Telephone: 415-353-8725; FAX: 415-353-8709
Hepatitis C etiology of liver disease is strongly associated with early acute rejection following liver transplantation
Article first published online: 23 JUL 2004
Copyright © 2004 American Association for the Study of Liver Diseases
Volume 10, Issue 8, pages 975–985, August 2004
How to Cite
McTaggart, R. A., Terrault, N. A., Vardanian, A. J., Bostrom, A. and Feng, S. (2004), Hepatitis C etiology of liver disease is strongly associated with early acute rejection following liver transplantation. Liver Transpl, 10: 975–985. doi: 10.1002/lt.20213
- Issue published online: 23 JUL 2004
- Article first published online: 23 JUL 2004
- NIH/INRSA T32 DK60414
Although recurrent hepatitis C (HCV) occurs universally after liver transplantation (LT), its tempo and severity are variable and unpredictable. Diagnosis and treatment of early acute rejection (EAR) likely affect the course of recurrent HCV disease. We have studied a contemporary cohort of LT recipients to reexamine risk factors for EAR. We hypothesized that HCV etiology may represent a significant risk factor for EAR for many reasons. First, recurrent disease commonly causes biochemical abnormalities prompting allograft biopsy. Second, overlapping histologic features of acute rejection and recurrent HCV ambiguity may result in diagnostic ambiguity. Finally, the biology of hepatitis may precipitate an antidonor response in addition to an antiviral response. Records of 285 adult recipients undergoing primary LT for cirrhosis between January 1, 1999, and December 31, 2002, were retrospectively reviewed. EAR was defined as a biopsy-proven or an empirically treated episode within 6 months of LT. Cox proportional hazards analysis identified donor, recipient, transplant, and posttransplant characteristics associated with EAR; Kaplan-Meier analysis was used to assess rejection by etiology. HCV cirrhosis was the etiology for 51% of all LT recipients. There were 135 episodes of EAR (127 biopsy proven) in 117 recipients for an overall incidence of 41%. Patient groups with HCV and cholestatic / autoimmune disease groups exhibited the highest incidence of rejection at 49%. Recipient gender, ethnicity, etiology, LT year, and posttransplant immunosuppression levels were risk factors for EAR in univariate analysis; HCV etiology and female gender remained robust risk factors in multivariate analysis. Interferon-based therapy did not impact the incidence or timing of EAR. In conclusion, HCV etiology is strongly associated with EAR. HCV allograft reinfection may create an immunologic environment predisposed to EAR. Alternatively, the association of HCV and EAR may result from an increased frequency of allograft biopsy and may be further exacerbated by inability to accurately diagnose EAR in the setting of recurrent HCV. (Liver Transpl 2004;10:975–985.)
In the United States and Europe, hepatitis C (HCV) cirrhosis is the most common indication for liver transplantation. Unlike other indications, recurrent disease is universal, although tempo and disease severity vary considerably. Historically, outcomes of transplantation for HCV cirrhosis matched those of transplantation for other etiologies of cirrhosis.1–5 However, more recent data suggest that HCV etiology is associated with increased risk of graft loss after transplantation6–8 and that the severity of recurrent disease has escalated over time.7, 9 Studies exploring the apparent accelerated natural history of HCV allograft infection have identified donor age, recipient gender, virologic factors, and immunosuppression regimens as critical factors.3, 8, 10–13 Although it may be possible to exert some control over which donor livers are transplanted into HCV recipients, clearly immunosuppression represents the variable that may be most easily altered in the interest of optimizing outcomes for HCV recipients.
The potential relationship between immunosuppression and severity of recurrent HCV is complex. Most transplant centers utilize established protocols for immunosuppression after LT with deviations motivated by specific indications. For example, immunosuppression may be intensified to treat rejection or reduced to ameliorate untoward side effects. In general, rejection has historically represented an easily treatable and thus relatively benign diagnosis. Previous data have gone so far as to suggest that in an episode of early rejection was associated with improved long-term outcomes.19, 20 However, in the HCV context, the diagnosis of rejection may be uniquely deleterious, as it typically precipitates intensification of immunosuppression, which is thought to worsen recurrent disease.10, 12, 21, 22 More recent data has complicated the picture further, suggesting that changes in immunosuppression rather than simply the total amount of immunosuppression impact upon recurrent HCV.9, 22, 23 Irrespective of the exact effects, it appears as if immunosuppression has the distinct potential of altering the tempo of recurrent disease.
We have undertaken the present study to identify the risk factors for early acute rejection (EAR) after LT and the relationship between HCV and EAR. We limited our analysis to the first 6 months after transplantation, a time period during which both acute rejection and HCV recurrence frequently occur. Although liver biopsy is widely regarded as the gold standard for diagnosis of rejection, unfortunately, in the setting of possible recurrent HCV disease, the transplant community has become increasingly aware that histological features of an alloimmune response may substantially overlap with those of an antiviral response, blurring the distinction between acute rejection and recurrent HCV disease.24, 25 We hypothesized that the increased frequency of allograft biopsy driven by abnormal liver tests compounded with the histological ambiguity between rejection and recurrent HCV disease might render HCV etiology as a significant risk factor for the diagnosis of EAR.
Patients and Methods
This study was approved by the University of California San Francisco Institutional Review Board and conforms to the ethical guidelines of the 1975 Declaration of Helsinki.
Patients and Data Collection
Recipient and Donor Parameters.
Between January 1, 1999 and December 31, 2002, a total of 400 orthotopic liver transplants were performed at the University of California San Francisco of which 362 were in adults. Within the adult group, transplants for fulminant liver disease (n = 24), noncirrhotic liver disease (n = 12; amyloidosis, oxalosis, non-hepatocellular carcinoma tumors, and polycystic liver disease), and retransplants (n = 28) were excluded. In addition, grafts that failed within 7 days secondary to recipient death or retransplantation (n = 10), and those diagnosed with hepatic artery thrombosis within 7 days (n = 3) were also excluded. Retrospective medical record review was performed for the 285 remaining recipients and their corresponding donors. Details regarding donor demographics and type, as well as recipient demographics, etiology and severity of liver disease, and the presence or absence of hepatocellular carcinoma in the explanted liver were collected. Any recipient with documentation of pretransplant and/or posttransplant serum HCV RNA positivity was included in the HCV group.
Cold ischemia time was defined as the interval from deceased donor cross-clamp to removal from cold storage for anastomosis. For cases of living donor LT where donor and recipient operations were simultaneously performed, no cold ischemia time was recorded. Warm ischemia time represents venous anastomotic time, defined as the interval between removal from cold storage and venous reperfusion.
Posttransplant information such as alanine aminotransferase and tacrolimus levels, allograft biopsies, immunosuppression, and anti-HCV treatment were collected. Pathology reports for all biopsies performed within 6 months of transplantation were reviewed.
At the University of California San Francisco, triple immunosuppression comprised of steroids, tacrolimus, and mycophenolate mofetil is routinely used. Methylprednisolone (1 g) is given intraoperatively, then rapidly tapered to 20 mg/day by posttransplant day 7 and then to 5 mg/day by posttransplant day 42. Mycophenolate mofetil (1 g po bid) is initiated immediately after transplantation and dosing adjusted as necessitated by hematological abnormalities or clinical symptoms. Nearly all patients (98%) were discharged on mycophenolate mofetil. Tacrolimus, typically initiated within 24–48 hours after transplantation but delayed in the setting of renal dysfunction, was orally dosed to achieve 12-hour trough levels of 8–12 ng/mL. If tacrolimus produced significant toxicity, most often neurotoxicity, cyclosporine was substituted and dosed to achieve 12-hour trough levels of 200–250 ng/mL. Tacrolimus and cyclosporine trough levels were obtained daily during hospitalization, but neither were routinely followed after initial hospital discharge. Sirolimus was also used during the study period beginning in November, 1999, for two primary indications: significant renal dysfunction as a bridge to calcineurin inhibitors, or hepatocellular carcinoma for potential antitumor benefit. In total, 43 recipients (15%) were exposed to sirolimus for greater than 1 month. Sirolimus substituted for tacrolimus in 21 (7%) recipients and for mycophenolate mofetil in 7 recipients (2%); it was added to standard triple therapy for 15 recipients (5%).
Diagnosis of Acute Rejection
During the study period, allograft biopsy was prompted by abnormal liver tests (rising or failing to normalize) in the absence of another diagnosis. Liver tests were monitored daily on an in-patient basis. Outpatient testing was performed twice weekly until 6 weeks after transplantation, weekly until 3 months after transplantation, and once every 2 weeks until 6 months after transplantation. During the early posttransplant period (8 weeks), ultrasound and Doppler examination of the allograft and its vasculature were obtained prior to or in conjunction with biopsy. The Banff schema was utilized to diagnose and grade acute allograft rejection.26 We reviewed all allograft biopsy reports within 6 months of transplant to identify rejection episodes. Any biopsy with a diagnosis of acute rejection, irrespective of treatment, was considered an episode of rejection. In addition to biopsy-proven rejection episodes, empiric treatment courses for rejection (defined below) were also included as rejection episodes. Biopsies demonstrating rejection within 30 days of an index acute rejection event (diagnostic biopsy or treatment) were not considered as separate episodes of rejection.
Treatment of Acute Rejection
Treatment of acute rejection was defined as bolus corticosteroids with or without an antilymphocyte preparation. First-line treatment for rejection was typically bolus intravenous methylprednisolone (1 g qd for 2 days) followed by oral prednisone recycle (rapid taper from 200 mg qd to 20 mg qd over 6 days, followed by a more gradual taper to 5 mg qd over 5 weeks). On occasion, particularly beyond 1 month after transplantation, recipients were treated with either oral steroid recycle or bolus intravenous corticosteroids alone rather than the combination. For the 18 recipients with untreated biopsy-proven rejection, both clinical assessment and histologic findings (minimal or mild acute rejection) were considered in the decision-making process, while clinical assessment alone was used for the 7 recipients with empiric treatment. Antilymphocyte therapy, initiated for severe and/or steroid-refractory rejection episodes, typically consisted of a 7- to 10-day course of OKT3 or rabbit thymoglobulin accompanied by bolus intravenous steroids followed by a slow taper of prednisone from 20 to 5 mg qd over five weeks. Alterations of baseline immunosuppression, such as increasing the dose of one or more immunosuppressive agents or conversion from one maintenance agent to another, were not considered treatment for acute rejection.
Descriptive statistics for the study cohort, including medians and ranges, were provided as appropriate. Kaplan-Meier analysis of EAR according to disease etiology was performed and compared using log-rank and Wilcoxon tests. Donor, recipient, transplant, and posttransplant factors were assessed in univariate and multivariate models for association with EAR, using Cox proportional hazards analysis. Etiology of liver disease as a predictor of EAR was analyzed in detail. Each disease etiology was first compared with all other etiologies and then individually compared to eliminate the influence of the specific composition of our cohort. Correlations between identified risk factors for EAR were determined by Spearman rank coefficients. All statistical analyses were performed using SAS, version 8.2 (SAS Institute, Cary, NC).
Recipient, Donor, and Transplant Characteristics
This study is a contemporary cohort of 285 adults undergoing primary LT for cirrhosis between January 1,1999, and December 31, 2002. Recipient, donor, and transplant characteristics are shown in Table 1. The most common indication for transplantation was HCV cirrhosis (146 of 285 recipients; 51%). Three recipients had both chronic HCV and hepatitis B (HBV) disease and were included in the HCV etiology group. The median Model for End-stage Liver Disease (MELD) score calculated from laboratory values obtained just prior to transplantation was 18 (mean, 20.7; range, 6-72). Child-Turcotte-Pugh Class A, B, or C, determined by clinical assessment and laboratory values on the day of admission for the hospitalization for LT, was seen in 17%, 42%, and 58% of recipients, respectively. Adult-to-adult living donor liver transplantation at the University of California San Francisco began in January, 2000. In total, 51 of 285 recipients (18%) underwent living donor LT; 4 of 47 recipients received whole grafts from living donors who were undergoing deceased donor transplantation for amyloidosis. Fourteen recipients (5%) underwent combined liver-kidney transplantation.
|Age (yrs)||Median (range)||40.0 (8.5–78.5)|
|African American||7 (2%)|
|Age (yrs)||Median (range)||52.2 (18.0–74.6)|
|African American||11 (4%)|
|PBC/ PSC/ AIH||41 (14%)|
|Pretransplant MELD†||Median (range)||18 (6–52)|
|Liver + Kidney||14 (5%)|
|Cold IT§ (hrs)||Median (range)||9.5 (3–20)|
|Warm IT (mins)||Median (range)||45 (24–99)|
|Tac level d5–7 (ng/mL)||Median (range)||6.8 (<3.0–22.4)|
|(n = 276)||>8.0||72 (33%)|
|IFN Start (days)||Median (range)||42.5 (6–178)|
|(n = 76)|
Diagnosis, Incidence, and Treatment of EAR
A total of 117 recipients experienced 1 or more episodes of EAR for an overall incidence of 41%. The incidence of EAR was 49% for HCV cirrhosis; 49% for primary biliary cirrhosis (PBC) / primary sclerosing cholangitis(PSC) / autoimmune hepatitis (AIH); 34% for cryptogenic cirrhosis; 27% for alcoholic cirrhosis; 19% for HBV cirrhosis; and 29% for miscellaneous diagnoses. Kaplan-Meier analyses of EAR by pretransplant diagnoses are shown in Figures 1, 2, and 3. The first episode occurred within 6 weeks of transplantation for 83 of the 117 recipients (71%). Sixteen recipients had more than 1 episode of EAR: 14 experienced 2 episodes, and 2 experienced 3 episodes. The etiology of liver disease for these 16 recipients were HCV cirrhosis (n = 8), cholestatic disease / autoimmune hepatitis (n = 4), cryptogenic cirrhosis (n = 2), alcoholic cirrhosis (n = 1), and miscellaneous diagnoses (n = 1).
Of the 135 EAR episodes, the majority (93%) were biopsy proven. Liver biopsy, typically triggered by abnormal liver tests, varied in frequency by disease etiology but not by year of transplantation. Recipients with HCV had a median biopsy frequency of 1.98 biopsies for the 6-month posttransplant period, essentially double that of the median for all etiologies combined (.96 biopsies / 6 months). Of the 127 biopsy-proven rejection episodes, 109 were treated (bolus intravenous corticosteroids and/or antilymphocyte therapy and/or oral corticosteroid recycle), while 18 were untreated (Table 2). The most common reason for treatment deferral, as documented in the medical record for 11 episodes, was clinical uncertainty regarding the biopsy's diagnostic accuracy in patients with HCV infection. Of the 18 untreated episodes, 13 occurred when the recipient was an outpatient. Of these 13, 9 episodes did precipitate an increase in baseline immunusuppression (tacrolimus, 6; mycophenolate mofetil, 2; or sirolimus, 1). For the remaining 5 episodes, which occurred during hospitalization, frequent changes in immunosuppression obscured our ability to identify those specifically motivated by biopsy results. Eight episodes of rejection were not biopsy proven but were defined by empiric courses of antirejection treatment. Five of these episodes had no associated biopsy; in 3 episodes, recipients had hematological contraindications to biopsy. The other 3 episodes had associated biopsies that were not diagnostic of rejection; in 2 cases, the biopsy was preliminarily interpreted as consistent with rejection, but the final pathology report was nondiagnostic of rejection; in the remaining case, biopsy was performed after initiation of treatment but did not confirm suspected rejection.
|Total EAR Episodes||N (%)||Diagnoses|
|135 (100)||79 (58)||56 (42)|
|Biopsy-proven rejection episodes||127 (94)||72 (53)||55 (41)|
|Clinical decision to forego treatment||11||11||0|
|Biopsy obtained during laparotomy for other indications showed rejection||2||2||0|
|Severe infection contraindicating treatment for rejection||2||0||2|
|Rapid normalization of liver tests||2||0||2|
|Unknown: chart unavailable for review||1||0||1|
|Biopsy-unproven rejection episodes||8 (6)||7 (5)||1 (1)|
|No biopsy performed||5||5||0|
|Treatment initiated; final biopsy interpretation not diagnostic of rejection||2||1||1|
|Treatment initiated; subsequent biopsy showed no rejection||1||1||0|
|Hazard Ratio||95% CI||P Value|
|Age (per year increment)||1.01||1.00–1.02||.081|
|Gender female vs. male||1.08||0.75–1.55||.68|
|Donor type deceased vs. living||1.47||0.87–2.49||.15|
|Age (per year increment)||1.00||0.98–1.01||.62|
|Gender female vs. male||1.60||1.11–2.31||.013|
|Ethnicity Asian vs. Caucasian||.48||0.24–0.96||.038|
|Liver disease etiology|
|HCV vs. all other etiologies||1.68||1.16–2.43||.0063|
|HCV vs. HBV||3.19||1.47–6.94||.0034|
|HCV vs. alcohol||2.18||1.00–4.73||.050|
|HCV vs. cryptogenic||1.59||.82–3.09||.17|
|HCV vs. AIH/ PBC/ PSC||1.00||.61–1.64||1.00|
|HBV vs. all other etiologies||.36||.17–.77||.0088|
|HBV vs. AIH/ PBC/ PSC||.31||.13–.74||.0082|
|HBV vs. cryptogenic||.50||.19–1.31||.16|
|HBV vs. alcohol||.68||.24–1.94||.47|
|AIH/ PBC/ PSC vs. all other etiologies||1.32||.82–2.14||.25|
|AIH/ PBC/ PSC vs. alcohol||2.18||.92–5.15||.077|
|AIH/ PBC/ PSC vs. cryptogenic||1.59||.75–3.40||.23|
|Alcohol vs. all other etiologies||.56||.26–1.20||.13|
|Alcohol vs. cryptogenic||.73||.28–1.92||.53|
|Pretransplant MELD ≥28||1.50||.97–2.30||.067|
|Year 2000 vs. 1999||2.10||1.71–3.77||.013|
|Year 2001 vs. 1999||1.57||.89–2.78||.12|
|Year 2002 vs. 1999||1.42||.80–2.52||.23|
|Cold IT* (per hour increment)||1.00||.93–1.07||.99|
|Warm IT (per minute increment)||1.00||.99–1.02||.61|
|POD 1 ALT||1.00||1.00–1.00||.45|
|Tac level d5–7 (per 1.0 decrement)†||1.06||1.02–1.11||.0073|
Of the 117 treatment courses given for biopsy-proven and biopsy-unproven EAR, 18 episodes (15%) in 17 recipients required antilymphocyte therapy, either OKT3 or rabbit thymoglobulin. The incidence of EAR requiring antilymphocyte therapy was 5% for HCV cirrhosis (8 of 146), 10% for cholestatic disease / AIH (4 of 41), 14% for cryptogenic cirrhosis (4 of 29), and 14% (1 of 7) for alcoholic cirrhosis. No recipient undergoing transplantation for HBV or miscellaneous diagnoses required antilymphocyte therapy.
Risk Factors Associated With EAR: Univariate and Multivariate Models
Donor, recipient, and transplant factors associated with EAR were identified by Cox proportional hazards analysis of time to EAR with censoring of deaths and graft failures (Table 3). Univariate models show that recipient female gender, transplantation in 2000 compared to 1999, and lower posttransplant tacrolimus levels on posttransplant days 5–7 (Tac level d5-7) were significantly associated with increased EAR, while Asian ethnicity was significantly associated with decreased EAR. HCV disease etiology was associated with an increased risk compared to all other disease etiologies (hazard ratio [HR] 1.68; P = .0063), to HBV (HR 3.19; P = .0034), and to alcohol-associated liver disease (HR 2.17; P = .050). HCV and PBC / PSC / AIH were the disease etiologies most strongly predisposed to EAR, while HBV and alcohol were etiologies least correlated with EAR.
In multivariate models (Table 4), HCV etiology was associated with approximately twice the risk for EAR compared to cryptogenic (HR 2.10; P = .042), alcohol (HR 2.12; P = .06), and HBV (HR 2.09; P = .10) etiologies but with comparable risk for EAR compared to PBC / PSC / AIH (HR 1.28; P = 0.36). Recipient gender, pretransplant MELD score, and transplant year remained independently associated with EAR. Asian ethnicity and HBV etiology were not independently predictive of EAR, since they were highly associated (23 of 27 Asians compared to 13 of 248 non-Asians had HBV; Fisher's exact test P value < .0001). A second multivariate analysis (Table 5), excluding 9 recipients who rejected prior to 7 days after transplantation and six recipients who received cyclosporine showed that recipient female gender and HCV etiology remained independent risk factors for EAR. However, transplant year was no longer significantly associated with EAR. Lower Tac d5-7 levels were significantly correlated with both higher pretransplant MELD score (Spearman rank correlation, −.34; P = .0000) and later transplant year (Spearman rank correlation, −.17; P = .019).
|Characteristics||Hazard Ratio||95% CI||P Value|
|Donor age (per year increment)||1.007||1.00–1.019||.22|
|Liver disease etiology||HCV vs. all others||1.71||1.17–2.52||.0060|
|Recipient gender||female vs. male||1.79||1.21–2.65||.0037|
|Recipient ethnicity||Asian vs. Caucasian||.52||.26–1.06||.22|
|Transplant year||2000 vs. 1999||2.22||1.23–4.02||.0085|
|2001 vs. 1999||1.98||1.11–3.53||.020|
|2002 vs. 1999||1.39||.78–2.48||.27|
|Characteristics||Hazard Ratio||95% CI||P Value|
|Donor age (per year increment)||1.01||1.00–1.02||.25|
|Liver disease etiology||HCV vs. all others||1.72||1.14–2.58||.0091|
|Recipient gender||female vs. male||1.71||1.13–2.60||.012|
|Recipient ethnicity||Asian vs. Caucasian||.48||.22–1.05||.067|
|Transplant year||2000 vs. 1999||1.67||.88–3.16||.11|
|2001 vs. 1999||1.61||.88–2.95||.13|
|2002 vs. 1999||1.11||.60–2.04||.74|
|Tac level d5–7 (per 1.0 decrement)||1.04||.99–1.09||.088|
Interferon (IFN)-Based Therapy and EAR in HCV Recipients
A separate analysis of HCV recipients was performed to assess whether IFN-based therapy exerted any impact on the incidence of EAR. During the study period, HCV treatment was utilized preemptively and for treatment of those with documented recurrent disease. A total of 76 (52%) of HCV recipients were exposed to IFN therapy (standard or pegylated), with and without ribavirin within 6 months of transplantation. Approximately one-third were initiated on a therapeutic basis after documentation of recurrent HCV while the remaining two-thirds were initiated on a prophylactic basis, either on or off study. Those receiving IFN preemptively typically initiated as early as 2 weeks after and as late as 4 months after transplantation (Table 1). Although there were differences between the groups who did and did not receive IFN-based therapy, neither achieved statistical significance. Specifically, there was no significant difference in either the incidence of rejection (41% rejection with IFN compared to 57% rejection without IFN; Fisher's exact test P value = .068) or the timing of the first rejection episode (Figure 3; log-rank P value = .095) between those who received and those who did not receive IFN-based therapy within 6 months of transplantation. There was also no difference in either the day of initiation (posttransplant day 70.1 vs. 60.6; P value = .29) or the duration of IFN-based therapy (85.6 vs. 102.5 days; P value = .13) for rejectors compared to nonrejectors.
Our retrospective study of a single-center cohort of liver transplant recipients managed under a standardized immunosuppression protocol has identified a strong association between HCV etiology and the diagnosis of EAR after LT. While there have been several previous studies published within the last five years exploring predictors of EAR,19, 27–31 including two specifically focused upon the etiology of liver disease,27, 30 our study is particularly notable for the large size of our cohort of cirrhotic patients (n = 285) transplanted within a short (4 years) and contemporary (1999-2002) time span. These features are critical to our study design, considering the dynamic landscape of LT. The last 10 to 15 years have witnessed substantial changes in allocation strategies, surgical techniques, donor and recipient characteristics, and immunosuppression options. The results of studies examining transplants that were either performed a decade ago or spanned a decade or more may have limited relevance to current transplant recipients.
We focused on the first 6 months after transplantation for two reasons. First, it has long been recognized that the vast majority of first acute rejection episodes occurs during this period.19, 29, 31, 32 The reported incidence of early rejection, most frequently defined as within 6 weeks of transplantation, range between 40% and 80%. Our overall EAR incidence of 41%, which encompasses 6 months after transplantation, falls at the lower end of this spectrum. This result is not unexpected, as we did not perform protocol biopsies. Second, it has more recently become accepted that clinically evident recurrent HCV frequently occurs within 6 months of transplantation. Although HCV viral loads decrease in the immediate peritransplant period, peripheral blood viral titers begin to increase steeply within days after transplantation, frequently peaking 2 to 4 months after transplantation at levels substantially higher than pretransplant values.15, 33–35 Protocol biopsy studies have diagnosed acute HCV in 70% of HCV recipients within 6 months of transplantation.33, 36, 37 The concordance in timing of acute rejection and recurrent HCV led us to postulate that HCV etiology may influence the diagnosis of EAR in several ways. First, HCV-positive recipients are more likely to develop abnormal liver tests, which trigger more frequent allograft biopsy. Second, substantial overlap in the histological appearance of acute rejection and recurrent HCV results in diagnostic ambiguity and uncertainty. And third, acute reinfection of the allograft may precipitate an antidonor response.
Our study did indeed identify a strong and independent association between HCV etiology and the diagnosis of EAR. In multivariate analysis, HCV etiology was associated with a higher risk for rejection compared to both HBV and alcohol etiologies but with a comparable risk of rejection compared to PBC / PSC / AIH. Time to EAR generated nearly superimposable curves for the HCV and PBC / PSC / AIH cohorts. Previous studies have consistently demonstrated that the cholestatic (PBC and PSC) and autoimmune (AIH) etiologies are associated with the highest risk of acute rejection after transplantation, while alcohol and HBV etiologies are associated with the lowest risk of acute rejection after transplantation.19, 27, 29, 31, 38, 39 In contrast, the risk posed by HCV etiology has been variably reported, if at all. Some have reported results concordant with ours indicating high risk,29, 38 while others have suggested an intermediate risk19, 31; several studies, unfortunately, have provided no information, as HBV and HCV etiologies were combined as posthepatitic cirrhosis.27, 30, 33 The discrepant results may arise from important differences in study demographics as well as immunosuppression strategies and utilization of protocol biopsies. In our study, the majority of EAR episodes were identified by biopsies for cause. We did find that, as a group, HCV recipients had the highest rate of biopsy within the first 6 months after transplantation. We surmise that HCV recipients had a higher frequency of abnormal liver tests. Since the diagnosis of EAR typically precipitated a course of treatment and therefore directly affected the recipient, we consider it important to elucidate the association between HCV etiology and EAR, irrespective of possible biopsy bias.
Our study also demonstrated a significant and independent association between recipient female gender and EAR. While there has been some suggestion that female recipients of male livers may be predisposed to chronic rejection,29, 40, 41 there has been no previous reports of gender predisposition to EAR. Our analysis also identified insufficient immunosuppression, identified by low average Tac d5-7 levels, as predisposing to EAR. Since our practice is to delay initiation of tacrolimus until return of renal function, we identified a strong correlation between low immunosuppression levels and high pretransplant disease severity, designated by a laboratory MELD score ≥28 immediately prior to transplantation. This association translated into a tendency for the sickest recipients to reject. Previous studies have reported contrary results in that recipients entering transplantation with greater nutritional and functional hepatocellular reserve experience more rejection.28, 31 In addition, Weisner and colleagues identified that low serum creatinine and the lack of edema or ascites—two measures that connote decreased disease severity —were significant and independent risk factors for EAR.19 None of these studies provided information regarding posttransplant immunosuppression doses or levels. Our results, however, suggest that recipients with high disease severity nevertheless can mount an immune response and reject if they are not receiving adequate immunosuppression.
Finally, we examined the potential impact of IFN-based therapy on the diagnosis of EAR in our HCV cohort. The antiviral activity of IFN and ribavirin, at least in part, stems from stimulation of both innate and adaptive immunity, which, in theory, may precipitate allogeneic immune responses. The general weight of the literature, however, does not support a causal relationship between either prophylactic or therapeutic antiviral therapy and acute rejection.42–47 Our study yielded concordant results in that the incidence and the timing of EAR did not differ between those who were and were not exposed to IFN-based therapy within 6 months of transplantation. Very recently, however, one report suggested that response to antiviral therapy results in improved hepatic function which enhances the metabolism of calcineurin inhibitors and thereby may precipitate acute rejection.48 This intriguing hypothesis certainly merits prospective consideration.
We have identified, in a large and contemporary cohort of adult recipients undergoing transplantation for cirrhosis, that HCV etiology is strongly and independently associated with the diagnosis of EAR. The association begs the question as to whether HCV recipients who face the unique risk of early recurrent disease are being incorrectly diagnosed with EAR secondary to increased frequency of allograft biopsy exacerbated by histologic ambiguity. While biopsy bias is difficult to address, our study questions the wisdom of relying on liver biopsy—the historical gold standard— to distinguish between acute rejection and recurrent HCV. Inter- and intra-rater reliability and the predictive value of liver biopsy in the diagnosis of acute rejection was strongly validated years ago in an era where HCV was not the predominant indication for transplantation and recurrent disease was not as rapidly progressive.49 Perhaps a similar study is necessary to determine the ability and accuracy of liver biopsy to identify EAR in the current climate of transplantation with the predominance of HCV and the well-recognized incidence of allograft reinfection.
An alternative explanation for the association of HCV and EAR is that HCV recipients may actually have a higher incidence of EAR. It is certainly easy to postulate that an active viral infection such as recurrent HCV may stimulate immune responses of not only antiviral but also antidonor specificity. Interestingly, studies of HCV infection in native livers have demonstrated that the binding of E2, an HCV envelope protein, to CD81, its putative receptor on T cells, delivers a costimulatory signal resulting in strongly enhanced T cell proliferation.50, 51 Furthermore, it has been suggested that HCV infection may upregulate expression of MHC antigens and adhesion molecules on biliary epithelium.52 Unfortunately, studies that carefully dissect the immune response in HCV-infected allografts are scarce. Zekry et al.53 have compared HCV-acute rejectors and non-HCV acute rejectors to normal, nondiseased livers. HCV-acute rejectors have significantly higher intrahepatic expression of both TH1 and TH2 cytokines (IFN-γ, interleukin 2, tumor necrosis factor alfa, interleukin 10, and interleukin 4) compared to the normal, nondiseased liver. However, non-HCV acute rejectors differed from nondiseased livers only in increased intrahepatic tumor necrosis factor alfa levels.53 These data suggest, albeit indirectly, that HCV infection of an allograft results in increased intrahepatic cytokine concentrations consistent with an immunologically active milieu. Clearly, careful studies that elucidate the specificity of the immune response— antiviral, antidonor, or both, and, if both, their relative proportion—may be critically important for accurate diagnosis in the difficult early, posttransplant period.
In conclusion, irrespective of the accuracy of the EAR diagnosis during our study period, the fact remains that the diagnosis of EAR routinely precipitates treatment of the recipient with bolus corticosteroids with or without a course of antilymphocyte therapy. While historically, the diagnosis and treatment of EAR was considered at least benign if not even beneficial, recent data suggests that treated rejection is associated with more severe HCV recurrence.9, 10, 12, 21, 22, 54 Since the transplantation community has become increasingly aware of the diagnostic and therapeutic dilemma posed by acute rejection within the context of recurrent HCV, many transplant centers are utilizing clinical and virologic data in addition to histology in determining a management plan.22, 32, 55, 56 Others have gone further and decided to reserve antirejection therapy only for those with definitive and severe rejection.22 The impact of these substantial changes in practice patterns is, as yet, unknown. We anxiously await studies that indicate whether a raised threshold for instituting antirejection treatment favorably alters the natural history of recurrent HCV.
- 49Reliability and predictive value of the National Institute of Diabetes and Digestive and Kidney Diseases Liver Transplantation Database nomenclature and grading system for cellular rejection of liver allografts. Hepatology 1995; 21: 408–416., , , , , , et al.
- 52Role of Liver Biopsy in Post-Transplant Hepatitis C. In: BaconBR, GoodmanZD, BruntEM, eds. Liver disease in the 21st century: clinico-pathologic correlates. Virginia: The American Association for the Study of Liver Diseases 2003 Liver Meeting Postgraduate course; 2003: 111–118..
Supplementary material for this article can be found on the LIVER TRANSPLANTATION website ( http://www.interscience.wiley.com/jpages/1527–6465/suppmat )
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.