Results of liver transplantation

Hepatocellular carcinoma (HCC) is the most frequently occurring primary malignant tumor1 and is today the fifth most common cancer in the world and the third cause of cancer-related death. Most patients with HCC have cirrhosis, which is the major clinical risk factor for hepatic cancer. It is primarily caused by infection with either hepatitis B or C virus, with a 5-year cumulative incidence ranging 15-20%.2 The clinical course of HCC and the survival rates of patients depend not just on the stage of the tumor but also on the underlying disease at the time of diagnosis and on the response to treatment. Liver resection (LR) and transplantation, together with locoregional ablation therapies, are currently the only potential curative treatment options available for HCC. To date, there have been no prospective studies comparing LR with orthotopic liver transplantation (OLT). Randomized controlled trials have not been performed because most patients evaluated for curative surgical procedures manifest distinctive traits (e.g., liver function, age) that inevitably address them to OLT rather than to LR.3 This diversity can be inferred from the recent experiences regarding the rate of transplantability, namely the rate of patients who are both transplantable and resectable; this appears to be about 20-25% of the total surgical population. In the Barcelona Clinic Liver Cancer system4 and in other classifications, LR and OLT are reserved for early-stage HCC (i.e., asymptomatic patients, affected by single HCC nodule ≤5 cm; or multiple tumors up to 3 nodules that are ≤3 cm in size), the only subgroup of patients that is known to receive an unquestionable benefit in survival from surgery. In the last few years, the number of patients with more advanced HCC referred for curative surgical procedures (both LR and OLT) has markedly increased, and predetermined criteria set up against the uncontrolled use of sophisticated surgical procedures are becoming hard to respect. The literature reports many different single-center experiences that on the one hand enroll patients for surgery whose disease is far beyond the definition of “early” HCC, and on the other hand seek to expand the current conventional entry criteria for resection or transplantation. HCC, hepatocellular carcinoma; LR, liver resection; OLT, orthotopic liver transplantation; UCSF, University of California, San Francisco; FAI, fractional allelic loss. LR surgery today seeks to broaden existing boundaries, taking into consideration the high expertise of surgeons in the field and the referral centers' desire for zero mortality. Recent prospective experiences enrolling patients with far more advanced tumors (single tumor of any size, providing that sufficient remnant liver exists; or multiple tumors up to 3 nodules, up to 3 cm, with a high percentage of bilobar extension, vascular invasion, and satellites) resulted in an acceptable 5-year survival rate of approximately 65%.5, 6 The characteristics of patients with cirrhosis and HCC who are eligible for resection will probably be routinely broadened in the future, likely by combining surgery with nonsurgical approaches, such as thermoablation. Moreover, the use of postoperative (adjuvant) regimens will help lengthen the recurrence-free survival time both early (<2 years) and late (>2 years) after resection.7 In terms of OLT, in the past 10 years, survival rates have steadily improved from initially disappointing 5-year survival rates ranging 18-40% in the early days, to a far better >70% rate, thanks to careful patient selection. Patient selection was improved by the introduction of the conventional Milan criteria in 1996.8 Patients who meet the Milan criteria today have radiologic evidence of a single tumor ≤5 cm in diameter, or 2-3 tumors each with a maximum diameter of ≤3 cm. The Milan criteria originated in a scientific and general context where liver transplantation for cancer was hardly accepted and transplantation was considered to be palliation. The Milan criteria were conceived to predict a good outcome in patients who met the criteria itself, rather than to predict a poor prognosis in patients who did not meet the criteria. The elements of the Milan criteria, namely size and number of tumors, have been shown by multivariate analysis to be the only independent variables predicting patient survival and tumor recurrence, with other biological prognostic factors playing a role only within the size and number limits (Fig. 1). These same results were confirmed worldwide, and the Milan criteria were subsequently adopted by the United Network for Organ Sharing staging system. Size of the overlapping population, Milano vs. University of California, San Francisco (UCSF), criteria. In recent years, as a consequence of the change in centers' surgery scenarios (in terms of immunosuppression, surgical expertise, dropout rate, etc.), and considering that many studies have suggested that tumor stage beyond the Milan criteria does not necessarily predict worse survival after OLT, numerous centers have advocated removal or expansion of the Milan criteria. The University of California, San Francisco (UCSF), criteria (single tumor nodule up to 6.5 cm; or 3 or fewer tumors, the largest ≤4.5 cm with the sum of the total tumor diameters ≤8 cm) best reflects the trend toward expanding the Milan criteria.9 Three recent articles summarizing European and American experiences, in addition to the single-center Milan experience, confirmed that even though the UCSF criteria have been independently validated in several studies, their applicability could be limited because most of the patients who meet the UCSF criteria also meet the Milan criteria, and the exact percentage of patients who meet just the UCSF criteria is negligible (estimated to be <10% of the each reporting center's overall transplanted population) (Fig. 2). Impact of University of California, San Francisco (UCSF), criteria. Analysis of 479 patients with hepatocellular carcinoma (HCC) and short waiting time. When applied to pre–orthotopic liver transplantation evaluation, UCSF criteria are associated with a 5-year survival <50%. Their applicability is therefore limited, despite similar survival rates compared with Milan criteria, when the explanted liver is taken into account. In a recent large French multicenter study,10 the application of the UCSF criteria compared with the Milan criteria in the pretransplantation setting (contemplating time on the waiting list close to zero) is associated with a very low survival rate (<50%) (Fig. 3). Despite this, the UCSF criteria has become the most used alternative measurement tool. Prospective validation of its potential has not yet been investigated. Metro ticketing paradigm. To try and overcome this problem, and to avoid competition among different proposals that seek to expand the Milan criteria, the vast area outside conventional criteria needs to be analyzed.11, 12 One such attempt reports patients outside the Milan criteria by plotting size (x-axis) and number of tumors (y-axis) on a Cartesian plane. By using this language of size and number, and following the so-called metro ticket paradigm, each spot on the graph represents the originating “city.” The longer the “trip” away from the conventional limits (e.g., the higher the number or size of nodules), the higher the price of the “ticket” (e.g., reduction in expected patient survival) (Fig. 4). From http://www.hcc-olt-metroticket.org. After exploring the area outside the Milan criteria, attention may be diverted to the definition of a predicting model far from any possible criteria limitations. This is what our group suggested to the international liver transplantation community during the 2006 International Liver Transplantation Society meeting held in Milan. We proposed a nonsponsored multicenter collection of patients who did not meet the Milan criteria. Thirty-five centers all around the world participated in our survey. We collected the biggest ever series of 1,142 patients who did not meet the Milan criteria (Fig. 5). Hepatocellular carcinoma (HCC) forecast chart showing search for extended criteria for orthotopic liver transplantation (OLT) in patients with HCC. Predicted survival is based on size and number criteria. From http://www.hcc-olt-metroticket.org. We plotted the results of size and number of tumors on a Cartesian plane. On the basis of the relationship between individual tumor characteristics and actual survival, we created a HCC forecast chart that can predict the 5-year survival rate (Fig. 6). This chart is likely destined to replace previous criteria to establish candidates for liver transplantation. This particular dynamic model can be implemented with other important variables in HCC and OLT that would not be considered otherwise; it can also change greatly from center to center. It may change the approach to patient listing (i.e., vascular invasion, dropout rate, donor availability, ethical and social issues), starting from the acceptable survival percentages adopted in different transplant realities. Decision analysis models for hepatocellular carcinoma (HCC). The last determining factor missing in most of the experiences that could actually be contemplated when analyzing results of surgical procedures is the exact measure of the difference in gain of life expectancy. When considered with all the other analyzed factors, this should be considered for adoption in the final decisional model. Such a model would work not just for single patients but for large cohorts of patients. Gain in life expectancy should be considered as the starting point for future prospective trials that seek to identify survival prediction models. This model should take into consideration other factors such as local conditions, score and prioritization policies for patients with HCC, graft quality, and above all patient condition and possible alternatives in therapeutic strategies (Fig. 7). It is likely that molecular markers of cancer will permit a refinement of the predictive model. This will add discriminatory power to the current staging systems and improve organ allocation schemes for patients with HCC. The current staging criteria for HCC takes into account tumor morphology, which is closely intertwined with tumor biology. But considering that morphology is not yet sufficiently defined by radiological procedures, and considering that it is often subject to evaluation errors, we must incorporate biological data that may allow desegregation of HCC patients from morphology-based staging systems and allow strict correlation of prognosis with biologic parameters such as vascular invasion, grading, and presence of microsatellites. To understand the biological behavior and identify genes associated with survival after OLT, Marsh et al.13 pioneered microdissection of explanted tissue and studied DNA mutations near 9 tumor-associated gene loci to create an index of cumulative mutational damage, termed the fractional allelic loss (FAI). They found that FAI and vascular invasion were the strongest independent predictors of tumor-free survival after OLT, even for cases exceeding conventional and expanded criteria. The major limitation of this study is that FAI cannot be recognized preoperatively and therefore cannot be included in any of the currently used staging systems. Nevertheless, the genotyping information can be preoperatively obtained by a biopsy of the lesion. Such a procedure could be provided to all HCC patients while awaiting discovery of molecular surrogate markers, with particular reference cases exceeding the size and number criteria.