Some investigators have proposed the use of a combination of markers, such as IL-6, which is an acute reactor, and CRP, which increases later in the course of sepsis [5, 8, 17]. In the present study, this combination did not offer better diagnostic value Selleckchem BI2536 than IL-6 alone. TNF-α at the higher cut-off level (>30 pg/ml) was found to be a good predictor of sepsis but not as precise as IL-6, confirming previous data [5, 8, 17]. Finally, IL-1b was proven to be a specific but not sensitive index of neonatal infection [8, 18]. The levels of all three cytokines decreased during the course

of the study, but remained higher in the sepsis and suspected infection groups compared with the control group. Ng et al. [5] found that the IL-6 levels decreased by 83% 48 h after the introduction of treatment in very low birthweight neonates with sepsis. In the present study, IL-6 was found to be reduced by 50% 2 days after the introduction of treatment in neonates with sepsis, while TNF-α was reduced to a lesser degree. More similar are the

findings of Santana-Reyes et al. [19], namely that full-term neonates with suspected infection had lower IL-6 levels than neonates with sepsis, but higher than controls at the beginning of clinical signs of infection [19]. In their study, in accordance with the present study, IL-6 levels remained higher than selleck kinase inhibitor baseline values in neonates with suspected and documented infection 3 days after the introduction of antibiotics. Although neonates with a very high clinical suspicion of sepsis, despite negative cultures, were not included in the present study, it cannot be certain that Suplatast tosilate all of the remaining neonates with suspected infection were infection-free. This may be the reason for the elevated infection indices in some neonates of this group. Studies in adults with sepsis have shown changes in the subpopulations of lymphocytes and particularly

of those lymphocytes participating in adaptive immunity. These changes involve decrease in T-helper cells – with CD4+ lymphopenia – and in B lymphocytes [11–13]. Few clinical studies have reported on lymphocyte subsets in neonates with infection, and those published provide inconsistent results. Sofatzis et al. [20] found lower mean CD3+, CD4+, CD18 and CD11a and CD4+/CD8+ ratio in 20 preterm and term neonates with sepsis, compared with 23 healthy control subjects, while Juretićet al. [21] also showed that preterm neonates with sepsis have lower CD3+ and CD4+ than uninfected premature neonates. Aygun et al. [22] found CD3+, CD4+ and CD8+ in 12 neonates with proven sepsis similar to controls in absolute numbers, but a lower percentage of total lymphocytes and CD4+. Conversely, Kotiranta-Ainamo et al.