Assessment of the Diagnostic Yield of Combined Cardiomyopathy and Arrhythmia Genetic Testing

Authors, Journal, Affiliations, Type, DOI

• Authors: Lisa M. Dellefave-Castillo, MS; Allison L. Cirino, MS; Thomas E. Callis, PhD; Edward D. Esplin, MD, PhD; John Garcia, PhD; Kathryn E. Hatchell, PhD; Britt Johnson, PhD; Ana Morales, MS; Ellen Regalado, PhD; Susan Rojahn, PhD; Matteo Vatta, PhD; Robert L. Nussbaum, MD; Elizabeth M. McNally, MD, PhD
• Journal: JAMA Cardiology, September 2022, Volume 7, Number 9
• Affiliations: Center for Genetic Medicine, Feinberg School of Medicine, Northwestern University, Chicago (Dellefave-Castillo, McNally); Cardiovascular Division, Brigham and Women's Hospital, Boston (Cirino); Invitae Corporation, San Francisco (Callis, Esplin, Garcia, Hatchell, Johnson, Morales, Regalado, Rojahn, Vatta, Nussbaum)
• Type: Original Investigation — retrospective cohort study
• DOI: https://doi.org/10.1001/jamacardio.2022.2455
• Conflict of interest: Multiple authors hold Invitae stock; study testing funded in part by industry sponsors Amicus Therapeutics and Alnylam Pharmaceuticals. Sponsors had no role in study design, data analysis, or manuscript preparation.

Overview

This retrospective cohort study (n = 4,782) evaluated a no-charge sponsored genetic testing program offering combined cardiomyopathy and arrhythmia panel testing (up to 150 genes) in patients with suspected genetic cardiomyopathy or arrhythmia across the US and Canada. A positive (P/LP) result was identified in 19.9% of patients, with 66% of positives having clinical management implications. Critically, 10.9% of positive findings would have been missed by disease-specific panels — particularly cross-disease diagnoses where a cardiomyopathy indication yielded an arrhythmia gene finding or vice versa. Cascade family testing of 958 relatives identified additional P/LP variants in 42%, demonstrating a substantial family multiplier effect.

Keywords

Cardiomyopathy, arrhythmia, genetic testing, diagnostic yield, hypertrophic cardiomyopathy, dilated cardiomyopathy, arrhythmogenic cardiomyopathy, long QT syndrome, cascade testing, variants of uncertain significance, pathogenic variants, clinical management

Key Takeaways

Study Design and Population

• Retrospective review of a no-charge, commercially sponsored genetic testing program (Invitae Corporation), July 2019–July 2020
• 4,782 unrelated patients referred by 1,203 clinicians; up to 150 cardiomyopathy + arrhythmia genes tested simultaneously
• Mean age 40.5 years; 53.3% male; 59.9% White, 11.9% Black/African American, 7.8% Hispanic (higher non-White representation than similar prior studies)
• 190 patients (4.0%) were post-mortem (decedent) cases; 41.9% of patients had multiple clinical reasons for referral
• Most common referral indications: HCM (41.5%), DCM (28.7%), LQTS (16.5%)
• Inclusion criteria deliberately broad — any level of clinician suspicion of genetic cardiomyopathy or arrhythmia

Overall Diagnostic Yield

• Positive result (P/LP): 19.9% (954 of 4,782 patients)
• 1,227 total P/LP variants identified; 42 (3.4%) were copy number variants; 7 CNVs were single-exon only (missed by standard Sanger sequencing)
• 89 patients (1.9%) had more than 1 P/LP variant; 8 patients (0.8%) had positive results in both cardiomyopathy and arrhythmia genes simultaneously
• 51.2% of patients had only VUS (no P/LP variant) — highlighting the substantial burden of uncertain results

Yield by Referral Indication

• Patients with high clinician index of suspicion had higher yield (25.7%) vs. low suspicion (9.0%), but those 11 low-suspicion positives confirm genetic findings can occur even in unlikely cases
• If AHA guidance were followed to exclude low/moderate suspicion: 14.4% (137/954) of positives would have been missed

Top Genes by P/LP Variant Frequency

• MYBPC3: 205 patients (16.7% of all P/LP variants)
• TTN: 143 patients (11.7%)
• MYH7: 111 patients (9.0%)

Clinical Management Implications

• 66.0% (630/954) of patients with positive results had potential clinical management implications
• Three management-relevant gene categories:
  1. Sarcomeric HCM adverse outcomes (ACTC1, MYL2, MYBPC3, MYH7, MYL3, TNNI3, TNNT2, TPM1): 279 patients (29.2% of positives) — earlier monitoring for AF, VT, HF
  2. LAMP2 (rapid deterioration, early transplant consideration): 4 patients (0.4%)
  3. Heightened arrhythmia risk (ABCC9, DES, DSP, FLNC, LMNA, PLN, RBM20, RYR2, SCN5A, TTN): 300 patients (31.4%) — more intensive cardiac monitoring and/or device intervention
  4. Targeted therapies (GAA, GLA, TTR): 57 patients (6.0%) — enzyme replacement therapy or TTR stabilisers/inhibitors
• Among all patients tested: 1 in 8 (13.2%) had a positive result that could inform prognosis or clinical management

Diagnoses Gained by Combined vs. Disease-Specific Panel Testing

• Among 689 patients with a single specific disease indication: 75 positive results (10.9%) would have been missed by disease-specific panels
• Breakdown of 75 gained diagnoses:
  • 44.0% (33/75): within-category gain (e.g., DCM tested but a different cardiomyopathy gene found)
  • 16.0% (12/75): cardiomyopathy indication + arrhythmia genetic finding
  • 20.0% (15/75): arrhythmia indication + cardiomyopathy genetic finding
  • 36.0% (27/75) total cross-disease gains (arrhythmia↔cardiomyopathy indication mismatch)
• LVNC had the highest proportion of cross-disease missed diagnoses: 14 of 16 positives (87.5%) were in genes not on the LVNC-specific panel — reflecting the poorly defined genetic landscape of LVNC
• Combined testing was statistically superior to disease-specific testing overall (P = .02), though individual subtype comparisons did not reach significance

Cascade Family Testing

• 306 of 954 patients (32.1%) with positive results had family members tested through the same program
• 958 family members tested; 402 (42.0%) received a positive result; 445 (46.5%) received a negative result (freed from continued intensive surveillance)
• Approximately 1 in 23 family members with positive results would have been missed if disease-specific panels had been used instead of combined panel
• Postmortem cascade testing: 190 decedents tested; 18 (9.5%) had positive results; cascade testing was performed in 61.1% of positive postmortem cases (vs. 32.1% overall) — emphasising value of post-mortem testing
• Family testing may be underestimated as data only captured testing at the same laboratory

Copy Number Variants

• 42 CNVs identified (3.4% of all P/LP variants); 7 (16.7%) were single-exon CNVs requiring targeted analysis
• 81.0% were in autosomal dominant genes; 14.3% in X-linked DMD gene; 4.8% in autosomal recessive genes (monoallelic)
• Clinical implication: standard Sanger sequencing and some NGS panels may miss these — comprehensive CNV analysis is part of combined panel testing

Limitations of the Document

• Clinical data not confirmed with medical records — diagnoses based on clinician-reported requisition form; some clinical information may be incomplete or inaccurate
• Invitae sponsorship and conflict of interest: Most co-authors hold Invitae stock; testing was conducted through a no-charge Invitae-sponsored program — selection bias toward patients who would not otherwise agree to for-cost testing; also potential bias in variant interpretation practices
• No comparator arm: There was no randomised or contemporaneous control group receiving disease-specific testing; yield gains from combined testing are estimated by re-analysing existing results under hypothetical panel restrictions
• Non-White patient underrepresentation: Black/African American and Hispanic patients are overrepresented compared to some prior studies but remain underrepresented relative to the US population; VUS rates are known to be higher in non-White patients due to reference database gaps — this was not analysed in detail
• Family testing underestimation: Only family testing at the same laboratory was captured; total cascade testing in the cohort may be substantially higher
• Single laboratory: Variant classification may differ at other laboratories using different frameworks or reference datasets
• Overall yield (19.9%) lower than a prior 105-gene combined panel study (~30%): Permissive inclusion criteria (any level of clinician suspicion) likely accounts for this difference; reflects real-world conditions rather than enriched referral populations

Key Concepts Mentioned

• concepts/Genetic-Testing-in-Cardiomyopathy — central concept; yield data by phenotype and combined vs. disease-specific panel comparison
• concepts/Cascade-Family-Screening — 42% positive rate in family members; 46.5% negative clearance rate
• concepts/Variant-Reclassification — VUS burden (51.2%); CPVT highest VUS rate (63.2%)

Key Entities Mentioned

• entities/MYBPC3 — most common gene in P/LP variants (16.7%)
• entities/TTN — second most common gene (11.7%); listed as heightened arrhythmia risk gene
• entities/LMNA — listed as heightened arrhythmia risk gene; gene-specific management implications
• entities/DCM — 19.1% yield; consistent with prior estimates
• entities/ARVC — lowest VUS rate (46.0%); 18.8% yield
• entities/HCM — highest yield in young patients (40.4% age 19–39)

Wiki Pages Updated

• wiki/sources/genetic-yield-jama-card-2022 (created)
• wiki/wikiindex (updated)
• wiki/concepts/Genetic-Testing-in-Cardiomyopathy (updated)
• wiki/concepts/Cascade-Family-Screening (updated)
• wiki/entities/LMNA (updated)
• wiki/entities/TTN (updated)
• wiki/entities/ARVC (updated)
• wiki/entities/DCM (updated)