university of groningen quality assessment of prenatal ...chapter 2.2 minimal volume of amniotic...
Post on 21-Feb-2021
1 Views
Preview:
TRANSCRIPT
University of Groningen
Quality assessment of prenatal cytogenetic diagnosisSikkema-Raddatz, Birgit
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.
Document VersionPublisher's PDF, also known as Version of record
Publication date:2005
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):Sikkema-Raddatz, B. (2005). Quality assessment of prenatal cytogenetic diagnosis: some guidelines forhandling amniotic fluid and chorionic villus material. s.n.
CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.
Download date: 14-07-2021
B.Sikkemaproefschrift2.doc
Quality assessment of prenatalcytogenetic diagnosis:
Some guidelines for handling amniotic fluid andchorionic villus material
B.Sikkemaproefschrift2.doc
The printing of the thesis was financially supported by
Cambrex Bio Science
Printing: Ponsen & Looyen BV, Wageningen
Cover design: Y. Swart
ISBN: 90-367-2206-3
© Copyright by B. Sikkema – Raddatz
B.Sikkemaproefschrift2.doc
RIJKSUNIVERSITEIT GRONINGEN
Quality assessment of prenatal cytogenetic diagnosis:
Some guidelines for handling amniotic fluid and chorionic villus
material
Proefschrift
ter verkrijging van het doctoraat in deMedische Wetenschappen
aan de Rijksuniversiteit Groningen
op gezag van de
Rector Magnificus, dr. F. Zwarts,
in het openbaar te verdedigen op
maandag 21 februari 2005
om 16.15 uur
door
Birgit Sikkema-Raddatz
geboren op 28 februari 1965
te Güstrow (Duitsland)
B.Sikkemaproefschrift2.doc
Promotor: Prof. dr. C.H.C.M. Buys
Copromotor: Dr. G.J. te Meerman
Beoordelingscommissie: Prof. dr. M.A. Ferguson-Smith
Prof. dr. M.J. Heineman
Prof. dr. N.J. Leschot
B.Sikkemaproefschrift2.doc
Habe nun, ach! Philosophie,
Juristerei und Medizin,
Und leider auch Theologie!
Durchaus studiert, mit heissem Bemühn.
Da steh ich nun ich armer Tor!
Und bin so klug wie nie zuvor;
Bilde mir nicht ein, was Rechts zu wissen,
Bilde mir nicht ein, ich könnte was lehren,
Die Menschen zu bessern und zu bekehren.
Auch hab ich weder Gut noch Geld,
Noch Ehr und Herrlichkeit der Welt;
Drum hab ich mich der Magie ergeben,
Dass ich nicht mehr, mit saurem Schweiss,
Zu sagen brauche, was ich nicht weiss;
Dass ich erkenne, was die Welt
Im Innersten zusammenhält.
(Aus Faust von J.W. Goethe)
Für Heike
B.Sikkemaproefschrift2.doc
B.Sikkemaproefschrift2.doc
Contents:
Chapter 1 Introduction...........................................................................................9
Chapter 1.1 Prenatal cytogenetic diagnosis………………………………………… 10
Chapter 1.2 Quality control and assessment in prenatal cytogenetic diagnosis… 16
Chapter 1.3 Scope of the thesis……………………………………………………… 27
Chapter 2 Cell culture of amniotic fluid cells .....................................................35
Chapter 2.1 Effect of variation in amniotic fluid specimens and of type of culturemedium on culture time and number of cell clones of primary in-situ cultures ofamniocytes ..............................................................................................................37
Chapter 2.2 Minimal volume of amniotic fluid for reliable prenatal cytogeneticdiagnosis..................................................................................................................49
Chapter 2.3 Absolute procedure to test the growth potential of medium in primaryamniotic fluid cell cultures and the influence of decreased oxygen tension............55
Chapter 3 Slide preparation of in-situ cell cultures from amniotic fluid cellsand chorionic villi ..................................................................................................71
Assessment of the influence of fixative, temperature, humidity and air flow onmetaphase quality in preparations from primary in-situ cultures of amniotic fluidcells and chorionic villi .............................................................................................73
B.Sikkemaproefschrift2.doc
Chapter 4 Chromosomal analysis in preparations from chorionic villi andamniotic fluid cell cultures ...................................................................................89
Chapter 4.1 A 46,XX,der(13;14)(q10;q10),+21 child born after a 45,der(13;14)(q10;q10) chromosomal finding in CVS...................................................................91
Chapter 4.2 Four years’ cytogenetic experience with the culture of chorionic villi..97
Chapter 4.3 Probability tables for exclusion of mosaicism in prenatal diagnosis..113
Chapter 5 Summary, future prospects, and suggested guidelines ...............121
Chapter 5.1 Summary………………………………………………………………… 122
Chapter 5.2 Future prospects…………………………………………………………124
Chapter 5.3 Suggested guidelines……………………………………………………129
Samenvatting .......................................................................................................135
Curriculum Vitae..................................................................................................145
List of publications..............................................................................................147
Dankwoord ...........................................................................................................149
Introduction
B.Sikkemaproefschrift2.doc
9
Chapter 1
Introduction
Chapter 1
B.Sikkemaproefschrift2.doc
10
1.1 Prenatal cytogenetic diagnosis
In many parts of the world prenatal diagnosis has been part of antenatal care for
already thirty to forty years. The purpose of prenatal diagnosis is to determine
whether a foetus, believed to be at risk for a certain disorder, actually has that
disorder and to inform the parents about the possible course, seriousness and
treatment of the disorder. On the basis of that information they can decide to
continue the pregnancy or not. In the Netherlands prenatal diagnosis is performed
in 5% -10 % of all pregnancies (Niermeijer et al., 1997; Nagel et al., 2004). Women
with an increased risk of bearing a foetus with a chromosomal abnormality can opt
for a prenatal cytogenetic diagnosis. The purpose is to detect or exclude a
chromosomal abnormality of the foetus. For that, sufficient metaphase
chromosomes of an appropriate quality are needed. Current routine procedures to
obtain cells for prenatal cytogenetic diagnosis are amniocentesis (AC) and
chorionic villus biopsy (CVB). In the Netherlands, 8370 ACs and 3230 CVBs were
performed in 2002 (Dutch Prenatal Working Party, 2002). Thus, second trimester
amniocentesis, normally performed between 16 – 18 weeks of pregnancy, when 15
–20 ml amniotic fluid is drawn under direct ultrasound guidance, is the preferred
technique in most centres for prenatal diagnosis. There is an estimated procedure-
related risk of abortion of 0.5%- 1% (NICHD, 1976). Because of the introduction of
first trimester serum screening combined with an early cytogenetic analysis in
Groningen, CVB is there, in contrast to other centres, the most used invasive
prenatal procedure for cytogenetic diagnosis. CVB is usually performed
transcervically or transabdominally between 10 –12 weeks of pregnancy, when
under ultrasound guidance 20 –100 mg chorion villi are obtained for karyotyping.
There is an estimated risk of procedure related abortion of 0.5% - 1.0% (Brambati
et al., 1991), i.e. similar as the risk associated with AC.
Introduction
B.Sikkemaproefschrift2.doc
11
1.1.1 Procedure of cytogenetic diagnosis on amniotic fluid cells
Amniocentesis was first used in Germany in the early 1880s to treat hydramnions
(Lambl, 1881). In the 1950s amniotic fluid was used for monitoring bilirubin levels
in foetuses with Rhesus isoimmunization (Bevis, 1952). This development was
followed by the determination of foetal sex (Fuchs and Riis, 1956) in amniotic fluid
(AF) cells. The first chromosomal analysis was performed in 1966 by Steele and
Breg, and the first chromosomal abnormality, a balanced translocation, was
detected by Jacobson and Barter in 1967. In 1968, both Valenti et al. and Nadler et
al. reported in utero detection of Down syndrome, with confirmation after elective
abortion. The relative safety and accuracy of the procedure was demonstrated in a
study on 142 patients during 155 pregnancies (Nadler and Gerbie, 1970). In the
same period, AF was used for prenatal detection of biochemical abnormalities
(Galjaard, 1972). Brock and Sutcliff (1972) found an association of a raised
concentration of alpha- foetoprotein in AF and open neural tube defect.
AF is suggested to be a mixture of transudate of the maternal plasma,
urine and other body secretions from the foetus, accumulating into the AF space
(Behrman et al., 1967). It contains cells that can be precipitated by low-speed
centrifugation. The cell pellet consists of at least three cell types that can be
morphologically distinguished in culture; fibroblast-like (F) cells, epithelioid (E) cells
and amniotic fluid (AF) type cells (Hoehn et al., 1975). The precise foetal origins of
these cells have still not been resolved. AF cells are thought to originate from
trophoblast tissue as it has been demonstrated that these cells produce chorionic
gonadotropin (Laundon et al., 1981), a product of trophoblast tissue. E- type
cultured cells probably come from foetal skin and AF- type cells are from
mesenchymal origin. Also kidney tissue might contribute to cells in amniotic fluid,
since it has been implicated as a source of trisomy 20 cells (Hsu el al., 1991). For
cell culture either the in-situ method or the flask method are applied as basic
techniques. For in-situ culture, cells are grown on coverslips and harvested in-situ,
on average after 6-14 days. The in-situ harvest method has first been described by
Cox et al. (1974), who used clones of AF cells grown and harvested in Petri dishes.
Chapter 1
B.Sikkemaproefschrift2.doc
12
The sides of the dish were removed and the bottom on which the cells were
growing was taped on a slide for microscopy. Peakman et al. (1977) modified this
method and grew cells on coverslips inside Petri dishes. The coverslips were
‘harvested in-situ’ in the dishes, removed and attached to slides for microscopy.
Cells cultured according to the flask method are harvested from a cell suspension,
which is obtained by trypsinisation of the cells growing attached to the bottom of
the flask. Since a proportion of cells is lost in this procedure, more cells are
needed, which implies a longer culture time.
Whatever the culture method, before harvesting cultures are treated
overnight with the metaphase arrester colcemid to obtain adequate numbers of
cells in metaphase. Hypotonic treatment is performed the following day and cells
are fixed several times with a mixture of methanol and glacial acetic acid. At last,
the fixative is removed and the cells are allowed to dry under specific climate
conditions to achieve well-spread metaphase chromosomes. After "ageing" either
overnight or for a short period at 900C, the slides are ready for histochemical
treatment in order to produce discriminating banding patterns on the
chromosomes.
From the stained slides a certain number of metaphases will be analysed. For
preparations from in-situ cultures a single metaphase per clone is used for
analysis.
1.1.2 Procedure of cytogenetic diagnosis on chorionic villi
In 1968, Mohr reported a new method as an alternative to second trimester
amniocentesis. In this method chorionic villi were aspirated transcervically. A test
series was made prior to elective abortion. Only half of the samples, however, were
found to contain chorionic tissue. Complications as puncture of the amniotic sac
and membranes aspirated instead of villi were encountered. Reports on larger
series (Hahnemann, 1974; Kullander and Sandahl, 1973) had to contend with
maternal bleeding and infections, absence of real time ultrasound and low success
Introduction
B.Sikkemaproefschrift2.doc
13
rate of chorionic villus cell culture. Still, interest in first trimester diagnosis and
improvement of ultrasound machines resulted in 1982 in the first report of placental
biopsies using ultrasound guidance (Kazy et al., 1982), followed by a publication of
Old et al. (1982) on genetic diagnosis of hemoglobinopathies. Simoni et al., (1983)
developed a method for a rapid, direct preparation of chromosomes that took
advantage of spontaneous mitoses in the rapidly dividing villi. This direct method
yielded a cytogenetic diagnosis within a couple of hours after sampling. Moreover,
maternal cell contamination, often encountered in chorionic tissue culture, was not
a problem anymore.
Chorionic villi have a central mesenchymal core with foetal blood vessels.
This core is covered by a trophoblastic epithelium which consists of two layers; an
inner layer of single cells, which is called the cytotrophoblast and a
"multinucleated" outer layer, which is called the syncytiotrophoblast. The
trophoblast outer layer already exists four to six days after fertilisation as a layer
that surrounds the blastocyst. After embedding of the blastocyst in the
endometrium, trophoblast cells will fuse, forming the syncytiotrophoblast layer.
Non-fusing trophoblast cells will grow into this syncytiotrophoblast layer. Stromal
cells originating from the extraembryonal mesoderm will invade the cytotrophoblast
after about 13 days. From this stage on, new placental tissue is formed by
proliferation and branching (Figure 1).
After aspiration of the chorionic villi, the biopsy is checked under a
microscope for quality and quantity of the cell material and carefully cleaned from
maternal decidua. Chorionic villi can either be cultured or used for direct
preparation.
Chapter 1
B.Sikkemaproefschrift2.doc
14
Figure 1: Early embryonic development, according to Crane and Cheung, 1988
Direct preparation of chorionic villi cells
For direct preparation, cells from the cytotrophoblast are used. The great
advantage is that culturing of trophoblast cells is not necessary for getting
metaphase cells. Thus, a cytogenetic result can be rapidly obtained. Colcemid is
added to the villi to obtain adequate numbers of cells in metaphase. After
approximately one hour, hypotonic treatment is performed and chorionic villi are
fixed by methanol and glacial acetic acid. The cells are released from the villi by a
brief exposure to 60% acetic acid. The suspension is spread over a slide using a
bent tip of a Pasteur pipette (Simoni et al., 1983). Since no culturing is involved, an
effect of maternal cell contamination is limited to a minimum. At least 10 mg
chorionic villi are needed. However, the banding quality of the chromosomes is
poor and sometimes not enough metaphases are available. Therefore, as an
Introduction
B.Sikkemaproefschrift2.doc
15
alternative to this method, a semi-direct preparation or short-term culture of
chorionic villi is used. The difference between direct and semi-direct methods is a
24h incubation of chorionic villi in fluorodeoxyuridine (Gibas et al., 1987). This
causes a block in DNA synthesis, which is subsequently released by adding
thymidine at high concentrations as well as colcemid, so that synchronisation is
reached. Although theoretically this should result in a higher yield of metaphases
and an improvement of the morphology of the chromosomes, the overall quality
improvement, however, appears to be minimal.
Culturing of chorionic villi cells
For culturing of chorionic villi, the mesenchymal core is used. After the first
attempts to culture trophoblast cells (Kullander et al., 1973), culture methods were
improved considerably by the introduction of hormone-supplemented medium
(Chang et al., 1982; Czepulkowski et al.,1986). To avoid maternal cell
contamination, the outer layer of chorionic villi is stripped with trypsin.
Mesenchymal cells are further desegregated by collagenase (Jackson, 1989) and
the resulting suspension is cultured. A cytogenetic diagnosis can be made after 7-
10 days. Very few mg of chorionic villi are sufficient for successful culturing.
However, for quality reasons at least four cultures in two different incubators
should be initiated. The percentage of maternal contamination turned out to be
relatively low, between 0.1% to 2.16% (Roberts et al., 1988; Ledbetter et al., 1990
and 1992). By the application of a long-term culture technique problems involved in
the direct method, namely poor banding quality and a low number of metaphases,
can be overcome.
For harvesting chorionic villus cultures, either the in-situ or the suspension
method can be used. The harvest procedure, as well as the procedures for slide
making, staining and analysis, are in principle the same as described for AF cell
cultures.
Chapter 1
B.Sikkemaproefschrift2.doc
16
1.2 Quality control and assessment in prenatal cytogenetic diagnosis
Quality assessment in cytogenetics is still somewhat in its formative stage. The
best developed quality control and assurance programmes can probably be found
in the United Kingdom and the United States, where every cytogenetic laboratory
must participate in external proficiency testing (Watson, 1997). This includes
inspection of the facilities and making cytogenetic diagnoses of different test
samples every other year. As a consequence, only qualified and certified personnel
are permitted to sign chromosome analysis reports. Moreover, in the United
Kingdom, the Association of Clinical Cytogeneticists (2001) has formulated
guidelines for cytogenetic diagnosis. In the Netherlands a report “Advice
concerning Genetic Counselling” published in 1977, described for the first time the
indications for post- and prenatal cytogenetic diagnosis and the requirements for
personnel and equipment. A report from the Working Party of the Dutch
Association of Clinical Cytogeneticists (Mellink et al., 2000) was published in 1995
and revised in 2003, formulating minimal standards for cytogenetic diagnosis. A
minimum number of metaphases for analysing and karyotyping, a minimum
number of chromosome bands, a maximum turn around time for different
indications and a minimum qualification of the personnel are prescribed. These
standards are, however, insufficient to cover all quality aspects of the cytogenetic
process. In particular, aspects of culturing and harvesting are missing in these
guidelines. Factors involved in handling material for cytogenetic diagnosis have
also not been spelled out, nor has the effect of such factors on cytogenetic quality
been considered. Nevertheless, such factors are essential in determining the
quality needed for adequate analysis. Therefore, the following gives an overview of
which factors are involved and what is their effect on quality during the process of
cell culturing, slide preparation and chromosomal analysis.
Introduction
B.Sikkemaproefschrift2.doc
17
1.2.1 Quality control of culturing AF cells and chorionic villi
Culture quality can be measured as the culture time required for slide preparation,
as the number of metaphases on a slide, as the number of culture failures and
more specifically for AF fluid cells; as the number of clones. Culture quality is
mainly dependent of the operator who performs AC or CVB, on the condition of the
submitted material and on the culture conditions. Although these aspects count for
AF cell cultures as well as for chorionic villus cultures, we will focus on the culture
of AF cells, because this has been more extensively investigated.
With respect to the operator, different studies (Silver et al., 1998; Horger et
al., 2001; Blackwell et al., 2002) have demonstrated that the caseload influences
the sampling efficiency and that successful aspiration of clear AF increases with
experience in amniocentesis. Therefore, only an experienced obstetrician should
perform amniocentesis (Elias et al., 1998). Failure to obtain AF is then less than 1
percent (Tabor et al., 1986). In order to be allowed to perform amniocentesis an
obstetrician in the Netherlands has to perform at least 30 amniocenteses per year
(Dutch Society of Obstetrics and Gynaecology, 1997). However, neither the
influence of the caseload of the operator on culture quality, nor the minimal amount
of AF necessary for a reliable cytogenetic diagnosis have been investigated. Well-
defined guidelines for the transport of AF (at which temperature and for how long)
are also missing.
For the submitted material, the gestational age at which the sample is
taken, the amount and appearance of AF as well as the reason for sampling might
influence the quality of culture. Analysis of the effect of these variations might
retrospectively explain culture failures or delay in culture time. Some publications
have appeared in which these aspects were investigated in relation to culture
quality. Although the total number of cells in AF increases with gestational age, the
number of viable cells decreases (Elejalde et al., 1990). Moreover, only a fraction
of these cells will form clones, on average no more than 10 cells/ ml AF (Hoehn et
al., 1974). At 7 to 9 weeks of gestation 7.7 to 12.2 clones were counted per ml AF
(Kennerknecht et al., 1992); at 14 to 16 weeks 3.1+/- 1.2 clones and at 24 to 32
weeks less than 1.5 clones were counted (Rischkind and Risch, 1990).
Chapter 1
B.Sikkemaproefschrift2.doc
18
A red or brown coloured appearance of AF due to the presence of haemoglobin
negatively influences cell culture quality according to Golbus et al. (1979), who
observed a higher percentage of foetal demise and culture failures among these
samples. In a series of Hankins et al. (1984), however, no advanced risk on
pregnancy nor foetal problems were detected. Chang and Jones (1991) as well as
Seguin and Palmer (1983) reported inhibition of cell growth in red coloured AF.
Persute and Lenke (1995) concluded that the risk of culture failure is higher in
cases of foetal aneuploidy. However, Reid et al. (1996) substantiated further that it
was not merely a higher percentage of foetal aneuploidies, but more in general the
reason of referral “structural abnormalities at ultrasound”, resulting in a higher
percentage of culture failures. However, for none of the above-mentioned
variations of the submitted AF the effect of these variations on cell culture quality
has been systematically investigated.
In most laboratories the in-situ method is used because of its reduced turn-
around time compared to the flask method. Furthermore, after harvesting one is
still dealing with clones, which is preferable for analysis. Various protocols describe
how many cultures need to be initiated, which type of culture medium has to be
used and what should be further environmental conditions during culturing (Priest
and Rao, 1997; Keagle et al.,1999; Saunders and Czepulkowski, 2001). Cell
culture quality has to be monitored at least by culture time and number of culture
failures. On average, 0.5% culture failure is considered as acceptable (Lam et al.,
1998).
Crucial for success and minimal duration of AF cell cultures is the (type of) culture
medium and its supplements, in particular the foetal calf serum. Mostly used are
commercial media, such as Chang medium, Amniochrome or Amniomax that
include growth factors and foetal calf serum. The growth potential of different lots
or types of medium can differ. To minimise culture failures, use of two different (lots
or types) of medium per sample is recommended, as well as culturing in separate
incubators. All the used reagents, particularly each lot of medium and serum, need
to be tested on sterility and ability to support growth (Priest and Rao, 1997).
However, an absolute and easy to use procedure for such testing is not available.
Introduction
B.Sikkemaproefschrift2.doc
19
In order to control all aspects that influence culture quality, protocols for collecting
AF and chorionic villi, guidelines for handling these and protocols for culturing are
necessary. To define such guidelines, knowledge and registration of the potential
influence of different factors on culture quality is essential. Some of these factors
have been well investigated and some guidelines are already existing. This is e.g.
the case for the influence of the operator on the material which is submitted and for
some factors during culturing. A systematic investigation of the effect of quality
variations in the submitted material on culture quality is missing, however, as are
guidelines for transport of the material and for testing the growth potential of the
culture medium. Therefore, there is a strong urge for investigating these factors in
order to formulate guidelines.
1.2.2 Quality control during slide preparation
Slide preparation quality can be measured as the degree of spreading of the
metaphase chromosomes and the presence or absence of cytoplasm over the
metaphases. It will likely depend mainly on conditions during harvesting (hypotonic
treatment and composition of fixative) and on climate conditions during the process
of slide preparation. Slide preparation is a rather complex process. Individual
laboratories have their own “tricks” to obtain appropriate metaphase quality.
Several authors have investigated factors involved in this process. Claussen et al.
(2002) found hypotonic treatment to be essential for obtaining well-spread
metaphase cells. Without such a treatment, chromosomes become preferentially
positioned close to each other in the cell centre. The slide preparation process
starts with a water-induced swelling of mitotic cells during evaporation of the
fixative from the slide (Claussen et al., 2002). Methanol from the fixative
evaporates early during drying. This is followed by a water-induced swelling
through the hygroscopic effect of the gradually evaporating acetic acid and leads to
a stretching of the chromosomes via their flattening (Hliscs et al., 1997). For this
process a relative humidity of at least 21% is needed, otherwise no swelling will
occur. Thus, from these studies it may be concluded that hypotonic treatment,
Chapter 1
B.Sikkemaproefschrift2.doc
20
evaporation of acetic acid and a certain percentage of relative humidity are
essential for well-spread metaphases.
Spurbeck et al. (1996) determined the optimal climate conditions for slide
preparation from cells from suspension and secondary AF cell cultures and found
that the largest metaphase area was reached at 55% relative humidity and a
temperature of 200C. Deng et al. (2003) demonstrated a strong influence of relative
humidity on spreading of cells from suspensions. However, in none of these
studies the effects of the investigated factors have been determined quantitatively.
Furthermore, in-situ cultures, for which the slide preparation process is much more
complicated than for cells from suspension due to variable retention of nuclear
matrix proteins over the metaphase plate, have not been investigated in these
studies. For primary in-situ cultures effects of different variables during harvesting
and slide making need, therefore, to be investigated in order to define optimal
harvest conditions and to establish rational guidelines.
1.2.3 Quality control of chromosomal analysis
The quality of analysis can be measured as sensitivity and specificity of diagnosis.
This in turn mainly depends on the banding resolution of the chromosomes and on
the number of analysed metaphases. The chromosomal analysis will reveal or
exclude structural and numerical abnormalities (in case of mosaicism with a given
likelihood for exclusion).
Some minimal banding resolution is one quality criterion for an accurate
analysis. The International System for Human Cytogenetic Nomenclature (ISCN)
defined a total number of bands in human metaphases in 1995. In the Netherlands,
the Working Party of the Dutch Association of Clinical Cytogeneticists formulated a
minimal banding quality for cytogenetic diagnosis. For prenatal diagnosis, the
minimum resolution has been determined at 400 bands (Mellink et al., 2003). This
is in general the resolution of metaphases from chorionic villus cultures.
Metaphases from direct preparations of chorionic villi usually have less than 400
bands. In metaphases from AF cell cultures up to 550 bands are visible.
Introduction
B.Sikkemaproefschrift2.doc
21
A chromosomal abnormality which is detected in only a proportion of the
investigated cells, is called chromosomal mosaicism. A mosaic pattern results from
an abnormal cell division, which leads to the presence of two or more cells with an
identical abnormality among the other (normal) cells. This can either reflect a true
foetal mosaicism or be confined to the extra embryonic tissue or represent a
culture artefact (pseudomosaicism). Pseudomosaicism is believed to usually
originate in vitro, without clinical consequences. Incidence and interpretation of
mosaicism is difficult and different for AF cells and chorionic villi.
Chromosomal analysis in AF cell cultures
Mosaicism can be excluded to a certain degree by analysing a given number of
metaphases. Hook (1977) published the number of cells that should be analysed
from peripheral blood cultures for exclusion of 10%, 20% and 30% mosaicism at
90%, 95% and 99% confidence levels, as derived from Newton's binomial formula.
Based on this model also the number of metaphases to be analysed from in-situ
cultures and flask cultures in order to exclude mosaicism in AF cell cultures were
similarly determined (Claussen et al., 1984; Richkind and Risch, 1990;
Featherstone et al., 1994). In current practice, for the flask method 20 cells from at
least two flasks and for the in-situ method 10 –15 cells from different clones should
be analysed. These guidelines allow exclusion of more than 20% mosaicism with a
confidence level of 95% to 99% (Rischkind and Risch; 1990).
In terms of a laboratory result, mosaicism can be classified at three
different levels (Hsu et al., 1992; Hsu and Benn, 1999): (1) a single cell is
abnormal; (2) two or more cells in the flask method, a single clone or two or more
clones within the same culture dish in in-situ cultures are abnormal; (3) two or more
cells/ clones from independent cultures are abnormal. Only the last situation is
considered to reflect true mosaicism. In AF cell cultures true mosaicism is rare with
an occurrence in 0.1% to 0.3 % of samples (Bui et al., 1984; Hsu; 1984; Worton
and Stern, 1984). The mosaic pattern is confirmed in the foetus in about 70% of
the cases. Pseudomosaicism of level (1) occurs in 2.47% to 7.10% and of level (2)
Chapter 1
B.Sikkemaproefschrift2.doc
22
in 0.64% to 1.10 % of cases, including more structural than numerical
abnormalities. In order to discriminate pseudomosaicism from a potentially true
mosaicism one can increase the number of analysed metaphases by analysing
metaphases in a culture dish different from the one in which the abnormal cells/
clones were present. Hsu et al. (1992; 1999) developed guidelines how to deal with
different chromosomal abnormalities. Depending on which abnormality was
detected, no analytical work up, moderate (12 extra clones) or extensive work up
(24 extra clones) should be done on metaphases in other dishes than the initial
dish in which the abnormal cells/ clones had been found (Table 1). The use of
these guidelines results in a reliable diagnosis of pseudomosaicism or true
mosaicism. In case of a work up sometimes, however, an insufficient number of
metaphases is available on the regularly harvested in-situ dishes and a back-up
culture needs to be trypsinised and analysed. In such a situation guidelines on how
many metaphases should be analysed from trypsinised back-up cultures are
missing.
Chromosomal analysis in chorionic villi
True mosaicism in chorionic villi can be confined to the placenta, to the foetus or
be present in both foetus and placenta, i.e. be a generalised mosaicism. Kalousek
and Dill (1983) demonstrated mosaicism confined to extra-embryonic tissue in two
out of 46 spontaneous abortions. Although foetus and placenta originate from the
same zygote, their chromosomal constitution can be different. Soon after the
introduction of CVB in routine practice, it turned out that the cytogenetic diagnosis
from direct preparations did not always reflect the chromosomal constitution of the
foetus. Reports appeared on false positive results (Bartels et al., 1986; Breed et
al., 1986; Leschot et al., 1987), followed by reports on false negative results
(Eichenbaum et al., 1986; Linton et al., 1986; Martin et al., 1986). Collaborative
studies (Ledbetter et al., 1992; ACC, 1994; Hahnemann and Vejerslev,1997)
showed mosaicism confined to the placenta in 1% to 2% of all cases.
Introduction
B.Sikkemaproefschrift2.doc
23
Table 1: Guidelines for work-up to exclude or detect pseudomosaicism or
mosaicism, from Hsu and Benn (1999)
Flask method In-situ method
A. Indications for extensive work-up1. Autosomal trisomy involving achromosome 21, 18, 13 or 2; 5, 8, 9, 12, 14, 1516, 20, 22 (SC, MC)
1. Autosomal trisomy involving a chromosome21, 18, 13 or 2; 5, 8, 9, 12, 14, 15 16, 20, 22(SCl, MCl)
2. Unbalanced structural rearrangements (MC) 2. Unbalanced structural rearrangements (MCl)
3. Marker chromosome (MC) 3. Marker chromosome (MCl)
B. Indications for moderate work-up4. Extra sex chromosome (SC, MC) 4. Extra sex chromosome (SCl, MCl)
5. Autosomal trisomy involving a chromosome 1,3, 4, 6, 7, 10, 11, 17, 19 (SC, MC)
5. Autosomal trisomy involving a chromosome 1,3, 4, 6, 7, 10, 11, 17, 19 (SCl, MCl)
6. 45,X (MC) 6. 45,X (SCl, MCl)
7. Monosomy (other than 45,X) (MC) 7. Monosomy (other than 45,X) (SCl, MCl)
8. Marker chromosome (MC) 8. Marker chromosome (MCl)
9. Balanced structural rearrangement (MC) 9. Balanced structural rearrangement (MCl)
C. Standard, no additional work-up10. 45,X (SC) 10. Unbalanced structural rearrangements (SCl)
11. Unbalanced structural rearrangements (SC) 12. Break at centromere with loss of one arm(SC)
13. Break at centromere with loss of one arm (SC) 13. All single- cell abnormalities
MC: multiple cells (single flask), MCl: multiple clones (single dish), SC: single cell, SCl: single clone
A model of Crane and Cheung (1988) can explain the discrepancy
between the chromosomal constitutions of foetus and placenta. According to their
model (Figure 2), from the first five cell divisions following fertilisation only two of
the 32 resulting cells will give rise to the formation of the embryo. All other cells are
responsible for the formation of the trophoblast and the extraembryonic tissues.
Chapter 1
B.Sikkemaproefschrift2.doc
24
Figure 2: Diagram of cell lineages arising from differentiation after the first five
divisions in a human embryo, according to Crane and Cheung (1988)
A possible failure (nondisjunction) in an early cell division may lead to an
abnormal cell in a specific cell line. A false positive result can be explained when in
a normal foetus a localised nondisjunction in the chorionic villi occurs, which may
cause an aberrant cell line in the direct preparation or culture, but not in the foetus.
A false negative result can be explained by the occurrence of a meiotic
nondisjunction, which may lead to a trisomic fertilised ovum, and a subsequent
mitotic nondisjunction in the chorionic villi leading to a normal diploid cell line in the
placenta, whereas the foetus remains fully trisomic.
Based on the position of the different cell lines and on the chromosomal
constitution of the foetus six major categories of mosaics can be distinguished
(Table 2).
Introduction
B.Sikkemaproefschrift2.doc
25
Table 2: Types and frequencies of placental-fetal discordance at chorionic villussampling, modified from Hahnemann and Verjerslev (1997)
Karyotype
Relativefrequencies (%)
Type Nature Trophoblast Villus core Embryo/foetusI CPM Abnormal Normal Normal 48
II CPM Normal Abnormal Normal 25
III CPM Abnormal Abnormal Normal 16
IV TFM Abnormal Normal Abnormal 0
V TFM Normal Abnormal Abnormal 4
VI TFM Abnormal Abnormal Abnormal 7
Trophoblast, preparations from direct/ short term culture; Villus core, preparations from long term
culture; CPM, confined placental mosaicism; TFM, true foetal mosaicism (generalised foetal-placental
mosaicism). An abnormal karyotype could be mosaic or non-mosaic. A normal embryonic/ foetal
karyotype could reflect normal biparental disomy or uniparental disomy.
This table also shows the relative frequencies of the different types of mosaicism
(Hahnemann and Vejerslev, 1997). The results of direct preparation alone (type I)
were less reliable than those of the method using culturing (type II). More cases of
mosaicism confined to the placenta were detected in the direct preparation alone,
which would in a number of cases lead to unnecessary amniocentesis as follow-up.
False negative cases occur at a very low frequency (Saura et al., 1998), but
(nearly) all of the cases were found in direct preparations (type IV versus type V).
The model of Crane and Cheung (1988) can also explain these differences.
Trophoblast cells (used in direct preparation) are forming the first line of
differentiation from totipotent cells of the morula and have therefore a greater
chance of failures of cell division. Mesodermal core cells (used in culture) reflect
the later segregating line of the extraembryonic mesoderm. These cells are closer
related to the cells that form the embryo than trophoblast cells. Therefore, the use
of a direct preparation alone should be avoided and reliability of the cytogenetic
diagnosis improved by the combined use of a direct preparation and a long-term
culture (Hahnemann and Vejerslev, 1997).
Chapter 1
B.Sikkemaproefschrift2.doc
26
Once an abnormality is detected, the question rises of the probability, i.e. a
calculated predictive value, of an abnormal foetal karyotype. In general, the
predictive value is high in case of a non-mosaic abnormality found in culture (83%-
100%), low in case of a mosaic abnormality in culture (24%) and zero in case of
a(n) (mosaic) abnormality detected only in a direct preparation (Hahnemann and
Vejerslev, 1997). Of special interest is the predictive value of a particular
chromosome abnormality for adequate counselling of the parents about follow-up
options, such as amniocentesis, ultrasound investigation or termination of
pregnancy. However, because of the small number of aberrant cases for most
chromosomes, predictive values are imprecise and additional cases are needed to
improve that.
Introduction
B.Sikkemaproefschrift2.doc
27
1.3 Scope of the thesis
The aim of prenatal cytogenetic diagnosis is to identify a possible chromosomal
abnormality of the foetus. The most important quality criteria for diagnosis are
sensitivity, i.e. the overall detection rate of abnormalities, and specificity, i.e. true
occurrence of the detected abnormality in the foetus. For a reliable result,
sensitivity and specificity need to be as close to 100%. Moreover, a result should
be given within the shortest time possible on a minimum of material and at the
lowest costs. To realise this, knowledge of all factors involved in the process of
diagnosis is needed, as is a determination of their effects. Based on this
knowledge, guidelines for handling prenatal material can be established. In
particular aspects of culturing and harvesting are missing in the previously
described cytogenetic procedures and factors involved in handling prenatal
material have not been spelled out nor has the effect of these factors on
cytogenetic quality been considered. This is particularly important in prenatal
cytogenetic diagnosis, where because of the risk for both the pregnant woman and
the foetus in obtaining the required material, the material is very precious.
Moreover, in case of a detected chromosomal abnormality, a following decision to
terminate or continue the pregnancy needs to be based on a reliable diagnosis.
The scope of the thesis is, therefore, to establish guidelines for quality controlled
processing of amniotic fluid cells and chorionic villi for purposes of prenatal
cytogenetic diagnosis. We will investigate which factors are of critical importance in
handling these materials and will quantify their influence. Depending on their
influence we will advice which factors need to be monitored carefully and we will
develop some new guidelines in case the existing ones are insufficient. Our goal is
to determine the limits between which reliable prenatal cytogenetic diagnosis is
possible. For this purpose we divide the cytogenetic process in three distinct main
parts, namely cell culture, slide preparation and chromosomal analysis.
Chapter 1
B.Sikkemaproefschrift2.doc
28
Culture quality can be measured as the culture time required for slide
preparation for analysis and for AFcell culture more specifically as the number of
clones and as the number of culture failures. The culture quality depends mainly on
the condition of the submitted material and on the culture conditions. Therefore, we
will first investigate the effect of variations in the submitted amniotic fluid (i.e.
volume, appearance) and in culture medium on cell culture quality (Chapter 2.1).
Second, we want to determine the minimal volume of AF required for a reliable
cytogenetic diagnosis, avoiding repeated amniocentesis (Chapter 2.2). Finally, we
want to develop an absolute procedure for testing the growth potential of reagents
involved in culturing and of other factors influencing cell culturing. Using such a
procedure, we want to test the influence of a decreased oxygen tension during
culturing and to determine the effect on culture quality (Chapter 2.3).
The quality of slide preparation can be measured as the degree of
spreading of the metaphase chromosomes and the presence or absence of
cytoplasm over the metaphases. It mainly depends on conditions during harvesting
(hypotonic treatment and composition of fixative) and the process of slide making
(climate conditions). We will investigate the effect of these variables on primary in
situ cultures (Chapter 3). Once we know the quantitative effects of ambient
temperature, relative humidity, air flow and composition of fixative, we will be able
to give some recommendations on how to achieve a consistent high quality of
preparations.
The quality of analysis can be measured as sensitivity and specificity of
diagnosis. It mainly depends on the banding resolution of the chromosomes and
the number of metaphases analysed. For amniotic fluid culture, probability tables
exist to determine the number of metaphases that should be analysed from either
the in-situ or the flask method of culturing in order to exclude a certain degree of
mosaicism. For trypsinised in-situ cultures, however, such a table is missing. We
will, therefore, construct a probability table to determine the number of metaphases
that need to be analysed in order to detect mosaicism when using trypsinised
back-up cultures complementing in-situ AF cell culture analysis (Chapter 4.3).
Introduction
B.Sikkemaproefschrift2.doc
29
For chorionic villi analysis, sensitivity and specificity need careful consideration,
starting from the selection of the cell type which should be analysed from the villi;
trophoblast cells in direct preparations, mesenchymal core cells in culturing or a
combination of direct preparation and culturing. Since routinely we are using
culturing only, we will investigate the results of that method in terms of success
rate, maternal cell contamination, chromosomal aberrations observed and
predictive value (false positive results). These results will be compared with those
from the literature on direct preparations alone and on the combination of the direct
preparation and the culturing method (Chapter 4.2).
Eventually, we will try to formulate guidelines for maintaining a good quality in
prenatal cytogenetic analysis of amniotic fluid culture and chorionic villi and
suggest possible further investigations (Chapter 5).
Chapter 1
B.Sikkemaproefschrift2.doc
30
1.4 LiteratureAdvice concerning genetic counselling. Report of health.1977. No. 77/14.
Association of Clinical Cytogeneticists (ACC) Working Party on Chorionic Villi in Prenatal Diagnosis.Cytogenetic analysis of chorionic villi for prenatal diagnosis: an ACC collaborative study of U.K. data.Prenat Diagn. 1994;14(5):363-79.
Association of Clinical Cytogeneticists (ACC) Working Party. Professional guidelines for clinicalcytogenetics. BMJ Publishing Group 2001. http://www.cytogenetics.org.uk/information/Guidelines.pdf
Bartels I, Rauskolb R, Hansmann I. Chromosomal mosaicism of trisomy 7 restricted to chorionic villi.Am J Med Genet. 1986; 25(1):161-2.
Behrman RE, Parer JT, De Lannoy CW Jr. Placental growth and the formation of amniotic fluid. Nature.1967; 214(89):678-80.
Bevis DCA. The antenal prediction of haemolytic disease of the newborn. Lancet 1953; 1: 395
Blackwell SC, Abundis MG, Nehra PC. Five-year experience with midtrimester amniocentesisperformed by a single group of obstetricians-gynaecologists at a community hospital. Am J ObstetGynecol. 2002; 186(6):1130-2.
Brambati B, Terzian E, Tognoni G. Randomized clinical trial of transabdominal versus transcervicalchorionic villus sampling methods. Prenat Diagn. 1991; 11(5):285-93.
Breed ASPM, Mantingh A, Govaerts LCP, Boogert A, Anders GJPA, Laurini R. Abnormal karyotype inthe chorion, not confirmed in a subsequently aborted foetus. Prenat Diagn. 1986; 6: 375-7.
Brock DJ, Sutcliffe RG. Alpha-fetoprotein in the antenatal diagnosis of anencephaly and spina bifida.Lancet. 1972; 2(7770):197-9.
Bui TH, Iselius L, Lindsten J. European collaborative study on prenatal diagnosis: mosaicism,pseudomosaicism and single abnormal cells in amniotic fluid cell cultures. Prenat Diagn. 1984;4 SpecNo:145-62.
Chang HC, Jones OW, Masui H. Human amniotic fluid cells grown in a hormone-supplementedmedium: suitability for prenatal diagnosis. Proc Natl Acad Sci U S A. 1982; 79(15):4795-9.
Chang HC, Jones OW. Enhancement of amniocyte growth on a precoated surface. Prenat Diagn. 1991;11(6):357-70.
Claussen U, Schafer H, Trampisch HJ. Exclusion of chromosomal mosaicism in prenatal diagnosis.Hum Genet. 1984; 67(1):23-8.
Claussen U, Michel S, Muhlig P, Westermann M, Grummt UW, Kromeyer-Hauschild K, Liehr T.Demystifying chromosome preparation and the implications for the concept of chromosomecondensation during mitosis. Cytogenet Genome Res. 2002; 98(2-3):136-46.
Cox DM, Niewczas-Late V, Riffell MI, Hamerton JL. Chromosomal mosaicism in diagnostic amnioticfluid cell cultures. Pediatr Res. 1974 ; 8(6):679-83.
Crane JP, Cheung SW. An embryogenic model to explain cytogenetic inconsistencies observed inchorionic villus versus fetal tissue. Prenat Diagn. 1988; 8(2):119-29.
Czepulkowski BH, Heaton DE, Kearney LU, Rodeck CH, Coleman DV. Chorionic villus culture for firsttrimester diagnosis of chromosome defects: evaluation by two London centres. Prenat Diagn. 1986;6(4):271-82.
Dutch Society of Obstetrics and Gynaecology. 1997. Guidelines for invasive prenatal diagnosis.http://www.nvog.nl
Dutch Prenatal Working Party. Annual report. 2002.
Deng W, Tsao SW, Lucas JN, Leung CS, Cheung AL. A new method for improving metaphasechromosome spreading. Cytometry. 2003; 51A(1):46-51.
Introduction
B.Sikkemaproefschrift2.doc
31
Eichenbaum SZ, Krumins EJ, Fortune DW, Duke J. False negative finding on chorionic villus sampling.Lancet 1986; 2: 391.
Elejalde BR, de Elejalde MM, Acuna JM, Thelen D, Trujillo C, Karrmann M. Prospective study ofamniocentesis performed between weeks 9 and 16 of gestation: its feasibility, risks, complications anduse in early genetic prenatal diagnosis. Am J Med Genet. 1990; 35(2):188-96.
Elias S, Simpson JL, Bombard AT. Amniocentesis and foetal blood sampling. In Genetic disorders andthe foetus, Milunsky A (ed.), The Johns Hopkins University Press, Baltimore and London,1998.
Featherstone T, Cheung SW, Spitznagel E, Peakman D. Exclusion of chromosomal mosaicism inamniotic fluid cultures: determination of number of colonies needed for accurate analysis. Prenat Diagn.1994; 14(11):1009-17.
Fuchs F. and Riis P. Antenatal sex determination. Nature 1956; 177: 330.
Galjaard H. Possibilities of quantitative histochemical analysis in prenatal detection and geneticclassification of hereditary metabolic disorders. In: Histochemistry and cytochemistry, T. Takeuchi (ed).Nakamishi Printing Kyoto 1972; 182-3.
Gibas LM, Grujic S, Barr MA, Jackson LG. A simple technique for obtaining high quality chromosomepreparations from chorionic villus samples using FdU synchronization. Prenat Diagn. 1987; 7(5):323-7.
Golbus MS, Loughman WD, Epstein CJ, Halbasch G, Stephens JD, Hall BD. Prenatal genetic diagnosisin 3000 amniocenteses. N Engl J Med. 1979; 300(4):157-63.
Hahnemann N. Early prenatal diagnosis; a study of biopsy techniques and cell culturing fromextraembryonic membranes. Clin Genet. 1974; 6(4):294-306.
Hahnemann JM, Vejerslev LO. European collaborative research on mosaicism in CVS (EUCROMIC)--foetal and extrafoetal cell lineages in 192 gestations with CVS mosaicism involving single autosomaltrisomy. Am J Med Genet. 1997; 70(2):179-87.
Hahnemann JM, Vejerslev LO. Accuracy of cytogenetic findings on chorionic villus sampling (CVS)--diagnostic consequences of CVS mosaicism and non-mosaic discrepancy in centres contributing toEUCROMIC 1986-1992. Prenat Diagn. 1997; 17(9):801-20.
Hliscs R, Muhlig P, Claussen U. The spreading of metaphases is a slow process which leads to astretching of chromosomes. Cytogenet Cell Genet. 1997; 76(3-4):167-71.
Hankins GD, Rowe J, Quirk JG Jr, Trubey R, Strickland DM. Significance of brown and/or greenamniotic fluid at the time of second trimester genetic amniocentesis. Obstet Gynecol. 1984; 64(3):353-8
Hoehn H, Bryant EM, Karp LE, Martin GM. Cultivated cells from diagnostic amniocentesis in secondtrimester pregnancies. I. Clonal morphology and growth potential. Pediatr Res. 1974; 8(8):746-54.
Hoehn H, Bryant EM, Karp LE, Martin GM. Cultivated cells from diagnostic amniocentesis in secondtrimester pregnancies. II. Cytogenetic parameters as functions of clonal type and preparative technique.Clin Genet. 1975; 7(1):29-36.
Hook EB. Exclusion of chromosomal mosaicism: tables of 90%, 95% and 99% confidence limits andcomments on use. Am J Hum Genet. 1977; 29(1):94-7.
Horger EO 3rd, Finch H, Vincent VA. A single physician's experience with four thousand six hundredgenetic amniocenteses. Am J Obstet Gynecol. 2001; 185(2):279-88.
Hsu LY, Perlis TE. United States survey on chromosome mosaicism and pseudomosaicism in prenataldiagnosis. Prenat Diagn. 1984; 4 Spec No:97-130.
Hsu LY, Kaffe S, Perlis TE. A revisit of trisomy 20 mosaicism in prenatal diagnosis--an overview of 103cases. Prenat Diagn. 1991; 11(1):7-15.
Hsu LY, Kaffe S, Jenkins EC, Alonso L, Benn PA, David K, Hirschhorn K, Lieber E, Shanske A, ShapiroLR, et al. Proposed guidelines for diagnosis of chromosome mosaicism in amniocytes based on dataderived from chromosome mosaicism and pseudomosaicism studies. Prenat Diagn. 1992; 12(7):555-73
Chapter 1
B.Sikkemaproefschrift2.doc
32
Hsu LY, Benn PA. Revised guidelines for the diagnosis of mosaicism in amniocytes. Prenat Diagn.1999; 19(11):1081-82.
ISCN 1995; An International system for human cytogenetic nomenclature. Mitelman F (ed); S. KargerBasel1995.
Jackson LG, Gibas LM, Coutino W. Chorionic villus. In Human chromosomes, manual of basictechniques . Verma RS, Babu A (eds.). Pergamon Oxford 1989:19-26.
Jacobson CB, Barter RH. Intrauterine diagnosis and management of genetic defects. Am J ObstetGynecol. 1967; 99(6):796-807.
Kalousek DK, Dill FJ. Chromosomal mosaicism confined to the placenta in human conceptions.Science. 1983; 221(4611):665-7.
Kazy Z, Rozovsky, Bakharev VA. Chorion biopsy in early pregnancy: a method of early prenataldiagnosis for inherited disorders. Prenat Diagn. 1982; 2: 39-45.
Keagle M, Gersen E, Gersen SL. Basic laboratory procedures. In: The principles of Clinical Genetics.Gersen S and Keagle M (eds.). Humana Press Inc. Totowa, NJ 1999.
Kennerknecht I, Baur-Aubele S, Grab D, Terinde R. First trimester amniocentesis between the seventhand 13th weeks: evaluation of the earliest possible genetic diagnosis. Prenat Diagn. 1992; 12(7):595-601.
Kullander S, Sandahl B. Foetal chromosome analyses after transcervical placental biopsies during earlypregnancy. Acta Obstet Gynecol Scand 1973: 52: 355-9.
Lam YH, Tang MHY, Sin SY, Ghosh A. Clinical significance of amniotic fluid cell culture failure. PrenatDiagn. 1998; 18: 343-7.
Lambl D. Ein seltener Fall von Hydramnions. Zentralbl Gynaekol 1881; 5: 329.
Laundon CH, Priest JH, Priest RE. The characterisation of hCG regulation in cultured human amnioticfluid cells. Prenat Diagn. 1981; 1: 269-75.
Ledbetter DH, Martin AO, Verlinsky Y, Pergament E, Jackson L, Yang-Feng T, Schonberg SA, GilbertF, Zachary JM, Barr M, et al. Cytogenetic results of chorionic villus sampling: high success rate anddiagnostic accuracy in the United States collaborative study. Am J Obstet Gynecol. 1990; 162(2):495-501.
Ledbetter DH, Zachary JM, Simpson JL, Golbus MS, Pergament E, Jackson L, Mahoney MJ, DesnickRJ, Schulman J, Copeland KL, et al. Cytogenetic results from the U.S. Collaborative Study on CVS.Prenat Diagn. 1992; 12(5):317-45.
Leschot NJ, Wolf H, Verjaal M, van Prooijen-Knegt, de Boer EG, Kanhai HH, Christiaens GC. Chorionicvilli sampling: cytogenetic and clinical findings in 500 pregnancies. Br Med J 1987; 295: 407-10.
Linton G, Lilford RJ. A false-negative finding on chorion villus sampling. Lancet 1986; 2: 630.
Martin AO, Sherman E, Rosinsky B, Bombard AT, Simpson JL. False negative finding on chorionic villussampling. Lancet 1986; 2: 391-2.
Mellink CH, de Pater JM, Poddighe PJ, Smeets DFCM. Quality of cytogenetic diagnosis: conditions,guidelines and control. 2003.
Mohr J. Foetal genetic diagnosis: development of techniques for early sampling of foetal cells. ActaPathol Microbiol Scand 1968; 73: 73-7.
Nadler H. Antenatal detection of hereditary disorders. Pediatrics 1968; 42: 912-8.
Nadler HL, Gerbie AB. Role of amniocentesis in the intrauterine detection of genetic disorders. N Engl JMed 1970; 282: 596-9.
Introduction
B.Sikkemaproefschrift2.doc
33
Nagel HTC, Knegt AC, Kloosterman MD, Wildschut HIJ, Leschot NJ, Vanderbussche FPHA. Invasiveprenatal diagnosis in the Netherlands, 1991-2000: number of tests, indications and detectedabnormalities. Ned Tijdschr Geneeskd 2004; 148 (31): 1538-43.
NICHD National Registry for Amniocentesis Study Group. Midtrimester amniocentesis for prenataldiagnosis: Safety and accuracy. JAMA 1976; 236: 1471.
Niermeijer MF, Los FJ. And Brandenburg H. Indications for prenatal diagnosis. In: Prenatal diagnosis.Brandenburg H (ed). Bohn Stafleu Van Loghum, Houten/ Diegem 1997; 1-9.
Old JM, Ward RHT, Kragozlu F, Petrou M, Modell B. Weaterall DJ. First- trimester foetal diagnosis forhaemoglobinopathies: three cases. Lancet 1982; 2(8313): 1413-6.
Peakman DC, Moreton MF, Robinson A. Prenatal diagnosis: techniques used to help ruling outmaternal cell contamination. J Med Genet 1977; 14: 37- 9.
Persutte WH, Lenke RR. Failure of amniotic- fluid- cell growth: is it related to foetal aneuploidy? Lancet1995; 345: 96-7.
Priest JH, Rao KW. Prenatal chromosome diagnosis. In The AGT Cytogenetics laboratory manual.Barch MJ, Knutsen T, Spurbeck (eds.). Lippincott – Raven Publishers, Philadelphia 1997.
Reid R, Sepulveda W, Kyle PAM, Davies G. Amniotic fluid culture failure: Clinical significance andassociation with aneuploidy. Obstet Gynecol 1996; 87: 588- 92.
Richkind KE, Risch NJ. Sensivity of chromosomal mosaicism detection by different tissue culturemethods. Prenat Diagn. 1990; 10: 519-27.
Roberts E, Duckett DP, Lang GD. Maternal cell contamination in chorionic villus samples assessed bydirect preparations and three different culture methods. Prenat Diagn. 1988; 8: 635-40.
Saunders K and Czepulkowski B. Culture of human cells for chromosomal analysis. In: Analyzingchromosomes. Czepulkowski B (ed.). BIOS Scientific Publishers Limited, Trowbridge, UK 2001.
Saura R, Roux D, Maugey-Laulon B, Taine L, Wen ZQ, Vergnaud A, Horovitz J. False negative resultsof trisomy 21 on direct analysis on chorionic villus sampling. Prenat Diagn. 1998; 18: 866-7.
Seguin LR, Palmer CG. Variables influencing growth and morphology of clones of cells from humanamniotic fluid. Prenat Diagn.1983; 3:107-16.
Silver RK, Russel TL, Kambich MP, Leeth EA, MacGregor SN. Midtrimester amniocentesis. Influence ofoperator caseload sampling efficiency. J Reprod Med 1998; 43 (3): 191-5.
Simoni G, Brambati B, Danesino C, Rossella F, Terzoli GL, Ferrari M, Fraccaro M. Efficient directchromosome analyses and enzyme determinations from chorionic villi samples in the first trimester ofpregnancy. Hum Genet. 1983;63(4):349-57.
Spurbeck JL, Zinsmeister AR, Meyer KJ, Jalal SM. Dynamics of chromosome spreading. Am J MedGenet. 1996; 61(4):387-93.
Steele M. and Breg WR. Chromosome analysis of human amniotic fluid cells. Lancet 1966; 1: 383-5.
Tabor A, Philip J, Madsen M, Bang J, Obel EB, Norgaard-Pedersen B. Randomised controlled trial ofgenetic amniocentesis in 4606 low-risk women. Lancet. 1986; 1(8493):1287-93.
Valenti C, Schutta EJ, Kehaty T. Prenatal diagnosis of Down's syndrome. Lancet. 1968; 2(7561):220.
Watson MS. Quality assurance and quality control in clinical cytogenetics. Current Protocols in HumGenet 1997; Suppl 14: 8.2.1- 8.2.7.
Worton RG, Stern R. A Canadian collaborative study of mosaicism in amniotic fluid cell cultures. PrenatDiagn. 1984; 4: 131-44.
Chapter 1
B.Sikkemaproefschrift2.doc
34
Cell culture of amniotic fluid cells
B.Sikkemaproefschrift2.doc
35
Chapter 2
Cell culture of amniotic fluid cells
Chapter 2
B.Sikkemaproefschrift2.doc
36
Cell culture of amniotic fluid cells
B.Sikkemaproefschrift2.doc
37
Chapter 2.1
Effect of variation in amniotic fluidspecimens and of type of culturemedium on culture time and numberof cell clones of primary in-situcultures of amniocytes
B.Sikkema-Raddatz, R.Suijkerbuijk, K.Bouman, C.H.C.M. Buys, G.J. te Meerman
Department of Medical Genetics, University of Groningen, The Netherlands
Chapter 2
B.Sikkemaproefschrift2.doc
38
2.1.1 Abstract
For a series of 579 consecutive amniotic fluid (AF) specimens received for prenatal
cytogenetic analysis within a year, we registered for each specimen the following
parameters: the indication for amniocentesis, which gynaecologist performed the
amniocentesis, the gestational age at amniocentesis, the volume and the sample
appearance (clear, with some blood, bloody or brown), the size of the cell pellet,
the lot and type of culture medium and the number of cell clones observed at the
first check after 5 or 6 days of culturing. Slides for cytogenetic analysis were made
when about 80% of the cell clones reached a medium size. The number of days
between initiation and end of the culture (total culture time) was registered, as was
the result of the cytogenetic analysis. The total number of clones at first inspection
and the total culture time as measures of cell culture quality were used as
dependent parameters. The influence of these parameters on cell culture quality
was statistically analysed by analysis of variance. The number of clones is
significantly influenced by the variability of admitted AF (volume, size of the pellet,
appearance). However, provided that at least 0.8 ml AF per culture dish is
admitted, this variability will in general not affect the procedure, since the number
of clones routinely found exceeds the number considered as minimal for a reliable
cytogenetic diagnosis. Culture time is only influenced by the type of culture
medium and the appearance of AF. Bloody or brown AF increased the culture
time, from on average 8.7 days to 11.5 days.
2.1.2 Introduction
Prenatal cytogenetic diagnosis by amniocentesis requires cell culture from amniotic
fluid (AF). The success rate of AF cell culture has dramatically increased over the
years, so that culture failure currently occurs in no more than 0.5% of the cases
(Lam et al., 1998). Introduction of the in situ method and optimisation of culture
media (Chang et al., 1982) have strongly reduced the time required for cytogenetic
diagnosis. The volume of the admitted AF specimen is rarely critical, as 3.2 ml
appear sufficient for a reliable cytogenetic diagnosis (Sikkema et al., 2002).
Cell culture of amniotic fluid cells
B.Sikkemaproefschrift2.doc
39
Normally, the specimens obtained consist of clear amniotic fluid, but sometimes
admixture of blood causes a bloody or brown appearance. It is speculated that this
might adversely influence cell culture quality as measured by the number of cell
clones and the total culture time (Seguin et al., 1983). Other factors that may affect
the quality of the cell culture are the experience of the gynaecologist performing
the amniocentesis, the gestational age at amniocentesis and the culture medium.
However, for none of these parameters the influence on cell culture quality has
been systematically investigated. Therefore, we retrospectively investigated the
effect of these parameters on the outcome of our standard AF cell culture
procedure for a period of one year by analysis of variance.
2.1.3 Materials and MethodsCulture procedure
Per patient approximately 16 ml AF was drawn and divided over two tubes. The
volume of AF was measured, the appearance was categorised as clear, containing
some blood, bloody or brown. After centrifugation the height of the pellet in a 15 ml
conical-based tube was ranked as small (< 1mm), normal (1 – 4 mm) or large
(>4mm). Amniotic fluid cells were cultured according to the in-situ method on 30
mm coverslips in 35 mm Petri dishes under 5% CO2 at 37 0C. From each tube two
cultures were started, using two different types of culture media. In one system
Amniomax (Invitrogen – Life Technologies, Breda, The Netherlands) from three
different lots was used. In the other system Amniochrome (Cambrex Company,
Verviers, Belgium) was used pure as well as diluted with foetal calf serum and α-
MEM (Cambrex Company, Verviers, Belgium). The latter system included three
different lots of Amniochrome and three different lots of serum. In total we have
used seven different compositions of culture medium by combining each lot of
Amniochrome with two different lots of serum and one lot of pure Amniochrome.
The medium was changed after 5 or 6 days and the cultures were checked
for cell growth under a phase contrast microscope (Zeiss, Germany). The number
Chapter 2
B.Sikkemaproefschrift2.doc
40
of clones was counted and the size of the clones was classified as "a cell group"
(from a minimum of 5 cells up to 50 cells), "medium" (more than 50 cells, but all
cells visible within one view at 65x magnification) and "large" (not all cells visible
within one view at 65x magnification). Cultures were checked every other day and
prepared for analysis of mitotic cells when about 80% of the clones were of
medium size. One of the four cultures per patient was not prepared for analysis
and kept as back-up until a chromosomal diagnosis was obtained.
Data registration
Data from 579 AF samples, received in a one-year period, were stored in a
datafile, with as independent variables per culture:
- gynaecologist who performed amniocentesis (N=9)
- indication for amniocentesis
- weeks of gestation (14 – 36 weeks)
- volume of AF from each tube (2ml – 15 ml )
- appearance of the AF (clear, some blood, bloody, brown)
- height of the pellet after centrifugation (small: < 1mm, cells yet visible; normal:
1 - 4mm; large: > 4mm in a 15 ml conical-based tube)
- culture medium: type and lot number of medium and serum
- chromosomal diagnosis
As dependent variables we used:
- the total number of clones at the first check (after 5 or 6 days of culture)
- total culture time (time between initiation of culture and preparation for
analysis)
Cell culture of amniotic fluid cells
B.Sikkemaproefschrift2.doc
41
Statistical analysis
Cell culture quality was defined per specimen as the total number of clones at the
first check and as the total culture time. For both parameters, analysis of variance
was used to assess the influence of the admitted AF, the gynaecologist and the
composition and lot of the culture medium. The data were analysed in SPSS10.
Analysis of variance and regression was used to assess the quantitative effect of
the different parameters.
2.1.4 ResultsVariance in the number of clones after the first check
The variance in the total number of clones at the first check is minor, but
significantly influenced by the diversity of the admitted AF (Table 1a). The same
result was calculated for the total number of clones. Of the variance of the number
of clones after the first check 4.7 % is explained by which gynaecologist performed
the amniocentesis. Different appearance of AF resulted in about 3% of the
variance in the number of clones. Size of the pellet and volume of AF explained
approximately 1.5% of the variance in the number of clones, whereas duration of
gestation had no significant influence. The total number of clones was significantly
different between the different lots of culture media, explaining up to 4.5% of the
variance (Table 1b). There was no significant difference in the total number of
clones between the different types of culture medium.
Chapter 2
B.Sikkemaproefschrift2.doc
42
Table 1a: Separate effect analysis of variance for the number of clones on thediversity of the admitted amniotic fluid
Effect Explainedvariance
(%)
Sum ofsquares
DF Meansquare
F ratio P (twotailed)
gynaecologist 4.7 3746.3 6 624.4 6.4 0.00
Error 79070.9 782 97.1
weeks of gestation 0.4 328.9 1 328.9 3.3 0.072
Error 76643.6 756 101.1
volume of AF 1.3 1045.1 2 522.6 5.2 0.006
Error 79070.9 782 100.0
size of the pellet 1.4 1084.8 2 542.4 5.4 0.005
Error 79070.9 782 99.9
appearance of the AF 2.9 2283.2 2 1141.6 11.6 0.00
Error 79070.9 782 98.4
Table 1b: Analysis of variance for the total number of clones on different lots andtypes of medium
Effect Explainedvariance
(%)
Sum ofsquares
DF Meansquare
Fratio
P (twotailed)
Amniomax versusAmniochrome with serum 1.60 401.6 1 401.6 3.9 0.048
Error 24471.2 241 101.5
Amniomax (differentbatches) 3.50 3326.4 2 1663.2 14.9 0.00
Error 102291.4 919 111.3
Amniochrome (differentbatches with/withoutserum)
4.50 4734.2 6 789.0 7.2 0.00
Error 100883.6 915 110.3
Amniochrome pure versusAmniochrome with serum 2.10 2193.1 1 2193.1 19.5 0.00
Error 103424.7 920 112.4
Cell culture of amniotic fluid cells
B.Sikkemaproefschrift2.doc
43
Variance in the total culture time
Approximately 14% of the variance in culture time is explained by the appearance
of the AF (Table 2). For bloody or brown AF (Table 3) three extra days of culturing
appeared necessary. Although the culture time became extended, the appearance
of AF had a minor influence on the number of clones.
About 6 % of variance in culture time is explained by duration of gestation
(Table 2). However, this effect is minor, as it appears statistically to be not
significant. We received 40 AF samples drawn after 24 weeks of gestation. After
sub-grouping the duration of gestation in less or more than 24 weeks, effects on
culture time and on number of clones were still not significant, though a minor
increase in culture time was detectable: 9.2 days for specimens taken after 24
weeks of gestation versus 8.5 days for specimens taken before 24 weeks of
gestations. The reason for amniocentesis for the specimens taken after 24 weeks
of gestation, were ultrasound abnormalities. Ten of them were chromosomally
abnormal and in one case a total culture failure occurred.
Of all the 579 specimens, 36 showed an abnormal karyotype (29 with unbalanced
and 7 with balanced abnormalities). Culture time appeared not to be influenced by
an abnormal chromosomal diagnosis (2% of the variance explained, statistically
not significant; data not shown).
Chapter 2
B.Sikkemaproefschrift2.doc
44
Table 2: Single effect analysis of variance for the total culture time on the diversityof the admitted amniotic fluid
Effect Explainedvariance
(%)
Sum ofsquares
DF Meansquare
F ratio P (twotailed)
gynaecologist 4.75 52.1 8 6.5 2.3 0.021
Error 1044.3 367 2.8
weeks ofgestation
6.10 66.8 21 3.2 1.1 0.351
Error 1029.5 354 2.9
volume of AF 5.30 57.9 13 4.5 1.6 0.097
Error 1038.5 362 2.9
size of the pellet 1.80 20.1 3 6.7 2.3 0.075
Error 1076.3 372 2.9
appearance ofthe AF 14.0 151.6 4 37.9 14.9 0.00
Error 944.8 371 2.5
Table 3: Mean effects on culture time
Appearance of amniotic fluid Number of samples Culture time in
days (±SD)clear 264 8.68 (± 1.4)
some blood 71 8.64 (± 1.2)
bloody 28 10.35 (± 3.1)
brown 11 11.45 (± 2.4)
SD, Standard deviation
A significant difference in culture time was present related to the type of
medium (Table 4), which explained 2.6% of the variance. This resulted in a one-
day delay in culture time for Amniochrome diluted with foetal calf serum and α-
MEM compared with Amniomax (Table 5). Therefore, more cultures were
harvested from the culture system with Amniomax. The total culture time is not
influenced by a lot effect for different batches of medium (Table 4).
Cell culture of amniotic fluid cells
B.Sikkemaproefschrift2.doc
45
Table 4: Analysis of variance for the total culture time on different lots and types ofmedium
Effect Explainedvariance
(%)
Sum ofsquares
DF Meansquare
F ratio P (twotailed)
Amniomax versusAmniochrome withserum
2.6 55.3 1 55.3 23.6 0.000
Error 1999.3 852 2.3
Amniomax (differentbatches) 0.6 7.1 2 3.5 1.2 0.301
Error 1089.3 372 2.9
Amniochrome(different batcheswith/without serum)
1.5 16.7 6 2.8 0.9 0.461
Error 1079.7 368 2.9
Amniochrome pureversus Amniochromewith serum
0.1 1.0 1 1.0 0.3 0.566
Error 1095.4 373 2.9
Table 5: Comparison between culture media
Amniomax Amniochrome with serum
Average culture time(days ±SD) 8.45 (± 2.1) 8.88 (± 1.7)
No. of harvested slides 478 376
Number of clones (±SD) 18.95 (± 9.5) 17.22 (± 11.0)
No. of clones at day 5 (±SD) 11.9 (± 7.5) 9.99 (± 7.6)
No. of clones at day 7 (±SD) 21.8 (± 9.5) 21.25 (± 12.7)
No. of clones at day 8 (±SD) 18.5 (± 8.5) 17.9 (± 8.1)
Growth maximum at day 7 day 8
SD, standard deviation
Chapter 2
B.Sikkemaproefschrift2.doc
46
2.1.5 Discussion
In this study we have systematically examined the influence of differences in AF
sample characteristics and culture medium on cell culture quality. For admitted AF,
bloody or brown AF had a significant effect (14% explained variance) on culture
time. Seguin et al. (1983) suggested from their results that bloody or brown AF
inhibited cell growth by restriction of clone size. Chang and Jones (1991) showed
that a grossly bloody specimen could inhibit growth by 30 percent. Our results
show that bloody or brown AF delayed the growth for three days. However, the
number of clones was minimally affected.
The volume of AF did not affect total culture time and affected the number
of clones to only a minimal extent. Our former results (Sikkema et al, 2002) already
demonstrated that the same number of clones resulted from 0.8 ml to 1.5 ml of AF
per in-situ dish as from using 5 ml or more. Seguin et al. (1983) determined that
the inoculating concentration of cells which was most efficient corresponded to
about 0.8 to 2.5 ml AF for flask cultures. They found that for producing clones small
sample volumes are more efficient. This saturation effect might explain the fact that
only small differences were measured in our results, since the minimal amount of
AF received was 2 ml.
The gestational age at amniocentesis had no influence on culture quality. This is in
line with results of Elejalde et al. (1990). Although the total number of cells
increases with gestational age, the number of viable cells decreases and only a
fraction of these cells is forming clones. According to our results, the size of the
pellet correlated positively with the duration of gestation. We therefore conclude,
that the number of clone-forming cells remains stable during the second and third
trimester of gestation.
Reports in the literature (Reid et al., 1996; Persutte et al., 1995) suggest an
association between culture failure and karyotype abnormality. Among our cases
there was only a single case of culture failure (IUGR at 35 weeks, karyotype in
fibroblasts 46,XY, del(4p)). Because of this near absence of failures, the
association could not be investigated in this study. We therefore measured the
Cell culture of amniotic fluid cells
B.Sikkemaproefschrift2.doc
47
variance in culture time for patients with normal and abnormal chromosomal
diagnosis. Since we did not find any difference between these two groups, we
conclude that culture delay or even culture failures are not associated with
chromosomal abnormalities. Lam et al. (1998) conclude that rather than being
associated with an abnormal karyotype, culture failure is associated with foetal
structural defects found by ultrasound in advanced pregnancies. Our results are in
line with these data. Although there was no significant variance in culture time
before or after 24 weeks of gestation, culture time was delayed to a minor extent in
the group of patients with AF sampling after 24 weeks. All of them had foetal
abnormalities at ultrasound. An abnormal karyotype occurred in 25%.
The number of clones was only slightly affected under all investigated
conditions. Obviously, the quality of AF culture medium has been so much
optimised by enrichment with growth factors (Chang et al., 1985) that when a
minimum of AF is cultured, most of the living cells in the fluid develop into a clone.
This idea is supported by experiments of Chang and Jones (1991) in which growth
saturation occurred after enrichment of culture medium with cell attachment
factors.
For the type of culture medium a significant difference in culture time was
found between Amniomax and Amniochrome (with serum). For Amniochrome, one
day delay of culture time was measured. Since the exact compositions of the
culture media are not available, we cannot identify what causes this difference.
Although culture time for Amniochrome is extended, we keep using this medium in
our routine diagnostic procedures in order to have two different and independent
culture systems, thus contributing to the overall reliability of the diagnosis.
In summary, current procedures of amniocentesis and of culturing of
amniocytes are sufficient for a reliable and consistent culture quality. The number
of clones is significantly influenced by the variability of admitted AF (volume, size of
the pellet), but provided that at least 0.8 ml AF per culture dish is admitted, this is
hardly of importance since the number of clones generally exceeds the number
considered as minimal for a reliable cytogenetic diagnosis. Culture time is only
Chapter 2
B.Sikkemaproefschrift2.doc
48
influenced by the type of culture medium and the appearance of AF. Bloody or
brown AF increases culture time for three days.
2.1.6 Literature
Chang HC, Jones OW, Masui H. Human amniotic fluid cells grown in a hormone-supplemented
medium: suitability for prenatal diagnosis. Proc Natl Acad Sci U S A. 1982 ;79(15):4795-9.
Chang HC, Jones OW. Reduction of sera requirements in amniotic fluid cell culture. Prenat Diagn.
1985;5(5):305-12.
Chang HC, Jones OW. Enhancement of amniocyte growth on a precoated surface. Prenat Diagn.
1991;11(6):357-70.
Elejalde BR, de Elejalde MM, Acuna JM, Thelen D, Trujillo C, Karrmann M. Prospective study of
amniocentesis performed between weeks 9 and 16 of gestation: its feasibility, risks, complications and
use in early genetic prenatal diagnosis. Am J Med Genet. 1990;35(2):188-96.
Lam YH, Tang MHY, Sin SY, Ghosh A. Clinical significance of amniotic-fluid-cell culture failure. Prenat
Diagn. 1998;18(4):343-7.
Persutte WH, Lenke RR. Failure of amniotic-fluid-cell growth: is it related to foetal aneuploidy? Lancet.
1995;14;345:96-7.
Reid R, Sepulveda W, Kyle PM, Davies G. Amniotic fluid culture failure: clinical significance and
association with aneuploidy. Obstet Gynecol. 1996;87(4):588-92.
Seguin LR, Palmer CG. Variables influencing growth and morphology of colonies of cells from human
amniotic fluid. Prenat Diagn. 1983;3(2):107-16.
Sikkema-Raddatz B, van Echten J, van der Vlag J, Buys CHCM, te Meerman GJ. Minimal volume of
amniotic fluid for reliable prenatal cytogenetic diagnosis. Prenat Diagn. 2002;22(2):164-5.
Cell culture of amniotic fluid cells
B.Sikkemaproefschrift2.doc
49
Chapter 2.2
Minimal volume of amniotic fluid forreliable prenatal cytogenetic diagnosis
B. Sikkema-Raddatz, J. van Echten, J. van der Vlag, C.H.C.M. Buys and G.J. te
Meerman
Department of Medical Genetics, University of Groningen, The Netherlands
Prenat. Diagn 2002; Vol. 22: 164-165
Chapter 2
B.Sikkemaproefschrift2.doc
50
Letter to the editor:
Second trimester amniocentesis and amniotic fluid (AF) culture is a well-known
method of prenatal cytogenetic diagnosis. Introduction of the in situ method and
optimised culture media has strongly reduced the culture time. In addition, culture
failures have been reduced to less than 0.5%. In current amniocentesis practice,
20 ml AF is drawn for prenatal cytogenetic diagnosis and usually at least three
cultures are established from this volume (Priest and Rao, 1997). However, in a
published amniocentesis series (Winsor et al.,1999), a volume of less than 8 ml
was obtained in 13% of the cases. Repeated amniocentesis is recommended when
less than 5 ml of AF is obtained (Priest and Rao, 1997).
This prompted us to investigate the minimal volume of AF required for a
reliable diagnosis, in order that repeated amniocentesis can be avoided.
In a pilot study we pippetted directly 0.25 ml, 0.50 ml, 0.75 ml, 1.0 ml and
1.5 ml AF of five different samples on Ø30 mm coverslips in Ø35 mm Petri dishes
and immediately added 2 ml of culture medium (Amniomax TM C-100, Life
Technologies B.V., The Netherlands). It appeared that from a volume of 0.25 ml AF
only one or two clones developed. A volume of 0.75 ml yielded more clones, but
fewer than eight, too few for a reliable diagnosis.
Therefore, volumes of 0.8 ml, 1.0 ml and 1.5 ml of AF were used for
culturing on Ø30 mm coverslips in Ø35 mm Petri dishes. From each of 12 samples,
randomly chosen within a certain time interval, these various volumes were directly
pipetted onto coverslips and 2 ml culture medium (Amniomax TM C-100, Life
Technologies) was immediately added. As a control, our standard procedure was
used, in which four cultures with an equivalent of about 5 ml AF are established.
Briefly, approximately 20 ml of AF is collected in two tubes and centrifuged at 1000
rpm for 10 min. Supernatants are removed until a volume of 0.3 ml remains. Then,
0.5 ml culture medium (Amniomax TM C-100, Life Technologies) is added. The mix
per tube is divided over two Ø35 mm Petri dishes containing Ø30 mm coverslips.
The following day, 2 ml medium is added to the culture.
Cell culture of amniotic fluid cells
B.Sikkemaproefschrift2.doc
51
In order to exclude changes in the proportion of AF and culture medium, a
second series of 0.8 ml, 1.0 ml and 1.5 ml volumes from 12 other AF samples was
studied. After centrifugation at 1000 rpm for 10 min, supernatants were removed
until a volume of 0.8 ml remained. This was pipetted onto Ø30 mm coverslips in
Ø35 mm Petri dishes. Than 2 ml culture medium was added. As a control, about 5
ml of AF was centrifuged as described earlier. The supernatant was removed until
0.8 ml remained. This was pipetted directly on a coverslip and 2 ml culture medium
was added.
After 6 days, both test and standard cultures were blindly checked for cell
growth. Based upon the number of clones, the appropriate day for in-situ
cytogenetic analysis was determined.
In both series we found on average 14 clones per Petri dish with 0.8 ml AF
(Table 1), sufficient for chromosomal diagnosis. Although the proportion of AF and
culture medium differed between the two series, this did not affect the number of
resulting clones. The appropriate day for analysis, however, varied with the amount
of AF. Culturing of 1.5 ml AF was comparable with the standard procedure,
whereas culturing of 0.8 ml resulted in a delay of two days for the appropriate day
for analysis.
The reliability of cytogenetic diagnosis was confirmed in a series of 52 samples. On
average 13 ±6 clones were obtained that could be analysed at Day 11 (controls
16±5 clones at Day 9). No culture failure was obtained.
Chapter 2
B.Sikkemaproefschrift2.doc
52
Table 1: Number of clones and appropriate day of analyses resulting from different
volumes of amniotic fluid (AF)
AF (ml) Culturing with different endvolumes (n= 12)
Culturing with a fixed end volume(n= 12)
Number ofclones (SD)
Day of analyses(SD)
Number of clones(SD)
Day of analyses(SD)
0.8 14 (±7.6) 11.4 (±1.8) 14 (±7.2) 12.1 (±1.6)
1.0 16 (±5.1) 11.3 (±1.8) 14 (±5.7) 11.3 (±1.7)
1.5 16 (±3.7) 11.5 (±1.5) 17 (±7.5) 9.9 (±1.6)
Control(approximately5ml)
15 (±5.6) 8.6 (±0.7) 18 (±4.9) 9.0 (±0.8)
SD, standard deviation; n, number of samples
Whereas Richkind and Risch (1990) obtained on average no more than
two clones per millilitre of AF, about four times more cells grew out to analysable
clones under our much more marginal conditions. The finding that an
approximately equal number of clones grow from about 1ml AF in the present
experiments and from 5 ml in the standard procedure might be attributed to an
inhibitory effect. Substances derived from dead cells, present in the amniotic fluid,
might inhibit the viable cells to grow out to clones. This effect would be stronger
when a higher number of dead cells is present.
The present study shows that using four cultures per diagnosis, repeated
amniocentesis will not be necessary as long as more than 3.2 ml (4x 0.8 ml) of AF
is available. Since the number of clones obtained in second trimester
amniocentesis in the present experiments is comparable to the number obtained
per millilitre in early amniocentesis as reported by Kennerknecht et al. (1992), it
might be worthwhile investigating the feasibility of also performing early diagnosis
with smaller volumes of AF.
Cell culture of amniotic fluid cells
B.Sikkemaproefschrift2.doc
53
Literature
Kennerknecht I, Baur-Aubele S, Grab D, Terinde R. First trimester amniocentesis between the seventh
and 13th weeks: evaluation of the earliest possible genetic diagnosis. Prenat Diagn. 1992;12(7):595-
601.
Priest JH and Rao KW. Prenatal chromosome diagnosis. The AGT Cytogenetic Laboratory Manual (3rd
edn), Barch MJ, Knutsen T and Spurbeck J (eds.) Lippincott- Raven: Philadelphia, PA 1997; Chapter 5,
206.
Richkind KE, Risch NJ. Sensitivity of chromosomal mosaicism detection by different tissue culture
methods. Prenat Diagn. 1990;10(8):519-27.
Winsor EJ, Tomkins DJ, Kalousek D, Farrell S, Wyatt P, Fan YS, Carter R, Wang H, Dallaire L, Eydoux
P, Welch JP, Dawson A, Lin JC, Singer J, Johnson J, Wilson RD. Cytogenetic aspects of the Canadian
early and mid-trimester amniotic fluid trial (CEMAT). Prenat Diagn. 1999;19(7):620-7.
Chapter 2
B.Sikkemaproefschrift2.doc
54
Cell culture of amniotic fluid cells
B.Sikkemaproefschrift2.doc
55
Chapter 2.3
Absolute procedure to test the growthpotential of medium in primaryamniotic fluid cell cultures and theinfluence of decreased oxygen tension
B. Sikkema-Raddatz, R. Suijkerbuijk, J. van der Vlag, M. Stoepker, C.H.C.M. Buys,
G.J. te Meerman
Department of Medical Genetics, University of Groningen, The Netherlands
Chapter 2
B.Sikkemaproefschrift2.doc
56
2.3.1 Abstract
We have developed an absolute procedure to test the performance of culture
medium in amniotic fluid in-situ cell cultures. By reducing the concentration of
medium until a 50% growth reduction was achieved, we expect that under these
sub-optimal conditions differences in culture quality will be more pronounced
(stress testing). Our test system is based on three cultures of primary amniotic fluid
(AF) cells from each of 10 individual patients (a total of 30 cultures), initiated with
medium diluted to a 30% end concentration. Absolute thresholds for acceptance
or rejection of new lots of medium were determined from 174 individual in-situ
cultures with different types of medium. We set the threshold for acceptance to at
least 22 successful cultures out of 30 and to an average number of 3 clones or
more. With these criteria the test system has a probability of less than 0.01 of false
rejection of a good lot of medium and sufficient power to detect downward
deviations in quality. We verified the validity of our stress testing system by
evaluating the effects of decreased oxygen tension during culturing and found that
decreased oxygen tension leads to an increase of the size of a clone and the
number of cells it contains, but with minimal effect on the number of clones only.
Maximal increase in cell number was observed under 5% oxygen tension
measured by growth curves.
2.3.2 Introduction
The quality of culture medium is crucial for the success and speed of a prenatal
chromosomal diagnosis. A milestone in AF cell culture was the introduction of
media supplemented with growth factors by Chang et al. (1982). The quality of the
cell cultures increased dramatically; the culture time was reduced from 20.3 days
to 12.8 days and culture failure from 1.53% to 0.33% (Salk et al., 1983). Each type
of AF cell culture medium contains a certain amount of serum, either included in a
commercial medium or added separately to a basic culture medium. The growth
potential of serum can vary between different lots (Pye, 1975), because serum
contains a complex mixture of growth promoting factors, since it is produced in
vivo. To maintain cell culture quality and to avoid culture failures, each lot of serum
Cell culture of amniotic fluid cells
B.Sikkemaproefschrift2.doc
57
and medium has to be tested for its growth potential (Priest and Rao, 1997). This is
particularly important in prenatal cytogenetic diagnosis, in which repeated
amniocentesis needs to be avoided.
To date, use of parallel cultures with new and old medium is the recommended
procedure (Priest and Rao, 1997). In our laboratory, growth of pooled secondary
AF cell cultures is compared between new and old medium by making growth
curves over several days. A new lot is only introduced in routine diagnostic practice
if growth in the new medium is better than or approximately equal to that in the old
medium. This type of testing is in principle subject to drift. It also may not yield the
most relevant quality indication, as the test is based on cells that have been in
culture already. We, therefore, wanted to develop an absolute test for primary AF
in-situ cell cultures which could be applied to left-over small fractions of patient
samples. Among individual AF cell cultures there is a high variability in number of
clones and culture time to harvest. The growth potential of serum is in particular
important for cultures with a minimal number of clones and slow growth.
Our absolute test for primary AF cell cultures should have fixed thresholds,
defined as a minimal number of clones and a minimal number of successful
cultures after certain days. Our starting point for developing the test is that
differences in quality of culture medium might be better detectable under sub-
optimal conditions, i.e. stress testing. The criterion for rejection should then be
based on the observed variability in culture success and number of clones, such
that false rejection should occur only rarely, e.g. in 1% of cases. Here we describe
the successful development of such a test. Our stress testing procedure is not only
suitable for testing the growth potential of culture medium, but can also be used to
investigate the influence on growth potential of any other factor involved in cell
culture. As an example we demonstrate its application to monitor the influence of
decreased oxygen tension on growth potential during culturing.
Chapter 2
B.Sikkemaproefschrift2.doc
58
2.3.3 Materials and MethodsStress testing
Amniotic fluid was obtained by second trimester amniocentesis between 15 weeks
and 20 weeks of gestation. Routine primary in situ cultures of AF cells were
initiated, leaving a small fraction (approximately 2.5ml) of the AF for stress testing.
We evaluated the effect of the following forms of different potential stress:
- Storage of AF; before culturing, AF was stored for 2 days at room temperature
(RT) or at 40C or AF was centrifuged and the pellet resuspended in PBS. This
mixture was stored either at RT or at 40C.
- Heat shock; before culturing, the received AF was placed in a waterbath at 400C
for 30 min or 60 min.
Storing or heat shock of AF before culturing showed no growth reduction (results
not shown).
- Dilution of culture medium; in order to have a comparable quality of AF in different
tests, pooled AF was used. A pool consisted of AF from at least three different
patients. This enabled us to test several conditions within a short time. In a first
series of experiments ten cultures were established from pooled AF. Culture
medium was diluted with PBS to end concentrations of 20%, 40%, 60% and 80%,
respectively.
Primary AF cell cultures were initiated from 0.8 ml AF and supplied with 2
ml diluted culture medium as described previously (Sikkema et al., 2002). Cultures
initiated with undiluted medium served as controls. Culture medium was refreshed
on days 6 and 8. The number of clones was counted under a phase contrast
microscope on days 6, 8 and 10 and compared with control cultures. The size of
the clones was classified as a " cell group" (from a minimum of 5 cells up to 50
cells), "medium" (more than 50 cells, but all cells visible within one view at 65x
magnification) and "large" (not all cells visible within one view at 65x magnification).
Growth reduction was defined as the reduction of the number of clones compared
with that in control cultures.
Cell culture of amniotic fluid cells
B.Sikkemaproefschrift2.doc
59
Finally, to obtain a good estimate of dilutions resulting in approximately 50%
growth reduction, end concentrations of culture medium of 60% and 30% were
tested with AF from 18 different pools (54 cultures).
Stress testing and determining thresholds for individual AF cell cultures
AF was used from all individuals from whom a minimum of 18ml fluid was received.
Experimental series, consisting of three cultures from each of 10 individual
patients, were initiated with 0.8 ml AF supplied with 2ml culture medium diluted to
an end concentration of 30%. Amniomax C-100 TM (Invitrogen – Life
Technologies, Breda, The Netherlands) and different lots of Amniochrome
(Cambrex Company, Verviers, Belgium) were tested. Media had been accepted
previously through growth curve testing and used in our routine diagnostic practice.
For acceptance or rejection of new lots of serum or medium, absolute thresholds
are defined as a minimal average number of clones and a minimal number of
successful cultures at 8 days. In total, six different series of 30 cultures resulting in
180 individual cultures were used. Six cultures were excluded from analysis
because of non-clonal growth of rapidly adhering cells. This left 174 cultures. The
number of clones was counted on days 6 and 8 and the number of successful
cultures was determined on day 8. Routine AF cell cultures from the same patients
served as a control for failures not due to stress testing. No culture failures were
observed under normal conditions. For validation of the thresholds, other media
with different growth potential were equally stress tested (Table 1). Medium with
known inferior quality as determined by growth curve testing (F10 supplemented
with 20% foetal calf serum), medium of unknown quality (Amniochrome 4MB0020)
and medium accepted in our routine practice were used in stress testing.
Chapter 2
B.Sikkemaproefschrift2.doc
60
Growth curve testing
The growth curve is used to compare a new lot of medium with the lot in use and is
based on secondary AFC cultures. About 14 days after initiating routine (primary)
cultures, five different back-up cultures are trypsinised. Cell suspensions of these
cultures are pooled and the number of cells is counted on a cell counter (Sysmex
F-820, Charles Goffin, Tiel, The Netherlands). The suspension is split to initiate 24
secondary cultures, of which12 are cultured on a new lot of medium and the other
12 on the lot in use. The medium is changed on day 6. Three cultures of each lot
are harvested on days 5, 6, 7 and 8. The number of cells in each culture is counted
and the mean and standard deviation is calculated. The new lot of medium is only
introduced into routine diagnostic practice if growth of the new lot is equal (within 1
standard deviation) or better than the lot in use. This indicates that drift of the
acceptance criterion towards higher and eventually insurmountable thresholds may
occur, as well as gradual worsening of quality.
Stress testing of decreased oxygen tension during culturing
The decreased oxygen tension was set at 2.5% and 5%, compared to the normal,
approximately 20% in the atmosphere. Stress testing was performed as described
before with individual AF cell cultures at 370C and a CO2 concentration of 5% in a
waterjacketed CO2 incubator (Forma Scientific, Thermo Electron Corporation,
USA). Parallel cultures from the same individuals were incubated at standard
atmospheric (20%) oxygen in a Steri- Cycle CO2 incubator.
A single lot number of Amniochrome, previously accepted in the routine, was used
for all cultures. The number of clones grown under 2.5% and 5% oxygen tension
was counted on days 6 and 8 and compared with the number of clones from
control cultures. Furthermore, growth curves were initiated as described before at
2.5%, 5%, and 10% and 20% oxygen tension (control condition), respectively, with
undiluted medium and with medium diluted to an end concentration of 30%. The
number of cells was counted and compared with the number of cells grown under
the control condition. To test for possible differences in growth potential between
Cell culture of amniotic fluid cells
B.Sikkemaproefschrift2.doc
61
both incubators, a series of 30 parallel cultures was initiated. No significant
differences in number of clones and/ or number of successful cultures were
detected.
2.3.4 ResultsMedium dilution as stress condition
Dilution of culture medium resulted in a proportional reduction of growth. In a series
of 18 pools (56 cultures), a 50% growth reduction was achieved at a dilution of
culture medium to an end concentration of 30%. The number of clones at day 8
was reduced from 10.1 ± 5.0 (mean ± S.D.) in controls to 5.3 ±3.2 (mean ± S.D.) in
the diluted medium. A dilution of culture medium to 30% end concentration has
therefore been used as the appropriate stress condition in the following
experiments.
Determining thresholds for stress testing of individual AF cell cultures
Thresholds for stress testing were determined from 174 individual primary in-situ
cultures from 6 different series, cultured in medium diluted to an end concentration
of 30%. On average 5.2 ± 5.0 (mean ± S.D.) clones were counted after 8 days of
culturing. Sixteen of the 174 cultures failed to grow, i.e. a 9.2% culture failure. The
average number of culture failures was 2.6 out of 30 within the 6 experimental
series. If we set the threshold for the minimally acceptable number of successful
cultures to 23 out of 30, the observed variability indicates that this would lead to
less than 1.0% wrongly rejected lots. In order to determine a threshold for the
minimal mean number of clones, we will accept that 1% of the rejected lots of
medium will be rejected wrongly. Considering the variability among the cultures,
this results in a threshold of on average 3 clones in 30 cultures. For the exact
derivation the reader is referred to the Appendix.
These thresholds were validated with different types and lots of medium
(Table 1). Basic medium F10 supplemented with 20% foetal calf serum was
Chapter 2
B.Sikkemaproefschrift2.doc
62
rejected under stress testing, as expected. The other media were tested as
acceptable.
Table 1: Results of stress testing of individual amniotic fluid in-situ cell cultures
from different types and lots of medium
Culture medium Day ofanalysis
No. of clones(mean ± S.D.)
No. of culturefailures
Acceptedfor routine
Amniomax (lot 1198269) day 6 3.1 ± 2.45 2
day 8 4.0 ± 2.9 1 yes
Amniochrome (lot 3MB0203) day 6 2.1 ± 1.3 2
day 8 3.0 ± 1.9 3 yes
Amniochrome (lot 4MB0020) day 6 4.0 ± 2.3 1
day 8 6.7 ± 5.3 0 yes
Amniochrome (lot 3MB0221) day 6 3.0 ± 2.7 1
day 8 5.0 ± 4.0 1 yes
F10 with 20% FCS day 6 1.0 ± 0.98 12
day 8 1.4 ±1.4 11 no
Stress testing of decreased oxygen tension during culturing
Using stress testing, the average number of clones at day 8 when cultured at 2.5%
O2 was 5.8 versus 4.8 for control cultures. This difference is significant (P< 0.036,
paired T- test, two sided). For 5% O2 we observed 7.7 clones versus 7.0 clones in
control cultures. This difference was not significant (P> 0.429, paired T- test, two
sided). The number of successful cultures at 2.5% O2 was 29 versus 27 in control
cultures and at 5% O2 28 versus 26 in control cultures under 20% O2. The number
of clones of medium or large size was higher under decreased oxygen tensions of
both, 2.5% and 5%.
Cell culture of amniotic fluid cells
B.Sikkemaproefschrift2.doc
63
Table 2: Analysis of variance for the effect of decreased oxygen tension on thegrowth of amniotic fluid cells
Dependent variable Explainedvariance
(%) 1)
Effect Sum ofsquares
DF Meansquare
F- ratio P (twotailed)
Growth (undilutedmedium)
Culturetime
22.0 3 7.3 128.6 0.000
20.0 O2 tension 5.7 1 5.7 100.5 0.000
Error 1.1 19 0.06
Total 28.8 23
Growth (medium of30% end-concentration)
Culturetime
12.1 3 4.0 52.3 0.00
36.0 O2 tension 7.8 1 7.8 101.1 0.000
Error 1.5 19 0.08
Total 21.4 231)variance in difference between number of cells cultured under 5% and 20% oxygen tension
To measure the increase in number of cells over time, growth curves were
made, both with decreased oxygen tension (2.5%; 5% and 10%) and 20% oxygen
(control concentration). A significant increase in the number of cells was measured
at all investigated conditions of decreased oxygen. At 5% (Figure 1a and b) the
effect was most pronounced, with a 2.5 fold average increase, explaining 20% of
the variance (Table 2). Within this condition, the largest increase was seen at days
5 and 6 (Figure 1a). In the same experiment, but using diluted medium with end
concentration of 30%, the increase was even 4.5 fold, explaining 36% of the
variance (Figure 1b).
Chapter 2
B.Sikkemaproefschrift2.doc
64
Figure 1a: Growth curves of secondary amniotic fluid cells with undiluted medium
0 5 6 7 8Day
0
10
20
30
40
Num
ber o
f cel
ls /m
l x 1
0.00
0
20%5%
Standard deviation I
Oxygen tension
Figure 1 b: Growth curves of secondary amniotic fluid cells with medium diluted to 30%
0 5 6 7 8Day
0
5
10
15
Num
ber o
f cel
ls /m
l x 1
0.00
0
5%20%
Standard deviation I
Oxygen tension
Cell culture of amniotic fluid cells
B.Sikkemaproefschrift2.doc
65
2.3.5 Discussion
In this study we describe an absolute method for testing the influence of any factor
involved in in-situ cultures of primary amniotic fluid cells on the growth potential.
Stress testing appears more sensitive to detect differences in growth potential than
testing without a stress condition, as is apparent from the stronger response of
cultures to a decreased oxygen tension. This supports the validity of using stress
testing as a method to determine acceptance or not of different lots of medium for
cell culture for prenatal diagnosis.
Whereas other absolute procedures to test the growth potential of medium,
such as plating efficiency (Goodheart et al., 1973), incorporation of radionuclides
during cell growth (Dvorakova and Starek, 1980) or an end point dilution method
(Shu-de and Hampson, 1985) require a high degree of technical skill or elaborate
equipment, the method we propose is easy to incorporate in daily routine.
Our stress testing is based on absolute thresholds for the minimum number
of clones and for the minimum number of successful cultures after 8 days of
culturing. This implies a high probability that no medium will be accepted with less
than the required quality, while rejection of lots that would have had sufficient
quality if tested more extensively, will be minimal. The acceptance criterion is
absolute and not subject to any drift. We recommend to use at least 30 cultures
from at least 10 different patients. This should be sufficient to detect important
differences in quality.
A further advantage of this method is the use of non-pooled primary AF in-
situ cell cultures. Therefore, the growth behaviour of the cells during testing is likely
more representative for the conditions present in routine cultures. To demonstrate
that the absolute method can readily be used for an assessment of the effect of
other factors than culture medium influencing growth potential and to demonstrate
the high sensitivity of stress testing, we investigated the effect of a decreased
oxygen tension during culturing. Two early reports (Brackerts et al., 1983; Held and
Soennichsen et al., 1984) have appeared in which reduced oxygen supply during
Chapter 2
B.Sikkemaproefschrift2.doc
66
culturing resulted in an up to 25% increase in the number of clones and in a
significant increase of the number of cells. In stress testing we could only confirm
this for the lowest oxygen tension (2.5%), but we did find that the size of the clones
was larger under both conditions of decreased oxygen. This was in agreement with
the results of growth curve testing which showed that the number of cells was 2.5-
fold increased, explaining 20% of the variance and under stress conditions even
36% of the variance when culturing under a 5% oxygen tension. The difference in
number of clones between the former results and ours might be explained by our
use of hormone-supplemented medium as introduced by Chang et al. in 1982,
whereas in the earlier experiments a basic medium supplemented with foetal calf
serum only was used. The hormone-supplemented medium already maximises the
growth of the viable cells from AF, giving a maximum number of clones (Chang et
al., 1991). Thus, with respect to the number of clones, the effect of the culturing
medium is much stronger than that of a decreased oxygen tension. The strongest
effect had a decreased oxygen tension on cultures with a small number of clones
and on the increase in number of cells during the first days of culturing, as seen in
our series. This finding may suggest that cell culturing under a decreased oxygen
tension appears to be more successful.
In general, to assess the growth potential of different lots of culture medium
for acceptance or rejection in routine diagnosis, we recommend stress testing. To
test the growth potential of a new type of medium or any other factor involved in
culturing, stress testing should be performed first. In case of an increased growth
potential, the number of cells can subsequently be determined very precisely, if
desired, by making a growth curve.
Cell culture of amniotic fluid cells
B.Sikkemaproefschrift2.doc
67
2.3.6 Literature
Brackertz M, Kubbies M, Feige A, Salk D. Decreased oxygen supply enhances growth in culture of
human mid-trimester amniotic fluid cells. Hum Genet. 1983; 64(4):334-8.
Chang HC, Jones OW, Masui H. Human amniotic fluid cells grown in a hormone-supplemented
medium: suitability for prenatal diagnosis. Proc Natl Acad Sci U S A. 1982; 79(15):4795-9.
Chang HC, Jones OW. Enhancement of amniocyte growth on a precoated surface. Prenat Diagn.
1991;11(6):357-70.
Dvorakova M, Starek M. A method for testing the growth promoting property of tissue culture media
using radionuclides. J Biol Stand 1980; 8: 107- 113.
Goodheart CR, Casto BC, Zwiers A. Plating efficiency for primary hamster embryo cells as an index of
efficacy of foetal bovine serum for cell culture. Appl Microbiol 1973; 26: 525- 528.
Held KR, Sonnichsen S. The effect of oxygen tension on colony formation and cell proliferation of
amniotic fluid cells in vitro. Prenat Diagn. 1984; 4(3):171-9.
Hoehn H, Bryant EM, Karp LE, Martin GM. Cultivated cells from diagnostic amniocentesis in second
trimester pregnancies. II. Cytogenetic parameters as functions of clonal type and preparative technique.
Clin Genet. 1975; 7(1):29-36.
Priest JH, Rao KW. Prenatal chromosome diagnosis. Barch MJ, Knutsen T, Spurbeck (eds.) The AGT
Cytogenetics laboratory manual, Lippincott – Raven Publishers Philadelphia 1997.
Pye D. Simple and rapid evaluation of serum for cell culture. J Biol Stand. 1975; 3(1):83-7.
Salk D, Disteche C, Stenchever MR, Mitchell L, Rodriguez M. Routine use of Chang medium for
prenatal diagnosis: Improved growth and increased chromosomal breakage. Am J Hum Genet 1983;
35: 151A.
Shu-de J, Hampson AW. A simple endpoint dilution method for evaluating serum used for cell culture. J
Biol Stand 1985;13: 303-308.
Sikkema-Raddatz B, van Echten J, van der Vlag J, Buys CHCM, te Meerman GJ. Minimal volume of
amniotic fluid for reliable prenatal cytogenetic diagnosis. Prenat Diagn. 2002; 22(2):164-5.
Chapter 2
B.Sikkemaproefschrift2.doc
68
Appendix
Determination of the thresholds:
To set a threshold for the minimum mean number of clones, we will reject a new lot
of medium if we find significantly less than the average of 5.2 clones which are
found growing after 8 days of culture. We accept that 1% of the rejected lots of
medium has by chance less than this number.
Justification:
K is the minimum mean number of clones found growing in 30 separate culture
experiments.
We assume that the mean number of clones is distributed according to a T
distribution with 29 degrees of freedom. The one-sided threshold of finding a value
occurring in less than 1% of the cases = -2.462. This relates to the acceptable false
rejection rate.
The standard deviation of the number of clones is 5. The standard deviation of the
mean number of clones in 30 experiments is therefore 5/(sqrt(30). The following
equations show the derivation of the minimum acceptable mean number of clones
(K- 5.2) / (5/√30) > -2.462
(K- 5.2) x 1.09 > -2.462
K> 5.2 – 2.258
K > 2.94
Considering the variability among the cultures, this results in a threshold of an
average of 3 clones in 30 cultures.
To set a threshold for the minimum of successful cultures, we observed that on
average 2.6 cultures failed in a series of 30. We think that this number of failures is
on average acceptable. The number of actual culture failures might be higher than
expected due to culture failures by chance. We again accept that 1 of the 100 lots
will be rejected falsely when the real number of culture failures is 2.6 out of 30. In
30 experiments the number of failures follows a binomial distribution with
Cell culture of amniotic fluid cells
B.Sikkemaproefschrift2.doc
69
probability 0.0833. The probability that we find more than 7 failed cultures is then <
0.01.
If we set the threshold for the minimally acceptable number of successful cultures
to 23 out of 30, this would lead to less than 1.0% wrongly rejected lots.
Chapter 2
B.Sikkemaproefschrift2.doc
70
Slide preparation of in-situ cell cultures from amniotic fluid cells and chorionic villi
B.Sikkemaproefschrift2.doc
71
Chapter 3
Slide preparation of in-situ cellcultures from amniotic fluid cells andchorionic villi
Chapter 3
B.Sikkemaproefschrift2.doc
72
Slide preparation of in-situ cell cultures from amniotic fluid cells and chorionic villi
proefdruk1.doc
73
Assessment of the influence offixative, temperature, humidity and airflow on metaphase quality inpreparations from primary in-situcultures of amniotic fluid cells andchorionic villi
Birgit Sikkema- Raddatz1; Harry Moes 1;2; Marian Stoepker1; Bauke de Jong1;
Charles H.C.M. Buys 1and Gerard J. te Meerman1
1 Department of Medical Genetics, University of Groningen, The Netherlands
2Present address:
Department of Obstetrics and Gynaecology, University Hospital Groningen, The
Netherlands
Chapter 3
B.Sikkemaproefschrift2.doc
74
3.1 Abstract
The quality of metaphase chromosomes in preparations from amniotic fluid (AF)
and chorionic villus sample (CVS) in situ cell cultures for prenatal cytogenetic
diagnosis depends on various factors at harvesting. To determine the quantitative
effect of fixative, ambient temperature, relative humidity and air flow, primary AF
cell cultures and primary CVS in situ cultures were harvested under a range of
different conditions in a chromosome harvest chamber. We applied analysis of
variance and regression to establish the relation of these conditions with the
degree of metaphase spreading and the presence of cytoplasm. We found that
ambient temperature alone had no effect on metaphase quality. Relative humidity
explained 18.6% of the variance in the presence of cytoplasm and 7.5% of the
variance in the degree of metaphase spreading in AF cell cultures. In CVS
cultures, these percentages were 7.2% and 4.2 %, respectively. In AF cell cultures,
air flow had a limited influence on metaphase quality. In CVS cultures, however,
this factor was responsible for 29.9% of variance. Different compositions of fixative
influenced the spreading of metaphases for 13.7% in AF cell cultures and 4.7% in
CVS cultures. We conclude that metaphase quality in preparations from primary in
situ cultures is only to a limited degree affected by the investigated factors at
harvesting. Within reasonable limits, conditions cannot be considered as critical.
Keywords:
prenatal diagnosis, in situ culture, metaphase quality; chromosome preparation
Slide preparation of in-situ cell cultures from amniotic fluid cells and chorionic villi
B.Sikkemaproefschrift2.doc
75
3.2 Introduction
Proper techniques for harvesting dividing cells and preparing slides for
chromosome analysis are critical for attaining the quality needed for correct
karyotype analysis (Lawce and Brown, 1991). In prenatal cytogenetic diagnosis a
reliable quality of (in-situ) preparations is essential, since failure of the analysis and
repeated removal of amniotic fluid and chorionic villi bear an increased risk of
foetal loss. An understanding of the factors influencing the process of chromosome
preparation is important for maintaining a good quality. Hliscs et al. (1997)
demonstrated that spreading of metaphases is a slow process, which leads to a
stretching of chromosomes via flattening. The process of spreading involves a
significant water-induced swelling of mitotic cells during evaporation of the fixative
from the slide (Claussen et al., 2002). The degree of spreading of metaphases
depends on several factors, among which the evaporation time of the fixative is
considered an imported one (Spurbeck et al., 1996). Evaporation time depends on
the composition of the fixative, on relative humidity and ambient temperature
(Spurbeck et al., 1996). The last two factors can be controlled by the use of a
special climate room (Lundsteen et al., 1985) or a chromosome harvest chamber.
The effect of several factors on metaphase spreading has been determined
from cell suspensions from different primary cells and cell lines (Deng et al., 2003,
Claussen et al., 2002; Henegariu et al., 2001) and secondary AF cell cultures
(Spurbeck et al., 1996). For prenatal cytogenetic analyses, however, primary in situ
cultures are the most used and sensitivity of spreading for metaphases of these
cells is unknown. Therefore, we investigated the effect of relative humidity, of
ambient temperature, of air flow and of the composition of fixative at harvesting of
primary in-situ cell cultures (amniotic fluid (AF) cells and chorionic villus sample
(CVS) cultures). The quality of metaphases was determined using as parameters
the degree of spreading and the presence of cytoplasm.
Chapter 3
B.Sikkemaproefschrift2.doc
76
3.3 Materials and Methods
Culturing
Primary AFcells were cultured according to the in situ method on 30 mm coverslips
in 35 mm Petri dishes. The culture medium was Amniomax-C100 (Invitrogen – Life
Technologies, Breda, The Netherlands) and Chang A (Irvine Scientific, USA) with
α-MEM. Cultures were checked for cell growth after 5 days. They were harvested
when approximately 10 clones were observed, which occurred on average after 8
days.
Chorionic villi were treated with trypsin (Cambrex Company, Verviers, Belgium)
and collagenase (Sigma-Aldrich Chemie Gmbh; Steinheim, Germany) before being
cultured. The cell suspension was then transferred to coverslips (30 mm in 35 mm
Petri dishes) and cultured in Amniomax and Chang with α-MEM. Cultures were
checked for cell growth after 4 days. When growth was adequate, cells were
harvested, on average after 6 days.
Harvesting
Cultures were treated with BrdU (final concentration in culture, 36 µg/ml) and
colcemid (final concentration in culture, 0.015 μg/ml) for 16 hours (overnight).
Subsequently, cultures were harvested with a Tecan 5031 robotic harvester
(Tecan, Maennedorf, Switzerland) using 0.8% (w/v) sodium citrate as hypotonic
solution and methanol and glacial acetic acid (3:1, v/v) as fixative. During the last
fixation step the cultures were transferred to the chromosome harvest chamber
(C30, Percival Scientific, Inc., Perry Iowa, USA). The fixative was removed by
suction and the cultures were allowed to dry.
Harvesting conditions
To quantify the effects of ambient temperature, relative humidity and air flow at
harvesting, cultures were harvested under a range of conditions. The ambient
temperature ranged from 200C to 300C (intervals of 20C). The relative humidity
ranged from 25% to 60% (intervals of 5%). The effect of air flow (0.2; 0.4; 0.6 and
Slide preparation of in-situ cell cultures from amniotic fluid cells and chorionic villi
B.Sikkemaproefschrift2.doc
77
0.7 m/s) was tested at established optimal conditions (a temperature of 260C and
35% relative humidity, respectively). Because of the technical limits of the harvest
chamber it was not feasible to investigate more extreme conditions, e.g. 200C and
25% relative humidity (Table 1).
Table 1a: Summary of the results of metaphase quality scoring in AF cell culturesat different climate conditions
Relative humidity (%)
25 30 35 40 45 50 60
Tempera-ture (0C)
20 not real. not real. not real. not real. 54 metaph. not determ. 55 metaph.
44% 16%
22 not real. not real. not real. 72 metaph. 71 metaph. not determ. not determ.
47% 56%
24 not real. not real. 58 metaph. 88 metaph. 50 metaph. not determ. 55 metaph.
67% 33% 48% 20%
26 not real. 81 metaph. 101metaph. 72 metaph. 72 metaph. 67 metaph. not determ.
63% 78% 57% 16% 7%
28 not real. 69 metaph. 69 metaph. 70 metaph. 59 metaph. 57 metaph. not determ.
37% 58% 64% 2% 24%
30 56 metaph. 60 metaph. not determ. 70 metaph. not determ. not determ. 60 metaph.
41% 42% 47% 4%
metaph., metaphases; not real.; not realisable because of technical limits of the harvest chamber;
not determ., not determined
56 metaph. means that 56 metaphases were screened
41% means that 41% of the metaphases had an optimum quality (well-spread, no cytoplasm)
Chapter 3
B.Sikkemaproefschrift2.doc
78
Table 1b: Summary of the results of metaphase quality scoring in CVS cultures atdifferent climate conditions
Relative humidity (%)
25 30 35 40 45 50 60
Tempera-ture (0C)
20 not real. not real. not real. not real. 60 metaph. not determ. 60 metaph.
17% 26%
22 not real. not real. not real. 60 metaph. 58 metaph. not determ. not determ.
18% 12%
24 not real. not real. 58 metaph. 84 metaph. 72 metaph. not determ. 60 metaph.
47% 29% 45% 5%
26 not real. 67 metaph. 96 metaph. 72 metaph. 72 metaph. 47 metaph. not determ.
42% 38% 25% 10% 19%
28 not real. 72 metaph. 72 metaph. 72 metaph. not determ. 60 metaph. not determ.
34% 37% 18% 13%
30 57 metaph. 60 metaph. not determ. not determ. 72 metaph. not determ. 60 metaph.
26% 46% 20% 12%
metaph., metaphases; not real., not realisable because of technical limits of the harvest chamber;
not determ.; not determined
57 metaph. means that 57 metaphases were screened
26% means that 26% of the metaphases had an optimum quality (well-spread, no cytoplasm)
The different ambient temperatures, relative humidities and air flow conditions were
set by adjusting the control panel of the Percival harvest chamber. Relative
humidity was calibrated with a psychrometer. The air flow was measured with an
anemometer.
The culture slides were aged at 900C for one hour and stained with Giemsa after
treatment with 0.1% pancreatine (w/v) (Sigma-Aldrich Chemie Gmbh; Steinheim,
Germany) by an automatic staining device (Midas, Merck, Whitehouse Station,
USA).
Slide preparation of in-situ cell cultures from amniotic fluid cells and chorionic villi
B.Sikkemaproefschrift2.doc
79
Quality scoring of metaphases
Metaphase quality was determined by scoring the metaphases for the degree of
spreading of chromosomes and for the presence of cytoplasm under different
conditions.
For each condition of ambient temperature and relative humidity a minimum of
50 metaphases on a total of five different culture slides were scored. Per slide a
minimum of 10 metaphases were screened. For AF cell cultures, only one
metaphase (the best as judged by visual observation) per clone was selected
(Table 1). For the different ambient temperatures and relative humidities a total of
1371 metaphases were screened for AF cell cultures and 1391 for CVS cultures.
For each condition of air flow a minimum of 60 metaphases on a total of five
different slides was scored. Per slide a minimum of 12 metaphases were screened.
For AF cell cultures, only one metaphase (the best as judged by visual
observation) per clone was selected. For the effect of the air flow a total of 413
metaphases (AF cell cultures) and 423 metaphases (CVS cultures) were
examined.
Scoring for the degree of spreading of a metaphase was as follows;
1: good (a maximum of 5 overlaps of chromosomes), 2: over-spread (not al
chromosomes in one view by 1,000 times magnification) and 3: bad (more than 5
overlaps of chromosomes).
Scores for the degree of presence of cytoplasm were;
1: good (no visible cytoplasm), 2: moderate (haze of cytoplasm) and 3: bad (cell
membrane still visible).
These well-defined scores allowed an independent application of the scoring
system by different observers. Two persons independently scored all metaphases.
Chapter 3
B.Sikkemaproefschrift2.doc
80
Different compositions of fixative
The prenatal cultures were harvested as described before and dried at 260C, 35%
relative humidity and 0.6 m/s air flow. Methanol and methanol- glacial acetic acid in
proportions of 1:1(v/v); 1:2(v/v); 1:3(v/v); 1:4(v/v) and 1:5(v/v) were used as
fixatives. For each condition a minimum of at least 60 metaphases on a total of
five different culture slides was scored. Per slide a minimum of 12 metaphases
were screened. For AF cell cultures, only one metaphase (the best as judged by
visual observation) per clone was selected.
For each condition the degrees of metaphase spreading and presence of
cytoplasm were determined as described before.
Statistical analysis
Analyses of variance and regression were used to assess the quantitative effect of
the conditions. The software package Systat 5.01 was used for statistical analyses.
The combined influence of ambient temperature and relative humidity, the
quadratic effects of ambient temperature alone and relative humidity alone, the
quadratic effect of air flow and the quadratic effect of the composition of fixative
were analysed. A general linear model was used to evaluate the data.
3.4 Results
Quantitative effects of ambient temperature, relative humidity and air flow
The effects of ambient temperature, relative humidity (Table 2) and air flow (Table
3) on the degree of presence of cytoplasm and on the degree of metaphase
spreading showed differences between AF cell and CVS cultures. In CVS cultures
ambient temperature and relative humidity together explained 9.1% of the variation
in cytoplasm and 4.2% of the variation in spreading. In AF cell cultures these
percentages were 20.2% and 8.0%, respectively. Relative humidity alone showed
the greatest influence; 18.6% explained variance in AF cell cultures and 7.2% in
Slide preparation of in-situ cell cultures from amniotic fluid cells and chorionic villi
B.Sikkemaproefschrift2.doc
81
CVS cultures. Ambient temperature alone had a minimal effect on metaphase
spreading and presence of cytoplasm. The air flow was the most important quality
factor for CVS cultures; 29.9% of the variance in cytoplasm was explained by this
factor. In contrast, AF cell cultures were hardly influenced by air flow.
Metaphase quality
No clearly optimal harvest conditions exist, as the effects of ambient temperature,
relative humidity and air flow on metaphase quality are relatively minor (Table 1).
There is no statistically significant interaction between the investigated factors.
For AF cell cultures, a low relative humidity resulted in a haze of cytoplasm and
a bad spreading. A high relative humidity resulted in bad quality metaphases,
visibility of the cell membrane and/ or a haze of cytoplasm and bad spreading. The
best quality was obtained with a relative humidity between 30% and 40%. Ambient
temperature had little effect on metaphase quality. High or low ambient
temperatures resulted in a lesser quality.
For CVS cultures, regression analysis showed only a very limited influence of
ambient temperature and relative humidity. Only 4% to 9% of the variation (Table
2) was due to the experimental conditions. Therefore, these conditions can be
considered as not critical under the investigated range of conditions. Air flow, the
third factor, did not affect the metaphase quality in AF cell cultures. In contrast to
that, air flow did influence metaphase quality in CVS cultures. In preparations from
these cultures, presence of cytoplasm was strongly reduced at 0.6 m/s (T=260C;
RH=35%)(Table3).
Chapter 3
B.Sikkemaproefschrift2.doc
82
Table 2: Results of regression analysis of non linear effects of temperature and/ or relativehumidity on metaphase quality
AF cell-cultures (n = 1371) CVS-cultures (n = 1391)
Effect on cytoplasm Coefficient (SD) P (two tailed) Coefficient (SD) P (two tailed)
Temperature -1.418 (1.138) 0.213 -5.864 (1.095) 0.000
Temperature x Temperature -0.012 (0.198) 0.951 0.998 (0.189) 0.000
Humidity -2.247 (0.304) 0.000 -0.995 (0.300) 0.001
Humidity x Humidity 0.187 (0.020) 0.000 0.077 (0.019) 0.000
Temperature x Humidity 0.329 (0.067) 0.000 0.166 (0.066) 0.012
Explained variance (%) 20.2 9.1
Humidity -0.094 (0.014) 0.000 -0.054 (0.013) 0.000
Humidity x Humidity 0.001 (0.000) 0.000 0.001 (0.000) 0.000
Explained variance (%) 18.6 7.2
Temperature -0.415 (0.098) 0.000 not determined not determined
Temperature x Temperature 0.008 (0.002) 0.000 not determined not determined
Explained variance (%) 1.4 not determined
Effect on spreading
Temperature -0.256 (0.926) 0.783 -0.547 (0.926) 0.555
Temperature x Temperature 0.117 (0.161) 0.466 0.069 (0.160) 0.666
Humidity -0.058 (0.247) 0.814 0.202 (0.253) 0.426
Humidity x Humidity 0.038 (0.016) 0.018 -0.024 (0.016) 0.138
Temperature x Humidity -0.047 (0.055) 0.391 0.041 (0.056) 0.460
Explained variance (%) 8.0 4.2
Temperature -0.309 (0.074) 0.000 -0.080 (0.072) 0.272
Temperature x Temperature 0.006 (0.001) 0.000 0.001 (0.001) 0.362
Explained variance (%) 1.3 0.7
Humidity -0.036 (0.011) 0.001 0.036 (0.011) 0.001
Humidity x Humidity 0.001 (0.000) 0.000 0.000 (0.000) 0.022
Explained variance (%) 7.5 4.2
Metaphase quality was measured for several climate conditions during slide preparation. Ambienttemperature ranged from 200C-300C and relative humidity from 25%-60%. Analysis of regressionbetween these conditions was performed to determine the effect on the degree of metaphase spreadingand the presence of cytoplasm.
Slide preparation of in-situ cell cultures from amniotic fluid cells and chorionic villi
B.Sikkemaproefschrift2.doc
83
Table 3: Analysis of variance of the effect of air flow on metaphase quality.
Dependentvariable
Effect Explainedvariance
(%)
Sum ofsquares
DF Meansquare
F -ratio
P (twotailed)
AFcell-cultures(n= 413)
Cytoplasm air flow 7.4 6.10 3 2.02 10.85 0.000
Error 76.17 409 0.19
Spreading air flow 1.9 1.96 3 0.65 2.66 0.048
Error 100.50 409 0.25
CVS-cultures(n = 423)
Cytoplasm air flow 29.9 38.59 3 12.86 59.64 0.000
Error 90.38 419 0.22
Spreading air flow 2.4 2.80 3 0.93 3.42 0.017
Error 114.45 419 0.27
Metaphase quality was measured for four different air flows during slide preparation; namely 0.2; 0.4;0.6 and 0.7m/s (at 35% relative humidity and 200C ambient temperature). Analysis of variance wasperformed between these conditions to determine the effect on the degrees of metaphase spreadingand presence of cytoplasm.
Different compositions of fixative
The effect of the composition of fixative on metaphase quality was different for AF
cell and CVS cultures (Table 4). In AF cell cultures, the composition of fixative
explained 13.7% of the variance in presence of cytoplasm and degree of
spreading. Spreading was most strongly affected, explaining 13.9% of the
variance. The presence of cytoplasm was not strongly dependent of these factors.
It explained only 1.5% of the variance. In CVS cultures, the composition of fixative
explained 4.7% of the variance in metaphase quality. Cytoplasm and spreading
were both minimally influenced (5% of the variance). The composition of fixative
was not critical for metaphase quality, although methanol alone gave a bad result.
Chapter 3
B.Sikkemaproefschrift2.doc
84
Table 4: Analysis of variance of the effect on the composition of fixative onmetaphase quality.
Dependentvariable
Effect Explainedvariance
(%)
Sum ofsquares
DF Meansquare
F- ratio P (twotailed)
AFcell cultures(n=904)
Cytoplasm fixative 1.5 0.02 4 0.004 1.76 0.135
Error 1.97 896 0.002
Spreading fixative 13.9 91.22 4 22.81 36.18 0.000
Error 566.67 899 0.63
Cytoplasm +Spreading
fixative 13.7 90.67 4 22.67 35.67 0.000
Error 571.26 899 0.64
CVS cultures(n=880)
Cytoplasm fixative 5.3 4.88 4 1.22 5.30 0.000
Error 200.95 872 0.23
Spreading fixative 5.1 22.86 4 5.71 11.81 0.000
Error 423.14 875 0.48
Cytoplasm +Spreading
fixative 4.7 35.54 4 8.89 10.75 0.000
Error 723.09 875 0.83
Metaphase quality was measured for five different compositions of fixative during harvesting; namely
methanol- glacial acetic acid 1:1; 1:2; 1:3; 1:4 and 1:5 (v/v) (at 35% relative humidity and 200C ambient
temperature). Analysis of variance was performed between these conditions to determine the effect on
the degrees of metaphase spreading and presence of cytoplasm.
3.5 Discussion
In this study we have quantified the effects of ambient temperature, relative
humidity, air flow and composition of fixative on metaphase quality in preparations
Slide preparation of in-situ cell cultures from amniotic fluid cells and chorionic villi
B.Sikkemaproefschrift2.doc
85
of primary in-situ cultures of CVS and AF cells. We found that our design was
sufficiently sensitive to detect effects. Most effects appeared to be minor. Whereas
in AF cell cultures, fixative, ambient temperature and relative humidity together are
of greater influence, in CVS cultures the air flow is more important. In the absence
of a statistically significant interaction, the effects can be considered as additive.
Relative humidity appeared in our study as the most important factor for
the degree of presence of cytoplasm (18.6% explained variance) and to a minor
extent as a factor influencing the degree of spreading (7.5% explained variance)
for AF cell cultures. Deng et al. (2003) determined relative humidity as the most
important factor for spreading (degree of presence of cytoplasm was not
measured) as well. Increasing the relative humidity resulted in better metaphase
quality, at least up to a certain degree, after which higher humidity no longer
resulted in better spreading. The latter effect might be explained by a fast increase
in water content of the fixative, so that a dramatic rehydration takes place, too fast
to result in a good spreading of the chromosomes (Deng et al., 2003).
Ambient temperature alone had no effect on metaphase quality in our study. An
explanation might be that evaporation of methanol from the fixative leads to a
cooling of the remaining fluid on the slide, a process that cannot be totally
compensated for by controlling ambient temperature (Claussen et al., 2002).
The composition of fixative seems to be of minor importance, as seen in
our results (14% explained variance for AF cells). Use of methanol only led to
hardly spread metaphases. It is clear that the hygroscopic effect of acetic acid is
needed for water-induced swelling. Higher concentrations of acetic acid resulted in
a better spreading. We routinely add pure glacial acetic acid in case of badly
spread metaphases to improve spreading. This effect has, however, not been
systematically investigated in the present study.
In CVS cultures, air flow has a substantial effect (29.9%) on the presence
of cytoplasm, whereas it has no effect in AF cell cultures. In contrast to the latter,
cultures of primary chorionic villi show a much more dispersed growth, often
Chapter 3
B.Sikkemaproefschrift2.doc
86
including larger cell clusters, resulting in an uneven evaporation of the fixative
during harvesting and in variations in metaphase quality. Therefore, more extreme
forces, such as produced by the air flow from a fan might lead to a more dynamic
washing of the cells by evaporating fixative and a thinner and more equal layer of
fixative (Hliscs et al., 1997). This may result in a better overall metaphase quality.
According to the results of Claussen et al. (2002), the potential of swelling of
different tissues is related to the lengths of the resulting chromosomes.
Preparations of primary CVS cultures show short chromosomes, comparable to
those prepared from bone marrow. This explains the very limited influence of the
investigated factors on the degree of spreading.
The difference in influences of ambient temperature, relative humidity and
composition of fixative as found in previous reports and in our study may be related
to differences in the type of cell culture. In cell suspensions, Deng et al. (2003)
found a strong influence of the relative humidity on the spreading of metaphases.
Claussen et al. (2002) also demonstrated that in cell suspensions the composition
of fixative had a great effect on metaphase spreading. However, in these studies,
the magnitude of the effect on metaphase quality has not been quantified.
During slide preparation, cells in suspension stick to the glass surface after
they have been dropped onto the slide. Evaporation of methanol from the fixative
causes a shrinking of the cells. Subsequently, a water-induced swelling
(rehydration) of the cells occurs by the hygroscopic effect of acetic acid. During this
last step, cells are flattening and the remaining fixative evaporates. The process of
swelling is crucial for elongation and spreading of the chromosomes. Slide
preparation from in-situ culture follows in principle the same steps. However, cells
from in-situ cultures are already sticking onto the slide. Within an AF cell clone for
instance, they do not grow as a homogeneous layer and swelling of mitotic cells
can be hampered by lack of space. Cells harvested in-situ thus, have a more
limited capacity to rehydrate. The influence of different parameters on slide
preparation may, therefore, in general be more restricted for in-situ cultures than
for cells in suspension.
Slide preparation of in-situ cell cultures from amniotic fluid cells and chorionic villi
B.Sikkemaproefschrift2.doc
87
In our experiments, we could not find statistically significant optimum
conditions for harvesting, because of the limited influence of the factors analysed.
In our routine, we recommend a higher than the standard concentration (25% v/v)
of glacial acetic acid for a better metaphase spreading and a relative humidity of
30% – 40% for a minimal presence of cytoplasm in metaphase preparations from
primary prenatal in-situ cultures. Spurbeck et al. (1996) determined the optimal
harvest conditions for secondary AF cultures to be an ambient temperature of 200C
and a relative humidity of 55%, based on the criterion of the size of metaphase
area. However, they recommend to use in practice 250C and 50% relative humidity,
because they found fairly frequently an over-spreading, indicating that metaphase
area alone is not properly reflecting metaphase quality. The discrepancy in relative
humidity for optimal harvesting between the results of Spurbeck et al. (1996) and
ours might be explained by their use of secondary in-situ cultures contrary to our
use of primary cultures. As previously discussed, primary in-situ cultures may have
a lower capacity of swelling in general, leading to a reasonable spreading at a
lower relative humidity than in secondary in-situ cultures or cell suspensions.
According to their own figures, they did not find a clear optimum for different
harvest conditions, but only small differences between the various harvesting
conditions they used. In that respect, their results are consistent with ours.
In conclusion, in general, metaphase quality of prenatal in-situ cultures is
only to a limited extent affected by ambient temperature, relative humidity, air flow
and the composition of fixative. Harvesting conditions are therefore not critical
within the investigated range and for good metaphase quality expensive measures
for fine controlling of ambient temperature, relative humidity and air flow are not
required.
Chapter 3
B.Sikkemaproefschrift2.doc
88
3.6 Literature
Claussen U, Michel S, Muhlig P, Westermann M, Grummt UW, Kromeyer-Hauschild K, Liehr T.
Demystifying chromosome preparation and the implications for the concept of chromosome
condensation during mitosis. Cytogenet Genome Res. 2002; 98(2-3):136-46.
Deng W, Tsao SW, Lucas JN, Leung CS, Cheung AL. A new method for improving metaphase
chromosome spreading. Cytometry. 2003; 51A(1):46-51.
Henegariu O, Heerema NA, Lowe Wright L, Bray-Ward P, Ward DC, Vance GH. Improvements in
cytogenetic slide preparation: controlled chromosome spreading, chemical ageing and gradual
denaturing. Cytometry. 2001; 43(2):101-9.
Hliscs R, Muhlig P, Claussen U. The spreading of metaphases is a slow process which leads to a
stretching of chromosomes. Cytogenet Cell Genet. 1997; 76(3-4):167-71.
Lawce HJ and Brown MG. Harvesting, slide- making, and chromosome elongation techniques; In Barch
M J (ed): The AGT Cytogenetics Laboratory Manual. Second Edition, Raven Press New York 1991: 31-
66.
Lundsteen C, Lind AM. A test of a climate room for preparation of chromosome slides. Clin Genet.
1985; 28(3):260-2.
Spurbeck JL, Zinsmeister AR, Meyer KJ, Jalal SM. Dynamics of chromosome spreading. Am J MedGenet. 1996; 61(4):387-93.
Chromosomal analysis in preparations from chorionic villi and amniotic fluid cell cultures
B.Sikkemaproefschrift2.doc
89
Chapter 4
Chromosomal analysis in preparationsfrom chorionic villi and amniotic fluidcell cultures
Chapter 4
B.Sikkemaproefschrift2.doc
90
Chromosomal analysis in preparations from chorionic villi and amniotic fluid cell cultures
B.Sikkemaproefschrift2.doc
91
Chapter 4.1
A 46,XX,der(13;14)(q10;q10),+21child born after a 45,der(13;14)(q10;q10) chromosomal finding inCVS
B. Sikkema- Raddatz, C.C. Verschuuren- Bemelmans, M. Kloosterman* and B. de
Jong
Department of Medical Genetics, University of Groningen, Groningen, The
Netherlands
*Department of Prenatal Diagnosis, Rijnstate Hospital Arnhem, Arnhem,
The Netherlands
Prenat Diagn 1997; 17: 1085-1089
Chapter 4
B.Sikkemaproefschrift2.doc
92
Letter to the Editor:
A 38- year- old gravida 5, para 2 woman underwent chorionic villus sampling
(CVS) at 10+3 weeks of gestation because she is a carrier of a Robertsonian
translocation (13;14). Chromosome analysis was performed after direct
preparation. A female chromosome pattern with a Robertsonian translocation
45,XX, der(13;14)(q10;q10) was found in 15 GTG-banded metaphases.
The pregnancy continued uncomplicated and a female infant with clinical
manifestations of Down syndrome was born spontaneously at 38 weeks.
Cytogenetic analyses of PHA- stimulated lymphocyte culture showed 46,XX, der
(13;14)(q10;q10), +21 (5 metaphases). Fluorescence in situ hybridisation (FISH)
with the probe YAC259H11 showed 3 signals for chromosome 21 in 100
metaphases.
Slides of the direct chorionic villus preparation were subsequently re-
examined. An additional 112 metaphases were seen, all without a trisomy 21.
There was no chorionic villus culture available.
Discordant findings between CVS and the foetal tissue have been documented
before. The majority of the reported cases have been false-positive results (CVS
abnormal, child normal). In a very few instances, a false-negative result (CVS
normal, child abnormal) has been reported. There are several mechanisms by
which the false-negative results may have arisen, including:
- a post-zygotic non-disjunction in the inner cell mass during embryogenesis
(Crane and Cheung, 1988);
- a twin pregnancy, with loss and re-absorption of the normal foetus;
- a mosaic situation, undetected because of the low mitotic index; and
- contamination of the villi with normal maternal cells (Pindar et al.,1992).
In 1987 Simoni et al. made a rough estimation about the incidence (1 per
thousand) of a false-negative result. Lilford et al. (1991) suggest a frequency of
Chromosomal analysis in preparations from chorionic villi and amniotic fluid cell cultures
B.Sikkemaproefschrift2.doc
93
0.001per cent using the (semi)direct method. The case described here is the first
false-negative result in our series of 7000 CVS patients by the direct and 1000
patients by the culture method.
Table 1: False-negative results in CVS for sex-chromosomal abnormalities
Foetus Direct CVS Culture Indication ReferenceFibroblasts:
47,XYY[23]/46,XY[39] 46,XY[85] 47,XXY[43]/46,XY[7] ? Callen et al. (1988)
Blood:
47,XXY[20]/46,XY[58] 46,XY[85] 47,XXY[58]/46,XY[7] Maternal age Eichenbaum et al.(1986)
Blood:
47,XXY[?]/46,XY[?] 46,XY[10] 47,XXY[?] Maternal age Linton and Lilford(1986)
Blood:
47,XXY[7]/46,XY[13] 46,XY[15] 47,XXY[10]/46,XY[5] ? Miny et al. (1991)
Amniotic fluid:
45,X[28]/46,XY[32] 46,XY[12] 45,X[21]/46,XY[5] Maternal age Smidt-Jensen andLind (1987)
Different tissues:
45,X[9]/46,XX[281] 46,XX[50] 45,X[50] Previouschild +21
Schlesinger et al.(1990)
Blood:
47,XXX[47]/46,XX[3] 46,XX[35] --- Maternal age Verjaal et al.(1987)
Chapter 4
proefdruk1.doc
94
Tab
le2
– F
alse
-neg
ativ
ere
sults
inC
VS
for
tris
omie
s18
and
21
Fet
usD
irect
CV
SC
ultu
reIn
dica
tion
Ref
eren
ce
Dif
fere
nttis
sues
:47
,XY
,+18
[47]
46,X
Y[3
7]47
,XY
,+18
[30]
Mat
erna
lag
eM
artin
etal
.(1
986)
Am
niot
icflu
id:
47,X
Y,+
18[9
]46
,XY
[15]
47,X
Y,+
18[5
]?
Min
yet
al.
(199
1)B
lood
:47
,XX
,+18
[10]
/46,
XX
[2]
46,X
X[1
0]47
,XX
,+18
[15]
/46,
XX
[15]
US
abno
rmal
ities
at30
wee
ksLe
scho
teta
l.(1
988)
Blo
od:
47,X
X,+
18[5
6]46
,XX
[16]
–U
Sab
norm
aliti
esat
31w
eeks
Wirt
zet
al.
(198
8)F
ibro
blas
ts:
47,X
X,+
18[3
0]45
,XX
[4]
46,X
X[1
1]M
ater
nal
age
Pin
dar
etal
.(1
992)
Blo
od:
47,X
Y,+
21[9
9]46
,XY
[50]
–M
ater
nal
age
Bar
tels
etal
.(1
989)
Am
niot
icflu
id:
47,X
Y,+
21[1
5]/4
6,X
Y[7
2]46
,XY
[25]
47,X
Y,+
21[9
0]/4
6,X
Y[1
0]?
Cal
len
etal
.(1
988)
Dif
fere
nttis
sues
:47
,XY
,+21
[300
]46
,XY
[110
]47
,XY
,+21
[100
]M
ater
nal
age
Lilfo
rdet
al.
(199
1)A
mni
otic
fluid
:47
,XX
,+21
[12]
/46,
XX
[24]
46,X
X[2
5]47
,XX
,+21
[11]
Mat
erna
lag
eM
iny
etal
.(1
988)
Blo
od:
47,X
X,+
21[6
]/46,
XX
[94]
46,X
X[1
00]
47,X
X,+
21[8
]/46,
XX
[24]
Pre
viou
sch
ild+
21N
isan
iet
al.
(198
9)B
lood
:47
,XX
,+21
[17]
46,X
X[2
0]–
Mat
erna
lag
eS
imon
ieta
l.(1
987)
Am
niot
icflu
id:
47,X
X,+
21[?
]46
,XX
[?]
47,X
X,+
21[?
]?
Vej
ersl
ev&
Mik
kels
en(1
989)
Blo
od:
46,X
X,d
er(1
3;14
),+21
[105
]45
,XX
,der
(13;
14)[
129]
–M
othe
rde
r(13
;14)
Pre
sent
ed c
ase
(199
7)
Chromosomal analysis in preparations from chorionic villi and amniotic fluid cell cultures
B.Sikkemaproefschrift2.doc
95
To our knowledge, 20 cases (excluding structural abnormalities) of a false-
negative finding including our case have been published: seven cases with a sex-
chromosomal abnormality (Table 1); five cases with trisomy 18; and eight cases
with trisomy 21 (Table 2). In five cases, the (semi)direct method only was
performed. In most of the cases (11 times), the culture method gave the right result
whereas the (semi)direct method gave a false-negative result. In three cases, the
(semi)direct method was normal, the culture was non-mosaic abnormal, and
chromosome studies of the blood of the child showed a mosaic sex- chromosomal
abnormality. In the case reported by Pindar et al. (1992), the (semi)direct as well
as the culture method gave a false-negative result (46,XX), whereas the child had
a 47,XX,+18 chromosomal pattern. By this new case it is again demonstrated that
the most favourable technique for accurate prediction is the combination of the
(semi)direct and culture method. According to these data, the culture method
alone is a very good second choice.
References
Bartels,I.; Hansmann,I.; Holland, U.; Zoll, B.; Rauskolb, R.(1989) Down syndrome at birth not detected
by first trimester chorionic villus sampling, Am. J. Med. Genet. 34, 606-607.
Callen, D.F.; Korban, G.; Dawson, G.; Gugasyan, L.; Krumins, E.J.M.; Eichenbaum,S.; Petrass,J.;
Purvis- Smith,S.; Smith, A.; Den Dulk, G.; Martin, N.(1988) Extra embryonic/ fetal karyotypic
discordance during diagnostic chorionic villus sampling, Prenat. Diagn. 8, 453- 460.
Crane, J.P., Cheung, S.W. (1988) An embryogenetic model to explain cytogenetic inconsistencies
observed in chorionic villus versus fetal tissue, Prenat. Diagn. 8, 119- 129.
Eichenbaum, S.Z.; Krumins, E.J.; Fortune, D.W.; Duke, J. (1986) False negative finding on chorionic
villus sampling, The Lancet, 391.
Leschot, N.L.; Wolf, G.; Weenink, G.H. (1988) False negative findings at third trimester chorionic villus
sampling, Clin. Genet. 34, 204- 205.
Lilford, R.J., Caine, A., Linton, G., Mason, G. (1991) Short term culture and false negative results for
Down's syndrome on chorionic villus sampling, The Lancet 337, 861.
Linton, G.; Lilford, R.J. (1986) False negative finding on chorionic villus sampling, The Lancet, 630.
Chapter 4
B.Sikkemaproefschrift2.doc
96
Martin, O.A.; Elias, S.; Rosinsky, B.; Bombard, A.T.; Simpson, J.L. (1986) False negative finding on
chorionic villus sampling, The Lancet, 391- 392.
Miny, P.; Basaran, S.; Horst, J.; Pawlowitzki, I.H.; Kim Nhan Ngo, T. (1988) False negative cytogenetic
result in direct preparations after CVS, Prenat. Diagn. 8, 633.
Miny, P.; Hammer, P.; Gerlach, B.; Tercanli, S.; Horst, J.; Holzgreve, W.; Eiben, B. (1991) Mosaicism
and accurancy of prenatal cytogenetic diagnosis after chorionic villus sampling and placental biopsies,
Prenat. Diagn. 11, 581-589.
Nisani, R.; Chemke, J.; Voss, R.; Appelman, Z.; Caspi, B.; Lewin, A.; Dar, H.; Reiter, A. (1989) The
dilemma of chromosomal mosaicism in chorionic villus sampling - direct versus long term cultures,
Prenat. Diagn. 9, 223- 226.
Pindar, L.; Whitehouse, M.; Ocraft, D. (1992) A rare case of a false negative finding in both direct and
culture of a chorionic villus sample, Prenat. Diagn. 12, 525- 527.
Schlesinger, C.; Raabe, G.; Ngo,T.; Miller, K. (1990) Discordant findings in chorionic villus direct
preparation and long term culture- mosaicism in the fetus, Prenat. Diagn. 10, 609- 612.
Simoni, G.; Fraccaro, M.; Gimelli,G.; Maggi, F.; Dagna Bricarelli, F. (1987) False positive and false
negative findings on chorionic villus sampling, Prenat. Diagn. 7, 671- 672.
Smidt-Jensen, S.; Lind, A.M. (1987) A case of first trimester chromosomal mosa-icism confined to the
cultivation of the gestational products, Clin. Genet. 32, 133- 136.
Vejerslev, L.O.; Mikkelsen, M.(1989) The European Collaborative study on mosaicism in chorionic villus
sampling: data from 1986 to 1987, Prenat. Diagn. 9, 575-588.
Verjaal, M.; Leschot, N.J.; Wolf, H.; Treffers, P.E. (1887) Karyotyping differences between cells from
placenta and other fetal tissues, Prenat. Diagn. 7, 343- 348.
Wirtz, A.; Seidel, H.; Brusis E.; Murken, J.(1988) Another false- negative finding on placental sampling,
Prenat. Diagn. 8, 321.
Chromosomal analysis in preparations from chorionic villi and amniotic fluid cell cultures
B.Sikkemaproefschrift2.doc
97
Chapter 4.2
Four years’ cytogenetic experiencewith the culture of chorionic villi
B. Sikkema- Raddatz, K. Bouman, C.C. Verschuuren- Bemelmans, M. Stoepker, A.
Mantingh*, J.R. Beekhuis# and B. de Jong
Department of Medical Genetics, University of Groningen
*Department of Obstetrics and Gynaecology, Antenatal Diagnosis Unit, University
Hospital Groningen, Groningen, The Netherlands
# Department of Obstetrics and Gynaecology, Isala Klinieken, Zwolle, The
Netherlands
Prenat Diagn 2000; 20: 950-955.
Chapter 4
B.Sikkemaproefschrift2.doc
98
4.2.1 Summary
In 1958 chorionic villus samples, investigated by culture method, we found 137 (7%)
abnormalities. The abnormal results were classified in certain abnormal (generalised
abnormal at high probability) and uncertain abnormal (potentially confined to the
placenta) results. Certain abnormal were 73 cases (3.7%). Uncertain abnormal were
64 cases (3.3%), in which confirmation studies were done in 47 cases. In 12 cases of
these 47, the abnormality was confirmed and in 35 cases (1.8%) the abnormality was
confined to the placenta. Among the latter cases, poor pregnancy outcome [(16%
intrauterine death (IUD), 6% intrauterine growth retardation (IUGR)] was increased.
Total maternal cell contamination was not seen.
The positive predictive value of all confirmed abnormal cases was 66%. The positive
predictive value was 100% for indications ‘ultrasound abnormalities’ and ‘carrier’ and
between 50 and 60% for all other indications. Predictive value among uncertain
abnormal cases was low (26%). However, the positive predictive value depends of the
type of abnormality. Therefore we conclude that the culture method for chorionic villi is
a good test for indications ‘ultrasound abnormalities’ and ‘carrier’ and reliable for all
other indications. Whether or not follow-up investigations should be offered to the
parents depends of the type of abnormality. We conclude that the culture method is
reliable for prenatal diagnosis and can be used as the sole investigated method.
Key words:
chorionic villus sampling (CVS); confined placental mosaicism (CPM); culture method;
predictive value
Chromosomal analysis in preparations from chorionic villi and amniotic fluid cell cultures
B.Sikkemaproefschrift2.doc
99
4.2.2 Introduction
Chorionic villus sampling (CVS) is widely used as a method for the prenatal
detection of chromosome abnormalities. Two approaches were developed for
cytogenetic diagnosis: a (semi)direct method (Simoni et al., 1983) and a culture
method (Niazi et al., 1981; Heaton et al., 1984). However, soon after the
introduction of CVS it transpired that cytogenetic diagnosis did not always reflect
the chromosomal constitution of the foetus. In these cases the aberration was
confined to the placenta, so-called confined placental mosaicism (CPM) (Kalousek
and Dill, 1983). Reports on false positive and false negative results have been
published (Brambati et al., 1985; Breed et al., 1986; Eichenbaum et al., 1986).
Various (collaborative) studies (Ledbetter et al., 1992; Association of Clinical
Cytogeneticists Working Party on Chorionic Villi in Prenatal Diagnosis, 1994;
Hahnemann et al., 1997a; Smith et al., 1999) described cytogenetic findings and
pregnancy outcome of the (semi)direct and/ or culture method. These studies
reported on CPM (false negative and false positive findings), maternal cell
contamination (MCC) and calculations of the predictive value of the method
employed. A combination of the (semi)direct and culture method was recommended
to achieve the highest predictive value. When using only one method, the culture
method should be the method of choice because of fewer false negative and/ or false
positive findings (Ledbetter et al., 1992).
Formerly in our laboratory we used the direct method for routine cytogenetic
diagnosis. In case of abnormalities, the culture was also investigated (Breed et
al.,1990).
Based on the results of collaborative studies and on a false negative result in our
laboratory (Sikkema- Raddatz et al., 1997) we decided to change our policy. We
skipped the direct method and now use only the culture method, without increasing
turn around time of diagnosis. For economic reason it is not possible to use both
culture and direct method.
Chapter 4
B.Sikkemaproefschrift2.doc
100
In the present study we describe our results with the culture method over a 4-year
period. We describe the CVS success rate, MCC, the aberrations, and calculate the
predictive value (false positive results).
4.2.3 Materials and Methods
Between 1994 and 1998 we received 1958 chorionic villus samples. The majority of
the samples (1723) came from the University Hospital in Groningen. The other 235
samples came from the hospital "Isala Klinieken" in Zwolle.
CVS was performed transcervically between 10 and 12 weeks of gestation as
described previously (Breed et al., 1990). Exceptionally, CVS was performed
transabdominally in cases of more than 12 weeks’ gestation. Chorionic villi were
treated with trypsin and collagenase before being cultured (according to the method of
Jackson et al., 1989). The cell suspension was then transferred to coverslips and
cultured in Amniomax and Chang with α-mem. After adequate growth, cultures were
harvested after 6 days on average. Karyotyping was routinely performed after
Pancreatin- Giemsa staining. Other additional staining techniques, such as DA-DAPI-
staining, AgNOR- staining, C- banding and FISH were used, for confirming a suspect
chromosomal diagnosis after Pancreatin- Giemsa staining or when a rapid diagnosis
was required.
At least 15 cells were analysed. This number was extended to 29 cells in case of
a single cell with a trisomy 8, 9, 13, 18, 20, 21 or 22, a supernumerary marker
chromosome (ESAC) or a single cell with gain or loss of a sex chromosome.
A normal result was given in cases of 46,XX or 46,XY, in case of the common
inv(9)(p11q13), in cases of less than 50% of tetraploid cells and/ or in cases of one
cell exhibited any abnormality. All other results were considered abnormal.
Abnormal cytogenetic results
Between the abnormal results a distinction was made in non- mosaic cases and
mosaic cases.
Chromosomal analysis in preparations from chorionic villi and amniotic fluid cell cultures
B.Sikkemaproefschrift2.doc
101
A cytogenetic result was defined as mosaic when two or more cells showed a
karyotype different from the karyotype(s) in the other cells. Mosaic monosomy was
considered when three cells all lacked the same chromosome. In cases of a mosaic
result the parents were told that further investigation in amniotic fluid was required to
evaluate the result in chorionic villi.
The non-mosaic results were classified as structural and numerical abnormal
results. In cases of a structural abnormal result parental karyotyping was done, if the
structural abnormality was not known to be familial. In cases of a numerical abnormal
result further investigation in amniotic fluid was carried out to evaluate the result in
chorionic villi. An exception was made in cases of trisomy 21 and in cases of
ultrasound abnormalities in combination with a cytogenetic abnormality.
Certain abnormal cytogenetic results
Abnormal results were defined as certain when a non-mosaic trisomy 21, a non-
mosaic sex chromosomal trisomy, a non-mosaic triploidy, a non-mosaic structural
abnormality, or a non-mosaic ESAC was found. In addition, a non-mosaic 45,X
pattern and a trisomy 18 in combination with ultrasound abnormalities were defined as
certain abnormal. Trisomy 13 was not seen in this study, but would have been
counselled similar to trisomy 18.
Uncertain abnormal cytogenetic results
Abnormal results were defined as uncertain when a non-mosaic trisomy 18 and a
45,X pattern without ultrasound abnormalities, a non-mosaic autosomal trisomy other
than 13, 18 or 21 and all kinds of mosaic patterns (both numerical and structural) were
found.
Chapter 4
B.Sikkemaproefschrift2.doc
102
Predictive value
Predictive value was defined as the probability of an abnormal karyotype of the foetus,
given that CVS showed either a mosaic and/ or non-mosaic abnormality.
The 95% confidence intervals (CI) of the predictive value (p) with a standard error
SE(p) were calculated as follows:
CI(p)= p ± 1.96 * SE(p) and SE(p)= √p(1- p) / n.
Follow-up studies
In cases of uncertain abnormal results, follow-up studies (chromosomal study of
(un)cultured amniocytes, parental peripheral blood karyotyping or ultrasound
investigations) were offered to the parents.
Amniotic fluid cells were cultured according to the in situ method. Clones (n= 29)
were analysed to exclude at least 10% of a mosaic pattern (95% CI). In some cases
fluorescence in situ hybridisation (FISH) was performed on metaphase chromosomes
and/or on uncultured amniocytes confirming suspect chromosomal results and/ or for
rapid diagnosis. DNA studies were performed to investigate the possibility of
uniparental disomy (UPD) in abnormal CVS cases in which chromosome 15 was
involved.
Peripheral blood lymphocytes were prepared according to standard techniques
for karyotyping.
First and second trimester ultrasound investigations were performed by a
gynaecologist subspecialized in prenatal diagnosis with the use of third level
equipment (Acuson TM Corporation, Mountain View, CA, USA, 128xP10).
In cases of termination of pregnancy (TOP) we always tried to confirm the
cytogenetic diagnosis in foetal fibroblasts. Information about the course and outcome
of pregnancy was received from the woman and/or the referring midwife or physician.
Birth weights of babies were categorised according to the tables of Kloosterman
(1970).
Chromosomal analysis in preparations from chorionic villi and amniotic fluid cell cultures
B.Sikkemaproefschrift2.doc
103
4.2.4 Results
From the 1958 samples, a successful diagnosis was obtained in 99,7%. Of the six
failures, four involved small samples (≤5 mg). In one case cells did not attach on
coverslips. In the final case cell growth was very poor; only two cells could be exami-
ned.
Normal karyotypes were found in 1815 cases. Follow-up revealed no false
negative results. Abnormal karyotypes were found in 137 cases (7%) [Table 1]. Single
cell abnormalities were counted in 362 cases (18.5%).
Certain abnormal cytogenetic results
A total of 73 cases (3.7%) were certain abnormal (Table 2), including 34 cases of
structural abnormalities and 39 cases of numerical abnormalities. Confirmation
studies of the 73 certain abnormal cases were done in 58 cases.
Uncertain abnormal cytogenetic results
A total of 64 cases (3.3%) were uncertain abnormal (Table 3): 16 non-mosaic and 48
mosaic cases (2.45%). Confirmation studies were done in 47 cases. In 12 of these
cases the abnormality was confirmed. Thus of the 47 cases of expected CPM, 35
cases (1.8%) were indeed confined to the placenta. Not confirmed were two cases of
trisomy 16 and 2 cases with a 45,X. In one of the 45,X cases, a 46,XX pattern was
found in foetal tissue. Maternal contamination could not be excluded entirely in this
tissue. In the other case, a mos 45,X/46,XX pattern was found in foetal tissue after
TOP. In the confirmation studies of the mosaic abnormalities, 3/5 cases with a mosaic
sex chromosomal abnormality and 1/4 cases with a mosaic ESAC were confirmed.
Chapter 4
proefdruk1.doc
104
Tab
le1
–N
umbe
rof
abno
rmal
resu
ltsfo
rdi
ffere
ntin
dica
tions
Non
-mos
aic
case
sM
osai
cca
ses
Tris
-47
,XX
X69
,XX
XT
ris-
Str
uctu
ral
rear
rang
emen
tsA
utos
omal
tris
omy
Indi
catio
nN
umbe
rof
case
sN
orm
alca
ses
omy
2147
,XX
Y47
,XY
Y69
,XX
Y69
,XY
Yom
y18
45,X
Unu
sual
tris
omy
Bal
-an
ced
Un-
bala
nced
ES
AC
13;1
8;21
Oth
erS
ex-
chro
mos
omal
Str
uc-
tura
lE
SA
CT
etra
-pl
oid
MA
36-3
99
MA
>40
8U
Sab
na5
Car
rierb
1D
NA
/BIO
c
Oth
erd
2A
ll
1212
373
47 33 114
173
1952
1154
342
32 14 110
163
1815
25
3 1 4
1 6 7
4 4 1 9
3 2 1 2 8
2 2
8 3 13 1 3 28
1 1 2 4
2 2
1 3 4
12 5 1 1 19
5 1 6
4 2 1 1 8
1 3 1 1 6
2 2 1 5
aF
etal
abno
rmal
ities
atul
tra
soun
d.bP
aren
t al c
arr ie
rshi
p fo
r st
ruct
ural
rea
rran
gem
ent.
c Cyt
ogen
etic
inve
stig
atio
nse
cond
ary
to D
NA
orbi
oche
mic
alin
vest
igat
ion.
dP
revi
ous
child
with
chro
mos
omal
abno
rmal
ityor
oth
erre
ason
s fo
rpre
nata
lcyt
ogen
etic
dia
gnos
is.
MA
,M
ater
nala
ge(y
ears
).
Chromosomal analysis in preparations from chorionic villi and amniotic fluid cell cultures
B.Sikkemaproefschrift2.doc
105
Predictive value
Of all 137 abnormal cases, confirmation studies were done in 105 cases. Of the 105
cases (58 certain and 47 uncertain abnormal cases) the abnormality was confirmed in
70 cases. Therefore predictive value for all confirmed cases was 66% (95% CI: 57-
75%) (Table 4).
The predictive value of the 58 certain abnormal cases was 100%. The predictive
value of the 47 uncertain abnormal cases was 26% (95 % CI: 14%- 38%). Predictive
values differed for different abnormalities (Table 3).
Predictive values for different indications are shown in Table 4. The indication ‘carrier’
resulted in nearly 100% because of one case of a non-mosaic inv(7) in combination
with a mosaic +inv(7) pattern. The mosaic +inv(7) was confined to the placenta and
not confirmed in amniotic fluid.
Maternal cell contamination (MCC)
There was no case of incorrect sex prediction. In 71 cases (3.6%) a XX/XY admixture,
an indication of MCC, was found. All male foetuses were carefully screened for XX-
cells (on average 100 metaphases). In most of these cases (n=51) less than ten cells
of MCC were seen. In 16 cases MCC was between 10 and 50%. It was remarkable
that this high percentage was seen in only one of the two culture systems in eight
cases. The other culture system was without MCC. In four cases MCC was > 50%, in
one case it was up to 70%.
Chapter 4
proefdruk1.doc
106
Table 2 – Certain abnormal results (generalised abnormal at high probability)
Abnormality
Numberof cases(%)
Abnormality confirmed/number of confirmationstudies
Positive predictivevalue(%)
Autosomal aneuploidyTrisomy 21 25 (1.3) 14/14 100Trisomy 18 1 (0.05) 1/1 100
Sexchromosomal aneuploidy45,X 2 (0.09)47,XXX 2 (0.09) 1/1 10047,XXY 2 (0.09) 2/2 100
Triploidy 7 (0.36) 6/6 100Structural abnormality
Familial balanced 13 (0.7) 13/13 100Familial unbalanced 2 (0.09) 2/2 100Possible de novo
Balanced 15 (0.8) 15/15 100Unbalanced 2 (0.09) 2/2 100ESAC 2 (0.09) 2/2 100
Total 73 (3.7) 58/58 100
Table 4 – Positive predictive value of all confirmed cases for different indications
IndicationNumberof cases
Number ofabnormal cases(%)
Abnormality confirmed/number of confirmationstudies
Positive predictivevalue (%)(95% CIe)
MA 36 ≥ 39 1212 58 (4.79) 27/49 55 (41-69%)MA ≥ 40 373 31 (8.33) 10/19 53 (31-78%)US abnormalitiesa 47 15 (31.91) 11/11 100 (41-100%)Carrierb 33 19 (57.58) 17/18 94 (83-100%)DNA/BIO c 114 5 (4.51) 2/3 66 (12-100%)Otherd 173 10 (5.78) 3/5 60 (17-100%)Total 1952 137 (7.00) 70/105 66 (57-75%)
aFetal abnormalities at ultrasound.bParental carriership for structural rearrangement.cCytogenetic investigation secondary to DNA or biochemical investigation. dPreviouschild with chromosomal abnormality or other reasons for prenatal cytogenetic diagnosis.e95% Confidence interval.MA, Maternal age (years).
Table 3 – Uncertain abnormal results (potentially confined to the placenta)
Abnormality
Numberof cases(%)
Abnormality confirmed/number of confirmationstudies
Positive predictivevalue(%)
95%confidence
Non-mosaicsAutosomal
Trisomy 18 8 (0.4) 6/6 100 20-100%Trisomy 16 2 (0.1) 0/2 0
Sexchromosomal45,X 6 (0.3) 2/4 50 1-99%
MosaicsAutosomal 23 (1.2) 0/19 0Sexchromosomal 6 (0.3) 3/5 60 13-100%Tetraploidy 5 (0.3) 0/3 0Structural 14 (0.7) 1/8 12.50 0-35.5%
Total 64 (3.3) 12/47 26 13.5-38.5%
Chromosomal analysis in preparations from chorionic villi and amniotic fluid cell cultures
B.Sikkemaproefschrift2.doc
107
Pregnancy outcome of uncertain abnormal cytogenetic results
Of the 64 uncertain abnormal cases confirmation studies were done in 47 cases. In 35
cases the abnormality was not confirmed in cytogenetic follow-up studies. Of these
cases 26 (74%) had a normal pregnancy outcome, four (11%) ended in an intra
uterine death (IUD), one pregnancy (3%) was terminated, three (9%) had intrauterine
growth retardation (IUGR) and one (3%) developed HELLP (haemolysis, elevated
liver enzymes, low platelet count) syndrome. The cytogenetic results of the four
pregnancies that ended in IUD were 45,X, trisomy 16, a mosaic tetraploidy and a
mosaic trisomy 7. The cytogenetic results of the three pregnancies with IUGR were a
mosaic trisomy 2, a mosaic trisomy 21 and a mosaic 45X/46,XX.
4.2.5 Discussion
This study confirms the results of previous studies in terms of cytogenetic findings and
pregnancy outcome. The negative predictive value is 100% since we had no cases
with false negative results. The positive predictive value of 66% in this study is
comparable to the results of Breed et al. (1990), the US collaborative study (Ledbetter
et al.; 1992) and Los et al. (1998), which were 64%, 68.4% and 67.8% respectively.
The positive predictive value of 75% in the EUCROMIC study (Hahnemann and
Vejerslev,1997) is somewhat higher. This is by far the largest study: 62 865 cases.
Therefore, the predictive value from the EUCROMIC study might be the most
accurate. However, comparison of the accuracy should be done with great caution
since the method used and the indication for CVS may differ.
Furthermore in the present study, positive predictive values were calculated for
different abnormalities. For trisomy 18 the 100% predictive value in our study is
according to results in the literature (Smith et al., 1999). At least 70 cases of trisomy
18 have been confirmed. As far as we aware with the culture method only one
discrepancy (Breed, 1992) has been described. Therefore we conclude that trisomy
18, detected in the culture method, should be counselled as a certain abnormal
cytogenetic result. For the (semi)direct method, however, several discrepancies (false
positive findings) have been described. Hahnemann and Vejerslev (1997b), for
Chapter 4
B.Sikkemaproefschrift2.doc
108
example, found seven cases in the EUCROMIC study. Unfortunately they did not
mention the total number of all trisomy 18 cases detected in the study. Therefore it
was not possible to calculate the positive predictive value. Breed (1992) gave a
predictive value of 80% (95% CI: 52.3- 94.9%) for the direct method. These figures
indicate that trisomy 18, detected by the (semi)direct method, is still an uncertain
abnormal cytogenetic result.
For 45,X we found at least a 50% positive predictive value. Since one case of
possible MCC and one case of a mosaic 45,X/45,XX pattern was not included in the
calculation, the predictive value might even be higher. The result of the present study
would then correspond with a predictive value of 78% for the culture method by
Pittalis et al. (1994). For the direct method 47% was given however.
For the mosaic cases the predictive value is very low in both the (semi)direct and
the culture method, although the number of mosaic cases seems to be lower in CVS
culture. Therefore in amniocytes in fewer cases a follow-up study is necessary. In the
present study in 35 cases (1.8%) a second prenatal test (amniocentesis) was
necessary. In the study of Los et al.(1998) this percentage was 2.1% for the
(semi)direct method.
Predicted value for indications ‘carrier’ and ‘US abnormalities’ have the expected
(nearly) 100%. All other indications have a predicted value of 50- 60%. Since we
found no differences in the predictive value for these indications, we do not advise
CVS only for women older than 40 years (Los et al., 1998) but for all woman of 36
years and older. Therefore we conclude that positive predictive value depends not on
the indication, but on the cytogenetic result found.
A CPM of 1.8% in the present study is consistent with frequencies of CPM of 1 -
2% in other studies (Association of Clinical Cytogeneticists Working Party on
Chorionic Villi in Prenatal Diagnosis, 1994; Leschot et al., 1996; Hahnemann and
Vejerslev, 1997; Los et al., 1998). We found seven cases with a mosaic structural
abnormality in the 35 CPM cases in our study. In these cases we speculate that the
aberration might be a culture artefact rather than a real abnormality confined to the
placenta. In such situations, especially in cases of balanced aberrations, in
Chromosomal analysis in preparations from chorionic villi and amniotic fluid cell cultures
B.Sikkemaproefschrift2.doc
109
combination with a pregnancy without any anomalies, it is arguable whether a follow-
up study in amniocytes should be advised, since the incidence of mosaic structural
abnormalities in the population is negligible (Leegte et al., 1998). Moreover, when a
mosaic structural abnormality is detected, the risk figure of mental retardation and/ or
congenital abnormalities is low. Taking this into consideration, in the present study
1.4% of CPM (28 cases) rather than 1.8% is the correct frequency.
In the present study, the overall frequency of MCC is high (3.6%) in comparison
to other studies. In the UK study (Association of Clinical Cytogeneticists Working
Party on Chorionic Villi in Prenatal Diagnosis,1994) 0.5% MCC and in the US study
(Ledbetter et al.,1992) 2.16% MCC at transcervical sampling was found. Only Smidt-
Jensen et al. (1993) found a higher percentage of MCC, namely 4.2%. The reason for
the high percentage in the present study could be that all male karyotypes were
carefully screened for XX cells. However, the percentage quoted in the collaborative
studies are not relevant for cytogenetic diagnosis. More important is the percentage of
MCC per patient. Only at a high percentage of MCC or total overgrowth of maternal
cells is it possible to miss the correct diagnosis in chorionic villi. In the present study
this was not the case. In the UK study, however, total MCC occurred in six individual
cases, while an overall percentage of 0.5% MCC was given.
In the present study we found IUD in 17% and IUGR in 5% of the CPM cases.
The risk of poor pregnancy outcome might be even higher if the mosaic structural
abnormal cases (probably culture artefacts) were excluded. This result confirms the
increased risk of poor pregnancy outcome described in the literature. Breed et
al.(1991) reported 16.6% IUD and Wapner et al. (1992) 8.6% IUD. However, Gold-
berg and Wohlferd (1997) were unable to correlate CPM with poor pregnancy
outcome. Since the number of CPM cases in the present study is small, the
percentage of poor pregnancy outcome is more an indication than a precise risk
figure. However, Morssink et al. (1996) measured elevated levels of maternal serum
human chorionic gonadotropin (MShCG) in pregnancies with CPM. This resulted in a
negative effect on placental function and foetal development.
Chapter 4
B.Sikkemaproefschrift2.doc
110
In conclusion, the reliability of the culture method alone is in the present study
comparable to the combination of (semi)direct and culture methods and more reliable
than the (semi)direct method alone. The predictive value is acceptable for all
indications. The predictive value for trisomy 18 and 45,X is higher in the culture
method than in the (semi)direct method alone. MCC should be closely investigated to
avoid maternal overgrowth. We therefore conclude that the culture method alone is
reliable for prenatal diagnosis and may be used as the sole investigative method.
Chromosomal analysis in preparations from chorionic villi and amniotic fluid cell cultures
B.Sikkemaproefschrift2.doc
111
4.2.6 References
Association of Clinical Cytogeneticists Working Party on Chorionic Villi in Prenatal Diagnosis .1994.
Cytogenetic analyses of chorionic villi for prenatal diagnosis: a collaborative study of U.K. data. Prenat Diagn
14: 363- 379.
Brambati B, Simoni G, Danesino C, Oldrini A, Ferrazzi E, Romitti L,Terzoli GL, Rosella F, Ferrari M,
Fraccaro M .1985. First trimester fetal diagnosis of genetic disorders: Clinical evaluation of 250 cases. J Med
Genet 22, 92- 99
Breed ASPM, Mantingh A, Govaerts L, Boogert A, Anders G, Laurini R. 1986. Abnormal karyotype in the
chorion, not confirmed in a subsequently aborted fetus. Prenat Diagn 6: 375- 377.
Breed ASPM, Mantingh A, Beekhuis JR, Kloosterman MD, Ten Bolscher H, Anders GJPA .1990. The
predictive value of cytogenetic diagnosis after CVS: 1500 cases: Prenat Diagn 10: 101 -110.
Breed ASPM, Mantingh A, Vosters J, Beekhuis JR, Lith JMM, Anders GJPA .1991. Follow-up and
pregnancy outcome after a diagnosis of mosaicism in CVS. Prenat Diagn 11: 577- 580.
Breed ASPM .1992. An evaluation of cytogenetic diagnosis by chorionic villus sampling: Reliability andimplications. STYX Publications
Eichenbaum SZ, Krumins EJ, Fortune DW, Duke J .1986. False negative finding on chorionic villus
sampling. Lancet 2: 391.
Goldberg JD, Wohlferd MM .1997. Incidence and outcome of chromosomal mosaicism found at the time of
chorionic villus sampling. AM J Obstet Gynecol 176: 1349- 1353.
Hahnemann JM and Vejerslev LO .1997. Accuracy of cytogenetic findings on chorionic villus sampling
(CVS)- Diagnostic consequences of CVS mosaicism and non- mosaic discrepancy in centres contributing to
EUCROMIC 1986- 1992. Prenat Diagn 17:9: 801- 820.
Hahnemann JM and Vejerslev LO .1997. European collaborative research on mosaicism in CVS
(EUCROMIC)- Fetal and extrafetal cell lineages in 192 gestations with cvs mosaicism involving single
autosomal trisomy. Am J Med Genet 70: 179- 187.
Heaton DE, Czepulkowski BH, Horwell DH, Coleman DV .1984. Chromosome analyses of first trimester
chorionic villus biopsies prepared by a maceration technique. Prenat Diagn 3: 279- 287
Jackson L, Gibas LM, Coutinho W.1989. Chorionic villus. In: Verma RS and Babu A (ed) Human
chromosomes, manual of basic techniques. Pergamon Press, Oxford: 19 - 26.
Kalousek DK, Dill FJ. 1983. Chromosomal mosaicism confined to the placenta in human conceptions.Science 221: 665- 667.
Chapter 4
B.Sikkemaproefschrift2.doc
112
Kloosterman GJ .1970. On intrauterine growth; the significance of prenatal care. Int J Gynaecol Obstet 8:
895- 912.
Ledbetter DH, Zachary JM, Simpson MS, Golbus MS, Pergament E, Jackson L, Mahoney MJ, Desnick R,
Schulman J, Copeland KL, Verlinsky Y, Yang-Feng T, Schonberg SA, Babu A, Tharapel A, Dorfman A,
Lubs HA, Rhodas GG, Fowler SE and de la Cruz F.1992. Cytogenetic results from the collaborative study on
CVS. Prenat Diagn 12: 137- 345.
Leegte B, Sikkema- Raddatz B, Hordijk R, Bouman K, van Essen, Castedo S, de Jong B .1998. Three casesof mosaicism for balanced reciprocal translocations. Am J Med Genet 79: 362- 365.
Leschot NJ, Schuring- Blom GH, van Prooijen- Knegt AC, Verjaal M, Hansson K, Wolf H, Kanhai HHH, van
Vugt JGM, Christiaens GCML .1996. The outcome of pregnancies with confined placental chromosome
mosaicism in cytotrophoblast cells. Prenat Diagn 16: 705- 712.
Los FJ, van den Berg C, van Opstal D, Noomen P, Braat APG, Galjaard RJH, Pijpers L, Cohen- Overbeek
TE, Wildschut HIJ and Brandenburg H .1998. Abnormal karyotypes in semi- direct chorionic villus
preparations of woman with different cytogenetic risks. Prenat Diagn18: 1023- 1040.
Morssink LP, Sikkema- Raddatz B, Beekhuis JR, de Wolf BTHM and Mantingh A .1996. Placental
mosaicism is associated with unexplained second – trimester elevation of MShCG levels, but not with
elevation of MSAFP levels. Prenat Diagn 16: 845- 851
Niazi M, Coleman DV, Loeffler FE .1981. Trophoblast sampling in early pregnancy. Culture of rapidly
dividing cells from immature placental villi. Br J Obtst Gynaecol 88: 1081- 1085
Pittalis MC, Dalpra L, Torricelli F, Rizzo N, Nocera G, Cariati E, Santarini L, Tibiletti MG, Agosti S, Bovicelli L,
Forabosco A. 1994. The predictive value of cytogenetic diagnosis after CVS based on 4860 cases with both
direct and culture methods. Prenat Diagn 14: 267- 278
Simoni G, Brambati B, Danesino C, Rosella F, Terzoli GL, Ferrari M, Fraccaro M, .1983. Efficient direct
chromosome analyses and enzyme determinations from chorionic villi samples in the first trimester of
pregnancy. Hum Genet 63: 349- 357.
Sikkema- Raddatz B, Verschuuren- Bemelmans CC, Kloosterman M and de Jong B. 1997. A
46,XX,der(13;14)(q10;q10),+21 child born after a 45,XX,der(13;14)(q10;q10) chromosomal finding in CVS.Prenat Diagn 17: 1085- 1089.
Smidt- Jensen S, Lind A-M, Permin M, Zachary JM, Lundsteen C, Philip J .1993. Cytogenetic analyses of
2928 CVS samples and 1075 amniocentesis from randomized studies. Prenat Diagn 13: 723- 740.
Wapner RJ, Simpson JL, Golbus MS, Zachary JM, Ledbetter DH, Desnick RJ, Fowler SE, Jackson LG, Lubs
H, Mahony RJ, Pergament E, Rhoads GG, Shulman JD, de la Cruz F .1992. Chorionic mosaicism:
association with fetal loss but not with adverse perinatal outcome. Prenat Diagn 12: 347- 355.
Chromosomal analysis in preparations from chorionic villi and amniotic fluid cell cultures
B.Sikkemaproefschrift2.doc
113
Chapter 4.3
Probability tables for exclusion ofmosaicism in prenatal diagnosis
Birgit Sikkema-Raddatz*, Sérgio Castedo*† and Gerard J. te Meerman*
*Department of Medical Genetics, University of Groningen, Groningen, The
Netherlands
† Department of Medical Genetics, IPATIUP, Medical Faculty of Porto, Portugal
Prenat Diagn 1997; 17 (2): 115-118
Chapter 4
B.Sikkemaproefschrift2.doc
114
4.3.1 Summary
The decision concerning the number of metaphases that need to be analysed to
detect mosaicism of a certain degree depends mainly, for the same confidence levels,
on the culture method used (in situ or flask methods). Several probability tables,
designed for either the in situ or the flask method, have been reported and can be
used to assist laboratories in making the decisions referred to above. However, there
are instances where part of the analysis is done using the in situ and flask methods. In
such situations, the previously published tables are of limited use. We have generated
new table that can be used in such situations, as well as in cases where only the flask
method is used.
Keywords:
amniotic fluid culture; chromosomal mosaicism; probability tables; prenatal diagnosis
4.3.2 Introduction
True chromosomal mosaicism is a rare finding in amniotic fluid cultures, but its
exclusion represents an obvious major concern for most laboratories involved in
prenatal diagnosis (Hsu et al., 1992).
Several tables have been generated to determine the optimum number of
metaphases to be counted to exclude (detect) chromosomal mosaicism for either the
in situ or flask culture method (Claussen et al., 1984; Cheung et al.,1990; Rischkind
and Risch; 1990; Featherstone et al., 1994). However, for laboratories (like ours)
using the in situ culture technique, there are instances where not enough colonies can
be analysed (or counted) to exclude mosaicism with the usual confidence. In these
situations, a back-up in situ culture is usually trypsinized and the respective cells
divided over culture flasks, from which chromosomal analysis will proceed further.
Although it is possible to count how many colonies were present in the coverslip
prior to trypsinization, after trypsinization it is no longer possible to know which
Chromosomal analysis in preparations from chorionic villi and amniotic fluid cell cultures
B.Sikkemaproefschrift2.doc
115
metaphases come from which colony. For the calculation of the number of cells that
need to be analysed from these cultures, the tables referred to above are of limited
use.
A similar difficulty can occur when using the flask method, in cases where a small
sample of amniotic fluid (hence with fewer cells than usual) is cultured.
We have calculated the number of metaphases that need to be analysed in order
to be 95 per cent certain that at least one cell of a desired number of colonies has
been observed. The same table can be used to determine the number of metaphases
required for accurate analyses using only the flask method.
4.3.3 Mathematical calculation
Assuming that a coverslip contains, prior to trypsinization, N equal-sized colonies of
cells, a sample of K cells is analysed after trypsinization and harvesting. The question
is: what is the probability that M colonies are represented by at least one cell in the
sample? For the exact derivation the reader is referred to the Appendix.
A Pascal program has been written to evaluate the recursive expression. The
result is stored as probability distributions for the number of colonies M from which
cells are observed, conditional on the number N of colonies present and the number
of cells K sampled. Then the number of colonies for which there is 95 per cent
certainty that they have been observed, given the number of cells sampled, is
computed. This computation is done downwards from the highest sample number
present, such that the lowest sample number is obtained for every entry in the table.
4.3.4 Results
Table 1 shows the number of metaphases that need to be analysed to be 95 per
cent certain that at least the horizontal number of different colonies have been
sampled, when the number of colonies present in the culture is given by the
vertical number.
Chapter 4
proefdruk1.doc
116
Tab
le1
– R
equi
red
num
ber
ofsa
mpl
esto
obse
rve
cells
from
Mdi
ffere
ntco
loni
esou
tof
Nco
loni
espr
esen
t.95
perc
ent
prob
abili
tyta
ble
No.
ofco
loni
esN
pres
ent
No.
ofco
loni
esM
need
ed
23
45
67
89
1011
1213
1415
1617
1819
2021
2223
2425
2627
2829
30
26
00
00
00
00
00
00
00
00
00
00
00
00
00
00
34
110
00
00
00
00
00
00
00
00
00
00
00
00
00
44
716
00
00
00
00
00
00
00
00
00
00
00
00
00
53
611
210
00
00
00
00
00
00
00
00
00
00
00
00
63
69
1427
00
00
00
00
00
00
00
00
00
00
00
00
73
58
1218
330
00
00
00
00
00
00
00
00
00
00
00
83
57
1115
2238
00
00
00
00
00
00
00
00
00
00
00
93
57
10*
1318
2644
00
00
00
00
00
00
00
00
00
00
010
35
79
1216
2231
510
00
00
00
00
00
00
00
00
00
011
35
79
1215
1925
3557
00
00
00
00
00
00
00
00
00
012
34
69
1114
1822
2939
630
00
00
00
00
00
00
00
00
013
34
68
1113
1620
2532
4470
00
00
00
00
00
00
00
00
014
34
68
1013
1619
2329
3648
760
00
00
00
00
00
00
00
015
34
68
1012
1518
2226
3240
5383
00
00
00
00
00
00
00
016
34
68
1012
1417
2024
2935
4458
900
00
00
00
00
00
00
017
34
68
912
1417
1923
2732
3948
630
00
00
00
00
00
00
018
34
67
911
1416
1922
2630
3542
5267
00
00
00
00
00
00
019
34
67
911
1316
1821
2428
3338
4656
720
00
00
00
00
00
020
34
67
911
1315
1820
2327
3136
4249
6077
00
00
00
00
00
021
24
67
911
1315
1720
2326
2934
3945
5364
820
00
00
00
00
022
24
57
911
1215
1719
2225
2832
3742
4857
6887
00
00
00
00
023
24
57
910
1214
1619
2124
2731
3539
4552
6073
00
00
00
00
024
24
57
910
1214
1618
2123
2630
3337
4248
5564
770
00
00
00
025
24
57
810
1214
1618
2023
2629
3236
4045
5159
6881
00
00
00
026
24
57
810
1214
1618
2022
2528
3135
3943
4854
6272
860
00
00
027
24
57
810
1213
1517
2022
2427
3033
3741
4651
5866
7690
00
00
028
24
57
810
1213
1517
1922
2427
2932
3640
4449
5461
6980
00
00
029
24
57
810
1113
1517
1921
2426
2932
3538
4247
5257
6473
840
00
030
24
57
810
1113
1517
1921
2326
2831
3437
4145
4955
6168
7788
00
0
*Exa
mpl
e:th
ean
alys
isof
ten
met
apha
ses
isne
cess
ary
tobe
atle
ast9
5pe
rcen
tsur
eth
atat
leas
tfiv
edi
ffer
entc
olon
ies
have
been
sam
pled
out
ofni
neeq
ual-s
ized
pres
ent.
colo
nies
Chromosomal analysis in preparations from chorionic villi and amniotic fluid cell cultures
B.Sikkemaproefschrift2.doc
117
4.3.5 Discussion
In an attempt to define the minimum number of metaphases which need to be
analysed for a reliable prenatal diagnosis, several calculations have been made taking
into account the culturing method used (in situ or flask methods) and the desired
confidence levels for exclusion of mosaicism (Claussen et al., 1984; Cheung et al.,
1990; Richkind and Risch, 1990; Featherstone et al., 1994).
Some of these calculations resulted in different probability tables, depending on
different assumptions concerning the culture method used. However, even for
cytogenetic laboratories (like ours) using the in situ method, there are instances where
not enough colonies can be analysed to allow exclusion of chromosomal mosaicism
with the usual confidence. In such situations, a back-up culture is usually trypsinized
and the cells are inoculated onto two new coverslips, one of which will be harvested
when a sufficient number of mitoses is observed.
Since this trypsinization step obviously disrupts colony integrity, the probability
tables used for the in situ method are no longer applicable to determine the number of
metaphases that still need to be analysed. On the other hand, since part of the
analysis has been done from an in situ culture, the published tables for the flask
method (refs.) are also not directly applicable in such circumstances.
We have therefore generated a new table that can be used to determine quickly
the number of metaphases that need to be analysed from a sample of amniotic fluid in
order to exclude chromosomal mosaicism with a predefined confidence level (95 per
cent). The table is of use for both the flask method alone and the situations referred to
above where part of the analysis is done from an in situ culture and part from a
trypsinized culture.
To illustrate the use of the table, we discuss a few applications. For a laboratory
using the in situ method and wishing to detect 20 per cent mosaicism with 95 per cent
confidence, 14 metaphases, derived from different colonies, need to be analysed
irrespective of the total fluid volume (Richkind and Risch, 1990). If, however, only ten
colonies could be analysed from the in situ culture, four additional colonies must be
sampled from the trypsinized culture to achieve a similar sensitivity level of mosaicism
Chapter 4
B.Sikkemaproefschrift2.doc
118
detection. The decision as to how many metaphases need to be analysed from the
trypsinized culture will obviously depend on the total number of colonies present prior
to trypsinization. This number can be obtained by counting the actual number of
colonies in the culture, or by using the estimation of Richkind and Risch (1990), i.e.,
assuming the formation of 1.5 or 2 colonies per ml of cultured amniotic fluid.
Supposing that there were eight colonies prior to trypsinization, seven metaphases
would need to be analysed to be 95 per cent confident that at least four different
colonies had been sampled.
The total sensitivity will be somewhat lower (90.25 per cent) since the two
sensitivities (95 per cent confidence to detect mosaicism and 95 per cent confidence
to sample different colonies) need to be multiplied.
Our calculations, like those of Claussen et al. (1984), Richkind and Risch (1990),
Cheung et al. (1990), and Featherstone et al. (1994), obviously assume equal colony
sizes and contributions to the pool of analysable metaphases in the trypsinized
culture. When there are colonies of different size, only medium or large-sized colonies
should be counted and used in the calculation since these are more likely to
contribute to the pool referred to above.
The probability table (Table 1) presented here can also be used in laboratory
setting using only the flask method. Accordingly, if one wants to be 95 per cent
confident that a mosaicism ≥ 20 per cent is detected, and if the total amount of fluid
collected was 20 ml, a total number of 23 metaphases have to be analysed, assuming
the presence of 30 "colony-forming" cells in the cultured sample (Richkind and Risch,
1990). Smaller amounts of amniotic fluid imply, for the same sensitivity and
confidence levels, the analysis of more metaphases.
The calculations presented here, and corresponding probability tables, provide
simple guidelines that can help laboratories standardise the confidence levels of
prenatal diagnosis, either when they rely on the flask method alone or when, despite
using the in situ method, they need to use trypsinized cultures to complete the
analysis.
Chromosomal analysis in preparations from chorionic villi and amniotic fluid cell cultures
B.Sikkemaproefschrift2.doc
119
4.3.6 References
Cheung S.W., Spitznagel E., Featherstone T., Crane J.P. (1990). Exclusion of chromosomal mosaicism in
amniotic fluid cultures: Efficacy of the in situ versus the flask techniques, Prenat. Diagn., 10, 41- 57
Claussen U., Schäfer H., Trampisch H.J. (1984). Exclusion of chromosomal mosaicism in prenatal
diagnosis, Hum. Genet., 67, 23- 28
Featherstone T., Cheung S.W., Spitznagel E., Peakman D. (1994). Exclusion of chromosomal mosaicism in
amniotic fluid cultures: determination of number of colonies needed for accurate analysis, Prenat. Diagn., 14;
1009- 1017
Hsu L.Y.F., Kaffe S., Jenkins E.C., Alonso L., Benn P.A., David K., Hirschhorn K., Lieber E., Shanske A.,
Shapiro L.R., Schutta E., Warburton D.(1992). Proposed guidelines for diagnosis of chromosome mosaicism
in amniocytes based on data derived from chromosome mosaicism and pseudomosaicism studies, Prenat.
Diagn., 12, 555- 573
Richkind K.E. and Risch N.J. (1990). Sensivity of chromosomal mosaicism detection by different tissue
culture methods, Prenat. Diagn., 10; 519- 527
Chapter 4
B.Sikkemaproefschrift2.doc
120
Appendix
The question is: what is the probability that M colonies are represented by at least one
cell in the sample?
Solution (assuming an equal contribution of the different colonies)
If one cell is sampled, one colony is present. We represent the probability that M
colonies are present in a sample of K cells, assuming N colonies by P(M,K,N). M may
vary from 1 (at least one colony is represented in the sample) to the lower of the
numbers K and N, denoted min (K,N) (there may not be more colonies than there are
samples and there may not be more colonies than there are colonies).
The computation of P(M,K,N) is recursive. Assume that we want to compute
P(M,K,N). This probability is related to the probability that in K-1 samples, M and M-1
colonies are present:
P(M,K,N)= (N - M + 1)/
N x P(M - 1,K - 1,N) + (M)/
N x P(M,K -1,N) if K≤N,
assuming that P(M,K,N)=0 if M > min(K,N) or M < 1.
The recursion stops at P(1,1,N)=1.0.
The factor (N - M + 1)/N relates to the probability that a cell is observed from the
not yet observed pool of cells, while the factor (M)/N is the probability of observing a
cell from an already observed pool.
Summary,future prospects, and suggested guidelines
B.Sikkemaproefschrift2.doc
121
Chapter 5
Summary, future prospects, andsuggested guidelines
Chapter 5
B.Sikkemaproefschrift2.doc
122
5.1 Summary
The scope of the thesis is to establish guidelines for quality controlled processing
of amniotic fluid cells and chorionic villi obtained for purposes of prenatal
cytogenetic diagnosis. We, therefore, have investigated which factors are of critical
importance and we have quantified their influence. We have introduced controlled
experimental variations in the workflow in order to find main and possibly
interactive effects. The results will be discussed with respect to the three parts that
can be distinguished in the cytogenetic process, namely cell culture, slide
preparation and chromosomal analysis.
Factors influencing the quality of cell culture
Amniotic fluid as obtained by amniocentesis following the guidelines of the Dutch
Society of Obstetrics and Gynaecology (1997) will be in general a good source for
culturing amniocytes for cytogenetic diagnosis (Chapter 2.1). The volume of the
admitted material, the reason for amniocentesis, gestational age at amniocentesis
as well as the chromosomal diagnosis, appeared as no major critical factors where
small variations would lead to unacceptable variation in outcome. We found that
only the variation in appearance of the amniotic fluid has a relatively major impact.
However, the effect of bloody or brown amniotic fluid was an extended culture time
rather than culture failure (Chapter 2.1). The minimal amount of amniotic fluid
required for culturing leading to a reliable cytogenetic result appeared to be 3.2 ml
(Chapter 2.2). Commercial hormone-supplemented media are generally of
sufficient quality for prenatal cytogenetic purposes. However, because of the
inclusion of in-vivo produced serum, intrinsic variations in quality may occur that
make acceptance testing necessary. To address this issue, we have developed an
acceptance test based on the principle of stress testing (Chapter 2.3). We
established the validity and sensitivity of this test by investigating the influence of
decreased oxygen tension during culturing and by comparison with standard
growth curve determination. A reduced oxygen tension during culturing caused a
Summary,future prospects, and suggested guidelines
B.Sikkemaproefschrift2.doc
123
2.5-fold increase in the number of AF cells within a clone. The number of clones,
however, is only slightly increased (Chapter 2.3).
Factors involved in culturing chorionic villi have not yet been systematically
investigated. To shorten the culture time and to increase the number of
metaphases per preparation, a systematic investigation of factors involved in this
process, similar to what we have done for amniocytes, is recommended.
Factors influencing the quality of slide preparation
In preparations from primary in-situ cultures of AF cells and chorionic villi we found
that metaphase quality is only to a limited degree affected by the composition of
fixative, the relative humidity, the ambient temperature and the air flow at
harvesting (Chapter 3). Variations in the conditions commonly used in prenatal
cytogenetic diagnosis have a limited influence and cannot be considered as critical
for the quality of slide preparation. Thus, expensive measures for fine control of
these conditions are not necessary at all. Within this limited influence it is the
composition of fixative which mostly affects chromosome spreading. The presence
or absence of cytoplasm over the metaphase is for amniotic fluid cells most
affected by the relative humidity and for chorionic villi by the air flow (Chapter 3).
We have not been able to discover factors that could lead to less variation in
quality.
We recommend in-situ harvesting rather than suspension type of harvesting,
since the turn-around time is much shorter and less chromosomal analyses have to
be made for a reliable diagnosis of in-situ cultures.
Factors influencing the quality of chromosomal analysis
All previously mentioned factors lead to the final quality of the chromosome
preparation; an appropriate number of metaphases on the slide, a sufficient length
of the chromosomes and within a short turn-around time. The required conditions
have already been well-defined in existing guidelines (Hsu et al., 1992; Hsu and
Chapter 5
B.Sikkemaproefschrift2.doc
124
Benn, 1999; ACC Working Party, 2001; Mellink et al., 2003). Following these
guidelines will result in an accurate and reliable detection or exclusion of structural
and numerical chromosomal abnormalities. However, an additional guideline for
the exclusion of chromosomal mosaicism in AF cell cultures would be desirable.
We, therefore, developed a probability table which may help laboratories to
standardise the confidence levels of prenatal diagnosis regarding mosaicism when,
despite routinely using the in-situ method, trypsinised cultures need to be used to
complete the analysis (Chapter 4.3). This probability table can of course also be
applied when using solely the flask method.
For chorionic villi, chromosomal analysis from both a direct preparation and a
culture is recommended. Provided that maternal contamination is limited, the
culture method alone is sufficiently reliable for prenatal diagnosis to be used as the
sole method (Chapter 4.2). The risk of a false negative result when using a direct
preparation has been calculated as 1-3 percent per 1000 and is negligible when
analysing preparations from cultures (Chapter 4.1). The risk of a false positive
result (positive predictive value) is in general higher in analysis from direct
preparations (Hahnemann and Vejerslev, 1997) and depends on the type of
abnormality (Chapter 4.2).
5.2 Future prospects
Amniocentesis and chorion villus biopsy will in the foreseeable future remain the
methods of choice in prenatal diagnosis. It is to be expected that the cytogenetic
diagnosis of AF cells and chorionic villi will increasingly be supplemented by
molecular techniques. For a more rapid test and a reduction of maternal anxiety,
detection of the most common aneuploidies, such as trisomies 13; 18; 21 and sex
chromosomal aneuploidies, interphase fluorescence in situ hybridisation (FISH)
has been introduced (Klinger et al.,1992; Bryndorf et al., 1997; D’Alton et al., 1997;
Pergament et al., 2000). In this technique, a set of chromosome-specific
fluorescence-labelled probes is hybridised to interphase nuclei of uncultured
prenatally obtained cells. The number of fluorescent signals in each nucleus
Summary,future prospects, and suggested guidelines
B.Sikkemaproefschrift2.doc
125
represents the copy number of the chromosome. A result is ultimately available
after two days, however, due to the significantly higher costs and labour
requirements this testing is offered only for a subset of prenatal referrals, mainly for
the indication ‘foetal structural abnormalities found upon ultrasound examination’.
More recently, quantitative molecular tests, such as quantitative fluorescent
(QF)-PCR or multiplex ligation-dependent probe amplification (MLPA), have
become available for high-throughput, reliable, fast and robust aneuploidy
detection (Mann et al., 2004; Slater et al., 2003). MLPA is a PCR-based method for
relative quantification of up to 40 DNA sequences by means of a simple single-tube
assay (Schouten et al., 2002). During the procedure, several sequence-specific
probe sets are hybridised to genomic DNA, covalently ligated into single probes
after successful hybridisation, and only than amplified using just one primer set.
Afterwards, the resulting PCR-products are analysed semi-quantitatively using a
capillary sequencer. In practice, a first study on MLPA for detection of aneuploidy
of chromosomes 13; 18; 21; X or Y in almost 600 AF samples, leading to no false
negative or false positive results, has been successfully conducted (Slater et al.,
2003).
QF-PCR involves the relative quantification of microsatellite alleles to determine
sequence copy number. Multiplex amplification using fluorescence labelled primers
is followed by size separation and allele peak measurement on a semi-automated
genetic analyser. In testing more than 7000 prenatal samples for trisomy 13; 18
and 21 no false positive nor false negative results were detected for non-mosaic
trisomies (Mann et al., 2004).
DNA-microarray analysis is another molecular technique which allows a
simultaneous quantitative analysis of genomic copy numbers, using many probes
consisting of mapped DNA fragments that are spotted on a glass slide (Pinkel et
al., 1998; Snijder et al., 2001). Microarray technology was initially conceived to
detect the expression of thousands of genes simultaneously, but quantification of
genomic copy numbers is also possible. The resolution of an array depends solely
on its contents. Presently, subtelomere-specific arrays (Veltman et al., 2002),
Chapter 5
B.Sikkemaproefschrift2.doc
126
chromosome-specific arrays (Veltman et al., 2004) and whole genome arrays, with
resolutions varying from 2Mb to tiling path density, are being used in diagnostic
practice. The comparative nature of the technique implies that only net gains and
losses can be found. Balanced rearrangements will remain undetected. In prenatal
diagnosis, routine microarray analysis of AF cells or chorionic villi is too labour-
intensive and too expensive to date. However, this technique can be used for
further investigation of small chromosomal unbalances at a higher resolution when
an apparently balanced karyotype is found in the group of referrals with
‘abnormalities at ultrasound’, or it can be used to determine whether a de-novo
structural abnormality is balanced or not.
Cytogenetic analysis of AF cells and chorionic villi is associated with
invasive techniques to obtain the material. These techniques carry a risk for the
mother and the foetus. Therefore, they need to be restricted to women whose
increased risk of having a foetus with an abnormality outweighs the risk of the
procedure. Introduction of a better risk estimator than maternal age will reduce the
number of invasive tests, especially in the group of women with advanced maternal
age (Beekhuis et al., 1994; Kornman et al., 1995). At the same time the detection
rate for Down syndrome for each procedure as well as the absolute number has
increased (Cornel et al., 1993). One approach is biochemical screening of maternal
serum possibly combined with a measurement of thickening of the nuchal fold at
ultrasound (nuchal translucency). This procedure for prenatal detection of Down
syndrome is estimated to result in a detection rate of 93% with 5.9% false positive
rate (Steenhouse et al., 2004). Only if this screening indicates an increased risk for
Down syndrome, subsequent cytogenetic diagnosis by use of AF cells or chorionic
villi should be performed. Recently, a model for contingent screening already in the
first trimester was recommended which avoids further testing in up to 75% of
unaffected pregnancies (Wright et al., 2004). In most European countries
screening for Down syndrome is offered to all pregnant women. In the Netherlands
the advice of the Health Council (2004) was in this line. However, the last
population screening act (Ministry of Health, 2004), passed after this advice,
allowed health care providers only to offer screening for Down syndrome to
Summary,future prospects, and suggested guidelines
B.Sikkemaproefschrift2.doc
127
pregnant woman older than 36 years of age. Pregnant women younger than 36
years of age have to be informed about screening, but the use is at their own
expense. Since most Down syndrome children are born to younger mothers, this
policy means a highly ineffective way of screening for Down syndrome.
Alternative non-invasive techniques for prenatal diagnosis that can be
offered to all pregnant women are actively explored. Isolation of intact foetal cells
from the maternal circulation has been extensively studied (Adinolfi, 1995; Bianchi
et al., 1999; Simpson and Elias, 1993). The major limitation in applying this
technique clinically, however, is the in general low number of foetal cells found in
the maternal circulation; approximately one foetal cell per ml of maternal blood
(Bianchi et al.; 1997, Hamada et al.; 1993). Furthermore, foetal cells can persist in
the circulation, sometimes many years after delivery (Bianchi et al.; 1996). As an
alternative, Lo et al. (1997) was the first to describe the presence of cell-free foetal
DNA in maternal plasma and serum. Cell-free foetal DNA is cleared from the
circulation in less than an hour (Lo et al., 1999). Several clinical applications for
cell-free foetal DNA have been described, such as foetal sex determination
(Sekizawa et al., 2001) and foetal Rhesus- D genotyping (Faas et al., 1998).
Maternal plasma/ serum analysis has the advantage of being rapid, reliable,
reproducible, and easily carried out for a large number of samples. The main
limitation at present appears to be the availability of uniquely foetal gene
sequences that identify and/or measure the presence of foetal DNA in both male
and female foetuses (Bianchi, 2004; Wataganara et al., 2004).
For the further future combining the isolation of foetal DNA from maternal
serum/ plasma with molecular techniques such as MPLA or whole genome
microarrays, would certainly be a desirable way to go, in order to perform prenatal
diagnosis without any risk to mother and foetus.
As an alternative to prenatal diagnosis, preimplantation genetic diagnosis
(PGD) was introduced at the beginning of the 1990s to prevent termination of
pregnancy in couples with a high risk for offspring affected by a sex-linked genetic
Chapter 5
B.Sikkemaproefschrift2.doc
128
disease (Sermon et al., 2004). It is a kind of early prenatal diagnosis, in which at
least one cell of an 8-cells in-vitro embryo is analysed for well defined genetic
defects. Only embryos free of this defect are replaced in utero. Recently, PGD for
aneuploidy screening has been applied to improve a low success rate of in-vitro
fertilisation (IVF) attributed to chromosomal aneuploidies in the embryos, as is
sometimes the case in woman older than age 36 –40 years (Wilton et al., 2002). In
the Netherlands, a blind randomised study has been started to answer this
question (Mastenbroek et al., 2004). However, PGD will be restricted to women
undergoing IVF and to a few couples with a high risk of an affected foetus for well-
defined diseases.
Reducing the costs of health care is an important issue since the budget is
not increasing at the same rate as the demands. From a medical and financial
point of view, a screening program is the most cost-effective policy (Wortelboer et
al., 2000). It leads to a better use of cytogenetic diagnosis, with more benefit and
less harm to the pregnant women and their babies. Furthermore the question rises
whether a complete cytogenetic diagnosis is essential for the indication ‘advanced
risk at screening’ (Hulten et al.,2003). A diagnosis for just the common
aneuploidies, such as 13; 18; 21; X and Y by a molecular technique, e.g. MLPA,
might be sufficient. In a retrospective study among 65,189 cases in the advanced
maternal age group 0.2% false negative or ‘clinically relevant’ abnormalities would
have been missed if women would have had the MLPA test for chromosomes 13;
18; 21, X and Y instead of karyotyping (Niewint et al., 2004). The false negative
rate could be further reduced by the use of an MLPA test with probes for the
subtelomeric regions of all chromosomes (Suijkerbuijk et al., 2004).
Although (prenatal) cytogenetics is moving into the direction of molecular
approaches, these developments by no means imply that ‘conventional’
cytogenetic diagnosis will disappear from the laboratories in the near future. Only
by regular karyotyping, an overview of the complete human genome is obtained
and balanced aberrations, such as translocations, inversions or insertions can be
visualised, whereas DNA techniques can only detect unbalanced aberrations.
Summary,future prospects, and suggested guidelines
B.Sikkemaproefschrift2.doc
129
In summary, possibilities for prenatal diagnosis will further increase in the
future. The choice of the method will remain difficult. On the one hand there is the
wish of pregnant women to get maximal certainty of a healthy baby, on the other
hand there is a limited health care budget. Moreover, with every introduction of a
new technique the question should be discussed whether the test should be
available to all pregnant women. Professionals in the field of prenatal diagnosis
should inform and advice the government, health care providers and pregnant
women on the implications of the different tests. Eventually, it is up to the society to
determine whether all questions in prenatal diagnosis should be answered, which
tests should be offered to whom and at which costs. The answer to these
questions might be different for different countries.
5.3. Suggested guidelines for handling amniotic fluid and chorionic villi
In general, our prenatal cytogenetic practice has proven to be robust and reliable.
Yet, based on our findings a few additional guidelines are formulated.
Guidelines for culturing AF cells:
It can be concluded that AF can be stored, if necessary, for at least two days either
at room temperature or in the refrigerator without any decrease of growth potential
(Chapter 2.3). The desired amount of AF is 6 ml. According to our standard
procedure for culturing, four in-situ cultures will be initiated with 1.5 ml AF per dish
(Chapter 2.2). However, 3.2 ml AF, 0.8 ml AF per in-situ dish, are sufficient for a
reliable diagnosis, while culture time is extended with two days. We suggest to
register a bloody or brown appearance of AF and AF sample volumes less than 3.2
ml for explaining a possible delay in culture time or culture failure (Chapter 2.1). A
new lot or type of culture medium needs to be subjected to stress testing before
introducing it in the routine practice (Chapter 2.3). The growth potential of AF cells,
as assessed by an increase of the number of cells within a clone, can be enhanced
2.5-fold under a decreased oxygen tension (Chapter 2.3).
Chapter 5
B.Sikkemaproefschrift2.doc
130
Guidelines for harvesting and slide preparation of in-situ cultures:
As demonstrated in Chapter 3, metaphase quality is only to a limited degree
affected by the composition of fixative, the relative humidity, the ambient
temperature and the air flow at harvesting. Within this limited influence, the
following recommendations can be given. (1) The higher the concentration of
acetic acid before slide preparation the better the spreading of chromosomes. (2)
The presence of cytoplasm in in-situ preparations is slightly influenced by relative
humidity and is minimal between 30% and 40%. (3) Extra air flow of 0.6 m/s further
decreases the presence of cytoplasm for chorionic villi preparations. Expensive
instruments in climate control during harvesting and slide preparation are not
necessary.
Guidelines for chromosomal analysis:
Guidelines for analysing preparations from AF cells and chorionic villi have already
been well defined. Therefore, only two additional suggestions are given. (1) The
table for exclusion of mosaicism, especially developed for preparations from
trypsinised in-situ AF cell cultures, should be used to complete analysis from in-situ
culture dishes in case an insufficient number of clones is obtained in the regularly
harvested dishes. (2) Metaphase preparations from cultured mesenchymal cells
from chorionic villi alone can be analysed for cytogenetic diagnosis provided that
maternal cell contamination is limited.
Summary,future prospects, and suggested guidelines
B.Sikkemaproefschrift2.doc
131
5.4 Literature
Adinolfi M. Non- or minimally invasive prenatal diagnostic tests on maternal blood samples or
transcervical cells. Prenat Diagn. 1995; 15: 889-96.
Association of Clinical Cytogeneticists (ACC) Working Party. Professional guidelines for clinical
cytogenetics. BMJ Publishing Group 2001. http://www.cytogenetics.org.uk/information/Guidelines.pdf
Beekhuis JR, de Wolf BTMH, Mantingh A, Heringa MP. The influence of serumscreening on the
amniocentesis rate in women of advanced maternal age. Prenat Diagn. 1994; 14: 199-202.
Bianchi DW, Zickwolf GK, Weil GJ, Sylvester S, De Maria MA. Male foetal progenitor cells persist in
maternal blood for long as 27 years postpartum. Proc Natl Acad Sci USA1996; 93: 705-8.
Bianchi DW, Williams JM, Sullivan LM, Hanson FW, Klinger KW, Shuber AP. PCR quantitation of foetal
cells in maternal blood in normal and aneuploid pregnancies. Am J Hum Genet. 1997; 61: 822- 9.
Bianchi DW. Foetal cells in the maternal circulation: feasibility for prenatal diagnosis. Br J Haematol.
1999 ; 105: 574-83.
Bianchi DW. Circulating foetal DNA: its origin and diagnostic potentials; A review. Placenta 2004; 25:
S93-S101.
Bryndorf T, Christensen B, Vad M, Parner J, Brocks V, Philip J. Prenatal detection of chromosome
aneuploidies by fluorescence in situ hybridisation: experience with 2000 uncultured amniotic fluid
samples in a prospective preclinical trial. Prenat Diagn. 1997; 17(4): 333-41.
Cornel MC, Breed ASPM, Beekhuis JR, Meerman te GJ, Kate ten, LP. Down syndrome: effects of
demographic factors and prenatal diagnosis on the future livebirth prevalence. Hum Genet. 1993; 92:
163-68.
D’Alton ME, Malone FD, Chelmow D, Ward BE, Bianchi DW. Defining the role of fluorescence in situ
hybridisation on uncultured amniocytes for prenatal diagnosis of aneuploidies. Am J Obstet Gynecol.
1997; 176: 769-76.
Dutch Society of Obstetrics and Gynaecology. 1997. Guidelines for invasive prenatal diagnosis.
http://www.nvog.nl
Faas BH, Beuling EA, Christiaens GC, dem Borne AE, van der Schoot CE. Detection of foetal RHD-
specific sequences in maternal plasma. Lancet 1998; 352 : 1196.
Hahnemann JM, Vejerslev LO. European collaborative research on mosaicism in CVS (EUCROMIC) –
foetal and extrafoetal cell lineages in 192 gestations with CVS mosaicism involving single autosomal
trisomy. Am J Med Genet. 1997; 70: 170-87.
Hamada H, Arinami T, Kubo T, Hamaguchi H, Iwasaki H. Foetal nucleated cells in maternal peripheral
blood: frequency and relationship to gestational age. Hum Genet. 1993; 91: 427-32.
Chapter 5
B.Sikkemaproefschrift2.doc
132
Health Council of the Netherlands. Prenatal screening (2); Down’s syndrome, neural tube defects. The
Hague: Health Council of the Netherlands, 2004; publication no. 2004/06.
Hulten MA, Dhanjal S, Pertel B. Rapid and simple prenatal diagnosis of common chromosome
disorders: advantages and disadvantages of the molecular methods FISH and QF-PCR. Reproduction
2003; 126(3): 279-97.
Hsu LY, Kaffe S, Jenkins EC, Alonso L, Benn PA, David K, Hirschhorn K, Lieber E, Shanske A, Shapiro
LR, et al. Proposed guidelines for diagnosis of chromosome mosaicism in amniocytes based on dataderived from chromosome mosaicism and pseudomosaicism studies. Prenat Diagn. 1992;12(7):555-73
Hsu LYF, Benn PA. Revised guidelines for the diagnosis of mosaicism in amniocytes. Prenat Diagn.
1999; 19: 1081-82.
Kornman LH, Beekhuis JR, Mantingh A, Morssink LP. Serumscreening for foetal Down syndrome in
pregnant women at advanced maternal age: less amniocentesis. Ned Tijdschr Geneeskd. 1995; 139:
1840-4.
Klinger K, Landes G, Shook D, Harvey R, Lopez L, Locke P, Lerner T, Osathanondh R, Leverone B,
Houseal T, et al. Rapid detection of chromosome aneuploidies in uncultured amniocytes by using
fluorescence in situ hybridization (FISH). Am J Hum Genet. 1992; 51(1):55-65.
Lo YM, Corbetta N, Chamberlain PF, Rai V, Sargent IL, Redman CW, Wainscoat JS. Presence of foetal
DNA in maternal plasma and serum. Lancet. 1997; 350(9076): 485-7.
Lo YM, Zhang J, Leung TN Lau TK, Chang AM, Hjelm NM. Rapid clearance of foetal DNA from
maternal plasma. Am J Hum Genet. 1999; 64: 218-24.
Mann K, Donaghue C, Fox SP, Docherty Z, Ogilvie CM. Strategies for the rapid prenatal diagnosis of
chromosome aneuploidy. Eur J Hum Genet. 2004; 12: 907-15.
Mastenbroek S, Engel C, van Echten-Arends J, Sikkema-Raddatz B, van Wassenaer, AG, de Vries WA,
Heineman MJ, Repping S, van der Veen F. Preimplantation genetic screening for numerical
chromosomal abnormalities in embryos from of women 35 years of age and older: first results in the
Netherlands. Ned Tijdschr Geneeskd. 2004; 148(50): 2486-90.
Mellink CH, de Pater JM, Poddighe PJ, Smeets DFCM. Quality of cytogenetic diagnosis: conditions,
guidelines and control.2003.
Ministry of Health. Population screening act. www.minvws.nl. 2004. IBE/E/ 2488853.
Nieuwint AWM, de Pater JM, Madan K. To what extent can the MLPA technique replace standard
chromosome analysis? 12th International Conference on Prenatal Diagnosis and Therapy. 2004;
Abstract 77.
Pergament E, Chen PX, Thangavelu M, Fiddler M. The clinical application of interphase FISH in
prenatal diagnosis. Prenat Diagn. 2000; 20: 215-220.
Summary,future prospects, and suggested guidelines
B.Sikkemaproefschrift2.doc
133
Pinkel D, Segraves R, Sudar D, Clark S, Poole I, Kowbel D, Collins C, Kuo WL, Chen C, Zhai Y,
Dairkee SH, Ljung BM, Gray JW, Albertson DG. High resolution analysis of DNA copy number variation
using comparative genomic hybridisation to microarrays. Nat Genet. 1998; 20: 207-11.
Rijnders RJ, van der Schoot CE, Bossers B, de Vroede MA, Christiaens GC. Foetal sex determination
from maternal plasma in pregnancies at risk for congenital adrenal hyperplasia. Obstet Gynecol. 2001;
98: 374- 8.
Schouten JP, McElgunn CJ, Waaijer R, Zwijnenburg D, Diepvens F, Pals G. Relative quantification of40 nucleic acid sequences by multiplex ligation-dependent probe amplification. Nucleic Acids Res.
2002; 30(12):e57.
Sekizawa A, Kondo T, Iwasaki M, Watanabe A, Jimbo M, Saito H, Okai T. Accuracy of foetal gender
determination by analysis of DNA in maternal plasma. Clin Chem. 2001; 47(10):1856-8.
Sermon K., van Steirteghem A., Liebaers I. Preimplantation genetic diagnosis. Lancet 2004; 363(15):
1633-41.
Simpson JL, Elias S. Isolating foetal cells from maternal blood. Advances in prenatal diagnosis through
molecular technology. JAMA 1993; 270: 2357-61.
Slater HR, Bruno DL, Ren H, Pertile M, Schouten JP, Choo KHA. Rapid, high throughput prenatal
detection of aneuploidy using a novel quantitative method (MLPA). J Med Genet. 2003; 40: 907-912.
Snijders AM, Nowak N, Segraves R, Blackwood S, Brown N, Conroy J, Hamilton G, Hindle AK, Huey B,
Kimura K, Law S, Myambo K, Palmer J, Ylstra B, Yue JP, Gray JW, Jain AN, Pinkel D, Albertson DG.
Assembly of microarrays for genome-wide measurement of DNA copy number. Nat Genet. 2001; 29:
263-4.
Steenhouse EJ, Crossley JA, Aitken DA, Brogan K, Cameron AD, Connor JM. First trimester combined
ultrasound and biochemical screening for Down syndrome in routine clinical practice. Prenat Diagn.
2004; 24 (10): 774-80.
Suijkerbuijk RF, Sikkema-Raddatz B, Dijkhuizen T, Dijkhuis J, van der Veen AY, Bouman K, Gerssen-
Schoorl KBJ. Evaluation of multiplex ligation-dependent probe amplification as a means to rapidly detect
unbalanced abnormalities of all chromosomes in prenatal cytogenetic analysis. 12th International
Conference on Prenatal Diagnosis and Therapy. 2004; Abstract 88.
Veltman JA, Schoenmakers EFPM, Eussen BH, Janssen I, Merkx G, van Cleef B, van Ravenswaaij
CM, Brunner HG, Smeets D, van Kessel AG. High-throughput analysis of subtelomeric chromosome
rearrangements by use of array-based comparative genomic hybridization. Am J Hum Genet.
2002;70(5):1269-76.
Chapter 5
B.Sikkemaproefschrift2.doc
134
Veltman JA, Yntema HG, Lugtenberg D, Arts H, Briault S, Huys EHLPG, Osoegawa K, de Jong P,
Brunner HG, Geurts van Kessel A, van Bokhoven H, Schoenmakers EF. High resolution profiling of X
chromosomal aberrations by array comparative genomic hybridisation. J Med Genet. 2004; 41(6):425-
32.
Wataganara T, Bianchi DW. Foetal cell-free nucleic acids in the maternal circulation: new clinical
applications. Ann N Y Acad Sci. 2004. 1022: 90-9.
Wilton L. Preimplantation genetic diagnosis for aneuploidy screening in early human embryos: a review.Prenat Diagn. 2002; 22: 312-8.
Wortelboer MJM, de Wolf BTHM, Verschuuren-Bemelmans CC, Reefhuis J, Mantingh A, Beekhuis JR,
MC Cornel. Trends in live birth prevalence of Down syndrome in the Northern Netherlands 1987-96: the
impact of screening and prenatal diagnosis. Prenat Diagn. 2000; 20: 709-13.
Wright DE, Bradbury I, Benn P, Cuckle H, Ritchie K. Contingent screening for Down syndrome is an
efficient alternative to non-disclosure sequential screening. Prenat Diagn. 2004; 24 (10): 762-6.
Samenvatting
B.Sikkemaproefschrift2.doc
135
Samenvatting
B.Sikkemaproefschrift2.doc
136
Ongeveer 1 op de 200 baby's wordt geboren met een chromosoomafwijking. De
gevolgen daarvan kunnen sterk uiteenlopen, van verwaarloosbaar (het kind zal er
in zijn of haar ontwikkeling geen last van ondervinden) tot aan 'niet levensvatbaar'.
Veel chromosomale afwijkingen blijken gepaard te gaan met hartafwijkingen,
ontwikkelingsachterstand en uiterlijke bijzonderheden.
Chromosomale afwijkingen kunnen in twee categorieën worden ingedeeld:
numerieke en structurele. Bij een numerieke afwijking is een chromosoom teveel of
te weinig aanwezig. Bij een structurele afwijking mist er materiaal van een
chromosoom of heeft een chromosoom juist extra materiaal. Ook kan het zijn dat
bepaalde stukken van chromosomen zijn uitgewisseld. Alles is wel aanwezig maar
zit niet op de juiste plaats. De ernst van de aandoening wordt over het algemeen
bepaald door de mate waarin chromosomaal materiaal is toe- dan wel afgenomen.
Door middel van prenatale cytogenetische diagnostiek is het mogelijk om
het overgrote deel van de chromosoomafwijkingen uit te sluiten en zwangeren op
dat punt gerust te stellen. Indien bij dergelijk onderzoek wel een
chromosoomafwijking wordt aangetoond, kunnen de toekomstige ouders na
grondige voorlichting over wat hen mogelijk te wachten staat, eventueel kiezen
voor afbreking van de zwangerschap. Ook al wordt er d.m.v. prenataal
chromosomenonderzoek geen afwijking gevonden, dan blijft er nog een algemeen
risico bestaan van ca. 2% - 3% op een kind met een erfelijke ziekte of aangeboren
afwijking berustend op een gen-afwijking, die niet aan de chromosomen te zien is.
De meest gebruikte methodes om uit foetaal materiaal cytogenetisch
onderzoek te verrichten zijn de vruchtwaterpunctie en de vlokkentest (Figuur 1).
Samenvatting
B.Sikkemaproefschrift2.doc
137
Figuur 1: Schematische weergave van een vroege zwangerschap tijdens de
afname van chorionvlokken (uit Genetics: Human Aspects; Mange & Mange,
Second Edition, Sinauer Associates, INC. Publishers Sunderland, pg. 504)
Bij de vruchtwaterpunctie wordt normaliter tussen de 16e en 18e week van de
zwangerschap 15- 20 ml vruchtwater afgenomen. Foetale cellen die in het
vruchtwater zijn terecht gekomen worden onder bepaalde omstandigheden in
kweek gebracht. Door deling groeien de vitale cellen uit tot een samenhangend
groepje cellen, een zogenaamde celkloon. Als er redelijk wat klonen van
voldoende grootte zijn, dit is na circa acht dagen, wordt een mitose-remmer aan de
kweken toegevoegd, zodat de cellen in het delingsstadium blijven steken. Om tot
een chromosoompreparaat te komen worden de cellen met een hypotone vloeistof
behandeld, bedoeld om de cellen te laten opzwellen en de chromosomen beter te
spreiden. Vervolgens worden de cellen met een mengsel van methanol en
azijnzuur gefixeerd. Na het verwijderen van het fixatief worden de cellen gedroogd.
Hierna worden de chromosomen bewerkt en gekleurd om een bepaald
bandenpatroon te verkrijgen. Aan de hand van dit bandenpatroon kunnen de
B.Sikkemaproefschrift2.doc
138
chromosomen in paren worden gerangschikt (karyotype)1, zodat eventuele
numerieke of structurele afwijkingen kunnen worden gedetecteerd. Het gehele
proces van afname van het vruchtwater tot diagnose duurt twee tot drie weken.
Bij een vlokkentest worden 5- 50 mg chorionvlokken (embryonaal weefsel
dat deel uitmaakt van de placenta) afgenomen, normaliter tussen de 10e en de
12e week van de zwangerschap. Na afname worden de chorionvlokken eerst van
aanhangend moederlijk weefsel ontdaan. De vlokken bestaan uit verschillende
celtypes. Van één daarvan, de cytotrofoblast cellen, kunnen via de zogenaamde
directe afwerking zonder kweek chromosoompreparaten gemaakt worden,
aangezien meestal voldoende cellen in deling zijn. Een cytogenetische diagnose is
dan reeds na enkele uren mogelijk, echter de kwaliteit van deze
chromosoompreparaten is slecht. Zodoende kunnen kleine structurele afwijkingen
over het hoofd worden gezien.
Als alternatief kan van een ander celtype, de mesenchymale cellen, een kweek
worden ingezet. Meestal worden hierbij eerst de cytotrofoblast cellen verwijderd en
de mesenchymale cellen met behulp van een enzym tot een suspensie (cellen
bevattende vloeistof) verteerd. Deze celsuspensie wordt in kweek gebracht. Het
proces om tot een chromosoompreparaat te komen is dan hetzelfde als bij
vruchtwatercellen. Een cytogenetische diagnose is na acht tot veertien dagen
mogelijk en de kwaliteit is dusdanig dat ook kleine structurele afwijkingen kunnen
worden gedetecteerd. Een nadeel van deze methode is de mogelijke bijmenging
van moederlijke cellen, waardoor in plaats van foetale cellen moederlijke cellen
1 Een mens heeft normaliter in elke cel 46 chromosomen; 22 paren van autosomen en één paar
geslachtschromosomen. Bij een vrouw zijn beide geslachtschromosomen X- chromosomen, bij de man
bestaan ze uit één X- en één Y-chromosoom. Chromosomen zijn tijdens een celdeling (mitose) het
meest gecondenseerd in de metafase. Zij zijn dan lineair van gestalte en kunnen als paar naar lengte,
positie van het centromeer en bandenpatroon gerangschikt worden. Deze ordening wordt karyotype
genoemd.
Samenvatting
B.Sikkemaproefschrift2.doc
139
zouden kunnen worden geanalyseerd. Vanwege de kans op discrepanties tussen
het chromosomenpatroon van de foetus en weefsel van de placenta wordt een
diagnose gebaseerd op de resultaten van cytotrofoblast en mesenchymale cellen
dan ook het betrouwbaarst geacht.
De belangrijkste kwaliteitscriteria ten behoeve van de diagnostiek voor het
genoemde foetale materiaal zijn de sensitiviteit, of wel de mate waarin het lukt een
in de foetale cellen aanwezige afwijking ook inderdaad aan te tonen; en de
specificiteit, die aangeeft in hoeverre kan worden vastgesteld dat een afwijking ook
daadwerkelijk voorkomt in de foetale cellen en niet stamt uit moederlijke cellen of
berust op een pas in de kweek ontstane afwijking. Het doel van prenatale
cytogenetische diagnostiek is immers om een mogelijke chromosomale afwijking
bij de foetus vast te stellen. Er is sprake van een optimale diagnostiek wanneer
sensitiviteit en specificiteit zo goed als 100% zijn. Bovendien dient een diagnose zo
snel mogelijk tegen zo laag mogelijke kosten tot stand te komen. Om dit te kunnen
realiseren is kennis nodig van alle factoren die hierbij een rol spelen en van de
mate van hun invloed. Gebaseerd op deze kennis kunnen richtlijnen worden
opgesteld met betrekking tot de wijze waarop vruchtwatercellen en chorionvlokken
bewerkt moeten worden om uiteindelijk tot betrouwbare prenatale diagnostiek te
komen.
Het cytogenetisch proces kan in de volgende de stappen gesplitst worden;
afname van materiaal, celkweek, preparaatbereiding en analyse. Voor de analyse
van chorionvlokken en vruchtwatercellen bestaan reeds richtlijnen, opgesteld door
de Werkgroep van de Nederlandse Vereniging voor Klinische Cytogenetica. Deze
richtlijnen zijn van voldoende kwaliteit om een betrouwbare diagnose te
waarborgen. Voor de andere stappen in dit proces zijn echter vrijwel geen
richtlijnen voorhanden. Deze aan de analyse voorafgaande stappen zijn wel van
groot belang voor de uiteindelijke kwaliteit van het chromosoompreparaat inzake
de hoeveelheid klonen en delende cellen en de lengte en structuur van de
chromosomen. Bovendien is de invloed van factoren die hierbij een rol zouden
B.Sikkemaproefschrift2.doc
140
kunnen spelen nog nooit systematisch onderzocht. Dergelijk onderzoek is wel van
belang, met name voor prenatale diagnostiek, omdat het verkrijgen van
vruchtwater of chorionvlokken gepaard gaat met een zeker risico voor zowel de
aanstaande moeder alsook voor de foetus en dus een gewaarborgde kwaliteit
gewenst is. Mislukkingen die herhaling van de procedure noodzakelijk zouden
maken, dienen daarom zo veel mogelijk te worden vermeden dan wel beperkt.
Het doel van dit proefschrift is om te trachten richtlijnen op te stellen voorhet hele proces dat vruchtwatercellen en chorionvlokken ondergaan om toteen juiste cytogenetische diagnose te komen.
In ons onderzoek hebben wij de invloed van mogelijk belangrijke factoren binnen
dit proces onderzocht om, al naar gelang de mate van hun invloed, te kunnen
bepalen welke factoren kritisch bewaakt dienen te worden en voor welke er
richtlijnen moeten worden opgesteld.
In een algemene introductie in hoofdstuk 1 wordt het proces beschreven
dat vruchtwatercellen en chorionvlokken ondergaan om tot een cytogenetische
diagnose te komen. Tevens worden bestaande richtlijnen voor dit proces
beschreven en wordt aangegeven wat reeds bekend is over de invloed van
verschillende factoren op de kwaliteit van het cytogenetisch proces.
Hoofdstuk 2 richt zich op het kweken van vruchtwatercellen volgens de in-
situ procedure (in-situ = ter plaatse 2). Hierbij werd middels retrospectief
onderzoek bekeken in welke mate de gynaecoloog, de hoeveelheid vruchtwater,
de samenstelling van het vruchtwater (met name al dan niet ook bloed bevattend),
de grootte van het neerslag van cellen na centrifugatie, de zwangerschapsduur bij
afname, de reden voor diagnostiek, de chromosomale diagnose en het type of de
2 In dit geval wil dit zeggen, dat vruchtwater in een kweekschaaltje wordt gebracht waarin op de bodemeen glaasje is geplaatst en dat de cellen van het vruchtwater uitzakken naar de bodem en zich terplaatse aan het glasoppervlak hechten. Ze kunnen daar dan eventueel tot klonen uitgroeien.
Samenvatting
B.Sikkemaproefschrift2.doc
141
geleverde partij van het kweekmedium van invloed zijn op de uiteindelijke kwaliteit
van de celkweken. Kwaliteit wordt in dit verband gemeten aan de totale
hoeveelheid klonen welke aanwezig zijn op de eerste dag van de beoordeling van
de celkweken en aan de tijdsduur van het in kweek brengen van de
vruchtwatercellen tot aan de preparaatbereiding (hoofdstuk 2.1). Uit ons onderzoek
is gebleken dat vruchtwater afgenomen volgens de richtlijnen van de Nederlandse
Vereniging voor Gynaecologie en Obstetrie over het algemeen van voldoende
kwaliteit is voor cytogenetisch onderzoek. Verder kwam naar voren dat de
hoeveelheid vruchtwater, de reden voor afname, de zwangerschapsduur bij
afname, alsmede de chromosomale diagnose niet van invloed zijn op de kwaliteit
van vruchtwater celkweken. Kleine veranderingen binnen deze factoren leiden niet
tot een onacceptabel resultaat. Daarentegen werd duidelijk, dat bloederig of bruin
vruchtwater een grote invloed heeft op de kwaliteit van de kweken. Hoewel de
kweken hierdoor weliswaar niet mislukken, wordt de kweekduur er met maar liefst
drie dagen door verlengd.
Verder hebben wij de minimale hoeveelheid vruchtwater bepaald die voor
een betrouwbare diagnose nodig is (hoofdstuk 2.2). Dit bleek 3,2 ml te zijn.
Commercieel medium voor vruchtwatercellen dat is aangevuld met hormonen, is
over het algemeen van voldoende kwaliteit voor het kweken. Omdat medium
tevens in-vivo geproduceerd serum bevat, kan de kwaliteit tussen verschillende
geleverde partijen uiteenlopen. Om deze reden hebben wij een acceptatietest voor
medium ontwikkeld, gebaseerd op het principe van een stresstest (hoofdstuk 2.3).
Om de validiteit en sensiviteit van deze test te kunnen bepalen, hebben wij als
voorbeeld de invloed van een verlaagde zuurstofspanning tijdens het kweken
onderzocht. De groei van vruchtwatercellen is m.b.v. de stresstest en groeicurven3
vergeleken onder gereduceerde zuurstofspanningen van 10%, 5% en 2,5% en bij
de 20% zuurstof in de normale atmosfeer. Het bleek dat een gereduceerde
3 Vanuit samengevoegde vruchtwatermonsters zijn verschillende kweken opgezet. De groei van dezekweken wordt op verschillende dagen gestopt om de hoeveelheid cellen te tellen en zo een groeicurveover die dagen op te stellen.
B.Sikkemaproefschrift2.doc
142
zuurstofspanning resulteerde in een 2,5 voudige toename in het aantal cellen per
kloon. Het aantal klonen nam slechts licht toe.
Hoofdstuk 3 richt zich op de preparaatbereiding. Tijdens de
preparaatbereiding van in-situ kweken4 van vruchtwatercellen en chorionvlokken is
de invloed bepaald van de samenstelling van het fixatief, de relatieve vochtigheid,
de temperatuur en de luchtstroom op de kwaliteit van chromosomen. Kwaliteit
wordt in dit verband gemeten als de mate van spreiding van de chromosomen en
de mate van het (niet) aanwezig zijn van cytoplasma. Hierbij heeft de
samenstelling van het fixatief de grootste invloed op de spreiding van de
chromosomen. Het beperken van de aanwezigheid van cytoplasma wordt bij
vruchtwatercellen gerealiseerd door het beïnvloeden van de luchtvochtigheid en bij
chorionvlokken door het beïnvloeden van de luchtstroom. We hebben echter
vastgesteld, dat de genoemde factoren maar in beperkte mate van invloed zijn op
de kwaliteit van de chromosomen. Variatie in deze condities leidt nauwelijks tot
kwaliteitsveranderingen. Desalniettemin verdient de in-situ preparaatbereiding
onze voorkeur boven de preparaatbereiding van celsuspensies5, omdat de in-situ
procedure veel sneller is en daarbij minder delende cellen nodig zijn voor een
betrouwbare diagnose.
Hoofdstuk 4 richt zich op de analyse van preparaten van vruchtwatercellen
en chorionvlokken. Voor de analyse bestaan reeds richtlijnen met betrekking tot de
hoeveelheid cellen die geanalyseerd moeten worden per indicatie en per gevonden
afwijking en inzake de benodigde bandenresolutie van de chromosomen. In
aanvulling hierop hebben wij in een tabel gedefinieerd hoeveel cellen geanalyseerd
moeten worden voor het kunnen uitsluiten van een chromosomaal
4 Dat wil zeggen dat de in-situ celkweken tijdens de preparaatbereiding op het glaasje blijven en na hetdrogen op een objectglaasje geplakt worden.
5 Dat wil zeggen dat de cellen meestal in een plastic flesje worden gebracht en op de bodem groeien.Wanneer bijna het gehele oppervlak met cellen is bedekt, worden de cellen met behulp van een enzymvan de bodem losgehaald (getrypsiniseerd). De zo onstaande celsuspensie wordt tot chromosoompreparaat verwerkt.
Samenvatting
B.Sikkemaproefschrift2.doc
143
mozaïekpatroon6 in getrypsiniseerde in-situ preparaten van vruchtwatercelkweken.
Deze tabel draagt bij aan de verdere standaardisering van de
betrouwbaarheidsgrenzen om een mozaïekpatroon uit te sluiten wanneer, de in-
situ kweek ten spijt, getrypsiniseerde kweken nodig zijn om de diagnose te
completeren (hoofdstuk 4.3). Deze tabel kan ook gebruikt worden wanneer in
flessen gekweekt wordt en vanuit een suspensie het preparaat bereid wordt van
waaruit cellen geanalyseerd worden.
In de literatuur wordt voor chorionvlokken de chromosomale analyse van
cytotrofoblast cellen (directe afwerking) en mesenchymale cellen (kweek)
aangeraden. Wij hebben aangetoond dat analyse uit kweekpreparaten alleen
voldoende betrouwbaar is mits bijmenging van moederlijke cellen beperkt is
(hoofdstuk 4.2). De kans op een fout-negatieve diagnose (foetus is chromosomal
afwijkend maar de geanalyseerde cellen zijn normaal) is bepaald op 1-3 procent
per 1000 wanneer slechts cytotrofoblast cellen geanalyseerd worden en is te
verwaarlozen wanneer mesenchymale cellen worden gebruikt (hoofdstuk 4.1). De
kans op een fout- positieve diagnose (foetus is chromosomal normaal maar de
geanalyseerde cellen zijn afwijkend) is over het algemeen hoger bij analyse van
cytotrofoblast cellen dan bij analyse van mesenchymale cellen en hangt af van de
gevonden afwijking (hoofdstuk 4.2).
Samengevat kan geconcludeerd worden dat onze cytogenetische
diagnostiek in vruchtwatercellen en chorionvlokken robuust en betrouwbaar is. De
volgende aanvullende conclusies en richtlijnen worden door ons in hoofdstuk 5
gegeven.
Vruchtwater kan zonder verlies van groeipotentiaal van de cellen twee
dagen in de koelkast of bij kamertemperatuur bewaard worden. De gewenste
hoeveelheid vruchtwater per patiënt is 6 ml. Volgens onze standaard procedure
voor in-situ kweken zouden vier kweken met 1,5 ml vruchtwater opgezet moeten
6 Mozaïekpatroon wil zeggen dat naast elkaar verschillende cellen met verschillende chromosomalesamenstelling voorkomen.
B.Sikkemaproefschrift2.doc
144
worden. Met 3,2 ml vruchtwater (0,8 ml per kweekschaaltje) is echter reeds (met
een twee dagen langere kweekduur) een betrouwbare diagnose te stellen. Het in
kweek brengen van minder dan 3,2 ml vruchtwater zou geregistreerd moeten
worden, evenals een bloederige of bruine samenstelling van het vruchtwater. Dit
om een mogelijke groeivertraging of mislukking te kunnen verklaren. Een nieuw
geleverde partij of type medium dient eerst op groeipotentiaal getest te worden
d.m.v. een stresstest voordat het in de diagnostiek wordt gebruikt. De
groeisnelheid van vruchtwatercellen binnen een kloon kan 2,5 keer vergroot
worden door een verlaagde zuurstofspanning tijdens het kweken.
Tijdens de preparaatbereiding is de chromosoomkwaliteit van in-situ
kweken van vruchtwatercellen en chorionvlokken slechts beperkt afhankelijk van
de samenstelling van het fixatief, de relatieve vochtigheid, de temperatuur en de
luchtstroom. Binnen deze beperkte invloed kan het volgende worden vastgesteld:
(1) hoe hoger de concentratie van azijnzuur in het fixatief voor het drogen van de
preparaten, hoe beter de spreiding van de chromosomen, (2) het minste
cytoplasma is zichtbaar wanneer de relatieve vochtigheid tijdens het drogen tussen
30% en 40% is, (3) een extra luchtstroom van 0,6 m/s kan de aanwezigheid van
cytoplasma in chorionvlokken preparaten verder beperken.
Analyse uit mesenchymale cellen van chorionvlokken (kweek) is voldoende
voor een betrouwbare diagnostiek mits bijmenging van moederlijke cellen beperkt
is. Het aantal te analyseren cellen uit getrypsiniseerde vruchtwater celkweken zou
middels de hiervoor ontwikkelde tabel bepaald moeten worden om zo nodig de
diagnostiek uit in-situ preparaten aan te vullen.
B.Sikkemaproefschrift2.doc
145
Curriculum Vitae
Birgit Sikkema-Raddatz was born the 28th of February 1965 in Güstrow (Germany).
In 1989 she graduated in Biology (specialisation Genetics) at the Martin Luther
University in Halle.
In 1990 she married Jan Sikkema, a Dutchman, which made her decide to move to
the Netherlands. That same year she started her career in the Department of
Medical Genetics (chair: Prof.dr. C.H.C.M. Buys) at the University of Groningen,
first as a technician and since 1991 as a cytogeneticist. In 1992 she became
manager of the laboratory for prenatal cytogenetics. In 2000 she switched to the
molecular cytogenetic laboratory where she now works as head of this group.
Birgit Sikkema-Raddatz is registrated as clinical cytogeneticist with the Dutch
Association of Clinical Cytogeneticists.
B.Sikkemaproefschrift2.doc
146
B.Sikkemaproefschrift2.doc
147
List of publications
Sijmons RH, Sikkema-Raddatz B, Kloosterman MD, Briet JW, de Jong, Leschot NJ.
46,XY,dup(10q) in direct CVS preparation and mosaic 48,XXXY,dup(10q) in CVS long-term
culture and fetal tissue. Prenat Diagn.1995; 15 (3): 285-90.
Sikkema-Raddatz B, Sijmons RH, Tan-Sindunata MB, van der Veen AY, Brunsting R, de
Vries B, Beekhuis JR, Bekedam DJ, van Aken, de Jong B. Prenatal diagnosis in two cases
of de novo complex balanced chromosomal rearrangements. Three-year follow-up in one
case. Prenat Diagn. 1995; 15 (5): 467-73.
Morssink LP, Sikkema-Raddatz B, Beekhuis JR, de Wolf BT, Mantingh A. Placental
mosaicism is associated with unexplained second-trimester elevation of MShCG levels, but
not with elevation of MSAFP levels. Prenat Diagn. 1996; 16 (9): 845-51.
Morssink LP, de Wolf BT , Kornman LH, Beekhuis JR, van der Hall TP, Sikkema-Raddatz B,
Mantingh A. The significance of placental pathology in pregnancies with unexplained
abnormal concentrations of MSAFP or MShCG. Early Hum Dev. 1996; 30; 47 Suppl: S63-5.
Sikkema-Raddatz B, Castedo S, te Meerman GJ. Probability tables for exclusion of
mosaicism in prenatal diagnosis. Prenat Diagn. 1997; 17 (2): 115-8.
Sikkema-Raddatz B, Verschuuren-Bemelmans CC, Kloosterman M, de Jong B. A
46,XX,der(13;14)(q10;q10),+21 child born after a 45,XX,der(13;14)(q10;q10) chromosomal
finding in CVS. Prenat Diagn. 1997; 17 (11): 1086-8.
B.Sikkemaproefschrift2.doc
148
Sikkema-Raddatz B and Sijmons RH. De novo balanced complex chromosomal
rearrangements in prenatal diagnosis. From: Chorionic villus sampling; Eds. Rao KA and
Nicolaides K; Jaypee Brothers, Medical Publishers (P) LTD, New Delhi1998.
Leegte B, Sikkema-Raddatz B, Hordijk R, Bouman K, van Essen T, Castedo S, de Jong.
Three cases of mosaicism for balanced reciprocal translocations. Am J Med Genet 1998;
79(5): 362-5.
Sikkema-Raddatz B, Bouman K, Verschuuren-Bemelmans CC, de Jong B. Trisomy 12
mosaicism in CVS culture confirmed in the fetus. Prenat Diagn. 1999; 19 (12): 1176-7.
Sikkema-Raddatz B, Bouman K, Verschuuren-Bemelmans CC, Stoepker M, Mantingh A,
Beekhuis JR, de Jong B. Four years' experience with the culture of chorionic villi. Prenat
Diagn. 2000; 20 (12):950-5.
Leegte B, Sikkema-Raddatz B, Hordijk R, Davelaar I, van der Veen, Cobben JM. Two
unbalanced segregation products due to a maternal t(7;16)inv(16). Prenat Diagn. 2001;
21(7): 550-2.
Sikkema-Raddatz B, van Echten J, van der Vlag J, Buys CHCM, te Meerman GJ. Minimal
volume of amniotic fluid for reliable prenatal cytogenetic diagnosis. Prenat Diagn. 2002; 22
(2): 164-5.
Westra JL, Boven LG, van der Vlies P, Faber H, Sikkema B, Schaapveld M, Dijkhuizen T,
Hollema H, Buys CHCM, Plukker JTM, Kok K, Hofstra RMW. A substantial proportion of
microsatellite-instable colon tumours carries TP53 mutations while not showing
chromosomal instability. Genes Chromosomes and Cancer 2004; accepted for publication
Mastenbroek S, Engel C, van Echten-Arends J, Sikkema-Raddatz B, van Wassenaer, AG,
de Vries WA, Heineman MJ, Repping S, van der Veen F. Preimplantatie Genetische
screening naar numerieke chromosoomafwijkingen bij embryo’s van vrouwen van 35 jaar en
ouder: de eerste resultaten in Nederland. Ned Tijdschr Geneeskd. 2004; 148(50): 2486-90.
proefdruk1.doc
149
Dankwoord
Allereerst wil ik mijn copromotor Dr. Gerard te Meerman bedanken daar hij aan de
wieg stond van deze promotie. Gerard, jouw enthousiasme voor dit onderzoek is
van begin tot eind onverminderd groot geweest en was voor mij een belangrijke
bron voor inspiratie. Hartelijk dank voor al jouw berekeningen, alle uitvoerbare en
niet uitvoerbare ideeën en jouw support op het praktisch vlak.
Veel dank ben ik ook verschuldigd aan mijn promotor Prof. Charles Buys.
Charles, bedankt voor jouw kritische blik ten aanzien van de resultaten van onze
experimenten waardoor vaak nieuwe invalshoeken konden ontstaan. Daarnaast
ben ik je zeer erkentelijk voor het lezen en redigeren van mijn manuscripten
waardoor ze aan helderheid toenamen.
Mijn direct leidinggevende Klasien Gerssen-Schoorl wil ik bedanken voor het
feit dat zij mij in mijn promotieplannen heeft gesteund waardoor ik in de
gelegenheid werd gesteld één en ander te realiseren. Anneke van der Veen wil ik
bedanken voor het feit dat zij zorg droeg voor de dagelijkse werkzaamheden
waardoor ik de handen vrij had voor mijn onderzoek.
Ron, Katelijne, Trijnie en Marga, bedankt voor jullie hulp in de vorm van ‘even
meedenken, even een manuscript lezen’ en ‘even een computerprobleem
oplossen’.
Tevens wil ik op deze plek Prof. Bauke de Jong bedanken. Bauke, dankzij het
vertrouwen dat jij in mij stelde door mij in 1991 aan te nemen en mij op te leiden tot
cytogeneticus heb jij de basis gelegd voor dit proefschrift.
B.Sikkemaproefschrift2.doc
150
Verder wil ik mijn collega’s van de FISH-groep bedanken; Gineke, Hanny, Harke,
Hendrika, Inge, Marjan en Yolanthe, met name het laatste half jaar hebben jullie
door je hoge inzet er voor gezorgd dat de ‘tent bleef draaien’. Yolanthe wil ik
tevens bedanken voor het ontwerpen van de cover van mijn proefschrift.
Uiteraard gaat ook een woord van dank naar de prenatale groep voor het
verrichten van alle proeven ten behoeve van mijn onderzoek. Annet, Anke, Eddy,
Fia, Janet, Jolanda, Jurjen, Lennart, Madeleine en Reina, jullie moesten een stapje
harder werken om mijn paranimfen Jakob en Marian gelegenheid te bieden om
zich aan mijn onderzoek te wijden.
Jakob en Marian; zonder jullie niet aflatend enthousiasme en inzet had ik het
onderzoek nimmer kunnen voltooien. Heel veel dank daarvoor.
I would like to thank the members of the reading committee for their willingness
and quick review of the manuscript.
Auch möchte ich mich bei allen Raddatz’en bedanken. Schliesslich habt Ihr auf
direkte und indirekte Weise dafür gesorgt, dass ich Biologie studiert habe und dass
alles was ich mir beruflich vorgenommen habe, auch gelungen ist.
Mijn echtgenoot Jan wil ik bedanken voor diens aandeel in het redigeren en voor
zijn morele ondersteuning.
Tot slot zijn er nog veel mensen die een bijdrage hebben geleverd aan het tot
stand komen van dit proefschrift zonder dat zij hier met hun naam zijn genoemd.
Jullie allen ook hartstikke bedankt.
Birgit
top related