Irradiation of the prostate and pelvic lymph nodes with an adaptive algorithm.

PURPOSE: The simultaneous treatment of pelvic lymph nodes and the prostate in radiotherapy for prostate cancer is complicated by the independent motion of these two target volumes. In this work, the authors study a method to adapt intensity modulated radiation therapy (IMRT) treatment plans so as to compensate for this motion by adaptively morphing the multileaf collimator apertures and adjusting the segment weights. METHODS: The study used CT images, tumor volumes, and normal tissue contours from patients treated in our institution. An IMRT treatment plan was then created using direct aperture optimization to deliver 45 Gy to the pelvic lymph nodes and 50 Gy to the prostate and seminal vesicles. The prostate target volume was then shifted in either the anterior-posterior direction or in the superior-inferior direction. The treatment plan was adapted by adjusting the aperture shapes with or without re-optimizing the segment weighting. The dose to the target volumes was then determined for the adapted plan. RESULTS: Without compensation for prostate motion, 1 cm shifts of the prostate resulted in an average decrease of 14% in D-95%. If the isocenter is simply shifted to match the prostate motion, the prostate receives the correct dose but the pelvic lymph nodes are underdosed by 14% ± 6%. The use of adaptive morphing (with or without segment weight optimization) reduces the average change in D-95% to less than 5% for both the pelvic lymph nodes and the prostate. CONCLUSIONS: Adaptive morphing with and without segment weight optimization can be used to compensate for the independent motion of the prostate and lymph nodes when combined with daily imaging or other methods to track the prostate motion. This method allows the delivery of the correct dose to both the prostate and lymph nodes with only small changes to the dose delivered to the target volumes.

Collagen quantitation by video-microdensitometry in rabbit atherosclerosis.

A video-microdensitometry method was developed to measure the mass fraction of collagen in portions of tissue sections. A frame grabber and computer are used to capture micrographs and calculate the collagen fraction at each pixel. The method is based on the stoichiometric staining of collagen by picrosirius red stain, and was calibrated for glutaraldehyde fixed tissue; a detailed map of collagen distribution throughout a six-month rabbit atherosclerotic lesion was made. The method is reliable, fast and easily reproduced, since it uses readily available equipment.

The role of taper on the distribution of atherosclerosis in the human infra-renal aorta.

The role of vessel taper on the distribution of atherosclerotic lesions was examined in 11 photographs of human infra-renal aortas, aged 42-76. The aortas, opened anteriorly and pinned flat with the luminal surface facing up, were analysed using Mocha Imaging software. The circumference was measured at intervals of 10 mm and fitted with a third order polynomial function. Taper was calculated for three segments within each aorta as the angle produced by the slope of the decrease in circumference with length. The percent surface area covered by fatty streaks correlated strongly to the coefficients of the polynomial function, while the raised lesions showed a significant increase with an increase in taper angles in the different segments. The results suggest taper plays a role in the distribution of both fatty streaks and raised lesions, however, the nature of the cause and effect remains to be determined.

A quantitative study of the localization of atherosclerotic lesions in the rabbit aorta.

A quantitative study of the size and location of early sudanophilic lesions has been carried out on the aortae of five rabbits. The atherosclerotic lesions were induced by feeding an average of 114 (+/- 3.7) egg yolks over an average period of 83 (+/- 1.7) days. The aortic lesions were visualuized by gross staining with Sudan III and measured by the polar coordinate method. The lesions were almost entirely around orifices: their size was directly proportional to the area of the associated ostium (P less than 0.005). The sudanophilic deposits were located downstream from the ostia in areas believed to experience high shear stresses. The area of the intercostal ostia increased as one proceeded down the thoracic aorta (P less than 0.005). A deviation from the distal distribution pattern was observed where local flow and shear stresses were disturbed. The coronary lesions completely surrounded the ostia, the coeliac lesions had significant proximal components and the left renal and inferior mesenteric lesions were skewed to the right. The study suggests that hemodynamic forces and particularly high shear stress is of considerable importance in both the initiation and localization of early atherosclerotic lesions.

Effect of unilateral nephrectomy of the localization of aortic sudanophilic lesions in cholesterol-fed rabbits.

The method of polar coordinate mapping of lesions on the aortic wall around orifices was used to study the effect of unilateral nephrectomy on sudanophilic lesions in rabbits. Four groups of six rabbits each were studied. Group I had a right nephrectomy with a short renal artery stump. Group II had a sham right nephrectomy. Group III had a left nephrectomy with a short renal artery stump and Group IV a sham left nephrectomy. All groups were allowed to recover for one week and were then fed rabbit chow enriched with 2% cholesterol and 6% corn oil for a four-week period. Groups I and III showed a marked alteration in lesions around the stump with both proximal and distal peaks. Group I also showed the development of skewing of the left renal lesion to the left, presumably because of a change in aortic flow due to removal of the proximal right renal flow. The right renal artery had a lesion skewed to the right; this was not altered by left nephrectomy and was probably due to coeliac flow. The data provides strong evidence that sudanophilic lesions in rabbit aortas are altered by local flow disturbances.

A new probability mapping method to describe the development of atherosclerotic lesions in cholesterol-fed rabbits.

A new probability mapping method was developed to quantify the size and location of lesions near aortic orifices. The precise location of any part of the lesion could be compared between rabbits. Colour photographs of lesions were projected onto a digitizing tablet, and coded as lesion or non-lesion. Next the orifices were warped onto a standard orifice, and then the lesion mapped to maintain the original length and angular location of the lesion from the edge of the orifice. This method, unlike the previously used polar mapping method, excludes neither absent lesions nor ones which surround more than one orifice. In contrast to other probability mapping methods it warps the orifice rather than the artery wall containing the lesion, and so is easier to use for correlation with histological studies. The eventual aim is to use the probability maps as a tool to estimate the age of various positions of the lesion and to identify areas for histological sampling. The method was used to describe the distribution of lesions in 21 rabbits fed a diet with cholesterol levels declining from 0.5% during the first week, to 0.25% during the next two weeks, to 0.125% for weeks 3-10, to 0.1% for weeks 11-24. This feeding protocol produces fatty lesions which are transformed into fibro-fatty and fibrous lesions with time.

A quantitative study of the development of sudanophilic lesions in the aorta of rabbits fed a low-cholesterol diet for up to six months.

Eighteen New Zealand white rabbits were placed on a diet containing 0.25% cholesterol and 6% corn oil for periods of up to 6 months. The animals were divided into 4 gours and sacrificed after 2, 3, 4, and 6 months. The aortae were removed, stained with Sudan III, and analyzed with the polar coordinate technique. While the periorificial lesions developed more slowly on this diet than on the diet with 2% cholesterol and 6% corn oil we had used previously, there was no differences in either the location or the shape of the lesions. In the descending thoracic aorta, lesions developed initially distal to orifices; however, significant lateral and proximal components were observed as atherogenesis progressed. The coronary lesions completely surrounded the ostia in all stages of development. The total area of the lesions was more related to time (r = 0.74, P less than 0.01), than to serum cholesterol (r - 0.51, P less than 0.05) or to cholesterol-time product (r = 0.69, P less than 0.01).

Artefacts of localization of atherosclerosis in pinned aortas.

Detailed studies of aortas from 8 rabbits showed that serious artefacts occur if the aortas are pinned at their in vivo dimensions, rather than fixed at physiological pressure. In the pinned aortas, the proximal parts of the branches were pulled up onto the aortic wall. This was more pronounced for the large branches of the abdominal aorta than for the smaller intercostal branches. This artefact caused atherosclerotic lesions, which had developed at the origin of the branch, to appear as if they were entirely on the aortic wall. We found that a marked change in the elastin pattern was present at the origin of the branch; this can be used to mark the true origin of the branch. With the pressure technique we found that lesions had different shapes and locations at different branch points.

The effect of the duration of cholesterol feeding on the development of sudanophilic lesions in the rabbit aorta.

Polar coordinate mapping was used to determine the rate of growth of individual sudanophilic lesions on the aortic wall around several major branches of the aortae of cholesterol fed rabbits. Four groups, with 6 8-month old male albino white rabbits in each, were used in the study. One group served as a control and the remaining 3 were fed a diet of 2% cholesterol and 6% heated corn oil mixed with ground rabbit pellets for 4, 8, and 10 weeks each. Animals were sacrificed, the aortae removed, stained with Sudan III, pinned at in vivo dimensions, and mapped by the polar coordinate method. No sudanophilic lesions were observed in the control animals. In the experimental groups, the early lesions, except the coronaries, were almost entirely distal to the orifices, and maintained roughly the same contour while spreading around the orifice. The coronary lesions completely encircled the orifices as described previously. As lesions progressed, they became elevated and often granular, so that the lesions themselves may have affected flow profiles around the orifices. Lesions around adjacent orifices were fused in 48% of the cases after 10 weeks on the diet, as opposed to 2% after 4 weeks on the diet. More prolonged experiments were not possible with this diet as the animals developed jaundice and diarrhea. Hemodynamically, these results suggest that early sudanophilic lesions in cholesterol-fed rabbits develop on the aortic wall in areas of high shear stress.

Local changes in collagen content in rabbit aortic atherosclerotic lesions with time.

A video-microdensitometry method was used to quantify collagen on a pixel by pixel basis at specific locations within atherosclerotic lesions, induced by feeding New Zealand White rabbits a declining cholesterol diet. Probability mapping (done previously) showed the lesions occurred in predictable locations so the age of any one area of a lesion could be estimated. The oldest part of the lesion was on the lateral lip of the flow divider as lesions always developed there first. Serial sections through three lesions showed the distribution of collagen within the plaque. The measurement of all the lesions showed that early lesions had low collagen levels: 3-6% for the intercostals and 8-11% for the renals. Collagen increased consistently with time on the diet to 11-15%, for the intercostal and approximately 17% for the renal. The aorto-renal lesions were more severe; thicker and with a higher percentage of collagen. The greatest collagen content was on the lateral lip of the flow dividers, where controls had intimal thickening. No fibrous caps were seen. This study shows conclusively that collagen is not evenly distributed in lesions, but develops progressively in specific locations of fatty plaques with the duration of cholesterol feeding. Intimal thickening predisposes to early collagen development.

Quantitative measurements of early atherosclerotic lesions on rabbit aortae from vascular casts.

We have measured the size and location of fatty atherosclerotic lesions in the cholesterol fed rabbit by scanning electron microscopic analysis of vascular casts. The aortic periorificial lesions examined in this study have been divided into proximal (upstream), distal (downstream) and lateral (to the sides of the branch). Lateral lesions developed after 1 week on diet, followed by distal lesions at 2 weeks, and occasional proximal lesions at 4 weeks. The area covered by the lesions shows a strong linear relationship (P less than 0.001) to the cross-sectional area of the branch after 2 weeks on the cholesterol diet.

The distribution of fibro-fatty atherosclerotic lesions in the aortae of casein- and cholesterol-fed rabbits.

Atherosclerotic lesions were induced in the aortas of 50 rabbits by feeding a semi-purified cholesterol-free casein diet or normal rabbit chow with a low level of added cholesterol for 6 or 10 months. Following perfusion fixation, the aortae from these animals were opened along their length, stained with oil red O and photographed en face. Orifice associated lesions were mapped by measuring radial lesion length at 10 degrees intervals circumferentially around ostia. Histology of these lesions revealed abundant smooth muscle cells surrounded by collagen and elastin in the extracellular matrix, typical of fibrous plaques, as well as oil red O staining lipid and some macrophage derived foam cells. These fibro-fatty lesions were found distal and lateral to ostia, at the same locations as fatty streaks seen in rabbits fed a 2% cholesterol diet for 1 week to 2 months in previous studies. The results of this study show that in moderately hypercholesterolemic rabbits fed an atherogenic diet for 6 to 10 months, advanced atherosclerotic plaques develop in the same location as the fatty streaks seen in short term experiments.

A digital-radiographic technique for in vitro tissue thickness measurement using iodine displacement.

A novel digital radiographic method has been developed for measuring the thickness of a tissue sample by iodine displacement. This simple, accurate method is useful both in medical research and in the comparison of pathological and clinical findings. Radiographic measurements of tissue samples in air suffer from the limitation that the quantity measured depends on the product of both thickness and radiographic attenuation coefficient. This technique allows one to obtain thickness measurements from a digital radiograph of a tissue sample suspended in a bath of radio-opaque contrast agent. The attenuation of the iodinated contrast agent is much higher than that of tissue or calcium. Thus the resulting image is determined largely by the contour of the surrounding bath, and is only slightly influenced by the composition of the tissue. This technique improves both the accuracy and precision of radiographic thickness determination. In this paper the iodine displacement technique is described and the accuracy and precision of thickness measurements in appropriate phantoms are quantified. This technique has been used with both image intensifier and screen-film based imaging systems to obtain thickness maps of calcified human aorta, with precision better than 4% and spatial resolution of 2.5 mm-1.

Hemodynamic method to predict whether hunterian ligation of the basilar artery will lead to thrombosis of basilar bifurcation aneurysms.

In some cases, basilar artery aneurysms cannot be repaired surgically and the basilar artery is occluded near the neck of the aneurysm to stop flow into the aneurysm. After the operation, the aneurysm can fill only by flow through the posterior communicating arteries (PCoAs). Hemodynamically if the flow were the same in both PCoAs and there were no phase lag in the pressures, there would be no pressure gradient for flow to go across the neck of the aneurysm and therefore the aneurysm would thrombose. We have assumed that the diameter of the artery is roughly proportional to the flow that goes through it chronically. We measured the diameters of the PCoAs in 25 patients who had hunterian ligation of the basilar artery. We also measured the maximal width, height, and depth of the aneurysms on angiograms obtained before and after operation. Eleven aneurysms thrombosed completely and had a diameter ratio of greater than 0.6; 10 aneurysms thrombosed partially and had a diameter ratio of 0.46 approximately 1.0; 4 aneurysms did not change and had a diameter ratio of less than 0.45. The ratio of the sizes of the PCoAs pre- and postoperatively was comparable in most cases, so we believe that it is possible to predict reasonably accurately from this simple measurement whether the aneurysm is likely to thrombose if the basilar artery is ligated.

Stenotic and occlusive disease of the venous drainage system of deep brain AVM's.

Angiograms obtained prior to treatment in 53 cases of deep-seated cerebral arteriovenous malformations (AVM's) were retrospectively analyzed with particular attention to the topography of the AVM nidus and the venous drainage. The location of the lesion was determined by a combination of angiography and computerized tomography. Twenty-seven AVM's involved the basal ganglia and thalamus, 12 were located in the corpus callosum, six were intraventricular, and eight involved the mesencephalon and brain stem. Forty-one patients (77.3%) presented with intracranial hemorrhage. Vessel wall irregularities and/or stenosis of the system of the vein of Galen were observed in 14 cases, and occlusion of the deep venous system was present in seven cases. These AVM's showed numerous collateral venous pathways through enlarged medullary and cortical regional veins. There was dominant participation of the basal vein of Rosenthal in all cases. Unique local hemodynamic factors produced by the convergence of the draining veins of the AVM's into the vein of Galen and straight sinus may lead to a higher incidence of stenosis and/or occlusion of the venous drainage. The relatively high incidence of intracranial hemorrhage in these deep-seated AVM's may suggest a relationship between an increased incidence of intracranial bleeding and impaired venous outlets.

Factors in the propagation of aortic dissections in canine thoracic aortas.

Factors were examined which altered the propagation of aortic dissections in canine aortas. Thoracic aortas were removed from sacrificed dogs from the distal end of the arch to the diaphragm. An intimal tear was created at the proximal end of the aorta. The dissection was propagated using a pulsatile pressure system with no flow. The aorta was perfused with a dilute black paint solution, which allowed both video monitoring of the extension of the dissection and measurement of the dissection rate. The dependence of the dissection rate on the variables peak pressure, (dP/dt)max and intimal tear depth was examined. The dissection rate was found to be dependent on (dP/dt)max (p less than 0.005) and the intimal tear depth, expressed as a percentage of wall thickness (p less than 0.01), but not on the peak pressure or intimal tear length. The equation relating the significant variables was log (dissection rate) = (-0.034) X % tear depth +(1.89 +/- 0.56) X (dP/dt)max -(4.3 +/- 1.8); r = 78. Thus a higher (dP/dt)max was associated with a more rapid dissection rate and a deeper intimal tear was associated with a slower dissection rate.

The strength of the aortic media and its role in the propagation of aortic dissection.

The elastic properties of the thoracic aorta are well known, but this is the first study of the inherent strength of the tunica media. The latter is crucial to understand how dissecting aneurysms occur. Pressure-volume (P-V) measurements were recorded as a dilute suspension of India ink was infused into the tunica media using a constant flow pump (0.21 or 0.88 ml min-1) attached to a 20 G needle inserted into the media. Tests done on 31 opened pig upper descending thoracic aortas showed the peak pressure to tear the media averaged 77.2 +/- 1.5 kPa (579 mm Hg). The initial slope of the P-V curve revealed the average distensibility of the media of 3.02 +/- 0.28 (MPa)-1. The work per unit area of tissue required to propagate a tear in the aorta was 15.9 +/- 0.9 mJ cm-2. These values were independent of the tear depth at the 95% confidence interval.

The use of ligament efficiency to model fenestrations in the internal elastic lamina of cerebral arteries. I--modelling scheme.

The spatial geometry of fenestrations (windows) in the internal elastic lamina from human cerebral arteries is being characterized by a single parameter termed 'ligament efficiency', which is a ratio of the solid band of material to the centre-to-centre spacing between two or more holes. As a result, the apparent random distribution of fenestrations with variable diameters may be represented as a uniform array of holes with a single diameter. The actual arrangement of fenestrations from three separate tissue specimens were replicated in thin latex sheets by transposing the image of the fenestrations from photomicrographs obtained with the scanning electron microscope. In a similar manner, the uniform array of holes with an equivalent ligament efficiency are modelled in latex sheets. The tensile (stress-strain) properties of the latex sheets representing the replication and model configurations were comparable for all three specimens, even though their individual ligament efficiencies were different. The close similarity between the elastic characteristics for the two configurations, verifies the application of ligament efficiency to represent the spatial geometry of a perforated material such as the fenestrated internal elastic lamina.

The use of ligament efficiency to model fenestrations in the internal elastic lamina of cerebral arteries. II--analysis of the spatial geometry.

During uniaxial extension of the latex model, the holes (both marked on the sheet and perforated) demonstrated a distinctive change in their shape, from circular to elliptical. Measurements of the axial and transverse diameters of a consistent row of holes, along with the width of the sample at distinct elongations were converted into an area, ligament efficiency (transverse and axial) and shape factor (eccentricity). Values for the expansion ratio of the holes and necking of the latex sheet were also computed. The perforations expanded more rapidly than with the holes only marked on the solid material, which translated into a more rapid increase for the axial diameter, area and eccentricity, while the axial ligament efficiency decreased more rapidly. The transverse diameter and transverse ligament efficiency remained essentially constant. Necking of the latex sheet was consistent for all of the specimens (both solid and perforated). The relative expansion ratios paralleled the relative changes for the standardized stresses presented in Part I.

Steady flow visualization in a rigid canine aortic cast.

Steady flow studies were conducted in a transparent canine aortic cast. The cast segment stretched from the aortic valve to beyond the renal arteries and included all major branches. Flow was visualized by analysis of dye streaklines. Flow rates for basal and exercising cardiovascular states were simulated. The Reynolds numbers in the ascending aorta for basal and exercising conditions were 900 and 1587 respectively. Aortic core flow was laminar in basal simulations. Disturbed flow commenced in the upper descending aorta with exercising flow rates. Separation zones existed along the inner curvature of the aortic arch and the proximal walls of the brachiocephalic, left subclavian, and coeliac arteries. Such zones may exist over a portion of the cardiac cycle. If either renal artery was occluded, then a vortex formed. This vortex is associated with high shear regions which correlate well with sites where sudanophilic lesions have been reported in cholesterol-fed nephrectomized rabbits.