Fig. 15.1
Pancreatography obtained by ERCP showing a small ventral pancreas (a) and two images showing increasing filling of the dorsal duct via the minor papilla (b, c)
Fig. 15.2
MRCP demonstrating pancreas divisum with dorsal pancreatic duct emptying into duodenum proximal to the major papilla where the common bile duct exits. (Courtesy: Dr. Nisha Sainani Brigham and Women’s Hospital, Boston, MA)
Endoscopic Ultrasound
EUS images the pancreaticobiliary ductal system without the need for contrast injection and has the added advantage of being able to examine the pancreatic parenchyma in detail which provides important additional information when looking for early chronic pancreatitis . Both radial and linear echoendoscopes are capable of diagnosing PD but reports vary on the diagnostic accuracy of EUS. The inconsistency in diagnosis results from the fact that while various criteria (imaging patterns) have been proposed to suggest the presence or absence of PD, any particular EUS examination may or may not detect those patterns in that specific patient. If certain patterns are seen, then PD can be definitively ruled in or out. With other criteria, increasing the number of diagnostic criteria needed to diagnose PD increases the specificity but decreases sensitivity for PD [20]. The EUS diagnostic criteria that have been evaluated include the absence of stack sign, presence of crossed duct sign, visualizing the pancreatic duct cross the dorsal-ventral anlage, and inability to follow the pancreatic duct from the major papilla to the pancreatic body [21]. The stack sign is assessed using a radial echoendoscope in the long position in the duodenal bulb with the tip at the apex of the bulb. In this position, the bile duct, pancreatic duct, and portal vein should be visualized running parallel to each other in a “stack.” The crossed duct sign is present when the dorsal duct is seen crossing the common bile duct (CBD) from the bulb with a radial echoendoscope. The absence of stack sign is suggestive but not diagnostic of PD with 50 % sensitivity and 97 % specificity. The presence of crossed duct sign seems consistent with PD, however, has limited sensitivity. Following the pancreatic duct is very useful for ruling out PD. If the duct can be definitively traced from the major papilla to the pancreatic body or from the body dipping at the genu toward the major papilla (Video 15.1), then PD is ruled out. If it cannot be seen, is that because PD is present or the duct is just tortuous and cannot be kept in the plane of imaging? Despite these potential problems, a recent study compared the sensitivity of EUS, CT, and MRCP for the diagnosis of PD and investigators found the sensitivity for EUS to be 86.7 %, which was significantly higher than that of CT and MRCP [22].
Although these investigations establish a diagnosis of PD in a large proportion of patients, PD should still remain in the differential diagnosis of patients with idiopathic recurrent acute pancreatitis despite nondiagnostic studies [19]. ERCP is the definitive test for PD, but in most cases, ERCP is reserved for those with PD in whom intervention is planned.
Case Continued
An MRCP was done and suggested the possibility of complete PD . The dorsal duct was not dilated and the biliary tract was normal. EUS using a radial echoendscope also confirmed the presence of PD as the stack sign could not be elicited and the dorsal duct could be seen crossing the CBD and entering the duodenum in the area of the minor papilla. The pancreatic parenchyma was normal revealing only echogenic strands and foci without shadowing. What should be done next?
ERCP
During ERCP, PD should be suspected if a pancreatogram cannot be obtained via the major papilla and is usually diagnostic when cannulation of the major papilla reveals a short ventral duct (Fig. 15.1). One must be careful not to misinterpret a complete duct cutoff, as seen in pancreatic cancer , with a ventral duct of PD. The ventral duct is 1–4 cm long, does not cross the spine, tapers into multiple side branches, drains promptly, and acinarization can occur quickly with contrast injection [1]. In a small group of patients with PD, the ventral duct may be absent and major papilla injection will only produce a cholangiogram [23]. These findings are suggestive but not diagnostic of PD as similar findings may be also seen in patients with benign or malignant processes that cause severe ductal obstruction. In cases where PD is suspected, minor papilla cannulation is mandatory for confirming the diagnosis.
Localization of Minor Papilla
The minor papilla is located 1–3 cm superior and anterior (cephalad and medial) to the major papilla and thus should be located in the right upper quadrant of the endoscopic field when facing the major papilla (Fig. 15.3) [1]. With the advent of newer generation video endoscopes , the minor papilla can be localized in most patients. When the minor papilla or the actual orifice cannot be identified, pancreatic juice flow can be stimulated by administering intravenous secretin. The orifice can then be identified by visualizing the resulting fountain of clear juice. In a minority of patients, even secretin stimulation is inadequate to visualize the orifice. In these cases, one can spray dilute methylene blue (1:10) in the general area of the minor papilla and then look for a clearing of the dye caused by the flow of pancreatic juice [1]. In patients where all attempts to localize the minor papilla have failed, methylene blue can be injected into the dorsal duct under EUS guidance and the flow of blue-colored pancreatic juice can help localize the minor papilla [24].
Fig. 15.3
a, b Endoscopic images showing the major and minor papilla ( arrows). The minor papilla is cephalad and medial to the major papilla
The endoscopic appearance of the minor papilla can be predictive of PD [25]. The presence of an enlarged minor papilla or an obviously patent orifice has been shown to be moderately predictive of the presence of PD . During intubation of the duodenum at ERCP, before the scope is advanced around the apex and shortened, one should look for the minor papilla as one gets a great view of the area cephalad and medial to the major papilla.
Cannulation of Minor Papilla
The minor papilla and orifice are quite small, and there is little if any intraduodenal segment. Cannulation requires precise technique and unique accessories. The endoscope should be in the “long” position [1]. This can be accomplished by turning left on the left-right knob, torqueing counter-clockwise and advancing the scope. Alternatively, one can begin in the stomach and advance the scope through the pylorus and begin the turn around the apex. Just past the apex, stop and look for the minor papilla rather than further advancing into the second portion and straightening the scope. The optimal accessories for cannulation are tapered cannulas or papillotomes with small caliber guidewires (0.021” or 0.018”). Place the guidewire 1–2 mm beyond the tip of the catheter and precisely insert the wire tip into the orifice. “Precision” is the operative word. One cannot “cram” the wire or catheter tip into the minor papilla. Several “abusive” attempts will cause edema and bleeding which render subsequent attempts futile. It is critical to appreciate the trajectory of the dorsal duct. Once the guidewire is insinuated, the cannula should be advanced flush to the orifice followed by contrast injection to opacify the dorsal duct. It is critical to obtain a ductogram to enable advancement of the guidewire in line with the course of the duct. The duct usually courses from right to left from the endoscope perspective. The initial guidewire trajectory will be perpendicular to the duct and the endoscopist must adjust the direction, under fluoroscopic guidance, to align with the dorsal duct to achieve deep cannulation. For therapy, the goal is always to insert the guidewire to the tail of the pancreas. The guidewire is guided under fluoroscopy with contrast in the duct to avoid advancing the guidewire into a side branch.
Types of Pancreas Divisum
Variations of the ductal anatomy in PD have been described elsewhere, and the clinical significance of these variants is the same as that of complete PD [7, 26]. Incomplete PD is characterized by the presence of a small, filamentous branch connecting the ventral with the dorsal duct. While the dorsal duct may opacify with injection into the major papilla, the connection is inadequate for drainage and most or all of the pancreatic secretions from the dorsal duct exit through the minor papilla [7, 26]. When only an isolated small segment of the dorsal pancreas drains via the minor papilla and the dorsal duct does not communicate with the ventral duct, patients have reverse PD. In this case, the majority of pancreas juice drains via the major papilla. This ductal variation has no clinical significance but an unaware endoscopist may repeatedly try to obtain a complete pancreatogram through the minor papilla suspecting PD [7]. A functional variant of PD has also been described where the entire pancreas including the uncinate process drains via the minor papilla and no pancreatic duct connects to the major papilla [26].
Identification of Patients with Minor Papilla Stenosis
A majority of patients with PD remain asymptomatic throughout their life. A subset will have pancreatic type pain or recurrent pancreatitis. The current understanding of the pathophysiology of PD and pancreatitis/pain hypothesizes the presence of ductal hypertension because of minor papilla stenosis as the main cause of symptoms. Conversely, relief of the obstruction by endoscopic or surgical therapy has been shown to relieve pain in the majority of patients with minor papilla stenosis. A number of imaging criteria have been suggested to identify patients with PD and minor papilla stenosis:
1.
An abnormal dorsal pancreatic duct with a normal ventral duct suggests minor papilla stenosis. Similarly, cystic dilatation of the terminal portion of the dorsal duct (Santorinicele) is also thought to suggest dorsal duct outflow obstruction [1, 7]. Although these findings on ERCP/MRCP are highly suggestive of minor papilla stenosis, they are infrequently observed.
2.
3.
4.
Collection of pure pancreatic juice from the dorsal duct after secretin stimulation and analyzing bicarbonate concentration as well as the volume of pancreatic juice secreted. This will help in diagnosing early CP localized to the draining area of dorsal duct [7].
5.
Manometry of minor papilla: There are only a handful of studies that have evaluated the minor papilla pressures as this is technically difficult and the normal pressures have not been established [1]. One study found that patients with PD had higher dorsal duct pressures compared to pressures at the major papilla of patients with normal pancreatic duct anatomy [28]. However, Satterfield et al found the basal and phasic pressures to be similar at both the major and minor papilla in 4 patients with PD and ARP [29].
6.
Once deep cannulation of the dorsal duct has been achieved, minor papilla stenosis can be subjectively gauged by observing resistance to sequential passage of 3, 4, and 5 Fr catheters [7].
7.
Trial of opening the minor papilla: The minor papilla orifice can be temporarily enlarged by placing a transpapillary stent . Such a strategy might be helpful in patients with daily or weekly symptoms but is problematic if the patient has infrequent episodes of pancreatitis because of the potential for ductal damage induced by the stent [1, 7]. Most pathology studies suggest that there is no sphincter associated with the minor papilla so injection of Botulinum Toxin is not helpful.
As evident above, there are a number of tests and imaging findings that may suggest minor papilla stenosis but their sensitivity and specificity have not been evaluated by prospective studies. Additionally, a high suspicion of a tight minor papilla does not always guarantee the success of endoscopic or surgical therapy [1, 7]. With current evidence, the presence of recurrent pancreatitis without a defined cause is the best indicator of minor papilla stenosis [30]. Some patients with typical pancreatic pain without pancreatitis will respond to minor papilla sphincterotomy , but this is in highly selected patients, and there is much less evidence to support this approach.
Endoscopic Therapy for Pancreas Divisum: Dilate, Stent, or Cut?
Endoscopic techniques for opening the minor papilla consist of dilatation by sequentially larger tapered catheters, balloons, or by stenting and sphincterotomy [1, 7]. However, balloon dilatation of a naïve papilla has been associated with a significant risk of pancreatitis as well as duct disruption and therefore should not be employed in the presence of a relatively normal or moderately dilated dorsal duct [7]. A recent small retrospective study from Asia does, however, suggest the safety and efficacy of balloon dilation in PD [31]. In their study of 16 patients with symptomatic PD from ARP or CP, balloon dilation of the minor papilla to 4 or 6 mm was successful with 85 % clinical improvement. Mean main pancreatic duct (MPD) diameter was 4.3 mm and no complications were observed from balloon dilation.
Dorsal duct stenting has been used as a therapeutic trial on both a short- and a long-term basis for relief of pain (Fig. 15.4). A prospective randomized study comparing long-term dorsal duct stenting (1 year with stent exchanged every 4 months) with conservative medical therapy in patients with PD and idiopathic recurrent pancreatitis reported significantly greater number of hospitalizations, emergency department visits, and pancreatitis episodes in patients treated conservatively [32]. Another study on the efficacy of long-term dorsal duct stenting reported the best results in patients presenting with ARP while only 13 % of patients presenting with pain alone had complete pain relief [33]. In a study of 48 patients with PD and CP, successful outcome was achieved in 96 % of patients [34]. In this study, the majority of the patients underwent a long-term ductal stenting and over a median follow-up of 67 months, 39 % required re-stenting for recurrence of symptoms. While the long-term pancreatic duct stenting may improve symptoms in PD, leaving pancreatic stents in place for weeks to months is associated with potentially serious complications including stent occlusion, migration, pancreatitis , duct perforation, and duct disruption [7]. The major concern however is stricturing of the main duct. Both ductal and parenchymal changes are seen relatively quickly after pancreatic stent placement. While these are of little significance in patients with advanced chronic pancreatitis, they can be disastrous in patients with normal pancreatic ducts [35]. Therefore, the long-term dorsal duct stenting is not advocated for patients with normal appearing ducts.
Fig. 15.4
Endoscopic image showing a pancreatic stent inserted across the minor papilla
Endoscopic sphincterotomy or papillotomy of the minor papilla is considered the first-line endoscopic therapy for patients with symptomatic PD (Fig. 15.5) [1, 7]. There are two main techniques: needle-knife papillotomy (NKP) and standard pull-type papillotomy (PTP) [1, 7]. The needle-knife technique is performed at the 10- to 12-o’clock position over a pancreatic stent. The stent provides a platform to avoid cutting too deeply and provides direction. It is also a safety valve by ensuring adequate drainage in case bleeding, or patient instability causes the procedure to be prematurely halted. The standard PTP is done using a minipapillotome or a standard pull-type sphincterotome using a freehand technique over a guidewire (Video 15.2). The landmarks for determining an adequate sphincterotomy are not standardized but the cut should be made to end flush with the duodenal wall. Sometimes, a minor papilla bulge is present, which makes it easier to determine where the minor papilla ends and the duodenal wall begins [36].
Fig. 15.5
a–d A series of images demonstrating minor papilla sphincterotomy using a tapered tip papillotome
The choice of the type of papillotomy is usually based upon personal preference of the endoscopist. Lehman et al. advocate the NKP because of its directional and depth control and safety, but recent studies have shown that both techniques are equally safe and effective [36–38]. Current evidence suggests that the two sphincterotomy techniques are comparable with respect to re-stenosis and re-intervention rates. The technique applied for sphincterotomy may depend on the ductal anatomy. For example, if a patient has a very tortuous duct requiring a short stent, it may be advantageous to perform sphincterotomy first with a papillotome followed by placement of a short stent. If the ductal anatomy can accommodate a longer, more stable stent, then stent placement followed by a NKP may be optimal.
Minor papilla sphincterotomy has a higher reported complication rate when compared with the major papilla including pancreatitis, bleeding, sepsis, and perforation [26, 37–39]. The fact that interventions on the minor papilla are much less common than those on the major papilla undoubtedly contributes to the higher complication rate. As expected in PD, cannulation of only the major papilla has a very low rate of pancreatitis (1.2 %) while dorsal duct cannulation confers higher risk of pancreatitis, and patients undergoing minor papilla sphincterotomy are at the highest risk for post-ERCP pancreatitis with rates reported between 8 and 11 % [38, 39].
Results of Endoscopic Therapy: Who Should Be Treated?
The studies evaluating the role of endoscopic therapy for PD are plagued with the limitations that are common to many studies on endotherapy and include small sample size, referral bias, patient heterogeneity, lack of control arm, short duration of follow-up, and the absence of blinding as well as randomization [7]. The majority of studies assessing the efficacy of endoscopic therapy have been retrospective case series with a small sample size and no control arm. There are no studies that prospectively compare endoscopic with surgical therapy. A recent pooled systematic analysis published in 2009 looked at the results of endoscopic therapy and reported that response rates were insignificantly lower than that of surgical therapy (69.4 % vs. 74.9 %, respectively; p = 0.106) [4]. The pooled analysis reconfirmed the findings of earlier studies that had shown the best results for endoscopic and surgical therapy in patients with PD and ARP. The following response rates were noted for endoscopic and surgical therapy respectively: 79.2 % and 83.2 % in patients presenting with ARP, 69.0 % and 66.7 % in patients with CP, and the lowest response rates of 54.4 % and 51.6 % in patients presenting with abdominal pain only. Thus, all these studies suggest that patients with PD and ARP are the best candidates for endoscopic or surgical therapy. Patients with CP and pancreatic-type pain have lower response rates, and therefore, careful case selection is critical to optimizing outcomes. One recent study has shown excellent response rates with aggressive endoscopic therapy even in patients with chronic pancreatitis although very few of these patients had pancreatic duct stones or strictures, supporting the importance of case selection [34]. While there is no consensus on the method of detecting restenosis after minor papillotomy, studies report restenosis rates of 10–20 % leading to reintervention in these patients. The restenosis rates are comparable between the two minor papillotomy techniques [38].
In the absence of a randomized study comparing endoscopic and surgical treatment for PD and the current evidence suggesting equal efficacy, endoscopic therapy being less invasive is currently considered the first-line treatment option for symptomatic patients with PD.
Case Continued
As the patient had PD diagnosed on both MRCP and EUS with multiple attacks of documented acute pancreatitis, ERCP with minor papillotomy was planned. Our policy is to offer pancreatic endotherapy to all patients with PD who have had a single attack of severe acute idiopathic pancreatitis or 2 attacks of mild acute pancreatitis . The pancreatogram obtained through the major papilla revealed a short ventral duct, and thereafter, the minor papilla was identified and cannulated with a metal tip catheter. The pancreatogram revealed a nondilated dorsal duct, and a NKP was performed over a 3 Fr pancreatic stent without inner flanges. Plain abdominal X-ray 3 weeks later documented stent dislodgement. The patient has been asymptomatic over 2 years of follow-up.
ERCP in Autoimmune Pancreatitis: Is There a Role?
Autoimmune pancreatitis (AIP) is an uncommon form of chronic pancreatitis that in some cases can be difficult to differentiate from pancreatic cancer [40]. There was initial enthusiasm that elevated serum IgG4 levels would be diagnostic for AIP, but this enthusiasm has tempered with the realization that it has poor diagnostic performance in certain populations such as Caucasian patients [40]. It is now known that there are two types of AIP (type 1 and type 2), and serum IgG4 levels are usually normal in type 2. Type 1 AIP, also known as lymphoplasmacytic sclerosing pancreatitis, is most common in Asian countries, has an older age of onset, and usually presents with painless obstructive jaundice. It is a multisystem disease and can affect the biliary tree, salivary glands, kidneys, and retroperitoneum [41]. Serum IgG4 levels are elevated, and many IgG4 + cells can be seen on immunohistochemistry in pathology specimens. Patients with type 1 AIP respond very well to steroids but recurrences are common. Type 2 AIP or idiopathic duct-centric chronic pancreatitis is more common in Europe and USA and occurs at an earlier age. Patients present either with obstructive jaundice or with acute pancreatitis , serum IgG4 levels are usually normal, and the disease is limited to the pancreas. Like type 1 AIP, these patients also respond well to steroids but recurrences are rare [41].
There is no single clinical, laboratory or imaging feature that is characteristic of AIP, and histology is the reference standard for the diagnosis. It is very difficult in routine clinical practice to obtain pancreatic tissue for histopathological analysis. Therefore, a number of groups have proposed criteria to establish the diagnosis [Japanese Pancreas Society (JPS), Korean criteria, Mayo Clinic HISORt criteria, and International Consensus Diagnostic Criteria and Algorithm (ICDC)] [40, 41]. These diagnostic criteria use a combination of imaging features, serology, evidence of other organ involvement, histology of the pancreas with immunohistochemistry, and dramatic response to steroids (Tables 15.1, 15.2, and 15.3).
Table 15.1
Japan Pancreas Society Diagnostic Criteria for AIP (Proposed in 2002 and revised in 2006)
Criteria | Diagnostic criteria |
|---|---|
Imaging | Diffuse narrowing of MPD with irregular wall (> 1/3 the length of entire pancreas) and diffuse enlargement of the pancreas |
In revised criteria, minimum extent (> 1/3 the length of the entire pancreas) of MPD removed | |
Diffuse narrowing of MPD and diffuse enlargement of pancreas changed to diffuse or segmental narrowing and diffuse or localized enlargement | |
Laboratory finding | Abnormally elevated levels of serum gamma globulin and/or IgG, or the presence of auto antibodies |
In the revised criteria, elevated serum IgG4 included | |
Histology | Marked lymphoplasmacytic infiltration and dense fibrosis |
In the revised criteria, changed to marked interlobular fibrosis with prominent infiltration of lymphocytes and plasma cells |
Table 15.2
Korean Diagnostic Criteria for AIP (Kim Criteria at Asan Medical Centre)
Criteria | Diagnostic criteria |
|---|---|
Imaging | (1) Diffuse enlargement (swelling) of pancreas |
(2) Diffuse or segmental irregular narrowing of the main pancreatic duct | |
Laboratory finding | (1) Elevated levels of IgG or IgG4 or |
(2) Detected autoantibodies | |
Histology | Fibrosis and lymphoplasmacytic infiltration |
Response to steroids | Present |
Table 15.3
HISORt Diagnostic Criteria for AIP (Mayo Clinic)
Criteria | Diagnostic criteria |
|---|---|
Histology | 1. Diagnostic (any one) |
a. Pancreatic histology showing lymphoplasmacytic sclerosing pancreatitis | |
b. Lymphoplasmacytic infiltrate with abundant (> 10 cells/HPF) IgG4-positive cells in pancreas | |
2. Supportive (any one) | |
a. Lymphoplasmacytic infiltrate with abundant (> 10 cells/HPF) IgG4-positive cells in extrapancreatic organ | |
b. Lymphoplasmacytic infiltrate with fibrosis in pancreas | |
Imaging | Typical imaging features |
1. CT/MRI: diffusely enlarged gland with delayed (rim) enhancement | |
2. ERCP: diffusely irregular and attenuated main pancreatic duct, atypical imaging features, pancreatitis, focal pancreatic mass, focal pancreatic duct stricture, pancreatic atrophy, pancreatic calcification
Related posts: Training in Endoscopic Retrograde Cholangiopancreatography
 Choledochal Cysts: Evaluation and Management
 Idiopathic Acute Pancreatitis and Sphincter of Oddi Dysfunction: Diagnostic and Therapeutic Role of ERCP and Sphincter of Oddi Manometry
 EUS in Pancreatic Tumors
 ERCP in Pregnancy
 Pediatric Endoscopic Retrograde Cholangiopancreatography
 Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
Get Clinical Tree app for offline access
|