|
|
||||||||
Guest Access | Sign In via User Name/Password |
|||||||||

* From the Departments of Respiratory Medicine (Dr. Snell), Physiotherapy (Ms. Munro), and Cardiothoracic Surgery (Dr. Smith), The Alfred; and Department of Epidemiology and Preventative Medicine (Mr. Bailey), Monash University, Victoria, Australia.
A list of contributing hospital is given in the Appendix.
Correspondence to: Prue E. Munro, BPhysio, Department of Respiratory Medicine, The Alfred, Commercial Rd, Prahran, VIC 3181; Australia; e-mail P.Munro{at}alfred.org.au
| Abstract |
|---|
|
|
|---|
Aims: To review patient selection, surgical activity, and patient outcomes related to LVRS in Australia and New Zealand.
Methods: Prospective data were voluntarily submitted by hospitals performing LVRS in Australia and New Zealand. Preoperative, surgical, perioperative, and follow-up variables were analyzed.
Results: Data were collected from 15 hospitals regarding 529 patients. Mean age (± SD) at surgery was 63 ± 7 years. Preoperatively, FEV1 was 29 ± 9% predicted, total lung capacity (TLC) was 138 ± 20% predicted, residual volume (RV) was 250 ± 64% predicted, and 6-min walk (6MW) distance was 327 ± 111 m. There has been a reduction in the overall number of cases and hospitals performing LVRS since 1999. Improvements in lung function following LVRS (ie, FEV1 increase of 38%, RV decrease of 27%, TLC decrease of 17%) and exercise capacity (ie, 6MW distance increase of 24%) appear to be maintained for approximately 3 years.
Conclusions: LVRS continues to be performed in Australia and New Zealand, predominantly in large tertiary teaching hospitals with similar outcomes to those described in the literature. It remains difficult to capture long-term lung function and survival outcomes in this population. Ongoing audit and RCTs are both required to resolve the confusion that still shrouds this procedure.
Key Words: evidence based medicine lung volume reduction outcomes assessment pulmonary emphysema registry
| Introduction |
|---|
|
|
|---|
The short-term outcomes of LVRS have been repeatedly demonstrated. LVRS has the potential to improve lung function, exercise performance, and quality of life in highly selected individuals with severe emphysema.7 Published operative mortality rates vary from 0 to 19%,8 and postoperative morbidity is relatively high. Optimal patient selection is still unknown.
LVRS has remained controversial as major questions regarding its efficacy and long-term outcomes still remain unanswered. The degree of benefit in terms of symptomatic relief, improvement in lung function, exercise capacity, and quality of life varies between individual patients. There is also a lack of long-term results describing the duration of any improvement obtained. Aside from these clinical questions, economic and ethical issues have also been raised regarding whether health-care systems can afford to offer what must be regarded as an expensive palliative treatment for a very common, often "self-induced" disease.9 10 11 12 13
A few small randomized controlled trials (RCTs) have compared medical management with LVRS,14 15 and the results of these studies have been consistent with the earlier case series. Large multicenter RCTs are underway in the United States,16 Canada,17 and United Kingdom,18 and will compare LVRS with best medical management, including pulmonary rehabilitation. It is hoped that they will be able to answer many of these complex questions, further defining the role of LVRS in the management of emphysema.
In the absence of current clear clinical guidelines, the Australian and New Zealand LVRS Database, operating under the auspices of the Royal Australasian College of Surgeons Australian Safety and Efficacy Register for New Interventional Procedures-Surgical, was created to audit local clinical practice and to monitor surgical activity and patient outcomes. Priority was given to the dissemination of data and facilitation of discussions among clinicians in an attempt to provide regional leadership to what had internationally developed into a sometimes confusing and at times a controversial management option. Comprehensive patient selection criteria have been developed using Australian and New Zealand data,19 and opportunities for the improvement in surgical and perioperative management have been identified.20
This report presents the results of these audit activities as a summary of the ongoing assessment of safety and efficacy of LVRS in Australia and New Zealand over a 6-year period. It also highlights some of the practical difficulties with the administration and utilization of voluntary registries as part of an evidence-based approach.
| Materials and Methods |
|---|
|
|
|---|
Table 1 shows the variables collected. Values were recorded preoperatively, at 3 months, 6 months, 9 months, and 12 months postoperatively, and then yearly following LVRS. The best value for each time period was used if multiple test results were available.
|
Data Analysis
The data set from September 1995 to August 2001 was analyzed. Patients who had LVRS following lung transplantation were excluded. Data analyses were conducted using SAS version 8.0 (SAS Institute; Cary, NC) and Excel 97 (Microsoft). Descriptive statistics are reported as mean ± SD unless otherwise specified. A repeated-measures generalized linear model was used to describe the change in lung function (n = 336) and 6-min walk (6MW) distance (n = 292) over time following LVRS. The multivariate models were adjusted for age, gender, and hospital. Where data are not complete for the variable of interest, actual numbers included in the analysis are specified.
| Results |
|---|
|
|
|---|
The number of cases performed per year dramatically increased from 1 case in 1995 to 128 cases in 1998. In 1999, the level of activity stabilized and then declined in 2000. Data for 2001 are incomplete (only available to August), but a trend toward a further decline in activity is clear (Fig 1 ).
|
|
|
Preoperative Pulmonary Rehabilitation
Pulmonary rehabilitation was undertaken preoperatively in 424 of 439 cases (97%).
Surgery
Four hundred seventy-six of 509 patients (93.5%), in whom the type of surgery was reported had bilateral LVRS. There was a significant increase in the proportion of unilateral LVRS performed in the period from 1998 to 2000 compared with from 1995 to 1997 (p = 0.045).
A midline sternotomy was used in 49%, thoracotomy in 43%, and video-assisted thoracoscopy (VAT) in 8%. Bovine pericardial strips (Biovascular; Minneapolis, MN) were used for reinforcement in 57% of cases, and Gore-tex (Gore and Associates; Phoenix, AZ) was used in 35% of cases.
Although there were no postoperative complications in 43% of patients, major (life threatening) and minor postoperative complications were reported in 29% and 28% of cases, respectively. Of those with major postoperative complications, 23% had an air leak > 7 days, 22% had sepsis, 19% required reintubation for respiratory failure, 12% underwent tracheostomy, 15% had cardiac events, and 9% had other complications. In those patients with a survival time of at least 90 days, intercostal catheters were in situ for 9.8 ± 9.3 days and length of hospital stay was 17 ± 14 days.
Lung Function
FEV1 percentage of predicted improved from 29 to 40% predicted at 3 months after surgery, then declined linearly at a rate of approximately 3% (0.10 L) per year (p < 0.0001) returning to baseline at 3 to 4 years. Vital capacity (VC) percentage of predicted improved in the first postoperative year, then also fell steadily by approximately 1.5% (0.14 L) per year (p < 0.009). RV percentage of predicted decreased from 250 to 183% predicted in the first 6 months, then increased at a rate of approximately 6%/yr thereafter (p = 0.0007). TLC percentage of predicted dropped from 138 to 115% in the first 3 months, then increased by approximately 1.5%/yr (p = 0.006).
6MW
There was a continued improvement in the 6MW distance for approximately the first year after LVRS. The distance walked appeared to remain above the baseline preoperative value for approximately 3 years. Overall, 6MW distance declined at a rate of 16 m per year (p = 0.0002) over the 5-year time period from the preoperative value.
Survival
Survival status was reported in 454 cases (85.8%). Three hundred ninety-six patients were alive, and 58 patients were dead. Thirty-day survival was 96.2%, and 90-day survival was 92.2%. Survival after surgery was 90.5%, 90.2%, 88.7%, and 79.4% at 1 year, 2 years, 3 years, and 4 years, respectively. Cause of death was specified in 48 cases. Respiratory failure accounted for 45% of deaths, sepsis accounted for 21%, and cardiac failure accounted for 17%.
| Discussion |
|---|
|
|
|---|
In comparison with European countries, to the end of 1999, Australia had a greater number of reported cases of LVRS (Fig 3 ). Australia also had a greater number of institutions performing LVRS; at 4.1 LVRS cases per million population per year, the level of servicing for the population was only higher in Switzerland and Austria24 (Fig 4 ). Similar US data are not available, but with the withdrawal of US Medicare funding for LVRS in 1993,25 it is probable that US levels of activity are considerably lower.
|
|
Notwithstanding, the number of LVRS cases performed appears to have significantly waned in the last 2 years. The reasons for the downturn in LVRS procedures performed in Australia and New Zealand are multifactorial. There has possibly been some diminution in the pool of ideal candidates, and there is always competition for diminishing resources; however, concern about the seemingly inevitable background morbidity and mortality of existing techniques, in the setting of a likely median duration of improvement of only 3 years, are the more significant issues.
Australian and New Zealand case selection, clinical results and survival outcomes closely follow those published results from large US centers.3 4 7 14 It is also notable that no significant "center effect" was demonstrable among the contributing groups, suggesting Australian and New Zealand centers have heeded the messages behind the US Medicare LVRS funding crisis.25
In the current series, the vast majority of patients (93.5%) underwent bilateral LVRS, usually via an open technique. The choice of surgical approach (open vs VAT) related to the surgeons experience and preference. Outcomes appear similar whatever the strategy, but further data are clearly needed.
One of the main limitations to our analyses is the voluntary nature of data reporting. Obtaining long-term follow-up data was challenging as individual centers had differing follow-up practices. There was considerable variability in the length of time patients were followed up for and the frequency and extent of physiologic and functional outcome testing performed. Long-term outcome data from this (and all other) LVRS studies is limited by the relatively small numbers of patients included in the analysis, the high proportion of patients unavailable for follow-up, and the lack of an adequate control group for comparison. In particular, this has the tendency for LVRS studies to underestimate long-term mortality.27
Selection bias appears to have been less problematic for the Australian and New Zealand LVRS Database than is often experienced by voluntary registries. A survey undertaken in 1999 estimated that the Australian and New Zealand LVRS Database captured approximately 76% of all identifiable cases of LVRS in Australia and New Zealand.28 We suspect that this has now improved to > 85% as the procedure has become more focused in the tertiary hospital setting. This represents a uniquely high level of cooperation with data transfer, and could be attributed to the relatively small medical community in Australia and New Zealand and the strong professional network links of key personnel with relevant professional bodies including the Royal Australasian College of Surgeons, Thoracic Society of Australia and New Zealand, and the Australian Physiotherapy Association.
The Australian and New Zealand LVRS Database has played a vital role in the evaluation and clinical evolution of LVRS in Australia and New Zealand. Although registries are inherently subject to greater bias than RCTs, they are still capable of generating reliable data and can reduce bias attributed to surgical skill.29 The existence of this registry has enabled individual hospitals to benchmark their own practice and outcomes (particularly mortality and morbidity) with other participating centers and with the published results. There is also the potential to compare registry reported mortalities with national death indexes to improve accuracy.30
Fessler and Wise8 provided a detailed review of the duration of benefit following LVRS and suggested that it was probably incorrect to assume there to be a linear decline in FEV1; however, in our analysis, we found that a linear model best described the change in lung function over time. Our results showed that FEV1 percentage of predicted improved from 29 to 40% in the first 3 months postoperatively, then declined at a rate of 100 mL/yr, returning to baseline at 3 to 4 years. Brenner and coworkers31 reported an average loss of FEV1 of 163 mL/yr in 180 patients who, like our group, had undergone a variety of LVRS procedures. The annual rate of decline of FEV1 in patients with COPD has been reported to be up to 80 mL/yr.32 Despite these data, it remains difficult to make direct comparisons between groups of LVRS patients with emphysema, as there may be important differences in smoking history, age, medical management, severity of disease, and other factors. RCTs will ultimately provide the best comparisons.
There are prospects for improvements in morbidity, mortality, and clinical outcomes, as the results of various large RCTs, such as the National Emphysema Treatment Trial, become available, and via the inevitable evolution to newer, less invasive LVRS technologies.33 34
Although RCTs are the scientific "gold standard" for evaluating new therapies, it can be argued that observational studies may be preferable in the setting of evolving patient selection, operative, and perioperative management practices as is the case with LVRS.29
Recently, the National Emphysema Treatment Trial research group published a report35
regarding patients who were at high risk of death after LVRS. This patient population was characterized by an FEV1
20% predicted and diffusing capacity of the lung for carbon monoxide (DLCO)
20% predicted or a homogenous pattern of disease. Based on their findings, they have now stopped enrolling this group of patients in their clinical trial. Of the 1,033 patients who had been randomized from 17 centers, 140 patients (13.6%) were identified as high risk. In comparison, of our 529 patients, only 8 patients (1.5%) met these "high-risk" criteria. We believe that this large difference in inclusion criteria for LVRS is likely to be due to the fact that the Australian and New Zealand centers had already clinically defined this group correctly and deemed them unsuitable candidates for surgery on the basis of a high likelihood of increased morbidity and mortality.
So who is responsible for providing evidence regarding the efficacy of a new procedure? Despite an apparent keenness to develop an evidence base for clinical medicine from professional bodies, hospital administrators, and individual clinicians, practical issues essential for the ongoing functioning of a database such as ours are often overlooked. Funding support and potential clashes with privacy laws and institutional ethics committees across different international and state borders are all very real problems; however, from a practical standpoint, maintaining a database for regular audit is an inexpensive way to monitor activity and outcomes when compared to the financial burden and restriction of an RCT. In an Australian context, groups like Australian Safety and Efficacy Register for New Interventional Procedures-Surgical are only just starting to provide assistance for this type of project.36
| Conclusion |
|---|
|
|
|---|
| Appendix |
|---|
|
|
|---|
New Zealand
Greenlane Hospital, Auckland.
| Footnotes |
|---|
Received for publication July 22, 2002. Accepted for publication February 26, 2003.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
S. Imfeld, K. E. Bloch, W. Weder, and E. W. Russi The BODE Index After Lung Volume Reduction Surgery Correlates With Survival. Chest, April 1, 2006; 129(4): 873 - 878. [Abstract] [Full Text] [PDF] |
||||
![]() |
A R Glanville Ethical and equity issues in lung transplantation and lung volume reduction surgery Chronic Respiratory Disease, January 1, 2006; 3(1): 53 - 58. [Abstract] [PDF] |
||||
![]() |
R. B. Gorman, D. K. McKenzie, J. E. Butler, J. F. Tolman, and S. C. Gandevia Diaphragm Length and Neural Drive after Lung Volume Reduction Surgery Am. J. Respir. Crit. Care Med., November 15, 2005; 172(10): 1259 - 1266. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. D. Yusen and B. Littenberg Integrating Survival and Quality of Life Data in Clinical Trials of Lung Disease: The Case of Lung Volume Reduction Surgery Chest, April 1, 2005; 127(4): 1094 - 1096. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |