Conclusion
Our findings do not support the hypothesis that pretreatment levels of MMP9 would be higher in tuberculous meningitis patients experiencing a poor treatment outcome. Further, MMP9 levels in the CSF did not correlate with blood-brain barrier permeability in patients with tuberculous meningitis.
Methods
We prospectively assessed the treatment outcome in a consecutive sample of human immunodeficiency virus-negative patients with tuberculous meningitis. We defined good outcome as survival without severe neurological disability (modified Rankin scale scores 0-2). We estimated levels of MMP9 and its tissue inhibitor (TIMP1) on pretreatment CSF samples. We used albumin index to assess blood-brain barrier permeability.
Results
We studied 40 patients (23 males [58%]) with tuberculous meningitis. Sixteen patients (40%) had stage 3 disease. On follow-up, 18 (45%) patients had a poor treatment outcome-15 patients died and 3 had severe neurological disability. Pretreatment MMP9 levels were not associated with treatment outcome (median [interquartile range], 254 [115-389] vs. 192 [60-383] ng/mL in good vs. poor outcome groups; P = 0.693). MMP9 levels did not correlate with the albumin index (Spearman's rho = 0.142; P = 0.381). However, MMP9 levels significantly correlated with CSF glucose levels (rho = -0.419; P = 0.007) and admission Glasgow coma scale score (rho = 0.324; P = 0.032). Likewise, TIMP1 levels also did not differ by treatment outcome (1239 [889-1511] vs. 1522 [934-1949] ng/mL; P = 0.201). MMP9/TIMP1 ratio that reflects net proteolytic activity was also not different between the two groups (0.191 [0.107-0.250] vs. 0.163 [0.067-0.34]; P = 0.625).