If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Correspondence and requests for materials should be addressed to Katherine Johansen Taber, Department of Medical Affairs, Myriad Women's Health, Inc, 180 Kimball Way, South San Francisco, CA 94080
The American College of Obstetricians and Gynecologists (ACOG) and the American College of Medical Genetics and Genomics (ACMG) suggest carrier screening panel design criteria intended to ensure meaningful results. This study used a data-driven approach to interpret the criteria to identify guidelines-consistent panels.
Methods
Carrier frequencies in >460,000 individuals across 11 races/ethnicities were used to assess carrier frequency. Other criteria were interpreted on the basis of published data. A total of 176 conditions were then evaluated. Stringency thresholds were set as suggested by ACOG and/or ACMG or by evaluating conditions already recommended by ACOG and ACMG.
Results
Forty and 75 conditions had carrier frequencies of ≥1 in 100 and ≥1 in 200, respectively; 175 had a well-defined phenotype; and 165 met at least 1 severity criterion and had an onset early in life. Thirty-seven conditions met conservative thresholds, including a carrier frequency of ≥1 in 100, and 74 conditions met permissive thresholds, including a carrier frequency of ≥1 in 200; thus, both were identified as guidelines-consistent panels.
Conclusion
Clear panel design criteria are needed to ensure quality and consistency among carrier screening panels. Evidence-based analyses of criteria resulted in the identification of guidelines-consistent panels of 37 and 74 conditions.
Carrier screening aims to identify couples at risk of having affected pregnancies by testing for disease-causing variants in the same autosomal genes in both members of the couple (for autosomal recessive conditions) or in a gene on the X chromosome (for X-linked conditions). The purpose of carrier screening is to identify these at-risk couples (ARCs) so that they may maximize their reproductive options.
Expanded carrier screening in reproductive medicine-points to consider: a joint statement of the American College of Medical Genetics and Genomics, American College of Obstetricians and Gynecologists, National Society of Genetic Counselors, Perinatal Quality Foundation, and Society for Maternal-Fetal Medicine.
In those screened preconceptionally, reproductive options include in vitro fertilization with preimplantation genetic testing for monogenic conditions, the use of a noncarrier donor gamete, adoption, or the avoidance of pregnancy altogether. In those screened during pregnancy, management options include prenatal diagnostic testing, pregnancy termination after a confirmed diagnosis, postnatal diagnostic testing, early preparation for the initiation of medical intervention if and when appropriate, and parental education about postbirth special care needs.
Expanded carrier screening in reproductive medicine-points to consider: a joint statement of the American College of Medical Genetics and Genomics, American College of Obstetricians and Gynecologists, National Society of Genetic Counselors, Perinatal Quality Foundation, and Society for Maternal-Fetal Medicine.
The American College of Obstetricians and Gynecologists (ACOG) currently recommends offering carrier screening to all couples, regardless of their race/ethnicity (ie, pan-ethnic carrier screening), for only 2 conditions, namely cystic fibrosis (CF) and spinal muscular atrophy (SMA) (Supplemental Table 1).
ACOG additionally recommends screening for 5 conditions, alpha-thalassemia, Hb beta chain-related hemoglobinopathy (including sickle cell disease), hexosaminidase A deficiency (Tay-Sachs disease), Canavan disease, and familial dysautonomia, for those of certain races/ethnicities (Supplemental Table 1).
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
Pan-ethnic screening for a large number of conditions, traditionally known as “expanded carrier screening” (ECS), but more recently referred to as simply “carrier screening,”
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
has been shown to more effectively identify carriers and ARCs across all races/ethnicities than does screening that is restricted to only certain races/ethnicities.
An empirical estimate of carrier frequencies for 400+ causal Mendelian variants: results from an ethnically diverse clinical sample of 23,453 individuals.
Comparing ethnicity-based and expanded carrier screening methods at a single fertility center reveals significant differences in carrier rates and carrier couple rates.
However, clearly defined carrier screening panel design criteria are needed to maximize identification of ARCs across all races/ethnicities, a goal that aligns with both ACOG’s and ACMG’s commitments to work toward the elimination of racial inequities that lead to disparate health outcomes.
American Association of Gynecologic Laparoscopists, American Board of Obstetrics and Gynecology, American College of Obstetricians and Gynecologists, et al. Joint statement - obstetrics and gynecology: collective action addressing racism. American College of Obstetricians and Gynecologists Published August 2020. https://www.acog.org/-/media/project/acog/acogorg/files/pdfs/news/commitmentendracism-historyobgyn-082720-v8.pdf. Accessed November 10, 2020.
Citing the vast number of conditions that could be screened, ACOG and ACMG have both recommended criteria for carrier screening panel inclusion. ACOG suggests 7 criteria, several of which should be met for a condition to be included on an ECS panel.
These criteria include the following: have a carrier frequency of 1 in 100 or greater, have a well-defined phenotype, have a detrimental effect on quality of life, cause cognitive or physical impairment, require surgical or medical intervention, have an onset early in life, and can be diagnosed prenatally (Table 1).
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
They address carrier frequency (high prevalence of carriers in the screened population), severity (phenotype severity may impact decision-making), genotype–phenotype relationship (predictable genotype–phenotype correlation), prenatal diagnostic capability (available prenatal diagnosis and reproductive options), and analytical validity (established analytical validity of screening methods) (Table 1).
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
High prevalence of carriers in the screened population. Carrier frequency threshold is ≥1 in 200 for autosomal recessive conditions and > 1 in 40,000 for X-linked conditions, in at least 1 subpopulation.
Subpopulations include 1 of 6 ancestral populations (African/African American, Hispanic, Ashkenazi Jewish, East Asian, non-Finnish European, and South Asian) and others that have at least 1% representation in the United States (including United States Territories).
2.
Genotype–phenotype relationship
Have a well-defined phenotype
Predictable genotype–phenotype correlation. A ClinGen gene–disease association level of at least moderate is recommended. In most cases, only pathogenic or likely pathogenic variants should be reported.
3.
Severity (quality of life)
Have a detrimental effect on quality of life
Phenotype severity that may impact decision-making. Condition severity categorizations of moderate, severe, and profound are recommended.
4.
Severity (impairment)
Cause cognitive or physical impairment
5.
Severity (intervention)
Require surgical or medical intervention
6.
Age of onset
Have the onset early in life
Not addressed
7.
Prenatal diagnostic capability
Can be diagnosed prenatally
Available prenatal diagnosis and reproductive options.
8.
Analytical validity
Not addressed
Established analytical validity of screening methods.
ACMG, American College of Medical Genetics and Genomics; ACOG, American College of Obstetricians and Gynecologists; ECS, expanded carrier screening.
a Subpopulations include 1 of 6 ancestral populations (African/African American, Hispanic, Ashkenazi Jewish, East Asian, non-Finnish European, and South Asian) and others that have at least 1% representation in the United States (including United States Territories).
For example, ACOG’s “have a well-defined phenotype” criterion does not offer detail on the mechanism by which it should be evaluated. Furthermore, it is unclear how many criteria should be met by each condition or if certain criteria must always be met. Variable interpretation of panel design criteria may explain why commercially available carrier screening panels have vastly different sizes, with limited overlap in condition inclusion,
ACMG acknowledged this problem in its Practice Resource, and for this reason, ACMG suggested specific, quantitative thresholds by which conditions should be evaluated (Table 1).
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
and leveraging carrier frequency data from >460,000 individuals across 11 ethnicities, we applied a data-driven approach to assess 176 conditions for their adherence to the criteria, resulting in the identification of a panel of conditions that conservatively meet the criteria.
Materials and Methods
Defining and applying the panel design criteria
ACOG and ACMG panel design criteria were categorized into combined criteria 1 to 8 as described in Table 1. Criteria were objectively defined and evaluated as follows, using consensus across published studies when possible. Criterion 1 was defined by carrier frequency thresholds described in ACOG and ACMG criteria and evaluated using carrier rates from 460,608 individuals across 11 races/ethnicities (as seen further on). Criterion 2 was defined as gene–disease association and evaluated by applying the ClinGen clinical validity framework, which quantifies gene–disease association by assessing the strength of available evidence,
as recommended by ACMG. Criteria 3 to 5 were defined as factors of severity and were evaluated by mapping criteria to disease traits reported in Arjunan et al.
Expanded carrier screening in reproductive medicine-points to consider: a joint statement of the American College of Medical Genetics and Genomics, American College of Obstetricians and Gynecologists, National Society of Genetic Counselors, Perinatal Quality Foundation, and Society for Maternal-Fetal Medicine.
Criterion 7 was defined as the capability to perform prenatal diagnostic testing for each condition. Criterion 8 was defined as published analytical sensitivity and analytical specificity for each condition.
Thresholds for each criterion were set on the basis of those explicitly cited by ACOG or ACMG
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
or, in the absence of an explicitly cited threshold, the degree to which conditions currently recommended by both ACOG and ACMG for screening in 1 or more race/ethnicity
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
(Supplemental Table 1) met the most conservative interpretations of each criterion. Criteria and their thresholds were then applied to 176 conditions on a commercially available ECS panel (Foresight, Myriad Women’s Health, Inc).
We retrospectively analyzed de-identified data of 495,194 individuals who were screened using next-generation sequencing for up to 176 conditions (see Supplemental Methods) during a 7-year period (January 1, 2012-February 2, 2020) and who had consented to research use of their data. To ensure carrier frequencies were representative of the general United States population, we excluded individuals who had a family or personal history of disease or reported consanguinity, resulting in a cohort of 460,608 individuals (Supplemental Table 3). Self-reported race/ethnicity was collected via the test requisition form. The races/ethnicities included in Supplemental Table 3 represent those included on the test requisition.
Carrier rate, ARC rate, and efficiency of ARC detection calculations
Brief descriptions of carrier rate, ARC rate, and efficiency of ARC detection calculations are included in this study but are described in detail in the Supplemental Methods. To calculate race/ethnicity-specific carrier rates, we computed race/ethnicity-specific allele frequencies for each pathogenic or likely pathogenic variant observed in our cohort. To determine ARC rates (Supplemental Table 4), we used modeled fetal disease risk, described in detail in Beauchamp et al
and defined as the probability of a disease affecting the conceptus of randomly selected male and female parents. Carrier rates and ARC rates for conditions with complicated inheritance were calculated as described in Ben-Shachar et al.
The square root of the carrier rate for X-linked conditions was used to enable direct comparison in carrier frequencies for autosomal recessive and X-linked conditions, except when calculating panel carrier rates and panel ARC rates. Panel carrier rates and panel ARC rates were calculated population-wide by weighting race/ethnicity-specific carrier and ARC rates according to data from the United States Census.
and ACMG’s statement that there should be a high prevalence of carriers in the screened population, with a carrier frequency threshold of ≥1 in 200 in any race/ethnicity
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
(Table 1). We therefore analyzed panel composition for both carrier frequency thresholds of ≥1 in 100 and ≥1 in 200.
Consistent with previous interpretations of ACOG’s “1 in 100” criterion, we interpreted it to mean a carrier frequency of ≥1 in 100 in any race/ethnicity.
The number of conditions in each race/ethnicity with carrier frequencies of ≥1 in 100 ranged from 10 in individuals of African or African American descent to 32 in individuals of AJ descent (Figure 1A, Supplemental Table 3). This threshold resulted in a panel of 40 conditions that included all 7 conditions currently recommended by both ACOG and ACMG for screening in 1 or more ethnicity
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
(Figure 1A, Supplemental Table 5). This 40-condition panel would capture 68.2% of carriers and 92.3% of ARCs compared with a 176-condition panel (Supplemental Table 6). At the more permissive carrier frequency of ≥1 in 200 in any race/ethnicity threshold supported by ACMG, the number of conditions ranged from 16 in individuals of Southeast Asian descent to 52 in individuals of AJ descent. This threshold resulted in a panel of 75 conditions (Figure 1A, Supplemental Table 5), capturing 82.0% of carriers and 96.6% of ARCs compared with a 176-condition panel (Supplemental Table 6).
Figure 1A. Number of conditions in each ethnicity that meet carrier frequency thresholds of ≥1 in 100 or ≥1 in 200. Conditions meeting the given threshold in each ethnicity are represented by boxes (with each box representing 1 condition). Within each ethnicity, conditions are listed vertically in order of decreasing carrier frequency. The conditions in larger boxes above the solid line are those recommended by both ACMG and ACOG for screening in 1 or more ethnicity and that meet the carrier frequency threshold. Green indicates conditions recommended in the corresponding ethnicity. B. Genotype–phenotype relationship and severity criteria met among 176 conditions. The first column represents the genotype–phenotype relationship criterion, indicating the ClinGen gene–disease association categorization of each condition. The next column represents the severity categorization as determined by Arjunan et al.
The next 3 columns indicate the 3 ACOG severity criteria related to quality of life, impairment, and intervention. Meeting 3, 2, and 1 criteria are indicated by dark purple, light purple, and pink colors, respectively. The sixth column indicates the “onset early in life” criterion. The conditions in larger boxes above the solid line are those recommended by both ACMG and ACOG for screening in 1 or more ethnicity. ACMG, American College of Medical Genetics and Genomics; ACOG, American College of Obstetricians and Gynecologists.
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
Although ACOG does not define how to quantify this relationship, ACMG recommends evaluating evidence for gene–disease association using the ClinGen framework that categorizes the strength of such evidence,
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
We therefore used ACMG’s recommended quantification and threshold for evaluating genotype–phenotype relationship. Of the 176 conditions examined here, 175 (99.4%) met the moderate or higher gene–disease association threshold
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
Applying this threshold resulted in 175 of 176 conditions (99.4%) having at least moderate severity (short-chain acyl-CoA dehydrogenase deficiency was categorized as mild) (Figure 1B).
We therefore also applied ACOG’s severity criteria and determined thresholds by examining which of these criteria were met by each of the 7 conditions recommended for screening by both ACOG and ACMG.
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
Among these 7 conditions, 6 had a detrimental effect on quality of life, 4 caused cognitive or physical impairment, and 3 required surgical or medical intervention (Figure 1B, Supplemental Table 5). Only 1 of the 7 conditions, SMA, met all 3 severity-related criteria; 4 conditions met 2 criteria; and 2 conditions met 1 criterion. No distinctive pattern could be seen in the severity-related criteria met by these conditions other than meeting at least 1 criterion. We therefore considered a condition to meet the collective severity-related threshold if it met at least 1 of ACOG’s 3 severity-related criteria.
Criterion 6 addresses age of onset (Table 1). ACOG specifies that conditions should have an onset early in life,
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
Therefore, to determine a conservative threshold for age of onset, we examined which of the 7 conditions recommended by both ACMG and ACOG for screening in 1 or more race/ethnicity
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
had an onset early in life. All 7 conditions had an onset in infancy or childhood (Figure 1B, Supplemental Table 5). We therefore set an age of onset threshold at infancy or childhood and considered conditions with an onset during adolescence not to meet this threshold.
Because previous studies have included age of onset as an indicator of severity,
we evaluated it together with the 3 other ACOG severity criteria, finding that 165 of 176 conditions (93.8%) had an onset in infancy or childhood and met at least 1 ACOG severity criterion, thus meeting the conservative severity threshold that is inclusive of infancy or childhood onset (Figure 1B, Supplemental Table 5). This 165-condition panel would capture 94.2% of carriers and 92.3% of ARCs compared with a 176-condition panel (Supplemental Table 6).
Prenatal diagnosis as a criterion for panel inclusion
Criterion 7 is based on both ACOG’s and ACMG’s statements that conditions included in ECS panels should have the capability to be diagnosed prenatally (Table 1).
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
Single-gene disorders, such as those screened by carrier screening panels, can be diagnosed by targeted genetic testing of amniocytes or chorionic villi if the particular variant carried by parents has been identified.
All 176 conditions evaluated here can be diagnosed prenatally (Supplemental Table 5), as confirmed by additional information received from diagnostic testing laboratories
Analytical validity as criterion for panel inclusion
Criterion 8 addresses analytical validity and is based on ACMG’s statement that analytical validity of the screening method should be established. We interpret this statement to refer to the screening method for each condition. All 176 conditions evaluated here have established and published analytical validity (Supplemental Table 5).
To determine the composition of a panel consistent with both ACOG and ACMG panel design criteria, we combined the thresholds for each criterion that were applied to all 176 conditions. For criteria that differed between ACOG and ACMG, we applied the more conservative and/or specific of the thresholds. The resulting thresholds were carrier frequency of ≥1 in 100 in any race/ethnicity, moderate or higher gene–disease association evidence level, meeting at least 1 of ACOG’s 3 severity criteria, having an onset in infancy or childhood, the ability for the condition to be diagnosed prenatally, and published analytical validity (Figure 2). Taken together, this guidelines-consistent panel would consist of 37 conditions (Figure 2, Table 2), would include all 7 conditions currently recommended by both ACMG and ACOG for screening in at least 1 race/ethnicity, and would capture 63.0% of carriers and 84.6% of ARCs compared with a 176-condition panel (Figure 3A, Supplemental Table 6).
Figure 2Summary of all panel design criteria met by each of 176 conditions. Conditions are sorted horizontally by carrier frequency of the ethnicity with the highest carrier frequency (shown at top), and the vertical axis indicates age of onset and American College of Obstetricians and Gynecologists (ACOG) severity criteria. The genotype–phenotype criterion is indicated by boxes colored red (definitive), orange (moderate) or clear (limited). As all 176 conditions evaluated can be diagnosed prenatally and have published analytical validity data, those criteria are not shown. Conditions within the dark gray box in the upper left represent the 37-condition conservative guidelines-consistent panel, ie, the conditions that meet carrier frequency of ≥1 in 100 in any ethnicity, moderate or definitive gene–disease association, at least 1 ACOG severity criterion, and onset in infancy or childhood. Conditions in both the dark gray and light gray boxes represent the 74-condition permissive guidelines-consistent panel, ie, conditions that meet carrier frequency of ≥1 in 200 in any ethnicity, moderate or definitive gene–disease association, moderate or higher severity, and an adolescent or earlier age of onset. Only 1 condition (short-chain acyl-CoA dehydrogenase deficiency) in the light gray box did not meet the moderate or higher overall severity criterion; this condition is denoted by a black border and an asterisk and is not included in the 74-condition panel.
X-linked condition. A 1 in 10,000 carrier frequency for X-linked conditions is equivalent to a 1 in 100 carrier frequency for autosomal recessive conditions.
X-linked condition. A 1 in 10,000 carrier frequency for X-linked conditions is equivalent to a 1 in 100 carrier frequency for autosomal recessive conditions.
X-linked condition. A 1 in 10,000 carrier frequency for X-linked conditions is equivalent to a 1 in 100 carrier frequency for autosomal recessive conditions.
X-linked condition. A 1 in 10,000 carrier frequency for X-linked conditions is equivalent to a 1 in 100 carrier frequency for autosomal recessive conditions.
X-linked condition. A 1 in 10,000 carrier frequency for X-linked conditions is equivalent to a 1 in 100 carrier frequency for autosomal recessive conditions.
X-linked condition. A 1 in 10,000 carrier frequency for X-linked conditions is equivalent to a 1 in 100 carrier frequency for autosomal recessive conditions.
X-linked condition. A 1 in 10,000 carrier frequency for X-linked conditions is equivalent to a 1 in 100 carrier frequency for autosomal recessive conditions.
X-linked condition. A 1 in 10,000 carrier frequency for X-linked conditions is equivalent to a 1 in 100 carrier frequency for autosomal recessive conditions.
X-linked condition. A 1 in 10,000 carrier frequency for X-linked conditions is equivalent to a 1 in 100 carrier frequency for autosomal recessive conditions.
X-linked condition. A 1 in 10,000 carrier frequency for X-linked conditions is equivalent to a 1 in 100 carrier frequency for autosomal recessive conditions.
X-linked condition. A 1 in 10,000 carrier frequency for X-linked conditions is equivalent to a 1 in 100 carrier frequency for autosomal recessive conditions.
X-linked condition. A 1 in 10,000 carrier frequency for X-linked conditions is equivalent to a 1 in 100 carrier frequency for autosomal recessive conditions.
X-linked condition. A 1 in 10,000 carrier frequency for X-linked conditions is equivalent to a 1 in 100 carrier frequency for autosomal recessive conditions.
A carrier frequency of 1 in 10,000 for X-linked conditions is equivalent to a carrier frequency of 1 in 100 for autosomal recessive conditions. A carrier frequency of 1 in 40,000 for X-linked conditions is equivalent to a of carrier frequency 1 in 200 for autosomal recessive conditions.
ACMG, American College of Medical Genetics and Genomics; ACOG, American College of Obstetricians and Gynecologists; Af/AA, African or African American; AJ, Ashkenazi Jewish; EA, East Asian; FrCa/Caj, French Canadian or Cajun; Hisp, Hispanic; ME, Middle Eastern; NE, Northern European; SEA, Southeast Asian; SE, Southern European.
a X-linked condition. A 1 in 10,000 carrier frequency for X-linked conditions is equivalent to a 1 in 100 carrier frequency for autosomal recessive conditions.
b Ethnicity with highest carrier frequency. Does not include Native American, Pacific Islander, Finnish, or Unknown ethnicity.
Figure 3A. Relative proportion of conditions, carriers, and ARCs detected for each of the indicated panels compared with that of a 176-condition panel. Carrier and ARC rates for panels are indicated in Supplemental Table 6 and were calculated population-wide using data from the United States Census to weight ethnicities. B. Panel ARC rate as a function of panel carrier rate. The dotted lines denote respective carrier frequency thresholds. C. Efficiency of ARC detection, which is the ratio of panel carrier rate to panel ARC rate for each of the indicated panels. ACMG, American College of Medical Genetics and Genomics; ACOG, American College of Obstetricians and Gynecologists; ARC, at-risk couple.
We also examined the composition of a panel meeting more permissive interpretations of certain criteria, including having a carrier frequency threshold of ≥1 in 200 in any race/ethnicity, a moderate or higher gene–disease association, at least moderate severity, age of onset during or before adolescence, the ability for the condition to be diagnosed prenatally, and published analytical validity. This more permissive, but still guidelines-consistent, panel would consist of 74 conditions (Figure 2) and would capture 81.4% of carriers and 96.6% of ARCs compared with a 176-condition panel (Figure 3A, Supplemental Table 6).
Because the goal of carrier screening is to maximize the number of ARCs identified per individual screened, we evaluated the number of new ARCs identified for different panels. As panels become larger in size because of inclusion of rare conditions, the number of additional ARCs identified decreases (Figure 3B). A panel with conditions that have a population-wide carrier frequency of ≥1 in 100 identifies 100 ARCs per 10,000 patients screened, and a panel with conditions that have a carrier frequency of ≥1 in 200 identifies 108 ARCs per 10,000 patients screened (Figure 3B). We also compared the efficiency of ARC detection, defined as the ratio of the carrier detection rate to the ARC detection rate for different panels; a panel with a lower ratio indicates more efficiency. The efficiency of ARC detection for the 37-condition guidelines-consistent panel was 31:1 (Figure 3C). The efficiency of ARC detection for the 74-condition guidelines-consistent panel was 35:1 (Figure 3C). Both panels were more efficient than the 7-gene ACOG/ACMG panel, for which the efficiency of ARC detection was 36:1.
Discussion
Equitable identification of carriers and ARCs across diverse populations is consistent with the joint statement by ACOG committing to eliminating racial inequities that lead to disparate health outcomes
American Association of Gynecologic Laparoscopists, American Board of Obstetrics and Gynecology, American College of Obstetricians and Gynecologists, et al. Joint statement - obstetrics and gynecology: collective action addressing racism. American College of Obstetricians and Gynecologists Published August 2020. https://www.acog.org/-/media/project/acog/acogorg/files/pdfs/news/commitmentendracism-historyobgyn-082720-v8.pdf. Accessed November 10, 2020.
ACMG cited the inequity and scientific flaws associated with race/ethnicity-based screening as justification for its recommendation that carrier screening for 112 conditions be offered to all individuals who are pregnant or considering pregnancy without regard to ethnicity.
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
Both ACOG and ACMG have published criteria that conditions should meet to be included on carrier screening panels. We considered both conservative and permissive interpretations of criteria to identify 2 guidelines-consistent panels. The conservative panel, identified by applying a carrier frequency threshold of ≥1 in 100 in any race/ethnicity (thus approximating ACMG’s tier 2 category
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
), moderate or higher gene–disease association, infancy or childhood onset, and descriptive severity criteria, resulted in the identification of a panel of 37 conditions (Figure 2, Table 1). We also identified a 74-condition permissive panel that met the carrier frequency threshold suggested by ACMG (≥1 in 200 in any race/ethnicity), moderate or higher gene–disease association, moderate or higher severity, and an age of onset during or before adolescence, thus approximating ACMG’s tier 3 category.
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
Both the 37-condition conservative panel and the 74-condition permissive panel identify a majority of ARCs in comparison with a 176-condition panel and show similar efficiency of ARC detection and therefore are reasonable clinical options for providers wishing to adhere to guidelines-based panel design criteria.
Previous publications have addressed interpretation of panel design criteria, but these were incomplete because they did not incorporate recent analyses that have used data-driven approaches to quantify each criterion.
evaluated genes from several commercially available panels for their adherence to a combination of ACOG and ACMG panel design criteria but did not include severity because a framework for evaluating severity had only recently been published.
In contrast, a 2018 study suggested that severity be used as a primary criterion for condition inclusion but did not provide a recommendation for how to evaluate severity.
interpreted both ACOG and ACMG criteria as broad categories addressing scope, impact, age of onset, and actionability but did not propose guidance on how to objectively define the criteria and evaluate conditions for their adherence to the criteria. Our study goes beyond these previous studies because it applies evidence-based definitions of all ACOG criteria and the most recent ACMG criteria to a large number of conditions.
The process of interpreting panel design criteria revealed the need for important clarifications to make them consistently applicable. For example, ACOG guidelines state that conditions included on carrier screening panels should meet several of 7 criteria,
and although ACMG provided specificity by suggesting thresholds by which to evaluate conditions, it too did not specify whether all or only certain criteria must be met.
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
For example, many of the conditions included on the ACMG tier 3 list have not been curated by ClinGen and therefore have unpublished genotype–phenotype association. Analysis of CF, a condition that has been recommended for pan-ethnic screening by both ACOG and ACMG for nearly 2 decades, suggests that it is not necessary for conditions to meet all criteria to be appropriate for panel inclusion because it met only 6 of the 8 collective criteria (Table 2). Explicit language in guidelines stating exactly which of the criteria must be met is needed to provide straightforward guidance on and to promote quality and consistency in panel design to laboratories offering carrier screening. Interpreting guidelines criteria as demonstrated in this study, in addition to carefully understanding how ACMG developed its recommended tier 3 list, provides a roadmap for guidelines-consistent iterations of carrier screening panels. Consistent criteria from both ACMG and ACOG would better facilitate implementation of guidelines by laboratories. Clear guidance is also important for providers who strive to offer guidelines-based care and for payers to ensure that the individuals they insure have access to guidelines-based care.
The ACMG recently applied its updated criteria to identify 112 conditions (113 genes counting HBA1 and HBA2 separately) that should be offered to all who are pregnant or considering pregnancy (tier 3 screening).
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
Using similar thresholds, we identified 74 conditions that met all permissive criteria applied in this study. This discrepancy in number of conditions is partially explained by differences in the conditions evaluated in this study versus those by ACMG. ACMG evaluated 415 autosomal recessive conditions, X-linked conditions listed in OMIM, ACMG’s secondary findings list,
ACMG SF v3.0 list for reporting of secondary findings in clinical exome and genome sequencing: a policy statement of the American College of Medical Genetics and Genomics (ACMG).
and a subset of conditions included in the recommended universal newborn screening panel. In contrast, our study evaluated only conditions for which data were published and/or available to assess adherence to all panel design criteria. Many of the conditions ACMG evaluated did not have specific severity classifications (291 conditions) or published ClinGen gene–disease associations (186 conditions) and therefore could not be evaluated in this study (Supplemental Table 7). Fifty-three conditions were included in both ACMG’s tier 3 panel and our permissive panel (Supplemental Figure 1). Fifty-nine were included on the ACMG tier 3 list but excluded from the permissive panel; of these, 46 did not have severity or ClinGen gene–disease association classifications available. Conversely, of the 21 conditions that were included in the permissive panel but excluded from the ACMG tier 3 list, 10 were not evaluated by ACMG.
The use of different databases—Genome Aggregation Database (gnomAD) v2.0.2 by ACMG
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
and our laboratory’s database for this study—to determine race/ethnicity-specific carrier frequencies also contributes to the discrepancy in the number of conditions on ACMG’s tier 3 list and our permissive panel. Among the conditions on the ACMG tier 3 list but not on the permissive panel, 13 had carrier frequencies of <1 in 200 in at least 1 ethnicity in the Myriad database. Eleven conditions on the permissive panel but not on ACMG’s tier 3 list had carrier frequencies of <1 in 200 in any ethnicity in the gnomAD database (or <1 in 40,000 prevalence for X-linked conditions). Differences in carrier frequencies may be due to mechanisms used to collect race/ethnicity information. GnomAD uses ancestry-specific single-nucleotide variation (formerly single-nucleotide polymorphisms) to sort individuals into categories,
whereas in our database, race/ethnicity is self-reported. The ancestry categories in gnomAD and the race/ethnicity categories in our database are only partially overlapping, resulting in the inability to compare some categories. Furthermore, poor representation of races/ethnicities can artificially increase or decrease carrier frequencies. To account for this possibility, we did not include races/ethnicities for which there were fewer than 1000 individuals; ACMG may have used a different threshold or mechanism to control for poor representation within gnomAD. Methods used to exclude those with known high risk could also result in carrier frequency differences. Our database excludes individuals who have reported high risk for carrier status, such as personal or family history of the condition. It is possible that carrier frequencies in our database could be artificially increased if those with high risk do not report such risk and are therefore inadvertently included in it. GnomAD controls for known high risk by excluding those diagnosed with severe genetic disease and their family members
Because panel size is most influenced by carrier frequency threshold, reanalyses and comparisons of carrier frequencies should be undertaken as databases grow and become more racially/ethnically diverse. We encourage laboratories to share carrier frequency data, as we have in this study, to enable robust analyses across diverse populations.
This study has limitations that should be noted. First, conditions other than those evaluated in this study may also meet guidelines criteria, making them appropriate candidates for carrier screening panel inclusion. We urge laboratories to apply the approach presented in this study to provide transparency on their panels’ adherence to criteria. Second, it is possible that neither ACOG nor ACMG intended for the criteria to be interpreted as presented in this study. However, thresholds were based on those explicitly cited by ACOG or ACMG or the degree to which conditions recommended for screening met criterion definitions. Finally, the measures previously discussed were taken to ensure that carrier frequencies were accurate and closely representative of the general population; however, we cannot be certain that they are not overestimates or underestimates.
Carrier screening panel design criteria that can be clearly interpreted and applied are needed to ensure quality and consistency of carrier screening offerings. The panels identified here are consistent with evidence-based interpretations of ACOG and ACMG panel design criteria and more effectively identify carriers and ARCs, regardless of their race/ethnicity, than past screening paradigms that relied on race and/or ethnicity to stratify care.
Data Availability
Carrier and at-risk couple frequencies reported in this study are included in Supplemental Tables 3 and 4. All other data included in this study have been previously published and have been appropriately referenced.
Conflict of Interest
Rotem Ben-Shachar, Katherine Johansen Taber, and Dale Muzzey are current employees and shareholders of Myriad Genetics, Inc. Aishwarya Arjunan, James Goldberg, Kristjan E. Kaseniit, and Raul Torres are former employees of Myriad Genetics, Inc. Haywood Brown is the Executive Medical Director of the Access to Equitable Carrier Screening Coalition, for which he receives compensation. No funding was provided for the purpose of this study. All work was conducted by authors in the course of their respective employment.
Acknowledgments
The authors wish to thank Terry Adkins, Krystal Brown, Bryan Dechairo, and Sarah Soto for valuable comments on the manuscript.
Expanded carrier screening in reproductive medicine-points to consider: a joint statement of the American College of Medical Genetics and Genomics, American College of Obstetricians and Gynecologists, National Society of Genetic Counselors, Perinatal Quality Foundation, and Society for Maternal-Fetal Medicine.
Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. Published online July 21, 2021. Published correction appears in Genet Med. August 27, 2021. https://doi.org/10.1038/s41436-021-01203-z.
An empirical estimate of carrier frequencies for 400+ causal Mendelian variants: results from an ethnically diverse clinical sample of 23,453 individuals.
Comparing ethnicity-based and expanded carrier screening methods at a single fertility center reveals significant differences in carrier rates and carrier couple rates.
American Association of Gynecologic Laparoscopists, American Board of Obstetrics and Gynecology, American College of Obstetricians and Gynecologists, et al. Joint statement - obstetrics and gynecology: collective action addressing racism. American College of Obstetricians and Gynecologists Published August 2020. https://www.acog.org/-/media/project/acog/acogorg/files/pdfs/news/commitmentendracism-historyobgyn-082720-v8.pdf. Accessed November 10, 2020.
ACMG SF v3.0 list for reporting of secondary findings in clinical exome and genome sequencing: a policy statement of the American College of Medical Genetics and Genomics (ACMG).
04136-8&id=gr1.jpg){kind=link}
04136-8&id=gr2.jpg){kind=link}
04136-8&id=gr3.jpg){kind=link}