How to manage hepatic vein tumour thrombus in hepatocellular carcinoma

See article in J. Gastroenterol. Hepatol. 1999; 14: 922–7. Recently, screening of high-risk populations for hepatocellular carcinoma (HCC) by using ultrasonography and serum alpha-fetoprotein levels has facilitated the early detection of HCC. However, at the time of diagnosis, some HCC patients still have advanced disease with hepatic vein tumour thrombus (HV-TT). Hepatocellular carcinoma has a tendency to involve vascular structures in the liver, such as the portal veins (PV) and the hepatic veins (HV). Although HCC involvement of the HV is less frequently observed compared with that of the PV, tumour thrombus (TT) extending into the inferior vena cava (IVC) and right atrium (RA) through the HV has been encountered. 1 Despite the progress of therapy for HCC, HCC patients with HV-TT generally have an extremely poor prognosis. A certain number of patients with HV-TT may develop secondary Budd–Chiari syndrome, pulmonary infarction, and/or lung metastasis, especially when TT extend into the IVC. 2 Further extension of HCC into the RA can cause cardiac failure and ball valve thrombus syndrome. 3 Because of the advances in imaging modalities, tumour growth through the HV can be diagnosed with certainty. On angiography, the threads and streaks sign, indicating feeder vessels of the TT, may be observed from the HV to the IVC and the RA. 4 In recent years, ultrasonography and computed tomography have become the most useful and most commonly used modalities for the diagnosis of HV-TT. 5 Magnetic resonance imaging may offer further detailed information on TT, such as the presence of vascular lumens between the TT and IVC wall. 6 In HCC patients with HV-TT, hepatic resection with removal of the TT is the only radical treatment that may facilitate prolonged survival, although such extensive surgery is applied only to patients with sufficient hepatic reserve to tolerate surgery. 7–9 Tumour thrombus removal without hepatic resection is also carried out to prevent pulmonary infarction and/or to ameliorate TT-related symptoms, such as oedema and ascites. 10,11 However, the effects of such a surgical procedure will be transient and limited, unless new extension of the TT is prevented by successive treatments for hepatic tumours. Liver transplantation is another surgical treatment for HCC, but the unfavourable outcomes reported in advanced HCC do not encourage liver transplantation for patients with HV-TT, given the serious shortage of donor organs. 12 The prognosis of patients after these surgical treatments is generally poor, even when hepatic resection with removal of TT is successfully performed. 9–11 The prognosis of such resected patients has been reported to depend on the extension of TT through the HV and/or coexistence of portal vein TT (PV-TT). 13,14 Patients with HV-TT alone show a better prognosis than those with TT extending into the IVC; the outcome of patients with IVC-TT is extremely poor because of early distant-organ metastasis. 13 The prognosis of patients with HV-TT can also differ considerably, depending on whether they have TT in the PV; the prognosis of patients with TT in both the PV and HV is much poorer, probably because they are also at high risk for intrahepatic metastasis through PV-TT. 14 Most patients with HV-TT are not treated surgically because of concomitant liver dysfunction, but receive non-surgical treatment. In some patients, only palliative treatments may be offered due to the advanced stage of HCC in the liver. Non-surgical treatments include transcatheter arterial embolization (TAE), chemotherapy, and radiation therapy. Successful non-surgical treatments may allow a small subset of patients with initially unresectable HCC to become candidates for surgical treatment. 15 Moreover, given marked tumour regression, the surgical procedure required for radical resection may become less invasive. However, at present, little anti-tumour effects can be expected from any non-surgical treatments in patients with HV-TT. Transcatheter arterial embolization, either alone or in combination with intra-arterial chemotherapy, has a marked anti-tumour effect against HCC because this neoplasm receives its blood supply through the hepatic artery. 16,17 The effect of TAE is dependent on the growth pattern of HCC; the effect is more apparent in nodular lesions, while infiltrative HCC does not respond as well as nodular HCC. 18 The anticancer effect of TAE in HCC patients with HV-TT is usually limited, although HV-TT is also fed from tumour vessels derived from the hepatic artery. The possible reason for this refractoriness of HV-TT to TAE is that most such cases show an infiltrative growth pattern macroscopically. 19 Segmental TAE, which has recently been reported to show a marked anti-tumour effect against PV-TT, should also be evaluated in patients with HV-TT, as it can enhance the anti-tumour effect and can be performed more safely than conventional TAE. 20–22 Chemotherapy for HCC, either systemically or intra-arterially, has been of limited value in clinical practice, because only a small portion of patients will obtain meaningful palliation. 6,23–27 In particular, patients with quite advanced HCC are inappropriate candidates for systemic chemotherapy, because the disease is more refractory to chemotherapy at that stage. 28 However, hepatic intra-arterial infusion (HAI) may have to be tested in patients with HV-TT, as HAI has been reported to show promising results in HCC patients with TT in the main portal branch. 29 Radiation therapy may be a good palliative therapy, either alone or in combination with other treatments for HCC with TT extending through the HV; radiation therapy for TT may induce tumour regression, resulting in palliation of TT-related symptoms. 30–32 This is the background to the paper by Kashima et al. published in a recent issue of the Journal of Gastroenterology and Hepatology. 33 They treated five patients with advanced HCC showing extensive tumour growth through the HV by HAI by using aclarubicin, mitomycin C and lipiodol. Although two patients died of cancer within several weeks after HAI, the remaining three demonstrated remarkable tumour regression. As a result, two of these three patients could undergo potentially curative resection, and surgical specimens revealed coagulation necrosis (complete necrosis in one, and partial in one) in both the hepatic tumours and the TT. Furthermore, the prognoses of the three patients were incredibly improved, even though all patients had TT in the IVC and/or the PV. Kashima et al. concluded that HAI using these agents might be an effective treatment for HCC with extensive tumour growth through the HV. However, several cautions are necessary in interpreting the results by Kashima et al. First, the number of patients enrolled in this trial was too small to accurately evaluate the anti-tumour and adverse effects of HAI. Moreover, according to intention-to-treat analysis, these effects must be evaluated in all patients who receive the therapy. Second, patient selection for HAI, which was not described in detail in the paper, must be optimized to achieve the potential benefits of HAI without inducing severe complications. The two patients with early death after treatment might not have been appropriate candidates for this therapy. Finally, the HAI method somewhat varied with each patient, in terms of the combined use of gelatin powder and mitomycin C. To define the real role of this HAI in patients with HV-TT, further trials, which include a large number of patients, are mandatory. Among HCC patients with HV-TT, disease status (extension of TT and associated liver disease) varies greatly with each patient. For patients who have less advanced tumour stage and better liver function, surgical treatment is the treatment of choice. In particular, patients with HV-TT alone may be the optimal candidates for surgical treatment; HV-TT can be removed by anatomical hepatic resection, and longer survival can be expected with the removal of both the hepatic tumours and HV-TT. In contrast, when TT extends into the IVC, and/or PV-TT coexists with HV-TT, the indication for such surgical treatment should be considered prudently. In these patients, a more drastic surgical procedure is required, and its effect on prognosis remains to be elucidated. 13,14 However, surgical removal of TT in the IVC and RA may be indicated only when hepatic tumours are under control or will be effectively controlled by non-surgical treatments. Patients who are unsuitable for resection may be enrolled in phase II trials to evaluate the anti-tumour effect and toxicity of various non-surgical treatments, if they are expected to be able tolerate treatment. Although none of the non-surgical treatments currently shows reliable anti-tumour effects in patients with HV-TT, TAE may be the most promising method of achieving significant tumour necrosis. However, for patients with TT in the main portal branch or poor hepatic reserve (Child’s grade C), TAE is not applicable. Such patients may undergo chemotherapy, especially HAI, in a clinical trial. Radiation therapy for TT may also be performed either alone or in combination with other non-surgical treatments to palliate TT-related symptoms. In addition, patients in this unresectable group who request treatment may be treated with more experimental therapies in a phase I setting.