The signal intensity patterns of benign osteoporotic compression fractures are usually variable. In the majority of cases, the signal-intensity changes are localised (Fig. 4a) and normal bone marrow is visualised in at least one area within the vertebral body (Sugimura et al. 1987). Indeed, most vertebral metastases do not result in compression fractures until the entire body is infiltrated by the tumour (Fig. 4b), causing destruction of the trabeculae, the cortex or both (Yuh et al. 1989). However, it has been reported that in 15 % of acute osteoporotic compression fractures (Fig. 4c) there is complete replacement with abnormal signal intensity (Shih et al. 1999). This might be due to spread to the entire vertebra of a reactive process induced by the fracture. Therefore, in acute osteoporotic fractures, when the height of the vertebral body is preserved and diffused signal-intensity changes are seen on T1-weighted, STIR, and contrast-enhanced T1-weighted images, the diagnosis may be very difficult, since these imaging features mimic those encountered in malignant compression fractures. In chronic osteoporotic fractures there is no abnormal signal intensity identified in the bone marrow of the compressed vertebra (Tan et al. 2002).

A cleft of fluid (demonstrated as a linear area of high signal-intensity) or gas within the vertebral body on T2-weighted and STIR images (Fig. 5a, b), have been described in association with osteoporotic fractures (Baur et al. 2002). These are considered to represent avascular necrosis or pseudoarthrosis. In particular, the “fluid sign” is seen adjacent to the fractured endplates and the signal is isointense to that of cerebrospinal fluid (Oka et al. 2005).

In acute benign compression fractures, a band like low signal intensity area on T1-weighted and T2-weighted images is commonly seen (Fig. 6a, b). This area is sited adjacent to the collapsed endplate and is encountered in 93 % of acute osteoporotic fractures as opposed to 44 % of metastatic ones (Jung et al. 2003).

Retropulsion of a posterior bone fragment is believed to be sensitive and specific for osteoporotic fractures. It is detected in 60 % of acute osteoporotic fractures as opposed to 11 % of metastatic collapses (Fig. 7). In cases of metastatic disease, retropulsion might be due to preexisting compression fractures that were secondarily involved by metastatic disease. It is also possible that metastatic disease can sometimes be associated with retropulsion of bone fragments (Kaplan et al. 1987).

Pedicle involvement has been reported as specific for malignancy (Fig. 8a–d). However, it can be found in many patients with osteoporotic fractures, particularly in the early phase. It has been reported that pedicle abnormal signal intensity and contrast enhancement are seen in 64 % of benign compression fractures and in 84.2 % of malignant ones, with no significant difference between groups (Ishiyama et al. 2010). In general, the criteria of pedicle infiltration by tumour include the presence of inhomogeneous diffuse signal-intensity changes or a mass effect.

Nonetheless, it should be stressed that signal-intensity changes in osteoporotic fractures are often focal and inhomogeneous as well. When pedicle involvement is the only sign, a diagnosis of malignant pathologic fracture should not be assumed. Sclerosis of the pedicle is seen during the healing process of microfractures and is considered to represent reactive change. This reflects the signal intensity patterns noted on MRI, which represent infiltration by inflammatory cells, granulation tissue or fibrosis.