Any immunosuppressive regimens after transplantation increases the overall risk of cancer.1 After orthotopic liver transplantation (OLT) performed in patients presenting with a hepatocellular carcinoma (HCC), recurrent malignancy occurs between 3% and up to 52%, depending on the stage of the disease at the time of OLT.2 The majority of recurrent carcinomas are detected more than 1 year after OLT with a median survival of less than 10 months following the diagnosis of the recurrence.3 Very few options are available to treat those recurrences with few successes reported with liver resection4, 5 and retransplantation.6
A new class of immunosuppressive drugs, inhibitors of the mammalian target of rapamycin (mTOR), exhibit simultaneously antiproliferative and immunosuppressive effects.7, 8 The mTOR protein is a central regulator of cell metabolism, proliferation, and survival9 (Fig. 1). It also modulates the innate and adaptive immune responses by the regulation of dendritic cells and T cells.10 Drugs like sirolimus, also known as rapamycin (trademarked as Rapamune by Wyeth), or everolimus (trademarked as Certican by Novartis) may have the potential to solve the two-sided Janus face of immunosuppression, with one face inducing tolerance of the graft and the other face causing cancer.
Is it possible to break the paradigm of effectively preventing rejection with an immunosuppressive regimen and its inherent risk of new or recurrent tumors? In this issue of HEPATOLOGY, Toso, Kneteman, and colleagues11 look at the impact of different immunosuppressive regimens on the long-term survival of recipients originally presenting with an HCC. To address this issue, these authors used a large cohort of patients taken from the Scientific Registry of Transplant Recipients (SRTR) database, which offers information about all candidates listed for OLT, donors, and recipients in the United States (www.ustransplant.org). The study included 2491 recipients originally presenting with an HCC and a control group of 12,167 patients including those who underwent transplantation for diseases other than HCC. The main message of this study is that sirolimus-based immunosuppression increased long-term survival in recipients presenting with an HCC. These findings were observed in both the univariate and multivariate analysis.
Six years ago, Kneteman et al. already reported excellent outcome on 40 OLT recipients with HCC who were subjected to a sirolimus-based immunosuppression.12 The 4-year patient survival was 81% with a recurrence rate of 5% only in patients meeting the Milan criteria (one tumor >5 cm or up to three tumors <3 cm in size).13 In patients with an HCC load extending the Milan criteria, the 4-year survival was still 77% with an incidence of tumor recurrence of 19%. In a follow-up study, adding 30 patients with a median follow-up more than 4 years,14 6% of the patients meeting the Milan criteria and only 17% of the patients beyond the Milan criteria developed a recurrence of HCC. Tumor-free 4-year survival was about 75 % regardless of whether the patients met the Milan criteria. Several recent retrospective studies have suggested results similar to the Edmonton, Canada, group of Toso, Kneteman, and colleagues.15-17
In their current study, 5-year survival of OLT recipients presenting with an HCC was 83% when receiving sirolimus compared to 69% in those treated with another regimen. Although the data suggesting an almost 15% improvement in 5-year survival in patients treated with an mTOR inhibition regimen are impressive and potentially of high clinical relevance, we must point out some pitfalls that are inherent to all retrospective or uncontrolled studies.18 In the HCC group, 109 patients treated with sirolimus were compared to 2382 patients treated with a different immunosuppressive protocol. Although at first glance, the HCC group on sirolimus appears comparable to the group not treated with sirolimus, important information is lacking regarding tumor staging and biology. For example, only information about the mean total tumor volume and the percentage of tumors exceeding >115 cm3 is available. The proportion of patients meeting the Milan criteria in each group is unknown. The data on alpha-fetoprotein (AFP) levels disclosed a trend toward higher levels in the HCC group not receiving sirolimus. In addition, AFP levels were not known in about one-third of the population studied. Unfortunately, data about the histological differentiation grade and vascular invasion is completely missing. These shortcomings, including likewise a higher proportion of patients meeting the Milan criteria in the HCC group receiving sirolimus and more aggressive tumor biology in the HCC group not receiving sirolimus, could easily confound the univariate and multivariate analysis. Several studies have shown that tumor staging as well as grading and vascular invasion are the most important factors affecting tumor recurrence in patients receiving OLT for treatment of HCC.2-4, 6, 19
The authors' statement about “the unique post-transplant effects of sirolimus in HCC patients”, i.e., that mTOR inhibition improves survival because it prevents tumor recurrence in in vitro and in vivo models,8, 20 may be plausible but cannot be proven with the presented data. The study did not assess the main outcome of interest, the recurrence-free survival, because the focus was on overall survival. There is also no information about the median follow-up after OLT. Therefore, the SRTR data may underestimate the HCC recurrence rate due to underreporting, and the unknown median follow-up additionally limits the tempting conclusion that sirolimus has oncological benefits in the HCC population.
A second observation was that induction therapy with anti-CD25 was associated with a 74% 5-year survival, whereas survival was reduced to 68% in those receiving a regimen omitting anti-CD25 antibodies. Anti-CD25 antibodies, such as daclizumab (trademarked as Zenapax by Roche) or basiliximab (trademarked as Simulect by Novartis), specifically target the alpha-chain of interleukin-2 receptors, CD25, which are expressed on T cells. These monoclonal antibodies are used as induction therapy for OLT to prevent acute rejection,21 and 11% of the patients were treated with these antibodies in this study.
The induction therapy with anti-CD25 antibodies surprisingly appears to offer a survival benefit in the HCC group, a finding that may tempt us to support the suspicion of a similar selection bias as for sirolimus. Data about graft rejection is lacking and would have been of interest to better interpret the influence of the degree of immunosuppression on the recurrence of HCC after OLT. Nevertheless, this data raises the question whether the depletion of CD25 could be another strategy to solve the Janus face of immunosuppression. This justifies further studies.
Taken together, the study provides important clinical information about a representative and large cohort of patients after OLT. This data is of great value to design definitive randomized controlled trials that test sirolimus in the HCC and non-HCC population. Because several currently recruiting randomized trials have already picked up the message of the study (www.clinicaltrials.gov), we are awaiting the final answer.