Transcript Septic Shock - Calgary Emergency Medicine

Septic Shock in 2004
Emergency Department
Strategies for Reducing Mortality
Moritz Haager PGY-3
May 13, 2004
Objectives


Brief review of basic pathophysiology
Overview of recent advances in treatment with focus
on those most relevant to ED care

Initial resuscitation
 Early goal directed therapy



Infection & source control
 Role of steroids
 Ventilatory strategies
Adjunctive pharmacologic therapies
 Activated Protein C
 Insulin therapy


Which fluid?
Which pressor?
Blood transfusions?
Epidemiology

Incidence variable but on the rise






~ 1/1000 – 260/1000 pts days
Larger # of elderly, HIV, chemotherapy, organ transplant,
and dialysis pts in addition to diabetics, alcoholics etc
Mortality ranges from 3% for pts w/ no SIRS criteria
to 46% for septic shock
Accounts for 215,000 deaths/yr in US = MI deaths or
9.3% of all deaths in 1995
Annual est cost in US $16.7 billion
Locally ~250 ICU admissions for sepsis per year
Definitions

Old

SIRS

T >38 or <36
HR > 90
RR > 20 or PCO2 < 32
WBC > 12 or < 4

Too sensitive & simplistic



Latest ACCP/SCCM
Consensus Definitions




Infection = invasion of organ system(s) by
microorganisms
Sepsis = systemic host response to infection
requiring > 1 signs & symptoms of sepsis
Severe sepsis = sepsis w/ organ failure
Septic shock = severe sepsis w/
cardiovascular failure requiring vasoactive
medications

Vincent & Jacobs. Curr Opin Infect Dis 16: 309-13. 2003
Vincent & Jacobs. Curr Opin
Infect Dis 16: 309-13. 2003
Classifications reflect disease
severity
Diagnostic category
SIRS criteria
none
2
3
4
Sepsis
Severe sepsis
Septic shock

Mortality (%)
3
7
10
17
16
20
46
McCoy & Matthews. Drotrecogin Alfa (Recombinant Human
Activated Protein C) for the treatment of severe sepsis. Clin
Ther 2003; 25: 396-421
PIRO Grading System

P – Predisposing factors


I – Infection


Organism, site(s), degree
R – Response


Age, comorbidities, immune status etc
Degree of host response as judged by clinical & laboratory
parameters
O – Organ dysfunction

Degree of organ involvement
Sepsis Etiology


> 90% bacterial etiology
 Gram negative ~42%
 Gram positive ~34%
 Anaerobes ~2-5%
 Mixed ~14%
Fungi ~5%
 Primarily Candida
 More common in ICU setting,
immunocompromised pts, steroids, diabetics


Llewelyn & Cohen. Diagnosis of infection in sepsis. Int Care Med. 2001; 27: S10-32
Bochd, Glauser, & Calandra. Antibiotics in sepsis. Int Care Med. 2001; 27: S33-48
Sources of Infection

Specific sites:








Respiratory
Blood
Abdomen
Urinary tract
Wounds & Skin
Other
36%
20%
19%
13%
7%
5%
Can be identified in ~92% of pts
Extremely important in choosing Abx

Bochd, Glauser, & Calandra. Antibiotics in sepsis. Int Care
Med. 2001; 27: S33-48
Not all that is febrile & shocky
is infectious…

Non-infectious causes of SIRS
 Tissue damage


Metabolic


Tumor lysis syndrome, lymphoma
CNS


Thyroid storm, adrenal insufficiency
Malignancy


Surgery, trauma, DVT, MI, PE, pancreatitis etc
SAH
Iatrogenic

Transfusion rxns, anesthetics, NMS etc

Llewelyn & Cohen. Diagnosis of infection in sepsis. Int Care
Med. 2001; 27: S10-32
INFLAMMATION
PATHOGENS
TNF-α, IL-1, IL-6, IL-7,
Proteases, Leukotrienes,
Prostaglandins
Bradykinin, Platelet activating
factors
Free oxygen radicals
Endotoxins, Exotoxins
Direct endothelial invasion
SEPSIS
TF EXPOSURE
ENDOTHELIAL INJURY
ANTI-COAGULANT
SYSTEM INHIBITON
↓ AT III, ↓ aPC, ↓ pS
↓ thrombomodulin
ACTIVATION
OF CLOTTING
CASCADE
FIBRINOLYTIC
SYSTEM INHIBITION
↑ PAI-1
PRO-COAGULANT EFFECT
MICROVASCULAR THROMBOSIS
MULTI ORGAN DYSFUNTION SYNDROME
A tale of 2 theories

Hyperimmune response theory



Sepsis is a state of uncontrolled inflammatory
response to infection
Multiple (unsuccessful) trials of anti-inflammatory
agents
Hypoimmune response theory

Sepsis leads to immunosuppression through
anergic & apoptotic mechanisms

Hotchkiss & Karl. The pathophysiology and treatment of
sepsis. N Eng J Med. 2003; 348: 138-50
Clinical Effects

Peripheral vasodilatation & capillary leak




Intravascular volume depletion
Myocardial depression
Hypermetabolic state – global tissue hypoxia
DIC – coagulation > fibrinolysis
Treatment of Septic Shock
Antibiotics
Infection
Control
Surgical Management
Septic
Shock
Steroids
rhAPC
Immunomodulatory
Therapies
Supportive
Care
EGDT
Ventilation
SCCM Guidelines for
Treatment of Septic Shock

Utilize






EGDT in 1st 6 hrs
Cultures before Abx
Source control
Aggressive rehydration with
colloid or crystalloid
Use dopamine or

norepinephrine for
refractory shock
Give stress dose steroids




Give rhAPC when
appropriate
Keep Hb 70-90
Use low TV’s & minimal
peak pressure & PEEP
vent strategy
Use insulin therapy
Avoid


Supranormal oxygenation
Bicarb

Dellinger et al. Surviving sepsis
campaign guidelines for management of
severe sepsis and septic shock. Crit
Care Med. 2004; 32: 858-73
Early Goal Directed
Therapy
SCCM Guidelines


“resuscitation…should not be delayed pending ICU
admission.”
Goals of resuscitation in 1st 6 hrs of recognition: (B)
 CVP: 8-12 mm Hg (12-15 if ventilated)
 MAP: > 65 mm Hg
 Urine output: > 0.5 ml/kg/hr
 SVO2 > 70%
 If unable to attain SVO2 >70% despite above then:


Transfuse to keep Hct > 30%
Dobutamine

Dellinger et al. Surviving sepsis campaign guidelines for
management of severe sepsis and septic shock. Crit Care Med.
2004; 32: 858-73
Rationale behind EGDT

Time is survival:




Goal is to achieve balance b/w O2 delivery & consumption
Standardized approaches to ED Tx have improved
outcomes in other Dz (e.g. MI)
Traditional parameters to guide resus (vitals, mental
status, urine output) appear to be too insensitive for
ongoing tissue hypoxia
Early observational trials found survivors to have
hemodynamic parameters that were both higher
than non-survivors as well as predicted
Earlier Trials

No consistent benefit from using goal-directed
therapy to optimize oxygen delivery in ICU
patients






Gattinoni et al. A trial of goal-directed hemodynamic therapy in
critically ill patients. N Eng J Med 1995; 333: 1025-32
Hayes et al. Elevation of systemic oxygen delivery in the treatment
of critically ill patients. N Eng J Med 1994; 330: 1717-22
Yu et al. Effect of maximizing oxygen delivery on morbidity and
mortality rates in critically ill patients: a prospective randomized
controlled study. Crit Care Med. 1993; 21: 830-8
Boyd et al. A randomized clinical trial of the effect of deliberate
perioperative increase of oxygen delivery on mortality in high-risk
surgical patients. JAMA. 1993; 270: 2699-707
Tuchschmidt et al. Elevation of cardiac output and oxygen delivery
improves outcome in septic shock. Chest 1992; 102: 216-20
Shoemaker et al. prospective trial of supranormal values of survivors
as therapeutic goals in high-risk surgical patients. Chest 1988; 94:
1176-86
Earlier Trials

Limitations:





Heterogeneous study populations
Small sample sizes & wide CI’s
Enrollment after ICU admission
Tended to focus on one intervention in
isolation
Most used PA catheters
Rivers et al. Early goal-directed therapy in
the treatment of severe sepsis and septic
shock. N Eng J Med. 2001; 345: 1368-77




Prospective RCT of 263 adult pts with sepsis
treated either with traditional care or a
standardized resuscitation protocol in the ED
All had arterial & central venous lines placed
– the EGDT group got a catheter capable of
continuous O2 sat measurement
EGDT discontinued once transferred to ICU –
all ICU staff blinded to pts assignments
Primary endpoint was mortality
Edwards PreSep Central
Venous Oximetry Catheter
EGDT Protocol
Rivers et al. 2001 (cont’d)

Found that EGDT did significantly better




In-hospital mortality 30.5% vs. 46.5%,
ARR 16%, NNT = 6; OR 0.58 (95%CI 0.38 – 0.87)
60d mortality 44.3% vs. 56.9%
 Primarily explained by reduction in sudden CVS
collapse deaths (10.3% vs. 21.0%)
Various secondary outcomes (labs & severity
scores) significantly better in EGDT group



EGDT pts spent longer time in the ED
EGDT survivors spent less time in hospital than
standard Tx survivors (14.6 d vs. 18.4 d)
Baseline SVO2 was 48% despite only 50% ventilated
Rivers et al. 2001 (cont’d)

Differences in EGDT group
 More fluid early (4.9 L vs. 3.5L)
 More transfusions (64.1% vs. 18.5%)
 More inotropic support (13.7% vs. 0.8%)
 Less use of pulmonary artery catheters
later in ICU stay (18% vs. 31.9%)
Controversies

Conflicts with earlier studies showing lack of
benefit from using hemodynamic goals




Hayes et al. N Eng J Med 1995; 330: 1717-22
Gattinoni et al. N Eng J Med 1995; 333: 1025-32
Different time points – all prior studies in ICU
setting
More heterogeneous patient populations
Controversies




Transfusion practice
 How does this fit with the TRICC trial?
Need for IJ or SC lines
Which part of protocol accounts for benefit?
How will this affect department flow
Supporting data

Retrospective study of pediatric sepsis



Early normalization of vitals associated with >9
fold improved odds of survival
Odds of mortality increase >2-fold with every hour
of ongoing shock
Only 45% of pts were adequately fluid
resuscitated

Han et al. Early reversal of pediatric-neonatal septic
shock by community physicians is associated with
improved outcome
Supporting data

Success of hemodynamic optimization
appears time-dependent



Meta-analysis of ICU pts
Studies instituting PAC goal-directed therapy later
than 12 hrs or after onset of organ failure failed to
show benefit
Studies that intervened early found to result in
significant mortality reduction of 23% (95%CI 1630)

Kern et al. Meta-analysis of hemodynamic optimization in
high-risk patients. Crit Care Med 2002; 30: 1686-92
Fluids in Sepsis
SCCM Guidelines



No evidence for choosing colloid over
crystalloids (Grade C)
Administer crystalloids as 500-1000 cc over
30 mins & repeat prn based on response (E)
Administer colloids at 300-500 cc over 30
mins & repeat prn based on response (E)
Crystalloids or colloids?



Controversial
Many heterogeneous studies
No evidence for superiority of one over other, but
trend towards increased mortality w/ colloids



Choi et al. Crystalloids vs. colloids in fluid resuscitation: A
systematic review. Crit Care Med. 1999; 27: 200-10
Shierhout & Roberts. Fluid resuscitation with colloid or
crystalloid solutions in critically ill patients: A systematic
review of randomized trials. BMJ. 1998; 316: 961-4
Crystalloids

Cheaper, easily available, less risk of anaphylactoid rxns,
resuscitate intra- & extravascular space
Vasopressors
Just need a little squeeze..
SCCM Guidelines

Should be used when







shock refractory to fluid resuscitation
Life-threatening hypotension (E)
Dopamine or norepinephrine are 1st line agents (D)
‘Renal dose’ dopamine does not work & should not be
used (B)
Invasive BP monitoring & central IV lines should be
placed as soon as possible (E)
Vasopressin may be considered as a 2nd line agent in
refractory shock (E)
Dobutamine may be considered in refractory shock felt
to be due to low cardiac output (E)
Does “renal dose” dopamine
work?

NO!!


DBRCT multicenter trial of 328 ICU pts
randomized to placebo or dopamine at
2ug/kg/min
No difference in mortality, peak serum creatinine,
need for renal replacement Tx, rise in serum
creatinine, or length of stay

ANZICS clinical trials group. Low-dose dopamine in
patients with early renal dysfunction: a placebocontrolled randomized trial. Lancet. 2000; 356: 2139-43
Norepinephrine or Dopamine
1st in septic shock?

Tons of animal data; very few clinical studies


Decreased mortality w/ norepinephrine vs. dopamine in
one NON-randomized trial
Theoretical benefits w/ norepinephrine
 Less tachycardia
 No effect on HPA or cerebral perfusion pressure
 Increased GFR
 Decreased lactate levels
 Improved splanchnic blood flow



Vincent & de Backer. Crit Care 2003; 7: 6-8
On the other hand dopamine is quickly available and
familiar
Bottom line = either will do as an initial pressor
Transfusion
Why do sick pts become
anemic?


95% of ICU pts are anemic by day 3 of ICU
stay
Mechanisms


Phlebotomy = 65 ml/day on average
Underproduction anemia


Blunted erythropoietin response secondary to
inflammatory cytokine production
Abnormal iron metabolism due to immune activation

Low iron levels & elevated ferritin
 Corwin et al. Transfusion practice in the critically ill. Crit
Care Med 2003; 31(S): S668-71
What should be the
transfusion threshold?

SCCM Guidelines



Transfuse to keep Hb > 70 g/L unless extenuating
circumstances (e.g. CAD) (B)
Based on TRICC trial
Rivers et al. 2001

Transfuse to keep hematocrit at least 30%
TRICC Trial

Multicenter RCT of 838 ICU pts w/ Hb <90

Randomized to

Liberal strategy


Restrictive strategy (study group)


Transfusion threshold 100 g/L – aim for 100 -120g/L
Transfusion threshold 70 g/L – aim for 70 – 90 g/L
Primary outcome

All cause mortality at 30 days

Herbert et al. A multicenter, randomized, controlled trial of
transfusion requirements in critical care. N Eng J Med.
1999; 340: 409-17
TRICC Trial

Results





No difference in 30d mortality
 ARR 4.6% (95%CI -0.84 – 10.2%)
No difference in 60d mortality
No difference in mortality in sepsis sub-group
Less sick pts (APACHE II score <20) did better
with restrictive strategy
 ARR 7.4% (95%CI 1.0 – 13.6%)
Conclusion

Restrictive strategy equivalent to, and possibly
better than keeping Hb > 100 g/L
Why the TRICC trial does not
contradict Rivers et al

Different patient population



Euvolemic pts
Enrolled within 72 hrs of ICU admission
Only 6% had Dx of sepsis, and only 26.5% had
any infection at all
Antibiotic Treatment
The war against bugs
SCCM Guidelines




Draw appropriate cultures first
Give antibiotics within 1 hr of recognition of septic
syndrome
Antibiotics should be broad-spectrum & chosen to
cover most likely organisms based on presentation
& local resistance patterns
Arrange for further diagnostic studies to rule out
surgically correctable foci of infection once
appropriate
Fatal Error

“Autopsy studies in
persons who died in the
intensive care unit show
that failure to diagnose
and appropriately treat
infections with antibiotics
or surgical drainage is the
most common avoidable
error”

Hotchkiss & Karl. The
pathophysiology and treatment
of sepsis. N Eng J Med. 2003;
348: 138-50
Do Antibiotics make a
difference?


Animal models indicate progressive increase in
mortality w/ each hour of delay to Abx
Few prospective RCT’s – most outcome data based
on retrospective analyses


ARR 16 – 26% when initial Abx were appropriate
Virtually all studies in ICU setting


Prospective cohort study of 406 pts w/ sepsis found
inadequate initial Abx Tx significantly increased risk of
death in non-surgical sepsis (OR 8.15; 95%CI 1.98-33.5)
Adequate Tx dec’d risk of death in surgical sepsis (OR
0.37, 95%CI 0.18-0.77)

Garnacho-Montero et al. Impact of adequate empirical antibiotic
therapy on the outcome of patients admitted to the intensive
care unit with sepsis. Crit Care Med 2003; 31: 2742-51
Mono- or Combination
Therapy?

Combination Tx: theoretical advantages




Broadens spectrum
Synergism
Decreases emergence of resistant strains
No good studies to document improved
outcomes

Paucity of relevant data and adequately powered
studies

Bochd, Glauser, & Calandra. Antibiotics in sepsis. Int Care
Med. 2001; 27: S33-48
Mono- or Combination
Therapy?

Meta-analysis




64 RCT’s (7586 pts) of monotherapy vs. Beta-lactam &
aminoglycosides combo
No difference in all-cause mortality, treatment failure, or
resistance development
Lack of benefit consistent in all subgroups analyzed
Significant increased nephrotoxicity w/ combo Tx NNH
15 (14-17)

Paul et al. Beta-lactam monotherapy versus beta-lactam
aminoglycoside combination therapy for sepsis in
immunocompetent patients: systematic review and metaanalysis of randomized trials. BMJ 2004; 238: 668
Local ID recommendations:
Quick reference guide

Undifferentiated febrile
shocky pt w/ no focus



Ceftriaxone

Gentamicin or quinolone
Meningitis


Ceftriaxone + macrolide
or resp quinolone
Urinary tract

Intraabdominal

Respiratory



Ceftriaxone +/vancomycin +/- ampicillin


Ancef + flagyl +
gentamicin (24 hr dosing)
Ceftriaxone + flagyl
Pip-tazo
Carbapenem
Necrotizing fasciitis

IVIG + penicillin +
clindamycin + surgery

Dr. Dan Gregson
personal communication
Blood cultures

Important to establish Dx for:



Guiding antimicrobial Tx
Guiding adjunctive Tx
Microbiological epidemiological surveillance



Llewelyn & Cohen. Diagnosis of infection in sepsis. Int Care
Med. 2001; 27: S10-32
We still frequently fail to obtain cultures prior to
initiating antibiotics in the ED
Yield of cultures ranges from 9-64%

Bochd, Glauser, & Calandra. Antibiotics in sepsis. Int Care Med. 2001;
27: S33-48
Blood cultures

Limitations



Colonization vs. infection
Prior antimicrobial Tx
Significance of



Rare or unfamiliar organisms
Mixed culture results
Organisms not usually associated w/ Dz

Llewelyn & Cohen. Diagnosis of infection in sepsis. Int
Care Med. 2001; 27: S10-32
Source Control
Getting to the root of the cause
SCCM Guidelines




Arrange for appropriate studies & consults
Choose intervention least disruptive to
unstable physiological status (e.g.
percutaneous drainage rather than surgery)
Source control should occur as soon as
possible after initial resuscitation
Remove lines & tubes if appropriate
Ya gotta do all the lil’ things
right…that’s what wins games

When inserting central
lines, chest tubes etc




Wear a sterile gown
Wear a mask
Prep & drape a huge
area
Communicate potential
‘dirty’ lines to admitting
service
Ventilatory Strategies
SCCM Guidelines

Avoid high tidal volumes & high plateau pressures
(B)





Target TV 6 ml/kg
Target end-inspiratory plateau pressure < 30 cm H20
Small levels of PEEP should be used to prevent
atelectasis (E)
Utilize permissive hypercapnea to help minimize TV
& plateau pressures if necessary (C)
In absence of contraindications, position intubated &
ventilated pts w/ HOB 45o to prevent VAP
Background

Traditional vent parameters


TV 10 – 15 ml/kg, minimal PEEP to maintain normal
PCO2, PO2 & pH
ARDS mortality as high as 90% in the 70’s (currently
30-40%)

Gattinoni et al. Physiologic rationale for ventilator setting in
acute lung injury / acute respiratory distress syndrome
patients. Crit Care Med. 2003; 31(S): S300-04
Mechanisms of VALI

Mechanisms of lung injury due to barotrauma &
volutrauma



Basic underlying effect due to high inspiratory pressures & resultant
shear forces on lung parenchyma
Large Vt's, low PEEP, high peak pressure  large gradient in lung
volume b/w inspiration & expiration  cyclical collapse & distention
of alveolar units
Lung heterogeneity – diseased portions collapse
while healthy alveoli become overdistended



Local inflammation
Disruption of alveolar-capillary barrier
Reduced clearance of edema fluid

Frank & Matthay. Science review: Mechanisms of ventilator-induced
injury. Crit Care. 2003; 7: 233-41
Mechanisms of VALI



Conventional ventilation increases
inflammatory mediators
Reducing TV’s to 6-8 ml/kg decreases cytokine
levels
Low TV’s and inspiratory pressures
significantly reduce mortality in human trials


Vincent et al. Reducing mortality in sepsis: new directions. Crit Care.
2002. 6: S1-18
Excessively low TV’s & airway pressures also
appear to lead to VALI likely due to propensity
for alveolar collapse

Brower et al. Lung-protective ventilation strategies in acute lung injury.
Crit Care Med 2003; 31(S): S312-16
Summary of trials
Frank & Matthay. Science review: Mechanisms of ventilatorinduced injury. Crit Care. 2003; 7: 233-41
ARDS-Net trial


Multicenter RCT of 861 adults with ARDS
Randomized within 36 hrs of intubation to:
 Control group



Low Vt group



Vt 12 ml/kg predicted body wt on AC mode
Adjusted Vt to keep plateau pressure b/w 45-50 cm H2O
Vt 6-8 ml/kg predicted body wt on AC mode
Adjusted Vt to keep plateau pressure b/w 25-30 cm H20
Followed for 180 days

Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes as
compared with traditional tidal volumes for acute lung injury and the acute respiratory
distress syndrome
. N Eng J Med. 2000; 342: 1301-8
ARDS-Net trial (cont’d)

Primary outcomes



In hospital mortality
Ventilator-free days in first 28 days
Secondary outcomes



Organ failure
Barotrauma
Plasma IL-6 levels
ARDS-Net trial (cont’d)

Results: Low Vt group had

Sig decreased mortality







31.0% vs. 39.8%
ARR 8.8% (95%CI 2.4-15.3%); NNT = 11
More ventilator free days
More organ-failure free days
More pts breathing w/o assistance at 28d
Greater decreases and lower absolute levels of
IL-6 at day 3
No difference in barotrauma
Criticisms

Vt or plateau pressure


Control group had higher plateau pressures
Recent meta-analysis argues that difference due
to increased mortality in control group & that
plateau pressures are to blame (despite lower
than average mortality in control group)

Eichacker et al. Meta-analysis of acute lung injury and
acute respiratory distress syndrome trials testing low tidal
volumes. Am J Resp Crit Care Med 2002; 166: 1510-14
Conclusions

General consensus in the literature that ARDS
trial results are valid, and that



VT should be 6-8 ml/kg
Plateau pressures should be kept to < 30 cm H2O
PEEP should be used to minimize alveolar collapse at
pressures as low as possible (start 5-10 cm H2O)
Steroids
SCCM Guidelines



IV hydrocortisone 200-300 mg/day for 7 days
should be given to adequately fluidresuscitated pts in refractory shock (C)
Doses of > 300 mg/day should not be used
(A)
Use dexamethasone in ED & consider use of
ACTH stim test to identify pts in need of
continued steroids (E)
Background

Anti-inflammatory effects


Basis for large dose (primarily methylprednisolone 30
mg/kg followed by 5 mg/kg }steroid trials in 80’s
2 large RCT’s failed to show benefit



Veterans administration. Effect of high-dose glucocorticoid
therapy on mortality in patients wit clinical signs of systemic
sepsis. N Eng J Med. 1987; 317: 659-65
Bone et al. A controlled clinical trial of high dose
methylprednisolone in the treatment of severe sepsis and
septic shock. N Eng j Med. 1987; 317: 653-58
Meta-analysis of 9 RCT’s found no benefit, and possibly
increased mortality w/ large dose steroids RR 1.13, 95%CI
0.99 – 1.29

Cronin et al. Corticosteroid treatment for sepsis: A critical
appraisal and meta-analysis of the literature. Crit Care Med.
1995; 23: 1430-39
Background

Concept of adrenal insufficiency

Stress steroid response essential
 Taking out adrenals increases septic & hemorrhagic shock
mortality in animals -- reversible with exogenous steroids
 Bilateral adrenal necrosis or infarction noted in ~30% of
septic pts at autopsy
 Multiple factors affect HPA axis during stress
 Studies of sepsis pts have shown that up to 42% have
adrenal or HPA dysfunction which correlates w/ increased
mortality
 Multiple studies document improved catecholamine
response in steroid-treated septic shock

Prigent et al. Clinical review: Corticotherapy in sepsis. Crit Care
2004; 8: 122-29
Annane et al. Effect of treatment with low doses of
hydrocortisone and fludrocortisone on mortality in
patients with septic shock. JAMA 2002; 288: 862-71

Multicenter DBRCT of 300 adult septic shock pts
tested with short corticotropin test & randomized to



Primary outcome


Placebo
Hydrocortisone 50 mg q6h IV & fludrocortisone 50 ug PO
OD for 7 days
28d survival
Secondary outcomes




28d survival in responders vs. nonresponders
28d, 1 yr, ICU, & hospital mortality
Time to vasopressor Tx withdrawal
Adverse events
Annane et al cont’d

Results


No significant difference in mortality for all pts
Non-responders treated w/ steroids had decreased
28d mortality


Less reliance on vasopressors



53 vs. 63%; ARR 10%, OR 0.54 (95%CI 0.31-0.97) NNT = 10
Non-responders: Median time to withdrawal 7 vs. 10 d; HR
1.91 (95%CI 1.29-2.84)
All pts: Median time to withdrawal 7 vs. 9 d; HR 1.54 (95%CI
1.10-2.16)
No significant differences in adverse events
Criticisms

Possible inclusion of true adrenal insufficiency


Use of fludrocortisone in addition to hydrocortisone



Not widely practiced
CORTICUS trial ongoing to evaluate hydrocortisone alone
in septic shock
Underpowered to detect harm in responders



High mortality rate in placebo group
Trend towards harm in responders needs clarification
Avoid steroids for all approach
Change of entry criteria during study

No analysis of pts recruited before & after
Diagnosis of adrenal
insufficiency

No clear cut normal range: Serum cortisol levels
variable & poorly reflective of biologic action during
stress
Inc’d production (up to 6x normal) & loss of diurnal variation
 Dec’d concentration & binding affinity of corticosteroidbinding globulin (CBG)
 Inc’d local concentration due to protease activity on CBG
 Up- or down-regulation of intracellular steroid receptors
Elevated and depressed cortisol levels are both associated
w/ increased morbidity & mortality



Cooper & Stewart. Corticosteroid insufficiency in acutely ill
patients. N Eng J Med. 2003; 348: 727-34
Adrenals & sepsis

Adrenal responsiveness is normally
maintained even during illness


25-59% of pts maintain corticotropin response
Possible causes of adrenal insufficiency



Drugs (etomidate, fluconazole)
Adrenal infarction / hemorrhage / abscess
Dysregulation of HPA axis by high levels of
inflammatory cytokines  steroid resistance

Cooper & Stewart. Corticosteroid insufficiency in acutely
ill patients. N Eng J Med. 2003; 348: 727-34
Diagnosis of adrenal
insufficiency

What is a normal serum cortisol during
stress?




Most controversial area
Nobody knows – no good studies to compare
methods of testing for adrenal insufficiency w/
accepted gold standards
Accounts for variation in incidence from 1.4 – 54%
Current diagnosis based on limited data &
consensus opinion on threshold cortisol levels &
“appropriate” response to ACTH stim test
Suggested diagnostic
approach


Draw a random cortisol level
Perform a ACTH stim test



Administer 250 ug of cosyntropin IV
Draw serum cortisol levels at 0, 30, and 60 min
Give dexamethasone 2-4 mg in ED


Does not interfere w/ ACTH stim test
Treatment should be stopped if test negative
 Serum cortisol levels >1242 nmol/L have been found to be
associated w/ significantly greater mortality
 Suggests that exogenous steroids could be harmful

Sam et al. Cortisol levels and mortality in severe sepsis. Clin
Endo. 2004; 60: 29-35
Interpreting results

Random cortisol




< 414 nmol/L (15 ug/dL) – suggestive of adrenal
insufficiency – start steroids
>938 nmol/L (34 ug/dL) – suggestive of steroid resistance
– replacement unlikely to help
414 – 938 nmol/L – base decision on ACTH stim test result
ACTH stim test


>250 nmol/L (9 ug/dL) change  adrenal insufficiency
unlikely
<250 nmol/L (9 ug/dL) change  suggestive of adrenal
insufficiency – start steroids

Cooper & Stewart. Corticosteroid insufficiency in acutely ill
patients. N Eng J Med. 2003; 348: 727-34
What about serum free
cortisol?

66 consecutive ICU pts w/ APACHE III > 15



Compared w/



Group 1: serum albumin ≤ 25 g/L
Group 2: serum albumin > 35 g/L
33 healthy volunteers
7 ICU pts w/ proven adrenal insufficiency
Looked at

Total & free cortisol levels at baseline & after
cosyntropin test
Conclusions




Critically ill pts have elevated cortisol levels
Free cortisol levels can change significantly
w/ less significant concomitant changes in
total cortisol
Total serum cortisol levels in pts w/
hypoproteinemia can be misleading
Suggest baseline free cortisol of < 52.4
nmol/L identifies pts at risk for adrenal
insufficiency
Steroid conclusions





Think of steroids in pts w/ apparent septic
shock refractory to standard treatment
Draw baseline cortisol levels & do ACTH stim
test
Use dexamethasone in the ED
Do NOT give steroids card blanche
Await trials on use of free cortisol
Recombinant Human
Activated Protein C
(rhAPC = Drotrecogin alfa = Xigiris®
aka superdrug)
SCCM Guidelines

rhAPC should be given to severely ill pts:





APACHE II score > 25
Sepsis-induced MOF
Septic shock
ARDS
In the absence of absolute or significant
relative contraindications (B)
Background

No pharmacologic agent shown to reduce in sepsis
mortality in phase III trials….











Ibuprofen
NAC
Anti-TNF-α mAb vs. placebo (NORASEPT II)
IL-1 receptor antagonist vs. placebo
PAF receptor antagonist vs. placebo
High dose steroids
Bradykinin antagonist (Deltibant)
Tissue factor pathway inhibitor
AT III vs. placebo (KYBERSEPT)
Etc, etc
…until now (maybe)
Multiple pharmacologic
actions



Anti-thrombotic
 Inhibits FVa & FVIIIa
Anti-fibrinolytic
 Inhibits PAI-1 & TAFI
 Decreases thrombin
production
Anti-apoptotic

Induces Bcl-2 & inhibitor
of apoptosis-1 gene
expression

Anti-inflammatory




Inhibits TNF-α, IL-1, IL6 production
Inhibits monocyte &
neutrophil migration
Inhibits lipid A
activation of
monocytes
Inhibits tissue factor
activation
McCoy & Matthews. Drotrecogin Alfa (Recombinant Human
Activated Protein C) for the treatment of severe sepsis. Clin
Ther 2003; 25: 396-421
Protein C as Tx

Most septic pts have low levels of protein C


Pharmacologic properties




Associated with increased M & M
Anti-inflammatory, anti-thrombotic, anti-fibrinolytic
Increased survival in primate model of septic shock
Improved outcomes suggested in non-randomized
trial of meningococcemia
Administration associated w/ dec’d levels of
proinflammatory mediators & D-dimer in humans
PROWESS Trial: Bernard et al. Efficacy and safety
of recombinant human activated protein C for
severe sepsis. N Eng J Med 2001; 344: 699-709


Multicenter DBRCT of 1690 adult pts w/
severe sepsis
Randomized to



rhAPC infusion @ 24 ug/kg/h for 96 hrs
Placebo
Primary outcome

All-cause mortality at 28d
Bernard et al. cont’d

Results

rhAPC significantly reduced mortality



28d mortality 24.7% APC vs. 30.8% placebo
ARR 6.1% (95% CI 1.9-10.4); NNT = 16
rhAPC had non-significant increase in risk of
serious bleeding

3.5% vs. 2.0% (p=0.06), NNH = 67
Sounds great, but don’t forget
to read the fine print…
Post hoc analyses


Pts w/ APACHE II scores <25 did worse w/
rhAPC than w/ placebo
Benefit dec’d w/ less organ dysfunction




ARR single organ system 1.7%
ARR multi-organ failure 7.4%
More benefit in pts w/ septic shock rather
than sepsis
Pts not in DIC did worse w/ rhAPC than w/
placebo

Dhainaut et al.
Drotrecogin alfa
(activated) in the
treatment of severe
sepsis patients with
multiple-organ
dysfunction: data
from the PROWESS
trial. Int Care Med
2003; 29: 894 - 903
Criticisms

Face Validity


When multiple other trials of anti-cytokine or antithrombotic meds have not worked why does this one?
Validity of the results
 Entry criteria where changed ½ way through
 Cell line used to produce rhAPC was changed
½ way through


Changes not mentioned anywhere in methods
Changes coincided w/ significant difference in
observed efficacy
Criticisms

External validity



Exclusion criteria extensive & included many pt
groups relevant to increasing incidence of sepsis
What is mortality beyond 28d? What is status of
survivors?
Other concerns


Sponsored by Eli Lilly
3 authors are Eli Lilly employees, 2 are stockholders,
5 have served as consultants leaving only 3/11
primary authors as having no ties
Remaining questions






What is the mortality benefit beyond 28d? What
about morbidity?
What is the best method to identify pts most likely to
benefit from rhAPC?
Can we use rhAPC in any of the pt populations
excluded from PROWESS?
Would giving rhAPC earlier = greater efficacy?
Would a longer Tx period = greater efficacy?
How does rhAPC interact w/ other existing or novel
sepsis therapies?
Cost

$335 Cdn per 5 mg vial


0.024 mg x 70g kg x 96 hrs = ~161 mg or 32 vials = $10,
800 Cdn per treatment
Is it cost-effective? Yes, if used selectively.


Cost per life-year gained
 APACHE II <25 $19, 723 USD
 APACHE II >25 $575,054 USD
Total cost to our system
 CHR ICU pharmacy budget 2001: $1.6 million USD
 Cost if rhAPC was used in pts w/ APACHE II > 25:
$482,800 USD

Manns et al. An economic evaluation of activated protein C for
severe sepsis. N Eng J Med. 2002; 347: 993-1000
Contraindications to rhAPC






Active internal bleeding
Recent (within 3 mo) hemorrhagic stroke
Recent (within 2 mo) intracranial or intraspinal
surgery, or severe head trauma
Trauma with an increased risk of life-threatening
bleeding
Presence of an epidural catheter
Intracranial neoplasm or mass lesion or evidence of
cerebral herniation
The Future

ADDRESS Trial




Placebo-controlled trial of rhAPC in lower-risk pts
w/ severe sepsis
Trials in pediatric populations
Trials examining use of heparin in conjunction
w/ rhAPC
Development of more defined criteria for
selecting pts likely to benefit from rhAPC
Insulin therapy
Sugar: its not just bad for your teeth..
SCCM Guideline

IV infusions of insulin should be used to
maintain serum glucose levels < 8.3 mmol/L
(D)
Van den Berghe et al. Intensive insulin
therapy in critically ill patients. N Eng J Med.
2001; 345: 1359-67


RCT w/ blinded outcome assessment of 1548 pts
admitted to surgical ICU
Randomized to



Intensive insulin Tx
 Continuous IV insulin initiated if glucose >6.1 mmol/L &
adjusted to maintain glu b/w 4.4 – 6.1 mmol/L
Traditional Tx
 Continuous IV insulin initiated if glucose >11.9 mmol/L &
adjusted to maintain glu b/w 10-11.1 mmol/L
Primary outcomes

All-cause ICU mortality
Van den Berghe et al cont’d

Results





ARR 3.4% or NNT = 29, adjusted RRR 32%
(95%CI 2-55%)
Greatest reduction in mortality due to decrease in
deaths due to MOF with septic focus (33 pts vs. 8
pts)
Decreased septicemia by 46% (95%CI 25-67%)
Decreased need for renal replacement, ARR 3.4%
Shorter ICU stay, less Abx, less vent support
Van den Berghe et al

Criticisms

Generalizability




Virtually all post-op pts – is this applicable to all “sick’
pts?
Does prophylactic value of glycemic control in
preventing sepsis translate into benefit as treatment
for primary sepsis?
Single center trial
Not really blinded
Is it the normoglycemia or the
insulin that provided the benefit?


Multivariate analysis of the original study data
suggests benefit primarily related to
prevention of hyperglycemia than to
exogenous insulin administration
Prevention of renal failure however appeared
to be associated more with insulin Tx

Van den Berghe et al. Outcome benefit of intensive
insulin therapy in the critically ill: Insulin dose versus
glycemic control. Crit Care Med. 2003; 31: 359-66
Conclusions

We can reduce mortality in sepsis in the ED:








EGDT:
ARDSNet vent strategy
Steroids
rhAPC
Insulin
Dex in Meningitis
Early appropriate Abx
Source control
ARR 16.0%
ARR 8.8%
ARR 10.0%
ARR 6.1%
ARR 3.4%
ARR 8.0%