Despite the urgent need for early diagnosis, there is still no single definitive test or clear-cut pathway for diagnosing sepsis.
Here, Simon Walker and Cathal McElgunn, CEO and CSO of MicrosensDx, discuss the limitations of current diagnostic approaches and how emerging biomarkers could help clinicians identify sepsis earlier and improve patient management. And, in a challenging landscape, they explore what it will take to translate the science into the clinical workflow.
Why is sepsis still so difficult to diagnose?
Simon Walker: When you consider that around 11 million people worldwide die from sepsis every year, it’s remarkable that the diagnostic landscape is still where it is. And although many more people survive sepsis, recovery can be long and life-changing, with consequences that can include amputation in the most severe cases.
From a diagnostic perspective, sepsis is incredibly difficult to identify. Clinicians still rely heavily on clinical observations, but the symptoms associated with sepsis are varied and nonspecific. A patient might have shivering, a severe chest infection, or an infected wound. Those symptoms could resolve within a few days, or the patient’s condition could escalate to sepsis. It can be very difficult to predict which way it will go.
The blood tests currently available typically identify inflammation, which can occur for many reasons – from infection and immune disorders to something as simple as strenuous exercise. What’s missing is a way to identify which of those inflammatory presentations is on a dangerous trajectory.
That challenge is compounded by the fact that obvious symptoms may not always be present. There may already be an underlying immune response in which the body’s reaction to infection becomes dysregulated and begins damaging its own organs. That can become systemic and very serious, very quickly.
If we could detect that process earlier and more specifically assess the risk of sepsis escalating, it would give clinicians an opportunity to intervene sooner and get the right treatment in place for patients.
What is the standard of care for sepsis diagnosis, and where do bottlenecks arise?
Simon Walker: Blood tests will typically be performed to assess inflammation and other markers. If infection is suspected, microbiological testing may also be initiated. Blood culture and subsequent identification of the organism can be very accurate, but the process can take 48 hours or more – and sepsis can progress rapidly during that time.
If an organism can be isolated, microbiology can help identify the most appropriate antibiotic to target the infection. But it isn’t always possible to identify a specific causative organism. That leaves us with a multifaceted diagnostic problem: detecting sepsis in the first place, identifying its underlying trigger, and determining the most appropriate treatment for that individual patient.
A test that provides an earlier indication of sepsis risk could allow clinicians to act sooner while the microbiological testing pathway runs in parallel. Ideally, it would also provide an objective measure of where the patient is in their response to sepsis. Does this patient need therapy or organ support escalated? How aggressively do we need to intervene? Or, equally importantly, do we not need to escalate treatment?
That last question can be particularly valuable. Unnecessary treatment carries both clinical and financial costs, and unnecessary antibiotic use contributes to antimicrobial resistance. So, there is a need not only to identify patients who require urgent intervention, but also to give clinicians greater confidence about when treatment may not be necessary.
How can emerging biomarker profiles help with risk stratification of patients with sepsis?
Cathal McElgunn: We already have biomarkers such as C-reactive protein (CRP), lactate, and procalcitonin in routine use, but their specificity in early undifferentiated presentations remains debated. Translating their kinetics into definitive clinical action continues to challenge routine practice. There is therefore a need for biomarkers that can grade severity and provide information about prognosis, disease progression, and response to treatment.
One biomarker we are investigating is the ratio of oxidized pentraxin 3 (PTX3) to total PTX3. Total PTX3 rises with inflammation; the oxidized fraction may carry additional information about the oxidative conditions that accompany a dysregulated host response.
Published research has shown that higher oxidized-to-total PTX3 ratios appear to correlate with more severe sepsis. Importantly, the ratio also appears to decline as patients improve. That raises the possibility of using the biomarker not only for risk stratification but also to follow the course of disease and potentially assess whether a patient is responding to treatment. We are taking the science forward to develop a diagnostic test for sepsis.
How could biomarker-based risk stratification influence patient management?
Cathal McElgunn: In our model, a patient arriving in the emergency department with a markedly elevated PTX3 level may be seriously unwell, which could help identify those needing closer monitoring. Measured serially, the oxidized-to-total ratio may indicate whether the host response is resolving or persisting.
Furthermore, it could help reduce the risk of a patient with nonspecific symptoms being sent home when there is an underlying sepsis process that has not yet become clinically apparent.
The evidence we have so far suggests that the biomarker may have value both in indicating the presence and severity of sepsis and in monitoring the patient’s subsequent response to treatment.
What makes developing an assay for oxidized PTX3 technically challenging, and how far has the work progressed?
Cathal McElgunn: The core technical hurdle is structural selectivity. The oxidized and non-oxidized forms of PTX3 share an identical primary amino acid sequence, so the published work relied on non-reducing Western blotting, which is far too slow and too manual for clinical use. Our development program is directed at an immunoassay format that resolves the same distinction, and we have an antibody pair in optimization with its linearity range established. We are also making progress on sample preparation.
How could your biomarker test help to provide a dynamic picture of treatment response?
Simon Walker: We envisage the biomarker test being used at multiple time points during treatment rather than providing a single snapshot.
If the oxidized-to-total PTX3 ratio is high, it could contribute to a clinician's decision to escalate care, alongside the patient’s other clinical findings. Once treatment has started, the test could then be repeated – perhaps several hours later – to provide another objective measure of how the patient is responding. If the oxidized PTX3 level and ratio have fallen, for example, that could indicate that the treatment is having an effect.
The opposite scenario could also be informative. If PTX3 levels and the oxidized-to-total ratio are low, that could provide additional objective context to reassure clinicians when routine observation is already indicated, rather than driving intervention.
We’re moving toward clinical trials, and that will be critical for establishing how the biomarker behaves across larger patient populations and at different time points.
How quickly could this type of biomarker test deliver a result, and how might it fit into the clinical workflow?
Cathal McElgunn: In the lab setting, we anticipate a turnaround time of under an hour, which fits well with serial monitoring on the ward or in ICU. The assay is intended to be portable to the automated, random-access immunoanalyzers already used in hospital central laboratories. This setup would allow standard acute blood tubes to be run alongside emergency panels with sub-hour turnaround. Beyond that, near-patient formats are imperative for frontline emergency triage.
Simon Walker: Speed is important because clinicians currently have to bring together many different observations – temperature, blood pressure, symptoms, and how the patient is presenting overall. We’ve spoken with clinicians and sepsis survivors about their experiences, and unfortunately there are cases where sepsis simply isn’t recognized early because the initial signs are non-specific.
A rapid test would complement clinical judgment, as it could provide another objective piece of information to consider alongside those observations. If it could be performed routinely and return a result quickly, it could give clinicians an additional tool for assessing whether a patient may be at risk of sepsis.
What will it take to translate that biomarker science into routine sepsis care?
Cathal McElgunn: We need to demonstrate that the biomarker reliably reflects the initial severity of sepsis and can subsequently monitor the patient’s response to treatment. Can it show that a patient is improving? Can it identify somebody whose condition is deteriorating and who may require more intensive care? We now need clinical trials to demonstrate conclusively that the patterns we have observed are reproducible across larger patient populations.
Simon Walker: Our initial feasibility study of around 50 patients next year will be an important technical milestone for assay optimization and clinical correlation. Toward the end of next year and into 2028, we expect to progress to studies involving hundreds of patients.
Alongside comparisons with existing biomarkers, including lactate levels and broader clinical risk scoring systems, a crucial part of that work will be understanding what the results mean across the spectrum of health and disease. We need to study healthy individuals, people with relatively minor illnesses or other causes of inflammation, and patients who are seriously ill with sepsis. One of our clinical partners describes an important group as the “poorly well” – people who feel extremely unwell but are ultimately going to recover without developing sepsis. We need to be able to distinguish those patients from somebody whose condition is progressing toward something much more serious.
While our primary focus is the adult acute-care pathway, the underlying biology could ultimately hold relevance for vulnerable groups such as oncology patients or neonates in future studies.
There are many possible applications, but we need the clinical data before we can establish where the test will provide the greatest value.
What is still holding back progress in sepsis diagnostics?
Simon Walker: Sepsis diagnostics is an incredibly important area, but it often doesn’t receive attention until the consequences become a major public health issue. As developers, we’re experiencing firsthand how challenging it can be to secure the investment needed to translate promising diagnostic science into clinical practice.
Diagnostics also have a wider role to play in hospital efficiency and health system resilience. Sepsis and antimicrobial resistance are closely connected: if we can identify patients who genuinely need antibiotics – and those who don’t – we can use the treatments we already have more responsibly. Better diagnostics can help us intervene earlier while also protecting those treatments for the future. That’s the space we believe this test could ultimately occupy.
Newsletters
Receive the latest pathologist news, personalities, education, and career development – weekly to your inbox.
