Share this article on:

CT Differentiation of Female Peritoneal Tuberculosis and Peritoneal Carcinomatosis From Normal-Sized Ovarian Cancer

Shim, Sang Woo MD; Shin, Shang Hun MD; Kwon, Woon Jung MD; Jeong, Yoong Ki MD; Lee, Jong Hwa MD

Journal of Computer Assisted Tomography: January 2017 - Volume 41 - Issue 1 - p 32–38
doi: 10.1097/RCT.0000000000000446
Abdominal Imaging

Objective: The purpose of this study was to analyze the computed tomography (CT) features of female peritoneal tuberculosis and peritoneal carcinomatosis from normal-sized ovarian cancer for their differentiation.

Materials and Methods: We analyzed the CT features of 18 female peritoneal tuberculosis and 17 peritoneal carcinomatosis with proven normal-sized ovarian carcinomas. Omental change, mesenteric change, parietal peritoneal thickening, lymph node enlargement, ascites, ovarian CT attenuation, and ovarian capsular change were analyzed.

Results: Heterogeneous parenchymal hyperattenuation and capsular change of the ovary were more frequently seen in cases of peritoneal carcinomatosis than in cases of female peritoneal tuberculosis (P = 0.002, P < 0.001, respectively). There were no statistically significant differences in the other CT features.

Conclusions: The differentiation of female peritoneal tuberculosis and peritoneal carcinomatosis with normal-sized ovarian cancer by CT may be a diagnostic challenge. Ovarian hyperattenuation and any prominent ovarian capsular change may facilitate the differentiation between these groups.

From the Department of Radiology, Ulsan University Hospital, University of Ulsan College of Medicine, Ulsan, Republic of Korea.

Received for publication January 18, 2016; accepted March 29, 2016.

Correspondence to: Jong Hwa Lee, MD, Department of Radiology, Ulsan University Hospital, University of Ulsan College of Medicine, 877 Bangeojinsunhwando-ro, Dong-gu, Ulsan 682-714, Republic of Korea (e-mail: jhlee@uuh.ulsan.kr).

The authors declare no conflict of interest.

This is an open-access article distributed under the terms of the Creative Commons Attribution-Non Commercial-No Derivatives License 4.0 (CCBY-NC-ND), where it is permissible to download and share the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Tuberculosis is a curable infectious disease caused by Mycobacterium tuberculosis. In recent years, this disease has emerged as a major public health problem in developed and underdeveloped countries. The incidence of tuberculosis is increasing due to poor socioeconomic conditions, the spread of HIV infection, the development of multi-drug resistance, and poor case findings.1,2 Although tuberculous peritonitis can be diagnosed clinically and radiologically in some instances, it can be difficult in patients who lack typical symptoms and laboratory data or in patients with normal chest films; thus, tuberculous peritonitis often mimics peritoneal carcinomatosis.3

Because both peritoneal tuberculosis and peritoneal metastasis may have similar clinical features in female patients, determining the correct preoperative diagnosis may be difficult. Pelvic peritoneal tuberculosis classically presents with dull abdominopelvic pain, menstrual disturbances of infertility, and constitutional symptoms such as low-grade fever, loss of appetite, loss of weight, and lethargy. It can clinically mimic an advanced abdominopelvic malignancy in females.3,4 Pelvic pain and mass, ascites, and increased CA-125 levels are well-known markers of both ovarian cancer and peritoneal tuberculosis.1,5–9

Extensive overlap between the known CT features of diffuse infiltration of the peritoneum, omentum, and mesentery make distinguishing between peritoneal tuberculosis and peritoneal carcinomatosis difficult or impossible.10

Previous reports have noted that since CT features and tumor marker such as CA-125 were not crucial in differential diagnosis, the final diagnosis should be performed through percutaneous or laparoscopic biopsy.11,12

Radiologically, if an ovary is significantly enlarged with a definite tumor, significant peritoneal infiltrations indicate peritoneal metastasis. However, when the ovary is normal-sized in any patient with peritoneal fat infiltration, there may be a difficulty in differential diagnosis between benign or malignant peritonitis, particularly in regions where tuberculosis is endemic. Early correct diagnosis is important for early treatment; a delayed diagnosis can lead to inappropriate treatment and increased morbidity and mortality.

We attempted to investigate the significance of various CT features related to peritoneal infiltrations and other findings associated with female peritoneal tuberculosis and peritoneal carcinomatosis from normal-sized ovarian cancer for their differential diagnosis.

Our study was performed to analyze the spectrum of CT findings from two groups of female peritoneal tuberculosis and peritoneal carcinomatosis with normal-sized ovarian cancer, and to statistically analyze the frequencies of each CT feature in the two patient groups for differential diagnosis.

Back to Top | Article Outline

STUDY SUBJECTS

This was a retrospective study approved by our institutional review board. Informed consent for review of the patient’s medical record and images was waived. We performed an electronic search of the databases for information obtained between January 2004 and January 2014 to identify pathologically or medicoclinically confirmed cases of female peritoneal tuberculosis, and ovarian cancer with peritoneal metastasis.

The inclusion criteria for enrollment of this study were pathological confirmation of peritoneal tuberculosis by percutaneous omental biopsy, or clinical improvement after medical treatment without pathological confirmation. Those of patients with normal-sized ovarian cancer with peritoneal carcinomatosis were surgicopathologic confirmation of primary ovarian cancer, initial interpretation of normal-sized ovaries less than 4 cm in length, omental and/or mesenteric infiltrations, and exclusion of any case with rim-enhancing lymphadenopathy with central necrosis on MDCT.

We identified a total of 144 patients with radiologic suspicions of the above two disease entities after a search of PACS data. And we carefully analyzed their medical records and images, and excluded 90 cases of peritoneal metastasis with ovarian cancer overt larger than 4 cm in size or inseparable from adjacent thickened peritoneum, as well as 19 cases of suspicious tuberculous peritonitis lacking clinical follow-up or histopathologic evidence. Therefore, the final sample in this study included 18 patients with female tuberculous peritonitis and 17 patients with normal-sized ovarian cancer with peritoneal metastasis.

Back to Top | Article Outline

IMAGING TECHNIQUES

Abdominopelvic CT examinations were performed with multidetector CT scanners including Siemens Sensation 16 and Sensation AS (Siemens Healthcare) as well as Philips Brilliance 256 iCT (Philips Healthcare). The scanning parameters included 100 kV (peak) and automatic tube current modulation. Images were typically acquired after oral administration of water (800 mL) and intravenous administration of nonionic contrast material (iopromide, Ultravist 300 or 340 mL/I/mL; Bayer Healthcare, with the total volume <120 mL, 3 mL/s) in the arterial and portal venous phases. The images were displayed in the axial plane with a slice thickness of 3 mm and reconstructed at 3-mm intervals, and in the coronal plane with a slice thickness of 3 mm and reconstructed at 3-mm intervals.

Back to Top | Article Outline

IMAGING ANALYSIS

The MDCT images were independently interpreted by two attending abdominal imaging radiologists with 20 and 21 years of experience. The radiologists were blinded to the patient demographics as well as to the clinical findings and the histopathologic results during the image interpretation. All images, including coronal multiplanar reformatted images, were reviewed on a picture archiving and communication system workstation system (Infinite Healthcare System). Reader disagreements, if any, were resolved during consensus meetings with reevaluation of the images together.

The analyzed CT features included omental change, mesenteric change, parietal peritoneal change, lymph node enlargement, ascites, ovarian capsular change, and ovarian parenchymal attenuation. Each analyzed CT feature related to the greater omentum, mesentery parietal peritoneum, and ovarian capsule and ovarian parenchymal attenuation was graded according to the severity using the following 4-point scale: 1, negative; 2, mild degree; 3, moderate degree; and 4, severe degree. Omental changes included nodular or smudged (infiltration with ill-defined) or cake-like (lobulated soft-tissue replacement) lesions based on the morphologic appearance. Mesenteric changes included nodular infiltrates, thickening of the mesenteric leaves, and loss of normal mesenteric configuration.10 Parietal peritoneal changes included smooth, lobulated, nodular thickenings. Lymph node enlargement larger than 10 mm in short-axis diameter was graded as negative or positive. Significant presence of ascites was graded as negative or positive. Ovarian capsular changes include smooth, nodular, and lobulated margins. In grading the scores for omental change, parietal peritoneal change, mesenteric change, ascites, and ovarian capsular change, a score 0 and 1 was regarded as negative and scores of 2 and 3 was regarded as positive. Lymph node enlargement was divided into negative and positive groups. Ovarian parenchymal attenuation was measured in the ROI with the largest parenchyma, excluding cyst or vessels as much as possible. Ovarian attenuation was acquired as the mean value of ROI of both ovarian parenchyma (Figs. 1–5).

Qualitative and quantitative differences in the CT findings between the two groups were statistically analyzed with Fisher exact and Student t-test using a software program (SAS Institute, Cary, NC). Fisher exact test was used to determine statistical differences in omental change, mesenteric change, parietal peritoneal change, lymph node enlargement, ascites, and ovarian capsular change. A Student t-test was used to determine differences in ovarian parenchymal attenuation. A P value of less than 0.05 was considered statistically significant.

Back to Top | Article Outline

RESULTS

Table 1 presents the estimated grading scores for each of the CT features in peritoneal tuberculosis and peritoneal carcinomatosis groups, respectively.

Omental abnormalities were noted in 14 patients (77.8%) with female tuberculous peritonitis group and in 15 patients (88.2%) with peritoneal carcinomatosis group. There was no significant difference between the two groups regarding the severity of the omental changes (P = 0.658).

Parietal peritoneal abnormalities were noted in 5 patients (27.8%) with female tuberculous peritonitis group and in 6 patients (35.3%) with peritoneal carcinomatosis group. There was no significant difference between the two groups regarding the severity of parietal peritoneal changes (P = 0.725).

Mesenteric abnormalities were noted in 13 patients (72.2%) with female tuberculous peritonitis group and in 8 patients (47.1%) with peritoneal carcinomatosis group. There was no significant difference between the two groups regarding the severity of mesenteric changes (P = 0.176). The mesenteric changes noted on CT were thickenings of the mesenteric leaflets with visceral peritoneal thickenings, heterogeneous smudged or soft tissue infiltrative lesions, or micronodular or macronodular lesions.

Ovarian capsular changes were noted in 1 patient (5.6%) with tuberculous peritonitis group and in 13 patients (76.5%) with peritoneal carcinomatosis group, and this difference was significant between the groups (P < 0.001).

Ovarian parenchymal attenuation ranged from 33 to 89 HU (mean = 57) in patients with female tuberculous peritonitis group and from 39 to 115 HU (mean = 76) in patients with peritoneal carcinomatosis group. Ovarian parenchymal attenuation in tuberculous peritonitis group was significantly lower in peritoneal carcinomatosis group (P = 0.002) (Table 2).

Back to Top | Article Outline

DISCUSSION

Abdominopelvic CT is often the initial imaging examination performed in patients with a wide variety of clinical symptoms. CT substantially aids in the identification of peritoneal disease. Proper recognition of the spectrum of CT findings of peritoneal disease allows identifications of disease extending into various sites of the peritoneum. The radiologist’s precise localization of all affected sites and the accurate description of the disease severity provide important guidance to clinicians for treatment.

Diseases of the peritoneum are common, and given the considerable overlap in their imaging appearances and the resultant potential for misinterpretation, they can present a diagnostic challenge. The dissemination of a primary or secondary peritoneal malignancy may result in a condition known as peritoneal carcinomatosis. The imaging manifestations of this condition may resemble those seen in the presence of various benign conditions such as peritoneal thickening due to peritoneal tuberculosis, active Crohn’s disease, diverticulitis, appendicitis, omental infraction, and severe pancreatitis, making differential diagnosis difficult. Accurate differentiation between these conditions is of crucial importance for selecting the most appropriate method of treatment.13

The abdomen is the most common focus of extrapulmonary tuberculosis. Peritonitis is the most common clinical manifestation of abdominal tuberculosis, affecting one-third of all patients.14 The incidence of peritoneal tuberculosis has declined, but still this disease remains a persistent problem in endemic areas or in immunocompromised patients.10 Peritoneal tuberculosis is thought to originate primarily from hematogenous spread; however, it may be secondary to a ruptured lymph node or gastrointestinal deposit or to fallopian tube involvement.14 The CT findings of peritoneal tuberculosis with varying degrees of omental and mesenteric involvements, ascites, and peritoneal thickening are nonspecific. If only diffuse infiltrations of the peritoneum, omentum, or mesentery are observed on CT, distinguishing between peritoneal tuberculosis and peritoneal carcinomatosis just with imaging findings related to peritoneal abnormalities is difficult or impossible because of the overlap of the imaging findings. Therefore, identification of the presence or absence of primary or secondary malignancy is crucial to determine the final diagnosis about the any peritoneal disease on CT.

There have been many reports dealing with peritoneal tuberculosis mimicking peritoneal carcinomatosis or advanced ovarian cancer. The presence of ascites and high serum level of CA-125 do not necessarily indicate that the clinical picture is malignant in reproductive women. CA-125 has a limited diagnostic value. The serum CA-125 level is found to be elevated in up to 82% of women with late-stage epithelial ovarian cancer. However, it may also be elevated in several benign pelvic pathologies such as pelvic inflammatory disease, uterine fibroids, endometriosis, and tuberculous peritonitis.11,12,15 The possibility of peritoneal tuberculosis should be considered in the differential diagnosis of patients with ascites and high serum CA-125 level. In one study, the sensitivity of AFB culture from ascites was only 12.8%.15

The CT features of female tuberculous peritonitis and peritoneal carcinomatosis are considerably similar in endemic areas, making differential diagnosis difficult. It is impossible to correctly differentiate these two diseases using clinical features, and hematologic as well as radiologic examinations without surgicopathological intervention. The majority of cases with peritoneal tuberculosis can be diagnosed by laparoscopic exploration with the use of frozen sections in conjunction with clinical features.16

Ha et al noted that the prevalence of macronodules in the mesentery was uncommon in peritoneal carcinomatosis.10 The common occurrence of mesenteric micro- or macronodules in tuberculous peritonitis appears to be because tuberculous peritonitis develops from the rupture of mesenteric lymph nodes seeded by the hematogenous or lymphatic routes from the primary lesion sites or by direct spread from the serosa by continuity with adjacent glands or structures. In our study, 50% of cases with peritoneal metastasis with normal-sized ovarian cancer revealed macronodules in mesentery, and there was no statistically significant difference in the prevalence of macronodules between the two groups. This result may indicate that the mechanism of peritoneal metastasis in any case of ovarian cancer would be direct spread through the tumor capsule of the ovary. Therefore, the prevalence of omental and parietal peritoneal abnormalities as well as mesenteric changes would not be significantly different between the groups due to their similar possible mechanism of disease development.

We have also focused on the ovarian morphology itself in the CT image analysis for differentiation between female peritoneal tuberculosis and peritoneal carcinomatosis with normal-sized ovarian cancer. There have been many previous reports dealing with peritoneal tuberculosis mimicking ovarian cancer. However, relatively few studies have presented the available CT images,2,12,17–19 and only one case report described the normal-sized both ovarian complex lesions with extensive omental and peritoneal stranding on CT scan.2

The definition of a normal-sized ovary would be difficult to determine because the normal ovary may have different shapes or sizes due to the presence or absence of ovulation-associated cysts. As a control group, we evaluated the size and attenuation of the ovaries according to the each decade of age composed of 20 women who have checked MDCT because of any purpose about other body parts. The mean largest size of normal ovaries without cysts was 3.2 cm in the third decade, 3.5 cm in the fourth, 3.2 cm in the fifth, 2.8 cm in the sixth, 2.1 cm in the seventh, and 2.1 cm in the eighth. Therefore, we considered the normal-sized ovary in the absence of cysts to be approximately 4 cm in length. The mean attenuation of the intact ovarian parenchyma in the absence of cysts was mean 55.5 HU in the third decade, 54.1 HU in the fourth, 57.3 HU in the fifth, 58.1 HU in the sixth, 61 HU in the seventh, and 63.3 HU in the eighth. The parenchymal attenuation tended to increase progressively, probably owing to the loss of follicles and stromal tissue, in addition to parenchymal atrophy.

The ovarian morphology may differ according to a patient’s age, menopause state, the menstrual cycle, and the presence or absence of follicles or cysts. We expected that the morphology of the ovarian capsule and the parenchymal attenuation would be most important in image analysis for significant differentiation between normal and abnormal ovaries even though various morphologic differences according to the various conditions may be present. Ovarian capsular changes were anticipated in both peritoneal tuberculosis and peritoneal carcinomatosis. However, our study revealed that the more significant ovarian capsular changes occurred in normal-sized ovarian cancer patients with peritoneal carcinomatosis than in female peritoneal tuberculosis patients. The ovarian parenchymal attenuation (mean 58 HU) in the control group was significantly lower than in peritoneal carcinomatosis group with normal-sized ovarian cancer (mean 76 HU) (P = 0.002). The ovarian parenchymal attenuation (mean 56 HU) in the female peritoneal tuberculosis group was significantly lower than in peritoneal carcinomatosis group with normal-sized ovarian cancer (mean 76 HU) (P = 0.002). The enhanced increased parenchymal attenuation appears to be an important CT feature in normal-sized ovarian cancer.

Among our study cases with peritoneal carcinomatosis with normal-sized ovarian cancer, the pathologic diagnosis was 15 serous papillary cystadenocarcinomas and 2 poorly differentiated adenocarcinomas. Serous papillary cystadenocarcinoma is known to account for the largest proportion of malignant ovarian tumors, representing 60 to 80% of all malignant epithelial ovarian tumors. Various macroscopic findings of mixed solid and cystic components are known CT features.20 However, there have been no reports dealing with minor CT features in normal-sized ovarian carcinoma such as capsular changes or parenchymal attenuation. We believe that normal-sized ovarian malignancy is a real CT feature of serous papillary cystadenocarcinomas of the ovary.

Our study has several limitations. First, the possibility of a selection bias should be considered because this study was a retrospective study. Our cases with female peritoneal tuberculosis did not have advanced peritoneal thickenings conglomerating or enlarging the adnexa. We believe that severe peritoneal thickenings encasing a normal-sized ovary could prevent the correct ovarian capsular analysis. We excluded cases with any size of ovarian cancer inseparable from margin of adjacent peritoneal thickening in this study. Second, less-experienced radiologists would require quite some learning time to understand the capsular change and parenchymal attenuation of the normal-sized ovary. Third, because the images were analyzed retrospectively by two experienced radiologists with consensus, we did not assess the interobserver variability. Fourth, conventional CT would be of limited use for the correct analysis of the detail of the capsular and parenchymal abnormalities because of its lower anatomical resolution and internal parenchyma conspicuity. Fifth, we did not consider the variation of the normal ovarian size or CT attenuation according to the decades of patient’s ages.

In conclusion, we presented various CT features for the differential diagnosis between female peritoneal tuberculosis and peritoneal carcinomatosis with normal-sized ovarian cancer. To our knowledge, no similar studies have been previously reported. In this study, there were high percentages of significant positive findings of omental and mesenteric abnormalities in each disease group, with no significant differences between two groups. There were significantly high percentages of significant ovarian capsular change and higher ovarian parenchymal attenuation only in the peritoneal carcinomatosis group. We consider the value of the significant ovarian capsular change and higher ovarian parenchymal attenuation to be useful for differentiation between female peritoneal tuberculosis and peritoneal carcinomatosis with normal-sized ovarian cancer.

Back to Top | Article Outline

REFERENCES

1. Chandir S, Hussain H, Salahuddin N, et al. Extrapulmonary tuberculosis: a retrospective review of 194 cases at a tertiary care hospital in Karachi, Pakistan. J Pak Med Assoc. 2010;60:105–109.
2. Imtiaz S, Siddiqui N, Ahmad M, et al. Pelvic-peritoneal tuberculosis mimicking ovarian cancer. J Coll Physicians Surg Pak. 2012;22:113–115.
3. Nance FC. Disease of the peritoneum, retroperitonum, mesentery, and omentum. In: Haubrich WS, Schaffner F, Berk JE, eds. Bockus gastroenterology. 5th ed. Philadelphia: Saunders; 1995:3061–3096.
4. Jana N, Mukhopadhyay S, Dhali GK. Pelvic tuberculosis with elevated serum CA125: a diagnostic dilemma. J Obstet Gynaecol. 2007;27:217–218.
5. Vagenas K, Stratis C, Spyropoulos C, et al. Peritoneal carcinomatosis versus peritoneal tuberculosis: a rare diagnostic dilemma in ovarian masses. Cancer Therapy. 2005;3:489–494.
6. Geoffrey ER, Andrew JL. Peritoneal tuberculosis mimicking advanced-stage epithelial ovarian cancer. Obstet Gynecol. 2007;110:1417–1419.
7. Bilgin T, Karabay A, Dolar E, et al. Peritoneal tuberculosis with pelvic abdominal mass, ascites and elevated CA 125 mimicking advanced ovarian carcinoma: a series of 10 cases. Int J Gynaecol Oncol. 2001;11:290–294.
8. Piura B, Rabinovich A, Leron E, et al. Peritoneal tuberculosis mimicking ovarian carcinoma with ascites and elevated serum CA-125: case report and review of literature. Eur J Gynaecol Oncol. 2002;23:120–122.
9. Nistal de Paz F, Herrero Fernández B, Pérez Simón R, et al. Pelvic-peritoneal tuberculosis simulating ovarian carcinoma: report of three cases with elevation of the CA 125. Am J Gastroenterol. 1996;91:1660–1661.
10. Ha HK, Jung JI, Lee MS, et al. CT differentiation of tuberculous peritonitis and peritoneal carcinomatosis. AJR. 1996;167:743–748.
11. Flores-Alvarez E, Tello-Brand SE, López-López F, et al. Peritoneal tuberculosis. Report of seven cases. Cir Ciruj. 2010;78:67–71.
12. Gosein MA, Narinesingh D, Narayansingh GV, et al. Peritoneal tuberculosis mimicking advanced ovarian carcinoma: an important differential diagnosis to consider. BMC Res Notes. 2013;6:88–91.
13. Marques DT, Tenório de Brito Siqueira L, Franca Bezerra RO, et al. Resident and fellow education feature: imaging evaluation of peritoneal disease: overview of anatomy and differential diagnosis. Radiographics. 2014;34:962–963.
14. Burrill J, Williams CJ, Bain G, et al. Tuberculosis: a radiologic review. Radiographics. 2007;27:1255–1273.
15. Bae SY, Lee JH, Park JY, et al. Clinical significance of serum CA-125 in Korean females with ascites. Yonsei Med J. 2013;54:1241–1247.
16. Patel SM, Lahamge KK, Desai AD, et al. Ovarian carcinoma or abdominal tuberculosis?—a diagnostic dilemma: study of fifteen cases. J Obstet Gynaecol India. 2012;62:176–178.
17. Nasiri R, Yazdi M, Naghibi S, et al. Pelvic peritoneal tuberculosis mimicking stage IV ovarian cancer: case report. J Family Reprod Health. 2008;2:163–166.
18. Zamłyński J, Olejek A, Oleś E, et al. Peritoneal tuberculosis mimicking ovarian cancer—diagnostic difficulties. Two cases reports. Pneumonol Alergol Pol. 2009;77:422–428.
19. Tan O, Luchansky E, Rosenman S, et al. Peritoneal tuberculosis with elevated serum Ca-125 level mimicking advanced stage ovarian cancer: a case report. Arch Gynecol Obstet. 2009;280:333–335.
20. Kawamoto S, Urban BA, Fishman EK. CT of epithelial ovarian tumors. Radiographics. 1999;19:S85–S102.
Keywords:

peritoneal tuberculosis; peritoneal carcinomatosis from normal-sized ovarian cancer; CT differentiation

Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.