AJR 2005; 184:531-538
© American Roentgen Ray Society
Thin-Section MRI with a Phased-Array Coil for Preoperative Evaluation of Pelvic Anatomy and Tumor Extent in Patients with Rectal Cancer
Takayuki Akasu1,
Gen Iinuma2,
Toshiyuki Fujita2,
Yukio Muramatsu2,
Ukihide Tateishi2,
Kunihisa Miyakawa2,
Tsutomu Murakami2 and
Noriyuki Moriyama2
1 Colorectal Surgery Division, National Cancer Center Hospital, 5-1-1, Tsukiji,
Chuo-ku, Tokyo 104-0045, Japan.
2 Diagnostic Radiology Division, National Cancer Center Hospital, Tokyo
104-0045, Japan.
Received February 8, 2004;
accepted after revision June 2, 2004.
Address correspondence to T. Akasu.
Supported in part by a Grant-in-Aid for Clinical Research for
Evidence-Based Medicine and a Grant-in-Aid for Cancer Research from the
Ministry of Health, Labor, and Welfare and a grant from the Foundation for
Promotion of Cancer Research in Japan.
Abstract
OBJECTIVE. The aim of our study was to assess the accuracy of
thin-section MRI performed with a phased-array coil as a technique for the
preoperative evaluation of pelvic anatomy and tumor extent in patients with
rectal cancer.
CONCLUSION. Thin-section MRI with a phased-array coil is accurate
and reliable for preoperative evaluation of pelvic anatomy and depth of
transmural tumor invasion. Thus, it may be helpful in the selection of the
appropriate treatment for patients with rectal cancer.
Introduction
The principal problems associated with rectal cancer treatment are tumor
recurrence and impairment of anorectal and genitourinary functions after
surgery. For a patient with rectal cancer to achieve a better prognosis and
quality of life, the extent of surgery should accurately reflect the disease
status. The internal and external anal sphincters, which are essential for
anorectal function, are adjacent to the rectum. The pelvic autonomic nervous
system—consisting of the hypogastric plexus, hypogastric nerves, and
pelvic plexuses—is essential for genitourinary functions and is adjacent
to the mesorectal fascia surrounding the mesorectum
[1]. The mesorectum is defined
as the lymphovascular, fatty, and neural tissue that is circumferentially
adherent to the rectum [2].
Therefore, excessive resection easily leads to unnecessary damage of anorectal
and genitourinary functions, whereas insufficient resection inevitably leads
to tumor recurrence. Indeed, reported incidences of permanent stoma, erectile
dysfunction, urinary dysfunction, and local recurrence generally are 34%
[3], 45%
[4], 58%
[5], and 22–27%
[6,
7], respectively. However, the
incidences of these outcomes in a series of patients who received ideal
treatment from experts were reported to be only 6%
[8], 13%
[9], 5%
[9], and 5–7%
[9,
10], respectively.
Treatment options should be selected according to the extent of the tumor.
In general, T1 tumors invading the superficial submucosa can be effectively
treated by local excision, which is minimally invasive and promises excellent
maintenance of anorectal and genitourinary functions
[11]. T1 tumors invading the
deep submucosa, T2 tumors invading the muscularis propria, or T3 tumors
invading the perirectal fat slightly but remaining within the mesorectal
fascia can be treated by mesorectal excision, which maintains good
genitourinary functions and fair anorectal function if the anal sphincter can
be preserved
[8–11].
Patients with T3 tumors invading the mesorectal fascia or T4 tumors invading
the neighboring organs require more radical surgery, and preservation of
genitourinary functions is more difficult.
Randomized controlled studies have shown that adjuvant preoperative
radiation therapy is effective for reducing local recurrence and prolonging
survival in patients with rectal cancers, especially those with T3 tumors or
node-positive cancer [6,
7]. Thus preoperative radiation
therapy is becoming standard treatment for advanced rectal cancer. However,
surgery alone can achieve local control in almost all T1 or T2 tumors and in
many cases in T3 tumors as well. In addition, radiation therapy is complicated
by toxicity [12], so the
adjuvant therapy adopted also should reflect the accurate disease status.
The extent of tumor spread is generally evaluated using digital
examination, endorectal sonography, CT, and MRI. The accuracy rates of
endorectal sonography in the evaluation of the depth of transmural tumor
invasion have been reported to be 82–88%
[13,
14], and the technique has
been described as superior to others for preoperative staging
[15–17].
However, endorectal sonography is not applicable for stenosing tumors; further
improvements are necessary for optimum tailoring of treatment for the
individual patient.
Recent advances in medical imaging have shown that thin-section MRI
performed with a phased-array coil is accurate and useful for preoperative
evaluation of the extent of rectal cancer
[18,
19]. Thus, we used a new
phasedarray coil that originally was developed to permit the early diagnosis
of pancreatic cancer. Our previous study [unpublished] showed that this coil
is superior to the conventional body coil, as indicated by the signal
intensity distributions. The purpose of this study was to evaluate accuracy of
thin-section MRI performed with this coil for the preoperative evaluation of
pelvic anatomy and tumor extent in patients with rectal cancer.
Subjects and Methods
Between June 2001 and April 2002, 34 consecutive patients with primary
rectal cancer proven by biopsy were examined with thin-section MRI using a
phased-array coil for the preoperative evaluation of tumor extent. The
patients were 25 men and nine women with a median age of 57 years (age range,
34–82 years). Of the 34 tumors in the patients, two were in the upper
rectum, or 10–15 cm from the anal verge; seven were in the middle
rectum, or 5–10 cm from the anal verge; and 25 were in the lower rectum,
or less than 5 cm from the anal verge. None of the patients received
preoperative radiation therapy. Informed consent was obtained from all
patients.
MRI was performed preoperatively and interpreted by one gastrointestinal
radiologist and one colorectal surgeon who were blinded to the findings of the
digital rectal examination, endorectal sonography, and CT. The resected
specimens were histopathologically examined by pathologists who were blinded
to the findings of the preoperative evaluation of tumor extent. The depth of
transmural tumor invasion was assessed according to the TNM classifications
[20]
(Table 1) for both MRI and
histopathologic examinations, and results were compared prospectively.
MRI Methods
The patients received a 150-mL glycerin enema before examination and were
placed in a supine, head-first position. No air insufflation was used, but an
intramuscular antispasmodic was administered. We used a 1.5-T whole-body
system (VISART/EX Scanner, Toshiba Medical Systems) and placed a wraparound
quadrature phased-array coil (Pancreatic QD paired array coil, Toshiba Medical
Systems) at the patient's pelvis. Initially, sagittal T2-weighted fast
spin-echo images (TR/TE, 4,000/120; echo-train length, 23; slice thickness, 6
mm; gap, 1.2 mm; signal averages; 4; matrix, 166 x 256; field of view,
15 x 15 cm) of the pelvis were obtained. These images were used to plan
T2-weighted thin-section axial imaging. Axial T2-weighted thin-section fast
spin-echo images (9,500/120; echo-train length, 23; slice thickness, 3 mm;
gap, 0 mm; signal averages; 4; matrix, 166 x 256; field of view, 15
x 15 cm) of the pelvis were then obtained.
MR Image Interpretation
One experienced gastrointestinal radiologist and one experienced colorectal
surgeon who had no knowledge of the clinical and histopathologic data
interpreted each MR image in consensus on the workstation monitor. Distance
was measured with electronic calipers. The reviewers assessed the
visualization of the rectal mucosa, submucosa, muscularis propria (inner
circular and outer longitudinal muscle layers), and mesorectal fascia; depth
of the transmural invasion by the tumor; mesorectal involvement by the tumor;
visualization of the branches of the named arteries such as the superior
rectal and the internal iliac arteries; visualization of the mesorectal and
extramesorectal lymph nodes; numbers of detected lymph nodes; and smallest
short-axis diameters of the lymph nodes.
The depth of transmural invasion by each tumor was categorized according to
the TNM classification [20]
(Table 1) and was assessed
according to the reported criteria
[18]
(Table 2). In accordance with
the findings of Brown et al.
[18], we did not regard the
presence of spiculation within the fat alone as sufficient evidence of
extramural invasion. Small interruptions of the outer contours of the muscle
coat were also not regarded as sufficient for diagnosis of a T3 lesion. To
further evaluate agreement in the assessment of invasion depth, reviewers
performed second interpretations after an interval of at least 4 months.
Histopathologic Study
All patients underwent radical surgery. The median interval between MRI and
surgery was 22 days (range, 1–55 days). Procedures performed were
mesorectal excision
[8–10]
in 30 patients (low anterior resection in 24 and abdominoperineal resection in
six), pelvic exenteration in three, and pelvic exenteration with partial
sacrectomy in one. Immediately after surgery, resected specimens were opened
on the side opposite the tumor and fixed in 10% formalin. After fixation, we
obtained serial slices through the whole tumor in Tis–T2 cases or
through more than two sections of the deepest part of the tumor in T3 or T4
cases. The slices were embedded in paraffin, sectioned, and examined
histologically after H and E staining. The depth of transmural tumor invasion
was classified according to the TNM classification
(Table 1)
[20].
Identification of the Pelvic Plexuses
Postoperative MR images were compared with ones obtained preoperatively in
two patients so that the exact locations of the pelvic plexuses—which
are essential for genitourinary function—could be identified. During
surgery, metal hemostatic clips had been applied to the cut ends of the middle
rectal arteries and veins on the inner surfaces of the pelvic plexuses. These
clips facilitated identification of the pelvic plexuses on postoperative MR
images.
Statistical Methods
The agreement regarding MRI-determined and histologically determined tumor
stage was assessed with the weighted kappa statistic, as was the agreement
between the first and second interpretations.
Results
All patients tolerated the thin-section MRI examination well. The total
scanning time was about 20 min. Although motion artifacts complicated findings
in five patients (15%), the images were of sufficient quality to allow
assessment. The histologic diagnoses were well-differentiated adenocarcinoma
in 11 patients, moderately differentiated adenocarcinoma in 16, poorly
differentiated adenocarcinoma in two, mucinous adenocarcinoma in four, and
linitis plastica carcinoma in one. The histologic transmural invasion depths
were pT1 in four patients, pT2 in nine, pT3 in 15, and pT4 in six. The
mesorectal fascia was involved in eight patients. The median tumor diameter
was 4.1 cm (range, 1.5–9.0 cm).
Visualization of the Pelvic Anatomy
In all patients, the rectal mucosa was visualized as a low-intensity layer;
the submucosa, as a high-intensity layer; the muscularis propria, as a
low-intensity layer; and the perirectal fat, as a high-intensity layer (Fig.
1A,
1B). However, the inner
circular muscle and outer longitudinal muscle layers could be distinguished
only in three patients (9%). The mesorectal fascia was consistently depicted
as a fine linear hypointense structure enveloping the mesorectum in all
patients (Fig. 2A). In all
patients, the internal and external sphincter muscles were shown as
low-intensity layers separated by a hyperintense intersphincteric plane
(Fig. 2C).
View larger version (130K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A. —64-year-old woman with pT3 rectal carcinoma. Unenhanced
T2-weighted fast spin-echo image shows rectal mucosa (m) as low-intensity,
submucosa (sm) as high-intensity, muscularis propria (mp) as low-intensity,
and perirectal fat (pf) as high-intensity layers. Signal intensity of tumor
(T) is higher than that of proper muscle layer but lower than that of
submucosa. Tumor is seen invading through muscularis propria
(arrowheads).
|
|
View larger version (162K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A. —42-year-old man with pT2 rectal carcinoma. Unenhanced
T2-weighted fast spin-echo image shows mesorectal fascia (arrowheads)
as fine linear hypointense structure enveloping mesorectum. Tumor (T) is
revealed as being confined in muscularis propria (mp) and was staged as
T2.
|
|
View larger version (158K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2C. —42-year-old man with pT2 rectal carcinoma. Unenhanced
T2-weighted fast spin-echo image shows internal sphincter muscle (i) and
puborectalis muscle (p) as low-intensity layers separated by hyperintense
intersphincteric plane.
|
|
The first, second, third, and fourth branches of the superior rectal artery
were seen as hypointense vascular structures in 34 (100%), 34 (100%), 31
(91%), and 11 patients (32%), respectively (Figs.
3A,
3B,
3C,
3D). The bilateral obturator
arteries branching from the internal iliac arteries were shown as hypointense
vascular structures in all patients (Fig.
3E).
View larger version (124K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A. —65-year-old man with rectal carcinoma. Unenhanced T2-weighted
fast spin-echo images reveal main trunk (A, arrowhead) and
first (B, arrowheads), second (C, arrowheads),
and third (D, arrowheads) branches of superior rectal artery
seen as hypointense vascular structures. e = external iliac artery, i =
internal iliac artery.
|
|
View larger version (136K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3B. —65-year-old man with rectal carcinoma. Unenhanced T2-weighted
fast spin-echo images reveal main trunk (A, arrowhead) and
first (B, arrowheads), second (C, arrowheads),
and third (D, arrowheads) branches of superior rectal artery
seen as hypointense vascular structures. e = external iliac artery, i =
internal iliac artery.
|
|
View larger version (136K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3C. —65-year-old man with rectal carcinoma. Unenhanced T2-weighted
fast spin-echo images reveal main trunk (A, arrowhead) and
first (B, arrowheads), second (C, arrowheads),
and third (D, arrowheads) branches of superior rectal artery
seen as hypointense vascular structures. e = external iliac artery, i =
internal iliac artery.
|
|
View larger version (141K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3D. —65-year-old man with rectal carcinoma. Unenhanced T2-weighted
fast spin-echo images reveal main trunk (A, arrowhead) and
first (B, arrowheads), second (C, arrowheads),
and third (D, arrowheads) branches of superior rectal artery
seen as hypointense vascular structures. e = external iliac artery, i =
internal iliac artery.
|
|
View larger version (141K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3E. —65-year-old man with rectal carcinoma. Unenhanced T2-weighted
fast spin-echo image reveals obturator lymph node (arrowhead) and
mesorectal lymph node (black arrow), displaying lower signal
intensity than that of perirectal fat but higher signal intensity than those
of arteries and veins. o = obturator artery.
|
|
The lymph nodes were identified as having lower signal intensity than the
perirectal fat but as having higher signal intensity than the arteries and
veins (Fig. 3E). In patients
with mucinous carcinoma, metastatic lymph nodes were shown as hyperintense
nodules alone or as hyperintense nodules within hypointense nodules. The
shapes of the lymph nodes were spherical or spheroidal, so that they could be
distinguished easily from vascular structures. The mesorectal lymph nodes were
apparent in all patients (Fig.
3E); the median number detected was five (range of nodes detected,
1–12). The median short-axis diameter of the smallest detected lymph
nodes was 2.7 mm (range, 1.3–8.3 mm). The iliac or obturator lymph nodes
were detected in only nine patients (33%)
(Fig. 3E); the median number
detected was 0 (range of nodes detected, 0–4). The median short-axis
diameter of the smallest detected lymph nodes was 0 mm (range, 0–8.2
mm).
Comparisons of preoperative and postoperative MR images showed the pelvic
plexuses to be located just outside the mesorectal fascia (Figs.
4A and
4B). However, even with metal
hemostatic clips applied during surgery, the plexuses themselves could not be
visualized on thin-section MRI.
View larger version (155K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4A. —42-year-old man with rectal carcinoma. Comparison of pre- and
postoperative MR images show pelvic plexuses are located just outside
mesorectal fascia. MR image obtained before surgery (A) shows pelvic
plexuses (white arrows). Postoperative MR image (B) shows one
of metal hemostatic clips that were applied to inner surfaces of pelvic
plexuses during surgery to mark their exact locations (black
arrow).
|
|
View larger version (147K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4B. —42-year-old man with rectal carcinoma. Comparison of pre- and
postoperative MR images show pelvic plexuses are located just outside
mesorectal fascia. MR image obtained before surgery (A) shows pelvic
plexuses (white arrows). Postoperative MR image (B) shows one
of metal hemostatic clips that were applied to inner surfaces of pelvic
plexuses during surgery to mark their exact locations (black
arrow).
|
|
Assessment of the Depth of Transmural Tumor Invasion
All rectal cancers were detected on thin-section MRI and, in most patients,
showed higher signal intensity than the proper muscle layer but lower signal
intensity than the submucosa (Fig.
1A). However, linitis plastica carcinoma showed signal intensity
as low as that of the proper muscle layer, and mucinous carcinoma showed a
signal intensity that was higher than that of the submucosa in parts of the
mucous lakes.
At the first interpretation, MRI staging agreed with the histologic staging
in 28 (82%) of 34 patients (weighted 
= 0.82; 95% confidence interval
[CI], 0.69–0.95). Detailed results of the MRI staging are shown in
Table 3. Sensitivity,
specificity, overall accuracy rate, positive predictive value, and negative
predictive value for detection of proper muscle invasion (T2) were 97%
(29/30), 100% (4/4), 97% (33/34), 100% (29/29), and 80% (4/5), respectively
(Fig. 2A,
2B,
2C). Those values for detection
of perirectal fat invasion (T3) were 95% (20/21), 77% (10/13), 88% (30/34),
87% (20/23), and 91% (10/11), respectively (Fig.
1A,
1B). For detection of adjacent
organ invasion (T4), the respective values were 100% (6/6), 96% (27/28), 97%
(33/34), 86% (6/7), and 100% (27/27). The values for detection of the
mesorectal fascia involvement were 100% (8/8), 100% (26/26), 100% (34/34),
100% (8/8), and 100% (26/26), respectively ( = 1.0) (Fig.
5A,
5B).
At the second interpretation, MRI staging agreed with the histologic
staging in 29 (85%) of 34 patients (weighted = 0.85; 95% CI,
0.74–0.97). Sensitivity, specificity, overall accuracy rate, positive
predictive value, and negative predictive value for detection of proper muscle
invasion (T2), adjacent organ invasion (T4), and mesorectal fascia involvement
were the same as those for the first interpretation. Those values for
detection of perirectal fat invasion (T3) were 95% (20/21), 85% (11/13), 91%
(31/34), 91% (20/22), and 92% (11/12), respectively. The agreement of the
first and second interpretations on the depth of transmural invasion depth was
good ( = 0.87; 95% CI, 0.73–1.0).
Of the six cases in which staging errors were encountered at the first
interpretation, four were overstaged, and two were understaged
(Table 3). Histologic review of
the specimens revealed that in three of the overstaged cases, the tumor
invaded close to the deeper uninvolved layer and reactive changes were present
in the connective tissue around the tumor, including inflammatory cell
aggregation, desmoplastic change, and hypervascularity (Fig.
6A,
6B). In addition, the deepest
part of the tumor was not sectioned vertically on MRI but was sectioned
obliquely, so that interpretation was difficult (Fig.
7A,
7B). Histologic review of the
two understaged cases revealed that they had only microscopic invasion beyond
the estimated involved layers and that reactive changes of the connective
tissue around the tumor were either only very slight or absent.
View larger version (138K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6A. —80-year-old man with pT2 rectal carcinoma. Tumor (T) was
overstaged as T3 because spiculation (arrowheads) was interpreted as
cancer invasion on unenhanced T2-weighted fast spin-echo image.
|
|
View larger version (58K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6B. —80-year-old man with pT2 rectal carcinoma. Photograph of
histologic specimen reveals tumor confined in muscularis propria (stage pT2).
However, reactive changes in connective tissue around tumor, including
desmoplastic change and hypervascularity (arrows), can affect MRI
findings and mimic tumor invasion.
|
|
View larger version (145K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 7A. —56-year-old woman with pT2 rectal carcinoma. Tumor (T) was
overstaged as T3 because site of deepest invasion (arrowheads) was
sectioned obliquely on MRI and mimicked cancer invasion beyond muscularis
propria (mp).
|
|
Discussion
As these results show, thin-section MRI performed with a quadrature
phased-array coil has sufficient accuracy to depict fine details of the rectal
wall (mucosa, submucosa, and muscularis propria), the anal sphincter, the
mesorectum (perirectal fat; superior rectal artery and vein and their
branches; lymph node; and mesorectal fascia), and the extramesorectal
structures (internal iliac artery and vein and their branches; and lymph node)
clearly in every patient. Fourth branches of the inferior mesenteric artery
and lymph nodes measuring 2 mm could be visualized in most patients. In
addition, although the pelvic plexuses per se could not be visualized on our
thin-section MRI, we identified their exact locations just outside the
mesorectal fascia via metal hemostatic clips placed on their inner surfaces
during surgery and comparisons of preoperative and postoperative MR
images.
Previous studies using similar instruments also provided precise images of
the rectal and pelvic anatomy
[18,
19]. Brown et al.
[18] reported that their
technique had an in-plane resolution of 0.6 x 0.6 mm and allowed
differentiation of the inner circular and outer longitudinal muscle layers. We
could distinguish the layers in only 9% of the patients, but such
differentiation is not clinically important because treatment for the tumor
invading the inner muscle is the same as that for the tumor invading the outer
muscle.
All intraluminal cancers measuring more than 1.5 cm were detected. Most
tumors showed a signal intensity that was higher than that of the proper
muscle layer but lower than that of the submucosa, as has been reported
previously [18,
19]. In addition, we found
that linitis plastica carcinoma had a signal intensity that was as low as that
of the proper muscle layer and that mucinous carcinoma had a signal intensity
higher than that of the submucosa in parts of the mucous lakes. These findings
are useful for predicting histologic diagnosis and may contribute to treatment
selection because they are risk factors for a poor prognosis
[21–23].
However, whether the histology of the tumor affects staging accuracy could not
be determined because of the limited number of patients studied.
In our prospective study, we performed unenhanced thin-section MRI (slice
thickness, 3 mm) on a 1.5-T scanner with a quadrature phased-array coil. The
depth of transmural tumor invasion and mesorectal fascia involvement were
predicted correctly in 82% and 100% of the patients, respectively. In their
retrospective evaluation, Beets-Tan et al.
[19] used contrast-enhanced
thin-section MRI (slice thickness, 3 mm) on a 1.5-T scanner with a quadrature
phased-array spine coil and reported that the depth of transmural tumor
invasion and mesorectal fascia involvement were predicted correctly in 83% and
100% of their patients, respectively. Brown et al.
[18] used unenhanced
thin-section MRI (slice thickness, 3 mm) on a 1.5-T scanner and a four-element
flexible wraparound surface coil and conducted a retrospective study that
found correct invasion depth assessment was attained in 100% of their cases.
Thus, thin-section MRI performed on a 1.5-T scanner with a phased-array coil
in general can be considered to provide moderate to good accuracy in the
prediction of invasion depth and good accuracy in the prediction of mesorectal
fascia involvement. These data are comparable to accuracy rates of
82–88% [13,
14] obtained with endorectal
sonography for the prediction of invasion depth. However, endorectal
sonography is not applicable for stenotic or obstructive tumors and cannot
visualize the mesorectal fascia and obturator space because of the limitations
of sonographic attenuation
[14]. In addition,
good-quality sonograms can be guaranteed only if the images are acquired by a
skilled operator [14].
Therefore, thin-section MRI can be concluded to be clinically more useful than
endorectal sonography.
As to reproducibility, we did not evaluate interobserver agreement, but
concordance between the first and second interpretations was good for both
invasion depth ( = 0.87) and mesorectal fascia involvement ( =
1.0). Brown et al. [18]
evaluated only interobserver agreement and reported good agreement between
experienced reviewers for invasion depth ( = 1.0). Beets-Tan et al.
[19] assessed both
intraobserver and interobserver agreement. For assessment of invasion depth,
intraobserver agreement was good ( = 0.8) for a radiologist experienced
in pelvic MRI but was only moderate ( = 0.49) for an inexperienced
radiologist; interobserver agreement was moderate ( = 0.53). In
contrast, intraobserver and interobserver agreements for the prediction of
involvement of circumferential resection margin
[24–26]
(the same as mesorectal fascia involvement in patients who undergo mesorectal
excision
[8–10])
were good, because intraclass correlation coefficients for the experienced
reviewer, inexperienced reviewer, and both reviewers were 0.99, 0.91, and
0.93, respectively. Therefore, examinations for invasion depth should be
interpreted by a reviewer experienced in pelvic MRI; involvement of the
circumferential resection margin or mesorectal fascia is more easily
interpretable.
Thin-section MRI is sufficiently accurate and reliable to provide
clinically useful information. Prediction of involvement of the mesorectal
fascia, adjacent organs, or circumferential resection margin is especially
important
[24–26].
Involvement of these structures requires surgery more radical than mesorectal
excision
[8–10],
preoperative adjuvant therapy, or both to reduce local recurrence and overall
recurrence [27]. Prediction of
an absence of such involvement allows performance of mesorectal excision alone
[8–10],
reducing the incidence and severity of anal and genitourinary dysfunctions
[9] and preventing toxicity
from unnecessary adjuvant radiation therapy
[28,
29], chemotherapy, or both.
Accurate prediction of invasion depth of T1 tumors ensures proper assignment
of candidates for local excision to enhance patient survival and quality of
life [11].
Although thin-section MRI is very accurate, it is not perfect. In our
series, two thirds of staging errors in invasion depth resulted from
overstaging and were most common with pT2 tumors, as has been reported for
endorectal sonography [13,
14]. Reactive changes in the
connective tissue around the tumor, including inflammatory cell aggregation,
desmoplastic change, and hypervascularity, mimic tumor invasion on MR images.
Such reactive changes have also been previously noted as a main cause of
overstaging on sonography [14,
30] and MRI
[18,
19]. Contrast enhancement may
be helpful for differentiating these reactive changes from true tumor
invasion. However, Beets-Tan et al.
[19], who used gadolinium as a
contrast medium, reported that MRI could not be used to distinguish reliably
between fibrosis with and fibrosis without tumor cells. The best results were
reported by Brown et al. [18],
who could differentiate between desmoplastic spiculation and true invasion.
Therefore, the best technique may be the one described in their report or may
involve more precise image acquisition and administration of effective
contrast material. In addition, the direction of MRI sectioning is important.
Obliquely sectioned images make contours of tumors obscure and interpretation
difficult, as seen in our study. This difficulty may be overcome by more
precise image acquisition and 3D data accumulation.
One third of the staging errors in our study involved underestimation that
was mostly attributable to microscopic invasion that is fundamentally
undetectable on MRI or difficulties in attaining a complete examination with
the 2D rather than 3D approach, so that we obtained not continuous images but
rather interrupted images. To reduce overstaging and understaging,
investigators need to address the possibility of using an image matrix smaller
than 166 x 256, a slice width thinner than 3 mm, techniques for
achieving a higher signal-to-noise ratio, 3D data accumulation, effective
contrast material, and a shorter scanning time. MRI with an endorectal coil
may have higher signal-to-noise ratio near the coil and produces better
visualization of the rectal wall structure
[31,
32]; however, its limited
field of view makes assessment of the mesorectal fascia and surrounding
structures difficult, and insertion of the coil is difficult in patients with
annular stenotic lesions. Therefore, approaches using thin-section MRI with a
phased-array coil still seem better.
Although our study concerned a relatively small number of patients, we
conclude that thin-section MRI with a phased-array coil is accurate and
reliable for the preoperative evaluation of the pelvic anatomy and the depth
of transmural tumor invasion Thus, it may be helpful in the selection of the
appropriate treatment for patients with rectal cancer. However, the accuracy
of this technique is not perfect, so further investigation to improve accuracy
is warranted. In addition, for validation, a multiinstitutional prospective
study is necessary.
References
- Havenga K, DeRuiter MC, Enker WE, Welvaart K. Anatomical basis of
autonomic nerve-preserving total mesorectal excision for rectal cancer.
Br J Surg 1996;83:384
–388[Medline]
- Nelson H, Petrelli N, Carlin A, et al. Guidelines 2000 for colon
and rectal cancer surgery. J Natl Cancer Inst2001; 93:583
–596[Abstract/Free Full Text]
- Dahlberg M, Glimelius B, Pahlman L. Changing strategy for rectal
cancer is associated with improved outcome. Br J Surg1999; 86:379
–384[Medline]
- Walsh P, Schlegel PN. Radical pelvis surgery with preservation of
sexual function. Ann Surg1988; 208:391
–400[Medline]
- Hojo K, Sawada T, Moriya Y. An analysis of survival and voiding,
sexual function after wide iliopelvic lymphadenectomy in patients with
carcinoma of the rectum, compared with conventional lymphadenectomy.
Dis Colon Rectum1989; 32:128
–133[Medline]
- Colorectal Cancer Collaborative Group. Adjuvant radiotherapy for
rectal cancer: a systematic overview of 8,507 patients from 22 randomized
trials. Lancet2001; 358:1291
–1304[Medline]
- [No authors listed] Improved survival with preoperative
radiotherapy in resectable rectal cancer: Swedish Rectal Cancer Trial.
N Engl J Med1997; 336:980
–987[Abstract/Free Full Text]
- MacFarlane JK, Ryall RD, Heald RJ. Mesorectal excision for rectal
cancer. Lancet1993; 341:457
–460[Medline]
- Enker WE. Potency, cure, and local control in the operative
treatment of rectal cancer. Arch Surg1992; 127:1396
–1401[Abstract]
- Enker WE, Thaler HT, Cranor ML, Polyak T. Total mesorectal excision
in the operative treatment of carcinoma of the rectum. J Am Coll
Surg 1995;181:335
–346[Medline]
- Akasu T, Kondo H, Moriya Y, et al. Endorectal ultrasonography and
treatment of early stage rectal cancer. World J Surg2000; 24:1061
–1068[Medline]
- Otchy DP, Nelson H. Radiation injuries of the colon and rectum.
Surg Clin North Am1993; 73:1017
–1035[Medline]
- Solomon MJ, McLeod RS. Endoluminal transrectal ultrasonography:
accuracy, reliability, and validity. Dis Colon Rectum1993; 36:200
–205[Medline]
- Akasu T, Sugihara K, Moriya Y, Fujita S. Limitations and pitfalls
of transrectal ultrasonography for rectal cancer staging. Dis Colon
Rectum 1997;40[suppl 10]:S10
–S15[Medline]
- Beynon J, Mortensen NJM, Foy DMA, Channer JL, Virjee J, Goddard P.
Pre-operative assessment of local invasion in rectal cancer: digital
examination, endoluminal sonography or computed tomography? Br J
Surg 1986;73:1015
–1017[Medline]
- Waizer A, Powsner E, Russo I, et al. Prospective comparative study
of magnetic resonance imaging versus transrectal ultrasound for preoperative
staging and follow-up of rectal cancer: preliminary report. Dis
Colon Rectum 1991;34:1068
–1072[Medline]
- Thaler W, Watzka S, Martin F, et al. Preoperative staging of rectal
cancer by endoluminal ultrasound vs. magnetic resonance imaging: preliminary
results of a prospective comparative study. Dis Colon
Rectum 1994;37:1189
–1193[Medline]
- Brown G, Richards CJ, Newcombe RG, et al. Rectal carcinoma:
thin-section MR imaging for staging in 28 patients.
Radiology1999; 211:215
–222[Abstract/Free Full Text]
- Beets-Tan RG, Beets GL, Vliegen RF, et al. Accuracy of magnetic
resonance imaging in prediction of tumour-free resection margin in rectal
cancer surgery. Lancet2001; 357:497
–504[Medline]
- Sobin LH, Wittekind CH (International Union Against Cancer), eds.
TNM classification of malignant tumours, 5th ed.
Baltimore, MD: WileyLiss, 1997
- Shirouzu K, Isomoto H, Morodomi T, Ogata Y, Akagi Y, Kakegawa T.
Primary linitis plastica carcinoma of the colon and rectum.
Cancer 1994;74:1863
–1868[Medline]
- Younes M, Katikaneni PR, Lechago J. The value of the preoperative
mucosal biopsy in the diagnosis of colorectal mucinous adenocarcinoma.
Cancer 1993;72:3588
–3592[Medline]
- Kanemitsu Y, Kato T, Hirai T, et al. Survival after curative
resection for mucinous adenocarcinoma of the colorectum. Dis Colon
Rectum 2003;46:160
–167[Medline]
- Adam IJ, Mohamdee MO, Martin IG, et al. Role of circumferential
margin involvement in the local recurrence of rectal cancer.
Lancet 1994;344:707
–711[Medline]
- Birbeck KF, Macklin CP, Tiffin NJ, et al. Rates of circumferential
resection margin involvement vary between surgeons and predict outcomes in
rectal cancer surgery. Ann Surg2002; 235:449
–457[Medline]
- Wibe A, Rendedal PR, Svensson E, et al. Prognostic significance of
the circumferential resection margin following total mesorectal excision for
rectal cancer. Br J Surg2002; 89:327
–334[Medline]
- Marijnen CA, Nagtegaal ID, Kapiteijn E, et al. Radiotherapy does
not compensate for positive resection margins in rectal cancer patients:
report of a multicenter randomized trial. Int J Radiat Oncol Biol
Phys 2003;55:1311
–1320[Medline]
- Otchy DP, Nelson H. Radiation injuries of the colon and rectum.
Surg Clin North Am1993; 73:1017
–1035
- Cedermark B, Johansson H, Rutqvist LE, Wilking N. The Stockholm I
trial of preoperative short term radiotherapy in operable rectal carcinoma: a
prospective randomized trial—Stockholm Colorectal Cancer Study Group.
Cancer 1995;75:2269
–2275[Medline]
- Hulsmans FJH, Tio TL, Fockens P, Bosma A, Tytgat GNJ. Assessment of
tumor infiltration depth in rectal cancer with transrectal sonography: caution
is necessary. Radiology1994; 190:715
–720[Abstract/Free Full Text]
- Schnall MD, Furth EE, Rosato EF, Kressel HY. Rectal tumor stage:
correlation of endorectal MR imaging and pathologic findings.
Radiology1994; 190:709
–714[Abstract/Free Full Text]
- Vogl TJ, Pegios W, Mack MG, et al. Accuracy of staging rectal
tumors with contrast-enhanced transrectal MR imaging.
AJR 1997;168:1427
–1434[Abstract/Free Full Text]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
