Peter Slinger, MD
The Toronto Hospital (General) and University of
Toronto
Objectives:
The routine isolation and separation of one of the body's
major paired organs, as is frequently necessary during lung
surgery, is unique in medicine. The evolution of elective
thoracic surgery was delayed more than half a century following
the introduction of ether because early anesthesiologists did not
have the technology to protect healthy lung regions from
secretions, hemorrhage or air leaks in diseased lungs. In the
mid-1930's several different practitioners pioneered the use of
endobronchial tubes (EBT), bronchial blockers (BB) and
double-lumen tubes (DLT) to achieve lung isolation1.
In spite of numerous modifications, these three methods have
persisted to the present as the basis of modern lung isolation
techniques. All of the early methods required extensive clinical
experience and rigid bronchoscopy to achieve satisfactory
endobronchial positioning of the tube or blocker. The
modification of the double-lumen tube by Carlens in 19492,
for split-lung function studies, was a landmark in the progress
of thoracic anesthesia. The Carlens tube was widely adopted by
anesthesiologists because it provided reliable lung isolation and
one-lung ventilation in the vast majority of patients and did not
require bronchoscopy for positioning.
The second half of this century has seen refinements of the
DLT from that of Carlens to a tube specifically designed for
intraoperative use (Robertshaw3) with larger,
D-shaped, lumens and without a carinal hook. Current disposable
polyvinyl chloride DLT's have incorporated high-volume
low-pressure tracheal and bronchial cuffs4. These
recent DLT refinements have 2 major drawbacks: 1. These tubes now
require fiberoptic bronchoscopy for positioning5,6. 2.
A satisfactory right-sided DLT has not yet been designed, to deal
with the short (average 2cm) and variable length of the right
main stem bronchus7.
Recently, there has been a revival of interest in BB's due to
several factors: design advances such as the Univent tube8,
greater familiarity of anesthesiologists with fiberoptic
placement of BB's9, and new indications for OLV such
as lung transplantation, and cost10.
Indications for one lung ventilation
Since it is impossible to describe one technique as best in
all indications for one-lung ventilation (OLV), the various
indications will be considered separately, the discussion will
proceed from clinically common to less frequent situations.
1) Elective pulmonary resection, right-sided. This is the commonest adult indication for OLV. The first choice is a left-DLT: There is a wide margin of safety in positioning left-DLT's7. With blind positioning the incidence of malposition can exceed 20% but is correctable in virtually all cases by fiberoptic adjustment11. A partial resection can proceed to a pneumonectomy, if required, without loss of lung isolation. There is continuous access to the non-ventilated lung (NV-lung) for suctioning, fiberoptic monitoring of position, and continuous positive airway pressure (CPAP). There are differences in the designs of the bronchial cuffs of the three most widely used disposable left-DLT's in North America. These differences in cuff design result in different mean bronchial cuff inflation volumes and pressures during one-lung isolation12.
Possible alternatives are: a) Single lumen EBT. A standard
7.5mm, 32cm length endotracheal tube (ETT) can be advanced over a
fiberoptic bronchoscope (FOB) into the left mainstem bronchus13.
It is impossible to double-check the positioning without impeding
ventilation and there is no access to the NV-lung. b) Univent
tube or BB. The BB can be placed external to or intra-luminally
with an ETT9. BB's are often unstable in the short
right mainstem bronchus.
2) Elective pulmonary resection, left-sided:
i) Not pneumonectomy. There is no obvious best choice, between a Univent and a left-DLT. It has been described that, even in left thoracotomies, the Univent blocker may migrate into the trachea and require replacement with a DLT10. The use of a left DLT for a left thoracotomy is rarely associated with obstruction of the tracheal lumen by the lateral tracheal wall and subsequent problems with gas exchange in the ventilated lung (V-lung). A right-DLT is an alternate choice but problems with lung isolation and/or positioning occur in up to a third of cases with right DLT's14.
ii) Left pneumonectomy. Again there is no completely
satisfactory choice. Any left pulmonary resection may
unforseeably become a pneumonectomy. When a pneumonectomy is
forseen, a right-DLT is the best choice. In 2/3 of cases a
right-DLT will function appropriately and it will not require
repositioning when the mainstem bronchus is clamped. A right-DLT
will permit the surgeon to palpate the left hilum during OLV
without interference from a tube or blocker in the left mainstem
bronchus. This is occasionally required when there is some
question regarding the extent of a tumor.
The disposable right DLT's currently available in North
America vary greatly in design depending on the manufacturer
(Mallinkrodt, Rusch, Sheridan). The Mallinckrodt design is
currently the most reliable. A new Rusch design is undergoing
clinical trials. All three designs include a ventilating
side-slot in the distal bronchial lumen for right upper lobe
ventilation. Positioning this slot can be time-consuming. The
commonest problem with these tubes is that they require
relatively high bronchial intra-cuff pressures (40-50cmH2O
vs. 20-30cmH2O for left DLT's). However, this is lower
than the range of pressures required by a Univent15 or
non-disposable DLTs4. Rarely, left lung isolation is
impossible in spite of extremely high pressures in the right-DLT
bronchial cuff. In these cases a Fogarty catheter can be passed
into the left main bronchus after estimation of depth with a FOB.
The surgeon can then palpate the balloon of the Fogarty catheter
in the bronchus to aid positioning. This avoids replacement of
the DLT. As an alternative, there is no clear preference among a
Univent, left-DLT or other bronchial blocker. These will all
require repositioning intraoperatively, but this usually is not a
major problem. A right single-lumen EBT is almost certain to
obstruct the right upper lobe with an attendant increased risk of
hypoxemia during OLV.
3) Thoracoscopy:
A rapidly increasing number of thoracic procedures can be done
with video-assisted thoracoscopy. Lung biopsies, wedge
resections, bleb/bullae resections, even some lobectomies can be
done using this technique. Thoracoscopic surgery under general
anesthesia usually requires OLV.16 During open
thoracotomy the lung can be compressed by the surgeon to
facilitate collapse prior to inflation of a bronchial blocker.
This is not possible during thoracoscopy. The operative lung
deflates more easily when the NV-lung lumen of a DLT is opened to
atmosphere than via the 2mm suction channel of a Univent tube. A
left DLT is preferred for thoracoscopy of either hemi-thorax.17
Spontaneous ventilation without lung isolation is an alternative
in many patients.18
4) Pulmonary hemorrhage.
Instances of life threatening pulmonary hemorrhage can occur
due to a wide variety of causes (Aspergillosis, Tuberculosis, PA
catheter trauma, etc.). The anesthesiologist is often called to
deal with these cases outside of the operating suite. The primary
risk for these patients is asphyxiation, and first line treatment
is lung isolation. There are several problems associated with
using any sort of bronchial blocker in the acute situation: i) It
is often not known which side to occlude. ii) Visualization below
the vocal cords to aid placement is difficult. iii) After the
blocker is placed there is no access to the involved lung to
monitor bleeding. In patients with pulmonary hypertension,
endobronchial blockade can lead to lobar rupture from continued
bleeding.19 A left-DLT avoids these problems.20
Even if passed initially into the right mainstem bronchus, lung
isolation is achieved and the hypoxemia due to obstruction of
only the right upper lobe is usually not significant. After
suctioning and resuscitation, further diagnosis and therapy can
then proceed in a controlled fashion. Tracheo-bronchial
hemorrhage from blunt chest trauma will usually resolve with
suctioning, only rarely is lung isolation necessary.21
PA catheter induced hemorrhage during weaning from bypass should
be dealt with by resumption of full bypass, bronchoscopy, and
lung isolation. Weaning may then preceed without pulmonary
resection in some cases.22,23
5) Bronchopleural fistula.
The anesthesiologist is faced with the triple problem of
avoiding tension pneumothorax, ensuring adequate ventilation, and
protecting the healthy lung from the fluid collection in the
involved hemi-thorax. Management depends on the site of the
fistula and the urgency of the clinical situation. For a
peripheral bronchopleural fistula in a stable patient, some form
of BB such as a Univent tube may be acceptable. For a large
central fistula, and in urgent situations, the rapidest and most
reliable method of securing one-lung isolation and ventilation is
a DLT. In life threatening situations such as early postoperative
dehiscence of a mainstem bronchial stump, a DLT can be placed in
the awake patients with direct FOB guidance.24
6) Purulent secretions.
Lung abscess, hydatid cysts, pneumonitis distal to an
obstructing carcinoma, and a variety of other diseases require
protection of uninvolved lung from secretions during thoracotomy.
Lobar or segmental blockade is the ideal. Loss of lung isolation
in these cases is not merely a surgical inconvenience, but may be
life threatening. Univent tubes can be used for lobar blockade.
Manipulation of the rigid Univent blocker into lower or middle
lobes can be difficult. A time consuming but more secure
technique in these cases is the combined use of a bronchial
blocker and a DLT.25 A Fogarty catheter can be passed
through the vocal cords as a BB then followed and positioned with
a FOB in a DLT. This "belt and suspenders" technique
allows OLV for easy surgical access in the non-ventilated
hemithorax while protecting the uninvolved segments/lobes of the
ipsilateral lung.
7) Non-pulmonary thoracic surgery.
Thoracic aortic and esophageal surgery require OLV. Since
there is no risk of V-lung contamination, a left-DLT or a Univent
tube are equivalent choices.
8) Bronchial surgery.
An intra-bronchial tumor, bronchial trauma, or a bronchial
sleeve resection during a lobectomy require that the surgeon have
intra-luminal access to the ipsilateral mainstem bronchus. Either
a single lumen EBT or a DLT in the V-lung is preferred. Some
surgeons find the exposure easier with an EBT than a DLT.26
9) Upper airway abnormalities.
It is occasionally necessary to provide OLV in patients who have abnormal upper airways due to previous surgery, trauma, etc. The Univent tube may be useful in some of these patients. The smallest available Univent (#6) corresponds to an ETT with an internal diameter (ID) of 8.5-9.0mm and can be passed through the narrowed upper airway or tracheostomy of some of these patients. The measured external circumference of a #6 Univent tube is 34 Fr, in spite of the 26 Fr marking on the tube itself. Thus, it is not much smaller than the commonly used 35 Fr DLT. Smaller DLT's (28 and 26 Fr) are available and their measured circumferences correspond to the marking on the tubes. DLT's smaller that 35 Fr will not permit passage of a F0B of the diameter commonly used to monitor positioning (3.5-4.0mm). An infant FOB (2mm) or a flexible ureteroscope can be used with the small DLT's.
An ETT designed for microlaryngoscopy (5-6mm ID and >30cm
length) can be used as an EBT, with FOB positioning, in the left
main-stem bronchus for a right thoracotomy. For left thoracotomy,
if the patient's trachea can accept a 7.0mm ETT, a Fogarty
catheter (8/10 Fr venous thrombectomy catheter with a 4cc
balloon) can be passed through the ETT via a fiberoptic
bronchoscopy adapter for use as a BB in the left main bronchus. A
4mm FOB can then be passed in the ETT beside the BB.27
For BB manipulation, it is useful to bend the distal 2cm of the
Fogarty guide wire 45, then leave the guide in situ during
placement.
10) Unilateral lung lavage.
An infrequent indication for OLV is uni-lateral lung lavage
for primary alveolar proteinosis28. A left-DLT is the
only technique that permits adequate access to the NV-lung for
lavage while providing a secure enough bronchial seal to protect
the V-lung.
11) Independent lung ventilation.
This is not truly OLV. However, there are situations such as
unilateral lung trauma29,30 and post pulmonary
embolectomy where independent ventilation of the lungs is
required. A DLT is required for these patients. The
anesthesiologist is usually the only person with adequate
experience in lung isolation techniques to provide this service
for the intensive care unit. Since these DLT's will often be in
place for a prolonged period it is important to select the best
available DLT.12,31
12) Lung transplantation.
The pathology, age, size and complicating conditions of
transplant recipients vary so widely it is difficult to
generalize about the best techniques for OLV. In patients with
copious thick secretion suctioning via a DLT was a major initial
problem that lead to a trial of bronchial blockers. The use of
segmental bronchial lavage via a FOB/ETT after induction then
replacement with a DLT has overcome this problem.
Airway trauma
Tracheal or bronchial rupture is a recognized complication of
DLT's. The frequency of this complication has been estimated to
be between 0.5-2 per 1000 cases.32 The infrequent case
reports with disposable DLT's have usually been attributed to:
using nitrous oxide as an anesthetic33 or to using an
inappropriately large tube.34 Tracheo-bronchial
rupture due to airway catheters occurs most often at the junction
of at the posterior (membranous) and lateral (cartilaginous)
walls of the airway. In the majority of reported cases, the
diagnosis has been made intra-operatively and only rarely has the
problem become evident postoperatively. Intraoperative diagnosis
necessitates immediate surgical repair. When the diagnosis is
delayed until the postoperative period, conservative treatment
with fiberoptic surveillance and antibiotic coverage is an option
if: the tracheo-bronchial wound is small, there are minimal
symptoms, and the patient can be weaned from positive pressure
ventilation. Several strategies should be followed to minimize
the airway trauma due to endobronchial tubes/blockers:
1) The majority of difficult endobronchial intubations can be
predicted from viewing the chest x-ray.35 Pericarinal
distortions due to tumor, previous surgery/infections or other
intrathoracic structures (especially the ascending aorta) are
usually evident on the preoperative x-ray film; even though they
may not be mentioned in the radiologist's report. There is no
substitute for the anesthesiologist assessing the film
him/herself prior to induction. The CT scan may also reveal
bronchial obstruction not evident on the plain chest x-ray.36,37
Alternatives to the standard practice of DLT placement via
laryngoscopy will depend on the abnormal anatomy. Endobronchial
intubation under direct vision using the FOB as a guide is one
alternative. Other possibilities include: use of a DLT in the
contralateral main bronchus (eg. right instead of left-DLT) or
use of a bronchial blocker instead of a DLT.
2) Use of an appropriate size tube. Too small a tube will make
lung isolation difficult. However, too large a tube is more
likely to cause trauma. Useful guidelines for DLT sizes in adults
are: females height < 1.6m (63in.) 35Fr.; females > 1.6m
37Fr.; males < 1.7m (67in.) 39Fr.; and males 1.7m 41Fr.
Although tracheo-bronchial dimensions correlate with height, the
correlation is weak.38 At all times it is best to
defer to direct FOB airway assessment and use of a smaller size
tube if undue resistance is encountered during intubation. The
average depth at insertion, from the teeth, for a left-DLT is
29cm in an adult and varies 1cm for each 10cm of patient height
above/below 170cm.39
3) Avoid nitrous oxide as an anesthetic agent. Nitrous oxide
will diffuse into the cuff of any airway tube/blocker that has
been inflated with air. Nitrous oxide 70% can increase the
bronchial cuff volume from 5 to 16ml intraoperatively, a volume
which has been demonstrated to be capable of rupturing the
mainstem bronchus of a cadaver.40 If nitrous oxide is
used as part of the anesthetic either: inflate the bronchial cuff
with the same nitrous/oxygen mixture, or inflate the cuff with
liquid, or use a continuous cuff inflation
pressure-monitor/release-valve system.41
4) Inflate the bronchial cuff/blocker only to the minimum
volume required for lung isolation. This minimal volume can be
determined by observing for the cessation of leak from the
non-ventilated lumen of a DLT, or the central lumen of a Univent
blocker, to a 1cm underwater drain as the ventilated lung is held
at a static inflation pressure (eg. 30cmH2O).42
Only, in cases at exceptional risk for contamination of the
ventilated lung (eg. bronchopulmonary lavage, lung abscess, etc.)
is a higher level of static inflation isolation-pressure (45cmH2O)
required.
5) The risk of major contamination of the ventilated lung with
blood, pus, etc., occurs only in the minority of thoracic
patients. In most thoracic cases the bronchial cuff/blocker only
needs to be inflated during the period when OLV is required for
surgical exposure. It is a false impression that inflating the
bronchial cuff stabilizes the DLT position when the patient is
turned to the lateral position43.
6) Compared to the oral mucosa, the tracheobronchial mucosa is
less firmly attached to its supporting cartilage. Thus, it is
much less resistant to trauma from foreign bodies in the airway.
Endobronchial intubation must be done gently and with fiberoptic
guidance if resistance is met. However, the majority of case
reports of iatrogenic endobronchial trauma do not describe any
difficulty in initial tube placement. A significant number of
case reports are from cases of esophageal surgery.32,33.
In patients with malignancies, particularly in the area of the
carina, the elastic supporting tissue may be weakened and
predisposed to rupture from DLT placement.
The lack of reports of trauma with the Univent may be due to
its relative newness. The bronchial cuff inflation pressure
required to achieve occlusion in vitro was significantly higher
for a Univent (> 150mmHg) than for a DLT (< 80mmHg)15.
The risk of trauma to the trachea and vocal cords increases with
the size of the airway catheter. Anesthesiologists should be
aware of the misleading circumference markings printed on the
Univent tubes. These tubes are actually much larger than the Fr
size printed on the tube indicates.
Other complications
1) Malpositioning. Initial malpositioning of DLT's with blind
placement can occur in over 20%44 of cases.
Verification and adjustment with FOB immediately prior to
initiating OLV is mandatory since these tubes will migrate during
patient positioning.45 It is now possible to
indirectly monitor tube position continuously using side-stream
spirometry.46
2) Early studies of the Univent tube vary on whether it is
easier47 or more difficult48 to position
than a DLT. Malpositioning after the start of OLV due to
dislodgement is more of a problem with bronchial blockers than
DLT's.
3) Hypoxemia during OLV. CPAP is the first line treatment of
hypoxemia and can be provided with either the Univent or DLT.
When re-inflation/deflation is required during OLV it is easier
to maintain the CPAP with a DLT. CPAP cannot be used with a BB or
EBT. Malposition of DLT's checked only by auscultation is
responsible for a large fraction of hypoxemia during OLV.49
4) Airway resistance. There is a common misconception that the
Univent tube offers much less airflow resistance than a
comparable DLT. On this basis, it is felt by some
anesthesiologists that patients who require postoperative
ventila-tion must have a DLT changed to an ETT, but a Univent
tube can be left in place. We have measured the resistance to
airflows in the range of mean flows commonly seen during
spontaneous ventilation (5-60 l/min, unpublished data). The
resistance from a 37 Fr DLT exceeds that of a #9 Univent by <
10% over this range. These flow resistances are both less than a
8.0mm ID ETT but both exceed a 41 Fr DLT or a 9.0mm ETT. For
short periods of postoperative ventilation and weaning, airflow
resistance is not a problem with a DLT.
Summary
The preference of anesthesiologists has oscillated between
bronchial blockers and single or double-lumen tubes for the past
fifty years, and no overall best method of providing OLV has yet
been found. Each of the available methods of lung isolation has
its strengths and weaknesses. The choice of technique for each
individual case will depend on an interplay of several factors
including: the indication for OLV, concurrent patient problems,
available equipment and skill/training of the anesthesiologist.
All of these factors are continually evolving.50,51
References