0021-972X/00/$03.

00/0
The Journal of Clinical Endocrinology & Metabolism
Copyright 2000 by The Endocrine Society

Vol. 85, No. 3

Printed in U.S.A.

Angiogenesis in Pituitary Adenomas and the Normal

Pituitary Gland
HELEN E. TURNER, ZSUSHA NAGY, KEVIN C. GATTER, MARGARET M. ESIRI,
ADRIAN L. HARRIS, AND JOHN A. H. WASS
Departments of Endocrinology (H.E.T., J.A.H.W.) and Neuropathology (Z.N., M.M.E.), Radcliffe
Infirmary; Department of Pharmacology, University of Oxford (Z.N.); and Departments of Cellular
Science (K.C.G.) and Molecular Angiogenesis Group (A.L.H.), Imperial Cancer Research Fund,
Institute of Molecular Medicine, John Radcliffe Hospital, Oxford, United Kingdom OX2 6HE
ABSTRACT
Angiogenesis is essential for tumor growth beyond a few millimeters in diameter, and the intratumoral microvessel count that represents a measure of angiogenesis has been correlated with tumor
behavior in a variety of different tumor types. To date no systematic
study has assessed pituitary tumors of different secretory types, correlating vascular count with tumor size. The vascular densities of
pituitary tumors and normal anterior pituitary were therefore assessed by counting vessels labeled using the vascular markers CD31
and ulex europaeus agglutinin I. One hundred and twelve surgically
removed pituitary adenomas (30 GH-secreting, 25 prolactinomas, 15
ACTH-secreting, and 42 nonfunctioning tumors) were compared with
13 specimens of normal anterior pituitary gland. The vascular counts

NGIOGENESIS IS the process of development of new

vessels from existing blood vessels and is crucial for
embryo development, wound healing, and the female reproductive cycle. However, angiogenesis has also been
shown to be required for tumor growth and metastasis (1).
On the basis of experiments showing that tumors implanted
into isolated perfused organs failed to develop, whereas the
same tumors implanted within 6 mm of blood vessels induced angiogenesis, grew, and metastasized (2, 3), Folkman
proposed that solid tumors are dependent on angiogenesis
for growth beyond a few millimeters in size, and that an
increase in tumor diameter required a corresponding increase in vascularization (4). In many human tumors, including breast, bladder, and stomach, angiogenesis assessed
by vascular counts, correlates with development of metastasis (5), poor prognosis (6), and poor survival (7, 8).
It was reported by Schechter in 1972 that the parenchyma
of pituitary tumors appeared less vascularized than autopsy
specimens of normal tissue (9). Jugenburg and colleagues
used immunostaining for factor 8-related antigen to assess
vascular density in a group of pituitary adenomas and carcinomas (10). They showed that pituitary adenomas had
lower vascular densities compared to nontumorous pituitary, but the relationship to tumor size was not studied, and
vascular hot spots were not positively identified for count-

Received September 2, 1999. Revision received November 23, 1999.

Accepted December 4, 1999.
Address all correspondence and requests for reprints to: Prof. J. A. H.
Wass, Department of Endocrinology, Radcliffe Infirmary, Woodstock
Road, Oxford, United Kingdom OX2 6HE.

in the normal anterior pituitary gland were significantly higher (P

0.05) than those in the tumors using both CD31 and ulex europaeus
agglutinin I. In addition, microprolactinomas were significantly less
vascular (P 0.05) than macroprolactinomas, although there was no
such difference between vascular densities of microadenomas and
macroadenomas producing GH. ACTH-secreting tumors were, like
microprolactinomas, of much lower vascular density than the normal
pituitary and other secreting and nonsecreting tumor types. In
marked contrast to other tumors, pituitary adenomas are less vascular than the normal pituitary gland, suggesting that there may be
inhibitors of angiogenesis that play an important role in the behavior
of these tumors. (J Clin Endocrinol Metab 85: 1159 1162, 2000)

ing. It has been suggested that as the most angiogenic tumor

clones will determine tumor behavior, the area of the highest
microvessel count (hot spot) should be positively identified
when assessing vascular density (5, 11, 12). Using factor
8-related antigen in 22 pituitary adenomas, the highest vascular counts occurred in FSH-expressing adenomas, and the
lowest were found in GH-secreting tumors (13). There was
no comparison of vascular density with tumor size or normal
pituitary tissue.
The object of this study was to assess the vascular densities
of a large number of carefully characterized pituitary tumors
of different secretory types and compare them with those of
normal pituitary, using two different endothelial markers
and assessing vessel hot spots (14, 15). In addition, we compared vascular density of macroadenomas and microadenomas to determine whether angiogenesis may play a role in
determining pituitary tumor size.
Materials and Methods
Specimen collection
One hundred and twelve surgically removed pituitary adenomas
were investigated. There were 30 GH-secreting tumors [22 macroadenomas (1 cm in diameter) and 8 microadenomas (1 cm in diameter)],
6 microprolactinomas, 19 macroprolactinomas, 15 ACTH-secreting tumors (Cushings disease), and 42 nonfunctioning pituitary adenomas (28
gonadotropin-positive and 14 negative on immunostaining). Thirteen
specimens of normal anterior pituitary gland obtained during surgery
for pituitary tumor (12) and at autopsy (1) were also studied. The tissue
has been fixed in 4% buffered formalin, dehydrated, and embedded in
paraffin. Histological examination and immunohistochemistry for anterior pituitary hormones had been performed previously and together
with the clinical, biological, and radiological data were used to fully
characterize each tumor type.

1159

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 August 2016. at 07:29 For personal use only. No other uses without permission. . All rights reserved.

1160

JCE & M 2000

Vol 85 No 3

TURNER ET AL.

Immunohistochemistry for CD31 and ulex europaeus

agglutinin I (UEAI)

Results
Comparison of different vascular markers

The streptavidin-biotin-peroxidase complex technique was used for

CD31, and the alkaline phosphatase/antialkaline phosphatase method
was used for UEAI.
Four-micron sections were mounted on aptes (3-aminopropyl triethoxy silane; Sigma, St. Louis, MO)-coated slides, dewaxed, and rehydrated. Endogenous peroxidase activity was blocked using 3% hydrogen peroxide for CD31 cases. Sections for CD31 staining were pretreated
using 0.1% trypsin at 37 C for 15 min followed by microwave pretreatment in sodium citrate buffer, pH 6. Slides for staining with UEAI did
not require pretreatment. Nonspecific primary antibody binding was
blocked using FCS at a dilution of 1:20. The primary antibodies were
applied for 60 min at room temperature. For CD31, the DAKO Corp.
antibody (Carpinteria, CA) was applied at a dilution of 1:20. The biotinylated UEAI (Vector Laboratories, Inc., Burlingame, CA) was used at
a dilution of 1:200. After three washes in phosphate-buffered saline,
biotinylated secondary antibody (Insight Biotechnology, Wembley,
Middlesex, UK) was applied at a 1:200 dilution for 30 min at room
temperature, followed by washes and then application of the appropriate avidin-biotin-peroxidase complex. The horseradish peroxidasestreptavidin complex (DAKO Corp.) was applied at 1:400 dilution for 30
min. The alkaline phosphatase/antialkaline phosphatase complex (Vector Laboratories, Inc.) was applied to the UEAI cases for 30 min, followed
by washes. Color development was performed with metal-enhanced
diaminobenzidine (Pierce Chemical Co., Rockford, IL) applied for 15
min to the CD31 cases and with fast red substrate applied for 20 min to
the UEAI cases. The slides were lightly counterstained with hematoxylin. Negative controls were performed where FCS replaced the primary
antibody.

Assessments of vascular density using mean Chalkley

count and grade were highly correlated (Spearman rank
correlation CD31 mean and grade, r 0.8; UEAI mean and
grade, r 0.9). Vascular densities measured using the two
different endothelial markers were different. UEAI consistently stained more vessels than CD31. Despite these differences, there was good correlation between the two markers
(CD31 mean and UEA1 mean: r2 21.8%; P 0.0002; CD31
grade and UEAI grade: r2 25.3%; P 0.0000). The intrarater tests for vascular count and vascular grade both gave
values of 0.6, indicating substantial agreement ( values,
0.6 0.8) between assessments made on different occasions
by the same observer.

Assessment of vascular density

Vascular density was assessed blindly by 1 examiner without prior
knowledge of tumor type or size. The Chalkley point technique was used
(14). The most vascular area of the tumor section was identified at low
power as the hot spot. A 25-point Chalkley eyepiece graticule was
orientated so that the maximum number of points was on or within areas
of highlighted vessels at 250. The mean of the counts for the 3 most
vascular areas was recorded. An overall subjective semiquantitative
grading system was also used (1 and 2, low and low moderate vascular
density; 3 and 4, high and very high vascular density). The counts and
grades were made by a single observer (H.E.T.), and 20% were checked
by a second blinded observer (K.C.G.), with 100% concordance for
grading into low and high vascular densities.

Vascular counts

Vascular counts in normal anterior pituitary were 7.1 1.1

with CD31, and 10.3 1.3 using UEAI. Vascular counts in
tumors were significantly lower than those in normal tissue
(Table 1). The semiquantitative vascular grades in normal
pituitary tissue were 3.8 0.4 (sd) with CD31 and 3.9 0.3
with UEAI. Vascular grades in tumors were also significantly
lower than those in normal tissue (Table 1 and Fig. 1, a and
b). Macroprolactinomas were the most vascular tumor, significantly more vascular than functionless macroadenomas
(P 0.05). Microprolactinomas and ACTH-secreting tumors
were the least vascular (P 0.05).
Tumor size

Microprolactinomas were significantly less vascular than

macroprolactinomas (Fig. 2), but there was no relationship
between vascular count and tumor size when GH-secreting
tumors were compared (Table 1). Bromocriptine treatment of
patients (n 14) with macroprolactinomas before surgery
was not related to the vascular density of the tumors. All
patients with microprolactinomas had received bromocriptine, but were resistant or intolerant.
Discussion

Statistical analysis
The Statgraphics software package (Manugistics, Rockville, MD) was
used. ANOVA was used for categorical data analysis and regression
analysis for continuous variables. Statistical analysis using the method
described by Landis and Koch was used to assess intra-rater reliability
on sections counted and graded by the same observer on different
occasions (16).

Pituitary tumors are less vascular than normal pituitary

tissue, suggesting that angiogenesis may be inhibited in these
tumors. This is consistent with the findings of Jugenburg and
colleagues (10), but is in marked contrast to studies in other
tissues, for example breast (11), prostate (17), and lung (18),
in which tumors are more vascular than respective normal

TABLE 1. Vascular density of normal pituitary and pituitary adenomas

Type of tumor

Mean UEAI
Chalkley count

Normal pituitary
Microprolactinoma
Macroprolactinoma
GH-secreting microadenoma
GH-secreting macroadenoma
Nonfunctioning
ACTH-secreting

10.3 (1.3), [12]

4.2 (1.5), [4]
7.1 (1.3), [17]
8.7 (0.8), [5]
7.1 (2.4), [17]
7.4 (2.3), [42]
5.0 (1.4), [14]

UEAI grade

3.9
1.6
2.7
2.6
3.0
2.9
2.2

(0.3),
(0.7),
(0.5),
(1.3),
(1.0),
(0.7),
(0.4),

[13]
[6]
[19]
[7]
[22]
[42]
[15]

Mean CD31
Chalkley count

7.4
3.5
6.3
4.9
5.8
5.4
3.1

(0.9),
(0.6),
(0.6),
(2.1),
(1.1),
(1.0),
(0.3),

[12]
[5]
[9]
[8]
[17]
[34]
[4]

CD31 grade

3.8
1.2
3.3
2.5
2.5
2.7
1.0

(0.4), [12]
(0.5), [5]
(0.5), [11]
(1.0), [8]
(0.7), [18]
(0.5), [34]
(0), [5]

Mean (SD), [number of cases].

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 August 2016. at 07:29 For personal use only. No other uses without permission. . All rights reserved.

ANGIOGENESIS IN THE PITUITARY

1161

FIG. 2. Comparison of vascular density between macroprolactinomas

and microprolactinomas. Vascular density was measured by CD31
expression. Macro, Macroprolactinomas; Micro, microprolactinomas.
The asterisk indicates mean values; error bars indicate the SEM.

FIG. 1. Comparison of vascular density between normal pituitary

and different tumor types. a, Vascular density as measured by CD31
expression. B, Vascular density as measured by UEA1 staining. Normal, Normal anterior pituitary gland; Acro, GH-secreting tumors;
NFA, nonfunctioning tumors; MacPRL, macroprolactinomas;
MicPRL, microprolactinomas; Cushings, ACTH-secreting tumors.
The asterisk indicates mean values; error bars indicate the SEM.

tissue. Although pituitary tumors are benign adenomas,

rather than carcinomas, precarcinoma of the cervix and
breast exhibit high levels of angiogenesis (19 21). The relatively common finding of incidental nonprogressive pituitary microadenomas in autopsy specimens or radiological
series (22) may be related to inhibition of angiogenesis, as it
is known that for tumors to enlarge, increased vascularization is required (4). Low vascular density and/or inhibition
of angiogenesis may also play a role in the usually slow
growth of pituitary adenomas (10). Alternatively, the low
growth rate of these tumors may not influence the metabolic
demand significantly, so that vascularization does not limit
growth. The proangiogenic growth factor vascular endothe-

lial growth factor has been identified in the pituitary gland

(23), as have various inhibitors of angiogenesis, including
16-kDa PRL (24, 25) and leukemia inhibitory factor (26). The
overall balance of these growth factors in the pituitary may
determine angiogenic phenotype.
Our results show for the first time that different pituitary
tumors vary in the relationship between size and vascular
density. There is no difference in vascular density between
GH-secreting macroadenomas and microadenomas, but microprolactinomas are significantly less vascular than macroprolactinomas. This fits with the clinical observation that
microprolactinomas rarely progress in size and are a distinct
clinical entity from macroprolactinomas, which may grow to
a considerable size, suggesting that they are not part of the
same pathological process (27). Macroprolactinomas have
been shown to have higher labeling indexes (as a measurement of proliferation) than microprolactinomas, measured
using Ki-67 and proliferating cell nuclear antigen (27). In
contrast to macroprolactinomas, up to one third of patients
with microprolactinomas will show spontaneous remission
(28). In contrast, different size GH-secreting tumors are clinically part of the same spectrum of disease.
Vascular counts determined using immunostaining for
CD31 and UEAI were clearly related, but the counts using
UEAI were higher than those using CD31. In addition, a
proportion of tumors did not stain with CD31 (possibly due
to differences in fixation), and UEAI was occasionally not
assessable because of extensive Golgi staining (29).
Unlike other sites of tumor formation, the anterior pituitary has a dual blood supply; the hypothalamo-pituitary
portal supply is the main source, carrying blood from the
median eminence with hypothalamic releasing and inhibitory factors, but there is an additional direct arterial supply
from the loral and capsular arteries (30). The source of the
blood vessels supplying the tumors is unclear, although
there are several reports suggesting that a direct arterial
supply may develop or perhaps predispose to pituitary tumor development (31). An angiographic study demonstrated
tumor vessels that arose directly from the arterial system

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 August 2016. at 07:29 For personal use only. No other uses without permission. . All rights reserved.

1162

TURNER ET AL.

(32), and an autopsy study of 22 microadenomas showed that

66% of the tumors had a direct extraportal arterial blood
supply (33). An animal model of estrogen-induced lactotroph hyperplasia and tumorigenesis in rats demonstrated
the development of a direct arterial blood supply (34, 35) that
was inhibited by bromocriptine (36). The tumor vasculature
detected in our study may therefore represent a completely
or partially de novo blood supply from the extraportal system.
Thus, although the tumors are less vascular overall, they may
have induced new vessel development from the systemic
circulation, altering oxygen delivery and escaping hypothalamic influences on hormone production. Further work is
required to differentiate the source of the tumoral blood
vessels in the different tumor types compared with the
mainly portal supply to the normal anterior pituitary gland.
The novel findings that pituitary adenomas are less vascular than normal anterior pituitary tissue, and that, depending on tumor type, size is related to vascular density suggest
that these tumors may provide useful information regarding
endogenous inhibitors of angiogenesis and their role in determining overall angiogenic phenotype and the resulting
tumor behavior.
References
1. Folkman J. 1990 What is the evidence that tumours are angiogenesis dependent? J Natl Cancer Inst. 82:4 6.
2. Gimbrone Jr MA, Leapman S, Cotran RS, Folkman J. 1972 Tumour dormancy
in vivo by prevention of neovascularisation. J Exp Med. 73:461 473.
3. Gimbrone Jr MA, Leapman S, Cotran RS, Folkman J. 1973 Tumour angiogenesis: iris neovascularisation at a distance from intraocular tumours. J Natl
Acad Inst. 50:219 228.
4. Folkman J. 1972 Anti-angiogenesis: new concept for therapy of solid tumours.
Ann Surg. 175:409 416.
5. Weidner N, Semple JP, Welch WR, Folkman J. 1991 Tumour angiogenesis and
metastasis: correlation in invasive breast carcinoma. N Engl J Med. 324:1 8.
6. Weidner N, Folkman J, Pozza F, et al. 1992 Tumour angiogenesis: a new
significant and independent prognostic indicator in early stage breast carcinoma. J Natl Cancer Inst. 84:18751887.
7. Maeda K, Chung Y-S, Takasuka S, et al. 1995 Tumour angiogenesis as a
predictor of recurrence in gastric carcinoma. J Clin Oncol. 13:477 481.
8. Bochner BH, Cote RJ, Weidner N, et al. 1995 Angiogenesis in bladder cancer:
relationship between microvessel density and tumour prognosis. J Natl Cancer
Inst. 87:16031612.
9. Schechter J. 1972 Ultrastructural changes in the capillary bed of human pituitary tumours. Am J Pathol. 67:109 126.
10. Jugenburg M, Kovacs K, Stefaneanu L, Scheithauer BW. 1995 Vasculature in
nontumorous hypophyses, pituitary adenomas and carcinomas: a quantitative
morphologic study. Endocr Pathol. 6:115124.
11. Horak ER, Leek R, Klenk N, et al. 1995 Angiogenesis, assessed by platelet/
endothelial cell adhesion molecule antibodies, as indicator of node metastasis
and survival in breast cancer. Lancet. 340:1120 1124.
12. Gasparini G. 1997 Prognostic and predictive value of intra-tumoural microvessel density in human solid tumours. In: Bicknell R, Lewis C, Ferrara N,
eds. Tumour angiogenesis. Oxford: Oxford University Press; 29 44.
13. Pawlikowski M, Pisarek H, Jaranowska M. 1997 Immunocytochemical in-

14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.

JCE & M 2000

Vol 85 No 3

vestigations on the vascularisation of pituitary adenomas. Endocr Pathol.

8:189 193.
Fox SB, Leek RD, Weekes MP, Whitehouse RM, Gatter KC, Harris AL. 1995
Quantification and prognostic value of breast cancer angiogenesis: Chalkley
count and computer image analysis. J Pathol. 177:275283.
Turner HE, Nagy ZS, Gatter KC, Esiri MM, Wass JAH, Harris AL. Proliferation, Bcl-2 expression, and angiogenesis in pituitary adenomas: relationship
to tumour behaviour. Br J Cancer. In press.
Landis JR, Koch GC. 1977 The measurement of observer agreement for categorical data. Biometrics. 33:159 174.
Weidner N, Carrol PR, Flax J, Blumenfeld W, Folkman J. 1993 Tumour
angiogenesis correlates with metastasis in invasive prostate carcinoma. Am J
Pathol. 143:401 409.
Macchiarini P, Fontanini G, Hardin MJ, Squartini F, Angeletti CA. 1992
Relation of neovascularisation to metastasis of non-small cell lung cancer.
Lancet. 340:145146.
Brem SS, Gullino PM, Medina D. 1978 Angiogenesis as a marker of preneoplastic lesions of the human breast. Cancer. 41:239 244.
Jensen-HM, Chen-I, DeVault-MR, Lewis-AE. 1982 Angiogenesis induced by
normal human breast tissue: a probable marker for precancer. Science.
218:293295.
Dobbs-SP, Hewett-PW, Johnson-IR, Carmichael-J, Murray-JC. 1997 Angiogenesis is associated with vascular endothelial growth factor expression in
cervical intraepithelial neoplasia. Br J Cancer. 76:1410 1415.
Turner HE, Moore NR, Byrne JV, Wass JAH. 1998 Pituitary, thyroid and
adrenal incidentalomas. Endocr Cancer. 5:131150.
Ferrara N, Henzel WJ. 1989 Pituitary follicular cells secrete a novel heparinbinding growth factor specific for vascular endothelial cells. Biochem Biophys
Res Commun. 161:851 858.
Sinha YN, Gilligan TA, Lee DW, Holligsworth D, Markoff E. 1985 Cleaved
prolactin: evidence for its occurrence in human pituitary gland and plasma.
J Clin Endocrinol Metab. 60:239 243.
Clapp C, Martial JA, Guzman RC, Rentier-Delrue F, Weiner RI. 1993 The
16-kilodalton N-terminal fragment of human prolactin is a potent inhibitor of
angiogenesis. Endocrinology. 133:12921299.
Ferrara N, Winer J, Henzel WJ. 1992 Pituitary follicular cells secrete an inhibitor of aortic endothelial cell growth: identification as leukaemia inhibitory
factor. Proc Natl Acad Sci USA. 89:698 702.
Delgrange E, Trouillas J, Maiter D, Donckier J, Turnaire J. 1997 Sex-related
difference in the growth of prolactinomas: a clinical and proliferation marker
study. J Clin Endocrinol Metab. 82:21022107.
Jeffcoate-WJ, Pound-N, Sturrock ND, Lambourne-J. 1996 Long-term follow-up of patients with hyperprolactinaemia. Clin Endocrinol (Oxf).
45:299 303.
Witt M, Klessen CH. 1987 Galactose and fucose binding sites in anterior
pituitary cells of the rat: detection by means of biotinylated lectins. Folia
Histochem Cytobiol. 25:115118.
Stanfield JP. 1960 The blood supply of the human pituitary gland. J Anat.
94:257273.
Schecter J, Goldsmith P, Wilson C, Weiner R. 1988 Morphological evidence
for the presence of arteries in human prolactinomas. J Clin Endocrinol Metab.
67:713719.
Powell DF, Baker LH, Laws ER. 1974 The primary angiographic findings in
pituitary adenomas. Radiology. 110:589 595.
Gorczyca W, Hardy J. 1988 Microadenomas of the human pituitary and their
vascularisation. Neurosurgery. 22:1 6.
Elias KA, Weiner R. 1984 Direct arterial vascularisation of estrogen-induced
prolactin-secreting anterior pituitary tumours. Proc Natl Acad Sci USA.
81:4549 4553.
Schechter J, Ahmad N, Elias K, Weiner R. 1987 Estrogen-induced tumours:
changes in the vasculature in two strains of rat. Am J Anat. 179:315323.
Elias KA, Weiner R. 1987 Inhibition of estrogen-induced anterior pituitary
enlargement and arteriogenesis by bromocriptine in Fischer 344 rats. Endocrinology. 120:617 621.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 August 2016. at 07:29 For personal use only. No other uses without permission. . All rights reserved.