Determinants of outcome in melanoma patients with cerebral metastases.

PURPOSE: To analyze prognostic factors, effects of treatment, and survival for patients with cerebral metastases from melanoma. PATIENTS AND METHODS: All melanoma patients with cerebral metastases treated at the Sydney Melanoma Unit between 1952 and 2000 were identified. From 1985 to 2000, patients were diagnosed and treated using consistent modern techniques and this cohort was analyzed in detail. Multivariate analysis of prognostic factors for survival was performed. RESULTS: A total of 1137 patients with cerebral metastases were identified; 686 were treated between 1985 and 2000. For these 686 patients, the median time from primary diagnosis to cerebral metastasis was 3.1 years (range, 0 to 41 years). A total of 646 patients (94%) have died as a result of melanoma. The median survival from the time of diagnosis of cerebral metastasis was 4.1 months (range, 0 to 17.2 years). Treatment was as follows: surgery and postoperative radiotherapy, 158 patients; surgery alone, 47 patients; radiotherapy alone, 236 patients; and supportive care alone, 210 patients. Median survival according to treatment received for these four groups was 8.9, 8.7, 3.4, and 2.1 months, respectively; the differences between surgery and nonsurgery groups were statistically significant. On multivariate analysis, significant factors associated with improved survival were surgical treatment (P <.0001), no concurrent extracerebral metastases (P <.0001), younger age (P =.0007), and longer disease-free interval (P =.036). Prognostic factors analysis confirmed the important influence of patient selection on treatment received. CONCLUSION: This large series documents the characteristics of patients who developed cerebral metastases from melanoma. Median survival was dependent on treatment, which in turn was dependent on patient selection.

Abduction strength following intramedullary nailing of the femur.

OBJECTIVES: To assess hip abductor function, strength and complaints following insertion of a femoral intramedullary nail. DESIGN: Retrospective clinical review. SETTING: Department of Orthopaedics, Adelaide Women's and Children's Hospital, Adelaide. Department of Orthopaedic Surgery and Trauma, Royal Adelaide Hospital, Adelaide, South Australia, Australia. PATIENTS: 1. 32 of 37 patients who had an intramedullary nail inserted for an isolated femoral shaft fracture at the Royal Adelaide Hospital between 1987 and 1990. 2. 14 of 18 patients who had closed femoral shortening for leg length discrepancy, at the Adelaide Women's and Children's Hospital between 1985 and 1987. Patients with pathology involving the abductor mechanism were excluded. 3. 40 asymptomatic controls. INTERVENTION: Intramedullary fixation for femoral shaft fractures or as part of closed femoral shortening. All procedures were performed on a traction table via a gluteal splitting approach with reamed nails. MAIN OUTCOME MEASUREMENTS: Complaints included, pain, stiffness, limp and diminished walking distance. Examination of abductor function and measurement of abductor strength. Radiological assessment at followup. RESULTS: Complaints included trochanteric pain (40%, 40%), thigh pain (10%, 8%) and limp (13%, 42%) in the femoral fracture and closed femoral shortening groups respectively. There was significant difference in the abduction strength (p < 0.01) and abduction ratio (p < 0.01) between the control and each treatment group. Abductor weakness correlated (r = 0.30) with the incidence of complaints. CONCLUSION: Pain, limp and weakness are common following insertion of a femoral intramedullary nail. Agluteal retracting approach may minimize abductor weakness.

Recurrent haemorrhage from a malignant brain stem glioma: a review of the mechanisms of tumoural haemorrhage.

The authors report a 23-year-old man who was diagnosed with a brain stem malignant glioma following his first episode of generalised seizure. This was subsequently complicated by three separate documented episodes of intracranial haemorrhage. The literature is reviewed and the pathophysiological mechanisms of tumoural haemorrhage are discussed. It is speculated that significant intratumoural arteriovenous shunting may lead to recurrent haemorrhages.