COMPUTED TOMOGRAPHIC CHARACTERIZATION OF MALIGNANT PERITONEAL MESOTHELIOMA

Document Sample
COMPUTED TOMOGRAPHIC CHARACTERIZATION OF MALIGNANT PERITONEAL MESOTHELIOMA Powered By Docstoc
					Tumori, 91: 394-400, 2005


COMPUTED TOMOGRAPHIC CHARACTERIZATION OF MALIGNANT
PERITONEAL MESOTHELIOMA

Tristan Dongbo Yan1, Namik Haveric2, Carlos Pablo Carmignani1, Christina M Bromley3, and Paul H Sugarbaker1

1Program
            in Peritoneal Surface Malignancy, Washington Cancer Institute, Washington Hospital Center, Washington, DC, USA;
2
    The Department of Radiology, Washington Hospital Center, Washington, DC, USA; 3Biostat Solutions, Mt. Airy, MD, USA



Aims and background: Peritoneal mesothelioma is a rare disease            analysis of 16 abdominopelvic anatomic sites, the vesical or
with a universally fatal outcome when managed in a tradition-             rectal uterine pouch was involved in 97% and the greater
al palliative manner. New approaches to treatment using cy-               omentum in 91%. These anatomic sites were the only ones
toreductive surgery and intraperitoneal chemotherapy sug-                 with a positive Z-score of >1. In the analysis of 9 ab-
gest that long-term survival is possible in selected patients.            dominopelvic regions, the central and pelvic regions had
Early recognition of this disease process with an orderly sur-            Z-scores >2 for large volume disease >5 cm. For CT interpre-
gical approach will begin to optimize treatment.                          tative findings class I, class II and class III was determined in
Methods: Thirty-three patients with malignant peritoneal                  approximately one-third in each category. Sixty-six percent of
mesothelioma had CT scans available for review. A Z-score                 the patients had ascites by CT.
was used to evaluate the incidence of cancer at a particular              Conclusions: Malignant peritoneal mesothelioma by CT evalua-
anatomic site as compared to a general incidence of disease               tion predominates in tumor mass within the central and pelvic
at all sites. CT was analyzed by abdominopelvic anatomic                  portions of the abdomen. Minimal, moderate, and extensive
sites (16), abdominopelvic regions (9), and for presence ver-             small bowel enlargements were seen in roughly one-third of
sus absence of disease in the chest. Interpretative CT find-              the patients. With the use of the Z-score and interpretative
ings (class 0-III) were determined for these 33 patients.                 small bowel findings a radiologic characterization of this dis-
Results: Eight of 33 patients had pleural abnormalities. In an            ease for primary radiologic diagnosis is possible.
Key words: Abdominopelvic anatomic sites, ascites and omental disease, abdominopelvic regions, computed tomography, peri-
toneal cancer index, peritoneal mesothelioma.




Introduction                                                              68 peritoneal mesothelioma patients who underwent this
                                                                          comprehensive treatment, age less than 53, female gen-
   Malignant peritoneal mesothelioma is a rare neo-                       der, low volume of disease, and adequate tumor eradica-
plasm, accounting for approximately one-fifth of all                      tion by surgery resulted in improved survival. In view of
mesothelioma. In most published series it is usually a                    an increasing interest in this disease and the demonstra-
rapidly fatal peritoneal surface malignancy. In the ab-                   tion of a treatment-related survival benefit, improved ra-
sence of special treatments the median survival is less                   diologic description for accurate diagnosis is necessary.
than 1 year1-7. Due to low the incidence of the disease,                  The purpose of this paper is to identify the characteristic
the index of clinical suspicion is reduced. Also, patients                preoperative CT appearances of the disease in order to
often present with non-specific symptoms. Consequent-                     facilitate a radiologic diagnosis.
ly, patients with peritoneal mesothelioma may go undi-
agnosed for long periods of time. The diagnosis can be                    Patients and methods
very difficult and is usually not made from clinical and
radiological parameters. Biopsy with histologic and im-                      Seventy-one patients with peritoneal mesothelioma
munohistochemical findings are required for definitive                    underwent cytoreductive surgery and perioperative in-
diagnosis8,9. The use of computed tomography has not                      traperitoneal chemotherapy by the same surgical team
been established as a diagnostic tool in patient with ma-                 at the Washington Cancer Institute, between 1989 and
lignant peritoneal mesothelioma, even though CT ap-                       2003. All of these patients had definitive diagnosis by
pearances of this disease have been described. Mostly                     histology prior to the operation. Five patients had multi-
radiologic reports include a single case or small collec-                 cystic peritoneal mesothelioma and were excluded from
tions of patients10-12.                                                   this study. Among the remaining patients with malig-
   With the new comprehensive treatment modality con-                     nant peritoneal mesothelioma, 33 patients with avail-
sisting of cytoreductive surgery and perioperative in-                    able preoperative abdominal, pelvic and chest CT for
traperitoneal chemotherapy, recent reports increased the                  assessment form the basis of this study. The median in-
median survival to 67 and 97 months respectively13-15. In                 terval between CT and the operation was 2 days (range
the experience at the Washington Cancer Institute with                    1 to 19 days).


Correspondence to: Dr Paul H Sugarbaker, Washington Cancer Institute, 106 Irving Street NW, Suite 3900, Washington, DC 20010, USA.
Tel +202-877-3908; fax +202-877-8602; e-mail Paul.Sugarbaker@medstar.n

Received May 17, 2005; accepted June 6, 2005.
CT OF PERITONEAL MESOTHELIOMA                                                                                                     395

Computed tomographic scans                                      CT assessment of LS was categorized into three
   CT scans were performed with a GE 9800 Hilight CT          groups: 0, no detectable disease; 1, minimal disease (tu-
scanner (General Electric, Milwaukee, WI) using 1-cm          mor diameter <0.5 cm); 2, moderate disease (tumor di-
continuous slice thickness. The radiologic studies were       ameter ≥0.5 cm and <5 cm) and 3, gross disease (tumor
performed following the administration of oral and in-        diameter >5 cm).
travenous contrast media in all patients. These were 2-       Interpretative CT findings of small bowel
second scan times with a 3- to 4-second interscan delay       and its mesentery
for breathing. Some patients had three to five slices ob-
tained between consecutive breath holds. All CT scans            As shown in Table 1, characteristic interpretative CT
were performed following angiodynamic bolus at a rate         appearances of the small bowel and its mesentery were
of 1 to 2 mL/second, for a total volume of 150 to 180         classified into four groups. Class 0 – indicated a normal
mL of iothalamate meglumine 60% (Connray 60 ®,                appearance. The jejunal and ileal vessels appeared as
Mallinckrodt, St. Louis, MO) or iohexol 240 (Omni-            round and curvilinear densities within the mesenteric
paque®, Sanofi Winthrop, New York, NY). A 30- to 60-          fat, central to the small bowel loops. Class I – indicated
second delay occurred between the initiation of contrast      that the small bowel and the mesenteric surfaces were
administration and the start of CT scanning. Precontrast      separated by fluid accumulation. The jejunal and ileal
bowel preparation included a total of 900 to 1,200 mL         vessels were easily identified within the mesenteric fat
of barium ingested at least six hours before CT, and rec-     (Figure 2). Class II – indicated a layering of solid tumor
tal barium contrast administered immediately before           on the surfaces of the small bowel and/or its mesentery.
CT. For clinical use, the CT scans were read by staff ra-     The peritoneal lining appeared thickened and enhanced.
diologist. For the current clinical research study, all CT    The mesenteric leaves remained separated by ascites
scans were reread in order to score abdominal and
pelvic tumor deposits by a standardized data sheet. The
radiologist was aware that all patients in the study had
clinical evidence of peritoneal mesothelioma, but was
                                                                                                       Abdominopelvic
masked to the operative findings.                                                                      regions
Chest CT                                                                                               0 Central
                                                                                                       1 Right upper
   Chest CT was evaluated in terms of presence vs. ab-                                                 2 Epigastrium
sence of the five parameters: pleural thickening, pleural                                              3 Left upper
effusion, pleural metastases, pleural lymph nodes and                                                  4 Left flank
                                                                                                       5 Left lower
pericardial thickening.                                                                                6 Pelvis
                                                                                                       7 Right lower
Abdominopelvic anatomic sites                                                                          8 Right flank
   The presence versus absence of peritoneal mesothe-
lioma was assessed in 16 anatomic sites: undersurface                                                  Lesion size score
of right hemidiaphragm, undersurface of left hemidi-                                                   LS 0 No tumor seen
aphragm, hepatic surface, splenic surface, stomach/duo-                                                LS 1 Tumor up to 0.5 cm
denum, lesser omentum, subpyloric space, pancreatic                                                    LS 2 Tumor up to 5.0 cm
surface, parietal peritoneum, greater omentum, small                                                   LS 3 Tumor >5.0 cm
bowel and its mesentery, large bowel and its mesentery,                                                     or confluence
right paracolic gutter, left paracolic gutter, rectovesical
or rectouterine pouch and retroperitoneum16. The inci-
dence and the location of any isolated large tumor mass
that could be interpreted as an epicentric mesothelioma
                                                              Figure 1 - The abdominopelvic regions 0-8 were scored for a lesion size
primary site were also recorded.                              0-3.

Abdominopelvic regions
   The abdominopelvic regions were defined by two
sagittal planes through the mid-clavicular line and two       Table 1 - Interpretative CT classification of small bowel and its
transverse planes, one through the anterior superior ili-     mesentery. The classification was recorded for abdominopelvic
                                                              regions 9 through 12
ac spines and the other through the most caudal point
of the costal margins17. Lesion size (LS) was evaluated       Interpretation CT    Presence       Tumor layering       Adherence of
                                                                classification     of ascites     of small bowel      leaves of small
in these abdominopelvic regions (AR): periumbilical                                              and its mesentery     bowel and its
region (AR-0), right upper quadrant (AR-1), epigastri-                                                                  mesentery
um (AR-2), left upper quadrant (AR-3), left flank                 Class 0             No                No                  No
(AR-4), left lower quadrant (AR-5), pelvis (AR-6),                Class I             Yes               No                  No
right lower quadrant (AR-7) and right flank (AR-8)                Class II            Yes               Yes                 No
(Figure 1).                                                       Class III           Yes               Yes                 Yes
396                                                                                                  TD YAN, N HAVERIC, CP CARMIGNANI ET AL


and the small bowel mesenteric vessels were identifi-                           Statistics
able (Figure 3). Class III – indicated progressive solid                           The Z-scores were used to evaluate the incidence of
tumor involvement with an adherence of adjacent loops                           tumor implants at a particular anatomic site.18 The Z-
of small bowel. The configuration of small bowel and
                                                                                score is a statistical method for measuring the frequen-
its mesentery appeared matted, distorted and grossly
thickened. Often segmental small bowel obstruction                              cy of tumor at a single site in standard deviation units
was present. Intraperitoneal fluid could be loculated.                          relative to the mean tumor frequency at all sites.
The small bowel mesenteric vessels were difficult to de-                           A sample Z-score:             Z=x-x
fine due to sporadic loss of mesenteric fat density (Fig-                                                         SD
ure 4).
   The small bowel was designated as class 0-3 in the                             Where x = tumor frequency at a single site; x = mean
abdominopelvic regions 9-12; proximal jejunum (AR-                              tumor frequency at all sites; and SD = one standard de-
9); distal jejunum (AR-10); proximal ileum (AR-11)                              viation.
and distal ileum (AR-12) as shown in Figure 5.
                                                                                  A Z-score greater than +1 or less than -1 indicates
Ascites                                                                         that the observed data is different from 68% of the pop-
   Intraperitoneal fluid was quantified as minimal when
only a trace of intraperitoneal fluid was found in the
pelvis or paracolic gutters; moderate when collected
around perihepatic and perisplenic spaces and marked
when present throughout the abdomen and pelvis. As-
cites in the lesser sac was recorded as present vs. absent.




                                                                                Figure 4 - Class III – indicated gross involvement of the small bowel and
                                                                                its mesentery by tumor. There was an adherence of adjacent loops of
                                                                                small bowel by tumor.




Figure 2 - Class I – indicated the separation of the small bowel and its
mesentery due to the presence of ascites fluid.




Figure 3 - Class II – indicated the presence of peritoneal thickening and en-
hancement due to coating of the small bowel and its mesentery by tumor.         Figure 5 - Abdominopelvic regions for the small bowel.
CT OF PERITONEAL MESOTHELIOMA                                                                                              397

ulation. Similarly, a Z-score of +2 or -2 indicates that          Table 3 - CT findings in 16 abdominopelvic anatomic sites in
                                                                  33 patients with malignant peritoneal mesothelioma. The rela-
the observed data is different from 95% of the popula-            tive involvement of sites was based upon the Z-score calcu-
tion and +3 or -3 indicates that the observed data is dif-        lated from standard deviation from the mean
ferent from 99.7% of the population.
                                                                  CT findings                           Cases   %    Z-score

Results                                                           Right hemidiaphragm                    25     76      0.28
                                                                  Left hemidiaphragm                     22     67    - 0.07
                                                                  Liver surface                          24     73      0.16
  Thirty-three patients with peritoneal mesothelioma              Splenic capsule                        22     67    - 0.07
and a preoperative CT were evaluated. All patients had            Stomach/duodenum                       18     55     -0.54
                                                                  Lesser Sac                             23     70      0.04
a histological diagnosis of malignant peritoneal                  Subpyloric space                       21     64    - 0.19
mesothelioma. Twenty-one patients (63%) were male                 Pancreatic surface                      6     18     -1.94 *
and 12 patients (36%) were females. The mean age was              Parietal peritoneum                    27     82      0.51
                                                                  Greater omentum                        30     91      0.86
49 at the time of diagnosis.                                      Small bowel and its mesentery          27     82      0.51
                                                                  Large bowel and its mesentery          28     85      0.63
Analysis of CT findings of the chest                              Right paracolic gutter                 29     88      0.74
                                                                  Left paracolic gutter                  28     85      0.63
   Among the 33 patients, none had pulmonary parachy-             Rectovesical/rectouterine pouch        32     97      1.1 *
mal metastases, medistinal lymph nodal involvement or             Retroperitoneal space                   0      0    - 2.6 *
pericardial thickening. A total of 8 patients had pleural         *Indicates a z-score greater than +1 or -1.
thickening, which was in the basal areas in all patients.
Five patients had pleural effusion and 4 out of these 5
patients also had pleural thickening (Table 2).                   cm mass on the parietal peritoneum in the left lower
Analysis by 16 abdominopelvic anatomic sites                      quadrant, occupying the left paracolic gutter and ex-
                                                                  tending into the left inguinal canal.
   The most common sites of disease occurrence were                  In an analysis of the 16 abdominopelvic anatomic
rectovesical or rectouterine pouch (97%), greater omen-           sites by Z-score, only the rectovesical/rectouterine
tum (91%) and right paracolic gutter (88%) (Table 3).             pouch had a positive score of more than one. The pan-
Twenty-five patients (76%) had tumor present on the               creatic surface had a Z-score of -1.94 and the retroperi-
undersurface of right hemidiaphragm vs 22 patients                toneal space a Z-score of -2.6.
(67%) on the undersurface of the left hemidiaphragm.
All 22 patients with disease on the left hemidiaphragm,           Analysis by 9 abdominopelvic regions
had concomitant right-sided disease. Disease on the he-             The abdominopelvic region 0 showed a large volume
patic and splenic surfaces was usually separated from             of disease in 70% of patients and an absent or low vol-
the disease of the undersurface of the diaphragm by in-           ume of disease in 9%. The right upper quadrant showed
traperitoneal fluid. At no time was there any intrahepat-         moderate volume of disease in 61%. This moderate vol-
ic or intrasplenic metastasis. Disease in the lesser sac          ume was also observed in the left flank (58%), left lower
was also common in this disease (70%). Despite great              regions (52%), right lower (53%) and right flank (58%).
volume of disease in some of the patients in and around           A large volume of disease was observed in the pelvis in
the hepatic hilum, no biliary tract obstruction was               67% of patients (Table 4).
found. Also, no matter how extensive the intraperitoneal            In an analysis of the 9 abdominopelvic regions by Z-
volume of disease, there was no extraperitoneal disease           score, large volume disease was more frequently ob-
extension to lymph nodes or systemic sites. Three pa-             served in the central and pelvic regions. Figure 6 shows
tients (2 ipsilateral and 1 bilateral) were found to have         the mesothelioma tumor distribution by Z-score for tu-
ureteral obstruction by pelvic tumors.                            mor volume greater than 5 cm. The central region
   Two patients were found to have large isolated tumor           which contains the omentum and the pelvis both had z-
masses at a specific abdominopelvic sites. One of them            scores greater than 2.
had a 7 cm tumor mass in the mesentery of the proximal
and distal segments of jejunum. It “clumped” the                  Analysis by interpretative CT findings of small bowel
mesentery together, however it did not cause any small            and its mesentery
bowel obstruction. The other patient had an isolated 12              A moderate volume of disease as estimated by a clas-
                                                                  sification of Z was observed in all four small bowel re-
                                                                  gions. If the small bowel classification was determined
Table 2 - Chest CT findings in 33 patients with malignant peri-
toneal mesothelioma. Four of the five patients with pleural       by the largest involvement at any of the 4 small bowel
effusion also had pleural thickening                              regions class 0 was seen in 3 patients (9%), class I in 9
CT Findings                           Cases            %
                                                                  (27%), class II in 9 (27%) and class III in 12 (36%).
Pleural effusion                        5              15         Ascites
Pleural thickening                      8              24
Pleural metastasis                      0               0           Sixty-six percent of the patients had ascites. Thirteen
Pleural lymph nodal involvement         0               0         patients (41%) had marked ascites and 10 of them
Pericardial thickening                  0               0         (31%) had ascites in both greater sac and lesser sac.
398                                                                                            TD YAN, N HAVERIC, CP CARMIGNANI ET AL


Table 4 - CT findings in 9 abdominopelvic regions in 33
patients with malignant peritoneal mesothelioma. The relative
involvement of the regions was based upon the Z-score
calculated from the standard deviation from the mean

Abdominopelvic       Lesion size      No.       %          Z-score of                                           Regions            Lesion size
region 0-8            of tumor                           lesion size 1-3                                        0 Central              2.2
                                                                                                                1 Right upper         -1.2
0 – central               -            -         -                                                              2 Epigastrium         -0.03
                          0            2         6           - 1.2*
                                                                                                                3 Left upper          -0.35
                          1            1         3           - 1.3*
                          2            6        18           - 0.51                                             4 Left flank          -1.2
                          3           23        70             2.2*                                             5 Left lower          -0.03
                                                                                                                6 Pelvis               2.01
1 – right upper           -            -         -                                                              7 Right lower          0.03
                          0            4        12           - 0.83                                             8 Right flank          0.51
                          1            6        18           - 0.51
                          2           20        61             1.7*
                          3            2         6           - 1.2*

2 – epigastrium           -            -         -
                          0            9        27           - 0.03
                          1            5        15           - 0.67
                          2            9        27           - 0.03
                          3            9        27           - 0.03

3 – left upper            -            -         -                         Figure 6 - Mesothelioma distribution in abdominopelvic regions 0-8 by
                          0            4        12           - 0.83        Z-score for tumor volume greater than 5 cm.
                          1            7        21           - 0.35
                          2           14        42             0.76
                          3            7        21           - 0.35

4 – left flank            -            -         -                         toneal cavity. These data show a distribution of this dis-
                          0            5        15             0.67        ease that is diffuse throughout the abdominal and pelvic
                          1            6        18             0.51
                          2           19        58             1.56*       surfaces in most patients. Only two patients had a large
                          3            2         6           - 1.2*        solitary mass that could have been misinterpreted as the
5 – left lower            -            -         -
                                                                           epicenter of a primary abdominal or pelvic cancer.
                          0            3        27           - 0.99        Large volume disease simultaneously present in ab-
                          1            3        27           - 0.99        dominopelvic regions 0 and 6 was a frequent finding.
                          2           17        52             1.24*
                          3            9        27           - 0.03        This may be a characteristic CT finding in peritoneal
                                                                           mesothelioma. Analysis of the distribution of the dis-
6 – pelvis                -            -         -                         ease by abdominopelvic anatomic sites supported this
                          0            1         3           - 1.31*
                          1            1         3           - 1.31*       impression. Our analysis by Z-score supported the fact
                          2            8        27           - 0.19        that a majority of patients will show large volume dis-
                          3           22        67             2.04*       ease in the mid-abdomen and pelvis. This CT finding
7 – right lower           -            -         -                         should raise a clinical suspicion that a patient with ma-
                          0            2         6           - 1.15*       lignant ascites may have malignant peritoneal mesothe-
                          1            4        12           - 0.83        lioma.
                          2           17        53             1.24*
                          3            9        27           - 0.03           The interpretative CT classification of malignant
                                                                           peritoneal mesothelioma present on small bowel and its
8 – right flank           -            -         -                         mesentery is an organized description of progressive in-
                          0            5        15           - 0.67
                          1            2         6           - 1.2*        volvement. It describes an increasing severity of the ag-
                          2           19        58             1.6*        gressive nature of the disease and the volume of cancer
                          3            6        18           - 0.51        that is present on the intestine and the mesentery. Thir-
0, No tumor; 1, tumor diameter <0.5 cm; 2, tumor diameter is 0.5 - 5 cm;   ty-three percent of these patients showed class III find-
3, tumor diameter > 5 cm; No., number of patients; NS, not significant.    ings on the small bowel. Previous work in this CT
*indicates a z-score greater than +1 or -1.                                analysis suggested that these patients had a poor prog-
                                                                           nosis and were not good candidates for combined surgi-
                                                                           cal and regional chemotherapeutic interventions19. This
                                                                           interpretative classification of the CT findings of small
Discussion                                                                 bowel should allow the radiologist to provide important
                                                                           information about the extent of disease on the small
   This study is the first of its kind to describe in a sys-               bowel and the functional deficit that may be expected to
tematic manner the images recorded on abdominal and                        occur within the near future in the absence of definitive
pelvic CT in this rare disease. This analysis allows the                   treatment14,15.
physician to begin to compare the findings seen in pa-                        It is important to note with the interpretative CT clas-
tients with peritoneal mesothelioma to the other dis-                      sification that a quantitative involvement of the small
eases which result in a spread of cancer within the peri-                  bowel and its mesentery is described. Minimal changes
CT OF PERITONEAL MESOTHELIOMA                                                                                                          399

Table 5 - CT findings in 4 portions of the small bowel and its              adenocarcinomas that tend to compartmentalize the
mesentery in 33 patients with malignant peritoneal mesothe-
lioma. A CT classification of ascites and solid tumor was on a              small bowel and isolate it from the large volume of tu-
0-3 scale. The relative involvement of the regions was based on             mor present elsewhere within the abdomen16. Of note is
a Z-score calculated from the standard deviation from the mean              the lack of a large volume disease in peritoneal mesothe-
Abdominopelvic       Small bowel     No.        %          Z-score of       lioma above the right hemidiaphragm characteristically
region 9-12          configuration                     classification 1-3   noted with pseudomyxoma peritonei. Table 6 contrasts
9 – upper jejunum         -            -         -
                                                                            the abdominal and pelvic CT in malignant peritoneal
                          0            6        18           - 0.51         mesothelioma and in pseudomyxoma peritonei.
                          1            7        21           - 0.35            The biology of this disease is a field defect on peri-
                          2           16        48             0.92
                          3            3        23           - 0.83         toneal surfaces. Our data from CT strongly supports the
                                                                            diffuse natural history of this disease. In an important
10 – lower jejunum        -            -         -                          contrast to other gastrointestinal and gynecologic malig-
                          0            6        18           - 0.51
                          1            7        21           - 0.35         nancies is the lack of a primary site. In only 2 of the 33
                          2           15        45             0.92         patients studied was there a single large focus of disease
                          3            4        39           - 0.83         that could be interpreted as an epicenter of the primary
11 – upper ileum          -            -         -                          malignancy. Also, none of the 33 patients had extraperi-
                          0            5        15           - 0.67         toneal involvement to lymph nodes or metastasis. Other
                          1            9        27           - 0.03
                          2           15        45             0.92         clinical studies of patients with early disease support the
                          3            3         9           - 0.99         lack of a localized primary site for this cancer22.
                                                                               The pathobiology of other cancers that involve the
12 – lower ileum          -            -         -
                          0            5        15           - 0.67         peritoneal surfaces has been previously published 16.
                          1            7        21           - 0.35         These studies suggest that each malignant process may
                          2           14        42             0.76         have its own characteristic pattern of distribution within
                          3            6        18           - 0.51
                                                                            the abdominal and pelvic space. Both the cancer cells
0, no tumor; 1, tumor diameter <0.5 cm; 2, tumor diameter is 0.5-5 cm; 3,   within the peritoneal cavity and the physical factors
tumor diameter >5 cm; No., number of patients.
                                                                            from the host go together to make up a complex patho-
                                                                            physiology that bring about the characteristic distribu-
                                                                            tion for a particular disease. The ability of cells to in-
                                                                            vade a peritoneal surface, the presence of a mucinous or
of the normal images occur with the class I findings.                       fluid vehicle to distribute the cells, the forces of gravity,
Ascites fluid separates the loops of small bowel and                        the primary site from which cancer cells are released,
small bowel mesentery but does not otherwise distort                        the flow of intraperitoneal fluid, and the occurrence of
the normal anatomy. In class II a layering of tumor is                      previous surgical procedures all have been identified as
imaged on the small bowel and mesenteric surface. This                      biologic or physical determinants of cancer cell distrib-
layering is usually attributed to many thousands of                         ution and progression within the peritoneal space. Ma-
small nodules that progress to form a distinctive coating                   lignant peritoneal mesothelioma has its own unique bi-
of these peritoneal surfaces. The tumor nodules may be                      ology and our findings suggest that it has a distinctive
of multiple different sizes so that there is some irregu-                   pattern of progression on peritoneal surface that may be
larity to this coating of the small bowel and its mesen-                    of great help in making a radiologic diagnosis of this
tery. In class III there are gross nodular thickenings of                   disease.
this peritoneal surface and evidence of invasion of the                        A thorough knowledge of the pathophysiology of the
mesentery by larger cancer nodules. Adjacent bowel                          multiple different types of peritoneal surface malignan-
loops appear adherent suggesting that surgical access to                    cy is essential to a knowledgeable radiologic interpreta-
the abdomen may be very limited. The normal architec-                       tion of the abdominal and pelvic CT. No longer should
ture of the bowel is beginning to disappear and evi-                        the description of peritoneal surface malignancy (carci-
dence of disruption of function as evidenced by seg-                        nomatosis) be merely present vs. absent. Both quantita-
mental obstruction of the bowel may be apparent.                            tive and qualitative descriptions are possible.
Sometimes these findings are more characteristic in
some portions of the bowel than in others.
   This analysis of CT findings in peritoneal mesothe-
lioma allows some important comparisons and contrasts                       Table 6 - Comparison of abdominal and pelvic CT of malignant
to be made with other diseases that predominantly affect                    peritoneal mesothelioma and pseudomyxoma peritonei
the peritoneal surfaces. Previous experience of our group                   Large volume disease      Malignant peritoneal   Pseudomyxoma
with the radiology of pseudomyxoma peritonei allows us                                                  mesothelioma            peritonei
to make some observations20,21. Certainly, the general-                     Beneath hemidiaphragms         Moderate            Extensive
ized pattern of involvement as seen with peritoneal                         Greater omentum                Moderate            Extensive
mesothelioma is in great contrast to the “redistributed                     Small bowel and small
                                                                               bowel mesentery                Yes                Spared
pattern” seen with pseudomyxoma peritonei syndrome.                         Pelvis                            Yes                 Yes
Also, the wide variety of images seen with the interpre-                    Ascites                       Serous fluid       Globular mucin
tative CT classification is in great contrast to mucinous                   Primary site                     None              Appendix
400                                                                                    TD YAN, N HAVERIC, CP CARMIGNANI ET AL


References

 1. Brenner J, Sordillo PP, Magill GB, Golbey RB: Malignant         12. Guest PJ, Reznek RH, Selleslag D, Geraghty R, Slevin M:
    peritoneal mesothelioma. Review of 25 patients. Am J Gas-           Peritoneal mesothelioma: the role of computed tomography
    troenterol, 75: 311-313, 1981.                                      in diagnosis and follow up. Clin Radiol, 45: 79-84, 1992.
 2. Antman K, Shemin R, Ryan L, Klegar K, Osteen R, Herman          13. Sebbag G, Yan H, Shmookler BM, Chang D, Sugarbaker PH:
    T, Lederman G, Corson J: Malignant mesothelioma: prognos-           Results of treatment of 33 patients with peritoneal mesothe-
    tic variables in a registry of 180 patients, the Dana-Farber        lioma. Br J Surg, 87: 1587-1593, 2000.
    Cancer Institute and Brigham and Women’s Hospital experi-       14. Feldman AL, Libutti SK, Pingpank JF, Bartlett DL, Beresnev
    ence over two decades, 1965-1985. J Clin Oncol, 6: 147-153,         TH, Mavroukakis SM, Steinberg SM, Liewehr DJ, Kleiner
    1988.                                                               DE, Alexander HR: Analysis of factors associated with out-
 3. van Gelder T, Hoogsteden HC, Versnel MA, de Beer PH,                come in patients with malignant peritoneal mesothelioma un-
    Vandenbroucke JP, Planteydt HT: Malignant peritoneal                dergoing surgical debulking and intraperitoneal chemothera-
    mesothelioma: a series of 19 cases. Digestion, 43: 222-227,         py. J Clin Oncol, 21: 4560-4567, 2003.
    1989.                                                           15. Sugarbaker PH, Welch LS, Mohamed F, Glehen O: A review
 4. Ascensio JA, Goldblatt P, Thomford NR: Primary malignant            of peritoneal mesothelioma at the Washington Cancer Insti-
    peritoneal mesothelioma. A report of seven cases and a re-          tute. Surg Oncol Clin N Am, 12: 605-621, 2003.
    view of the literature. Arch Surg, 125: 477-481, 1990.          16. Carmignani P, Sugarbaker TA, Bromley CM, Sugarbaker PH:
 5. Goldblum J, Hart WR: Localized and diffuse mesotheliomas            Intraperitoneal cancer dissemination: Mechanisms of the pat-
    of the genital tract and peritoneum in women. A clinico-            terns of spread. Cancer Metastases Rev, 22: 465-472, 2003.
    pathological study of nineteen true mesothelial neoplasms       17. Jacquet P, Sugarbaker PH: Current methodologies for clinical
    other than adenomatoid tumors, multicystic mesotheliomas            assessment of patients with peritoneal carcinomatosis. J Exp
    and localized fibrous tumors. Am J Surg Pathol, 19: 1124-           Clin Cancer Res, 15: 49-58, 1996.
    1137, 1995.                                                     18. Weyn B, Jacob W, da Silva VD, Montironi R, Hamilton PW,
 6. Eltabbakh GH, Piver MS, Hempling RE, Recio FO, Intengen             Thompson D, Bartels HG, Van Daele A, Dillon K, Bartels
    ME: Clinical picture, response to therapy, and survival of          PH: Data representation and reduction for chromatin texture
    women with diffuse malignant peritoneal mesothelioma.               in nuclei from premalignant prostatic, esophageal, and
    J Surg Oncol, 70: 6-12, 1999.                                       colonic lesions. Cytometry, 41: 133-138, 2000.
 7. Kerrigan SA, Turnnir RT, Clement PB, Young RH, Churg A:         19. Yan TD, Haveric N, Carmignani P, Chang D, Sugarbaker PH:
    Diffuse malignant epithelial mesotheliomas of the peri-             Abdominal computed tomographic scans in the selection of
    toneum in women. Cancer, 94: 378-385, 2002.                         patients with malignant peritoneal mesothelioma for compre-
 8. Sugarbaker PH, Acherman YIZ, Gonzales-Moreno S, Ortega-             hensive treatment with cytoreductive surgery and periopera-
    Perez G, Stuart OA, Marchettini P, Yoo D: Diagnosis and             tive intraperitoneal chemotherapy. Cancer, 103: 839-849,
    treatment of peritoneal mesothelioma: the Washington Can-           2005.
    cer Institute experience. Semin Oncol, 29: 51-61, 2002.         20. Archer AG, Sugarbaker PH, Jelinek JS: Radiology of peri-
 9. Alcorn KW, Yan H, Shmookler BM, Sugarbaker PH: Differ-              toneal carcinomatosis. In: Peritoneal carcinomatosis: princi-
    ential diagnosis of the simultaneous occurrence of colon car-       ples of management, Sugarbaker PH (Ed), 263-288, Kluwer,
    cinoma and peritoneal mesothelioma by immunohistochem-              Boston, 1996.
    istry. Surg Rounds, 23: 411-417, 2000.                          21. Jacquet P, Jelinek JS, Chang D, Koslowe P, Sugarbaker PH:
10. Reuter K, Raptopoulos V, Reale F, Krolikowski FJ, D’Orsi            Abdominal computed tomographic scan in the selection of
    CJ, Graham S, Smith EH: Diagnosis of peritoneal mesothe-            patients with mucinous peritoneal carcinomatosis for cytore-
    lioma: computed tomography, sonography, and fine-needle as-         ductive surgery. J Am Coll Surg, 181: 530-538, 1995.
    piration biopsy. AJR Am J Roentgenol, 140: 1189-1194, 1983.     22. Sugarbaker PH, Yan H, Grazi R, Shmookler B: Early local-
11. Ros PR, Yuschok TJ, Buck JL, Shekitka KM, Kaude JV:                 ized peritoneal mesothelioma as an incidental finding at la-
    Peritoneal mesothelioma. Radiologic appearances correlated          paroscopy: Report of a case and implications regarding natur-
    with histology. Acta Radiol, 32: 355-358, 1991.                     al history of the disease. Cancer, 89: 1279-1284, 2000.