0
Original Research: Disorders of the Pleura |

Derivation and Validation of a CT Scan Scoring System for Discriminating Malignant From Benign Pleural EffusionsCT Scan Scoring System for Pleural Effusions FREE TO VIEW

José M. Porcel, MD, FCCP; Marina Pardina, MD; Silvia Bielsa, MD; Antonio González, MD; Richard W. Light, MD, FCCP
Author and Funding Information

From the Department of Internal Medicine (Drs Porcel and Bielsa) and the Department of Radiology (Drs Pardina and González), Pleural Diseases Unit, Arnau de Vilanova University Hospital, Biomedical Research Institute of Lleida, Lleida, Spain; and the Division of Allergy, Pulmonary, and Critical Care (Dr Light), Vanderbilt University Medical Center, Nashville, TN.

CORRESPONDENCE TO: José M. Porcel, MD, FCCP, Department of Internal Medicine, Arnau de Vilanova University Hospital, Avda Alcalde Rovira Roure 80, 25198 Lleida, Spain; e-mail: jporcelp@yahoo.es


FUNDING/SUPPORT: The authors have reported to CHEST that no funding was received for this study.

Reproduction of this article is prohibited without written permission from the American College of Chest Physicians. See online for more details.


Chest. 2015;147(2):513-519. doi:10.1378/chest.14-0013
Text Size: A A A
Published online

BACKGROUND:  Chest CT scanning has become an integral part of the workup for undiagnosed pleural effusions. We aimed to develop a CT scan-based scoring system for differentiating between benign and malignant pleural effusions.

METHODS:  A number of chest CT scan abnormalities were compared between 228 patients with benign and 115 with malignant effusions (derivation cohort). A logistic regression analysis was used to identify the independent predictors of malignancy and generate CT scan scores, with more points assigned to those findings associated with higher β-coefficient values. The diagnostic accuracy of the CT scan scoring system was calculated for the derivation cohort and further evaluated in two independent populations (n = 80 and 42, respectively) by two radiologists.

RESULTS:  CT scan scores predicting malignancy included any pleural lesion (ie, nodule, mass, or thickening) ≥ 1 cm (5 points); the presence of liver metastases, an abdominal mass, or a lung mass or nodule ≥ 1 cm (3 points each); and the absence of either pleural loculations, pericardial effusions, or cardiomegaly (2 points each). In the first validation cohort, a sum score of ≥ 7 yielded a sensitivity of 88% (95% CI, 73%-95%), specificity of 94% (95% CI, 83%-98%), likelihood ratio positive of 13.8 (95% CI, 4.6-41.5), likelihood ratio negative of 0.13 (95% CI, 0.05-0.33), and area under the receiver operating characteristics curve of 0.919 (95% CI, 0.849-0.990). Moreover, 69% of 42 patients with pathologically unconfirmed malignant effusions from a second independent cohort would have been correctly labeled by the predictive score.

CONCLUSIONS:  A simple CT scan-based scoring system can help physicians to separate malignant from benign pleural effusions.

Figures in this Article

Clinicians often encounter patients with pleural effusions that remain undiagnosed after the initial workup, which includes a thoracentesis. One of the main concerns is to rule out malignancy, since it is well known that pleural fluid cytology has a sensitivity of only 60%.1 Another concern is to avoid subjecting frail patients to unnecessary invasive procedures and selecting just those who may benefit the most from such interventions.

Chest CT scans furnish clues for a potentially underlying pathology in patients with pleural effusions (eg, pulmonary infiltrates, masses or emboli, mediastinal lymphadenopathies, or pericardial involvement). In addition, several CT scan features have been described in previous studies that may aid in the differentiation of malignant from benign pleural conditions. For example, images may reveal pleural nodularity and thickening, which are considered highly suggestive of a malignant cause.1 However, the evidence supporting the discriminatory role of CT scanning relies upon just a few series in which the results were not validated in an independent population and often included a large percentage of mesotheliomas, which are more likely to exhibit abnormalities of the pleural surfaces.28

In the present study, in addition to addressing the previous shortcomings we aimed to develop a simple CT scan scoring system, based on easily identifiable radiologic data, to be used by pulmonologists as a tool for separating malignant effusions from benign pleurisy.

Patient Selection

From June 2008 to December 2011, all adult patients with pleural effusions who underwent both a diagnostic thoracentesis and a contrast-enhanced chest CT scan at our institution, being determined as necessary by the attending physicians, were enrolled in this prospective study. Obtaining a definite cause of the pleural effusion was required for final inclusion. This population allowed for the generation of a pleural malignant-benign discriminating CT scan score, which was subsequently validated in a different set of patients recruited prospectively during 2012. Two different cohorts were further analyzed. One (ie, validation cohort) was composed of a stratified sample of consecutive patients with confirmed benign and malignant pleural effusions, to closely mirror the derivation cohort with respect to the distribution of causes. The other included patients with highly probable, but unconfirmed, malignant effusions based on the following criteria: demonstration of a primary tumor or clinical evidence of tumor dissemination (eg, liver metastases); exclusion of other potential causes for the effusion; and negative pleural fluid cytologic examination, which was considered to be a false-negative result. In this subgroup of patients, a definitive diagnosis of malignancy through a pleural biopsy was not pursued, because the oncologist did not anticipate a change in treatment strategy in the event of a histopathologic confirmation of the pleural cancer, because of one of the following reasons: (1) the patient had a poor performance status, and, consequently, only palliative care was indicated; or (2) adjuvant chemotherapy and/or radiotherapy was already being considered as a beneficial treatment. The institutional ethics committee of our hospital (CEIC, University Hospital Arnau de Vilanova, ID 974/2011) approved the study protocol, and the participants signed written informed consent.

Diagnostic Criteria

An effusion was categorized as malignant if malignant cells were detected upon cytologic examination of the pleural fluid or biopsy specimens. Tuberculous pleuritis was diagnosed if Ziehl-Neelsen stains or Lowenstein-Jensen cultures of pleural fluid, sputum, or pleural biopsy specimens were positive or if granulomas were found in the parietal pleural biopsies. TB was considered probable in patients with lymphocytic exudates containing high pleural adenosine deaminase levels (> 35 U/L) that cleared in response to antituberculous therapy. A parapneumonic effusion was defined as any effusion associated with bacterial pneumonia, lung abscess, or bronchiectasis, and empyema described the presence of pus within the pleural space. The diagnosis of heart failure was based on history, physical examination, chest radiographs, ECG, echocardiogram (if available), and response to diuretics. Other causes of pleural effusions followed well-established clinical criteria. All patients with benign conditions were followed up by means of clinical examination and imaging until the complete resolution of the pleural effusion.

CT Scan Examination

All CT imaging studies were performed with a 16-multiple detector CT scanner (Brilliance 16; Koninklijke Philips N.V.), using the following scan parameters: collimation 16 × 1.5 mm, tube rotation time 0.75 s, pitch 0.938, 120 kV, and section thickness of 3 mm with a 1.5-mm reconstruction interval. Patients were given 120 mL of nonionic contrast material (iobitridol, Xenetic 300; Guerbet Group; or iopromide, Ultravist 300; Bayer AG) at 3 mL/s injected through an IV cannula (Abbocath 20G) with a mechanical automated injector (Med-tron). Images, which were acquired after a 60-s delay from the start of injection, extended from the lung apices to the level of the adrenals and were viewed with a picture archiving and communication system with standard lung (level, −450 Hounsfield units [HU]; width, 1,600 HU) and mediastinal (level, 40 HU; width, 400 HU) windows.

The CT scan features evaluated for the discriminating analysis were: (1) focal pleural thickening (ie, a visible pleural line of ≥ 3 mm, limited to a single lobe); (2) diffuse pleural thickening, which was further characterized by whether it enveloped the entire perimeter of the hemithorax at any CT scan slice (circumferential thickening); (3) pleural nodules (ie, focal pleural thickening with a nodular contour and < 3 cm in length); both nodules and thickening were assessed in terms of their precise size and anatomic location (eg, parietal, visceral, mediastinal, or diaphragmatic); (4) pleural-based masses (≥ 3 cm); (5) pleural contrast enhancement, including the “split pleural sign,” whereby enhanced visceral and parietal pleural surfaces are seen separated by pleural fluid; (6) increase in attenuation of extrapleural fat; (7) pleural calcifications; (8) pleural fluid attenuation values expressed in HU; (9) pleural loculations (ie, an effusion that is compartmentalized, or has septations or a convex shape facing the lung parenchyma, or is accumulated in a nondependent portion); (10) effusion’s side (bilaterality); (11) volume of pleural effusions using a previously reported formula9; (12) thoracic lymph node enlargement > 1 cm in the short-axis diameter; (13) parenchymal lung lesions (ie, nodules, masses, atelectasis, alveolar or interstitial patterns); (14) pericardial effusion; (15) cardiomegaly (cardiothoracic ratio > 0.5 in axial images); (16) liver metastases; (17) abdominal masses displayed at the upper abdominal slices from chest CT scan; (18) hepatomegaly (craniocaudal liver span > 15.5 cm in the midclavicular line); and (19) dilation of the inferior vena cava (diameter > 1.7 cm) measured just above the entrance of the suprahepatic veins.

Interpretation of chest CT imaging in the derivation cohort was done by a thoracic radiologist blinded to the final diagnosis. The same investigator, together with a general radiologist, both with 20 years of experience, applied the resulting CT scan scoring system to the validation cohorts, also in blinded fashion, with disagreements resolved by consensus. For bilateral pleural effusions, the CT scan findings were recorded from the side of diagnostic thoracentesis or biopsy.

Statistical Analysis

Categorical and continuous data are expressed as numbers (percentages) and medians (25th and 75th percentiles), respectively. Between-group comparisons were performed with the χ2, Fisher exact, Kruskal-Wallis, and Mann-Whitney U tests, as appropriate. Receiver operating characteristic (ROC) curve analysis helped to decide the optimal cutoff points for continuous variables based on their highest diagnostic accuracy. A logistic regression analysis with backward conditional method served to select those radiologic variables entering the scoring system. Weight values to each variable were assigned proportionally to the magnitude of the logistic equation’s coefficients. The diagnostic performance of the derived score for labeling malignant pleural effusions in both the derivation and validation cohorts was calculated through 2 × 2 contingency tables (sensitivity, specificity, likelihood ratios [LRs]) and the area under ROC curve (AUC). Interobserver agreement about radiologic signs was expressed using the κ statistic. The statistical significance level was set at 0.05 (two-tailed). All analyses were conducted with the SPSS version 18.0 statistical software (IBM).

Study Population

Baseline characteristics of the 343 patients who composed the derivation cohort are shown in Table 1. Patients were aged 69 (53-80) years, 59% were men, and 34% had malignant effusions, which were metastatic in 94% of the cases. Lung cancer was the leading cause (36.5%) of pleural malignancies. The validation cohort consisted of 80 patients, with a median age of 71 (55-83) years, of whom 41% had pleural malignancy, mostly (97%) of metastatic origin. The generated CT scan score was also applied to an additional sample of 42 patients (median age, 73 [58-85] years, 52% men) with probable malignant effusions from the following tumor primaries: 21 lung; six breast; two each of lymphoma, kidney, melanoma, pancreas, colon, and stomach; and one each of sarcoma, unknown primary, and GI stromal tumor.

Table Graphic Jump Location
TABLE 1 ]  Baseline Characteristics of the Derivation and Validation Cohort

Data are expressed as No. (%), No., or median (quartiles) as appropriate.

CT Scan Data From the Derivation Cohort

Most of the evaluated radiologic variables were found to be significantly different between malignant and benign pleural effusions in the bivariate analysis, with the exception of bilateral effusions, circumferential pleural thickening, pleural calcifications, pleural enhancement after contrast infusion, CT scan attenuation values of the pleural fluid, and hepatomegaly (Table 2). Pleural nodularity or thickening ≥ 1 cm (cutoff established by ROC analysis) was the most consistent feature arguing for malignant effusions (Figs 1, 2). Three patients with benign conditions (two TB and one sarcoidosis) exhibited pleural nodules < 1 cm. Two of those three patients (one TB and one sarcoidosis) also had focal pleural thickening < 1 cm. Diffuse pleural thickening was present in 11 parapneumonics/empyemas, which was circumferential for seven, and measured > 1 cm in one. The presence of any pleural lesion ≥ 1 cm, either nodules, masses, or thickening, had a sensitivity of 43% (95% CI, 34%-52%), specificity of 99.6% (95% CI, 98%-100%), LR positive of 97.2 (95% CI, 13.6-694.6), and LR negative of 0.58 (95% CI, 0.49-0.68) for identifying malignant effusions.

Table Graphic Jump Location
TABLE 2 ]  Chest CT Scan Findings in the Derivation Cohort

Data are expressed as No. (%) or median (quartiles) as appropriate.

a 

Pleural fluid volume was calculated by a formula from Reference 9.

b 

Abdominal masses were located in the peritoneum (7), suprarenal glands (4), pancreas (2), kidney (2), lymph nodes (2), spleen (1), and stomach (1).

Figure Jump LinkFigure 1 –  CT scan showing a malignant pleural nodule (arrowhead).Grahic Jump Location
Figure Jump LinkFigure 2 –  Radiologic appearance of malignant pleural thickening (arrowheads).Grahic Jump Location

Of interest, CT scan showed pleural loculations in 78 (76%) parapneumonics/empyemas and 13 (59%) tuberculous, 40 (35%) malignant, and 16 (29%) heart failure-related effusions. Thoracic lymphadenopathies were seen in 35%, 41%, 50%, and 15% of these respective causes. As expected, cardiomegaly and dilation of the inferior vena cava were prevalent (55% and 100%, respectively) in heart failure, also being characterized by a low attenuation coefficient (decreased density) of the pleural fluid (−4 HU; quartiles, −7 to 1).

Generation of a CT Scan Score

All radiologic findings that displayed statistical significance in the bivariate analysis were processed using a multivariable logistic regression model. The model selected nine as being predictive of malignancy. However, two of them, namely obstructive atelectasis and the volume of pleural effusion, were then excluded because they are not readily recognizable by an inexperienced pulmonologist. The remaining seven parameters helped to establish weight scores (Table 3) as follows: any pleural lesion ≥ 1 cm (5 points), liver metastases (3 points), abdominal mass (3 points), lung mass or lung nodule ≥ 1 cm (3 points), absence of pleural loculations (2 points), no pericardial effusion (2 points), and nonenlarged cardiac silhouette (2 points). Thus, CT scan scores ranged from 0 to 20.

Table Graphic Jump Location
TABLE 3 ]  CT Scan Scoring System for Predicting Malignant Pleural Effusions
a 

Nodule, mass, or thickening.

Table 4 shows the performance of predictive scores in the derivation group. As the cutoff value was raised, the specificity of the rule increased, whereas the sensitivity concurrently decreased. At the best cutoff of 7 points, the scoring system yielded 74% sensitivity (95% CI, 65%-81%), 92% specificity (95% CI, 88%-95%), LR positive 9.4 (95% CI, 5.9-14.8), LR negative 0.28 (95% CI, 0.21-0.39), and AUC 0.908 (95% CI, 0.873-0.942) for discriminating malignant and benign pleural effusions. The respective median CT scan scores in patients with heart failure, parapneumonics, TB, and malignancy were 4 (2-6), 4 (4-6), 6 (4-7), and 9 (6-11). If we had only considered patients with exudative effusions and a negative first cytologic fluid analysis, which reflects a more typical scenario in which CT scan is ordered, a CT scan score ≥ 7 would have had 74% sensitivity (95% CI, 55%-87%), 91% specificity (95% CI, 85%-95%), LR positive 8.44 (95% CI, 4.5-16), and LR negative 0.28 (95% CI, 0.15-0.54) for labeling malignant pleurisy.

Table Graphic Jump Location
TABLE 4 ]  Operating Characteristics of the CT Scan Scoring System for Identifying Malignant Effusions (Derivation Cohort)

LR = likelihood ratio.

Validation of the CT Scan Scoring System

The operating characteristics of a CT scan score ≥ 7 in the validation cohort (33 malignant and 47 benign new effusions) were: sensitivity 88% (95% CI, 73%-95%), specificity 94% (95% CI, 83%-98%), LR positive 13.8 (95% CI, 4.6-41.5), LR negative 0.13 (95% CI, 0.05-0.33), and AUC 0.919 (95% CI, 0.849-0.990). The predictive model was then evaluated in 42 patients with a clinical diagnosis of malignant effusions whose pleural fluid cytology was considered to be falsely negative. Ten (24%) exhibited pleural lesions ≥ 1 cm, 22 (52%) lung masses or nodules ≥ 1 cm, eight (19%) abdominal masses, and six (14%) liver metastases, whereas the absence of pleural loculations, pericardial effusions, and cardiomegaly characterized 24 (57%), 40 (95%), and 36 (86%) cases, respectively. Overall, 29 (69%) patients scored ≥ 7, a figure that did not differ significantly from that of the confirmed malignant effusions from the derivation cohort (74%, P = .54). For most of the parameters composing the CT scan score, the interobserver agreement was substantial, as indicated by a κ statistic > 0.6 (Table 5).

Table Graphic Jump Location
TABLE 5 ]  Interobserver Agreement for CT Scan Variables (Validation Cohort)

We have developed and validated a CT scan scoring system to assess the probability of malignancy in patients with pleural effusions. The proposed method not only appears to be simple and feasible but also has high interobserver reliability and accurately distinguishes between benign and malignant pleural effusions, as reflected by an AUC of 0.919. From a maximum sum of 20 points, a score ≥ 7 would prompt consideration of a malignant process, whereas a score of < 5 would mitigate against it. Of note, the discriminative properties of the CT scan scoring system remained unchanged for the subgroup of patients in whom no diagnosis had been obtained after an initial thoracentesis (ie, exudative effusion and a first negative pleural fluid cytologic result), the population for which CT scan is most commonly ordered in routine clinical practice.

Although the CT scan findings of pleural effusions secondary to malignancy,10 empyema,11 or TB12 have been described individually, few previous studies have analyzed the value of CT scan for the differential diagnosis of benign and malignant causes of pleural effusions.28 Unfortunately, most were retrospective,24,68 did not perform a multivariate analysis to select the relevant CT scan discriminative data,25,7,8 and, most importantly, did not validate their findings using an independent population.28

Some CT scan features, namely pleural nodularity, pleural thickening ≥ 1 cm, circumferential thickening (pleural rind) and mediastinal pleural involvement have been consistently reported as highly suggestive of a malignant process. The sum of seven series, totaling 457 patients with malignant pleural effusions (35% mesotheliomas) and 475 with benign pleural conditions (50% TB),28 yields respective mean sensitivities and specificities for the preceding CT scan data as follows: pleural nodules, 39% (range, 18%-53%) and 95% (range, 87%-100%)28; pleural thickening ≥ 1 cm, 34% (range, 7%-47%) and 87% (range, 64%-98%)2,68; mediastinal pleural involvement, 46.5% (range, 14%-74%) and 84% (range, 68%-97%)24,68; and pleural rind, 32% (range, 7%-54%) and 88.5% (range, 63%-100%).28 The operating characteristics of pleural nodules and thickening in our large prospective series compared similarly with these figures. Nevertheless, our findings showed a trend toward a lower proportion of pleural rind (4%) in malignant effusions, probably due to the low number of mesothelioma cases included. Furthermore, some classic malignant-related radiologic data, such as pleural mediastinal involvement, lost significance as predictors of malignancy due to the strong influence of pleural nodularity or thickening in the logistic regression model.

It was found that benign conditions, such as TB and parapneumonic effusions, rarely demonstrated some features typically associated with metastases to the pleura, that is, pleural nodularity and thickening. These particular CT scan findings are also helpful in guiding pleural biopsies to the sites more likely to yield a positive diagnosis, in cases where pleural fluid analyses are not contributory.13 In this sense, pleural lesions easily amenable to biopsy occurred in 43% of confirmed and 24% of cytologically negative malignant effusions. The generated scoring system outperformed the predictive capability of any individual parameter. Its diagnostic threshold (7 points) was exceeded by about three-fourths of patients with malignant pleurisy, regardless of the results of pleural fluid cytology.

This study is subject to limitations. It can be argued that the applicability of the scoring system may depend on the prevalence of specific causes of pleural effusions in the population in which it is to be implemented. Our derivation and validation cohorts accurately reflected the relative proportion of pleural effusion causes in our country.14 For example, lung cancer and mesothelioma accounted for 37% and 3% of 840 malignant effusions, respectively, whereas TB was the cause of 9% of 3,077 pleural effusions in a recent series from our hospital.14 In the derivation cohort, these percentages were 36.5%, 5%, and 6.5%, respectively. However, in other geographic areas, the incidence of mesothelioma or TB may substantially change and so too may the accuracy of the CT scan score. In most of the previous benign-malignant CT scan discrimination studies, there was indeed an overrepresentation of mesotheliomas2,3,57 and TB68 in the comparison groups, pointing to a selection bias. A second potential limitation relates to radiologic interpretation. Although the radiologist was blinded to the clinical diagnosis, the assessment of the pleural surfaces may have been biased in those patients having additional CT scan features that suggested malignancy (eg, liver metastases, lung masses or nodules). Moreover, the requirement for both CT scan and pleural fluid data among the inclusion criteria might appear to introduce a potential selection bias. However, knowledge of pleural fluid analyses is essential for definitively classifying patients who undergo a CT scan examination as having a definite malignant or benign effusion. Finally, the scoring system should be viewed as an aid for constructing a differential diagnosis of pleural effusions, which does not preclude obtaining a cytohistologic confirmation of malignancy. It may be considered a diagnostic weight to estimate the probability of malignancy, in combination with other clinical findings. A prospective validation of this CT scan score in large cohorts is warranted. To conclude, a simple chest CT scan scoring system, which includes the evaluation of pleural nodularity or thickening, liver metastases, abdominal masses, lung nodules or masses, pleural loculations, pericardial effusions, and cardiomegaly, may be reliably used for the differential diagnosis of pleural effusions.

Author contributions: J. M. P. is the guarantor of this study. J. M. P. contributed to conception and design, acquisition of data, drafting the manuscript, and approving the final version; M. P. contributed to acquisition of data, analysis and interpretation of data, and approving the final version; S. B. contributed to research design, analysis and interpretation of data, and approving the final version; A. G. contributed to analysis and interpretation of data and approving the final version; and R. W. L. contributed to critical revision of the manuscript for important intellectual content and approving the final version.

Financial/nonfinancial disclosures: The authors have reported to CHEST that no potential conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article.

AUC

area under the receiver operating characteristic curve

HU

Hounsfield units

LR

likelihood ratio

ROC

receiver operating characteristic

Porcel JM, Light RW. Pleural effusions. Dis Mon. 2013;59(2):29-57. [CrossRef] [PubMed]
 
Leung AN, Müller NL, Miller RR. CT in differential diagnosis of diffuse pleural disease. AJR Am J Roentgenol. 1990;154(3):487-492. [CrossRef] [PubMed]
 
Hierholzer J, Luo L, Bittner RC, et al. MRI and CT in the differential diagnosis of pleural disease. Chest. 2000;118(3):604-609. [CrossRef] [PubMed]
 
Arenas-Jiménez J, Alonso-Charterina S, Sánchez-Payá J, Fernández-Latorre F, Gil-Sánchez S, Lloret-Llorens M. Evaluation of CT findings for diagnosis of pleural effusions. Eur Radiol. 2000;10(4):681-690. [CrossRef] [PubMed]
 
Traill ZC, Davies RJ, Gleeson FV. Thoracic computed tomography in patients with suspected malignant pleural effusions. Clin Radiol. 2001;56(3):193-196. [CrossRef] [PubMed]
 
Metintas M, Ucgun I, Elbek O, et al. Computed tomography features in malignant pleural mesothelioma and other commonly seen pleural diseases. Eur J Radiol. 2002;41(1):1-9. [CrossRef] [PubMed]
 
Yilmaz U, Polat G, Sahin N, Soy O, Gülay U. CT in differential diagnosis of benign and malignant pleural disease. Monaldi Arch Chest Dis. 2005;63(1):17-22. [PubMed]
 
Kim JS, Shim SS, Kim Y, Ryu YJ, Lee JH. Chest CT findings of pleural tuberculosis: differential diagnosis of pleural tuberculosis and malignant pleural dissemination. Acta Radiol. 2014;55(9):1063-1068. [CrossRef] [PubMed]
 
Mergo PJ, Helmberger T, Didovic J, Cernigliaro J, Ros PR, Staab EV. New formula for quantification of pleural effusions from computed tomography. J Thorac Imaging. 1999;14(2):122-125. [CrossRef] [PubMed]
 
O’Donovan PB, Eng P. Pleural changes in malignant pleural effusions: appearance on computed tomography. Cleve Clin J Med. 1994;61(2):127-131. [CrossRef] [PubMed]
 
Waite RJ, Carbonneau RJ, Balikian JP, Umali CB, Pezzella AT, Nash G. Parietal pleural changes in empyema: appearances at CT. Radiology. 1990;175(1):145-150. [CrossRef] [PubMed]
 
Yilmaz MU, Kumcuoglu Z, Utkaner G, Yalniz O, Erkmen G. Computed tomography findings of tuberculous pleurisy. Int J Tuberc Lung Dis. 1998;2(2):164-167. [PubMed]
 
Maskell NA, Gleeson FV, Davies RJ. Standard pleural biopsy versus CT-guided cutting-needle biopsy for diagnosis of malignant disease in pleural effusions: a randomised controlled trial. Lancet. 2003;361(9366):1326-1330. [CrossRef] [PubMed]
 
Porcel JM, Esquerda A, Vives M, Bielsa S. Etiology of pleural effusions: analysis of more than 3,000 consecutive thoracenteses. Arch Bronconeumol. 2014;50(5):161-165. [CrossRef] [PubMed]
 

Figures

Figure Jump LinkFigure 1 –  CT scan showing a malignant pleural nodule (arrowhead).Grahic Jump Location
Figure Jump LinkFigure 2 –  Radiologic appearance of malignant pleural thickening (arrowheads).Grahic Jump Location

Tables

Table Graphic Jump Location
TABLE 1 ]  Baseline Characteristics of the Derivation and Validation Cohort

Data are expressed as No. (%), No., or median (quartiles) as appropriate.

Table Graphic Jump Location
TABLE 2 ]  Chest CT Scan Findings in the Derivation Cohort

Data are expressed as No. (%) or median (quartiles) as appropriate.

a 

Pleural fluid volume was calculated by a formula from Reference 9.

b 

Abdominal masses were located in the peritoneum (7), suprarenal glands (4), pancreas (2), kidney (2), lymph nodes (2), spleen (1), and stomach (1).

Table Graphic Jump Location
TABLE 3 ]  CT Scan Scoring System for Predicting Malignant Pleural Effusions
a 

Nodule, mass, or thickening.

Table Graphic Jump Location
TABLE 4 ]  Operating Characteristics of the CT Scan Scoring System for Identifying Malignant Effusions (Derivation Cohort)

LR = likelihood ratio.

Table Graphic Jump Location
TABLE 5 ]  Interobserver Agreement for CT Scan Variables (Validation Cohort)

References

Porcel JM, Light RW. Pleural effusions. Dis Mon. 2013;59(2):29-57. [CrossRef] [PubMed]
 
Leung AN, Müller NL, Miller RR. CT in differential diagnosis of diffuse pleural disease. AJR Am J Roentgenol. 1990;154(3):487-492. [CrossRef] [PubMed]
 
Hierholzer J, Luo L, Bittner RC, et al. MRI and CT in the differential diagnosis of pleural disease. Chest. 2000;118(3):604-609. [CrossRef] [PubMed]
 
Arenas-Jiménez J, Alonso-Charterina S, Sánchez-Payá J, Fernández-Latorre F, Gil-Sánchez S, Lloret-Llorens M. Evaluation of CT findings for diagnosis of pleural effusions. Eur Radiol. 2000;10(4):681-690. [CrossRef] [PubMed]
 
Traill ZC, Davies RJ, Gleeson FV. Thoracic computed tomography in patients with suspected malignant pleural effusions. Clin Radiol. 2001;56(3):193-196. [CrossRef] [PubMed]
 
Metintas M, Ucgun I, Elbek O, et al. Computed tomography features in malignant pleural mesothelioma and other commonly seen pleural diseases. Eur J Radiol. 2002;41(1):1-9. [CrossRef] [PubMed]
 
Yilmaz U, Polat G, Sahin N, Soy O, Gülay U. CT in differential diagnosis of benign and malignant pleural disease. Monaldi Arch Chest Dis. 2005;63(1):17-22. [PubMed]
 
Kim JS, Shim SS, Kim Y, Ryu YJ, Lee JH. Chest CT findings of pleural tuberculosis: differential diagnosis of pleural tuberculosis and malignant pleural dissemination. Acta Radiol. 2014;55(9):1063-1068. [CrossRef] [PubMed]
 
Mergo PJ, Helmberger T, Didovic J, Cernigliaro J, Ros PR, Staab EV. New formula for quantification of pleural effusions from computed tomography. J Thorac Imaging. 1999;14(2):122-125. [CrossRef] [PubMed]
 
O’Donovan PB, Eng P. Pleural changes in malignant pleural effusions: appearance on computed tomography. Cleve Clin J Med. 1994;61(2):127-131. [CrossRef] [PubMed]
 
Waite RJ, Carbonneau RJ, Balikian JP, Umali CB, Pezzella AT, Nash G. Parietal pleural changes in empyema: appearances at CT. Radiology. 1990;175(1):145-150. [CrossRef] [PubMed]
 
Yilmaz MU, Kumcuoglu Z, Utkaner G, Yalniz O, Erkmen G. Computed tomography findings of tuberculous pleurisy. Int J Tuberc Lung Dis. 1998;2(2):164-167. [PubMed]
 
Maskell NA, Gleeson FV, Davies RJ. Standard pleural biopsy versus CT-guided cutting-needle biopsy for diagnosis of malignant disease in pleural effusions: a randomised controlled trial. Lancet. 2003;361(9366):1326-1330. [CrossRef] [PubMed]
 
Porcel JM, Esquerda A, Vives M, Bielsa S. Etiology of pleural effusions: analysis of more than 3,000 consecutive thoracenteses. Arch Bronconeumol. 2014;50(5):161-165. [CrossRef] [PubMed]
 
NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

Find Similar Articles
CHEST Journal Articles
PubMed Articles
  • CHEST Journal
    Print ISSN: 0012-3692
    Online ISSN: 1931-3543