Rejection Under Alpha Interferon Therapy in Liver Transplant Recipients


*Corresponding author: Jérôme Dumortier,


Interferon alpha (IFN) is the corner stone drug for the treatment of recurrent hepatitis C (HCV) in liver transplant (LT) recipients. One of its serious potential adverse effects is acute and chronic rejection. The aim of this study was to review our experience using IFN-based therapy, in order to examine the incidence and the risk factors for rejection, and the outcome of patients who developed rejection.

Between September 1990 and December 2004, 70 LT recipients were treated. Patients started antiviral treatment 16 (1–137) months after LT. Histological follow-up was available in all patients according to protocol biopsies. Rejection was diagnosed and graded according to Banff classification.

Twenty-one percent of patients developed acute rejection (5 mild, 9 moderate and 1 severe) during IFN-based therapy. Patients were treated for 8 (1–15) months prior to rejection. Previous history of acute rejection before IFN therapy and treatment with pegylated-IFN was significantly associated with rejection (p = 0.04 and p = 0.02, respectively). The rejection was successfully treated in 87% of patients. No chronic rejection or graft losses were observed.

Acute rejection under IFN-based therapy often occurs in LT recipients, but early diagnosis with protocol biopsies and early treatment can lead to a favorable outcome.


Liver cirrhosis due to hepatitis C virus (HCV) is one of the main currently accepted indications for liver transplantation (LT) worldwide (1–3). However, recurrence of HCV infection is almost universal and histological recurrence is observed in 80% of HCV-infected grafts within 5 years after LT (2,4). After 5 years, up to 30% of the patients transplanted with HCV infection might develop cirrhosis (5,6); moreover, long-term graft and patient survivals are significantly reduced in patients undergoing LT for HCV-related cirrhosis compared with other groups (7,8). At the present time, combination antiviral therapy using alpha interferon (IFN) and ribavirin has been shown to be useful for viral clearance and improvement of histological liver damage (9–11). However, after LT, the use of antiviral therapy has been limited by poor tolerability and adverse effects. One of the serious reported complications of IFN after LT is the precipitation of acute or chronic rejection and subsequent graft loss (12,13). However, the risk and severity of rejection due to IFN in liver transplant recipients are still controversial (10,11,14–29).

In the present study, we reviewed our experience using IFN-based therapy with long histological follow-up according to protocol biopsies, in order to examine the incidence and the risk factors for rejection, and the outcome of patients who developed rejection.

Patients and methods


From September 1990 to December 2004, LT was performed in 635 adult patients in our center; 127 patients (20%) had anti-HCV antibodies, confirmed using a third-generation recombinant immunoblot assay (RIBA). The median age of the recipients was 51 (13–68) years. Eighty-eight of the 127 patients (69%) were males. Sera from 111 patients were screened for HCV genotypes using the Inno-lipa technique (Ingen); 81 of 111 (73%) had genotype 1 HCV. Fifty-seven (45%) patients received IFN-based therapy before LT. All patients received blood group compatible grafts from deceased or living donor. Thirteen (10%) patients received living donor grafts, while the rest received deceased donor grafts. The median age of the donors was 33 (14–63) years. Forty-eight (38%) patients had an hepatocellular carcinoma on the explanted liver. Follow-up was conducted by measuring serum alanine aminotransferase (ALT) levels, serum quantitation of HCV-RNA by signal amplification procedure (Versant-HCV b-DNA 3.0-UI, Bayer Diagnostics, Puteaux, France), and by performing liver biopsies (every 6 months when ALT levels were within the normal ranges, every 3 months when ALT levels were higher than the upper normal limit).

Immunosuppressive regimen

Initial immunosuppressive regimen was based on a calcineurin-inhibitor: cyclosporine or tacrolimus. Patients initially receiving cyclosporine (Novartis Pharma; Rueil-Malmaison, France) were treated with a dose of 2.0–4.0 mg/kg/day orally in two divided doses with target trough whole blood concentrations of 200–250 ng/mL for the first month posttransplant, followed by target trough blood concentration of 100–200 ng/mL thereafter. Patients initially receiving tacrolimus (Fujisawa, Nanterre, France) were treated with a dose of 0.08–0.12 mg/kg/day orally in two divided doses, with target trough whole blood concentrations of 10–15 ng/mL for the first month posttransplant, followed by 5–10 ng/mL thereafter. In addition, patients received 500 mg IV methylprednisolone after reperfusion. Starting on postoperative day 1, methylprednisolone was tapered from 200 mg to 20 mg within 5 days; thereafter, methylprednisolone was maintained at 20 mg/day and then tapered by 2.5 mg/month to a maintenance dose of 0–5 mg/day after 6 months posttransplantation. Azathioprine or mycophenolate mofetil was either administrated as a part of initial triple immunosuppressive regimen, or introduced during follow-up as maintenance immunosuppressive agent. Long-term follow-up was conducted every month by using blood concentrations of tacrolimus or cyclosporine.

Histological study

Percutaneous liver biopsies (LBs) were performed using 18-gauge needles. A total of 765 LBs were performed in these 70 patients (mean of 10.9 LBs per patient). No mortality occurred after liver biopsy; three (0.39%) cases of severe complications were observed (one pneumothorax and two bleeding episodes).

All the available tissue material was reviewed independently by two observers. Only biopsy samples with a number of portal tracts >10 and/or a size >15 mm were retained for histopathological examination and diagnosis (74% of liver biopsies performed). Sections stained with hematoxylin-eosin-saffron, chromotrope and Perls reaction were available in all cases; in some cases, immunodetection of biliary markers, such as cytokeratins 7 and 19, has been performed. The following data were recorded for each biopsy: (a) length of the sample and number of portal tracts, (b) liver architecture, (c) presence and number of intrahepatic bile ducts; evidence, severity and distribution of epithelial biliary lesions; presence and intensity of ductular reaction; (d) evidence, composition, distribution and density of inflammatory infiltrates; (e) evidence, pattern and distribution of hepatocellular necrosis; (f) evidence, severity and topography of steatosis; (g) evidence and distribution of vascular lesions; (h) evidence, pattern and severity of fibrosis; (i) other features.

Piecemeal necrosis was defined as the eosinophilic necrosis of isolated hepatocytes located in the immediate periportal area of the hepatic lobule and surrounded by inflammatory cells coming from the adjacent portal tract. Endothelialitis was defined as the presence of inflammatory endothelial lesions of portal and/or centrilobular veins, such as the detachment of the endothelial lining from the basement membrane by inflammatory cells or the focal destruction of the endothelial lining by inflammatory cells; adhesion of inflammatory cells to the luminal or abluminal surfaces of endothelial cells, but without evidence of cell detachment or other endothelial cell lesions, was not considered sufficient for a diagnosis of endothelialitis. Ductopenia was defined as previously, as the absence of visible bile duct in at least 50% of the portal tracts present in the tissue sample; immunodetection of biliary cytokeratins was performed in all cases to confirm the diagnosis.

The histological diagnosis of chronic hepatitis due to graft reinfection by HCV was based on the combination of the following intraportal and intralobular lesions: (a) evidence of hepatocellular necrosis (piecemeal necrosis, intralobular eosinophilic necrosis or apoptotic bodies), (b) chronic inflammatory infiltrate (including intraportal inflammatory infiltrates predominantly made of lymphocytes, sometimes associated with lymphoid nodules and intralobular inflammatory cells, including lymphocytes, histiocytes and polymorphonuclear neutrophils) and (c) portal and periportal fibrosis. In addition, epithelial biliary lesions had to be focal and mild. No evidence of endothelialitis had to be found. Mild to moderate steatosis may be present. The histological staging and grading of chronic HCV-related graft hepatitis was performed according to the METAVIR scoring system (30).

The histological diagnosis of acute rejection was based on the following combination of lesions: (a) portal inflammatory infiltrate containing eosinophils, (b) presence of biliary epithelial lesions and endothelialitis, (c) absence of prominent piecemeal necrosis, (d) possible presence of centrilobular lesions, including central vein endothelialitis and pericentrilobular inflammatory infiltrates. The histological grading of acute rejection was performed according to the Banff system (31,32).

In some cases, the differential diagnosis between mild acute rejection and HCV-related lesions was difficult (Figure 1). The diagnosis of acute rejection was favored when, as previously described (31), the portal inflammatory infiltrate contained a significant proportion of eosinophils, epithelial biliary lesions were more than occasional, endothelialitis and centrilobular lesions were present, interface hepatitis was absent and no prominent piecemeal necrosis and/or intralobular hepatocellular necrosis was present. On the contrary, a diagnosis of HCV hepatitis was favored when the portal inflammatory infiltrate was devoid of eosinophils, associated with evidence of interface hepatitis and with only occasional epithelial biliary lesions; endothelialitis, as defined before, was absent. Initial discordances between the two observers were observed in six of 70 (8.6%) cases; a final consensual decision was reached in all cases by the common reviewing of all available sections and the strict application of the criteria underlined above.

Figure 1.

Comparative histopathological features of chronic recurrent hepatitis C (A, B) and acute rejection (C, D, E, F). In this typical case of chronic recurrent hepatitis C after (A, B), prominent inflammatory infiltrates are visible in portal tracts (A, B); they are associated with piecemeal necrosis (A, arrow); there is no evidence of endothelialitis, neither in portal veins (A, PV) nor in centrilobular veins (B, CLV); biliary lesions are mild (A); there is no evidence of centrilobular inflammation (B). In this typical case of acute rejection (C, D), a mild inflammatory infiltrate, composed of lymphocytes and eosinophils, is visible in the portal tract (C); endothelialitis is prominent along the portal vein (A, PV); no evidence of piecemeal necrosis is visible; biliary epithelial cell lesions are present (C, arrow); in addition to portal lesions, an inflammatory infiltrate is visible in the centrilobular area (D, arrow) along with endothelialitis along the centrilobular vein (D, CLV). In this case (E, F) raising a problem of differential diagnosis between chronic recurrent hepatitis C and acute rejection, a dense inflammatory infiltrate is visible in the portal tract (E); hepatocellular lesions, which may be consistent with a diagnosis of chronic hepatitis C, are visible at the interface between the portal tract and the lobule (broken arrow); however, the presence of marked biliary alterations (E, arrow) and of vascular lesions (E, arrowhead) are in keeping with a diagnosis of acute rejection; this diagnosis is further supported by the presence of a centrilobular inflammatory infiltrate and of centrilobular vein endothelialitis (F, CLV). HE staining; original magnification: ×240.

Antiviral therapy

Indication for antiviral therapy was histological evidence of HCV hepatitis, as previously defined, on two consecutive liver biopsies (interval of 1–3 months) with a necro-inflammatory grade (A) ≥2 and/or a fibrosis score (F) ≥1 and absence of ongoing rejection. During antiviral therapy, the immunosuppressive regimen was maintained at target levels.

Before 1996, antiviral therapy consisted of IFN monotherapy: IFN (Schering-Plough) was initiated using low doses (1 MU subcutaneously three times a week for 2 weeks, 2 MU for the next 2 weeks, and 3 MU three times a week for the following months). Between 1996 and 2000, antiviral therapy consisted of a combined regimen of IFN-ribavirin (Schering-Plough) beginning with low doses (IFN as previously described and ribavirin 400 mg daily for 2 weeks, 800 mg for the next 2 weeks and 1200 mg for the following months). Since 2000, antiviral therapy consisted of a combined regimen of pegylated interferon (PEG-IFN) and ribavirin (Schering-Plough) beginning with low doses (ribavirin was used as described previously; PEG-IFN was given at 0.5 mg/kg/week weekly for 4 weeks, then at 1 mg/kg/week for the following months).

Every month, haemoglobin, white cell count, platelets, prothrombin time, blood urea nitrogen, serum creatinine, bilirubine and liver blood enzymes were performed. At 1, 3, 6, 12 and 18 months, and every 6 months thereafter, HCV-RNA assay and liver biopsy were performed. Antiviral therapy was maintained as possible for 12 months or more, according to efficacy and tolerance. Doses were adjusted according to leukopenia, thrombocytopenia or anemia.

Biochemical response was defined by serum ALT normalization, virological response as undetectable serum HCV-RNA at the end of treatment. Sustained virological response (SVR) was defined by undetectable serum HCV-RNA 6 months after the end of treatment and histological response as an improvement of liver histological lesions graded according to the METAVIR score.

Statistical analysis

Categorical variables were expressed as percentages, and continuous variables were presented as median (minimum and maximum). We examined the predictive value of the following parameters: age at LT, recipient gender, donor age, living or deceased donor, time of treatment after LT, IFN and ribavirin dosage, initial serum ALT level, HCV genotype, initial HCV RNA levels, immunosuppressive regimen and initial liver damage. Variables were compared using the Fisher's exact test or the Mann-Whitney test, and a p value less than 0.05 was considered statistically significant.


Outcome of HCV recurrence under antiviral therapy

Between September 1990 and December 2004, 70 LT recipients were treated in our institution with IFN for recurrent HCV. The main clinical and virological characteristics of these patients at LT, at the beginning of antiviral therapy and during treatment are summarized in Table 1. The median delay between LT and the beginning of the treatment was 16 (1–137) months; 57 of 70 (81%) treatments were initiated 6 months after LT. At the beginning of the treatment, serum ALT was elevated in 60 (86%) patients. Histological examination disclosed moderate inflammatory activity and fibrosis according to the METAVIR score (mean score was A1.5 F1.4). Fibrosis score were ≥F1 in 67 (96%) patients (F1, 39 (56%); F2, 24 (34%); F3, 4 (6%)); only three patients, all with HCV genotype 2–3 and with an inflammatory grade A2, were treated before the occurrence of fibrosis.

Table 1.  Comparison of patients who developed or not rejection under IFN therapy
 All patients N = 70Rejection N = 15No rejection N = 55p value
  1. IFN = interferon; PEG = pegylated interferon; RBV = ribavirin; MMF = myclofenolate mofetil; ALT = alanine aminotransferase.

  2. 1Some patients received IFN + RBV and then PEG + RBV.

  3. 2ALT were normal in 6 of 15 (40%) patients at time of acute rejection.

  4. 3HCV RNA at time of rejection were negative in 5(15 patients (we indicate the mediane of still positive HCV RNA).

Characteristics of patients at LT
 Median age at liver transplantation (years)51 (29–66)51 (38–62)51 (29–66)0.94
 Sex ratio (female/male), n (%)21/49 (30/70%)6/9 (40/60%)15/40 (27/73%)0.91
 Renal clearance (mL/min)110 (8–260)106 (8–158)111 (6–260)0.36
 HCV genotype (1/other), n (%)47/23 (67/33%)9/6 (60/40%)38/17 (69/31%)0.51
 Time of the transplantation (before/after 1998), n (%)32/38 (46/54%)5/10 (33/67%)49/51%0.28
 Median donor age (years)35 (14–63)35 (15–63)39 (14–63)0.62
 Deceased donor/living donor, n (%)64/6 (91/9%)14/1 (93/7%)50/5 (91/9%)0.77
 Previous acute rejection, n (%)30 (43%)10 (67%)20 (36%)0.04
Immunosuppressive regimen
  Ciclosporine/tacrolimus, n (%)28/42 (40/60%)5/10 (33/67%)23/32 (42/58%)0.55
  Azathioprine/MMF, n (%)24/23 (34/33%)4/4 (27/27%)20/19 (36/35%)0.48/0.56
  Steroids, n (%)63 (90%)15 (100%)48 (87%)0.15
Characteristics of patients at the beginning of the antiviral therapy
 Median delay between LT and antiviral treatment (months)16 (1–137)19 (3–98)14 (1–137)0.88
 Renal clearance (mL/mn)59 (28–153)57 (28–153)62 (35–131)0.58
 Median ALT level (internationnal Units)102 (23–729)81 (22–369)104 (25–729)0.21
 Median HCV RNA level (millions copies/mL)1.2 (0.002–23)1.5 (0.1–23)1.0 (0.002–21)0.24
Metavir score before antiviral therapy
  AA1.5 (1–3)A1.5 (1–3)A1.5 (1–3)0.96
  FF1.4 (1–3)F1.3 (1–2)F1.5 (1–3)0.70
Immunosuppressive regimen
  Ciclosporine/tacrolimus, n (%)22/48 (31/69%)4/11 (27/73%)18/37 (33/67%)0.65
  Azathioprine/MMF, n (%)7/32 (10/46%)0/6 (04/0%)7/26 (13/47%)0.15/0.62
  Steroids, n (%)44 (63%)9 (60%)35 (64%)0.8
Characteristics of patients during antiviral therapy
 Type of antiviral therapy (IFN/IFN-RBV/PEG-RBV) (n)17/27/432/4/135/23/30 
 Median duration of antiviral therapy (months)15 (1—62)15 (1.5–50)14 (1–62)0.74
 Median ALT level (International Units) at time of rejection2 115 (15–357) 
 Median HCV RNA level (millions copies/mL) at time of rejection3 0.2 (10−4–4.7) 
 Negative HCV RNA at the end of treatment, n (%)31 (44%)5 (33%)26 (34%)0.34

Seven patients received IFN monotherapy and 63 patients received combined therapy (19 combined IFN-ribavirin therapy, 35 combined PEG-IFN-ribavirin therapy and 8 combined IFN-ribavirin then PEG-IFN-ribavirin therapy). The median duration of antiviral treatment was 15 (1–62) months with 50 of 70 (71%) patients who received antiviral therapy for at least 12 months. During antiviral therapy (except in cases of histologically proven rejection episodes), the same immunosuppressive regimen was maintained in all patients to obtain blood trough level close to the upper limit of therapeutic target. Fifty-seven (81%) patients had biochemical response. Serum HCV-RNA was undetectable in 31 patients (44%) at the end of treatment. SVR was 23 of 70 (33%). From the different variables that were examined to determine whether they could predict SVR, completion of treatment for at least 12 months and HCV genotype other than 1 were significantly associated with viral clearance (p = 0.01 and p = 0.02, respectively). Activity grade was improved in 38 of 70 (54%) patients and fibrosis score in 17 of 70 (24%) patients.

Incidence, features and outcome of rejection under IFN

Fifteen of the 70 patients (21%) developed acute rejection during IFN-based therapy (representing 15% of acute rejection per year of IFN therapy). The number of LBs displaying histological features of rejection under IFN was 22 (from 271 LBs under IFN for the entire cohort, i.e., 8.1%). Acute rejection was mild in five (33%) patients, moderate in nine (60%) patients, and severe in one (7%) patient.

Ten of the 15 patients with acute rejection (67%) had previous history of acute rejection (all cases have been retrospectively reviewed to confirm the diagnosis) with a mean of 2.3 (1–5) biopsies with rejection before IFN. In this group, the median delay between the last available liver biopsy without evidence of rejection and the beginning of antiviral therapy was 4.7 (0.5–20.5) months. Four of the 15 (27%) patients in the acute rejection group versus nine of 55 (16%) of the remaining patients began antiviral therapy less than 6 months after LT (p = 0.36). Patients were treated for 8 (1–15) months prior to rejection. Acute rejection occurred in six of 15 patients within 6 months after the beginning of the treatment, and four of 15 acute rejections occurred after 12 months of treatment. Rejections occurred under IFN alone (2/15) or PEG-IFN-ribavirin (13/15). At the time of rejection, median ALT levels were 115 (10–357) UI/L, alkaline phosphatase levels were 138 (49–485) and total bilirubin levels were 15 (11–41). ALT level was in the normal range (thus defining subclinical rejection) in six of 15 (40%) patients with histological signs of rejection; in these patients, rejection was mild in four of six (67%) and moderate in two of six (33%). The severity of their rejection was lower than this of patients with increased ALT levels (in whom rejection was mild in one of nine (11%), moderate in seven of nine (78%) and severe in one of nine (11%), but the difference was not statistically significant p = 0.075). At time of rejection, the immunosuppressive regimen consisted in cyclosporine in four of 15 (27%) patients (mean blood trough level of 125 (89–164) ng/mL) or tacrolimus in 11 of 15 (73%) of the patients (mean blood trough level of 10.3 (4.4–17.1) ng/mL).

Baseline variables evaluated as possibly contributing to rejection are shown in Table 1. Previous history of acute rejection before IFN therapy and treatment with PEG-IFN were significantly associated with acute rejection under IFN-based therapy (p = 0.04 and p = 0.02, respectively). Therefore, the risk of rejection under IFN-based therapy was almost 3-fold higher for patients with a history of prior rejection. Otherwise, there was no significant relationship between acute rejection and age of recipient patient, donor age, living donor, time of treatment after transplant, renal clearance, immunosuppressive regimen, pretreatment ALT, HCV RNA level, initial necro-inflammatory activity or fibrosis scores.

Treatment of rejection and influence of rejection on HCV outcome

Rejection was treated by temporarily stopping IFN in 7 (47%) patients and/or by increasing immunosuppressive treatment in 12 (80%) patients and/or by adding boluses of methylprednisolone in 3 (20%) patients; all three patients had marked increases in ALT levels (>300 UI/L) and alkaline phosphatase levels (>400 UI/L); rejection was histologically severe in 1 and moderate in 2. No patient was treated for steroid-resistant rejection by OKT3. The rejection was controlled by the treatment (as defined by ALT normalization and absence of histological criteria of rejection in the subsequent biopsy) in 13 (87%) patients (2 (1–28) months); at the end of the period of study, signs of rejection persisted in only one patient, 10 months after the first rejection episode; the patient was still under IFN treatment which was not discontinued. One patient with SVR and normal liver tests had no liver biopsy since February 2000. Five of the six patients in the subclinical rejection group resolved rejection in a mean delay of 2.1 (1–28) months; similarly, resolution of rejection occurred in eight of nine patients in the clinical rejection group after 1.75 (1–28) months (p = 0.15). No chronic rejection was observed and no graft loss was observed after a follow-up of 3 (0.2–9.3) years after acute rejection. At histopathological examination of the last available liver biopsy, no patient presented with bile duct lesions or with ductopenia; two patients had criteria of acute rejection, with presence of centrilobular lesions and portal inflammatory infiltrates containing eosinophils. Serum HCV-RNA was undetectable in five of 15 (33%) patients at the end of treatment and SVR in four of 15 (27%).

The follow-up (Table 2) showed a difference (but not statistically significant) between acute rejection group and no rejection group, in levels of ALT and SVR. No significant differences were noted in mean fibrosis score, occurrence of severe fibrosis (F3–4), graft loss and death in patients with and without rejection. We found no significant differences in outcome between patients with increased ALT at rejection and those with only subclinical rejection (Table 3).

Table 2.  Follow-up of the 70 liver transplant patients after treatment with IFN according to presence of histologic evidence of rejection on posttreatment liver biopsy
 All patients N = 70Rejection N = 15No rejection N = 55p value
  1. IFN = interferon; ALT = alanine aminotransferase; SVR = sustained virological response.

Median time of follow-up after liver transplantation (years)6.4 (1–14)5.5 (2–14)7.1 (1–14)0.26
Normal ALT level, n (%)57 (81%)12 (80%)46 (84%)0.74
SVR, n (%)23 (33%)4 (27%)19 (34%)0.56
Fibrosis score Metavir after antiviral therapyF1.7 (0–4)F1.7 (0–4)F1.8 (0–4)0.68
Severe fibrosis (F3–4), n (%)20 (28%)4 (27%)16 (29%)0.85
Graft loss, n (%)2 (3%)0 (0%)2 (4%)0.45
Death, n (%)12 (17%)3 (20%)9 (16%)0.74
Table 3.  Follow-up of the 15 liver transplant patients with histological rejection under IFN according to presence of increased ALT at time of rejection versus subclinical rejection (normal ALT)
 Subclinical rejection N = 6Clinical rejection N = 9p value
  1. IFN = interferon; ALT = alanine aminotransferase; SVR = sustained virological response.

Histological grading of rejection (mild/moderate/severe)4/2/01/7/10.075
Persistent rejection at the end of follow-up, n (%)1 (17%)1 (11%)0.76
Median time to resolved rejection (months)2.1 (1–28)1.8 (1–28)0.15
Normal ALT level, n (%)5 (83%)7 (78%)0.79
SVR, n (%)2 (33%)2 (22%)0.63
Fibrosis score Metavir after antiviral therapyF1.3 (0–4)F1.9 (0–4)0.38
Severe fibrosis (F3–4), n (%)1 (17%)3 (33%)0.47
Death, n (%)1 (17%)2 (22%)0.79
Median time of follow-up after liver transplantation (years)6.6 (2.9–9.2)5.0 (2.4–14)0.56


The combination therapy of IFN and ribavirin is a widely accepted treatment for LT recipient with recurrent hepatitis C. Most of the centers now use PEG-IFN-ribavirin which seems more efficient than IFN-ribavirin. Antiviral therapy leads to an SVR in 25–45% of the patients (11,17–19). Patients with SVR have better long-term outcome in comparison to patients with ongoing infection, based on the activity score and, maybe, also on fibrosis regression (9). However, side effects of IFN therapy, such as poor haematological tolerance and depression, are common in the LT population. Moreover, the risk of rejection in transplant patients treated with IFN has been well recognized in renal transplant recipients (33,34), but is more controversial in LT recipients. Since 1994, different studies have reported that IFN after LT may precipitate acute or chronic rejection and subsequent graft loss (12,13,16). Then, several studies have supported the long-term efficacy and safety of IFN in this population. The reported incidence of rejection in LT recipients treated with IFN is very disparate and ranges between 0% and 35% (10,11,14–29).

In this study, the diagnosis of acute rejection was histological. Problems of differential diagnosis between HCV infection and acute rejection in this particular setting have been underlined in previous papers and in the Banff classification. We paid therefore a particular attention to this point: only validated histological criteria have been used and a double examination of all liver biopsies has been performed.

In the present study, based on frequent protocol biopsies, we found that (1) acute rejection under IFN-based therapy often occurs in LT patients (21%, among which six of 15 patients with normal ALT), (2) most cases corresponded to mild or moderate grades of rejection (93%) and (3) an early treatment usually permitted favorable outcome, without significantly reducing the efficiency of antiviral therapy. In the absence of ‘control group’ (patients with rejection that would have not been treated), our results cannot be interpreted as evidence that treatment of rejection under IFN must be systematically applied. Some acute rejection episodes (maybe even more in subclinical rejection) may have improved without specific treatment. However, it can be argued that not to treat an ongoing rejection can be deleterious since it has been demonstrated in several studies that occurrence of rejection episodes is associated with a more aggressive progression of recurrent hepatitis C and with a higher risk to develop chronic rejection and graft loss (12,16). In the same way, in the absence of a control group (patients with histological HCV recurrence and not treated), we could not state the incidence of acute rejection in patients with recurrent HCV infection and without antiviral therapy. However, the constitution of a real control group would imply not to treat cases of recurrent HCV hepatitis after LT, which must be treated according to current evidence and recommendations (9,35,36).

Several features may explain the high incidence of histological rejection after IFN therapy in the present study. It may probably be the reflect of the performance of frequent protocol biopsies (21% of treated patients developed histological features of acute rejection under IFN, but this represented only 8% of LBs with rejection from all LBs from the entire cohort of patients under IFN), disclosing subclinical rejection even in patients with normal ALT (as observed in 40% of our patients with histological features of acute rejection). Most of the studies only reported the results of LBs at the beginning and at the end of antiviral treatment, and, sometimes, in the case of increased liver enzymes. This probably led to underestimate the incidence of rejection. Moreover, the incidence of rejection in these studies without liver biopsy during antiviral therapy can be more underestimated in non responding patients, when the increase in liver enzymes would have been falsely related to recurrent hepatitis C. In previous studies, patients with previous acute rejection were usually excluded from IFN therapy (15). Actually, the present study confirms that a previous history of acute rejection before IFN therapy is a major risk factor for the occurrence of subsequent acute rejection under IFN therapy (p = 0.04). However, our data also demonstrate that such rejection episodes usually present favorable evolution. In consequence, our experience does not support the concept that a previous history of rejection must be considered a contraindication to antiviral treatment. Our data underline the importance of a liver biopsy before starting antiviral treatment, in order to eliminate any ongoing histological rejection. The delay after LT when initiating antiviral treatment is probably greatly relevant; the risk of rejection is higher in the first month posttransplant (37). A ‘lower’ immunosuppressive level, when patients have been weaned of steroids and/or mycophenolate mofetil or azathioprine may contribute to an increased evidence of rejection (15). The median delay between LT and antiviral treatment was 16 months in our series, shorter than some studies (for example 54 months in the randomized study of Samuel and colleagues, using IFN and ribavirin, in which they observed only one of 28 episodes of rejection in the treated group (10)). In the present study, we found no significant relation between the time to treatment after LT and evidence of rejection, and this delay was even longer in acute rejection group (19 months vs. 14 months in nonacute rejection group). The duration of treatment with IFN is probably another explanation for a higher incidence of acute rejection in our study. Seventy-one percent of our patients received antiviral therapy for more than 12 months. If we had stopped treatment after 48 weeks as in most previous studies, the rate of acute rejection under IFN treatment would have been of 11 of 70. It would possibly be more pertinent to report rate of acute rejection per year of IFN therapy (16% in our study). The type of antiviral therapy might modify the prevalence of rejection. It has been suggested that adjunction of ribavirin might decrease the risk of rejection, since a lower number of cases of rejection has been reported in LT patients treated with IFN plus ribavirin therapy than with IFN alone (2/15 vs. 0/15 in our study). The immunomodulatory action of the ribavirin might explain these results (38). On the contrary, PEG-IFN might theoretically increase the risk of rejection due to its long-lasting effects and its renal clearance, which is frequently decreased in LT patients (from 110 (8–260) at LT to 59 (28–153) mL/mn at the beginning of IFN after LT in our study). We previously reported a prevalence of 25% of mild rejection episodes in a series of 20 patients treated with PEG-IFN-ribavirin (11). In the present study, 61% of the patients received PEG-IFN-ribavirin and we found that 13 of 15 (87%) patients who developed acute rejection after IFN-based therapy were treated with PEG-IFN-ribavirin (vs. 30/55 (55%) in nonacute rejection group, p = 0.02). It may have been suggested that the risk of rejection under IFN therapy may depend on the type and level of immunosuppressive regimen used at the time of antiviral treatment, however, we find no evidence supporting this hypothesis in the present study. During antiviral therapy, the immunosuppressive regimen must be maintained at usual levels, with blood concentrations close to the therapeutic targets, as in our series.

In addition to the incidence of rejection under IFN therapy, the evaluation of long-term clinical and histological consequences of this complication is of great relevance. In our series, a favorable outcome was usually observed: most cases of rejection under IFN had a rapidly favorable course and were successfully treated. However, some patients had persistence of histological rejection for several months. It may probably be due, in part, to our strategy to not systematically stop IFN-based therapy with, consequently, a higher risk for persistent rejection. We took this attitude in cases of mild or moderate rejection with normal ALT levels, in which only the immunosuppressive regimen was modified. The occurrence of rejection episodes as well as the use of potent immunosuppressive agents for their treatment (including methylprednisone boluses, OKT3) has been associated with a more aggressive progression of HCV infection in several previous studies (2,3,39,40). One could speculate that mild rejection, as we observed, might be in part responsible for progressive fibrosis after viral clearance. If this is indeed the case, then it would be important to perform repeated routine liver biopsies during treatment to identify and treat rejection. However, the clinical benefit of repeated LBs must be balanced by taking into account the potential risks of the procedure; the morbidity and mortality rates of liver biopsy are respectively evaluated to be 0.3–0.6% and 0–0.01% (41–43). Our results suggest that early diagnosis with protocol biopsies and early treatment of acute rejection might explain the high rate of favorable evolution observed in our series, as compared to previous ones (12,15,16). No chronic rejection, graft loss or significant increase of fibrosis were observed in the acute rejection group. These results were obtained despite a reduced efficacy of the antiviral therapy with only 27% of SVR in the group of patients who developed acute rejection during IFN treatment.

We think important to take into account even the subclinical rejection episodes, since, in our study, their outcome was comparable to that of clinical rejection episodes. However, a prospective study including a higher number of patients with a longer follow-up period would be necessary to definitely rule out a negative impact of rejection and its treatment on fibrosis progression and to evaluate the possible correlations between the severity of rejection and the outcome of recurrent hepatitis C on SVR and fibrosis progression. Despite the possible side effects of IFN-based therapy in LT recipients, we privileged the option of an early antiviral treatment when recurrent hepatitis C is still a mild disease. This option has been, and remains, debated. A recent report, by Futagawa et al., based on UNOS Registry data, compared the outcome of patients during two successive 5-year periods (1992–1996 vs. 1997–2002); in accordance with previous studies, it showed a decrease in graft survival during the second period. Among the causes of graft failure, the incidence of immunologic graft loss was decreased whereas the incidence of disease recurrence was increased. Such data strongly support the idea that early treatment of HCV recurrence would have a positive impact on the overall survival of LT recipients (35).

In conclusion, this study demonstrates that histological acute rejection under IFN-based therapy often occurs in LT patients; however, it is usually mild and its evolution is favorable. This encourages to continue the use of IFN-based combination antiviral therapy for the treatment of recurrent hepatitis C after LT; however, special attention must be paid to patients with previous history of acute rejection, which constitutes a major risk factor for the occurrence of acute rejection under antiviral IFN therapy.