Anti-cytokine strategies in acute pancreatitis: pathophysiological insights and clinical implications.

Abstract

The clinical presentation of acute pancreatitis varies significantly from mild self-limiting discomfort to a severe life-threatening condition. Once the disease process is initiated, the severity of the disease is largely determined by a complex network of activated inflammatory mediators such as cytokines, proteolytic enzymes, reactive oxygen species, and many more which render the local injury to a systemic disease with multiple organ dysfunction, sepsis, and considerable mortality. Remarkable progress in diagnostic modalities, intensive care technologies, and organ preserving surgical techniques have decreased mortality of severe acute pancreatitis during the past decades. However, the treatment of acute pancreatitis still remains largely supportive and no specific approach exists to prevent evolving complications. A large body of clinical and experimental evidence suggests that cytokines are key factors in the pathomechanism of local and systemic complications of acute pancreatitis. Targeting cytokine activity as therapeutic approach to acute pancreatitis is a challenging concept and the results of modulating activation of TNF-alpha, IL-1beta, IL-2, IL-10, PAF and various chemokines has indeed been promising in the experimental setting even if tested under therapeutic conditions. However, experience from a limited number of clinical trials on anti cytokine strategies in acute pancreatitis has remarkably emphasized that translating successful experimental observations into reproducible clinical associations seems to be difficult.

Full text

Page 1

106
Rau BM, et al.
Roczniki Akademii Medycznej w Białymstoku · Vol. 50, 2005 · Annales Academiae Medicae Bialostocensis
* CORRESPONDING AUTHOR:
Department of General, Visceral and Vascular Surgery
University of the Saarland
Kirrberger Strasse
66421 Homburg/Saar, Germany
Tel: +49 6841 162 2630; Fax: +49 6841 162 3132
e-mail: bettina.rau@uniklinikum-saarland.de (Bettina M. Rau)
Received 21.06.2005 Accepted 21.06.2005
Anti-cytokine strategies in acute pancreatitis:
pathophysiological insights and clinical implications
Rau BM*, Krüger CM, Schilling MK
Department of General, Visceral and Vascular Surgery, University of the Saarland, Homburg/Saar, Germany
Abstract
The clinical presentation of acute pancreatitis varies
significantly from mild self-limiting discomfort to a severe
life-threatening condition. Once the disease process is initi-
ated, the severity of the disease is largely determined by a
complex network of activated inflammatory mediators such
as cytokines, proteolytic enzymes, reactive oxygen species,
and many more which render the local injury to a systemic
disease with multiple organ dysfunction, sepsis, and consid-
erable mortality. Remarkable progress in diagnostic modali-
ties, intensive care technologies, and organ preserving
surgical techniques have decreased mortality of severe acute
pancreatitis during the past decades. However, the treat-
ment of acute pancreatitis still remains largely supportive
and no specific approach exists to prevent evolving compli-
cations. A large body of clinical and experimental evidence
suggests that cytokines are key factors in the pathomecha-
nism of local and systemic complications of acute pancrea-
titis. Targeting cytokine activity as therapeutic approach to
acute pancreatitis is a challenging concept and the results
of modulating activation of TNF- , IL-1 , IL-2, IL-10, PAF
and various chemokines has indeed been promising in the
experimental setting even if tested under therapeutic condi-
tions. However, experience from a limited number of clinical
trials on anti cytokine strategies in acute pancreatitis has
remarkably emphasized that translating successful experi-
mental observations into reproducible clinical associations
seems to be difficult.
Key words: cytokines, acute pancreatitis, anti-cytokine
approaches.
Introduction
Acute pancreatitis usually takes an uneventful course with
complete restitutio ad intergrum and mortality rates of less
than 1%. In contrast, in about 25% of all patients the disease
takes a severe course evolving to a potentially life-threatening
condition with considerable morbidity and mortality [1]. It has
been largely shown that the hallmark of severe acute pancrea-
titis is the development of pancreatic necrosis which is still the
essential determinant of further complications [2,3]. Depending
on the extent of intra- and extrapancreatic necrosis pancreatic
infections and remote organ failure frequently arise as major
complications in the subsequent course of the disease and con-
siderably add to mortality [2-6]. Nowadays, a multidisciplinary
approach of prolonged intensive care management and delayed
organ sparing surgical protocols has decreased the mortality of
severe acute pancreatitis to about 20% to 30% in the past 20
years [7,8]. Despite the introduction of novel therapeutic con-
cepts such as early ERCP in biliary acute pancreatitis, prophy-
lactic antibiotics, and enteral nutrition since the early 90ies [7]
the management of acute pancreatitis still remains largely sup-
portive and despite all efforts, a break-trough in lowering mor-
tality in patients with severe attacks has not been achieved [3-6].
Currently, no disease-specific medical treatment has ever been
proven to overcome relevant complications such as pancreatic
necrosis, infection of necrosis or organ failure effectively.
Dissatisfaction with persistently high mortality rates along
with an improved understanding of the underlying patho-
mechanism of acute pancreatitis have made pancreatologists
to pursue the search for alternative therapeutic approaches.
More than 100 years ago Chiari proposed the first pathophysi-
ological concept of acute pancreatitis by “autodigestion” of
the gland via its own enzymes [9]. Since the mid 80ies, the
pancreatic proteinase-antiproteinase imbalance was thought

Page 2

107
Anti-cytokine strategies in acute pancreatitis: pathophysiological insights and clinical implications
to play a key role in the pathogenesis of acute pancreatitis.
Hence, trypsinogen activation is believed to be one of the earli-
est pathophysiological events which triggers a cascade of other
pancreatic proenzymes such as chymotrypsinogen, type I pro-
phospholipase A2, procarboxypeptidase B, or proelastase [10].
According to the “autodigestion” theory by Chiari, premature
trypsinogen activation within the acinar cells has been found
in various experimental models of acute pancreatitis [10,11].
Subsequently, significant amounts of trypsinogen and other pro-
teases have been measured in the interstitial space as well as in
the systemic circulation with a positive correlation to the extent
of pancreatic tissue destruction and overall disease severity [11].
However, trypsinogen activation is only a temporary event in
acute pancreatitis and most recent experimental studies have
questioned the prevailing opinion of its dominating pathophysi-
ological role [12]. These recent findings would at least in part
explain the failure of antiproteinase therapy [13] or inhibiting
pancreatic enzyme secretion [14] in decreasing complications
and associated mortality in human acute pancreatitis.
An alternative pathophysiological concept was proposed
by Rinderknecht in 1988. According to his theory, a complex
network of inflammatory mediators released by activated
leukocytes was suggested as key factor for rendering the local
pancreatic insult into a systemic disease with distant organ fail-
ure [15]. In the subsequent years a growing number of clinical
studies convincingly showed that acute pancreatitis is reflected
by a large array of circulating inflammatory variables such as
cytokines, chemokines, reactive oxygen species, adhesion mol-
ecules, acute phase proteins, and others [16,17]. In patients with
an overt systemic inflammatory response and subsequent organ
failure the quantitative release of nearly all mediators measured
was significantly higher than that observed in patients with mild
disease [18,19]. As a clinical consequence, measurement of
specific inflammatory mediators offered a new interesting alter-
native to an easier severity stratification of acute pancreatitis
compared with expensive imaging procedures or clinical staging
scores [20]. Beyond the diagnostic and prognostic implica-
tions the “mediator hypothesis” was further substantiated by
a growing number of experimental studies which ultimately
lead to the establishment of the so-called “three” hit theory
(Fig. 1). Herein, various pro- and antiinflammatory mediators
are considered as important link between the initial local insult
(first hit), the systemic host response and organ failure (second
hit), and subsequent septic complications (third hit) in acute
pancreatitis. By either inhibiting leukocyte activation or directly
targeting leukocyte derived inflammatory mediators cytokines
Figure 1. Schematic overview of the inflammatory cascade in acute pancreatitis. Activation of various leukocyte subsets and endothelium
at the site of injury release various pro- and antiinflammatory cytokines, chemokines, and other mediators. An overt and sustained activa-
tion of proinflammatory mediators leads to systemic inflammatory response syndrome (SIRS) which may further proceed to multi organ
dysfunction syndrome (MODS), infection of necrosis and sepsis
2nd hit
Infected Necrosis Sepsis
MODS
3rd hit
1st hit
Acinar Cell Damage
Protease Activation
TAP, CAPAP, PLAP/PROP
Pro-inflammatory Mediators
TNF-α, IL-1β, IL-18, IL-6, IL-2
Adhesion Molecules, PAF
Nitric Oxide
Oxygen Free Radicals
Chemokines
IL-8,
RANTES
GRO-α
ENA-78
MCP-1
Anti-inflammatory Mediators
IL-10
IL-1ra
IL-11
SIRS
Acute Phase Response, Pyrexia, Tachycardia, Tachypnoea
Gut ischemia, Bacterial translocationPulmonary / Renal failure, Shock
Attraction / Activation of Mediator Releasing Cells
PMN-LeukocytesMonocytes / MacrophagesEndotheliumLymphocytes

Page 3

108
Rau BM, et al.
have been recognized as central determinants of severity
in acute pancreatitis and emerged as interesting targets for
a potential therapeutic approach (Tab. 1).
The Role of Cytokines
Cytokines are a family of low molecular weight proteins
(16-28 kDa), which have been extensively investigated in
inflammatory conditions including acute pancreatitis. More
than 30 different cytokines have been identified so far. With few
exceptions, cytokines are not constitutively expressed in normal
tissues and upregulation is usually initiated following external
stimuli such as injury or stress in various cell types [16,17].
All cytokines cause their effects via highly specific membrane
bound cell-surface receptors and have pleiotropic activities on
a variety of target cells. On a functional basis cytokines can be
divided into two groups: the pro- and the anti-inflammatory
cytokines. There is immense redundancy within the system in
such that many cytokines share similar biological effects and in
the absence of another, they can fill the gap. Currently, there is
no more doubt about the detrimental role of many cytokines in
promoting local tissue destruction and mediating distant organ
complications in acute pancreatitis and inflammatory disorders
in general.
Tumor necrosis factor- (TNF- )
and Interleukin-1
Tumor necrosis factor- (TNF- ) and interleukin-1 (IL-1 )
belong to the so-called “first order” proinflammatory cytokines.
Activated macrophages and polymorphonuclear leukocytes
have been thought to be the primary site and major source of
TNF- and IL-1 synthesis for years [18,21]. However, more
recent studies have emphasized the impact of acinar cells in
contributing to the synthesis and release of TNF- and other
cytokines [22-25]. A primary involvement of TNF- in acute
pancreatitis was shown in both clinical [26] and experimental
studies [27-30] already since the early 90ies. Besides an early
rise of systemic TNF- concentrations [27-29] organ-specific
and time dependent upregulation of TNF- mRNA and protein
levels in the pancreas but also in distant organs such as lung and
liver provided the first evidence that cytokines are important
mediators of systemic complications and survival [29-31]. In the
experimental setting TNF- antagonism by either anti TNF-
antibodies or TNF- receptor blockade almost uniformly
revealed protective effects on local intrapancreatic damage,
systemic severity, and mortality [32-36]. An effective amelio-
ration of pancreatitis associated pulmonary damage could be
shown by alternative anti TNF strategies using inhibitors of p38
mitogen-activated protein (MAP) kinases of nuclear factor B
(NF B) [23,37,38]. Interestingly, the protective effect of TNF-
antagonism on disease severity and mortality was still observed
in a therapeutic study design after the systemic effects already
had fully developed [35].
Similar observations have been made for IL-1 , the second
of the “first order” cytokines. As observed for TNF- , organ-
specific expression of IL-1 is an early feature in experimental
acute pancreatitis and is found in both the pancreas and distant
organs [21,31] and correlates with the severity of the model
studied. However, in contrast to an overt local overexpression,
systemic IL-1 concentrations remain relatively low [19,39].
Unlike TNF- , IL-1 synthesis and release has been mainly
demonstrated by activated leukocyte populations [21,18]
and no direct effect of this cytokine on acinar cell viability or
function has ever been demonstrated [40,41]. Blockade of the
IL-1 receptor by pharmacological agents or targeted genetic
disruption revealed a significant reduction of intrapancreatic
damage, systemic severity, and mortality in every established
pancreatitis model similar to that observed by blocking TNF-
[36,42-45]. A recent interesting alternative approach to inhibit
IL-1 activation in acute pancreatitis included the inhibition of
caspase-1, formerly termed interleukin 1 -converting enzyme
Table 1. Therapeutic effects of cytokine modulating approaches in experimental acute pancreatitis
CDE – choline deficient ethionine supplemented; TC – Taurocholate; GDOC – Glycodeoxycholic acid; ND – not determined
Author TargetModel
Delay of drug
administration
Intrapancreatic damage Distat organ damageMortality
Norman et al. [35]
Norman et al. [43]
Norman et al. [45]
Paszkowski et al. [48]
Kusske et al. [75]
Rongione et al. [76]
Zou et al. [80]
Mayer et al. [89]
Formela et al. [99]
Hofbauer et al. [101]
Foitzik et al. [102]
Bhatia et al. [117]
Bhatia et al. [118]
Bhatia et al. [120]
TNF-
IL-1
IL-1
IL-1 /ICE
IL-10
IL-10
IL-10
IL-2
PAF
PAF
PAF
CINC
RANTES
MCP-1
CDE, mouse
Cerulein, mouse
CDE, mouse
TC, rat
CDE, mouse
Cerulein, rat
TC, rat
Cerulein, mouse
Ischemia, rat
Duct ligation, opossum
GDOC, rat
Cerulein, rat
Cerulein, mouse
Cerulein, mouse
1.5 days
1 hour
1.5 days
12 hours
33 hours
2 hours
30 minutes
6 hours
30 minutes
2 days
6 hours
1 hour
1 hour
1 hour
edema
necrosis, edema
necrosis, edema
necrosis
histologic score
histologic score
necrosis
histologic score
histologic score
necrosis
microcirculation
no effect
no effect
necrosis
no effect
ND
lung
lung
ND
ND
lung , liver
lung
ND
lung
lung, kidney function
lung
lung
ND
decrease
ND
decrease
decrease
decrease
ND
decrease
ND
ND
ND
decrease
ND
ND
ND

Page 4

109
Anti-cytokine strategies in acute pancreatitis: pathophysiological insights and clinical implications
(ICE). Targeting ICE activity by a specific synthetic inhibitor led
to a dramatic amelioration of severity and mortality irrespective
of the model used [46-48]. Of specific interest is the fact that
the protective effects on overall severity and mortality were still
present after a therapeutic window of 12 hours following induc-
tion of severe acute pancreatitis [48]. By comparing anti TNF-
and anti IL-1 strategies, it becomes clearly evident that both
cytokines share striking similarities in their pathophysiological
functions and closely control the regulation of their own and
each other [44,49]. This was convincingly shown by Denham
et al. [36] who could not demonstrate any additive protective
effects by combined genetic disruption of TNF- and the IL-1
receptor in two models of murine acute pancreatitis.
Surprisingly, in contrast to their outstanding pathophysi-
ological importance both cytokines play no role as biochemical
markers for a reliable severity assessment of acute pancreatitis
in the clinical setting. Is has been largely shown that TNF-
measurements are difficult, because they are substantially
hampered by intermittent TNF- release and a short plasma
half-life of less than 20 minutes. Similar observations have been
made for IL-1 , which shows an early and transicient increase
in most severe cases only [19,50-52]. The soluble TNF receptor
complex as well as the IL-1 receptor antagonist (IL-1RA) are
more stable than the cytokines itself and thus easier to measure.
Although TNF receptors [50,53] and IL-1ra [19,50,51,54] were
found to correlate with severe acute pancreatitis and associated
organ failure they are no candidate parameters for a meaningful
clinical application.
In contrast to the extensive investigations of both cytokines
in experimental respect no study has ever been conducted inves-
tigating TNF- or IL-1 antagonism in clinical acute pancreatitis.
However, interesting insights can be drawn from a number of
sepsis trials. Acute pancreatitis, especially in its severe form,
shares striking similarities with sepsis and septic shock. The
clinical feature of multi system organ failure and the inflam-
matory mediator profile are indistinguishable in each of these
conditions and suggest a common pathogenic mechanism, albeit
as a result of different inflammatory stimuli. Some large rand-
omized multicenter trials on anti TNF- and IL-1 antagonism
in patients with sepsis have demonstrated overall disappointing
results. The use of an anti TNF- antibody in patients with sep-
sis failed to reduce 28 day mortality in two phase III trials, the
North American Sepsis Trial (NORASEPT) and the Interna-
tional Sepsis Trial. Likewise, fusion proteins for p75 TNF-R and
p55 TNF-R were ineffective as well [55,56]. A post hoc analysis
of a controlled trial of human recombinant IL-1 RA in patients
with sepsis syndrome revealed an increase in survival time in
a subgroup of patients with multi organ failure [57]. However,
this observation again could not be confirmed by a subsequent
trial [58].
Interleukin-18
Interleukin-18 (IL-18), formerly called interferon- -induc-
ing factor, is a novel proinflammatory cytokine playing an
important role in the Th-1 response, primarily due to its ability
to induce IFN- production in T-cells and natural killer cells [59].
IL-18 shares striking similarities with IL-1 concerning struc-
ture and function. Both are synthesized as biologically inactive
precursors requiring proteolytic cleavage into their mature form
by caspases-1/ICE. Moreover, the biological activity of IL-18 is
closely related to that of IL-1 : IL-18 induces the gene expres-
sion and synthesis of TNF, IL-1, and several chemokines by
means of a putative IL-18 receptor complex which is a member
of the IL-1R family as well [59]. IL-18 has gained considerable
attention since the striking protective effects of caspase-1 inhi-
bition have been reported by a large number of experimental
studies in various inflammatory conditions [60] including acute
pancreatitis [46-48]. Under therapeutic conditions, caspase-1
antagonism has been more effective in reducing pancreatitis
related severity and mortality [48] than has IL-1 antagonism
[43,45] in severe experimental models. Therefore, caspase-1
mediated activation of IL-18 may well explain the better results
of blocking caspases-1 activity [61]. In fact, by comparing the
dynamics of systemic IL-1 and IL-18 concentrations an inter-
esting observation became evident supporting this theory: IL-1
revealed a temporary and moderate increase during the very
first days after onset of symptoms in severe disease only [19]. In
contrast, IL-18 was released in much higher concentrations with
maximum levels during the second week after disease onset in
patients with persisting multiorgan system failure [62,63]. The
effectiveness of delayed ICE treatment could therefore be a re-
sult of inhibited generation of mature IL-18 rather than IL-1 .
Interestingly, neutralizing IL-18 activity by monoclonal antibod-
ies has indeed proven to decrease intrapancreatic damage more
effectively than neutralizing IL-1 activity in cerulein-induced
pancreatitis in mice [64]. Since a therapeutic inhibition of IL-18
has not been investigated in any model of acute pancreatitis so
far this interesting cytokine will need further investigation.
Interleukin-6
Interleukin-6 (IL-6) is produced by a wide range of cells
including monocytes/macrophages, endothelial cells, and
fibroblasts in response to potent proinflammatory stimuli such
as endotoxin, IL-1 or TNF- . IL-6 is the primary inducer of
the acute-phase response in various inflammatory conditions
[65]. The clinical value of IL-6 for an early and accurate severity
stratification of acute pancreatitis has been recognized since the
very first reports on cytokine measurements in human acute
pancreatitis appeared in the literature [66,67]. Hence, a large
number studies have addressed this issue which uniformly
confirmed that IL-6 is an earlier marker of severity than the cur-
rently established “gold standard” C-reactive protein [19,67]. In
terms of predicting pancreatitis associated complications, IL-6
was found to be an excellent predictor of remote organ failure
[18,19,53]. In contrast to the exhaustive clinical investigation
of IL-6, only few studies have ever addressed the role of this
cytokine as potential target for modulating disease severity.
From the limited number of experimental studies prophylactic
inhibition or genetic deletion of IL- 6 had a deleterious [68,69]
rather than a protective effect [70] on disease severity and
mortality [68-70]. Therefore, the presence of a physiologic IL-6
mediated inflammatory response seems to be necessary for local

Page 5

110
Rau BM, et al.
and systemic damage control in acute pancreatitis. However, in
an overall sense, these observations preclude IL-6 as driving
force for the initiation or propagation of organ-specific compli-
cations of acute pancreatitis.
Interleukin-10
Interleukin-10 (IL-10) is a potent anti-inflammatory
cytokine expressed by almost all cells but primarily released
by activated monocytes/macrophages and Th-2 lymphocytes.
This cytokine exerts its antiinflammatory properties through
inhibition of various proinflammatory cytokines and adhesion
molecules on the transcriptional and post-transcriptional level.
In addition, IL-10 induces the synthesis of natural cytokine
antagonists such as IL-1RA and TNF- receptors [71]. In
human acute pancreatitis, circulating levels of IL-10 were
found to correlate with the severity of the disease [19,51] and
with organ failure or death in some, but not all of the studies
[72,73]. Interesting pathophysiological aspects of this cytokine
arose from a number of experimental studies which uniformly
demonstrated a protective effect of IL-10 in several models of
acute pancreatitis [74-80]. Irrespective whether IL-10 activity
was blocked [79], the IL-10 gene was genetically disrupted
[78] or the cytokine activity was augmented [74-77,80] organ
specific damage in the pancreas, the lung and the liver as well as
mortality were significantly reduced. Of specific interest was the
observation that the protective effects were still observed when
active IL-10 or genetic transfection was started in a therapeutic
fashion after acute pancreatitis had been induced [75,76,80].
Convincing experimental evidence of IL-10 modulating
approaches has driven the design of clinical trials with the aim
to reduce the occurrence of post-ERCP pancreatitis by prophy-
lactic recombinant IL-10 administration. In 2001 Deviere et al.
[81] published a single center, double blind controlled study in
patients undergoing ERCP, which showed that IL-10 was able
to decrease the incidence of post-ERCP pancreatitis independ-
ently from other risk factors as well as the length of hospital
stay. Inconclusive results came from an US-american trial in
which only a trend toward a reduced incidence of post-ERCP
pancreatitis and hospital stay was found [82]. A recent meta-
analysis including four randomized clinical trials in 294 patients
receiving recombinant IL-10 and 259 patients receiving placebo
before ERCP could show that IL-10 significantly reduces the
risk of post-ERCP pancreatitis [83]. However, the ultimate ben-
efit of IL-10 treatment in preventing post-ERCP pancreatitis is
not definitely proven and still needs further evaluation.
Interleukin-2
Increasing clinical evidence suggests that an impaired
immune function contributes to the progression of acute
pancreatitis. However, cellular immune functions constitute
a complex network and seem to have distinct roles in the early
toxic and the late septic stages of acute pancreatitis [19,84-86].
Interleukin-2 (IL-2) is a product of activated Th-1 lymphocytes
and plays a central role in normal immune function. Clinical and
experimental observations have pointed out that the activation
of the T-cell system within the inflammatory cascade of acute
pancreatitis enhances pancreatic tissue injury [87], the inflam-
matory response [88-90], and mortality [88]. The release of the
soluble IL-2 receptor shows a close correlation with persisting
organ complications during the later stages of the disease [19,86]
with peak levels predicting a lethal outcome [19]. In diet induced
acute pancreatitis in mice a significant reduction of IL-2 produc-
tion with a consecutively enhanced susceptibility to endotoxin-
induced mortality was found during the later stages of the disease
which could be reversed by in vivo therapy with recombinant IL-2
[90]. These experimental data are well in line with clinical obser-
vations and strongly suggest that an impaired immune function
increases the risk of subsequent septic complications. Interest-
ingly, the administration on the immunostimulant Levamisole,
which is known to potentiate IL-2 production, effectively
decreased the incidence of pancreatic infections in a cat model
of severe acute pancreatitis [91]. In contrast to the late effects of
IL-2 deficiency and immunoparalysis the deleterious conse-
quences of an overt IL-2 mediated T-cell response during the
early course of the disease could be emphasized in moderate to
severe models of murine acute pancreatitis [87-89]. T and B-cell
deficient mice with acute pancreatitis exhibit significantly lower
pulmonary damage [88]. The immunosuppressant FK506 which
inhibits IL-2 production on the transcriptional level effectively
decreased early local and systemic disease severity [89,92],
even if given therapeutically after induction of pancreatitis [89].
However, opposite results were shown by another study in which
FK506 significantly worsened survival in diet-induced murine
pancreatitis [93]. As a result of the obvious controversies even in
the experimental setting and the diverse effects of IL-2 during dif-
ferent stages of acute pancreatitis the general concept of immu-
nomodulation as a potential therapeutic target is yet attractive
but remains inconclusive and is not ready to be transferred to
clinical application.
Platelet-Activating Factor
Platelet-activating factor (PAF) is a lipid that functions as
a proinflammatory cytokine since it induces platelet activa-
tion and aggregation, neutrophil and monocyte activation,
chemotaxis, and vascular effects in terms of vasodilatation and
increased vascular permeability [94]. Upon activation leuko-
cytes, platelets, and endothelial cells are major sources of PAF
release. PAF synthesis and secretion is closely related to TNF-
and IL-1 in a synergistic manner [94].
A pathophysiological implication of PAF in acute pan-
creatitis could be first demonstrated by an Italian group in 1989.
Administration of PAF into the superior pancreaticoduodenal
artery of rabbits induced classical morphologic and biochemi-
cal changes of acute pancreatitis in a dose dependent manner
within 24-72 hours of injection [95]. These exciting observations
have made a number of groups to pursue the role of this novel
cytokine in acute pancreatitis [96-103]. Surprisingly, the course
of PAF levels has never been investigated in human acute
pancreatitis, however, an increase of PAF concentrations in
pancreas, lung, ascites, and plasma was found in experimen-