Earlier charcoal haemoperfusion in fulminant hepatic failure.

Charcoal haemoperfusion with a prostacyclin infusion for platelet protection was carried out daily in the treatment of 76 patients with fulminant hepatic failure. in the 31 patients who had been referred early and in whom the serial haemoperfusion was started while signs of grade III encephalopathy were still evident remarkable survival rates were obtained-70% for patients with paracetamol poisoning and 65% for the group overall. Cerebral oedema developed significantly less frequently in this group than in those patients in whom haemoperfusion was started later in the course of the disease, when signs of grade IV encephalopathy were already apparent (49% and 78% respectively, p less than 0.05), and this was likely to have been a major factor in their improved survival. Biocompatibility of the system was excellent, and both platelet and white-cell counts at the end of perfusion periods were the same as pre-perfusion values.

Controlled trial of dexamethasone and mannitol for the cerebral oedema of fulminant hepatic failure.

A controlled trial of 44 patients was undertaken to evaluate the use of dexamethasone (32 mg stat, 8 mg qds) in preventing, and intravenous mannitol (1 g/kg) in reversing the cerebral oedema of fulminant hepatic failure. Diagnosis of cerebral oedema was based on intracranial pressure recordings or the presence of defined clinical signs. Cerebral oedema developed in 34 patients with similar frequency in those treated with and without dexamethasone (16 of 21 and 18 of 23 respectively). In those 34 patients episodes of cerebral oedema resolved significantly more frequently in the 17 patients who received mannitol than in the 17 patients who did not (44 of 53 and 16 of 17 respectively, p less than 0.001). Dexamethasone did not affect survival but among patients who developed cerebral oedema those who received mannitol had a significantly better survival than those who did not receive it (47.1% and 5.9% respectively, p 0.008, Fisher's one-tail test).

Clinical monitoring of intracranial pressure in fulminant hepatic failure.

Cerebral oedema is the commonest immediate cause of death in fulminant hepatic failure and an investigation was carried out to determine the value of monitoring intracranial pressure (ICP) and to examine the effects of ICP of dexamethasone therapy and mannitol administration. ICP values in 10 patients at the time of insertion of a subdural pressure transducer (grade IV encephalopathy) averaged 15.5 +/- SD 14.8 mmHg. Despite dexamethansone therapy, which had been started on admission, rises in ICP were subsequently observed in seven of the eight patients who died. In the two patients who survived, the highest reading were 47 and 35 mmHg. Mannitol consistently reversed or arrested ICP rises when pressure was < 60 mmHg. ICP monitoring provides additional information in the managment of patients and is essential if mannitol therapy is to be used.

Prostacyclin to prevent platelet activation during charcoal haemoperfusion in fulminant hepatic failure.

Adverse effects associated with hypotension and the appearance of platelet aggregates in the circulation complicate charcoal haemoperfusion of patients with fulminant hepatic failure. In an attempt to avoid these difficulties the platelet protective effect of prostacyclin (PGI2) given intravenously before and continuously during haemoperfusion was evaluated with an improved charcoal column design. Two of the six patients who underwent haemoperfusion without PGI2 had hypotension, which in one was associated with a striking rise in Swank screen filtration pressure necessitating discontinuation of haemoperfusion after an hour. No platelet losses were observed in the six patients treated with haemoperfusion and PGI2 infusion, and there was significant protection from platelet activation, as assessed by the prevention of release into plasma of the platelet-specific protein beta-thromboglobulin.

Intracranial pressure in pigs with surgically induced acute liver failure.

Cerebral edema has now been noted to occur frequently in patients dying of fulminant hepatic failure. In the present study, intracranial pressure was monitored in an animal model of acute liver failure. Acute liver failure was induced surgically by hepatic devascularization. Serial monitoring of the electroencephalogram revealed progressive slowing of the frequency with decreasing amplitude. Elevation of the blood ammonia was also observed from baseline values of 64 +/- 12 SE to 744 +/- 97 mumol/liter. Monitoring of the intracranial pressure with a subdural pressure transducer demonstrated a progressive and reproducible rise from 12.8 +/- 2.5 mm Hg immediately after the operation to a mean value of 51.6 +/- 11.8 mm Hg just before death 6--12 hr later. At autopsy, the brains of the test animals were found to be swollen with flattened cortical gyri. In the control animals, intracranial pressure rose slightly but returned toward normal levels (8.0 +/- 2.5 mm Hg) 8 hr after laparotomy and remained normal until their death. There was a statistically significant difference between intracranial pressure levels of the test animals and those of the controls (P less than 0.01). Intravenous methylprednisolone (2.0 g initially followed by 0.5 g every 2 hr) administered immediately before and after hepatic devascularization prevented rises in intracranial pressure but had no effect when given 4 hr after operation. The early and progressive increase in intracranial pressure was an unexpected finding, and an assessment of such a sequence in patients with fulminant hepatic failure is currently in progress.

Effects of cerebral oedema and arterial hypotension on cerebral blood flow in an animal model of hepatic failure.

The effects of arterial hypotension and a raised intracranial pressure on cerebral blood flow (CBF) have been investigated in an animal model of hepatic failure. Arterial hypotension was associated with a fall in CBF in the animals with liver failure but not in the controls. Significant differences in mean CBF between the two groups of animals could be demonstrated when the systolic blood pressure was in the 30-60, 60-90, and 90-120 mmHg range, but not in the 120-150 mmHg range. A raised intracranial pressure also resulted in a fall in CBF in the animals with liver failure, and a significant difference could be demonstrated between the two groups when the intracranial pressure was in the 20-40 mmHg range but not in the 0-20 mmHg range. Furthermore, in the animals with liver failure the cerebral metabolic rate for oxygen (CMRO(2)) fell as the CBF fell, there being a highly significant correlation between these two parameters. In the controls no such relation existed. Treatment with neither charcoal haemoperfusion nor high dose corticosteroids affected the fall in cerebral blood flow that occurred during arterial hypotension in the animals with liver failure. Corticosteroids, however, did reduce the fall in cerebral blood flow associated with a high intracranial pressure. These results suggest a disruption of the cerebral circulatory responses in hepatic failure. They also raise the possibility that CMRO(2) and cerebral blood flow may be maintained at normal levels in hepatic encephalopathy if cerebral oedema and arterial hypotension can be prevented.

Biocompatibility of coated and uncoated charcoal during haemoperfusion in healthy dogs.

The biocompatibility of two commercially available charcoal columns, one containing coated and the other uncoated but immobilized charcoal, was compared during four haemoperfusions with each in eight healthy greyhounds. Reductions in arterial levels of platelets (49% and 42% respectively) and leucocytes (both 21%) were similar. Microaggregates, detected by the Swank screen filtration pressure technique, were found in blood leaving the columns during three of the four perfusions with each column. Another twelve perfusions with the uncoated column were carried out with the addition of one or other of the following three agents which inhibit platelet aggregation: sulfinpyrazone, dipyridamole, or citrate-phosphate-dextrose. With none of these were platelet losses less as compared with the four perfusions in which heparin only was used. However, rises in screen filtration pressure were less pronounced. In other perfusions, where its dosage was varied, heparin was shown to reverse, and in large doses delay, the appearance of micro-aggregates. Thrombus in the column itself may be a source of microaggregates, but platelet aggregation in the absence of thrombus deposition may be responsible. The relation of these findings to micro-aggregate formation, which has constituted a clinical problem during charcoal haemoperfusion in humans with fulminant hepatic failure, is considered.

Treatment of fulminant hepatic failure by polyacrylonitrile-membrane haemodialysis.

24 patients with fulminant hepatic failure who had deteriorated to grade-IV coma were treated by repeated periods of haemodialysis with a polyacrylonitrile membrane. 9 patients fully recovered consciousness, and 8 (33%) survived to leave hospitals. These results are to be compared with those of conservative management alone (15% survival in 53 cases) and those obtained initially with charcoal haemo-perfusion (38%). Of the 16 treatment failures, cerebral oedema was found at necropsy in 13 (18%). Whether this would have been less of problem if treatment had been started earlier in the course of the illness remains to be determined.

A dog model of fulminant hepatic failure produced by paracetamol administration.

Oral administration of graded doses of paracetamol to dogs produced hepatic necrosis with some similarities to the clinical syndrome seen in man following a paracetamol overdose. Coma, with raised levels of arterial ammonia, was produced and the aspartate aminotransferase levels became markedly elevated in 2 animals who survived more than 24 h. However, the extent of the hepatic necrosis and the time of survival following paracetamol administration were too variable for this model to be of value for the testing of new methods of temporary liver support. When paracetamol was given by intraperitoneal injection many of the animals died of respiratory distress. Significant methaemoglobinaemia was detected, which was associated with a reduction in the arterial partial pressure of oxygen and was partly reversed by the administration of methylene blue.

Artificial liver support based on haemoperfusion of adsorbents.

The need for a device capable of performing the excretory functions of the liver during acute and chronic liver failure is described. Experience of artificial liver support systems is reviewed. The results of patients treatment with adsorbent haemoperfusion are discussed and problems of system design and biocompatibility are outlined.

Resin column perfusion with whole blood or plasma separated by the continuous flow celltrifuge.

1. The aim of this study was to define the factors influencing plasma separation from the continuous flow celltrifuge and to evaluate plasma as an alternative to whole blood for perfusion of exchange resins as part of a system of artificial liver support. 2. Studies in vitro showed the importance of packed cell volume, centrifugal force and duration of centrifugation on the degree of plasma separation. From these data it was possible to calculate plasma flow rates likely to be obtained from the celltrifuge when used in vivo. These predicted values correlated closely with plasma flow rate obtained in twenty-six studies in dogs. 3. Comparison of whole blood perfusion with plasma perfusion of exchange resins in another series of dog experiments showed that with whole blood perfusion there was often a considerable rise in pressure across the resin column but that this did not occur with plasma perfusion. 4. Measurements of platelet losses in the same series of experiments showed a 50% reduction of arterial platelet counts over a 31/2 h period of perfusion when whole blood was perfused. Although the fall was lower with plasma perfusion, the difference was not statistically significant. 5. Use of the celltrifuge provides a means of resin perfusion free of the mechanical difficulties of whole blood perfusion, but platelet losses still remain a problem.