Clinical Evaluation of tNGS for Diagnosis of DR-TB (Seq&Treat)

Multicentre Clinical Trial to Assess the Performance of Culture-free, End-to-end Targeted NGS (tNGS) Solutions for Diagnosis of Drug Resistant TB (DR-TB)

Current rapid molecular assays for detection of drug-resistant TB from direct clinical samples have important limitations. They are not suited for high-throughput settings; can only be used to detect a limited number of target gene regions and are not ideal for detection of phenotypic resistance conferred by mutations across large gene regions (e.g. pyrazinamide).

Culture-free, end-to-end targeted NGS (tNGS) Solutions for Diagnosis of Drug Resistant TB can offer higher throughput and greater accuracy across more TB drugs than current WHO endorsed molecular assays, and a significantly faster time to result than phenotypic drug susceptibility testing (DST).

Evidence regarding the clinical diagnostic accuracy and operational characteristics of tNGS solutions is needed to comprehensively evaluate tNGS for diagnosis of drug-resistant TB among patients who have been diagnosed with TB, and will be critical to inform global and national policy.

Study Overview

• Status: Completed
• Conditions: Tuberculosis, Multidrug-Resistant
• Intervention / Treatment: Diagnostic test: targeted Next Generation Sequencing (tNGS)

Detailed Description

In recent years, tuberculosis (TB) control efforts have been complicated by the rise and spread of MDR-TB, or TB that is resistant to the first-line drugs isoniazid (INH) and rifampicin (RIF), and XDR-TB, or MDR-TB that has developed additional resistance to a fluoroquinolone (FQ) and any of the second-line injectable compounds (SLI) [amikacin (AMK), kanamycin (KAN) and/or capreomycin (CAP)] The rapid diagnosis and appropriate treatment of M/XDR-TB is essential to prevent significant morbidity, mortality and further transmission of disease. For treatment of uncomplicated MDR-TB, the World Health Organization (WHO) recently endorsed a 6-9 month treatment regimen, replacing conventional 18-24 month regimens. The FQs and SLIs are key components of the 6-9 month regimen, and so it is necessary to rule-out resistance to these compounds prior to treating patients with the shorter regimen. For longer MDR-TB treatment regimens, WHO no longer endorses the use of CAP or KAN, but the regimens still rely heavily on FQs, with AMK and PZA listed as recommended Group C drugs, thus ensuring that the rapid diagnosis of all of these drugs is still relevant and critical to the timely treatment of DR-TB4.

The current standard for comprehensive DR-TB diagnosis is culture-based DST (liquid culture, e.g. MGIT or solid culture, e.g. LJ), which can only be implemented in custom-built Biosafety Level 3 (BSL3) laboratories. The global response to the need for comprehensive DST in the last two decades has therefore been focused on the scale-up of BSL3 facilities and MGIT liquid culture systems to expand global capacity. While this enabled some standardization of DST workflows and resulted in a foundation for a global network of reference TB laboratories, it has come at considerable cost and effort in terms of infrastructure needs, significant hands-on time and expertise for routine use. Moreover, the slow growth of Mycobacterium tuberculosis (Mtb) in culture means it takes 6-8 weeks to get results from conventional culture-based DST and data has shown that often culture comes too late to have an impact on the patient outcomes. More recently, the global response to the public health need for universal DST has been a wide rollout and broad adoption of rapid molecular TB tests with limited DST capabilities to support early treatment decisions and complement traditional phenotypic DST, e.g. (Cepheid Xpert MTB/RIF or Ultra, Hain GenoType MTBDRplus). However, the inherent design of probe- or array-based molecular diagnostics limits their coverage to a limited number of specific mutations included in the initial design-locked assay, and prevents them from integrating new mutations and new drug resistance prediction without going back to complete product development and validation. Next generation sequencing (NGS) technology has the potential to revolutionize global comprehensive DST. Targeted NGS (tNGS) can be used to interrogate more than numerous genome-wide targets in the Mycobacterium tuberculosis (MTB) that contain mutations associated with resistance, and differentiate them from mutations that do not confer resistance, providing better performance than probe or array-based molecular diagnostics.

For DR-TB diagnosis, tNGS offers huge potential gains over culture-based methods and other molecular diagnostics in terms of speed, ease of use and comprehensive coverage. However, the uptake of sequencing for DR-TB patient care in LMICs has been limited due to (i) the absence of a standardized and complete solution directly from clinical samples to data analysis and interpretation (i.e. an end-to-end solution); (ii) lack of verification and validation of such commercial end-to-end sequencing solutions to drive policy and guidelines for use of sequencing as an in vitro diagnostic; and (iii) perceived technical and cost barriers. The WHO and FIND developed a technical guide on requirements and utilization of sequencing for clinical diagnosis of DR-TB and a consensus-based Target Product Profile (TPP) for sequencing. The standard components of an End-to-end solution include: DNA extraction from sputum sample, library preparation, sequencing and data analysis/interpretation for clinical result reporting. Based on the published TPP guidelines, a number of manufacturers have developed end-to-end solutions for DR-TB diagnosis. These sequencing solutions require rigorous analytical testing, laboratory validation, and clinical evaluation in order to ensure that they meet technical as well as operational target product profile criteria.

The current study is part of a two-phase project to assess performance of culture-free, tNGS end-to-end solutions for drug resistant TB diagnosis, and to recommend their use in diverse clinical settings.

• Phase 1 is an analytical study performed at the manufacturer site. It is a blinded laboratory analytical validation of tNGS end-to-end solutions using a well-characterized panel of MTB strains.
• Phase 2 (the current study) is a multi-centre, clinical trial to evaluate the diagnostic accuracy and performance of up to 3 tNGS end-to-end sequencing solutions in settings of intended use.

The two study phases will provide complementary bodies of evidence to address the study objectives and determine whether tNGS solutions meet the diagnostic performance criteria outlined in the NGS Target Product Profile (TPP) for clinical diagnosis of drug resistant TB. The purpose of this clinical trial is to confirm accuracy estimates observed in the laboratory validation study and ensure that tNGS diagnostic performance characteristics are consistent at sites of intended use.

The focus of this protocol is the multi-centre clinical evaluation of the diagnostic accuracy and technical performance of End-to-end tNGS solutions for DR-TB diagnosis.

• Study Type: Observational
• Enrollment (Actual): 800

Contacts and Locations

• Study Contact: Name: Pamela Nabeta, MD; Phone Number: +41 22 710 27 87; Email: Pamela.Nabeta@finddx.org
• Study Contact Backup: Name: Andres De la Rossa, PhD; Phone Number: +41 79 881 47 05; Email: Andres.DelaRossa@finddx.org

Study Locations

• Georgia
  • Tbilisi, Georgia, 380060
    • National Center for Tuberculosis and Lung Diseases
• India
  • Maharashtra
    • Mumbai, Maharashtra, India, 40016
      • Hinduja Hospital and Medical Research Centre
• South Africa
  • Sandringham
    • Johannesburg, Sandringham, South Africa, 2192
      • National Institute for Communicable Diseases & Wits Health Consortium

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: 18 years and older (Adult, Older Adult)
• Accepts Healthy Volunteers: No

• Sampling Method: Probability Sample

Study Population

Adults with pulmonary TB, confirmed by Xpert MTB/RIF or Ultra, who are also at risk for or proven to have drug resistant TB, will be invited to participate in the trial. Interested individuals meeting these criteria will be referred to study personnel for screening. Both HIV-positive individuals and HIV-negative individuals will be included in this study. Depending on the sites individuals may be recruited at outpatient clinic settings as well as inpatient hospital settings.

Description

Inclusion Criteria:

1. A TB-positive result by Xpert MTB/RIF OR Xpert MTB/RIF Ultra (i.e. "MTB DETECTED") at or prior to enrollment, AND

2. At risk for drug resistant TB based on at least one of the following risk factors:

   A. A positive RIF-resistance result by Xpert MTB/RIF OR Xpert MTB/RIF Ultra (i.e."RIF resistance DETECTED") OR B. Not responding TB treatment with positive sputum smear or culture after ≥ 3 months of standard TB treatment. OR C. Previously diagnosed with Rif-resistant/MDR-TB and failed TB treatment with positive sputum smear or culture after ≥ 3months of a standard MDR-TB regimen OR D. Previously received >1 month of treatment for a prior TB episode OR E. Close contact with a known drug-resistant TB case AND

3. Willing to provide sputum AND
4. 18 years of age and older (or legal adult age corresponding to the site) AND
5. Provision of signed informed consent

Exclusion Criteria:

1. Have started treatment for current TB episode more than 7 days prior to date of enrolment (i.e. must have been on treatment for less than 7 days for this TB treatment episode at enrolment) OR
2. Institutionalized or imprisoned

Study Plan

How is the study designed?

Design Details

• Observational Models: Cohort
• Time Perspectives: Cross-Sectional

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Point estimates, with 95% confidence intervals, of sensitivity and specificity for RIF resistance detection for each of up to 3 tNGS solutions	Outcomes 1.1 to 1.5 will be evaluated on the PP population. Point estimates (with 95% CI) of sensitivity and specificity will be derived based on pooled data for all sites. In order to assess whether there is an association between the results and the clinical site from where the samples were collected, the estimates of sensitivity and specificity for each site will be compared with each other by use of a Pearson's chi-squared test, at a significance level of 5%, adjusted by Bonferroni correction for the total number of tests performed. A random effect model will be used if it is believed that a high site-effect is present: this will be assessed by the evaluation of heterogeneity using Cochran's Q (with a significance level set at 0.1) and the I2 statistics (with a value >0.5).	January 2021 - March 2021
Point estimates, with 95% confidence intervals, of sensitivity and specificity for INH resistance detection for each of up to 3 tNGS solutions	Outcomes 1.1 to 1.5 will be evaluated on the PP population. Point estimates (with 95% CI) of sensitivity and specificity will be derived based on pooled data for all sites. In order to assess whether there is an association between the results and the clinical site from where the samples were collected, the estimates of sensitivity and specificity for each site will be compared with each other by use of a Pearson's chi-squared test, at a significance level of 5%, adjusted by Bonferroni correction for the total number of tests performed. A random effect model will be used if it is believed that a high site-effect is present: this will be assessed by the evaluation of heterogeneity using Cochran's Q (with a significance level set at 0.1) and the I2 statistics (with a value >0.5).	January 2021 - March 2021
Point estimates, with 95% confidence intervals, of sensitivity and specificity for fluoroquinolone (moxifloxacin, levofloxacin) resistance detection for each of up to 3 tNGS solutions	Outcomes 1.1 to 1.5 will be evaluated on the PP population. Point estimates (with 95% CI) of sensitivity and specificity will be derived based on pooled data for all sites. In order to assess whether there is an association between the results and the clinical site from where the samples were collected, the estimates of sensitivity and specificity for each site will be compared with each other by use of a Pearson's chi-squared test, at a significance level of 5%, adjusted by Bonferroni correction for the total number of tests performed. A random effect model will be used if it is believed that a high site-effect is present: this will be assessed by the evaluation of heterogeneity using Cochran's Q (with a significance level set at 0.1) and the I2 statistics (with a value >0.5).	January 2021 - March 2021
Point estimates, with 95% confidence intervals, of sensitivity and specificity for second-line injectable (amikacin, capreomycin, kanamycin) resistance detection for each of up to 3 tNGS solutions	Outcomes 1.1 to 1.5 will be evaluated on the PP population. Point estimates (with 95% CI) of sensitivity and specificity will be derived based on pooled data for all sites. In order to assess whether there is an association between the results and the clinical site from where the samples were collected, the estimates of sensitivity and specificity for each site will be compared with each other by use of a Pearson's chi-squared test, at a significance level of 5%, adjusted by Bonferroni correction for the total number of tests performed. A random effect model will be used if it is believed that a high site-effect is present: this will be assessed by the evaluation of heterogeneity using Cochran's Q (with a significance level set at 0.1) and the I2 statistics (with a value >0.5).	January 2021 - March 2021
Point estimates, with 95% confidence intervals, of sensitivity and specificity for pyrazinamide (PZA) resistance detection for each of up to 3 tNGS solutions	Outcomes 1.1 to 1.5 will be evaluated on the PP population. Point estimates (with 95% CI) of sensitivity and specificity will be derived based on pooled data for all sites. In order to assess whether there is an association between the results and the clinical site from where the samples were collected, the estimates of sensitivity and specificity for each site will be compared with each other by use of a Pearson's chi-squared test, at a significance level of 5%, adjusted by Bonferroni correction for the total number of tests performed. A random effect model will be used if it is believed that a high site-effect is present: this will be assessed by the evaluation of heterogeneity using Cochran's Q (with a significance level set at 0.1) and the I2 statistics (with a value >0.5).	January 2021 - March 2021

Secondary Outcome Measures

Outcome Measure	Time Frame
Point estimates, with 95% confidence intervals, of sensitivity and specificity for additional second-line resistance detection (bedaquiline, linezolid, clofazimine, streptomycin) for up to 3 tNGS solutions.	January 2021 - March 2021
Comparison of drug-specific point estimates of sensitivity and specificity (with 95% confidence intervals) for up to 3 tNGS solutions against Hain MTBDRplus/sl drug-specific results.	January 2021 - March 2021
Comparison of overall test success rate (defined as the total number of full profiles i.e. calls across all drug targets) in up to 3 tNGS solutions against success rate in Xpert MTB/RIF and Hain MTBDRplus/sl.	January 2021 - March 2021
Summary of technical performance characteristics in up to 3 tNGS solutions including invalid and indeterminate rates, ease of use metrics, and other operational characteristics	January 2021 - March 2021

Collaborators and Investigators

• Sponsor: Foundation for Innovative New Diagnostics, Switzerland

Study record dates

Study Major Dates

• Study Start (Actual): April 16, 2021
• Primary Completion (Actual): January 31, 2023
• Study Completion (Actual): June 30, 2023

Study Registration Dates

• First Submitted: January 20, 2020
• First Submitted That Met QC Criteria: January 20, 2020
• First Posted (Actual): January 27, 2020

Study Record Updates

• Last Update Posted (Actual): October 4, 2023
• Last Update Submitted That Met QC Criteria: October 3, 2023
• Last Verified: October 1, 2023

More Information

Terms related to this study

• Additional Relevant MeSH Terms: Infections; Bacterial Infections; Bacterial Infections and Mycoses; Gram-Positive Bacterial Infections; Actinomycetales Infections; Mycobacterium Infections; Tuberculosis; Tuberculosis, Multidrug-Resistant
• Other Study ID Numbers: TB038

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No