Edwin F. Donnelly

Phase I Study of Intravenous Vascular Endothelial Growth Factor Trap, Aflibercept, in Patients With Advanced Solid Tumors

PURPOSE Vascular endothelial growth factor (VEGF) Trap (aflibercept) is an angiogenesis inhibitor comprising portions of the extracellular domains of human VEGF receptors 1 and 2 fused to the Fc portion of human immunoglobulin G. This phase I study was designed to evaluate the safety, pharmacokinetics, and pharmacodynamics of VEGF Trap administered intravenously (IV) every 2 weeks. PATIENTS AND METHODS Patients with refractory solid tumors or non-Hodgkins lymphoma with adequate organ function were eligible. Pharmacokinetic/pharmacodynamic markers included measurement of plasma VEGF bound to VEGF Trap and free VEGF Trap. Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) was incorporated to measure the biologic effects of the drug on tumor vascularity and permeability. RESULTS The study enrolled 47 patients at doses ranging from 0.3 to 7.0 mg/kg IV every 2 weeks. Dose-limiting toxicities were rectal ulceration and proteinuria at the 7.0 mg/kg dose. Other mechanism-specific toxicities included hypertension. On the basis of these observations and on pharmacokinetics, the recommended phase II dose of VEGF Trap as a single agent is 4 mg/kg every 2 weeks. Three RECIST (Response Evaluation Criteria in Solid Tumors) -defined partial responses were observed, one at the 3.0 mg/kg and two at the 7.0 mg/kg dose level. Maximum plasma concentration of free VEGF Trap increased proportionally with dose. Maximal VEGF-bound VEGF Trap complex levels were reached at doses > or = 2.0 mg/kg. Changes in volume transfer constant measured by DCE-MRI at baseline and at 24 hours after administration indicate a possible dose-related change in this pharmacodynamic marker. CONCLUSION IV VEGF Trap was well tolerated at the dose levels tested. Pharmacodynamic and pharmacokinetic markers were indicative of VEGF blockade.

Integrin-mediated targeting of drug delivery to irradiated tumor blood vessels

The objective of this study was to target drug delivery to radiation-induced neoantigens, which include activated receptors within the tumor vasculature. These responses include posttranslational changes in pre-existing proteins, which can be discovered by phage-displayed peptide libraries administered to mice bearing irradiated tumors. Phage-displayed peptides recovered from irradiated tumors included the amino acid sequence RGDGSSV. This peptide binds to integrins within the tumor microvasculature. Immunohistochemical staining of irradiated tumors showed accumulation of fibrinogen receptor alpha(2b)beta(3) integrin. We studied tumor targeting efficiency of ligands to radiation-induced alpha(2b)beta(3). Radiopharmaceuticals were localized to irradiated tumors by use of alpha(2b)beta(3) ligands conjugated to nanoparticles and liposomes. Fibrinogen-conjugated nanoparticles bind to the radiation-activated receptor, obliterate tumor blood flow, and significantly increase regression and growth delay in irradiated tumors. Radiation-guided drug delivery to tumor blood vessels is a novel paradigm for targeted drug delivery.

Increased and Prolonged Pulmonary Fibrosis in Surfactant Protein C-Deficient Mice Following Intratracheal Bleomycin

Recent reports have linked mutations in the surfactant protein C gene (SFTPC) to familial forms of pulmonary fibrosis, but it is uncertain whether deficiency of mature SP-C contributes to disease pathogenesis. In this study, we evaluated bleomycin-induced lung fibrosis in mice with genetic deletion of SFTPC. Compared with wild-type (SFTPC+/+) controls, mice lacking surfactant protein C (SFTPC-/-) had greater lung neutrophil influx at 1 week after intratracheal bleomycin, greater weight loss during the first 2 weeks, and increased mortality. At 3 and 6 weeks after bleomycin, lungs from SFTPC-/- mice had increased fibroblast numbers, augmented collagen accumulation, and greater parenchymal distortion. Furthermore, resolution of fibrosis was delayed. Although remodeling was near complete in SFTPC+/+ mice by 6 weeks, SFTPC-/- mice did not return to baseline until 9 weeks after bleomycin. By terminal dUTP nick-end labeling staining, widespread cell injury was observed in SFTPC-/- and SFTPC+/+ mice 1 week after bleomycin; however, ongoing apoptosis of epithelial and interstitial cells occurred in lungs of SFTPC-/- mice, but not SFTPC+/+ mice, 6 weeks after bleomycin. Thus, SP-C functions to limit lung inflammation, inhibit collagen accumulation, and restore normal lung structure after bleomycin.

Quantified Color Doppler Sonography of Tumor Vascularity in an Animal Model

This study was designed to evaluate the accuracy of a system to quantitate tumor vascularity with amplitude (power) color Doppler sonography two‐ and three‐dimensionally. The vascularity of 20 transplanted murine tumors was determined with quantitated amplitude color Doppler sonography both two‐ and three‐dimensionally and compared to tumor vascularity estimated by histologic examination. Serial examinations were performed 15, 30, 45, and 60 min after the injection of the exotoxin CM‐101 and saline solution to assess changes in tumor vascularity. Three‐dimensional amplitude color Doppler sonography best depicted the overall vascularity of tumor when compared to histologic estimation of vessel density. However, neither two‐ nor three‐dimensional amplitude color power angiography correlated well to the microvessel count, probably a reflection of the difference in the method for vessel quantification using sonographic versus histologic techniques. Three‐dimensional amplitude Doppler sonography correlated better with counts of large vessels (> 100 microm) as opposed to small vessels (> 15 microm). Time‐activity curves showed no difference in tumor flow at the times measured in the experimental group injected with CM‐101 or when compared to saline solutions in either the peripheral or central portions of the tumor. This three‐dimensional amplitude color Doppler sonographic system affords global quantification of tumor vascularity and flow that may, in turn, be useful in determining the probability of malignancy (by determination of branching patterns and vessel regularity) or tumor response or both to treatment.

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.

Three-dimensional color Doppler sonography before and after fibroid embolization.

Three‐dimensional color Doppler sonography was performed within 1 hour, 1 day, 3 months, and 6 months of fibroid embolization in 20 patients who had a total of 31 fibroids greater than 2 cm in average dimension. The greatest decrease in vascularity occurred 1 day after the procedure, whereas the greatest volume change was found at the 3 month follow‐up examination. In about one half of the patients scanned, depiction of fibroid vascularity by color Doppler sonography was found to improve the delineation of the size, location, and extent of myometrial involvement. Hypervascular fibroids (12 of 31) tended to decrease in size after treatment more than isovascular (10 of 31) or hypovascular ones (9 of 31). Additional investigations that are similar to this one will be needed to determine if three‐dimensional color Doppler sonography can be used to predict those who will be responders, partial responders, or nonresponders to embolotherapy.

Quantification of the effect of kVp on edge‐enhancement index in phase‐contrast radiography

This study was performed to measure the dependence of edge-enhancement in polychromatic phase-contrast radiography on x-ray tube operating voltage. Measurements of edge enhancement were made at tube voltages from 40 to 86 kVp using a tungsten anode x-ray tube with a nominal focal spot size of 100 micrometers. A relatively weak attenuating, sharp edge consisting of a thin lucite sheet (3 mm) in air was imaged utilizing phase-contrast radiography (PC-R). PC-R images were acquired at different radiographic techniques in which x-ray tube voltage was varied from 40 to 86 kVp. The image receptor was a single emulsion x-ray mammography cassette. Optical density profiles across the edge of the object were obtained using a film digitizer and edge-enhancement indices were calculated. Increasing kVp resulted in a gradual decrease of the edge-enhancement index. Even at the highest kVp (86), however, important edge-enhancement effects were evident. While there is some degradation in the edge-enhancement effect of phase-contrast radiography at higher kVps, the decrease from 40 to 86 kVp is relatively small (11%). Our results suggest that further investigation into the role of phase-contrast imaging at higher kVp values for the purpose of patient dose reduction while still realizing the advantage of phase-contrast effects for improved soft-tissue detectability is warranted.

Quantification of the effect of system and object parameters on edge enhancement in phase-contrast radiography.

The purpose of this study was to evaluate the effects of system parameters (focal spot size, tube voltage, geometry, detector resolution, and image noise) and object characteristics (edge gradient/ shape, composition, thickness, and overlying attenuating material) upon the edge enhancement effect in phase-contrast radiography. Each variable of interest was adjusted and images of a 3 mm lucite phantom were obtained with the other variables remaining constant. A microfocus x-ray source coupled to a CCD camera with an intensifying screen was used to acquire the digital images. Two parameters of image analysis were used to quantify the effects. The edge enhancement index (EEI) was used to measure the absolute degree of edge enhancement, while the edge enhancement to noise ratio (EE/N) was used to measure the conspicuity of the edge enhancement relative to image noise. Little effect on EEI was seen from tube voltage, object thickness, overlying attenuating material, while focal spot size and system geometry demonstrated measurable effects upon the degree of edge enhancement. It was also shown that while the edge enhancement effect over straight edges is highly dependent upon how the edge aligns with the x-ray beam, rounded edges, which better model biological objects, do not suffer from this dependence and the EEI reaches its maximal level at any alignment. Decreasing detector resolution diminished the EEI slightly, but even with pixel sizes of 0.360 x 0.360 mm edge enhancement effects were readily visible. The effect of image noise on EE/N was evaluated using different exposure times showing an expected improvement with longer exposure time with EE/N approaching a plateau at 5 min. Many of the parameters that will go into the design of a future PC-R imaging system have been quantified in terms of their effect on the degree of edge enhancement in the acquired image. These results, taken together, indicate that either a specimen or even clinical breast imaging system could be created with currently available technology. The major limitation to a clinical system would be the low x-ray flux from the microfocal x-ray source.

Characterization of the phase-contrast radiography edge-enhancement effect in a cabinet x-ray system

The purpose of this study was to demonstrate that a commercially available cabinet x-ray system is capable of phase-contrast radiography (PC-R) and to evaluate the effect of different system parameters on the degree of edge enhancement. An acrylic plastic edge phantom was imaged at different tube potentials (25-60 kV) and in different geometries (variable object-to-detector distances, R(2), at a constant source-to-detector distance, R(1) + R(2)). In addition, the effect of noise on the perceived edge enhancement was studied as a function of exposure time. Our results show that a modest degree of phase contrast can be achieved in an unmodified cabinet x-ray system. In addition, the particular system evaluated allowed low-noise PC-R images to be obtained with short (6 s or less) exposures. These results suggest that with appropriate geometric choices PC-R is already available to a wide range of research scientists for use in both small-animal and human-specimen experiments.

Three-dimensional Color Doppler Sonography and Uterine Artery Arteriography of Fibroids Assessment of Changes in Vascularity Before and After Embolization

Objective. To assess the accuracy of three‐dimensional color Doppler sonography and uterine artery arteriography in depicting changes in fibroid vascularity before and after embolization. Methods. Preembolization and postembolization three‐dimensional color Doppler sonography and selective uterine artery arteriography were retrospectively compared in 15 patients who underwent uterine artery embolization for treatment of symptomatic fibroids. Three‐dimensional color Doppler sonography was performed by using a scanner with color power angiographic imaging capability. Vascularity was quantified by using an estimation of power‐weighted pixel density as described by our group in previously published studies. Uterine artery arteriography was performed by using a standard selective microcatheter embolization technique. For purposes of comparison, fibroids were classified as either hypervascular or hypovascular relative to myometrial vascularity before and minutes to several hours after uterine artery embolization. Changes in fibroid vascularity (i.e., from hypervascular to hypovascular) as depicted by three‐dimensional color Doppler sonography were compared with those shown on uterine artery arteriography and classified as being in agreement or disagreement. Results. In 13 (87%) of 15 patients there was agreement; in 2 (13%) of 15 there was disagreement. In both cases of disagreement, three‐dimensional color Doppler sonography showed collateral flow not depicted by uterine artery arteriography. The mean reduction in quantitated vascularity after uterine artery embolization was 44% (range, 19%–78%). Conclusions. Three‐dimensional color Doppler sonography accurately depicts fibroid vascularity and in some cases can reveal collateral flow not depicted by uterine artery arteriography.