The role of ¹⁸F-FDG PET/CT in patients with serous cavity effusion of undetermined origin: a retrospective clinical study

Introduction

The serous cavity encompasses the chest, abdominal, and pericardial cavities. Under normal circumstances, a minimal quantity of fluid resides in the serous cavity, serving to lubricate and shield organs and tissues. Pathological conditions such as inflammation, trauma, tumors, autoimmune, and metabolic diseases can lead to an accumulation of fluid in the serous cavity, termed serous cavity effusion (SCE) (Rossi et al., 2015; Gecmen et al., 2018). Based on the underlying cause, SCE can be benign or malignant. Common benign causes include cardiac insufficiency, decompensation of liver cirrhosis, non-specific inflammation, renal failure, tuberculosis, and similar conditions (Lee et al., 2014). Malignant serous cavity effusion arises in the advanced stages of various malignancies due to tumor invasion of blood vessels, lymphatic vessels, or lymph node metastasis, leading to the blockage of serous effusion absorption or direct tumor cell invasion of blood vessels, enhancing capillary permeability (Porcel, 2018). While puncturing the serous cavity for effusion extraction and cytological examination is the gold standard for determining the nature of SCE, its sensitivity is only 30%–50%, and it poses certain risks as an invasive procedure. Traditional imaging methods, including ultrasound, computed tomography (CT), and magnetic resonance imaging (MRI), can identify the presence and estimate the amount of serosal effusion. They aid in differentiating between benign and malignant diseases of the chest and abdominal cavity by examining signs of benign and malignant effusion. However, ultrasonography, CT, and MRI primarily focus on local anatomical structures, making it challenging to identify small lesions and hidden metastases, limiting their ability to differentiate and diagnose the nature of serosal effusion (Sugiyama et al., 2011). Positron emission tomography (PET)/CT imaging, combining anatomical CT with functional PET, has played a pivotal role in diagnosing, staging, and assessing tumor prognosis post-treatment. It is increasingly valued for identifying and diagnosing the nature of serosal effusion (Zhang et al., 2009; Li et al., 2013; Porcel, 2018). This study aims to investigate the utility of positron emission tomography with fluoro-18 fluorodeoxyglucose integrated with computed tomography (¹⁸F-FDG PET/CT) in localizing and qualitatively diagnosing primary lesions in unexplained serous cavity effusion, as well as assessing the diagnostic performance for serous cavity metastasis in malignant tumors.

Material and methods

This retrospective study received approval from the Ethics Committee of Affiliated Hospital of Zunyi Medical University (approval number: KLL-2022-166), and all methods were performed in accordance with the relevant guidelines and regulations.

Results

Detection of primary lesions

Among the 134 patients with unexplained serous cavity effusion, 98 cases were abnormal. Of the 98 abnormal scans, 91 were eventually diagnosed with malignant lesions. For the remaining 36 scans, 26 benign lesions were accurately identified, and 10 malignant lesions were missed by PET/CT. The positive predictive rate (PPV) was 92.9%, and the negative predictive rate (NPV) was 72.2%. Typical true positive and true negative cases are demonstrated in (Figs. S1 and S2). False-negative results were pathologically confirmed in 10 cases, including two cases of unknown origin (as shown in Fig. 1), one case of pericardial mesothelioma, one case of pleural mesothelioma, one case of lung cancer, two cases of advanced adenocarcinoma of the fallopian tubes, one case of liver cancer, one case of gastric mucosa-associated lymphoid tissue (MALT) lymphoma, and one case of gallbladder signet ring cell carcinoma. Seven cases were confirmed falsely positive after surgery or biopsy, including two cases of ovarian theca fibroma showing focal increased ¹⁸F-FDG uptake on PET/CT, misdiagnosed as ovarian cancer, two cases of organized pneumonia showing strongly increased ¹⁸F-FDG uptake, misdiagnosed as lung cancer, one case of tuberculous pericardium misdiagnosed as malignant mesothelioma (as shown in Fig. 2), one case of tuberculous peritonitis misdiagnosed as ovarian cancer with peritoneal metastasis (as shown in Fig. 3), and one case of Castleman’s disease, misdiagnosed as lymphoma. For the primary cause of pericardial effusion, PET/CT demonstrated sensitivity, specificity, and accuracy of 87.5%, 84.6%, and 85.7%, respectively. In cases of pleural effusion, the sensitivity, specificity, and accuracy were 91.4%, 86.4%, and 90.2%, respectively. For ascites, PET/CT exhibited sensitivity, specificity, and accuracy of 88.4%, 80.0%, and 86.8%, respectively. Overall, the sensitivity, specificity, and accuracy of PET/CT in diagnosing the etiology of serous cavity effusion were 90.1%, 78.8%, and 85.7%, respectively (refer to Table 1). The SUVmax of primary malignant lesions was significantly higher than that of benign lesions (12.83 ± 6.64 vs. 4.48 ± 3.16, P < 0.001) (see Fig. 4).

Table 1:

Clinical and epidemiologic features of patients with serous cavity effusion of undetermined origin.

Parameters	Pericardial effusion (N = 21)a	Pleural effusion (N = 92)b	Ascites (N = 53)c	Total (N = 134)
Gender
Man	12 (57.1%)	51 (55.4%)	11 (20.8%)	60 (44.8%)
Woman	9 (42.9%)	41 (44.6%)	42 (79.2%)	74 (55.2%)
Age (years)	61.9 ± 13.6	59.4 ± 14.5	56.5 ± 12.7	58.8 ± 13.4
Primary lesions
Tumor lesions	8 (38.1%)	70 (70.6%)	43 (81.1%)	101 (75.4%)
Malignant tumors	6 (28.6%); including mesothelioma (N = 2), lung cancer (N = 2), liver cancer (N = 1), lymphoma (N = 1)	64 (65.2%); including lung cancer (N = 38), lymphoma (N = 9), ovarian cancer (N = 9), mesothelioma (N = 2), thymic carcinoma (N = 2), prostate cancer (N = 1), rectal cancer (N = 1), unknown origin (N = 2).	40 (75.5%), including ovarian cancer (N = 25), lymphoma (N = 5), adenocarcinoma of unknown origin (N = 2), intestinal cancer (N = 2), pancreatic cancer (N = 1), gallbladder carcinoma (N = 2), mesothelioma (N = 1), liver caner (N = 2)	94 (70.1%)
Benign tumors	2 (9.5%); including thymoma type A (N = 1) and theca fibroma (N = 1)	6 (5.4%); including ovarian sex gonad stromal tumor (N = 4), thymoma type A (N = 2).	3 (5.7%); ovarian sex gonad stromal tumor (N = 3)	7 (5.2)
Non-tumorous lesions	13 (61.9%)	22 (23.9%)	10 (18.9%)	33 (24.6%)
Nonspecific inflammationd	6 (28.6%)	11 (12.0%)	1 (1.9%)	13 (9.7%)
Tuberculosis	2 (9.5%)	5 (5.4%)	4 (7.5%)	5 (3.7)
Others	5 (23.8%); including dilated cardiomyopathy (N = 2); hypothyroidism (N = 1); heart failure (N = 2)	6 (6.5%); including hypothyroidism (N = 1), heart failure (N = 3), liver cirrhosis (N = 2).	5 (9.4%); including liver cirrhosis (N = 4), Castleman’s disease (N = 1).	15 (11.2%)

DOI: 10.7717/peerj.19495/table-1

Notes:

aIncluding 13 cases of pericardial effusion combined with pleural effusion and 1 case of pericardial effusion combined with pleural effusion and ascites.

bIncluding 13 cases of pleural effusion combined with pericardial effusion, 17 cases of pleural effusion combined with ascites and 1 case of pericardial effusion combined with pleural effusion and ascites.

cIncluding 17 cases of ascites combined with pleural effusion, and 1 case of ascites combined with pericardial effusion and pleural effusion.

dIncluding three patients with pericardial effusion, five with pleural effusion, and one with both pericardial and pleural effusion, all underwent re-examination after anti-inflammatory treatment. The subsequent follow-up showed that the lesions had disappeared, and no recurrence was observed for more than half a year. T and PET/CT in Detection of Primary Lesion.

Discussion

Both benign and malignant conditions can lead to serous cavity effusion. In this study, malignant tumors causing pericardial effusion included mesothelioma, lymphoma, while benign conditions comprised heart failure, tuberculous pericarditis, and type A thymoma. Malignant tumors associated with pleural effusion included lung cancer, ovarian cancer, and lymphoma, with benign conditions primarily featuring tuberculous pleurisy and non-specific inflammation. The leading causes of ascites involved ovarian cancer, gastrointestinal and biliary cancers, whereas benign conditions encompassed tuberculous peritonitis, non-specific inflammation, and ovarian sex cord stromal tumors.

Routine examination methods for serous cavity effusion include biochemical analysis, cytological examination, and traditional imaging techniques such as X-ray, B-ultrasound, and CT. However, the positive rate of cytological examination for serous cavity effusion is low, and the sensitivity and specificity of tumor markers in blood and serous cavity effusion are limited (Naito, Shimada & Yuki, 2022). X-ray, ultrasound, and CT have a restricted role in the differential diagnosis of diseases affecting the pericardium, pleura, and peritoneum. Some patients exhibit atypical medical histories and clinical manifestations, making it challenging to determine the etiology of serous effusion through various routine examinations (Strange et al., 2023). Malignant tumor cells demonstrate an increased rate of glycolysis, a characteristic reflected in strong ¹⁸F-FDG uptake on PET/CT. Serous cavity effusion is a multisystem disorder resulting from various conditions throughout the body. PET/CT can provide whole-body imaging, allowing for the simultaneous localization of primary and metastatic lesions, showcasing high sensitivity in various malignant tumors (Zhang et al., 2009; Chen et al., 2016). Consequently, PET/CT serves as a suitable imaging modality for diagnosing unexplained serous cavity effusion.

For clinical doctors and patients, early diagnosis of the primary cause of serous cavity effusion is crucial for formulating correct clinical treatment plans and predicting prognosis. While a few previous studies have discussed the role of PET/CT in detecting the primary causes of pleural and ascites, revealing good diagnostic performance (Chen et al., 2016; Yang et al., 2019; Lu et al., 2022; Simó et al., 2023), our study is the first to systematically evaluate the role of PET/CT in the diagnosis of unexplained serous cavity effusion and further assess its diagnostic performance in pericardial effusion, pleural effusion, and peritoneal effusion. Overall, our study revealed that PET/CT in locating the primary disease site of serous cavity effusion had high diagnostic performance, with a sensitivity of 90.1%, specificity of 78.8%, and accuracy of 87.3%. Further analysis based on the location of the effusion showed that the sensitivity, specificity, and accuracy of PET/CT in determining the primary cause of pericardial effusion were 87.5%, 84.6%, and 85.7%, respectively. It is noteworthy that PET/CT correctly identified a patient with pericardial effusion induced by ovarian thecoma, who was also accompanied by pleural and ascitic fluid, known as Meigs syndrome. After undergoing surgery to remove the tumor, all serous cavity effusions, including pericardial effusion, disappeared, confirming our diagnosis. For determining the primary cause of pleural effusion, the sensitivity, specificity, and accuracy of PET/CT were 91.4%, 86.4%, and 90.2%, respectively, similar to the results of Yang et al. (2019). Regarding ascites, PET/CT demonstrated a sensitivity of 88.4%, specificity of 80.0%, and accuracy of 86.8% in determining the primary cause, higher than previously reported (Kolesnikov-Gauthier et al., 2005; Zhang et al., 2009; Han et al., 2018). The positive interpretation of lesions may be the reason for the high accuracy of our research results. In previous studies, only SUVmax and SUVmax combined with serum Ca125 were used as parameters for positive interpretation. In addition to using SUVmax as a positive reference, our study also included anatomical abnormalities only on CT as possible positive lesions.

Due to only two patients having metastasized to the pericardium, no further subgroup analysis was conducted. Previous studies have reported that the sensitivity of ¹⁸F-FDG PET/CT in diagnosing pleural and peritoneal metastasis was 84.4% and 86.4%, respectively (Zhang et al., 2009; Du, Zhu & Yu, 2022), similar to our current study’s sensitivity of 84.3%. These results are higher than those of Suzuki et al. (2004) and Turlakow et al. (2003) with sensitivity ranging from 66.7% to 78%. The specificity and accuracy of PET/CT in our study were 93.0% and 88.3%, respectively, both higher than in previous studies, with specificity and accuracy of 80% and 84.6% for pleural metastasis and 83.3% and 85.0% for peritoneal metastasis, respectively (Zhang et al., 2009; Du, Zhu & Yu, 2022). The possible reason is that in our study, we combined patient clinical data and tumor marker examination results to improve the accuracy of diagnosis.

The current study results demonstrate that PET/CT has high diagnostic performance in detecting primary lesions of unknown serous cavity effusion and detecting serous cavity metastasis, despite the presence of false-negative and false-positive lesions. For primary lesions, there are 10 false-negative and seven false-positive patients. False-negative results occur in mesothelioma, lung adenocarcinoma, fallopian tube adenocarcinoma, hepatocellular carcinoma, MALT lymphoma, and gallbladder signet ring cell carcinoma. Possible causes include low expression of glucose transporter 1 in these pathological types of tumors, resulting in lower uptake of ¹⁸F-FDG on PET/CT (Soyluoglu & Ozdemir Gunay, 2023). Moreover, the high glucose-6-phosphatase activity in the tissues of highly differentiated hepatocellular carcinoma can dephosphorylate the phosphorylated ¹⁸F-FDG, resulting in low ¹⁸F-FDG uptake in PET/CT, leading to false negatives (Sun et al., 2009). Compared to true positive tumors, the smaller tumor volume of these primary tumors is another possible reason for false negativity. False-positive cases occur in ovarian theca fibroma, organizing pneumonia, tuberculous serositis, and Castleman’s disease. From these cases, it can be seen that inflammation and granuloma are the main sources of false positives in our study, and macrophages and neutrophils in inflammatory or granulomatous lesions are metabolically active, resulting in excessive production of glycolytic enzymes that are presented as increased uptake of ¹⁸F-FDG on PET/CT (Murakami et al., 2013). Regarding serous cavity metastasis, false negatives are derived from ovarian adenocarcinoma, rectal adenocarcinoma, and lung adenocarcinoma, which may also be caused by the small size of the lesion resulting in low ¹⁸F-FDG uptake on PET, and the serous membrane thickening is not obvious on CT. In three cases of false negative serous cavity metastasis, bacterial infection occurred in the serous cavity, resulting in strong uptake of ¹⁸F-FDG on PET, leading to erroneous judgment.

As one of the commonly used PET/CT parameters, SUVmax serves as a semi-quantitative indicator of glucose consumption, aiding in distinguishing benign and malignant diseases in most cases. ROC curve analysis revealed that the area under the curve of SUVmax is the largest, significantly surpassing those under the curves of Ca125, Ca199, and CEA. This indicates that PET/CT is the optimal method for differentiating malignant from benign serous cavity effusion. However, the distribution figure of SUVmax indicates that elevated SUVmax is also present in some benign diseases, particularly in non-specific infections and tuberculosis. Similarly, certain malignant tumors, such as the false negative pathological types found in our study (e.g., signet ring cell carcinoma, well-differentiated liver cancer, and MALT lymphoma), exhibit lower SUVmax values, overlapping with benign lesions. Therefore, when using the parameter SUVmax of PET/CT to differentiate benign and malignant serous cavity effusion, it is necessary to combine the patient’s clinical history, serum tumor markers, and anatomical changes on CT to enhance diagnostic accuracy.

A limitation of the current research is that, when evaluating serous cavity metastasis using PET/CT, due to the limited number of patients with pericardial metastasis, subgroup analysis by pericardium, pleura, and peritoneum was not conducted. Additionally, not conducting pathological biopsy of all lesions when evaluating serous cavity metastasis is another limitation, and some lesions were identified through follow-up results of CT, MRI, or PET/CT. However, conducting pathological biopsies on all lesions is not ethical and is not feasible.

Conclusion

¹⁸F-FDG PET/CT proves to be an effective imaging modality for detecting undetermined pleural effusion, including pericardial effusion, pleural effusion and ascites. It demonstrates high diagnostic performance in differentiating primary disease from benign to malignant and evaluating serous cavity metastasis.

Supplemental Information