Joshua A. Waters
C. Max Schmidt
Presentation
A 57-year-old, previously healthy woman is referred to the surgical clinic with a newly diagnosed 2-cm cystic lesion within the body/tail of the pancreas. The cyst lesion was identified incidentally on a computed tomography (CT) scan performed during a workup for hematuria. The patient is without any significant gastrointestinal complaint, has no history of pancreatitis, and denies weight loss or diarrhea. Her physical examination is unremarkable. The patient is physically active and works full-time as an elementary school teacher.
Differential Diagnosis
The differential diagnosis of pancreatic cystic lesions can be subdivided into inflammatory and neoplastic. Pancreatic pseudocysts are inflammatory and generally arise as sequelae of acute pancreatitis. The majority of incidentally discovered asymptomatic pancreatic cystic lesions fall into the neoplastic category. This group can then be further divided into cystic lesions with or without malignant potential. Serous cystic neoplasms (SCNs), lymphoepithelial pancreatic cysts, and simple pancreatic cysts are regarded as benign entities, whereas intraductal papillary mucinous neoplasms (IPMNs), mucinous cystic neoplasms (MCNs), solid pseudopapillary neoplasms, and cystic neuroendocrine tumors (pNETs) represent cystic neoplasms with varying degrees of malignant potential. The most common neoplastic cystic lesions encountered in practice will be IPMN, MCN, and SCN.
Workup
Although a pancreatic cystic lesion may be detected as an “incidental finding” as in this case, a complete history and physical examination will often reveal subtle but characteristic symptoms and signs. In taking a history, patient demographics, specifically gender and age, should be considered in the diagnosis and treatment of pancreatic cystic lesions. Symptoms and signs caused by a pancreatic cystic lesion are important to characterize because they may increase the likelihood of the lesion being malignant. Alternative causes of symptoms/signs must be ruled out since these should not impact this risk assessment. Symptoms/signs are typically abdominal/back pain, pancreatitis, steatorrhea, unintentional weight loss, and new-onset diabetes. These findings may be due to mass-effect, inflammation, ductal obstruction, or pancreatic (exocrine or endocrine) failure. Persistent or recurrent upper abdominal pain with/without radiation to the back may represent pain of pancreaticobiliary origin. In IPMN, this is potentially caused by intermittent pancreatic ductal obstruction with mucus that is most common when the main pancreatic duct is involved. Such pain may also occur in MCN and SCN possibly due to the effects of inflammation or local pressure on the pancreatic parenchyma and ductal system. Sequelae of pancreatic exocrine failure may occur, characterized by steatorrhea and unintentional weight loss that may result in malnutrition. In protracted untreated cases, the cachexia that results may be significant enough to be confused with cancer cachexia. Exocrine failure occurs most commonly in patients with main duct IPMNs but may occur in side-branch IPMNs, MCNs, or SCNs. Often, exocrine failure presents with more subtle findings such as bloating or foul-smelling flatulence. These symptoms will resolve with administration of oral pancreatic enzyme supplementation. Pancreatic endocrine failure manifested by glucose intolerance or diabetes may indeed occur and should be suspected particularly if there is no other obvious explanation (e.g., high body mass index or positive family history of diabetes).
Finally, a social and family history is important to delineate further risks of pancreatic cancer (e.g., tobacco, obesity, family) to arrive at a cumulative risk assessment for the patient. Some of pancreatic cystic lesions can be seen a part of familial syndromes (e.g., SCN and von Hippel-Lindau syndrome; IPMN and familial pancreatic cancer).
Physical exam findings aside from abdominal tenderness are uncommon. A palpable abdominal mass, enlarged supraclavicular lymph nodes, cachexia, or jaundice may occur, but are uncommon except in cases of advanced pancreatic malignancy.
The primary diagnostic modalities for identification and characterization of pancreatic cysts are cross-sectional imaging and endoscopy. Thin-slice helical CT with dualphase intravenous contrast administration or magnetic resonance cholangiopancreatography (MRCP) with gadolinium and secretin injection are both useful in the characterization of pancreatic cystic lesions. MRCP is more sensitive at detecting small lesions in part because it produces a detailed pancreatic ductogram (Figure 1). The presence of ductal connectivity suggests an IPMN. Multifocality in the absence of a history of pancreatitis such as in this case is virtually diagnostic of IPMN. The presence of a central stellate scar is classic for microcystic SCN but is unreliable for macrocystic serous lesions. Cystic dilation of the main pancreatic duct is typical of main duct–involved IPMN but may also be present in other cystic lesions (MCN, pNET) and is associated with high rates of malignant transformation. Intracystic mural nodules are characteristic of IPMNs and MCNs and also serve as predictors of malignancy. Mucus or dependent debris may resemble mural nodularity in static imaging, so suspected mural nodules on CT or MRI–MRCP should be confirmed by a dynamic test such as endoscopic ultrasound (EUS).
FIGURE 1 • This MRCP ductogram demonstrates multifocal side-branch IPMN with the largest lesion situated within the body/tail junction of the pancreas. Additional small dilated branch ducts may be seen within the pancreatic head and throughout.
EUS is more invasive than cross-sectional imaging, but allows for dynamic visualization. If there are contraindications to MRI–MRCP, EUS should be strongly considered. In addition, EUS allows fine needle aspiration (FNA) of cyst fluid, intracystic mural nodules, and adjacent masses. Endoscopic retrograde cholangiopancreatography (ERCP) is uncommonly used due to its higher risk profile, but ERCP in combination with pancreatic ductoscopy with directed biopsies may be indicated in patients with main pancreatic duct dilation to establish a diagnosis of neoplasia versus pancreatitis and to assist in operative planning in patients with diffusely dilated main pancreatic duct and suspected main duct–involved IPMNs.
Cyst fluid cytopathology, second only to surgical pathology, remains the “gold standard” and most accurate and specific predictor of malignancy. Unfortunately, it is not as sensitive, and this is most often due to low cellularity. Cyst fluid may also be assessed for biochemical (carcinoembryonic antigen [CEA], amylase) and molecular (DNA) analyses. Pancreatic cyst fluid CEA >192 ng/mL predicts a mucinous lesion (MCN or IPMN) but has no value in determination of malignant character. Elevated pancreatic cyst fluid amylase suggests ductal connectivity (IPMN or pseudocyst), but in ductal obstruction (e.g., mucin plug), it may be consistent with serum levels even in the presence of ductal connectivity. Molecular analyses may detect KRAS mutations that predict a mucinous lesion (compliments CEA). Quantity of DNA and number of DNA mutations may predict malignant potential.
Finally, serum studies may be useful in the workup of the incidental pancreatic cyst. Serial elevations of serum hemoglobin A1C (or fasting glucose) or cancer antigen 19-9 may predict malignant progression. Alkaline phosphatase elevation, particularly in pancreatic head cysts, may indicate early biliary obstruction, which correlates with malignancy in pancreatic head cystic lesions. Finally, elevated pancreatic inflammation markers, serum amylase and lipase, even in the absence of clinical pancreatitis, may indicate a pancreatic cyst with greater malignant potential.
Presentation Continued
Upon further clinical interview and exam, this patient recalls that she has experienced intermittent but prolonged epigastric pain radiating to her back over the past several months that she has attributed to dyspepsia. An MRI–MRCP demonstrates a 2-cm multiloculated cyst located in the body/tail junction of the pancreas with a 5-mm mural nodule, a ductal connection, and a main pancreatic duct that is 4 mm in close proximity to the cyst. Also identified on the MRCP is a 5-mm cyst within the head of the pancreas. EUS–FNA was performed on the 2-cm pancreatic body/tail cyst with the biopsy needle directed at the 5-mm mural nodule (Figure 2). Cyst fluid analysis reveals a CEA level of 429 ng/mL and an amylase of 2,000 IU. Cyst fluid cytopathology is negative for high-grade atypia but demonstrates mucinous epithelium and abundant extracellular mucin. DNA analysis of the pancreatic cyst fluid reveals high DNA quantity and a high clonality KRAS mutation, the combination reported as a pancreatic cyst with aggressive behavior. Serum amylase and lipase on random check are elevated.
FIGURE 2 • EUS demonstrating a discreet unilocular cyst within the pancreatic gland with a mural nodule.
Diagnosis and Treatment
In a patient with side-branch IPMN without attributable symptoms, concerning radiographic features, or suspicious cytopathologic findings, surveillance in lieu of surgery is an acceptable approach. Surveillance may be less acceptable, however, in young patients due to the anticipated length of surveillance. This is due both to the cumulative risk of malignant transformation over time as well as cost to the individual and society of prolonged surveillance.
In this patient, based upon imaging features (main pancreatic duct connection and cyst multifocality), pancreatic cyst fluid markers (CEA > 192 ng/mL and cyst fluid amylase > serum amylase), and cytopathology (mucinous epithelium), the diagnosis of the tail of pancreas cystic lesion is likely to be a side-branch IPMN. Mucin contamination may occur due to the transgastric route utilized for the EUS–FNA leading to the possibility of false positive results, but there are enough data in this case for the diagnosis to stand without a definitive cytopathologic diagnosis.
Upon arriving at IPMN as the most likely diagnosis for this pancreatic cyst, the remainder of the management of this lesion hinges on the formation of an oncologic risk assessment. IPMNs represent a wide spectrum of potential to progress to invasive carcinoma. Consequently, accurate assessment of this risk allows for optimal prevention of pancreatic cancer development. Various predictors of this potential have been derived from numerous retrospective clinical studies focusing on clinical factors associated with the presence of, or eventual progression to, carcinoma within these cysts. The most basic distinction when developing an oncologic risk assessment relies on the presence or absence of main pancreatic duct involvement. Main duct–involved IPMNs have been reliably associated with elevated rates of invasive cancer, up to 50% to 60%, whereas side-branch IPMNs represent a risk of harboring invasive cancer of 10% to 20%. An additional radiographic/endoscopic finding that has been associated with an increased rate of invasive cancer is the presence of a mural nodule within the cystic lesion. The presence of a mural nodule within a branch-type IPMN has been validated as an independent predictor of invasive carcinoma. Finally, concerning cyst cytology (i.e., high-grade atypia) is useful as a predictor only when positive, as it is not a sensitive test. The surgical treatment of IPMN involves segmental resection of the affected pancreatic segment. This is often complicated by the fact that IPMN presents as multifocal lesions affecting more than one region of the pancreatic gland (as demonstrated in this case by the synchronous 5-mm cyst within the head of the pancreas). Additionally, the substantial postoperative morbidity of pancreatic resection further highlights the importance of oncologic risk stratification and preoperative characterization of IPMN.
In the described patient scenario, the presence of symptoms, possible early main pancreatic duct involvement (4-mm MPD diameter), and the presence of a mural nodule all elevate the risk of malignancy. The combination of these findings in the setting of a relatively young and fit patient would typically not be acceptable from an oncologic risk standpoint and would warrant consideration for resection of the pancreatic segment containing the most concerning lesion—in this case the 2-cm tail cyst containing a mural nodule.
Surgical Approach
In this case, the risk of malignancy significantly outweighs the risk of surgical resection. Thus, the patient is offered segmental resection targeted to the tail of the pancreas. The additional cyst in the head of the pancreas is almost certainly a side-branch IPMN. Despite this and the patient’s young age, total pancreatectomy is not recommended in this setting. The pancreatic head cyst has no concerning radiographic, cytopathologic, or molecular features. Thus, we would recommend surveillance to avoid the inevitable complications of pancreatic endocrine and exocrine failure associated with total pancreatectomy.
A minimally invasive distal pancreatectomy is one approach particularly well suited for cystic lesions of the distal pancreas. Although randomized clinical trials are not available to compare outcomes of minimally invasive to open distal pancreatectomy, case control studies suggest minimally invasive distal pancreatectomy is associated with decreased blood loss, decreased hospital length of stay, and a quicker recovery compared to open distal pancreatectomy. Preoperatively, the patient may be prepared with splenic vaccinations in case the spleen is sacrificed. The spleen may be salvaged in the setting of premalignant lesions of the distal pancreas, while it should be taken en bloc with the pancreatic resection in the setting of cancer. Spleen preservation may take one of two forms, either vessel-preserving or vessel-ligating (Warshaw) techniques.
The operation is commenced by performing a vertical incision above (laparoscopic) or below (robotic) the umbilicus. The laparoscopic video camera is introduced into the peritoneum using an open, needle localization or trocar visualization technique. Resection for IPMN, as in operations for known cancer, requires a thorough laparoscopic exploration to rule out metastatic disease to liver and visceral surfaces. Four additional 10 to 12-mm trocars are placed under direct laparoscopic vision, two in the left and two in the right hemiabdomen. The lesser sac is entered with a laparoscopic energy device through the greater gastrocolic omentum preserving the gastroepiploic vasculature. The posterior wall of the stomach is then retracted cephalad to expose the dorsal aspect of the pancreas in the retroperitoneum. Often a segmental portion of the splenic artery is visualized as it courses anteriorly from the celiac trunk and makes a tortuous course along the superior border of the pancreas. If these landmarks are not readily appreciated (e.g., fatty pancreas/peripancreatic retroperitoneum), laparoscopic ultrasound may help with localization. A plane is established along the inferior border of the pancreas using a laparoscopic energy device. Care must be taken to identify and preserve the splenic vein as the plane inferior to the pancreas is developed. The splenic vein is dissected free from the overlying pancreas using a laparoscopic energy device to seal and transect splenic vein branches to the pancreas. A vessel loop is placed around the splenic vein to retract the vein posteriorly to facilitate dissection. With the splenic vein retracted posteriorly, the proximal splenic artery underneath the neck of the pancreas is often visualized where it branches from the celiac trunk. A vessel loop may likewise be placed around the splenic artery to facilitate its dissection from the pancreas. Laparoscopic energy devices are less reliable on arterial branches, so the surgeon should have a low threshold for suture ligation.
Spleen preservation is preferred except in cases where there is a high index of suspicion preoperatively or intraoperatively for invasive cancer. In addition to spleen preservation, splenic artery and vein preservation are also desirable as case control studies suggest superior outcomes compared to splenic vessel sacrifice. Splenic vessel preservation when unsuccessful fails most often as dissection of the splenic vessels approaches the splenic hilum. Deliberate and meticulous dissection in this area is important, as upstream pancreatic tail anatomy can be quite variable particularly if the cystic lesion is located in this area. Foreshortening and disfigurement of the pancreas from inflammation may also complicate this dissection. Optimal preoperative imaging to understand the relationship between the upstream pancreas and the splenic vasculature is critical. These images should be consulted liberally intraoperatively. Laparoscopic ultrasound may also be used but is likely to be less helpful in the splenic hilum. Once the pancreatic tail is separated from the splenic vasculature, the pancreatic neck is transected. In thin pancreata, a stapling device, with or without buttress material, may be employed successfully. In thick or firm pancreata, transaction should be performed with laparoscopic energy device followed by oversewing (absorbable suture) of the pancreatic neck and pancreatic duct (if visualized). A closed suction drain although often used has no proven benefit. If a closed suction drain is placed, optimally the drain is removed early postoperatively, but if the output is >50 mL/d, drain amylase should be tested to guide timing of removal.
Special Intraoperative Considerations
One of the most challenging aspects in the operative management of IPMN comes as a result of its commonly multifocal and multicentric character. Even when prepared with the most robust preoperative imaging, the surgeon is often greeted with difficult intraoperative decisions. During resection for suspected IPMN, intraoperative frozen section of the proximal margin should be obtained routinely. Even in patients with a radiographically normal-appearing pancreatic ductal system at the site of transection, involvement with mucinous papillary epithelium may be noted. When the intraoperative margin is negative, no further action is necessary. When the main duct is involved with IPMN at the margin, reresection should be undertaken. The management of low-grade main duct IPMN at the resection margin when further resection would require total pancreatectomy (particularly in cases of normal sized main pancreatic duct) is more controversial and should be considered on a case-by-case basis, weighing such things as the fitness and projected life expectancy of the patient. In side-branch IPMN, a commonly multifocal condition, it is important to remove the most threatening lesion(s), but try to preserve the remainder of the gland. Low-grade side-branch IPMN at the margin need not be re-resected unless this involves the lesion targeted preoperatively for removal. High-grade or invasive IPMN at the margin regardless of duct localization (branch vs. main) mandates re-resection to negative margins in fit candidates, even to the extent of total pancreatectomy. In the case presented, the operative goal is targeted resection of the highest-risk side-branch lesion. Due to the multifocal nature of this IPMN, we anticipate side-branch IPMN will be left in the remnant pancreatic head. As always, the risks of additional resection should be weighed against the resulting morbidity including pancreatic exocrine/endocrine failure (Table 1).
TABLE 1. Key Technical Steps to Laparoscopic Distal Pancreatectomy
Postoperative Management
The most common serious morbidity associated with resection of the distal pancreas is related to leak of exocrine pancreatic secretions from the cut end of the pancreatic gland. Although the clinical impact of pancreatic fistulae (PF) varies, the overall rate of PF in distal pancreatectomy is between 15% and 25%. The subset of patients undergoing distal pancreatectomy in the setting of pancreatic cyst is at particularly high risk for fistula as they often have a soft pancreatic gland texture and a small pancreatic duct. Traditionally, intra-operatively placed closed suction drains are left at the cut edge of pancreas. Contemporary practice suggests that this may offer limited benefit or even detriment. PF may present late, in some cases after removal of operatively placed drains, or even after discharge, consequently the patient and clinician should be vigilant for clinical signs including fever, tachycardia, delayed return of bowel function, or worsening abdominal pain.
The long-term surveillance for IPMN varies based on the status of the remnant gland and the pathology of the primary lesion. Patients with unifocal low-grade IPMNs with a radiographically negative or positive remnant should undergo annual surveillance with history/physical, cross-sectional imaging (MRI–MRCP, CT) and serum studies. Patients with a positive low-grade main duct margin, patients with high-grade dysplasia on pathology, and patients who develop a new lesion in a previously negative remnant are at higher risk for malignancy and should undergo at least semiannual surveillance with the studies listed above. Surveillance in these high-risk groups should also include EUS–FNA with cytopathology and molecular analyses on an annual or biannual basis to monitor the remnant gland for recurrence or progression of existing unresected IPMN.
Case Conclusion
The patient ultimately elects to undergo robotic distal pancreatectomy given the relative risk of malignancy. The operation is completed robotically with splenic vessel and spleen preservation, and the patient is discharged from the hospital on postoperative day 3. Permanent pathology reveals an isolated branch-type IPMN with high-grade dysplasia (Figure 3). In light of the high-grade dysplasia, this patient is placed in a high-risk surveillance protocol to monitor the remnant pancreas, which includes a small unresected cystic lesion in the pancreatic head.
FIGURE 3 • Pathologic cross section of a distal pancreatectomy specimen, which demonstrates a cystic lesion lined with papillary epithelium and an associated mural nodule projecting into the cyst.
TAKE HOME POINTS
· Pancreatic cystic lesions are commonly identified incidentally on cross-sectional imaging.
· Differentiation of pancreatic cystic lesions is important to determine malignant potential.
· Signs/symptoms are often subtle, but their presence elevates malignant potential.
· Main pancreatic duct dilation and intracystic mural nodules both elevate malignant potential.
· Cytopathology is highly specific for malignancy, but has low sensitivity.
· Frozen section pathology should be obtained to guide intraoperative decision making.
· Side-branch IPMN are commonly multifocal. Operative resection, if indicated, should focus on removal of high-risk lesion(s) with preservation and surveillance of the pancreatic remnant.
· In IPMN, a positive main duct margin, high-grade dysplasia on initial pathology, or a new IPMN in a previously negative remnant that occurs during surveillance following segmental pancreatectomy are all risk factors for subsequent malignancy.
· Minimally invasive spleen-preserving distal pancreatectomy is the operation of choice for cystic lesions of the pancreatic body and tail.
SUGGESTED READINGS
Kooby DA, Gillespie T, Hawkins WG, et al. Left-sided pancreatectomy: a multicenter comparison of laparoscopic and open approaches. Ann Surg. 2008;248:438–446.
Merchant NB, Parikh AA, Kooby DA. Should all distal pancreatectomies be performed laparoscopically? Adv Surg. 2009;43:283–300.
Schmidt CM, White PB, Waters JA, et al. Intraductal papillary mucinous neoplasms: predictors of malignant and invasive pathology. Ann Surg. 2007;246(4):644–654.
Sohn TA, Yeo CJ, Cameron JL, et al. Intraductal papillary mucinous neoplasms of the pancreas: an updated experience. Ann Surg. 2004;239:788–797.
Waters JA, Schmidt CM. Intraductal papillary mucinous neoplasm—when to resect? Adv Surg. 2008;42:87–108.