| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Sharp and Children's MRI Center, Sharp Memorial Hospital, 7901 Frost St., San
Diego, CA 92123.
2 Department of Radiology, Sharp Memorial Hospital, San Diego, CA.
3 Department of Surgical Oncology, Sharp Memorial Hospital, San Diego, CA.
4 Department of Pathology, Sharp Memorial Hospital, San Diego, CA.
Received February 6, 2007;
accepted after revision September 17, 2007.
Address correspondence to R. N. Low
(rlow{at}ucsd.edu).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsdiscussionReferences |
|---|
MATERIALS AND METHODS. Twenty-two patients underwent preoperative MRI including T1-weighted, T2-weighted, immediate gadolinium-enhanced, and delayed gadolinium-enhanced imaging. Two observers reviewed the images for peritoneal tumor at 13 sites, tumor size and distribution, and degree of tumor enhancement. Peritoneal tumor sites were recorded at surgery. Cytoreduction was categorized as complete or suboptimal. Surgical specimens were classified as disseminated peritoneal adenomucinosis tumors, intermediate-grade tumors, or peritoneal mucinous carcinomatosis tumors.
RESULTS. Surgery confirmed 232 tumor sites. Delayed gadolinium-enhanced MRI was the most accurate of the MR techniques, with a sensitivity, specificity, and accuracy of 89%, 87%, and 89%, respectively, for observer 1 and 82%, 87%, and 83% for observer 2 (p < 0.001). Surgical cytoreduction was complete in 14 patients and suboptimal in eight. MRI findings predicting suboptimal cytoreduction included a large (> 5 cm) mesenteric mass, which was present in 75% of the patients in the suboptimal cytoreduction group and 0% of those in the complete cytoreduction group; diffuse mesenteric tumor (88% and 0%, respectively); tumor encasement of mesenteric vessels (88% and 0%); or diffuse small-bowel serosal tumor (75% and 0%). Histopathology results showed six disseminated peritoneal adenomucinosis tumors, four intermediate tumors, and 11 peritoneal mucinous carcinomatosis tumors. The specimens for the remaining patient were not available for histopathologic analysis. Qualitatively, the 11 peritoneal mucinous carcinomatosis tumors showed greater enhancement than the liver, whereas six disseminated peritoneal adenomucinosis and the four intermediate tumors showed less enhancement than the liver. Quantitatively, the mean tumor-to-liver contrast for disseminated peritoneal adenomucinosis and intermediate tumors was 0.67 compared with 1.53 for peritoneal mucinous carcinomatosis tumors (p < 0.0001).
CONCLUSION.Of the MR techniques evaluated, delayed gadolinium-enhanced MRI was the most accurate for the staging and classification of mucinous appendiceal neoplasms and provided prognostic information useful for patient selection.
Keywords: disseminated peritoneal adenomucinosis • MRI • mucinous adenocarcinoma • peritoneal mucinous carcinomatosis • pseudomyxoma peritonei
|
TopAbstractIntroductionMaterials and MethodsResultsdiscussionReferences |
|---|
The spectrum of disease of pseudomyxoma peritonei syndrome may be separated into three clinical pathologic categories as described by Ronnett and colleagues [2]. Disseminated peritoneal adenomucinosis is a benign condition arising from appendiceal adenomas, whereas peritoneal mucinous carcinomatosis is characterized by architectural and cytologic features of adenocarcinoma. Peritoneal mucinous carcinomatosis arises from appendiceal or intestinal mucinous adenocarcinomas. An intermediate category occurs with features in-between those of disseminated peritoneal adenomucinosis and peri toneal mucinous carcinomatosis. The classi fication of pseudomyxoma peritonei determines the clinical course and long-term survival. The age-adjusted 5-year survival for patients with disseminated peritoneal adenomucinosis is 84% compared with 37.5% for patients with intermediate features and 6.7% for those with peritoneal mucinous carcinomatosis [2].
At the time of pseudomyxoma peritonei diagnosis, this group of patients is very heterogeneous including benign and frankly malignant forms of mucinous appendiceal tumors. In a series of 410 patients with appendiceal tumors, only 217 were found to have disseminated peritoneal adenomucinosis at histopathologic evaluation [8]. The exact definition of "pseudomyxoma peritonei" is controversial, with some authors reserving the term for patients with benign adenomucinosis. Currently, it is not possible to make this distinction at clinical presentation. The term "pseudomyxoma peritonei syndrome," therefore, includes the spectrum of possible histopathologic entities facing the surgeon and radiologist evaluating the patient with a mucinous appendiceal neoplasm.
|
|
|
Helical CT is currently the best and most widely used preoperative imaging technique in patients with mucinous appendiceal neoplasms. However, the limited contrast range of CT makes it difficult to distinguish mucin, ascites, and solid peritoneal tumor [16]. Incomplete depiction of the extent of tumor on preoperative imaging can adversely affect patient management if the results of imaging are used to select patients for surgical cytoreduction. For example, patients with unresectable disease may be subjected to unnecessary laparotomy if preoperative imaging underestimates the extent of tumor.
Current preoperative imaging also does not provide information about tumor histology, which could affect the surgeon's decision to proceed with laparotomy and tumor debulking. High-grade peritoneal mucinous carcinomatosis tumors are more likely to be invasive, thus complicating complete surgical resection of the tumor. To our knowledge, preoperative imaging with CT or MRI has not been used to predict the histologic type of mucinous tumor.
MRI with gadolinium enhancement has been shown to be useful for depicting perito-neal metastases in patients with ovarian cancer and other tumors with intraperitoneal dissemination [17–22]. In patients with pseudo-myxoma peritonei syndrome, we hypo thesized that on fat-suppressed gadolinium-enhanced images only the solid cellular tumor would enhance, whereas the benign mucinous material and ascites would not enhance (Fig. 1A, 1B, 1C); and that the peritoneal lesions from disseminated peritoneal adenomucinosis and intermediate tumors might show less pronounced enhancement than the more cellular peritoneal metastases from peritoneal mucinous carcinomatosis tumors (Fig. 2A, 2B, 2C). We undertook this study in 22 patients with mucinous appendiceal neoplasms to determine the accuracy of MRI for preoperative tumor staging and classification and as a tool for predicting the feasibility of complete surgical cytoreduction.
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsdiscussionReferences |
|---|
This retrospective study was approved by our institution's investigational review board, which waived requirements for written informed consent. HIPAA compliance was established. After the study, all patient-identifying data were removed from the study records to protect patient confidentiality.
MRI
All patients underwent abdominal and pelvic MRI on a 1.5-T imager: 12 were
imaged on an Intera scanner (Philips Medical Systems) (30 MT/m, 150 MT/m/s)
and 10 were imaged on a Signa LX scanner (GE Healthcare) (23 MT/m, 120
MT/m/s). Oral contrast material was administered starting 1 hour before the
examination and included dilute barium sulfate (1,350 mL; Readi-Cat 2, E-ZEM)
or a similar volume of psyllium hydrophilic mucilloid (0.8 g/kg; Metamucil,
Procter & Gamble) mixed in 1–1.5 L of water. Water (500–1,000
mL) was administered rectally through a balloon-tipped barium enema
catheter.
Body coil imaging of the abdomen and pelvis was performed with axial gradient-echo T1-weighted, fat-suppressed T2-weighted, and immediate and delayed fat-suppressed gadolinium-enhanced gradient-echo sequences. The time for MR examination was 30–40 minutes. T1-weighted images were not obtained in two patients.
The imaging parameters for the T1-weighted images included the following: TR range/TE, 140–160/4.4; number of excitations (NEX), 1; 256 x 192 matrix; 7-mm slice thickness; and 3-mm interslice gap. The time of acquisition was 12–24 seconds per 24 slices. For the fat-suppressed T2-weighted images, we used either a breath-hold fast spin-echo sequence (TR/TE, 2,500/77; NEX, 1; 256 x 192 matrix; chemical fat suppression; 7-mm slice thickness; 3-mm interslice gap) that required a 25-second breath-hold for each 12 slices or a respiratory-triggered turbo spin-echo sequence with selective partial inversion-recovery (SPIR) fat suppression (1,600/70; NEX, 2; 512 x 204 matrix; 7-mm slice thickness; 3-mm interslice gap) with a time of acquisition of 3 minutes 28 seconds for 24 slices.
The fat-suppressed gadolinium-enhanced images were obtained in the axial plane im mediately after and 5 minutes after IV injection of 0.2 mmol/kg of gadolinium. Delayed imaging in the coronal plane was also performed.
The selection of pulse sequences differed for the scanners. On the Philips scanner, immediate gadolinium-enhanced imaging used a 3D T1 High Resolution Isotropic Volume Examination (THRIVE) acquisition with SPIR fat suppression requiring a 21-second breath-hold for each 50 slices and was performed using the following para meters: 3.5/1.7; 512 x 192 matrix; NEX, 1; 6-mm slice thickness with 50% slice overlap; and 10° flip angle. Delayed 2D gradient-echo imaging was performed with ProSet-selective (Philips Medical Systems) water excitation, requiring a 17-second breath-hold for each 8 slices, using the following parameters: 144/5; 512 x 228 matrix; NEX, 1; 8- to 10-mm contiguous slice thickness; and 70° flip angle.
On the GE scanner, immediate and delayed dynamic gadolinium-enhanced imaging used a breath-hold fat-suppressed 2D spoiled gradient-echo sequence requiring a 24-second breath-hold for each 12 slices and was performed using the following parameters: TR range/TE range, 140–165/1.9–2.6; 512 x 192 matrix; NEX, 1; 8- to 10-mm-thick contiguous slices; 16- to 20-kHz receiver bandwidth; and 70° flip angle. Delayed gadolinium-enhanced 2D gradient-echo imaging was used on both the GE and Philips scanners. In our experience, high-resolution 2D gradient-echo images provide sharper image detail and better soft-tissue contrast than 3D gradient-echo images.
Surgical Exploration
Surgical exploration and cytoreductive surgery of the abdomen and pelvis
were performed in all 22 patients after MR examination. All of the
laparotomies were performed by an oncologic surgeon. During laparotomy, visual
inspection and palpation of all abdominal organs, peritoneal surfaces, and
bowel serosa were performed. All suspicious nodules and masses were biopsied
and sent for histopathologic evaluation. Cytoreductive surgeries included
omentectomy and major peritonectomy procedures as required combined with
large- and small-bowel resections, gastrectomy, and splenectomy with the
intention of removing all visible peritoneal and visceral tumors. Peritoneal
washings were obtained for cytologic evaluation. The location, extent, and
volume of abdominal and pelvic tumor and sites free of tumor were recorded
from the operative notes and interviews with the surgeon.
The surgeon attempted to achieve complete surgical resection of all gross tumors. On the basis of the operative reports and interviews with the surgeon after the operation, each case was categorized as complete cytoreduction if there was no gross residual tumor or as suboptimal cytoreduction if there was gross residual tumor.
Histopathologic Evaluation
Histopathologic specimens for 21 patients were retrospectively reviewed by
one pathologist blinded to prior results. For the remaining patient, the
original specimens were not available. Each case was categorized as
disseminated perito neal adenomucinosis, intermediate, or peritoneal mucinous
carcinomatosis on the basis of the microscopic features of the submitted
specimens. Cases were categorized as disseminated peritoneal adenomucinosis if
there was abundant extracellular mucin and only scant simple to focally
proliferative epithelium with little cytologic atypia or mitotic activity or
as peritoneal mucinous carcinomatosis if the peritoneal tumors showed mucin
combined with moderate to abundant cellular material composed of proliferative
mucinous epithelium showing cytologic and architectural features of
adenocarcinoma [2]. Cases were
categorized as intermediate type if the features were in-between those of
disseminated peritoneal adenomucinosis and those of peritoneal mucinous
carcinomatosis.
Review of MR Images
MR images were reviewed separately and independently by two radiologists.
The T1-weighted, T2-weighted, immediate gadolinium-enhanced, and delayed
gadolinium-enhanced images were separately reviewed for each patient to
determine which sequence provided the most useful information regarding tumor
staging.
Peritoneal tumor staging—For each of the four image types, the reviewers determined the presence or absence of peritoneal tumor at 13 anatomic sites, including the right subphrenic space, left subphrenic space, gastrohepatic liga ment, lesser sac, right subhepatic space, perihepatic fissures, right paracolic gutter, left paracolic gutter, omentum, small-bowel mesentery, colon, small-bowel serosa, and pelvis. Features that were used to determine the presence of peritoneal tumor for each image type included mass effect on T1-weighted images, mass effect or heterogeneity of soft tissue on T2-weighted images, and enhance ment of peritoneal masses on the immediate or delayed gadolinium-enhanced spoiled gradient-echo images. On the unenhanced T1- and T2-weighted images, mass effect, heterogeneity, or both were used to distinguish solid peritoneal lesions of pseudomyxoma peritonei syndrome from simple ascites or mucin. The presence of ascites and mucin was also noted for each of the four image types.
Prediction of the extent of surgical cyto-reduction—MR images were reviewed to determine the size of the tumors at each anatomic location. The tumors were categorized as < 1, 1–2, > 2–5, or > 5 cm.
The distribution of mesenteric tumor was further categorized as solitary, multifocal, or diffuse involvement of the small-bowel mesentery. Encasement of mesenteric vessels by tumor, if present, was noted. The size and distribution of tumors involving the small-bowel serosa were also noted. The distribution of small-bowel tumor was categorized as focal, multifocal, or diffuse if more than 50% of the small-bowel loops were involved by peritoneal tumors. These MRI findings were correlated with the postoperative findings of adequate versus suboptimal surgical cytoreduction of all gross tumors.
Pseudomyxoma Peritonei Tumor Classification
Qualitative assessment—Using the delayed gadolinium-enhanced
images, the observers independently scored the degree of maximal enhancement
of the dominant peritoneal mass as follows: 1, no enhancement; 2, enhancement
less than or equal to liver parenchyma; 3, enhancement greater than liver
parenchyma; or 4, enhancement equal to intravascular gadolinium. For cases
with varied degrees of enhancement in the peritoneal lesions, the mass with
the greatest degree of enhancement was assessed. Nondistended segments of
bowel were excluded from this assessment because enhancement of collapsed
bowel can be mis leading. Linear enhancement of nonthickened peritoneal
surfaces was also excluded.
Quantitative assessment—A quantitative analysis of the degree of maximal tumor enhancement was also performed. On a computer workstation, the delayed gadolinium-enhanced images were evaluated by placing separate regions of interest (ROIs) over the dominant enhancing mass, liver parenchyma, and intravascular gadolinium in the ab dominal aorta or inferior vena cava. Intrahepatic vessels were avoided when selecting the placement of the liver ROI. The mean signal intensity was measured for tumor, liver, and IV gadolinium in the abdominal aorta or inferior vena cava.
The tumor-to-liver contrast was calculated by dividing the mean signal intensity of the tumor by the mean signal intensity of the liver. The tumor-to-intravascular gadolinium contrast was calculated by dividing the mean signal intensity of the enhancing tumor by the mean signal intensity of IV gadolinium in the abdominal aorta or inferior vena cava.
Correlation of MR Images with Surgical and Histopathologic Findings
A site-by-site comparison of MR images and operative and histopathologic
findings was performed. Operative and histopathologic reports were reviewed
for the presence of peritoneal tumors at each of the 13 anatomic sites.
Information from the surgeon and the operative reports was used to determine
whether tumor was present at each of the anatomic sites. Both the original
written pathology reports available in all 22 patients and the results of the
retrospective review of specimens in 21 patients were used. Based on these
comparisons, per-patient and per-site sensitivity, specificity, and accuracy
were calculated for each of the four types of MR images. Histopathologic tumor
classification was compared with the qualitative and quantitative analyses of
the degree of gadolinium tumor enhancement.
Statistical Analysis
The McNemar test of correlated proportions was used to analyze per-site and
per-patient peritoneal tumor detection for the four types of MR images. The
histologic classification versus the degree of gadolinium enhancement on
delayed MR images was evaluated using the Fisher's exact test. A 2 x 2
contingency table was constructed. The disseminated peritoneal adenomucinosis
tumors and intermediate-grade tumors were combined in one category and were
compared with the peritoneal mucinous carcinomatosis tumors. Gadolinium
enhance ment levels 1 and 2 were combined as were gadolinium enhancement
levels 3 and 4.
The results of the quantitative analysis of tumor enhancement were analyzed by comparing means for disseminated peritoneal adenomucinosis and intermediate-grade tumors with high-grade peritoneal mucinous carcinomatosis tumors using the nonparametric Mann-Whitney test. MRI features used to predict adequate versus inadequate surgical resection were tested using a chi-square test for independence. Interobserver agreement was tested using the kappa analysis. In all cases, a two-tailed p value is reported with the null hypothesis rejected for p values less than 0.05.
|
|
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsdiscussionReferences |
|---|
|
|
|
|
|
Table 1 shows the per-site sensitivity for the two observers in detecting cellular peritoneal tumors. For both observers, the sensitivity and accuracy of the delayed gadolinium-enhanced images were superior to T1-weighted, T2-weighted, and immediate gadolinium-enhanced images (p < 0.001, McNemar test). The immediate gadolinium-enhanced images were superior to the T1-weighted images (p < 0.001) but not to the T2-weighted images (p > 0.05), and the T2-weighted images were superior to the T1-weighted images (p < 0.001, McNemar test).
|
On a per-patient basis, peritoneal tumor was depicted on T1-weighted images in 80% (16/20) and 70% (14/20) of patients, T2-weighted images in 82% (18 of 22 patients) for both observers, immediate gadolinium-enhanced images in 82% (18 of 22 patients), and delayed gadolinium-enhanced images in 100% (22 of 22 patients). There was no significant difference among the four MR image types for depicting tumor on a per-patient basis (p > 0.05, McNemar test).
Prediction of the Extent of Surgical Cytoreduction
Surgical cytoreduction achieved complete resection of all gross tumors in
14 patients (64%). In the remaining eight patients (36%), resection was
incomplete with gross tumor remaining after surgical cytoreduction. The MRI
findings that best predicted suboptimal surgical cytoreduction were the
presence of large (> 5 cm) small-bowel mesenteric tumors, which were
present in 75% of the suboptimal and 0% of the complete cytoreduction groups
(p < 0.004) (Fig.
3A,
3B,
3C,
3D,
3E,
3F); diffuse mesenteric tumor
infiltration (88% and 0%, respectively; p < 0.0002); tumor
encasement of mesenteric vessels (88% and 0%; p < 0.0001); or
diffuse small-bowel serosal tumor involving > 50% of the small intestine
(75% and 0%; p < 0.001) (Table
2).
|
Mesenteric tumor—In the eight patients with suboptimal surgical cytoreduction, mes enteric tumors measured < 1 cm in one patient, > 2–5 cm in one patient, and > 5 cm in six patients (Fig. 3A, 3B, 3C, 3D, 3E, 3F). Mesenteric tumor distribution was diffuse in seven patients and multifocal in one patient (p < 0.0002). All seven patients with diffuse mesenteric tumor also showed tumor encasement of mesenteric vessels. One patient with unresectable tumor had small mesenteric tumors (< 1 cm) but also presented with diffuse small-bowel serosal tumor (> 50% of small intestine) that precluded complete surgical tumor resection.
In the 14 patients with adequate surgical cytoreduction, MR images showed
normal small-bowel mesentery in one patient, small tumors measuring 
2 cm
in 10 patients, and moderate tumors measuring > 2–5 cm in three
patients. Mesenteric tumor distribution was focal in four patients and
multifocal in nine patients (p < 0.0002). In none of the 14
patients with adequate surgical cytoreduction was there a large (> 5 cm)
mesenteric tumor mass, diffuse mesenteric tumor infiltration, or encasement of
mesenteric vessels by tumor.
Small-bowel serosal tumor—Table 2 shows the size and distribution of small-bowel tumor. Tumor involvement of the small-bowel serosa was present in 19 patients. Patients with complete cytoreduction tended to have smaller serosal tumors than those with suboptimal cytoreduction, but this difference was not statistically significant. The distribution of small-bowel tumor was focal in two patients, multifocal in 11 patients, and diffuse in six patients. All six patients with diffuse tumor involving > 50% of the small bowel had tumor that was unresectable at the time of surgery. Small-volume small-bowel serosal tumor was not depicted on MR images in three patients (Fig. 5). For all other anatomic locations, the presence and size of tumor did not correlate with the adequacy of surgical cytoreduction.
|
The high-grade peritoneal mucinous carcinomatosis tumors exhibited a more complex and heterogeneous MR appearance than the other two categories of tumor (Fig. 4A, 4B, 4C, 4D), with masses showing marked gadolinium enhancement combined with mucin and other soft-tissue masses showing mild to moderate gadolinium enhancement. The areas of marked enhancement correlated with cellular adenocarcinoma mixed with mucin and varying amounts of fibrosis (Fig. 4A, 4B, 4C, 4D).
Qualitative analysis of tumor enhancement—Table 3 shows the qualitative evaluation of the degree of maximal tumor enhancement by the two observers. Enhancement of the dominant enhancing peritoneal mass is compared with histologic classification. Peritoneal mucinous carcinomatosis tumors showed more marked gadolinium enhancement (Figs. 1A, 1B, 1C and 4A, 4B, 4C, 4D), whereas the disseminated peritoneal adenomucinosis and intermediate-type tumors showed less intense enhancement (Figs. 2A, 2B, 2C and 3A, 3B, 3C, 3D, 3E, 3F).
|
All patients with mild gadolinium enhancement (level 1 or 2) had disseminated peritoneal adenomucinosis or intermediate tumors (Table 3). Marked gadolinium tumor enhancement (level 4) was noted by at least one observer in 11 patients. Ten of those 11 patients had a peritoneal mucinous carcinomatosis tumor at histopathologic evaluation. Moderate tumor enhancement (level 3) was noted in five patients by at least one observer. Four had peritoneal mucinous carcinomatosis tumors and one had a disseminated peritoneal adenomucinosis tumor. Level 3 or 4 tumor enhancement of the disseminated peritoneal adenomucinosis and intermediate tumors was due to focal areas of enhancing bowel in two patients. The remaining tumor masses in these patients showed only mild level 2 enhancement.
The difference in the degree of maximal tumor enhancement between the disseminated peritoneal adenomucinosis and intermediate-type tumors versus the peritoneal mucinous carcinomatosis tumors was statistically significant (p < 0.0001, Fisher's exact test). Combining the scores for the two observers on delayed gadolinium-enhanced images, a threshold of level 3 or 4 gadolinium enhancement had 100% sensitivity (22/22), 95% accuracy (40/42), and 92% positive predictive value (22/24) for distinguishing peritoneal mucinous carcinomatosis tumors from disseminated peritoneal adenomucinosis and intermediate tumors.
Quantitative analysis of tumor enhancement—Figure 6 shows the results of the quantitative evaluation of the dominant enhancing tumor compared with the liver parenchyma and intravascular gadolinium. The mean tumor-to-liver contrast for disseminated peritoneal adenomucinosis and intermediate tumors was 0.67 ± 0.13 (range, 0.48–0.96) compared with 1.53 ± 0.28 (range, 1.26–2.19) for peritoneal mucinous carcinomatosis tumors (p < 0.0001). The tumor-to-intravascular gadolinium contrast was 0.53 ± 0.09 (range, 0.40–0.72) for the disseminated peritoneal adenomucinosis and intermediate tumors and 0.98 ± 0.09 (range, 0.86–1.19) for the peritoneal mucinous carcinomatosis tumors (p < 0.0001).
|
Interobserver Agreement
The two radiologists agreed in 1,041 (84%) of 1,240 evaluations. For
determining the presence or absence of peritoneal tumor, agreement between the
two observers was excellent (
= 0.66). For determining tumor presence
at individual sites, there was moderate to excellent interobserver agreement
with kappa values as follows: small-bowel serosa, 0.524; small-bowel
mesentery, 0.545; colon, 0.480; right subphrenic space, 0.638; left subphrenic
space, 0.719; right subhepatic space, 0.545; gastrohepatic ligament, 0.794;
lesser sac, 0.480; omentum, 0.777; right paracolic gutter, 0.668; left
paracolic gutter, 0.677; pelvis, 0.629; and perihepatic fissures, 0.685. For
determining the degree of maximal gadolinium tumor enhancement, agreement was
excellent ( = 0.78).
|
TopAbstractIntroductionMaterials and MethodsResultsdiscussionReferences |
|---|
Currently the best method to assess the operability of pseudomyxoma peritonei tumor is contrast-enhanced CT. Jacquet et al. [28] described favorable and unfavorable CT features and found that segmental small-bowel obstruction or a small-bowel or mesenteric tumor larger than 5 cm in diameter are features associated with subsequent incomplete resection of pseudomyxoma peritonei tumor. These patients present significant operative challenges that may lead to extensive small-bowel resections that result in the possible development of short-bowel syndrome and significantly increased postoperative morbidity.
In our study, the MRI features that were indicators of a poor prognosis for surgical resection included large mesenteric masses > 5 cm, diffuse mesenteric tumor infiltration, mesenteric vascular encasement, or diffuse small-bowel serosal tumors. Our current practice is to avoid attempting surgical resection in patients in whom MRI shows extensive tumor involving the small-bowel mesentery or small-bowel serosa. The ability of gadolinium-enhanced MRI to directly show enhancing diffuse peritoneal and serosal tumor increases the accuracy of this assessment.
Tumor classification also affects the likelihood of successful surgical cytoreduction and survival in patients with mucinous appendiceal malignancy [29–33]. Peritoneal mucinous carcinomatosis tumors are prone to deep invasion of the peritoneal surfaces, mesentery, and bowel serosa, resulting in incomplete surgical cytoreduction. Gadolinium-enhanced MR images may allow one to preoperatively predict the histologic type of tumor using either a qualitative or a quantitative evaluation of the degree of delayed gadolinium enhancement.
This MR assessment of tumor classification as a function of gadolinium enhancement reflects the cellularity of the peritoneal tumors. Peritoneal mucinous carcinomatosis tumors contain some areas that are more highly cellular and show more intense enhancement than the less-cellular disseminated peritoneal adenomucinosis and intermediate-type tumors. Certainly on a per-site basis, not all microscopic sites of invasive adenocarcinoma showed marked gadolinium enhancement. Qualitatively, each case of peritoneal mucinous carcinomatosis contained some highly cellular tumors showing level 3 or 4 gadolinium enhancement, whereas disseminated peritoneal adenomucinosis and intermediate tumors showed less-pronounced gado linium enhancement. A quantitative measurement of tumor-to-liver contrast can be used to distinguish disseminated peritoneal adenomucinosis and intermediate tumors with ratios of < 1.0 (mean, 0.67) from disseminated peritoneal adenomucinosis tumors with ratios of > 1.0 (mean, 1.53).
At most institutions, MDCT is the first examination used and therefore plays an important role in evaluating patients with mucinous appendiceal neoplasms. The speed and excellent spatial resolution of CT are ideally suited for abdominal imaging. However, the more limited contrast resolution of CT can present challenges in depicting subtle peritoneal tumors [19, 34]. Coakley et al. [34] noted that the sensitivity of helical CT for peritoneal tumors < 1 cm was only 25–50% compared with 85–93% for all tumors. MRI with gadolinium provides images with superior soft-tissue contrast in which small peritoneal tumors and carcinomatosis are routinely depicted due to their enhancement with IV gadolinium [19–22]. The sensitivity of delayed gadolinium-enhanced images for depicting subtle peritoneal tumor is well established. In one study, the sensitivity of gadolinium-enhanced MR images for depicting peritoneal tumors < 1 cm was 85–90% compared with 22–33% for CT [19]. The overall sensitivity of MRI for depicting peritoneal tumors of all sizes was 84% in our study compared with 54% for CT [19].
The limitations of our study should be acknowledged. Our study was retrospective and included a relatively small number of patients with pseudomyxoma peritonei syndrome. A study with a larger number of patients is needed to confirm our observations. A direct comparison of gadolinium-enhanced MRI and helical CT would be useful but was beyond the scope of this retrospective study. Our observations regarding tumor classification and maximal gadolinium enhancement may be altered by fibrosis or postoperative changes, which also may result in areas of gadolinium enhancement. Limitations due to technical factors may have introduced bias in detection rates. For instance, the delayed gadolinium-enhanced 2D gradient-echo images used a thicker slice thickness than the immediate 3D gradient-echo images.
In conclusion, in the evaluation of patients with mucinous appendiceal neoplasms, MRI provides important information regarding preoperative staging, tumor classification, and patient selection for surgical resection. Delayed gadolinium-enhanced MR images are most accurate for depicting cellular peritoneal tumors in pseudomyxoma peritonei syndrome. Qualitatively and quantitatively disseminated peritoneal adenomucinosis and intermediate tumors show less intense gadolinium enhancement than peritoneal mucinous carcinomatosis tumors. At our institution, all patients with mucinous appendiceal malignancy are evaluated before laparotomy with gadolinium-enhanced MRI.
|
TopAbstractIntroductionMaterials and MethodsResultsdiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  A. C. Westphalen, F. V. Coakley, J. Kurhanewicz, G. Reed, Z. J. Wang, and J. P. Simko Mucinous Adenocarcinoma of the Prostate: MRI and MR Spectroscopy Features Am. J. Roentgenol., September 1, 2009; 193(3): W238 - W243. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. N. Low, C. P. Sebrechts, R. M. Barone, and W. Muller Diffusion-Weighted MRI of Peritoneal Tumors: Comparison With Conventional MRI and Surgical and Histopathologic Findings--A Feasibility Study Am. J. Roentgenol., August 1, 2009; 193(2): 461 - 470. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) and tumor-to-intravascular gadolinium signal (
). Measurements
of mean signal intensity were made on workstation by placing region of
interest over enhancing tumor, liver parenchyma, and abdominal aorta or
inferior vena cava.