A crucial component of an EUS training program is achieving proficiency in gastrointestinal tumor staging. Where available, EUS has become the standard of care in staging several gastrointestinal malignancies including esophageal, gastric, rectal, and pancreatic cancers . Accurate imaging of the lesion and recognition of surrounding lymphadenopathy , in particular the celiac axis region for upper tract cancers, are critical to the diagnosis and correct staging of mucosally based tumors. Evaluation of rectal cancers should include intubation of the sigmoid colon and identification of the iliac vessels. A prospective study reported competent intubation of the esophagus, stomach, and duodenum was achieved in 1–23 procedures (median 1–2), with visualization of the esophageal or gastric wall in 1–47 procedures (median 10–15) [19]. Adequate evaluation of the celiac axis region required 8–36 procedures (median 10–15).

Unfortunately, there are limited studies addressing the learning curve for staging mucosal tumors of the gastrointestinal tract. Only two studies have evaluated the learning curve in esophageal cancer staging . Fockens et al [20] reported that adequate staging accuracy was achieved only after 100 examinations, while Schlick and colleagues [21] reported an 89.5 % T stage accuracy after a minimum of 75 cases. A survey of the American Endosonography Club in 1995 suggested an average of 43 cases for esophageal imaging, 44 for gastric, and 37 for rectal [19]. Once competence is achieved in one anatomic location, the threshold number of procedures for other anatomical locations may be reduced depending on the skill and training of the endosonographer. The ASGE currently recommends a minimum of 75 supervised cases, at least 2/3 in the upper GI tract, before competency for evaluating mucosal tumors can be assessed [26].

Determining the accuracy of tumor staging by a trainee is an important aspect of trainee evaluation as EUS staging potentially alters the clinical neoadjuvant/adjuvant treatment plan and surgical decisions. Studies in endosonographic staging of esophageal cancer suggested that at least 75–100 procedures were required before an acceptable level of accuracy was achieved [20, 21]. Ideally, the accuracy of EUS staging should be compared to a gold standard such as surgical pathology; however, surgical specimens are not always readily available, and any preoperative chemoradiation therapy may affect staging. In these circumstances, staging by a skilled and competent endosonographer should be considered the gold standard comparison for the trainee. The trainee should achieve accuracy in tumor staging comparable to published medical literature (Table 21.4) [8]. Appropriate documentation of all EUS procedures in a training log, along with review of surgical pathology results, will further assist in improving the accuracy of tumor staging. Furthermore, the implications of EUS findings in staging gastrointestinal malignancies must be incorporated into the whole treatment plan for each patient (i.e. surgical, medical, and/or radiation oncology referrals).

Evaluation of subepithelial lesions has become a common indication for EUS. Discriminating between neoplasms, varices, enlarged gastric folds, and extrinsic compression from extramural masses can be performed with traditional echoendoscopes or catheter-based ultrasound probes. With the advent of the catheter-based probes, some practitioners have developed competency in subepithelial abnormalities without achieving competence in other indications for EUS. Although no studies are available for determining the threshold number of cases required to accurately assess subepithelial abnormalities, the ASGE Standards of Practice Committee currently recommends a minimum of 40–50 supervised cases [30].

Most endosonographers will agree that accurate imaging and interpretation of images of the pancreaticobiliary system including the gallbladder, bile duct, pancreatic duct, and ampulla are more technically challenging than evaluating mucosal and subepithelial lesions . For this reason, a larger volume of supervised pancreaticobiliary cases are required before competence can be adequately assessed. A multicenter, 3-year prospective study reported that adequate imaging of the pancreatic and bile ducts required 13–135 cases (median 55), while imaging of the pancreatic parenchyma required 15–74 cases (median 34) [31]. Adequate assessment of the ampulla required 13–134 cases (median 54). Although technical competence in pancreaticobiliary imaging may be achieved in less than 100 cases, a survey from the American Endosonography Club suggests that interpretive competence of pancreatic images may require additional procedures (120 cases) [19]. Other expert opinion suggests a higher threshold of 150 cases before assessing interpretative competence [27]. Currently, the ASGE Standards of Practice Committee recommends a minimum of 75 pancreaticobiliary cases before competency can be assessed [26].

EUS-FNA has emerged as an important diagnostic tool for obtaining tissue from intramural lesions, peri-gastrointestinal adenopathy, pancreatic lesions, and others [32]. Training in EUS-FNA requires knowledge of basic endoscopic ultrasound principles along with mastery of the skills necessary for obtaining and interpreting EUS images. Understanding and appreciating the complexity and risk that EUS-FNA adds to the procedure is critical for successful training. A recent study suggested that introducing trainees to EUS-FNA from the onset of training is a safe and feasible way to maximize exposure to FNA during training [24]. This is the first reported study evaluating the safety and diagnostic yield of EUS-FNA by attending supervised advanced endoscopy trainees. It found similar diagnostic yield comparing attending versus fellow FNA passes when the trainee is supervised. Unfortunately, the minimum number of FNA cases needed to achieve competence has not been studied. Due to the lack of literature supporting a threshold number for EUS-FNA, these numbers were adopted from the guidelines set forth for therapeutic ERCP. The similarities between EUS and ERCP including use of side-viewing instruments and combined endoscopic and radiologic imaging led to these recommendations. The current recommendation suggests that the trainee perform a minimum of 50 EUS-FNA procedures, split between non-pancreatic and pancreatic FNA [8]. It is generally agreed that EUS-FNA of pancreatic lesions carries higher complexity and risk for potential complications than other anatomical sites. Therefore, the number of pancreatic FNAs is considered separately from other anatomical locations. Competence in EUS-FNA of pancreatic lesions requires demonstrating proficiency in at least 75 pancreaticobiliary EUS cases in addition to 25 supervised FNA of pancreatic lesions. For non-pancreatic lesions (i.e. intramural lesions, lymph nodes, and ascites), the trainee should perform at least 25 supervised FNA cases before competency can be assessed [26]. Large clinical studies are needed to further assess the validity of these recommendations.