Surgery and Soft-Tissue Tumours

Objective: In randomised Dutch Bone Metastasis Study on the palliative effect of single versus multiple fraction irradiation on painful bone metastases all impending fractures were excluded from participation. During intensive follow-up 35 fractures occurred in 1157 patients. A subset of 102 patients with a femoral lesion, in which 8 fractures occurred within 12 weeks after irradiation (8%), was used to assess the pre-treatment fracture risk. Methods: Pre-treatment radiographs were collected. A radiologist, orthopaedic surgeon and radiation oncologist independently and without knowledge of the outcome after treatment (fracture yes/ no) scored size and extent in mm, radiographic appearance/structure/aspect, localisation, percentage of cortical involvement and estimated risk of fracturing. Known risk factors (size >2.5 cm, cortical involvement >50% or continuing pain after irradiation) and an adjusted Mirels’ scoring system* were applied to evaluate effectiveness (table1). Results: If a score of 9 points as Mirels suggested was applied sensitivity for predicting a fracture was high but specificity very low (table2). The positive predictive value of Mirels was low (<10%). Size >2.5 cm, cortical involvement >50% or continuing pain were not predictive of fracturing. Conclusion: In this study reliability of accepted risk factors and scoring systems in predicting fracturing in patients with a femoral bone metastasis is low. The use of Mirels’ scoring system will lead to excessive surgical overtreatment (positive predictive value <10%). Also size >2.5 cm, cortical involvement >50% or continuing pain after irradiation are not specific enough. Development of a more sensitive scoring system is ongoing.