| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, University "Federico II," Via Pansini 5,
Via Manzoni 214/0, Napoli 80123, Italy.
2 Department of Radiology, New York University School of Medicine, New York,
NY.
Received May 1, 2004;
accepted after revision August 16, 2004.
Address correspondence to M. Imbriaco
(mimbriaco{at}hotmail.com).
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Seventy-one patients (41 men, 30 women; mean age, 63 years; range, 29-80 years) with suspected pancreatic tumor underwent MDCT. Scanning was performed on an MDCT scanner with 0.5-sec gantry rotation and acquisition of 4 slices per rotation. Unenhanced scanning was followed by one set of scanning in the caudocranial direction from the inferior hepatic margin to the diaphragm with a scanning delay of 60 sec after the IV injection of 150 mL of contrast material delivered at 3 mL/sec. Two reviewers independently scored images in a blinded fashion for the presence of tumor and assessment of resectability. Receiver operating characteristic analysis was performed.
RESULTS. A final histopathologic diagnosis derived from surgical findings was obtained in 42 patients; in the remaining 29 patients, percutaneous fine-needle aspiration biopsy coupled with a 1-year clinical follow-up to determine development of local, regional or distant neoplasm served as gold standard proof of diagnosis. Final diagnosis was pancreatic cancer in 40 patients (27 ductal adenocarcinoma, nine mucinous cystoadenocarcinoma, two neuroendocrine tumors, one lymphoma, and one papillary cystoadenocarcinoma) and chronic pancreatitis in 31. The mean tumor size was 2.4 cm (range, 4-1 cm). Values for the area under the curve (Az) for the assessment of tumor detection were 0.97 for reviewer 1 and 0.96 for reviewer 2 (p = not significant). Az values for tumor resectability were 0.90 for reviewer 1 and 0.90 for reviewer 2 (p = not significant). No statistically significant differences were observed between superior mesenteric artery and vein opacification with the hepatic parenchyma enhanced at a time closer to the peak hepatic enhancement, optimizing the detection of hepatic lesions.
CONCLUSION. Thin-section single-phase MDCT is an accurate technique for the diagnosis and assessment of resectability in patients with a suspected pancreatic neoplasm. This technique provides optimal tumor-to-pancreas contrast and maximal pancreatic parenchymal and peripancreatic vascular enhancement. It allows visualization of the entire liver and the whole upper abdomen during the portal phase for accurate identification of liver metastases and peritoneal seeding.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Single-detector helical CT is established as a highly accurate imaging technique for the diagnosis and staging of pancreatic ductal adenocarcinoma. Imaging protocols take advantage of the speed of acquisition and ability to obtain thin sections to optimize differentiation of the tumor, normal pancreatic tissue, and surrounding blood vessels. This technique has proven to be highly specific for the determination of nonresectability when it enables identification of tumor extension into the adjacent peripancreatic structures. However, the technique is not as accurate in depicting resectable tumor, with a negative predictive value of 79% [4].
Several scanning protocols have been described for pancreatic cancer imaging using helical CT technology [4-10]. Most authors recommend a dual-phase or biphasic technique in which the pancreas is scanned twice after a single injection of contrast material: once during the pancreatic parenchymal phase, which provides the best tumor contrast and good vessel opacification, and a second time during the portal venous phase to scan for liver metastases [5, 9]. In a previous article, a different helical CT protocol was proposed for studying patients with suspected pancreatic carcinoma that allows the study of the pancreas in a single phase of acquisition in the reverse direction of traditional scanning protocols [11], from the inferior hepatic margin to the diaphragm using single-detector CT technology. This technique has been shown to be highly reproducible and to have a high diagnostic accuracy in comparison with the dual-phase technique for the diagnosis and assessment of resectability in patients with suspected pancreatic carcinoma [11]. With the introduction of MDCT, it is now possible to image a larger volume of anatomy in a much shorter time, with no sacrifice in image quality.
In this prospective study, we aimed to determine the diagnostic value of single-phase acquisition in the reverse direction, from the inferior hepatic margin to the diaphragm, using MDCT for tumor detection and assessment of resectability in patients with suspected pancreatic cancer.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CT Technique
All scanning was performed on an MDCT scanner (MX8000, Marconi) with
0.5-sec gantry rotation and the acquisition of 4 slices per rotation. Ten to
fifteen minutes before the examination, patients were given 500 mL of water
for demarcation of the stomach and duodenum and delineation of the pancreatic
head region. An 18- or 20-gauge catheter was placed in an antecubital vein.
Glucagon was not administered. Unenhanced scanning of the pancreas was
initially performed to define the craniocaudal extent of the pancreas using
the following parameters: 4 x 2.5 mm detector configuration, 3-mm
reconstruction interval, pitch of 1, 120 kVp, 200 mAs, and 35-cm field of
view. Then one acquisition through the pancreas and upper abdomen was
performed using the following parameters: caudocranial direction (from the
inferior hepatic margin to the diaphragm), 4 x 1 mm detector
configuration, 1.25-mm reconstruction interval, pitch of 1, 120 kVp, 260-280
mAs, and 35-cm field of view; with a scanning delay of 60 sec after the IV
injection of 150 mL of nonionic contrast material (Ultravist [iopromide],
Schering) with an iodine content of 370 mgI/mL delivered at 3 mL/sec. Mean
scanning time from the inferior hepatic margin to the diaphragm was 18.4
± 5 sec (range, 10-28 sec). Imaging interpretation was performed
directly at a dedicated workstation (Kayak PC, Hewlett Packard) using a
software package with a volume-rendering algorithm (Vitrea 2.2, Vital Images).
Two reviewers independently evaluated the images for the presence of tumor and
assessment of resectability. In addition, attenuation values of the tumor, the
normal pancreatic parenchyma, the liver, the hepatic veins, and the superior
mesenteric artery and vein were also measured.
CT Image Interpretation
CT scans were interpreted separately and in a random order by two
experienced abdominal radiologists. At the time of interpretation, the
reviewers were unaware of findings from other imaging tests, clinical history,
or surgical or histopathologic findings. The results of the interpretation
were recorded on forms produced for the project. Observers evaluated the
abnormal findings in the pancreas using a 5-point confidence scale: 1 =
definitely benign, 2 = probably benign, 3 = indeterminate, 4 = probably
malignant, and 5 = definitely malignant. Interpretations scored as
indeterminate were considered positive in the specificity and sensitivity
calculations and therefore were attributed to definite categories. In
addition, for assessment of tumor resectability, observers scored the presence
of resectable or unresectable tumor using the following scores: 1 = definitely
resectable, 2 = probably resectable, 3 = indeterminate, 4 = probably
unresectable, and 5 = definitely unresectable.
The reviewers were told that lesions should be classified as unresectable if evidence was seen of any or all of the following findings: peripancreatic vascular invasion (defined as encasement of the celiac, hepatic, or superior mesenteric arteries or the portal or superior mesenteric vein); extrapancreatic invasion of adjacent tissues and organs other than duodenum; presence of hematogenous or distant lymph node metastases (other than peripancreatic nodes); or signs of peritoneal carcinomatosis. A vessel was considered to be involved if it showed a focal reduction in caliber, circumferential (> 180°) encasement by tumor, or frank thrombosis. Involvement of the splenic artery or veins was not considered an absolute contraindication to resection unless the tumor extended into the splenoportal confluence or involved other splanchnic vessels. A distant lymph node exceeding 1.5 cm was considered to be involved.
A tumor was defined as resectable if it was considered completely removable with surgery. Contrast-enhanced mean CT attenuation values (in Hounsfield units) of the normal pancreatic parenchyma, liver, pancreatic tumor, superior mesenteric artery, and superior mesenteric vein were obtained by means of regions of interest (ROI) analysis. These measurements were made in the normal pancreatic parenchyma using 1-cm-diameter ROIs in a region adjacent to the tumor rather than in an area remote from the tumor site where normal parenchymal enhancement may vary slightly. Attenuation values for the pancreatic tumor were obtained in hypodense portions of the tumor (cystic areas were avoided). In the hepatic parenchyma, ROI measurements were obtained at three sections: the upper, middle, and lower liver. In each section, two ROIs were measured and the mean attenuation value was subsequently calculated in the three sections. For the quantitative analysis, we considered only patients with tumors (n = 40) and did not include patients with chronic pancreatitis.
Statistical Analysis
The findings of imaging studies were classified as true-positive,
true-negative, false-positive, or false-negative for malignancy on the basis
of histopathologic results. A receiver operating characteristic (ROC) analysis
was obtained for the two radiologists who were blinded to the histopathologic
results. The ROC analysis was performed using the SPSS program (Statistical
Package for the Social Sciences), which plotted the true-positive fraction
against the likelihood of a false-positive image and calculated the value of
the area under the curve (Az) and its SE. The area under
each ROC curve was used to indicate the overall performance of the
single-phase technique for the assessment of tumor detection and resectability
for each reviewer. Factors with Az values greater than
0.80 were considered to have a good discriminatory power
[12,
13]. The areas obtained for
the two reviewers were compared using the z-test. Sensitivity,
specificity, and diagnostic accuracy were determined using standard criteria.
Confidence levels of 3-5 were regarded as positive for malignant lesion;
conversely, confidence levels of 1-2 were regarded as negative. For analysis
of interobserver variability in the detection of tumor and assessment of
resectability, analysis with weighted kappa statistics was used to measure the
degree of agreement between the two reviewers. A kappa value of less than 0.4
was considered to indicate poor agreement; 0.4-0.75, good agreement; and
greater than 0.75, excellent agreement
[14]. Comparison between
figures of merit was obtained using the McNemar test. CT attenuation values
for the normal pancreatic parenchyma, pancreatic tumor, liver, and superior
mesenteric artery and vein were compared using a paired two-tailed Student's
t test. Values for p of less than 0.05 were considered
statistically significant.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
The results for assessment of tumor detection and resectability are shown in Table 2. Sensitivities and specificities for the two reviewers for the assessment of tumor detection were, respectively, 95% and 94% for reviewer 1, and 93% and 90% for reviewer 2 (p = not significant). Sensitivities and specificities for the assessment of tumor resectability were, respectively, 94% and 89% for reviewer 1, and 90% and 78% for reviewer 2 (p = not significant). The results of ROC analysis are reported in Table 3. The Az values for the assessment of tumor detection were 0.97 for reviewer 1 and 0.96 for reviewer 2 (p = not significant). The Az values for the assessment of tumor resectability were 0.90 for reviewer 1 and 0.90 for reviewer 2 (p = not significant); ROC curves are shown in Figures 1A, and 1B. A high reproducibility was observed for single-phase MDCT, with an overall score for agreement between the two reviewers for assessment of tumor detection of 62% (44/71), with a weighted kappa value of 0.7 (or > 0.41); score agreement for assessment of tumor resectability was 63% (25/40), with a weighted kappa value of 0.7 (or > 0.41) (Table 4). Table 5 shows the quantitative measurements of enhancement of the normal pancreatic parenchyma, liver, pancreatic tumor, and superior mesenteric artery and vein in 40 patients with pancreatic tumors. The mean CT attenuation difference between the tumor and normal pancreatic parenchyma was 70 ± 3 H (mean difference ± SE) (p < 0.0001). No statistically significant differences were observed between arterial and venous peripancreatic vessel opacification, with the hepatic parenchyma enhanced at a time closer to the peak hepatic enhancement optimizing the detection of hepatic lesions. In particular, the mean hepatic CT attenuation value was 112 ± 20 H in the upper liver, 111 ± 19 H in the middle liver, and 111 ± 21 H in the lower liver, with an overall mean hepatic CT attenuation value of 111 ± 20 H. Finally, the mean CT attenuation value for the hepatic veins was 195 ± 14 H. Figures 2, 3, 4 show examples of pancreatic cancer detected using the single-phase technique.
|
|
|
|
|
|
|
|
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
A recent article suggested using a single phase of acquisition beginning at the bottom of the liver and scanning in a cranial direction so that scanning terminates at the apex of the diaphragm. Scanning begins 50 sec after the IV administration of contrast material [11]. This technique has shown comparable diagnostic accuracy with the dual-phase approach and high reproducibility. Because of the lower radiation burden and the lower cost, a single-phase "bottom-up" acquisition has been suggested as the protocol of choice when evaluating patients with suspected pancreatic cancer [11].
In 1992, Megibow [22] was the first to propose a reverse direction of the traditional scanning protocols for imaging patients with pancreatic adenocarcinoma using a dynamic incremental bolus CT technique. In that study, the first section was acquired at the level of the transverse duodenum (superior margin of L3), and subsequent sections were acquired as the patient moved in a caudocephalad direction. This protocol ensured that data acquisition occurred when the major arteries (celiac, hepatic, splenic, superior mesenteric, gastroduodenal, pancreaticoduodenal) and veins (portal, splenic, mesenteric) were intensely opacified. Furthermore, the pancreas itself was intensely enhanced [22].
To our knowledge, this protocol using an MDCT technique has never been proposed before. The technique used in our study offers several advantages over conventional dual-phase helical CT. The imaging protocol is less complicated and time-consuming for the technologist in terms of both image acquisition and recording on film. Film costs, demands on the equipment, and radiation dose to the patients are also reduced. In addition, the technique allows good opacification of both venous and arterial peripancreatic structures. Finally, another important advantage of our protocol is that the liver and the whole upper abdomen can be fully evaluated during the portal phase, which is known to be the optimal temporal window for identification of liver metastases and other ancillary tumor findings such as ascites in patients with pancreatic cancer.
In a previous article, Tublin et al. [23] observed that both the volume and the rate of contrast material administration play an important role in determining the degree of pancreatic enhancement. In particular, it was shown that both the peak enhancement and the time to peak enhancement of the pancreas are directly related to the rate of contrast material injection. In that study, at an injection rate of 2.5 mL/sec, peak enhancement of the pancreas (65 H) was achieved in 69 sec, whereas at an injection rate of 5 mL/sec, peak enhancement of 84 H was achieved in 43 sec. Using the formula suggested by Tublin et al. in our study, with an injection rate of 3 mL/sec the peak enhancement of the pancreas was reached at 60 sec. In addition, in agreement with previous data by Kim et al. [24], we also found the contrast material volume and injection rate to be directly related to pancreatic parenchymal enhancement. The larger the volume, the better the pancreatic parenchymal enhancement [24]. As can be seen from the quantitative data, a high degree of enhancement (in Hounsfield units) of the normal pancreatic parenchyma and a high tumor- to-gland attenuation difference was achieved. In addition, maximal enhancement of both arterial and venous peripancreatic vessels was obtained with no statistically significant difference observed between these vascular structures with the single-phase technique.
Evaluation of the status of these vessels is of major importance in determining resectability of tumors in the pancreatic head and neck. Finally, the hepatic parenchyma was enhanced at a time closer to the peak hepatic enhancement, optimizing the detection of hepatic lesions. Our results are consistent with those observed by Fletcher et al. [19]. In that study, which included only patients with pancreatic adenocarcinoma, the authors showed that although tumor-to-gland attenuation differences were highest in the pancreatic phase, images obtained in the hepatic phase were equivalent for tumor detection, were best for determining the presence of vascular invasion, provided a homogeneous appearance of the superior mesenteric vein, and allowed the liver to be scanned at the time of peak hepatic enhancement. In our study, we found that by scanning in the equivalent of the late pancreatic phase and early hepatic phase, we can maximize the benefits of each phase of enhancement and provide a complete evaluation of the pancreas, tumor, vessels, and presence of metastatic disease. In addition, in our study, using an injection rate of 3 mL/sec (150 mL of nonionic contrast material) with a scanning delay of 60 sec and a mean scanning duration of 18.4 ± 5 sec, we found that portal venous phase images show the highest liver attenuation and give greater conspicuity for identification of hepatic metastases. Furthermore, the overall mean hepatic CT attenuation value obtained in our study (111 ± 20 H) was also comparable to the mean hepatic CT attenuation value (112 ± 13 H) obtained using a single-detector, single-phase CT technique [11].
As previously shown by Megibow [22], potential problems exist with the reverse-direction technique. Variants in pancreatic morphology or scoliosis will occasionally result in the inferior margin of the gland being below L3. In addition, the patient may have a Riedel lobe of the liver, and the tip of the right lobe will not be included. To overcome this problem, we prefer to start our acquisition from the bottom of the liver through the entire diaphragm. In our series, no patient was excluded from the study because of lack of inclusion of the entire pancreas or liver on the scan.
Our study had several other limitations. First, we did not attempt to compensate for patients' physiologic differences by means of a test dose technique or by using a semiautomated scanning delay. We standardized all scanning to begin 60 sec after the start of the contrast injection. Further studies are required to determine if pancreatic enhancement can be optimized using a technique such as SmartPrep (software, GE Healthcare). A single-phase helical CT acquisition performed with a scanning delay of 60 sec after the beginning of IV contrast administration may not be the ideal method for the detection of hypervascular pancreatic tumors such as islet cell tumors. In these particular cases, with a high clinical and laboratory suspicion, a dual-phase protocol with an early arterial phase (20-25 sec) acquisition should be preferred to the single-phase technique. Second, another potential limitation of our study is the small number of pancreatic adenocarcinomas compared with other series [7, 8, 20, 21]; further studies including only patients with adenocarcinoma are warranted to confirm our findings. Finally, the amount of iodinated contrast material used for each study was not adjusted for each patient on the basis of their weight.
In conclusion, the results of our study show that for detection and assessment of tumor resectability in patients with suspected pancreatic carcinoma, a single-phase acquisition in a caudocranial fashion might be a useful alternative to the standard dual-phase protocol when using MDCT.
Acknowledgments
We thank Graciana Diez-Roux for critically reviewing this paper; Fabio
Garbino, Ugo Salzano, and Bruno Fenderico for their valuable technical
assistance; and Carmela Imparato for editing the manuscript.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  H. Kondo, M. Kanematsu, S. Goshima, T. Miyoshi, Y. Shiratori, M. Onozuka, N. Moriyama, and K. T. Bae MDCT of the Pancreas: Optimizing Scanning Delay with a Bolus-Tracking Technique for Pancreatic, Peripancreatic Vascular, and Hepatic Contrast Enhancement Am. J. Roentgenol., March 1, 2007; 188(3): 751 - 756. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Ichikawa, S. M. Erturk, H. Sou, H. Nakajima, T. Tsukamoto, U. Motosugi, and T. Araki MDCT of pancreatic adenocarcinoma: optimal imaging phases and multiplanar reformatted imaging. Am. J. Roentgenol., December 1, 2006; 187(6): 1513 - 1520. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Imbriaco Reply Am. J. Roentgenol., January 1, 2006; 186(1): 267 - 267. [Full Text] [PDF]
|
|
|
|
|
|
S. M. Erturk Value of the Single-Phase Technique in MDCT Assessment of Pancreatic Tumors Am. J. Roentgenol., January 1, 2006; 186(1): 266 - 267. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
