| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Pulmonary, Allergy, and Critical Care Division (Drs. Kotloff, Hansen-Flaschen, Lipson, Tino, and Arcasoy), Division of Nuclear Medicine (Dr. Alavi), and Division of Cardiothoracic Surgery (Dr. Kaiser), University of Pennsylvania Medical Center, Philadelphia, PA.
Correspondence to: Robert M. Kotloff, MD, FCCP, Pulmonary, Allergy, and Critical Care Division, 838 West Gates, University of Pennsylvania Medical Center, 3400 Spruce St, Philadelphia, PA 19104; e-mail: kotloff{at}mail.med.upenn.edu
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: We performed a retrospective analysis of 128 patients who underwent bilateral LVRS. An apical perfusion fraction (AP%), defined as the percentage of total lung perfusion to the apical one third of both lungs, was derived for each patient by quantitative scintigraphy technique. Pulmonary function testing and 6-min walk test (6MWT) data were obtained preoperatively and 3 to 6 months postoperatively.
Results: The mean (± SD)
improvement in FEV1 was 309 ± 240 mL, 209 ± 293 mL,
and 116 ± 224 mL for patients with an AP% of 
10%, 11 to 20%,
and > 20%, respectively (p = 0.01, analysis of variance
[ANOVA]). The likelihood of experiencing an increase in
FEV1 
200 mL was 68% for those with an AP% 10%
but only 31% for those with an AP% > 20%. Preoperative and
postoperative 6MWT data were available for 109 of 128 patients.
Improvement was 250 ± 252 feet, 205 ± 299 feet, and 77 ± 200
feet for patients with AP% 10%, 11 to 20%, and > 20%,
respectively (p = 0.04, ANOVA). While 50% of those with an AP%
10% improved their 6MWT by 180 feet, only 21% of those with
an AP% > 20% did so.
Conclusion: This retrospective analysis suggests that quantification of apical perfusion by nuclear scintigraphy assists in predicting the likelihood of short-term functional improvement after LVRS.
Key Words: COPD • emphysema • lung volume reduction surgery • perfusion scintigraphy
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
As initially conceived by Brantigan et al9 and subsequently refined by Cooper et al,10 LVRS is deemed to be best suited for patients with heterogeneously distributed disease, with resection targeted at the most severely diseased portions of lung tissue. The rationale for this strategy is straightforward. The resected tissue is presumed to contribute little to overall lung function and gas exchange, and its loss should not, therefore, adversely impact the patient. As importantly, resection of nonfunctioning, space-occupying tissue permits the expansion of remaining healthier tissue, thus enhancing the overall elastic recoil of the remaining lung.11
Preliminary data from several centers support the notion that patients with heterogeneously distributed and, in particular, apically predominant emphysema derive the greatest functional benefit from LVRS.4 12 13 Identification of such patients has relied on a number of radiographic techniques, including standard chest radiography,13 14 CT imaging,8 15 16 17 and lung perfusion scintigraphy.12 15 To date, however, these techniques have either employed quantitative analysis that is not widely available (ie, CT morphometry) or semiquantitative scoring systems that require an experienced reader and are subject to interobserver variability.
To address these potential shortcomings, we have devised a simple surrogate marker for apically predominant emphysema—derived from computer analysis of lung perfusion scintigraphy images—that we have called apical perfusion fraction (AP%). The purpose of this study was to determine the utility of the AP% in predicting short-term functional improvement following LVRS.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Patient selection criteria employed by our program have been previously published.1 Briefly, patients who underwent LVRS had evidence of severe airflow obstruction, were severely hyperinflated with a residual volume (RV) [by body plethysmography] > 200% of predicted, and had areas of hypoperfused lung by perfusion scintigraphy that could be targeted for resection. All patients selected for surgery were required to complete a 6-week course of preoperative pulmonary rehabilitation and a second 6-week course following surgery.
Quantitative Perfusion Scintigraphy
Prior to surgery, all patients underwent quantitative perfusion
lung scintigraphy using standard techniques. Approximately 4 mCi of
99mTc-labeled macroaggregated albumin was IV
administered to the supine patient; images were subsequently obtained
in eight standard projections with the patient in an erect position.
Since posterior views allow best overall estimates of pulmonary
perfusion, we routinely used this projection for generation of
quantitative perfusion data. The boundaries of the lungs were defined
by enhancing the peripheral perfusion of the lungs (including the upper
lung zones) by reducing the upper threshold for displaying the
intensity of pixel count in the entire image set. The lung zones were
defined by assigning two rectangles whose perimeters abutted on the
outermost pixel in the superior, medial, inferior, and lateral
boundaries of the lungs. Thereafter, the rectangles were equally
divided into upper, middle, and lower zones by assigning two horizontal
lines within each rectangle (Fig 1
). The counts in each of the six lung zones were individually
determined, and the percentage of perfusion to each lung zone was
calculated by dividing the counts of the lung zone of interest by the
total counts in both lungs, multiplied by 100. The AP% was calculated
as the sum of the percentage of perfusion to the right and left upper
lung zones (Fig 1)
.
|
Clinical Outcomes
For the purpose of this analysis, patients were classified into
three groups based on preoperative AP%: 10%, 11 to 20%, and
> 20%. The primary outcome measures were the change in
FEV1 and 6MWT distance at 3 to 6 months after
LVRS, reflecting changes in the degree of airflow obstruction and
exercise tolerance, respectively. In addition to examining the absolute
magnitude of change in these parameters, we determined the percentage
of patients in each AP% group who derived a "clinically significant
benefit," which we defined as an increase of 200 mL (and 12%) in
the FEV1 or 180 feet in the 6MWT distance.
These thresholds are based on standards previously established in the
literature.18
19
Statistical Analysis
StatView statistical program (version 5.0 for Windows; SAS
Institute; Cary, NC) was used for statistical analysis. Data are
expressed as mean ± SD. Analysis of variance (ANOVA), followed (when
significant) by the Fisher’s Protected Least Significant Difference
post hoc test, was used to compare differences in continuous
variables among the three AP% groups. 
2
analysis was employed to examine differences in categorical variables.
A p value < 0.05 was considered statistically significant.
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
10%, 59
patients (46.1%) had an AP% from 11 to 20%, and 35 patients (27.3%)
had an AP% > 20%. There was no significant difference in
baseline characteristics among the three groups (Table 1)
. Since
not all of the preoperative arterial blood gas values were obtained
with patients breathing room air, data on preoperative
PO2 were not included.
Following LVRS, the mean improvement in
FEV1 for the entire group of 128 patients was
210 ± 270 mL. The association between AP% and change in
FEV1 is shown in Figure 2
. Patients with an AP% 10% realized an improvement in
FEV1 of 309 ± 240 mL, those with a score of 11
to 20% improved by 209 ± 293 mL, and those patients with an AP%
> 20% improved by a mean of 116 ± 224 mL (p = 0.01, ANOVA). To
further investigate the utility of the AP% in predicting
response to LVRS, we identified the percentage of patients in each
group that achieved a "clinically meaningful" improvement in
FEV1. This was defined as an increase in
FEV1 of > 200 mL (in all cases, this was also
> 12% of baseline value). As shown in Table 2
, a favorable response was achieved by 68% of patients with an AP%
10%, 47% of those with a score of 11 to 20%, and only 31% of
those with a score > 20% (p = 0.01).
|
|
10%, 11 to 20%, and > 20%, respectively (p = 0.04, ANOVA).
Defining a clinically meaningful response as an increase in walk
distance of > 180 feet, 50% of those with an AP% 10%, 43% of
those with an AP% from 11 to 20%, and 21% of those with an AP%
> 20% realized a favorable outcome (Table 2
; p = 0.05).
|
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
10% achieved a greater degree of
improvement in FEV1 and 6MWT distance than those
who had better relative perfusion to the apices. The notion that topographic heterogeneity of emphysema is critically important to the success of LVRS dates back to the original description of the procedure by Brantigan et al,9 who emphasized the importance of targeting the most severely diseased and presumably functionless areas of lung while leaving behind healthier tissue whose function could be enhanced by providing room for expansion within the thorax. The requirement for heterogeneity of emphysema distribution was adopted by Cooper et al10 in their resurrection of LVRS in the early 1990s. The subsequent incorporation of this requirement into standard selection criteria utilized by a majority of centers preceded any scientific verification of its validity and any standardization of its measurement.
Utilizing quantitative perfusion scintigraphy, we have chosen to assess heterogeneity in a functional rather than anatomic fashion by assessing regional variations in blood flow within the lung. The presumption in using this approach in patients with advanced emphysema is that areas of relative hypoperfusion correspond to geographic regions that are most severely diseased. Quantitative perfusion scintigraphy has a number of features that make it an attractive method of assessing LVRS candidates: it is widely available, requires no breath-holding or patient effort, and provides a digital output that is readily amenable to quantitative analysis. We devised a simple, objective, computer-generated index of apical perfusion, the AP%, representing the percentage of total perfusion that is directed to the upper one third of both lungs. We chose to focus on apical perfusion based on preliminary data from several centers that upper lobe predominant disease was the most favorable pattern of disease heterogeneity with respect to the functional outcome of LVRS. This "ideal" pattern of disease distribution would be reflected in a low AP%. In contrast, a high AP% provides less information about disease heterogeneity and distribution, since it could result either from homogeneously distributed disease (with relatively even perfusion to all lung zones) or from focal regions of severe disease in the middle or lower lung zones.
Our finding that lung scintigraphy provides meaningful predictive information with respect to functional outcomes is in agreement with previous studies. Wang and colleagues12 studied 96 patients and utilized a semiquantitative scoring of scintigraphy patterns. Similar to our results, they found that the greatest degree of improvement in FEV1 occurred in association with upper-lung-zone-predominant disease. In a study of 70 patients, Thurnheer and colleagues15 utilized a visual analysis that categorized perfusion scan patterns into markedly heterogeneous, intermediately heterogeneous, and homogeneous categories. The magnitude of improvement in FEV1 was 57 ± 8%, 38 ± 9%, and 23 ± 9% for the markedly heterogeneous, intermediately heterogeneous, and homogeneous categories, respectively, although the differences among the three groups were not statistically significant and the correlation with outcomes was not as strong as with CT scoring. In contrast to our quantitative, computer-generated technique of scoring the scintigraphy scans, these two studies relied on observer-generated assessment. This not only mandates the involvement of highly experienced interpreters but also introduces the problem of interobserver variability. For example, in the study by Thurnheer et al,15 there was agreement among at least five of the six observers in only 55% of studies. Using an apical perfusion score similar to ours in a preliminary study of 20 patients, Oey and colleagues20 reported a significant correlation between this score and postoperative functional and health-status outcomes.
Other investigators, employing morphologic rather than functional measures of emphysema severity and distribution, have drawn conclusions similar to our own. Utilizing sophisticated visual scoring systems of chest radiographs, two groups demonstrated that the degree of heterogeneity and, in particular, an upper-lobe-predominant distribution pattern of emphysema, were associated with a more favorable improvement in FEV1 and 6MWT distance.13 14 Similar results were noted by others with the use of semiquantitative scoring of CT scans15 16 21 and quantitative CT morphometry.17 The information derived from functional and anatomic imaging may prove to be complementary rather than redundant. In a direct comparison of perfusion scintigraphy and CT imaging, Thurnheer and coworkers15 observed heterogeneity of perfusion in 16 of 22 patients with ostensibly homogeneous disease as assessed by CT scoring. This suggests that scintigraphy may be more sensitive than visual assessment of CT images in detecting subtle differences in regional distribution of disease. Conversely, CT offers several potential advantages over scintigraphy. CT imaging permits confirmation that poorly perfused areas identified by scintigraphy are, in fact, a result of emphysematous destruction of the involved region of lung. This is critically important in light of evidence that LVRS is suited only for patients whose airflow obstruction is due to emphysema and not to those whose disease is intrinsic to the airways.6 CT scans also permit identification of other pulmonary processes, such as lung nodules, bronchiectasis, or chronic infection, which would influence the suitability of the patient for LVRS or direct the targeted resection of abnormal areas.
While we have demonstrated a correlation between AP% and functional
outcome, the limitations of this score in selection of patients must be
acknowledged. One-third of patients with an AP% 10% failed to
realize an improvement in FEV1 of 200 mL and
one half of patients failed to increase their 6MWT distance by 200
feet. Conversely, nearly a third of patients with an AP% > 20% did
experience such an improvement in FEV1 and one
fifth of patients achieved a clinically significant improvement in 6MWT
distance. Assessment of heterogeneity by chest radiography and CT
imaging have yielded similarly crude correlations with functional
outcomes, suggesting that the imprecise predictive value of the AP% is
not due to limitations inherent in the use of perfusion scintigraphy.
It is likely that heterogeneity is only one of several factors critical
to the outcome of LVRS. In addition to augmenting elastic recoil of the
remaining lung, LVRS may also effect improvement by reducing the degree
of hyperinflation, allowing the diaphragm to ascend to a more
mechanically favorable position.22
23
Thus, factors such
as the degree of preoperative hyperinflation and the volume of tissue
removed may also dictate the magnitude of functional
improvement.15
An additional potential limitation of the apical perfusion fraction is
that it cannot identify patients with markedly heterogeneous but
basilar predominant disease, a condition encountered in patients with
1-antitrypsin deficiency as well as in some
without this condition. In our series, there was an insufficient number
of patients with basilar hypoperfusion (ie, 10% total
perfusion to the lower thirds) to allow analysis of their response to
LVRS. However, observations by multiple other
investigators4
12
13
14
have suggested that functional
outcomes in this group of patients are inferior to those achieved in
patients with apically distributed disease, although there is at least
one study24
suggesting equivalent outcomes. Additional
experience is required to settle this issue; certainly the quantitative
scoring system we have proposed could be easily adapted to detect
basilar predominant disease if this distinction proves to be critical.
Finally, the incomplete nature of our database must be acknowledged as
a potential limitation. Fifteen patients died within the first 6 months
postoperatively and were excluded from our analysis. Including these
patients as "nonresponders" would not have significantly altered
the predictive utility of the AP% (Table 2)
. Analyzed in this fashion,
the percentage of patients deriving a clinically meaningful improvement
in FEV1 would have been 59% (vs 68%), 47% (vs
47%), and 25% (vs 31%) for the 10% AP%, 11 to 20% AP%, and
> 20% AP% groups, respectively. Similarly, the percentage of
patients deriving a clinically meaningful improvement in 6MWT distance
would have been 42% (vs 50%), 40% (vs 41%), and 16% (vs 21%) for
the 10% AP%, 11 to 20% AP%, and > 20% AP% groups,
respectively. There was also incomplete follow-up of survivors; one
fourth of survivors did not undergo pulmonary function testing within
the initial 6 months, and one third of patients did not undergo
follow-up 6MWTs in this time frame. Since relatively equal proportions
of patients were lacking follow-up data in each AP% group, any bias
introduced by the incomplete data should be minimized. We did not
report data beyond the initial 6 months because our intent was not to
examine the durability of response but only the magnitude of response
to LVRS. Numerous studies2
25
have confirmed that peak
functional effects occur by 6 months following surgery, justifying our
selection of this time point.
In conclusion, we have developed a simple, quantitative index of apical perfusion derived by lung scintigraphy that assists in predicting the likelihood of short-term functional response to LVRS. Further study is needed to verify and refine the approach described herein and to compare it to other imaging modalities. Derivation of receiver operating characteristic curves from a larger, prospectively acquired database will be necessary to determine the most appropriate cutoff values for AP% that would optimize the predictive value of the index. Additionally, more information is needed on the relationship between zones of hypoperfusion assessed by scintigraphy and zones of emphysema assessed by CT imaging. The ongoing National Emphysema Treatment Trial has incorporated quantitative perfusion scintigraphy and CT imaging in the evaluation of all participants and should provide an unprecedented opportunity to study the merits of preoperative imaging in the selection of candidates for LVRS.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Received for publication January 18, 2001. Accepted for publication May 3, 2001.
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. Vaughan, I. Oey, A. Nakas, A. Martin-Ucar, J. Edwards, and D. Waller Is there a role for therapeutic lobectomy for emphysema? Eur. J. Cardiothorac. Surg., March 1, 2007; 31(3): 486 - 490. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. C. Mineo, E. Pompeo, D. Mineo, P. Rogliani, C. Leonardis, and I. Nofroni Results of unilateral lung volume reduction surgery in patients with distinct heterogeneity of emphysema between lungs J. Thorac. Cardiovasc. Surg., January 1, 2005; 129(1): 73 - 79. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. F. Gelb and R. J. McKenna Jr Lung Volume Reduction Surgery Update Chest, April 1, 2003; 123(4): 975 - 977. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
