The
n e w e ng l a n d j o u r na l
of
m e dic i n e
A New Era in Noninvasive Prenatal Testing To the Editor: In their Perspective article, Morain et al. (Aug. 8 issue)1 say that commercial companies have not met reasonable standards for demonstrating clinical validity. This is not correct. In five independent cohorts, researchers have examined 463 pregnancies with fetuses with Down’s syndrome, 187 with trisomy 18, and 37 with trisomy 13. To maximize confidence in sensitivity estimates, all were high-risk pregnancies (in a general population, more than 250,000 pregnancies would have had to be studied). Five professional organizations, including the American Congress of Obstetricians and Gyneco logists,2 recommend offering such testing for “high-risk” pregnancies. The Perspective article also implies that a positive predictive value (PPV) of 62% might be inadequate for decision making. Using the authors’ numbers, we modeled the performance of cell-free fetal DNA (cfDNA) testing in 200,000 pregnancies at a risk level of 1 in 200. We assumed that the cfDNA test had a sensitivity of 99.9% and a specificity of 99.7%. Of the pregnant women tested, only 1596 (0.8%) would be candidates for diagnostic studies, and 999 of 1000 cases would be detected. More than 198,000 women would have a negative screen, which would result in approximately 660 fewer procedure-related fetal losses. Suggesting that testing is unreliable because it is not approved by the Food and Drug Administration (FDA) may discourage its valid use in avoiding unnecessary invasive diagnostic testing.
prenatal testing. N Engl J Med 2013;369:499-501. [Erratum, N Engl J Med 2013;369:683.] 2. American College of Obstetricians and Gynecologists Committee on Genetics. Committee opinion no. 545: noninvasive prenatal testing for fetal aneuploidy. Obstet Gynecol 2012;120:1532-4. DOI: 10.1056/NEJMc1311604
To the Editor: Morain et al. do not cite the results of a postmarketing clinical laboratory study in which nearly 6000 samples were tested at the Verinata Health laboratory.1 In this initial laboratory population, there was a 4.8% detection of aneuploidy, which is similar to the percentage detected in our prior clinical validation trial in which only samples from high-risk pregnancies were used.2 Data from the study show a PPV of 0.994 and a negative predictive value of 0.9996 for chromosome 21. Furthermore, Morain et al. did not mention the major advantage of cfDNA testing, which is that more than 90% of women tested receive a result of “no aneuploidy.” In the past 18 months, this testing has already had a great effect on clinical care by substantially reducing the number of invasive procedures needed to clarify a positive screening test. We agree with the authors that results from ongoing prospective studies of patients with a lower risk of fetal aneuploidy are needed to further inform safe and effective clinical implementation.3,4 Amy J. Sehnert, M.D. Richard P. Rava, Ph.D.
Edward R. Ashwood, M.D.
Verinata Health Redwood City, CA [email protected]
ARUP Laboratories Salt Lake City, UT
Diana W. Bianchi, M.D.
Glenn E. Palomaki, Ph.D. Women and Infants Hospital Providence, RI [email protected] Drs. Ashwood and Palomaki report participating in a clinical validation study of maternal plasma DNA testing for aneuploidy sponsored by Sequenom; Dr. Ashwood reports that his laboratory refers specimens to a commercial laboratory for noninvasive prenatal testing; Dr. Palomaki reports participating in a current clinical validation study of maternal plasma DNA testing for aneuploidy sponsored by Natera and serving as a consultant for Natera. No other potential conflict of interest relevant to this letter was reported.
2164
1. Morain S, Greene MF, Mello MM. A new era in noninvasive
Tufts Medical Center Boston, MA Drs. Sehnert and Rava report being employees of Verinata Health, and Dr. Bianchi reports receiving consulting fees for serving on the Verinata Clinical Advisory Board. No other potential conflict of interest relevant to this letter was reported. 1. Futch T, Spinosa J, Bhatt S, de Feo E, Rava RP, Sehnert AJ.
Initial clinical laboratory experience in noninvasive prenatal testing for fetal aneuploidy from maternal plasma DNA samples. Prenat Diagn 2013;33:569-74. 2. Bianchi DW, Platt LD, Goldberg JD, Abuhamad AZ, Sehnert AJ, Rava RP. Genome-wide fetal aneuploidy detection by maternal plasma DNA sequencing. Obstet Gynecol 2012;119:890-901. 3. ClinicalTrials.gov. Comparison of Aneuploidy Risk Evalu
n engl j med 369;22 nejm.org november 28, 2013
The New England Journal of Medicine Downloaded from nejm.org on August 13, 2015. For personal use only. No other uses without permission. Copyright © 2013 Massachusetts Medical Society. All rights reserved.
correspondence ations (CARE). 2012-2013 (http://clinicaltrials.gov/ct2/show/ NCT01663350). 4. Idem. Investigation of the impact of noninvasive prenatal testing for fetal aneuploidy on utilization of prenatal diagnostic procedures and pregnant women’s views. 2013 (http://clinicaltrials.gov/ct2/show/NCT01708746). DOI: 10.1056/NEJMc1311604
the testing was unreliable because it is not FDAapproved. We did point out that current regulations permit marketing of laboratory-developed tests without the need to demonstrate their clinical validity. The study by Futch et al. cited by Sehnert et al. had not yet been published when we finalized our article. This is the type of large study that will provide the data needed to calculate all the performance characteristics that we called for in our article. These published data are reassuring, but we note that the delivery dates for only half of the investigators’ tested cohort had passed by the time of publication. Thus, ascertainment of outcomes is incomplete, and positive and negative predictive values are not yet calculable. Michael F. Greene, M.D.
The Authors Reply: The studies cited by Ashwood and Palomaki regarding cfDNA noninvasive prenatal testing techniques certainly were necessary to demonstrate the sensitivity and specificity of the various techniques for the detection of the studied fetal trisomies, and the studies were conducted efficiently. We maintain, however, that these exploratory studies alone do not provide information on all the performance characteristics (positive and negative predictive values) of the testing that is sufficient for provid- Massachusetts General Hospital ers to make informed decisions about whether to Boston, MA offer the testing or for patients to make decisions Michelle M. Mello, J.D., Ph.D. about whether to accept it. Mathematical model- Harvard School of Public Health ing of these predictive values from preliminary Boston, MA studies in an artificially composed population Stephanie Morain, M.P.H. with a disease frequency on the order of 1 in 8 is Harvard University very different from a demonstration of the true Cambridge, MA performance of the testing in a validation study Since publication of their article, the authors report no furof actual use in a large population with a disease ther potential conflict of interest. frequency of 1 in 200. We did not suggest that DOI: 10.1056/NEJMc1311604
The Great East Japan Earthquake and Out-of-Hospital Cardiac Arrest To the Editor: The most powerful earthquake ever recorded in Japan, with a magnitude of 9.0 on the Richter scale, struck the northeast region of Japan on March 11, 2011 (Fig. 1A).1,2 Earthquakes are known to increase the risk of sudden cardiac arrest.3 Using a nationwide populationbased database, we assessed the association between this earthquake and the occurrence of outof-hospital cardiac arrest of presumed cardiac origin.4,5 Our study involved adult patients with out-ofhospital cardiac arrest of presumed cardiac origin who were transported to medical institutions in Iwate, Miyagi, and Fukushima — the three prefectures that were most damaged by the earth-
quake (Fig. 1A). The study included data from 4 weeks before to 8 weeks after March 11 in each year from 2005 to 2011. Weekly changes in the occurrence of out-of-hospital cardiac arrest after the earthquake were assessed with the use of Poisson regression models to calculate risk ratios comparing the observed numbers of patients with out-of-hospital cardiac arrests in each week with the expected numbers of such events. Expected numbers of events were estimated with the use of data from 2005 to 2010 on the number of events, according to the weekly average ambient temperature, the week, and the year during that period. Risk ratios were also calculated according to sex and age group.
n engl j med 369;22 nejm.org november 28, 2013
The New England Journal of Medicine Downloaded from nejm.org on August 13, 2015. For personal use only. No other uses without permission. Copyright © 2013 Massachusetts Medical Society. All rights reserved.
2165
n e w e ng l a n d j o u r na l
of
m e dic i n e
A New Era in Noninvasive Prenatal Testing To the Editor: In their Perspective article, Morain et al. (Aug. 8 issue)1 say that commercial companies have not met reasonable standards for demonstrating clinical validity. This is not correct. In five independent cohorts, researchers have examined 463 pregnancies with fetuses with Down’s syndrome, 187 with trisomy 18, and 37 with trisomy 13. To maximize confidence in sensitivity estimates, all were high-risk pregnancies (in a general population, more than 250,000 pregnancies would have had to be studied). Five professional organizations, including the American Congress of Obstetricians and Gyneco logists,2 recommend offering such testing for “high-risk” pregnancies. The Perspective article also implies that a positive predictive value (PPV) of 62% might be inadequate for decision making. Using the authors’ numbers, we modeled the performance of cell-free fetal DNA (cfDNA) testing in 200,000 pregnancies at a risk level of 1 in 200. We assumed that the cfDNA test had a sensitivity of 99.9% and a specificity of 99.7%. Of the pregnant women tested, only 1596 (0.8%) would be candidates for diagnostic studies, and 999 of 1000 cases would be detected. More than 198,000 women would have a negative screen, which would result in approximately 660 fewer procedure-related fetal losses. Suggesting that testing is unreliable because it is not approved by the Food and Drug Administration (FDA) may discourage its valid use in avoiding unnecessary invasive diagnostic testing.
prenatal testing. N Engl J Med 2013;369:499-501. [Erratum, N Engl J Med 2013;369:683.] 2. American College of Obstetricians and Gynecologists Committee on Genetics. Committee opinion no. 545: noninvasive prenatal testing for fetal aneuploidy. Obstet Gynecol 2012;120:1532-4. DOI: 10.1056/NEJMc1311604
To the Editor: Morain et al. do not cite the results of a postmarketing clinical laboratory study in which nearly 6000 samples were tested at the Verinata Health laboratory.1 In this initial laboratory population, there was a 4.8% detection of aneuploidy, which is similar to the percentage detected in our prior clinical validation trial in which only samples from high-risk pregnancies were used.2 Data from the study show a PPV of 0.994 and a negative predictive value of 0.9996 for chromosome 21. Furthermore, Morain et al. did not mention the major advantage of cfDNA testing, which is that more than 90% of women tested receive a result of “no aneuploidy.” In the past 18 months, this testing has already had a great effect on clinical care by substantially reducing the number of invasive procedures needed to clarify a positive screening test. We agree with the authors that results from ongoing prospective studies of patients with a lower risk of fetal aneuploidy are needed to further inform safe and effective clinical implementation.3,4 Amy J. Sehnert, M.D. Richard P. Rava, Ph.D.
Edward R. Ashwood, M.D.
Verinata Health Redwood City, CA [email protected]
ARUP Laboratories Salt Lake City, UT
Diana W. Bianchi, M.D.
Glenn E. Palomaki, Ph.D. Women and Infants Hospital Providence, RI [email protected] Drs. Ashwood and Palomaki report participating in a clinical validation study of maternal plasma DNA testing for aneuploidy sponsored by Sequenom; Dr. Ashwood reports that his laboratory refers specimens to a commercial laboratory for noninvasive prenatal testing; Dr. Palomaki reports participating in a current clinical validation study of maternal plasma DNA testing for aneuploidy sponsored by Natera and serving as a consultant for Natera. No other potential conflict of interest relevant to this letter was reported.
2164
1. Morain S, Greene MF, Mello MM. A new era in noninvasive
Tufts Medical Center Boston, MA Drs. Sehnert and Rava report being employees of Verinata Health, and Dr. Bianchi reports receiving consulting fees for serving on the Verinata Clinical Advisory Board. No other potential conflict of interest relevant to this letter was reported. 1. Futch T, Spinosa J, Bhatt S, de Feo E, Rava RP, Sehnert AJ.
Initial clinical laboratory experience in noninvasive prenatal testing for fetal aneuploidy from maternal plasma DNA samples. Prenat Diagn 2013;33:569-74. 2. Bianchi DW, Platt LD, Goldberg JD, Abuhamad AZ, Sehnert AJ, Rava RP. Genome-wide fetal aneuploidy detection by maternal plasma DNA sequencing. Obstet Gynecol 2012;119:890-901. 3. ClinicalTrials.gov. Comparison of Aneuploidy Risk Evalu
n engl j med 369;22 nejm.org november 28, 2013
The New England Journal of Medicine Downloaded from nejm.org on August 13, 2015. For personal use only. No other uses without permission. Copyright © 2013 Massachusetts Medical Society. All rights reserved.
correspondence ations (CARE). 2012-2013 (http://clinicaltrials.gov/ct2/show/ NCT01663350). 4. Idem. Investigation of the impact of noninvasive prenatal testing for fetal aneuploidy on utilization of prenatal diagnostic procedures and pregnant women’s views. 2013 (http://clinicaltrials.gov/ct2/show/NCT01708746). DOI: 10.1056/NEJMc1311604
the testing was unreliable because it is not FDAapproved. We did point out that current regulations permit marketing of laboratory-developed tests without the need to demonstrate their clinical validity. The study by Futch et al. cited by Sehnert et al. had not yet been published when we finalized our article. This is the type of large study that will provide the data needed to calculate all the performance characteristics that we called for in our article. These published data are reassuring, but we note that the delivery dates for only half of the investigators’ tested cohort had passed by the time of publication. Thus, ascertainment of outcomes is incomplete, and positive and negative predictive values are not yet calculable. Michael F. Greene, M.D.
The Authors Reply: The studies cited by Ashwood and Palomaki regarding cfDNA noninvasive prenatal testing techniques certainly were necessary to demonstrate the sensitivity and specificity of the various techniques for the detection of the studied fetal trisomies, and the studies were conducted efficiently. We maintain, however, that these exploratory studies alone do not provide information on all the performance characteristics (positive and negative predictive values) of the testing that is sufficient for provid- Massachusetts General Hospital ers to make informed decisions about whether to Boston, MA offer the testing or for patients to make decisions Michelle M. Mello, J.D., Ph.D. about whether to accept it. Mathematical model- Harvard School of Public Health ing of these predictive values from preliminary Boston, MA studies in an artificially composed population Stephanie Morain, M.P.H. with a disease frequency on the order of 1 in 8 is Harvard University very different from a demonstration of the true Cambridge, MA performance of the testing in a validation study Since publication of their article, the authors report no furof actual use in a large population with a disease ther potential conflict of interest. frequency of 1 in 200. We did not suggest that DOI: 10.1056/NEJMc1311604
The Great East Japan Earthquake and Out-of-Hospital Cardiac Arrest To the Editor: The most powerful earthquake ever recorded in Japan, with a magnitude of 9.0 on the Richter scale, struck the northeast region of Japan on March 11, 2011 (Fig. 1A).1,2 Earthquakes are known to increase the risk of sudden cardiac arrest.3 Using a nationwide populationbased database, we assessed the association between this earthquake and the occurrence of outof-hospital cardiac arrest of presumed cardiac origin.4,5 Our study involved adult patients with out-ofhospital cardiac arrest of presumed cardiac origin who were transported to medical institutions in Iwate, Miyagi, and Fukushima — the three prefectures that were most damaged by the earth-
quake (Fig. 1A). The study included data from 4 weeks before to 8 weeks after March 11 in each year from 2005 to 2011. Weekly changes in the occurrence of out-of-hospital cardiac arrest after the earthquake were assessed with the use of Poisson regression models to calculate risk ratios comparing the observed numbers of patients with out-of-hospital cardiac arrests in each week with the expected numbers of such events. Expected numbers of events were estimated with the use of data from 2005 to 2010 on the number of events, according to the weekly average ambient temperature, the week, and the year during that period. Risk ratios were also calculated according to sex and age group.
n engl j med 369;22 nejm.org november 28, 2013
The New England Journal of Medicine Downloaded from nejm.org on August 13, 2015. For personal use only. No other uses without permission. Copyright © 2013 Massachusetts Medical Society. All rights reserved.
2165
![[Non-invasive prenatal testing: a new era has already come].](/assets/img/hold.png)
