566 THE JOURNAL OF BONE AND JOINT SURGERY Tumor Comparative analysis of risk factors for pathological fracture with femoral metastases RESULTS BASED ON A RANDOMISED TRIAL OF RADIOTHERAPY Y. M. Van der Linden, P. D. S. Dijkstra, H. M. Kroon, J. J. Lok, E. M. Noordijk, J. W. H. Leer, C. A. M. Marijnen From Leiden University Medical Centre, Leiden, The Netherlands  Y. M. Van der Linden, MD, Radiation Oncologist  E. M. Noordijk, MD, Radiation Oncologist  C. A. M. Marijnen, MD, Radiation Oncologist Department of Clinical Oncology  P. D. S. Dijkstra, MD, Orthopaedic Surgeon Department of Orthopaedic Surgery  H. M. Kroon, MD, Radiologist Department of Radiology  J. J. Lok, PhD, Statistician Department of Medical Statistics Leiden University Medical Centre, Albinusdreef 2, 2300 RC, Leiden, The Netherlands.  J. W. H. Leer, MD, Radiation Oncologist Department of Radiotherapy, St Radboud University Medical Center, PB 9101 6500 HB, Nijmegen, The Netherlands. Correspondence should be sent to Dr Y. M. Van der Linden. ©2004 British Editorial Society of Bone and Joint Surgery doi:10.1302/0301-620X.86B4. 14703 $2.00 J Bone Joint Surg [Br] 2004;86-B:566-73. Received 13 June 2003; Accepted after revision 18 September 2003 A number of risk factors based upon mostly retrospective surgical data, have been formulated in order to identify impending pathological fractures of the femur from low-risk metastases. We have followed up patients taking part in a randomised trial of radiotherapy, prospectively, in order to determine if these factors were effective in predicting fractures. In 102 patients with 110 femoral lesions, 14 fractures occurred during follow-up. The risk factors studied were increasing pain, the size of the lesion, radiographic appearance, localisation, transverse/axial/circumferential involvement of the cortex and the scoring system of Mirels. Only axial cortical involvement >30 mm (p = 0.01), and circumferential cortical involvement >50% (p = 0.03) were predictive of fracture. Mirels’ scoring system was insufficiently specific to predict a fracture (p = 0.36). Our results indicate that most conventional risk factors overestimate the actual occurrence of pathological fractures of the femur. The risk factor of axial cortical involvement provides a simple, objective tool in order to decide which treatment is appropriate. A pathological fracture in a metastatic lesion in the femur can be most distressing and cause considerable morbidity. Metastatic lesions with a high risk of fracture require surgical sta- bilisation using prophylactic osteosynthesis. Painful, low-risk lesions can be treated con- servatively using external beam radiotherapy, chemotherapy, hormonal therapy or regular infusions with bisphosphonates. However, it is difficult to differentiate between low- and high-risk lesions on their radiographic appear- ance. Attempts have been made to formulate objective risk factors for impending fracture, mostly using surgical and retrospective data, in order to decide which lesions need prophy- lactic osteosynthesis and which can be treated conservatively. Factors mentioned include the size of a lesion (>25 mm), 1-3 a radiographic osteolytic appearance, 1-9 the percentage of circumferential cortical involve- ment (>50%), 3,7,8,10-14 and increasing local pain. 1-4,7-10,14-17 In 1989, Mirels 7 proposed a scoring system for the prediction of fracture in which several radiographic and clinical factors were combined into a single score. However, the majority of the patients in these studies had presented with a fracture or had undergone prophylactic osteosynthesis. Little is known about the natural behaviour of similar lesions without surgical fixation. It is possible that strict application of these risk factors leads to surgical overtreatment in patients who only have a limited life expectancy. We therefore studied the prognostic value of conventional risk factors and the scoring system of Mirels in 102 patients with femoral metastases treated conservatively. They received external beam radiotherapy as part of a large, prospective, randomised multicentre trial 18 which was designed to assess the pallia- tive effect on pain of a single fraction of 8 Gy as opposed to six fractions of 4 Gy in bone metastases. A total of 1157 patients were ran- domised with a median follow-up of 21 months. There were no major differences between the two radiation schedules with regard to pain, overall survival, the quality of life, consumption of analgesics or side-effects of treatment. Although the patients with a fem- oral lesion were considered by the treating physician to have a low risk of pathological fracture, 14 fractures occurred during follow- up. We have reviewed all the patients with fem- oral metastases in the trial in order to evaluate the risk of pathological fracture of the femur. Patients and Methods Patient selection and follow-up. Between March 1996 and September 1998, 1157 patients with painful bone metastases from solid tumours were randomised to treatment with either a single fraction of 8 Gy (n = 579) or six frac-