Current diagnostic methods for sepsis that are in clinical use lack precision, often leading to misdiagnoses. Although microbiological cultures can aid diagnosis, they usually require 2–3 days for bacterial growth and are therefore unsuitable for early diagnosis and treatment. Now, Daniel Irimia and colleagues report that patients with sepsis can be distinguished from those without sepsis based on differential neutrophil motility patterns inside microfluidic devices.

The researchers previously observed a sepsis-specific spontaneous motility signature of isolated neutrophils. They have now developed an assay in which neutrophils from a diluted whole-blood sample autonomously pass through a filter that largely prevents other cells from entering the migration channels. This approach enables the assessment of neutrophil motility in their physiological and biochemical environment, without the need for isolation procedures and chemoattractant gradients. Five motility parameters (neutrophil count, average distance travelled and the frequency of oscillation, pausing and reverse migration events) were identified as characteristic for sepsis using a machine learning-based approach. These parameters were used to develop a scoring system that stratified critically ill patients suspected of sepsis with 98% specificity and 97% sensitivity.

“Our findings suggest that neutrophils have an important role in the pathology of sepsis, which could be exploited using interventions that target neutrophils in addition to current antimicrobial and supportive therapies,” says Irimia.

Next, the researchers plan to investigate the underlying molecular mechanisms of spontaneous neutrophil motility in sepsis, as well as to validate and refine the sepsis score and threshold in a larger and more diverse cohort of at-risk patients.