Download Visualizing lung function with positron emission tomography

J Appl Physiol 102: 448 – 458, 2007.
First published October 12, 2006; doi:10.1152/japplphysiol.00763.2006.
Invited Review
HIGHLIGHTED TOPIC
Physiological Imaging of the Lung
Visualizing lung function with positron emission tomography
R. Scott Harris1 and Daniel P. Schuster2
1
Pulmonary and Critical Care Unit, Department of Medicine, Massachusetts General Hospital,
Boston, Massachusetts; and the 2Department of Internal Medicine and the Mallinckrodt
Institute of Radiology, Washington University School of Medicine, St. Louis, Missouri
ventilation; perfusion; tracer; permeability
THE PURPOSE OF THIS PAPER is to describe the technology of,
applications for, and information obtained from positron emission tomography (PET) in the study of lung function. The
range of applications of PET and the variety of information that
has been obtained have expanded significantly in the past
decade, and much of this new material has not been included in
previous reviews (22, 23, 45, 53, 70).
PET has very high sensitivity, such that picomolar concentrations of labeled compounds can be readily detected in body
tissues. This allows for the ability to image radiolabeled
substances that target gene expression or cell surface receptors.
Current technology allows for a resolution of ⬃1 cm3 in the
lungs, a size that would be expected to contain a few acini. The
short half-lives of many of the positron-emitting compounds
helps keep radiation exposure to subjects low. This, coupled
with the desirable biological and chemical properties of many
PET radionuclides, has made it an ideal technology to study
lung function. A summary of methods and compounds used to
study lung function is given in Table 1.
A positron, emitted from a decaying radionuclide, travels a
few millimeters in tissue and then combines with an electron in
what is termed an “annihilation event,” yielding two ␥-rays of
equal energy (511 keV) at 180° from each other. These ␥-rays
then travel outward in opposite directions until they impact
detectors opposite each other in a ring around the camera. The
␥-rays are detected when they reach a scintillator material in
the scanning device, creating a burst of light that can be
detected. These electronically recorded coincidences eliminate
Address for reprint requests and other correspondence: R. S. Harris, Assistant Professor of Medicine, Massachusetts General Hospital, Pulmonary and
Critical Care Unit, Bulfinch 148, 55 Fruit St., Boston, MA 02114 (e-mail:
[email protected]).
448
the need for collimation, since the source of the ␥-rays occurred somewhere along the path between the two detectors.
However, the ␥-rays are subject to a small amount of scatter,
which contributes to noise. Another source of noise is random
coincidences, where two ␥-rays are detected simultaneously
but whose source was different. The contribution to noise of
these two effects is estimated to be ⬍5% of total coincidences.
The ␥-ray energy can be transferred to body tissues due to
interactions with atoms along their path. Because of these
interactions, the radioactivity of the source will be underestimated. An important advantage for PET is that this attenuation
can be corrected for using an image of regional density called
a transmission scan. This is performed by having a radioactive
positron source (usually 68Ge/68Ga) rotate around the subject.
A scan is also acquired without the subject in the camera, and
the difference in counts between the two scans represents the
reduction in counts due to tissue density. From the raw data,
corrected for energy attenuation by body tissues, one can use
reconstruction algorithms similar to those used for computed
tomography (CT) scanning to generate axial slices or a threedimensional image of the region of interest. The resulting
image is called an emission scan to denote that the source of
the radioactivity was within the body, as opposed to the
transmission scan, where the source is outside the body.
The resolution of PET systems is dependent on the detector
size, the detector coding process, annihilation photon acollinearity, positron range, and reconstruction algorithm. It is
usually quantified by a term called full-width half-maximum,
which is the width of an imaged point source at one-half of its
maximum value. As PET resolution improves (mainly due to
improvements in detector design), sensitivity can be degraded,
since the sensitivity depends on the number of annihilation
events per voxel and improved resolution means smaller voxel
8750-7587/07 $8.00 Copyright © 2007 the American Physiological Society
http://www. jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on May 5, 2017
Harris RS, Schuster DP. Visualizing lung function with positron emission tomography. J Appl Physiol 102: 448 – 458, 2007. First published October 12, 2006;
doi:10.1152/japplphysiol.00763.2006.—Positron emission tomography (PET) provides
three-dimensional images of the distributions of radionuclides that have been inhaled or
injected into the lungs. By using radionuclides with short half-lives, the radiation
exposure of the subject can be kept small. By following the evolution of the distributions of radionuclides in gases or compounds that participate in lung function, information about such diverse lung functions as regional ventilation, perfusion, shunt, gas
fraction, capillary permeability, inflammation, and gene expression can be inferred.
Thus PET has the potential to provide information about the links between cellular
function and whole lung function in vivo. In this paper, recent advancements in PET
methodology and techniques and information about lung function that have been
obtained with these techniques are reviewed.
Invited Review
449
VISUALIZING LUNG FUNCTION WITH PET
Table 1. Summary of functional PET methods
Q̇Cv៮ ⫽ V̇Aeff共13N兲CA ⫹ ␭NQ̇CA
Ĉ共t兲 ⫽ Afe⫺␣ft ⫹ Ase⫺␣st
Physiological
Parameter
Gas content
V̇A
Q̇
V̇A/Q̇
Blood volume
Plasma volume
PTCER
Inflammation
Gene expression
Compound and Method
Tracer Half-life
Ge/68Ga rotating pin source
(transmission scan)
Bolus injection 13N-N2-saline
13
N-N2 gas equilibration washout
19
Ne gas equilibration washout
81m
Kr gas equilibration washout
Bolus injection 13N-N2-saline
68
Ga-labeled particles
Bolus injection 13N-N2-saline
Continuous infusion 13N-N2-saline
Continuous infusion H215O
Inhalation of 11CO
Inhalation of C15O
11
C methyl albumin injection
68
Ga-transferrin injection
11
C-methylalbumin injection
[18F]fluorodeoxyglucose injection
Intratracheal delivery of
adenovirus carrying the
mHSV1-tk gene, followed by
intravenous administration of
9-(4-[18F]-fluoro-3hydroxymethylbutyl)guanine
270.8 days/67.6 min
68
9.97 min
9.97 min
17.4 s
13 s
9.97 min
67.6 min
9.97 min
9.97 min
2.04 min
20.4 min
2.04 min
20.4 min
67.6 min
20.4 min
110 min
110 min
V̇A, alveolar ventilation; Q̇, perfusion; PTCER, pulmonary transcapillary
escape rate; mHSV1-tk, mutant variants of the herpes simplex virus type I
thymidine kinase.
volume. To improve signal to noise, one can lengthen the
image collection time, increase the dose of administered tracer,
or filter the image. Therefore, there is a trade-off between
resolution and sensitivity, such that spatial resolution may be
killed to enable more accurate quantification, or vice versa.
MODELING AND VALIDATION
(1)
where Cv៮ and CA are the mixed-venous and alveolar concentrations of tracer, respectively, and ␭N is the blood-gas partition
coefficient for nitrogen. This can be rearranged to yield:
共V̇ A/Q̇兲eff共13N兲 ⫽ Cv៮ /CA ⫺ ␭N
(2)
The mixed-venous content is obtained by imaging the right
heart during the infusion, and the alveolar concentration is
obtained by imaging the lung, after the blood volume is
subtracted from the image using a separate 11C-labeled carbon
monoxide inhalation (54). This model assumes the V̇A/Q̇ is
uniform within a voxel, mixed-venous concentration is supplied to all lung regions at the same concentration as the right
heart, V̇A and Q̇ are nonpulsatile, and the tracer is in equilibrium with the alveolus.
Using a modification of the continuous infusion technique,
Venegas and colleagues developed the bolus intravenous inJ Appl Physiol • VOL
(3)
where Af and As are amplitude parameters that are proportional
to Q̇ to fast and slow compartments, respectively; ␣f and ␣s
describe the specific V̇A of the corresponding compartments;
and t is time. Using this method, the spatial information of Q̇
and V̇A was preserved, showing that apparent large uniform
regions of hypoventilation during bronchoconstriction actually
contained units with normal ventilation (Fig. 1). In a noise
analysis, this method was found to yield parameters that
deviated by ⬍1% from noise-free data for up to 32-fold of
expected noise levels (36).
For any new measurement technique, it is important to be
able to validate the new measurements with established
techniques. Richard et al. (60) found a good correlation
between PET measurements of Q̇ using a continuous intravenous infusion of H215O with radiolabeled microspheres
(r ⫽ 0.79, P ⬍ 0.001). The same group (62) showed that
102 • JANUARY 2007 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on May 5, 2017
With PET, the tracer kinetics constitute a dynamic system
that can be analyzed with a mathematical model. For example,
Rhodes et al. (55) used the simple assumption of conservation
of mass of the tracer at the alveolar-capillary interface (Fick
principle) during the constant intravenous infusion of dissolved
13
N-N2 to calculate regional effective (eff) alveolar ventilationperfusion (V̇A/Q̇) in a volume element with the following
formula:
jection technique first used in animals (39, 84) and then
extended to humans (20, 44, 88). With this technique, the
subject undergoes a brief 20- to 30-s apnea as the tracer bolus
(usually ⬃30 ml of 13N-labeled N2-saline with activity 30
mCi) is infused over 10 s. As the tracer arrives into the
pulmonary capillaries, nearly all 13N-N2 diffuses out into the
aerated alveolar spaces due to nitrogen’s low blood-gas partition coefficient (␭N ⫽ 0.015 at 37°C) and remains there until
breathing resumes. The plateau of activity during the apnea is
proportional to regional Q̇. When the subject resumes ventilation, the tracer washes out in proportion to regional-specific
V̇A, or V̇A per unit volume. However, in lungs with atelectasis
or edema, the injected 13N-N2 is not retained in nonaerated
units during breath hold, but is reabsorbed by shunting blood.
In addition, single-compartment analysis of the tracer washout
in these regions is not accurate, since one cannot separate
clearance due to ventilation from clearance due to shunt.
Therefore, a three-compartment tracer kinetic model was developed to assess regional Q̇, regional shunt fraction, and
regional-specific V̇A (13). The model was tested and found to
provide regional parameters of pulmonary function in regions
of lung, the size of which approaches the intrinsic resolution
for PET images of 13N-N2 in the lung (⬃7 mm for a multiring
camera). This model was later refined (48) and compared with
data obtained during PET imaging and found to accurately
describe arterial 13N-N2 kinetics.
The reliability of PET in accurately recovering parameters of
interest depends on the number of counts. When capturing
rapidly changing kinetics, one must use shorter collection
times that necessarily lower the signal-to-noise ratio. One way
to improve the reliability of parameter estimation is to combine
neighboring voxels or low-pass filtering, but this lowers spatial
resolution. Another method pioneered by Kimura et al. (32, 33)
and improved by Layfield and Venegas (34) is that of principle
component analysis, whereby voxels are grouped by their
physiological behavior rather than their spatial proximity. If
the number of discrete values is large enough, the reduction in
precision due to grouping is not physiologically significant.
This method was used to develop images of Q̇ and V̇A from
tracer kinetics data obtained with the bolus method of injected
13
N-N2-saline in a bronchoconstricted human subject. The
tracer content (Ĉ) of each voxel was described by the sum of
two exponential decays:
Invited Review
450
VISUALIZING LUNG FUNCTION WITH PET
V̇A, Q̇, SHUNT, AND GAS CONTENT
Rhodes et al. (54), using the 13N-N2 continuous infusion
technique in 12 subjects, found a vertical gradient in V̇A/Q̇
(⫺1.77%/cm). They measured a mean V̇A/Q̇ of 0.8 and 0.76 in
the right and left lungs, respectively. In some subjects, they
also noted areas of very low V̇A/Q̇ in the most dependent parts
of the lung. These areas were associated with delayed tracer
washout when the infusion was stopped that washed out
quickly with increased tidal volume, suggesting that these were
areas of regional airway closure. This group (5, 6) used
equilibration washout of trace amounts of 19Ne corrected for
blood volume using inhaled 11CO and the continuous infusion
of 13N-N2-saline to measure the gravitational gradients of
blood volume, regional V̇A, and Q̇. They found that variations
in the vertical direction could explain 65% of the total variation
in blood volume, alveolar volume, and V̇A, and 61% of the
variation in Q̇. The vertical gradient for Q̇ was 11%/cm. The
relationship between alveolar volume and V̇A suggests that
elastic properties of lung tissue explains V̇A. They also found
a relationship between blood volume and regional Q̇, suggesting that regional transit times are nearly equal in the gravitational direction.
Fig. 1. Physiological parameters (perfusion parameters Af and As on left, ventilation parameters ␣f and ␣s on right for fast and slow components, respectively)
from a bronchoconstricted subject with asthma calculated by using the grouped parameter estimation method. These results show that bronchoconstriction is focal
in nature, but, even within heavily constricted regions, well-ventilated lung units remain. Also apparent is the significant perfusion (yellow in bottom left image)
to compartments with minimal ventilation (black in bottom right image). [Reproduced from (34) with permission from The Association of University
Radiologists.]
J Appl Physiol • VOL
102 • JANUARY 2007 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on May 5, 2017
PET measurements of V̇A using an equilibration washout of
tracer amounts of 13N-N2 gas compared favorably with
ventilation estimated using changes in lung density from the
transmission scans. Vidal Melo et al. (90) evaluated the
ability of the 13N-N2-saline bolus technique of V̇A/Q̇ measurement to predict arterial blood gases in animal models of
pulmonary embolism, acute lung injury, and bronchoconstriction (Fig. 2). From this analysis, it was clear that, to
match the arterial blood-gas data, it was necessary to consider heterogeneity in V̇A/Q̇ at a level smaller than the
resolution of the PET images (⬃2.2 cm3). This was done by
analyzing the tracer washout and, depending on the behavior, treating the voxel as a one- or two-compartment washout. These results were supported by another study (38) of
intravenous methacholine-induced bronchoconstriction in
sheep, where, in addition to large, contiguous regions of
hypoventilation, there was a substantial subresolution V̇A/Q̇
heterogeneity that had to be taken into consideration to
account for bimodal V̇A/Q̇ distributions. These studies emphasize that, although PET data obtained with the 13N-N2
bolus technique is limited in spatial resolution, one can
extract information from below the resolution of the camera
because of the information contained in the kinetics of the
PET tracer.
Invited Review
VISUALIZING LUNG FUNCTION WITH PET
451
Musch et al. (44), using the 13N-N2-saline bolus technique, studied the effect of the prone and supine positions on
the topographical distribution of pulmonary V̇A and Q̇ in
healthy subjects breathing room air. They found that both
V̇A and Q̇ had vertical gradients that favored the dependent
regions in both positions. The vertical gradients explained,
on average, 24% of the Q̇ heterogeneity and 8% of the V̇A
heterogeneity. The squared coefficient of variation for Q̇
was similar in both positions, whereas squared coefficient of
variation for V̇A was lower in the prone position compared
with supine. Perhaps most important in this study was the
great variability among the six subjects in the response of Q̇
to the change in position. All subjects demonstrated a
redistribution of Q̇ toward ventral, dependent lung on turning prone, but the magnitude of this redistribution varied
substantially. This may mean that factors that make the
vertical gradient of Q̇ much more homogeneous in quadriJ Appl Physiol • VOL
peds (such as the geometry of the vascular tree and/or the
distribution of vasoactive compounds) may not be as important for bipeds. This also may explain discrepancies in
the literature using other techniques (42, 46).
PET has been used to study the V̇A and Q̇ changes after
lung injury in both animal models (15, 43, 61, 65, 74, 75, 85,
92) and humans (73). Schuster et al. (73) found that, unlike
some animal models of lung injury, in seven patients with
acute lung injury/acute respiratory distress syndrome
(ARDS), there was not a significant change in the Q̇ distribution compared with healthy controls. This may be due to
a blunting of hypoxic pulmonary vasoconstriction by substances such as endotoxin that are released during the
initiation of lung injury (14, 15). However, this pattern of Q̇
was also noted in a sheep model of smoke inhalation injury
(92), where a progressive increase in shunt was noted in
dependent regions up to 4 h after injury, suggesting pro-
102 • JANUARY 2007 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on May 5, 2017
Fig. 2. Regional perfusion and end-of-washout lung images, tracer kinetics of the whole lung field, and positron emission tomography (PET)-derived alveolar
ventilation-perfusion (V̇A/Q̇) distributions for single examples of normal sheep and sheep after pulmonary embolism, saline lung lavage, and bronchoconstriction.
Images are tomographic sections viewed in the craniocaudal direction from top to bottom. Animals were prone for normal, bronchoconstriction, and pulmonary
embolism studies and supine for lung lavage. In the supine position, the left side in the image corresponds to the left side in the animal. Note the different scales
for images. Regions of unperfused lung are seen after embolism. After lung lavage, there is redistribution of perfusion and an increase in residual tracer at end
of washout. The early peak and fast drop to plateau in lung lavage tracer kinetics indicates the presence of intrapulmonary shunt. There is significant tracer
retention in large areas after bronchoconstriction. [Reproduced with permission from Vidal Melo et al. (90)].
Invited Review
452
VISUALIZING LUNG FUNCTION WITH PET
diffusing capacity for carbon monoxide (DLCO), they had lower
tissue density, lower peripheral vascular volume, lower blood
flow per centimeter cubed thorax, and higher V̇A/Q̇ than
patients with high DLCO. For those patients with a high DLCO,
they had high tissue density, lower V̇A, and higher Q̇ per tissue
volume. These findings support the notion that some patients
with COPD have predominantly airway destruction vs. airway
inflammation and edema.
Bronchoconstriction has been studied using PET in both
animals (38, 87) and humans (20, 88) using the 13N-N2-saline
bolus infusion technique. A consistent finding in these studies
with methacholine-induced bronchoconstriction is the formation of large-sized (subsegmental to segmental) regions of
hypoventilation (ventilation defects) (Fig. 3). Associated with
these ventilation defects is a substantial reduction in Q̇ (Fig. 3),
estimated to prevent a 33% increase in heterogeneity of the log
SD of V̇A/Q̇ (20). Despite the large size of these regions,
careful analysis of the washout within these ventilation defects
shows that it consists of both normal and hypoventilating units
(Fig. 1). Therefore, the formation of these defects cannot be
attributed solely to large airway constriction (83). To interpret
those findings, a computational model was formulated and
used to test whether the clustering of severely constricted small
airways could be explained with existing knowledge of lung
structure and smooth muscle and airway physiology. The
model incorporated the bistable behavior of airway constriction
proposed by Anafi and Wilson (1) for a terminal bronchiole
into an airway tree structure and included the mechanical
interaction of each airway with the surrounding lung tissues
during breathing. The combination of local bistable behavior
with short-range to long-range competing interactions in the
model predicted patchy self-organized ventilation defects compatible with those seen in the PET images. Although ventilation in asthma was known to be heterogeneous, the reasons
why had not previously been elucidated.
Fig. 3. PET images of perfusion and end-washout (End w/o) before (baseline) and after nebulized methacholine (bronchoconstriction) for a subject with mild
asthma. Each row contains a selected slice of the lung from apex (top) to base (bottom) over a 10-cm span of lung. In each slice, the left lung is on the left, and
the right lung is on the right. Each panel of images is mean-normalized with the color scale representing the 13N-N2 activity. In the perfusion images, note the
vertical gradient in perfusion from ventral to dorsal regions at baseline, but the marked heterogeneity of perfusion during bronchoconstriction. Almost all activity
is washed out at baseline, but large areas of tracer retention are seen in the end w/o images during bronchoconstriction, and these areas correspond to areas of
low perfusion. [Reproduced with permission from Harris et al. (20)].
J Appl Physiol • VOL
102 • JANUARY 2007 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on May 5, 2017
gressive alveolar derecrutiment or flooding in these zones.
Attempts at alveolar recruitment with sustained high-pressure inflations may lead to an increase in shunt blood flow,
if the inflation does not restore aeration to atelectatic regions
(43). In lung injury characterized by substantial atelectasis,
however, recruitment achieved by the prone position (with
the abdomen unsupported) results in improved dorsal lung
aeration and preservation of dorsal blood flow, resulting in
reduced shunt and improved V̇A/Q̇ matching (65). In an
oleic acid injury model of lung injury in pigs, there was a
shift of Q̇ to more ventral regions in the prone position (61),
perhaps because this model is characterized by more alveolar flooding than derecruitment; thus the dorsal regions
were less aerated. The addition of inhaled nitric oxide
enhanced the ventral Q̇ redistribution, regardless of position
(prone vs. supine), but the addition of the vasoconstrictor
almatrine had no effect in either position.
Hypocapneic pneumoconstriction has been proposed as a
homeostatic mechanism to match V̇A and Q̇ (80) and confirmed
by planar positron imaging following pulmonary artery occlusion in dogs (81). Vidal Melo et al. (89) studied the effect of
autologous blood clot (7-mm diameter ⫻ 7-mm length) embolism on Q̇ and V̇A in sheep using the 13N-N2-saline bolus
method and 13N-N2 equilibration washout. They found that
there was an increase in specific V̇A in perfused areas after
embolism, representing a shift in ventilation from zones affected by the clot likely secondary to hypocapneic pneumoconstriction. This study showed, for the first time, that autologous clots causing scattered subsegmental regions of absent Q̇
resulted in a difference of ventilation between areas of clot and
no clot of 100%.
Brudin et al. (4) compared regional V̇A, Q̇, and V̇A/Q̇
measured in 10 subjects with chronic obstructive pulmonary
disease (COPD) using inhaled 19Ne and a continuous infusion
of 13N2. They reported marked differences in these variables,
depending on the clinical type of COPD. For those with a low
Invited Review
VISUALIZING LUNG FUNCTION WITH PET
453
Fig. 4. Transmission (top set) and pulmonary transcapillary
escape rate (PTCER) images (bottom set) in a normal subject
and a patient with acute respiratory distress syndrome (ARDS).
Transmission images have been scaled to units of density
(g/ml), and PTCER images are in units of 10-4/min. Note that
density in the ARDS patient shows a marked ventral-dorsal
gradient (white vs. black arrows), but the increase in permeability (PTCER) is present throughout the lung. [Reproduced
with permission from Sandiford et al. (67)].
Reviews of acute lung injury often emphasize the nonbarrier
functions of the pulmonary endothelium (49, 51, 91). Even so,
barrier function per se is essential to preserving the most
important purpose of the lungs: the adequate exchange of
respiratory gases. Not surprisingly, then, measuring the severity of damage to the endothelial barrier is a common method of
quantifying “acute lung injury” (51, 68).
One approach to evaluating barrier function–and one that
can be applied to imaging methods–is to measure the rate at
which protein moves across the endothelial barrier, from
vascular to extravascular compartments. When PET imaging
is used for this purpose, the goal is to measure the flux (i.e.,
the time-dependent behavior) of radiolabeled proteins
across the pulmonary endothelium (40). This is quantified as
the pulmonary transcapillary escape rate (PTCER) (40).
Thus, as endothelial barrier function deteriorates, PTCER
increases.
Both 68Ga-transferrin and 11C-methylalbumin have been
used as the protein tracers for PET imaging of PTCER (77).
The theory, validation, and limitations of this approach have
been described elsewhere (16, 29, 40, 41, 68, 69, 77, 86).
Furthermore, as PET imaging studies have repeatedly shown
(7, 29 –31, 72), measures of barrier function (such as PTCER)
are a nonspecific index of lung injury, as an increase in PTCER
has been measured by PET imaging in diverse settings, including acute pneumonia, active interstitial lung disease, the reimplantation response, and acute rejection after lung transplantation, in addition to ARDS. Thus such measures as PTCER are
an index of functional not structural lung injury.
The value of an imaging approach to quantifying lung injury
is illustrated by a study by Palazzo et al. (50), who used PET
imaging to measure PTCER in an in vivo canine model of
unilateral pulmonary ischemia-reperfusion injury. They found
that PTCER increased on the ischemic side more than on the
nonischemic side, but both sides of the lungs were increased
compared with control lungs (i.e., with no ischemia on either
J Appl Physiol • VOL
side), suggesting that injury in one lung can lead to similar, but
less severe, injury in the contralateral lung, a finding that has
been observed in an analogous clinical setting (acute unilateral
pneumonia, see below). Subsequently, in a series of studies,
Hamvas et al. (16 –18) used PET imaging to show that the
lungs are remarkably resistant to ischemic tissue injury. In each
of these studies, PTCER was used as the primary biomarker of
lung injury, and the ability to make measurements regionally
within the lungs over time was used to advantage in the study
design.
Similar benefits can be found in PET studies of ARDS (7,
29, 67). In addition, these studies illustrate the value of being
able to measure more than one variable with imaging, allowing
region-specific correlations among these variables. For instance, Calandrino et al. (7) reported that, while PTCER and
extravascular density (EVD) [a close correlate to extravascular
lung water (EVLW)] were both elevated in ARDS patients,
they correlated poorly with one another on a regional basis.
[EVLW can also be measured directly with PET, as reviewed
elsewhere (68, 70, 71, 78).] Furthermore, even as EVD returned to normal, PTCER remained elevated (although decreased relative to the baseline measurement). These results
suggested that injury to lung tissue may be “subclinical” but,
nevertheless, present, even after pulmonary edema has actually
resolved.
In another study in this same patient population, Sandiford et
al. (67) examined the regional distribution of PTCER and EVD
more closely and found ventral-dorsal gradients for EVD in
supine patients (similar to observations made by X-ray CT),
but no ventral-dorsal gradient for PTCER (Fig. 4). Again,
functional injury was present even in lung regions that appeared to be free of structural injury (measured as an increase
in EVD).
The finding that the lungs of ARDS patients are more
diffusely involved than what might otherwise be assumed from
just structural imaging with X-ray CT could help explain why
the lungs of ARDS patients are so vulnerable to ventilator-
102 • JANUARY 2007 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on May 5, 2017
ALVEOLAR-CAPILLARY PERMEABILITY
Invited Review
454
VISUALIZING LUNG FUNCTION WITH PET
induced lung injury during ARDS: radiographically “normal”
lung (i.e., lung with a normal EVLW content) in nondependent
lung regions may still be abnormal and, therefore, vulnerable to
mechanical stresses caused by mechanical ventilation. Data
from the PET imaging studies just cited suggest that nondependent portions of lungs in the ARDS patient are especially
“at risk” because they demonstrate subclinical evidence of
injury, which can be made manifest by inappropriate ventilator
use.
INFLAMMATION
J Appl Physiol • VOL
GENE EXPRESSION
“Structural imaging” includes those methods, like X-ray CT
and magnetic resonance imaging, that depict anatomic structures. Spatial rather than temporal resolution is the critical
variable determining the value of such techniques. In contrast,
“functional imaging,” as with many of the PET imaging
methods described thus far, is commonly used to characterize
a transport process (like pulmonary blood flow, ventilation,
vascular permeability, or rate of glucose uptake). For these
methods, spatial resolution may be important but sometimes
102 • JANUARY 2007 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on May 5, 2017
Despite the apparent importance of inflammation in the
pathogenesis of such diverse conditions as acute pneumonia,
ARDS, cystic fibrosis (CF), and chronic obstructive lung
disease, among many others, no consensus exists about an
acceptable biomarker of the inflammatory response for lung
diseases. Molecular imaging methods (57), however, offer
an exciting opportunity to develop and validate imaging
biomarkers as adjuncts to diagnosis, tests of treatment
efficacy, and treatment monitoring. Ideally, molecular imaging methods would be able to combine the attractiveness
of noninvasive imaging with the specificity for the inflammatory response afforded by analysis of material obtained
from the lower airways.
[18F]fluorodeoxyglucose ([18F]FDG), the most widely used
PET tracer in clinical practice, is a glucose analog in which
fluorine-18 is substituted for the hydroxyl group at the
second carbon position (50a, 94). Entry into cells occurs via
the same glucose transporter family of membrane transporters used by glucose, but, unlike glucose, FDG cannot be
metabolized after it is phosphorylated by hexokinase. Thus
FDG remains trapped after it is taken up by the cell. As
fluorine-18-labeled FDG accumulates in tissue cells, the
concentration of radioactivity builds, eventually reaching a
point at which it can be detected and quantified by an
appropriately calibrated PET camera. FDG-PET imaging
has been used frequently in various studies to identify
pulmonary inflammation, but its potential use as a quantitative biomarker of the inflammatory response is still under
development.
Most studies involving the use of [18F]FDG suggest that
increased uptake by the lungs is specific for neutrophilic
infiltration. Animal studies evaluating [18F]FDG uptake in
models of pneumonia, pancreatitis-induced lung injury, and
sepsis-induced lung injury (19, 21, 24), as well as recent
clinical studies (8, 9), not only show increased [18F]FDG
uptake by the lungs after the insult, but also show by tissue
autoradiography that [3H]deoxyglucose uptake was specifically
limited to neutrophils.
Chen and Schuster (10) hypothesized an increased rate of
[18F]FDG uptake by the lungs in the setting of acute lung
injury and, indeed, found that the rate of [18F]FDG uptake by
the lungs in animals exposed to endotoxin was eight times
higher than that seen in normal controls and four times higher
than the uptake in animals exposed only to oleic acid, a
common method of experimentally producing acute lung injury
(71). The rate of [18F]FDG uptake correlated with in vitro
measurements of deoxyglucose uptake in neutrophils accessible by bronchoalveolar lavage.
Despite such data, increased lung uptake of [18F]FDG is not
necessarily limited to neutrophils. In a recent study in mice, the
increased uptake of [18F]FDG by the lungs in response to
endotoxin was diminished only by ⬃50% in response to
neutrophil depletion (93). And significant increases in the lung
uptake of [18F]FDG can occur during inflammatory lung diseases not characterized by neutrophilic infiltration, such as
sarcoidosis (3, 35). Nevertheless, the available evidence at this
point strongly suggests that, in the appropriate experimental or
clinical context, an increased uptake of [18F]FDG is a highly
reliable biomarker of the neutrophilic inflammatory burden
within the lungs.
Jones et al. (26 –28) were the first to suggest that FDGPET imaging could be used to study pulmonary inflammation in patients. More recently, FDG-PET imaging was used
in a study of 20 adult patients with stable CF (8). The
imaging results were grouped according to the patients’ rate
of decline in pulmonary function during the previous 4 yr
(66). The rate of [18F]FDG uptake was higher in CF patients
than in normal controls, and the highest rates of [18F]FDG
uptake occurred in those patients with the most rapid decline
in lung function. Additionally, the imaging signal correlated
strongly with neutrophil numbers in bronchoalveolar lavage
fluid. These results not only imply that the rate of [18F]FDG
uptake may be useful as a biomarker of active lung inflammation, but that it may also predict the onset of more rapid
deterioration in pulmonary function in CF patients.
A previous study by Taylor et al. (82) suggested the
possibility that FDG-PET imaging could be used to quantify
focal inflammatory responses in the airways of the lungs
induced by bronchoscopic deposition of grass pollen or
other allergens. Therefore, Chen et al. (9) used FDG-PET
imaging in a dose-escalation study in a recently reported
model of focal, limited lung inflammation, induced by the
direct bronchial instillation of small amounts of endotoxin
into the airway of a single lung segment in normal humans
(47). An example of the images obtained is shown in Fig. 5.
The rate of [18F]FDG uptake increased postendotoxin in
the right lung of each research volunteer in the high-dose
group.
Altogether, these various studies suggest FDG-PET imaging
may be useful as a biomarker of the inflammatory response, for
instance in testing new anti-inflammatory agents or in quantifying clinical responses to treatment.
While the use of [18F]FDG-PET is the most frequent example to date of molecular imaging of pulmonary inflammation,
other approaches are also being developed (25). These include
markers of macrophage infiltration, adhesion molecule expression, and apoptosis, among others.
Invited Review
VISUALIZING LUNG FUNCTION WITH PET
455
Fig. 6. PET images (top), light microscopic
(middle), and corresponding fluorescence
micrographs (bottom) obtained in one rat
infected with AdCMVnull (panel 1) and 3
rats, respectively, infected with 1 ⫻ 1010,
5 ⫻ 1010, and 1 ⫻ 1011 VP of AdCMVmNLS-sr39tk-egfp (panels 2, 3, and 4 from
left to right). The PET images are transverse
slices obtained at the midchest level. Light
and fluorescent micrographs (⫻10 magnification) represent identical (ID) lung areas in
the left lung from adjacent sections. Exposure time for each fluorescent micrograph
was the same (960 ms). Fluorescence was
mainly seen in distal parenchymal (alveolar)
cells, rather than in airway epithelium or
microvessel endothelium. VP, viral particles.
[Reproduced with permission from Richard
et al. (64)].
J Appl Physiol • VOL
102 • JANUARY 2007 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on May 5, 2017
Fig. 5. Transmission (top row) and [18F]fluorodeoxyglucose (FDG)-PET (bottom row) images from one healthy volunteer, before and 24 h after direct bronchial
instillation of 4 ng/kg of endotoxin into the right middle lobe. An area of increased density (presumably due to inflammation) and [18F]FDG uptake are clearly
visible (arrows). Also shown is a subtraction [18F]FDG image, in which activity of the before and after images are normalized for injected dose, and the “before”
image is subtracted from the “after” image. [Reproduced with permission from Chen et al. (9)].
Invited Review
456
VISUALIZING LUNG FUNCTION WITH PET
J Appl Physiol • VOL
In conclusion, PET has emerged as an important technology
in the noninvasive study of regional V̇A, Q̇, shunt, gas content,
and alveolar-capillary permeability and is already showing
promise as a tool for imaging inflammation and gene expression. Linking structure to function in complex biological systems has become increasingly important for biomedical research. Making this link requires understanding of how the
components interact to produce a group behavior that cannot be
predicted from that of isolated components. PET has the
potential of integrating gene activation, cell signaling, and cell
trafficking to whole lung organ function, thus advancing our
understanding of pulmonary physiology.
REFERENCES
1. Anafi RC, Wilson TA. Airway stability and heterogeneity in the constricted lung. J Appl Physiol 91: 1185–1192, 2001.
2. Blasberg R, Tjuvajev J. Molecular-genetic imaging: current and future
perspectives. J Clin Invest 111: 1620 –1629, 2003.
3. Brudin L, Valind S, Rhodes C, Pantin C, Sweatman M. Fluorine-18
deoxyglucose uptake in sarcoidosis measured with positron emission
tomography. Eur J Nucl Med 21: 297–305, 1994.
4. Brudin LH, Rhodes CG, Valind SO, Buckingham PD, Jones T,
Hughes JM. Regional structure-function correlations in chronic obstructive lung disease measured with positron emission tomography. Thorax
47: 914 –921, 1992.
5. Brudin LH, Rhodes CG, Valind SO, Jones T, Hughes JM. Interrelationships between regional blood flow, blood volume, and ventilation in
supine humans. J Appl Physiol 76: 1205–1210, 1994.
6. Brudin LH, Rhodes CG, Valind SO, Jones T, Jonson B, Hughes JM.
Relationships between regional ventilation and vascular and extravascular
volume in supine humans. J Appl Physiol 76: 1195–1204, 1994.
7. Calandrino FS Jr, Anderson DJ, Mintun MA, Schuster DP. Pulmonary
vascular permeability during the adult respiratory distress syndrome: a
positron emission tomographic study. Am Rev Respir Dis 138: 421– 428,
1988.
8. Chen DL, Ferkol TW, Mintun MA, Pittman JE, Rosenbluth DB,
Schuster DP. Quantifying pulmonary inflammation in cystic fibrosis with
positron emission tomography. Am J Respir Crit Care Med 173: 1363–
1369, 2006.
9. Chen DL, Rosenbluth DB, Mintun MA, Schuster DP. FDG-PET
imaging of pulmonary inflammation in healthy volunteers after airway
instillation of endotoxin. J Appl Physiol 100: 1602–1609, 2006.
10. Chen DL, Schuster DP. Positron emission tomography with [18F]fluorodeoxyglucose to evaluate neutrophil kinetics during acute lung injury.
Am J Physiol Lung Cell Mol Physiol 286: L834 –L840, 2004.
11. Dharmarajan S, Hayama M, Kozlowski J, Ishiyama T, Okazaki M,
Factor P, Patterson GA, Schuster DP. In vivo molecular imaging
characterizes pulmonary gene expression during experimental lung transplantation. Am J Transplant 5: 1216 –1225, 2005.
12. Dharmarajan S, Schuster DP. Molecular imaging of pulmonary gene
expression with positron emission tomography. Proc Am Thorac Soc 2:
549 –552, 514 –516, 2005.
13. Galletti GG, Venegas JG. Tracer kinetic model of regional pulmonary
function using positron emission tomography. J Appl Physiol 93: 1104 –
1114, 2002.
14. Gust R, Kozlowski J, Stephenson AH, Schuster DP. Synergistic hemodynamic effects of low-dose endotoxin and acute lung injury. Am J Respir
Crit Care Med 157: 1919 –1926, 1998.
15. Gust R, McCarthy TJ, Kozlowski J, Stephenson AH, Schuster DP.
Response to inhaled nitric oxide in acute lung injury depends on distribution of pulmonary blood flow prior to its administration. Am J Respir
Crit Care Med 159: 563–570, 1999.
16. Hamvas A, Kaplan JD, Markham J, Schuster DP. The effects of
regional pulmonary blood flow on protein flux measurements with PET.
J Nucl Med 33: 1661–1668, 1992.
17. Hamvas A, Palazzo R, Kaiser L, Cooper J, Shuman T, Velazquez M,
Freeman B, Schuster DP. Inflammation and oxygen free radical formation during pulmonary ischemia-reperfusion injury. J Appl Physiol 72:
621– 628, 1992.
102 • JANUARY 2007 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on May 5, 2017
less so than with anatomic imaging, as regions of an organ with
similar function (which is often the case in the lungs) can be
adjacent to one another (thus may not need to be distinguished
from one another spatially). However, as the derivation of the
functional parameter frequently requires a mathematical analysis of the time-dependent behavior of some contrast agent or
tracer, the temporal resolution of the imaging method is usually
of special importance.
The value of molecular imaging is directly dependent on its
“specificity,” i.e., the ability to assume that the accumulation of
a tracer or contrast agent into a tissue is entirely dependent on
the expression or activity of a specific molecule, be it an
enzyme, receptor, membrane transporter, etc. While this specificity may be conferred by the contrast agent or tracer itself
(for instance, a highly specific radiolabeled ligand for a receptor), nonspecific uptake of such probes is almost invariably a
factor that reduces the ratio of imaging signal-to-background
noise.
In contrast, very high degrees of specificity can be achieved
by using modern molecular biology methods to introduce
reporter genes into target tissues, the products of which trap a
contrast agent in sufficient concentrations to generate an imaging signal that can be detected and recorded by PET (or
other) instrumentation. Using such an approach, molecular
imaging methods are freed from the time-consuming process of
developing, implementing, and validating the pharmacokinetics and pharmacodynamics of the imaging probe (the contrast
agent or tracer) and instead become simply dependent on the
ability to link the reporter gene to the target gene, a process that
can be achieved with great efficiency using modern molecular
tools. Broad overviews that cover general concepts, strategies,
and imaging platforms for molecular imaging can be found
elsewhere (2, 11, 12, 37, 52, 56, 76, 79).
Thus far, studies involving PET reporter gene (PRG) imaging
in the lungs have almost exclusively used mutant variants of the
Herpes simplex virus (HSV) type I thymidine kinase (mHSV1-tk)
as the “PRG” and 9-(4-[18F]-fluoro-3-hydroxymethylbutyl)guanine ([18F]FHBG) as the “PET reporter probe” (PRP) (11, 58, 59,
63, 64). All of these studies have used intratracheal delivery of
adenovirus carrying the mHSV1-tk gene, followed by intravenous
administration of [18F]FHBG before imaging. The [18F]FHBG is
trapped in cells, if it is phosphorylated by an appropriate kinase.
Mammalian kinases have a restricted substrate specificity relative
to the HSV viral kinase. Thus mammalian thymidine kinase
will not phosphorylate, or trap, [18F]FHBG intracellularly. In
contrast, cells expressing the mHSV1-tk gene produce a thymidine kinase capable of phosphorylating (and therefore trapping) [18F]FHBG. As the [18F]FHBG accumulates in these
tissues, an imaging signal is generated from the radiation
emitted by fluorine-18 decay.
The feasibility of using this PRG-PRP combination as a
reporter system to monitor transgene therapy in the lungs was
demonstrated by Richard et al. (64) using previously transfected normal rodent lungs with various doses of adenovirus
carrying the mHSV1-tk gene administered intratracheally (Fig.
6). Using this approach, they then showed that gene delivery in
a surfactant vehicle is at least 40% greater than when saline is
used as the vehicle (59), providing proof of principle that
imaging could be used to monitor gene delivery in clinical
trials of gene therapy.
Invited Review
VISUALIZING LUNG FUNCTION WITH PET
J Appl Physiol • VOL
43. Musch G, Harris RS, Vidal Melo MF, O’Neill KR, Layfield JD,
Winkler T, Venegas JG. Mechanism by which a sustained inflation can
worsen oxygenation in acute lung injury. Anesthesiology 100: 323–330,
2004.
44. Musch G, Layfield JD, Harris RS, Melo MF, Winkler T, Callahan RJ,
Fischman AJ, Venegas JG. Topographical distribution of pulmonary
perfusion and ventilation, assessed by PET in supine and prone humans.
J Appl Physiol 93: 1841–1851, 2002.
45. Musch G, Venegas JG. Positron emission tomography imaging of regional pulmonary perfusion and ventilation. Proc Am Thorac Soc 2:
522–527, 508 –509, 2005.
46. Nyren S, Mure M, Jacobsson H, Larsson SA, Lindahl SG. Pulmonary
perfusion is more uniform in the prone than in the supine position:
scintigraphy in healthy humans. J Appl Physiol 86: 1135–1141, 1999.
47. O’Grady NP, Preas HL, Pugin J, Fiuza C, Tropea M, Reda D, Banks
SM, Suffredini AF. Local inflammatory responses following bronchial
endotoxin instillation in humans. Am J Respir Crit Care Med 163:
1591–1598, 2001.
48. O’Neill K, Venegas JG, Richter T, Harris RS, Layfield JD, Musch G,
Winkler T, Melo MF. Modeling kinetics of infused 13NN-saline in acute
lung injury. J Appl Physiol 95: 2471–2484, 2003.
49. Orfanos S, Mavrommati I, Korovesi I, Roussos C. Pulmonary endothelium in acute lung injury: from basic science to the critically ill.
Intensive Care Med 30: 1702–1714, 2004.
50. Palazzo R, Hamvas A, Shuman T, Kaiser L, Cooper J, Schuster DP.
Injury in nonischemic lung after unilateral pulmonary ischemia with
reperfusion. J Appl Physiol 72: 612– 620, 1992.
50a.Pauwels EK, Sturm EJ, Bombardieri E, Cleton FJ, Stokkel MP.
Positron-emission tomography with [18F] fluorodeoxyglucose. Part I.
Biochemical uptake mechanism and its implication for clinical studies.
J Cancer Res Clin Oncol 126: 549 –559, 2000.
51. Pittet J, Mackersie R, Martin T, Matthay M. Biological markers of
acute lung injury: prognostic and pathogenetic significance. Am J Respir
Crit Care Med 155: 1187–1205, 1997.
52. Piwnica-Worms D, Schuster DP, Garbow JR. Molecular imaging of
host-pathogen interactions in intact small animals. Cell Microbiol 6:
319 –331, 2004.
53. Rhodes CG, Hughes JM. Pulmonary studies using positron emission
tomography. Eur Respir J 8: 1001–1017, 1995.
54. Rhodes CG, Valind SO, Brudin LH, Wollmer PE, Jones T, Buckingham PD, Hughes JM. Quantification of regional V̇A/Q̇ ratios in humans
by use of PET. II. Procedure and normal values. J Appl Physiol 66:
1905–1913, 1989.
55. Rhodes CG, Valind SO, Brudin LH, Wollmer PE, Jones T, Hughes
JM. Quantification of regional V̇A/Q̇ ratios in humans by use of PET. I.
Theory. J Appl Physiol 66: 1896 –1904, 1989.
56. Richard JC, Chen DL, Ferkol T, Schuster DP. Molecular imaging for
pediatric lung diseases. Pediatr Pulmonol 37: 1–11, 2004.
57. Richard JC, Chen DL, Ferkol T, Schuster DP. Molecular imaging for
pediatric lung diseases. Pediatr Pulmonol 37: 286 –296, 2004.
58. Richard JC, Factor P, Ferkol T, Ponde DE, Zhou Z, Schuster DP.
Repetitive imaging of reporter gene expression in the lung. Mol Imaging
2: 342–349, 2003.
59. Richard JC, Factor P, Welch LC, Schuster DP. Imaging the spatial
distribution of transgene expression in the lungs with positron emission
tomography. Gene Ther 10: 2074 –2080, 2003.
60. Richard JC, Janier M, Decailliot F, Le Bars D, Lavenne F, Berthier V,
Lionnet M, Cinotti L, Annat G, Guerin C. Comparison of PET with
radioactive microspheres to assess pulmonary blood flow. J Nucl Med 43:
1063–1071, 2002.
61. Richard JC, Janier M, Lavenne F, Berthier V, Lebars D, Annat G,
Decailliot F, Guerin C. Effect of position, nitric oxide, and almitrine on
lung perfusion in a porcine model of acute lung injury. J Appl Physiol 93:
2181–2191, 2002.
62. Richard JC, Janier M, Lavenne F, Tourvieille C, Le Bars D, Costes N,
Gimenez G, Guerin C. Quantitative assessment of regional alveolar
ventilation and gas volume using 13N-N2 washout and PET. J Nucl Med
46: 1375–1383, 2005.
63. Richard JC, Zhou Z, Chen DL, Mintun MA, Piwnica-Worms D,
Factor P, Ponde DE, Schuster DP. Quantitation of pulmonary transgene
expression with PET imaging. J Nucl Med 45: 644 – 654, 2004.
64. Richard JC, Zhou Z, Ponde DE, Dence CS, Factor P, Reynolds PN,
Luker GD, Sharma V, Ferkol T, Piwnica-Worms D, Schuster DP.
102 • JANUARY 2007 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on May 5, 2017
18. Hamvas A, Schuster DP. Bronchial and reverse pulmonary venous blood
flow protect the lung from ischemia-reperfusion injury. J Appl Physiol 77:
731–736, 1994.
19. Hansson L, Ohlsson T, Valind S, Sandell A, Luts A, Jeppsson B,
Wollmer P. Glucose utilisation in the lungs of septic rats. Eur J Nucl Med
26: 1340 –1344, 1999.
20. Harris RS, Winkler T, Tgavalekos N, Musch G, Melo MF, Schroeder
T, Chang Y, Venegas JG. Regional pulmonary perfusion, inflation, and
ventilation defects in bronchoconstricted patients with asthma. Am J
Respir Crit Care Med 174: 245–253, 2006.
21. Hartwig W, Carter EA, Jimenez RE, Jones R, Fischman AJ, Fernandez-Del Castillo C, Warshaw AL. Neutrophil metabolic activity but not
neutrophil sequestration reflects the development of pancreatitis-associated lung injury. Crit Care Med 30: 2075–2082, 2002.
22. Hughes JM. Positron emission tomography and the lung. Br J Hosp Med
56: 86 – 89, 1996.
23. Hughes JM, Brudin LH, Valind SO, Rhodes CG. Positron emission
tomography in the lung. J Thorac Imaging 1: 79 – 88, 1985.
24. Iyer M, Berenji M, Templeton N, Gambhir S. Noninvasive imaging of
cationic lipid-mediated delivery of optical and PET reporter genes in
living mice. Mol Ther 6: 555–562, 2002.
25. Jones H. Inflammation imaging. Proc Am Thorac Soc 2: 545–548,
513–514, 2005.
26. Jones H, Schofield J, Krausz T, Boobis A, Haslett C. Pulmonary fibrosis
correlates with duration of tissue neutrophil activation. Am J Respir Crit
Care Med 158: 620 – 628, 1998.
27. Jones H, Sriskandan S, Peters A, Pride N, Krausz T, Boobis A, Haslett
C. Dissociation of neutrophil emigration and metabolic activity in lobar
pneumonia and bronchiectasis. Eur Respir J 10: 795– 803, 1997.
28. Jones HA, Clark RJ, Rhodes CG, Schofield JB, Krausz T, Haslett C.
Positron emission tomography of 18FDG uptake in localized pulmonary
inflammation. Acta Radiol Suppl 376: 148, 1991.
29. Kaplan JD, Calandrino FS, Schuster DP. A positron emission tomographic comparison of pulmonary vascular permeability during the adult
respiratory distress syndrome and pneumonia. Am Rev Respir Dis 143:
150 –154, 1991.
30. Kaplan JD, Trulock EP, Anderson DJ, Schuster DP. Pulmonary vascular permeability in interstitial lung disease. A positron emission tomographic study. Am Rev Respir Dis 145: 1495–1498, 1992.
31. Kaplan JD, Trulock EP, Cooper JD, Schuster DP. Pulmonary vascular
permeability after lung transplantation. A positron emission tomographic
study. Am Rev Respir Dis 145: 954 –957, 1992.
32. Kimura Y, Hsu H, Toyama H, Senda M, Alpert NM. Improved
signal-to-noise ratio in parametric images by cluster analysis. Neuroimage
9: 554 –561, 1999.
33. Kimura Y, Senda M, Alpert NM. Fast formation of statistically reliable
FDG parametric images based on clustering and principal components.
Phys Med Biol 47: 455– 468, 2002.
34. Layfield D, Venegas JG. Enhanced parameter estimation from noisy PET
data. I. Methodology. Acad Radiol 12: 1440 –1447, 2005.
35. Lewis P, Salama A. Uptake of fluorine-18-fluorodeoxyglucose in sarcoidosis. J Nucl Med 35: 1647–1649, 1994.
36. Marcinkowski A, Layfield D, Tgavalekos N, Venegas JG. Enhanced
parameter estimation from noisy PET data. II. Evaluation. Acad Radiol 12:
1448 –1456, 2005.
37. Massoud T, Gambhir S. Molecular imaging in living subjects: seeing
fundamental biological processes in a new light. Genes Dev 17: 545–580,
2003.
38. Melo MF, Harris RS, Layfield JD, Venegas JG. Topographic basis of
bimodal ventilation-perfusion distributions during bronchoconstriction in
sheep. Am J Respir Crit Care Med 171: 714 –721, 2005.
39. Mijailovich SM, Treppo S, Venegas JG. Effects of lung motion and
tracer kinetics corrections on PET imaging of pulmonary function. J Appl
Physiol 82: 1154 –1162, 1997.
40. Mintun MA, Dennis DR, Welch MJ, Mathias CJ, Schuster DP.
Measurements of pulmonary vascular permeability with PET and
gallium-68 transferrin. J Nucl Med 28: 1704 –1716, 1987.
41. Mintun MA, Warfel TE, Schuster DP. Evaluating pulmonary vascular
permeability with radiolabeled proteins: an error analysis. J Appl Physiol
68: 1696 –1706, 1990.
42. Mure M, Nyren S, Jacobsson H, Larsson SA, Lindahl SG. High
continuous positive airway pressure level induces ventilation/perfusion
mismatch in the prone position. Crit Care Med 29: 959 –964, 2001.
457
Invited Review
458
65.
66.
67.
68.
69.
70.
71.
73.
74.
75.
76.
77.
78.
79.
80.
Imaging pulmonary gene expression with positron emission tomography.
Am J Respir Crit Care Med 167: 1257–1263, 2003.
Richter T, Bellani G, Scott Harris R, Vidal Melo MF, Winkler T,
Venegas JG, Musch G. Effect of prone position on regional shunt,
aeration, and perfusion in experimental acute lung injury. Am J Respir Crit
Care Med 172: 480 – 487, 2005.
Rosenbluth DB, Wilson K, Ferkol T, Schuster DP. Lung function
decline in cystic fibrosis patients and timing for lung transplantation
referral. Chest 126: 412– 419, 2004.
Sandiford P, Province MA, Schuster DP. Distribution of regional
density and vascular permeability in the adult respiratory distress syndrome. Am J Respir Crit Care Med 151: 737–742, 1995.
Schuster D. The evaluation of pulmonary endothelial barrier function:
quantifying pulmonary edema and lung injury. In: Pulmonary Edema,
edited by Matthay M and Ingbar D. Boca Raton, FL: Dekker, 1997, p.
121–161.
Schuster D, Stark T, Stephenson J, Royal H. Detecting lung injury in
patients with pulmonary edema. Intensive Care Med 28: 1246 –1253,
2002.
Schuster DP. Positron emission tomography: theory and its application to
the study of lung disease. Am Rev Respir Dis 139: 818 – 840, 1989.
Schuster DP. ARDS: clinical lessons from the oleic acid model of acute
lung injury. Am J Respir Crit Care Med 149: 245–260, 1994.
Schuster DP. What is acute lung injury? What is ARDS? Chest 107:
1721–1726, 1995.
Schuster DP, Anderson C, Kozlowski J, Lange N. Regional pulmonary
perfusion in patients with acute pulmonary edema. J Nucl Med 43:
863– 870, 2002.
Schuster DP, Haller J. Regional pulmonary blood flow during acute
pulmonary edema: a PET study. J Appl Physiol 69: 353–361, 1990.
Schuster DP, Haller JW, Velazquez M. A positron emission tomographic comparison of diffuse and lobar oleic acid lung injury. J Appl
Physiol 64: 2357–2365, 1988.
Schuster DP, Kovacs A, Garbow J, and Piwnica-Worms D. Recent
advances in imaging the lungs of intact small animals. Am J Respir Cell
Mol Biol 30: 129 –138, 2004.
Schuster DP, Markham J, Welch MJ. Positron emission tomography
measurements of pulmonary vascular permeability with Ga-68 transferrin
or C-11 methylalbumin. Crit Care Med 26: 518 –525, 1998.
Schuster DP, Mintun MA, Green MA, and Ter-Pogossian MM. Regional lung water and hematocrit determined by positron emission tomography. J Appl Physiol 59: 860 – 868, 1985.
Serganova I, Blasberg R. Reporter gene imaging: potential impact on
therapy. Nucl Med Biol 32: 763–780, 2005.
Severinghaus JW, Swenson EW, Finley TN, Lategola MT, Williams J.
Unilateral hypoventilation produced in dogs by occluding one pulmonary
artery. J Appl Physiol 16: 53– 60, 1961.
J Appl Physiol • VOL
81. Simon BA, Tsuzaki K, Venegas JG. Changes in regional lung mechanics
and ventilation distribution after unilateral pulmonary artery occlusion.
J Appl Physiol 82: 882– 891, 1997.
82. Taylor IK, Hill AA, Hayes M, Rhodes C, O’Shaughnessy K,
O’Connor B, Jones H, Hughes J, Jones T, Pride N, Fuller R. Imaging
allergen-invoked airway inflammation in atopic asthma with [18F]-fluorodeoxyglucose and positron emission tomography. Lancet 347: 937–940,
1996.
83. Tgavalekos NT, Tawhai M, Harris RS, Musch G, Vidal-Melo M,
Venegas JG, Lutchen KR. Identifying airways responsible for heterogeneous ventilation and mechanical dysfunction in asthma: an image functional modeling approach. J Appl Physiol 99: 2388 –2397, 2005.
84. Treppo S, Mijailovich SM, Venegas JG. Contributions of pulmonary
perfusion and ventilation to heterogeneity in V̇A/Q̇ measured by PET.
J Appl Physiol 82: 1163–1176, 1997.
85. Velazquez M, Schuster DP. Pulmonary blood flow distribution after
lobar oleic acid injury: a PET study. J Appl Physiol 65: 2228 –2235, 1988.
86. Velazquez M, Weibel ER, Kuhn C 3rd, Schuster DP. PET evaluation of
pulmonary vascular permeability: a structure-function correlation. J Appl
Physiol 70: 2206 –2216, 1991.
87. Venegas JG, Schroeder T, Harris S, Winkler RT, Melo MF. The
distribution of ventilation during bronchoconstriction is patchy and bimodal: a PET imaging study. Respir Physiol Neurobiol 148: 57– 64, 2005.
88. Venegas JG, Winkler T, Musch G, Vidal Melo MF, Layfield D,
Tgavalekos N, Fischman AJ, Callahan RJ, Bellani G, Harris RS.
Self-organized patchiness in asthma as a prelude to catastrophic shifts.
Nature 434: 777–782, 2005.
89. Vidal Melo MF, Harris RS, Layfield D, Musch G, Venegas JG.
Changes in regional ventilation after autologous blood clot pulmonary
embolism. Anesthesiology 97: 671– 681, 2002.
90. Vidal Melo MF, Layfield D, Harris RS, O’Neill K, Musch G, Richter
T, Winkler T, Fischman AJ, Venegas JG. Quantification of regional
ventilation-perfusion ratios with PET. J Nucl Med 44: 1982–1991, 2003.
91. Ware LB, Matthay MA. The acute respiratory distress syndrome. N Engl
J Med 342: 1334 –1349, 2000.
92. Willey-Courand DB, Harris RS, Galletti GG, Hales CA, Fischman
A, Venegas JG. Alterations in regional ventilation, perfusion, and
shunt after smoke inhalation measured by PET. J Appl Physiol 93:
1115–1122, 2002.
93. Zhou Z, Kozlowski J, Goodrich AL, Markman N, Chen DL, Schuster
DP. Molecular imaging of lung glucose uptake after endotoxin in mice.
Am J Physiol Lung Cell Mol Physiol 289: L760 –L768, 2005.
94. Zhuang H, Alavi A. 18-Fluorodeoxyglucose positron emission tomographic imaging in the detection and monitoring of infection and inflammation. Semin Nucl Med 31: 47–59, 2002.
102 • JANUARY 2007 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on May 5, 2017
72.
VISUALIZING LUNG FUNCTION WITH PET