1 Techniques of chorion villus sampling

1

Techniques of chorion villus sampling

R. H. T. W A R D

DEVELOPMENT OF M E T H O D S

Although diagnostic first trimester fetal diagnosis was described as recently as 1982, attempts to develop an obstetric technique for obtaining chorionic villi were made nearly a decade earlier by two Scandinavian groups following on the idea of transcervical first trimester diagnosis suggested by Mohr in 1968, and Hahnemann and Mohr in 1969. Both of these groups used hysteroscopy and were remarkably successful at obtaining villi from patients before termination of pregnancy. However, infection was responsible for the failure to culture in 50% of 39 cases (Kullander and Sandahl, t 973), and Hahnemann (1974) encountered considerable problems, including poor visibility due to maternal bleeding. Suspected damage to the fetal membranes resulted in the loss of 10 out of 27 pregnancies sampled up to 11 days before therapeutic abortion. Nevertheless, villi were obtained in 60% of the cases and it is interesting to contemplate how these groups would have fared had real time ultrasound been available to them. As it was, amniocentesis had at that time become established as an acceptable procedure and the possibilities for this approach to first trimester diagnosis were not explored further for some years in the Western World.

However, in China (Anshan, 1975) a simple method of chorionic villus sampling (CVS) without ultrasound was used for diagnostic fetal sexing in 100 pregnancies. The instrument used was a metal cannula 3 mm in diameter with an inner suction tube. Four of the 66 continuing pregnancies subsequently aborted and the sex prediction was incorrect in six of the fetuses. Although there were strict criteria for exclusion, the number of fetal losses is unexpectedly low in view of a figure of about 10% for spontaneous abortion in the first trimester even in the younger women (Gilmore and McNay, 1985; Gustavii, 1984). For a time fetal sexing in this way was discontinued, but the approach has been revived recently for genetic diagnosis using a plastic catheter. The value of this work, which received little publicity in the West until the late 1970s, was that it encouraged others to reevaluate the potential for first trimester diagnosis.

Baillibre's Clinical Obstetrics and Gynaecology--Vol. 1, No. 3, September 1987 489

490 R. H. T. WARD

Table 1. Early attempts at diagnostic chorionic villus sampling.

Authors, centre, year

Weight of tissue Number

Obstetric obtained of technique (Range rag) cases

Anshan, China, 1975 Kazy et al, Russia, 1982 Old et al, UK, 1982 "{ Brambati and Simoni, Italy, 1983 J Goossens et at, France, 1983

Metal cannula, no ultrasound Not known 100 Biopsy forceps and ultrasound 10-25 26

3 Portex catheter and ultrasound 30-60 1

Biopsy forceps and ultrasound 10-20 5

In the meantime, a different approach to early fetal diagnosis, that of endocervical lavage to obtain exfoliated trophoblastic cells from behind the cervical mucus plug, was advocated (Rhine et al, 1975; Rhine and Milunsky, 1979). However, whilst the method was simple and probably carried a negligible risk to the pregnancy, fetal karyotyping was possible in less than 50% of cases and subsequently proved unreliable (Goldberg et al, 1980). The increasing demand for fetal diagnosis for high risk conditions along with progress in DNA technology, particularly of the haemoglobinopathies (Williamson et al, 1981) and the evolution of ultrasound, was then to generate enough interest to reexplore the other first trimester approaches.

The first diagnostic CVS series for genetic disease was reported from Moscow (Kazy et al, 1982) and subsequently diagnostic series were published from England for the haemoglobinopathies (Old et al, 1982); from Italy for chromosome and metabolic disorders (Brambati and Simoni, 1983; Simoni et al, 1983) and from France for sickle-cell disease (Goossens et al, 1983) (see Table 1).

All of these groups used ultrasound-guided transcervical methods which have subsequently been adopted and modified by a rapidly increasing number of obstetricians. The transabdominal method was developed later and as yet experience of this approach is limited.

TECHNIQUES

Counselling Before undertaking any procedure for fetal diagnosis, couples should be carefully counselled, ideally by a geneticist. This is particularly important when experience of the first trimester techniques is still limited. In order to offer early diagnosis it is necessary to make practitioners and other health workers aware of the need for early hospital referral. A system that has been successful in achieving this is the introduction of an early counselling clinic (Knott et al, 1986) at which patients can be seen with a minimum of delay. As a significant number of pregnancies will be non-viable in patients of 38 years of age and over when seen before 11 weeks, an ultrasound scan is best performed before counselling. Apart from confirming viability (as assessed from fetal

TECHNIQUES OF CHORION VILLUS SAMPLING 491

movements and heart rate) the gestational age is determined (by measurement of crown-rump length) and multiple pregnancy excluded. At the same visit, after counselling, if CVS is planned by the transcervical route, a cervical swab can be taken and appropriate treatment given for any local infection. In addition, the Rhesus group of the woman should be determined to identify the Rhesus-negative patient who will require anti-D immunoglobulin after CVS.

Ultrasound for CVS

Most obstetricians require the assistance of an ultrasonographer during CVS although this is not necessary for transabdominal sampling, or if the single operator technique (Rodeck et al, 1983a) is used for transcervical aspiration. Optimum equipment is required for visualizing procedures within the pelvis but whether a sector scanner or a linear array machine is used is a matter of personal preference. A special transducer placed in the vagina has been advocated (Schaaps and Lambotte, 1985) and may have a place for use during CVS in the obese patient. Whatever approach is selected for CVS, careful scanning should be done, firstly to confirm viability and secondly to identify the location and limits of the placental site. For the former, as well as observing the fetal heart activity it is wise to confirm that the gestational sac has increased in size proportionately since the initial counselling clinic and that a twin pregnancy has not been overlooked. The definitive placenta or chorion frondosum is the thickest and most echogenic area around the gestational sac (Figure 1). Occasionally the site may not be obvious and confusion can arise with a marked decidual reaction. Whilst it is helpful to identify the endometrial canal, this may be inadvertently obliterated by pressure either from the transducer head or an overfull bladder. In order to be confident of the location of the placenta, the umbilical cord insertion should be found; locating the yolk sac does not give a guide to the position of the placenta. The limits of the placenta should be defined in a transverse scan as there is often a major lateral component at this stage of pregnancy. If, despite these measures, the placenta cannot be localized it is probably wise to consider postponing any attempt at CVS for a few days.

For the transcervical procedure itself, when the obstetrician is assisted by an ultrasonographer it is essential that a plan is discussed beforehand for each case. A reliable method of communication needs to be agreed so that the sampling device is readily located within the cervical canal and followed from the internal os to the proposed sampling site (Figure 2). The tip of the device should then be located with a transverse scan to confirm that it is well within placental tissue (Figure 3).

After the procedure, patients appreciate the reassurance of being shown the fetal heart activity and most obstetricians offer a further ultrasound examination between 16 and 18 weeks.

Techniques for CVS

All CVS procedures can be performed on an outpatient basis and neither premeditation or local anaesthetic are necessary. The available techniques are summarized in Table 2.

492 R. H. T. WARD

~ Placenta (Chorion Frondosurn)

Cord insertion

Yolk sac

Fetus

Figure 1. Longitudinal ul trasound scan showing chorion frondosum, cord insertion and yolk sac at 9 weeks gestation.

Transcervical methods: aspiration

This method remains the most popular approach although a variety of catheters and cannulae are now used for the ultrasound-guided technique described using the Portex catheter (Old et al, 1982; Ward et al, 1983). This technique was based on the blind sampling method used in China (Anshan, 1975) and similar work undertaken by Loeffier (Horwell et al, 1983). The lithotomy position is used (or, better, a colposcopy chair). After swabbing the external genitalia, the cervix and upper vagina are exposed either with a bivalve speculum or Sim's speculum and cleaned with Aqueous Betadine. At this stage a bimanual examination is carried out to determine both the position of the uterus and whether there is any significant lateral deviation. After reinserting the speculum and applying more aqueous Betadine to the cervix, scanning is recommenced. In a few cases the cervical canal and internal os can be readily identified, but the passage of a uterine sound clarifies the


Figure 2. Longitudinal ultrasound scan showing Portex catheter advanced to sampling site. Note full bladder.

s i tua t ion (Figure 4). The catheter , with in t roducer bent appropr i a t e ly , is passed gent ly into the cervical canal (a t enacu lum is not no rma l ly necessary) and di rec ted towards the p lacenta l edge. N o resistance should be encounte red

Figure 3. Transverse ultrasound scan showing tip of Portex catheter in right-sided posterior placenta.

494 R.H. T, WARD

Table 2. Techniques for chorionic villus sampling.

Transcervical

Direct vision and biopsy

Transabdominal

Blind or ~ B i o p s y Ultrasound-guided ~ - " - - - - A s p i r a t i o n

Fetoscope Chofionscope

Ultrasound-guided ~ ' - ~ S i n g l e needle " ' D o u b l e needle

at any stage. If the placenta is posterior in an anteverted uterus, the catheter can be rotated through 180 ° when it reaches the internal os. The passage of the catheter is readily followed on ultrasound. Ideally the tip should be placed well into the placental tissue (chorion frondosum) equidistant between the chorionic plate and decidua (Figure 5). The placenta beneath the umbilical cord insertion is best avoided because of increased vascularity. As the obturator is removed the catheter should be advanced slightly, thus fixing it within the placenta (Figure 6). The sample is taken by applying up to 5 ml of suction with a 20 ml syringe containing 3-5 ml of heparinized culture medium. This is accompanied by a slight ' to and fro' movement (known as 'hoovering' or 'pull and suck') which is necessary to avulse the tips of the villi. The sample can be seen within the catheter, which shows up ultrasonically as a parallel lined structure. The suction is maintained until the catheter is removed. The contents of the syringe are flushed down the catheter into a sterile petri dish

Figure 4. Longitudinal ultrasound scan at 9 weeks showing uterine sound in cervical canal to level of internal os.


Figures $ and 6. Longitudinal ul trasound scan showing catheter advanced to sampling site before and after removal of trocar (obturator).

and immediately examined under a dissecting microscope. This allows for immediate confirmation that an adequate sample has been obtained and any contamination with decidua can be avoided at this stage by careful selection. The amount of villi obtained is readily estimated with experience.

The success rate for one aspiration is between 80 and 85%. This depends on experience and, as demonstrated by Brambati (Table 3), has also led many

496 R . H . T . W'ARD

Table 3. Chorion villus sampling learning curve.

Ist attempt

Ist 100 Diagnostic cases 59% Success 2nd 100 Diagnostic cases 80% Success 3rd 100 Diagnostic cases 86% Success

(Brambati, 1984)

Data of International Fetoscopy Working Group, Beaune, September 1984.

obstetricians to restrict aspiration attempts to two, which will be successful in a further 10-12%. However, it is also prudent to limit the number of sampling attempts as this is one factor that correlates with increasing fetal loss (see Figure 7). The Portex (Trophocan) catheter has been used extensively worldwide, but many operators have utilized essentially the same technique with a variety of other catheters (see Table 4). Most of these catheters are designed for other purposes but Holzgreve has designed a special polyethylene catheter with barium added (Angiomed) to improve sonographic visualization (Hotzgreve and Miny, 1985).

The most widely used metal cannula is that devised by Rodeck for single operator use (Rodeck et al, 1983a). It is made from silver and is 23 cm in length with a finger plate which acts as a suction control valve. An obturator protrudes 1-2 m m from the cannula tip. After bending the cannula according

2 0 %

u) oo o

10%

I

(N=1884)

1 2 5 or>

Number of attempts

Figure 7. Fetal loss~ortex catheter (3 centres).


Table 4. Plastic catheters used for chorionic villus sampling.

Portex 1.7 mm x 21 cm (Trophocan) (Brambati, Wapner, Ginsberg) Intracath, 16 gauge (Golbus, Elias) Desert Co., 8 inch Sandy, Utah Angiomed 1.98 mm x 24 cm (Contains 23 % (Hotzgreve) barium sulphate Umbilical catheter 8 French 2.8 mm gauge

(Mackenzie)

Data presented at International Fetoscopy Work- ing Group, Rotterdam, September 1986.

to the uterine position and implantat ion site, it is guided by the opera tor with one hand whilst holding the transducer enclosed in a sterile plastic bag in the other. Villi are obtained after removing the stylet by at taching a neonatal mucus aspirator connected to a theatre pump, the suction being controlled by a finger over the hole in the cannula, and cont inuous ul t rasound surveillance being maintained throughout .

Ano the r metal cannula has recently been described by Liu (available through Femcare Ltd., Not t ingham) with the opening at the tip laterally set. A round-ended tip may minimize the risk o f contamina t ion o f the villus specimen (Figure 8). The metal cannulae require careful cleaning and

Figure 8. The Liu sampler for chorionic villi (with kind permission of Mr David Liu).

498 R, H. T. WARD

Table 5. Chorionic villus sampling techniques--'Portex' cannula.

Number Failures Fetal loss (%)

Chicago (Ginsberg) 1395 12 2.3 Milan (Brambati) 1500 10 2.7 Philadelphia (Wapner) 2107 7 1.8

5002 0.6% 2.2

Data presented at International Fetoscopy Working Group, Rotterdam, September 1986.

sterilization between cases and are suitable for up to 20 procedures before they have to be discarded. Tables 5 and 6 summarize experience of the transcervieal catheter techniques.

Transcervical methods: biopsy forceps

Fine biopsy forceps introduced through the cervix were used first by Kazy and colleagues in 1982 and have since been popularized in the West by Dumez (Goossens et al, 1983). He uses rigid straight forceps 20 cm in length and 2 mm in diameter (Storz Ref. 8591A). The tip when closed is smooth and rounded and only the upper jaw opens, having the effect of deflecting the fetal membranes at the time of biopsy (Figure 9). A tenaculum is an essential part of the technique, so that the forceps, after being gently inserted through the cervix, can be advanced to the placental site by suitable manipulation of the uterus. The biopsy is taken by opening the jaw in the placental substance close to the chorionic plate. Decidual contamination is uncommon (10%) and in 90% of cases the weight of the villus sample is in excess of 10 rag. Dumez (Dumez et al, 1985) finds no correlation between either the placental site or uterine position with the number of insertions required to obtain a sample. Vaginal spotting occurs rarely following the procedure. Although the posterior placenta in a retroverted uterus or anterior placenta in an anteverted uterus may present potentially difficult access, by modifying the bladder fullness and traction on the tenaculum a 100% success rate has been achieved.

Table 6. Chorionic villus sampling techniques--transcervical catheter.*

Number Failures Fetal loss(%)

london (Rodeck and Warren) 386 7 3.4 San Francisco (Golbus) 1968 21 5.0

Data presented at International Fetoscopy Working Group, Rotterdam, September 1986, * See Table 4 and text for details,


Figure 9. Biopsy forceps (with kind permission of Dr Yves Dumez).

The technique is considered to be reliable, and the forceps are strong and easy to clean and sterilize. A similar technique is used elsewhere in France (Rouquet in Paris), in Rome (Pachi) and in Cagliari (Monni). Table 7 summarizes the experience from three of these centres. A similar biopsy forceps, but with a curved shank, is also now in use at Queen Charlotte 's Maternity Hospital, London (Rodeck, personal communication).

Transcervical methods." brush

A specially designed brush has been suggested as an alternative way of obtaining villi transcervically. Using such an instrument Liu (Liu et al, 1983) found with practice that villi were collected in 84% of cases immediately prior to therapeutic abortion. The brush, which is 21 cm in length, 1.2 cm long and

Table 7. Chorionic villus sampling techniques-- transcervical biopsy forceps.

Number Failures Fetal loss(%)

Paris (Dumez) 380 1 3.7 Cagliari (Monni) 440 1 3~0 Rome (Pachi) 265 7 6.8

1085 0.7% 4.1

Data presented at International Fetoscopy Working Group, Rotterdam, September 1986.

5 0 0 k . H . T . WARD

Figure 10. System of brush and cannula for CVS (with kind permission of Mr David Liu).

3.0 mm wide (Figure 10) is delivered to the biopsy site by a metal introducer and cannula set (Rocket of London). Rotating the brush through 180 ° within the placenta entwines the filamentous villi. The brush is withdrawn into the introducer before the latter is removed from the uterus. The teasing out of the villi does not appear to affect satisfactory karyotyping but contamination with mucus or decidua occurs in two-thirds of the cases.

Transcervical methods--endoscopy

The fetoscope was used to obtain villi with biopsy forceps by several investigators including Rodeck (Gosden et al, 1982), Brambati (1983) and Gustavii (1983). The instrument usually used was the 1.7 mm Olympus

Table 8, Chorionic villus sampling techniques--transcervical endoscopic biopsy,

Number Failures (%) Fetal loss (%)

Gustavii-- 135 0.7 6.7 Fetoscopy Ghirardini-- 70 10 10 Chorionscope

Data from CVS Newsletter, September 1986, Editor: L. Jackson.


Figure 11. Needle guide for transabdominal CVS with access on either side for both 18 and 20 gauge outer needle (with kind permission of Mr Darryl Maxwell).

Selfoscope housed within a custom-built cannula 3.0 x4.7 mm in outer diameter.

The cannula has two attachments; one for infusion of physiological saline and the other covered with a rubber cap for the introduction of biopsy forceps. The method involves passing the fetoscope and cannula into the extra- amniotic space with a continuous saline infusion running. An opening is made through the decidua at the base of the amniotic sac, by pushing the forceps with the jaws closed into the placenta. After this manoeuvre the specimen is taken under direct vision but the site can also be checked with ultrasound. Afterwards the endoscope is removed, followed by the forceps, and the specimen is transferred into culture media for initial examination. Finally, after releasing the saline, the cannula is removed. Although an elegant technique and successful in the experts' hands it is noteworthy that none of the operators (Rodeck, Gustavii or Brambati) who used it in the early days of CVS have persisted with it.

However, biopsy under direct vision using the 'chorionscope' is still advocated by Ghirardini et al (1985). The endoscope (Wolf D 7134) has an outer diameter of 2.7 mm housed in two sheaths with an outer diameter of 4 mm. The assembled chorionscope has a collecting chamber (3.5 x 2.5 ram) close to the tip capable of obtaining a villus sample of 30 rag. Up to 5 ml of fluid can be injected between the optic part and the sheath to facilitate visualization and the villi are collected into the chamber by aspiration with a syringe. The chamber can then be closed, trapping the villi. It is not necessary to select villi from the chorion frondosum but villi from degenerate chorion

5 0 2 R . H . T . WARD

laeve must be avoided. The best gestational age for sampling is between 9 and 11 weeks. The experience of Gustavii and Ghirardini is summarized in Table 8.

Time will tell whether this technique is practical and safe enough to consider for routine CVS.

Transabdominal technique: aspiration

CVS by transabdominal aspiration is a natural extension of amniocentesis and placentacentesis (Alvarez, 1964), but the credit for developing a reliable technique must go to Hahnemann and his colleagues who advocated this approach in the early days of CVS (Smidt-Jensen and Hahnemann 1984; Smidt-Jensen et al, 1984). With a double needle system and using a special needle guide attached to the transducer head (Figure 11), which is enclosed in a sterile plastic bag, a predetermined needle pathway is delineated as a line parallel to the chorionic plate (Figure 12). The puncture line and placenta are readily lined up if necessary by adjusting the fullness of the bladder, although under normal circumstances the bladder is empty. The abdominal skin is cleaned with alcoholic chlorhexidine and sterile acoustic medium applied. The 15 cm needle guide (18 G and 1.7 mm inner bore) is inserted in one movement into the uterine wall and then advanced to the edge of the placenta. The stylet is removed and substituted with a 20 cm sampling needle (20 G and 0.7 mm or 0.8 mm inner bore) which protrudes about 1 cm beyond the tip of the guide needle. Villi are aspirated into heparinized buffered culture medium using a 20 ml syringe with a gentle to and fro movement of the sampling needle.

Figure 12. Transabdominal ul trasound scala to show predetermined needle pathway. Echogenic site of sample seen on pathway to the left.

TECHNIQUES OF CHORION VILLUS SAMPLING

Table 9. Chorionic villus sampling techniques--transabdominal needle aspiration.

503

Number Failures (%) Fetal loss (%)

Aalborg (Smidt-Jensen & Hahnemann) 375 1.9 3.2 Copenhagen (Bang) 165 4.2 1.8

Data presented by Hahnemann at International Fetoscopy Working Group, Rotterdam, September 1986.

Repeat aspiration may be needed, particularly if the 0.7 mm inner bore needle is used. The earliest practical time for sampling is 9 weeks but the method can be used into the second trimester and indeed has been advocated in the third trimester for rapid karyotyping (Nicolaides et al, 1986). Other needle combinations can be used such as 17 or 19 G or 20 and 22 G but greater suction and several aspirations may be required to achieve an adequate sample with the narrow gauge needles. Contamination with blood is common, but with decidua is rare. If autoclaving is not readily available, the needle guide can be sterilized by immersion in alcohol.

An elegant method using a piezo electric crystal at the end of the styler of the guide needle used for amniocentesis (McDicken et al, 1984) has been adapted for transabdominal CVS (W. E. MacKenzie, presentation at Second Interna- tional CVS Meeting, Birmingham 1985).

Ideally, any needle guide system should be tested beforehand in a water path to confirm that the needle travels exactly along the predetermined pathway.

A double needle guided system is not essential for transabdominal CVS and single 'freehand' needle can be used successfully. Although this does not produce such a 'rigid' system the overall needle size is smaller, and this is the method favoured by Brambati (personal communication, 1986). Villi can also be obtained by transabdominal aspiration using a similar vacuum system (Rizzo, personal communication, 1986) to that described by Rodeck for transcervical aspiration.

Transabdominal CVS is associated with minimal discomfort to the patient and vaginal spotting is uncommon.

Post C V S care

Initially it was the practice to observe patients overnight following CVS but this is not necessary, although they should be advised to rest for the day of the procedure and avoid strenuous activity including coitus for a week.

Maternal alpha-fetoprotein blood screening (MSAFP) for neural tube defects is still appropriate after CVS and can be conveniently timed with a follow-up ultrasound scan at 17 weeks. Although the MSAFP levels rise immediately after CVS by transcervical aspiration, they do return to normal (Ward et al, 1985; Warren et al, 1985; Mariona et al, 1986). Hahnemann (personal communication, 1986) reports a similar rise in 31% of patients after

504 R . H . T . WARD

Table 10. Relative contraindications to chorionic villus sampling.

Threatened abortion Cervicitis or vaginitis Fibroids Multiple pregnancy Rhesus sensitization Obesity

transabdominal aspiration. Thus it is prudent to offer Rhesus negative patients anti-D after CVS and subsequently at 28 and 34 weeks.

Vaginal spotting is comparatively common after transcervical aspiration, occurring in about 50% of patients. However, should the bleeding persist or be heavy, or leakage of fluid or pyrexia occur, patients should be told to report back immediately.

Success of CVS techniques

Two aspects need to be considered; the successful collection of a sample and successful testing in the laboratory. Varying amounts of villi are required depending on the indication. For fetal karyotyping and biochemical investigation 10 mg is usually sufficient but for DNA analysis, the quantity needed is between 20 and 50 mg, although diagnosis has been possible on much smaller amounts.

Whilst the amount of villi obtained may not influence the rate of fetal loss other factors that undoubtedly do are the amount of trauma, which is reflected by the number of attempts at CVS, and maternal age. Other variables between populations will certainly exist and too much should not be concluded by direct comparisons of experience with the different CVS techniques. The relative failure and fetal loss rates are shown for each method (Tables 5-9). After the initial learning process, failure to achieve a satisfactory sample occurs in less than 2% of patients (range 0.5-1.6%). However, some selection may have already removed certain patients with a greater than average risk, for instance patients with a threatened abortion or with an IUCD in situ. The relative contraindications to CVS are shown in Table 10 and except for cervicitis and vaginitis apply to all methods of CVS. If time allows, local infection can be treated to permit transcervical sampling but it may be more appropriate to elect for transabdominal aspiration. Similarly, the exact position of myomata will dictate the best approach for CVS.

In the presence of vaginal bleeding it may be wise to postpone CVS until at least the tenth week as this will avoid sampling in many of the pregnancies which may initially have a viable fetus but will inevitably abort. There now appears little doubt that the optimum time for transcervical aspiration is between 9 and 10 weeks. Early experience from San Francisco describes an 11.8% fetal loss when CVS was performed at less than 9 weeks and 24% when


Figure 13. Longitudinal ultrasound scan of a twin pregnancy at i0 weeks showing thick septum.

performed at 12 weeks or more. This compared with a 3.5% loss when sampling was performed between 9 and 11 weeks (Hogge et al, 1985).

Certainly for transcervical aspiration techniques villi are less easily obtained after 11 weeks and may result in a higher incidence of local haemorrhage. Small haematomas are also seen in a substantial number of patients after second trimester transabdominal aspiration. The latter appear to be of less prognostic significance than those following transcervical aspiration. The explanation may be that the latter technique may introduce bacteria more readily which will flourish in the presence of a haematoma.

CVS in multiple pregnancy

Experience of CVS in multiple pregnancies is still limited but does raise several specific problems. These include the difficulty in identifying and confirming the origin of the samples obtained and the possibility of contamination between two separate placentae. Little difficulty should arise in these respects if a thick septum is present (Figure 13) and if the sampling sites are conveniently separated. Careful ultrasound scanning with a video recording is essential to allow for review or in considering selective termination later. Both Wapner and Brambati (personal communications, 1986) suggest that a combination of transcervical and transabdominal CVS may be appropriate.

Clearly, for safe selective termination accurate identification of the affected fetus is essential. Fetocide can be performed by intrathoracic injection of potassium chloride (Rodeck, personal communication, 1986). Alternatively, suction aspiration of the affected fetus but without disturbing the trophoblast

506 R.H.T. W'ARD

Table 11. Complications of chorionic villus sampling.

Chorioamnionitis Damage to fetal membranes Haematoma formation Rhesus isoimmunization Placental abruption IUGR Premature rupture of membranes Prematurity

can be undertaken with a Number 10 suction catheter attached to a 50 ml syringe (Dumez, personal communication, 1986).

Appropriate follow-up may be required for confirmation of the normality of the surviving twin.

Complications following CVS (Table 11)

The immediate risk to the pregnancy is of haemorrhage, trauma to the amniotic sac and chorioamnionitis. Complete rupture of the amniotic sac is rare in the experienced operator's hands but partial damage to the chorion may permit bulging of the amnion through the deficiency, ultimately leading to rupture of the amnion and loss of amniotic fluid. This sequence of events may take many weeks and may account for a number of the late abortions at 18-20 weeks or premature rupture of membranes following transcervical CVS. In addition, chorioamnionitis could well play a part. However, the aetiology of late abortions that occur after transabdominal aspiration is unclear. There was initial concern that prematurity, placental abruption and intrauterine growth retardation may be associated with CVS but this has not

Table 12. Complications of chorionic villus sam- piing--experience from three centres.

San Francisco, Chicago, Philadelphia 3000 Cases CVS (December, 1985)

First "l 3 Acute infection 0.1% trimester ~ 8 Haematoma

82 Vaginal bleeding Second ) I20ligohydramnios Trimester J 6 Premature RM Third ) 2 Rh sensitization aggravated trimester k (both previously sensitized)

Data presented by Pergament at 2nd International Chorion Villus Sampling Meeting, Birmingham.


been realized. The pregnancy complications after transcervical aspiration from three groups are summarized in Table 12, as this aspect of CVS is covered in greater detail elsewhere in this volume.

Indications for CVS

The indications for CVS fall broadly into four categories.

Chromosomal disorders

This group includes women of advanced age, those with an earlier fetal chromosome aberration and those with a known parental translocation.

X-linked disorders

Whilst we have hitherto relied on fetal sexing as the only investigation in such situations, an increasing number of these conditions are becoming amenable to DNA analysis.

Metabolic disorders and fetal infection

Other conditions diagnosable using DNA analysis

In patients where the indication is other than for a chromosomal disorder fetal karyotyping should if possible be done. Apart from extending the scope for fetal diagnosis for that couple, the finding of any abnormal fetal karyotype will explain some of the subsequent spontaneous losses after CVS.

Choice of method

Comparatively few operators have had experience of more than one method of CVS although initially different transcervical techniques were tried by Rodeck et al (1983b) and Brambati (Simoni et al, 1983). The reinvestigation of the endoscopic approach followed the pioneering work using hysteroscopy, but experience now suggests that as a method it is impractical for large scale use. The other transcervical methods of aspiration and biopsy with forceps have their own advocates, many of whom have tried other methods before settling for one. Certainly, results to date do not suggest that there is any single method which is superior to the others. Indeed, as obstetricians will inevitably choose a method that they feel most confident with, both approaches will continue to be used in the foreseeable future. A small randomized study has been reported by Bovicelli and his colleagues (1986) in which they compared the two approaches with patients between 9 and 13 weeks. In each group of 60 patients there was one failure to sample and two spontaneous abortions. More than two insertions were necessary for transcervical aspiration (23%) compared with 13 % for the transabdominal group, but the direct preparation for fetal karyotyping was less successful for the transabdominal group (52% compared with 82%). They concluded that although the samples from

508 R . H . T . WARD

transabdominal aspiration were consistently smaller and therefore may be inappropriate for DNA analysis, it was a less uncomfortable procedure and quicker and easier than transcervical sampling. It should be noted that in this study transcervical aspiration was done up to 13 weeks, which is recognized as inappropriate for this method. Larger randomized studies would be valuable, although there are few centres that can offer the same level of expertise in both main methods. However, particularly those couples with a high genetic risk ought to be able to choose CVS should they present too late for safe transcervical aspiration, and fetal diagnostic centres will need to develop transabdominal CVS even if they have hitherto relied on transcervical techniques. In the meantime, until there has been greater experience o f transabdominal aspiration the claim that it has 'all the advantages of transcervical aspiration but none of the drawbacks' (Maxwell et al, 1986) cannot be substantiated.

S U M M A R Y

Whilst randomized studies into the safety of chorionic villus sampling (CVS) are a l ready under way the technique is now offered as an acceptable alternative to amniocentesis in many diagnostic centres. In counselling, the obstetrician can now quote a risk to the pregnancy of 2-4% which, even if the inevitable losses before 16 weeks are excluded, represents probably at least twice the risk of amniocentesis.

The evolution of the obstetric procedures has meant that the transcervical approach to CVS has been the most popular to date but there is now increasing interest in transabdominal aspiration as it minimizes the possibility of infection.

The two best known transcervical methods are aspiration with a plastic or metal cannula and biopsy with rigid forceps. The majority of aspirations have been performed using the Portex cannula technique. Using this, three centres (Milan, Chicago and Philadelphia) have had experience of over 5000 cases with a failure rate of less than 1% and a minimal fetal loss of 2.2%. However, the proportion of fetal losses may be between 4 and 7% by the time complete obstetric follow-up is available (Brambati et al, 1985).

A similar technique has been used with a variety of cannulae. The experience of the first 1000 cases from San Francisco (Hogge et al, 1986) led them to conclude that CVS by this technique was acceptably safe but that continuing investigation was needed before CVS was offered routinely as an alternative to amniocentesis.

The only other transcervical technique that is practical for routine use is biopsy with rigid forceps. The failure and fetal loss rates associated with this method are comparable to the aspiration technique.

The pioneering work of Hahnemann and his colleague Smidt-Jensen has established transabdominal aspiration as a reliable alternative approach. Its main advantage should be to minimize the risk of infection which is inherent in all transcervical techniques. In addition, it does not need to be confined to 9-11 weeks gestation and thus offers couples the possibility of diagnosis


between 12 and 14 weeks rather than wait ing for amniocentesis at 16 weeks. The qual i ty and quan t i ty of the sample depends on the size of the needle used and the technique is said to have high pat ient acceptability. Nevertheless, fetal losses occur following the procedure (1.8-3.2%) and it may not always be successful (failure rate 1.9-4.2%).

Diagnost ic problems exist in dealing with villi ra ther than amniocytes which may result f rom materna l con t amina t ion or the differences (biochemical and of the fetal karyotype) between the two. These need to be covered in pre l iminary counselling. It is hoped that the randomized studies will ascertain whether any long- term effects on the child may be associated with CVS, bu t with an increasing n u m b e r of children born after CVS there are no apparen t problems.

In the meant ime, even amongs t women with a comparat ively low genetic risk there is an increasing desire to consider CVS rather than wait for anmiocentesis , and indicat ions to date suggest that a l though the overall risk to the pregnancy is greater, this is often acceptable because of the oppor tun i ty of early diagnosis.

R E F E R E N C E S

Alvarez H (1964) Morphology and physiopathology of the human placenta. Obstetrics and Gynaecology 28:813.

Anshan Department of Obstetrics and Gynaecology (1975) Fetal sex prediction by sex chromatin of chorionic villi cells during early pregnancy. Chinese Medical Journal I: 117-126.

Bovicelli L, Rizzo N, Montacuti V & Morandi R (1986) Transabdominal versus transcervical routes for chorionic villus sampling. Lancet ii: 290.

Brambati B & Simoni G (1983) Diagnosis of fetal trisomy 21 in first trimester. Lancet i: 586. Brambati B, Simoni G, Danesino C et al (1985) First trimester fetal diagnosis of genetic disorders:

clinical evaluation of 250 cases. Journal o f Medical Genetics 22: 92-99. Dumez Y, Goossens M, Goue J et at (t985) Chorionic villi sampling using rigid forceps under

ultrasound control. In Fraccaro Met al (eds) First Trimester Fetal Diagnosis, pp 38-45. Berlin: Springer- Verlag.

Ghirardini G, Camurri L, Gualerzi C et al (1985) Chorionic villi sampling by means of a new endoscopic device. In Fraccaro Met al (eds) First Trimester Fetal Diagnosis, pp 54~59. Berlin: Springer-Verlag.

Gilmore DH & McNay MB (1985) Spontaneous fetal loss rate in early pregnancy. Laneet i: 107. Goldberg MF, Chen ATL, Young WA & Reidy JA (1980) First trimester fetal chromosome

diagnosis using endocervical lavage: a negative evaluation. American Journal of Obstetrics and Gynecology 138: 436-440.

Goossens M, Dumez Y, Kaplan Let al (1983) Prenatal diagnosis of sickle-cell anaemia in the first trimester of pregnancy. New England Journal o f Medicine 309:531-533.

Gosden JR, Mitchell AR, Gosden CM, Rodeck CH & Morsman JM (1982) Direct vision chorion biopsy and chromosome-specific DNA probes for determination of fetal sex in first trimester prenatal diagnosis, Lancet ii: 1416-1419.

Gustavii B (1983) First-trimester chromosomal analysis of chorionic villi obtained by direct vision technique, Lancet ii: 507-508.

Gustavii B (I984) Chorionic biopsy and miscarriage in the first trimester. Lancet i: 562. Hahnemann N (1974) Early prenatal diagnosis; a study of biopsy techniques and cell culturing

from extraembryonic membrane. Clinical Genetics 6: 294-306. Hahnemann N & Mohr J (1969) Antenatal fetal diagnosis in genetic disease. Bulletin of the

European Society of Human Genetics 3: 47.

510 R . H . T . WARD

Hogge WA, Schonberg SA & Golbus MS (1985) Prenatal diagnosis by chorionic villus sampling: lessons of the first 600 cases. Prenatal Diagnosis 5: 393~400.

Hogge WA, Schonberg SA & Golbus MS (1986) Chorionic villus sampling: experience of the first 1000 cases. American Journal of Obstetrics and Gynecology 154: 1249-1252.

Holzgreve W & Miny P (1985) Improved echogenicity of the catheter for chorionic villi sampling. In Fraccaro Met al (eds) First Trimester Fetal Diagnosis', pp 63 64. Berlin: Springer-Verlag.

Horwell DH, Loeffler FE & Coleman DV (I983) Assessment of a transcervical aspiration technique for chorionic villus biopsy in the first trimester of pregnancy. British Journal of Obstetrics and Gynaeeology 90: 196-198.

Jackson L (1986) Chorion Villus Sampling Newsletter, September 1986. Kazy Z, Rozovsky IS & Bakharev FA (1982) Chorion biopsy in early pregnancy: a method for

early prenatal diagnosis for inherited disorders. Prenatal Diagnosis 2: 39--45. Knott PD, Ward RHT & Lucas MK (1986) Effect of chorionic villus sampling and early

pregnancy counselling on uptake of prenatal diagnosis. British Medical Journal 293: 479- 480.

Kullander S & Sandahl B (1973) Fetal chromosome analysis after transcervical placental biopsy during early pregnancy. Acta Obstetricia et Gynecologiea Scandinavica 52: 355-359.

Liu DTY, Mitchell J, Johnson J & Wass DM (1983) Trophoblast sampling by blind transcervical aspiration. British Journal of Obstetrics and Gynaecology 90:1119 1123.

Liu DTY, Symonds M, Jeavons B & Norman S (1985) Transcervical chorionic villus biopsy with a brush. Prenatal Diagnosis 5: 349-355.

Mariona FG, Bhatia R, Syner FN & Koppitch F (1986) Chorionic villi sampling changes maternal serum alpha-fetoprotein. Prenatal Diagnosis 6: 69-73.

Maxwell D, Lilford R, Czepulkowski B, Heaton D & Coleman D (1986) Transabdominal chorionic villus sampling. Lancet i: 123-126.

McDicken WN, Anderson T, MacKenzie WE, Dickson H & Scrimgeour JB (t984) Ultrasonic identification of needle tips in amniocentesis. Lancet ii: 198-199.

Mohr J (1968) Foetal genetic diagnosis: development of techniques for early sampling of fetal cells. Acta Pathologica, Microbiologica et Immunologica Scandinavica 73:73 77.

Nicolaides KH, Soothill PW, Rodeck CH & Warren RC (1986) Why confine chorionic villus (placental) biopsy to the first trimester? Lancet i: 543 544.

Old JM, Ward RHT, Karagozlu F, Petrou M & Modell B (1982) First trimester fetal diagnosis for haemoglobinopathies: three cases. Lancet ii: 1413 1416.

Rhine SA & Milunsky A (1979) Utilization of trophoblast for early prenatal diagnosis. In Milunsky A (ed,) Genetic" Disorders of the Fetus: Diagnosis and Treatment, pp 527-539. New York: Plenum Press.

Rhine SA, Cain JL & Cleary RE (1975) Prenatal sex detection with endocervical smears. Successful results with utilising Y-body fluorescence. American Journal of Obstetrics and Gynecology 122: 155.

Rodeck CH, Morsman JM, Nicholaides KH, et al (1983a) A single operator technique for first trimester chorion biopsy. Lancet ii: 1340.

Rodeck CH, Morsman JM, Gosden CM & Gosden JR (1983b) Development of an improved technique for first trimester microsampling of chorion. British Journal of Obstetrics and Gynaecotogy 90: I 113-I 118.

Schaaps JP & Lambotte R (1985) Ultrasonic observation of pregnancy during the first trimester using a vaginal approach. In Fraccaro M et al (eds) First Trimester Fetal Diagnosis, pp. 78 91. Berlin: Springer-Verlag.

Simoni G, Brambati B & Danesino C (1983) Efficient direct chromosome analysis and enzyme determinations from chorionic villi samples in the first trimester of pregnancy. Human Genetics 63:349 357.

Smidt-Jensen S & Hahnemann N (1984) Transabdominal fine needle biopsy from chorionic villi in the first trimester. Prenatal Diagnosis 4: 163w169.

Smidt-Jensen S, Hafinemann N, Jensen PKA & Therkelsen AJ (1984) Experience with transabdominal fine needle biopsy from chorionic villi in the first trimester: an alternative to amniocentesis. Clinical Genetics 26:272 274.

Ward RHT, Modell B, Petrou M, Karagozlu F & Douratsos E (1983) Method of sampling chorionic villi in first trimester of pregnancy under guidance of real time ultrasound. British Medical Journal 286:1542-1544.


Ward RHT, Grudzinskas JG, Bolton AE et al (1985) Fetoptacental products as a prognostic guide following chorionic villi sampling. In Fraccaro M e t al (eds) First Trimester Fetal Diagnosis, pp 74-77. Berlin: Springer-Verlag.

Warren Re , Butler J, Morsman JM, McKenzie C & Rodeck CH (1985) Does chorionic villus sampling cause feto-maternal haemorrhage? Lancet i: 69 I.

Williamson R, Eskdale J, Coleman DV et al (1981) Direct gene analysis of chorionic villi: a possible technique for first trimester antenatal diagnosis of haemoglobinopathies. Lancet ii: 1125-1 I27.

Documents

1 Techniques of chorion villus sampling