cmv hepatitis after liver transplantation: incidence, clinical course, and long-term follow-up

CMV Hepatitis After Liver Transplantation: Incidence, Clinical Course, and Long-Term Follow-Up

Daniel Seehojiet.,. N& Rayes, Stefan G. Tullius, ' Christian A. ~chmidt, ' UYP. Neumann, ' Cornelia ~adke,' Utz Settrnacher, ̂ Andrea R. Miiller, '

Thomas Steinmiilh, ,' and Peter Neuhaus"

C y t a m c g u l ~ (CMV) hepatitis is hi u the most kquent manifatdon of CMV tirsue invasivc direau Pfter livcr transplantation. Its correlation with HLA- matching, hepatic may thmmbdi, and chmnic rejection is still conuovenid. Risk factors, incidence, d i n i d counc, and compliutiow of CMV hepatitis were respectively analyzedin a 12-yeuseriu of 1,146 consecutive livcr vpnsplantatiow in 1,054 patients. AU patients received only low-dose acydovk but no pnfydovir prophylaxis. CMV infection was disgnod by virhl culture, pp65 antigenemin, or by polymerpae chain reaction (PCR). CMV hepatitis wu pmvcn by liver biopsy. Trent- mcnt of CMV disew consisted of inmvcoou, gauciclovk for a minimum of 14 &p. Long-turn follow-up of patients included monthly motine laborntory vduu and mutine livcr biopsiu 1, 3 and 5 yenrs after uuuplantn- tion. CMV hepatitis wu a rare went h e r livcr unnsplan- tntion, with n t o d incidence of 2.10% (24 cnsu). It apr significantly more kqucnt in CMV semucg~tive (5.2%) than in seropositive recipients (0.7%). The leading indication in patients with CMV hepatitis wpr HCVcirrhosis (n = 8). The maximum number of pp65 positive white blood cells was 82 * 23 per 10,000 cells. Most courses manifated u isolated he~atitist onh 2 ~atients had dia- - , seminated disease. N i c oj24 patients h d received OKT3 monocloud antibodiu because of steroid-ruistant rejection before CMV hepatitis. In seronegative patients with CMV hepatitis, 71% revealed 1 or 2 HLA DR matchu, in contrast to 32% in paticots without CMV hepatitis. One-, 3-, and 5-yeat paft S U M V ~ wpr 7896,6596, and 59% in patients with CMV hepatitis compared with 88%. El%, and 79% in patients without Chronic rejection was observed in one patient, but already bdore onset of CMV hepatitis. Beoeath DCR-constellation and OKT3 treatment ai risk facton, HHU DR-matched BrPfLs and HCV seem to h r m a n i b t i o n of CMV hepatitis ha liver tr~nsplsntntion. Long-term complications of CMV hcpa-

Fmm r l n 'DqumnmrofGmorrl Vmcnul, und TmmpkznrSugr3, rhe tDpmnmr o fHmaro~undOncob~. andrhr #Dcpamntnr of Parholog, Chnrirl Gmpw Viwhmv. HumboUt Uniwrrip of Rmlin.

titis were not ob.cnrd and aped.lly no comlztion with chmnic rejection wu found. (Livrr Tranrpl 200218: 1138-1146.)

e ~nc~dence of cytomcgalovirus (CMV) infection T h ' after ' orthotopic liver transplantation (OLT) depends on several factors, in particular the use of different diagnostic tools for CMV.12 the donor and recipient semstam, the periopeative prophylaxis,3" and the immunosuppressive and antirejection management.' Therefore, in different series, the incidence of CMV infection after OLT ranges from 25% to 85%.1"-8A high percentage ofCMV infections remain asymptomatic, whereas a smaller number of patients develops symptoms such as fever, leukopenia, or thrombocytopenia, the so-called CMV viral syndrome. Only a minority develops organ manifestation (i.e., tissue invasive disease). This is divided into isolated disease, affecting only one organ, and disseminated disease, ddined as localized disease in two or more noncontiguous sites. Tissue inwive disease after OLT most frequently a&ts the liver graft.' In older series, the incidence of CMV hepatitis after OLT is reported to be as high as 7% to 17%.9-11 The occurrenceof CMV hepatitis is thought to be supported by HLA-DR- marched g&'O caused by MHC-restricted lympho- cyte responsiveness to CMV. This increased frequency of CMV hepatitis is theorized ro be one of the revons that HLA-DR-matched livers were found to have lower survival rates in several studies.12.'3 In addition, CMV infection and, especially, CMV hepatitis are suspected to enhance chronic allograft rejection, although this remains controversia1.l"~" Most reports about CMV hepatitis after OLT are dated from the early 1990s or even before'J0.1' and do not represent current immu-

B d n , G m y . nosuppressive regimens and periopeative management Ad& nprinr nqum m Dnnirl Srrhfi., MD. Lkpumnmr of

GmnuL Vi1cera4 und Trawpbnr S u m , Charitl Gmpw Viwhow, of liver transplantation. Furthermore, long-tenn

~ ~ ~ ~ ~ b , , ~ phe 1, 0-13353 B&. cmnv j-*hhnc 0049 reprts are not available. The incidence of CMV disease , . 30"450552&1; Fm OO493045055229W;E-mil~~nirls~hofir@ is d e ~ n d e n t on the net state of immunosuo~ression.~~ . . tknrt.& and, because immunosuppressive and antiviral drugs C o p M t 0 2002 by rbe Amm'un hmiruion fir tht Sndy of I is- nirrnrx have changed during the last decades, a series of 1,200

consecutive liver transplantations within the last 12 years was andyxd retrospectively concerning ind-

1138 Liver Tramp&ntnrion, Vol8, No I 2 (Dtcmber), 2OOZ:pp 1138-1 I46

dence, risk factors, clinical features, and long-term outcome of CMV hepatitis.

Patients and Methods

Patient Population

Bctween September 1988 and April 2000, a total of 1,200 liver transplantations were performed at our center. Fifty-four transplantations inchildren younger than 16 years of agewere excluded from the analysis. Thus 1,146 liver transplantations in 1,054 adult patients retrospectively were analyzed. Indica- tions for liver transplantation were viral hepatitis (n = 3821, alcoholic cirrhosis (n = 230), cholestatic liver disease (n = 144), autoimmune cirrhosis (n = 461, cryptogenic cirrhosis (n - 81), cryptogenic acute liver failure (n = 27), and other less common causes. Orthotopic liver transplantation was performed with standard techniques in most cases with extra- corporeal bypass." Immunosuppression consisted of tacrolimus (n = 597) or cyclosporine (n = 549) plus prednisolone. In addition, azathio~rine (0 = 564) or mycophenolate mofetil (n = 103) were added to the immunosuppressive regimens. An induction therapy with antithymocyte globulin (ATG) or intedeukin-2 (IL2)-receptor antibodies was performed in 379 and 216 patients, respectively. Rejection epi- sodes were treated with methylprednisolone (500 mg per day intravenously for 3 days). In case of steroid-resistant rejection,

prophylaxis for 6 weeks. CMV prophylaxis with CMV hyper- imm~no~lobuline (Cytotect, Biotest, Dreieich, Germany, 1 mUkg bodyweight) at postoperative days 1 and 14 was used until 1996. Thereafter, no CMV prophylaxis was applied, especially no gancyclovir prophylaxis was used during the wholestudy period. Between 1993 and 1996, half ofthepp65 positive asymptomatic patients received pre-emptive therapy with intravenous gancydovir; between 1996 and 2000,50% of the pp65-positive, asymptomatic patients received pre- emptive therapy with oral gancyclovir, both within random- ized studies.20.2' Treatment of CMV disease consisted of intravenous gancydovir (Cymeven, Roche, Germany, 5 mglkg bodyweight twice daily or adapted to renal function) for a minimum of 14 days. Until 1996, CMV Ig was given additionally at a dosage of 1 mUkg bodyweight every other day.

Labomtory Values and Long-Term Follow-Up

Biochemical and hematologic parameters and HLA typing were determined with standard laboratory methods. After primary hospitalization, patients were seen on a regular basis in our outpatient clinic. In addition, liver biopsies were taken at least 1, 3, and 5 years after transplantation, or in case of abnormal laboratory values. Doppler ultrasound was performed regularly at primary hospitalization and at least yearly thereafter. If hepatic artery thrombosis (HAT) was suspected, an angiogtaphy was initiated.

patients were treated'with 5 mg/d of monoclnnal antibody Statistid Analysis OKT 3 (Orthodone, Cilag, Germany) for 5 to 7 days.

Data are presented as mean and standard error of mean. For CMV-Surveillance comoarison of values within erouos. the Wilcoxon test for

viral cultures and polymerase chain reaction (PCRjof b l o i , categoric variables was conducted with ,f test. Multivariate

urine, and throat specimens. Since 1993, CMV-pp65 anti- analysis of risk factors for CMV hepatitis was performed with eeneminwlqdererminedwirh techniouc (Clonah, binary logistic regression analysis. Actuarial patient survival n........- ~~~ ~~ ~ L , -

Biotest, Dreieich, Germany) as described p t~ ious ly . '~ Blond rates were calculated with the Kaplan-Meier estimation. Dif-

was investigated weekly during primary hospita[ization after ferences between groups were log rank test. Dif-

and addidonally i f ~ M V infection was suspected, CMV ferences were considered significant if the P-value was less than .05. All statistic analyses were performed with SPSS 10.0 infection was defined as pp65 antigenemia of at least 0.5 .. . -. . .. . . . -

uositive cells per 10,000 leukocytes. CMV disease was classi- (S1'SS Inc, Chicago, IL).

fied as CMVsyndrome (antigenemia plus fever, leukopenia or thrombocytopenia) or tissue invasive disease (antigenemia plus hepatitis, pneumonia, gastroenteritis, or involvement of other organs). In the case of clinically relevant elevation of liver enzymes (transaminases, cholestatic liver enzymes, or both), a percutaneous liver biopsy was initiated and investigated histomorphologically and immunohistochemically for CMV hepatitis.19 In the case of suspected involvement of other organs, respective biopsies were taken and investigated accordingly.

Antiviral Prophylaxis and Treatment

All patients received oral low dose acyclovir (200 mg three times daily or adapted to renal function) as herpes simplex

Results

Incidence and Virologic Data of CMV Hepatitis

In total, after 254 out of 1,146 transplantations (22%) a C M V infection evolved. I t was asymptomatic in 50.4% of the patients; a CMVviral syndrome occurred in 28.0% and a tissue invasive disease in 21.7%. CMV hepatitis was diagnosed in 24 patients of the study population (2.1%, Table I), either by liver biopsy (n = 20), or by clinical and laboratory course (n = 4), if liver biopsy was contraindicated. Of the 1,146, 746 transplantation patients were seropositive for CMV IgG

.Y .. .- .. 5 Z > 3

Y

.- 2 L

.P x .- e: " z U i

Y .-

0- .- E

.

m

e m

I1 E I I I I g I I I I I l l "5 2 5

" = .- 2 .; u ci

ci

0

v o : 2 'C 3 .s .g

n -

u a z

; , $ I I I I I I I I I ~ I I I I I ~ I I I I I I I I X 0

P.

m m m ~ w m ~ o m m r , , ~ m n \ r m ~ ~ m m n m m m

+ 1 + 1 1 1 1 + 1 + 1 1 I I 1 + 1 I I I f 1 I I

I I 1 + 1 + + 1 I + + + + + + + I I l + + + + +

w .- - 3 5 2 .c .- A E . = z a 6 8 M

z " - " + rn - 2 2 u.. 2 2 + 0 2 VCC>>.gu-+Gu& > > > " 2 T > $

4 $ 2 4 8 $ m u ,, U U U ~ u . 5 8 3 0 ." < & ! z f f i ~ > z z m ~ 5 ? z 2 z Z Z Z ~ Z Z - ; ~ < Z E ~

2 &- m .- A E c C : 4 Z r i a -b ̂

0 F'

.. g > , E u

" $ 3 0 0 2

g = * z g - n

2 % .

80 . O . r ,j g C M .- 2 6

& %

2": .9 + 2 d

.Z .. .- .- .j a o LL 5 + - + $ 5 a g -

0 0

: &B < G

Z .- - 2

4 1 4 1 1 1 I I I I I 1 2 I I I 1 2 1 I I I I I

I 1 1 1 I I I I I I I I I I I 0.

+ I I I I I I I I I I + + ? w m * Z "

* O h + -.. c 2 G 2 z g z

I I I I I I I I I + + - N " N m $ ~ f ; , N ~ ~ n

s * * +9 o n o m ~ m w - a o c . m m , , o n o m n c . n o ~

W ~ W = ~ ~ . W W ~ - N - - - - . a n n e m m h ~ b m - ~

E E E E E L L E L E E r E E E L E E E E E E E L , , , , , , , , , , ~ ~ - m ^ ~ ^ N ^ ~ ~ ^ N ^ . + N ^ ~ N ^ w - - - ~ ~ m ~ ~ m n n - a ~ n m m m m ~ ~ m n n m ~

- m m , m w - m c . = 22, Z , 2 > z m o - N m \ f N N N N N

a c 'E % 2 6 u 0,

." c .z ra c

2 5 a Y - C

; a ' Z 6j .- a ', $ < > ; z u" E Z % -. a + 2 . 5

2 0 "

'2 + v -

LY * .G 'g > e 3 .+ Y C - * % > 2 5

e of CMV HI :panns in Dlf

- AU CMV Seropositive CMV Semneprive

(n = 24) (n = 5) (n = 18)

Hepatitis C 4.2% 1.6% (21127) 12.2% (6149) Hepatitis B 1.1% 0.6% (11154) 3.6% (1128) Alcohol 1.3% - 3.5% (3183) PBC 2.2% - 7.4% (2127) PSC 1.9% - 3.7% (1127) Auraimmune 2.2% - 5.9% (1117)

NOTE (Differenccr were nor smrisrirnlly signifimr by $ tut.)

before OLT (65%); 348 were seronegative (30%). In 52 patients (5%), no preoperative CMV s tam was available. From the 24 cases of CMV hepatitis, 18 were observed in seronegative patients and only 5 in seropositive recipients; in one patient, CMVstatus before OLT was unknown. Thus, the incidence of CMV hepatitis was 5.2% in seronegative recipients and 0.7% in seropositive recipients (P < .001 by ,$ test). In patients with CMV-hepatitis the leading indication for OLT (Table 1) was hepatitis C (n = 8), followed by alcohol cirrhosis (n = 3), hepatitis B (n = 2), acute liver failure (n = 2) and primary biliary cirrhosis (PBC) (n = 2). The incidence of CMV hepatitis was higher in hepatitis C patients (4.2%) than in patients transplanted for other indications, although this did not reach statistic significance (Table 2). Most cves of CMV hepatitis occurred between week 4 and 8 aftcr OLT (Fig. 1). Only three cases were seen after week 8, two of them during HCV reinfection. Twenty-one cases of CMV hepatitis were observed More 1996, only 3 afterwards.

The mean number of pp65-positive cells was 11 5 3 per 10,000 cells at the time of first diagnosis of CMV infection. The mean maximum number of positive cells

per 10,000 was 82 t 23 (Table 1). All except two early cases had isolated CMV hepatitis without involvement ofother organs (Table 1). RecurrenceofCMVhepatitis was observed in 2 patients 6 weeks (patient 1) and 9 months (patient 3) after primary infection; both had a third episode of CMV infection without hepatic involvement. Recurrent CMV infection with CMV syndrome but without hepatitis was seen in 2 other patients (13 and 18).

Clinical pad Biochemid Course

Squential laboratory values within 2 weeks before and after diagnosis of CMV hepatitis were available in 20 of the24 patients. In theother four, only occasional values were available, and they were therefore excluded from the following analysis. The biochcmid profile during CMV hepatitis was characterized by a marked elevation of tnnsaminases, with a mean of 3.7 times the baseline value of alanine aminotransferase (ALT) within 1 week before start of intravenous gancyclovir. The mean AP was oily 1.8 times elevated at time of diagnosis, but elevation of alkaline phosphatase (AP) and gammaglu-

weeks after liver transplantation

Figure 1. Timing of CMV hepatitis after OLT: a maximum incidence between week 4 and week 8 was found. From the three late cases two had simultaneous HCV reinfection.

-I4 .I a 0 3 7 10 14 21

days bBfm Isner stan ofganciclovir treatment

Figure 2. Courses ofALT and AP activity before and after start of intravenous ganciclovir treatment of CMV hepatitis. (#: p < 0.05 vs. treatment s ta t (=O) by Wdcoxon test.)

Figure 3. G u r s a of white blood cell count and platelet count before and after start of intravenous ganciclovir treatment for CMV hepatitis (#: p < 0.05 vs. treatment start (=0) by Wilcoxon tat).

tamyl uansfuw (GGT) persisted longer than ALT and aspartate vnhomnsfense elevation (Fig. 2). The peripheral blood count revealed a 35% decrease of platelets, which evolved before dugnosis of CMV h c p atitis and in some patients wen before pp65 antigenemia. Platelet count markedly increased with the onset of intravenous gancydovir treatment (Fig. 3). In contrast, a significant leukopenia was observed only in four patients; the mean white bkod -t did not decrease signifiantly before or a f e r diagnosis of CMV hepatitis (Fig. 3). Fwqofgratcr than 38.5"Cwar observed only in seven patients; mean temperature at diagnosis was slightly elevated at 37.6 t 0.4"C.

Imm-ppnsrion

From the 24 patients with CMV hepatitis, five received a dual, i 1 a triple, and eight a quadruple primary immunosuppression. ~ h ; used ddneurin inhibitor was cydosporh A in 15 and tacrolimus in nine patients. Fifrecn and one patients also received azathioprine and

'able 3. lnflr ;uppression ( dV Infection

mycophenolate mofetil, mpmivdy. An induction therapy wy petformed wkh ATG or antilymphocyte globulin (ALG) in 8 patients, with anti-112 receptor antibodies in five patients. The general immunosuppressive rrgLNn showed a trend during the study period h n t t-riple or quadruple immunosuppmJion in the arly phase towards dual immunosuppression (Table 3). Sudnic andyais mraled a significantly lower incidenoe ~f CMV infection in patients with dual immunosuppression, but no lower incidence ofCMV hepatitis in this grmp (Table 3). Likewk, the type of ddneurin inhibitor had no significant impact on the incidence of CMV hepatitis. Also, an antibody induction therapy with ATG or 112-receptor antagonists had no statistic influence on rhc development of CMV hepatitis.

H U Matching

B n a w CMV hepatitis was obauvcd mostly in CMV negative patienu, the influence of HLA compatibility on CMV hepatitis was analyr~d in the 348~semneptive patients. A complete HLA typing of donor and tecipi* ent war available in 278 patients (80%); in the remain- ing 70 ptienu, donor or racipient HLA typing wu incomplete; therefore, they we= excluded from the analysis.

For MHC c l w I matches, no differences were observed at HLA A and HLA B marches between patients with and without CMV hepatitis (Table 4). In contrast, HLA DR matches were significantly mo* frequent in patients with CMV hepatitis than in patients without. Ftom the 14 analyzed patients, 10 (71%) had one or two HLA DR matches. In contrast, only 24 of 72 (33%) patients with CMV infection but

Triple or Qu; lepariris

~ence of Primary Immuno: Dual Versus' Ch and CMV H

- Dual 'I'riplelQ~adru~le

lrnmunoruppresrion Immunosuppression P'

1988-1992 18.2"h 81.8% 1993-1995 21.8% 78.2% 1996-2000 38.9% 61.1% <.01

Paricnr withour CMV infection 30.0% 70.0% Patient with CMV infection 17.8% 82.2% <.01

Parienr wirhour CMV hepatitis 27.5% 72.5% Patienr wirh CMV hepatitis 20.8% 79.2% .47

'By 2 rcst.

- Table 4. InHuence of HIA Marcha on CMV Infection and CMV Hepatitis in CMV Seroncgarivc Reripk

CMV Heparib Other CMV Infcnion No CMV Infcnion H U Matcher (D = 14) (n = 72) (n = 192) P

A 0 57.1% 59.796 58.3% I 42.9% 36.1% 35.4% 2 - 4.2% ' ' 6.3% .83

B 0 64.3% 75.0% 71.9% 1 35.7% 25.0% 25.0% 2 - - 3.1% .47

DR 0 28.6% (n = 4) 66.7% in = 48) 68.2% (n = 131) 1 50.0% (n = 7) 3 1.9% (n = 23) 28.6% (n = 55) 2 21.4% (n = 3) 1.4% (n = I) 3.1% (n = 6) ,001 oQ

0 85.7% 80.6% 81.3% I 7.1% 15.3% 14.6% 2 7.1% 4.2% 4.2% .93

Missing HLA wing n = 4118 (22%) n = 16/88 (18%) n = 501242 (21%)

'By ,$ test.

without hepatic manifestation revealed HLA DR matches. Also, in seronegative recipients without CMV infection, only 32% had one or two HIA DR matches. In addition, from the 10 analyzed patients with two HLA DR matches, four developed CMV infmion. In thm of there four patients (75%), CMV hepatitis evolved. The HLA D Q matches showed no differences in patients with and without CMV hepatitis.

Acate Rejcction/Chronic Rejcaion

Of 24 patients, 14 (58%) had a histologically proven mjecrion before occurrence of CMV hepatitis that was mated with steroids. In nine of thw 24 cases (38%), OKT3 monodonal antibodies were used after steroids because of persistent rejection. In one patient (No; 2), rejection persisted after OKT3 treatment, and the patient died bcause of infectious compliutions after rejection treatment. CMV hepatitis in this patient occurred after 0 K T 3 treatment.

In one patient (patient lo), chmnic rejmion was diagnosed before CMV hepatitis. This patient was u~nsplanted for PSC and retransplanted because of HAT, and chronic rejection with ductopenia was diag- noad in the second graft. Aftetwards, while waiting for asecond retransplantauon because ofchronic rejection, CMV hepatitis evolved. In patient 16, liver biopsy 3 years after OLT revealed alterations of bile ducts without ductopenia. This biopsy was suspicious of early

chronic rejection, but it remains to be confirmed in the further course. All other patients were followed up with regular liver biopsies, but no signs of chronic rejection were detected. Other histologic abnormalities were caused by underlying lim disease, especially viral h e p atitis or rmmence of alcohol abuse.

In parallel to the persistent elevation of cholestatic liver enzymes for several months, in sequential liver biopsies an inllammation of the small bile ducts was detectable several weeks long& than inflammation of the liver parenchyma.

The following paramems were analyzed, but univariate pnalysis showed no statistic correlation with the development of CMV hepatitis: preoperative Child-Pugh score, cold ischemic time, intraoperative transfusions, operating time, duration of postoperative ventilation, and mranrphtation. In the multivariate analysis, the following three factors were significant risk factors for the development of CMV hepatitis: CMV-high risk constellation (D+R-), OKT3 treatment, and HLA-DR matches.

Hepatic Army Thromboat (HAT)

Hepatic artery thrombosis was found in 3 of the 24 patients (12.5%). In patient 7, HAT was diagnosed 12

no CMV heparmJ (n=11221

5 :: 60%

CMV hepalilis (n=24)* I y88n aRer liver transplanlatian

Figure 4. Actuarial grift S u ~ v a l of patients with CMV hepatitis (interrupted line) in comparison with patients without CMV hepatitis (solid line) by Kaplan Meier estimation (p = 0.04 by log rank test). The difference was not significant (p = 0.33) if patients with pre-terminal graft failure already before development of CMV hepatitis were excluded (gray line).

days after transplantation, which was before onset of CMV hepatitis. In patient 13, HAT and CMV hepatitis were diagnosed simultaneously, and in patient 17, who was finally retransplanted because of HAT, CMV hepatitis was diagnosed before HAT. All others had normal perfusion of the hepatic artery during the follow-up period.

Survival and Long-Term FoUoa-Up

One-, 5-, and 10-year graft survival rates in patients with CMV hepatitis were 78%,65%, and 59%. respectively, compared with 8S0/0, 81%, and 79% in patients without (Fig. 4). This difference was significant by log rog rank test (P = .04), but it was not significant, if three patients in which CMV hepatitis occurred during terminal gnft dysfunction were excluded from the analysis (P = .33, Fig. 4). In total, 3 patients ofthe 24 were retransplanred. Indications were hepatic artery thrombosis (17), chronic rejection (10). and destructive bacterial cholangitis (1); one of them died aftn retransplantation (10). In the patient with chronic rejection, CMV hepatitis occurred late after onset of chronic rejection, when the patient was already listed for retransplantation.

Seven patients with CMV hepatitis died within the follow-up period of 6 to 122 months (mean, 76 month;). Reasons for death were septic multiorgan failure (2). CMV pneumonia (4), de novo HBV infection (1 1). recurrence of alcohol abuse (6). HCV recurrence (13). and recurrent carcinomas (18, 19). No grafi loss caused by CMV hepatitis was observed but one early patient (4) died of a generalized CMV infection. In this case, mild CMV hepatitis occurred a few days before

death, when the patient was already mechanically ven- tilated because of severe CMV pneumonia.

Discussion

The present analysis reports long-term follow-up, ind- dence, and clinical course of CMV hepatitis after liver transplantation. CMV hepatitis was found to be a rare event after OLT with a tonl incidence of 2%, which mostly manifested in CMV IgG-seronegative recipients of CMV-positive grafa. A subclinical hepatic involvement in further ases cannot be ruled out, because only patients with elevation of liver enzymes underwent liver biopsy, but this is not clinically relevant. Additional risk fictors for CMV hepatitis were OKT3 treatment,' HLA DR matched grafts, and OLT for HCV cirrhosis. The low incidence of CMV hepatitis is in contrast to earlier reports, in which a frequency of up to 17% was found.9,'O This difference might be explained by reduced net immunosuppression with increasing use of dual immunosuppression and resulting reduced frequency of CMV infection. However, the use of dual immunosuppression lowered only the overall incidence of CMV infection, but not of CMV hepatitis, which was influenced by other factors (see previously in this article). In addition the use of pre-emptive therapy, which was used at our center since 1993 in 50% ofthe asymptomatic pp65-positive patients, might have con- tributed to the low incidence of CMV hepatitis, especially in the later study period. The very low incidence after 1996 in the present series is attributable to a com- bination of different fictors, including a b a l a n d immunmupprasion, p-ptive therapy and advanm in the monitoring of CMV infection.

The liver was the most common siteofCMVdisease after 0LT,'t6 but one has to consider that other sites are less commonly investigated, for example the gastrointestinal uact (GIT), where Alexander et al22 found 33% prcvalenceof gastrointestinal CMV infection in routine endoscopic procedures. Severe disseminated disease was seldom due to an effective antiviral treatment of CMV infectionlhepatitis with intravenous gancydovir. This is underlined by prompt dedine of symptoms within 2 to 3 weeks and no graft failure antibutable to CMV hepatitis.

In accordance with earlier repom? the incidence of CMV hepatitis in our patients was almost 10 times higher in seronegative than in seropositive recipients. If CMV hepatitis was observed in sero-positive recipients, then ir occurred most commonly in patients with simulraneous recurrence of viral hepatitis. This comor- biditiy also has been reported by other authors for hep-

16. Neuhaus P, Klupp J, LangrehrJM, Neumann U, Gebhardt A, ct al. Quadruple acmlimus-based induction therapy includingaza- thioprine and ALG does nor significantly improve outcome afrer liver transplanration when compared wirh standard induction with tacrolimur and steroids: rcrulrs ofa prospecrive, randomised trial. 'Sranrphnntion 2000;69:2343-2353.

17. Pichlmayr R, Ncuhaur I', Ringc B. Wonigcir K, Burdclrki M, et

al. Developmenrs in liver transplantarion. Jpn J Surg 1985:15:

26. Diicke WD, Priisch S. Fietze E, Kimrl V, Zuckcrman H, cr al. Cytomegalovirus reacrivation and rumor necrosis factor. Lancer L994;343:268-269.

27. O'Gndy JG, Alexander GJ. Surhcrlrnd S. Donaldron 1'T. Har- vey F, er d. Cymmegalovirur infection and donor recipient HLA anrigenes: Independent co-factors in pathogenesis of vanishing bileduct syndrome =her liver transplantation. hncct 1988;2: 302-305. ~-

409-41 9. 28. Laurenschlager I. Hiickerstedr K, Jalanko H, Loginov R. Salmela 18. Schmidt CA, Oettle H, I'eng R, ct al. Comparison of pulymense K, et al. Persistant cytomegalovirus in livcr allograft, with

chain reaction from plasma and butTycoat wirh antigen detection chronic rejection. Hcparology 1997;25:190-194. and uccurence of immunoglobulin M for the demonrtrarion of cymmegalovirur infection after liver transplantarion. Transplan-

29. Evans PC, Coleman N, Wreghirt TG, Wighr DGD, Alexandre

ration 1995591 133-1 138. GJM. Cytomrgalovirus infecrion of bile duct epithelial cells,

19. Colina F. J u u NT, Moreno E, Bailertin C, Farina J, cr al. hepatic anery and venous endorhclium in relation to

Histological diinasis of cyromegalovirur hepatitis in liver allo- chronic rejection oflivrrgrahs. J Heparology 1999:31:913-120.

gahs. J Clin Pathol 1995;48:35 1-357. 30. Grundy JE, Reid MF. The effect of primly and seconda~

20. Raves N. Oertlc H. Schmidr CA. Lohmann R Sreinmiiller T. er infection wirh cyromegzlovirur on the host response to alloanti- , . al. Preemptive rhenpy in CiMV-antigen positive patients afrcr livcr rranrplantarion: A prospecrive trial. Ann Transplanration 19W:4:12-17.

21. Ray- N, Sechofcr D. Schmidt CA, Oetrle H. Miiller AR, er al. Prospective randomised trial to usess the value of prccmprivc oral therapy for CMV infection following liver transplantarion. Tran~~lantnrion 2001;72:881-885.

22. Alexander JA, Cuellar RE, Fadden RJ, Genovcse JJ. Gavder JS, cr d. C y t o m ~ ~ l u v i r w infection of the upper gastrointestinal tract bcforc and ahcr livcr rran~~lanration. Tansplancarion 1988;46:378-382.

23. Rosen HG, Chou S, Curless CL, Grctch DK, Flora KD, et al.

genes. Transplant Proc 1985;17:592-594. 31. Paya CV, Wiesncr RH, Hcrmans PE. Lanon-Keller JJ, llsrrup

DM, er al. Lack of association bcrwccn Cytomcgalovirus infecrion, HLA matchingand the vanishing bile duct syndmme afrcr liver transplantation. Hepatology 1992;16:66-70.

32. Settrnacher U, Stangc B, Haasc R, Heise M, Sreinmiiller T , er aI. Arrrrial complicacianr aher livcr transplantation. Tnnrplanr Inr 2000;13:372-378.

33. Madalosso C, de Suoza NF, llstrup DM, Wiesner KH. Krom RAF. Cyromegalovinrr and irr asrocktion with hepatic arrrry rhromborir after liver tnnsplantation~. Tnnrplanrarion I998;66: ,n< -07 L-,L1.L-/,.

Cytomegdovirus viremia: Risk factor for livcr allognh cirrhosis 34. van Dam-Mieru MCE, Mullcr AD, van Hinrbcrgh VWM. aher liver rnnrplznrarion for hrpatitis C. Transplantation 1997; Mullrrr WJHA, Bomanr PHH, ct al. 'fie procoagulant response 64721 -726.

of cytamegdovirus infected rndothrlial cells. Thromb Haemost 24. Singh N, Zecvi A, Gayowski T. Marino IR. Lareonset cyromcg-

alovirus disease in liver transplant recipients: dc novo rcactiva- 1992;68:364.

in rrcurrent hepatitis C, .rnnsplanr 1:308.31 35. Paya VC, Wiesner K, Hermans I'E, hrson-Keller JJ. Ilrrrup

25. Larrea E, Garcia N, Qian C, Civeira MP, PictroJ. Tumor necro- DM, et al. Kirk factors for cytomrrgaloviru\ and severe bacterial

sir facror-alpha gene expression and the response to interfcmn in infections following liver ransplantation: A prospecrive multi-

chronic heparirir C: Hepatology 1996;23:210. variare rime-dependent analysis. J Hepatol 1993;18:185-195.

cmv hepatitis after liver transplantation: incidence, clinical course, and long-term follow-up

Documents