Black swans or red herrings - Inflammatory derangement after cardiac arrest

Could the post-inflammatory state that follows cardiac arrest be a mechanistic black swan? In this issue of Resuscitation, Langeland and colleagues help us understand the potential scope of inflammation after cardiac arrest, with their description of the post-cardiac arrest inflammatory milieu in relationship to arrest characteristics, hemodynamic status, and short-term outcomes¹.

In this prospective observational study of adult survivors of out of hospital cardiac arrest, the early inflammatory signals interleukin (IL-6), IL-8, terminal complement activation (sC5b-9) and markers of endothelial damage (syndecan-1) sampled on admission associated with time to return of spontaneous circulation (ROSC) and serum lactate level in multivariable linear regression. Using Swan-Ganz catheter monitoring, they also observed that high levels of these markers associated with key post-resuscitation circulatory parameters including cardiac output, systemic vascular resistance, vasopressor need and fluid administration. Among these, elevated sC5b-9 and IL-6 and lower levels of syndecan-1 remained associated with 30-day mortality, even after adjustment for levels of other biomarkers.

Black swan events share three key features; they are initially unanticipated, retrospectively predictable and have extreme impact. Since January 1^st, 1990, <1% of articles catalogued in PubMed under the medical subject heading of “cardiac arrest” can also be found under “inflammation” (400 of 43,323 on November 24^th, 2021). This suggests that at least in terms of how the pathobiology of cardiac arrest is conceptualized, consideration of the post-resuscitation phase as an inflammatory state is uncommon. However, this may be artifact of anchoring bias, rather than biologic reality.

While sporadically discussed in the literature, increasing evidence suggests that inflammatory derangements after cardiac arrest are ubiquitous in survivors. Cross referencing “inflammation” and “cardiac arrest” on PubMed, reveals that nearly 40% of the articles identified were published in the past 5 years. Viewed through this emerging literature, it seems reasonable to conclude inflammation after cardiac arrest is a likely fait accompli, with few things being more robust triggers as life-threatening circulatory arrest and reperfusion. In a seminal report on post-ROSC inflammatory syndrome, Adrie et al showed that elevated plasma levels of IL-6 and IL-8 could discriminate survivors, and that IL-6 associated with vasopressor need, in what they termed the “sepsis-like” post-cardiac arrest syndrome.² In fact, IL-6 has been a consistent predictor of poor neurologic outcome and mortality after cardiac arrest in observational studies³ and randomized controlled trials^4,5. In a contemporary observational study published in this journal of in-hospital cardiac arrest in adults treated according to an ACLS-guided post-resuscitative care, higher early levels of IL-6, IL-8, IL-10, and tumor necrosis factor (TNF) receptor-1 were associated with unfavorable outcome.⁶

Knowledge of general immunopathology suggests that beyond cytokines, other indicators of inflammation, microvascular dysfunction, cellular damage, and capillary leak should be altered early after cardiac arrest. And in fact, reports by Langeland¹ and others, show markers of endothelial injury including syndecan-1⁷, a cell surface heparan sulfate proteoglycan, endothelin⁸, a paracrine regulator of vascular bed vasoconstriction⁹, and thrombomodulin¹⁰, a membrane protein with anticoagulant properties¹¹, are also altered post-resuscitation and are variably associated with outcome¹⁰. Likely, the mechanistic link between endothelial damage and hemodynamic dysfunction are not specific to cardiac arrest, as glycocalyx degradation markers associate with total volume of intravenous fluid resuscitation¹² and overall mortality¹³ in sepsis, trauma¹⁴ and acute respiratory distress syndrome.¹⁵

Alongside this early cytokine and endothelial response, Langeland et al revealed a clear relationship between an activated complement response, and outcomes after cardiac arrest. Complement is part of the innate immune response and a circulating protease surveillance system, that can be quickly activated and amplified by cellular damage to recruit leukocytes, promote cytokine synthesis, and clear cellular debris. Reports of the impact of cardiac arrest on complement are limited, but available evidence suggests that the pathway is also acutely activated after ROSC^16–18. Langeland et al also show that the soluble membrane attack complex (terminal complement complex or sC5b-9), the final step in pathway synthesis, associated with 30-d mortality¹. Chablan et al¹⁸ similarly showed increased complement activation in patients with poor neurologic outcome, and that sC5b-9 independently associated with poor outcome among survivors with a time to ROSC of <25 minutes.

The third condition of the black swan classification is a large effect size, which remains to be definitively satisfied. And while it is reasonable to hypothesize that ongoing systemic inflammation contributes to secondary damage after cardiac arrest, this has not been definitively demonstrated. The associative studies reported to date link inflammatory, endothelial and complement markers to outcome, but fail to discern whether these changes are epiphenomena or drive pathobiology. These inflammatory and endothelial markers possess prognostic utility, but their ability to outperform biomarkers of injury such as troponin, or neuro-biomarkers such as neurofilament light, neuron specific enolase, or tau¹⁹ also remains to be defined.

A higher bar remains as to whether these inflammatory pathways are viable therapeutic targets after cardiac arrest. Surprisingly, studies of targeted temperature management and therapeutic hypothermia suggest that inflammatory pathways are not uniformly inhibited by cooling,²⁰ though measurable differences are seen in neonates²¹. Similarly, studies of glucocorticoid administration after cardiac arrest have failed to demonstrate improvement in neurologic outcomes.²² A more specific approach to targeted immunomodulation in adult survivors of cardiac arrest using the IL-6 inhibitor tocilizumab showed a clear ability to reduce systemic inflammation, without demonstrable effects on mortality or neurologic outcomes, though the study was underpowered for these endpoints with larger studies ongoing.²³

Beyond manipulating systemic inflammation after cardiac arrest, strategies targeting neuroinflammation also deserve exploration. Precision therapies may be required given the marked regional differences in brain cytokine levels that have been observed in preclinical models of cardiac arrest.²⁴ For example, TNFα-mediated neuronal death may represent an immediate or delayed therapeutic target, of greater relevance in the striatum rather than the hippocampus, since the striatum develops extremely high levels of TNFα early after cardiac arrest.^25,26 Fortunately for therapy development, in recent pre-clinical studies, the two key phenotypes of cardiac arrest (cardiac vs. asphyxial origin) produced similar systemic and brain cytokine responses.²⁷

Preclinical or clinical data for complement inhibition in cardiac arrest are more limited. Preclinical stroke models suggest that inhibition of the complement pathway can limit perilesional complement deposition, microgliosis and cognitive deficits without affecting regenerative responses²⁸. This area, however, remains to be more fully explored in cardiac arrest.

Langeland et al¹ have provided new evidence linking an increased inflammatory response to metrics of cardiovascular dysfunction and outcome after cardiac arrest. Study limitations include the relatively small sample size, single center experience, and an over-representation of patients with non-cardiac etiologies of arrest along with pulmonary aspiratory events in those who suffer 30-d mortality. These limitations may confound the findings and limit their generalizability. Other potential limitations include the inherent complexity of immune signaling, and the complex interplay of individual inflammatory markers in injury and repair that limits our ability to predict the effects of therapies in the absence of dedicated mechanistic study.

Still, Langeland et al¹ suggest that inflammatory mediators may represent prognostic biomarkers, bio-mediators, and potential therapeutic opportunities. Moving forward, it is exciting to see ongoing pre-clinical and clinical efforts to increase our understanding of systemic and local inflammatory processes after cardiac arrest and test therapies, mirroring studies of established targets such as optimization of CPR and post-ROSC hemodynamics, mitigation of cell death in heart and/or brain²⁹, and enhancement of beneficits of hypothermia.³⁰

Many potential new therapies are emerging including both devices and drugs targeting both systemic and neuro-inflammation. Beyond the antibody-mediated approaches in ongoing trials, novel approaches are being evaluated to mitigate the detrimental effects of IL-6 using inhibitors of trans IL-6 signaling,³¹ cytokine removal strategies,³² therapies blocking microRNA-mediated inflammatory triggering,³³ cellular therapy to modify the inflammatory milieu,³⁴ and small molecules to modulate neuroinflammation, including novel intranasal delivery strategies.³⁵ However, proof that targeting inflammation will meaningfully effect either systemic dysfunction or neurological outcome remains controversial.³⁶ Given the desperate need for new therapies, considerable work remains to determine if the inflammatory signals that follow ROSC are mediators or bystanders – swans or herrings.