Note
Use of INNO-LIPA assay for rapid identification of mycobacteria

https://doi.org/10.1016/S0732-8893(03)00042-7Get rights and content

Abstract

Using 106 clinical isolates of mycobacteria, we showed that INNO-LIPA Mycobacteria assay is an excellent tool to rapidly identify the most frequently isolated nontuberculous mycobacteria, in one procedure. It may be used as an additional technique to AccuProbe assay, which remains the fastest and the cheapest tool for a rapid and accurate identification of the M. tuberculosis complex.

Introduction

Although Mycobacterium tuberculosis remains the most frequent agent of mycobacterial diseases, nontuberculous mycobacteria (NTM) are being recovered with increasing frequency, particularly in respiratory specimens of patients with underlying chronic lung disease, in HIV-infected patients, and in nosocomial infections Astagneau et al 2001, Juffermans et al 1998, O’Brien et al 2000, Olivier et al 2001, Phillips and Fordham von Reyn 2001. Recently, new automated nonradiometric systems, such as the Bactec MGIT 960 (Becton Dickinson, France) and the MB BacT/Alert 3D system (Organon Teknika, France) (Alcaide et al., 2000) have contributed to the more widespread use of cultures in liquid media. For rapid identification, the AccuProbe assay (GenProbe, BioMérieux, France) covers a limited set of species (M. tuberculosis complex, M. avium complex, M. avium, M. intracellulare, M. kansasii, M. gordonae). However, each probe has to be tested in a single assay. The new INNO-LIPA Mycobacteria assay (Innogenetics, Ghent, Belgium) allows the identification of the mycobacterium genus and the above listed species plus three additional species (M. xenopi, M. scrofulaceum, M. chelonae). The INNO-LIPA method is based on the reverse hybridization principle, in which the mycobacterial 16S-23S ribosomal RNA spacer region is amplified by polymerase chain reaction (PCR). Amplicons are subsequently hybridized with oligonucleotide probes arranged on a membrane strip and detected by a colorimetric system. A positive reaction is revealed by clearly visible lines representing the genus and species of mycobacteria. The method allows different probe testing in a single assay. The aim of the present study was to evaluate the value of this new assay in identifying mycobacterial strains compared to Accuprobe assays and/or to classic biochemical tests generally used in a clinical laboratory.

The study was performed in two microbiology laboratories, using 106 cultures of clinical specimens detected positive with the new BacT/Alert 3D system (71 studied retrospectively in lab 1, and 35 studied prospectively in lab 2). Specimens were decontaminated by the N-acetyl-L-cysteine NaOH procedure. For culture, 0.5 mL of sediment was inoculated into the MB/BacT vials and 0.2 mL onto Löwenstein-Jensen slants incubated at 37°C for eight weeks and three months, respectively.

When bottles were flagged positive by the instrument, Ziehl Neelsen staining, and subculture onto Lowenstein-Jensen slants were performed. Aliquots of broth culture were taken for identification, one of them was frozen at –20°C until results were known. For identification by INNO-LIPA, the amplification was carried out using a Perkin-Elmer 9600 thermocycler in the presence of Taq polymerase (Perkin Elmer Cetus, France) following the manufacturer’s recommendations. Amplification products were kept at –20°C. Hybridization was performed automatically in lab 1 (Auto-LIPA, Innogenetics) and manually in lab 2, as described by the manufacturer. For identification by Accuprobe, the choice of specific probes was made on the basis of the morphologic characteristics by microscopic examination especially for M. tuberculosis complex, which is easily recognized by its characteristic cord formation (Badak et al., 1999) and clinical features.

All of the 106 specimens studied grew mycobacteria in the MB BacT/Alert media. The average number of days required for the detection of mycobacteria with the BacT/Alert 3D system was 13.5 (range 6-28) for M. tuberculosis, 6 (range 3-9) for the M. avium complex, and 36 (range 30-47) for M. xenopi isolates. The manufacturer’s protocol concerning the delay of incubation of 6 weeks was found to be insufficient for one clinically significant isolate of M. xenopi which was detected only after 47 days of incubation. However, when compared to the isolates obtained on the Löwenstein-Jensen slants, no M. xenopi strain was missed, suggesting that the new algorithm developed for the MB BacT/Alert 3D system indeed improves the detection of slow-growing species such as M. xenopi (Alcaide et al., 2000).

LIPA amplicons were detected in 105 of the 106 patient samples after the first amplification. The negative specimen was positive when the test was repeated on the frozen aliquot. After the hybridization step, the line probe 2 specific for the genus mycobacteria was positive in all cases. Concordant results with usual identification tests were obtained for 59/59 isolates of M. tuberculosis complex, 25/25 isolates of M. xenopi, 6/6 isolates of M. gordonae, 1/1 isolate of M. chelonae and 14/15 isolates of MAIS complex (M. avium, M. intracellulare, M. scrofulaceum, intermediates, M. malmoense and M. hoemophilum), at the complex level (probe 9) (Table 1). Twelve (80%) of these 15 MAIS complex isolates were accurately identified at the species level with the LIPA (probe 10 for M. avium or probe 11 for M. intracellulare) and the Accuprobe M. avium or M. intracellulare specific assays: seven as M. avium and five as M. intracellulare.

Some discrepancies were observed with results obtained with LIPA-MAIS complex (probe 9) and AccuProbe M. avium complex assay. Three isolates (20%) from three different patients were only reactive with the LIPA-MAIS complex (probe 9). Two of them were accurately identified as M. intracellulare by the Accuprobe assay. These results confirm the minor discrepancies between INNO-LIPA and Accuprobe assays in identifying M. intracellulare at the species level, as recently reported by others Miller et al 2000, Scarparo et al 2001, Tortoli et al 2001. However, contrary to Accuprobe, no misidentification was noted with the INNO-LIPA assay. Indeed, the strain found negative with the M. avium complex and the M. avium and M. intracellulare specific Accuprobe assays (Table 1) was identified as M. intracellulare serotype 7 by the 16S rRNA gene sequence determination, according to the method developed by Kirschner et al., 1993. This clinically relevant strain was isolated from the bone marrow of a transplant patient.

At the time of a positive signal in MB/BacT bottles, all 106 cultures contained enough organisms to enable observation of acid fast bacilli and the possible presence of cord formation by microscopy (Badak et al., 1999). In our experience, on the basis of this morphologic characteristic, the selection of the appropriate Accuprobe was always successful for the identification of M. tuberculosis complex. Accuprobe assay, which is easier to perform, (2 h vs 6 h for INNO-LIPA) is cheaper and faster than INNO-LIPA for a rapid and accurate identification of M. tuberculosis complex in a routine clinical laboratory. Without specific morphologic characteristics to guide probe selection testing, multiple specific AccuProbe tests may be necessary. Moreover, the Accuprobe assay is available for only a limited number of species. In our study, INNO-LIPA Mycobacteria correctly identified all NTM isolates at the species or complex level, while the AccuProbe assay failed to identify one M. intracellulare even at the M. avium complex level.

The availability of a M. xenopi probe in the INNO-LIPA assay represents an important improvement. Indeed, the presumptive identification by incubating vials at three different temperatures with the radiometric system Koemoth et al 1991, Marx et al 1995 cannot be performed with the new automated nonradiometric systems. M. xenopi is an emerging pathogen in HIV-infected patients, in non HIV-infected patients with underlying chronic lung disease, and in nosocomial infections, causing colonization and pseudo-outbreaks Astagneau et al 2001, Juffermans et al 1998, Phillips and Fordham von Reyn 2001. This slow-growing species, which is not easily and rapidly identified by classic biochemical tests (Marx et al., 1995), represented 29.2% of our isolates. A rapid and accurate diagnosis of infections is extremely useful when selecting the most appropriate therapeutic agents.

In conclusion, if the morphologic characteristics observed by microscopic examination and/or the clinical features favor tuberculosis, we suggest first using the AccuProbe M. tuberculosis complex assay. The INNO-LIPA assay could then be used in a routine clinical laboratory skilled in molecular biology techniques as a complementary assay, to rapidly identify the most frequently isolated NTM in a single procedure.

Section snippets

Acknowledgements

The authors thank Organon Teknika and Innogenetics for providing us respectively with the BacT/Alert 3D media and the Auto-Lipa system; Sandrine Rozé, and François Casanova for technical assistance.

References (13)

  • P. Astagneau et al.

    Mycobacterium xenopi spinal infections after discovertebral surgeryinvestigation and screening of a large outbreak

    Lancet

    (2001)
  • C.E. Marx et al.

    Laboratory and clinical evaluation of Mycobacterium xenopi isolates

    Diagn Microbiol Infect Dis

    (1995)
  • F. Alcaide et al.

    Evaluation of the Bactec MGIT 960 and the MB/BacT systems for recovery of mycobacteria from clinical specimens and for species identification by DNA Accuprobe

    J Clin Microbiol

    (2000)
  • F.Z. Badak et al.

    Cord formation in MB/BacT medium is a reliable criterion for presumptive identification of Mycobacterium tuberculosis complex in laboratories with high prevalence of M. tuberculosis

    J Clin Microbiol

    (1999)
  • N.P. Juffermans et al.

    Mycobacterium xenopi in HIV-infected patientsan emerging pathogen

    AIDS

    (1998)
  • P. Kirschner et al.

    Genotypic identification of mycobacteria by nucleic acid sequence determinationreport of two-year experience in a clinical laboratory

    J Clin Microbiol

    (1993)
There are more references available in the full text version of this article.

Cited by (19)

  • Mycobacterium Nontuberculosis Species

    2018, Principles and Practice of Pediatric Infectious Diseases
  • Mycobacterium Species Non-tuberculosis

    2012, Principles and Practice of Pediatric Infectious Diseases, Fourth Edition
  • New methods for mycobacteria identification

    2006, Enfermedades Infecciosas y Microbiologia Clinica
View all citing articles on Scopus
View full text