Lawrence T. Fitzgerald

The effect of overall treatment time on local control in patients with adenocarcinoma of the prostate treated with radiation therapy

Studies of patients treated with radiation therapy for squamous cell carcinoma of the head and neck have demonstrated that when all other variables are constant, protraction of the overall treatment time leads to a decreased probability of local control. Few data exist on the effect of overall treatment time on local control following irradiation of tumors that are generally thought to be slowly proliferating, such as adenocarcinoma of the prostate. This analysis was undertaken to determine the time-dose relationships for local control of prostatic adenocarcinoma at the University of Florida. All patients were treated at least 5 years prior to the date of analysis. For patients with Stage A2 disease, a tumor dose of 6500 cGy in 7 to 7.5 weeks to 7000 cGy in 8 weeks resulted in local control in 17/17 patients (100%). For patients with Stage B1 disease, the local control rate was 14/16 (88%) with an overall treatment time of less than or equal to 8 weeks versus 1/3 in patients who received split-course treatment in greater than 8 weeks (p = .097). For patients with Stage B2, C1, and C2 disease who received greater than or equal to 6500 cGy, the 5-year rate of local control was lower when overall treatment time was protracted beyond 8 weeks. Results were as follows: B2 (62 patients), less than or equal to 8 weeks, 88%, versus greater than 8 weeks, 55%, p = .002; C1 (87 patients), less than or equal to 8 weeks, 88%, versus greater than 8 weeks, 73%, p = .052; Cs (33 patients), less than or equal to 8 weeks, 81%, versus greater than 8 weeks, 65%, p = .056. Stratification by tumor grade of patients with Stage B1, B2, C1, and C2 disease who received greater than or equal to 6500 cGy demonstrated significantly lower local control rates for all grade categories when the overall treatment time was protracted beyond 8 weeks. Five-year local control rates (life-table method) for overall treatment time less than or equal to 8 weeks versus greater than 8 weeks were as follows: well differentiated, 93% versus 73% (p = .003); moderately differentiated, 86% versus 69% (p = .017); and poorly differentiated, 75% versus 59% (p = .046). These data suggest that tumor repopulation during excessively protracted treatment may be a clinically significant factor in patients with adenocarcinoma of the prostate.

Adenocarcinoma of the prostate treated with external-beam radiation therapy: 5-year minimum follow-up

This is a retrospective analysis of 225 patients with localized adenocarcinoma of the prostate who were treated with continuous-course external-beam radiation therapy at the University of Florida between October 1964 and August 1982. All patients were treated 5 or more years prior to the date of analysis, and 30% were eligible for 10-year follow-up. Hormonal treatment was used only in the management of recurrent disease. Ten-year results by stage were as follows: local control--A, 96%; B1, 92%; B2, 51%; C1, 57%; C2, 76%. Relapse-free survival--A, 96%; B1, 58%; B2, 38%; C1, 43%; C2, 61%. Absolute survival--A, 69%; B1, 29%; B2, 47%; C1, 35%; C2, 50%. Freedom from distant metastasis--A, 100%; B1, 55%; B2, 71%; C1, 65%; C2, 77%. Tumor grade was an important prognostic variable in most of the subgroups analyzed. The 5-year rate of distant metastasis was significantly greater in patients with stage C disease when the biopsy was made by transurethral resection of the prostate (TURP) rather than by needle biopsy. In stage B patients, the biopsy method was not prognostically important. For the 48 patients who developed recurrent tumor in the prostate gland, with or without concurrent metastasis, the 5-year absolute survival rate calculated from the date of recurrence was 26%, compared with 10% for the 34 patients who developed distant metastasis alone. Severe complications developed in 5 out of 225 patients (2%) and included three severe rectal injuries, one bladder neck contracture, and one femoral head necrosis. Moderate complications developed in 48 out of 225 patients (21%), with rectosigmoiditis (8%) and hematuria (5%) being the most common problems. For both moderate and severe complications, there was a clear trend toward an increasing complication rate with increasing dose. The method of diagnosis appeared to be a factor in the development of urinary incontinence following irradiation: needle biopsy, 0/106 (0%); TURP, 3/112 (3%); subtotal prostatectomy, 1/7 (needle biopsy versus TURP or prostatectomy, p = 0.076). The frequency of peripheral edema following irradiation was influenced by a history of surgical disruption of the pelvic lymphatics: staging lymphadenectomy, 2/16 (13%), versus no lymphadenectomy, 0/209 (p = 0.005). A comparison with other series is presented.

Cadmium Telluride Gamma Camera

A small 4 cm X 4 cm prototype CdTe gamma camera has been constructed which demonstrates sensitivity and spacial resolution superior to the standard Anger camera. The camera is based on an array of linear solid state detectors contained within a laminar collimator which confines the field of view to one dimension. The array is rotated stepwise through 180° and data from each position recorded and computer processed. This technique leads to a major enhancement in sensitivity relative to the multihole collimator. Clinical images have been obtained and a larger CdTe camera is planned.

Rotating laminar emission camera with Ge-detector: further developments

A germanium (Ge) camera with laminar collimator-detector geometry, ROLEC, simulating a 195 X 195 mm2 detection area, has been constructed and tested. The detector consists of five separate Ge blocks grooved into distinct electrical channels. Results show that ROLEC measuring times have to be about 3 X greater than Anger camera measuring times to achieve a 5% signal-to-noise ratio with each device. Both spatial and energy resolution of the ROLEC exceed that of the Anger camera but sensitivity variations along the detector strips pose a significant problem in development of a clinically useful device.

Compton scatter axial tomography with x-rays: SCAT-CAT.

A method of extracting information from the backscattered field produced in parallel beam x-ray computed tomography (CT) is presented. A calculational model to predict the backscattered field based on Compton scattering is described, and the model is verified by measurements of simple phantoms. The phantoms tested--cylinders of polymethylmethacrylate (PMM) with air gaps and aluminium rods placed internally--are irradiated on a scanning assembly, built to simulate a first generation CT scanner with a transmission and a scatter detector (the SCAT-CAT). Data from the transmission detector are reconstructed by traditional CT methods to provide a transmission image; it is the data from the backscatter detector which are analysed in this study. After verification of the model for the scattered field calculations, a method of extracting information from the scattered field is developed, based on ratios of scatter signals from non-uniform to uniform phantoms. This method is demonstrated for predicted data of a simulated phantom and for measured data of the same and two additional phantoms. The method is very sensitive to air gaps in the phantoms because of the relative electron density of air with respect to PMM; it is not as sensitive to aluminium rods for the same reason. Various methods of applying the scattered field information to produce an image representing a simulated phantom are considered, and a preferred method is chosen to reconstruct scattered field data into an image for the three phantoms studied.

A Germanium Laminar Emission Camera

A 30 elenent laminar Collimator with a segmented high purity Ge detector gave spatial resolutions to 1.5 mm for 2 dimensional Tc99m imaging. Detector, collimator and imaging aspects are discussed.

Rotating laminar emission camera with GE‐Detector: An analysis

Design criteria for a germanium detector emission camera with laminar collimator-detector geometry are reported. In this design, parallel plates above a grooved germanium slab form detector channels which accumulate data in multiple projections through 180 degrees. Activity distribution is determined through mathematical reconstruction from the projections. The laminar design has greater efficiency than a hole-collimated device due to a favorable open area ratio and due to collimation in only one dimension. The relative sensitivity at the center of the detector strip increases with detector length and distance from collimator face. Spatial resolution depends upon the design geometry (septal penetration and scatter within the crystal are of such small magnitude that degradation of the point spread function is minimal). A 30 channel protype device (ROLEC) shows good correlation with theoretically determined PSF and signal-to-noise ratio (SNR). Measuring times projected for a clinical sized device (300 mm diam) are about 4.6 times that of the Anger camera (with same SNR) and high resolution collimator.

Rotating laminar emission camera with GE-detector: experimental results.

Experimental results of a prototype rotating laminar emission camera (ROLEC) for nuclear medicine imaging are reported. A 11.5 mm thick, 45 x 45 mm high-purity germanium detector is segmented into 30 1.47 mm wide parallel channels and collimated with 39 mm high parallel plates. Projection data acquired at multiple angular orientations as the detector-collimator assembly is rotated about its center are mathematically reconstructed to image the activity distribution. The spatial resolution of the ROLEC is at least twice as good, at all distances, as that of gamma cameras with high resolution collimators. The better energy resolution of the germanium enhances the detection and resolution of the ROLEC in comparison to gamma cameras with NaI(T1) crystals, the relative superiority increasing with greater volumes and with greater depths. Adequate sensitivity is maintained while achieving these improvements in spatial resolution and in practice; ROLEC images are acquired in less time than pinhole collimator images with gamma cameras.

Analysis of errors in three-dimensional reconstruction of radium implants from stereo radiographs.

The present method of steroe shift for three-dimensional (3-D) reconstruction of radium implants has been found to produce displacement errors. Analysis shows that the maximum displacement error decreases as the shift (S) increases with fixed focal spot-film distance (H), while conversely for a fixed shift distance the maximum displacement error with increased focal spot-film distance. The authors suggest that sufficiently accurate 3-D reconstruction may be achieved when H greater than or equal to 100 cm and H/S less than 2.

A coded aperture device for on-line imaging with megavoltage photon beams

A prototype coded aperture device based upon a liquid-filled strip ion chamber has been investigated as an on-line imaging unit for portal imaging of megavoltage photon beams. The strip ion chamber was used to collect integrated data for a series of pixels along a line across the field of view. By translating the detector, in equal steps in a direction perpendicular to its length, the data for a complete projection of the image were collected. Multiple-projection data, corresponding to different angular views, were collected by sequentially rotating the detector in equal angular steps and then translating through the field. The portal image was reconstructed from the multiple-projection data by a convolution-filtered back-projection algorithm. The phantom images taken with a cobalt-60 radiation source were found to be of similar contrast to film images. The signal-to-noise ratio, including the noise propagation of the reconstruction process for an image of 101 ? 101 pixels, was 207:1. For a translational step of 2 mm the spatial resolution of the system was found to be 4.2 mm, as determined from the full width at half maximum of the point spread function.