- Top of page
- Supporting Information
In the last US national conference on liver transplantation for hepatocellular carcinoma (HCC), a continuous priority score, that incorporates model for end-stage liver disease (MELD), alpha-fetoprotein and tumor size, was recommended to ensure a more equitable liver allocation. However, prioritizing highest alpha-fetoprotein levels or largest tumors may select lesions at a higher risk for recurrence; similarly, patients with higher degree of liver failure could have lower postoperative survival. Data from 300 adult HCC recipients were reviewed and the proposed HCC-MELD equation was applied to verify if it can predict post-transplantation survival. The 5-year survival and recurrence rates after transplantation were 72.8 and 13.5%, respectively. Cox regression analysis confirmed HCC-MELD as predictive of both postoperative survival and recurrence (p < 0.001). The 5-year predicted survival and recurrence rates were plotted against the HCC-MELD-based dropout probability: the higher the dropout probability while on waiting list, the lower the predicted survival after transplantation, that is worsened by hepatitis C positivity; similarly, the higher the predicted HCC recurrence rate after transplantation. The HCC priority score could predict the postoperative survival of HCC recipients and could be useful in selecting patients with greater possibilities of survival, resulting in higher post-transplantation survival rates of HCC populations.
- Top of page
- Supporting Information
Allocation rules for patients with hepatocellular carcinoma (HCC), waiting for liver transplantation under the model for end-stage liver disease (MELD)-based policy, are still a difficult issue in continuous evolution. Since March 2005, the United Network for Organ Sharing (UNOS) policy states that patients with a T1 HCC do not receive extra MELD points whereas a MELD score of 22 is given to patients with a T2 HCC (1,2). The UNOS priority for HCC patients was based on the replacement of the native MELD score, assessing the risk of death due to liver failure by an estimation of the risk of tumoral progression. However, HCC patients appear to be advantaged in the current system, and the question was raised whether added priority is necessary (3). Adjustments to allocation policy have all been aimed at maintaining equity in access to deceased donor livers for both HCC and non-HCC liver transplant patients while maintaining respectable outcomes for all recipients. However, the current HCC policy does not seem to adhere to the general principles for liver allocation adopted with the introduction of the MELD score; in fact HCC patients still have better access to the liver transplant donor pool compared with non-HCC patients (3,4). For this reason the development of a more dynamic score has been highly endorsed (4).
In the last US national conference of transplant physicians, surgeons and policy makers, a calculated continuous HCC priority score, that incorporates the MELD score, alpha-fetoprotein (AFP) level and tumor size, was recommended on the basis of waiting list removal data registered on the Organ Procurement and Transplantation Network (OPTN) (3,4). This model was first suggested by Freeman in 2006, who developed the HCC-MELD equation (3): both Cox modeling and competing risk analysis showed that the risk of dropout increases linearly with increased severity of liver disease as well as with an increase of AFP and tumor size (3,4). However, these features are also well-known risk factors for postoperative survival, thus prioritizing the highest AFP level or the largest tumor may select HCC lesions at a higher risk for recurrence; similarly, patients with a higher degree of liver failure could have lower postoperative survival, especially in cases with the highest MELD scores (5). The first issue that has to be clarified is whether the HCC priority score could also predict the postoperative survival of HCC recipients. The second issue to investigate, in the case of a relationship between the HCC priority score and postoperative survival, is what is the probability of dropout that could be accepted with respect to a predicted reduction of postoperative survival. These two objectives are the aims of this report, since the balance between dropout risk while on the waiting list and the predicted survival after liver transplantation of HCC patients has still to be clarified.
- Top of page
- Supporting Information
The primary end-point of the present study was to test whether the risk of dropout from the waiting list for primary transplantation of HCC patients could also predict survival after liver transplantation. The probability of dropout of HCC candidates was calculated by means of the HCC-MELD formula proposed by Freeman in 2006, so that the probability of dropout has to be considered as a hypothetical HCC-MELD-based dropout probability, and defined as 1–0.920 exp [0.09369 × (MELD − 12.48) + 0.00193 × (AFP − 97.4) + 0.1505 (maximum tumor size − 2.59)] (3). The HCC-MELD formula was chosen because it was built from the largest liver transplant candidate database available (OPTN), externally validated (6) and reflecting our allocation policy since March 2003 (7). Briefly, in our transplant program, patients were listed in order of their MELD score, and an extra score, based on the HCC stage, downstaging protocol and for every month on the list, was added to the native MELD score. We also applied the UNOS modification of January 2004 that removed all increased priority for stage 1 tumors. The MELD score was calculated on the basis of immediate preoperative blood biochemistry as well as serum AFP; maximum tumor size was considered at the last clinical observation before transplantation. In particular, in our transplant program, HCC patients were followed while on the waiting list with a computed tomography (CT) scan or magnetic resonance imaging (MRI) every 3 months, so that data regarding tumor burden were not older than this time point from the surgical procedure.
The testing population was represented by 300 adult patients with a preoperative diagnosis of HCC and cirrhosis who underwent primary whole liver transplantation from January 1, 1997 to December 31, 2009. Exclusion criteria adopted to select the final study population were postoperative diagnosis of HCC (incidental cases = 33), combined kidney–liver transplant procedures (one case) and postoperative diagnosis of combined hepato-cholangiocarcinoma (one case). The preoperative diagnosis of HCC was based upon the European Association for the Study of the Liver (EASL) and the American Association for the Study of Liver Diseases (AASLD) guidelines (8,9). During the waiting time, a multidisciplinary team defined adequate preoperative treatments for each patient according to tumor stage, HCC location and liver function. Patients with a single nodule or two nodules located in the same segment and preserved liver function underwent liver resection; in the case of multiple nodules and/or clinical signs of liver dysfunction patients were submitted to trans-arterial chemoembolization (TACE); radiofrequency ablation (RFA) was applied in the case of deep and/or multiple nodules smaller than 3 cm; percutaneous ethanol injection (PEI) was preferred when nodules were adjacent to a major vessel. In our transplant program, HCC patients were considered still transplantable if they did not exceed Milan criteria (single ≤5 cm or up to three nodules, each of them ≤3 cm without major vascular invasion or distant metastasis). From January 1, 2003, 33 patients initially not meeting the Milan criteria were included on the waiting list after a downstaging protocol, as previously reported (10). Tumor grade was scored using the modified nuclear grading scheme outlined by Edmondson and Steiner (11,12). In all cases, tumor grade was defined by the poorest degree of differentiation identified within the tumor upon pathologic analysis of the entire specimen. Microscopic vascular invasion (MVI) was defined by the presence of tumor emboli within either the central hepatic vein, the portal vein, or the large capsular vessels (13). Fulfillment of the Milan criteria at pathological examination was also considered.
Postoperatively, calcineurin inhibitors (tacrolimus or cyclosporine) were used as immuno-suppressants in the majority of patients; m-TOR inhibitors were administrated to patients as primary immunosuppressant drugs or in combination with reduced doses of calcineurin inhibitors in the case of side-effects of calcineurin inhibitors and/or histological evidence of HCC with a high risk of recurrence (14). The follow-up after OLT included determination of serum AFP level and hepatic US every 3 months, and abdominal CT scan and chest X-ray every 6 months. In the case of suspected tumor recurrence, CT and/or MRI were performed. In selected cases, bone scan, positron emission tomography scan, angiography of the celiac trunk with Lipiodol injection and, if necessary, liver biopsy were performed.
The impact of the HCC-MELD score on postoperative outcome was investigated using both the Kaplan–Meier method together with the log-rank test and the Cox regression model. Follow-up collection ended on January 1, 2011. Since HCC-MELD incorporates both tumor and liver function variables, the primary endpoint of the study was patient survival rather than recurrence rate; risk-adjusted survival was thus calculated accounting for potential confounding covariates that were those proving significantly related to postoperative survival at univariate analysis. Univariate and multivariate Cox regression analyses were also performed on tumor recurrence rate. The predicted 5-year patient survival and recurrence rates were then calculated by means of the Cox regression analysis: S(t) = So(t)e(βx), where S(t) is the survival at time t, and So(t) is the baseline survival at time t (5 years); So(t) is raised to a power of e(βx), where βx is the sum of the individual coefficients for each term in the model multiplied by their actual values. A significance level of 0.05 was used in all analyses. The statistical analysis was performed using SPSS Version 10.0 software for PC computer (SPSS, Chicago, IL, USA).
- Top of page
- Supporting Information
The baseline demographical, clinical and tumor characteristics of the whole study population are reported in Table 1 together with the main donor and intraoperative characteristics. Distribution of the HCC-MELD scores is reported in Figure 1; in particular, the median HCC-MELD score was 0.093 that corresponds to a HCC-MELD-based predicted the 3-month dropout probability of 9.3%, equivalent to a standard MELD score of about 20. During a median follow-up of 3.6 years (range 1 month–12 years), 85 patients died (28.3%). The 1-, 3- and 5-year survival rates after transplantation were 87.7, 78.1 and 72.8%, respectively. The 10-year patient survival was 58.9%. Recurrence of HCC was the leading cause of death accounting for 32 cases (37.6%), hepatitis recurrence for 18 cases (21.1%), infection for 17 cases (20.0%) and miscellaneous for the remaining proportion. Death most frequently occurred in HCV-positive patients (63 cases out of 85; 74.1%) rather than HCV-negative patients (22 cases; 25.9%). In particular, HCC recurrence was the cause of death in 25 out of 63 HCV-positive patients (39.7%) and in 7 out of 22 HCV-negative patients (31.8%; p = 0.613). Hepatitis recurrence was more frequently the cause of death in HCV-positive patients (16 cases; 25.4%) rather than in HCV-negative patients (2 cases; 9.1%; p = 0.107). As regards HCC recurrence, 36 patients developed tumor relapse during follow-up (thus, 4 patients were still alive with tumor recurrence at the end of clinical observation): the 1-, 3- and 5-year HCC recurrence rates were 2.3, 6.3 and 13.5%, respectively. Survival curves after liver transplantation, in relationship to the median HCC-MELD value, are reported in Figure 2. Patients with an HCC-MELD score below 9.3% had 1-, 3- and 5-year survival rates of 91.3, 87.4 and 80.5%, respectively, whereas patients with an HCC-MELD score ≥ 9.3% had significantly lower survival rates of 84.0, 69.0 and 63.3%; respectively (Figure 2A; p = 0.001). The 1-, 3- and 5-year HCC recurrence rates of patients with lower HCC-MELD scores were 2.1, 5.3 and 7.5%, respectively and of patients with higher HCC-MELD scores were 9.3, 17.0 and 19.1%, respectively (Figure 2B; p = 0.002).
Table 1. Baseline characteristics of the study population
|Variables||All HCC Patients (no. = 300)|
|Recipient age (years) (median; range)||57 (32.5–70.0)|
|Recipient gender (M/F)||254/46 (84.7%/15.3%)|
| HCV||167 (55.7%)|
| HBV||74 (24.7%)|
| HBV and HCV||13 (4.3%)|
| Alcohol||35 (11.7%)|
| Other||11 (3.7%)|
| Serum bilirubin (mg/dL) (median; range)||2.45 (0.90–58.20)|
| Serum creatinine (mg/dL) (median; range)||1.00 (0.10–3.60)|
| INR (median; range)||1.42 (0.90–5.20)|
| Serum AFP (ng/dL) (median; range)||8 (1–6826)|
| AFP > 500 ng/mL||11 (3.7%)|
|Preoperative MELD score (median; range)||15 (6–45)|
| 6–10||63 (21.0%)|
| 11–14||76 (25.3%)|
| 15–20||94 (31.3%)|
| 21–30||59 (19.7%)|
| > 30||8 (2.7%)|
|Preoperative maximum tumor size (cm) (median; range)||2.5 (1.0–5.0)|
| 1||129 (43.0%)|
| 2||89 (29.7%)|
| 3||82 (27.3%)|
| T1||26 (8.7%)|
| T2||274 (91.3%)|
| TACE only||125 (41.7%)|
| TACE + Percutaneous therapies||38 (12.7%)|
| Percutaneous therapies only||47 (15.7%)|
| Hepatic resection||17 (5.7%)|
| None||73 (24.3%)|
|Downstaging procedures||62 (20.7%)|
|HCC-MELD score (median; range)||0.093 (0.034–1.000)|
|Donor gender (M/F)||160/140 (53.3%–46.7%)|
|Donor age (years) (median; range)||63 (14–95)|
| <18 years||5 (1.7%)|
| 18–39 years||47 (15.7%)|
| 40–49 years||30 (10.0%)|
| 50–59 years||49 (16.3%)|
| ≥ 60 years||169 (56.3%)|
|CVA as cause of donor death||200 (66.7%)|
|Donor HBcAb positive||72 (24.0%)|
|Cold ischemic time (hours) (median; range)||6 (2.5–13.0)|
|Presence of histological complete tumor necrosis||58 (19.3%)|
|Histologic maximum tumor size (cm) (median; range)||2.5 (0.8–6.5)|
|Histologic tumor number|
| 1||105 (35.0%)|
| 2||58 (19.3%)|
| 3||28 (9.3%)|
| >3||51 (17.0%)|
|Histologic tumor grading*|| |
| G1–G2||82 (27.3%)|
| G3–G4||160 (53.3%)|
|Presence of MVI*||129 (43.0%)|
|Histologic Milan criteria fulfilled||195 (65.0%)|
Figure 1. Distribution of HCC-MELD values in the study population. HCC-MELD score is the hypothetical 3-month predicted probability of HCC candidates to drop out from the waiting list. For the purpose of the study, the HCC-MELD formula was calculated on the basis of immediate preoperative blood biochemistry and serum AFP; maximum tumor size was established not more than 3 months before liver transplantation.
Download figure to PowerPoint
Figure 2. Kaplan–Meier survival curves in relationship to the median HCC-MELD value calculated at transplant (9.3%). (A) Patient survival and (B) tumor recurrence rate after transplantation were both significantly affected by the HCC-MELD score (p < 0.001 and 0.002, respectively).
Download figure to PowerPoint
Table 2 reports univariate Cox regression analysis on patient survival after transplant. Preoperative variables found to be significantly related to survival were female recipient gender (p = 0.036); HCV positivity (p = 0.002); serum AFP (p < 0.001); preoperative MELD score (p = 0.047), HCC-MELD score (p < 0.001) and cold ischemic time (p = 0.014). Histological variables found to be related to survival were tumor grading (p = 0.008) and MVI (p = 0.001). Multivariate Cox regression analysis, performed on preoperative available variables, confirmed HCC-MELD score (p < 0.001), HCV positivity (p = 0.003) and cold ischemic time (p = 0.015) as independent predictors of survival after liver transplantation (Supporting Table S1). Regarding covariates in relationship to the HCV status, the median MELD score of HCV-positive patients was 15 (range 7–37) and of HCV-negative patients it was 14 (range 6–45; p = 0.463); the median HCC-MELD score of HCV-positive patients was 8.9 (range 3.4–71.8) and of HCV-negative patients it was 9.4 (range 3.6–100; p = 0.245). The prevalence of G3-G4 tumors was similar between HCV positive (99 out of 147 cases without complete necrosis at histological examination; 67.3%) and HCV-negative patients (61 cases out of 95; 64.2%; p = 0.677). The prevalence of MVI was also similar (54.4% in HCV positive and 51.6% in HCV-negative patients; p = 0.694). The baseline 5-year survival of the whole study population, at mean of covariates, was 76.9% with a β-coefficient for the HCC-MELD score of 2.62. Stratifying Cox regression for HCV serology status, the baseline 5-year survival for HCV-negative patients was 83.7% and for HCV-positive patients it was 70.6% when adjusted for cold ischemic time. Relationships observed between the HCC-MELD score and the 5-year predicted survival after transplantation are reported in Figure 3A; the higher the probability of dropout while on the waiting list, the lower the predicted survival after transplantation, that is worsened by HCV-positivity. With a 3-month HCC-MELD-based dropout probability of 15%, the predicted 5-year survival after liver transplantation was about 68% but dropped to about 60% if the patient was HCV positive and rose to about 77% for HCV-negative patients.
Table 2. Univariate Cox regression model on patient survival after liver transplantation
|Variables||Exp (B)||95% C.I.||p-Value|
|Recipient age (years)||1.03||0.99–1.06||0.117|
|Recipient gender female||1.78||1.04–3.06||0.036|
|Diagnosis of HCV||2.34||1.38–3.97||0.002|
|Serum AFP (log 10; ng/dL)*||1.99||1.51–2.63||<0.001|
|Preoperative MELD score*||1.03||1.01–1.07||0.047|
|Preoperative maximum tumor size (cm)||1.02||0.99–1.03||0.087|
|Preoperative tumor number||1.03||0.85–1.26||0.741|
|Preoperative bridge therapies||1.55||0.88–2.73||0.132|
|Donor gender male||1.01||0.64–1.59||0.977|
|Donor age (years)||1.01||0.99–1.02||0.595|
|CVA as cause of donor death||1.29||0.79–2.12||0.314|
|Donor HBcAb negative||1.13||0.65–1.96||0.671|
|Cold ischemic time (hours)||1.09||1.02–1.16||0.014|
|Histologic complete necrosis||0.54||0.27–1.09||0.087|
|Histologic maximum tumor size (cm)||1.06||0.83–1.37||0.635|
|Histologic tumor number||1.04||0.85–1.28||0.701|
|Tumor grading G3–G4||1.04||1.01–1.07||0.008|
|Presence of MVI||2.99||1.72–5.21||0.001|
|Histologic Milan criteria fulfilled||0.48||0.30–1.15||0.113|
Figure 3. Results of the multivariate Cox regression models on patient survival and tumor recurrence rate after transplantation and relationship to the HCC-MELD score. (A) 5-year predicted patient survival was adjusted for cold ischemic time and stratified for HCV status; (B) 5-year predicted recurrence rate was adjusted for the tumor number. With respect to a HCC-MELD-based 3-month dropout probability of 15%, the predicted 5-year survival after liver transplantation was 67.8%, dropping to 59.8% if the patient was HCV positive and rising to 76.9% for an HCV-negative patient; the predicted recurrence rate was 18.9%.
Download figure to PowerPoint
A further Cox regression analysis was conducted to identify predictors of tumor recurrence after transplantation; results of univariate analysis are outlined in Table 3. Preoperative variables found to be significantly related to HCC recurrence were serum AFP (p < 0.001); preoperative tumor size (p = 0.043) and number (p = 0.006) as well as the HCC-MELD score (p < 0.001). Histological variables found to be related to survival were tumor number (p = 0.001), grading (p = 0.008), MVI (p = 0.001) and pathological Milan criteria (p = 0.001). Multivariate Cox regression analysis, performed on preoperative available variables, confirmed the HCC-MELD score (p < 0.001) and tumor number (p = 0.010) as independent predictors of HCC recurrence after liver transplantation (Supporting Table S1). The baseline 5-year recurrence-free survival of the whole study population, adjusted for tumor number, was 89.1% with a β-coefficient for the HCC-MELD score of 3.95. The relationship observed between the HCC-MELD score and the 5-year-predicted recurrence rate after transplantation is reported in Figure 3B; the higher the probability of dropout while on the waiting list, the higher the predicted HCC recurrence rate after transplantation. With a 3-month HCC-MELD-based dropout probability of 15%, the predicted 5-year recurrence rate was about 19%. However, it should be noted that the MELD score was not significantly related to HCC recurrence rate at univariate analysis (as expected); consequently, the relationship between the HCC-MELD score and recurrence rate has to be considered as the sole expression of tumor size and serum AFP.
Table 3. The Cox regression model on HCC recurrence rate after liver transplantation
|Variables||Exp (B)||95% C.I.||p-Value|
|Recipient age (years)||1.01||0.96–1.05||0.953|
|Recipient gender male||1.95||0.88–4.30||0.100|
|Diagnosis of HCV||2.32||0.86–5.26||0.132|
|Serum AFP (Log 10; ng/dL)*||3.28||2.19–4.90||<0.001|
|Preoperative MELD score||1.02||0.97–1.07||0.433|
|Preoperative maximum tumor size (cm)*||1.30||1.01–1.81||0.043|
|Preoperative tumor number||1.43||1.11–1.85||0.006|
|Preoperative bridge therapies||1.71||0.81–3.14||0.133|
|Donor gender female||1.52||0.77–2.99||0.228|
|Donor age (years)||0.99||0.98–1.01||0.880|
|CVA as cause of donor death||0.66||0.33–1.30||0.229|
|Donor HBcAb positive||1.14||0.53–2.44||0.731|
|Cold ischemic time (hours)||0.98||0.84–1.16||0.852|
|Histologic complete necrosis||0.04||0.01–1.24||0.065|
|Histologic maximum tumor size (cm)||1.29||0.99–1.67||0.057|
|Histologic tumor number||1.29||1.15–1.45||0.001|
|Tumor grading G3–G4||1.07||1.02–1.12||0.004|
|Presence of MVI||6.01||2.81–12.70||0.001|
|Histologic Milan criteria fulfilled||0.14||0.06–0.34||0.001|
- Top of page
- Supporting Information
The introduction of the MELD allocation system allowing priority scores for patients with limited-stage HCC has resulted in a consistent increase in the proportion of liver transplant recipients who have HCC: in the United States, more than 25% of liver donor organs are allocated to patients with HCC, many of whom have low MELD scores (3,4). In our geographical region (Italy), the proportion of HCC is even higher, up to 45% (15) being much more endemic in this area (16). Although it is frequently quoted that the survival after transplantation of HCC patients within the Milan criteria is similar to the survival of patients without HCC, recent evidence showed that this does not appear to be true when adjusting for the severity of the underlying liver disease (5). The current study suggests that prioritizing HCC candidates on the basis of the MELD score, serum AFP protein and tumor diameter can lead to a predictable worse post-transplantation survival since the same prognostic factors are involved in determining both dropout from the waiting list and post-transplantation survival (3,5). Consequently, the dropout probability that could be accepted with respect to a predictable reduction of postoperative survival is the most important issue to be considered in optimizing the policy that guides liver transplantation for HCC.
Based on the HCC-MELD equation, a patient meeting listing criteria, with a calculated MELD score of 15, a maximum tumor size of 3 cm and an AFP of 500 ng/mL has a dropout probability of 21.7%, equivalent to a standard MELD score of 24; consequently, this patient should receive high priority because of the high risk of dropout (3). The present study showed, however, that the 5-year survival that could be expected for this patient is only 62.9%, that is quite below the survival of both HCC and nonmalignant recipient populations reported by UNOS database (5,17). In the presence of HCV positivity, the predicted 5-year survival drops to 54.2% and in the absence of HCV infection it rises to 73.1%. The stratification of post-transplantation survival by HCV status is fundamental for a correct prediction of post-transplant survival; in fact, it is well known that HCV patients have a worse post-transplantation survival in comparison to non-HCV patients and regardless of HCC presence (18,19). Therefore, the patient with the HCC-MELD-based dropout risk of 21.7% will have a satisfactory survival if HCV negative, but whether a 5-year survival of 54.2% could be considered acceptable is debatable. From an urgency point of view (sickest first), the above-mentioned patient should surely receive the highest priority. From a utility point of view (maximize life expectancy across liver recipients), a 54% 5-year survival seems to be quite low. From a survival benefit point of view a 54.2% 5-year survival could be acceptable because the survival benefit concept is aimed at ameliorating the intention-to-treat survival since listing. However, the use of a very limited resource, as liver donors are, undertakes to also pursue an acceptable postoperative survival that has still to be defined (20).
On the other hand, the present analysis suggests that there can be room for expanding the Milan criteria in some specific conditions. Considering a patient with a MELD score of 12, a maximum tumor size of 8 cm and a normal AFP of 10 ng/mL, the predicted 5-year survival after transplantation is 68.4%, dropping to 60.6% if the patient is HCV positive and rising to 77.3% for an HCV-negative patient. The presence of a large tumor size and a low serum AFP is consistent with a slow tumor growth because it is well differentiated and there is no MVI (21), recognized risk factors for tumor recurrence after liver transplantation (22,23). One method for selecting this type of patient would be to incorporate a time-from-diagnosis variable to provide more weight for the slower and less aggressive tumors in balance with the urgency criteria (5); alternatively, histologic evaluation of specimens obtained by preoperative fine-needle biopsy should be used to identify patients exceeding Milan criteria with a ‘good biology’ (24). The absence of hepatitis C infection will further improve the possibility of survival and a 5-year survival of 77.3% is surely acceptable for a patient exceeding Milan criteria and will not cause any harm to the remaining waiting list candidates (25). It should be noted, however, that since the HCC-MELD equation was derived from a population meeting Milan criteria for the majority of patients and our study population consisted of patients within the same criteria, the model prediction of survival can be overestimated and the issue will require further investigations.
In March 2003, in developing our priority policy, we already assumed that the fixed score proposed by UNOS would not satisfactorily take into account the role of the underlying liver disease in patients with HCC (7). However, our priority score is not continuous and is based upon native MELD, UNOS TNM stage and waiting-time. The proposed HCC-MELD equation (3) and the new guidelines from the US national conference (4) gives us the opportunity to have a valuable tool to further optimize both allocation system and post-transplant prediction of survival in HCC patients. The present analysis can well represent an external and independent validation set for the new US proposed HCC priority rules that provide the confirmation that this could be the right way to obtain both a fairer allocation system and an improvement in survival after transplantation. It should be noted however, that a European validation of the HCC-MELD equation was never provided and results from the present study must be taken with caution in considering this equation as a definitive system in scenarios different from USA. More recently, a recalculation of hazard coefficient was proposed on the basis of competing risk analysis (26). In the waiting list for liver transplantation scenario, a patient may experience an event other than the one of interest (dropout) which alters the probability of experiencing the event of interest, such as transplantation (competing risk event). In this setting, the calculation of the cumulative incidence of the specific event of interest (dropout) is more informative (27). Unfortunately, a clear formula with the new hazard coefficient was not provided, making a comparison with the HCC-MELD equation impossible. It should also be noted that the concordance statistics for the Cox model and competing risk model were identical and equal to 0.70. Consequently, the HCC-MELD formula was the only tool readily available to test the hypothesis of the present study.
In conclusion, the HCC priority score could predict postoperative survival of HCC recipients and could be useful to improve transplant results by identifying patients with a good or a poor outcome. Factors affecting dropout risk and post-transplantation survival are essentially the same and further investigations should be aimed at assessing the survival benefit of HCC patients considering these features.