Appendix 2. Interpretation of Xpert assay results

Invalid/Error/No result

Perform a 2^nd test on a new specimen.

MTB not detected

• Child with presumed PTB: perform a 2^nd test on a new (respiratory or stool) specimen.
• Adult: based on clinical judgment, perform a 2^nd test and/or a culture on a new specimen.

MTB detected

No Rif resistance detected

Perform Xpert MTB/XDR (or LPA or pDST) to detect H resistance (a) Citation a. For all patients if possible, and at least those with high risk of H resistance (patients with previous TB treatment with H, or contact with a TB case resistant to H, or from an area with a prevalence of resistance to H ≥ 3%). ; adjust treatment according to DST.

MTB detected

Rif resistance detected

Evaluate risk factors for rifampicin resistance (RR):

• High risk of RR (b) Citation b. Patients with previous TB treatment with R, or contact with a TB case resistant to R, or from an area of high prevalence of resistance to R. : treat for MDR/RR-TB.
• Low risk of RR (c) Citation c. Patients with no previous TB treatment with R, or contact with a TB case resistant to R, and from an area of low prevalence of resistance to R. : perform a 2^nd test on a new specimen (d) Citation d. A 2^nd test is necessary because in a population with a prevalence of resistance to rifampicin < 5%, the positive predictive value of one test is < 80%, i.e. > 20% of rifampicin resistant results are false positive. . If 2^nd test shows:
  • R susceptibility: treat for DS-TB.
  • R resistance: treat for MDR/RR-TB.

For patients with MDR/RR-TB, perform:

• Xpert MTB/XDR (or LPA) and pDST (and genome sequencing if available) for resistance to other TB drugs.
• If discordant results with pDST (R resistance with Xpert, R susceptibility with pDST): treat for MDR/RR-TB [1] Citation 1. World Health Organization. Xpert MTB/RIF implementation manual: technical and operational ‘how-to’; practical considerations. Geneva 2014. https://apps.who.int/iris/bitstream/handle/10665/112469/9789241506700;jsessionid=44788E3067C3F9DBF836E8BB6BB0F253?sequence=1 .

MTB detected

Rif resistance indeterminate

• Xpert MTB/RIF:
  • Perform a 2^nd test on a new specimen. If still "indeterminate", treat for DS-TB while investigating RR.
  • Perform pDST (or other gDST) to confirm or rule out RR.
  • Perform Xpert MTB/XDR (or LPA or pDST) to detect H resistance (a) Citation a. For all patients if possible, and at least those with high risk of H resistance (patients with previous TB treatment with H, or contact with a TB case resistant to H, or from an area with a prevalence of resistance to H ≥ 3%). .
• Xpert MTB/RIF Ultra:
  • Send an extraction of the raw results (gxx file) to a reference laboratory for identification of possible mutations (interpretation of melting curves).
  • If not feasible or still "indeterminate": proceed as for Xpert MTB/RIF.

MTB detected "trace"

Rif resistance indeterminate

(Xpert Ultra)

• Patients with HIV infection, children and EP specimens: a "trace" result should be considered as positive.
• Adults with history of TB in the previous 5 years: a "trace" result cannot be interpreted, culture should be performed.
• No interpretation of RR is possible.
• If suspected resistance to R or other TB drugs: perform pDST (or other gDST). Adjust treatment according to DST.  
• Do not test the specimen with Xpert MTB/XDR as the Xpert MTB/XDR has a higher detection limit than Ultra.

Invalid/Error/No result

Perform a 2^nd test on a new specimen.

MTB detected

After a positive Xpert MTB/RIF, an "MTB detected" result is expected because Xpert MTB/XDR and Xpert MTB/RIF have similar detection limit.

MTB not detected

No resistance detected

• After a positive Xpert MTB/RIF: perform a 2^nd test on a new specimen. If the 2^nd test is negative, perform culture and pDST.
• After a "trace" result with Ultra, a negative result is expected because the Xpert MTB/XDR has a higher detection limit than Ultra.

MTB detected

No resistance detected

• Treat according to the result of Xpert MTB/RIF or Ultra.
• The test does not rule out all resistances (some resistance-conferring mutations are not detected by Xpert MTB/XDR, e.g. only 30% of Eto resistance conferring mutations are detected).

MTB detected

• Low INH resistance detected
• INH resistance detected
• Low FLQ resistance detected
• FLQ resistance detected
• ETH resistance detected
• AMK, KAN and/or CAP resistance detected

Evaluate risk factors of resistance for each drug:

• High risk of resistance (e) Citation e. Patients with previous TB treatment with the drug or contact with a TB case resistant to the drug, or from an area of high prevalence of resistance to the drug. : consider as resistant to the drug.
  • If low-level H resistance detected (inhA mutation and no katG mutation): H^h can be used, but not counted as a likely effective drug.
  • If low-level resistance to FQs detected: Mfx^h can be used, but not counted as a likely effective drug [2] Citation 2. World Health Organization. WHO operational handbook on tuberculosis. Module 4: treatment - drug-resistant tuberculosis treatment. Geneva 2020.
    https://www.who.int/publications/i/item/9789240006997 .
  • Resistance to Eto can be detected (inhA mutation). However, a negative result does not rule out resistance.
  • Perform pDST for resistance to other TB drugs.
• Low risk of resistance (f) Citation f. Patients with no previous TB treatment with the drug or contact with a TB case resistant to the drug, and from an area of low prevalence of resistance to the drug. : perform a 2^nd test on a new specimen (g) Citation g. A 2^nd test is necessary because in a population with a prevalence of resistance < 5%, the positive predictive value of one test is < 80% (i.e. > 20% of resistant results are false positive). . If the 2^nd test shows:
  • Drug susceptibility: treat with the drug.
  • Drug resistance: consider as resistant (see above for "High risk of resistance").

MTB detected

Drug resistance indeterminate (h) Citation h. No "indeterminate" result is given for Eto.

Perform a 2^nd test on a new specimen. If still "indeterminate": treat with likely effective drug(s) while investigating resistance with pDST (or other gDST, e.g. second-line LPA, genome sequencing).