Bioneer Q-RFIA Clinical Evaluation

Multicentre Clinical Study to Assess the Performance of the Bioneer IRONqPCR ™ RFIA Kit for INH-, RIF-, FQ- and AG-Resistance Detection

Background and rationale Evidence of the clinical diagnostic accuracy and operational characteristics of the Bioneer IRONqPCR ™ RFIA Kit is needed to comprehensively evaluate Bioneer RFIA validity and inform global and national policy decision-making.

The Bioneer IRONqPCR ™ RFIA Kit. The Investigational Use Only (IUO) product is intended for the detection of MTB and mutations associated with RIF, INH, FQ and AMG drug resistance. The assay which is run as a kit cartridge physically consists of two parts; a nucleic extraction body part and the PCR plate. The cartridge is intended to be used with the IRON qPCR system.

The IRON qPCR system provided for this study should only be used for testing IUO product and study samples. The IRON qPCR instrument developed by Bioneer is intended to be used at lower healthcare systems. The instrument has a small footprint ((30.5(W) x 27(D) x 39.5(H) cm), weight (< 15 kg) and can be operated on a battery charge. The system has the capability to fully automate the detection of 40 targets and to provide sample-to-result within 30 minutes.) The investigational product will be strictly accounted for, including receipt and inventory, storage, use during the trial, and return or disposal, as detailed in the Study Manual for this study.

Other diagnostic tests (2nd comparators and reference tests) provided for use in this Trial are:

• Xpert MTB/RIF Ultra assay
• Xpert MTB/XDR assay
• Hain MTBDRplus assay
• Hain MTBDRsl assay Instructions for use of the investigational product(s) are provided in the IP package inserts.

The Bioneer IRONqPCR ™ RFIA Kit assay detects Mtb and mutations associated with INH, RIF, FQ and aminoglycoside resistance direct from sputum. In line with this intended use, clinical samples will be directly tested by the Bioneer Accupower Q-RFIA assay. However, the assay will also be used to test the culture isolate. This additional Bioneer IRONqPCR ™ RFIA Kit assay testing will allow for a comparison of performance and non-determinant rates between testing directly from the sample and testing from the culture isolate. It will also support resolution of potential discordant LPA and NGS results from the initial Bioneer IRONqPCR ™ RFIA Kit test result, as these tests will also be performed from the culture isolate (discordance may arise as a result of infections with a mix of different strains, with only one strain being grown out in culture). Furthermore, as all assays will be performed on the same culture, there is also no risk of any assay being tested on a different quality sample.

In order to efficiently assess the diagnostic performance of the Bioneer IRONqPCR ™ RFIA Kit on IRON qPCR, participants will be asked to provide ≥3ml of sputum (either a single sputum or two consecutively collected and pooled sputa). Spot sputa should be pooled and homogenized. For samples where only one specimen was collected from the participant, the sample needs to be homogenized. Briefly, following a mechanical homogenisation step the sputum sample will be split and the following tests will be performed:

• Fluorescence smear microscopy (FM smear)
• Xpert MTB/RIF Ultra
• Xpert MTB/XDR assay
• Bioneer Accupower Q-RFIA
• Liquefaction/decontamination by NALC-NaOH followed by re-suspension on PBS and inoculation on MGIT and Löwenstein-Jensen culture media
• MTB complex confirmation of positive cultures by either MPT64/MPB64 antigen detection or PCR (GenoType) from each positive culture.
• Phenotypic DST to test INH, RIF, fluoroquinolone (moxifloxacin and levofloxacin), AMK, and KAN resistance will be done on the first culture positive for MTB complex per patient.
• Molecular testing from all MTB-complex confirmed cultures using GenoType MTBDRplus and MTBDRsl
• Whole genome sequencing will only be done from samples where the Bioneer result (from either the sputum or culture testing) is discordant with phenotypic DST
• Bioneer Accupower Q-RFIA from all MTBC confirmed cultures

All diagnostic test results (smear, Bioneer IRONqPCR ™ RFIA Kit on direct sample, Xpert MTB/RIF Ultra on direct sample, Xpert MTB/XDR on direct sample, MGIT, LJ, Bioneer Accupower RFIA on culture isolate, MTBDRplus, MTBDRsl, and phenotypic DST results) will be recorded by the laboratory technicians for each site on the laboratory CRF.

Retrospective specimens selected from different biobanks will also be processed. In the event where routine testing has already been done on the retrospective specimens, FM smear and Xpert MTB/RIF or Ultra does not have to be repeated. Prior to the study start the FIND study manager will evaluate the diagnostic data linked to each specimen to determine which testing needs to be done or repeated for each site. Xpert MTB/XDR test will be performed for every selected specimen.

Reference Standard Test and Index Test Procedures Culture (LJ and MGIT), phenotypic DST and smear microscopy will be performed at the testing sites based on local SOPs. Whole genome sequencing will be done at a central location following a standardised protocol to be provided by FIND.

Each participating laboratory will undergo a remote laboratory evaluation to assess the quality management system of the laboratory and more specifically to ensure standardized and high-quality performance of routine tests including smear, culture and DST. Results from reference standard testing will be recorded by technicians blinded to index test results, eliminating the risk of review bias. Presence of MTB complex will be confirmed for all patients from a positive culture using an MPT64-based rapid test or WHO-approved NAAT on the culture isolate.

A culture reference standard may be imperfect in paucibacillary (smear-negative) TB. This can lead to bias in the accuracy estimates of the index test: (i) sensitivity may be overestimated when index test and reference standard both yield negative results in a patient with TB; and (ii) specificity may be underestimated when the index test gives a positive result in a patient with TB who tests negative with the reference standard. A biased sensitivity estimate is acceptable as it still represents a meaningful parameter (the proportion of culture-positive TB that is also detected by the index test). However, since near-perfect specificity is usually required of TB assays, bias in specificity can be problematic. To mitigate the risk of spurious index test "false-positive" results, a total of two cultures (one MGIT and one LJ culture) will be done in all cases.

A composite reference standard of phenotypic DST and NGS will be used for resistance detection to ensure high confidence in the reference standard for this study. Using phenotypic DST as a reference standard may be imperfect for mutations with minimum inhibitory concentrations around the breakpoint of the respective drug. This can lead to bias in the accuracy estimates of the index test where (i) sensitivity may be overestimated when index test and reference standard both yield negative results for resistance; and (ii) specificity may be underestimated when the index test gives a positive result for resistance where phenotypic DST gives a negative result. To mitigate this risk sequencing will be done for all samples with a discrepant result between the index assay and phenotypic DST. As part of this study, sequencing results will only be available for results that are discordant between the index assay and pDST. Using sequencing only for discrepant samples can also lead to bias in the accuracy estimates for the index assay for cases where both the index and phenotypic DST results are negative for resistance and where sequencing would have shown the presence of a mutation. This risk is however minimal as clinically resistant mutations with a susceptible phenotype result is rare and their genotypic basis not well understood. In such cases the phenotypic DST result is still considered as the gold standard. Therefore, we considered use of the composite reference standard the approach with the lowest risk of bias overall and will use this for the primary analysis.

Phenotypic MGIT DST will be done on the culture isolate for each clinical sample. MGIT DST will be done on site to determine culture drug-susceptibility to INH, RIF, moxifloxacin, levofloxacin, AMK, and KAN at the revised WHO critical concentrations15. Whole genome sequencing will be done for samples where the Bioneer assay result (from either the sputum or culture testing) is discordant with phenotypic DST. Whole genome sequencing will be done on DNA extracted from the cultured sample. The output for both phenotypic DST and NGS results are automated, and require no manual operator interpretation. Testing results, when obtained, will be input onto the laboratory CRF.

The index test will be performed on both the direct clinical sample as well as the positive culture, to ensure high Bioneer IRONqPCR ™ RFIA Kit diagnostic performance direct from sputum and to provide a basis for comparison if any discordants are seen between the initial Bioneer IRONqPCR ™ RFIA Kit result and the phenotypic DST and additional molecular test results (NGS and LPA). The output for the Bioneer IRONqPCR ™ RFIA Kit is automated and does not require manual interpretation. As with the reference test, all Bioneer IRONqPCR ™ RFIA Kit test results will be noted on the laboratory CRF.

GCP statement This study will be conducted in compliance with the protocol, the Declaration of Helsinki, ICH-GCP E6 (R3) guidelines, ISO 20916:2019 and in compliance with all applicable regulations.