C horionic V illus Sam pling

1.1.2.1. B ackground

An alternative method of prenatal diagnosis, possible in the first trimester, was first proposed by Mohr in 1968. In the first 12 weeks of pregnancy the gestation sac is enveloped in a coat of chorionic villi, the total volume of which exceeds the total volume of the fetus. By the analysis of these trophoblast villi, removed by endoscopic biopsy, prenatal diagnosis could be achieved: Chorionic Villus Sampling (CVS). Chorionic villi are mitotic derivatives of the zygote and afford the advantage that most will disappear naturally by about 14 weeks leaving only a portion to become the definitive placenta. It was proposed that removal of some of these actively growing villi will not compromise subsequent fetal development. Mohr obtained samples of extra embryonic tissue from first and second trimester pregnancies (8-20 weeks). Villi were pulled into a hole in the side of a transcervically inserted ‘hysteroscope’ barrel using vacuum and cut from the chorion by a knife inside the tube.

The first blind transcervical vacuum biopsies were reported from China in 1975. These were carried out at 6-14 weeks’ gestation, and 4% induced miscarriage. Successful sex diagnosis was achieved in 94% of continuing pregnancies. Ultrasound guidance was first utilised by Kazy in 1982, and the procedure was adapted to that of today by Old et al., (1982) who used a 1.5mm transcervical catheter under real-time ultrasound guidance and suction to aspirate villi. A 90% sampling success rate was achieved on samples from pregnant women between 7-14 weeks’ gestation. By 1983 transcervical CVS had been used to obtain fetal material for the correct diagnosis of Tay-Sachs disease, sickle cell anaemia (HbS), Duchenne muscular dystrophy and argininosuccinicacidura in the first trimester of pregnancy (Ward et al,. 1983; Rodeck

et al, 1983; Simoni et a l, 1983; Silverman et al, 1992). Between 9-12 weeks of fetal gestation the uterine cavity is not completely filled by the amniotic sac. This is the ideal time for the passage of the catheter transcervically into the developing placenta.

Chorionic villi can also be obtained using a transabdominal ultrasonically- guided needle (Smidt-Jensen et al, 1984). This method is preferred for women with genital herpes, cervical polyps, or a markedly retroverted uterus. A needle is passed through the mother’s abdomen and uterine wall to gain access to the chorionic villi. Chorionic villi are removed by a suction biopsy with a single or two-needle (one within the other) approach. This transabdominal procedure causes patients no subsequent bleeding, but more cramping than with the transcervical approach. It is now more commonly used than the transcervical method.

Chorionic Villus Sampling, either transcervical or transabdominal, is now routinely carried out as a prenatal diagnostic procedure between 10 and 13 weeks of fetal gestation. The procedure has been found to have no adverse perinatal effects (Williams et a l, 1987). Direct metaphase preparations are also possible allowing a preliminary aneuploidy assessment after 24 hours. Even allowing for lengthy cell cultures and karyotyping, any ensuing TOP can be completed relatively early during pregnancy. Large amounts of DNA can be obtained and chorionic villi are better than amniocytes for fluorescent in situ hybridisation (FISH) analysis (Evans et a l, 1992). With the advent of high resolution ultrasound CVS can theoretically be performed routinely transabdominal ly from 6 weeks gestation using the freehand ultrasound- guided single needle aspiration technique (Brambati et a l, 1988). Several refinements have also recently been proposed including the use of biopsy forceps (Dumez et a l,

1984) jKkniblc ■ng.cjJlcjay^tym {Mnx.welWiil,_] Q8d) and an automatic puncturing apparatus (Popp and Ghirardini, 1990). CVS has now become an established procedure throughout the world providing material for fetal karyotyping and DNA analysis. CVS cannot however be used in assays for which amniotic fluid is essential, such as

measurement of the alpha-fetoprotein concentration. CVS is also the ideal technique for first-trimester prenatal diagnosis in multiple pregnancies, sampling for each individual placenta. In such cases, there is no greater risk of pregnancy loss than that associated with singleton pregnancies (Pergament et al, 1992). Problems however |may arise if the same placenta is erroneously sampled twice.

1.1.2.2. R isks

1.1.2.2.a. Procedure Related A bortions

Randomised trials have demonstrated that a woman assigned to undergo first trimester chorionic villus sampling has a 0.5 to 4.6% lower chance of a successful pregnancy outcome than a woman assigned to a second -trimester amniocentesis (MRC, 1991; Smidt-Jensen et al., 1991; Lippman et al., 1992. Table 1.1). A trial conducted by the National Institute of Child Health and Human Development estimated the procedure-related rate of fetal loss following CVS exceeded that for amniocentesis by 0.8% (Rhoads et al., 1989). It was found that the risks of spontaneous abortion is however significantly increased (up to 10.8%) among women in whom 3 or 4 attempts are made to place the transcervical catheter. These estimates comparing fetal loss due to CVS with fetal loss due to amniocentesis give a fair indication of the relative increase in risk to the fetus of the CVS technique. Due to the higher average age of women undergoing CVS compared to the general population, the calculated CVS procedure related pregnancy loss may be artificially high when compared to the background pregnancy loss rate at the same stage of gestation (Jahoda et al., 1991).

1.1.2.2.b. B leeding

Using the transabdominal procedure bleeding occurs in less than 1% of patients. With the transcervical method, however, vaginal bleeding occurs in 15-25% of women (Silverman et al., 1992; Rhoads et al., 1989). Subchorionic haematoma formation immediately after transcervical CVS has been shown to occur in 4% of patients (Brambati et al., 1987), but this is only rarely associated with adverse outcome and usually disappears after 16 weeks. Direct vascular injury of small branches of the utero-placental or unbilico-placental circulations may also lead to a retro-placental haemotoma and /or subchorionic haemorrhage and subsequently to a miscarriage. Most bleeding problems occur due to accidental placement of the catheter tip into the

vascular decidua basalis underlying the chorion frondosom and is related to operator experience and in particular to the number of attempts needed to obtain a sufficient villous sample (Rhoads et al., 1989; Silverman et al., 1992).

1.1.2.2.C. Infection

Intrauterine infection and chronic amniotic fluid leakage are two main medium-term complications of CVS, occurring between a few days and 3 weeks after the procedure. It has been found that 30% of catheters used for CVS have bacteria on them following the operation (Scialli et al., 1985). This has raised concern about the possibility of inducing vaginal flora into the uterus by the transcervical passage of the catheter, resulting in chorioamnionitis. However no association between these bacteria and chorioamnionitis has been found (Garden et al., 1985). Intrauterine infection after CVS is a rare (<0.1 %) but serious complication which can lead to maternal septic shock (Barela et al., 1986) and immediate evacuation of the uterine contents (Fisk and Anderson, 1987).

1.1.2.2.d. R upture o f m em branes

Rupture of the amniotic membranes due to CVS is a very rare occurrence. However, delayed rupture, due to injury of the chorion allowing it to become exposed and damaged, or low grade chorioamnionitis occurs in 0.3% of cases (Hogge et al., 1986).

1.1.2.2.e. R hesus Sensitisation

After CVS an acute rise in maternal serum alphafetoprotein levels has consistently been reported (Schulman et al., 1990) indicating the occurrence of fetal- maternal bleeding. This bleeding in most cases is asymptomatic, and serum levels return to normal ranges by 16-18 weeks. Fetal-maternal bleeding only becomes important in Rh (D) negative women, where a volume as low as 0.1ml of Rh-positive fetal blood in the maternal circulation can cause Rh sensitisation (Zipursky & Israels, 1967). This potential complication can be avoided if all Rh-negative mothers,

undergoing CVS, receive Rh(D) immune globulin subsequent to the procedure.

1.1.2.2.f. P lacental M osaicism

Most cells cultured from the CVS technique are fibroblasts originating from the placenta. A complete genetic identity between the fetal and extraembryonic tissues

cannot however always be assumed. The inner cell mass of a developing pregnancy is represented by approximately 16 cells of a 64 cell blastocyst. Confined placental mosaicism, in which the placenta is karyotypically different from the embryo, may arise as a consequence of mitotic non-disjunction in one of the other 48 extraembryonic cells early in development (Simoni and Sirchia 1994). This situation may also be exaggerated by the preferential allocation of abnormal cells to the trophoectoderm as seen in studies of mouse chimeras (James and West, 1994). Confined placental mosaicism is thought to occur in approximately 1% of human pregnancies (Silverman et al., 1992) resulting in both false negative and false positive diagnosis. Of 62865 karyotyped chorionic villus samples that were reported to EUCROMIC between 1986- 1992, 98.5% showed either a true normal karyotype (94.8%) or a true non-mosaic chromosomal aberration (3.7%) (Hahnemann & Vejerslev, 1997). True fetal mosaicism was diagnosed in about 0.15% and confined placental mosaicism in 1.0%. False positive non-mosaic aberrations were observed in 0.15% and false negative CVS results in 0.03%. Only 0.15% were unclassifiable. Ledbetter et al., (1990) reported the frequency of pseudomosaicism (artefact mosaicism in CVS culture metaphases, not present in the fetus or placenta) to be 1.8%, and indicated that the results of the direct method of chromosome preparation might be less representative of the fetal status than those from the full culture method. It has been shown that between 1-10% of chorionic villus samples need follow-up amniocentesis for an unambiguous diagnosis (Ledbetter et al., 1992; Lippman et al., 1992; Canadian Collaborative CVS-Amniocentesis Trial Group, 1989).

Amniotic cells are derived in the main from the developing fetus. Confined placental mosaicism is therefore not a problem when assessing amniocentesis samples. True fetal mosaicism, reflected in amniocenteses cultures is rare (0.25%) (Hsu and Perl is, 1984). Smidt-Jensen et al., (1993) confirmed that maternal tissue contamination and chromosomal mosaicism/pseudomosaicism occur in 1 and 0.4% respectively and early amniocentesis samples are altered by inherent mosaicism problems 10 times less often.

1.1.2.2.g. M aternal contam ination

Contamination of CVS samples with maternal cells is a potential problem, particularly with transcervical CVS. Matemal-cell contamination is uncommon however in experienced cytogenetic laboratories, where villus specimens are

meticulously separated from blood and decidual cells before cytogenetic analysis. In such situations maternal contamination is seen to occur at a frequency between 1.9- 3.8% (Ledbetter et al., 1990; 1992; Lippman et al., 1992).

1.1.2.2.h. F etal A bnorm alities

In 1991 Firth et al., reported an increased incidence of severe limb

abnormalities in babies whose mothers had undergone transabdominal CVS prior to 9 completed weeks of gestation (56-66 days). The abnormalities were rare congenital disorders and included varying degrees of limb deficiency, hypoglossia-hypodactylia syndrome and cavernous hemangiomas. Occurring in the normal population at a frequency of one incidence per 17500 live births, (Froster-Iskenius et al., 1988), these abnormalities occurred in approximately 1.5% of offspring from women undergoing early CVS. The babies concerned were found to have normal karyotypes. The

deformities sustained were consistent with incomplete morphogenesis and could be due to some kind of vascular insult, disrupting the normal development of the fetus. These observations were confirmed in subsequent studies (Burton et al., 1993; Gruppo Italiano Diagnosi Embrio-Fetali, 1993; Firth, 1994), including cases of transcervical CVS beyond the tenth week of gestation.

Some doubt however has been cast on these findings when further review showed that four of the cases reported by Firth and co-workers were not limb reduction defects (one case each of club foot or finger deformities, and two cases of annular constrictions without limb reduction defects); there had also been one duplication of cases published independently by different groups (Froster and Jackson, 1996). In their assessment of 138996 CVS follow-ups, Froster and Jackson concluded that there was no difference from the background population in the overall frequency or pattern distribution of limb deficiencies, with no correlation between gestational age at CVS. The majority of cases in this study however were collected after 10 weeks of gestation (mean 71.5 days), and complete absence of one or more limbs was observed in 4 cases, all collected between 53 and 63 days. The authors agreed that CVS should be avoided before 60 days gestation. This report was also questioned with regard to sample bias, either by failure of centres to report cases not felt to fit the spectrum of limb defects related to CVS or by the exclusion of cases with limb defects in patterns of multiple malformations owing to pregnancy termination or inadequate description (Evans and Hamerton, 1996).

Most authors now agree that very early CVS is associated with an overall 10- fold increase in the incidence of limb reduction effects and oromandibular-limb hypogenesis. This pattern and a much higher (2-3 times) incidence in limb reduction deformities before 9 weeks compared with sampling after 9.5 weeks suggest a causal relationship (Brambati et al., 1992; Rodeck 1993). CVS is now only carried out after 8- 9 weeks’ gestation, subsequent to the period of fetal organogenesis, apart from

exceptional cases.

1.1.2.3. T ranscervical CVS versus T ransabdom inal CVS.

Transabdominal CVS is associated with less secondary bleeding than transcervical CVS. In cases of a posterior placenta or in obese patients, the transcervical approach is probably safer and should be considered for CVS.

Conversely, if the amniocentesis is planned and the and-the placenta is anterior and obstructing, a transabdominal CVS would be more appropriate and probably safer (Jauniaux and Rodeck 1995). Transcervical CVS has the added advantage that there is no need to enter the peritoneal cavity or transverse sensitive structures. In practised hands both the transabdominal and transcervical techniques are equally safe (Jackson et al., 1992a; Jahoda et al., 1991).