Routine Culture–Resistant Mycobacterium tuberculosis Rescue and Shell-Vial Assay, France

We used shell-vial assay with a medium that buffered rifampin to isolate routine culture–resistant Mycobacterium tuberculosis bacteria from cerebrospinal fluid and rifampin-containing intervertebral disc and vertebral corpus of a patient in treatment for Pott’s disease and disseminated tuberculosis. Whole-genome sequencing confirmed M. tuberculosis lineage 4 (Euro-American) strain.

C ulturing Mycobacterium tuberculosis from clinical specimens confirms the viability of mycobacteria and enables drug susceptibility testing (1). Routinely used culture protocols may fail to isolate M. tuberculosis from vertebral biopsy specimens in 17%-50% of cases (2) and from cerebrospinal fluid (CSF) specimens in >80% of cases (3). However, the shell-vial culture assay (4) has demonstrated a high sensitivity for the isolation of mycobacteria, representing an alternative method for growing M. tuberculosis (5,6).
We report a case of disseminated tuberculosis (TB) documented by culturing M. tuberculosis strain P7739 from a patient who was previously treated with antituberculous drugs. We used the shell-vial assay, even though strain P7739 resisted standard cell-free culture techniques. The Ethics Committee of Institut Hospitalier Universitaire Méditerranée Infection (Marseille, France) approved this study (no. 2016-024, October 19, 2016).
A 47-year-old man who was HIV negative and had no previous history of TB received a diagnosis of Pott's disease with systemic tuberculosis on the basis of clinical symptoms of spondylodiscitis, myelitis, meningitis, and pulmonary miliary infection. In January 2017, the patient suffered lumbar pain; a computer tomodensitometry scan showed corporeal bone defects in the L1 vertebra. Ten months later, magnetic resonance imaging revealed T12-L1 vertebral spondylodiscitis with a paravertebral abscess in the right iliopsoas muscle. We performed 2 biopsies of the vertebral corpus of T12 and L1 and 1 biopsy of the intervertebral disc; we also collected CSF and sputum. These clinical specimens remained negative for M. tuberculosis using microscopic examination after Ziehl-Neelsen staining, real-time PCR (GenExpert, https://www.cepheid. com), culture in liquid medium BBL Mycobacteria Growth Indicator Tube (Becton Dickinson, https://www.bd.com), and in solid culture media including Coletsos medium (bioMérieux, https://www.biomerieux.com). Thirteen days later, the neurologic condition of the patient deteriorated with meningeal syndrome. Examination by magnetic resonance imaging showed a triventricular hydrocephalus and transependymal periventricular resorption, which led to an emergency external ventricular bypass. We strongly suspected TB, so we introduced TB treatment (900 mg rifampin, 300 mg isoniazid, 1,800 mg pyrazinamide, and 1,200 mg ethambutol daily). We performed a second round of clinical sampling from bronchoalveolar fluid, T12 vertebral corpus and T12-L1 intervertebral discs, and CSF 9 days after the initiation of TB treatment. Microscopic examination after Ziehl-Neelsen staining remained negative, as did results from the GenExpert assay, except for the detection of a rifampin-susceptible M. tuberculosis complex mycobacterium in 3 vertebral biopsy specimens and the intervertebral disc specimen. Culturing in MGIT tubes and on Coletsos remained negative after 8 weeks of incubation.
Gouriet et al. reported that 11.5% of Mycobacterium sp. isolates are cultured only in a cell culture assay (6). We cultured M. tuberculosis using the shell-vial assay after incubating for 17-28 days, in accordance with previous studies (5). Our findings confirm the use of the shell-vial assay in diagnosis of tuberculosis in patients in whom TB is suspected and whose specimens do not grow on conventional media, especially after initiating TB treatment in patients before sampling. Our observations suggest that cell culture medium buffers the anti-TB activity of clinical specimens, thus enabling growth of M. tuberculosis mycobacteria that are no longer exposed to anti-TB activity.

About the Author
Dr. Fellag is a veterinarian and a PhD candidate at Aix-Marseille University, Marseille, France. His research interests include the diagnosis and the natural history of tuberculosis in humans and animals.