Computed tomography-based structural analysis for predicting fracture risk in children with benign skeletal neoplasms: comparison of specificity with that of plain radiographs

Abstract

Background: The decision whether to treat benign skeletal lesions surgically can be difficult to make. The purpose of this study was to validate our previously published method of predicting fracture risk with use of quantitative computed tomography-based structural analysis.

Methods: We prospectively studied a group of children who presented to a major children's hospital with a benign appendicular skeletal lesion between 2002 and 2007. As in our previous study, the resistance of the affected bone to compressive, bending, and torsional loads was calculated with rigidity analysis performed with the use of serial transaxial quantitative computed tomography data obtained along the length of the bone containing the lesion and from homologous cross sections through the contralateral, normal bone. At each cross section, the ratio of the structural rigidity of the affected bone to that of the normal, contralateral bone was determined.

Results: Forty-one patients who had not received surgical treatment for the skeletal lesion met the criteria for our study. Thirty-four (83%) of these individuals completed our activity questionnaire at least two years after the quantitative computed tomography-based rigidity analysis. None of the patients for whom no increased fracture risk had been predicted by the rigidity analysis sustained a fracture, even though they had not received surgical treatment.

Conclusions: Many considerations other than the predicted fracture risk are factored into the decision of whether to treat a benign skeletal lesion. However, this study indicated that quantitative computed tomography-based rigidity analysis is more specific (97% specificity) than criteria based on plain radiographs (12% specificity) for predicting the risk of a pathologic fracture since fracture risk indices based on lesion size alone fail to account for the compensatory remodeling of the host bone that occurs in response to the presence of the lesion in a growing child.

Level of evidence: Prognostic Level I. See Instructions to Authors for a complete description of levels of evidence.

Fig. 1

Lower-extremity computed tomography, with a hydroxyapatite phantom, showed a nonossifying fibroma with destruction of 61% of the cortex. The left tibia was predicted to be at risk for fracture with standard radiographic methods but not with quantitative computed tomography-based rigidity analysis. This patient was not treated surgically and had not sustained a fracture over a six-year period.

Fig. 2

Axial (EA), bending (EI), and torsional (GJ) rigidity were calculated, with use of the indicated algorithms, from transaxial computed tomography images of the affected and contralateral, unaffected bone. The ratio of the rigidity of the affected bone normalized by that of the contralateral limb at a homologous cross section was calculated. The bone was predicted to be at risk of fracture if EA, EI, or GJ was ≤65% of that on the contralateral side. da = pixel size.

Fig. 3

Fracture prediction and treatment status of the patient cohort. Seventeen patients were treated surgically while forty-one patients were not. Fifteen of the seventeen patients who were treated surgically responded to our treatment questionnaire regarding the reasons why they elected to have surgery. Of the forty-one patients who underwent surgery, we obtained follow-up information from thirty-four (83%), six could not be contacted, and one refused to participate in the study. qCT = quantitative computed tomography-based structural rigidity analysis, and X-ray = plain radiography.