Creating guidelines that turn evidence into better diagnoses and treatment

Biomedical research has transformed our ability to turn symptoms into answers in the form of an accurate diagnosis. Over the past few decades, molecular profiling, pharmacogenomics and new computational tools have given physicians unprecedented ways to understand disease. However, that progress has also created a challenge: How do clinicians keep up with a rapidly expanding body of evidence and turn it into clear, practical decisions for patients?

To meet the challenge, experts, including many St. Jude researchers, play leading roles in writing and updating consensus guidelines across pediatric diseases to keep clinical diagnosis and treatment up to date with the latest discoveries.

“The purpose of diagnostic testing is to reveal the true character of a patient’s tumor and to guide us to the most effective treatment,” said David Ellison, MD, PhD, Department of Pathology chair. “We update our global guidelines for diagnosis as our understanding of cancer biology grows, improving our ability to anticipate how tumors will behave and predict which treatments will work best.”

Turning new biology into better brain tumor diagnosis

Pediatric brain tumors illustrate why that process is so important. For years, tumors were only classified by how they appeared under the microscope. However, molecular pathology, which uses DNA and RNA sequencing and methylation profiling, a map of a set of chemical marks on DNA, has revealed biological differences that histological methods could not fully capture.

Ellison has helped lead that shift through his work on the World Health Organization Classification of Tumours of the Central Nervous System. Since his first involvement leading up to the fourth edition in 2016, the field has advanced rapidly, making it possible to distinguish more specific tumor subtypes, including important differences between adult and pediatric brain cancers. He helped incorporate those insights into the fifth edition.

“I was motivated because, 30 years ago, I could see there were differences between adult and pediatric tumors down the microscope,” Ellison said. “We could finally include strong evidence of their molecular differences and the distinctive character of pediatric tumors.” 

Still, discovering new molecular tumor types is not enough to change a guideline. The key question is whether the distinction is clinically useful. “We ask if it has distinct clinical features or biological behavior, or if it has genetic alterations that allow you to have a targeted therapy, Ellison explained. “In other words, we want to make sure that we are only including things that will be useful on the clinical frontline.”

Guiding diagnosis and treatment through pharmacogenomics

The same principle guides pharmacogenomics, which looks at how medicines will act based on an individual’s genetics. In oncology, that can mean predicting whether a patient is likely to metabolize a drug quickly or slowly, affecting both efficacy and toxicity.

Kelly Caudle, PharmD, PhD, BCPS, FCCP, Clinical Pharmacogenomics Implementation Consortium (CPIC) principal investigator and director, and Department of Pharmacy & Pharmaceutical Sciences associate member, helps maintain and update guidelines for cancer drugs through CPIC.

“Our guidelines go all the way from raw genetic data to interpreting it into clinical action,” Caudle explained. “Our bodies influence our responses to cancer drugs, both efficacy and toxicity. Therefore, we keep up to date with the most recent evidence to have ‘living’ guidelines to support the most appropriate treatment for a patient with the lowest number of likely side effects.”

St. Jude helped establish this field decades ago through work showing that variants in the gene TPMT could dramatically alter how children tolerate mercaptopurine, a key drug in pediatric cancer treatment. Adjusting mercaptopurine dosing based on TPMT status reduced toxicity and helped set a precedent for using genetics to personalize therapy, including more recent updates in 2025.

However, research evidence alone does not guarantee a useful guideline. Recommendations must also work in real clinical settings.

Making guidance practical

That challenge has become clear in efforts to guide fluoropyrimidine dosing based on the DPYD gene. The evidence supported reducing the dose and then increasing it as tolerated, but implementing that recommendation comes with its own challenges.

“While the evidence was sound for decreasing the dosage, then increasing it to tolerability, that turned out to be difficult and confusing for physicians to do,” Caudle explained.

That kind of gap between evidence and practice is common. An approach may seem straightforward in a research setting but prove harder to apply in a busy clinic or hospital. In those cases, guideline groups must revisit the recommendation, sometimes changing how they present it, adding education or refining the guidance so it is more usable.

The same balance between evidence and practicality shapes survivorship care, where the challenge is not curing disease, but helping survivors avoid or detect late effects of treatment without overburdening them with unnecessary testing.

Refining surveillance for survivors

Matthew Ehrhardt, MD, MS, Department of Oncology, co-leads survivorship guideline development for the Children’s Oncology Group and is a member of the International Guideline Harmonization Group. His work focuses on when survivors of childhood cancer should receive screening for treatment-related complications, including heart disease and second cancers.

“When I started developing cardiomyopathy surveillance guidelines for survivors of childhood cancer, we knew the existing recommendations were too cumbersome for frontline physicians to use efficiently,” Ehrhardt said. “To improve this, we first asked how frequently we needed to screen various at-risk groups, which led to generation of evidence that then informed new guidelines. That set the precedent for identifying gaps and then generating evidence that leads to definitive guideline changes to better serve patients.”

The change in cardiomyopathy surveillance guidelines alone spared roughly 40% of survivors from unnecessary screening. That matters because survivors already face more frequent and earlier surveillance than their peers, which can be physically, emotionally and logistically taxing.

“The bulk of the guidelines remain dependent on a thorough history and physical examination,” Ehrhardt explained. “However, the rest of the recommendations are all about diagnostic testing: echocardiograms for heart health, along with breast, colorectal and thyroid cancer screenings and more.”

“Screening can put a great burden on patients, so we only want to include guidelines that have a clinically meaningful impact,” Ehrhardt said, echoing Ellison’s emphasis on utility.

Building guidance when datasets are limited

For some rare disorders, however, the challenge is different. Instead of sorting through an overwhelming volume of evidence, experts must create guidance despite not having enough patients to support large randomized clinical trials. That is the case for rare inherited bone marrow failure syndromes. One example is Diamond-Blackfan anemia (DBA), which is a multi-system disorder with anemia and involvement of other organ systems.

Marcin Wlodarski, MD, PhD, Department of Hematology, leads efforts to develop and maintain guidelines for its diagnosis, treatment and long-term surveillance. “We wanted to standardize best practices for how to diagnose, manage and treat patients, and do long-term surveillance, to improve patient outcomes,” Wlodarski said. “Experts from many centers across the world worked to come to a consensus.”

To achieve this, Wlodarski gathered physicians, nurses, scientists and patients to build consensus recommendations from the available evidence and from broad experiences. The consensus building was based on the existing evidence and expertise from national registries and was published in The Lancet Haematology in 2024. Based on this consensus manuscript, the St. Jude team created an easily accessible resource for rapid dissemination of knowledge. Physicians across the globe can rapidly access details on every aspect of DBA: www.stjude.org/DBA. 

Like pharmacogenomic guidelines, those recommendations continue to evolve as new research emerges. Regardless of the disease, all guideline efforts eventually face a similar question: How can the best guidance be made useful across very different care settings?

Adapting evidence to a global context

Most guidelines are developed around care in highly resourced settings, such as the United States or Western Europe. In countries with more limited resources, following these standards is not always feasible. Therefore, St. Jude investigators are participating in programs to help adapt guidelines into a more resource-conscious format, such as the St. Jude Global-supported tool Adapted Resource and Implementation Application (ARIA), which is custom-made to provide clinical decision support in resource-limited settings.

“We ask, ‘How can people use WHO-defined molecular tumors if they don’t have molecular tests?’” Ellison said. “So, we are finding ways to do these tests with histochemistry, the traditional and much more inexpensive method to classify tumors, as well as giving advice through ARIA on best practices when you don’t have all the diagnostic and treatment resources available that we have at St. Jude.”

Ehrhardt and his colleagues are taking a similar approach in survivorship care. “We are working with content and context experts across the globe to provide a framework of variably-resourced screening options,” he said. “Then, providers can choose which option works best in their specific context.”

In those cases, the gold standard remains the goal when possible. However, providing a range of endorsed options can still help clinicians deliver the best care available within their setting.

Guiding the future of diagnosing catastrophic diseases of childhood

From brain tumor diagnosis to pharmacogenomics, survivorship care and rare blood disorders, St. Jude investigators are helping shape how evidence becomes action. Their work ensures that guidelines do more than summarize discoveries. They help make those discoveries usable, reliable and relevant for patients.

“For many pediatric diseases, St. Jude has developed excellent internal standards, and we have worked to create general standards from there that we have published with experts worldwide,” Wlodarski said. “The ultimate goal is to use our experience and research to ensure that every child receives outstanding care, no matter where they are.”