NEW YORK – Researchers estimate that making genetic screening for pediatric cancer-associated gene variants standard for newborns could reduce cancer deaths before age 20 by half, according to a study published in Genetics in Medicine in March.
Despite such findings, though, there will likely need to be further reductions in sequencing costs before state screening programs will embrace such an effort, experts said.
Each year in the US around 4 million infants are screened for at least 30 rare genetic and other disorders, such as hearing loss and congenital heart defects, using dried blood spots. Each year, about 3,000 newborns with severe disorders are identified, according to the US Centers for Disease Control and Prevention. However, state-based newborn screening programs currently don't test for pediatric cancer predisposition genes. Researchers led by Jennifer Yeh, an associate professor of pediatrics at Harvard Medical School, set out to explore the benefits and costs of implementing such a program.
In a paper published last month in Genetics in Medicine, Yeh and colleagues modeled data from clinical studies, the National Center for Biotechnology Information's ClinVar database, the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) database, and the Broad Institute's Genome Aggregation Database (gnomAD) to determine the incidence of cancer-linked pathogenic or likely pathogenic genetic variants in newborns and estimate the proportion of young patients who would likely develop a cancer predisposition syndrome (CPS) before age 20.
"Although randomized controlled trials are the gold standard for studying screening outcomes, they are often infeasible because of the long follow-up time and very large sample size that would be required," said Yeh. "This is especially true for rare diseases, such as childhood cancer. We therefore used simulation modeling to estimate the potential benefits and harms of newborn genetic screening for pediatric cancer predisposition syndromes."
For their analysis, researchers categorized pathogenic or likely pathogenic variants in RET, RB1, TP53, DICER1, SUFU, PTCH1, SMARCB1, WT1, APC, ALK, and PHOX2B, which are associated with increased risk of childhood medullary thyroid cancer, retinoblastoma, adrenocortical carcinoma, choroid plexus, rhabdomyosarcoma, osteosarcoma, rhabdoid tumors, pleuropulmonary blastoma, medulloblastoma, neuroblastoma, Wilms tumor, and hepatoblastoma.
They found that among a typical US newborn cohort of 3.7 million, about 1,800 would develop a CPS malignancy before age 20 under usual care. However, if these newborns had been universally screened for cancer risk genes, 1,564 patients with pathogenic variants would have been identified, 232 of whom would likely develop cancer before age 20.
These young patients with pathogenic cancer predisposition genetic variants would be candidates for further surveillance with more frequent ultrasounds or MRI screenings, which could identify cancer earlier. That surveillance and early detection could result in a 53.5 percent decrease in cancer deaths in patients with pathogenic or likely pathogenic heterozygous variants, translating to a 7.8 percent decrease in cancer deaths in the entire newborn cohort before age 20, Yeh and colleagues estimated.
The researchers also explored whether universal newborn genetic screening and subsequent surveillance was cost effective. They found that the universal screening model had an incremental cost-effectiveness ratio (ICER) of $244,860 per life year, if there was full adherence to the screening.
However, without full adherence, the authors found that the universal screening program would be less cost effective. If there was 90 percent compliance with screening, for example, the ICER would increase to $270,260 per life year; at 70 percent compliance, it would jump to $321,000 per life year. Yeh and colleagues noted that adherence to surveillance and screening, as well as parental uptake of screening in these programs, needs to be studied more.
An ICER of $100,000 per life year is usually the benchmark for value, which means these study results alone may not support such a universal screening program, said David Veenstra, associate director of the Comparative Health Outcomes, Policy & Economics (CHOICE) Institute at the University of Washington, who was not an author on this study.
"One challenge facing newborn cancer predisposition screening is that so many decades pass between the time of the screening and the time of the benefit in the future," Veenstra added. "Because in economic evaluations future events are discounted, this can make interventions like this less favorable."
Yeh and colleagues assumed in their analysis that, on average, it would cost $55 to add an 11-gene next-generation sequencing panel to existing genetic tests already being conducted within state-based newborn screening programs. "We assumed that this cost reflected the incremental cost of adding the panel to a [newborn screening] program with existing infrastructure for genetic testing," the researchers wrote in the study. The cost analysis also included costs for surveillance and cancer treatment based on estimates from the published literature and national databases and the cost of parental time lost from work.
The researchers speculated that the cost of NGS would continue to decrease, which would improve the cost-effectiveness of universal screening. Ultimately, they suggested that universal screening could meet the $100,000 per life year threshold as the cost of genetic testing falls. Their analysis suggests that if the cost of testing for the 11 cancer predisposition genes could be around $20, then the ICER would fall below $100,000 per life year.
Yeh said that researchers and health systems have previously suggested using higher cost-effectiveness thresholds when considering care and treatment for rare diseases, such as childhood cancers. Using a higher threshold could make this universal screening program for cancer cost effective sooner without the need to wait for the costs of genetic screening to decrease.
The researchers based their cancer care cost estimates on a 2018 study that estimated the cost of childhood cancer treatment to be around $170,000 for the initial and continuing phases and $365,000 for the final phase of treatment.
The costs of surveillance, like physician's office visits, ultrasounds, or MRIs, were pulled from a database compiled by the US Centers for Medicare & Medicaid Services. The cost for conducting a brain MRI, for example, was $842, and an abdominal ultrasound was $240, which quickly add up when spread over thousands of patients and many years of surveillance.
Within a universal screening program, Yeh noted that even newborns who would never develop cancer despite harboring cancer risk variants would undergo more frequent surveillance using these costly interventions, which adds to health system costs. This process, in turn, may heighten anxiety in parents, knowing their child could develop cancer. The present analysis did not account for that issue, Yeh said, but the authors may explore the effect of parental anxiety in future research.
"With any population-based screening program, there will be some drawbacks," Yeh said. "A strength of our modeling approach was to estimate the number of variant carriers that would need to be followed for the benefits associated with surveillance to be realized."
The researchers acknowledged in the paper that adding this NGS cancer screening to the current newborn screening program would be complex, since states having different processes for expanding existing screening programs. But they argued that universal implementation into states' existing newborn screening programs would be the most effective way to make this type of testing broadly available.
Offering genetic screening separately from existing newborn screening programs "would create added burden and require additional resources for implementation," which the study authors did not analyze, they wrote.
The researchers acknowledged that more resources and money would be needed to implement the infrastructure for a nationwide universal screening program, along with support for families after receiving test results. A pathogenic cancer risk variant in a newborn could have implications for the parents' future reproductive planning and could necessitate cascade testing to assess genetic cancer risk in other relatives.
"While our findings are suggestive, using newborn genetic screening for pediatric CPS as an example, our study demonstrates how advances in genetics can be applied to populations and what the implications might be for public health," the authors wrote.