Why sepsis attacks the barriers that protect us

---

Reference: October 2026 | Issue 10 | Vol 12 | Page 55

---

Sepsis is a disorder of the host response in which infection triggers processes that damage organs and impair recovery. Inflammatory activation and ineffective antimicrobial immunity can coexist, making the description of a cytokine storm incomplete. Endothelial and epithelial barriers provide a useful focus for understanding how this dysregulated response becomes tissue injury.

Glycocalyx degradation, disruption of cellular junctions, and interactions between platelets, neutrophils, and coagulation can undermine vascular integrity, while epithelial injury compromises organ-specific functions. These observations support investigating barrier protection as an adjunct to antimicrobial treatment and organ support. This article examines the rationale, selected experimental evidence, and the patient selection and safety questions that must be resolved before barrier-directed approaches can enter clinical practice.

Why treating infection may not prevent organ dysfunction

If we consider a 67-year-old woman with community-acquired pneumonia. She is tachypnoeic, hypoxaemic, and newly confused. Following urgent referral, she receives antimicrobial treatment and physiological support. Over the next 24 hours, however, her oxygen requirement increases, urine output falls, and hypotension develops. She requires critical care despite treatment that appears appropriate for the infection.

This scenario introduces a familiar question. How does an infection in one organ become dysfunction of several organs? Persistent infection, inadequate antimicrobial coverage, and complications remain possible explanations. But deterioration also directs attention towards the biological response to infection and the tissue structures that response can damage.

Sepsis is defined as life-threatening organ dysfunction caused by a dysregulated host response to infection.¹ Bloodstream infection is not required. The definition invites us to consider mechanisms through which the host response undermines organ function, as well as the microorganisms that initiate illness. One potentially modifiable link is the failure of tissue barriers. The argument for targeting these barriers is that preserving their integrity might limit the conversion of immune activation into organ injury while antimicrobial treatment addresses the infection. This is a rationale for therapeutic investigation, with established sepsis care remaining essential.²

Immune dysregulation

Recognition of microbial components activates immune pathways that recruit inflammatory cells and support pathogen clearance. This response needs to occur in the appropriate location and at an intensity the surrounding tissue can tolerate. Injured cells release additional danger signals, allowing tissue damage itself to sustain inflammatory activation.

The description of sepsis as a cytokine storm captures excessive inflammation but gives an incomplete account of immune function. In the Genetic and Inflammatory Markers of Sepsis (GenIMS) study, patients hospitalised with pneumonia frequently had simultaneous proinflammatory and anti-inflammatory cytokine responses. High concentrations of both interleukin-6 and interleukin-10 were associated with the greatest mortality risk.³

These measurements describe inflammatory activity; they do not, by themselves, establish functional immunoparalysis. Human tissue studies also demonstrate impaired immunity. Boomer and colleagues identified reduced immune-cell responsiveness, lymphocyte depletion, and increased inhibitory signalling in patients who died with sepsis.⁴ Although these findings cannot define the immune state of every patient, they help explain why inflammatory activation and ineffective antimicrobial defence can coexist.

The therapeutic problem is therefore more complex than reducing inflammation. A treatment must address a damaging process without aggravating an existing deficit in host defence. Tissue barriers offer a complementary focus because they are the interfaces at which circulating inflammatory signals, immune cells, and organ physiology meet. Preserving these interfaces could help maintain tissue function even while the immune response remains dysregulated.

Vascular barrier injury

The endothelium forms the interface between blood and tissues. It regulates permeability and leukocyte recruitment while contributing to vascular tone and haemostasis. During infection, changes in endothelial behaviour can affect several of these functions together. Its role in sepsis extends beyond being a passive surface exposed to circulating cytokines.

The endothelial glycocalyx is a carbohydrate-rich layer on the vessel’s luminal surface. It contributes to vascular homeostasis and regulates interactions between circulating cells and the vessel wall. Damage to this layer can change the accessibility of endothelial adhesion molecules, creating conditions that favour inflammatory-cell attachment.

Schmidt and colleagues demonstrated that heparanase-mediated degradation of the pulmonary endothelial glycocalyx promoted neutrophil adhesion and lung injury in experimental sepsis.⁵ The significance is the feedback this identifies. Inflammation damages the endothelial surface, and that damage facilitates further inflammatory recruitment. Barrier injury can consequently participate in amplification of the response.

Cellular junctions provide another point of vulnerability. VE-cadherin is an important component of endothelial adherens junctions. Changes in signalling and junctional organisation can increase permeability, allowing fluid and plasma proteins to move into surrounding tissues. In experimental SARS-CoV-2 studies, Nader et al linked αVβ3 integrin signalling to disruption of endothelial junctions and showed protection with integrin antagonism. These findings establish a mechanistic example in viral infection that holds true in bacterial induced sepsis also.⁶ This biology gives capillary leak a more precise meaning.

The structures controlling movement between blood and tissue have become dysfunctional. The resulting oedema can compromise tissue function, although permeability injury is one contributor among several, including altered vascular regulation and cellular dysfunction. For therapeutic development, the distinction between endothelial activation and destructive barrier failure matters. Leukocyte trafficking is necessary for infection control. The goal should be to preserve regulated barrier function while allowing an effective local immune response, rather than preventing all endothelial responses to infection.

Epithelial dysfunction

A focus on tissue barriers should include the epithelium as well as the endothelium. The lung illustrates why. Its alveolar-capillary interface separates circulating blood from the airspaces in which gas exchange occurs. Endothelial permeability and alveolar epithelial function influence different aspects of fluid movement and recovery. Protecting one side of this interface may be insufficient if the other remains severely injured. Alveolar epithelial function includes clearing fluid from the airspaces.

In a study of 79 patients with acute lung injury or acute respiratory distress syndrome, Ware and Matthay found impaired alveolar fluid clearance in the majority. Maximal clearance was associated with lower mortality and a shorter duration of mechanical ventilation.⁷ This observational study does not prove that restoring clearance improves outcomes, but it highlights the clinical relevance of epithelial function alongside vascular permeability.

For a patient with pneumonia, this widens the therapeutic question. Could an adjunctive treatment help preserve the interface needed for gas exchange and the epithelial functions needed to resolve oedema? Answering this requires measurements of both permeability and tissue function. A reduction in one marker of vascular injury would not establish recovery of the complete alveolar-capillary barrier.

The pulmonary example also helps define the limits of a broad barrier argument. Different organs have different epithelial and endothelial arrangements. A mechanism demonstrated in the lung cannot automatically be assumed to operate identically in the kidney or brain. The case for tissue barrier protection is strongest when the affected interface and its contribution to organ injury are specified.

Platelets and neutrophils

Barrier dysfunction occurs within a wider immune and haemostatic response. Platelets and neutrophils can cooperate to contain microorganisms, but these interactions may also promote vascular obstruction when they become extensive or poorly controlled. Clark and colleagues demonstrated that platelet Toll-like receptor 4 signalling promoted neutrophil extracellular trap formation in septic blood.⁸ These traps, commonly called NETs, contain DNA, histones, and antimicrobial proteins. They can capture microorganisms and provide a scaffold for coagulation.

McDonald and colleagues subsequently demonstrated NET-platelet interactions promoting thrombin generation and microvascular occlusion in mouse sepsis models.⁹ These studies connect an antimicrobial mechanism with a potential loss of tissue perfusion. They also caution against treating all immunothrombosis as pathological – its local containment function is part of host defence. Endothelial injury, permeability changes, and coagulation disturbances need to be considered together when evaluating an intervention’s effects.

Human microcirculatory imaging has shown reduced and heterogeneous small-vessel perfusion in sepsis, with more marked abnormalities in patients with poorer outcomes.¹⁰ This helps explain why a satisfactory blood pressure may coexist with impaired perfusion within tissues. It does not establish that barrier protection alone will correct those abnormalities. Trials should assess whether preserving vascular integrity translates into improved organ function, rather than assuming that less leakage necessarily means better perfusion.

The case for barrier protection

The evidence supports treating barrier failure as a candidate therapeutic target. It identifies vulnerable structures, mechanisms of injury, and experimental opportunities to modify them. A barrier-directed approach could complement pathogen clearance by preserving the tissue environment in which recovery must occur. The angiopoietin-Tie2 pathway provides one example.

Han and colleagues developed an angiopoietin-2-binding antibody that activated the endothelial receptor Tie2. In three mouse sepsis models, the intervention protected vascular integrity and improved survival, with effects on the glycocalyx, vascular leakage, and organ damage.¹¹ This is experimental evidence that modifying endothelial signalling can influence the course of sepsis. Whether those effects can be reproduced safely in patients remains a clinical question.

The glycocalyx offers another mechanistic focus. The heparanase experiments demonstrate that degradation of the endothelial surface can contribute to inflammatory recruitment and lung injury.⁵ They support investigating ways to limit destructive glycocalyx loss. They do not establish that replacing glycocalyx components or inhibiting degradation will provide the same benefit in human sepsis. Integrin signalling provides a further route through which infection can affect cellular junctions. Our experimental work on αVβ3 and endothelial barrier disruption in SARS-CoV-2 infection illustrates the possibility of interrupting an injury pathway at the tissue interface.⁶ It sits within the broader case for barrier protection and should be evaluated alongside other approaches. Evidence from a particular pathogen and experimental model cannot establish efficacy across the heterogeneous sepsis population.

The proposed advantage of barrier protection is preservation of organ function during infection. Its compatibility with antimicrobial defence must be demonstrated. Endothelial and epithelial signalling also supports normal trafficking and repair, so dose, treatment duration and biological context could determine whether an intervention is protective or harmful.

Experience with other host-directed approaches reinforces the need for clinical testing. In the SCARLET randomised trial, recombinant soluble thrombomodulin did not significantly reduce 28-day mortality in patients with sepsis-associated coagulopathy.¹² This was a coagulation-directed intervention, rather than a direct test of junctional protection, but it illustrates the distance between a compelling vascular mechanism and an effective sepsis treatment. The appropriate claim is therefore specific.

Barrier protection merits investigation as an adjunctive therapeutic strategy because barrier injury contributes to processes relevant to organ dysfunction. The magnitude of that contribution, its reversibility, and the consequences of modifying it need to be established in the patients selected for treatment.

Identifying patients and treatment window

A barrier-directed treatment should be tested in patients whose illness involves the mechanism it is intended to modify. Sepsis includes different organisms, infection sites, and stages of injury. Host-response genomic studies in pneumonia-associated sepsis have identified biological subgroups with different characteristics and outcomes.¹³ A common syndrome definition does not guarantee a common treatment-responsive state.

Biomarkers may help, but prognosis and treatment selection are different tasks. In a cohort of 757 patients with sepsis, higher early plasma angiopoietin-2 concentrations were associated with acute respiratory distress syndrome and mortality.¹⁴ This supports investigating endothelial biomarkers for trial enrolment. It does not establish a threshold at which Tie2-directed or other barrier therapies will work.

Soluble urokinase plasminogen activator receptor, or suPAR, has also been studied for prognostic assessment alongside APACHE II.¹⁵ Its prognostic value does not establish that it specifically identifies barrier failure or predicts benefit from integrin modulation. A companion diagnostic would need to demonstrate that it helps identify patients who benefit from a particular treatment.

Timing deserves equal attention. Preventing junctional disruption before extensive organ injury and promoting recovery after established damage are different therapeutic aims. Trials should define which aim they are testing, confirm target engagement, and measure patient-centred outcomes such as organ-support requirements and survival. Improved laboratory permeability or a lower biomarker concentration can support a mechanism without establishing clinical benefit.

The sepsis research agenda

Returning to the patient with pneumonia, barrier biology offers a way to connect increasing oxygen requirements and tissue oedema with the host response. It also suggests a therapeutic question that is relevant before infection-associated injury becomes extensive – can vulnerable tissue interfaces be preserved while infection is treated?

Current management still depends on prompt antimicrobial treatment, source control where indicated, repeated assessment, and appropriate organ support.² Investigational barrier therapies would be added to that care. They should be judged on whether they improve meaningful outcomes without compromising infection control or recovery.

The case for prioritising this work rests on the position of tissue barriers between immune activation and organ function. Understanding immune dysregulation is essential, but preserving the structures through which that response affects tissues may offer an additional opportunity to prevent injury. Barrier protection deserves a defined place in sepsis research, supported by evidence of the relevant mechanism, a realistic treatment window, and trials designed to establish benefit for patients.

Key messages

• Immune activation and impaired antimicrobial defence can coexist in sepsis.
• Endothelial and epithelial dysfunction provide complementary targets for understanding infection-associated organ injury.
• Experimental studies support investigating barrier protection alongside treatment of infection.
• Successful translation requires patient selection, evidence of target engagement, and preservation of effective host defence.
• Clinical benefit must be demonstrated through meaningful patient outcomes.