Osteopontin expression predicts overall survival after liver transplantation for hepatocellular carcinoma in patients beyond the Milan criteria

Background & Aims

Microarray data showed that osteopontin overexpression predicts early HCC-recurrence after liver resection. Osteopontin (OPN) expression could serve as a predictor of HCC-recurrence after OLT.

Methods

Osteopontin expression was investigated immunohistochemically in a unique population of 125 HCC-patients undergoing OLT between 1982 and 2002, including 81 patients (65%) outside the Milan criteria. Multivariate analysis of factors associated with median overall survival (OS) and time to recurrence (TTR) was performed.

Results

Osteopontin was expressed in 40/125 (32%) of the HCCs. Overall survival post-OLT at 1, 2, 3, 5 years was 77%, 62%, 52%, and 43% (median survival 37 months). Overall survival was significantly longer without expression of OPN (p <0.05; (median OS: 56 vs. 23 months). The same was true for median TTR (p = 0.008). Outside Milan criteria, patients without OPN-expression had better prognosis (median OS: 37.8 vs. 19.2 months, p = 0.003). Tumor recurrence in patients transplanted outside Milan criteria occurred in 43% (23 of 54) of patients without and 70% (19 of 27, p = 0.018) of patients with OPN-expression after a median TTR of 83.5 vs. 13.9 months. On multivariate analysis, vascular invasion and OPN-expression were independently associated with OS and TTR in HCC-patients after OLT.

Conclusions

Immunohistochemically detectable Osteopontin in HCC is an independent predictor of tumor recurrence and survival in patients beyond Milan criteria undergoing OLT.

Introduction

Hepatocellular carcinoma (HCC) is the most common primary liver malignancy worldwide, being responsible for more than 600,000 deaths annually [1]. HCC occurs mainly in chronic liver diseases resulting from chronic hepatitis virus infection, alcoholic liver disease, or other etiologies leading to cirrhosis. In spite of remarkable progresses in diagnostic and therapeutic techniques [2] patients with HCC often have a dismal prognosis [3], [4] attributable to a number of factors. The poor prognosis of HCC is often the result of late diagnosis or a high rate of recurrence after surgery due to intra-hepatic metastases, while extra-hepatic metastases are less common [5], [6].

Orthotopic liver transplantation (OLT) is the most effective therapy for patients with HCC and offers potential cure of HCC in selected patients. The initial experience with OLT for HCC was limited to those with extensive tumors and was associated with dismal outcome as a result of aggressive tumor recurrence after OLT [7], [8]. In 1996, Mazzaferro et al. confirmed that long-term survival can be achieved with transplantation by carefully selecting patients who satisfy specific tumor criteria [9]. According to these criteria, patients who suffer from one tumor nodule <5 cm or three nodules <3 cm are suitable for OLT. The Milan criteria for HCC have been widely and successfully used for selecting patients for OLT. However, pathobiologic features of the tumors have not been incorporated into the selection criteria, potentially excluding some patients with larger tumors who might benefit from OLT [10].

Tumor characteristics predicting post-OLT recurrence remain a matter of debate. Several different studies have reported that tumor size, positive lymph nodes, poorly differentiated histological features, and bilobular disease show varying correlations with tumor recurrence. Only vascular invasion predominates as a clear predictor of poor outcome [11], [12], [13]. In recent years, much effort has been made to characterize the molecular biology of HCC [14] and correlate it to the outcome. However, specific predictive markers for recurrence and metastasis after OLT are still missing.

A recent report using gene array technology showed that osteopontin (OPN) mRNA overexpression predicts early recurrence of HCC in Chinese patients after hepatic resection. OPN overexpression could thus serve as an unfavorable prognostic factor in this population and a useful marker for predicting early recurrence of HCC [15], [16]. OPN plasma levels have been reported to be elevated in HCC, and has superior diagnostic accuracy as a tumor marker compared to AFP or PIVKAII [17]. Preoperative OPN plasma levels predict tumor recurrence and survival in HBV-associated HCC-patients undergoing tumor-resection [18].

Osteopontin (OPN) is a secreted, 44–66-kDa adhesive glycophosphoprotein that has been involved in both normal conditions, such as bone development and immune system regulation, and in pathologic processes, such as inflammation, cell transformation, tumor invasiveness, and tumor metastasis [19], [20], [21]. OPN is expressed in a number of different cell types, including osteoclasts, arterial smooth muscle cells, macrophages, T lymphocytes, and various types of epithelial cells. In normal liver tissue, OPN is expressed in bile duct epithelium, stellate cells, and Kupffer cells [20], [22] but not in normal hepatocytes [17].

In the last decade, several studies have defined an important role for OPN in carcinogenesis and metastasis. OPN was first implicated in malignancy by in vitro studies detecting increased levels of OPN-expression after cell transformation [23]. It was also observed that tumor cells with high metastatic potential had increased OPN-expression [24]. The expression of OPN associated with tumorigenesis and tumor progression was also confirmed in experimental models and clinical studies of various human cancers, including breast, colon, stomach, kidney, and lung [25], [26], [27], [28], [29], [30]. OPN promotes VEGF-dependent tumor growth and neoangiogenesis [31]. Recently, the whole SIBLINGs-family of proteins, to which OPN belongs, has been implicated in various stages of tumor development and progression [32].

The prognostic role of OPN-expression for tumor recurrence in HCC has not been investigated in Western patients without Hepatitis B-associated liver disease and in patients with HCC undergoing liver transplantation. The prognostic value of OPN-expression in patients transplanted outside the criteria would be of particular interest for expanding the criteria for patient selection. However, this is a difficult task since only a few patients are knowingly transplanted outside the criteria. We provide here data of a unique patient population with two-thirds of the liver transplant recipients transplanted outside the Milan criteria.