Noninvasive Prenatal Testing: Impact on Genetic Counseling, Invasive Prenatal Diagnosis, and Trisomy 21 Detection Joseph R. Wax, MD,1 Angelina Cartin,1 Ren ee Chard, MSc,1 F. Lee Lucas, PhD,2 Michael G. Pinette, MD1 1
Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Maine Medical Center, Portland, Maine 04102 2 Center for Outcomes Research and Evaluation, Maine Medical Center, Portland, Maine 04102 Received 23 July 2014; accepted 6 September 2014
ABSTRACT: Purpose. The aim of this study was to compare rates of genetic counseling, invasive prenatal diagnosis, and trisomy 21 detection among women at increased risk for aneuploidy, before versus after the availability of noninvasive prenatal testing (NIPT). Methods. This institutional review board–exempt retrospective study included all women who had an ultrasound (US) examination between 10 0/7 and 21 6/7 weeks’ gestation and were eligible for NIPT (ie, age 35 years, US findings suggestive of increased aneuploidy risk, positive aneuploidy screen, prior trisomic fetus, parental balanced translocation with increased risk for trisomy 13 or 21) between June 1, 2012 and February 1, 2013. NIPT was performed by a single laboratory after patients received genetic counseling. We also identified a comparison group of women evaluated between December 1, 2010 and November 30, 2011, who would have been eligible for NIPT had it been available. The two groups were compared for maternal demographics, aneuploidy risk factors, rates of genetic counseling, invasive diagnostic procedures, and trisomy 21 detection. Results. The before-NIPT and after-NIPT groups contained 1,464 and 1,046 subjects, respectively. All 33 fetuses with trisomy 21 in the two groups were identified by positive aneuploidy screening. After the introduction of NIPT, genetic counseling for aneuploidy risk increased (adjusted odds ratio [aOR], 1.77 [1.49–2.11]; p < 0.0001) and the overall invasive diagnosis (aOR, 0.42 [0.32–0.55]; p < 0.0001), including amniocentesis (aOR, 0.37 [0.27–0.52], p < 0.0001), decreased, whereas the prenatal diagnosis of trisomy 21 remained similar (88% versus 100%; p 5 0.86). Correspondence to: J.R. Wax C 2014 Wiley Periodicals, Inc. V
VOL. 43, NO. 1, JANUARY 2015
Conclusions. NIPT in clinical practice uses more genetic counseling resources but requires significantly fewer invasive procedures to maintain the detection C 2014 Wiley Periodicals, Inc. rates of trisomy 21. V J Clin Ultrasound 43:1–6, 2015; Published online in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jcu.22243 Keywords: aneuploidy screening; cell-free fetal DNA; Down syndrome; trisomy 21; ultrasonography
INTRODUCTION
N
oninvasive prenatal testing (NIPT) analyzes cell-free fetal DNA in the maternal circulation to provide highly efficient screening for trisomy 21 in pregnancies at increased risk for fetal aneuploidy.1–4 NIPT offers higher detection and lower screen-positive rates than existing aneuploidy screening tests do, and it should reduce the invasive diagnosis rate while maintaining or increasing trisomy 21 detection. Two studies demonstrated fewer invasive diagnostic procedures following the clinical introduction of NIPT. However, neither report evaluated the effect on trisomy 21 detection.5,6 Although NIPT has been validated by different research methods, experience suggests that new aneuploidyscreening methods may not demonstrate performance that is as robust as expected when introduced into clinical practice.7 Several professional organizations have published recommendations on the clinical use of NIPT, including patient selection and counseling.8–11 These groups uniformly call for pretest counseling by a genetic counselor or other 1
WAX ET AL
trained individual, with genetic counseling available to any screen-positive patient. The effect of these recommendations on genetic counseling resources remains unknown. Therefore, we performed a retrospective cohort study to evaluate whether clinically implementing NIPT increases genetic counseling utilization, reduces invasive diagnostic procedures (amniocentesis or chorionic villus sampling), or affects trisomy 21 detection. PATIENTS AND METHODS
This institutional review board–exempt study included all women with a singleton pregnancy seen in a single maternal–fetal medicine practice for an ultrasound (US) examination between 10 0/7 and 21 6/7 weeks’ gestation from June 1, 2012 to February 1, 2013, and who were eligible for NIPT according to current guidelines. NIPT was available to all women regardless of payor status. Women were considered to be screen-positive for increased aneuploidy risk if they had exhibited any NIPT eligibility criteria, which included maternal age 35 years at delivery, positive serum aneuploidy screen, US findings suggestive of increased aneuploidy risk, prior pregnancy with a trisomic fetus, or parental balanced translocation with increased risk for fetal trisomy 13 or 21.9–11 US findings considered suggestive of aneuploidy included first-trimester nuchal translucency 3.5 mm or cystic hygroma colli. Secondtrimester features included nuchal skin-fold thickness 6 mm, absent nasal bone, short humerus less than the fifth percentile for gestational age, short femur less than the fifth percentile for gestational age, echogenic intracardiac focus, echogenic bowel, double-bubble sign, congenital heart defect, or other major congenital anomaly. All US examinations were performed by registered diagnostic medical sonographers in an American Institute of Ultrasound in Medicine– accredited practice, and results were interpreted by one of two maternal–fetal medicine physicians. NIPT was ordered and performed by a single commercial laboratory using massively parallel shotgun sequencing. All patients in both study periods had been offered nondirective counseling by a certified genetic counselor employed by our practice for increased aneuploidy risk. However, genetic counseling was a prerequisite to NIPT in our practice, consistent with current guidelines.8–11 We identified a comparison group of women with singleton pregnancies seen in our office for US examination between 10 0/7 and 21 6/7 2
weeks’ gestation from December 1, 2010 through November 30, 2011, who would have been eligible for NIPT had it been available. The two groups were compared for maternal demographics, aneuploidy risk factors, receipt of genetic counseling for aneuploidy risk, amniocentesis, chorionic villus sampling, and population-based trisomy 21 detection rates by screening tests and invasive prenatal diagnosis. Other than the introduction of NIPT, no practice changes were introduced during the study period. Data were abstracted from several prospectively ascertained databases, including (1) genetic counseling, (2) prenatal screening, (3) US, and (4) prenatal and postnatal genetic diagnosis. To ensure complete trisomy 21 ascertainment, all subjects were cross-referenced to pediatrics genetics records that included all live births with trisomy 21 and to the fetal–neonatal death database. Among fetal and neonatal deaths without chromosomal analysis, a normal autopsy or physical examination served as a proxy for a normal karyotype. Categorical data, including rates of trisomy 21 detection by invasive prenatal diagnosis, were compared using the v2 or Fisher’s exact test. Continuous variables were compared using the two-tailed unpaired t test. We used logistic regression analysis to test the effect of NIPT availability on outcomes, controlling for potential confounders. Statistical significance was set at p < 0.05. RESULTS
A total of 1,464 subjects before and 1,046 subjects after NIPT availability were included. Figure 1 is a flow diagram for the 1,046 subjects after NIPT availability. There were no significant differences in the groups’ demographics. The after-NIPT group had significantly fewer positive serum-based aneuploidy screening tests than the before-NIPT group did (115 versus 223; p 5 0.002). However, 100 of 166 women (60.2%) who underwent NIPT had NIPT as a first-line aneuploidy screen without a prior serum-based test. The frequency of other positive aneuploidy screening tests, which also comprised NIPT eligibility criteria, was similar by group (Table 1). US findings associated with increased aneuploidy risk were not significantly different by study group, except for increased frequencies of short femur in the before-NIPT groups and thickened nuchal skin fold in the after-NIPT groups (Table 2). JOURNAL OF CLINICAL ULTRASOUND
NIPT IN CLINICAL PRACTICE
FIGURE 1. Flow diagram illustrates the distribution of noninvasive prenatal diagnosis (NIPT)-eligible patients when NIPT was available. IUFD, intrauterine fetal death; TOP, termination of pregnancy; CHD, congenital heart disease; SAB, spontaneous abortion.
TABLE 1 Patients’ Characteristics
Characteristic
TABLE 2 Ultrasound Findings Suggesting Increased Risk of Aneuploidy
Before NIPT,
After NIPT,
n 5 1,464
n 5 1,046
p Value
Ultrasound Finding Maternal age, years (mean 6 SD) Parity, n 0 1 Unknown NIPT eligibility risk factors*, n (%) Maternal age 35 years US Positive screen Prior trisomy Parental translocation
34.5 6 5.9
458 988 20
34.6 6 5.5
354 678 14
After NIPT
n 5 1,464 (%)
n 5 1,046 (%)
p Value
391 (26.7)
280 (26.8)
1.0
12 (0.8)
6 (0.6)
0.51
19 (1.3)
7 (0.7)
0.14
7 (0.5) 8 (0.5) 272 (18.6)
3 (0.3) 13 (1.2) 215 (20.1)
0.51 0.05 0.11
33 (2.3) 0 (0) 26 (1.8) 66 (4.5) 13 (0.9) 5 (0.3)
19 (1.8) 2 (0.2) 11 (1.1) 45 (4.3) 2 (0.2) 1 (0.1)
0.51 0.17 0.16 1.0 0.03 0.26
0.68
0.17
1,023 (69.8)
742 (70.9)
0.53
391 (26.7) 223 (15.2) 25 (1.7) 0 (0)
280 (26.8) 115 (11.0) 15 (1.4) 1 (0.1)
1.0 0.002 0.59 0.42
Abbreviations: NIPT, noninvasive prenatal testing; US, ultrasound. *Data add up to more than the total number of patients in each group because some patients had more than one indication.
All known fetuses with trisomy 21, 25 in the before-NIPT and 8 in the after-NIPT groups (prevalences 1.7% and 0.8%; p 5 0.04) were detected by positive aneuploidy screening on the basis of history, serum tests, or US findings. Ninety-five percent confidence intervals for 100% trisomy 21 detection before NIPT, after NIPT, and overall were 87–100%, 66–100%, and 90–100%, respectively. In the before-NIPT and after-NIPT groups, trisomy 21 was prenatally diagnosed in 88% and 100% of cases by amniocentesis (n 5 10 versus n 5 3) and chorionic villus sampling (n 5 12 versus n 5 5) and postnatally by neonatal blood in 12% (n 5 3 verVOL. 43, NO. 1, JANUARY 2015
Before NIPT
Ultrasound finding present, n First trimester, n Nuchal translucency 3.5 mm Cystic hygroma Second trimester, n* Absent nasal bone Nuchal fold 6 mm Echogenic intracardiac focus Cardiac anomaly Double bubble Echogenic bowel Other anomaly Short femur Short humerus
Abbreviation: NIPT, noninvasive prenatal testing. *Data add up to more than the total number of patients in each group because some patients had more than one finding.
sus n 5 0); p 5 0.86. Significantly more women underwent amniocentesis after a positive serum-based screen before versus after NIPT: 62 of 152 (40.8%) versus 5 of 56 (8.9%); p < 0.0001. Likewise, significantly more women had chorionic villus sampling after a positive serum-based screen before versus after NIPT: 17 of 85 (20.0%) versus 1 of 50 (2.0%); p 5 0.003. NIPT was accepted by 166 (15.9%) eligible subjects, including 54 of 316 (17.1%) with public 3
WAX ET AL TABLE 3 Genetic Counseling by Indication, Before versus After NIPT Before NIPT (n 5 1,464) Indication* Age 35 years US finding Positive screen Prior trisomy Prior translocation
After NIPT (n 5 1,046)
Counseling
No Counseling
Counseling
No Counseling
p Value
385 (37.6) 165 (42.2) 208 (93.3) 14 (56.0) 0
638 (62.3) 226 (57.8) 15 (6.7) 11 (44.0) 0
249 (33.6) 198 (70.7) 99 (86.1) 7 (46.7) 1 (100)
493 (66.4) 82 (29.3) 16 (13.9) 8 (53.3) 0
0.08
Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Maine Medical Center, Portland, Maine 04102 2 Center for Outcomes Research and Evaluation, Maine Medical Center, Portland, Maine 04102 Received 23 July 2014; accepted 6 September 2014
ABSTRACT: Purpose. The aim of this study was to compare rates of genetic counseling, invasive prenatal diagnosis, and trisomy 21 detection among women at increased risk for aneuploidy, before versus after the availability of noninvasive prenatal testing (NIPT). Methods. This institutional review board–exempt retrospective study included all women who had an ultrasound (US) examination between 10 0/7 and 21 6/7 weeks’ gestation and were eligible for NIPT (ie, age 35 years, US findings suggestive of increased aneuploidy risk, positive aneuploidy screen, prior trisomic fetus, parental balanced translocation with increased risk for trisomy 13 or 21) between June 1, 2012 and February 1, 2013. NIPT was performed by a single laboratory after patients received genetic counseling. We also identified a comparison group of women evaluated between December 1, 2010 and November 30, 2011, who would have been eligible for NIPT had it been available. The two groups were compared for maternal demographics, aneuploidy risk factors, rates of genetic counseling, invasive diagnostic procedures, and trisomy 21 detection. Results. The before-NIPT and after-NIPT groups contained 1,464 and 1,046 subjects, respectively. All 33 fetuses with trisomy 21 in the two groups were identified by positive aneuploidy screening. After the introduction of NIPT, genetic counseling for aneuploidy risk increased (adjusted odds ratio [aOR], 1.77 [1.49–2.11]; p < 0.0001) and the overall invasive diagnosis (aOR, 0.42 [0.32–0.55]; p < 0.0001), including amniocentesis (aOR, 0.37 [0.27–0.52], p < 0.0001), decreased, whereas the prenatal diagnosis of trisomy 21 remained similar (88% versus 100%; p 5 0.86). Correspondence to: J.R. Wax C 2014 Wiley Periodicals, Inc. V
VOL. 43, NO. 1, JANUARY 2015
Conclusions. NIPT in clinical practice uses more genetic counseling resources but requires significantly fewer invasive procedures to maintain the detection C 2014 Wiley Periodicals, Inc. rates of trisomy 21. V J Clin Ultrasound 43:1–6, 2015; Published online in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jcu.22243 Keywords: aneuploidy screening; cell-free fetal DNA; Down syndrome; trisomy 21; ultrasonography
INTRODUCTION
N
oninvasive prenatal testing (NIPT) analyzes cell-free fetal DNA in the maternal circulation to provide highly efficient screening for trisomy 21 in pregnancies at increased risk for fetal aneuploidy.1–4 NIPT offers higher detection and lower screen-positive rates than existing aneuploidy screening tests do, and it should reduce the invasive diagnosis rate while maintaining or increasing trisomy 21 detection. Two studies demonstrated fewer invasive diagnostic procedures following the clinical introduction of NIPT. However, neither report evaluated the effect on trisomy 21 detection.5,6 Although NIPT has been validated by different research methods, experience suggests that new aneuploidyscreening methods may not demonstrate performance that is as robust as expected when introduced into clinical practice.7 Several professional organizations have published recommendations on the clinical use of NIPT, including patient selection and counseling.8–11 These groups uniformly call for pretest counseling by a genetic counselor or other 1
WAX ET AL
trained individual, with genetic counseling available to any screen-positive patient. The effect of these recommendations on genetic counseling resources remains unknown. Therefore, we performed a retrospective cohort study to evaluate whether clinically implementing NIPT increases genetic counseling utilization, reduces invasive diagnostic procedures (amniocentesis or chorionic villus sampling), or affects trisomy 21 detection. PATIENTS AND METHODS
This institutional review board–exempt study included all women with a singleton pregnancy seen in a single maternal–fetal medicine practice for an ultrasound (US) examination between 10 0/7 and 21 6/7 weeks’ gestation from June 1, 2012 to February 1, 2013, and who were eligible for NIPT according to current guidelines. NIPT was available to all women regardless of payor status. Women were considered to be screen-positive for increased aneuploidy risk if they had exhibited any NIPT eligibility criteria, which included maternal age 35 years at delivery, positive serum aneuploidy screen, US findings suggestive of increased aneuploidy risk, prior pregnancy with a trisomic fetus, or parental balanced translocation with increased risk for fetal trisomy 13 or 21.9–11 US findings considered suggestive of aneuploidy included first-trimester nuchal translucency 3.5 mm or cystic hygroma colli. Secondtrimester features included nuchal skin-fold thickness 6 mm, absent nasal bone, short humerus less than the fifth percentile for gestational age, short femur less than the fifth percentile for gestational age, echogenic intracardiac focus, echogenic bowel, double-bubble sign, congenital heart defect, or other major congenital anomaly. All US examinations were performed by registered diagnostic medical sonographers in an American Institute of Ultrasound in Medicine– accredited practice, and results were interpreted by one of two maternal–fetal medicine physicians. NIPT was ordered and performed by a single commercial laboratory using massively parallel shotgun sequencing. All patients in both study periods had been offered nondirective counseling by a certified genetic counselor employed by our practice for increased aneuploidy risk. However, genetic counseling was a prerequisite to NIPT in our practice, consistent with current guidelines.8–11 We identified a comparison group of women with singleton pregnancies seen in our office for US examination between 10 0/7 and 21 6/7 2
weeks’ gestation from December 1, 2010 through November 30, 2011, who would have been eligible for NIPT had it been available. The two groups were compared for maternal demographics, aneuploidy risk factors, receipt of genetic counseling for aneuploidy risk, amniocentesis, chorionic villus sampling, and population-based trisomy 21 detection rates by screening tests and invasive prenatal diagnosis. Other than the introduction of NIPT, no practice changes were introduced during the study period. Data were abstracted from several prospectively ascertained databases, including (1) genetic counseling, (2) prenatal screening, (3) US, and (4) prenatal and postnatal genetic diagnosis. To ensure complete trisomy 21 ascertainment, all subjects were cross-referenced to pediatrics genetics records that included all live births with trisomy 21 and to the fetal–neonatal death database. Among fetal and neonatal deaths without chromosomal analysis, a normal autopsy or physical examination served as a proxy for a normal karyotype. Categorical data, including rates of trisomy 21 detection by invasive prenatal diagnosis, were compared using the v2 or Fisher’s exact test. Continuous variables were compared using the two-tailed unpaired t test. We used logistic regression analysis to test the effect of NIPT availability on outcomes, controlling for potential confounders. Statistical significance was set at p < 0.05. RESULTS
A total of 1,464 subjects before and 1,046 subjects after NIPT availability were included. Figure 1 is a flow diagram for the 1,046 subjects after NIPT availability. There were no significant differences in the groups’ demographics. The after-NIPT group had significantly fewer positive serum-based aneuploidy screening tests than the before-NIPT group did (115 versus 223; p 5 0.002). However, 100 of 166 women (60.2%) who underwent NIPT had NIPT as a first-line aneuploidy screen without a prior serum-based test. The frequency of other positive aneuploidy screening tests, which also comprised NIPT eligibility criteria, was similar by group (Table 1). US findings associated with increased aneuploidy risk were not significantly different by study group, except for increased frequencies of short femur in the before-NIPT groups and thickened nuchal skin fold in the after-NIPT groups (Table 2). JOURNAL OF CLINICAL ULTRASOUND
NIPT IN CLINICAL PRACTICE
FIGURE 1. Flow diagram illustrates the distribution of noninvasive prenatal diagnosis (NIPT)-eligible patients when NIPT was available. IUFD, intrauterine fetal death; TOP, termination of pregnancy; CHD, congenital heart disease; SAB, spontaneous abortion.
TABLE 1 Patients’ Characteristics
Characteristic
TABLE 2 Ultrasound Findings Suggesting Increased Risk of Aneuploidy
Before NIPT,
After NIPT,
n 5 1,464
n 5 1,046
p Value
Ultrasound Finding Maternal age, years (mean 6 SD) Parity, n 0 1 Unknown NIPT eligibility risk factors*, n (%) Maternal age 35 years US Positive screen Prior trisomy Parental translocation
34.5 6 5.9
458 988 20
34.6 6 5.5
354 678 14
After NIPT
n 5 1,464 (%)
n 5 1,046 (%)
p Value
391 (26.7)
280 (26.8)
1.0
12 (0.8)
6 (0.6)
0.51
19 (1.3)
7 (0.7)
0.14
7 (0.5) 8 (0.5) 272 (18.6)
3 (0.3) 13 (1.2) 215 (20.1)
0.51 0.05 0.11
33 (2.3) 0 (0) 26 (1.8) 66 (4.5) 13 (0.9) 5 (0.3)
19 (1.8) 2 (0.2) 11 (1.1) 45 (4.3) 2 (0.2) 1 (0.1)
0.51 0.17 0.16 1.0 0.03 0.26
0.68
0.17
1,023 (69.8)
742 (70.9)
0.53
391 (26.7) 223 (15.2) 25 (1.7) 0 (0)
280 (26.8) 115 (11.0) 15 (1.4) 1 (0.1)
1.0 0.002 0.59 0.42
Abbreviations: NIPT, noninvasive prenatal testing; US, ultrasound. *Data add up to more than the total number of patients in each group because some patients had more than one indication.
All known fetuses with trisomy 21, 25 in the before-NIPT and 8 in the after-NIPT groups (prevalences 1.7% and 0.8%; p 5 0.04) were detected by positive aneuploidy screening on the basis of history, serum tests, or US findings. Ninety-five percent confidence intervals for 100% trisomy 21 detection before NIPT, after NIPT, and overall were 87–100%, 66–100%, and 90–100%, respectively. In the before-NIPT and after-NIPT groups, trisomy 21 was prenatally diagnosed in 88% and 100% of cases by amniocentesis (n 5 10 versus n 5 3) and chorionic villus sampling (n 5 12 versus n 5 5) and postnatally by neonatal blood in 12% (n 5 3 verVOL. 43, NO. 1, JANUARY 2015
Before NIPT
Ultrasound finding present, n First trimester, n Nuchal translucency 3.5 mm Cystic hygroma Second trimester, n* Absent nasal bone Nuchal fold 6 mm Echogenic intracardiac focus Cardiac anomaly Double bubble Echogenic bowel Other anomaly Short femur Short humerus
Abbreviation: NIPT, noninvasive prenatal testing. *Data add up to more than the total number of patients in each group because some patients had more than one finding.
sus n 5 0); p 5 0.86. Significantly more women underwent amniocentesis after a positive serum-based screen before versus after NIPT: 62 of 152 (40.8%) versus 5 of 56 (8.9%); p < 0.0001. Likewise, significantly more women had chorionic villus sampling after a positive serum-based screen before versus after NIPT: 17 of 85 (20.0%) versus 1 of 50 (2.0%); p 5 0.003. NIPT was accepted by 166 (15.9%) eligible subjects, including 54 of 316 (17.1%) with public 3
WAX ET AL TABLE 3 Genetic Counseling by Indication, Before versus After NIPT Before NIPT (n 5 1,464) Indication* Age 35 years US finding Positive screen Prior trisomy Prior translocation
After NIPT (n 5 1,046)
Counseling
No Counseling
Counseling
No Counseling
p Value
385 (37.6) 165 (42.2) 208 (93.3) 14 (56.0) 0
638 (62.3) 226 (57.8) 15 (6.7) 11 (44.0) 0
249 (33.6) 198 (70.7) 99 (86.1) 7 (46.7) 1 (100)
493 (66.4) 82 (29.3) 16 (13.9) 8 (53.3) 0
0.08
