Children with mTBI had a reduction in activated neutrophils, evidenced by reduced CD11b expression in the mTBI cohort reaching a significant decrease at 10–14 days post-injury. In parallel, total monocytes were reduced acutely with the reduction seen in intermediate monocytes. The TLR4 expression on neutrophils was highly upregulated in the acute phase and downregulated by 10–14 days, below that of control levels.
Activated CD11b neutrophils play an important role in blood-brain barrier disruption and cytokine release in traumatic brain injury [21]. Infiltrating CD11b + neutrophils contribute to the size of the lesion in severe TBI [22]. Although severe TBI in this cohort had modest increases in CD11b, in this study the response in mTBI is reduced systemic CD11b expression. The dynamics of neutrophil survival differ with age [23]. This is exemplified in a neonatal study which showed an increase in both CD11b and TLR4 expression in infants with neonatal encephalopathy [24] which persisted over the first week of life with dysfunctional immune responses in later childhood [25] [26]. Elderly people have decreased CD11b expression, a phenomenon contributing to immune senescence [27] compromising their immune response. Age differences may be a factor in this decreased responsiveness, or the type of injury presenting, as mild TBI is distinct from severe TBI. Our cytokine analysis shows trends of immune suppression, consistent with adult studies of blast TBI [28] rather than the typical elevation of IL-8 seen in severe TBI [29]. There may have been an initial peak in immune activation that was not captured in this cohort as the highest neutrophil count in severe TBI is in the first three hours [30].
Systemic TLR4 expression was activated in mTBI and sTBI, especially in the latter group. However, TLR4 expression was suppressed in the subacute phase of mild traumatic brain injury. The rise of TLR4 production in the severe cohort was significant. The TLR4/NF-κB pathway has been demonstrated to play an important role in traumatic brain injury [31]. The TLR4 complex is expressed on the cell surface of the neutrophil and plays an important role in targeting inflammatory cytokine gene transcription. Tang et al. found that IL-1β and IL-6 were significantly lower in the wounds of TLR4-deficient mice [32]. When the receptor for TLR4 is activated, immune mediators released from TLR4-expressing cells may have neurotoxic effects [33].
Total monocytes were depleted in the initial period following mTBI, within the intermediate subset, with a recovery seen by 10–14 days. This is opposite to the effect seen in severe TBI where the monocytes were increased in the intermediate population, not reaching statistical significance. Monocytes aid recovery in spinal cord decompression [34] and their depletion in the animal model results in poorer motor results [34]. There is discrepancy in the literature, as a murine model shows depleted monocytes in the first day following TBI that persists for a month [35], with human studies showing elevation of the monocyte population in those with severe TBI [30]. The monocyte population has not been documented in those with mTBI.
The innate immune response is a significant contributor to inflammation after TBI regulated by the inflammasomes. Activation of IL-1β, the end product of inflammasome activation, includes pathways via TLR4/NF-κB, via Nod-like receptor pathways including NLRP3 and NLRP1 and via AIM1 pathways. These pathways activate the inflammasome via caspase driving cytosolic NF-κB transcription and IL-1β production, driving pyroptosis. The interleukin-1 family has autocrine properties and can self- propagate its own signal with IL-1α having local autocrine effects and IL-1β having paracrine effects [36]. Our study did not demonstrate an upregulation of the NLRP3 pathway in the mild form of TBI in children. The pathway was downregulated, and more so if there was a prior concussion. NLRP1 was also downregulated, which persisted into week two. The AIM pathway was upregulated on presentation. ASC was not statistically higher and was low in those previously suffering mTBI. ASC and IL-18 biomarker has previously proven to be a useful in determining outcome in severe TBI [37]. In our study IL-18 was depressed however, not raised. This may be accounted for in the mild and different nature of an mTBI to sTBI or by age.
IL-1β mRNA was significantly upregulated at presentation with injury. At 10–14 days from injury, it was higher still showing an important role of inflammasome at the time of injury, but especially in the subacute phase, indicating it may have a long therapeutic window and mechanistic role in prolonged symptoms. Human tissue samples around brain contusion show an IL-1α acute spike following brain injury, while IL-1β shows a much more gradual increase that is thought to represent a portion of the delayed cytokine response to TBI [38]. It is a potent stimulus for other cytokine release including TNF-α and IL-6. IL-1β is a target for therapies to prevent secondary injury in TBI.
A mouse model of chronic repetitive mTBI in adolescence demonstrated IL-1β and IL-18 production with prominent inflammasome activation. Mice with deficiencies in caspase and interleukin receptors in that study had better outcomes at a year [39]. Prior concussion carries higher risk of prolonged recovery time [40]. Increased inflammasome activation did not account for this in our study, however there were low numbers in our groups with heterogenous injuries. Further evaluation of these groupings is warranted.
Overall there is a hyper-responsiveness of inflammasome-related cytokine release in response to LPS stimulation in the TBI population, but the interleukins do not respond in the subacute phase. This may represent priming by the injury event [41] which has been suggested to last one week [42]. Previous animal studies of severe TBI have demonstrated that pre-conditioning with LPS increased both pro- and anti-inflammatory cytokines, and their effect attenuated the extent of histological injury [42]. These models of injury have contusions with a penumbra of reduced cerebral blood flow. These brain injuries are different to mTBI, further studies looking at the preconditioning response are warranted in mTBI models and patients.
IL-18 plays a role in downregulating TNF-α [43]. At low doses, LPS causes a decrease in IL-18 production and enhances the innate immune response to sepsis, at higher doses of LPS there is an increase of IL-18 production which influences death from sepsis [44]. IL-18, an end-product of the inflammasome, was significantly decreased in mTBI and the normal decrease of IL-18 in response to LPS not seen in the TBI cohorts. There were similarities with IL-33 with reduced expression in the controls and a marked response seen with the same LPS stimulus in severe TBI. IL-33 does not respond by priming and has enhanced secretion in response to attempt to prime [45]. The response of the mTBI group in the subacute phase fits that of priming described in the literature for the cytokines associated with the inflammasome complex.
The inflammasome is a potential safe target of therapy. Numerous safe therapies modify NLRP3 including pioglitazone [12], rapamycin [46] and Omega three [14] and are under consideration as therapeutics. In our cohort this was not a dominant pathway. AIM1 was the only pathway upstream of IL-1β in our study that was upregulated. One small trial in adults with sTBI showed demonstrated the effect of interleukin-1 antagonist Anakinra in adults but was not powered to show clinical differences in outcomes [47].
Mild traumatic brain injury induces changes in the innate immune system. These effect of mTBI on the immune system is prolonged. This is evidenced by the modified response of the inflammasome complex to LPS stimulation at two weeks.
Mild traumatic brain injury, like burns, are sterile injuries that induce systemic immune changes. It would be judicious to profile this response in these children at a later time- point out from injury to see if the immune activation quiesces. The clinical period of a year was noted by Eisenberg and colleagues to be the period within which it was critical not to get a repeat concussion [8] Treating serum with LPS and measuring the responsiveness of a number of key cytokines may demonstrate the window for which the injury persists.
The current literature is based on animal models of more severe injury.
Our findings are inconsistent with some severe TBI studies. Severe TBI usually involves gross intracranial bleeding. Furthermore, there are immune differences between adults and children. Finnerty and colleagues described how children after burn injury had a depressed response of important pro- and anti-inflammatory relative to adults
[48] which persist for up to 3 years [49]. Animal studies demonstrate persistent immune dysfunction in sTBI [4]. Numerous studies have demonstrated the chronic changes associated with repeated head injury [9], that support the hypothesis of inflamma-aging [50].
The injuries in our study population were heterogenous. Only 40% had vomiting and some of the children only had isolated symptoms[16]. In the clinical setting it is impossible to quantify the volume of trauma that was sustained. The numbers in the study measuring the effect of previous mTBI on inflammatory cascade were small. A limitation of this study was the absence of data on analgesia and quantifying if there was an effect from paracetamol and ibuprofen.
This study showed that the pathway in mild TBI includes a concerted response by the innate immune system with TLR4 pathway signalling, inflammasome activation and alterations in neutrophil activation and monocyte populations. The changes are sustained and result in altered systemic immune responses at two weeks from injury supported by the changes in LPS responsiveness.