Adjuvant Chemotherapy for Lung Cancer
http://www.100md.com
《新英格兰医药杂志》
To the Editor: The International Adjuvant Lung Cancer Trial (IALT) Collaborative Group (Jan. 22 issue)1 concludes that their study demonstrates that "cisplatin-based adjuvant chemotherapy improves survival among patients with completely resected non–small-cell lung cancer." The IALT Collaborative Group is to be commended for a study that represents an important milestone in our approach to the treatment of such patients. However, in examining Figure 2 of the article, we find that 34.5 percent of patients with stage I disease who received chemotherapy died, as compared with 35.1 percent of the control group. The corresponding rates for patients with stage II disease were 53.5 percent and 56.8 percent and, for those with stage III disease, 62.6 percent and 70.1 percent. Thus, it seems that there is an inverse correlation between the stage of disease and the efficacy of adjuvant chemotherapy, with little if any benefit for patients with stage I disease. Since 0.8 percent of the patients who received adjuvant chemotherapy died of toxic effects of the treatment, we question whether the benefit of adjuvant chemotherapy has been shown to outweigh the risk for patients with early-stage disease.
Nadeem Ikhlaque, M.D.
Wright State University School of Medicine
Dayton, OH 45435
Michael A. Baumann, M.D.
Department of Veterans Affairs
Dayton, OH 45428
michael.baumann@wright.edu
References
The International Adjuvant Lung Cancer Trial Collaborative Group. Cisplatin-based adjuvant chemotherapy in patients with completely resected non-small-cell lung cancer. N Engl J Med 2004;350:351-360.
To the Editor: The IALT investigators conclude that cisplatin-based adjuvant chemotherapy improves survival in patients with completely resected non–small-cell lung cancer. However, the survival advantage associated with adjuvant chemotherapy was not the same for all the patients in the study. If we consider the hazard ratios, for example, patients older than 64 years of age and those with a World Health Organization (WHO) performance status of 2 seem not to have benefited from the adjuvant chemotherapy. Indeed, in patients with a WHO performance status of 2, adjuvant chemotherapy had a detrimental effect. Likewise, adjuvant chemotherapy was less likely to have a beneficial effect in patients with stage I or II disease and a tumor classification of T1 than in those with stage III disease and a tumor classification of T2 or higher. It seems, therefore, that cisplatin-based adjuvant chemotherapy is indicated only in patients with completely resected non–small-cell lung cancer who are 64 years of age or younger and have a good performance status and in patients who are at high risk for relapse.
Abdullah Büyük?elik, M.D.
Güng?r Utkan, M.D.
Bülent Yal?in, M.D.
Ankara University School of Medicine
06100 Ankara, Turkey
drabdullah@superposta.com
To the Editor: The statistical methods used in IALT raise some questions. A total of 3300 patients were to be included to demonstrate a 5 percent increase in survival. Because of the low rate of enrollment, the trial was stopped after 56.6 percent of the planned number of patients had been recruited (1867 patients, 973 of whom died). For this reason, the final data should have been treated as they would in an interim analysis — that is, with a correction in the P value. In fact, two interim analyses were scheduled after 320 and 640 deaths, with a P value of less than 0.001, instead of 0.05, considered to indicate a significant result. A conservative correction would require a P value of less than 0.017 for three interim analyses, such as those conducted in IALT. Stricter methods would require a more stringent significance level,1,2 but in IALT, the analysis of five-year survival (the main end point) resulted in a P value of 0.03. The conclusions are that differences in survival could have happened by chance and that the benefit of adjuvant chemotherapy for non–small-cell lung cancer remains undetermined.
Jaime Feliu, M.D.
Enrique Espinosa, M.D.
Hospital La Paz
28029 Madrid, Spain
eespinosa00@terra.es
José Javier Sánchez, M.D.
Universidad Autónoma
28029 Madrid, Spain
References
O'Brien PC, Fleming TR. A multiple testing procedure for clinical trials. Biometrics 1979;35:549-556.
Haybittle JL. Repeated assessment of results in clinical trials of cancer treatment. Br J Radiol 1971;44:793-797.
To the Editor: In IALT, 932 patients received cisplatin-based adjuvant chemotherapy, and 935 patients served as controls. The difference in survival between the groups at five years was 4.1 percentage points in favor of chemotherapy. The authors state that this small difference is statistically significant. I think it is not. Why? One hundred forty-eight centers in 33 countries participated. The heterogeneity of the centers (e.g., Instituto Privado de Radioterapia, Mendoza, Argentina, and University Hospital of Troms?, Oslo, Norway) must translate into heterogeneity of care. With overall survival as the end point, the time from relapse to death plays an important role. Palliative treatment (e.g., with expensive antibiotics) from the time of relapse to death certainly differs among Macedonia, Morocco, Sweden, and the United States. The small difference of survival at five years may just be a reflection of geographic variation. In his accompanying editorial,1 Blum suggests that cisplatin-based adjuvant chemotherapy is a new standard of care. On the grounds of this study, Blum's suggestion is certainly premature.
Christian Sauter, M.D.
University of Zurich
CH-8057 Zurich, Switzerland
cesauter@bluewin.ch
References
Blum RH. Adjuvant chemotherapy for lung cancer -- a new standard of care. N Engl J Med 2004;350:404-405.
To the Editor: In their study of adjuvant chemotherapy for lung cancer, the IALT investigators found an improvement in survival of 4.1 percentage points, which means that 25 patients have to be treated for just 1 to benefit. Therefore, the suggestion that this therapy represents "a new standard of care" cannot be accepted without supporting data on quality of life, data from a cost–benefit analysis, or both.
Graeme W. Morgan, F.R.A.N.Z.C.R.
Royal North Shore Hospital
Sydney 2065, Australia
The authors reply: Dr. Sauter raises the question of heterogeneity related to the numbers of participating centers (148) and countries (33). In our analysis, we compared patients who received chemotherapy and control patients only within each center; the reported benefit with chemotherapy is the average of the benefit observed at each center. Moreover, the effect of adjuvant chemotherapy on disease-free survival is not modified by the effectiveness of palliative treatments given after relapse. Finally, the large geographic basis of our study broadens the extent to which the results may be generalized.
Dr. Feliu and colleagues question our statistical rules and suggest that a P value of 0.017 should have been used for the two interim analyses and for the final one. The trial was not stopped because of the results of the interim analyses, of which the steering committee was unaware, but because of the low accrual. Therefore, the analysis presented should indeed be considered the final analysis. We chose Haybittle's rule, which requires more extreme results (P<0.001) to stop a trial early, but used a P value of 0.05 for the last analysis.
Drs. Ikhlaque and Baumann and Dr. Büyük?elik and colleagues believe that they can identify subgroups of patients who do not benefit from chemotherapy. The correct way to analyze such data is not to perform subgroup analyses, but to study the interaction between covariates and the effect of chemotherapy.1 In our study, none of the tests of interaction between covariates and the effect of treatment were significant, either for overall survival or for disease-free survival. Interaction tests have a low power. To improve this power, a pooled analysis of four large randomized trials, the Lung Adjuvant Cisplatin Evaluation project, is under way.
Dr. Morgan would like to see information from quality-of-life and cost–benefit analyses before chemotherapy is adopted as a standard of treatment for non–small-cell lung cancer. The most commonly used chemotherapy in our trial — three to four cycles of cisplatin and etoposide — is inexpensive, it is administered over a period of two to three months, and its toxic effects are well known and manageable. Furthermore, adjuvant chemotherapy was accepted as a standard treatment for breast and colorectal cancer before quality-of-life and cost-effectiveness information became available. IALT validates the concept of adjuvant chemotherapy in non–small-cell lung cancer. It opens the way for a new generation of trials investigating drugs and timing.
Ariane Dunant, M.S.
Jean-Pierre Pignon, M.D., Ph.D.
Thierry Le Chevalier, M.D.
Institut Gustave-Roussy
94805 Villejuif, France
References
Assmann SF, Pocock SJ, Enos LE, Kasten LE. Subgroup analysis and other (mis)uses of baseline data in clinical trials. Lancet 2000;355:1064-1069.
Nadeem Ikhlaque, M.D.
Wright State University School of Medicine
Dayton, OH 45435
Michael A. Baumann, M.D.
Department of Veterans Affairs
Dayton, OH 45428
michael.baumann@wright.edu
References
The International Adjuvant Lung Cancer Trial Collaborative Group. Cisplatin-based adjuvant chemotherapy in patients with completely resected non-small-cell lung cancer. N Engl J Med 2004;350:351-360.
To the Editor: The IALT investigators conclude that cisplatin-based adjuvant chemotherapy improves survival in patients with completely resected non–small-cell lung cancer. However, the survival advantage associated with adjuvant chemotherapy was not the same for all the patients in the study. If we consider the hazard ratios, for example, patients older than 64 years of age and those with a World Health Organization (WHO) performance status of 2 seem not to have benefited from the adjuvant chemotherapy. Indeed, in patients with a WHO performance status of 2, adjuvant chemotherapy had a detrimental effect. Likewise, adjuvant chemotherapy was less likely to have a beneficial effect in patients with stage I or II disease and a tumor classification of T1 than in those with stage III disease and a tumor classification of T2 or higher. It seems, therefore, that cisplatin-based adjuvant chemotherapy is indicated only in patients with completely resected non–small-cell lung cancer who are 64 years of age or younger and have a good performance status and in patients who are at high risk for relapse.
Abdullah Büyük?elik, M.D.
Güng?r Utkan, M.D.
Bülent Yal?in, M.D.
Ankara University School of Medicine
06100 Ankara, Turkey
drabdullah@superposta.com
To the Editor: The statistical methods used in IALT raise some questions. A total of 3300 patients were to be included to demonstrate a 5 percent increase in survival. Because of the low rate of enrollment, the trial was stopped after 56.6 percent of the planned number of patients had been recruited (1867 patients, 973 of whom died). For this reason, the final data should have been treated as they would in an interim analysis — that is, with a correction in the P value. In fact, two interim analyses were scheduled after 320 and 640 deaths, with a P value of less than 0.001, instead of 0.05, considered to indicate a significant result. A conservative correction would require a P value of less than 0.017 for three interim analyses, such as those conducted in IALT. Stricter methods would require a more stringent significance level,1,2 but in IALT, the analysis of five-year survival (the main end point) resulted in a P value of 0.03. The conclusions are that differences in survival could have happened by chance and that the benefit of adjuvant chemotherapy for non–small-cell lung cancer remains undetermined.
Jaime Feliu, M.D.
Enrique Espinosa, M.D.
Hospital La Paz
28029 Madrid, Spain
eespinosa00@terra.es
José Javier Sánchez, M.D.
Universidad Autónoma
28029 Madrid, Spain
References
O'Brien PC, Fleming TR. A multiple testing procedure for clinical trials. Biometrics 1979;35:549-556.
Haybittle JL. Repeated assessment of results in clinical trials of cancer treatment. Br J Radiol 1971;44:793-797.
To the Editor: In IALT, 932 patients received cisplatin-based adjuvant chemotherapy, and 935 patients served as controls. The difference in survival between the groups at five years was 4.1 percentage points in favor of chemotherapy. The authors state that this small difference is statistically significant. I think it is not. Why? One hundred forty-eight centers in 33 countries participated. The heterogeneity of the centers (e.g., Instituto Privado de Radioterapia, Mendoza, Argentina, and University Hospital of Troms?, Oslo, Norway) must translate into heterogeneity of care. With overall survival as the end point, the time from relapse to death plays an important role. Palliative treatment (e.g., with expensive antibiotics) from the time of relapse to death certainly differs among Macedonia, Morocco, Sweden, and the United States. The small difference of survival at five years may just be a reflection of geographic variation. In his accompanying editorial,1 Blum suggests that cisplatin-based adjuvant chemotherapy is a new standard of care. On the grounds of this study, Blum's suggestion is certainly premature.
Christian Sauter, M.D.
University of Zurich
CH-8057 Zurich, Switzerland
cesauter@bluewin.ch
References
Blum RH. Adjuvant chemotherapy for lung cancer -- a new standard of care. N Engl J Med 2004;350:404-405.
To the Editor: In their study of adjuvant chemotherapy for lung cancer, the IALT investigators found an improvement in survival of 4.1 percentage points, which means that 25 patients have to be treated for just 1 to benefit. Therefore, the suggestion that this therapy represents "a new standard of care" cannot be accepted without supporting data on quality of life, data from a cost–benefit analysis, or both.
Graeme W. Morgan, F.R.A.N.Z.C.R.
Royal North Shore Hospital
Sydney 2065, Australia
The authors reply: Dr. Sauter raises the question of heterogeneity related to the numbers of participating centers (148) and countries (33). In our analysis, we compared patients who received chemotherapy and control patients only within each center; the reported benefit with chemotherapy is the average of the benefit observed at each center. Moreover, the effect of adjuvant chemotherapy on disease-free survival is not modified by the effectiveness of palliative treatments given after relapse. Finally, the large geographic basis of our study broadens the extent to which the results may be generalized.
Dr. Feliu and colleagues question our statistical rules and suggest that a P value of 0.017 should have been used for the two interim analyses and for the final one. The trial was not stopped because of the results of the interim analyses, of which the steering committee was unaware, but because of the low accrual. Therefore, the analysis presented should indeed be considered the final analysis. We chose Haybittle's rule, which requires more extreme results (P<0.001) to stop a trial early, but used a P value of 0.05 for the last analysis.
Drs. Ikhlaque and Baumann and Dr. Büyük?elik and colleagues believe that they can identify subgroups of patients who do not benefit from chemotherapy. The correct way to analyze such data is not to perform subgroup analyses, but to study the interaction between covariates and the effect of chemotherapy.1 In our study, none of the tests of interaction between covariates and the effect of treatment were significant, either for overall survival or for disease-free survival. Interaction tests have a low power. To improve this power, a pooled analysis of four large randomized trials, the Lung Adjuvant Cisplatin Evaluation project, is under way.
Dr. Morgan would like to see information from quality-of-life and cost–benefit analyses before chemotherapy is adopted as a standard of treatment for non–small-cell lung cancer. The most commonly used chemotherapy in our trial — three to four cycles of cisplatin and etoposide — is inexpensive, it is administered over a period of two to three months, and its toxic effects are well known and manageable. Furthermore, adjuvant chemotherapy was accepted as a standard treatment for breast and colorectal cancer before quality-of-life and cost-effectiveness information became available. IALT validates the concept of adjuvant chemotherapy in non–small-cell lung cancer. It opens the way for a new generation of trials investigating drugs and timing.
Ariane Dunant, M.S.
Jean-Pierre Pignon, M.D., Ph.D.
Thierry Le Chevalier, M.D.
Institut Gustave-Roussy
94805 Villejuif, France
References
Assmann SF, Pocock SJ, Enos LE, Kasten LE. Subgroup analysis and other (mis)uses of baseline data in clinical trials. Lancet 2000;355:1064-1069.