| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Paul Strickland Scanner Centre, Mt. Vernon Hospital, Northwood, United
Kingdom.
2 Intestinal Imaging Centre, St. Mark's Hospital, Middlesex, HA1 3UJ, United
Kingdom.
3 Present address: Specialist Radiology, University College Hospital, 235 Euston
Road, London, NW1 2BU, United Kingdom.
Received January 10, 2005;
accepted after revision April 27, 2005.
This research was supported in part by a grant from the Royal College of
Radiologists, London, United Kingdom.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Ten patients (mean age, 67 years; six men, four women) with histologically proven colorectal cancer were examined prospectively using 4-MDCT. Perfusion studies (cine mode; 4 x 5 mm collimation; 1 acquisition/s; 65 seconds total) were performed through the tumor epicenter after IV bolus contrast administration (iopamidol 340, 100 mL; 5 mL/s) and repeated within 48 hours. Quantitative values for blood volume, blood flow, mean transit time, and permeability were determined using commercial software. Two regions of interest were studied on the axial image: one within the tumor and another within the left gluteal muscle. Measurement reproducibility was assessed using Bland-Altman statistics.
RESULTS. For the tumor, the mean difference (95% limits of agreement) was -0.04 mL/100 g tissue (-2.50, 2.42); 8.80 (-50.5, 68.0) mL/100 g tissue/min; -0.99 (-8.19, 6.20) seconds; and 1.20 (-5.42, 7.83) mL/100 g tissue/min for blood volume, blood flow, mean transit time, and permeability, respectively. For muscle, the mean difference (95% limits of agreement) was 0.02 (-1.40, 1.43), 6.60 (-11.2, 24.3), -3.76 (-16.87, 9.35), and 1.30 (-4.68, 7.28), respectively.
CONCLUSION. Quantitative perfusion measurements are reproducible. Measurements from tumor are less variable than from skeletal muscle.
Keywords: cancer • colon • colorectal cancer • dynamic CT • MDCT • perfusion CT
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Assessment of measurement variability is essential for therapeutic monitoring. The value of any quantitative perfusion measurement actually reflects a combination of factors: the true measurement value, the intrinsic variability of the method used to make the measurement, biologic variation such as cardiac output, and observer variability [4]. For these reasons, an assessment of baseline measurement variability is necessary so that a distinction can be made between measurement error and a true therapeutic response. This may be achieved by assessment of the intrinsic variability found in healthy tissue (e.g., skeletal muscle) as distinct from any tumor [5]. Furthermore, although it may be possible to compensate for variations in physiologic factors such as cardiac output by normalizing tumor measurements with reference to uninvolved tissue [5, 6], such a procedure assumes that healthy tissue variability is low.
We wanted to determine the reproducibility of colorectal cancer perfusion measurements and to compare them with those obtained from healthy skeletal muscle, in an attempt to quantify variability and so define the limits likely to signify a therapeutic response.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CT Scanning
After a 4-hour fast, 1,000 mL of a water-soluble contrast agent (2-4%
meglumine and sodium diatrizoate, Gastrografin, Bracco) was ingested to
opacify the small bowel 30 minutes before CT according to normal practice in
our institution. With the patient lying supine on the CT scanner table, an
18-gauge venous cannula was sited in the right antecubital fossa. The
spasmolytic hyoscine N-butylscopolammonium bromide (20 mg Buscopan;
Boehringer Ingelheim) was administered IV to abolish bowel peristalsis.
Patient movement was also minimized by placing a restraining band around the
abdomen.
|
|
For the dynamic study, a pump injector (Percu-Pump Touch Screen, EZ-EM) was used to inject IV 100 mL of iopamidol 340 (Niopam 340, Bracco) at a rate of 5 mL/s. Four contiguous slices, each collimated to 5 mm, were obtained at 1-second intervals through the midpoint of the tumor using a cine mode acquisition (120 kV; 60 mA; scanned FOV, 50 cm; matrix, 512 x 512 mm; effective dose, 8 mSv [7]). Acquisition commenced 5 seconds after the start of IV injection to allow acquisition of a baseline unenhanced image and continued for 65 seconds.
The dynamic study was followed by a diagnostic portal venous phase abdominopelvic study that started 75 seconds after the commencement of IV injection using the following parameters: slice thickness, 5 mm; interval, 2.5 mm; HiSpeed mode (22.5 mm/s); pitch, 1.5; 120 kV; 280 mA; 0.6-second rotation speed; scanned FOV, 50 cm; and matrix, 512 x 512 mm. This diagnostic study was used to determine the local and distant stage of the tumor, and a radiologic report was issued on the basis of these images according to day-to-day clinical practice.
All patients returned within 48 hours of the initial study for a second dynamic study to assess measurement reproducibility. Only the unenhanced planning scan and 65-second dynamic study were performed on the second occasion. The second unenhanced planning scan was compared with that used for the initial study so the tumor level examined for each study was similar. IV spasmolytic and contrast agent were administered exactly as previously, and data were acquired using technical parameters identical to the initial dynamic study.
Image Analysis
Image analysis was performed by a single radiologist experienced in
perfusion analysis. All 20 dynamic studies (10 patients; two paired studies
each) were analyzed on a stand-alone workstation (Advantage 4.1, GE Healthcare
Technologies) using commercial software based on deconvolution analysis (Body
Tumor Perfusion 3.0, GE Healthcare Technologies).
|
|
|
|
The arterial input was determined by selecting a circular region of interest (ROI) from the control panel and placing this freehand using a mouse within either the iliac or femoral arteries, whichever was best visualized in the chosen slice plane. Change in arterial attenuation over the 65-second acquisition was determined via automatic generation of an arterial input curve along with perfusion maps for all of the tissues within the scan plane, within the processing threshold selected. Four perfusion maps were generated: blood volume, blood flow, mean transit time, and permeability. In an attempt to reduce confounding, the ROI used for the arterial input was saved by the software so the exact same-size ROI could be placed automatically in the same location in subsequent analyses.
An ROI was then drawn freehand around the peripheral margin of the tumor using an electronic cursor and mouse. Care was taken to exclude perirectal or pericolonic fat and also intraluminal gas. To ensure the tumor ROI remained within the tumor boundaries on serial images, a cine loop of the serial images of the perfusion acquisition was viewed to gauge the degree and margins of patient movement during acquisition. A time-attenuation curve for the selected tumor tissue and the four perfusion parameters within the tumor ROI were then derived automatically (Figs. 1A, 1B, 1C, 1D, 1E, and 1F). Mean values for the four tumor perfusion parameters—blood volume, blood flow, mean transit time, and permeability—for each individual patient were recorded.
Another circular ROI was selected and placed freehand using the mouse within the left gluteal muscle (which was also included within the selected 5-mm axial image), so mean values for skeletal muscle blood volume, blood flow, mean transit time, and permeability could be derived. Measurements from skeletal muscle were recorded for each individual patient. As for the arterial input ROI, the skeletal muscle ROI was saved using the software so it could be used again.
Image analysis was performed by the same investigator for the second set of dynamic scans obtained from each of the 10 patients. Analysis was performed exactly as previously, notably selecting the same 5-mm axial image used for the initial analysis and recalling the ROIs saved by the software.
Statistical Analysis
The mean and SD of blood volume, blood flow, mean transit time, and
permeability measurements for both tumor and muscle were determined for each
of the paired sets of studies. Measurement reproducibility was assessed using
a variety of methods.
Agreement between replicate measurements was assessed using the Bland-Altman method [8]. The mean difference, SD of the differences, and 95% limits of agreement (i.e., mean difference - 2 x SD and mean difference + 2 x SD) were calculated for each of the four perfusion parameters for both tumor and skeletal muscle.
Measurement error and repeatability were assessed for each of the four
perfusion parameters. An estimate of the precision of the measurement was
derived by calculating the within-subject SD, using the formula dSD / 
2
(where dSD is the square root of the mean squared difference).
Measurement error relative to the size of the parameters was estimated via the within-subject coefficient of variation, which were calculated by dividing the within-subject SD by the overall mean of the parameter studied and then expressed as a percentage.
The repeatability coefficient, which represents the threshold value below
which the absolute differences between two measurements on the same patient is
expected to lie for 95% of the measurement pairs, was assessed using the
formula 1.96 x dSD, and the 95% confidence limits for spontaneous change
for an individual patient were calculated using the formula (1.96 x dSD)
/ n, where n = 1, and expressed as a percentage of the mean
parameter value [4].
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Table 1 summarizes the mean difference, SD of the differences, and 95% limits of agreement for the two sets of tumor measurements. Table 2 summarizes the within-subject SD, within-subject coefficient of variation, and repeatability coefficient. Thus, for a single patient undergoing therapy, a measurement change of at least ± 38%, 65%, 97%, and 48% is required to be significant at the 5% level for blood volume, blood flow, mean transit time, and permeability, respectively.
|
|
Skeletal Muscle Measurements
The mean (SD) values for intramuscular blood volume, blood flow, mean
transit time, and permeability were 1.8 (0.8) mL/100 g tissue, 14.1 (11.2)
mL/100 g tissue/min, 19.9 (9.8) seconds, and 4.8 (3.4) mL/100 g tissue/min,
respectively, for the first set of measurements, and 1.8 (0.9) mL/100 g
tissue, 7.6 (6.0) mL/100 g tissue/min, 23.7 (7.7) seconds, and 3.5 (2.6)
mL/100 g tissue/min, respectively, for the second set of measurements.
Table 1 summarizes the mean
difference, SD, and 95% limits of agreement.
Table 2 shows the
within-subject SD, within-subject coefficient of variation, and repeatability
coefficient.
Skeletal muscle measurements, in particular blood volume, blood flow, and permeability, showed a greater variability compared with those obtained from tumor. Measurement error relative to the size of the parameter studied (within-subject coefficient of variation) was larger for skeletal muscle than for tumor. For example, for blood volume measurements, the within-subject coefficient of variation was 14% for tumor but 27% for muscle (Table 2).
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Functional CT measurements are attractive because they directly quantify tumor perfusion and are increasingly accessible. However, although extensively assessed in the cranial circulation [11-15], at the time of writing, little published data exist relating to measurement variability elsewhere. Reproducibility and measurement error are important considerations when considering therapeutic response because this will be contingent on a change in magnitude of the measurements made. It follows that the difference between repeated baseline measurements should be relatively small so they are overwhelmed by any changes in perfusion as a consequence of therapy. If not, measurements may merely represent the so-called noise inherent to the technique rather than reflecting a true therapeutic response [16]. Indeed, clinical therapeutic trials that rely on functional measurements now advocate assessment of baseline reproducibility to ensure that any monitored change is really caused by the therapeutic agent being examined [17].
A variety of factors contribute to measurement variability, which can be divided broadly into those that are intrinsic and those that are extrinsic. Intrinsic factors include tumor heterogeneity, physiologic components including cardiac output and volume of distribution, and the intrinsic measurement variability of the technique used. Extrinsic factors include observer variability, which reflects the experience in making these measurements and is acceptable for colorectal cancer perfusion measurement [18], and variability related to acquisition parameters including kilovoltage, milliampere, and slice collimation.
The 95% limits of agreement that we used provide an overall measure of reproducibility, or variability, that directly reflects clinical interpretation. These limits represent the boundaries within which the true measurement is expected to lie 95% of the time; the narrower the limits, the more precise the measurement being made. We found that the 95% limits of agreement were acceptable for all four tumor perfusion parameters (blood volume, blood flow, mean transit time, and permeability) and acceptable for day-to-day therapeutic assessment. This has been achieved in part by optimizing the extrinsic factors contributing to variability. Scans were supervised directly by a radiologist experienced in functional CT; the same radiographic technicians performed each scan; identical scan parameters were used; a pump injector was used to administer the IV contrast agent via the same-bore cannula placed in the same vein; a single experienced observer analyzed both sets of data, and identical ROIs were used whenever possible.
The within-subject SD, within-subject coefficient of variation, and repeatability coefficient provide complementary information. The repeatability coefficient, expressed as a percentage of the mean, provides a measure of values over and above that which measured change can be regarded as significant in therapeutic assessment and does not merely reflect measurement error. For example, our results predict that a measured change equal to or greater than 38% in blood volume, 65% in blood flow, 97% in transit time, and 48% in permeability would reflect a real change in a single patient because of therapy rather than just reflecting measurement noise. These data also indicate that tumor blood volume and permeability may be the least variable parameters.
We found that the 95% limits of agreement, within-subject SD, within-subject coefficient of variation, and repeatability for skeletal muscle tissue indicate considerable variation in these measurements, more so than for tumor, in particular for blood volume, blood flow, and permeability. This may seem surprising but is actually comparable to reproducibility data from dynamic MRI, where ktrans, a measurement akin to permeability, is known to be more variable in muscle than tumor [5, 17]. The lower mean and narrower range of skeletal muscle perfusion measurements compared with tumor may be contributory because identical changes in perfusion for tumor and skeletal muscle elicit a greater relative change in the latter. Skeletal muscle blood flow is also variable as a result of physiologic regulatory mechanisms [19] and is affected, for example, by the degree of exercise. This may have contributed to the greater variability in skeletal muscle measurements, although the level of exercise should have been similar before both scans in each patient because the patients were seated in the waiting room for an hour before scanning. It has been suggested that normalizing tumor perfusion measurements by reference to healthy tissue (e.g., skeletal muscle) may help compensate for physiologic factors such as cardiac output and blood pressure and thus make the absolute measurement more reliable. However, our results indicate that skeletal muscle is unsuitable because of its wide measurement variability.
Measurement variability of 15-30% has been reported for perfusion CT in the cranial circulation of both animals and humans [20]. The cranial circulation is ideal for the mathematic modeling used for quantitative perfusion measurement because of its unique blood-brain barrier. A relative lack of movement also reduces measurement error. Nabavi et al. [11] reported a variability of 35% and 31%, respectively, for white and gray matter in dogs; Cenic et al. [12, 13] reported a variability of 32.5% in healthy rabbits and 13.2% in rabbits with intracranial tumors. Gillard et al. [14] measured cranial perfusion from the middle cerebral artery territory on subsequent days in seven patients with glioma undergoing radiation therapy. Linear correlation between the two measurements was good, with a Pearson coefficient of 0.88. However, it should be kept in mind that simple correlation does not confer agreement. Rather, it is a measure of linear association, and high correlation can be present when agreement is actually poor [8, 16].
Reproducibility studies have been performed for a variety of tumors using dynamic contrast-enhanced MRI as opposed to CT. Jackson et al. [21] reported a coefficient of variation of 7.7% in ktrans, a measurement akin to permeability, in nine patients with cerebral glioma. In a study of a variety of extracranial tumors, Galbraith et al. [17] found that a change of -45% to 83% in ktrans was required for significance in a single patient, with a coefficient of variation of 29%. Our MDCT data are similar to, if not better than, that achieved in these MR studies. This is perhaps unsurprising because quantification with MRI is technically challenging because artifact influences arterial input measurement, and there is lack of a linear relationship between MRI signal intensity and contrast agent concentration, with an at times unpredictable magnification effect [22]. Reproducibility data from one tumor type are unlikely to be directly transferable to another, however, and further work in this area is required.
In summary, we have shown that quantitative colorectal cancer perfusion measurements using dynamic contrast-enhanced MDCT are broadly reproducible, and this reproducibility may be comparable and possibly better than dynamic contrast-enhanced MRI. We have also defined the limits of variability for colorectal cancer perfusion measurements made by MDCT, which can be used to help define the future criteria necessary for a therapeutic response.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  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. L. Buckley, V. Goh, and S. Halligan Are Measurements from Two Commercial Software Packages Interchangeable? Possibly, If Like Is Compared with Like Radiology, February 1, 2008; 246(2): 642 - 643. [Full Text] [PDF]
|
|
|
|
 |
|
 S. C. Faria, C. S. Ng, K. R. Hess, S. Phongkitkarun, J. Szejnfeld, D. Daliani, and C. Charnsangavej CT Quantification of Effects of Thalidomide in Patients with Metastatic Renal Cell Carcinoma Am. J. Roentgenol., August 1, 2007; 189(2): 378 - 385. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Goh, S. Halligan, and C. I. Bartram Quantitative Tumor Perfusion Assessment with Multidetector CT: Are Measurements from Two Commercial Software Packages Interchangeable? Radiology, March 1, 2007; 242(3): 777 - 782. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
