Control of breathing during

mechanical ventilation

D. Georgopoulos, Professor of Medicine,

Intensive Care Medicine Department,

University Hospital of Heraklion,

University of Crete,

Heraklion, Crete, Greece

Disclosures

Covidien

Draeger

Research grants

Basic principles of control of

breathing

Multi-sources

Medullary respiratory controller

Spinal motor neurons

Pmus = V’xR + ΔVxE

Volume-time profile

VentilationPaO2, PaCO2, pH

Diseases

Therapy

Pres Pel

Chem

ical

fee

dbac

k

Ref

lex a

nd M

echan

ical

feed

bac

k

Behavioral

feedback

Multi-sources

Medullary respiratory controller

Spinal motor neurons

Paw +Pmus = V’xR + ΔVxE

Volume-time profile

VentilationPaO2, PaCO2, pH

Diseases

Therapy

+ Paw = Prs + Pel

Volume-time profile

Mechanical

Chemical

Reflex

Behavioral

Feedback

Response of Pmus

to ventilator delivered

breath

Response of

ventilator to Pmus

Pmus

Ventilator and Patient

factors

Georgopoulos D: Principles and Practice of Mechanical Ventilation

(ed. Tobin) 2013: pp 805-820

Response of ventilator to Pmus

1) The mode of mechanical ventilation

2) The mechanics of the respiratory system

3) The characteristics of Pmus waveform

OutputInput

Μodes of assisted mechanical

ventilation

• Assist volume control (AVC, VT constant)

• Pressure support or control (PS, Pressure constant)

• Proportional modes (PAV, NAVA. Neither pressure

nor volume are pre-set but the patient effort drives

ventilator pressure)

Proportional modes will help us to clarify the issue of control of

breathing in mechanically ventilated patients

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

0 2 4 6 8 10 12

VT

(L)

Patient effort [PTPPmus (cmH2O.sec)]

PAV, NAVA

PS

AVC

Interaction between ventilator delivered volume

and patient effort with

different modes of support at high assist levels

Figure based on Data from: 1) Mitrouska et al. Eur Respir J 1999

2) Colombo et al. Intensive Care Med 2008

3) Meric et al. Respir Physiol Neurobiol 2014

Response of ventilator to Pmus

1) The mode of mechanical ventilation

2) The mechanics of the respiratory system

3) The characteristics of Pmus waveform

OutputInput

How do these factors affect the response of the ventilator?

(i.e. output)

-0,8

-0,6

-0,4

-0,2

0

0,2

0,4

0,6

0,8

-2 3 8 13 18

-5

0

5

10

15

20

25

30

35

-2 3 8 13 18

-5

0

5

10

15

20

-2 3 8 13 18

Time (sec)

Flo

w (

l/se

c)P

aw (

cmH

2O

)P

es (

cmH

2O

)

5 sec

Fr = 33 b/min

Georgopoulos et al. Intensive Care Med 2006;32:34

A patient with dynamic hyperinflation due to COPD on PSV

Ventilator rate = 12 b/min

Kondili et al. Br J Anaesth 2003;91:106

Double triggering due to expiratory asynchrony

between patient and ventilator

Ventilator rate = 2 x patient’s rate

Response of ventilator to Pmus

MessageDirect extrapolation of the ventilator output to the

patient’s control of breathing system (i.e. input) may be

misleading, guiding inappropriate therapeutic decisions

OutputInput

+ Paw = V’xR + VxE

Volume-time profile

Mechanical

Chemical

Reflex

Behavioral

Feedback

Response of Pmus

to ventilator delivered

breath

Response of

ventilator to Pmus

Pmus

Ventilator and Patient

factors

Georgopoulos D: Principles and Practice of Mechanical Ventilation

(ed. Tobin) 2013: pp 805-820

Chemical – Reflex

feedback

Chemical feedback

Prevents derangement

in arterial blood gasses

and acid-base balance

(hypoxia, hypercapnia,

hypocapnia)

Reflex feedback

Mainly prevents over-distention

(Hering-Breuer reflex)

VV

Pmus

ChemicalReflex

Polacheck et al. J Appl Physiol 1980;49:609

Cunningham et al. Handbook of Physiology. The Respiratory

system. Vol. 2. 1986; pp. 475–528.

Does mechanical ventilation

affect the operation (effectiveness)

of these feedback systems?

ChemicalReflex

Wakefulness

Sleep (and sedation)

Respiratory response to CO2 during

mechanical ventilation in awake humans

0

50

100

150

200

250

300

350

20 25 30 35 40 45 50 55

% o

f b

ase

lin

e

PaCO2

Pmus

Fr

Data from Georgopoulos et al. Am J Respir Crit Care Med 1997;156:146

f is relatively insensitive to CO2

over a wide range of PCO2

the ventilatory response to CO2 is expressed

mainly by the intensity of respiratory effort

VT/effort per breath

(neuroventilatory coupling per

breath) and not per min is the main

determinant of the respiratory

response to chemical feedback

0

100

200

300

400

20 30 40 50 60

% o

f b

ase

lin

e

PaCO2

• Consequences?

Chemical feedback during mechanical

ventilation - Wakefulness

VT/effort per breath is important

0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

PAV PS AVC

VT/P

TP

-Pm

us

I

(l/c

mH

2O

. sec)

* *

+

+

VT/effort per breath (neuroventilatory coupling) with low and

high respiratory drive at various modes of support (at high assist)

Data from Mitrouska et al. Eur Respir J 1999; 13:873

Low drive

High drive (CO2 challenge)

Hypocapnia

Hypercapnia

and/or distress

PETCO2

22 mmHg

PETCO2

21 mmHg

PETCO2

31 mmHg

Independent on assist level: RR constant on both modes

VT increased ≈ 500 ml VT constant

PtCO2 decreased

by ≈ 10 mmHg

(severe hypocapnia)

PtCO2 constant

Meric et al. Respir Physiol Neurobiology 2014;195:11

PS NAVA

Chemical feedback during mechanical

ventilation - Wakefulness

Consequences

• Modes of support that permit the

intensity of patient effort to be expressed

on the VT delivered by the ventilator

increase the effectiveness of chemical

feedback to regulate PaCO2

ChemicalReflex

Meric et al. Respir Physiol Neurobiology 2014;195:11

Mitrouska et al. Eur Respir J 1999; 13:873

Effective

Chemical feedback during mechanical

ventilation - Wakefulness

Consequences

• Modes of ventilation that impair the

neuroventilatory coupling (VT/effort)

may cause either hypocapnia or

hypercapnia and/or distress.

ChemicalReflex

Meric et al. Respir Physiol Neurobiology 2014;195:11

Mitrouska et al. Eur Respir J 1999; 13:873

Ineffective

Chemical feedback during mechanical

ventilation -Sleep

Observations

• The respiratory rhythm is critically depended on PaCO2

• A drop in PaCO2 by 3-4 mmHg causes apnea

(Datta et al. J Physiol 1991; 440:17, Mezza et al. J Appl Physiol 1998; 84:3)

0

2

4

6

8

10

12

14

16

18

20

10 15 20 25 30 35 40 45 50 55 60 65 70

Ventila

tion (

l/m

in)

PaCO2 (mmHg)

Wakefulness

Sleep

Apnea (Apneic threshold)

During sleep breathing frequency changes minimally

with decreasing PaCO2 up to a point (apneic threshold)

0

2

4

6

8

10

43 44 45 46 47 48

0

5

10

15

20

25

43 44 45 46 47 48

f

Pm

us

PaCO2

Meza et al. J Appl Physiol 1998;85:1928

Meza et al. J Appl Physiol 1998;84:3

16 14 14

Apnea Apnea Apnea

0 3 6

8

During sleep the chemical feedback

controls ventilation exclusively by

changing the respiratory effort per

breath (and not by changing

frequency)

Even during sleep VT/effort per breath

(neuroventilatory coupling per breath) and not

per min is the main determinant of the

respiratory response to chemical feedback

0

2

4

6

8

10

43 44 45 46 47 48

0

5

10

15

20

25

43 44 45 46 47 48

f

Pm

us

PaCO2

16 14 14

Apnea Apnea Apnea

0 3 6

8

Chemical feedback during mechanical

ventilation -Sleep

Consequences

• Modes of support that decrease the VT in

response to any reduction in Pmus

promote ventilatory stability

ChemicalReflex

Meza et al. J Appl Physiol 1998;85:1928

Meza et al. J Appl Physiol 1998;84:3

Chemical feedback during mechanical

ventilation -Sleep

Consequences

• At high assist, modes of ventilation that

impair the neuroventilatory coupling

(VT/effort) may cause apnoeas and trigger

periodic breathing

ChemicalReflex

Meza et al. J Appl Physiol 1998;85:1928

Meza et al. J Appl Physiol 1998;84:3

MESSAGE

The operation (effectiveness) of chemical

feedback during mechanical ventilation depends

mainly on the mode of mechanical ventilatory

support

+ Paw = V’xR + VxE

Volume-time profile

Mechanical

Chemical

Reflex

Behavioral

Feedback

Response of Pmus

to ventilator delivered

breath

Response of

ventilator to Pmus

Pmus

Ventilator and Patient

factors

Georgopoulos D: Principles and Practice of Mechanical Ventilation

(ed. Tobin) 2013: pp 805-820

Reflex feedback

Prevents over-distention

(i.e. Hering-Breuer)

VV

Pmus

ChemicalReflex

If effort drives the ventilator

(PAV, NAVA) this feedback

may control VT

↓Tineural with increasing volume

Polacheck et al. J Appl Physiol 1980;49:609

Paw

EAdi

Volume

Pes

↑↑ assist No assist

(for one breath)

ChemicalReflex

Grasselli et al. Intensive Care Med 2012;38:1224

Rabbits ventilated with NAVA

Theoretically

• Mechanically ventilated patients who express the feedback systems of control of breathing (CHEMICAL and REFLEX feedback) may choose a breathing pattern that limit the risks of lung damage.

ChemicalReflex

Animal data

Allowing animals with acute lung

injury to “control” their ventilatory

pattern (using NAVA) is at least as

protective (and probably more)

as a low VT strategy

(i.e. protective strategy)

Rabbit brain does a good job

Brander et al. Intensive Care Med. 2009;35:1979

Mirabella et al. Crit Care 2014;18:R22

Human brain?

108 patients mechanically ventilated on controlled modes andrecovering from acute respiratory failure were permitted to select theirown breathing pattern by placing them on PAV+ for a maximum periodof 48 hours

2016 Mar 17;228:69-75`

Individual driving pressure (ΔP) were calculated in these patients andexamined how it related to ΔP when the same patients were ventilatedpassively using the currently accepted lung-protective strategy

Studies in passively ventilated ARDS patients have shown that drivingpressure (but not VT) is associated with adverse outcomes

Amato MB et al. N Engl J Med 2015

Laffey et al. Intensive Care Medicine 2016

Sherpa Neto et al. Intensive Care Med. 2016

Human brain?

2016 Mar 17;228:69-75

Main results:

Critically ill patients control the driving pressure to low levels [10

cmH20 (8-12), median (IQR)] bybsizing VT to individual respiratory

system compliance using appropriate feedback mechanisms aimed at

limiting the degree of lung stress

-13

-11

-9

-7

-5

-3

-1

1

3

5

7

9

11

4 6 8 10 12 14 16 18 20 22

Ch

ang

e in

ΔP

(cm

H2O

)

ΔP

PA

V+

-Δ

PC

MV

ΔP during CMV (cmH2O)

87% (58/67)

91% (59/65)

Georgopoulos et al. Respir Physiol Neurobiology 2016 Mar 17;228:69-75

When the lung protective strategy results in high driving pressure (≥15 cmH2O),

the patients’ control of breathing system decreased the driving pressure in the

majority (87%) of measurements.

Message

The patients’ control of breathing system is adept

at protecting the lungs by preventing high driving

pressure, while not unnecessarily restricting tidal

volume when this has no protective value.

-13

-11

-9

-7

-5

-3

-1

1

3

5

7

9

11

4 6 8 10 12 14 16 18 20 22

ChemicalReflex

Provided that:

1) The mode of support does not interfere

with their operation

Doorduin et al. Am J Respir Crit Care Med 2016; Oct 17

Brochard et al. Am J Respir Crit Care Med. 2016 Sep 14

2) Load-independent high respiratory drive

that may override the protective mechanisms,

IS NOT AN ISSUE

Patients selection is crucial

(be very careful in patients with severe

metabolic acidosis, delirium, ongoing

sepsis, etc)

Response of

ventilator to Pmus

Response of Pmus

to ventilator delivered

breath

Direct extrapolation of the ventilator

output to the patient’s control of

breathing system (i.e. input) may be

misleading, guiding inappropriate

therapeutic decisions

OutputInput

Commands

ChemicalReflex

The control of breathing

system is damaged by the

ventilator