Better testing holds the promise of improved patient care
By Patrick R. Murray, PhD
Sepsis is the body’s overwhelming and life-threatening response to infection, which can lead to tissue damage, organ failure, and death. In spite of such advances in modern medicine as vaccines, antibiotics, and acute care, this devastating disease takes the lives of approximately 258,000 Americans each year.1 Sepsis remains the primary cause of death from infection.
In the developing world, sepsis accounts for 60% to 80% of the lives lost every year, affecting more than 6 million newborns and children annually. More than 100,000 women contract sepsis in the course of pregnancy and childbirth.
Over the past decade, the number of hospital admissions for sepsis following healthcare-associated infections and community-acquired infections has increased up to three-fold. By comparison, hospital admissions for stroke and myocardial infarction remained stable over the same period.2 The US Agency for Healthcare Research and Quality lists sepsis as the most expensive condition treated in US hospitals, costing more than $20 billion in 2011 and increasing annually by 11.9%.3
The objectives of this article are to increase awareness of this life-threatening disease, to look at the diagnostic tools that are currently available, and to identify what actions are needed to improve diagnosis and decrease sepsis-related morbidity and mortality.
AUTOMATING CULTURE
At a fundamental level, the most effective means of managing septic patients requires early recognition, confirmatory diagnostics, and rapid directed antimicrobial therapy and systemic support. Early recognition must be capable of differentiating between the body’s systemic inflammatory response (SIRS), which may be triggered by a variety of infectious and noninfectious insults, and sepsis—the body’s inflammatory response to an infection. Inflammation markers such as C-reactive protein, procalcitonin, and elevation of neutrophil levels are used to identify patients at increased risk for progressing from SIRS to sepsis. However, these are relatively nonspecific biomarkers, and they may be more useful for assessing the adequacy of antimicrobial therapy in a septic patient.
A more comprehensive approach for the early prediction of sepsis can be achieved by assessing the expression of multiple genes.4 However, it must be recognized that the technical challenges inherent in this approach are significant: multiple gene targets must be measured quantitatively, analytical performance must be sufficient to guide treatment decisions, and the results must be timely enough for intervention to make a difference. These challenges suggest the need for a diagnostic that can be performed both quickly and at the point of care. The ability to differentiate among bacterial, fungal, and viral etiologies would also be an attractive feature, because this information can help refine empiric therapy.
Blood cultures offer an integral diagnostic for confirming sepsis. Guidelines and best practices for preparing blood cultures are well established, including techniques for skin disinfection, the number and timing of collections, volume of blood processed, selection of culture media, and incubation conditions and duration.5,6
Significant improvements in processing blood cultures were observed with the introduction of automated platforms. Currently, three platforms are used in the United States:
- The Bactec FX system by Becton Dickinson, Franklin Lakes, NJ.
- The BacT/Alert 3D and Virtuo systems by bioMérieux, Durham, NC.
- The VersaTrek system by Thermo Scientific, Waltham, Mass.
Each system actively monitors commercial blood culture bottles for the metabolic activity of bacteria and yeasts, providing results significantly earlier than manual methods in which bottles were visually examined once a day. All three companies offer aerobic and anaerobic blood culture bottles. Becton Dickinson and bioMérieux also produce blood culture bottles for pediatric patients and patients receiving antibiotics. Becton Dickinson produces lytic media for the release of intracellular organisms.
TURNING UP THE VOLUME
A number of factors may adversely affect obtaining positive blood cultures in septic patients, including contamination of the culture, the timing and number of blood collections in patients with intermittent septicemia, and the selection of the culture system and media. However, two factors are most influential—the volume of blood collected and the presence of antibiotics. Over the past 40 years, a number of studies have documented the need to process a large volume of blood. More than a decade ago, Cockerill et al. reported a 30% increase in positive cultures when the volume of blood was increased from 10 ml to 20 ml per culture.7
Three methods can be used to measure the amount of blood collected for culture:
- Weighing the culture bottles before and after collection (1 ml of blood weighs approximately 1 gram).
- Measuring the increased height of the blood-broth in the bottle either manually (each ml of blood is equivalent to approximately a 0.4 mm increase in height) or with an automated sensor (such as the bioMérieux Virtuo).
- Measuring the metabolic activity of blood cells in inoculated bottles (Becton Dickinson Bactec FX).
Weighing the culture bottles is the most accurate of these methods, but it isn’t impractical to perform this process manually. Moreover, all three methods measure the volume of blood retrospectively—after the culture has been collected. To ensure that the proper volume of blood is being collected, a prospective method needs to be developed.
Although the bottles are under vacuum to facilitate the inoculation of the culture bottles, this vacuum cannot reliably deliver the required volume of blood because an adequate vacuum must be maintained over the life of the bottle.
Additionally, although it is recommended that blood cultures should be collected before the patient receives antibiotics, the reality is that a significant proportion of patients have blood cultures collected after antibiotics have been initiated. To counteract this problem, in the 1980s Becton Dickinson introduced media supplemented with hydrophobic and ion exchange resins. More recently, bioMérieux has also done so.
Resin-supplemented media are clearly superior to nonsupplemented media or charcoal-supplemented media for patients receiving antibiotics. Nevertheless, it should be recognized that such resins may not remove all classes of antibiotics—failing to clear antibiotics such as the extended spectrum cephalosporins and carbapenems, for instance—so they should be considered a good but not comprehensive solution.8,9
ALTERNATIVE PLATFORMS
An underappreciated problem with blood cultures results from delays between the time the culture is collected and when it is received in the lab, when it is accessioned and placed into the incubation system, and when positive cultures are removed from the incubator and processed. Efforts to eliminate the first two types of delay have been made by placing small incubator systems, such as the Bactec FX40, in high-volume collections sites that can be remotely monitored from the central laboratory (eg, intensive care units, emergency departments).
However, the third problem, of processing delays, has not been adequately addressed. It is recognized that positive blood cultures may remain in the incubator for significant periods of time during busy day shifts or in the evening and night when technologists may not available to process the cultures. The solution to this problem is not simply the automated removal of bottles, but the automated processing of bottles for the identification of the organisms and performance of antimicrobial susceptibility tests (ASTs). Such an automated approach to blood culture processing remains to be achieved.
In the past decade we have seen a paradigm shift in organism identification, moving from biochemical testing to the use of MALDI-TOF mass spectrometry. Through the use of two commercial platforms—the Bruker Biotyper and the bioMérieux Vitek MS—organisms from positive blood cultures can be identified in less than 1 hour and at a fraction of the cost previously encountered.10
An alternative to these platforms is the Abbott Plex-ID. Species-specific housekeeping genes (genes whose products are required for bacterial and fungal survival) are amplified by polymerase chain reactions and then the amplified products are separated and identified by electrospray ionization–time of flight (ESI-TOF) mass spectrometry. Although this system can identify a wide range of organisms, the current version of this platform does not offer advantages over the MALDI-TOF instruments in terms of cost, time to results, or ease of use.
Unfortunately, a dramatic shift in rapid time to results for ASTs has not occurred. Molecular multiplex tests, such as the BioFire FilmArray and Nanosphere Verigene systems, can be used with positive blood culture broths to identify the most common pathogens (23 with the FilmArray BCID panel; 30 with three Verigene panels) and antimicrobial resistance genes (mecA, vanA/B, KPC with FilmArray; mecA, vanA/B, CTX-M, and 5 carbapenemase genes with Verigene).
However, the ability of such systems to identify organisms is more limited than with MALDI-TOF, and their detection of resistance markers is also very limited. Furthermore, these tests identify which antibiotics cannot be used—but not which antibiotics can be used. Resistance to antibiotics such as the beta-lactams can be mediated by the presence of permeability barriers or enhanced efflux of the antibiotic, as well as the presence of beta-lactamases. So the absence of beta-lactamases is insufficient to predict the drug will be active. It is likely that only phenotypic tests will provide a comprehensive susceptibility profile.
The difficulty with current phenotypic tests such as the Beckman Coulter MicroScan, Becton Dickinson Phoenix, and bioMérieux Vitek 2 is that end-points (that is, growth or no growth of the organism in the presence of the antibiotic) are measured after the organism has multiplied sufficiently, so growth is detected by an increase in turbidity or metabolic activity. The ideal next-generation susceptibility test will have to measure the effect of the antibiotic on the organism upon exposure rather than after a series of cell divisions.
Despite an understanding of the pathogenesis of sepsis, knowledge of best practices for collecting and processing blood cultures, and significant improvements in culture systems through the introduction of automated commercial systems, the value of blood cultures for the diagnosis of sepsis is still limited. Only 5% to 15% of collected cultures are positive, and less than half of septic patients have a positive culture.11 Whether this low rate is related to the sensitivity of blood cultures or the intermittent presence of small numbers of organisms in the blood stream is unclear.
Alternatives to growth-based cultures that both detect and identify organisms directly in blood samples include products by both Roche and T2 Biosystems. The Roche SeptiFast system is a probe-based real-time PCR platform that targets ribosomal sequences of the 25 most commonly isolated organisms. It is CE-marked for the European market, but is not cleared by the FDA. Introduction of the Septi-Fast system promised earlier time to results; however, this opportunity was not realized. Because specimens must be processed in batches by highly skilled technologists, the assay time is a minimum of 6 hours, and alternative assays such as culture must still be processed.12
T2 Biosystems developed the T2Candida panel, which is FDA-cleared and CE-marked. The panel is designed to detect five species of Candida directly in blood samples. It does this with high sensitivity and specificity, with a 3- to 5-hour processing time, and is superior to traditional culture techniques.13 The one difficulty that remains with this product is that this also is an assay that must currently be complemented with culture-based systems for detection of non-Candida yeasts (eg, Cryptococcus) and bacteria.
CONCLUSION
When it comes to diagnosing sepsis, today’s clinical laboratories have access to sophisticated, automated blood culture systems that have significantly improved detection and time to results. We have moved from biochemical identification to mass spectrometry and probe-based molecular methods. And we have introduced molecular probes for detecting antimicrobial resistance markers, and for detecting and identifying selected organisms directly in blood samples.
What we have not done is realized the goal of physicians caring for septic patients. We do not have biomarkers that permit accurate identification of septic patients in the early phase of their disease, and we cannot provide specific diagnostics to guide initiation of therapy. These are challenges that will require innovative solutions.
But there are short-term changes that can significantly improve patient outcomes. We can improve collection of the proper volume of blood, further eliminate the antimicrobial effects of antibiotics, and use automation and molecular tools to eliminate the delays currently associated with processing positive blood cultures. The path forward may be long, but significant achievements are within our grasp.
Patrick R. Murray, PhD, is senior director of worldwide scientific affairs at BD Life Sciences. For further information contact CLP chief editor Steve Halasey via shalasey@allied 360.com.
REFERENCES
- What is sepsis? [online]. San Diego: Sepsis Alliance, 2015. Available at: www.sepsisalliance.org. Accessed August 24, 2015.
- Sepsis facts [online]. Jena, Germany: Global Sepsis Alliance, 2015. Available at: http://world-sepsis-day.org/?met=showcontainer&vprimnaviselect=3&vseknaviselect=1&vcontainerid. Accessed August 24, 2015.
- Torio CM, Andrews RM. National inpatient hospital costs: the most expensive conditions by payer, 2011. HCUP statistical brief no. 160. Healthcare Cost and Utilization Project Statistical Briefs [online]. Rockville, Md: US Agency for Healthcare Research and Quality, 2013. Available at: www.hcup-us.ahrq.gov/reports/statbriefs/sb160.jsp. Accessed August 24, 2015.
- Sweeney TE, Shidham A, Wong HR, Khatri P. A comprehensive time-course-based multicohort analysis of sepsis and sterile inflammation reveals a robust diagnostic gene set. Sci Transl Med. 2015;13(7):287ra71; doi: 10.1126/scitranslmed.aaa5993.
- Baron EJ, Weinstein MP, Dunne WM, Yagupsky P, Welch DF, Wilson DM. Cumitech 1C, Blood Cultures IV. Baron EJ, ed. Washington, DC: American Society for Microbiology Press, 2005.
- Wilson ML, Mitchell M, Morris AJ, et al. Principles and procedures for blood cultures; approved guideline. CLSI document M47-A. Wayne, Pa: Clinical and Laboratory Standards Institute, 2007.
- Cockerill FR 3rd, Wilson JW, Vetter EA, et al. Optimal testing parameters for blood cultures. Clin Infect Dis. 2004;38(12):1724–1730.
- Zadroga R, Williams DN, Gottschall R, et al. Comparison of 2 blood culture media shows significant differences in bacterial recovery for patients on antimicrobial therapy. Clin Infect Dis. 2013;56(6):790–797; doi: 10.1093/cid/cis1021.
- Kirn TJ, Mirrett S, Reller LB, Weinstein MP. Controlled clinical comparison of BacT/alert FA plus and FN plus blood culture media with BacT/alert FA and FN blood culture media. J Clin Microbiol. 2014;52(3):839–843; doi: 10.1128/jcm.03063-13.
- Clark AE, Kaleta EJ, Arora A, Wolk DM. Matrix-assisted laser desorption ionization–time of flight mass spectrometry: a fundamental shift in the routine practice of clinical microbiology. Clin Microbiol Rev. 2013;26(3):547–603; doi: 10.1128/cmr.00072-12.
- Murray PR, Masur H. Current approaches to the diagnosis of bacterial and fungal bloodstream infections in the intensive care unit. Crit Care Med. 2012;40(12):3277–3282; doi: 10.1097/ccm.0b013e318270e771.
- Warhurst G, Dunn G, Chadwick P, et al. Rapid detection of healthcare-associated bloodstream infection in critical care using multipathogen real-time polymerase chain reaction technology: a diagnostic accuracy study and systematic review. Health Technol Assess. 2015;19(35)1–142; doi: 10.3310/hta19350.
- Mylonakis E, Clancy CJ, Ostrosky-Zeichner L, et al. T2 magnetic resonance assay for the rapid diagnosis of candidemia in whole blood: a clinical trial. Clin Infect Dis. 2015;60(6):892–899; doi: 10.1093/cid/ciu959.