An 88-Year-Old Woman with a 33-Year History of Idiopathic Portal Hypertension Presenting with Hepatocellular Carcinoma Treated with Carbon-Ion Radiotherapy

Introduction

Idiopathic portal hypertension (IPH), also called non-cirrhotic portal fibrosis or hepatic portal sclerosis, is a disease of unknown cause characterized by findings of portal hypertension without histopathologic cirrhosis [1]. Clinically, IPH is characterized by the features of portal hypertension: esophageal varices (EVs), splenomegaly, preserved liver function, and patency of the hepatic veins and portal vein. IPH has been reported in many parts of the world, especially in developing countries [2]. IPH is common in India and Japan, with an incidence of 5.6% in India [3] and 7.9 per million in Japan [4]. However, it accounts for only 3–6% of patients with portal hypertension in Western Europe [5]. The main complication of IPH is gastroesophageal variceal bleeding, and the prognosis for IPH is generally good once this bleeding is controlled [6]. EVs are effectively treated by endoscopic injection sclerotherapy, while splenectomy is also indicated in cases of gastric varices or marked hypersplenism [2].

Hepatocellular carcinoma (HCC) accounts for 90% of primary liver tumors and is the fifth most common cancer worldwide [7]. Approximately 80–90% of HCCs are associated with cirrhosis. The main etiology of HCC is chronic infection with hepatitis B virus and hepatitis C virus (70% of cases), followed by excessive alcohol consumption, non-alcoholic fatty liver diseases, aflatoxin, and rare inborn errors of metabolism such as Wilson’s disease [7]. The annual incidence of HCC from hepatitis viral cirrhosis and alcoholic cirrhosis is 2.0–4.8% [8]. HCC is largely asymptomatic in the early stages of the disease. In early-stage HCC, alpha-fetoprotein (AFP) is a useful biomarker, and des-gamma-carboxyprothrombin (DCP) and lens culinaris agglutinin-reactive fraction of AFP-L3 may also be useful [7]. The diagnosis of HCC is based on imaging studies and usually does not require biopsy. When the tumor diameter is greater than 1 cm, the diagnostic probability exceeds 90% if contrast computed tomography (CT) or contrast magnetic resonance imaging (MRI) shows non-peripheral washout images in the portal or delayed phases, hyperintense nonmarginal images in the arterial phase, and smooth enhanced capsular images [7,8]. Treatment for HCC is staged by the Barcelona Clinic Liver Cancer (BCLC) system [9] based on the tumor burden, severity of liver disease using the Child-Pugh classification, and patient performance status. The treatment options are ablation, resection, transplantation, arterial embolization, and chemotherapy.

HCC in patients with IPH is extremely rare, with only a few cases reported to date, and the causal relationship is not clear. Herein, we describe the case of an 88-year-old Japanese woman with a 33-year history of IPH who presented with HCC and was treated with carbon-ion radiotherapy (CIRT).

Case Report

An 88-year-old Japanese woman presented to our hospital with hematemesis 33 years ago. She had undergone colorectal cancer resection 10 years previously, at which time unevenness of the liver and splenomegaly were noted but were not investigated or treated. She had no history of alcohol consumption or blood transfusion and no family history of autoimmune or hematological diseases. Upon presentation, she had mild anemia and thrombocytopenia, with normal liver function test results. Furthermore, she tested negative for hepatitis B virus surface antigen, hepatitis C virus antibody, and antibodies for autoimmune liver diseases (Table 1). Ultrasound examination and CT revealed mild atrophy of the right lobe of the liver, enlargement of the left lobe, and marked splenomegaly. The results of tests for viral and autoimmune liver disease were negative. Endoscopy revealed moderately enlarged EVs with a red-color sign, prompting endoscopic injection sclerotherapy. We considered it likely that she had IPH and followed her up.

Further endoscopic injection sclerotherapy was performed for recurrent EVs over the next 4 years, achieving temporary shrinkage. Recurrence of the red-color sign and an increase in gastric varices prompted us to perform Hassab’s procedure 4 years after the initial presentation. A liver biopsy taken at the same time showed preserved lobular architecture, round fibrous enlargement of the portal area, dilatation of vessels in the portal area, and focal dilated sinusoids, but scant evidence of hepatocellular inflammation (Figure 1), all of which are characteristic of IPH [1,10], leading to a diagnosis of IPH.

The patient’s subsequent course was good; however, she gradually developed hypoalbuminemia, and after 5 years, she did not develop edema or ascites, but persistent hypoalbuminemia (approximately 3.5 g/dl, normal reference range 4.1–4.9 g/dl) led to her being started on branched-chain amino acid granules.

Another 3 years later, she had a cerebral infarction but had no sequelae. She was started on clopidogrel 75 mg/day. Another 2 years later, she developed mild encephalopathy with frequent daytime somnolence and hyperammonemia (NH3 105 μmol/L, normal reference range 7–40 μmol/L). She was then started on lactulose, which achieved normalization and stabilization of her ammonia concentrations and resolution of her somnolence.

Thirty years after our patient’s initial presentation, her serum AFP concentration, which had previously been less than 5.0 ng/mL, increased to 14.7 ng/mL (normal reference range <10.0 ng/ml) and her DCP concentration increased from less than 40 mAU/mL (normal reference range <40 mAU/mL) to 63 mAU/mL, whereas the AFP-L3 concentration was only 10% (normal reference range <10%). In the same month, CT showed a 24-mm-diameter HCC in segment 8 (Figure 2). By another 4 months later, her AFP concentration had further increased to 199.7 ng/mL and her DCP concentration had increased to 2120 mAU/mL. Portal vein thrombosis (PVT) was diagnosed by repeat CT (Figure 3); however, no arterioportal shunts were evident in the portal phase. By that time, our patient was 88 years old, 145 cm tall, weighed 45 kg, was awake for more than half the day, could take care of herself, and had a Child-Pugh score of 6. She was diagnosed with BCLC stage A and underwent CIRT at 60 Gy (relative biological effectiveness) in 4 fractions at the Kanagawa Cancer Center in December 2021. One week later, her D-dimer concentration was 12.9 μg/mL (normal reference range <0.5 μg/mL) and antithrombin-III (AT-III) was 88% (normal reference range 80–120%), and thrombolytic therapy was initiated with edoxaban at 15 mg/day for 7 days, followed by 30 mg/day. However, from the 11th day of the 30 mg/day regimen, she had clinically significant subcutaneous hemorrhaging and black stools, so the dose was reduced to 15 mg/day. This achieved resolution of her bleeding tendency, so the same dose was continued. Four weeks after commencing edoxaban treatment, a CT scan showed no evidence of PVT. Furthermore, 4 months after completing CIRT, CT scans showed shrinkage of the HCC, and her AFP concentration had normalized to less than 5.0 ng/mL and DCP concentration to less than 40 mAU/mL. Edoxaban was continued. Almost 3 years after CIRT treatment, there is no evidence of recurrence of HCC or PVT on imaging (Figure 4), and AFP and DCP concentrations remain normal.

Discussion

The present case is instructive in the following respects. First, our findings indicate that HCC can occur in older adult patients who are long-term survivors of IPH. Second, extremely favorable results can be achieved in older adult patients using CIRT and direct oral anticoagulants for HCC and PVT, respectively.

IPH is a rare disease of unknown etiology. An increasing incidence of IPH has resulted in greater awareness of it in recent years in Western Europe. Background factors known to be associated with IPH include HIV infection, other chronic infections, immune disorders, hematological disorders, malignancies, therapeutic drugs (eg, azathioprine, arsenic), and genetic predisposition [5,11]. However, as in the present case, many patients lack these background factors. Thus, IPH should be included in the differential diagnoses for patients with hepatosplenomegaly of unknown cause and thrombocytopenia. Our patient was diagnosed with IPH based on the presence of portal hypertension symptoms such as EVs and splenomegaly, the absence of cirrhosis on liver biopsy, histological findings consistent with IPH, and the absence of known causes of liver disease such as hepatitis virus, autoimmune liver diseases, alcohol consumption, drugs, and obesity [2].

Several studies on the long-term prognosis of IPH have been reported in the last 25 years but, to our knowledge, no cases have been followed up for more than 30 years and no long-term survivors with IPH have been reported to develop liver cancer. Hillaire et al [12] followed up 28 patients with IPH for an average of 7.6 years (maximum 21 years) and noted 4 deaths, but no cases of liver cancer. All deaths were attributed to liver failure complicated by acute hepatitis or chronic renal failure, and the authors concluded that IPH itself is a relatively benign disease [12]. Schouten et al [11] followed up 62 patients with IPH for an average of 7.5 years (maximum 26 years), 23 of whom died during the study period. None of the deaths were due to HCC, and there were only 4 liver-related deaths [11], indicating that patients with IPH rarely die from liver-related complications. Siramolpiwat et al [6] followed up 62 patients for an average of 6.1 years (maximum 15 years) and reported 1 liver transplant and 6 direct or indirect liver-related deaths, none of which were associated with HCC. These data indicate that development of HCC in patients with IPH is extremely rare. Furthermore, Okuda et al [13] identified only 1 case of IPH in 56 autopsy cases with histologically confirmed non-cirrhotic liver cancer.

To our knowledge, there are only 3 reported cases of HCC developing during follow-up of IPH. Isobe et al [14] reported a case of a 69-year-old woman who developed HCC after 6.5 years of follow-up and underwent hepatic resection. Hidaka et al [15] reported a case of a 75-year-old woman who developed HCC after 13 years of follow-up, but was not treated for HCC due to her poor condition, and she died of pneumonia 1 year later. In both cases, histological findings in the liver were consistent with IPH, revealing no evidence of cirrhosis. The patients in these 2 cases, like ours, were Japanese older adult women; therefore, the possible factors contributing to the development of HCC are genetic factors, aging, and parenchymal changes associated with the course of IPH. Gokce et al [16] also reported a case of a 38-year-old woman with hepatoportal sclerosis who developed HCC after 2 years of follow-up and underwent radiofrequency ablation (RFA), but later underwent liver transplantation due to multiple metastases. The case reported by Gokce et al [16] is distinct from the others because the patient was relatively young at the onset of HCC; the liver tissue displayed pericentral fibrosis [16], which is a hallmark of alcoholic liver injury and non-alcoholic steatohepatitis [17]. However, the patient’s body mass index and alcohol consumption history were unknown. In the present case, a liver biopsy showed minor changes consistent with IPH 26 years before the diagnosis of HCC. However, our patient’s liver function gradually deteriorated thereafter and she developed hepatic encephalopathy and hypoalbuminemia and presumably had histological changes in her liver, such as fibrosis and reduced parenchymal volume.

In the present case, although we did not confirm the diagnosis of HCC histologically, we believe that the imaging findings and changes in AFP and DCP concentrations strongly support the diagnosis [7]. In terms of the pathogenesis of HCC, the absence of factors such as hepatitis viruses and alcohol consumption, and the very old age of onset of HCC (88 years) suggests that age-related changes in the liver, such as increased reactive oxygen species generation due to increased lipofuscin, genomic instability, telomere exhaustion, epigenetic changes, impaired protein homeostasis, and mitochondrial depletion and dysfunction, all played major roles in carcinogenesis [18]. On CT imaging, there was mild cardiac enlargement (Figure 2), which had been consistent for some time, and there was no dilatation of the inferior vena cava or hepatic veins on imaging and no signs of cardiac failure. Therefore, we considered her mild cardiac enlargement to be inconsequential and not related to hepatic pathology. Genetic factors and the ability to provide long-term follow-up may have also contributed to the fact that most reports of HCC in IPH to date have occurred in Japan.

Regarding treatment of our patient’s HCC, we considered that less invasive RFA or microwave ablation would be preferable because of her BCLC stage A and advanced age [19]. Furthermore, because the lesion was located just below the hepatic S8 dome and close to the hepatic limbus, the risk of complications was presumably high and she would have been at greater risk of recurrent cerebral infarction if antithrombotic drugs were withdrawn. We selected CIRT to minimize these risks. CIRT has advantageous physical and radiobiological properties compared with conventional radiotherapy – the carbon-ion beam has a unique physical property called the Bragg peak and can deliver the maximum amount of energy to the tumor while minimizing energy deposition in the surrounding normal tissue. Thus, CIRT allows for high-dose irradiation of the target, severely damaging DNA while reducing radiation exposure to organs at risk [20]. However, few facilities can offer CIRT because of its high costs. The indications for CIRT in patients with HCC have not been established. However, it is used in patients with impaired liver function who are expected to experience significant liver toxicity from photon radiation therapy, patients with large tumors, patients who require re-irradiation, and patients whose HCC is close to critical organs such as major vessels. Because CIRT is a relatively new therapy, there are currently limited data from large-scale clinical trials of patients with HCC, and there are various uncertainties about factors such as bowel movements and bowel gas when the tumor is located near the digestive tract [21]. In Japan, CIRT has been actively administered. For example, Fujita et al [22] reported a series of 69 patients with a single HCC of ≤5 cm or 2–3 HCCs of ≤3 cm. In their patient group, CIRT achieved a lower cumulative subregional recurrence rate than RFA and achieved an equivalent cumulative local recurrence rate and progression-free and overall survivals [22]. In a multicenter study, Shibuya et al [23] reported treating 174 patients with HCCs (mostly older patients, median age 73 years) with a maximum tumor diameter of 10.3 cm (median 3.0 cm), and mostly with a single HCC, with 48–60 Gy (relative biological effectiveness) in 2 or 4 fractions. This treatment achieved a 3-year overall survival rate of 87.7%, local control rate of 81.0%, and Grade 3–4 toxicity of only 5.7% [23]. Similarly, our patient had no adverse events, and a favorable outcome was achieved. We therefore believe that this CIRT treatment should be widely used in the future.

Regarding the incidence of PVT in patients with IPH in Western Europe, Hillaire et al [12] reported a 12.9% incidence during a mean follow-up of 7.5 years, Schouten et al [11] reported a 46% incidence during a mean follow-up of 7.6 years, and Siramolpiwat et al [6] reported a 36% incidence in 69 cases during a maximum follow-up of 10 years. In Japan, Matsutani et al [24] followed 22 Japanese patients with IPH for a mean period of 12 years and identified 9 cases (41%) of PVT, a higher incidence than that observed in a control group with cirrhosis (10%, 7/70 cases). However, given that our patient had no evidence of PVT for 30 years following the onset of IPH and that PVT was first detected several months after the onset of HCC, we postulate that the PVT was caused by the HCC rather than the IPH. The incidence of venous thromboembolism has been reported to be 4–7 times higher in patients with cancer than in patients without cancer [25,26], and the incidence of PVT in patients with HCC is reportedly 20–65% [27]. In other words, the combination of induction of the blood coagulation cascade by tissue factors overexpressed by carcinogenesis, vascular endothelial damage by cancer cell-derived cytokines, and stagnation of blood flow in the portal vein, together with her extremely old age, would all have contributed to the development of PVT in our patient [28]. In the treatment of PVT, AT-III replacement has been shown to be capable of correcting abnormalities in the coagulation fibrinolytic system without increasing the risk of bleeding. Additionally, Hidaka et al [29] reported a good response rate of 56% with AT-III treatment in patients with PVT and an AT-III below 70%, mainly in patients with cirrhosis. However, our patient’s AT-III concentration was within the normal range (88%), encouraging us to use other treatment options. Nagaoki et al [30] reported favorable results in a study of patients with PVT that had developed into cirrhosis, with danaparoid sodium for 2 weeks followed by edoxaban for 6 months, achieving complete resolution of PVT in 70% of their patients. However, we excluded implementation of the abovementioned options as our patient was already taking clopidogrel, an antiplatelet drug, and there is a lack of data on the combination of clopidogrel and danaparoid, as well as concerns about the likely decline in our aged patient’s physical and mental function if she was hospitalized for twice-daily intravenous infusions. Thus, we considered that the optimal means of treating her PVT was with oral anticoagulants, which are a low-risk, minimally invasive option. Because she was already taking the antiplatelet drug clopidogrel, adding a vitamin K antagonist, a more conventional anticoagulant, could have increased her risk of bleeding [31]. We chose to administer edoxaban, a direct oral factor Xa inhibitor, accompanied by careful monitoring, because this agent is more effective and carries less risk of bleeding than traditional anticoagulation therapy [32]. Transient, mild subcutaneous and gastrointestinal hemorrhage occurred in the early phase of treatment, but resolved with dose adjustment, and recanalization was achieved within 3 months. The superiority of edoxaban over a vitamin K antagonist for treatment of PVT has recently been reported. Tadokoro et al [33] reported a prospective study of 61 patients with cirrhosis and PVT treated with edoxaban or a vitamin K antagonist for approximately 600 days and found that the overall rate of thrombolysis was significantly better in the edoxaban group (76.7%) versus the vitamin K antagonist group (29.4%). Thus, direct oral anticoagulants may be a promising option for the treatment of PVT in the future.

Conclusions

This report focused on the rare condition of IPH and HCC co-existing in a patient followed up for more than 30 years. Although the association between IPH and HCC is not clear, this report highlights the importance of continuous clinical follow-up in patients with chronic liver disease.