Spanos [8] described a study in 55 patients with BM and a negative direct CSF examination. The median neutrophil count in the CSF of these patients was 512/mm3, compared with 1,520/mm3 in the CSF of 134 patients with BM and Lenalidomide supplier a positive direct CSF examination; the difference between these groups was statistically significant. Ray et al. [17] reported a mean neutrophil count of 428/mm3 in 18 patients with BM and a negative direct CSF examination. In this study, neutrophil count had a sensitivity of 78% and a specificity of 75% for the differential diagnosis of BM. In our study, this parameter was among those most poorly discriminating between BM and VM, with a sensitivity of 80% and a specificity of 85% at a diagnostic cut-off level of 118/mm3.Low CSF glucose levels (2 to 2.

5 mmol/L) and a low CSF/serum glucose ratio (on the order of 0.30 to 0.40) have been classically described as a feature of BM [18-22]. Spanos [8] reported a median CSF glucose level of 3.4 mmol/L (CSF/serum glucose ratio, 0.45) in patients with BM and a negative direct examination of the CSF, compared with 1.7 mmol/L (CSF/serum glucose ratio, 0.23) in 134 patients with BM and a positive direct CSF examination. Similar values were seen in the patients included in our study, although CSF glucose levels were more markedly reduced.Elevated CSF protein levels are also seen in patients with BM, mean values ranging from 1 to 2.5 g/L [8,17,21,23]. With regard to the differential diagnosis of BM, the reported sensitivity of this parameter ranged from 63% to 86% with a specificity of 94% to 98% [17,21].

Normal CSF protein values were found in 10% of patients with BM and a negative direct CSF examination [8]. In our study, a CSF protein level of 1.88 g/L gave a sensitivity of 89% and a specificity of 93% for the diagnosis of BM. Normal levels of protein in the CSF were seen in 5% of the patients with BM in our study.Although several studies have demonstrated the value of determining lactate levels in the CSF for the differential diagnosis of BM, none of these focused specifically on the diagnostic value of this parameter in adult patients with BM and a negative direct CSF examination [21,23-27]. Furthermore, its use in this context was not recommended by the last-but-one French consensus conference on the management of BM, notably on the grounds of the disparity of the diagnostic cut-off levels reported and the methods used to determine these [28].

In addition, the selection of the control groups compared with the BM group could be a confounding factor with regard to interpretation Drug_discovery of the results. In some studies, the BM group was compared with heterogeneous groups of patients with pathologic conditions likely to induce elevated lactate concentrations in the CSF (for example, brain tumor, subdural hemorrhage, trauma, and coma of metabolic origin) [19,27].