Human Reproduction vol.12 no.7 pp.1577–1581, 1997
First trimester development of human chorionic villous
vascularization studied with CD34
immunohistochemistry
Elisabeth A.te Velde1, Niek Exalto2, Peet Hesseling3
and Hans C.van der Linden1,4
1Department
of Pathology, Academic Hospital Vrije Universiteit
Amsterdam, PO Box 7057, 1007 MB Amsterdam, 2Department of
Obstetrics and Gynaecology, Spaarne Hospital Haarlem,
3Department of Pathology, Spaarne Hospital Heemstede,
The Netherlands
4To
whom correspondence should be addressed
Normal chorionic villous vascularization is essential for the
undisturbed development of pregnancy. Defective vasculogenesis may play a role in pathological pregnancy. To assess
pathological chorionic villous vascularization, normal vascularization has to be defined first. Few data are available
on this topic. The aim of this study was therefore to
investigate normal chorionic villous vascularization in
ultrasound-dated first trimester pregnancies from week 5
menstrual age to week 12 (n J 41), using quantitative
CD34 immunohistochemistry. Two important processes
in chorionic villous vascularization were quantitatively
illustrated: (i) maturation, reflected by an increase of the
total number of luminized vessels as opposed to nonluminized haemangioblastic cords and (ii) margination,
due to a decrease of villous stromal area and an increase
of total villous vascular area. The percentage of villous
stromal area occupied by vascular elements (area difference
%) increased from 0.7% in week 5–2.5% in week 10.
Therefore, the area of the villous stroma occupied by
vascular elements increases and the vessels are situated
closer to the trophoblastic layer suitable for fetal–maternal
exchange. There was also a trend in increased number of
peripheral vessels (2.0 in week 5 to 4.6 in week 10),
supporting both developmental mechanisms. In conclusion,
in exactly dated normal human first trimester pregnancies,
development of the chorionic villous vascular system seems
to be mostly characterized by maturation of luminized
vessels from primitive haemangioblastic cords, and margination to a situation of peripherally located vessels.
Key words: abortion/CD34/chorionic villi/immunohistochemistry/vascularization
Introduction
Normal chorionic villous vascularization is essential for the
undisturbed development of pregnancy. On routine histological
sections, deficient chorionic villous vascularization was found
in cases of spontaneous abortion, e.g. blighted ova and macerated embryos (Meegdes et al., 1988).
© European Society for Human Reproduction and Embryology
To assess the pathology of chorionic villous vascularization,
normal vascularization has to be defined first. As reviewed by
Castellucci and Kaufmann in 1995, there is no systematic
report dealing with early chorionic villous vascularization in
human placental material (Castellucci and Kaufmann, 1995).
Only limited data are available on the human and macaque
situation based on electron microscopic studies. In these studies
on normal chorionic villous vascularization, the first embryonic
capillaries have been reported to appear between days 18 and
20 post-conception (p.c.) (Kaufmann and Castellucci, 1995).
Demir et al. (1989) identified haemangioblastic cell cords and
primitive capillary sprouts without blood cells at day 21–22
p.c., but blood cells were present in the capillary lumina only
at day 28 p.c. There were, however, no convincing signs of a
continuous vessel system and thus of functional embryonic
circulation. It seems that haemangioblastic cells locally differentiate from the mesenchyme and form cords or aggregates.
Capillary formation takes place by dilatation of intercellular
clefts. Nevertheless, in these studies no distinction was made
between pathological material and normal controls.
In haematoxylin and eosin (H&E) sections, haemangioblastic
cells can hardly be seen until they form strings of endothelium.
However, specific immunohistological markers may allow
identification of haemangioblastic cells at an earlier stage.
CD34 is a transmembrane protein, with an extracellular
region made up of an amino-terminal mucin domain and a
globular domain, which is related to immunoglobulin (Ig)
domains. It was first described as being expressed on the
earliest human haematopoietic progenitor cells. The pattern of
expression of CD34 structure suggests that it plays an important
role in early haematopoiesis (Sutherland and Keating, 1992).
The monoclonal antibody against CD34 could therefore provide
advantages over currently available antibodies reacting with
endothelial cells, such as factor VIII, CD31 and Ulex europaeus. CD34 immunohistochemistry was used to study early
chorionic villous vasculogenesis in a pilot study (Exalto and
te Velde, 1994). In six legally induced abortions and 15
pathological pregnancies the vascularization was described. In
this study, capillary formation was seen only in the presence
of an embryo. In blighted ova and molar pregnancies only
haemangioblastic cords were stained.
To allow optimal maternal–fetal exchange of oxygen and
nutrition between the intervillous maternal blood and the villi
at the functional transfer site, the mean maternal–fetal diffusion
distance must be as small as possible. In the chorionic villi,
peripheral vessels are therefore needed (Burton and Feneley,
1992; Kaufmann and Castellucci, 1995). The mechanisms of
progressive margination of the capillaries is a subject of
interest. The capillaries in the villus are positioned closer to
1577
E.A.te Velde et al.
Figure 1. Vessels and cords are seen stained with CD34, in stem
villi, as well as the remaining villi. Magnification 382.5. Menstrual
age 12 weeks.
Figure 2. Number of peripheral vessels in relation to menstrual age
during first trimester of pregnancy.
Figure 4. Mean villous stromal area in relation to menstrual age
during first trimester of pregnancy.
Figure 5. Percentage of villous stromal area occupied by vascular
elements in relation to menstrual age during first trimester of
pregnancy (area difference).
vascularization patterns in normal first trimester pregnancies
has been described.
The aim of this study was therefore to investigate normal
chorionic villous vascularization in exactly dated first trimester
pregnancies, using standardized quantitative CD34 immunohistochemistry.
Materials and methods
Figure 3. Total vascular elements in relation to menstrual age
during first trimester of pregnancy.
the villous surface with advanced maturation, since the mean
villous diameter decreases and the dynamics and volume of
the villous vessels change. No detailed quantification of villous
1578
Case selection
Records of patients who underwent legal abortion at the Department
of Gynaecology and Obstetrics at the Spaarne Hospital, locations
Haarlem and Heemstede, The Netherlands, in the years 1990–1991
were retrieved.
Ultrasound and clinical data were used to distinguish between
normal and defective embryonic development. Cases in which the
abortion was performed for medical reasons or with embryonic or
placental abnormalities were excluded. The crown–rump length (CRL)
and the dates of ultrasound examination and curettage were used to
calculate the age in days of menstrual age (MA) of the embryo.
Cases were grouped around complete weeks (63 days).
After fulfilling the criteria, a random representative sample was
Chorionic villous vascularization
Table I. Mean number per 10 villi of central (Vc) and peripheral (Vp) vessels, and mean number of central
(Cc) and peripheral (Cp) cords, and total number of vascular elements of the 41 patients studied, sorted on
menstrual age
Patient
number
Menstrual
age (week)
Menstrual age
(day)
Vc
Vp
Cc
Cp
Total
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
5
5
6
6
6
7
7
7
7
7
7
7
8
8
8
8
8
8
8
8
8
9
9
9
9
9
9
9
9
10
10
10
10
10
10
11
11
12
12
12
12
35
38
44
45
45
47
47
49
51
51
51
51
53
54
54
54
55
55
56
58
58
60
61
62
62
65
66
66
66
68
69
69
69
72
73
79
80
81
81
81
85
0.9
0.8
1.1
0.2
1.4
0.1
1.3
0.4
0.2
1.8
1.8
1.1
1.1
0.5
2.1
1.6
0.8
1
0.4
1.5
0.5
1
1.8
0.8
1
0.4
1.3
2
0.5
1.5
0.8
0.7
0.9
1.8
1.8
3.4
1.4
0.4
0.8
0.8
1.2
2.6
1.4
3.7
0.7
3.4
1.3
1.6
4.3
2.1
2.9
1.7
3.9
3.6
3.9
5.1
3.6
3
4.9
2.3
2.9
2.9
5.6
1.6
4
4.3
1.5
4
3.6
1
2.6
2.1
3.8
4.5
5.9
8.5
3.9
4.4
3.8
2.2
3.4
5.6
1.2
2.1
1
0.8
1.3
2.1
1.4
1.1
1.7
2.3
1.2
1.6
1.8
1.7
0.5
0.7
0.8
1.4
2.3
1.1
2
1.7
2.4
0.8
1.9
1.4
1.2
2.5
0.7
0.6
3.8
1.2
1.1
0.6
0.5
0.4
2.2
1.8
1.2
0.6
1.4
2.9
3
2.3
4.4
2.5
4.3
3.5
3
2.9
3.4
1.9
2
2.3
4
1.7
2.4
1.6
2.7
3.4
1.9
3
3.1
2.3
2.8
3.3
3
2.7
3.8
5.6
4.9
3.1
5.3
4.6
2.4
2.9
1.1
3.8
3
4
7.1
2.1
7.6
7.3
8.1
6.1
8.6
7.8
7.8
8.8
6.9
10.4
6.6
8.6
8.8
10.1
9.4
8.3
6.2
10
8.4
7.4
8.4
11.4
8.1
8.4
10.5
6.3
9.2
11.9
7.8
9.6
9.8
11
11.1
10.7
13.7
8.8
11.8
9
8.2
11.9
10.3
Table II. Mean number per villus of central (Vc) and peripheral (Vp) vessels, and mean number of central
(Cc) and peripheral (Cp) cords, and mean number of total vascular elements of the 41 patients studied,
sorted on menstrual week. Standard error (SE)
Menstrual age
(week)
Vc
SE
Vp
SE
Cc
SE
Cp
SE
Total
SE
5
6
7
8
9
10
11
12
0.9
0.9
1.0
1.1
1.1
1.3
2.4
0.8
0.1
0.6
0.7
0.6
0.6
0.5
1.4
0.3
2.0
2.6
2.5
3.6
3.2
4.6
4.2
3.8
0.9
1.7
1.2
0.9
1.6
2.4
0.4
1.4
1.7
1.0
1.6
1.4
1.6
1.3
1.3
1.3
0.6
0.3
0.5
0.6
0.7
1.3
1.3
0.5
3.0
3.0
3.0
2.6
3.3
3.9
2.5
4.1
0.1
1.2
0.9
0.8
1.0
1.2
1.9
2.2
7.5
7.6
8.1
8.6
9.2
11.0
10.3
9.9
0.2
1.3
1.3
1.2
1.9
1.5
2.1
1.6
drawn for every week of gestation; thus in 41 cases routinely
phosphate-buffered formalin fixed paraffin blocks were retrieved. The
age of the 41 patients undergoing legal abortion ranged between 17
and 43 years (mean 30). Duration of their pregnancies varied from 5
to 12 weeks MA. It was not possible to study vascularization in
normal human chorionic villi before the 5th week MA, because these
1579
E.A.te Velde et al.
Table III. Morphometrical data
Menstrual age
(week)
Villous stromal
area (µm2)
Standard
error
Vascular area
(µm2)
Standard
error
Area
difference %
5
6
7
8
9
10
11
12
18 548.5
19 868.3
15 986.8
14 745.0
14 453.0
10 517.8
10 470.0
9 520.8
2513.8
8013.4
3287.5
2860.7
3270.7
5960.0
4002.2
3708.9
135.5
177.0
225.0
313.7
325.3
252.8
226.5
166.0
51.6
7.0
78.0
133.5
92.1
98.0
136.5
52.5
0.70
0.90
1.40
2.10
2.25
2.50
2.26
1.70
patients did not present themselves to our clinic, or after the 12th
week, since no legal abortions were performed after this period.
Immunohistochemistry
Sections 4 µm thick were cut and mounted on 3-aminopropyltriethoxy-silane coated slides. Incubation with monoclonal mouseanti-CD34 antibody (Biogenex, San Ramon, CA, USA) was performed
at room temperature for 1 h, after blocking endogenous peroxidase.
Detection of the primary antibody was performed using biotinylated
rabbit anti-mouse antibody (DAKO A/S, Copenhagen, Denmark)
and streptavidin–biotin horseradish peroxidase complex (sABC/HRP,
DAKO A/S, Denmark). The peroxidase reaction was visualized using
diaminobenzidine/H2O2 (0.05% w/v/0.03% v/v).
Analysis of vasculogenesis
Slides were examined at a magnification 3400 (field diameter
450 µm) by one trained observer (E.A.T.V.), blind to the age of the
pregnancy. For each case ten mesenchymal or immature intermediate
villi (without stromal connective tissue fibres to rule out stem villi)
were evaluated. This number appeared to be sufficient to obtain stable
running means for the different variables assessed as described
below. No attempt was made to orientate the villi for sectioning or
morphometry.
In these, ‘cords’, defined as clusters of CD34-positive haemangioblastic cells without lumen formation, as well as ‘vessels’, defined
as clusters of CD34-positive cells with an obvious lumen, were
counted to depict the process of ‘maturation’ (Figure 1).
It was noted whether these cords and vessels were peripherally or
centrally located, to illustrate the process of margination. ‘Peripherally’ was defined as situated against the trophoblastic surface of the
villus. ‘Centrally’ was defined as without any connection to the
trophoblast.
Morphometrical analysis
Morphometrical measurements were performed using the QPRODIT
interactive video-overlay system (Leica, Cambridge, UK). The system
comprised an IBM-compatible microcomputer with a video overlay
board, a computer mouse and a charge coupled device colour camera
mounted on a standard light microscope. Contours of the stroma of
the previously mentioned 10 villi and all the included vascular
elements were traced manually on the computer monitor with a
mouse-controlled cursor at an on-screen magnification of 3600.
The following features were calculated: area of the villous stroma,
without the trophoblastic layer, and all the vessels and cords, as well
as the percentage of villous stroma occupied by vascular elements.
We only calculated the stromal area without the trophoblastic layer,
since the trophoblast develops in a special way, not described in
this study.
1580
Statistical techniques
Linear regression analysis was performed to assess correlations
between the age of the pregnancy on the one hand and the numbers
of central and peripheral cords and vessels, the areas of villous stroma
and the percentage of villous stroma occupied by vascular elements
on the other.
Results
Insight in the overall distribution of all data concerning vascular
cords, vessels and localisation is given in Table I; the means
of the measurements per 10 villi are given. The means of
these variables were calculated per week of menstrual age and
are shown in Table II. The mean number of central vessels
(Vc) per villus increased from 0.9 at 5 weeks up to maximally
2.4 at week 11. The mean number of peripheral vessels (Vp)
per villus increased from 2.0 in week 5 to 4.6 in week 10
(Table II, Figure 2). The central cords (Cc) appeared to be
stable at ~1.5 per villus. The peripheral cords (Cp) per villus
did not vary much over the period studied. Figure 3 shows
that the total number of vascular elements per villus gradually
increased with the duration of pregnancy, ranging from 7.5 at
5 weeks to .10 at 10 weeks and later (r 5 0.89, P 5 0.003)
(Table II).
The morphometrical results are shown in Table III. The mean
villous stromal area decreased with duration of pregnancy, from
~19 000 µm2 in week 5 to 9500 µm2 in week 12 (r 5 –0.96,
P , 0.001) as depicted in Figure 4. The mean villous vascular
area increased from about 135 µm2 in week 5 to 325 µm2 in
week 9. The percentage of villous stromal area occupied by
vascular elements (area difference %) increased from 0.7%
in week 5 to 2.5% in week 10 (Table III, Figure 5).
Discussion
Two different processes may be discerned in development of
blood vessels: vasculogenesis, defined as the development of
blood vessels from in-situ differentiating endothelial cells, and
angiogenesis, defined as the sprouting of capillaries from preexisting vessels (Rissau et al., 1988). We believe that it is the
former process that is mostly observed during the first trimester
in the development of chorionic villi. Therefore, to study early
vasculogenesis of chorionic villi, the primitive haemangioblastic cells from which the capillaries differentiate must be
Chorionic villous vascularization
studied. To this end, the use of a sensitive immunohistochemical
assay based on CD34 which stains these progenitor cells was
proven to be quite helpful.
In the vasculogenesis of chorionic villi, several processes
have been suggested such as maturation of vessels from
haemangioblastic cords to luminated capillaries and the margination of vascular elements. This study provides for the first
time quantitative data to support these observations.
Lumen formation is essential for the fetal–maternal oxygen
and nutrition exchange. In the normal first trimester chorionic
villi, the total number of vessels indeed increased, especially
the peripheral vessels (Tables I and II, Figures 2 and 3),
indicating the maturation of cords into vessels. Therefore,
since the vessels derive from cords, we expected the number
of cords to decrease with duration of pregnancy. In fully
matured placentae no cords are found, although some isolated
CD34 positive cells that are actually part of vessels may be
observed due to cutting artefacts, as we have been able to
show in a pilot experiment using serial slides (unpublished
results). In our study the number of cords up to week 12
appears to remain stable around 4.5 and did not evidently
decrease. Possibly, the decrease only starts in the second
trimester of pregnancy.
The stromal area decreased, together with an increase of
total vascular area, resulting in margination of the vessels, and
the number of peripheral vessels increased considerably (from
two to about four from the tenth week on) (Figure 2 and
Table II). This may result in a functional state, since the
principal site of fetal–maternal transfer is against the trophoblastic lining of the villi (Castellucci and Kaufmann, 1995;
Kaufmann and Castellucci, 1995).
The time of onset of fetal–maternal oxygen and nutrition
exchange has recently been the subject of discussion. Jauniaux
et al. did not demonstrate blood flow in the intervillous space
before the 12th week of pregnancy. An untimely initiation of
blood flow may be of importance, being the final mechanism in
cases of spontaneous abortion (Jauniaux et al., 1994; Jauniaux
1996). It has also been stated that these findings provide no
evidence for absence of intervillous blood flow in the first
trimester (Moll, 1995).
It seems that the embryo develops in a hypoxic environment,
and that placental function is linked with oxygen availability
in early pregnancy (Rodesch et al., 1992; Wheeler et al.,
1995). The fetal–maternal exchange from intervillous blood to
peripherally located luminized vessels can take place only
when there is a real blood flow in the intervillous space. Even
as early as in week 5, we found well-developed luminized
peripheral vessels. If there is no intervillous blood flow at that
time, the function of these vessels in maternal–fetal exchange
during first trimester of pregnancy may need to be further
investigated.
In routine pathology practice, correct histological classification of tissue from abortion proved to be worthwhile (Fox,
1993; Novak et al., 1990; Hustin et al., 1996). In spontaneous
abortion material in routine practice, deficient vascularization
may be found or even avascular villi are said to be seen
without immunohistochemical staining. Defective chorionic
villous vascularization is reflected by disturbed maturation
and margination and may play a role in the mechanism of
spontaneous abortion. The CD34 immunoquantification may
prove to be a sensitive method to assess the abnormal vasculogenesis in histological material of spontaneous abortions. This
will be the subject of further studies.
In conclusion, in exactly dated normal human first trimester
pregnancies, our data suggest that development of the chorionic
villous vascular system seems to be mostly (but not necessarily
exclusively) characterized by the maturation of luminized
vessels from primitive haemangioblastic cell cords, and margination of vessels due to decrease of villous stromal area and
the increase of the total vascular area.
Acknowledgements
The authors wish to thank Dr J.te Velde for his initiating work on
the subject and Dr P.J.van Diest for his advice and assistance with
this study.
References
Burton, G.J. and Feneley, M.R. (1992) Capillary volume fraction is the
principal determinant of villous membrane thickness in the normal human
placenta at term. J. Dev. Physiol., 17, 39–45.
Castellucci, M. and Kaufmann, P. (1995) Basic structure of the villous trees.
In Bernirschke, K. and Kaufmann, P. (eds), Pathology of the Human
Placenta. Springer-Verlag, New York.
Demir, R., Kaufmann, P., Castellucci, M. et al. (1989) Fetal vasculogenesis
and angiogenesis in human placental villi. Acta Anat., 136, 190–203.
Exalto, N. and te Velde, J. (1994) Vascular system development in early
human chorionic villi. Proceedings 10th Annual Meeting of ESHRE,
Brussels. Hum. Reprod., 9, 10.
Fox, H. (1993) Histological classification of tissue from spontaneous abortions:
valueless exercise? Histopathology, 22, 599–600.
Hustin, J., Kadri, R., and Jauniaux, E. (1996) Spontaneous and habitual
abortion – a pathologist’s point of view. Early Pregnancy Biol. Med., 2,
85–95.
Jauniaux, E. (1996) Intervillous circulation in the first trimester: the phantom
of the color Doppler obstetric opera. Ultrasound Obstet. Gynecol., 8, 73–76.
Jauniaux, E., Zaidi, J., Jurkovic, D. et al. (1994) Comparison of colour Doppler
features and pathological findings in complicated early pregnancy. Hum.
Reprod., 12, 2432–2437.
Kaufmann, P. and Castellucci, M. (1995) Development and anatomy of the
placenta. In Fox, H. (ed.), Obstetrical and Gynaecological Pathology.
Churchill Livingstone, New York, p. 1437.
Meegdes, H.L.M., Ingenhoes, R., Peeters, L.L.H. and Exalto, N. (1988) Early
pregnancy wastage: relationship between chorionic vascularization and
embryonic development. Fertil. Steril., 49, 216–220.
Moll, W. (1995) Invited commentary: absence of intervillous blood flow in
the first trimester of human pregnancy. Placenta, 16, 333–334.
Novak, R.W., Malone, J.M. and Robinson, H.B. (1990) The role of the
pathologist in the evaluation of first trimester abortions. Pathol. Am.,
297–311.
Risau, W., Sariola, H., Zerwes, H.G. et al. (1988) Vasculogenesis and
angiogenesis
in
embryonic-stem-cell-derived
embryoid
bodies.
Development, 102, 471–478.
Rodesch, F., Simon, P.H., Donner, C. and Jauniaux, E. (1992) Oxygen
measurements in endometrial and trophoblastic tissue in early pregnancy.
Obstet. Gynecol., 80, 283–285.
Sutherland, D.R. and Keating, A. (1992) The CD34 Antigen: structure,
biology, and potential clinical applications. J. Hematother., 1, 115–129.
Wheeler, T., Elcock, C.L. and Anthony, F.W. (1995) Angiogenesis and the
placental environment. Placenta, 16, 289–296.
Received on January 28, 1997; accepted on April 28, 1997
1581