Chapter 13:         When is karyotyping indicated?


13.1         The “At-risk” Patient

Genetic counseling may be offered to pregnant women in many instances.  Advanced maternal age is the one we are most familiar with, and the risk for aneuploidy and fetal abnormality increases as the woman ages due to the increasing incidence of meiotic non-disjunction38,139.  The aneuploidy rate for advancing maternal age is shown in Table 20.  Age-based screening for Down syndrome was advocated from the 1960’s on, but it has not been successful in substantially reducing the number of T21 babies60 because only 7% (1982) to 12% (1993) of pregnancies occur in women 35 years or older56, although this rate is obviously rising.  Despite the higher risk, only about 20% of all Down Syndrome pregnancies occur in this age group60. 

 

 

Maternal Age (yr) at Delivery

Aneuploidy Rate (%)

35

0.78

36

0.97

37

1.2

38

1.5

39

1.9

40

2.4

41

3.1

42

3.9

43

4.9

44

6.3

45

8.0

49

21.1

Table 20:  Maternal age-specific aneuploidy rates at time of amniocentesis. (From: Callen38.)

 

 

As the rate of aneuploidy also varies with gestational age due to spontaneous fetal demise, Snijders177 has developed more comprehensive tables which correlate the risks at various maternal and gestational ages for the common chromosomal disorders.  In her tables the risk of T18 in a 29-year-old woman, 16 weeks pregnant, is 1/2371.   The table for T18 is reproduced in the Appendix. 

 

13.2         Biochemical and 1st Trimester Ultrasound Tests

There is increasing utilization of biochemical tests from maternal serum taken in pregnancy98, 41.  Tests include Alpha Feto Protein (AFP), free and total beta-subunit of Human Chorionic Gonadatrophin (b-hCG), unconjugated oestriol (uE3) and Pregnancy Associated Plasma Protein A (PAPP-A).  In T21, AFP tends to be reduced and b-hCG raised192.  In T18, AFP, oestriol and b-hCG are often low, although there may be a bi-modal distribution (some low, some high) for AFP and b-hCG90.  These tests are usually performed in the second trimester.  PAPP-A and free b-hCG may be more useful in the first trimester26.  These test tend to have low to mid-range (35 - 50%) sensitivities and fairly high specificity (95%) depending upon the cut-off levels chosen.

 

These biochemical tests have been shown to be useful in discriminating chromosomal abnormalities, either alone or in combination, at various stages in pregnancy and in certain age groups26.  In combination with ultrasound, biochemical tests may be particularly effective147.  Thiagarajah184 suggest selective amniocentesis when the AFP is abnormal only in the presence of an abnormal ultrasound.  Of all the ultrasound soft signs, nuchal translucency in the first trimester has the highest sensitivity, according the data generated by Pandya149, but these are high-risk patients.  Some have recommended that a first trimester screening program of nuchal translucency measurements by ultrasound and biochemical markers to triage for early amniocentesis or CVS would have the highest pick up of aneuploid fetuses of any other test or combination of tests147,98.  Roberts et al227 criticized such a suggestion, as they felt the high rate of spontaneous loss in fetuses with nuchal oedema made this a merely a screening test for potential miscarriages. 

 

An abnormal ultrasound by itself has been considered a suitable indication for amniocentesis197, but many would consider this to be extreme over-caution, as the risk would depend upon the nature of the abnormality, the age of the patient and the certainty of the diagnosis.  CPC and two-vessel cord are two cases in point, where amniocentesis should not be offered on these isolated findings.  Nadel124 suggest that their “Sonographic Scoring Index” can be used in older women to reduce the utilisation of amniocentesis by adjusting the risk of aneuploidy for structurally normal fetuses.  Druzin70 considered offering amniocentesis to all pregnant women, regardless of age.

 

13.3         Risk of karyotyping

Degrees of operator skill, patient age, and gestational age will affect the actual risk of the karyotyping procedure157.  The total spontaneous loss rate in Victoria in 1988 attributable to amniocentesis (over and above the background rate) was 1.3% or 1/7794.  The CVS loss rates for 1987 to 1988 was 2.9% or 1/3494.  In discussing risks, many authors quote a risk from amniocentesis at  0.5%, or 1 in 200, which is approximately the risk for Down syndrome in a 37 year old woman.  Experienced practitioners and institutions report results as good or better than this.  Clearly there is discrepancy in skill from practitioner to practitioner and from institution to institution, and also in the quality of the follow-up and reporting of fetal losses, figures which greatly affect the estimation of risk94.  The only randomized controlled trial of amniocentesis was reported by Tabor183, who found a 1% loss rate.  The risks of amniocentesis are discussed by Ramsay and Fisk,157 and they conclude that this is the rate which they quote to their patients.  Other studies they quote indicate losses of 0.3% and 1.5% above controls.  Nicolaides also quotes a figure of 1%129.

 

There are several different methods of obtaining fetal cells for karyotyping.  These are:

·        conventional amniocentesis, which is performed at 16 weeks;

·        early amniocentesis, which is performed at 12 - 13 weeks;

·        chorionic villus sampling, which is performed at 9 - 10 weeks;

·        late chorionic villus sampling (placental biopsy), in the 2nd and 3rd trimesters;

·        fetal blood sampling by cord puncture, in the 2nd and 3rd trimesters;

·        fetal cell capture from maternal blood sampling;

·        fetal cell capture from cervical flushing.

 

Each of these carries a different risk and the selection of methods may depend on various factors.  For example, amniocentesis is not advisable in patients with oligohydramnios, and fetal blood sampling may be indicated in cases with IUGR or hydrops, where fetal blood gases measurements are required, despite its higher loss rate of about 1.4%3.  The last two methods mentioned above are essentially experimental techniques, which may offer opportunity in the future for risk-free karyotyping. 

 

13.4         A Balance of Risks

There is a balance of risks sought in prenatal diagnosis.  The risk of a chromosomal problem is weighed against the risk of fetal loss attributable to the procedure.  It is hard to find any author who will commit to a definition of what constitutes an ethically supportable balance of risks in prenatal diagnosis, but a common justification in many articles is that:

 “… the risk of aneuploidy be equal to or more than the risk of miscarriage from the test” (Nicolaides129.) 

 

Thus if the risk of aneuploidy is greater than or equal to 1/200, then the decision to perform amniocentesis can be supported.  Most authors in papers on CPC and amniocentesis use this argument, which equates to saying diagnosis of at least one chromosomally abnormal fetus is needed to balance the loss of one healthy fetus.  The estimation of risks therefore is crucial.

 

Nicolaides is of the belief that not only is the risk of amniocentesis is understated, but that the risks of aneuploidy are also often overstated by disregarding the effects of the maternal age, and that therefore the equation has been biased in favour of performing amniocentesis in younger women, and: 

“… that medically imposed decisions based on arbitrary equations of the burdens of miscarriage against those of the birth of a chromosomally abnormal baby are contrary to the principles of informed consent.”129  

 

Despite this opinion, there is increasing utilization of karyotyping amongst the “at risk” pregnant population, and more mothers are taking up the option, particularly in the over 35 age group95.  A significant proportion of pregnant women do not have karyotyping performed however.  Halliday95, studying the Victorian population, found that multiparous women are less likely to have a test, as are women of Asian birth, people from country areas, and those who give birth outside private hospitals, even though karyotyping is free in Victoria for women over 36.  In a low-risk, American, privately insured, population Druzin70 found that 77.5% of women refused the offer of amniocentesis.  Halliday95 suggested that the “inverse care law” might apply in many situations, whereby people from lower socio-economic groups, from non-English speaking communities and without easy access to secondary or tertiary health services (i.e. from country areas) are under-investigated and under-treated.  The fact that karyotyping does have an associated and significant pregnancy loss must force to us examine why we would allow healthy fetuses to be placed at risk in order to detect aneuploid ones.  Indeed, another question arises.  Why are we performing ultrasound in the first place? 

 

13.5         The “Routine Scan” has become the “Routine Anomaly Scan”

 As many as 97% of women in Victoria will have an ultrasound at some time during their pregnancy 93 according to one analysis.  Apparantly ultrasound is considered a very important part of antenatal care.  In 1990-1991 in Australia, obstetric ultrasound generated a Medicare bill of over $65m while the total bill of all other maternity services was $55m.  Why do doctors consider it necessary to order these scans, and to utilise so much of our health resources on them? 

 

Despite recommendations from the ASUM, RACOG and RACR and many others elsewhere, that: 

“an ultrasound examination should only be undertaken after discussion between the doctor and woman (or couple) as to the potential benefits and possible implications of an abnormal finding”,141 

 

many women seem to present for ultrasound under the assumption that the scan is being merely done to determine the sex of the baby and to promote bonding.  According to Green89, there are differences in the “agendas” of ante-natal care-givers and care-recipients.   DeCrespigny243 lists the four main purposes of an ultrasound examination as:

 

1.    establish dates

2.    check for multiple pregnancy

3.    locate the placenta

4.    check for fetal malformations.

 

Daly-Jones 245  suggested that most women who present for the 18-20 week routine ultrasound “did not fully comprehend the reasons for their scan and most were unaware of the risk of having a sonographically detectable abnormality.” 

 

In recent years, as equipment and training have dramatically improved, health professionals increasingly see the role of the routine ultrasound as a screening test for fetal abnormalities.  It was not long ago that ultrasound was considered “still too primitive to catch many physical defects.”22  The task of checking the “good” things – confirming gestational number (twins, etc), gestational age,  placental position, amniotic fluid and extrauterine problems – while still a crucial part of antenatal care, has arguably become reduced in its significance as part of the entire examination.  The 18-20 week scan has recently come to be referred to as the “routine anomaly scan.”  Controversies such as the CPC discussion have arisen out of this change in focus to the fetus as an independent patient.

 

13.6         Trials of Routine Ultrasound

Large studies such as the RADIUS201,207,208 two stage ultrasound study, the Helsinki Trial210, the Stokholm Trial,241 and meta-analysis of several studies206, have shown that routine scanning, as opposed to referred ultrasound for specific clinical concerns, while being effective in detecting these “good” things, has no measurable benefits in terms of  many traditional pregnancy and neonatal outcomes.  The RADIUS study concluded that routine ultrasound screening should not be performed in America due to the high added costs to the health care system.  Several of these studies have shown improved perinatal mortality due to the increased detection of anomalous fetuses and their subsequent early termination210, 206,241.  The initial reports of the RADIUS study, surprisingly failed to classify major fetal anomalies as adverse outcomes in the majority of cases215. 

 

According to Skupski, “The ability of routine ultrasound to detect fetal anomalies is central to consideration of its clinical value.231  Berkowitz states that: 

“assuming… that epidemiological evidence of a favorable cost: benefit ratio is required for any screening test, the rate of detection of anomalies in the RADIUS study does not appear to support a national policy of routine ultrasonography, despite the fact that it may be of psychological benefit to many women and have a favorable effect on the outcome of some pregnancies.  This should be accepted as a challenge by those who provide ultrasound examination to pregnant women…”213

 

In summary, according to the rising paradigm, routine ultrasound scanning at 18-20 weeks only confers benefits to the outcome of pregnancies when fetal anomalies are detected, and the challenge has been given for sonographers to demonstrate how well they are scanning. 

 

13.7         How Well Are We Scanning?

Even this benefit is only obtained when the scanning is performed at a suitable level of expertise, and the RADIUS201 study has been especially criticized for basing its conclusions on relatively poor quality scanning65,211,214,215,216,217,231.  It has quoted its overall detection rate for anomalies at 35%, with only 16% of anomalies being detected at the earlier scan (18-20wks), and 13% detection in non-tertiary centres, that is mainly private obstetric clinics.  Other studies have also shown low rates of detection.  Stoll181 blames the poor results (34.5%) in his study of chromosome disorders to “the inadequate qualification of some operators and  … the insufficient duration of the routine examination” rather than inadequacies of the equipment. 

 

Some studies however, and select groups within larger studies (including the RADIUS study), have shown good detection rates which would be consistent with cost-effectiveness of routine ultrasound as a screening tool compared to routine maternal blood testing, for example65.  In the Helsinki Trial210, studies were done at either of two centres.  At the City Hospital in the study, the rate of detection of anomalies was 36%, whereas at the University Hospital, the rate of detection was 76.9%.  An earlier study from Finland by Rosendahl205, detected 58.1% of malformed fetuses. 

 

In a study by Chitty et al204, from London and Luton, the scans were performed in the maternity department of a district general hospital.  “All scans were performed by radiographers who have between one and 10 (average two) years’ experience of obstetric ultrasonography.  They are supervised by an obstetrician and a radiologist, neither of whom have any particular skills in obstetric ultrasonography.”  The sensitivity of their scanning was 74% for all abnormalities, and 83% for lethal or severe abnormalities. 

 

Shirley et al171 from Middlesex detected 61% of all abnormalities and 73% of major or lethal abnormalities, again with scans performed by radiographers with the British DMU, even though only 15 minutes was available for each scan in this study.  A study by Goncalves202 from Nashville detected 53% of all abnormalities and 89% of lethal conditions. 

 

The detection of fetal abnormalities is likely to continue to improve into the future with improved training and skill of sonographers and sonologists and the inevitable improvements in equipment.  In centres where obstetric ultrasound is performed by qualified or well supervised student sonographers with a medical imaging background, using high quality equipment, a detection rate of about 75% should be expected, as evidenced in Chitty’s survey204.  As mentioned, Stoll was concerned with the duration of the routine scan181.  In Australia, 20 to 30 minutes are commonly allocated for an obstetric scan.  With the growing list of fetal anatomy which it is expected to be checked, many centres have moved towards a 40 or 45 minute scan.  It has been emphasized that a thorough anatomy survey should be an essential part of every routine ultrasound examination and that the concept of a limited “Level 1” and a detailed “Level 2” examinations are no longer recommended or relevant212,231. 

 

It is crucial for sonographers who perform the 18-20 week scan to be aware of the many fetal abnormalities which ultrasound has the potential to detect, but there is a range of significance of fetal anomalies, some being more crucial to fetal survival or obstetric management than others.  The concept has arisen therefore of “hard” and “soft” signs in relation to the risk for chromosomal abnormalities of the fetus.

 

13.8         Hard Signs and Soft Signs

The search for fetal structural anomalies often produces ambiguous or controversial findings.  As various anecdotal reports and larger series produce more and more information for sonographers to interpret in their clinical situations, new abnormalities are associated with aneuploidy.  No doubt some of these reports show mere coincidental associations.   At what point should sonographers take these reports more seriously, and treat a finding such as CPC as a real association?  Two terms have come into vogue. 

 

“Hard signs” are fetal structural anomalies that have an established and strong association with chromosomal abnormalities to the point where the balance of risks is such that amniocentesis is warranted.  Crane, in Chapter 3 of Callen38, provides a list of suggested hard signs. 


ISOLATED FETAL ANOMALIES WARRANTING AMNIOCENTESIS

Holoprosencephaly

Ventriculomegaly

Agenesis of the corpus callosum

Thickened nuchal fold or nuchal oedema

Cystic hygroma

Congenital heart disease

Oesophageal atresia

Duodenal atresia

Diaphragmatic hernia

Omphalocele

Non-immune hydrops

Generalised arthrogryposis

Table 21: Hard signs, from Callen38

 

 

“Soft signs” are fetal structural anomalies in which the balance of risks remains weak or controversial, and caution is deemed to the better option. 

 

When several soft signs are combined in the one fetus, the risk for aneuploidy increases steeply.  The presence of a single ultrasound detected abnormality has been estimated to place the fetus at a 2% risk for aneuploidy.  Individual anomalies carry a specific risk for aneuploidy however and this may range up to 70% for nuchal thickening, to less than 0.25% for isolated choroid plexus cysts.   Nicolaides133 found that the risk increased markedly when multiple abnormalities were present as shown in Table 22.

 

 

No. of defects

No. with chromosome abnormality / total with defect (%)

any

            301/2086   (14)

> 2

              276/958   (29)

> 3

              223/488   (48)

> 4

              153/248   (62)

> 5

                93/133   (70)

> 6

                  58/80   (72)

> 7

                  33/40   (82)

> 8

                  22/24   (92)

Table 22: Frequency of chromosome abnormalities and number of ultrasound-detected defects.  (Adapted from Nicolaides133)

 

 

Benacerraf17,18 and others124 have developed and refined a scoring system for aneuploidy (particularly T21). They ascribe points to various abnormalities detected on ultrasound and tally up the score to determine the risk.  The current system is shown below.

 

 

Nuchal fold ³ 6mm

2

Structural defect

2

Short femur

1

Short humerus

1

Pyelectasis

1

Echogenic bowel

1

Choroid plexus cyst

1

Table 23: Sonographic Scoring Index. (From Nadel124)

 

 

They suggest that amniocentesis should be performed when the score is 2 or higher. 

 

13.8         The “Isolated” Sign

When an abnormality is suspected and subsequently confirmed, the sonographer and sonologist must determine if it constitutes a soft or a hard sign.  But in either circumstance, there will considerable increase in risk if any other abnormalities are found.  It is in the search for the second, third and fourth abnormalities that the quality of the sonographic examination will become evident.  An abnormality may be considered “isolated” only when the other soft signs have been adequately searched for and excluded.  There is an immediate danger when scanning abnormal fetuses in concentrating on the initial finding to the detriment of the quality of the remaining parts of the examination.  If the sonographer is aware of this effect, the so called “sunburst phenomenon”, the examination can be completed fully.  The use of a checklist, which guides the sonographer in the complete evaluation of the fetus is strongly advocated.  The checklist used at the Geelong Hospital is provided in the Appendix.  Further discussion on this checklist is provided in Chapter 15.

 

13.9         Having Found a Chromosomal Sign…?

Of the 2,560 malformed fetuses and infants in the PDCU report for 199356, 280 had a chromosomal abnormality; a rate of 11%.  In 105 (37%) of these cases, the pregnancy was terminated.  After stillbirths, neonatal and infant deaths, 134 chromosomally abnormal infants were still alive.  It is sobering to think that 89% of malformed fetuses do not have a chromosomal problem.  Other malformations reported include anencephaly, spina bifida, cardiac defects, bony dysplasias, multi-system disorders and metabolic disorders such as thalassemia.  The causes for these disorders may be genetic, environmental and sporadic mutation of unknown origin.  The relationship of neural tube defects such as spina bifida and low levels of maternal serum folate and vitamin B12 is well established242, for example. 

 

If the ultrasound detected sign is considered “hard”, or if multiple “soft” signs are discovered, the patient would normally be offered non-directive genetic counselling.

 

13.10       What Does the Knowledge of Karyotype Provide?

The knowledge that there is a chromosomal disorder underlying a hard sign does provide more information for genetic counseling, aiding in the explanation of the prospects for the pregnancy and the longer term quality of life, severity of handicap, presence of mental retardation etc. of the surviving person41,44,60,244.  When severe autosomal-based or structural abnormalities are present, a high rate of acceptance of termination has been evident89,97.  Absence of a chromosomal cause may make counseling less clear cut97,156.  An example is a low spina bifida defect in a chromosomally normal fetus, where there is a possibility of normal intelligence and only a mild to moderate decrease in quality of life, given adequate surgery and social support.  Skeletal and metabolic abnormalities that may affect the fetus, similarly present no chromosomal disorder in the usual tests, but may exhibit instead Mendelian inheritance patterns, or be triggered by some unknown factor.  Diagnosis of problems like thalassemia, cystic fibrosis, Gaucher’s disease, etc., may one day rely on amniocentesis and analysis of one specific portion of a chromosome using markers yet to be developed or currently under development.

 

But changes in societal attitudes towards prenatal testing and persons with disabilities must also be considered121, 232.  There is a pressure to expand prenatal testing and concomitant pressure, both economic and social, to reduce the institutionalised support for disabled persons.  The burden of care for the disabled appears to fall more heavily on the family, usually the mother, as governments provide less and in some cases, no social support156.  In this society, where women have equal economic and social responsibilities with men, the perception of coercion into prenatal testing and the threat of no support for a disabled child, it not surprising that there has been a strong feminist response against what may be interpreted as a less than subtle introduction of eugenics121,232,233 , or even as a backlash against feminism and the autonomy of women. 

13.11       Trisomy 18 and Amniocentesis

In fetuses with multiple abnormalities, T18 is the most common aneuploidy244.  In T18, the prospects are not good for the fetus, and many obstetricians consider that it is effectively a lethal disease24.  It has been estimated that up to 70% die between 16 weeks and term.  Gross91 quotes a loss rate of 67.5%, based on the work of Hook.  Snijders177 quotes an expected loss rate of 74% and an observed loss rate of 68% based on the work of Warburton.  Therefore most of those T18 fetuses diagnosed after ultrasound will not survive to term, even without destructive intervention.  When they are live born, about 50% will die in the first week, with only 5 - 10% of these living a year244.  Long term survivors have mental retardation, development delay, slow progress and require intensive and continual care44.  Of T18 fetuses alive at 16 weeks, only 1.5 to 3% are destined to be alive at one year of age.  In 1991, 36 T18 fetuses were reported to the malformations unit of the PDCU93 in Victoria.  It would be expected that 0.5 to 1 infant would survive to one year from this group.   T18 does not provide a major economic strain on the community.

 

As T18 fetuses experience growth retardation in the 3rd trimester and difficulties during delivery, obstetrical intervention may be considered when the condition is unsuspected179.  Caesarean section is contra-indicated in aneuploid pregnancies.  The frequent finding of structural and cardiac defects may mean that corrective surgery is contemplated for the live-born T18 infant in the neonatal period24.  In a child with T18, many would consider that such surgery would cause unnecessary pain and suffering on the infant who has such a poor prognosis, as well as being poor utilization of scarce health resources24.  These situations should be avoided if possible.  Advance warning of a chromosomal abnormality from a prenatal test would be invaluable, as would reduction of the problem by prior termination of affected fetuses. 

 

Screening tests for lethal and uncommon conditions (such as T18) have been criticized for the physical risk of the test itself, the psychological effects of the test results, the physical effects of the test results and ensuing follow-up tests (in this case amniocentesis), and the cost effects of testing on the health care system199.  Most of the chromosomal screening protocols have been targeted for Down Syndrome, as it is more frequent and has a higher survival rate.  T18, while very common at amniocentesis, attracts less public attention than T21 because its low birth and survival rates235, 236 and the severity of mental retardation in survivors mean that sufferers are not publicly visible.  The high number of false positive results and the possibility than another abnormality than that screened for will be found in tests like serum analysis and ultrasound can have profound and negative effects on the screened population234,199

 

Is it cost effective to screen for T18 in a group of patients considered to be at increased risk due to the presence of a soft sign such as CPC?  A method of analysing the effects of different amniocentesis protocols is undertaken in the next chapter.

13.12       Conclusion

The RADIUS study and its commentators, in effect demonstrated that ultrasound screening is only worth doing if it is done well.  Sonographer training and accreditation, choice of ultrasound machines, and the time allocation for scanning are the scan quality factors over which we have some control.  Government decisions about health service reimbursement may in the future affect the routine 18-20 week scan, fuelled by information of the poor performance of ultrasound. 

 

 



Commonwealth Government, Health Insurance Commission data.