993. Rapid phenotypic antibiotic susceptibility testing through RNA detection
Session: Oral Abstract Session: Diagnostics and Why They Matter
Friday, October 6, 2017: 11:30 AM
Room: 05AB

Culture-based antibiotic susceptibility testing, the gold standard, is too slow to guide early antibiotic selection, while newer genotypic methods require comprehensive knowledge of resistance mechanisms to predict phenotype. Quantitative measurement of key antibiotic-responsive transcripts offers a rapid, phenotypic assay for assessing antibiotic susceptibility, agnostic to the genetic basis for resistance.


We performed RNA-Seq on Klebsiella pneumoniae and Acinetobacter baumanii treated with ciprofloxacin, gentamicin, or meropenem for 0, 10, 30, and 60 minutes. For each, we identified 50 responsive transcripts whose expression levels differ most between susceptible and resistant organisms upon antibiotic exposure. We measured their expression using a multiplexed fluorescent RNA hybridization assay (NanoString) in 69 clinical isolates, including a “test set” of multidrug-resistant strains from the CDC, in an 8-hour assay. Gene expression data from test strains were compared against known susceptible and resistant isolates to generate a transcriptional susceptibility metric. We also designed NanoString probes to detect 5 carbapenemase genes (KPC-2, KPC-3, NDM-1, OXA-48, and CTX-M15).


Across all bacteria-antibiotic pairs tested, a susceptibility metric derived from these transcriptional assays correctly grouped isolates in 167 of 173 tests (Table 1), with only 1 of 88 resistant isolates misclassified as susceptible. 5 of 6 incorrectly grouped isolates were within one dilution of the breakpoint MIC, including the misclassified resistant isolate.

Table 1

RNA Signature result




Actual (MIC)











We also detected all 5 targeted carbapenemase genes.

Conclusion: We demonstrate phenotypic antibiotic resistance detection based on fluorescent RNA detection in an 8-hour assay. We have previously published proof-of-concept studies that this assay may be run on a positive blood culture bottle with minimal sample processing. By coupling this phenotypic assay with detection of genetic resistance determinants (demonstrated for carbapenemases) in a single assay, strains with unexplained resistance can be prioritized for further study.

Roby Bhattacharyya, MD PhD1,2, Jamin Liu, BS2, Peijun Ma, PhD2, Nirmalya Bandyopadhyay, PhD2, Jonathan Livny, PhD2 and Deborah Hung, MD PhD2, (1)Department of Medicine, Division of Infectious Diseases, Massachusetts General Hospital, Boston, MA, (2)Broad Institute, Cambridge, MA


R. Bhattacharyya, None

J. Liu, None

P. Ma, None

N. Bandyopadhyay, None

J. Livny, None

D. Hung, None

<< Previous Abstract | Next Abstract

Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 4th with the exception of research findings presented at the IDWeek press conferences.