Correspondence

Chemotherapy plus Surgery versus Surgery Alone in Non-Small-Cell Lung Cancer

N Engl J Med 1994; 330:1756-1757June 16, 1994

Article

To the Editor:

We are concerned about several aspects of the study by Rosell et al. (Jan. 20 issue)1. The title is incorrect and misleading: the trial did not compare preoperative chemotherapy plus surgery with surgery alone; the trial compared preoperative chemotherapy plus surgery plus radiotherapy with surgery plus radiotherapy.

The rates of survival and disease-free survival were extremely poor in the surgery-plus-radiotherapy group. These rates would be expected among patients with metastatic non-small-cell lung cancer but are inferior to what would be expected in patients with resectable stage III disease. In several randomized trials, reviewed by Van Houtte,2 the five-year survival rate ranged from 21 to 36 percent for patients with N1 or N2 squamous-cell lung cancer treated with surgery and postoperative radiotherapy. These results seem substantially better than the 9 percent survival rate in the surgery-plus-radiotherapy group, anticipated by Rosell et al. However, none of their patients treated with surgery plus radiotherapy survived longer than two years. In other studies,3 the two-year survival rate among patients with unresectable stage IIIA or IIIB tumors treated with radiotherapy alone was between 15 percent and 33 percent. Nowhere in the paper do Rosell et al. discuss their poor results.

We are also concerned about the analysis of the results. According to the Results section, six patients in the group treated with chemotherapy, surgery, and radiotherapy (one patient with liver metastasis, one with a partial response who refused surgery, and four who did not have a complete resection) and three patients in the group treated with surgery and radiotherapy were never rendered free of disease by treatment. Yet in Fi1gure 1 disease-free survival is shown to be 100 percent at time 0, and it starts dropping only after three months in the group receiving chemotherapy, surgery, and radiotherapy. Since at the bottom of the figure it is clearly indicated that 30 patients were able to be evaluated in each group at time 0, the authors should clarify what they meant by “disease free,” when the evaluation was done, and how they scored an event with regard to the construction of these curves. There is a similar problem with the overall-survival curves.

Veronique Cocquyt, M.D.
Wilfried De Neve, M.D., Ph.D.
Simon J.-P. Van Belle, M.D., Ph.D.
University Hospital Ghent, Ghent, B 9000, Belgium

3 References

1
Rosell R, Gomez-Codina J, Camps C, et al. A randomized trial comparing preoperative chemotherapy plus surgery with surgery alone in patients with non-small-cell lung cancer. N Engl J Med 1994;330:153-158
Full Text | Web of Science | Medline

2
Van Houtte P. Postoperative radiotherapy for lung cancer. Lung Cancer 1991;7:57-64
CrossRef

3
Murren JR, Buzaid AC. Chemotherapy and radiation for the treatment of non-small-cell lung cancer: a critical review. Clin Chest Med 1993;14:161-171
Web of Science | Medline

To the Editor:

The ability of Rosell et al. to demonstrate a difference between treatment with chemotherapy and surgery and treatment with surgery alone in the management of stage IIIA non-small-cell lung cancer hinges on the similarity of the patients in the two treatment groups at the time of randomization. The authors point out a difference in the proportion with N2 disease between the groups: 25 of the 30 patients in the chemotherapy-plus-surgery group had stage N2 disease at the time of randomization, as compared with only 19 patients in the surgery group. The charitable explanation for this difference is that it resulted from the effects of chemotherapy. The alternative explanation is that the stage was incorrectly determined initially. If so, some of the patients in fact had stage I or II disease, rather than stage III disease, with a five-year survival rate after surgery on the order of 40 percent.1

The authors state that all the patients with N2 disease as determined by computed tomography underwent mediastinoscopy and biopsy but do not give the results of these procedures. What proportion of patients thought to have N2 disease at the time of randomization were found to have N2 disease by biopsy? As the results stand, the patients in the chemotherapy-plus-surgery group may have had a survival advantage from the outset of the study.

Nicholas Chanarin, M.R.C.P.
Southampton General Hospital, Southampton SO9 4XY, United Kingdom

1 References

1
Ihde DC, Minna JD. Non-small cell lung cancer. 1. Biology, diagnosis, and staging. Curr Probl Cancer 1991;15:61-104
Web of Science | Medline

To the Editor:

We are concerned that the survival advantage found by Rosell et al. in the chemotherapy-plus-surgery group may have resulted from an inadvertent bias in randomization. As they note, several groups have reported that K-ras mutations portend a poor prognosis in patients with non-small-cell lung cancer1,2. In their analysis of surgically resected specimens, Rosell et al. found K-ras mutations in the tumors of 10 of 24 patients (42 percent) in the surgery group but in only 3 of 20 patients (15 percent) in the chemotherapy-plus-surgery group. Thus, the improved survival in the latter group may have simply reflected the lower prevalence of tumors with the K-ras mutation.

This concern would be unwarranted if the frequency of K-ras mutations was similar in the two groups before therapy and the lower frequency of mutations in the resected tumors of the patients who had received chemotherapy resulted from the chemotherapy itself. If the cells with ras mutations are the fastest-growing population of cells in the tumor, they may be the most sensitive to chemotherapy and may thus be selectively eliminated by this treatment.

To determine whether the disproportionate distribution of tumors with ras mutations among the patients in the two groups resulted from the chemotherapy, the authors should have assessed the ras genotype before chemotherapy in the tumor tissue obtained initially by biopsy, as well as in the resected tumor tissue. The patients could then have been stratified according to ras genotype before randomization, or at least initial ras mutations could have been analyzed as an independent variable. This approach would also have clarified whether chemotherapy can alter the ras-mutation genotype of tumors.

Nancy E. Mills, M.D.
Charles L. Fishman, M.D.
Daniel R. Jacobson, M.D.
New York University Medical Center, New York, NY 10016

2 References

1
Slebos RJC, Kibbelaar RE, Dalesio O, et al. K-ras oncogene activation as a prognostic marker in adenocarcinoma of the lung. N Engl J Med 1990;323:561-565
Full Text | Web of Science | Medline

2
Rosell R, Li S, Skacel Z, et al. Prognostic impact of mutated K-ras gene in surgically resected non-small cell lung cancer patients. Oncogene 1993;8:2407-2412
Web of Science | Medline

To the Editor:

The results reported by Rosell et al. should be regarded with skepticism. Previous trials of combined therapy for lung cancer have reported, at most, modest improvements in survival as compared with surgery or radiation alone1,2. Rosell et al. defined their primary end point as the five-year survival rate, which they expected to be approximately 9 percent, but they failed to indicate the improvement they sought to detect. A study sample of at least 300 patients would be required to detect a doubling of this five-year survival probability. Their sample of 60 patients is sufficient only to detect an unprecedented level of benefit.

The results of this trial are premature. No patient has been followed for five years. The advantage in terms of disease-free survival in the chemotherapy-plus-surgery group appears to be impressive but represents an absolute difference of only five events. The difference in overall survival must be based on an even smaller number of events. Experience suggests that these differences will diminish with further follow-up3,4.

Finally, the authors inspected their results repeatedly: at least four analyses can be inferred (at 12, 18, and 24 months and at the time of the report). The failure to enroll patients during the last six months of the trial suggests widespread knowledge of the early analyses. Multiple analyses inflate the type I error, making the reported P values too small. No criterion for stopping the trial is noted by the authors. The decision to stop the trial early, based on knowledge of the accumulating results, further exaggerates both the treatment effect and the reported P value.

Sue-Anne McLachlan, M.B., B.S.
Martin Stockler, M.B., B.S.
Ontario Cancer Institute–Princess Margaret Hospital, Toronto, ON M4X 1K9, Canada

for the Medical Oncology Residents and Fellows

4 References

1
Dillman RO, Seagren SL, Propert KJ, et al. A randomized trial of induction chemotherapy plus high-dose radiation versus radiation alone in stage III non-small-cell lung cancer. N Engl J Med 1990;323:940-945
Full Text | Web of Science | Medline

2
Martini N, Kris MG, Flehinger BJ, et al. Preoperative chemotherapy to stage IIIa (N2) lung cancer: the Sloan-Kettering experience with 136 patients. Ann Thorac Surg 1993;55:1365-1374
CrossRef | Web of Science | Medline

3
Grilli R, Oxman AD, Julian JA. Chemotherapy for advanced non-small-cell lung cancer: how much benefit is enough? J Clin Oncol 1993;11:1866-1872
Web of Science | Medline

4
Stewart LA, Parmar MKB. Meta-analysis of the literature or of individual patient data: is there a difference? Lancet 1993;341:418-422
CrossRef | Web of Science | Medline

Author/Editor Response

Dr. Rosell replies:

To the Editor: The five-year survival for patients with stage IIIA, N2 disease is variable but tends to be less than 5 percent for patients treated with surgery plus radiotherapy1. These results do not differ substantially from those obtained with irradiation alone2. Dr. Cocquyt and colleagues question whether the way the disease-free-survival curves were constructed was appropriate. Failure-free survival was calculated from the date of randomization to the date of disease progression, relapse after a response, or death from any cause. The three-year survival in the chemotherapy-plus-surgery group is close to the 28 percent survival reported by the Sloan-Kettering group1 and agrees with the results of the M.D. Anderson randomized trial,3 in which the survival benefit was similar to that in our study. In both those trials the patients received two or three cycles of preoperative chemotherapy and two or three cycles of chemotherapy after surgery.

We stated in the Methods section that 25 patients (83 percent) in the chemotherapy-plus-surgery group had mediastinoscopy. In 8 of these 25 patients (32 percent) the tumor was assigned a lower stage at the time of thoracotomy, which means that the initially positive mediastinal node sites were found to be negative after chemotherapy. This rate of “downstaging” is similar to the 41 percent reported by Kim et al.4.

We found no differences in survival according to the K-ras status of the tumors, as stated in our paper. With respect to the difference in the frequency of K-ras mutations in the two groups, the alternative hypothesis that chemotherapy may eliminate K-ras-positive malignant cells should be examined in future studies.

We agree with McLachlan and Stockler that the probability of a type I error may become substantial if too few patients are analyzed. However, we found marked differences, and it was not ethically acceptable to continue to enroll patients. Because interim analyses may be misleading, interim results in our trial were reviewed by a multidisciplinary committee rather than by the study physicians. The differences in these results were significant (P<0.001, two-sided), which permitted early cessation of the trial.

Rafael Rosell, M.D.
University Hospital Germans Trias i Pujol, 08916 Badalona, Spain

4 References

1
Martini N, Kris MG, Flehinger BJ, et al. Preoperative chemotherapy for stage IIIa (N2) lung cancer: the Sloan-Kettering experience with 136 patients. Ann Thorac Surg 1993;55:1365-1374
CrossRef | Web of Science | Medline

2
Hazuka MB, Bunn PA Jr. Controversies in the nonsurgical treatment of stage III non-small cell lung cancer. Am Rev Respir Dis 1992;145:967-977
Web of Science | Medline

3
Roth JA. Surgical approaches to locally advanced, potentially resectable non-small cell lung cancer. Presented at IASLC Workshop: Controversies in staging and treatment of locally advanced non small cell lung cancers, Brussels, Belgium, August 1993. abstract.

4
Kim DH, Lynch TJ, Mentzer SJ, et al. Multimodality therapy of patients with stage IIIA, N2 non-small-cell lung cancer: impact of preoperative chemotherapy on resectability and downstaging. J Thorac Cardiovasc Surg 1993;106:696-702
Web of Science | Medline

Citing Articles (1)

Citing Articles

1
Alain Depierre, Virginie Westeel, Pascale Jacoulet. (2002) Gemcitabine induction chemotherapy in non–small cell lung cancer. Seminars in Oncology 29:3, 55-60
CrossRef