Bloodstream infections leading to sepsis are a leading cause of morbidity and mortality in the United States. Defined as a “life-threatening organ dysfunction caused by a dysregulated host response to infection,” sepsis afflicts almost two million Americans annually leading to over 270,000 deaths, accounting for as many as a quarter of all hospital deaths.
For septic patients, prompt intervention with the appropriate antimicrobial is known to be a key determinant of outcomes. In order to select the most appropriate antimicrobial, clinicians must first identify the infection’s etiological agent. However, due to numerous technical challenges, including the very low concentrations of the infecting pathogen (often only a few cells in a milliliter of blood), current diagnostic approaches require an initial culturing step to grow a sufficient number of cells for downstream processes to identify the pathogen and assess its antimicrobial susceptibility. Unfortunately, this initial culture step suffers from several weaknesses.
First and foremost, cultures are slow, often requiring days for results. Additionally, some infections fail to grow, particularly if antimicrobial treatment has already been applied. Numerous studies, including the recently completed FABLED study, have shown that prior antimicrobial use reduces culture sensitivity by around 50%. Fastidious microorganisms, host-interactions, and mixed infections all present additional challenges and thus even for patients with severe sepsis or septic shock, blood cultures are only positive roughly half the time.
With the probability of survival reduced by almost 10% hourly in the absence of appropriate antimicrobials, clinicians cannot wait for cultures. Treatment with multiple broad-spectrum antimicrobials is generally initiated in the absence of microbial diagnostic confirmation. Unfortunately, these antimicrobials may miss the target in up to a third of cases. Furthermore, the extensive use of antimicrobials has been shown to increase the likelihood of antibiotic-associated adverse drug events, including C. difficile infections. Finally, this approach exemplifies poor antimicrobial stewardship supporting the rise of antibiotic-resistant strains and jeopardizes our antimicrobial arsenal.
Though imperfect, culture-based diagnostics are tremendously powerful. Once an infection has been characterized, clinicians can select the most appropriate antimicrobial while de-escalating unnecessary treatments. This can improve care, reduce cost, and improve antimicrobial stewardship. Indeed, it has been demonstrated that the antimicrobial intervention is modified from the initial, empiric intervention, in roughly two-thirds of cases once microbiology is available.
However, considering the limitations of cultures, significant efforts have been made to develop culture-free diagnostics. While the majority of these approaches have only begun (if at all) to enter the market, they are anticipated to support improved patient care and improved antimicrobial stewardship. Broadly speaking, these efforts can be broken down into three approaches:
(1) Interpreting our body’s physiological response to the presence of an infection; commonly referred to host response,
(2) Direct sequencing of DNA in blood, and
(3) Direct detection pathogenic microorganisms from patient blood.
The body’s host response can serve as an amplified signal of an infection and biological markers are in abundance and readily identifiable. As a result, tests that measure these biomarkers are often low-cost, of relatively low-complexity, and have a relatively fast turnaround time. This approach may inform us of the presence of the infection, and in some cases, may also indicate the nature of the infection (bacterial, fungal, or viral). Procalcitonin, white-blood cell counts/morphology, and mRNA expression profiling stand out as biomarkers of wide interest. While these strategies have cost and speed advantages, they do not identify the infecting pathogen; critical information for guiding antimicrobial treatment. As the performance of these tests improves, we anticipate that they could be more widely for use as a screening tool, allowing clinicians to quickly assess for the presence of a bloodstream infection prior to going down the path of tests that, while more informative, have a somewhat longer turnround time.
Leveraging impressive advances made in sequencing, approaches utilizing next-generation sequencing (NGS) of DNA in blood have recently been developed. While potentially providing a “hypothesis-free” test menu, these tests face multiple challenges. First, due to the complexity of sample-preparation and NGS, these tests are not available on-site. Blood samples must be shipped to a dedicated lab (1 day) prior to processing (up to 3 days), resulting in a turnaround time too slow for immediate intervention. Further, by detecting any and all DNA in the blood rather than only microbial intracellular DNA, this approach is apt to find DNA not associated with an ongoing infection. Thus, while in some cases sensitivity may be high (>90%), specificity is low when compared to culture (40%). With improvements in concordance to clinical standards, we believe this approach will be well suited to deconvolve complicated cases where a definite diagnosis has been elusive.
The direct detection of pathogens from blood could provide critical microbiological information within the timeframe to impact care. Unlike the broad-based sequencing of the DNA found in blood, targeted detection of these pathogens via their genomic material would be more in-line with today’s culture-based standards. To date, though, direct detection techniques have demonstrated limited test menu size or insufficient sensitivity. Better sample preparation and detection strategies are needed. The relatively large volume of blood that must be sampled to ensure the detection of infections at low concentrations overwhelms and inhibits enzymatic amplification processes for DNA such as PCR. Meanwhile, the analytical specificity of common DNA-based detection techniques is often not sufficient to discriminate between broad menus of potential pathogens. As these technologies improve, we expect direct detection approaches to serve as a complement to host response, assisting clinicians to rapidly identify appropriate antimicrobials.
In the coming years, as the approaches described above advance and mature, clinicians will be presented with a larger and more capable toolbox to aid in the diagnosis of sepsis; directly overcoming the drawbacks of culture-based diagnostics. Better host response tests will help identify septic patients sooner. Direct detection of infecting pathogens will enable a timely the transition to appropriate antimicrobials. DNA sequencing will help identify rare or complicated infections. Together these technologies will help reduce the immense burden of sepsis.