The proximal femur is the most frequent operative site for metastasis, but there is no consensus between internal fixation and hip replacement. The present multicenter retrospective observational study sought: (1) to compare early clinical results between internal fixation and hip replacement for proximal femoral metastasis (PFM), and (2) to assess events affecting survival.

The study hypothesis was that internal fixation and hip replacement give comparable clinical results, operative site complications rates and survival.

The series comprised 309 cases, 10 of which were bilateral, in 182 females and 117 males, with a mean age of 67.2±11.5 years and 62.5±13.2 years, respectively. Primaries were mainly breast (118; 38.2%), lung (85; 25.5%) or kidney (40; 12.9%). PFM was revelatory in 114 cases (36.9%). There was visceral involvement in 142 patients (46%), multiple peripheral bone involvement in 212 (68.6%), and spinal involvement in 134 (43.4%). There were 124 pathologic fractures (40%), 51 of which were revelatory. Metastases were cervicocephalic in 135 cases (43.7%), metaphyseal in 166 (53.7%) and both in 8 (2.6%). PFM was osteolytic in 90% of cases, managed by hip replacement in 161 cases and internal fixation in 148 (12 screwed plates, 136 nails). Seventy-seven patients had postoperative radiation therapy.

After hip replacement (n=144), walking was normal in 35 cases (24.3%), impaired but unassisted in 53 (36.8%), with 1 forearm crutch in 24 (16.6%), 2 crutches or a frame in 26 (18%), and impossible in 6 (4.1%). After nailing (n=125), results were respectively 38 (30.4%), 47 (37.6%), 15 (12%), 18 (14.4) and 7 (5.6%). Recovery of normal walking capacity did not significantly differ according to technique (p=0.162); nor did pain or function. Recovery of normal walking capacity was better after preventive surgery (p<0.001). Perioperative complications comprised: 10 cases of severe blood loss, 7 pulmonary embolisms, 6 digestive hemorrhages, and 5 lung infections. Secondarily, there were 11 infections (7 after hip replacement, 2 after nailing, 2 after plate fixation), 7 progressive osteolyses and 5 fractures. Complications rates were significantly higher with plate fixation, with no difference between nailing and hip replacement. Survival did not significantly differ between hip replacement (12 months [95% CI: 7–19]), nailing (7 months [95% CI: 6–11]) and plate fixation (16 months [95% CI: 6-not calculable]).

Clinical results and survival were comparable between hip replacement and nailing, confirming the study hypothesis, in agreement with the literature. Each technique has its indications. Patients with severely impaired walking capacity benefited greatly from surgery. The importance of preventive surgery was highlighted.

IV, retrospective study.