Second-Line Anti-Tuberculosis Drug Concentrations for Susceptibility Testing in the MODS Assay
Total Page:16
File Type:pdf, Size:1020Kb
Eur Respir J 2013; 41: 1163–1171 DOI: 10.1183/09031936.00059812 CopyrightßERS 2013 Second-line anti-tuberculosis drug concentrations for susceptibility testing in the MODS assay Sean Patrick Fitzwater*,#, G. Andrew Sechler*,", Oswaldo Jave+, Jorge Coronel*, Alberto Mendoza1, Robert H. Gilman", Jon S. Friedlande and David A.J. Moore*,#,e,** ABSTRACT: Multidrug-resistant tuberculosis (TB) threatens TB control worldwide. The AFFILIATIONS microscopic observation drug susceptibility (MODS) assay is a low-cost, high-performance TB *Laboratorio de Investigacio´nde Enfermedades Infecciosas, diagnostic tool for rapid liquid culture and direct isoniazid and rifampicin drug susceptibility Universidad Peruana Cayetano testing (DST). The objective of this study was to explore the potential for extending the MODS Heredia, Lima, assay to rapid second-line DST and to identify critical concentrations of candidate drugs for +Estrategia Sanitaria Nacional de prospective testing. Prevencio´n y Control de la Tuberculosis, Ministerio de Salud, Sputum samples from 94 TB culture-positive patients receiving second-line TB agents were Lima, and cultured following standardised MODS protocols, with a range of titrations of antimicrobial drugs 1Laboratorio de Referencia Nacional added. Critical concentrations were determined using a modified Kaplan–Meier survival curve de Mycobacteria, Instituto Nacional analysis. de Salud (INS), Lima, Peru. #Dept of International Health, Johns ? -1 Candidate critical concentrations were determined for capreomycin (10 mg mL ), ciprofloxacin Hopkins Bloomberg School of Public -1 -1 -1 -1 (1.25 mg?mL ), cycloserine (40 mg?mL ), ethambutol (10 mg?mL ), ethionamide (5 mg?mL ), Health, Baltimore, MD, and kanamycin (5 mg?mL-1), para-aminosalicylic acid (10 mg?mL-1) and streptomycin (10 mg?mL-1). No "Dept of Internal Medicine and cut-off point was identified for the other second-line drugs or for pyrazinamide. Paediatrics, Mt Sinai School of Medicine, New York, NY, USA. At particular concentrations of some second-line TB drugs this novel Kaplan–Meier analysis eDept of Infectious Diseases and clearly differentiated populations that were susceptible or resistant. These candidate critical Immunity and Wellcome Centre for concentrations should now be tested in a range of epidemiological settings to define the Clinical Tropical Medicine, Imperial performance of direct, second-line TB DST with MODS, offering potential low-cost second-line TB College London, London, and **LSHTM TB Centre and Dept of DST capacity. Clinical Research, Faculty of Infectious and Tropical Diseases, KEYWORDS: Diagnostics, extensively drug-resistant tuberculosis, multidrug resistance, second- London School of Hygiene and line drug susceptibility testing, tuberculosis Tropical Medicine, London, UK. CORRESPONDENCE D.A.J. Moore ver 1.7 million deaths are attributed to Union and South Africa, where the potentially Dept of Clinical Research, Faculty of tuberculosis (TB) annually, the leading devastating impact of XDR-TB–HIV co-infection Infectious and Tropical Diseases O curable cause of death from infectious has been demonstrated [2–6]. London School of Hygiene and disease in the world [1]. The impact of TB is Tropical Medicine greatest in resource-constrained settings, where The emergence of XDR-TB is an indirect indica- Keppel Street the poor suffer disproportionately. Despite pro- tion of programmatic failure to adequately London WC1E 7HT gress towards the targets set out by the United diagnose and treat MDR-TB [7]. Wider availabil- ity of rapid second-line DST that performs UK Nations Millennium Development Goals and the E-mail: [email protected] Stop TB Partnership, the expansion of robust TB reliably and cost-effectively in resource-limited settings is needed to address this threat [8, 9]. The diagnostic capacity remains a challenge [1, 2]. Received: nonproprietary microscopic observation drug April 10 2012 Drug-resistant TB threatens progress towards TB susceptibility (MODS) assay, a high-performance Accepted after revision: control [1, 2]. Globally the burden of multidrug- diagnostic technique for the liquid culture-based Aug 02 2012 resistant (MDR) TB is estimated at 440 000 cases, detection of TB and direct DST for MDR-TB, has First published online: and extensively drug-resistant (XDR) TB has been the potential to address this need. Developed in Aug 16 2012 found in 58 countries [1, 2]. Estimates of XDR-TB Peru, MODS has been shown to be a rapid, prevalence are, in part, limited by the availability reliable, low-cost method for DST of isoniazid of drug susceptibility testing (DST) for second- and rifampicin, but studies have not yet assessed European Respiratory Journal line TB drugs. However, XDR-TB accounts for performance in direct DST for second-line drugs Print ISSN 0903-1936 c .10% of MDR-TB cases in the former Soviet [10]. Additionally, data on ethambutol (EMB), Online ISSN 1399-3003 EUROPEAN RESPIRATORY JOURNAL VOLUME 41 NUMBER 5 1163 TUBERCULOSIS S.P. FITZWATER ET AL. streptomycin (STM) and pyrazinamide (PZA) DST in MODS wells was significantly impaired by the lower pH. Subsequent have failed to clearly demonstrate high performance at the PZA experiments were limited to pH 6.8. drug concentrations employed [10–12]. Inoculated plates were incubated and examined for growth as Evaluation of MODS performance for direct DST for these previously described [11, 25, 26]. Once TB growth was three drugs and for second-line agents will require a large, observed in a drug-free well, all other wells were examined. prospective, multicentre study comparing MODS with existing In the event of contamination or inconsistent growth, the reference standards in a range of epidemiological settings. original sample was decontaminated again and re-cultured. Because MODS is a direct drug susceptibility test, the M. tuberculosis was confirmed using spoligotyping or PCR [27]. evaluation needs to be performed using fresh sputum samples, Inconsistent growth was defined as M. tuberculosis growth in a not a bank of stored strains with known DST profiles. In order well with a higher drug concentration than a well that showed to design such a trial it is necessary to first define the candidate no growth. Isolates were sent for reference first- and second- critical concentrations for each drug, which was the primary line DST by the proportion method at the Peruvian National goal of this study. Using the receiver operating curve (ROC) Mycobacteria Reference Laboratory (at INS, Lima). methodology [13] and a novel Kaplan–Meier survival curve analysis, we aimed to explore the inhibitory capability of a Statistical analysis wide range of concentrations of each of the drugs tested and to Data were analysed using Stata 9.11 (StataCorp, College define the concentrations in MODS at which strains defined as Station, TX, USA) and Excel 2003 (Microsoft; Microsoft Corp., susceptible or resistant by the reference standard were best Redmond, WA, USA). The sensitivity and specificity of the differentiated. Thus, identified candidate critical concentra- differing concentrations was calculated using the proportion tions may then be prospectively evaluated in a larger trial as method as the reference. ROC curves were constructed from described above. this data. A Kaplan–Meier survival analysis was performed using increasing drug concentrations in place of the usual time MATERIALS AND METHODS variable. The survival failure event for each series was the Field methods point at which a sample yielded a positive culture at all lower Candidate participants with confirmed MDR-TB were identi- drug concentrations, but not higher concentrations. Series with fied from databases of the Peruvian Instituto Nacional de growth at the highest drug concentration were censored at that Salud (INS; Lima, Peru) and selected for inclusion if they were point. Individual drug concentration series that did not follow scheduled to receive specific MDR therapy, the most recent this pattern due to inconsistent growth or contamination at sputum culture was positive and resistance to any second-line critical concentrations were withheld from the analysis. Critical agent had been demonstrated previously. Consenting partici- concentrations for each drug were chosen from the Kaplan– pants provided a sputum sample of o5 mL, which was Meier survival curves by identifying the junction point transferred to the Universidad Peruana Cayetano Heredia between the rapid sterilisation phase of the curves and the (UPCH) laboratory in Lima, where it was stored refrigerated more level later portions of the curves. Curves that did not and processed, generally within 24 h of collection. follow the expected pattern were stratified by auramine smear status and proportion method resistance. The results of the Laboratory methods ROC and survival analyses were compared to determine the Sputum samples were processed according to the MODS most appropriate candidate critical concentration. methodology [14], with modifications to incorporate seven different concentrations of capreomycin (CAP), ciprofloxacin Ethical review (CIP), cycloserine (CS), EMB, ethionamide (ETO), kanamycin The institutional review board of UPCH (Lima) approved the (KM), para-aminosalicylic acid (PAS), PZA and STM. Sputum study protocol and the informed consent form, and the samples were decontaminated by the NaOH–N-acetyl-L-cysteine National TB Control Programme at the Peruvian Ministry of method and reconstituted in supplemented Middlebrook 7H9 Health, Lima, granted