|
|
||||||||
CORRESPONDENCE: M. Decramer, Respiratory Division, University Hospital, Katholieke Universiteit Leuven, Herestraat 49, 3000, Leuven, Belgium. Fax: 32 16346803. E-mail: Marc.Decramer@uz.kuleuven.ac.be
Keywords: chronic obstructive pulmonary disease, exercise capacity, health status, lung volume reduction surgery, rehabilitation, selection of candidates
Received: January 29, 2003
Accepted May 5, 2003
The present studies were supported by the Research Foundation of the Catholic University of Leuven (Leuven, Belgium) 98/27 and the Fund for Scientific Research-Flanders grant numbers G.0175.99 and G.0237.01 and Lifeline grant number 7.0007.00.
| ABSTRACT |
|---|
|
|
|---|
It remains, however, difficult to predict accurately which patients will improve after rehabilitation and which will not. Factors that may contribute to this prediction are: baseline peak exercise ventilation/maximal voluntary ventilation, maximal inspiratory pressure, peripheral muscle force, and 6-min walking distance. Several studies have clearly shown that training effects are as pronounced in patients with severe as in those with moderate airflow obstruction. This is the most significant insight in this area of the 1990s.
Lung volume reduction surgery may also be of benefit in patients with end-stage COPD. It is clear that lung function, exercise capacity and health status improve after this procedure, although the results of only six randomised studies are currently available.
It remains difficult to accurately predict which patients will benefit from the procedure. From a model analysis, the most important action mechanism appears to be resizing of the lungs. Only in patients with an increased residual volume/total lung capacity ratio are beneficial effects expected. The model analysis, however, did not fit the experimental data very well in a recent publication. Emphysema heterogeneity is also likely to be related to the response. The question remains as to whether or not lung volume reduction surgery accelerates the decline in forced expiratory volume in one second in the long run. The technique is undergoing considerable progress as numerous new surgical and endoscopic procedures are currently being developed. The results of these procedures have not yet been systematically evaluated.
Chronic obstructive pulmonary disease (COPD) is without question the most common cause of respiratory failure. Ideally, it should be detected and treated in the early stages before respiratory failure occurs 1. The only treatment modality that may reduce the progression of COPD is smoking cessation 2. Nevertheless, many other treatments, such as bronchodilators 3–5, inhaled corticosteroids 6–10, lung transplantation and lung volume reduction surgery (LVRS) 11, 12, may improve lung function. The effect of all these treatments combined, as usually performed in practice, has never systematically been studied. Most patients do not follow appropriate treatments for their disease because they underestimate it and consider it self-inflicted 13, 14.
Several treatment options are currently available for end-stage COPD. These include: pharmacological treatment, rehabilitation, oxygen therapy, LVRS, and lung transplantation. These treatments are increasingly successful, which has resulted in a clear improvement in survival in patients receiving long-term oxygen treatment since the 1980s 15. Most of these treatment modalities are discussed in the present supplement, with lung transplantation being discussed by Lau and Patterson 16. The present contribution reviews two treatment modalities that are complementary in the treatment of COPD patients: rehabilitation and LVRS. The first of these primarily improves exercise capacity and health status but does not alleviate airflow obstruction. The second has at least the potential of improving pulmonary function substantially in some patients.
For each of the treatment modalities, the potential benefits, selection of candidates, practical organisation and potential future developments are discussed in order.
| Respiratory rehabilitation |
|---|
|
|
|---|
|
At present, there has been no clear demonstration of improved survival after a pulmonary rehabilitation programme. There are, however, three indications that survival might improve after pulmonary rehabilitation. First, Ries et al. 18 found greater survival after a rehabilitation programme during a 6-yr follow-up, but the difference did not reach significance. The power, however, of the study to detect a difference in survival was too low. Secondly, Griffiths et al. 23 found a similar tendency, but the number of patients was too low for it to reach significance. Thirdly, Troosters et al. 28 found a tendency for greater survival in patients who were admitted to a training programme after hospitalisation for an acute exacerbation of COPD. These three indications, however, are insufficient to fully demonstrate improved survival after pulmonary rehabilitation. Moreover, it will be difficult to design a large-scale trial examining this in the future as randomisation to a control group would no longer be ethically acceptable since the benefits of pulmonary rehabilitation have been clearly demonstrated.
Finally, it was unequivocally demonstrated that the benefits achieved after pulmonary rehabilitation are relatively inexpensive, particularly when the reduction in healthcare resource utilisation caused by rehabilitation is taken into account. Indeed, Griffiths et al. 26 demonstrated that the cost of the programme was outweighed by the reduction in healthcare resource utilisation such that the incremental cost of adding rehabilitation to standard care was negative. The treatment is thus clearly cost-effective.
Selection of candidates
In general, it is very difficult to appropriately select candidates for a pulmonary rehabilitation programme on a scientific basis. The relation between functional variables before rehabilitation and response to training is weak and does not allow prediction of the response to rehabilitation. Nevertheless, some variables are, albeit weakly, related to response and may be used as a guideline on which to base clinical decisions.
Some prerequisites are necessary before starting a rehabilitation programme. Medical therapy should be optimised before starting the programme 30. Optimal medical therapy for COPD is now based on international consensus guidelines 1. In addition, most programmes require that the patient stops smoking before the programme or that they are at least enrolled in a smoking cessation programme. However, no data at present show that the effects of rehabilitation would be smaller in smokers as compared to nonsmokers 31.
ZuWallack et al. 32 examined the determinants of improvement in 12-min walking distance (12MWD) in a 6-week programme in 50 COPD patients. They found that the initial 12MWD, breathing reserve, initial maximal oxygen consumption and oxygen pulse were related to the improvement in 12MWD. On multiple regression analysis, the significant contributors to the improvement in 12MWD were forced expiratory volume in one second (FEV1) and initial 12MWD. Several later studies, however, demonstrated that FEV1 was not well related to the effect. Maltais et al. 33 showed that the training effects were similar in a group of COPD patients with an FEV1 of > or <40% of the predicted value in a 12-week programme. Casaburi et al. 34 showed clear training effects in an 8-week lasting rehabilitation programme in COPD patients with an FEV1 of <1 L and who were barely able to raise their lactate levels during exercise (peak lactate 2.2 mM). These patients showed clear increases in maximal oxygen consumption, reductions in ventilation and cardiac frequency at a given work rate and changes in oxygen uptake kinetics indicative of a training effect. Similarly, Troosters et al. 29 found that a measure of ventilatory limitation (peak exercise ventilation/maximal voluntary ventilation), initial maximal inspiratory pressure (PI,max) and peripheral muscle force were related to the response in 58 patients who followed a 6-month outpatient training programme. In these patients, FEV1 or other pulmonary function measures were not related to the training response. The factors related to response are summarised in table 2
.
|
There is some evidence that psychosocial factors may be related to adherence to the programme and also to its effects 37. Better social support improved adherence to the programme.
In general, candidates for a pulmonary rehabilitation programme are patients who, after smoking cessation, are well motivated to attend the programme and have severely impaired health status and functional status. Contraindications are either diseases that interfere with the rehabilitation process itself, e.g. rheumatoid arthritis, or put a patient at excessive risk during training, e.g. ischaemic heart disease. The latter contraindication is only relative since cardiologists admit such patients into rehabilitation programmes. The key concept is that any level of airflow obstruction is eligible for pulmonary rehabilitation. In other words, a patient with severe deconditioning and muscle weakness and an FEV1 of 30% pred or the same patient with an FEV1 of 60% pred are equally good candidates for a pulmonary rehabilitation programme 30.
Practical organisation
The programme may be administered on an outpatient or inpatient basis. In terms of cost-effectiveness, outpatient programmes are better, as they do not incur the cost of hospital admission. There is also some evidence on home programmes, which produce smaller effects than clinical programmes 38.
The programme administered is a multidisciplinary programme that consists of education, exercise training, peripheral muscle training, ventilatory muscle training, occupational therapy, dietary advice and psychosocial support. Exercise training and psychosocial support are the most effective components of the programme 39. The modality of exercise training is usually endurance training, although interval-type training may also be effective 40. Endurance training in COPD patients at a load of 60% of maximum for 20 min is possible, even in patients with severe airflow obstruction, although it may take several weeks before the appropriate training load and duration can be achieved 33. Exercises usually include walking on a treadmill, cycling, arm ergometry, stepping and peripheral muscle training.
Peripheral muscle resistance training has also been shown to be useful in the rehabilitation of COPD patients. Indeed, several studies showed that it clearly improved muscle force and health status in COPD patients 41–44. A recent study demonstrated that endurance training and peripheral muscle resistance training were equally effective in improving exercise capacity, muscle force and health status in patients with COPD and muscle weakness 44. Peripheral muscle stimulation also appears effective in COPD patients 45, 46. It is clear that these treatment modalities will play an increasingly important role in COPD rehabilitation in the future. The role of ventilatory muscle training is not yet clearly defined 47, 48, and more research is required in this area.
As ventilatory limitation is usually related to the response to training, it may be worthwhile to attempt to reduce the impact of ventilatory limitation. This could be attempted in three different ways. First, administration of oxygen is expected to reduce the ventilatory requirements during exercise. An improvement of the effects of pulmonary rehabilitation with oxygen has been suggested, but not unequivocally demonstrated 49–51. Secondly, heliox, a gas mixture in which oxygen and helium replace nitrogen, has also been tried. The lower density of helium reduces the work of breathing and improves ventilatory capacity. However, no significant effect on the training effect was found, although a trend was present 52. Finally, Hawkins et al. 53 demonstrated that noninvasive ventilation during exercise training enhanced the training effects. This is due to unloading of the ventilatory system such that greater training loads can be obtained. The procedure, however, is tedious in the sense that it is difficult to adjust ventilation to the increased ventilatory demands during exercise such that, at present, it is unclear what its place is in clinical routine.
The role of dietary intervention in COPD patients is clearly beyond the scope of the present article and is not discussed here. There is a need for education since COPD patients underestimate their disease and do not usually follow the proper treatments for their disease 12. General education programmes, however, on the whole do not appear effective 18. More targeted education programmes may be more effective.
Although the average programme has an occupational therapy component, its role has never clearly been established in this context. More specifically, it has not been clearly demonstrated whether rehabilitation improves the activities of daily living in a randomised controlled design, nor whether potential improvement is related to occupational therapy, exercise training or both. Whether programmes should be directed at those specific activities of daily living that pose problems in particular patients is also unclear.
After a rehabilitation programme, some maintenance is likely to be required to maintain the effects achieved during the rehabilitation period, although no randomised trials on this are currently available. Several modalities may be followed. Home exercise with or without supervision by a physiotherapist may be of help. COPD exercise or sports clubs may help as well. At present, it is not clear whether a rehabilitation programme automatically leads to a greater activity level after the programme, although the persistent improvement is usually attributed to increased activity.
Future developments
The question as to whether the effects of rehabilitation may be enhanced by pharmacological therapy remains an interesting one. Several treatments have been studied in this respect. A study on growth hormone therapy in 16 COPD patients demonstrated that, with 3 weeks of this therapy combined with rehabilitation, lean body mass increased, but outcome variables such as PI,max, handgrip strength and subjective wellbeing were not affected 54. The 6MWD even decreased significantly with this therapy. A large study with anabolic steroids was also conducted in 217 COPD patients. Treatment with nutritional intervention and nandrolone decanoate combined with rehabilitation was shown to increase PI,max more than rehabilitation alone 55 but not more than rehabilitation combined with nutritional intervention. The gain in fat-free mass was also greater with the combined treatment. Again, there was no difference between nutritional intervention and anabolic steroids and nutritional intervention alone. Testosterone therapy was shown to have an additional effect on muscle strength in COPD patients with low testosterone levels in combination with resistance training 56.
There is clear evidence that muscle weakness in COPD patients may be related to increased circulating levels of tumour necrosis factor-
or interleukin-6 or –8 27, 57, 58. Levels of these cytokines are elevated in stable patients with recent weight loss 58 or during exacerbations of COPD 27, 59. Recent evidence demonstrates that muscle weakness in the latter patients correlates with serum interleukin-8 levels 59. As a consequence, treatment with antitumour necrosis factor-
and anti-interleukin-6 and –8 may be considered in the future, not only to affect the progression of the disease but also to reduce muscle weakness. The response to a rehabilitation programme can be influenced by expression of these cytokines. Also, serum levels of insulin-like growth factor-I correlate with muscle weakness in COPD patients 59.
| Lung volume reduction surgery |
|---|
|
|
|---|
Surprisingly, few randomised studies have been performed. The National Emphysema Treatment Trial (NETT), which is still running, aims to compare the effects of LVRS to those of continued rehabilitation, and was intended to be the definitive randomised controlled study on survival and exercise capacity after LVRS 66. The study has a duration of 5 yrs with a fixed end-point and a variable follow-up period. The original recruitment target of 4,500 patients was reduced to 2,300 patients and, in June 2001, 1,033 patients were randomised. The results of this trial will be available soon. The results of six randomised studies are already available 67–72. They are summarised in table 3
. Full results have been published for only three of these trials. They showed that LVRS reduced hyperinflation (RV and TLC) and improved FEV1 and forced vital capacity (FVC) compared to control. In addition, the 6MWD or shuttle walk test increased significantly and quality of life improved. The effects on the latter variables were substantial. In general, inspiratory muscle function (PI,max and maximal transdiaphragmatic pressure) also improved, confirming the results of more-targeted studies 74–76. The improvements in inspiratory muscle function are probably due to geometric factors rather than intrinsic improvements in inspiratory muscle function 77, 78. LVRS is unique in the sense that it is the only procedure that causes substantial improvements in lung function in end-stage emphysema, besides lung transplantation. Rehabilitation improves exercise capacity and health status as well, but does not affect pulmonary function (see above).
|
Several hypotheses have been formulated to explain the beneficial effects observed after the procedure. A comprehensive discussion is clearly beyond the scope of the present publication, and has already been published 79. The hypotheses explaining the improvement include: improved lung mechanics, increased lung elastic recoil 80, 81, reduced intrinsic positive inspiratory pressure 82, reduced hyperinflation, increased respiratory muscle strength due to geometric factors 74–78, improved ventilation/perfusion matching, and reduced pulmonary vascular resistance. There is evidence that the major mode of action is resizing of the lungs 83–86. This means that, in emphysema, lung mechanics deteriorate because the lung is disproportionately expanded compared to the chest wall. The intervention restores the balance between lung size and chest wall size and hence lung mechanics improve. The intervention thus primarily increases vital capacity by reducing RV more than TLC. The increase in vital capacity is the prime determinant of the increase in FEV1. Together with increased elastic recoil, this improved lung mechanics reduces the work of breathing. This action mechanism is derived from several model analyses and experimental data. As a consequence, it is expected that postoperative improvement in FEV1 is largely determined by the improvement in vital capacity and is related to the preoperative RV/TLC ratio. Prospective testing of this model, however, has yielded relatively disappointing results. Indeed, Fessler et al. 87 demonstrated that the RV/TLC ratio was related to the postoperative change in FVC in a group of 13 patients, but, in a larger group of 78 patients, this relationship was not present. It thus remains difficult to predict the improvement on the basis of preoperative functional variables. Recently, Ingenito et al. 88 demonstrated that patients who respond tend to show a postoperative reduction in RV, whereas those who do not tend not to show this reduction. Their results are depicted schematically in figure 1
. This proved to be a better predictor of the effect than the increase in lung elastic recoil, confirming the reasoning developed above and the theoretical model of Fessler and Permutt 83. This model was also confirmed in a recent animal model 86.
|
Another fundamental question is to what extent LVRS may affect the rate of decline in FEV1. As, in general, a relationship between the rate of decline and the progression of emphysema is expected, this is an important question. Although some indications for accelerated decline after LVRS are present in the sense that, in many studies, the decline observed after LVRS appeared to exceed the expected decline in COPD patients 91, this has never clearly been substantiated directly. Kesten et al. 92 found accelerated decline over 1–2 yrs in six of 17 patients for whom they had follow-up spirometric data available. Gelb et al. 89 found similar rates of decline as before in five patients in whom they had follow-up spirometric data over a mean of 3.8 yrs before and after the operation. The decline in FEV1 was 116 mL·yr–1 before the operation and 78 mL·yr–1 afterwards. This observation thus refutes the accelerated decline after LVRS, but is based on a very small number of patients.
Selection of candidates
As clarity regarding the action mechanism is still absent to a large extent and since the size mismatch between the lungs and chest wall is not easy to measure noninvasively, it remains difficult to accurately predict the postoperative response from preoperative functional data. Indeed, although several factors were shown to be significantly related to response, the models, in general, only explain a relatively limited fraction of the variance such that they do not allow proper prediction of the response in an individual patient. Indeed, Brenner et al. 93 analysed the variables related to response in a study comparing buttressed stapled resections with laser coagulation. Buttressed stapled resections were clearly more effective than laser coagulation. Smoking history and TL,CO were the only significant predictors of outcome. FEV1 and blood gas levels did not add to the predictive power, nor did computed tomography (CT) scores. The model explained a total of only 45% of the variance observed. Similarly, the model of Thurnheer et al. 94, based on RV/TLC ratio and CT visual analogue score, explained only
40% of the variance in postoperative response. This, in general, means that accurate prediction of the response is not possible on a scientific basis.
Several studies have examined the relationship of preoperative functional variables to the response to surgery using univariate analysis. In general, clinical, physiological and imaging factors were examined. Several of these factors were related to the response. They are summarised in table 4
. For the clinical criteria, age >75–80 yrs is usually associated with poor outcome 64, 93, although this has been debated in other studies 95. Severe comorbidity is usually considered to be an exclusion criterion. Particular attention has been paid to the coexistence of nonsmall cell lung cancer 96 and coronary artery disease 97. This is a very specific problem and is not addressed further in the present article. The presence of severe pulmonary hypertension with a mean pulmonary artery pressure of >35 mmHg is usually considered to be a contraindication 98, although the procedure may significantly improve right ventricular function 99. Patients with
1-antitrypsin deficiency seem to respond less favourably in the long run 12, 100–102.
|
Besides these variables, identification of candidates has primarily focused on measures of hyperinflation. TLC 98, RV and RV/TLC 87, 94 ratio have been correlated to response. According to the model of Fessler and Permutt 83, the latter ratio is expected to be the best predictor of postoperative increase in FVC and FEV1. Although several studies show a relationship between RV/TLC ratio and outcome, in general these relationships have been relatively disappointing. An arterial carbon dioxide tension of >6.7–7.3 kPa (>50–55 mmHg) has also been related to poor outcome 12, 104, 105, although this is still debated by other groups 73, 106–108. Finally, a low preoperative exercise capacity (6MWD <200 m) has been related to poor outcome 105.
Findings with imaging have been related to the response to LVRS as well. Generally, heterogeneous emphysema with upper lobe predominance is the best indication for LVRS. McKenna et al. 64 analysed the results of 138 patients undergoing bilateral LVRS. Upper lobe-predominant emphysema was found in 106 (77%) patients, lower lobe-predominant emphysema in 10 (7%) and diffusely homogeneous emphysema in 22 (16%). The greatest improvement in FEV1 (73%) was noted in patients with upper lobe emphysema, whereas it was smallest in patients with diffuse emphysema. In the latter patients, however, a 38% improvement in FEV1 was still observed. Similar results were obtained by Weder et al. 109. They used a qualitative emphysema CT scoring system and demonstrated that, 3 months after a bilateral thoracoscopic procedure, FEV1 improved by 81% in the group with markedly heterogeneous emphysema, 44% in the group with intermediately heterogeneous emphysema and 31% in the group with homogeneous emphysema. Improvement in dyspnoea, however, was similar in the three groups. The same group also published follow-up results after 2 yrs 110. This confirmed the greater initial improvement in the heterogeneous group, but also demonstrated faster decline in this group. Dyspnoea worsened in all three groups, but it remained improved at 24 months compared to baseline. Survival was greatest in the heterogeneous group, but the functional data at onset were best in this group as well. These results were confirmed by Wisser et al. 111.
Thurnheer et al. 94 compared the qualitative assessment of emphysema heterogeneity on high-resolution CT (HRCT) with that on lung perfusion scintigraphy in the prediction of response in 70 patients undergoing bilateral video-assisted LVRS. Patients with homogeneous perfusion experienced less short-term improvement in FEV1 (23%) compared with those with intermediately heterogeneous perfusion (38%) or markedly heterogeneous perfusion (57%). However, on multivariate analysis, preoperative RV/TLC ratio (partial r=0.24) and heterogeneity on HRCT (partial r=0.28) were clearly more powerful predictors than heterogeneity on perfusion scintigraphy (partial r=0.01). The contribution of scintigraphy was not significant on multiple regression analysis. Ingenito et al. 112 further confirmed the superiority of physiological criteria over upper/lower perfusion ratio on scintigraphy in predicting the response of FEV1 in a cohort of 50 patients undergoing bilateral LVRS. Finally, quantitative CT data have also been used recently to predict FEV1 and exercise response to bilateral LVRS in 21 patients. The presence of emphysema in the rind area of the lung was shown to be a better predictor of response than its presence in the core of the lung 113. This is probably due to the fact that emphysematous rind zones are more accessible to surgery than the core zones. This is, however, unlikely to lead to a fundamentally better prediction of postoperative response than in the past as the best r2 obtained in this study was still
0.40, similar to previously published predictions based on physiological variables (see above).
Although several of the above-mentioned studies substantiate a better response in patients with heterogeneous emphysema, a number of caveats should be added to that conclusion. First, patients with homogeneous emphysema, in general, improve as well but less than those with heterogeneous emphysema. Secondly, the difference in response also appears to depend upon the variable studied. Weder et al. 109 found greater FEV1 response in heterogeneous emphysema, but no greater response in dyspnoea. Finally, the best conceptual model for explaining the improvement in lung mechanics after LVRS also allows for improvement in patients with homogeneous emphysema 83.
Taken together, it is not possible to properly select candidates for LVRS on a scientific basis. As the most plausible mechanism of action appears to be resizing of the lungs, better ways of assessing the mismatch between chest wall size and lung size might contribute to a better selection of candidates. Some variables may be used as a clinical guideline, but they do not allow fine discrimination between responders and nonresponders. The most significant variables appear to be RV/TLC ratio and emphysema heterogeneity on imaging studies.
Practical organisation
An exhaustive description of the surgical procedure is beyond the scope of the present article, and several excellent reviews have recently been published on this topic 114–117. The surgical approach has varied and has included: median sternotomy, standard thoracotomy, and unilateral or bilateral video-assisted thoracoscopic surgery. Some authors also promote laser ablation, but comparative trials have shown this technique to be inferior 93, 118. In general, bilateral procedures produce better effects than unilateral ones 60. Several studies demonstrated benefits of the video-assisted thoracoscopic surgical procedures over median sternotomy, although similar improvements in lung function were obtained.
The short-term surgical mortality ranges 0–15%. With proper patient selection and surgical procedure, current mortality is likely to be <5%. Postoperative complications are: postoperative air leaks, subcutaneous emphysema, pneumonia, pneumothorax, and prolonged mechanical ventilation. As postoperative air leak is a frequent complication, some groups have developed no-cut plication procedures to reduce the incidence of air leaks 119.
Future developments
Although LVRS definitely improves the condition of some patients, the controversy continues. This is exemplified by a statement made by the editor of the The New England Journal of Medicine in a recent editorial commenting on the results of the high-risk group of the NETT trial, "...in my opinion, it does not make sense for anyone to undergo LVRS outside a controlled trial..." 120. Conversely, LVRS has without question offered at least temporary palliation to patients for whom little other treatment was available.
Future developments will be directed primarily at developing application modalities that do not exhibit the relatively high postoperative mortality of the present surgical procedures. Technological developments that simplify surgical procedures are one example 119, 121 and bronchoscopic procedures another 122. The former include no-cut plication procedures 119, 121 and the latter procedures in which bronchoscopic collapse is obtained which is consolidated with tissue glue 123 or the implantation of valves 122. These technological developments have been shown to be feasible in animals. Recent evidence shows that they may also be applicable to patients with emphysema 122. At present, however, no randomised studies are available and so the effects still need to be better demonstrated. In addition, the long-term results of such procedures are largely unknown. These need to be available before their use in the treatment of emphysema can be recommended.
| ACKNOWLEDGEMENTS |
|---|
|
|
|---|
| REFERENCES |
|---|
|
|
|---|
levels and weight loss in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1994;150:1453–1455.[Abstract]
1-antitrypsin deficiency versus smoker's emphysema. Eur Respir J 1998;12:1028–1032.[Abstract]
1-antitrypsin emphysema. Eur Respir J 1999;14:928–933.