Molecular markers and hepatocellular carcinoma: lending a helping hand in liver transplantation?

Currently, hepatocellular carcinoma (HCC) is the third most common cause of cancer death worldwide; 82% of the cases occur in developing countries, with 55% occurring in China [1]. Liver transplantation (LT), a potentially curative therapeutic modality of HCC, allows radical extirpation of the cancer and restores normal liver function. There is increasing evidence that, in the long term, LT will be the best therapeutic option for cirrhosis-associated HCC [2]. Unfortunately, the recurrence rate is high in HCC patients who undergo LT, which markedly reduces the long-term and medium-term survival of HCC patients [3]. Therefore, risk estimation of post-transplant tumor recurrence is an essential element in selecting a subset of HCC patients for LT.

In clinical practice, this risk estimation is based on the number and size of the nodules, and microscopic/macroscopic vascular invasion; and selection criteria for LT have been established to select HCC patients according to these clinical features. The Milan criteria proposed by Mazzaferro and his colleagues – a single lesion of 5 cm or smaller in diameter, or three lesions each smaller than 3 cm in diameter, and without macroscopic vascular invasion – have been accepted worldwide for identifying candidates with good prognosis and low recurrence rates [4]. The Milan criteria were chosen in an attempt to minimize the difference in outcomes between transplanted HCC patients and those who did not have HCC, and these criteria have been adopted for organ allocation in the USA and in Europe [5,6]. Despite their proven utility, many recent data have suggested that these criteria are too restrictive, and many patients with tumor burdens outside the Milan criteria may potentially benefit from LT. Moreover, limiting LT to HCC patients within the Milan criteria may be unjust for those outside of the criteria who could have a favorable outcome with LT.

Several expansions of the Milan criteria have been proposed. Yao and coworkers at the University of California, San Francisco (UCSF; CA, USA) reported that LT candidates with a solitary tumor of 6.5 cm in diameter or smaller, or three or fewer nodules with the largest lesion no larger than 4.5 cm in diameter, and a total tumor diameter no larger than 8 cm (the so-called UCSF criteria) achieved excellent results following LT that were not different from those for HCC patients within the Milan criteria [7]. Although the UCSF criteria have been validated by several groups [8], these criteria have not been accepted unanimously, which may be related to the results of a large multicenter study that suggested the UCSF criteria to be associated with a 5-year survival rate of less than 50% when applied to preoperative evaluation [9]. In addition, the UCSF criteria allow LT for large tumors, which have been proven to be associated with post-transplant metastatic recurrence [10]. Lastly, both the Milan and the UCSF criteria exclude those patients with more than three nodules, who may obtain good outcomes if the tumors are of a reasonable size. Toso and his colleagues proposed an expansion of the current allocation criteria for LT for the treatment of HCC. They performed an overview of 6478 adult recipients and found that using a composite patient selection score, which combined levels of a-fetoprotein – a well-demonstrated and diagnostic protein in HCC – and the total tumor volume, was more effective for the prediction of post-transplant patient survival for candidates with HCC than some other staging systems, including the Milan criteria [11].

The limitation of using imaging to assess tumor size and vascular invasion, which is the basis of the Milan and the UCSF criteria, has led to the realization that preoperatively available clinical and radiological variables are imperfect surrogates for predicting HCC metastatic potential. With our deeper understanding of tumors and developing technology, increasing attention is being paid to effective prognostic molecular markers for selecting HCC patients undergoing LT. Until now, many molecules, including genes, mRNAs and proteins, have been confirmed to be useful predictors of the post-transplant recurrence of HCC patients, for example calpain small subunit-1 (Capn4) [12], retinoblastoma gene [13] and albumin mRNA [14].

Hepatocellular carcinoma metastatic recurrence is a multistage process involving the acquisition of aggressive phenotypes, reduction of intercellular adhesion in the primary tumor, degradation of the extracellular matrix (both in the primary and metastatic sites), proliferation and angiogenesis of metastases. The molecular mechanisms associated with post-transplant HCC recurrence are quite different to those of other potentially curative therapies, such as surgical resection or radioablation. First, the abnormal liver is replaced by a normal one in the treatment of HCC. Thus, metastatic cancer cells originate from the blood or other organs but not from the liver itself. Second, the application of an immunosuppressant, a good blood supply for the liver, and a noncirrhotic liver will increase the risk of post-transplant tumor metastatic recurrence in LT. Therefore, a system should be established to elucidate the mechanisms of HCC recurrence after LT.

More recently, investigators have utilized genomic technologies to identify highly aggressive tumors with poor prognosis after transplantation and to select patients who, despite having HCC outside of the Milan criteria, are associated with acceptable disease-free survival rates. Schwartz and colleagues analyzed the allelic imbalance (AI) of 18 microsatellites (MS) in 70 patients transplanted for HCC. In this study, a fractional AI (FAI) rate, defined as the number of MS demonstrating AI divided by the total number of informative MS, was calculated for each patient. They found that the AI in nine out of the 18 MS correlated with recurrence. An FAI of more than 0.27 and macrovascular invasion were the two independent predictors of recurrence in patients with tumors that were outside of the Milan criteria; the probability of recurrence at 5 years was 85% with a FAI rate of 0.27 or more versus 10% when the FAI rate was less than 0.27. This study emphasized the utility of the FAI rate in patients outside of the Milan criteria but with a favorable outcome of LT [15].

Bai and colleagues recently used a comparative proteomic analysis to identify the molecules underlying tumor invasion and metastasis after LT for HCC [12]. They discovered 52 proteins that were differentially expressed by at least twofold in the recurrence group compared with the nonrecurrence group. Among them, Capn4 was dramatically overexpressed in metastatic tissues from HCC patients undergoing LT, whereas it was downregulated in the nonmetastatic tissues. Furthermore, a tissue microarray in 192 LT cases revealed that Capn4 was correlated significantly with the invasive phenotype of HCC, and univariate and multivariate analyses indicated that Capn4 was an independent prognostic factor for post-transplant tumor recurrence and survival of HCC patients. Shi et al. found that the X-linked inhibitor-of-apoptosis protein (XIAP) could be used to predict HCC recurrence and patient survival after treatment, and may be a molecular target subject to the intervention of metastasis and recurrence [16]. It is noteworthy that XIAP expression could provide additional prognostic values outside the Milan criteria in advanced cases, owing to the fact that, among these patients, those with increased XIAP expression had significantly shorter disease-free survival. The main advantage of molecular markers such as Capn4 and XIAP is that, unlike the Milan and UCSF criteria, they do not include a rigid limit on the number of tumors. Such a limit prevents transplantation for patients with more than three lesions, even if the lesions are very small and would be associated with a good post-transplant outcome. However, it is important to recognize that the data analyzed in those studies were based on primary HCC samples, which are not readily available in the preoperative setting and particularly at listing. Therefore, there remain significant challenges in the translation of these molecular biomarkers into practical clinical application.

Predictive markers allow us to predict post-transplant HCC metastatic recurrence more accurately and to make early diagnoses; they also offer more and better molecular targets for therapy. There are many targeted agents being tested in Phase II and III clinical trials for the treatment of HCC patients. For example, the multiple-molecular-targeted drug sorafenib has shown promising therapeutic effects and its efficacy was confirmed in a Phase III trial in HCC [17]. Sorafenib blocks tyrosine protein kinase receptors, such as VEGF receptor 2, VEGF receptor 3, and PDGF receptor, which are associated with angiogenesis [18,19]. Phase II and III clinical trials of oral sorafenib for advanced HCC have been completed [17,20]. Capn4 and XIAP are both potential targets for therapy, and this provides us with a new direction for further research into these molecules.

Molecular-targeted therapy has gained some clinical accomplishments, which boost our ability to manage HCC. Although most of the molecular-targeted drugs have not yet been used in the area of LT for HCC, surgery and molecular-targeted therapy theoretically support each other, and their combination may be a breakthrough milestone.

Financial & competing interests disclosure

This work was supported by grants from the National “11th Five-Year” Key Technology R&D Program of China under contract no. 2006BAI02A04, the State 863 High Technology R&D Project of China under contract no. 2007AA02Z479, and Specialized Research Fund for the Doctoral Program of Higher Education under contract no. 20070246112. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.