Effects of dexamethasone and cox inhibitors on intracranial pressure and cerebral perfusion in the lipopolysaccharide treated rats with hyperammonemia anoxic ischemic encephalopathy


Brain edema and intracranial hypertension frequently evolves in patients with acute liver failure (ALF)[ 1– 3]. Besides the presence of hyperammonemia there is now accumulating evidence that pro-inflammatory cytokines are also of pathophysiological importance [ 4– 9].

Pro-inflammatory mediators, such as the tumor necrosis factor alpha (TNF-α) and prostanoids (products of the cyclo-oxygenases (COX)), are upregulated in response to bacterial endotoxins (i.E. Lipopolysaccharide (LPS)), decrease the vascular tone and causes peri-vascular edema in the brain [ 10– 13]. In ALF patients severe infections is known to aggravate the stage of encephalopathy and increase intracranial pressure (ICP) [ 10].Anoxic ischemic encephalopathy the reason on how infection affects the brain in liver failure remains essentially unknown [ 4; 14] but suggest that anti-inflammatory interventions could be of therapeutic value. In support of this view it has been shown that induction of mild hypothermia can prevent (and reverse) development of high cerebral blood flow (CBF) and ICP in ALF [ 3]. To further explore the effect of anti-inflammatory therapy we here used an established experimental model [ 15] to investigate if development of cerebral hyperemia and high ICP induced by co-administration of LPS and NH3. We here test if the cerebral hemodynamics can be prevented by the two COX inhibitors indomethacin and diclofenac, as well as by dexamethasone.Anoxic ischemic encephalopathy

Experimental preparation

After a period of at least 5 days the animals were anesthetized with pentobarbital (50 mg/kg i.V.) and all efforts were made to minimize suffering. Catheters (PE-50) were placed in the femoral artery, femoral veins and in the peritoneum. Heparin (500 IU) was administered through the arterial catheter. The arterial catheter was connected to a pressure transducer which was reset at the mid-level of the body. The rats were tracheotomized and mechanically ventilated (hallowell EMC; E-vet. Haderslev, denmark) with a frequency of 60 per minute and a tidal volume of 5–10 ml. The respirator was connected to a capnograph allowing the expiratory CO 2 level to be monitored. The temperature was monitored by an intra-abdominally placed thermometer and maintained at 37 ± 0.2°C by the aid of a heating blanket.Anoxic ischemic encephalopathy

The rat was fixed in a stereo-tactic instrument; a scalp incision was made and two small boreholes were drilled. One borehole was used for the placement of a catheter (PE-10) in cisterna magna. The catheter was connected to a pressure transducer, which was reset at the mid-level of the body. The other borehole was used to place a laser doppler probe (probe 407; perimed, stockholm, sweden) in the brain cortex for continuous measurement of blood velocity using laser doppler flowmetry (LDF). Flowmetry was performed using a periflux laser doppler system 5000 monitor (perimed, stockholm, sweden). Continuous recordings of arterial blood pressure, ICP and LDF were stored in a computer using the software perisoft (perimed, stockholm, sweden).Anoxic ischemic encephalopathy this software was used to calculate the relative change in LDF, which was used as an index of relative change in regional CBF, and to measure the average values of ICP and mean arterial blood pressure (MAP) at consecutive time intervals of 5 min. The cerebral perfusion pressure (CPP) was calculated according to the formula CPP = MAP—ICP.

Experimental procedure

Before the experiment was started, a stable baseline of MAP, ICP and LDF was secured. Arterial blood levels of O 2, CO 2, ph, K + and na + were measured by the use of a blood analyzer (ABL 505; radiometer, copenhagen, denmark) before and every 30 min during the experiment. Once a stable baseline was obtained LPS (E. Coli 0127 B8, sigma-aldrich) dissolved in sterile saline (1 mg/ml) or vehicle alone was injected i.P. (1 mg/kg) and subsequently i.V. (1 mg/kg).Anoxic ischemic encephalopathy ammonium acetate (NH3) at a concentration of 140 μmol/kg/min or saline alone was then administered i.V. To the animals by an infusion rate of 2.3–2.6 ml/hour and the chronometer was started.

Indomethacin (confortid; activas, copenhagen, denmark) or diclofenac (sigma-aldrich) 10 mg/kg were administered i.V. 15 min after start, while dexamethasone (fortecortin; merck, darmstadt, deutschland) 2 mg/kg was administered i.V. 30 min before start. Sterile saline was used as vehicle in the control group.

The experiment was terminated after 70 min. Arterial blood was sampled and the animal was decapitated. Blood samples were centrifuged and the plasma were frozen in liquid nitrogen in heparin containing tubes.Anoxic ischemic encephalopathy


The results of this study show that LPS induces a rise in both CBF and ICP, which are associated with a decrease in CPP, i.E. A decrease in the cerebrovascular resistance. An even more pronounced increase in both CBF (33 vs. 121%) and ICP (136 vs. 286%) was observed if also hyperammonemia was induced. These findings are in accordance with a previously conducted study using the same experimental model [ 15]. Also the effect of a superimposed inflammation induced by LPS on the brain in bile duct ligated rats shows similar results [ 5].

In this study we speculated that the increase in CBF and ICP induced by LPS with or without hyperammonemia could be explained by the influence of different pro-inflammatory mediators upon the brain vasculature causing modulation of the blood-brain barrier (BBB) with subsequently increased permeability and relaxation of the microvessel endothelial cells that control vascular tone [ 16].Anoxic ischemic encephalopathy in support of this view hyperemia has been reported to develop in naïve rats in response to iv infusion of TNF-α [ 14]. In the present study, the TNF-α plasma concentration increased following exposure to LPS which then make TNF-α a plausible candidate. This is in accordance with previously conducted studies showing that the plasma TNF-α concentration increases very fast following LPS injection [ 17]. Products of the COX enzymes, which are known to be activated in response to LPS, are also able to modulate the BBB microvessel endothelial cells and the smooth muscle cells of the brain arterioles [ 18]. The high plasma level of the stable prostaglandine metabolite 6-keto PGF 1α found in the present model confirms that the COX pathway is activated.Anoxic ischemic encephalopathy interestingly, the addition of ammonia, does not course any additional elevation in the TNF-α or 6-keto PGF 1α plasma concentrations indicating that the effect of hyperammonemia is not mediated through an elevation of the prostanoids or TNF-α concentrations.

The main finding in the present study was that the two unspecific COX inhibitors indomethacin and diclofenac prevent the increase in ICP in the LPS group, and the increase in CBF plus ICP in rats exposed to both LPS and ammonia. This neuroprotective effect of indomethacin and diclofenac in the LPS + NH3 group is associated with an almost complete prevention of plasma 6-keto PGF 1α formation, and seems in accordance with an experimental study that showed that indomethacin can prevent the development of hyperemia in portacaval administered rats exposed to NH3 infusion [ 19].Anoxic ischemic encephalopathy contrary to the two COX inhibitors, dexamethasone did not prevent the increase of CBF and ICP induced by LPS and NH3. This finding may be explained by a weaker effect of dexamethasone in inhibiting the formation of vasoactive prostanoids as we found that the plasma level of 6-keto PGF 1α is only partly reduced compared to the LPS+NH3 group. We cannot refuse that administration of dexamethasone earlier than 30 minutes before the study was initiated might not have allowed the fully effect of dexamethasone. Further study will be needed.

Dexamethasone inhibits phospholipase A2 leading to a reduction in the free aracidonic acid, the substrate of the COX enzymes which explain an only partly reduction in 6-keto PGF 1α formation [ 20].Anoxic ischemic encephalopathy the observation, however, that the small but significant increase of CBF in the group treated with LPS alone is not reduced by the indomethacin and diclofenac, in contrary to ICP, may indicate that the mechanism underlying the development of high ICP and CBF are different when hyperammonemia is also present.

It could be questioned whether the effect of indomethacin is related to inhibition of COX or mediated by another mechanism, as hypercapnia-induced brain vasodilation can be prevented by indomethacin, while diclofenac fails to have same effect [ 21; 22]. Obviously, the finding that diclofenac has the same effect as indomethacin in this study favors the involvement of vasoactive prostanoids in causing brain edema and high ICP since the paco2 levels in the study animals remained stable.Anoxic ischemic encephalopathy our results also indicate that the synergistic effect between ammonia and endotoxin on the cerebral hemodynamic depends on a high level of vasoactive prostanoids. However, the observation that the two COX inhibitors have no effect on the LPS induced increase in the TNF-α plasma level suggests that TNF-α is not a pro-inflammatory mediator that actually carries out, at least not alone, the changes in cerebral hemodynamics seen in the present model. A more likely role of TNF-α is that it works via a signal molecule placed longer downstream on the inflammatory cascade pathway, such as prostacycline [ 18; 23]. Clearly more study is needed to firmly establish the key role of TNF-α in mediating cerebral vasodilatation and high ICP, since high levels of TNF-α play a major role in ALF gut-liver-brain axis and neurological complications [ 24; 25].Anoxic ischemic encephalopathy

In conclusion the results of the present study show that LPS (with and without hyperammonemia), induces an increase in the plasma concentration of both TNF-α and 6-keto PGF 1α, which correlate with a rise in both CBF and ICP. The COX inhibitors indomethacin and diclofenac prevent the increase in 6-keto PGF 1α plasma concentration but not in TNF-α measured in the LPS + NH3 group. Furthermore, indomethacin and diclofenac, but not dexamethasone, prevents the synergistic effect of LPS and NH3 on CBF and ICP, which suggests a role of vasodilatory prostanoids in the development of cerebral hyperemia and high ICP in the naïve rat with concomitant systemic inflammation and hyperammonemia.

Taken together our data supports the thesis that anti-inflammatory intervention, that is COX inhibition, is a serious candidate for treatment of HE in ALF patients with hyperammonemia.Anoxic ischemic encephalopathy one could speculate that treatment with COX inhibitors as well as dexamethasone would be beneficial. However, we couldn’t demonstrate any changes in the cerebral haemodynamics in the LPS + NH3 setup by dexamethasone and therefore we didn’t test the potential synergistic effect of NSAID and dexamethasone.

All physicians should however consider the potential detrimental impacts administrating steroids and COX inhibitors. It is widely known that prednisolone have a tendency to mask signs of infection, sometimes even reactivate infections [ 28], which is a major cause of morbidity and mortality in ALF patients. COX inhibitors contain several negative physiological effects in humans as well as in animals.Anoxic ischemic encephalopathy for example they increase the risk of gastrointestinal haemorrhage by inhibiting the cytoprotective effect of prostaglandins (mucus production, bicarbonate secretion e.G.) thereby thinning the mucosal layer and increasing the risk of ulceration. Also, ALF patients often have thrombocytopenia and administration of COX inhibitors reduces platelet activation through decreased thromboxane A2 synthesis/activity inhibiting platelet aggregation resulting in impaired blood clotting ability.

Though sodium and potassium levels remained stable in our experimental model of ALF it is important to remember that cirrhotic patients with acute liver failure often suffer from impaired kidney function due to complex circulatory changes involving splanchnic vasodilatation, decreased effective circulatory volume, activation of the renin-angiotensin-aldosterone system and renal vasoconstriction (hepatorenal syndrome).Anoxic ischemic encephalopathy by constricting the afferent arteriole in the glomerulus and reducing the glomerular filtration COX inhibitors could worsen kidney function even more rapidly.

Overall this study suggest that nsaids could be of importance for future treatment regimes in ALF by bringing the theory of the gut-liver-brain axis into account, in respect of bacterial translocation, gram negative bacteremia, liver macrophage activation and proinflammatory cytokines production [ 26, 27]. We speculate that early anti-inflammatory intervention by COX inhibition could prevent the detrimental consequences of brain oedema and intracranial hypertension in ALF patients.