsequently,
formation of the corpus luteum is associated with intense angiogenesis (Fig.
I) 9 . By the mid-luteal phase, when the corpus luteum (CL) is fully functional,
blood flow is among the greatest of any tissue in the body. Proliferation of
endothelial cells continues at a moderate level, decreasing towards the end of
the cycle. By the mid-luteal phase, the microvascular tree is established. It
appears that the rescue of the CL in early pregnancy is not associated with a
further spurt of intense angiogenesis since chorionic gonadotrophin (hCG)
does not influence endothelial cell proliferation (Fig. I) 9 .
Angiogenesis is regulated in the reproductive tract and elsewhere by at
least 20 angiogenic growth factors and inhibitors identified to date. A key
player is vascular endothelial growth factor (VEGF) also known as
vascular permeability factor10. VEGF is a member of the platelet-derived
growth factor family, bemg a dimeric protein existing as 5 different homodimeric isoforms generated by alternative splicing from a single gene. The
secreted forms are composed of 121 and 165 amino acids and act upon
788
British Medical Bulletin 2000,56 (No 3)
Angiogenesis
300 - . a )
glow1000J
150- b)
1 loo's
E 50z
0J
Early
Mid
Late
Rescued
Fig. 1 (a) Number of endothelial cells and (b) proliferating endothehal cells per field in
the human corpus luteum during the early, mid and late luteal phase and after rescue by
hCG administration. Data are means ± SEM; *shows significant differences between
groups Note the intense proliferation during the early luteal phase and the resultant
increase in endothelial cell number. From Rodger etaP with permission.
vascular endothelial cells in which its receptor tyrosine kinases (Flt-1
and KDR) are expressed selectively. VEGF is expressed in the ovary in a
regulated manner11'12 and its importance in ovarian angiogenesis has
been established following studies in which its action has been inhibited
during the ovulatory cycle (see below).
Particular attention is being given to the inter-relationship of VEGF
with a novel family of angiogenic regulators, the angiopoietins, which
consist of about 500 ammo acids and act via another receptor tyrosine
kinase, Tie-2, on the endothelial cells. Angiopoietins are relatively highly
expressed in the ovary, uterus and placenta12. Since they appear to play
a major role in stabilisation of blood vessels on the one hand and
endothelial cell death on the other, the angiopoietins are attractive
candidates as regulators of these divergent processes as they occur in the
female reproductive tract. When active angiogenesis is occurring in the
presence of VEGF, Ang-2 acts in concert to enhance angiogenesis, while
Ang-1 is involved in the process of maturation and stabilisation of new
blood vessels12. Paradoxically, in circumstances of reduced VEGF
expression, Ang-2 may act to destabilise blood vessels and induce
vascular regression by competitive inhibition of Ang-112'13.
The temporal changes in the pattern of expression of VEGF and
angiopoietin mRNA and in the rat ovary have shown that VEGF is
British Medical Bulletin 2000, 56 (No 3)
789
Human reproduction- pharmaceutical and technical advances
expressed in the pre-ovulatory follicle and in the hormone-producing
cells of the corpus luteum12. Ang-2 is present in theca of the preovulatory follicle and in the endothelial cell sprouts of the developing
corpus luteum, suggesting an angiogenic role in concert with VEGF at
this time. Intriguingly, Ang-2 is highly expressed in regressing corpora
lutea, when VEGF is declining, implying its involvement in the luteolytic
process via destabilisation of blood vessels. Similarly, quantitative PCR
in the bovine revealed the Ang-2 :Ang-l mRNA ratio increases during
luteolysis14. While luteolysis in the normal cycle is associated with
endothelial cell loss, stabilisation and maturation of luteal blood vessels
may be crucial to the maintenance of the corpus luteum in early
pregnancy4. The mechanisms involved in these divergent processes are
not understood. If luteolysis is associated with an increase in Ang-2
expression and blood vessel destabilisation perhaps 'rescue' of the
corpus luteum is associated with an increased expression of Ang-1
resulting in recruitment of pericyte cells which stabilise the blood
vessels. Control of luteal function differs markedly between species and
it is not yet known whether similar inter-relationships between positive
and negative angiogenic mechanisms exist in the primate corpus luteum.
Two further natural inhibitors of angiogenesis have been discovered
recently, both being fragments of larger molecules; angiostatin is 98%
identical to an internal fragment of plasminogen,15 while endostatin
corresponds to the C-terminal fragment of collagen XVIH16. Both proteins
inhibit tumour growth, but their role in reproductive physiology is
unknown. Since it appears that an 'angiogenic switch' regulates the
divergent states of follicular development, luteal rescue and luteolysis, and
endometrial function, it is of importance to localise and quantify the
temporal changes in the angiostatin/endostatin system in these situations
to explore the hypothesis that production of these proteins is increased in
the regressing state and suppressed in association with rescue.
Changes in the extracellular matrix (ECM) are an essential part of the
angiogenic process and are regulated by metalloproteinases and their
inhibitors. Cell adhesion to the ECM is primarily through integrins, a family
of heterodimeric transmembrane proteins17. These mediate cell adhesion to
the ECM leading to intracellular signalling events that regulate cell survival,
proliferation and migration18. During angiogenesis, it is thought that a
number of integrins expressed on the surface of activated endothelial cells
regulate critical adhesive interactions with a variety of ECM proteins.
A complete understanding of these complex regulatory mechanisms
within the ovary and uterus will allow systematic targeting in pathologies
which have a vascular component. While basic studies strive to achieve
this goal, targeting of putative key regulators has been conducted both in
the laboratory and clinic during the last decade, particularly in the area of
solid tumour biology.
790
British Medical Bulletin 2000,56 (No 3)
Angiogenesis
Manipulation of angiogenesis
Development of anti-angiogenic agents
Over 300 endogenous, natural or synthetic anti-angiogenic agents have
been described19. The initial development involved screening for such
activity in compounds developed for other indications. A second group
required the design of specific antagonists to known positive angiogenic
factors and their receptors, while a third class have been identified from
tumour sources as having negative effects on angiogenesis (Table 1).
Numerous studies have shown that inhibition both by broad-spectrum
or systematic targeting of the angiogenic pathway can have profound
suppressive effects upon tumour growth in animal models16-20"22. These
include the effects of neutralisation of VEGF in an ovarian cancer
model23. Most of the clinical trials on angiogenesis inhibitors have
employed first generation compounds19 such as the fumagillin analogue,
TNP-470, but in common with many of these compounds its mechanism
of action is multi-factorial and is still under investigation.
Specific angiogenic peptide or receptor antagonists are being developed
for clinical application. A humanised monoclonal antibody to VEGF is
currently on clinical trial and the VEGF receptor is also being targeted
using a soluble truncated Flt-1 receptor or tyrosine kinase inhibitors, the
latter having the advantage of oral activity. Combination treatments such
Table 1 Examples of anti-angiogenic agents
First generation
Suramin
Fumagillin derivative (TNP-470)
Thahdomide
Medroxyprogesterone
Anti-oestrogens
Inhibition of positive regulators
VEGF inhibitors
VEGF immunoneutrahsing antibodies
Truncated soluble receptor
Tyrosine kinase receptor antagonists, eg Sugen 5416
VEGF receptor immunoneutrahsing antibodies
VEGF diphtheria toxin
Ribozymes
Angiopoietin/Tle receptor antagonists
Metalloproteinase inhibitors (e g Marimastat)
Integrin inhibitors (e.g Vitaxin)
Administration of negative regulators
Angiogemn
Angiostatin
Thrombospondin
Endostatin
British Medical Bulletin 2000, 56 (No 3)
791
Human reproduction: pharmaceutical and technical advances
as VEGF inhibitor followed by Ang-2 may prove to be exceptionally
effective in preventing angiogenesis13.
The involvement of the ECM in angiogenesis has led to development of
metalloproteinase inhibitors and neutralising antibodies capable of inhibiting specific integrins. The therapeutic application of such approaches is
currently under investigation in the clinic.
Initial studies in experimental tumour models using negative regulators
of angiogenesis, such as recombinant angiostatin, were encouraging in
that they demonstrated that tumours may be treated successfully with
these agents without generatmg resistance15'16-22, hi addition, prolonged
tumour dormancy was observed post-treatment. The challenge of
extending these results to man is eagerly awaited.
As increasing numbers of specific antagonists are developed, it should
become possible to target systematically key steps in the angiogenic
pathway at selected periods of the reproductive cycle using suitable
animal models and their effects upon angiogenesis in the ovary, uterus
and placenta determined. This potential for inhibition of physiological
angiogenesis is only just beginning to be explored and three of the first
experiments are reviewed below.
Rodent studies
In the reproductive system, the pioneering study underlining the importance
of angiogenesis was carried out in the mouse using TNP-47024. Females
were treated at various times after mating and it was shown that TNP-470
prevents pregnancy by interfering with the process of deciduahsanon and
placental and yolk sac formation. In mice with regular oestrous cycles
treated for 16 days, mating was abolished, there was a marked decrease in
endometrial stromal and glandular cell proliferation, uterine blood vessel
development was severely restricted, and the ovaries contained fewer and
smaller corpora lutea than cyclic controls.
The importance of angiogenesis in the immature rat ovary was shown
in studies in which VEGF was blocked using a truncated Fit receptor25.
Follicular development was induced by gonadotrophin treatment and
resulted in multiple ovulations and elevated serum progesterone in
control rats, whereas those treated with antagonist starting 4-6 h prior
to gonadotrophin administration showed small antral follicles and only
a few corpora lutea. The latter were relatively avascular and often
demonstrated central ischaemic necrosis. While the uteri of the gonadotrophin-treated rats demonstrated morphological evidence of steroid
stimulation, such changes were largely blocked in antagonist-treated
animals, presumably secondary to the comparative lack of steroid
stimulation, but also in part as a result of inhibition of uterine VEGF.
792
British Medical Bulletin 2000,56 (No 3)
Angiogenesis
Primate studies
The mechanisms which regulate luteal function in rodents and primates are
markedly different, hi the non-fertile cycle of the rodent, the corpus luteum
is active for less than a day, while that of primates is functional for 2 weeks
prior to its regression. In addition, in many non-primates, regression of the
corpus luteum is mediated by a specific luteolysin, a phenomenon not
apparent in the human. In the rat model, since VEGF antagonist treatment
was initiated prior to the onset of follicular hyperstimulation,25 it was not
possible to dissociate the luteal inhibition from the consequences of
inhibition of angiogenesis in the developing follicle. Experiments outlined
below targeted the corpus luteum specifically by treating animals during
the normal cycle at the time of ovulation.
When macaques or marmosets were treated with TNP-470, there were
no apparent effects upon angiogenesis or luteal function26 in contrast to
the results in the mouse. The reason for this difference is unclear, but
could be related to dose or route of administration of the compound or
real species differences. The effects of 'first generation' compounds may
be comparatively weak relative to the more specific antagonists being
developed. These primate results emphasise the need for caution in
extrapolating results from the rodent to the human.
Luteal angiogenesis is suppressed by neutralisation of VEGF 27 .
Marmosets were treated with a monoclonal antibody to human VEGF
starting at the time of ovulation and continuing for 3 days (early luteal
phase group) or 10 days (mid-luteal group). Treatment effects were
determined by measuring plasma progesterone concentrations, endothelial cell proliferation index using bromodeoxyuridine (BrdU)
incorporation, and endothelial cell number using the specific cell marker,
factor Vin. Neutralisation of VEGF during the early luteal phase inhibited
the intense angiogenesis associated with luteal formation (Fig. 2a,b). This
resulted in blockade of development of the normally extensive capillary
bed since, in the mid-luteal phase treated animals, the endothelial cell
numbers were reduced (Fig. 2c,d). Luteal function, as judged by secretion
of progesterone, was markedly compromised by the treatment, being
reduced by 60% in comparison with controls and demonstrating the
principle of manipulation of ovarian function by this approach. Further
studies using this model should allow the identification of the essential
components of the angiogenic pathway and provide a vehicle for
translating these findings into safe and effective clinical application.
Information on stimulation of angiogenesis within the reproductive
system, e.g. by recombinant protein or by gene transfer, is currently limited.
However, the feasibility of this approach has been demonstrated in clinical
trials in patients with critical limb ischaemia and myocardial ischaemia19.
British Medical Bulletin 2000; 56 (No 3)
793
Human reproduction: pharmaceutical and technical advances
Fig. 2 Photomicrographs of early luteal phase marmoset corpora lutea showing
(a) proliferating endothehal cells in microvessels and capillaries (dark staining nuclei) using
BrdU incorporation as a marker, (b) reduced proliferation following anti-VEGF treatment;
(c) localisation of endothehal cells (dark staining) using Factor VIII as a marker in the midluteal corpus luteum; and (d) reduction in the vascular tree following anti-VEGF treatment.
(Scale bar = 200 |im). SE Dickson and HM Fraser unpublished.
Clinical studies and possible applications
Anti-angiogenic compounds have been used primarily for treatment of
cancer, but are also envisaged for non-malignant pathologies which have a
neovascular component (e.g. rheumatoid arthritis, psoriasis and retinal
neovascularisation)20*22. Results from the use of 31 agents which have
entered clinical trials at over 140 centres have been reviewed19. In addition
to treatment efficacy, such studies allow evaluation of side-effects and these
have turned out to be more wide-spread than anticipated. This may suggest
that, in tissues in which there appears to be no angiogenesis, angiogenic
regulators have a more general 'housekeeping' role. The introduction of
drugs having a selective site of action may dissociate specific from nonspecific adverse effects on the vasculature or overcome them altogether.
Paradoxically, such observations could shed light on novel physiological
sites and/or pathways of vascular regulation. In any event, it is important
that future trials incorporate a multidisciplinary team approach19 which
includes a reproductive component. Ultimately, as with any other drug, a
risk-benefit assessment will have to be undertaken prior to commencement
of treatment.
794
British Medical Bulletin 2000,56 (No 3)
Angiogenesis
To date, there have been no reports on the effects of regulators of
angiogenesis on the reproductive system in women, although it is acknowledged that it offers numerous potential targets both for manipulation of
fertility and treatment of pathologies involving a vascular component1^'24'25. Manipulation of angiogenesis indicates a potentially powerful approach to either promoting or inhibiting the normal processes of
folliculogenesis, ovulation and corpus luteum function in the ovary, and
to the control of menstruation and implantation in the endometrium. It is
unlikely that for fertility control, one would aim to suppress angiogenesis
for prolonged periods as the risk of disruptive side-effects would be
unacceptable. Targeting anti-angiogenic treatment to the folhcular phase
would probably compromise follicular development by inhibition of the
thecal vasculature. Treatment during the early luteal phase would
suppress progesterone production and probably inhibit implantation.
Despite the fact that the functioning corpus luteum is essential for the
establishment of pregnancy, there is no agent currently in use which can
prevent the supporting effect of hCG during the conceptual cycle. Thus,
the use of angiogenesis inhibitors during the postovulatory period could
have potential for development.
One clinical condition in which inhibition of follicular angiogenesis would
be indicated is ovarian hyperstimulation syndrome (OHSS) which occurs in
a proportion of patients undergoing ovulation induction. The pathophysiology of OHSS is complex, but increased ovarian angiogenesis and
permeability are major factors28. An effective treatment for this potentially
fatal condition has proved elusive and it is likely that the introduction of antiangiogenic approaches will offer a major clinical advantage.
Angiogenesis is implicated to varying degrees in numerous pathological
processes of the endometrium including dysfunctional uterine bleeding,
endometrial response to exogenous hormonal treatments, bleeding
associated with intra-uterine devices, uterine leiomyomata, endometriosis
and endometrial carcinoma7. The pathophysiology of the vasculature in
these conditions is currently the focus of active research. For example,
VEGF has been implicated in endometriosis5. Inhibition of angiogenesis
should have potential in the treatment of these conditions, possibly as an
adjunct to existing therapies operating via a separate pathway.
Administration of angiogenic agents to increase blood vessel development could be of use in stimulation of follicular development, for the
treatment of inadequate luteal function, early pregnancy failure, preeclampsia and intra-uterine growth retardation.
Studies in the female reproductive tract are gathering momentum as
the potential for control of angiogenesis in normal and pathological
conditions come under scrutiny. This is an exciting field in which
advances are likely in the near future and this review has given some
indication as to the prospects for further developments.
British Medical Bulletin 2000, 56 (No 3)
795
Human reproduction: pharmaceutical and technical advances
References
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
796
Folkman J. Angiogenesis in female reproductive organs. In: Alexander NJ, D'Arcangues C (Eds)
Steroid Hormones and Uterine Bleeding. Washington DC: AAAS Press, 1992; 143-58
Reynolds LP, Kilhlea SD, Redmer DA. Angiogenesis in the female reproductive system. FASEB
J 1992; 6: 886-92
Redmer DA, Reynolds LP. Angiogenesis in the ovary. Rev Reprod 1996; 1: 182-92
Augustin HG. Development of the vascular system of the corpus luteum. In: Risau W, Rubanyi
G (Eds) Morphogenesis of Endotbeltum Reading, UK Harwood, 2000; 237-254
Smith SK. Angiogenesis, vascular endothelial growth factor and the endometnum. Hum
Reprod Update 1998; 4: 509-19
Rees M, Hague S, Oehler MK, Bicknell R. Regulation of endometnal angiogenesis. Climateric
1999; 2: 52-8
Abulafia O, Sherer DM Angiogenesis of the endometnum. Obstet Gynecol 1999; 94: 148-53
Suzuki T, Sasano H, TaKaya R, Fukaya T, Yajima A, Nagura H. Cyclic changes of vasculature
and vascular phenotypes in normal human ovaries. Hum Reprod 1998; 13: 953-9
Rodger FE, Young FM, Fraser HM, Dhngworth PJ. Endothelial cell proliferation follows the
mid-cycle luteimzing hormone surge, but not human chononic gonadotrophin rescue, in the
human corpus luteum Hum Reprod 1997; 12: 1723-9
Ferrara N, Davis-Smyth T The biology of vascular endothelial growth factor Endocrine Rev
1997; 18: 4-25
Gordon JD, Mesiano S, Zaloudek CJ, Jaffe RB. Vascular endothelial growth factor localization
in human ovary and fallopian tubes: possible role in reproductive function and ovanan cyst
formation. / Clm Endocrinol Metab 1996; 81 353-9
Maisonpierre PC, Sun,C, Jones PF et al. Angiopoietin-2, a natural antagonist for Tie2 that
disrupts in vtvo angiogenesis. Science 1997, 277: 55-60
Hanahan D. Signaling vascular morphogenesis and maintenance. Science 1997, 277: 48-50
Goede V, Schmidt T, Kimmina S, Kozian D, Augustin HG. Analysis of blood vessel maturation
processes during cyclic ovarian angiogenesis. Lab Invest 1998; 78: 1385-94
O'Reilly MS, Holmgren L, Chen C, Folkman J. Angiostatin induces and sustains dormancy of
human primary tumors in mice. Nat Med 1996; 2: 689-92
O'Reilly MS, Boehm T, Shing Y et al. Endostatin: an endogenous inhibitor of angiogenesis and
tumor growth Cell 1997; 88: 277-85
Bazzoni G, Dejana E, Lampugnani G. Endothelial adhesion molecules in the development of
the vascular tree: the garden of forking paths. Curr Opm Cell Biol 1999; 11: 573-81
Ehcein BP, Cheresh DA. The role of av integnns during angiogenesis: insights into potential
mechanisms of action and clinical development./ Clm Invest 1999, 103: 1227-30
Thompson WL), Li WW, Maragoudakis M. The clinical manipulation of angiogenesis:
pathology, side-effects, surprises, and opportunities with novel human therapies. / Pathol 1999;
187: 503-10
Auerbach W, Auerbach R. Angiogenesis inhibition: a review. Pharm Ther 1994; 63: 265-311
Skobe M, Rockwell P, Goldstein N, Vosseler S, Fusenig NE. Halting angiogenesis suppresses
carcinoma cell invasion. Nat Med 1997; 3: 1222-7
Jones A, Bicknell R Angiogenesis and invasion. In- Souhami RL, Tannock I, Hohenberger P,
Honot JC (Eds) Oxford Textbook of Oncology. Oxford. Oxford University Press, 2000; In
press
Mesiano S, Ferrara N, Jaffe RB Role of vascular endothelial growth factor in ovanan cancer:
inhibition of ascites formation by lmmunoneutralization. Am J Pathol 1998; 153: 1249-56
Klauber N, Rohan RM, Flynn E, D'Amato RJ. Cntical components of the female reproductive
pathway are suppressed by the angiogenesis inhibitor AGM-1470. Nat Med 1997; 3: 443-6
British Medical Bulletin 2000;56 (No 3)
Angiogenesis
25 Ferrara N, Chen H, Davis-Smyth T et al. Vascular endothehal growth factor is essential for
corpus luteum angiogenesis Nat Med 1998; 4: 336-40
26 Fraser HM, Dickson SE, Morris KD, Erickson GF, Lunn SF. The effects of the angiogenesis
inhibitor TNP-470 on luteal establishment and function in the primate. Hum Reprod 1999; 14:
2054-60
27 Fraser HM, Dickson SE, Morris KD, Lunn SF, Carroll V, Bicknell R. Suppression of luteal
angiogenesis by neutralization of vascular endothehal growth factor in the primate.
Endocrinology 2000, 141: 995-1000
28 Schenker JG. Clinical aspects of ovarian hyperstimulation syndrome. Eur J Obstet Gynecol
1999; 85 13-20
British Medical Bulletin 2000, 56 (No 3)
797