Ultrasonographic detection of the second-trimester fetus with trisomy 18 and trisomy 21 Norman Ginsberg, MD, Alan Cadkin, MD, Eugene Pergament, MD, PhD, and Yury Verlinsky, PhD Chicago, Illinois Biparietal diameter/femur length ratio and nuchal thickness were found to be sensitive indicators for the prenatal detection of trisomy 18 and trisomy 21. A biparietal diameter/femur length ratio >1.5 SO above the control mean correctly identified 5 of 11 (46%) fetuses with trisomy 21 and 3 of 4 (75%) fetuses with trisomy 18. Nuchal thickening (6 mm or more) correctly identified 5 of 12 (41%) fetuses with trisomy 21 and 2 of 4 (50%) fetuses with trisomy 18. The sensitivity and specificity of the biparietal diameter/femur length ratio in detecting either aneuploidy was 53% and 93%, respectively, whereas a thickened nuchal fold had a sensitivity of 44% and a specificity of 100%. The combined use of the two ultrasonographic measurements had an overall sensitivity of 81% and a specificity of 93%. Prospective ascertainment of these two trisomies appears warranted in low-risk populations. (AM J OSSTET GVNECOL 1990;163:1186-90.)
Key words: Trisomy 18, trisomy 21, ultrasonographic screening, fetus
Prenatal genetic studies to detect chromosomally abnormal conceptions have generally been limited to women of advanced maternal age and those who have experienced a previous conception with a chromosomal aberration. However, it is young mothers with completely normal pregnancy histories who subsequently are delivered of the largest number of chromosomally abnormal newborns. Although their individual risk for such an occurrence is very small, women younger than 34 years of age have >80% of all affected pregnancies. For this reason less than 20% of all trisomy 21 pregnancies could potentially be detected prenatally by means of chorionic villus sampling or amniocentesis.' An important benefit to the screening of pregnancies for open neural tube defects is the recognition that a low concentration of maternal serum u-fetoprotein (AFP) may be associated with trisomy 21.2 Critical assessment of the effectiveness of these screening programs reveals their disappointing low sensitivity (12% to 21 %) as well as high false-positive rate (7% to 9%)."" Recently the combination of maternal serum AFP, (3subunit human chorionic gonadotropin, and serum estriol have shown promise for greater detection of Down syndrome than maternal serum AFP alone!-6 From the Reproductive and Medical Genetics Section, Department of Obstetrics and Gynecology, Illinois Masonic Medical Center. Received for publication August 7, 1989; revised March 9,1990; accepted June 2, 1990. Reprint requests: Norman Ginsberg, MD, 30 N. Michigan Ave., Chicago, 1L 60602. 611122875
1186
It has been reported that biparietal diameter/femur length ratio and nuchal skin thickness measurements can identify 50% to 70% of trisomy 21 pregnancies in the second trimester,7.8 but this has been disputed. 9 . 10 This study confirmed that a combination of both measurements may be the most sensitive, noninvasive approach to the prenatal identification of trisomy 21 and can be successfully applied to the detection of trisomy 18 as well.
Material and methods The ultrasonographic equipment used in this study consisted of an ATL Ultramark-4, Diasonics Wide Vu, and Acuson-128. All transducers were 3.5 mHz with a combination of arc sector and linear array. The study group consisted of 12 cases of trisomy 21 and four cases of trisomy 18 previously diagnosed by amniocentesis. The control group comprised 212 randomly selected cases with a normal karyotype as determined by chorionic villus sampling or amniocentesis. The control group was used to establish a gestational curve for the biparietal diameter/femur length ratio. The control group biparietal diameter/femur length ratio then became the basis for comparison against the trisomic fetuses. The criteria for inclusion into the control group were either a first-trimester ultrasonographic evaluation confirming gestational age (determined by crown-rump length), or a second-trimester ultrasonographic evaluation confirming that the gestational age did not vary from the stated menstrual age by more than 1.5 weeks (determined by biparietal diameter and femur length measurements).'!' '2 All pa-
Volume 163 Kumber 4. Part I
tients who underwent chorionic villus sampling had a similar ultrasonographic evaluation as those undergoing amniocentesis. The biparietal diameter was measured with electronic calipers from leading edge to leading edge at the level of the thalami. The femus was measured in a plane that visualized the greater trochanter to the end of the ossified shaft. When there was a question regarding the exact placement of the markings by the electronic calipers, the acoustical shadow created by the ossified bone was used. For each of the trisomies the number of cases with a biparietal diameter/femur length ratio> 1.5 SDs above the mean of the control for a given gestational age was determined. The selection of 1.5 SD above the mean of the controls was used to confirm the work of Lockwood et al. 7 as well as to check its applicability to trisomy 18. The nuchal thickness was measured at both the level of the occiput, at the midpoint of the cerebellum in the transverse plane, as well as in the dorsal midsagittal plane at the same level, and was considered abnormal if >6 mm. One case of trisomy 21 was not included in the biparietal diameter/femur length analysis because the length of the femur was not measured in the original ultrasonographic evaluation; it was included in the analysis of nuchal thickness. All abnormal cases were independently evaluated from videotapes by two of the authors (N. G. and A. C.) to assure measurement accuracy. Any differences in measurement were reevaluated and based on this review. Two femur length measurements were changed and one previously undiagnosed nuchal thickening was detected. The control group was also reviewed and no changes were made.
Results The control value data are summarized in Table I. A biparietal diameter/femur length ratio> 1.5 SDs above the control population mean was present in five of 11 (46%) fetuses with trisomy 21 and three of four (75%) fetuses with trisomy 18; there were 14 fetuses (6.6%) whose biparietal diameter/femur length ratios exceeded 1.5 SDs from their respective ratio means in the control population. Five of the 12 (41 %) fetuses with trisomy 21 and two of four (50%) fetuses with trisomy 18 had thickened (>6 mm) nuchal folds. There were no cases of occipital nuchal thickening 2:6 mm in the 212 fetuses that comprised the normal population. If the data on the biparietal diameter/femur length ratio are combined with those of fetuses with a thickened nuchal fold nine of 12 (75%) fetuses with trisomy 21 and all four fetuses with trisomy 18 are identified, with a false-positive rate of 6.6%.
Ultrasonographic detection of trisomy 18 and 21 fetuses
1187
Table I. Control value for normal biparietal diameter/femur length ratio Gestational age (wh)
Normal biparietal diameter/femur length ratio (mean + 1.5 SD)
14
1.96 1.83 1.72 1.66 1.65 1.63 1.64
15 16 17
18
19
20
The sensitivity (true-positive results) and specificity (true-negative results) of the two ultrasonographic measurements in the detection of either aneuploidy of chromosome 18 or 21 are summarized in Table II. The sensitivity and specificity of the biparietal diameter/femur length ratio was 53% and 93%, respectively. A thickened nuchal fold had a sensitivity of 44% and a specificity of 100%. The combined use of the two ultrasonographic measurements had an overall sensitivity of 81 % and a specificity of 93%.
Comment The biparietal diameter/femur length ratio and occipital nuchal thickness measurements can be used to detect a significant number of trisomic fetuses. When the biparietal diameterlfemur length ratio with a 1.5 SD cut-off was applied, 75% of the fetuses with trisomy 18 were detected. A nuchal thickening of 2:6 mm was present in 50% of the fetuses with trisomy 18. By combining the use of these ultrasonic measurements, all four of these chromosomally abnormal fetuses were identified between 14 and 17 weeks' gestation. The results of this study demonstrated remarkable agreement with those previously reported by Lockwood et aU and Benacerraf et al. B despite the fact that each study established different normal values for the biparietal diameter / femur length ratio for each week of gestation. This strengthened the recommendation that each investigator needs to establish his or her own set of normal biparietal diameter/femur length ratios, which is dependent on their ultrasonographic equipment and their measurement techniques. 7 The value of nuchal thickening in identifying fetuses with trisomy 21 in the second trimester of pregnancy has been questioned. 9 In this study a nuchal thickening of 2:6 mm was uniformly associated with aneuploidy and, in contrast to the previous report to Toi et al.,9 there were no false positives in the control group. Moreover, the demonstration of nuchal thickening was clearly more indicative of the presence of a chromosomal abnormality than was a biparietal diameter / femur length ratio> 1.5 SDs from the control mean. Mea-
1188
Ginsberg et al.
October 1990 Am J Obstet Gynecol
Fig. 1. Nuchal thickening at the level of the cerebellum. Arrows indicate skin line wrapping around parietal temporal complex.
Table II. Distribution of normal and abnormal biparietal diameter/femur length ratios and nuchal thickness for aneuploid (trisomies 18 and 21) and normal fetuses Sign
Biparietal diameter/femur length Abnormal* Normal Nuchal thickness Abnormalt Normal One or more signs Abnormal Normal
Aneuploid fetuses 8
7 (Sensitivity 53%) 7 9
Normal fetuses 14 198
(Specificity 93%) 0 212
(Sensitivity 44%)
(Specificity 100%)
13 3
14 198
(Sensitivity 81 %)
(Specificity 93%)
*Ratio value> 1.5 SDs from normal mean. tNuchal thickness> 6 mm.
surement was taken in a plane at the level of the occiput, as described by Benacerraf et al. l:l from the outer skin line to the outer edge of the occipital bone. A thickened curvilinear extension of the skin line toward the level of the convexity of the parietal temporal complex (Fig.
1) associated with the abnormal thickening. This curvilinear line should not be confused with the head abutting against the amniotic membrane, particularly when there is an ultrasonographically residual chorioamniotic space. If there was any question as to whether
1190 Ginsberg et al.
a program of ultrasonographic screening and followup amniocentesis in a low-risk population would cost approximately $150,000 to detect one case of trisomy 18 or 21, or about $100,000 less than the cost per case when amniocentesis is performed for advanced maternal age or low maternal serum AFP level. If the approach described in this article were to be widely applied, it might be possible to considerably reduce the number of undiagnosed trisomies before delivery. This approach may be particularly helpful in cases with a low maternal serum AFP level to decide the most appropriate management plan. Prospective ascertainment of trisomies 18 and 21 now appears warranted to determine if ultrasonographic evaluation would effectively predict chromosomally abnormal gestations in low-risk populations. The possibility that trisomy 13 gestations might also be identified by ultrasonographic assessment of biparietal diameter/femur length ratio and nuchal thickness should likewise be investigated.
REFERENCES 1. Adams M, Erickson ], Leyde P, Oakley G. Down's syndrome: recent trend in the United States. ]AMA 1981; 246:758-60. 2. Cuckles HS, Walk N], Lindebaum RH. Maternal serum a-fetoprotein measurement: a screening test for Down's syndrome. Lancet 1984; 1:926-9. 3. Spencer K, Carpenter P. Screening for Down's syndrome using a-fetoprotein: a retrospective study indicating caution. Br Med] 1985;290: 1940-3. 4. Canick ]A, Knight G], Palomaki GE, et al. Low second trimester maternal serum unconjugated oestriol in pregnancies with Down's syndrome. Br ] Obstet Gynaecol 1988;95:330-3. 5. Petrocik E, Wassman ER, Kelly]C. Prenatal screening for
October 1990 Am J Obstet Gynecol
6.
7. 8. 9. 10.
11.
12. 13. 14. 15. 16.
17.
Down syndrome with maternal serum human chorionic gonadotropin levels. AM ] OBSTET GYNECOL 1989; 161: 1168-73. Bogart MH, Pandian MR,]ones OW. Abnormal maternal serum chorionic gonadotropin levels in pregnancies with fetal chromosome abnormalities. Pernat Diagn 1987; 7:623-30. Lockwood C, Benacerraf B, Krinsky A, et al. A sonographic screening method for Down syndrome. AM] OBSTET GYNECOL 1987;157:803-8. Benacerraf BR, Gelman R, Frigoletto FD. Sonographic identification of second trimester fetuses with Down's syndrome. N Engl] Med 1987;317:1371-5. Toi A, Simpson GF, Filly RA. Ultrasonically evident fetal nuchal skin thickening; is it specific for Down syndrome? AM] OB5TET GYNECOL 1987;156:150-3. Lynch L, Berkowitz GS, Chitkara U, Wilkins lA, Megakej KE, Berkowitz RL. Ultrasound detection of Down syndrome: is it really possible? Obstet Gynecol 1989;73:26770. Sabbagha RE, Barton BA, Barton BF, Kingas E, Orgill], Turner ]H. Sonar biparietal diameter. II. Predictive of three fetal growth patterns leading to a closer assessment of gestational age and neonatal weight. AM] OBSTET GyNECOL 1976; 126:485-90. ]eanty P, Rodesch F, Delbeke D, Dumont ]E. Estimation of gestational age from measurements of fetal long bones. ] Ultrasound Med 1984;3(suppl 2):75-9. Benacerraf B, Frigoletto F, Cramer D. Down syndrome: sonographic sign for diagnosis in the second-trimester fetus. Radiology 1987;163:81-3. Elias S, Annas GJ. Routine prenatal genetic screening. N Engl] Med 1987;317:1409-10. Hook EB, Cross PK. Schreinemacher DM. Chromosonal abnormality rates at amniocentesis and in live-born infants. ]AMA 1983;249:2034-8. Hook EB. Chromosome abnormalities and spontaneous fetal death following amniocentesis: further data and association with maternal age. Am] Hum Genet 1983; 35:110-6. Merkatz IR, Nitowsky HM, Macri ]N, Johnson WE. An association between low maternal serum alphafetoprotein and fetal chromosomal abnormalities. AM] OB5TET GYNECOL 1984; 148:886-94.
Volume 163 Number 4, Part 1
Ultrasonographic detection of trisomy 18 and 21 fetuses
1189
Fig. 2. Nuchal thickening in the dorsal midsagittal plane. Arrows point to skin line. Hypoechoic area represents areolar edema.
the measured thickening was actually present, the fetus was examined in the dorsal midsagittal plane (Fig. 2). When the angle of the transducer was caudal to a line drawn from the lower margin of the orbit to the auricular point and extended to the center of the occipital bone (Reid's baseline), this plane ruled out any falsepositive findings. The fetus, when viewed in the dorsal midsagittal plane with high-resolution ultrasonographic equipment, usually demonstrated a double line in the region of the nuchal thickening. This skin line was a hypoechoic stripe that represented edematous areolar tissue; below this the muscles of the neck were seen. This hypoechoic area added to the thickening in the nuchal region. With the head in normal flexion, the skin bulges at the occiput even if the double line cannot be seen in the dorsal midsagittal plane. It has been suggested that all pregnant women be screened during the second trimester for trisomy 21," although caution has been urged in making such a practice widespread. 11 The positive predictive value of biparietal diameter/femur length ratio and nuchal thickness is an estimate of the probability that a fetus identified by these
ultrasonic measurements as having a high risk of trisomy 21 or 18 actually is aneuploid for one of these chromosomes. The general population incidence for trisomy 21 and 18 at term is 1 per 1000 and 1 per 7407, respectively, whereas the fetal loss rate between the second trimester and term is 29% and 67%, respectively.15. 16 The resulting positive predictive value in this study would be 1147. Thus 46 amniocenteses on chromosomally normal gestations would be necessary for each diagnosis of either trisomy 18 or 21. Amniocentesis currently is performed for advanced maternal age when the risk of a live-born infant with trisomy 21 is 1/38515 and for a low maternal serum AFP value when the risk of a second-trimester gestation with trisomy 21 is 1 per 250.17 Therefore about one eighth to one third as many amniocenteses generated from ultrasonographic screening evaluations are required to detect either trisomy 18 or 21 compared with advanced maternal age and low maternal serum AFP level. Ultrasonography is noninvasive and widely available in the United States; such screening would not require extensive technical training or the use of overly expensive ultrasonographic equipment. Finally, we estimate that
Key words: Trisomy 18, trisomy 21, ultrasonographic screening, fetus
Prenatal genetic studies to detect chromosomally abnormal conceptions have generally been limited to women of advanced maternal age and those who have experienced a previous conception with a chromosomal aberration. However, it is young mothers with completely normal pregnancy histories who subsequently are delivered of the largest number of chromosomally abnormal newborns. Although their individual risk for such an occurrence is very small, women younger than 34 years of age have >80% of all affected pregnancies. For this reason less than 20% of all trisomy 21 pregnancies could potentially be detected prenatally by means of chorionic villus sampling or amniocentesis.' An important benefit to the screening of pregnancies for open neural tube defects is the recognition that a low concentration of maternal serum u-fetoprotein (AFP) may be associated with trisomy 21.2 Critical assessment of the effectiveness of these screening programs reveals their disappointing low sensitivity (12% to 21 %) as well as high false-positive rate (7% to 9%)."" Recently the combination of maternal serum AFP, (3subunit human chorionic gonadotropin, and serum estriol have shown promise for greater detection of Down syndrome than maternal serum AFP alone!-6 From the Reproductive and Medical Genetics Section, Department of Obstetrics and Gynecology, Illinois Masonic Medical Center. Received for publication August 7, 1989; revised March 9,1990; accepted June 2, 1990. Reprint requests: Norman Ginsberg, MD, 30 N. Michigan Ave., Chicago, 1L 60602. 611122875
1186
It has been reported that biparietal diameter/femur length ratio and nuchal skin thickness measurements can identify 50% to 70% of trisomy 21 pregnancies in the second trimester,7.8 but this has been disputed. 9 . 10 This study confirmed that a combination of both measurements may be the most sensitive, noninvasive approach to the prenatal identification of trisomy 21 and can be successfully applied to the detection of trisomy 18 as well.
Material and methods The ultrasonographic equipment used in this study consisted of an ATL Ultramark-4, Diasonics Wide Vu, and Acuson-128. All transducers were 3.5 mHz with a combination of arc sector and linear array. The study group consisted of 12 cases of trisomy 21 and four cases of trisomy 18 previously diagnosed by amniocentesis. The control group comprised 212 randomly selected cases with a normal karyotype as determined by chorionic villus sampling or amniocentesis. The control group was used to establish a gestational curve for the biparietal diameter/femur length ratio. The control group biparietal diameter/femur length ratio then became the basis for comparison against the trisomic fetuses. The criteria for inclusion into the control group were either a first-trimester ultrasonographic evaluation confirming gestational age (determined by crown-rump length), or a second-trimester ultrasonographic evaluation confirming that the gestational age did not vary from the stated menstrual age by more than 1.5 weeks (determined by biparietal diameter and femur length measurements).'!' '2 All pa-
Volume 163 Kumber 4. Part I
tients who underwent chorionic villus sampling had a similar ultrasonographic evaluation as those undergoing amniocentesis. The biparietal diameter was measured with electronic calipers from leading edge to leading edge at the level of the thalami. The femus was measured in a plane that visualized the greater trochanter to the end of the ossified shaft. When there was a question regarding the exact placement of the markings by the electronic calipers, the acoustical shadow created by the ossified bone was used. For each of the trisomies the number of cases with a biparietal diameter/femur length ratio> 1.5 SDs above the mean of the control for a given gestational age was determined. The selection of 1.5 SD above the mean of the controls was used to confirm the work of Lockwood et al. 7 as well as to check its applicability to trisomy 18. The nuchal thickness was measured at both the level of the occiput, at the midpoint of the cerebellum in the transverse plane, as well as in the dorsal midsagittal plane at the same level, and was considered abnormal if >6 mm. One case of trisomy 21 was not included in the biparietal diameter/femur length analysis because the length of the femur was not measured in the original ultrasonographic evaluation; it was included in the analysis of nuchal thickness. All abnormal cases were independently evaluated from videotapes by two of the authors (N. G. and A. C.) to assure measurement accuracy. Any differences in measurement were reevaluated and based on this review. Two femur length measurements were changed and one previously undiagnosed nuchal thickening was detected. The control group was also reviewed and no changes were made.
Results The control value data are summarized in Table I. A biparietal diameter/femur length ratio> 1.5 SDs above the control population mean was present in five of 11 (46%) fetuses with trisomy 21 and three of four (75%) fetuses with trisomy 18; there were 14 fetuses (6.6%) whose biparietal diameter/femur length ratios exceeded 1.5 SDs from their respective ratio means in the control population. Five of the 12 (41 %) fetuses with trisomy 21 and two of four (50%) fetuses with trisomy 18 had thickened (>6 mm) nuchal folds. There were no cases of occipital nuchal thickening 2:6 mm in the 212 fetuses that comprised the normal population. If the data on the biparietal diameter/femur length ratio are combined with those of fetuses with a thickened nuchal fold nine of 12 (75%) fetuses with trisomy 21 and all four fetuses with trisomy 18 are identified, with a false-positive rate of 6.6%.
Ultrasonographic detection of trisomy 18 and 21 fetuses
1187
Table I. Control value for normal biparietal diameter/femur length ratio Gestational age (wh)
Normal biparietal diameter/femur length ratio (mean + 1.5 SD)
14
1.96 1.83 1.72 1.66 1.65 1.63 1.64
15 16 17
18
19
20
The sensitivity (true-positive results) and specificity (true-negative results) of the two ultrasonographic measurements in the detection of either aneuploidy of chromosome 18 or 21 are summarized in Table II. The sensitivity and specificity of the biparietal diameter/femur length ratio was 53% and 93%, respectively. A thickened nuchal fold had a sensitivity of 44% and a specificity of 100%. The combined use of the two ultrasonographic measurements had an overall sensitivity of 81 % and a specificity of 93%.
Comment The biparietal diameter/femur length ratio and occipital nuchal thickness measurements can be used to detect a significant number of trisomic fetuses. When the biparietal diameterlfemur length ratio with a 1.5 SD cut-off was applied, 75% of the fetuses with trisomy 18 were detected. A nuchal thickening of 2:6 mm was present in 50% of the fetuses with trisomy 18. By combining the use of these ultrasonic measurements, all four of these chromosomally abnormal fetuses were identified between 14 and 17 weeks' gestation. The results of this study demonstrated remarkable agreement with those previously reported by Lockwood et aU and Benacerraf et al. B despite the fact that each study established different normal values for the biparietal diameter / femur length ratio for each week of gestation. This strengthened the recommendation that each investigator needs to establish his or her own set of normal biparietal diameter/femur length ratios, which is dependent on their ultrasonographic equipment and their measurement techniques. 7 The value of nuchal thickening in identifying fetuses with trisomy 21 in the second trimester of pregnancy has been questioned. 9 In this study a nuchal thickening of 2:6 mm was uniformly associated with aneuploidy and, in contrast to the previous report to Toi et al.,9 there were no false positives in the control group. Moreover, the demonstration of nuchal thickening was clearly more indicative of the presence of a chromosomal abnormality than was a biparietal diameter / femur length ratio> 1.5 SDs from the control mean. Mea-
1188
Ginsberg et al.
October 1990 Am J Obstet Gynecol
Fig. 1. Nuchal thickening at the level of the cerebellum. Arrows indicate skin line wrapping around parietal temporal complex.
Table II. Distribution of normal and abnormal biparietal diameter/femur length ratios and nuchal thickness for aneuploid (trisomies 18 and 21) and normal fetuses Sign
Biparietal diameter/femur length Abnormal* Normal Nuchal thickness Abnormalt Normal One or more signs Abnormal Normal
Aneuploid fetuses 8
7 (Sensitivity 53%) 7 9
Normal fetuses 14 198
(Specificity 93%) 0 212
(Sensitivity 44%)
(Specificity 100%)
13 3
14 198
(Sensitivity 81 %)
(Specificity 93%)
*Ratio value> 1.5 SDs from normal mean. tNuchal thickness> 6 mm.
surement was taken in a plane at the level of the occiput, as described by Benacerraf et al. l:l from the outer skin line to the outer edge of the occipital bone. A thickened curvilinear extension of the skin line toward the level of the convexity of the parietal temporal complex (Fig.
1) associated with the abnormal thickening. This curvilinear line should not be confused with the head abutting against the amniotic membrane, particularly when there is an ultrasonographically residual chorioamniotic space. If there was any question as to whether
1190 Ginsberg et al.
a program of ultrasonographic screening and followup amniocentesis in a low-risk population would cost approximately $150,000 to detect one case of trisomy 18 or 21, or about $100,000 less than the cost per case when amniocentesis is performed for advanced maternal age or low maternal serum AFP level. If the approach described in this article were to be widely applied, it might be possible to considerably reduce the number of undiagnosed trisomies before delivery. This approach may be particularly helpful in cases with a low maternal serum AFP level to decide the most appropriate management plan. Prospective ascertainment of trisomies 18 and 21 now appears warranted to determine if ultrasonographic evaluation would effectively predict chromosomally abnormal gestations in low-risk populations. The possibility that trisomy 13 gestations might also be identified by ultrasonographic assessment of biparietal diameter/femur length ratio and nuchal thickness should likewise be investigated.
REFERENCES 1. Adams M, Erickson ], Leyde P, Oakley G. Down's syndrome: recent trend in the United States. ]AMA 1981; 246:758-60. 2. Cuckles HS, Walk N], Lindebaum RH. Maternal serum a-fetoprotein measurement: a screening test for Down's syndrome. Lancet 1984; 1:926-9. 3. Spencer K, Carpenter P. Screening for Down's syndrome using a-fetoprotein: a retrospective study indicating caution. Br Med] 1985;290: 1940-3. 4. Canick ]A, Knight G], Palomaki GE, et al. Low second trimester maternal serum unconjugated oestriol in pregnancies with Down's syndrome. Br ] Obstet Gynaecol 1988;95:330-3. 5. Petrocik E, Wassman ER, Kelly]C. Prenatal screening for
October 1990 Am J Obstet Gynecol
6.
7. 8. 9. 10.
11.
12. 13. 14. 15. 16.
17.
Down syndrome with maternal serum human chorionic gonadotropin levels. AM ] OBSTET GYNECOL 1989; 161: 1168-73. Bogart MH, Pandian MR,]ones OW. Abnormal maternal serum chorionic gonadotropin levels in pregnancies with fetal chromosome abnormalities. Pernat Diagn 1987; 7:623-30. Lockwood C, Benacerraf B, Krinsky A, et al. A sonographic screening method for Down syndrome. AM] OBSTET GYNECOL 1987;157:803-8. Benacerraf BR, Gelman R, Frigoletto FD. Sonographic identification of second trimester fetuses with Down's syndrome. N Engl] Med 1987;317:1371-5. Toi A, Simpson GF, Filly RA. Ultrasonically evident fetal nuchal skin thickening; is it specific for Down syndrome? AM] OB5TET GYNECOL 1987;156:150-3. Lynch L, Berkowitz GS, Chitkara U, Wilkins lA, Megakej KE, Berkowitz RL. Ultrasound detection of Down syndrome: is it really possible? Obstet Gynecol 1989;73:26770. Sabbagha RE, Barton BA, Barton BF, Kingas E, Orgill], Turner ]H. Sonar biparietal diameter. II. Predictive of three fetal growth patterns leading to a closer assessment of gestational age and neonatal weight. AM] OBSTET GyNECOL 1976; 126:485-90. ]eanty P, Rodesch F, Delbeke D, Dumont ]E. Estimation of gestational age from measurements of fetal long bones. ] Ultrasound Med 1984;3(suppl 2):75-9. Benacerraf B, Frigoletto F, Cramer D. Down syndrome: sonographic sign for diagnosis in the second-trimester fetus. Radiology 1987;163:81-3. Elias S, Annas GJ. Routine prenatal genetic screening. N Engl] Med 1987;317:1409-10. Hook EB, Cross PK. Schreinemacher DM. Chromosonal abnormality rates at amniocentesis and in live-born infants. ]AMA 1983;249:2034-8. Hook EB. Chromosome abnormalities and spontaneous fetal death following amniocentesis: further data and association with maternal age. Am] Hum Genet 1983; 35:110-6. Merkatz IR, Nitowsky HM, Macri ]N, Johnson WE. An association between low maternal serum alphafetoprotein and fetal chromosomal abnormalities. AM] OB5TET GYNECOL 1984; 148:886-94.
Volume 163 Number 4, Part 1
Ultrasonographic detection of trisomy 18 and 21 fetuses
1189
Fig. 2. Nuchal thickening in the dorsal midsagittal plane. Arrows point to skin line. Hypoechoic area represents areolar edema.
the measured thickening was actually present, the fetus was examined in the dorsal midsagittal plane (Fig. 2). When the angle of the transducer was caudal to a line drawn from the lower margin of the orbit to the auricular point and extended to the center of the occipital bone (Reid's baseline), this plane ruled out any falsepositive findings. The fetus, when viewed in the dorsal midsagittal plane with high-resolution ultrasonographic equipment, usually demonstrated a double line in the region of the nuchal thickening. This skin line was a hypoechoic stripe that represented edematous areolar tissue; below this the muscles of the neck were seen. This hypoechoic area added to the thickening in the nuchal region. With the head in normal flexion, the skin bulges at the occiput even if the double line cannot be seen in the dorsal midsagittal plane. It has been suggested that all pregnant women be screened during the second trimester for trisomy 21," although caution has been urged in making such a practice widespread. 11 The positive predictive value of biparietal diameter/femur length ratio and nuchal thickness is an estimate of the probability that a fetus identified by these
ultrasonic measurements as having a high risk of trisomy 21 or 18 actually is aneuploid for one of these chromosomes. The general population incidence for trisomy 21 and 18 at term is 1 per 1000 and 1 per 7407, respectively, whereas the fetal loss rate between the second trimester and term is 29% and 67%, respectively.15. 16 The resulting positive predictive value in this study would be 1147. Thus 46 amniocenteses on chromosomally normal gestations would be necessary for each diagnosis of either trisomy 18 or 21. Amniocentesis currently is performed for advanced maternal age when the risk of a live-born infant with trisomy 21 is 1/38515 and for a low maternal serum AFP value when the risk of a second-trimester gestation with trisomy 21 is 1 per 250.17 Therefore about one eighth to one third as many amniocenteses generated from ultrasonographic screening evaluations are required to detect either trisomy 18 or 21 compared with advanced maternal age and low maternal serum AFP level. Ultrasonography is noninvasive and widely available in the United States; such screening would not require extensive technical training or the use of overly expensive ultrasonographic equipment. Finally, we estimate that
