Mikrobiyoloji bulteni, cilt.55, ss.113-124, 2021 (SCI-Expanded)
Brusellosis is the world's most common bacterial zoonotic infection and is still endemic in many developing countries. The clinical appearance of brucellosis is not specific, but the course and severity of the infection varies. In this study, it was aimed to determine the relationship between laboratory parameters and clinical response and organ involvement in patients with Brucellosis diagnosed with specific diagnostic tests. In the study, 100 patients without previous diagnosis of Brucellosis who have admitted to the Department of Infectious Diseases and Clinical Microbiology and had positive Brucella tube agglutination tests and whose clinic was compatible with Brucellosis were evaluated prospectively. Patients were invited to be checked in the 1st, 2nd and 6th weeks. Patients with hip pain and low back pain were evaluated with sacroiliac magnetic resonance imaging (MRI) and lumbar MRI for sacroiliitis and spondylodiscitis. Patients with liver and bone marrow involvements, sacroiliitis, spondylodiscitis and orchitis were recorded as the patients with organ involvement. After six weeks, the decline of the complaints was considered as a clinical response. In the 6th week of the treatment, it was observed that platelet distribution width (PDW) and mean platelet volume (MPV) levels were lower in patients with a clinical response compared to the patients with no clinical response which was statistically significant (p= 0.01, p= 0.02). Platelet and platelet lymphocyte ratio (PLR) level in patients with organ involvement in the 1st and 6th weeks of the treatment, were observed to be lower than the patients without organ involvement which was statistically significant (week 1: p= 0.001, p= 0.01; week 6: p= 0.03, p= 0.01). Among patients with organ involvement and non-organ involvement, the area under the curve was 66% in the ROC curve analysis for PLR at the onset of the treatment. When the cut-off value was taken as 128.8%, the sensitivity was 55%, and the specificity was 78%. Depending on the level of platelet at the beginning of the treatment, in the ROC curve analysis carried out among patients with organ involvement and non-organ involvement, the area under the curve was 73% and when the cut off value was taken as 256000, the sensitivity was 71%, and the specificity was 68%. In the 6th week of the treatment, the area under the curve was observed as 67% in the ROC curve analysis of the PDW level among patients from whom clinical responses were received and not received. When the cut-off value was taken as 10.75%, the sensitivity was 65%, and the specificity was 70%. In the ROC curve analysis of the MPV value for clinical response, the area under the curve was 66%, and when the cut-off value was taken as 9.95, the sensitivity was observed as 52%, and the specificity was 74%. As a result, in the evaluation of the clinical response, which is important in the termination of the treatment in patients with Brucellosis, the use of MPV and PDW values in the evaluation of organ involvement and platelet level and PLO in the follow up, are cheap, easily accessible biomarkers that can be used clinically.