Keywords

Introduction

Adenomas are the most common benign tumor of the ampulla of Vater (Fig. 54.1), with an incidence ranging from 0.04 to 0.12 % in autopsy series [1]. Ampullary adenomas, which may occur sporadically or more commonly, in association with familial polyposis syndromes, are classified histologically as tubular, villous, or tubulovillous. As they are thought to follow a similar adenoma-carcinoma sequence well-established in colonic adenocarcinomas [2], when identified, ampullary adenomas should be considered for resection . This is particularly the case for the villous subtype, which tend to be more aggressive and more prone to malignant degeneration with a reported incidence of such ranging between 22 and 56 % [1].

Fig. 54.1
figure 1

Diagrammatic illustration of the papilla of Vater. Different primary sites of neoplastic lesions are shown: ampullo-biliary segment (Ab), ampullo-pancreatic segment (Ap), ampullo-pancreatico-biliary segment of the common channel (Ac), ampulloduodenum (Ad). Neighboring structures are the choledochal duct (Dc), pancreatic duct (Pd), pancreatic head (Ph), and duodenum (D) (Reproduced with the permission of Springer Science + Business Media. Zhou et al. J Hepatobiliary Pancreat Surg. 2004;11:301–309)

Full size image

Although pancreaticoduodenectomy is considered standard treatment for ampullary carcinomas, controversy exists regarding the most appropriate method for treating ampullary adenomas, with several studies suggesting that many of these tumors may be successfully treated with open transduodenal or endoscopic excision.

Studies in support of pancreaticoduodenectomy cite the risk of occult malignancy and the risks of recurrence as two major reasons to pursue a more radical resection [3]. In addition to decreased morbidity and mortality compared with pancreaticoduodenectomy, proponents of transduodenal ampullectomy and in particular endoscopic resection insist that the risks of recurrence is relatively low and that in the event of recurrence or the presence of occult malignancy, that the majority of these lesions can subsequently be treated with open or endoscopic re-excision or pancreaticoduodenectomy [4–6].

Therefore, the aim of this chapter is to compare the morbidity and mortality and risk of recurrence between pancreaticoduodenectomy , transduodenal resection , and endoscopic ampullectomy in the treatment of villous adenomas of the ampulla of Vater . This chapter will focus on sporadic ampullary adenomas and attempts will be made to identify those factors that may assist in determining which method of treatment would provide the most benefit for each individual patient.

Search Strategy

A literature search of English language publications from 1994 to 2014 was used to identify published data on the management of villous adenomas of the ampulla of Vater with either pancreaticoduodenectomy or open or endoscopic ampullectomy . This was accomplished using the PICO outline (Table 54.1). Databases searched were PubMed, Embase, and Web of Science. Terms used individually and in combination in the search were “ampullary villous adenoma,” “ampulla of Vater adenoma,” “pancreaticoduodenectomy,” “transduodenal ampullectomy,” and “endoscopic ampullectomy.” Articles that addressed the differences in morbidity and mortality , risk of recurrence, and factors that influence the success of each treatment modality were of particular interest. Case reports, publications written in languages other than English, and those that fell out of the aforementioned date range were excluded. Fourteen retrospective cohort studies, two prospective cohort studies, and one systematic review were included in our analysis. The data was classified using the GRADE system.

Table 54.1 PICO table for management of villous adenomas of the ampulla of Vater
Full size table

Morbidity and Mortality

Although historically pancreaticoduodenectomy was the mainstay of treatment for ampullary adenomas, there has been a push towards using transduodenal or endoscopic resection as a primary means of treating these benign lesions. This shift in treatment paradigm was influenced by the increased morbidity and mortality associated with pancreaticoduodenectomy. Although perioperative mortality associated with pancreaticoduodenectomies in experienced hands at high-volume centers is less than 2 %, the procedure continues to have an operative morbidity ranging from 30 to 50 %. Furthermore, these complications, which include delayed gastric emptying, pancreatic fistula s, and biliary leak, tend to be more severe and of longer duration in comparison to those encountered with open or endoscopic ampullectomy [7].

A number of studies have shown that transduodenal ampullectomy is associated with lower rates of morbidity compared to pancreaticoduodenectomies (Table 54.2). In a study by de Castro et al. which compared the short-term outcomes and long-term survival in 145 patients undergoing either local or transduodenal (LR) resection versus pancreaticoduodenectomy (PD) for ampullary neoplasms, the mean operative time (LR 141 min ± 34.7 vs. PD 278 min ± 81.9, p < 0.001) and hospital length of stay (LR 13 days ± 6.7 vs. PD 23 days ± 21.8, p = 0.032) were significantly shorter for those undergoing local resection. Furthermore, although mortality was nearly equal for both modalities, perioperative morbidity was found to be significantly lower in the local resection group (LR 27 % vs. PD 52 %, p = 0.035) [4].

Table 54.2 Overview of the morbidity , mortality , and risk of recurrence with transduodenal resection of ampullary adenomas
Full size table

These findings were similar to a retrospective review by Clary and colleagues, for which mean operative times (LR 169 min vs. PD 268 min, p = 0.04), estimated blood loss (LR 192 mL vs. PD 727 mL), average length of stay (LR 10 days vs. 25 days, p < 0.01), and overall complication rates (LR 29 % vs. PD 78 %, p < 0.01) were lower for those patients undergoing transduodenal resection for ampullary neoplasms [5].

Endoscopic mucosal resection (EMR) and the use of ablative therapies like argon plasma coagulation, laser, and bipolar electrocautery, are among the techniques employed for the endoscopic removal of ampullary adenomas. Even when performed by the most experienced endoscopists, the complications seen after endoscopic papillectomy are higher compared to other endoscopic procedures and include acute pancreatitis , bleeding and less commonly cholangitis , papillary stenosis, and perforation [6]. In an effort to reduce some of these complications, pancreatic or biliary sphincterotomy is often performed after papillectomy along with placement of a pancreatic stent. In a prospective randomized study of prophylactic stent placement following papillectomy, there was a statistically significant decrease in the rate of pancreatitis in patients who received a stent (unstented 33 % vs. 0 % stented, p = 0.02) [8].

Studies comparing endoscopic excision to operative resection with either transduodenal resection or pancreaticoduodenoctomy consistently show that endoscopy is associated with lower rates of morbidity and mortality (Table 54.3). Furthermore, although not without risk, the periprocedural complications of endoscopy are typically mild and short-lived.

Table 54.3 Overview of the morbidity , mortality , and risk of recurrence with endoscopic resection of ampullary adenomas
Full size table

In a study by Onkendi and colleagues, there was a 58 % complication rate in those undergoing open resection with either transduodenal ampullectomy or pancreaticoduodenectomy compared to only 29 % in those undergoing endoscopic excision (p < 0.001). In addition, these complications were less severe with the most common being that of gastrointestinal bleeding for which all were managed conservatively. Of the 130 patients treated endoscopically, only 2 sustained more significant complications of ampullary obstruction or perforation, both of which were successfully managed nonoperatively as well [9].

Recurrence

Despite a mean success rate of 82.2 % [10] and lower morbidity , systematic reviews of endoscopic ampullectomy continue to demonstrate higher rates of recurrence, ranging from 0 % to 30 % [11]. These findings are summarized in Table 54.3. The same holds true for transduodenal excisions as illustrated in Table 54.2. In this study, among the 50 patients with benign adenomas managed by transduodenal excision, 17 (32 %) experienced a recurrence at 5 years and 43 % at 10 years compared to none of the patients in the pancreaticoduodenectomy group. Although the majority of the recurrences were benign and amendable to endoscopic resection , 4 of the 17 were characterized by invasion, requiring subsequent pancreaticoduodenectomy [12].

Onkendi et al. found a fivefold greater risk of recurrence after endoscopic resection compared to operative resection with either approach (32 vs. 6 %, p = 0.006) but when comparisons were only made between those patients undergoing open versus endoscopic ampullectomy , there was no difference in the rate of recurrence (33 vs. 32 %, p = 0.49). Furthermore, the majority of recurrences occurred in adenomas greater than 3.6 cm in size, ones containing foci of high-grade dysplasia, carcinoma in situ, and those cases in which more than one endoscopic procedure was required to obtain a complete excision [9].

These findings emphasize the necessity for repeat endoscopic examinations although the exact frequency and duration of surveillance endoscopy has yet to be established.

Risk Factors

As previously eluded to, certain factors may decrease the likelihood of achieving a complete excision and increase the incidence of recurrence in those undergoing less invasive means of resection .

Ridtitid and colleagues discovered that for those patients undergoing endoscopic ampullectomies between 1995 and 2012 at a large tertiary medical center, that the presence of jaundice (27.7 % vs. 4.5 %, p < 0.0001) and intraductal extension were associated with higher rates of incomplete resection (31.3 % vs. 9 %, p = 0.0002). In addition, tumors that were able to be removed en bloc as opposed to piecemeal had a significantly higher probability of being completely excised (57.5 % vs. 22.9 %, p < 0.001) [13]. Although there was no significant difference in this study with regards to tumor size, this may have been affected by the piecemeal fashion in which many of the tumors were removed.

There are no definitive guidelines as to the size above which an attempt at endoscopic excision should be avoided. Although it has been advised by many that open or endoscopic excision should not be attempted for lesions greater than 3 cm [9, 14] there have been reports of endoscopic success for tumors greater than 4–5 cm and up to 7.5 cm with transduodenal resection [15, 16]. However, the majority of tumors presenting with pancreaticobiliary symptoms, intraductal extension, and ones not amendable to en bloc resection, tend to be large in size [13]. Furthermore, size is an important predictor of endoscopic success with the highest rates achieved for lesions less than 24 mm in a large, multicenter study by Catalano et al. [17].

These characteristics, along with adenomas that are firm, ulcerated, and friable are more commonly seen in tumors harboring high-grade dysplasia, carcinoma in situ, and foci of invasive malignancy. These findings therefore suggest that the most appropriate treatment for such ampullary tumors may be that of a full oncologic resection with a pancreaticoduodenectomy [9].

Although a few, small series have reported technical success in the endoscopic removal of adenomas with high-grade dysplasia and foci of well-differentiated T1 adenocarcinoma [18–21], the reality remains that such lesions have higher rates of incomplete resection and recurrence when managed endoscopically.

This notion is supported by a study by Kim et al. which demonstrated a co-existence of cancer in 50 % of patients with pre-procedural high-grade dysplasia undergoing endoscopic ampullectomy , compared to only 15.7 % in those with low-grade dysplasia. Likewise, the rates of recurrence in the high versus low-grade dysplasia groups were 80 % and 5.2 % respectively [22].

The same argument can be made for invasive malignancies treated with transduodenal resection . In fact, Roggin et al. showed a 0 % recurrence-free survival after 2 years in the open ampullectomy group versus 48 % in those undergoing pancreaticoduodenectomy (95 % CI 37–60 %). In addition, the 2-year estimated disease free survival was 58 % versus 78 % (95 % CI 22–95 %) in the open ampullectomy and pancreaticoduodenectomy groups respectively [23]. This is likely due to the presence of lymphovascular invasion and intraductal infiltration observed in 20–40 % of T1 tumors [21].

Recommendations

Adenomas of the ampulla of Vater are rare tumors with tremendous potential for malignant degeneration into ampullary carcinomas. Therefore, when encountered, effort should be made to excise these lesions prior to the development of dysplasia or invasive cancer . Despite the three available treatment options available, the fact that we lack a clear consensus on how to approach ampullary adenomas was the inspiration for this chapter.

Based on a thorough review of current literature available on this topic, we recommend surgical excision rather than endoscopic excision for the following:

  1. 1.

    Lesions >3 cm in size (evidence quality moderate; weak recommendation).

  2. 2.

    Evidence of high-grade dysplasia, carcinoma in situ, or foci of invasion (evidence quality moderate; strong recommendation).

  3. 3.

    Lesions that are firm, ulcerated, or friable on endoscopic evaluation (evidence quality moderate; strong recommendation).

  4. 4.

    Presence of intraductal extension (evidence quality low; weak recommendation).

  5. 5.

    Lesions accompanied by pre-procedural jaundice or pancreatitis (evidence quality low; weak recommendation).

Furthermore, in the absence of endoscopists skilled in performing ampullectomies and in which there is failure of complete resection with endoscopic attempts, surgical excision is advised. Although there is significantly less morbidity associated with transduodenal resection, we favor surgical resection with pancreaticoduodenectomy as a result of the high recurrence rate observed with open ampullectomy. Endoscopic resection should be attempted therefore, in smaller lesions including those with evidence of low-grade dysplasia and which lack the aforementioned high-risk characteristics. Limiting endoscopic excision to adenomas 3 cm or less will ensure successful removal with a decreased risk of recurrence. Open or endoscopic ampullectomies may prove to be better options, however for patients whose underlying health and comorbidities render them incapable of tolerating more extensive surgery . If either approach is taken, patients should be followed closely for recurrence with surveillance endoscopy (Fig. 54.2).

Fig. 54.2
figure 2

Algorithm for the management of ampullary adenomas. EGD Esophagogastroduodenoscopy, ERCP Endoscopic Retrograde Cholangiopancreatography, EUS Endoscopic Ultrasound, HGD High-grade dysplasia, LGD Low-grade dysplasia, Tis Carcinoma in situ

Full size image

While there are no established guidelines on the frequency and duration of endoscopic surveillance , we support the recommendations made by the Standards of Practice Committee of the American Society for Gastrointestinal Endoscopy. The committee suggests an initial surveillance exam 1–6 months following the initial ampullectomy followed by repeat endoscopies every 3–12 months for at least 2 years with periodic exams thereafter based on symptoms [24]. If feasible, it may be of benefit to extend this surveillance period beyond 2 years in consideration of the findings by Farnell et al. who observed recurrences 5–10 years after initial treatment [12].

Finally, if the lesion appears amendable to endoscopic excision, consideration should be made towards the placement of prophylactic pancreatic stents to decrease the incidence of post-procedural pancreatitis [8].