Pancreatic cystic lesions (PCLs) are known precursors to pancreatic cancer, one of the deadliest types of cancer worldwide. Surgical removal or pancreatectomies remain the central approach to managing precancerous high-risk PCLs. Endoscopic ultrasound (EUS)-guided therapeutic management of PCLs is a novel management strategy for patients with prohibitive surgical risks. Various ablation techniques have been explored utilizing EUS-guided fine needle injection (FNI) of alcohol and chemotherapeutic agents.
Pancreatic malignancy is one of the deadliest types of cancer worldwide. With an 11.5% 5-year survival, it is responsible for approximately 8.2% of all cancer-related deaths in the US 
. Globally, it is the seventh-leading cause of cancer-related death, and by 2030, it will be the second-leading cause of cancer-related death in the US 
. While numerous established risk factors predict the development of pancreatic cancer, including older age, male sex, diabetes mellitus, obesity, and diets high in fats, the two most dominant are cigarette smoking and family history. The lack of an adequately specific screening test for the general population contributes to delayed diagnoses and increased mortality. Pancreatic cystic lesions (PCLs) are known precursors to pancreatic cancer and provide an opportunity for early intervention to prevent progression to pancreatic malignancy.
The prevalence of PCLs has been increasing rapidly in recent years, largely due to improvements in cross-sectional imaging and an aging population 
. These heterogeneous lesions are often discovered incidentally and pose significant uncertainty in management due to their varying malignant potential, with over 50% of incidentally discovered PCLs being premalignant or malignant 
. Accurate diagnosis of PCL type and risk stratification of premalignant lesions is crucial in determining whether observation or surgical resection is most appropriate. In patients with prohibitive surgical risks, local ablation of the PCL can be attempted.
Neoplastic PCLs have two main subtypes—mucinous and non-mucinous. Mucinous PCLs include intraductal papillary mucinous neoplasms (IPMNs) and mucinous cystic neoplasms (MCNs). Non-mucinous PCLs include serous cystic neoplasms (SCNs), solid pseudopapillary neoplasms (SPNs), and cystic neuroendocrine neoplasms (cystic-NETs). SCNs are the most common benign pancreatic neoplasm, with a malignant transformation rate of 1 to 3% 
. SPNs are most commonly seen in young females and make up 0.3 to 2.7% of all pancreatic tumors. Margin-negative surgical resection is generally curative; however, these lesions do have the potential to recur or metastasize 
. Pancreatic NETs arise from the islets of Langerhans and have two main subtypes—cystic and solid. Cystic-NETs make up 6.5–36.1% of all pancreatic NETs and are considered to have less aggressive behavior than solid-NETs 
. Pseudocysts are the most common non-neoplastic PCLs 
. They are inflammatory lesions that result from leakage of pancreatic fluid in the setting of acute or chronic pancreatitis 
. Unlike true cysts or cystic neoplasms, they lack an epithelial lining and may spontaneously resolve.
2. EUS-Guided Ethanol Ablation
The discovery of EUS-FNI opened the door to the potential to utilize the technique in managing PCLs. The first of those to explore EUS-FNI in treating PCLs was Gan et al., who aspirated the lesion with a needle and lavaged the cyst cavity with ethanol 
. A total of 25 patients were treated with ethanol lavage, 23 patients completed follow-up at 6–12 months, and 8 patients (35%) experienced complete cyst ablation. Although this study in and of itself did not show a remarkable resolution rate in PCL with ethanol lavage, it was a milestone study for many reasons. Firstly, it demonstrated that ethanol lavage of PCLs with EUS-FNI can be safely conducted without significant adverse events. Secondly, it proved the feasibility of EUS-FNI to be used in treating PCLs. Following this monumental study, multiple investigators lavaged PCLs using varying concentrations of EtOH (80–99%) and saw up to 85% of the complete resolution of PCLs 
. However, these follow-up studies also revealed some clinically significant adverse events associated with ethanol lavage. Dewitt et al. reported that 20% of patients experienced abdominal pain after the procedure, and 4% acquired post-ablation pancreatitis 
. The etiology of acute pancreatitis was thought to be secondary to the extravasation of ethanol out of the cyst into the pancreatic parenchyma and duct, causing inflammation.
3. EUS-Guided Chemoablation and Inclusion Criteria
The first group to explore alternative injectables to achieve an increased rate of complete resolution of PCLs with EUS-FNI was Oh et al. 
. Their choice of agent was paclitaxel, a chemotherapeutic agent that inhibits cell replication. Based on how local injection chemotherapy is used to treat localized tumors in other organ systems, such as endobronchial lesions of lung cancer and advanced ovarian cancer, Oh et al. hypothesized that paclitaxel injection into the cyst cavity would eradicate the cyst. Oh et al. employed the following inclusion criteria: (1) unilocular or oligolocular cystic tumors (oligolocular cyst was defined as having two to six locules within a cyst in a later study by Oh et al.), (2) indeterminate cystic neoplasms despite evaluation by EUS-FNA, and (3) cystic tumors that increased in size during the observation period 
. Moyer et al. used similar inclusion criteria, although they defined the eligible PCL size to be between 1 and 5 cm 
. Oh et al. defined their exclusion criteria as (1) cystic tumors that had the typical morphology of serous cystadenomas and pseudocysts (therefore including only mucinous and indeterminate cysts), (2) evidence of communication between the PCL and the main pancreatic duct, (3) overt carcinomas with peripancreatic invasion, and (4) patients with a bleeding tendency (pro-thrombin time >1.5 international normalized ratio or a platelet count <50,000/microliter) 
4. Cyst Selection
Given the variation in pancreatic cysts discussed in the introduction, the question of which types of PCLs would respond best to chemoablation has become central to the investigation of this technique. As discussed previously, each group of researchers utilized varying inclusion criteria for the PCLs they would treat with EUS-FNI chemoablation. The general consensus was the first to treat cystic lesions that are not currently of malignant quality, since those cysts would be better candidates for surgical removal. Additionally, if the patients were not good surgical candidates, EUS-FNI chemoablation could be considered. The primary purpose of EUS-FNI chemoablation is to prevent PCLs from evolving into overt malignancy. Therefore, studies have found it most effective to ablate cysts that can potentially turn malignant during the disease process, specifically IPMNs and MCNs. Indeterminate PCLs have also been included in the studies due to their unknown potential for malignancy.
Historically, the method of classifying PCLs to evaluate candidacy for EUS-FNI chemoablation has been with cross-sectional imaging and cystic fluid studies. PCLs with a honeycomb appearance on imaging were identified as SCAs, and those with parenchymal changes on imaging could be identified as pseudocysts. Cyst sizes were calculated based on cross-sectional imaging as well, both to evaluate inclusion and to determine the response to therapy. Cyst fluid was aspirated for analysis prior to ablation. Moyer et al. classified MCN as lesions with CEA greater than 200 ng/mL and amylase less than 800 U/L and IPMN as lesions with CEA greater than 200 ng/mL and amylase greater than 800 U/L 
. Oh et al. classified SCA as lesions with CEA less than 5 ng/mL and pseudocysts when CEA was less than 5 ng/ML but amylase greater than 800 U/L. Lesions with fluid studies that did not meet any of these criteria were classified as indeterminate 
. FNA of the cystic component and evaluation of its cytology lend a sensitivity of 51% and specificity of 94% for the diagnosis of malignant pancreatic cystic lesions, with its lower sensitivity due to sampling error in meta-analyses 
EUS-guided needle-based confocal laser endomicroscopy (EUS-nCLE) allows real-time microscopic visualization of the PCL epithelium 
. In addition to providing an accurate diagnosis of PCLs, EUS-nCLE has been studied for differentiating IPMNs with high-grade versus low-grade dysplasia 
. Cyst fluid molecular analysis by next-generation sequencing analysis allows accurate diagnosis of cyst type and also reliably identifies IPMNs with advanced neoplasia 
. Most of the published studies on EUS-guided chemoablation have yet to utilize accurate diagnostic tools, such as EUS-nCLE and cyst fluid molecular analysis, in PCL diagnosis and risk stratification.
Additional diagnostic measures with EUS include contrast-enhanced EUS (CE-EUS) and through-the-needle biopsy (TTNB). CE-EUS can distinguish mural nodules from mural clots in IPMNs and can potentially identify high-grade dysplasia or invasive carcinoma among mural nodules 
. Lastly, TTNB can offer higher quality histological samples than FNA to offer a more accurate diagnosis of PCLs 
5. Chemotherapy Agent Selection, Efficacy, and Safety
There have been a number of anti-tumor agents, including alcohol, oncolytic viruses, brachytherapy, and chemotherapy drugs, trialed for targeted treatment of PCLs using the EUS-FNI technique 
. The primary chemotherapeutic agents used for the management of PCLs are paclitaxel and gemcitabine, sometimes in combination with each other or preceded by alcohol lavage.
Paclitaxel binds to tubulin and promotes the formation of aberrant mitotic spindles that disrupt mitosis. Its hydrophobic nature makes it an ideal candidate for intralesional delivery without significant extravasation 
. Various formulations of injectable paclitaxel for intralesional delivery have been tested in animal models and have demonstrated high, sustained local concentrations of the drug without significant extravasation into the surrounding tissue 
. Therefore, it has been the chemotherapeutic agent of choice in the vast majority of studies evaluating the chemoablation of PCLs. A second agent, gemcitabine, has also been used in combination with paclitaxel in some studies. Gemcitabine is a nucleoside analog that promotes apoptosis of rapidly dividing malignant cells 
. The two agents synergize in animal models, as paclitaxel reduces cytidine deaminase, the primary enzyme that metabolizes gemcitabine 
Five studies have assessed the safety and efficacy of ethanol lavage with paclitaxel injection in the management of PCL. Of the 256 PCLs treated in these studies, 170 (66%) were completely resolved, and 46 (18%) were partially resolved on follow-up imaging 
. DeWitt et al. also reported the presence of K-ras mutations in aspirated cyst fluid, and noted elimination of these mutations in 8 of 11 PCLs (73%) with baseline mutations after EUS-guided ethanol lavage with paclitaxel injection (EUS-EP) 
. The most commonly reported procedure-related adverse events were abdominal pain (n
= 4, 2%), pancreatitis (n
= 12, 5%), hyperamylasemia (n
= 6, 2%), and fever without bacteremia (n
= 2, <1%) 
. The only predictors of cyst resolution were a cyst diameter <35 mm and volume <22 mL 
6. Durability of Chemoablation
As the various studies described above established the safety and efficacy of EUS-guided chemoablation for the management of PCLs, attention has turned to the durability of this intervention. Choi et al. published a study in 2017 assessing the longer-term outcomes of EUS-EP in a larger study population of 164 patients, including 71 with MCNs, 16 with SCAs, 11 with IPMNs, 3 with pseudocysts, and 3 with indeterminate PCLs. After undergoing the same intervention procedure as used in the previous studies, 114 patients had complete cyst resolution, with partial resolution in 31, and persistent cysts in 13. After the extended follow-up period of an average of 72 months, only 2 of the patients with complete cyst resolution had evidence of recurrence. Two morphologic features of the original PCLs—absence of septa and smaller (≤35 mm) cyst size—predicted resolution 
Lester et al. also sought to establish the durability of chemoablation for the management of PCLs by conducting an analysis of patients who participated in the CHARM trial and had follow-up imaging at least 12 months after the original study’s post-treatment response assessment. A radiologist reviewed each patient’s series of imaging, including baseline, 12-month follow-up, and long-term follow-up at least an additional 12 months later. Cysts were measured in multiple dimensions to calculate volume, and the response to EUS-guided chemoablation with paclitaxel and gemcitabine was determined by a reduction in volume. In total, 36 of the original 39 were included in the analysis. The mean long-term follow-up was 36.5 months from the original intervention, with a range of 20–78 months. Of the 23 patients determined to have complete resolution of their PCL in the CHARM trial, 20 demonstrated sustained resolution, whereas the other three had cyst volumes that fell just above the cutoff of 94% reduction in the original volume required to be classified as complete resolution. Additionally, 4 patients who initially had a partial response and 1 who had no response were found to have complete resolution at this later follow-up, while 2 initial non-responders achieved a partial response. Only one patient had a regression from partial response to no response 
. With this data, the authors concluded that EUS-guided chemoablation is effective in preventing PCL progression and, therefore, has the potential to spare patients from invasive surgical resection. This reinforces the utility of this technique as an option for patients who are unable or unwilling to undergo pancreatectomy.