Radiographic Characteristics of Mediastinal and Hilar Lymph Nodes in Sarcoidosis

INTRODUCTION Sarcoidosisis a multisystemic granulomatous disease whose etiology is unknown. The diagnosis of sarcoidosisis made by exclusion of other causes in addition to compatible clinical, radiological and histopathological findings (1,2). Its most frequently encountered form of stage-I sarcoidosis characterized by bilateral hilar lymphadenopathy (3). Symmetrical hilar lymphadenopathy (LAP) is a significant characteristic that distinguishes sarcoidosis from disease that may progress with mediastinal and hilar lymph nodes such as lymphoma, fungal infection and tuberculosis (4,5). Unilateral hilar LAP is seen in only about 3-5%of sarcoidosis cases (6).

Tissue biopsy is required to show granulomas that does not histopathologically contain caseification necroses (7). However, in stage-I patients, sarcoidosis can be diagnosed without tissue sampling after excluding other causes (8). The 2019 sarcoidosis guideline of BTS (British Thoracic Society) also stated that, in patients with Lofgren syndrome (Bilateral hilar lymphadenopathy, erythema nodosum, fever, arthritis) with a low probability of alternative diagnosis, biopsy is required only when radiologically atypical findings are revealed during follow up (9). However, in this case, the suspicion of whether or not the correct decision is made remains a highly disturbing issue for both the physician and the patient. Biopsy should be performed if the clinical condition is not typical, and other causes could not be excluded. Bronchoscopy, which is frequently used for biopsy, is an invasive procedure although its complication rate is low. In relation to bronchoscopy, complications such as pneumothorax, hemorrhage and infection may be experienced (10,11).Therefore, it is important to investigate if it is possible to prevent patients from unnecessary invasive procedures. This study aimed to investigate the diagnostic value of thoracic computerized tomography image characteristics in the diagnosis of sarcoidosis without pathological sampling.

MATERIAL-METHOD The study was planned by retrospectively analyzing the data of patients who received EBUS for sampling mediastinal lymph nodes and were subjected to other diagnostic methods such as mediastinoscopy and thoracotomy in the case of non-diagnostic results of EBUS at the Department of Pulmonology at the Faculty of Medicine at KTÜ between 1 January 2013 and 1 July 2019. The study was started after obtaining local ethics board approval, and only included patients with definitive histopathological diagnosis. The patients whose computerized thoracic tomography images at first admission were not available in the archives of our hospital were excluded (Figure 1).

The computerized thoracic tomography images were examined in terms of characteristics as mediastinal lymph node localization, size, density, homogeneity, necrosis, calcification and hilar symmetry. Parenchymal lesions were compared in terms of nodule, ground-glass, reticular opacity, consolidation and the distributions of these lesions. For computerized thoracic tomography imaging, the Somatom (Siemens, Forchhim, Germany) device at our hospital was used. For EBUS imaging and sampling, an Olympus EVIS EXERA II CV-180 device was used.

The assessment of lymph node localization was made based on the lymph node map by Wang (2R, 2L, 4R, 4L, etc.). Measurements of lymphadenopathy density were made by considering the largest lymph node and by the Hounsfield Unit (HU) in a rectangular region determined in the axial plane contacting the LAP borders from four corners in the tomography cross-section with the lymphadenopathy showing the broadest area. The maximum, minimum and mean densities measured at the marked region were recorded. The measurements were made by the same person in all patients (Figure 2).

Statistical analysis Kolmogorov-Smirnov test was used to test the normal distribution of the continuous variables. The data characterized by a normal distribution are expressed as mean±standard deviation. Student's t-tests was used for the comparison of the data which had a normal distribution. Mann-Whitney-U test was used for the comparison of the non-normally distributed data. The discrete variables were compared using Chi-squared test. The parameters that were potential predictors of sarcoidosis were analyzed using logistic regression analysis. Multivariate logistic regression analysis was used as a stepwise backward LR method from predictive factors with a significance of ≤0.05 in the univariate analysis. The diagnostic accuracies of the parameters for identifying sarcoidosis were assessed using the area under the receiver operating characteristic (ROC) (AUC). P <0.05 was considered to be statistically significant. The data were analyzed using the SPSS statistical software (version 13.01, serial number 9069728, SPSS Inc., Chicago).