- Email: [email protected]
Reasons for trisomy 13 or 18 births despite the availability of prenatal diagnosis and pregnancy termination
Early Human Development 73 (2003) 53 – 60 www.elsevier.com/locate/earlhumdev
Reasons for trisomy 13 or 18 births despite the availability of prenatal diagnosis and pregnancy termination Bradford D. Gessner * Alaska Division of Public Health, P.O. Box 240249, 3601 C Street, Anchorage, AK 99524, USA Accepted 13 May 2003
Abstract Background: Few studies have evaluated the reasons why lethal chromosomal anomalies continue to occur despite the importance of this question for maximizing perinatal care. Aims: To determine why trisomy 13 or 18 births continue to occur in Alaska. Study design: Case series involving review of maternal and infant medical records. Subjects: All 28 known infants and fetuses that died with trisomy 13 or 18 during 1992 – 2001 and their mothers. Outcome measures: The proportion of mothers that declined or received a variety of routine prenatal tests, the results of prenatal testing, and the impact of testing on decisions related to pregnancy. Results: Seventeen women declined pregnancy termination or amniocentesis, 10 had no prenatal risk factors and were not offered these procedures, and one woman had an amniocentesis but was not offered pregnancy termination. Twenty-six women had z 1 prenatal ultrasounds; for 17 women, these were interpreted as normal throughout pregnancy (n = 11) or until after 30 weeks gestation (n = 6) despite substantial fetal malformations. Fourteen of 15 women with an abnormal ultrasound had an amniocentesis compared to one of eight women whose only risk factor was advanced maternal age. Conclusions: Most trisomy 13 or 18 deliveries occurred to women who declined amniocentesis or pregnancy termination. Failure to identify abnormalities on prenatal ultrasound may have contributed to the decision not to have these procedures. D 2003 Elsevier Ireland Ltd. All rights reserved. Keywords: Trisomy 13 or 18; Prenatal diagnosis; Pregnancy termination
* Tel.: +1-907-269-3446; fax: +1-907-269-3432. E-mail address: [email protected] (B.D. Gessner). 0378-3782/03/$ - see front matter D 2003 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/S0378-3782(03)00072-0
54
B.D. Gessner / Early Human Development 73 (2003) 53–60
1. Introduction For over 10 years, ultrasound evaluation followed by amniocentesis confirmation has allowed early prenatal diagnosis of trisomy 13 and 18, both lethal conditions; additionally, triple test (alpha-fetoprotein, total beta human chorionic gonadotrophin, and estriol) has aided with screening for trisomy 18. Alaska abortion laws allow for pregnancy termination throughout pregnancy when lethal conditions such as these exist. Thus, continued delivery of infants and late-gestation fetuses with these conditions could represent suboptimal case management. Alternatively, deliveries could occur because mothers declined to have diagnostic tests performed or their pregnancies terminated, they presented late or not at all for prenatal care, or they did not have relevant risk factors. Potential interventions differ for each of these possibilities. The current study analyzed infant deaths and fetal demises associated with trisomy 13 or 18 that occurred in Alaska during 1992– 2001. The goal of the study was to evaluate the reasons why deliveries associated with trisomy 13 and 18 occurred despite the availability of prenatal diagnosis and pregnancy termination. Trisomy 13 and 18 were selected for evaluation because of their lethal outcome and well-established and easily evaluated indications for amniocentesis (maternal age and abnormal ultrasound or triple test).
2. Methods Data came from records gathered as part of the Alaska Maternal – Infant Mortality Review (AMIMR) process. AMIMR is a surveillance project in which the State of Alaska has enlisted qualified professionals to review medical and other records for the purpose of developing public health recommendations. AMIMR includes all fetal and infant deaths that occur to Alaska residents regardless of whether the deaths occur in Alaska. AMIMR is prohibited from communicating information about specific individuals outside of the review process and from using results for any purpose other than generating population-based recommendations. As part of routine public health surveillance, no institutional review board oversees the project and informed consent is not sought. Infant and fetal deaths were identified through matched birth and death certificates from the Bureau of Vital Statistics. For each infant and fetal death, a file was created that included birth and death certificates, infant and maternal medical records, and autopsy reports. Data from each of these sources were abstracted using a standard form and entered into a computer database. The AMIMR contained information on all documented cases of trisomy 13 or 18 leading to fetal or infant death during 1992– 2001. One fetal death occurring during 1992 and reported by vital statistics as due to trisomy 13 or 18 did not have records available for review; because several deaths ascribed to trisomy 13 or 18 on death certificates actually had different chromosomal abnormalities upon review of the medical records, this case was not included in any analysis, including analysis of incidence. All analyses were performed with SPSS version 10.0 statistical software.
B.D. Gessner / Early Human Development 73 (2003) 53–60
55
3. Results 3.1. Case characteristics During 1992– 2001, 25 infant deaths and three late-gestation fetal demises (delivered at 31, 35, and 39 weeks gestation) associated with trisomy 13 (n = 14) or 18 (n = 14) were identified (Table 1). None of the included cases were associated with medical induction for pregnancy termination. Ten cases occurred during 1992– 1993, six each during 1994 – 1995 and 1996– 1997, four in 1998 – 1999, and two during 2000– 2001 (4.4, 2.9, 3.0, 2.0, and 1.0 per 10,000 live births, respectively) (X2 for trend, 4.6; p = 0.033). The median maternal age at delivery was 33 years (range, 19 –46). Religious affiliation was noted for 14 women and all belonged to a Christian denomination including Catholicism (4), Protestantism (8), or unspecified (2). All but one case was associated with major anomalies that were documented pre- or postnatally including 22 with complex cardiac malformations and 10 with central nervous system malformations such as a variety of central nervous system cysts, corpus collosum anomalies, and neural tube defects. Other major anomalies included cleft lip or palate (5), polycystic kidneys (5), and omphalocele (3). Prenatal care began at 20 weeks gestation or earlier for 27 women—including 18 who began care during the first trimester—while one woman received no prenatal care. Thirteen women received prenatal care in Anchorage, the state’s major urban center, three at one of Alaska’s military bases, and the remainder from one of seven other cities in Alaska. One woman received prenatal care from a lay midwife while the remainder received care from a physician, most commonly an obstetrician. 3.2. Triple test Four women had abnormalities detected on ultrasound early in pregnancy and did not have the opportunity to have a triple test. Nine women refused the triple test throughout pregnancy. Two women with pregnancies associated with trisomy 18 had a triple test performed and results were abnormal for one; an additional four women with pregnancies associated with trisomy 13 had a normal triple test but the triple test is not designed to detect trisomy 13. The remaining nine women, seven of whom were less than 35 years of age, had no medical record documentation that they had been offered a triple test. 3.3. Prenatal ultrasound One woman received no prenatal care while a second had incomplete prenatal records available. The remaining 26 women had one or more documented prenatal ultrasounds (Table 1). Eleven of these 26 women had ultrasounds interpreted as normal throughout pregnancy including nine who had at least one ultrasound performed at 15 weeks gestation or later. An additional six women did not have an abnormal ultrasound until after 30 weeks gestation (all within 5 weeks of delivery); five of these women had had at least one earlier ultrasound that was interpreted as normal. During 1992– 1996, 12 of 15 women (80%) with at least one documented second- or third-trimester ultrasound had one or more of
56
B.D. Gessner / Early Human Development 73 (2003) 53–60
Table 1 Infant and fetal deaths associated with trisomy 13 or 18 (Alaska, 1992 – 2001) Case Trisomy Gestation Maternal Triple type at diagnosis age test result
Gestation at 1st prenatal US
Gestation at 2nd prenatal US
1 2 3 4 5
13 13 13 13 13
17 21 22 28 Postnatal
23 32 41 22 38
15 21 21 17 34
– 29 – 24 –
6 7
13 13
Postnatal Postnatal
20 31
8
13
Postnatal
22
9
13
Postnatal
27
10
13
Postnatal
41
11
13
Postnatal
34
12
13
Postnatal
19
13 14 15 16
13 13 18 18
Postnatal Postnatal 18 18
36 27 35 38
17 18 19 20 21 22 23 24 25
18 18 18 18 18 18 18 18 18
19 27 27 29 34 36 40 Postnatal Postnatal
37 40 38 19 25 22 30 43 38
26
18
Postnatal
39
27
18
Postnatal
30
28
18
Postnatal
46
Not done Normal Refused Not done Normal
Gestation abnormality identified on US
15 21 21 24 No abnormality seen Normal 17 29 37 Not done 16 33 No abnormality seen Not done 19 26 No abnormality seen Refused 7 – No abnormality seen Refused 27 – No abnormality seen Refused 16 – No abnormality seen Normal 3 17 No abnormality seen Refused 34 – 34 Unknown Unknown Unknown Unknown Not done 19 – 19 Not done 15 – No abnormality seen Not done 21 – 21 Not done 27 – 27 Not done 7 12 27 Normal 10 20 23 Not done 7 16 32 Not done 12 34 34 Not done 16 30 39 Refused 24 29 31 Refused 9 – No abnormality seen Refused 11 37 No abnormality seen Abnormal 17 – No abnormality seen No PNCc No PNCc No PNCc No PNCc
Second-trimester amniocentesis
Performed Performed Performed Performed Declined Not offered Not offered Declined Not offered Declined Not offered Not offered Performeda Not offered Performed Performed Performed Performed Performed Performed Not offeredb Not offeredb Not offeredb Performeda Declined Not offered Declined No PNCc
a Initially refused second-trimester procedure; performed only after abnormalities identified on ultrasound just before delivery. b No prenatal risk factors during second trimester; amniocentesis performed after abnormal ultrasound during the third trimester. c Prenatal care.
B.D. Gessner / Early Human Development 73 (2003) 53–60
57
these ultrasounds interpreted as normal compared to three of nine women (33%) during 1997 –2001 (the remaining four women did not have a documented second-trimester ultrasound). Five women who did not obtain a diagnosis until delivery had one or more ultrasounds at greater than 17 weeks gestation interpreted as normal. All five delivered an infant with major anomalies including (1) truncus arteriosis, microcephaly, and polyhyramnios; (2) tetralogy of Fallot and polydactyly; (3) double outlet right ventricle/hypoplastic left heart and agenesis of the corpus collosum; (4) atrial septal defect/patent ductus arteriosis, cleft palate, and polydactyly; and (5) ventricular septal defect, cleft palate, agenesis of the corpus collosum, and myelomeningocele. The specialty of the person who interpreted the ultrasounds was available for 20 women. Eleven initial ultrasounds were read by a radiologist, three by a perinatologist, four by other physicians, and three by nonphysicians. All women who had abnormalities identified at less than 30 weeks gestation had an initial ultrasound interpreted by a radiologist or perinatologist. However, a radiologist also read the ultrasounds for six women whose results were interpreted as normal. 3.4. Amniocentesis Seventeen women had risk factors identified early in pregnancy—including age z 35 years or abnormalities identified on ultrasound—and were offered an amniocentesis during the second trimester. Ten of these women consented to an amniocentesis at less than 30 weeks gestation (median, 21 weeks; range, 17 to 29) while seven declined amniocentesis throughout pregnancy (n = 5) or initially declined and then consented immediately before a near-term delivery (n = 2). The remaining 11 women were not offered a second-trimester amniocentesis. One had no prenatal care, and thus implicitly refused amniocentesis, while 10 had no prenatal risk factors identified early in pregnancy and thus no reason to be offered an early amniocentesis. Three of these women eventually had an amniocentesis after abnormalities were identified late during pregnancy. 3.5. Pregnancy termination Nine of the 10 women who received a diagnosis at less than 30 weeks gestation explicitly refused pregnancy termination; among women for whom this was known, this decision was not associated with religious affiliation. The 10th woman had an abnormal ultrasound at 23 weeks gestation, an amniocentesis delayed until 29 weeks gestation, and was subsequently told that she would have to carry the infant to natural delivery. 3.6. Summary Seventeen women (61%) declined pregnancy termination, declined an amniocentesis until immediately before delivery, or did not present for prenatal care and thus implicitly declined these procedures. Ten women had no known risk factors and thus were not
58
B.D. Gessner / Early Human Development 73 (2003) 53–60
offered an amniocentesis or pregnancy termination. The last woman had a delayed amniocentesis and was not offered pregnancy termination. Women were more likely to consent to an amniocentesis following an abnormal ultrasound than solely for advanced maternal age. Of the 15 women with abnormalities detected on ultrasound, 14 agreed to an amniocentesis shortly after abnormalities were identified while the last woman had abnormalities identified the day before delivery and thus did not have the opportunity to have this procedure. By contrast, only one of nine women z 35 years of age agreed to an amniocentesis in the absence of abnormal ultrasound results.
4. Discussion Current optimal management of pregnancies associated with trisomy 13 or 18 involves four steps: presentation for care during the first trimester, screening, definitive diagnosis, and appropriate education and counseling regarding pregnancy termination. All but one woman in the current study presented for prenatal care at 20 weeks gestation or earlier. Thus, it is unlikely that emphasis on earlier prenatal care will alter the management of trisomy 13 or 18 deliveries in Alaska. Screening will rely on multiple strategies such as universal use of the triple test [1,2], first [3– 6] and second [7 –9] trimester ultrasound evaluation, and newer technologies [10,11]. In the current study, many low-risk women did not have documentation that a triple test had been offered or performed. Additionally, many ultrasounds were interpreted as normal, including those performed late in pregnancy; this occurred despite skilled personnel performing most evaluations and fetuses having major malformations that could have been identified early in pregnancy. In theory, then, better implementation of existing screening strategies—through education, technical training, or better equipment—could influence the management of pregnancies associated with trisomy 13 or 18, including the number of unexpected (and potentially unwanted) deliveries. Screening results may influence a woman’s decision to obtain a definitive diagnosis by amniocentesis. In the current study, all women who had abnormalities identified on ultrasound early enough agreed to an amniocentesis while few women agreed for advanced maternal age alone [12]; indeed, a prenatal ultrasound interpreted as normal may have convinced older women who otherwise would have agreed to an amniocentesis to decline this procedure. Triple test results were not associated with a woman’s decision to have an amniocentesis as most based this decision on ultrasound results alone or were not offered the test. The final step in optimal pregnancy management is education and counseling regarding pregnancy termination. Almost two thirds of women in the current study decided not to obtain a diagnosis or knew the diagnosis and elected to continue the pregnancy. To the extent that women made their decisions based on a clear understanding of risks and benefits, this reflects successful case management. It is possible, though, that in addition to suboptimal use of screening tests, suboptimal prenatal counseling contributed to the decision not to obtain an amniocentesis or pregnancy termination. The available data argue against this since only one case was identified where the medical records documented that
B.D. Gessner / Early Human Development 73 (2003) 53–60
59
the physician provided incorrect information, specifically regarding abortion availability in Alaska. However, more subtle omissions in counseling—such as the positive economic and mental health aspects of prenatal diagnosis even when pregnancy termination is not considered an option [13]—would not necessarily have been recorded in the medical record. Combined trisomy 13 and 18 birth rates in Alaska for the last 4 years of the study were lower than the combined of 2 to 4 per 10,000 live births reported from most other states [14]. Moreover, a substantial decline in the occurrence of trisomy 13 or 18 deaths occurred in Alaska during the study period despite no change in the number of live births to women over 35 years of age [15]; by comparison, overall US rates have remained constant during 1992 –1998. This suggests that some of the issues raised by the current study may have already been addressed. For example, the first perinatologist in Alaska arrived during 1994 and a second came during 1997. This led to improvements in diagnostic, therapeutic, counseling, and education services. For example, one perinatologist in Anchorage is certified in nuchal translucency thickness measurement during first-trimester ultrasound. Diagnostic improvements may help explain our finding that since 1996, fewer trisomy 13 and 18 cases occurred in the context of a second trimester sonogram interpreted as normal [16]. Alaska’s small population and substantial perinatal care regionalization make such improvements in perinatal care readily available to most eligible women. The current study had several limitations. No population-based information was available for comparison on the proportion of all eligible women who agreed to have a triple test, amniocentesis, or abortion. Similarly, records for infants born with nonlethal but major abnormalities, such as trisomy 21, by definition were not included in the AMIMR database; thus, it is unknown if the current results reflect the management of women carrying fetuses with other types of major anomalies. Finally, it should be noted that all deaths following birth at > 20 weeks gestation, including those resulting from medical intervention, will be counted as a fetal or neonatal death and thus included in determination of incidence rates. Although this was not an issue in the current study, this implies that reductions in the incidence of trisomy 13 and 18 births will not occur with pregnancy termination during the second half of pregnancy. The data presented here raise the possibility that decreases in the number of unexpected outcomes and potentially unwanted deliveries might occur with better implementation of current screening strategies (particularly improved ultrasound training and equipment) and implementation of new technology. The substantial decline in the rate of trisomy 13 or 18 births seen during the study period may indicate that these interventions are being adopted. The cost-effectiveness of potential intervention strategies needs to be evaluated [17], particularly given that the majority of women in the current study who did not receive a definitive prenatal diagnosis were low-risk. References [1] Hogge WA, Fraer L, Melegari T. Maternal serum screening for fetal trisomy 18: benefits of patient-specific risk protocol. Am J Obstet Gynecol 2001 (Aug.);185:289 – 93. [2] Benn PA, Leo MV, Rodis JF, Beazoglou T, Collins R, Horne D. Maternal serum screening for fetal trisomy 18: a comparison of fixed cutoff and patient-specific risk protocols. Obstet Gynecol 1999;93:707 – 11.
60
B.D. Gessner / Early Human Development 73 (2003) 53–60
[3] Schwarzler P, Carvalho JS, Senat MV, Masroor T, Campbell S, Ville Y. Screening for fetal aneuploidies and fetal cardiac abnormalities by nuchal translucency thickness measurement at 10 – 14 weeks of gestation as part of routine antenatal care in an unselected population. Br J Obstet Gynaecol 1999;106:1029 – 34. [4] Wayda K, Kereszturi A, Orvos H, Horvath E, Pal A, Kovacs L, et al. Four years experience of first-trimester nuchal translucency screening for fetal aneuploidies with increasing regional availability. Acta Obstet Gynecol Scand 2001;80:1104 – 9. [5] Souter VL, Nyberg DA. Sonographic screening for fetal aneuploidy: first trimester. J Ultrasound Med 2001;20:775 – 90. [6] Whitlow BJ, Chatzipapas IK, Lazanakis ML, Kadir RA, Economides DL. The value of sonography in early pregnancy for the detection of fetal abnormalities in an unselected population. Br J Obstet Gynaecol 1999;106:929 – 36. [7] Shipp TD, Benacerraf BR. Second trimester ultrasound screening for chromosomal abnormalities. Prenat Diagn 2002;22:296 – 307. [8] Nyberg DA, Souter VL. Sonographic markers of fetal trisomies: second trimester. J Ultrasound Med 2001;20:655 – 74. [9] Benacerraf BR, Nadel A, Bromley B. Identification of second-trimester fetuses with autosomal trisomy by use of a sonographic scoring index. Radiology 1994;193:135 – 40. [10] Spencer K, Ong C, Skentou H, Liao AW, Nicolaides HK. Screening for trisomy 13 by fetal nuchal translucency and maternal serum free beta-hCG and PAPP-A at 10 – 14 weeks of gestation. Prenat Diagn 2000;20:411 – 6. [11] Tul N, Spencer K, Noble P, Chan C, Nicolaides K. Screening for trisomy 18 by fetal nuchal translucency and maternal serum free beta-hCG and PAPP-A at 10 – 14 weeks of gestation. Prenat Diagn 1999;19:1035 – 42. [12] DeVore GR, Romero R. Genetic sonography: a cost-effective method for evaluating women 35 years and older who decline genetic amniocentesis. J Ultrasound Med 2002;21:5 – 13. [13] Ralston SJ, Wertz D, Chelmow D, Craigo SD, Bianchi DW. Pregnancy outcomes after prenatal diagnosis of aneuploidy. Obstet Gynecol 2001;97:729 – 33. [14] Birth defects surveillance data from selected states, 1995 – 1999. Teratology 2002;66:S129 – 211. [15] Alaska Bureau of Vital Statistics. Annual reports for 1992 through 1999. Alaska Department of Health and Social Services, Division of Public Health. [16] DeVore GR. The genetic sonogram: its use in the detection of chromosomal abnormalities in fetuses of women of advanced maternal age. Prenat Diagn 2001;21:40 – 5. [17] Smith-Bindman R, Hosmer W, Feldstein VA, Deeks JJ, Goldberg JD. Second-trimester ultrasound to detect fetuses with down syndrome: a meta-analysis. JAMA 2001;285:1044 – 55.
Reasons for trisomy 13 or 18 births despite the availability of prenatal diagnosis and pregnancy termination Bradford D. Gessner * Alaska Division of Public Health, P.O. Box 240249, 3601 C Street, Anchorage, AK 99524, USA Accepted 13 May 2003
Abstract Background: Few studies have evaluated the reasons why lethal chromosomal anomalies continue to occur despite the importance of this question for maximizing perinatal care. Aims: To determine why trisomy 13 or 18 births continue to occur in Alaska. Study design: Case series involving review of maternal and infant medical records. Subjects: All 28 known infants and fetuses that died with trisomy 13 or 18 during 1992 – 2001 and their mothers. Outcome measures: The proportion of mothers that declined or received a variety of routine prenatal tests, the results of prenatal testing, and the impact of testing on decisions related to pregnancy. Results: Seventeen women declined pregnancy termination or amniocentesis, 10 had no prenatal risk factors and were not offered these procedures, and one woman had an amniocentesis but was not offered pregnancy termination. Twenty-six women had z 1 prenatal ultrasounds; for 17 women, these were interpreted as normal throughout pregnancy (n = 11) or until after 30 weeks gestation (n = 6) despite substantial fetal malformations. Fourteen of 15 women with an abnormal ultrasound had an amniocentesis compared to one of eight women whose only risk factor was advanced maternal age. Conclusions: Most trisomy 13 or 18 deliveries occurred to women who declined amniocentesis or pregnancy termination. Failure to identify abnormalities on prenatal ultrasound may have contributed to the decision not to have these procedures. D 2003 Elsevier Ireland Ltd. All rights reserved. Keywords: Trisomy 13 or 18; Prenatal diagnosis; Pregnancy termination
* Tel.: +1-907-269-3446; fax: +1-907-269-3432. E-mail address: [email protected] (B.D. Gessner). 0378-3782/03/$ - see front matter D 2003 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/S0378-3782(03)00072-0
54
B.D. Gessner / Early Human Development 73 (2003) 53–60
1. Introduction For over 10 years, ultrasound evaluation followed by amniocentesis confirmation has allowed early prenatal diagnosis of trisomy 13 and 18, both lethal conditions; additionally, triple test (alpha-fetoprotein, total beta human chorionic gonadotrophin, and estriol) has aided with screening for trisomy 18. Alaska abortion laws allow for pregnancy termination throughout pregnancy when lethal conditions such as these exist. Thus, continued delivery of infants and late-gestation fetuses with these conditions could represent suboptimal case management. Alternatively, deliveries could occur because mothers declined to have diagnostic tests performed or their pregnancies terminated, they presented late or not at all for prenatal care, or they did not have relevant risk factors. Potential interventions differ for each of these possibilities. The current study analyzed infant deaths and fetal demises associated with trisomy 13 or 18 that occurred in Alaska during 1992– 2001. The goal of the study was to evaluate the reasons why deliveries associated with trisomy 13 and 18 occurred despite the availability of prenatal diagnosis and pregnancy termination. Trisomy 13 and 18 were selected for evaluation because of their lethal outcome and well-established and easily evaluated indications for amniocentesis (maternal age and abnormal ultrasound or triple test).
2. Methods Data came from records gathered as part of the Alaska Maternal – Infant Mortality Review (AMIMR) process. AMIMR is a surveillance project in which the State of Alaska has enlisted qualified professionals to review medical and other records for the purpose of developing public health recommendations. AMIMR includes all fetal and infant deaths that occur to Alaska residents regardless of whether the deaths occur in Alaska. AMIMR is prohibited from communicating information about specific individuals outside of the review process and from using results for any purpose other than generating population-based recommendations. As part of routine public health surveillance, no institutional review board oversees the project and informed consent is not sought. Infant and fetal deaths were identified through matched birth and death certificates from the Bureau of Vital Statistics. For each infant and fetal death, a file was created that included birth and death certificates, infant and maternal medical records, and autopsy reports. Data from each of these sources were abstracted using a standard form and entered into a computer database. The AMIMR contained information on all documented cases of trisomy 13 or 18 leading to fetal or infant death during 1992– 2001. One fetal death occurring during 1992 and reported by vital statistics as due to trisomy 13 or 18 did not have records available for review; because several deaths ascribed to trisomy 13 or 18 on death certificates actually had different chromosomal abnormalities upon review of the medical records, this case was not included in any analysis, including analysis of incidence. All analyses were performed with SPSS version 10.0 statistical software.
B.D. Gessner / Early Human Development 73 (2003) 53–60
55
3. Results 3.1. Case characteristics During 1992– 2001, 25 infant deaths and three late-gestation fetal demises (delivered at 31, 35, and 39 weeks gestation) associated with trisomy 13 (n = 14) or 18 (n = 14) were identified (Table 1). None of the included cases were associated with medical induction for pregnancy termination. Ten cases occurred during 1992– 1993, six each during 1994 – 1995 and 1996– 1997, four in 1998 – 1999, and two during 2000– 2001 (4.4, 2.9, 3.0, 2.0, and 1.0 per 10,000 live births, respectively) (X2 for trend, 4.6; p = 0.033). The median maternal age at delivery was 33 years (range, 19 –46). Religious affiliation was noted for 14 women and all belonged to a Christian denomination including Catholicism (4), Protestantism (8), or unspecified (2). All but one case was associated with major anomalies that were documented pre- or postnatally including 22 with complex cardiac malformations and 10 with central nervous system malformations such as a variety of central nervous system cysts, corpus collosum anomalies, and neural tube defects. Other major anomalies included cleft lip or palate (5), polycystic kidneys (5), and omphalocele (3). Prenatal care began at 20 weeks gestation or earlier for 27 women—including 18 who began care during the first trimester—while one woman received no prenatal care. Thirteen women received prenatal care in Anchorage, the state’s major urban center, three at one of Alaska’s military bases, and the remainder from one of seven other cities in Alaska. One woman received prenatal care from a lay midwife while the remainder received care from a physician, most commonly an obstetrician. 3.2. Triple test Four women had abnormalities detected on ultrasound early in pregnancy and did not have the opportunity to have a triple test. Nine women refused the triple test throughout pregnancy. Two women with pregnancies associated with trisomy 18 had a triple test performed and results were abnormal for one; an additional four women with pregnancies associated with trisomy 13 had a normal triple test but the triple test is not designed to detect trisomy 13. The remaining nine women, seven of whom were less than 35 years of age, had no medical record documentation that they had been offered a triple test. 3.3. Prenatal ultrasound One woman received no prenatal care while a second had incomplete prenatal records available. The remaining 26 women had one or more documented prenatal ultrasounds (Table 1). Eleven of these 26 women had ultrasounds interpreted as normal throughout pregnancy including nine who had at least one ultrasound performed at 15 weeks gestation or later. An additional six women did not have an abnormal ultrasound until after 30 weeks gestation (all within 5 weeks of delivery); five of these women had had at least one earlier ultrasound that was interpreted as normal. During 1992– 1996, 12 of 15 women (80%) with at least one documented second- or third-trimester ultrasound had one or more of
56
B.D. Gessner / Early Human Development 73 (2003) 53–60
Table 1 Infant and fetal deaths associated with trisomy 13 or 18 (Alaska, 1992 – 2001) Case Trisomy Gestation Maternal Triple type at diagnosis age test result
Gestation at 1st prenatal US
Gestation at 2nd prenatal US
1 2 3 4 5
13 13 13 13 13
17 21 22 28 Postnatal
23 32 41 22 38
15 21 21 17 34
– 29 – 24 –
6 7
13 13
Postnatal Postnatal
20 31
8
13
Postnatal
22
9
13
Postnatal
27
10
13
Postnatal
41
11
13
Postnatal
34
12
13
Postnatal
19
13 14 15 16
13 13 18 18
Postnatal Postnatal 18 18
36 27 35 38
17 18 19 20 21 22 23 24 25
18 18 18 18 18 18 18 18 18
19 27 27 29 34 36 40 Postnatal Postnatal
37 40 38 19 25 22 30 43 38
26
18
Postnatal
39
27
18
Postnatal
30
28
18
Postnatal
46
Not done Normal Refused Not done Normal
Gestation abnormality identified on US
15 21 21 24 No abnormality seen Normal 17 29 37 Not done 16 33 No abnormality seen Not done 19 26 No abnormality seen Refused 7 – No abnormality seen Refused 27 – No abnormality seen Refused 16 – No abnormality seen Normal 3 17 No abnormality seen Refused 34 – 34 Unknown Unknown Unknown Unknown Not done 19 – 19 Not done 15 – No abnormality seen Not done 21 – 21 Not done 27 – 27 Not done 7 12 27 Normal 10 20 23 Not done 7 16 32 Not done 12 34 34 Not done 16 30 39 Refused 24 29 31 Refused 9 – No abnormality seen Refused 11 37 No abnormality seen Abnormal 17 – No abnormality seen No PNCc No PNCc No PNCc No PNCc
Second-trimester amniocentesis
Performed Performed Performed Performed Declined Not offered Not offered Declined Not offered Declined Not offered Not offered Performeda Not offered Performed Performed Performed Performed Performed Performed Not offeredb Not offeredb Not offeredb Performeda Declined Not offered Declined No PNCc
a Initially refused second-trimester procedure; performed only after abnormalities identified on ultrasound just before delivery. b No prenatal risk factors during second trimester; amniocentesis performed after abnormal ultrasound during the third trimester. c Prenatal care.
B.D. Gessner / Early Human Development 73 (2003) 53–60
57
these ultrasounds interpreted as normal compared to three of nine women (33%) during 1997 –2001 (the remaining four women did not have a documented second-trimester ultrasound). Five women who did not obtain a diagnosis until delivery had one or more ultrasounds at greater than 17 weeks gestation interpreted as normal. All five delivered an infant with major anomalies including (1) truncus arteriosis, microcephaly, and polyhyramnios; (2) tetralogy of Fallot and polydactyly; (3) double outlet right ventricle/hypoplastic left heart and agenesis of the corpus collosum; (4) atrial septal defect/patent ductus arteriosis, cleft palate, and polydactyly; and (5) ventricular septal defect, cleft palate, agenesis of the corpus collosum, and myelomeningocele. The specialty of the person who interpreted the ultrasounds was available for 20 women. Eleven initial ultrasounds were read by a radiologist, three by a perinatologist, four by other physicians, and three by nonphysicians. All women who had abnormalities identified at less than 30 weeks gestation had an initial ultrasound interpreted by a radiologist or perinatologist. However, a radiologist also read the ultrasounds for six women whose results were interpreted as normal. 3.4. Amniocentesis Seventeen women had risk factors identified early in pregnancy—including age z 35 years or abnormalities identified on ultrasound—and were offered an amniocentesis during the second trimester. Ten of these women consented to an amniocentesis at less than 30 weeks gestation (median, 21 weeks; range, 17 to 29) while seven declined amniocentesis throughout pregnancy (n = 5) or initially declined and then consented immediately before a near-term delivery (n = 2). The remaining 11 women were not offered a second-trimester amniocentesis. One had no prenatal care, and thus implicitly refused amniocentesis, while 10 had no prenatal risk factors identified early in pregnancy and thus no reason to be offered an early amniocentesis. Three of these women eventually had an amniocentesis after abnormalities were identified late during pregnancy. 3.5. Pregnancy termination Nine of the 10 women who received a diagnosis at less than 30 weeks gestation explicitly refused pregnancy termination; among women for whom this was known, this decision was not associated with religious affiliation. The 10th woman had an abnormal ultrasound at 23 weeks gestation, an amniocentesis delayed until 29 weeks gestation, and was subsequently told that she would have to carry the infant to natural delivery. 3.6. Summary Seventeen women (61%) declined pregnancy termination, declined an amniocentesis until immediately before delivery, or did not present for prenatal care and thus implicitly declined these procedures. Ten women had no known risk factors and thus were not
58
B.D. Gessner / Early Human Development 73 (2003) 53–60
offered an amniocentesis or pregnancy termination. The last woman had a delayed amniocentesis and was not offered pregnancy termination. Women were more likely to consent to an amniocentesis following an abnormal ultrasound than solely for advanced maternal age. Of the 15 women with abnormalities detected on ultrasound, 14 agreed to an amniocentesis shortly after abnormalities were identified while the last woman had abnormalities identified the day before delivery and thus did not have the opportunity to have this procedure. By contrast, only one of nine women z 35 years of age agreed to an amniocentesis in the absence of abnormal ultrasound results.
4. Discussion Current optimal management of pregnancies associated with trisomy 13 or 18 involves four steps: presentation for care during the first trimester, screening, definitive diagnosis, and appropriate education and counseling regarding pregnancy termination. All but one woman in the current study presented for prenatal care at 20 weeks gestation or earlier. Thus, it is unlikely that emphasis on earlier prenatal care will alter the management of trisomy 13 or 18 deliveries in Alaska. Screening will rely on multiple strategies such as universal use of the triple test [1,2], first [3– 6] and second [7 –9] trimester ultrasound evaluation, and newer technologies [10,11]. In the current study, many low-risk women did not have documentation that a triple test had been offered or performed. Additionally, many ultrasounds were interpreted as normal, including those performed late in pregnancy; this occurred despite skilled personnel performing most evaluations and fetuses having major malformations that could have been identified early in pregnancy. In theory, then, better implementation of existing screening strategies—through education, technical training, or better equipment—could influence the management of pregnancies associated with trisomy 13 or 18, including the number of unexpected (and potentially unwanted) deliveries. Screening results may influence a woman’s decision to obtain a definitive diagnosis by amniocentesis. In the current study, all women who had abnormalities identified on ultrasound early enough agreed to an amniocentesis while few women agreed for advanced maternal age alone [12]; indeed, a prenatal ultrasound interpreted as normal may have convinced older women who otherwise would have agreed to an amniocentesis to decline this procedure. Triple test results were not associated with a woman’s decision to have an amniocentesis as most based this decision on ultrasound results alone or were not offered the test. The final step in optimal pregnancy management is education and counseling regarding pregnancy termination. Almost two thirds of women in the current study decided not to obtain a diagnosis or knew the diagnosis and elected to continue the pregnancy. To the extent that women made their decisions based on a clear understanding of risks and benefits, this reflects successful case management. It is possible, though, that in addition to suboptimal use of screening tests, suboptimal prenatal counseling contributed to the decision not to obtain an amniocentesis or pregnancy termination. The available data argue against this since only one case was identified where the medical records documented that
B.D. Gessner / Early Human Development 73 (2003) 53–60
59
the physician provided incorrect information, specifically regarding abortion availability in Alaska. However, more subtle omissions in counseling—such as the positive economic and mental health aspects of prenatal diagnosis even when pregnancy termination is not considered an option [13]—would not necessarily have been recorded in the medical record. Combined trisomy 13 and 18 birth rates in Alaska for the last 4 years of the study were lower than the combined of 2 to 4 per 10,000 live births reported from most other states [14]. Moreover, a substantial decline in the occurrence of trisomy 13 or 18 deaths occurred in Alaska during the study period despite no change in the number of live births to women over 35 years of age [15]; by comparison, overall US rates have remained constant during 1992 –1998. This suggests that some of the issues raised by the current study may have already been addressed. For example, the first perinatologist in Alaska arrived during 1994 and a second came during 1997. This led to improvements in diagnostic, therapeutic, counseling, and education services. For example, one perinatologist in Anchorage is certified in nuchal translucency thickness measurement during first-trimester ultrasound. Diagnostic improvements may help explain our finding that since 1996, fewer trisomy 13 and 18 cases occurred in the context of a second trimester sonogram interpreted as normal [16]. Alaska’s small population and substantial perinatal care regionalization make such improvements in perinatal care readily available to most eligible women. The current study had several limitations. No population-based information was available for comparison on the proportion of all eligible women who agreed to have a triple test, amniocentesis, or abortion. Similarly, records for infants born with nonlethal but major abnormalities, such as trisomy 21, by definition were not included in the AMIMR database; thus, it is unknown if the current results reflect the management of women carrying fetuses with other types of major anomalies. Finally, it should be noted that all deaths following birth at > 20 weeks gestation, including those resulting from medical intervention, will be counted as a fetal or neonatal death and thus included in determination of incidence rates. Although this was not an issue in the current study, this implies that reductions in the incidence of trisomy 13 and 18 births will not occur with pregnancy termination during the second half of pregnancy. The data presented here raise the possibility that decreases in the number of unexpected outcomes and potentially unwanted deliveries might occur with better implementation of current screening strategies (particularly improved ultrasound training and equipment) and implementation of new technology. The substantial decline in the rate of trisomy 13 or 18 births seen during the study period may indicate that these interventions are being adopted. The cost-effectiveness of potential intervention strategies needs to be evaluated [17], particularly given that the majority of women in the current study who did not receive a definitive prenatal diagnosis were low-risk. References [1] Hogge WA, Fraer L, Melegari T. Maternal serum screening for fetal trisomy 18: benefits of patient-specific risk protocol. Am J Obstet Gynecol 2001 (Aug.);185:289 – 93. [2] Benn PA, Leo MV, Rodis JF, Beazoglou T, Collins R, Horne D. Maternal serum screening for fetal trisomy 18: a comparison of fixed cutoff and patient-specific risk protocols. Obstet Gynecol 1999;93:707 – 11.
60
B.D. Gessner / Early Human Development 73 (2003) 53–60
[3] Schwarzler P, Carvalho JS, Senat MV, Masroor T, Campbell S, Ville Y. Screening for fetal aneuploidies and fetal cardiac abnormalities by nuchal translucency thickness measurement at 10 – 14 weeks of gestation as part of routine antenatal care in an unselected population. Br J Obstet Gynaecol 1999;106:1029 – 34. [4] Wayda K, Kereszturi A, Orvos H, Horvath E, Pal A, Kovacs L, et al. Four years experience of first-trimester nuchal translucency screening for fetal aneuploidies with increasing regional availability. Acta Obstet Gynecol Scand 2001;80:1104 – 9. [5] Souter VL, Nyberg DA. Sonographic screening for fetal aneuploidy: first trimester. J Ultrasound Med 2001;20:775 – 90. [6] Whitlow BJ, Chatzipapas IK, Lazanakis ML, Kadir RA, Economides DL. The value of sonography in early pregnancy for the detection of fetal abnormalities in an unselected population. Br J Obstet Gynaecol 1999;106:929 – 36. [7] Shipp TD, Benacerraf BR. Second trimester ultrasound screening for chromosomal abnormalities. Prenat Diagn 2002;22:296 – 307. [8] Nyberg DA, Souter VL. Sonographic markers of fetal trisomies: second trimester. J Ultrasound Med 2001;20:655 – 74. [9] Benacerraf BR, Nadel A, Bromley B. Identification of second-trimester fetuses with autosomal trisomy by use of a sonographic scoring index. Radiology 1994;193:135 – 40. [10] Spencer K, Ong C, Skentou H, Liao AW, Nicolaides HK. Screening for trisomy 13 by fetal nuchal translucency and maternal serum free beta-hCG and PAPP-A at 10 – 14 weeks of gestation. Prenat Diagn 2000;20:411 – 6. [11] Tul N, Spencer K, Noble P, Chan C, Nicolaides K. Screening for trisomy 18 by fetal nuchal translucency and maternal serum free beta-hCG and PAPP-A at 10 – 14 weeks of gestation. Prenat Diagn 1999;19:1035 – 42. [12] DeVore GR, Romero R. Genetic sonography: a cost-effective method for evaluating women 35 years and older who decline genetic amniocentesis. J Ultrasound Med 2002;21:5 – 13. [13] Ralston SJ, Wertz D, Chelmow D, Craigo SD, Bianchi DW. Pregnancy outcomes after prenatal diagnosis of aneuploidy. Obstet Gynecol 2001;97:729 – 33. [14] Birth defects surveillance data from selected states, 1995 – 1999. Teratology 2002;66:S129 – 211. [15] Alaska Bureau of Vital Statistics. Annual reports for 1992 through 1999. Alaska Department of Health and Social Services, Division of Public Health. [16] DeVore GR. The genetic sonogram: its use in the detection of chromosomal abnormalities in fetuses of women of advanced maternal age. Prenat Diagn 2001;21:40 – 5. [17] Smith-Bindman R, Hosmer W, Feldstein VA, Deeks JJ, Goldberg JD. Second-trimester ultrasound to detect fetuses with down syndrome: a meta-analysis. JAMA 2001;285:1044 – 55.