Journal of the Korean Radiological Society 1995 : 32(6) : 937 - 942

Intraperitoneal Fluid Collection : CT Characteristics in Dertermining the Causes1

Mi Young Kim, M.D. , Chang Hae 5uh , M.D. , Won Kyun Chung , M.D. ,

Chong 500 Kim, M.D .2, Ki Chul Choi , M.D.2

Purpose: Abdominal CT scans in patients with intraperitoneal fluid were retrospectively studied to identify characteristic features useful for differential diagnosis of various causes.

Materials and Methods: One hundred and seventy patients with intraperitoneal fluid collection were classified as categories of hepatic disease, carcinomatosis,

and infectious disease. We analyzed sites of fluid collection , the presence of per­itoneal thickening , omental and mesenteric fat infiltration, and Iymph node enlargment.

Results: Intraperitoneal fluid was present in subhepatic space, subphrenic space, paracolic gutter, mesentery, and fossa ofthe gallbladder in decreasing or­der of frequency. Fluid in the gallbladder fossa was the most frequent in hepatic diseases. The fluid collection in subhepatic and subphrenic space was less fre ­quent in infectious diseases . Peritoneal thickening was noted in infectious dis­eases, and carcinomatosis . Omental fat infiltration and enlarged Iymph nodes were the most frequent in carcinomatosis (58% and 44% , respectively), whereas,

mesenteric fat infiltration and enlarged Iymph nodes were the most common in infectiousdiseases (61% , and 26% , respectively).

Conclusion: The location of peritoneal fluid collection showed some lesion specific characteristics, and CT features of fat infiltration and enlarged Iymph nodes of peritoneum, omentum, and mesenterywere helpful for differential diag­nosis between carcinomatosis and infectious diseases.

Index Words : Ascites Peritoneum , CT Peritoneum , fluid

INTRODUCTION

Several radiologic modal ities are used to assess intraperitoneal lesions associated with ascites, includ­ing ultrasonography (US) , CT, or MR. The goals of these modalit ies are to determine the location , amount , and cause of intraperitoneal fluid. CT was proven to be accurate in the evaluation of location or assessment of intraperitoneal flu id amount (1 , 2). To date, evaluation of the cause of the site -specific intraperitoneal fluid

'0epartment olRadiology. lnha University Hosp ital 20epartment 01 Oiagnostic Radiology. Chonbuk National University Medical School Received September 4,1995 ; Accepted June 12, 1995 Address reprint requests to: Oepartment 01 Rad i이。gy ， Inha Un iversi ty Hospi tal ~ 3309.327 , Taepyung.dong, Suchung.gu , Sungnam City , Kyunggi . d。

461.192, Korea Fax. 82- 342- 755- 2812 TeL 82- 342- 720-5222

collection has not been undertaken. Although CT findings such as peritoneal nodules , omental or mes­enteric masses, tethered bowelloops , and contrast en­hancement or thickened peritoneum may suggest ma­lignancy, peritoneal fluid per se occuring in a variety of intraperitoneal disease usually do not provide a clue to a specific diagnosis (1 -5). Can abdominal CT accu­rately help differentiating intraperitoneal lesions ac­companied with fluid? This study attempted to estab­lish lesion - specific CT features of intraperitoneal fluid collection which may be valuable in management of the patients with intraperitoneal fluid.

MATERIALS and METHODS

The study population consisted of 170 patients , 112 men and 58 women raging in age from 7 to 86 years (mean , 49 years ). AII patients had taken CT scan , and

” / %

Journal of the Korean Radiological Society 1995; 32(6) : 937-942

a b Fig. 1. Pancreatic cancer with liver metastasis and peritoneal seeding in a 73-year-old woman. a. Contrast enhanced CT scan shows sharply delined enhancing mass or Iymph nodes 01 peritoneum (arrowheads). Low attenuated pancreatic tail mass (arrows) and mul­tiple liver metastases are also noted. b. CT scan shows omental cakes with irregular lat infiltration (arrowheads) and m비tiple

Iymphnodes (arrows)

the films were evaluated at the department of radiology of 1 nha Hospital , between January 1992 and September 1994. Final diagnosis included hepatoma (n=46) , liver cirrhosis (n=33) , carcinomatosis (n=36) , infectious dis­ease (n =23) , traumatic hemoperitoneum (n =15) , pan­creatitis (n=12) , intestinal obstruction (n=3) , and Bu­dd-Chiari syndrome (n=2). The diseases encounter­ed were classified into 1) hepatic disease, 2) carci­nomatosis , or 3) infectious disease. Hepatic diseases included hepatoma and liver cirrhosis. Carcinomatosis included stomach , biliary , pancreatic , ovarian , gall­bladder, and rectal cancers in 12, 7, 6, 4, 3, and 2 patients, respectively , and unknown metastatic adeno­carcinoma in 2 patients. The diseases of infectious ori­gin included tuberculosis , panperitonitis or abscess with visceral perforation , and diverticulitis in 13, 9, and 1 patients, respectively. The diagnosis was confirmed by operation followed by pathologic examination of the lesions, and peritoneoscope or ultrasound guided bi­opsy. Clinical data and surgical findings were compiled from the pati ents’ medical records.

The CT scanners used were a GE 9800 (GE Medical System , Milwaukee, Wisconsin , USA) and a Shimadzu SCT 2000 T (Kyoto , Japan). Unenhanced and contrast enhanced scans were obtained in all patients , with 5 to 10mm collimation at intervals of 10mm, from the dia­phragm to iliac crest with axial scans. AII patients received 250 cc of oral contrast material (370 mg/ml , Gastrografin diluted 1 : 40 in water , Schering , Seoul , Korea) , and 100-150ml of intraveneous contrast me­dia (iodine 300 mg/ml , lopromide, SChering , Seoul , Korea).

Two radiologists(MYK , CHS) jointly interpreted the CT findings without prior knowledge of the patient’s specific diagnosis. CT scans were reviewed and ev­aluated for the following find ings : (a) the locations of

peritoneal fluid , such as subphrenic space, subhepatic space, paracolic gutter, fossa of the gallbladder and lesser sac , (b) the CT features suggesting lesional spread pattern , such as peritoneal thickening and Iy­mph nodes (Fig. 1 a) , omental fat infiltration and Iymph nodes (Fig. 1 b) , mesenteric fat infilatration (Fig. 2a) and Iymph nodes (Fig. 2b). Peritoneum was considered to be thickened , if it was twice the normal thickness with or without abnormal enhancement. Although a Iym­phadenopathy over 10 mm in diameter is generally considered as abnormal , Iymph node over 5 mm in di­ameter was considered to be enlarged for this study in order to include small Iymph nodes of peritoneum or omentum. Diagnostic criterion of fat infiltation was lin­ear fibrous densities throughout the omentum and mesentery associated with thickening and indistinct­ness of vasculature. The diseases thus reviewed were compared in terms of the location of fl uid , and involve­ment of peritoneum , omentum , and mesentery

The statistical significance was obtained using the chi -square test (or Fisher’s exact test where indi­cated). For statistical analysis , P value less than .05 was considered sign ificant

RESUlTS

The fluid was present in subhepatic space in 132 (78%) , subphrenic space in 131 (77 %), paracolic gutter in 121 (71 %), mesentery in 66 (39%) , fossa of the ga­Ilbladder in 50 (29 %), lesser sac in 17 (10 %) , ret­roperitoneum in 15 (9 %) , and plerual space in 35 (21 %) patients. Intraperitoneal fluid of subhepatic and subphrenic space was seen in 85/87 %, 81/83 %, 61/ 48 % in hepatic disease, carcinomatosis , and infectious disease, respectively. Fluid in paracolic gutter was detected in 75% , 67 % , 57 % of the hepatic diseases,

- 938 -

Mi Young Kim, et al : Intraperitoneal Fluid Collection

a b Fig. 2. Generalized ascites with tuberculous peritonitis in a 32-year-old woman. a. Contrast enhanced CT scan shows irregular mes­enteric fat infiltration with increased linear fibrous densities (arrows). b. CT scan demonstrates mesenteric Iymph node (arrow) and ir­regular mesenteric fat infiltration.

carcinomatosis , and infectious diseases, respectively. Fluid in the mesentery was noted in 44%, 42%, and 39% ofthe hepatic disease, carcinomatosis , and infec­tious diseases, respectively. Fluid in the gallbladder fossa was present in 41 %, 28 %, and 13 % of the hepatic diseases, carcinomatosis , and infectious diseases The frequency of fluid collection in subhepatic space , subphrenic space, and fossa of gallbladder was signifi­cantly different among the disease groups (p < 0.05) (Fig.3).

Comparison of spread pattern of the lesions involv­ing peritoneum , omentum , and mesentery is shownin Fig. 4. Peritoneal thickening was detected in 26% and 19% of the infectious diseases and carcinomatosis , re­spectively. Omental fat infiltration was present in 58 %, 38%, and 30% of carcinomatosis , hepatic diseases, and infectious diseases, respectively. Mesenteric fat infiltration was noted in 61 %, 42% , and 32% of the in­fectious disease, carcinomatosis , and hepatic disea­ses , respectively. Peritoneal Iymph nodes were de­tected in 8% of the carcinomatoses. Omental Iymph node was detected in 44% and 13 % of the carcino­matoses and .infectious diseases. Mesenteric Iymph node was noted in 26% and 17% of the infectious dis­eases and carcinomatosis. Incidences of peritoneal thickening , enlarged omental Iymph nodes, presence of mesenteric fat infiltration , and enlarged mesenteric Iymph nodes were significantly different among the di­sease groups (p<0.05).

DISCUSSION

The major role ofCT is not onlyto makethe diagnosis of intraperitoneal fluid collection but also to identify the cause of ascitic fluid. Hepatic ascites associated with mass or cirrhosis is often advanced and clinically evi­dent before the onset of intraperitoneal fluid. There-

fore , it is important to differentiate the peritoneal seed­ing of carcinoma from benign process such as infec­tious or inflammatory condition. Although various CT features have been used to differentiate malignant from benign ascites (1 -3) , these findings are less spe­cific for differential diagnosis of peritoneal lesions (5 9).

The location of intraperitoneal fluid highly correlates with the volume, peritoneal pressure , position of the patient, region of origin , rate of fluid accum비 ation ，

presence of adhesions , density of the fluid , forces of gravity, negative subdiaphragmatic pressure , and per­itoneal reflections (2 , 4, 1 이 In our study , common sites of the fluid collection were subphrenic space, subhe­patic space, and paracolic gutter, and these results suggested that the locations of intraperitoneal fluid were mainly depended on patient’s position and grav­ity

There are some lesion specific characteristics in the locations of intraperitoneal fluid. The fluid of the ga­Ilbladder fossa was most commonly seen in hepatic diseases. Increased portal pressure might have cau­sed this fluid collection probably secondary to the ed­ema of the gallbladder brought about by disturbance of the drainage of cystic vein into the portal venous sys­tem (1 1). Tsujimoto et al. (12) reported that gallbladder wall thickening is a useful sign of benign ascites; patients with liver cirrhosis showed the gallbladder wall with >4 mm thickness. In infectious diseases, ab­normal fluid appeared to spare the subhepatic or subphrenic space probably due to the interference of natural fluid flow by.intraperitoneal adhesion or fibrous septation. In carcinomatosis , the common site of in­traperitoneal fluid may be subjected to increased inci­dence of metastasis , as tumor seeding depends on the natural flow of ascites within the peritoneal recesses (10 , 13). Buy et al. (14) found that the common sites of

939 -

Journal of the Korean Radiologica l Society 1995 : 32(6) : 937- 942

% 100

80

60

40

20

0 SH SP PG

딩 Hepatic disease

~ Carcinomatosls 디 Infection

MS GB

SH: Subhepatic space, SP: Subphrenic space, PG : Paracolic gutter, MS: Mesentery, GB: Gallbladder fossa

Fig . 3. Histogram of comparing hepatic disease, carcinoma tosis, and infectious disease. The intraperitoneal fluid was pres­ent in subhepatic space, subphrenic space, paracolic gutter, mesentery, and fossa of gallbladder in decreasing order of fre­quency. The fl ui d collection in su bphren ic and subhepatic space were less frequent in infectious diseases. Fluid in gallbladder fossa was most frequent in hepatic diseases

ovarian tumor deposition were the right subphrenic space, greater omentum , and pouch of Douglas.

Gore et al. (15) reported the fluid collection in the greater and lesser sacs to be due to malignant dis­eases , whereas fluid collection was primarily seen in the greater sac and not in the lesser omental bursae in the cases with benign transudates. However , fluid in the lesser sac is not considered a differential feature of malignant or benign ascites , because a wide range of diseases show fluid in the lesser sac (16). In our study , 21 percent of intraperitoneal lesions with fluid were associated with pleural effusion. Pleural effusion may reflect direct tumor invasion , pleural tuberculosis , or pleuroperitoneal communication. Peritoneal scintigra­phy following intraperitoneal administration of Tc-99m sulfur colloid may identify the source of the effusion ; if activity is demonstrated in the pleural space, this indicates the presence of a pleuroperitoneal communi­cation (17). A sudden pleural efffusion is also a com­mon sequelae of pulmonary embolism associated with maligancy (18)

The peritoneum , omentum , and mesentery are com­monly involved by various intraperitoneal diseases. Peritoneal thickening or enhancement are detected in peritoneal carcinomatosis, mesothelioma, and tuber­culosis (5 , 12). Fat infiltration in the omentum reflects invasion oftumor , edema, fibrosis , infectious or inflam­matory processes (7 , 8). In this study , peritoneal thick­ening was detected in the infectious diseases and car­cinomatosis, and omental fat infiltration was seen in carcinomatosis , hepatic disease, and infectious or in ­flammatory disease in decreasing order of frequency.

- 940

% 80

며 Hepatic dlsease ria Carcìn omatosis 口 Infectio n

60+--"-"" " •-- --_ ... ~-→1

40

20

PT PN OF ON MF MN PT: Peritoneal thickening, PN: Peritoneallymph nOde, OF: Omental lat intiltration, ON: Omental Iymph nOde, MF: Mesenteric ’at in’iltration, MN: Mesenteric lymph nOde

Fig. 4. Histogram shows percentage of fat infiltration and Iymph nodes in peritoneum, omentum, and mesentery. Omental fat infil­tration and Iymph nodes are most frequent in carcinomatosis ,

and mesenteric fat infiltration and Iymph nodes are most com. monly presented in infectious disease

Omental fat infiltration in liver cirrhosis seems to be due to the edema possibly caused by portal hyperten­sion , interfering omental vein drainage into the portal system via superior mesenteric and splenic veins Mesenteric fat infiltration is seen in metastatic dis­eases , inflammatory lesions, and mesenteric vascular diseases (8). In this study , mesenteric fat infiltration was noted most commonly in the infectious diseases, followed by carcinomatosis and hepatic diseases

Most peritoneal solid masses are secondary tumors in overwhelming majority (7). In this study , enlarged peritoneal Iymph nodes were present in carcinoma­tosis , but not in infectious disease. The classic omental cake due to replacement of the omental fat by tumor results in thick, confluent softtissue masses closely ad­herent to the anterior surface of the transeverse colon (8). Omental Iymph nodes in our cases were most prominent in carcinomatosis. Mesenteric Iymph nodes were more common in infectious disease than in carci­nomatosis. Hulnick et al. (19) reviewed 27 patients with abdominal tuberculosis and reported that there was a tendency for Iymphadenopathy to be prominent in pe­ripancreatic and mesenteric compartments. Lymph nodes in carcinomatosis were evenly distributed in the peritoneum , omentum , and mesentery, whereas in in­fectious diseases, they tended to be confined between mesenteries. Lymph nodes in peritoneum and omen­tum virtually ensured the diagnosis ofcarcinomatosis ; however , mesenteric Iymph nodes highly suggested infectous cause.

Although CT features are cons idered to have poten­tial for differentiating various diseases associated with intraperitoneal fluid , they are not sufficiently specific to obviate pathologic confirmation. However, the location of intraperitoneal fluid , and lesional involvement of peritoneum , omentum , and mesentery may be helpf비 for the differential diagnosis of intraperitoneallesions with ascites.

In conclusion , several findings regarding peritoneal

fluid collection appear to be helpful for specific diag­

nosis of intraperitoneal lesions. In hepatic diseases, fluid in gallbladder fossa was more frequent, and less

commonly involved peritoneum , omentum , and mes­

entery. In carcinomatosis , the fluid showed a diffuse

pattern in the peritoneal cavity. This contrasts with in­

fectious processes where fluid tended to spare sub­

phrenic and subhepatic spaces. Omental fat infitration

and Iymph node enlargement were frequently obser­

ved in carcinomatosis , whereas, mesentric fat infil­

tration and Iymph node enlargement were most com­

monly associated with infectious diseases

Mi Young Kim, et a/: Intraperitoneal Fluid Collection

and Inflmmatory Processes 01 the Peritoneum , Omentum and

Mesentery: Diagnosis with CT. RadioGraphics 1992 ; 12 : 1051-

1068

8. Gore RM , Levine MS, Lauler 1. Textbookofgastrointestinal radi­

ology. ln : Silverman PM , CooperC eds. Mesentery andOmental

lesions. Philadelphia: Saudeners, 1994: 2367-2381

9. Goerg C, Schwerk WB. Peritoneal Carcinomatosis with Ascites

AJR1991 ; 156 : 1185-1187

10. Proto AV, Lane EJ , Marangola JP. A new concept 01 ascitic fluid

distribution. AJR 1976; 126 : 974-980

11. Colli A, Cocciolo M, Buccino G, et al. Thickening 01 the Gall­

bladderWall in Ascites. JClin Ultrasound1991 ; 19 : 357-359

12. Tsujimoto F, Miyamoto Y, Tada S : Differentiation 01 Benign Irom

Malignant Ascites by Sonographic Eva’uation 01 Gallbladder

Wal l. Raiology 1985 ; 157 ‘ 503-504

13‘ Mayer MA. Dynamic rediology of the abdomen. 2nd ed. In

REFERENCES Mayer MA ed. Intraperitoneal spread of malignancies. New

York : Springer , 1982 : 55-104

1. Jolles H, Coulam CM. CT 01 ascites: Dillerential diagnosis. AJR 14. Buy JN , Moss AA, Ghossain MA, et al. Peritoneallmplants Irom

1980 ; 135: 315-322 Ovarian Tumors : CT Findings. Radiology 1988 ; 169 : 691-694

2. Gore RM , Levine MS, Lauler 1. Textbook 01 gastrointestinal radi- 15. Gore RM , Gallen PW, Filly RA : Lesser Sac Fluid in Predicting the

ology. In: Gore RM ed. Ascites and peritoneal fluid collection . Etiology 01 Ascites: CT Findings. AJR 1982; 139 ’ 71-74

Philadelphia: Saudeners , 1994: 2352-2366 16. Jellrey RB , Federle MP, Goodman PC. Computed Tomography

3. Seltzer SE. Ananlysis 01 the the tetjered-bowel sign on abdomi- olthe Lesser Peritoneal Sac. Radiology 1981 ; 141 : 117-122

nal CT as a predictor 01 malignant ascites. Gastrointest Radiol 17. Recoskie MJ, Picard D, Picard M, Carrier L, Chartrand R

1987 ;12 ‘ 245-249 Demosntration 01 Abnormal Peritoneal Communication in Pa-

4. Jellrey RB. CT demonstration 01 peritoneal implants. AJR 1980 ; tients with Ascites. Clin Nucl Med1990 ; 15: 97-100

135 : 323-326 18. Jacobson AF , Cerqueira MD , Breitz HB , Whitley MA, HiganoCS.

5. Walkey MM , Friedma n AC , Sohotra P, Radecki PD. CT Mani- Pleuraoperitoenal Communication Associated with Malignant

lestations 01 Peritoneal Carci nomatosis. AJR 1988 ; 150 : 1035- Ascites, A Potential Cause lor New Pleural Effusion Suggestive

1041 olPulmonary Embolism. ClinNucl Med 1990 ; 15 : 317-320

6. Nelson RC , Chezmar JL, Hoel MJ , Buck DR. Sugarbaker PH. Per- 19. Hulnick DH , Megibow AJ , Naidich DP , Hilton S, Cho KC ,

itoneal Carcinomatosis: Preoperative CT with Intraperitoneal Balth

7. Hamrick-Turner JE , Chiechi MV , Abbitt PL , Ros PR. Neoplastic

- 941 -

Journal of the Korean Radiological Society 1995 ; 32(6) : 937-942

대 한 방사 선 의 학회 지 1995; 32(6) : 937-942

복수를동반한복강내 병변:원언에 따른 CT소견1

1 인하대학교 의과대학 진단방사선과학교실

2전북대학교 의과대학 진단방사선과학교실

김미영 · 서창해·정원균 · 김종수2 • 최기철2

목 적 : 복수를 돔반한 다앙한 병변들의 감별 진 단은 임상적으로 중요한 의의가 있으며， 저자틀은 복부 CT 소견을 후향적

으로 분석함으로써 감별에 유용한 특징적 소견을 알아보고자 한다.

대상 및 방법 .복수가 있었던 170 예의 복강내 병변을 간성 질환， 악성종양 전 01 (carcinomatosis) , 염증성 질환르로 대별하

였다. CT 상 각 병변에 따른 복수의 분포 위치를 비교하였고， 복막 비후， 대망과 장간막의 지 방침습， 복막， 대망， 장간막에 돔

반된 임파절 비대 소견을 분석하였다.

결 과 : 복수의 분포를 분석한 결과， 간하， 횡격막하의 빈도가 가장 높았으며， 대장주위구， 장간막， 담낭와의 빈도순으로 나

타났다. 담낭와 복수는 간성 질환에서 빈도가 가장 높았으며， 간하와 횡격막하의 복수는 염증성 병변에서 상대적으로 낮은

빈도를 보였다. 복막 비후는 염증성 병변과 악성종양 전이에서 동반되었으며， 대망의 지방조직 침습과 임파절 비대는 악성

종앙전이에서 가장 높은 빈도를 보인 반면 (각각 58%, 44%), 장간막의 지방조직 침습과 임파절 비대는 염증성 병변에서 가

장 높게 나타났다 (각각 61 %, 26%).

결 론:복수의 분포는 병변에 따른 위치별 특성이 있었으며， CT상 복막， 대망， 장간막의 지방조직 징습과 임파절 비대 소

견은 악성종양 전이와， 염증성 병변의 감별진 단에 유용한 것으로 사료된다.

- 942 -