Table 3.
Incremental contribution of genetic risk factors to risk prediction of melanoma when added to a traditional risk factor model, based on published risk estimates
| Risk Factor Model | AUC (95% CI) | Change in AUC From Base Model | P-Value1 | Hosmer- Lemeshow P-Value |
Improvement in Sensitivity NRI (95% CI)2 |
Improvement in Specificity NRI (95% CI)2 |
Overall Improvement in Classification NRI (95% CI)2 |
|---|---|---|---|---|---|---|---|
| Australia (N = 1,035) | |||||||
| Base model with traditional risk factors3 | 0.72 (0.69 to 0.75) | 0.13 | |||||
| + MC1R | 0.73 (0.70 to 0.76) | 0.011 | 0.05 | 0.47 | –0.02 (–0.10 to 0.06) | 0.35 (0.26 to 0.43) | 0.33 (0.21 to 0.45) |
| + Pigmentation pathway | 0.74 (0.71 to 0.77) | 0.019 | 0.008 | 0.58 | 0.11 (0.03 to 0.19) | 0.26 (0.18 to 0.35) | 0.38 (0.26 to 0.50) |
| + Nevus pathway | 0.72 (0.69 to 0.75) | 0.001 | 0.54 | <0.01 | –0.02 (–0.11 to 0.06) | 0.07 (–0.02 to 0.16) | 0.04 (–0.08 to 0.17) |
| + Telomere, senescence, and other pathway | 0.72 (0.69 to 0.75) | 0.002 | 0.36 | 0.14 | –0.06 (–0.14 to 0.02) | 0.16 (0.07 to 0.25) | 0.10 (–0.02 to 0.22) |
| + All SNPs4 | 0.74 (0.71 to 0.77) | 0.023 | 0.003 | 0.23 | 0.13 (0.05 to 0.21) | 0.29 (0.20 to 0.38) | 0.42 (0.30 to 0.54) |
| Leeds (N = 1,460) | |||||||
| Base model with traditional risk factors3 | 0.65 (0.62 to 0.68) | 0.70 | |||||
| + MC1R | 0.67 (0.64 to 0.70) | 0.014 | 0.02 | 0.34 | –0.10 (–0.16 to –0.03) | 0.27 (0.18 to 0.35) | 0.17 (0.07 to 0.28) |
| + Pigmentation pathway | 0.68 (0.66 to 0.71) | 0.031 | 0.0005 | 0.76 | 0.09 (0.02 to 0.15) | 0.22 (0.13 to 0.30) | 0.30 (0.20 to 0.41) |
| + Nevus pathway | 0.65 (0.62 to 0.68) | 0.000 | 0.98 | 0.81 | –0.03 (–0.10 to 0.03) | 0.06 (–0.03 to 0.14) | 0.02 (–0.08 to 0.13) |
| + Telomere, senescence, and other pathway | 0.66 (0.63 to 0.69) | 0.004 | 0.33 | 0.19 | –0.01 (–0.07 to 0.06) | 0.10 (0.02 to 0.19) | 0.10 (–0.01 to 0.21) |
| + All SNPs4 | 0.68 (0.65 to 0.71) | 0.028 | 0.002 | 0.10 | 0.09 (0.03 to 0.16) | 0.19 (0.11 to 0.28) | 0.29 (0.18 to 0.39) |
Abbreviations: AUC, area under the receiver operating characteristic curve; CI, confidence interval; NRI, net reclassification improvement; SNP, single nucleotide polymorphism.
Chi-square P-value for the difference in the AUC when compared with the base model.
Based on continuous NRI. Improvement in sensitivity is calculated from reclassification of case individuals improvement in specificity is calculated from reclassification of control individuals, and overall improvement combines the improvements in sensitivity and specificity.
Traditional factors include hair color, skin color, eye color, freckling as an adult, skin photosensitivity, self-reported nevi, sunbed use, keratinocyte cancer personal history, first degree family history of melanoma, vacation sun exposure, and blistering sunburns as a child, as well as the demographic and study design factors of age, sex, city of recruitment, and European ancestry.
Added as a polygenic risk score, comprising 45 SNPs in 21 genes. The SNPs in each pathway can overlap; the pigmentation pathway includes 14 genes (31 SNPs); nevus pathway includes 7 genes (13 SNPs); and telomere, senescence, and other pathways includes 5 genes (9 SNPs).