| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, University of Vienna, Währinger Gürtel
18-20, A-1090 Vienna, Austria.
2 Department of Surgery, University of Vienna, A-1090 Vienna, Austria.
3 Department of Medical Computer Sciences, University of Vienna, A-1090 Vienna,
Austria.
4 Department of Clinical Pathology, University of Vienna, A-1090 Vienna,
Austria.
Received May 28, 2002;
accepted after revision February 13, 2003.
Address correspondence to A. G. Maier.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Thirty-nine rectal cancer patients (20 men, 19 women) underwent transrectal sonography performed with a 10-MHz endoanal probe and MR imaging (1.0 T or 1.5 T) using a whole-body coil. After rectal application of a superparamagnetic iron oxide MR contrast agent, T1- and T2-weighted images and gadolinium-enhanced double-contrast images were obtained. The results of examinations were compared with the histology of resected specimens.
RESULTS. Histopathology showed four stage T1, 11 stage T2, 18 stage T3, and six stage T4 tumors using the TNM staging system. Nodal metastases were seen in 16 patients. Transrectal sonography could not be performed in 11 patients because of the high location of the tumor. In the remaining 28 patients, the accuracy of transrectal sonography for T stage was 64% overall (patients not receiving radiation, 69%; patients receiving radiation, 60%) and 70% for N stage. In 39 patients, double-contrast MR imaging correctly identified the T stage with an accuracy of 64% overall (patients not receiving radiation, 75%; patients receiving radiation, 53%) and the N stage with an accuracy of 62%. The assessment of rectal wall penetration (Dukes' classification A versus B) revealed a sensitivity, specificity, and accuracy of 93%, 71%, and 82%, respectively, for transrectal sonography and 100%, 60%, and 85% for MR imaging.
CONCLUSION. If it is technically feasible, transrectal sonography is an accurate method for staging rectal cancer. In proximal or stenotic tumors, double-contrast MR imaging is the method of choice. Diagnostic accuracy of transrectal sonography and MR imaging is high for predicting bowel wall penetration.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Accurate preoperative staging is important for the planning of an appropriate treatment strategy. In particular, the tumor stage determines whether radiation or chemotherapy should be used in addition to surgery. There is an increasing trend toward preoperative radiotherapy to improve local tumor control by "down-staging" in patients with stages T3 and T4 disease. Preoperative irradiation prolongs patient survival by reducing the rate of local recurrence [5, 6].
Transrectal sonography has been the method of choice for preoperative staging of rectal tumors. Transrectal sonography is a good method for assessment of tumor T stages, with a reported accuracy of 67–96% [7–9]. However, it has limitations in the detection and characterization of lymph nodes. Moreover, transrectal sonography is highly operator-dependent and, because of the rigid sonographic transducer, cannot be used for tumors in the upper rectum or for stenotic tumors.
MR imaging using a transrectal coil has been shown to be effective in revealing the layers of the rectal wall, which is a prerequisite for accurate staging. However, MR using a transrectal coil has the same limitations as does transrectal sonography with regard to the field of view and the assessment of stenotic tumors [10, 11].
Superparamagnetic iron oxide MR contrast agents have been shown to be effective and safe after oral ingestion for delineation of the intestinal tract [12, 13]. A double-contrast MR characterization of the rectum can be obtained by rectal application of a superparamagnetic iron oxide, together with the IV administration of gadolinium chelates. Recent studies have shown that this technique is promising in the staging of rectal cancer [14, 15].
The purpose of our study was to assess the accuracy of double-contrast MR imaging compared with transrectal sonography in the preoperative staging of rectal cancer.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
At histopathologic examination of the resected specimens, stage T1 tumors were found in four patients, stage T2 tumors in 11, stage T3 tumors in 18, and stage T4 tumors in six patients. No metastatic lymph nodes were found in 22 patients. Metastatic lymph nodes were present in 16 patients, 10 having stage N1 and six having stage N2 tumors. In two patients, lymph node involvement could not be assessed histopathologically because of transanal endoscopic microsurgery.
Transrectal Sonography
All transrectal sonography was performed with a commercially available
sonographic scanner (Diagnostic Ultrasound Systems 3535, B&K Medical,
Gentofte, Denmark) using a mechanical rotating endoprobe (type 1850, 10 MHz)
with a focal zone (penetration depth) of 2–4 cm, an axial resolution of
0.4 mm, and a lateral resolution of 1.0 mm. A 360° radial display of the
rectal wall and surrounding tissues was provided. A latex balloon was attached
to the end of the probe over the transducer. After a cleansing enema, the
patient was placed in the lithotomy position. The sonographic probe was
inserted by an experienced radiologist without external guidance, the balloon
was filled with degassed water, and the probe was slowly withdrawn during
scanning of the rectal wall.
The transrectal sonography classification of tumor invasion into bowel wall corresponds to the pathologic T classification of the TNM system [2]. At transrectal sonography, stage T1 represents mucosal or submucosal disease seen as an irregular and thinned middle echogenic layer (submucosa) with a hypoechogenic tumor mass. Stage T2 is disease involving the hypoechoic muscularis propria, stage T3 is defined by extension into the hyperechoic perirectal fat, and stage T4 represents extension into adjacent organs or pelvic sidewall structures. In addition, the penetration of the rectal wall (stages T1 and T2 versus stages T3 and T4) as clinically important information was assessed because that information translates into Dukes' classification A versus Dukes' classification B [16].
For the N staging, the presence of sonographically visible hypoechoic regional lymph nodes, regardless of size, was considered indicative of metastatic disease. By definition, in stage N1, one to three perirectal lymph nodes are affected; in stage N2, four or more lymph nodes are affected.
MR Imaging
MR imaging was performed on a 1.0-T unit (NT 1.0, Philips, Best, The
Netherlands; n = 18) and a 1.5-T unit (Magnetom Vision, Siemens,
Erlangen, Germany; n = 21) using a whole-body coil. Preparation of
the patients before the examinations included a cleansing enema. Before the
start of the MR examination, all patients received an enema consisting of
200–400 mL of a superparamagnetic iron oxide MR contrast agent (Abdoscan
[ferristene], Amersham Health, Oslo, Norway; or Lumirem [ferumoxsil], Guerbet,
Aulnay-sous-Bois, France). The rectal administration was performed manually
using an enema bag or a syringe with attached tubing. The total volume of
superparamagnetic iron oxide applied (median, 320 mL) varied depending on the
patient and the extent of tumor masses. Application was stopped when the
patient started to feel uncomfortable.
N-butylscopolammonium bromide (Buscopan, Boehringer Ingelheim, Ingelheim, Germany) was given IV at a dose of 40 mg immediately before administration of the superparamagnetic iron oxide. In case of contraindications or if the patient was allergic to Buscopan, IV glucagon at a dose of 0.5–1.0 mg was used instead. MR imaging was begun immediately after the rectal contrast agent was applied. T2-weighted turbo spin-echo images in the sagittal plane and T1-weighted spin-echo images in the sagittal and paraaxial planes, oriented perpendicular to the axis of the rectum at the level of the tumor, were obtained before and after the IV administration of 0.1 mmol/kg of gadodiamide (Omniscan, Amersham Health).
The parameters for the Philips NT 1.0 scanner for the T2-weighted turbo spin-echo sequences were TR/TE, 3787/99; number of acquisitions, 2; slice thickness, 5 mm; field of view, 250 mm; matrix, 286 x 512; and acquisition time, 4 min 6 sec. The parameters for the T1-weighted spin-echo sequences were 627/15; number of acquisitions, 2; slice thickness, 5 mm; field of view, 230 mm; matrix, 224 x 512; and acquisition time, 5 min 18 sec.
The parameters for the Siemens Magnetom Vision for the T2-weighted turbo spin-echo sequences were 5677/132; number of acquisitions, 1; slice thickness, 6 mm; field of view, 350 mm; matrix, 240 x 256; and acquisition time, 3 min 7 sec. The parameters for the T1-weighted spin-echo sequences were 760/17; number of acquisitions, 2; slice thickness, 6 mm; field of view, 350 mm; matrix, 230 x 256; and acquisition time, 6 min 17 sec.
Staging of rectal tumors was performed according to the TNM system using the following criteria: location of the detectable primary tumor and infiltration of mucosa, muscles, perirectal fat, or adjacent organs [2]. On MR images, tumors show a higher signal intensity than does the hypointense muscle layer; but T2-weighted and contrast-enhanced T1-weighted images show a lower signal intensity than the hyperintense mucosa and submucosa. On MR images, in stage T1 a thickening of mucosa and submucosa with preserved muscularis propria is seen. Stage T2 shows an irregular or thickened muscularis propria layer. In stage T3, the muscularis propria presents as disrupted; in stage T4, the tumor extends into adjacent organs.
For the N staging with MR imaging, the presence of detectable regional lymph nodes, regardless of size, was considered suggestive of metastatic disease.
Image Assessment and Verification
Preoperative transrectal sonography and MR imaging results were compared
with histopathologic results from resected specimens from all patients. For
histopathology, all surgical specimens were dissected transversally. After
overnight fixation in neutral buffered formalin (7.5%), the entire tumor was
transversely dissected into 4-mm-thick slices. All tumor slices were blocked
sequentially, dehydrated in a series of graded alcohol, and embedded in
paraffin wax. Two 4-µm-thick sections of each paraffin block were stained
with H and E.
The depth of local tumor spread was classified histopathologically according to the TNM stage as follows: stage T1 is tumor confined to the submucosa, T2 is tumor confined to the muscularis propria, T3 is a tumor that penetrates the rectal wall and extends into perirectal fat or subserosa, and T4 is a tumor that invades adjacent organs or structures or penetrates the serosa.
Real-time transrectal sonograms were interpreted by one reviewer. Hard-copy images of the studies were reviewed by a second radiologist. Discrepancies were resolved by consensus. MR images were interpreted by two observers in consensus, both of whom were unaware of the results of the sonographic studies.
Statistical Analysis
Sensitivity, specificity, and corresponding 95% confidence intervals (CIs)
were calculated for bowel wall penetration and assessment of the presence of
nodal disease. The stratified chi-square trend test and the exact chi-square
trend test were used to test the relationship between transrectal sonography
and MR imaging results, stratified by the true results of T and N staging. A
p value of less than 0.05 was considered statistically significant.
Fisher's exact test and the chi-square test were used to test differences in
accuracy between transrectal sonography and MR imaging with regard to T and N
staging; and between the groups of patients who underwent preoperative
irradiation and those who did not, for transrectal sonography and MR imaging.
A p value of less than 0.05 was considered to indicate statistical
significance; all tests were two-sided. Statistical software (SAS Institute,
Cary, NC) and StatXact-4 software (Cytel Software, Cambridge, MA) were used
for statistical evaluation.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
|
|
Regarding penetration of the rectal wall (stages T1 and T2 vs stages T3 and T4), transrectal sonography showed a sensitivity of 93% (95% CI, 66.1–99.8%), a specificity of 71% (95% CI, 41.9–91.6%), and an accuracy of 82% (Dukes' classification).
In 10 patients with discrepancies between transrectal sonography and the histopathologic results (overstaging in seven, under-staging in three), six patients had undergone preoperative radiation. Of these six patients, three stage T1 tumors were overstaged as T2 tumors, and three stage T2 tumors were misdiagnosed as T3 on sonography.
Of the 28 patients, 15 had undergone preoperative radiotherapy (25 Gy, n = 7; 50 Gy, n = 8) and 13 had not. Accuracy for patients who underwent preoperative radiotherapy was 60% (9/15) for T staging and 73% (11/15) with regard to bowel wall penetration (Dukes' classification). For the patient group without irradiation, the accuracy was 69% (9/13) and 92% (12/13), respectively. The difference did not yield statistical significance (p = 0.71).
Double-contrast MR imaging.—Overall, MR imaging correctly staged 25 of 39 tumors for an accuracy of 64% for T staging (Table 2). MR tumor staging was correct in 12 of 18 patients examined at the 1.0-T MR unit and in 13 of 21 patients examined at the 1.5-T MR unit (accuracy, 67% vs 62%; p = 0.54, not significant). Disease was overstaged in 10 patients (seven who had preoperative radiation) and understaged in four. One small cancer, stage T1, was not detected on MR imaging (Fig. 4).
|
|
With regard to penetration of the rectal wall, double-contrast MR imaging showed a sensitivity of 100% (95% CI, 88.3–100%), a specificity of 60% (95% CI, 32.3–83.7%), and an accuracy of 85% (Dukes' classification).
In total, 19 patients underwent preoperative radiotherapy (25 Gy, n = 10; 50 Gy, n = 9), and 20 did not. Accuracies for patients having radiotherapy were 53% (10/19) for T stage and 68% (13/19) with regard to bowel wall penetration (Dukes' classification). For the patient group without irradiation, the accuracies were 75% (15/20) and 100% (20/20), respectively. However, these differences did not reach statistical significance (p = 0.16).
In the 28 patients who underwent both modalities, the overall accuracies of MR imaging were 57% (T staging) and 79% regarding bowel wall penetration (Dukes' classification; 1.0 T: 77%, 1.5 T: 80%; p = 0.59, not significant) and for transrectal sonography, 64% and 82%, respectively.
No statistically significant difference was seen between transrectal sonography and MR imaging with regard to T staging (p = 0.6).
N Staging
The detailed results of N staging by transrectal sonography and
double-contrast MR imaging are shown in Tables
3 and
4. Transrectal sonography was
correct in 19 of 27 cases (one patient did not undergo lymph node dissection)
for an accuracy of 70%. With regard to the clinically important presence or
absence of nodal disease, transrectal sonography showed a sensitivity of 92%
(95% CI, 64.0–99.8%), a specificity of 71% (95% CI, 41.9–91.6%),
and an accuracy of 81%.
|
|
Of the 37 patients with sufficient histopathologic workup, MR imaging was correct in 23 (accuracy, 62%). With regard to nodal involvement, MR imaging showed sensitivity, specificity, and accuracy of 81% (95% CI, 54.4–96.0%), 62% (95% CI, 38.4–81.9%), and 70%, respectively.
No statistically significant difference was seen between transrectal sonography and MR imaging with regard to N staging (p = 0.34).
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Many studies have emphasized the importance of accurate preoperative tumor staging because patient outcome and survival depend on early tumor detection and the best-suited operative procedure, chosen with the help of the proper preoperative tumor classification [16, 19, 20].
A variety of imaging techniques—MR imaging, endosonography, and CT—are available, but mixed results concerning efficacy and applicability have been reported [21]. Transrectal sonography shows the best results for T staging, with accuracy rates of up to 93%; CT accuracy is low, with rates of 33–77%, but improves in advanced stages. In cases of possible tumor extension into adjacent organs, cross-sectional imaging is indicated, and MR imaging is the most accurate [21]. MR imaging with endorectal surface coils has shown encouraging results, although accuracy rates were no higher than 81% [9]. Blomqvist et al. [22] reported no better results when using an endorectal coil with a pelvic phased array coil. However, the patient sample was small in that study. The excellent depiction of sphincteral infiltration in tumors of the lower rectum is an advantage. MR imaging with endorectal surface coils has the same limitations as transrectal sonography: insufficient tumor visualization for tumors located high in the rectum and for stenotic tumors.
Our study compares the efficacy of transrectal sonography and double-contrast MR imaging in the staging of rectal cancer and attempts to point out the advantages and disadvantages of these two methods.
Transrectal sonography shows the rectal wall as rings of different echogenicity. Transrectal sonography can define the depth of tumor invasion by the disruption of the different wall layers with an accuracy of 66–95% [7–9, 23, 24]. In contrast, our results show a moderate accuracy of 64%. However, the analysis of our patient group reveals that most tumors that were not correctly staged were in patients who had undergone preoperative radiation therapy.
Patients with tumors considered to be stage T3 by transrectal sonography and clinical assessment were referred for preoperative radiation therapy, which renders correlation between pre-radiotherapy transrectal sonography and postradiotherapy surgical specimens difficult because of shrinkage of the rectal carcinoma and possible postradiation edema of the rectal wall, as well as regression of perirectal lymph nodes [22, 25, 26]. These changes are the reason for some cases of overstaging by double-contrast MR imaging and transrectal sonography. The result is an artificially low accuracy for transrectal sonography in the differentiation between stages T1 and T2 tumors and between stages T2 and T3 tumors, even though transrectal sonography is expected to yield good accuracy, especially in defining the earlier tumor stages (Fig. 5A, 5B).
|
|
Accuracy for T staging in patients who underwent radiotherapy before surgery decreases to 53–59% for transrectal sonography and 47–65% for MR imaging [22, 25]. Our study shows an accuracy of 53% (10/19) for double-contrast MR imaging and 60% (9/15) for transrectal sonography for assessing the exact T stage. Accuracies for patients who did not undergo preoperative radiation therapy were higher: 69% for transrectal sonography and 75% for MR imaging. However, for staging according to the Dukes' classification, the sensitivities for bowel wall penetration (Dukes' A vs B) were 93% and 100% of the cases for transrectal sonography and MR imaging, respectively.
A limiting factor for the applicability of transrectal sonography is the tumor location: in 11 patients (28%) the tumor could not be evaluated with this method because it was located too high in the rectum (Fig. 2). Moreover, the depth penetration of transrectal sonography is limited. Therefore, transrectal sonography was unable to detect infiltration of the levator ani muscle in one patient.
Despite the larger field of view of MR imaging with a phased array coil, this method is not superior to transrectal sonography in early tumor stages because of its lower spatial resolution (Fig. 4). Particularly for tumors confined to the rectal wall (stages T1 and T2), MR imaging lacks precision in differentiating between stages T1 and T2 tumors and in defining early infiltration into the perirectal fat (stage T3). In our study, transrectal sonography showed a higher specificity than MR imaging for the penetration of the rectal wall (71% vs 60%).
MR imaging has no limitation in the assessment of advanced tumor stages (Fig. 3A, 3B). In our study, all tumors located in the upper rectum could be identified on MR imaging. MR imaging can detect perirectal organ infiltration [9, 10, 14] as well as infiltration of the levator ani muscle because of its multiplanar imaging capability, which has a major impact on subsequent management and can influence the choice of surgical intervention. Abdominoperineal resection, rather than a sphinctersaving anterior resection, is the method of choice in cancers involving the levator ani muscle [27]. However, our group had only a small number of advanced-stage tumors.
With regard to the clinically and therapeutically important penetration of the bowel wall, previous studies have shown sensitivity, specificity, and accuracy for double-contrast MR imaging of 97–100%, 50–60%, and 82–90%, respectively [14, 28]. Our results are in the same range: 100%, 60%, and 85%, respectively. Transrectal sonography showed equally high sensitivity and accuracy (93%, 82%) but an even higher specificity of 71%.
Lymph node assessment remains an unsolved problem in the preoperative staging of rectal cancer for both transrectal sonography and MR imaging. Neither method can assess the cause of lymph node enlargement with a high degree of certainty. Although we used a larger field of view (250–350 mm) to assess not only the perirectal but also the parailiac lymph nodes on MR imaging, accuracy was greater for transrectal sonography (70%) than for MR imaging (62%). With regard to the presence or absence of nodal involvement, both methods showed an increase of diagnostic accuracy (81%, transrectal sonography; 70%, MR imaging). In the literature, accuracy ranges between 61% and 83% for transrectal sonography [21, 29] and between 43% and 95% for MR imaging [14, 21]. Alternatives, such as lymph node–specific MR contrast agents, have shown promising results in lymph node staging of pelvic tumors [29–31].
The main limitation of our study is that it is difficult to obtain exact correlation between transrectal sonography, MR imaging, and surgical specimens after preoperative irradiation. However, preoperative radiation therapy in presumed stage T3 tumors is now the standard of care at many institutions.
Accuracy of MR imaging was limited by the use of a whole-body coil. The use of a dedicated phased array coil would have resulted in higher spatial resolution and probably decreased motion artifacts caused by shortening of acquisition time. Both types of coils do not have the limitations of transrectal coils, such as limited field of view and limited imaging of stenotic tumors.
In conclusion, transrectal sonography is of value in the preoperative staging of rectal cancer. Moreover, it is cheaper and more readily available than MR imaging and should therefore be performed as the standard examination in early-stage tumors in the lower part of the rectum. Double-contrast MR imaging is the method of choice in patients suspected of having advanced disease and in patients with tumors located in the upper rectum. Preoperative irradiation is an important limiting factor for both methods. However, both methods are accurate in predicting the presence of bowel wall penetration.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  C MESSIOU, A G CHALMERS, K BOYLE, D WILSON, and P SAGAR Pre-operative MR assessment of recurrent rectal cancer Br. J. Radiol., June 1, 2008; 81(966): 468 - 473. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X. M. Zhang, H. L. Zhang, D. Yu, Y. Dai, D. Bi, M. R. Prince, and C. Li 3-T MRI of Rectal Carcinoma: Preoperative Diagnosis, Staging, and Planning of Sphincter-Sparing Surgery Am. J. Roentgenol., May 1, 2008; 190(5): 1271 - 1278. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. J. A. Margolis, J. M. Hoffman, R. J. Herfkens, R. B. Jeffrey, A. Quon, and S. S. Gambhir Molecular Imaging Techniques in Body Imaging Radiology, November 1, 2007; 245(2): 333 - 356. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. P. Dhamanaskar, W. Thurston, and S. R. Wilson Transvaginal sonography as an adjunct to endorectal sonography in the staging of rectal cancer in women. Am. J. Roentgenol., July 1, 2006; 187(1): 90 - 98. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
