Optimized molecular resolution of cross-contamination alerts in clinical mycobacteriology laboratories

• Published in: BMC microbiology 2008-02, Vol.8 (1), p.30-30, Article 30
• Main Authors: Martín, Ana, Herranz, Marta, Lirola, Miguel Martínez, Fernández, Rosa Fernández, Bouza, Emilio, García de Viedma, Darío
• Format: Article
• Language: English
• Subjects: Diagnosis; Diagnostic Errors; False Positive Reactions; Humans; Interspersed Repetitive Sequences; Methods; Minisatellite Repeats; Mycobacterium tuberculosis; Mycobacterium tuberculosis - genetics; Mycobacterium tuberculosis - isolation & purification; Polymerase Chain Reaction - methods; Polymerase Chain Reaction - standards; Polymorphism, Restriction Fragment Length; Prospective Studies; Restriction fragment length polymorphism analysis; Risk factors; Specimen Handling - methods; Specimen Handling - standards; Tuberculosis; Tuberculosis - diagnosis; Tuberculosis - microbiology
• ISSN: 1471-2180; 1471-2180
• DOI: 10.1186/1471-2180-8-30

The phenomenon of misdiagnosing tuberculosis (TB) by laboratory cross-contamination when culturing Mycobacterium tuberculosis (MTB) has been widely reported and it has an obvious clinical, therapeutic and social impact. The final confirmation of a cross-contamination event requires the molecular identification of the same MTB strain cultured from both the potential source of the contamination and from the false-positive candidate. The molecular tool usually applied in this context is IS6110-RFLP which takes a long time to provide an answer, usually longer than is acceptable for microbiologists and clinicians to make decisions. Our purpose in this study is to evaluate a novel PCR-based method, MIRU-VNTR as an alternative to assure a rapid and optimized analysis of cross-contamination alerts. MIRU-VNTR was prospectively compared with IS6110-RFLP for clarifying 19 alerts of false positivity from other laboratories. MIRU-VNTR highly correlated with IS6110-RFLP, reduced the response time by 27 days and clarified six alerts unresolved by RFLP. Additionally, MIRU-VNTR revealed complex situations such as contamination events involving polyclonal isolates and a false-positive case due to the simultaneous cross-contamination from two independent sources. Unlike standard RFLP-based genotyping, MIRU-VNTR i) could help reduce the impact of a false positive diagnosis of TB, ii) increased the number of events that could be solved and iii) revealed the complexity of some cross-contamination events that could not be dissected by IS6110-RFLP.