Figure 15.1 Trypsin encoded by the PRSS1 gene. (From Emw [own work], CC BY-SA 3.0 [http://creativecommons.org/licenses/by-sa/3.0] or GFDL [http://www.gnu.org/copyleft/fdl.html], via Wikimedia Commons.)

The first causative genetic mutation (protein serine 1 [PRSS1]) was reported in 1996 by Whitcomb et al. [6] in families from Kentucky and West Virginia (Figure 15.1). Others followed (Table 15.1), but PRSS1 mutations appear to be responsible for up to 80% of all cases, and in particular, there are two PRSS1 point mutations (p.Arg122His and p.Asn29Ile) that are most common, accounting for 90% of mutations in affected individuals. What proportion of children with HP express mutations is difficult to establish, though, and it varies with the population. In one Polish study, PRSS1 mutations were only found in about 10% of children with CP, reducing to 2.5% in those with acute recurrent pancreatitis [7]. By contrast, in a nationwide survey of Japanese patients with HP, PRSS1 mutations accounted for about 40% of the families involved.

PRSS1 encodes for a cationic trypsinogen, and such mutations cause molecular instability and early cleavage to trypsin, and hence autodigestion. Interestingly, mutations in PRSS2 appear to protect against the development of pancreatitis and mitigate intrapancreatic trypsin activity [8].

The second most frequent genetic cause in HP, at least in the West, appears to be due to mutations in the SPINK1 gene (e.g. N34S, also found in 1%–2% of the normal population) [9,10]. Certainly in the Japanese study referred to above, SPINK1 mutations accounted for about 30% of their affected families [8]. SPINK1 mutations (especially IVS3+2T>C) were found in 57% of a large group of 75 Chinese children with CP [11]. SPINK1 codes for the most important protease inhibitor of premature trypsinogen activation and consequent intrapancreatic trypsin activity [8,10].

Mutations in the CFTR gene may also contribute to the aetiology of CP, but were recognised only relatively recently in 1998 [12]. There are more than 1500 mutations recognised (classified into five groups), but the most common, at least in cystic fibrosis, is F508del, accounting for more than 60% of all mutated alleles in caucasians. Still, it has been shown that other rarer CFTR mutations are actually more common in non–cystic fibrosis chronic pancreatitis [13,14].

Loss-of-function mutations have also been recognised in the gene CPA1, coding for carboxypeptidase A1. This is one of the pancreatic metalloproteases, which hydrolyse C-terminal peptide bonds in dietary polypeptide chains. This was identified in about 3% of patients of German origin with nonalcoholic pancreatitis compared with 0.1% of controls. The incidence in those with an onset at <10 years had an even greater incidence of almost 10% [15].

Table 15.1 Genetic causes of CP

At one stage, it was felt that affected individuals with HP had an abnormally hypertrophied sphincter of Oddi and high intraductal pressures. This feature, if present, is perhaps now explained as prematurely activated trypsin passing through the sphincter, producing chronic inflammation, scarring and stenosis [16], and thereby deriving some benefit from a surgical sphincterotomy.

Perhaps due to its early onset, patients with HP appear to be predisposed to other more serious complications of CP. So, portal and splenic vein thrombosis appear more commonly. Importantly, these patients also have an increased lifetime risk of cancer, estimated at about 50-fold higher than in the normal population. By comparison, a 20-fold increased risk is observed in those with alcoholic CP [17]. The relationship is most obvious in those with PRSS1 mutations, but whether this is still true for those with SPINK1 or indeed CFTR mutations is not known.

15.4  AUTOIMMUNE PANCREATITIS

Autoimmune pancreatitis (AIP) is a type of CP with a suspected autoimmune aetiology. The concept was first proposed by Yoshida et al. in a paper published in 1995 [18], and the key immunological hallmark, elevated immunoglobulin G4 (IgG4), was first recognised by Hamano et al. in patients in 2001 [19]. It has also been known by a number of names, including lymphoplasmacytic sclerosing pancreatitis, alluding to the characteristic histological features of a dense infiltrate of plasma cells and so forth [20].

AIP has been suggested to account for about 2% of all cases of CP in adults and has a distinct elderly male (2:1) predominance. Children fitting the description have also been reported in a few series [21–24], but it needs to be looked for carefully and diagnosed critically, as it is certainly rare in this age group.

There are now believed to be two distinct variants:

1.  Type 1 AIP characterised by storiform^* type fibrosis, obliterative phlebitis and elevated numbers of IgG4 +ve plasma cells (typically >50 per high-power field). The type 1 variant of AIP is the pancreatic manifestation of IgG4-related disease; thus, both pancreatic and extrapancreatic recurrences are common. It is steroid responsive.

2.  Type 2 AIP characterised by granulocytic epithelial lesions and only occasional IgG4-bearing plasma cells. Bile duct involvement is uncommon in this type and is probably unrelated to IgG4-related disease. Disease recurrence is uncommon.

Extrapancreatic disease is a feature and might include sclerosing cholangitis, primary biliary cirrhosis, ulcerative colitis, Sjorgren’s^† syndrome, Hashimoto’s thyroiditis^‡ and retroperitoneal fibrosis.

15.4.1  Clinical features

AIP in younger patients seems to have distinct clinical features, such as abdominal and back pain, usually without jaundice. Serum amylase levels may be raised. Computed tomography (CT) scan shows diffuse enlargement of the pancreas with loss of lobular architecture (‘featureless’) and maybe a rind of hypoattentuation. Calcification and pseudocyst formation are distinctly uncommon.

The key diagnostic element, at least in adult AIP, is elevated serum levels of IgG4 (>135 mg/dL). These are seen in more than 90% of patients. However, it is not necessarily raised in younger patients [21,22], and confirmation of diagnosis has often relied upon open biopsy [21,24].

The main medical treatment of AIP has been oral steroid therapy (typically prednisolone, starting at 30 mg/day and tapered by 5 mg every 1–2 weeks). To prevent relapses, continued maintenance therapy with prednisolone 2.5–5 mg/day is sometimes required. Recently, mycophenolate mofetil has been used in a child with resistant features [25].

Table 15.2 illustrates the King’s College Hospital experience with six children (female-to-male ratio of 5:1) with suspected AIP seen between 2006 and 2013.

15.5  TROPICAL CALCIFIC PANCREATITIS

This entity was first described in 1959 in Indonesia [26], but it appears to be a worldwide problem in specific areas in Asia, Africa and South America. The region of Kerala in south India appears to have the highest prevalence in the world [27]. A number of factors have at one time or another been suggested as important in aetiology, including

1.  Chronic protein calorie malnutrition. This is believed to impair pancreatic function, perhaps making it more susceptible to toxins such as smoking and alcohol.

2.  Chronic cyanide toxicity due to overreliance on cassava^§ as a foodstuff. This contains cyanogenic glucosides which are activated releasing hydrogen cyanide (HCN) by an enzyme found in the root (linamaraze).

3.  Micronutrient deficiency (e.g. vitamins C and A).

4.  Genetic predisposition – SPINK1 and CFTR gene mutations [27].

Table 15.2 King’s College Hospital (2002–2013) experience with AIP (n = 6) Age at presentation 11.5 (9–13)^a years

^a Median (range).

The disease is characterised by a mean age of onset of about 12 years, abdominal pain and weight loss with an abnormal glucose tolerance. Radiographic imaging shows obvious pancreatic calcification and stones (often up to 1–2 cm in diameter). These patients have a high incidence of insulin-dependent but ketosis-resistant diabetes mellitus labelled ‘fibrocalculous pancreatic diabetes’ [28]. They also have an exceptionally high incidence of pancreatic cancer starting from around their 40s, affecting particularly the body and tail. This has been summarised as ‘pain in childhood, diabetes in puberty and death at the prime of life’.

There is no specific treatment, at least in the areas in which it is predominant. Diabetes can be hard to manage, as there is still pancreatic β-cell reserve, a low glucagon reserve and decreased adipose tissue, making blood glucose monitoring difficult and hypoglycaemia common [29,30]. Sometimes, surgical decompression (see below) is offered, or occasionally partial or subtotal pancreatectomy, particularly for those with intractable pain.

15.6  INVESTIGATIONS

Comprehensive investigation of a child with suspected CP should demonstrate evidence of biochemical pancreatic inflammation using serial lipase and amylase levels – although it has long been recognised from adult practice that end-stage disease may be accompanied by low levels. This should also include immunoglobulin estimation (specifically IgG4) and metabolic screening for lipid, calcium and copper abnormalities.

The pancreas is evaluated initially by ultrasound (US) looking at parenchymal irregularity and duct size and exclusion of biliary abnormalities, but magnetic resonance (MR), specifically magnetic resonance cholangiopancreatography (MRCP), will be required. Definitive delineation of the characteristics of the ampulla, main and accessory ducts can only be reliably accomplished by endoscopic retrograde cholangiopancreatography (ERCP) in children, although increasingly endoscopic US and MRCP may give approximately the same information (Figure 15.2). What these latter tests fail to do is provide any prognostic information on therapeutic endoscopic stenting, which in our practice is often the key indication for definitive surgery (Figure 15.3). Long-standing pancreatic disease may lead to vitamin deficiencies (D, A, K, E) and should be excluded.

15.6.1  Pancreatic exocrine function tests

15.6.1.1  NONINVASIVE TESTS: FAECAL ELASTASE-1 MEASUREMENT

Faecal elastase-1 (FE-1) is a quantification enzyme-linked immunosorbent assay (ELISA) test using a monoclonal antibody to human pancreatic elastase E1 in faecal specimens. Concurrent use of enzyme supplements does not affect the results, as these are of nonhuman origin. The results correlate well with the older standard, but invasive tests such as the secretin–pancreozymin test and the secretin–caerulein test tend to be less sensitive to mild exocrine deficiency [31]. For example, Naruse et al. [32] looked at the value of FE-1 measurements in a heterogenous population of children by comparing the test with an invasive secretin stimulation test. Using a cut-off value of >200 μg/g, the specificity of this test was >90% for detection of exocrine failure, with a sensitivity of about 60% for detecting definite CP and 76% for calcifying pancreatitis.

There may be other issues; thus, low values in infants should be interpreted with caution as they normally do not produce much pancreatic proteolytic enzyme or amylase. Similarly, transient low values may also occur in some cases of acute viral or bacterial enteritis and in children with mal-absorption due to untreated coeliac disease [33].

It is stable outside the body for at least 7 days, and samples can be collected at home and posted. Commercial sources include ScheBo Biotech (Germany) and Genova Diagnostics (United States).

15.6.1.2  MISCELLANEOUS TESTS

Faecal fat estimation over a measured period (e.g. 72 h) is a fairly crude index of exocrine function, and it has been suggested that more than 90% of lipase action has to be lost to be reflected in abnormal values.

Estimation of faecal chymotrypsin was at one time the routine test for exocrine evaluation [34], but it has now been superseded completely by FE-1 estimation. One crucial difference was that enzyme supplementation had to be stopped for at least 3 days prior to the test because of colorimetric overlap. Subsequent comparison tests with FE-1 also showed increased day-to-day variation and reduced sensitivity and specificity [35].

Figure 15.2 MRCP (left) and ERCP (right) showing dilated pancreatic ducts (arrowed). (Reproduced with permission from Ford K, et al., European Journal of Pediatric Surgery 2016, 26: 232.)

Figure 15.3 ERCP showing dilated and tortuous pancreatic duct (left) and ERCP showing stented pancreatic duct (right). (Reproduced with permission from Ford K, et al., European Journal of Pediatric Surgery 2016, 26: 232.)

The bentiromide (or N-benzoyl-L-tyrosyl-para-aminobenzoic acid [NBT-PABA]) test involved the ingestion of bentiromide. This is split by pancreatic chymotrypsin, liberating PABA, which is then absorbed by the small intestine, conjugated in the liver and excreted in the urine. A 6-hour collection of urine is obtained to determine the concentration of PABA conjugates. In adults, the recovery of 50% or more of the administered dosage in the urine during this 6-hour collection time is considered to be normal.

The pancreolauryl test has a similar principle in that fluorescein dilaurate is given by mouth. This is hydrolysed by specific pancreatic arylesterases to lauric acid and free fluorescein. The fluorescein is absorbed in the small intestine, conjugated in the liver and again excreted in the urine to be collected and measured. To evaluate the subject’s baseline absorption, conjugation and excretion, the test had to be repeated 2–3 days later with free fluorescein, and then recovery rates on both days expressed as a ratio.

Secretin-enhanced magnetic resonance cholangiopancreatography (SMRCP) is a possible tool of the future that is able to estimate various aspects of pancreatic exocrine function in addition to imaging the parenchyma and ducts. Intravenous secretin allows a ‘pancreatic flow rate’ to be calculated. So in one Chinese study of 53 subjects (both normal adults and patients with CP), ‘flow rates’ of about 8 mL/min were regarded as normal, and those with severe CP could only achieve flow rates of around 4 mL/min [36]. There was also good correlation with FE-1 values estimated concurrently.

15.6.2  Invasive tests

Invasive tests involving nasoduodenal or endoscopic aspiration have never been popular in paediatric practice. Most were based upon the stimulation of the exocrine pancreas by a variety of means, with the original being simply a standardised meal, and known as the Lundh test. This was replaced when cholecystokinin (CCK), and secretin became commercially available [37,38], and more recently ceru-lean. Clearly, both hormones stimulate different elements of pancreatic juice, so it is possible to measure volume output, bicarbonate output and concentration with secretin and a variety of enzymes (e.g. amylase) with CCK used as the secretagogue. The best correlation with histologically graded CP (adults) seems to be with maximal bicarbonate secretion, then amylase output and then total volume [38]. For example, achieving a bicarbonate concentration of >80 mmol/L in any of the samples collected is considered normal.

One recent report in children describes a typical test as involving upper gastrointestinal (GI) endoscopy with a suction tube passed through the endoscope and used to collect aliquots of ongoing duodenal secretions. Baseline collection aliquots were collected over 5- to 10-min intervals and then intravenous secretin given (0.2 mg/kg body weight) [31]. Elastase was the preferred enzyme measured.

15.7  MEDICAL MANAGEMENT

CP can probably be managed medically in most children, although of course this depends on the degree and character of the symptoms. Still, it is unlikely that recurrent painful episodes will be tolerated if there is a surgical alternative. There needs to be strict attention to appropriate analgesia requirements, and nutritional and growth failure should be treated with dietetic and interventional techniques. The details are, however, beyond the scope of this chapter [39].

15.7.1  Pancreatic enzyme replacement therapy

Pancrelipase (Creon ™^* [AbbVie Inc., North Chicago, Illinois]) is the most common enzyme supplement used for exocrine deficiency and is sourced from porcine pancreas in a slow-release formulation. Although provided as a capsule, for paediatric use, these can be opened and the microspheres mixed with acidic soft food (e.g. apple sauce and yogurt) or fruit juice.

The usual dosage is 1000 lipase units/kg (<4 years) and 500 lipase units/kg (>4 years), both per meal. The adult dose is 25,000–40,000 units/meal.

Side effects are uncommon, but they can cause oral irritation if chewed or perianal enzymatic irritation if too much passes through the GI tract. Fibrosing colonopathy was a specific complication described on its introduction in the 1980s and typically presented as right-sided colon strictures in children with cystic fibrosis [40]. Since changes to the formulation, it has become exceptionally rare.

15.8  SURGERY FOR CHRONIC PANCREATITIS

We believe that prior to consideration for surgical intervention it is important to demonstrate the anatomical indication appropriately and always attempt ERCP to delineate ductal anatomy. Typically, if there is a dilated, irregular duct, often with stones and debris, we would then go on to endoscopic stricture dilatation and stenting using an indwelling pancreatic stent (e.g. Zimmon™ stent, Cook Medical). Relief of symptoms (usually pain) then should predict those that would benefit from definitive surgery. If there is no response to stenting, then it is unlikely that surgery directed at improving duct drainage will be effective. Most recent surgical series tend to follow this route, although in reality, failure to achieve endoscopic visualisation of the duct or failure to control symptoms seems to be the route to surgery in many series [41].

There are few reports of endoscopic intervention only in children [42,43]. With increasing access to skilled paediatric endoscopists, this proportion will probably increase in the future. Iqbal et al. [44] reported a combined surgical and endoscopic experience from the Mayo Clinic. The endoscopic-only group (n = 38) had a higher rate of recurrence, more hospitalisations, increased narcotic use and more procedure-related pancreatitis than the surgical group.

Operations for CP have been classified into three types:

1.  Decompressive, for example, Puestow procedure, DuVal procedure, and Frey procedure

2.  Resection, for example, Whipple-like procedure and Beger operation

3.  Operations aiming to denervate or relieve the neurogenic aspects of CP

The procedure that is most commonly used in children is a lateral pancreaticojejunostomy, and it is actually the Partington and Rochelle modification of the operation first described by Charles Puestow in 1958^† [44,45]. Puestow’s technique was to separate the pancreas and splenic vessels (usually with splenectomy), open the dilated duct longitudinally and invaginate the whole of the body and tail into a Roux loop. What most now refer to as a ‘Puestow procedure’ leaves the pancreas in situ and anastomoses the opened-out duct to the Roux loop.

The key requirements are persistent pancreatic pain (usually intermittent) and a dilated main duct. HP is the usual underlying cause in most recent reported series in children, although it is not invariable.

Box 15.1 illustrates the steps of this operation. There are perhaps some modifications that can be used to extend the objective of opening out strictured parts of the duct, although actual reports in children are uncommon. The least invasive is the DuVal procedure [47] (Figure 15.4), draining only the transected tail, but it is prone to recurrence. The most invasive involves extending the duct incision medially to open the main and uncinate ducts lying in the pancreatic head, and is usually referred to as a Frey procedure [48,49]. Rarely, there is significant bile duct involvement or stricture in paediatric CP, and consideration should then be given to a concurrent hepaticojejunostomy. One of our children developed an initial bile duct stricture postradiation for lymphoma requiring hepaticojejunostomy, and then some years later a pancreatic duct stricture requiring a modified Puestow procedure.

Table 15.3 illustrates reported outcomes in 11 published series [41,44,49–58]. The largest series of cases dating back to 1973 has been from the Mayo Clinic, reported initially by Moir et al. in 1992 [51] and more recently by Iqbal in 2009 [44]. Virtually all reported series are from the United States, and there is clearly wide variation in reported outcome. The most common initial procedure is the Puestow operation, with only a single series describing the more radical Frey variant [49] (Table 15.4). What is apparent is that surgery is being performed increasingly in younger children. Actual complications from the surgery are few beyond actual failure to achieve their goal. The composite leak rate is small, and actual complications are rarely reported. Splenic infarction requiring splenectomy and subphrenic abscess has been reported [52]. Longer-term issues with exocrine failure requiring enzyme supplementation and endocrine failure requiring insulin occur in about 10% of cases, although sometimes the surgery reverses steatorrhea and so forth.

Figure 15.4 Schematic illustration of DuVal operation.

Despite an untoward, uneventful surgical procedure in most, it may not cure their symptoms, and the overall composite success rate is more than 90%.

A few surgeons have adopted a more radical approach in the first instance to CP in children. Notable is a series reported from a pancreas clinic in Bochum, Germany, involving six adolescents, of which three underwent a duodenal-sparing resection of the pancreatic head (Beger procedure) and the remainder had other types of pancreatic resection (neck, tail, etc.) [55].

Pancreatic surgeons have resisted laparoscopic surgery for longer than most specialists, but a small series of modified Puestow procedures have now been reported from Long Li’s group in Beijing illustrating feasibility and safety [57]. To reduce the technical problems of laparoscopic suturing, Meehan and Sawin from Seattle, Washington, have also reported this using a robot in a 14-year-old boy [59]. Endoscopic US was used in their case to localise the dilated pancreatic duct.

A radical alternative to retrograde duct procedures is total pancreatectomy with islet cell autotransplantation [60–62]. There are two American groups with the most experience in this technique and approach: Minneapolis and Cincinnati. Thus, Bellin et al. from Minneapolis reported the outcome of 24 children and adolescents during a 16-year period [60], with a further update on an additional 51 only 6 years later [61]. In summary, at 1-year posttransplant, 56% were insulin independent, but still, 40% took narcotics for pain control. Harvesting of salvageable islet cells in diseased pancreas is lower than normal glands, and perhaps one should not wait too long in the disease process for this alternative. By comparison, the Cincinnati group reported 14 adolescents, of which 4 had already undergone previous pancreatic operations. Predisposing genetic mutations were present in six. After about 1 year, all but four required regular insulin therapy, with a significant reduction in narcotic use. The longer-term rate of insulin independence may be even lower. A recent systematic review of five long-term reports of this procedure in both adults and children suggested that the risk of an insulin requirement could be more than 90% at 8 years [63].

So, if removal of the entire pancreas cannot provide the definitive pain-free solution, perhaps strategies should be sought targeted at the pain or at least the perception of pain. The afferent pain pathway from the pancreas starts in the coeliac plexus and runs to both sympathetic trunks on either side of the spine. Splanchnic nerves arise from the thoracic sympathetic ganglia: the greater (T6–T9), the lesser (T10–T11) and the least splanchnic (T12) nerves. Splanchnic nerves are usually easily visualised during thoracoscopy, and this is the current recommended approach for safe division [65]. Still, adult experience suggests that this may not offer perfect pain control, and there is a significant incidence of recurrent symptoms over time exacerbated in those who have already undergone endoscopic and surgical intervention – the likeliest scenario in children [66].

One complication that is rarely mentioned in this context but is certainly predictable is the long-term risk of cancer during adulthood. The scale of the problem is difficult to ascertain, as most studies are in adults who develop their disease as adults, often due to alcohol and which may not be the same as those developing CP during childhood. Still, there are estimates of a 15- to 20-fold increase in risk of adenocarcinoma compared with those without CP [67].

15.9  CONCLUSIONS

The diagnosis of CP is often straightforward once the disease is considered. Up to a half will have a recognised causal genetic mutation, although this does not seem to predict natural history or response to treatment. Perhaps genetic knowledge is most useful in acknowledging the full potential for severe disease and expediting effective interventional treatment. Full evaluation of the functional capacity of the pancreas and precise definition of ductal anatomy using US and MRCP complete the picture. ERCP and endoscopic stenting should identify that cohort of patients best served by duct decompression surgery.

Stay updated, free articles. Join our Telegram channel

Join