Introduction

While lesions of the ampulla of Vater are relatively rare, increasingly, gastroenterologists are finding, managing, and resecting ampullary and periampullary lesions. Neoplasms of the ampulla of Vater are indeed uncommon in the general population with multiple historic autopsy studies showing an overall prevalence of between 0.04 and 0.12% [1,2,3••, 4]. Though seemingly insignificant compared with the prevalence of colon cancer, breast cancer, and lung cancer, periampullary neoplasms hold an important but under-recognized place in the realm of gastrointestinal malignancies. Periampullary tumors now account for 5% of all newly diagnosed gastrointestinal tumors [5].

Periampullary tumors generally develop later in life and most commonly are brought to clinical attention in the sixth to eighth decades of life; they likely arise years or decades earlier and grow silently without clinical detection [6,7,8]. In recent years, there has been observed a significant increase in the incidence or periampullary lesions. This has been attributed to increased endoscopic screening of high-risk patients and increased use of upper endoscopy for unrelated indications [9••, 10,11,12,13,14]. Thankfully, this has not only resulted in an increased incidence but has also resulted in the detection of lesions at an earlier point in their development, when they are more amenable to endoscopic resection [6, 15].

While most ampullary lesions are currently discovered incidentally, a significant portion of patients will still present with a myriad of symptoms including biliary obstruction, jaundice, pancreatitis, non-specific abdominal symptoms, weight loss, abdominal pain, dyspepsia, malaise, or anorexia [3, 4, 16, 17, 18••]. No specific symptom or constellation of symptoms has been shown to be pathognomonic for ampullary lesions, and so, physicians should keep these lesions in mind when seeing a patient with atypical abdominal symptoms.

Ampullary adenomas can occur in the setting of genetic predisposition, as is the case in patients with familial adenomatous polyposis syndrome (FAP) or, more commonly, occur sporadically [3, 9, 19]. Similar to what is now widely known to occur with colon polyps, ampullary adenomas progress through an adenoma to carcinoma sequence, which allows lead time for discovery and excision of these lesions prior to malignant transformation [14, 18, 20, 21, 22••].

In this article, we will look at the latest in diagnosis, management, and surveillance of ampullary adenomas and describe our own experience at a large, tertiary referral cancer hospital.

Familial Adenomatous Polyposis Syndrome

Familial adenomatous polyposis syndrome is a well-described condition resulting from an autosomal dominant genetic mutation in the adenomatous polyposis coli (APC) gene [23, 24]. APC is a tumor suppressor gene located on long arm of chromosome 5 [24, 25]. This mutation results in the growth of adenomatous polyp in multiple locations throughout the body, but is most well-known for causing prolific growth in colonic polyps. These numerous polyps result in a nearly 100% chance of developing colon cancer, which is why prophylactic colectomy is now the standard of care [26, 27].

While the focus of the general population and most physicians has been on how FAP impacts an individual’s risk of developing colon cancer, most are unaware the syndrome also conveys a significantly increased risk for cancers of the duodenal, pancreas, biliary tree, brain, and thyroid [28]. The duodenum is the second most common site of polyp formation in FAP patients with a lifetime risk of developing duodenal adenomas approaching 100% in some studies [28,29,30,31,32,33,34,35,36,37,38]. Duodenal adenomas typically show up approximately 15 years after the development of colon polyps and have a tendency to be found in clusters around and just distal to the ampulla of Vater [29, 31, 38,39,40].

Patients with FAP have a 100–330-fold increase risk of duodenal adenocarcinoma compared with the general population with an absolute lifetime risk of developing duodenal adenocarcinoma or ampullary carcinoma of 3–5% [10, 29, 30, 38, 41, 42]. Duodenal and periampullary cancers have become a leading cause of death for FAP patients since prophylactic colectomy became the standard of care [28, 42,43,44]. Thankfully, screening and early resection of these adenomas have become more widespread. The current European Society of Gastrointestinal Endoscopy (ESGE) recommendations suggest starting endoscopic duodenal surveillance at age 25 and continuing at intervals determined by the characteristics of previously found polyps [45]. Adequate evaluation of the duodenum can be obtained with a forward-viewing and side-viewing endoscope with or without the use of chromoendoscopy [22, 23]. Effective implementation of screening recommendations like the use of a side-viewing duodenoscope with random biopsies has resulted duodenal and ampullary adenoma detection rates as high as 70% in some studies [34, 46, 47]. However, taking a biopsy at the ampulla should be practiced with caution as it can cause pancreatitis. Thus, we recommend biopsying the ampulla only when there is a clear indication.

One 2012 study longitudinally followed 71 patients with FAP. The mean follow-up time was 4.5 years during which time 70 out of 71 had duodenal adenomas discovered; there were 13, greater than 1 cm in size. Forty-three had ampullary or periampullary adenomas and 17 patients underwent APC ablation with 100% technical success rate. No patients were found over the study period to have adenocarcinoma [42]. The management of ampullary lesions occurring in patients with hereditary syndromes is similar to those occurring sporadically. However, several key differences exist including the following: lesions in FAP patients are typically detected as part of intensive endoscopic surveillance and, due to their typically indolent nature; lesions in FAP patients are often monitored for progression in size and pathology while sporadic lesions are typically resected very shortly after discovery.

Endoscopic Management

While endoscopic management of ampullary and periampullary adenomas seems a daunting task, the 2015 American Society for Gastrointestinal Endoscopy (ASGE) guidelines for the role of endoscopy in ampullary and duodenal adenomas attempt to simplify and streamline recommendations for management. The ASGE recommends ERCP prior to resection attempts. These guidelines also recommend endoscopic ultrasound prior to resection as EUS has been shown to be superior to computed tomography in the accuracy of primary tumor staging [48,49,50,51,52,53,54,55,56]. We, at MD Anderson Cancer Center, first cannulate the pancreatic duct with a guidewire only without injection. Once we know where the pancreatic duct is located, we then perform, in most cases, en bloc resection.

The ASGE guidelines suggest that lesions smaller than 1 cm, in the absence of concerning features, can typically be resected without the need for EUS, but in our practice, we typically employ EUS for the evaluation of all ampullary lesions [57]. While EUS is superior for the evaluation of primary tumor staging, CT/PET and MRI should be completed prior to resection attempts as these modalities have been shown to be superior in their ability to detect seasonal small, distant metastases, and nodal metastases, respectively [55, 56].

Biopsies prior to primary resection have become standard practice for most physicians performing papillectomy; however, significant discordance rates exist between sample biopsies and final pathology from resection specimens. Several studies have examined concordance estimates that vary between 45 and 80%. A disturbingly high false-negative rate of 16 to 60% is seen in patients with adenocarcinoma. This fact is clinically relevant as endoscopic resection should only be attempted in patients with adenocarcinoma that is staged as Tis while other stages should be referred for surgical resection. This presents a challenging clinical question as this discordance results in a relatively high number of patients undergoing ampullectomy for what is thought to be adenoma but results in a final diagnosis of adenocarcinoma, and physicians and patients are then faced agonizing decision whether endoscopic surveillance or surgery is the right next step. No clear recommendations exist for this difficult clinical scenario [1, 7, 58,59,60,61,62,63,64].

Advanced analysis of biopsy specimens for K-ras mutations, immunostaining, MicroRNA, and flow cytometry is available but their clinical utility has yet to be established and is not available at all institutions [20, 65,66,67,68,69,70,71,72].

The current recommendations suggest the removal of ampullary adenomas < 4–5 cm in maximal diameter is advisable. Resection of ampullary adenocarcinoma has been described and resection of Tis tumors can be curative although the most recent ASGE guidelines conclude that, “although endoscopic removal of ampullary adenocarcinoma has been described, this cannot be endorsed for routine management.” [21, 73,74,75]. Certain endoscopic features suggestive of possible malignancy like non-lifting with injection, firmness, ulceration, and areas of depression have been proposed as indications to refer for surgical resection [21, 76].

Evidence of intraductal lesion extension is often considered to be an indication for referral to surgery, although intraductal RFA is changing this somewhat [4, 77, 78]. While injection for lifting of lesions prior to resection is commonplace in other areas of the gastrointestinal tract, injection of ampullary lesions does not enjoy such widespread acceptance. Most publications, guidelines, and expert opinion pieces leave the decision up to individual endoscopist but note that injection has the potential to be counterproductive in some lesions and may actually cause central depression due to the tethering effect of the ductal mucosa [1]. Two recent studies showed complete resection in 81% without injection vs. 50% with injection and no difference in the relative rates of bleeding or pancreatitis [22, 79]. Recent guidelines determined there was insufficient data to recommend mandatory lifting of ampullary lesions [4, 57, 78].

Techniques and Strategies to Improve Endoscopic Resection

Although the technique for optimal resection is largely a matter of personal preference, endoscopist comfort, and available equipment, some strategies have been shown to be superior in ampullary resection. There is no evidence for one type of electrocautery setting over another or one electrocautery generator over another [4, 74, 77, 78, 80, 81].

If the pancreatic duct (PD) stenting is desired, several strategies exist to aid in the PD stent placement. Some centers describe pancreatic duct cannulation and injection of a dilute solution of methylene blue prior to ampullary resection to aid in the identification of the PD for subsequent stenting [22, 82]. Other endoscopists encourage stenting of the pancreatic duct prior to resection. Still, others recommend wire-guided resection whereby the PD is first cannulated, and then, the snare is advanced over the wire and resection occurs “over the wire” [83, 84]. Following resection, the “donut” of tissue can be cut off around the wire with a needle knife, if it does not fall off on its own.

While a multitude of options and opinions exist on the “right way” to go about resection, we have found that cannulation of the pancreatic duct prior to resection without injection is helpful to obtain and fluoroscopic roadmap as to the trajectory of the PD. We then remove the guidewire, resect the ampulla, and then reattempt cannulation based on the information gained from the initial cannulation (Figs. 1 and 2). This ensures en bloc resection (Fig. 3).

Fig. 1
figure 1

En bloc resection of ampullary mass was performed. The specimen was grabbed by the snare that was used for resection

Full size image
Fig. 2
figure 2

Selective cannulation of the pancreatic duct was performed post-ampullectomy. Note that the orifices of the pancreatic duct and bile duct were separated post-ampullectomy

Full size image
Fig. 3
figure 3

Arrow, ampullary os. en bloc resection was performed. The specimen was pinned in anatomical position with 12 o’clock pointing to the proximal end

Full size image

To Stent or Not to Stent?

The topic of ductal stenting has been a point of the controversy of late. Several papers recommend prophylactic pancreatic duct stenting as it has been shown to decrease the rate of post-procedure pancreatitis [85,86,87]. Still, other sources have suggested that PD stenting may not be necessary at all [18, 88]. While the role of pancreatic duct stenting related to post-procedure pancreatitis may still be in question, its potential benefits of decreasing late ductal stenosis and allowing for safer use of coagulation therapies seem to justify the use of prophylactic PD stenting [4, 21, 75, 80, 89]. In our practice, we do routinely place PD stents and then remove them about 4 weeks later. This also allows for early inspection of the ampullectomy site. We have found no compelling evidence to support pancreatic sphincterotomy, but we perform biliary sphincterotomy, carefully watching the landmarks post ampullectomy [57, 76, 77, 81, 86, 89] (Figs. 4 and 5).

Fig. 4
figure 4

The pancreatic duct was stented. Biliary sphincterotomy was performed and a plastic stent was placed

Full size image
Fig. 5
figure 5

Fluoroscopic image of PD stent and cannulation of the bile duct

Full size image

We recommend that retrieval of specimens should be done as quickly as possible to avoid peristalsis removing them from endoscopic reach. Once the specimen is grabbed with the snare that was used to perform ampullectomy, the specimen is brought into the stomach; the specimen can be completely brought out or be dropped in the stomach until it is removed at the end of the procedure. Adequate administration of glucagon can help decrease motility of the small bowel but, again, the degree of impact is debatable and should be determined by individual physicians.

Endoscopic Resection Success Rates

While individual endoscopist experience with ampullectomy varies, the quoted technical success for complete resection also varies quite significantly from 46 to 92% [4, 21, 74, 75, 78, 80, 89,90,91]. One difference to be acutely aware of is some of the studies who quote success rates toward the higher end of this range define success as complete endoscopic resection, regardless of how many sessions are required. While these numbers may be promising, recurrence rates are also noted to be between 2 and 33% and can occur late following resection [4, 21, 92,93,94]. Though this is certainly a broad range, more recent analyses with larger sample size have somewhat narrowed this range from 8 to 19%.

Patient and lesion factors play a large role in the likelihood of technical success, and so, patient selection plays a critical role in identifying cases in which success is probable. Predictors of success are male sex, age > 48, lesion size < 24 mm, and absence of FAP [1].

Potential procedural complications include perforation, pancreatitis, bleeding, ductal stenosis, and cholangitis (though few reports of this are noted). Overall complication rates are low (between 9 and 21%) and post-procedure pancreatitis is also infrequently reported (between 5 and 19%) [4, 21, 92,93,94]. It is worth noting when discussing this with patients that the adverse event rate for surgical resection is higher than that for endoscopic resection [95]. To minimize the adverse event rate for endoscopic resection, most endoscopists keep patients on a liquid diet 24 to 48 h and also recommend a proton pump inhibitor twice a day for at least 2 weeks.

What Really Is Proper Post Procedure Screening?

In the review of the literature and in clinical practice, it seems nearly universal that all patients are asked to follow up in 1 week to 1 month. Frequently, if this first endoscopy shows no residual lesion, routine endoscopic follow-up is scheduled at either 3-month or 6-month intervals [4, 21, 74, 75, 89, 91, 96, 97]. This typically is continued to 24 months but some institutions continue screening for much longer periods. This topic certainly requires further study to determine the optimal timing and duration of post-procedure surveillance.

As aforementioned, we routinely bring the patient back to remove the PD stent in 4 weeks, and then the 3 months from the ampullectomy to assess the ampullectomy site. If there are no residual tissues or recurrence, the next endoscopy is done in 1 year. We then annually follow the patients at least for 5 years.

Conclusions

Lesions of the ampulla, specifically ampullary adenomas, are increasingly seen in clinical practice with the proliferation of routine upper endoscopy. As such, it is of paramount importance that gastroenterologists familiarize themselves with this important topic and are well versed in both lesions occurring as part of a larger, heritable syndrome (like FAP) and also sporadically. Greater than two decades of experience with endoscopic ampullectomy have shown this procedure to be safe and effective for the management of most ampullary adenomas. Recent studies have shown complete endoscopic resection rates as high as 92% with a relatively low complication rate. Pre-resection biopsy to final pathology concordance and recurrence rates remain the areas of continued emphasis with room for improvement; however, effective and diligent use of post-resection surveillance and introduction of new modalities such as intraductal RFA are making endoscopic ampullectomy more promising than ever.