Table 2.
Disease | Genes | Variants | Observations, n | +Dx, n (% of observations) | −Dx -Sx, n (% of −Dx observations) | −Dx +Sx, n (% of −Dx observations) |
---|---|---|---|---|---|---|
Amyotrophic lateral sclerosis | 4 | 19 | 21 | 1 (4.8) | 15 (80) | 5 (20) |
Cardiomyopathy | 21 | 151 | 234 | 80 (34) | 124 (81) | 30 (19) |
Colorectal cancer | 7 | 16 | 19 | 2 (11) | 15 (88) | 2 (12) |
Endometrial cancer | 1 | 1 | 13 | 1 (7.7) | 8 (67) | 4 (33) |
Familial breast cancer | 4 | 64 | 168 | 62 (37) | 83 (78) | 23 (22) |
Familial hypercholesterolemia | 1 | 7 | 11 | 5 (45) | 3 (50) | 3 (50) |
Monogenic diabetes | 1 | 1 | 1 | 1 (100) | – | – |
Prostate cancer | 1 | 2 | 2 | 2 (100) | – | – |
Retinitis pigmentosa | 15 | 42 | 175 | 3 (1.7) | 164 (95) | 8 (4.7) |
A total of 644 observations of 303 pathogenic/likely pathogenic or predicted LoF (P/LP/LoF) variants carried by 614 individuals were identified for nine diseases. Observations were first categorized by whether the disease corresponding to the variant was clinically diagnosed (+Dx) or not clinically diagnosed (−Dx) in the individual with the variant. Electronic health records (EHRs) of clinically undiagnosed individuals were then evaluated and categorized by whether evidence of symptomatic disease was present (+Sx) or absent (−Sx). This resulted in three distinct groups: (1) clinically diagnosed (+Dx), (2) no clinical diagnosis and no EHR evidence of disease (−Dx −Sx), and (3) no clinical diagnosis but EHR evidence of disease (−Dx +Sx). n (% of observations), number and percentage of the observations of individuals with P/LP/LoF variants; n (% of −Dx observations), number and percentage of the observations of individuals with P/LP/LoF variants lacking a clinical diagnosis. Genes, number of disease predisposition genes containing the P/LP/LoF variants; Variants, number of P/LP/LoF variants; –, not applicable, as all observations for monogenic diabetes and prostate cancer were diagnosed with the corresponding disease.