Efficacy and safety of endoscopic ultrasonography-guided interventional treatment for refractory malignant left-sided liver tumors: a case series of 26 patients

This study aimed to compare the efficacy and safety of EUS-guided ethanol injection and 125I seed brachytherapy for malignant left-sided liver tumors which were difficult for trans-abdominal intervention. The study protocol was registered at Clinicaltrials.gov (NCT02816944). Twenty-six patients were consecutively and prospectively hospitalized for EUS-guided interventional treatment of refractory malignant left-sided liver tumors between June 2014 and June 2016. Liver masses were detected using EUS in 25 of 26 (96.2%) patients. EUS-guided interventional treatment was completed uneventfully in 23 of 26 (88.5%) patients using anhydrous ethanol injection (n = 10) or iodine-125 seed implantation (n = 13). Six months later, complete response was achieved in 15 of 23 (65.2%) patients and partial response in 8 of 23 (34.8%) patients. Patients with tumor residual have second-look EUS-guided interventional treatment (n = 5), radiotherapy (n = 2) or surgical resection (n = 1). Complete response was achieved after repeated interventional treatment in 3 of 5 patients who underwent second EUS-guided intervention; 2 patients required additional surgical resection but one succeed. No significant complications occurred. Therefore EUS-guided 125I seed brachytherapy is an effective and safe treatment modality for radical operation or promising palliative control of malignant left-sided liver tumors refractory to trans-abdominal intervention.

in liver cancer patients. These two treatment regimens are normally performed percutaneous access. In a few cases, thermal ablation performed in open surgery or under laparoscopy. While percutaneous intervention is generally thought to be minimally invasive, it may not be feasible for some patients with left hepatic lobe atrophy, lesion was located at the surgical margin of the left liver or in the caudate lobe, gastrointestinal flatulence and abdominal skin surgical scar. At this time, the conventional trans-abdominal ultrasound has difficulty to display lesions or even unable to discover it.
Endoscopic ultrasonography (EUS) is an advanced endoscopic technique primarily used for the characterization and biopsy of the gastrointestinal tract, particularly for preoperative staging of malignant diseases, including oesophageal, gastric, and pancreatic cancers 11 . As left lobe of the liver is located in front of the lower esophagus, stomachus cardiacus and in proximity to the gastric lesser curvature, it is accessible on EUS, have better image quality and more adjacent to the focus. Therefore it is possible to interventionally treat left-sided liver tumors through EUS, especially when other percutaneous intervention is not deemed feasible or safe. The present work aimed to compare the efficacy and safety of EUS-guided ethanol injection and 125 I seed brachytherapy for malignant liver tumors located in the left lobe that are refractory to conventional percutaneous intervention. (A) presence of a 1.3-cm, low-density liver mass located in the left lobe on CT scan (white arrows). US image failed to show the lesions due to the gas ahead; (B) identification of a 1.5 cm × 1.3 cm low-echodensity mass on EUS (white arrows); (C) EUS-guided implantation of iodine-125 particles (white arrows); (D) obvious downsizing of the liver disease on follow-up CT scan at 1 months (white arrows); and (E) disappearance of the liver disease on follow-up MRI scan taken after 12 months.

Results
Demographic and clinical data. Overall, 26 patients were enrolled in this study, including 17 men and 9 women, aged 31-75 years. The basic characteristic of included patients was summarized in Table 1. The aetiologies of the hepatic tumors were chronic hepatitis (n = 6), liver cirrhosis (n = 11), and primary gastrointestinal malignancy (n = 9). Twenty-one patients had a single left-sided liver tumor, and 5 patients had more than one liver tumor with major disease located in the left lobe but multiple tumors in the right lobe. Liver tumors were occurred in left liver shadowed by gastrointestinal (GI) gas (n = 12; Fig. 1), resection margin (n = 7; Fig. 2),  caudate lobe (n = 3; Fig. 3), significantly atrophic left lobe (n = 2), and remarkable abdominal skin scar (n = 2). All tumors were not detectable or poorly visualised on percutaneous ultrasound scan because of the above reason. In addition, a systematic review by searching from PubMed up to August 2016 was performed. The summary of 11 studies of ethanol ablation and 125 I brachytherapy of the hepatocellular carcinoma was presented in Table 2 12-17 and Table 3   Tumor response and mid-term follow-up data. All patients were followed up for 1-2.5 years as scheduled. At 12 months after treatment, 12 of 13 (92.3%) patients achieved complete tumor response in iodine-125 particle implantation ( Fig. 1) while 3 of 10 (30%) in anhydrous ethanol ablation ( Fig. 3 and Table 4). Residual tumor was detected in 8 patients after anhydrous ethanol ablation (n = 2) or iodine-125 particle implantation (n = 6); these patients were further treated with second-look EUS-guided interventional treatment (n = 5) and radiotherapy (n = 2) and surgical resection (n = 1). Complete response was achieved after again EUS-guided treatment in 3 of 5 patients, whereas the remaining 2 patients required surgical resection but only one succeed. Recurrence was observed in 2 patients at 1 year after anhydrous ethanol ablation. No clinically significant late-onset complications occurred.

Discussion
Interventional EUS, namely, EUS-guided interventional treatment, has been reported for drainage of pancreatic 22 or bile duct (pseudo) cysts 23 or even complex choledochoduodenostomy 24 . This newly emerging interventional treatment incorporating ethanol or chemotherapy injection, brachytherapy, and/or photodynamic therapy has also been piloted in preclinical studies and a small series of clinical studies for the treatment of malignant pancreatic, lower common bile duct, and periampullary diseases 25 . In contrast, EUS is used relatively less frequently for the treatment of hepatic tumors. Awad et al. 26 reported that EUS could detect small-sized primary and secondary liver cancers previously undetectable on CT scans, and simultaneous fine-needle biopsy aided in the diagnosis of deep-seated liver tumors for preoperative staging. Singh et al. 27 also reported, in a small series of prospective patients, that EUS exhibited a higher specificity and positive predictive value than CT and percutaneous ultrasound scanning for metastatic liver tumors located in hepatic segments II and III. Furthermore, interventional EUS has been attempted for transgastric drainage of liver abscess 28 and infectious biloma after radiofrequency-ablated hepatocellular carcinoma 29 . To the best of our knowledge, the present work is the first prospective study comparing the therapeutic role of EUS-guided ethanol injection and 125 I seed brachytherapy in the treatment of left-sided liver malignancies. Percutaneous ultrasonography-guided interventional treatment normally requires an acoustic window to clearly display the liver tumor and allow subsequent ethanol injection, high-intensity focused ultrasound 30 , or other ablation therapy. However, in some patients, such as those enrolled in the current study, hepatic tumors cannot be detected or clearly visualised on percutaneous ultrasound scanning because of distorted liver anatomy from a previous operation, a deep-seated location such as the caudate lobe, at the edge of liver, at the surgical margin that usually affected by gastrointestinal gas or a pre-existing abdominal scar. For left-sided liver tumors, especially when located in proximity to the lesser curvature of the stomach, transgastric EUS can exclude the interference of intestinal gas and provide safe liver access for ethanol injection and iodine-125 implantation. Our results showed that EUS-guided iodine-125 brachytherapy has a better prognosis and less recurrence for left-sided liver tumors with a good safety profile comparing to ethanol ablation. Moreover, iodine-125

Author
Year Country  implantation is recommended for the treatment of larger tumors because of a higher risk of residual or recurrent tumor. In previous study, it was reported that the 1-, 3-and 5-year recurrence-free rates of the 125 I brachytherapy after resection of HCC were 94.12%, 76.42%, and 73.65%, and their overall survival rates were 94.12%, 73.53%, and 55.88%, respectively 9 , which was also similar to another study 21 . This may result from that a high-dose short radiation emitted by 125 I could be continuous within the the target tissues for a long half-life of 59.4 days, which would trigger tumor cell apoptosis and drop proliferation of tumor cells by gathering cells in the radiosensitive cell cycle phase (G2/M) 20,31 . The radiation declined sharply with the distance from the seeds and had limited injury to surrounding normal tissues. Meanwhile, 125 I implantation also could excite the anti-tumor immune response in HCC patients by adding CD3+ and CD4+ immunocytes and facilitating Th2/Th1 deviation 32 . Thus these findings indicated that EUS-guided 125 I brachytherapy could be a safe and feasible way to effectively target tumor cells and minimize injury to healthy tissues. There are some limitations regarding EUS-guided interventional treatment of liver cancer. First, interventional EUS requires a more delicate ultrasound system and greater operator expertise; therefore, this interventional technique is less feasible in general ultrasonographic practice. Second, transgastric EUS can only access a relatively small portion of the left lobe; an overly large puncture angulation failed the pre-planned interventional therapy in 2 of our patients. Third, implantation of radioactive particles is relatively complex through a EUS puncture needle compared to the percutaneous approach.

Characteristics of patients
In conclusion, EUS-guided iodine-125 brachytherapy is an effective alternative to ethanol injection for radical operation or promising palliative control of refractory malignant left-sided liver tumors. This technique is minimally invasive and associated with a good safety profile, and residual tumor can be treated by repeated sessions of interventional EUS. The long-term efficacy and safety of interventional EUS have yet to be investigated in large-scale, comparative studies, especially studies evaluating patients' survival and quality of life.

Patients and Methods
Patient selection. Between June 2014 and June 2016, 26 patients with primary or metastatic left-sided liver tumors diagnosed on abdominal ultrasound, CT, and/or MRI scans, who were ineligible for surgical treatment due to unresectable disease or high risk of surgical morbidity, were consecutively and prospectively referred to our ultrasound centre for palliative treatment. The included patients were randomly divided into two groups: 10 in the anhydrous ethanol group and 13 in the iodine-125 implantation group. Inclusion criteria were as follows: American Society of Anaesthesiologists physical status class I and II; relatively well confined, single hepatic tumor or extensive hepatic tumor (in the scenario of multiple foci) located in Couinaud segments I-III refractory to  EUS and interventional treatment procedures. All patients underwent routine hematologic, clinical biochemical, coagulative function, virologic, and oncologic assays prior to interventional EUS. All interventional procedures were performed by an assigned interventional treatment team led by a board-certified interventional physician, including resident interventional physicians, radiologists, and research nurses. Left-sided liver EUS was performed as previously described 28 . Briefly, the patient was placed in the left recumbent position and sedated using intravenous propofol. After excluding the presence of oesophagogastricvarices using conventional endoscopy, a Pentax endoscopic ultrasound system (Pentax Medical, Tokyo, Japan) incorporating a Hitachi colour Doppler ultrasound scanner (Hitachi Medical, Tokyo, Japan) was used for EUS of the left-sided liver. An endoscopic ultrasound injection needle (Cook Medical, Bloomington, IN, USA) was used for injection of anhydrous ethanol (22G) for well-encapsulated primary hepatocellular carcinoma or implantation of iodine-125 particles (19G; Cook Medical, Bloomington, IN, USA) for hepatocellular carcinoma, cholangiocellular carcinoma or metastatic liver cancer.
Postoperative care, follow-up procedure, and main outcome measures. Patients were closely followed after EUS and treated with antimicrobial, haemostatic, and hepatoprotective agents as indicated. All patients were followed up at the outpatient clinic using abdominal ultrasound, CT, and/or MRI at 1, 3, 6, and 12 months postoperatively. The main outcome measures were frequency of detection by EUS, tumor response as defined by the Response Evaluation Criteria in Solid Tumors 30 at 6 months, and interventional EUS-associated morbidities. In this study, all statistical analysis was performed using x 2 , Fisher exact test and Student t test by spss 13.0 software.