Chun-Po Yen, Julie A. Matsumoto, Max Wintermark, Lucia Schwyzer, Avery J. Evans, Mary E. Jensen, Mark E. Shaffrey and Jason P. Sheehan
The objective of this study was to evaluate the incidence, severity, clinical manifestations, and risk factors of radiation-induced imaging changes (RIICs) following Gamma Knife surgery (GKS) for cerebral arteriovenous malformations (AVMs).
A total of 1426 GKS procedures performed for AVMs with imaging follow-up available were analyzed. Radiation-induced imaging changes were defined as newly developed increased T2 signal surrounding the treated AVM nidi. A grading system was developed to categorize the severity of RIICs. Grade I RIICs were mild imaging changes imposing no mass effect on the surrounding brain. Grade II RIICs were moderate changes causing effacement of the sulci or compression of the ventricles. Grade III RIICs were severe changes causing midline shift of the brain. Univariate and multivariate logistic regression analyses were applied to test factors potentially affecting the occurrence, severity, and associated symptoms of RIICs.
A total of 482 nidi (33.8%) developed RIICs following GKS, with 281 classified as Grade I, 164 as Grade II, and 37 as Grade III. The median duration from GKS to the development of RIICs was 13 months (range 2–124 months). The imaging changes disappeared completely within 2–128 months (median 22 months) following the development of RIICs. The RIICs were symptomatic in 122 patients, yielding an overall incidence of symptomatic RIICs of 8.6%. Twenty-six patients (1.8%) with RIICs had permanent deficits. A negative history of prior surgery, no prior hemorrhage, large nidus, and a single draining vein were associated with a higher risk of RIICs.
Radiation-induced imaging changes are the most common adverse effects following GKS. Fortunately, few of the RIICs are symptomatic and most of the symptoms are reversible. Patients with a relatively healthy brain and nidi that are large, or with a single draining vein, are more likely to develop RIICs.
Ching-Hsiao Cheng, R. Webster Crowley, Chun-Po Yen, David Schlesinger, Mark E. Shaffrey and Jason P. Sheehan
Gamma Knife surgery (GKS) has emerged as the treatment of choice for small- to medium-sized cerebral arteriovenous malformations (AVMs) in deep locations. The present study aims to investigate the outcomes of GKS for AVMs in the basal ganglia and thalamus.
Between 1989 and 2007, 85 patients with AVMs in the basal ganglia and 97 in the thalamus underwent GKS and were followed up for more than 2 years. The nidus volumes ranged from 0.1 to 29.4 cm3 (mean 3.4 cm3). The mean margin dose at the initial GKS was 21.3 Gy (range 10–28 Gy). Thirty-six patients underwent repeat GKS for residual AVMs at a median 4 years after initial GKS. The mean margin dose at repeat GKS was 21.1 Gy (range 7.5–27 Gy).
Following a single GKS, total obliteration of the nidus was confirmed on angiograms in 91 patients (50%). In 12 patients (6.6%) a subtotal obliteration was achieved. No flow voids were observed on MR imaging in 14 patients (7.7%). Following single or repeat GKS, total obliteration was angiographically confirmed in 106 patients (58.2%) and subtotal obliteration in 8 patients (4.4%). No flow voids on MR imaging were observed in 18 patients (9.9%). The overall obliteration rates following one or multiple GKSs based on MR imaging or angiography was 68%. A small nidus volume, high margin dose, low number of isocenters, and no history of embolization were significantly associated with an increased rate of obliteration. Twenty-one patients experienced 25 episodes of hemorrhage in 850 risk-years following GKS, yielding an annual hemorrhage rate of 2.9%. Four patients died in this series: 2 due to complications of hemorrhage and 2 due to unrelated diseases. Permanent neurological deficits caused by radiation were noted in 9 patients (4.9%).
Gamma Knife surgery offers a reasonable chance of obliterating basal ganglia and thalamic AVMs and does so with a low risk of complications. It is an optimal treatment option in patients for whom the anticipated risk of microsurgery is too high.
Jay Jagannathan, Chun-Po Yen, Dibyendu Kumar Ray, David Schlesinger, Rod J. Oskouian, Nader Pouratian, Mark E. Shaffrey, James Larner and Jason P. Sheehan
This study evaluated the efficacy of postoperative Gamma Knife surgery (GKS) to the tumor cavity following gross-total resection of a brain metastasis.
A retrospective review was conducted of 700 patients who were treated for brain metastases using GKS. Forty-seven patients with pathologically confirmed metastatic disease underwent GKS to the postoperative resection cavity following gross-total resection of the tumor. Patients who underwent subtotal resection or who had visible tumor in the resection cavity on the postresection neuroimaging study (either CT or MR imaging with and without contrast administration) were excluded. Radiographic and clinical follow-up was assessed using clinic visits and MR imaging. The radiographic end point was defined as tumor growth control (no tumor growth regarding the resection cavity, and stable or decreasing tumor size for the other metastatic targets). Clinical end points were defined as functional status (assessed prospectively using the Karnofsky Performance Scale) and survival. Primary tumor pathology was consistent with lung cancer in 19 cases (40%), melanoma in 10 cases (21%), renal cell carcinoma in 7 cases (15%), breast cancer in 7 cases (15%), and gastrointestinal malignancies in 4 cases (9%). The mean duration between resection and radiosurgery was 15 days (range 2–115 days). The mean volume of the treated cavity was 10.5 cm3 (range 1.75–35.45 cm3), and the mean dose to the cavity margin was 19 Gy. In addition to the resection cavity, 34 patients (72%) underwent GKS for 116 synchronous metastases observed at the time of the initial radiosurgery.
The mean radiographic follow-up duration was 14 months (median 10 months, range 4–37 months). Local tumor control at the site of the surgical cavity was achieved in 44 patients (94%), and tumor recurrence at the surgical site was statistically related to the volume of the surgical cavity (p = 0.04). During follow-up, 34 patients (72%) underwent additional radiosurgery for 140 new (metachronous) metastases. At the most recent follow-up evaluation, 11 patients (23%) were alive, whereas 36 patients had died (mean duration until death 12 months, median 10 months). Patients who showed good systemic control of their primary tumor tended to have longer survival durations than those who did not (p = 0.004). At the last clinical follow-up evaluation, the mean Karnofsky Performance Scale score for the overall group was 78 (median 80, range 40–100).
Radiosurgery appears to be effective in terms of providing local tumor control at the resection cavity following resection of a brain metastasis, and in the treatment of synchronous and metachronous tumors. These data suggest that radiosurgery can be used to prevent recurrence following gross-total resection of a brain metastasis.
Rod J. Oskouian Jr., Christopher I. Shaffrey, Richard Whitehill, Charles A. Sansur, Nader Pouratian, Adam S. Kanter, Ashok R. Asthagiri, Aaron S. Dumont, Jason P. Sheehan, W. Jeffrey Elias and Mark E. Shaffrey
The purpose of this study was to evaluate the results obtained in patients who underwent anterior stabilization for three-column thoracolumbar fractures.
The authors retrospectively reviewed available clinical and radiographic data (1997–2006) to classify three-column thoracolumbar fractures according to the Association for the Study of Internal Fixation (AO) system, neurological status, spinal canal compromise, pre- and postoperative segmental angulation, and arthrodesis rate.
The mean computed tomography–measured preoperative spinal canal compromise was 48.3% (range 8–92%), and the mean vertebral body height loss was 39.4%. The mean preoperative kyphotic deformity of 14.9° improved to 4.6° at the final follow-up examination. Although this angulation had increased a mean of 1.8° during the follow-up period, the extent of correction was still significant compared with the preoperative angulation (p < 0.01). There were no cases of vascular complication or neurological deterioration.
Contemporary anterior spinal reconstruction techniques can allow certain types of unstable three-column thoracolumbar fractures to be treated via an anterior approach alone. Compared with traditional posterior approaches, the anterior route spares lumbar motion segments and obviates the need for harvesting of the iliac crest.