Driving Rapid AST by Blood Culture ID PCR

We are witnessing a change in the current standard of care related to antimicrobial susceptibility testing (AST), shifting from the traditional 16-24 hour colony-based AST, where bacteria must first be cultured on solid media before susceptibility testing, to direct testing from positive blood cultures, which bypasses the need for prolonged incubation and accelerates results.

The arrival of direct blood culture AST solutions is challenging our current standards by necessitating adjustments in laboratory workflows, validation protocols, and clinical interpretation. These solutions introduce the need for robust quality control, potential limitations in detecting polymicrobial infections, and ensuring accurate phenotypic correlation. However, it seems to be only a matter of time before they become the default method for routine blood culture workup. Although there are a few significant caveats, such as polymicrobial growth requiring mitigation plans through a continuous quality assurance process, including culture growth to capture the infrequent scenario when a secondary organism escapes detection by PCR, rapid AST may co-exist with rapid blood culture ID (BCID). The presence of absence of resistant genes through BCID, appropriated validated and corelated with phenotypic methods, should inform if a rapid AST is required as part of the Laboratory Stewardship program.

Gram-negative BCID panels are comprehensive enough for identifying the most prevalent Enterobacterales and the most significant non-fermenters. Among these targets, ESBL (CTX-M) and carbapenemase genes such as KPC, NDM, OXA48, VIM, and IMP are commonly included in the list of PCR targets. This comprehensiveness, along with its rapid turnaround time, is crucial for determining when to implement rapid AST.

With appropriate correlation and validation, the prevalence of CTX-M in E. coli, K. pneumoniae, K. oxytoca, and P. mirabilis can be determined. By correlating the presence of CTX-M with a confirmatory test, such as double-disk diffusion, its prevalence relative to SHV/TEM genotypes, and even AmpC, can be established at the species level.

This data has been analyzed and is being used to drive our implementation of rapid AST in blood cultures at AdventHealth. We have identified that approximately 3% of ESBLs isolated from blood cultures do not carry a CTX-M gene, establishing a CTX-M prevalence of 97% in the four Enterobacterales species recommended for testing by the CLSI. Further review demonstrated variation in the prevalence of non-CTX-M ESBLs among these species: E. coli (2.9%), K. pneumoniae (3.1%), P. mirabilis (2.2%), and K. oxytoca (100%). It is important to emphasize that although all K. oxytoca ESBLs are non-CTX-M, their presence is minimal, representing less than 1% of all ESBLs detected in blood cultures.

Understanding the prevalence of circulating ESBL genes presents a critical opportunity to enhance the effectiveness and clinical impact of BCID testing. The next essential data point involves assessing the prevalence and correlation of carbapenemase genes to further refine treatment decisions.

We have established the correlation between KPC, NDM, and OXA48 among Enterobacterales. Although there are non-carbapenemase producing (non-CP-CRE) isolates, these have not been identified in blood cultures across our facilities. The presence and absence of these genes have correlated accurately at 100% with phenotypic CRE patterns using colony-based AST.

So, what does all this mean? And how is this information used?

Prevalence and Cumulative Susceptibility for Resistant Genes 

This information can guide the development of a structured multistep algorithm to optimize testing and ensure adherence to Laboratory Stewardship initiatives.

Rapid and direct AST from blood culture is not inexpensive, especially when compared to traditional AST panels. However, its implementation can be optimized to be cost-effective, rational, and clinically meaningful, beginning with the correlated and cumulative antimicrobial susceptibility data.

If 30% of E. coli in blood cultures are ESBL producers and the presence of a carbapenemase gene correlates phenotypically at 100%, then the detection of a CTX-M gene and the absence of a carbapenemase gene in E. coli should routinely exclude the need for a rapid AST panel if hospital guidelines recommend carbapenem therapy, such as meropenem. Since the absence of a carbapenemase gene correlates at 100%, this information should be leveraged to rapidly guide antimicrobial escalation. 

With a similar approach, the presence of KPC should indicate treatment with agents such as meropenem/vaborbactam, ceftazidime/avibactam, or imipenem/relebactam, which collectively have a cumulative potency of 98% against KPC-producing Enterobacterales in our institutions. The absence of NDM and OXA48 genes is considered reliable based on historical data demonstrating a high negative predictive value (NPV).

In these circumstances, traditional AST after 16-24 hours of incubation is still required for comprehensive susceptibility testing and additional agents if necessary. The rapid AST solution in this scenario will not provide significant added value unless the epidemiology of these resistances is unknown or not well defined, or if access to anti-KPC agents is limited beyond the rapid AST panel.

Rapid and Direct Blood AST, the Perfect Tool for De-Escalation and IV-to-Oral Conversion 

In 70% of E. coli cases that are non-ESBL and non-CP-CRE, this method demonstrates its greatest utility. In a positive blood culture where BCID indicates the presence of E. coli without any detected resistance genes, there is a 97% likelihood of it being a non-ESBL strain and a 100% likelihood of it being non-CRE. In this scenario, rapid AST is warranted for two main reasons:

• Since susceptibility to ampicillin/sulbactam and cefazolin cannot be inferred by resistance genes alone, rapid AST provides a fast 5-7 hour result with actionable data supporting de-escalation from third- or fourth-generation cephalosporins or piperacillin/tazobactam. Additionally, determining ciprofloxacin susceptibility may enable IV-to-PO conversion. It is crucial to recognize that patient clinical condition and risk assessment play a significant role in guiding de-escalation, and these culture results alone should not be the sole decision-making factor.
• Since 2.9% of E. coli ESBL cases are not CTX-M mediated, susceptibility testing for cefoxitin, ceftriaxone, ceftazidime, and cefepime provides an additional level of phenotypic screening and confirmation for the presence of SHV/TEM and AmpC-producing E. coli.  

A similar approach for each detected organism by the BCID can be designed and codified into a protocol. In cases of an E. cloacae complex, where the prevalence of CTX-M is extremely low, but they are known to carry a clinically significant AmpC, the absence of a carbapenemase gene would not warrant a rapid AST, as cefepime remains the standard treatment for non-carbapenemase-producing E. cloacae complex.

For organisms such as S. marcescens, where the expression of AmpC is less prevalent than in other Enterobacterales, the absence of a carbapenemase gene warrants rapid AST with an expanded panel, including ceftriaxone and piperacillin-tazobactam, to support possible de-escalation from cefepime following an appropriate risk assessment. 

BCID and Rapid AST the Perfect Duo

The arrival and expansion of rapid and direct AST from positive blood cultures, utilizing microscopic visual bacterial reaction and metabolic growth analysis, will necessitate a revision of our traditional broth microdilution methods. We must evaluate whether single-cell observation or metabolic analysis is as accurate or even more accurate than our current reference AST methods.

Rapid AST should be integrated from both a Laboratory and Antimicrobial Stewardship perspective. It should be combined with a rapid identification method to maximize its utility and provide clinically actionable information efficiently.

Routine rapid AST for all Gram-negative cultures should be reserved for situations where the prevalence of resistance and organism variability is unknown, or when phenotypic correlation is unreliable. In other situations, when BCID resistance data is available and integrated into our laboratory procedures and protocols, it becomes a powerful driving force that not only impacts patient care and safety but also aligns with current Laboratory Stewardship efforts.

#microbiology #clinicallaboratory #pharmacy #antimicrobialstewardship #laboratorystewardship #antimicrobialsusceptibility #antimicrobialresistance #infectiousdiseases #clinicallabscientists #infectionprevention #infectioncontrol #epidemiology #mdro #amr #rapidast #pcr #rapiddiagnostic #moleculardiagnostic #medicalaffairs #medicalscienceliaison #ceo #cmo #cxo #leadership #molecularepidemiology #escmid #idsa #asm #cdc #bcid