**3. Rationale and indications for surgical intervention**

Surgery is indicated in chronic pancreatitis for the relief of pain, to manage complications and to resect confirmed or suspected neoplastic disease (Bornman PC S Afr Med J 2010). Two theoretical principles underlie the rationale for surgery to alleviate pain in CP. The first utilises the ductal / parenchymal tissue hypertension and inflammatory-neural theories on the pathogenesis of pain in CP. It postulates that surgical decompression of the main pancreatic duct will alleviate interstitial hypertension thereby improving parenchymal perfusion and acidosis (Patel AG Gastroent 1995) with consequent reduction of inflammatory stimulation and influx of mediators into damaged nerves (Salim AS HPB Surg 1997). The second principle focuses on removal of pathologically inflamed parenchyma together with altered neural tissue in particular that within the head, which is considered the "pacemaker" of disease. Emphasis has also been placed on the importance of addressing diseased side ducts, thereby limiting the possibility of recurrence (Beger HG World J Surg 1990).

The objectives of surgery for pain in CP are effective and durable relief of symptoms while preserving endocrine and exocrine function, thereby restoring the patient's quality of life. The potential for morbidity and mortality as well as recurrence should be low. Based on these objectives and the principles outlined above, a number of procedures have been developed. Essentially, these procedures fall into a spectrum covering three broad categories. At one end of the spectrum are drainage procedures which focus on decompressing the main pancreatic duct by establishing a new pancreatic-enteric communication which bypasses any native obstruction to pancreatic outflow. At the other end of the spectrum are resectional procedures which aim to remove diseased ductal and neural tissue within chronically inflamed parenchyma. Over time, a number of modified and hybrid procedures have evolved which attempt to retain the advantages while limiting the disadvantages of both the former 2 categories.

Little data exists to guide decision making regarding the optimal timing of surgery to alleviate pain in CP. There are two schools of thought. The first suggests that conservative non-surgical management should be pursued for as long as possible in order to avoid morbidity and side effects that may be associated with surgical intervention, in particular pancreatic insufficiency. They argue that the long term outcome of surgery is no different from medical management, and that the "pancreatic burn-out syndrome" is likely responsible for pain relief observed after surgery (Amman RW gastroenterol 1984). In contrast to this, others have argued that pain relief is better when surgical drainage is carried out earlier rather than later (Nealon WH Ann Surg 1993). It also remains controversial whether surgery can delay the natural course of the disease in terms of deterioration in pancreatic function (Warshaw AL Gastroenterol 1980; Nealon WH Ann Surg 1993; Jalleh RP Ann Surg 1992). Both arguments appear to have merit. While it seems

uncommon (Bornman PC in Hepatobiliary and pancreatic surgery – a companion to

Enzyme rich fluid collections may erode into vascular structures resulting in false aneurysms with bleeding into the cyst and pancreatic duct, peritoneum or retroperitoneum.

Surgery is indicated in chronic pancreatitis for the relief of pain, to manage complications and to resect confirmed or suspected neoplastic disease (Bornman PC S Afr Med J 2010). Two theoretical principles underlie the rationale for surgery to alleviate pain in CP. The first utilises the ductal / parenchymal tissue hypertension and inflammatory-neural theories on the pathogenesis of pain in CP. It postulates that surgical decompression of the main pancreatic duct will alleviate interstitial hypertension thereby improving parenchymal perfusion and acidosis (Patel AG Gastroent 1995) with consequent reduction of inflammatory stimulation and influx of mediators into damaged nerves (Salim AS HPB Surg 1997). The second principle focuses on removal of pathologically inflamed parenchyma together with altered neural tissue in particular that within the head, which is considered the "pacemaker" of disease. Emphasis has also been placed on the importance of addressing diseased side ducts, thereby limiting the possibility of recurrence (Beger HG World J Surg

The objectives of surgery for pain in CP are effective and durable relief of symptoms while preserving endocrine and exocrine function, thereby restoring the patient's quality of life. The potential for morbidity and mortality as well as recurrence should be low. Based on these objectives and the principles outlined above, a number of procedures have been developed. Essentially, these procedures fall into a spectrum covering three broad categories. At one end of the spectrum are drainage procedures which focus on decompressing the main pancreatic duct by establishing a new pancreatic-enteric communication which bypasses any native obstruction to pancreatic outflow. At the other end of the spectrum are resectional procedures which aim to remove diseased ductal and neural tissue within chronically inflamed parenchyma. Over time, a number of modified and hybrid procedures have evolved which attempt to retain the advantages while limiting

Little data exists to guide decision making regarding the optimal timing of surgery to alleviate pain in CP. There are two schools of thought. The first suggests that conservative non-surgical management should be pursued for as long as possible in order to avoid morbidity and side effects that may be associated with surgical intervention, in particular pancreatic insufficiency. They argue that the long term outcome of surgery is no different from medical management, and that the "pancreatic burn-out syndrome" is likely responsible for pain relief observed after surgery (Amman RW gastroenterol 1984). In contrast to this, others have argued that pain relief is better when surgical drainage is carried out earlier rather than later (Nealon WH Ann Surg 1993). It also remains controversial whether surgery can delay the natural course of the disease in terms of deterioration in pancreatic function (Warshaw AL Gastroenterol 1980; Nealon WH Ann Surg 1993; Jalleh RP Ann Surg 1992). Both arguments appear to have merit. While it seems

**3. Rationale and indications for surgical intervention** 

the disadvantages of both the former 2 categories.

specialist surgical practice. 2009).

b. Pseudoaneurysms

1990).

foolhardy to offer surgical intervention with it's attached risk of morbidity and even mortality in patients whose symptoms might be controlled by medical means, it seems equally unreasonable to persist with a conservative approach in anticipation of pain relief, delaying surgery until narcotic addiction has developed and the outcomes from surgery may be worse (Warshaw AL gastroenterol 1984). In the absence of good evidence to guide decision making, it seems most appropriate that the decision regarding timing of surgery be individualized on a patient to patient basis. Surgical intervention should be performed only once an adequate trial of medical therapy has failed to control symptoms and the patient has been counseled regarding the risks and benefits of both modalities.

Patients referred for surgery for relief of intractable symptoms of CP should be evaluated by experienced clinicians working in a high volume, multi-disciplinary environment. All other treatment options should have been exhausted or considered not appropriate. Cross sectional imaging should be conducted to clearly delineate pancreatic morphology and detect local complications or features suggestive of neoplastic disease. The presence of portal hypertension, particularly as a result of portal or superior mesenteric vein thrombosis should be noted, as this may preclude surgical intervention (Bornman PC S Afr Med J 2010). With careful patient selection and modern surgical strategies, surgery may offer effective pain relief in over 90% of patients at 5 year follow up (Beger HG Ann Surg 1989).

In considering intervention for complications of CP, the clinical picture is paramount in decision making. Biliary obstruction may be asymptomatic, detected only biochemically or during imaging for other indications. In addition, there may be transient jaundice as a result of oedema during acute flares of the disease. The above are not indications for intervention. It must be remembered that once the biliary system has been entered, either percutaneously, endoscopically or surgically, this once sterile system should be considered contaminated with the risk of sepsis developing should obstruction recur in the future. On the other hand, persistent biliary obstruction of sufficient duration may result in secondary biliary cirrhosis, atrophy and deterioration in hepatic function. (Abdallah A HPB 2007). Obstruction longer than 4 weeks should arouse concern and warrants intervention. Decompression by means of endoscopic stenting should only be considered as a temporary bridge to surgery, in acute cholangitis or where patient factors preclude surgery (Bornman PC S Afr Med J 2010). Duodenal obstruction on the other hand typically represents either advanced fibrosis or a clinically significant pseudocyst, neither of which are likely to resolve before progression or further complications develop. Intervention is therefore indicated. Pseudocysts in CP are less likely to resolve than their acute counterparts and thus more often require drainage. The indications for drainage are the presence of symptoms or complications. Although size alone is not a criterion for intervention, cysts larger than 6cm are more likely to be symptomatic and require treatment (Bornman PC S Afr Med J 2010). Percutaneous procedures are generally not favoured for these lesions due to an increased risk of failure, introducing sepsis or creating an external fistula. Endoscopic drainage is associated with a success rate of 65-95% and a low complication rate and is preferred to surgery due to its less invasive nature. (Beckingham IJ Br J Surg 1997). Strict morphological criteria are required however, relating to cyst maturity, intra-luminal bulging, wall thickness (less than 10mm) and vascularity, particularly in the presence of portal hypertension. To this end, careful cross sectional imaging and endoscopic ultrasound are important adjuncts in assessing patients for this modality of treatment. Transmural drainage may be transduodenal or transgastric depending on the best route into the cyst while transpapillary drainage is an

The Surgical Management of Chronic Pancreatitis 437

In the modern era, pylorus preservation as in a Pylorus Preserving Pancreaticoduodenectomy (PPPD) has been shown to result in less pain and nausea and improved quality of life when compared with the traditional Whipples pancreaticoduodenectomy (Mobius C Langenbecks Arch Surg 2007). This procedure can be performed with a mortality of 5-10% and morbidity of 20-40% and improves pain and quality of life in both the short and long term in up to 90% of patients (Bachmann K Best Pract and res Clin Gastro 2010). There are however a number of disadvantages relating to the sacrifice of functional pancreatic parenchyma and the nondiseased duodenum and common bile duct. The loss of natural bowel continuity and reduced endocrine and exocrine function result in side effects and reduced quality of life (Izbicki JR Ann Surg 1998 (228); Koninger J Surgery 2008). In order to allow organ preservation and reduce adverse effects, duodenum preserving resections of the pancreatic head (DPPHR) were developed. The Beger procedure was introduced in 1980 and was the first to include these principles (Beger HG Chirurg 1980). It consists of a subtotal resection of the head following transection of the pancreas above the portal vein. The Pancreas is then drained by an end-toside or end-to-end pancreaticojejunostomy using a Roux-en-Y loop. Physiological gastroduodenal passage and CBD continuity are therefore preserved. This procedure could be performed with low mortality (0-3%) and morbidity (15-32%) and long term pain relief in 75- 95% of patients (Izbicki JR Ann Surg 1995, Buechler MW J Gastrointest Surg 1997Frey CF Ann Surg 1994). The Frey procedure (Frey CF Pancreas 1987) subsequently combined an LPJ (as described by Partington and Rochelle) with a limited duodenum preserving excision of the head. Following exploration of the main pancreatic duct well into both the head and the tail, the head is cored out leaving a small cuff of parenchyma along the duodenal wall. This results in a lesser resection of the head than that described by Beger. In further contrast to the Beger operation, the pancreas is not divided over the SMV/portal vein complex making it an easier operation to perform. Care is taken not to enter the CBD. Drainage of the resection cavity within the head and from the opened main pancreatic duct within the body and tail is obtained with an LPJ using a Roux-en-Y loop (Frey CF Pancreas 1987). Good results have been obtained with substantial pain relief in more than 85% of patients while mortality is less than 1% and morbidity 9-39% (Izbicki JR Ann Surg 1995, Izbicki JR Ann Surg 1998, Beger HG Ann Surg 1989). Endocrine & exocrine function are well preserved and the operation may control complications such as CBD stenosis, duodenal stenosis and internal pancreatic fistulas. The Frey operation is currently the most widely performed operation for patients with an inflammatory mass in the head together with pancreatic duct dilatation while the Beger procedure is reserved for patients where the main pancreatic duct is not dilated (Bornman PC

Two further modifications of the above procedures have been described. The Hamburg operation employs subtotal excision of the pancreatic head including the uncinate process( a more extensive resection than the Frey operation but comparable to Beger's procedure) together with a V-shaped excision of the ventral aspect of pancreas into the pancreatic duct. Pancreatic-enteric continuity is re-established with an LPJ using a Roux-en-Y loop (comparable to the Partington-Rochelle and Frey reconstructions).This operation combines aspects of the Frey and Beger procedures, without transection of gland over SMV/portal vein. The extent of resection is customized to pancreatic morphology while the V-shaped excicion creates a trough-like new ductal system allowing better drainage of ductal side branches (Izbicki JR Ann Surg 1998, 227, Bachmann K Med Sci Monit 2008). In the Berne operation, an extensive duodenum-preserving resection of the head is performed (as in the

S Afr Med J 2010).

alternative option when communication with the pancreatic duct can be demonstrated. Surgery is indicated when endoscopic intervention fails or is not appropriate due to cyst morphology or patient factors. Surgical drainage of a pseudocyst may also be employed as part of an intervention planned for treatment of pain or additional complications. Pancreatic ascites is an uncommon but serious complication of CP which is managed in the first instance with paracentesis, nutritional support and endoscopic stenting of the pancreatic duct (Kozarek RA Gastrointest Endosc Clin North Am 1998; Bornman PC in Hepatobiliary and pancreatic surgery – a companion to specialist surgical practice 2009). Use of a somatostatin analogue remains controversial. Surgery is reserved for failures of conservative treatment. Bleeding from gastric varices related to segmental portal vein thrombosis is uncommon, thus the authors recommend intervention only once there is proven bleeding from gastric varices. Haemorrhage related to a pseudoaneurysm is best dealt with via selective angiography and embolisation due to the hazards of surgery in this setting. Surgery is reserved for failure of angiographic treatment.
