Authors: EMİNE KARTAL BAYKAN, ENGİN ŞEBİN, ÖMER KARAŞAHİN, AHMED RAMİZ BAYKAN, SERKAN CERRAH, HASAN GÖĞEBAKAN, CAN SEVİNÇ, MUSTAFA KAHRAMAN, YASEMİN COŞKUN YAVUZ
Abstract: Background/aim: Biochemical markers are needed to show lung involvement in COVID-19 disease. Galectin-3 is known to play a key role in the inflammation and fibrosis process. We aimed to evaluate the predictive role of galectin-3 levels for pneumonia in patients with COVID-19. Materials and methods: Total of 176 patients with COVID-19, confirmed with reverse transcriptase polymerase chain reaction, admitted to the Erzurum Regional Training and Research Hospital was analyzed. The study was designed as a cross sectional. The baseline data of laboratory examinations, including galectin-3 were collected at the time of diagnosis. CT images evaluated by a single radiologist according to the recommendation of the Radiological Society of North America Expert Consensus Document for pulmonary involvement. The severity of COVID-19 pneumonia was assessed using the total severity score. Results: The mean galectin-3 level in patients with typical pneumonia was found to be significantly higher than those patients with atypical (p < 0.01) and indeterminate appearance (p < 0.01) and patients without pneumonia (p < 0.01). The severity of lung involvement was significantly associated with Galectin-3 levels (p < 0.01 r: 0.76). Stepwise logistic regression model showed that the levels of ferritin (odds ratio [OR] = 0.05, p: 0.08) and galectin-3 (OR = 0.1, p < 0.01) were significantly and independently associated with typical pneumoniain COVID-19 patients. When COVID-19 patients were evaluated in terms of typical pneumonia, we determined a cut-off value of 18.9 ng/mL for galectin-3 via ROC analysis (87% sensitivity; 73% specificity; area under curve (AUC): 0.89; p < 0.001). Conclusion: Galectin-3 was found as a diagnostic tool for COVID-19 associated typical pneumonia and as an indicator of both pneumonia and its severity.
Keywords: Galectin-3, covid-19, pneumonia
Full Text: PDF