Introduction

Surgery is sometimes thought of as the last resort in chronic pancreatitis management. But for patients with advanced disease, medical and endoscopic treatments are frequently associated with inadequate symptom relief. Up to 50% of all patients may undergo surgical intervention at some time during their disease [1, 2]. Operative approaches for management chronic pancreatitis have evolved—improved understanding of pathophysiology, advances in surgical technique, and better patient selection have all contributed toward the increased utility of surgery in the treatment of chronic pancreatitis.

Indications for Surgery

The primary therapy for chronic pancreatitis is simply symptom management. However, while symptoms of exocrine insufficiency (steatorrhea and malnutrition) and endocrine insufficiency (diabetes mellitus) in chronic pancreatitis can be treated medically, the recurrent and intractable episodes of abdominal pain represent the most common indication for endoscopic and surgical intervention [3]. The main mechanism of pain in chronic pancreatitis is thought to be a “parenchymal hypertension” due to pancreatic ductal obstruction. This understanding of the pathophysiology of the pain serves as the rationale for endotherapies to decompress the duct by papillotomy, stone removal, dilation of strictures, or stent placement [4, 5]. While these procedures often provide temporary relief of symptoms, several prospective randomized clinical trials have demonstrated that endoscopic therapies are inferior to surgery with respect to pain control and weight gain [6, 7].

A numbers of local complications of chronic pancreatitis may require surgery (Table 1). Though the exact mechanism is unclear, inflammation leads to the formation of pancreatic duct stones (pancreatolithiasis), which often obstruct the pancreatic duct and results in multiple areas of ductal stenosis and dilatation. Inflammatory masses at the head of the pancreas may cause jaundice from biliary ductal obstruction or duodenal stenosis leading to gastric outlet obstruction. It is widely appreciated that chronic pancreatitis patients have at least a threefold higher risk of cancer. Fine needle aspiration is more prone to false-negative results in this setting, so it can be difficult to “rule out” cancer in the patient with jaundice and an inflammatory mass in the head of the gland [8, 9]. Thus, subsets of patients with head-dominant chronic pancreatitis undergo resection because of the inability to exclude an underlying cancer. Other rare but severe local complications that may require surgery include vascular erosion and resultant gastrointestinal hemorrhage or a pseudoaneurysm that cannot be resolved via interventional radiology. Finally, sizeable symptomatic pancreatic pseudocysts that are not accessible to endoscopic drainage are an occasional indication for surgery.

Table 1 Indications for surgery in chronic pancreatitis
Full size table

Timing of Surgical Intervention

The goal of surgical treatment for chronic pancreatitis is to relieve symptoms and preserve as much of the pancreatic parenchyma as possible. However, supportive treatment of chronic pancreatitis in the early stage is important to correct exocrine and endocrine insufficiency, provide nutritional supplementation, and treat pain. When patients require additional therapy beyond these initial supportive measures, they are generally referred to gastroenterologists for endoscopic treatment. Typically, patients with proximal pancreatic duct stenosis without calcifications and without an inflammatory mass or patients with pancreatic pseudocysts may be candidates for endoscopic therapy first and evaluation of surgery only if endoscopic treatment fails. If endoscopic attempts have been unsuccessful in the management of symptoms and ductal obstruction after one year or if local complications described above develop, then surgery should be considered before the appearance of nutritional or metabolic disorders. The loss of pancreatic parenchyma from ductal obstruction is progressive and irreversible. In addition, persistent pancreatic pain may become pancreas-independent over time and lead to narcotic dependence. Therefore, it is desirable that a pancreatic surgeon be involved in the assessment of the patient reasonably early in their course of disease. The success of surgery for chronic pancreatitis relies heavily on identifying patients at the appropriate time in the course of their disease.

Surgical Spectrum

The principal surgical strategies available for chronic pancreatitis are resection procedures, drainage procedures, and a mix of the two (Table 2). One end of the spectrum involves classic pancreatic head resection procedures such as the Whipple pancreaticoduodenectomy (PD) or its pylorus-preserving variant (ppPD). These “oncologic” procedures treat the inflammatory pseudotumor of the pancreatic head that is often the dominant morphologic abnormality in chronic pancreatitis. This inflammatory mass may be accompanied by biliary or duodenal obstruction and is thought by some to be the “pacemaker,” propagating pain development in the disease [10]. Simple drainage procedures represent the other end of the spectrum, a less invasive strategy aimed at decompressing a dilated pancreatic duct. The commonly known “Puestow” procedure, a longitudinal pancreaticojejunostomy (LPJ), has been the standard drainage procedure for treatment of chronic pancreatitis for more than 50 years [11]. In the middle of the spectrum lie duodenal-preserving pancreatic head resections (DPPHRs). These combine a partial resection to address inflammatory disease in the proximal pancreas (without interruption of the gastroduodenal and biliary outflow as in PD) and drainage of the distal pancreas [12, 13]. Therefore, whether the pain is from perineural inflammation in the proximal pancreas or ductal hypertension in the distal pancreatic duct, the DPPHR procedures aim to address both potential causes. For local complications limited to the tail of the pancreas, distal pancreatectomy is indicated in a small portion of patients, but there is a significant risk of symptomatic recurrence if untreated disease is left behind in the head of the pancreas. Broader anatomical resections such as subtotal and total pancreatectomy (TP) have been generally abandoned for the treatment of chronic pancreatitis due to the high morbidity associated with resultant exocrine and endocrine insufficiency except in centers that perform islet cell autotransplantation (IAT). Over the last decade, there has been increasing interest in total pancreatectomy with islet cell autotransplantation (TP-IAT) as a surgical option for patients with recurrent chronic pancreatitis and minor morphologic change or patients considered at high-risk for malignancy as in hereditary pancreatitis [14]. Although TP potentially eliminates the source of inflammation, pain, and cancer risk by complete removal of the offending tissue, the resultant brittle diabetes can be very difficult to manage. Hence, with technical improvements in the isolation and preparation of islet cells after pancreatectomy, the addition of IAT has emerged as a viable strategy to preserve beta cell function and lessen risk of diabetes when combined with TP. Select centers performing these procedures report insulin independence in up to 24–40% of patients [15].

Table 2 Surgical spectrum for the treatment of chronic pancreatitis
Full size table

Resection Procedures

The classic Whipple PD and ppPD procedures were initially conceived for malignancies in the pancreatic head. The operations are identical except that classic PD includes resection of the distal stomach while the ppPD was subsequently introduced to preserve the antrum for the presumed physiologic benefits of retaining the pylorus. The reconstruction involves three anastomoses to restore gastrointestinal continuity and the flow of pancreaticobiliary secretions. The pancreatic anastomosis is the crux of the operation, and a pancreaticojejunostomy leak can be a major cause of morbidity and prolonged hospital stay. Although the consistency of the gland in chronic pancreatitis is quite firm (associated with a decreased leak rate), the main duct can be 2–3 mm in a gland with diffuse sclerosis which makes the anastomosis technically challenging. In high-volume centers, with increasing safety and reduction of the mortality rate of the operation to less than 5%, these procedures have also been used for patients with chronic pancreatitis [16, 17]. Long-term follow-up after PD has offered an improvement in the quality of life and lasting pain relief in greater than 75–85% of patients [18, 19]. The disadvantages of PD in these cases are the sacrifice of surrounding non-diseased organs with loss of natural bowel continuity; morbidity associated with pancreatic anastomotic leak, delayed gastric emptying and reduction in pancreatic endocrine and exocrine function can approach 20–40% [20].

In TP-IAT, a pylorus-preserving total pancreatectomy is performed (usually with splenectomy) and the pancreas is immediately placed in a cold sterile preservation solution and transported on ice to an islet cell isolation laboratory. While processing is performed using an enzymatic digestion process, the remaining two anastomoses are completed to restore gastrointestinal and biliary continuity. Islet cell infusion is performed through the splenic vein stump into the portal venous system [15]. Although PD and TP provide the benefit of a complete resection in some patients with unrecognized pancreatic adenocarcinoma, for most patients with chronic pancreatitis, resection of the duodenum is arguably not strictly necessary.

Drainage Procedures

Treatment of chronic pancreatitis with surgical drainage procedures may be carried out in selective patients with main duct dilatation and without a proximal inflammatory mass. The rationale behind duct drainage alone is the presumption that pain is generated from elevated pressures in the obstructed duct. Stasis of pancreatic fluid can also result in the formation of calculi, often found in the dilated duct during surgery [21]. The most common procedure to achieve decompression of the pancreatic duct for a stenosis to the left of the gastroduodenal artery is the modified Puestow LPJ drainage procedure [22]. The dilated main pancreatic duct is identified by intraoperative ultrasound and confirmed by needle aspiration of the pancreatic fluid. The pancreatic duct is often described as having a “chain-of-lakes” appearance (Fig. 1) with progressive segmental stenosis that develops along the longitudinal axis [23]. Therefore, a longitudinal (rather than end) opening of the pancreatic duct is created and extended to an area without stenosis. A Roux-en-Y limb of small bowel is created and opened longitudinally to match the length of the opened pancreatic duct, and a single layer anastomosis is performed as shown in Fig. 2B [22, 24]. In some cases of combined biliary and pancreatic stricture, the pancreaticojejunostomy can be combined with a hepaticojejunostomy.

Fig. 1
figure 1

Anatomical changes in chronic pancreatitis. The pancreatic duct is often described as having a “chain-of-lakes” appearance with areas of dilation and segmental stenosis. There is resultant atrophy of the pancreas and stasis of pancreatic fluid can lead to the development of intraductal calculi

Full size image
Fig. 2
figure 2

Modified Puestow pancreaticojejunostomy. A longitudinal incision is made along the length of the dilated pancreatic duct as in 2A. A roux limb of jejunum is then brought up to create a side-to-side anastomosis for the longitudinal pancreaticojejunostomy as in 2B and allow for decompression of the pancreatic duct

Full size image

This is a safe and efficient technique with very low operative morbidity and mortality. Pancreatic leak rate is low (<5%) in selected patients with a fibrotic gland. As there is no loss of functional pancreatic tissue, exocrine and endocrine function is also well preserved in this procedure. The primary success rate of the Puestow has been good with over 80–85% of patients achieving pain relief in the short term [22]. There have been no prospective randomized trials to compare the Puestow LPJ versus pancreatic resection procedures for chronic pancreatitis, but the recurrence of pain in those patients who have undergone LPJ is often attributed to persistent or recurrent disease in the head of the pancreas [25]; therefore, the strategy of drainage alone may not have been the ideal surgical solution for those patients in the first place. In patients with ductal dilation alone (>7 mm), the Puestow procedure is simple, avoids extra dissection around the pancreatic head, and is effective in long-term improvement of pain symptoms in the absence of proximal inflammatory disease.

The Frey procedure (named for Charles Frey and described by Frey and Smith) is arguably a variant of the Puestow LPJ [25]. The authors describe their procedure as LR-LPJ (local resection-longitudinal pancreaticojejunostomy) and claim its advantage in patients with less severe inflammation in the head combined with an obstructed pancreatic duct. Although often classified as a “resection” procedure by the authors (to distinguish the Frey from the Puestow), the mechanism of the Frey procedure is essentially drainage carried out proximally for further decompression of the duct. The local resection does not serve so much to excise the pancreatic head inflammation but is essentially a limited “coring out” of the pancreas parenchyma to facilitate better exposure of the proximal duct for LPJ and prevent the natural tendency of the dense pancreatic head tissue to want to close the longitudinal incision made in this part of the duct. A similar single anastomosis using a Roux limb to drain both the proximal and distal duct is performed as in the Puestow (Fig. 3).

Fig. 3
figure 3

Intraoperative photograph of the Frey procedure. The surgical instrument is placed in the distal pancreatic duct. The head of the pancreas has been cored out (blue suture) to allow for continued exposure of the proximal pancreatic duct. A Roux limb of small intestine is brought up to the longitudinal incision made in the pancreatic duct, and the posterior row of the pancreaticojejunostomy has been completed

Full size image

In patients with evidence of ductal disease (dilatation and stones) and a pseudocyst, addressing the ductal hypertension alone will also effectively treat the pseudocyst. Nealon and Walser elegantly demonstrated this in a series of 103 patients that underwent LPJ with or without cystojejunostomy [26]. They demonstrated nearly identical resolution of the pseudocyst and pain relief, with no recurrences in those that had LPJ alone.

Duodenum-Preserving Pancreatic Head Resections

Duodenum-preserving pancreatic head resections (DPPHRs) were introduced as middle ground to address chronic pancreatitis with an inflammatory mass in the head of the gland but with the intent to preserve anatomy of the foregut. The first duodenum-preserving resection of the pancreatic head was introduced by Hans Beger in 1980 consisting of a subtotal resection of the pancreatic head with transection of the neck of the pancreas at the level overlying the portal and superior mesenteric vein [12]. The pancreas is then reconstructed by two anastomoses: an end-to-end or end-to-side Roux-en-Y pancreaticojejunostomy to the distal pancreas and an end-to-side pancreaticojejunostomy to the remnant of pancreatic tissue on the inner aspect of the duodenum. The main complications of this procedure include the risk of ischemia of the duodenum due to inadequate perfusion of the posterior branch of the gastroduodenal artery and risk of pancreatic leak from either of the two anastomoses. Several European studies have shown the Beger procedure to be a safe procedure with minimum alteration of pancreatic insufficiency and very effective long-term pain relief of 80–85% maintained after a median follow-up of 5 years [27].

The Berne modification of the DPPHR (described by Markus Büchler) represents a simplification of the Beger procedure by eliminating the division of the pancreas anterior to the level of the portal mesenteric vein confluence [13]. Creation of the retropancreatic tunnel in preparation for pancreatic resection can often be tedious in the setting of inflammation and portal hypertension. The resection of the pancreatic head is performed with identical extent compared to the Beger procedure, but reconstruction is completed with one single anastomosis between a jejunal Roux limb and the pancreatic resection edge (short-range pancreaticojejunostomy). By eliminating the need for a second anastomosis, the Berne modification has been shown to have shorter operating times and shorter hospital stay [28]. Notably, neither the Beger nor the Berne has been adopted outside of very limited European centers.

Comparison of PD and DPPHR

There has been considerable interest in comparing DPPHR procedures to PD and to each other (Beger versus Berne). The Beger and Frey procedures also have been compared in several randomized controlled trials (RCTs), because the Frey procedure is sometimes categorized with DPPHR because it includes a “local resection” of the pancreatic head. The most cited initial RCTs comparing PD and DPPHR are two studies from 1995, comparing Beger versus PD [29] and another comparing Beger versus ppPD [30]. Both claimed benefits of the DPPHR in producing better pain relief and better weight gain postoperatively. However, these convincing early advantages were not maintained over more recent studies comparing PD and DPPHR with long-term follow-up demonstrating equal pain management, quality of life, and endocrine and exocrine function [31, 32]. In the RCT comparing the Beger and Berne techniques, the study demonstrated equal outcomes in pain control, quality of life, and metabolic parameters between the two procedures. As mentioned above, the only differences were that the Berne procedure could be performed significantly faster (by 46 min) and led to a shorter hospital stay (11 versus 15 days) [28]. Similarly, the Beger and Frey RCTs, despite being two different surgical strategies, also demonstrate comparable outcomes over the long term [33–35]. The most recently published protocol for the ChroPac Trial pertains to the first large multicenter RCT to compare DPPHR and PD with the primary outcome represented by quality of life at 24 months after surgery [36]. The trial finished enrollment and randomization of 250 patients by 2013, and their last follow-up was at the end of 2015 with no results published to date. It is hoped that the trial will provide a better idea of the short-term outcomes and morbidity rates associated with these procedures.

Conclusion

When pain relief and an improvement in quality of life cannot be achieved by medical therapy alone, current evidence suggests that surgery is the treatment of choice and can provide lasting long-term improvement. Endoscopic therapy may be effective for internal drainage of pancreatic pseudocysts and for selective cases of proximal ductal stenoses in patients without significant calcifications. Initial endoscopic strategies may be warranted but often require frequent reinterventions, and if not sufficiently effective, early surgical referral is recommended.

For decades, PD served as the standard operation for chronic pancreatitis and still is used in some cases where a malignancy cannot be ruled out in the setting of pancreatitis. The patient with pancreatic, biliary, and duodenal stenosis is also often managed with PD when feasible, although sometimes the inflammatory mass in the head of the pancreas and surrounding changes can make for a treacherous dissection. The DPPHR procedures evolved as an alternative and have proven effective in limited studies, but have not been adopted broadly.

The Puestow lateral pancreaticojejunostomy, particularly as modified by Frey, continues to serve a valuable role in patients with chronic pancreatitis associated with ductal dilatation. There is relief of ductal hypertension, and the long-term outcome is usually successful for the patient’s quality of life. Irrespective of the surgical strategy, when carried out at experienced centers, these techniques are safe and effective, all with comparable results in patients with chronic pancreatitis. The decision of which procedure to employ should take into consideration the anatomical changes of the pancreas and the experience/comfort of the surgeon to provide the best operative outcome and long-term improvement for the patient.

Key Messages

  • Advances in surgical technique and better patient selection have led to increased utility of surgery in the treatment of chronic pancreatitis.

  • Regardless of technique, if performed in a high-volume center on well-selected patients, all procedures approach good pain control in approximately 85% of patients over the long term.

  • The DPPHR techniques were introduced as a new alternative to surgical resection of the head of the pancreas but never achieved global status.

  • The Puestow LPJ is probably the most common procedure used internationally for relieving pain in the setting of a dilated duct given its simplicity and low morbidity.

  • The Puestow LPJ can also be combined with a hepaticojejunostomy or a gastrojejunostomy in the case of biliary or duodenal stenosis to avoid difficult dissection around the pancreatic head in the case of severe inflammation.