Curative treatment is recommended as the first-line treatment for patients with single HCC regardless of tumor diameter [1, 7]. In clinical practice, however, patients with single tumors unfit for curative treatment are usually treated by TACE, based on clinical judgment. According to the current guidelines, TACE is the first line non-curative treatment for intermediate stage patients . No evidence of a beneficial impact of TACE in patients with single HCC is reported in the guidelines since the trials upon which the guidelines are built , for the most part, included patients with multiple nodules of HCC. Accordingly, TACE is frequently performed outside the current treatment guidelines in a considerable percentage of patients with a single nodule, according to a “stage migration strategy” .
Only a few studies have evaluated the efficacy of TACE in patients with a single nodule [5, 17, 18] and a valid comparison with previous data reported in the literature is very difficult, due to the different criteria used for the evaluation of tumor response, TACE procedure, the selectivity of technique and the expertise of the radiological center. This fact led to the investigation of the overall survival and clinical determinants of survival in patients with a single nodule who represent approximately half (45%) of the total cohort of patients who underwent a first TACE cycle in our Interventional Radiology Unit (156/344) (Figure 1). This number is fully comparable to a very large Japanese series in which patients with single tumors were 46% of those who underwent TACE , and some other studies [14, 19] which showed high heterogeneity of patients routinely undergoing TACE, including 35-50% of patients with single tumors, even those <5 cm. Furthermore, the vast majority of the studies investigating the efficacy of TACE excluded patients with advanced liver disease, PVT and impaired PS; therefore, there was also no evidence of the impact of TACE in those categories of patients . The allocation policy and the impact of TACE in patients with impaired liver function (namely CPT-B patients) has already been described  and, in the present study, the aim was to evaluate the impact of bland PVT and slight impairment of PS on overall survival after TACE.
The median overall survival of the entire patient population, after the exclusion of patients who underwent LT who were generally long term survivors, was 36.0 months with 1-, 3- and 5-years survival rates of 85%, 50% and 26%, respectively. These data are slightly lower than those observed in a large Japanese series  reporting 1-, 3- and 5-years survival rates of 91%, 66% and 53%, respectively in patients treated with TACE for a single nodule of HCC (even though no information regarding possible subsequent LT was reported). As expected, when comparing these results with those reported in the metanalysis of Llovet et al. (median survival of 20 months in patients who underwent TACE) in which the vast majority of patients had multinodular HCC , the median overall survival was considerably higher despite the large presence of CPT-B patients in our series. On the basis of survival analysis, TACE treatment indeed represents a valid therapeutic option for patients with single HCC who are not eligible for curative treatment, as has also been shown by recent series of BCLC-A patients from Barcelona and from Pisa [17, 22]. Such data also supported the use of the stage migration policy from the early to the intermediate HCC stage.
When assessing the clinical predictors of survival, tumor diameter >3 cm, and particularly >5 cm (beyond the MC), lack of complete radiological tumor response, AFP ≥14.5 ng/mL, the presence of ascites before TACE and a MELD increase ≥1 point the day after TACE were found to be independently associated with shorter survival at multivariate analysis. These data are in agreement with the fact that life expectancy depends not only upon tumor treatment efficacy, but also on the underlying severity of liver disease and patients with worsening hepatic function after TACE; with a MELD score increase ≥1 point, they are at risk of liver failure.
The presence of bland PVT in patients with HCC represents a challenging therapeutic issue. In recent decades, some authors  have considered the presence of PVT to be a contraindication for TACE due to the risk of liver function deterioration and hepatic infarct  but patients with PVT may not present technical and safety contraindications to TACE if a selective/superselective procedure is performed [20, 25]. In fact, more recent studies have demonstrated that TACE could be a safe treatment option for HCC patients with PV occlusion especially when performed in a selective manner , and that TACE could have a survival benefit over conservative treatment [27, 28]. In our Hospital, patients with bland thrombosis are candidates for TACE if they have preserved liver function, limited tumor burden, contraindications to other treatment and a selective approach is feasible. Interestingly, despite the limited number of patients with bland PVT (n = 17) who underwent TACE, our results showed that the presence of bland PVT, either lobar or segmental, has no negative impact on overall survival when TACE is performed with a selective or superselective approach.
The BCLC staging system includes the ECOG PS  evaluation regarding the assessment of tumor stage. In patients with HCC, the classic determination of PS is not able to differentiate between cancer- or cirrhosis- related symptoms  and the subjective assessment of “how the patient feels” can be related to cirrhosis as well as to cancer. In our study population, 15 patients with PS-1 were formally classified to be in the advanced stage (BCLC-C) (Table 2) but, since the tumor diameter was ≤5 cm, the likelihood of having cancer-related symptoms could be considered very low. Accordingly, these patients in BCLC-C had a significant and theoretically paradoxical better survival (28 months) than patients in the BCLC-B stage (6 months), as all the latter had large tumors (diameter >5 cm) (Tables 2, 5). It could be speculated that, in case of symptoms of uncertain tumor relation, the tumor burden should be considered the driving force for treatment allocation. On the other hand, considering patients with the same tumor burden (within the MC), PS-1, and consequently the same liver function, this certainly impacts survival so that BCLC-C patients (PS-1) had significantly worse survival with respect to BCLC 0-A patients (PS-0) (Figure 3).
To the best of our knowledge, the only data on tumor radiological response of conventional TACE in patients with single unresectable HCC is that of Malagari et al. . Our study showed notably higher objective response rates (CR + PR) of 90% vs. 59.6% and superior rates of CR (64%) and PR (26%) (Table 3) as compared to the 4.8% CR and 54.8% PR reported by Malagari. Our data appeared consistent with those reporting per-nodule TACE efficacy (mimicking patients with a single nodule)  where similar CR and PR rates were reported (64% and 36%, respectively) and tumor diameter ≤5 cm was again found to be a statistical predictor of complete response .
We acknowledge that the response rate in our study might be overestimated since the assessment of radiological response was usually made with CT, which may underestimate the residual tumor due to the interference of Lipiodol . The availability of MRI was not sufficient to offer this technique as a standard procedure after TACE to all patients. Nevertheless, our data are of current interest since, despite the introduction of TACE using drug-eluting beads [30, 31], clinical trials comparing TACE with drug-eluting beads and conventional TACE did not show significant differences in tumor response and overall survival [19, 32] and, nowadays, conventional TACE is still for the most part used.
A number of studies have demonstrated that the repetition of TACE increases tumor response and prolongs survival , but it is necessary to select the best TACE candidates who could benefit from treatment and eventually subsequent cancer retreatment in order to avoid overtreatment and detrimental effects on liver function. The issue of proper patient selection for retreatment after TACE has become more stringent in recent years due to the availability of alternative treatments such as sorafenib , and radioembolization . To this end, the worsening of laboratory tests 24 hours after treatment was evaluated and a significant modification in serum albumin, bilirubin, the INR and the MELD score after TACE treatment was documented (Figure 2). As expected, it was also found that patients with a CPT score increase ≥1 point were more likely to undergo a single TACE cycle vs. multiple cycles (70% vs. 30%, P = 0.006) Furthermore, both a CPT and a MELD score ≥1 point increase were found to be associated with a significantly worse prognosis (Table 4). Such findings do not affect the initial choice of recommending TACE, but seem to alert clinicians to consider the risk that patients will be no more candidate for future repeated TACE in case of early CPT score worsening after the procedure, deserving an even more careful assessment of treatment strategy.