AJR 2004; 182:1233-1240
© American Roentgen Ray Society
Small (
2 cm) Hepatic Lesions in Colorectal Cancer Patients: Detection and Characterization on Mangafodipir Trisodium–Enhanced MRI
Kyoung Won Kim1,
Ah Young Kim,
Tae Kyoung Kim,
Seong Ho Park,
Hyun Jin Kim,
Young Kyung Lee,
Mi-Suk Park,
Hyun Kwon Ha,
Pyo Nyun Kim,
Jin Cheon Kim and
Moon-Gyu Lee
1 All authors: Department of Radiology, University of Ulsan-Asan Medical Center,
388-1, Pungnap-dong, Songpa-ku, Seoul 138-736, Korea.
Received April 10, 2003;
accepted after revision October 29, 2003.
Address correspondence to M.-G. Lee
(mglee{at}amc.seoul.kr).
Abstract
OBJECTIVE. The purpose of this study was to evaluate whether
mangafodipir trisodium (MnDPDP)–enhanced MRI improves the detection and
characterization of small (
2 cm) hepatic lesions in patients with
colorectal carcinoma, compared with helical CT.
SUBJECTS AND METHODS. Sixty-nine patients who had or were suspected
of having focal liver lesions underwent helical CT and MnDPDP-enhanced MRI and
constituted the study population. Two experienced radiologists independently
reviewed CT and MR images for the number of hepatic lesions seen and whether
the lesion appeared to be benign or metastatic; their interpretations were
correlated with the reference diagnoses, including histopathologic diagnoses
in 35 patients. The lesions were categorized as small ( 2.0 cm in
diameter) or large (> 2.0 cm). The differences between MnDPDP-enhanced MRI
and helical CT with regard to the detection rates for hepatic lesions and
metastases and with regard to the false-positive rates for hepatic metastases
were analyzed using the McNemar test. The performances of MnDPDP-enhanced MRI
and helical CT in indicating metastases of focal liver lesions were analyzed
using receiver operating characteristic curves.
RESULTS. No statistically significant differences were seen between
MnDPDP-enhanced MRI and helical CT in the detection of all hepatic lesions
(p = 0.383) and small lesions (p = 0.210). However,
concerning the differentiation between benign and metastatic lesions,
MnDPDP-enhanced MRI was superior to helical CT both for all hepatic lesions
(p = 0.023) and for small lesions (p = 0.015), and remained
better when the analyses were restricted to patients with histopathologic
confirmation (p = 0.023 for both). MnDPDP-enhanced MRI changed the
diagnosis of hepatic metastasis in nine (13.0%) of 69 patients. Of 12
metastases that were found on MnDPDP-enhanced MRI and missed on helical CT, 11
lesions (91.7%) were small. MnDPDP-enhanced MRI showed a significantly greater
detection rate than helical CT for small hepatic metastases (p =
0.022). MnDPDP-enhanced MRI was better when the analyses were restricted to
patients with histopathologic confirmation (p = 0.043).
CONCLUSION. Although MnDPDP-enhanced MRI is equal to helical CT in
detection of both all hepatic lesions and small lesions in patients with
colorectal carcinoma, it is superior to CT in characterization of the
lesions.
Introduction
Helical CT is a primary imaging technique for preoperative staging of
patients with known or suspected colorectal carcinoma. However, although an
accurate assessment of hepatic metastases before surgery in patients with
colorectal carcinoma is important to avoid unnecessary hepatic surgery or
other treatment, sometimes the detection and characterization of hepatic
metastases are limited with helical CT alone, especially when the lesion is
small. In these cases, MRI is now generally accepted as the principal
technique for additional imaging evaluation of the liver. Many now believe
that, for both focal liver lesion detection and characterization,
contrast-enhanced MRI with gadolinium chelates
[1–3],
iron oxide compounds
[4–10],
and manganese chelates [11,
12] is superior to
contrast-enhanced multiphasic CT. However, although diagnostic difficulty in
liver lesion detection and characterization is often encountered with helical
CT, especially when the lesion is small, previous studies with
contrast-enhanced MRI using liver-specific contrast agents have not focused on
its value in assessing small hepatic lesions.
Of MRI contrast agents, mangafodipir trisodium (MnDPDP) has a special
affinity for hepatocytes. After IV administration, the MnDPDP chelate
dissociates slowly and manganese is taken up by the hepatocytes, which leads
to an increase in signal intensity of normal liver parenchyma on the
T1-weighted image caused by T1 shortening and thereby to an increase in
contrast between normal and abnormal tissue
[13–15].
The purpose of this study was to evaluate whether MnDPDP-enhanced MRI improves
the detection and characterization of small ( 2 cm) hepatic lesions,
compared with helical CT, in patients with colorectal carcinoma.
Subjects and Methods
Study Population
The study was approved by our institutional review board and informed
consent was obtained from each patient. Inclusion criteria were diagnosis of
colorectal carcinoma at presentation or a history of colorectal carcinoma and
suspected focal hepatic lesions on helical CT performed for preoperative
staging or for the evaluation of the presence or absence of tumor recurrence
during the follow-up period after radical surgery. Exclusion criteria included
contraindications to MRI (e.g., aneurysm clip, pacemaker), pregnancy or
lactation, or the presence of other existing or past malignancy. Between
August 2000 and January 2002, 69 consecutive patients were enrolled in this
study. Their mean age was 55.9 years (range, 31–77 years); 42 (61%) were
men and 27 (39%) were women. The time between helical CT and MnDPDP-enhanced
MRI ranged from 24 hr to 47 days (mean, 17 days).
Of these 69 patients, 33 underwent surgery; histopathologic findings,
together with the combination of careful surgical inspection, palpation of the
liver, and, sometimes, intraoperative sonography (n = 12),
constituted the reference diagnoses in these patients. In two patients,
reference diagnoses were histopathologic results of specimens obtained by
percutaneous liver biopsies. The average time between MnDPDP-enhanced MRI and
surgery or biopsy was 9 days (range, 1–56 days). In the remaining 34
patients, the reference diagnosis was formed by means of interpretation of all
available data in each patient, including all follow-up imaging (sonography,
CT, MRI, and other imaging if available) and levels of serum carcinoembryonic
antigen. Duration of follow-up was a minimum of 18 months and a maximum of 27
months.
Imaging Methods
CT examinations were performed with a helical scanner (Somatom Plus-S or
Somatom Plus 4, Siemens) (n = 48) or a multidetector scanner
(LightSpeed QX/i, General Electric Medical Systems) (n = 21). Images
were obtained with a biphasic protocol during the hepatic arterial phase and
the portal venous phase in 24 patients, or with a monophasic protocol during
the portal venous phase in the remaining 45 patients. Each patient received
100–120 mL of iopromide (Ultravist 370, Schering) through an 18-gauge
angiographic catheter inserted in a forearm vein using a mechanical injector
at a rate of 3 mL/sec. Scanning was performed with the following parameters:
120 kVp and 130–150 mAs; slice thickness, 5–8 mm; reconstruction
interval, 5–8 mm; and table pitch, 1:1. The hepatic arterial and portal
venous phase scans were obtained at 30 and 70 sec, respectively, after the
initiation of the injection of contrast material. Each helical acquisition
through the liver was obtained in the craniocaudal direction and was
accomplished in a single breath-hold (after maximal inspiration).
MRI was performed with a 1.5-T unit (Magnetom Vision, Siemens). After
initial scout images were obtained, unenhanced T1-weighted spoiled
gradient-echo imaging (fast low-angle shot [FLASH]) was performed with the
following parameters: TR/TE, 149/4.1; flip angle, 80°; breathhold, 19 sec;
19 sections; section thickness, 8 mm; intersection gap, 2 mm; field of view,
350 mm; and matrix, 132 x 256. T2-weighted imaging was performed with
both true fast imaging with steady-state free precession (FISP) and HASTE
sequences. Parameters for true FISP sequences were as follows: 4.8/2.3; flip
angle, 70°; breath-hold, 17 sec; 19 sections; section thickness, 8 mm;
gap, 2 mm; field of view, 350 mm; and matrix, 150 x 256. For the HASTE
sequences, parameters were as follows: TR/effective TE, infinite/134; echo
spacing, 4.6 msec; echo-train length, 104; flip angle, 150°; breath-hold,
20 sec; 15 sections; section thickness, 8 mm; gap, 2 mm; field of view, 350
mm; and matrix, 192 x 256. MnDPDP (Teslascan, Nycomed Imaging), 0.005
mmol/kg of body weight (0.5 mL/kg; maximum dose, 15 mL), was administered
IV (slow infusion at 2.5 mL/min). Coupled MnDPDP-enhanced MR images were
obtained at 30 min and 2 hr after administration of the contrast agent with
the FLASH sequence using the same protocol as described before. In all
participants, MRI acquisition through the liver was accomplished in the
craniocaudal direction during a single breath-hold.
Image Interpretation
All helical CT scans and MR images were collected and organized for image
interpretation. Every study had at least one lesion. Two radiologists (5
years' experience each) who were unaware of any clinical, laboratory, or other
imaging information except that the patients presented with or had a history
of colorectal carcinoma, participated in the image analysis. The MR and CT
images were independently reviewed and interpreted by each observer at two
discrete sessions separated by an 8-month interval. All CT and MR images were
evaluated on a PACS (picture archiving and communicating system) (Radpia,
Hyundai Information Technology) with the annotations masked. All MR images
(i.e., those obtained both before and after MnDPDP enhancement) were combined
for MRI interpretation. Each observer recorded the number of hepatic lesions
seen, their size, and the segmental location according to the Couinaud
classification. Each lesion was categorized as being small ( 2.0 cm in
diameter) or large (> 2.0 cm). The reviewers were also asked to decide
whether the lesions appeared to be benign or metastatic and to rate their
confidence level for diagnosing the presence of metastasis on a 5-point scale
(definitely not, probably not, possibly, probably, definitely) on the basis of
previous reports that showed the characteristic appearance of the most common
focal liver lesions on the helical CT scan
[16,
17] or the MnDPDP-enhanced MR
image
[13–15,
18,
19]. If more than five lesions
were present, the organizer assigned the most important five lesions
clinically, primarily according to their sizes: lesions that ranged from 1.5
to 2.5 cm were selected.
Results of the interpretations by the two observers were collected and
compared with each other. For cases in which the two observers reached
agreement, their agreement was accepted as the final interpretation. In case
of disagreement, a consensus panel consisting of the two original radiologists
plus a third radiologist (9 years' experience), who was also blinded, decided
the final interpretation.
Statistical Analyses
All statistical analyses were performed, both for all patients and for
those with histopathologic confirmation, using standard statistical software
(SPSS for Windows [Microsoft], version 10.0). For each imaging technique, the
detection rates for identification of focal hepatic lesions and hepatic
metastases and the false-positive rates were calculated using the reference
diagnosis as the standard reference. The differences between MnDPDP-enhanced
MRI and helical CT with regard to the detection rates for hepatic lesions and
metastases and with regard to the false-positive rates for hepatic metastases
(percentage of patients with at least one false-positive diagnosis) were
analyzed using the McNemar test, with significance being a p value of
less than 0.05. The diagnostic performances of MnDPDP-enhanced MRI and helical
CT in indicating metastases of focal liver lesions were compared with each
other using a nonparametric receiver operating characteristic (ROC) curve
analysis [20,
21]. According to the 5-point
scale, "definitely not" and "probably not" were
considered benign and "possibly," "probably," and
"definitely" were considered metastatic for the calculation of the
detection rates. The area under an ROC curve (Az) and its
95% confidence interval were calculated for both MnDPDP-enhanced MRI and
helical CT, and the two curves were compared. A p value of less than
0.05 was regarded as statistically significant.
Results
A total of 181 lesions were found in 69 patients (average, 2.6 lesions per
patient). The lesions ranged from 0.2 to 12.5 cm (mean, 1.8 cm) in diameter.
The size distribution of benign and metastatic lesions is shown in
Table 1. Sixty-two lesions were
in the left lobe, and the other 119 were in the right lobe. Of these, 113
metastases were found in 49 patients (average, 2.3 metastases per patient).
The other lesions were diagnosed as 52 cysts, nine cirrhotic nodules, six
hemangiomas, and one focal fat deposition. Histopathologic information was
available for the reference diagnosis in 67 lesions (37%) in 35 patients,
including 58 metastases in 29 patients. For the other 114 lesions (63%),
including 55 metastases in 13 patients (average, 4.2 metastases per patient),
biopsy was considered unethical because the lesions were considered definitely
benign or because the patient's condition was inoperable as a result of
multiple metastatic lesions that showed characteristic appearances on both
helical CT and MnDPDP-enhanced MRI. Fifty-four patients had at least one small
hepatic lesion. A total of 137 small hepatic lesions were found in 54 patients
(average, 2.5 lesions per patient). Their mean diameter was 1.1 cm (range,
0.2–2.0 cm). Forty-six lesions were in the left lobe, and 91 were in the
right lobe. Of these, 77 metastases were found in 37 patients (average, 2.1
metastases per patient). The other lesions were diagnosed as 49 cysts, six
cirrhotic nodules, four hemangiomas, and one focal fat deposition.
Histopathologic information was available in 48 lesions (35%) in 24 patients,
including 40 small hepatic metastases in 22 patients. Illustrative cases are
shown in Figures 1A,
1B,
1C,
1D,
2A,
2B,
2C,
3A,
3B,
3C,
3D,
3E,
3F,
3G,
3H,
4A,
4B,
4C.
View larger version (119K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A. —57-year-old woman with hepatic metastasis from colorectal
carcinoma. CT scan during portal venous phase shows hypoattenuating mass
(arrow) with faint peritumoral rim enhancement (arrowhead)
in left lobe of liver.
|
|
View larger version (110K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1C. —57-year-old woman with hepatic metastasis from colorectal
carcinoma. Fast low-angle shot (FLASH) T1-weighted image 30 min after
administration of mangafodipir trisodium (MnDPDP) shows no enhancement of
lesion. Also, peritumoral rim enhancement is not seen.
|
|
View larger version (123K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1D. —57-year-old woman with hepatic metastasis from colorectal
carcinoma. FLASH T1-weighted image 2 hr after administration of MnDPDP also
shows no intratumoral or peritumoral enhancement. Final histopathologic
diagnosis was single liver metastasis from colorectal carcinoma.
|
|
View larger version (189K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A. —54-year-old man with hepatic metastasis from colorectal
carcinoma. Helical CT scan during portal venous phase shows hypoattenuating
nodule (arrow) that is too small to characterize in right lobe of
liver.
|
|
View larger version (152K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2C. —54-year-old man with hepatic metastasis from colorectal
carcinoma. On fast low-angle shot T1-weighted image at 2 hr after
administration of mangafodipir trisodium, intense peritumoral rim enhancement
(arrowhead) is seen that is suggestive of hepatic metastasis. Final
histopathologic diagnosis was single liver metastasis from colorectal
carcinoma.
|
|
View larger version (176K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A. —49-year-old man with hepatic cyst and metastasis from
colorectal carcinoma. Helical CT scan during hepatic arterial phase shows
small hypoattenuating nodule (arrow) in liver segment IV.
|
|
View larger version (165K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3B. —49-year-old man with hepatic cyst and metastasis from
colorectal carcinoma. Helical CT scan during hepatic arterial phase also shows
poorly defined lesion (arrowheads) with faint high attenuation in
segment V.
|
|
View larger version (151K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3E. —49-year-old man with hepatic cyst and metastasis from
colorectal carcinoma. True fast imaging with steady-state free precession
(FISP) T2-weighted MR image shows segment IV lesion with bright high signal
intensity (arrow), suggestive of hepatic cyst.
|
|
View larger version (168K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3F. —49-year-old man with hepatic cyst and metastasis from
colorectal carcinoma. True FISP T2-weighted MR image shows segment V lesion
with intermediate high signal intensity (arrowhead), suggestive of
malignancy.
|
|
View larger version (175K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3G. —49-year-old man with hepatic cyst and metastasis from
colorectal carcinoma. On fast low-angle shot (FLASH) T1-weighted image at 2 hr
after administration of mangafodipir trisodium (MnDPDP), lesion in segment IV
(arrow) shows no enhancement.
|
|
View larger version (175K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3H. —49-year-old man with hepatic cyst and metastasis from
colorectal carcinoma. On FLASH T1-weighted image at 2 hr after administration
of MnDPDP, lesion in segment V (arrowhead) also shows no enhancement.
Histopathologic diagnoses were hepatic cyst for segment IV lesion and hepatic
metastasis for segment V lesion.
|
|
View larger version (154K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4C. —57-year-old woman with small hepatic metastases from
colorectal carcinoma. On fast low-angle shot T1-weighted image at 2 hr after
administration of mangafodipir trisodium, two tiny nodules
(arrowheads) are seen in segment VII of liver. Final histopathologic
diagnoses of surgically resected specimen were two small hepatic
metastases.
|
|
Although the detection rates for both all hepatic lesions and all
metastases were slightly higher on MnDPDP-enhanced MRI than on helical CT, the
differences were not statistically significant (p = 0.383 and 0.143,
respectively) (Table 2).
MnDPDP-enhanced MRI found 12 metastases in nine patients that CT missed, and
CT found five metastases in three patients that MRI missed. Concerning the
differentiation of benign and metastatic lesions, the two techniques showed a
statistically significant difference (p = 0.023). The difference
remained when the analysis was confined to the patients with histopathologic
confirmation (p = 0.024). The Az values and their
95% confidence intervals for MnD-PDP-enhanced MRI and helical CT are shown in
Table 3. The distributions of
confidence ratings for ROC analysis according to lesion size are shown in
Table 4. Disagreements existed
regarding the confidence ratings in 54 lesions (29.8%) on helical CT and in 32
lesions (17.7%) on MRI. Three false-positive diagnoses of hepatic metastases
were made only on MRI, four were made only on CT, and three were made on both
examinations. The average difference in the false-positive rate for detection
of hepatic metastases between MnDPDP-enhanced MRI (8.7%) and helical CT
(10.1%) was not statistically significant (p = 1.000). False-positive
diagnoses on MRI were related to hepatic hemangiomas (n = 4) or
cirrhotic nodules (n = 2). On helical CT, the reasons for
false-positive diagnoses were hepatic cysts (n = 4), cirrhotic
nodules (n = 2), and a hemangioma (n = 1).
View this table:
[in this window]
[in a new window]
|
TABLE 4 Distribution of Confidence Ratings for Receiver Operating Characteristic
Analysis According to Lesion Size
|
|
Although the detection rate of small hepatic lesions was slightly higher on
MnDPDP-enhanced MRI than that on helical CT, the difference was not
statistically significant (p = 0.210). However, with regard to the
detection of small hepatic metastases, the detection rate on MnDPDP-enhanced
MRI was significantly higher than that on helical CT both for all patients
(p = 0.022) and for those with histopathologic confirmation
(p = 0.043) (Table 2).
MnDPDP-enhanced MRI found 11 metastases in eight patients that CT missed, and
CT found two metastases in two patients that MRI missed. For the
differentiation between benign and metastatic lesions, the two techniques
showed a statistically significant difference (p = 0.015). The
difference remained when the analysis was confined to the patients with
histopathologic confirmation (p = 0.024). The Az
values and their 95% confidence intervals for MnDPDP-enhanced MRI and helical
CT are shown in Table 3.
Disagreements occurred about the confidence ratings in 45 small hepatic
lesions (32.8%) on helical CT and 24 lesions (17.5%) on MRI. Two
false-positive diagnoses of hepatic metastases were made only on MRI, four
were made only on CT, and two were made on both examinations. The average
difference in the false-positive rate between MnDPDP-enhanced MRI (7.4%) and
helical CT (11.1%) for the detection of small hepatic metastases was not
statistically significant (p = 0.687). False-positive diagnoses on
MRI were related to hepatic hemangiomas (n = 3) or a cirrhotic nodule
(n = 1). On helical CT, the reasons for false-positive diagnoses were
hepatic cysts (n = 4), a cirrhotic nodule (n = 1), and a
hemangioma (n = 1).
The detection rate for identification of large hepatic lesions on
MnDPDP-enhanced MRI was 93.2% (41/44 lesions) and on helical CT was 95.5%
(42/44 lesions). For large hepatic metastases, the detection rate was 91.7%
(33/36 lesions) on MnDPDP-enhanced MRI and 97.2% (35/36 lesions) on helical
CT. MnDPDP-enhanced MRI found one metastasis in one patient that CT missed.
However, MRI missed three metastases in one patient that CT found, which
ranged from 2.1 to 2.4 cm in diameter: two lesions were located in the
subcapsular area of liver segment VI abutting the hepatic flexure of the colon
and the other one, in segment IV, abutting the gallbladder. One false-positive
diagnosis was made on both examinations and one only on MRI. False-positive
diagnoses on MRI were related to a hepatic hemangioma (n = 1) or a
cirrhotic nodule (n = 1). On helical CT, the reason for
false-positive diagnosis was a cirrhotic nodule (n = 1).
Discussion
Complete resection of hepatic metastases has been shown to increase
survival in patients with metastatic colorectal cancer
[22]. To determine
resectability, accurate knowledge of the number, location, and size of
metastases is crucial. Although CT during arterial portography and
intraoperative sonography are the currently accepted imaging methods of
assessing the liver in this setting, both techniques are invasive and
therefore not routinely performed. On the other hand, CT and MRI, the two
principal imaging techniques for the evaluation of the liver, have undergone
marked technical advances during the past few years. With the advent of
helical and multidetector scanners, CT is the most frequently used imaging
technique for the preoperative depiction of focal liver lesions, with
sensitivity reported between 74% and 85% in the detection of metastases
[8,
23]. However, even with
multidetector scanners, detection of liver lesions is often limited when the
lesion is small [24]. Being
able to characterize lesions as benign or metastatic is also important.
However, although approximately 90% of hepatic lesions measuring 1 cm or
smaller in patients with a known primary malignancy are benign, these lesions
are often deemed too small to characterize on helical CT
[25]. Detecting more benign
lesions while not detecting more metastatic lesions on helical CT would not be
of great clinical value.
For MRI, liver-specific contrast agents such as iron oxide compounds and
manganese chelates have been developed, and recent investigations have focused
on their value in the detection and characterization of focal hepatic lesions
[4–12,
26]. Of these, MnDPDP is a
hepatocyte-specific paramagnetic agent that causes a predominantly T1
shortening effect. In patients with colorectal carcinoma, IV administration of
MnDPDP theoretically results in increased metastasis-to-liver contrast because
the normal liver shows increased signal intensity on T1-weighted images and
metastatic lesions do not. Also, metastatic liver tumors sometimes show
rimlike enhancement of variable degree that is infrequently seen in benign
lesions and that is attributed to several causes
[27]. However, in spite of
these theoretic benefits, controversy still exists about the difference
between the performance of MnDPDP-enhanced MRI and that of helical CT in the
detection and characterization of liver lesions
[12,
26]. Although the diagnostic
dilemma of liver lesion detection and characterization is often encountered
with helical CT, especially when the lesion is small, little is known about
the performance of MnDPDP-enhanced MRI for the assessment of small hepatic
lesions in patients with colorectal carcinoma.
In our experience, MnDPDP-enhanced MRI did not show statistically
significant differences from helical CT in the detection of all hepatic
lesions (p = 0.383) and the detection of small ones (p =
0.210). However, for the differentiation of benign and metastatic lesions,
MnDPDP-enhanced MRI was superior to helical CT, both for all hepatic lesions
(p = 0.023) and for small ones (p = 0.015), and it remained
better when the analyses were restricted to patients with histopathologic
confirmation (p = 0.023 for both). MnDPDP-enhanced MRI changed the
diagnosis of hepatic metastasis in nine (13.0%) of 69 patients. Of 12
metastases that were found on MnDPDP-enhanced MRI and missed on helical CT, 11
(91.7%) were small. MnDPDP-enhanced MRI showed a significantly greater
detection rate than did helical CT for small hepatic metastases (p =
0.022), and the rate was better when the analyses were restricted to the
patients with histopathologic confirmation (p = 0.043). Therefore,
detecting more small metastatic lesions on MnDPDP-enhanced MRI would have
great clinical value in patients with colorectal carcinoma for evaluating the
liver before surgery. In our experience, the improved results with
MnDPDP-enhanced MRI were assumed to be largely attributed to the improved
confidence level in their diagnoses with MRI, and this improvement was
frequently seen when the lesion was small. The number of disagreements in the
confidence ratings for helical CT (n = 45) exceeded that for
MnDPDP-enhanced MRI (n = 24) when the lesion was small, and this
pattern paralleled the uncertainty in the confidence scores shown in
Table 4. On the other hand, in
our experience the diagnostic performance on MnDPDP-enhanced MRI is subject to
interobserver variability. Although MnDPDP-enhanced MRI was superior to
helical CT in the characterization of the focal hepatic lesions for observer 1
and for the consensus review, no significant difference was seen in
characterization of either all or small hepatic lesions for observer 2 in our
study. This finding may be due to the fact that observer 3 was more senior and
that observer 2 was not as experienced as the other two observers with
MnDPDP-enhanced MRI.
Our study had several limitations. First, not all the subjects of this
study underwent state-of-theart CT or dual-phase helical CT. However, because
hepatic arterial phase scans in general are not necessarily required for
preoperative evaluation of patients with colorectal carcinoma and because
repeated CT examination was also deemed unethical, the results shown in this
study may be important for the assessment of the liver in these patients in
the routine practice of radiology. Second, histologic proof was not available
for nonresected segments because of ethical considerations. The reference
standard we used was based on all available follow-up imaging data combined
with all available clinical data. Because a suboptimal standard of reference
may result in a too-small fraction of small lesions and overestimation of the
detection rate [28], some
small lesions were probably present that were not detected either on
preoperative imaging or at surgery. In our experience, the statistical
differences between the two techniques remained when the analyses were
restricted to patients with histopathologic confirmation. Third, the section
thickness used for MRI was 8 mm, whereas the section thickness used for CT was
5–8 mm. The use of the thicker sections in MRI was based on
considerations of an adequate signal-to-noise ratio. Also, although CT was
performed with a 512 x 512 matrix, MRI was performed with a 132 x
256 matrix. If voxel sizes were identical for MRI and CT, an improvement with
greater statistical significance would have been seen for MRI. In addition, we
did not use fat suppression for MnDPDP-enhanced T1-weighted sequences although
that is generally recommended. Finally, our study of MnDPDP is of limited use,
considering the recent trend toward the use of gadolinium-based liver-specific
contrast agents.
In conclusion, the results of our study show no significant difference
between MnDPDP-enhanced MRI and helical CT for detecting focal hepatic lesions
in patients with colorectal carcinoma. However, MnDPDP-enhanced MRI is
superior to helical CT in characterization of focal hepatic lesions, and it
not infrequently changes the diagnosis, especially when the lesion is small,
although this finding may be subject to interobserver variability.
MnDPDP-enhanced MRI, with heavily T2-weighted sequences, is recommended for
the evaluation of focal hepatic lesions in patients with colorectal carcinoma
because it improves the characterization of these lesions and the detection of
small hepatic metastases in these patients.
References
- Whitney WS, Herfkens RJ, Jeffrey RB, et al. Dynamic breath-hold
multiplanar spoiled gradient-recalled MR imaging with gadolinium enhancement
for differentiating hepatic hemangiomas from malignancies at 1.5 T.
Radiology1993; 189:863
–870[Abstract/Free Full Text]
- Semelka RC, Shoenut JP, Ascher SM, et al. Solitary hepatic
metastasis: comparison of dynamic contrast-enhanced CT and MR imaging with
fatsuppressed T2-weighted, breath-hold T1-weighted FLASH, and dynamic
gadolinium-enhanced FLASH sequences. J Magn Reson
Imaging 1994;4:319
–323[Medline]
- Semelka RC, Martin DR, Balci C, Lance T. Focal liver lesions:
comparison of dual-phase CT and multisequence multiplanar MR imaging including
dynamic gadolinium enhancement. J Magn Reson Imaging2001; 13:397
–401[Medline]
- Hagspiel KD, Neidl KFW, Eichenberger AC, Weder W, Marincek B.
Detection of liver metastases: comparison of superparamagnetic iron
oxide-enhanced MR imaging at 1.5T with dynamic CT, intraoperative US and
percutaneous US. Radiology1995; 196:471
–478[Abstract/Free Full Text]
- Ros PR, Freeny PC, Harms SE, et al. Hepatic MR imaging with
ferumoxides: a multicenter clinical trial of the safety and efficacy in the
detection of focal hepatic lesions. Radiology1995; 196:481
–488[Abstract/Free Full Text]
- Seneterre E, Taourel P, Bouvier Y, et al. Detection of hepatic
metastases: ferumoxides-enhanced MR imaging versus unenhanced MR imaging and
CT during arterial portography. Radiology1996; 200:785
–792[Abstract/Free Full Text]
- Oudkerk M, van den Heuvel AG, Wielopolski PA, Schmitz PIM, Borel
Rinkes IHM, Wiggers T. Hepatic lesions: detection with ferumoxide-enhanced
T1-weighted MR imaging. Radiology1997; 203:449
–456[Abstract/Free Full Text]
- Ward J, Naik KS, Guthrie JA, Wilson D, Robinson PJ. Hepatic lesion
detection: comparison of MR imaging after the administration of
superparamagnetic iron oxide with dual-phase CT by using alternative free
response receiver operating characteristic analysis.
Radiology1999; 210:459
–466[Abstract/Free Full Text]
- Reimer P, Jahnke N, Feibich M, et al. Hepatic lesion detection and
characterization: value of nonenhanced MR imaging, superparamagnetic iron
oxide-enhanced MR imaging, and spiral CT-ROC analysis.
Radiology2000; 217:152
–158[Abstract/Free Full Text]
- Arbab AS, Ichikawa T, Sou H, et al. Ferumoxides-enhanced
double-echo T2-weighted MR imaging in differentiating metastases from nonsolid
benign lesions of the liver. Radiology2002; 225:151
–158[Abstract/Free Full Text]
- Torres CG, Lundby B, Sterud AT, McGill S, Gordon PB, Bjerknes HS.
MnDPDP for MR imaging of the liver: results from the European phase III
studies. Acta Radiol1997; 38:631
–637[Medline]
- Oudkerk M, Torres CG, Song B, et al. Characterization of liver
lesions with mangafodipir trisodium-enhanced MR imaging: multicenter study
comparing MR and dual-phase spiral CT. Radiology2002; 223:517
–524[Abstract/Free Full Text]
- Hamm B, Vogl TJ, Branding G, et al. Focal liver lesions: MR imaging
with MnDPDP—initial clinical results in 40 patients.
Radiology1992; 182:167
–174[Abstract/Free Full Text]
- Rummeny EJ, Torres CG, Kurdziel JC, Nilsen G, Op de Beeck B, Lundby
B. MnDPDP for MR imaging of the liver: results of an independent image
evaluation of the European phase III studies. Acta
Radiol 1997;38:638
–642[Medline]
- Imam K, Bluemke DA. MR imaging in the evaluation of hepatic
metastases. Magn Reson Imaging Clin N Am2000; 8:741
–756[Medline]
- van Leeuwen MS, Noordzij J, Feldberg MAM, Hennipman AH, Doornewaard
H. Focal liver lesions: characterization with triphasic spiral CT.
Radiology1996; 201:327
–336[Abstract/Free Full Text]
- Oliver JH III, Baron RL. Helical biphasic contrast-enhanced CT of
the liver: technique, indications, interpretation, and pitfalls.
Radiology1996; 201:1
–14[Abstract/Free Full Text]
- Kane PA, Ayton V, Walters HL, et al. MnDPDP-enhanced MR imaging of
the liver: correlation with surgical findings. Acta
Radiol 1997;38:650
–654[Medline]
- Hahn PF, Saini S. Liver-specific MR imaging contrast agents.
Radiol Clin North Am1998; 36:287
–297[Medline]
- Hanley JA, McNeil BJ. The meaning and use of the area under a
receiver operating characteristic (ROC) curve.
Radiology1982; 143:29
–36[Abstract/Free Full Text]
- Hanley JA, McNeil BJ. A method of comparing the areas under
receiver operating characteristic curves derived from the same cases.
Radiology1983; 148:839
–843[Abstract/Free Full Text]
- Greenway B. Hepatic metastases from colorectal cancer: resection or
not. Br J Surg1988; 75:513
–519[Medline]
- Valls C, Andia E, Sanchez A, et al. Hepatic metastases from
colorectal cancer: preoperative detection and assessment of resectability with
helical CT. Radiology2001; 218:55
–60[Abstract/Free Full Text]
- Haider MA, Amitai MM, Rappaport DC, et al. Multidetector row
helical CT in preoperative assessment of small ( 1.5 cm) liver metastases:
is thinner collimation better? Radiology2002; 225:137
–142[Abstract/Free Full Text]
- Schwartz LH, Gandras EJ, Colangelo SM, Ercolani MC, Panicek DM.
Prevalence and importance of small hepatic lesions found at CT in patients
with cancer. Radiology1999; 210:71
–74[Abstract/Free Full Text]
- Braga HJV, Choti MA, Lee VS, Paulson EK, Siegelman ES, Bluemke DA.
Liver lesions: manganese-enhanced MR and dual-phase helical CT for
preoperative detection and characterization— comparison with receiver
operating characteristic analysis. Radiology2002; 223:525
–531[Abstract/Free Full Text]
- Ni YC, Marchal G, Yu J, et al. Experimental liver cancers:
MnDPDP-enhanced rims in MR-microangiographic-histologic correlation study.
Radiology1993; 188:45
–51[Abstract/Free Full Text]
- van Erkel AR, Pijl MEJ, van den Berg-Huysmans AA, Wasser MNJM, van
de Velde CJH, Bloem JL. Hepatic metastases in patients with colorectal cancer:
relationship between size of metastases, standard of reference, and detection
rates. Radiology2002; 224:404
–409[Abstract/Free Full Text]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
|
|
 
D.-M. Koh, G. Brown, Z. Meer, A. R. Norman, and J. E. Husband
Diagnostic Accuracy of Rim and Segmental MRI Enhancement of Colorectal Hepatic Metastasis After Administration of Mangafodipir Trisodium
Am. J. Roentgenol.,
February 1, 2007;
188(2):
W154 - W161.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
H. J. Kim, K. W. Kim, J. H. Byun, H. J. Won, Y. M. Shin, P. N. Kim, M.-S. Lee, and M.-G. Lee
Comparison of mangafodipir trisodium- and ferucarbotran-enhanced MRI for detection and characterization of hepatic metastases in colorectal cancer patients.
Am. J. Roentgenol.,
April 1, 2006;
186(4):
1059 - 1066.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
H. I. Khalil, S. A. Patterson, and D. M. Panicek
Hepatic Lesions Deemed Too Small to Characterize at CT: Prevalence and Importance in Women with Breast Cancer
Radiology,
June 1, 2005;
235(3):
872 - 878.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
