The development of fractionated stereotactic radiotherapy (SRT) has been hampered by slow and cumbersome relocation techniques. The Medtronic Sofamor Danek (MSD) frameless system is promoted as a fast, relocatable system. It relies upon the integrity of a bite tray system that is molded to the patient's upper dentition. Precise relocation of the bite tray is the key to accurate external placement of the fiducial reference frame. The optimum construction method for a stable reproducible tray is not described. We undertook a study to identify factors, which might influence the integrity of the bite tray system. Reprosil Fast Set Putty was used to construct three bite tray conditions to include: teeth only, teeth and alveolar sulcus, and teeth, alveolar sulcus, and the hard palate. The influence of dentition was also assessed: full dentition, partial dentition, and edentulous. Stability of the bite tray system was tested by assessing minimum displacement forces from known locations around a centrally placed maxilla jig. Dentition is important to stability, with progressively less displacement forces required for partial and edentulous conditions. Extension of the surface area of the bite tray to include teeth, gums, and also hard palate substantially increased stability across all dental conditions. Impression techniques that incorporate the teeth, alveolar sulcus, and the hard palate, result in significantly increased stability and are recommended to ensure maximal resistance to deviation.
Approximately 20% of patients with subfoveal choroidal neovascularization qualify for laser photocoagulation, a treatment associated with both disease control and reduced visual acuity. No proved treatment exists for other patients. Single-fraction proton therapy was investigated as an alternative for two groups of patients: 21 patients received 8 GyE; 27 received 14 GyE. Follow-up by fluorescein angiography, visual acuity, contrast sensitivity, and reading speed were done (mean duration: 16 months). Actuarial lesion control at 21 months was 36% for 8-GyE patients and 89% for 14-GyE patients; 77% of patients with controlled lesions achieved improved/stable visual acuity, compared to 44% with uncontrolled lesions. Actuarial mean visual loss for proton-treated maculas was zero at 24 months. No treatment-related morbidity supervened, based on Radiation Therapy Oncology Group criteria. Proton therapy of 14 GyE in one fraction appears to be more effective in controlling neovascular macular degeneration than 8 GyE in one fraction, for both laser-ineligible and -eligible patients.
It is recognized that multiple-shot radiosurgery using the Leksell Gamma Knife (LGK) can result in larger penumbra dose and less target dose homogeneity compared with a single shot treatment. In addition, the number of shots used increases significantly with increasing volume, which makes planning and optimization time consuming and complex. Although such complexity may not affect the accuracy of treatment delivery itself, it is difficult to assess the dosimetrical merits of a given plan. To deal with this complicated process, we have studied shot-related effects and target shape as separate issues. Shot-related effects include homogeneity, penumbra, shot weight, and prescription isodose line “shift.” Target shape is analyzed using morphological tools. Target volume and shape, as represented by the three-dimensional (3-D) medial axis transform or skeleton, are used to determine optimal shot positions, sizes, and number. These parameters are determined from successive 3-D skeletons of the unplanned target volume. Essentially, target shape is replicated by the arrangement of shots compromising the 3-D skeleton, forcing a conformal dose distribution. This planning process has been automated for multishot treatment. A clinical case in 3-D demonstrates this approach as a replacement for manual planning when a plan with multiple shots is required.
This study evaluates prognostic factors influencing survival outcomes for 50 patients with permanent125 iodine-125 implants in the primary treatment of non-GBM high-grade gliomas. Stereotactic treatment planning aimed to encompass the contrast-enhancing rim of the tumor visualized by CT, with an initial dose rate of 0.05 Gy/hour with 125I, delivering 100 Gy at 1 year and 103.68 Gy at infinity. Survival was evaluated using the Kaplan–Meier method for unvariate analysis and the Cox regressional method for multivariate analysis. In addition to the implant, 31 patients received external radiation therapy (5000 to 6000 cGy) before the implant; 10 patients were implanted without additional external beam radiation, and 9 patients underwent external radiation therapy before implant placement. With a mean follow-up of 40.76 months (range 3.47–87 months); 1−, 3−, and 5-year survival were 78.5% (± .05%), 58.7% (± .07%), and 56.2% (± .07%) respectively. Since 56.2% of the patients were alive at 5 years, median survival has not been reached yet. Second surgery was performed following the implant in 19 patients. Findings were tumor recurrence in 11 patients (22.5%), radiation necrosis in 7 patients (14.3%), and brain abcess in 1 patient (2%). Age, sex, tumor location, side of brain, tumor volume, Karnofsky, and neurological status were correlated with survival outcome. Favorable prognostic factors were age younger than 45 years, superficial tumor location, and preoperative Karnofsky greater than 70. Surgical treatment of patients with non-GBM high grade gliomas combined with external beam radiation and permanent 125I implants represent a valuable alternative for the treatment of patients with malignant gliomas, allowing patients good quality of life and long survival.
We report a retrospective study on the use of the permanent iodine-125 (125I) implants in the management of low-grade gliomas. From July 1988 to July 1997, 16 patients with low-grade gliomas underwent permanent 125I implants in the management of their lesions. There were 7 males and 9 females ranging in age from 4 to 48 years (mean 19). The location was in the cerebral hemisphere in 7 patients, brainstem in 5 patients and thalamus/basal ganglia in 4 patients. Prior to brachytherapy, 9 patients underwent surgical resection and 7 patients underwent stereotactic biopsy procedures. Fourteen patients were treated as part of the initial management and 2 were recurrent. The histological diagnosis was: 9 WHO grade II astrocytomas, 3 oligodendrogliomas, 2 gemistocytic astrocytomas, 1 pilocytic astrocytoma, and 1 ependymoma. The tumor volume ranged from 0.7 to 33.4 cc (mean 8.4). Stereotactic treatment planning was used to encompass the contrast-enhancing rim of the tumor visualized by computerized tomography with an initial dose rate of 0.05 Gy/hour with 125I. The total activity ranged from 0.8 to 20.5 mCi. With a median follow-up period of 35 months (range, 4–105 months), the 2- and 5-year survival rates were 93.7% and 87.5%, respectively. Three patients underwent reoperation after implants, two of three had recurrent disease, and one had radiation necrosis. Permanent 125I implants appear to be safe and effective as a part of the multimodality management of low-grade gliomas.
A rare case of isolated liver metastases in a patient with recurrent glioblastoma multiforme (GBM) treated with surgery, chemotherapy, external beam irradiation, and salvage stereotactic radiosurgery is described. Our review of literature has revealed only 8 other cases of isolated liver metastases in patients with GBM in the last three decades. Recent data suggest that stereotactic radiosurgery may alter the prognosis of specific subsets of patients with GBM. The potential for combined modality therapy, including radiosurgery to impact on the survival of GBM patients, and, hence, the natural history of this disease, is discussed.
The authors report a case of a 69-year-old man with metastatic brain tumors who died of spontaneous intracerebral hemorrhage 3 days after γ-knife surgery. He had been suffering from lung cancer with multiple systemic metastasis. Preoperative magnetic resonance images showed two well-defined round lesions with intratumoral hemorrhage in the left frontal and right occipital lobe. There was no bleeding tendency in the hematological examination and the patient was normotensive. γ-Knife surgery was performed on both lesions in a single session. However, the patient died of massive intracerebral hemorrhage from the left frontal lesion 3 days after the surgery. There have been no previous reports of mortality resulting from spontaneous intracerebral hemorrhage after γ-knife surgery in metastatic brain tumors documented in the literature. It is likely that the two events, γ-knife surgery and spontaneous intracerebral hemorrhage, occurred separately and were not associated. However, it is worth noting that there is a possibility of bleeding after γ-knife surgery, especially in a metastatic brain tumor with preexisting intratumoral hemorrhage as in our case.
Dosimetry of the small radiation fields used in radiosurgery is often difficult because of finite detector size and loss of lateral electronic equilibrium. However, small-field dosimetry is critical in radiosurgery where a relatively high dose in a single fraction is often delivered. Radiosurgery dosimetry varies significantly in small fields depending upon the choice of detector. Small-volume (0.015 cm 3 –0.125 cm 3 ) cylindrical ion chambers, parallel-plate ion chamber, films (Kodak and CEA), thermoluminescent dosimeter (TLD), diamond detector, and Monte Carlo simulation were used to study the small-field (≤ 4 cm) dosimetry of a radiosurgery unit, which consists of Radionics hardware with cone sizes of 12.5 mm to 40 mm with a 6-MV linear accelerator. Results indicate that, in general, ion chambers and TLDs are not suitable for beam profiles. The diamond detector provides better resolution when faced parallel, rather than the perpendicular, to the beam axis. The dose profiles measured with laser-film densitometer are found to be moderately sensitive to the aperture sizes (200 μm–470 μm). The profiles generated using film and diamond detectors are nearly identical. The cone factors measured using film, diamond detectors, and TLD varies (+ 5%) depending upon the uncertainty of the measurements; however, the measured cone factors differ significantly (≤ 10%) among various ion chambers. The Pinpoint (0.015 cm 3 ) ion chamber provides reliable data, which differs from other detectors. Monte Carlo data are in between all other measured data. The large variation in SRS dosimetry suggests that there should be a national protocol to review the dosimetry in small fields.
Some lesions situated in sensitive areas of the brain are potentially dangerous to treat with any modality. This study reviews one institution's experience with the use of stereotactic radiosurgery (SRS) for brainstem and thalamic metastases to evaluate its efficacy in treating such lesions. Between October 1989 and January 1998, 20 patients (9 men, 11 women) underwent linear accelerator SRS for metastases in the brainstem or thalamus. A retrospective chart and radiographic analysis was performed on these patients. The mean patient age at the time of SRS was 55.9 years (range 34–76). The median dose of SRS was 1600 cGy (range 1200–2000) to the 80% isodose line, although the isodose line varied somewhat. Of the 20 lesions, median patient survival from the time of SRS was 27.2 weeks (mean 39.7, range 5.4–216). Reliable evaluation of neurological status after SRS was obtainable in 17 patients and radiographical follow-up in 12. The rate of clinical control of symptoms was 88.2% (15/17) and radiographical control was 100% (12/12). Complications occurred in two patients (11.8 %). Our results indicate outcomes that are similar to those in a previous report of SRS for metastatic lesions of the brainstem and thalamus, with survival shorter than that for other cerebral lesions. SRS for metastases of the brainstem and thalamus is a safe and effective treatment option, but does not offer as favorable an outcome as SRS for lesions in other areas of the brain.
We developed a C-arm multi-axis, stereotactic linear accelerator radiosurgical system combined with 3-dimensional (3D) dose planning software. The linac beam generator is mounted on a C-arm frame that allows rotation parallel and diagonal to the patient table. As a result of the new structural design, the positional deviation of the beam from mechanical isocenter during rotation of the gantry is almost less than 0.5 mm. The system can provide 3D stereotactic radiosurgery using precessional convergent irradiation (PCI) without movement of the patient table, which improves accuracy of the irradiation. A satisfactory dose gradient suitable for radiosurgery with fewer arcs was achieved by PCI than was possible with noncoplanar multi-converging arcs (MCA). The 3D dose planning system can accurately optimize the dose plan even for an irregularly shaped target. We have treated 60 cases since 1996 with a reduction of total treatment time.
This retrospective analysis details the experience of a tertiary care center with survival and results for patients with recurrent glioblastoma multiforme (GBM) treated with stereotactic radiosurgery (SRS). Between August 1990 and June 1999, 23 patients were treated for recurrent GBM with SRS using either modified 6-MV linear accelerator (linac) or γ-knife. Twenty-two patients (96%) had an initial histological diagnosis of GBM, while 1 patient had an initial diagnosis of anaplastic astrocytoma that was biopsied at recurrence and found to have upgraded to GBM. The median Karnofsky performance score at the time of SRS was 80; the median age was 53. The median-treated tumor volume was 9.9 cm 3 , with a dose of 15 Gy delivered to the 60% isodose line. Median progression-free survival was 4.7 months. Median survival time after SRS was 10.3 months. No prognostic factors were found to be significant. Patients with a KPS of 80 or higher had longer median survival times than those with lower KPS scores, but this was not found to be statistically significant. Our results suggest that for selected patients with recurrent GBM, SRS appears to be an appropriate salvage therapy.
This paper describes the use of a MD-55 radiochromic film to measure proton doses and dose distributions in a homogeneous lucite phantom. Radiosurgery dose distributions were measured for the plateau and for the Bragg peak techniques with narrow proton beams. Measured distributions were compared to the dose distributions calculated with a 3-D image-based treatment planning system developed at Loma Linda University Medical Center, California and modified to accommodate to treatment delivery apparatus used at the Institute of Theoretical and Experimental Physics, Moscow, Russia. The results of this comparison indicate agreement within the accuracy of the radiochromic film technique.
The risk of brain recurrence after whole-brain radiotherapy for metastatic disease has been infrequently reported, but quoted to be as high as 50%. These values are in the context of autopsy series or clinical trials and thus may not represent experience in routine clinical practice. This retrospective analysis was undertaken: (1) to confirm recursive analysis-based prognostic factors contributing to survival; and (2) to investigate efficacy of Gamma Knife Stereotactic Radiosurgery (GKSRS) in the poorest prognosis population. The records of 85 unselected, consecutive patients receiving whole-brain radiotherapy for brain metastases at Naval Medical Center San Diego were analyzed. Median dose of whole brain radiotherapy was 30 Gy in 10 fractions. These records were categorized by Radiation Therapy Oncology Group (RTOG) categories based on several prognostic factors. Further, analysis of 153 consecutive patients (178 treatments of 396 lesions) receiving GKSRS at the San Diego Gamma Knife Center was made for patients with Karnofsky Performance Status (KPS) < 70 (RTOG class 3). Twenty patients were identified. Median dose of GKSRS was 17.5 Gy at the 61% mean isodose line for all GKSRS patients. Six patients experienced clinical brain recurrences after whole-brain radiotherapy; the actuarial risk of clinical brain recurrence after whole-brain radio therapy approached 30% at 3 years, although few patients survived this long. Survival of patients after whole brain radiotherapy alone by RTOG category mirrored findings of other centers. Patients in class 3 ( n = 26) had a median survival in our population of 59 days. For the 20 GKSRS patients with similar performance status and clinical follow-up, the median survival was longer, but not significantly so (162 days; p = 0.467). In the off-protocol setting, clinical brain recurrence of metastatic disease after whole-brain radiotherapy occurs not infrequently in patients with otherwise long survival. The RTOG criteria well describes an extremely poor-prognosis group of patients for whom the single fraction of stereotactic radiosurgery may optimize palliation as well as or better than longer courses of conventional palliative radiotherapy.
The use of radiosurgery in treatment of the glaucomas is a completely new way of treating advanced stages of this treacherous disease. Up to now, the most common indications for the LGK therapy in ophthalmology at our department are malignant melanoma, vascular diseases, intraocular and intraorbital metastasis, and glioma of the optic nerve. Our pilot study shows that focal irradiation of the ciliary body in advanced glaucoma can ameliorate the main symptoms and thus prevent the otherwise inevitable enucleation of the eye. We treated in 12 patients, 13 very painful eyes, who were blind or with very diminished vision, and where the conventional treatment was ineffective. The ocular pain was alleviated in all patients and their intraocular pressure was reduced or normalized. There were no early side effects. Further studies are needed to elucidate the best treatment parameters and the involved pathophysiological mechanisms. However, it is evident that gamma knife radiosurgery can be considerably extended in this way.
Background: The objective of this retrospective study is to determine the value of radiosurgery in the management of arteriovenous malformations (AVM) in the pediatric age group. Methods: From January 1994 through January 1999, thirty-one children with arteriovenous malformations (AVMs) were treated with radiosurgery. All patients were treated on an outpatient basis at the same institution by the same team. The Leksell Gamma Knife unit was used. Workup included angiography, MRI, and MRA. Follow-up ranged from 7 months to 67 months, with a median of 33 months. Minimum doses of radiation, depending on the size of the lesion, ranged from 20 Gy to 25 Gy. Treatment volumes for all the vascular malformations ranged from 0.6 cc to 17 cc with a mean volume of 4.7 cc. The mean number of isocenters was 4.8. Results: Total obliteration of AVM nidus was obtained in 22 of 31 (71%) patients, while 9 patients had partial obliteration. Stabilization of the benign lesions was obtained in all the patients treated. None had rebleeding after the procedure and, as of this writing, no patient required retreatment. Conclusion: Radiosurgery is an effective noninvasive and safe therapeutic modality for the management of vascular malformations independent of location, size, or grade.
Brain metastases from hepatocellular carcinoma are rare. Limited data from case reports indicates that these tumors may be relatively resistant to conventional whole brain radiotherapy. We report a case of hepatocellular carcinoma metastatic to the brain treated with stereotactic radiosurgery. This patient's tumors showed an excellent response to this modality and were successfully controlled for 13 months when she died of disseminated lung metastases.
We have analyzed a group of 163 patients with pituitary adenoma treated by the GK for 5 years. They were followed up for 12–60 months (median 24 months) after irradiation. An antiproliferative effect has been achieved using the minimal dose 16–35 Gy with median of 20 Gy to the tumor margin in all of our patients who were controlled by MRI (n = 126 patients). One-half of these adenomas noticeably decreased in size. This antiproliferative dose was safe for the surrounding structures. The effect was not dependent on the gender or age of the patient, on previous fractionated radiotherapy, or on the applied dose. A significant dependence was found on the invasive growth of adenoma, on the use of a modern planning system (GammaPlan) in relation to the older Kula system, and on previous microsurgery. The hormonal normalization has been achieved in 50.4% from 133 hypersecreting adenomas (39/91 = 43% of acromegalics, 11/13 = 85% of patients with Cushing's disease, 2/9 = 22% of patients with Nelson's syndrome, 11/18 = 61% of prolactinomas). The median latency of the effect was 12 months. The minimal dose to the margin of hypersecreting adenomas was 10--45 Gy, median 35 Gy. The hormonal normalization was not dependent on the gender, on previous surgical resection, fractionated radiotherapy, or on the volume of adenoma. The dependence of the treatment result could be proved in relation to the age, the invasive growth of the tumor, the level of its hypersecretion, the size of adenoma, the radiation dose to the tumor, and the use of a modern planning system (GammaPlan). Rare side effects were provoked only by increasing the dose to influence the hypersecretion. This led to partial or complete hypopituitarism in 3.7% of patients and left one patient with homonymous hemianopia (0.6%). A longer follow-up should disclose the possible late effects of radiation and its potency of preventing adenoma recurrences. Radiosurgery by GK has the same value for pituitary adenomas as other treatment procedures: microsurgery, pharmacotherapy and fractionated radiotherapy--therefore, it can contribute to optimizing the treatment results by combining several of the available treatment methods.
When the size of an arteriovenous malformation (AVM) in any one direction is more than 4 cm, an attempt is made to treat it in two stages, separated by 6 months. The first stage treats one-half of the AVM and the remainder is treated in the second stage, thus giving rise to a higher prescription dose for each stage. Therefore, the AVM, as a whole, could be given a higher dose than if the entire AVM were treated in a single fraction. In order to reproduce, in each stage, the isocenter coordinates determined from a pair of orthogonal films, two techniques are applied. One is the use of BrainLAB's mask system, which maintains the isocenter accuracy within ± 1 mm between the two stages. The other is the use of four 3-mm long titanium screws (fiducial markers) embedded in the patient's skull. The coordinate transformation matrix established for the fiducial markers between the first- and second-treatment stages can be applied to the respective isocenter coordinates. Thus, the original isocenter can be reestablished in the coordinate system of the second-treatment stage. Application of this method was first tested with four 5-mm diameter aluminum balls (BBs) attached as fiducial markers to the external surface of the Rando head phantom and three 3-mm tungsten balls embedded as isocenters inside the skull. One patient case was also studied in terms of predicted accuracy, since the isocenters cannot be pinpointed in terms of anatomical structure (like the three tungsten balls). The two-staged radiosurgical approach with small screws embedded in the skull maintains the accuracy required for stereotactic radiosurgery, thereby facilitating the treatment of large AVMs. Since rigid fixation of the head, as is used with traditional stereotactic radiosurgery, is not used with this two-staged approach, treatment does not have to be delivered shortly after diagnostic images are acquired for treatment planning. This gives the physicist, radiation oncologist, and neurosurgeon additional time to optimize the treatment plan, if necessary. By dividing the treatment into two stages, it is hoped that a higher dose of radiation can be safely delivered to large AVMs, thereby increasing the likelihood of cure.
