Introduction

Neoplasms of the major duodenal papilla (MDP) are rare gastrointestinal tumors arising from MDP ampulla distal to confluence of common bile duct and pancreatic duct. Incidence of these neoplasms is 0.2—0.5 per 100,000, i.e. 0.3—0.5% of all gastrointestinal tumors [1—3]. Benign adenomas are found in 65-76% of cases, malignant adenocarcinoma — in 10—23% of patients [4]. Some authors consider adenoma as a precancerous condition. Incidence of malignant transformation is similar to that for colorectal cancer (about 30%) [5].

In most cases (53-93%), MDP neoplasms do not have clinical manifestations that significantly complicates their early diagnosis [6—8]. Tumor enlargement is followed by obstructive jaundice (7.1—86.0%), pain syndrome (6.9—50.9%), acute pancreatitis (1.9—6.3%) and cholangitis (0.3—1.6%) [3, 6, 9].

Currently, the main diagnostic methods are endoscopic examination with biopsy, endoscopic ultrasonography (EUS) and retrograde cholangiopancreatography (ERCP). Combination of these methods makes it possible to assess the type and extent of lesion, submucosal invasion, lymph node metastases and intraductal invasion [10].

EUS seems to be a more informative method for diagnosis of small tumors and can diagnose invasive adenocarcinoma that is not recognized by ultrasound and computed tomography (CT). Moreover, EUS ensures assessing the relationship with periampullary structures, spread to pancreatobiliary ducts, regional lymph nodes and stage of disease [11—16].

To date, duodenoscopy with biopsy is mandatory to verify the diagnosis [17]. However, false-negative results are obtained in 16—30% of cases, false-positive results — in 12—38% of cases [18, 19]. Thus, endoscopic papillectomy (EP) is considered as a variant of total biopsy for the final diagnosis [20].

Computed tomography (CT) is one of the methods for differential diagnosis of biliopancreatoduodenal tumors and usually used to clarify relationships of tumor with adjacent vessels and organs [21].

Laboratory diagnosis of tumor includes serum CA 19—9 as a specific marker of the neoplasm. However, this method is not specific because its increase can also be observed in patients with bile stasis in the bile ducts [22].

To date, resection is the only radical method for MDP neoplasms [3, 23]. Various resections via laparotomy or laparoscopic approach have been developed (pancreaticoduodenectomy, duodenotomy followed by papillectomy). However, these procedures are often complicated by wound infection (17.3—41.7%), leakage of pancreaticodigestive (3.7—21.4%) or biliodigestive (1.8—6.3%) anastomosis, intra-abdominal bleeding (1.8—16.7%) and abscesses (6.9—15%) in early postoperative period. Postoperative mortality is up to 10% [17, 24—27].

In recent years, minimally invasive resection of MDP neoplasms (intraluminal endoscopy) have been actively introduced. The advantage of this approach is low incidence of postoperative complications (bleeding 7.2—18%, acute post-manipulation pancreatitis 4.4—9%, retroduodenal perforation 1.8—8.8%). Postoperative mortality is less than 1% [6, 7, 23, 28].

However, there is no consensus on clear indications for endoscopic treatment of MDP neoplasms and optimal surgical strategy.

The purpose of the study was to develop the indications and effectiveness of treatment of patients with obstructive jaundice following MDP neoplasms.

Material and methods

There were 26 patients with MDP neoplasms at the Sklifosovsky Research Institute for Emergency Care for the period 2015—2020 (11 men (42.3%) and 15 women (58.7%) aged 68.5 (53; 89) years).

The diagnosis was confirmed by complaints, instrumental (duodenoscopy, EUS, ultrasound and CT), laboratory (serum total, direct bilirubin, alpha-amylase, CA 19-9 marker) data and verified by histological examination.

During endoscopic examination, we determined location and dimensions of tumor, examined its surface in "white light" and NBI mode (Narrow Band Imaging) and performed biopsy. In NBI mode, benign MDP neoplasms were characterized by compact small rounded fossae or large elongated fossae. Malignant neoplasms were characterized by irregular or amorphous fossae and vascular patterns. MDP biopsy was also performed during duodenoscopy.

EUS was used to analyze dimensions, invasion of tumor into duodenal wall, diameter of the common bile duct, dimensions and structure of regional lymph nodes, adjacent tissue and vessels. Tumor vascularization was assessed using Doppler ultrasound. In case of abnormal vascularization, we found active blood flow in the neoplasm with large afferent artery. Regional lymphadenopathy was also a sign of malignant growth.

During abdominal ultrasound, we analyzed calculi in the bile ducts and gallbladder, their dimensions, signs of biliary (enlargement of CBD over 10 mm and lobar ducts over 6 mm) and intraductal pancreatic (pancreatic duct enlargement over 3 mm) hypertension, level of common bile and pancreatic duct obstruction.

CT was used to evaluate extrahepatic bile ducts, duodenum, pancreas, peripancreatic tissue and lymph nodes.

We applied the classification by V.D. Fedorov et al. (2000) to assess hyperbilirubinemia (mild hyperbilirubinemia — total bilirubin < 100 µmol/L, moderate hyperbilirubinemia — 101—200 µmol/L, severe hyperbilirubinemia — over 201 µmol/L) [29]. Serum alpha-amylase was valuable to assess severity of acute pancreatitis in patients with appropriate complaints.

Percutaneous transhepatic microcholecystostomy or transpapillary interventions were performed for temporary decompression of the bile ducts and pancreatic duct. Lithoextraction with bilioduodenal stenting was performed for choledocholithiasis, nasobiliary drainage — for suppurative cholangitis. Pancreatoduodenal stenting was performed in patients with acute pancreatitis.

Radical surgical treatment of MDP neoplasms included endoscopic papillectomy, transduodenal excision of adenoma, pylorus-sparing pancreaticoduodenectomy with antecolic pancreaticogastroenterostomy and Roux-en-Y hepaticoenterostomy.

The indications for endoscopic papillectomy were MDP adenoma with mild-to-moderate epithelial dysplasia, location within mucous and submucous layers of the duodenum and no invasion in muscle layer and distal third of the common bile duct, no damage to regional lymphatic vessels and signs of abnormal vascularization.

Endoscopic resection was performed using a loop method in a mixed mode (Endocut). En-bloc resection was performed in patients with tumor < 2.5 cm, fragmentation — for larger neoplasms. After that, we examined resection edges in NBI mode to identify residual tissue of adenoma. If necessary, additional resection of abnormal mucosa was performed. The next stage was stenting of common bile duct and pancreatic duct in standard fashion. We applied plastic bilioduodenal stents 8.5 and 10 Fr (length 5 and 9 cm, respectively) and pancreatic stents 5 and 7 Fr (length 4 and 6 cm, respectively).

The indications for surgical treatment of MDP neoplasms were invasion of muscular layer of the duodenum or distal parts of the common bile duct, as well as abnormal vascularization and severe epithelial dysplasia. In this case, we performed transduodenal excision of tumor (for benign growth within duodenal wall) [30] or pylorus-sparing pancreaticoduodenectomy with antecolic pancreaticogastroenterostomy and Roux-en-Y hepaticoenterostomy (for invasion of common bile duct). Lymph node dissection was not performed in patients with MDP malignancies and no regional or distant metastases (stage T0, I, II N0 M0). Enlargement of regional lymph nodes required their dissection (within liver hilum, hepatoduodenal ligament, around the pancreas and duodenum).

The indications for palliative care were MDP malignancies and regional or distant metastases (stage T III NI, MI), severe concomitant diseases. In these cases, we performed bilioduodenal stenting only.

Results

Twenty-four (92.2%) out of 26 patients had symptoms of obstructive jaundice (itching, icteric skin and sclera, dark urine). In 2 (8.3%) of these patients, the above-mentioned signs were combined with symptoms of purulent cholangitis (hyperthermia, chills). Two (7.8%) out of 26 patients had signs of acute pancreatitis (upper abdominal pain, hyperthermia, nausea, vomiting).

Ultrasound data of biliary hypertension were obtained in 23 (88.4%) patients. In 7 patients (30.4%), bile duct enlargement was combined with calculi in the common bile duct; in 4 (17.3%) patients — with multiple calculi in the gallbladder. Three (11.6%) patients had pancreatic duct enlargement (including 2 patients with acute pancreatitis).

According to endoscopic data, tumor dimension ≥ 2.5 cm was observed in 12 patients (46.1%), less than 2.5 cm — 14 (53.9%) patients. In 16 (61.5%) cases, tumor surface was represented by adenomatous tissue with contact bleedings (Fig. 1).

Fig. 1. Adenoma of the major duodenal papilla (endoscopic image).

In 10 (38.5%) patients, growth of tumor with tuberous surface and contact bleedings was determined within MDP that allowed us to assume malignancy.

According to histological data, benign neoplasm was revealed in 16 (61.5%) patients including 11 (42.2%) ones with MDP adenoma and mild-to-moderate epithelial dysplasia, 2 (7.6%) patients with severe dysplasia, 2 (7.6%) ones with catarrhal papillitis, 1 (4, 1%) patient with Brunner's gland hamartoma. Adenocarcinoma was detected in 10 (38.5%) patients (low differentiation — 5 (19.3%) patients, moderate differentiation — 3 (11.6%) patients, high differentiation — 2 (7.6%) patients).

EUS revealed tumor growth within mucous membrane in 6 patients (23.1%), spread to muscular layer in 6 (23.1%) cases, spread to distal third of common bile duct in 14 (53.6%) cases (Fig. 2).

Fig. 2. Endoscopic US scan of neoplasm of the major duodenal papilla.

EUS signs of abnormal vascularization and regional lymphadenopathy were detected in 9 (36.4%) patients. In 17 (63.6%) cases, these signs were absent (Fig. 3).

Fig. 3. Endoscopic US scan of neoplasm of the major duodenal papilla with abnormal tumor vascularization (arrow).

EUS of pancreatobiliary zone revealed regional lymph node lesion in 6 (23%) patients. In 20 (77%) patients, regional lymphadenopathy was not observed.

Hyperbilirubinemia was observed in 19 (73.1%) cases (mild — 12 (63.1%) cases, moderate — 4 (21.1%) cases, severe — 3 (15.8%) patients). In 7 (26.9%) cases, serum bilirubin was normal.

In 2 patients with clinical and instrumental signs of acute pancreatitis, serum alpha-amylase was up to 626.0 U/L. In other cases, this indicator was normal.

CT was performed in 13 (50%) patients. Signs of biliary and/or pancreatic hypertension were revealed in 7 (54.1%) cases, MDP tumor — 2 (15.5%) cases, intraluminal duodenal tumor — 1 (7.6 %) patient, distal block of common bile duct and suspected neoplasm in this zone — 1 (7.6%) patient, pancreatic head tumor followed by common bile duct compression and biliary hypertension — in another 1 (7.6%) patient. One (7.6%) patient had no signs of hepatopancreatobiliary lesion.

Serum CA 19-9 was analyzed in 8 (30.7%) patients for differential diagnosis of benign and malignant MDP neoplasms. Four (50%) patients with adenomatous tumor had normal level of serum CA 19-9, while other 4 (50%) patients with malignant neoplasms had 2-9-fold increase of this marker.

Sensitivity of instrumental and laboratory diagnostic methods is presented in Table 1.

Sensitivity of various instrumental and laboratory diagnostic methods for ampullary tumors

Thus, EUS is characterized by the highest sensitivity (100%).

At the first stage, bilioduodenal stenting was carried out in 12 (46.1%) patients to resolve hyperbilirubinemia, papillosphincterotomy followed by lithoextraction and bilioduodenal stenting — 4 (15.3%) patients, nasobiliary drainage — 2 (7.6%) patients, pancreatoduodenal stenting — 2 (7.6%) patients. Six (23.4%) patients underwent percutaneous transhepatic microcholecystostomy. In all cases, laboratory parameters decreased in 5-7 days after surgery.

After the first surgical stage, 8 (30.7%) out of 26 patients refused further treatment due to clinical improvement and were discharged.

In 18 patients, the second surgical stage was performed after 4-8 days and normalization of serum bilirubin.

Eight (44.5%) patients underwent endoscopic submucosal papillectomy. En-bloc resection was carried out in 4 (50%) cases (tumor dimension < 2.5 cm), fragmentation — in other 4 (50%) patients (tumor dimension > 2.5 cm). Stenting of common bile and pancreatic ducts was achieved in 7 (87.5%) out of 8 cases. Cannulation of pancreatic duct was not possible in 1 (12.5%) case. In 2 (25%) cases, calculi in common bile duct required preliminary lithoextraction.

Intraluminal approach was not followed by postoperative complications. All patients were discharged after 5—7 days.

Four (22.2%) patients underwent tumor resection via laparotomy: transduodenal excision of adenoma — 1 (25%), pylorus-sparing pancreaticoduodenectomy — 2 (50%), pylorus-sparing pancreaticoduodenectomy with regional lymph node dissection — 1 (25%). One patient required re-laparotomy in postoperative period for pancreatojejunostomy leakage and bleeding from biliodigestive anastomosis and pancreatojejunostomy. This one was discharged in 32 days after redo surgery.

Six (33.3%) patients underwent bilioduodenal stenting as palliative treatment.

In all patients, endoscopic examination after 3 and 6 months did not reveal recurrent tumor growth. Biliary stents were removed (Fig. 4).

Fig. 4. Endoscopic image. Follow-up survey after 3 months (arrow indicates bilio-duodenal stent).

Discussion

Despite asymptomatic course of MDP neoplasms, these tumors can nevertheless result such urgent conditions as obstructive jaundice (hyperbilirubinemia in 73% of cases) and acute pancreatitis (7.6%). Therefore, a special vigilance is required in diagnosis and treatment of this disease in patients with urgent pathology.

EUS is a highly specific diagnostic method to determine tumor dimensions, invasion of adjacent structures and vascularization. These data are essential to choose an adequate surgical approach [17]. CA 19-9 marker increase is specific only for malignant MDP neoplasms (44%).

Considering literature data and own results, we propose the following algorithm for surgical treatment of patients with MDP neoplasms (Fig. 5).

Fig. 5. Surgical algorithm for ampullary tumors.

In our opinion, invasion of muscle layer or distal parts of common bile duct, as well as abnormal vascularization are indications for open surgery or palliative endoscopic treatment. On the contrary, tumor location within mucous and muscle layers without invasion of distal third of common bile duct and no blood flow enhancement suggest endoscopic papillectomy.

Endoscopic approach can be considered not only as a highly effective method for obstructive jaundice relief (100%), but also as final minimally invasive method of surgical treatment (100%) with minimal risk of postoperative complications in patients with benign MDP neoplasms.

Small sample size requires further research to obtain significant results.

Our data may be classified as a grade of recommendations C and level of evidence IV (according to the Oxford Center for Evidence-Based Medicine classification).

Conclusions

1. EUS is a highly specific method for diagnosis of MDP neoplasms that is essential to choose an adequate surgical approach.

2. Endoscopic treatment of obstructive jaundice is highly effective in patients with MDP neoplasms: complete clinical and laboratory regression of jaundice was achieved in 16 (66.6%) patients; other 8 (33.4%) patients reduced direct bilirubin level by more than 2 times.

3. Endoscopic treatment of benign MDP neoplasms can be considered as minimally invasive approach with minimal risk of postoperative complications.

The authors declare no conflicts of interest.