About 1 in 20 children with leukemia carries an inherited genetic risk, and for most of them nobody knew until after they got sick. A recent study sequenced the genomes of 181 children with newly diagnosed acute leukemia. Nine had inherited risk, and six of those nine had never been recognized. In eight children, the findings led to tailored surveillance. In four, they changed treatment.

In Destiny’s Child No Longer: Rewriting Genetic Fate , I made the case for reading every child’s genome at birth or before. The Swedish results support that case. They also show its main weakness so far: we sequence the whole genome but read only a small part.

Leukemia is the most common cancer in children and teens . It accounts for about one in four new childhood cancer cases in the United States. Sweden diagnoses about 100 children with leukemia each year .

Most childhood leukemia comes from genetic mutations inside blood cells as a child grows. Some children, though, are born with a “germline” change, meaning a variant present in every cell of the body. These are often passed down in families. Germline risk is significant in three ways. Some chemotherapy regimens are more dangerous for these children. They may face a higher risk of a second cancer. Their parents and siblings may also carry the same variant without knowing it. So far, 45 genes are linked to inherited childhood leukemia risk.

Earlier studies estimated that 4% to 10% of children with blood cancers have such a predisposition. Most of those studies tested children already suspected of inherited risk, which can inflate the numbers.

The team at Karolinska Institutet offered testing to every child diagnosed with acute leukemia at all six pediatric cancer centers in Sweden between December 2021 and April 2024. Of the 213 children diagnosed during that period, 181 participated . The group closely matched the national population by age and disease type.

The strategy had three parts. Doctors conducted a structured clinical review of each child’s family history and physical features. Each child’s genome was sequenced from a skin sample, then analyzed using a list of 189 established cancer predisposition genes. When a child carried a harmful variant, the team also compared it with the DNA of the child’s leukemia cells.

Of the 181 children, nine carried a germline predisposition. Six had conditions linked to leukemia. Three more children carried variants in genes known mainly for adult cancers of the colon, breast and other organs. Those findings led to cascade testing of relatives and to adult surveillance plans.

What The Gene List Missed

The study’s most important finding was something the standard gene list missed. There was a child who showed clear signs of a genetic disorder—such as having too much fluid in the brain, slow development and weak muscles—but when their sequence was checked against the usual list of 189 genes that are known to increase cancer risk, they didn’t find anything unusual. In other words, even though the child had symptoms that suggested a genetic problem, the genetic test came back negative.

The team then sequenced the child together with both parents and searched the full catalog of known disease genes. That wider search found a new variant that causes Tatton-Brown-Rahman syndrome. The diagnosis carries a roughly 250-fold higher risk of blood cancer. The panel left out the variant because it was created in 2020, when evidence linking the gene to leukemia was only emerging.

About 10% of the children had a close relative with childhood cancer, and 9% had relatives with blood cancers. Yet no child with a positive family history carried a harmful variant in any gene on the list. The study concluded that “undiscovered leukemia predisposition syndromes likely exist”. This means focused panels likely won’t capture the full catalog of inherited risk. As a result, reliance on a predefined gene list is a key limitation. These methods could also miss mosaic variants present in fewer than about 20% of cells.

These findings offer a core lesson for genomic medicine. The study generated whole genomes, then deliberately looked at fewer than 200 genes. Large newborn programs do something similar. England’s Generation Study sequences whole genomes and analyzes only about 400 genes tied to treatable childhood conditions. Narrow analysis keeps interpretation manageable, but it means the answers can be no better than the list they come from. Unfortunately, these lists will always lag behind the science.

The Case For Sequencing At Birth

Timing matters as much as scope. In the study, germline DNA came mostly from cultured skin cells. That’s a good choice in leukemia, where blood is contaminated by cancer cells, but culturing takes three to four weeks . That lag time can be significant when a family is making treatment decisions for their child. A genome read at birth, before any cancer, would already be on file when a diagnosis came. It would also be ready to reanalyze as new genes are discovered.

Newborn genome screening is already possible. In New York’s GUARDIAN study , sequencing succeeded for 99.6% of the first 4,000 newborns, and 72% of families agreed to take part. Genome sequencing identified 120 babies with serious conditions . Standard newborn screening caught only 10 of them.

There are still real costs to weigh. Broader analysis finds more variants of uncertain meaning. Sequencing both parents, as this study did, can help sort harmful changes from harmless ones. Still, reading genomes without clinical context can reduce the quality of interpretation.

Systematic genome testing for children with leukemia is feasible nationwide and could change care. Larger studies that combine long-read trio sequencing , transcriptomics, epigenomics and other tools to find new predisposition genes are still needed. Reading a genome early and fully interpreting it, then reanalyzing it as knowledge grows, would let families learn of hidden risk before a crisis rather than after.