ročník 12,2004 č.3
Zprávy z kongresů

Zpráva z kongresu

Ve dnech 23.- 26.6. 2004 jsme se zúčastnili v pořadí již XXXVI. Evropského pankreatického klubu (EPC), který se konal v lázeňském letovisku XXY poblíž italské Padovy. Sympozium zorganizoval profesor S. Pedrazzoli se spolupracovníky (M. Plebani, C. Pasquali, C. Sperti, D. Basso, P. Fogar a F. Navagali) z Istituto di Semeiotica Chiurgica of Padova. Během dvou a pů dne účastníci vyslechli 40 sdělení, 7 "State of the art lectures". 4 sympozia, program mladých výzkumníků a shlédli 158 posterů. Nejvíce posterů bylo přijato z Polska (33) a Německa (24).
Kromě toho formou satelitních sympozií byli informováni o probíhajících evropských studií: EUROPAC, ESPAC and Pan-Cro-INF, Pro-INF AISP.
Celkový počet registrovaných dosáhl počtu 320. Z České republiky sympozium navštívilo 5 účastníků (Dítě, P., Šálek, C., Strnad, R., Froněk, J., Ryska, M.) s 3 postery.
Vzhledem k tomu, že sympozium bylo po obsahové stránce velmi zajímavé, uvádíme ve zkrácené podobě zprávu z kongresu, kterou vypracoval °Ake Andrén-Sandberg (Stavanger, Norsko).
Příští rok se uskuteční XXXVII. Kongres EPC v Grazu ve dnech 6 - 8.července.

Froněk, J., Ryska, M. (se svolením autora)

Chronic pancreatitis

Maybe there was no "break through" regarding chronic pancreatitis, but the knowledge is gradually being deepened.

Stellate cells
Increasing evidence suggests that pancreatic stellate cells, PSCs, are the major mediators of pancreatic fibrogenesis and the predominant cell type for matrix synthesis in the pancreas. The pancreatic stellate cells have been shown to be morphologically similar to hepatic stellate cells, the principal effector cells in liver fibrosis. The pancreatic stellate cells are situated at the base of pancreatic acini, with their extended cytoplasmatic processes encircling the basal aspects of acinar cells. In the healthy pancreas, the stellate cells exist in a quiescent state and can be identified by the presence of vitamin A-containing lipid droplets in the cytoplasm and by immunostaining for cytoskeletal proteins such as desmin and glial fibrillary acidic protein, GFAP. Electron microscopy reveals a prominent rough endoplasmatic reticulum, collagen fibrils, and vacuoles (lipid droplets) surrounding a central nucleus. Max Bachem, Ulm, now discussed the role of the stellate cells in fibrinogenesis. In normal cells theses cells are quiet, but if they are activated in acinar cells injury (in contact with collagen, in oxidatide stress etc) they strongly produce extracellular matrix, they can move, proliferate, contract and they can attract other inflammatory cells. It is possible to see them outgrow in vitro from isolated acinini. There are good markers for stellate cells versus for example fibroblasts and fibrocytes. Ethanol by itself and actealdehyde can also directly induce the stellate cells. Significant lipid-peroxidation is associated with activated stellate cells. There is also a paracrine induction of the stellate cells.
The strong desmoplasia in human pancreatic cancer was seen also ex vivo. Stimulation of PSCs increase motility by carcinoma cell supernatants in cell lines and the PSC moves against the cancer cells. Also a stimulation by PEGF induced proliferation of the stellate cells. This means that concerning the pancreas stimulation of stellate cells induce:
o increased motility
o proliferation
o synthesis of extracellular matrix.

In vivo when animals had tumors on both sides of the body there were bigger tumors in combination with PSC (volume, weight etc) and this could not only be due to extracellulary matrix, as the cellularity also was also increased.

Surgery

Markus Büchler, Heidelberg, spoke on the role of surgery in chronic pancreatitis. He started by once again giving light to Warshaw's article in Gastroenterology (1984; 86: 987-9): "Patients, patience, and the impatient surgeon." Büchler underlined that he exact date for surgery is important: not too early, not too late. The indications for surgery in Heidelberg are:
o duodenal obstruction
o common bile duct obstruction
o pancreatic main duct obstruction
o vascular obstruction
o symptomatic pseudocysts
o pain
o suspicion for malignancy
o (inflammatory mass in the pancreatic head).

The goal for surgery in chronic pancreatitis should be:
o organ preserving operations
o remove complications of chronic pancreatitis
o preserve endocrine function
o good quality of life
o good short term surgical results (no complications)
o good long term surgical results (freedom from pain).

He then turned to the randomized studies, and discussed several of them in detail.

First author, year of publication Operative procedures Number of patients Pain relief (%) New onset of diabetes (%) Follow-up (months)
Klempa, 1995 Whipple
Beger
21
22
70
100
29
9
36-60
Büchler, 1995 Beger
PPPD
20
20
75
40
n.r.
n.r.
6
Izbicki, 1995 Beger
Frey
20
22
95
89
no change
no change
18
Izbicki, 1998 Frey
PPPD
31
30
90
87
n.r.
n.r.
24
Evans, 1997 Whipple
Beger
15
18
70
70
30
60
60
Jimenez, 2000 PPPD
Whipple
39
33
60
70
10
12
35
30
Makowiec, 2004 PPPD
Beger
43
43
equal
equal
equal
equal
32
32
Tab. 1 - Clinical effectiveness of pancreatic resection for chronic pancreatitis - randomized studies n.r. = not reported

There is now also a randomised study from Leipzig of Whipple versus Beger operation, but these results are not yet totally available. According to Büchler evidence based surgery now show that the Beger operation is better than Whipple's for chronic pancreatitis. As the Beger operation has the same results as the Frey operation this means that we can omit the lateral operation, which is now done in a modification - "the Bernese operation". This has been described of Gloor et al in Dig Surg 2001, by Farkas et al in Br J Surg and by Andersen et al in Arch Surg. He also pointed out that it is today 25 years after Beger's first duodenum-preserving pancreatic head resection.
He also discussed the Dite randomized study of endoscopy versus surgery for pain in chronic pancreatitis published in Endscopy (see below). The initial results were similar, whereas after 5 years surgery is better (endotherapy can be offered as a first-line treatment). In the material from Magdeburg, recently published in Gastroenterology it was shown that bile duct obstruction was well treated by endoscopic stent in only 8 percent of patients if there were calcifications but in 55 percent of patients in non-calcification pancreatitis.
In Heidelberg it had now been done 841 pancreatic operations during 2 years, which at the moment means 45 operations per months: 70 percent of them resections. The reoperation rate was 6 percent, the mortality 0,7 percent and the fistula rate 4 percent. For chronic alcoholic pancreatitis it was performed 192 operations. Morbidity was 20 percent (medical morbidity 7 %), and no mortality.
In a randomized study of endoscopy (n=35) and surgery (n=36) the endoscopists performed endoscopic papillosphincterotomy in 48 percent of their patients, papillosphincterotomy and pancreatic duct drainage in 39 percent and in 13 percent stents were inserted in both ducts after papillotomy. The surgeons performed duodenum-preserving pancreatic head resection in 43 percent of cases, pancreatoduodectomy in 33 percent, distal pancreatectomy in 7 percent and Partington-Rochelle anastomosis in 20 percent.Total complication rate was in both methods similar, 7 and 11 percent, respectively, but the surgical complications were more severe. After 5 years of follow-up complete absence of pain and increase of body weight were significantly better in the group treated surgically (1).

Quality of life

In a study of 43 patients with chronic pancreatitis the mean health-related quality of life as assessed by the Short Form-36 questionnaire the scores were lower compared to a control group in all domains. Statistical differences were observed in general health perception, physical functioning, role-physical and vitality. There was a negative correlation between scores and pain in all domains. The worst quality of life was scored by retired and disabled patients, as well as in unemployed persons in almost all SF-36 domains (2).
In a study of 104 consecutive patients operated on for chronic pancreatitis a pancreatoduodenectomy was chosen when a head mass was present in the pancreatic head and pancreatic cancer could not be ruled out (n=48); otherwise a duodenum-preserving pancreatic head resection was performed. Quality of life before and after the operation with EORTC QLQ-C30. The total pain score significantly decreased after both types of surgical procedures. During the follow-up period median global quality of life improved by 30 percent in the duodenum-preserved group and by 23 percent in the group operated on with a pancreatoduodenectomy. Postoperatively the figures for morbidity and mortality after the bigger operation, but the differences were not significant (3).
Using EORTC QLQ-30 and PAN26 82 patients with chronic pancreatitis were evaluated, 29 following ESWL and 53 after endoscopic stenting. Average follow-up time was 1.2 years. Increase in body weight was 1.2 kg and return to working capacity was 82 percent. There was a significant improvement in quality of life scores for pain, general physical state, daily activity, digestive function and fatigue, with no significant differences after different treatments (4).
In a multicentre study of 301 patients with chronic pancreatitis due to alcohol abuse EORTC QLQ-30 and PAN26 was tested. Apart from the jaundice scale, all component scales met accepted standards for internal consistency. The convergent validity of the two pain scales from C30 and PAN26 is high (0.79). However, as it not is 1, it is suggested that these scales are measuring some different aspects of pain and pain perception (5).

Pancreatic Cancer

Epidemiology of pancreatic cancer
In the Utah Population Database 55 large kindreds at increased risk for pancreatic cancer was identified, each kindred had at least five cases. Among the kindreds 219 pancreatic cancers were found. It was found an increased risk for several cancers:
Type of cancer Expected Observed Observed/Expected ratio (95 % confidence interval)
Prostate 629 913 2.9 (2.7-3.1)
Breast 378 657 3.5 (3.2-3.8)
Colon 167 371 2.2 (5.4-6.4)
Lung 162 267 1.6 (1.4-1.8)
Melanoma 69 285 4.2 (3.7-4.7)

There was also an increased risk for bladder cancer, lymphoma, ovarian, thyroid, brain, esophageal, and cervical cancer (6).
Data collected over 15 years (1985-2000) from the West-Midlands in England (a population of 5.5 million) from the cancer registry and the geographical information system within their unit, including four variables as proxy indicators of socioeconomic status. There was marginal overall decrease in the incidence of pancreatic cancer in men (12 %) and a stable overall incidence in women over 15 years. This, however, masked the preferential increase in the most affluent men and women compared with their more deprived counterparts. Thus, the increase in the affluent men rose by 89 percent with an increase in age standardised rates of 8.9 per 100000 in the mid 80s to 12.4 by 2000. The corresponding figures for the most affluent women were 41 percent, 5.3 and 7.5, respectively. The affluent groups had a 3-5 percent survival advantage over their deprived counterparts but the differences were not significant at the 95 percent confidence interval level. If all men and women with pancreatic cancer had survival rates equivalent to those experienced by the most affluent men and women, 55 deaths would have been avoided in each 5 year period.

Diagnosis Years MenWomen
Most affluent incidence rate Most deprived incidence rate Most affluent incidence rate Most deprived incidence rate
1985-1987 8.92 12.89 5.34 9.61
1992-1994 9.48 12.04 7.51 8.01
1998-2000 12.40 11.91 7.52 6.51

This implies that a modest overall decrease in incidence of pancreatic cancer over 15 years masks the preferential increase noted in the affluent population and decrease in the deprived. The reason for the temporal changes remains unclear (7).
Fourteen new cases with metastases to the pancreas were reported, in four of them pancreas was the only site of metastasis. The interval between primary treatment for renal cancer and occurrence of pancreatic metastases averaged 11 years. Metastases were multifocal in 36 percent of patients (8).

Molecular biology of pancreatic cancer
In a State-of-the-art lecture on apoptosis signalling in pancreatic cancer cells - suitable targets for therapy, Holger Kalthoff from Kiel, started by showing the two different apoptotic ways (type I and type II): death receptor (extrinsic) and mitochondrial (intrinsic) pathways. It should also be understood that there is probably not one regulating step, but instead several concerted steps, from the caspaces and downstream). Today it is well described several experimental models where one gene has been altered i some way, which of course give possibilities to study exactly this factor, but it might not is applicable to the concerted "normal" situation. From the theories it can now be better understood the chemotherapy resistance - and maybe better ways to circumvent this problem is possible. For example, gemcitabine induces the mitochondrial pathway of apoptosis.

Diagnosis of pancreatic cancer
In a consecutive series of 75 patients with pancreatic cancer treated the last 12 months the most frequent first symptom was dyspepsia (n=27), followed by pain (n=24) and jaundice (n=22). Mean time from the onset of the first symptom to presentation was 33 I 28 days. The mean number of procedures (CT, ultrasonography, ERCP, EUS and MRI) was 4.4 I 1.5. The most frequent algorithm of imaging methods was ultrasound - CT - ERCP. Mean time of the diagnostic process to specific treatment was 3 I 3 months (36).
In a study of 112 patients 2000-2003 (37) with pancreatic head tumors 70 percent had a malignant tumor. The patients were investigated with 18F-fluro-2-deoxy-D-glucose (FDG) positron emission tomography (PET) preoperatively:

  FDP-PET CT
Sensitivity (%) 73 89
Specificity (%) 60 65
Positive predictive value (%) 80 90
Negative predictive value (%) 49 61
The sensitivity and specificity for detecting small volume metastatic disease was:

  FDP-PET CT
Sensitivity (%) 22 20
Specificity (%) 91 94

Staging of pancreatic cancer In 43 consecutive patients who underwent a pancreatoduodenectomy for pancreatic adenocarcinoma endoscopic ultrasonography accurately predicted tumor size in 9 of 25 patients that had preoperative chemoradiation and 14/17 that only had an operation without neoadjuvant treatment. Preoperative endoscopic ultrasound overestimated tumor size in patients with pancreatic adenocarcinoma after neoadjuvant chemoradiation (38).
In a study of 15 consecutive patients with pancreatic cancer endoscopic ultrasonography showed a significantly higher accuracy than helical CT for locoregional staging (39).
In 27 patients operated on for pancreatic adenocarcinoma 11 hade radical operations. Comparison of colour Doppler sonograms and operative findings show that sensitivity of colour Doppler sonography for detection of vascular involvement was 81 percent, whereas its specificity was 83 percent. The sensitivity of endosonography was 91 percent and its specificity 86 percent (40).
From a prospectively collected database 1996-2003, 66 patients with pancreatic cancer were investigated. The level of hyperbilirubinemia was not different in patients with high or low CA 19-9 in serum. CA 19-9 (cut off value 150 U/ml) as a predictor of resectability, evaluated by laparoscopy showed:

  Straight CA 19-9 CA 19-9 adjusted for hyperbilirubinemia
Sensitivity (%) 36 63
Specificity (%) 100 55
Positive predicted value (%) 100 88
Negative predictive value (%) 24 22

It was predicted that preoperative use of CA 19-9 from a theoretically point of view can reduce the need for laparoscopy in up to 30 percent of the patients (41).

Preoperative drainage in pancreatic cancer
In a prospective study of 345 patients 2000-2003 with pancreatic cancer, 67 percent presented with jaundice, of which 146 were expected to be resectable and underwent biliary drainage. Complications to the drainage occurred in 14 percent of these patients: acute pancreatitis 13, cholangitis 3, empyema 1, pneumonia 1, intraabdominal collection 3, peritoneal seeding 1, death 2. In 4 percent resection was abandoned or modified because of complications of biliary drainage. The authors concluded that preoperative biliary drainage causes an unacceptable level of complications (42).
The authors also investigated 111 patients with pancreatic cancer and found a median increase in bilirubin levels of 13 micromol/L/day in the jaundiced patients. The rate of increase did not vary between resectable and not-resectable tumors. They also predicted the number of days before levels were above 300 micromol/L, which arbitrarily was mentioned as a limit for biliary drainage (43):

Presenting with 50 micromol/L 23 days to reach 300 micromol/L
150 14 days
250 4 days

Neoadjuvant treatment
Preoperative chemoradiation was given to 61 patients with resectable pancreatic cancer. However, 35 percent of the patients were not resected due to tumor progression. Major histopathological responses were noted in 9 patients, including three complete responses with tumors in the head of the pancreas (49 of 61 patients). Local control rate was similar in patients with and without major pathologic response. Survival in patients with major response was significantly higher than in non-responding patients or with minor response (44).
In 45 patients with locally advanced pancreatic cancer without distant metastases radiotherapy, 50 Gy in two courses, concomitant with intravenous 5-FU 25 mg/kg/24 hours continuously on the first 4 days of each treatment course. They were compared to 38 patients without any such treatment. Eleven of 38 evaluable patients (29 %) showed a partial response and 29 percent had a stable disease. A second look operation was performed in 8 out of 11 patients with a radiological response, and in three patients the tumor could be resected. The median survival in the radiotherapy group was 10 months which compares favorable to an 8 months median survival when no treatment was given (45).

Laparoscopic pancreatic cancer surgery
Laureano Fernández-Cruz, Barcelona, Spain, gave a State-of-the-art lecture: "Laparoscopic surgery in pancreatic diseases." He stated that percutaneous catheters, endoscopic procedures and videoscopic surgery now are included in the pancreatic surgeons's armamentarium. Today laparoscopic methods are used in benign and malignant disorders of the pancreas:

o Benign disorders Chronic pancreatitis
Acute pancreatitis
o Benign-apperaring tumors Neuroendocrine tumors
Cystic tumors
o Malignant tumors Diagnoses and staging
Palliation of complications

For necrosectomy in severe acute pancreatitis there are today several options:

o conventional treatment:
necrosectomy followed by sump
drainageand re-laparotomy as needed
mortality 6-42%
o lavage method (Beger's technic) mortality 10-20%
o open management:
necrosectomy followed by scheduled
re-laparotomy though a marsupialized
open abdomen
mortality 15-30%
o open retroperitoneal approach

Fernández-Cruz advocated videoscopic-assisted retroperitoneal debridement techniques (VARD) in the management of infected pancreatic necrosis and fluid collections even though the method is only in the early stages of the development. Concepts underlying the development of VARD are:
o the effective debridement of particulate material not drained through catheter
o creating a large continuous collections by disrupting located collections, which a catheter can then drain more effectively
o necrosectomy performed in a minimal invasive fashion by using small incisions
o definitive treatment of the patient without the need for a crossover to laparotomy

Fernandez-Cruz discussed the pseudocysts from a surgical point of view. He stated that by open internal drainage there is complication in about one case in four. External drainage is not recommended anymore as the fistulas may persist for a very long time and be very troublesome for the patients. In most centres the endoscopic approach can be used in about 85 percent of cases, leaving 15 percent for surgery. Regarding the surgical approach the principles are the same for laparoscopic and open surgery, and as the results and complication risk is about the same he preferred the laparoscopic technique.
He reported his results of anterior gastropseudocystostomy in 5 cases of acute pancreatitis and 8 cases of chronic pancreatitis. In another 2 cases with chronic pancreatitis he performed an intraluminal anastomosis and in another two a distal pancreatectomy. He had had two complication requiring reoperation. However, the median follow-up was only one year. He always started the laparoscopic operation with an injection of contrast to rule out communication between the pseudocyst and the ductal system. One advantage of a surgical approach is the possibility to take out necrotic material - especially in acute pancreatitis. He also gave antibiotics before the procedure and took bacterial cultures peroperatively.
For pancreatic pseudocysts there are now larger series showing that the laparoscopic approach is worthwhile:

  Mortality Recurrence Morbidity
External drainage 6 % 22 % 
Internal drainage 0-10 % 0-17 % 25 %
Percutaneous aspiration 3 % 54 % 16 %

Fernández-Cruz recommended laparoscopic management in patients with acute pancreatic pseudocysts for internal drainage at the appropriate time in the illness as it allows exploration of the cyst cavity, debridement and establishment of dependent drainage. However, reoperation may be necessary for the presence of infected necrosis not resolved.
Other indications for laparoscopic surgery are:

o Benign appearing tumors Cystic neoplasms
Inflammatory tumors
Neuroendocrine tumors
o Malignant tumors Exocrine cancers
Neuroendocrine

Operations that may be done laparoscopically are pancreaticoduodenectomy, tumor enucleation and distal pancreatectomy. Alfred Cuschieri stated "The probable benefit of laparoscopic surgery over open surgery depends on the ratio of access trauma to procedural trauma."

Pancreatoduodenectomy
Of 172 patients undergoing pancreatoduodenectomy 1994-2003 16 patients (9 %) were reoperated for postoperative hemorrhage. This occurred in 23 percent of irradiated patients and 6 percent in non irradiated patients. Pancreatic leak occurred in 56 percent of the bleeding patients. Sentinel bleeding was noted in half of the patients with a mean delay of 10 days after the operation. Overall mortality after hemorrhage was 56 percent. Mortality rates of axial (hepatic artery and mesenteric vessels) or lateral (pancreatic remnant and splenic vessels) bleeding were 88 and 25 percent, respectively. Completion of pancreatectomy was achieved in 75 percent without rebleeding (46).
In a non-randomised series it was compared 27 patients with Billroth-I reconstructions after pancreatoduodenectomy with 27 patients reconstructed according to Billroth-II. The differences were small, but maybe there were some advantages for B-II with respect to postoperative oral ingestion, and avoiding bypass operation at the time of recurrence (47).
In a series of 138 pancreatoduodenectomies performed 1994-2003 there were 29 duodenum-preserving and 109 standard operations for a resection rate of 52 percent but a resection rate with R0-resections of 22 percent. Hospital mortality was 4.2 percent: circulatory and respiratory insufficiency (n=3), cerebral circulation insufficiency (1), intraoperative bleeding (n=1), mesenteric artery prothesis infection (1), myocardial infarction (1) and pancreatic fistula and sepsis (1). Postoperative complications were noticed in 25 percent of the patients. In 17 percent of patients there was a delayed gastric emptying. The average survival rate after an R0 resection was 25 months compared with 15 months after R1. Long-time survival was 8 months after bypass operation (48).
For 346 patients who underwent pancreatectomy for invasive adenocarcinoma of the head of the pancreas 1978--2002 venous invasion, tumors less than 2 cm in largest diameter, extrapancreatic nerve plexus invasion, lymph node metastases, and dissected peripancreatic tissue margin were independent prognostic factors (49).
For 300 patients operated with pancreatoduodenectomy in three multicentre randomized studies the mortality was 9 percent. The two independent risk factors were age > 70 years and extended resection. The morbidity rate was 31 percent. Two risk factors were defined: size of main pancreatic duct inferior to 3 mm and extended resection. The rate of pancreatic fistula was 17 percent (50).

Diabetes in pancreatic diseases
The association between pancreatic cancer and abnormal glucose tolerance has been known for more than a century and was actually reported as early as 1893 by Marallie. Moreover, patients with pancreatic adenocarcinoma have a high incidence of overt diabetes, i.e. the prevalence of treated diabetes in pancreatic carcinoma is significantly greater than controls. Compared to patients with other gastrointestinal cancers and controls the patients with pancreatic cancer are hyperglycemic and hyperinsulinemic, i.e. they have an impaired glucose tolerance. There are observations that an abnormal glucose tolerance or frank diabetes mellitus develop in up to 80 percent of the patients shortly before the clinical manifestation or is diagnosed at the first clinical admission. Johan Permert, Stockolm, Lucio Gullo, Bologna, and Charlotte Erlanson-Albertson, Lund, chaired a symposium on diabetes pancreatic diseases.
Physiopathological aspect of pancreatic diabetes was covered by A Avogaro, Padova. He pointed out that one of the physiologic problem for the resected patients are not only the lack of glucagone but also the lack of epinephrine. Moreover, the patients have lower insulin sensitivity, i.e. an insulin resistance. The low glucagon secretion gives a low glucagogenesis and a higher blood lactate, 1.5 times higher than for other individuals of the same age. They also have higher amino acid concentrations. Pancreatic polypeptide inhibits the liver glucose production by insulin, and may therefore also be of importance. The lack of pancreatic polypeptide gives a blunted response to insulin. Low glucose absorption Increase insulin insensitivity further with a reduced phasic glucose output.
There has been a lot of interest in diabetes in pancreatic cancer, according to Daniela Basso, Padova, over the years. If studied close enough only one tenth of all pancreatic cancer patients have a normal glucose metabolism. Also small tumors may lead to glucose abnormalities. Of 822 pancreatic cancers < 2 cm 8 percent had diabetes. There is a clear association of recent onset diabetes to pancreatic cancer, but diabetes may also predispose to pancreatic cancer. However, there is no correlation to stage or size of the tumor.
It is interesting to find that pancreatic cancer cells stimulate amylin but not insulin. There is defined a pancreatic cancer-associated diabetogenic factor - a low molecular peptide that is being purified at the moment.
Diabetes in chronic pancreatitis was covered by Lucio Gullo, Bologna. In Ammans report from 1984 75 percent of the 145 patients had diabetes. In a new study by Gullo (in press) 44 percent of 135 patients had diabetes after a mean of 8 years. It is more frequent in alcoholic chronic pancreatitis, after surgery and by time. In multivariate analysis calcification was a significant risk for diabetes but not for insulin-dependent diabetes.
In chronic pancreatitis there is a destruction of the islands of Langerhans. However, there is also a reduced stimulus on endocrine pancreatic cells. The insulin resistance is localized to peripheral tissue rather than liver, but in these patients insulin resistance is frequently also associated with overweight, alcohol etc. There is no difference in frequency of retinopathy compared with diabetes type I and the same is true for peripheral aterosclerosis.
Engraftment and islet cell transplantation was discussed by Leif Gunnar Jansson, Uppsala, Sweden. The problem is the vascularization of the islets after the transplantation, i.e. engraftments. The islets need 5-6 ml blood/min/gram tissue, i.e. 5 times higher than exocrine tissue. When insulin is secreted there is an increased blood flow, and all conditions increasing blood flow might stimulate the vascularization. Today many follows the Edmonton protocol which means that 12 000 islets per kg body weight are transplanted. It is acceptable to do multiple transplants, and after these procedures 80 percent is functioning after 1 year.
Revascularization a rapid process, initiated within 3-7 days and are finished after 7-14 days. Two different terms are used: Vasculogenesis, which means formation of vessels from angioblasts and angiogenesis which means sprouting from existing vessels. It is also interesting to find that endothelials cells are not all from host, but also some from donor, after transplantation. This vascularization is compared by cyclosporine and exocrine pancreatic juice. After transplantation:
o blood flow decreases
o capillary blood pressure increases
o pO2 pressure decreases
o vascular density decreases, i.e. angiogenesis impaired
o anaerobic mertabolism with pH decreasing
o lymphatic capillaries (normally none) transport high concentrations of insulin.

Pancreas transplantation was discussed by Franco Mosca, Pisa, Italy. There are several surgical risks that can be lessened by:
o donor selection
o graft selection
o bench-work transplantation
o pre- and postoperative good vascularization.

Today the portal-enteric techniques are most common. However, there was still venous thrombosis in 8 percent of 112 patients; arterial thrombosis in 1 percent and relaparotomy in 14 percent. Survival of the patients is after 36 months 98 percent, the pancreas transplants are functioning in 90 percent of cases, similar as after one year. Vascular thrombosis remains the Achilles heal of pancreatic transplantation. Ultrasonography-Doppler is a good way to follow these patients up. Better results may be achieved by:
o meticolous surgical techniques
o implementation of prophylactiuc antithrombotic measures
o strict post-operative surveillance.

Perspectives of artificial pancreas were given by Wolfgang Kerner, Ulm, Germany. He discussed type I diabetes which might be a T-cell mediated beta-cell destruction. The alternatives are then:
o intensified conventional insulin therapies
o insulin pump therapy
o pancreas-kidney transplantation
o experimental
- islet transplantation
- artificial pancreas

Artificial pancreas started already with C Weller in 1960 with a continuous bedside glucose monitoring with a modified autoanalyser. In the year 2000 industry started delivering machines for measurement in subcutaneous tissues
o glucos sensor
- function at least 1 year
- needle sensors commercial available in 2004
- transdermal may soon be available
o insulin pump
o algorithm for insulin delivered control
- infusion into peritoneum,
portal vein or inferior vena cava.

The glucose sensors are at the moment the weakest point together with the loop between implanted sensor and pump.
In another report it was identified a peptide at m/z of 5005 that significantly correlated with the presence of pancreatic cancer-associated diabetes. A negative association found between pancreatic cancer-associated diabetes and some peptides, suggests that diabetes mellitus might be consequent not only to the diabetogenic effects of new tumor product(s), but also to lacking production of some anti-diabetogenic substances (51).

Palliative treatment
In 144 consecutive patients with unresectable pancreatic cancer 199-2003 all patients with surgical bypass were excluded from the study. A self-expandable metallic enteral stent was employed if duodenal obstruction was symptomatic, confirmed by endoscopy and barium studies. The actuarial median survival was 8 months in those with metastatic disease and 12 months in those with locally advanced pancreatic cancer. Biliary stenosis occurred in 89 percent of patients with cancer of the pancreatic head and duodenal stenosis in 19 percent, respectively. When a self-expandable metallic stent was first inserted (n=59), a single stent was sufficient in 69 percent of cases (median duration of patency 6 months). Duodenal stenting was successful in 96 percent of patients and 92 percent of them required only one stent (median duration of patency 6 months). Twenty-one patients (15 %) had either concomitant or metachronous development of biliary and duodenal obstruction - combined stenting was successful in 89 percent of cases. No major complication or death occurred related to endoscopic treatment (52).

Cachexia in pancreatic cancer
Despite long and widespread interest in the topic, there is no standard definition for cachexia, which is derived from the Greek words kakos meaning bad and hexis meaning condition. Cachexia is one of the most distressing complications of patients suffering from cancer and also one of the most common symptoms of advanced cancer, least of all in pancreatic cancer. Anorexia is the result of a failure of usual appetite signals whereas cachexia is the debilitating state of involuntary weight loss. Cancer causes disengagement of the normal balance of energy intake to caloric expenditure in the face of increasing catabolism. Yet weight loss does not always correlate with anorexia, and the pathophysiology may be different. It usually consists of a combination of anorexia, tissue wasting, protein calorie malnutrition and loss of compensatory increase in feeding. This state is sometimes referred to as "the cancer anorexia- -cachexia syndrome". It is said to be present at the time of diagnosis in 15-40 percent of "ordinary" cancer patients but is more common in pancreatic cancer patients. Weight loss is common in patients with pancreatic cancer in all clinical stages of the disease and anorexia and cachexia are present in the majority of patients with advanced-stage cancer. The spectrum of clinical presentation ranges from predominant anorexia to predominant sarcopenia (muscle loss). Weight loss is not accounted for by diminished nutritional intake alone in that both symptoms occur in most patients. Clinically, the two main symptoms may be inseparable and their effects difficult to measure. Anorexia may consist of appetite loss, satiety, combined satiety and appetite loss, or altered food preferences. Satiety may occur during meals, indicating a lack of gastric accomodation, or afterwards, as a result of gastroparesis and delayed antropyloric transit. The difference may be seen clinically as premature discontinuation of feeding, due to satiety from poor gastric accomodation or postprandial bloating or nausea from antropyloric motility delay.
Stacey Wyke, a co-worker of Tisdale, Birmingham, gave a state-of-the-art lecture on physiopathology of cachexia in pancreatic cancer, together with symptoms like emaciation, debilitation, anemia and malnutrition. Clinically it is seen a massive loss of body weight with extensive breakdown of both fat and skeletal muscle. Death is inevitable at 30 percent weight loss, irrespectively of the cause. The patients and the relatives often ask for TPN, but it has repeatedly been shown that this increase only fat and water content of the body. In pancreatic cancer 85 percent of patients have lost weight on a cachexia level (15 %) at diagnosis. Cachexia not only is responsible for much of the symptomatology of pancreatic cancer but also reduces the response to drug therapy. Much interest has been given to associated changes in cytokines, such as:
o TNF-alpha
o interleukin 1 and 6
o interferon-gamma
o leukemia inhibiting factors.

Urine from cachexia patient contains a factor inducing cachexia in patients, and in all types of cancer patients. The factor is rather well characterized and is usually called proteolysis-inducing factor, PIF. It is a sulphated glycoprotein that acts directly to stimulate muscle protein breakdown, through inducing an increased expression of the ubiquitin-proteasome proteolytic pathway. This pathway is considered to be the most important in intracellular protein degradation, and expression is elevated in skeletal muscle of pancreatic cancer patients with a weight loss greater than ten percent. Administration of PIF to normal mice produces about 10 percent weight loss in 24 hours araising entirely through the loss of skeletal muscle mass. PIF not only increases protein degradation in skeletal muscle, but also inhibits protein synthesis (by 50 %). PIF is ATP-dependent.
The catabolic action of PIF is attenuated by the polyunsaterated fatty acid eicosapentaenoic acid, EPA, that is present in fish. EPA prevents muscle breakdown by interfering with key regulatory steps leading to increased gene expression of proteasome subunits and ubiquitin-conjugating enzymes. One of the most important steps inhibited by EPA is the nuclear accumulation of the transcription factor Nuclear Factor-kB, NFkB. However, EPA has no effect on the inhibition of protein synthesis by PIF. In clinical studies 0.5 mg/kg EPA blocked protein degradation with weight change after EPA preparation also in advanced pancreatic cancer. When combined with a high energy, high protein, nutritional supplement weight gain was seen, and this was entirely due to an increase in lean body mass. The supplement well tolerated, had no major side effects and increased the quality of life (53).
It was in another study showed that interleukin-6 is elevated in the serum and tissue of pancreatic cancer patients with cachexia but not in non-cachectic patients. Mononuclear cells of patients with cachexia are sensitized and can be triggered by an interleukin-6 positive cancer cell to produce high amounts of interleukin-6 (54)¨.

Intraductal papillary mucinous tumors, IPMT
It is well documented that IPMT have a high risk potential to malignant transformation, but the range reported is high: 20-70 percent. In France 106 patients were studied with respect to the duration of disease and compared to symptoms and risk of malignancy. The median age at first symptom was 61 years. Initial sign included pain (n=28), acute pancreatitis (10), jaundice (n=6), steatorrhoea (n=4), but it was an incidental finding in 39 persons. IPMT involved the main pancreatic duct in 53 patients (15 with a diameter of the duct > 10 mm) with or without involvement of branch ducts, and exclusively in branch ducts in 53 patients (4 of them with a branch duct diameter of more than 30 mm). Median duration was 21 (0-241) months, and median follow-up of patients not undergoing surgery was (0-78) months from diagnosis. The involvement of main pancreatic duct was the only parameter associated with malignancy (high grade dysplasia 45 % a<t 2 years and 50 % cancer risk at 5 years vs 10 % and 15 % in patients with involvement of branch ducts only. Female gender, acute pancreatitis, and size of duct diameter were not associated with malignancy. Moreover, the authors challenged the audience by stating that IPMT may be managed non-operatively (55).
Fluid from patients with IPMT was obtained in 47 patients by endoscopic fine-needle aspiration (77 %), aspiration during ERCP (17 %) or peroperatively (6 %). Median amylase and lipase values (range) were 10600 U/ml (0-424500) and 12800 (10-52000), respectively. Median tumor marker values were: CEA 100 ng/ml, CA 19-9 5400 U/ml, CA 72-4 8 U/ml, and mucine M1 56 U/ml. Marker profiles were identical to those observed in mucinous cystadenomas in 23-38 percent of the patients. CEA levels over 400 ng/ml were found in 27 percent of benign and 29 percent of malignant IMPTs. CA 19-9 was the most accurate marker in predicting malignancy and had a sensitivity of 75 percent, specificity of 91 percent, and positive and negative predicted values of 75 and 91 percent, respectively (56).
In 82 patients operated on for IPMT there were endocrine tumors as well in 4 patients. The reverse was true in 4 of 90 patients. The pancreatic endocrine tumors varied from 11 to 30 mm in diameter; there were 2 in the head and 2 in the tail of the pancreas. They were all four positive for chromogranin A, two were positive for glucagon, 1 for insulin and 1 for somatostatin. The frequency of the association of these tumors seems to be two high to be just at random and opens for three hypotheses (57):
o a paracrine interaction between the two neoplasms
o an alteration of a common cellular precursor
o a transdifferentiation from one cell type into the other.
In another study 38 consecutive IMPT patients, 1988-2003, were followed up 40 + 28 months after diagnosis. Presence of measurable solid tissue component within the IMPT at CT, endoscopic ultrasonography or MRI appeared as the unique malignant predictive criteria after an univariate analysis (58).

Pharmacolological treatment of pancreatic cancer
Update on chemotherapy of pancreatic cancer was given by Generoso Uomo, Napoli, Italy. He pointed out that hard endpoints for efficacy of chemotherapy such as survival are difficult to overcome. Traditional endpoints for efficacy include objective response rate, but that is in many ways more inexact. As palliation of symptoms may occur even in the absence of an improvement in survival it has instead been proposed "Clicinal Benefit Response", which once again has an in-build uncertainty. Moreover, intrinsic difficulties of chemotherapy studies relates to:
o variation in trial size
o small size
o variation of patients
o variation of doctors.

Also, locally advanced and surgically unresectable and "metastatic" pancreatic cancers are very different and should not be evaluated together in treatment studies. There has been evaluated a lot of drugs for pancreatic cancer, such as mitomycin C, doxirubicine, BCNU, CCNU, methyl-CCNU etc. None of these have been really beneficial to the patients. Today there are only two drugs that have verified effects: 5-FU and gemcitabine. These have also been tried in combinations:
FAM: 5-FU, adriamycin, mito
SMF: strptozotosin, mitomycin-C, 5FU
Gemcitabin gives a clinical benefit response in 1/3, with a tumor response rate of 15 percent, and stable disease in 45 percent of cases. In a retrospective analysis of 3000 patients with advanced pancreatic cancer given gemcitabine there was an overall tumor response in 12 percent of patients and a clinical benefit response of 18 percent. However, the potentially palliative benefit must always be weight against possible side effects.
Future directions of chemotherapy of pancreatic cancer were discussed by Volker Heinemann, Munich, Germany:
o as standard chemotherapy: gemcitabine
o pharmacokinetic improvements of gemcitabine: fixed dose rate infucion
o synergistic chemotherapy: gemcitabine + 5FU, capecitabine, ALIMTA, cisplatin, irinotecan etc.
(not statistically better than gemcitabine for survival)
o combination chemotherapy with targeted therpy: inhibition of Ras farnesyl transferase, metalloproteinase inhibitors, anti-VEGF-antibodies, anti-EGF-receptor antibody - Cetuximab - herceptine in Her-2/neu overexpressing tumors, traceva (Eroltinib), bevacizumab blocking angiogenesis, COX-dependent inhibition (Celecoxib)
o salvage therapy (second line) or 3rd or 4th line:
G-FLIP, T-GX, LIMTA, rubithecan (Orathecin).

The role of radiotherapy for localized pancreatic cancer was discussed by Morganti Alessio, Campobasso, Italy.
o interstitial brachytherapy effective but with risk of fistulas
o IORT - little effect in clinical studies
o chemoradiotherapy gives up to 50 percent response according to a review by Wannerbo 2000.
o postoperative radiotherapy
o chemoradiation prolong survival in unresectable tumors, but it must be understood that most patients with pancreatic cancer die from a metastatic, not local, disease
o new regimens must be tried aiming to improve the resectability rate
o never use radiation by itself, always compared to chemotherapy (or surgery).

Prodrug activation by living cells
Local conversion of prodrugs by encapsulated cells have been tried recently also in pancreatic cancer. Here was now presented an attempt to use human HEK293 cells genetically modified to overexpress the prodrug-converting enzyme cytochrome P450/2B1. The cells were encapsulated in cellulose sulphate capsules of 700 microm in diameter to protect the cells from the hosts immune system. Cells of several pancreatic tumor lines were effectively killed by activated ifosfamide released from co-cultivated encapsulated cells (60).
Rapamycin
The mammalian target of rapamycin, mTOR, plays a central role in the cell proliferation. Dysregulation of mTOR signalling occurs in diverse human tumors. An mTOR inhibitor have in most human pancreatic cancer cell lines an antiproliferative effect in vitro and in vivo models of pancreatic cancer (61).
Zoledronoic acid T-cells bearing the Gamma9/delta2 T cell receptor, TCR, constitute 2-10 percent of peripheral blood T-lymphocytes. They have recently raised much interest as non-MHC restricted effector cells against a variety of tumors. Gamma/Delta T cells are known to be stimulated by phosphoantigens without the need of professional antigen presenting cells. Now apoptotic and anti-proliferative effects of two bisphosphonates, pamidronate and zoledronic acid, against eight different ductal pancreatic carcinoma cell lines showed that zoledronic acid has even at concentrations which could be achieved by normal dosage an anti-proliferative and -apoptotic effect. Inhibition of proliferation correlated significantly with susceptibility against Gamma/Delta T cells (62).

Pancreatic cancer vaccines and other gene therapy
One promising alternative to other forms of palliative treatment is vaccine therapy. In particular, autologous granulocyte-macrophage colony-stimulating factor (GM-CSF)-secreting tumor vaccines have cured established tumors in mice and shown promising results in patients with prostate and renal cell carcinoma and melanoma. In theory, autologous tumor cells would be the best source of immunizing proteins, since they would likely display all of the relevant tumor antigens for inducing antitumor immunity in the patients. However, this approach poses important technical problems. First, an autologous vaccine depends on the availability of adequate numbers of tumor cells, which are rarely available because of the reactive processes that are found infiltrating the tumor cells of many common cancers. Second, and autologous vaccine requires de novo gene transfer for treatment of each patient, which is labour intensive and may cause variable GM-CSF expression levels between different patient vaccines.
Gustav Gaudernack, Oslo, Norway, gave a State-of-the-art lecture: "Targeting k-ras mutations and telomerase in pancreatic cancer by therapeutic vaccines." He explained that the general strategy in his laboratory was to direct the immune response towards genetic alterations leading to cancer and used tumors with high immunogenicity. They preferred antigens that are present in a broad spectrum of cancers - much can be gained from comparisons between the effects on pancreatic cancer and other gastrointestinal cancer. It is important to have a broad HLA binding profile. It might not be so nice to tell, but of course the short survival of the unresected patients - 98 percent will die within one year - make the evaluation of treatment attempts interesting.
As mutant K-Ras is found in 90-100 percent of adeocarcinomas of pancreas and as telomerase in all tumors are different compared to normals it has been tempting to try these peptides as a basis for vaccine production. Immortalization of somatic cells requires reactivation of telomerase (hTERT). A cocktail of ras-peptides was tried as early as 1993 in the Oslo studies (also in colorectal cancers):
o no side effects
o prolonged survival of immunological responders.

In Oslo they have at the Radium Hospital several ongoing studies. Usually they activated vigorously the first week to get an inflammatory situation. In their last 48 patients (6 vaccines) immunity was induced in 54 percent of the patients, and as they had tried different doses they had some clues to witch dosage to use in the future (the smallest and highest doses gave an inferior immogen response). They have worked with two adjuvant studies CTN 9502 (n=10) and CT 98010 and found a respons immunogenically in 85 percent of the patients. A combination of ras and telomerase is now being considered as the next step.
Evaluation of lentiviral vectors for human pancreatic cancer gene therapy in five pancreatic cancer cell lines showed that lentiviral vectors are capable of permanent integration into the host's DNA with little toxicity. Two vectors, EIAV and HIV-1 based, resulted in efficient and stable transduction of pancreatic cancer cell lines (63).
Nicole Vaysse, Toulouse, France, discussed basic principles of gene therapy and stated that gene therapy is the use of genetic manipulation for disease treatment. The developing strategies are to use the specificity and activity of the own products of the body. It is interesting to note that for example expression of tumor-suppressor genes is different in different cells: normal cells - no effect; tumor cells: growth arrest, apoptosis, and bystander effect. Among the tumor suppressor genes that have been approached for gene therapy are p16/INKA4, Smad4, RB, p53, and PTE. It has also been tried suppressor genes such as mda-2 and sssr-2. These are probably safe to use but have bystander effects. If instead blocking of oncogene expressions are attempted the same results as for tumor suppressors can be expected but there are no bystander effects. It has also been tried to kill tumor cells by suicide genes: e.g. to use a prodrug and an enzyme to produce toxins to other cells. Examples of this is gamcyclovir that is activated by HSV-tk and then blocks DNA synthesis.
There are also experimental therapy aiming at killing tumor cells by oncolytic viruses with virus replication and lysis and then spreading to other cells. Also, strategies directed to the host are possibble, protecting normal tissue from the toxic effects of chemotherapy. There are also new combinations:
- blocking oncogenes + forced expression of tumor suppressor genes
o antiangiogenetic targets + cytokine therapy
o synergizing + chemotherapy or radiation.

There are several practical ways to deliver genes into cells:
o viral systems: adenovirus, AAV, herpes, lentivirus, retrovirus, rSV40
o non-viral system: local injection, naked DNA, liposomes
o liposomes plus PEI (complexed DNA).

Retroviral transduction infects only dividing cells. This may be overcome by:
o modified envelope proteins
o use of specific cell surface receptors
o tissues including endothelial cells
o induction of some diseases
o ionising radiation or hypoxia to condition cells.

As these gene therapies are potentially dangerous. Therefore, preclinical studies to evaluate risks and benefits are needed. Unfortunately no single mouse model is likely to expose the full spectrum of human malignancies. The perspectives according to Vaysse is:
o several specific hurdles need to be overcome
o multifactoral nature of the disease is one of the cause preventing the discovery of new targets
o combinational approaches to gene therapy is probably needed.

The issue of applications of gene therapy to pancreatic cancer was covered by Mattias Löhr, Mannheim. He stated that very few drugs which work convincingly in pancreatic cancer are available. In the Cochran Library for 2004, there are no entries. Historically it can be mentioned that the 14 september 1990 W. French Anderson had the first replacement therapy, i.e. 14 years ago. On the other hand Jesse Gelsinger, 18 years old, in 1999 died of a high dose of viruses in an experimental setting. GM-CSF vaccination in Johns Hopkins gave a heavy immune response, disease-free survival increased.
Ex vivo gene therapy is interesting, so is therapeutic vaccination with dendritic cells. However, from a clinical point of view suicidal gene-therapy may be closer as there is no virus replication.
Virus tumor injection has succeeded in head-and-neck tumors in humans as has adenovirus for treatment of metastases.
Löhr has worked with genetically modified cells that activate ifosfamid by P450. The cells are microencapsulation and have a local mode of action, e.g. vascular way. He summed up by saying that success in treating human pancreatic cancer xenografts in mice does not necessarily predict success in humans.
In an extensive and comprehensive report M. Sunamura from the S. Matsuno group in Sendai, Japan, described oncolytic viral therapy as a novel strategy for pancreatic cancer treatment. They have used adenovirus E1 gene products, in addition to transactivating other early gene promotors, prepare the cellular environment for optimal viral replication by associating with a number of key cell cycle proteins. Replication-selective viruses may overcome the limitations of gene transfer of conventional adenoviral vectors. Viral replication in a small fraction of tumor cells leads to amplification and extension of the anti-tumor effect of gene expression. Cell death is due exclusively to viral replication and cell lysis. In Sendai it has been constructed a double-mutant, replication-selective adenovirus containing a mutation in the RB-binding motif of the E1A region and a deletion of large E1B-55kDa, suggesting a wider therapeutic potential given that most pancreatic cancers have abnormalities in both TP53 and RB pathways. The therapeutic effect of AxE1AdB on subcutaneous tumor model and in selected human head-and-neck cancers (together with cytostatics) have been significant (64).
It was also discussed the anti-tumor and anti-metastasis effect of a recombination adenovirus expressing the proapoptotic gene TRAIL (TNF-related apoptosis inducing ligand, APO-2 ligand)-Bax in combination with gemcitabine for the treatment of pancreatic cancer in vitro and in an experimental model. The number of metastasis and tumor volume were significantly reduced after combined administration of Ad-TRAIL and Ad-Bax and the effect was more pronounced in combinations with gemcitabine, including survival (65).
Optimised retroviral vectors for suicide gene therapy in pancreatic cancer was also reported with an improved promoter conversion (ProCon)-retroviral vector system. Upon transduction of target cells with ProCon vectors the retroviral promoter is replaced by a tissue-specific promoter, which regulates the expression of the transgene in these cells. Unfortunately, tissue specificity of retroviral vectors often is associated with reduction in viral titer or shut down of transgene expression (66).
It was shown that short interfering RNAs, siRNA, targeting specific gene expression was more effective than the antisense phosporothidate oligonucleotide for antitumor activity against a specific K-ras mutation, K-ras mut12 (67).
The suicide gene cytosine deaminase, able to convert the 5-fluorocytosine into 5-fluorouracil has been used for gene therapy. The co-transfection of cytosine deaminase and 5PRT from Saccaromyces cerevisiae enhance 5-fluorocytosine sensitivity of pancreatic cancer cell lines, possibly because the enzyme encoded by 5PRT may in part counteract the 5-FU degradation observable in some pancreatic cancer cells (68).

Endocrine pancreatic tumours

Endocrine pancreatic tumors, a subgroup of the neuroendocrine tumors, share a number of features. Each type is histologically similar to each other and to carcinoid tumors. They are classified as APUDomas (Amino-Precursor Uptake and Decarboxylation) and share cytochemical features with melanoma, pheochromocytoma, carcinoid tumors and medullary thyroid carcinoma. Except for insulinoma, each is malignant in most cases, whereas more than 90 percent of insulinomas are benign, intrapancreatic, and solitary. The pancreatic tumors may be either functioning and associated with a specific clinical syndrome due to hormone overproduction or non-functioning. They are considered functioning if they are associated with a clinical syndrome due to eutopic or ectopic hormone release and non-functioning if they are not associated with clinical symptoms due to hormone overproduction. In this latter category are included tumors that release pancreatic polypeptide, neurotensin, HCG -alpha and -beta subunits and cromogranin A. Their incidence is 0.3-0.4 per 100 000 population, with non-functioning tumors forming the biggest group (15-40 %), followed by gastrinomas. All of these tumors other than insulinomas are malignant in over 50 percent of cases and so treatment must be directed at both tumor growth and hormone excess. The most common functioning types are gastrinomas described in 1955 by Zollinger and Ellison, with severe peptic ulcer disease and extreme hypersecretion of gastric acid. The second most common entity is insulin producing tumors encompassing a broad range of diagnostic and therapeutic features, the so-called Whipple's triad, which consists of the characteristic symptoms of hypoglycemia, associated low blood sugar levels and an immediate relief following ingestion of glucose. The incidence of insulinomas is 0.8-0.9 cases per million persons and year. Gastrinomas is considered to be 1.5 times as common as insulinomas. Other well-described clinical syndromes are the Verner-Morrison syndrome, described in 1958: severe secretory diarrhea, hypokalemia and hypochlorhydria (WDAH). Another distinct syndrome is the glucagonoma syndrome with the necrolytic migratory erythema of the skin. The last two clinical entities are rare.
In a symposium moderated by Kjell Öberg, Uppsala, Sweden, Gunter J. Kreis, Graz, Austria, and Claudio Pasquali, Padova, Italy, on endocrine pancreatic tumors, Paul Komminoth, Baden, Switzerland, spoke on "Genetics of endocrine pancreatic tumours" he showed that there emerges a pattern. For example:

MEN1 11q13 >80 % risk of malignancy
VHL 3q25 5-10 % risk of malignancy
NF1
(neurofibromatosis type 1)
17q11.2 very rarely tumors
TSC2
(tuberous sclerosis)
 very rarely tumors

MEN1 is found also in sporadic tumors - not only in the MEN syndrome - in some reports 13 percent of the patients with endocrine tumors had this mutation. Non-functioning tumors hade a lower frequency, and insulinoma had mutations in 8 percent of the patients with alleleic deletions 39 percent. VHL-patients, on the other hand, had this mutation in only 1 of 75 cases. Many chromosomes are involved with both gains and losses. The bigger tumor and the more malignant, the more losses compared to gains. Also functioning small (< 2 cm) tumors usually have markers of malignancy and of metastases, such as mutations in 3p23, 6q22, 9p, 11q13, which seems to have more losses not picked up otherwise. There seems also to be a role for mutations in VHL, p16, PTEN, K-ras, p53 (1-3 %). However, taken together the molecular pathogenis of pancreatic tumors is only partially understood.
The pathologist's point of view was given by Cesare Bordi, Parma, Italy. There is nowadays a rather new WHO classification of neuroendocrine tumors taking in account.
o site of origin
o combination of pathology and clinical features
o grading histopathological.

There are well described criteria for differentiating well from poorly differentiated endocrine carcinomas.
Wouter W. de Herder, Rotterdam, The Netherlands, gave his view on "Radiolabeled somatostatin analogues in localisation and treatment". These membrane receptors can be used for visualization and for intraoperative detection. The first use for localization was done in Rotterdam already in 1988. It must be understood, however, that all endocrine tumors are not receptor-positive. Gastrinomas > 2.0 cm (n=25) were detected in 96 percent of cases with somatostatin receptor scintigraphy, smaller tumors are seen less well. In the clinic OctreotideScan®-positiv patients also reacts to doses of somatostatin. The higher uptake, the better result when treated with 111In-DTPA-octereotide. There are antitumor effects and symptomatic improvement. Toxicity is limited, and concerning the low renal toxicity - amino acid infusion protects. Today treatment is given to patients with progressive disease, e.g. chemoembolization of liver metastases. In the future probable intervention is also done earlier.
Kjell Öberg spoke on "Advances in medical treatment". There are several good targets for treatment:

proliferation Ki-67
angiogenesis bFGF, VEGF
cell adhesion CD 44, V6, V9
apoptosis bcl-2, bax
receptors SST1-5
chromogranin fragments
growth factors 

It must be noticed that there is a heterogeneity inside also a small tumors regarding the factors above. Also, different tumors express different factors. p-chromoganinen A + B is found in about 94 percent of cases. Regarding insulinomas a majority of patients present with metastatic disease but vague symptoms. There is a mean delay of symptoms of 4 years and a wide spectrum regarding survival time. Some patients may present with life-threatening symptoms. There is often a good clinical status despite metastases.
Surgery is still the best treatment if possible, but otherwise chemotherapy is tried: streptozotosin, cisplatin + etopside. Also alfa-interferon and combinations with retinoids have been tried. If cure is not possible reduction of the tumor burden may be a good possibility. This might be made with radiotherapy and as symptomatic treatment diazoxide and omeprazole may be tried. Median survival after chemotherapy is 19 months.
Recombinant interferon gave a biochemical response in 44 percent and tumor reduction in 11 percent of 372 patients. Flu-like symptoms were experienced in 80 percent and fatigue 70 percent. From a scientifical point of view it is seen an upregulation of SSSR-2, stimulation of cytotoxic T-cells and a downregulation of EGF.
Octreotide is given i.a. dose of 100-1500 microgram/day (100-500 microgram x 2-3) and there has then been seen a tumor response in 0-15 percent of cases and biochemical response in 30-60 percent. When a combination of interferon and octreotide was given for progressive diseased a biochemical response was seen in 63 percent of cases.
For the future Öberg predicted a more individualized treatment based on tumor biology and molecular genetics. Probably tyrosinekinase will be new targets together with new somatostatin analoges and inhibitors of angiogenesis. Glivec is an interesting drug as is rapamycine, that blocks a new pathway and blocks proliferation and apoptosis.
Göran Äkerström, Uppsala, Sweden, discussed the challenge of surgical treatment of endocrine pancreatic tumors. He was not so enthusiastic of laparoscopic surgery as it gives a higher risk of fistualization and other complications.
Regarding duodenal gastrinomas this location is found in 60 percent. Many of them do not stain for gastrin and is then difficult to find. Lymph-node metastases are more easily seen, and are sometimes mistaken as the primary tumor. Whipple-type surgery is however rarely needed. These tumors are almost always malignant, but have despite that very different prognosis. Thirty percent of these tumors are non-aggressive, and very slowly growing.
Glucagonoma are usually distal pancreatic lesions, but frequent with lymph node metastases. Non-functioning sporadic endocrine pancreatic tumors are also more commonly found, especially in young persons without cachexia. Operation may be indicated despite presence of some liver metastases. These tumors are not as invasive exocrine tumors, which mean that they can frequently be dissected from artery and veins, possibly with a reconstruction with a graft. The survival is good if the tumor is removed, 65 percent are alive after 5 years. If liver metastases are present surgery gives no survival benefit.
MEN1 patients should be explored regarding the pancreas, even with no image shown tumor, as 30-50 percent already have metastases.