3.10 Biopsies, laboratory tests on body fluids and other biological tests

3.10.1 Biopsies and fine needle aspirate cytology (FNAC)

Biopsies of lymph nodes, bone and pleural lining are often not feasible in resourceconstrained settings given the technical skill and laboratory resources required. The cytology of the lymph nodes from FNAC is easier to perform. Specific granulomatous tissue, the presence of giant Langhans' cells, and/or caseous necrosis strongly correlate with TB. AFBs are not always present. For the procedure for FNAC, see Appendix 4.

Note: molecular tests can be used on the specimens obtained from FNAC of lymph nodes.

3.10.2 Laboratory tests on body fluids

The diagnosis of some EPTB localisations can be supported or confirmed by a combination of tests performed in respective body fluids.

Table 3.2 - Summary of findings suggestive of TB in body fluids



Ascitic fluid

  • Typically translucent yellow-coloured liquid.
  • Exudate rich in lymphocytes, usually > 300 white cells/mm3; Rivalta test positive (Appendix 5).
  • Serum-ascites albumin gradient (SAAG) < 1.1 g/dl is consistent with TB (and many other conditions) while a SAAG > 1.1 g/dl makes peritoneal TB unlikely31.
  • Adenosine desaminase can be used as a surrogate marker for TB in peritoneal fluid (Appendix 6).
  • The search for M. tuberculosis by microscopy is most often negative.

Pleural fluid

  • Typically straw-coloured.
  • Proteins ≥ 30 g/l (Rivalta test, Appendix 5).
  • Rich in white cells (1,000-2,500/mm3), with predominant lymphocytes
  • Adenosine desaminase can be used as a surrogate marker for TB in pleural fluid (Appendix 6).
  • Microscopy for M. tuberculosis is most often negative.
  • Xpert MTB/RIF in pleural fluid has a moderate sensitivity, and therefore, is not recommended.

Cerebrospinal fluid

  • Clear, hyper-concentrated liquid.
  • Proteins > 0.40 g/l (Pandy test, Appendix 5).
  • Glucose diminished: < 60 mg/l.
  • CSF glucose/blood glucose < 0.5.
  • Between 100 and 1,000 white blood cells/ml, of which over 80% are lymphocytes.
  • M. tuberculosis can be found by CSF direct microscopy in less than 10%.
  • Xpert MTB/RIF has a moderate sensitivity that can be increased following centrifugation. Centrifugation is recommended if facilities for efficient and safe centrifugation exist (high-speed centrifuge and biosafety cabinet).
  • In HIV+ patients, cryptococcal meningitis is a concern. Perform the antigen test with cryptococcal antigen on serum and CSF (CrAgLFA).


  • A culture or molecular testing, after centrifugation, are the only measures to confirm diagnosis.
  • The search for M. tuberculosis in urinary microscopy is almost always negative.
  • Xpert MTB/RIF has a moderate sensitivity. Priority should be given to patients with CD4 counts < 50 due to demonstrated higher sensitivity in this group32.
  • The LAM assay is useful in patients with CD4 < 200 (Section 3.10.3).

3.10.3 Other biological examinations

New TB diagnostic tests are in development for point-of-care use. These antigen-detection assays are based on detecting liporabinomannan (LAM): a carbohydrate cell wall antigen that is excreted in the urine of TB patients. The performances of the LAM urine assay for most populations are poor. An exception is the sensitivity of the LAM assay in patients with CD4 counts < 20033,34,35. The test may have some utility where advanced HIV-associated immunodeficiency is common.

Sedimentation rate is almost always higher but this examination is very non-specific. A normal sedimentation rate makes TB less likely but still possible.

C-reactive protein is also generally increased but this test also is very non-specific.

There exist commercial rapid blood tests for “serological diagnosis of TB”, but they are so far not very reliable in diagnosing active TB and should not be used.