Michael S. Muhlbauer

PURPOSE: To evaluate the impact of primary tumor site, age at diagnosis, extent of resection, and histology on progression-free survival (PFS) in pediatric low-grade astrocytoma. PATIENTS AND METHODS: Medical, pathologic, and imaging information were reviewed for 142 children (ages 2 months to 19 years) with low-grade astrocytoma treated between January 1984 and July 1994. Gross total resection (GTR) was attempted for cerebellar and cerebral hemisphere tumors, with biopsy or less aggressive resection used predominantly for tumors in other sites. Surgery was followed by observation in 107 cases, radiation therapy in 31, and chemotherapy in four. RESULTS: The overall survival rate was 90% +/- 3% (SE) at 4 years. PFS was significantly better for patients with cerebellar and cerebral hemisphere tumors (n = 75) than those with tumors in all other sites (P = .0006). Within the former group, there was no significant difference in PFS for patients in whom GTR was achieved versus those with incomplete resections (4-year estimates, 89% and 77%, respectively). Histology (juvenile pilocytic v astrocytoma not otherwise specified [NOS]) was not related to PFS in an analysis that controlled for tumor site and patient age. Patients younger than 5 years at diagnosis had a significantly poorer PFS than older children, regardless of histology (P < .03) or tumor site (P < .002). Treatment for progressive/recurrent disease was effective in a majority of patients, but appeared more successful in patients with hemispheric than thalamic or hypothalamic tumors. CONCLUSION: The overall survival in this series of pediatric low-grade astrocytomas is excellent. Age at diagnosis and tumor location, but not histology, had a significant impact on PFS. Efforts to improve treatment outcome should focus on young patients (< 5 years) and on those with central midline tumors. The majority of patients with completely resected hemispheric tumors were monitored without further therapy, which supports attempted GTR of cerebral and cerebellar hemisphere low-grade astrocytoma.

OBJECT: Surgical site infection (SSI) is a serious and costly complication of spinal surgery. There have been several conflicting reports on the use of intrawound vancomycin powder in decreasing SSI in spine surgery. The purpose of this study is to answer the question: "Does intrawound vancomycin powder reduce the rate of SSIs in spine surgery?" METHODS: A comprehensive search of multiple electronic databases and bibliographies was conducted to identify clinical studies that evaluated the rates of SSI with and without the use of intrawound vancomycin powder in spine surgery. Independent reviewers extracted data and graded the quality of each paper that met inclusion criteria. A random effects meta-analysis was then performed. RESULTS: The search identified 9 retrospective cohort studies (Level III evidence) and 1 randomized controlled trial (Level II evidence). There were 2574 cases and 106 infections in the control group (4.1%) and 2518 cases and 33 infections (1.3%) in the treatment group, yielding a pooled absolute risk reduction and relative risk reduction of 2.8% and 68%, respectively. The meta-analysis revealed the use of vancomycin powder to be protective in preventing SSI (relative risk = 0.34, 95% confidence interval 0.17-0.66, p = 0.021). The number needed to treat to prevent 1 SSI was 36. A subgroup analysis found that patients who had implants had a reduced risk of SSI with vancomycin powder (p = 0.023), compared with those who had noninstrumented spinal operations (p = 0.226). CONCLUSIONS: This meta-analysis suggests that the use of vancomycin powder may be protective against SSI in open spinal surgery; however, the exact population in which it should be used is not clear. This benefit may be most appreciated in higher-risk populations or in facilities with a high baseline rate of infection.

BACKGROUND: Radiation-induced necrosis (RIN) of the brain is a complication associated with the use of aggressive focal treatments such as radioactive implants and stereotactic radiosurgery. In an attempt to treat patients with central nervous system (CNS) RIN, ten patients received hyperbaric oxygen treatment (HBOT). METHODS: Patients presented with new or increasing neurologic deficits associated with imaging changes after radiotherapy. Necrosis was proven by biopsy in eight cases. HBOT was comprised of 20-30 sessions at 2.0 to 2.4 atmospheres, for 90 minutes-2 hours. Sites of RIN included the brain stem (n = 2), posterior fossa (n = 1), and supratentorial fossa (n 7). Histologic types included brain stem glioma (n = 2), ependymoma (n = 2), germinoma (n = 2), low grade astrocytoma (n = 1), oligodendroglioma (n = 1), glioblastoma multiforme (n = 1), and arteriovenous malformation (n = 1). RESULTS: Initial improvement or stabilization of symptoms and/or imaging findings were documented in all ten patients studied and no severe HBOT toxicity was observed. Four patients died, with the cause of death attributed to tumor progression. Five of six surviving patients were improved by clinical and imaging criteria; one patient was alive with tumor present at last follow-up. CONCLUSIONS: HBOT may prove to be an important adjunct to surgery and steroid therapy for CNS RIN.

A retrospective analysis of 76 civilian craniocerebral gunshot wounds treated over a 20-month period is presented. The authors report a 62% mortality rate and conclude that the admission Glasgow Coma Scale (GCS) score is a valuable prognosticator of outcome. Other important findings were: patients with a GCS score of 3 invariably died, with or without surgical intervention; and the presence of intracranial hematomas, ventricular injury, or bihemispheric wounding was associated with a poor outcome. Standardized methods of data reporting should be adopted in order to allow multicenter trials or comparisons that might lead to management practices that could improve results.

A total of 2259 children with solid malignant tumors were treated at St. Jude Children's Research Hospital between the years 1962 and 1987. Of these, 112 (5%) developed spinal epidural metastasis with spinal cord compression during the course of their disease process. Metastatic epidural spinal cord compression was caused most commonly by Ewing's sarcoma and neuroblastoma, followed by osteogenic sarcoma, rhabdomyosarcoma, Hodgkin's disease, soft-tissue sarcoma, germ-cell tumor, Wilm's tumor, and (rarely) hepatoma. There was no significant difference in outcome between patients with small-cell tumors (neuroblastoma, Hodgkin's disease, and germ-cell tumors) who received only chemotherapy and/or radiation therapy and the patients with similar lesions who received a decompressive laminectomy alone or prior to chemotherapy and/or radiation therapy. Patients with spinal cord compression from metastatic sarcoma (Ewing's sarcoma, soft-tissue sarcoma, osteogenic sarcoma, and rhabdomyosarcoma) showed a significant improvement with decompressive laminectomy alone or before medical therapy, compared to those who received radiation therapy and/or chemotherapy without posterior decompression. Pediatric tumors invade the spinal canal via the neural foramen, compressing the spinal cord in a circumferential manner, allowing decompressive laminectomy (posterior approach) to be an effective surgical approach. Sixty-six percent of children who had no evidence of motor or sensory function below the level of the compression became ambulatory after surgical decompression and medical treatment, regardless of tumor type.