contribution to the study of one-lung ventilation …
TRANSCRIPT
CONTRIBUTION TO THE STUDY OF ONE-LUNG
VENTILATION DURING THORACIC SURGERY
Thesis submitted to obtain the degree of Doctor in Medical Sciences
Ghent University, Medical School
Department of Anesthesia
2004
Promoter:
Prof.dr. Eric Mortier
Co-Promoter:
Prof.dr. Jan Poelaert
Laszlo L. Szegedi
This thesis is dedicated to the memory of Gizella I. Bardoczky, M.D., PhD,
D.A.B.A., who introduced me to the exciting world of thoracic anesthesia
ACKNOWLEDGEMENTS
This work would not have been possible without the support of many
people and to them I am extremely grateful:
- The promoters of this thesis: Professor Eric Mortier, chairman of
the Department of Anesthesiology of the Ghent University Hospital, that in
his around the clock schedule still had the time to help me, patiently but
firmly directing my steps towards concretization of this work, and
Professor Jan Poelaert, for his patient help performing cardiac output
measurements and for his scientific expertise. Their experience, wisdom
and valuable advices, made this work possible.
- Professor Alain A. d’Hollander, an understanding, supporting and
encouraging mentor and may I say friend, giving me the opportunity to
follow my own way in clinical research. He stimulated and nurtured the
project from the beginning to the end, spent long hours with me to
exchange ideas and opinions, talk over projects, discussing results. His
constructive ideas were inspiring.
- Professors P. Calle, J. Decruyenaere, F. De Baets, G. Joos, L.
Leybaert, M. Struys, J. Van de Voorde, E. Vandermeersch, F. Vermassen,
members of the examination and accompanying commission, for their
passionate criticism, which changed my intrinsic “operating system”.
I am grateful to all my surgical colleagues and nursing staff from the
Ghent University Hospital and Erasme University Hospital in Brussels.
They were victims of my time consuming work and yet were able to put
up with me during the long hours of these studies and beyond.
I wish to express my gratitude to Professors Luc Barvais and Philippe Van
der Linden and Dr. Anne Ducart, whose criticism during preparation of
some of the manuscripts has been of great help.
I would like to thank all my colleagues from the Department of
Anesthesiology of the Ghent University Hospital, who accepted me and my
work as a friend and not as an intruder: Nadia Den Blauwen, Luc
Herregods, Luc De Baerdemaeker, Piet Cosaert and all the others.
I thank Mrs. Brigitta Locy, Fanny Martens, Hilde De Brabandere, Frieda
Haven, Oona Sinnaeve and Mr. Geert Mets, for their patience (for listening
at my monologues) and for their redactional help.
I am expressing my warm thanks to all my friends, being always there,
encouraging and following the development of these studies with the
keenest interest.
Finally, but before all, I thank my parents, for their love and support.
5
Contents
Abbreviations 8
1. Introduction 11
1.1. Evolution of anesthesia for thoracic surgery and one-lung
ventilation 12
1.2. Development of endobronchial intubation and one-lung ventilation
16
1.3. Methods of lung separation 22
1.3.1. Double-lumen tubes 22
1.3.2. Which size tube for which patients ? 24
1.3.3. Positioning of double-lumen tubes 25
1.3.4. Auscultation and inspection 26
1.3.5. The air bubble method 26
1.3.6. Fiberoptic bronchoscopy 29
1.3.7. Fiberoptic bronchoscopy for left-sided double-lumen tube
positioning 29
1.3.8. Fiberoptic bronchoscopy for right-sided double-lumen tube
positioning 31
1.3.9. Bronchial blockers 32
1.3.10. Single-lumen endobronchial tubes 35
1.3.11. The ABC's of lung isolation 36
1.4. One-lung anesthesia: how it works? 39
1.5. References 42
1.6. Summary 46
2. Aims of the work and general methodology 47
2.1. Aims of the work 48
2.2. General methodology used in the presented studies 52
2.3. References 56
3. Pathophysiology of one-lung ventilation 59
3.1. Introduction 60
3.2. Factors influencing hypoxic pulmonary vasoconstriction 63
6
3.3. Effects of anesthetic agents on hypoxic pulmonary 67
vasoconstriction 67
3.4. Carbon dioxide elimination during one-lung ventilation 70
3.5. Other factors influencing blood flow distribution during one- 71
lung ventilation 71
3.6. Gravity 72
3.7. Pathology caused by endotracheal tubes 74
3.8. Pathology of the patients 78
3.9 Auto-positive end-expiratory pressure during one-lung 80
ventilation 80
3.10. Pathophysiology caused by ventilatory management of one- 84
lung ventilation 84
3.11. Is intrinsic positive end-expiratory pressure harmful? 88
3.12. References 91
3.13. Summary 104
4. Airway pressure changes during one-lung ventilation 105
4.1. Introduction 106
4.2. Methods 107
4.3. Results 110
4.4. Discussion 114
4.5. References 118
4.6.Summary 119
5. Two-lung and one-lung ventilation in patients with chronic obstructive
pulmonary disease: the effects of position and FiO2 121
5.1. Introduction 122
5.2. Methods 123
5.3. Results 126
5.4. Discussion 133
5.5. References 141
5.6.Summary 145
6. Dependent versus nondependent lung ventilation of COPD patients in
supine and lateral positions 147
7
6.1. Introduction 148
6.2. Materials and methods 150
6.3. Statistical analysis 154
6.4. Results 155
6.5. Discussion 162
6.6. References 170
6.7.Summary 175
7. The effects of acute isovolemic hemodilution on oxygenation during
one-lung ventilation 177
7.1. Introduction 178
7.2. Materials and methods 180
7.3. Statistical Analysis 184
7.4. Results 185
7.5. Discussion 191
7.6. References 196
7.7. Summary 199
8. Intrinsic positive end-expiratory pressure during one-lung ventilation of
patients with pulmonary hyperinflation. Influence of low respiratory rate
with unchanged minute volume. 201
8.1. Introduction 202
8.2. Materials and methods 204
8.3. Statistical analysis 208
8.4. Results 209
8.5. Discussion 212
8.6. References 216
8.7. Summary 220
9. General discussion and conclusions 221
9.1. Summary of the results 222
9.2. Limitations of the studies and future perspectives 227
9.3. References 230
10. End remarks 231
11. Summary 233
8
Abbreviations
a/A ratio: Ratio of the arterial and alveolar oxygen tension
ASA: American Society of Anesthesiologists
BSA: Body surface area
bpm: Breaths per minute
CI: Cardiac index
CO: Cardiac output
COPD: Chronic obstructive pulmonary disease
Cst,rs: Static compliance of the respiratory system
DLT: Double lumen endobronchial tube
DPH: Dynamic pulmonary hyperinflation
D-OLV: Dependent one-lung ventilation
EIP: End-inspiratory pause
ETCO2: End-tidal carbon dioxide
FEV1: Forced expired volume in one second
FiO2: Fractional concentration of oxygen in inspired gas
FOB: Fiberoptic bronchoscopy
FRC: Functional residual capacity
FVC: Forced vital capacity
Hb: Hemoglobin level
Hct: Hematocrit
HPV: Hypoxic pulmonary vasoconstriction
IEF: Interrupted expiratory flow
9
IH: Isovolemic hemodilution
IPPB: Intermittent positive pressure breathing
IRB: Institutional Review Board
MAC: Minimum alveolar concentration
MAP: Mean arterial blood pressure
MV: Minute ventilation
ND-OLV: Nondependent one-lung ventilation
OLV: One-lung ventilation
PaCO2: Partial pressure of carbon dioxide in arterial blood
PAO2: Alveolar partial pressure of oxygen
PaO2: Partial pressure of oxygen in arterial blood
P(A-a)O2: Alveolar-arterial oxygen tension gradient
P(a-ET)CO2 Arterial to end-tidal carbon dioxide tension difference
PEEP: Positive end-expiratory pressure
PEEPi: Intrinsic positive end expiratory pressure
Pel,rs: Elastic recoil pressure of the respiratory system
PH: Pulmonary hyperinflation
Ppeak: Peak inspiratory airway pressure
Pplateau: Plateau (end-inspiratory) airway pressure
PvO2: Partial pressure of oxygen in venous blood
Qs/Qt: Shunt ratio
RR: Respiratory rate
RV: Residual volume
SaO2: Saturation of hemoglobin with oxygen in arterial blood
10
SsvcO2: Oxygen saturation of blood collected in the superior vena cava
SvO2: Saturation of hemoglobin with oxygen in venous blood
TE: Expiratory time
TI: Inspiratory time
TLC: Total lung capacity
TLV: Two-lung ventilation
TTOT: Total cycle time
VA/Q: Ventilation-perfusion ratio
VC: Vital capacity
Vr: Relaxation volume
VT: Tidal volume
11
1. Introduction
12
1.1. Evolution of anesthesia for thoracic surgery and one-lung ventilation
Safe anesthesia specifically for thoracic surgery is a relatively new
development in the history of modern anesthetic practice.
Prior to 1900, surgeons rarely ventured into the chest. Thoracotomy was a
hazardous and uncommon operation. It had been recognized for centuries
that respiration rapidly became ineffective once the chest was open,
although the reasons for this were not understood. It has also been known
since 1550 that animals could be kept alive with an open chest using
positive pressure ventilation, but it was not known how to achieve this
effectively in humans. It had been realized that maintenance of a
differential pressure across the lungs during thoracotomy helped to delay
respiratory failure. This led to the development of ingenious positive
pressure headboxes and negative pressure opening chambers with an
airtight seal around the patient’s neck; the airways remained at a
constant positive pressure with respect to the open thoracic cavity while
the patient breathed spontaneously under anesthesia. Positive pressure
ventilation by insufflation of air using bellows and tracheal intubation with
various design of metal tube had been advocated for resuscitation for
drowning and asphyxia in the eighteenth and early nineteenth centuries,
but artificial ventilation was not widely used in anesthesia until the mid-
twentieth century.
13
Inhalational anesthesia had been successfully administered through
tracheal tubes by 1900, although the techniques were difficult and it was
not widely used by the medical profession of the day. The tracheal tubes
available were difficult to insert, and even though laryngoscopes had been
described, they were not in common use. Both tracheal and bronchial
insufflation of anesthetic gas under pressure had been tried for thoracic
work. Anesthesia had even been performed for successful thoracotomy,
using a combination of a metal resuscitation tube and bellows ventilation
with the Fell-O’Dwyer apparatus in 1899. None of these techniques were
widely accepted at the time, although they all contained the roots of
modern thoracic anesthesia (1,2).
Simultaneous advances in several areas were necessary for modern
thoracic anesthesia to develop. The development of safe anesthetic agents
and apparatus for their delivery was vital. An understanding of the
anatomy of the lower airway, the mechanism of normal and pathological
pulmonary gas exchange and lung ventilation was necessary. The concept
of modern “balanced anesthesia”, using a muscle relaxant, was invaluable.
The advances in materials science that allowed the development of rubber
endotracheal and endobronchial tubes, reliable and safe mechanical
ventilators and, more recently, plastic tubes and fiber optics were crucial.
It should also be remembered that concomitant medical advances in areas
such as asepsis, antibiosis and public health were also of great
significance, as the focus of chest disease gradually changed during the
period under consideration from chronic infective lung disease towards
14
mostly malignant disease. Diagnosis methods and surgical techniques
were advancing rapidly at the same time, and increasingly challenging
operations frequently highlighted the inadequacy of the available
anesthetic apparatus (1,2).
The first necessary advance for thoracic anesthesia was tracheal
intubation. While there are many claims to early successes in this field,
the accepted pioneers of tracheal intubation were Ivan Magill and Stanley
Rowbotham, during and after the First World War in England (1,2).
During the same period, American anesthetists were developing also
techniques for tracheal intubation. The idea of attaching an inflatable
balloon or cuff to the tracheal tube was not new, having been described
several times since 1871, but Arthur Guedel and Ralph Waters in the
United States were the first formally to describe the “closed endotracheal
technique for the administration of anesthesia” in 1928. To make their
point, they proved that they were able to anesthetize a dog and
completely immerse both dog and anesthetic apparatus in a tank of water
with no ill effect! (1,2,3).
Thoracic surgery in the 1920s consisted primarily of surgery for
tuberculosis, which was treated by collapsing the affected lung, or for
empyema, which required rib resection and drainage of the infected
cavity. The patients were in poor health, continually producing copious
quantities of purulent sputum. The lung disease would remain unilateral
as long as the patient could cough effectively, but one of the risks of
surgery was the spread of the pus into the “good” lung as the cough reflex
15
was inhibited and the patient positioned for the operation. Anesthesia was
either local or inhalational, with spontaneous respiration throughout,
although anesthetic circuits were beginning to incorporate a reservoir bag
that could be squeezed. Tracheal intubation alone was not the answer to
the secretion problem, but further developments were limited by
anatomy, the lung pathology and the materials available to manufacture
tubes capable of occluding a bronchus safely whilst maintaining a route for
ventilation (2).
16
1.2. Development of endobronchial intubation and one-lung
ventilation
The concept of endobronchial intubation had its roots in the late
nineteenth century. Experimental physiology work on dogs published by
Head in 1889 described endobronchial intubation using a long, cuffed
metal bronchial tube, attached to a short tracheal tube. However, the first
description of “closed endobronchial anesthesia”, or true one-lung
anesthesia, came in 1932, from Joseph Gale and Ralph Waters of Madison,
Wisconsin USA, as a direct extension of earlier cuffed endotracheal tube
work (3).
As with infectious diseases, the development of antibiotics completely
changed the clinical course of infective chest diseases. The treatment of
lung abscess, bronchiectasis, and empyema was revolutionized in the
post- World War II period by the use of sulfonamides (2) and the
availability of penicillin for civilian use (2). The discovery of streptomycin
in 1943, para-amino-salicylic acid in 1946, and isoniazid in 1952 had
equally profound effects on the treatment of tuberculosis. Because of the
discovery of these antibiotics, pulmonary tuberculosis and bronchiectasis
were treated surgically less far often, whereas resection for malignant
disease became more and more common, a trend that has continued to
the present time (4,5).
Major anesthesia and non-anesthesia related advances in the first half of
the twentieth century contributed to the development of intrathoracic
surgery. The major non-anesthesia related advance was the development
17
of techniques for dissecting and ligating individual structures at the hilum
of the lung. Prior to 1930, lung resection was only occasionally performed
using a quickly applied total lesion snare or a tourniquet technique. The
technique required two surgical stages: one to snare the lesion, and then
another several days later to remove the necrotic tissue. Consequently the
technique was fraught with dangers on infection, hemorrhage, and air
leak, resulting always in significant morbidity and mortality. The first
successful pneumectomy using the snare technique was carried out by
Rudolf Nisson in Germany in 1931 for bronchiectasis; he was followed in
1932 by Haight and Alexander. In 1933 Evarts Graham performed a left
snare pneumectomy on a physician with squamous cell carcinoma, and
the individual survived 30 years (2,6,7); before, no patient had survived a
total pneumectomy for malignant tumor of the lung.
Although everyone recognized that Dr. Graham’s success was a milestone,
the development of the individual structure (bronchus, pulmonary artery,
pulmonary vein) ligation technique in 1929 and the 1930s was much more
important because it greatly reduced the incidence and risk of
postoperative bronchial leaks and tension pneumothorax, pulmonary
hemorrhage, and infection from residual necrotic tissue. The individual
ligation technique for lobectomy was first extensively used by Harold
Brunn in 1929 (2,4,8,9) and then subsequently by Edward Churchill
starting in 1938 (10). However, the missing surgical detail in allowing the
individual structure ligation technique to work well was the routine
postoperative use of closed chest thoracostomy drainage with intrapleural
18
suction. The first report of pleural drainage after lobectomy was in 1929
(2,9,10), and for lung abscess in 1936 (3,11). By the end of World War II,
intercostal closed thoracostomy with intrapleural suction had become the
standard primary treatment for most thoracic injuries (12).
The truly important anesthesia-related advances, aside the primordial lung
separation techniques, consisted in the development of intermittent
positive pressure breathing (IPPB). The introduction of IPPB for the
management of intrathoracic operations first required the development of
laryngoscopy, endotracheal intubation, and adequate bellows or pump
machinery. Prior to this time, the primary difficulty in performing
intrathoracic procedures was known as the “pneumothorax problem”: as
soon as the chest of a spontaneously breathing patient was opened, the
lung in question not only collapsed but also moved up and down violently
with each struggling breath (mediastinal flap and paradoxical respiration);
within a few minutes the patient become cyanotic and hypotensive, and
unless the chest was closed quickly, the patient would die. The use of
IPPB (and the catastrophic Copenhagen polio epidemic of 1952) led to a
crash production program of Engström volume ventilators. In 1955, Björk
and Engström in Sweden first described the use of their ventilator for
postoperative respiratory care of poor risk thoracic surgery patients (2,4).
The development of double-lumen tubes (DLT) was a response to the fast
growing capabilities in thoracic surgery, which now required faster, surer
and simpler lung separating methods. The Björk and Carlens
bronchospirometric DLT was first used during anesthesia in 1950. The
19
next decades were followed by the introduction of different types of DLTs
with variable capabilities (13,14) (Table 1.1.).
Until the advent of fiberoptic bronchoscopy (FOB), the method of
placement of these tubes within the tracheobronchial tree was blind.
The 1970 witnessed a phenomenal explosive increase in monitoring
patients intra- and postoperatively. The majority of monitors used actually
were introduced in this period.
Positive end-expiratory pressure (PEEP) was introduced into clinical
practice almost 40 years ago (15) and has proved to be a rapid and
relatively high-benefit, low risk method for increasing the oxygenation
capability of severely diseased lungs during thoracic anesthesia.
Direct examination of the tracheobronchial tree with flexible FOB has
greatly facilitated the diagnosis, staging and management of pulmonary
neoplasm as well as many other lung diseases (16). With particular
reference to thoracic anesthesia, the FOB allows for placement of DLTs
and endobronchial blockers with great precision and accuracy and at
relative low risk for the patients.
Prevention of lung carcinoma and efforts to discontinue smoking will
certainly continue to increase and be vigorous in the future. Interest will
continue to be high in multimodal therapy (chemotherapy, irradiation,
surgical excision) of carcinomas. A promising approach is the
administration of immunotherapy. Lasers may be used to remove lesions
without damage to the surrounding tissues.
20
Anesthesia for thoracic surgery has gone through a remarkable evolution
over the last 60 years and moved from performing simple chest wall
surgery in poorly monitored patients to performing extremely complicated
intrathoracic procedures, with control over each lung independently, and
on more and more compromised patients.
The last reported mortality rates of a pneumectomy and lobectomy are
only 6 and 3 per cent respectively (17).
Considering that the first modern pneumectomy was performed only six
decades ago, thoracic anesthesia represents a subspecialty that has had a
rapid and dramatic development. The focus in today’s research will
determine to a large extent the types of advances that will be made in the
future.
21
Table 1.1. Development of double-lumen tubes (modified from Benumof JL. History of Anesthesia
for Thoracic Surgery. In Benumof JL (ed): Anesthesia for thoracic surgery, ed 2. Philadelphia, WB
Saunders, 1995, pp 1-14, with permission)
Date Name Distinctive characteristics
1950 Carlens Double lumen catheter with two inbuilt curves; tracheal
and a bronchial cuff for left main bronchus, carinal hook
and cross-sectional shape: oval in horizontal plane
1959 Bryce-Smith Modification of the Carlens catheter with no carinal hook
1960 Bryce-Smith and Salt Right-sided version of the Bryce-Smith tube, possessing
slit in endobronchial cuff
1960 White Right-sided version of the Carlens catheter, possessing slit
in endobronchial cuff and a carinal hook
1962 Robertshaw Right and left double-lumen tubes; modification of the
Carlens catheter with a larger lumen and hence low
resistance to gas flows; slotted right endobronchial cuff;
no carinal hooks; cross-sectional shape: D-shaped in the
horizontal plane
1979 National Catheter
Corporation
Right and left Robertshaw-type disposable double-lumen
tube made of clear tissue implantable plastic with low-
pressure, high-volume cuffs
22
1.3. Methods of lung separation
1.3.1. Double-lumen tubes
In the past, bronchial blockers and single lumen endobronchial tubes have
been used to achieve lung isolation. These tubes are seldom used today
due to technical obstacles, inability to remove secretions from the
nonventilated lung, and less than satisfactory performance. In modern
practice, DLTs are most widely used. In 1949 Carlens designed the first
DLT for selective pulmonary spirometry (Figure 1.1.). The tube was left-
sided. It had a carinal hook and consisted of two round lumens bundled
together. This design had a large cross-sectional deadspace that
presented a high resistance to airflow and made suctioning difficult.
Figure 1.1. The Carlens double-lumen tube and its placement
23
In modern practice, disposable DLTs of a Robertshaw design are most
widely employed. These tubes have a fixed curvature, are without carinal
hook to avoid tracheal laceration, and reduce the likelihood of kinking. The
internal diameter is large and is composed of two D-shaped lumens that
reduce the deadspace and the resistance to airflow at equivalent external
diameters. Numerous manufacturers produce disposable DLTs of
Robertshaw design (Figure 1.2.). They are made of polyvinyl chloride
(PVC) and are available in French (Fr) sizes 28, 32, 35-41. While the basic
features are similar for the products of all manufacturers, there may be
some differences in cuff shape and location. The design of the bronchial
cuff is of the “rolling-pin” type (“O”- shaped lumen) for the Rüsch DLTs
Figure 1.2. A left-sided double lumen tube of Robertshaw design (Rüsch).
(Duluth, Georgia, USA) and donut shaped for the Broncho-Cath DLTs
manufactured by Mallinkrodt (Athlone, Ireland). Tubes made by Sheridan
and Portex have a round, ball-shaped bronchial cuff. The bronchial cuff is
commonly colored blue to permit easy identification by fiberoptic
bronchoscopy. The internal lumens of the 35, 37, 39 and 41 Fr PVC DLTs
24
are 5, 5.5, 6 and 6.6 mm, respectively. The internal diameter of the DLTs
is important when fiberoptic bronchoscopy or the use of a tube exchanger
are considered. A DLT with an internal diameter of at least 4 mm is
recommended for bronchoscopy.
1.3.2. Which size tube for which patients ?
A 37 Fr DLT can be used in most adult females, while a 39 Fr DLT is used
in the average adult male. It is important to keep in mind that during
insertion the disposable PVC DLTs are not designed to be pushed until
resistance is encountered; such action may result in a high incidence of
upper lobe obstruction. DLTs are grouped in French (external
circumference); since they are not circular in cross section, it is not useful
to describe them in diameter. If the tube is too large it will cause trauma
and cuff over inflation with inadequate isolation. If the tube is too small it
will be easily dislocated with malposition. It was also suggested that a
chest X-ray is used to determine, at the level of the clavicles, the size of
the trachea and the left and/or right mainstem bronchus. It was showed
that a tracheal diameter ranging from 15 to 18 mm will accommodate a
DLT from 35 to 41. The left-sided DLTs are most widely used in clinical
practice. Hanallah et al. (18) suggested a method of choice of the size of
the DLT, based upon sex and height of the patient. Brodsky et al. (19)
calculated that the average tube depth, measuring from the central
incisor, is 29 cm for an adult of medium height (170 cm) and will increase
25
or decrease by approximately 1 cm for each 10 cm change in height.
Recently, Mallinckrodt introduced two new sizes of DLT: a 32 Fr DLT for
small adults, and a 28 Fr DLT for adolescents.
1.3.3. Positioning of double-lumen tubes
As with tube size, there is little agreement among anesthesiologists who
do a large volume of thoracic cases on the optimal method to place a DLT.
Commonly used variations following laryngoscopy include: tube rotation
with or without the stylet; tube advancement with or without the stylet;
head in neutral position or right lateral rotation/flexion; and fiberoptic
bronchoscopy during or immediately after intubation or after patient
positioning.
Repositioning is then done under direct vision with the bronchoscope and
lateral flexion or rotation of the patient's neck is occasionally required at
this stage to get a misplaced DLT to enter the left main bronchus.
Bronchoscopy will need to be repeated after patient positioning
Until several years ago, following the insertion of a DLT, auscultation and
inspection were the only reliable methods to check DLT position. With
severe chronic obstructive pulmonary disease (COPD) it is often practically
impossible to verify the DLT position with auscultation. The subsequent
introduction and current availability of fiberoptic bronchoscopy confirm the
need for this technique to correct the fine malpositions of the DLT.
26
1.3.4. Auscultation and inspection
Auscultation and inspection are carried out as follows. After intubation, the
tracheal cuff should be inflated first and equal bilateral breath sounds
should be confirmed with both cuffs inflated. The professional should
develop his own method to check for correct positioning of the DLT. A
simple way consist of verifying that the tip of the bronchial lumen is
located in the desired bronchus. The tracheal lumen is clamped at the
level of the connector while the patient is being ventilated. Usually,
inspection will reveal unilateral ascent of the one hemithorax and
unilateral ventilation is confirmed by auscultation.
1.3.5. The air bubble method
The use of FOB to determine DLT position does not provide evidence or
guarantee that the two lungs are functionally separated (i.e. against a
fluid and/or air pressure gradient). There are times, such as during the
performance of unilateral pulmonary lavage, when the anesthesiologist
must be absolutely certain that functional separation has been achieved.
Complete separation of the two lungs by the bronchial cuff can be
demonstrated by clamping the connecting tube to the tracheal lumen,
proximal to the suction port and attaching a small tube (i.e. intravenous
extension tubing) to the open tracheal suction port (by appropriate
27
adaptors)(20). The free end of this tube is submerged in a beaker of
water. When the intubated lung (left or right) is selectively ventilated or
exposed to any desired distending pressure, and the bronchial cuff is
sealed, no air will escape around this cuff and out the open suction port of
the tracheal lumen; thus no bubbles will be observed passing through the
beaker of water. When the respective lung is ventilated or exposed to any
desired distending pressure, and the bronchial cuff is not adequately
sealed, air will escape around the bronchial cuff and out the tracheal
suction port; thus air bubbles will be observed passing through the beaker
of water. The Figure 1.3. illustrates the air bubble method for detection of
cuff seal/leak.
28
Figure 1.3. Illustration of the bubble-method for checking for air leaks in a left-sided double-
lumen endotracheal tube. (From Benumof JL. Separation of the two lungs (Double-lumen tube and
bronchial blocker). In: Benumof JL (ed): Anesthesia for Thoracic Surgery. Ed 2. Philadelphia, WB
Saunders, 1995, pp 330-389; with permission.)
29
1.3.6. Fiberoptic bronchoscopy
The introduction of FOB into clinical practice was perhaps the most
important advance in confirming the proper positioning of DLT. The
importance of FOB was described in a recent report in the UK (21). The
report made disturbing reading. Problems with DLT were a feature in 30%
of deaths reported, even though patients were managed by senior
anesthesiologists. Problems ranged from requirements for multiple tube
changes to prolonged periods of hypoxia and hypoventilation. Of particular
note in the report is the fact that no anesthesiologist reported the use of
FOB to confirm the correct position of the DLT (21). Several sizes of FOB
are available for clinical use, with external diameters of 5.6, 4.9 or 3.9
mm (Olympus, Pentax, Storz). The 3.9 mm diameter FOB can easily be
passed through a 37 Fr DLT or larger tubes, while it fits tightly through a
35 Fr tube.
1.3.7. Fiberoptic bronchoscopy for left-sided double-lumen tube
positioning
When the bronchoscope is passed down the right lumen of the left-sided
tube, the endoscopist should see a clear straight-ahead view of the
tracheal carina and the upper surface of the blue left endobronchial cuff
just below the tracheal carina. Excessive pressure in the endobronchial
30
cuff, as manifested by tracheal carinal deviation to the right and
herniation of the endobronchial cuff over the carina, should be avoided.
When the FOB is passed down the left lumen of the left-sided tube, the
endoscopist should see a very slight left luminal narrowing and a clear
straight-ahead view of the bronchial carina off in the distance (22) (Figure
1.4.).
Figure 1.4. Fiberoptic bronchoscopy for positioning of left-sided double-lumen tube (DLT).
31
1.3.8. Fiberoptic bronchoscopy for right-sided double-lumen tube
positioning
When the FOB is passed down the left (tracheal) lumen, the endoscopist
should see a clear, straight-ahead view of the tracheal carina and the right
lumen going off into the right mainstem bronchus. When the fiberoptic
bronchoscope is passed down the right (bronchial) lumen, the endoscopist
should see the bronchial carina off in the distance; when the FOB is flexed
laterally and cephalad and passed through the right upper lobe ventilation
slot, the right upper bronchial orifice should be visualized (22) (Figure
1.5.).
Figure 1.5. Fiberoptic bronchoscopy for positioning of right-sided double-lumen tube (DLT).
32
1.3.9. Bronchial blockers
As the clinical indications for lung isolation have expanded, the limitations
of DLTs have become more evident and there has been a renewed interest
in finding alternative methods to provide lung isolation. Because of the
fixed distances and sizes of the tracheal and bronchial orifices and cuffs of
DLTs they function optimally in patients with normal airway anatomy but
have very limited adaptability in patients with abnormal upper or lower
airways. Bronchial blockers are useful in many of these situations.
Blockers can be placed through or external to existing single-lumen
endotracheal (ET) tubes and this can obviate the need for a tube change
both at the start or at the end of surgery, which may be useful in trauma
or other difficult airway cases. With single-lumen tubes 7.5 mm internal
diameter (ID) or larger, both the blocker and a 4-mm diameter fiberoptic
bronchoscope scope can be passed intra-luminally for positioning. With
smaller ID ET tubes the blocker is passed through the glottis beside the ET
tube. Blockers can provide either lobar or lung isolation. The major
problem with all blockers to date is that the reliability of isolation is not as
good as with a DLT. Bronchial blockers tend to become dislodged intra-
operatively (23) and this is particularly a concern in the presence of
infected secretions. Also, they do not allow easy access to the non-
ventilated lung for suctioning, to verify position, or to aid deflation. This
can be a problem during thoracoscopic surgery when there is less access
33
for the surgeon in the operative hemithorax and less tolerance for
incomplete collapse.
General principles to increase the isolation success rate with blockers are:
1) position the deflated blocker in the correct bronchus while the patient is
supine before turning, as it can be difficult to manoeuver the blocker into
the nondependent bronchus in the lateral position; 2) use a blocker
preferentially for left vs. right thoracotomies and position the blocker as
distal as possible in the mainstem bronchus; this increases the margin of
safety to maintain isolation if the blocker moves intra-operatively; 3) when
possible use blockers for non-pulmonary intra-thoracic procedures such as
esophageal or vascular cases as there is less lung manipulation and less
accidental dislodgement of the blocker; 4) use blockers for open
thoracotomies vs. thoracoscopies as the surgeon can more easily move an
incompletely deflated lung out of the field if the chest is open; 5) use a
video camera and monitor attached to the bronchoscope to position the
blocker as this allows an assistant to help in controlling the bronchoscope.
Positioning and placing a blocker requires four hands.
The most widely used blocker in the past two decades has been an 8F
Fogarty venous embolectomy catheter with a 4 or 10 ml balloon (24). This
is a closed-end catheter and is not specifically designed as a blocker.
There has been a large effort in the industry to improve on this. The
Univent tube is a single-lumen tube with a blocker enclosed in a separate
channel within the tube (25). Recently, the Univent blocker (Figure 1.6.)
has been redesigned to make it more flexible and easier to manipulate.
34
The Univent tubes are silicon and tend to be relatively stiff and, although
they have been described for use in difficult airways (26), they have not
received universal acceptance.
Figure 1.6. The Univent bronchial blocker (BB)
The newest commercially available blocker is the Arndt Catheter (Cook
Inc., Bloomington, USA).(27) (Figure 1.7.). The blocker balloon has a
modified pear shape which seems to make it more stable in the bronchus.
As with the Univent, there is a narrow suction channel in the blocker. Also,
it comes with an elegantly designed three-way connector that permits
separate air-tight access for the catheter, the bronchoscope and the
35
anesthetic circuit. However the wire guide-loop, that is used to position
the catheter in the bronchus with a bronchoscope, can be awkward to
manipulate particularly with an ET tube < 8 mm ID. The newer blockers
are generally more expensive than Fogarty catheters or DLTs.
Figure 1.7. The Arndt endobronchial blocker
1.3.10. Single-lumen endobronchial tubes
The original method of lung isolation was the distal advancement of a
single-lumen tube into the desired mainstem bronchus (28). Some
anesthesiologists continue to use a standard 32 cm long 7.5 mm ID ET
tube advanced with bronchoscopic guidance as an endobronchial tube
when needed. Naturally, there is no access to the non-ventilated lung for
suctioning, for continuous positive airway pressure (CPAP), to confirm
positioning, etc. Some surgeons prefer single-lumen endobronchial tubes
36
to DLTs for procedures involving the carina since they are more flexible,
permitting better mobilization and exposure (29). Bilateral endobronchial
tubes can be used with lesions close to the carina (30).
1.3.11. The ABC's of lung isolation
Which ever method of lung isolation is chosen for a particular patient in a
specific clinical situation, there are several general principles of lung
isolation that should be followed to improve the safety and reliability of the
procedure. These are refered as the ABC's of lung isolation. They are:
know the tracheo-bronchial Anatomy, always use the fiberoptic
Bronchoscope and examine the Chest x-ray and CT scan preoperatively.
Anatomy
Lung isolation requires a thorough knowledge of bronchial anatomy. Just
as advances in invasive monitoring have mandated that anesthesiologists
develop a more complete understanding of vascular anatomy to achieve
reliable central venous access, lung isolation requires detailed knowledge
of bronchial anatomy. Minor variations in subsegmental anatomy are
common. However, the anatomy of the lobar and segmental bronchi is
consistent enough to use as a guide for positioning endobronchial tubes
and blockers.
The left mainstem bronchus is narrower (mean adult diameter 13 mm)
than the right (16 mm) and makes a more acute angle at the carina (45
vs. 30) this means that it is generally more difficult to get a tube or
37
blocker into the left side. However, the left mainstem bronchus is longer
(48 ± 8 mm) than the right (21 ± 8 mm) and thus there is a larger
margin of safety in positioning tubes or blockers on the left.
Bronchoscope
It was pointed out in a recent editorial that it would be very difficult to
defend an anesthesiologist who was involved with a complication during
thoracic surgery if a DLT or blocker was placed without confirmation by
bronchoscopy (31). Similarly, it would be difficult for an institution where
elective thoracic surgery is done on a regular basis to defend itself if
suitable fiberoptic equipment was not made available (31).
The tracheal carina and the orifices of all lobes should be verified each
time the bronchoscope is used. The ability to use a fiberoptic
bronchoscope to recognize normal and abnormal tube placement is a skill
that all anesthesiologists who manage thoracic cases should possess. Like
any skill it is best learned under appropriate guidance in elective
situations.
Chest x-ray and CT scan
All anesthesiologists are familiar with the clinical assessment of the upper
airway for the difficulty of endotracheal intubation. In a similar fashion,
each thoracic surgical patient should be assessed for the difficulty of
endobronchial intubation. The single most important predictor of difficult
endobronchial intubation is the plain chest x-ray (32). Major abnormalities
of tracheal or bronchial anatomy due to congenital malformations,
distortion by tumor, etc., are usually readily visible. The anesthesiologist
38
must view the chest x-ray him/herself prior to induction since neither the
radiologist's nor the surgeon's report of the x-ray is made with the specific
consideration of lung isolation in mind. It is also worthwhile to examine the
CT scan of the chest when available since endobronchial problems that can
lead to problematic lung isolation that may not be evident on the plain
chest film can sometimes be seen on the CT scan (33).
39
1.4. One-lung anesthesia: how it works?
In the context of one-lung anesthesia and ventilation, the lung collapse
must be both complete and well tolerated by the patient. Although his
concept seems simple, a number of clinical details frequently make the
difference between success and failure. Lung isolation or lung separation
implies the ability to ventilate one lung independent from the other and to
restrict passage of blood or fluids (watertight seal) from one lung to the
other. One-lung ventilation implies not only functional lung separation, but
also adequate ventilation and oxygenation. Thus, the three endpoints of
one-lung ventilation are: optimal position of the DLT or bronchial blocker,
functional lung separation, and adequate ventilation and oxygenation
(Figure 1.7.).
Various overlapping of these subsets can and do occur. For example,
adequate position of the DLT and bronchial blocker does not ensure
functional lung separation, and adequate OLV can be achieved without
optimal DLT and bronchial blocker position. By identifying the exact nature
of difficulties, the anesthesiologist can implement therapy without wasting
time on DLT repositioning, cuff volume manipulations or ventilation
changes when they are not part of problem. Table 1.2. describes the
clinical problem in each area of Figure 1.8. and lists examples of causes
and solutions.
40
Figure 1.8. Overlap of clinical endpoints in one-lung anesthesia
(A description of the clinical problems in each area of this figure is listed in the table below)
Optimal DLT position Adequate OLV and oxygenation Functional lung separation
Table 1.2. Clinical conditions during one-lung anesthesia
Area Example Typical solution
A No airtight cuff seal More air in cuff or larger DLT
B Left DLT too far in Position DLT optimally
C Right DLT cuff occluding right upper
bronchus
Position DLT optimally
D Hypoxemia – Obstruction of the
ventilating lumen of the DLT
100%FiO2/CPAP/PEEP/TLV/Consider
alternative lung separation technique
E No problem
(The Figure 1.8. and the Table 1.2. are reproduced from: Benumof JL. Anesthesia for thoracic
surgery, ed 2. Philadelphia, WB Saunders, 1995, with permission)
41
In the settings of functional lung separation, the nondependent, operative
lung is frequently suctioned, but this maneuver does not greatly assist in
facilitating atelectasis for optimal surgical exposure. It is important to
distinguish between a lung deflating slowly and one being ventilated. The
distinction can be made visually or with the cuff seal method (positive
pressure test or bubble method).
The primary mechanism by which the lung becomes atelectatic is
absorption atelectasis, which is achieved most rapidly with 100% oxygen
rather than, air-oxygen mixture and when the amount of the gas to be
resorbed is the least (34). Suction is of limited utility during good lung
separation because the trapped gas in the lung is distal to collapsible
airways (try applying suction to a Penrose drain). If, despite best efforts,
complete lung separation cannot be accomplished and gas is introduced
into the ipsilateral lung with each breath, then continuous suction may be
helpful in evacuating the gas as it continually enters.
In the patient with a difficult airway who requires lung separation, the
concern for lung separation is secondary to securing the airway.
It should be remembered that one-lung anesthesia adds to the technical
complexity of anesthesia. It should not be pursued at all cost, and never
justifies making patients hypoxic.
42
1.5. References
1. Barry JES, Adams AP. The development of thoracic anaesthesia. In
Hurt R. (ed.) The History of Cardiothoracic Surgery from Early
Times. Parthenon Publishing, 1996
2. Benumof JL. History of Anesthesia for Thoracic Surgery In
Benumof JL (ed): Anesthesia for thoracic surgery, ed 2.
Philadelphia, WB Saunders, 1995, pp 1-14
3. Gale JW, Waters RM. Closed endobronchial anesthesia in thoracic
surgery. J Thorac Surg 1932; 1: 432-7
4. Lee G. History and equipment: the evolution of endobronchial
apparatus for one-lung ventilation and anaesthesia. In Ghosh S,
Latimer RD (eds): Thoracic anesthesia: principles and practice, ed
1, Oxford, Butterworth-Heinemann, 1999, pp 1-23
5. Gothard JWW, Branthwaithe MA. History. In Gothard JWW,
Branthwaithe MA (eds): Anesthesia for thoracic surgery. Oxford,
Blackwell Scientific Publications, 1982, pp 1-8
6. Brewer LA III. The first pneumectomy. J thorac Cardiovasc Surg
1984; 88: 810-26
7. Graham AE, Singer JJ. Successful removal of the entire lung
carcinoma from the bronchus. JAMA 1933; 101: 1371-4
8. Brunn H. Surgical principles underlying one-stage lobectomy. Arch
Surg 1929; 18: 490
9. Meyer JA: Hugh Morriston Davies and and lobectomy for cancer,
1912. Historical vignette. Ann Thorac Surg 1988; 46: 472-4
43
10. Churchill E, Belsey RHR. Segmental pneumectomy in
bronchiectasis: Lingula segment of upper lobe. Ann Surg 1939;
109: 481-99
11. Wilkins EW Jr. Acute putrid abscess of the lung. Classics in thoracic
surgery. Ann Thorac Surg 1987; 44: 560-1
12. Wagner RB, Slivko B. Highlights of the history of nonpenetrating
chest trauma. Surg Clin North Am 1989; 69: 1-14
13. Carlens E. A new flexible double-lumen catheter for
bronchospirometry. J Thorac Surg 1949; 18: 742-6
14. Robertshaw FL. Low resistance double-lumen endobronchial tubes.
Br J Anaesth 1962; 34: 576-9
15. Ashbaugh DG, Bigelow DB, Petty TL et al. Acute respiratory
distress in adults. Lancet 1967; 2: 319
16. Sackner MA. Bronchofiberoscopy. Am Rev Resp Dis 1975; 111:
62-88
17. Ginsberg JR, Hill DL, Eagan TR, et al. Modern thirty-day operative
mortality for surgical resections in lung cancer. Thorac Cardiovasc
Surg 1983; 86: 654-8
18. Hannallah MS, Benumof JL, McCarthy PO, et al: Comparison of
three techniques to inflate the bronchial cuff of left polyvinyl
chloride double-lumen endobronchial tubes. Anesth Analg 1996;
82: 867-69
44
19. Brodsky JB, Benumof JL, Ehrenwerth G, Ozaki GT. Depths of
placement of left double-lumen endobronchial tubes. Anesth Analg
1991; 73: 570-2
20. Benumof JL. Separation of the two lungs (double-lumen tube and
bronchial blocker intubation). In Benumof JL ed. Anesthesia for
thoracic surgery. Philadelphia. WB Saunders. 1995
21. Sherry K. Management of patients undergoing oesophagectomy.
In: The Report of the National Confidential Inquiry into
Perioperative Deaths. Gray AJG (ed). London. 1996-97, 57-61
22. Smith B, Hersch N, Ehrenwerth J. Sight and sound: can double-
lumen endotracheal tubes be placed accurately without fiberoptic
bronchoscopy? Br J Anaesth A987; 58: 1317
23. Campos JH, Reasoner DK, Moyers JR. Comparison of a modified
double-lumen endotracheal tube with a single-lumen tube with
enclosed bronchial blocker. Anesth Analg 1996; 83: 1268–72
24. Ginsburg RJ. New technique for one-lung anesthesia using an
endobronchial blocker. J Thorac Cardiovasc Surg 1981; 82: 542–6
25. Inoue H. New device for one lung anesthesia, endotracheal tube
with movable blocker. J Thorac Cardiovasc Surg 1982; 83: 940
26. Ransom ES, Carter SL, Mund GD. Univent tube: a useful device in
patients with difficult airways. J Cardiothorac Vasc Anesth 1995; 9:
725–7
45
27. Arndt GA, Buchika S, Kranner PW, DeLessio ST. Wire-guided
endobronchial blockade in a patient with a limited mouth opening.
Can J Anesth 1999; 46: 87–9
28. Kubota H, Kubota Y, Toyoda Y, Ishida H, Asada A, Matsuura H.
Selective blind endobronchial intubation in children and adults.
Anesthesiology 1987; 67: 587–9
29. Newton JR, Grillo HC, Mathisen DJ . Main bronchial sleeve resection
with pulmonary conservation. Ann Thorac Surg 1991; 52: 272–6
30. Lobato EB, Risley WP, Stolzfus DP. Intraoperative management of
distal tracheal rupture with selective bronchial intubation. J Clin
Anesth 1997; 9: 155–8
31. Pennefather SH, Russel GN. Editorial III. Placement of double-
lumen tubes: time to shed light on an old problem. Br J Anaesth
2000; 83: 308-10
32. Saito S, Dohi S, Tajima K. Failure of double-lumen endobronchial
tube placement: congenital tracheal stenosis in an adult.
Anesthesiology 1987; 66: 83–6
33. Bayes J, Salter EM, Hadberg PS, Lawson D. Obstruction of a
double-lumen tube by a saber-sheath trachea. Anesth Analg 1994;
79: 186–8
34. Nunn JF. Nunn’s applied respiratory physiology, 4th ed. Oxford.
Butterworth-Heinemann, 496
46
1.6. Summary
The history of thoracic anesthesia and endobronchial intubation reflects
contemporany progress in other areas of medicine, and the associated
technology. Thoracic anesthesia represents a subspeciality that has had a
rapid and dramatic development. The double-lumen tube remains the
standard means of producing lung separation, but other possibilities
should be considered also. Thoracic anesthesia has moved from
performing simple chest wall surgery without a secure airway in poorly
monitored patients who were deeply anesthetized with inherently
dangerous techniques to performing extremely complicated intrathoracic
procedures, with firm control over each lung independently, in very well
monitored patients who are anesthetized with safe techniques.
47
2. Aims of the work and general methodology
48
2.1. Aims of the work
Any operation is a team effort, and nowhere is this team effort more
fundamental than in thoracic surgery. Increasingly complex operations are
performed on more and more seriously compromised patients, because of
the development of new surgical techniques and the anesthesiologists’
awareness of surgical needs and requirements to provide a safe surgical
field. The management of some problematic patients having thoracic
surgery is among the most difficult challenges facing the alliance of the
surgical team and the anesthesiologist.
In order to facilitate thoracic surgery, the single most important and
valuable anesthetic technique is one-lung ventilation (OLV). In the
technique of OLV, one lung is mechanically ventilated while the other is
either occluded, or open to the atmosphere. The Chapter 3 of this thesis
gives a review of the possible pathophysiological problems encountered
during OLV.
Double-lumen endotracheal tubes (DLT) are widely used for OLV during
anesthesia and in the critical care settings. Properly used it provides
ample exposure by collapsing the surgical lung. This allows for less
retraction and better anatomical exposure of lung parenchyma and hilar
structures. There are special circumstances where the correct position of
the DLT is critical (hemorrhage, bronchopleural fistula, lung
transplantation, unilateral bronchoalveolar lavage), but failure to collapse
49
the would be operated lung can compromise any thoracic operation.
Impaired ventilation of the dependent lung, inappropriate deflation of the
nondependent lung, may not only cause a slowly deteriorating oxygen
saturation, but also influence the success of the surgery. Complications
due to DLT malposition are one of the main causes of adverse events
occurring in as many as 20-30 % of patients during thoracic anesthesia
(1). These problems usually manifest themselves as cross-ventilation of
the two lungs during OLV, the inability to deflate the operated lung and
collapse of one or more lobes of the ventilated lung. Despite fiberoptic
bronchoscopic confirmation of DLT position, during surgical manipulation
the DLT moves easily and quickly out from the correct position, producing
obstruction, and a modification in the measured ventilatory data
(inspiratory and expiratory airway pressures, tidal volume). Accordingly, it
is possible that these modified ventilatory data may help to determine the
correct positioning of the DLT or promptly detect intraoperative
displacement.
In the first study (Chapter 4), it was examined how the displayed
ventilatory data are influenced during OLV and how they contribute in
assessing the position of the DLTs.
A variety of thoracic procedures like double-lung transplantation (2), lung
volume reduction surgery (3,4,5), or minimally invasive coronary artery
surgery (6) are routinely performed with patients in the supine instead of
the traditional lateral position. During these interventions, OLV is
50
inherently required, generally inducing some degree of hypoxemia. The
major protective mechanism against hypoxemia is hypoxic pulmonary
vasoconstriction (HPV) (7,8), but in the lateral position, gravity-induced
blood flow redistribution must also be considered (7,9). Data comparing
gas exchange in the supine versus lateral position during OLV are lacking
(10) as well as the evidence that gravity must be considered also as a
major protective mechanism against hypoxemia in patients with pre-
existing pulmonary hyperinflation. In the second and third (Chapters 5
and 6) study, we compare the influence of position (supine vs. lateral) on
gas exchange during OLV in order to demonstrate that gravity plays an
important, major role in blood flow redistribution during OLV in patients
with pulmonary hyperinflation.
The isovolemic hemodilution (IH) technique lowers a patient's hemoglobin
at the start of the surgery, so fresh units of the patients' blood are
available when needed. Data concerning the influence of IH on arterial
oxygenation during OLV for thoracic surgery are lacking. The surgical
importance of such a study is also not negligible: it may reproduce an
acute hemorrhage during dissection, compensated with i.v. solutions when
packed red cells are not immediately available. The purpose of the study
presented in the chapter 7 of this thesis is to assess the effects of mild to
moderate isovolemic hemodilution on gas exchange during OLV, and to
see if there is any difference in patients with normal lung function and
51
patients with pulmonary hyperinflation and COPD, who are hemodiluted
during OLV.
During mechanical ventilation, conditions that impede expiratory flow
(increased airway resistance and the additional resistance of the
endotracheal tube) (11,12), or inadequate ventilatory settings (13) may
predispose to dynamic pulmonary hyperinflation (DPH) and intrinsic
positive end-expiratory pressure (PEEPi).
There are few studies about the mechanical characteristics of the
respiratory system during OLV (14,15) for thoracic surgery in patients
with preoperative pulmonary hyperinflation. The occurrence and
magnitude of PEEPi and DPH during thoracic surgery has been already
established; however, the ventilatory methods of lowering or avoiding this
phenomena, as well as the clinical importance of PEEPi and DPH during
thoracic surgery has not been evaluated yet. Thus, in the last part of this
work, (Chapter 8) we focused our research on how PEEPi may be
influenced during OLV by manipulating ventilatory variables.
52
2.2. General methodology used in the presented studies The below presented studies were performed after Institutional Review
Board (IRB) (Erasme University Hospital, Brussels, Belgium and Ghent
University Hospital, Ghent, Belgium) approval and informed and written
consent, in patients (n=167) with normal lung function and in patients
with stable chronic obstructive pulmonary disease with variable degrees of
airflow obstruction and pulmonary hyperinflation, scheduled for elective
lung surgery.
Preoperatively, the pulmonary function was tested in the sitting position,
including spirometry and static lung volumes determined by
plethysmography (MasterScreen Body™, Jaeger and Toennies™, Erich
Jaeger GmbH, Hoechberg, Germany).
Following the definition of the European Respiratory Society (16), patients
whose preoperative functional residual capacity (FRC) exceeded 120% of
the predicted value were considered as patients with pulmonary
hyperinflation.
Patients with cardiac or renal function impairment were not included in the
studies.
Anesthesia was conducted in a standardized manner. All patients were
given alprazolam 0.5 mg p.o., approximately 60 to 90 minutes before
arrival in the operating theatre.
A thoracic epidural catheter was inserted at the mid-thoracic level (Th6-
Th9) to assure analgesia during and after surgery. A test dose of 3 ml 2%
lidocaine with epinephrine 1/200000 was used; the initial dose (8 ml 2%
53
lidocain with 100 µg fentanyl) was given only after the end of the studies
followed by a continuous infusion of bupivacaine 0.5%, at a rate of 2-5
ml/h.
Induction and maintenance of general anesthesia was achieved with
fentanyl (100-150 mcg) and propofol (initial dose 2 mg/kg, followed by
continuous infusion of 3-5 mg.kg-1.h-1) and in some cases inhaled
isoflurane (not exceeding 1 MAC). Pancuronium or cisatracurium was used
to allow tracheal intubation and to maintain neuromuscular blockade
throughout surgery. The neuromuscular blockade was assessed by regular
measurements of post-tetanic count during the procedure.
Electrocardiogram, invasive arterial blood pressure, central venous
pressure and arterial oxygen saturation were monitored continuously.
Expired end-tidal carbon dioxide tension (ETCO2), flow-volume and
pressure-volume loops were also continuously followed.
In all patients, the bronchus of the dependent lung was intubated with a
disposable DLT (Broncho-cath™, Mallinckrodt Laboratories, Athlone,
Ireland). Tube size was chosen according to a formula based on patients’
sex and height (17). The position of the DLT was ascertained by using
FOB, in the supine and in the lateral positions.
The dependent lung is defined as being the lower lung (in lateral
decubitus), while the nondependent lung is the upper lung when the
patient is in lateral position. In the studies presented in this thesis, the
dependent lung’s bronchus was always the intubated one; except one
study (Chapter 6), the dependent lung was always the ventilated one also.
54
In the first three studies, ventilation was monitored with the Ultima SV™
respiratory monitor (Datex Instrumentarium, Helsinki, Finland).
Mechanical ventilation was with a Siemens Servo 900 C (Siemens Elema;
Solna, Sweden) constant flow ventilator. In the last two studies, for
ventilation and monitoring purposes, the Datex ADU3 mechanical
ventilator and S/5 Anesthesia Monitor were used (Datex-Ohmeda Division,
Instrumentarium Corp., Datex-Ohmeda, Finland).
In all cases, a baseline tidal volume of 10 ml/kg at a ventilatory rate of 10
breaths per minute was used. Inspiratory time was 33% of total cycle
time (TTOT), and end-inspiratory pause was 10% of the total cycle time.
External positive end-expiratory pressure (PEEP) was set to zero. The
ventilator and the breathing system were carefully checked for potential
leaks before each patient. End-inspiratory and end-expiratory occlusions
were performed to determine the mechanical characteristics of the
respiratory system [peak inspiratory airway pressure (Ppeak), end-
inspiratory plateau pressure (Pplateau), and intrinsic positive end-expiratory
pressure (PEEPi)] (Figure 2.1.).
Blood gas samples were analyzed immediately after they were drawn, and
temperature corrected. Arterial and central venous oxygen saturation
were measured with a co-oximeter (Synthesis 350, Instrumentation
Laboratory, Milano, Italy).
55
Figure 2.1. Pressure-time curve during mechanical ventilation. 1.Peak inspiratory pressure,
2.Plateau inspiratory pressure, 3.Positive end-expiratory pressure.
A transesophageal echocardiograph probe was used for one study; it was
inserted after tracheal intubation to measure the cardiac output with the
method of the effective aortic valve area (18).
All investigations were performed with closed chest, before the surgical
procedure.
The temperature of the patients was kept constant during the studies (and
thereafter during the surgery) by using an air convection system.
Time
Paw
12
3
56
2.3. References
1. Cohen E: Anesthetic management of one-lung ventilation. In E
Cohen (ed): The practice of thoracic anesthesia. Philadelphia, JB
Lippincott, 1995, pp 308-40.
2. Pasque MK, Cooper JD, Kaiser L, et al.: Improved technique for
bilateral lung transplantation: rationale and initial clinical
experience. Ann Thorac Surg 1990;49:785–91.
3. Cooper JD, Patterson GA, Sundaresan RS, et al.: Results of 150
consecutive bilateral lung volume reduction procedures in patients
with severe emphysema. J Thorac Cardiovasc Surg
1996;112:1319–30.
4. Lima O, Ramos L, DiBiasi P, Judice L: Median sternotomy for
bilateral resection of emphysematous bullae. Thorac Cardiovasc
Surg 1981; 82:892–7.
5. Urschel HC, Razzuk MA.: Median sternotomy as a standard
approach for pulmonary resection. Thorac Surg 1986; 41:130–4.
6. Wasnick JD, Acuff T: Anesthesia and minimally invasive
thoracoscopically assisted coronary artery bypass: a brief clinical
report. J Cardiothorac Vasc Anesth 1997; 11:552–5.
7. Benumof JL: Special respiratory physiology of the lateral decubitus
position, the open chest, and one-lung ventilation. In JL Benumof
(ed): Anesthesia for thoracic surgery- 2nd ed, Philadelphia, WB
Saunders, 1995, pp 123-151.
57
8. Marshall C, Marshall BE: Site and sensitivity for stimulation of
hypoxic pulmonary vasoconstriction. J Appl Physiol 1983; 55:711-
6.
9. Arborelius M, Lundin G, Svanberg L, et al.: Influence of unilateral
hypoxia on blood flow through the lungs in man in lateral position.
J Appl Physiol 1960; 15: 595-7.
10. Fiser WP. Friday CD, Read RC: Changes in arterial oxygenation and
pulmonary shunt during thoracotomy with endobronchial
anesthesia. J Thorac Cardiovasc Surg 1982; 83:523-31.
11. Pepe PE, Marini JJ: Occult positive end-expiratory pressure in
mechanically ventilated patients with airflow obstruction. Am Rev
Respir Dis 1982; 126:166-70.
12. Rossi A, Polese G, Brandi G, et al: Intrinsic positive end-expiratory
pressure. Intensive Care Med 1995; 21: 522-36.
13. Scott LR, Benson MS, Bishop MJ: Relationship of endotracheal tube
size to auto-PEEP at high minute ventilation. Respir Care 1986;
31: 1080-2.
14. Larsson A, Malmkvist G, Werner O: Variations in lung volume and
compliance during pulmonary surgery. Br J Anaesth 1987; 59:
585-91.
15. Slinger PD, Hickey DR, Lenis SG, Gottfried SB: Intrinsic PEEP
during one-lung ventilation. Anesth Analg 1989; 68:S269.
16. Siafakas NM, Vermeire P, Pride NB, Paoletti P, Gibson J, Howard P,
Yernault JC, Decramer M, Higenbottam T, Postma DS, et al.
58
Optimal assessment and management of chronic obstructive
pulmonary disease (COPD). The European Respiratory Society Task
Force. Eur Respir J. 1995;8:1398-420.
17. Hannallah MS, Benumof JL, McCarthy PO, et al: Comparison of
three techniques to inflate the bronchial cuff of left polyvinyl
chloride double-lumen endobronchial tubes. Anesth Analg 1996;
82: 867-9.
18. Poelaert J, Schmidt C, Van Aken H, et al. A comparison of
transoesophageal echocardiographic Doppler across the aortic
valve and thermodilution technique for estimating cardiac output.
Anaesthesia 1999; 54: 128-36.
59
3. Pathophysiology of one-lung ventilation
Modified from: Szegedi LL. Pathophysiology of one-lung ventilation. Anesthesiology
Clinics of North America 2001; 19: 435-53
60
3.1. Introduction There are absolute indications for one-lung ventilation (OLV) (e.g.,
hemothorax, unilateral lavage, unilateral cyst), but most procedures using
double-lumen tubes (DLT) are relative indications to facilitate surgical
exposure. The adequate surgical exposure facilitates the dissection and
reduces operative time.
During OLV, the nondependent, nonventilated lung is excluded from the
ventilation, with all the tidal volume (VT) directed into the dependent lung.
In this situation, the distribution of perfusion is the major determinant of
the degree of venous admixture. The blood flow through the operative
lung becomes a right-to-left shunt in addition to that, which exists in the
ventilated lung (1). Given the same inspired oxygen concentration (FiO2)
and hemodynamic and metabolic status, OLV results in a much larger
alveolar-arterial oxygen tension difference P(A-a)O2 and lower arterial
oxygen partial pressure (PaO2) than during two-lung ventilation.
In estimating the degree of shunt that is created by OLV when it is
performed in the lateral decubitus position, on average, 40% of cardiac
output perfuses the nondependent (upper) lung and the remaining 60%
the dependent (lower) lung (1). Mechanisms that tend to decrease the
percent of cardiac output perfusing the nondependent, nonventilated lung
are passive (e.g., mechanical-like gravity, surgical manipulation, amount
of pre-existing lung disease) or active (e.g., hypoxic pulmonary
vasoconstriction (HPV)) (Figure 3.1.) (1).
61
Figure 3.1. Determinants of blood flow distribution during one-lung ventilation. The major
determinants of blood flow to the nondependent lung are gravity, surgical interference with blood
flow, the amount of nondependent lung disease, and the magnitude of nondependent lung hypoxic
pulmonary vasoconstriction. The determinants of dependent lung blood flow are gravity, amount of
dependent lung disease, and dependent lung hypoxic pulmonary vasoconstriction. RV = right
ventricle. (From Benumof JL: Special respiratory physiology of the lateral decubitus position, the
open chest, and one-lung ventilation. In JL Benumof (ed): Anesthesia for Thoracic Surgery, ed 2.
Philadelphia, WB Saunders, 1995, pp 123-151; with permission.)
The normal response of the pulmonary vasculature to atelectasis is an
increase in pulmonary vascular resistance (in the atelectatic lung), and
the increase in atelectatic lung resistance is almost entirely caused by
hypoxic pulmonary vasoconstriction. HPV is a protective reflex mechanism
that diverts blood flow away from the atelectatic lung. With an intact HPV
response, the transpulmonary shunt through the nondependent lung
decreases approximately to 23% of the cardiac output (Figure 3.2.).
62
Figure 3.2. Two-lung ventilation versus one-lung ventilation. Typical values for fractional blood
flow to the nondependent and dependent lungs as well as PaO2 and shunt fraction (Qs/Qt) for the
two conditions are shown. The Qs/Qt (approximately 10% of total cardiac output) during two-lung
ventilation is assumed to be distributed equally between the two lungs (5% to each lung). The
essential difference between two-lung and one-lung ventilation is that during one-lung ventilation
the nonventilated lung has some blood flow and, therefore, an obligatory shunt, which is not
present during two-lung ventilation. The 35% of total flow perfusing the nondependent lung, which
was not shunt flow, was assumed to be able to reduce its blood flow by 50% by HPV. The increase
in Qs/Qt from two-lung to one-lung ventilation is assumed to be solely caused by an increase in
blood flow through the nonventilated, nondependent lung during one-lung ventilation.(From
Benumof JL. Special physiology of the lateral decubitus position, the open chest, and one-lung
ventilation. In: JL Benumof (ed): Anesthesia for Thoracic Surgery, ed 2. Philadelphia, WB
Saunders, 1995, pp 123-151; with permission.)
63
3.2. Factors influencing hypoxic pulmonary vasoconstriction
Most of the blood flow reduction in an acutely atelectatic lung is caused by
HPV; re-expansion and ventilation of a collapsed lung with nitrogen does
not increase the blood flow to the lung, whereas ventilation with oxygen
restores all of the blood flow to precollapse value (2).
The predominant stimulus is the alveolar oxygen partial pressure (PAO2).
The mixed venous oxygen partial pressure (PvO2) also is believed to have
a role in the degree of HPV. The HPV response is maximal when the PvO2
is normal, and is decreased by high or low values of PvO2 (1,3). Low PvO2
lowers the PAO2 in the ventilated lung, causing offsetting and competing
HPV in the ventilated lung, whereas increased PvO2 (e.g., sepsis) causes
the oxygen tension to increase in the nonventilated lung, thereby
inhibiting HPV. The stimulus oxygen partial pressure (PsO2) is defined as if
the sensor were at a discrete site in the precapillary arteriole influenced
by both the alveolar and the mixed venous oxygen tensions. It may be
estimated using the following equation (4):
PsO2 = PvO20.39 x PAO2
0.61
Concerning the time course of HPV response, many experimental and
clinical studies were performed, with contradictory results. Benumof (5)
showed that intermittent hypoxic challenges potentiate the hypoxic
vasoconstriction in the left lower lobe of open-chest dog lungs, but the
64
preparation and manipulation of the animals required considerable
instrumentation and manipulation, which may have interfered with the
HPV. Carlsson et al (6) found maximal HPV response within 15 minutes of
hypoxia in a human study, which agrees with observations in animal
studies; Tucker and Reeves (7) could not maintain HPV during acute
hypoxia in anesthetized dogs. During one-lung hypoxia in dogs, Domino et
al (8) studied HPV in closed chest dogs and found a maximal level from
the very first hypoxic challenge, concluding that time factor should not be
a hindrance to manipulative studies on the HPV response, once a maximal
response has been evoked, normally in 10 to 15 minutes. Results obtained
concerning the influence of time on HPV vary widely.
Arterial carbon dioxide partial pressure (PaCO2), especially at high values
influences the level of HPV response (9); however many discrepancies
exist in the literature (10, 11) (Figure 3.3.). Both hypocapnia and
hypercapnia inhibit the HPV response. Hypocapnia can directly
pharmacologically dilate the hypoxic lung. In addition, hypocapnia is
mostly achieved by increasing minute ventilation and thus airway pressure
in the ventilated lung. The increased airway pressure in the ventilated
lung, may selectively increase ventilated lung vascular resistance, thereby
inhibiting nonventilated lung HPV. Hypercapnia may directly vasoconstrict
the ventilated lung, thereby increasing ventilated lung pulmonary vascular
resistance, which will inhibit nonventilated lung HPV. In addition,
hypercapnia may possibly be achieved by decreasing minute ventilation,
which would decrease airway pressure in the ventilated lung, which would
65
decrease ventilated lung pulmonary vascular resistance, which would
enhance nonventilated lung pulmonary vascular resistance (1).
Figure 3.3. Schematic diagram of the effect of changes in CO2 and regional hypoxic pulmonary
vasoconstriction (HPV). .(From Benumof JL. Special physiology of the lateral decubitus position, the
open chest, and one-lung ventilation. In: JL Benumof (ed): Anesthesia for Thoracic Surgery, ed 2.
Philadelphia, WB Saunders, 1995, pp 123-151; with permission.)
Cardiac output variations can influence HPV, through changes in
pulmonary vascular pressure. Benumof and Wahrenbrock (12)
demonstrated in a canine, open-chest model that the HPV response was
abolished when left-atrial pressure reached 25 mm Hg.
The distribution of alveolar hypoxia is not a determinant of the amount of
HPV because all regions of the lungs respond to alveolar hypoxia with
vasoconstriction (1).
66
Almost all of the studies with the very potent systemic and pulmonary
vasodilator drugs have shown inhibition of HPV, or have a clinical effect
(increased shunt, decreased PaO2) that is consistent with decreased HPV
(1).
The vasoconstrictor drugs constrict the dependent, normoxic lung vessels
preferentially, increasing its vascular resistance and diminishing blood flow
from the nonventilated lung (13). The most studied was the dopamine
which seems to be a reasonable cardiovascular stimulant to use in
patients with pulmonary disease (14, 15).
Drugs such as almitrine bismesylate, nitric oxide and leukotrienes may
increase the HPV response in a dose-dependent manner. Nonsteroidal
anti-inflammatory drugs have been reported to produce contradictory
effects on HPV (16-19).
A decreased FiO2 in the dependent lung will cause an increase in the
vascular resistance of this lung, decreasing blood flow diversion from the
nondependent towards the dependent lung. FiO2 changes cause secondary
changes in PaO2, and tertiary changes in PvO2, which is also, a major
determinant of HPV (20).
67
3.3. Effects of anesthetic agents on hypoxic pulmonary vasoconstriction
The effect of inhalational anesthetic agents on HPV has been debated by
various investigators. Particular concerns have arisen concerning species
differences and the applicability of in vitro data to in vivo or clinical
situations (21).
In an isolated rat lung model, halothane, enflurane and isoflurane were
found to depress HPV in a dose dependent manner. The concentrations in
minimum alveolar concentration (MAC) units for the rat at which 50%
depression of HPV occurred were respectively 0.47, 0.60 and 0.56 for
halothane, isoflurane and enflurane, respectively (22, 23).
Bjertnaes in 1978 (24) was one the first to suggest that inhalational
anesthetics inhibit HPV in humans (Figure 2.4.).
Figure 3.4. Schematic diagram showing the effect of 1 MAC isoflurane anesthesia on shunt during
one-lung ventilation (1LV) of normal lungs. A normal HPV response decreases the blood flow to the
nondependent lung by 50%, so that the dependent/nondependent lung blood flow ration is now
approximately 20%/80% (middle). Administration of 1 MAC isoflurane should cause a 21%
decrease in the HPV response, which would decrease the 50% blood flow reduction to a 40% blood
flow reduction in the HPV response. (From Benumof JL, Anesthesiology 1986; 64: 419, with
permission).
68
The distribution of blood flow to the test lung was determined by
pulmonary perfusion scanning or lung scintigraphy using human serum
albumin macro aggregates labeled with either 99mTc or 131I. After
spontaneous breathing of room air occurred, anesthesia induction
proceeded with thiopental, fentanyl, and pancuronium. Unilateral hypoxia
was achieved by ventilating the test lung with 100% N2, and the other
lung with 100% O2. After ipsilateral administration of either diethyl ether
or halothane, blood flow to the test lung increased significantly. The
researchers concluded that diethyl ether and halothane inhibit HPV and
that this contributes to the development of arterial hypoxemia during
anesthesia in humans.
Recent studies in humans show that isoflurane seems to be less inhibitory
than enflurane or halothane and equivalent to sevoflurane or desflurane.
In addition, there is an interindividual variability with volatile anesthetic
agents, this explaining why some patients do very well with an
inhalational agent, while other appear to be susceptible to hypoxemia.
This latter group may actually benefit from a change to an injectable
agent.
Buckley and colleagues (25) studied N2O using a dog model of global
hypoxia and suggested HPV was enhanced by N2O. Animal models using
global hypoxia result in systemic hypoxemia, which has profound
hemodynamic effects on the whole animal, however, and therefore must
be interpreted cautiously. Most subsequent studies have found a small but
consistent inhibition of HPV by N2O.
69
Generally, intravenous anesthetic techniques have not been shown to
provide better oxygenation than the newer volatile anesthetics in < 1MAC
concentrations (26). Pentobarbital has been reported having no influence
on HPV. However, there has been some controversy concerning this issue.
Susmano et al (27) reported that in intact dogs anesthetized with
pentobarbital, HPV was dampened compared to those that received
fentanyl-droperidol anesthesia. In this study, however, high doses of
anesthetics were used. Intravenous drugs in clinical use, such as
narcotics, benzodiazepines, local anesthetics (28), droperidol, appear to
have no significant effect on the HPV response. Ketamine, which appears
to have no effect on HPV, does not impair arterial oxygenation (29).
Recently, claims have been made that propofol also has no effects on the
HPV response (30, 31).
70
3.4. Carbon dioxide elimination during one-lung ventilation
If the patients’ lung function is between normal limits, the same minute
volume that is used during two-lung ventilation may be used during OLV
for carbon dioxide elimination. The end-tidal carbon dioxide decreases
when initiating OLV, at least for the first 5 to 10 minutes, because the
dependent lung is hyperventilated relative to its perfusion. Thereafter, the
HPV will increase the perfusion of the ventilated lung, and the relative
hyperventilation will decrease (1). If the lung or the lungs have
preexisting disease causing V/Q mismatch, however, the situation may be
different. PaCO2 may decrease if the diseased lung with a large alveolar
deadspace is no longer ventilated. When the two lungs are diseased,
during OLV, the ventilated lung can no longer cope with the whole CO2
production, and the PaCO2 rises.
71
3.5. Other factors influencing blood flow distribution during one- lung ventilation
Surgical compression or retraction may also contribute to passive
reduction of nondependent lung blood flow (1,32). These manipulations
also may cause release of some vasodilating prostaglandins or
endothelium derived relaxing factor, which may increase shunt fraction
(33,34); thus surgical manipulation during one-lung ventilation may have
variable and inconsistent effects on blood flow deviation towards the
dependent lung. Ligation of pulmonary vessels for lobectomy or
pneumectomy in the nondependent lung reduces blood flow to this lung
greatly.
Another factor that may influence hypoxia during OLV is the side of
surgery. Left thoracotomy has a better PaO2 during OLV than right
thoracotomy, because the left lung normally receives 10 percent less
cardiac output than the right lung (35).
72
3.6. Gravity
Gravity is a major, pharmacologically and physiologically independent
determinant of regional blood flow distribution. The extent of blood flow
redistribution depends on the local relationship between pulmonary
arterial, venous and airway pressures. In the lateral position, regional
blood flow increases from the nondependent to the dependent thoracic
wall (36).
Because of the latest developments in cardiothoracic surgery, various
procedures are performed with patients in the supine, instead of the
traditional lateral, position. Double-lung transplantation, lung volume
reduction surgery, and minimally invasive coronary artery surgery
routinely are performed with patients in the supine position (37 - 40).
Median sternotomy has been recommended for pulmonary resection (41).
During these interventions, OLV is required inherently, generally inducing
some degrees of hypoxemia. Fiser et al (42) studied the period of OLV
with patients in supine and lateral positions, and did not find changes in
arterial oxygen tension after 10 to 20 minutes of OLV when the patients
were turned into the lateral position. A recent study (43) (see Chapter 5 in
this thesis) performed on patients with stable chronic obstructive
pulmonary disease (COPD), with mild to moderate degrees of pulmonary
hyperinflation, scheduled for elective pulmonary resection, showed a
significantly better oxygenation during OLV in the lateral than in the
supine position. Another study (44) (see Chapter 6 in this thesis),
separates the effects of gravity from other factors during OLV, suggesting
73
that the gravitational factor may be more important in blood flow
redistribution during OLV than previously believed.
74
3.7. Pathology caused by endotracheal tubes In some circumstances OLV is critical, but failure to isolate the operated
lung may compromise any thoracic operation. In practice, DLTs are the
most used for lung separation, but there are also other devices that may
be good choices, such as the Univent tube, a normal endotracheal tube, a
Fogarty catheter, or the most recently introduced Arndt endobronchial
blocker (45,46).
In recent reports, the bronchial lumen of the DLT offered less resistance
than the tracheal one (47,48). This difference is most likely a result of the
manufacturing process, to offer the least resistance for ventilation of the
dependent lung during OLV (48). Despite this advantageous design, most
textbooks advocate the use of left-sided DLTs for most cases requiring
OLV because of easier positioning (45, 46). Insertion of a left-sided DLT
for a left-sided thoracotomy, implies dependent lung ventilation through
the tracheal lumen, imposing a significantly higher resistance to air flow
than the bronchial lumen – which may cause an increased peak
inspiratory pressure (Ppeak) and most importantly an increased level of
auto-positive end-expiratory pressure (PEEPi) - during OLV.
The right DLTs have higher resistance than the left ones, probably
because of the design characteristics of their distal, bronchial lumen; the
lateral aspect of the bronchial cuff is fenestrated for ventilation of the right
upper bronchus.
There is a considerable difference in resistance between the smaller (35 or
37 F) and the larger (39 or 41 F) DLTs at the same airflow. The clinical
75
interest of this step-up increase in air flow resistance is illustrated in one
study (49), where 83% of the patients whose trachea was intubated with
the smaller (35 or 37 F) DLT showed the presence of interrupted
expiratory flow (IEF), whereas only 52% of the patients in whom the
larger DLTs were inserted, exhibited IEF. Changes in DLT size would be
expected to have major effects on the in vivo resistance of DLTs. The in
vivo resistance of the DLTs, however, may exceed the values obtained in
vitro (50) because of turbulence, tube deformities, and secretions.
There are special circumstances where the correct position of the DLT is
critical (e.g., hemorrhage, bronchopleural fistulae, lung transplantation).
Impaired ventilation of the dependent lung or inappropriate deflation of
the nondependent lung may cause not only slowly deteriorating oxygen
saturation but may influence the success of the surgery. Complications
caused by DLT malposition are one of the main causes of adverse events
that occur in as many as 20-30% of patients during thoracic anesthesia
(51). These problems usually manifest as cross-ventilation of the two
lungs during OLV, the inability to deflate the operated lung, and collapse
of one or more lobes of the ventilated lung. During a prospective analysis
of 234 intubations (52), the rate of complications were arterial oxygen
saturation below 90% in 9% of cases, Ppeak above 40 cm H2O in 9% of
cases, poor lung isolation in 7% of cases, and air trapping in 2% of cases.
Despite confirmation of DLT position by fiberoptic bronchoscopy, during
surgical manipulation, the DLT moves easily and quickly out from the
76
correct position, producing obstruction and increased inspiratory airway
pressures.
The disadvantages of the use of DLTs for lung separation are related to
the fact that the lumens of a double-lumen tube are narrow. The older,
rubber-made DLTs have low total flow resistance (53, 54). It commonly is
taught, however, that despite the reduced effective diameter of the DLT
during OLV, the increased airway resistance easily can be overcome by
positive pressure ventilation (55). The small diameter of the bronchial
lumen, however, imposes an increased resistance to passive expiration,
and may promote the development or the amplification of dynamic
pulmonary hyperinflation and PEEPi.
In case of hypoxemia, or inability to deflate the operative lung, the correct
position of the DLT must be verified first. The onset of hypoxemia,
however, is seldom instantaneous when the DLT is moved from the
correct position. Some way of continuous monitoring of DLT position
would be the ideal method to diagnose an incorrect tube position
promptly. Monitoring the inspiratory airway pressure differences, even if
they are statistically significant for DLT malposition, is useful, but these
pressure differences cannot be used as a single value in clinical decision
making (see Chapter 4 in this thesis)(56). In addition to the Ppeak
displayed on the manometer of most ventilators, stand-alone respiratory
monitors and some new ventilators are displaying additional ventilatory
data continuously in digital format - such as inspiratory and expiratory
tidal volumes (VTI, VTE), Ppeak, end-inspiratory (plateau) pressure (Pplateau),
77
end-expiratory pressure (PEEP), and quasistatic compliance of the
respiratory system (Cst,rs). By automatic processing of these data, flow-
volume (FV) and pressure-volume (PV) curves are obtained (57). When
the DLT moves from the ideal position, the measured ventilatory data not
only are modified but this deviation may cause a change in the pattern of
the displayed curves. Accordingly, it is possible that this modified pattern
may help to determine the correct positioning of the DLT or may detect
intraoperative displacement promptly.
78
3.8. Pathology of the patients
Concurrent lung disease is the rule rather than the exception in patients
undergoing thoracic surgery. The majority of patients scheduled for lung
resection are stable patients with chronic obstructive pulmonary disease
(COPD), with variable degrees of pulmonary hyperinflation.
COPD is defined by the European Respiratory Society as reduced
maximum expiratory flow and slow forced emptying of the lungs, which is
slowly progressive and mostly irreversible to present medical treatment
(58). The only positive requirement for diagnosis of COPD is abnormal
spirometry, but there are other causes of airway obstruction.
Extrathoracic airway obstruction, localized forms of intrathoracic airway
obstruction (e.g., bronchial carcinoma) and most specific causes of
widespread intrathoracic airway obstruction (e.g., cystic fibrosis) are
excluded from the definition of COPD; however, they may coexist (e.g.,
bronchial tumor and COPD). COPD has an uncertain cause; however,
smoking plays a predominant role. The morphologic definition of
emphysema as destructive enlargement of peripheral air spaces remains
widely accepted. Tumors may reduce perfusion to that lung. A patient with
COPD may be hypoxemic, and he or she has pulmonary hyperinflation on
preoperative pulmonary function findings. At moderate and severe levels
of disease, hypercapnia also may be present. Because these patients are
chronically hypoxemic, HPV is already maximal (59). OLV of these patients
will no more be influenced by the amount of HPV. If the nondependent
lung is not diseased, and has a normal amount of blood flow, collapse of
79
such a normal lung may be associated with a higher blood flow and shunt
at the nonventilated lung; during OLV, patients with normal lung function
would desaturate more than patients with previously diseased lungs.
Because most patients are smokers, they may present coexisting (80 %
of cases) ischemic heart disease (60). Because most patients scheduled
for lung surgery have lung cancers, heart disease resulting from
therapeutic radiation also may be present (61). Some patients with
preoperative chemotherapy may present antineoplastic drug pulmonary
toxicity. Bleomycin may induce interstitial lung disease, and may sensitize
the lung to oxygen-induced lung injury. The bleomycin-high FiO2
relationship, however, is controversial (1, 62, 63).
80
3.9 Auto-positive end-expiratory pressure during one-lung ventilation The equilibrium point for lung volume, where lung elastic recoil is offset by
the resting tone of the chest wall and diaphragm is called functional
residual capacity (FRC). In normal subjects at rest, the end-expiratory
lung volume (FRC) corresponds to the relaxation volume (Vr) of the
respiratory system (i.e., lung volume at which the elastic recoil pressure
of the respiratory system is 0) (64).
In patients with COPD, FRC may be increased markedly above predicted
FRC because of static (i.e., loss of elastic lung recoil) and dynamic (i.e.,
increased airway resistance and expiratory flow limitation) factors. The
term pulmonary hyperinflation defines this increase in the FRC above
normal (65). Because the increased FRC corresponds to Vr, elastic recoil
pressure is 0 in these patients (66). The increased air-flow resistance
reduces the rate of lung emptying, and, in certain conditions, expiration
duration may not be long enough. Inspiration begins before the system
has reached its Vr. In addition to the static hyperinflation caused by the
loss of the elastic recoil, these patients develop dynamic pulmonary
hyperinflation (DPH) (67, 68). In this situation, elastic recoil pressure is
no longer 0, but becomes positive.
This phenomenon also may occur during passive mechanical ventilation.
When the time required to complete the passive exhalation is longer than
the available expiratory time imposed by the ventilatory settings, the end-
81
expiratory lung volume exceeds the relaxation volume of the respiratory
system (68). As a result of the increased lung volume, alveolar pressure
remains positive throughout the course of expiration until interrupted by
the next inflation cycle. This is comparable to the use of positive-end
expiratory pressure (PEEP), and has been termed to as auto (69) or
intrinsic (70) positive end expiratory pressure (intrinsic PEEP, PEEPi, auto-
PEEP). Because intrinsic PEEPi is not recorded on the pressure manometer
of the ventilator, it is also named occult PEEP (71). Because the
phenomenon occurs unintentionally, it is sometimes called inadvertent
PEEP (72).
DPH and PEEPi are common in mechanically ventilated individuals with
advanced COPD (73, 74), in whom the passive expiration is prolonged by
increased airway resistance (i.e., dynamic airway collapse and expiratory
flow limitation). DPH is not restricted to patients with end-stage COPD;
expiratory flow limitation and external factors can play an important role
in the development of PEEPi (74,70).
Because PEEPi is not recorded on the pressure manometer of the
ventilator, it frequently remains undetected. If the expiratory port of the
ventilator circuit is occluded immediately before the onset of the next
inspiration, the pressure in the lungs and ventilator circuit may equilibrate
and the level of PEEPi will be displayed on the ventilator manometer (71).
Alternatively, the level of PEEPi can be estimated from continuous
recordings of airway pressure and flow during mechanical ventilation (70,
54). End-expiratory airway occlusion can be performed manually at the
82
expiratory port of the ventilator (71), or it can be performed with
ventilators (e.g., Siemens 900 C Servo Ventilator, Solna, Sweden)
equipped with the end-expiratory hold for the rapid occlusion of the
expiratory port exactly at the end of the tidal expiration (54). Observing a
plateau in airway pressure obtained this way provides evidence of
respiratory muscle relaxation, the absence of leaks in the circuit and
equilibration between alveolar and airway opening pressures.
As explained previously, PEEPi can be measured by applying an end-
expiratory occlusion, but most of the ventilators used during anesthesia
and OLV are not equipped with a flow-time curve display, and cannot
perform an end-expiratory occlusion. To identify patients with PEEPi,
without the need to interrupt OLV, the flow-volume curves of an online
respiratory monitor may be examined (Figure 3.5.)
Figure 3.5. A normal flow-volume curve (left) and a flow-volume curve illustrating interrupted
expiratory flow (right). Arrowheads on the traces indicate the direction of the flow. The arrow
indicates the gap on the expiratory part of the curve.(From Bardoczky GI, d’Hollander AA, Cappello
M, et al: Interrupted expiratory flow on automatically constructed flow-volume curves may
determine the presence of intrinsic positive end-expiratory pressure during one-lung ventilation.
Anesth Analg 1998; 86: 880-4; with permission.)
83
This method has the advantage over the end-expiratory method by being
applicable in operating room settings when anesthesia ventilators are
used. The presence of an interrupted expiratory flow may suggest the
presence of clinically relevant PEEPi with a reasonable accuracy (75)
(Figure 3.5.).
In patients with COPD and pulmonary hyperinflation, PEEPi occurs
commonly during the period of OLV and only occasionally in patients with
normal lungs. PEEPi occurs during OLV more in the lateral as compared to
the supine position; otherwise, the occurrence and magnitude of PEEPi is
influenced mainly by the pre-existing pulmonary hyperinflation (functional
residual capacity and residual volume). The preoperative forced expiratory
volume in one second (FEV1) alone is poorly predictive of its occurrence
(76) (Figure 3.6.).
Figure 3.6. The magnitude of intrinsic PEEP measured in 20 patients during dependent lung
ventilation in the supine and lateral position. PEEPi N: patients with normal static lung volumes.
PEEPi PH: patients with pulmonary hyperinflation. PEEPI was present in 8 of 11 hyperinflated
patients in the supine position, and in all 11 in the lateral position. PEEPi was present in 8 of 9
hyperinflated patients in the supine position and in all 11 in the lateral position. PEEPi was
demonstrable in only 1 of 9 normal patients in the supine position and in none in the lateral
position. (From Bardoczky GI, Yernault JC, Engelman EE, et al: Intrinsic positive end-expiratory
pressure during one-lung ventilation for thoracic surgery: The influence of preoperative pulmonary
function. Chest 1996; 110: 180-4; with permission.)
84
3.10. Pathophysiology caused by ventilatory management of one- lung ventilation
To facilitate thoracic surgery, the most important and valuable anesthetic
technique used is OLV. In the technique of OLV, one-lung is ventilated
mechanically while the other is occluded or open to the atmosphere. For
ventilatory management of OLV, it is recommended that a dependent lung
tidal volume similar during two-lung ventilation be used, and respiratory
rate be adjusted (55). Respiratory frequencies around 15 breaths per
minute commonly are used; a recent textbook recommended that patients
should be ventilated with a tidal volume of 10-12 ml/kg at a rate to
maintain a PaCO2 of 35 ± 3 mmHg (51). These guidelines, however, fail to
consider that most patients scheduled for lung surgery have chronic
obstructive pulmonary disease with variable degrees of air-flow limitation
and pulmonary hyperinflation. During mechanical ventilation of these
patients, conditions that impede expiratory flow (e.g., increased airway
resistance and additional resistance of the endotracheal tube) (54, 71) or
inadequate ventilatory settings (77) may predispose to DPH and PEEPi.
Two important changes are taking place during OLV. Initiating OLV
decreases the potential lung volume participating in ventilation. Because
the VT is fixed, a decrease in the potential ventilated lung volume implies
a larger VT delivered to the remaining ventilated lung. By clamping the
tracheal lumen of the DLT and ventilating through the bronchial lumen,
the diameter of the airway is decreased, and thus the resistance to flow
increases significantly. With unaltered ventilatory settings, a larger VT has
85
to be exhaled through the bronchial lumen of the DLT, which represents
significantly higher resistance to exhalation. The significantly larger
proportion of patients who develop PEEPi during the period of OLV is not
unexpected.
The volume, frequency and timing of gas delivered to the dependent lung,
might have important, disease-specific effects on the cardiovascular and
respiratory systems (78).
In the presence of COPD and DPH, an increase in respiratory rate (RR) or
VT, or manipulation of the expiratory time (TE) promotes the development
or enhancement of PEEPi and DPH. Earlier studies performed during OLV
have assessed mainly the effect of VT changes (79, 80) or the isolated
effect of altered RR with unchanged VT (81). DPH and PEEPi are dynamic
phenomena that depend on the pre-existing lung disease (76) and the
actual conditions of ventilation. By changing inspiratory time (Ti), and
thereby TE, or by altering RR at the same minute volume, inspiratory
airway pressures and PEEPi can be deliberately manipulated. The Ti allows
a reduction in inspiratory flow and thus Ppeak. The increased TI/TTOT implies
an equivalent decrease in TE associated with incomplete lung emptying.
Any change in Ppeak induced by reducing inspiratory flow does not alter the
risk for barotrauma if it is not combined with a decrease of the alveolar
pressure, reflected by Pplateau. The most frequently used TI/TTOT during
anesthesia and mechanical ventilation is 33% with RR of 10 breaths per
minute (bpm) and end-inspiratory pause (EIP) of 10%. Shortening TI/TTOT
86
increases the duration of expiration and, in turn, reduces the level of
PEEPi, producing a high Ppeak, and unaltered Pplateau.
Increasing the RR with the same minute ventilation has two opposite
effects (see Chapter 8 in this thesis) (82). The delivered VT decreases, and
less time is needed for passive deflation of the lung to the resting end-
expiratory volume. Expiratory time decreases and less time is available to
empty the VT (which, in this circumstance, also is decreased). Increasing
the RR significantly decreases the Ppeak and the Pplateau. The effect of TE on
PEEPi depends on the rate of passive lung deflation, which is determined
by the elastic recoil (Pel,rs) stored during the preceding inflation (83). The
main determinant of DPH and PEEPi is TE. Considering a specific
respiratory system, for a constant RR, PEEPi will increase with increasing
Ti fraction. For a fixed Ti fraction, PEEPi will increase with RR.
There are no clear guidelines established for VT during OLV. VT below 10
ml/kg generally are discouraged because of concerns about dependent
lung atelectasis (55, 51, 84). General anesthesia in patients without
significant lung disease causes a decrease in arterial oxygenation because
of the development of atelectasis in the dependent lung (1, 85). Patients
with COPD, however, do not develop atelectasis or a decrease in FRC
during anesthesia (86), presumably because of long-standing
hyperinflation. High inspiratory airway pressures (79, 87) causing
increased dependent lung vascular resistance in some patients is believed
to explain the lack of consistently improved oxygenation during large tidal
volume OLV.
87
End-inspiratory pause (EIP) is a ventilatory pattern characterized by a
period of zero-flow occurring at the end of inspiration, resulting in a
prolongation of inspiration without gas flow. EIP is used widely during
mechanical ventilation in the operating room and critical care settings. In
one study, EIP may improve gas exchange and the efficiency of ventilation
while decreasing microatelectasis in patients with acute respiratory failure
of various origins (88). In another study (89) of mechanically ventilated
patients with acute exacerbation of COPD, the addition of EIP to the
inspiratory cycle was not associated with significant alteration of gas
exchange but led to a decrease in expiratory flow and further
hyperinflation because of decreased TE. In a more recent study (90),
during the period of OLV in the lateral position with closed chest of
patients with pre-existing pulmonary hyperinflation, the magnitude of
PEEPi increased (EIP shortened the TE) and oxygenation decreased
significantly, whereas the efficacy of ventilation was not changed by the
addition of an end- inspiratory pause to the ventilatory pattern. By
altering EIP, there was a significant negative correlation between the
PEEPi present and the PaO2, and a close association was found between
the change in PEEPi and the ensuing P(A-a)O2 changes.
88
3.11. Is intrinsic positive end-expiratory pressure harmful?
The PEEPi-induced increase in lung volume can cause compression of
intra-alveolar vessels, and increase the pulmonary vascular resistance in
the dependent lung, and, in turn, may cause the diversion of blood flow
from the ventilated lung towards the nonventilated lung (1, 55). The
right-to-left shunt increases and PaO2 decreases. This mechanism may be
similar to the one observed during application of an external PEEP to the
dependent lung during OLV (1, 55). Overdistension and hyperinflation as a
consequence of generally used ventilatory regimen during OLV is believed
to be one of the factors promoting the development of postpneumectomy
pulmonary edema (91) and acute lung injury (92).
A low level of PEEPi, however, seems not to be so harmful. Patients with
DPH and PEEPi may have a favorable end-expiratory lung volume during
OLV, and a diminution of dependent lung shunt (93). To appreciate the
presence of PEEPi is important in the ventilatory management of patients
scheduled for lung transplantation or for resection of emphysematous
bullae and who have end-stage obstructive airway disease with severe
pulmonary hyperinflation. To sustain a long period of OLV during the
surgery may be difficult for these patients. They may have a much higher
level of PEEPi, and small alterations of the ventilatory pattern may be
beneficial to decrease the degree of excessive dynamic hyperinflation or
may be detrimental when the change of settings results in an excessive
level of PEEPi (94).
89
An external PEEP may be applied during OLV of patients with low PaO2
(55, 94, 95). The beneficial effects of the external PEEP, if any, during
OLV in patients without pulmonary hyperinflation is believed to be
because of an increased FRC, which helps to prevent airway closure at the
end of expiration. An increased FRC above normal may alter the vascular
resistance. The pulmonary vascular resistance is least at lung volumes
near the ideal FRC, but increases rapidly once the lung volume increases
or decreases from this ideal value (Figure 3.7.). This is due to the
anatomical structure of the pulmonary circulation (alveolar and
extraalveolar vessels).
Figure 3.7. Effects of lung volume expansion on the caliber and resistance of the alveolar (ALV)
and extraalveolar (EA) pulmonary vessels. Note that the net effect on total resistance (TOTAL)
results in the lowest value at functional residual capacity (FRC). Values for total resistance increase
toward residual volume (RV) and total lung capacity (TLC). (From Gal TJ Anatomy and physiology
of the respiratory system and pulmonary circulation.. In: Kaplan JA, Slinger PD (eds): Thoracic
Anesthesia, ed 3. Philadelphia, Churchill Livingstone, 2003, pp 57-71; with permission.)
90
As at moderate to severe degrees of COPD, some patients are moderately
hypercapnic while breathing spontaneously (58), full compensation of
hypercapnia in these patients during OLV is not necessary.
Tugrul et al. (96) compared classic volume-controlled and pressure-
controlled ventilation during one-lung anesthesia. It was found that during
pressure-controlled OLV, Ppeak decreased significantly and PaO2 increased
significantly, showing the potential benefits of pressure-controlled OLV.
Terminating OLV and re-establishing TLV may be followed by pulmonary
edema. Its occurrence has been related to the surgical procedure,
duration of OLV, and barotrauma during re-inflation (94, 95). Inflation of
the nondependent lung should be done slowly, manually and gradually,
with an inflation pressure that does not exceed 30 cmH2O.
The management of some problematic patient having thoracic surgery
remains among the most difficult challenges for the anesthesiologist. OLV
adds to the complexity of the anesthetic technique and to the associated
level of risks, and must not be pursued at all costs. This technique, even
when applied with totally correct management skills and a perfect
knowledge of physiology and pathophysiology of OLV, does not represent
the ultimate panacea for some highly diseased patients, who may require
other perioperative therapies (e.g., partial extracorporeal circulation and
carbon dioxide removal). Continuing the OLV never justifies rendering the
patients hypoxemic.
91
3.12. References
1. Benumof JL: Special respiratory physiology of the lateral decubitus
position, the open chest, and one-lung ventilation. In JL Benumof
(ed): Anesthesia for thoracic surgery- 2nd ed, Philadelphia, WB
Saunders, 1995, pp 123-151
2. Benumof JL: Mechanism of decreased blood flow to atelectatic lung. J
Appl Physiol 1978; 46: 1047-8
3. Domino KB, Glasser SA, Wetstain L, et al: Influence of PvO2 on blood
flow to atelectatic lung. Anesthesiology 1982; 57:A471
4. Marshall C, Marshall BE: Site and sensitivity for stimulation of hypoxic
pulmonary vasoconstriction. J Appl Physiol 1983; 55: 711-6
5. Benumof JL: Intermittent hypoxia increases lobar hypoxic pulmonary
vasoconstriction. Anesthesiology 1983; 58: 399-404
6. Carlsson AJ, Bindslev L, Santesson J, et al: Hypoxic pulmonary
vasoconstriction in the human lung: the effect of prolonged unilateral
hypoxic challenge during anaesthesia. Acta Anaesthesiol Scand 1985;
29: 346-51
7. Tucker A, Reeves JT: Non-sustained pulmonary vasoconstriction during
acute hypoxia in anesthetized dogs. Am J Physiol 1977; 42: 889-99
8. Domino K, Chen L, Alexander C, et al: Time course and responses of
sustained hypoxic pulmonary vasoconstriction in the dog.
Anesthesiology 1984; 60: 562-6
92
9. Benumof JL, Mathers JM, Wahrenbrock EA: Cyclic hypoxic pulmonary
vasoconstriction induced by concomitant carbon dioxide changes. J
Appl Physiol 1976; 41: 466-9
10. Grover RF, Wagner WW, McMurtry IF, et al: Pulmonary circulation.
In Handbook of physiology. The cardiovascular system. Peripheral
Circulation and organ blood flow. Bethesda, MD, American
Physiological Society, 1983, pp 103-136
11. Nattie EE: Gas exchange in acid-base disturbances. In Handbook of
physiology. The respiratory system. Gas exchange. Bethesda, MD,
American Physiological Society, 1983, pp 421-438
12. Benumof JL, Wahrenbrock EA: Blunted hypoxic vasocontsriction by
increased lung vascular pressures. J Appl Physiol 1975; 38: 846-50
13. Marin JLB, Orchard C, Chakrabarti MK, et al: Depression of hypoxic
pulmonary vasoconstriction in the dog by dopamine and isoprenaline.
Br J Anaest 1979; 51: 303-12
14. Lejeune P, Leeman M, Deloof T, et al: Pulmonary hemodynamic
response to dopamine and dobutamine in hyperoxic and in hypoxic
dogs. Anesthesiolgy 1987; 66: 49-54
15. Lejeune P, Naeije R, Leeman M, et al: Effects of dopamine and
dobutamine on hyperoxic and hypoxic pulmonary vascular tone in
dogs. Am Rev Respir Dis 1987; 136: 29-35
16. Frostell CG, Blomquist H, Hedenstierna G, et al: Inhaled nitric oxide
selectivity reverses human hypoxic pulmonary vasoconstriction without
causing systemic vasodilation. Anesthesiology 1993; 78: 427-35
93
17. Hales CA, Rouse E, Slate JL: Influence of aspirin and indomethacin
on variability of alveolar hypoxic vasoconstriction. J Appl Physiol 1978;
45:33
18. Lejeune P, Deloof T, Leeman M, et al: Multipoint pulmonary vascular
pressure/flow relationships in hypoxic and normoxic dogs: effects of
nitrous oxide with and without cyclooxygenase inhibition.
Anesthesiolgy 1988; 68: 92-9
19. Melot C, Naeije R, Rothchild T, et al: Improvement in ventilation-
perfusion matching by almitrine in COPD. Chest 1983; 83: 528-33
20. Benumof JL, Pirlo AF, Johanson I, et al: Interaction of PVO2 with
PAO2 on hypoxic pulmonary vasoconstriction. J Appl Physiol 1981;
51:871-4
21. Benumof JL. Choice of anesthetic drugs and techniques. In Benumof
JL ed. Anesthesia for thoracic surgery. Philadelphia. WB Saunders.
1995; pp 300-330
22. Sykes MK, Loh L, Seed RF et al. The effect of inhalational
anaesthetics on hypoxic pulmonary vasoconstriction and pulmonary
vascular resistance in the perfused lungs of the dof and the cat. Br J
Anaesth 1972; 44: 776
23. Marshall C, Lindgren L, Marshall BE. Effects of halothane, enflurane,
and isoflurane on hypoxic pulmonary vasoconstrictiuon in rat lung in
vitro. Anesthesiology 1984; 61: 304
94
24. Bjertnaes LJ. Hypoxia-induced pulmonary vasoconstriction in man:
inhibition due to diethyl-ether and halothane anesthesia. Acta Anesth
Scand 1978; 22: 570-9
25. Buckley MJ, McLaughlin JS, Fort LIII, et al. effects of anesthetic
agents on pulmonary vascular resistance during hypoxia. Surg Forum
1964; 15: 183-92
26. Chen L, Marshall BE. Hypoxic pulmonary vasoconstriction and choice
of anesthesia. In. Cohen E ed. Thoracic anesthesia. Philadelphia. JB
Lippincott. 1995, pp 111-144
27. Susmano A, Passovoy M, Carleton RA. Comparison of the effects of
two anesthetic agents on the production of hypoxic pulmonary
hypertension in dogs. Am Heart J 1972; 84:203-10
28. Bindslev L, Cannon D, Sykes MK. Effect of lignocaine and nitrous
oxide on pulmonary vasoconstriction with dog constant flow perfused
left lower lobe preparation. Br J Anaesth 1986; 58: 315-21
29. Rees D, Galnes GY. One-lung anesthesia – a comparison of
pulmonary gas exchange during anesthesia with ketamine or
enflurane. Anesth Analg 1984; 63: 521-7
30. Naeije R, Lejeune P, Leeman M, et al. Effects of propofol on
pulmonary and systemic arterial pressure-flow relationship in
hyperoxic and hypoxic dogs. Br J Anaesth 1989; 62: 532-7
31. Van Keer L, Van Aken H, Vandermeersch E, et al. Propofol does not
inhibit hypoxic pulmonary vasoconstriction in humans. J Clin Anesth
1989; 1: 284-7
95
32. Alfery DD, Benumof JL, Trousdale FR: Improving oxygenation during
one-lung ventilation in dogs: the effects of positive end-expiratory
pressure and blood flow restriction to the non-ventilated lung.
Anesthesiology 1981; 55:381-5
33. Anderson HW, Benumof JL: Intrapulmonary shunting during one-
lung ventilation and surgical manipulation. Anesthesiology 1981; 55:
A377
34. Ishibe Y, Marshall C, Marshall BE: Hypoxic pulmonary
vasoconstriction inhibited by lung manipulation in rabbits.
Anesthesiology 1988; 69: A139
35. Slinger P, Suissa S, Adam J, et al: Predicting arterial oxygenation
during one-lung ventilation with continuous positive airway pressure to
the non-ventilated lung. J Cardiothorac Vasc Anesth 1990; 4: 436-40
36. Arborelius M, Lundin G, Svanberg L, et al: Influence of unilateral
hypoxia on blood flow through the lungs in man in lateral position. J
Appl Physiol 1960; 15: 595-7
37. Cooper JD, Patterson GA, Sundaresan RS, et al. Results of 150
consecutive bilateral lung volume reduction procedures in patients with
severe emphysema. J Thorac Cardiovasc Surg 1996; 112: 1319-30
38. Lima O, Ramos L, DiBiasi P, et al: Median sternotomy for bilateral
resection of emphysematous bullae. J Thorac Cardiovasc Surg 1981;
82: 892-7
96
39. Wasnick JD, Acuff T: Anesthesia and minimally invasive
thoracoscopically assisted coronary artery bypass: a brief clinical
report. J Cardiothorac Vasc Anaesth 1997; 11: 552-5
40. Pasque MK, Cooper JD, Kaiser L, et al: Improved technique for
bilateral lung transplantation: rationale and initial clinical experience
Ann Thorac Surg 1990; 49:785-91
41. Urschel HC, Razzuk MA: Median sternotomy as a standard approach
for pulmonary resection. Ann Thor Surg 1986; 41: 130-4
42. Fiser WP. Friday CD, Read RC: Changes in arterial oxygenation and
pulmonary shunt during thoracotomy with endobronchial anesthesia. J
Thorac Cardiovasc Surg 1982; 83: 523-31
43. Bardoczky GI, Szegedi LL, d’Hollander AA, et al.: Two-lung and one-
lung ventilation in patients with chronic obstructive pulmonary
disease: the effects of position and FiO2. Anesth Analg 2000; 90: 35-
41
44. Szegedi LL, Barvais L, d”Hollander AA : Dependent vs.
nondependent lung ventilation of COPD patients in supine and lateral
positions. Anesthesiology 1999; 91: 1344
45. Benumof JL: Separation of the two lungs (double-lumen tube and
bronchial blocker intubation). In JL Benumof (ed): Anesthesia for
thoracic surgery- 2nd ed, Philadelphia, WB Saunders, 1995, pp 330-
405
46. Brodsky JB: Separation of the lungs. In Cohen E (ed): The practice
of thoracic anesthesia. Philadelphia, JB Lippincott, 1995, pp 271-307
97
47. Chiaranda M, Rossi A, Manani G, et al: Measurements of the flow-
resistive properties of double-lumen bronchial tubes in vitro.
Anesthesia 1989; 44: 335-40
48. Hannallah MS, Benumof JL, McCarthy PO, et al: Comparison of three
techniques to inflate the bronchial cuff of left polyvinyl chloride double-
lumen endobronchial tubes. Anesth Analg 1996; 82: 867-9
49. Bardoczky GI, d’Hollander AA, Yernault JC, et al: On-line expiratory
flow-volume curves during thoracic surgery: the occurrence of auto-
PEEP. Br J Anaesth 1994; 72: 25-8
50. Wright PE, Marini JJ, Bernard GR. In vitro versus in vivo comparison
of endotracheal tube airflow resistance. Am Rev Resp Dis 1989; 140:
10-16
51. Cohen E: Anesthetic management of one-lung ventilation. In E
Cohen (ed): The practice of thoracic anesthesia. Philadelphia, JB
Lippincott, 1995, pp 308-340
52. Hurford WE, Alfille PH: A quality improvement study of the
placement and complications of double-lumen endobronchial tubes. J
Cardiothorac Vasc Anesth 1993; 7: 517-20
53. Hammond JE, Wright DJ: Comparison of the resistances of double-
lumen endobronchial tubes. Br J Anaesth 1984; 56:299-302
54. Rossi A, Polese G, Brandi G, et al: Intrinsic positive end-expiratory
pressure. Intensive Care Med 1995; 21: 522-36
98
55. Benumof JL: Conventional and differential lung management of one-
lung ventilation. In Benumof JL (ed): Anesthesia for thoracic surgery.
2nd ed, Philadelphia, WB Saunders, 1995, pp 406-431
56. Szegedi LL, Bardoczky GI, Engelman EE, et al: Airway pressure
changes during one-lung ventilation. Anesth Analg 1997; 84:1034-7
57. Bardoczky G, Engelman E, d’Hollander A: Continuous spirometry: an
aid to monitoring ventilation during operation. Br J Anaesth 1993; 71:
747-51
58. Siafakas NM, Vermeire P, Pride NB, et al: ERS Consensus
Statement. Optimal assessment and management of chronic
obstructive pulmonary disease (COPD). Eur Respir J 1995; 8: 1398-
420
59. Voelkel NF: Mechanism of hypoxic pulmonary vasoconstriction. Am
Rev Respir Dis 1986; 133: 1186-95
60. Fielding JE: Smoking and women: the tragedy of the majority. N
Engl J Med 1987; 317; 1343-5
61. Arsenian MA: Cardiovascular sequelae of therapeutic thoracic
radiation. Progr Cardiovasc Dis 1991; 33:299-312
62. Selvin BL.: Cancer chemotherapy: Implications for the
anesthesiologist. Anesth Analg 1981; 60; 425-34
63. Rudders RA, Mensley GT: Bleomycin pulmonary toxicity. Chest
1976; 63; 626-8
99
64. Milic-Emili J: Respiratory mechanics in COPD. In Milic-Emili J (ed):
Applied physiology in respiratory mechanics. Milano, Springer, 1998,
pp 95-106
65. Rossi A, Ranieri MV: Positive end-expiratory pressure. In: Tobin MJ
(ed): Principles and practice of mechanical ventilation. New York,
McGraw-Hill, 1994, pp 259-303
66. Similowski T, Derenne JP, Milic-Emili J: Respiratory mechanics
during acute respiratory failure of chronic obstructive pulmonary
disease. In Derenne JP, Whitelaw WA, Similowski T (eds): Acute
respiratory failure in chronic obstructive disease. New York, Marcel
Dekker, 1996, pp 40-64
67. DeTroyer A, Pride NB: The chest wall and respiratory muscles in
chronic obstructive pulmonary disease. In Roussos C (ed): The thorax.
Part C, Disease, New York, Marcel Dekker Inc, 1995, pp 1975-2005
68. Kimball WR, Leith DE, Robins AG: Dynamic hyperinflation and
ventilator dependence in chronic obstructive pulmonary disease. Am
Rev Respir Dis 1982; 126: 991-5
69. Brown DG, Pierson DJ: Auto-PEEP is common in mechanically
ventilated patients: a study of incidence, severity and detection. Respir
Care 1986; 31:1080-2
70. Rossi A, Gottfried SB, Zocchi L, et al: Measurement of static
compliance of the total respiratory system in patients with acute
respiratory failure during mechanical ventilation. Am Rev Respir Dis
1985; 131:672-7
100
71. Pepe PE, Marini JJ: Occult positive end-expiratory pressure in
mechanically ventilated patients with airflow obstruction. Am Rev
Respir Dis 1982; 126: 166-70
72. Siembruner G: Inadvertent positive end-expiratory pressure in
mechanically ventilated newborn infants: detection and effect on lung
mechanics and gas exchange. J Pediatr 1986; 108: 589-95
73. Marini JJ, Culver BH, Kirk W: Flow resistance of exhalation valves
and positive end-expiratory pressure devices used in mechanical
ventilation. Am Rev Respir Dis 1985; 131:850-4
74. Bernasconi M, Ploysongsang Y, Gottfried SB: Respiratory compliance
and resistance in mechanically ventilated patients with acute
respiratory failure. Intensive Care Med 1988; 14: 547-53
75. Bardoczky GI, d´’Hollander AA, Cappello M, et al: Interrupted
expiratory flow on automatically constructed flow-volume curves may
determine the presence of intrinsic positive end-expiratory pressure
during one-lung ventilation. Anesth Analg 1998; 86: 880-4
76. Bardoczky GI, Yernault JC, Engelman EE, et al: Intrinsic positive
end-expiratory pressure during one-lung ventilation for thoracic
surgery. The influence of preoperative pulmonary function. Chest
1996; 110: 180-4
77. Scott LR, Benson MS, Bishop MJ: Relationship of endotracheal tube
size to auto-PEEP at high minute ventilation. Respir Care 1986; 31:
1080-2
101
78. Slutsky AS: Consensus conference on mechanical ventilation. Int
Care Med 1994; 20: 64-79
79. Flacke JW, Thompson DS, Read RC: Influence of tidal volume and
pulmonary artery occlusion on arterial oxygenation during
endobronchial anesthesia. Southern Med J 1976; 69:619-26
80. Katz JA, Laverne RG, Fairley HB, et al: Pulmonary gas exchange
during endobronchial anesthesia: effects of tidal volume and PEEP.
Anesthesiology 1982; 56: 164-71
81. Torda TA, McCulloch CH, O’Brien HD, et al: Pulmonary venous
admixture during one-lung ventilation. Anesthesia 1974; 29: 272-9
82. Szegedi LL, Barvais L, Sokolow Y, Yernault JC, d’Hollander AA.
Intrinsic positive end-expiratory pressure during one-lung ventilation
of patients with pulmonary hyperinflation. Influence of low respiratory
rate with unchanged minute volume. Br J Anaesth 2001; 88: 56-60
83. D’Angelo E, Prandi E, Milic-Emili J: Dependence of maximal
expiratory flov-volume curves on the time course of preceeding
inspiration. J Appl Physiol 1993; 75: 1155-9
84. Slinger P: New trends in anesthesia for thoracic surgery including
thoracoscopy. Can J Anaesth 1995; 42: 77-84
85. Hedenstierna G, Baehrendtz S, Klingstedt C, et al: Ventilation and
perfusion of each lung during differential ventilation with selective
PEEP. Anesthesiology 1984; 61: 369-76
102
86. Gunnarson L, Tokics L, Lundquist H, et al.Chronic obstructive
pulmonary disease and anesthesia: formation of atelectasis and gas
exchange impairement. Eur Respir J 1991; 4:1106-16
87. Khanam T, Branthwaite MA: Arterial oxygenation during one-lung
anesthesia. Anesthesia 1973; 28: 132-8
88. Fuleihan SF, Wilson RS, Pontoppidan H: Effect of mechanical
ventilation with end-inspiratory pause on blood gas exchange. Anesth
Analg 1976; 55: 122-30
89. Georgopoulos D, Mitrouska I, Markopoulou K, et al. : Effects of
breathing patterns on mechanically ventilated patients with chronic
obstructive pulmonary disease and dynamic hyperinflation. Intensive
Care Med 1995; 21:880-6
90. Bardoczky GI, d’Hollander AA, Rocmans PA, et al: Respiratory
mechanics and gas exchange during one-lung ventilation for thoracic
surgery: the effects of end-inspiratory pause in stable COPD patients. J
Cardiothorac Vasc Anesth 1998; 12: 137-41
91. Slinger PD: Perioperative fluid management for thoracic surgery:
the puzzle of postpneumectomy pulmonary edema. J Cardiothorac
Vasc Anesth 1995; 9: 442-51
92. Williams EA, Evans TW, Goldstraw P: Acute lung injury following
lung resection: is one-lung anesthesia to blame? Thorax 1996; 51:
114-6
93. Slinger P: Predicting arterial oxygenation during one-lung
anesthesia. Can J Anaesth 1992; 39: 1030-5
103
94. Cohen E, Eisenkraft JB: Positive end-expiratory pressure during one-
lung ventilation improves oxygenation in patients with low arterial
oxygen tensions. J Cardiothorac Vasc Anesth 1996; 10: 578-82
95. Yokota K, Toriumi T, Sari A, et al: Auto-positive end-expiratory
pressure during one-lung ventilation using a double-lumen
endobronchial tube. Anesth Analg 1996; 82: 1007-10
96. Tugrul M, Çamci E, Karadeniz H, et al: Comparison of volume
controlled with pressure controlled ventilation during one-lung
anaesthesia. Br J Anaesth 1997; 79: 306-10
97. Desiderio DP, Meister M, Bedford RF: Intraoperative re-expansion
pulmonary edema. Anesthesiology 1987; 67: 821-3
98. Waller DA, Turner N: Re-expansion pulmonary oedema.
Anesthesiology 1989, 44: 446-7
104
3.13. Summary
During one-lung ventilation there is a decrease in arterial oxygenation.
This is due mainly to an increase in right-to-left shunt due to continuing
blood flow through the nonventilated lung. Fortunately there are some
major factors that tend to diminish this shunt (e.g., hypoxic pulmonary
vasoconstiction, gravity, surgical manipulation). Pre-existing lung disease,
as well as some pharmacological and physiological factors may affect also
the redistribution of pulmonary blood flow. One-lung ventilation adds to
the complexity of anesthetic technique and to the associated level of risk.
The anesthesiologist should have a logical plan for dealing with ventilatory
problems and hypoxia during one-lung ventilation. Knowledge of
physiology and pathophysiology of one-lung ventilation is mandatory in
management of the increasing number of highly diseased patients.
105
4. Airway pressure changes during one-lung
ventilation
Modified from: Szegedi LL, Bardoczky GI, Engelman E, d’Hollander AA. Airway Pressure
Changes During One-Lung Ventilation Anesth Analg 1997; 84: 1034-7
106
4.1. Introduction
Continuous methods to determine double-lumen endotracheal tube (DLT)
position during thoracic surgery have yet to be explored. The usefulness
of the continuously displayed pressure-volume and flow-volume curves
has been already established in the assessment of DLT position during
thoracic surgery (1). However, the change in configuration of the
pressure-volume and flow-volume curves that accompanies DLT
malposition may be easily missed when the curves are not constantly
observed.
Increased peak (Ppeak) and plateau (Pplateau) inspiratory airway pressures at
the beginning of one-lung ventilation (OLV) may also suggest malposition
of the DLT. The acceptable upper limit of the Ppeak at the onset of OLV,
however, seems to be controversial. Cohen (2) proposed Ppeak increases
above 40 cmH2O as a sign of a malpositioned tube. Slinger (3) choose 45
cmH2O as admissible Ppeak during OLV. Another approach is to compare
the pressure values during OLV with those observed during two-lung
ventilation (TLV). According to the report of Ovassapian (4), the tube is
correctly positioned when the Ppeak during OLV does not exceed 150% of
the baseline value observed during TLV.
The aims of this prospective study were to investigate changes in Ppeak and
Pplateau when changing from TLV to OLV in the supine position in patients
ventilated with DLTs and to evaluate three tests commonly used as
markers of malpositioned DLTs.
107
4.2. Methods
After institutional approval and informed consent were obtained, we
studied 55 consecutive patients scheduled for lung surgery in the supine
position. Demographic and intraoperative data of the patients are shown
in Table 4.1.. General anesthesia was induced and maintained with
continuous infusion of propofol and inhaled isoflurane. Muscle paralysis
was attained with pancuronium bromide.
Table 4.1. Demographic data and preoperative pulmonary function studies of 51 patients during
lung surgery.
Age (yr) 60.27 ± 12.66
Weight (kg) 69.50 ± 12.83
Height (cm) 168.50 ± 7.97
FEV1 (% pred) 75.45 ± 20.31
FRC (% pred) 134.72 ± 33.41
RV (% pred) 145.03 ± 42.58
Thoracotomies (right/left) (n) 24/27
DLT sizes (35/37/39/41) 1/15/21/14
Data are presented as mean ± SD.
Abbreviations: FEV1 = forced expiratory volume in 1s, % pred = percentage of the
predicted value, FRC = functional residual capacity, RV = residual volume, DLT = double-
lumen endobronchial tube.
108
Electrocardiogram, invasive arterial blood pressure, and arterial oxygen
saturation were monitored continuously. The trachea was intubated with a
disposable DLT (Broncho-cath™, Mallinckrodt Laboratories, Athlone,
Ireland) using a classical method of intubation without fiberoptic
bronchoscopy (FOB). Tube size was chosen according to a formula based
on patients’ sex and height (5). Ventilation was monitored with the Ultima
SV™ respiratory monitor (Datex Instrumentarium, Helsinki, Finland).
Mechanical ventilation was with a Siemens 900 C (Siemens Elema, Solna,
Sweden) constant flow ventilator with a tidal volume of 10 ml/kg at a
ventilatory rate of 10/min. Inspiratory time was 33%, and end-inspiratory
pause was 10% of the total cycle time. Ventilatory settings were kept
constant during the study. In the supine position, Ppeak and Pplateau were
recorded while the two lungs were ventilated (TLV). Then, the tracheal
lumen of the DLT was clamped (OLV), and ventilatory data were recorded
again. After data collection, FOB was performed first through the tracheal
then through the bronchial lumen.
The position of DLTs were evaluated according to the criteria described by
Smith et al. (6). In left-sided intubations through the tracheal lumen, the
bronchial cuff was below the level of the carina. Through the bronchial
lumen, the bronchial carina was not visualized, which indicates left upper
lobe obstruction. In right-sided intubations, the blue bronchial cuff was
below the level of the carina when the FOB was introduced through the
tracheal lumen. Through the bronchial lumen it was impossible to enter
109
the right upper lobe through the bronchial opening slot, which indicates
right upper lobe obstruction.
Data are presented as mean ± SD. The χ² test was used to compare the
incidence of malpositions according to the side of the DLTs. Wilcoxon’s
signed rank test was used to analyze the data. From our data, sensitivity
and specificity were calculated (7) using cutoff values of three commonly
used tests reported in the literature (2-4). A true positive result was
defined as occurring when the DLT was malpositioned and the pressure
was above the cutoff value; a false-positive result was defined as
occurring when the pressure was above the cutoff value but the DLT was,
in fact, well positioned; a true negative result was defined as occurring
when the DLT was well positioned and the pressure was below the cutoff
value; a false-negative result was defined as occurring when the pressure
was below the cutoff value but the DLT was malpositioned. Standard
formulas (7) were used to calculate the sensitivity (True positive / True
positive + False negative), specificity (True negative / True negative +
False positive), positive predictive value (True positive / True positive +
False positive), negative predictive value (True negative / True negative +
False negative), and diagnostic accuracy (True positive + True negative /
True positive + True negative + False positive + False negative). A
probability value of less than 0.05 was considered significant.
110
4.3. Results
Of the 55 DLTs inserted using a classical method, subsequent FOB
revealed 36 well-positioned and 19 malpositioned DLTs. Of the 19
malpositioned DLTs, 15 were inserted too far into the respective mainstem
bronchus, 1 was found to be twisted to the other side, and other 3 were
placed too far out in the trachea. As these types of malpositions do not
influence the magnitude of airway pressure changes in the same way as
malpositions placed in too far, only the 15 DLTs placed too far in were
included in further analysis. Malpositions were equally distributed between
the left-sided and right-sided DLTs (Table 4..2.).
Table 4.2. Position of the tubes according to side of DLT
Correctly positioned
DLTs (n=36)
Malpositioned DLTs
(n=15)
Total
Left 17 7 24
Right 19 8 27
Total 36 15 51
DLT = double-lumen endobronchial tube
Inspiratory airway pressures increased significantly in all cases when
switched from TLV to OLV with unchanged ventilatory settings. This
increase was observed both in the well-positioned and in the
malpositioned DLTs. The magnitude of pressure increase was nevertheless
different between the two groups.
111
When the DLT was well-positioned, baseline TLV Ppeak increased by a mean
of 55.1% (from 17.2 ± 4.1 cmH2O to 26.7 ± 6.6 cmH2O), and Pplateau Pplateau
increased by a mean of 41.9% (from 12.3 ± 3.3 cmH2O to 17.5 ± 4.7
cmH2O). In contrast, when the DLT was malpositioned, pressure increases
were significantly larger: Ppeak augmented by a mean of 73.4% (from 21.0
± 5.8 cmH2O to 36.5 ± 8.2 cmH2O) and Pplateau by a mean of 76.2% (from
13.7 ± 3.2 cmH2O to 24.2 ± 6.8 cmH2O).
When the individual Ppeak (Figure 4.1.) and Pplateau (Figure 4.2.) values
were divided according to the FOB-confirmed position of the DLT,
considerable overlap was evident in the pressure values of the 36
correctly positioned DLTs; overlap was less evident in the 15 incorrectly
positioned DLTs.
Using the pressure limit of 45 cmH2O recommended by Slinger (3), only
13.3% of the malpositions would have been detected, whereas the 40
cmH2O limit would have identified 40% of the malpositioned DLTs in this
study. Ppeak values greater than 40 cm H2O were not observed in the 36
well-positioned DLTs. Of the 15 incorrectly positioned DLTs, Ppeak was
greater than 40 cm H2O in 6 cases (40%); in two patients (13.3%) Ppeak
greater than 45 cm H2O was observed (Table 4.3.). If the cutoff value of
pressure was altered to 35 cm H2O, 66.7% of the malpositions would have
been identified (Table 4.3.).
During transition to OLV in 34 cases (67.9%), increases in Ppeak were
greater than 150% of the values observed during TLV. Despite that, use
of this marker would have detected 73.3% of the malpositions; the low
112
specificity (28.8%) (Table 4.4.) may indicate an increased FP rate.
Indeed, of these 34 cases, 23 DLTs (67.6%) were well-positioned and 11
(32.4%) were malpositioned.
Table 4.3. Sensitivity, Specificity, Positive Predictive Value, Negative predictive Value, and
Diagnostic Accuracy of Selected Values of Peak Airway Pressures (Ppeak)
35 cmH2O 40 cmH2O 45 cmH2O
Sensitivity 0.67 0.4 0.13
Specificity 0.89 1 1
Positive predictive
value
0.71 1 1
Negative
predictive value
0.86 0.8 0.73
Diagnostic
accuracy
0.82 0.82 0.74
Table 4.4. Sensitivity, Specificity, Positive Predictive Value, Negative predictive Value, and
Diagnostic Accuracy of Selected Values of Peak Airway Pressure as Percentages of the Value
Observed During Two-Lung Ventilation Increase during one-lung ventilation 140 % 150 % 160 %
Sensitivity 0.8 0.73 0.6
Specificity 0.25 0.36 0.53
Positive predictive value, 0.31 0.32 0.35
Negative predictive value 0.75 0.76 0.76
Diagnostic accuracy 0.41 0.47 0.55
Figure 4.1. Individual values of the peak inspiratory pressures (Ppeak) during one-lung ventilation
(OLV) and two-lung ventilation (TLV) in 51 patients. **p<0.01. DLT= double-lumen endobronchial
tube.
113
Figure 4.2. Individual values of the plateau inspiratory pressures (Pplateau) during one-lung
ventilation (OLV) and two-lung ventilation (TLV) in 51 patients. **p<0.01. DLT= double-lumen
endobronchial tube.
114
4.4. Discussion
This investigation examines the changes in inspiratory airway pressures
during transition from TLV to OLV in patients with a DLT. It was found
that, although the differences in pressures related to position (correct or
incorrect) were statistically significant, they cannot by themselves be used
for clinical decision making due to their low sensitivity (single-pressure
values) or poor diagnostic accuracy (relative-pressure values).
During mechanical ventilation, Ppeak and Pplateau pressures are commonly
available for monitoring purposes. Ppeak is the maximum pressure during
lung inflation, and its increase may be a sign of conditions associated with
decreased compliance and/or increased flow resistance (8). In intubated
patients, Ppeak is also influenced by the flow-resistive properties of the
endotracheal tube (8). Pplateau is defined as the end-inspiratory pressure
during a period of at least 0.5 seconds of zero gas flow (9). As there is no
flow during this period, Pplateau is not influenced by the resistive
characteristics of the patient-ventilator system (9) and, therefore,
increases of Pplateau do not provide information about the position of the
DLT.
In this study of 51 patients, the onset of OLV was associated with a
significant increase in both Ppeak (Figure 4.1.) and Pplateau (Figure 4.2.)
independent of whether the DLT was well-positioned or malpositioned. The
proportion of pressure increase, however, was significantly higher in the
malpositioned DLTs. Nevertheless, inspiratory airway pressures were quite
115
variable during TLV and OLV, with a considerable overlap of both Ppeak and
Pplateau.
Pressure monitoring may or may not detect an initially malpositioned DLT,
depending on the cutoff value of pressure used. Three commonly used
tests are reported in the literature as markers of malpositioned DLTs (2-
4).
Using a cutoff value of 45 cm H2O (3), only 13.3% of the malpositions
that occurred in this study would have been detected. The 40 cm H2O
pressure limit (2) seems to be more effective, as 40% of the malpositions
of this study would have been identified by using this cutoff value. Low
sensitivity of a diagnostic test indicates high false negative rate (7).
Indeed, in this study, 49 of the 51 patients had Ppeak values less than 45
cm H2O, and 45 of the 51 patients had Ppeak less than 40 cm H2O during
OLV. Using the lower limit of 35 cm H2O to detect DLT malposition
improved the sensitivity of the method to 66.7%, with only a slight
decline in specificity. Decreasing the pressure limit further (Table 4.3.)
decreased the specificity of the method and increased the FP rate to a
higher (55.5%) level.
Because a single value of Ppeak may not reflect the baseline TLV
characteristics of the patient-ventilator system, Ovassapian (4) relates the
value of Ppeak during OLV to that observed during TLV and considers the
limit of acceptable Ppeak to be less than 150% of the TLV value. Using the
reference value of Ovassapian (4), 73.3% of the malpositions that
occurred in this study would have been identified. Unfortunately, in 63.8%
116
of well-positioned DLTs, Ppeak increase was also greater than 150% of TLV
pressure (high false positive rate).
In conclusion, although the pressure differences related to position are
statistically significant, as a single value they cannot be used for clinical
decision making. A valuable feature in monitoring peak airway pressure is
the adjustable high alarm: when appropriately set, the alarm may warn
about a position-related problem. An appropriately set high-pressure
alarm associated with pressure-volume and flow-volume curves (1) may
improve the discriminating feature of both monitoring modalities.
Remarks
As mentioned before, of the initially 19 malpositioned DLTs, 15 were
inserted too far in the respective mainstem bronchus, 1 was found to be
twisted to the other side, and other 3 were placed too far out in the
trachea. As these types of malpositions do not influence the magnitude of
airway pressure changes in the same way as malpositions placed too far
in, only the 15 DLTs placed too far in were included in further studies.
However, some discussion is needed, given that the above mentioned
malpositions are also possibilities that may occur during DLT insertion. If
initiating OLV in case of a DLT placed too far out, thus in the trachea, no
pressure changes will be observed; the pressure will remain the same as
during TLV. This will give us a a higher false negative result (pressure
below the cutoff value, but the DLT is malpositioned), and consequently a
117
high sensitivity and high negative predictive value. However, in this case,
OLV monitoring doesn’t make sense anymore, as the two lungs remains
ventilated.
In case of the twisted tube to the other side, it was a right-sided DLT
inserted in the left mainstem bronchus. In this particular case, if the tube
is not positioned too far in the left mainstem bronchus, OLV is perfectly
possible, however inversed (ventilation of the would be operated lung via
the bronchial lumen and the other lung via the tracheal lumen). The other
theoretical situation that can occur by inversing a tube, is a left-sided DLT
inserted in the right mainstem bronchus. In this case, however, during
OLV, we will observe pressure changes similar to that observed with a
tube placed too far in, because obstruction of the right upper bronchus
orifice will occur; inversed OLV, like in the situation of a right-sided DLT
positioned in the left mainstem bronchus is no more possible.
118
4.5. References
1. Bardoczky G, Levarlet M, Engelman E, et al. Continuous spirometry
for detection of double-lumen endobronchial tube displacement. Br J
Anaesth 1993; 70: 499-502
2. Cohen E. Anesthetic management of one-lung ventilation. In: Cohen
E, ed. The practice of thoracic anesthesia. Philadelphia: JB
Lippincott, 1995: 308-40
3. Slinger P. New trends in anaesthesia for thoracic surgery including
thoracoscopy. Can J Anaesth 1995; 42: R77-84
4. Ovassapian A. Flexible bronchoscopic positioning of right-sided
double-lumen endobronchial tubes. J Bronchol 1995; 2: 12-9
5. Hannallah MS, Benumof JL, McCarthy PO, et al. Comparison of three
techniques to inflate the bronchial cuff of left polyvinyl chloride
double-lumen tubes. Anesth Analg 1993; 77: 990-4
6. Smith GB, Hirsch NP, Ehrenwerth J. placement of double-lumen
endobronchial tubes. Br J Anaesth 1986; 58: 1317-20
7. Griner PF, Mayewski RJ, Mushlin AI, et al. Selection and
interpretation of diagnostic tests and procedures: principles and
applications. Ann Intern Med 1981; 94: 557-600
8. Marini JJ. Lung mechanics determination at the bedside:
instrumentation and clinical application. Respir Care 1990; 35: 669-
96
9. Slutsky AS. Consensus conference on mechanical ventilation.
Intensive Care Med 1994; 20: 64-79
119
4.6.Summary
This investigation analyzed the changes in inspiratory airway pressures
during transition from two-lung to one-lung ventilation in patients
tracheally intubated with a double-lumen endotracheal tube, using a
classical method of intubation without fiberoptic bronchoscopy. All patients
were anesthetized in a standardized manner. Peak and plateau inspiratory
airway pressures were recorded with an on-line respiratory monitor before
and after clamping the tracheal limb of the double-lumen tube. The
position of the double-lumen tube was evaluated by fiberoptic
bronchoscopy with the patients in supine position. It was established that,
although the pressure differences related to position are statistically
significant, as a single value they cannot be used for clinical decision
making.
120
121
5. Two-lung and one-lung ventilation in patients
with chronic obstructive pulmonary disease:
the effects of position and FiO2
Modified from: Bardoczky GI, Szegedi LL, d’Hollander AA, Moures JM, de Francquen P,Yernault JC.
Two-lung and one-lung ventilation in patients with chronic obstructive pulmonary disease: the
effects of position and FiO2 . Anesth Analg. 2000; 90: 35-41
The authors dedicated this article to the memory of Dr. Gizella Bardoczky, who passed away in
July, 1998.
122
5.1. Introduction
As a result of the latest technical developments in cardiothoracic surgery,
a variety of thoracic procedures are performed with patients in the supine
instead of the traditional lateral position. Double-lung transplantation (1),
lung volume reduction surgery (2,3), and minimally invasive coronary
artery surgery (4) are routinely performed with patients in the supine
position. Median sternotomy has even been recommended for pulmonary
resection (5). During these interventions, one-lung ventilation (OLV) is
inherently required, generally inducing some degree of hypoxemia. The
major protective mechanism against hypoxemia is hypoxic pulmonary
vasoconstriction (HPV) (6), but in the lateral position, gravity-induced
blood flow redistribution must also be considered (7,8). Data comparing
gas exchange in the supine versus lateral position during OLV are lacking
(9). The aims of our study were to analyze the role of the position (supine
versus lateral) and the effect of various fraction of inspired oxygen (FiO2)
values during OLV of patients with chronic obstructive pulmonary disease
(COPD) who are scheduled for lung surgery.
123
5.2. Methods
After approval from the ethics committee of the hospital, informed consent
was obtained from 24 consecutive patients scheduled for elective thoracic
surgery. All subjects were studied in the supine and lateral positions both
during two-lung ventilation (TLV) and OLV, before surgery. Patients were
randomly assigned to receive FiO2 of 0.4 (Group 0.4 = eight patients), 0.6
(Group 0.6 = eight patients), or 1.0 (Group 1.0 = eight patients)
throughout the study. The patients had mild pulmonary hyperinflation
(functional residual capacity > 120%). All patients were premedicated with
midazolam 3 to 5 mg IM, approximately 30 min before the induction of
anesthesia. In the operating room, all patients had an epidural catheter
inserted around the T7 level, and after the administration of a test dose,
an epidural anesthesia was started with 8 mL of 2% lidocaine with 100 µg
fentanyl and was maintained with a continuous infusion of epidural
bupivacaine 0.5% (2–5 mL/h). The patients were anesthetized with a
variable-rate (3–6 mg.kg-1.min-1) continuous infusion of propofol and
inhaled isoflurane (0.4%–0.6%) in 40% or 60% oxygen in air or pure
oxygen. Muscle relaxation was achieved with pancuronium. In all patients,
the bronchus of the dependent lung was intubated with a double-lumen
endotracheal tube (DLT) (Broncho-cathTM; Mallinckrodt Laboratories,
Athlone, Ireland) of an appropriate size determined by the height of the
patient (10). The correct position of the DLT was determined with
fiberoptic bronchoscopy in both the supine and the lateral positions.
124
The electrocardiogram, heart rate, systemic arterial blood pressure,
noninvasive arterial oxygen saturation, and end-tidal CO2 were
continuously monitored. Constant tidal volume of 10 mL/kg was delivered
with a Siemens 900C ventilator (Siemens® ElemaTM, Solna, Sweden)
throughout the study. The ventilatory pattern consisted of a volume-
controlled, square-wave flow pattern, at a rate of 10 breaths/min.
Inspiratory time (TI/TTOT) was 0.33 and end-inspiratory pause was 10% of
the total respiratory cycle. All measurements were performed with zero
applied end-expiratory pressure. Ventilatory variables were kept constant
during the study, both during TLV and OLV. This investigation was
performed with the chest closed and before the surgical procedure.
After intubation in the supine position and fiberoptic control of the correct
DLT position, the two lungs were ventilated with the above-described
ventilatory pattern for 15 minutes. End-inspiratory and end-expiratory
occlusions were performed to determine the mechanical characteristics of
the respiratory system (peak inspiratory airway pressure, end-inspiratory
plateau pressure, and intrinsic positive end-expiratory pressure) and
arterial blood gas samples were drawn. Then, the tracheal lumen of the
DLT was clamped, and the would-be operated nondependent lung was
allowed to deflate to atmospheric pressure. After 15 min of OLV and data
collection, TLV was restored with unaltered ventilatory settings, and the
patient was turned to the lateral decubitus position. TLV and turning and
positioning of the patients generally lasted approximately 30 min. At the
end of this period, ventilatory data were recorded, an arterial blood gas
125
sample drawn, and the nondependent lung collapsed again, now with the
patient in the lateral decubitus position. After 15 min, respiratory
mechanics and gas exchange data were collected again. PaO2, alveolar-
arterial oxygen tensions difference [P(A-a)O2] were calculated by using
standard equations.
Demographic data and preoperative pulmonary functions of the three
groups were comparable (Kruskall-Wallis test). The data obtained in the
two positions were compared with the Wilcoxon matched pair test. The
data caused by FiO2 differences were analyzed with the Kruskall-Wallis
test. Values of P < 0.05 were considered as statistically significant. Data
were presented as median (range).
126
5.3. Results
Demographic characteristics, preoperative pulmonary function tests, and
the values of the preoperative arterial oxygen and carbon dioxide tensions
of the 24 patients are shown in Table 5.1.. The differences were not
statistically significant among the three groups. Right- and left-sided
thoracotomies were comparable in the three groups and among the groups
(Kruskall-Wallis test). Initiating OLV with unaltered ventilatory settings
increased peak inspiratory pressures significantly in all three groups
(Tables 5.2.–5.4.).
Table 5.1. Demographic Characteristics, Preoperative Pulmonary Function Studies, and
Preoperative Arterial Blood Gas Values of 24 Patients During Thoracic Surgery
Group 0.4 (n = 8)
Group 0.6 (n = 8)
Group 1.0 (n = 8)
Age (yr) 59 (46–62) 60 (24–72) 64 (44–78)
Height (cm) 175 (169–183) 169 (154–180) 167 (162–184)
Weight (kg) 75 (51–105) 69 (55–86) 65 (56–98)
FEV1 (% predicted) 76 (35–81) 79 (51–84) 80 (53–85)
FRC (% predicted) 152 (122–195) 145 (121–182) 145 (122–181)
RV (% predicted) 170 (128–177) 156 (131–196) 151 (124–193)
TLC (% predicted) 115 (99–131) 115 (98–138) 112 (77–135)
PaCO2 (mm Hg,
room air, supine) 37 (31–41) 37 (29–43) 40 (36–47)
PaO2 (mm Hg, room
air, supine)
78 (66–90)
82 (68–102)
72.5 (64–95)
Data are median (range).
Abbreviations: FiO2 : fraction of inspired oxygen, ; FEV1 : forced expiratory volume in 1 s, ; FRC :
functional residual capacity, ; RV : residual volume, ; TLC : total lung capacity.
127
5.2. Respiratory Mechanics and Gas Exchange Data Obtained During TLV and OLV with Patients in
the Supine and in the Lateral Positions, FiO2 = 0.4 (n = 8)
Supine
Lateral
TLV
OLV
TLV
OLV
Ppeak (cmH2O) 21 (12–35) 29 (17–40) 20 (12–35) 29 (17–42)
Pplateau
(cmH2O) 14 (7–23) 18 (9–26) 13 (9–19) 18 (12–24)#
PEEPi (cmH2O) 1.5 (0–2) 2.5 (0–6) 1 (0–3) 3 (0–8)
PaCO2
(mmHg) 38 (34–47) 39 (32–49) 40.5 (35–47) 38.5 (36–45)
PaO2 (mmHg) 124 (81–240) 63 (57–144)* 132 (102-296) 101 (72–201)*
##
P(A-a)O2
(mmHg)
118 (12–170)
181 (94–197)*
117 (16–166)
131 (44–177)
##
Data are median (range).
Abbreviations: TLV : two-lung ventilation, OLV : one-lung ventilation, Ppeak : peak inspiratory
airway pressure, Pplateau : end-inspiratory airway pressure, PEEPi : intrinsic positive end-expiratory
pressure, P(A-a)O2 : alveolar-arterial oxygen tension difference.
*p < 0.02, #p < 0.05 (OLV versus TLV).
##p < 0.02 (lateral versus supine OLV).
128
Table 5.3. Respiratory Mechanics and Gas Exchange Data Obtained During TLV and OLV with
Patients in the Supine and in the Lateral Positions, FiO2 = 0.6 (n = 8)
Supine
Lateral
TLV
OLV
TLV
OLV
Ppeak (cm H2O) 18 (15–22) 23 (16–35) 19 (12–32) 24 (16–32) #
Pplateau (cm
H2O) 13 (11–17) 15 (12–23) 14 (9–17) 15 (11–30)
PEEPi (cm
H2O) 1 (0–4) 2 (0–6) 1.5 (0–4) 3 (0–6)
PaCO2 (mm
Hg) 40 (36–52) 39 (31–46) 38 (35–46) 41 (35–51)
PaO2 (mm Hg) 256 (172–391) 155 (114–235)
*# 289 (225–353)
268 (162–311)
##
P(A-a)O2 (mm
Hg)
136 (101–198)
196 (163–254)
*#
126 (97–213)
151 (123–234)
##
Data are median (range).
Abbreviations: TLV : two-lung ventilation, OLV : one-lung ventilation, Ppeak : peak inspiratory
airway pressure, Pplateau : end-inspiratory airway pressure, PEEPi : intrinsic positive end-expiratory
pressure, P(A-a)O2 : alveolar-arterial oxygen tension difference.
*p < 0.02 , #p < 0.05 (OLV versus TLV).
##P < 0.02 (lateral versus supine OLV).
129
Table 5.4. Respiratory Mechanics and Gas Exchange Data Obtained During TLV and OLV with
Patients in the Supine and in the Lateral Positions, FiO2 = 1.0 (n = 8)
Supine
Lateral
TLV
OLV
TLV
OLV
Ppeak (cm H2O) 17 (16-21) 24 (14-32)* 19 (15-24) 24 (17-29)#
Pplateau (cm
H2O)
15 (8-18)
15 (8-19) 16 (9-19) 18 (11-22)
PEEPi (cm
H2O) 1.5 (0-3) 2 (0-4) 1.5 (0-4) 2.5 (0-6)
PaCO2 (mm
Hg) 39 (33-48) 38 (36-52) 38 (34-50) 40 (34-51)
PaO2 (mm Hg) 472 (232-591) 301 (216-422)# 492 (336-600) 486 (288-563)
##
P(A-a)O2 (mm
Hg)
185 (151-382) 267 (212-445)# 173 (144-376) 190 (153-396)
##
Data are median (range).
Abbreviations: TLV : two-lung ventilation, OLV : one-lung ventilation, Ppeak : peak inspiratory
airway pressure, Pplateau : plateau inspiratory airway pressure, PEEPi : intrinsic positive end-
expiratory pressure, P(A-a)O2 : alveolar-arterial oxygen tension difference.
*p < 0.02 , #p < 0.05 (OLV versus TLV).
##P < 0.02 (lateral versus supine OLV).
130
Effects of the position
During the period of TLV, the position of the patient (supine versus lateral)
did not significantly influence the variables related to gas exchange in the
three groups of patients (Tables 5.2.-5.4.).
In contrast, in the OLV assessment stages, the PaO2 values in the lateral
position were always significantly higher than in the supine position:
Group 0.4, 63 (57–144) vs. 101(72–201) mm Hg (P < 0.02); Group 0.6,
155 (114–235) vs. 268 (162–311) mm Hg (P < 0.02); Group 1.0, 301
(216–422) vs. 486 (288–563) mm Hg (P < 0.02) (Wilcoxon’s matched pair
test). The values of P(A-a)O2 showed similarly significant changes, but in
the opposite direction (Tables 5.2.-5.4. and Figure 5.1.).
While turning the patients from the supine into the lateral position, no
changes were observed regarding the variables related to the mechanical
properties of the respiratory system (peak inspiratory airway pressure,
end-inspiratory airway pressure, and intrinsic positive end-expiratory
pressure), during neither TLV, nor OLV (not significant, Wilcoxon matched-
pair test).
131
Figure 5.1. Individual modifications in arterial oxygenation secondary to changes in position in the
three groups of patients during thoracic surgery. The bold lines represent the median values.
Effects of FiO2
In the three groups of patients ventilated with different FiO2 values, the
differences of PaO2 values observed between the supine and lateral TLV
periods were not significant. In contrast, the increases of PaO2 values
observed between the lateral and supine OLV were significantly higher
when the FiO2 was higher (P < 0.02, Kruskall-Wallis test) (Figure 5.2.).
PaO2 was decreased and P(A-a)O2 was increased during OLV in
comparison with the TLV values in both the supine and the lateral
positions and at every level of FiO2. However, these changes were more
pronounced in the supine position (Group 0.4: 61 [17–132] vs. 19 [5–39])
mm Hg; Group 0.6: 68 [9–150] vs. 14 [4–44] mm Hg; Group 1.0: 107
[5–268] vs. 65 [18–205] mm Hg; P < 0.05, Wilcoxon’s matched pair test).
132
When turning the patients from the supine to the lateral position, PaO2
increased significantly when comparing the periods of OLV. These PaO2
increases (lateral-supine) were significantly larger when higher FiO2 values
were used (Figure 5.2.).
When comparing the periods of TLV, no significant differences in PaO2
increases (lateral-supine) were found, regardless of the FiO2 values
administered (Figure 5.2.).
There were no significant changes in mean arterial pressure or heart rate
in any of the patients during the study period.
Figure 5.2. Values of arterial oxygen tension increases between lateral and supine positions,
during TLV and OLV in the three groups of patients. The bold lines represent the median values.
TLV = two-lung ventilation, OLV = one-lung ventilation.
133
5.4. Discussion
In this study, we investigated arterial oxygenation with the patient in the
supine and in the lateral positions during TLV and OLV in patients with
mild pulmonary hyperinflation scheduled for thoracic surgery. We found
significantly higher PaO2 and lower P(A-a)O2 when OLV was performed in
the lateral position.
Recent developments in thoracic surgery have widened the scale of
surgical interventions and changed certain traditions, including the patient
positioning. Single-lung transplantation (11) and resection of unilateral
emphysematous bullae (3) are performed with the patient in the lateral
decubitus position, while double-lung transplantation, lung volume
reduction surgery, and minimally invasive coronary artery surgery are
performed with the patient placed in the supine position (1,2,12). In both
conditions, to facilitate the surgeon’s task, variable periods of OLV are
required during these procedures.
To explain the preservation of blood oxygenation in the presence of a
large fraction of atelectatic lung, as during OLV, in experimental or clinical
conditions, many factors have been mentioned, including principally HPV
(8,13), gravity (11), or local mechanical forces (12, 14). We have found
only one study comparing oxygenation during OLV in the supine and
lateral positions (9).
When OLV is performed and one of the lungs is excluded from ventilation,
an obligatory right-to-left shunt occurs through the nonventilated lung
that is not present during TLV, and arterial oxygenation decreases (7). The
134
pulmonary vessels in that nonventilated, hypoxic area respond by
increasing resistance to flow. This reflex HPV of vessels perfusing hypoxic
alveoli diverts blood flow from nonventilated lung units (7,13) to
ventilated regions and attenuates hypoxemia by actively reducing the
perfusion of nonventilated lung tissue.
In the supine position, during anesthesia and controlled mechanical
ventilation of both lungs, there are generally no significant differences in
the perfusion between the two lungs as both are exposed to the same
pressures of gravity (11). Starting OLV will initiate right-to-left shunt, and
as a consequence, PaO2 will decrease and P(A-a)O2 will increase. Indeed,
in this study, when OLV was initiated in the supine position, PaO2
decreased from 124 (81–240) to 63 (57–144) mm Hg (P < 0.02) in Group
0.4, from 255 (172–391) to 155 (114–235) mm Hg (P < 0.02) in Group
0.6, and from 472 (232–591) to 300.5 (216–422) mm Hg in Group 1.0.
Every patient’s P(A-a)O2 increased significantly during OLV (Tables 5.2.-
5.4.).
However, during TLV in the lateral position, the position of the patient
reduces perfusion of the upper lung caused by gravitational diversion of
blood flow to the dependent lung (7,8). As a result of anesthesia and
muscle relaxation, the distribution of ventilation also changes in the lateral
position because the applied positive-pressure ventilation displaces the
diaphragm preferentially at the nondependent part. Traditionally, this
discrepancy is thought to be disadvantageous (7). However, in our
patients, the change in position from the supine to lateral decubitus
135
position resulted in a modest change in PaO2 during TLV (124 [81–240]
and 132 [102–296] mm Hg in Group 0.4; 256 [172–391] and 289 [225–
353] mm Hg in Group 0.6; and 472 [232–591] and 492 [336–600] mm Hg
in Group 1.0, respectively). This is in agreement with the findings of Boldt
et al. (15) and Rehder et al. (16), who also reported no difference in PaO2
between the lateral and the supine positions with TLV.
Inducing OLV when the patient is in the lateral position activates HPV and
reduces the further perfusion of the collapsed lung. Hence, we found that
PaO2 was significantly greater when OLV was initiated after turning the
patients into the lateral decubitus position (Tables 5.2. –5.4.and Figures
5.1. – 5.2.).
Fiser et al. (9) studied the period of OLV in both the supine and the lateral
positions and did not find changes in arterial oxygen tension after 10 to 20
minutes of OLV when the patients were turned into the lateral position.
However, in the study of Fiser et al. (9), OLV was initiated in the supine
position and maintained continuously, even during the period of
positioning and turning the patient, with an FiO2 of 1.0.
The most important mechanism for reducing blood flow of an atelectatic
lung is HPV (6,13). When OLV is induced in the lateral position, the blood
flow of the nondependent lung is already reduced by gravitational forces,
and HPV further reduces blood flow. In contrast, in the supine position,
both lungs are equally exposed to an identical gravitational force; thus
during OLV, the reduction of blood flow depends solely on the strength of
the HPV.
136
Here, consideration should be given to possible limitations in our
experimental methods.
First, concerning the patients included in the study, the presence of
chronic airflow obstruction may be associated with better PaO2 during OLV
possibly because of dynamic hyperinflation resulting in an increased
functional residual capacity and intrinsic positive end-expiratory pressure
in the dependent lung (17). In contrast, in patients with severe COPD,
HPV may not be an important protective mechanism, as these patients
already have an increased pulmonary arterial pressure and reduced
pulmonary vascular bed. The amount of disease in the nondependent lung
is also a significant determinant of the amount of blood flow to the
nondependent lung. If the nondependent lung is severely diseased, there
may be a fixed reduction in blood flow to this lung preoperatively, and the
collapse of such a diseased lung may not further increase the shunt. PaO2
during TLV may also be a determining factor of oxygenation during OLV
(17).
Second, we did not study OLV for long periods (more than two hours) with
high FiO2 values, but according to the study of Barker et al. (18), late
hypoxemia occurs (mean OLV time, 170 minutes) when ventilating with
100% oxygen, which may be in part caused by absorption atelectasis.
High FiO2 can lead to arteriovenous shunting in areas of airway closure
and, further, cause absent ventilation in lung units with low
ventilation/perfusion ratios.
137
Third, if we consider that the lateral TLV/OLV always followed the supine
TLV/OLV sequence, one can argue that a time effect could influence the
present results. Concerning the time course of the HPV response, many
experimental and clinical studies have been performed, with sometimes
contradictory results. Benumof (19) showed that intermittent hypoxic
challenges potentiated the hypoxic vasoconstriction in the left lower lobe
of open-chest dog lungs, but the preparation and manipulation of the
animals required considerable instrumentation and manipulation, which
may have interfered with the HPV. Carlsson et al.(20) found maximal HPV
response within 15 minutes of hypoxia in a human study, which agrees
with observations in animal studies. Tucker and Reeves (21) were unable
to maintain HPV during acute hypoxia in anesthetized dogs. During one-
lung hypoxia in dogs, Domino et al. (22) studied HPV in closed-chest dogs
and found a maximal level from the very first hypoxic challenge. Thus,
there is a wide variation in the results obtained concerning the influence of
time on HPV. In our study, the experimental procedure was almost
identical to the procedure performed by Carlsson et al. (20) or by Domino
et al. (22), who concluded that the time factor should not be a hindrance
to manipulative studies on the HPV response once a maximal response
has been evoked, normally in 10–15 minutes.
Fourth, if we consider that, after the period of supine TLV, we have only
declamped the tracheal limb of the DLT and closed it, without sighing the
nondependent lung, the 5 mL/kg gas distributed to that lung will not
reverse the amount of atelectasis. So, our study may be comparable to
138
the study of Fiser et al.(9), but as significant changes were found in the
lateral position, we can argue that if HPV was maximal after the first 15
min of hypoxia, there is another factor, probably gravity, that could
contribute to flow redistribution. Nevertheless, in the particular conditions
of the present study - patients with mild COPD, closed chest, mixed
locoregional/general balanced anesthesia - increased HPV responses in the
lateral position cannot be excluded as explanations of PaO2 values
observed in the second OLV episode.
PaCO2, especially at high values, influences the level HPV response (23).
Thus, the PaCO2 values of the studied patients were always maintained
within normal limits in the different periods of blood gas measurements.
Another factor that may influence hypoxia during OLV is the side of the
surgery. Left thoracotomy has a better PaO2 during OLV than right
thoracotomy, because the left lung normally receives 10 percent less
cardiac output than the right lung (17). In our study, there were no
statistical differences concerning the side of surgery in the three groups or
among the three groups.
Gravity is a major, pharmacologically and physiopathologically,
independent determinant of regional pulmonary blood flow distribution.
The extent of blood flow redistribution depends on the local relationship
between pulmonary arterial, venous, and airway pressures. In the lateral
position, regional blood flow increases from the nondependent to the
dependent thoracic wall (8). Unfortunately, the design of our study did not
139
permit us to distinguish between the direct effects of HPV on blood flow
and the effects of gravitational redistribution of blood flow.
The differences observed among the three groups of patients supports the
hypothesis of Benumof et al.(24), who demonstrated, on a canine left
lower lobe preparation, that if FiO2 changes cause secondary changes in
PaO2, and in the mixed venous oxygen tension, then the mixed venous
oxygen tension is a new and important determinant of the magnitude of
HPV. This suggests that when one compartment FiO2 is 1.0 and the other
compartment is hypoxic, HPV in the hypoxic compartment is maximal.
The method used to ventilate the dependent lung is an important
determinant of the blood flow distribution during OLV. High airway
pressures can compress lung vessels, diverting blood flow from ventilated
to nonventilated regions. However, hypoventilation of the dependent lung
during OLV is associated with lower airway pressure, and the ventilated
lung pulmonary vascular resistance may decrease, thus promoting HPV in
the nonventilated lung (7). This mechanism of improved PaO2 was unlikely
in the present investigation, as ventilatory settings were kept constant,
inspiratory airway pressures were not changed, and the mechanical
characteristics of the dependent lung remained unaltered (Tables 5.2. – 5.
4.) after changing the position. These findings indicate that the improved
PaO2 cannot be attributed to an altered HPV caused by change in
ventilatory pattern.
Surgical compression and retraction may also contribute to the passive
reduction of nondependent lung blood flow (4). However, in our study,
140
there was no mechanical effect on lung parenchyma, as the data were
collected before chest opening. Therefore, surgical manipulation could not
influence the amount of shunt occurring.
We conclude that, in addition to HPV, the augmented redistribution of
perfusion caused by gravitational forces is probably responsible for the
higher PaO2 during OLV in the lateral position. The finding of significantly
lower PaO2 values occurring when OLV was initiated in the supine position
may predict more frequent intraoperative hypoxemia when thoracic
surgery requiring OLV is performed with patients in the supine position. If
severe hypoxemia occurs during OLV in the lateral position, higher FiO2
values might be used.
141
5.5. References 1. Pasque MK, Cooper JD, Kaiser L, et al. Improved technique for
bilateral lung transplantation: rationale and initial clinical experience.
Ann Thorac Surg 1990;49:785–91.
2. Cooper JD, Patterson GA, Sundaresan RS, et al. Results of 150
consecutive bilateral lung volume reduction procedures in patients
with severe emphysema. J Thorac Cardiovasc Surg 1996;112:1319–
30.
3. Lima O, Ramos L, DiBiasi P, Judice L. Median sternotomy for bilateral
resection of emphysematous bullae. Thorac Cardiovasc Surg
1981;82:892–7.
4. Wasnick JD, Acuff T. Anesthesia and minimally invasive
thoracoscopically assisted coronary artery bypass: a brief clinical
report. J Cardiothorac Vasc Anesth 1997;11:552–5.
5. Urschel HC, Razzuk MA. Median sternotomy as a standard approach
for pulmonary resection. Thorac Surg 1986;41:130–4.
6. Benumof JL. One-lung ventilation and hypoxic pulmonary
vasoconstriction: implications for anesthetic management. Anesth
Analg 1985;64:821–33.
7. Benumof JL. Physiology of the lateral decubitus position, the open
chest and one-lung ventilation. In: Kaplan JA, ed. Thoracic
anesthesia. New York:Churchill Livingstone, 1993:193–221.
142
8. Arborelius M, Lundin G, Svanberg L, Defares JG. Influence of
unilateral hypoxia on blood flow through the lungs in man in lateral
position. J Appl Physiol 1960;15:595–7.
9. Fiser WP, Friday CD, Read RC. Changes in arterial oxygenation and
pulmonary shunt during thoracotomy with endobronchial anesthesia.
J Thorac Cardiovasc Surg 1982;83:523–31.
10. Hannalah MS, Benumof JL, McCarthy PO, et al. Comparison of three
techniques to inflate the bronchial cuff of left polyvinylchloride
double-lumen tubes. Anesth Analg 1993;77:990–4.
11. Rehder K, Wenthe FM, Sessler AD. Function of each lung during
mechanical ventilation with ZEEP and with PEEP in man anesthetized
with thiopental-meperidine. Anesthesiology 1973;39:597–606.
12. Gayes JM, Emery RW. The MIDCAB experience: a current look at
evolving surgical and anesthetic approaches. J Cardiothorac Vasc
Anesth 1997;11:625–8.
13. Marshall BE. Hypoxic pulmonary vasoconstriction. Acta Anaesthesiol
Scand 1990;34:37–41.
14. Alfery DD, Benumof JL, Trousdale FR. Improving oxygenation during
one-lung ventilation in dogs: the effects of positive end-expiratory
pressure and blood flow restriction to the non-ventilated lung.
Anesthesiology 1981;55:381–5.
15. Boldt J, Muller M, Uphus D, et al. Cardiorespiratory changes in
patients undergoing pulmonary resection using different anesthetic
management techniques. J Cardiothorac Vasc Anesth 1996;7:854–9.
143
16. Rehder K, Knopp TJ, Sessler AD, Didier EP. Ventilation-perfusion
relationship in young healthy awake and anesthetized-paralyzed man.
J Appl Physiol 1979;47:745–53.
17. Slinger P, Suissa S, Adam J, Triolet W. Predicting arterial oxygenation
during one-lung ventilation with continuous positive airway pressure
to the nonventilated lung. Cardiothorac Anesth 1990;4:436–40.
18. Barker SJ, Clarke C, Trivedi N, et al. Anesthesia for thoracoscopic
laser ablation of bullous emphysema. Anesthesiology 1993;78:44–50.
19. Benumof JL. Intermittent hypoxia increases lobar hypoxic pulmonary
vasoconstriction. Anesthesiology 1983;58:399–404.
20. Carlsson AJ, Bindslev L, Santesson J, et al. Hypoxic pulmonary
vasoconstriction in the human lung: the effect of prolonged unilateral
hypoxic challenge during anaesthesia. Anaesthesiol Scand
1985;29:346–51.
21. Tucker A, Reeves JT. Non-sustained pulmonary vasoconstriction
during acute hypoxia in anesthetized dogs. Am J Physiol
1977;42:889–99.
22. Domino K, Chen L, Alexander C, et al. Time course and responses of
sustained hypoxic pulmonary vasoconstriction in the dog.
Anesthesiology 1984;60:562–6.
23. Benumof JL, Mathers JM, Wahrenbrock EA. Cyclic hypoxic pulmonary
vasoconstriction induced by concomitant carbon dioxide changes. J
Appl Physiol 1976;41:466–9.
144
24. Benumof JL, Pirlo AF, Johanson I, Trousdale FR. Interaction of PVO2
with PAO2 on hypoxic pulmonary vasoconstriction. J Appl Physiol
1981;51:871–4.
145
5.6.Summary
The effects of position and fraction of inspired oxygen on oxygenation
during one-lung ventilation for thoracic surgery were evaluated. Three
groups of patients, at three different fractions of inspired oxygen, were
compared during two and one-lung ventilation in supine and in lateral
positions. Arterial oxygen tension was decreased in all three groups during
one-lung ventilation in comparison with the two-lung ventilation values,
but this decrease was significantly less in the lateral, as compared with
the supine position.
146
147
6. Dependent versus nondependent lung
ventilation of COPD patients in supine and
lateral positions
Under review: Szegedi LL, d’Hollander AA, Barvais L, De Baerdemaeker L, Mortier EP. Dependent
versus nondependent lung ventilation of COPD patients in supine and lateral positions. 1999 ASA
Meeting Abstracts; Published as abstract in Anesthesiology, October 1999
148
6.1. Introduction Newer pulmonary surgical techniques are performed in the supine position
instead of the “classical”, lateral one (1-5). During these procedures,
variable periods of one-lung ventilation (OLV) are required, generally
inducing some degrees of hypoxemia. The normal response of the
pulmonary vasculature to atelectasis is an increase in pulmonary vascular
resistance (in the atelectatic lung), and the increase in atelectatic lung
resistance is almost entirely caused by hypoxic pulmonary
vasoconstriction (HPV) (6). In the lateral position, the gravity-induced
blood flow redistribution is also regularly considered (7-10).
Other studies, have either contradicted the gravitational model or
questioned the primordial role of this model as the perfusion to each lung
was not changed with a shift from the supine to the lateral position,
concluding that the redistribution of pulmonary blood flow in the lateral
position is dominated either by the vascular structure or other factors
excluding gravity (11-20).
In a recent study (7), it was demonstrated, that patients with chronic
obstructive pulmonary disease (COPD), undergoing lung surgery with OLV,
had a better arterial oxygen tension (PaO2) during OLV in the lateral as
compared to the supine position, and that, independently of the fraction of
inspired oxygen used (FiO2). The explanation was, that in addition to HPV,
the augmented redistribution of perfusion caused by gravitational forces is
probably the responsible for the higher PaO2 during OLV in the lateral
position.
149
The aims of this clinical study design were to try to separate the effects of
gravity from other factors that may influence arterial oxygenation during
OLV of patients with mild to moderate pulmonary hyperinflation scheduled
for lung surgery.
150
6.2. Materials and methods
Patient selection
After Institutional Review Board approval, informed consent was obtained
from 30 consecutive patients scheduled for elective right thoracotomy, for
localized lung tumor resection. Before surgery, pulmonary function was
tested in the sitting position, including spirometry and static lung volumes
determined by plethysmography (MasterScreen Body™, Jaeger and
Thoennies™, Erich Jaeger GmbH, Hoechberg, Germany). All the patients
had stable, mild to moderate degrees of chronic obstructive pulmonary
disease (COPD) (forced expired volume in 1s (FEV1) 55 to 79% of the
predicted value) (21) with pulmonary hyperinflation (functional residual
capacity, FRC >120%). Preoperative isotopic measurements of the
right/left lung perfusion ratios of the lungs were done in order to exclude
restrictive disorders due to the tumor (i.v. administration of 99mTc
macroaggregates of albumin, Macrotec, Sorin Biomedica, Saluggia, Italy).
Patients with significant right/left lung perfusion inequalities (more than 5
% right/left discrepancies) were not included in the study.
Anesthesia
Anesthesia was conducted in a standardized manner.
151
All patients were premedicated with alprazolam 0.5 mg p.o.,
approximately 60 to 90 minutes before arrival in the O.R.. In order to
assure per and postoperative analgesia, an epidural catheter at mid-
thoracic level (Th5-Th8) was inserted and tested (test dose of 3ml 2%
lidocain with epinephrine 1/200000) before the induction of general
anesthesia; the loading dose (8 ml 2% lidocain with 100µg fentanyl) was
given only after the end of the study. General anesthesia was induced and
maintained with variable rate continuous infusion of propofol (loading dose
2 mg/kg followed by continuous infusion of 3-6 mg .kgq1. min.-1).
Cisatracurium was used to allow tracheal intubation and to maintain
neuromuscular block throughout surgery; the neuromuscular block was
assessed by regular measurements of post-tetanic count during the
procedure.
Monitoring and ventilation
In all patients, the bronchus of the dependent left lung was intubated with
a double-lumen endotracheal tube (DLT) (Broncho-cath™, Mallinckrodt
Laboratories, Athlone, Ireland) of an appropriate size determined by the
height of the patient (22). The correct position of the DLT was ascertained
with fiberoptic bronchoscopy both in the supine and the lateral positions.
All subjects were studied with closed chest in the supine and in the lateral
position during OLV. Patients were randomly assigned to two groups,
according to the side of OLV; either the dependent (lower, left) lung
152
(Group D-OLV, n=15) or the nondependent lung (upper, right) (Group
ND-OLV, n=15) was ventilated.
Electrocardiogram, invasive radial arterial and central venous pressures,
and arterial oxygen saturation were continuously monitored. The Datex
Ultima SV capnometer (Datex Instrumentarium, Helsinki, Finland) was
used to measure expired end-tidal carbon dioxide tension (ETCO2), and
display flow-volume and pressure-volume loops. Constant tidal volume of
10 ml/kg was delivered with a Siemens 900C (Siemens® Elema™, Solna,
Sweden) ventilator throughout the study. The ventilatory was a volume-
controlled square-wave flow pattern, at a rate of 10 breaths/min..
Inspiratory time (TI) was 33% and end-inspiratory pause (EIP) was 10%
of the total respiratory cycle (TTOT). All measurements were performed
with zero applied end-expiratory pressure. Ventilatory variables were kept
constant during the study, both during TLV and OLV, in supine and in
lateral positions. This investigation was performed with closed chest,
before the surgical procedure.
Study protocol
After intubation in the supine position and fiberoptic control of the correct
DLT position, the two lungs were ventilated with the above-described
ventilatory pattern for 15 minutes. End-inspiratory and end-expiratory
occlusions were performed to determine the mechanical characteristics of
the respiratory system [peak inspiratory airway pressure (Ppeak), end-
153
inspiratory plateau pressure (Pplateau), and intrinsic positive end-expiratory
pressure (PEEPi)] and arterial and central venous blood gas samples were
drawn. Then, the tracheal or bronchial lumen of the DLT was clamped,
and the dependent or nondependent lung respectively, was allowed to
deflate to atmospheric pressure. After 15 minutes of OLV and data
collection, TLV was restored with unaltered ventilatory settings, and the
patient was turned to the left lateral position. TLV, turning and positioning
of the patients generally lasted approximately 15 minutes. At the end of
this period, ventilatory data were recorded, and blood gas samples were
drawn, and the studied lung (dependent or nondependent) was collapsed
again, now in the lateral position. After 15 minutes, respiratory
mechanics and gas exchange data were collected again.
154
6.3. Statistical analysis
Statistical analysis was done with the GraphPad InStat software package,
version 3.05 for Windows 95/NT (GraphPad Software Inc., San Diego
California USA, www.graphpad.com).
First, the assumption that the differences are sampled from a Gaussian
distribution was verified using the method of Kolmogorov and Smirnov.
Thereafter, demographic data and preoperative pulmonary functions of
the two groups were compared with the Student’s unpaired t-test with
Welch correction.
In each group, the data obtained in the two positions were compared with
the Student’s paired t-test.
Differences between the two groups in the same position were analyzed
with the Student’s unpaired t-test with Welch correction.
Values of p<0.05 were accepted as statistically significant. Data are
presented as mean ± SD.
155
6.4. Results
Patients
Demographic characteristics, preoperative pulmonary function tests,
isotopic investigations results and the values of the preoperative arterial
oxygen and carbon dioxide tensions of the 30 patients are shown in Table
6.1.. The differences observed between the two groups of patients were
statistically not significant.
Table 6.1. Demographic characteristics, preoperative pulmonary function results and preoperative
arterial blood gas values of the 30 patients.
Group D-OLV Group ND-OLV
Age(yr) 63.33 (12.08) 63.60 (11.72)
Height (cm) 168.13 (6.95) 168.40 (8.09)
Weight (kg) 74.06 (13.40) 76.20 (11.32)
FEV1 (% predicted) 69.60 (10.30) 67.20 (9.00)
TLC (% predicted) 127.10 (19.40) 122.40 (16.30)
FRC (% predicted) 149.43 (25.54) 152.80 (23.95)
RV (% predicted) 152.90 (43.63) 150.53 (32.87)
PaCO2 (room air, supine) 39.05 (3.80) 39.69 (2.96)
PaO2 (room air, supine) 79.33 (19.12) 76.80 (15.03)
Q left/right (%) 50 (46-54) 50 (46-54)
Data are expressed as Mean (SD)
Abbreviations: D-OLV: dependent one lung ventilation; ND-OLV: nondependent one lung
ventilation; FEV1: Forced expiratory volume in 1s; FRC: Functional residual capacity; RV: Residual
volume ; TLC: Total lung capacity; PaCO2: arterial carbon dioxide tension; PaO2: arterial oxygen
tension; Q: perfusion of the lung
156
TLV periods
During TLV periods, no significant differences in ventilatory mechanics
data or blood gases were found, whether comparing the two positions
inside the groups or comparing the two groups in the supine or lateral
positions (Table 6.2.).
Table 6.2. Respiratory mechanics and gas exchange data obtained during TLV in the supine and in
the lateral position observed in the two groups. FiO2 = 1.0 .
Group D-OLV (n=15) Group ND-OLV (n=15)
TLV Supine
TLV Lateral
TLV Supine
TLV Lateral
Ppeak (cmH2O) 19.29 (5.10) 19.41 (3.98) 19.44 (5.30) 18.16 (3.76)
Pplateau (cmH2O) 15.10 (4.11) 17.40 (4.43) 16.80 (5.23) 15.00 (2.75)
PEEPi (cmH2O) 1.86 (1.46) 1.17 (0.93) 1.81 (2.37) 1.41 (0.99)
PaCO2 (mmHg) 39.63 (2.11) 39.76 (2.30) 40.14 (2.90) 38.72 (1.96)
PaO2 (mmHg) 314.75 (66.45) 315.37 (67.59) 303.43 (78.41) 302.87 (78.61)
P(A-a)O2 (mmHg) 130.87 (49.67) 134.38 (55.49) 143.18 (33.11) 164.77 (54.55)
SsvcO2 (%) 80.11 (5.83) 84 .17 (5.70) 79.43 (5.50) 78.67 (8.54)
Data are expressed as Mean (SD)
Abbreviations: D-OLV: dependent one lung ventilation; N-DOLV: nondependent one lung
ventilation; TLV: Two-lung ventilation; PaCO2: Arterial carbon dioxide tension; Pa O2: Arterial
oxygen tension; P(A-a) O2: Alveolar-arterial oxygen tension difference; PEEPi: intrinsic positive
end-expiratory pressure; Ppeak: Peak inspiratory airway pressure; Pplateau: End-inspiratory airway
pressure; Ssvc O2: Oxygen saturation of blood collected in the superior cava vein.
157
TLV vs. OLV
Initiating OLV, without altering the ventilatory settings used during TLV,
increased Ppeak and Pplateau significantly (p<0.05; Student’s paired t-test),
either in supine or in lateral positions, in both groups of patients as
compared to the periods of TLV (Table 6.2. and Table 6.3.). In these
conditions, significant arterial oxygen tension (PaO2) decreases were
registered in both groups of patients in the supine [from 314.75 (66.45)
to 215.68 (58.34) mmHg (p<0.01) in the D-OLV group, and from 303.43
(78.41) to 224.56 (53.50) mmHg (p<0.01) in the ND-OLV group
(Student’s paired t-test)] and in the lateral position also: from 315.37
(67.59) to 274.18 (77.66) mmHg (p<0.01; Student’s paired t-test) in the
D-OLV group, and from 302.87 (78.61) to 196.81 (82.03) mmHg in the
ND-OLV group (p<0.01; Student’s paired t-test). The alveolar-arterial
oxygen tension difference [P(A-a)O2] increased (p<0.01; Student’s paired
t-test) when switching from TLV to OLV (Table 6.2. and 6.3.). Arterial
carbon dioxide tension (PaCO2) and oxygen saturation of blood collected in
the superior cava vein (SsvcO2) did not change (Table 6.2. and 6.3.).
OLV periods
In the D-OLV group, the PaO2 values observed in the lateral position were
significantly higher than those noted in the supine position: 274.18
158
(77.66) vs. 215.68 (58.34) mmHg (p<0.01; Student’s paired t-test)
(Table 6.3.). At the contrary, in the ND-OLV Group, the PaO2s in the
lateral position were lower as compared to those in the supine position:
196.81 (82.03) vs. 224.56 (53.50) mmHg (p< 0.05; Student’s paired t-
test) (Table 6.3.). PaCO2 and SsvcO2 did not change significantly when
the two positions where compared (Table 6.3.).
159
Table 6.3. Respiratory mechanics and gas exchange data obtained during OLV in the supine and in
the lateral position observed in the two groups. FiO2 = 1.0 .
Group D-OLV (n=15) Group ND-OLV (n=15)
OLV Supine OLV Lateral OLV Supine OLV Lateral
Ppeak (cmH2O) 30.57 (5.76) § 29.66 (5.70) § 32.10 (4.26) § 33.13 (5.38) §
Pplateau (cmH2O) 20.64 (3.29) § 20.09 (3.98) § 19.85 (2.56) § 20.25 (3.38) §
PEEPi (cmH2O) 2.71 (2.49) § 1.91 (0.99) 3.00 (2.14) § 2.80 (0.44)
PaCO2 (mmHg) 39.03 (2.92) 38.72 (2.76) 38.01 (3.50) 41.18 (6.97)
PaO2 (mmHg) 215.68 (58.34)
§§
274.18 (77.66)
**
224.56 (53.50)
§§
196.81 (82.03) §§
* * ##
P(A-a)O2 (mmHg) 386.29 (101.89)
§§
207.20 (76.18)
**
353.12 (107.27)
§§
431.72
(141.51)##++
SsvcO2 78.05 (6.64) 76.20 (7.70) 77.68 (3.83) 76.28 (5.28)
Data are expressed as Mean (SD)
Abbreviations: D-OLV: dependent one lung ventilation; N-DOLV: nondependent one lung
ventilation; OLV: One-lung ventilation; Ppeak: Peak inspiratory airway pressure; Pplateau: End-
inspiratory airway pressure; PEEPi: intrinsic positive end-expiratory pressure; PaCO2: Arterial
carbon dioxide tension; PaO2: Arterial oxygen tension; P(A-a)O2: Alveolar-arterial oxygen tension
difference; SsvcO2: Oxygen saturation of blood collected in the superior cava vein
§, §§ p<0.05, p<0.01 intragroup TLV vs OLV comparisons (Student’s paired t-test) ++ p<0.01 (Group ND-OLV, lateral vs. supine) (Student’s paired t-test)
** p<0.01 (Group D-OLV, lateral vs. supine) (Student’s paired t-test) ## p<0.01 (Group D-OLV vs. Group ND-OLV, lateral) (Student’s unpaired t-test with Welch
correction)
160
Dependent lung vs. nondependent lung ventilation
Initiating OLV in the supine or in the lateral position increased Ppeak and
Pplateau, in both groups of patients (Table 6.2. and 6.3.). When comparing
the Ppeak and Pplateau values of the two groups, during OLV in supine or
lateral position, no differences were found. For the PaO2 values recorded,
no differences were observed between the D-OLV and N-DOLV patients
placed in the supine position (Table 6.3. and Figure 6.1.). However, in the
lateral position, the PaO2 was significantly higher in the D-OLV group as
compared to the N-DOLV group – 274.18 (72.18) vs. 196.81 (82.03)
mmHg (p<0.01; Student’s unpaired t-test with Welch correction) (Table
6.3. and Figure 6.1.). The SsvcO2 and the PaCO2 did not change
significantly when comparing the two groups (Table 6.3.).
161
Figure 6.1. Individual values of the PaO2 in the two groups of patients, during OLV, in the supine
and lateral positions
Abbreviations: OLV: One-lung ventilation;D-OLV: Dependent lung ventilation; ND-OLV:
Nondependent lung ventilation; PaO2: Arterial oxygen tension
162
6.5. Discussion
In this study, arterial oxygenation during dependent and nondependent
OLV was investigated in the supine and in the lateral positions in patients
with mild degrees of pulmonary hyperinflation scheduled for thoracic
surgery, in order to try to separate the effects of gravity on the
redistribution of pulmonary blood flow. The main findings were a
significantly higher PaO2 and lower P(A-a)O2 when dependent-lung OLV
was performed in the lateral position, as compared to the supine position
or as compared to the ventilation of the nondependent lung in the lateral
position.
Recent developments in thoracic surgery have widened the scale of
surgical interventions and changed certain traditions, including the
patient’s positioning. Single lung transplantation (23) and resection of
unilateral emphysematous bullae (24) are performed in the lateral
decubitus position, while double-lung transplantation, lung volume
reduction surgery and minimally invasive coronary artery surgery are
performed in the supine position (1-5,24). In both conditions, to facilitate
the surgeon’s task, variable periods of OLV are required during these
procedures.
To explain the preservation of blood oxygenation in the presence of large
fraction of atelectatic lung, as during OLV, in experimental or clinical
conditions, many factors have been mentioned, including principally:
hypoxic pulmonary vasoconstriction (HPV) (6,25), gravity (8,9) or local
mechanical forces (8,26).
163
In a recent study (7) comparing oxygenation during OLV of the dependent
lung, it was found that in the lateral position the PaO2 was greater as
compared to the supine position, independently of the FiO2 used. As the
main protective mechanism during OLV, the hypoxic pulmonary
vasoconstriction was believed to be position independent, the findings of
the study were explained by the presence of an other factor which would
deviate blood flow from the non-ventilated to the ventilated lung; however
this factor was thought to be gravity, it was not clearly demonstrated and
isolated.
In the present study, the authors tried to imagine a method which could
separate the gravity from other factors (if the factor that facilitate blood
flow redistribution during OLV in the lateral position is really gravity !), by
comparing OLV of the dependent and nondependent lungs.
When OLV is undertaken and one of the lungs is excluded from
ventilation, an obligatory right to left shunt occurs through the non-
ventilated lung, that is not present during two-lung ventilation, and
arterial oxygenation decreases (6,8). The pulmonary vessels in that non-
ventilated, hypoxic area respond by increasing resistance to flow. This
reflex HPV of vessels perfusing hypoxic alveoli diverts blood flow from
non-ventilated lung units (8,25,27) to ventilated regions and attenuate
hypoxemia by actively reducing the perfusion of nonventilated lung tissue.
In the supine position, during anesthesia and controlled mechanical
ventilation of both lungs, generally there are no significant differences in
the perfusion between the two lungs as both are exposed to the same
164
level of gravity. For a given FiO2, OLV results in a larger P(A-a)O2 and a
low PaO2 as compared with TLV. This has been demonstrated by several
studies in which patients served as their own controls (6, 8, 27).
Initiating OLV will be source of right to left shunt and as a consequence,
PaO2 will decrease and P (A-a) O2 increase, as in the present study (Table
6.2. and 6.3.).
If we are considering the gravitational model, on average, 40% of blood
flow perfuses the nondependent lung and the remaining 60% the
dependent lung (27,28). The total shunt during TLV (up to 10% of the
cardiac output) is assumed to be distributed equally between the two
lungs (27,28); 5% for each lung. The 35% (40% - 5% shunt flow) of the
total blood flow perfusing the nondependent lung which is not ventilated,
is presumed to be reduced by half by an intact HPV (thus blood flow for
this lung will be 35/2 + 5% ≈ 22,5%). The situation will change if the
nonventilated lung is the lower lung. Theoretically, the blood flow in this
lung, with unaltered HPV will be 55/2 + 5% ≈ 32.5%. This result may
explain the findings of this study, where the PaO2 decreased in the lateral
position during OLV of the nondependent lung, in comparison with OLV of
the dependent lung, and this decrease is due primarily to the gravitational
redistribution of blood flow; if gravity is absent, the situation would be the
same as in the supine position.
Recently, the importance of the gravitational pulmonary blood flow
distribution has been challenged by new studies using single photon
emission computed tomography scanning high resolution computed
165
tomography or 99MTechnetium labeled macroaggregates detected by a
planar gamma counter. These techniques are reconstructing regional
blood flow by dissecting the lung into a large number of small cubic pieces
and determining the perfusion of each of the cubes using radioactive or
fluorescent microspheres (11-20). These studies suggest that while
pulmonary blood flow tends to follow a vertical distribution, the variability
of blood flow within each isogravitational plane is greater than the change
with gravity, thus they suggest that pulmonary blood flow is not primarily
distributed on gravitational basis, and the anatomic properties of
pulmonary vasculature (a centro-peripheral distribution, in isogravitational
planes) are more important than gravity in blood flow distribution (11-20).
However, all the above mentioned studies were done on animals and not
in humans; moreover, the patients included in our study were COPD
patients with pulmonary hyperinflation and not patients with normal lung
function, and in these patients there is a remodeling of the pulmonary
vasculature and a destruction of the pulmonary parenchymal tissue. The
optimal situation would be to investigate first in these patients the precise
distribution of blood flow in the lungs.
Another difference between our study and the other experimental settings
is that in our study there is an induced acute hypoxic condition,
represented by OLV.
Unfortunately, the design of our study did not permit us to measure shunt
fraction; it was a pure clinical study on patients undergoing thoracic
surgery, and according to institutional structures, monitoring and
166
measurements were limited, with no justification for more invasive (and
thus more dangerous monitoring) in these patients.
No pulmonary artery catheter was inserted to sample mixed venous
blood; however, some studies demonstrated that monitoring of central
venous blood oxygen may be as useful as monitoring of mixed venous
blood oxygen (29,30). For the same reasons, we could not perform shunt
measurements.
There is always the possibility that the intensity of HPV is influenced by a
time effect. Benumof (31) showed that intermittent hypoxic challenges
may potentate the HPV response in an animal model. In our study design,
there were, indeed, two sequences of deflating/inflating of the lung,
however, the oxygenation was not better in neither of the groups at the
second period of OLV (Table 6.2. and Figure 6.1.). Our arguments are
similar to the findings of Carlsson et al (32), who obtained maximal HPV
within 15 minutes of hypoxia, while Domino et al. (33) obtained a
maximal response after the very first hypoxic challenge during OLV in
closed chest dogs.
Moreover, there is no possibility to reproduce the recent imaging
techniques studies in clinical conditions.
Due to anesthesia and muscle relaxation, the distribution of ventilation
will also change in the lateral position because the applied positive
pressure ventilation displaces the diaphragm preferentially at the
nondependent part. This may be in the favor of our findings.
167
The presence of chronic airflow obstruction may be associated with better
PaO2 during OLV possibly because of dynamic hyperinflation resulting in
an increased FRC and intrinsic PEEP in the ventilated lung (34). In severe
COPD patients, HPV may not be an important protective mechanism, as
these patients have already an increased pulmonary arterial pressure and
reduced pulmonary vascular bed. The amount of disease in the
nondependent lung is also a significant determinant of the amount of
blood flow to the nondependent lung. If the nondependent lung is severely
diseased, there may be already a fixed reduction in blood flow to this lung
preoperatively, and collapse of such a diseased lung may not cause more
increase in shunt; thus, we didn’t included in our study patients with
significant unilateral restrictive disorders, as shown by the preoperative
perfusion measurements of the lungs, given that the PaO2 during TLV may
be also a determinant factor of oxygenation during OLV (34).
The most important mechanism to reduce blood flow of an atelectatic lung
is HPV (6,8). When OLV is induced in the lateral position, the blood flow of
the nondependent lung is reduced already by gravitational forces and HPV
will induce further reduction of blood flow. In contrast, in the supine
position, both lungs are equally exposed to an identical gravitational force,
thus during OLV the reduction of blood flow will depend solely on the
strength of the HPV.
PaCO2 may influence the level HPV response (35), but in our study, the
PaCO2 values of the studied patients were always within normal limits in
168
the different periods of blood gas measurements, thus they could not
influence our results.
The side of surgery may also influence oxygenation during OLV. Left
thoracotomy has a better PaO2 during OLV than right thoracotomy
because the left lung normally receives 10 percent less cardiac output
than the right lung (34). In the present study, on patients scheduled for
elective right thoracotomies, there were no statistical differences
concerning the preoperative right/left lung perfusions; moreover, there
were only right thoracotomies, and even so oxygenation was better during
dependent OLV in comparison with nondependent OLV.
The method used to ventilate the dependent lung is an important
determinant of the blood flow distribution during OLV. High airway
pressures can compress lung vessels, diverting blood flow from ventilated
regions to nonventilated regions. On the other hand, hypoventilation of
the dependent lung during OLV is associated with lower airway pressure
and the ventilated lung pulmonary vascular resistance may decrease, thus
promoting HPV in the nonventilated lung (34). This mechanism of
improved PaO2 was unlikely in the present investigation, as ventilatory
settings were kept constant, inspiratory airway pressures were not
changed and the mechanical characteristics of the dependent and
nondependent lungs remained unaltered (Tables 6.2. and 6.3.) after
changing the position. These findings indicate that the improved PaO2 can
not be attributed to an altered HPV due to change in ventilatory pattern.
169
Surgical compression and retraction may also contribute to the passive
reduction of nondependent lung blood flow (8,26). However, in the
present investigation, there was no mechanical effect on lung
parenchyma, as the data were collected before chest opening, therefore
surgical manipulation could not influence the amount of shunt occurring.
During TLV in the lateral position, the position of the patient reduces
perfusion of the upper lung due to gravitational diversion of blood flow to
the dependent lung which is thought to be disadvantageous (9,10).
However, in agreement with the findings of Boldt et al. (36), and Rehder
et al. (37), in our patients, the change of position from supine to lateral
decubitus resulted in no significant changes in PaO2 during TLV in both
studied groups.
In conclusion, in accordance with the findings of our study, during OLV of
COPD patients with pulmonary hyperinflation, gravity remains beside HPV
a major factor in blood flow redistribution from the dependent towards the
nondependent lung.
170
6.6. References
1. Pasque MK, Cooper JD, Kaiser L, et al. Improved technique for
bilateral lung transplantation: rationale and initial clinical
experience. Ann Thorac Surg 1990;49:785-91.
2. Cooper JD, Patterson GA, Sundaresan RS, et al. Results of 150
consecutive bilateral lung volume reduction procedures in patients
with severe emphysema. J Thorac Cardiovasc Surg 1996; 112:1319-
30.
3. Lima O, Ramos L, DiBiasi P, Judice L. Median sternotomy for
bilateral resection of emphysematous bullae. J Thorac Cardiovasc
Surg 1981; 82:892-7.
4. Wasnick JD, Acuff T. Anesthesia and minimally invasive
thoracoscopically assisted coronary artery bypass: a brief clinical
report. J Cardiothorac Vasc Anaesth 1997;11: 552-5.
5. Urschel HC, Razzuk MA. Median sternotomy as a standard approach
for pulmonary resection. Ann Thor Surg 1986; 41: 130-4.
6. Benumof JL. One-lung ventilation and hypoxic pulmonary
vasoconstriction: implications for anesthetic management. Anesth
Analg 1985;64:821-33.
7. Bardoczky GI, Szegedi LL, d'Hollander AA, Moures JM, de Francquen
P, Yernault JC. Two-lung and one-lung ventilation in patients with
chronic obstructive pulmonary disease: the effects of position and
F(IO)2. Anesth Analg. 2000; 90 :35-41.
171
8. Benumof JL. Physiology of the lateral decubitus position, the open
chest and one-lung ventilation. In: Kaplan JA ed. Thoracic
Anesthesia. New York, Churchill Livingstone; 1993: 193-221.
9. Arborelius M, Lundin G, Svanberg L, Defares JG. Infulence of
unilateral hypoxia on blood flow through the lungs in man in lateral
position J Appl Physiol 1960; 15: 595-7.
10. Fiser WP. Friday CD, Read RC. Changes in arterial oxygenation and
pulmonary shunt during thoracotomy with endobronchial anesthesia.
J Thorac Cardiovasc Surg 1982; 83: 523-31.
11. Glenny RW, Polissar L, Robertson HT. Relative contribution of
gravity to pulmonary perfusion heterogeneity. J Appl Physiol 1991;
71: 2449-52
12. Glenny RW, Robertson HT. Fractal modeling of pulmonary blood flow
heterogeneity. J Appl Physiol 1991; 70:1024-30.
13. Glenny RW, Robertson HT. Fractal properties of pulmonary blood
flow: Characterization of spatial heterogeneity. J Appl Physiol 1990;
69:532-45.
14. Glenny RW. Spatial correlation of regional pulmonary perfusion. J
Appl Physiol 1992;72:2378-86.
15. Hakim TS, Dean GW, Lisbona R. Effect of body posture on spatial
distribution of pulmonary blood flow. J Appl Physiol 1988; 64: 1160-
70.
16. Hakim TS, Lisbona R, Dean GW. Gravity-independent inequamity in
pulmonary blood flow in humans. J Appl Physiol 1987; 63: 1114-21.
172
17. Mure M, Domino KB, Robertson T, Hlastala MP, Glenny RW.
Pulmonary blood flow does not redistribute in dogs with reposition
from supine to left lateral position. Anesthesiology 1998; 89: 483-
92.
18. Walther SM, Domino KB, Glenny RW, Polissar NL, Hlastala MP.
Piulmonary blood flow distribution has a hilar-to-peripheral gradient
in awake, prone sheep. J Appl Physiol 1997; 82: 678-85.
19. Chang H, Lai-Fook SJ, Domino KB, Schimmel C, Hildebrandt J,
Roberston HT, Glenny RW, Hlastala MP. Spatial distribution of
ventilation and perfusion in anesthetized dogs in the lateral
positures. J Appl Physiol 2002; 92: 745-62.
20. Glenny RW, Lamm WJE, Albert RK, Robertson HT. Gravity is a minor
determinant of pulmonary blood flow distribution. J Appl Physiol
71:620-9; 1991.
21. Siafakas NM, Vermeire P, Pride NB, Paoletti P, Gibson J, Howard P,
Yernault JC, Decramer M, Higenbottam T, Postma DS, et al. Optimal
assessment and management of chronic obstructive pulmonary
disease (COPD). The European Respiratory Society Task Force. Eur
Respir J. 1995;8:1398-420.
22. Hannalah MS, Benumof JL, McCarthy PO, et al. Comparison of
three techniques to inflate the bronchial cuff of left polyvinilchloride
double-lumen tubes. Anesth Analg 1993; 77: 990-4.
173
23. Conacher ID. Isolated lung transplantation: A review of problems
and guide to anesthesia. Br J Anaesth 1988; 61:468-74.
24. Gayes JM, Emery RW. The MIDCAB experience: A current look at
evolving surgical and anesthetic approaches J Cardiothorac Vasc
Anesth 1997;11: 625-8.
25. Marshall BE. Hypoxic pulmonary vasoconstriction. Acta Anaesthesiol
Scand 1990;34: S94: 37-41.
26. Aalto-Setala M, Heinonen J, Salorinne Y. Cardiorespiratory function
during thoracic anaesthesia: Comparison of two-lung ventilation and
one lung ventilation with and without PEEP. Acta Anaesthesiol Scand
19:287-95; 1975
27. Cohen E (ed) The practice of thoracic anesthesia. Philadelphia, PA,
JB Lippincott Company, Chapter 5, 1995.
28. Cohen E. Physiology of the lateral position and one-lung ventilation.
Chest Surgery Clinics of North America, Volume 7, Number 4, 1997.
29. Schou H, Perez de Sa V, Larsson A: Central and mixed venous
oxygen correlate well during acute normovolemic hemodilution in
anesthetized pigs. Acta Anaesthesiol Scand 1998; 42: 172-7.
30. Ladakis C, Myrianthefs P, Karabinis A et al.: Central venous and
mixed venous oxygen saturation in critically ill patients. Respiration
2001; 68: 279-85.
31. Benumof JL. Intermittent hypoxia increases lobar hypoxic pulmonary
vasoconstriction. Anesthesiology 1983; 58: 399-404.
174
32. Carlsson AJ, Bindslev L, Santesson J, Gottlieb I, Hedenstierna G.
Hypoxic pulmonary vasoconstristion in the human lung: the effect of
prolonged unilateral hypoxic challenge during anaesthesia. Acta
Anaesthesiol Scand 1985; 29: 346-51.
33. Domino K, Chen L, Alexander C, et al. Time course and responses of
sustained hypoxic pulmonary vasoconstriction in the dog.
Anesthesiology 1984; 60: 562-6.
34. Slinger P, Suissa S, Adam J, Triolet W. Predicting arterial
oxygenation during one-lung ventilation with continuous positive
airway pressure to the nonventilated lung. J Cardiothorac Anesth,
1990; 4: 436-40.
35. Benumof JL, Mathers JM, Wahrenbrock EA. Cyclic hypoxic pulmonary
vasoconstriction induced by concomitant carbon dioxide changes. J
Appl Physiol 1976; 41: 466-9.
36. Boldt J, Muller M, Uphus D, Padberg W, et al. Cardiorespiratory
changes in patients undergoing pulmonary resection using different
anesthetic management techniques. J Cardiothorac Vasc Anesth
1996; 7: 854-9.
37. Rehder K, Knopp TJ, Sessler AD, Didier EP. Ventilation-
perfusion relationship in young healthy awake and anesthetized-
paralyzed man. J Appl Physiol 1979; 47: 745-53.
175
6.7.Summary
This study compared arterial oxygenation in lateral and supine positions
during one-lung ventilation of the dependent (lower) vs. the
nondependent (upper) lung. Arterial oxygenation was better in the lateral
position as compared to the supine one; arterial oxygenation was better
when the dependent (lower) lung was ventilated as compared to the
ventilation of the nondependent (upper) lung; this finding demonstrates
that gravity remains a major factor that influences blood flow
redistribution during one-lung ventilation.
176
177
7. The effects of acute isovolemic hemodilution on
oxygenation during one-lung ventilation
Modified from: Szegedi LL, Van der Linden Ph, Ducart A, Cosaert P, Poelaert J, Vermassen F,
Mortier EP, d’Hollander AA. The effects of acute isovolemic hemodilution on oxygenation during
one-lung ventilation. Anesth Analg 2004; 99
This article is preceded by an editorial, written by Dr. Andre Lewin.
178
7.1. Introduction
One-lung ventilation (OLV) is a useful technique during thoracic surgery
(1). One lung is mechanically ventilated while the other is either occluded
or left open to the atmosphere. The challenge for the anesthesiologist
during OLV is to maintain an acceptable oxygenation. Indeed, blood flow
through the nonventilated lung becomes a right-to-left shunt in addition to
that which exists in the ventilated lung. OLV will therefore result in a lower
arterial oxygen partial pressure (PaO2) than during two-lung ventilation
(TLV). Fortunately several mechanisms that tend to decrease the percent
of cardiac output perfusing the nonventilated lung. These mechanisms are
either passive (e.g., mechanical-like gravity, surgical manipulation,
amount of pre-existing lung disease) or active (e.g., hypoxic pulmonary
vasoconstriction (HPV)) (1).
Increasing awareness of potential severe complications associated with
blood transfusion has prompted the development of blood conservation
strategies. Among these strategies, acute isovolemic hemodilution (IH), is
a safe procedure (2). The IH technique lowers a patient's hemoglobin at
the start of the surgery, so fresh units of the patients' blood are available
when needed (2).
Few studies have investigated the effects of hemodilution on pulmonary
gas exchange, in particularly during OLV. Acute hemodilution may
improve gas exchange efficiency of the normal lung (3,4), as evidenced by
higher PaO2 and lower alveolar-arterial oxygen partial pressure difference,
but this effect is not consistently observed. Experimentally, the
179
relationship between anemia and PaO2 is not well defined. Deem et al (4)
analyzed the studies over the relationship between anemia and PaO2.
These studies have examined the effects of hemodilution or blood
transfusion in the absence of known lung disease, suggesting either an
inverse relationship between PaO2 and Hct, or a direct relationship, or an
inconsistent effect. Moreover, only a few studies included concurrent
controls for the effects of time and/or anesthesia (4).
The purpose of this study was to assess the effect of IH on arterial
oxygenation during OLV. Patients scheduled for thoracic surgery may have
a compromised preoperative lung function. Therefore, we evaluated the
effect of IH on arterial oxygenation in patients with normal preoperative
lung function, and in patients with stable chronic obstructive pulmonary
disease (COPD), with moderate degrees of pulmonary hyperinflation. A
third group of patients with COPD, who did not undergo IH was added to
serve as control for the effects of time and/or anesthesia.
180
7.2. Materials and methods
Our IRB approved the investigation and informed consent was obtained
from 47 adult ASA class II-III patients with a hemoglobin level >140 g/l.
The patients were scheduled for tumor surgery through right thoracotomy
and requiring one-lung ventilation.
Preoperatively, the pulmonary function was tested in the sitting position,
including spirometry and static lung volumes determined by
plethysmography (MasterScreen Body™, Jaeger and Toennies™, Erich
Jaeger GmbH, Hoechberg, Germany).
Based upon their preoperative pulmonary function, patients were included
either in a group with normal lung function (Group NL; n=17) or a group
with stable COPD and mild to moderate pulmonary hyperinflation
(functional residual capacity (FRC) above 120% of the predicted value)
(Group COPD; n=17). After preliminary results on 12 patients in each
group, a third group of patients with stable COPD was added (Group
CTRL; n=13). The assignment of the patients either to the Group COPD or
the Group CTRL was randomized. The rationale for including a concurrent
control group (Group CTRL) was to eliminate the effects of time and/or
anesthesia that could influence our results. Hence, these patients
underwent the same sequences of OLV, but no IH.
Patients with cardiac or renal function impairment were not included in the
study.
181
Anesthesia was conducted in a standardized manner. All patients were
given alprazolam 0.5 mg p.o., approximately 60 to 90 minutes before
arrival in the operating theatre.
A thoracic epidural catheter was inserted at the mid-thoracic level (Th6-
Th9) to assure analgesia during and after surgery. A test dose of 3 ml 2%
lidocaine with epinephrine 1/200000 was used; the initial dose (8 ml 2%
lidocain with 100 µg fentanyl) was given only after the end of the study
followed by a continuous infusion of bupivacaine 0.5%, at a rate of 2-5
ml/h.
Induction and maintenance of general anesthesia was achieved with
fentanyl (100 mcg) and propofol (initial dose 2 mg/kg, followed by
continuous infusion of 3-5 mg.kg-1.h-1). Cisatracurium (initial dose of 0.15
mg/kg and top up doses of 0.05 mg/kg) was used to allow tracheal
intubation and to maintain neuromuscular blockade throughout surgery.
The neuromuscular blockade was assessed by regular measurements of
post-tetanic count during the procedure.
A three-lead electrocardiogram, invasive radial arterial and central venous
pressures and arterial oxygen saturation were continuously monitored.
Expired end-tidal carbon dioxide tension (ETCO2), flow-volume and
pressure-volume loops were also continuously displayed (S/5 Anesthesia
Monitor AM, Datex-Ohmeda Division, Instrumentarium Corp., Datex-
Ohmeda, Finland).
In all patients, the bronchus of the dependent left lung was intubated with
a double-lumen endotracheal tube (DLT) (Broncho-cath™, Mallinckrodt
182
Laboratories, Athlone, Ireland) of an appropriate size determined by the
height of the patient (5). The correct position of the DLT was ascertained
with fiberoptic bronchoscopy.
A constant tidal volume of 10 ml/kg was delivered with an S/5 ADU
ventilator throughout the study. The ventilatory pattern consisted of
volume-controlled, square-wave flow pattern, at a rate of 10
breaths/min., with a fraction of inspired oxygen (FiO2) in air of 0.5.
Inspiratory time : expiratory time (I:E) was 1:2 and end-inspiratory pause
(EIP) was 10% of the total respiratory cycle. End-expiratory pressure was
set to zero. Ventilatory variables were kept constant during the study,
both during TLV and OLV.
A transesophageal echocardiograph probe was inserted after tracheal
intubation to measure the cardiac output with the method of the effective
aortic valve area (6). Blood gas samples were analyzed immediately after
they were drawn, and temperature corrected. Arterial and central venous
oxygen saturation were measured with a co-oximeter.
This investigation was performed with closed chest, before the surgical
procedure, with the patients in supine position. After intubation in the
supine position and fiberoptic bronchoscopic control of the correct DLT
position, the tracheal lumen of the DLT was clamped, and the
nonventilated right lung was allowed to deflate to atmospheric pressure.
After 15 minutes of OLV with the above-described ventilatory settings,
end-inspiratory and end-expiratory occlusions were performed to
determine the mechanical characteristics of the respiratory system [peak
183
inspiratory airway pressure (Ppeak), end-inspiratory plateau pressure
(Pplateau), and intrinsic positive end-expiratory pressure (PEEPi)] and
arterial and venous blood gas samples were drawn and analyzed. The
cardiac output was measured.
After these baseline measurements (before hemodilution), the lung was
sighed manually (insufflation pressure up to 40 cm H2O (7)), and TLV was
restored with unaltered ventilatory settings. An IH was done during the
period of TLV by simultaneous withdrawal of blood and infusion of 6%
hydroxyethyl starch 130/0.4 at equal volumes. The exchange was
standardized to 500 ml for each patient.
Thereafter, OLV was restored for a period of 15 minutes. At the end of this
period, ventilatory data were again recorded, arterial and venous gas
samples drawn and analyzed, and cardiac output measured (after IH).
In Group CTRL the sequences of OLV were the same, but no IH was done.
The temperature of the patients was kept constant during the whole study
by using an air convection system.
184
7.3. Statistical Analysis
Statistical analysis was performed with the GraphPad InStat software
package, version 3.05 for Windows 95/NT.
First, the assumption that data are sampled from populations with
identical standard deviations (SD) was tested using the method of
Bartlett. The assumption that the differences are sampled from
populations that follow Gaussian distribution was verified using the
method of Kolmogorov and Smirnov. Thereafter, differences between the
groups and inside each group were analyzed with the One-way Analysis of
Variance (ANOVA) with the Tukey-Kramer Multiple Comparisons Test.
Values of p<0.05 were accepted as statistically significant. Data are
presented as mean ± SD.
185
7.4. Results
There was a slight, but statistically significant, difference when comparing
the weight of the patients from the Group NL either to the Group COPD or
to the Group CTRL, but no differences in body surface area (Table 7.1.).
Table 7.1..Demographic data, preoperative pulmonary function tests and preoperative hemoglobin
levels and hematocrit of the patients.
Group NL
(n=17)
Group COPD
(n=17)
Group CTRL
(n=13)
Age (years) 57 ± 12 63 ± 9 64 ± 9
Height (cm) 175 ± 5 171 ± 7 173 ± 11
Weight (kg) 79 ± 16* 70 ± 14 68 ± 18
BSA (m²) 1.9 ± 0.20 1.9 ± 0.21 1.8 ± 0.17
FEV1(% predicted) 87 ± 8 * 78 ± 11.7 76 ± 7.2
FRC (% predicted) 108 ± 7.8 ** 149 ± 15.9 143 ± 18.8
RV (% predicted) 114 ± 13** 160 ± 26 156 ± 19
Hb preoperative (g/dl) 15 ± 0.8 15 ± 0.6 15 ± 0.3
Hct preoperative (%) 44 ± 2.2 45 ± 3.3 44 ± 1.3
Data are mean ± SD
*p<0.05, **p<0.01, as compared to Group COPD and Group CTRL
One-way Analysis of Variance (ANOVA) with Tukey-Kramer Multiple Comparisons Test.
Abbreviations: Group NL: Group of patients with normal lung function; Group COPD: Group of
patients with chronic obstructive pulmonary disease; Group CTRL: Control group, patients with
chronic obstructive pulmonary disease, who had the same sequences of one-lung ventilation, but
who did not undergo hemodilution; BSA: Body surface area; FEV1: Forced expiratory volume in
one second; FRC: Functional residual capacity; RV: Residual volume; Hb: Hemoglobin level; Hct:
Hematocrit
186
Forced expired volumes in one second (FEV1) were significantly lower,
FRC and residual volumes (RV), were significantly higher in the Group
COPD and in the Group CTRL as compared to the Group NL (Table 7.1.).
Inspiratory airway pressures (Ppeak and Pplateau) were comparable in all
groups before and after IH. There was a lower PEEPi in Group NL as
compared either to Group NL or CTRL (p<0.05, ANOVA with Tukey-
Kramer Multiple Comparisons Test). IH did not significantly affect these
parameters (Table 7.2.).
Table 7.2. .Inspiratory airway pressures and intrinsic positive end-expiratory pressure values in
the three groups during OLV
Group NL (n=17) Group COPD (n=17) Group CTRL (n=13)
Before IH After IH Before IH After IH OLV 1 OLV 2
Ppeak
(cmH2O)
33 ± 6 35 ± 5.3 33 ± 6.4 33. ± 7.3 34 ± 8.3 33 ± 7.1
Pplateau
(cmH2O)
26 ± 4.4 26 ± 5.1 26 ± 6.8 26 ± 6.3 26 ± 8.9 26 ± 7.4
PEEPi
(cmH2O)
2.4 ± 1.5 * 2.2 ± 1.6 # 3.2 ± 0.8 3.5 ± 1.5 3.5 ± 1.2 3.7 ± 1.2
Data are mean ± SD
*p<0.05, as compared to the values in Group COPD and Group CTRL at all periods of time studied
# p<0.05, as compared to the values in Group COPD and Group CTRL at all periods of time studied
One-way Analysis of Variance (ANOVA) with the Tukey-Kramer Multiple Comparisons Test.
Abbreviations: Group NL: Group of patients with normal lung function; Group COPD: Group of
patients with chronic obstructive pulmonary disease; Group CTRL: Control group, patients with
chronic obstructive pulmonary disease, who had the same sequences of one-lung ventilation, but
who did not undergo hemodilution; IH: Isovolemic hemodilution; Ppeak: peak inspiratory airway
pressure; Pplateau: Plateau (end-inspiratory) airway pressure; PEEPi: Intrinsic positive end-
expiratory pressure
187
IH resulted in a significant and similar decrease in hemoglobin
concentration (or hematocrit) in the Group COPD and the Group NL (Table
7.3.). No hemodilution was done in the Group CTRL and thus no
hemoglobin concentration (or hematocrit) changes were found. Cardiac
output, mean arterial pressure, and central venous pressure remained
stable during the exchange procedure (in the Group NL and the group
COPD) and in Group CTRL (no IH).
Arterial partial pressure of carbon dioxide (PaCO2) remained constant and
similar in all the three groups with or without IH. (Table 7.3.).
In Group NL, IH was not associated with a significant change in PaO2. In
Group COPD and CTRL, baseline PaO2 was slightly, but statistically
significantly lower than in Group NL. IH resulted in a significant decrease
in PaO2 in the Group COPD (p<0.01, ANOVA with Tukey-Kramer Multiple
Comparisons Test); no significant changes were found in the Group CTRL
(no IH was performed in this group) at the time periods studied (Table
7.3. and Figure 7.1.).
188
Table 7.3.Hemoglobin, hematocrit, cardiac output, and blood gas values in the three groups
during OLV.
Group NL (n=17) Group COPD (n=17) Group CTRL (n=13)
Before IH After IH Before IH After IH OLV 1 OLV 2
Hb (g/l) 15 ± 0.8 11.6 ± 0.8
## , ‡‡
15 ± 0.6 11 ± 1.2
## , ‡‡
15 ± 0.3 15 ± 0.6
Hct (%) 44± 2.2 36 ± 2.3
## , ‡‡
45 ± 3.3 36 ± 4 ## ,
‡‡
44 ± 1.3 45 ± 2.6
MAP
(mmHg)
81 ± 14 76 ± 16 82 ± 18 83 ± 16 78 ± 16 78 ± 12
CVP
(mmHg)
4.1 ± 2.5 4.2 ± 2.2 3.7 ± 3 3.9 ± 3.2 3.6 ± 3 3.5 ±3.4
CO
(l/min)
3.7 ± 0.8 3.7 ± 0.7 3.6 ± 0.6 3.7 ± 0.5 3.8 ± 0.4 3.8 ± 0.4
CI(l/min/
m²)
2 ± 0.4 2 ± 0.3 1.9 ± 0.4 1.9 ± 0.4 2 ± 0.5 2 ± 0.3
SvcO2 (%) 79 ± 13 79 ± 13 79 ± 7 77 ± 7 79 ± 12 77 ± 14
SaO2 (%) 99 ± 2 98 ± 2 97 ± 2 95 ± 2 98 ± 2 98 ± 2
PaCO2
(mmHg)
41 ± 5.4 40 ± 5.8 41 ± 5.4 40 ± 4.3 38 ± 7.2 39 ± 4.4
PaO2
(mmHg)
140 ± 26* 138 ± 26* 119 ± 21 86 ± 16
##,‡‡,††
120 ± 20 118 ± 23
Data are mean ± SD
*p<0.05, as compared to the values in Group COPD and Group CTRL at all periods of time studied
##p<0.01, as compared to the values before IH in the same group
‡‡p<0.01, as compared to the values in Group CTRL at OLV1 and OLV2
††p<0.01, as compared to the values in Group NL before and after IH
One-way Analysis of Variance (ANOVA) with the Tukey-Kramer Multiple Comparisons Test.
189
Abbreviations: Group NL: Group of patients with normal lung function; Group COPD: Group of
patients with chronic obstructive pulmonary disease; Group CTRL: Control group, patients with
chronic obstructive pulmonary disease, who had the same sequences of one-lung ventilation, but
who did not undergo hemodilution; IH: Isovolemic hemodilution; OLV 1: First period of one-lung
ventilation in the Group CTRL; OLV 2: Second period of one-lung ventilation in the Group CTRL;
Hb: Hemoglobin level; Hct: Hematocrit; MAP: Mean arterial pressure; CVP: Central venous
pressure; CO: Cardiac output; CI: Cardiac index; ScvO2: Central venous blood saturation in
oxygen; SaO2: Arterial blood saturation in oxygen; PaCO2: Partial pressure of carbon dioxide in
arterial blood; PaO2: Partial pressure of oxygen in arterial blood
190
Figure 7.1. Individual values of the partial pressure of oxygen in arterial blood (PaO2) in the three
groups of patients, at the two time periods studied. The PaO2 decreased significantly after
isovolemic hemodilution in Group COPD (n=17), but no decreases were found neither in Group NL
(n=17), nor in Group CTRL (n=13) (One-way Analysis of Variance with the Tukey-Kramer Multiple
Comparisons Test).
Abbreviations: Group NL: Group of patients with normal lung function; Group COPD: Group of
patients with chronic obstructive pulmonary disease; Group CTRL: Control group, patients with
chronic obstructive pulmonary disease, who had the same sequences of one-lung ventilation, but
who did not undergo hemodilution; PaO2: Partial pressure of oxygen in arterial blood; IH:
Isovolemic hemodilution; OLV 1: First period of one-lung ventilation in the Group CTRL; OLV 2:
Second period of one-lung ventilation in the Group CTRL
0
20
40
60
80
100
120
140
160
180
200
Before IH After IH Before IH After IH O LV 1 OL V 2
Group NL Group CO PD Group CTRL
Pa
O2 (
mm
Hg
)
**NS NS
**p<0.01ANOVA and Tukey post test
191
7.5. Discussion
This is the first study evaluating the effects of mild IH on arterial
oxygenation during OLV in humans. In patients with normal preoperative
lung function, mild IH did not influence arterial oxygenation, while in
patients with mild to moderate degrees of preoperative pulmonary
hyperinflation and stable COPD, PaO2 decreased significantly. In both
groups, IH did not influence pulmonary mechanics during OLV. All the
studied parameters remained stable over time in Group CTRL.
HPV is the major protective mechanism that diverts blood flow away from
the atelectatic lung (1). Although HPV is an intrinsic mechanism of the
pulmonary vasculature (8), several studies indicate that it may be
influenced by red blood cells. McMurtry et al (9) reported that isolated rat
lungs perfused with plasma had rapidly decaying HPV when compared
with lungs perfused with blood. Another study reports similar results in
isolated rat, cat or rabbit lungs (10).
Deem et al (11) studied the effect of anemia on intrapulmonary shunt
during left lung atelectasis in rabbits and concluded that isovolemic
anemia has a deleterious effect on pulmonary gas exchange, possibly
through attenuation of HPV. Another study (12) suggests that, in rabbits
with previous lung injury induced by gas embolism, IH results in improved
oxygen exchange, without finding a clear mechanism for this
improvement. The same authors report an improvement in gas exchange
in the normal rabbit lung as a result of an improvement in overall
ventilation/perfusion matching (13). Kleen et al (14) suggest that severe
192
acute normovolemic hemodilution in dogs, causes different alterations in
the heterogeneity of regional pulmonary blood flow in hyperoxic
conditions. Other authors (15) found no influence of changes in Hb
concentration on HPV efficiency.
In the present study, pre-existing lung disease, as demonstrated by
altered preoperative pulmonary function tests, was associated with a
higher baseline PEEPi, and a slightly lower PaO2 (in Group COPD and
Group CTRL).
The most important finding of the present study was a significant
decrease in PaO2 after IH during OLV of the patients with compromised
lung function. COPD patients are chronically hypoxic, and thus in such
patients HPV may be already maximal, and no more protective during
OLV, as these patients already have an increased pulmonary vascular
resistance and decreased pulmonary vascular bed (16). Other
mechanisms that tend to decrease the degree of venous admixture during
OLV include gravitational effect and surgical manipulation (1). In the
present. study, performed with closed chest, and with patients in the
supine position, these potential factors were absent.
The maintenance of tissue oxygen delivery during an acute reduction in
red blood cell concentration depends on both an increase in CO and an
increase in blood oxygen extraction (17). However, in our study, CO
remained constant with the mild IH (Table 7.2.). Anesthesia significantly
reduces the cardiac output response associated with acute IH. This could
193
be related to the effects of the anesthetics on the autonomic and the
cardiovascular systems (18).
The higher PEEPi values (Table 7.2.) during OLV, observed in the Group
COPD as compared to the Group NL, could have induced enough
hyperinflation to disrupt the ventilation/perfusion matching. However, this
observation is unlikely, given that in the Group CTRL (COPD patients who
did not undergo IH) the PEEPi values were comparable to those measured
in the Group COPD, and the PaO2 changes in the Group CTRL were not
significant.
If the PaO2 was lower in the Group COPD after IH than in the Group CTRL
(without IH), and there were no differences in the oxygen saturation of
blood collected in the superior cava vein or in cardiac index, then the
shunt must have increased in the Group COPD; even though this is
difficult to explain, this may be the only explanation to the findings of the
present study.
Some consideration should be given to the limitations in our study. First,
this study was a pure clinical study; according to institutional structures,
monitoring and measurements were limited. For example, pulmonary
artery catheter could not be inserted, so we were not able to perform
shunt measurements. Second, given the absence of data on the subject
and the patient population studied (stable COPD with mild to moderate
degrees of pulmonary hyperinflation), only mild degree of IH was studied.
Further studies are required to evaluate the effect of more profound IH in
patients with normal preoperative lung function. It is surprising that this
194
mild hemodilution/anemia was associated with this degree of gas
exchange impairment in the patients with COPD, given that changes in
intrapulmonary shunt in the experimental model reported by Deem et al
(11) were not realized until the hematocrit was in the mid-twenties. The
different results in the experimental model and this clinical study could be
for any number of reasons, including the species difference and the
presence of COPD, but it raises the question as to whether the observed
decrease in PaO2 was in fact due to hemodilution. Although somewhat
unlikely (18), the study protocol leaves the possibility that the observed
change in arterial oxygenation occurred already following IH during TLV.
Third, our results could have been influenced by the repeated
endotracheal tube clamping maneuvers. Benumof (19) showed that
intermittent hypoxic challenges may potentate the HPV response in an
animal model. In our study design, there were, indeed, two sequences of
deflating/inflating of the lung; however, the oxygenation was not better in
either of the groups at the second period of OLV. This is in accordance
with the studies of Carlsson et al (20), who found maximal HPV within 15
minutes of hypoxia, and Domino and co-workers (21) who observed a
maximal response after the very first hypoxic challenge during OLV in
closed-chest dogs. In order to eliminate the changes over time in
oxygenation during OLV, a third group of patients (Group CTRL) with
altered preoperative pulmonary function was studied. Sequences of OLV
done without IH did not result in significant difference in arterial
oxygenation (Table 7.3.).
195
Fourth, the study was done with the patients in the supine position and
with closed chest, instead of the usual clinical situation of lateral position
with open chest. These differences may affect extrapolation of the results,
because of the absence of the gravitational effect in COPD patients and
the absence of surgical manipulation, factors which may contribute to
blood flow redistribution during OLV (1).
In conclusion, mild IH significantly altered arterial oxygenation during OLV
in patients with preoperative compromised lung function. Although anemia
may be less tolerated by COPD patients in these conditions, this does not
indicate that a more aggressive transfusion approach is required in these
patients, indeed PaO2 could be easily raised by increasing the FiO2, which
is common practice during OLV.
Acute IH did not appear to be a safe alternative technique to blood
transfusion in patients with altered pulmonary function undergoing
thoracic surgery. Further studies are required to better define the
adequate transfusion trigger in this particular population.
196
7.6. References
1. Benumof JL. Special respiratory physiology of the lateral decubitus
position, the open chest and one-lung ventilation. In: Benumof JL, ed.
Anesthesia for thoracic surgery. Philadelphia: WB Saunders, 1995,
123-51.
2. Goodnough LT, Brecher ME, Kanter MH, AuBuchon JP. Transfusion
medicine. First of two parts - blood transfusion. N Engl J Med 1999;
340: 438-47.
3. Deem S, Alberts MK, Bishop MJ, et al. CO2 transport in normovolemic
anemia: complete compensation and stability of blood CO2 tensions. J
Appl Physiol 1997; 83: 1240-6.
4. Deem S, Swenson ER, Alberts MK, et al. Red-blood-cell augmentation
of hypoxic pulmonary vasoconstriction. Hematocrit dependence and the
importance of nitric oxide. Am J Resp Crit Care Med 1998; 157: 1181-
6.
5. Hannallah MS, Benumof JL, McCarthy PO, et al. Comparison of three
techniques to inflate the bronchial cuff of left polyvinyl chloride double-
lumen tubes. Anesth Analg 1993; 77: 990-4.
6. Poelaert J, Schmidt C, Van Aken H, et al. A comparison of
transoesophageal echocardiographic Doppler across the aortic valve
and thermodilution technique for estimating cardiac output.
Anaesthesia 1999; 54: 128-36.
197
7. Rothen HU, Sporre B, Engberg G, et al. Re-expansion of atelectasis
during general anaesthesia: a computed tomography study. Br J
Anaesth 1993; 71: 788-95.
8. Voelkel NF. Mechanism of hypoxic pulmonary vasoconstriction. Am Rev
Respir Dis 1986; 133: 1186-95.
9. Mc Murtry IF, Hookway BW, Roos SD. Red blood cells play a crucial role
in maintaining vascular reactivity to hypoxia in isolated rat lungs. Chest
1977; 71: 253-6.
10. Hakim TS, Malik AB. Hypoxic vasoconstriction in blood and plasma
perfused lungs. Respir Physiol 1988; 72: 109-21.
11. Deem S, Bishop MJ, Alberts MK. Effect of anemia on intrapulmonary
shunt during atelectatis in rabbits. J Appl Physiol 1995; 79: 1951-7.
12. Deem S, McKinney S, Polissar NL, et al. Hemodilution during venous
gas embolisation improves gas exchange, without altering V(A)/Q or
pulmonary blood flow distributions. Anesthesiology 1999; 91: 1861-72.
13. Deem S, Hedges RG, McKinney S et al. Mechanism of improvement in
pulmonary gas exchange during isovolemic hemodilution. J Appl
Physiol 1999; 87:132-41.
14. Kleen M, Habler O, Hutter J et al. Hemodilution and hyperoxia locally
change distribution of regional pulmonary perfusion in dogs. Am J
Physiol 274 (Heart Circ Physiol 43) 1995; H520-8.
15. Brimioulle S, Lejeune P, Naeije R. Effects of hypoxic pulmonary
vasoconstriction on pulmonary gas exchange. J Appl Physiol 1996; 81:
1535-43.
198
16. Slinger P. New trends in anaesthesia for thoracic surgery including
thoracoscopy. Can J Anaesth 1995; 42: R77-84.
17. Van der Linden P. Anemic hypoxia. In: Sibbald WJ, Messmer K, Fink
MP, eds. Tissue oxygenation in acute medicine. Berlin, Heidelberg:
Springer Verlag, 1998:116-27.
18. Ickx BE, Rigolet M, Van Der Linden PJ. Cardiovascular and metabolic
response to acute normovolemic anemia. Effects of anesthesia.
Anesthesiology 2000; 93:1011-6.
19. Benumof JL. Intermittent hypoxia increases lobar hypoxic pulmonary
vasoconstriction. Anesthesiology 1983; 58:399-404.
20. Carlsson AJ, Bindslev L, Santesson J, et al. Hypoxic pulmonary
vasoconstriction in the human lung: the effect of prolonged unilateral
hypoxic challenge during anaesthesia. Acta Anaesthesiol Scand 1985;
29:346-51.
21. Domino K, Chen L, Alexander C, et al. Time course and responses of
sustained hypoxic pulmonary vasoconstriction in the dog.
Anesthesiology 1984; 60:562-6.
199
7.7. Summary
The effects of isovolemic hemodilution on oxygenation during one-lung
ventilation were compared. It was found that oxygenation didn’t change
after isovolemic hemodilution in patients with normal lung function;
however, it decreased in patients with altered lung function. In patients
with pulmonary hyperinflation and chronic obstructive pulmonary disease,
care should be taken when bleeding occurs during one-lung ventilation,
because these patients don’t tolerate acute blood loss.
200
201
8. Intrinsic positive end-expiratory pressure
during one-lung ventilation of patients with
pulmonary hyperinflation. Influence of low
respiratory rate with unchanged minute
volume.
Modified from: Szegedi LL, Barvais L, Sokolow Y, Yernault JC, d’Hollander AA. Intrinsic positive
end-expiratory pressure during one-lung ventilation of patients with pulmonary hyperinflation.
Influence of low respiratory rate with unchanged minute volume.Br J Anaesth 2002; 88: 56-60
202
8.1. Introduction
To facilitate thoracic surgery, one-lung ventilation (OLV) is a valuable
technique. One lung is mechanically ventilated while the other is either
occluded, or left open to the atmosphere. Generally a dependent lung tidal
volume (VT) similar to that used in two-lung ventilation is recommended
(1). Respiratory frequencies around 15 bpm are commonly used during
OLV (1). Other authors recommend that the dependent lung should be
ventilated with a tidal volume of 10-12 ml/kg and the respiratory rates
(RR) to maintain the arterial partial pressure of carbon dioxide (PaCO2) at
normal values (2). These suggestions do not consider that many patients
having lung surgery have chronic obstructive pulmonary disease (COPD)
with airflow limitation and pulmonary hyperinflation. During the OLV of
these patients, intrinsic positive end-expiratory pressure (PEEPi) is often
present, and can be increased by an unfavorable ventilatory pattern (3).
Moreover, COPD patients have increased physiological dead space and can
develop hypercapnia.
During the mechanical ventilation of these patients’ lungs, conditions that
impede expiratory flow (increased airway resistance, and the resistance of
the double-lumen tube (DLT) (4,5)), or poor settings of the ventilator may
cause dynamic pulmonary hyperinflation (DPH) and PEEPi. In addition,
alveolar overdistension by severe DPH in the ventilated lung may divert
blood flow to the nonventilated lung, and impair arterial oxygenation.
203
Increases in RR, VT, or a reduction of expiratory time (TE) would
encourage the development of PEEPi and DPH (7). Severe DPH causes
circulatory depression and pulmonary barotraumas (4,7,8).
The settings for mechanical ventilation in patients with COPD during OLV
to avoid excessive DPH and hypercapnia have not been clearly identified.
Studies performed during OLV have mainly assessed the effect of VT
changes (9,10), or the isolated effect of altered RR with unchanged VT
(11). However, with constant RR, a greater VT would promote PEEPi (6),
while a smaller VT would not be adequate because of dependent lung
atelectasis and hypercapnia. If a constant VT is chosen, with changes in
RR, then an increase of RR would cause PEEPi and a decrease in RR would
increase hypercapnia because of reduced minute ventilation. We tested if
a ventilatory pattern, which combined a reduction in RR with constant
minute volume (with a proportional VT increase), could control both
hypercapnia and PEEPi during OLV of patients with chronic obstructive
pulmonary disease.
204
8.2. Materials and methods
The Institutional Review Board approved the investigation and informed
consent was obtained from 15 patients undergoing elective thoracotomy
for tumor resection and requiring OLV. Before surgery, pulmonary function
was tested in the sitting position, including spirometry and static lung
volumes determined by plethysmography (MasterScreen Body™, Jaeger
and Toennies™, Erich Jaeger GmbH, Hoechberg, Germany). Only patients
whose preoperative functional residual capacity (FRC) exceeded 120% of
the predicted value were included in the study. Patient details and
pulmonary function values are shown in Table 8.1..
Table 8.1. Patient details and preoperative pulmonary function tests. FEV1, forced expiratory
volume in 1s.
Mean (n=15) SD
Age (yr) 63 9
Height (cm) 173 8
Weight (kg) 78 16
FEV1 (% predicted) 68 12
FRC (%predicted) 162 42
RV (%predicted) 183 62
PaCO2 (mmHg) room air, supine 38 4
PaO2 (mmHg) room air, supine 77 10
Abbreviations: FEV1: Forced expiratory volume in 1 s; FRC: Functional residual capacity; RV:
Residual volume; PaCO2 : Arterial carbon dioxide partial pressure; PaO2 : Arterial oxygen partial
pressure
205
In all patients, anesthesia was induced and maintained with a variable
rate continuous infusion of propofol (loading dose 2mg/kg) followed by
continuous infusion of 3-5 mg/kg/h and inhalation of isoflurane (<0.5
MAC). Epidural analgesia at the mid-thoracic level (Th7-Th8) (test dose of
3 ml 2% lidocaine with epinephrine 1/200000 followed by a continuous
infusion of bupivacaine 0.5% 2-5 ml/h) was started before induction of
general anesthesia and maintained continuously during surgery.
Pancuronium (loading dose of 80 µg/kg and top up doses of 20 µg/kg)
was used to allow tracheal intubation and to maintain neuromuscular
block throughout surgery. The neuromuscular block was assessed by
regular measurements of post-tetanic count during the procedure. A three
lead electrocardiogram, invasive radial arterial pressure and arterial
oxygen saturation were monitored continuously. The Datex® Ultima™ SV
capnometer (Datex Instrumentarium, Helsinki, Finland) was used to
measure expired end-tidal carbon dioxide tension (ETCO2), flow-volume
and pressure-volume loops. In all patients, the bronchus of the dependent
lung was intubated with an appropriate size (based on sex and height of
the patients)(12), DLT (Broncho-cath™, Mallinckrodt Laboratories,
Athlone, Ireland). This was correctly positioned using fiberoptic
bronchoscopy, first in the supine position and then in the lateral position.
The patients’ lungs were ventilated with 50% oxygen in air, with a
constant inspiratory flow ventilator (Siemens Servo 900 C; Siemens
Elema; Solna , Sweden). Initial settings were VT 10 ml/kg and a rate of 10
bpm. External positive end-expiratory pressure (PEEP) was set to zero.
206
Inspiratory time (TI) was 33% of total cycle time (TTOT) and end-
inspiratory pause was 10% (of TTOT). The ventilator and the breathing
system were inspected carefully and checked for potential leaks. After
turning the patient in the lateral position and checking of the correct
position of the DLT by bronchoscopy, the tracheal limb of the DLT was
clamped and opened to the atmosphere to allow lung collapse. Three
ventilatory rates of 5, 10, and 15 bpm (RR5, RR10, RR15) were applied in
a random sequence and maintained for 15 min. After each15-min period,
end-inspiratory and end-expiratory occlusions were performed to measure
elastance and intrinsic positive pressure, and arterial blood gas samples
were drawn. The occlusions were maintained until the airway pressure
was stable (6). The airway pressure was recorded with an ink recorder
(Gould Brush 2600, Gould Inc., Instrument Systems Division, Cleveland,
OH, USA) with a paper speed of 25 mm/s. Blood gas samples were
analyzed immediately after they were drawn, then corrected according to
the patient’s temperature, with the Synthesis® 350 (Instrumentation
laboratory, Milan, Italy) blood gas analyzer.
From the recorded data, the elastic recoil pressure (Pel,rs) of the dependent
lung-thorax was determined by subtracting the recorded PEEPi from the
static end-inspiratory pressure. Static compliance of the respiratory
system (Cst,rs) was obtained by dividing the expiratory tidal volume by
Pel,rs. The exhaled volume was corrected for the compliance of the
respiratory circuit (7ml/cm H2O) to eliminate errors from the volume
compressed in the tubing (13).
207
The study was performed in the lateral position, before the surgical
procedure.
208
8.3. Statistical analysis
We used the Kolmogorov-Smirnov one-sample test to check that the
samples were normally distributed. Repeated measures ANOVA was used
to test for differences between RRs. Pair wise comparisons between
groups were made, using Student’s paired t-test with the Bonferroni
adjustment for multiple comparisons, using the GB-Stat 6.0 for IBM and
compatibles (Dynamic Microsystems, Inc., 1996). P-values < 0.05 were
considered significant. Data are presented as mean (SD).
209
8.4. Results
All the patients had mild pulmonary hyperinflation. The pulmonary
function tests and blood gas values are shown in Table 8.1.. The
perioperative measured and calculated variables at the different RRs are
shown in Table 8.2..
With a reduced RR and unchanged minute volume, VT (and TTOT and TE)
increased. These changes were inversely proportional to the changes in
the RR (i.e. x2 or X1.5). Peak inspiratory airway pressure (Ppeak) and
Pplateau were significantly greater when the RR was decreased (p<0.01,
repeated measures ANOVA).
Pel,rs increased at lower RR (p<0.01, repeated measures ANOVA), but Cst,rs
did not change significantly (Table 8.2.). During dependent lung
ventilation, PEEPi values greater than 10 cm H2O were found in only two
patients (at RR 10 and RR 15). Zero PEEPi values were recorded in four
patients at RR 5 and another one at RR 10. Despite the larger VT at lower
RR, PEEPi was less (p<0.01, Bonferroni t-test), but without statistical
significance when comparing RR 15 and RR 10 (Table 8.2. and Figure
8.1.).
Arterial partial pressure of oxygen was not affected (Table 8.2.). PaCO2
was less (p<0.01, Bonferroni t-test) at RR 5 in comparison with RR 15 or
RR 10. End-tidal carbon dioxide partial pressure (ETCO2) increased
(p<0.01, repeated measures ANOVA) when the RR was reduced (Table
8.2.). Consequently, the arterial to end-tidal carbon dioxide partial
210
pressure difference (P(a-ET)CO2) was significantly less when the RR was
small (Table 8.2.).
Table 8.2. Respiratory mechanics and blood gas values during the three RRS. Values are mean
(SD). Student’s t-test with Bonferroni adjustement for multiple comparisons. *p<0.01 comparing
RR 10 with RR 15. ‡p<0.01 comparing RR 5 with RR 10. #p<0.01 comparing RR 5 with RR 15. Pel,rs, elastic recoil pressure.
RR 5
RR 10
RR 15
p-values
repeated
measures
ANOVA
VT (ml) 1234 (197)# 623 (97) ‡ 433 (81)* <0.01
Ppeak (cmH2O) 33 (5)# 26 (6)‡ 24 (5)* <0.01
Pplateau (cmH2O) 21 (5)# 17 (6) ‡ 14 (5)* <0.01
PEEPi (cmH2O) 3 (3)# 5 (4) ‡ 6 (4) <0.01
Pel,rs (cmH2O) 18 (6)# 11 (6) ‡ 7 (5)* <0.01
Cst,rs (ml/cmH2O) 82 (41) 75 (38) 97 (83) NS
ETCO2 (cmH2O) 36 (6)# 35 (5) ‡ 33 (5)* <0.01
PaCO2 (mmHg) 39 (4)# 41 (4) ‡ 42 (4) <0.01
PaO2 (mmHg) 200 (58) 193 (65) 191 (69)* NS
P(a-ET)CO2 2 (3)# 7 (5) ‡ 8 (5)* <0.01
Abbreviations: VT: Tidal volume; Ppeak: Peak inspiratory airway pressure; Pplateau: plateau
inspiratory airway pressure; PEEPi: positive end-expiratory pressure; Pel,rs: Elastic recoil pressure
of the respiratory system; Cst,rs: Static compliance of the respiratory system; ETCO2: End-tidal
carbon dioxide tension; PaCO2: Arterial carbon dioxide partial pressure; PaO2: Arterial oxygen
partial pressure; P(a-ET)CO2: Arterial to end-tidal carbon dioxide partial pressure difference.
211
Figure 8.1. Flow-volume loop measured with the Datex Ultima SV Capnometer (Datex
Instrumentarium, Helsinki, Finland) of one patient when changing the respiratory rate.
Abbreviations: PEEPi: Intrinsic positive end-expiratory pressure; RR: Respiratory rate
212
8.5. Discussion
Our main findings were a decrease of PEEPi when RR was reduced, along
with a decrease in PaCO2 and P(a-ET)CO2, and with an increased Ppeak,
Pplateau, and Pel,rs (p<0.01, repeated measures ANOVA) (Table 8.2. and
Figure 8.1.).
Co-existing lung disease is rule rather than exception in patients
undergoing lung surgery. The volume, frequency and timing of gas
delivered to the dependent lung can have important, disease-specific
effects on the cardiovascular and respiratory systems (14).
Until now, a low RR has not been studied during OLV in patients with
pulmonary hyperinflation in a randomized fashion. Robinson and co-
workers (15) describe two ventilator-dependent patients with cystic
fibrosis who could not be managed with conventional ventilation during
sequential double-lung transplantation. Lung implantation was facilitated
by OLV at a slow rate (6bpm) with a long TI (5s) and a long TE (5s),
implantation of donor lung, by decreasing hypercapnia and reducing
pulmonary arterial pressure.
Lowering the RR with unchanged minute volume implies an increase in VT.
In patients without significant pulmonary disease, OLV during general
anesthesia decreases arterial oxygenation because of atelectasis in the
dependent lung (9). However, patients with COPD do not develop
atelectasis and decrease in FRC during anesthesia (16), presumably
because of long standing hyperinflation.
213
In obstructive airway disease, lung injury with alveolar leaks may be
caused by dynamic hyperinflation during mechanical ventilation. Some
authors believe that increased inspiratory airway pressures are necessary
to overcome airway resistance when high inspiratory flow rates are used,
as a large part of the inspiratory pressure is not transmitted to the alveoli
(17,18). Because alveolar distending volume is not readily measured
clinically, Pplateau measured during an inspiratory pause is generally
accepted as a reasonable estimate of peak alveolar pressure (19, 20).
During positive pressure ventilation, the Pplateau below which lung injury is
unlikely is approximately 35 cm H2O, commonly believed to correspond to
an alveolar pressure of approximately 30 cm H2O. For safer OLV
conditions, Slinger (21) suggested limiting Pplateau to 25 cm H2O Given this
range of opinion, careful monitoring of patients managed with this
approach is mandatory. In the present study, the Pplateau was generally
less than this value (except a single value of 26 cm H2O at RR 5) despite
the VT increases from 433 (81) ml at RR 15 to 623 (97) ml at RR 10 and
1234 (197) ml at RR 5. The duration of OLV periods in this study was
short and the patients were anesthetized and paralyzed, but concern could
arise from long-term use of low-rate ventilation.
At large values of VT, an increased PEEPi might be expected, because in
addition to the degree of airflow obstruction, one of the major
determinants in the development of PEEPi is the VT to be exhaled in a
fixed fraction of time (22). Nevertheless, besides VT, the main
determinant of dynamic hyperinflation and PEEPi in COPD patients is the
214
absolute value of TE and not the TI: TTOT ratio per se (4). Therefore,
considering a given respiratory system with its specific compliance and its
expiratory and inspiratory resistances, for a fixed Ti fraction, PEEPi will
increase when RR is raised (reduction of TE) and decrease when the RR is
reduced (increase of TE). Small changes in TE did not affect PEEPi because
the volume expelled per unit time near the end of exhalation is very small
(23). In patients with COPD and acute respiratory failure, while PEEPi was
increased by TE shorter than 3 s, prolonging TE more than 3 s had little
effect on PEEPi (6). Both studies were of mechanically ventilated patients
with acute exacerbation of COPD, and the effect of prolonging TE on gas
exchange were not assessed. In the present study of stable COPD patients
during OLV in the lateral position, with constant inspiratory time fraction
(TI= 33% of TTOT), a reduced RR and increased VT, increased the TE from
2.6 s at RR 15 to 4 s at RR 10 and 8 s at RR 5 bpm, respectively. In
contrast to the findings of Rossi (6), because of the parallel increase of VT
in our study, a decrease of PEEPi was observed only at very long TE
associated with the RR 5 (Table 8.2. and Figure 8.1.).
An insignificant increase of PaO2 when RR was decreased (and VT
increased) supports the results of previous studies (9-11). Increased
inspiratory airway pressures causing increased vascular resistance in the
dependent lung of some patients could explain the lack of consistently
improved oxygenation during high VT ventilation (9, 24).
In this study, the high RR (15 bpm) and a small VT ventilation caused
hypercapnia with a mean PaCO2 of 43 mm Hg. When ventilation was
215
performed at low RR (5 bpm) and high VT, mean PaCO2 decreased
significantly to 39 mm Hg. This value approached the preoperative values
(38 (4) mm Hg). The reduced P(a-ET)CO2 suggests altered intrapulmonary
gas distribution at low RR, but differences in ETCO2 may also simply
reflect the fact that, with a significant positive slope of the capnogram,
prolonging expiration itself will increase ETCO2 without necessarily
reflecting any change in gas exchange.
This ventilatory pattern – lowered RR (and increased VT) with constant
minute volume – may represent a simple way to reduce PEEPi and
hypercapnia during OLV in patients with pulmonary hyperinflation, as
changing RR or VT alone increases either the PEEPi or the PaCO2. Careful
monitoring of these patients for the risk of barotraumas is mandatory.
However, given the lack of significant effect on oxygenation (the hallmark
of OLV), it is difficult to recommend the technique of high VT, low RR OLV
for routine practice. This ventilatory management should be reserved for
patients with severe COPD in whom PEEPi and hypercapnia would possibly
complicate OLV.
216
8.6. References
1. Benumof JL. Physiology of the lateral decubitus position, open chest
and one-lung ventilation. In: Kaplan JA, ed. Thoracic Anesthesia, 2nd
Edn. New York: Churchill-Livingstone Inc, 1993; 193-221
2. Cohen E. Anesthetic management of one-lung ventilation. In: Cohen
E, ed. The practice of thoracic anesthesia. Philadelphia: JB Lippincott,
1995: 308-40
3. Bardoczky GI, Yernault JC, Engelman EE, et al. Intrinsic positive end-
expiratory pressure during one-lung ventilation for thoracic surgery.
The influence of preoperative pulmonary function. Chest 1996; 110:
180-4
4. Pepe PE, Marini JJ. Occult positive end –expiratory pressure in
mechanically ventilated patients with airflow obstruction. Am Rev
Respir Dis 1982; 126: 166-70
5. Scott LR, Benson MS, Bishop MJ. Relationship of endotracheal tube
size to auto-PEEP at high minute ventilation. Respir Care 1986; 31:
1080-2
6. Rossi A, Poleses G, Brandi G, Conti G. Intrinsic positive end-
expiratory pressure (PEEPi). Intensive Care Med 1995; 21: 522-36
217
7. Rossi A, Polese G, Milic-Emili J. Mechanical ventilation in the passive
patient. Theory and clinical investigation. In: derenne JP, Whitelaw
WA, Similowski T, eds. Acute Respiratory Failure in Chronic
Obstructive Pulmonary Disease. New York: Marcel Dekker, 1996;
709-46
8. Gottfried SB, Rossi A, Milic-Emili J. Dynamic hyperinflation, intrinsic
PEEP and the mechanically ventilated patient. Int Crit Care Dig 1986;
5: 30-3
9. Flacke JW, Thompson DS, Read RC. Influence of tidal volume and
pulmonary artery occlusion on arterial oxygenation during
endobronchial anesthesia. South Med J 1976; 69: 619
10. Katz JA, Laverne RG, Fairley HB, Thomas AN. Pulmonary oxygen
exchange during one-lung anesthesia: effect of tidal volume and
PEEP. Anesthesiology 1982; 56: 164-71
11. Torda TA, McCullogh CH, O’Brien HD, et al. Pulmonary venous
admixture during one-lung ventilation anaesthesia. Anaesthesia
1974; 29: 272-9
12. Hannallah MS, Benumof JL, McCarthy PO, et al. Comparison of three
techniques to inflate the bronchial cuff of left polyvinyl chloride
double-lumen tubes. Anesth Analg 1993; 77: 990-4
13. Chatburn RL. Classification of mechanical ventilators. In: Tobin MJ,
ed. Principles and Practice of Mechanical Ventilation. New York:
McGraw-Hill, 1994; 37-65
218
14. Slutsky AS. Consensus conference on mechanical ventilation.
Intensive Care Med 1994; 20: 64-79
15. Robinson RJ, Shennib H, Noirclerc M. Slow-rate, high pressure
ventilation: a method of management of difficult recipients during
sequential double-lung transplantation for cystic fibrosis. J Heart Lung
Transplant 1994; 13: 779-84
16. Gunnarson L, Tokics L, Lundquist H, et al. Chronic obstructive
pulmonary disease and anesthesia: formation of atelectasis and gas
exchange impairment. Eur Respir J 1991; 4: 1106-16
17. Dreyfuss D, Saumon G. Role of tidal volume, FRC, and end-inspiratory
volume in the development of pulmonary edema following mechanical
ventilation. Am Rev Respir Dis 1993; 148: 1194-203
18. Tuxen DV. Independent lung ventilation. In: Tobin MJ, ed. Principles
and Practice of Mechanical Ventilation. New York: McGraw-Hill, 1994;
571-88
19. Marini JJ, Ravenscraft SA. Mean airway pressure: physiologic
determinants and clinical importance – Part 1: Physiologic
determinants and measurements. Crit Care Med 1992; 20: 1461-72
20. Marini JJ, Ravenscraft SA. Mean airway pressure: physiologic
determinants and clinical importance – Part 2: Clinical implications.
Crit Care Med 1992; 20: 1604-16
21. Slinger P. New trends in anaesthesia for thoracic surgery including
thoracoscopy. Can J Anaesth 1995; 42: R77-84
219
22. Similowski T, Derenne JP, Milic-Emili J. Respiratory mechanics during
acute respiratory failure of chronic obstructive pulmonary disease. In:
Derenne JP, Whitelaw WA, Similowski T, eds. Acute Respiratory
Failure in Chronic Obstructive Pulmonary Disease. New York: Marcel
Dekker, 1996; 40-64
23. Smith TC, Marini JJ. Impact of PEEP on lung mechanics and work og
breathing in severe airflow obstruction. J Appl Physiol 1988; 65:
1488-99
24. Khanam T, Brandwaite MA. Arterial oxygenation during one-lung
anaesthesia. Anaesthesia 1973; 28: 132-8
220
8.7. Summary
During one-lung ventilation of patients with chronic obstructive pulmonary
disease, three respiratory rates with unchanged minute volume were
assessed. At low respiratory rate, end-tidal carbon dioxide pressure was
reduced. Intrinsic positive end-expiratory pressure was reduced even with
larger tidal volumes at low rate ventilation, because of increased
expiratory time at this rate. This technique, however, should be kept for
the ventilatory management of some highly difficult patients, in whom
other ventilatory methods gave negative results.
221
9. General discussion and conclusions
222
9.1. Summary of the results
These studies were performed in patients with normal lung function and in
patients with stable chronic obstructive pulmonary disease (COPD) with
variable degrees of airflow obstruction and pulmonary hyperinflation (PH),
scheduled for lung surgery. All the studied patients (n=167) remained
hemodynamically stable throughout the studies, and major hypoxemic
event (arterial partial pressure of oxygen < 90 %) complications due to
double-lumen tube malposition did not occur. Written informed consent
was obtained from all the patients before inclusion in a study. All the
patients were extubated at the end of the surgical procedure in the
operating room.
The presented studies have explored the three endpoints of one-lung
ventilation: optimal position of the double-lumen tube with functional lung
separation, and adequate ventilation and oxygenation.
The double lumen endotracheal tube (DLT) remains the standard means of
producing lung separation in adult thoracic anesthesia. During thoracic
surgery, the change from two-lung ventilation (TLV) to OLV should be
carefully monitored. In the first study we have evaluated the usefulness of
a continuous method to determine DLT position during thoracic surgery,
by analyzing the changes in inspiratory pressures during transition from
two-lung to one-lung ventilation in patients tracheally intubated with a
DLT using a classical method of intubation without fiberoptic bronchoscopy
(FOB). Peak (Ppeak) and plateau (Pplateau) inspiratory airway pressures were
recorded with an on-line respiratory monitor before and after clamping the
223
tracheal limb of the DLT. The position of the DLTs was evaluated by FOB
with the patient in the supine position. Of the 51 intubations, the DLT was
malpositioned in 15 cases (29.5%). Ppeak and Pplateau increased significantly
when switched from two-lung to one-lung ventilation in both correctly and
incorrectly positioned DLTs. When the DLT was in a correct position, Ppeak
increased by a mean of 41.9%. When the DLT was malpositioned, this
increase was significantly larger (74.9% and 68.8%, respectively).
Three tests commonly used as markers of malpositioned DLTs were
evaluated based on data of this study, and it was established that,
although the pressure differences related to position are statistically
significant, as a single value they cannot be used for clinical decision
making. The “golden standard” in the positioning of DLTs remains the
FOB.
The major protective mechanism against hypoxemia during OLV is the
hypoxic pulmonary vasoconstriction (HPV). Other, passive mechanisms,
like gravity, surgical manipulation or preexisting lung disease may play
also an important role in blood flow redistribution during OLV. We
compared the effects of position and fraction of inspired oxygen (FiO2) on
oxygenation during thoracic surgery in 24 consenting patients randomly
assigned to receive an FiO2 of 0.4, 0.6, and 1.0 %. Arterial oxygen tension
was decreased in all three groups during one-lung ventilation in
comparison with the two-lung ventilation values, but the decrease was
significantly less in the lateral, compared with the supine position.
224
We conclude that, compared with the supine position, gravity augments
the redistribution of perfusion as a result of hypoxic pulmonary
vasoconstriction, when patients are in the lateral position, which explains
the higher PaO2 during OLV.
As some controversy still exists between diverse experimental designs and
methods about the magnitude of the gravity among the factors influencing
regional pulmonary blood flow in diverse situations, the aims of the third
study were to separate the effects of gravity on blood flow redistribution
and HPV during OLV of dependent and nondependent lung in the supine
and lateral positions.
Thirthy consenting adult patients scheduled for elective right thoracotomy,
with mild degrees of pulmonary hyperinflation, with comparable
preoperative left/right lung perfusion scintigraphy, were randomly
assigned to ventilate either the dependent (D) lung (Group D-OLV, n=15)
or nondependent (ND) lung (Group ND-OLV, n=15). The study was
performed in standardized conditions before surgery. During OLV
episodes, the PaO2 increased in the lateral position as compared to the
supine one in D-OLV, but, at the contrary, for ND-OLV Group, PaO2 was
decreased
This study points out that in stable COPD patients with mild pulmonary
hyperinflation, gravity plays a major (and not minor) role in blood flow
redistribution during OLV in the lateral position.
Data concerning the influence of isovolemic hemodilution (IH) on arterial
oxygenation during OLV for thoracic surgery are lacking.
225
We studied 47 patients, with a hemoglobin >14 g/dl, scheduled for lung
surgery [17 with normal lung function (Group NL), 17 with chronic
obstructive pulmonary disease (COPD) (Group COPD), and 13 with COPD
as control for time/anesthesia effect (Group CTRL)]. Anesthesia was
standardized. The tracheas were intubated with a double-lumen tube.
Ventilatory settings and FiO2 remained constant. The study was done in
the supine position, before surgery. OLV was initiated for 15 minutes. Two
lung ventilation was re-instituted and IH was performed (500 ml) with an
identical volume of hydroxyethyl starch administered. Subsequently, OLV
was again performed for 15 minutes. In the Group CTRL the same
sequences of OLV were done without IH. At the end of each period of OLV,
pulmonary mechanics and blood gases were recorded. Data were analyzed
by ANOVA (mean±SD). In the Group NL and the Group CTRL, the arterial
oxygen partial pressure (PaO2) remained constant, while it decreased in
Group COPD from 119±21 mmHg before IH, to 86±16mmHg after IH
(p<0.01). Mild IH impairs gas exchange during OLV in COPD patients, but
not in patients with normal lung function.
Thus, mild to moderate IH may be a safe method for blood preservation
during thoracic surgery and OLV for patients with normal lung function;
care should be taken when IH is performed (or by bleeding) in stable
COPD patients with mild to moderate pulmonary hyperinflation, since they
don’t tolerate anemia during OLV.
Recently, it has been recognized that the volume, frequency and timing of
gas delivered to the dependent lung, might have important effects on the
226
respiratory system. In the last study, lung mechanics and gas exchange
during OLV of patients with chronic obstructive pulmonary disease were
evaluated, using three respiratory rates (RR) and unchanged minute
volume.
We studied 15 patients about to undergo lung surgery, during anesthesia,
and placed in the lateral position. Ventilation was with constant minute
volume, inspiratory flow and FiO2. For periods of 15 min., RR of 5, 10 and
15 bpm were applied in a random sequence and recordings were made of
lung mechanics and an arterial blood gas sample was taken. PaO2 changes
were not significant. At the lowest RR, PaCO2 decreased. PEEPi was
reduced even with larger tidal volumes, most probably caused by
increased expiratory time at the lowest RR. Thus, a reduction in RR
reduces PEEPi and hypercapnia during OLV in anesthetized patients with
chronic obstructive lung disease.
227
9.2. Limitations of the studies and future perspectives
Providing safe perioperative management for thoracic surgical patients
remains one of the most consistently challenging areas both technically
and intellectually within the specialty of Anesthesia. The practice of
thoracic anesthesia is continually evolving. In the last half of the 20th
century, thoracic surgery for malignancies became the pre-eminent
procedure, and the last decade has seen the beginnings of surgery for
end-stage lung disease. The attendant advances in anesthetic and
perioperative care that have accompanied these recent changes in the
surgical spectrum have made it such that almost any patient with a
respectable lesion can be considered “operable”.
Some consideration should be given to the limitations in our studies.
First, the studies were pure clinical studies; according to institutional
structures, monitoring and measurements were limited. For example,
pulmonary artery catheter could not be inserted, so we were not able to
perform shunt measurements.
Second, given the absence of data on the subject and the patient
population studied (stable COPD with mild to moderate degrees of
pulmonary hyperinflation), we had to limit our investigations. The different
results in the experimental animal models and the clinical studies
presented above, could be for any number of reasons, including the
species difference and the presence of COPD.
228
Third, our results could have been influenced by the repeated
endotracheal tube clamping maneuvers. Benumof (1) showed that
intermittent hypoxic challenges may potentate the HPV response in an
animal model. Carlsson et al (2), found maximal HPV within 15 minutes of
hypoxia, and Domino and co-workers (3) observed a maximal response
after the very first hypoxic challenge during OLV in closed-chest dogs. In
order to eliminate the changes over time in oxygenation during OLV, a
control group of patients with altered preoperative pulmonary function
was studied. Two sequences of inflating/deflating the lung did not result in
significant difference in arterial oxygenation (4).
Fourth, the presented studies were done on the patients with closed chest,
instead of the usual clinical situation with open chest. These differences
may affect extrapolation of the results, because of the absence of surgical
manipulation, which may contribute to blood flow redistribution during
OLV (5).
Further studies, are needed to elucidate the mechanism of these findings,
but we have to keep in mind the (severe) limitations of such studies in
clinical conditions. It would be difficult to imagine the experimental studies
on animals, in clinical settings. However, even if time consuming and
more invasive, the future studies have to include necessarily shunt
measurements by using inert gas technique or radiological markers for the
perfusion of the lungs.
229
We should keep in mind the availability, for lung separation, of other
devices than double-lumen tubes (Arndt endobronchial blocker, Univent
tube), which are less invasive and easier to place, but they have known
disadvantages. Airway pressure changes, resistance to flow, and
ventilatory leaks caused by such lung separation devices were never
investigated.
Pain management is yet not well studied, as well as its relationship with
outcome. One should not forget the possibility of occurrence of post-
pneumectomy lung edema or lung edema after inflating the lung.
The outcome after thoracic surgery has to be further investigated in a
prospective, randomized manner.
230
9.3. References
1. Benumof JL. Intermittent hypoxia increases lobar hypoxic pulmonary
vasoconstriction. Anesthesiology 1983; 58:399-404.
2. Carlsson AJ, Bindslev L, Santesson J, et al. Hypoxic pulmonary
vasoconstriction in the human lung: the effect of prolonged
unilateral hypoxic challenge during anaesthesia. Acta Anaesthesiol
Scand 1985; 29:346-51.
3. Domino K, Chen L, Alexander C, et al. Time course and responses of
sustained hypoxic pulmonary vasoconstriction in the dog.
Anesthesiology 1984; 60:562-6.
4. Szegedi LL, Van Der Linden Ph, Ducart A, et al. The effects of acute
isovolemic hemodilution on oxygenation during one-lung Ventilation.
Anesth Analg 2004; 99
5. Benumof JL. Special respiratory physiology of the lateral decubitus
position, the open chest and one-lung ventilation. In: Benumof JL,
ed. Anesthesia for thoracic surgery. Philadelphia: WB Saunders,
1995, 123-51.
231
10. End remarks
Despite the ever-lengthening list of procedures for which one-lung
anesthesia seems to be indicated, it should always be remembered that
one-lung anesthesia adds considerably to the complexity of anesthetic
technique and also to the associated level of risk. Prolonged attempts at
endobronchial intubation may put the patient at risk – at best from
damage caused by trauma and at worst from hypoxia. It is also too easy
to be so distracted by the technology of one-lung anesthesia or its
complexity that a falling blood pressure or profound bradycardia goes
unnoticed. The anesthesiologist should have a logical plan for dealing with
ventilatory problems and hypoxia during one-lung ventilation. One-lung
ventilation should not be pursued at all costs, because never justifies
rending a patient hypoxic.
232
233
11. Summary
One-lung ventilation is required to control the distribution of ventilation if
the airways to the opposite lung are to be opened, or if the surgical
procedure is facilitated by collapse of that lung. The technique is also used
to prevent spread of secretions from one-lung to the other, and as a route
for therapeutic unilateral lung lavage. One-lung ventilation is inevitably
associated with gross changes in respiratory physiology. One-lung
ventilation implies not only functional lung separation, but also adequate
ventilation and oxygenation. This work examined the three endpoints of
one-lung ventilation: optimal position of the double-lumen tube, functional
lung separation, and adequate ventilation and oxygenation.
The double lumen tube remains the standard means of producing lung
separation in adult anesthesia. The usefulness of a continuous method to
determine double-lumen tube position during thoracic surgery was
evaluated (Chapter 4), by analyzing the changes in inspiratory pressures
during transition from two-lung to one-lung ventilation, and it was
established that, although the pressure differences related to position are
statistically significant, as a single value they cannot be used for clinical
decision making, but they remain as a warning sign for double–lumen
tube displacement.
During one-lung ventilation, the nondependent lung is excluded from the
ventilation, with all the tidal volume directed into the dependent lung. In
this situation, the distribution of perfusion is the major determinant of the
234
degree of venous admixture. The blood flow through the operative lung
becomes a right-to-left shunt in addition to that, which exists in the
ventilated lung. Mechanisms that tend to decrease the percent of cardiac
output perfusing the nondependent, nonventilated lung are passive (e.g.,
mechanical-like gravity, surgical manipulation, amount of pre-existing
lung disease) or active (e.g., hypoxic pulmonary vasoconstriction).
Increasingly complex operations are performed on patients with seriously
compromised lung function, because of the development of cardiothoracic
surgery, anesthesia and medicine. Concurrent lung disease is the rule
rather than exception in patients undergoing thoracic surgery.
A variety of surgical procedures are performed with patients in the supine,
instead of the traditional lateral position. A study (Chapter 5) performed
on patients with stable chronic obstructive pulmonary disease, showed a
significantly better oxygenation during one-lung ventilation in the lateral
than in the supine position. As a logical consequence of these results, we
separated (Chapter 6) the effects of gravity from other factors influencing
perfusion redistribution during one-lung ventilation, suggesting the
greater importance of the passive gravitational factor in blood flow
redistribution during one-lung ventilation than previously believed.
The normal response of the pulmonary vasculature to atelectasis is an
increase in pulmonary vascular resistance (in the atelectatic lung) and the
increase in atelectatic lung resistance is due almost entirely to hypoxic
pulmonary vasoconstriction, a protective reflex mechanism that diverts
blood flow away from the atelectatic lung. Although hypoxic pulmonary
235
vasoconstriction is an intrinsic property of the pulmonary vasculature,
several studies indicate that the red blood cells may also play a role in the
mechanism of its occurrence. Previous animal studies suggested that
severe anemia is associated with remarkable stability of pulmonary gas
exchange and, that the gas exchange efficiency of the normal lung may be
improved with acute hemodilution. The effect of mild isovolemic
hemodilution on oxygenation during one-lung ventilation in patients with
normal preoperative lung function, and in patients with stable chronic
obstructive pulmonary disease was studied (Chapter 7). The patients with
compromised lung function are more sensible to blood loss than patients
with normal lung function.
The ventilatory parameters to be used during one-lung ventilation,
recommended by textbooks, fail to consider that the majority of patients
scheduled for lung surgery have chronic obstructive pulmonary disease
with variable degrees of airflow limitation and pulmonary hyperinflation.
During mechanical ventilation of these patients, conditions that impede
expiratory flow or inadequate ventilatory settings may predispose to
dynamic pulmonary hyperinflation and intrinsic positive end-expiratory
pressure. By changing inspiratory time, and thereby expiratory time, or by
altering the respiratory rate at the same minute volume, inspiratory
airway pressures and intrinsic positive end-expiratory pressure can be
deliberately manipulated (Chapter 8).
Further studies are needed to elucidate the complex mechanisms of these
findings, but we have to keep in mind the limitations of such
236
investigations in clinical conditions (limited monitoring and measurement
possibilities, coexisting pathology).
237
Samenvatting
Eén-long ventilatie is vereist voor de controle van de distributie van
ventilatie indien de luchtwegen van de andere long operatief geopend
worden of indien de heelkunde vergemakkelijkt wordt door het collaberen
van de zieke long. Deze techniek wordt ook gebruikt om de verspreiding
van longsecreties van een aangetaste long naar de gezonde long te
vermijden, alsook voor een therapeutische longspoeling. Eén-long
ventilatie is overmijdelijk geassocieerd met grote veranderingen in
respiratoire fysiologie. Eén-long ventilatie vereist niet alleen functionele
longscheiding, maar ook adequate beademing en oxygenatie. In dit werk
worden de drie eindpunten van één-long ventilatie onderzocht: optimale
positionering van de dubbel-lumen tube, functionele longscheiding en
adequate ventilatie en oxygenatie.
De dubbel-lumen tube blijft de standaard voor longscheiding bij
volwassenen. Het praktisch nut om continu de positionering van een
dubbel-lumen tube gedurende de thoraxchirurgie te evalueren is
bestudeerd (Hoofdstuk 4) door de analyse van de veranderingen in
inspiratoire drukken tijdens overschakeling van twee-longen naar één-
longventilatie. Het was duidelijk dat, alhoewel de drukverandering
gecorreleerd aan de positionering statistisch significant was, deze meting
alleen niet gebruikt kan worden voor klinische besluitvorming, maar ze
blijft waarschuwen voor dubbel-lumen tube verplaatsing.
238
Tijdens één-long ventilatie wordt de niet-dependente long niet beademd,
terwijl het totale teugvolume gestuurd wordt naar de dependente long. In
dit geval is de distributie van de perfusie de voornaamste determinant van
percentage veneuze bijmenging. De doorbloeding door de geopereerde
long wordt een rechts-links shunt toegevoed aan deze die reeds bestaat in
de geventileerde long. De mechanismen die pogen de perfusie van de
niet-geventileerde long te verminderen zijn passief (bijvoorbeeld
zwaartekracht, chirurgische manipulatie, graad van vooraf bestaande
longaandoening) of actief hypoxische pulmonaire vasoconstrictie).
Meer en meer ingewikkelde operaties worden uitgevoerd op patiënten met
ernstig belaste longfunctie. Geassocieerd longlijden is eerder regel dan
uitzondering bij patiënten die longchirurgie ondergaan.
Verschillende chirurgische ingrepen worden eerder in ruglig uitgevoerd
dan in de klassieke zijlig. Onze studie (Hoofstuk 5) uitgevoerd bij
patiënten met chronisch obstructief longlijden toonde een betere
oxygenatie aan gedurende één-long ventilatie in zijlig ten overstaan van
in ruglig. Ten gevolge van deze bevindingen hebben wij de invloed van de
zwaartekracht afgeplitst van andere factoren die de perfusie beheersen
(Hoofdstuk 6), wijzend op de grotere rol van de zwaartekracht in de
perfusieherverdeling gedurende één-long ventilatie dan vroeger gedacht.
Het normale antwoord van het pulmonaire vasculaire bed op atelectase is
een toename in pulmonaire vasculaire weerstand (in de atelectatische
long). De weerstandstoename in de atelectatische long is voornamelijk het
gevolg van hypoxische pulmonaire vasocontrictie, een beschermende
239
reflex die het bloed afleidt van de niet beademende long. Alhoewel de
hypoxische pulmonaire vasoconstrictie een intrinsieke eigenschap is van
de pulmonaire vaten, wijzen verschillende studies erop dat de rode
bloedcellen daarin ook een rol spelen. Studies tonen aan dat anemie
geassocieerd is met een opmerkelijke stabiliteit van pulmonaire
gasuitwisseling bij dieren. Bij de mens met normale longen verbetert
acute hemodilutie de gasuitwisseling. De invloed van milde isovolemische
hemodilutie op oxygenatie tijdens één-long ventilatie bij patiënten met
normale preoperatieve longfunctie, alsook bij patiënten met chronisch
obstructief longlijden werd bestudeerd (Hoofstuk 7). De patiënten met
longaandoeningen zijn gevoeliger voor bloedverlies dan patiënten met
normale longfunctie.
De ventilatieparameters aanbevolen in leerboeken van één-long
beademing houden geen rekening met het feit dat de meeste patiënten
voor longheelkunde ook chronisch obstructief longlijden hebben. Dit is
gekenmerkt door wisselende longstroombeperking en longhyperinflatie.
Gedurende één-long beademing van deze patiënten, kunnen belemmering
van de expiratoire flow of onjuiste instelling van de ventilator aanleiding
geven tot dynamische pulmonaire hyperinflatie en intrinsiek positieve
eindexpiratoire druk. Door aanpassing van de inspiratoire tijden en
daaraangekoppeld de expiratoire tijd, of door verandering van de
respiratoire frequentie met behoud van het minuutvolume, kunnen de
inspiratoire luchtwegdrukken en de intrinsiek positieve eindexpiratoire
druk geregeld worden (Hoofstuk 8).
240
Verdere studies zijn nodig om het complex mechanisme van deze
bevindingen te verklaren rekeninghoudend met de beperkingen van
dergelijk onderzoek in klinische omstandigheden (beperkingen van de
monitorapparatuur en meettechnologie, beperkingen tgv de onderliggende
pathologie).
241
Résumé
La ventilation unipulmonaire est nécessaire pour le contrôle de la
distribution de la ventilation en cas d’ouverture des voies aériennes du
côté opposé, ou si l’intervention chirurgicale est facilitée par le collapsus
de ce poumon. Cette technique est aussi utilisée afin de prévenir la
contamination par sécrétions d’un poumon sain, ou comme le seul moyen
pour des actes thérapeutiques comme le lavage broncho-pulmonaire
unilatérale. La ventilation unipulmonaire est inévitablement associée à des
changements physiologiques respiratoires importants, impliquant non
seulement une séparation fonctionnelle des poumons, mais aussi une
ventilation et une oxygénation adéquate. Dans ce travail les trois aspects
principaux de la ventilation unipulmonaire ont été examinés: le
positionnement optimal des tubes à double lumière, la séparation
pulmonaire fonctionnelle, la ventilation et l’oxygénation adéquate.
En anesthésie chez l’adulte, le standard utilisé pour isoler les poumons
reste le tube à double lumière. Un moyen continu de détection du
déplacement des tubes à double lumière a été évalué (Chapitre 4) par
l’analyse du changement des pressions inspiratoires au moment de la
transition de la ventilation bipulmonaire et unipulmonaire. Même si les
valeurs relatives aux différences de pression sont statistiquement
significatives, la pression inspiratoire seule ne peut être utilisée en clinique
pour le diagnostique de mauvais positionnement du tube à double
lumière, mais reste un indicateur précieux.
242
Pendant la ventilation unipulmonaire, le poumon nondépendant est exclu
de la ventilation, avec tout le volume courant dirigé vers le poumon
dépendant. Dans cette situation, la proportion du mélange veineux est
déterminée par la distribution de la perfusion. Le flux sanguin qui traverse
le poumon opéré devient un shunt droit-gauche, en plus de ce qui existe
déjà dans le poumon ventilé. Les mécanismes qui tendent à diminuer le
débit cardiaque destiné au poumon non ventilé, peuvent être passifs (par
exemple d’origine mécanique comme la gravité, la manipulation
chirurgicale, l’affection pulmonaire pré-existante) ou actifs (la
vasoconstriction pulmonaire hypoxique).
Grâce aux progrès dans le domaine de la chirurgie, l’anesthésie et la
médecine cardiothoracique, des opérations de plus en plus complexes
peuvent être réalisées sur des patients ayant des fonctions pulmonaires
altérées. En fait, la maladie coexistante est plutôt la règle que l’exception
chez les patients en chirurgie thoracique.
Une large gamme des procédures chirurgicales sont fait avec les patients
en décubitus dorsal au lieu de décubitus latéral classique. Une étude
(Chapitre 5) effectuée sur des patients avec broncho-pneumopathie
chronique obstructive, a montré une meilleure oxygénation pendant la
ventilation unipulmonaire en position latérale que dorsale. A la vue de ces
résultats, nous avons séparé (Chapitre 6) les effets de la gravité des
autres facteurs pouvant influencer la redistribution de la perfusion
pendant la ventilation unipulmonaire, nous suggérant que le facteur
243
gravitationnel passif joue un rôle plus prépondérant dans la redistribution
de la perfusion, qu’on ne le croyait auparavant.
La réponse normale des vaisseaux pulmonaires à l’atélectasie est une
augmentation de la résistance vasculaire pulmonaire (dans le poumon
atélectasié) et cette augmentation est due presque entièrement à la
vasoconstriction pulmonaire hypoxique (un mécanisme reflex protecteur
qui dévie le flux sanguin de régions qui ne sont pas ventilées). Quoique la
vasoconstriction pulmonaire hypoxique soit une propriété intrinsèque des
vaisseaux pulmonaires, des études suggèrent que les globules rouges
peuvent avoir un rôle important dans son mécanisme. Dés études
animales ont suggéré que l’anémie sévère est associée avec une stabilité
remarquable des échanges gazeux pulmonaires et que l’efficacité des
échanges gazeux du poumon normal peut être améliorée avec une
hémodilution aiguë.
Nous avons étudié les effets de l’hémodilution isovolémique légère sur
l’oxygénation pendant la ventilation unipulmonaire chez des patients avec
des fonctions préopératoires normales et chez des patients avec broncho-
pneumopathie chronique obstructive (Chapitre 7). Les patients ayant une
fonction pulmonaire altérée sont plus sensibles aux pertes sanguines que
ceux avec des poumons normaux.
Les recommandations des ouvrages de référence pour les paramètres à
utiliser pour la ventilation unipulmonaire ne prennent pas en considération
que la majorité des patients de chirurgie pulmonaire ont une broncho-
pneumopathie chronique obstructive, avec des degrés variables de
244
limitation du flux expiratoire et une hyperinflation pulmonaire. Pendant la
ventilation unipulmonaire de ces patients, les conditions qui limitent le flux
expiratoire ou des réglages inadéquats du ventilateur, peuvent favoriser
l’apparition d’une hyperinflation pulmonaire dynamique et de la pression
positive intrinsèque en fin d’expiration. On peut volontairement manipuler
la pression positive intrinsèque en fin d’expiration et les pressions dans les
voies aériennes, en changeant le temps inspiratoire, et de ce fait le temps
expiratoire, ou en modifiant la fréquence respiratoire, tout en conservant
le même volume minute (Chapitre 8).
Des études complémentaires sont nécessaires pour élucider les
mécanismes complexes de ces résultats, mais il faut garder à l’esprit les
limitations de ces recherches dans des conditions cliniques (limitation de
mesures et de monitoring invasif, pathologie coexistante, temps).