Modeled Fetal Risk of Genetic Diseases Identified by Expanded Carrier Screening

Importance  Screening for carrier status of a limited number of single-gene conditions is the current standard of prenatal care. Methods have become available allowing rapid expanded carrier screening for a substantial number of conditions.

Objectives  To quantify the modeled risk of recessive conditions identifiable by an expanded carrier screening panel in individuals of diverse racial and ethnic backgrounds and to compare the results with those from current screening recommendations.

Design, Setting, and Participants  Retrospective modeling analysis of results between January 1, 2012, and July 15, 2015, from expanded carrier screening in reproductive-aged individuals without known indication for specific genetic testing, primarily from the United States. Tests were offered by clinicians providing reproductive care.

Exposures  Individuals were tested for carrier status for up to 94 severe or profound conditions.

Main Outcomes and Measures  Risk was defined as the probability that a hypothetical fetus created from a random pairing of individuals (within or across 15 self-reported racial/ethnic categories; there were 11 categories with >5000 samples) would be homozygous or compound heterozygous for 2 mutations presumed to cause severe or profound disease. Severe conditions were defined as those that if left untreated cause intellectual disability or a substantially shortened lifespan; profound conditions were those causing both.

Results  The study included 346 790 individuals. Among major US racial/ethnic categories, the calculated frequency of fetuses potentially affected by a profound or severe condition ranged from 94.5 per 100 000 (95% CI, 82.4-108.3 per 100 000) for Hispanic couples to 392.2 per 100 000 (95% CI, 366.3-420.2 per 100 000) for Ashkenazi Jewish couples. In most racial/ethnic categories, expanded carrier screening modeled more hypothetical fetuses at risk for severe or profound conditions than did screening based on current professional guidelines (Mann-Whitney P < .001). For Northern European couples, the 2 professional guidelines-based screening panels modeled 55.2 hypothetical fetuses affected per 100 000 (95% CI, 51.3-59.3 per 100 000) and the expanded carrier screening modeled 159.2 fetuses per 100 000 (95% CI, 150.4-168.6 per 100 000). Overall, relative to expanded carrier screening, guideline-based screening ranged from identification of 6% (95% CI, 4%-8%) of hypothetical fetuses affected for East Asian couples to 87% (95% CI, 84%-90%) for African or African American couples.

Conclusions and Relevance  In a population of diverse races and ethnicities, expanded carrier screening may increase the detection of carrier status for a variety of potentially serious genetic conditions compared with current recommendations from professional societies. Prospective studies comparing current standard-of-care carrier screening with expanded carrier screening in at-risk populations are warranted before expanded screening is adopted.

Genetic testing of prospective parents to detect carriers of specific inherited recessive diseases is part of routine obstetrical practice. Current recommendations by professional organizations are to test for a limited number of individual diseases in part based on self-reported racial/ethnic background (Table 1).^1,2 Multiplex platforms simultaneously assaying many potentially pathogenic variants on each sample have become available, allowing rapid expanded carrier screening for a large number of conditions. These advances could facilitate screening for an expanded number of conditions independent of racial/ethnic background.

Expanded carrier screening was first made available in late 2009. A 2013 review³ of this initial experience examining 23 453 individuals found that 24.0% of individuals were carriers for at least 1 of 96 serious recessive conditions. This information in part led the American College of Medical Genetics and Genomics (ACMG), the American Congress of Obstetricians and Gynecologists (ACOG), the National Society of Genetic Counselors, the Perinatal Quality Foundation, and the Society for Maternal-Fetal Medicine to issue a joint statement on expanded carrier screening, suggesting a need for more data on “the frequency of variants in previously untested racial and ethnic groups” to facilitate general panethnic testing.²

The purpose of this study was to assess expanded panethnic carrier screening in 15 self-reported racial/ethnic categories using either targeted genotyping or next-generation sequencing for up to 94 profound or severe conditions.

Box Section Ref ID

The Western Institutional Review Board granted exemption from further review of this study because the work was performed only on deidentified existing medical records. As recommended by the ACMG,⁴ a generic consent form supplemented by online information about each disease was used. The consent document included information regarding the benefits and limitations of screening, as well as consent to allow use of anonymized testing data for research purposes.

Individuals in this study had expanded carrier screening for genetic disorders. The test requisition form also included an indication for testing. To minimize bias, individuals with known personal or family history of a genetic disorder, those with infertility, and those selecting other as the reason for testing were excluded from the analysis (Figure).

Tests were offered and ordered by the individual’s physician, typically a general obstetrician or a specialist in reproductive endocrinology or infertility, maternal-fetal medicine, or genetics. Testing was primarily funded by a third-party payer, the tested individual, or a combination of the 2. Test requisition forms required individuals to self-report their racial/ethnic background, choosing 1 of 15 possible categories. Racial/ethnic categories included an option for unknown and an option for mixed ethnicity. These options were chosen on the basis of specific professional guidelines (eg, Ashkenazi Jewish), known genetic founder populations (eg, Finnish), or US Census categorizations (eg, Hispanic). Posttest genetic counseling was available at no additional cost.

Conditions were classified as having profound, severe, moderate, or mild severity based on a literature review and the algorithm described in Lazarin et al.⁵ Use of this method aimed to overcome the challenges associated with classifying rare, unfamiliar conditions by assigning severity based on widely recognized characteristics. The assumption was made that the pathogenic or likely pathogenic variants occurring in the genes corresponding to profound or severe conditions were of clinical importance. Genetic variants with known mild phenotype, uncertain significance, known incomplete penetrance, or all 3 were also excluded (eTable 6 in the Supplement). Of 110 genes tested that are associated with 112 conditions, 18 conditions are associated with moderate, mild, or highly variable phenotypes and were excluded from the analysis, leaving 94 profound or severe conditions to be analyzed (eTable 9 in the Supplement).

Diseases were further categorized based on the inclusion in the carrier screening guidelines issued by the ACMG or the ACOG and by the inclusion in the ACMG core newborn screening panel⁶ (eTable 9 in the Supplement).

All samples were tested using the Family Prep Screen (Counsyl), which screened for carrier status in up to 110 genes causing autosomal or X-linked recessive conditions (eTable 9 in the Supplement). For most genes, 2 testing platforms were available. The first was a targeted genotyping test using TaqMan assays on the Fluidigm 96.96 platform to genotype 417 disease-causing variants. The second was a next-generation sequencing test using custom hybrid capture followed by sequencing on the Illumina HiSeq 2500 to test for variants in select exons and intronic regions of the same genes tested by the targeted genotyping assay. Both technologies have demonstrated analytical sensitivity and specificity greater than 99%.^7,8

Two genes were assayed separately for both platforms: (1) the SMN1 exon 7 deletion causing spinal muscular atrophy was tested using quantitative polymerase chain reaction and (2) the FMR1 5′UTR expansion causing fragile X syndrome was tested using polymerase chain reaction with capillary electrophoresis.⁹ Samples were accepted as whole blood, saliva, or extracted DNA. The laboratory was certified under the Clinical Laboratory Improvement Amendments (05D1102604), accredited by the College of American Pathologists Laboratory Accreditation Program (7519776), and received a permit from New York state (8535).

Physicians could select the testing platform used and could customize the set of conditions screened; therefore, some individuals were not tested for all conditions. The number tested for each condition is detailed in the Supplement data file on the tab “chromosome frequencies.”

Individual variants were classified according to ACMG guidelines, which was the consensus standard in the field at the time of the analysis.¹⁰ Consistent with these guidelines, most variants required literature evidence to be considered reportable. However, novel, putative loss-of-function variants could be considered for pathogenicity without specific literature evidence. Specific details and classification categories are described in section 3 in the Supplement. The laboratory protocol was to report only known and likely pathogenic variants. Variants of uncertain significance and benign variants were not considered in this analysis.

Additional details regarding the statistical methods used appear in section 4 in the Supplement. Data from the targeted genotyping and next-generation sequencing assays were unified to estimate the chromosome carrier frequency (the probability that a random chromosome carries a pathogenic variant) for each condition in each racial/ethnic category. Risk was defined as the probability of a hypothetical fetus being affected and homozygous or compound heterozygous for pathogenic variants for a given condition. Risk was computed using the carrier frequencies accounting for varying inheritance patterns.

Risk of fragile X syndrome was weighted by allele size, and only male fetuses were assumed to be potentially affected. The 95% CIs were derived by repeating risk calculations on 100 000 samples from the carrier frequency posterior distributions to compute a posterior distribution over risk and taking the 2.5 and 97.5 percentiles. Distributions over ratios of risks (eg, spinal muscular atrophy vs cystic fibrosis in eTable 1 in the Supplement and between positive rates for different racial/ethnic categories) were computed by dividing 100 000 pairs of computed risks from the relevant individual disease and guideline or panel posterior distributions.

All calculations were performed using Python version 2.7.9 (Python Software Foundation) and the numpy (version 1.10.2), scipy (version 0.15.1), and sympy (version 0.7.6) libraries. Unless otherwise stated, significance testing used the 2-sided Mann-Whitney test on samples from the posterior distribution of risk or ratios of samples as appropriate. The level of significance used was α = .05, with Bonferroni correction for multiple hypothesis testing.

The results compare the prevalence of potential conditions using modeled risks. This metric allows simpler comparison of a large number of conditions than do commonly used metrics like carrier frequency and carrier couple frequency (section 4.3.1 in the Supplement provides details on the limitations of these metrics). The main analysis results are for couples in which both members are of the same racial/ethnic category (computations for all pairings are included in section 7 in the Supplement). Carrier frequency, carrier couple frequency, and modeled risk are provided in the Supplement for all diseases and pairings of racial/ethnic categories.

Between January 1, 2012, and July 15, 2015, 430 584 individuals underwent expanded carrier screening (including data from 23 453 individuals that were previously reported³). After excluding patients with testing indications other than for routine carrier testing, 346 790 individuals remained (Figure and eTable 3 in the Supplement). The main analysis of data includes 11 of 15 self-reported racial/ethnic categories in which at least 5000 individuals were tested under this indication. The data from racial/ethnic categories with fewer than 5000 individuals are included in section 7 in the Supplement.

There were 308 668 tests performed by targeted genotyping and 38 122 by next-generation sequencing (Table 2). Of the tested individuals, 78.9% were female. Among those tested by next-generation sequencing, females were relatively less common (60.3%; Fisher exact P < .001). The largest racial/ethnic category was mixed or other Caucasian (n = 85 451). The smallest racial/ethnic category included in the main analysis was Southeast Asian (n = 6263). Finnish was the smallest racial/ethnic category overall; however, there were only 213 individuals so these data are not included in the main analysis but are included in the Supplement.

For each racial/ethnic category, the risks of particular sets of diseases appear in Table 3. Among African or African American couples, 5.2 fetuses per 100 000 (95% CI, 3.7-6.9 per 100 000) were modeled to be affected by cystic fibrosis, whereas there were 1.1 fetuses per 100 000 (95% CI, 0.7-1.8 per 100 000) affected by a condition on the ACMG-recommended panel for Ashkenazi Jewish couples. The columns in Table 3 are disjoint. Among African or African American couples, there were 3.1 fetuses per 100 000 (95% CI, 2.2-4.3 per 100 000) modeled to be affected by a condition on the ACMG core newborn screening panel (panel did not include cystic fibrosis or hemoglobinopathies; Table 3). Similarly, there were 19.2 fetuses per 100 000 (95% CI, 14.7-24.7 per 100 000) modeled to be affected by a profound or severe condition in the expanded carrier panel, which did not include any of the conditions appearing in columns 2 through 7 of Table 3. Ratios of risk are presented in eTable 1 in the Supplement.

The ACOG recommends universally offering cystic fibrosis carrier screening. Similarly, the ACMG recommends offering universal carrier screening for spinal muscular atrophy. Although in the Northern European population only 26% (95% CI, 23%-30%) as many hypothetical fetuses were modeled for spinal muscular atrophy vs cystic fibrosis (P < .001), in other racial/ethnic categories the ratio of spinal muscular atrophy to cystic fibrosis was much higher (eTable 1 in the Supplement). It was 67% (95% CI, 49%-88%) among Hispanic couples, 106% (95% CI, 70%-154%) among African or African American couples, and up to 1389% (95% CI, 450%-3594%) among East Asian couples. The smallest absolute risk of spinal muscular atrophy was found in African or African American couples with 5.3 affected fetuses per 100 000 (95% CI, 4.2-6.6 per 100 000). The largest risk of spinal muscular atrophy was found among Southern European couples (15.1 affected fetuses per 100 000 [95% CI, 11.7-19.0 per 100 000]).

Fragile X syndrome was modeled to be more common than spinal muscular atrophy in all racial/ethnic categories other than Southeast Asian couples (P < .001 for all comparisons). Among Northern European couples, 28.2 fetuses per 100 000 (95% CI, 22.2-35.1 per 100 000) were modeled to be affected by fragile X syndrome compared with 11.5 fetuses per 100 000 (95% CI, 10.3-12.6 per 100 000) modeled to be affected by spinal muscular atrophy.

Fragile X syndrome also was modeled to be more common than cystic fibrosis in most racial/ethnic categories, with the exception of Northern European and mixed or other Caucasian couples (P < .001 for all). For example, among Ashkenazi Jewish couples, 54.7 per 100 000 fetuses (95% CI, 38.3-74.8 per 100 000) were modeled to be affected by fragile X syndrome vs 36.2 per 100 000 fetuses (95% CI, 31.4-41.5 per 100 000) modeled to be affected by cystic fibrosis. The risk of fragile X syndrome for non-Ashkenazi European couples was 65% to 124% (95% CI, 50%-218%) of the risk of cystic fibrosis.

Severe and profound disorders not included in either carrier or newborn screening guidelines were more common than cystic fibrosis or spinal muscular atrophy in all racial/ethnic categories (Table 3; P < .001 in all racial/ethnic categories). The highest prevalence of this subset of conditions was found in Ashkenazi Jewish couples in which 115.4 per 100 000 fetuses (95% CI, 105.8-125.9 per 100 000) were modeled to be affected by 1 of these diseases; the lowest prevalence was found in the African or African American couples, with a corresponding rate of 19.2 per 100 000 fetuses (95% CI, 14.7-24.7 per 100 000).

The total modeled risk of profound and severe conditions was partitioned for each racial/ethnic category according to the guideline status and severity of the condition by which a fetus would be affected (Table 4). Among African or African American couples, 87% (95% CI, 84%-90%) of fetuses would be affected by a condition within the guideline recommendations for that category; 2% (95% CI, 2%-2%) would be affected by a profound condition outside the guideline recommendations; and 11% (95% CI, 8%-14%) would be affected by a severe condition outside the guideline recommendations. In addition, among African or African American couples, 1 in 275 fetuses (95% CI, 1 in 255 to 1 in 297 fetuses) were modeled to be affected by a tested profound or severe condition (Table 4). These data appear cumulatively in Table 5.

In most racial/ethnic categories, the cumulative risk of severe and profound conditions outside the guideline recommendations was greater than the risk identified by guideline-based panels (Table 5; P < .001 for all racial/ethnic categories). Among the Ashkenazi Jewish population, which has the most extensive ethnicity-specific panel, 55% (95% CI, 52%-59%) of fetuses with profound or severe phenotypes would not be identified using a guidelines-based panel (392.2 per 100 000 fetuses [95% CI, 366.3-420.2 per 100 000] affected by a profound or severe condition and 174.8 per 100 000 fetuses [95% CI, 164.2-185.9 per 100 000] affected by conditions detectable by the ACOG and ACMG panel guidelines; Table 3). Among Middle Eastern couples, 91% (95% CI, 87%-94%) of affected fetuses would not be identified using a guidelines-based panel (193.8 per 100 000 fetuses [95% CI, 149.0-248.1 per 100 000] affected by a profound or severe condition and 17.7 per 100 000 fetuses [95% CI, 12.9-23.6 per 100 000] affected by conditions detectable by ACOG and ACMG panel guidelines; Table 3).

Diseases not included in the guideline recommendations were minor contributors to total risk for African or African American couples (13%; 95% CI, 10%-16%) and for Southeast Asian couples (21%; 95% CI, 13%-30%) (Table 3). This was because the ACOG guidelines for these 2 racial/ethnic categories include testing for hemoglobinopathies. Hemoglobin variants in these racial/ethnic categories (HbS variant in African or African American couples and HBA variants in Southeast Asian couples) had higher frequencies than any other pathogenic variants tested (eg, nearly 1 in 10 for HbS in African or African American couples). Excluding hemoglobinopathies from consideration (Table 3 and eTable 2 in the Supplement) resulted in a pattern similar to other racial/ethnic categories.

Among African or African American couples, 82% (95% CI, 76%-86%) of risk for nonhemoglobinopathy profound and severe disease fell outside both the ACOG and ACMG guidelines (47.0 per 100 000 fetuses [95% CI, 35.9-60.9 per 100 000] would be affected by a nonhemoglobinopathy condition outside the guideline recommendations vs 10.5 per 100 000 fetuses [95% CI, 8.6-12.6 per 100 000] affected by cystic fibrosis or spinal muscular atrophy). Among Southeast Asian couples, 81% (95% CI, 71%-89%) of this risk fell outside guideline recommendations (42.6 per 100 000 fetuses [95% CI, 26.5-65.2 per 100 000] would be affected by a nonhemoglobinopathy condition outside the guideline recommendations vs 9.7 per 100 000 fetuses [95% CI, 6.6-13.6 per 100 000] affected by cystic fibrosis or spinal muscular atrophy).

In all racial/ethnic categories except African or African American and Southeast Asian, an expanded carrier panel including the severe and profound diseases outside the guideline recommendations would be expected to detect at least twice as many hypothetical fetuses compared with the ACOG and ACMG panel recommendations (Table 5 and eTable 1 in the Supplement; P < .001 for all racial/ethnic categories). Among these racial/ethnic categories, the smallest proportional change but largest absolute change was found among Ashkenazi Jewish couples. The ACOG and ACMG panel recommendations detected 174.8 hypothetical fetuses per 100 000 (95% CI, 164.2-185.9 per 100 000), but the rate including all profound and severe conditions was 392.2 hypothetical fetuses per 100 000 (95% CI, 366.3-420.2 per 100 000) (Table 5). The largest proportional change was in East Asian couples, among whom the rate was 7.6 per 100 000 fetuses (95% CI, 5.9-9.5 per 100 000) for the ACOG and ACMG panel recommendations but the expanded carrier screening positive rate was 129.9 per 100 000 fetuses (95% CI, 107.0-156.2 per 100 000). Both of these differences were statistically significant (Mann-Whitney P < .001 for both).

The last row of Table 5 shows the largest ratio among pairs of racial/ethnic categories in modeled positive rates. Both the cumulative risk detected by the ACOG and ACMG panel recommendations and the fraction of severe and profound disorders identified by this panel varied significantly among different racial/ethnic categories (P < .001 for all pairs). The ACOG and ACMG guideline panels exhibited a 42.2-fold difference (95% CI, 32.6- to 54.0-fold difference) in the rate of detecting an affected fetus between the highest frequency category (African or African American) and the lowest frequency category (East Asian) (Table 5; P < .001). An expanded panel including profound and severe conditions not included in the guideline recommendations also had different positive rates between categories (P < .001 for all pairs), but reduced the disparity to a maximum of 4.2-fold difference (95% CI, 3.6- to 4.9-fold) between Ashkenazi Jewish and Hispanic couples.

This study of 346 790 individuals undergoing expanded carrier screening provides insights on carrier frequencies for many rare conditions in a large, diverse, albeit selected population. The findings showed that an expanded testing panel identified more hypothetical fetuses at risk for severe or profound phenotypes than did testing based on current screening guidelines. This was not only because expanded carrier screening included additional disorders but also because guideline-based testing was based in part on self-identified racial/ethnic categories. The data further suggested that the guidelines recommended by the ACOG and the ACMG at the time of the study did not perform equally between racial/ethnic categories, resulting in differing residual risk among different racial/ethnic categories. In particular, expansion of the screening panel reduced the maximum intercategory positive rate disparity from 42.2-fold to 4.2-fold, suggesting that much of the difference in residual risk may be driven by panel composition rather than true variation in the interethnic rate of genetic disease. Therefore, even though current guidelines target a number of diseases prevalent in those of European descent (such as cystic fibrosis), they do not identify risk for other conditions that may be important to diverse populations. Expanded carrier screening revealed that many non-European racial/ethnic categories have a risk of a profound or severe genetic disease that may not be detected by the guidelines in place at the time of this analysis.

It has been challenging to determine which disorders to include in prenatal screening guidelines. Suggestions include that universally screened diseases should have a relatively severe natural course and, according to ACMG, that they should be of a nature “that most at-risk patients and their partners would consider having a prenatal diagnosis.”^2,11 The effect of early diagnosis on the potential health of the child is also an important consideration because early interventions have been shown to improve outcomes for some disorders but have not been demonstrated for others.¹² Previous methods, such as those used by the ACMG to determine its core newborn screening panel, have found categorizing the severity of each disease particularly challenging.⁶ This study used the general characteristics of each disorder to classify it as mild, moderate, severe, or profound, using a method shown to be both time efficient and highly concordant with prior classification methods.⁵ However, this classification is based on limited available data regarding the phenotypic spectrum of these disorders. For many of the rarer conditions, genotype-phenotype correlations have been derived from very small known populations of affected individuals who manifest severe disease, and applicability to a general population is unknown.

The difficulty in choosing conditions to be screened is demonstrated by the discord between national organizations regarding population screening for single-gene disorders. The ACOG recommends only cystic fibrosis for universal screening, whereas the ACMG recommends both cystic fibrosis and spinal muscular atrophy testing to be offered to all individuals. The carrier frequency of spinal muscular atrophy is well established to be between 1 in 50 and 1 in 100, significantly lower than the cystic fibrosis carrier frequency in those with European race/ethnicity.^3,13 However, the much lower frequency of cystic fibrosis carriers without European race/ethnicity indicates spinal muscular atrophy is relatively more common among those with nonwhite race/ethnicity.

This study also brings additional data to the debate on population screening for fragile X syndrome.^14-17 The characterization of the FMR1 allele size distribution for each racial/ethnic category allowed calculation of size-adjusted risk for fragile X syndrome (ie, risk calculation accounting for the variable probability of full mutation transmission depending on FMR1 CGG repeat count). In every race/ethnicity category other than Southeast Asian race/ethnicity, fragile X syndrome was modeled to be more common than spinal muscular atrophy (P < .001), and in most race/ethnicity categories fragile X syndrome was modeled to be more common than cystic fibrosis (eg, per 100 000 fetuses in Ashkenazi Jewish couples, 54.7 affected by fragile X syndrome vs 36.2 affected by cystic fibrosis). Research suggests that early intervention may improve outcomes in children affected by fragile X syndrome, but also that most of these children are not diagnosed until after the period in which intervention can help.^18-22 This large population risk and potential treatability suggests reconsideration of fragile X syndrome population (carrier or newborn) screening.

Limitations of this study included the use of US professional guidelines for many of the comparisons. However, the disease-specific frequencies provided in the Supplement should allow comparisons with other screening approaches. For comparative purposes, the study analyzed an artificial construct by calculating disease frequencies in an affected hypothetical fetus resulting from random mating (within a racial/ethnic category in the main analysis included in the article and across racial/ethnic categories in the Supplement). Disease frequencies observed in actual populations may differ from the numbers reported herein for a number of reasons, including the uncertainty inherent in modeling assumptions, the racial/ethnic admixture common in many modern societies, and the fact that the study population was drawn from an available retrospective source rather than in a random prospective manner.

In addition, the analysis adopted assumptions regarding the outcomes associated with putatively pathogenic variants in rare conditions. Mounting evidence illustrates the difficulty in correlating genotype and phenotype for certain rare and common conditions.^23-25 The current study did not prospectively assess the actual outcomes of the individuals tested. Therefore, before assertions regarding the clinical utility of broadly testing for these variants can be made with certainty, additional data are needed from unselected diverse populations on the phenotypic spectrum and for the health consequences of pathogenic variants associated with rare conditions.

This study did not examine the costs of expanded carrier screening or traditional screening panels because the rapidly declining cost of next-generation sequencing has made this difficult to assess. A recent study of a 14-gene panel found next-generation sequencing testing to be cost-effective compared with no screening or targeted mutation analysis.²⁶

In a population of diverse races and ethnicities, expanded carrier screening may increase the detection of carrier status for a variety of potentially serious genetic conditions compared with current recommendations from professional societies. Prospective studies comparing current standard-of-care carrier screening with expanded carrier screening in at-risk populations are warranted before expanded screening is adopted.

Corresponding Author: Imran S. Haque, PhD, 180 Kimball Way, South San Francisco, CA 94080 (research@counsyl.com).

Author Contributions: Dr Wapner had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Haque, Lazarin, Evans, Wapner.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Haque, Lazarin, Goldberg, Wapner.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Haque, Wapner.

Obtaining funding: Evans.

Administrative, technical, or material support: Haque, Lazarin, Goldberg, Wapner.

Study supervision: Evans, Goldberg, Wapner.

Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Drs Haque, Kang, Evans, Goldberg, and Mr Lazarin are employees of Counsyl, a laboratory providing expanded carrier screening. Dr Wapner reported receiving grants paid to his institution from Natera Inc, Sequenom, Ariosa Diagnostics Inc/Roche, Illumina Inc, and KellBenx; receiving personal fees from Natera Inc, Sequenom, Ariosa Diagnostics Inc/Roche, and Illumina Inc; and being on the board of directors for Perinatal Quality Foundation, which is an unpaid position. No other disclosures were reported.

Funding/Support: This study was funded by Counsyl, a laboratory providing expanded carrier screening.

Role of the Funder/Sponsor: The nonauthor, nonacknowledged individuals employed by the sponsor had no role in design and conduct of the study; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. Sponsor employees (laboratory personnel) were involved in the sample handling and laboratory processing for collection of the genetic data comprising the original medical record that was analyzed for this study, but nonauthor sponsor employees had no role in the collection, management, analysis, and interpretation of the data covered by this study that was exempted by an institutional review board.

Previous Presentations: Parts of this work have previously been presented at the annual meetings of the European Society of Human Genetics, June 6-9, 2015, Glasgow, Scotland (poster presentation); the American Society of Human Genetics, October 6-10, 2015, Baltimore, Maryland (poster presentation); the American Society for Reproductive Medicine, October 17-21, 2015, Baltimore, Maryland (oral platform presentation); and the International Society of Prenatal Diagnosis, July 12-15, 2015, Washington, DC (oral platform presentation).

Additional Contributions: We thank the following Counsyl employees for their assistance, none of whom received additional compensation for their role in this work: Dale Muzzey, PhD, and Greg Hogan, PhD, for helpful feedback on data presentation, and Marianna Raia, MS, CGC, and Holly Bellerose, MS, CGC, for their work in classifying the severity of diseases on the panel.