Effective control of SLDR requires rapid diagnosis and prompt treatment. However, access to diagnostic assays for SLDR remains a challenge, especially in highly endemic countries. The present study is a retrospective study that shows the greater rapidity of the DNA bio-chip assay over the phenotypic methods. In most developing countries, like India, DST is usually performed on solid/liquid medium, wherein results are available after 8 to 18 weeks. During this time many patients with MDR-TB or XDR-TB may have died, transmitted their disease, or both. Conventional DST for second-line drug testing is far less standardized than DST for first-line drugs. Our bio-chip assay detects the presence of resistance to FQ and SLID of M. tuberculosis isolates within a short span of 1 day by detecting point mutations present in, gyrA, rrs and eis genes. This could help clinicians initiate appropriate and timely therapy and implement infection control measures to curb the spread of resistance.

Although nucleic acid amplification tests (NAATs) are widely used in TB diagnostics, they are limited by the concentration and quality of DNA used for PCR, which depends on the bacterial load in the sample and DNA isolation method [28, 29]. The vital tasks in developing NAATs are increasing the analytical and clinical sensitivity as well as specificity for resistance detection to a wide range of anti-TB drugs. The targeted analysis of the M. tuberculosis genome for DR detection requires multiplex amplification of several genomic targets as M. tuberculosis DR is associated with multiple mutations in multiple genes. Although PCR for a single target achieves high sensitivity, multiplexing lowers the analytical sensitivity. In the case of hybridization-based NAATs, the efficiency of the amplification stage is significant for the overall performance and sensitivity of the assay. We standardized multiplex PCR using an asymmetric PCR approach for gyrA, and symmetric PCR for rrs and eis genes, which increased the sensitivity of the biochip as a higher signal can be achieved for gyrA probes.

While next generation sequencing (NGS) provides a comprehensive view and can detect novel or rare mutations, DNA biochips are advantageous for rapid, affordable, and targeted detection of known resistance mutations in TB, as for SLDR detection using our SLDR bio-chip, making them suitable for clinical use in resource-limited or high-burden areas. PCR based NAAT, like Xpert MTB/RIF assay and its successor the Xpert MTB/RIF Ultra (Cepheid, Sunnyvale, CA, USA) are used extensively worldwide to detect M. tuberculosis and rifampicin resistance but can’t assess resistance to isoniazid and second-line drugs [30]. Multiplex hybridization-based NAATs, such as GenoType MTBDRplus, CapitalBio microarray, TB-Biochip microarray and BluePoint MtbDR microarray are available, but for detecting resistance to first line drugs [31,32,33,34]. Second line DST is also technically demanding and requires advanced laboratory infrastructure [35, 36]. Economic and logistical challenges limit second-line drugs DST. Genotype MTBDRsl Ver 2.0, LPA for FQ and SLID resistance detection is available for rapid detection of SLDR but its use is restricted to central or national TB laboratories. Till date, no DNA bio-chip is available in the market for the detection of SLDR and there are limited studies on the development of DNA bio-chip/ microarray for mutation studies for SLDR [37]. Our SLDR bio-chip assay targets both FQs and SLIDs and can be performed in the laboratory with simple equipments like a hybridization oven and chemiluminescence imager. A unique feature of our bio-chip is the development of bio-chip on PC-TEM that provides higher immobilization capacity than planar substrates like glass on which most of the bio-chips are developed. Compared to other microporous supports like nitrocellulose filter membranes, PC-TEM provides higher pore density, uniformly sized pores, and flat, smooth surfaces for uniform spotting.

FQ resistance is correlated with mutations in the quinolone resistance-determining region, predominantly in gyrA and in gyrB, and a 90% accuracy of prediction of FQ resistance could be achieved when mutations are detected at the 90th and 94th codons. The resistance to SLIDs can be predicted by the nucleotide changes in rrs gene, with key mutations at A1401G and G1484T. Therefore, the developed bio-chip targeted codons 90 and 94 in gyrA gene and 1401 and 1484 base in rrs gene. The sensitivity and specificity of the developed bio-chip for detection of FQ resistance ranged from 75-100% and 96.7%-100%, respectively. 6 of 8 OFX resistant strains were correctly identified by biochip analysis (sensitivity - 75%). Sequencing analysis shows the presence of gyrA D94N and G88A mutation, probes for which were not present on the bio-chip resulting in loss of sensitivity. The specificity of FQ was 96.7% as 2 of 61 isolates sensitive to OFX show the presence of gyrA D94G by bio-chip analysis. The presence of these mutations was confirmed by sequencing. Loss of specificity is due to incorrect results obtained by culture DST. Similarly, the sensitivity and specificity of SLID detection ranged from 90.9-100% and 96.7-100% respectively. Loss of sensitivity was due to the absence of probes for the mutation found in the isolate and loss of specificity was due to incorrect results obtained by culture DST. The vast majority of molecular tests cannot achieve a performance of 100% sensitivity and 100% specificity, mainly because TB drug resistance mechanisms are not completely understood. The analytical sensitivity of the method, or the minimal detectable bacterial genome equivalents that can be reliably detected by biochip assay is 250 per µl.

This study demonstrates the successful development and evaluation of an in-house DNA biochip for the rapid detection of mutations conferring resistance to FQ and SLIDs. One of the primary strengths of biochip assay is its significantly reduced turnaround time. Conventional culture-based DST requires weeks to yield results, whereas this biochip assay provides actionable results in one day. This rapid detection is critical for initiating timely and appropriate treatment regimens, especially in cases of SLDR TB. Our biochip assay can be used easily in any laboratory utilizing molecular biology techniques for analysis of M. tuberculosis, by adding a simple and inexpensive hybridization incubator and chemi documentation imaging system. Also, bio-chip is cost-effective compared to commercial assays for detecting SLDR and fluorescence-based DNA bio-chip assays which necessitates use of expensive equipment and reagents. This makes bio-chip particularly suitable for implementation in low-resource settings where TB burden is often highest.

The biochip is designed to cover clinically relevant and high-prevalence mutations in key resistance-associated genes (gyrA, rrs, and eis promoter). This targeted approach enabled the accurate identification of the most common resistance-conferring mutations, such as gyrA D94G and rrs A1401G, that are well-established markers of resistance to FQ and SLID respectively. Moreover, the biochip’s modular design offers adaptability for future applications, such as incorporating new resistance markers or tailoring probes to region-specific mutations as the mutation landscape evolves.

However, there are important limitations that warrant consideration. The biochip assay relies on a limited mutation spectrum, detecting only those mutations for which probes have been designed. This inherently excludes rare, novel, or complex mutations that may also confer resistance, potentially resulting in false-negative results in such cases. In present study, four isolates were missed due to presence of rare mutations. Also, bio-chip assay requires chemiluminescence imaging and temperature-controlled hybridization equipment, which may limit its deployment in primary health care facilities without molecular infrastructure. Second, the assay workflow—though streamlined—still involves multiple technical steps that require trained personnel and stringent quality control to avoid cross-contamination or assay failure.

Further studies with a larger and more geographically diverse panel of M. tuberculosis isolates are needed to validate biochip performance across different settings. Automation of the hybridization and detection processes could further improve the throughput and reduce the hands-on time. Future directions include the integration of this biochip into a comprehensive diagnostic platform for rapid and simultaneous detection of first- and second-line drug resistance, as well as the exploration of its potential for direct detection of mutations from sputum samples to further reduce the diagnostic delay.