Risk assessment is reshaping pediatric cancer care beyond diagnosis

All stories start in a similar way. They introduce characters, establish the setting and provide the first hints about what lies ahead. However, crucial clues are often hidden in this first chapter, details that only become meaningful as the story unfolds. For physicians treating childhood cancer, diagnosis can similarly contain clues that shape a patient’s care.

What was once simply identifying a disease has increasingly become an opportunity to understand what may come next. Advances in genomics, molecular profiling and computational biology help researchers uncover clues about how cancer is likely to behave, how a patient may respond to treatment and even what health challenges may emerge years or decades later.

“Diagnosis is increasingly becoming a data-driven navigation system for pediatric cancer care,” said Xin Zhou, PhD, Department of Computational Biology. “By integrating genomic, molecular and clinical data from large patient populations, computational methods and data portal infrastructures help clinicians identify risk with greater precision and tailor therapy to the individual patient.”

That transformation has changed the role of risk assessment in cancer care. Rather than focusing solely on identifying a disease, physicians are increasingly using information gathered at diagnosis to forecast future chapters of a patient’s journey, from treatment decisions to survivorship.

Reading between the lines

Historically, pediatric cancer diagnoses relied on what clinicians could directly observe. Tumors were classified based on their appearance under a microscope, while characteristics such as age, symptoms and laboratory values helped guide treatment decisions. 

“The concept of diagnosis in pediatric oncology hasn’t really changed, but modern technology has allowed us to expand on diagnosis,” said Giles Robinson, MD, Division of Neuro-Oncology director within the Department of Oncology.

Researchers have long recognized that pediatric cancers are rare and require careful classification. What has changed is the ability to examine those cancers at an unprecedented level of detail.

“When we look inside these tumors, we see more differences and nuances among pediatric cancers than we originally knew existed,” Robinson said. “This complexity is both confounding but also highly informative because it helps us explain why two patients with seemingly similar looking tumors can have very disparate responses to treatment.” 

Advances in next-generation sequencing and molecular profiling have revealed that cancers once considered single diseases are often composed of many biologically distinct subtypes. Those differences can influence treatment response, prognosis and therapeutic opportunities. 

“A more comprehensive molecular diagnosis provides a more complete view of the tumor genome,” said Lu Wang, MD, PhD, Department of Pathology. “Identifying not only the primary driver alterations but also additional genetic changes that may be targetable or that may cooperate with the driver alteration to influence a patient’s response to standard therapies.” 

Integrated approaches that combine whole-genome and whole-transcriptome sequencing help physicians identify abnormalities and patterns that may influence treatment planning. “Together, these approaches provide deeper insight into tumor biology and the clinical significance of genetic findings,” Wang said.

Today, physicians use diagnosis not only to classify disease, but also to uncover the biology that shapes its course.

When the story changes

Even the most thorough opening chapter cannot reveal everything about a story — or a course of treatment. Some of the most important information emerges over time. One of the most significant advances in pediatric leukemia care has been the development of minimal residual disease (MRD) testing, which can detect tiny numbers of leukemia cells that remain after treatment — cells that are invisible to traditional diagnostic methods. 

Ching-Hon Pui, MD, Department of Oncology, and colleagues helped demonstrate that MRD levels measured during treatment were among the strongest predictors of patient outcomes, outperforming many traditional clinical risk factors. “Incorporation of MRD into risk stratification enabled treatment intensification for patients with persistent leukemia and de-escalation for those with rapid disease clearance,” Pui explained, “thereby improving survival while reducing treatment-related toxicity.”

MRD fundamentally changed how physicians assess risk. Instead of relying solely on information gathered at diagnosis, clinicians could refine treatment strategies based on how a patient’s leukemia responds to therapy in real time. The success of MRD-guided therapy has driven the development of new approaches across oncology, including liquid biopsies, single-cell DNA sequencing, single-cell multi-omics, circulating tumor DNA and other biomarkers that allow physicians to monitor disease more precisely and adapt treatment as disease status evolves.

Risk assessment is no longer a one-time decision made at diagnosis. It is a dynamic process that evolves alongside the patient and the disease itself.

Changing the course of the story 

For decades, much of cancer research focused on identifying patients at the highest risk of relapse. Today, physicians recognize that identifying low-risk patients can be just as important.

“Risk assessment is fundamental to improving therapy,” Robinson said. “While cancer therapy has advanced, many curative treatments still rely on highly toxic and nonspecific therapies that take a huge toll on children.”

Patients identified as low-risk often have cancers that are particularly sensitive to treatment. That information creates opportunities to reduce therapy intensity safely while preserving excellent outcomes.

“When we identify patients as ‘low-risk,’ this generally means that they have good outcomes or prognosis from therapy,” Robinson said. “When I see this, I see it as an opportunity to lower doses of therapy and reduce the long-standing toxicity to these kids.”

The motivation behind those decisions extends beyond the immediate goal of a cure.

“The knowledge we have gained from survivorship cohorts, such as the Childhood Cancer Survivor Study and the St. Jude Lifetime Cohort Study, has been essential in informing our understanding of cancer treatment-related late effects,” said Matt Ehrhardt, MD, MS, Departments of Oncology and Epidemiology & Cancer Control.

Findings from these studies have informed treatment protocols that intentionally reduce cumulative doses, lower treatment intensity or eliminate specific therapies when possible to lessen long-term toxicity.

The chapters that follow

For survivors, risk takes on a different meaning. “At diagnosis, risk often refers to risk of recurrence if all patients were treated similarly,” Ehrhardt said.

In survivorship, physicians focus on the risk of chronic health conditions, treatment-related complications and premature mortality. “Some of the treatments required to cure cancer also place patients at increased risk for side effects years, or even decades after treatment,” Ehrhardt explained.

For example, anthracycline chemotherapy agents can increase the risk of cardiomyopathy long after treatment is complete. Understanding those risks allows clinicians to tailor screening programs, surveillance strategies and interventions to the needs of individual survivors.

Risk assessment, therefore, does not end when treatment ends.

In some diseases, physicians are already beginning to incorporate long-term considerations into treatment selection itself. 

“We are at a point with some diseases, such as Hodgkin lymphoma, where there are multiple, equally efficacious treatment approaches,” Ehrhardt said. “By understanding these risks, we can now begin to consider the selection of treatments that are associated with risk profiles that are most favorable to an individual patient.”

The goal is no longer simply to help patients survive cancer. It is to help them live healthier lives after cancer.

A story still being written

Researchers are now integrating genomics, epigenomics, single-cell analyses, proteomics, immune profiling, liquid biopsies and artificial intelligence into clinical decision-making. 

“The most powerful advances are not coming from any single technology,” Zhou said, “but from our ability to connect diverse data types and transform them into actionable knowledge that improves both survival and survivorship.”

That broader view of risk represents one of the most important shifts in pediatric oncology.

“Personalized medicine does not only impact treatment choices,” Ehrhardt said. “It starts at diagnosis but impacts the entire survivorship journey.”

Every patient story begins with a diagnosis. Yet, unlike the opening chapter of a book, physicians are no longer reading it simply to understand what is happening in the moment. They are reading it for clues about every chapter that follows. 

The story can still take unexpected turns, and not every outcome can be predicted. However, today, physicians can read far more between the lines than ever before. With each new insight, they are helping write stories defined not only by survival, but by the opportunity for children to thrive long after cancer becomes part of their past.