Development and evaluation of a rapid PLP-NASBA assay for TB resistance profiling in routine labs

Thesis Abstract

The emergence of multidrug-resistant tuberculosis (MDR-TB) and extensively drug-resistant TB (XDR-TB) poses critical diagnostic delays and treatment failures in routine laboratory settings, underscoring the need for rapid, accurate resistance profiling that can be integrated into existing workflows. This study addresses the gap by developing and evaluating a rapid PLP-NASBA (pyrophosphate-locked Polymerase Chain Reaction–Nucleic Acid Sequence-Based Amplification) assay to profile TB drug resistance directly from clinical sputum specimens and decontaminated culture isolates, aiming to shorten turnaround times and improve therapeutic decision-making in routine laboratories. The primary aim is to determine the performance of the PLP-NASBA assay for detecting canonical mutations associated with isoniazid, rifampicin, fluoroquinolone, and second-line injectable drug resistance, and to compare its diagnostic accuracy with the standard phenotypic susceptibility testing (DST) and line probe assays (LPA). Specific objectives include (1) to optimize PLP-NASBA primer-probe sets targeting katG, inhA promoter, rpoB, gyrA, gyrB, embB, rrs, and eis loci; (2) to evaluate analytical sensitivity and specificity using quantified reference strains and a blinded panel of 300 clinical specimens (200 sputum samples and 100 culture isolates) representing varying resistance profiles; (3) to assess diagnostic concordance with MGIT DST and Xpert MTB/RIF Ultra as reference comparators; (4) to assess workflow metrics including total turnaround time, reagent costs, and ease of integration in routine laboratory settings; (5) to model the potential impact on treatment regimens using scenario analysis informed by Bayesian latent class models. The methodology adopts an explanatory sequential mixed-methods design anchored in the Technology Acceptance Model (TAM) and the Diffusion of Innovations framework, ensuring that technical performance is evaluated alongside user acceptability and operational feasibility. A cross-sectional diagnostic accuracy study will be conducted in a tertiary reference laboratory, enrolling consecutive clinical specimens. Analytical validation will involve 50 synthetic positive controls and 50 negative controls to establish limit of detection, specificity, and cross-reactivity. Clinical validation will process 300 specimens, with PLP-NASBA results blinded to the reference test outcomes to avoid review bias. Data collection instruments include standardized PLP-NASBA reaction kits, validated DNA/RNA extraction protocols, and electronic data capture forms. Data analysis will use descriptive statistics to summarize performance characteristics, with sensitivity, specificity, positive and negative predictive values computed against composite reference standards. Cohen’s kappa will quantify agreement with MGIT DST and LPA; McNemar’s test will compare paired proportions. Regression analyses will quantify the relationship between parasite load and assay sensitivity, while time-to-result analyses will evaluate throughput improvements. A probabilistic sensitivity analysis will assess the robustness of results under varying pre-test probabilities. The study anticipates that the PLP-NASBA assay will demonstrate superior turnaround time (?6 hours from sample receipt) and high concordance (kappa >0.8) with established resistance profiles, particularly for rifampicin and isoniazid resistance, with variable performance for second-line markers depending on mutation prevalence. Expected findings include precise estimates of analytical sensitivity down to 10^2 CFU/mL for key loci, specificity exceeding 98%, and a reduction in overall diagnostic time by at least 48 hours compared with phenotypic DST. The study contributes to knowledge by providing a validated rapid molecular diagnostic tool tailored for routine laboratories, offering evidence on implementation feasibility, cost-effectiveness, and potential to inform timely, individualized therapy for TB patients. The conclusion will emphasize that PLP-NASBA can complement existing methods, enabling rapid initial resistance profiling and guiding early regimen optimization, with policy recommendations for integration into TB diagnostic algorithms and guidelines for training and quality assurance in diverse laboratory contexts. Recommendations will address scalability, external validation in high-burden settings, and ongoing surveillance of missing resistance mutations to maintain assay relevance.