Neurological Complications Following Lver Transplantation Daniel P. Stein, MD,' &chard J. Lederman, MD, PhD,* David P. Vogt, MD,t William D. Carey, MD,$ and Thomas A. Broughan, MDt

Neurological complications occurred i n 4 (19%) of 2 1 consecutive patients (Group 11) undergoing orthotopic liver transplantation, compared with a 47% (9119) incidence in our initial series (Group I). In G r o u p 11, the neurological problems included new recurrent headaches a n d delayed intracerebral hemorrhage (1 patient), partial third nerve palsy and brachial plexopathy (1 patient), and ataxic dysarthria with encephalopathy (2 patients). Seizures, noted in 8 of 5, neurologically affected G r o u p I patients, were not encountered in G r o u p II. Of the 4 patients in t h e entire series with the cerebrocerebellar syndrome, 2 had partial recovery after stopping treatment with cyclosporine, 1 stabilized when cyclosporine was discontinued b u t later worsened when rechallenged, and 1 had full recovery but died follolwing a second transplantation. Brain magnetic resonance images appeared normal in 3 of the 4 patients. Complication!; affecting vision included cortical blindness in 2 patients and occipital lobe hemorrhage in 1. All completely recovered. Survival was comparable for patients with and those without neurological complications (69% and 63%, respectively).. Immediate withdrawal of cyclosporine at t h e onset of a change in mental status or dysarthria and improvement i n intra- and postoperative management may have contributed to t h e decreased incidence of neurological complications. Stein DP, Lederman RJ, Vogt DP, Carey WD, Broughan TA. Neurological complications following liver transplantation, Ann Neurol 1092;J 1:644-040

-

Liver transplantation has become a practical and effective option for patients with acute and chronic irreversible hepatic disease. Neurological complications occur in 8.4 to 47.Op4 [ 1-51 of transplant recipients and can affect many aspects of nervous system function. Alterations in mental status can range from mild encephalopathy to akinetic mutisrn or coma, and from mild anxiety or depression to psychosis with visual hallucinations and delusions { 1-61. Seizures, myoclonus, tremor, quadriplegia, cerebellar dysfunction o r cortical blindness, brachial plexopathy, poly- o r mononeuropathy, stroke, and central nervous system (CNS) infection are other potential complications { 1-5, 7-12]. A previous report on the first 19 patients undergoing orthotopic liver transplantation at the Cleveland Clinic Foundation described the form, frequency, and potential causes of postoperative neurological dysfunction [3}. This article briefly reviews the features of our initial series and presents our experience with the subsequent 2 1 patients undergoing orthotopic liver transplantation. T h e effect of neurological cornplications on survival is also discussed. Patients and Methods Forty consccutive pntients undergoing 46 transplantations between November 1984 and July 1988 were divided into

From the Ilc~pc~rtments of 'Neurology, i-General Surgery, and $Gastroentenhpy, Clevelaiiit Clinic Founitation, Cleveland, OH.

RetciveLt

Se,,

,99 I . and

(),

p u b l i c ~ t i o nDec

I,

644 Lopyright

1W1

C,

ill

revised forin

for

two groups. Our initial series [ 5 } ,Group I, included 2 3 transplantations in 19 patients (10 males, 9 females), 18 adults (19-52 years old) and 1 child Group 11, the subsequent 21 patients (10 males, 11 females) undergoing 2 3 transplantations between September 1986 and July 1988, included 17 adults (19-57 ycars old) and 4 children. Patient data, including the incidence of neurological complications and the 1-year survival rate, are displayed in Table 1. Examination by a neurologist was a standard part of the pretransplantation assessment. However, 1 patient (Patient 2) who subsequently developed a neurological complication was not seen by a neurologist preoperatively. Orthotopic liver transplantation has become a relatively standardized, although demanding operative procedure { 131. The operation may be thought of in three stages: dissecrion; anhepatic stage (removal of the diseased liver and implantation of the donor liver), and reperfusion. In many centers, venovenous bypass is used routinely during the anhepatic stage of the procedure. Heparin-bonded cannulas placed in the saphenous and portal veins return blood via a centrifugal pump to the axillary vein through another piece of heparinbonded tubing 1141. The routine use of venovenous bypass during the anhepatic stage of thc procedure has been challenged. Many centers, including the Cleveland Clinic Foundation, have found that bypass is necessary for hernodynamic stability in only 10V of patients [ 151. Venovenous bypass was routinely used in the first 2 1 patients of our series and not in any of the subsequent l!,

Address correspondence to Dr Stein, Neurornuscular Discdses Section, National Institute of Neurological Llisorders and Stroke, National Institutes of Health, 9000 Rockville Pike, Building 10, Rooin 4N248,Bethrsda, M1> 20892.

1092 b i the Americm Neurological Association

Table I , Patient Data: 46 Transplantations in 40 Patients, from Nowmbev 1984

Patients without neurological complications Patients with neurological complications Total

t o J74ly

1988

Group I 11184-8186

Group I1 9/86-7188

Total

1-Yr Survival Rate

10

17

27

17 (63%)

7 (47%) 19

patients. An analysis of hernodynamic parameters revealed that cardiac output decreased an average of 40% and systemic vascular resistance increased an average of 90% during the anhepatic stage of the procedure when bypass was not used (David Vogt, unpublished data from our center, 1989). However, blood pressure, pulse, and urine output were maintained in spite of these temporary hemodynamic changes. The intraoperative transfusion requirements, postoperative renal function, and survival rate were not significantly different between patients who underwent bypass and those who did not. No patient in either group had a known intraoperative air embolic event. Intravenous corticosteroids and cyclosporine (Cy) were used for immunosuppression in the early postoperative period. Cy was titrated to a level of 150 to 200 ngiml by measuring daily trough blood levels using high-pressure liquid chromatography. Intravenous methylprednisolone sodium succinate (Solu-Medrol) was tapered in 40-mglday decrements over 5 days from an initial dose of 200 mgiday. Prednisone, 20 mg by mouth each day, was then given as maintenance. Levels of liver enzymes, bilirubin, electrolytes, total protein, albumin, magnesium, calcium, and phosphorus, as well as complete blood cell count and coagulation parameters, were determined every day until the time of hospital discharge. Cholesterol was not measured until October 1987, following a report linking lower levels to neurological complications {l). Cholesterol was thus measured in only 1 neurologically affected patient (Patient 13).

Results Clinical Data Of the 40 patients, 13 (33%;) had neurological complications, t h e type and frequency of which are summarized in Table 2. T h e rate of complications in G r o u p I, 47% (9119), exceeded that in G r o u p 11, 19% (41 2 1). This difference, however, was not significant (p = 0.09). Several patients had more than o n e neurological complication (Table 3). Of the 1.3 patients with neurological complications, 4 died within 1 year after transplantation, a 69% survival rate. No o t h e r patients in this g r o u p died during the second postoperative year. I n the g r o u p without neurological complications, t h e survival rate was 63% at 1 year and 5 9 9 ' at 2 years following transplantation. The underlying liver diseases in the neurologically

4 (19%) 21

9 (69%)

13 (33%)

26 (65%)

40

Table 2. Neurological Complicatiom

Central nervous system Seizure Cerebrocerebellar syndrome Coma Cortical blindness Delusions with visual hallucinations Psychotic behavior without hallucinations Headache Intracerebral hemorrhage Tremorimyoclonus Peripheral nervous system Brachial plexopathy Poly neuropathy Partial third nerve palsy

Group I

Group I1

8 2 2 2 1

0 2 0 0 1

2

1

2 0 1

1 1 1

1 1 0

1 0 1

affected patients were varied and mirrored the hepatic diseases in t h e patients without neurological dysfunction. Chronic active hepatitis was most common ( n = 4), followed by sclerosing cholangitis (n = S), cryptogenic cirrhosis (n = 3), Wilson's disease ( n = 2 ) , and alcoholic cirrhosis (n = 1). Pretransplantation neurological abnormalities included hepatic encephalopathy in 2 (Patients 1 and 13) and Guillain-Barre syndrome in 1 (Patient 10). Four patients developed a multifocal neurological disorder, which we have characterized as a cerebrocerebellar syndrome, including altered sensorium, aphasia, dysarthria, ataxia, quadriparesis, and seizures. This was a monophasic illness with partial recovery and significant residua in 2 (Patients 3 and 71, and was a fluctuating disorder with mild residua in 1 (Patient 13). In Patients 3 and 13, neurological symptoms clearly improved when the dosage of C y was lowered or discontinued. Symptoms in Patient 13 later worsened when Cy was restarted. The fourth patient with cerebrocerebellar dysfunction (Patient 11) had mild encephalopathy and dysarthria b u t also had worsening failure of the first hepatic graft. She seemed t o recover from t h e neurological disorder b u t subsequently died following a second liver transplantation.

Stein et al: Liver Transplant Complications

645

Table 3 Clrnziai Datu -

Patient No

Age (yr)

8 0 Groul’ I1 10 I1 12

13

sex

Complications

F M M M M M M F M

Seizure, delusions, visual hallucinations Seizure Seizure, coma, cerebrocerebellar syndrome, polyneuropathy, cortical visual cicfcct Seizure Brachial plexopathy Seizure, myoclonus, headache Cerebrocerebellar syndrome, manic event, seizure, coma Psychotic depression, seizure, headache, cortical blindness Seizure

F F M M

Headache, intracercbral hemorrhage Cerebrocerebellar syndrome, paranoid ideation, tremor Brachial plexopathy, partial third nerve palsy Cerebrocerebellar syndrome, visual hallucinations, delusions -

Two brief case histories illustrate the major features of this complication. P A T I F N ’ ~7.

A 42-year-old man underwent liver transplantation for chronic active hepatitis. From the third to the ninth postoperative d a y . progressive dysarthria, agraphia, and lethargy developed. Seizure activity began on the eleventh postoperative day and continued sporadically over the next 4 days despite treatment with phenytoin and intermittent diazepam. Cy was stopped on the twelfth postoperative day. By postoperative day 19, he was in a state of vigilant unresponsiveness. The patient stabilized over the next several weeks and subsequently became more responsive. He was able to communicate, dthough his speech was extremely slurred and often unintcllipible. He had marked weakness as well as ataxia of the extremities. By the sixth postoperative month, he could ambuliite with a walker, feed himself, and communicate with moderate dysarthria. At the time of writing, he had had no further seizures, hut moderate ctysarthria and ataxia with mild weakness persisted.

13. A 5)-year-old man with a history of cirrhosis of uncertain origin, insulin-&pendent diabetes, and aortic stenosis had two episodes of- hepatic encephalopathy prior to transplanration. Eight days postoperatively, h e was more lethargic; at 9 days, he was alert and oriented, with moderate memory impairment and mild dysarthria. He had visual hallucinations and delusions, which were controlled with thiothixene. Cy was discontinued and he was maintained on azathioprine and prednisone. Three months following transplantation, the liver function worsened and he was again started o n Cy. Two months later, his mental status and speech were again worse and the Cy was discontinued. O n neurological reevaluation 10 months after transplantation and 3 months after stopping Cy for the second time, he had moderate memory impairment, mild dementia, moderate ataxic dpsarthria. and normal motor and sensory function. PATIENT

Seizures were the most commonly observed compli-

646 Ann‘ils of Neurology

Vol 3 I

No 6 June l‘Y12

cation overall, occurring in 8 ( 2 0 5 ) of the ,irO patienrs. T h e seizures were usually generalizeci, a i d almost always multiple or recurrent. Onset occurred during the first postoperative week in 4 patients, during the second postoperative week in 3, anti at 8 weeks following transplantation in 1. O n e patient (Patient 3 ) had a scizure prior to transplantation. This was thought t o he secondary to alcohol withdrawal; a preoperative eleitroencephalogram (EEG) and computed tomography (CT) scan in this patient appeared normal. All patients with seizures were in Group 1. O n e patient with multiorgan failure, who had a seizure just prior t o death, was not considered to have a neurological complication. Abnormalities in vision included hallucinations, cortical blindness, and in Patient 13, transient bilateral cecocentral scotomata. In Patient 8, cortical blindness followed recurrent generalized seizures and resolved over 24 hours. I n Patient 6, recurrent, throbbing headache with nausea and vomiting subsided after stopping treatment with Cy but later recurred when Cy was resrarteil. This patient then developed a chronic pain syndrome with an addiction to oxycodone. In Patient 10, headache and hypertension heralded a left occipiral lobe hemorrhage. Peripheral nervous system complications of orthotopic h e r transplantation included brachial plexopathy in 2 patients and sensorimotor polyneuropathy in 1 patient. In Patient 5 , the affected plexus was ipsilateral to an axillary seroma, which developed at the site ~ u e J for venovenous shunting. T h e plexopathy in Patient 12 was axon ioss in type, involving the left lateral and posterior cords. Venovenous shunting was not employed in this patient. In Patient 3, the sensorimotor polyneuropathy, axon loss in type, became apparent 5 weeks following transplantation as the patient recovered from a comatose state complii:-ated by peritonitis.

Table 4.Labaratov Data

Patient No.

Cerebrospinal Fluid Analysis

Electroencephalography

Computed Tomography

NL NL NL 20 WBCs Protein, 95 mgidl ND ND NL Protein, 75 mg/dl 10 WBCs Protein, 56 mg/dl

Gen slow Gen slow Gen slow Gen slow BlED ND NL Gen slow NL ND ...

NL NL Cortical atrophy NL

ND ND ND NL

ND Delta coma ND Gen slow

Left occipital hemorrhage ND ND Cortical atrophy, basal ganglia infarctions

Magnetic Resonance Imaging

Group I 1 2

3 4 5 h 7 8

9

ND ND ND ND ... ND ND NL ND ND ...

...

ND NL NL NL NL ...

Group I1 10 11

12 13

NL

=

normal; Gen

=

generalized; ND

=

not done; BIED

=

bilateral independent epileptiform discharges; WBCs

Lboratov Studie.r Table 4 lists the cerebrospinal fluid (CSF), EEG, CT, and magnetic resonance imaging (MRI) findings in the neurologically affected patients. Cy levels were not abnormally high in 12 of the 13 affected patients. In Patient 1, however, following an intravenous bolus of Cy and the occurrence of generalized seizure activity, the blood level was 3,015 ng/ml. The use of high-dose steroids and the presence of hypertension, fluid retention, or hyperosmolality could not be uniformly associated with the neurological complications. Serum magnesium levels in all patients and cholesterol levels in 1 patient (Patient 13) were normal during neurologically symptomatic periods. Autopsy Data A postmortem examination was performed on 2 patients with neurological abnormalities. PATIENT 1. Death was caused by an overwhelming cytomegalovirus infection 2 months following transplantation. Examination of the brain and spinal cord revealed diffuse subarachnoid hemorrhage with areas of demyelination surrounding blood vessels. Alzheimer type I1 astrocytes were seen in the globus pallidus and glial nodules were seen in cortical gray matter. Glial proliferation was also seen in the inferior olives and areas of the cerebellum. A single intranuclear cytomegaloviral inclusion was noted in a cerebellar vessel.

11. Death occurred 10 days following the second transplantation in the setting of disseminated intravascular coagulation. Brain weight was 1,190 gm. There was a yellow-

PATIENT

ND ND NL Cortical atrophy, basal ganglia infarctions =

whire blood cells.

green discoloration of the dura. Except for Alzheimer type I1 astrocytes (prominent in the basal ganglia), there were no other significant findings in the CNS.

Discussion Orthotopic liver transplantation is currently performed in at least 85 medical centers in the United States. In 1990 there were 2,656 liver transplantations, compared with 1,199 in 1987 (M. A. Wirtz, United Network of Organ Sharing, personal communication, 1791). The frequency of neurological complications in our patients is within the range reported by various groups [I-51. Mortality at 1 and 2 years following the transplantation did not seem related to the occurrence of neurological complications. An unusual cerebrocerebellar syndrome was observed in 4 patients. The morbidity from this condition was significant in 3 and at least partially reversed following discontinuation of Cy in 1. We believe that Cy is causally related to both this syndrome and perhaps other neurological abnormalities observed in these patients. This hypothesis is supported by the following: 1. There have been similar reports of neurological dysfunction in renal [6, 16-18}, bone marrow [19-221, and cardiac 123, 241 transplant recipients treated with Cy. 2. Clinical recovery and resolution of white matter abnormalities seen on CT [ l , 18, 22) or MRI [ l , 17, 23, 241 may follow a reduction or discontinuation of Cy. 3. In several instances, discontinuation and rechal-

Stein et al: Liver Transplant Complications

647

lenge with Cy were associated with recovery from and relapse of neurological dysfunction, respectively [ l , 8, 163. Cy is a lipophilic, cyclic endecapeptide with a molecular weight of approximately 1,203 Da C251. Cy selectively affects the adaptive immune response through the inhibition of helper-inducer T-cell activities and the promotion of suppressor-cell function 1251. The principal Cy metabolite (M17) has been found in abnormally high amounts in patients with neurological dysfunction when Cy levels were considered “therapeutic” 1261. The mechanism(s) by which Cy could cause neurological complications is unknown. We suggest the following two possibilities: 1. Primarily, by direct toxic effect on cellular processes within the exposed areas. 2. Secondarily, through ischemia caused by Cy’s effect on the intracranial vasculature. Our knowledge of the effects of Cy on the vascular endothelium is derived mainly from research on the pathogenesis of renal toxicity, which occurs in nearly all patients taking Cy {25]. Cy alters the balance of prostacyclin and thromboxane such that vasoconstriction predominates. This effect may be mediated through an augmentation of thromboxane A> generation, as evidenced by the capacity for thromboxane synthetase inhibitors to reduce Cy-induced renal dysfunction [251. Angiographic visualization of the cerebral vessels or attempts to treat neurologically affected patients with vasoactive drugs have not been documented, to our knowledge. The association between neurological abnormalities and Cy in the setting of hypomagnesemia [20), hypocholesterolemia { I], aluminum overload 1271, hypertension, high-dose corticosteroid therapy {28), or beta-lactam antibiotic therapy has been well documented. We were unable, however, to substantiate a cause-effect relarionship among any of these variables in our patients. In a report from Martinez and associates [9}$complete neuropathological studies on 5 5 adults and 30 children following orthotopic liver transplantation revealed various forms of cerebral ischemia, most commonly global, various infectious processes, and very common astrocyrosis with prominent Alzheimer type I1 cells suggesting hepatic or other metabolic encephalopathy. Among this group, 61 patients had encephalopathy, 14 had coma of sudden onset, 30 had seizures, and 15 had focal abnormalities. Subsequent reports C7, 10, 29) correlated the occurrence of seizures, mental status changes, and focal deficits with ischemic or hemorrhagic strokes, niyelinolysis, and CNS infection. Bra-

648 Annals of Neurology Vol 31 No 6 June 1992

chial plexus injury, as observed in 2 of our patients, was the subject of a recent report by Katirji 112). Development of alternative immunosuppressants, such as FK 506, will provide options for both primary treatment and rescue therapy in those patients who do not tolerate Cy. The neurological side effects a?9soc1’ ated with FK 506 include headache, paresthesias, and tremor {30]. Cy remains a mainstay in our immunosuppressive armamentarium. Its use in the liver transplant recipient has resulted in an increase in the 1-year survival rate by about 20% [ 3 11. Nonetheless, our data suggesit that neurological complications may be lessened or reversed by discontinuation of Cy following the ap:pearance of mental status changes or cerebellar dysfunction. Routine preoperative neurological evaluation and careful postoperative examination will allow further definition of the causes and consequences of neurological dysfunction in the liver transplant recipient. i

Presented at the annual meeting of the American Neurological Association, Atlanta, GA, October 16, 1990

References 1. deGroen PC, Aksamit AJ. Rakela J. et al. Central i1ervou.s system toxicity after liver transplantation: the role of cyclosptxine and cholesterol. N Engl J Melt 198-,317:861-866 2. Rowley H A , K,&u D A , Ascher NL, ct a]. Neurologic findings in 100 consecutive liver transplant recipients. Neurology l990;4O(suppi): 181 3. Vogt DP, Lederman RJ, Carey W D , et al. Neurologic complications of liver transplantation. Transplantation 1988;45:1057I061 4. Busuttil RW, Colonna JO, Hiatt JR, ct al. The first 100 liver transplants at UCLA. Ann Surg 198’;206:387-402 5. Krom RAF, Wiesner R H , Rertke SR, et al. The first 100 liver transplantations at the Mayo Clinic. Mayo Clin Proc 1980;64:

84-04 6. Katirji MB. Visual hallucination and cyclosporine. Transpldntation 1987;43:768-769 7. Wszolek ZK, McComb RD, Pfeiffer RF, et al. Pontine and

8.

9. 10.

11.

12.

13.

extrapontine myelinolysis following liver transplantarion: relationship to serum sodium. Transplantation 1989;48:1006- 1012 Deierhoi M H , Kalayoglu M, Sollingcr H W , et al. Cyclosporine neurotoxicity in liver transplant recipients: report of three cases. Transplant Proc I988;2O: 1 10- 1 18 Martinez AJ, Estol C , Faris AA. Neurologic complications of liver transplantation. Neurol Clin 1988;6:327-348 Estol CJ, Lopez 0, Brenner RP, et al. Seizures after liver transplantation: a clinicopathologic study. Neurology 1980;W: 1297- 130 1 Bird GLA, Meadows J, Goka J, et al. Cyclosporin-associated akinetic mutism and extrapyramidal syndrome after liver tt-ansplantation. J Neurol Neurosurg Psychiatry 199)0;53: 1068- 1O? 1 Katirji MB. Brachial plexus injury following liver transp1;inration. Neurology 1989;39:736-’18 Starzl TE, Iwatsuki S, Esquivel CO, et al. Refinements in the surgical technique of liver transplantation. Semin Liver Dis

1985;5:349-355

14. Shaw BW, Martin DJ, Marquez JM, et al. Advantages of venous bypass during orthoropic transplantation of the liver. Sernin Liver Dis 1985;5:344-348 15. Wall WJ, Grant DR, Duff J H , et al. Liver transplantation without venous bypass. Transplantation 1987;43:56-61 16. Wilczek H, Ringden 0, Tyden G. Cyclosporin-associated central nervous system toxicity after renal transplantation. Transplantation 1984;39: 110 17. Scheinman SJ. Reinirz ER, Petro G, et al. Cyclosporine central neurotoxicity following renal transplantation: report of a case using magnetic resonance imaging. Transplantation 1990;49: 215-216 18. Berden JHM, Hoitsma AJ, M e n JI, et al. Severe central nervous system toxicity associated with cyclosporine. Lancer 1985;1:219-220 19. Atkinson K, B i g s J, Darveniza P, et al. Spinal cord and cerebellar-like syndromes associated with the use of cyclosporine in human recipients of allogeneic marrow transplants. Transplant Proc 1985;17: 1673- 1675 20. Thompson CB, June C H , Sullivan KM, e t al. Association between cyclosporin neurotoxicity and hypomagnesaemia. Lancet 1984;2:1116-1 120 21. Noll RB, Kulkarni R. Complex visual hallucinations and cyclosporine. Arch Neurol 1984;4 1:329-330

22. Rubin AH, Kang H. Cerebral biindness and encephalopathy with cyclosporine-A toxicity. Neurology 1987;37: 1072-1076 23. Lane RJM, Roche SW, Leung AAW, et al. Cyclosporin neurotoxicity in cardiac rransplant recipients. J Neurol Neurosurg Psychiatry 1988;51:1434-1437 24. Hughes RL.Cyclosporine-related central nervous system toxicity in cardiac transplantation. N Engl J Med 1990;323:420-421 25. Kahan BD. Cyclosporine. N Engl J Med 1989;321:1725-1738 26. Kunzendorf U, Brockmoller J, Jochimsen F, et al. Activity of cyclosporin metabolites M17 and M1. Transplant Proc 1990; 22: 1697- 1699 27. Nordal KP, Talseth T, Dahl E, er al. Aluminum overload, a predisposing condition for epileptic seizures in renal transplant patients treated with cyclosporin? Lancet 1985;2: 153-154 28. Polson RJ, Powell-Jackson PR, Williams R. Convulsions associated with cyclosporin-A in transplant recipients. Br Med J 1985;230:1003 29. Estol CJ, Pessin MS, Martinez AJ. Cerebrovascular complications after orthoropic liver transplantation: a clinicoparhologic study. Neurology 1991;4 1:815-819 30. Shapiro R, Fung JJ, Jain P, et al. The side effects of FK506 in humans. Transplant Proc 1990;22:35-36 31. deGroen PC. Cyclosporine: a review and its specific use in liver transplantation. Mayo Clin Proc 1989;64:680-689

Stein et al: Liver Transplant Complications

649