Published online before print March 16, 2006, 10.1148/radiol.2392050568
(Radiology 2006;239:547-553.)
© RSNA, 2006
Lung Cancer Perfusion at Multi–Detector Row CT: Reproducibility of Whole Tumor Quantitative Measurements1
Quan-Sing Ng, MBBS, MRCP,
Vicky Goh, MA, MRCP, FRCR,
Heinz Fichte, Dipl Inf,
Ernst Klotz, Dipl Phys,
Pat Fernie, DCR(R),
Michele I. Saunders, MD, FRCP, FRCR,
Peter J. Hoskin, MD, FRCP, FRCR and
Anwar R. Padhani, FRCP, FRCR
1 From the Marie Curie Research Wing (Q.S.N., M.I.S., P.J.H.) and Paul Strickland Scanner Centre (V.G., P.F., A.R.P.), Mount Vernon Hospital, Rickmansworth Rd, Northwood, Middlesex HA6 2RN, England, and Siemens Medical Solutions, Forchheim, Germany (H.F., E.K.). Received April 6, 2005; revision requested June 6; revision received June 10; accepted June 22. Supported by a pump-priming grant from the Royal College of Radiologists, London, England.
Address correspondence to V.G. (e-mail: vicky.goh{at}paulstrickland-scannercentre.org.uk).
 |
ABSTRACT
|
|---|
Institutional review board approval and informed consent were obtained for this study. The aim of the study was to prospectively assess, in patients with lung cancer, the reproducibility of a quantitative whole tumor perfusion computed tomographic (CT) technique. Paired CT studies were performed in 10 patients (eight men, two women; mean age, 66 years) with lung cancer. Whole tumor permeability and blood volume were measured, and reproducibility was evaluated by using Bland-Altman statistics. Coefficient of variation of 9.49% for permeability and 26.31% for blood volume and inter- and intraobserver variability ranging between 3.30% and 6.34% indicate reliable assessment with this whole tumor technique.
© RSNA, 2006
|
INTRODUCTION
|
|---|
Treatment options for lung cancer to date have included surgery and radical radiation therapy for potentially curable tumors and chemotherapy and palliative radiation therapy for more advanced tumors. Prognosis has remained poor, particularly for advanced disease, with a 5-year survival rate of 2% for stage IV disease. However, manipulation of the tumor vasculature has emerged as a promising therapeutic strategy for patients with advanced lung cancer; antiangiogenic agents that stop new tumor vessel formation and vascular targeting agents that destroy existing tumor blood vessels are currently under evaluation (1).
Functional imaging techniques have been used increasingly to provide an in vivo assessment of tumor vascularity and have been used successfully for therapeutic assessment in clinical trials that involve vascular targeting and antiangiogenic drugs and demonstrate proof of principle of drug mechanism (2–5). Tumor perfusion measurement with multi–detector row computed tomography (CT) is more readily available and provides more robust quantification than does magnetic resonance (MR) imaging. To date, however, multi–detector row CT assessment has been limited to a single tumor level (which comprises a combination of contiguous transverse sections) with a maximum coverage of 4 cm with 64–detector row CT scanners. This assessment may be only a limited sample of the entire region of interest (ROI) and is unrepresentative of the tumor as a whole. Ideally, whole tumor perfusion should be measured. Thus the aim of our study was to prospectively assess, in patients with lung cancer, the reproducibility of a quantitative whole tumor perfusion CT technique that we developed.
|
MATERIALS AND METHODS
|
|---|
Siemens Medical Solutions supported this study by providing the software used for analysis. Two authors (E.K., H.F.) are employees of Siemens Medical Solutions and developed the software. Other authors were in control of the data and information submitted for publication.
Subjects
This study was carried out within a phase IB clinical trial of combretastatin A4 phosphate (Oxigene, Waltham, Mass) in combination with radiation therapy. The study design called for the measurement of reproducibility, against which we would then assess treatment response. We obtained institutional review board approval, which included explicit approval for radiation exposure of patients for research purposes as required under the Ionising Radiation (Medical Exposure) Regulations, or IR(ME)R. Each patient gave written informed consent, which included information about the radiation exposure at the CT examinations and at the subsequent radiation therapy. Ten consecutive patients (eight men, two women; mean age, 66 years; range, 52–79 years) with histologically proved, inoperable non–small cell lung cancer (mean size, 8.3 cm; range, 5.3–11.8 cm; stage III cancer in seven patients; stage IV cancer in three patients) were enrolled prospectively into the study.
Imaging
Patients were scanned with a 16–detector row CT scanner (Sensation 16; Siemens Medical Solutions, Forchheim, Germany). No additional patient preparation was required over and above that for a routine thoracic CT examination. An 18-gauge cannula was placed into an antecubital fossa vein while the patient lay supine on the scanner table. An initial nonenhanced breath-hold helical study was performed to ensure that the entire lung tumor was encompassed, and the following imaging parameters were used: 80 kV; 120 mAs; table feed, 30 mm; rotation time, 0.5 second; detector width, 1.5 mm; reconstruction width, 2 mm; field of view, 500 mm; matrix, 512 mm. This nonenhanced study provided baseline images and was used to plan the subsequent perfusion study. All studies were performed by using breath hold at tidal expiratory volume, as the same tumor coverage was more achievable with a breath hold in expiration than a breath hold in inspiration for this patient group.
The level of the aortic arch was identified from these studies. A single transverse scan was obtained at the level of the aortic arch to allow an ROI to be placed within the aorta for a subsequent contrast material–enhanced bolus-tracking study (CARE bolus; Siemens Medical Solution). One hundred eight milliliters of iobitridol (Xenetix 300; Guerbet, Paris, France), which contained 300 mg of iodine per milliliter, was administered by using a dual-head pump injector (Injekttron CT2; Medtron, Saarbrucken, Germany). The bolus infusion was administered at a decreasing rate (32 mL at 4 mL/sec for 8 seconds, 16 mL at 2 mL/sec for 8 seconds, and 60 mL at 1 mL/sec for 60 seconds) and was followed by a saline flush (20 mL at 1 mL/sec for 20 seconds). The rationale for the contrast material–infusion protocol was to maintain a more constant intravascular concentration of contrast material, to minimize the concentration gradient between the intravascular and extravascular spaces, to optimize conditions for Patlak analysis, and to improve signal-to-noise ratio.
The single-level bolus-tracking study at the level of the aortic arch was performed at the same time as contrast agent administration. The perfusion study was triggered when peak aortic enhancement was identified on the bolus-tracking enhancement graphs visualized on the console screen. This perfusion study, which encompassed the entire tumor, consisted of eight breath-hold helical acquisitions in expiration with the following parameters: 80 kV; 120 mAs; table feed, 30 mm; rotation time, 0.5 second; detector width, 1.5 mm; reconstruction width, 2 mm; field of view, 500 mm; matrix, 512 mm. Eighty kilovolts was chosen in preference to the standard 120 kV to minimize radiation dose, as available data from the cranial circulation show no substantial change in image quality at 80 kV despite a lower dose (6). A further advantage of using 80 kV for perfusion assessment is the better x-ray absorption of iodine at a lower kilovoltage. The patient was asked to breathe in and out and then was asked to stop breathing between each acquisition. Total dynamic acquisition time varied from patient to patient but was approximately 90 seconds. Within 24 hours, this perfusion study was repeated by using identical technical parameters to allow assessment of reproducibility.
Data Postprocessing and Analysis
Data were transferred to a dedicated workstation (Leonardo; Siemens Medical Solutions). There were images from 20 perfusion studies for evaluation (10 patients, two studies each). Evaluation was performed initially by one observer (Q.S.N.) with 2 years of experience in chest CT. Each perfusion study consisted of nine helical studies (one baseline study and eight contrast-enhanced studies) that required postprocessing prior to quantitative perfusion analysis. For each study, the 2-mm transverse images were reformatted into 10-mm transverse sections that overlapped by 5 mm by using three-dimensional software (3D analysis; Siemens Medical Solutions) to permit section-by-section analysis of the whole tumor within a clinically acceptable time of 15 minutes.
These reformatted scans were checked by an observer (Q.S.N.) to ensure that the entire tumor had been included and, by comparing the position of the tumor to those of adjacent anatomic structures, that each of these reformatted transverse images corresponded to a similar position along the z-axis of the patient at all nine studies. Each reformatted 10-mm transverse image from a similar position along the z-axis of the patient from each of the nine helical studies was then saved as a separate series on the workstation for further analysis (Fig 1). Thus, depending on the tumor size, a number of series were obtained for every patient for each of the two dynamic studies to encompass the entire tumor; each series consisted of one baseline nonenhanced image and eight subsequent dynamic contrast-enhanced transverse images at the same tumor level.
View larger version (102K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1: CT series shows reformatted nonenhanced image and eight contrast-enhanced images from a single 10-mm transverse level; all images are at similar position along z-axis of patient.
|
|
For each patient, the series of reformatted dynamic images that encompassed the whole tumor were loaded into prototype software (Functional CT; Siemens Medical Solutions). The arterial input was determined from the bolus-tracking images for each patient. A circular ROI of mean area 1.5 cm2 was placed within the aorta by using an electronic cursor and the mouse by the observer (Q.S.N.). An arterial time-attenuation curve was generated automatically, and this arterial input information was saved automatically by the software for subsequent calculation of permeability and blood volume by using Patlak analysis (7).
An ROI was drawn freehand around the tumor by an observer (Q.S.N.) using an electronic cursor and mouse, and care was taken to exclude surrounding air and atelectatic lung tissue. Subsequent software processing was automated: For each pixel within the selected ROI, the change in Hounsfield units over time after contrast material administration was transformed by using Patlak analysis to generate colored parametric maps of permeability (milliliters per 100 mL per minute) and relative blood volume (milliliters per 100 mL) for the tumor ROI (Fig 2). Each pixel location within the functional map corresponded to a single quantitative perfusion value that resulted from the mathematical calculation of the data at that location (ie, analysis was on a pixel-by-pixel basis).
View larger version (46K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2: Left: CT image from a single 10-mm transverse section shows ROI drawn freehand around lung tumor by using mouse and electronic cursor. Right: Corresponding colored parametric map shows tumor permeability.
|
|
This process was repeated for each contiguous transverse level, until the entire tumor had been covered (Fig 3). A global value representing the perfusion parameter for the entire tumor was calculated by taking the median perfusion value of all individual pixels involved. This entire process was repeated for the second perfusion study to allow assessment of reproducibility. To determine inter- and intraobserver variability, images from all 20 studies were reanalyzed by a second independent observer (V.G.), who had 6 years of experience with chest CT, and were reanalyzed by both observers after 4 weeks to minimize recall bias. All 20 studies were technically adequate, included the entire tumor, and were analyzed successfully.
View larger version (120K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3: Colored parametric maps show tumor permeability values at multiple transverse levels that encompass whole lung tumor. Each pixel location within tumor ROI corresponds to a single quantitative perfusion value.
|
|
Statistical Analysis
Statistical analysis was performed by using statistical software (StatsDirect, version 2.3.8; Sale, Cheshire, England). Initial analysis was performed to confirm that the statistical assumptions required for repeatability analysis were upheld. Kendall 
was used to establish any correlation between the absolute difference and the mean; if the difference appeared to increase when the mean parameter value increased, the data were transformed by natural logarithm.
Bland-Altman statistics were used to determine the reproducibility of the two perfusion studies (8,9). The mean, standard deviation, mean difference, and 95% limits of agreement were established. Within subject coefficient of variation, repeatability coefficient, and intraclass correlation coefficient were determined also. From the repeatability coefficient, the 95% limits of change that might occur spontaneously in an individual patient, as well as for a group of 10 patients, was established. This potentially represents the degree of change required for a therapeutic response to be significant at the 5% level. Bland-Altman statistics were also performed to assess inter- and intraobserver agreement. The mean difference, 95% limits of agreement, within subject coefficient of variation, and intraclass correlation coefficient were assessed.
|
RESULTS
|
|---|
Tumor permeability values were transformed by natural logarithm because the difference in mean values increased when the mean parameter value increased (Kendall , P < .02). Both permeability and blood volume (Table 1; Figs 4, 5) showed good agreement between studies as demonstrated by a coefficient of variation of 9.49% for permeability and 26.31% for blood volume. Inter- and intraobserver agreement (Table 2) was good for both parameters as indicated by a coefficient of variation between 3.30% and 6.34% and by an intraclass correlation coefficient of greater than 0.97.
View this table:
[in this window]
[in a new window]
|
Table 1. Reproducibility Statistics of Whole Tumor Permeability and Blood Volume Measurements Calculated from Paired Perfusion CT Scans Obtained in Patients with Lung Cancer
|
|
View larger version (14K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4a: (a) Scatter plot of permeability measurements from the two studies with line of perfect agreement, which represents line all points would lie on if both measurements gave the same reading. Plot provides visual demonstration of how reproducible the two sets of measurements are. The closer the measurements lie to this line of equality, the better the agreement. (b) Bland-Altman agreement plot of difference between the two studies against mean of permeability median values from the two studies. Mean difference is indicated by solid line. Two outer dotted lines represent 95% limits of agreement, which define range within which most differences between repeated permeability measurements made on the same subject will lie. There was an obvious outlier, which we cannot explain by any technical factors; the outlier may be related to intrinsic tumor perfusion variability in this patient.
|
|
View larger version (20K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4b: (a) Scatter plot of permeability measurements from the two studies with line of perfect agreement, which represents line all points would lie on if both measurements gave the same reading. Plot provides visual demonstration of how reproducible the two sets of measurements are. The closer the measurements lie to this line of equality, the better the agreement. (b) Bland-Altman agreement plot of difference between the two studies against mean of permeability median values from the two studies. Mean difference is indicated by solid line. Two outer dotted lines represent 95% limits of agreement, which define range within which most differences between repeated permeability measurements made on the same subject will lie. There was an obvious outlier, which we cannot explain by any technical factors; the outlier may be related to intrinsic tumor perfusion variability in this patient.
|
|
View larger version (11K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5a: (a) Scatter plot of blood volume measurements from the two studies, with line of perfect agreement, which represents line all points would lie on if both measurements gave the same reading. Plot provides visual demonstration of how reproducible the two sets of measurements are. The closer the measurements lie to this line of equality, the better the agreement. (b) Bland-Altman agreement plot of difference between the two scans against mean of blood volume median values from the two scans. Mean difference is indicated by solid line. Two outer dotted lines represent 95% limits of agreement, which define range within which most differences between repeated permeability measurements made on the same subject will lie. There was an obvious outlier, which we cannot explain by any technical factors; the outlier may be related to intrinsic tumor perfusion variability in this patient.
|
|
View larger version (20K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5b: (a) Scatter plot of blood volume measurements from the two studies, with line of perfect agreement, which represents line all points would lie on if both measurements gave the same reading. Plot provides visual demonstration of how reproducible the two sets of measurements are. The closer the measurements lie to this line of equality, the better the agreement. (b) Bland-Altman agreement plot of difference between the two scans against mean of blood volume median values from the two scans. Mean difference is indicated by solid line. Two outer dotted lines represent 95% limits of agreement, which define range within which most differences between repeated permeability measurements made on the same subject will lie. There was an obvious outlier, which we cannot explain by any technical factors; the outlier may be related to intrinsic tumor perfusion variability in this patient.
|
|
View this table:
[in this window]
[in a new window]
|
Table 2. Inter- and Intraobserver Agreement for Whole Tumor Permeability and Blood Volume Measurements Calculated from Paired Perfusion CT Scans Obtained in Patients with Lung Cancer
|
|
As well as demonstrating acceptable reproducibility, our results could be used to assess whether therapeutic response is meaningful. Any increase in whole tumor permeability over 28.53% or a reduction of more than 22.20% as a consequence of treatment would be over intrinsic measurement variability and would be a statistically significant difference for a single patient (ie, the probability of this change being random is less than 5%). Correspondingly, for a group of 10 patients, as in this cohort, an increase in whole tumor permeability over 8.27% or a reduction of more than 7.64% would be a statistically significant difference. Similarly, a change in blood volume of more than ±72.88% for a single patient or a change in blood volume of more than ±23.06% for a group of 10 patients would be a statistically significant difference.
|
DISCUSSION
|
|---|
Clinical assessment of vascular targeting and antiangiogenic drugs has highlighted problems with use of conventional response criteria to quantify therapeutic effect, as tumor shrinkage may not occur and therapeutic effect may be underestimated with these drugs (9–12). Consequently, functional assessment has been advocated, as direct quantification of tumor vascular function is possible (13). CT is an attractive modality, as commercial software makes functional assessment clinically accessible. Moreover, CT perfusion measurements have been correlated with histologic markers of angiogenesis in lung cancer (14,15).
CT perfusion measurements are intrinsically variable because of a combination of internal and external factors, which include cardiac output, technique variability, observer variability, and tumor heterogeneity. Although it may be possible to compensate for some external factors (eg, observer variability by using the same observer), it is more difficult to compensate for intrinsic variation in tumor perfusion. Tumor vasculature exhibits both spatial and temporal heterogeneity (16,17). Although temporal variability is unavoidable, whole tumor assessment may compensate for spatial variability and potentially improve reproducibility. To date there have been few data on tumor measurement reproducibility at CT, which is clearly relevant to therapeutic assessment in which a magnitude of change is measured to assess effect.
Tumor coverage has been restricted previously to, depending on detector configuration, a number of contiguous sections at a single transverse level. For example, with a 16–detector row CT scanner, a number of contiguous transverse images can be acquired, which gives a maximum composite coverage of 2.4 cm. This coverage has been problematic for quantitative assessment of perfusion in lung lesions, not least because of image misregistration from respiratory motion during image acquisition. For example, Miles et al reported that six of 16 CT perfusion studies could not be analyzed because of respiratory motion (18).
With our described helical acquisition technique, it was possible to assess the entire tumor, and analysis was successful in all 20 studies. We found that both permeability and blood volume measurements were reproducible, although permeability measurements were less variable than blood volume measurements. This lack of variability may be related to the mathematical basis of Patlak analysis (7), which has also been used in other perfusion CT applications (eg, to assess kidney function and lymphoma disease activity) and in nuclear medicine studies (19,20). Please refer to the Appendix for further information on this analysis method.
Our results are comparable to published CT data of reproducibility for both animal and human subjects (21,22). Our variation coefficients of 9.49% for permeability and 26.31% for blood volume are similar to variability in cranial perfusion of 12%–35% in dogs (21) and of 13%–33% in rabbits (22) and in tumor perfusion of 14%–24% in rabbits (23). Indeed, our results compare favorably with reproducibility of single-level dynamic contrast-enhanced MR imaging in a variety of human tumors (24) in which the coefficient of variation in log10 Ktrans, a measurement akin to permeability, was 24%.
Our results also demonstrate that the variability is within the accepted limits for therapeutic agents currently being assessed in clinical trials. In a phase IB study that used this technique, consistent increases in tumor permeability after radiation therapy and reduction in blood volume after administration of combretastatin A4 phosphate were demonstrated (25). In a completed phase I study of combretastatin A4 phosphate assessed by using single-level dynamic MR imaging, mean log10 Ktrans value was reduced by 37% for nine patients after drug administration (4). Therefore, the level of therapeutic change seen is greater than the measurement variability of our whole tumor technique. Furthermore, there was good agreement between and within observers, and our results compare favorably with those of previous studies of observer variability in which intraclass correlations greater than 0.75 were achieved (26,27); this bodes well for clinical utility of this technique.
Although we have shown that this technique is reproducible and that the level of reproducibility is in line with those in the current literature, whole tumor evaluation has several limitations as a clinical technique. First, Patlak analysis–based software is not yet commercially available, though this is likely to change as manufacturers implement this analysis method. Second, this technique does not permit whole tumor blood flow measurement, because the greater temporal resolution required (eg, one dynamic image per second) is not achievable with current scanners; this may change with technologic advances. Third, although perfusion CT has demonstrated proof of principle in the evaluation of vascular-modulating drugs (2,25), to our knowledge changes in perfusion CT measurements have not been correlated with a patient's likelihood of treatment success and have not been shown to predict survival. Finally, this was a single-center study, and these results may not have been so readily achieved in a multicenter setting. However, quality assurance and baseline reproducibility studies to assess intrapatient measurement reproducibility prior to treatment may address the problem of using this technique in a multicenter setting.
In summary, we have developed a helical multi–detector row CT technique that permits quantitative whole tumor perfusion assessment and with which good measurement reproducibility can be achieved. Whole tumor assessment is a step in the direction of improving the reliability of tumor perfusion measurements, and further evaluation of this technique is ongoing.
|
APPENDIX
|
|---|
Patlak analysis uses a two-compartment model and describes the one-way transfer of contrast material from the intravascular space to the extracellular space. At any time point, the tissue concentration of contrast material is equivalent to the sum of the intravascular and extravascular concentrations of contrast material as denoted by the following equation: Ct = rbv · bt + K · 
bt · dt, where Ct is the concentration of contrast material within the tissue, rbv is the relative blood volume of the tissue, bt is the concentration of contrast material in blood, and K is the permeability or transfer constant. Dividing the equation by bt produces the linear equation Ct/bt = rbv + K · bt · dt/bt. By plotting this graphically, permeability, or K, can be derived from the gradient of the slope of this line, and relative blood volume can be derived from the y-intercept (Fig A1). Thus, minimal changes in the gradient will produce greater changes in the y-intercept, and relative blood volume measurements will show greater variability when they are compared with permeability.
|
ADVANCES IN KNOWLEDGE
|
|---|
- Reproducibility of a whole tumor quantitative perfusion measurement technique with CT was demonstrated.
- Range of permeability and blood volume measurements in lung cancer and the percentage change in permeability and blood volume needed to show a significant response at the 5% level were documented.
|
FOOTNOTES
|
|---|
Abbreviations: ROI = region of interest
See Materials and Methods for pertinent disclosures.
Author contributions: Guarantor of integrity of entire study, V.G.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, Q.S.N., V.G.; clinical studies, Q.S.N., V.G., P.F.; statistical analysis, Q.S.N., V.G., A.R.P.; and manuscript editing, Q.S.N., V.G., H.F., E.K., A.R.P.
|
References
|
|---|
- Eskens FA. Angiogenesis inhibitors in clinical development: where are we now and where are we going? Br J Cancer 2004;90:1–7.[CrossRef][Medline]
- Willett CG, Boucher Y, di Tomaso E, et al. Direct evidence that the VEGF-specific antibody bevacizumab has antivascular effects in human rectal cancer. Nat Med 2004;10:145–147.[CrossRef][Medline]
- Anderson HL, Yap JT, Miller MP, Robbins A, Jones T, Price PM. Assessment of pharmacodynamic vascular response in a phase I trial of combretastatin A4 phosphate. J Clin Oncol 2003;21:2823–2830.[Abstract/Free Full Text]
- Galbraith SM, Maxwell RJ, Lodge MA, et al. Combretastatin A4 phosphate has tumor antivascular activity in rat and man as demonstrated by dynamic magnetic resonance imaging. J Clin Oncol 2003;21:2831–2842.[Abstract/Free Full Text]
- Stevenson JP, Rosen M, Sun W, et al. Phase I trial of the antivascular agent combretastatin A4 phosphate on a 5-day schedule to patients with cancer: magnetic resonance imaging evidence for altered tumor blood flow. J Clin Oncol 2003;21:4428–4438.[Abstract/Free Full Text]
- Wintermark M, Maeder P, Verdun FR, et al. Using 80 kVp versus 120 kVp in perfusion CT measurement of regional cerebral blood flow. AJNR Am J Neuroradiol 2000;21:1881–1884.[Abstract/Free Full Text]
- Patlak CS, Balsberg RG, Fernstmacher JD. Graphical evaluation of blood-to-brain transfer constants from multiple-time uptake data. J Cereb Blood Flow Metab 1983;3:1–7.[Medline]
- Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–310.[CrossRef][Medline]
- Bland JM, Altman DG. Applying the right statistics: analyses of measurement studies. Ultrasound Obstet Gynecol 2003;22:85–93.[Medline]
- World Health Organization. WHO hand book for reporting results of cancer treatment. WHO offset publication no. 48. Geneva, Switzerland: World Health Organization, 1979.
- Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to evaluate the response to treatment in solid tumors. J Natl Cancer Inst 2000;92:205–216.[Abstract/Free Full Text]
- Kerbel R, Folkman J. Clinical translation of angiogenesis inhibitors. Nat Rev Cancer 2002;2:727–739.[CrossRef][Medline]
- Miles KA, Charnsangavej C, Lee FT, Fishman EK, Horton K, Lee TY. Application of CT in the investigation of angiogenesis in oncology. Acad Radiol 2000;7:840–850.[CrossRef][Medline]
- Yi CA, Lee KS, Kim EA, et al. Solitary pulmonary nodules: dynamic enhanced multi-detector row CT study and comparison with vascular endothelial growth factor and microvessel density. Radiology 2004;233:191–199.[Abstract/Free Full Text]
- Tateishi U, Nishihara H, Tsukamoto E, Morikawa T, Tamaki N, Miyasaka K. Lung tumors evaluated with FDG-PET and dynamic CT: the relationship between vascular density and glucose metabolism. J Comput Assist Tomogr 2002;26:185–190.[CrossRef][Medline]
- Chaplin DJ, Hill SA. Temporal heterogeneity in microregional erythrocyte flux in experimental solid tumors. Br J Cancer 1995;71:1210–1213.[Medline]
- Kimura H, Braun RD, Ong ET, et al. Fluctuations in red cell flux in tumor microvessels can lead to transient hypoxia and reoxygenation in tumor parenchyma. Cancer Res 1996;56:5522–5528.[Abstract/Free Full Text]
- Miles KA, Griffiths MR, Fuentes MA. Standardized perfusion value: universal CT contrast enhancement scale that correlates with FDG-PET in lung nodules. Radiology 2001;220:548–553.[Abstract/Free Full Text]
- Miles KA, Kelley BB. CT measurements of capillary permeability within nodal masses: a potential technique for assessing the activity of lymphoma. Br J Radiol 1997;70:74–79.[Abstract]
- Miles KA, Leggett DA, Bennett GA. CT derived Patlak images of the human kidney. Br J Radiol 1999;72:153–158.[Abstract]
- Nabavi DG, Cenic A, Dool J, et al. Quantitative assessment of cerebral hemodynamics using CT: stability, accuracy, and precision studies in dogs. J Comput Assist Tomogr 1999;23:506–515.[CrossRef][Medline]
- Cenic A, Nabavi DG, Craen RA, Gelb AW, Lee TY. A CT method to measure hemodynamics in brain tumors: validation and application of cerebral blood flow maps. AJNR Am J Neuroradiol 2000;21:462–470.[Abstract/Free Full Text]
- Purdie TG, Henderson E, Lee TY. Functional CT imaging of angiogenesis in rabbit VX2 soft tissue tumour. Phys Med Biol 2001;46:3161–3175.[CrossRef][Medline]
- Galbraith SM, Lodge MA, Taylor NJ, et al. Reproducibility of dynamic contrast-enhanced MRI in human muscle and tumors: comparison of quantitative and semi-quantitative analysis. NMR Biomed 2002;15:132–142.[CrossRef][Medline]
- Ng QS, Goh V, Hoskin PJ, Carnell D, Meer K, Padhani A. Tumor antivascular effects of combretastatin A4 phosphate (CA4P) in combination with radiotherapy measured using dynamic contrast enhanced computed tomography (DCE-CT) [abstract]. Radiother Oncol 2004;73(suppl 1):322.
- Fiorella D, Heiserman J, Prenger E, Partovi S. Assessment of the reproducibility of postprocessing dynamic CT perfusion data. AJNR Am J Neuroradiol 2004;25:97–107.[Abstract/Free Full Text]
- Goh V, Halligan S, Hugill JA, Bassett P, Bartram CI. Quantitative assessment of colorectal perfusion using MDCT: inter- and intra observer agreement. AJR Am J Roentgenol 2005;185:225–231.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
G. Petralia, L. Preda, S. Raimondi, G. D'Andrea, P. Summers, G. Giugliano, F. Chiesa, and M. Bellomi
Intra- and Interobserver Agreement and Impact of Arterial Input Selection in Perfusion CT Measurements Performed in Squamous Cell Carcinoma of the Upper Aerodigestive Tract
AJNR Am. J. Neuroradiol.,
June 1, 2009;
30(6):
1107 - 1115.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
M.-S. Park, E. Klotz, M.-J. Kim, S. Y. Song, S. W. Park, S.-W. Cha, J. S. Lim, J. Seong, J. B. Chung, and K. W. Kim
Perfusion CT: Noninvasive Surrogate Marker for Stratification of Pancreatic Cancer Response to Concurrent Chemo- and Radiation Therapy
Radiology,
January 1, 2009;
250(1):
110 - 117.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
V. Goh, S. Halligan, A. Gharpuray, D. Wellsted, J. Sundin, and C. I. Bartram
Quantitative Assessment of Colorectal Cancer Tumor Vascular Parameters by Using Perfusion CT: Influence of Tumor Region of Interest
Radiology,
June 1, 2008;
247(3):
726 - 732.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
D. A. Torigian, S. S. Huang, M. Houseni, and A. Alavi
Functional Imaging of Cancer with Emphasis on Molecular Techniques
CA Cancer J Clin,
July 1, 2007;
57(4):
206 - 224.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. van den Hoff, V. Goh, E. Klotz, and Q. S. Ng
Assessment of Lung Cancer Perfusion by Using Patlak Analysis: What Do We Measure?
Radiology,
June 1, 2007;
243(3):
907 - 908.
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
