The Microcirculation in Sepsis

The Microcirculation in sepsis

Daniel De BackerDepartment of Intensive CareErasme University Hospital

Brussels, Belgium

• Microcirculatory perfusion is a key determinant of tissue perfusion.

• Microvascular perfusion is under control of different mechanisms than systemic hemodynamics.

• O2 transport is driven at microcirculatory level by diffusion more than by convection.

TISSUE PERFUSION: Key points

Trzeciak et alCrit Care 9:S20;2005

The density of capillaries is a primary determinant of tissue oxygenation

Saldivar-E et alAJP 285:H2064;2003

The density of capillaries is a primary determinant of tissue oxygenation

Adaptation to chronic hypoxia is characterized by an increased capillary density

MICROCIRCULATION

DDB USI

r4 1V =

P8 L

.. . .

Determinants of microvascular blood flow

Blood flow is adapted to local metabolic needs through local vasodilation and upstream changes in vasomotor tone

Neural control

Hungerford-JE et alFASEB 14:197;2000

Perivascular sympathetic

nerves10 µm

Neural control Beach-JM et alAJP 275:H1489;1998

Arteriole

Capillaries

Microcirculatory alterations in experimental sepsis

Pig

Gut serosa

OPS, 5x probe

Pig

Gut serosa (sepsis)

OPS, 5x probe

Pig

Liver

OPS, 5x probe

Pig

Liver (sepsis)

OPS, 5x probe

Sheep

Brain

SDF, 5x probe

Sheep

Brain (sepsis)

SDF, 5x probe

Hoffmann et alCCM 32,1011;2004

Hamster skinfold LPS

DECREASED CAPILLARY DENSITY IN EXPERIMENTAL SEPSIS

ENDOTOXIC SHOCK MODEL

ANOVA: in sepsis all times p<0.0001 vs baseline

Evolution of sublingual microcirculation in septic shock

Verdant et alCCM 37;2875,2009

Sham

Sepsis

PigsCholangitis

Evolution of microcirculation in unresuscitated septic shockPranskunas A et alCrit Care 16:R83;2012

Pigs E Coli infusion

MICROVASCULAR ALTERATIONS IN EXPERIMENTAL SEPSIS

• Decreased capillary density• Heterogeneity of perfusion

HETEROGENEITY OF CAPILLARY PERFUSION IN EXPERIMENTAL SEPSIS Ellis C et al

AJP 282:H156;2002

rats

SPACE

CONTROL

SINUSOIDAL BLOOD FLOW IN ENDOTOXIC SHOCK

After Pannen et al.AJP 271: H1953; 1996

%

0

10

20

30

0 10 20 30 40 50 >50

Fraction ofperfusedcapillaries

ENDOTOXIN

0

10

20

30

40

0 10 20 30 40 50 >50

%

Sinusoidal flow pl/s

Fraction ofperfusedcapillaries

Increase heterogeneity of renal perfusion in sepsis

RatsLPS

Legrand et alICM 2011

EXPERIMENTAL STUDIES IN SEPSIS

Microvascular blood flow alterations are frequent

decreased vascular densityabsent or intermittent flow in capillariesheterogeneity between areas

DDB USI

Branemark et Urbaschek Angiology 18:667;1967Lam et al. JCI 94: 2077; 1994Farquhar et al. J Surg Res 61: 190; 1996Madorin et al CCM 27:394;1999Ellis et al AJP 282:H156;2002Verdant et al CCM 37:2875;2009Secor et al ICM 2010

Different models (LPS, CLP, live bacteria,…)Various species (rats, mice, hamsters, pigs, sheep…)Various organs (skin, gut, liver, lung, kidney, heart, brain…)

Cardiomyocytes rats

Bateman et alAJP 293/H448;2007

CTRL LPS

Consequences…

0

10

20

30

40

50

60

70

80

Gut mucosa

SEPSIS CTRL

-10

-5

0

5

10

15

p<0.01

%

%

Hemorrhage

p<0.01

Humer et alJAP 81: 895; 1996

CRITICAL O2ER

CHANGES IN CAPILLARY TRANSIT-TIME HETEROGENEITY

MICROCIRCULATORY ALTERATIONS AND EO2CRIT

Potential mechanisms ?

PATHOPHYSIOLOGY OF MICROVASCULAR ALTERATIONS

Triggered by inflammatory mediatorsTNF (Vicaut E JCI 87:1537;1991)

ENDOTHELIAL ALTERATIONS

ctrl

5 dayspost LPS

Leclers J et alCCM 28:3672;2000

rabbits

25-35% vessels

Loss of neural control Tyml-K et alAJP 281:H1397;2001

Change in diameter and

communication rate (CR500)

between 500 µm distant

microvessels(retrograde

communication)

B: cremaster muscle (mice)C: Endothelial microlayer

Enhanced response to vasoconstrictor substances Pannen B et alAJP 26:180;2001

CTRL

HEM

CTRL + ET

HEM + ET

Rats

2 µm

Endothelialcell

Glycocalyx

0.2 µm

GLYCOCALYX ALTERATIONS van den Berg et alCirc Res 92:592;2003

Marechal X et alShock 29:572;2008

Role of glycocalyx in microvascular alterations

Rats, ileum

µm

Gupta A et alAJP 293:F245;2007

LPS promotes rolling and adhesion in renal microcirculation

Two photons videomicroscopy

McCuskey et alCardiovasc Res32:752:1996

Rouleau formation

Altered Red blood cell deformability

Eichelbronner et alICM 29:709;2003

ETX promotes adhesion of RBC to endothelium

HUVEC

*

Croner et alCrit Care 10:R15;2006

Microthrombi ?

Rats, CLP

VasoconstrictionPlatelet/WBC interaction with endotheliumMicrothrombi not frequent

Secor et alICM 36:1928; 2010

Mice, CLP

De Backer et alAnnals Intensive Care 2011

Potential mechanisms

ALSO IN HUMANS ?

DDB USI

Vascular density(all vessels)

++ p <0.01 vs volunteers

2.5

3

3.5

4

4.5

5

5.5

6

6.5

7n/mm

Volunteers(10)

Septic patients(50)

++

De Backer et alAJRCCM 166:98;2002

DDB USI

Percentage of vessels perfused(large vessels)

60

65

70

75

80

85

90

95

100%

Volunteers(10)

Septic patients(50)

MICROCIRCULATORY ALTERATIONS IN SEPTIC PATIENTS


DDB USI

0

20

40

60

80

100

Volunteers(10)

Severe sepsis(50)

+++

%

+++ p <0.001 vs volunteers

Percentage of vessels perfused(small vessels)

De Backer et al AJRCCM 166:98-104;2002

Emergency department

Trzeciak et alAnn Em Med 49:1579;2007

N=26

0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

Heterogeneity index

CTRL SEPSIS

P<0.01

Alterations of sublingual microcirculation in patients with sepsis

• De Backer et al AJRCCM 2002• Spronk et al Lancet 2002• Sakr et al CCM 2004• De Backer et al CCM 2006• De Backer et al CCM 2006• Creteur et al ICM 2006• Boerma et al CCM 2007• Trzeciak et al Ann Emerg Med 2007• Sakr et al CCM 2007• Trzeciak et al ICM 2008• Boerma et al ICM 2008• Dubin et al Crit Care 2009• Buchele et al CCM 2009• Boerma et al CCM 2010• Ospina et al ICM 2010• Spanos et al Shock 2010• Pottecher et al ICM 2010• Morelli et al Crit Care 2010• Ruiz et al Crit Care 2010• Dubin et al J Crit Care 2010• Morelli et al ICM 2011• Edul et al CCM 2012• Kanvundis et al ICM 2012• Hernandez et al CCRP 2012• Pranskunas et al ICM 2013

• ↓ total vascular density• ↓ perfusion of capillaries

(no flow or intermittent flow)

• Preserved venular perfusion• Heterogeneity between areas

( close by a few microns)

baseline StO2 (%) delta StO2 (%)

slope (%/sec)

0

1

2

3

4

5

6

7

8

sepsis n = 72

volunteers n = 18

icu control n = 18

Slo

pe

(%

/sec

)

*

* p < 0.001 vs volunteers and ICU control

Creteur et alICM 2007

Endothelial reacitivity is impaired in sepsis

Alterations of NIRS vasoreactivity test in patients with sepsis

• Girardis et al ICM 2003• De Blasi et al ICM 2005• Pareznik et al ICM 2006• Podbregar et al Crit Care 2007• Doerschung et al JAP 2007• Creteur et al ICM 2007• Skarda et al Shock 2007• Nanas et al Aenesth Intens Care 2009• Payen et al Crit Care 2009• Donati et al Crit Care 2009• Mesquida et al ICM 2009• Mozina et al Crit Catre 2010• Georger et al ICM 2010• Shapiro et al Crit Care 2011• Soga T et al Emerg Med J 2013

Relationship with outcome ?

Severe sepsis (n=252)

Association with outcome De Backer et alCCM 41:791;2013

Edul et alCCM 2012

Vessel density, proportion of perfused vessels and heterogeneity but not velocity differ between survivors and non survivors

N=3N=15


Association with outcome De Backer et alCCM 41:791;2013

De Backer et alCCM 41:791;2013


Doerschug et alAJP 293:1065;2007

Severe sepsis 24 / volunteers 15

Shapiro et alCrit Care 15 R223; 2011

50 5860168 pts at emergency dpt

33 pts with septic shockTrzeciak et alICM 34:2210; 2008

• 1st SDF evaluation within 3 hours after EGDT initiation• 2nd SDF evaluation 3 to 6 hours after EGDT initiation• SOFA changes between 0 and 24 h

DDB USI

EVOLUTION OF MICROCIRCULATORY ALTERATIONS IN SEPTIC PATIENTS

Death after shock

ANOVAp<0.01* * *

Sakr et alCCM 32:1825;2004

Top et alCCM 39:8; 2011

Persistent microcirculatory alterations in pediatric sepsis

N=3N=15

0

1

2

3

4

5

6

7

8

Baseline 24hrs 48 hrs

Nonsurvivors (24/52)

Survivors (28/52)

*

*analysis of variance: p < 0.01

Slo

pe

(%/s

ec)

n = 52 septic patients Creteur et alICM 2007

Alterations of NIRS vasoreactivity test in patients with sepsis

Vallee F et alChest 138:1062;2012

Septic shock (n=46)

Ear Lobe PCO2

But isn’t it just the consequence of the altered global hemodynamics ?

Subtitle: could you detect it using hemodynamic measurements, clinical assessment or biomarkers?

Cardiac indexMean arterial pressure

LactateSvO2

mmHg L/min.M²

% mEq/L

40

50

60

70

80

90

100

1

2

3

4

5

6

7

20

30

40

50

60

70

80

90

Septic shockSevere sepsis

p<0.01

0

2

4

6

8

10

12

Septic shockSevere sepsisDDB USI


De Backer et al AJRCCM 166:98-104;2002



Association with systemic variables ?


Early severe sepsis (n=204)

Association with systemic variables ?

Primary event or adaptive phenomenon ?

Ellis C et alAJP 282:H156;2002

In perfused capillaries, O2 extraction is INCREASED in sepsis

Single perfused capillary

Cardiomyocytesrats

Bateman et alAJP 293/H448;2007

During LPS, ICAM-1 expression and HIF gene

induction are heterogeneous, and inversely related to flow.

Microcirculatory alterations are

associated with renal impairment

Wu L et alAJP 292:F261; 2007

Alteration in redox potential are proportional to microcirculatory alterations

Wu L et alAJP 292:F261; 2007

Mice / LPSPeritubular capillaries

Microcirculatory alterations are associated with renal hypoxia (co-localized with NADH)

Wu L et alAJP 292:F261; 2007

CTRL

LPS

Peritubular capillaries

Mice

NADH Hypoxia

Fink T et alShock 2013

Liver microvascular perfusion and redox state are inversely related

Rats / Fecal peritonitis / pretreatment / absence of shock

Brain microcirculation alteration precedes the loss of function

Rosengarten B et alCrit Care 13:R139;2009

Rats / LPS

R2 = 0,80

0

20

40

60

80

100

20 40 60 80

Baseline

% well perfused capillaries

Psl

CO

2gap

=> Microcirculatory alterations are primary events rather than secondary to altered cellular metabolism.

Creteur et alICM 32:516;2006

Expected(flow adapted to metabolism)

Stagnation of waste products

b-adrenoceptor stimulation improved liver microvascular perfusion and redox state

Rats / Fecal peritonitis


Sham Sepsis DobuSham Sepsis Dobu

Change in Lactate

%

Change in capillary perfusion

mEq/LDOBU 5 mcg/kg.min

-1

-0,8

-0,6

-0,4

-0,2

0

0,2

0,4

0 10 20 30 40 50


Effects of fluids

Influence of timing of fluid resuscitation

RatsLPS

Legrand et alICM 37:1534; 2011

20

30

40

50

60

70

80

90

100

Baseline Fluids

Late stage(>48h)N=23

Early stage(<24h)N=37

Proportion of perfused small vessels

%

$ p<0.01 fluids vs baseline and + p<0.01 late vs early

+ $

Microvascular effects of fluid challenge in patients with septic shock

Ospina et alICM 35:949;2010

Sensitivity Specificity Pos pred val

Neg pred val

Correct classifica

tion

MAP 0.55 0.50 0.42 0.64 0.52

CI 0.42 0.68 0.42 0.68 0.58

DPP 0.50 0.56 0.42 0.64 0.53

The microvascular response to fluids cannot be predicted by global hemodynamic changes (including indices of fluid responsiveness) !


20

30

40

50

60

70

80

90

100

Late, RL

Late, alb

Early, RL

Early, alb

Baseline After fluids

Proportion of perfused vessels

$ p<0.01 fluids vs baseline Coll vs cryst p = NS

%$ $

The time of administration but not the type of fluid influenced the microvascular response


Pottecher et alICM 2010

Futier E et alCrit Care 15:R214;2011

High risk surgery (n=42)

Fluids improve microvascular reactivity (NIRS ascending slope)

Futier E et alCrit Care 15:R214;2011

High risk surgery (n=42)

Relationship with systemic response

Vallée et alChest 138:1062;2010

Ear lobe PCO2

Relationship with systemic response

Effects of RBC transfusions

EFFECTS OF RED BLOOD CELL TRANSFUSIONS

N=35 RBC Baseline

Capillary density

Ca

pill

ary

de

nsi

ty (

/mm

)

7

6

5

4

3

2


Baseline capillary perfusion, %

De

lta

ca

pill

ary

pe

rfu

sio

n,

%

EFFECTS OF RED BLOOD CELL TRANSFUSIONS

N=35


Vasoactive agents ?

Secchi et alJourn Crit Care 12:137;1997

Rats, gut villi

ShamLPS

LPS + Dobu 2.5

Dobutamine in experimental sepsis

Secchi et alCCM 29:597;2001

Rats, liver sinusoids

Dobutamine in experimental sepsis

b-adrenoceptor stimulation improved liver microvascular perfusion

Rats / Fecal peritonitis / pretreatment / absence of shock


Capillary Perfusion

Dobutamine 5 mcg/kg.min De Backer et alCCM 34:403;2006

21 patients in septic shock

Change in Lactate

%

Change in capillary perfusion

mEq/LDOBU 5 mcg/kg.min

-1

-0,8

-0,6

-0,4

-0,2

0

0,2

0,4

0 10 20 30 40 50


Beta-adrenergic stimulation blunt PAF mediated leucocyte adhesion and rolling

HUVEC

Red blood cell White blood cell/platelet

Arterioles Capillaries

BETA(vasodilatation)

BETA(WBC / PLT)

Decrease rolling et adhesion

Effects of adrenergic agents on the microcirculation

Vasopressor agents ?

Influence of blood pressure on microvascular perfusion ?

Cerebral microcirculation

Sheep, peritonitis

Taccone FS et alCrit Care 14 R140; 2010

Cerebral microcirculation and metabolism

Sheep, peritonitis

Taccone FS et alCCM (in press)



No clear cut-off….

Blood pressure targets

Impact of vasopressor agents

0

20

40

60

80

100

120

Arter

ioles

Aorta

Capill

aries

Arter

ies

Venul

es

Veins

MAP (mmHg)

Vasodilation

Ctrl

Vasoconstiction

Impact of vasopressors on the microcirculation(Norepinephrine vs Vasopressine)

Hamster, control condition

Friesenecker et alCrit Care 10:R75;2006

Maier-S et alBJA 2009

Phenylephrine impairs microvascular perfusion in CPB

Impact of vasopressors on the microcirculation

Rats, LPS, gut muscularis

Nacul F et alAnesth Analg 110:447;2010

No fluidsDoses:NE 5 mcg/kg.minEPI 5 mcg/kg.minPHE 10 mcg/kg.minDOPA 20 mcg/kg.minDobu 12 mcg.kg.min

Normotensive sepsis

Red blood cell White blood cell/platelet

Arterioles Capillaries

BETA(vasodilatation)

BETA(WBC / PLT)

Decrease rolling et adhesion

Effects of adrenergic agents on the microcirculation

ALPHA(vasoconstriction)

Flow

Mean arterial pressure

Kidney

Muscle

Liver

Brain / heart

Different vascular beds have different vascular resistances

Gut

DDB USI

Does correction of hypotension result in an improved tissue perfusion ?

Impact of vasopressors on the microcirculation(Norepinephrine vs Vasopressin)

Rats, LPS

Baseline value

Shock value

Nakajima et alCCM 34:1847;2006

MAP 46 71 70 mmHg

Fries et alCCM 36:1886; 2008

Norepinephrine improved slightly

microvascular perfusion but not

µPO2

Rats / CLPBuccal mucosa

Correction of hypotension improves microvascular reactivity (NIRS)

MAP 54 => 77 mmHg

Georger et alICM 36:1882;2010

DDB USI

What is the optimal blood pressure target for the microcirculation ?

Jhanji et alCCM 37:1961;2009

p<0.05

p=0.45

N=16

Impact of MAP/NE on microvascular perfusion

Dubin et alCrit Care 2009

N=20


Thooft et alCrit Care 2011


Density of perfused small vessels

65 (baseline) 75 85 656

8

10

12

level of mean arterial pressure, mmHg

FC

D s

mal

l ve

ssel

s, n

/mm

*

*

Thooft et alCrit Care 2011


65 (baseline) 75 85 650

100

200

300

400

500 * *

level of mean arterial pressure, mmHg

Asc

endi

ng S

lope

%/m

in

* *

• Vasopressor agents have a dual effect on the microcirculation: on the one hand vasopressors decrease microvascular perfusion by constriction of precapillary sphincters.

• On the other hand, achievement of a minimal perfusion pressure is needed to preserve organ blood flow and microcirculatory perfusion.

• Optimal pressure targets are variable and should be individualized.

Vasopressors and the microcirculation

Vasodilatory agents ?

65

70

75

80

85

90

95

100

BASE

Proportion of perfused vessels(all vessels)

++ p <0.01 vs base

TOPICAL ACETYLCHOLINE(10-2 M)

Patients with septic shock (n = 11)

%

DDB USI



Spronk et alLancet 360:1395;2002

Effects of nitroglycerin

8 pts with septic shock

Boerma E et alCCM 38:93-100;2010


70 pts with severe sepsis


70 pts with severe sepsis

But can normal be more normal than normal ?

Boerma E et alCCM 38:93-100;2010

Proportion of Perfused Small Vessels

0

20

40

60

80

100

ENALAPRILATPLACEBO

Enalaprilat

Placebo

n=8n=8

n=8n=8

n=6n=7

n=7n=7

n=6n=7

* p<0.05 vs baseline and before drugp=0.83 for trend in enalaprilat groupp=0.006 for trend in placebo groupp=0.48 for group/time interaction

* *

PP

V s

mal

l ves

sels

, %

ACE inhibitors?Salgado D et alShock 2011

SheepCLP

But no impact on • outcome• organ function

Magnesium sulfate?Pranskunas A et alBMC Anesth 11:12; 2011

14 pts severe sepsis

Modulation of endothelial function ?

Vitamin CTyml K et alCCM 33,1823;2005

Rat / muscleCLP

Ascorbate 7.6g/100g BW 1h or 24h post CLP

Vitamin CThe effect is related to endothelial NOS

Tyml K et alCCM 2008

Mice / muscleFeces in peritoneum

Ascorbate 10-200 mg/kg 6h post peritonitis

BH4 (tetrahydrobiopterin)

Effect on NO metabolism at endothelial level

Schmidt S and Alp NClinical Science113: 47; 1997

He X et alCCM 20:2833; 2012

SheepCLP


He X et alCCM 20:2833; 2012

SheepCLP


Improved• outcome• organ function

• Multiple experimental and clinical studies suggest that microvascular alterations play a key role in the pathophysiology of sepsis and in the development of sepsis-induced organ failure.

CONCLUSIONS

• These alterations are due to several factors (endothelial dysfunction, interaction with circulating cells) that make unlikely that classical hemodynamic resuscitation can be effective in restoring an adequate microcirculation.

• Modulation of endothelial NO synthase seems promizing.

Health & Medicine

The Microcirculation in Sepsis