ANTENATAL ULTRASOUND FOR FETAL ANOMALIES: Importance of Perinatal Autopsy

Susan Shen-Schwarz, MB, BS, Carol Neish, BS, and Lyndon M. Hill, MD u Division of Perinatal Pathology, Department of Pathology, and Division of Ultrasound, Department of Obstetrics and Gynecology; Magee-Women's Hospital and University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15213

Sixty-one instances of congenital anomalies identified prenatally by ultrasound were reviewed to determine whether autopsy provided important additional information. An important finding was defined as one which would affect: I ) genetic counseling; 2) diagnosis of a syndrome; 3) determination of etiology or pathogenetic mechanism of the anomaly,. or 4) interpretation of severity of the anomalies. In 28 cases (46%), post-mortem examination provided such information. All of these infants had multiple anomalies; correlations with oh'gohydramnios and poor fetal outcome were noted. Autopsy provided no additional meaningful information in 30 cases (49%), the majority (77'70) Df whom had isolated anomalies. In 3 cases (5 %), due to tissue autolysis, autopsy provided less information than the previous ultrasound. Although most fetal anomalies are readily diagnosed by ultrasound, we found that postmortem examination is still necessary: 1) to conjrm a prenatal diagnosis; 2) to delineate multiple anomalies; 3) when the ultrasound examination is limited by olcfohydramnios; and 4) to obtain tissue for microscopic examination, cytogenetic and biochemical analysis, if these studies have not been performed prenatally. KEY WORDS: autopsy, developmental anomalies, prenatal ultrasound, prenatal diagnosis.

INTRODUCTION Accurate prenatal diagnosis of fetal anomalies is essential for making decisions on pregnancy management options and for the determination of recurrence risks. ',' Traditionally, the autopsy has been the ultimate method by which fetal anomalies are detected and studied. Increasingly prenatal ultrasound has been utilized for the same purpose, and in some instances has replaced the autopsy in this capacity. Few s t ~ d i e have s ~ ~addressed ~~~ the overall accuracy of prenatal ultrasound Presented at the Society for Pediatric Pathology Annual Meeting in Washington, D.C., February 29, 1988, where it received the award for the best presentation by a trainee (C.N.). Lab Invest 1988;58:6P. Address reprint requests to Susan Shen-Schwarz, M.D., Department of Pathology, Children's Memorial Hospital, 2300 Children's Plaza, Chicago, Illinois 60614.

Pediatric Pathology, 9: 1-9, 198.9 Copyrip.ht @ I989 by Hemisphere Publishing Corporation

1

2

S. SHEN-SCHWARZ E l AL.

for the diagnosis of fetal anomalies. The purpose of this study is to determine the role of perinatal autopsy in cases of fetal anomalies previously evaluated by a level I1 ultrasound examination.

MATERIALS AND METHODS From the 1986 perinatal autopsy records at Magee-Womens Hospital (MWH), we identified 61 cases of developmental anomalies which had had a targeted (level 11) ultrasound examination in the antenatal period, Fetuses with chromosomal aberration who were studied prenatally but who did not have structural anomalies, were excluded from this review. The majority of the cases (51/61) were examined in the division of ultrasound M W H under the supervision of a perinatologist with expertise in ultrasonography. Patients wcre scanned in a semi-recumbent position with a stateof-thc-art real-time unit (GE 2800, R T 3000, General Electric Corporation, Rancho Cordova, California), utilizing either a 3.5- or 5.0-mHz transducer. The fetal lie, assessment of gestational age, amniotic fluid volume, and placcntal site were determined. A systematic organ review was performed in order to rule out any sonographically detectable abnormalities. The indications for Level I1 antenatal ultrasound were: elevated maternal serum alphafetoprotein, intrauterine fetal death, abnormal findings at routine screening for assessment of' gestational age, advanccd maternal age, and others. In a small number of cases, serial ultrasound studies werc performed to follow progression of fetal anomaly. The remaining cases (7/61) were performed in diagnostic ultrasound facilities outside the hospital. In our institution, termination of pregnancy after the diagnosis of a severe congenital anomay was by intraamniotic saline instillation to ensure fetal demise. In cases of antepartum fetal death, vaginaljrectal prostaglandin was usc-d to induce labor. A complete post-mortem examination was performed in all cases, sometimes without the knowledge of prenatal ultrasound findings. If the result of' antenatal karyotyping was not available because amniocentesis was not performed or was unsuccessful, or because amniotic cell growth was still uncertain, then placental chorionic tissue, fetal lung, kidney, and spleen were submitted for chromosomal analysis. Final diagnosis of the fetal anomalies was made using the information obtained from photography, radiographic studies, gross and microscopic descriptions, and whenever possible, in concert with out reproductive geneticists who provided a detailed family history, cytogenetic and/or biochemical analysis, and access to a computerized database for syndrome identification (Birth Defect Information Service). In this retrospective study, each case was reviewed to determine whether additional important information was gained from the autopsy. Important

PRENATAL ULTRASOUND AND AUTOPSY

3

information was defined as those findings that would affect one or more of the following: 1) genetic counselling; 2) diagnosis of a syndrome; 3) determination of etiology or pathogenetic mechanism of the anomaly; 4) interpretation of severity of the anomaly. Each case was classified into one of three categories according to the following criteria: Class A: autopsy provided important information as defined above. Class B: autopsy confirmed ultrasound diagnosis but no important information was obtained. Class C : autopsy provided less information than the ultrasound findings, and could neither confirm nor refute the antenatal diagnosis. The following variables were examined to determine whether they influence the accuracy of prenatal ultrasound: location of prenatal ultrasound ( M W H or elsewhere), gestational age, fetal viability and amount of amniotic fluid at time of prenatal ultrasound, as well as the mode of death, gestational age at autopsy and type of anomalies (isolated versus multiple). Statistical analysis used to test these variables included Fisher’s exact test and chi-square with Yates correlation. A probability (f) value of