HCC is associated with the presence of intralesional APS regarding the common portal vein in approximately 60% , or the main portal vein branches in 30% of patients according to data reported by Okuda et al .
The presence of a high flow APS with hemodynamically significant consequences, results in the creation or worsening of portal hypertension, with potentially life threatening complications such as oesophageal varices rupture, refractory ascites, and liver encephalopaty [1–3].
The diagnosis of APS is reached through the execution of a contrast enhanced abdomen CT, revealing the absence of atypical HCC radiologic features, a clear enhancement of the portal vein axis contemporary to the hepatic artery during the arterial phase of the examination and a lack of enhancement during the arterial phase, since the blood flow to the lesion is reduced because of the APS [8, 9]. The diagnosis of multifocal HCC with APS makes two issues arising: first, the treatment of APS and, secondly, the treatment of HCC out of Barcelona criteria .
Treatment options range from medical therapy alone (Sorafenib), interventional radiology procedures (TAE to treat the APS and/or transarterial chemoembolization for the treatment of hepatic multifocal lesions) with or without medical therapy, to radiological treatment of the shunt (APS-TAE) and subsequent surgical liver resection [11–14].
In this case report, in decision-making process two sets of issues were taken into account: in the one hand, the issue of APS in patients with HCC and, in the other the treatment of hepatocellular malignancy out of Barcelona criteria.
Regarding the first point, the emergency treatment of acute bleeding from esophageal varices may be made by endoscopic ligation or sclerotherapy, and finally by TIPS or surgical shunt if the control of portal hypertension is not effective .
It is remarkable that in patients with APS the control of portal hypertension is consequent to the treatment of the shunt itself. Even if the rates of recanalization are high, as reported in the literature , in this case it was decided to proceed with embolization of arterial blood vessels feeding the shunt, with the dual aim of reducing the risk of portal hypertension related complications (also aggravated by portal vein thrombosis) and of gaining better hemostatic control intraoperatively.
The need for a double session of embolization was established by radiological finding of shunt persistence due to arterial vessels originating from the left gastric artery.
Control of portal hypertension resulted in ascites resolution and re-classification of the patient in class A5 according to Child Pugh: if not, the patient would have taken two points for ascites resulting in B7. This means that transient complications of portal hypertension tend to overestimate patient's class of risk, especially regarding postoperative liver failure. After shunt treatment indeed, the patient was reported to the appropriate class of Child and during the postoperative course, the patient developed only mild and transient signs of liver failure. Ishii et al. noted that in the presence of APS the green indocyanin retention test proved to be unreliable , since it overstates functional parenchymal impairment .
Timing for surgery has to be decided in a case by case analysis, considering the balance between complete acute complication resolution and the risk of ischemic or septic complications and APS recanalization.
A proper evaluation of the hepatic functional reserve, prevents inappropriate exclusion of patients from the only potentially curative treatment, because of risk overestimation.
In the present case report, the patient presented with multifocal disease, portal hypertension and malignant PVT: the long term prognosis of these patients is quite poor, so that Barcelona criteria do not indicate surgery as the treatment of choice, since the 5-year survival rate is less than 50% [10, 18]. Despite of this, the most experienced surgical centers continue to perform liver resection out of Barcelona indications , suggesting that good long-term results are achievable, better than those reported after non-curative treatments .
In particular, patients with PVT have a median survival of 2.7 months without treatment . On the contrary, following surgery, several authors report 3 years survival rates ranging between 0 and 43% depending on the series [21–24]. Furthermore, the presence of portal hypertension is not a contraindication itself to invasive approach or a predictor of outcome, as suggested by Capussotti  and other authors [26, 27].