Discrepancies in Xpert tuberculosis testing

The announcement, in 2010,1Boehme CC Nabeta P Hillemann D et al.Rapid molecular detection of tuberculosis and rifampin resistance.N Engl J Med. 2010; 363: 1005-1015Crossref PubMed Scopus (1618) Google Scholar of a rapid, accurate nuclear acid amplification test to diagnose tuberculosis was a remarkable leap forward for tuberculosis diagnostics. The GeneXpert tuberculosis assay (Cepheid, Sunnyvale, CA, USA) provides a result in 2 h, is both highly sensitive and specific, and is a major improvement over conventional smear microscopy. Moreover, simultaneous testing for rifampicin resistance guides immediate and appropriate treatment for patients with rifampicin, or more extensive, resistance. Before the introduction of rapid molecular assays, the turnaround time for laboratory phenotypic sensitivity results was 6–8 weeks. This time lag sometimes led to inadequate or excessive treatment and, because appropriate treatment was delayed, community transmission continued, with worsened patients' outcomes, particularly among those with rifampicin-resistant tuberculosis.2Chaisson LH Duong D Cattamanchi A et al.Association of rapid molecular testing with duration of respiratory isolation for patients with possible tuberculosis in a US hospital.JAMA Intern Med. 2018; 178: 1380-1388Crossref PubMed Scopus (17) Google Scholar, 3Getnet F Demissie M Worku A et al.Delay in diagnosis of pulmonary tuberculosis increases the risk of pulmonary cavitation in pastoralist setting of Ethiopia.BMC Pulm Med. 2019; 19: 201Crossref PubMed Scopus (6) Google Scholar, 4Chen Y Yuan Z Shen X Wu J Wu Z Xu B Time to multidrug-resistant tuberculosis treatment initiation in association with treatment outcomes in Shanghai, China.Antimicrob Agents Chemother. 2018; 62: e02259-e02317Crossref PubMed Scopus (14) Google Scholar However, in The Lancet Microbe, Jean Claude Semuto Ngabonziza and colleagues5Ngabonziza JCS Decroo T Migambi P et al.Prevalence and drivers of false-positive rifampicin-resistant Xpert MTB/RIF results: a prospective observational study in Rwanda.Lancet Microbe. 2020; 1: e74-e83Summary Full Text Full Text PDF Scopus (11) Google Scholar suggest that the Xpert MTB/RIF (Xpert) assay might be a victim of its own success. Earlier tuberculosis diagnosis attributable to widespread availability of Xpert testing seems to identify patients with less advanced disease, who have lower bacterial density per unit of sputum volume than do those with more advanced disease, but Ngabonziza and colleagues' data suggest that lower bacillary loads greatly reduce the ability of the Xpert assay to accurately diagnose rifampicin-resistant tuberculosis. Their study5Ngabonziza JCS Decroo T Migambi P et al.Prevalence and drivers of false-positive rifampicin-resistant Xpert MTB/RIF results: a prospective observational study in Rwanda.Lancet Microbe. 2020; 1: e74-e83Summary Full Text Full Text PDF Scopus (11) Google Scholar was nested within a larger trial of novel diagnostics (DIAMA; NCT03303963), during which discordant results were noted between initial and repeat Xpert testing, prompting the investigation described in their Article. An initial spot sputum specimen was obtained from patients presenting with tuberculosis symptoms at 68 peripheral sites in Rwanda and was subjected to the Xpert assay; patients who had a positive result for rifampicin resistance on the initial Xpert test were referred to one of two central drug-resistant tuberculosis treatment clinics, at which a second Xpert test was done on overnight sputum samples, which usually have a larger volume than spot samples. In patients with discordant initial and repeat Xpert results, a third (spot) sample was used for Sanger sequencing of both the rifampicin resistance determining region and non-rifampicin resistance determining region of the rpoB gene. If this test did not yield a result, it was repeated on culture isolates, when available. Furthermore, next-generation sequencing and whole-genome sequencing were used on sputum DNA extracts or culture isolate DNA to detect rifampicin resistance mutations. Ngabonziza and colleagues report very high rates of discordance between initial and repeat Xpert results; indeed, only 54 (35%) of 154 participants had concordant results. On repeat Xpert assay, 53 (34%) participants were found to be tuberculosis-negative and 46 (30%) rifampicin-sensitive. Similar results were reported by our group in South Africa.6Kenaope L Ferreira H Seedat F Otwombe K Martinson NA Variava E Sputum culture and drug sensitivity testing outcome among Xpert MTB/Rif positive, rifampicin resistant sputa: a retrospective study—not all rifampicin resistance is MDR.J Glob Antimicrob Resist. 2019; (published online Nov 13.)DOI:10.1016/j.jgar.2019.11.008PubMed Google Scholar Discordance in Xpert results was more frequent in samples with a very low bacillary load (adjusted odds ratio 118·4 [95% CI 22·2–631·4] for very low vs high bacillary load). Low bacillary loads could result in inadequate amplification of sequences that are present on specific Xpert probes, thereby leading to failure of attachment and causing false-positive results for rifampicin resistance. 67 (67%) of 100 discordant samples had sequencing results, of which only ten (15%) were found to have confirmed rifampicin resistance. Surprisingly, there was no difference between HIV-infected (who are likely to have negative results) and HIV-seronegative individuals, despite tuberculosis in individuals with HIV co-infection being associated with low Mycobacterium tuberculosis bacterial loads.7Beynon F Theron G Respeito D et al.Correlation of Xpert MTB/RIF with measures to assess Mycobacterium tuberculosis bacillary burden in high HIV burden areas of Southern Africa.Sci Rep. 2018; 85201Crossref PubMed Scopus (14) Google Scholar Ngabonziza and colleagues were unable to sequence 25 (47%) of 53 samples that were tuberculosis-negative on repeat Xpert, and these samples would be false positive for both rifampicin resistance and tuberculosis. Of the 28 samples for which sequencing data were available, only six had evidence of rifampicin resistance, whereas the other 22—almost all of which had very low bacterial loads—were drug susceptible. Having tuberculosis diagnostic waters muddied by overcalling of rifampicin resistance in a high proportion of patients has huge safety, stigma, and cost implications for both patients and health systems. Replacement of the Xpert assay with the next version, Xpert Ultra, could reduce the risk of false-positive rifampicin resistance results, since the updated assay has improved test characteristics, particularly in patients with paucibacilliary tuberculosis.8Dorman SE Schumacher SG Alland D et al.Xpert MTB/RIF Ultra for detection of Mycobacterium tuberculosis and rifampicin resistance: a prospective multicentre diagnostic accuracy study.Lancet Infect Dis. 2018; 18: 76-84Summary Full Text Full Text PDF PubMed Scopus (253) Google Scholar A study similar to Ngabonziza and colleagues' is urgently needed to ascertain if the shortcomings of the Xpert assay reported also affect the Xpert Ultra assay. In the interim, clinicians should be advised that, before starting treatment appropriate for drug-resistant tuberculosis, a high-volume sputum sample should be subjected to a repeat Xpert test, particularly in patients whose initial result is uncertain because of low or trace bacillary loads. If rifampicin resistance is confirmed by the repeat Xpert test, treatment for drug-resistant tuberculosis can be initiated, but additional resistance tests for isoniazid, aminoglycosides, and quinolones will still be required when considering short or long treatment courses for rifampicin resistant tuberculosis.9World Health OrganizationWHO rapid communication: key changes to the treatment of drug resistant tuberculosis.https://www.who.int/tb/publications/2019/rapid_communications_MDR/enDate: 2019Date accessed: January 10, 2020Google Scholar However, with discordant results, either tuberculosis-negative or rifampicin-susceptible, closer individual clinical review is warranted to ascertain whether treatment for multidrug-resistant tuberculosis is needed, and clinicians should be willing to consider de-escalation of therapy if rifampicin susceptibility is subsequently confirmed. An ambitious but not unachievable goal is the adoption of a new single regimen with efficacy against both drug-resistant and drug-susceptible tuberculosis.10Tweed CD Dawson R Burger DA et al.Bedaquiline, moxifloxacin, pretomanid, and pyrazinamide during the first 8 weeks of treatment of patients with drug-susceptible or drug-resistant pulmonary tuberculosis: a multicentre, open-label, partially randomised, phase 2b trial.Lancet Respir Med. 2019; 7: 1048-1058Summary Full Text Full Text PDF PubMed Scopus (38) Google Scholar If achieved, it would remove reliance on resistance testing. Until then, the gradual shift to earlier diagnosis of tuberculosis reported by Ngabonziza and colleagues will continue, and is probably pronounced in settings where high value is placed on symptom screening, contact tracing, and immediate, increased testing for tuberculosis. Studies such as this one by Ngabonziza and colleagues should be repeated, elsewhere. We declare no competing interests. Prevalence and drivers of false-positive rifampicin-resistant Xpert MTB/RIF results: a prospective observational study in RwandaThe Xpert testing algorithm should include an assessment of bacillary load and retesting in case rifampicin resistance is detected on a paucibacillary sputum sample. Only when rifampicin resistance has been confirmed on repeat testing should multidrug-resistant tuberculosis treatment be started. When rifampicin resistance has not been confirmed on repeat testing, we propose that patients should be given first-line anti-tuberculosis drugs and monitored closely during treatment, including by baseline culture, pDST, and further Xpert testing. Full-Text PDF Open Access