roèník 11,2003 è.4
Reports

35th meeting of the European Pancreatic Club - Liverpool, June 18-22, 2003

Ake Andrén-Sandberg

The 35th meeting of EPC was held in Liverpool under the guidance of the Danish (but living in England longer than most Englishmen; sic!) physiologist Ole Peterson and the surgeon John Neoptolemos. As usual it is a limited group of dedicated persons who know each other well and are happy to have the possibility the discuss pancreatology among friends at least once a year. The program was as good as ever, and the organisation perfect. Maybe we missed some of the most well - known continental pancreatologist, but the size never influence the quality of a pancreatic meeting! This year there was on the other hand more basic scientist than usual as the last 2 days was occupied by an "International pancreas basic science workshop: beyond frontiers".
Liverpool was founded in 1207 and got its Royal Charter from King John, but was a small town until the 18th century when it became the second largest harbour in England due to business with Africa and the West Indies ("triangle trade"). The first railway in England was built 1830 between Liverpool and Manchester, and at that time there was a strong immigration of Irish into the town - the Irish has ever since made important landmarks in the town. Liverpool is today for foreigners probably best known as "home of the Beatles" and its two Premier League clubs (Liverpool FC and Everton). However, the Merseyside town has also a lot of other things to offer, such as galleries, museums, theatres, and restaurants, and is the European capital of culture in 2008. The Royal Liverpool Philharmonic Orchestra was founded already in 1840, whereas John, Paul, Georg and Ringo had their first number one hit not until 1962! Including the suburbs about 1.5 million people now live in Liverpool. The town once was known for its shipping, with more than 10 km of docks. The importance of this has decreased drastically since the 1950s and there is today no more coal - mining in the area and the shipping is streamlined in quite another way than 50 years ago. The Albert Dock near the centre was rebuilt in the 1980s and is today one of the most well - known attraction for the tourists. The Anglican Cathedral is the largest in England, and was built 1904-78.
The beginnings of the modern University of Liverpool can be traced back to the founding of the Liverpool Royal Institution in 1814 and the Liverpool Royal Institution School of Medicine and Surgery in 1834.
The social program, including the accompanying persons' program, was excellent with a special highlight at the garden party in Neston by the Neoptolemos: food, friendship, performance was fantastic as well the liquid challenge of the pancreas.

Pancreatic physiology

John Williams from Ann Arbor, Michigan, discussed growth of acinar cells. Pancreatic growth occurs in a number of conditions, for example with normal body growth in young animals, but also with prolonged demand for digestive enzymes. An experimental model is trypsin inhibitor feeding model with soya inhibitor or camostate, which increase plasma CCK from 1 pM to about 2-4 days. Pancreatic wet weight can be expected to double in about 10 days. In CCK - receptor knocks out mice the increase of endogenous and exogenous CCK is totally abolished. Calcineurin is a potential target for increasing intracellular free Ca2+ and is required for growth of heart and muscles. It has been shown that camostate feeding induces calcineurin in the pancreas. Cyclosporin A blocks pancreatic growth induced by camostate. However, camostate feeding induces multiple pathways for pancreatic growth. An activation of the CCK-A receptor leads to transcriptional regulation of gene expression of c-fos, c-jun and c-myc mRNA in isolated rat pancreatic acini. On the other hand, camostate feeding induce all these but also EGR-1 protein in about 2 hours and is normalised again in 4 hours. Pancreatic growth induced by endogeneous CCK is dependent on calcineurin, but also other pathways are possible.
The physiologist Yoshio Maruyama from Sendai (Japan) spoke on GTP-binding proteins, i.e. the proteins next to the receptor. That was difficult to follow for a clinicians, even though he was a fascinating, artistic speaker.
It has been shown in healthy volunteers that there is an immediate regulation of pancreatic lipase secretion by ingested fat (1).

Acute pancreatitis

There are at least five major classes of heat shock protein (HSP) according to Saluja, Boston. It seems as if HSP has a cytoprotective role in pancreatitis. It seems as if the HSP blocks the redistribution of enzymes - lysozymes and digestive enzymes - in patients with an insult to the pancreas potentially leading to acute pancreatitis. This means that if HSP-protein can be upregulated clinically it may be possible to use these findings also in the clinical situation, for example before ERCP.
Bile - induced pancreatic acinar cell damage was discussed by Shmuel Muallem from Dallas and started - of course - with Opie's report from 1901. Nowadays it is well accepted that there is no pressure difference between the pancreatic duct and the bile duct, which means that the discussion must be deepened. If bile is added to pancreatic cells there will be an influx of calcium into the pancreatic acinar cells in a slow process (slow = minutes). It seems that the bile acids can open up calcium channels, but it is also possible that bile can to inhibit the calcium - pump. Chelation of calcium prevents bile acids - induced cell damage and death. On the other hand, increasing calcium intracellulary activates trypsinogen, which may be the starting point of acute pancreatitis.
Nordback and co - workers showed that "Alcohol Use Disorders Identification Tesat (AUDIT) gave 22 + 9 points to patients with alcoholic pancreatitis (n=38) and 4 + 8 to those with gallstone pancreatitis (n = 8) (2).
Prophylactic antibiotics with meropenem in a randomized multicenter prospective study of 20 + 21 patients did not reduce neither the rate of infectious complications nor mortality (3).
In a population of 659 consecutive patients undergoing 897 ERCP procedures there were post - ERCP pancreatitis in 6 percent. Pancreatic secretory trypsin inhibitor, PSTI, concentrations increased clearly within hours after the procedure in patients with ERCP - induced pancreatitis.The concentrations of both trypsinogen-1 and -2 increased also sharply and rapidly, however, trypsinogen-2 concentrations were higher than those of trypsinogen-1. The concentrations of trypsinogen-1 decreased, also, much more rapidly. Trypsin-2-alpha1-antitrypsin complex increased slower than the other measured variables (4). Serum matrix metalloproteinase-9, MMP-9, allows a good early prediction of pancreatitis - associated pulmonary complications, but does not discriminate between oedematous and necrotising disease (5).

Chronic pancreatitis

David Whitcomb, Pittsburg, gave a state-of-the-art lecture on "What can genetics tell us about the pathophysiology of chronic pancreatitis". He saluted Michael Steer's article in New England Journal of Medicine 1995; 332: 1482-90 where it was stated that "chronic pancreatitis remains an enigmatic process of uncertain pathogenesis, unpredictable clinical course, and unclear treatment". Whitcomb than started by describing hereditary pancreatitis as an unusual form of acute and chronic pancreatitis that runs in families following an autosomal dominant pattern. Acute pancreatitis in seen in 80 percent with the mutated gene, chronic pancreatitis in 50 percent of those with acute pancreatitis, and pancreatic cancer in >40 percent of those with chronic pancreatitis. Most cases are caused by gain-of-function mutations in the trypsinogen gene.
The normal is that enterokinase activates trypsinogen in the gut lumen and this then starts the cascade that change solid food to liquid. What is interesting is how trypsin both can activate itself and inactivate itself. Also, there are several different ways for the acinus to protect itself from activated trypsin:
- proenzymes
- low calcium
- compartmentalized
- trypsin inhibitor
- zymogen autolysis
- remote activation
- a1- antitrypsin and b2 microglobulin

According to Somogyi et al (Gastroenterology, 2001:120: 708-17) the etiologies of chronic pancreatitis are:
- Toxic-metabolic
Alcohol
Hypertriglyceridemia
Hypercalcemia
Pharmacological substances
Organophosphates
Scorpion toxins
Methylene chloride
- Mechanical
Choledocholithiasis
Periampullary obstruction
Congenital malformation
- Miscellaneous
Vascular
Infections
Hereditary
Tropical pancreatitis
Cystic fibrosis
- Suspected
SOD
Pancreas divisum
Autoimmune

The etiologies of chronic pancreatitis are nearly identical to the etiologies of recurrent acute pancreatitis. This observation is consistent with the concept that chronic pancreatitis is associated with recurrent acute pancreatitis. Whitcomb has proposed his own classification of chronic pancreatitis (Gastroenterology 2001; 120: 682-707), called "TIGAR-O":
- Toxic-Metabolic
Alcoholic
Tobacco smoking
Hypercalcemia
Chronic Renal Failure
- Idiopathic
Tropical - Genetic
Autosomal dominant
Autosomal recessive/polygenic
- Autoimmune
- Recurrent and severe acute pancreatitis associated chronic pancreatitis
- Obstructive

The lesson from the studies on hereditary pancreatitis is that susceptibility to acute pancreatitis is associated with loss of protection from prematurely activated trypsinogen and susceptibility to chronic pancreatitis is associated with recurrent acute pancreatitis. Cystic fibrosis is the most common lethal genetic defect. It is an autosomal recessive , where the CFTR gene mutations are well described. Despite that genetic and functional testing - are not there - yet. Pancreatitis is a feature of atypical cystic fibrosis. CFTR is a gene regulating chloride and bicarbonate channels. The mutations may be classified into 6 functional groups: - class 1-3 not function
- class 4-5 some function
- class 6 - alter other channels

Pancreatic injury can occur through 4 mechanisms:
- CFTRsev/CFTRsev (0 %) = cystic fibrosis
- CFTRsev/CFTRm-v (<10 %) = susceptibility to idiopathic chronic pancreatitis (~ 80x)
- class 6 mutations (5 %) = susceptibility to idiopathic chronic pancreatitis
- CFTRm-v / other factors (genetic or environmental)

Once trypsinogen is secreted into the calcium-rich pancreatic fluid the "autolysis" mechanism is lost. Pancreatic CFTR is involved in flushing the pancreatic duct of digestive enzymes. Failure could lead to autodigestion. Cystic fibrosis causes typical chronic pancreatitis (with protein-filled ducts) not pancreatic agenesis or atrophy and begins in infancy at the time of trypsin expression, and is detected by elevated serum trypsinogen. Pancreatitis in atypical cystic fibrosis is often associated with CFTRmv mutations plus factors that further reduce duct clearance (pancreas divisum).
Regarding alcohol and chronic pancreatitis (Etemad and Whitcomb. Gastroenterology 2001; 120: 682-707) it is known that alcohol preceded chronic pancreatitis in 55-80 percent of patients. Onset occurs after ~150 g/d x 15-20 years, but the ranges for this are largely unknown. However, it is very interesting that only 5-10% of alcohol consumers develop pancreatitis. There is no "threshold" for alcohol toxicity. I should also be borne in mind that 30 percent of "alcoholics" in the USA had "social" or "uncertain" intake. That may be one of the reasons that the epidemiological relationship between alcohol consumption and chronic pancreatitis is weak, or at least weaker than expected. Animals fed alcohol seldom develop pancreatitis. The clinical course and progression differs among patients. Compared with white patients, blacks have 2-3 times bigger likelihood to be hospitalized for chronic pancreatitis than alcoholic cirrhosis in the US. According to Whitcomb (Best Pract Res Clin Gastroenterol 2002; 16: 347-63) chronic pancreatitis requires both a trigger and a continued injury, such as alcohol or hyperparathyroidism - the SAPE hypothesis. Therefore, chronic pancreatitis should be regarded as a complex disorder requiring susceptibility, a triggering event, and factors driving pro- and antiinflammatory pathways leading to fibrosis.
Jeremy Wilson from Sidney started his lecture on pathophysiology by stating that there is a paradox in alcoholic pancreatitis:
- not all alcoholics develop chronic pancreatitis
- the risk of developing pancreatitis increases with the amount of alcohol consumption.

The individual susceptibility can not be explained by HLA, detoxifying enzymes, CYP2E1 polymorphism, ADH genotype, TNF-alpha polymorphism, alpha-1-antitrypsin etc. The drinker's pancreas increases the content of protein (digestive enzymes) in the acinar cell but decrease the membrane stability. However, there must be some kind of trigger factor as well. What is known is that some transcription factors are increased, leading to more cytokines, which seems to be working in an autocrine fashion. It seems to be a necrosis-fibrosis sequence functioning. The pancreatic stellate cells are involved in this process and can be induced by some cytokines, but also of ethanol and oxidant stress. There is no overwhelming evidence that smoking is a risk factor as has been proposed by some researchers (also at this meeting). The initial injury may be to the acinar cell, the ductal cells (protein plugs due to lithostatin) both or neither - this is still not fully understood. There is also an increased risk with N34S SPINK1 mutation is very high: 40-600 times increased relative risk depending on how many mutations are found.
Christhoph Beglinger, Basel, told us that his speech on medical treatment could be given almost exactly in the same way ten years ago - development is slow. Trapnell in 1979 reported that pain was related to alcohol consumption in 75 percent of patients with chronic pancreatitis. However, this statement has been challenged, and today it is not clear if pain and immediate alcohol consumption is related. Enzyme therapy treatment gave pain relief in 36/49 patient in three studies (73 %). However, there was no effect in four other studies, with a larger number of patients. The conclusion is that from a scientific point of view there is little support for the study. In the clinic it is not effective - at least in alcoholics (who has so many other problems and who is so difficult to evaluate as they may have some positive social and economical effects of the pain). If there is steatorrhoea there should be 25 - 40.000 unites of lipase per meal; if less enzymes is given the diarrhoea is reduced by only 50 percent. In some patients a proton pump inhibitor can make the enzymes more effective. Paracetamol, NSAIDs and tricyclic antidepressives should always be tried before opiates. If there should be a priority between stop drinking and stop smoking, drinking is most important (the patient compliance may be low if the patients have to stop both immediately).
Myriam Delhaye from Brussels gave her opinion on endoscopic treatment of chronic pancreatitis. Rösch et al. has in Endoscopy in 2002 shown that endoscopy is of value if the main pancreatic duct is obstructed. In Brussels 123 consecutive patients 1987-1989 has been followed up after main pancreatic duct clearence. The immediate pain relief (74-95 %) was related to technical success whereas the long-term results were related to the use of opiate before the treatment, the pain severity and the duct morphology (strictures). Rudi Amman of Zürich rightly noted that most patients with long-standing chronic pancreatitis have dilated main pancreatic duct without outflow obstruction - and pain. John Neoptolemos also stated that patients with main pancreatic duct stones always have parenchymal disease, which make the evaluation of removal of ductal stones difficult.
Claudio Bassi (Verona) gave his expertise on drainage operations. He stated that it is important to treat the disease and not only the morphology of the disease. Pain is probably related both to high pressure in the pancreatic duct and gland, and to neuritis. The main disadvantage with drainage operation is the risk of missing a resectable pancreatic tumour. However, most patients with enlarged ducts do very well after a drainage operation, and do not need further in-patient admissions. Bassi - and the audience over all - did not use preoperative stenting as a prognostic instrument, i.e. that it is not known if a positive outcome of stenting indicate a positive outcome of drainage procedures.
From Jacob Izbicki's group in Hamburg was given an overview of resection procedures in chronic pancreatitis. There was nothing new, but the speaker had an interesting discussion on quality of life. There are no randomised clinical trials in chronic pancreatitis. So, up to now there are no preferences for any operations concerning quality of life in the long - term results (7 years). Markus Büchler (Heidelberg) did not agreed on this, as the short-term results are better after smaller operations. Also the frequency of diabetes is higher after Whipple operations.
One of the headline speakers was Markus Büchler who in a special state-of-the-art spoke on advanced chronic pancreatitis. He started by saying that chronic pancreatitis is primarily not a surgical disease, but surgery is used when medical treatment is insufficient. This means that surgery is not an alternative to medical treatment, but a complement in the most severe or special cases. Büchler is impressed with the finding starting with Dale Bockman, and then continued many years, where the nerve fibres in the chronic pancreatitis is enlarged and infiltrated by immune cells: the more infiltration, the more pain. Nerve growth factor and its receptor has also been found overexpressed in the metaplastic ductal cells, ductal complexes and the dying acinar cells in chronic pancreatitis. It was also found cytotoxic cells and chemokines different from normal pancreas. The neural alterations are independent on etiology: alcoholic, idiopathic and tropical pancreatitis. With gene array analysis of 5600 genes there were 152 increased in both pancreatic cancer and chronic pancreatitis, and 5 increased only in chronic pancreatitis. They must be further characterized. To Büchler there are strict indications for surgery in chronic pancreatitis:
- benign enlargement of head of pancreas,
- suspicion of pancreatic cancer,
- obstruction of bile duct,
- obstruction of duodenum,
- obstruction of main pancreatic duct,
- large retropancreatic vascular structures,
- pseudocysts > 6 cm,
- otherwise intractable pain.

Büchler advocated for chronic pancreatitis a modification of the Beger and the Frey operation and called it the Bern operation. It has been tried by the Farka's Hungarian group in 30 patients and published in British Journal of Surgery and is a combination of the best parts of the Beger and the Frey operations.
It can also be mentioned that Büchler and his group have now worked in Heidelberg from October 2001. Up to June 2003 there has been done 443 pancreatic operations; i.e. more than 25 pancreatic operations per month lately, 2/3 are resections. For the 300 resections the mortality is 0.7 percent, pancreatic fistula rate 3 percent and reoperation rate 5 percent.
It was also reported some early effects of eradication of Helicobacter pylori in patients with chronic pancreatitis with a decreased pain and dyspeptic symptoms. Blood levels of trypsin decreased and C-peptide increased (6). It has also been found that Helicobacter pylori lipopolysaccharide increased pancreatic secretion as well as plasma gastrin and CCK levels. Suspension of live Helicobacter pylori given intraduodenally significantly raised both pancreatic secretion and plasma gastrin level (7).

Stenting of biliary strictures

On of the most important reports during the meeting came from Magdeburg and was dedicated to success and failure of endoscopic stenting of biliary strictures (not pancreatic!) in chronic pancreatitis. Altogether 61 patients with symptomatic common bile duct strictures caused by alcoholic chronic pancreatitis were treated by endoscopic stent insertion for one year with scheduled stent exchanges every three months. After the treatment period all patient were included in a special follow-up program. Initial endoscopic drainage was successful in all cases, with complete resolution of obstructive jaundice. However, after one year from the initial stent insertion the obstruction has resolved in only 31 percent of cases and the patients could do without stenting. During a follow-up period of 40 months (range 18 - 66 months) 16 of these 19 patients had no recurrence, giving a long-term success rate of 26 percent. Of the 45 patients that needed definitive therapy 12 were treated with repeated plastic stent insertion or by insertion of a metal stent (n = 3), whereas 30 patients underwent surgery. Of the 39 patients with calcification of the pancreatic head only 8 percent were successfully treated by a 1-year period of plastic stent insertion, whereas in 13 of 22 patients without calcifications (59 %) this treatment was successful. Regarding these figures should, however, be remembered that they relate only to the jaundice problem; i.e. some of the successful patients might need other treatment modalities, e.g. surgery, due to the pain (8).

Enzyme supplementation

It was advocated in a study of 31 patients with severe chronic pancreatitis and exocrine pancreatic insufficiency that oral pancreatic enzyme supplementation cannot be correctly optimized based on clinical evaluation. Symptoms and signs as serum levels of liposoluble vitamins often remain abnormally low despite a theoretically adequate oral enzyme substitution (9).

Quality of life

In a follow-up study of 110 patients treated with endoscopy and 108 treated with surgery for chronic pancreatitis, EORTC-QLQ-C30/PAN-26 questionnaires were used. Although the patients before endocsopic treatment had better nutritional status and quality-of-life scores than the surgically treated patients the average results were the same after an average of 25 - 28 months (10).

Pancreatic pseudocyst

In four years time 48 patients were treated for pancreatic pseudocysts after chronic (n = 26), acute (n = 19) and equivocal (n = 3) cases. It was performed transmural gastropseudocystostomy in 26 cases, transmural- and transpapillary duodenopseudocystostomy in 16 cases and unsuccessful drainage in 6 cases. There were 1 bleeding, 12 septic conditions and 2 severe pancreatitis accounting for a 29 percent complication rate. Four cases were operated on due to complications. In 88 percent of the cases the pseudocyst disappeared completely and the stents were removed after 4 months (11).

Pancreatic cancer

Epidemiology
In a combined effort between Sweden and Latvia it was shown that there was an increased mortality rate of pancreatic cancer from 3.9 to 4.5 (1998-2001) per 100,000 persons in the 420 persons related to those with pancreatic cancer compared to the 420 controls. This means that in Latvia hereditary factor is responsible for less than 2 percent of all pancreatic cancer. On the other hand there was a correlation to smoking (46 %) and asthma (5 %) (12).
In a population-based case-control study of 15.001 pancreatic cancers reported to the Swedish Cancer Registry 1987-1999 were compared to 77000 randomly collected controls matched for age and gender. Primary sclerosing cholangitis (odds ratio 2.6 -3.2, 95 % confidence interval 1.8 - 5.8), chronic pancreatitis (odds ratio 6.9 - 14, 95 % confidence interval 3.3 - 22), acute pancreatitis (odds ratio 3.2 - 4.7, 95 % confidence interval 2.2 - 6.1), gall - stone disease (odds ratio 1.4, 95 % confidence interval 1.2 - 1.7), diabetes mellitus (odds ratio 1.6 - 1.7, 95 % confidence interval 1.4 - 1.9), and obesity (odds ratio 1.6 - 1.7, 95 % confidence interval 0.7 - 3.7) were associated with increased risk of pancreatic cancer, whereas alcoholic abuse, hypertension, and peptic ulcer disease were without association. This means that pancreatic and biliary inflammatory disorders and diseases associated with hyperinsulinemia are associated with increased risk of pancreatic cancer (13).
Families with pancreatic ductal cancer in more generations are termed "familial pancreatic cancer". In a study of the European Registry of Hereditary Pancreatitis and Familial Pancreatic Cancer, EUROPAC, 99 men and 98 women with pancreatic cancer in 81 families median male age was 57 years and median femal age 66. Transmission from the mother gave a significantly survival for males in a group of 10 families also afflicted by gastric cancer. The children with cancer were younger than their parents when they developed cancer in 83 percent of cases (14).
In a Rumanian consecutive series of 106 patients operated on for chronic pancreatitis 3 males with a follow-up of 9,11 and 14 years were found to have pancreatic cancer (15).
The carbohydrate antigen CA 19-9, the 2-3 sialylated Lewis antigen, is a marker for pancreatic carcinoma. It has been shown that in the patients with pancreatic cancer the expression of the Le antigen on erythrocytes could be diminished. In these patients therefore the prevalence of Le(a-b) is increased. In a study of 57 patients with pancreatic diseases and 200 healthy controls 39 percent of the patients were Le(a-b-) which was significantly more than in the controls. However, there was no significant correlation between serum levels of CA 19-9 and the Lewis phenotypes (16).
Diagnosis
Measuring of M2 pyruvate kinase isoenzyme, M2-PK, had a lower sensitivity and specificity in diagnosing pancreatic cancer compared to CA 19-9 in a study of 103 consecutive patients with pancreatic diseases (17).
Surgery
Preoperative CRP was obtained for all patients undergoing resection for pancreatic cancer in Glasgow 1993-2001. Significant positive relationships between elevated CRP (> 10 mg/L) and larger tumour size, greater incidence of vascular invasion, and poorer tissue differentiation. Median survival in the group with CRP < 10 mg/L was 17 months compared to 8 months for the rest (18).
In a study 1994-2002 altogether 120 patients with a pancreatic head resection 119 had a pancreatogastrostomy in an effort to decrease the risk of fistula from the pancreatic remnant. The postoperative complication rate was 33 percent (wound infection 9 %, upper gastrointestinal bleeding 4 %, pancreatic leakage 3 %, and biliary leak 3 %). Hospital mortality rate after pancreatic head resection was 5 percent (19). In a study from one of the biggest Hungarian universities 326 operations were performed for pancreatic and other periampullary tumors 1991-2001. There were 105 resections. The most common complication seen in the resection group was insufficiency of the anastomosis (5 patients). The mortality was 3.9 percent. In the palliative group, transitory bile leakage was the most frequent observed complication (27 patients). Mortality was 4.1 percent. The average survival rate in the radically operated group was 17 months, in the palliative group 7 months and in the explorative group 2 months (20).
In a study of 161 patients undergoing resection for pancreatic cancer status of lymph node 8a was available in 73. Of them 18 percent had metastatic disease. Tumour type did not affect lymph node 8a status. However, involvement of 8a was an independent prognostic factor; those with a negative lymph node had a median survival of 417 days whereas the positives had a median survival of 175 days (21).
In Prague 72 patients 1998-2002 underwent palliative surgery and 13 had a resection for pancreatic cancer. Within the early postoperative course after palliative resection one patient (8 %) died. Early postoperative morbidity was 61 percent. The 1-year survival was 27 percent, 2-year 18 percent. In the palliative by-pass group 4 percent died of complications to the operation. Early postoperative morbidity was 36 percent. The 1-year survival rate was 35 percent, 2-year 17 percent. The authors concluded that bypass procedures and palliative resections for pancreatic cancer have nonsignificant differences concerning morbidity, immediate postoperative mortality and long-term survival (22). In Gdansk, Poland, 159 patients had a resection out of 263 with pancreatic cancer and 41 with periampullary cancer. Hospital mortality rate after resections was 3.8 percent and pancreatic leakage 3.3 percent, but there was only one patient who died due to pancreatic leakage. Long-term survival after a radical resection was 21 months and after palliative resections 11 months. After bypass procedures the long-term survival was 6 months (23).
In a study of 93 patients with R0 resections for pancreatic and other periampullary adenocarcinomas the most important independent negative prognostic variables for disease free survival were age (< 60 years), preoperative pain, tumour positive lymph nodes and perineural invasive growth (24).
Thirty-four patients with R1 or R2 resection for pancreatic cancer were given 50 Gy external radiation combined with 5-FU in a radiotherapy sensitizing dose and were compared to a group of 21 patients that did not receive chemoradiation. Local recurrence was found significant less common in the group that received adjuvant therapy (6 vs 16), whereas the incidence of metastatic disease did not differ. In the treatment group development of both local and distant disease was found in 3 patients versus 13 in the group that received surgery alone, which is also a significant difference. The median survival time did not differ between the two groups, 13 versus 14 months, but two patients are still alive in the treatment group after 10 and 6 years, respectively (25).
In a retrospective study of 281 patients with pancreatic cancer those who received warfarin had a median survival of 11 months (n = 170) as compared to 8 months in those who did not (n = 111). There was no significant difference in the number of thrombotic events or bleeding episodes between both groups, and stage of disease did not play a significant role in the prolongation of life with warfarin (26).
In a Finish study Nordback and co-workers showed with 20 consecutive patients undergoing pylorus-preserving pancreaticoduodenectomy - 18 of the for malignant disease - that 35 percent of the patients developed delayed gastric emptying, DGE. This was than defined as the need of nasogastric tube or recurrent vomiting still on the tenth postoperative day - all patients had a gastrografin examination on day 6. None of the DGE-patients had leakage or obstruction in the duodenojujunal anastomosis. Thirty percent of the patients developed pancreatic or splenic complications: one open and 3 subclinical pancreatic fistulas (a fistula vas defined as 50 ml or more amylase rich fluid in drain output on day 5), and 4 had pancreatitis - three with pancreatic or biliary fistula). Pancreatic or splenic complications occurred in 6/7 of the DGE-patients compared to 1/13 in the non-DGE group, which is a significant difference (27).
In Finland the national hospital discharge database was used to study the annual number of pancreaticoduodenectomies ICD-9 code 6706 and ICD-10 JLC30 related to different type of hospitals and year. After intense advocating of centralising this type of surgery the number of operations has increased by 50 percent 1990 to 2001 but the number of hospitals has decreased by 30 percent. This means that the proportion of the patients being operated on in hospitals doing more than 10 Whipple operations per year has increased from 16 to 44 percent (28).
Basic and experimental studies
In 50 kg pigs a biodegradable stent were used after a hepaticojejunal anastomosis in 15 animals. Their liver clearance as measured with 99mTc dynamic biligraphy was improved compared to 15 animals treated without the stents at 12 and 24 weeks (29).
Polyunsaterated (omega-6) fats such as arachidonic and linoleic acid, are associated with an increased incidence and growth of pancreatic cancer. Cyclooxygenases and 5 - and 12 - lipoxygenases seams to play key roles in promoting pancreatic cancer growth. 15 - lipoxygenase - 1 was now investigated and it was found to be expressed in all normal tissue areas with strong granular cytoplasmic staining in normal ductal cells, pseudoducts and centroacinar cells while no staining was seen in islets, cancer cells and PanIn lesions. This means that 15-LOX-1 is downregulated in pancreatic cancer and already in PanIn lesions, confirming observations in colorectal and esofageal cancers. Its metabolite 13-S-HODE suppresses cell proliferation and induces apoptosis. Indeed, tumour cells may downregulate 15-LOX-1 expression to avoid apoptosis (30). In a study of athymic mice GFP-expressing human pancreatic cancer cells were significantly inhibited by a combination of gemcitabine and leukotriene B4 antagonists both regarding tumour growth and prevention of metastases (31).
One monosaturated fatty acid extracted from sea cucumber abolishes proliferation of pancreatic cancer cells in vitro, induces apoptosis in vitro, markedly inhibits growth of pancreatic cancers xenografted into atymic mice and induces apoptosis in pancreatic cancer in vivo (32).
DPC4 inactivation is one of the more common genetic alterations in pancreatic carcinoma. In 22 patients with pancreatic cancer and 18 with chronic pancreatitis undergoing pancreatic resection a loss of DPC4 expression was seen in 54 percent of cancer cases, which was significantly more than the 22 percent in chronic pancreatitis. No significant association between expression of DCP4 and tumor size, gender or age was observed (33). In 23 pancreatic cancer patients and 21 patients with chronic pancreatitis K - ras mutations was shown in 74 and 43 percent, respectively, whereas p16 mutations (A148T and I49S) was found in 78 and 33 percent, respectively. K - ras and p16 mutations together were observed in 70 and 14 percent, respectively. No significant association between theses mutations and tumor size, gender or age was seen (34).
It has been tested a new DNA - based substance, a 5'3'- terminal modified oligodesoxynucleotide with a novel sequence, "EVITON", showing a significant inhibition of cell proliferation and induction of apoptosis in pancreatic cancer cell lines known to be highly resistant to death receptor mediated apoptosis and in an orthotopic xenograft model (35).
Expression of the cholecystokinin-2/gastrin receptor (CCK-2R) in the murine pancreas leads to a loss of intercellular adhesion and altered cell fate in vivo. This may lead to synergy in conjunction with chemical carcinogenesis promoting dedifferentiation, neoplastic progression and aggresiveness of neoplastic lesions (36).
Pancreatic cancer cell lines were all found to express the receptor for glial cell line-derived neutrophic factor, GDNF. All human pancreatic cancer cell lines were stimulated by GDNF and increased the invasive and adhesive ability for extracellular matrix proteins, and antibodies had the reverse effect (37).
ADAM17/TACE expression is regulated in vivo in pancreatic cancer models by the microenvironment and in vitro by interleukin-6 as well as by the ADAM17/TACE target molecule TNF - alpha in a time - dependent manner (38).
Stimulation with interleukin - 1beta and TNF - alpha, but not with interleukin - 6 significantly enhanced the adhesion of three pancreatic tumour cell lines in a time- and concentration dependent manner. Antibodies against interleukin-1beta and TNF-alpha blocked this enhancement (39).
Both in vitro and experimental findings suggest that the gene SEL1L (suppressor enhancer line) play a regulatory role in pancreatic carcinoma, maybe through the TGF-beta pathway of signal transduction (40).
The metabolic and cachectic effects of elevated microcirculating levels of islet amyloid peptide, IAPP, that are observed in the majority of pancreatic cancer patients were investigated in pair-fed rats. IAPP subcutaneously reduced food intake by 45 percent as well as body weight. Insulin and leptin were decreased. Glucose and lactate were unchanged but free fatty acid and triglycerids were decreased (41). The effects of IAPP may be mediated by the melanocortinnaxis and not by neuropeptide Y, NPY, in the brain (42).
Recently a new pancreatic cell type was charachterized as a nestin - positive stem cell. These nestin-positive cells have extended proliferative capacity in culture and are capable of differentiating into multiple cell types. Nestin RNA expression has been seen in all 10 cell lines and adenocarcinoma specimens tested and the specificity has been confirmed by sequencing the RT-PCR product (43).
In vitro low oxygen levels activate Hypoxia inducible factor-1, HIF-1, which upregulate vascular endothelial growth factor, VEGF. Genistein inhibits hypoxic activation of HIF-1, thereby inhibiting tumour angiogenesis. In an in vitro study in human pancreatic cancer cell lines inhibition of HIF-1 downregulated VEGF production, leading to a decreased microvessel density in tumour xenografts (44). The urokinase receptor, uPAR, is under transcriptional control of HIF-1 and it has further been shown that hypoxic upregulation of uPAR is caused by an actual increase in transcription of uPAR mRNA synthesis. This mechanism also increases the potential of metastatic tumour growth of pancreatic cancer, both in vitro and in vivo in an uPAR dependent manner (45).
The novel anticancer therapeutics 17AAG and geldanamycin target the chaperone heat shock protein 90, HSP90. This protein holds client proteins in an inactive state before release or targeted destruction, thus maintaining a constant pool of protein ready for activation. Combination of geldanamycin and 5FU synergise to kill pancreatic cell lines after 24 hours, but synergy is lost after 48 hours. 5-fluorouracil pre-treatment sensitises cells to geldanamycin, but pretreated cells are significantly more resistant to geldanamycin plus 5FU. Geldanamycin pre-treatment sensitises cells to 5FU and geldanamycin plus 5FU. Semistarvation induces nuclear HSP90 localisation. In vivo nuclear HSP90 is seen in pancreatic cancer but not in normal pancreas, and is associated with larger tumours. This means that in vivo localisation of HSP90 may lead to resistance to apoptotic stimuli, equivalent to prolonged semistarvation (46).
Inhibition of false proliferative signals is one of the major targets for drug development today. Tyrosinase kinase enzymes are considered among the key molecular targets in this respect (47). Receptor tyrosine kinases mediate growth factor signals from the cell surface. Increased expression of receptors of EGF, PDGF and VEGF has often been correlated with a more aggressive behaviour and a poor prognosis (48). EGF receptor subtypes, overexpressed in pancreatic cancer, may be inhibited by a great number of ATP-like, heterocyclic compounds (49). Flow cytometri can be used to show the different effects of kinase inhibitors on the cell cycle and apoptosis (50).
Quality of life
In a study of 14 patient radically operated on for pancreatic cancer were compared to 12 who had a palliative resection, 17 patients who just had a bypass and 11 who only had a laparotomy. All patients completed the questionnaires EORTC QLQ-C30/PAN26 before treatment and 1, 2, 3 and 6 months after the operation. Both parts of the questionnaires had satisfactory psychometric properties for measuring quality of life. Of interest is that not only the radically operated but also patients operated palliatively noticed some improvement in their quality of life, and explorative laparotomy did not aggravate the patient's quality of life. However, in patients with bypass anastomoses some life functions and the symptoms of the disease aggravated (51).
In an English study of quality of life after pancreatic resection 34 patients were studied 2000-2003, 22 of them with ductal pancreatic cancer. They were given a quality-of-live questionnaire at 6 weeks, 3, 6 and 12 months after the operation. Median (+ SD) global quality of life scores were 50 (+ 21) at 6 weeks, 58 (+ 21) at 3 months, 67 (+ 23) at 6 months, and 71 (+ 23) at 12 months, respectively. The median fecal fat were at the same time-points 7.6, 7.0, 6.7, and 5.7 gram per day, and 41 percent of patients had stool fat more than 7 grams per day at 12 months (52).
Endocrine pancreatic tumours
In an Hungarian study of 17 patients with endocrine pancreatic cancers endoscopic ultrasonography, EUS, localised 11 insulinomas but in three cases the examination was true negative. Intraoperative ultrasonography accurately identified 16 cases. The false negative was a 4 mm insulinoma (53).
In the West Midlands' cancer registry there were identified 43 cases of pancreatic neuroendocrine tumours 1996-2000, i.e. an incidence of 1.6 per million per year. The median survival was 760 days in those undergoing resection and 834 days overall. In the North Western cancer registry the incidence was 2.1 per million per year and in the Merseyside and Cheshire registry was it 2.1 (54). However, calculation of these incidences using only cancer registry data in England underestimates rates by at least 50 percent (55).
M2 pyruvate kinase isoenzyme, M2-PK, is expressed by undifferentiated and proliferating tissue. In 24 healthy subjects the tumour M2-PK was abnormally high in 25 percent with a cut-off value of 5.6 U/mL, in 42 percent of 17 patients resectable endocrine pancreatic tumours, in 67 percent of 9 patients with tumors without metastases and in 78 percent of those with distant metastases. Chromogranin A concentration was normal in all healthy subjects, whereas they were abnormally high in 77, 89 and 96 percent, respectively in the groups mentioned above. This means that M2-PK has a sensitivity similar to that of chromogranin A but a significantly lower specificity (56).

Other rare tumours

Gangliocytic paragangliom is a rare tumor usually located in periampullary region. It has an uncertain histogenesis and behavior. In two presented cases the speciments appeared as ulcerated polyps 1.2 and 1.7 cm in diameter, respectively, grey-white, smoth and with nodular aspects. Histological features consisted of epithelioid cells, ganglion-like cells and spindle cells. Immunohistochemistry was positive for cytokeratin, chromogranin, S-100 protein, and NSE. Gastrointestinal bleeding and obstructive jaundice seems to be presenting symptoms (57).
In 22 intraductal papillary - mucinous tumours of the pancreas the immunohistochemical analyses showed negative staining for p53 in all patients. Expression of alpha6-integrin subunit was significantly strong in adenocarcinoma and moderate dysplasia tissues of those tumours (58).

ESPAC-1

Results have in 2001 been published in the Lancet, but there were now up-dated results: 79 percent of the patients (434/550) are dead with a median follow-up of 44 months for the primary endpoint of 2-year survival. Evidence of a chemotherapy benefit exist (median survival 22 months with chemotherapy vs 15 months without) (59).

ESPAC-2

To the study on intraarterial cytostatics in pancreatic cancer 46 patients have been recruited, but only 9 completely followed the whole protocol. This means that this is a difficult protocol both for the patients and the doctors and the recruitment of new patients is relatively slow. However, one must be thankful that someone has the will and the power to continue with the trial as the results are eagerly waited for.

ESPAC-3

Opened in July 2000, now 260 patients included. Effective from June 2003 the steering committee has for ethical reasons decided to stop the arm without treatment - based on results from ESPAC1 and ESPAC3 combined. This is most interesting: adjuvant treatment with cytostatics is better than no treatment also for pancreatic cancer (as for breast cancer and colon cancer Duke C). It is also interesting that the combined results will be analysed also in the future. Obviously all pancreatic surgeons resecting pancreatic cancer are obliged to give adjuvant treatment now on a routibe basis, or - better - to include resected patients in ESPAC3. It pays off!

EUROPAC

The subgroup confined to studies around hereditary pancreatitis and hereditary issues in pancreatic cancer plans to start several studies. One of them will be a randomized study on the influence of antioxidant/magnesium, with pain and use of analgesics as the end points. The study will be lead by professor M Lerch, Greifswald, and J Neoptolemos, Liverpool.

There are also studies done - and in progress - on the genetic issues. First published in 1996 and 1997 BRCA2 mutations have been found in 5-7 percent of "sporadic cancer". CDKN2a (p16) has been reported to be mutated in patients with pancreatic cancer and malignant melanoma. In a study BRCA2 was found in 12-19 percent of patients with familial pancreatic cancer. CDKN2a on the other hand was of no importance if not melanoma was also present. The "major" gene mutation is yet to be found in pancreatic cancer. A putative locus may be at 4q32-34 according to I Ellis of Liverpool, which means that there are 133 characterized putative (defined) genes. There are ongoing studies to further analysing this.
Genetic counselling for families with inherited pancreatic disease was approached by I Ellis (Liverpool). This is very interesting from a medical point of view, but contain some difficult ethical issues that is only partially discussed so far. Today it is perfectly possible to find mutations on a routine basis on fertilized human eggs on the 6-8 cells stadium, but more difficult to decide what to do with these results.
In November 2003, 7-9, the 4th International Symposium in Inherited Diseases of the pancreas will be held in Chicago.

Personal summary

This was my 25th consecutive EPC-meeting (the first was in Copenhagen 1979) so I think that I have some possibility to make comparisons, and I may state that the 2003 meeting in Liverpool was among the top EPC-meetings of the last 25 years, no doubt. My high-lights were:
- that it is now proven by ESPAC that adjuvant treatment is of value after radically intended pancreatoduodenectomy for pancreatic cancer,
- that the well-done study by gastroenterologist from Magdeburg showed that stenting in chronic pancreatitis should be avoided in most cases,
- the many good abstracts from Eastern Europe,
- the Neoptolemos' garden party.

Up-coming meetings

Next year at about the same time we will visit Padua in Italy (or rather just north of the town) June 23-26 (www.padova2004.org). As it is organised by Professor Sergio Pedrazzoli we can expect well organised meeting filled with science and Italian culture at its very best. The year after that we go to Newcastle and then it is Tampere in Finland. On the list further on may be Brno/Prague, Lodz, Stockholm and others. The future seems thrilling!
Of special interest is also the meeting following this year's meeting of the American Pancreatic Association: "the 4th International Symposium on inherited diseases of the pancreas" in Chicago November 7-9. For further information look at www.pancreas.org or www.american-pancreatic-association.org.
The next meeting of the International Association of Pancreatology will be held together with the yearly meeting of the Japan Pancreas Society in Sendai, Japan, July 11-14, 2004.

Address for correspondence:

Ake Andrén-Sandberg, MD, PhD
Department of Gastrointestinal Surgery
Sentraljukehuset i Rogaland
4069 Stavanger
Norway