148Medicine and Health / Rhode Island

�The Smallest Patient: Foundations in Fetal Medicine

A CME Issue: Introduction

Until the latter part of the twentieth century, the fetus washidden from outside scrutiny and little could be pre-

dicted about gestation. Within decades, fetal diagnosis becamea reality, largely (but not solely) due to the advent of ultra-sonography. Thus, a new specialty was born, as normal andabnormal fetal development could be followed, and more andmore accurate diagnosis of fetal conditions could be made.Advance knowledge of fetal diseases led to the quest for fetaltreatment, which today ranges from counseling of future par-ents and alterations in mode or place of delivery to more orless invasive procedures on the fetus himself.

The first invasive fetal intervention was performed in 1963by Liley, in New Zealand. His intrauterine transfusion of afetus with rhesus isoimmunization is generally considered tomark the beginnings of fetal medicine. That same year, KarlissAdamsons performed the very first (although unsuccessful)open fetal operation, for the same condition. Dr. Adamsonswent on to become the first chairman of Obstetrics and Gy-necology in the Brown University Program in Medicine. Al-most forty years later, fetal medicine has come of age, andBrown Medical School is one of the first medical schools inthe country to have a Program in Fetal Medicine. This pro-gram, and the Multidisciplinary Antenatal Diagnosis AndManagement conferences (affectionately dubbed MADAM)on which it is based, emphasize how fetal medicine has evolvedin this relatively short time: whereas care of the pregnancy wastraditionally assumed by the obstetrician alone, managementof the fetus as a patient has become the responsibility of a largegroup of health professionals, including maternal-fetal medi-cine specialists, neonatologists, pediatric surgical and medical

François I. Luks, MD, PhD, Stephen R. Carr, MD, Lewis P. Rubin, MD

Rhesus(Rh) isoimmunization is thedevelopment of maternal IgG an-

tibodies against fetal Rh red blood cell(RBC) antigens. Transplacental passageof maternal Rh antibody (IgG anti-D)may result in hemolytic disease of thenewborn (HDN), a condition charac-terized by hemolysis, severe fetal ane-mia and generalized fetal edema(hydrops fetalis). Although other redcell antigens may elicit a maternal im-mune response resulting in HDN, theRh blood group system is the mostcommon and important.

�Michael P. Plevyak, MD, and Stephen R. Carr, MD

Rhesus Isoimmunization

INCIDENCE OF RH ISOIMMUNIZA-TION

The Rh system is comprised of 5antigens: D, C, c, E, e. The presence ofD confers Rh positivity while its absencesignifies Rh negativity. Approximately15% of whites are Rh negative comparedwith 6% of blacks and 1% of Asians.1

Rh isoimmunization occurs when Rh-positive fetal RBCs gain access to thematernal circulation and trigger antibodyproduction. Without prophylaxis, nearly17% of Rh-negative women become im-munized as a result of their first ABO-compatible, Rh-positive pregnancy.2 The

risk of immunization decreases to 1-2%with ABO-incompatibility.3 Spontane-ous fetomaternal hemorrhage (FMH) isthe most common cause of Rh isoim-munization. The rate of FMH increaseswith gestational age occurring in 6.7%,15.9% and 28.9% in the first, secondand third trimesters, respectively, and in50% of pregnancies at delivery.4 The in-cidence of Rh isoimmunization increaseswith the volume of FMH. Three percentof patients become immunized if the vol-ume of FMH is less than 0.1 ml; 14%and 22% become immunized with FMHvolumes of greater than 0.1 ml and 0.4

specialists, geneticists and genetic counselors, radiologists andperinatal pathologists.

The present issue on fetal medicine highlights thismultidisciplinary approach, as a wide array of fetal conditionsand their treatment are discussed. The articles reflect the exper-tise of these and other members of the Brown Program in FetalMedicine, and are published in parallel with a state-of-the-artconference (“The Smallest Patient: Foundations in Fetal Medi-cine”) organized on the Brown University campus. Together,these initiatives highlight the spirit of collaboration and thehigh level of expertise present in Rhode Island.

Francois I. Luks, MD, PhD, is an Associate Professor of Sur-gery and Pediatrics and co-director of the Program in Fetal Medi-cine, Brown Medical School.

Stephen R. Carr, MD, is an Associate Professor of Obstetricsand Gynecology and co-director of the Program in Fetal Medicine,Brown Medical School.

Lewis P. Rubin, MD, is an Associate Professor of Pediatrics andco-director of the Program in Fetal Medicine, Brown Medical School.

CORRESPONDENCE:François I. Luks, M.D., Ph.D.Division of Pediatric SurgeryHasbro Children’s Hospital2 Dudley Street, Suite 180Providence, RI 02905phone: (401) 421-1939fax: (401) 868-2319e-mail: Francois_Luks@brown.edu

149Vol. 84 No. 5 May 2001

Without treatment after birth, these new-borns become severely jaundiced and areat risk of developing kernicterus or bi-lirubin encephalopathy. Severe disease,characterized by hydrops fetalis, occursin 20-25% of affected fetuses. Nearly50% develop hydrops prior to 34 weeksof gestation.

ANTENATAL MANAGEMENT

The Rh-negative woman with a nega-tive antibody screen

A blood type and antibody screenmust be obtained at the first prenatal visitof every pregnant patient. Among Rh-negative mothers, the paternal ABOgroup and Rh status, including zygosityin the case of Rh positivity, should bedetermined. If the partner is Rh positive,then a maternal antibody screen is re-peated at 28 weeks and anti-D immuneglobulin is administered if the antibodyscreen is negative. Serial screening hasbeen recommended every 6 weeks from20 weeks until delivery by some.6 In or-der to prevent isoimmunization, Ameri-can College of Obstetricians andGynecologists (ACOG) recommendsthe administration of anti-D immuneglobulin to Rh-negative, unimmunizedwomen at 28 weeks of gestation andwithin 72 hours of delivery (of an Rh-positive baby) and other potentially sen-sitizing events (e.g. amniocentesis,threatened abortion, abruptio placentaand abdominal trauma).7 Postpartumanti-D immune globulin prophylaxis has

reduced the alloimmunization rate by90%.8

The Rh-negative woman with a posi-tive antibody screen

The severity of Rh hemolytic dis-ease may be assessed in the immunizedpatient by several different means whichinclude: past obstetric history, maternalantibody titer, amniotic fluid (AF) bi-lirubin quantification, ultrasonographyand percutaneous umbilical blood sam-pling. A past history of hydrops fetalis,fetal demise, intrauterine transfusion orneonatal exchange transfusion impliesthat the present pregnancy is or will beseverely affected because Rh disease ef-fects on the fetus generally recur orpresent progressively earlier or more se-verely with each successive Rh positivepregnancy. Obstetric history does notpertain to the first sensitized pregnancyin which the risk of hydrops is 8-10%.

The maternal antibody titer shouldbe evaluated at the first prenatal visit inRh-sensitized gravidae, at 16-18 weeksand then every 2-4 weeks.6,9 The fetus isnot at significant risk for severe diseaseas long as the titer remains < 1:16.9 Atiter of 1:16 requires further evaluationbecause the risk of hydrops increases withRh antibody titers. Once the critical ti-ter is achieved or hydrops occurs the util-ity of the antibody titer in managingsubsequent pregnancies is low.

Bilirubin, a product of hemolysis, isexcreted into the AF by the fetus. The rela-

Table 1. Outcome based on ∆ OD 450 Liley graph zone locationZone Disease Severity Fetal/Neonatal Status1 Mild or no hemolytic disease 10% probability neonatal exchange transfusion required2 Intermediate disease disease severity increased as the _ OD 450 readings approached zone 33 Severe hemolytic disease Severe anemia; hydrops fetalis; fetal death probably within 7 -10 days

Modified from Liley AW. Liquor amnii analysis in management of pregnancy complicated by Rhesus immunization. Am JObstet Gynecol 1961;82:1359-70.

ml, respectively.5 The incidence of FMHis 2.5% following amniocentesis, 5% af-ter spontaneous abortion and up to 20-25% after an induced abortion.6

PATHOGENESIS OF HEMOLYTIC

DISEASE OF THE NEWBORN

The primary maternal immune re-sponse to a challenge by fetal RBC anti-gens tends to develop slowly, ischaracterized by low avidity and is mostlyIgM, which does not cross the placenta.For these reasons and the fact that mostcases of isoimmunization occur at birth,the index pregnancy is usually unaffected.The secondary antibody response is usu-ally IgG which readily crosses the pla-centa. Subsequent maternal exposure toRh-positive blood is characterized byrapid development and increased avid-ity of the IgG antibodies. Once in thefetal circulation, maternal anti-D anti-bodies bind fetal RBC surface antigensand are sequestered and destroyed in thefetal spleen. This leads to fetal anemia,which stimulates erythropoietin produc-tion. When RBC destruction exceedsmarrow production, extramedullary sites,such as the liver and spleen, are recruitedfor RBC production. Hepatomegaly andhepatocellular damage result in portal hy-pertension and hypoalbuminemia whichif sufficiently severe, may result in fetalascites, fetal pericardial and pleural effu-sions, and fetal skin edema or placentalthickening.6

SEVERITY OF HEMOLYTIC DISEASE

OF THE NEWBORN

Bowman defined 3 categories basedon severity of disease.6 Approximately50% of affected newborns have mild dis-ease and do not require treatment. Innewborns with mild disease, hemoglo-bin levels remain above 11 g/dl and se-rum indirect bilirubin levels do notexceed 20 mg/dl. Moderate disease isnoted in 25-30% of affected neonates.

Table 2. Management based on ∆ OD 450Liley graph zone location

Zone Timing of subsequent intervention

1 Repeat amniocentesis in 3-4 weeks

2 Repeat amniocentesis in 1-2 weeks; increased frequency as valuesrise; fetal blood sampling if value falls in upper 65% of modified zone 2

3 Immediate fetal blood sampling

Modified from Grannum P, Copel JA. Prevention of Rh isoimmunization andtreatment of the compromised fetus. Semin Perinatol 1988;12:324-35.

150Medicine and Health / Rhode Island

tive concentration of bilirubin in the AFcan be quantified by spectrophotometry.Optical density (OD) readings, obtainedover the 700 to 350 nm wavelength range,are plotted on semilog paper with wave-length on the horizontal axis and OD onthe vertical axis. The readings at 550 nmand 365 nm are connected by a line. Thedeviation from linearity of the OD read-ing at 450 nm (the OD 450 value) is themeasurement from the actual reading at450 nm to where the connecting line in-tersects 450 nm (Figure 1). The OD 450value is then plotted on a graph with ges-tational age as the linear coordinate (Fig-ure 2). Liley10 examined amniotic fluidwithin a week of birth from 101 Rh im-munized pregnancies at 28 weeks of ges-tation. The OD 450 was correlated withcord hemoglobin and the degree of HDN.Liley described three zones (Table 1 andFigure 2).

The timing of the first amniocente-sis is based on the maternal antibody ti-ter or the past obstetric history. The initialamniocentesis for OD 450 evaluationmay occur as early as 18 weeks of gesta-tion if a prior pregnancy was complicatedby hydrops, stillbirth or neonatal ex-change transfusion. Subsequent amnio-centeses are planned based on the prior

OD 450 values. Table 2provides a general guidelinefor timing of interventions.

The predictive value ofOD 450 measurements forsevere disease obtained af-ter 27 weeks of gestation isapproximately 95%.11 AOD 450 value may not beas predictive of severe dis-ease when obtained prior to26 weeks. According toBowman, fetal blood sam-pling with direct fetal he-moglobin determination isrecommended for OD 450values in upper zone 2 or

zone 3 obtained prior to 26 weeks of ges-tation.5 The risks of amniocentesis in-clude: miscarriage, FMH, membranerupture, preterm labor and infection.Fetomaternal hemorrhage is minimizedby performing amniocentesis under ul-trasound guidance to avoid transplacen-tal access. According to the ACOG, themidtrimester “risk of abortion second-ary to amniocentesis is 1 in 200 or less.”12

Ultrasonography provides anoninvasive means of diagnosing fetal hy-drops and assessing fetal well being.Sonographic attributes claimed to have aclinical association with severe fetal anemiaprior to the onset of fetal hydrops include:increased umbilical vein diameter, placen-tal thickness, bowel wall visualization, peri-cardial effusion, enlarged fetal heart,enlarged liver and spleen, decreased headcircumference to abdominal circumferenceratio and increased amniotic fluid volume.However, none of these observations, aloneor in combination, are as accurate predict-ing hydrops as serial amniotic fluid OD450 measurements.13 More recently, fetalmiddle cerebral artery peak systolic veloci-ties (MCA-PSV) determined by Dopplerflow studies have been found to estimatethe degree of fetal anemia in Rh disease.Mari and colleagues utilized MCA-PSV to

prospectively evaluate 111 fetuses at risk foranemia. An elevated MCA-PSV was de-fined as a value 1.50 multiples of the me-dian (MoM) for gestational age. Thesensitivity of an increased MCA-PSV forthe prediction of moderate or severe ane-mia was 100% in the absence or presenceof hydrops accepting a false positive rate of12%.14 Moderate anemia was defined as ahemoglobin concentration from less than0.65 to 0.55 times the median for gesta-tional age and severe anemia as a hemoglo-bin less than 0.55 times the median. Thesedata suggest that MCA-PSV measure-ments, while non-invasive, provide an ac-curate means of determining the degree offetal anemia in pregnancies complicated byRh isoimmunization.

The most accurate means of deter-mining fetal hemoglobin concentrationrequires percutaneous umbilical bloodsampling (PUBS). During the PUBS pro-cedure, a needle is inserted into the lu-men of the umbilical vein underultrasound guidance. Once access is ob-tained, the fetal hemoglobin concentra-tion can be determined and packed redblood cells transfused if necessary. Poten-tial candidates for this procedure includepatients with OD 450 measurements inhigh zone 2 or zone 3, sonographic evi-dence of fetal hydrops or MCA-PSV ≥1.50 MoM. The risks associated withPUBS are listed in Table 3. The proce-dure-related pregnancy loss rate is approxi-mately 1-1.4%.15, 16

MANAGEMENT OF DELIVERY

Among fetuses with ∆ OD 450measurements at or below the middle ofzone 2 and normal sonography labor anddeliver at term may be anticipated. Al-ternatively, labor may be induced withdocumentation of fetal lung maturity anda favorable cervix. The management ofthe severely affected fetus depends on thegestational age at the time of diagnosis.If severe disease occurs after 34 weeks,

Table 3. Complications of PUBSComplications

Cord Cord Fetal Pretermhemorrhage hematoma bradycardia Infection Delivery

Rate 23-53% 17% 3-12% 1% 5-6%

Modified from GhidiniA, Sepulveda W, Lockwood CJ, Romero R. Complications of fetal blood sampling. Am J ObstetGynecol 1993;168:1339-44.

Figure 1.6

151Vol. 84 No. 5 May 2001

then delivery is indicated. Prior to 34 weeks, in utero red celltransfusion is recommended. Fetal blood transfusion mayoccur through the intraperitoneal (IP)or direct intravascular(IV) routes. Compared to the IP route, direct IV transfusionis preferred because of a significantly improved survival rate(76% vs 88%) and lower complication rate (3.5% vs 0.8%).5,17

Complications of in utero transfusion are listed in Table 4.Short and long-term neurologic and physical developmentamong those who have had in utero transfusions appears to becomparable to non-anemic, birth age-matched controls.6,18

Major CNS sequelae occur in 4%, of which 25% is secondaryto prematurity and 75% attributed to HDN.18

SUMMARY

Rh isoimmunization is a potentially preventable conditionthat occasionally is associated with significant perinatal morbidityor mortality. Disease severity may be assessed using the modalitiesdescribed above and frequently, invasive techniques are requiredto determine the risk of severe disease. Doppler flow studies ap-pear to offer accurate, noninvasive means of evaluating fetal risk,which may allow for a decrease in invasive diagnostic procedures.The Rh isoimmunized patient, managed by an experienced team,can anticipate a favorable pregnancy outcome.

REFERENCES1. Grannum P, Copel JA. Semin Perinatol 1988;12:324-35.2. Mollison PL, Engelfriet CP, Contreras M. Blood Transfusion in Clinical

Medicine, 9th edn. Oxford: Blackwell Scientific Publications, 1993.3. Woodrow JC. Rh isoimmunization and its prevention. Series Hematologia

Vol III. Copenhagen, Munksgaard, 1970.4. Cohen F, Zueler WW, Gustafson DC, Evans MM. Blood 1964;23:621-

46.5. Zipursky A, Israels LG. Can Med Assoc J 1967;97:1245-9.6. Bowman JM. Hemolytic disease (erythroblastosis fetalis). In: Creasy

RK, Resnik R, eds. Maternal-fetal medicine. 4th ed. Philadelphia: Saunders,1999:736-67.

7. ACOG Practice Bulletin: Prevention of Rh D Alloimmunization. No.4, 1999.

8. Freda VJ, Gorman JG, Pollack W, Bowe E. NEJM 1975;292:1014-6.9. ACOG Educational Bulletin: Management of isoimmunization in preg-

nancy. No. 227, 1996.10. Liley AW. Am J Obstet Gynecol 1961;82:1359-70.11. Bowman JM, Pollock JM. Pediatrics 1965;35:815-3512. ACOG Technical Bulletin. Antenatal Diagnosis of Genetic Disorders

(number 108). Washington, DC: 1987.13. Whitecar PW, Moise KJ. Obstet Gynecol Surv 2000;55:240-50.14. Mari G, Deter RL, Carpenter RJ, et al. NEJM 2000;342:9-14.15. Daffos F, Capella-Pavlovsky M, Forestier F. Am J Obstet Gynecol

1985;153:655-60.16. GhidiniA, Sepulveda W, Lockwood CJ, Romero R. Am J Obstet Gynecol

1993;168:1339-44.17. Harman CR, Bowman JM, Manning FA, et al. Am J Obstet Gynecol

1990;162:1053-19.18. Halitsky V. Sequelae in children who survived in utero fetal transfusion

only. In Tejani N, ed. Obstetrical Events and Developmental Sequelae. BocaRaton, FL: CRC press, 1990.

Michael P. Plevyak, MD, is a Clinical Instructor in Obstet-rics and Gynecology, Brown Medical School.

Stephen R. Carr, MD is an Associate Professor of Obstetricsand Gynecology and co-director of the Program in Fetal Medicine,Brown Medical School.

CORRESPONDENCE:Michael P. Plevyak, MDDivision of Maternal-Fetal MedicineWomen & Infants’ Hospital101 Dudley StreetProvidence, RI 02903phone: (401) 274-1122 x 2345fax: (401) 453-7622e-mail: mplevyak@wihri.org

Table 4. Complications of intrauterine transfusionMaternal Fetal Fetomaternal

Preterm membrane rupture Overtransfusion Transplacental hemorrhageInfection Exsanguination

Placental abruption Cord hematomaPreterm labor Bradycardia

Emergency cesarean section Fetal injury or demise

Modified from Bowman JM. Hemolytic disease (erythroblastosis fetalis). In: Creasy RK, Resnik R, eds. Maternal-fetalmedicine. 4th ed. Philadelphia: Saunders, 1999:736-67.

Figure 2. Modified from Liley ∆ OD450 reading zone boundariesbefore 24 weeks of gestation. From Bowman J. Rhesus hemolytic

disease. In Wald NJ ed. Antenatal Screening. 2nd ed. Oxford,Oxford University Press. By permission.

152Medicine and Health / Rhode Island

�Lewis P. Rubin, MD

Alpha-thalassemia Major: Antenatal Diagnosis and Management

The thalassemias are among the mostcommon single gene disorders in

humans. Their high frequency in specificpopulations appears to be a result of se-lection by resistance to malaria.1

Thalassemias result from inefficient syn-thesis of α-globin (α-thalassemia), β-globin (β-thalassemia) or, more rarely, β-and δ-globin chains (δβ-thalassemia).

Without medical intervention, thelethal form of α-thalassemia, homozy-gous α-thalassemia major or hemoglo-bin Bart’s disease, is not compatible withintrauterine survival through the thirdtrimester. This disorder is a frequent causeof perinatal deaths and the principal causeof fetal hydrops in South China, Thai-land, Vietnam, Laos, Cambodia, Malay-sia, the Philipines and other parts ofSoutheast Asia.2,3 During the past sev-eral decades, increased emigration fromAsian regions where α-thalassemia-1 isprevalent has heightened awareness ofthis disorder in the United States. An es-timated 5-6% of persons of SoutheastAsian origin living in the United Statescarry an α-thalassemia deletion, and car-riage rates in specific Asian ethnic groupsmay be much higher.4,5

DEVELOPMENTAL ASPECTS OF

HEMOGLOBIN SYNTHESIS: THE

THALASSEMIAS

The timing of onset and clinicalcourse of α-thalassemia major followsfrom the normal developmental pro-gramming of hemoglobin chain synthe-sis. Each tetrameric hemoglobinmolecule consists of two different globinchain pairs, each containing one hememolecule. Balanced expression of α-globin and non-α-globin chains is nec-essary for normal hemoglobin synthesisand red cell function. The duplicated α-globin genes lie in close proximity (~3.7kilobases [kb] apart) on chromosome16p13.3. The α-globin gene cluster spans30 kb and consists of one embryonic ζ-globin gene (ζ2), two α-globin genes (α2and α1) and four pseudogenes. Thus,each diploid cell contains four α-globingene copies. Deletion of two α-globin

chains can either occur on the same chro-mosome (cis type) or as two single genedeletions, one on each chromosome(trans type). Cis-type α-thalassemia trait(α-thalassemia-1, α0-thalassemia) is mostfrequent in individuals of East Asian de-scent (and in some Mediterranean popu-lations) and carries the risk for fetaldisease, while the trans-type deletions aremore common in individuals of Africandescent.

During the first 6 to 7 weeks ofembyrogenesis, zeta (ζ) chains are tran-siently expressed from the α-globin genecluster in preference to α-globin chains.Thereafter, one of the hemoglobin chaintypes always is an α-globin chain. (seeTable). During development, α-globinchains are incorporated into fetal hemo-globin (Hb) F (α2/γ2) and embryonicHb Gower 2 (α2/ε2). Normally, 12 to16 weeks after birth the postnatal globinchain switch takes place from δ to β re-sulting in a shift to the predominant adultHb, Hb A (α2/β2). This phenomenonof postnatal hemoglobin switching pre-cludes appearance of β-thalassemias be-fore or at the time of birth.

β-Thalassemia also is commonlycaused by point mutations of the β-globin gene whereas α-thalassemias usu-ally result from deletions of one or moreα-globin genes. The α-thalassemia silentcarrier state has the genotype of threenormal genes (-α/αα). α-Thalassemiatrait results from two normal genes in cis(--/αα) or trans orientation (-α/-α). InHb H disease, which is characterized bya microcytic hemolytic anemia, there isone normal gene (--/-α). No normal α-globin genes (--/--) cause homozygous α-thalassemia major, or Hb Bart’s hydropsfetalis. In α-thalassemia major, the ab-sence of α-globin chain production leads

to excess γ chain synthesis and forma-tion of the unstable and physiologicallyineffective Hb Bart’s (γ4).

HB BART’S HYDROPS FETALIS:CARRIER DIAGNOSIS AND SCOPE OF

THE PROBLEM

A diagnosis of α-thalassemia-1should be considered whenever a womanof East Asian, Southeast Asian or Filipinabackground has microcytosis (MCV <80fL, frequently <75 fL [nl 80-100]), whichsometimes is accompanied by erythro-cytosis and a mild hypochromic anemia.Other causes of anemia such as iron de-ficiency, chronic blood loss and β-thalas-semia may co-exist with α-thalassemiaand should be evaluated. Other labora-tory findings in the α-thalassemia-1 car-rier state are hypochromia (MCH <27pg [nl 27-34]), a normal Hb electro-phoresis (i.e., normal or low Hb A2 level),and presence of precipitated Hb H (β4)inclusions after supravital staining withthe redox reagent brilliant cresyl blue.Adult carriers of the dual α-globin dele-tion (--SEA, see below) have minuteamounts of persistent embryonic ζ-globin chains incorporated into Hb. De-tection by immunocytological stainingof peripheral blood smears with anti-hu-man ζ-globin antibodies has high posi-tive predictive value.6

When both prospective parents havemicrocytosis, identification of the α-globin cluster genotypes by DNA test-ing becomes important for geneticcounseling and prenatal diagnosis. If bothparents are heterozygous for the cis-typeα-globin gene deletion, each pregnancycarries a 25% risk for Bart’s hemoglobin-opathy. In Southeast Asia, the commonα-thalassemia mutation is a 20.5 kb de-letion of both α-globin genes, sparing the

153Vol. 84 No. 5 May 2001

embryonic ζ-globin gene (--SEA). Indi-viduals of Filipino or Thai ancestry maycarry more extensive α-globin gene clus-ter deletions (--SEA, --FIL) that also involveloss of the ζ-globin gene. In these in-stances, the absence of functional embry-onic hemoglobins in homozygous nullfetuses probably induces early pregnancyloss, before onset of the hydrops syn-drome would occur.

A second indication for genetic test-ing in a risk population is a previous ob-stetric history of either fetal losses orhydrops. Given the molecular heteroge-neity of α-thalassemia mutations, Asian-Americans may carry one or moredifferent mutations. Optimal genetictesting and counseling for individuals atrisk for α-thalassemia, therefore, requireknowledge of the frequencies of differ-ent mutations and the specificity of spe-cific testing methods to detect thesemutations. Although more than 20 typesof α-globin gene deletions have beendescribed, three mutations (--SEA, -α3.7,and --FIL) account for most mutant alle-les present in North Americans of mixedSoutheast Asian ancestry.5 DNA hybrid-ization probes have been designed to de-tect α-globin and γ-globin deletions andthe two large deletions (--THAI, --FIL) thatremove the entire ζ-α-globin gene com-plex. Genotypes are assigned by compar-ing the observed restriction fragmentlength patterns on genomic Southernblots to previously defined patterns of α-globin gene deletions. Polymerase chainreaction (PCR)-based diagnosis for ho-mozygotes has been limited by technicaland interpretive concerns stemming fromhigh guanine/cytosine (GC) content andextensive α-globin cluster sequence ho-mology. Recently, however, developmentof multiplex-PCR assays capable of de-tecting the common α-globin gene clus-ter deletions has resolved some of theseissues.

HB BART’S HYDROPS FETALIS:ANTENATAL DIAGNOSIS AND

MANAGEMENT

Early in pregnancy, production ofζ-globin and ε-globin chains results insynthesis of embryonic Hb Gower 1 (ζ2/ε2) and Hb Portland 1 (ζ2/γ2), whichpermit oxygen dissociation to the devel-oping fetus. Later in the first trimester,

when the fetus with α-thalassemia ma-jor switches from ζ- to α-globin chainexpression, Hb Bart’s (γ4) predominatesin the fetal circulation. Hb Bart’s has anabnormally high oxygen affinity and noBohr effect and cannot deliver oxygentaken up in the placenta to fetal tissues.The relatively unstable Hb Bart's mol-ecules also can precipitate and shortenred cell survival. Hypoxia and anemia (6-8 g Hb/dl at 12-13 weeks7) lead to ex-cess extramedullary erythropoiesis, highoutput cardiac failure, impaired hepaticsynthetic function, hepatic blood flowobstruction and, inevitably, fetal hydrops.Half of fetuses with Bart’s hemoglobin-opathy die between 23 and 28 weeks ges-tation, and live-born infants usually diesoon after birth, despite aggressive resus-citative efforts.8 Hb Bart's hydrops fetalisalso significantly increases maternal riskfor preeclampsia, polyhydramnios, pla-cental retention and postpartum bleed-ing, the last complication probablyrelated to placental hypertrophy.2,9

For these reasons, prenatal diagno-sis of Bart’s hemoglobinopathy in the firsttrimester is often undertaken to allowpregnancy termination. The fetal diag-nosis can be made by DNA analysis ofchorionic villi or amniocytes obtained bychorionic villus biopsy or amniocentesisor by Hb studies obtained bycordocentesis. This is the principal indi-cation for cordocentesis in SoutheastAsia.10 The large number of affected preg-nancies that present to perinatal centersin Hong Kong and Thailand also hasaided early sonographic diagnosis ofevolving Bart’s hydrops fetalis. In high-risk pregnancies at 10-14 weeks of gesta-tion, placental thickening11 and elevatedfetal cardiothoracic ratio12 appear to havegood positive predictive value.

Although Hb Bart’s disease is almostalways fatal in utero or shortly after birth,there are rare instances of survival intochildhood.13-16 These infants were all de-livered quite prematurely after exposureto the harmful consequences of Hb Bart’shydrops fetalis. Since intrauterine trans-fusion (IUT) can reverse the hydropscaused by fetal isoimmune hemolyticanemia, we reasoned that IUT might bea successful approach in this disorder aswell. Reversal of fetal hydrops, anemiaand hypoxia ought to improve fetal de-

velopment and permit delivery ofhealthier infants at >34 weeks of gesta-tion. The first intrauterine exchangetransfusions for salvaging a fetus with HbBart’s fetal hydrops were performed atWomen & Infants Hospital in 1994.8

This patient was a Filipina-Americanwoman whose previous pregnancy re-sulted in intrauterine demise of a hydro-pic fetus at 28 weeks. Pathologic andDNA analysis had confirmed α-thalas-semia major and Southern blotting re-vealed both parents to be α-thalassemiacarriers with the genotype, αα/α-SEA. Theparents received genetic counseling and,in the next pregnancy, the mother un-derwent amniocentesis at 16 weeks,which again confirmed fetal Hb Bart’sdisease. After considering their options,this couple elected IUT. During a courseof three exchange IUTs, fetal ascites re-solved and fetal Hb concentrations nor-malized. A 2.1 kg male was electivelydelivered at 34 weeks and was begun ona program of hypertransfusion. Sincethen, Dr. Stephen Carr and the BrownFetal Medicine Program have similarlymanaged another Hb Bart’s pregnancy.This Cambodian-American couple firstpresented to medical attention in mid-pregnancy. After IUT, a non-hydropicmale infant was delivered at 35 weeks ofgestation. Several other affected coupleshave chosen pregnancy termination.Since our first case, there have been sev-eral additional reports describing success-ful management of Hb Bart’s fetalhydrops by IUT and scheduled deliv-ery.17-19

The approach taken at Women &Infants Hospital for this disorder beginswith genetic counseling (preferably be-fore pregnancy), identification of preg-nancies at risk and prenatal diagnosis.The maternal risks of carrying a fetuswith Hb Bart’s hydrops into the thirdtrimester make medical intervention de-sirable. In our experience, couples oftenopt for pregnancy termination, especiallywhen their α-thalassemia carrier statuswas unknown before pregnancy.

Before a couple chooses pregnancymaintenance with IUTs, they meet withseveral specialists experienced with thiscondition. The perinatologist reviews therisks of intrauterine intervention, theneonatologist describes the range of out-

154Medicine and Health / Rhode Island

comes and neonatal interventions, and apediatric hematologist reviews chronichypertransfusion and bone marrowtransplantation (BMT). TheMultidisciplinary Antenatal DiagnosisAnd Management (MADAM) programaims for preterm delivery at 33-35 weekswith minimized use of neonatal inten-sive care. Long-term management, as inβ-thalassemia major, requires regularblood transfusions and iron chelation.Transfusion dependency is expensive andcarries risks over time of central venouscatheter complications and iron accumu-lation with myocardial, pancreas or liverdamage, growth deficiency, hypogo-nadism, infection, and folic acid defi-ciency.

For these reasons, when intrauter-ine therapy is begun, we also begin hu-man leukocyte antigen (HLA) typing tosearch for a matched sibling donor. Thefirst successful HLA-matched allogeneicsibling donor BMT for α-thalassemiamajor took place in a 21-month-old girlin Hong Kong in 1997.19 Our first HbBart’s survivor is an only child, but thesecond child treated in utero has an HLA-identical sibling. In 1999, at age twoyears, he also underwent successful HLA-identical sibling donor BMT.

In these Hong Kong and Providencecases, liver biopsies before BMT showediron accumulation and hemosiderosis.After BMT, neither child has requiredblood transfusions. They also show ac-celerated developmental catch-up, al-though their physical growth remainsalong the 3rd percentile. These HLA-identical BMT recipients show mixed α-globin gene cluster chimerism, but noevidence of graft-versus-host disease(GVHD). Despite the chimerism, pe-ripheral blood cell counts of these twochildren have remained normal.

Westgren et al20 and Hayward et al21

have reported intrauterine stem cell trans-plantation in two fetuses with with Bart'shemoglobinopathy, the former with do-nor fetal liver cells (a source of extramed-ullary erythroid precursors), the latterwith haploidentical paternal CD34 cells.In each instance, the child's subsequentblood transfusion dependency was notreduced and the donor cells did not showa survival advantage compared with en-dogenous stem cells. At present, then,

and until the level of donor chimerismcan be increased, BMT in early child-hood remains the best long-term inter-vention for survivors with α-thalassemiamajor when an HLA-identical relateddonor is available. A second option,BMT from an unrelated partially mis-matched donor, may emerge as a feasiblealternative for those α-thalassemia ma-jor survivors who do not have an HLA-identical sibling. BMT across HLAboundaries has been undertaken for morecommon hemoglobinopathies, such as β-thalassemia major and clinically severesickle cell disease.

We previously have raised ethicalconcerns about embarking on an expen-sive and technology-intensive therapeu-tic course for these fetuses.8,22 Inconsidering long-term outcomes andquality of life measures, uncertaintieshave focused on associated congenitalanomalies and the potential for develop-mental compromise from intrauterinehypoxia.

Transverse limb reduction defectshave been reported frequently in fetusesaffected by Hb Bart’s disease and oc-curred in 8% of cases in one series fromHong Kong.23 The severity and numberof involved limbs varies. Our first patienthas a terminal transverse defect with ab-sence of most of the distal left foot. Atbirth, other limbs showed partial cuta-neous syndactyly of several digits. Thesefindings, fortunately, have not led tomajor motor disability (M. Msall, MD,personal communication). The patho-genesis may be terminal arteriolar occlu-sion by the abnormally large and poorlydeformable red cells which are inducedin Hb Bart’s fetuses by megaloblasticerythrocytosis. Sonographic diagnosis ofmajor terminal limb defects is sometimespossible at 10-12 weeks of gestation.24

Embryoscopy at the time ofcordocentesis and fetal MRI also mayhave a diagnostic role and assist in pa-rental counseling.

In addition, apparently, all male sur-vivors with homozygous α-thalassemiahave hypospadias. Hypospadias, ambigu-ous genitalia or misassignment of femalesex has been common in autopsy orsonographic series of stillborn fetuseswith this condition.22 Our two patientshave undergone hypospadias repairs with

good results. We have speculated that thehypospadias, like the terminal limb de-fects, may be due to ischemic tissue dis-ruption, in this instance, at the corpusspongiosum.22 Hb Bart’s-associated hy-pospadias alternatively might be due toin utero edema leading to failure of fu-sion of the urogenital folds25 or to a de-fect or deletion of another, unidentifiedgene falling within the α-globin genecluster.17

The intrauterine hypoxia has raisedconcerns about long-termneurodevelopmental outcomes in survi-vors with homozygous α-thalassemia.8,13-

15 Fortunately, as more survivors arereported, normal development, at leastinto school age, appears the most com-mon outcome. Surprisingly, this seemsto be true whether or not intrauterinetherapy was undertaken.14,16,18 Thesefindings are similar to the wider experi-ence with isoimmune hemolytic anemia,the most common indication for IUTin the United States. Isoimmunehemolytic and Hb Bart’s fetuses both areaffected by severe intrauterine anemia,although the conditions differ in timingof onset of severe disease and degree oftissue hypoxia. Nevertheless, long-termfollow-up of children treated with IUTfor hemolytic disease also indicates nor-mal developmental outcome can be ex-pected.26

In summary, without medical inter-vention α-thalassemia major is a uni-formly lethal fetal or, occasionally,neonatal disease. Current antenatal man-agement options are pregnancy termina-tion for maternal safety or, in selectedinstances, IUT and prudent delivery tim-ing. Life-long hypertransfusion for sur-vivors probably carries similarmorbidities and lifestyle constraints tothose for individuals with β-thalassemiamajor. HLA-identical related donorBMT during early childhood can cureα-thalassemia major. In the near future,intrauterine stem cell transplantationmay emerge as the therapy of choice forthis genetic disease.

REFERENCES1. Allen SJ, O’Donnell A, Alexander NDE, et

al. Proc Natl Acad Sci USA 1997; 94:14736-41.

2. Liang ST, Wong VCW, So WWK, et al.Br JObstet Gynaecol 1985; 92:680-4.

3. Lau TK, Li CY. Austral N Z J Obstet Gynaecol

155Vol. 84 No. 5 May 2001

1994; 34:416-21.4. Chui DHK, Waye JS. Blood 1998; 91:2213-

22.5. Hofgartner WT, Keefe SFW, Tait JF. Am J

Clin Pathol 1997; 197:576-81.6. Tang W, Luo HY, Eng B, et al. Lancet 1993;

342:1145-7.7. Lam YH, Tang MHY, Lee CP, Tse HY. Ul-

trasound Obstet Gynecol 1999b; 13:48-51.8. Carr S, Rubin L, Dixon D, et al. Obstet

Gynecol 1995; 85:876-9.9. Nakayama R, Yamada D, Steinmiller V, et al.

Obstet Gynecol 1986; 67:176-80.10. Tongsong T, Wanapirak C, Kunavikatikul C,

et al. Prenat Diagn 2000; 20:224-8.11. Ghosh V, Tang MHY, Lam YH, et al. Lancet

1994; 344:988-9.12. Tongsong T, Wanapirak C, Sirichotiyakul S,

et al. J Ultrasound Med 1999; 18:807-11.13. Beaudry MA, Ferguson DJ, Pearse K, et al. J

Pediatr 1986; 108:13-6.14. Bianchi DW, Beyer EC, Stark AR, et al. J

Pediatr 1986; 108:716-8.15. Lam TK, Chan V, Fok TF, et al. Acta

Haematol 1992; 88:198-200.

16. Singer ST, Styles L, Bojanowski J, et al. JPediatr Hematol Oncol 2000; 22;539-42.

17. Dame C, Albers N, Hasan C, et al. Eur JPediatr 1999; 158:217-20.

18. Fung TY, Lau TK, Tam WH, Li CK. PrenatDiagn 1998; 18:838-41.

19. Chik KW, Shing MMK, Li CK, et al. JPediatr 1998; 132:1039-42.

20. Westgren M, Ringden O, Eik-Nes S, et al.Transplantation 1996; 61:1176-9.

21. Hayward A, Ambruso D, Battaglia F, et al.Fetal Diagn Ther 1998; 13:8-14.

22. Abuelo DN, Forman EN, Rubin LP. Am JMed Genet 1997; 68:158-61.

23. Chitayat D, Babul R, Wyatt P, et al. Am JMed Genet 1997; 68:162-7.

24. Lam YH, Tang MH. Ultrasound ObstetGynecol 2000; 16:587-9.

25. Fung TY, Kin LT, Kong LC, Keung LC. AmJ Med Genet 1999; 82:225-7.

26. Hudon L, Moise KJ, Hegemier SE, et al. AmJ Obstet Gynecol 1998; 179:858-63.

Lewis P. Rubin, MD, is an AssociateProfessor of Pediatrics, and co-director ofthe Fetal Medicine Program, Brown Medi-cal School.

CORRESPONDENCE:Lewis P. Rubin, MDDepartment of PediatricsWomen & Infants Hospital101 Dudley StreetProvidence, RI 02905phone: (401) 274-1122 x1245fax: (401) 453-7571e-mail: Lewis_Rubin@brown.edu

�Farjaad M. Siddiq, MD, and Anthony A. Caldamone, MD

Fetal Obstructive Uropathy: Diagnosis and Management

In recent years fetal ultrasound exami-nation has become a routine part of

obstetrical care. In addition, improve-ments in ultrasound resolution have fa-cilitated diagnosis of fetal anomalies thatpreviously would have gone undetecteduntil postnatal symptomatic presenta-tion. A significant fetal anomaly can bedetected in approximately 1% of prena-tal ultrasounds, and approximately 20%of these anomalies are genitourinary.1

Hydronephrosis represents 50% of theseprenatally detected genitourinary abnor-malities, but only in a minority of casesis it associated with significant obstruc-tion resulting in renal dysplasia and pul-monary hypoplasia. It is important,therefore, to distinguish physiologicalfrom pathological hydronephrosis.

Evaluation of fetal renal function aswell as surgical intervention for fetal geni-tourinary obstruction is now possible. Al-though this technology has made thetreatment of fetal obstructive uropathypossible, the results have not been as suc-cessful as once hoped. Not only is ourunderstanding of diagnosis and patientselection incomplete, intervention carriesa high complication rate and renal andpulmonary sequelae may not be pre-vented. The purpose of this article is toreview the current approaches to the di-

agnosis and management of prenatallydetected hydronephrosis and fetal ob-structive uropathy.

EMBRYOLOGY

Renal development is a result of theinteraction between the ureteral bud andthe metanephric blastema. It is initiatedin the fifth week by the ureteral bud,which arises from the mesonephric ductand penetrates the metanephric blastema.The ureteral bud then undergoes severaldivisions and forms the renal collectingsystem by the 20th week of gestation. Si-multaneously, nephron development oc-curs at an exponential rate.Approximately 80% of nephrons havedeveloped by mid second trimester andnephrogenesis is complete by the 36thweek of gestation.

Fetal urine production begins at theeighth week and replaces the placenta asthe major source of amniotic fluid by 18weeks. The developing kidneys maintainappropriate amniotic fluid volumethroughout the remainder of gestation butcontribute minimally to fetal electrolyteand fluid management. At term, fetalurine production may approach 51ml/h.2

Adequate amniotic fluid volumeleads to normal pulmonary developmentand function. Lung development begins

at three to four weeks and developmentof the respiratory structures continuesuntil term. Development of the tubularstructures, however, is completed by the24th week. Low amniotic fluid volume,or oligohydramnios, interferes with nor-mal lung development due to extrinsiccompression of the lungs or loss of ad-equate internal pulmonary stenting.3 Theresult is pulmonary hypoplasia which isthe major cause of neonatal mortality inobstructive uropathy. Severe oligohydram-nios also results in compression deformi-ties of the head, thorax, and extremities.

FETAL URINARY TRACT OBSTRUC-TION

Fetal urinary tract obstruction of-ten presents as hydronephrosis. The dif-ferential diagnosis of fetal hydronephrosisincludes multicystic dysplastic kidney,vesicoureteral reflux, duplication anoma-lies, infravesical obstruction, ureterovesi-cal junction obstruction, or ureteropelvicjunction obstruction. Obstruction ofboth renal units (i.e., posterior urethralvalves or bilateral upper tract obstruction)may result in oligohydramnios with itssubsequent complications.

Genitourinary obstruction also leadsto fetal renal injury and dysplasia. Ex-perimental models have shown that mod-

156Medicine and Health / Rhode Island

erate ureteric obstruction leads to a de-crease in renal blood flow (RBF), glom-erular filtration rate (GFR), andpotassium excretion. When the obstruc-tion is relieved, renal function improvesbut not to baseline. Severe obstructionmay produce a rapidly progressive de-crease in RBF and GFR, which may notrecover.4 In 1971, Beck demonstratedthat ureteral obstruction in the fetal lambled to ipsilateral renal fibrosis, focal pa-renchymal disorganization, and other his-tologic changes consistent with renaldysplasia.5 In the early 1980s, Harrison etal created a model of obstructive fetal ur-opathy by placing a constricting devicearound the urethra of fetal lambs. Bychanging the timing and duration of theinsult during gestation, they demonstratedthat both the time of the obstruction aswell as its duration were key in producingrenal dysplasia and greatly influenced itsseverity. Lesions created early in gestationand maintained for a longer period causedthe most severe injury. Furthermore, theydemonstrated that genitourinary decom-pression in the fetal lambs could restoreamniotic fluid volume and prevent pul-monary hypoplasia.3, 6-8

These experiments suggested that inutero surgical intervention in fetuses withprenatally detected hydronephrosis andoligohydramnios may improve clinicaloutcome. Although technically feasible,early operative results were not as success-ful as expected. Often renal insufficiencypersisted and intervention did not preventpulmonary hypoplasia, leading to fetal orearly neonatal demise. In one series of 73fetuses in 1986, 60% of fetuses died de-spite intervention, implying irreversiblefetal renal damage.9 This led to the devel-opment of prognostic criteria and carefulfetal selection for intervention.

ULTRASOUND FINDINGS INOBSTRUCTIVE UROPATHY

The normal kidneys and bladdercan now be evaluated as early as 16 to 17weeks of gestation, while markedly di-lated systems can be seen as early as 12to 14 weeks.10 Hydronephrosis is themost common genitourinary abnormal-ity seen on prenatal ultrasound. Often adistended bladder with bilateral ureteraldilation is evident, implying bladder out-let obstruction. The degree of collecting

comes isotonic. Fetuses in whom thereis an elevation in urinary sodium con-centration to greater than 100mEq/L,chloride concentration greater than90mEq/L, and urine osmolality greaterthan 210mOsm/L are at high risk of ir-reversible renal dysplasia and poor post-natal renal function, even withintervention. Urinary calcium concentra-tion greater than 8mg/dl is the most sen-sitive indicator of dysplasia. Total proteingreater than 20mg/dl and PO4 greaterthan 2mMol/L also correlate with renalinjury.14 In addition, elevated fetal urinarylevels of β2-microglobulin to greater than2mg/L may reflect proximal renal tubu-lar dysfunction. β2-microglobulin is a lowmolecular weight protein that is filteredthrough the glomerular basement mem-brane and is almost entirely resorbed inthe proximal renal tubule.15 β2-microgloblin may be elevated in spite ofnormal urine electrolytes in fetuses withnormal amniotic fluid volume. This find-ing has been shown to correlate with worsepostnatal renal function at one year of agewhen compared to similar fetuses with lowurinary β2-microglobulin levels.16

Individual urinary parameters are notvery accurate in predicting presence orabsence of renal dysplasia. However, whenused in combination, these parametersgreatly improve sensitivity and specificityin diagnosing underlying renal damage.Due to normal deviation in renal func-tion, urinary electrolytes can vary result-ing in inaccurate prediction of fetal renalfunction. Furthermore, a single samplingfrom stagnant urine in an obstructed sys-tem may be inaccurate as osmotic gradi-ents and urothelial secretions may changeits composition. The first sampled urinemay, indeed, be isotonic leading to a mis-diagnosis of severe renal dysplasia. How-ever, sequential vesicocentesis which showsprogressively hypotonic urine in combi-nation with sub-threshold last urine val-ues improves diagnostic precision andindicates a reversible renal injury.17

SELECTION CRITERIA FOR FETAL

INTERVENTION

Survival for fetuses with unilateral re-nal obstruction is almost 100% as the con-tralateral kidney compensates for the loss.Bilateral hydronephrosis, however, repre-sents a potentially more serious situation,

system dilation seen on ultrasound mustbe correlated with gestational age, be-cause it can vary during gestation. Aminimal degree of hydronephrosis is of-ten physiologic, resulting in a normalneonatal examination, and requires no fur-ther intervention. It has therefore becomenecessary to establish thresholds of renalpelvic diameter that distinguish betweenphysiologic and pathologic hydronephro-sis. Currently, a renal pelvic anteroposte-rior diameter (APD) of >6mm at <20wks,>8mm at 20-30 weeks’, and >10mm at>30 weeks’ gestation is considered signifi-cant.4 In addition to renal collecting sys-tem dilation, the ultrasound shouldevaluate renal parenchymal appearance forecho density and/or cortical cysts, unilat-eral or bilateral involvement, amnioticfluid volume, fetal size and weight for age,external genitalia and gender, other organsystem abnormalities, and bladder size,thickness, and cycling.

Renal parenchymal appearance onprenatal ultrasound does not always cor-relate with renal function. Nevertheless,the presence of cortical cysts andhyperechoic parenchyma is a sign of re-nal dysplasia. In one study the presenceof renal cortical cysts had a 100% speci-ficity for renal dysplasia but it had a lowsensitivity of 60%.11 Thus the absenceof cortical cysts does not ensure the ab-sence of dysplasia. When evaluated sepa-rately, increased echogenicity of the renalcortex was less sensitive and specific thanthe presence of cortical cysts.12 Thesesonographic findings are, therefore, onlyuseful when present and their absenceprovides no prognostic information re-garding renal function. Similarly, amni-otic fluid volume is only prognostic atthe extremes.13 Thus a more sensitivemeans of assessing fetal renal function isnecessary to facilitate appropriate selec-tion of fetuses with obstructive uropa-thy for prenatal treatment.

EVALUATION OF FETAL RENAL

FUNCTION

Fetal renal function can be evalu-ated by studying the fetal urine biochem-istry and the rate of bladder filling afterultrasound guided vesicocentesis. Thehealthy fetus makes hypotonic urine. Asrenal injury progresses, proximal tubu-lar function declines and the urine be-

157Vol. 84 No. 5 May 2001

and it may be due to infravesical obstruc-tion. It may lead to oligohydramnios, whichis potentially life threatening and is a strongpredictor of an adverse outcome. There-fore, fetuses with unilateral hydronephro-sis are excluded from in utero intervention.14

Fetal intervention aims to relieve uri-nary obstruction (preserve renal function)and to restore normal amniotic fluid vol-ume (prevent pulmonary hypoplasia). Ifrenal injury is severe and irreversible attime of diagnosis, surgical intervention islikely to be unsuccessful as oligohydram-nios will persist with resultant pulmonaryhypoplasia and probably fatal neonatal res-piratory failure. Therefore, criteria for inutero intervention in obstructive uropa-thy indicative of adequate renal functionhave been used as selection criteria for fur-ther evaluation and intervention. Thesecriteria are summarized in Table 1.

APPROACH TO THE FETUS WITH

HYDRONEPHROSIS

Using the selection criteria above,fetuses with hydronephrosis can begrouped into three categories: 1) fetuseswith severe irreversible renal injury un-likely to benefit from intervention, 2)fetuses with adequate renal function tobenefit from prenatal genitourinary de-compression, and 3) fetuses with goodprognoses without intervention to avoidunnecessary and risky intervention. Analgorithm suggesting management of thefetus with antenatally detected hydro-nephrosis is presented in Figure 1.18

Fetuses with bilateral hydronephro-sis, oligohydramnios, and poor renalfunction that fail to improve on serialvesicocentesis fall into Category 1 andlikely will not improve after prenatal in-

tervention. Fetuses with bilateral hydro-nephrosis in whom assessment demon-strates good renal function based onsonographic findings, urinary biochem-istry and post-vesicocentesis bladder fill-ing fall into Category 2 and should beconsidered for decompression. Fetuseswith bilateral hydronephrosis and ad-equate amniotic fluid volume through-out gestation fall into Category 3 and areunlikely to benefit from in utero inter-

vention. Their pulmonary developmentshould not be compromised and theirgenitourinary abnormalities are best ad-dressed postnatally.

FETAL INTERVENTION

Percutaneous vesicoamniotic shuntplacement, open fetal surgery, andfetoscopic surgery have been reported asapproaches to decompress the obstructedfetal genitourinary system. Ultrasoundguided percutaneous shunt placement isthe most widely used technique to dateand several series of prenatal shunt place-ment have been reported. The results,however, have been mixed and difficultto interpret due to wide variations in pa-tient selection, operative technique, andoutcome measurement. In a recent report,Freedman et al, reviewed the records of55 consecutive patients presenting be-tween 1987 and 1994 using the selectioncriteria described above.19 They concludedthat when evaluated by specific diagnosis,intervention appeared to provide out-comes in these high-risk fetuses that were

Table 1. Criteria for prenatal intervention

Ultrasound Findings Bilateral HydronephrosisNo renal cortical cystsNormal renal parenchymal echogenicityAdequate bladder filling after vesicocentesisOligohydramnios

Urine BiochemistryNa <100 mEq/LCl <90 mEq/LOsm <210 mOsm/LCa <8 mg/dlPO4 <2 mMol/LTotal Protein <20 mg/dlβ2-microglobulin <2 mg/L

Figure 1. Algorithm for management of fetal hydroephrosis. (From Gloor JM18)

158Medicine and Health / Rhode Island

comparable to those for disease detectedpostnatally. However, the study failed toanswer whether these comparable out-comes were due to shunting or the natu-ral history of the obstructive disease processthat did not necessarily need intervention.Strictly speaking, therefore, the benefit ofprenatal genitourinary decompression re-mains questionable. The use ofvesicoamniotic shunting must also be tem-pered by a 45% complication rate.20 Thisincludes inadequate shunt drainage orshunt migration, premature labor, urinaryascites, chorioamnionitis, and iatrogenicgastroschisis.14 Nevertheless, it may be rea-sonable to assume that some benefit maybe gained by prenatal intervention in thishigh-risk population.

Open surgical decompression of thefetal bladder had been suggested in fe-tuses less than 24 weeks’ gestation asshunt placement is technically most dif-ficult in early gestation fetuses with se-vere oligohydramnios. Bilateralcutaneous uteterostomies were first re-ported in a 21-week fetus in 1981.21 Al-though technically successful, the fetusdied at birth due to pulmonary hypo-plasia. There are several other reports inthe literature of open surgical decompres-sion, but the results have not been betterthan percutaneous shunting.14 Further,open procedures are riskier with compli-cations including infection, premature la-bor, fetal death, and surgical failure. Themother receives no direct benefit but as-sumes the inherent risk of general anes-thesia, laparotomy, and hysterotomy.

Fetoscopic surgery is the newest ap-proach to decompression of the fetalgenitourinary tract. A cutaneousvesicostomy using an argon laser has beenreported in a 17-week fetus.22 It requiresno hysterotomy as fetal cystoscopy istechnically possible through a 0.7mmfiberoptic endoscope. As technology im-proves, endoscopic intervention holdspromise for the future.

Termination of the pregnancy maybe considered in fetuses with severe re-nal dysplasia and oligohydramnios thatcannot be reversed by intervention. Thesefetuses have a uniformly poor outcomeand most die at birth or soon thereafter.Termination may also be considered ifother organ systems are affected or thereare karyotypic abnormalities.

Early delivery may be helpful if thereis new onset third trimester oligohydram-nios (28 to 32 weeks). The fetal lungs canbe induced to mature with steroids andlung maturity can be assessed by measur-ing the amniotic lecithin-sphingomyelinratio. There is a high risk of perinatal res-piratory distress. After 35 to 36 weeks’gestation delivery can be induced in thefetus with oligohydramnios with little pul-monary risk. There are, however, currentlyno studies documenting improved out-comes after early delivery.23

CONCLUSION

The advent of routine prenatal ul-trasound examination has resulted in anincrease in the detection of fetal hydro-nephrosis. Although fetal hydronephro-sis is not synonymous with urinaryobstruction, findings on ultrasound andfetal urinalysis can suggest obstructiveuropathy. Intervention should be limitedto chromosomally normal fetuses withbilateral hydronephrosis, oligohydram-nios, and adequate renal function.

The basic structure of the kidneysand lungs is complete by 12 weeks. Thediagnosis of obstructive uropathy, how-ever, often is not made until well into thesecond trimester when injury may alreadybe irreversible. Thus it is understandablewhy the results of fetal intervention havenot been uniformly successful. As the sur-gical and diagnostic technology improves,successful intervention earlier in gestationmay become feasible with better neonataloutcomes.

REFERENCES1. Elder JS. Pediatr Clin N Am

1997;44:1299-321.2. Rabinowitz R, Peters MT, Vyas S, et al.

Am J Obstet Gynecol 1989;161:1264.3. Harrison MR, Nakayama DK, Noall R,

de Lorimier AA.J Pediatr Surg1983;18:247-56.

4. Shokeir AA, Nijman, RJM.BJU Int2000;85:987-94.

5. Beck AD. J Urol 1971;105:784-9.6. Harrison MR, Ross NA, Noall R. J

Pediat Surg 1982;17:965-74.

7. Glick PL, Harrison MR, Noall RA, etal. J Pediatr Surg 1983;18:681-.

8. Glick PL, Harrison MR, Adzick NS, etal.J Pediatr Surg 1984;19:649-75.

9. Manning Fa, Harrison MR, Rodeck C.NEJM 1986;315:336-40.

10. Bronstein M, Yotte JM, Brandes JM, etal. Prenat Diagn 1990;10:653-66.

11. Crombleholme TM, Harrison MR,Golbus MS, et al. Am J Obstet Gynecol1990;162:1239-44.

12. Mahoney BS, Filly RA, Callen PW, etal. Radiol 1984;152:143-9.

13. Cendron M, D’Alton ME,Crombleholme TM. Semin Perinatol1994;18:163-81.

14. Coplen DE. J Urol 1997;157:2270-7.15. Burghard R, Fordjani N, Bald R. Fetal

Ther 1987;2:188-96.16. Muller F, Dommergues M, Mandelbrot

L, et al. Obstet Gynecol 1993;82:813-20.17. Johnson MP, Corsi P, Evans MI, et al.

Am J Obstet Gynecol 1995;173:59-65.18. Gloor JM. Mayo Clin Proc

1995;70:145-52.19. Freedman AL, Bukowski TP, Gonzalez

R, et al. J Urol 1996;156:720-4.20. Elder JS, Duckett JW, Snyder HM.

Lancet 1987;2:1007-10.21. Harrison MR, Golbus Ms, Filly RA, et

al. NEJM 1982;306:591-3.22. Macmahon RA, Renou PM, Shekelton

PA, et al. Lancet 1992;340:1234.23. Coplen DE. AUA Update Series

2000;XIX:130-6.

Farjaad M. Siddiq, MD, is a resi-dent in the Department of Urology, BrownMedical School.

Anthony A. Caldamone, MD, is aProfessor of Urology and Chief of the Divi-sion of Pediatric Urology, Brown MedicalSchool.

CORRESPONDENCE:Anthony A. Caldamone, M.D.Division of Pediatric UrologyHasbro Children’s Hospital2 Dudley Street, Suite 170Providence, RI 02905phone: (401) 444-5795fax: (401) 444-6947e-mail: Anthony_Caldamone@brown.edu

159Vol. 84 No. 5 May 2001

�Arlet G. Kurkchubasche, MD

The Fetus With an Abdominal Wall Defect

Antenatal detection of abdominal walldefects has impacted the perinatal

care of both the expectant mother andof the fetus. Prenatal referral to tertiarycare centers that can provide for the sur-gical needs of the infant has also allowedfor focused management from the obstet-ric perspective to identify the unique prob-lems associated with these pregnancies.With advances in maternal-fetal medicine,obstetrics and neonatal surgery and theincreasing availability of in utero interven-tions it is essential to determine which cur-rent therapeutic interventions result inoptimized outcomes and where future in-vestigational efforts should be directed. Inthis age of information technology weneed to provide expectant parents with re-liable and useful information.

Although not specifically elucidated,the etiologies of omphalocele and gas-troschisis are likely widely discrepant,based not only on the spectrum of associ-ated anomalies in the fetus but also thediffering demographics of the maternalpopulations. This dichotomy extends tothe postnatal period in terms of operativemanagement and morbidity and mortal-ity. Vital to appropriate counseling andstratification of risk therefore, is the abil-ity to make a specific diagnosis for a fetuswith an abdominal wall defect.

On sonogram, the presence of a de-fect to the right of the umbilicus, witheviscerated bowel that is not containedwithin a membrane is consistent with gas-troschisis. The fetus with omphalocele hasan absent abdominal wall subjacent to thecord insertion site with a membrane usu-ally containing the protuberant liver andeviscerated intestine. Localization of thedefect is helpful particularly to avoid di-agnostic errors associated with the rare rup-tured omphalocele that masquerades asgastroschisis.

OMPHALOCELE

Approximately 20% of anterior ab-dominal wall defects are omphaloceles(abdominal wall defects at the level ofthe umbilicus, usually covered by a mem-

brane) Antenatal evaluation of the fetuswith omphalocele focuses on the associ-ated conditions. These may include le-thal chromosomal anomalies(particularly trisomy 13 and 18), con-genital cardiac defects, other upper mid-line/thoracic defects such as in thePentalogy of Cantrell (sternal, diaphrag-matic, pericardial defects with ectopiacordis and omphalocele) or the lowermidline OEIS complex (omphalocele,exstrophy, imperforate anus, spinal de-fect). Other associated conditions includeBeckwith-Wiedeman syndrome, cleftlip/palate and cryptorchidism. The inci-dence of associated anomalies (exclud-ing intestinal malrotation, which isuniformly present in those with largedefects) is reported as high as 69%.1

In a recent study of 23 fetuses orinfants with the pre or postnatal diagno-sis of omphalocele, 21 fetuses had anantenatal diagnosis made by 18 weeksgestation.2 In 18 pregnancies, the diag-nosis was correct. (two false positives, and3 false negatives). Associated anomalieswere correctly identified in 12 but incor-rectly reported in 8. There were 13 ter-minations including 2 trisomy 18s andone trisomy 13. Two fetal deaths followedamniocentesis. Of the 10 live births, 9had their ventral defect repaired with aone-year survival rate of 89%.

When providing antenatal counsel-ing to parents, this information needs tobe relayed within the appropriate con-text. Those liveborn infants with anomphalocele and without additional life-threatening anomalies have a lesion thatis amenable to surgical therapy with goodoutcomes.

SURGICAL CONSIDERATIONS

Repair of the omphalocele providesspecific challenges to the infant and sur-geon, but in over 50% cases, a primaryrepair can be achieved. The spectrum ofdefects ranges from the “hernia of thecord,” which could potentially be re-duced and closed at the bedside, toomphalocele minor with a fascial defect

of up to 4cm and to omphalocele majortypically with defects from 4 cm to 8 cm.Given the closed nature of the defect,with liver and intestines enclosed in peri-toneum and amnion, postnatal manage-ment can initially focus on the potentiallylethal malformations. Once these havebeen identified and addressed, usuallywithin 24-48 hours, decisions can bemade regarding surgical closure. In con-trast to gastroschisis, the intestinal tractis usually normal, but the size of the de-fect and the liver may provide major im-pediments to complete fascialapproximation. Viscero-abdominal dis-proportion refers to the discrepancy be-tween the current abdominal capacityand the extra-abdominal volume of evis-cerated organs. Aggressive reduction intothe abdomen may result in compromisedhepatic or visceral perfusion requiringurgent decompression. Infants with verylarge defects may require staged closureto allow for gradual expansion of the ab-dominal wall. This may involve: 1) pri-mary coverage with skin flaps withsubsequent ventral hernia repair, 2)staged closure using a silo, with or with-out excision of the sac or 3) topical treat-ment may induce sufficient woundcontraction with epithelialization toachieve closure for subsequent ventralhernia repair. Infants with lethal cardiacor chromosomal disorders can be man-aged nonoperatively with topical therapy.

Extended hospital courses and com-plications are primarily limited to thosewith defects measuring greater than 8 cmin diameter. Even in this group the sur-gical mortality was only 8%.1 In the ab-sence of associated severe anomalies theseinfants can have an uncomplicated coursewith a normal long-term quality of life.Less optimal outcomes are determinedprimarily by the nature of the chromo-somal defect and the complexity of asso-ciated cardiac and other organ systemdefects. With improvements in the re-construction of these complex anoma-lies, this will further reduce mortality andimprove quality of life. As such, the an-

160Medicine and Health / Rhode Island

tenatal assessment by a multidisciplinaryteam including perinatalogists, neona-tologists, geneticists, cardiologists andsurgeons will have critical impact on thedecision to continue the pregnancy.

GASTROSCHISIS

In a gastroschisis, the defect is al-ways to the side (and almost always tothe right) of the umbilicus, which is nor-mal. A gastroschisis is never covered by amembrane. The perinatal managementof infants with gastroschisis is distinctfrom those with omphalocele. Whereasthe size of the ventral defect and associ-ated anomalies dictate prognosis inomphalocele, the relevant parameters inthe infant with gastroschisis are relatedto the condition of the newborn and theintestine. Short bowel syndrome with itsattendant risks remains one of the sig-nificant complications of the diagnosisof gastroschisis.

Fetuses with gastroschisis tend to besmall for gestational age and are born toyoung primiparous women, often afterpreterm labor. Although the specific fac-tors leading to this congenital malforma-tion have not been elucidated, the focushas rested on environmental and poten-tially nutritional factors. Studies from theCalifornia birth defects monitoring pro-gram have proposed that a lowprepregnancy body mass may representa risk factor for offspring with gastroschi-sis.3 These investigators suggest that ab-normal levels of 3 nutrients (low alphacarotene, low total glutathione and highnitrosoamines) are potential candidatesfor further investigation.4 Much ofthe clinical and basic science investiga-tion into gastroschisis has tried to iden-tify factors that contribute to theintestinal wall thickening and formationof a peel over the serosal surface, the find-ings that most impede reduction of theintestine into the abdominal cavity andthat are thought to contribute to thedysmotility encountered postoperatively.Conventional wisdom attributes thesechanges to exposure to amniotic fluid,although not all infants with gastroschi-sis exhibit the serosal peel. A recent ani-mal study has sought to differentiatebetween urinary and gastrointestinalwaste products in amniotic fluid, and hasimplicated components of meconium as

the more significant sources of inflam-mation.5 Saline amnioinfusion per-formed both in an animal model and ina small cohort of patients with gastroschi-sis and severe oligohydramnios was foundto be associated with less inflammatorypeel as compared to non-amnioinfusedinfants with gastroschisis.6, 7,8 These con-cerns have been the premise for advocat-ing early delivery of these infants,particularly when visceral distension isnoted to be progressive, suggesting an un-derlying intestinal obstruction. Vascu-lar etiologies of the intestinal atresias andof the inflammatory changes have beenproposed and may be related to constric-tion of the mesentery by the approximat-ing fascial edges as evidenced in fetusesborn with antenatal detection of gas-troschisis and consequent jejunal atresiaor congenital short bowel syndrome(SBS) without abdominal wall defect.Based on the premise that the amnioticinsult to the intestine is cumulative anda function of time, preterm induction oflabor was considered prudent so as to en-hance the ability to achieve primary clo-sure. In the current literature, norandomized prospective series exists tosupport this intervention and preliminaryevidence from our series of inborn pa-tients in whom no attempt was made toinduce early labor suggests that there isno beneficial effect to early delivery andthat term infants recover as well if notbetter than their preterm counterparts.Premature labor however, remains a fea-ture associated with gastroschisis and maynot be an avoidable event in approxi-mately 30% of patients.9 Debate in theperinatal management of gastroschisis hasalso revolved around the mode of deliv-ery with Cesarean section advocated bymultiple centers. Vaginal delivery, how-ever, has been shown to be safe in mul-tiple recent studies and general consensuswould indicate that a trial of labor is ap-propriate and that Cesarean sectionshould be reserved for obstetric indica-tions only.10,11,12

OPERATIVE MANAGEMENT

Antenatal counseling by a pediatricsurgeon will focus on the immediate sur-gical care to be delivered to an infant withexposed viscera that are at risk for fur-ther vascular compromise. The options

for acute management range from op-erative intervention either in the deliv-ery room or in the operating room. Theexposed intestine has a variable degree ofinflammatory peel. When extensive, thismay prohibit identification of an intesti-nal atresia. In virtually all cases, thebowel length appears shortened, with athickened mesentery. Sedation and pa-ralysis with expansion of the lateral ab-dominal wall may enable completereduction of the viscera and permit fas-cial closure. If not feasible then a silo,typically spring-loaded and no longerrequiring fascial sutures, can be inserted.This can also be accomplished at the bed-side with minimal sedation. A recent pro-spective trial of routine insertion of a siloas compared to emergency operatingroom closure provided favorable resultsfor the routine insertion of the silo withreduced number of days to extubation,to full feeds and to home discharge.13 Thepostoperative course of these infants istypically marked by a prolonged ileus,during which they rely on parenteral nu-trition support. When intestinal conti-nuity has not become evident after severalweeks, contrast studies are performed todelineate the anatomy and to exclude thepossibility of an occult atresia. By thistime much of the inflammatory peel,which may have been present initially,will have resolved and now allows forintestinal resection and anastomosis toestablish continuity. Short bowel syn-drome may occur as a consequence ofatresias or after postnatal hypoperfusioninsults to the intestine or even florid ne-crotizing enterocolitis. With appropriatenutritional management focusing onmeasures to avoid cholestasis, these in-fants can be transitioned to full enteralfeedings.

The use of promotility agents hasnot been shown to be useful in expedit-ing normal motility.12 Motility agents andacid suppression therapy however mayplay a role in a significant number of in-fants who have evident gastroesophagealreflux.14 Although both omphalocele andgastroschisis are associated with intesti-nal malrotation, the occurrence of gas-troesophageal reflux during the first yearof life is reported to be higher inomphalocele than gastroschisis.

161Vol. 84 No. 5 May 2001

REFERENCES1. Dunn JCY, Fonkalsrud EW. Improved sur-

vival of infants with omphalocele. Am J Surg1997;173: 284-7.

2. Holland AJ, Ford WD, Linke RJ, et al. In-fluence of antenatal ultrasound on the man-agement of fetal exomphalos. Fetal DiagnosisTherapy 1999;14:223-8

3. Lam PK, Torfs CP, Brand RJ. A low preg-nancy body mass index is a risk factor for anoffspring with gastroschisis. Epidemiol1999;10: 717-21.

4. Torfs CP, Lam PK, Schaffer DM, et al. Asso-ciation between motherís nutrient intake andtheir offspringís risk of gastroschisis. Teratol1998;58:241-50.

5. Akgur FM, Ozdemir T, Olguner M, et al.An experimental study investigating the ef-fects of intraperitoneal human neonatal urineand meconium on rat intestines. Res ExperimMed 1998;198:207-13.

6. Luton D, de Lagausie P, Guibourdenche J, etal. Influence of amnioinfusion in a model ofin utero created gastroschisis in the pregnantewe. Fetal Diagnosis Therapy 2000;15:224-8.

7. Sapin E, Mahieu D, Borgnon J, et al. Trans-abdominal amnioinfusion to avoid fetal de-mise and intestinal damage in fetuses withgastroschisis and severe oligohydramnios. JPediatr Surg 2000;35: 598-600.

8. Luton D, de Laugausie P, Guibourdenche J,et al. Effect of amnioinfusion on the outcomeof prenatally diagnosed gastroschisis. FetalDiagnosis Therapy 1999;14:152-5.

9. Anteby EY, Sternhell K, Dicke JM. The fe-tus with gastroschisis managed by a trial oflabor: antepartum and intrapartum compli-cations. J Perinatol 1999;19:521-4

10. How HY, Harris BJ, Pietrantoni M, et al. Isvaginal delivery preferable to cesarean deliv-ery in fetuses with a known ventral wall de-fect? Am J Obstet Gynecol 2000;182:1527-34.

11. Snyder CL. Outcome analysis for gastroschi-sis. J Pediatr Surg 1999; 34:1253-6.

12. Kumar RK, Shi EC Duffy B. Cisapride andCaesarian section: their role in babies withgastroschisis. J Pediatr Child Health 1999;35:181-4.

13. Minkes RK, Langer JC, Mazziotti MV, et al.Routine insertion of a silastic spring-loadedsilo for infants with gastroschisis. J PediatrSurg 2000;35:843-6.

14. Koivusalo A, Rintala R, Lindahl H. Gastroe-sophageal reflux in children with a congeni-tal abdominal wall defect. J Pediatr Surg1999;34:1127-9.

Arlet G. Kurkchubasche, MD, is anAssistant Professor of Surgery and Pediat-rics, Brown Medical School.

CORRESPONDENCE:Arlet G. Kurkchubasche, MDDivision of Pediatric SurgeryHasbro Children’s Hospital2 Dudley Street, Suite 180Providence, RI 02905phone: (401) 421-1939fax: (401) 868-2319e-mail:Arlet_Kurkchubasche@brown.edu

�Lloyd R. Feit, MD

Fetal Cardiac Arrhythmias: Diagnosis and Management

In the last two decades, the subspecialtyof fetal cardiology has grown dramati-

cally. Prenatal diagnosis and manage-ment have evolved from basicdescriptive reports to more invasive andexperimental therapeutics. The man-agement of fetal arrhythmias has seenthe greatest opportunity for in uterotherapy. This paper will review currentdiagnostic and management strategiesfor the fetus with a cardiac arrhythmia.Optimal outcome relies on a well inte-grated team approach to the maternal-fetal patient set. Careful considerationof the risk-benefit ratio of any therapyas it relates to both expectant motherand fetus must be discussed. Accuratediagnosis of fetal arrhythmias requiresknowledge of, and experience with,highly specialized ultrasonographictechniques, and is crucial prior to con-sideration of fetal therapy. Any attemptat developing a rational treatment planrequires a working knowledge of thelikely electrophysiologic principles ofthe suspected arrhythmia, the pharma-cology of the pertinent antiarrhythmicagents, as well as their pharmacokinet-ics when utilized in the pregnant pa-tient.

DEFINITION, INCIDENCE &ETIOLOGY

One may define a fetal arrhythmiaas any irregular rhythm, or any sustainedregular rhythm whose rate falls outsidethe normal fetal range of 120-160 bpm.This excludes the well-characterized ab-normalities of fetal heart rate seen in thecourse of labor and delivery, as well asthose associated with fetal distress forother reasons. It has been estimated thata cardiac arrhythmia is detected in ap-proximately 1-4% of fetuses,1of whichroughly 10% are considered potentiallyserious.2 Table 1 lists the more commonindications for fetal arhythmia evalua-tion.

Cardiac arrhythmias may begrouped into three main categories: iso-lated extrasystoles, tachyarrhythmias,

(sustained fetal heart rate greater than160-180 beats per minute) andbradyarrhythmias (sustained fetal heartrates below 110-120 beats per minute).The critical initial step in the manage-ment of the fetus with a suspected ar-rhythmia is accurate diagnosis.

DIAGNOSTIC TECHNIQUES

Cremer3 first described the fetal elec-trocardiogram (ECG) in 1906. Althoughtechnically feasible, the quality and in-tegrity of the signal obtained has pre-vented its routine application in clinicalpractice. Indirect methods presume theelectrical activity of the heart by evalua-tion of the physical consequences of atrialand ventricular contraction. By their in-ferential nature these techniques are ob-viously more prone to errors in diagnosis.

Table 1. Indications for Fetal Arrhythmia EvaluationSuspected arrhythmia

Non-immune hydrops fetalis

Structural congenital heart disease(esp heterotaxy syndromes, corrected transposition)

Fetal cardiac tumors

Maternal collagen vascular disease

Maternal medications/toxins that may predispose fetus to arrhythmia

162Medicine and Health / Rhode Island

Atrial systole (and the inferred atrialdepolarization, or p wave on the ECG)may be detected by evaluation of atrio-ventricular (AV) valve motion or atrialwall directly. Ventricular wall motion (oropening of the semilunar valves) are me-chanical events that indirectly representventricular depolarization (the QRScomplex of the ECG). By careful analy-sis of this association, one can deduce thespecific arrhythmia in a manner similarto that used with a surface ECG. Simi-larly, Doppler interrogation of blood flowserves as a reliable recorded physical con-sequence of chamber depolarization.Evidence of atrial depolarization may bedetermined when assessing the flow pat-tern across either of the AV valves. Dop-pler evaluation in either of the greatarteries will represent ventricular depo-larization. Simultaneous recording ofvelocities in the pulmonary artery andpulmonary vein can also determine fetalrhythm.

ISOLATED EXTRASYSTOLES

The vast majority of fetal arrhythmiasnoted in routine obstetric care are benignextrasystoles; isolated atrial or ventricularpremature contractions that if infrequent,have no physiologic significance for thedeveloping fetus or newborn. Isolated ex-trasystoles generally comprise about 80%-90% of suspected arrhythmias for whichfetuses are referred to tertiary centers forevaluation.6 These extrasystoles typicallyare perceived as pauses or ‘dropped beats’noted during routine auscultation of fetalheart rate. Determination of the originof the premature beat may de difficult. Ifincessant, or if occurring in a recogniz-able recurring pattern, premature extra-systoles may rarely be indicative of moresignificant pathology and complete fetalechocardiography is indicated.

Sustained arrhythmias may be clas-sified into two groups (Table 2) basedon the observed fetal heart rate (FHR):tachyarrhythmias, are those in which theFHR is greater than 160 bpm, orbradyarrhythmias, in which the FHR isless than 120 bpm. Any fetus with a sus-tained arrhythmia should have a com-plete level II ultrasound including a fetalechocardiogram to detect any potentialassociated defects.

mia. The atrial flutter rate is typically inthe range of 400-600 bpm and is associ-ated with some degree of AV block andlower ventricular rates. For example, anatrial rate of 500 bpm with accompany-ing 2:1 AV block would result in a ven-tricular rate of 250 bpm. The degree ofAV block may be variable, resulting inan irregular ventricular response. The ex-tremely high atrial rates make accuraterecording with either Doppler or m-mode techniques far more challenging.Atrial fibrillation is exceedingly rare, andis characterized by a paucity of organizedatrial depolarization or contraction andis always associated with a variable andirregular ventricular response.

VENTRICULAR TACHYCARDIA

Ventricular tachycardia (VT) is veryrare in the prenatal setting, accountingfor less than 3-5% of all sustainedtachyarrhythmias. The electrophysiologicmechanism usually involves an irritableventricular focus related to ischemia, car-diomyopathy, myocarditis, maternaldrug use or electrolyte abnormality. Thefocus may be associated with cardiacrhabdomyoma seen in tuberous sclero-sis, but is not usually associated with anyspecific form of structural heart disease.The rates encountered have been widelyvariable (180-400 bpm). The hallmarkof diagnosis is the presence of atrioven-tricular dissociation, where the ventricu-lar rate exceeds the atrial rate, with noclear relationship between ventricular andatrial depolarization/contraction.

BRADYARRHYTHMIAS

The causes for sustainedbradyarrhythmias include sinus bradycar-dia and high grade (second or third de-gree) heart block. (Table 2) As notedearlier for sinus tachycardia, sinus brady-cardia is rarely a pathologic process in andof itself, but is more likely a physiologicresponse to some other process, mostcommonly increased vagal tone from avariety of causes. Postnatal experience hasshown that newborns may tolerate sinusrates in the 50-70 bpm range with nountoward effects.

Congenital heart block is usuallynoted on routine prenatal screening asan unusually low heart rate, generally inthe 60-80 bpm range. Evaluation finds

TACHYARRHYTHMIAS

These are by far the most commonof the sustained arrhythmias seen in thepractice of fetal cardiology, and are madeup of a wide-ranging group of distinctdiagnoses, listed in Table 2. The etiol-ogy of fetal tachyarrhythmias generallymirrors that seen in the newborn infant,with some important differences. As ageneral rule, sinus tachycardia is not apathological diagnosis in and of itself, butrather the physiologic consequence ofanother pathologic process affecting thefetus such as anemia, thyrotoxicosis orinfection. At higher rates (200-230 bpm)it may be difficult to differentiate be-tween sinus tachycardia and some of theother relatively “slower” (and less com-mon) supraventricular tachycardias suchas ectopic focus atrial tachycardia, orjunctional ectopic tachycardia. Mostpathologic arrhythmias, (ie not sinus ta-chycardia) will begin and end abruptly,while sinus tachycardia occurs moregradually. This may be observed whileperforming a vagal maneuver, such asexerting moderate head pressure.

REENTRANT SUPRAVENTRICULAR

TACHYCARDIA

The most common type oftachyarrhythmia seen in clinical practiceis reentrant, or reciprocating supraven-tricular tachycardia (SVT), which makesup approximately 65-85% of all sus-tained fetal tachycardias.7 Reentrant SVTis rarely associated with structural heartdisease. As their names imply, both formsof reentrant tachycardia require a one-to-one relationship of atrial and ventricu-lar depolarization, and this should becarefully sought when attempting tomake either of these diagnoses in utero.Rates may range from 220 bpm to as highas 300 bpm; the more typical rate ob-served has been 240-250 bpm.

ATRIAL FLUTTER/ATRIAL FIBRILLA-TION

Atrial flutter and atrial fibrillationare relatively uncommon, comprisingapproximately 10-15% in many series.Like reentrant SVT, they are not usuallyassociated with any specific form of struc-tural heart disease, though any conditionthat results in dilated atria predisposesthe fetus to development of atrial arrhyth-

163Vol. 84 No. 5 May 2001

the atrial rate to be normal (120-160bpm) accompanied by a slower ventricu-lar rate. Congenital heart block may ei-ther be second degree or (morecommonly) third degree, which is alsotermed complete heart block. Seconddegree heart block is defined as the in-termittent loss of normal atrioventricu-lar (AV) conduction. Complete or thirddegree heart block is the more commonlyencountered form of sustained bradycar-dia, where there is no conduction what-soever of any electrical impulse from theatria to the ventricles. It is useful to pointout the difference between complete AVblock, where the atrial rate is alwaysgreater than the ventricular rate, and AVdissociation, (seen with ventricular tachy-cardia) where the ventricular rate is fasterthan the atrial rate.

Both second and third degree AVblock may be broadly grouped into twocategories: those with and those withoutassociated structural heart disease. Ap-proximately half of all cases of completeheart block are associated with structuralheart disease,8 and include those defectswith atrial isomerism (heterotaxy syn-dromes), AV septal defects or L-transpo-sition of the great arteries (ventricularinversion) as part of their cardiac mal-formation. The other group, those withgrossly normal cardiac anatomy are al-most invariably associated with mater-nal collagen vascular disease. The mothermay be asymptomatic at the time fetalheart block is diagnosed. Prognosisis significantly worse in fetal heart blockwhen associated with structural heartdisease, hydrops fetalis or very low ven-tricular escape rate (<50 bpm).

THERAPY

The decision to treat the fetus witha cardiac arrhythmia is a difficult onebased on many factors. Initially, onemust decide whether the arrhythmia isassociated with poor outcome, which isgenerally manifested by the developmentof fetal hydrops and its associated highmortality. If the fetus is hydropic at pre-sentation, there is general agreement thatantiarrhythmic therapy is warranted,though optimal route of administrationis highly variable among practitioners(oral vs. intravenous vs. transumbilical).In general, sustained arrhythmias are far

more likely to progress to fetal hydrops,though the duration of the arrhythmiaafter which hydrops is likely to developis unknown for any given arrhythmia atany given rate. As with non-sustainedarrhythmias, it is not known what per-centage of normal sinus rhythm in a 24-hour period is required to prevent thedevelopment of hydrops. One must con-sider the likelihood that the arrhythmiain question will respond to the proposedtherapy, based on past experience andreports in the literature from many cen-ters. Potential maternal morbidity relatedto proposed therapy must enter into thedecision before attempting transplacen-tal therapy. A thorough understandingof the proarrhythmic and myocardial de-pressant effects of nearly all antiarrhyth-mic drugs for both the mother and fetusis essential.

Maternal monitoring during theinitiation phase of therapy should, at aminimum, include daily and/or continu-ous electrocardiographic monitoring forpotential proarrhythmic effects, as wellas surveillance specifically tailored to theagent used. Route of administration mustalso be considered. If transplacentaltherapy is reasonable, the maternal in-travenous route is generally chosen forthe initial (loading) phase, with changeto oral therapy once the arrhythmia hasbeen controlled. Alternatively, direct fe-tal therapy via cordocentesis is feasible,and may be indicated if severe hydropsis present, or if placental edema is thoughtto interfere with transplacental therapy.Parental concerns, social circumstancesand informed consent as to the parents’understanding of thepotential risks andbenefits must also bedetermined. It is im-portant to rememberthat if fetal lung ma-turity can be docu-mented, or facilitatedby the administrationof antenatal steroids,delivery of the baby isusually the therapy ofchoice, since treat-ment of the arrhyth-mia is much safer andmore reliable whenadministered to the

infant directly. If transplacental ortransumbilical therapy has been unsuc-cessful, and/or maternal morbidity pre-cludes continuation of antiarrhythmictherapy, knowledge of institutional out-come statistics for various birthweightswill be of great importance in decisionmaking. The myriad factors at play againhighlight the optimum outcomeachieved by a coordinatedmultidisciplinary fetal team.

ISOLATED EXTRASYSTOLES

As mentioned above, isolated atrialor ventricular premature contractions arebenign and require no therapy save forwatchful observation for the unlikely as-sociation with a sustainedtachyarrhythmia.

TACHYARRHYTHMIAS

Supraventricular tachycardiaWhen considering therapy for sus-

tained SVT, thought should be given tothe likely electrophysiologic mechanism.Treatment is generally directed at slow-ing conduction in one or both of theavailable pathways (AV node or accessoryconnection) in the reentrant circuit,thereby “breaking the electrical loop.”Additional efficacy may be gained bychoosing an antiarrhythmic drug thatalso suppresses atrial extrasystoles.

Digoxin has been the first linetherapy of choice for many years. It actsprimarily by slowing conduction at theAV node, but also reduces the frequencyof atrial extrasystoles. It is also the onlyantiarrhythmic agent without any nega-tive inotropic effects. Generally, initia-

Table 2. Sustained Arrhythmias

TachyarrhythmiasSinus tachycardiaSupraventricular tachycardia

AV (atrioventricular) reentrant tachycardia (AVRT)(eg WPW syndrome)

AV Node reentrant tachycardia (AVNRT)Atrial flutter/fibrillationJunctional tachycardia

Ventricular tachycardia

BradyarrhythmiasSinus bradycardiaHigh grade (2nd or 3rd degree) AV block

Maternal collagen vascular diseaseStructural congenital heart disease

164Medicine and Health / Rhode Island

tion of digoxin therapy is accomplishedby intravenous loading or “digitalization”consisting of 1.5 - 2.0 mg in the first 24-36 hours of therapy. Maternal digoxintoxicity (nausea, vomiting, malaise,blurry vision, arrhythmia, etc.) can beminimized by close attention to mater-nal renal function, serum potassium con-centration and common druginteractions. Maintenance dosing shouldbe based on maternal trough digoxin lev-els, therapeutic effect, and toxicity, butcommonly fall in the range of 0.5-1.5mg per day due to the high maternalglomerular filtration rate seen duringpregnancy. Dosing will be guided by dailytrough drug level as transplacentaldigoxin pharmacokinetics are highly vari-able. The goal of therapy is to achieve amaternal serum concentration near thetoxic range of 2.0-2.3 ng/ml if well tol-erated by the mother.

If digoxin fails to achieve conversionto sinus rhythm, a second drug may beadded to the regimen. The most com-monly utilized classes of drugs have beenthe class IA agents (procainamide) andclass IC agents (flecainide). Amiodarone(mixed class II and class III action) andsotalol have also been used with efficacy.Calcium channel blockers are generallyconbtraindicated. Each of these medica-tions has its own list of potential risksthe full review of which is beyond thescope of this paper. Perhaps the mostimportant factor in guiding the use ofsecond line antiarrhythmics in this situ-ation is personal physician experience.The most effective agent to add to orsubstitute for digoxin when it alone failshas not yet been established.

Intrauterine treatment of atrial flut-ter or fibrillation is aimed at control ofventricular response rate. Again, digoxinhas the broadest experience as first linetherapy, with second line drugs used asin reentrant SVT. Atrial flutter may beparticularly resistant to pharmacotherapy,prompting important consideration forearly delivery and electrical cardioversion.

VT is much less common in the fe-tus than are other tachyarrhythmias, andrequires the documentation of AV dis-sociation for its diagnosis. Unlike the su-praventricular arrhythmias however,digoxin is not the drug of first choice,and in fact may aggravate the ventricular

tachycardia by precipitating ventricularfibrillation, a frequently lethal event.Drugs of first choice in this scenariowould be those in class IB, such aslidocaine or mexilitene. Alternatives in-clude procainamide and amiodarone.

Complete heart blockThe fetus with CCHB on the basis

of maternal connective tissue diseasewithout hydrops does not requiretherapy, but should be followed closelyfor the development of hydrops. Thebradycardic fetus with hydrops is at ex-tremely high risk for fetal or neonataldemise regardless of the presence of as-sociated structural heart disease and thuswarrants attempts at in utero therapy orearly delivery. Plasmapheresis and steroidadministration have been variably suc-cessful. The fetus with structural heartdisease and heart block is at high risk ofdeveloping hydrops, and must be closelyobserved. Theoretically, fetal ventricular(or even sequential atrioventricular) pac-ing would seem the logical treatment ofchoice. Both techniques have been ac-complished via maternal hysterotomyand fetal sternotomy in the fetal lamb,but have not yet been successfully appliedto the human fetus. The future may bein the use of fetal thoracoscopy to reducethe morbidity associated with exterioriz-ing the fetus. Early delivery may be thetreatment of choice even if the fetus ispreterm, if other therapeutic options havebeen exhausted.

SUMMARY

The diagnosis and management offetal cardiac arrhythmias requires com-plex skills and knowledge, and has had agreat impact on the care of infants withcongenital heart disease and their fami-lies. Optimal benefits will be derivedfrom a thoughtful team approach, withskillful internal communication, and es-pecially when parental involvement is en-couraged in the decision making process.

REFERENCES1. Southall DP, Richards JM, Hardwick RA, et

al. Prospective study of fetal heart rate andrhythm patterns. Arch Dis Child1980;55:506-11.

2. Reed KL. Fetal arrhythmias: Etiology, diag-nosis, pathophysiology, and treatment. SeminPerinat 1989;13:294-304.

3. Cremer M. Uber die direkte Ableitung derAktonstrome des menschlichen Herzens vonOesophagus and uber dasElektrodkardiogramm des Fotus. MunchenMedWochnschr 1906; 53:811-6.

4. Kleinman CS, Hobbins JC, Jaffe CC, et al.Echocardiographic studies of the human fe-tus: Prenatal diagnosis of congenital heartdisease and cardiac arrhythmias. Pediatrics1980;65:1059-67.

5. Reed KL, Sahn DL, Mark GR, et al. CardiacDoppler flows during fetal arrhythmias:Physiologic consequences. Obstet Gynecol1987; 70:1-6.

6. Kleinman CS, Copel JA, Weinstein EM, etal. In utero diagnosis and treatment of fetalsupraventricular tachycardia. Semin Perinatol1985;9:113-29.

7. Bergmans MGM, Jonker GJ, Kock HCLV.Fetal supraventricular tachycardia: review ofthe literature. Obstet Gynecol Surv 1985;40:61-8.

8. Schmidt KG, Ulmer HE, Silverman NH, etal. Perinatal outcome of fetal complete atrio-ventricular block: a multicenter experience. JAm Coll Cardiol 1991;17:1360-6.

9. Hansmann M, Gembruch U, Bald R, et al.Fetal tachyarrhythmias: transplacental anddirect treatment of the fetus- a report of 60cases. Ultrasound Obstet Gynecol 1991;1:162-70.

10. Kleinman CS, Copel JA: Electrophysiologi-cal principles and fetal antiarrhythmictherapy. Ultrasound Obstet Gynecol1991;1:286-97.

11. Chameides L, Truex RC, Vetter V, et al. As-sociation of maternal systemic lupus erythe-matosus with congenital complete heartblock. NEJM 1977; 297:1204-7.

12. Ward RM. Maternal drug therapy for fetaldisorders. Semin Perinatol 1992;16:12-20.

13. Meijboom EJ, van Engelen AD, van de BeekEW, et al. Fetal arrhythmias. Curr OpinCardiol 1994;9:97-102.

Lloyd R. Feit, MD, is an AssociateProfessor of Pediatrics and Chief of Pediat-ric Echocardiography, Brown MedicalSchool.

CORRESPONDENCE:Lloyd R. Feit, MDDivision of Pediatric CardiologyHasbro Children’s Hospital593 Eddy StreetProvidence, RI 02903phone: (401) 444-4232fax: (401) 444-6378e-mail: Lfeit@lifespan.org

165Vol. 84 No. 5 May 2001

�Stephen R. Carr, MD, and Michael P. Plevyak, MD

Contemporary Evaluation and Management ofTwin-Twin Transfusion Syndrome

Twinning rates vary by twin typeand location throughout the

world. The rate of monozygotic twin-ning is quite constant worldwide at 3.5/1000 deliveries while dizygotic twinningrates range from 49/1000 deliveries inNigerian populations to 1.3/1000 de-liveries in Japanese populations. Thirtyper cent of monozygotic twins resultfrom division of the embryo within thefirst 72 hours after fertilization and re-sult in monochorionic-diamniotictwins. These are the twins that are atrisk for twin-twin transfusion syndrome.Twin pregnancies evidence greater mor-bidity and mortality than do singletonpregnancies, and among twin pregnan-

cies monozygotic twins have greatermorbidity and mortality than do dizy-gotic twins. The morbidities seen morecommonly in monochorionic twins in-clude structural defects and twin-twintransfusion. Twin-twin transfusion syn-drome (TTTS) is the consequence ofunbalanced blood flow from one twin(the donor) to the other (the recipient)across transplacental vascular commu-nications and results in the polyhydram-nios/oligohydramnios sequence and caninclude growth discordance. These vas-cular communications (of which thereare four types) are present in nearly100% of monochorionic twins but oc-cur only rarely in dichorionic twins.1, 2

The diagnosisof TTTS has in thepast been based onneonatal findings,but these findingsare unreliable forthe prenatal detec-tion of the syn-drome. Currentstrategies use ultra-sound to determineboth the presenceand severity of pre-natally diagnosedTTTS. Table 1 liststhe salient points inultrasound diagno-sis of TTTS.

MANAGEMENT

OF TTTSOnce diag-

nosed, untreatedTTTS results inmorbidity and mor-tality that exceeds70%.3 Despite en-thusiasm for differ-ent modalities,treatment of TTTSis associated withsurvival rates of only

60-70%. Even more troubling is theincreased incidence of cerebral palsyand other cerebral impairment (from20-40%) in the surviving co-twinwhen one of a set of monochorionictwins dies in utero.4, 5

Several treatment modalities havebeen used to treat TTTS. Those receiv-ing the most attention include serialamnioreduction, septostomy, andfetoscopic laser ablation ofchorioangiopagus vessels (FLOC).

AMNIOREDUCTION

Amnioreduction is the removal oflarge quantities of amniotic fluid fromthe polyhydramniotic sac of the recipi-ent twin. This is accomplished usingan 18 or 20 gauge needle under ultra-sound guidance and is performed fromone to several times. Althoughamnioreduction does not address thepostulated cause of TTTS, it is postu-lated to result in decreased pressure onthe transplacental vascularanastamoses, which increases placen-tal compliance, thus reducing thepreload and afterload in the hearts ofboth twins.6 The reduction in intrau-terine volume also appears to decreasethe incidence of preterm labor, a ma-jor contributor to the morbidity ofTTTS. Proponents of amnioreductionpoint to itís simplicity and successrates; one recent trial of aggressiveamnioreduction reported 57% survivalof both twins at 24 months of age and70% survival at 24 months of age of atleast one twin.7 Preterm prematurerupture of membranes complicates 8%of pregnancies treated with serialamnioreduction.8

SEPTOSTOMY

Septostomy (deliberate creation ofa defect in the membrane separatingthe two twins) has been proposed as amethod to equalize the pressures in thetwins’ amniotic sacs. In this technique

Table 1. Ultrasound diagnosis oftwin-twin transfusion syndrome (TTTS)

A. Monochorionic twin gestation

1. Twins of same gender

2. Thin inter-twin membrane

3. Single placenta

B. Discordance in amniotic fluid volume

1. “Donor” twin with decreased fluid (deepest vertical pocket < 2.0 cm is consistent with severe TTTS).

2. “Recipient” twin with increased amniotic fluid volume (deepest vertical pocket > 8.0 cm is consistent with severe TTTS).

C. Other ultrasound findings

1. Appearance of a “stuck” twin that does notchange intrauterine position regardless ofmaternal position.

2. Small or non-visualized bladder in the “donor”twin

3. Large bladder in the recipient twin

4. Abnormal Doppler findings

a. pulsatile umbilical venous blood flow

b. absent or reversed end-diastolicumbilical arterial blood flow

c. tricuspid valve regurgitation.

5. Hydrops: ascites, pleural or epicardial effusions

166Medicine and Health / Rhode Island

a 20 or 22 gauge needle is introducedto the uterine cavity using ultrasoundguidance in such a way as to deliber-ately breach the amniotic membraneoverlying the smalleroligohydramniotic sac. A recent studyof 12 patients with severe TTTStreated with intentional septostomyyielded 75% survival of both twins todelivery and 92% survival of at leastone twin.9 The authors, and others,postulate that deliberate septostomyresults in an equilibration of amnioticfluid volumes around both twins as aresult of hydrostatic pressure differ-ences between the sacs that may be toosmall to measure. As inamnioreduction this technique doesnot directly address the postulatedcause of TTTS, the transplacental vas-cular anastamoses, but offers tempo-rizing measures in an attempt toprolong the pregnancy to the pointwhere survival ex utero is possible.

FETOSCOPIC LASER ABLATION OF

CHORIOANGIOPUS VESSELS

Fetoscopic laser ablation ofchorioangiopagus vessels (FLOC) isthe only proposed intervention forTTTS that directly addresses the pos-tulated etiology of TTTS, that is, the

transplacental vascular communica-tions. This technique, first describedby De Lia et al,10 uses a fetoscopically-directed neodynium-YAG laser tophotocoagulate those transplacentalvascular communications felt to becontributing to the TTTS. Initial useof this technology involved ablation ofall vessels crossing the vascular equa-tor of the placenta, but more recentlythere has been more selective ablationinvolving only those vessels thought tobe contributing to the TTTS. Datafrom recent series indicate 69% sur-vival of both twins, 82% survival of atleast one twin, and 4.3% significanthandicap in survivors of an in-uterodemise.11

CHOICE OF TREATMENT

Direct comparison of the efficacyof the different available interventionshas not been accomplished. Compari-sons extant in the TTTS literature typi-cally compare case series using oneintervention with case series using a dif-ferent intervention. One such recentcomparison12 compared outcomes in73 cases of severe TTTS treated withFLOC in one center with 43 cases ofsevere TTTS treated with serialamnioreduction in another center.

(Table 2; * denotes p < 0.05)There are currently two random-

ized trials underway in attempts to de-termine the most effective interventionfor treatment of TTTS. TheEUROFOETUS consortium ( http://www.eurofoetus.org) randomizescases diagnosed with severe TTTS be-tween serial amnioreduction andFLOC. At the University of NorthCarolina13 the trial is randomizing be-tween serial amnioreduction andseptostomy. The relatively infrequentnature of severe TTTS cases near anyone center, and the (usually) stronglyheld opinions of a given treatmentteam has rendered recruitment intothese trials more time-consuming thanmight have been originally anticipated.

Timing the intervention is of para-mount importance in the treatment ofTTTS. As in any medical procedure,the risks of the procedure itself mustbe weighed against the risk of the dis-ease that is being treated. The knownrisks of any of the interventions forTTTS (preterm premature rupture ofmembranes, infection, preterm labor,placental abruption, hemorrhage andfetal death) must be acknowledged indeciding when to intervene. Quinteroet al14 presented a classification schemabased on their experience (Table 3).The group recently published their re-sults following FLOC using their stag-ing schema15 (Table 4).

COMPLICATION OF TTTSConsideration of the morbidity

following diagnosis and treatment ofTTTS is of equal importance to sur-vival. In cases of TTTS treated withserial amnioreduction that are compli-cated by in-utero demise of one twin,

Table 2. Comparison of FLOC and amnioreduction

FLOC Serial amnioreduction

2 survivors 42% 42%≥ 1 survivor 79%* 61%*SAb 12% 7%Double fetal loss 3%* 19%*Neonatal deaths 6% 14%Abnormal brain scan survivor(s) 6%* 18%*Birth weight (donor) 1750 grams* 1145 grams*Birth weight (recipient) 2000 grams 1560 grams*p<0.05

Table 3. Staging of severe TTTSaccording to Quintero RA

Stage I: MVP > 8 cm in recipient and < 2 cm in donor

Stage II: stage I and bladder not seen in the donor twin

Stage III: stage II and critically abnormal Doppler findings(absent or reversed end-diastolic flow in theumbilical artery, pulsatile umbilical venous flow,reverse ductus venosus flow)

Stage IV: stage III and hydrops

Stage V: stage IV and demise

Table 4. Survival by stage of severeTTTS (Quintero RA et al)

Stage No Survivors > 1 Survivor Total

I 3/19 (15.8%) 16/19 (84.2%) 19

II 4/24 (16.7%) 20/24 (83.3%) 24

III 4/23 (17.4%) 19/23 (82.6%) 23

IV 0/4 (0.0%) 4/4 (100.0%) 4

Total 11/70 (15.7%) 59/70 (84.3%) 70

167Vol. 84 No. 5 May 2001

neurological handicap is seen in ap-proximately 30% of survivors.16 Casesof TTTS treated with FLOC that arecomplicated by in-utero demise of onetwin experience neurological handicapin 4.2%17 (compared with the 18%incidence of neurological handicapseen in singleton survivors followingserial amnioreduction.12 Taken at facevalue, these data suggest that whileoverall survival appears to differ littlebetween the different interventions,there is a lesser risk of neurologic mor-bidity in survivors following FLOCthan in survivors following serialamnioreduction.

There have also been reports oflimb reduction anomalies and intesti-nal atresia associated with TTTS. Table5 summarizes published cases of struc-tural anomalies associated with mono-chorionic twinning. The etiology ofthese defects remains unclear. Hecheret al18 suggested that polycythemia andan arterial steal syndrome were theprobable etiology of necrotic toes de-tected prior to FLOC for severe TTTS.Margono et al19 suggested that theirfindings of thrombosis of the transpla-cental vascular connections and necro-sis of the right foot of the survivingtwin were consistent with a throm-boembolic phenomenon. Lundvall etal20 found necrosis in the right lowerleg of the recipient twin 27 days afterFLOC. Post mortem examinationfound a thrombus in the right com-mon iliac artery, “presumably the re-sult of polycythemia.” Scott andEvans21 documented a case of severeTTTS managed with serialamnioreduction that resulted in 2 live

twins. At time of delivery the recipienttwin was found to have left lower legnecrosis that was associated with poly-cythemia (Hb 26.8 g/dL, Ht 89%).The authors concluded that the necro-sis was the result of hyperviscosity dueto polycythemia. Dawkins et al6 re-ported a pregnancy with severe TTTSmanaged with amnioreduction x 6 overa nine week period. Delivery was at 32weeks. The recipient twin was bornwith Hb 25.9 g/dL, Ht 72% and gan-grene of the left lower leg. Arul et al22

presented two cases of severe TTTStreated with FLOC. In both cases therewas demise of the donor twin and inboth survivors ileal atresia was notedafter birth. The authors suggest threepossible etiologies for these findings:hypoperfusion or hyperviscosity asso-ciated with TTTS could cause mesen-teric ischemia; death of the donorcould affect the hemodynamics of thesurvivor, causing mesenterichypoperfusion; a shower of emboli orthromboplastins could be released intothe fetal circulation. Van Allen et al23

described two sets of monochorionictwins. The first case documented asingleton demise at 12 weeks EGA. Atbirth the surviving twin had cleft lipand palate and terminal limb reductionwith ring constrictions of the left handand both feet. In the other case single-ton demise was documented at 18weeks EGA. At birth the survivor wasfound to have ring constrictions of theleft hand digits and left big toe. Therewas no evidence of amniotic bands ineither case. The authors suggest thatthese findings were the result of vascu-lar disruption in the co-twin. In our

case TTTS was first diagnosed at 14weeks and was treated with serialamnioreduction x 7. FLOC was per-formed at 23+ weeks EGA. Pretermpremature rupture of membranes oc-curred at 26+ weeks EGA and cesar-ean delivery was at 28+ weeks. At timeof delivery necrosis of the left lowerextremity of the recipient was seen. Thetoe-heel length of the necrotic limb of3.2 cm is consistent with 19 weeks 4days gestation, which is prior to eitherthe FLOC or the amnioreductions.Polycythemia was not universally seenin these cases, but is the most commonfinding among these pregnancies withsevere TTTS affected with structuralanomalies. Polycythemia could resultfrom the elevated atrial natriuretic pro-tein found in recipient twins and theresulting diuresis, and would lead tohyperviscosity. This hyperviscositywould result in greater incidence ofthrombosis. It remains unclear why thelower extremities are more affected bysuch a thrombotic diathesis.

CONCLUSION

Monochorionic twins presentunique challenges above and beyondthose associated with multiple gesta-tion. There have been several develop-ments in the evaluation and treatmentof twin-twin transfusion syndrome, butthese twins remain at high risk. In spiteof these advances, these twins are atcontinued risk for anomalies that ap-pear to result from hypoperfusion.These anomalies have been seen inmonochorionic twins that have under-gone serial amnioreductions, FLOC orno intervention at all. This suggests

Table 5. Ischemic complications of severe TTTS

EGA (wk) EGA (wk)atat onset intervention intervention lesion outcome

Scott 1995 24 24, 25, 26 amnio x 3 gangrene L leg 2 liveHecher 1994 22 22 FLOC gangrene L foot 1 live (recipient)Margono 1992 28 none none gangrene R foot 1 live (recipient)Dawkins 1995 23 23-32 amnio x 6 gangrene L leg 2 liveLundvall 1999 17+ 17+ FLOC gangrene R leg TOPArul 2001 19 19 FLOC ileal atresia 1 live (recipient)

20 20 FLOC ileal atresia 1 live (recipient)Van allen 1992 12 none none CL/CP, L hand, both feet 1 live

18 none none L hand, L toe 1 livecurrent 14 14 - 23s amnio x 7; FLOC gangrene L leg 2 live

168Medicine and Health / Rhode Island

that the tendency towards these defectsis intrinsic to monochorionic twinssuffering from TTTS, and is not re-lated to the interventions that havebeen used in attempts to mitigate theimpact of TTTS. The care ofmonoamniotic twins affected by TTTScontinues to require coordinated careby a team of highly trained individu-als.

REFERENCES1. Robertson EG, Neer KJ. Placental in-

jection studies in twin gestation. Am JObstet Gynecol 1983;147:170-4.

2. Blickstein I. The twin-twin transfusionsyndrome. Obstet Gynecol1990;76:714-21

3. Urig MA, Clewell WH, Elliott JP.Twin-twin transfusion syndrome. AmJ Obstet Gynecol 1990;163:1522-6.

4. Cincotta RB, Gray PH, Phythian G,et al. Long term outcome of twin-twintransfusion syndrome. Arch Dis ChildFetal Neo 2000;83:F171-6.

5. Pharaoh POD, Adi Y. Consequencesof in-utero death in a twin pregnancy.Lancet 2000;355:1597-602.

6. Dawkins RR, Marshall TL, RogersMS. Prenatal gangrene in associationwith twin-twin transfusion syndrome.Am J Obstet Gynecol 1995;172:1055-7.

7. Mari G, Detti L, Oz U, Abuhamad A.Long term outcome in twin-twintransfusion syndrome treated with se-rial aggressive amnioreduction. Am JObstet Gynecol 2000;1183:211-7.

8. Moise KJ. Polyhydramnios: problemsand treatment. Semin Perinatol1993;17:197-209.

9. Saade GR, Belfort MA, Berry DL, etal. Amniotic septostomy for the treat-ment of Twin Oligohydramnios-Poly-hydramnios sequence. Fetal DiagnTher 1998;13:86-93.

10. De Lia JE, Cruikshank DP, Keye WR.Fetoscopic laser occlusion ofchorioangiopagus vessels in severe twintransfusion syndrome. Obstet Gynecol1990;75:1046-53.

11. De Lia JE, Kuhlmann RS, Lopez KP.Treating previable twin-twin transfu-sion syndrome with fetoscopic lasersurgery: outcomes following the learn-ing curve. J Perinat Med 1999;27:61-7.

12. Hecher KH, Plath H, Bregenzer T, etal. Endoscopic laser surgery versus se-rial amniocenteses in the treatment ofsevere twin-twin transfusion syn-drome. Am J Obstet Gynecol1999;180:717-24.

13. Dorman K, Saade GR, Smith H,Moise KJ. Use of the world wide webin research: randomization in a multi-center clinical trial of treatment fortwin-twin transfusion syndrome.Obstet Gynecol 2000;96:636-9.

14. Quintero RA, Morales WJ, Allen MH,et al. Staging of Twin-twin transfusionsyndrome. J Perinatol 1999;19:550-5.

15. Quintero RA, Morales WJ, Allen MH,et al. Staging of Twin-Twin Transfu-sion syndrome. Frontiers in FetalHealth 2000;2:10-6.

16. Mahoney BV, Petty CN, Nyberg DA,et al. The “stuck twin” phenomenon:ultrasonic findings, pregnancy out-come and management with serialamniocenteses. Am J Obstet Gynecol1990;163:1513-22.

17. Ville Y, Hecher K, Gagnon A, et al.Endoscopic laser coagulation in themanagement of severe twin-to-twintransfusion syndrome. Br J ObstetGynaecol 1998;105:446-53.

18. Hecher K, Ville Y, Nicholaides K.Umbilical artery steal syndrome anddistal gangrene in a case of twin-twintransfusion syndrome. Obstet Gynecol1994;83:862-5.

19. Margono F, Feinkind L, Minkoff HL.Foot necrosis in a surviving fetus asso-ciated with twin-twin transfusion syn-drome and monochorionic placentathat received no intervention. ObstetGynecol 1992;79:867-9.

20. Lundvall L, Skibsted L, Graem N.Limb necrosis associated with twin-twin transfusion syndrome treatedwith YAG-laser coagulation. ActaObstet Gynecol Scand 1999;78:49-50.

21. Scott F, Evans N. Distal gangrene in apolycythemic recipient fetus in twin-twin transfusion. Obstet Gynecol1995;86:677-9.

22. Arull GS, Carroll S, Soothill PW,Spicer RD. Intestinal complications as-sociated with twin-twin transfusionsyndrome after antenatal laser treat-ment: report of two cases. J PediatrSurg 2001;36301-2.

23. Van Allen MI, Siegel-Bartelt J, DixonJ, et al. Constriction bands and limbreduction defects in two newbornswith fetal ultrasound evidence for vas-cular disruption. Am J Med Genet1992;44:598-604.

Stephen R. Carr, MD, is an Asso-ciate Professor of Obstetrics and Gyne-cology and co-director of the Program inFetal Medicine, Brown Medical School.

Michael P. Plevyak, MD, is a Clini-cal Instructor in Obstetrics and Gynecol-ogy, Brown Medical School.

CORRESPONDENCE:Stephen R. Carr, MDDivision of Maternal-Fetal MedicineWomen & Infants’ Hospital101 Dudley StreetProvidence, RI 02903phone: (401) 274-1122 x 2348fax: (401) 453-7622e-mail: scarr@wihri.org

169Vol. 84 No. 5 May 2001

�François I. Luks, MD, PhD

Fetal Surgery

In the last five decades, increasedknowledge of fetal physiology and

ever-improving diagnostic capabilitieshave paved the way for fetal medicine.As more conditions are discovered inutero, the possibility of intervening be-fore birth has become real. For most fe-tal conditions, however, the bestapproach is still postnatal intervention.Exceptionally, a structural condition maydeteriorate through late gestation andplace the fetus’s survival at risk. In these,even the earliest postnatal treatment maycome too late. It is for these rare condi-tions that fetal surgery may be consid-ered.

The first recorded operative inter-vention on a human fetus was reportedin 1963,1 the same year that Liley, con-sidered the father of modern fetal therapy,performed his first percutaneous intrau-terine transfusion. Reports of fetal sur-gery were few and unsuccessful until thelate 1980s when Michael Harrison andhis team at University of California SanFrancisco applied their experience withanimal models to clinical situations.2, 3

Fetal surgery is a formidable under-taking, and should be reserved for a se-lect group of conditions. Several criteriamust be fulfilled before surgical interven-tion can be considered: 1) accurate diag-nosis should be possible, and 2) thecondition should be differentiated fromother, non-surgical diseases; 3) the natu-ral history of the condition should bepredictable with reasonable accuracy, sothat no fetuses are needlessly operated;4) the condition should be lethal or se-verely debilitating if left untreated, justi-fying this very aggressive form oftreatment; and 5) surgical correction ofthe condition should be technically fea-sible.4 This further implies that fetal sur-gery should only be performed at a fewspecialized centers wheremultidisciplinary expertise is available.

With such stringent conditions, fewdiseases have been considered surgicallycorrectable before birth, and some indi-cations have since been abandoned. An

early example was obstructive hydroceph-alus not associated with chromosomal orother systemic conditions. It is possibleto place a ventriculoamniotic shunt, inan attempt to relieve intracranial pres-sure caused by aquaductal stenosis.5 Un-fortunately, the technical feasibility of theprocedure was offset by the frequent find-ing that the fetus would pull out theshunt. Moreover, increased survival offetuses with ventriculoamniotic shuntresulted in a higher rate of neurologicsequelae in this group than in the un-treated fetuses, many of whom would diebefore term. This prompted a morato-rium on this type of procedures.

Bilateral urinary tract obstruction,most often caused by posterior urethralvalves in male fetuses, appeared to be apromising indication. Extensive experi-mentation in the fetal lamb establishedthe feasibility of vesical decompression,by percutaneous cystostomy or operativevesicostomy. Thus, the onset of hydro-nephrosis, renal dysplasia and renal fail-ure could be averted, and sufficientamniotic fluid could be restored to avoidpulmonary hypoplasia and neonataldeath. Unfortunately, patient selectionproved to be difficult.6 Establishing re-nal function is still frustratingly unreli-able, and even successfully shuntedfetuses often go on to develop end-stagerenal failure in infancy or early childhood.Consequently, few fetuses undergoshunting of the bladder as a means tosalvage renal function.

Certain thoracic lesions, such ascongenital cystic adenomatoid malfor-mations (CCAM) of the lung and se-questrations, can sometimes causemassive compression of the ipsi- and con-tralateral lungs, jeopardizing pulmonaryfunction at birth. Rarely, these fetusesmay even develop hydrops secondary tomediastinal shift and impairment ofvenous return. Operative interventionhas had some success in the hands of fe-tal surgeons in San Francisco and, morerecently, Philadelphia. Through an hys-terotomy, the fetal chest is incised and

the affected lobe resected.7 Fortunately,only a minority of fetuses will requireantenatal intervention, because up to85% of these lesions spontaneously re-gress later in gestation.8

Congenital diaphragmatic hernia(CDH) is probably the most publicizedindication for fetal surgery. When it oc-curs early in gestation, herniation of ab-dominal viscera into the chest causescompression of the lung, leading to pul-monary hypoplasia and immaturity atbirth. Although repair of the hernia isrelatively straightforward, as many as40% of infants born with this conditiondie in the neonatal period, most com-monly from refractory pulmonary hyper-tension. Extensive experimentation withvarious animal models has greatly in-creased our knowledge of normal andabnormal lung development. Thus, sev-eral investigators have demonstrated thatreduction of the viscera before birth couldallow catch-up growth of the lung andnormalization of lung function at birth.9

However, positive results in fetal lambsand nonhuman primates could not betranslated into clinical successes: Openfetal surgery proved to be much toostressful and invasive, as the gravid uterusis exquisitely sensitive to surgical trauma.In addition, reducing the liver oftencaused acute kinking of the ductus veno-sus and interruption of placentofetalblood flow.10 Furthermore, the small ab-dominal cavity could often not accom-modate this sudden increase in content,necessitating the placement of a tempo-rary silo. As a result, more than 60% offetuses operated for diaphragmatic her-nia died intraoperatively or secondary topremature labor and pulmonary insuffi-ciency.

In addition to the poor results withopen fetal surgery, antenatal interventionfor diaphragmatic hernia was becomingmore difficult to justify as postnatal re-sults improved. The use of extracorpo-real membrane oxygenation (ECMO)and novel ventilatory strategies has greatlyincreased survival, which is now upward

170Medicine and Health / Rhode Island

of 70% in some centers. Thus, the barhas been raised for fetal surgery.

Two developments, in the last de-cade, have given new hope to the futureof fetal surgery. First, the popularizationof laparoscopy and other minimally in-vasive procedures paved the way for en-doscopic fetal surgery.11, 12 This approachto the fetus, which is kept in utero dur-ing the operation, proved to be much lessaggressive to the uterus as well.13, 14 Tech-niques are still being refined, and minia-turization has allowed the use oftelescopes and instruments of 2 mm di-ameter and less. One of the most com-mon applications of this new technologyis laser ablation of placental vessels intwin-to-twin transfusion syndrome.

Second, a little-known phenom-enon was applied to the treatment of dia-phragmatic hernia in utero: when thetrachea of the fetus is occluded, lung fluidis trapped and lung growth is accelerated.Thus, elegant experiments in fetal lambswith diaphragmatic defects showed howtracheal occlusion caused gradual pulmo-nary hyperplasia, which reduced the vis-cera into the abdomen.15 At birth, theanimals had normal respiratory function.Clinical application soon followed, withmixed results. Fetal tracheal occlusion viathe open surgical technique yielded a33% survival rate,16 but endoscopic fetalsurgery appears less traumatic (up to 75%survival). The optimal technique of tra-cheal occlusion is still a matter of debate,as it must be complete, atraumatic andreversible at birth. The exact timing andduration of tracheal occlusion is likely tobe important, to minimize exaggeratedgrowth (which can lead to hydrops) andavoid the side effects of accelerated lunggrowth. Indeed, several investigators havedemonstrated that prolonged tracheal oc-clusion causes severe surfactant defi-

ciency.17-19

Finally, the technique of trachealocclusion-has been refined as well. Withthe use of endoscopic surgery, it is nowpossible to place a tracheal clip or, betteryet, to perform endotracheal occlusionby means of fetal bronchoscopy.20 Thus,a condition that previously required agenerous hysterotomy, exteriorization ofthe fetus, fetal laparotomy and thorac-otomy, may be replaced with an endo-scopic procedure performed through asingle, 3 mm port into the gravid uterus.Early results are encouraging, and appearto compare favorably with open fetal sur-gery for diaphragmatic hernia (Table 1).It is important to realize that this sub-group of patients has a much worse prog-nosis than most infants born with CDH.Nevertheless, postnatal results are im-proving steadily, and prognostic indica-tors are still not optimal. Therefore, fetalsurgery for diaphragmatic hernia, and formost other surgical conditions of the fe-tus, remains a semi-experimental under-taking. As technology improves and ourunderstanding of fetal conditions grows,it is likely that more fetal conditions willbe amenable to in utero interventions.

REFERENCES1. Adamsons K Jr. Fetal surgery. NEJM

1966;275:204-6.2. Harrison MR, Jester JA, Ross NA. Correction

of congenital diaphragmatic hernia in utero. I.The model: intrathoracic balloon produces fa-tal pulmonary hypoplasia. Surg 1980;88:174-82.

3. Harrison MR, Bressack MA, Churg AM, et al.Correction of congenital diaphragmatic herniain utero. II. Simulated correction permits fetallung growth with survival at birth. Surg1980;88:260-8.

4. Harrison MR, Adzick NS. The fetus as a pa-tient. Surgical considerations. Ann Surg1991;213:279-91.

5. Glick PL, Harrison MR, Halks-Miller M, et al.Correction of congenital hydrocephalus in uteroII: Efficacy of in utero shunting. J Pediatr Surg1984;19:870-81.

6. Glick PL, Harrison MR, Golbus MS, et al.Management of the fetus with congenital hy-dronephrosis II: Prognostic criteria and selec-tion for treatment. J Pediatr Surg1985;20:376-87.

7. Adzick NS, Harrison MR, Crombleholme TM,et al. Fetal lung lesions: management and out-come. Am J Obstet Gynecol 1998;179:884-9.

8. Roggin KK, Breuer CK, Carr SR, et al. Theunpredictable character of congenital cystic lunglesions. J Pediatr Surg 2000;35:801-5.

9. Adzick NS, Outwater KM, Harrison MR, etal. Correction of congenital diaphragmatic her-nia in utero. IV. An early gestational fetal lambmodel for pulmonary vascular morphometricanalysis. J Pediatr Surg 1985;20:673-80.

10. Harrison MR, Adzick NS, Flake AW, et al.Correction of congenital diaphragmatic herniain utero: VI. Hard-earned lessons. J Pediatr Surg1993;28:1411-7.

11. Deprest JA, Luks FI, Peers KH, et al. Intrauter-ine endoscopic creation of urinary tract obstruc-tion in the fetal lamb: a model for fetal surgery.Am J Obstet Gynecol 1995;172:1422-6.

12. Luks FI, Deprest JA, Vandenberghe K, et al. Amodel for fetal surgery through intrauterine en-doscopy. J Pediatr Surg 1994;29:1007-9.

13. Luks FI, Peers KH, Deprest JA, et al. The effectof open and endoscopic fetal surgery onuteroplacental oxygen delivery in the sheep. JPediatr Surg 1996;31:310-4.

14. van der Wildt B, Luks FI, Steegers EA, et al.Absence of electrical uterine activity after endo-scopic access for fetal surgery in the rhesus mon-key. Eur J Obstet Gynecol Reprod Biol1995;58:213-4.

15. DiFiore JW, Fauza DO, Slavin R , et al. Experi-mental fetal tracheal ligation reverses the struc-tural and physiological effects of pulmonaryhypoplasia in congenital diaphragmatic hernia.J Pediatr Surg 1994;29:248-56.

16. Flake AW, Crombleholme TM, Johnson MP,et al. Treatment of severe congenital diaphrag-matic hernia by fetal tracheal occlusion: clinicalexperience with fifteen cases. Am J Obstet Gynecol2000;183:1059-66.

17. Benachi A, Delezoide AL, Chailley-Heu B, etal. Ultrastructural evaluation of lung matura-tion in a sheep model of diaphragmatic herniaand tracheal occlusion. Am J Respir Cell Mol1999;Biol 20:805-12.

18. Bin Saddiq W, Piedboeuf B, Laberge JM, et al.The effects of tracheal occlusion and release ontype II pneumocytes in fetal lambs. J PediatrSurg 1997;32:834-8.

19. De Paepe ME, Papadakis K, Johnson BD, et al.Fate of the type II pneumocyte following tra-cheal occlusion in utero: a time-course study infetal sheep. Virchows Arch 1998;432:7-16.

20. Papadakis K, Luks FI, Deprest JA, et al. Single-port tracheoscopic surgery in the fetal lamb. JPediatr Surg 1998;33:918-20.

Table 1. Results of fetal surgery for congenital diaphragmatic hernia(University of California, San Francisco and Childrenís Hospital of Philadelphia)

Date N Survival (%) CommentOpen fetal surgery(“classic” repair)10 1993/97 18 7 (39) 1 late neonatal death

Open fetal surgery(tracheal clip):

San Francisco 1998 28 6 (21) 1 died in infancyPhiladelphia16 2000 15 5 (33)

Endoscopic tracheal clip 1998 16 11 (69) 3 vocal cord damages

171Vol. 84 No. 5 May 2001

�CME Background Information

This CME activity is sponsored by Brown Medical School.

TARGET AUDIENCE

This enduring material is designed for physicians licensed in Rhode Island.

CME OBJECTIVES

After completing this CME activity, the primary care physician will be able to meet the following objectives:1. Describe the management of the fetus with anemia.2. Describe the work-up and treatment of the fetus with a urinary tract anomaly.3. Describe the role of prenatal counseling in the management of abdominal wall defects.4. Understand the treatment principles of fetal cardiac anomalies.5. Describe the role of invasive fetal intervention in twin-twin transfusion syndrome, and other surgical conditions of the fetus.

NEEDS ASSESSMENT

Traditionally the purview of obstetricians, fetal medicine today is multi-disciplinary: ob/gyn, pediatrics, surgery and their subspecialties all participate in the care of the unbornpatient. These papers are intended for residents, fellows and attending physicians in obstetrics and gynecology, perinatology and maternal-fetal medicine, neonatology and pediatrics,pediatric surgical specialties, genetics, and for medical students, genetic counselors, midwives, obstetrical nurses and clinical social workers, to present up-to-date information on fetalmedicine.

ACCREDITATION STATEMENT

Brown Medical School is accredited by the Accreditation Council for Continuing Medical Education (ACCME) to provide continuing medical education for physicians.

CREDIT DESIGNATION

Brown Medical School designates this education activity for 2 hours in category 1 credit toward the AMA Physician’s Recognition Award. Credit can be obtained by readingthe issue and completing the following quiz. The estimated time for completion of this activity is 2 hours.

DATE OF ORIGINAL RELEASE

This issue was published May 2001. This actvity is eligible for CME credit through April 2002.

FACULTY DISCLOSURE

In accordance with the disclosure policy of Brown Medical School as well as standards set forth by the Accreditation Council on Continuing Education (ACCME), authorshave been asked to disclose (1) any significant financial or any other relationship with the manufacturers(s) or any commercial products(s) and/or provider(s) of commercial servicesdiscussed in any educational presentation and (2) with any commercial supporters of this activity.

The intent of this policy is not to prevent an author with a potential conflict of interest from making a presentation. It is merely intended that any potential conflict shouldbe identified openly so that the reader may form his/her own opinion.

Anthony A. Caldamone, MD: Grant/Research: Curis, Inc.The following authors have disclosed that they have no commercial relationships to report.Francois I. Luks, MD, PhD, Lewis P. Rubin, MD, MPH, Stephen R. Carr, MD, Arlet G. Kurkchubasche, MD, Michael Plevyak, MD, Lloyd R. Feit, MD

TO OBTAIN CREDIT

To obtain credit, please submit answer grid and $25 fee to Office of Continuing Medical Education, Brown University. Respondents must receive a score of 70 or higher for credit.

CME REGISTRATION FORM PRINT OR TYPE

Name _____________________________________________________________

Address ___________________________________________________________

City, State, Zip _____________________________________________________

Phone ( ) ____________________________________________________

Fax ( ) ___________________________ e-mail _______________________

___Hospital ___Private Practice ___Resident ___Intern ___Other

DEADLINE FOR SUBMISSION

For credit to be received, please mail your registra-tion with $25 fee to Office of Continuing Medical Educa-tion, Brown Medical School, Box G-A2, Providence, RI02912. Submit your answers no later than April 30, 2002.

KEEP A COPY FOR YOUR FILES.Retain a copy of your answers and compare them

with the correct answers, which will be made available uponrequest, and receipt of submission requirements.

EVALUATIONPlease evaluate the effectiveness of the CME activity on a scale of 1 to 5 (1 being poor; 5 being excellent) by circling your choice.

Poor Excellent1. Overall quality of this CME activity 1 2 3 4 52. Content 1 2 3 4 53. Format 1 2 3 4 54. Faculty 1 2 3 4 55. Achievement of educational objectives* Describe the management of the fetus with anemia. 1 2 3 4 5* Describe the work-up and treatment of the fetus with a urinary tract anomaly. 1 2 3 4 5* Describe the role of prenatal counseling in the management of abdominal wall defects. 1 2 3 4 5* Understand the treatment principles of fetal cardiac anomalies. 1 2 3 4 5* Describe the role of invasive fetal intervention in twin-twin transfusion syndrome, and other surgical conditions of the fetus. 1 2 3 4 56. This material was presented without evidence of commercial bias. yes noPlease comment on the impact that this CME activity might have on your practice of medicine.______________________________________________________________________________________________________________________________________________________________________________________________________________

172Medicine and Health / Rhode Island

1. Which of the following criteria is NOT rou-tinely used to diagnose severe twin-twin trans-fusion syndrome (TTTS) in utero?A) Polyhydramnios in one twin and oligo-

hydramnios in the otherB) A difference in hematocrit between the

twins of more than 20%C) Inability to visualize the bladder in the

donor twinD) Tricuspid valve regurgitation in the recipi-

ent twinE) Same gender in both twins

2. Compared with serial amniodrainage,fetoscopic laser occlusion of chorioangiopagusvessels (“FLOC”) to treat twin-twin transfu-sion syndrome:A) Significantly increases the survival of both

twinsB) Significantly increases the survival of the

donor twinC) Significantly decreases the risk of prema-

turityD) Significantly decreases the morbidity of

the surviving twin

3. Which of the following is NOT a known com-plication of twin-twin transfusion syndrome?A) Cerebral palsyB) Intestinal atresiaC) Limb necrosisD) Aortic stenosisE) Ascites

4. Fetal surgery for a congenital anomaly can onlybe justified if:A) The natural history of the disease, if left

untreated, is knownB) Surgical correction after birth would sub-

ject the child to multiple and/or complexoperative interventions

C) It is performed as part of a clinical researchstudy

D) The anomaly can be completely correctedbefore birth

5. Which of these conditions has never been con-sidered an indication for fetal surgery?A) Congenital diaphragmatic herniaB) Esophageal atresiaC) Posterior urethral valvesD) Congenital cystic adenomatoid malfor-

mation of the lungE) Aquaductal stenosis

6. Which of the following fetal urine samples is apredictor of poor renal function at birth?A Osmolarity 180 mOsm/L, Na 100 mEq/LB) Na 80 mEq/l, Cl 70 mEq/LC) Cl 90 mEq/L, β2-microglobulin 4 mg/LD) β2-microglobulin 2 mg/L, osmolarity 210

mOsm/L

7. Which fetus might be a candidate for in uterobladder decompression (vesicoamniotic shunt-ing)?A) A 33–week-old fetus with hydronephro-

sis and normal amniotic fluid volumeB) A 27-week-old fetus with hydronephro-

sis, oligohydramnios and isotonic urineC) A 29-week-old fetus with bladder disten-

sion, oligohydramnios and normal fetalurinary electrolytes

D) A 30-week-old fetus with bladder disten-sion, oligohydramnios and an abnormalkaryotype

8. Currently, the best available treatment for al-pha-thalassemia major (hemoglobin Bart’s dis-ease) is:A) Intrauterine stem cell transplantationB) Intrauterine blood transfusions followed

by bone marrow transplantation in earlychildhood

C) Frequent postnatal blood transfusions andchelating agents

D) There is no treatment available; the dis-ease is uniformly fatal

9. Which statement about alpha-thalassemia iscorrect:A) If two parents each are heterozygous for a

cis-type -globin gene deletion, the risk ofan offspring with -thalassemia major (HbBart’s disease) is 25%

B) If two parents each are heterozygous for atrans-type -globin gene deletion, the riskof an offspring with -thalassemia major(Hb Bart’s disease) is 50%

C) A fetus with -thalassemia major (Hb Bart’sdisease) usually develops hydrops and diesearly in the first trimester

D) alpha-thalassemia major usually manifestsitself in utero, since half the globin chainsin normal fetal hemoglobin are of the type

10.Which statement about digoxin to treat fetalarrhythmias is true:A) Digoxin is the first-line therapy for ven-

tricular tachycardia (VT)B) Digoxin is the first-line therapy for su-

praventricular tachycardia (SVT)C) Digoxin is the first-line therapy for ven-

tricular extrasystole (VES)D) Digoxin is the first-line therapy for atrial

flutterE) B and D are both true

11.Which of the following is NOT currently a treat-ment option for fetal arrhyth ias:A) Maternal plasmapheresisB) Maternal digoxin administrationC) Fetal pacemakerD) Fetal digoxin administrationE) Early delivery

12.Which of the following statements about ab-dominal wall defects is true:A) Preterm delivery for gastroschisis prevents

edema and thickening of the intestinalwall (“peel”) caused by prolonged expo-sure to amniotic fluid

B) Cesarean section is indicated in gastroschi-sis to prevent further damage to the ex-posed intestines

C) Even in the absence of associated anoma-lies, omphalocele carries a poor prognosis

D) Gastroschisis typically occurs in mothersover 30 years of age

E) None of the above

13.Which syndrome is NOT associated withomphalocele:A) Pentalogy of Cantrell (including pericar-

dial, diaphragmatic, pleural and cardiacdefects)

B) OIES complex (including cloacal exstro-phy and imperforate anus)

C) VATER association (including vertebralanomalies, tracheo-esophageal fistula andimperforate anus)

D) Beckwith-Wiedemann syndrome (includ-ing macroglossia and hypoglycemia)

E) Trisomy 18

14.Which of the following may cause rhesus isoim-munization of a Rh-negative mother:A) Induced abortionB) Placental abruptionC) Abdominal traumaD) A and BE) All of the above

15.According to the recommendations of theAmerican College of Obstetrics and Gyne-cology (ACOG), which of the followingpatients may require administration of anti-D immune globulin to prevent Rh-isoim-munization:A) Rh-negative mother with negative Rh-

antibody screen and an Rh-negative fe-tus

B) Rh-negative mother with positive Rh-antibody screen, before amniocentesis

C) Rh-negative mother with negative Rh-antibody screen, within 72 hours ofdelivery

D) Rh-negative mother with positive Rh-an-tibody screen, at 28 weeks gestation

FETAL MEDICINE CME QUESTIONS – CIRCLE THE CORRECT ANSWER.