Lung-Volume Reduction Surgery for End-Stage Chronic Obstructive Pulmonary Disease: Health Technology Assessment: Number 10

Abstract

Lung-volume reduction surgery (LVRS) has been proposed as a palliative treatment for selected patients with diffuse emphysema and end-stage chronic obstructive pulmonary disease who have failed conventional therapy. A number of surgical techniques have been used that are designed to reduce lung volume by surgical resection or laser plication. These techniques are designed to restore previous compromised lung elastic recoil so that expiratory airflow obstruction is reduced, respiratory mechanics are improved, and disabling dyspnea is relieved.

Preliminary data derived from both published and unpublished information indicate some favorable short-term benefits. However, objective postoperative data are available for only a small proportion of patients, and long-term followup data are not available. In addition, these surgeries are associated with significant morbidity (and a 6 percent [approximate] surgical mortality) and prolonged hospital stays in a substantial percentage of patients.

Patient selection criteria are heterogeneous and in flux, and controversy continues concerning the most appropriate surgical techniques for various categories of patients. The current data do not permit a logical and scientifically defensible conclusion regarding the risks and benefits of LVRS.

Foreword

The Center for Health Care Technology (CHCT) evaluates the risks, benefits, and clinical effectiveness of new or unestablished medical technologies. In most instances, assessments address technologies that are being reviewed for purposes of coverage by federally funded health programs.

CHCT's assessment process includes a comprehensive review of the medical literature and emphasizes broad and open participation from within and outside the Federal Government. A range of expert advice is obtained by widely publicizing the plans for conducting the assessment through publication of an announcement in the Federal Register and solicitation of input from Federal agencies, medical specialty societies, insurers, and manufacturers. The involvement of these experts helps ensure inclusion of the experienced and varying viewpoints needed to round out the data derived from individual scientific studies in the medical literature.

CHCT analyzed and synthesized data and information received from experts and the scientific literature. The results are reported in this assessment. Each assessment represents a detailed analysis of the risks, clinical effectiveness, and uses of new or unestablished medical technologies. If an assessment has been prepared to form the basis for a coverage decision by a federally financed health care program, it serves as the Public Health Service's recommendation to that program and is disseminated widely.

CHCT is one component of the Agency for Health Care Policy and Research (AHCPR), Public Health Service, Department of Health and Human Services.

• Thomas V. Holohan, M.D., FACP Director Center for Health Care Technology

• Clifton R. Gaus, Sc.D. Administrator Agency for Health Care Policy and Research

• Questions regarding this assessment should be directed to:

• Center for Health Care Technology

• AHCPR

• Willco Building, Suite 309

• 6000 Executive Boulevard

• Rockville, MD 20852

• Telephone: (301) 594-4023

Introduction

Chronic obstructive pulmonary disease (COPD) is the nonspecific terminology commonly used to describe the spectrum of various diseases causing limitation of expiratory airflow, e.g., asthma, chronic bronchitis, and emphysema.(1,2) There is considerable overlap in the specific syndromes of chronic bronchitis or emphysema, and emphysematous COPD describes the syndrome characterized by permanent enlargement of the distal airspaces, with destruction of their walls, and reduction in lung elastic recoil, airway collapse, and irreversible airflow obstruction.(3,4) This emphysema is most often panlobular.(5)

There has been a dramatic increase in the prevalence of COPD and a 22 percent increase in its death rate in the past 20 years, making this condition the fourth leading cause of death in the United States (approximately 85,000/year), with most patients dying within 2 years after medical therapy was judged to be no longer effective.(3,6,7) A further indication of the seriousness of this disease is reflected in data indicating a 3-year mortality of 23 percent for ambulatory patients without evidence of other serious disease, and a substantial hospital mortality (24 percent) for patients admitted to an intensive care unit with an acute exacerbation of COPD.(7,8)

Although the pathophysiology and the exact cause of death in patients with severe COPD is not completely understood, it has been reported that patients commonly die of "respiratory failure" due to "problems with ventilatory mechanics and gas exchange."(9)

The majority of emphysema patients have diffuse disease; however, a small proportion have regional large or giant bullae, in which cases surgical excision or decompression have been reported to achieve a better outcome than that in patients with more generalized disease.(10-13) Paraseptal emphysema is most often associated with bulla formation and represents the predominant lesion in cases most suitable for bullectomy.(5,13,)

Rationale

Chronic and disabling dyspnea is the primary indication for surgery in patients with emphysema, and bullectomy or lung transplantation has been proposed as the only surgical treatment that may benefit selected patients with COPD.( 2-6,14) The basic rationale for bullectomy was initially proposed as the removal of space-occupying lesions to permit more lung to become functional.(15) Subjective and objective improvement after bullectomy has been widely reported and was attributed, in part, to the reexpansion of compressed lung, improved respiratory mechanics, and redistribution of ventilation to more normal lung.(16- 18) In some cases, the improvement is transient because remaining emphysematous areas may gradually form additional bullae after the initial surgery.(4,11,19-25) Whether, in fact, reexpansion of compressed lung tissue actually results in improved postoperative ventilation has been questioned.(10) More recent explanations for the benefits of bullectomy include the restoration of normal architecture and mechanical linkage with the chest wall, permitting deflated rather than compressed lung to regain its elasticity, as the primary benefit of bullous surgery rather than mere removal of space-occupying lesions.( 14,26,27) Indeed, more than 20 years ago, loss of elastic recoil was regarded as a primary mechanism for airflow limitation in COPD.(28) Recent reports have also suggested that lung-volume reduction surgery (LVRS) favorably affects inspiratory mechanics by restoring a more normal configuration of the diaphragm and improving its strength.( 16,18,29)

Published Surgical Experience

Since the early part of this century, various surgical procedures have been implemented for the treatment of COPD, including thoracoplasties, excision of bullae, resection of peripheral lung tissue, lobectomies, and a variety of autonomic denervations, all designed, one way or another, to reduce excessive thoracic volume, permit reexpansion, and restore elastic tension of more normal lung regions, and reduce bronchospasm. (15,30-33) A more novel approach, carotid body resection, designed to relieve dyspnea and applied to thousands of patients with asthma or COPD, was previously assessed, found not to be safe and effective, and virtually abandoned.(34)

Observations were made that surgically reducing the volume of overdistended and inelastic emphysematous lung would reduce expiratory obstruction and increase pulmonary ventilation and could result in long-lasting symptomatic relief of dyspnea. However, the enthusiasm for such surgery was tempered by the significant morbidity and mortality (almost 20 percent) associated with the procedure, and it did not gain favor in the general medical community.(33,35,36)

A 1977 retrospective review of 118 consecutive emphysema patients surgically treated at Toronto General Hospital apparently laid the foundation for the revival of interest in the surgical approach to emphysema, culminating in the dramatically increased interest in current treatments involving lung volume reduction by various techniques almost 20 years after this encouraging report.(37,38) On the basis of the rationale of removing avascular space-occupying areas of emphysematous lung to allow compressed lung tissue to expand and fill the pleural space so the residual elastic tissue will maintain the patency of the terminal bronchioles, Delarue and associates described the results of surgery in 47 patients.(37) Various surgical techniques were used, including lobectomy, bullectomy, and limited resections. Long-term improvement was seen in 45 percent, and perioperative mortality was 21 percent. Forced expiratory volume in 1 second (FEV₁) improved 15 percent and forced vital capacity (FVC) improved 31 percent. Severity of dyspnea was reduced from grade 3 to grade 1 (methodology not described). The 25 patients in the early group (1950- 1964) experienced a 32 percent mortality, whereas 48 percent of patients were reported as improved. The later group of 22 patients (1965-1974) treated by excision of destroyed lung tissue exhibited a 9 percent mortality rate. Nine patients were not improved, three exhibited transient improvement longer than 12 months, and 10 were improved longer than 24 months.

In a review of four large series of bullectomies involving 547 patients, operative mortality ranged from 0-15 percent, reflecting variable patient selection criteria (especially age and severity of disease), with the higher mortality in older patients with more diffuse disease.(39) The major causes of mortality were respiratory insufficiency and pulmonary infection. Reported morbidity ranged from 14-44 percent and were primarily persistent air leaks and infections.

Local excision and plication had been the procedure of choice and postoperative improvements in pulmonary function tests (PFTs) were related to the severity of the diffuse disease, with the best results seen in cases of paraseptal or periacinal emphysema (isolated bullae), and the worst results seen in panacinar or panlobular emphysema (multiple bullae).(5) No improvements were seen where bullae occupied less than one third of the lung, and long-term decline in lung function did not differ from the decline seen with normal aging. In the group of patients with large localized lesions, marked early improvement in PFT was sustained for at least 5 years, then declined in some patients to preoperative levels within 7-10 years. In a report of 15 patients on whom operations were performed for severe dyspnea and diffuse disease, FEV₁ declined an average 101 mL/ yr vs. 28 mL/yr for age-matched normals.(22) In the previously cited review, of 21 late deaths (in 82 cases), 12 were related to lung disease.(39) In nine patients, the last recorded FEV₁ was < 1 L. Mean followup of 9 years suggested surgery may have prolonged life. In a large French surgical series of 423 cases, 11 percent of patients with diffuse disease died in the 1st year, and 65 percent had sustained subjective and PFT improvement at 2 years and 50 percent at 3 years.(39) The consensus of experience suggests that removal of large bullae does not hasten the progression of diffuse emphysema,(5, 11,40) and most patients achieve significant subjective and some PFT improvement for a number of years before experiencing a gradual return of dyspnea.(40,41)

Although no single test is specific for assessment of pulmonary function in patients with emphysematous bullae, sustained or transient postoperative improvement reflects the overall function of the remaining lung tissue and may, with some difficulty, be predicted on the basis of preoperative PFTs.(42- 47) The postoperative symptomatic improvement of most patients does not always correlate with the commonly seen improvement in some PFT, especially tests of airway obstruction (FEV₁ and FVC).(10)

Single or double lung transplantation has become a therapeutic option in selected younger patients with advanced COPD, with an operative mortality in the range of 18 percent, and a 1-year actuarial survival of 68.4 percent.(6,48,49) A recent comparison of early results on exercise tolerance, PFT, and mortality after LVRS in 33 patients with results after transplantation in 66 patients suggested that LVRS may be a "suitable alternative in selected patients eligible for transplantation."(50) This was a retrospective analysis of results after surgery, and comparable outcome measures were evaluated; however, selection criteria of patients were variable, with established criteria being more strict for lung transplantation candidates than those for prospective volume-reduction patients.

As originally described by Abbott and associates,(33) Crenshaw and Rowles,(15) and later refined by Brantigan and Mueller,(32) the unilateral or bilateral operation consisted of reducing the volume of the lung by removing the least functioning areas by means of resection and/or plication, generally around the periphery of the lobe, and surgically denervating the lung. Ideally, the lung volume was reduced to the point at which the remaining lung filled the pleural cavity on full expiration. The procedure permitted local expansion of the adjacent lung tissue, restored a more nearly normal negative relative pressure within the thorax during inspiration, allowed elevation of the diaphragm, improved chest wall motion, and increased ventilation.

The resection of peripheral lung tissue as described by Brantigan and Mueller( 32) was not primarily concerned with the removal of pathologic tissue, but was rather "directed at the restoration of a physiologic principle," i.e., restoration of the more normal circumferential pull on the smaller airways, a more negative relative intrapleural pressure, and improved pulmonary capillary bed circulation. (32,51-53) Brantigan et al(51) emphasized the effects of lung volume reduction on restoring the circumferential pull on the bronchioles, the loss of which was regarded as an important cause of expiratory obstruction in emphysema.

The denervation purportedly resulted in a reduction in bronchospasm and tenacious bronchial secretions. Pulmonary artery pressure (PAP) was reduced (by capillary dilation) and pulmonary circulation increased, all of which served to improve ventilation and relieve dyspnea.(15,30, 32,33,54)

Recently, a similar surgical approach, albeit without denervation, was implemented by Cooper and associates(36) for selected patients with severe COPD with the goal of relieving thoracic distention and improving respiratory mechanics. Although rigorous selection criteria have not been defined, following Brantigan's concept, patients were selected who had an overdistended chest, specific regions of hyperexpansion, relatively absent ventilation, and perfusion in those regions, as well as flattening of the diaphragm.

In this study involving 20 patients with severe COPD, 14 patients required supplemental oxygen (02) during exercise, and six were candidates for lung transplantation. All patients underwent formal supervised exercise rehabilitation for a minimum of 6 weeks (details not provided), and preoperative assessments included PFTs and measurements of the degree of dyspnea and quality of life (QOL).

The surgical plan was to reduce the overall volume of each lung 20-30 percent by resecting the more destroyed portions of lung which, in most patients, predominately involved the upper lobes. The operation was done through a median sternotomy, and followup ranged from 1-15 months (mean 6.4 months). There was no mortality, and prolonged air leak was the most frequently seen morbidity (11/20). Objective and subjective assessments were made at 1, 3, 6, and 12 months postoperatively.

Pulmonary function tests at 6 months for eight patients and between 1 and 6 months for the 12 others indicated an 82 percent increase in FEV₁, a 27 percent increase in FVC, and significant reductions in total lung capacity (TLC), residual volume (RV), and trapped gas. Only two of 14 patients evaluated at 3 or more months after surgery required O₂ during exercise, compared with 14 of 20 patients requiring O₂ before surgery. The mean distance in the 6-minute-walk test increased from 958 feet before exercise rehabilitation, to 1,220 feet just before operation, to approximately 1,400 feet at 3 months postoperatively (15 patients studied), and to approximately 1,600 feet at 6 months (eight patients). Postoperative measurements of dyspnea, using the Modified Medical Research Council (MMRC) Dyspnea Scale and the Mahler Dyspnea Index, indicated moderate improvement, and measurements of QOL, using the Nottingham Health Profile and the Medical Outcomes SF-36 Health Survey (MOS SF-36 Health Survey), were reported as significantly improved. These beneficial effects are thought to be derived from both improved respiratory mechanics and improved ventilation/perfusion distribution in the remaining lung.

Recent publications have described extensive experience and early results with video-assisted thoracoscopic techniques (VATS), with or without lasers, believed by some to be a less morbid and equally effective treatment of diffuse bullous emphysema. (16,55,56)

Patients selected for surgery were regarded as having end-stage emphysema with intractable dyspnea and a poor QOL. The operative procedures were primarily unilateral thoracoscopic laser bullectomy, although lung reduction with staples has more recently become increasingly popular.(38,56-60) More than 500 cases have been reported with variable followup times; outcome measures have included subjective reports and some pulmonary function test results.

Table 1. Unilateral thoracoscopic surgery with postoperative (more...)

Table 1. Unilateral thoracoscopic surgery with postoperative outcomes at 1- 6 months

Author (measurements)	No. patients	Techniques	Dyspnea Scale Borg/MMRC	FEV1	FVC	DLCO mL/min/mm Hg	PO2 mm Hg	Mortality
Little(38) (3 months)		Unilateral thoracoscopic: YAG laser	7.6/3.6 a 62/2.9 b (32)	0.74 L a 0.85 L b (28)	1.82 L a 2.2 L b (28)	10.3 a 13.7 b (10)	59.3 a 63.3 b (22)	5.5% (operative)
Brenner(56) (1-3 months)	24	Unilateral thoracoscopic: C02 laser	NR	0.82 L a 1.18 L b (19)	2.18 L a 2.96 L b (19)	8.13 a 8.27 b (19)	NR	12% (peri-operative)
McKenna(59) (6 months)	72	Unilateral thoracoscopic: YAG laser (33) Staples (39)	24% improved more than one grade b 66% improved more than one grade b	13.4% improvement b 32.9% improvement b (72)	6% increase b 21% increase b (72)	NR	NR	11% (within 3 months)
Wakabayashi(58) (3-6 months)	443	Unilateral thoracoscopic: YAG laser	87.4% improved b (191)	23.6% of predicted a 31.0% of predicted b (202)	55.8% of predicted 67.4% of predicted (203)	23.6% of predicted a 33.7% of predicted b (123)	65.0 a 66.0 b (96)	4.8% (operative)

a Preoperative.

b Postoperative.

Abbreviations: DLCO = carbon monoxide diffusing capacity of the lungs; FEV₁ = forced expiratory volume in 1 second; FVC = forced vital capacity; MMRC = modified medical research council scale; NR = not reported; PO₂ = partial pressure of oxygen; YAG = yttrium-argon-garnet laser.

Number in parentheses = number of patients.

Information Obtained From Other Sources

As part of the assessment process, the Center for Health Care Technology (CHCT) routinely publishes a notice in the Federal Register soliciting information from interested parties regarding the technology under review. Letters may also be written directly to individuals, institutions, or medical centers using the technology. The Federal Register notice regarding LVRS was published on November 15, 1995, and letters were written to 17 institutions and professional organizations known to be involved with LVRS. These letters requested information regarding the risks, benefits, and costs associated with this treatment. Such solicitations provided the opportunity for the public and the professional medical community to participate in the process; material obtained is often cited or reproduced in the completed assessments. In the case of LVRS for COPD, it was apparent that far more patients had received operations than were represented in the published medical literature, and it was believed that review of such unpublished material was thus very important. In the solicitations, 27 institutions responded and provided opinion, information, or draft reports to CHCT. That material is reviewed in this section of the assessment; because this information is not presently available in the medical literature for consideration by interested parties, the review was accomplished with more emphasis upon the citation of detailed data than is ordinarily provided in the assessment process.

The National Jewish Center for Immunology and Respiratory Medicine stated that the paucity of published data concerning the safety and efficacy of LVRS, and absence of claims that this surgery prolonged life, suggest that LVRS should not be recommended outside of a randomized controlled clinical trial.

The University of California, Los Angeles Medical Center listed the following questions they believed were unanswered concerning LVRS:

• What are the best procedures?

• What is the mechanism for improvement?

• What are selection/exclusion criteria?

• Which preoperative tests are required?

• What are the long-term results?

The University of California, San Diego Medical Center stated that the rational, potential role for LVRS should be as one component of an evaluation and treatment program that has comprehensive pulmonary rehabilitation as the first step. Evaluation of LVRS would best be accomplished by a randomized clinical trial or a matched observational study comparing the effect of pulmonary rehabilitation with or without surgery.

The University of Washington School of Medicine reported their belief that it is not yet clear that the results of LVRS generated in academic centers will translate into similar results in community hospitals. Lung-volume reduction surgery should at present be regarded as experimental surgery and be limited to those centers capable of doing the research. Who might benefit and who might be at the greatest risk must be defined further.

The Harvard Medical School, Deaconess Hospital believes that the early enthusiastic reports of LVRS will be tempered by time and wider application, and only a controlled trial will answer the question of the usefulness of the operation.

The California Thoracic Society reported that they view LVRS as an addition to a pulmonary rehabilitation program rather than an independent procedure, and a review of the indications, risks, and benefits in a controlled trial is indicated before widespread application of these surgeries.

The American Thoracic Society forwarded correspondence noting that their board of directors had reviewed LVRS. They believe that although early experience has "created a long list of unanswered questions" regarding issues such as indications for LVRS, preoperative assessment, surgical technique, and long-term efficacy, they do not consider the procedure to be "experimental." The Society recommended that the Health Care Financing Administration establish a "trial period" during which all LVRS procedures will be performed "only on patients who are part of carefully planned, well-defined protocols," and that this could be accomplished by creation of a national registry.

Table 2. Inclusion criteria, lung reduction surgery (more...)

Table 2. Inclusion criteria, lung reduction surgery

Criteria	Temple Univ.	Mayo Clinic	Rose Medical Center	St. Louis Univ.	Yale Univ.	Medical Center of Central Mass.	Hartford Thoracic and Cardiovascular Group	Baylor Univ.	Columbia Univ.	Chapman Medical Center
FEV1, percent of predicted	<30	<35	<40	<35			<30-35	<50
Preoperative pulmonary rehabilitation program	Yes	Yes	Yes	Yes	Yes	Yes				a
RV, percent of predicted			>150			>150
Hyperinflation	Yes				Yes				Yes
Poor diaphragm movement					Yes	Yes			Yes
TLC, percent of predicted	>120	>120						>110
Bullous emphysema	Yes			Yes			Yes
V-P defects	Yes			Yes	Yes
O2 dependence
Rehabilitation "potential"							Yes
Bilateral emphysema on CXR or CT			Yes
Max inspiration pressure, percent of predicted	<50
Nonbullous emphysema		Yes
NYHA Class III-IV	Yes
6-min walk

Table 2. Inclusion criteria, lung reduction surgery (more...)

Table 2. Inclusion criteria, lung reduction surgery (continued)

Criteria	Foote Hospital, Jackson, MS	Midwest Pulmonary Consultants, Kansas City	Univ. of PA	Univ. of Louisville	Washington Univ., St. Louis, MO	Wakabayahi Institute	St. Paul Heart and Lung Center, MN	Univ. of Pittsburgh
FEV1, percent of predicted	20-35	15-30	<40	<35	15-35	<35		<30
Preoperative pulmonary rehabilitation program	Yes	Yes	Yes		Yes		Yes
RV, percent of predicted	>150	>200	>200	>200
Hyperinflation		Yes		Yes		Yes
Poor diaphragm movement	Yes
TLC, percent of predicted					>120
Bullous emphysema						Type 3 or Type 4 bullae
V-P defects			Yes
O2 dependence	Yes
Rehabilitation "potential"
Bilateral emphysema on CXR or CT
Max inspiration pressure, percent of predicted
Nonbullous emphysema
NYHA Class III-IV
6-min walk						<1,200 ft

a The only criteria specified were: "decreased" quality of life; "marked decrease" in dyspnea index; and under "maximal medical management."

Abbreviations: CT = computed tomography; CXR = chest x-ray; FEV₁ = forced expiratory volume in 1 second; NYHA = New York Heart Association; RV = residual volume; TLC = total lung capacity; V-P = ventilation-perfusion.

Table 3. Exclusion criteria, lung reduction surgery (more...)

Table 3. Exclusion criteria, lung reduction surgery

Criteria	Yale Univ.	Rose Medical Center	Mayo Clinic	Chapman Medical Center	Temple Univ.	Baylor Univ.
Smoking	Yes	Yes	Yes	Yes	Yes	Yes
Metastatic cancer	Implied	Yes	Yes	Yes	Implieda	Yes
PAP increase	Yes	<40 mm	(Relative)		<50 mm	"Elevated"
Serious comorbid disease (e.g., CVD)	Yes	Yes	Yes	Yes	Implieda	Yes
Intrathoracic cancer	Isolated nodule in operative area permissible	Yes	Yes	Yes	Isolated nodule permissible	Yes
Age	<75	<80	(Relative)	<75
Bronchitis/bronchospasm	Yes	Yes				Yes
PCO2 elevation	Yes (<50 mm)	<55 mm	(Relative)	<55 mm
Active steroid use	"Low dose" <10 mg prednisone/day permitted	<20 mg/day prednisone
Ventilator dependency	Yes			Yes
Morbid obesity	Yes
Psychiatric or social dysfunction		Yes	(non-compliant)		Yes
Wheelchair or bed-bound	Not ambulatory	Yes
Severe hypoxemia (PiO2/FiO2)			(Relative)		Yes
Prior chest surgery or pleurodesis			Yes
Chest wall deformity		Yes
Debilitation, severe cachexia	Yes				Yes (<80 ideal body wt)
Cannot perform 6-min walk			Yes
Excessive sputum production
Lung disease other than emphysema			Yes
Incapable of independent living		Yes
Presence of lung mass				Yes
<5-cm bulla				Yes
FEV1 <15 predicted						Yes
Drug dependence			Yes

Table 3. Exclusion criteria, lung reduction surgery (more...)

Table 3. Exclusion criteria, lung reduction surgery (continued)

Criteria	Hartford Thoracic and Cardiovascular Group	Columbia Univ.	Medical Center of Central Mass.	Univ. of Louisville	Washington Univ.	Univ. of Penn.
Smoking	Yes	Yes	Yes	Yes	Yes	Yes
Metastatic cancer		Yes		Yes	Implied
PAP increase	.<35-40 mm		"Significant increase"	<35 mm	<35 mm	<50 mm
Serious comorbid disease (e.g., CVD)	Yes			Yes	Yes
Intrathoracic cancer					Implied
Age				<75	<75	<75
Bronchitis/bronchospasm		Yes		Yes	Yes
PCO2 elevation				<55 mm	<50 mm	<55 mm
Active steroid use	<10-15 mg/day prednisone		??b	"High dose"
Ventilator dependency	Yes			Yes
Morbid obesity		Yes		Yes		Yes
Psychiatric or social dysfunction						Yes
Wheelchair or bed-bound				Yes
Severe hypoxemia (PiO2/FiO2)				PO2 <45
Prior chest surgery or pleurodesis	Yes				Yes	Yes
Chest wall deformity				Yes	Yes
Debilitation, severe cachexia						Yes
Cannot perform 6-min walk				<30 improvement after rehab		<600 ft after rehab
Excessive sputum production		Yes				Yes
Lung disease other than emphysema
Incapable of independent living
Presence of lung mass
>5-cm bulla
FEV1 <15 predicted
Drug dependence

Table 3. Exclusion criteria, lung reduction surgery (more...)

Table 3. Exclusion criteria, lung reduction surgery (continued)

Criteria	Wakabayashi Institute	St. Paul Heart and Lung Center	Univ. of Pittsburgh	Foote Hospital, Jackson, Miss.	Midwest Pulmonary Consultants, Kansas City	St. Louis Univ.
Smoking	Yes	Yes	Yes	Yes	Yes	Yes
Metastatic cancer	Implied	Yes	Implied	Implied	Implied
PAP increase		±	>55 mm	Yes		>50
Serious comorbid disease (e.g., CVD)	Yes	Yes	Yes		Yes
Intrathoracic cancer	Implied	Yes		Implied	Implied
Age		>75			>70	>80
Bronchitis/bronchospasm	Yes			Yes
PCO2 elevation						>50
Active steroid use		>20 mg/day	>20 mg/day
Ventilator dependency	Yes		Yes
Morbid obesity			Yes			Yes
Psychiatric or social dysfunction	Yes
Wheelchair or bed-bound	Yes		Yes
Severe hypoxemia (PiO2/FiO2)	PO2 <40
Prior chest surgery or pleurodesis
Chest wall deformity
Debilitation, severe cachexia
Cannot perform 6-min walk
Excessive sputum production
Lung disease other than emphysema	Yes
Incapable of independent living
Presence of lung mass
<5-cm bulla
FEV1 <15 predicted
Drug dependence

a Implied, based upon exclusion of "significant" extrapulmonary end-organ dysfunction expected to limit life.

b Inclusion criteria included "no active recent steroid use," but relative exclusion criteria included 10 mg/day or greater of prednisone.

Abbreviations: CVD = cardiovascular disease; FEV₁ = forced expiratory volume in 1 second; FiO₂ = fraction of oxygen in inspired air; PAP = pulmonary artery pressure; PCO₂ = partial pressure of carbon dioxide; PiO₂ = partial pressure of inspiratory oxygen; PO₂ = partial pressure of oxygen; rehab = rehabilitation.

Yale University provided information to the effect that they perform median sternotomies for LVRS in patients with severe COPD and have completed approximately 30 cases. They stated their inclusion and exclusion criteria (see Tables 2 and 3) and described an ongoing study of the procedure. They expressed their opinion that Medicare and private insurers should cover the procedure because it "improves the QOL of many of the patients." No specific data were provided documenting the observed risks or benefits of LVRS or costs of the procedure.

The Mayo Clinic forwarded a letter describing their belief that LVRS is of benefit to a number of patients with end-stage emphysema and provided their inclusion and exclusion criteria (see Tables 2 and 3). The Mayo Clinic reported they had performed LVRS via median sternotomy in 27 patients. Two deaths occurred in this series. Although individual patient-specific data were not provided, overall 6-month followup evaluations were reported to have shown a mean improvement of 42 percent in FEV₁ and an improvement of 55 percent in 6-minute-walk distance from the preoperative (postpulmonary rehabilitation) levels. The Mayo Clinic correspondents believe that a prospective, randomized trial is the most appropriate evaluative process for LVRS. No cost data were provided.

The Wakabayashi Institute responded with two documents that noted that 1,426 "bullectomies or pneumoplasties" had been performed by laser thoracoscopy on 1,257 patients in several different institutions (UC California, Irvine Medical Center [IMC]; Chapman General Hospital [CGH]; and Western Medical Center/ Santa Ana). They gave patient inclusion and exclusion criteria as cited in Tables 2 and 3. Further information was provided on the patient population operated at IMC, cited in the text (page 18 of correspondence) as numbering 614 patients; however, data provided in tabular form (page 6) indicated that 627 patients had undergone operations at IMC. Reported preoperative PFTs did not match either number (i.e., 648 preoperative PaO₂ and 650 FEV₁ studies). The 614 patients were reported as having been subject to totals of 680 procedures (page 6) or 663 procedures (page 19). Tracheostomies were reported as necessary in both 17 and in 18 patients. A 3-month mortality rate of 6.3 percent was reported for 663 "cases" (presumably, this referred to the number of procedures and not patients) operated at IMC; a separate mortality rate of 6.1 percent was reported for 668 "cases" operated at a different medical center (CGH). Nonfatal complications included 45 bronchopleural fistulae, 65 delayed pneumothoraces (denominator unclear in context), and unspecified numbers of loculated pneumothorax, infection, gastrointestinal (GI) complications, and "others." A risk- factor analysis was presented that concluded that variables positively associated with the 3-month mortality included "end-stage cardiomyopathy," ventilator dependency, PaO₂ < 40 mm, RV < 100 percent predicted, TLC < 80 percent predicted, and nonambulatory status. The document noted that only 50 percent of patients responded to the request to provide followup studies; the followup period was described as "up to 20 months." Patients were mailed "standardized questionnaires" (not further described); 475 were reported as having stated their breathing was better. Preoperative and postoperative PFTs were reported as available an unspecified time period after surgery in 290 patients; however, only FVC (percent predicted) were provided for that number. FEV₁ were reported for 289 patients, PaO₂ for 178, and PCO₂ for 177. A subgroup analysis was performed on 110 patients whose PFTs were performed at IMC. Five-year survival data of "the entire series over a 5-yea period" (May 12, 1989 to June 30, 1995) were presented for patients who had undergone unilateral and bilateral operations. The data enumerated survivors among a group of undefined "eligible patients," including 772 subjected to unilateral surgery and 202 subjected to bilateral surgery. A second table gave 5-year survival statistics for 732 patients, "corrected" for exclusion of operative mortality. Because of the variation in numbers of patients reported, and the uncertainty of the significance of the outcomes given for all the various subsets of patients referenced, pre- and postoperative data were not further analyzed for this center. The Wakabayashi Institute reported that "Irvine Medical Center does not have the data to evaluate or compare costs at this time."

The Midwest Pulmonary Consultant Group of Kansas City, MO reported that they had performed 91 LVRS procedures over a 1-year period. Operative techniques were not described. Their inclusion and exclusion criteria are noted in Tables 2 and 3. They provided summary data indicating that the mean increase in FEV₁ was 45 percent, although the postoperative followup time was not specified. The reported mortality was 6.6 percent, and average length of stay (LOS) was 14 days. No cost data were provided.

Table 4. W.A. Foote Hospital: Improvement at 3-6 months (more...)

Table 4. W.A. Foote Hospital: Improvement at 3-6 months (percent)

		Postoperative
Parameter	Preoperative	3 months	6 months
FEV1 (percent predicted)	28 (55)	50 "improved" (42)	60 "improved" (20)
RV (percent predicted)	197 (54)	64 "improved" (42)	68 "improved" (20)
6-min walk	820b [200-1,484] (43)	<20 increase in 42 of patientsc (43)	<20 increase in 63 of patients (16)
"Dyspnea"a	NR	72 "improved" (32)	65 "improved" (20)

a Not further defined.

b Assumed to be in feet; not specified.

c That is, 58 percent of patients had <20 percent improvement in the 6-minute-walk distance.

Abbreviations: FEV₁ = forced expiratory volume in 1 second; NR = not reported; RV = residual volume.

Number in parentheses = number of cases.

Table 5. Data from the Medical Center of Central Massachusetts (more...)

Table 5. Data from the Medical Center of Central Massachusetts

		Postoperative
Parameter	Preoperative	3 months	6 months
Fatigue scale	NR	20/32 improveda	13/20 improved
Dyspnea scale	NR	23/32 improved	13/20 improved
6-min-walk mean distance	774 ft (55)	904 ft (42)b	1,164 ft (20)

a Measured by Guyatt's CRQ.

b Only 18 of 42 cases improved their walk distance.

Abbreviation: NR = not reported.

Number in parentheses = number of cases.

Since December 1994, the Center has performed LVRS on 58 patients; 55 left the hospital -- 36 went home, 10 went to a subacute care unit, and three died. Three patients were ventilator-dependent preoperatively and six postoperatively; five were later weaned. Fifty cases were performed via median sternotomy. The Center provided no cost data.

The Cardiothoracic and Vascular Surgeons Ltd., St. Joseph's Hospital, Phoenix, AZ provided inclusion and exclusion selection criteria (see Tables 2 and 3). They further provided a protocol used at St. Joseph's, and noted that they perform LVRS via median sternotomy. The enclosed informed consent form includes the comment that there is a "10-15 percent possibility of mortality" from the operation. They provided summary cost data, and reported that the average surgical fee is $3,750,13 13 and average hospital cost is $48,666.

The Cardiothoracic Associates of Sandusky, OH reported having completed 12 LVRS procedures, with one death in the series. Excluding the single death, 10 procedures were accomplished by thoracoscopy, and one by thoracotomy. All patients were required to complete a 6-week preoperative pulmonary rehabilitation program. They reported an average LOS of 10.1 days. The Associates used the MOS SF-36 Health Survey to evaluate QOL, and reported seven of nine cases showed "significant improvement," but did not provide specific data on scale scores. Two patients are presently less than 3 months postsurgery. The Associates did not provide cost data.

Table 6. Temple University: 3-month followup, nine (more...)

Table 6. Temple University: 3-month followup, nine cases

Parameter	Preoperative mean	3-month postoperative mean
FEV1 (L)	0.67 ± 0.2	0.93 ± 0.34
TLC (L)	7.64 ± 1.5	6.08 ± 1.06
6-min walk (m)	236 ± 74	377 ± 55
Total exercise time (min)	5.2	8.3

Abbreviations: FEV₁ = forced expiratory volume in 1 second; TLC = total lung capacity.

Three ventilator-dependent patients were reported as weaned and at home. Of eight patients with pulmonary nodules, carcinoma was found in two, one nodule was not located at operation, and five lesions were benign; one patient died of metastatic disease. Temple University's patient inclusion and exclusion criteria are summarized in Tables 2 and 3. The Sickness Impact Profile (SIP) was used to evaluate QOL in 10 LVRS patients (mean age 58 years). The overall SIP score was reported as having exhibited increases ranging from 24-100 percent in eight cases, and decreases from 25-150 percent in two. The authors noted that "the overall variability in the perception of their disability...was large." Although they expressed belief in the utility of LVRS, they noted that "long-term benefits of this procedure and the risk-benefit ratio remain to be determined;" and, as regards the improved physiologic and functional measures, "further investigation is required to determine the stability of these...improvements over time." The correspondents stated in strong terms their belief that a randomized, prospective trial was critical in the evaluation of LVRS. No cost data were provided by Temple.

Table 7. University of Pennsylvania: 3-month postoperative (more...)

Table 7. University of Pennsylvania: 3-month postoperative results

Procedure	FEV1 percent predicted preoperative	FEV1 percent predicted postoperative	RV percent predicted preoperative	RV percent predicted postoperative	6-min walk preoperative	6-min walk postoperative
Median sternotomy (60)	27 (39)	38 (39)	237 (36)	159 (36)	1,032 ft (27)	1,189 ft (27)
VATS, unilateral (12)	28 (11)	34 (11)	235 (11)	210 (11)	862 ft (4)	1,017 ft (4)
VATS, bilateral (22)	21 (16)	34 (16)	265 (13)	198 (13)	1,024 ft (10)	1,335 ft (10)

Abbreviations: FEV₁ = forced expiratory volume in 1 second; RV = residual volume; VATS = video-assisted thoracoscopic surgery.

Number in parentheses = number of cases.

The correspondents noted that 10 percent of the sternotomy and 18 percent of the VATS cases exhibited decreased FEV₁ postoperatively; 43 percent of the former and 45 percent of the latter had increases equal to or greater than 40 percent. About one-third of cases had less than a 20 percent improvement in FEV₁; however, no preoperative characteristics appeared to predict response. An abstract that had been included contained the statement that "We performed. ...excision of emphysematous bullous disease in 17 patients...." This again raises the issue as to whether or how LVRS is distinctly different from bullectomy. Patient selection criteria are seen in Tables 2 and 3. No cost data were supplied by the University of Pennsylvania.

Table 8. Columbia University: 3-month followup data, (more...)

Table 8. Columbia University: 3-month followup data, lung-volume reduction surgery

	No. cases	FEV1 (percent predicted)	FVC (percent predicted)	6-min walk (ft)
Preoperative	53	22 ± 8	49 ± 17	605 ± 346
Postoperative	53	34 ± 14	64 ± 17	987 ± 335

Abbreviations: FEV₁ = forced expiratory volume in 1 second; FVC = forced vital capacity.

The mean duration of followup was 7.1 months, with a range of 1-16 months. Kaplan-Meier actuarial survival was reported as 90 percent at 3 months and 83 percent at 1 year. The median LOS was 17 days; 54 percent of patients had "prolonged" air leaks. Seven patients required tracheotomy, and five required continued intubation. The perioperative mortality rate was 7 percent; six deaths occurred within 30 days of surgery, and five deaths occurred 1.5-7 months postoperatively. Fifteen patients had pulmonary or mediastinal tumors found during preoperative workup, and nine were operated on with tumor resection and LVRS. Five had malignant disease (three non-small-cell and one bronchoalveolar carcinoma, and one invasive thymoma). The respondents noted that in this group there was one operative death and no late deaths after a mean followup of 8 months. Although "severe hypercapnia," FEV₁ < 500 cc, failure of pulmonary rehabilitation, and daily steroid dose >10 mg/day of prednisone seemed to be associated with slightly worse survival, no statistically significant relationships were found. Patient selection criteria are noted in Tables 2 and 3. Columbia also provided additional data after the deadline had been established in the Federal Register notice and correspondence. The information added four cases to the previous total of 85, and provided results for "the best 50 percent of patients followed 3 months or more." This material did not substantively alter the significance of the summarized data, nor did it appear that conclusions that might be drawn from that evidence would be affected. Columbia provided no cost data.

Table 9. Data from Rose Medical Center: Groups A and (more...)

Table 9. Data from Rose Medical Center: Groups A and B

			Postoperative
Group	Preoperative	3 months	6 months	12 months
Group A (laser)
No. cases	120	90	51	29
FEV1 (percent predicted)	26	30	29	28
RV (percent predicted)	239	214	223	257
No. cases	119	88	49	27
PO2 (mm)	50	49	50	52
PCO2 (mm)	41	39	36	36
Group B (staple)
No. cases	97	40	12
FEV1 (percent predicted)	26	29	32
RV (percent predicted)	256	220	170
No. cases	97	40	11
PO2 (mm)	49	50	55
PCO2 (mm)	38	37	35

Abbreviations: FEV₁ = forced expiratory volume in 1 second; PO₂ = partial pressure of oxygen; PCO₂ = partial pressure of carbon dioxide; RV = residual volume.

Table 10. Dyspnea index (0 = none; >10 = severe; (more...)

Table 10. Dyspnea index (0 = none; >10 = severe; 20 = extreme)

			Postoperative
Group	Preoperative	3 months	6 months	12 months
A	15.5 (68)	11.5 (71)	11.0 (46)	10.2 (23)
B	14.3 (92)	10.5 (41)	10.7 (12)

Number in parentheses = number of cases.

Although followup was more limited for the stapled than laser-treated patients, the data provided did not appear to favor either technique.

Information regarding patient inclusion and exclusion criteria is included in Tables 2 and 3. The Rose Medical Center also provided QOL data based upon patients' scores on the SF-36 Health Survey. Scale scores were separately provided for the laser-treated group (63 patients preoperative, and 69, 41, and 23 cases for 3, 6, and 12 months followup, respectively) and for the staple-treated group (93 preoperative, and 40 and 11 cases at 3 and 6 months followup, respectively). Three-month followup thus encompassed 109 of the 272 cases. In general, scores improved after surgery. However, this instrument has been subject to extensive psychometric analyses regarding its validity, reliability, precision, and standardization. The SF-36 Health Survey Manual and Interpretation Guide (1993; Medical Outcomes Trust, Boston, MA) indicates that for measuring changes in one group over time, the sample sizes needed to assure the reliability of each of the scale differences reported were, in the large majority of cases, much greater than those accumulated by the Center. The Rose Center reported a perioperative mortality of 5.6 percent; whereas increasing age was reported to have had a significant negative effect on survival, high dyspnea scores and low O₂ saturation did not. No predictive value regarding perioperative death was found for PFTs, arterial blood gas (ABG) results, or dosage of prednisone. The Rose Medical Center noted that the hospital costs for LVRS averaged $21,158. This figure did not include physicians' or surgeons' fees or the costs of preoperative outpatient evaluative studies.

Table 11. University of Pittsburgh: 3-month postoperative (more...)

Table 11. University of Pittsburgh: 3-month postoperative followup

	Unilateral			Bilateral
	Preoperative	Postoperative	Preoperative	Postoperative
No. cases	43	43	17	17
FEV1 (percent predicted)	31 ± 10	39 ± 13	27 ± 8	38 ± 10
TLC (L)	8 ± 1.6	7.3 ± 1.6	7.2 ± 1.2	6 ± 1.6
RV (L)	5 ± 1.1	4.1 ± 1	4.8 ± 0.9	3.3 ± 0.9
PO2 (mm)	66 ± 10	66 ± 11	62 ± 12	69 ± 12
Transitional dyspnea index		"Improved"		"Improved"

Abbreviations: FEV₁ = forced expiratory volume in 1 second; PO₂ = partial pressure of oxygen; RV = residual volume; TLC = total lung capacity.

A separate draft report noted that of 95 cases referred for lung transplant, 35 transplant candidates received LVRS. The description of the procedures chosen was somewhat confusing; the numbers of patients subjected to the various surgical techniques do not match unless one assumes that patients undergoing operation by median sternotomy were in the one instance included in the category of "endostapler resection" cases. The thoracoscopic surgery was unilateral in 26 and bilateral in five cases. Followup evaluation was reported at 3 months in 30 of the 35 patients. These data are not reproduced herein, because they appear to represent a subgroup of the larger number represented in the table. Preoperative variables were analyzed, and none appeared to have statistically significant associations with outcomes. The authors concluded, however, that the use of the laser alone and preoperative hypercarbia were associated with poorer outcomes (although they did not include hypercarbia in their exclusion criteria).

A separate short draft report on a subset of 20 cases operated at Pittsburgh observed that the 6-minute-walk distance increased from a preoperative value of 819-916 feet postoperatively. An additional unpublished paper provided a review of the current state of the art concerning LVRS, but no primary data. The authors remarked that "long-term followup studies will be necessary to define its [LVRS] ultimate utility." They also noted that it is not yet clear which surgical approach provides the best benefit/risk ratio, nor what the duration of any improvements is likely to be. The review concluded by noting that long-term improvements "have not been established." The University of Pittsburgh provided no information on costs.

Table 12. Baylor University: Increases in forced expiratory (more...)

Table 12. Baylor University: Increases in forced expiratory volume in 1 second (percent predicted)

		Percent increase
Procedure	3 months	6 months	12 months
Laser	10.4 (31)	11.3 (18)	12.3a (15)
Laser and resection	19.1 (12)	7.2 (5)	-6 (1)
Thoracoscopic resection	9.3 (6)	17 (4)
Median sternotomy	37.7 (29)	64.2 (11)
Total	21.8 (78)	26.7 (38)	12.2 (15)

a Note that, based upon the reported mean preoperative value, a 12 percent increase in FEV₁ ( percent predicted) at 12 months corresponds to an absolute increase from 28-31.4 percent.

Number in parentheses = number of cases.

Table 13. Emphysema and lung-volume reduction surgery (more...)

Table 13. Emphysema and lung-volume reduction surgery

Characteristics	Data
Agea	65.7 ± 10
FEV1 (percent predicted)a	28 ± 11.5
TLC (percent predicted)a	130 ± 22
6-min walk (ft)a	840 ± 351
PO2 (mm)a	64 ± 12
PCO2 (mm)a	44 ± 8
Length of stay	21 ± 15 days
Mortality	5
Morbidity	43

a Preoperative.

Abbreviations: FEV₁ = forced expiratory volume in 1 second; PCO₂ = partial pressure of carbon dioxide; PO₂ = partial pressure of oxygen; TLC = total lung capacity.

Baylor also noted interest in "a randomized study which has been proposed at the local Veterans Administration Hospital," which they believe will provide important information regarding results obtained with medical therapy vs. medical therapy with surgery. Also included in the material submitted was the Institutional Review Board approval for a record review of 203 patients who have undergone LVRS at Baylor.

Patient survey forms, provided by the Baylor correspondents, included seven forms entitled "Lung Volume Reduction Postoperative Evaluation" and 23 forms entitled "Baylor College of Medicine/The Methodist Hospital Emphysema Patient Questionnaire." It was difficult to form any conclusions based on the small number of surveys (representing 3-11 percent of the total number of patients subjected to LVRS) and the fact that not all items were completed in the survey forms. Notwithstanding, five of 23 respondents noted O₂ dependence preoperatively and O₂ independence postoperatively; eight reported that dyspnea was (somewhat) relieved by LVRS, 10 were able to climb a single flight of stairs after LVRS, and 13 regained the ability to walk one block on level ground. However, perceptions of overall health status were not consistently related either to symptoms or to functional capabilities.

Baylor provided no information on the costs of LVRS.

The Chapman Medical Center of Orange, CA stated that they had demonstrated the safety and effectiveness of LVRS "through scientific trials spanning a 300-patient population base" over 2 years. The summary statement of the Chapman correspondents included the following comments:

• Mortality rate of 3.5 percent in "the high-risk group" (not specifically defined).

• Lower morbidity with staple device "laser bullectomy."

• Bilateral staple reduction is the procedure of choice.

• Bilateral staple approach results in O₂ independence in 80 percent of patients, and steroid independence in 86 percent.

Table 14. Data from the Chapman Medical Center

Table 14. Data from the Chapman Medical Center

Procedure	Mean LOS	O2 independence a	FEV1 improvement at 6 months b	Operative mortality
Laser	10.6	52%	0.09 L a (13 ± 6%)	0 b
Staple	13.7	88%	0.35 L c (33 ± 5%)	2.5%

a No. of cases: 26.

b Three laser patients died postdischarge.

c No. of cases: 36.

Abbreviations: LOS = length of stay; FEV₁ = forced expiratory volume in 1 second.

The authors noted that their procedure of choice is now bilateral staple LVRS; they prefer thoracoscopy to median sternotomy.

Table 15. Outcomes, unilateral/bilateral lung-volume (more...)

Table 15. Outcomes, unilateral/bilateral lung-volume reduction surgery

	1-yr mortality	D/C steroid use	D/C O2 use	Gr 3/4 dyspnea postoperative	Mean improvement in FEV1
Unilateral	17	54	36	44	57a
Bilateral	1	68	68	12	31a

a However, tabular data (Table 5) in this report listed FEV₁ improvement as 22 percent (27-33 percent predicted) for unilateral cases and 52 percent (25-38 percent predicted) for bilateral cases.

Abbreviations: D/C = discontinue; Gr = grade.

It was difficult to be certain what the followup period was for these cases. The text refers to the fact that postoperative PFT results did not differ between "initial" followup (mean 80 days) and "subsequent 6-month followup" (mean 163 days), and that steroid independence was evaluated at followup "1- 4 months postoperatively."

The MOS-36 Health Survey and the dyspnea index were reported as measured before surgery and 6 months after surgery. However, the MOS SF-36 Survey scores were not reported, nor were the actual numbers of patients evaluated at specific time periods after LVRS. None of the data included the precise duration of followup.

An additional separate enclosed draft report stated, in the introduction section, that LVRS was performed "in 300 cases." In the results section, the document specified that surgery was performed "on 286 patients" but provided results in terms of changes in FEV₁ in only 79 patients at "3-12 months following LVRS." No outcomes data were provided for the remaining 207 patients operated. Chapman provided no cost data.

St. Louis University Health Sciences Center provided information on LVRS in 75 patients. The majority (71 cases) underwent single unilateral (66) or sequential bilateral (5) thoracoscopic stapling procedures. Bovine pericardium was not routinely used. Four patients were operated on using the median sternotomy technique. The data submitted indicated that, as with most institutions performing LVRS, patient selection and operative technique have evolved as experience has accumulated. Current patient inclusion and exclusion criteria were provided and appear in Tables 2 and 3. Preoperative evaluation included PFTs, chest CT scan, ventilation-perfusion scans, and echocardiograms. Patients with "target zones," defined as areas of bullous emphysema with diminished or absent perfusion, were accepted for operation if they met the other selection criteria. Preoperative pulmonary rehabilitation was required and was followed by reassessment with PFTs, 6-minute walk, ABG, and exercise stress test with collection of exhaled gas.

Table 16. St. Louis University Medical Center: Unilateral (more...)

Table 16. St. Louis University Medical Center: Unilateral thoracoscopic lung-volume reduction surgery^a

		Preoperative			3 months postoperative
Variable	Sample size	Mean (SD)	Range	Sample size	Mean (SD)	Range
Age	60	62 ± 9	31-78	Same	Same	Same
FVC (percent predicted)	60	58 ± 16	23-104	45	68 ± 17	32-111
FEV1 (percent predicted)	60	25 ± 10	8-50	45	32 ± 11	12-67
MVV (percent predicted)	60	24 ± 10	8-54	45	32 ± 14	10-65
RV (percent predicted)	58	254 ± 79	106-438	26	202 ± 50	121-305
DLCO (percent predicted)	57	33 ± 14	3-68	34	40 ± 21	7-130
PCO2	59	38 ± 7	27-57	28	36 ± 8	25-58
PO2	59	64 ± 10	40-92	28	69 ± 11	49-93
6-min walk (ft)	59	804 ± 349	0-1,601	32	983 ± 309	0-1,700

a Followup duration was reported as "3-6 months," because specific numbers followed at 6 months were not available (personal communication, Mar 13, 1996). All are presented as 3-month followup.

Abbreviations: DLCO = carbon monoxide diffusing capacity of the lungs; FEV₁ = forced expiratory volume in 1 second; FVC = forced vital capacity; MVV = maximum voluntary ventilation; PCO₂ = partial pressure of carbon dioxide; PO₂ = partial pressure of oxygen; RV = residual volume.

The baseline dyspnea index (BDI) questionnaire was reported to have shown moderate to severe dyspnea preoperatively. Postoperative subjective improvement was assessed using the transition dyspnea index (TDI), and 90 percent of patients were reported as having shown "substantial" improvement, although the sample size so evaluated at 3 months was not specified.

Several abstracts and draft documents were included, many describing LVRS as experience and numbers of cases accumulated and thus adding little to the more comprehensive data. One draft document, describing unilateral VATS lung reduction with stapling in 50 cases, reported on 25 who had been followed for at least 3 months. In that subset, 10 of 21 (48 percent) O₂-dependent patients were able to discontinue O₂.

The preference for unilateral LVRS, and the results obtained with that procedure, contrast with information supplied and conclusions drawn by other correspondents such as the University of Pennsylvania and the Chapman Medical Center, but is similar to the practice and experience of institutions such as the Rose Medical Center (vide supra). Descriptions of the surgery, selection criteria that included "bullous emphysema" (identified by CT [computed tomography] scan), and submitted drafts that described VATS in "patients with bullous and nonbullous end-stage emphysema" serve to make the distinction between lung-volume reduction surgery and bullectomy somewhat obscure. The St. Louis University Medical Center provided no cost data.

Several centers provided information after the submission deadline CHCT had established in the Federal Register notice and correspondence.

The Cardiac Surgical Associates at United Hospital, St. Paul, MN noted that they have initiated a program that has completed three LVRS cases using the median sternotomy approach. There was one postdischarge death in this group. They reported that the total hospital cost for the first case was $39,000 and $29,000 for the third. They require a 6-week preoperative pulmonary rehabilitation program, and included their patient selection criteria (Tables 2 and 3).

The University of Louisville submitted a draft paper on LVRS, describing 100 patients treated with bilateral surgery via a median sternotomy. Of that number, 51 have been followed for at least 6 months postoperatively. Five patients who underwent a unilateral procedure were excluded from analysis. A preoperative pulmonary rehabilitation program was completed before surgery, as was bicycle ergometer exercise testing. The authors stated that dyspnea was assessed using the MRC Dyspnea Scale, and QOL was evaluated by administering the SF-36 Health Survey. After pulmonary rehabilitation, patients were required to demonstrate at least 30 percent improvement in 6-minute-walk distance, or no O₂ desaturation during the walk, to qualify for LVRS. Echocardiography was used to evaluate ventricular function and PAP. A history of coronary disease or ectopy during rehabilitation was regarded as an indication for a dobutamine echocardiogram, which if abnormal was followed by cardiac catheterization. Five patients underwent coronary angiography before LVRS, and five others were subject to combined coronary bypass (CABG) and LVRS. No data were provided that indicated improved outcomes resulted from these interventions.

Table 17. Data from the University of Louisville

Table 17. Data from the University of Louisville

Characteristics	No. patients
Prior thoracic surgery	4
O2 at rest	59
O2 during exercise	12
On oral prednisone	52
Male	65
Female	35
Age (yr)	65 (47-78)

There were eight operative deaths, and two additional deaths during followup (both of these deaths were in patients who had become ventilator-dependent postoperatively); one surviving patient was ventilator-dependent after surgery. The authors reported that complications occurred in 45 patients and noted that the average LOS was 14 days.

Table 18. University of Louisville: Lung-volume reduction (more...)

Table 18. University of Louisville: Lung-volume reduction surgery outcome measures at 3-12 months

			Postoperative
	Preoperative	3 months	6 months	12 months
FEV1 (percent predicted)	19 (100)	34 (76)a	38 (51)a	(6)
FVC (percent predicted)	53	54	69
TLC (L)	8.3	7.0	6.8
RV (percent predicted)	290	198	191
MVV L/min (percent predicted)	24	33	36
VO2 at rest (mL/min)	303 ± 99	372 ± 101	390 ± 105
Work output (watts)	30 ± 9	44 ± 10	47 ± 10
6-min walk (ft)	1,106b	1,308	1,425
PaO2	59 ± 5	68 ± 6	72 ± 8
PCO2	48 ± 4	43 ± 5	40 ± 4

a Calculated from authors' statement that 6 patients had been followed for 12 months, 45 for 6 months, and 25 for 3 months.

b Six-minute-walk distance after pulmonary rehabilitation.

Abbreviations: FEV₁ = forced expiratory volume in 1 second; FVC = forced vital capacity; MVV = maximum voluntary ventilation; PaO₂ = partial pressure of arterial oxygen; PCO₂ = partial pressure of carbon dioxide; RV = residual volume; TLC = total lung capacity; VO₂ = volume of oxygen consumption per unit of time.

Number in parentheses = number of cases.

The draft document remarked that the median 6-minute-walk distance before pulmonary rehabilitation was 855 feet; after rehabilitation, but before LVRS, this increased to 1,106 feet. The median values at 3 and 6 months after LVRS were reported as 1,308 and 1,425 feet, respectively. Although no data were provided as to whether any postoperative treatment, such as continued pulmonary rehabilitation was provided, the implication was that the benefit was solely due to LVRS. At 6-month followup, seven of 51 patients required O₂ at rest, and nine required O₂ on exertion; as noted above, preoperatively, 59 patients required O₂ at rest, and 12 required it on exertion. The overall proportion of O₂-dependent patients appeared to have been much reduced; if the intervention had no effect on such requirements, approximately 30 patients would have been expected to need O₂ postoperatively, although the 10 percent mortality rate may have included a disproportionate number of the O₂-dependent cases. However, information was not provided as to which of the 51 evaluated at 6 months had been O₂-dependent preoperatively, so the actual conversion rate could not be ascertained.

The authors also remarked that "subjective improvement" as measured by the SF-36 Health Survey instrument was "significantly better at 3 and 6 months postoperative after LVRS compared with 1 year before LVRS." However, specific SF-36 scale scores were not provided, nor was it ever detailed in the body of the report that SF-36 surveys were administered 1 year before patients were subjected to surgery. No cost data were provided.

The University of South Florida reported LVRS in 32 patients; 14 were performed as bilateral and 18 as unilateral thoracotomies. The mortality rates were 20 percent and 6 percent for the bilateral and unilateral procedures, respectively. Mean preoperative FEV₁ was 22 percent of predicted. No followup pulmonary function or functional capability data were provided, but the respondents stated that 92 percent of the patients "experienced improved quality of life" although the QOL measurement techniques were not specified. They noted that they "believe strongly that preoperative rehabilitation is essential." The unilateral thoracotomy approach is now preferred because of the better early preoperative results and the option of a second operation at a later date. The respondents provided no cost data.

Table 19. Washington University: Lung-volume reduction (more...)

Table 19. Washington University: Lung-volume reduction surgery data

		Postoperative
Characteristics	Preoperative	6 months	12 months
FEV1 (percent predicted)	24 (120)	38 (65)	37 (35)
FVC (percent predicted)	70 (120)	88 (65)	83 (35)
TLC (percent predicted)	148 (120)	124 (65)	124 (35)
RV (percent predicted)	291 (120)	202 (65)	205 (35)
Percent patients on continuous O2	55 (120)	8 (65)	9 (35)
Percent patients on O2 with exercise	90 (120)	34 (65)	29 (35)
MRC Dyspnea Scale scores	3.0 (120)	1.0 (65)	1.3 (35)

Abbreviations: FEV₁ = forced expiratory volume in 1 second; FVC = forced vital capacity; MRC = Medical Research Council; RV = residual volume; TLC = total lung capacity.

Number in parentheses = number of cases.

Sixty-day mortality was reported as 2.5 percent, with late mortality (to 9 months) as an additional 2.5 percent. Detailed morbidity data were not provided, but eight patients required reintubation; five of those later required tracheostomy. Patient inclusion and exclusion criteria were offered and are contained in Tables 2 and 3.

The report stated that, in view of the association between life expectancy and FEV₁ measurements in COPD patients, "it may be anticipated" that the improvement in FEV₁ "should be translated into an improved survival." Further, an "informal registry" for bilateral LVRS has been established at Washington University that included 130 procedures from 30 other centers. The overall hospital mortality has averaged 6 percent; no other followup data are as yet available.

Cooper's report briefly addressed alternative surgical techniques, and concluded that most centers reporting experience with laser LVRS "report insufficient benefit to justify the significant risks involved." The report concluded that: "laser ablation therapy" in LVRS has not been documented to show consistent significant benefit, and the technique "should be considered experimental at best;" a thoracoscopic approach is "feasible" but additional experience is necessary to determine the risk/ benefit ratio compared with that for median sternotomy; and a unilateral approach may be appropriate for selected patients. Finally, Cooper noted that "what remains to be determined" included: the role of the procedure, appropriate selection of patients, "longevity" of the result, impact upon subsequent pulmonary-related medical costs, and effects upon life expectancy.

Included in the report was a separate document from the Nursing Division of Barnes Hospital, entitled "Hospital Costs of a Lung Volume Reduction: A Comparison of 43 Lung Volume Reduction Patients and 264 Other Patients in DRG Category 75." This document noted that LVRS are about 45-60 percent more expensive than "the average thoracic surgery cases," and went on to elaborate increases in categories such as operating room time, medications, supplies, laboratory testing, etc. However, no dollar figures were provided, and this account concluded with the statement that "There is no set price for lung volume reduction because of the variability between patients in our patient population."

The AATS/Cooper report also contained information from several centers in addition to Washington University; all of these data had been obtained by CHCT directly from the centers concerned.

Discussion

Unpublished Data

Given the limited published data available on the risks and benefits of LVRS, we believed that the acquisition of data directly from institutions and individuals performing the surgery was important to this assessment process. The Health Care Financing Administration shared our concern that the published data might not prove adequate to permit evidence-based conclusions on important issues such as the overall clinical effectiveness of LVRS, the mortality and morbidity of surgery, the appropriate patient or institutional selection criteria, the effects of the surgery on COPD mortality rates, and changes in QOL and functional capabilities which might be occasioned by LVRS.

It was disappointing that only a small number of organizations were willing to provide data; several large institutions that are known to have performed many LVRS operations chose not to respond and directly participate in the assessment process. Discussions with correspondents indicated that in some cases there existed institutional concerns that citation in the assessment could be discomfiting if their practices or outcomes appeared to differ from those of other groups; nevertheless, the same situation was obtained regarding publication in the peer-reviewed literature. It is not the intent of this assessment to disparage any organizations that have submitted information; it is, however, the responsibility of the assessment process to critically review objective data, irrespective of the source. The responding institutions and individuals deserve great credit for choosing to participate in this evaluative process, particularly in the circumstance wherein they possess the bulk of the information, which would otherwise be unavailable to the public.

Notwithstanding, the quality of the evidence was generally weak. Some correspondence provided only statements of opinion, occasionally accompanied by very general descriptions of LVRS at the institution. In many other cases, it was not possible to discern critical details such as the specific numbers of surgical procedures, the precise pre- and postoperative physiologic and functional measurements used, and the duration of followup and specific numbers of cases followed for definite time intervals. Most followup was of short duration, a number of measurements reported as statistically significant were clinically trivial (e.g., 2 mm Hg improvement in PaO₂), there were few reliable data concerning changes in QOL due to LVRS, and errors in manuscripts were not uncommon and made interpretation of some material difficult (vide supra).

Although measures of central tendency (means or medians) were understandably emphasized, relevant deviations were rarely reported; there were, however, some exceptions. Data provided by the University of Pennsylvania indicated that the VATS surgical technique resulted in an improvement of 10 percent in mean FEV₁ at 3 months, but further detailed that 18 percent of patients exhibited reductions in FEV₁. The fact that nearly one in five cases may be worse after surgery is at least as important as the demonstration that mean predicted FEV₁ increased from 24-34 percent. The W.A. Foote Hospital reported data collected by their surgeon at a previous appointment at the Medical Center of Central Massachusetts (vide supra); they noted that the mean 3-month postoperative improvement in the 6-minute-walk test was 130 feet, but also included data which showed that less than half of the patients (about 42 percent) tested after surgery actually improved their distance walked. Temple University reported that, of 10 patients evaluated with the SIP, eight exhibited improved scores after LVRS, but two worsened. The ranges noted by St. Louis University for pre- and postoperative 6-minute-walk distances (0-1,601, and 0-1,700 feet, respectively) indicated that one or more cases had zero exercise tolerance after surgery. It remains unclear whether these and similar anomalous results simply represent the extremes of a typical distribution of patient responses to a particular intervention, or are due to factors such as specific patient characteristics, distinctions between LVRS techniques, technical aspects of the surgery, or the kind and degree of postoperative medical treatments. Such information is obviously critical in evaluating the risks and the benefits of LVRS, yet few respondents provided those details.

Table 20. Surgical procedures utilized for lung-volume (more...)

Table 20. Surgical procedures utilized for lung-volume reduction surgery

Procedure	Approximate total no. patients reported
Open procedures
Median sternotomy	410
Thoracotomy, unspecified	28
Thoracotomy, bilateral	44
Thoracoscopya
Unilateral thoracoscopy	512
Bilateral thoracoscopy	143
Thoracoscopy, otherwise unspecified	55
Thoracoscopy, bilateral sequential	181
Thoracoscopy, laser	1,452

a Note that categories are not necessarily mutually exclusive; i.e., some unilateral, bilateral, or unspecified thoracoscopies may have used lasers and not specifically so stated; for some laser thoracoscopy procedures, details were not provided as to whether they were uni- or bilateral.

A variety of surgical approaches were utilized. Not all respondents described their preferred technique(s) or provided information as to whether they utilized unilateral or bilateral surgery, performed resections with laser or stapling methods, etc. Thus, Table 20 represents a plausible estimation of the types and numbers of procedures reported, but cannot be considered to approach a precise enumeration.

Reference to Tables 2 and 3 will reveal that there appears to be significant heterogeneity in patient selection criteria. The only unanimous exclusion criteria was active or recent smoking. Fourteen of 18 respondents excluded metastatic cancer (although this was not specifically stated but was inferred in 7); 14 also excluded patients with increased PAP. However, within these criteria there were differences of opinion (e.g., maximum permissible PAP varied from 35 to 50 mm). Thirteen reported that they excluded patients with serious comorbid disease, and 11 respondents excluded those with intrathoracic cancer. Half excluded cases with bronchospasm or bronchitis. A minority (one-third or less) of the respondents specifically excluded ventilator-dependent patients, those with morbid obesity, patients with psychiatric or social dysfunction, wheelchair or bed-bound candidates, and patients on steroids of varying dosages. Exclusion criteria included both bullous and nonbullous emphysema. As regards inclusion criteria, only two were specified by more than half of the respondents: specific levels of FEV₁, and required preoperative pulmonary rehabilitation. For FEV₁, requirements ranged from < 30 percent predicted to < 50 percent predicted. Despite the arguments that some have made regarding the specific mechanical benefits of LVRS in terms of high residual volumes and poor air movement (vide supra), requirements of increased RV, hyperinflation, and poor diaphragmatic movement were specified by less than a third of the centers. Oxygen dependence was specified as a criterion by only one institution.

Thus, both the operative procedures of choice and specific patient inclusion and exclusion criteria varied. Different patients have been subjected to different procedures, and there are few data suggesting which patient or surgical characteristics have been associated with superior outcomes. It appeared that such variation was based upon what was believed, at a given institution, to be the most prudent clinical judgement.

Although many correspondents reported improvements in various PFTs, the claimed improvements that were often considered to be of greatest importance related to patients' perceptions of functional status and QOL. It is therefore troubling that relatively few data were available which specifically addressed that issue.

Table 21. Summary data: Mortality and morbidity

Table 21. Summary data: Mortality and morbidity

Characteristic	Range	Mean	Median	Sample size
Perioperative mortalitya	0 (9) to 20 (14)	6.6	7.7	1,019
Morbidity	43-68	55	53	407
Length of stay	10-19 days	15	15	515

a Mortality: Because it was not always clear how "perioperative" mortality was defined, deaths were not included unless it was obvious that they were directly attributable to the surgery itself.

Number in parentheses = number of cases.

Table 22. Summary data: Forced expiratory volume in (more...)

Table 22. Summary data: Forced expiratory volume in 1 second

Characteristic	Range	Mean	Median	Sample size
Mean preoperative FEV1 (percent predicted)	19-31	26	27	1,062
Postoperative percent improvement in FEV1 for bilateral surgery	13-79a	43	42	598 (at 3 months)

a For example, a 20 percent improvement in a preoperative FEV₁ of 30 percent predicted would result in a change from 30-36 percent.

Table 23. Preoperative oxygen use

Table 23. Preoperative oxygen use

Institution	No. patients using O2/no. reported operated	Percentage of O2 use
Washington University	90/100	90
University of Pittsburgha	29/35	83
Wakabayashi Institute	505/614	82
St. Louis University	37/50	74
Medical Center Central Massachusetts	41/58	71
University of Louisville	71/100	71
Temple University	16/23	70
Columbia University	59/85	69
Chapman Medical Centera	86/166	52
Totals and mean	934/1,231	76

a Note that not all institutions specified the prevalence of preoperative O₂ use for all of the patients operated at their center; e.g., Pittsburgh reported 100 thoracoscopically treated patients, but preoperative O₂ use was reported only for a subset of 35. Chapman Medical Center stated LVRS had been performed on a "300-patient population base" but detailed preoperative O₂ use only within a subset of 166.

Given the heterogeneity of patient selection and institutional practices and the limited number of institutions that provided detailed information, there are obvious hazards inherent in combining the data from the various sources, and one cannot assume that such a process would necessarily provide a reliable illustration of the current state of LVRS. Nevertheless, the exercise is both interesting and revealing. We attempted to summarize, as well as possible, characteristics of LVRS provided by respondents to our solicitations. However, not all provided categories of information or sufficient detail for inclusion in the analysis, and consequently the summary still represents only a sample of current clinical practices. Specific categories of interest are detailed in Tables 21 and 22.

Not all respondents reported information as to the proportion of patients preoperatively dependent on O₂, and when reported, such data were not always available for all the cases operated at the institution; moreover, it was not consistently clarified whether such dependence was continuous or intermittent, e.g., with exercise. Given such limitations, by combining the material nine respondents provided and by postulating that any level of preoperative use defined O₂ dependence, it was found that 934 of 1,231 patients who underwent operations, or approximately 76 percent, were so dependent.

Thus, the best estimate from the available information was that about 24 percent of patients subjected to LVRS were not using any O₂ before surgery, whereas at some centers this proportion reached almost 50 percent. These disparities are difficult to reconcile, and it is unclear if the observed variation results from differences in patient selection criteria, operation on patients with divergent severity of COPD, or other factors.

Operative effects upon O₂ dependency were more difficult to calculate. A number of institutions provided only the overall percentage of patients requiring O₂ preoperatively and the percentage dependent postoperatively; this did not permit estimation of the proportion actually rendered O₂-independent, because the number of patients reported at followup was most often notably smaller than the preoperative sample size. Moreover, the duration of followup varied. For those instances wherein patients' status at 3-month followup were clearly documented as having gone from O₂ dependence to independence (only approximately 145 cases), the reported percentages so benefiting ranged from 36-68 percent.

Table 24. Summary data: 6-minute walk

Table 24. Summary data: 6-minute walk

	Preoperative weighted combined mean (range)	3-month postoperative weighted combined mean (range)
Distance (ft)	906 (605-1,106)	1,118 (904-1,308)
Sample size	288	251

Pre- and postoperative measures of QOL and/or functional capacity were used. A number of different measures of dyspnea were employed; the validity and reliability of some of these are as yet uncertain (vide supra), and there are no data that permit cross-estimations from one to another measurement technique. A wide variety of instruments have been used to evaluate QOL; which of those are most reliable and valid for patients with COPD has yet to be established, although fairly large samples have been evaluated using some instruments, such as the SF-36 Health Survey. Although the submitted material implied that the surgery itself was responsible for improvements reported, in many instances patients appeared to have been provided postoperative pulmonary rehabilitation and other treatments, which could have significantly contributed to the benefits claimed. The most widely used functional measure represented in the submitted material was the 6-minute-walk distance; however, few institutions provided detailed postoperative data. The mean values, ranges, and sample sizes calculated from data supplied by six institutions are summarized in Table 24.

Detailed 3-month followup data were provided for only 251 cases. Moreover, measurements were quite variable, perhaps reflecting differences in patient selection criteria. For example, the mean preoperative distance of 1,106 feet reported by one institution was greater than the mean postoperative distances reported by three of the remaining five institutions.

Table 25. Summary data: Duration of followup and number (more...)

Table 25. Summary data: Duration of followup and number of cases at specified times^a

		Postoperative
Preoperative	3 months	6 months	12 months
3,090	568	281	85

a Note: Data were included only if specific numbers were reported at particular time intervals after surgery (e.g., information reported as "139 patients with followup between 1 and 4 months" were not considered for inclusion). The type or comprehensiveness of postoperative evaluation was not considered.

For the most part, followup after LVRS has been quite limited. Although it was not always possible to be certain precisely how many cases were followed for specific intervals after surgery, Table 25 provides the best estimate from the data submitted.

Thus, the cases with clearly documented followup evaluation at 3, 6, and 12 months after LVRS represent approximately 18, 9, and 3 percent, respectively, of the operated population reported. This presents a major obstacle to an objective assessment of the risks and benefits of LVRS. The proportion of cases followed even over fairly short periods is quite small compared with the number operated, making extrapolation of the reported results quite tenuous. Incomplete postoperative evaluation will distort or bias the results in the circumstance that patients with poorer outcomes are more likely to fail to return for followup and are not included in the analyses.

Summary and Conclusions

In considering all of the published and unpublished information obtained for this assessment, it cannot reasonably be concluded at this time that the objective data permit a logical and a scientifically defensible conclusion regarding the risks and the benefits of LVRS as currently provided. Patient selection criteria are heterogeneous and in flux both within and between institutions; serious disagreements exist as to the most appropriate surgical techniques for various categories of patients; followup data are meager, and only a small fraction of operated cases have been followed for even as long as 6 months; benefits have been inconsistently and often imprecisely documented; and it is as yet unclear whether changes in PFTs, patient functional capacity, O₂ dependency, QOL, or any combination of such measures, are most significant. Improvements in survival have not yet been demonstrated, nor have measures such as FEV₁ been shown to be reliable surrogates for survival post LVRS; and the available evidence is inadequate to explain the wide variations that have been reported for some measures of outcome.

Notwithstanding, the data suggest that an as-yet undefinable proportion of patients with severe COPD may have realized some benefit from the procedure. If the surgery could be accomplished without undue morbidity or mortality, a prospective trial of LVRS under uniform protocol requirements with comprehensive long-term postoperative followup data is both ethically supportable and scientifically essential.

Appendix A. List of Abbreviations

• AATS = American Association for Thoracic Surgery

• ABG = Arterial blood gas

• BDI = Baseline dypnea index

• CABG = Coronary artery bypass graft

• COPD = Chronic obstructive pulmonary disease

• CRQ = Guyatt's Chronic Respiratory Disease Questionnaire

• CT = Computerized tomography

• CXR = Chest x-ray

• DLCO = Carbon monoxide diffusing capacity of lungs

• FEV1 = Forced expiratory volume in 1 second

• FiO2 = Fraction of oxygen in inspired air

• FVC = Forced vital capacity

• IMC = Irvine Medical Center

• LOS = Length of stay

• LVRS = Lung-volume reduction surgery

• MOS = Medical Outcomes Survey

• MRC = Medical Research Council

• MVV = Maximum voluntary ventilation

• NYHA = New York Heart Association

• PaO2 = Partial pressure of arterial oxygen

• PAP = Pulmonary artery pressure

• PCO2 = Partial pressure of carbon dioxide

• PFT = Pulmonary function test

• PiO2 = Partial pressure of inspiratory oxygen

• PO2 = Partial pressure of oxygen

• QOL = Quality of life

• RV = Residual volume

• SF-36 = Short-form 36 questionnaire constructed from the MOS

• SIP = Sickness Impact Profile

• TDI = Transition dyspnea index

• TLC = Total lung capacity

• VATS = Video-assisted thoracoscopic surgery

• VO2 = Volume of oxygen consumption per unit of time

• YAG = Yttrium-argon-garnet laser

Appendix B. List of Contributors

• American Association for Thoracic Surgery

• American Thoracic Society

• Baylor University

• California Thoracic Society

• Cardiac Surgical Associates at United Hospital, St. Paul, MN

• Cardiothoracic Associates of Sandusky, OH

• Cardiothoracic and Vascular Surgeons Ltd., St. Joseph's Hospital, Phoenix, AZ

• Chapman Medical Center of Orange, CA

• Columbia University

• Hartford Thoracic and Cardiovascular Group

• Harvard Medical School, Deaconess Hospital

• Mayo Clinic

• Medical Center of Central Massachusetts

• Midwest Pulmonary Consultant Group of Kansas City, MO

• National Jewish Center for Immunology and Respiratory Medicine

• Rose Medical Center of Denver, CO

• St. Louis University Health Sciences Center

• Temple University

• University of California, Los Angeles Medical Center

• University of California, San Diego Medical Center

• University of Louisville

• University of Pennsylvania

• University of Pittsburgh

• University of South Florida

• University of Washington School of Medicine

• W.A. Foote Hospital of Jackson, MI

• Wakabayashi Institute

• Yale University