Comparison of manual and automatic segmentation methods for brain structures in the presence of space-occupying lesions: a multi-expert study.

The purpose of this work was to characterize expert variation in segmentation of intracranial structures pertinent to radiation therapy, and to assess a registration-driven atlas-based segmentation algorithm in that context. Eight experts were recruited to segment the brainstem, optic chiasm, optic nerves, and eyes, of 20 patients who underwent therapy for large space-occupying tumors. Performance variability was assessed through three geometric measures: volume, Dice similarity coefficient, and Euclidean distance. In addition, two simulated ground truth segmentations were calculated via the simultaneous truth and performance level estimation algorithm and a novel application of probability maps. The experts and automatic system were found to generate structures of similar volume, though the experts exhibited higher variation with respect to tubular structures. No difference was found between the mean Dice similarity coefficient (DSC) of the automatic and expert delineations as a group at a 5% significance level over all cases and organs. The larger structures of the brainstem and eyes exhibited mean DSC of approximately 0.8-0.9, whereas the tubular chiasm and nerves were lower, approximately 0.4-0.5. Similarly low DSCs have been reported previously without the context of several experts and patient volumes. This study, however, provides evidence that experts are similarly challenged. The average maximum distances (maximum inside, maximum outside) from a simulated ground truth ranged from (-4.3, +5.4) mm for the automatic system to (-3.9, +7.5) mm for the experts considered as a group. Over all the structures in a rank of true positive rates at a 2 mm threshold from the simulated ground truth, the automatic system ranked second of the nine raters. This work underscores the need for large scale studies utilizing statistically robust numbers of patients and experts in evaluating quality of automatic algorithms.

Unilateral radiation pneumonitis in sheep: physiological changes and bronchoalveolar lavage.

Radiation pneumonitis is a life-threatening result of therapeutic thoracic irradiation, yet its mechanisms are poorly understood. We studied the effects of unilateral lung irradiation (3,000 rad) in sheep from the immediate response to the later development of radiation pneumonitis. We defined radiation pneumonitis by its diagnostic clinical feature, radiographic infiltration of the irradiated zone with a straight margin corresponding to the radiation port. The immediate response in the few hours after irradiation was characterized by cough, labored respiration, hypoxemia (arterial PO2 decreased 19 Torr), mild pulmonary hypertension (pulmonary arterial pressure increased 20%), and lymphopenia. Hemodynamics and gas exchange returned to normal by day 2 but became abnormal again before or during radiation pneumonitis at 32 +/- 2 days. Respiratory distress, hypoxemia, and pulmonary hypertension recurred during radiation pneumonitis. Bronchoalveolar lavage during radiation pneumonitis contained increased neutrophils (19 +/- 4%, control = 7%), increased protein (0.27 +/- 0.1 g/dl, control = 0.12 +/- 0.03), and severely impaired ability to lower surface tension. Alveolar macrophages from both lungs during unilateral radiation pneumonitis exhibited impaired generation of superoxide after phorbol myristate (only a 30% increase). Normal control alveolar macrophages increased superoxide production after stimulation greater than 400%. We conclude that unilateral lung irradiation in sheep causes a mild immediate response followed by radiation pneumonitis at 1 mo. Unilateral radiation pneumonitis in this model is associated with ipsilateral neutrophilic alveolitis, increased bronchoalveolar lavage protein, and impaired surfactant function, as well as bilateral functional abnormalities of alveolar macrophages.

Corticosteroids prevent acute lung dysfunction caused by thoracic irradiation in unanesthetized sheep.

We sought to determine the effect of corticosteroid therapy in a new acute model of oxidant lung injury, thoracic irradiation in awake sheep. Sheep were irradiated with 1,500 rads to the whole chest except for blocking the heart and adjacent ventral lung. Seven experimental sheep were given methylprednisolone (1 g intravenously every 6 h for four doses) and thoracic irradiation; control sheep received only irradiation. In irradiated control sheep, lung lymph flow increased from baseline (7.6 ml/h) to peak at 3 h (13.2), and lung lymph protein clearance increased from 5.1 to 9.7 ml/h. Mean pulmonary artery pressure increased in the irradiated control sheep from 19 to 32.4 cm H2O, whereas the lung lymph thromboxane concentration increased from 0.09 to 6.51 ng/ml at 3 h. Arterial oxygen tension in irradiated control sheep fell gradually from 86 mm Hg at baseline to 65 mm Hg at 8 h. Methylprednisolone administration significantly prevented the increase in lung lymph protein clearance, mean pulmonary artery pressure, and lung lymph thromboxane concentration. Methylprednisolone also prevented the fall in arterial oxygen tension after thoracic irradiation, but did not prevent a further decrease in lymphocytes in blood or lung lymph after radiation. We conclude that corticosteroid therapy prevents most of the acute physiologic changes caused by thoracic irradiation in awake sheep.

Simultaneous radiation and chemotherapy for advanced carcinoma of the cervix.

Six patients with poor prognosis carcinoma of the cervix were treated with external radiation therapy simultaneously with cisplatin, bleomycin, and vincristine. Toxicity was very mild with nausea and vomiting and mild myelosuppression being the major toxicities. At a median of 36 months follow-up, four of six patients are alive, three with no evidence of disease. The median survival after diagnosis is 25+ months. The data suggest that radiation therapy and cytotoxic therapy administered together in patients with advanced cervix carcinoma is well tolerated. Further study to determine therapeutic efficacy is warranted.

Diamide exposure, thermal resistance, and synthesis of stress (heat shock) proteins.

Chinese hamster ovary (CHO) cells were treated with the thiol oxidant diamide for 1 hr at 37 degrees, incubated in diamide-free medium for 4 hr at 37 degrees, and then exposed to hyperthermic treatment (43 degrees) or assayed for the presence of 110, 90 and 66 kilodalton (kD) stress (heat shock) proteins. Cellular inactivation produced by the hyperthermic treatment was measured using colony formation as the end point. Low concentrations of diamide, which did not result in depletion of intracellular GSH, induced a moderate degree of protection against thermal toxicity but did not affect the pattern of protein synthesis. Exposure to 0.4 mM diamide, which reduced intracellular GSH concentrations by 50-60%, significantly reduced the rate of hyperthermic cellular inactivation. This occurred coincidentally with the synthesis of stress proteins of approximate molecular weights of 110, 90 and 66 kD. Furthermore, this concentration of diamide protected cells from thermal inhibition of protein synthesis. These results indicate that thiol oxidation by diamide can induce both the development of thermal resistance to cellular inactivation and the synthesis of stress proteins.

Acute effects of thoracic irradiation on lung function and structure in awake sheep.

To investigate the acute physiological and structural changes after lung irradiation, the effects of whole-lung irradiation were investigated in fourteen sheep. Ten sheep were prepared with vascular and chronic lung lymph catheters, then a week later were given 1,500 rad whole-lung radiation and monitored for 2 days. Four sheep were given the same dose of radiation and were killed 4 h later for structural studies. Lung lymph flow increased at 3 h after radiation (14.6 +/- 2.1 ml/h) to twice the base-line flow rate (7.5 +/- 1.3), with a high lymph-to-plasma protein concentration. Pulmonary arterial pressure increased twofold from base line (18 +/- 1.6 cmH2O) at 2 h after radiation (33 +/- 3.8). Cardiac output and systemic pressure in the aorta did not change after lung radiation. Arterial O2 tension decreased from 85 +/- 3 to 59 +/- 4 Torr at 1 day after radiation. Lymphocyte counts in both blood and lung lymph decreased to a nadir by 4 h and remained low. Thromboxane B2 concentration in lung lymph increased from base line (0.07 +/- 0.03 ng/ml) to peak at 3 h after radiation (8.2 +/- 3.7 ng/ml). The structural studies showed numerous damaged lymphocytes in the peripheral lung and bronchial associated lymphoid tissue. Quantitative analysis of the number of granulocytes in peripheral lung showed no significant change (base line 6.2 +/- 0.8 granulocytes/100 alveoli, 4 h = 10.3 +/- 2.3). The most striking change involved lung airways. The epithelial lining of the majority of airways from intrapulmonary bronchus to respiratory bronchiolus revealed damage with the appearance of intracellular and intercellular cell fragments and granules. This new large animal model of acute radiation lung injury can be used to monitor physiological, biochemical, and morphological changes after lung radiation. It is relevant to the investigation of diffuse oxidant lung injury as well as to radiobiology per se.

Magnetic resonance imaging of the breast: functional T1 and three-dimensional imaging.

Thirty-two patients with breast abnormalities and 10 normal volunteers were examined with magnetic resonance imaging (MRI) utilizing a 0.5-Tesla superconducting magnet imaging system. Sagittal, coronal, and transverse breast imaging with the 50-cm body coil and with a specialized 16 X 8 cm breast coil allowed the evaluation of a single breast. Soft-tissue contrast was clearly apparent using proton density-weighted, T1 weighted, calculated T1 images. In addition, the feasibility and potential clinical utility of three-dimensional color imaging was evaluated using computer-based interactive algorithms for edge detection of the breast and for abnormal structures within the breast. MR scan data were correlated with conventional xeromammography, ultrasound, clinical, data, and pathological specimens. In vivo and in vitro breast imaging was evaluated in a feasibility study. It was concluded that MR imaging is a most sensitive modality for the identification of breast lesions, especially when the adjacent normal tissue is primarily fat. There does appear to be significant overlap in soft-tissue contrast when the adjacent breast parenchymal tissue is fibrocystic in nature. The T1 parameter may be a relatively specific measure for identifying a malignant breast mass.

Role of glutathione in cell survival after hyperthermic treatment of Chinese hamster ovary cells.

Chinese hamster ovary cells cultured in vitro were exposed to hyperthermic treatment at 43 degrees C or 45 degrees C. Depletion of 99.5% of the intracellular glutathione (GSH) by addition of 0.1 mM diethylmaleate (DEM) increased thermal sensitivity. This effect, however, was dependent upon oxygen tension. Decreasing the oxygen tension reduced the effect. Thermal cytotoxicity was further augmented by continued depletion of GSH after hyperthermic treatment. Once cells became thermally tolerant, depletion of GSH did not affect thermal sensitivity. The development of thermotolerance was triggered by exposing Chinese hamster ovary cells to a thermal dose consisting of either 30 min at 43 degrees or 10 min at 45 degrees. After a 4-h incubation at 37 degrees, the cells were reheated at 43 degrees or 45 degrees, respectively. A triggering dose of 30 min at 43 degrees is a nonlethal treatment, while 10 min at 45 degrees kills 50% of the cells. Cells were exposed to 0.1 mM DEM during the triggering and development phases of tolerance. Depletion of GSH by addition of DEM did not inhibit the development of thermal tolerance when triggered by the 30-min-43 degrees thermal dose. However, exposure to DEM did inhibit the development of thermal tolerance if the triggering dose consisted of 10 min at 45 degrees. These results were interpreted to mean that GSH depletion will inhibit the development of thermotolerance only after a sufficiently toxic thermal dose.

Glutathione pool size affects cell survival after hyperthermic treatment.

Intracellular glutathione (GSH) concentrations were titrated in Chinese hamster ovary cells by exposure to various concentrations of diethylmaleate (DEM). The various steady state levels of GSH obtained were maintained throughout the experimental time course. Cells were incubated at 42 degrees after DEM addition in order to produce thermal dose response curves using colony formation as the end point. The slope of the dose response curve was subsequently determined and compared to the intracellular GSH concentration. This comparison indicated Chinese hamster ovary cells contain multiple reservoirs of GSH which in turn regulate thermal toxicity in a stepwise manner. Removal of 50% or less of the GSH did not affect thermal sensitivity. A small increase in sensitivity occurred when 50 to 80% of the GSH was removed. Removal of greater than 80% of the GSH increased thermal toxicity significantly. The facts that 10 and 20 microM DEM produce extensive GSH depletion and only small changes in survival imply that a threshold concentration of GSH must be removed before thermal toxicity is affected.

Ultrarapid recovery from lethal effects of bleomycin and gamma-radiation in stationary-phase human diploid fibroblasts.

An ultrarapid phase of cellular recovery, as measured in liquid holding type experiments, was studied in stationary-phase human fibroblasts exposed to bleomycin or cobalt-60 gamma-irradiation yielding comparable levels of cell killing. This rapid recovery was both faster and considerably greater in magnitude after bleomycin treatments. Bleomycin survival curves were multiphasic, indicating the presence of treated cells with varying sensitivities either at the beginning of treatments or as a result of resistance which developed during the treatment period. The amount of both ultrarapid (within 2 to 10 min) and slower recovery was dose dependent after irradiation with 200 to 800 rads or 30-min exposures to bleomycin (5 to 100 micrograms/ml). Following bleomycin treatments resulting in surviving fractions of 1 to 2%, survival increased up to 8-fold after only 2 min of posttreatment incubation. This rapid increase in survival was followed by a slower increase over time periods up to 3 h. In contrast, the rates of cellular recovery after gamma-irradiation were more gradual from 0 to 3 h. Recovery at all posttreatment intervals was always greater after bleomycin than after gamma-treatments, following doses yielding 1 to 50% survival. The ultrarapid component of cellular recovery after bleomycin treatments may have implications for both clinical cancer management and cellular studies directed toward determining mechanisms of action of bleomycin.

Decreased intracellular glutathione concentration and increased hyperthermic cytotoxicity in an acid environment.

Chinese hamster ovary (CHO) cells were heated at either pH 7.2 to 7.4 or 6.7 to 6.8 in order to determine if conditions which suppress the development of thermotolerance (pH 6.7 to 6.8) reduce intracellular levels of glutathione (GSH). When the pH of the growth medium was reduced from 7.2 to 6.7, a 25 to 30% reduction in GSH was observed in cells maintained at 37 degrees. Cells heated at 42 degrees in medium adjusted to pH 6.7 had lower levels of GSH compared to cells heated at pH 7.2. Cells were also heated for 1 hr at 43 degrees and then incubated at 37 degrees for up to 9.5 hr prior to GSH measurement. The GSH levels of cells treated at pH 7.3 increased approximately 20% above control, whereas treatment at pH 6.7 resulted in a 20% reduction compared to control. Chinese hamster ovary cells were exposed to 5 mM buthionine sulfoximine (BSO) prior to and during 42 degrees heat treatment. BSO exposure at either pH 7.3 or 6.8 reduced the GSH concentration to approximately 65% of control and increased thermal cytotoxicity. The thermal sensitivity of cells incubated at 42 degrees and pH 7.3 was compared to that of cells incubated at pH 6.8. Decreasing the pH from 7.3 to 6.8 increased sensitivity by a factor of 1.87 in the absence of BSO, whereas decreasing the pH in the presence of BSO increased sensitivity by only 1.50. In summary, these results suggest that the increase in thermal sensitivity observed when Chinese hamster ovary cells are heated in acid medium is due partly to the depletion of GSH.

Patient risk from interproximal radiography.

Computer simulation methods for determining patient dose from dental radiography have demonstrated that patient risk from a two-film interproximal examination ranges from 1.1 X 10(-8) to 3.4 X 10(-7) using 90-kVp beams, depending on film speed, projection technique, and age and sex of the patient. Further, changing from a short-cone round-beam to a long-cone technique with rectangular collimation reduces risk by a factor of 2.9, independent of other factors.

Radiosensitivity and thermosensitization of thermotolerant Chinese hamster cells and RIF-1 tumors.

CHO cells subline HA-1 were made thermotolerant by a priming heat treatment (43 degrees C, 30 min). Later, 4, 16, or 24 hr, they were either irradiated or heated (43 degrees C, 30 min) and irradiated. Thermotolerance had no effect on the radiation sensitivity of the cells as measured by the D0 value of the clonogenic survival curve. However, the N value of the curve (width of shoulder) showed a significant increase at 24 hr, indicating an increased capacity to accumulate sublethal damage. This indicates that the fractionation schedule 43 degrees C, 30 min + 37 degrees C, 24 hr + 43 degrees C, 30 min + X ray required approximately 100 rad more radiation than 43 degrees C, 30 min + X ray to reduce survival to the same level. The same priming treatment was given to RIF-1 tumors growing in C3H mice. Later, 24 hr, when the tumors were either irradiated or heated (43 degrees C, 30 min) and irradiated, it was found that thermotolerance had no effect on the radiosensitivity of the cells as measured by in vitro assay. However, thermal radiosensitization was not apparent 24 hr after the priming treatment.

Fraction size and the repair of potentially lethal radiation damage in a human melanoma cell line. Possible implications for radiotherapy.

Studies performed on a human melanoma line in exponential growth showed radiation survival parameters similar to those of other tumors (n = 1.2, D0 = 151). In plateau-phase cultures, however, this tumor cell line was efficient in the repair of potentially lethal radiation damage, especially when large dose fractions were employed (there was a 48-fold recovery after 1,200 rad [12 Gy]). Thus, extrapolation from survival data generated only on exponentially growing cells may be misleading when determining the ultimate surviving fraction in some tumors. Large-fraction radiotherapy may not be effective in tumors that contain cells analogous to plateau-phase cells.