Portal de Periódicos da CAPES

The National Emphysema Treatment Trial: a paradigm for future surgical trials

2001; Elsevier BV; Volume: 72; Issue: 2 Linguagem: Inglês

10.1016/s0003-4975(01)02396-7

1552-6259

Douglas E. Wood, Malcolm M. DeCamp,

Pleural and Pulmonary Diseases

Over the past century, a variety of surgical procedures have been proposed to improve the symptoms and quality of life in patients with severe emphysema. Costochondrectomy and thoracoplasty were performed in an attempt to alter the configuration of the chest cavity, and phrenic nerve ablation or induced pneumoperitoneum were attempted to restore the curvature of the diaphragm. Improvement of pulmonary blood flow was the intent of pleurodesis, and there were other techniques of airway stenting or autonomic denervation [1Deslauriers J. History of surgery for emphysema.Semin Thorac Cardiovasc Surg. 1996; 8: 43-51PubMed Google Scholar]. Each of these procedures enjoyed a period of popularity, at least among their surgical champions, who believed firmly, yet wrongly, that their intervention provided some benefit to patients with severe emphysema. The history of surgeryhas many such procedures that were initially popular and promoted in the absence of sound scientific data, only to wither under more rigorous examination. With the benefit of historical and scientific perspective, as academic thoracic surgeons we should anticipate these criticisms of our innovations and answer them with scientifically sound confirmation or disproof of new surgical hypotheses.Doctor Otto Brantigan described the techniques and physiologic principles underlying what is now known as lung volume reduction surgery (LVRS). [2Brantigan O. Mueller E. Kress M. A surgical approach to pulmonary emphysema.Amer Rev Resp Dis. 1959; 80: 194-201PubMed Google Scholar] The modern experience with LVRS was first described by Dr Joel Cooper and his colleagues. The Washington University group reported their first 20 LVRS patients with an objective improvement in forced expiratory volume in 1 second (FEV1) of 82% [3Cooper J.D. Trulock E.P. Triantafillou A.N. et al.Bilateral pneumectomy (volume reduction) for chronic obstructive pulmonary disease.J Thorac Cardiovasc Surg. 1995; 109: 106-116Abstract Full Text Full Text PDF PubMed Scopus (888) Google Scholar].When one examines the initial seven publications on LVRS, totaling 738 patients, the results are encouraging. The improvement in FEV1 averaged 61%, the 6-minute walk increased 46%, and there was a 62% rate of liberation from oxygen dependence. These results were accompanied by an improvement in dyspnea scale and quality of life, at a cost of 2.5% to 10% operative mortality and an 11- to 17-day hospitalization.It would appear that these data are adequate to support the acceptance of LVRS as a procedure with objective measures of efficacy, and with acceptable morbidity and mortality. With these data, some would argue that a randomized clinical trial is not needed and is, in fact, unethical. They would argue that it is unethical to randomize patients to medical therapy when surgery appears to produce such dramatic objective and subjective improvement.Although it may appear that the case is compelling for acceptance of LVRS without a randomized trial, there are a number of criticisms of the published reports. First of all, the majority of these reports involve small numbers of patients and vague or variable patient selection criteria. Second, all of these series were case controlled, with no randomized assessments to confirm that the outcomes were truly attributable to the surgical procedure, rather than other aspects of patient selection and medical care of a highly select patient subgroup. Third, all authors reported mean results, with most not providing standard deviation data in the objective outcomes. Where standard deviations were provided, they were often greater than the mean improvement, representing patients with negative outcomes or minimal improvement. Careful review of these data reveal that from 20% to 50% of patients actually had negligible objective improvement in FEV1 in these initial series [4Kotloff R.M. Tino G. Bavaria J.E. et al.Bilateral lung volume reduction surgery for advanced emphysema.Chest. 1996; 110: 1399-1406Crossref PubMed Scopus (150) Google Scholar, 5Argenziano M. Thomashow B. Jellen P.A. et al.Functional comparison of unilateral versus bilateral lung volume reduction surgery.Ann Thorac Surg. 1997; 64: 321-327Abstract Full Text PDF PubMed Scopus (66) Google Scholar, 6Ingenito E.P. Evans R.B. Loring S.H. et al.Relation between preoperative inspiratory lung resistance and the outcome of lung-volume reduction surgery for emphysema.N Engl J Med. 1998; 338: 1181-1185Crossref PubMed Scopus (120) Google Scholar]. Fourth, and probably most significantly, most of these series were weakened by incomplete follow-up, allowing follow-up selection bias to cloud the interpretation of the results. Short-term follow-up obscured the potential long-term deleterious effects of LVRS, and incomplete follow-up may have failed to include patients with poor outcomes who either suffered late mortality or refused to return for follow-up assessments. Finally, it was clear that there might also be an unintentional but systematic selection bias, with some centers with poor outcomes failing to publish their results. Although several major academic centers reported positive outcomes and low operative morbidity and mortality, the unreported outcomes were largely unknown. Published mortality after LVRS, although mostly in the range of 2.5% to 10%, was as high as 19% in some reports, and in unpublished experience, a mortality as high as 50% has been reported [7Fessler H.E. Wise R.A. Lung volume reduction surgery is less really more?.Am Rev Respir Crit Care Med. 1999; 159: 1031-1035Crossref PubMed Scopus (52) Google Scholar, 8Naunheim K.S. Ferguson M.K. The current status of lung volume reduction operations for emphysema.Ann Thor Surg. 1996; 62: 601-612Abstract Full Text PDF PubMed Scopus (73) Google Scholar].The only cohort of LVRS patients with complete follow-up is from Medicare data on 722 patients who had claims submitted for LVRS between October 1995 and January 1996. This revealed a mortality at 3, 12, and 18 months of 14%, 23%, and 28%, respectively [9Health Care Financing Administration. Report to Congress. Lung volume reduction surgery and Medicare coverage policy: implications of recently published evidence. Washington DC: Department of Health and Human Services, 1998.Google Scholar]. Obviously, it is unknown whether this high mortality represented the poor natural history of these emphysema patients, or whether it represented a significant negative outcome from LVRS that was not apparent in reports from individual centers. Acute hospitalizations actually increased in patients after LVRS. In the year before LVRS, these patients had 197 hospitalizations and approximately 2,300 hospital days, but in the year after LVRS, this doubled to 304 hospitalizations accounting for almost 5,700 hospital days, excluding the hospitalization for the LVRS [9Health Care Financing Administration. Report to Congress. Lung volume reduction surgery and Medicare coverage policy: implications of recently published evidence. Washington DC: Department of Health and Human Services, 1998.Google Scholar].As the preliminary experience with LVRS accumulated, the National Institutes of Health (NIH) convened a conference to evaluate LVRS and concluded that there was a “…need for systematic study of LVRS selection criteria and long-term efficacy” [10Weinmann G.G. Hyatt R. NHLBI Workshop Summary Evaluation and research in lung volume reduction surgery.Am J Respir Crit Care Med. 1996; 154: 1913-1918Crossref PubMed Scopus (54) Google Scholar]. The Agency for Health Care Policy and Research (AHCPR) concluded that there were inadequate scientific data regarding efficacy and risk of LVRS and that “…a prospective trial of LVRS…is both ethically supportable and scientifically essential” [11Holohan FV, Handelsman H. Lung volume reduction surgery for end-stage chronic obstructive pulmonary disease. Washington, DC, Health Technology Assessment, Agency for Health Care Policy and Research. US Department of Health and Human Services 1996, Publication #96-0062.Google Scholar].It is this setting in which the National Emphysema Treatment Trial (NETT) was initiated. The NETT is a prospective, randomized, multiinstitutional clinical trial that compares medical therapy for emphysema to medical therapy plus LVRS. The trial is supported by an historic collaboration between the Health Care Financing Administration (HCFA; Medicare) and the NIH, with the cost of allowable clinical services reimbursed by Medicare, and the NIH supporting the cost of the associated research questions. Seventeen clinical centers are involved in the trial and have a recruitment goal to achieve randomization of 2,500 patients in the 5 years of the study. The primary outcome measures are survival and maximum exercise capacity, and secondary outcome measures examine quality of life, disease-specific symptoms, cost-effectiveness, pulmonary function, gas exchange, oxygen utilization, radiologic assessments, and psychomotor function.The major question of whether such a prospective randomized trial is appropriate and needed is whether there is a legitimate professional uncertainty regarding the efficacy of LVRS within the medical community. Review of the data regarding LVRS has resulted in strong opinions, but these opinions include both LVRS proponents as well as LVRS skeptics.What limits our ability to conduct a randomized trial in general? We cannot perform a legitimate randomized clinical trial in the absence of supporting pilot data, either from phase II clinical trials or from laboratory research. We cannot ethically conduct a trial if there is clear or perceived efficacy for the new therapy within the medical community as a whole. A trial may also be limited by factors of reimbursement, with hospitals, physicians, and industry producing incredible innovation, but at a cost, and with an incentive to increase personal reimbursement. This is opposed by HCFA, the insurance industry, and public policy figures who may desire proof of efficacy at best, or stifle innovation at worst, in their efforts to control health care costs. Desperate patients clamoring for any intervention at any cost add an emotional and political component that may interfere with a legitimate clinical trial.Trials may be opposed by important individuals or groups who have a bias that prevents their participation, or have an incentive not to participate. Some have suggested that randomized trials are inherently unpopular with surgeons. Two separate editorials published in Lancet have suggested that surgeons are under pressure to remain “fashionable” in the procedures that they perform, with market forces from the public, industry, and hospital administrators to innovate and attract patients potentially superceding the establishment of efficacy in a randomized trial [12Horton R. Surgical research or comic opera questions, but few answers.Lancet. 1996; 347: 984-985Abstract PubMed Scopus (303) Google Scholar, 13Surgical research under scutiny.Lancet. 1993; 342: 187-188Abstract PubMed Scopus (30) Google Scholar].Doctor Richard Jonas has carefully examined this deficit in randomized trials in cardiothoracic surgery and concluded that “surgeons have been content to publish selected case series reflecting the results of perhaps one or two senior surgeons in a highly selected time frame. Control groups are often selected that are historical or from the literature where perhaps a completely different spectrum of patients has been treated. Surgical techniques are usually not standardized. …Case series provide the weakest evidence for assessing the efficacy of a treatment or for establishing causation. Some critics suggest that case series are not research at all and that unplanned observations are hopelessly invalid with the potentially lethal flaws of uncontrolled bias…” [14Jonas R.A. Randomized evaluations of treatment of pediatric cardiovascular disease a surgeon’s perspective.Prog Pediat Cardiol. 1997; 7: 97-104Abstract Full Text PDF Scopus (2) Google Scholar].The critical criteria for the legitimate conduct of a prospective randomized trial exists for lung volume reduction surgery. Phase II clinical trials exist to support the consideration of LVRS. A state of professional uncertainty, or equipoise, regarding LVRS clearly exists within the medical community. Some surgeons have opposed the trial due to their strong belief that the current nonrandomized data are adequate to establish efficacy and to define the contraindications, morbidity, and mortality of the procedure. However, we as surgeons suffer a lack of credibility when we challenge what, to others, appears to be an appropriate clinical trial. Some would gently criticize surgeons as not having the personality or culture to submit patients to randomized trials, or the patience to await the outcomes of randomized trials. Stronger criticisms may be that surgeons have a conflict of interest, with a self-interest in expanding a new surgical technique for the benefits of procedure-related reimbursement, expansion of relationships with industry, and development of an academic reputation.The National Emphysema Treatment Trial is an exciting model for large-scale clinical research in the future. The collaboration between the NIH and Medicare provides the combination of scientific merit along with the resources to fund large-scale clinical trials. This trial, and potential future trials, provide the opportunity for cardiothoracic surgeons to continue our clinical and scientific leadership. Perhaps there is a role within our professional organizations to initiate and organize new clinical trials, funded by a combination of NIH and Medicare dollars, as has been accomplished in the National Emphysema Treatment Trial. This could be a new and fertile ground of clinical investigation as well as a new role for our professional organizations.It is important that we continue to maintain our responsibilities as clinical scientists. First of all, we must continue to innovate, and our specialty can feel pride and satisfaction in our strong history of clinical innovation. Second, we must evaluate our innovations, and accept that there may be differences in opinions or conclusions regarding the efficacy of a new procedure. The presence of legitimate differences should drive us to conduct scientifically valid trials, even if they may be opposed to our biases or to our self-interests. All of this is with the ultimate goal of providing the best care for our patients, protecting them from unproven therapies, advocating for them to assure access to clinical trials and new therapies, and trying to lead, rather than disrupt, the struggle to maintain the fine line between ethical patient care and scientific clinical investigation. Over the past century, a variety of surgical procedures have been proposed to improve the symptoms and quality of life in patients with severe emphysema. Costochondrectomy and thoracoplasty were performed in an attempt to alter the configuration of the chest cavity, and phrenic nerve ablation or induced pneumoperitoneum were attempted to restore the curvature of the diaphragm. Improvement of pulmonary blood flow was the intent of pleurodesis, and there were other techniques of airway stenting or autonomic denervation [1Deslauriers J. History of surgery for emphysema.Semin Thorac Cardiovasc Surg. 1996; 8: 43-51PubMed Google Scholar]. Each of these procedures enjoyed a period of popularity, at least among their surgical champions, who believed firmly, yet wrongly, that their intervention provided some benefit to patients with severe emphysema. The history of surgeryhas many such procedures that were initially popular and promoted in the absence of sound scientific data, only to wither under more rigorous examination. With the benefit of historical and scientific perspective, as academic thoracic surgeons we should anticipate these criticisms of our innovations and answer them with scientifically sound confirmation or disproof of new surgical hypotheses. Doctor Otto Brantigan described the techniques and physiologic principles underlying what is now known as lung volume reduction surgery (LVRS). [2Brantigan O. Mueller E. Kress M. A surgical approach to pulmonary emphysema.Amer Rev Resp Dis. 1959; 80: 194-201PubMed Google Scholar] The modern experience with LVRS was first described by Dr Joel Cooper and his colleagues. The Washington University group reported their first 20 LVRS patients with an objective improvement in forced expiratory volume in 1 second (FEV1) of 82% [3Cooper J.D. Trulock E.P. Triantafillou A.N. et al.Bilateral pneumectomy (volume reduction) for chronic obstructive pulmonary disease.J Thorac Cardiovasc Surg. 1995; 109: 106-116Abstract Full Text Full Text PDF PubMed Scopus (888) Google Scholar]. When one examines the initial seven publications on LVRS, totaling 738 patients, the results are encouraging. The improvement in FEV1 averaged 61%, the 6-minute walk increased 46%, and there was a 62% rate of liberation from oxygen dependence. These results were accompanied by an improvement in dyspnea scale and quality of life, at a cost of 2.5% to 10% operative mortality and an 11- to 17-day hospitalization. It would appear that these data are adequate to support the acceptance of LVRS as a procedure with objective measures of efficacy, and with acceptable morbidity and mortality. With these data, some would argue that a randomized clinical trial is not needed and is, in fact, unethical. They would argue that it is unethical to randomize patients to medical therapy when surgery appears to produce such dramatic objective and subjective improvement. Although it may appear that the case is compelling for acceptance of LVRS without a randomized trial, there are a number of criticisms of the published reports. First of all, the majority of these reports involve small numbers of patients and vague or variable patient selection criteria. Second, all of these series were case controlled, with no randomized assessments to confirm that the outcomes were truly attributable to the surgical procedure, rather than other aspects of patient selection and medical care of a highly select patient subgroup. Third, all authors reported mean results, with most not providing standard deviation data in the objective outcomes. Where standard deviations were provided, they were often greater than the mean improvement, representing patients with negative outcomes or minimal improvement. Careful review of these data reveal that from 20% to 50% of patients actually had negligible objective improvement in FEV1 in these initial series [4Kotloff R.M. Tino G. Bavaria J.E. et al.Bilateral lung volume reduction surgery for advanced emphysema.Chest. 1996; 110: 1399-1406Crossref PubMed Scopus (150) Google Scholar, 5Argenziano M. Thomashow B. Jellen P.A. et al.Functional comparison of unilateral versus bilateral lung volume reduction surgery.Ann Thorac Surg. 1997; 64: 321-327Abstract Full Text PDF PubMed Scopus (66) Google Scholar, 6Ingenito E.P. Evans R.B. Loring S.H. et al.Relation between preoperative inspiratory lung resistance and the outcome of lung-volume reduction surgery for emphysema.N Engl J Med. 1998; 338: 1181-1185Crossref PubMed Scopus (120) Google Scholar]. Fourth, and probably most significantly, most of these series were weakened by incomplete follow-up, allowing follow-up selection bias to cloud the interpretation of the results. Short-term follow-up obscured the potential long-term deleterious effects of LVRS, and incomplete follow-up may have failed to include patients with poor outcomes who either suffered late mortality or refused to return for follow-up assessments. Finally, it was clear that there might also be an unintentional but systematic selection bias, with some centers with poor outcomes failing to publish their results. Although several major academic centers reported positive outcomes and low operative morbidity and mortality, the unreported outcomes were largely unknown. Published mortality after LVRS, although mostly in the range of 2.5% to 10%, was as high as 19% in some reports, and in unpublished experience, a mortality as high as 50% has been reported [7Fessler H.E. Wise R.A. Lung volume reduction surgery is less really more?.Am Rev Respir Crit Care Med. 1999; 159: 1031-1035Crossref PubMed Scopus (52) Google Scholar, 8Naunheim K.S. Ferguson M.K. The current status of lung volume reduction operations for emphysema.Ann Thor Surg. 1996; 62: 601-612Abstract Full Text PDF PubMed Scopus (73) Google Scholar]. The only cohort of LVRS patients with complete follow-up is from Medicare data on 722 patients who had claims submitted for LVRS between October 1995 and January 1996. This revealed a mortality at 3, 12, and 18 months of 14%, 23%, and 28%, respectively [9Health Care Financing Administration. Report to Congress. Lung volume reduction surgery and Medicare coverage policy: implications of recently published evidence. Washington DC: Department of Health and Human Services, 1998.Google Scholar]. Obviously, it is unknown whether this high mortality represented the poor natural history of these emphysema patients, or whether it represented a significant negative outcome from LVRS that was not apparent in reports from individual centers. Acute hospitalizations actually increased in patients after LVRS. In the year before LVRS, these patients had 197 hospitalizations and approximately 2,300 hospital days, but in the year after LVRS, this doubled to 304 hospitalizations accounting for almost 5,700 hospital days, excluding the hospitalization for the LVRS [9Health Care Financing Administration. Report to Congress. Lung volume reduction surgery and Medicare coverage policy: implications of recently published evidence. Washington DC: Department of Health and Human Services, 1998.Google Scholar]. As the preliminary experience with LVRS accumulated, the National Institutes of Health (NIH) convened a conference to evaluate LVRS and concluded that there was a “…need for systematic study of LVRS selection criteria and long-term efficacy” [10Weinmann G.G. Hyatt R. NHLBI Workshop Summary Evaluation and research in lung volume reduction surgery.Am J Respir Crit Care Med. 1996; 154: 1913-1918Crossref PubMed Scopus (54) Google Scholar]. The Agency for Health Care Policy and Research (AHCPR) concluded that there were inadequate scientific data regarding efficacy and risk of LVRS and that “…a prospective trial of LVRS…is both ethically supportable and scientifically essential” [11Holohan FV, Handelsman H. Lung volume reduction surgery for end-stage chronic obstructive pulmonary disease. Washington, DC, Health Technology Assessment, Agency for Health Care Policy and Research. US Department of Health and Human Services 1996, Publication #96-0062.Google Scholar]. It is this setting in which the National Emphysema Treatment Trial (NETT) was initiated. The NETT is a prospective, randomized, multiinstitutional clinical trial that compares medical therapy for emphysema to medical therapy plus LVRS. The trial is supported by an historic collaboration between the Health Care Financing Administration (HCFA; Medicare) and the NIH, with the cost of allowable clinical services reimbursed by Medicare, and the NIH supporting the cost of the associated research questions. Seventeen clinical centers are involved in the trial and have a recruitment goal to achieve randomization of 2,500 patients in the 5 years of the study. The primary outcome measures are survival and maximum exercise capacity, and secondary outcome measures examine quality of life, disease-specific symptoms, cost-effectiveness, pulmonary function, gas exchange, oxygen utilization, radiologic assessments, and psychomotor function. The major question of whether such a prospective randomized trial is appropriate and needed is whether there is a legitimate professional uncertainty regarding the efficacy of LVRS within the medical community. Review of the data regarding LVRS has resulted in strong opinions, but these opinions include both LVRS proponents as well as LVRS skeptics. What limits our ability to conduct a randomized trial in general? We cannot perform a legitimate randomized clinical trial in the absence of supporting pilot data, either from phase II clinical trials or from laboratory research. We cannot ethically conduct a trial if there is clear or perceived efficacy for the new therapy within the medical community as a whole. A trial may also be limited by factors of reimbursement, with hospitals, physicians, and industry producing incredible innovation, but at a cost, and with an incentive to increase personal reimbursement. This is opposed by HCFA, the insurance industry, and public policy figures who may desire proof of efficacy at best, or stifle innovation at worst, in their efforts to control health care costs. Desperate patients clamoring for any intervention at any cost add an emotional and political component that may interfere with a legitimate clinical trial. Trials may be opposed by important individuals or groups who have a bias that prevents their participation, or have an incentive not to participate. Some have suggested that randomized trials are inherently unpopular with surgeons. Two separate editorials published in Lancet have suggested that surgeons are under pressure to remain “fashionable” in the procedures that they perform, with market forces from the public, industry, and hospital administrators to innovate and attract patients potentially superceding the establishment of efficacy in a randomized trial [12Horton R. Surgical research or comic opera questions, but few answers.Lancet. 1996; 347: 984-985Abstract PubMed Scopus (303) Google Scholar, 13Surgical research under scutiny.Lancet. 1993; 342: 187-188Abstract PubMed Scopus (30) Google Scholar]. Doctor Richard Jonas has carefully examined this deficit in randomized trials in cardiothoracic surgery and concluded that “surgeons have been content to publish selected case series reflecting the results of perhaps one or two senior surgeons in a highly selected time frame. Control groups are often selected that are historical or from the literature where perhaps a completely different spectrum of patients has been treated. Surgical techniques are usually not standardized. …Case series provide the weakest evidence for assessing the efficacy of a treatment or for establishing causation. Some critics suggest that case series are not research at all and that unplanned observations are hopelessly invalid with the potentially lethal flaws of uncontrolled bias…” [14Jonas R.A. Randomized evaluations of treatment of pediatric cardiovascular disease a surgeon’s perspective.Prog Pediat Cardiol. 1997; 7: 97-104Abstract Full Text PDF Scopus (2) Google Scholar]. The critical criteria for the legitimate conduct of a prospective randomized trial exists for lung volume reduction surgery. Phase II clinical trials exist to support the consideration of LVRS. A state of professional uncertainty, or equipoise, regarding LVRS clearly exists within the medical community. Some surgeons have opposed the trial due to their strong belief that the current nonrandomized data are adequate to establish efficacy and to define the contraindications, morbidity, and mortality of the procedure. However, we as surgeons suffer a lack of credibility when we challenge what, to others, appears to be an appropriate clinical trial. Some would gently criticize surgeons as not having the personality or culture to submit patients to randomized trials, or the patience to await the outcomes of randomized trials. Stronger criticisms may be that surgeons have a conflict of interest, with a self-interest in expanding a new surgical technique for the benefits of procedure-related reimbursement, expansion of relationships with industry, and development of an academic reputation. The National Emphysema Treatment Trial is an exciting model for large-scale clinical research in the future. The collaboration between the NIH and Medicare provides the combination of scientific merit along with the resources to fund large-scale clinical trials. This trial, and potential future trials, provide the opportunity for cardiothoracic surgeons to continue our clinical and scientific leadership. Perhaps there is a role within our professional organizations to initiate and organize new clinical trials, funded by a combination of NIH and Medicare dollars, as has been accomplished in the National Emphysema Treatment Trial. This could be a new and fertile ground of clinical investigation as well as a new role for our professional organizations. It is important that we continue to maintain our responsibilities as clinical scientists. First of all, we must continue to innovate, and our specialty can feel pride and satisfaction in our strong history of clinical innovation. Second, we must evaluate our innovations, and accept that there may be differences in opinions or conclusions regarding the efficacy of a new procedure. The presence of legitimate differences should drive us to conduct scientifically valid trials, even if they may be opposed to our biases or to our self-interests. All of this is with the ultimate goal of providing the best care for our patients, protecting them from unproven therapies, advocating for them to assure access to clinical trials and new therapies, and trying to lead, rather than disrupt, the struggle to maintain the fine line between ethical patient care and scientific clinical investigation.