About Us

Periosteal Reaction

The periosteum is a membrane several cell layers thick that covers almost all of every bone. About the only parts not covered by this membrane are the parts covered by cartilage. Besides covering the bone and sharing some of its blood supply with the bone, it also produces bone when it is stimulated appropriately. What does it take to make this happen? Practically anything that breaks, tears, stretches, inflames, or even touches the periosteum. So, when some anonymous process stimulates this reactive bone formation, eventually we see evidence of it on some imaging study.

Once we spot this reactive new bone, how do we deal with it? In the best of all possible worlds, one would be able to look at the pattern of periosteal reaction and then give a precise histological diagnosis. Alas, this is not that kind of world. We can’t give a precise histological diagnosis. But wait — it gets worse! We can’t even tell for sure if the underlying process is benign or malignant! As it turns out, about all we can do is say with some confidence whether the process is a benign or an aggressive one. Why is this? Well, the periosteum is a fairly promiscuous tissue, and puts on a similar response to all comers. The main determinant of how the new bone formation looks is how fast the abnormal process grows, and has little to do with any intrinsic properties of the periosteum. Therefore, any differences in the pattern of periosteal reaction must arise in the disease process itself — not in the periosteum. Again, evidence of the speed at which these processes are growing is the main thing we look for when assessing periosteal reaction. Knowledge of this speed will help us to differentiate these processes into two broad categories.

With slow-growing processes, the periosteum has plenty of time to respond to the process. That is, it can produce new bone just as fast as the lesion is growing. Therefore, one would expect to see solid, uninterrupted periosteal new bone along the margin of the affected bone.


solid periosteal reaction along the cortex of a bone

figure after Ragsdale, et al 1981

However, with rapidly growing processes, the periosteum cannot produce new bone as fast as the lesion is growing. Therefore, rather than a solid pattern of new bone formation, we see an interrupted pattern. This interrupted pattern can manifest itself in several ways, depending on just how steadily the lesion grows. If the lesion grows unevenly in fits and starts, then the periosteum may have time to lay down a thin shell of calcified new bone before the lesion takes off again on its next growth spurt. This may result in a pattern of one or more concentric shells of new bone over the lesion. This pattern is sometimes called lamellated or “onion-skin” periosteal reaction.

Single LamellaMultiple Lamellae

lamellated periosteal reaction

figure after Ragsdale, et al 1981

If the lesion grows rapidly but steadily, the periosteum will not have enough time to lay down even a thin shell of bone, and the pattern may appear quite different. In such cases, the tiny fibers that connect the periosteum to the bone (Sharpey’s fibers) become stretched out perpendicular to the bone. When these fibers ossify, they produce a pattern sometimes called “sunburst” or “hair-on-end” periosteal reaction, depending of how much of the bone is involved by the process.

SunburstHair on end

“sunburst” and “hair-on-end” periosteal reaction

figure after Ragsdale, et al 1981

q6AP Lat Knee

Osteosarcoma of the distal femur, demonstating dense tumor bone formation and a sunburst pattern of periosteal reaction.

Another pattern seen in rapidly growing processes is called the Codman’s triangle. This is a bit of a misnomer, since there really is not a complete triangle. When a process is growing too fast for the periosteum to respond with even thin shells of new bone, sometimes only the edges of the raised periosteum will ossify. When this little bit of ossification is seen tangentially on a radiograph, it forms a small angle with the surface of the bone, but not a complete triangle. So, when a process is growing too fast for even the Sharpey’s fibers to ossify, one may only see a soft tissue mass arising from the bone, perhaps with small Codman’s triangles at its margins.


a Codman’s triangle

figure after Ragsdale, et al 1981

Soooooo…… what is the significance of all of these patterns? Well, we can usually differentiate lesions into one of two categories: benign vs. aggressive processes. If we see a solid pattern of periosteal reaction, we can be fairly confident that we are dealing with a benign process. How confident? In normal everyday practice, my estimate is that you can be about 90 – 95 % confident in this rule(9), but your mileage may vary. As with many rules in medicine, there are some caveats associated with the use of this rule. The main caveat with this rule is that benign processes and malignant processes may coexist. The usual way that this may manifest is when there is a fracture or infection in the same area as a tumor. In this case, you may see a fairly complex pattern of periosteal reaction that demonstrates some elements that look benign and some that look very aggressive.


a complex pattern of periosteal reaction

figure after Ragsdale, et al 1981

The take home point here is that complex patterns like this may be very misleading, and should be interpreted with caution. In general, though, the more aggressive the pattern of periosteal reaction, the greater the chance that you are dealing with a malignancy.

Here are partial lists of causes of both solid and aggressive periosteal reaction:

Causes of Solid Periosteal Reaction

  • infection
  • benign neoplasms
    • osteoid osteoma
  • eosinophilic granuloma
  • hypertrophic pulmonary osteoarthropathy
  • deep venous thrombosis (lower extremity)

Causes of Aggressive Periosteal Reaction

  • osteomyelitis
  • malignant neoplasms
    • osteosarcoma
    • chondrosarcoma
    • fibrosarcoma
    • lymphoma
    • leukemia
    • metastasis


  1. Ragsdale BD, Madewell JE, Sweet DE. Radiologic and pathologic analysis of solitary bone lesions. Part II: Periosteal reactions.Radiol Clin N Am 1981;19:749-783.