Comparator-Dependent Discrimination of sST2, Galectin-3, and Copeptin in Pediatric Chest Pain: A Prospective Exploratory Study
This prospective study demonstrates that while soluble suppression of tumorigenicity 2, galectin-3, and copeptin showed near-perfect discrimination between pediatric patients with chest pain and healthy controls, they failed to distinguish between those with suspected myocarditis and those with noncardiac chest pain, highlighting that their apparent diagnostic utility is heavily dependent on the comparator group selected.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Chest pain in children is a common reason for families to visit the emergency room, yet the vast majority of these cases turn out to be harmless. Most often, the discomfort stems from muscle strain, anxiety, or minor respiratory issues rather than a problem with the heart itself. The real challenge for doctors lies in identifying the rare child whose pain signals a serious condition like myocarditis, an inflammation of the heart muscle that can be life-threatening. To avoid missing these dangerous cases, physicians currently rely on a combination of listening to the patient, examining them, and running tests like electrocardiograms and blood work. However, because heart problems are so rare in this group, doctors worry that testing every child is inefficient and can lead to unnecessary follow-up procedures. This has sparked interest in finding specific substances in the blood, known as biomarkers, that could act as early warning signals, pointing directly to heart inflammation without the need for more invasive testing.
A team of researchers in Turkey set out to test three such substances: soluble suppression of tumorigenicity 2, galectin-3, and copeptin. These are molecules that the body releases during stress or when the heart is under strain. The scientists wanted to know if measuring these levels in the blood could help distinguish between a child with a suspected heart inflammation and a child with chest pain caused by something else entirely. They recruited 66 children between the ages of 6 and 17 who had come to the hospital with chest pain that was not caused by an injury. Alongside these patients, they also recruited 22 healthy children who had no symptoms at all to serve as a baseline for comparison. The researchers collected blood samples from everyone and measured the levels of the three biomarkers, but they kept these results hidden from the doctors treating the children to ensure the medical care was not influenced by the study data.
The study followed a strict process to determine the true cause of each child's pain. After the initial visit, a panel of pediatric cardiologists reviewed the medical history, physical exams, and standard test results to classify the children. Seven of the children were identified as having clinically suspected myocarditis or myopericarditis, while the remaining 59 were diagnosed with noncardiac causes, such as muscle pain or anxiety. When the researchers compared the biomarker levels between these two groups of sick children, the results were surprisingly flat. The levels of all three substances overlapped so much that it was impossible to tell who had the heart inflammation and who did not based on the blood test alone. In statistical terms, the ability of these markers to separate the two groups was no better than flipping a coin. The confidence intervals around these measurements were wide, meaning the study could not rule out a small effect, but it found no evidence that these biomarkers were useful for sorting out which sick child needed urgent heart care.
The picture changed dramatically when the researchers looked at the data differently. When they compared all 66 children with chest pain against the 22 healthy children with no symptoms, the biomarkers performed almost perfectly. The levels of the three substances were significantly higher in the group with chest pain than in the healthy group, creating a clear separation between the two. This result, however, was not a sign that the tests were good at diagnosing heart disease. Instead, it highlighted a critical flaw in how medical tests are sometimes evaluated. The markers were simply detecting that something was wrong in the body of a child with chest pain, regardless of whether the cause was a heart issue or a muscle strain. They could tell the difference between a sick child and a well child, but they could not tell the difference between a child with a dangerous heart condition and a child with a benign one.
This distinction is vital for understanding the value of a medical test. A test that looks impressive when comparing sick patients to healthy people can be completely useless when doctors need to make a decision among patients who are all already sick and presenting with the same symptom. The researchers concluded that these three biomarkers do not offer a reliable way to triage children with chest pain in an emergency setting. The near-perfect separation seen between the symptomatic and healthy groups was an artifact of the study design, reflecting the broad difference between illness and wellness rather than the specific nature of the disease. The study serves as a reminder that for a biomarker to be truly helpful in the clinic, it must be tested against the difficult, real-world question of distinguishing between different causes of the same symptom, not just against a group of healthy people. Without this rigorous comparison, a test might appear accurate while failing to solve the actual problem doctors face every day.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.