DNA Sequencing of MDR TB in Eastern Siberia

The multidrug-resistant tuberculosis (MDR-TB) epidemic in the Irkutsk oblast of Eastern Siberia is arguably the most severe in the world. The investigators have found that in the Irkutsk TB Dispensary, the referral hospital for the oblast (state), patients with HIV and primary MDR-TB (no prior history of TB treatment) suffered poor outcome during the inpatient phase of treatment. It was also found that genotypic drug-resistance mutations only partially explain the extent of phenotypic resistance that is so widespread in Irkutsk. Suboptimal serum anti-TB drug concentrations amplify such phenotypic resistance and work in other settings has demonstrated the importance of optimizing pharmacokinetics to improve TB outcome. Such foundational work in anti-TB/antiretroviral drug concentrations is lacking for drugs used in the treatment of MDR-TB, and absent in the HIV-infected population. Furthermore, the investigators have found that HIV infection in Irkutsk was associated with the Beijing TB genotype similar to other high burden settings, and that overall mortality was more frequent in those with the M. tuberculosis sublineage Beijing MIT17. Thus taken together the investigators hypothesize that the high early TB mortality in Irkutsk is due to a combination of complex MDR and extensively drug resistant (XDR)-TB that is incompletely diagnosed and incompletely treated, poor anti-TB/antiretroviral pharmacokinetics that lead to low efficacy and increased circulation of hyper-transmissible and virulent M. tuberculosis sublineages.

For subjects suspected of TB, sputum samples or other leftover sputum/blood specimens will be screened by GeneXpert and if positive, cultured for TB and the cultured specimen subjected to drug-susceptibility testing by conventional qualitative resistance and minimum inhibitory concentration (MIC), sequencing for drug-resistance mutation and spoligotyping for sublineage identification. Primary analysis will include standard sensitivity/specificity of each drug-mutation compared to conventional qualitative resistance and then median/range MIC values among isolates with/without mutation for improved discrimination. Sublineage analysis will be by mycobacterial interspersed repetitive unit/variable-number tandem repeats (MIRU24-VNTR) and correlated with drug-resistance by conventional susceptibility, MIC and the sequencing results for drug-resistance mutations. The transmission capacity of sublineages will be evaluated by the level of their clustering via the 24-loci MIRU typing.

For subjects ultimately initiating TB treatment, the pharmacokinetic parameters to be tested against the MIC for each TB drug in the patient's regimen include peak (Cmax) and area under the concentration-time curve (AUC). Parameters will be compared to the expected ranges for each drug and expressed as both a population value and at the individual level (proportion of patients below the expected range for each drug). The clinical outcome of TB treatment failure will be compared to the proportion of subjects with a baseline (2 week) Cmax/MIC below the minimal target (lowest µg/ml concentration/ highest MIC of an isolate still considered susceptible) for 2 key drugs in the regimen. It is anticipated that TB treatment failure to be more frequent in subjects treated for drug-susceptible TB when Cmax/MIC values are lower than target for both rifampin and isoniazid, and in subjects treated for MDR-TB when values are below the minimal target for both ofloxacin and kanamycin. A binary logistic regression model will be used to determine the possible risk factors to TB treatment failure in addition to the pharmacokinetic parameters.

The investigators aim for drug-resistance and sublineage analysis to be performed on M. tuberculosis isolates from 250 subjects, 200 of whom will undergo pharmacokinetic study.

Between the Irkutsk Dispensary and the Irkutsk AIDS Center, it is estimated that more than 700 subjects would be eligible. The majority enrolled from the Irkutsk Dispensary will have M. tuberculosis positive specimens (anticipated ~175 total enrolled) and complete pharmacokinetic analysis. It is expected that fewer subjects (anticipated ~25) of those enrolled from the referral Irkutsk AIDS Center will commence TB treatment at Irkutsk Dispensary and complete pharmacokinetic analysis. Sample size is restricted by the exploratory/observational focus of the R21 funding source, and thus estimated upon subjects recruited/enrolled over a 12 month period.

All microbiologic testing will be performed onsite in the TB laboratory at the Irkutsk Dispensary. DNA samples will be further analyzed at the Institute Epidemiology and Microbiology/Russian Academy of Medical Sciences in Irkutsk. The subset of subjects positive for M. tuberculosis and initiating TB treatment at the Irkutsk Dispensary will have pharmacokinetic sampling (for research) for plasma anti-TB drug concentrations at 2 weeks and 12 weeks from the initiation of anti-TB treatment (see outline below). If applicable, antiretroviral concentrations (efavirenz or nevirapine) will also be performed. In addition, all subjects will have a structured interview and review of the hospital/clinic chart for the following data: demographics (age, sex, country/region of origin), prior TB history including prior TB treatment regimens, site of TB disease (pulmonary, central nervous system (CNS), lymph node, bone/joint, genitourinary or other), medical comorbidities other than HIV (diabetes, chronic kidney disease, liver disease, smoking, alcohol or injection drug use), current chest radiograph abnormalities, and baseline laboratory values (CBC, chemistries, liver function tests, erythrocyte sedimentation rate (ESR) and if applicable most recent CD4+ T cell count and HIV-1 RNA level). For the subset initiating anti-TB treatment, follow-up chart review will be performed at the time of subject discharge or death, to determine TB treatment outcome.

Further detail of pharmacokinetic sampling: All subjects initiating anti-TB treatment will have venous blood samples drawn for pharmacokinetic study at 2 weeks and 12 weeks after initiation. The 2 week time point allows steady state accumulation of anti-TB drug levels, and the 12 week mark compares changes in response to treated disease, improved malabsorption or weight gain and/or initiation of antiretroviral (anti-HIV) therapy.

All medication will be directly administered and observed by nursing staff in the fasting state per hospital routine. All anti-TB medications are given as a morning dose. For the antiretrovirals (as may be assayed in a subset of patients at the 12 week mark), nevirapine is dosed q12 hours and for efavirenz as an evening dose, whereby for efavirenz the C2hr level (relative to the timing of anti-TB meds) serves as the trough level and the C12hr as the peak. This sampling strategy minimizes the patient blood draws, and is similar to prior pharmacokinetic studies. A maximum of 10 ml will be obtained in heparinized tubes at each draw as up to 6 drugs will be required to be assayed by HPLC. Blood will be immediately centrifuged onsite at the Dispensary and plasma stored in sealed screw-cap tubes at -80°C with the subject's study identification number and week/time of blood draw. Batched plasma specimens will then be shipped to the University of Virginia for high performance liquid chromatography (HPLC) testing per established protocols.

Further detail of sputum specimen collection: A trained study nurse using airborne precautions will collect expectorated sputum from enrolled subjects. The sputum will be collected in a sealable container labeled with subject's study ID number and securely transported on the same day to the Irkutsk Dispensary TB Laboratory for GeneXpert MTB/RIF testing as available. As GeneXpert is a validated, WHO-approved test for M. tuberculosis screening, these results will be communicated directly to the treating physician by the trained study nurse, per standard of care. The remainder of investigations, including MIC testing, sequencing for drug-resistance mutation and spoligotyping for sublineage identification, will be experimental.