Total Pancreatectomy With Islet Autotransplantation

Total Pancreatectomy with Islet Autotransplantation

Chronic pancreatitis represents irreversible morphologic and functional changes in the pancreas from persistent or recurrent inflammation. Although an uncommon disease in childhood and adolescence, with an incidence of 4 to 13 cases per 100,000, the incidence is comparable with that seen in adults. The general concepts regarding presentation and management are well covered in Chapter 82. The clinical presentation and course of chronic pancreatitis are similar to that seen in adults. Thus the presentation in younger years is also one of abdominal pain, and definitive management focuses on elimination of pain to improve quality of life. There is a particular form of idiopathic chronic pancreatitis, termed tropical pancreatitis, which can present as recurrent abdominal pain in adolescence occurring in endemic regions. ^, An important consideration in management should focus on causes of pancreatitis that can be eliminated to avoid ongoing inflammation. The etiology of pancreatitis is of particular interest when considering the differences between adult and pediatric patients. In adults the principle causes are alcohol and biliary tract disease, whereas in children hereditary and idiopathic causes predominate, especially in patients requiring surgery. Other causes of pancreatitis in children include a preponderance of patients with trauma, congenital anatomic variations including pancreatic divisum, annular pancreas, and choledochocysts. A minority of children, 10% to 30%, can develop acute recurrent pancreatitis. ^, Children with acute recurrent pancreatitis will most frequently have an idiopathic or hereditary etiology, and 42% of these patients will go on to develop chronic pancreatitis. The natural history of chronic pancreatitis would be of great interest in the pediatric population. The conflicting outcome variables of possible interest would include eventual relief of pain, so-called burn-out, versus risk of developing narcotic dependence and irreversible exocrine and endocrine insufficiency during the period of medical management. These competing outcomes are expected to likely play out over the course of a longer life expectancy in children than in the adult population. The natural history data for chronic pancreatitis in adults are illustrative for their shortcomings. Some European centers have tried to investigate the long-term success of complete pain relief with medical management alone. Data from these centers suggest that any sustained pain relief for chronic pancreatitis requires decades to achieve, requires removal of the initiators of pancreatitis (alcohol in adults), and is likely dependent on the type of pancreatic pain (type A vs. type B). Type A pain pattern is typically short lived with episodes lasting less than 10 days separated by long pain-free intervals of several months or longer. In distinction, type B pain pattern is characterized by prolonged periods of persistent pain that may last for months, exacerbated by severe, worsening pain superimposed on the baseline pain. In addition, any sustained pain relief will only predominantly occur at the expense of addiction and complete functional insufficiency. ^, This has not led to a groundswell to treat adults with medical therapy alone where durable pain relief without substantial quality of life consequences has been the general experience of patients and treating physicians.

The differences in etiologies are of importance when considering the natural history of pancreatitis. In adults the major causes of chronic pancreatitis are potentially mutable, especially the elimination of alcohol. In the pediatric population, the most common causes of chronic pancreatitis often cannot be altered because they include idiopathic and hereditary etiologies. Chronic idiopathic pancreatitis when it develops in juveniles may be best characterized as early onset of the severe type of pain but with better preservation of pancreatic function over time compared with adults. Hereditary pancreatitis is better characterized and is of greater concern for long-term consequences. Predisposing germline mutations have been identified in trypsinogen, trypsinogen inhibitory pathways, and bicarbonate secretory pathways. The most common hereditary cause of pancreatitis is a mutation of the cationic trypsinogen gene (PRSS1), which has a penetrance of greater than 90%. The median age of symptom onset in hereditary pancreatitis is 10 years, yet the median age of diagnosis is at 19 years. It is hoped that greater recognition of pancreatitis in children and its etiologies can lead to earlier diagnosis and therapy. There are no clinical or morphologic differences in patients with hereditary chronic pancreatitis compared with pancreatitis of other etiologies, but half will ultimately die from pancreatic adenocarcinoma. Long-term follow-up of patients with hereditary pancreatitis is important when considering total pancreatectomy and islet autotransplantation in terms of beta cell function. Children with chronic pancreatitis (CP) rarely have diabetes as a complication, present in 5% of patients. In distinction in patients with hereditary pancreatitis, 5% of patients will develop diabetes by 10 years after onset of symptoms and 18% by 20 years. Therapeutic interventions, endoscopic or surgical, should be considered when an accurate diagnosis has been established and significant alterations in quality of life are occurring, which in children includes inability to remain in school. Pediatric patients with chronic pancreatitis frequently require hospitalization, averaging five times per year, significantly impacting schooling and the entire family. In children, narcotic dependence is not required to justify intervention, and treatment to prevent addiction is advisable. Long-term success with endoscopic therapy has been demonstrated in selected patients, including in the pediatric population. Patients with pancreatic divisum have shown reintervention for pain in 39% and no need for surgical salvage in 45/47 (95.7%) patients followed by a median of 67 months in one series. Similarly, in another series, endoscopic stenting specifically in pediatric patients showed an 81% improvement in pain and a need for subsequent surgery in only 12% of patients at a median follow-up of 61 months, although objective measures of pain assessment were lacking.

There are numerous surgical options available to treat patients who have failed medical management of chronic pancreatitis. The plethora of surgical options should highlight the need to tailor the appropriate operation for each patient and not be a reflection of uncertainty as to how to surgically manage the disease, including in children. ^, All surgical options available in adults are options for the pediatric population and entail one of two categories: resectional or drainage. The drainage procedures are aimed at decompressing the ductal system either by ablating any sphincter obstruction or by surgical bypass to the duct directly. Surgical procedures of the sphincter include both the minor or major ampullae; the former is classically associated with pancreatic divisum and the latter with sphincter of Oddi dysfunction. The diagnostic study suggesting inadequate sphincter emptying is secretin-enhanced magnetic resonance imaging (MRI). Endoscopic sphincterotomy has substantially reduced the frequency of surgical sphincteroplasty, although surgery may still be necessary for recurrent stenosis postendoscopic drainage, or when the orifice is too small to allow access for endoscopic therapy. ^, Paramount in determining if a patient is a candidate for surgical therapy directed at a sphincter is the character of the pancreatic duct and length of sphincter stenosis. Multiple strictures in the main duct will not respond to sphincter-directed procedures alone, and the length of obstruction at the sphincter must be the level of the duodenal wall, not within the pancreatic parenchyma, for these procedures to be efficacious. The length of obstruction and diameter of the dorsal duct are critically important for therapy directed at the minor sphincter where it enters the duodenum at a right angle, making sphincteroplasty difficult without coring out the sphincter complex. In addition, the maximal size of the opening can be made no larger than the diameter of the duct. In properly selected patients, the results of surgical sphincteroplasty are acceptable, making it an appropriate initial operative approach for some pediatric patients with pancreatic divisum. Ductal decompression, by suturing the main pancreatic duct, with or without incorporating a chronic pseudocyst, is more commonly done and includes lateral longitudinal pancreaticojejunostomy (Partington-Rochelle modification of the Puestow procedure).

The modified Puestow procedure is the most frequent decompressive procedure done in children and is applicable to patients with firm fibrotic glands with a dilated duct due to multiple strictures without an inflammatory mass. Initial results from this operation in children are good, but long-term pain control is difficult to assess due to both poor and biased follow-up. In some series, follow-up is absent in half the cohort of operated patients. At best, pain relief is achieved in 50% to 80% of patients, and this result is better than what can be achieved in comparable patients managed with endoscopic therapy. Unfortunately, should duct decompression fail and total pancreatectomy with islet autotransplantation (TPIAT) is required, the islet yield and insulin independence are significantly impaired. In pediatric patients, endoscopic treatment is associated with more admissions for recurrent pancreatitis, with comparable pain response with a tailored surgical approach.

Resectional treatment inherently includes a component of ductal decompression, but some component of parenchymal loss is inherent. These procedures involve a proportional loss of the body and tail of the gland, or the pancreatic head. The most limited type of pancreatic resection is the Duval procedure, where a small portion of the tail is removed and the remaining gland is sutured to a defunctionalized limb of jejunum principally for ductal drainage. This procedure is infrequently done but may be indicated for distal outflow obstruction where the duct is uniformly dilated and a sphincteroplasty is not technically achievable, particularly in selected patients with pancreatic divisum. Distal pancreatectomy up to and including subtotal pancreatectomy can be considered for focal pancreatitis of that portion of the gland. Focal pancreatitis necessitating a distal pancreatectomy is not common. In one series, only 12% of patients undergoing distal pancreatectomy had the procedure for focal pancreatitis, and the procedure resulted in substantial loss of gland volume. Expected loss of insulin secretion of 50% to 55% is expected for major resections of the pancreas, either body or head. Pancreatic head resections involve complex operations that either preserve the duodenum (Beger, Bern, or Frey procedures) or remove the duodenum and distal bile duct (pancreaticoduodenectomy, Whipple procedure). All remove an inflammatory head mass, considered to be the location responsible for the elicitation of pain, the “pancreatic pain pacemaker.” They all achieve excellent pain relief in patients with head-centric disease, and the Whipple procedure removes areas secondarily affected such as duodenal and biliary strictures. These procedures have been used in appropriately selected children with chronic pancreatitis. ^{, ,}

The ultimate option for management of pancreatic pain in chronic pancreatitis is total pancreatectomy. This has the advantage of completely removing the offending source of pain and leaves no potential for recurrent disease. Clearly irrevocable, it has immediate and long-term consequences. Fortunately the operative consequences of total pancreatectomy have improved so that its morbidity and mortality outcomes are comparable to pancreaticoduodenectomy (Whipple procedure), but the long-term metabolic morbidity of metabolic bone disease and exocrine and endocrine insufficiency appropriately caution its application. The major endocrine insufficiency of diabetes mellitus in the setting of no islet cell mass is meant to be ameliorated by the use of islet cell autotransplantation. Although the clear indication for total pancreatectomy and islet autotransplantation is to improve quality of life through pain reduction and return to normal activity, this is supplemented by return of the islet cell mass, which may prevent the onset of diabetes or improve glucose control.

Paramount in importance is proper patient selection, particularly in afflicted children with chronic pancreatitis where decisions regarding the long-term success of medical management for chronic pancreatitis and durability of islet cell function are incompletely known. A multidisciplinary team assessment of prospective candidates is ideal, and at our center this includes involvement of the primary treating physician or pediatrician, gastroenterologist, surgeon, clinical psychologist specializing in chronic pain, and an endocrinologist. A decision to proceed with TPIAT assumes consensus by all members of the team as well as the patient and family. Etiology of pancreatitis weighs mightily in the decision-making because hereditary pancreatitis is as immutable as the operation of total pancreatectomy. Although the chief indication is for pain relief, it is important to realize that degree of symptoms does not clearly correlate with morphologic changes in the pancreas. Indeed, patients typically candidates for this procedure have “minimal change pancreatitis,” which belies their incapacitation and frequent hospitalizations despite a clear lack of radiologic abnormalities. Minimal change pancreatitis can therefore be difficult to diagnose and requires an appreciation of the entity, as well as investigational tests beyond endoscopic retrograde cholangiopancreatography (ERCP), magnetic resonance cholangiopancreatography (MRCP), and computed tomography (CT) imaging, which may not be as commonly available as endoscopic ultrasound (EUS) and pancreatic function testing. EUS allows the identification of specific parameters used to grade the extent of chronic pancreatitis in a reproducible manner and is less dependent of subjective interpretation. These EUS findings include hyperechoic parenchymal foci, strands, hypoechoic lobules, cysts, main duct irregularity, ductal dilation, hyperechoic duct walls, visible side branches, and calcifications. The standard diagnosis of chronic pancreatitis is made when five of these nine features are found. Some authors believe a better sensitivity to specificity balance to diagnose minimal change chronic pancreatitis may require finding only three of the nine features. Pancreatic function testing is aimed at quantifying pancreatic bicarbonate secretion into the duodenum with secretin stimulation. This has progressed at our institution to endoscopic duodenal aspiration, and a bicarbonate level of less than 80 mg/L is suggestive of exocrine dysfunction. ^, Supportive objective data are valuable in making the diagnosis of minimal change chronic pancreatitis that must also be supported by clinical features of chronic pancreatic-type pain. This might be difficult in the pediatric population, where the incidence of chronic pancreatitis is low and familiarity with the disease is uncommon. The indications for TPIAT in chronic pancreatitis are the same in children as in adults, with pain being responsible for all reported cases. There is a heightened need to correctly identify pediatric patients for consideration of all surgical options because the associated narcotic dependence can impair normal growth and development and important childhood activities such as school attendance. It may be difficult to know what time frame prior to surgical intervention is appropriate, but the reported median time to surgical intervention in pediatric patients is 4 years. Patients who have failed prior surgical treatment are candidates for TPIAT, understanding that the islet yield is less than for an unaltered gland. A key factor in selecting patients for TPIAT is assessing the extent of islet dysfunction. Diabetic patients are candidates for the operation, providing their beta cell mass is sufficient to produce C-peptide. ^{, ,}