WHAT CAN GENOMICS DO FOR SEPSIS?
- Ruth Wright
- Sep 16, 2020
- 5 min read
Sepsis is a life-threatening condition caused by an overreaction of a person's immune system to a infection which causes harm to distant parts of the body. In cases where there is also sepsis-induced organ dysfunction or tissue hypoperfusion the classification is upgraded to 'severe sepsis'. Sepsis can affect ~245,000 people of any age and accounts for more than 48,000 deaths in the UK annually.
With in an acute hospital setting the NICE guidelines recommend treating suspected sepsis patients with broad-spectrum antimicrobials at the maximum safe dose within the first hour of suspecting sepsis. However, reports show that if inappropriate antimicrobials are administered within these first 6 hours, patient survival reduces 5-fold. Around 40% of survivors also experience significant long-term complications which, further increases the economic burden faced by healthcare systems.
A diagnosis is key to implementing effective treatment programs, as it limits complications such as the development of antimicrobial resistance, antibiotic overuse, and ultimately improves patient outcome. Unfortunately, sepsis often presents with cold or flu-like symptoms which, makes it difficult to diagnose.
What is the current strategy?
The current International Sepsis guidelines recommend a blood culture (BC) test be performed before administering any antimicrobial therapy. A BC is viewed as the 'gold standard' for clinical practice while also being a cheap and easy test to perform. However, there are several limitations to BC testing which limit its usefulness.
turnaround time
minimum sample volume
sensitivity and specificity
resistance identification
Turnaround Time: A BC test relies on the growth of the infecting microbial so, the turnaround time will vary depending on the microbial growth rate. Also, some microbes (e.g. Tuberculosis) are ‘unculturable’ meaning they are unable to grow in culture and as such BCs are completely ineffective. Even in the case of faster-growing microbes, BC is unable to provide results within the recommended 1-3hr window for antimicrobial administrations. Consequently, broad-spectrum antimicrobials are administered within the first hour and treatment is only tailored once BC results are available.
Minimum Sample Volume: Another challenge with BC testing is the recommended blood sample volume necessary to perform the test. A single BC test needs one aerobic and one anaerobic inoculation bottle of patient blood. 2 to 4 BC tests are needed before antimicrobial treatment can be tailored so, when 10mL of blood is needed for each inoculation bottle, 40mL to 80mL of blood needs to be sampled to achieve 80-96% detection. Therefore, for pediatric or neonatal patients where blood samples are limited to 1-5mL this approach is not applicable.
In addition to this, a study of BC sample volumes from several counties found considerable variation in sample volume and 13% of the time there was suboptimal volumes obtained. There also appears to be a high contamination rate among positive BS tests.
These points along with others begin to explain the observed low rates of true positives and thus low sensitivity that is often observed with BC tests.
To combat this individual biomarkers are implemented in most current diagnostic approaches. However, these only provide yes or no answers. Other parallel haematological analysis techniques are used in clinical practice but show low sensitivity and specificity, which is why BC remains the 'gold standard'.
The ideal diagnostic test for sepsis was characterized by Sinha et al. in a 2018 review article, which has been illustrated in the above image. There are 9 characteristics including fast turn around time, detection of a broad range of microbes, polymicrobial infections, and unknown/emerging pathogens, etc.
Moving molecular
Moving away from culture-based methods, many rapid molecular technologies have been developed for direct from blood detection of microbial infections. The majority of these are post-culture technologies however, a small selection are pre-culture polymerase chain reaction (PCR) based methods. These PCR methods are not restricted by an organisms’ ability to grow in culture nor the turnaround time of BC. Iridica Plex ID by Abbott Molecular, SeptiFast by Roche Diagnostics, and SepsiTest by Molzyme are 3 such technologies. The image bellow details some of the strengths and limitations (in red) of each of these 3 technologies.
The PCR amplification step in all 3 tests also brings about an increased risk of contamination. This can result in false-positive test results which, in combination with low sensitivity, means negative tests results remain non-actionable. It should also be noted that all 3 platforms are bulky with extremely high up-front costs. Though several steps in each protocol can be automated none of them are fully automated. For these reasons, the application of these molecular tests is difficult in noncentralized clinical settings.
The new world of sequencers
The emergence of Next Generation Sequencing (NGS) and Third Generation Sequencing (TGS) technologies offers potential enhancements to the molecular diagnostic for sepsis. The HiSeq2500 by Illumina is an example of an NGS technology, which has been applied to the diagnosis of sepsis.
Cell-free (cf) DNA are small fragments of DNA that are expelled by a dying cell and are found freely circulating in the bloodstream. When a patient's infection progressed to sepsis the cfDNA in their bloodstream will be DNA from the patient's cells as well as DNA of the pathogen causing the infection. Using cfDNA, Grumaz et al. developed a Sepsis Indicating Quantifier (SIQ) score to discriminate microbial cfDNA from host cfDNA, thus diagnosis sepsis. Illumina’s HiSeq2500 was used to sequence the cfDNA from plasma samples from 7 septic patients and successfully identified gram-positive and gram-negative bacteria in agreement with BC and other non-BC tests. This method also identified species within samples where previous BC tests were negative, suggesting that an NGS-based approach may have better sensitivity and specificity in comparison to standard BC tests. Although, this study only comprised a cohort of 7 patients, more recent research, with larger cohorts, have indicated therapy guided by NGS-based diagnosis may be more beneficial than standard BC tests.
cfDNA has a very short half-life, which means that irrespective of whether cfDNA is released from calls alive or dead, the microbial cfDNA levels will correlate with the infection dynamics. A short half-life also reduces the chance a test will produce a false-positive result. Several studies have also proven the utility of cfDNA sequencing by NGS for identification of antimicrobial resistance. However, the applicability is still limited because currently available databases limited to genes which confer resistance. So, resistance resulting from single base mutations, expression changes, or post-translational modifications will be missed.
However, NGS approaches still have several limitations which prevent it from being used in clinical practice. Firstly, specialized tools and analytical expertise are required. Secondly, there is also yet to be standardization within clinical labs to allow this very precise method to be effectively implemented. Moreover, the greatest rate-limiting step in NGS-based protocols is sequencing (~16 hours) with the entire protocol calculated to take ~24 hours. Lastly, the cost of NGS is substantial, e.g. the HiSeq2000 instrument costs $690,000 UDS with each 30x human genome costing $6,000.
In light of the limitations of NGS technologies, Third-generation sequencing technologies have been create. The MinION by Oxford Nanopore Technologies is a small USB-powered device that achieves single-molecular sequencing by electrophoretically driving DNA/RNA molecules through nanopores. It has shown potential application for pathogen identification directly from blood in 4-6 hours and under 4 hours when in combination with a 16S PCR amplification step. It has also been shown to effectively identify polymicrobial pathogens by sequencing of genomic DNA mixtures containing 20 bacterial strains. The deceased turnaround and small size of the MinION offers the potential for a bedside set up which could greatly improve patient care. The MinION also has a significantly lower up-front cost with the MinION device, 3 flow cells, 2 reagent kits, and the necessary software currently being offered at $1,000 USD.
For application on whole-blood, the technology still requires some improvement including further validation and optimization of bioinformatics pipelines for identification of organisms, resistance genes and/or mutations. It also needs to be automated and standardization to prevent carryover contamination.
Closing remarks
Sepsis is life-threatening and due to non-specific symptoms, it is difficult to diagnose with certainty. Guidelines recommend treatment within a 1-hour window of initial identification however, treatment with inappropriate antimicrobials can significantly reduce patient survival and promote antimicrobial resistance.
Comments