Key words
emphysema - thoracic imaging - endoscopic lung volume reduction
Introduction
Chronic obstructive pulmonary disease (COPD) is one of the most common diseases worldwide,
having a prevalence of approximately 9 % in Germany [1]. The disease is diagnosed according to clinical and functional findings, and classified
by severity (GOLD stage 1 – 4) and into groups according to GOLD guidelines [2].
Even patients with similar degrees of impaired pulmonary function yield completely
different morphological pictures through radiological imaging, both in terms of the
extent ("quantifying") and the prevailing type of morphological changes ("qualifying").
Distinctions must be made between emphysema- and airway-predominant "phenotype", with
classification being made with the aid of computed tomography [3]
[4]
[5]
[6]. Imaging diagnostics have become increasingly relevant in recent years, since different
strategies exist for the two patient groups which open new perspectives particularly
for the emphysema group.
Modern emphysema therapy includes, above all, endoscopic lung volume reduction methods
(ELVR). In particular, valves and coils are seeing increased use, thus requiring radiologists
to become familiar with the different methods and diagnostic criteria.
Phenotyping COPD: Emphysema?
Phenotyping COPD: Emphysema?
The severity and course of COPD is determined by gathering the patient’s history and
testing lung function. Important parameters are forced exhalation (FEV1, forced expiratory
volume in one second), FEV1/FVC quotient (Tiffeneau index) and diffusion capacity.
The ITGV (intrathoracic gas volume) with RV (residual volume) and TLC (total lung
capacity) are used to detect hyperinflation. To ascertain their normal routine physical
capacity, patients undergo a standardized 6-minute walking test.
Severity staging according to GOLD is founded on the FEV1 limitation based on values
following bronchodilation (GOLD 1 FEV1 ≥ 80 %, GOLD 2 FEV1 < 80 % ≥ 50 %, GOLD 3 FEV1 < 50 %
≥ 30 %, GOLD 4 < 30 % of target). Patients are can additionally be classified into
groups A through D according to symptoms and exacerbations per year. However, the
morphological information indicating whether emphysema is present and, if so, to what
extent and how it is distributed can be obtained only through medical imaging. According
to the EvA study, it is possible to classify cases as E- (emphysema) or A- (airway
disease) type by measuring lung density and bronchial wall thickness [7], most cases being a mixed picture with varying degrees of both components.
Airway type
The chronic inflammatory process reshapes the major and minor airways.
The bronchial walls become thicker (chronic bronchitis) and either narrow (obstruction)
or widen (bronchiectasis), while the airways become unstable.
When the minor airways are involved (small airway disease), indirect and direct signs
of bronchiolitis are present. In the case of inflammatory (exudative, intraluminal)
bronchiolitis with thickening of the bronchiolar wall and displacement of the lumen
by secretion, centrilobular, micronodular densifications, which have the appearance
of buds on a branch (Tree-in-bud phenomenon) appear as direct signs. With constructive,
obliterative bronchiolitis, the minor airways are obstructed, the lumen of the bronchioles
collapse upon expiration and air can no longer be expelled (air trapping). This is
more clearly visible on expiration images when mosaic-like hypodense areas with trapped
air remain (indirect sign), while normally the density increases in healthy lung parenchyma
during exhalation. Comparison of inspiratory and expiratory images additionally provides
information on the stability of the tracheobronchial system.
Emphysema type
Emphysema is a morphological-structural diagnosis defined as irreversible expansion
of the air spaces distal to the terminal bronchioles with destruction of the walls
of the affected alveoli, but without any scarring ([Fig. 1]). Lung performance declines as a result of a loss of gas exchange surface as well
as a progressive hyperinflation. LVR measures can be used only for patients suffering
from an advanced stage (GOLD III through IV) depending on the type, extent and distribution
of their emphysema. It is therefore necessary to first address the classification
and subtypes of emphysema.
Fig. 1 Healthy a and abnormal b alveoli on an electron microscopic image. Emphysema (on the right) leads to irreversible
destruction of the alveolar walls, which results in a reduced area for gas exchange.
(Courtesy of PulmonX).
Radiological classification of emphysema
Radiological classification of emphysema
Anatomy
Emphysema is classified based on distribution within the secondary lobule, the smallest
connective tissue anatomical subunit in the lungs. The lobule is polygonal in shape,
measuring approximately 1 – 2.5 cm and is surrounded by connective tissue (interlobular)
septa [8]. Under normal, healthy conditions, the lobule can only be detected on CT through
inference. Running through the center of the lobule are the artery and the accompanying
bronchiolus. Each lobule comprises on average 12 acini, each acinus in turn containing
roughly 200 alveoli. While terms lobular and acinar are often used synonymously in
describing a type of emphysema, the reference to the lobule is more appropriate from
a radiological viewpoint, since only this part is visible.
There are three main types of emphysema [9]
[10]: centrilobular, paraseptal (formally known as distal acinar) and panlobular emphysema
([Fig. 2]). The Fleischner Society is currently developing a new method of classifying emphysema
based on severity [11]. While further descriptive terms exist, they do not constitute unique subtypes:
the (large) bullous emphysema, scar emphysema (vicarious emphysema) and a compensatory
overexpansion emphysema (e. g. following contralateral pneumonectomy). The term “elderly
emphysema” is considered to be obsolete and should not be used to describe the physiological
alveolar hyperinflation not involving destruction in the senile lung [12].
Fig. 2 Types of emphysema in relation to the secondary pulmonary lobule (HRCT sections).
a Centrilobular emphysema (CLE). The process starts centrilobular with small focal
lucencies without visible walls. The polygonal shape of some totally affected secondary
lobules is clearly visible. The artery is seen as a white dot in the center (arrow).
The interlobular septa build the margins containing pulmonary veins and lymphatics.
b Paraseptal emphysema (PSE). Pulmonary lobules are affected peripheral, paraseptal
and subpleural. There are lucencies in a single layer with focal hyperinflation, in
this case subpleural along the dorsal circumference. c Panlobular emphysema (PLE) due to alpha-1-antitrypsin deficiency. This type is characterized
by uniform destruction of the entire pulmonary lobule with homogenous involvement
of large parts of the lungs, mainly lower lobes. Normal parenchyma is barely visible,
vessels are rarefied, lung is hyperinflated.
Centrilobular emphysema
Centrilobular emphysema (CLE, adopted from the Fleischner Society) is the most common
type. It appears mainly in smokers, affecting primarily the superior lobe. The process
begins in the center of the secondary lobule. On CT images it morphologically appears
as round density reductions without a wall that look like small "holes" in the surrounding
initially still healthy lung parenchyma. This finding progresses toward the periphery
until the entire secondary lobule is involved. In advanced cases, multiple lobules
are affected, which can progress to confluent involvement of the entire lobe and extensive
parenchymal destruction ([Fig. 3]).
Fig. 3 Centrilobular emphysema of various severities. a Mild, with some small spotty centrilobular areas of emphysema, about 0.5 to 5 % of
the upper lobe, surrounded by normal parenchyma. b Moderate, with less affected regions scattered next to larger areas of destruction.
c Confluent, with centrilobular and lobular parts, without relevant hyperexpansion,
or destruction of the lung architecture. d Advanced destructive emphysema with marked hyperextension and destruction of the
lung anatomy not caused by alpha-1-antitrypsin deficiency (derived from Fleischner
Society).
Paraseptal emphysema
With paraseptal emphysema (PSE) the peripheral, subpleural lobules are selectively
affected. Density reductions and hyperinflations are visible along the pleura. The
superior lobes are affected above all, e. g. resulting from scarring post-infectious
shrinkage processes. This type of emphysema appears even in young nonsmokers and is
frequently the cause of spontaneous pneumothorax.
Panlobular emphysema
Panlobular emphysema (PLE) is typical in patients with alpha-1 antitrypsin deficiency
and the term is reserved for this condition (for cases of advanced, originally centrilobular
emphysema, the term confluent emphysema is more appropriate)
The entire lobule exhibits relatively homogeneous, panlobular destruction. In this
context, the “pan” refers to the secondary lobule (pan-lobular) not to the lobe. This
is important to note, since the terms lobule (the small subunit) and lobe (the pulmonary
lobe) should not be confused in everyday clinical activity. The lobules are inflated
and the blood vessels rarefied. Distribution within the lobes is not localized or
focal, but rather uniform and usually extensive. It is the inferior lobes that are
primarily affected.
Bulla
A bulla is a sharply demarcated air-filled lesion greater than one centimeter in diameter
(smaller lesions are called "blebs" = bubbles) with a delicate wall measuring less
than one millimeter thick [13]. The disease is referred to as bullous emphysema when bullae are prominent. Giant
bullae can measure more than 10 cm and take up a large volume, while compressing the
surrounding lung parenchyma.
What every radiologist should know about emphysema therapy
What every radiologist should know about emphysema therapy
Standard therapy for COPD, comprises, in addition to strict smoking cessation, medications
such as bronchodilators and, if necessary, steroids, physical therapy, rehabilitation
measures, later long-term oxygen therapy and, if necessary, non-invasive ventilation.
However, the lung tissue already transformed by emphysema is irreversibly damaged.
Due to the absent gas exchange surface and progressive dynamic pulmonary hyperinflation,
patients in advanced stages suffer from serious dyspnea even under the most minor
physical strain. In addition there are various (above all cardiovascular) comorbidities.
Only after maximum conservative therapy has been exhausted, is a lung volume reduction
(LVR) discussed as final stage therapy.
Principle of lung volume reduction
The idea of LVR, according to which the non-functional or pathologically hyperinflated
“bloated” portions of the lung are (surgically removed) or deactivated, has existed
as form of emphysema therapy for years. The objective is to achieve decompression
to restore breathing space for the less affected portions of the lungs. The flattened
diaphragm, which has been pressed downward, can recover and regain its rounded form.
Breathing mechanics are boosted, the elastic resilience of the lungs improved and
dyspnea decreased.
Surgical LVR
Surgical LVR was performed as early as the 1950 s. At that time, the surgery was in
no way equivalent to a simple bullectomy. The high peri- and postoperative mortality
initially prevented the procedure from becoming established. However, it became the
subject of research once again in the 1990 s. The NETT study compared bilateral surgical
LVR (usually performed by means of median sternotomy) with drug therapy between populations
of 608 and 610 patients, respectively [14]. The 90-day mortality rate for the surgical group was clearly elevated at 7.9 %
versus 1.3 %, especially in cases of homogenous emphysema and forced exhalation (FEV1)
values below 20 % of the target. On the other hand, the patients with superior lobe
emphysema and not capable of performing physical exercise benefited.
Subsequently developed minimally invasive, bronchoscopic LVR approaches are currently
the subject of intensive study and examination [15]
[16]
[17]. In the wake of encouraging results, focus has shifted back to surgical treatment
of emphysema (LVRS, lung volume reduction surgery) following years of restrained use.
Video-assisted thoracoscopic surgery (VATS) and median sternotomy exhibit comparable
efficacy and safety, while involving a shorter hospital stay than VATS [18]. Study approaches are “staged VATS” (in which both side undergo two consecutive
unilateral surgical interventions), a “non-resectional” LVR (clamping without actual
resection of lung tissue) or laser resection. Thus far, no results are available for
direct comparison between LVRS and endoscopic LVR, and the significance of LVRS is
currently not clearly defined.
Endoscopic lung volume reduction (ELVR)
Endoscopic lung volume reduction (ELVR)
Over the last 10 years different bronchoscopic methods ([Fig. 4]) aimed at collapsing or shrinking the emphysematous lung tissue have been developed
These methods are an option only for patients with advanced stages of the disease
and require an exhaustive diagnosis of lung function. Derived from the VENT-study
[19], the inclusion criteria comprises, among other factors, stage 3 (through 4) COPD
according to GOLD with high-grade pulmonary hyperinflation (residual volumes over
200 %). In principle, distinction must be made between blocking and non-blocking as
well as between reversible and irreversible procedures. Currently established methods
are compared in [Table 1].
Fig. 4 Valves and coils. a Endobronchial Zephyr® valve (courtesy of PulmonX). b Intrabronchial Spiration® valve (courtesy of Olympus). c RePneu coils (courtesy of PneumRx). Bronchoscopic use is demonstrated in the bottom
row.
Table 1
ELVR- Comparison of valves and coils.
|
valves
|
coils
|
|
technique
|
one-way valves allow air to leave but not enter the lung, causing a collapse of the
target lobe;
dynamic hyperinflation is reduced
|
shape-memory coils with mechanical action and tissue compression;
elastic recoils are restored, air trapping is reduced
|
|
method
|
blocking
|
non-blocking
|
|
reversibility
|
reversible, removable
|
questionable
|
|
before ELVR
|
|
type of emphysema
|
heterogenous emphysema
|
both heterogeneous and homogeneous emphysema
|
|
target area
|
important
|
less important
|
|
fissures
|
integrity is crucial
|
not relevant
|
|
collateral Ventilation (CV)
|
CV is the main problem
|
non interfering
|
|
contraindication
|
large parenchymal destruction
|
pulmonal hypertension, giant bullae,
large parenchymal destruction
|
|
after ELVR
|
|
atelectasis
|
atelectasis should be achieved
|
no atelectasis
|
|
residual parenchyma in the target lobe
|
gets lost
|
is preserved
|
|
pneumothorax
|
often
|
rare
|
|
hemoptysis
|
rare
|
often
|
|
migration
|
possible
|
rare
|
Valves
Via the working canal of a bronchoscope, multiple valves are implanted in the segmental-,
if necessary subsegmental bronchi as well, of a pulmonary lobe. Placement of a valve
takes only a few minutes. A valve mechanism then prevents new air from flowing into
the lungs, while still allowing the "old" air to flow out, thus collapsing the pulmonary
lobe ([Fig. 5]). The goal is complete atelectasis of the treated lobe. As of now two different
types of valves for this blocking, reversible method are available on the European
market.
Fig. 5 Imaging before and after ELVR with valves. a HRCT before implantation of valves. b HRCT after implantation of valves. Afer placement of 4 valves into the left lower
lobe the patient developed a lobe atelectasis as expected (thick white arrow). This
section shows one of the valves in a good position with the “narrow neck” at the proximal
end (black arrow). On the right side fissural and mediastinal shift is visible (thin
white arrows). c Chest radiography pre-treatment (same patient). Normal middle position of the mediastinum,
depression and flattening of the diaphragm. d Chest radiography after implantation of valves. On the left hilar region the filigran
mesh of the valve is hardly seen on hard copy films (black arrows) but good on monitors.
Left lobe atelectasis and mediastinal shift to the left (thick white arrow). The actual
level of the left hemidiaphragma is slightly more apical as a sign of reexpansion
(thin white arrow).
Endobronchial valves (EBV)
EBV (Zephyr® valves from PulmonX) comprises a silicone body surrounded by a nitinol wire cage
([Fig. 4a]). The end pointing in the proximal direction (toward the trachea) has a narrower
diameter (“neck”) with a “fish mouth” valve. The initial results were published in
2003 [20]. The data of the prospective VENT study in 2006 [29] initially showed only moderate benefits. In the meantime, the critical importance
of patient selection has become clear. According to the EuroVent-Study [21], predictors of successful EBV therapy are a heterogeneous lung emphysema, correct
placement and intact fissures.
Intrabronchial valves (IBV)
IBV (Spiration® valves from Olympus) resemble small umbrellas ([Fig. 4b]). They are anchored via the distal feet (“anchoring hooks”), and the leaves attach
to the walls of the bronchia. At the proximal end is a type of “stem” by which the
valve can be gasped for removal if necessary. The unilateral complete occlusion of
a lobe appears to be superior to a bilateral only partial occlusion [22].
According to user instructions, both types of valves are conditionally safe with MRI,
i. e. a patient with this implant can safely undergo an MRI examination with a static
magnetic field up to 3 Tesla [23]
[24].
Metal spirals, “coils”
Wire spirals made of nitinol (a biocompatible nickel/titanium alloy) with a shape
memory (RePneu® Coils from PneumRx) are used ([Fig. 4c]). With the aid of a special implantation system, the wires measuring roughly 10
to 15 cm long are first introduced into the segmental bronchia under fluoroscopy,
usually 10 units (if necessary more) being implanted in the selected lobes. Upon being
released, the wires resume their original spiral shape ([Fig. 6]), thereby pulling the bronchia together with the lung tissue to the hilum through
mechanical force. A certain compression of the pulmonary parenchyma is achieved, the
actual goal being the improvement of the elastic restoring forces. These non-blocking,
debatably reversible methods do not result in atelectasis. The latest results show
this method to be effective on heterogeneous and homogenous emphysema [25]. The product received the CE label in October 2010 and is MRI-compatible up to 3.0
Tesla [26].
Fig. 6 Chest radiography before and after ELVR with coils. a Pre-ELVR. b After treatment with coils in the left superior lobe. c After treatment with coils in the superior lobes, bilateral.
Additional methods: Vapor, gel and stents
Other irreversible, non-blocking methods include the introduction of hot water vapor
(bronchoscopic thermoablation) or hydrogel foam (“bronchial adhesives”, polymer LVR
or BioLVR) to systematically induce inflammatory stimulation in the most intensely
affected area of the lung [27]. The goal is an acute, but “controlled” inflammatory process with subsequent shrinkage
of the tissue through scarring or fibrosis. Although a meta-analysis [28] showed the best results for BioLVR, this method is currently no longer available.
Thermoablation is still considered to be experimental, since there have been no major
studies [29]. Endoscopic creation of artificial airways by means of needle perforations (airway
bypass) has been suspended due to complications (EASE-study) [30].
What pulmonologists expect from radiological imaging and reporting?
What pulmonologists expect from radiological imaging and reporting?
X-ray/CT/MRI?
According to the guidelines of the German Respiratory Society (Deutschen Gesellschaft
für Pneumologie), a chest X-ray in two views should be performed for general diagnostics
when COPD is initially diagnosed. The validity of the criteria for emphysema already
defined in 1965 was confirmed once again by Miniati et al. 2008 [31]. We routinely see flat, low-lying diaphragmatic arches, increased lung transparency,
and enlarged retrosternal space as well as expanded intercostal spaces, a barrel chest
and, as the case may be, a teardrop-shaped, narrow heart silhouette. However, when
dealing with mild forms of emphysema, this method is not sufficient for determining
distribution and is in no way suited for testing whether ELVR is indicated.
The lungs present a challenge for MRI, since they have a very low proton density compared
to the brain, liver or musculature, and the margins between the air and parenchyma
lead to susceptibility artifacts. Breathing and heartbeat additionally create motion
artifacts, thereby necessitating rapid sequences, parallel imaging as well as breathing
and EKG triggering [32]. In the majority of cases, there is agreement between MRI and CT in classifying
emphysema and evaluating the severity thereof [33]. On the other hand, the morphological and functional magnetic resonance imaging
of COPD with visualization of ventilation, breathing dynamics and perfusion is becoming
increasingly prominent [34]. Dynamic MRI during continued breathing is excellent at showing how severely diaphragmatic
mobility is restricted is cases of emphysema [35]. The phenotyping of COPD using MRI and low-dose CT is currently being compared in
a multicentric, national cohort study (Cosyconet) [36]. However, the lengthy examinations and long periods of having to hold their breath
still remain difficult for patients who already suffer from dyspnea due to their underlying
disease.
Without a doubt, CT of the lungs is currently the method of choice. Pulmonologists
expect a non-contrast, high-resolution spiral CT taken during inspiration. If disease
of the minor airways is suspected, sequential scans during expiration should additionally
be performed.
The technical prerequisite is a 3 rd generation or above multi-detector CT with at
least 16 lines, the standard being 64-lines (if necessary 40 lines). According to
definition, slice thickness for a HRCT should be less than 1.5 mm [8]. A slice collimation below 1 mm (depending on manufacturer, e. g. 0.6 mm) with as
low of a rotation period as possible (≤ 0.5 s), an increment of 0.7 and an overlap
of 30 – 50 % is recommended. In CT protocols of the thorax there are considerable
differences in current exposure time product (40 mAs to 200 mAs) [37]. Low-dose data sets with a current exposure time product of just 30 to 50 mAs allow
both a visual and quantitative evaluation of the emphysema [38]. In contrast, a higher dosage is necessary for the analysis of the airways particularly
in the case of software-assisted evaluations. For example, the median effective current
exposure time product in the COPD Gene Study was 180 mAs [39]. In our scan protocol, tube current was 100 mAs and tube voltage 120 KV. To reduce
radiation exposure, automatic dosage modulation was employed. Iterative image reconstruction
techniques are also highly promising. While the image quality suffers when dosage
is reduced by 50 % and data sets are reconstructed by means of filtered rear-projection
(for example, because of the increased noise pattern the differentiation between residual
lung parenchyma and bulla was compromised), iterative reconstruction technique allows
a comparable image quality to be achieved at half the radiation dose [40].
Generating sagittal and coronal 3 D reconstructions is especially important, with
reconstructed slice thickness of ≤ 1.5 mm also being ideal in this case. A differentiated
evaluation of findings focusing on the following diagnosis- and therapy-relevant criteria
is required.
Quantification of the emphysema
In routine clinical practice, the individual patient's complaints do not accurately
foreshadow the morphology or severity of the emphysema revealed through CT, although
the extent of the emphysema correlates with the severity of COPD at least in cases
of centrilobular and panlobular, but not paraseptal type. In addition to the purely
quantitative description in the report, an assessment that is as standardized, objective
and reproducible to the extent possible is required. Lung density measurements using
Hounsfield units (HU) and an estimate of the extent of emphysema as a percentage (e. g.
0 %, below 5 %, up to 25 %, up to 50 %, up to 75 %, over 75 %) are helpful in this
respect [42]. Visual assessment is performed separately for the individual lobes of the lungs.
The examiner-dependent variability remains problematic [41]
[42]
[43].
CT-densitometry can be performed semiquantitatively or quantitatively (computer-assisted)
[44]. Evaluation is reportedly made easier with specialized software which shows emphysematous
areas marked in color (e. g. as color map) on the basis of a prescribed density limit
value and generates a table summarizing density values. The average lung density (expressed
in HU) can be computed from the density values of all lung voxels, while segmentation
of the lung margins is fully automatic. The emphysema index can be computed as a quotient
of emphysema volume and lung volume (expressed in percent). Several programs are already
commercially available ([Fig. 7]). However, there are major differences in the different types of software, among
which are the evaluation and presentation of data. While the results were still disappointing
in 2006, e. g. the considerable amount of time required and the poor correlation between
human eyes and the machine [45], by 2014 it was more than clear that qMDCT (CT densitometry using multidetector-row
computed tomography) represents a diagnostic gain [46]. However, it remains unclear how strongly patient-related factors (age, inspiration
depth) influence the measured values. In addition, a uniform threshold value for density
measurement has not yet been established (initially -910 HU, then -970 HU, now usually
-950 HU) [47]. Moreover, the computer-based segmentation of the individual lobes of the lung is
still not reliable [48], and areas with increased density such as dystelectases and infiltrates are problematic.
Serious differences in terms of evaluation appear depending on the CT scanner used,
the reconstruction parameters and, above all, the software manufacturer [48].
Fig. 7 Software analysis of emphysema. Areas with low attenuation are depicted in orange
indicating centrilobular emphysema. Upper image shows axial HRCT scan, 3D-rendering
is seen in lower image. a Heterogeneous distribution, all lobes are affected, no explicit predominance. b Marked predominance of emphysematous regions in both upper lobes, possible target
areas for ELVR. (Courtesy of Fraunhofer MEVIS, Bremen).
The indication of the “PD 15" value – the lung density of the 15th percentile when
showing the relative frequency of all measured voxel densities in a histogram – is
highly promising. With emphysema, the low density values shift the HU distribution
curve to the left and into the negative region. The EXACTLE study examining alpha-1
antitrypsin deficiency demonstrated that CT densitometry can represent a valid endpoint
in a longitudinal study [49]. Unfortunately, no broadly available, reproducible standard has yet been established
for software-based diagnosing of emphysema.
Target area
Only with the aid of radiology can the distribution of emphysema and the heterogeneity
thereof be evaluated. A multidisciplinary approach must be employed to determine whether
or not a target area exists. Emphysema being prominent in a lobe of the lung with
hyperinflation or disproportionately high volume, e. g. visible in the form of a displacement
of the lobe fissures and mediastinum, would be ideal. In contrast, giant bullae or
an excessively extensive destruction of the parenchyma if anything impede endoscopic
measures, since the presence of consecutively larger hollow spaces poses the risk
of uncontrolled tissue tearing.
The quantity and quality of the residual parenchyma are not irrelevant. The idea behind
the valves is to remove all air from the target lobes and achieve atelectasis of a
lobe to reduce as much volume as possible. In this process, however, the lobe is completely
“deactivated”, and its remaining parenchyma is no longer available for gas exchange.
With the coils, in contrast, the residual tissue is “preserved” and can continue to
function for gas exchange, since the target lobe remains ventilated. In cases of homogenously
distributed emphysema, valves are at least not recommended. In addition, pre-interventional
lung perfusion scintigraphy is recommended for estimating the loss of function following
treatment of the target area. MRI of lung perfusion can also be expediently used to
diagnose diseases of the airways and lung parenchyma [50]. For planning LVR, the areas of the lungs with impaired function can be visualized
with the advantage of higher spatial resolution, combined imaging of morphology and
function as well as absence of radiation exposure. As a result, this method can also
be used to assess the course of the disease following LVR.
Fissure analysis: The interlobular fissures
CT examinations performed following valve implantations have shown that the desired
atelectasis is frequently not achieved, can have a delayed onset or is only temporary
[51]. The main reason for this would be what is referred to as collateral ventilation
(CV), i. e. even if the valves are placed correctly and good occlusion is established,
a retrograde reventilation of the treated lobe occurs via parenchymal bridges from
the neighboring lobe [52]. This process hinders the actual value reduction. It is now known that CV is an
important selection criterion. Radiology provides answers by allowing analysis of
the interlobular fissures ([Fig. 8]). A normal left lung has only one fissure, while the right has an oblique fissure
(lower fissure) and horizontal fissure (upper fissure, running horizontally between
the superior lobe and middle lobe). Visualizing the pulmonary fissures as continuous,
sharp lines requires a collimation of 0.5 to 1.0 mm [53] along with sagittal and coronal multiplanar reconstructions. A radiologically imaged
continuous “complete” fissure is an indirect sign of an absent or minor CV. In a small
study involving 25 patients, 20 of 21 pulmonary lobes having collateral ventilation
also showed a defect in the fissure, with sensitivity being 95 %. However, specificity
was only 44 %, i. e. 7 of 16 evaluated lobes without collateral ventilation had no
fissure interruption on CT [54].
Fig. 8 Fissural analysis as a selection criterion for ELVR (HRCT, sagittal reconstruction).
a Complete fissure. Visualization of the interlobular fissure as a contiguous sharp
fine line, no obvious parenchymal bridges between superior and inferior lobe, thus
no clear sign for collateral ventilation in this section. b Incomplete fissure. The fine interlobular line is disrupted in the apical portions.
The course of the fissure is no longer visible (arrow). Broad parenchymal bridges
between upper lobe and inferior lobe, collateral ventilation is very likely. c Absent atelectasis after ELVR. Five endobronchial valves were implanted in the left
inferior lobe. In this CT scan four of them are visible (arrows). A complete atelectasis
of the inferior lobe was not achieved, probable retrograde ventilation due to collateral
ventilation.
What constitutes a complete or incomplete fissure remains problematic, as neither
has yet been uniformly defined in terms of CT morphology. As a result, the reported
occurrence of incomplete fissures greatly fluctuates in the literature between 20 %
and 87 %. Automatic fissure analysis methods are currently being developed [55].
A computer-assisted analysis of 573 CT-examinations recently showed that roughly 90 %
of all examined persons have incomplete fissures regardless of whether COPD is present
or not [56]. For the individual fissures, integrity was around 82 % for the oblique fissures
(bilateral) and 62 % for the horizontal fissure, regardless of severity of COPD. Integrity
of 90 – 100 % was defined as “complete” (continuous, intact fissure) and was present
in only a quarter of patients, left 25 %, right 26 % (oblique fissure) and 14 % (horizontal
fissure). Contradicting these findings, an analysis of 250 CT images with a by far
higher portion of intact fissures showed the left fissure to be incomplete in only
24.4 % of cases (thus appearing continuous in three quarters of all patients), and
the right fissure incomplete in 35 % of cases. [57].. There is a consensus regarding at least the horizontal fissure, which very frequently
appears to be uninterrupted, as well as regarding the perihilar region, which is more
difficult to evaluate and is where parenchymal bridges tend to form. The clinical
significance of “small” parenchymal bridges (appearing in approximately one third
of patients [58]) as well as accessory fissures (appearing in 16 % of patients, primarily in the
superior lobes [57]) on CV has not been clarified.
Collateral ventilation can also be measured through bronchoscopy using a balloon catheter.
This can be performed in a procedure immediately prior to ELVR. A reduction of lung
volume can be forecast with an accuracy of approximately 75 % using the Chartis® evaluation system. However, this method does not replace CT. A prospective study
is currently being conducted to determine whether the clinical benefit can be predicted
when a homogenous emphysema with intact fissure is diagnosed using CT [60].
Overall, fissure analysis using CT must be viewed cautiously for the time being. In
addition to a thorough scanning in all three planes, a uniform classification is urgently
required for defining an intact fissure and assessing collateral ventilation.
Routine clinical experiences with ELVR
Routine clinical experiences with ELVR
For adequate treatment of emphysema, radiologists and pulmonologists must work together
as a team. While careful patient selection on the basis of clinical findings and imaging
is critical, it is not yet sufficiently practiced at all intuitions. In the meantime,
the results following ELVR are increasingly coming to our attention in cases of, for
example, valves becoming dislocated or having to become removed due to complications.
Among the frequent complications following ELVR are exacerbation of COPD and infections,
including pneumonia. Caution is thus urged for patients with hypersecretory bronchitis,
as the valves can be displaced by the secretion. Before undergoing ELVR, patients
should be free of bronchopulmonary infection, have no serious concomitant disease
and be mobile and clinically stable. The patient must also be able to tolerate a pneumothorax
in the pulmonologist's medical opinion, since the occurrence thereof must be anticipated
as a result of the at times enormous tractive forces following ELVR. On radiological
images, these forces are indicated by post-interventional migration of the fissures
and the mediastinum. Extensive tissue destruction and giant bullae can be problematic
prior to ELVR, constituting a contraindication for the insertion of coils. Instead
of valves, surgical LVR would be a better option to discuss. Because of the risk of
bleeding, coils are also contraindicated for patients on anticoagulation therapy or
suffering from pulmonary hypertension. Post-interventional hemoptysis is typical following
the implantation coils, yet is usually easily managed. In addition to patient selection,
good management of complications in a center is the key to success.
Summary
Today, the implantation of valves or coils is well established at many centers. Technically
relatively simple to perform, ELVR remains a purely symptomatic treatment for carefully
selected advanced stage patients, which should ideally be performed at experienced
centers and tracked in studies. Because the primarily promising therapy procedure
prompts a lack of caution, an interdisciplinary discussion among pulmonologists, thoracic
surgeons and radiologists is key for optimized patient selection. Finally, a checklist
was created which is intended to aid during radiological diagnostic testing before
and after ELVR ([Tab. 2]).
Table 2
Checklist for Imaging for ELVR.
|
checkpoint
|
|
note
|
|
image aquisition
|
high-resolution multidetector CT of the lungs without contrast
spiral technique, contiguous, slice thickness 1.0 to 1.5 mm
3 D reconstructions in axial, coronal and sagittal orientation
submission of images via CD/DVD
|
|
phenotyping
|
airway-type
|
bronchitis? bronchiolitis? mucus? bronchiectasis? instability of airways?
|
|
emphysema-typ
|
centrilobular, paraseptal, panlobular
giant bullae?
|
|
extension: mild, moderat, confluend, large destructions
if available: specified as a percentage
if available: automatic quantitative assessment
|
|
distribution: homogeneous, heterogeneous
if available: heterogenity-/emphysema-index
|
|
before ELVR
|
target area
|
yes (where?)/no
|
|
fissure-integrity
|
major and minor fissur: contiguous line? gaps?
careful inspection in all directions including MPR
|
|
contraindications
|
pulmonal hypertension? pneumonia?
|
|
relevant pathologies
|
Lung cancer, pleural thickening, pleural effusion, …
|
|
after ELVR
|
ELVR-device
|
valves or coils visible?
number of devices compliant to anamnesis?
implantation in correct side/ correct segment bronchi?
occlusion of the bronchi? migration/aspiration/dislocation?
correct position? EBV: narrow end (“neck“) points to the proximal
IBV: the “5 arm anchor tip” points to the distal
coils: about 2 cm distance to pleura
|
|
LVR-effect
|
volume reduction effect?
atelectasis?
mediastinal shift? fissural shift?
reexpansion of the diaphragma?
|
|
complications?
|
pneumothorax? infiltration? hemorrhage?
|
|
new target area?
|
yes/No
|
