ultrasound-guided procedures in obstetrics

Ultrasound-Guided Proceduresin Obstetrics
Christian A. Chisholm, MD*, James E. Ferguson II, MD, MBA
KEYWORDS

� Ultrasound � Obstetrics � Diagnosis � Complications

KEY POINTS

� Amniocentesis and chorionic villus sampling are the most common techniques for obtaininga prenatal karyotype on a fetus at risk; in appropriately-trained hands, either can be performedwith an acceptably low risk of complications.

� Ultrasound guidance is used for nearly all invasive diagnostic procedures in obstetrics, and seems toimprove the odds for a successful procedure, but may not reduce the risk of procedure-related preg-nancy loss.

� Ultrasound guidance is essential for fetal therapeutic interventions such as fetal blood transfusion,vesicoamniotic shunt placement, and thoracentesis.

com

INTRODUCTION

Ultrasound is used to provide guidance for inva-sive procedures during pregnancy because it isthe primary imaging modality used in the antenatalperiod. In addition to the documented safety ofultrasound and the absence of ionizing radiationexposure to the fetus, ultrasound offers the uniqueadvantage of real-time imaging, allowing contin-uous ultrasound monitoring of invasive diagnosticprocedures. This quality of ultrasound monitoringallowed the initial development and further refine-ment of fetal therapeutic procedures.

Amniocentesis was the first procedure per-formed for invasive fetal diagnosis, for Rh diseasein the late 1950s, and for genetic diagnosis in thelate 1960s. Although preprocedure mapping ofan amniotic fluid pocket was possible using staticscanning, real-time ultrasound guidance was notintroduced until the 1980s. With this advancecame the opportunity to introduce additionaldiagnostic techniques including chorionic villussampling, fetal blood sampling, and aspiration offetal urine to assess fetal renal function. Althoughintrauterine fetal transfusion was the earliest fetal

Disclosures: None.Department of Obstetrics and Gynecology, The Univer22908, USA* Corresponding author.E-mail address: [email protected]

Ultrasound Clin 7 (2012) 325–335doi:10.1016/j.cult.2012.03.0061556-858X/12/$ – see front matter � 2012 Elsevier Inc. Al

therapeutic intervention (with intraperitoneal trans-fusion again dating to the era before ultrasoundguidance was available), several additional thera-peutic interventions for the fetus have been devel-oped in the modern era.

In most centers, genetic counseling is an inte-gral part of a patient’s decision to undergo invasiveprenatal diagnosis testing, whether by amniocen-tesis or chorionic villus sampling. A genetic coun-seling encounter allows the opportunity to gathera detailed family history and to provide counselingregarding the relative risks of invasive diagnostictesting and noninvasive screening. The core prin-ciple of genetic counseling is its nondirectivenature, in which the aim is to provide adequateinformation to facilitate patient decision making.

AMNIOCENTESIS

There are several current indications for amniocen-tesis (Box 1). Amniocentesis was first proposedas a routine consideration for women older than35 years in the late 1970s by the National Institutesof Health. The threshold of age 35 years, whichpersists today in the International Classification

sity of Virginia, PO Box 800712, Charlottesville, VA

l rights reserved. ultrasound.th

eclinics.

mailto:[email protected]

http://dx.doi.org/10.1016/j.cult.2012.03.006

http://ultrasound.theclinics.com

Box 1Indications for amniocentesis

Advanced maternal age

Abnormal genetic screening

� First-trimester screening

� Second-trimestermaternal serumscreening

Family history of genetic disorder

� Prior child with fetal aneuploidy or othergenetic disorder

� Other family history of a disorder forwhich prenatal diagnosis by DNA analysisis available

Abnormal ultrasound findings

Increased maternal serum a-fetoprotein

Red cell or platelet alloimmunization

Fetal lung maturity testing

Evaluation for intra-amniotic infection

Evaluation for rupture of fetal membranes (dyetest)

Therapeutic reduction of amniotic fluid volumefor symptomatic hydramnios

Chisholm & Ferguson326

of Diseases (ICD-10) diagnosis code for advancedmaternal age, was selected in recognition ofepidemiologic data showing a higher risk for fetalDown syndrome in women more than 35 yearsold compared with those age 30 to 34 years. Thisage threshold to offer amniocentesis also tookinto account the relative availability of trained oper-ators and cytogenetic laboratories, and made anattempt to balance the risk of an abnormal findingagainst the risk of a procedure-related loss. Thisthreshold criterion is still used as the basis forestablishing criteria to offer invasive testing withother forms of screening (1:270 risk or higher ofDown syndrome).1

However, the risk of fetal aneuploidy is a contin-uous variable that increases with maternal age,therefore the use of maternal age as the solescreening criterion for fetal aneuploidy has majorlimitations. It is widely recognized that most(70%–80%) aneuploid conceptions occur inwomen less than 35 years old.A substantial goal of new advances in prenatal

diagnosis is the reduction in procedure-relatedpregnancy losses, in conjunction with enhanceddetection rates of fetal aneuploidy. Noninvasivescreening tests such as first-trimester geneticscreening or second-trimester maternal serumscreening offer women less than 35 years old theopportunity to be identified as having an increased

risk for fetal Down syndrome and thus the optionto pursue invasive diagnostic testing. Screeningtests also present the opportunity for womenmore than 35 years old to undergo screeningbefore making a decision about invasive diag-nostic testing, and perhaps opt out of diagnostictesting (and its attendant risk) if their likelihood ofaneuploidy after screening is reduced. An impor-tant element of counseling women more than35 years old regarding screening is a discussionof the false-negative and false-positive rates asso-ciated with screening. The recent developmentof noninvasive techniques to identify alterationsin the rate of single-nucleotide polymorphisms(SNPs) in maternal blood raises the possibility ofidentifying fetal Down syndrome and other fetalaneuploidies with high precision and few false-positives by noninvasive means.As new noninvasive tests become commercially

available, the indications for invasive prenataldiagnosis will decrease, and the number of proce-dures being performed will decrease in parallel.The rate of procedure-related loss is also ex-pected to decrease once this developmentbecomes a reality. A downward trend over timein procedure volumes has already been reportedby some investigators.2,3 As the number of womenwith an indication for prenatal diagnostic testingdeclines, it will be essential for specialists inmaternal-fetal medicine and perinatal geneticsto continue to receive training and maintaintheir skills in invasive prenatal diagnostic proce-dures, which have largely become the domainof subspecialists, and are rarely performedcurrently by general obstetrician-gynecologists orby radiologists.

Technique

In the modern era, continuous ultrasound moni-toring/guidance of amniocentesis is consideredessential. The use of ultrasound allows continuousguidance and visualization of the needle into theamniotic fluid within the gestational sac, as wellas identification of fetal movements and changein fetal position, either of which could result in fetalinjury if the needle were inserted without ultra-sound visualization. Ultrasound guidance allowsthe identification of membrane tenting, a possibleexplanation for absence of fluid return. Amniocen-tesis studies published during the era of contin-uous ultrasound guidance have not showna reduction in the pregnancy loss rate attributableto amniocentesis compared with studies of pro-cedures performed before ultrasound guidancewas routine, although they have shown a reductionin the rate of obtaining bloody amniotic fluid and

Ultrasound-Guided Procedures in Obstetrics 327

the number of needle insertions required to obtaina fluid sample. Fetal injury at the time of amniocen-tesis is rare. After patient counseling is completeand informed consent has been obtained, theabdomen is prepared with an antiseptic solution;we prefer the use of chlorhexidine-alcohol, whichis allowed to dry on the abdomen before perform-ing the procedure. Povidone-iodine solution is anacceptable alternative. Coupling gel is applied tothe ultrasound transducer, which is inserted intoa sterile probe cover or sterile glove, and sterilecoupling gel is applied to the abdomen. Underreal-time ultrasonography, an appropriate needleinsertion site is selected. The ideal insertion siteis a pocket of amniotic fluid that is free of fetalparts and does not require traversing the placenta.In cases in which the placenta is implanted anteri-orly, placental penetration may be unavoidable;however, ultrasound guidance makes it possibleto avoid the edges of the placenta as well as thecentral area where the umbilical cord inserts and,ideally, to avoid traversing large vessels on theplacental surface. Color Doppler may aid in identi-fying these vessels. These precautions reduce thelikelihood of a bloody sample. We use a 22-gaugespinal needle, selecting one with a modified echo-genic tip in obese women or when oligohydram-nios is present to enhance visualization of theneedle tip. Some operators prefer a 20-gaugeneedle. The needle is inserted along the plane ofthe transducer (coplanar), with or without the useof a needle guide, to allow visualization of the nee-dle path. It is advanced through the myometrium(Fig. 1) and into the amniotic sac, where the styletis withdrawn. The usual quantity of amniotic fluidwithdrawn for genetic analysis is 20 to 30 mL,with the first 1 to 2 mL either discarded or usedfor a-fetoprotein determination, to avoid potentialcontamination of the amniotic fluid cell culturewith maternal cells on the needle tip from theprocess of insertion. After needle removal, fetal

Fig. 1. Amniocentesis needle in place.

heart rate is confirmed and the site of myometrialor placental penetration is visualized to ensurethere is no bleeding. Rh immune globulin is admin-istered if the woman is Rh negative and unsensi-tized. She is advised to avoid strenuous activityfor 1 to 2 days after the procedure, to anticipatemild cramping for several hours, and to reportvaginal leakage of watery amniotic fluid, vaginalbleeding, malodorous vaginal discharge, lowerabdominal tenderness, or fever. Depending onthe test requested, preliminary cytogenetic resultsmay be available in as little as 1 to 2 days if fluores-cent in-situ hybridization techniques are used, ormay take as long as 2 to 3 weeks with conventionalculture techniques or if DNA analysis is to beperformed.

In the case of multiple gestations, each amnioticsac is sampled separately. Most operators instill1 to 2 mL of indigo carmine dye after withdrawingthe amniotic fluid sample from the first and eachsubsequent fetus as an additional measure to becertain that each fetus is sampled separately. If,during the second needle insertion into the sacof the other fetus in twins, the fluid obtained isblue, then the first fetus has been sampled twiceand reinsertion is necessary to sample the second.Some operators have advocated a single-needleinsertion technique, with intentional penetrationof the intertwin membrane to sample the secondfetus. Neither approach has been subjected torigorous study, although we prefer to sampleeach sac separately.

Complications

The greatest concern for most couples decidingwhether to undergo amniocentesis is the likelihoodof fetal loss. The maternal age threshold andscreening cutoff of age 35 years, or 1 in 270 riskof Down syndrome, has been retained in partbecause it is thought to reflect a balance betweenthe likelihood of abnormal findings and the risk of aprocedure-related loss. Historically, most centershave counseled their couples that the risk ofmiscarriage after amniocentesis is up to 0.5%(unless they have center-specific data on whichto base their counseling).

The first reports describing the rate ofamniocentesis-related pregnancy loss came frombefore the era of high-resolution real-time ultra-sound that could be used for guidance. In a UnitedStates study of 2000 women, fetal loss betweenthe time of the procedure and delivery occurredin 3.5% of women compared with 3.2% ofcontrols, a nonsignificant difference.4 A similarrate of 3.2% without a control group was reportedin a Canadian study.5 Subsequently, a Danish


study published in 1986 showed amiscarriage rateof 1.7% in the amniocentesis group comparedwith 0.7% in controls (P<.01), in a highly selectedpopulation.6 More recent reports have questionedthe association between amniocentesis and anincreased risk of pregnancy loss. In a study fromThailand published in 1998 involving 2256 womenundergoing amniocentesis with matched controls,no difference was found in the rate of spontaneousabortion after amniocentesis (1.8%) versus con-trols (1.4%), (P>.05). No other complicationseemed to be increased in the amniocentesisgroup.7 A more recent Thai study8 showed similarresults in a cohort of 5051 pregnant womenreferred for prenatal counseling and followedprospectively, with subsequent comparison ofthose who underwent amniocentesis with thosewho declined. There were no differences in fetalloss rates before 24 weeks, before 28 weeks, orin the rate of preterm delivery. Towner andcolleagues9 reported similar results in a cohort ofmore than 32,000 women from the California StateMaternal Serum Screening Program, with amiscarriage rate that was no different betweenwomen who chose amniocentesis and those whodid not. In addition, Odibo and colleagues10

reported on a 16-year experience at a singlecenter encompassing more than 11,000 womenundergoing amniocentesis and nearly 40,000women in the same gestational period who didnot undergo a diagnostic procedure. A significantdifference between this study and the othersmentioned earlier is that the control group repre-sented an unselected population of women havingultrasound at the same gestational age as thoseundergoing amniocentesis, rather than age-matched controls who were referred for, but didnot undergo, amniocentesis. The controls had alower mean maternal age and lower proportionof women more than 35 years of age, as well asother risk factors such as prior family history ofgenetic disorders. In this study, the fetal loss ratebefore 24 weeks’ gestation was 0.97% for theamniocentesis group and 0.84% for the groupnot undergoing a procedure, suggesting that thefetal loss rate before 24 weeks that was attribut-able to amniocentesis was 0.13%, or 1 in 769procedures. This result represents a substantiallylower risk than that typically used for prenataldiagnosis counseling.An analysis of these reports suggests that when

performed at the traditional gestational range of16 to 20 weeks, the risk of miscarriage associatedwith amniocentesis is 0.5% or less, allowingwomen to make an informed decision under-standing the risk of the procedure and weighingit against her estimated risk for diagnosis of a fetal

abnormality, as well as her desire to receive diag-nostic information before the birth of her child.Less clear is the miscarriage rate attributable to

amniocentesis in a twin gestation. In a recentlypublished study from Greece, the investigatorscompared the outcomes of 120 women under-going amniocentesis in a twin gestation with6150 undergoing amniocentesis in a singletongestation, with advanced maternal age being themost common indication for the procedure. Inthis study, the women with twin pregnancieswere statistically of lower parity and more likelyto have become pregnant as a result of in vitrofertilization. They reported a miscarriage rate of0.24% for singletons and 0% in twins with miscar-riage defined as pregnancy loss before 24 weeks;this difference was not statistically significant.11

Conversely, Cahill and colleagues12 reported theoutcomes of twin pregnancies undergoing andnot undergoing amniocentesis at WashingtonUniversity, and showed a 3.2% rate of pregnancyloss in those having an amniocentesis comparedwith 1.4% in those who did not; they concludedthat the attributable risk of fetal loss in twin preg-nancies undergoing amniocentesis was 1.8%.This study has the same potential limitations asthe Odibo and colleagues10 study of amniocen-tesis in singletons, including younger maternalage and mean gravidity in women not having anamniocentesis.12 Given these reports, it may beadvisable to counsel women undergoing amnio-centesis with twins of the potential of a higherrate of pregnancy loss than for amniocentesisin singletons. Counseling women with twins ismade more complicated by the potential fordiscordant karyotypic results in dizygotic twinsand the challenging management decisions in atwin pregnancy with 1 normal and 1 abnormalfetus.Early gestational age at amniocentesis has been

consistently implicated as a risk factor for anincreased rate of pregnancy loss. In severalstudies in which amniocentesis was performedas early as 11 to 13 weeks’ gestation as an alterna-tive to a chorionic villus sampling, the miscarriagerate was reported to be as much as 4 times highercompared with women undergoing amniocentesisin the traditional gestational age range of 16 to20 weeks.13–15 Likewise, the occurrence of bloodyamniotic fluid is increased significantly. Earlyamniocentesis is mentioned here only for perspec-tive, in that it is rarely offered clinically, particularlywith the availability of more data regarding thesafety of chorionic villus sampling (discussedlater).Other complications of amniocentesis are less

common and include amniotic fluid leakage and


vaginal bleeding. In most cases, these are tran-sient events that cease spontaneously. The occur-rence of amniotic fluid leakage that results inoligohydramnios is of greater concern becausethe presence of a sufficient quantity of amnioticfluid is essential for fetal lung development.Several investigators have proposed techniquesfor ultrasound-guided intra-amniotic instillationof thrombogenic substrate to serve as a patchover the site of apparent rupture, with variablesuccess.16 Epidemiologic data suggest the possi-bility of a mild increase in the risk of congenitaltalipes equinovarus (club foot) in women under-going amniocentesis compared with those whodo not,17 as well as for placental abruption.18

In summary, women undergoing second-trimester amniocentesis may be informed of alow risk of complications, with a miscarriage risknot exceeding 0.5% for singleton pregnancies,but potentially higher for twin gestations. Asalways, the most important considerations inassuring the safest possible execution of theprocedure include adherence to appropriate indi-cations for the procedure, the presence of a skilledoperator, and the use of ultrasound guidance.

CHORIONIC VILLUS SAMPLING

Chorionic villus sampling (CVS) is an ultrasound-guided biopsy of the chorion frondosum of thedeveloping placenta. Indications for CVS are listedin Box 2. Retrieval of chorionic villi allows isolationof cells from the rapidly dividing and highly cellularcytotrophoblast for karyotype analysis. CVS ismost commonly performed between 10 and 13weeks’ gestation by the transcervical route. Trans-abdominal procedures can be performed at any

Box 2Indications for CVS

Advanced maternal age

Abnormal first-trimester screening

Family history of genetic disorder

� Prior child with fetal aneuploidy or othergenetic disorder

Other family history of a disorder for whichprenatal diagnosis by DNA analysis is available

Contraindications:

Maternal red cell alloimmunization

Possibly maternal chronic viral infection(human immunodeficiency virus, hepatitis)

Active infection at site of needle or catheterplacement

gestational age after 10 weeks and provide analternative means of obtaining a karyotype in preg-nancies affected by severe oligohydramnios.Ultrasound guidance for CVS is essential for local-ization of the placenta and assurance of aspirationof an adequate tissue sample. Most often, theapproach to CVS is dictated by operator prefer-ence and experience along with patient choice;however, there are certain circumstances in whichonly 1 approach is appropriate, and thereforeoperators should remain proficient in both tech-niques. A posterior placenta is typically onlyaccessible by transcervical CVS unless the uterusis also sharply anteverted. A fundal placenta istypically only accessible transabdominally. Trans-abdominal sampling is more likely to be used inmultifetal gestations to avoid contamination ofone cell culture with another, which could occuras a result of sampling multiple placentas throughthe cervix.

Technique

Both approaches are preceded by a completetransabdominal ultrasound examination includingplacental localization, assessment of uterine posi-tion, and a basic assessment of fetal well-beingincluding gestational age. Neither should beattempted if infection is present on the maternalabdomen in the area of needle insertion, or in thecervix in the case of a transcervical approach.The woman maintains a full bladder until theprocedure is completed; the bladder provides anessential sonographic window for adequateimaging during the procedure. For transcervicalprocedures, the woman assumes a dorsallithotomy position, the external genitalia may beprepared with povidone-iodine solution, and avaginal speculum is inserted. The cervix andvaginal walls are prepared with additionalpovidone-iodine and the operator changes glovesbefore starting the procedure. The transcervicalCVS catheter is a flexible polyethylene catheterover a malleable aluminum obturator with a blunttip. The operator shapes the catheter to assumethe curvature of the cervical canal and uterusand guides the catheter tip into the placenta. Thesonographer follows advancement of the tip pastthe internal cervical os until it is placed intoa long axis of the placenta, ideally about midwaythrough the thickness (Fig. 2). The obturator isremoved, and a 10-mL syringe containing 5 mLof nutrient medium is attached. Negative pressureis applied while the catheter is withdrawn. Visualinspection of the sample typically allows con-firmation of the presence of an adequate sampleof chorionic villi (typically 20–30 mg). In some

Fig. 2. CVS catheter in place in a posterior placenta.


cases, examination under a dissecting micro-scope is necessary to assure an adequate samplesize before the procedure is completed. A secondsample may be performed if the first sample sizeis small. It is important that only chorionic villi(representing fetal tissue) are cultured rather thandecidual cells (maternal tissue), because theycan occasionally be aspirated.Transabdominal CVS is preceded by sterile

abdominal preparation with either povidone-iodine or chlorhexidine-alcohol. A coplanar needleinsertion technique similar to a transabdominalamniocentesis is used, and a 19-gauge or 20-gauge needle is used. Once the needle has beenplaced into a long axis of the placenta, many oper-ators use a pistol grip for the syringe, and villi arethen aspirated into the syringe similarly to thetranscervical technique. A to-and-fro motion ofthe needle tip within the placenta is made severaltimes to ensure removal of an adequate numberof villi because the needle diameter is smallerthan the transcervical CVS catheter. Again, thesample is inspected for adequacy.After either approach, fetal heart tones are veri-

fied and Rh immune globulin administered if thewoman is Rh negative. She is instructed to avoidstrenuous activities for 24 to 48 hours and to alerther provider if she experiences vaginal bleeding inexcess of spotting, watery discharge, severecramping, lower abdominal tenderness, or fever.Results from direct analysis of mitotic figures incytotrophoblast cells may be available in as soonas 2 days, with a full culture leading to results in 8to 14 days. The time to achieve results is, in part,related to the size of the sample (milligrams of villi).

Complications

When considering complications of CVS, it isimportant to carefully review the reports of

procedure-related loss rates and to put thoseinto the context of the gestational age at whichchorionic villus sampling is performed. Becausethis procedure is typically performed at 10 to 13weeks of gestation, the background rate of preg-nancy loss (the rate of pregnancy loss irrespectiveof whether an invasive procedure is performed) ishigher compared with amniocentesis at 15 to 18weeks. This difference is borne out in most reportscomparing the loss rates of these 2 procedures. Itis important to provide reasonable estimations ofthe risk of loss to patients because they makedecisions about which procedure they wish topursue based in part on the estimated likelihoodof pregnancy loss related to the procedure, aswell as their desire to have results as early in thepregnancy as possible. The typical loss ratedescribed for CVS is 0.5% to 1%, or 1 in 200 to1 in 100 procedures.19

Several recent publications have reported onthe rates of pregnancy loss for CVS comparedwith amniocentesis. Mujezinovic and Alfirevic20

published a systematic review of 16 studies oftransabdominal CVS and 29 studies of amniocen-tesis. Each of the included studies reported on theoutcomes of at least 100 procedures. They re-ported on pregnancy loss rates within 14 days ofthe procedure by 24 weeks’ gestation, and totalpregnancy loss. The rates for amniocentesiswere 0.6%, 0.9%, and 1.9% respectively. Incomparison, the rates for CVS were 0.7%, 1.3%,and 2%. These differences were not statisticallysignificant.20 The investigators commented thatthe major limitation of the studies included inthis systematic review was the lack of properlyselected controls. Caughey and colleagues21 re-ported in 2006 on the comparative experienceof nearly 10,000 CVS procedures and 30,000amniocentesis procedures at a single center. Therate of pregnancy loss for both proceduresdecreased with time and the difference in lossrate between procedures became nonsignificantafter 1993.More recently, Tabor and colleagues2 in 2009

reported a national registry study over a 10-yearperiod in Denmark including 32,852 amniocentesisand 31, 355 transabdominal CVS procedures. Theloss rates were 1.4% and 1.9% respectivelywhen a procedure-related loss was defined as amiscarriage or intrauterine fetal death before 24weeks’ gestation.2 These experienced investiga-tors concluded that the difference in loss ratebetween the procedures was possibly attributableto the gestational age difference at the time theprocedures were performed. They also observedthat the loss rates were lower in centers perform-ing a higher volume of procedures during the

Box 3Indications for fetal blood sampling

Fetal karyotyping (when a rapid karyotype isneeded or when a karyotype cannot beobtained by amniocentesis caused byoligohydramnios)

Fetus at risk for genetic disorder for whichprenatal diagnosis by DNA analysis is available

Fetus at risk for anemia or thrombocytopenia

Suspected or known fetal infection (parvovirus,cytomegalovirus, toxoplasma)

Determination of fetal acid-base status (uncom-monly performed)


study period. In addition, Odibo and colleagues22

reported a cohort of 5243 women undergoingCVS at a single center over 16 years, andcompared them with a control group of womenhaving an ultrasound showing a viable fetus atthe same gestational age. The control group wasnot comparable with the CVS group in that thecontrol women were younger, had their ultrasoundat a later gestational age, were more likely to bewhite, and had experienced fewer prior miscar-riages. They did not detect a difference in the fetalloss rate between these groups.

There have been few studies addressing thedifference in loss rate attributable to transabdomi-nal versus transcervical CVS. Jackson andcolleagues23 reported the results of a trial of3999 women randomly assigned to undergo trans-abdominal or transcervical CVS. The mean gesta-tional age at the time of procedure in both groupswas 10.9 weeks, and 80% of procedures in bothgroups were performed at 10 or 11 weeks. Nodifferences were seen in loss rate or live birthrate by route. In another report, Silver andcolleagues24 described a series of more than1000 CVS procedures performed between 9 and12 weeks’ gestation, and reported a loss ratefor transcervical CVS of 5.2% compared with2.9% for transabdominal CVS. This apparentdifference in loss rate did not reach statisticalsignificance.

No discussion of CVS is complete withoutmention of the controversy surrounding limbreduction defects. In 1991, Firth and colleagues25

reported that, among 289 women undergoing CVSat less than 66 days’ gestation (approximately 9.5weeks), there were 5 cases of severe limb abnor-malities and oromandibular-limb hypogenesissyndrome. Although this finding had not been re-ported by other observers, approximately 1 yearlater, Burton and colleagues26 reported 4 casesof minor limb reduction defects from the first 500CVS procedures performed in a newly establishedCVS program. The CVS procedures in this serieswere all performed after 10 weeks’ gestation. Inthis cohort, the CVS-related pregnancy loss ratewas 8% compared with the 1% to 2% reportedby most other centers. There was a significantsubsequent reduction in patient interest in CVSbecause of concerns about limb reduction abnor-malities. However, large population-based regis-tries have failed to substantiate concerns aboutlimb reduction abnormalities for CVS performedafter 10 weeks’ gestation. The World Health Orga-nization registry of 138,996 prospectively ascer-tained cases of CVS included 77 fetuses withlimb reduction defects. This finding correspondedwith an incidence of limb reduction defects in

CVS-exposed pregnancies of 5.2 to 5.7 per10,000, which compared favorably with the back-ground rate of limb reduction abnormalities in thepopulation of 4.8 to 5.97 per 10,000. The authorsof these reports concluded that CVS after 10weeks’ gestation was not a risk factor for fetallimb reduction defects.27,28

The final complication attributable to CVS is thatof either not obtaining results or obtaining non-diagnostic results. Culture failure is reported tooccur in up to 1% of CVS samplings and the likeli-hood of culture seems to be inversely related tothe sample size. In addition, mosaicism (the pres-ence of 2 distinct cell lines in the cell culture,usually in substantially different proportions) isfound in 1.3% of CVS cases.29 Although mosai-cism is nearly always confined to the placenta,amniocentesis or other diagnostic techniques arenecessary to exclude the unlikely possibility thatthe mosaicism also affects fetal tissue ratherthan simply the placenta.

FETAL BLOOD SAMPLING

CVS is exclusively a diagnostic procedure, andnearly all amniocentesis procedures are per-formed for diagnostic purposes as well. Fetalblood sampling serves as a diagnostic tool butalso provides access for fetal therapeutic interven-tions by transfusion of blood products and admin-istration of pharmacologic agents and holdspromise for other future interventions such asgene therapy. Indications for fetal blood samplingare listed in Box 3. Because of the technicallychallenging nature of fetal blood sampling, aswell as the small number of indications, it is essen-tial that this procedure be performed by individualswith current knowledge in the techniques andcontraindications.


Technique

Fetal blood sampling can be performed as early as20 to 22 weeks of gestation (potentially earlier inskilled hands), and as late as term; the uppergestational age limit is often determined whena threshold is reached at which the risk of a compli-cation of the procedure is thought to exceed therisks to the fetus of preterm delivery, often approx-imately 34 weeks. When this procedure is beingperformed before fetal viability, the ambulatorysetting is appropriate; however, when fetal bloodsampling (with or without transfusion) is to be per-formed at a gestational age when the fetus ispotentially viable ex utero, and when interventionby emergency cesarean would be considered inthe event of a complication of the procedure,then it is prudent to do so in an inpatient setting,most frequently in the labor and delivery unit wherean emergency cesarean is readily available.Before performing the procedure, a detailed

ultrasound examination is performed with specialattention to the location of the placenta andthe placental cord insertion site. After obtaininginformed consent and thoroughly preparing theabdomen with either chlorhexidine-alcohol orpovidone-iodine, the procedure is initiated. Withthe transabdominal ultrasound transducer in asterile sleeve, an appropriate needle insertionsite is selected. The most common access pointsfor fetal blood sampling and or transfusion are theplacental insertion of the umbilical cord, the intra-hepatic portion of the fetal umbilical vein, or a freeloop of umbilical cord. The placental insertion siteis particularly favored with an anterior or fundalplacenta and is also a good choice because ofits relative stability (Fig. 3). However, traversingthe placenta usually results in a small amount offetomaternal transfusion and, if the indication forthe fetal blood sampling is alloimmunization, thismay expose the mother to additional fetal blood

Fig. 3. Placental cord insertion to anterior placenta inadvance of fetal blood sampling.

and accelerate the immune response. A free loopof umbilical cord is the least stable option, becausefetal movements may dislodge the needle.Local anesthetic is administered at the needle

insertion site and a 22-gauge spinal needle isinserted under continuous ultrasound guidance.The required skill level of the sonographerproviding guidance for fetal blood sampling issubstantially higher than that required for amnio-centesis. The target for the needle insertion issmaller and therefore the margin of error isnarrower. A suitable amniotic fluid pocket foramniocentesis is generally 2 � 2 cm or larger,making it easier for the sonographer to adjust thetransducer to identify the needle. In the case offetal blood sampling, the sonographer mustmaintain continuous visualization of the umbilicalvein at the proposed insertion site and assist theoperator in redirection of the needle until it reachesthe target. Confirmation of umbilical vein punctureis obtained by removing the stylet from the needle,attaching a heparinized 1-mL syringe, and obtain-ing blood return (Fig. 4). At the conclusion of theprocedure, the stylet is replaced, the needle iswithdrawn, and fetal heart motion is confirmed.The puncture site of the umbilical vein is observedto ensure that the bleeding stops. This usuallyoccurs in less than 30 seconds. For nonimmunizedRh-negative patients, Rh immune globulin isadministered and, if the fetus is at a potentiallyviable gestational age, fetal monitoring is initiatedfor several hours to assure fetal well-being afterthe procedure.Most operators performing fetal blood transfu-

sion perform the transfusion at the same settingas the diagnostic fetal blood sampling, ratherthan performing 1 procedure to diagnose fetalanemia and a second to treat the anemia withtransfusion. This method represents a logicalrisk-reduction strategy because each separateprocedure has a procedure-related loss rate of

Fig. 4. Needle in lumen of umbilical vein during fetalblood sampling.

Fig. 5. Needle placement in fetal chest for diagnosticfetal thoracentesis.


approximately 1%. However, this requires a coor-dinated effort between the perinatal team and theblood bank to have appropriately cross-matched,CMV-negative, irradiated, highly packed (hemato-crit 75%) red blood cells available at the initiationof the procedure, and to have precalculated theappropriate volume to be transfused based onthe sonographic estimated fetal weight, theestimated fetal placental blood volume, and hypo-thetical starting hematocrit. Fetal neuromuscularblockade with pancuronium or vecuronium maybe used for transfusion procedures because oftheir increased duration, but may not be necessaryfor diagnostic-only fetal blood sampling. Neuro-muscular blockade may also not be necessary ifthe needle insertion does not pass throughthe amniotic sac. When fetal blood samplingand/or transfusion is being performed forplatelet alloimmunization, it is critically importantto have cross-matched platelets available.Streaming of blood from the needle insertion siteafter fetal blood sampling is common, occurringin 40% to 50% of cases and typically lasting lessthan 60 seconds. However, in a fetus with severealloimmune thrombocytopenia, the normal hemo-static mechanisms are impaired and blood lossfrom the needle puncture site can be substantialunless platelets are transfused.

Complications

The pregnancy loss rate for fetal blood samplinghas generally been reported to range from 1% to3%.30,31 The likelihood for loss is related to theexperience of the operator, the indications for theprocedure, the technique used, and the gestationalage. Because of this high loss rate, many operatorsplace an upper gestational age limit at which theywould no longer perform fetal blood samplingbut would rather move to delivery in a fetus withan indication for fetal blood sampling. This gesta-tional age is often set at approximately 34 weeks,when the risk to the fetus of complications orpreterm may be lower than the risk of complica-tions from the procedure. Ghidini and colleagues32

reported a fetal loss rate of 2.7% in low-risk fetusesundergoing fetal blood sampling, defining low riskas those fetuses in which no abnormality wasfound at the time of the procedure. Among thehigh-risk fetuses, which included those withchromosomal abnormalities, severe fetal growthrestriction, fetal infections, and fetal hydrops, theloss rate was substantially higher at 9.4%.

OTHER ULTRASOUND-GUIDED PROCEDURES

Early success of fetal blood transfusion as atherapeutic intervention has led to the promise of

other interventions to improve pregnancy outcomeand avoid preterm delivery. A wide variety ofultrasound-guided fetal interventions have beenreported with variable success. For example, fetalpleural effusions can be drained by ultrasound-guided needle placement into the fetal thorax,allowing expansion of the lung (Fig. 5). Althoughhypothetically this reduces the likelihood ofpulmonary hypoplasia, there is a lack of adequateprospective studies to support this contention. Ina similar manner, fetal ascites could be aspiratedfor both diagnostic and therapeutic purposes.Fetal skin and muscle biopsy have been describedfor the diagnosis of specific disorders, sucha congenital ichthyosis and muscular dystrophy.

Perhaps the most well established ultrasound-guided fetal therapeutic procedure after fetalblood transfusion is vesicoamniotic shunting forfetuses with bladder outlet obstruction (posteriorurethral valves, urethral atresia). A fetus withbladder outlet obstruction may be diagnosed earlyin pregnancy by the findings of a massively dis-tended urinary bladder and decreased or absentamniotic fluid. Although the urinary obstructionultimately leads to kidney damage, most untreatedfetuses with bladder outlet obstruction die frompulmonary hypoplasia caused by the absence ofamniotic fluid to facilitate normal lung develop-ment. The purpose of vesicoamniotic shunting isto bypass the obstructed urethra and allow fetalurine to drain through a shunt that traverses thebladder wall and the anterior abdominal wall intothe amniotic sac to restore normal amniotic fluidvolume.33 Before shunting, fetal urine is aspiratedunder ultrasound guidance for analysis of urineelectrolytes to assess prognosis for long-termrenal function. Vesicoamniotic shunting has metwith fairly consistent success and is generallyaccepted in fetal medicine as an appropriate ther-apeutic intervention for fetuses with bladder outlet


obstruction, although it may ultimately be replacedby laparoscopically guided, minimally invasivelaser ablation of the posterior urethral valves. Adiscussion of that procedure as well as other diag-nostic or therapeutic techniques is beyond thescope of this article.

REFERENCES

1. Wapner RJ, Jenkins TM,KhalekN. Prenatal diagnosis

of congenital disorders. In: Creasy RK, Resnick R,

Iams JD, et al, editors. Creasy & Resnick’s maternal-

fetal medicine. 6th edition. Philadelphia: Saunders

Elsevier; 2009. p. 222.

2. Tabor A, Vestergaard CH, Lidegaard O. Fetal loss

rate after chorionic villus sampling and amniocen-

tesis: an 11-year national registry study. Ultrasound

Obstet Gynecol 2009;34:19–24.

3. Nakata N, Wang Y, Bhatt S. Trends in prenatal

screening and diagnostic testing among women

referred for advanced maternal age. Prenat Diagn

2010;30:198–206.

4. National Institute of Child Health and Development

National Registry for Amniocentesis Study Group.

Mid-trimester amniocentesis for prenatal diagnosis:

safety and accuracy. JAMA 1976;236:1471.

5. Simpson LE, Dallaire L, Miller JR, et al. Prenatal

diagnosis of genetic disease in Canada: report of

a collaborative study. Can Med Assoc J 1976;15:

739–46.

6. Tabor A, Philip J, Madsen M, et al. Randomized

controlled trial of genetic amniocentesis in 4606

low-risk women. Lancet 1986;1:1287–93.

7. Tongsong T, Wanapirak C, Sirivatanapa P, et al.

Amniocentesis-related fetal loss: a cohort study. Ob-

stet Gynecol 1998;92:64–7.

8. Pitukkijronnakorn S, Promsonthi P, Panburana P,

et al. Fetal loss associated with second trimester

amniocentesis. Arch Gynecol Obstet 2010;284(4):

793–7.

9. Towner D, Currier RJ, Lorey FW, et al. Miscarriage

risk from amniocentesis performed for abnormal

maternal serum screening. Am J Obstet Gynecol

2007;196:608.e1–5.

10. Odibo AO, Gray DL, Dicke JM, et al. Revisiting the

fetal loss rate after second-trimester genetic amnio-

centesis. A single center’s 16 year experience. Ob-

stet Gynecol 2008;111:589–95.

11. Kalogiannidis I, Petousis S, Prapa S, et al. Amnio-

centesis-related adverse outcomes in diamniotic

twins: is there a difference compared to singleton

pregnancies? Eur J Obstet Gynecol Reprod Biol

2011;155:23–6.

12. Cahill AG, Macones GA, Stamilio DM, et al. Preg-

nancy loss rate after mid-trimester amniocentesis

in twin pregnancies. Am J Obstet Gynecol 2009;

200:257, e1–6.

13. Brumfield CG, Lin S, Conner W, et al. Pregnancy

outcome following genetic amniocentesis at 11-14

versus 16-19 weeks’ gestation. Obstet Gynecol

1996;88:114–8.

14. Saltvedt S, Almstrom H. Fetal loss rate after second

trimester amniocentesis at different gestational age.

Acta Obstet Gynecol Scand 1999;78:10–4.

15. The Canadian Early and Mid-Trimester Amniocen-

tesis Trial (CEMAT) Group. Randomized trial to

assess safety and fetal outcome of early and mid-

trimester amniocentesis. Lancet 1998;351:242.

16. Deprest J, Emonds MP, Richter J, et al. Amniopatch

for iatrogenic rupture of the fetal membranes. Prenat

Diagn 2011;31:661–6.

17. Cardy AH, Torrance N, Clark D, et al. Amniocentesis

in the second trimester and congenital talipes equi-

novarus in the offspring: a population-based record

linkage study in Scotland. Prenat Diagn 2009;29:

613–9.

18. Minna T, Mika G, Tiina L, et al. Risk for placental

abruption following amniocentesis and chorionic

villus sampling. Prenat Diagn 2011;31:410–2.

19. Centers for Disease Control. Chorionic villus

sampling and amniocentesis: recommendations for

prenatal counseling. MMWR Recomm Rep 1995;

44(Suppl):1–12.

20. Mujezinovic F, Alfirevic Z. Procedure-related compli-

cations of amniocentesis and chorionic callous

sampling. A systematic review. Obstet Gynecol

2007;110:687–94.

21. Caughey AB, Hopkins LM, Norton ME. Chorionic

villus sampling compared with amniocentesis and

the difference in the rate of pregnancy loss. Obstet

Gynecol 2006;108:612–6.

22. Odibo AO, Dicke JM, Gray DL, et al. Evaluating the

rate and risk factors for fetal loss after chorionic

villus sampling. Obstet Gynecol 2008;112:813–9.

23. Jackson LG, Zachary JM, Fowler SE, et al.

A randomized comparison of transcervical and

transabdominal chorionic-villus sampling. N Engl J

Med 1992;327:594–8.

24. Silver RK, MacGregor SN, Muhlbach LH, et al.

A comparison of pregnancy loss between transcer-

vical and transabdominal chorionic villus sampling.

Obstet Gynecol 1994;83:657–60.

25. Firth HV, Boyd PA, Chamberlain P, et al. Severe limb

abnormalities after chorion villus sampling at 56-66

days’ gestation. Lancet 1991;337:762–3.

26. Burton BK, Schulz CJ, Burd LI. Limb anomalies

associated with chorionic villus sampling. Obstet

Gynecol 1992;79:726–30.

27. Froster UG, Jackson L. Limb defects and chorionic

villus sampling: results from an international registry

1992-94. Lancet 1996;347:489–94.

28. Kuliev A, Jackson L, Froster U, et al. Chorionic villus

sampling safety. Report of World Health Organiza-

tion/EURO meeting in association with the Seventh


International Conference on Early Prenatal Diag-

nosis of Genetic Diseases. Tel Aviv, Israel, May 21,

1994. Am J Obstet Gynecol 1996;174:807–11.

29. Goldberg JD, Wohlferd MM. Incidence and outcome

of chromosomal mosaicism found at the time of

chorionic villus sampling. Am J Obstet Gynecol

1997;176:1349–53.

30. Buscaglia M, Ghisoni L, Bellotti M, et al. Percuta-

neous umbilical blood sampling: indication changes

and procedure loss rate in a nine years’ experience.

Prenat Diagn 1996;11:106–13.

31. Weiner CP, Okamura K. Diagnostic fetal blood

sampling-technique-related losses. Fetal Diagn Ther

1996;11:169–75.

32. Ghidini A, Sepulveda A, Lockwood CJ, et al. Compli-

cations of fetal blood sampling. Am J Obstet Gyne-

col 1993;168:1339–44.

33. Clark TJ, Martin WL, Divakaran TG, et al. Prenatal

bladder drainage in the management of fetal

lower urinary tract obstruction: a systematic review

and meta-analysis. Obstet Gynecol 2003;102:

367–82.

ultrasound-guided procedures in obstetrics

Documents