| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Gastrointestinal Imaging |
1 From the Division of Abdominal Imaging and Intervention, Department of Radiology, Massachusetts General Hospital, 55 Fruit St, White 270 F, Boston, MA 02114. From the 2003 RSNA Annual Meeting. Received February 13, 2004; revision requested April 15; revision received May 13; accepted June 15. Address correspondence to D.V.S. (e-mail: dsahani@partners.org).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSION APPENDIXREFERENCES |
|---|
MATERIALS AND METHODS: The institutional review board approved this study, and informed prior consent was obtained from participants. Perfusion CT of rectal cancer was performed with four-section multi–detector row CT in 15 patients (13 men, two women; mean age, 62.1 years; age range, 46–84 years). Five patients with prostate cancer served as controls. All patients with rectal cancer underwent 6–8 weeks of chemotherapy and radiation therapy followed by surgery. In nine patients, perfusion CT was repeated after completion of chemotherapy and radiation therapy. Contrast medium–enhanced dynamic CT was performed with a static table position for 45 seconds, and the data were analyzed by using commercial software to calculate tissue blood flow (BF), blood volume, mean transit time (MTT), and vascular permeability–surface area product. Perfusion parameters of normal rectum and tumor were compared. Perfusion parameters before and after chemotherapy and radiation therapy were compared. A tumor was considered to have responded if its stage at pathologic analysis indicated regression compared with the preoperative stage. Baseline perfusion values were compared between responders and nonresponders. Statistical analysis was performed with the Student t test.
RESULTS: Rectal cancer showed higher BF and shorter MTT compared with those of normal rectum (P 
.05). After chemotherapy and radiation therapy, tumors showed significant reduction in BF and increase in MTT (P .05). There was a significant difference in baseline BF and MTT values between responders and nonresponders (P .05). Tumors in three patients with high initial BF and short MTT showed poor response.
CONCLUSION: Perfusion CT of rectal cancer can enable assessment of tumor vascularity and perfusion changes that result from chemotherapy and radiation therapy. In this small patient sample, tumors with initial high BF and short MTT values tended to respond poorly to chemotherapy and radiation therapy.
© RSNA, 2005
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSION APPENDIXREFERENCES |
|---|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSION APPENDIXREFERENCES |
|---|
Patient Population
The institutional review board approved this prospective study, and informed consent was obtained from each participant prior to the study. Patients were included in this study if the tumor extended beyond the rectal wall without distant metastases. Patients were excluded from the study if the tumor was stage T1 or T2 or if the patient had renal disease that precluded contrast-enhanced CT (serum creatinine level, >1.9 mg/dL). Between July 2001 and January 2003, 15 consecutive patients who met the inclusion criteria and agreed to participate in the study were enrolled. The study cohort included 13 men and two women (mean age, 62.1 years; age range, 46–84 years) in whom pathologic analysis at biopsy confirmed a diagnosis of nonmucinous adenocarcinoma of the rectum. According to local tumor staging based on endorectal magnetic resonance (MR) imaging (n = 12) and endorectal ultrasonography (n = 3), 13 patients had stage T3 disease and two patients had stage T4 disease at the time of presentation. All patients underwent preoperative chemotherapy and radiation therapy. Three-dimensional conformal radiation therapy was given at a dose of 1.8 Gy per day, 5 days per week, for 6 weeks, with a total radiation dose of 45 Gy. In addition, 5-fluorouracil was administered via peripheral venous infusion at a dose of 225 mg/m2 per day, 5 days per week, for 6 weeks. At the end of 6 weeks, 4–5 Gy pelvic radiation was given for therapy. Total mesorectal fascial excision was finally performed within 1–3 weeks after completion of chemotherapy, and the specimen was evaluated for pathologic staging. Any tumor that showed a regression in local stage at pathologic analysis as compared with that at preoperative imaging (Table 1) was considered a responder to chemotherapy and radiation (n = 12). If there was no difference in tumor stage or if there was an increase in local tumor stage, the tumor was considered a nonresponder (n = 3).
|
Five additional patients, all men (mean age, 62.6 years; age range, 55–71 years) who had no rectal disease but who had prostate cancer confined to the prostate, also underwent perfusion CT and served as controls. These patients had received no cancer-related therapy prior to CT, and they were not consecutive. These patients underwent perfusion CT for the evaluation of prostate cancer at our hospital during a pilot study that was approved by the institutional review board and for which informed consent was obtained. The same parameters were measured in the five patients with a normal rectal wall as were measured in the patients with rectal cancer. These five patients served as controls for the estimation of perfusion parameters in normal rectal wall for this age group.
Perfusion CT Technique
Perfusion CT was performed with a four-section multi–detector row CT scanner (LightSpeed QX/i; GE Medical Systems, Milwaukee, Wis). CT scanning of the pelvis was performed without oral or intravenous contrast medium to localize the tumor. The tumor was localized on the nonenhanced CT scan, and a 2-cm tumor region of interest (ROI) was selected for cine imaging. This area was chosen by an experienced gastrointestinal radiologist (D.V.S., who had 11 years of experience in radiology and 5 years of experience in gastrointestinal radiology) and on the basis of visible tumor volume. Dynamic study of this area was performed at a static table position during intravenous injection of 125 mL of iopamidol (Isovue 300; Bracco Diagnostics, Princeton, NJ) containing 300 mg of iodine per milliliter. The following parameters were used: 1-second gantry rotation time, 100–120 kVp, 200–240 mA, 10-second scanning delay from the start of injection, 45-second duration of transverse data acquisition (four sections per gantry rotation), and 5-mm reconstructed section thickness. The images were reconstructed at 1-second temporal intervals. We chose a 45-second duration of scanning for three reasons: First, it would limit the radiation dose to the subject. Second, since the patient has to keep still during scanning, the shorter time would help to prevent image artifacts caused by patient motion. Third, the level of precision with which the parameters of our tracer kinetics model were estimated was 5%–15% (13,16), which was judged acceptable.
On completion of the study, the data were transferred to an image processing workstation (Advantage Windows 4.0; GE Medical Systems) and analyzed by using software (CT Perfusion 2.0; GE Medical Systems). The analysis is based on the principles discussed in the Appendix. The parameters generated by the software are blood flow (in milliliters per 100 g of wet tissue per minute), blood volume (in milliliters per 100 g of wet tissue), mean transit time (in seconds), and permeability–surface area product (in milliliters per 100 g of wet tissue per minute). As discussed in the Appendix, for the derivation of the functional maps of blood flow, blood volume, mean transit time, and permeability–surface area product, the arterial input curve of contrast medium concentration Ca(t) is required. In all the CT scans, Ca(t) was obtained from an ROI in the right external iliac artery (area range, 18–25 mm2) by one of the authors (S.P.K., with 8 years of experience in general radiology and 1 year in gastrointestinal radiology). In addition, a second reference ROI (area range, 18–25 mm2) within a vessel large enough to be fully representative of contrast enhancement (measured in Hounsfield units) for 100% blood was identified to correct for any partial-volume averaging that may have occurred in the obtained Ca(t) (14). In our study, the left external iliac artery was used for partial-volume averaging correction. Using these reference curves, we calculated maps of blood flow, blood volume, mean transit time, and permeability–surface area product. The CT data from the five control patients were similarly analyzed. The ROI (area range, 175–676 mm2) was drawn along the rectal wall for computation of the perfusion parameters.
Statistical Analysis
The tumor ROIs (area range, 293–1300 mm2) within the rectum were hand drawn (S.P.K.) both for each map type and within each map type (blood flow, blood volume, mean transit time, and permeability–surface area product) for all four anatomic section locations available for each patient. Representative parameter values were then averaged across the four sections.
Perfusion parameters in control patients (patients with prostate cancer confined to the prostate) were similarly obtained from normal rectum and compared with the perfusion parameters in patients with rectal tumors to identify any perfusion differences between tumor and normal rectum. The perfusion parameters before and after radiation therapy were compared to investigate whether radiation therapy caused changes in the perfusion parameters. Baseline perfusion values (before initiation of chemotherapy and radiation therapy) were compared between responders and nonresponders. Statistical analysis was performed by using the Student t test for comparison of two data sets, and P values were calculated for each comparison (Excel 2000; Microsoft, Richmond, Va). P .05 was considered to indicate a statistically significant difference.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSION APPENDIXREFERENCES |
|---|
.05), with high blood flow in rectal tumors compared with that in normal rectum. The mean transit time was significantly shorter in rectal tumors (P .05) (Figure). There was no significant difference in blood volume or in permeability–surface area product between tumors and normal rectum (Table 2).
|
|
|
|
|
.05) and increase in mean transit time (P .05) after chemotherapy and radiation therapy. Of the nine patients, seven showed decrease in blood flow and increase in mean transit time (Fig). Two patients showed a 13% increase in blood flow and a 10% increase in mean transit time. No statistically significant difference in blood volume or in permeability–surface area product was observed between CT scans obtained before and those obtained after chemotherapy and radiation therapy (Table 3).
|
.05). Scans in the group of patients who had tumors with initial high blood flow and shorter mean transit time (n = 3) showed a poor response to chemotherapy and radiation therapy. Comparison of other parameters, however, did not indicate any statistically significant difference (Table 4). Tumor response and nonresponse are defined in Table 1.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSION APPENDIXREFERENCES |
|---|
A correlation between contrast enhancement measures and microvessel density has been reported in renal and lung cancers (18–20). High tumoral microvessel density has been demonstrated to be associated with high blood flow by means of uptake of the radiotracer technetium Tc 99m sestamibi in breast carcinoma (21). In addition, newly developed tumor vessels demonstrated increased fenestrations of basement membrane that resulted in increased permeability to large molecules. Thus, the use of functional imaging techniques to assess tissue perfusion and permeability may offer a means to study the angiogenic process in tumors.
Axel (22) first described a method of assessing tissue perfusion by using dynamic contrast-enhanced CT. Because of the technical limitations of early CT scanners, this method was not widely available. With the advent of spiral CT and multi–detector row CT scanners, the technique became more popular and new indications emerged (22,23). The CT perfusion software used in this study models the distribution of contrast medium in tumor by using a distributed model (in which contrast medium concentration is nonuniform) for the intravascular space and a compartmental model (in which contrast medium concentration is uniform) for the extravascular space (11–15). This hybrid model allows the simultaneous determination of blood flow, blood volume, vascular mean transit time, and capillary permeability–surface area product, which is its main advantage over the compartmental model. The compartmental model can determine only blood volume and F · E, the product of blood flow (F) and extraction efficiency (E) (12–14).
Perfusion CT in Rectal Cancer
There are no published reports about the utility of perfusion CT in rectal cancer. However, many articles have been published about studies of perfusion CT and perfusion MR imaging in tumors of the cervix (24), liver (7), lung (8), heart (25), pancreas (9), spleen, lymph nodes (26), musculoskeletal system (27), and head and neck (10). In most of these studies, researchers focused on the detection of metastases, differentiation of benign from malignant lesions, and assessment of therapeutic response.
Comparison of Perfusion Parameters between Normal Rectum and Rectal Cancer
Compared with normal rectum, rectal cancer consistently showed high blood flow and shorter mean transit time. There was no significant difference in blood volume and permeability in rectal cancer compared with those in normal rectum. These results suggest that angiogenesis in the rectal tumors studied stimulated the opening of a significant number of arteriovenous shunts rather than the acquisition of a new vascular supply. The arteriovenous shunts have very low resistance to flow, which results in markedly increased blood flow and shorter mean transit time. These vessels facilitate the direct passage of blood from the arterial to the venous outlets without passage through the exchange vessels (capillaries) and, thus, result in no change in permeability–surface area product (28). We found that the duration of data collection for the arterial input concentration curve Ca(t) and the tissue curve Q(t) affected the accuracy and precision with which the permeability–surface area product was determined. If only first-pass data are acquired, then blood flow, blood volume, and mean transit time can be accurately determined with precision of about 15% (13).
Perfusion Changes after Chemotherapy and Radiation Therapy
Previous studies demonstrated increase in fractional vascular volume and contrast medium clearance per unit of volume (a measure of permeability) after radiation therapy, which may represent a hyperemic response to radiation therapy (29). Similarly, Mayr et al (24) studied cervical cancer with perfusion MR imaging and found that high tumor perfusion before radiation therapy and increasing or persistent high perfusion early in the course of therapy were favorable signs. In our study, however, there was a consistent decrease in blood flow and a consistent increase in mean transit time after chemotherapy and radiation therapy. These results may be attributable to the addition of chemotherapeutic agents in our treatment regimen and to the interval between therapy and subsequent perfusion imaging. Many of the present chemotherapeutic agents are cytotoxins that are capable of damaging the vascular endothelium, an effect that would have prevented the hyperemic response generally observed after radiation therapy. The timing of perfusion imaging after radiation therapy has a substantial effect on perfusion values, as early changes after radiation therapy result in leaky capillaries and late changes result in fibrosis with few vessels. This fact also explains the consistent increase in mean transit time in all of our patients. In our study, perfusion CT was performed 1–2 weeks after chemotherapy and radiation therapy. In addition, seven of nine patients showed a significant decrease in blood flow after chemotherapy and radiation therapy. Two patients showed a mild increase in blood flow; these two patients, however, also showed an increase in mean transit time.
Prediction of Response on the Basis of Baseline Perfusion Study
In numerous prior studies, measures of tumor perfusion have been correlated with response to therapy. Results of these studies have not always been in agreement. It is logical to think that high perfusion values, which indirectly suggest a high rate of angiogenesis and microvessel density within the tumor, may indicate a high grade of tumor and, therefore, may indicate a poor response to therapy and/or a worse prognosis. There are many publications that indicate a relationship between tumor angiogenesis, as determined by microvessel density, and prognosis (24,30–32). In our study, tumors with high blood flow and low mean transit time at baseline perfusion imaging showed a poor response. DeVries et al (33) reported similar findings in their study, in which pretherapeutic perfusion MR imaging indexes of rectal cancer were compared with therapeutic response after chemotherapy and radiation therapy. A high blood perfusion index (mL/min/100 g tissue) in tumors was associated with a poor response. Sokmen et al (34) observed similar findings in their analysis and showed that increased tumor vascularity was associated with decreased survival among patients with rectal cancer. High perfusion values in association with a poor response to chemotherapy and radiation therapy can be explained by large numbers of intratumoral arteriovenous shunts with a high perfusion rate and low exchange of oxygen (35). Our results also suggest high intratumoral arteriovenous shunts without the development of new blood vessels. High perfusion could also be a result of intrinsic high angiogenic activity of tumor or a secondary response to tissue hypoxia (36,37). Comparison of the results from different CT and/or MR perfusion studies, however, has to be made cautiously, as the values measured are dependent on the mathematic model and the pharmacokinetics of the contrast medium used. Thus, the application of different models to the same data may well yield different perfusion values.
A low vascularity, however, was related to a high local recurrence rate in cervical cancer after radiation therapy (38). In liver, high arterial perfusion peripheral to metastases was associated with better survival (39). Similarly, poor perfusion of the portal venous system was associated with a poor prognosis (40). These findings suggest that an assessment of response on the basis of perfusion values should be individualized to the type of tumor, the treatment modality, and the mathematic model used to evaluate perfusion parameters.
Limitations in Our Study
There were several limitations in our study. First, we did not study tumor permeability, which might have served as an independent predictor of tumor grade or response. Second, the size of the study population in this pilot study was too small to permit a power analysis. Third, the tumoral response was based on a comparison between initial radiologic findings and final pathologic findings, a method that could have caused an inadvertent bias, as the stage of some tumors would have been under- or overestimated because of the limitations inherent in imaging modalities. An alternative method based on imaging would involve measurement of tumor volume before and after therapy. This approach was avoided because of the irregular shape of some tumors and the limited view of tumors afforded by four-section repeated imaging. In addition, peristalsis in the rectum can introduce motion artifacts, which may interfere with calculation of perfusion values.
In conclusion, perfusion CT of rectal cancer can be useful for assessing tumor vascularity and changes in perfusion after chemotherapy and radiation therapy. In our small patient sample, tumors with initial high blood flow and short mean transit time tended to respond poorly to chemotherapy and radiation therapy. To obtain more conclusive results, however, a larger patient population must be studied. It should be emphasized that the results of our study are specific to the analytic methods and software employed.
|
APPENDIX |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSION APPENDIXREFERENCES |
|---|
|
|
|
|
|
|
Figure A2 is a plot of F · R(t) or the blood flow–scaled impulse residue function. It lends itself to the following interpretation: If a bolus of contrast medium is injected directly into the arterial inlet of the tissue so that the arterial concentration curve, Ca(t), is held at unity for a very short period (approaching 0 seconds), the total mass of contrast medium delivered to the tissue is numerically equal to F. The blood flow–scaled impulse residue function, F · R(t), which reflects the mass of contrast medium remaining in the tissue, therefore, would reach a height of F immediately and maintain this height for a duration equal to the vascular mean transit time of the tissue, Vb/F. The shaded area in Figure A2 is, therefore, the blood volume Vb, according to the central volume principle (44,45). After a time equal to Vb/F, unextracted contrast medium starts to leave the tissue, F · R(t) drops to a height of F · E, and thereafter contrast medium in the interstitial tissue diffuses back into the intravascular space (capillary) and is washed out by blood flow. This portion of F · R(t) is described by a decreasing monoexponential function with a rate constant equal to FE/Ve. With Ca(t) and Q(t) measured with CT, F · R(t) can be determined by model deconvolution (46) according to Equations (A3) and (A4), and the parameters F, Vb, MTT, and E(PS) can be determined as discussed previously.
|
|
FOOTNOTES |
|---|
T.Y.L. is a consultant with GE Healthcare (formerly GE Medical Systems) on their CT Perfusion software product. He is also supported in his research by grants from GE Healthcare.
Author contributions: Guarantor of integrity of entire study, D.V.S.; study concepts, D.V.S.; study design, D.V.S., C.G.W., S.S.; literature research, S.P.K., D.V.S.; clinical studies, C.G.W., D.V.S.; data acquisition, D.V.S., S.P.K.; data analysis/interpretation, S.P.K., L.M.H., D.V.S.; statistical analysis, S.P.K., L.M.H.; manuscript preparation, S.P.K., D.V.S., T.Y.L., L.M.H.; manuscript definition of intellectual content, D.V.S., S.P.K.; manuscript editing, D.V.S., P.F.H., S.P.K., L.M.H., T.Y.L.; manuscript revision/review, all authors; manuscript final version approval, D.V.S.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSION APPENDIXREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  V. Goh, S. Halligan, F. Daley, D. M. Wellsted, T. Guenther, and C. I. Bartram Colorectal Tumor Vascularity: Quantitative Assessment with Multidetector CT--Do Tumor Perfusion Measurements Reflect Angiogenesis? Radiology, September 23, 2008; (2008) 2492071365. [Abstract] [Full Text]
|
|
|
|
 |
|
 N. I. Sainani, O. A. Catalano, N.-S. Holalkere, A. X. Zhu, P. F. Hahn, and D. V. Sahani Cholangiocarcinoma: Current and Novel Imaging Techniques RadioGraphics, September 1, 2008; 28(5): 1263 - 1287. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Sabir, R. Schor-Bardach, C. J. Wilcox, S. Rahmanuddin, M. B. Atkins, J. B. Kruskal, S. Signoretti, V. D. Raptopoulos, and S. N. Goldberg Perfusion MDCT Enables Early Detection of Therapeutic Response to Antiangiogenic Therapy Am. J. Roentgenol., July 1, 2008; 191(1): 133 - 139. [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. 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]
|
|
|
|
|
|
R. G. Sheiman, A. Sitek, V. Goh, and S. Halligan CT Perfusion Imaging: Know Its Assumptions and Limitations Radiology, February 1, 2008; 246(2): 649 - 650. [Full Text] [PDF]
|
|
|
|
 |
|
 A. X. Zhu, N. S. Holalkere, A. Muzikansky, K. Horgan, and D. V. Sahani Early Antiangiogenic Activity of Bevacizumab Evaluated by Computed Tomography Perfusion Scan in Patients with Advanced Hepatocellular Carcinoma Oncologist, February 1, 2008; 13(2): 120 - 125. [Abstract] [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]
|
|
|
|
|
|
M. Bellomi, G. Petralia, A. Sonzogni, M. G. Zampino, and A. Rocca CT Perfusion for the Monitoring of Neoadjuvant Chemotherapy and Radiation Therapy in Rectal Carcinoma: Initial Experience Radiology, August 1, 2007; 244(2): 486 - 493. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Hakime, H. Peddi, A. U. Hines-Peralta, C. J. Wilcox, J. Kruskal, S. Lin, T. de Baere, V. D. Raptopoulos, and S. N. Goldberg CT Perfusion for Determination of Pharmacologically Mediated Blood Flow Changes in an Animal Tumor Model Radiology, June 1, 2007; 243(3): 712 - 719. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. V. Sahani, N.-S. Holalkere, P. R. Mueller, and A. X. Zhu Advanced Hepatocellular Carcinoma: CT Perfusion of Liver and Tumor Tissue--Initial Experience Radiology, June 1, 2007; 243(3): 736 - 743. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Provenzale Imaging of Angiogenesis: Clinical Techniques and Novel Imaging Methods Am. J. Roentgenol., January 1, 2007; 188(1): 11 - 23. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V Goh, S Halligan, L Gartner, P Bassett, and C I Bartram Quantitative colorectal cancer perfusion measurement by multidetector-row CT: does greater tumour coverage improve measurement reproducibility? Br. J. Radiol., July 1, 2006; 79(943): 578 - 583. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. K. Kalra, W. C. Small, and W. E. Torres A 45-Second CT Perfusion Protocol for Rectal Cancers May Not Be Adequate to Infer Vascular Permeability-Surface Area Products Radiology, February 1, 2006; 238(2): 755 - 756. [Full Text] [PDF]
|
|
|
|
|
|
C. Kremser, T. R. Trieb, W. Judmaier, A. F. DeVries, D. Sahani, S. Kalva, and P. F. Hahn Assessing Tumor Perfusion and Treatment Response in Rectal Cancer Radiology, February 1, 2006; 238(2): 756 - 757. [Full Text] [PDF]
|
|
|
|
