Results

LONGITUDINAL VARIATION IN HCV RNA LEVEL AND TYPE OF IMMUNOSUPPRESSION

The median pretransplant HCV RNA level was 0.4 (<0.2–5.5) × 10⁶ Eq/ml, at one week post-transplant 0.5 (<0.2–11.4), at two weeks 1.35 (<0.2–26.0), at three months 7.6 (<0.2–210.0), and at 12 months 8.7 (0–2000.0) × 10⁶ Eq/ml. Only one serum sample at 12 months post-transplant was found to be negative by both bDNA and PCR, while all other bDNA negative samples were positive by PCR. Mean preservation time was 1307 (40) days (range 10–2555) and was not correlated with HCV RNA levels (r=−0.03, p=0.69).

The pretransplant HCV RNA level was not related to the post-transplant HCV RNA level at any time point. Serum HCV RNA at one or two weeks post-transplant was not related to the type of initial immunosuppression (fig 1) or to any other variable.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Serum HCV RNA levels pretransplant and at one week (in 35 patients), two weeks (in 38 patients), and three and 12 months post-transplant in 47 patients transplanted for HCV cirrhosis in relation to the type of initial immunosuppressive therapy. Initial immunosuppressive therapy was triple (cyclosporin, azathioprine, prednisolone) in 31, double (cyclosporin and prednisolone) in five, and single (eight cyclosporin; three tacrolimus) in 11 patients. Box and whisker plots express medians, and interquartile and overall ranges. The outlying values are plotted individually. HCV RNA levels are presented in a logarithmic scale.

Levels of HCV RNA at three months were significantly associated only with the type of initial immunosuppression. In particular, three month HCV RNA levels were higher in the patients with single (median, 18.0 × 10⁶ Eq/ml) than in those with triple (median, 4.7 × 10⁶ Eq/ml; p=0.002) or double initial immunosuppressive therapy (median, 4.4 × 10⁶ Eq/ml; p=0.054) (table 2; fig1). In the single therapy group, three month HCV RNA levels were relatively higher in the patients treated with cyclosporin (median, 43.5 × 10⁶ Eq/ml; range, 8.8–210.0) than in those treated with tacrolimus (median, 10.6 × 10⁶ Eq/ml; range, 8.7–25.0), but the difference was not statistically significant (p=0.15). Average doses and levels of cyclosporin were similar in the three groups at three and 12 months after transplant (data not shown). Tacrolimus was only used in the monotherapy group.

View this table:

• View inline
• View popup

Table 2

Serum HCV RNA levels at three and 12 months after transplantation for HCV cirrhosis in relation to type of initial immunosuppressive therapy

HCV RNA levels decreased between three and 12 months post-transplant in all patients with initial single immunosuppressive therapy (12 month median, 4.9 × 10⁶ Eq/ml; p=0.003 compared with three month values). In contrast, it increased in 19 (61%) and decreased in only 11 (35%) patients with triple therapy (12 month median, 10.5 × 10⁶ Eq/ml; p=0.08 compared with three month values) and did not change significantly (increased in two and decreased in three) in the patients with initial double therapy (12 month median, 5.0 × 10⁶ Eq/ml; p=0.89 compared with three month values) (table2; fig 1).

HCV RNA levels at 12 months were significantly correlated with previous duration of prednisolone treatment (r=0.37, p=0.01). They were relatively higher in patients with triple therapy compared with those with single initial immunosuppressive regimen (p=0.058) (table 2; fig 1). There was no significant difference in the 12 month HCV RNA levels between patients treated with cyclosporin (median, 7.4 × 10⁶ Eq/ml) and tacrolimus (median, 2.7 × 10⁶ Eq/ml) single therapy (p=0.41).

HCV genotype was not associated with serum HCV RNA levels at any time point (data not shown).

HCV RNA LEVELS AND ACUTE HEPATITIS C

Acute hepatitis C was diagnosed in 23 (49%) of the 47 patients at a median of 3 (0.5–11) months. In particular, it was diagnosed in 39% of the patients with triple, 80% of those with double, and 64% of those with single initial immunosuppressive therapy. No patient developed severe cholestatic or histologically severe acute hepatitis; seven patients had moderate and 16 mild acute hepatitis. In patients with initial triple therapy, acute hepatitis C was diagnosed in five (55%) of the nine with azathioprine discontinuation (in all cases within one to two months after azathioprine withdrawal) and in seven (32%) of the remaining patients. Acute hepatitis was diagnosed at a median of 3.0 (1.0–11.0) months in patients with triple therapy, 3.8 (1.5–7.0) months in those with double therapy, and 1.5 (0.5–4.5) months in those with single therapy as initial immunosuppression (p=0.10 for comparison between patients with single and patients with double or triple therapy) (table 1).

Serum HCV RNA levels at three months post-transplant were significantly higher in the 23 patients with acute hepatitis C (median, 17.8 × 10⁶ Eq/ml; range, <0.2–98.0) compared with the 24 patients without (median, 5.0 × 10⁶ Eq/ml; range, <0.2–210.0; p=0.03). In contrast, HCV RNA levels at 12 months did not differ between the two groups (fig 2). In the 23 patients with acute hepatitis C, timing of acute hepatitis was not correlated with level of viraemia at three months (r=0.05, p=0.83) or 12 months (r=0.19, p=0.38). In particular, HCV RNA levels at three months were similar in the 11 patients with acute hepatitis before three months (median, 17.7 × 10⁶Eq/ml; range, <0.2–90.0) and in the 12 patients with acute hepatitis after three months (median, 18.0 × 10⁶ Eq/ml; range, 1.2–98.0; p=0.93).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Serum HCV RNA levels pretransplant and at one week (in 35 patients), two weeks (in 38 patients), and three and 12 months post-transplant in 47 patients transplanted for HCV cirrhosis in relation to the occurrence of acute hepatitis C. Box and whisker plots express medians, and interquartile and overall ranges. The outlying values are plotted individually. HCV RNA levels are presented in a logarithmic scale.

Multivariate analysis showed that three month HCV RNA levels were significantly associated only with the type of initial immunosuppressive regimen (β (SE): 0.30 (0.12), p=0.017) and not with occurrence of acute hepatitis, while 12 month HCV RNA levels were significantly associated only with previous duration of steroid treatment (β (SE): 0.14 (0.05), p=0.004) and not with the type of initial immunosuppression.

HISTOLOGICAL LESIONS AT 12 MONTHS AFTER TRANSPLANT

At 12 months post-transplant, necroinflammatory activity was absent or minimal in 16 (38%) and mild or moderate in 26 (62%) (grading score: 4 in six, 5 in 12, 6–8 in seven, and 13 in one) of the 42 patients who had a liver biopsy. Fibrosis was present (fibrosis score >0) in all but one patient, but no patient had fully established cirrhosis (fibrosis score = 6). Fibrosis score was ⩽2 in 32 (76%) and ⩾3 in 10 (24%) (3 in seven and 4–5 in three) patients. Chronic rejection was not diagnosed in any of the 47 patients.

Serum HCV RNA levels at any time point were not associated with necroinflammatory activity at 12 months (fig 3). No association was also found between HCV RNA levels and piecemeal necrosis, confluent necrosis, focal lytic necrosis, apoptosis, and focal inflammation, or portal inflammation score (data not shown). In contrast, 12 month HCV RNA was significantly correlated with the fibrosis score (r=0.56, p<0.001); the fibrosis score was not associated with HCV RNA at any other time point (fig4).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Serum HCV RNA levels pretransplant and at one week (in 35 patients), two weeks (in 38 patients), and three and 12 months post-transplant in 47 patients transplanted for HCV cirrhosis in relation to chronic hepatitis grading score. Box and whisker plots express medians, and interquartile and overall ranges. The outlying values are plotted individually. HCV RNA levels are presented in a logarithmic scale.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

Serum HCV RNA levels pretransplant and at one week (in 35 patients), two weeks (in 38 patients), and three and 12 months post-transplant in 47 patients transplanted for HCV cirrhosis in relation to fibrosis or chronic hepatitis staging score. Box and whisker plots express medians, and interquartile and overall ranges. The outlying values are plotted individually. HCV RNA levels are presented in a logarithmic scale.

High necroinflammatory activity was significantly associated with the occurrence of post-transplant acute hepatitis C. Chronic hepatitis grading score was higher in 20 patients with previous acute hepatitis C (5.1 (2.6)) than in 22 patients without (3.6 (2.0); p=0.038) and this was mainly due to significantly higher score for focal lytic necrosis, apoptosis, and focal inflammation and in part due to higher score for piecemeal necrosis (table 3). There was no difference in necroinflammatory activity between patients with mild (5.0 (2.1)) and patients with moderate (5.1 (3.6)) previous acute hepatitis C (p=0.44).

View this table:

• View inline
• View popup

Table 3

Histological findings at 12 months post-transplant in 42 patients transplanted for HCV cirrhosis in relation to previous acute hepatitis C

The extent of fibrosis at 12 months was also significantly associated with the occurrence of acute hepatitis C. Fibrosis was more severe in patients with previous acute hepatitis C (table 3) and a fibrosis score ⩾3 was found in 40% (8/20) of patients with and in only 9% (2/22) of patients without previous acute hepatitis C (p=0.03). There was no difference in the extent of fibrosis at 12 months between patients with mild (2.4 (1.0)) and patients with moderate (2.3 (1.5)) previous acute hepatitis C (p=0.65). HCV genotype was not found to be associated with chronic hepatitis grading or staging score.

Multiple regression analysis showed that the fibrosis score was significantly associated with the levels of 12 month HCV RNA (β (SE): 0.52 (0.16), p=0.002) and relatively with previous acute hepatitis (β (SE): 0.53 (0.27), p=0.056).

Discussion

Our study has the largest number of patients evaluated by repeated measurement of viraemia levels at certain time points post-transplant and the advantage of a blinded evaluation of liver histology based on an objective classification system. Although our patients were not randomised to the different immunosuppressive regimens, the type of treatment was not selected on the basis of patient characteristics but it changed according to study protocols. There was no significant difference in the pretransplant patient or viral characteristics among the three treatment groups. We included a relatively small number of patients treated with double or single therapy, but our study is the first which included patients treated immediately after transplant with only a single immunosuppressive drug.

HCV recurrence after transplantation, defined as detectable serum HCV RNA by PCR, was observed in 100% of the patients in this study. Longitudinal variation of serum HCV RNA levels was similar overall to that reported by others in liver transplant patients.5 ,6However, we found that patients initially treated with cyclosporin or tacrolimus monotherapy had significantly higher HCV RNA levels at three months post-transplant than the patients treated with combinations of immunosuppressive drugs. The explanation for this finding is not clear. The number of intravenous courses of methylprednisolone as antirejection treatment must not have had an impact on this difference in the three month viraemia levels, as the number of rejection episodes and methylprednisolone courses were similar among the patients treated with different immunosuppressive regimens. Occurrence of acute hepatitis C may have been a confounding factor for the difference of the three month viraemia levels between patients treated with different immunosuppressive regimens. Although we tried to adjust the HCV RNA levels for occurrence and timing of acute hepatitis C, occurrence of acute hepatitis C may have been underestimated. It has been suggested that the type of initial immunosuppression and courses for rejection may be associated with the timing and severity of HCV recurrence.30 Another possibility could be that liver injury during the initial phase of HCV infection may be more severe in patients under less immunosuppression (or in patients treated without steroids) and that serum HCV RNA levels could reflect increased release of HCV RNA from destroyed hepatocytes. Liver HCV RNA levels have been shown to be particularly high during the acute hepatitis phase in HCV transplant patients and to decline thereafter, suggesting that a host’s response may be involved in the pathogenesis of HCV infection.31 Earlier and more severe acute hepatitis in patients under less immunosuppression would further support the hypothesis of immunologically mediated liver injury during the acute hepatitis phase. A more severe acute hepatitis phase in HCV transplant patients with less depressed immune systems may be associated with a more favourable long term course of HCV infection.15 In immunocompetent patients, an efficient T helper lymphocyte response during the acute phase of HCV infection has been suggested as a critical determinant of the course of the disease.32 ,33

HCV RNA levels at 12 months post-transplant were found to be significantly associated with previous duration of prednisolone treatment. The prednisolone start daily dose (1 mg/kg or 50–75 mg) was relatively high in our study, as some centres, including ours, now start with 20 mg of prednisolone daily. We do not know whether a lower start dose of prednisolone would have had a different effect on the viraemia levels. Steroids have been shown to increase HCV viraemia levels in other studies in transplant5 as well as in non-transplant patients.16 However, direct stimulation of viral replication through a steroid responsive element on the viral genome such as that found in hepatitis B virus infection34seems unlikely. HCV RNA levels did not differ at one or two weeks between patients treated with or without steroids and, at three months, they were significantly higher in patients treated without steroids. The association of HCV RNA levels at 12 months, when only one patient remained on steroids, and duration of previous steroid treatment also favour the involvement of immune mediated mechanisms in the pathogenesis of HCV related liver injury. Withdrawal of steroids may be associated with an immunological attack and increased lysis of infected hepatocytes with subsequent release of a significant number of virions into the circulation.

At 12 months post-transplant, necroinflammatory activity of chronic hepatitis was absent or minimal (grading score 0–3) in 38% and mild (grading score 4–8) in 60%, but fibrosis was already present in all but one patient and it was significant (staging score ⩾3) in 24%.

Necroinflammatory activity at 12 months was not associated with concentrations of serum HCV RNA at any time point in this study. Thus, we did not confirm the finding from a smaller study (18 patients) that serum HCV RNA levels within the first two weeks post-transplant are associated with the activity of disease at 12 months.6 The presence of similar viraemia levels in our patients with or without histologically active hepatitis further suggests involvement of immune mechanisms and is against a direct cytopathic effect in the pathogenesis of HCV related liver injury.

Higher necroinflammatory activity was strongly associated with an occurrence of previous post-transplant acute hepatitis C. A relation of previous acute lobular hepatitis and severity of HCV infection has also been described by others,9 although different definitions of disease severity have been used. Although the type of initial immunosuppression was not associated with disease activity at 12 months after transplant, its effect on disease activity later on cannot be evaluated in the present study. Recently it has been shown that the number of methylprednisolone courses as antirejection treatment within the first year post-transplant was significantly associated with development of chronic active hepatitis at two years despite no association at one year post-transplant.18

We found a strong correlation between fibrosis and HCV RNA levels at 12 months post-transplant. Gane et al have also reported an association between the viraemia level and histological severity of HCV recurrence, although they used a different classification of histological lesions and did not score for fibrosis; there was also a variable follow up ranging from six to 69 months.5 Data from immunocompetent patients suggest that progression of fibrosis is the most important predictor of the outcome of chronic hepatitis C.14 If this is also the case in the post-transplant setting, then the association between the viraemia level and the extent of early fibrosis may be extremely important and higher 12 month HCV RNA levels may be associated with worse long term outcome of post-transplant HCV infection. Furthermore, the extent of fibrosis at 12 months post-transplant was also associated with the occurrence of previous acute hepatitis. More severe hepatocyte injury during the acute hepatitis phase in the patients with, compared with those without a diagnosis of post-transplant acute hepatitis C, might induce fibrogenesis. In the patients with a diagnosis of acute hepatitis C, however, we did not find an association between the severity of acute hepatitis and the extent of fibrosis at 12 months. It was recently reported that a more severe initial HCV recurrence may be associated with subsequent development of cirrhosis.35 ,36Although the results of the latter studies seem to be in contrast with some of our data, different definitions of initial HCV recurrence and longer follow up times may be responsible for the seemingly discrepant findings.

The role of HCV genotype in the course of post-transplant HCV infection has been controversial. Genotype 1b or 1 has been associated with worse disease progression in some studies,3 ,7 ,37 ,38 but no association between HCV genotype and course of HCV infection has been reported by others.39 Similarly, we did not find any association between HCV genotype and activity or stage of disease at 12 months post-transplant. However, a relation between HCV genotype and histological severity later on cannot be excluded. HCV genotype was also not found to be related to serum HCV RNA levels.

In conclusion, we found that HCV RNA levels at three months after transplant are higher in patients treated with single initial immunosuppressive therapy but HCV RNA levels at 12 months are higher in patients on steroids for longer. There was also a trend for an association of higher HCV RNA levels at 12 months with triple compared with single initial immunosuppressive therapy (p=0.058). HCV viraemia at 12 months seems to be particularly important, as its levels are strongly correlated with the extent of fibrosis. If our findings are confirmed by other studies and particularly if an association between the type of immunosuppressive therapy and the long term histological outcome is established, then initial immunosuppression in patients transplanted for HCV cirrhosis may be modified towards less intense regimens. Furthermore, if an association between the occurrence of acute hepatitis C and the histological outcome of post-transplant HCV infection is confirmed, then early antiviral treatment at the time of acute hepatitis may be beneficial. Studies with longer follow up are certainly needed to evaluate whether different types of initial immunosuppressive regimens have a different effect on the histological outcome of post-transplant HCV infection.