For reprint orders, please contact email@example.com
Use of antibiotics in severe
Expert Rev. Anti Infect. Ther. 8(3), 317–324 (2010)
Jan J De Waele Infectious complications in severe acute pancreatitis are an important problem and determine
Department of Critical Care outcome in patients who survived the first inflammatory hit of the disease. Timely diagnosis
Medicine, Intensive Care Unit of infected pancreatic necrosis is often challenging, but should not delay adequate treatment,
1K12-C, Ghent University which consists of source control and antibiotic treatment. Prophylactic antibiotics are not
Hospital, De Pintelaan 185, effective in reducing the incidence of (peri)pancreatic infection in patients with severe acute
9000 Ghent, Belgium pancreatitis (or necrotizing pancreatitis). The only rational indication for antibiotics at this
Tel.: +32 9332 2775 moment is documented infection. The spectrum of empiric antibiotics should cover both
Fax: +32 9332 4995
Gram-negative, Gram-positive and anaerobic microorganisms (also keeping in mind exposure
to nosocomial microorganisms), and local ecology should be taken into account. Fungal
infections are often present, and antifungal coverage should be considered, especially if
multiple risk factors for invasive candidiasis are present. Currently, no tools are available to
guide antimicrobial treatment.
Keywords : acute pancreatitis • infected pancreatic necrosis • necrotizing pancreatitis • severe acute pancreatitis
• walled-off necrosis
Acute pancreatitis is an inflammatory disease the disease to its more severe form. So far, the
of the pancreas, that can be either localized or Ranson and Imrie score  (which includes five
affect the whole pancreas. Clinical manifesta- parameters that are evaluated over the first 48 h
tions consist of acute epigastric pain, often asso- after hospitalization), the Acute Physiology and
ciated with referred pain in the back, nausea Chronic Health Evaluation (APACHE) II score
and vomiting. (a severity score often used in critical care based
The process is mainly caused by alcohol on data collected in the first 24 h) and C-reactive
abuse or biliary tract stones  . Other causes protein levels at 48 h  have proven to be the
include trauma (including endoscopic retro- most robust predictors of disease severity, but will
grade cholangiopancreatography), drugs, soon be replaced in clinical practice  . Several
hyperlipemia and viral infections. In a consid- new biomarkers [9,10] and scoring systems [11–14]
erable number (up to 20%) of patients, no clear have been devised to predict outcome at an early
cause can be identified. stage, but few of them have been studied on a
Pancreatitis is a disease with a typically unpre- larger scale. Prediction of severity of disease or
dictable course. Most of the patients develop only other relevant outcome parameters at admission
mild pancreatitis, which is self-limiting, resolves remains very difficult.
within 3–5 days, and carries a low morbidity Treatment of SAP is largely supportive, as
and mortality rate  . There is no necrosis pres- few pharmacological options are available  .
ent, only pancreatic edema, and systemic effects In patients with SAP, infections and other com-
are limited to a systemic inflammatory response plications frequently arise and determine the
syndrome (a combination of fever, tachycardia, further course of the disease. Common local
tachypnea or leukocytosis)  . complications include peripancreatic fluid col-
Some of these patients may progress to severe lections, which may compress adjacent organs
acute pancreatitis (SAP), which is associated and cause biliary obstruction or ileus, pseudo-
with organ failure and/or local complications  . cysts with associated problems such as bleeding
Patients with the fulminant variant of the disease or rupture, and thrombosis of peripancreatic
develop multiple organ dysfunction syndrome veins. Although organ dysfunction, which can
(MODS) within 24–72 h after admission [4,5] . be a problem, particularly early in the course
It remains a challenge to predict progression of of the disease, seems to be the most important
www.expert-reviews.com 10.1586/ERI.10.3 © 2010 Expert Reviews Ltd ISSN 1478-7210 317
Review De Waele
outcome determinant  , infection of necrotic pancreatic paren- Laboratory markers are also of limited use in diagnosis of
chyma and/or peripancreatic structures is the most feared com- infection. Usually, these markers will be elevated, but discern-
plication after the patient has survived the first weeks. Infection ing between infected and noninfected necrosis will rarely be
can occur at any stage of the disease, and should be managed possible based on these findings. The presence of concomitant
by a multidisciplinary team of intensivists, gastroenterologists, infections often blurs the picture. Procalcitonin has been stud-
surgeons, radiologists and infectious disease specialists, keeping ied extensively as an early predictor of subsequent infection  ,
the dynamic nature of the primary process in mind, as this will but its role as a diagnostic tool to confirm the presence of IPN
largely determine the therapeutic strategy  . remains to be elucidated.
The diagnosis of infection is often considered based on clinical
Definitions findings, but in most patients, confirmation through abdominal
In 1992, a committee of international experts laid down a set CT scan is required to confirm this, often combined with fine nee-
of definitions related to different stages of and entities related dle aspiration of suspected areas (Figure 1) . Necrosis may develop
to acute pancreatitis, which have been used both for research not only in pancreatic parenchyma but also in peripancreatic
and in clinical practice  . These definitions were found to have fat and connective tissue. The presence of only limited areas of
a number of limitations and an update is expected in the near pancreatic necrosis does not rule out IPN, and often there is a
future  . Although pancreatic parenchymal necrosis is a straight- relationship between the extent of necrosis and the likelihood
forward diagnosis, it is expected to be extended to the infection of of (subsequent) infection and other outcome variables  . In
peripancreatic tissue. Over time, physicians came to realize that case of suspected infection, diagnostic percutaneous aspiration of
the necrosis of pancreatic tissue is not easily classified in different abdominal fluid at the bedside can be used as a first-line strategy.
clear-cut stages, but that this is a dynamic process: pancreatic Close consultation with the radiologist to guide the fine-needle
necrosis may evolve to an acute postnecrotic collection and even- aspiration is essential, as well as immediate direct examination
tually develop into walled-off necrosis. At each of these different of the fluid obtained in the microbiology laboratory – cultures
stages, the pancreatic substrate may become infected. In the new should be ordered, but not awaited to decide on further therapy.
definitions, the term ‘pancreatic abscess’ will no longer be used, as
it was found that pure abscesses are a rare finding in SAP patients Timing of infection
– to some extent, necrotic parenchyma is present at most stages. Although recent data are often blurred by the use of antibiotics,
whatever the indication, data from prospective studies show that
Infected pancreatic necrosis the incidence of infection peaks in the third to fourth week. In a
Mechanisms of infection recent study of more than 700 patients in The Netherlands, the
Generally, the mechanisms that lead to infected pancreatic authors found that the diagnosis of IPN was made after a median
necrosis (IPN) are poorly understood. Bacterial translocation of 26 (interquartile range: 17–37) days after admission  . This
from the gut is believed to be the most important mechanism, study confirms the findings from an older study, in which increas-
although it has only been documented in animal models  . ing infection was observed from the first week to the third week,
Factors involved may include the ileus and the occurrence of peaking at around the end of the third week  . It should be
intra-abdominal hypertension  – a recently described phenom- added, however, that in some patients infection occurs early, espe-
enon that is also linked to bacterial translocation and exacerbates cially in cases with a history of protracted abdominal pain prior to
decreased perfusion pressure to the gut  . In some patients, hospitalization or prior documented episodes of acute pancreatitis.
direct contamination, for example due to GI tract perforation, In the latter population, infection may occur in areas with fluid
may also be involved. Hematogenous spread from concomitant collections or contained parenchymal necrosis.
infections may occur more frequently than previously thought.
In a recent study, Besselink et al. reported a high incidence of Treatment of IPN
bacteremia and pneumonia in the early stage of SAP  ; in 60% The treatment of IPN consists of both source control and anti-
of patients with both bacteremia and pancreatic infection, the biotic therapy. Although antibiotic therapy has generated much
same organism was isolated at the initial culture. Reflux from discussion, it should be realized that source control is the most
the duodenum is another mechanism that may lead to IPN and important determinant of outcome in most patients. Source
may be especially relevant in patients undergoing endoscopic control consists of all physical measures intended to eliminate
retrograde cholangiopancreatography. a source of infection, to control ongoing contamination, and to
restore premorbid anatomy and function  . When applied to
Clinical features of IPN IPN, elimination of the infectious focus often proves to be dif-
The clinical features of infections are often difficult to distinguish ficult due to the peculiar anatomic relations of the pancreas and
from the systemic inflammatory response syndrome that is present associated necrosis. Source control is based on three principles:
in patients with SAP. These included fever, tachypnea and tachy- drainage, debridement, and restoration of anatomy and function.
cardia. In addition, abdominal signs are often nonspecific, as SAP When applied to IPN in SAP patients, drainage of infected
per se causes severe abdominal pain; surinfection of the inflamed fluid collections and debridement of infected necrotic tis-
or necrotic tissue rarely causes new abdominal signs or symptoms. sue are the most important elements. Adequate drainage of
318 Expert Rev. Anti Infect. Ther. 8(3), (2010)
Use of antibiotics in severe acute pancreatitis Review
fluid collections can be obtained using
CT-guided percutaneous drainage, but
debridement of necrotic tissue requires
more aggressive, often surgical strategies.
Recently, minimally invasive techniques
have been developed and successfully
applied in patients with IPN [25,26] . A trial
comparing an open surgical strategy with
a more conservative approach combining
percutaneous drainage and minimally
invasive techniques in a step-up fashion has
recently been completed, and the results
are eagerly awaited  . The different ele-
ments of obtaining drainage and debride-
ment can be applied at different points in
time. Debridement is easier when necrosis
is ‘organized’ and damage to surround-
ing structures can be minimized. This is
especially relevant later in the course of
the disease, when the end product of pan-
creatic necrosis or ‘walled-off necrosis’ can
become infected, and form into what was
previously often referred to as a ‘pancreatic
abscess’. These collections are particularly
amenable to percutaneous drainage.
Antibiotic options in SAP
When considering antibiotics for patients
with SAP it is important to know the Figure 1. CT-guided fine-needle aspiration of the retroperitoneal area,
suspected for infected pancreatic necrosis based on impacted gas bubbles.
microorganisms that normally cause IPN
and the pharmacokinetics of the antibiotics
in this setting. the incidence of antibiotic-resistant infections in IPN to be as
high as 52%; the majority of these infections were nosocomial
Antibiotic penetration in the pancreas superinfections, occurring after surgery and after prolonged
Adequate penetration of antibiotics in pancreatic tissue is a exposure to broad-spectrum antibiotics  .
prerequisite for the use of antibiotics in this setting. This has Interesting data on this topic have emerged from two blinded
been studied mostly with the prophylactic use of antibiotics in studies on antibiotic prophylaxis. The study by Isenmann et al.,
mind, mostly in animal settings and in non-necrotic pancreatic which compared ciprofloxacin and metronidazole with placebo,
tissue. However, a number of studies have evaluated the pen- found that in the intervention group, 78% of the organisms
etration of antibiotics in pancreatic necrosis; in these studies, that caused infections (including extrapancreatic infections)
carbapenems, fluoroquinolones, piperacillin/tazobactam and a were resistant to ciprofloxacin  . Dellinger et al. found that
number of cephalosporins were found to reach adequate tissue five out of six isolates from pancreatic infection that were tested
levels [28,29] . were resistant to the study drug (meropenem)  . Table 1 sum-
marizes the organisms isolated from the combined placebo
Microbiology of IPN and intervention groups of these studies. From this table, it is
The microorganisms involved in infection of pancreatic necro- clear that the organisms recovered from patients in the placebo
sis are most often enteric Gram-negative bacteria, although an group are different from those recovered from the interven-
increase in Gram-positive infections has been described  . This tion group: notably, more infections with nosocomial Gram-
observation and also observations of the organisms described negative organisms (such as Pseudomonas, Acinetobacter and
in most series, may have been blurred by the extensive use of Enterobacter spp.) were found, and more enterococcal infections
antibiotic prophylaxis, often broad-spectrum antibiotics such as were present in patients who were given antibiotic prophylaxis.
carbapenems or quinolones. It should also be noted that an important number of patients in
The use of these agents may cause a selection of often resistant the placebo group were switched to antibiotics on suspicion of
Gram-positive and Gram-negative organisms, and this problem pancreatic infection or extrapancreatic infection, so the effect
has been increasingly recognized  . Previously, we described of selection may even be underestimated.
Review De Waele
The presumed beneficial effects of pro-
Table 1. Pooled microbiological data from the two randomized
phylactic antibiotics – albeit on question-
controlled trials comparing antibiotic prophylaxis with placebo.
able end points in most studies – have not
Placebo groups Intervention groups been confirmed in the only two blinded
combined (n) combined (n) randomized controlled trials to have been
Total number of patients 106 108 performed recently. The largest trial, com-
Patients with infected pancreatic necrosis 11 (10.4%) 16 (14.8%) paring a combination of ciprofloxacin and
metronidazole with placebo in predicted
Gram-positive microorganisms 10 14
severe pancreatitis, did not show a difference
• Coagulase-negative staphylococci 4 5
• Enterococci 2 7 in incidence of pancreatic infection, extra-
• Staphylococcus aureus 4 1 pancreatic complications or mortality, and
• Streptococcus viridans 0 1 was stopped after an interim analysis  .
Gram-negative microorganisms 5 13 More recently, a large, multicenter, rando-
• Escherichia coli 4 4 mized controlled trial has shown no effect
• Pseudomonas spp. 0 3 of prophylactic meropenem in patients with
• Enterobacter spp. 0 2 SAP  . Whereas it is often suggested that,
• Proteus spp. 1 2 although there is no reduction in mortality,
• Acinetobacter spp. 0 2 the use of prophylactic antibiotics defers
Anaerobes 2 0 infection beyond the second to third week,
Fungi 2 3 there is no evidence for this from the litera-
• Candida albicans 1 3 ture. In the study from Dellinger et al., the
• Candida glabrata 1 0 use of meropenem did not delay the pancre-
Data from [32,33]. atic infection  . In the first study, a con-
siderable number of patients (46%) in the
Indications for antibiotics in SAP control group (that should have no antibiotic), was switched to open
Prophylactic use of antibiotics antibiotic treatment, and therefore is not really a control group. In
Prophylactic antibiotics have been the most intensely debated the Dellinger study, this was not the case.
topic in the treatment of patients suffering from SAP ever since Another frequent observation on the studies performed to date
their first reported use in the 1980s. Although once considered a is that antibiotics are started too late to have any effect on the inci-
life-saving intervention based on a number of small unblinded tri- dence of infection. Apart from the fact that early infection is rare
als, and eagerly adopted by the medical community, this practice (IPN peaks in the third and fourth week after admission), a recent
proved not to demonstrate any benefit in two controlled rand- trial randomizing between early initiation of antibiotics or no
omized trials, the only blinded studies that have been performed antibiotics found no effect on mortality or peripancreatic infec-
to date [32,33] . tion rate, but the overall (pancreatic and nonpancreatic) infection
In the 1980s and 1990s, a number of clinical studies reported rate was reported to be lower  . The total cost of antibiotics in
on the use of different antibiotics to reduce the incidence of the intervention group was approximately double of the cost in the
pancreatic infection. Only one trial demonstrated an effect on placebo groups (€20,400 vs €10,200) and the clinical relevance
mortality. In a small, noncontrolled study with 60 patients, at this point is not clear as any details on the type of infections
Sainio et al. studied the use of cefuroxim and reported a and consequences were reported. A recent study from China in
reduced mortality rate in treated patients when compared with which patients received imipenem within 72 h after the start of
patients who had not received prophylactic antibiotics (1 out symptoms showed no effect on morbidity, the need for surgery
of 30 vs 7 out of 30; p=0.03)  . The incidence of pancreatic and mortality  . The reported pancreatic infection rates in the
infection was no different, however, and the high number of prophylactic antibiotic studies are summarized in Figure 2 .
patients that did receive antibiotics makes interpretation dif- To date, no trial has undeniably shown an effect on mortality
ficult. Peripancreatic coagulase-negative staphylococcal infec- and, in the randomized trials, no effect on pancreatic infections
tions were frequent; in addition, the high number of catheter- was found; in addition, studies with early administration could
related infections suggests problems with intravenous catheter not show any advantage and, therefore, the use of prophylactic
management in these patients. antibiotics – although widely practiced – cannot be supported
Similar studies found no effect on mortality and different end in patients with pancreatic necrosis. However, this remains a
points are used in every paper, which makes comparison very controversial issue and although recommended by some socie-
difficult. Pederzoli et al. compared imipenem with placebo, and ties [39,40] and experts in the field [41,42] , in more recent guidelines
found a decrease in the incidence of pancreatic sepsis, but no effect antibiotic prophylaxis is no longer endorsed [5,43] .
on mortality  . Bassi et al. randomized patients to either treat- When patients develop signs of severe sepsis or septic shock,
ment with pefloxacin or imipenem, and – not surprisingly – also it is rational to empirically administer antibiotics while further
found no difference in outcome  . diagnostic techniques are used to confirm infection. Pneumonia
320 Expert Rev. Anti Infect. Ther. 8(3), (2010)
Use of antibiotics in severe acute pancreatitis Review
and bacteremia in particular appear to
frequently complicate the early course 40
of the disease  , and administration of
antibiotics should not be delayed. If the
Infection rates (%)
suspected infection cannot be confirmed,
antibiotics should be discontinued.
Therapeutic use of antibiotics
The choice of empirical antibiotic treat- 10
ment for any infection should first be
guided by the microorganisms that are
expected to cause the infection. For patients 0
with SAP this may prove to be difficult to Sainio34 Pederzoli35 Isenmann32 Dellinger33 Rokke37 Xue38
judge. The microbiology of IPN seems to
be determined by previous antimicrobial Intervention
treatment, irrespective of the infection that Control
these antibiotics were used for. Recent evi-
dence shows that extra-abdominal infec-
tions are very frequent in the early course Figure 2. Pancreatic infection rates in studies on the use of prophylactic antibiotics.
of the disease, with pneumonia and bac-
teremia occurring much more often in the first 2 weeks after to aim for negative abdominal/drain cultures. The clinical status
admission  . Moreover, as IPN typically occurs only after 1 of the patient is probably one of the best markers for adequate
or 2 weeks of hospitalization – often in a critical care setting – therapy. When the patient’s status is improving and markers
this is a disease where nosocomial organisms are often involved. of inflammation are stable, stopping antibiotic therapy may be
Knowledge of the local epidemiology of the hospital and the safely attempted.
unit is therefore essential to guide antibiotic prescription. This
often results in broad-spectrum empiric Gram-negative and Consequences of antibiotic therapy
Gram-positive coverage, often including extended-spectrum b Apart from the changes in microbiology as described previously,
lactamase (ESBL)-producing and -resistant Gram-positive micro- the increased incidence in fungal infections also seems to be
organisms. Antifungal coverage is indicated in these patients as related to exposure to antibiotics. In the last 10 years, a number
multiple risk factors for invasive candidiasis are often present of studies have reported an incidence of fungal infections in up
and Candida spp. are often isolated from IPN at some stage  . to approximately 50% of patients with IPN [44,48,49] . Some stud-
For obvious reasons, it is pivotal to obtain a sample of the ies have demonstrated that invasive candidiasis is associated with
infected necrotic areas that may allow de-escalation at a later increased mortality  , whereas other studies could not find a
stage. However, it should be considered that nosocomial super- difference  . Several studies have demonstrated a link between
infection is a frequent finding, and colonization at other sites may this and the exposure to often prolonged courses of antibiotic
prompt continued broad-spectrum antibiotic coverage. treatment [44,48] . It should be said, however, that in patients with
Duration of antibiotic treatment is notoriously difficult to IPN, multiple other factors for invasive candidiasis are often
decide. When source control is adequate, duration can be limited present, such as prolonged intensive care unit stay, Candida colo-
to 7–10 days as for severe complicated intra-abdominal infections, nization, presence of central venous catheters, administration of
but as source control may often be incomplete, prolonged therapy total parenteral nutrition and the need for abdominal surgery.
may be necessary. Currently, there is no biomarker that may help SAP is nevertheless an additional risk factor for invasive fungal
to limit the duration of antibiotic treatment. Procalcitonin has infections, and prophylactic treatment with antifungals should be
emerged as a useful tool to guide antibiotic treatment in pneu- considered in patients with infected necrosis who have multiple
monia, and in intra-abdominal infections there is emerging evi- of the above risk factors  . Fluconazole seems to be an adequate
dence that procalcitonin can help to guide the adequacy of the agent for prophylaxis in this setting.
therapeutic strategy (both surgery and antibiotic therapy) [45,46] .
Data for IPN are lacking. Conclusion
After 2–3 weeks of antibiotic treatment, the situation often In conclusion, we propose a rational use of antimicrobials in
mimics the problem of tertiary peritonitis. Tertiary peritonitis patients with SAP in terms of indication, spectrum and duration.
is an ill-defined situation of ongoing peritoneal inflammation Although the principles of rational antibiotic use may be clear, it is
with continued positive microbiology, although it is difficult to often difficult to apply these in clinical practice in the treatment of
discern infection from colonization with associated persistent patients with SAP. The only rational indication to administer anti-
inflammation  . Especially with the use of prolonged abdomi- biotics is documented infection; the use of prophylactic antibiotics
nal lavage through multiple abdominal drains, it may be illusive – although very popular but poorly supported by clinical studies
Review De Waele
– should be avoided. When infection occurs, broad-spectrum will become more easily available on a broader scale. Although the
agents – including nosocomial organisms in cases of prolonged search for strategies to prevent infection will continue to include
exposure – are logical empiric choices. Treatment duration can prophylactic antibiotics, chances are small that the ineffectiveness
best be decided on a patient-to-patient basis – adequacy of source of this strategy will be refuted; enteral nutrition may emerge as
control is an important consideration in this decision. one of the most important mechanisms to prevent IPN. There
are no new drugs in the pipeline that will be particularly suited
Expert commentary to treat these infections, but applying new insights in the phar-
The role of prophylactic antibiotics will be discussed at eternam macokinetics of these agents in critically ill patients will allow
unless appropriately organized studies are undertaken. A number further optimization of antibiotic treatment.
of conditions to make such a study successful should be met before
embarking on such a journey. First of all, selection bias should be Financial & competing interests disclosure
avoided when including patients in the study. There is evidence The author has served as a consultant for Bayer and Weyth. The author has
that, so far, patients for intervention studies have been selectively no other relevant affiliations or financial involvement with any organization
recruited, and that the most severely ill patients (that could have or entity with a financial interest in or financial conflict with the subject
benefited most from an intervention) may not have been included matter or materials discussed in the manuscript apart from those disclosed.
in the studies. The slow recruitment and early termination of one No writing assistance was utilized in the production of this manuscript.
of the randomized controlled studies  could point in that direc-
tion. The second challenge is the recruitment of the patients at the Key issues
highest risk of infection. Better risk stratification using biomarkers • Early risk assessment for late infectious complications is
or other predictors that are only used to predict IPN is imperative. important to diagnose infection early.
Once established infection has been confirmed, both antibiotics • In the first 2 weeks, pneumonia and other infections are more
and tailored source control measures should be implemented as prevalent than infected pancreatic necrosis.
soon as possible, and tailored to the individual patients. When • Infected pancreatic necrosis typically develops 2 weeks or later
selecting empiric antibiotic therapy, this means that the local after admission.
ecology (both on the unit and patient level) should guide the • When suspected, diagnosis of infection should be confirmed and
treatment, and that Gram-negative, Gram-positive, anaerobic and the area of infection sampled for direct examination and culture.
probably also fungal microorganisms should be covered. Tailoring • Source control and antibiotics are the cornerstones of the
is even more important when applying source control measures; treatment of infection. Source control is a prerequisite for the
effectiveness of antibiotic therapy.
important elements such as the exact localization and extent of
the necrosis, as well as the consistency of the pancreatic necrosis • Prophylactic antibiotics do not reduce pancreatic infection rate or
mortality in patients with predicted severe disease or
will determine the most optimal strategy. Minimally invasive necrotizing pancreatitis.
techniques are logical first steps, but formal surgery should not
• Documented infection is the sole indication for antibiotics in
be delayed when these prove to be ineffective. patients with pancreatitis.
• Antibiotic treatment should cover enteric bacteria, including
Five-year view Gram-positive, Gram-negative and anaerobic organisms.
Early risk stratification, but more importantly early diagnosis of • Treatment duration should be decided on an individual basis.
IPN, will continue to be improved, most certainly with a role for When source control is adequate, a course of 7–10 days should
new biomarkers such as procalcitonin or different interleukins that be sufficient.
References on Acute Pancreatitis, Atlanta, GA, 6 Ranson JH, Rifkind KM, Roses DF et al.
Papers of special note have been highlighted as: September 11 through 13, 1992. Arch. Prognostic signs and the role of operative
• of interest Surg. 128(5), 586–590 (1993). management in acute pancreatitis. Surg.
•• of considerable interest Gynecol. Obstet. 139(1), 69–81 (1974).
4 Isenmann R, Rau B, Beger HG. Early
1 Pandol SJ, Saluja AK, Imrie CW et al. severe acute pancreatitis: characteristics of 7 Wilson C, Heads A, Shenkin A et al.
Acute pancreatitis: bench to the bedside. a new subgroup. Pancreas 22(3), 274–278 C-reactive protein, antiproteases and
Gastroenterology 132(3), 127–151 (2007). (2001). complement factors as objective markers
•• Comprehensive review of acute 5 Nathens AB, Curtis JR, Beale RJ et al. of severity in acute pancreatitis.
pancreatitis, from basic pathophysiology Management of the critically ill patient Br. J. Surg. 76(2), 177–181 (1989).
to surgical management. with severe acute pancreatitis. Crit. Care 8 Mofidi R, Patil PV, Suttie SA et al. Risk
2 Mitchell RM, Byrne MF, Baillie J. Med. 32(12), 2524–2536 (2004). assessment in acute pancreatitis. Br.
Pancreatitis. Lancet 361(9367), •• Consensus document discussing the J. Surg. 96(2), 137–150 (2009).
1447–1455 (2003). most relevant issues in the management • Timely review of risk stratification tools
3 Bradley EL 3rd. A clinically based of severe acute pancreatitis from a in acute pancreatitis.
classification system for acute pancreatitis. critical care perspective.
Summary of the International Symposium
322 Expert Rev. Anti Infect. Ther. 8(3), (2010)
Use of antibiotics in severe acute pancreatitis Review
9 Stimac D, Fisic E, Milic S et al. 20 De Waele JJ, Leppaniemi AK. Intra- 31 De Waele JJ, Vogelaers D, Hoste E et al.
Prognostic values of IL-6, IL-8, and IL-10 abdominal hypertension in acute Emergence of antibiotic resistance in
in acute pancreatitis. J. Clin. pancreatitis. World J. Surg. 33(6), infected pancreatic necrosis. Arch. Surg.
Gastroenterol. 40(3), 209–212 (2006). 1128–1133 (2009). 139(12), 1371–1375 (2004).
10 Mofidi R, Suttie SA, Patil PV et al. The 21 Besselink MG, van Santvoort HC, 32 Isenmann R, Runzi M, Kron M et al.
value of procalcitonin at predicting the Boermeester MA et al. Timing and impact Prophylactic antibiotic treatment in patients
severity of acute pancreatitis and of infections in acute pancreatitis. Br. J. with predicted severe acute pancreatitis: a
development of infected pancreatic Surg. 96(3), 267–273 (2009). placebo-controlled, double-blind trial.
necrosis: systematic review. Surgery • Large study on the timing of pancreatic Gastroenterology 126(4), 997–1004 (2004).
146(1), 72–81 (2009). and extrapancreatic infections in • Randomized controlled study on the use
11 Ueda T, Takeyama Y, Yasuda T et al. acute pancreatitis. of prophylactic antibiotics.
Utility of the new Japanese severity score Dellinger EP, Tellado JM, Soto NE et al.
22 Isenmann R, Rau B, Beger HG. Bacterial 33
and indications for special therapies in Early antibiotic treatment for severe acute
infection and extent of necrosis are
acute pancreatitis. J. Gastroenterol. 44(5), necrotizing pancreatitis: a randomized,
determinants of organ failure in patients
453–459 (2009). double-blind, placebo-controlled study.
with acute necrotizing pancreatitis. Br. J.
12 Ueda T, Takeyama Y, Yasuda T et al. Surg. 86(8), 1020–1024 (1999). Ann. Surg. 245(5), 674–683 (2007).
Simple scoring system for the prediction
23 Beger HG, Bittner R, Block S et al. • Randomized controlled study on the use
of the prognosis of severe acute
Bacterial contamination of pancreatic of prophylactic antibiotics.
pancreatitis. Surgery 141(1), 51–58
necrosis. A prospective clinical study. 34 Sainio V, Kemppainen E, Puolakkainen P
Gastroenterology 91(2), 433–438 (1986). et al. Early antibiotic treatment in acute
13 Harrison DA, D’Amico G, Singer M. The
24 Schein M, Marshall J. Source control for necrotising pancreatitis. Lancet
Pancreatitis Outcome Prediction (POP)
surgical infections. World J. Surg. 28(7), 346(8976), 663–667 (1995).
Score: a new prognostic index for patients
638–645 (2004). 35 Pederzoli P, Bassi C, Vesentini S et al.
with severe acute pancreatitis. Crit. Care
Med. 35(7), 1703–1708 (2007). • Comprehensive review of the basics of A randomized multicenter clinical trial of
source control and the antibiotic prophylaxis of septic
14 Singh VK, Wu BU, Bollen TL et al.
practical application. complications in acute necrotizing
A prospective evaluation of the bedside
Becker V, Huber W, Meining A et al. pancreatitis with imipenem. Surg.
index for severity in acute pancreatitis 25
Infected necrosis in severe pancreatitis – Gynecol. Obstet. 176(5), 480–483 (1993).
score in assessing mortality and
intermediate markers of severity in acute combined nonsurgical multi-drainage with 36 Bassi C, Falconi M, Talamini G et al.
pancreatitis. Am. J. Gastroenterol. 104(4), directed transabdominal high-volume Controlled clinical trial of pefloxacin
966–971 (2009). lavage in critically ill patients. Pancreatology versus imipenem in severe acute
9(3), 280–286 (2009). pancreatitis. Gastroenterology 115(6),
15 Mofidi R, Duff MD, Wigmore SJ et al.
van Santvoort HC, Besselink MG, 1513–1517 (1998).
Association between early systemic 26
inflammatory response, severity of Horvath KD et al. Videoscopic assisted 37 Rokke O, Harbitz TB, Liljedal J et al.
multiorgan dysfunction and death in retroperitoneal debridement in infected Early treatment of severe pancreatitis with
acute pancreatitis. Br. J. Surg. 93(6), necrotizing pancreatitis. HPB (Oxford) imipenem: a prospective randomized
738–744 (2006). 9(2), 156–159 (2007). clinical trial. Scand. J. Gastroenterol.
Besselink MG, van Santvoort HC, 42(6), 771–776 (2007).
16 De Waele J, Vogelaers D, Decruyenaere J 27
et al. Infectious complications of acute Nieuwenhuijs VB et al. Minimally invasive 38 Xue P, Deng LH, Zhang ZD et al. Effect
pancreatitis. Acta Clin. Belg. 59(2), 90–96 ‘step-up approach’ versus maximal of antibiotic prophylaxis on acute
(2004). necrosectomy in patients with acute necrotizing pancreatitis: results of a
necrotising pancreatitis (PANTER trial): randomized controlled trial.
17 Bollen TL, van Santvoort HC,
design and rationale of a randomised J. Gastroenterol. Hepatol. 24(5), 736–742
Besselink MG et al. The Atlanta
controlled multicenter trial (2009).
Classification of acute pancreatitis
revisited. Br. J. Surg. 95(1), 6–21 (2008). [ISRCTN38327949]. BMC Surg. 6, 6 39 Toouli J, Brooke-Smith M, Bassi C et al.
(2006). Guidelines for the management of acute
•• Critical review of the use of the
28 Otto W, Komorzycki K, Krawczyk M. pancreatitis. J. Gastroenterol. Hepatol.
1992 Atlanta classification in 17(Suppl.), S15–S39 (2002).
Efficacy of antibiotic penetration into
pancreatic necrosis. HPB (Oxford) 8(1), 40 Uhl W, Warshaw A, Imrie C et al. IAP
18 Cicalese L, Sahai A, Sileri P et al. Acute 43–48 (2006). Guidelines for the surgical management
pancreatitis and bacterial translocation. of acute pancreatitis. Pancreatology 2(6),
29 Adam U, Herms S, Werner U et al. The
Dig. Dis. Sci. 46(5), 1127–1132 (2001). 565–573 (2002).
penetration of ciprofloxacin into human
19 Strobel O, Wachter D, Werner J et al. pancreatic and peripancreatic necroses in 41 Baron TH, Morgan DE. Acute
Effect of a pneumoperitoneum on acute necrotizing pancreatitis. Infection necrotizing pancreatitis. N. Engl. J. Med.
systemic cytokine levels, bacterial 29(6), 326–331 (2001). 340(18), 1412–1417 (1999).
translocation, and organ complications in
30 Howard TJ, Temple MB. Prophylactic 42 Dervenis C, Johnson CD, Bassi C et al.
a rat model of severe acute pancreatitis
antibiotics alter the bacteriology of infected Diagnosis, objective assessment of
with infected necrosis. Surg. Endosc.
necrosis in severe acute pancreatitis. J. Am. severity, and management of acute
20(12), 1897–1903 (2006).
Coll. Surg. 195(6), 759–767 (2002).
Review De Waele
pancreatitis. Santorini consensus 47 Nathens AB, Rotstein OD, Marshall JC. 51 Gloor B, Muller CA, Worni M et al.
conference. Int. J. Pancreatol. 25(3), Tertiary peritonitis: clinical features of a Pancreatic infection in severe pancreatitis:
195–210 (1999). complex nosocomial infection. World the role of fungus and multiresistant
43 Banks PA, Freeman ML. Practice J. Surg. 22(2), 158–163 (1998). organisms. Arch. Surg. 136(5), 592–596
guidelines in acute pancreatitis. Am. J. 48 Vege SS, Gardner TB, Chari ST et al. (2001).
Gastroenterol. 101(10), 2379–2400 Outcomes of intra-abdominal fungal vs. 52 Blot SI, Vandewoude KH, De Waele JJ.
(2006). bacterial infections in severe acute Candida peritonitis. Curr. Opin. Crit.
44 De Waele JJ, Vogelaers D, Blot S et al. pancreatitis. Am. J. Gastroenterol. 104(8), Care 13(2), 195–199 (2007).
Fungal infections in patients with severe 2065–2070 (2009).
acute pancreatitis and the use of 49 Kochhar R, Ahammed SK, Chakrabarti A
prophylactic therapy. Clin. Infect. Dis. et al. Prevalence and outcome of fungal Affiliation
37(2), 208–213 (2003). infection in patients with severe acute • Jan J De Waele, MD, PhD
45 Novotny AR, Emmanuel K, Hueser N pancreatitis. J. Gastroenterol. Hepatol. Department of Critical Care Medicine,
et al. Procalcitonin ratio indicates 24(5), 743–747 (2009). Intensive Care Unit 1K12-C, Ghent
successful surgical treatment of 50 Hoerauf A, Hammer S, Muller-Myhsok B University Hospital, De Pintelaan 185,
abdominal sepsis. Surgery 145(1), 20–26 et al. Intra-abdominal Candida infection 9000 Ghent, Belgium
(2009). during acute necrotizing pancreatitis has Tel.: +32 9332 2775
a high prevalence and is associated with Fax: +32 9332 4995
46 Lepouse C, Murat O, Nicolai F et al.
increased mortality. Crit. Care Med. firstname.lastname@example.org
Postoperative procalcitonin kinetics: an
indicator for therapeutic strategy in 26(12), 2010–2015 (1998).
peritonitis? Intensive Care Med.
34(Suppl. 1), S123 (2008).
324 Expert Rev. Anti Infect. Ther. 8(3), (2010)