Methods: We conducted a retrospective chart review over 3 years to evaluate the utility of universal PCR-sequencing when ordered at the discretion of clinicians at our institution. A total of 163 fluid or tissue samples from various anatomic sites (most common: CNS, lung, lymph nodes) were collected for identification of bacterial, mycobacterial, and/or fungal organisms, by universal PCR sequencing of conserved regions of the 16S, hsp65/rpoB, and/or 23S/ITS genes, respectively.
Results: In this study, 26 /163 samples (16%) were true positives, and 8/163 (5%) were false positives. The rate of positive results varied by type of tissue collected (range 6-33%, lowest for CNS tissue, highest for cardiac and orthopedic tissue). Positive rates were significantly increased when organisms were visible on microbiology smears (50% versus 12%, p value < 0.01) and when high levels of inflammation were seen on histology (25% versus 4%, p value 0.02). There was poor concordance between culture and PCR results: 3 of 13 (23%) culture-positive samples were also PCR-positive, while only 3 of 21 (15%) PCR-positive results were also culture-positive. Notably, the proportion of true positive results was not significantly different depending on whether the tissue sent for sequencing was fresh versus formalin-fixed paraffin-embedded (FFPE), or from a fine-needle aspirate versus a tissue biopsy. However, 7 of the 8 samples with a false positive result were FFPE tissue.
Conclusion: Taken together, our data offer a window into understanding the clinical utility of universal PCR sequencing in a real-world setting, and may help to guide recommendations and set expectations regarding the value of this emerging technology in clinical practice.
J. Babik, None