Long-Term Longitudinal Study of Intrahepatic Hepatitis C Virus
Replication After Liver Transplantation
VINCENT DI MARTINO,1 FRANÇOISE SAURINI,1 DIDIER SAMUEL,1 MICHELE GIGOU,1 ELISABETH DUSSAIX,2 MICHEL REYNÈS,3
HENRI BISMUTH,1 AND CYRILLE FÉRAY1
Recurrence of hepatitis C after liver transplantation is com-
mon and can lead to severe liver diseases. Although immuno-
suppression and high levels of viremia suggest a direct patho-
genicity of hepatitis C virus (HCV), the relations between
viral replication and long-term histological course are still
unknown. Thirty-three patients with a mean histological fol-
low-up of 3.5 years (3 months - 8.6 years) were analyzed.
Nineteen patients were infected by genotype 1b. Liver HCV
RNA was determined in parallel with the quantitation of an
internal control (28S ribosomal RNA) by competitive poly-
merase chain reaction (PCR). Lobular hepatitis (LH) and
chronic active hepatitis (CAH) occurred in 27 and 19 patients,
respectively. Levels of liver HCV RNA determined in 84 biop-
sies were higher in cases of LH than in the other patterns (82
{ 123 vs. 19 { 38; P õ .01) and were unrelated to the
genotype. Progression from LH to CAH was associated with
a highly significant decrease of liver HCV RNA (P Å .006),
which was not observed in patients with stable histology.
Among patients with CAH, those infected by genotype 1b had
more severe liver damage and lower levels of liver HCV RNA
than others (P Å .04). Multivariate analysis showed that high
levels of liver HCV RNA at the time of the first posttrans-
plantation biopsy was an independent predictor of CAH (P
Å .01). After liver transplantation, the progression to CAH
together with a decrease of liver HCV RNA suggests that a
host’s response is involved in the long-term viral pathogenic-
ity. This response may be stronger and liver disease more
severe in patients with high levels of replication at the time
of LH and in those infected by genotype 1b. (HEPATOLOGY
1997;26:1343-1350.)
Cirrhosis caused by hepatitis C virus (HCV) is a major
indication for liver transplantation (27% in the US in 1993).1
HCV infection still remains after transplantation and recur-
rent hepatitis is frequent.2 In our liver transplantation center,
Abbreviations: HCV, hepatitis C virus; LH, lobular hepatitis; CAH, chronic active
hepatitis; PCR, polymerase chain reaction; rRNA, ribosomal RNA; cDNA, complemen-
tary DNA.
From the 1Centre Hépato-Biliaire, 2Laboratoire de virologie, 3Service d’Anatomo-
Pathologie, Hôpital Paul Brousse, Assistance Publique-Hôpitaux de Paris, Villejuif,
Faculté de Bicêtre, Université Paris Sud, France.
Received November 7, 1996; accepted July 8, 1997.
Supported by the following institutions: Association pour la Recherche sur le Can-
cer, Direction de la Recherche Clinique (Assistance Publique-Hôpitaux de Paris), and
Institut National de la Santé et de la Recherche Médicale, Paris, France.
Address reprint requests to: Cyrille Féray, M.D., Hôpital Paul Brousse, 14 avenue
Paul Vaillant-Couturier, Villejuif, 94800, France. Fax: 33-1-45-59-38-57.
Copyright q 1997 by the American Association for the Study of Liver Diseases.
0270-9139/97/2605-0038$3.00/0
1343
AID Hepa 0013 / 5p28$$$241 10-13-97 20:58:29
the 5-year actuarial rate of HCV-related chronic active hepati-
tis is 75%3 and the 7-year actuarial rate of HCV-related cir-
rhosis is 15% (Feray C, unpublished data, March 1997). HCV
infection in this setting is characterized by an intense viral
replication.4-6 Both intense replication and accelerated histo-
logical course suggest that HCV is able to directly induce
liver damage after transplantation. However, the relations
between the level of HCV replication and the severity of liver
disease or the occurrence of histologically-defined hepati-
tis,6,7 are far to be clear.4 Some patients have high levels of
serum HCV RNA without evidence of liver damage whereas
others have severe chronic hepatitis in spite of low levels of
serum HCV RNA.4 One reason for these discrepancies may
be that serum HCV RNA does not perfectly reflect the intra-
hepatic HCV replication,8,9 either because of extrahepatic
sources of HCV replication or because of an alteration of
serum preservation with time.10 Another important point is
that most studies on HCV replication after transplantation
are cross-sectional. For these reasons, it would be useful to
longitudinally assess the level of intrahepatic HCV RNA and
its relation with the long-term histological outcome of HCV-
related recurrent liver disease.
Our hypothesis is that a longitudinal and quantitative anal-
ysis of intrahepatic HCV replication after liver transplanta-
tion could show variations related to the histological course
and could enlighten the pathogenesis of recurrent hepatitis
C. A positive relation between intrahepatic HCV replication
and liver damage should suggest a direct pathogenicity of
HCV. Conversely, a decrease of intrahepatic HCV RNA to-
gether with a worsening of liver damage should suggest the
involvement of an host’s response, that could be immune-
mediated even in the context of transplantation. Our hypoth-
esis should also be tested in view of HCV genotype. Indeed,
a detrimental influence of genotype 1b has been reported
after liver transplantation in some centers5,11-13 but not in
all.14,15 The aim of this work was thus to determine the rela-
tions between the levels of intrahepatic HCV RNA, the long-
term histological outcome of recurrent hepatitis C and HCV
genotypes after liver transplantation.
PATIENTS AND METHODS
Patients. The records of 33 patients transplanted for HCV-related
cirrhosis between 1985 and 1993 were reviewed. These 33 patients
belonged to a previously published series of 60 patients in whom
a detrimental influence of genotype 1b was shown.5 They were
selected for the availability of at least two frozen liver biopsies
with well-conserved liver RNA, and the absence of any pattern of
rejection on these available biopsies. Patients with steroid resistant
rejection or chronic rejection or with less than two biopsies without
hepas WBS: Hepatology
1344 DI MARTINO ET AL.
signs of rejection, were thus excluded. From the 60 patients pre-
viously studied, 21 were excluded because not enough biopsies
were available, and 6 because of persistent rejection. Patients ex-
cluded were similar to patients included with respect to age, gender,
and genotype distribution.
All patients were HCV RNA-positive and hepatitis B surface anti-
gen -negative both before and after transplantation, and none had
detectable biliary obstruction or thrombosis of hepatic artery. Their
immunosuppressive regimen associated prednisolone, azathioprine
and ciclosporine, as previously reported.16 Genotype analysis (In-
noLipa assay) showed that 19 patients were infected by HCV geno-
type 1b, 9 by genotype 1a, 3 by genotype 2a, 1 by genotype 3c,
and 1 by an unclassified genotype. None of the patients received
ribavirine or interferon during the follow-up.
Samples. Percutaneous liver biopsies were performed because of
liver tests abnormalities (in 27 patients) or, in case of normal liver
tests, systematically every year (in 6 patients). A total of 37 serum
samples harvested at the time of liver biopsy in 19 patients were
available for quantitative analysis of HCV RNA (Fig. 1).
Methods
Histology. Histological evaluation of the liver was based on 5-
to 30-mm long pieces of liver harvested from percutaneous or
transjugular biopsy specimens using 0.8 to 1.4 gauge needles.
The specimens were immediately cut in two parts: one (1 to 6
mm) was frozen in liquid nitrogen and stored at 0807C, in a
distant location until 1992 and then in our laboratory; the other
was fixed in 10% formalin buffer and stained with hematoxylin/
eosin. The slides were coded by the same pathologist (MR) with-
out knowledge of genotype or of quantitation of serum or liver
HCV RNA.
Liver samples were distributed among the following four distinct
histological patterns: 1) normal histology; 2) lobular hepatitis (LH)
defined by hepatocytes necrosis and lobular inflammation without
piecemeal necrosis or fibrosis; 3) chronic active hepatitis (CAH)
defined by piecemeal necrosis and portal or interportal fibrosis; and
4) cholestasis defined by cholangitis with interlobular bile duct
neutrophil infiltration and no clear pattern of associated rejection
or hepatitis. CAH was scored for activity (A0 to A3) and fibrosis
AID Hepa 0013 / 5p28$$$241 10-13-97 20:58:29
HEPATOLOGY November 1997
FIG. 1. Main characteristics and
84 analyzed biopsies in 33 liver
transplantation recipients infected
by HCV. LH: lobular hepatitis; CAH:
chronic active hepatitis.
(F0 to F4) according to the French METAVIR coding system.17 To
ensure that histological liver damage was specifically related to
HCV, biopsy specimens with clear pattern of rejection were ex-
cluded.
Quantification of Serum HCV RNA. Amplification and quantification
of serum HCV RNA were performed using Monitor assay (Roche
diagnostics, Branchburg, NJ), according to the manufacturer’s in-
structions.
Quantification of liver HCV RNA by Competitive PCR. We used an
original home-made competitive polymerase chain reaction (PCR)
method: liver HCV-RNA and cellular RNA (28S ribosomal RNA)
were quantified simultaneously, to normalize the quantifications of
liver HCV RNA and to compare biopsy specimens stored for a large
range of time.
Competitive PCR is based on coamplification of the nucleic acid
to be measured with a defined amount of nucleic acid as internal
standard. When this internal standard is quasi-identical to the mea-
sured target (i.e., differs by only a short deletion or insertion, or
point mutations), the amplification yields of the target and internal
standard are similar18 permitting quantification in PCR run to satu-
ration. When applied to cellular RNA (e.g., liver) most investigators
using competitive PCR give results according to the weight of the
biopsy or the optical density of the total extracted RNA.8,19,20 The
quantity of b-actin messenger RNA can also be used,21 but its level
may be influenced by fibrosis. Our approach was based on parallel
quantitation of HCV RNA and ribosomal RNA (fraction 28S) in the
same competitive PCR. Indeed, 28S ribosomal RNA (28S rRNA) is
abundant and constant in all cells.22 The final amount of cellular
HCV RNA was given by the ratio between competitive PCR for
HCV and competitive PCR for the ribosomal RNA. This permitted
the level of HCV RNA to be normalized and to compare samples
despite the instability of RNA and the variability of extraction, re-
verse transcription, and amplification procedures. Finally, this nor-
malization permitted the use of a DNA competitor. Indeed, as the
conditions of reverse transcription were identical for both HCV and
ribosomal RNA, the use of versatile RNA competitors could be
avoided.
Competitors for HCV RNA and 28S rRNA were purified PCR
products (Fig. 2). For the construction of the HCV competitor,
hepas WBS: Hepatology
HEPATOLOGY Vol. 26, No. 5, 1997 DI MARTINO ET AL. 1345

FIG. 2. Schematic representation of the competitive PCR method used to quantify of HCV RNA and 28S rRNA. Quantitation of HCV RNA by
competitive PCR (left side): The (A) competitor and (B) target are by products of the 5* untranslated part of HCV RNA. Antisense primers used for the
reverse transcription have a different hydridization part (a or b) and a common sequence (c) that does not match to HCV RNA. For competitive PCR, a
known amount of (A) purified DNA competitor is coamplified with different dilutions of the (B) HCV cDNA to be measured. Competitive semi-nested
PCR is performed using a common antisense primer (c) and two sense primers (d and e). Quantitation of 28S rRNA by competitive PCR (right side):
The (A) competitor and (B) target are produced by reverse transcription using two different antisense primers. Both have the same hybridization part (a)
and a common 5* part (c) but antisense primer used for the construction of the competitor has a 11 mer-long insertion (i) between these two parts.

FIG. 3. Amounts of liver 28S rRNA and HCV RNA by competitive PCR in two samples (S1 and S2). Fivefold dilutions of 28S HCV cDNAs were
separately coamplified with known amounts of 28S and HCV DNA competitors. Measures of the area under the curve for each DNA peak through image
analysis software permit to calculate the ratio between amplified competitor and dilution of cDNA.

AID Hepa 0013 / 5p28$$0013 10-13-97 20:58:29 hepas WBS: Hepatology

1346 DI MARTINO ET AL.
a mixture of 30 different HCV RNA-positive sera (containing 20
genotype 1b sera), was used. After amplification and loading on
low-melting-point gel, the desired band was visualized and purified.
A 105 dilution of the eluted band was aliquoted and used as the
HCV internal standard in all experiments. Similar experiments with
cellular RNA were performed for the construction of the 28S rRNA
competitor.
Competitive PCR. Extraction of liver RNA from frozen liver graft
biopsies used the guanidium thiocyanate procedure.23 Reverse tran-
scription of HCV RNA and 28S ribosomal RNA (rRNA) was per-
formed in two distinct tubes with 5 mL of extracted RNA solution,
10 pmol of antisense primers, and Moloney murine leukemia virus
reverse transcriptase according to the manufacturer’s instructions.
Co-amplification of serial five-fold dilutions of HCV-complemen-
tary DNA (cDNA) and fixed amount of internal standard was per-
formed using semi-nested PCR. Briefly, two compartments were
created in the same tube by using 25 mL of high-density oil (dimeth-
ylpolysiloxane AK 350; Waker-Chemie GmbH, Munich, Germany)
permitting outer amplification first; then, after brief centrifugation,
inner amplification using the internal primer. This method, together
with routine use of the uracil N glycosylase method, avoids contami-
nation.
Measurement of PCR Products by Video Capture. PCR products were
electrophoresed on 3% agarose gel, then stained with ethidium
bromide (Fig. 3). Using video capture (The Imager, Appligène,
Strasbourg, France) then image analysis (NIH-Image, National Insti-
tutes of Health, Bethesda, MD), it was possible to measure the area
under the curve for each DNA peak (with a precision of 10%),
which is proportional to the quantity of amplicons.
Expression of Final Results. Final quantification was given by the
ratio between quantitation of HCV cDNA (in HCV competitor
unit) and quantitation of 28S cDNA (in 28S competitor unit).
Unit of quantification was thus arbitrary and called ‘‘competitor
unit.’’
Validation of the Quantification of Liver HCV RNA and 28S rRNA. We
first tested if saturation of PCR influenced the ratio between target
and competitor. A five-fold dilution of a mixture of HCV cDNA
(66%) and HCV competitor (33%) was coamplified. The mean ratio
between the two types of amplicons was 0.5, with a standard devia-
tion of { 9.8%. The same experiment was applied to our quantita-
tive method of 28S rRNA with similar results.
Statistical Analysis
The correlation between levels of liver HCV RNA and serum
HCV RNA load was studied using linear regression. A cross-sec-
tional study was performed by pooling all values of liver HCV-
RNA and using an independent t test and regression analysis to
compare levels of liver HCV RNA according to genotype, histology,
sex, age, dose of steroids, and time since transplantation. For
the longitudinal analysis, Student’s t test or Wilcoxon’s test and
regression analysis were used to compare serial levels of HCV RNA
and their relation to genotype, sex, age, dose of prednisolone,
histology, and time since transplantation. The relations between
the level of intrahepatic HCV RNA on the first available biopsy
and subsequent progression to chronic active hepatitis were deter-
mined by using the proportional hazard Cox’s model, taking the
date of the first biopsy as the starting date and the occurrence of
CAH as censoring event.
RESULTS
Histological follow-up lasted a mean of 3.5 years (range,
3 months-8.6 years) Twenty-nine of 33 patients developed
recurrent hepatitis and 4 patients remained with a normal
liver histology during the whole follow-up. The histological
course is summarized in Fig. 1. Twenty-seven patients devel-
oped histologically proven LH. Among them, 21 patients had
at a subsequent biopsy showing either normal histology (in
AID Hepa 0013 / 5p28$$$241 10-13-97 20:58:29
HEPATOLOGY November 1997
1 patient), cholestasis (in 1 patient), LH (in 2 patients), or
CAH in the 17 others. In 2 additional patients, CAH was
detected without previous detection of lobular hepatitis.
Therefore, CAH was diagnosed in 19 patients after a mean
follow-up of 43 months (8-6.2 years), of whom 4 progressed
to cirrhosis (data not shown). Among the 29 patients who
had recurrent hepatitis, 5 developed jaundice during the first
year, which resolved in 2 to 6 months and was associated
histologically with a neutrophil infiltration of bile ducts.
A total of 84 liver biopsy specimens were studied in the
33 patients. At least two biopsy specimens were available in
all patients. The mean time between transplantation and the
first and the last biopsies were respectively 15 months (range,
10 days-4.6 years) and 3.5 years (range, 3 months-8.6 years).
They were grouped according to histological features as fol-
lows: 15 normal in 8 patients (mean delay posttransplanta-
tion: 2.1 years, 23 days-8.6 years), 34 LH in 27 patients
(mean: 1.6 years, 27 days-5.4 years), 28 CAH in 19 patients
(mean: 3.2 years, 8 months-6.2 years), and 7 cholestasis in
5 patients (mean: 3 months, 1-8 months). Among the 29
patients with recurrent hepatitis, 17 patients were studied
both at the time of LH and CAH. For those patients, the
mean delay between the two biopsy prodedures was 2.1 years
(range, 2 months-5.1 years). In 17 patients, two successive
biopsies showed the same histological pattern (5 normal, 5
LH, and 7 CAH) with a mean interval of 1.25 year (range,
20 days-4 years).
Levels of Liver HCV RNA and Histological Course. All the 84
liver biopsies were HCV RNA positive, and quantitation
of liver HCV RNA could be determined in all 84 biopsies.
Mean levels of liver HCV RNA were significantly higher
in case of LH than in all other histological patterns (82 {
123 vs. 19 { 38; P õ .01). Mean levels of liver HCV RNA
were similar in CAH (22 { 49), cholestasis (19 { 15), and
normal histology (14 { 21) (Fig. 4A). When comparing
the first and last measurements of liver HCV RNA, levels
of liver HCV RNA were significantly higher in the first
biopsy than in the last, both in all patients (P õ .05) and
in the 29 who developed recurrent hepatitis (P õ .04)
(Fig. 4B and 4C). The fall in liver HCV RNA level was
more marked in the 17 patients in whom the levels of liver
HCV RNA could be compared at the time of LH and on a
subsequent biopsy showing CAH (P Å .006) (Fig. 5A).
There was a strong correlation (r Å .74, P õ .0002) be-
tween the two values of liver HCV RNA measured on two
successive biopsies showing the same histology (mean in-
terval between the two biopsies: 1.25 years; 1 month - 4
years). In this situation, the amount of liver HCV RNA
had no significant variation with time (Fig. 5B, 5C, and
5D). This showed that intrahepatic HCV replication was
stable in individuals remaining histologically stable.
Prognostic Value of Liver HCV RNA. To test the level of liver
HCV RNA as a predictor of CAH, we analyzed the actuarial
rate of CAH in 33 patients, determined from the date of the
first quantitation. There was no significant relation between
the time from transplantation and the level of liver HCV
RNA in the first biopsy (r Å .18, P Å .30). In contrast, there
was a strong correlation between the first quantitation and
the subsequent occurrence of CAH (P õ .01). Multivariate
analysis (Cox model; Table 1) identified three independent
factors positively related to the occurrence of CAH: 1) LH
on the first biopsy, 2) the level of liver HCV-RNA in the first
hepas WBS: Hepatology
HEPATOLOGY Vol. 26, No. 5, 1997 DI MARTINO ET AL. 1347

FIG. 4. Quantitation of liver

HCV RNA: relations with histology
and genotype. (A) Histology (84 bi-
opsies from 33 patients). LH: lobu-
lar hepatitis; CAH: chronic active
hepatitis. (B) Genotype (1b or not)
in the first and the last available bi-
opsies (n Å 33); and (C) genotypes
(1b or not) in the first biopsies with
lobular hepatitis (n Å 27) and in
the last biopsies with chronic active
hepatitis (n Å 19).

biopsy, and 3) young age. Sex, genotype 1b, and the time
between transplantation and the first biopsy were not inde-
pendent factors. No patient-related effect was observed.
Influence of Genotype 1b. In spite of a comparable duration
of follow-up, genotype 1b lead to more severe CAH as judged
on the activity score (genotype 1b Å 2.1 { 0.8; other geno-
types Å 1.2 { 0.8, P Å .01) and the fibrosis score (genotype
1b Å 2.9 { 1; other genotypes: 1.5 { 0.9, P õ .001), whereas
neither the activity score nor the fibrosis score was related
to levels of liver HCV RNA in the relevant biopsies. No
relation was found between genotype 1b and liver HCV RNA
levels measured in the first available biopsy (58 { 206 vs.
70 { 158; NS) or in the first biopsies showing LH (57 {
203 vs. 63 { 155; NS) (Fig. 4B and 4C). In contrast, levels
of liver HCV RNA were significantly lower at the end of
follow-up in case of genotype 1b both in the general group
(15 { 27 vs. 37 { 82; P Å .05) and in the 19 patients who
developed CAH (7 { 10 vs. 22 { 58; P Å .04) (Fig. 4B and
4C). These relations were not related to age, sex, or the
time between transplantation and analyzed biopsy through
multivariate analysis (data not shown).
Influence of Steroid Dose. Level of ciclosporinemia was
maintained within the same values for each patient and
showed poor intraindividual and interindividual variability
and could not be analyzed. Therefore the only available
marker of immunosuppression was the dose of prednisolone.
The dose of prednisolone (mg/kg) did not significantly influ-
ence the level of liver HCV RNA, significantly decreased with
time (r Å 0.64, P õ 1005) and was significantly lower at
the time of chronic active hepatitis (0.15 vs. 0.21, P Å .004).
The relations between liver HCV RNA, concomitant histolog-
ical findings, and HCV genotype previously described were
not modified when taking account of the dose of predniso-
lone in multivariate analysis.
Comparison of Concomitant Liver and Serum HCV RNA. All the
37 available serum samples harvested at the time of liver
biopsy were PCR positive and were quantified. For 12 of 19
concerned patients, serum was available both at the time of
LH and CAH. A correlation was found between liver HCV
RNA and serum HCV RNA (r Å .57, P Å .0002). However,
in 7 of 37 cases, apparent discrepancies were observed, that
concerned low values of serum HCV RNA (Fig. 6). Multiple
regression analysis showed that the level of serum HCV RNA
was independently correlated both to the level of liver HCV
RNA (P Å .0002) and to the duration of serum conservation
(P Å .05). A degradation of liver RNA with time was also
observed when comparing the quantitation of liver 28S rRNA
(normalized to the optically defined concentration of liver
RNA) according to the date of sampling. As the frozen liver
biopsies changed location in 1992, we compared 28S rRNA

FIG. 5. Variation of liver HCV RNA according to the histological course. (A) A significant decrease of liver HCV RNA was observed between lobular
hepatitis (LH) and subsequent chronic active hepatitis (CAH) in 17 patients (U test, P Å .006). Conversely, the level of liver HCV RNA had no significant
variation when comparing two consecutive biopsies showing the same histological pattern ([B]lobular hepatitis, [C] chronic active hepatitis, or [D] normal
histology) in 17 patients. As the time elapsing the two consecutive biopsies was similar in (A) the first case and in others, these results suggest that a
decrease of liver HCV RNA was not related to time, but was specifically associated with the occurrence of chronic active hepatitis.

AID Hepa 0013 / 5p28$$$241 10-13-97 20:58:29 hepas WBS: Hepatology

1348 DI MARTINO ET AL. HEPATOLOGY November 1997

TABLE 1. Multivariate Analysis of Factors Influencing the Occurrence of sponds to a pattern of lobular hepatitis without fibrosis nor
Chronic Active Hepatitis piece-meal necrosis and is characterized by a particularly
Cox Model b SE RR RR CI 95% P high level of intrahepatic viral replication. The second stage
is marked by progression to chronic active hepatitis and is
Sex 0.310 0.628 1.36 0.40-4.67 .62
characterized by a significant decline in intrahepatic HCV
Age 00.084 0.036 0.92 0.86-0.99 .02
Type 1b 00.111 0.488 0.89 0.34-2.33 .82
replication. High levels of intrahepatic replication suggest
Abnormal histology on that direct pathogenicity is the leading mechanism in the
first biopsy 3.241 1.197 25.56 2.45-266.77 .007 earlier stage of recurrent hepatitis. At the time of recurrent
Time between lobular hepatitis, intrahepatic HCV replication was not
transplantation and higher in genotype 1b, demonstrating that the particular
1st biopsy 00.000 0.001 1.00 0.99-1.01 .89 pathogenicity of this genotype in this setting5,11-13 is not
Level of liver HCV caused by a higher level of intrahepatic replication but rather
RNA on first biopsy 0.006 0.002 1.01 1.00-1.02 .01 to intrinsic properties. Despite progression to more severe
NOTE. Presence of histological lesions and high level of liver HCV hepatic lesions characterized by portal mononuclear infiltra-
RNA on the first biopsy, and young age, were independently related to tion and fibrosis, the levels of replication clearly decreased,
subsequent progression to chronic active hepatitis (SE Å standard error, suggesting that a host’s response (likely immune) was able
RR Å Relative Risk, CI Å Confidence Interval). n Å 33; x2 Å 22.04; df both to control intrahepatic HCV replication and to induce
Å 6; P Å .001. liver damage.
Particularly high levels of viremia have been described
during recurrent HCV infection.4,6,7 Posttransplantation vire-
mia was found to be a mean of 10-fold higher than pretrans-
quantification before and after this date. The mean quantifi- plantation viremia. These elevated viremia values have been
cation of 28S rRNA was significantly lower in samples har- reported during short-term follow-up (1-2 years posttrans-
vested before 1992 (1.12 { 1.36 vs. 1.89 { 2.25; P õ .03). plantation) and could correspond to the invasive phase of
recurrence and lobular hepatitis in most patients. However,
DISCUSSION relations between the level of HCV replication and histology
Recurrent hepatitis is frequent after liver transplantation are far from clear in liver transplantation. No relation be-
in anti-HCV positive patients. Persistent HCV infection is tween hepatitis and level of serum HCV RNA was re-
almost universal in this setting, and recurrent lobular or mild ported,4,14 while two successive series3,6 showed a strong rela-
chronic hepatitis occurs in more than 75% of patients during tion between viremia and concomitant recurrent hepatitis.
the 4 years following the transplantation.3 Half progress to However, none of these reports showed a long-term decline
CAH marked by fibrosis, piecemeal necrosis, and lympho- in viral replication. These discrepancies could be caused by
mononuclear infiltration of portal tracts in the 2 years follow- the degradation of serum HCV RNA with time.10 Indeed,
ing the onset of lobular hepatitis. In spite of numerous series serum harvested at the time of lobular hepatitis being more
in this setting, neither long-term longitudinal quantification ancient than those harvested at the time of chronic hepatitis,
of HCV replication nor quantification of liver HCV-RNA are decreasing of replication could be underestimated. Such phe-
available. This is the first quantitative study of intrahepatic nomenon was observed in this study as well in the serum as
replication of HCV in transplantation recipients. Using serial in the liver. Normalization of liver HCV RNA by the quantifi-
measurements of liver HCV RNA in a long-term period, we cation of 28S rRNA permitted to control this important bias.
show that, after liver transplantation, recurrent hepatitis C This is probably the main reason why our comparison be-
can be divided in two successive stages. The first stage corre- tween the level of liver HCV-RNA and concomitant viremia

FIG. 6. A significant relation

was observed between the level of
liver HCV RNA (assessed by com-
petitive PCR) and concomitant vire-
mia (measured by MONITOR assay)
in 37 paired samples (r Å .57; P Å
.0002). In 7 of 37 cases (19%) (s)
discrepancies between viremia and
liver HCV RNA were observed.

AID Hepa 0013 / 5p28$$$241 10-13-97 20:58:29 hepas WBS: Hepatology

HEPATOLOGY Vol. 26, No. 5, 1997
showed clear discrepancies in almost 20% of cases. Such
discrepancies are in accordance with the results of previous
studies comparing HCV-RNA levels in serum and liver in
immunocompetent subjects.8,20 Other theoretical reasons for
such discrepancies between measurements of liver and serum
HCV RNAs could be the influence of extrahepatic sources of
HCV replication, the variable amounts of immunoglobulin-
complexed HCV,24 or technical variations between meth-
ods.25
The most evident mechanism explaining the high levels
of replication observed after liver transplantation is immuno-
suppression. Indeed, it has been shown in this setting that
steroid boluses significantly enhance HCV replication.6 Our
results indicate that the dose of prednisolone were not corre-
lated to the level of liver HCV RNA, but we must underline
that patients with history of persistent or steroid-resistant
rejection were excluded from this study. Another theoretical
mechanism is the allogenic context of liver transplantation,
which impairs the antiviral immune response mounted be-
fore transplantation. A drastic reduction in this preexisting
immunity should favor HCV replication and, thus, the ex-
pression of the intrinsic cytopathic effects of viral proteins.
Although relations between the severity of recurrent hepatitis
and human leukocyte antigen matching have been sug-
gested,26 this was not borne out by our preliminary data or
those of a large series from the UK.12 Whatever the role
of histocompatibility, we observed a long-term decline in
intrahepatic HCV replication in spite of progression to CAH,
at the same time of a decline in the amount of immunode-
pression. This strongly suggests that effective immunity is
acquired against liver epitopes of HCV along the years. The
decline of intrahepatic HCV replication was particularly
marked in patients infected by genotype 1b. This genotype
could induce a stronger immune response than other geno-
types, explaining that genotype 1b leads to more severe CAH
both assessed by higher scores of fibrosis and activity. On
the other hand, we observed a strong correlation between
the first determination of intrahepatic HCV replication and
the risk of subsequent progression to CAH. These data are
in keeping with those of a previous study showing that the
early posttransplantation level of HCV viremia was related to
subsequent progression to chronic active hepatitis.27 Taken
together, these data support the notion that the level of liver
HCV RNA at the time of lobular hepatitis is related to the
risk of CAH, whereas genotype 1b is related with more severe
CAH. Finally, our results give a rational justification for early
eradication of HCV that should stop this process and have
beneficial effect in the long-term. Combination of interferon
and ribavirin seems to be both efficient and safe after liver
transplantation.28 Another potentially interesting approach
should be the neutralization of HCV through hyper-immune
immunoglobulins. However such immunoglobulin prepara-
tions are not yet available.
Acknowledgment: We thank Pierre Rufat for his helpful
critical review of the statistical analysis, and Produits ROCHE
(Neuilly, France) for providing the kits for Monitor Assay.
REFERENCES
1. Di Bisceglie A. Liver transplantation for hepatitis C: the promise and
the challenge. HEPATOLOGY 1995;22:660-661.
2. Wright TL, Donegan E, Hsu H, Ferrel L, Lake J, Kim M, Combs C, et
AID Hepa 0013 / 5p28$$$241 10-13-97 20:58:29
DI MARTINO ET AL. 1349
al. Recurrent and acquired hepatitis C viral infection in liver transplant
recipients. Gastroenterology 1992;103:317-322.
3. Féray C, Gigou M, Samuel D, Paradis V, David M, Reynes M, Bréchot
C, et al. The course of hepatitis C infection after liver transplantation.
HEPATOLOGY 1994;20:1137-1143.
4. Chazouillères O, Kim M, Ferrel L, Robert J, Ascher N, Wilber J, Neuwald
P, et al. Quantitation of hepatitis C virus RNA in liver transplant recipi-
ents. Gastroenterology 1994;106:994-999.
5. Féray C, Gigou M, Samuel D, Mishiro S, Reynes M, Bismuth H,
Bréchot C. Influence of the genotypes on the severity of recurrent
liver disease after liver transplantation. Gastroenterology 1995; 108:
1089-1097.
6. Gane E, Naoumov N, Qian K, Mondelli M, Maertens G, Portmann
B, Lau JYN, et al. A longitudinal analysis of hepatitis C virus replica-
tion following liver transplantation. Gastroenterology 1996; 110:
167-177.
7. Duvoux C, Pawlotsky JM, Cherqui D, Tran Van Nhieu J, Métreau JM,
Fagniez PL, Duval J, et al. Serial quantitative determination of hepatitis
C virus RNA levels after liver transplantation. A useful test for diagnosis
of hepatitis C virus reinfection. Transplantation 1995;60:457-461.
8. Sakamoto N, Enomoto N, Kurosaki M, Marumo F, Sato C. Detection
and quantification of hepatitis C virus RNA replication in the liver. J
Hepatol 1994;20:593-597.
9. Cabot B, Esteban JI, Martell M, Genesca J, Vargas V, Esteban R, Guardia
J, et al. Structure of replicating hepatitis C virus (HCV) quasispecies in
the liver may not be reflected by analysis of circulating HCV virions. J
Virol 1997;71:1732-1734.
10. Halfon P, Khiri H, Gerolami V, Bourliere M, Feryn JM, Reynier P,
Gauthier A, et al. Impact of various handling and storage conditions
on quantitative detection of hepatitis C virus RNA. J Hepatol 1996;25:
307-311.
11. Gordon ED, Zein NN, Poterucha JJ, Gross JB, Gossard AA, Steers JL,
Wiesner RH, et al. Hepatitis C genotype 1b in liver transplant recipients
is associated with an increased incidence and severity of recurrence
[Abstract]. Gastroenterology 1995;108:A 1074.
12. Gane EJ, Bernard CB, Portmann C, Naoumov NV, Smith HM, Underhill
JA, Donaldson PT, et al. Long-term outcome of hepatitis C infection
after liver transplantation. N Engl J Med 1996;334:815-820.
13. Alberti AB, Belli LS, Silini E, Bellati G, Asti M, d’Amico M, Rondinara
GF, et al. HCV genotypes and severe hepatitis C recurrence after liver
transplantation: report of 103 cases [Abstract]. HEPATOLOGY 1996;24:
297A.
14. Zhou S, Terrault NA, Ferrell L, Hahn JA, Lau JYN, Simmonds P, Roberts
JP, et al. Severity of liver disease in liver transplantation recipients with
hepatitis C virus infection: relationship to genotype and level of viremia.
HEPATOLOGY 1996;24:1041-1046.
15. Boker KHW, Dalley G, Bahr MJ, Maschek H, Tillman HL, Trautwein
C, Oldhaver K, et al. Long-term outcome of hepatitis C virus infection
after liver transplantation. HEPATOLOGY 1997;25:203-210.
16. Gugenheim J, Samuel D, Saliba F, Castaing D, Bismuth H. Use of flexible
triple-drug immunosuppressive therapy in liver transplantation. Trans-
plant Proc 1987;19:3805-3807.
17. The French METAVIR Cooperative Study Group. Intraobserver and
interobserver variations in liver biopsy interpretation in patients with
chronic hepatitis C. HEPATOLOGY 1994;20:15-20.
18. Pannetier C, Delassus S, Darche S, Saucier C, Kourilsky P. Quantitative
titration of nucleic acids by enzymatic amplification reactions run to
saturation. Nucl Acid Res 1993;21:577-583.
19. Hagiwara H, Hayagashi N, Mita E, Takehara T, Kasahara A, Fusa-
moto H, Kamada T. Quantitative analysis of hepatitis C virus RNA
in serum during interferon alfa therapy. Gastroenterology 1993;
104:877-883.
20. Mc Guinness PH, Bishop A, Painter DM, Chan R, McCaughan GW.
Intrahepatic hepatitis C RNA levels do not correlate with degree of liver
injury in patients with chronic hepatitis C. HEPATOLOGY 1996;23:676-
687.
21. Grassi G, Pozzato G, Moretti M, Giacca M. Quantitative analysis of
hepatitis C virus RNA in liver biopsies by competitive reverse transcri-
tion and polymerase chain reaction. J Hepatol 1995;23:403-411.
22. Barbu V, Dautry F. Northern blot normalization with a 28S rRNA oligo-
nucleotide probe. Nucleic Acid Res 1989;17:7115.
23. Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid
guadinium thiocyanate-phenol-chloroform extraction. Anal Biochem
1987;162:156-159.
24. Shimizu Y, Hijikata M, Iwamoto A. Neutralizing antibodies against hep-
hepas WBS: Hepatology
1350 DI MARTINO ET AL.
atitis C virus and the emergence of neutralisation escape mutant viruses.
J Virol 1994;68:1494-1500.
25. Collins ML, Zayati C, Detmer JJ, Daly B, Kolberg JA, Cha TA, Irvine
BD, et al. Preparation and characterization of RNA standards for use in
quantitative branched DNA hybridization assays. Annal Biochem 1995;
226:120-129.
26. Gretch DR, Wile M, Gaur L, Lesniewski R, Nelson K, Strong M,
Johnson R. Donor-recipient match at the HLA-DQB locus is associ-
ated with recrudescence of chronic hepatitis following liver trans-
AID Hepa 0013 / 5p28$$$241 10-13-97 20:58:29
HEPATOLOGY November 1997
plantation for end stage hepatitis C [Abstract]. H EPATOLOGY 1993;
18:108A.
27. Gretch DR, Bacchi CE, Corey L, DelaRosa C, Lesniewski RR, Kowdley
K, Gown A, et al. Persistent hepatitis C virus infection after liver
transplantation: clinical and virological features. HEPATOLOGY 1995;
22:1-9.
28. Bizollon T, Ducerf C, Trepo C. New approaches to the treatment of
hepatitis C virus infection after liver transplantation using ribavirin. J
Hepatol 1995;23(Suppl):22-25.
hepas WBS: Hepatology