Work

Bedside Thwaites’ score in the diagnosis, cerebro-spinal fluid biochemistry in predicting severity and outcomes of TB meningitis in Sri Lanka

Year: 2025

Type: article

Abstract: Tuberculous meningitis (TBM) poses a significant diagnostic challenge as detection of Mycobacterium tuberculosis (MTB) in cerebrospinal fluid (CSF) is difficult. Thwaites’ clinical criteria based on age, length of illness, WBC count, CSF neutrophil proportion has been utilized in the prediction of TBM. CSF biochemistry has been shown to predict disease severity and treatment outcome. Patients treated as TBM at central chest clinic, Sri Lanka over past two-years, were recruited. Thwaites’ criteria were applied to assess bedside prediction. CSF biochemistry (CSF-B) was categorized as Typical: Protein 50-500 mg/dl, cells 50–500/mm3, CSF/Plasma glucose concentration <50%; and, values above (Increased) or below (Decreased). Admission CSF-B were correlated with three British Medical Research Council-TBM severity grades (BMC-TBM-SG) within the cohort. Post 3-month-intensive-phase-treatment (P3MnIPRx) CSF-B were correlated to clinical outcome using modified Ranking Scale(mRS). Sensitivity of CSF-TB culture, TB-geneXpert, and likely reasons for their negative yield were evaluated. There were 87 TBM patients: 20-Definite, 67-highly probable; 59% males, mean (SD) age 45yrs (±20.1). Twaites system had 97.7% diagnostic sensitivity. On admission BMC-TBM-SG were 31 in 1, 34 in 2, and 22 in 3. CSF cells: typical in 53%, low in 35%, increased in 12%. None of CSF protein, CSF/Plasma glucose ratio:<50%, Lymphocytic reaction >50% of total CSF-White cells] correlated with BMC-TBM-SGs (table 1). Positive CSF-AFB smear in 2.5%, CSF-TB culture in 16.3% and CSF-TB-geneXpert in 10.2% (9/87). Mean(SD) sample volume of CSF received for CSF-TB-GeneXpert was 0.61ml (±0.28). In 31, Mean (SD), Adenosine deaminase (ADA) pre-treatment was 10.64 (±13.87), and in 18 P3MnIPRx ADA was 9.6 (±6.5). 38 patients had second CSF-B after 68 days (median) (range: 46-82). Compared to 1st CSF-B, Protein (mg/dl) <50 in 26%, 50-500 in 71%, and >500 in 3% (p=0.32), hypoglycorrhachia in 68%. Change in total no of cells (p = 0.006) and CSF/plasma glucose (p = 0.009). Treatment outcome: mRS 0-2 in 59%, mRS >3 in 24%, and mortality was 17.2%. Following P3MnIPRx clinical outcome was significantly correlated with post-treatment CSF/Plasma glucose ratio (p = 0.043) and not to pre-treatment CSF/plasma glucose ratio (p = 0.361). Thwaites’ criteria appears promising in early bedside prediction of TBM. CSF was typical in 34.4% and atypical in majority as described by Patel et al (2010) & Thomas et al (2023). While pre-treatment CSF-B didn't correlate with severity or clinical outcome, post-treatment CSF/Plasma glucose correlated with the clinical outcome at three months intensive phase treatment. The low sensitivity of CSF-TB-geneXpert indicated probable inadequacy of 0.6 ml uncentrifuged CSF. Thwaites’ criteria seems helpful for bedside prediction of TBM. Pre-treatment CSF-B didn't correlate with the severity, or outcome of treatment. Follow up CSF hypoglycorrhachia correlated with post-intensive-phase-treatment outcome. Uncentrifuged, low CSF volume is likely to reduce sensitivity of CSF-TB-GeneXpert.

Source: International Journal of Infectious Diseases

DOI: 10.1016/j.ijid.2024.107451

Authors: Dr Nimasha Ekanayaka, Dr Bimsara Senanayake, Dr Amitha Fernando, Dr Nilanthi Senanayake, Dr Chandima Hettiarachchi

Institutions: National Hospital of Sri Lanka, University of Colombo

Language: English

Topic: Infectious Diseases and Tuberculosis

Subfield: Surgery

Field: Medicine

Domain: Health Sciences

SDG: Good health and well-being

Open Access status: gold

Estimated APC paid: $2,900