Here we addressed an adenosine-based mechanism for mortality and morbidity following severe TBI by testing the hypotheses that an acute trauma-induced surge in adenosine (Clark, et al., 1997
) influences acute and chronic outcomes after TBI, and that transient blockade of adenosine with the non-selective adenosine receptor antagonist caffeine would ameliorate acute and chronic consequences of severe TBI. We provide evidence that in severe TBI induced by a lateral fluid percussion, (i) lethal outcome correlates with apnea duration, (ii) lethal apnea can be prevented by antagonizing the effects of adenosine with a single acute dose of caffeine, (iii) caffeine rescue does not worsen neurological outcome, (iv) EEG bursts occur within 4 weeks after severe TBI, and (v) EEG-bursts can be ameliorated by a single dose of caffeine, suggesting a role of adenosine in their pathogenesis. These findings are of direct therapeutic significance and suggest a novel and unexpected therapeutic potential for caffeine.
The acute administration of caffeine before an injury in caffeine-naïve subjects is generally believed to promote injury and to have pro-convulsant effects (Boison, 2010
); pretreatment of rats with caffeine caused significant, dose-dependent mortality after a cortical contusion injury (Al Moutaery, et al., 2003
). While acute adenosine receptor blockade prior to TBI can promote injury, post-injury treatment with caffeine, as demonstrated here for the first time, can completely prevent trauma-associated lethal apnea, however without aggravating behavioral outcome parameters. These findings (i)
support our hypothesis that lethal apnea following a severe TBI is causally linked to a surge of adenosine triggered by the injury and (ii)
demonstrate beneficial effects of post-injury caffeine on morbidity.
Injury to the brain in general, such as TBI, stroke, or excessive seizures, are known to result in a surge of adenosine (Clark, et al., 1997
, During and Spencer, 1992
, Gouder, et al., 2004
, Pignataro, et al., 2008
). In further support of our hypothesis, lethal apnea and post-ictal brain shutdown in sudden, unexplained death in epilepsy (SUDEP) (Hirsch, 2010
) has been attributed to excessive levels of adenosine due to deficiencies in the metabolic clearance of adenosine (Shen, et al., 2010
). In those studies, an acute dose of 40mg/kg caffeine was shown to extend the life-span of mice following excessive seizures, suggesting a mechanistic relationship between SUDEP and lethal post-ictal apnea. To make therapeutic use of post-injury caffeine, it is important to rule out that caffeine, while preventing mortality, might negatively impact neurological outcome. In studies of gross neurological and motor functions we show that the animals that received caffeine show a slight but non-significant improvement compared to non-caffeinated controls, suggesting that there are no overt deleterious effects of the acute caffeine treatment.
Severe TBI is associated with an increased risk of PTE (Lowenstein, 2009
). Studying the mechanisms of PTE is complicated by its long latency, often months or years after the traumatic event. Early electrophysiological studies using acute hippocampal slices demonstrate increased excitability at 1 week (Santhakumar, et al., 2000
) and as late as 15 weeks (Golarai, et al., 2001
) after brain injury. More recently, the progression to spontaneous seizures has been demonstrated in vivo after FPI (D’Ambrosio, et al., 2004
), with 92% of rats demonstrating electrographic seizure activity at 8 weeks, progressing to generalized seizures in 50% of the survivors by 1 year after injury (Kharatishvili, et al., 2006
). The incidence, frequency, and latency of these seizures demonstrate the face validity of the FPI model for the study of PTE, yet the logistics of the model make it difficult to efficiently propose and test hypotheses. We examined hippocampal electrographic activity at 4 weeks following FPI, and have found epileptiform bursts in the severe injury group, consistent with other studies (D’Ambrosio, et al., 2009
). Remarkably, a single acute dose of caffeine given immediately after the FPI was linked to a significant reduction in burst duration 4 weeks after the injury. This finding indicates potential disease-modifying consequences of early intervention with adenosine signaling at an early time point following FPI. More work is certainly needed to investigate the mechanistic relationship between an early surge of adenosine (blocked here at least partly with caffeine) and subsequent epileptogenesis. While it is not clear from our studies whether these bursts would develop into spontaneous generalized seizures, they provide evidence of early epileptiform activity and constitute a rational early target for evaluating interventions to modify the long-term outcome after TBI.
Caffeine, a non-selective antagonist of adenosine receptors at doses normally reached during human caffeine consumption, is the most widely used psychoactive substance, with a well-understood pharmacodynamic and pharmacokinetic profile (Fredholm, et al., 1999
). By antagonizing the function of adenosine, which acts as an endogenous anticonvulsant and neuroprotectant of the brain (Dragunow, 1986
, Dragunow and Faull, 1988
, Dunwiddie, 1980
, Ribeiro, 2005
, Ribeiro, et al., 2003
), the acute use of caffeine is generally thought to aggravate neuronal injury and to promote epileptic seizures (Boison, 2010
). Using a model of closed head injury in female rats, the high mortality associated with pre-injury caffeine was delayed beyond the acute period evaluated in our study (Al Moutaery, et al., 2003
). It is important to point out that in the present study we specifically assessed the immediate acute apnea-related mortality that has not been assessed in previous studies. This distinction might be model-dependent, since not all models of TBI recreate the immediate apnea-related mortality studied here.
In contrast, the chronic use of caffeine is generally thought to be neuroprotective, at least in part by effect inversion and adenosine receptor desensitization, at least under certain dosages (Fredholm, 1997
, Jacobson, et al., 1996
). In line with the neuroprotective effects of chronic caffeine, 3 weeks of caffeine administration to mice provided profound neuroprotection following a cortical contusion injury, whereas an acute dose of caffeine in the same model was without effect (Li, et al., 2008
). The detrimental effects of acute caffeine are likely based on the blockade of A1
Rs. This notion is supported by findings that TBI in A1
R knockout mice led to lethal status epilepticus (SE) (Kochanek, et al., 2006
). Likewise, A1
R knockout mice subjected to an excitotoxin succumbed to lethal SE (Fedele, et al., 2006
). In human populations the use of an A1
R antagonist for the treatment of acute heart failure with renal impairment was associated with seizures as one of the observed side effects (Cotter, et al., 2008
). Our current findings demonstrate for the first time that a single acute dose of caffeine, when given immediately after
the injury prevents lethal outcome and are in apparent contrast to previous data discussed above.
The time-point of acute caffeine administration may contribute to the observed differences in pre- and post-injury caffeine administration. In vitro
, stretch injury of neurons limits the effect of caffeine on calcium-induced calcium release (Weber, et al., 2002
), suggesting that pre-injury caffeine in a caffeine-naïve neuron is additive with injury, but post-injury caffeine is not. In vivo
studies examining the effect of a single bolus of caffeine on outcome after TBI considered a single time point, 30 minutes, prior to injury, sufficient for the caffeine to be well distributed throughout the brain, yet the relatively short half-life of caffeine (0.8 hours) suggests that some clearance has occurred prior to injury (Bonati, et al., 1984
). In addition, caffeine when given prior to the injury is likely to affect the entire brain, whereas caffeine perfusion into the injured brain might be compromised in the most severely affected brain areas. This could be a likely explanation why the acute dose of caffeine after the injury did not worsen morbidity. However, full penetration of caffeine into brainstem is likely, since this region is not directly affected by the lateral fluid percussion injury. In our experiments, we found that post-injury caffeine treatment rapidly restored spontaneous breathing; the reduced hypoxia as a result of limiting apnea may surpass any consequent negative effects of continued A1
R blockade. The relatively low affinity and rapid clearance of caffeine also may serve to limit the detrimental effects of A1
R blockade demonstrated in A1
R knockout mice (Kochanek, et al., 2006
). Based solely on its actions as an adenosine receptor antagonist, we would predict that caffeine delivered to caffeine-naïve subjects immediately prior to TBI might act similarly to post-injury administration to prevent lethal apnea. Of more significant clinical importance, the influence of chronic caffeine consumption on A1
R and A2A
R expression (Svenningsson, et al., 1999
) indicates that we must consider the protective effects of a single bolus of caffeine in both caffeine naïve and chronically caffeinated subjects.
The fact that caffeine is rapidly absorbed in the gut, readily crosses the blood-brain barrier, and is a relatively weak, non-selective antagonist may partly account for its post-injury efficacy without negative consequences. Adenosine has evolved to maintain homeostasis across organ systems and within the brain across neurotransmitters (Boison, 2008
, Boison, et al., 2011
, Fredholm, 2007
); to maintain stability, there are likely many as yet unrecognized compensatory mechanisms activated in response to typical environmental challenges. The most successful therapeutics may be those that blunt the acute effects of a traumatic event, and then allow endogenous compensatory mechanisms to fully function. Caffeine is a widely consumed psychoactive substance (Barone and Roberts, 1996
), and chronic consumption is known to affect adenosine receptor expression (Svenningsson, et al., 1999
). Sleep disruption, another common fact of modern life and likely contributor to risk of TBI, may further alter adenosine receptor expression (Basheer, et al., 2004
, Elmenhorst, et al., 2009
). While chronic caffeine consumption is associated with favorable outcome after TBI (Sachse, et al., 2008
), further studies to examine the combined effects of chronic pre-injury caffeine treatment with a protective post-injury caffeine treatment are essential to support the translation of the current findings to broader therapeutic use.
Our results are the first to demonstrate that caffeine limits apnea duration and prevents mortality when administered rapidly following severe TBI, without negative consequences. As apnea is a major cause of pre-hospital mortality, this finding presents a major therapeutic opportunity for first responders in both civilian and military environments. We also demonstrate that a single post-injury dose of caffeine can have long-lasting effects on electrographic burst durations, indicating that caffeine might beneficially influence processes involved in posttraumatic epileptogenesis, an interesting observation that warrants further experimentation. In conclusion, our studies show that, at a safe dose and without adverse neurological outcome, caffeine has the potential to prevent lethal apnea following TBI, and may reduce brain excitability long-term. Since caffeine is a well characterized and widely consumed drug, our findings present a translatable strategy to reduce acute lethal outcome after severe TBI.