Int J Gastrointest Interv 2020; 9(4): 135-141
Published online October 31, 2020 https://doi.org/10.18528/ijgii200036
Copyright © International Journal of Gastrointestinal Intervention.
1Department of Internal Medicine, Ulsan University Hospital, University of Ulsan College of Medicine, Ulsan, Korea
2Department of Gastroenterology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
Correspondence to:* Corresponding author. Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, 88-Olympic-ro 43-gil, Songpa-gu, Seoul 05505, Korea.
E-mail address: firstname.lastname@example.org (D.-W. Seo).
Hoonsub So and Dongwook Oh contributed equally to this work as first authors.
This is an open-access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted noncommercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Endoscopic ultrasound (EUS)-guided fine-needle aspiration was introduced in the early 1990s. EUS has evolved from a diagnostic modality to a therapeutic tool for patients with various pancreatic neoplasms. Recent advances in EUS-guided interventions include drainage and ablation. EUS-guided treatment provides a minimally invasive option for patients with pancreatic neoplasms instead of surgery or the percutaneous approach. This review aimed to provide an overview of the current EUS-guided ablation treatments, such as ethanol ablation and radiofrequency ablation, for treating various pancreatic tumors.
Keywords: Ablation, Endoscopic ultrasound, Pancreatic neoplasms
Pancreatic neoplasms have a wide spectrum of biological behaviors from benign to malignant. Pancreatic cancer has a poor prognosis, with a 5-year overall survival rate of < 5% and a median survival of < 6 months.1 Surgical resection can provide the only chance of a cure, with 5-year overall survival rates of 18%–24%; however, only one-fifth of patients have a resectable disease status.2 The outcomes of chemotherapy or radiation therapy are unsatisfactory, with most patients with pancreatic cancer experiencing only a small benefit. Benign pancreatic neoplasms, such as pancreatic cystic lesions (PCLs) and neuroendocrine tumors (NETs), are increasingly encountered in clinical practice owing to the widespread use of cross-sectional imaging.3 As surgical resection is associated with substantial morbidity (20%–40%) and a mortality rate of 2%, the management of these tumors is challenging.4,5 Therefore, new advances in the treatment of pancreatic neoplasms are required.
Over the past two decades, endoscopic ultrasound (EUS)-guided intervention has evolved from a diagnostic method to a therapeutic modality. As EUS-guided treatment is effective and safe, it can be an alternative for patients with pancreatic neoplasms, especially for those who are poor surgical candidates. This review aimed to provide an overview of the latest published evidence and innovations in EUS-guided ablation treatment of pancreatic neoplasms.
EUS has enabled injecting an ablative agent directly into the pancreas. Ethanol is a commonly used an ablating agent because it is cheap and widely available. With its low viscosity, it can be easily injected and reaspirated using a 22-gauge fine-needle aspiration (FNA) needle. Direct injection of ethanol induces coagulation necrosis through cellular dehydration, protein denaturation, and vascular occlusion. The typical indications for EUS-guided ethanol ablation are pancreatic NETs and pancreatic cystic neoplasms. EUS-guided celiac plexus neurolysis (CPN) with ethanol will not be discussed here.
Pancreatic NETs are rare, accounting < 2% of all pancreatic tumors.6 These tumors are being increasingly diagnosed; however, the management of tumors measuring < 2 cm and low-grade, incidentally discovered tumors is debated.7,8 The treatment for pancreatic NETs should be tailored according to the patient’s age, performance, and symptoms; location of the disease; and experience of the center as surgical resection of the pancreas could lead to serious adverse events. Considering that these tumors are diagnosed before the age of 60 years in many patients, the risk of surgical adverse events should be minimized. Long surveillance can also cause considerable anxiety, leading to deterioration of quality of life. Therefore, EUS-guided ethanol ablation can be a good option.
Careful examination with diagnostic contrast-enhanced harmonic EUS (CE-EUS) helps in evaluating the extent of the mass. After calculating the tumor volume with computed tomography (CT) and EUS, the maximal targeting ethanol volume is decided. If the tumor is located close to a vessel or the main pancreatic duct, the volume could be reduced to one-third of the calculated volume based on the endosonographer’s discretion. Thereafter, the tumor is punctured with a 22-gauge FNA needle without a side hole. For ablation, 95%–99% ethanol is most commonly used. During the injection, the endosonographer can observe a hyperechoic cloud under EUS in real time. The injection is terminated when the hyperechoic cloud reaches the tumor border. Additional puncture can be performed if there are inadequately ablated areas. Mixing lipiodol can help evaluate complete response on CT after the procedure. CE-EUS can also help evaluate residual viable tumors.
Although most studies had a small sample size, they demonstrated safety and efficacy (Table 1).9–16 Park et al15 conducted the largest case series with 11 patients with 14 tumors (10 nonfunctional tumors and 4 insulinomas). The tumor size ranged from 8 to 19 mm. After a single treatment session, 7 patients (7/13, 53.8%) showed complete response and 2 patients with insulinomas became asymptomatic. Levy et al13 and Qin et al14 treated 5 and 4 insulinomas with EUS-guided ethanol injection, respectively, and reported symptomatic improvement in all cases. All other cases of insulinoma showed improved symptoms and/or no recurrence.9–12,16 EUS-guided ethanol therapy seems to be highly valuable for functional NETs, such as insulinoma, especially for small tumors measuring < 2 cm. The greatest concerns are the risk of progression after incomplete ablation during follow-up and the risk of pancreatitis. As it is impossible to judge complete response through pathologic evaluation, there is always a risk of incomplete ablation even with a radiologic complete response. Furthermore, injected ethanol does not evenly distribute within the tumor capsule, and the endosonographer can only depend on hyperechoic blush during the procedure. Therefore, researchers believe that the ideal indication for ethanol injection would be a tumor with a capsule measuring ≤ 15 mm. Most studies have only reported mild pancreatitis as an adverse event. In a patient with a functional NET, ablation caused necrotizing pancreatitis, requiring laparoscopic necrosectomy.10 Direct injection of ethanol into the pancreatic parenchyma has a risk of causing severe pancreatitis. This may be related to a direct cytotoxic effect on the ductal epithelium, activation of zymogen, and/or pancreatic ductal hypertension. In the case series by Park et al,15 pancreatitis occurred after injection of > 2 mL ethanol in a single session. The endosonographer should keep the needle inside the tumor, should not inject too much ethanol, and should minimize ethanol leakage into the surrounding normal area.
Table 1 . Summary of Representative Published Data on EUS-Guided Ethanol Ablation for Neuroendocrine Tumors.
|Study||No. of patients||Indication||Size (mm)*||Needle (gauge)||Ethanol (%)||Session||Result||Adverse events|
|Jürgensen et al9 (2006)||1||Insulinoma||13||22||95||1||No recurrence at 34 mo||Mild pancreatitis|
|Muscatiello et al10 (2008)||1||2 nonfunctional NET||7, 11||NA||40||2||No recurrence at 18 mo||Necrotizing pancreatitis|
|Deprez et al11 (2008)||1||Insulinoma||NA||NA||98||1||No recurrence at 24 mo||Mild pancreatitis, duodenal wall hematoma and ulceration|
|Vleggaar et al12 (2011)||1||Insulinoma||9||25||96||1||Decreased size, asymptomatic at 6 mo||None|
|Levy et al13 (2012)||5||Insulinoma||8, 16, 18, 20, 21||22, 25||95, 98, 99||2 (||3 asymptomatic, 2 improved symptoms||None|
|Qin et al14 (2014)||4||Insulinoma||5.4, 10, 10, 11.8||25||95||1||4 asymptomatic||None|
|Park et al15 (2015)||11||10 nonfunctional NET, 4 insulinoma||12.3 (8–19)||22||99||1 (||9 complete response, 4 partial response, 1 excluded||Mild pancreatitis, pancreatic duct stricture|
|de Sousa Lages et al16 (2017)||1||Insulinoma||12||NA||95||1||1 asymptomatic||None|
EUS, endoscopic ultrasound; NET, neuroendocrine tumor; NA, not available..
*Number or mean (range)..
Many types of pancreatic cystic tumors are diagnosed yearly, and their types and clinical courses vary. Some have the potential for malignant transformation. Intraductal papillary mucinous neoplasms (IPMNs) and mucinous cystic neoplasms (MCNs) have a risk of malignant transformation, although the exact risk is unclear.17 Most cysts are small (< 1 cm), and only 0.8% cysts measure > 2 cm.18 Resected IPMNs can show malignancy in up to 25% cases; however, the risk of malignant transformation in unresected IPMNs is markedly low.19 Clinicians attempt to remove risky cystic neoplasms before malignant transformation occurs by evaluating the lesion with respect to size, morphology, main pancreatic duct diameter, symptoms, etc.20 However, it is true that there are not many things to do rather than doing regular image surveillance. Finally, EUS has brought therapeutic progress to this field by enabling ablation with ethanol and chemotherapy infusion.
Careful EUS examination before ablation is necessary to determine the anatomical and morphologic features of lesions. Especially, the suitable location for the procedure (stable scope position) and the cyst characteristics (size, septation, wall thickness, mural nodule, communication with the pancreatic duct) should be evaluated.21 CE-EUS may help avoid the overdiagnosis of branch duct IPMNs with mural nodules.22 Further, EUS-guided FNA is performed for cyst evacuation. We prefer a 22-gauge needle in most cases. A 19-gauge needle can be used if the mucin takes too much time to aspirate; however, it has a risk of leakage. A 25-gauge needle is not appropriate for aspiration and injection. The amount of aspirated fluid should be recorded, and the fluid should be sent for further examination (e.g., amylase and carcinoembryonic antigen examination). Ethanol or other ablative agent is injected into the cyst in a volume equal to that of the originally aspirated fluid. Clinicians should keep in mind that an approximately 0.8 mL of fluid is held inside the 22-gauge FNA needle itself (19-gauge needle: 1.1 mL), and this should also be considered when determining the injection volume.23 Lavage of the cyst, which entails repeated injection and aspiration with an ablative agent into and from the cystic cavity, is performed 3–5 times depending on the situation. Instead of lavage, retention (a single injection and removal after 3–5 minutes) can be adopted when the cyst is too large for lavage. At the end of the session, the ablative agent should be completely evacuated.21
The main purpose of EUS-guided pancreatic cyst ablation is preventing death from the malignant progression of pancreatic cysts. The mechanism of EUS-guided ablation is believed to be epithelial denudation with fibrosis and atrophy of the epithelium induced by contact with an ablative agent.24–26 However, there is no consensus regarding the risk of malignant transformation, and no randomized trial has compared intervention (ablation or surgery) and watchful waiting. Most published studies were not large enough to show a survival benefit (Table 2).3,25–33 A recent position statement on EUS-guided ablation of pancreatic cystic neoplasms was published to provide answers to the clinical questions.34 It recommended EUS-guided ablation for patients who are unfit for surgery or who refuse surgery, but have a reasonable life expectancy. The cyst should be unilocular or oligolocular and should have a premalignant potential. The number of locules could be up to 6. It is reasonable to ablate an enlarging cyst measuring > 2 or 3 cm.34 Oh et al26 reported that lesions with a diameter of < 3.5 cm are more likely to show complete resolution. In this perspective, MCNs would be the most ideal candidate for ablation. However, in a recent study by Moyer et al,27 no relationship between variables (initial diameter < 2.5 cm, locularity, carcinoembryonic antigen level, presumed diagnosis) was found in terms of complete resolution. Five studies used ethanol only25,28–31; four studies used ethanol and paclitaxel3,26,32,33; and one study used ethanol, paclitaxel, and gemcitabine injection.27 The outcomes of the studies showed a complete resolution rate of 33%–79%. Hence, the effect seems to be rather durable. DeWitt et al33 reported a durable effect over a 27-month median follow-up in patients in whom complete ablation was achieved. In a study by Choi et al,35 98.3% patients in whom complete ablation was achieved showed a durable effect over a 6-year follow-up.
Table 2 . Summary of Representative Published Data on EUS-Guided Ethanol Ablation for Pancreatic Cystic Neoplasms.
|Study||No. of patients||Indication (%)*||Size (mm)†||Needle (gauge)||Ablative agent||Complete resolution rate||Session||Adverse events‡|
|Studies that used ethanol for ablation|
|Gan et al28 (2005)||25||MCN (52), IPMN (16), SCN (12), pseudocyst (12), indeterminate cyst (8)||19.4||22||5%–80% ethanol||35%||1 (||None|
|DeWitt et al25 (2009)||25||MCN (40), IPMN (40), SCN (12), PC (8)||10–41||22||80% ethanol||40.9%||1 (||Pancreatitis (|
|DiMaio et al29 (2011)||13||IPMN (100)||10–32||22||80% ethanol||38%||2 (||None|
|Caillol et al31 (2012)||13||MCN (100)||10–50||NA||99% ethanol||85%||1 (||None|
|Gómez et al30 (2016)||23||IPMN (66), MCN (17), nonmucinous (17)||14.9–49.3||19, 22||0%–79% ethanol||9%||1 (||Pancreatitis (|
|Studies that used ethanol and paclitaxel for ablation|
|Oh et al3 (2008)||14||MCN (14), SCN (21), lymphangioma (21), indeterminate cyst (44)||17–52||22||99% ethanol + paclitaxel (3 mg/mL)||79%||1 (||Pancreatitis (|
|Oh et al32 (2009)||10||MCN (30), SNC (40), indeterminate cyst (30)||20–68||22||99% ethanol + paclitaxel (3 mg/mL)||60%||1 (||Pancreatitis (|
|Oh et al26 (2011)||47||MCN (19), SCN/PC (32), indeterminate (49)||17–68||22||99% ethanol + paclitaxel (3 mg/mL)||62%||1 (||Pancreatitis (|
|DeWitt et al33 (2014)||22||IPMN (55), MCN (27), SCN (18)||15–43||22||100% ethanol + paclitaxel (2 mg/mL)||50%||1 (||Pancreatitis (|
|Studies that used ethanol, paclitaxel, and gemcitabine for ablation|
|Moyer et al27 (2017)||39||MCN (23), IPMN (69), indeterminate (8)||16–42||19, 22||3 mg/mL paclitaxel + 19 mg/mL gemcitabine with/without 80% ethanol||64%||Pancreatitis (|
EUS, endoscopic ultrasound; MCN, mucinous cystadenoma; IPMN, intraductal papillary neoplasm; SCN, serous cystadenoma; PC, pseudocyst; NA, not available..
*Presumed diagnosis according to various modalities..
†Mean or range..
‡Abdominal pain was not considered an adverse event..
Abdominal pain is the most common adverse event. However, acute pancreatitis, peritonitis, and venous thrombosis have also been reported. Overall, adverse events occur in 2%–10% cases, and most are managed with conservative care.23 Ethanol seems to be the culprit of adverse events. Therefore, Moyer et al27 conducted a prospective, randomized, controlled trial to evaluate an alcohol-free pancreatic cyst ablation protocol for mucinous-type pancreatic cysts. A total 39 patients were managed with EUS-guided pancreatic cyst lavage with either 80% ethanol (control) or normal saline (alcohol-free group). Finally, an admixture of paclitaxel and gemcitabine was infused into the cyst in both groups. The complete ablation rate was 67% in the alcohol-free group and 61% in the saline group 1 year after the procedure. The rates of serious adverse events in the control and alcohol-free groups were 6% (1/18) and 0% (0/21), respectively, implying reduced procedure-related adverse events. Another study used lauromacrogol, a sclerosant with a mild anesthetic effect, as the sole ablative agent and reported a complete resolution rate of 37.9%.36
The limitation of cyst ablation, similar to solid tumor ablation, is the risk of incomplete ablation. Even if the cyst shows complete resolution, sustained surveillance is needed as metachronous malignancy in the remaining pancreas is not uncommon.37,38
Radiofrequency ablation (RFA) uses high-frequency alternating current, which generates energy as heat, resulting in coagulative necrosis and cellular apoptosis in the target tissue.39 RFA can be applied using monopolar or bipolar probes. The monopolar probe is used in a closed-loop system that includes an energy generator, an electrode, a dispersive electrode (ground pad), and the patient. High-current density energy heats the target tissue through the electrode. The ground pad closes the electric circuit and disperses the energy over a large area to reduce possible injury to the skin.39 In a bipolar RFA system, a ground pad is not necessary because the current oscillates two interstitial electrodes. The bipolar probe confines the current flow to the area between the electrodes, reducing the heat-sink effect (perfusion-mediated cooling).40 Therefore, heat injury of the lesion occurs more quickly with less damage to the surrounding normal tissue.39,41
RFA has been used percutaneously and intraoperatively to treat various tumors, including hepatocellular carcinoma, renal cell carcinoma, and lung cancer.2 However, percutaneous RFA could not be applied in lesions with interposition of organs and/or vessels. EUS-guided RFA (EUS-RFA) offers real-time imaging of the target lesion, in which RFA may enable safe tissue ablation. Recently, several reports have demonstrated that EUS-RFA is technically feasible, safe, and relatively effective for the treatment of pancreatic tumors such as unresectable pancreatic cancer and benign solid pancreatic tumors.2,42,43 EUS-RFA can also potentially be used for ablating PCLs.42,44 The application of EUS has an advantage in terms of minimal invasiveness and tolerability.
Currently, several types of probes, including the 19-gauge EUS-FNA needle electrode (Radionics, Burlington, MA, USA), HabibTM EUS-RFA catheter (EMcision, London, UK), and EUSRA RF electrode (STARmed, Goyang, Korea), are available for EUS-RFA. Among these probes, EUSRA has an internal cooling system, which helps prevent charring of the electrode surface, thus enabling efficient transmission of heat.
Before EUS-RFA, administration of prophylactic antibiotics is usually required. After identifying the target lesion, the needle electrode is inserted into the target lesion under EUS guidance. The echogenic needle tip is positioned at the far end of the target lesion. RFA usually starts at the right distal part of the target lesion on the EUS image while the needle electrode is withdrawing, the needle electrode is subsequently reinserted, and ablation is repeated at the left side of the previous ablation site on the EUS view. After inserting the needle electrode, the radiofrequency generator is activated to deliver ablation power. Energy delivery is controlled automatically (owing to a change in tissue impedance) or manually, depending on the system. Energy is delivered after the location of the tip of the needle electrode has been confirmed to be within the margin of the lesion using EUS. The RFA is repeated until the hyperechoic zone around the electrode is sufficiently covering the entire tumor. For larger lesions, the fanning technique can be applied to further ablate different areas within the same lesion.
The potential adverse events of EUS-RFA are thermal injury (including burns of the gastric wall), bowel injury, and peritonitis. Procedural adverse events are closely related to the duration of RFA.2 To prevent thermal injury to adjacent organs, some technical precautions, such as maintenance of a 5-mm minimum safety margin from surrounding vessels and use of a step-up approach in the case of larger lesions (> 2 cm in diameter), are required.43
For treatment response evaluation, CE-EUS has several advantages over CT, including lack of radiation and real-time visualization and detection of residual viable tumors. A recommended follow-up protocol after ablation is to perform CE-EUS within a week to detect residual tumors. Immediately after ablation, hyperemia develops around the ablation zone owing to tissue damage and the resultant inflammatory response. This inflammatory reaction often demonstrates a uniform rim of enhancement, which, unlike residual viable tumors, persists throughout the different enhancement phases. Therefore, postprocedure CE-EUS is recommended after at least 5 to 7 days.
EUS-RFA was first described in an animal model by Goldberg et al45 in 1999; they used this procedure to complete coagulation necrosis in a porcine pancreas. In 2009, Varadarajulu et al46 reported successful coagulation necrosis of the targeted areas using EUS-RFA of the liver in a porcine model.
In 2015, Pai et al44 reported the first human pilot study evaluating the feasibility and safety of EUS-RFA in patients with pancreatic neoplasms (PCL,
Table 3 . Summary of Representative Published Data on EUS-RFA for Pancreatic Neoplasms.
|Study||No. of patients||Indication (||RF device||Mean tumor size (range)||Application power and time||Mean RF sessions (range)||Technical success (%)||Clinical outcome (||Adverse events (|
|Pai et al44 (2015)||6||Mucinous cyst (4), IPMN (1), microcystic adenoma (1), neuroendocrine tumor (2)||Habib EUS-RFA catheter||PCL: 36.5 (24–70), NET: 27.5 (15–40)||5–25 W, 90–120 sec||1.3 (1–2)||100||2 cyst resolution, 4 cyst reduction, 50% reduction in 2 NETs with vascular changes||Mild abdominal pain (2)|
|Song et al2 (2016)||6||Locally advanced pancreatic cancer (4), metastatic pancreatic cancer (2)||EUSRA||38 (30–90)||20–50 W, 10 sec||1.3 (1–2)||100||Necrosis at the ablation site||Mild abdominal pain (2)|
|Lakhtakia and Seo39 (2017)||3||Insulinoma||EUSRA||19 (14–22)||50 W, 10–15 sec||1||100||Complete resolution of hypoglycemia||None|
|Scopelliti et al48 (2018)||10||Locally advanced pancreatic cancer||EUSRA||49.2 (35–75)||20 or 30 W, 100–560 sec||1.4 (1–2)||100||Necrosis at the ablation site||Mild abdominal pain (2), ascites (2), peripancreatic effusion (2)|
|Choi et al43 (2018)||10||NET (10), solid pseudopapillary neoplasm (2), insulinoma (1)||EUSRA||20 (8–28)||50 W||1.6 (1–3)||100||Radiologic complete response (7)||Mild abdominal pain (1), acute pancreatitis (1)|
|Barthet et al42 (2019)||29||IPMN (16), MCN (1), NET (14 lesions in 12)||EUSRA||PCL: 28 (9–60), NET: 13.1 (10–20)||50 W||NA||100||NET: radiologic complete response (12), PCL: complete response (11), > 50% reduction (1)||Acute pancreatitis (1), jejunal perforation (1), main pancreatic duct obstruction (1)|
EUS, endoscopic ultrasound; RFA, radiofrequency ablation; RF, radiofrequency; PCL, pancreatic cystic lesion; IPMN, intraductal papillary neoplasm; NET, neuroendocrine tumor; NA, not available..
In a recent preliminary study, 6 patients with unresectable pancreatic cancer underwent EUS-RFA.2 The tumors were located in the pancreatic head (
EUS-RFA has been proposed as a treatment option for benign pancreatic lesions such as PCLs and NETs. In a study by Choi et al,43 10 patients with benign solid pancreatic tumors (nonfunctional NET,
Recently, EUS-RFA has been attempted for pain alleviation in patients with pancreatic cancer. In a study by Zin et al,49 EUS-RFA of the celiac ganglia provided better pain relief in a patient with pancreatic cancer in whom initial treatment with EUS-guided CPN had failed. In a recent randomized controlled trial, EUS-CPN and EUS-RFA were compared for palliation of pain in patients with pancreatic cancer.50 A total of 26 patients underwent EUS-RFA (
Most adverse events after EUS-RFA are associated with thermal injury to the pancreatic parenchyma and surrounding structures, such as blood vessels, bile ducts, the stomach, and the duodenum. Although all recent preliminary experience demonstrates that EUS-RFA is relatively safe, there are concerns regarding postprocedural adverse events. The real-time visualization of the needle probe during EUS helps avoid puncturing of vital structures and imaging of the thermal effect localized within the tumor helps reduce major adverse events, unlike percutaneous and intraoperative RFA methods. Moreover, the lack of an internal cooling mechanism in percutaneous or intraoperative probes could result in unintended thermal injury to surrounding structures. EUS-RFA in pancreatic lesions has milder adverse events, even in the absence of internal cooling probes.
Although the available data are limited, published reports have demonstrated that EUS-RFA is a technically feasible, safe, and relatively effective modality for the treatment of different types of pancreatic tumors.
EUS-guided ablation is a promising, minimally invasive treatment that may provide targeted, individualized therapy for patients who are not candidates for surgical resection. Given the promising results of preliminary studies, EUS-guided ablation treatment can potentially change the clinical practice in the management of pancreatic neoplasms. Large-scale prospective randomized controlled trials are needed to verify the role of EUS-guided ablation in patients with pancreatic neoplasms.
No potential conflict of interest relevant to this article was reported.
© The Society of Gastrointestinal Intervention. Powered by INFOrang Co., Ltd.