Mark H. Phillips

Heavy-Charged-Particle Radiosurgery of the Pituitary Gland: Clinical Results of 840 Patients

Since 1954, 840 patients have been treated at Lawrence Berkeley Laboratory with stereotactic charged-particle radiosurgery of the pituitary gland. The initial 30 patients were treated with proton beams; the subsequent 810 patients were treated with helium ion beams. In the great majority of the 475 patients treated for pituitary tumors, marked and sustained biochemical and clinical improvement was observed. Variable degrees of hypopituitarism developed in about one-third of patients treated solely with radiosurgery. In the earlier years of the program, 365 patients underwent radiosurgery to treat selected systemic diseases by inducing hypopituitarism. Focal temporal lobe necrosis and cranial nerve injury occurred in about 1% of patients who were treated with doses less than 230 Gy.

Comparison of different radiation types and irradiation geometries in stereotactic radiosurgery

Recent interest in stereotactic radiosurgery of intracranial lesions, and the development of stereotactic irradiation techniques has led to the need for a systematic and complete comparison of these methods. A method for conducting these comparisons is proposed and is applied to a set of currently-used stereotactic radiosurgical techniques. Three-dimensional treatment planning calculations are used to compare dose distributions for several different radiation types and irradiation geometries. Calculations were performed using charged particles (H, He, C, and Ne ions) and the irradiation geometry currently used at Lawrence Berkeley Laboratory. Photons in the Gamma Knife configuration and the Heidelberg Linac arc method are used. The 3-dimensional dose distributions were evaluated by means of dose-volume histograms and integral doses to the target volume and to normal brain. The effects of target volume, shape and location are studied. The charged particle dose distributions are more favorable than those of the photon methods. The differences between charged particles and photons increase with increasing target volume. The differences between different charged particle species are small, as are the effects of target shape and location.

Stereotactic frame for neuroradiology and -charged particle bragg peak radiosurgery of intracranial disorders

The application of heavy charged particle Bragg peak radiosurgery for the treatment of intracranial vascular and other disorders requires a system of precise patient immobilization and stereotactic localization of defined intracranial targets. The process of using stereotactic neuroradiological procedures (including cerebral angiography, CT scanning and magnetic resonance imaging) for target definition and localization, and complex treatment planning constrain such a system to be adaptable and reusable. This paper describes a removable stereotactic frame-mask system that is used to immobilize and reposition the patient during stereotactic neuroradiological procedures and charged particle radiosurgery. It consists of four parts--(a) a plastic mask for immobilizing the patients head; (b) a lucite-graphite mounting frame; (c) a set of fiducial markers; and (d) interfaces between the frame for immobilization and fixation to various diagnostic and therapeutic patient couches. The relationship between each component and the radiosurgical procedure is discussed. This system has proven to be safe, reliable, and noninvasive and it does not require fixation to the bones of the face or skull. When integrated into the radiosurgical treatment planning and localization procedures developed at Lawrence Berkeley Laboratory, it is capable of reliably repositioning the patient to 1 mm in each of three planes and contouring the intracranial target reliably to this accuracy. The application of this stereotactic system in heavy charged particle radiosurgery of intracranial arteriovenous malformations is described in other reports.

MRI and PET of delayed heavy-ion radiation injury in the rabbit brain

Magnetic resonance imaging (MRI) and positron emission tomography (PET) techniques were used to obtain in vivo scans of delayed (30 GyE helium ion, 230 MeV/u) radiation injury in rabbit brain. T2-weighted (T2W) MRI scans demonstrated alterations that were restricted primarily to the white matter tracts and the deep perithalamic and thalamic regions. Quantitative measurements of T2 and T1 values demonstrated wide variations in absolute values. However, paired comparisons in hemibrain-irradiated rabbits revealed significant increases in T2 (p less than 0.001) and T1 (p less than 0.01) in irradiated versus unirradiated brain. Gadolinium DTPA (GdDTPA) enhanced MRI and 82Rubidium (82Rb) PET detected focal regions of blood-brain barrier (BBB) disruption restricted to the deep white matter and thalamic regions. Sequential GdDTPA enhanced MRI scans showed the spreading of the tracer from the initial site of contrast enhancement. 18Fluorodeoxyglucose (18FDG) PET studies demonstrated the markedly depressed metabolic profiles of irradiated brain. Histological findings of tissue edema and necrosis correlated well with the in vivo imaging abnormalities. These initial studies demonstrate that the irradiated rabbit brain is a suitable animal model for examining the delayed effects of radiation injury in the brain.

Stereotactic helium ion bragg peak radiosurgery for angiographically occult intracranial vascular malformations

Between July 1983 and July 1989, we treated 35 patients with surgically inaccessible, symptomatic angiographically occult vascular malformations (AOVMs) using stereotactic heavy-charged-particle radiosurgery. AOVMs were located in the brainstem (19), thalamus or internal capsule (9), basal ganglia (3), deep cerebral hemisphere and motor area (3), or cerebellopontine angle (1). All patients presented with clinical and radiological evidence of previous hemorrhage, usually with multiple episodes of hemorrhage. Treatment volumes ranged from 80 to 15,200 mm3 and treatment doses from 7.7 to 34.6 Gy. Mean follow-up was 40 months, with 31 patients followed for at least 2 years. Clinical outcome was excellent in 46%, good in 34% and poor in 14%; 6% died. Twenty-seven patients in excellent and good condition prior to treatment remained stable or improved neurologically. Two patients initially in poor condition, who had previously received conventional radiotherapy, died at 9 and 14 months after treatment, respectively. Six patients experienced recurrent hemorrhage 2-60 months following treatment. Three of these patients made a complete recovery. Although a larger number of treated patients must be followed over longer periods of time, stereotactic heavy-particle radiotherapy may be a valuable treatment modality for surgically inaccessible intracranial AOVMs.

Stereotactic helium ion bragg peak radiosurgery for intracranial arteriovenous malformations : detailed clinical and neuroradiologic outcome

89 patients with angiographically documented arteriovenous malformations were treated with helium ion Bragg peak radiation. The rate of complete angiographic obliteration 2 years after radiation was 94% in those lesions smaller than 4 cm3 (2.0 cm in diameter), 75% for those 4-25 cm3 and 39% for those larger than 25 cm3 (3.7 cm in diameter); at 3 years after radiation, the corresponding obliteration rates were 100, 95 and 70%. Major clinical complications occurred in 10 patients (8 permanent, 2 transient) between 3 and 21 months after treatment; all were in the initial stage of the protocol (higher radiation doses). 10 patients bled from residual malformation between 4 and 34 months after treatment. Seizures were improved in 63% and headaches in 68% of patients. Excellent or good clinical outcome was achieved in 94% of patients. Compared to the natural history and risks of surgery for these difficult malformations, we consider these results encouraging. Heavy-charged-particle radiation is a valuable therapy for surgically inaccessible symptomatic cerebral arteriovenous malformations. The current procedure has two disadvantages: the prolonged latent period before complete obliteration and the small risk of serious neurological complications.

The multiple Coulomb scattering of very heavy charged particles

An experiment was performed at the Lawrence Berkeley Laboratory BEVALAC to measure the multiple Coulomb scattering of 650-MeV/A uranium nuclei in 0.19 radiation lengths of a Cu target. Differential distributions in the projected multiple scattering angle were measured in the vertical and horizontal planes using silicon position-sensitive detectors to determine particle trajectories before and after target scattering. The results were compared with the multiple Coulomb scattering theories of Fermi and Molière, and with a modification of the Fermi theory, using a Monte Carlo simulation. These theories were in excellent agreement with experiment at the 2 sigma level. The best quantitative agreement is obtained with the Gaussian distribution predicted by the modified Fermi theory.

Heavy charged-particle stereotactic radiosurgery: Cerebral angiography and CT in the treatment of intracranial vascular malformations

A method is described for stereotactic localization of intracranial arteriovenous malformations (AVM) and for calculating treatment plans for heavy charged-particle Bragg peak radiosurgery. A stereotactic frame and head immobilization system is used to correlate the images of multivessel cerebral angiography and computed tomography. The AVM is imaged by angiography, and the frame provides the stereotactic coordinates for transfer of this target to CT images for the calculation of treatment plans. The CT data are used to calculate the residual ranges and compensation for the charged-particle beam required for each treatment port. Three-dimensional coordinates for the patient positioner are calculated, and stereotactic radiosurgery is performed. Verification of the accuracy of the stereotactic positioning is obtained with computer-generated overlays of the vascular malformation, stereotactic fiducial markers, and bony landmarks on orthogonal radiographs immediately prior to treatment. Using these procedures, the accuracy of the repositioning of the patient at each of a series of imaging and treatment procedures is typically within 1 mm in each of three orthogonal planes.

Cerebrovascular and metabolic perturbations in delayed heavy charged particle radiation injury.

Focal heavy charged particle irradiation of the rabbit brain created defined lesions which were observable by nuclear magnetic resonance (NMR) and positron emission tomography (PET) imaging techniques. The lesions appeared approximately 9-11 months after left partial hemibrain irradiation with 30 Gy (230 MeV/u helium ions), and were restricted to the white matter tracts and deep perithalamic and thalamic regions. 82Rubidium PET and Gadolinium DTPA enhanced NMR imaging were used to detect blood-brain barrier perturbations. 18Fluordeoxyglucose PET studies demonstrated widespread decreases in cerebral glucose uptake in the cortex and thalamus of the irradiated hemisphere. NMR and PET imaging results correlated well with histological findings. Rabbits irradiated with 15 Gy did not demonstrate any abnormalities in the brain with sequential NMR scans through 14 months post-irradiation.

Heavy-Charged-Particle Radiosurgery for Intracranial Arteriovenous Malformations

We have treated over 400 patients with symptomatic inoperable intracranial arteriovenous malformations (AVMs) with stereotactic heavy-charged-particle Bragg peak radiosurgery at the University of California at Berkeley in a collaborative program with Stanford University Medical Center and the University of California Medical Center, San Francisco. A long-term dose-searching clinical trial protocol has been developed and we have followed more than 250 patients for more than 2 years. Initially, radiation doses ranged from 45 to 35 GyE, and now doses of 25, 20, 15 and, under special circumstances, 10 GyE, depending on a number of factors, are being evaluated. The characteristics of charged-particle beams provide a relatively homogeneous dose distribution with the 90% isodose contour to the periphery of the lesion. When the entire arterial phase of the AVM core is included in the treatment field, the rates for complete obliteration 3 years after treatment are: 90-95% for volumes less than or equal to 4 cm3; 90-95% for volumes greater than 4 and less than or equal to 14 cm3; and 60-70% for volumes greater than 14 cm3. The total obliteration rate for all volumes up to 70 cm3 is approximately 80-85%. For complete radiation-induced obliteration there is a relationship of dose and volume primarily, and location secondarily. Results on relationships between dose, AVM obliteration, and complications and sequelae of the radiosurgical procedure are presented and discussed.