Universal newborn screening (NBS) has been shown to decrease the mortality of many severe pediatric-onset diseases. Genetic testing for pediatric cancer predisposition syndromes (CPS) could augment these newborn screening programs, but what are the costs? Are there benefits?

A new study senior-authored by Institute Associate Professor Ann Wu, MD, MPH, explores these questions in a new study in Genetics in Medicine: Universal newborn genetic screening for pediatric cancer disposition syndromes: model-based insights. Dr. Wu is also Director of the PRecisiOn Medicine Translational Research (PROMoTeR) Center at the Institute, which identifies and evaluates genomics practices to improve individual and population health. The study team included the Institute's Natasha K. Stout, PhD, Kurt Christensen, PhD, MPH, Pamela McMahon, PhD, and Christine Lu, MSc, PhD, and collaborators from Boston Children's Hospital, Brigham and Women's Hospital and Broad Institute, Center for Genomic Medicine, Massachusetts General Hospital, Geisinger Genomic Medicine Institute, and Dana-Farber Cancer Institute.

We spoke with Dr. Wu to learn more about the study. 

Q: Tell us about universal newborn screening. How can genetic testing augment this screening?

A: Newborn screening in the US varies by state, but typically tests for diseases such as cystic fibrosis, sickle cell disease, and metabolic disorders. Testing for pathogenic variants in genes that increase risks of pediatric cancers is technologically feasible and could identify children with these cancers at an earlier stage and improve survival and outcomes. The benefits and potential harms of testing for pathogenic variants is unknown, and randomized clinical trials are not likely due to the rarity of pediatric cancers.


Q: Your team developed the Precision Medicine Policy and Treatment (PreEMPT) Model to estimate the potential risks and benefits of population-based genetic screening for specific pathogenic germline variants. Can you describe to us what decision models are, and how, applied to public health questions, they can help inform clinical practice?

A: Randomized trials are unlikely to be performed for interventions targeting diseases that are diagnosed in only a handful of children per birth cohort. Decision models are a way to combine available data and understand the likelihood of observing a net benefit from an intervention. Decision models allow teams of interdisciplinary researchers to integrate data from multiple sources (genomic databases, disease registries, clinical guidelines, and other outcomes research literature) and then to simulate the world both as it is (status quo) and with a new hypothetical test. In our case, we simulated the life histories of a year of US births in which universal sequencing of newborns was used to identify pathogenic variants in genes that increase risks of pediatric cancers. We then simulated clinical events (surveillance of individuals with known variants, treatment of disease, survival by stage, etc.) and tallied up benefits and harms accruing to children under the hypothetical state with universal sequencing versus the status quo. We include uncertainty in input estimates and simulate multiple cohorts so that our predictions include ranges. We can also vary assumptions and find important sources of uncertainty that would change the conclusions, and help direct future research efforts. ​

Testing for pathogenic variants in genes that increase risks of pediatric cancers is technologically feasible and could identify children with these cancers at an earlier stage and improve survival and outcomes.

Q: What did study findings show? Was your team surprised by these findings?

A: Although pediatric cancer is rare, we found that universal sequencing resulted in fewer cancer deaths and fewer adult survivors at risk of radiation-related excess mortality. We found that if the cost of universal sequencing falls to $20 per child, universal sequencing to identify pathogenic germline variants in one of 11 genes would cost less than $100,000 per year of life saved, a commonly accepted benchmark for interventions that are cost-effective.


Q: What are some of the ethical concerns?

A: Most screening tests will identify individuals who require follow up tests but ultimately do not develop the disease in question. In our study, we estimated the number of children who would be identified as carrying a pathogenic variant associated with cancer and face surveillance tests, but who do not develop cancer. Similarly, other children with none of the variants included in the screening test will develop cancer, and are not helped by the test; this could change over time as additional variants are identified and added to the test. We postulated a ‘best case’ scenario of universal testing and follow up of positives, and did not consider any issues related to access to testing and how that would vary in disadvantaged populations.

Our study may help policymakers as they evaluate the tradeoffs of adding genetic sequencing to newborn screening programs.

Q: What are the potential implications of this study for public health? What future studies are you and your team considering to expand on this work?

A: Our study may help policymakers as they evaluate the tradeoffs of adding genetic sequencing to newborn screening programs. We are currently evaluating genetic sequencing of other pediatric diseases, as well as combinations of diseases.
 

Follow Dr. Wu on Twitter @Asthma3Ways | About the PRecisiOn Medicine Translational Research Center (PRoMoTeR) |
Follow PRoMoTeR on Twitter @PROMoTeR_DPM