If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Experience with sirolimus (SRL)-based immunosuppression following orthotopic liver transplantation (OLT), which was initially approved for use in renal transplantation by the Food and Drug Administration in 1999, is rapidly accumulating.1 In an effort to reduce calcineurin inhibitor (CNI) toxicity, SRL-based therapy is increasingly being applied. Preliminary observations in the late 1990s suggested that SRL is well tolerated and effective following OLT.2 SRL in combination with CNIs may reduce the incidence of acute cellular rejection (ACR)3 and dramatically lower the required levels of either tacrolimus or cyclosporine.4 Although SRL is associated with an improvement in renal function,5 some concerns have been raised over the association of SRL-based therapy and hepatic artery thrombosis (HAT) following OLT.3, 6
Tumor recurrence following OLT for end-stage liver disease (ESLD) and concomitant hepatocellular carcinoma (HCC) occurs in approximately 20% of recipients.7 Numerous experimental reports have documented significant antineoplastic effects of SRL in a variety of animal models.8–10 However, clinical observations in this select group are not exactly parallel. Although several retrospective series suggest comparable long-term outcomes with SRL-based immunosuppression in recipients with HCC,11, 12 a recent prospective initiative noted a high rate of incisional hernias and dyslipidemia.13 Despite the expanding application of SRL therapy post-OLT, it is unclear whether SRL has a direct oncological benefit. The current study sought to quantify long-term survival following OLT for ESLD and HCC and to document the potential influence of SRL on cancer recurrence and posttransplant complications.
ACR, acute cellular rejection; CAD, cadaveric; CIT, cold ischemic time; CNI, calcineurin inhibitor; Cr, creatinine; CT, computed tomography; ESLD, end-stage liver disease; F, female; HAT, hepatic artery thrombosis; HCC, hepatocellular carcinoma; LDLT, living donor liver transplant; M, male; MMF, mycophenolate mofetil; MRI, magnetic resonance imaging; OLT, orthotopic liver transplantation; SRL, sirolimus; UCHSC, University of Colorado Health Sciences Center; WIT, warm ischemic time.
PATIENTS AND METHODS
From January 2000 to June 2007, 97 patients underwent OLT for end-stage disease and concomitant HCC at the University of Colorado Health Sciences Center (UCHSC). Of those, 45 patients received SRL, in conjunction with CNIs, as a component of their primary immunosuppression regimen post-OLT. Conversely, 52 patients received the standard immunosuppression regimen including CNIs (cyclosporine A or tacrolimus), mycophenolate mofetil (MMF), and corticosteroids. Twenty patients who were converted to SRL-based therapy post-OLT, secondary to CNI toxicity, were excluded from this study. A retrospective review was conducted with the approval of the UCHSC Institutional Review Board. Sources of data for the patients in this study included the University of Colorado Transplant Database and individual patient medical records.
Diagnosis and Patient Evaluation
Preoperative diagnosis of all patients included a history and physical examination as well as laboratory studies. Preoperative diagnosis of HCC was based on abdominal computed tomography (CT) or magnetic resonance imaging (MRI) of the abdomen. Our institution does not routinely perform liver biopsies for diagnosis of malignancy. Extrahepatic metastasis was excluded on the basis of chest/abdominal CT, MRI, and bone scintigraphy prior to OLT. Final confirmation of the diagnosis of HCC was made by pathological examination of the explanted liver. Patients who demonstrated metastatic disease, including lymph node involvement or vascular invasion of the portal vein, hepatic vein, or vena cava identified by CT or MRI, were not considered for transplant. Cancer recurrence was diagnosed via CT, MRI, or positron emission tomography and routine chest X-ray.
Patients receiving SRL as a primary component of their post-OLT immunosuppression (n = 45) were started on 2 mg/day within 2 weeks of surgery. Patients transplanted between January 2000 and July 2002 were given a bolus dose of 6 mg at day 0 and started on 2 mg/day thereafter. All patients in this study received CNIs, with or without SRL, as part of their primary immunosuppression regimen. Until July 2001, recipients were treated with cyclosporine A or tacrolimus plus a 3-day corticosteroid taper (methylprednisolone, 1 g, day 0; 0.5 g, day 1; and 0.5 g, day). After July 2001, cyclosporine A was discontinued, and patients received tacrolimus exclusively, followed by the 3-day steroid taper. Patients receiving SRL did not receive MMF. Patients not receiving SRL were treated with MMF. After July 2002, patients transplanted with HCC and a post-OLT creatinine (Cr) over 1.5 were generally treated with SRL. Conversely, those with a Cr below 1.5 were treated with MMF. Major complications in the first 12 months post-OLT were evaluated. Major complications were defined as an episode of documented or presumed ACR, steroid-resistant rejection, HAT, portal vein thrombosis, biliary stricture/leak, incisional hernia, and wound infection and/or delayed healing.
Post-OLT explants were serially sectioned and examined by an experienced UCHSC faculty pathologist. Recorded characteristics included the following: anatomic location, tumor size by maximum diameter (cm) of each lesion, anatomic distribution, and presence of gross and/or microscopic vascular invasion. Suspicious or enlarged lymph nodes were also evaluated for metastatic disease. Tumors were graded by degree of differentiation as follows: well-differentiated (grade 1), moderately differentiated (grade 2), poorly differentiated (grade 3), or necrotic.14 Tumor stage was determined by the revised pathologic tumor-node-metastasis (PTNM) classification.15
The time-dependent outcomes of patient survival and HCC recurrence–free survival were compared with the log rank test when confounding factors were being ignored, and they were compared with a Cox proportional hazards model (multivariate) to control or adjust for potential confounding factors. Adjusted risk ratios (hazard ratios) were computed on the basis of the Cox model, and potential confounding factors that were significant with a liberal P < 0.15 criterion were retained for use in the adjustment. Survival curves were computed with the Kaplan-Meier method. The Poisson rate homogeneity test was used when P values were computed for comparing rates of acute graft rejection and major complications within the first 12 months post-OLT. In an effort to control for differential follow-up, rates were compared instead of the simple proportion (Fisher's test) with respect to graft rejection or major complications because not all patients had a full 12 months of follow-up. Because the baseline Cr varies patient to patient, the nonparametric Wilcoxon rank sum test was used to compute P values for comparing means and medians of continuous variables including the baseline Cr and change in Cr from baseline at 6 and 12 months. Fisher's exact test was used to compare non–time-dependent proportions for binary outcomes. For other nominal outcomes, P values were computed with the chi-square test.
Histopathology of the Explanted Liver
The majority of patients in this study were stage I (51%) by explant evaluation (Table 1). Twenty-eight percent were stage II, and 21% were stage III. No patients were transplanted with stage IV disease during this time period. Similarly, most recipients had well-differentiated lesions (50%), with 32% being moderately differentiated and only 5% being poorly differentiated. Approximately 12% were necrotic and not able to be evaluated secondary to pre-OLT ablative therapy. Thirty-three percent of the lesions were multifocal. Interestingly, only 12% exhibited vascular invasion. The preoperative Child's classification was also determined with 36 patients to be Child's C.
Table 1. Patient and Tumor Characteristics (n = 97)
This study includes 83 males and 14 females with ESLD and HCC (Table 1). Not surprisingly, the majority of patients in this cohort suffered from hepatitis C (n = 66). Additional diagnosis included hepatitis B (n = 10), autoimmune hepatitis (n = 2), hemochromatosis (n = 1), nonalcoholic steatohepatitis (n = 2), cryptogenic cirrhosis (n = 8), alcoholism (n = 7), and primary sclerosis cholangitis (n = 1). Comparing the 2 treatment groups, we found that patients receiving SRL had a mean age of 54 ± 6.1 versus 54 ± 6.6 for those treated with CNIs exclusively (P = 0.73). The mean follow-up times were 36 and 25 months in the SRL and CNI groups, respectively. For the cohort overall, 80 recipients received cadaveric grafts, with 17 grafts derived from a living donor. There was no difference in baseline Cr (1.3 versus 1.1, P = 0.31) or warm ischemic time [WIT (minutes); 35 versus 36, P = 0.65] between the treatment groups. However, cold ischemic time [CIT (hours)] was significantly higher in patients treated with CNIs only (7.8 versus 6.3, P = 0.009).
Cumulatively, recurrence of HCC occurred in 12 patients (12.4%). At the writing of this article, 2 patients remain alive with lung metastasis. The remaining 10 patients are dead, all as a direct result of HCC recurrence. Sites of tumor recurrence include the liver allograft (6), lungs (4), brain (1), and omentum (1). Overall, 9 patients who suffered an HCC recurrence had stage III lesions by explant. Two patients were stage I, and 1 patient was stage II. Fifty percent of the lesions demonstrated vascular invasion on final evaluation (6/12).
With respect to group comparison, 3 patients with tumor recurrence were treated with SRL. All 3 of these recipients had stage III lesions, with 2 of 3 demonstrating vascular invasion. Conversely, 9 patients were treated with CNIs, of which 6 are stage III. Four of these 9 patients had vascular invasion. None of the recipients with HCC recurrence were treated with cyclosporine A as a component of the primary immunosuppressive regimen. At the time of last follow-up, 7 patients in the SRL treatment are dead. Of the remaining 38 recipients, 23 are still on SRL-based immunosuppression (60%).
Overall, 21 deaths were documented in this cohort during the study period. Of those, 10 are attributed to recurrent HCC. Conversely, 11 patients died of non-HCC causes including sepsis (4), lung cancer (2), HCV recurrence (1), congestive heart failure (1), HAT (1), primary nonfunction (1), and cardiac arrest (1). By group comparison, 3 non–tumor-related deaths occurred in the SRL group versus 7 in the CNI group. Two of the 3 patients treated with SRL died in the first 24 months. However, 6 recipients in the CNI treatment group died within the first 2 years of transplant in the absence of tumor recurrence. In fact, all 6 patients died in the first 90 days postoperatively, 3 from sepsis alone. One recipient in the CNI treatment group died of lung cancer 4 years post-OLT.
Survival Analysis and Post-OLT Complications
Overall survival at 1 and 5 years post-OLT, for patients treated with SRL, is 95.5% and 80%, respectively. Conversely, survival in patients treated with CNIs exclusively at the same time intervals is 83% and 62%. The mortality risk ratio (SRL/CNIs) is 0.672 (95% confidence interval = 0.417 to 1.052, P = 0.087). Similarly, HCC recurrence–free survival at 1 and 5 years post-OLT in patients treated with SRL is 93% and 78.8% (Fig. 1). For those in the CNI group, HCC recurrence–free survival at the same time points is 75% and 54%. The mortality/HCC recurrence risk ratio (SRL/CNIs) is 0.622 (95% confidence interval = 0.385 to 0.954, P = 0.03).
To be more certain that the aforementioned risk ratios are not due to a lack of comparability, Cox model–based adjusted risk ratios (SRL/CNI) were computed. The 12 potential covariates considered for adjustment were gender, age, class, largest tumor diameter, nodularity, differentiation, vascular invasion (yes/no), pTNM stage, graft type (cadaveric or living donor liver transplant), baseline Cr, WIT, and CIT. For patient survival/mortality, gender, age, nodularity, vascular invasion, pTNM stage, and WIT were simultaneously significant at P < 0.15 and were therefore used for adjustment. For HCC recurrence–free survival/recurrence or death, gender, nodularity, differentiation (borderline effect, P = 0.16), vascular invasion, pTNM stage, and baseline Cr (borderline effect, P = 0.16) were simultaneously significant at P < 0.15 and were used for adjustment. The model-based adjusted mortality risk ratio (SRL/CNIs) is 0.667 (P = 0.11), and the adjusted mortality/HCC recurrence risk ratio is 0.503 (P = 0.006), similar to the unadjusted rates.
Short-Term Complications and Renal Function
Overall, 19 of 97 patients had at least 1 documented, or presumed, episode of ACR in the first 12 month post-OLT (Table 2). By treatment group comparison, 9 patients were treated for ACR in the SRL group (20%). Ten patients in the CNI-treated group suffered at least 1 episode of ACR (19.6%). The corresponding graft rejection rates per 100 person-months of follow-up are 1.8 and 2.1, respectively (P = 0.77). Additionally, 11% treated with SRL had at least 1 major complication, aside from ACR, in the first 12 months post-OLT. Thirteen percent of patients treated with CNIs suffered at least 1 major complication. The corresponding complication rates per 100 person-months are 1.0 and 1.44, respectively (P = 0.53). The incidence of HAT and wound complications was no different between treatment groups.
Baseline Cr levels were comparable at the time of OLT. The mean Cr level in the SRL group is 1.36 versus 1.25 mg/dL in the CNI group (P = 0.32). However, Cr levels fluctuated significantly over 6 and 12 months post-OLT (Table 3). In patients treated with SRL, the mean Cr level decreased to 1.14 mg/dL at 6 months post-OLT, but for patients treated with CNIs, it modestly increased to 1.31 mg/dL (a mean change from baseline of −0.22 versus +0.61 mg/dL, P = 0.0021). Similarly, the mean Cr level in the SRL group at 12 months is 1.08 versus 1.6 mg/dL in the CNI group (a mean change from baseline of −0.28 versus +0.35 mg/dL, P < 0.0001).
Although numerous experimental initiatives have repeatedly demonstrated an oncological benefit to SRL therapy, the cumulative clinical experience with HCC is modest at best.16 Kneteman and colleagues11 recently reported a prospective pilot study of 40 patients with HCC treated with an SRL-based protocol post-transplant. Categorizing recipients within the Milan criteria or beyond, the authors concluded that both 1-year survival and 4-year survival were not significantly different between groups. Furthermore, 24 of 35 surviving recipients were on SRL monotherapy at last follow-up. A large retrospective analysis from the University of Southern California compared SRL monotherapy to SRL plus CNIs.12 No differences were observed with respect to 12-month survival or rate of rejection. However, although the need for dialysis and average serum Cr were higher in the SRL treatment group at the time of transplant, these rates were comparable at 30 days and 1 year, respectively. Finally, a prospective cohort including 70 patients with HCC noted no difference in short-term tumor-free survival in patients within or beyond the Milan parameters.13 All patients received SRL-based therapy post-OLT. Although cancer recurrence overall was reasonably low (11%), nearly 50% suffered from delayed wound healing or incisional hernia.
Preliminary efforts suggest that SRL-based therapy may be associated with a substantial reduction in renal toxicity and ACR. However, some serious concerns have evolved. In addition to delayed wound healing, infection, and incisional hernias, several studies have reported a higher incidence of arterial thrombosis in patients treated with SRL.3, 6 Although these observations are indeed of concern, this has not been our institutional experience. Previous analysis of OLT recipients treated with SRL showed comparable rates of vascular and wound complications.17, 18 Specifically, of 170 patients treated with SRL-based immunosuppression, the prevalence of wound complications was 12.4% versus 13.9% in historic controls. Furthermore, the prevalence of HAT was 5.3% in recipients treated with SRL compared to 8.3% in the control group.
The current study makes several observations. Overall, SRL-based immunosuppression may have a positive influence on cancer recurrence–free survival following OLT. Undergoing transplantation in the same time interval, the 2 treatment groups are similar with respect to age, gender, and multiple pathologic variables. CIT is the only significantly different factor and likely does not completely account for the observed survival differences. This may reflect the higher number of living donors in the SRL group with a shorter CIT. The incidence of cancer recurrence is much higher in the CNI-treated cohort. This may account for the dramatic difference in long-term outcome as 10 of the 12 patients that suffered a recurrence are dead as a direct result. No differences in the incidence of vascular complications, acute rejection, wound infection, or delayed healing were noted. There was 1 case of HAT in each group. Finally, patients treated with SRL experienced a modest improvement in renal function at both 6 and 12 months post-OLT.
Although these data are encouraging, we acknowledge several limitations. Previously, retrospective analysis of our institutional experience with OLT for HCC failed to show a survival benefit in patients treated with SRL-based therapy.18 However, the previous group was heterogeneous in that it included 20 patients converted to SRL at various time points post-transplant. The current study excludes that cohort and updates our experience with SRL as a primary immunosuppressive agent. Although this group is a larger and more uniform sample, it is still relatively small. Finally, these data appear to corroborate the nephron-sparing effects of SRL shown by other groups.5 Unfortunately, we do not have documented changes in glomerular filtration for each recipient. As such, we have attempted to quantify alterations in renal function exclusively by serum Cr, which is clearly a less optimal approach.
At present, the clinical experience with SRL-based therapy suggests that this agent is generally safe and reasonably effective in the post-OLT setting. However, the degree of oncological benefit, if any, is unclear. Cumulatively, these data suggest a potential survival benefit with SRL-based therapy in patients undergoing OLT for end-stage disease and concomitant malignancy. Furthermore, no differences were observed with respect to rejection or major complications including arterial thrombosis and incisional concerns.
The authors acknowledge Dr. Jeff Gornbein, Dr.P.H., for his assistance with the statistical analysis.