To the Editor:
Walsh et al. base a strong plea for the addition of combined chemotherapy and radiotherapy to the standard surgical treatment of adenocarcinoma of the esophagus (Aug. 15 issue) (1) on the results of a randomized trial of 113 patients. In Table 4 of their article, under the subhead "intention to treat," they excluded 24 of 55 patients (43.6 percent) from the surgery group and 27 of 58 patients (46.6 percent) from the multimodal-therapy group, reporting the results for 62 of the 113 randomized patients (55 percent). With the addition of just one more death to the multimodal-therapy group, the P value would have changed from 0.01 to 0.03, and with two deaths it would have changed to >0.05. Analysis of the 62 patients included would have a power of less than 70 percent to detect a 10 percent improvement in survival over the base-line value of 5 percent.
We have recently reported a meta-analysis of randomized trials and historical-control studies of cisplatin-based chemotherapy (2) in esophageal cancer. The meta-analysis of 12 randomized trials at two years (odds ratio, 0.96; 95 percent confidence interval, 0.75 to 1.22; P>0.1) and of 8 trials at three years (odds ratio, 0.74; 95 percent confidence interval, 0.53 to 1.02; P = 0.07) does not show that the addition of chemotherapy had any significant effect on survival. Updated results, which included two reports (Walsh et al. (1) and Pouliquen et al. (3)) published since the initial analysis, revealed an odds ratio of 0.72 at three years (95 percent confidence interval, 0.53 to 0.97; P = 0.05). This meta-analysis included 1130 patients and 868 deaths and had 95 percent power and confidence to detect a 10 percent benefit in survival over a base-line value of 20 percent. The meta-analysis included squamous carcinoma and adenocarcinoma, although most patients in the trials had squamous carcinoma, which is reported to be more sensitive to chemotherapy than adenocarcinoma. (4) Only two trials (5,6) in the meta-analysis showed benefit, reaching significance in one. (5) Both these trials had extremely poor survival in the control group (<5 percent), similar to that reported by Walsh et al., whereas the survival rates in the chemotherapy group were similar in all the trials (about 20 percent). We believe that the evidence from the published, randomized trials is too weak to affect clinical practice.
Rajan A. Badwe, M.S.
Tata Memorial Hospital
Parel, Bombay 400 012, India
Jayant S. Vaidya, M.S., D.N.B.
Royal Marsden Hospital
London SW3 6JJ, United Kingdom
Mehul S. Bhansali, M.S., D.N.B.
Tata Memorial Hospital
Parel, Bombay 400 012, India
References
1.
Walsh TN, Noonan N, Hollywood D, Kelly A, Keeling N, Hennessy TPJ. A comparison of multimodal therapy and surgery for esophageal adenocarcinoma. N Engl J Med 1996;335:462-7.
Return to Text
2.
Bhansali MS, Vaidya JS, Bhatt RG, Patil PK, Badwe RA, Desai PB. Chemotherapy for carcinoma of the esophagus: a comparison of evidence from meta-analysis of randomized trials and historical control studies. Ann Oncol 1996;7:355-9.
Return to Text
3.
Pouliquen X, Levard H, Hay JM, McGee K, Fingerhut A, Langlois-Zantin O. 5-Fluorouracil and cisplatin therapy after palliative surgical resection of squamous cell carcinoma of the esophagus: a multicenter randomized trial. Ann Surg 1996;223:127-33.
Return to Text
4.
Wolfe WG, Vaughn AL, Seigler HF, Hathorn JW, Leopold KA, Duhaylongsod FG. Survival of patients with carcinoma of the esophagus treated with combined-modality therapy. J Thorac Cardiovasc Surg 1993;105:749-55.
Return to Text
5.
Herskovic A, Martz K, Al-Sarraf M, et al. Combined chemotherapy and radiotherapy compared with radiotherapy alone in patients with cancer of the esophagus. N Engl J Med 1992;326:1593-8.
Return to Text
6.
Roth JA, Pass HI, Flanagan MM, Graeber GM, Rosenberg JC, Steinberg S. Randomized clinical trial of preoperative and postoperative adjuvant chemotherapy with cisplatin, vindesine, and bleomycin for carcinoma of the esophagus. J Thorac Cardiovasc Surg 1988;96:242-8.
Return to Text
Walsh et al. compared multimodal therapy (radiotherapy and chemotherapy plus radical surgery) with radical surgery alone in patients with adenocarcinoma of the esophagus. Thirty-two percent of the multimodal-therapy group survived for three years, as compared with 6 percent of the surgical group. More important, 25 percent of the multimodal-therapy group were tumor-free at the time of surgery.
The most important message of this study is that radical surgery should not be performed on patients who are tumor-free after radiotherapy and chemotherapy. One unfortunate patient who was tumor-free at surgery died after the operation. The esophagus should be reassessed by endoscopy and ultrasonography before surgery is undertaken. Indeed, the results make a strong case for abandoning radical surgery for esophageal cancer, in that only 6 percent of the surgical group survived for three years, a result that is counterbalanced by the 6 percent rate of operative mortality and, probably, a poor quality of life. As Wilke and Fink point out in their editorial, (1) "the overall prognosis after surgical resection alone is dismal."
Peter Gaffney, F.R.C.S.I.
Mallow General Hospital
Mallow, Ireland
References
1.
Wilke H, Fink U. Multimodal therapy for adenocarcinoma of the esophagus and esophagogastric junction. N Engl J Med 1996;335:509-10.
Return to Text
In the study by Walsh et al., the similarity of the two study groups is questionable because of insufficient data on preoperative tumor staging. If endosonography is not available as the staging diagnostic procedure of choice, at least computed tomography should be used. Computed tomography, however, was performed only in "selected patients."
The withdrawal of 10 patients from the multimodal-therapy group, as compared with 1 patient from the surgery group, represents a selection that favors the outcome of the multimodal-therapy group. In the evaluation of survival, the deterioration of the patient's condition during chemotherapy must be taken into account, no matter whether this was caused by combined chemotherapy and radiotherapy or emerged in the natural course of the disease. The withdrawal of patients with complications and rapid progression of disease in one group results in improved outcome in that group.
The increase in median survival in the multimodal-therapy group was five months. Surgery was performed two months after the beginning of multimodal therapy, and even later in some cases. The average delay between diagnosis and surgery is not stated (but it must have been at least two months). Thus, the average gain in survival after surgery in the multimodal-therapy group decreases to less than three months. The quality of life of a patient who is awaiting major surgery, with its multiple hospital stays and outpatient visits, nausea, and vomiting, must be considered inferior. If we are able to prolong survival, we also must consider the quality of life during that time. In our view, the study by Walsh et al. does not provide sufficient support to recommend neoadjuvant combined chemotherapy and radiotherapy for adenocarcinoma of the esophagus outside of clinical studies.
Eberhard M. Funk, M.D.
Jens Witte, M.D.
Zentralklinikum
86156 Augsburg, Germany
To the Editor:
Badwe et al. appear to have misunderstood the data presented in Table 4 of our article. The data on survival at three years specifically refer to patients who were enrolled for a minimum of three years. The reported Kaplan-Meier analysis refers to all randomized patients. The results unambiguously favor multimodal therapy.
Badwe et al. also suggest that the evidence in published randomized trials is too weak to affect clinical practice, but their meta-analysis is not relevant, since it concerns chemotherapy only. Furthermore, it is fundamentally flawed, since the results for adenocarcinoma and squamous carcinoma are pooled even though these tumors have dissimilar biologic behavior, different incidences of lymph-node involvement, different response rates after neoadjuvant therapy, and different prognoses for patients with complete responses. (1,2) In addition, their meta-analysis used historical controls -- a practice not recommended in view of the potential for bias.
Gaffney suggests that the results of our trial make a strong case for abandoning surgery for esophageal carcinoma. Ideally, resection should be reserved for patients with residual disease. In 33 percent of our patients, residual disease was confined to the esophageal wall and probably would have progressed without resection. Current staging techniques, including endoscopy and ultrasonography, are too unreliable after chemotherapy and radiotherapy to be used to select patients for surgery. (3)
Funk and Witte suggest that our staging methods created the potential for the unequal distribution of tumors between the groups. Optimal staging would require laparoscopy in addition to endosonography (which has at best a sensitivity of 85 percent), but this could have had a negative effect on recruitment. It is unlikely that the difference in the incidence of positive lymph nodes (42 percent in the multimodal-therapy group and 82 percent in the surgery-only group) could have occurred by chance after randomization. Concern about the effect of the withdrawal of patients on outcome is unfounded, since we analyzed and reported the results on the basis of the intention to treat and treatment actually received.
Funk and Witte assume that the quality of life of our patients was markedly diminished during treatment. Although we did not measure quality of life, the introduction of 5-hydroxytryptamine3 blockers for nausea, colony-stimulating factor for myelosuppression, and outpatient radiotherapy allowed patients to have a nearly normal lifestyle during treatment. From the patient's perspective, survival from the date of diagnosis is most relevant. The real benefit, however, is seen in the 33 percent of patients who can hope to live full and productive lives after multimodal therapy, a hope denied to all but 7 percent of patients treated by surgery alone.
Thomas N. Walsh, M.D.
Beaumont Hospital
Dublin 9, Ireland
Alan Kelly, Ph.D.
Trinity College
Dublin 2, Ireland
Thomas P.J. Hennessy, M.D.
St. James Hospital
Dublin 8, Ireland
References
1.
Wolfe WG, Vaughn AL, Seigler HF, Hathorn JW, Leopold KA, Duhaylongsod FG. Survival of patients with carcinoma of the esophagus treated with combined-modality therapy. J Thorac Cardiovasc Surg 1993;105:749-55.
Return to Text
2.
Walsh TN, Noonan N, Hollywood D, Kelly A, Keeling N, Hennessy TPJ. Multimodality treatment plus surgery versus surgery alone for oesophageal squamous cell carcinoma. In: Peracchia A, Rosati R, Bonavina L, Fumagalli U, Bona S, Chella B, eds. Recent advances in diseases of the esophagus. Milan, Italy: Monduzzi Editore, 1995:451-6.
Return to Text
3.
Esophageal carcinoma. In: Rosch T, Classen M. Gastroenterologic endosonography. Stuttgart, Germany: Georg Thiem Verlag, 1992:45-62.
Return to Text