Targeted genomic sequencing of newborn dried blood spots may have identified pathogenic or likely pathogenic cancer-predisposition variants in about 7% of pediatric patients who subsequently developed a solid or central nervous system malignancy by 8 years. The findings could provide an estimate of the proportion of early-onset cancers in this population associated with germline variants detectable at birth.
Investigators used the Michigan Cancer Surveillance Program and newborn screening resources to identify 1,948 patients born in Michigan from 1987 to 2020 who developed a malignant solid or central nervous system tumor by 8 years and had an archived newborn dried blood spot available for research. They performed targeted next-generation sequencing of 11 autosomal dominant cancer-predisposition genes selected for their associations with early childhood cancers, the availability of gene-specific surveillance guidelines, and the potential benefit of early detection. Adequate DNA was extracted from 1,943 of the samples.
Pathogenic or likely pathogenic germline variants were detected in 7% (n = 132) of the patients in the cohort. RB1 accounted for 69 variants, followed by TP53 with 24, SMARCB1 with eight, and WT1 with seven. Variants were also identified in RET, SUFU, PTCH1, DICER1, APC, and PHOX2B; no pathogenic or likely pathogenic variants were detected in ALK. Further, 130 patients developed a tumor type with a well-established association with the affected gene.
Based on about 3.6 million Michigan births, the investigators estimated that about 1 in 27,000 newborns subsequently developed an early-onset solid or brain malignancy in the presence of a detected cancer-predisposition variant.
Pathogenic or likely pathogenic variants were found in all 6 patients who developed medullary thyroid carcinoma, 40% of those who developed retinoblastoma, 30% with choroid plexus carcinoma, 20% with adrenocortical carcinoma, 17% with pineoblastoma, and 11% with lung or pleural malignancy and medulloblastoma. The investigators reported strong gene-tumor specificity, with RB1, RET, and WT1 variants confined to patients who developed associated tumors.
Germline RB1 variants were detected in 68 of 168 patients who developed retinoblastoma. Among 51 patients with known bilateral disease, 80% (n = 41) carried an RB1 variant vs. 23% (n.= 27/117) of those with unilateral or unrecorded laterality. Patients with RB1 variants received a diagnosis at a median age of 9 months compared with 23 months among those without a detected germline RB1 variant. There were no statistically significant differences in stage distribution, primary treatment, or overall survival according to RB1 variant status.
Across the full cohort, patients with detected cancer-predisposition variants received a cancer diagnosis at a median age of 14 months vs. 32 months among those without variants. Among 126 carriers with available second-cancer data, 9% (n = 11) developed a second cancer compared with 5% (n = 85/1,755) of noncarriers, a difference that was not statistically significant.
Among the 1,780 patients with cancers other than retinoblastoma, 4% (n = 64) carried a pathogenic or likely pathogenic germline variant. The investigators found no statistically significant differences in stage distribution or primary treatment according to variant status in this group. Overall survival was lower among carriers.
The study had several limitations. The investigators used a phenotype-first approach and did not prospectively determine whether genomic newborn screening improved cancer detection, treatment, survival, or other clinical outcomes. They lacked an ideal large comparison cohort of newborns tested using identical genomic and analytic methods and known to remain cancer-free through early childhood.
The investigators cautioned that the findings should be extrapolated carefully to populations with different ancestry or demographic characteristics. Registry data did not establish how many patients already had a clinically recognized cancer-predisposition syndrome or family history of cancer, and analyses by race and ethnicity were limited by suppressed demographic data and small subgroup sizes. Additionally, the panel included just 11 genes associated with solid and brain tumors and excluded autosomal recessive cancer-predisposition syndromes and leukemia-predisposition genes, which likely underestimated the overall number of genetically at-risk newborns. The study did not evaluate the economic implications of newborn screening for cancer risk.
“Our data support newborn screening for selected cancer-risk genes,” wrote lead study author Lisa Diller, of the Department of Pediatric Oncology at Dana-Farber Cancer Center, and colleagues. Prospective research would be needed to determine the clinical and public health effects of implementing such screening.
The study authors reported no conflicts of interest.
Source: Nature Communications
