To the Editor—Rapid blood-culture identification techniques are increasingly common in hospitals across the United States. Rapid diagnostics can quickly identify resistance genes in bacteria that may otherwise have taken days to be identified, thus shortening the time until patients are placed on the appropriate transmission-based precautions and antimicrobial treatment. Because shortened time to appropriate therapy has been linked to improved mortality, this change has critical implications for patient care. ^{Reference Falcone, Bassetti and Tiseo1} Although numerous studies have described early identification of resistance genes that are later confirmed by traditional antimicrobial susceptibility testing (AST), few have reported cases of gram-negative bacteria with resistance genes identified by rapid tests but with AST showing antibiotic susceptibility. We present a case of a patient with a Klebsiella pneumoniae carbapenemase (KPC) gene identified by the ePlex Blood Culture Identification (BCID) panel (GenMark, Carlsbad, CA) whose subsequent AST showed susceptibility to carbapenems.

A man aged in his thirties with a medical history of necrotizing pancreatitis and type 2 diabetes mellitus was admitted to the intensive care unit for abdominal pain and diabetic ketoacidosis. Blood cultures collected on admission became positive 12 hours later for gram-negative rods. A known abdominal fluid collection was postulated as the source of his bacteremia. BCID detected Klebsiella pneumoniae and a KPC gene, and the patient was placed on contact precautions. Subsequent AST using the Vitek2 GN74 card (bioMérieux, Durham, NC) showed Klebsiella pneumoniae with susceptibility to meropenem (minimum inhibitory concentration [MIC] ≤ 0.25 µg/mL) and ertapenem (MIC ≤ 0.5 µg/mL) (Supplementary Table 1 online). Given this discrepancy, the isolate was sent to our investigational clinical microbiology core, where testing by Carba NP (performed according to guidelines ² ) showed carbapenemase activity. Gradient diffusion testing revealed a main population of bacteria susceptible to meropenem; however, satellite colonies were noted within the zone of inhibition (Fig. 1). Further testing in a research laboratory showed heteroresistance to meropenem. The patient was transitioned to ceftazidime-avibactam, and his blood cultures cleared within 48 hours.

Figure 1. Gradient diffusion susceptibility test of meropenem showing colonies within the zone of inhibition.

This phenomenon (presence of resistance genes on PCR, and antibiotic susceptibility on AST) has been well documented in Staphylococcus and other gram-positive species. ^{Reference Huang, Melnik, Bogaerts, Evrard and Glupczynski3} Suggested mechanisms for this include empty mec cassettes or multiple populations of bacteria yielding conflicting results. Few prior studies have mentioned similar discrepancies in Enterobacterales. ^{Reference Tansarli and Chapin4–Reference Marschall, Tibbetts, Dunne, Frye, Fraser and Warren7} Bratu et al ^{Reference Bratu, Mooty and Nichani5} described a multihospital outbreak of carbapenem-resistant Enterobacterales (CRE) in which many isolates appeared susceptible to imipenem (despite possessing KPC genes) when a lower inoculum was used during standard broth microdilution testing. ^{Reference Bratu, Mooty and Nichani5} Other studies suggest that when using more modern AST techniques, an unexpressed carbapenemase gene may lead to apparent susceptibility on standard AST despite presence of the resistance gene. ^{Reference Tansarli and Chapin4,Reference Emira, Madkour, Seif and Dwedar6} Our findings suggest that heteroresistance, whereby an established subpopulation of resistant bacteria proliferates under antibiotic pressure, may also play a role in these discrepant results.

The Centers for Disease Control and Prevention (CDC) recommends that healthcare workers use gowns and gloves when caring for patients with CRE and that these patients be placed in single-bed rooms when available (ie, contact precautions). Prior to instituting BCID, this patient would not have been identified as harboring CRE and would not have been placed under contact precautions. Surveillance for CRE in cultures, along with isolating and placing patients with CRE under contact precautions, have been shown to reduce the transmission of this class of organisms. ^{Reference Tomczyk, Zanichelli and Grayson8,Reference Ben-David, Masarwa and Fallach9} Our findings suggest that relying solely on AST to guide isolation decisions may miss some carbapenemase-producing CRE, potentially increasing the chance of undetected patient-to-patient transmission. As rapid diagnostics become more prevalent, more discrepancies between gene detection and AST will be identified. Further analysis should be undertaken to determine the transmission risk of unexpressed carbapenemase genes and their implications for infection control and prevention.