Clinical Value and Risk Analysis of Intraoperative Frozen Section Diagnosis in 313 Pediatric Bone Lesions: A Single-Center Study

In the operating room, time is a precious and often scarce resource. When a surgeon encounters an unexpected growth in a child's bone, the decision to cut deeper, remove a limb, or simply scrape away the surface depends entirely on knowing what that growth is. Is it a harmless cyst, a stubborn infection, or a dangerous cancer? Waiting days for a final laboratory report is often not an option, as the surgery must proceed in a single session. To bridge this gap, pathologists use a technique called a frozen section. This involves flash-freezing a tiny piece of tissue, slicing it thinly, staining it with dye, and examining it under a microscope while the patient is still on the table. It is a high-stakes race against the clock, requiring the pathologist to make a rapid judgment that will dictate the rest of the operation. While this method works well for soft tissues like skin or organs, bone presents a unique and stubborn challenge. Bone is hard, often brittle, and difficult to slice cleanly without damaging the delicate cellular structures inside. For children, whose bones are still growing and whose diseases can be rare and confusing, getting this quick diagnosis right is even more critical, yet the margin for error feels dangerously small.

A team of researchers at the Jiangxi Women's and Children's Medical Center recently set out to measure just how reliable this rapid diagnosis is when applied to pediatric bone lesions. They looked back at 313 cases of children, ranging from ten months to seventeen years old, who underwent surgery for bone problems between 2015 and 2026. In every single case, the medical team compared the quick, on-the-spot diagnosis given during the operation with the final, detailed diagnosis made weeks later after the tissue was processed in the standard, slower way. The goal was to see how often the quick guess matched the final truth, and more importantly, to understand exactly where and why the quick diagnosis might fail. The results offered a reassuring picture of high accuracy, but also a clear map of the specific traps that can catch even experienced doctors off guard.

The study found that the rapid frozen section diagnosis was correct in the vast majority of cases. Out of the 313 children, the quick diagnosis matched the final result 95.2% of the time. This means that in nearly every instance, the surgeon received the right information to make the best possible decision for the child's treatment. However, the researchers were careful to look closely at the small fraction of cases where the two diagnoses did not agree. In about 4.8% of the cases, the quick diagnosis was either partially correct or completely wrong. When the tumor was malignant, or cancerous, the agreement rate was slightly lower, sitting at 90.6%, which highlighted that cancerous bone lesions are the most difficult to identify quickly. The researchers did not view these errors as random failures but as specific, predictable problems that could be solved with better preparation and technique.

One of the most common sources of confusion was the visual similarity between different types of cells. The researchers found that the quick diagnosis often stumbled when trying to tell the difference between a specific type of inflammatory condition called Langerhans cell histiocytosis and a common bone infection known as osteomyelitis. Both conditions look like a chaotic mix of inflammatory cells under the microscope, and in the rush of a frozen section, it is easy to mistake one for the other. In a few cases, a harmless infection was mislabeled as the cellular condition, or vice versa. Another tricky area involved giant cells, which are large, multi-nucleated cells that can appear in both benign cysts and more serious tumors. Without the full context of the final lab report, these giant cells could lead a pathologist to misdiagnose a simple cyst as a more complex tumor, or confuse a benign growth with a malignant one.

The physical nature of the bone tissue itself also played a major role in these diagnostic challenges. Bone is hard and often comes out of the body in small, shattered fragments. If the surgeon or the pathologist does not grab a piece that truly represents the whole problem, the quick diagnosis will be based on incomplete information. The study noted that in several cases, the tissue sample was simply too small or too damaged by the freezing process to show the clear features needed for a definitive answer. Sometimes, the tissue was so necrotic, or dead, that the cells looked fragmented and unreadable, leading the pathologist to see only inflammation rather than a tumor. In these instances, the lack of clear, healthy-looking cells made it impossible to distinguish between a benign process and a dangerous cancer.

To combat these risks, the researchers outlined a set of practical strategies that improved their success rate. The most effective tool was not a new machine or a faster stain, but better communication and preparation. Before the surgery even began, the pathologists needed to know the full story: the child's age, the location of the lesion, and what the X-rays or scans showed. When the pathologist, the surgeon, and the radiologist share this information, the pathologist can use the quick microscope view to confirm what they already suspect, rather than guessing in the dark. They also emphasized the need for better sampling techniques, ensuring that the piece of tissue sent for freezing was large enough and representative enough to show the true nature of the lesion. By improving the quality of the tissue slices and reducing the artifacts caused by freezing, the clarity of the image under the microscope improved significantly.

The study concluded that while diagnosing bone lesions in children during surgery is inherently difficult, it is a highly valuable tool when used correctly. The high rate of agreement between the quick and final diagnoses proves that the method works, provided the team understands its limitations. The errors that did occur were not due to a flaw in the concept of frozen sections, but rather to specific, identifiable pitfalls like poor sample quality or the confusing overlap of cell types. By recognizing these traps and ensuring that clinical history and imaging are always part of the conversation, medical teams can minimize the risk of error. The work confirms that with careful attention to detail and strong teamwork, the rapid diagnosis can safely guide surgeons through the complex landscape of pediatric bone disease, ensuring that every child receives the right treatment at the right time.