Gamma Knife surgery in the management of radioresistant brain metastases in high-risk patients with melanoma, renal cell carcinoma, and sarcoma.

OBJECT: The purpose of this study was to examine the results of using Gamma Knife surgery (GKS) for brain metastases from classically radioresistant malignancies. METHODS: The authors retrospectively reviewed the records of 76 patients with melanoma (50 patients), renal cell carcinoma (RCC; 23 patients), or sarcoma (3 patients) who underwent GKS between August 1998 and July 2007. Overall patient survival, intracranial progression, and local progression of individual lesions were analyzed. RESULTS: The median age of the patients was 57 years (range 18-85 years) and median Karnofsky Performance Scale (KPS) score was 80 (range 20-100). Sixty-two patients (81.6%) had uncontrolled extracranial disease. A total of 303 intracranial lesions (average 3.97 per patient, range 1-27 lesions) were treated using GKS. More than 3 lesions were treated in 30 patients (39.5%). Median GKS tumor margin dose was 18 Gy (range 8-30 Gy). Thirty-seven patients (48.7%) underwent whole brain radiation therapy. The actuarial 12-month rate for freedom from local progression for individual lesions was 77.7% and was significantly higher for RCC compared with melanoma (93.6 vs 63.0%; p = 0.001). The percentage of coverage of the prescribed dose to target volume was the only treatment-related variable associated with local control: 12-month actuarial rate of freedom from local progression was 71.4% for lesions receiving >or= 90% coverage versus 0.0% for lesions receiving < 90% (p = 0.00048). Median overall survival was 5.1 months after GKS and 8.4 months after the discovery of brain metastases. Univariate analysis revealed that KPS score (p = 0.000004), recursive partitioning analysis class (p = 0.00043), and single metastases (p = 0.028), but not more than 3 metastases, to be prognostic factors of overall survival. The KPS score remained significant after multivariate analysis. Overall survival for patients with a KPS score >or= 70 was 7.1 months compared with 1.3 months for a KPS score 3 metastases. Higher rates of local tumor control were achieved for RCC in comparison with melanoma, and this may have an effect on survival in some patients. Although outcomes generally remained poor in this study population, these results suggest that GKS can be considered as a treatment option for many patients with radioresistant brain metastases, even if these patients have multiple lesions.

Gamma knife surgery for trigeminal neuralgia: improved initial response with two isocenters and increasing dose.

OBJECT: The authors sought to evaluate the initial response of trigeminal neuralgia (TN) to gamma knife surgery (GKS) based on the number of shots delivered and radiation dose. METHODS: Between September 1998 and September 2003, some 63 patients with TN refractory to medical or surgical management underwent GKS at Upstate Medical University. Ten patients had multiple sclerosis and 25 patients had undergone prior invasive treatment. Gamma knife surgery was delivered to the trigeminal nerve root entry zone in one shot in 27 patients or two shots in 36 patients. The radiation dose was escalated to less than or equal to 80 Gy in 20 patients, 85 Gy in 21 patients, and greater than or equal to 90 Gy in 22 patients. Pain before and after GKS was assessed using the Barrow Neurological Institute Pain Scale and the improvement score was analyzed as a function of dose grouping and number of shots. Sixty patients were available for evaluation, with an initial overall and complete response rate of 90% and 27%, respectively. There was a greater improvement score for patients who were treated with two shots compared with one shot, mean 2.83 compared with 1.72 (p < 0.001). There was an increased improvement in score at each dose escalation level: less than or equal to 80 Gy (p = 0.017), 85 Gy (p < 0.001), and greater than or equal to 90 Gy (p < 0.001). Linear regression analysis also indicated that there was a greater response with an increased dose (p = 0.021). Patients treated with two shots were more likely to receive a higher dose (p < 0.001). There were no severe complications. Five patients developed mild facial numbness. CONCLUSIONS: Gamma knife surgery is an effective therapy for TN. Initial response rates appear to correlate with the number of shots and dose.

Determination of 137Cs dosimetry parameters according to the AAPM TG-43 formalism.

Traditional treatment planning systems calculate dose distributions around 137Cs intracavitary sources by interpolating stored dose rate tables or by Sievert-type integrals. Some of the recently introduced planning systems, such as the Varian BrachyVision and Eclipse (Varian Medical Systems, Palo Alto, CA), have discontinued the use of tables and have implemented instead the AAPM TG-43 formalism as a brachytherapy dosimetry calculation algorithm. In this work we present the dosimetry parameters for 137Cs intracavitary sources as determined according to the TG-43 formalism. With the availability of the TG-43 parameters, the commissioning of a 137Cs source in any current brachytherapy planning system is a straightforward task for a clinical physicist.

A dwell position verification method for high dose rate brachytherapy.

Misplacement of dwell positions is a potential source of misadministration in high dose rate brachytherapy. In this work we present a dwell position verification method using fluoroscopic images. A mobile C-arm fluoroscopic machine is used to take a snapshot of the treatment machine's check cable as it reaches the most distal dwell position. This fluoroscopic image is displayed side-by-side with a treatment planning image on a dual monitor relay station at the HDR treatment console. Any discrepancy between the check cable's position on the verification image and the intended dwell position on the planning image can be identified immediately, thus avoiding the possibility of treating the wrong target volume.

Evaluation of two water-equivalent phantom materials for output calibration of photon and electron beams.

Two commercially available water-equivalent solid phantom materials were evaluated for output calibration in both photon (6-15 MV) and electron (6-20 MeV) beams. The solid water 457 and virtual water materials have the same chemical composition but differ in manufacturing process and density. A Farmer-type ionization chamber was used for measuring the output of the photon beams at 5- and 10-cm depth and electron beams at maximum buildup depth in the solid phantoms and in natural water. The water-equivalency correction factor for the solid materials is defined as the ratio of the chamber reading in natural water to that in the solid at the same linear depth. For photon beams, the correction factor was found to be independent of depth and was 0.987 and 0.993 for 6- and 15-MV beams, respectively, for solid water. For virtual water, the corresponding correction factors were 0.993 and 0.998 for 6- and 15-MV beams, respectively. For electron beams, the correction factors ranged from 1.013 to 1.007 for energies of 6 to 20 MeV for both solid materials. This indicated that the water-equivalency of these materials is within +/- 1.3%, making them suitable substitutes for natural water in both photon and electron beam output measurements over a wide energy range. These correction factors are slightly larger than the manufacturers' advertised values (+/- 1.0% for solid water and +/- 0.5% for virtual water). We suggest that these corrections are large enough in most cases and should be applied in the calculation of beam outputs.

On the use of C-arm fluoroscopy for treatment planning in high dose rate brachytherapy.

Treatment planning for brachytherapy requires the acquisition of geometrical information of the implant applicator and the patient anatomy. This is typically done using a simulator or a computed tomography scanner. In this study, we present a different method by which orthogonal images from a C-arm fluoroscopic machine is used for high dose rate brachytherapy treatment planning. A typical C-arm is not isocentric, and it does not have the mechanical accuracy of a simulator. One solution is to place a reconstruction box with fiducial markers around the patient. However, with the limited clearance of the C-arm this method is very cumbersome to use, and is not suitable for all patients and implant sites. A different approach is adopted in our study. First, the C-arm movements are limited to three directions only between the two orthogonal images: the C-orbital rotation, the vertical column, and the horizontal arm directions. The amounts of the two linear movements and the geometric parameters of the C-arm orbit are used to calculate the location of the crossing point of the two beams and thus the magnification factors of the two images. Second, the fluoroscopic images from the C-arm workstation are transferred in DICOM format to the planning computer through a local area network. Distortions in the fluoroscopic images, with its major component the "pincushion" effect, are numerically removed using a software program developed in house, which employs a seven-parameter polynomial filter. The overall reconstruction accuracy using this method is found to be 2 mm. This filmless process reduces the overall time needed for treatment planning, and greatly improves the workflow for high dose rate brachytherapy procedures. Since its commissioning nearly three years ago, this system has been used extensively at our institution for endobronchial, intracavitary, and interstitial brachytherapy planning with satisfactory results.

The linear-quadratic model and fractionated stereotactic radiotherapy.

PURPOSE: To determine the dose per fraction that could be used when gamma knife or linear accelerator-based stereotactic treatments are delivered in 2 or more fractions. METHODS AND MATERIALS: The linear-quadratic (LQ) model was used to calculate the dose per fraction for a multiple-fraction regimen which is biologically equivalent to a given single-fraction treatment. The results are summarized in lookup tables. RESULTS AND CONCLUSION: The tables can be used by practicing clinicians as a guide in planning fractionated treatment. For the large doses used in typical stereotactic treatments and for small fraction numbers, the model is not very sensitive to the value of the alpha/beta ratio in the LQ model. A simple rule of thumb is found that for two-fraction and three-fraction treatments the dose per fraction is roughly two-thirds and one-half of the single-fraction treatment dose, respectively.