SHINE-TB demonstrated 100% specificity (6/6) and 100% sensitivity (7/7) when testing against patient samples. We also found a positive correlation between our Cas12a and Cas13a signal, indicating in some samples both Cas12a and Cas13a are inhibited. (16/21)

Finally, we tested our assay’s ability to detect Mtb in real clinical samples. By partnering with CIDEIM, we were able to test 14 blinded patient samples that had been previously evaluated using MGIT culture and GeneXpert Ultra using our novel extraction method and SHINE-TB. (15/21)

We then moved to validate our finalized assay. We found that while sputum matrix causes a loss in sensitivity compared to a media matrix, our LoD was ~69 CFU/mL for H37Rv and 80.5 CFU/mL for BCG in sputum, similar to the gold-standard diagnostic (Xpert). (14/21)

TB predominantly affects under-resourced communities, so we adapted the assay to make sure it was compatible with both lyophilization and lateral flow. We found high sensitivity in both of these formats. (13/21)

The gold standard sample type for TB is sputum. Due to its viscous and inconsistent nature, we added a parallelized Cas12a internal control that targets human DNA so that poor sample extraction or inhibition are reported as undetermined instead of as false negatives. (11/21)

Our Cas13a multitarget assay also had no cross reactivity with nontuberculosis mycobacteria or other pathogens, demonstrating the Cas13a assay has both high sensitivity and specificity. Thus, we focused on Cas13a for detection of TB. (10/21)

As different strains of TB may have varying copies of IS6110 and IS1081, we thought it was important to target both of these sequences in a single assay to improve potential sensitivity. With this Cas13a multitarget assay, we found no loss in sensitivity detecting H37Rv. (9/21)

After optimizing Cas13a-based detection of IS6110 and IS1081, we performed a head-to-head comparison with Cas12a using assays targeting the same IS6110 region. Cas13a detected down to 1 copy, while Cas12a signal dropped off below 64 copies/µL with synthetic IS6110. (8/21)

We began by designing a Cas12a based diagnostic to detect multicopy insertion sequences IS6110 and IS1081, exclusive to the Mtb complex (MTBC). For simplicity, we created a one-pot assay that combined amplification and detection. (6/21)

Owen Dunkley and I developed SHINE-TB to bridge the current gap between widespread accessibility and diagnostic reliability in TB diagnosis. This CRISPR-Dx leverages both Cas13a and Cas12a for highly sensitive detection of both Mtb and a human DNA control. (5/21)