Genome screening, reporting, and genetic counseling for healthy populations

Abstract

Rapid advancements of genome sequencing (GS) technologies have enhanced our understanding of the relationship between genes and human disease. To incorporate genomic information into the practice of medicine, new processes for the analysis, reporting, and communication of GS data are needed. Blood samples were collected from adults with a PCR-confirmed SARS-CoV-2 (COVID-19) diagnosis (target N = 1500). GS was performed. Data were filtered and analyzed using custom pipelines and gene panels. We developed unique patient-facing materials, including an online intake survey, group counseling presentation, and consultation letters in addition to a comprehensive GS report. The final report includes results generated from GS data: (1) monogenic disease risks; (2) carrier status; (3) pharmacogenomic variants; (4) polygenic risk scores for common conditions; (5) HLA genotype; (6) genetic ancestry; (7) blood group; and, (8) COVID-19 viral lineage. Participants complete pre-test genetic counseling and confirm preferences for secondary findings before receiving results. Counseling and referrals are initiated for clinically significant findings. We developed a genetic counseling, reporting, and return of results framework that integrates GS information across multiple areas of human health, presenting possibilities for the clinical application of comprehensive GS data in healthy individuals.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00439-022-02480-7.

Introduction

Rapid advancements in the field of genomic medicine have prompted discussions about the integration of technologies, such as genome sequencing (GS), into clinical and personalized medicine (Prokop et al. 2018). In the clinical context, GS can increase diagnostic rates and alter medical management in previously undiagnosed pediatric and adult patients (Wise et al. 2019). GS may also be used opportunistically to identify hereditary predisposition to several conditions for the purposes of preventative screening and management (de Wert et al. 2021). GS may reveal primary findings that explain an aspect of a patient’s medical or family history, or secondary findings (SF) unrelated to patient history. Currently, The American College of Medical Genetics and Genomics (ACMG) recommends the return of SF related to risk for medically actionable conditions (Miller et al. 2021). However, as genomic knowledge continues to grow, so do the possibilities for the use of GS data outside of current recommendations and standards of care. In addition, the cost of GS and analysis continue to decline (Weymann et al. 2017), presenting the possibility for it to be used more broadly. Finally, relevant stakeholders support the return of SF from GS. Systematic reviews of studies focusing on return of SF show that the majority of patients, research participants, and society at large wish to learn all types of SF from genetic investigations, including findings considered to be actionable and non-actionable (Delanne et al. 2019). Motivations for learning GS results range from establishing a diagnosis or explanation for symptoms in patients to understanding future disease risks for currently healthy individuals and their families (East et al. 2019). Healthcare professionals exhibit similar support for the return of actionable SF, but take a more cautious approach to the disclosure of non-actionable SF (Delanne et al. 2019). Professional societies and organizations acknowledge the need for stringent protocols and informed consent surrounding the return of SF through research (Lewis et al. 2021), as well as careful consideration of the risks (e.g., psychological harm, disparities in accessibility, and genetic discrimination) and benefits (e.g., prevention of serious genetic disease) of opportunistic GS in the general population (de Wert et al. 2021).

To date, more comprehensive approaches to GS, analysis, and reporting have been applied in the research context. In addition to clinically significant SF related to personal disease risks and carrier status for genetic disorders, research participants have the opportunity to learn about pharmacogenomic variants that alter drug metabolism and impact medications (Cochran et al. 2021; Vassy et al. 2015). Information on blood type, platelets, and red blood cell antigens has also been reported to research participants undergoing GS (Vassy et al. 2015). Advancements in genome-wide associations studies (GWAS) have allowed for calculations of polygenic risk scores (PRS) pertaining to risk for multifactorial conditions like coronary artery disease and type 2 diabetes. Studies have taken different approaches to communicating PRS to research participants. PRS for common conditions have been presented as both absolute and relative risks, descriptively and visually, through physical reports as well as interactive web portals (Marjonen et al. 2021; Linderman et al. 2016; Brockman et al. 2021). Direct-to-consumer (DTC) companies, such as 23andMe (https://www.23andme.com/), promise to provide information about patients’ future health by presenting results related to PRS and risk for monogenic conditions such as hereditary cancers and cardiomyopathies. Reports may also include information on genetic ancestry and traits, such as taste, ear wax type, and athletic ability, all for a cost of approximately $200 USD (Bellcoss et al. 2012). However, DTC testing is often based on SNP-chip genotyping and rarely involves pre- or post-test genetic counseling, increasing the risk of misdiagnosis, false diagnosis, and failure to receive appropriate medical care because of misinterpretation of results by users (Horton et al. 2019). To our knowledge, there is no process published to date encompassing the return of all of these elements from GS data to participants in a single report.

The GENCOV study presents a unique opportunity to explore a process for the return of comprehensive GS results in a large cohort of ostensibly healthy individuals. We describe our framework for genetic counseling, reporting, and return of research GS results beyond the scope of current clinical recommendations and discuss opportunities to explore the clinical utility and impact on health outcomes in the future.

Materials and methods

A cohort of ostensibly healthy individuals ≥ 18 years of age from the GENCOV study (target N = 1500) provided blood samples at baseline (hospitalized inpatients only), 1, 6, and 12 months after PCR-positive SARS-CoV-2 (COVID-19) diagnosis. GS was performed. An intake survey was administered at baseline (before pre-test counseling) to collect pertinent medical history information. A detailed protocol describing participant recruitment and study design is published elsewhere (Taher et al. 2021).

Genome analysis and interpretation

GS was performed using next-generation sequencing on the Illumina NovaSeq 6000 platform through The Center for Applied Genomics (TCAG) at the Hospital for Sick Children in Toronto, Ontario. Data processing and analysis is performed using two different pipelines: (1) the Franklin by Genoox platform (genoox.com; GRCh37/hg19 alignment) and (2) locally following Genome Analysis Toolkit best practices (GATK 3.7; GRCh38/hg38 alignment). The Franklin by Genoox platform is used primarily for the purposes of variant filtration, analysis, and identification of variants for reporting. The secondary, local process uses Python and bash scripts to apply the same filtering parameters (outlined below and in Supplementary File 6) to a file of variants compiled by following a small-variant annotation pipeline described by Nalpathamkalam et al. (2014). Filtration parameters were developed and applied exclusively to capture substitutions, small or large insertions, deletions, duplications, or indels (1) previously classified as likely pathogenic or pathogenic in ClinVar®; (2) with aggregated and/or internal allele frequencies ≤ 5%; and, (3) in genes with established gene-disease relationships. Copy-number variants (CNVs) were compiled separately following the CNV annotation pipeline described by Trost et al. (2018). CNVs were filtered for inclusion based on the presence of known exonic and/or OMIM morbid genes, as well as the International Standards for Cytogenomic Arrays (ISCA) Consortium triplo- or haplo-sensitivity scores (Riggs et al. 2012). CNVs with > 90% overlap with known benign, polymorphic CNV regions were excluded (Supplementary File 7). CNV size (i.e., > 20 kb) and frequency (i.e., < 1%) were considered in the review process. Variants were further filtered by quality parameters such as confidence of calls, availability of published evidence, and participant preferences for SF.

Comprehensive gene panels were developed to aid in variant filtration and assessment. The medically actionable gene panel was developed according to the ACMG list of reportable SF versions 3.0 and v3.1 (Miller et al. 2021; ACMG SF v3.1, unpublished), The Clinical Genome Resource (ClinGen) (Rehm et al. 2015), and Reble et al. (2021), for a total of 208 genes. Reporting of medically actionable variants was not limited to the genes in this list and was subject to inclusion upon review by the study team. All other genes (n = 6226) associated with carrier status and/or rare Mendelian conditions were included in comprehensive disease panels developed through referencing publicly available gene data sets (Table 1), including ClinGen (Rehm et al. 2015), Clinical Genomics Database (Solomon et al. 2013), Online Mendelian Inheritance in Man® (OMIM), BabySeq (Ceyhan-Birsoy et al. 2017), and The Gene Curation Coalition (GenCC) (DiStefano et al. 2022) gene databases. For the complete gene database, please refer to Supplementary File 1.

Table 1.

Summary of publically available data sources used in the GENCOV study

Database name	Data type	Use in present study
The Clinical Genome Resource (ClinGen) (Rehm et al. 2015)	- Clinical validity, usefulness, and pathogenicity of genomic variants	- Development of comprehensive gene panels - Variant analysis and interpretation
The Clinical Genomics Database (CGD) (Solomon et al. 2013)	- Known gene-disease relationships organized by affected organ system(s) - Clinical utility including availability of medical interventions	- Development of comprehensive gene panels
Online Mendelian Inheritance in Man® (OMIM)	- Known human genes and associated phenotypes for established Mendelian disorders	- Development of comprehensive gene panels - Variant analysis and interpretation - Disease and familial risk information - General resource for letter/report writing and genetic counseling
BabySeq (Ceyhan-Birsoy et al. 2017)	- Catalog of genes with putative pediatric relevance	- Development of comprehensive gene panels
The Gene Curation Coalition (GenCC) (DiStefano et al. 2022)	- Gene-disease validity - Curated and harmonized	- Development of comprehensive gene panels
Pharmcogenomics Knowledge Base (PharmGKB) (Hewett et al. 2002)	- Genomic, phenotypic, and clinical information related to pharmacogenetics (variability in drug response due to inherited genetic differences)	- Development of pharmacogenomic output/tables - Variant analysis and interpretation
Clinical Pharmacogenetics Implementation Consortium (CPIC) (Relling et al. 2011)	- Clinical guidelines for medication prescription and dosage based on pharmacogenomic findings	- Development of pharmacogenomic output/tables - Variant analysis and interpretation
HapMap3 (Altshuler et al. 2010)	- Population reference data set of common human DNA variants	- Genetic ancestry estimation and output
1000Genomes*	- Population reference data set of common human DNA variants	- Alternative dataset for genetic ancestry estimation
GeneReviews® (Adam et al. 1993–2022)	- Summarized information on inherited conditions, including clinical descriptions, diagnosis, management, and genetic counseling	- Variant analysis and interpretation - Disease and familial risk information - General resource for letter/report writing and genetic counseling
ClinVar® variant database of the National Center for Biotechnology Information	- Pathogenicity of variants and phenotypic relationships - Submitter information and supporting data	- Variant analysis and interpretation
Genome Aggregation Database (gnomAD®)	- GS and exome sequencing data as well as calculated allele frequencies from unrelated control individuals	- Variant analysis and interpretation

^*URL: https://www.internationalgenome.org; GS Genome sequencing, SNP Single-nucleotide polymorphism

The study team followed a standard operating procedure for novel variant assessment, which included stepwise processes for applying pre-defined filters (Supplementary Files 6 and 7) and accumulating evidence for or against pathogenicity based on ACMG guidelines (Richards et al. 2015; Riggs et al. 2020). Variants determined to be benign, likely benign, or of uncertain significance were excluded from the final report. CNVs were reported in the context of personal disease risks only. Publicly available sources including GeneReviews® (Adam et al. 1993–2022) and OMIM® (https://www.omim.org/) were used to compile disease and familial risk information.

Genetic counseling and patient-facing materials

An intake survey was designed to collect participant demographic and health history data, as well as baseline knowledge of and attitudes toward GS (Table 2). Surveys were designed and administered by email using a secure web-based platform called Novi survey (https://novisurvey.net/). The survey was developed based on a general genetics intake and the International Severe Acute Respiratory and emerging Infection Consortium (ISARIC) COVID-19 case report form (URL: https://isaric.org/wp-content/uploads/2021/02/ISARIC-WHO-COVID-19-CORE-CRF_EN.pdf). Medical history questions were reviewed by relevant stakeholders (GC, Geneticists, Nurse Practitioners, Physicians, and Biochemists) and selected for inclusion based on applicability to the GENCOV study cohort.

Table 2.

Data variables collected through the online intake survey

Variable type	Examples of variables collected*
Demographics	Age, sex, gender, education/income level, ethnicity, employment
Medical and health history	Previous genetic testing, hereditary conditions, autoimmune/inflammatory conditions, diabetes (type 1/2), hepatic/renal/cardiac/lung issues, gastrointestinal disorders, cancer, neurological or neuromuscular disorders, blood/iron disorders, viral infections, blood pressure, body mass index, smoking habits, diet, medications
COVID-19 symptoms and vaccination status	Symptom type (e.g., cough, sore throat) as well as date of onset and duration, type (e.g., Pfizer) and number of vaccine doses, date of dose and resulting onset/duration of post-vaccination symptoms (if applicable), hospitalization status and emergency department admission
Knowledge and health outcomes	Knowledge questions pertaining to GS and COVID-19 serology, satisfaction with decision pertaining to GS information, attitudes toward genetics and healthcare, generalized hospital anxiety and depression and quality of life scales
Participant preferences	Initial preferences for learning COVID-19 serology results and SF from GS

^*Data collection occurs before pre-test genetic counseling; GS: Genome sequencing

A digital genomics platform was provided to participants before pre-test counseling (www.geneticsadviser.com) (Shickh et al. 2022). A PowerPoint presentation was developed by the study GC for use during pre-test counseling webinars (Supplementary File 2). The content of the presentation was developed to reflect the different categories of SF (medically actionable, drug reactions, PRS for common health conditions, carrier status, and other rare genetic conditions). Additional slides were developed to outline GENCOV-specific information (i.e., how genetic variation relates to COVID-19 susceptibility/symptom severity, logistics about return of results, and follow-up). Pre-test counseling webinars are conducted twice weekly using a HIPPA-compliant healthcare Zoom account (zoom.us). Each session was approximately 30 min in length and consists of a maximum of 20 participants. Personal Zoom or telephone appointments were also conducted upon request.

A consultation letter (Supplementary File 3) was developed by the study GC with input from relevant stakeholders (i.e., clinical/molecular Geneticist) for the purposes of summarizing pertinent findings from the GS report. For participants who receive results counseling, the consultation letter contains a summary of the discussion with the GC and plans for clinical follow-up if applicable. Consultation letters are forwarded to the family physician and/or referral clinic with a copy of the genomic report with participant consent.

Development of the comprehensive genome report

Comprehensive genome reports were modeled off of clinical reports from the Sinai Health System Laboratory (Toronto, ON), and reflect the current recommendations for reporting of clinical GS results (Miller et al. 2021; Green et al. 2013). In addition to monogenic disease risks and carrier status, reports include expanded information related to pharmacogenomics, PRS, genetic ancestry, as well as blood, human leukocyte antigen (HLA), and viral lineage.

Reporting of pharmacogenomics variants was based on the Pharmcogenomics Knowledge Base (PharmGKB) (Hewett et al. 2002), and includes annotations of variant-drug interactions, clinical implications, and dosing recommendations. The PharmGKB output is appended to star alleles called from Stargazer v1.0.8 using participant GS data (Lee et al. 2019). The genotypes of 17 pharmacogenes were identified, including structural variant analysis in CYP2D6. Additionally, the genotypes of two HLA loci, HLA-A and HLA-B, as well as rs12777823 were included. Custom Python scripts were used for assigning appropriate PharmGKB recommendations for each pharmacogene based on genotyping results. Genotypes that did not meet the Clinical Pharmacogenetics Implementation Consortium (CPIC) (Relling et al. 2011) and/or PharmGKB criteria for Level A and Level 1A evidence for a given gene–medication association were excluded from the final report.

Reporting of PRS results was based on previously outlined patient experiences and preferences (Brockman et al. 2021), as well as previously validated, ancestry-adjusted PRS assays for six common conditions (type 2 diabetes, coronary artery disease, atrial fibrillation, as well as breast [female only], prostate [male only], and colon cancer) (Hao et al. 2022). Raw scores were adjusted for ethnicity using ancestry-informative principal components to calculate an adjusted PRS (Hao et al. 2022).

Reads for 22 HLA loci, including HLA-A, -B, -C. -DQA1, -DQB1, -DPB1, and -DRB1, were extracted and aligned to the HLA v2 database from the HLA-VBSeq package, which was used to estimate the most probable HLA genotype from GS data to a four digit resolution (Mimori et al. 2019). HLA alleles associated with increased autoimmune disease risk were identified through a review of literature available through PubMed®. The final list included seven autoimmune diseases with significant HLA associations (type 1 diabetes, celiac disease, rheumatoid arthritis, ankylosing spondylitis, Behcet’s disease, multiple sclerosis, and Graves’ disease). An HLA–disease association database was developed with specific HLA haplotypes/genotypes, literature references, risk and/or odds ratios (OR), p values, and interpretation of findings (Supplementary file 4). HLA calls from HLA-VBSeq were assigned to the appropriate autoimmune disease association using custom Python scripts.

ADMIXTURE software (Alexander et al. 2009) was used to estimate patient ancestry against the HapMap3 dataset, which contains 1.6 million common single-nucleotide polymorphisms (SNPs) in 1184 reference individuals from 11 populations (Altshuler et al. 2010). Bash scripting and PLINK software were used to remove duplicate SNPs as well as chromosome and position mismatches. Variants used for ancestry estimation were found to be in linkage disequilibrium (LD) with an r2 > 0.2 in a 50 kb window. Variants were excluded if they fell within genomic ranges of known high-LD structure. In-house scripts were used to generate a visual summary of the output.

Red blood cell (RBC) and platelet antigens were predicted using participant GS data, described elsewhere (Lane et al. 2016). Predicted ABO and Rh blood types, rare RBC, and/or platelet antigens, in addition to implications for blood donation and transfusion, were presented following a framework for reporting of GS results for a generally healthy individual (Vassy et al. 2015).

SARS-CoV-2 viral lineage was determined by viral GS at the Ontario Institute of Cancer Research. Viral lineage follows standardized Phylogenetic Assignment of Named Global Outbreak (Pango) lineages and World Health Organization (WHO) nomenclature. Raw, de-identified viral and GS data were shared with open and controlled access databases with participants’ consent as outlined in Table 3. Finally, all reporting elements were compiled automatically into a final report using scripts developed in-house. Multiple quality control steps were taken to assure that participant data were correctly compiled according to study ID(s).

Table 3.

Mandatory and optional data-sharing with controlled and open access databases

Database name	Description	Location	Data sharing*		Access type
			Mandatory	Optional	Open	Closed
HostSeq Database at TCAG	Anonymized participant GS and clinical data	Canada	✓			✓
Global Initiative on Sharing Avian Flu Data (GISAID)	Anonymized raw COVID-19 viral sequencing data, host sex and age	Germany	✓			✓
National Center for Biotechnology Information (NCBI)	Anonymized raw COVID-19 viral sequencing data	United States	✓		✓
Host Genetics Initiative (HGI) at the Broad Institute	Anonymized participant GS and clinical data	United States		✓		✓

^*Informed consent obtained from participant for data-sharing; TCAG The Centre for Applied Genomics, GS Genome sequencing

Results

A complete outline of the final workflow for pre-test counseling and return of results is summarized in Fig. 1 and described in detail below.

Fig. 1.

Workflow for genetic counseling and return of GS results through the GENCOV study. (1) Study participants complete an online intake survey to collect information on medical history and COVID-19 symptoms, as well as an educational digital genomics platform wherein they are able to indicate initial preferences for return of SF from GS. The participant completes pre-test counseling with the study GC, either in group format by teleconference or by personal telephone and/or video appointment. The study GC confirms and records participants’ preferences for SF from GS. (2) GS data are filtered and analyzed based on participant preferences. A final GS report is compiled and reviewed by the study team. (3) The GC and Geneticist meet to review reports to determine which participants require results counseling and clinical follow-up. Individuals with clinically significant findings meet with the study GC (and Geneticist, where applicable) by teleconference to discuss their results and the plan for follow-up. The GC writes the consultation letter and initiates clinical referral(s). The family physician is forwarded a copy of the report, letter, and referrals with participants’ consent. Result counseling appointments and referrals are not initiated for other findings; however, participants will receive a copy of their report along with a letter summarizing their results with the option to speak with the study GC if they have questions about their report. †Clinically significant findings include monogenic disease risks for rare Mendelian conditions as well as medically actionable conditions. Other findings include carrier status results, PRS for common conditions, pharmacogenomics, ancestry, HLA blood group, and viral lineage, GC Genetic counselor, SF Secondary findings, GS Genome sequencing

Intake and pre-test genetic counseling

Before a participant’s pre-test genetic counseling appointment, the GC reviews intake survey data as well as completion status and responses to the digital genomics platform. The GC takes note of relevant information related to medical history and initial SF preferences. Participants are provided a secure login for the digital genomics platform, as well as a link to register for the group webinar, before their pre-test counseling appointment. Before starting the webinar, the GC provides an overview of the session and describes measures taken to ensure participant confidentiality. Next, the GC outlines general information on genes and hereditary conditions, potential SF, and other research-related information that may be included in the final report, implications and limitations of GS, as well as possibilities for follow-up. The webinar concludes with an anonymous question and answer session. Upon completion of the webinar, the GC follows up briefly with each attendee by phone to answer any additional questions as well as confirm preferences for SF and return of results. Preferences and a record of completion of pre-test counseling are recorded in the participant’s study file.

Comprehensive genome report

A comprehensive genomic report was developed for return of results through the GENCOV study. The report sections, including SF from GS and additional research findings, are summarized in Fig. 2. A mock example of a complete report can be viewed in Supplementary File 5.

Fig. 2.

Summary of the elements of the comprehensive GS report: (1) Clinically significant findings related to the participant’s risk for hereditary conditions (monogenic disease risks, including medically actionable and rare Mendelian conditions); (2) findings relevant to reproductive planning (carrier status); (3) pharmacogenomic variants; (4) PRS for common conditions; (5) HLA genotype; (6) genetic ancestry; (7) blood group genotype; (8) COVID-19 viral lineage; (9) testing methodologies and limitations; and (10) informational appendices. PRS Polygenic risk score, Rh Rhesus, GS Genome sequencing, HLA Human leukocyte antigen

The first page of the report states the participant’s elected SF categories and provides a concise summary of the genomic findings. Pathogenic and likely pathogenic variants associated with personal disease risks and reproductive planning are presented in a table format that includes the gene and transcript, variant coding and protein change, zygosity, associated disease, inheritance pattern, control population frequency, and variant classification. Detailed variant interpretations, as well as specific disease information and familial risk are provided. Recommendations are made for genetic counseling and clinical follow-up of research results. Likely benign, benign and variants of uncertain significance were not included in the report.

Variants in genes with established evidence at the time of analysis (PharmGKB levels 1A and CPIC levels A) for altered drug metabolism are reported. The gene name, genotype, rsID, medications, and metabolizer phenotype are included for non-normal metabolizer phenotypes (e.g., poor or rapid metabolizer). An expanded pharmacogenomics table with specific implications and dosing recommendations, as well as general information on gene-medication interactions is provided in the appendix.

PRS for common health conditions are reported dichotomously as an “average” or “increased” polygenic risk for disease. An increased polygenic risk is equivalent to a greater than two-fold increased risk for the condition (Hao et al. 2022). PRS results are supplemented with information on the general population lifetime risk and a description of the condition, as well as an appendix with additional resources for participants including informational websites and suggestions for lifestyle modifications that may help to mitigate risk.

The participant’s predicted HLA genotype (e.g., HLA-A*03:01) is provided and known autoimmune disease associations are indicated based on HLA type where applicable. If there are no known disease associations for the participant’s HLA genotype, they are considered “average risk.” If the participant has an HLA genotype with a known disease association, the table will include an “increased” risk for the associated disease along with evidence from the literature (sample size and OR) and a brief description of the disease. Blood group genotyping includes the participant’s ABO blood type and Rhesus (Rh) antigen in addition to expanded genotyping, which includes rare red blood cell antigens (e.g., KDAS +) and human platelet antigens. Reports indicate the relevance of HLA and blood group type to COVID-19 susceptibility and outcomes, as well as blood product donation and transfusions. Links are provided to external resources for rare blood type registration and stem cell donation in Canada.

The top ten ancestry associations are reported in the form of a color-coded pie chart with percentage values representing each population group. The chart is accompanied by a legend that defines the population group name and associated color. The majority of the participant’s estimated genetic ancestry is clearly stated above the figure (e.g., “By looking at multiple genetic changes present in your DNA, we estimate the majority of your genetic ancestry to be Chinese (Reference population: Chinese in Metropolitan Denver, Colorado (CHD))”).

Return of results and clinical follow-up

Participants are given the option to receive their report by standard mail or by secure file transfer over email. The report is accompanied by the consultation letter, which summarizes the primary research findings and, if applicable, plans for clinical follow-up. The study GC and clinical Geneticist meet to review pertinent cases and develop a structure for referrals and follow-up before meeting with the participant over Zoom or telephone conference. Results counseling is provided to participants with clinically significant findings, including risks for medically actionable and rare Mendelian conditions. During the results counseling session, the study GC collects a detailed medical and family history, delivers results, and discusses the plan for follow-up. The family physician is copied on the final report and any referrals made by the study team with the participant’s consent. Counseling and referrals for results related to carrier status, pharmacogenomics findings, PRS, and/or additional research findings (e.g., ancestry, HLA, blood group, etc.) are not initiated by the study team; however, participants may request an appointment with the study GC to discuss their results if they wish.

Discussion

We developed a genetic counseling and reporting framework for the return of comprehensive GS results to a large cohort of ostensibly healthy individuals. Although aspects of our report have been returned to participants before, never have they been presented collectively within a single, all-inclusive document and accompanied by a counseling and referral structure. Particularly novel is the return of a genetic ancestry estimate and HLA genotype within the research context. Individuals may seek ancestry testing to discover information about their ethnic background or genealogy. Determining genetic ancestry is important, because there are often discrepancies between a patient’s self-reported ethnicity and genetic ancestry, which can impact risk calculations in the context of reproductive counseling (Shraga et al. 2017). Therefore, clinical applications of reporting genetic ancestry may include identifying founder populations with high carrier frequencies for autosomal recessive conditions to ensure accurate risk assessments (Kirkpatrick and Rashkin 2017). HLA type has been reported specifically in the context of pharmacogenomic interactions to avoid adverse reactions to medications (e.g., HIV-positive patients with HLA*B57:01 are hypersensitive to Abacavir), but also has clinical use in the context of organ transplantation, blood product donations, and transfusions (Fung and Benson 2015). For instance, HLA-matching improves graft and host survival, as well as reduces alloimmunization rates in solid organ and hematopoietic stem cell transplantation (Fung and Benson 2015). Universal reporting and registration of HLA and blood group genotype, alongside genetic ancestry, may increase the identification of unrelated organ and stem cell donors, especially for non-White/European individuals. HLA type has also been associated with increased risk for autoimmune conditions such as celiac disease (Schweiger et al. 2016). Reporting of HLA type may provide an at least partial underlying explanation for a patient’s chronic condition.

Counseling and educational materials were important given the novelty and complexity of the information being returned to participants. Due to limited time and resources, it was not feasible to provide results counseling to all study participants. Patient-facing materials are helpful when it comes to increasing understanding of genomic information (Dwyer et al. 2021). To ensure appropriate understanding and interpretation of results, we developed appendices with additional information and resources on each section (i.e., ancestry, HLA, pharmacogenomics, blood type, etc.). For example, the PRS appendix contains sections titled “What is a polygenic risk score?” and “How can I reduce my risk?” as well as a list of external resources and websites with information on common health conditions like breast and prostate cancer. Likewise, we describe what HLA and genetic ancestry are in lay language, along with important limitations of the testing and results. We anticipate that this will reduce the amount of time the study GC spends counseling participants who otherwise do not require clinical follow-up. In addition to detailed, patient-friendly appendices, the consultation letter is provided to all participants, regardless if they meet with the GC or not, which summarizes the primary SF (i.e., pharmacogenomics results, carrier status, and monogenic disease risks). The consultation letter also suggests that participants contact their family physician should they have questions about their current medications in light of pharmacogenomics findings, or if they are actively family planning and wish to seek a referral to a local Genetics clinic for reproductive counseling. In the future, we aim to assess the understandability and overall acceptance of our reports and patient-facing materials through participant surveys and interviews. We also plan to assess participants’ uptake of clinical referrals, healthcare services (genetic and non-genetic), and screening recommendations after completion of the GENCOV study. This type of longitudinal follow-up is essential to understand the impact of GS results on health outcomes and behaviors (e.g., disease diagnoses and smoking cessation), as well as the use of healthcare resources. We may also adapt our reporting template, educational materials, and overall counseling framework to improve participants’ understanding and overall experiences post-return of GS results.

There are several limitations to our reporting. Repeat expansion disorders were not included in the report. Testing of the SMN1 gene associated with spinal muscular atrophy was also not included.We are evaluating the use of tools like ExpansionHunter (https://github.com/Illumina/ExpansionHunter/) and SMN Copy Number Caller (https://github.com/Illumina/SMNCopyNumberCaller) for detecting and reporting these types of results in the future. Our current process for variant filtration relies in part on an external software program, Franklin by Genoox. Our local custom filtration process will be compared to Franklin to establish concordance of the two pipelines. Variant interpretation processes were developed by our laboratory previously and these were followed to classify variants according to the ACMG guidelines (Lerner-Ellis et al. 2015; Richards et al. 2015). Auto-classification by Franklin is compared to the standard variant interpretation as determined by our variant assessment procedures; thus, both algorithmic based classification as well as manually reviewed and interpreted variants are evaluated in parallel. Assessing and establishing the performance parameters of GS data management and analysis tools like Franklin by Genoox are required to integrate such genomic technologies in the clinical healthcare setting. Currently, we do not have health outcome data on our cohort and are therefore unable to calculate specific risk percentiles or ORs for PRS for common health conditions. Therefore, it did not seem appropriate to model PRS using visuals reflective of continuous risk estimates (i.e., a bell curve with a threshold value), despite reports suggesting that these, along with percentiles and verbal explanations, may be better understood by research participants (Brockman et al. 2021). We may consider amending PRS results in the future after assessing health outcome data and calculating ORs for common conditions in our cohort. We chose to use HapMap3 over other commonly used datasets like 1000Genomes (https://www.internationalgenome.org) to generate an ancestry estimate due to reduced sample processing time. We recognize that 1000Genomes contains a larger number of reference populations and may increase the specificity of ancestry estimations. We plan to validate our genetic ancestry results against the 1000Genomes dataset in the future. Ancestry estimates and PRS are overall less accurate for individuals of non-White/European ethnicity given the lack of representation within publicly available datasets and genomic research studies. The GENCOV study population is recruited from within the Greater Toronto Area of Ontario, Canada, which represents a substantially diverse population. Correlations between self-reported and estimated genetic ancestry as well as contribution of GS data from non-White/European individuals to genomic datasets may help to identify discrepancies and potentially improve the accuracy of ancestry estimations and PRS calculations in the future. More importantly, improving ethnic diversity in genomic datasets is crucial to ensuring that GS is equitable, especially as we move toward using GS more broadly in the general population. The traditional standard for HLA genotyping is PCR-based sequencing methodologies (Mimori et al. 2019). Compared to PCR-based methods, HLA-VBSeq is 66% and 52% accurate at determining approximate and exact HLA class I/II genotype to 4-digit resolution from GS data, respectively (Bauer et al. 2018). Therefore, clinical testing is recommended to confirm HLA genotype and phase at this time. Although, for applications like solid organ transplantation, complete matching of HLA loci is not necessarily required and lower resolution HLA genotyping may be sufficient alongside confirmation of ABO compatibility (Fung and Benson 2015).

In summary, we developed an analysis and reporting structure for GS performed in healthy individuals. Our framework is in alignment with ideas about the future of genomic medicine, which involves the availability of information beyond what is currently considered clinically actionable to patients. Although more research is required to evaluate both the personal and medical implications of returning comprehensive GS data to healthy individuals, the integration of genomic data across multiple aspects of health, including knowledge of risks for hereditary diseases, pharmacogenomics interactions, PRS, as well as genetic ancestry and HLA/blood group genotyping is a step toward more personalized healthcare.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

Conceptualization: SCa, EF, MCh, CM, HF, YB, MCl, JT, and JL-E; data curation: EF, MC; methodology: SCa, EF, MCh, LS, TJP, JS, EG, ML, LH, WL, AN, JT, and JLE; project administration: GM, SCh, CYJF, and JTJL-E; software: EF, MCh, TJP, JS, ML, LH, WL, and JL-E; Visualization: SCa, EF, MCh, GMa, HF, ML, LH, WL, AN, and JLE; writing-original draft: SCa; writing-review & editing: SCa, EF, MCh, GMa, SCh, CM, HF, YB, LS, TJP, JS, SA, NA, EB, AB, BB, HC, MD, LD, DDI, SMF, CYJF, A-C G, LWG, DK, ZK, EL, TL, TM, AMc, SLM, GMo, DR, HS, SS, AT, IW, NZ, EG, ML, LH, WL, AN, JT, and JLE; supervision: JT, JL-E; funding acquisition: HF, TJP, JS, LS, JT, and JL-E; recruitment & data collection: SA, EB, AB, BB, HC, MD, LD, SMF, CYJF, A-CG, LWG, DK, ZK, EL, TL, TM, AMc, SLM, GMo, DR, SS, AT, IW, NZ, GMa, SCa, and SCh; data analysis: NA, DDI, HS, GMo, GMa, SCa, and SCh.

Funding

This work was supported by the Canadian Institutes of Health Research (Funding Reference Number VR4-172753).

Data availability

Raw, de-identified GS, viral, and clinical data will be submitted to and available for access through the databases listed in Table 3.

Declarations

Ethics declaration and consent to participate

This study is approved by the Mount Sinai Hospital research ethics board (Study ID: 424901). All institutions involved in recruitment of study participants received local ethics board approval. Informed consent was obtained from all participants in the study. Participant data were de-identified.

Competing interests

The authors have no relevant financial or non-financial interests to disclose.