Nutritional Support in Acute Pancreatitis
What are the Key Issues?
Stephen A. McClave, MD
Professor of Medicine
University of Louisville School of Medicine
Louisville, Kentucky
Objectives

To know how our perspective toward nutritional support in
acute pancreatitis in the past differs from that of today.

To learn which factors are important in promoting tolerance
to artificial nutritional support.

To understand how to utilize route, timing, dose, and
content to optimize outcome from nutritional therapy.
Objectives
•
•
•
To understand the benefits of enteral nutrition (EN) in acute
pancreatitis and the timing of the “window of opportunity”
during which feeds should be started.
To learn the risks and consequences of inadvertently
stimulating the pancreas with EN in acute pancreatitis.
To evaluate the benefits and compare actual experience of
gastric feeding with jejunal feeds in acute pancreatitis.
Introduction

Benefit of early EN on disease process and on patient outcome
is dramatic

Consequences of inadvertent pancreatic stimulation minimal
With vigilence, little chance of doing net harm

Any signs of symptom exaccerbation or inflamation in
response to EN ameliorated by subtle adjustments in
feeding strategy
Introduction

Narrow window of opportunity possibly 48-72 hrs
Potential for EN to ↓disease severity, ↓complications
Delays may result in loss of chance for EN to improve outcome

Vast majority of severe pancreatitis patients may tolerate feeds
Minimizing duration of ileus may improve tolerance

Dropping a Dobhoff NG tube simplest most expedient strategy
Attains rapid access
Quickens time to initiation of feeds
Involves minimal expertise
Fascilitates delivery of EN
Controversial Study
Gastric Feeds in Severe
Acute Pancreatitis

Eatock PRCT nasogastric vs nasojejunal
Glasgow
Severe pancreatitis (AP II >6, 25% mort)
EN initiated within 72 hrs onset pain
Reached goal feeds in mean 36 hrs
NG
No significant differences:
CRP levels
Pain scores
AP II scores Mortality
Hosp LOS
Days to PO

NJ
Conclusion: NG feeds can be considered as therapeutic option
Amer J Gastro 2005 Feb
Double-Edged Sword
Reduce Stress with EN
(Gut Integrity)
Increase Stress with EN
(Pancreatic Stimulation)
•
•
•
•
Why are we asking for trouble?
What is the risk of stimulating the pancreas ?
What is the benefit from providing EN ?
How strong is the evidence for benefit from EN ?
Pancreatic Rest: What Does it Mean?
Reduced level of stimulation that allows resolution of inflamation
Basal versus subclinical output
Poor management strategy alone
Clinical guidance by symptoms
Feedback monitor for tolerance
Benefit of Providing EN





Maintain gut integrity (Less bacterial challenge, endotoxemia)
Set tone for systemic immunity (Oppose Th1 thru Th2 stimulation)
Attenuate stress response, disease severity (CRP, glucose, TAC)
Faster resolution of disease process (Duration SIRS, Nutrit Rx, LOS)
Fewer complications (Infection, surgical intervention, possibly MOF)
EN
Impact on Outcome
Parameters
Infection
PN
by 52%
McClave SA, Chang WK, Heyland DK, Dhaliwal R (JPEN In Press)
Impact on Outcome
Parameters
Hospital LOS by 3.94 Days
McClave SA, Chang WK, Heyland DK, Dhaliwal R (JPEN In Press)
Impact on Outcome Parameters
Organ Failure (MOFS)
MOFS
EN
12.9%
(13/101)
PN
26.7%
(32/120)
by 41%
Signif
p=0.18
RR=0.59
McClave SA, Chang WK, Heyland DK, Dhaliwal R (JPEN In Press)
Impact on Outcome
Parameters
Need for Surgical
Intervention by 52%
PE Marik, GP Zaloga (BMJ 2004;328:1407)
EN vs No Nutrition Rx
Initial Admissions
(
) = EN (n=13)
(
) = Stand (n=14)
IL- 6
CRP
TNF
•
Powell (Brit J Surg 2000;87:1375)
EN vs No Nutrition Rx
Post-op for Complications of
Acute Pancreatitis
Mortality
by 74%
Mortality
EN
STD
p = 0.06
McClave SA, Chang WK, Heyland DK, Dhaliwal R (JPEN In Press)
Consequences of
Providing EN
Warning: Early advance
to oral diet will increase
late complications
(abdominal abscess)
Ranson (Surg 1997;82:99)
Three potentially adverse scenarios result from EN provision
to patients with acute severe pancreatitis
Shhhhh!!
First Scenario:
Silent Stimulation
of Secretion
•
•
Example: Patients from O’Keefe study
Occurs in 100% of patients
(iu/h)
TrypsinAmylase
Healthy vol EN
Healthy vol PN
439(27)
266(49)*
Patients EN
209(33)
Patients PN
34(5)‡
( ‡ p<0.05, * p<0.005 )
1 O’Keefe
Lipase
11,791(1106)
1,064(272)*
9,165(3787)
674(137) ‡
(Gastro 2003;122:A34)
610(61)
76(14)*
235(73)
13(2)‡
Second Scenario: Exacerbation of Symptoms
Example: Jejunal Formula versus PO Clear Liquids 1
Uncomplicated exacerbation of sx in 21.0% 2
1
McClave (JPEN 1997;21:14)
2
Levy (Gut 1997;40:262)
Third Scenario: Exacerbation of Disease Process
Example: Jejunal versus Gastric Infusion 1
WBC Count
Lipase
Amylase
Exacerbation of disease process in 4.3% 2
1 McClave
(JPEN 1997;21:14)
2
Levy (Gut 1997;40:262)
Who Needs Nutritional Rx?
Correlation to Disease Severity
Intestinal Permeability
(% Urinary Excretion PEG 3350)
Controls
0.009
Mild
0.008
Severe
No MOFS
MOFS
0.040 *
0.160 *
Ammori ( J Gastrointest Surg 1999;3:252 )
* p<0.001
Who Needs Nutritional Rx ?
APACHE II < 9
Rans Crit < 2
Degree of panc
CT scan
Mortality
Complications
PO diet in 7d
Management
mild/mod
no necrosis
0%
6%
81%
supportive
APACHE II > 10
Rans Crit > 3
severe
necrosis
19%
38%
0%
EN/PN
Not exclusions: Necrosis, pseudocyst, ascites, surgery
Exclusions: Intolerance
Sax (Amer J Surg 1987;153:117) Wilson (Brit J Surg 1990;77:1260)
Agarwal (Amer J Gastro 1991;86:1385)
Identifying Patients
with Severe Pancreatitis
He looks
good to
me…
Clinical Assessment
APACHE II Score >9
Ranson Criteria >2
Admission
Sensitivity
48-72 Hours
Sensitivity
34-44%
63%
N/A
44-66%
75-82%
75%
Sensitivity higher for biliary >> ethanol etiologies
Wilson (Brit J Surg 1990;77:1260)
Larvin (Lancet 1989;2:201)
Who Needs Nutritional Rx?
Correlation to Disease Severity
McClave1
(n=32)
* p<0.05
% Severe
Attenuate stress
19%
RC
Glucose
Time to Resolution
% Complications
% Septic Complications
1JPEN
Windsor2
(n=34)
38%
CRP*
AII*
SIRS *
Abou-Assi 3 Kalferentos4
(n=33)
(n=38)
35%
11.8
100%
6.2*
1997;21:14 2 Gut 1998;42:431
3 Gastro 2001;120:A469
4 Brit J Surg 1997;84:1665
75%
50%
44% *
28% *
Formula
Selection
Study Pts
•
•
•
•
1 Tiengou
Controls
Small peptide/MCT formula (n=30) 1
Hosp LOS
23.0d*
27.0d
Fish oil formula
Hosp LOS
Durat EN
Complics
(n=28) 2
13.1d*
10.6d*
42%
19.3d
17.6d
64%
Arginine/fish oil formula (n=15) 3
ICU LOS
8.6d
34.8d
Hosp LOS
27.2d
38.4d
Clinical Significance
Below Lig Treitz – Tolerate STD
Gastric – Content is tolerance factor
Pharmaconutrit – Fears of SIRS
(JPEN 2006;30:1) 2 Lasztity (Clin Nutrit 2005;24:198)
3 Hallay (Hepatogastroent 2001;48:1488)
*p<0.05
Gastric vs Jejunal Feeding

Time to initiation of EN signif less for gastric feeds 1
Gastric initiated mean 16 hrs (range 12-20 hrs) earlier
Eventually post-pyloric feeds “catch up” (time to goal, % goal)

Track record of intragastric feeding in acute pancreatitis is good!
McClave Louisville Study 2
One patient jejunal-gastric displacement - SIRS
Responded immediately to replacement back to jejunum
Eatock Glasgow Study 3
70.4% Tolerated >75% goal kcal within 48 hrs (vs 77.2% NJ)
Pain in 2/27 - No ▲infus rate, CRP, AII scores, analgesia
Kumar Indian Study 4
One patient each group experienced pain (no ▲amylase)
Partial PN required first week only in 6 NG (4NJ)
1Crit
Care 2003;7:R46
2
JPEN 1997;21:14
3
AJG 2005;100:432
4
J Clin Gastr 2006;40:431
Potential Changes in Strategy
In Response to Intolerance
NG
NJ

Divert level of EN infusion lower in GI tract
Infusion >40cm below Ligament Treitz no stimulation 4

Change content of formula to ↓stimulation
Louisville study – Peptamin (33% fat) 1
Glasgow study – Pepti-2000 LF (9% fat) 2
Indian study – Peptamin (33% fat) 3
1 JPEN
1997;21:14 2 AJG 2005;100:432 3 J Clin Gastro 2006;40:431
4 O’Keefe (Gastro 2003;122:A34)
Change in Content
Aspirate
Isolated duodenal fistula2
Acute Pancreatitis1
Volume Bicarb
Vivonex
Criticare
Osmolite
+27%
0%
-7%
Feed
-24%
-21%
-65%
Amylase Lipase
-62%
-25%
-84%
+4461%
+1317%
+21,283% *
( * p<0.05) 1Parekh (S African J Surg 1993;31:57) 2Grant (JPEN 1987;11:302)
What Factors
Affect Tolerance?

Level of infusion

Content

Duration of ileus

Institutional experience and expertise

Individual variation
Tolerance
Effect of Duration of Ileus
•
Prospective non-randomized series of 102 acute pancreatitis pts 1
Subset
Duration of Ileus
Achieve Tolerance EN
Group 1 (n=11)
Group 2 (n=8)
Group 3 (n=83)
•
> 6 days
< 5 days
< 2 days
0% (PN)
50%
92%
Early onset feeding EN within 48 hours thru 2 studies: 2
Maintains gut function, improves tolerance
Fewer problems with ileus, gastric stasis
1
Cravo (Clin Nutrit Suppl 1989;8:14)
2 Eatock
(AJG 2005;100:432)
Tolerance: Institutional
Experience and Expertise

Windsor Study in 34 patients PN vs EN 1
Some degree of ileus in 5/16 on EN
Required decreased rate for 2 to 4 days

Schneider Study in 69 ICU pts on EN protocol (prospective)
Mean APACHE II = 18 (range 4-40)
Results
% Total Patients
Mortality
EN alone
EN/PN
PN Alone
None
25% (17/69)
28% (19/69)
14% (10/69)
33% (23/69)
24%
60%
-
1Gut
1998;42:431
2BritJSurg
2
2000;87:362
Tolerance: Individual
Variation

Intolerance of gastric feeds involving
a single patient in McClave Study 1

Nasogastric feeds tolerated in 27 pts as well
as nasojejunal feeds in Eatock Study 2

Intolerance of nasojejunal feeds
Louisville case - 10cm below Lig of Treitz
Richmond case - Exaccerbation on NJ feeds 3
1
McClave (JPEN 1997;) 2 Eatock, Imrie (Gastro 2001;120:A469)
3 O’Keefe (Clin Gastro Hepat 2003;1:315)
Window of Opportunity
Early vs Delayed Enteral Nutrition
Two meta-analyses: Early (<36 hrs) vs delayed (>36 hrs) EN
Infection reduced 55% (p=0.0006) 1
Hospital LOS shortened 2.2 days (p=0.0004) 1
Mortality decreased 48% (p=0.08) 2
1
Marik (CCM 2001;29:2264)
2
Heyland (JPEN 2003;27:355)
Window of Opportunity
Early vs Delayed EN in Pancreatitis

Six PRCTs EN vs PN randomized/feeding within 48hrs
Five showed impact on outcome:
Infectious morbidity ↓ (Abou-Assi, Kalfarentzos, Olah)
Shorter hosp LOS (Gupta)
Less overall complications (Kalfarentzos)
Duration dz process ↓ , nutrit Rx ↓ (Abou-Assi, Gupta)
Faster resolution SIRS (Windsor)
One showed no effect on outcome
McClave (mean Ranson Criteria 1.1)

One PRCT EN vs PN randomized/feeding after 4 days (Louie)
Mean Ranson Criteria 4.7-5.0
No effect on any clinical outcome parameters
McClave (JPEN 2006;30:143)
Is PN a Dead Issue?

Early TPN may be a liability

PRCT in 54 pancreatitis patients 1
Controls
Early TPN
LOH
10 days
16 days *
Cath sepsis
1.5%
10.5% *
Complications
no difference
•
Up to 47% pts may still need TPN 2
1Sax
(Amer J Surg 1987;153:117)
2Schneider (Gut 1998;42:431) *p<0.05
Options for Nutrition Support
in the Individual Patient
Options in acute
pancreatitis based on:
Disease severity
Timing
Tolerance
Standard Rx
(Do nothing)
EN
PN
Priorities of “Nutritional Management”
100
(
Benefit
(%)
) = Immune
Modulation
50
(
) = Protein/calorie Provision
0
1
2
3
4
Time (days)
5
6
7
Is PN a Dead
Issue?
Xian-Li PRCT of PN in “severe” acute pancreatitis (after resuscitation)
Mortality
Complications
Panc infection
Hosp LOS (d)
Group I
STD (n=23)
Group II
PN (n=21)
Group III
PN/Glut (n=20)
43.5%
21
8
39.1+10.6
14.3% *
11 *
5
28.6+6.9*
0.0% *
4#
0 *#
25.3+7.6*
(*p<0.05 Groups II or III vs Group I)
(#p<0.05 Group III vs Group II)
Clin Nutrit Suppl 2004;1:43
Management Algorhythm
Place NG in ER
Start Peptide/MCT Feeds
At 48 Hrs – RC/APACHE II
Mild to Moderate Dz
<2RC, <9AII
Severe Dz
>3RC, >10AII
Tolerates NG EN
NG EN Intolerance
Switch to NJ feeds
Start PN if intolerant
> 5 days
Advance to Oral Clear Liquids
Conclusions
Gold Standard