Meta-Analysis of Postoperative Adjuvant Chemotherapy With Tegafur-Uracil in Non–Small-Cell Lung Cancer
http://www.100md.com
《临床肿瘤学》
Faculty of Engineering, Tokyo University of Science, Tokyo
The West Japan Study Group for Lung Cancer Surgery, Kyoto
The Japan Lung Cancer Research Group, Fukuoka
Northeast Japan Study Group for Lung Cancer Surgery, Sendai
Osaka Lung Cancer Study Group, Osaka
Adjuvant Chemotherapy for Lung Cancer Research Group, Nagoya
ABSTRACT
PURPOSE: Recent clinical trials have shown the efficacy of platinum-based adjuvant chemotherapy for completely resected non–small-cell lung cancer (NSCLC). In Japan, many clinical trials of adjuvant chemotherapy with tegafur-uracil (UFT) have been conducted, and some trials showed positive results while others showed negative results. Thus, we performed a meta-analysis to assess the efficacy of postoperative adjuvant chemotherapy with UFT in NSCLC.
METHODS: Among nine trials of postoperative adjuvant UFT-containing chemotherapy, six trials comparing surgery alone with surgery plus UFT were identified. Of six trials, two were three-arm trials including cisplatin-based chemotherapy followed by UFT, and data from that arm were not included in the meta-analysis.
RESULTS: Of 2,003 eligible patients, most (98.8%) had squamous cell carcinoma or adenocarcinoma, and most had stage I disease; the tumor classification was T1 in 1,308 (65.3%), T2 in 674 (33.6%), and the nodal status was N0 in 1,923 (96.0%). The two treatment groups did not differ significantly in major prognostic factors. The median duration of follow-up was 6.44 years. The survival rates at 5 and 7 years were significantly higher in the surgery plus UFT group (81.5% and 76.5%, respectively) than in the surgery alone group (77.2% and 69.5%, respectively; P = .011 and .001, respectively). The overall pooled hazard ratio was 0.74, and its 95% CI was 0.61 to 0.88 (P = .001).
CONCLUSION: This meta-analysis showed that postoperative adjuvant chemotherapy with UFT was associated with improved 5- and 7-year survival in a Japanese patient population composed primarily of stage I adenocarcinoma patients.
INTRODUCTION
Non–small-cell lung cancer (NSCLC) accounts for 75% to 85% of primary lung cancer, which is the leading cause of cancer deaths in most industrialized countries. Surgery is the most effective therapeutic modality for cure, but the postoperative survival of patients undergoing thoracotomy for NSCLC remains unsatisfactory, with 5-year survival rates less than 80% even in pathologic (p-) stage I disease.1,2 Thus, effective adjuvant therapy to improve postoperative survival (ie, 5-year overall survival rates exceeding 80% in early stage NSCLC) should be established.
Many clinical trials of postoperative adjuvant therapy have clearly indicated that postoperative radiation does not improve survival even when it can reduce the incidence of local recurrence.3 In addition, the efficacy of postoperative chemotherapy had not been demonstrated in most trials, such as the Adjuvant Lung Project Italy,4 whereas a meta-analysis conducted in 1995 showed that cisplatin-based chemotherapy might improve survival.5 Recently, however, a large-scale trial conducted by the International Adjuvant Lung Cancer Trial (IALT) group showing that cisplatin-based postoperative adjuvant chemotherapy can improve the survival of completely resected p-stage I-IIIA NSCLC patients (5-year overall survival rates, 44.5% for the chemotherapy group and 40.4%, respectively; P < .03) has been reported.6 More recently, two additional trials more clearly showing the efficacy of postoperative chemotherapy for completely resected NSCLC have been reported7,8; the JBR.10 trial conducted by the National Cancer Institute of Canada (NCIC) has shown that cisplatin in combination with vinorelbine can improve the survival of p-stage IB-II patients (5-year survival rates, 69% for chemotherapy group and 54% for surgery-alone group, respectively; hazard ratio [HR], 0.69; 95% CI, 0.52 to 0.92; P = .011),7 and the Cancer and Leukemia Group B (CALGB) 9633 trial has shown that carboplatin in combination with paclitaxel can improve the survival of p-stage IB patients (4-year survival rates, 71% and 59%, respectively; HR, 0.62; 95% CI, 0.41 to 0.95; P = .028).8 These results have clearly demonstrated the efficacy of platinum-based chemotherapy following complete resection for NSCLC.
In Japan, many randomized clinical trials of adjuvant chemotherapy with tegafur-uracil (UFT) in completely resected NSCLC have been conducted. UFT is an oral fluorouracil (FU) derivative chemotherapeutic drug composed of tegafur and uracil; tegafur is a prodrug of FU, and uracil is added to inhibit degradation of FU through inhibition of dihydropyrimidine dehydrogenase, which is the enzyme responsible for FU catabolism. Thus, oral UFT administration can achieve a certain plasma FU concentration for a long time. Direct antitumor effects of UFT are minimal in most malignant tumors, including NSCLC,9 but UFT is widely used in Japan in postoperative adjuvant therapy because of its mild toxicity characteristics. In NSCLC, some trials, including the West Japan Study Group for Lung Cancer Surgery (WJSG) second trial10 and the Japan Lung Cancer Research Group (JLCRG) trial,11 have shown that postoperative UFT administration can improve the survival of completely resected NSCLC patients, but others failed to show the efficacy. Since meta-analysis of randomized controlled trials can most accurately evaluate a therapy,5,12-14 we conducted a meta-analysis to gain insight into the efficacy of postoperative UFT administration for completely resected NSCLC.
METHODS
Inclusion Criteria of Trials
Trials were eligible for inclusion in the present meta-analysis if they were randomized controlled trials of patients with completely resected NSCLC. Patients in the control group had to receive surgery alone, and patients in the treated group had to receive UFT alone following surgery; trials comparing surgery alone with surgery plus intravenous chemotherapy and UFT were not included. Finally, trials were eligible only if the postoperative follow-up duration was at least 5 years.
Protocol for the Meta-Analysis
In August 2002, a protocol for the meta-analysis, describing its rationale, statistical methodology, and rules for publication was produced in Tokyo University of Science, and distributed to the principal investigators of the six trials. Investigators were asked to provide individual data for every randomly assigned patient, whether eligible or not, assessable or not, and properly followed up or not. Information items requested for every patient were as follows: patient identification, date of surgery, eligibility, treatment allocated by random assignment, age, sex, histological type, pathological tumor size, pathological lymph node metastases, date of death, or last visit. Survival was calculated from the date of surgery to the date of death, regardless of the cause of death. Toxicity data were not collected, because detailed analysis of adverse reactions can be found in the respective published reports of individual trials.10,11,15
Identification of Trials
Trials were identified by searches of MEDLINE (1966 onward). Experts in lung cancer were also asked to help to identify trials. As a result, a total of nine prospective randomized adjuvant trials of UFT-containing chemotherapy, all conducted in Japan, were identified.10,11,15-21 Among them, one trial (the Hirosaki University trial) was excluded because a surgery-alone arm was not employed.16 Two trials (the Chubu-Japan trial and the WJSG third trial) were excluded because the chemotherapy arm was not UFT alone but cisplatin-based chemotherapy followed by UFT17,18; cisplatin and doxorubicin followed by UFT was employed as adjuvant chemotherapy in the Chubu-Japan trial,17 and cisplatin, vindesine, and mitomycin followed by UFT was employed in the WJSG third trial.18
Finally, a total of six trials conducted by five different groups that met the inclusion criteria of the present meta-analysis were identified as follows (Table 1 and Fig 1): (1) The WJSG second trial, Wada et al10; (2) the WJSG fourth trial, Nakagawa et al19; (3) The Northeast Japan Study Group for Lung Cancer Surgery trial, Endo et al15; (4) The Osaka Lung Cancer Study Group trial, Tada et al20; (5) Adjuvant Chemotherapy for Lung Cancer Study Group (ACTLC) trial, Imaizumi21; and (6) The JLCRG trial, Kato et al.11 No trial was excluded due to inadequate follow-up. Among the six trials included in the present meta-analysis, the WJSG-second trial and the ACTLC trial were three-arm trials consisting of a surgery-alone arm, surgery-plus-UFT arm, and surgery plus cisplatin-based chemotherapy followed by UFT arm10,21; patient data in the third arm were not included in the present meta-analysis.
Data
Detailed trial protocols and data on individual patients of each trial were provided by the principal investigators of the trials. Eligibility criteria for inclusion common to all six trials included in the meta-analysis were as follows: (1) histologically confirmed NSCLC; (2) p-stage I to III disease; (3) complete resection achieved; (4) age less than 76 years; (5) no previous anticancer treatment (radiotherapy, chemotherapy, endocrine therapy, or immunotherapy); (6) no multiple cancer; (7) written or oral informed consent. In all six studies, confirmation of patients' eligibility and random assignment were performed by telephone or facsimile at a registration center.
Statistical Analysis
Two major approaches for meta-analysis are the fixed effect model and the random effect model. We performed analyses based on both approaches. However, since heterogeneity among the studies was not observed, these two models showed almost identical results. Therefore, we only describe the former statistical model here and do not show the result based on the random effect model.
The following analysis based on individual patient data, was performed using the Statistical Analysis System (SAS version 8.2), mainly the PHREG and LIFETEST procedures. In the present meta-analysis, the patients who were eligible in each trial were primarily analyzed in order to ensure consistency with the results of individual trials, but all randomized patients, that is, according to intent-to-treatment (ITT) principle, were also analyzed to confirm the robustness of the results. The primary end point of the present study was overall survival, defined as the interval from operation to death from any cause. Data on patients who were alive were censored on the date on which they were last known to be alive.
Comparability Between Two Groups
Comparability for major demographic factors (age, sex) and important prognostic factors (histologic type, pathologic T and N stage) between the surgery alone and the surgery plus UFT groups in all six trials was evaluated. The distributions of these factors were described using the contingency table, and an adequate statistical test, depending on the distribution form, was performed to detect differences of distribution between the two groups.
Methods for Meta-Analysis
Employing the following Cox's proportional hazard model (a), the hazard ratio between the surgery alone and surgery plus UFT group for each trial was calculated. Suffix k stands for each trial. h(t) and hk(t) show hazard function and the baseline hazard function in the k trial, respectively. Variable UFT is 0, 1 dummy variable, depending on whether a patient belonged to the surgery alone or surgery plus UFT group. ?k corresponds to the log hazard ratio of each trial. Heterogeneity for the effect of UFT among six trials was tested assuming the null hypothesis ?1 = ?2 = ?3 = ?4 = ?5 = ?6 (df = 5). Overall pooled effect was estimated based on the fixed effect model (b) and tested, setting the null hypothesis ? = 0 (df = 1). This test is essentially identical to a trial-stratified log-rank test.
Cumulative survival rates in all six trials were estimated according to the Kaplan-Meier method. Adjusted analysis was performed adding important prognostic factors to Cox's proportional hazard model (b). Subset analysis and test for interaction between UFT and other factors were also conducted, according to the major demographic variables of age (69 years or younger v 70 years or older), sex, histological type, and p-T stage (pT1 v pT2). All statistical tests were two-sided. P < .05 was considered to indicate statistically significant differences.
RESULTS
Trials Included and Numbers of Patients
Between December 1985 and March 1997, a total of 2,082 patients (1,046 patients in the surgery-alone group and 1,036 patients in the surgery-plus-UFT group) were enrolled in the six studies, of whom 1,002 patients (95.8%) in the surgery alone group and 1,001 patients (96.6%) in the surgery plus UFT group were eligible and were analyzed. It should be noted that one trial (the JLCRG trial) contributed almost one half of the patients included in the present meta-analysis and was restricted to stage I adenocarcinoma patients.
Clinical Characteristics of Patients
The most common histologic type was adenocarcinoma (1,679 patients, 83.8%), followed by squamous cell carcinoma (299 patients, 14.9%). Most patients had p-stage I disease: 1,308 patients (65.3%) had p-T1 disease, and 674 (33.6%) had p-T2 disease, and 1,923 (96.0%) had no nodal involvement (N0). There was no major difference between the surgery-alone group, and the surgery-plus-UFT group in any major prognostic factor including age, sex, histologic type, pT factor, and pN factor (Table 2).
Postoperative Survival
The median follow-up duration for all eligible patients was 6.44 years after operation. The 5-year overall survival rate in the surgery plus UFT group was 81.8%, which was 4.6% higher than that in the surgery-alone group (5-year survival rate, 77.2%). At 7 years, overall survival rates were 69.5% and 76.5% in the surgery-alone group and in the surgery plus UFT group, respectively, showing more significant survival benefit in the surgery plus UFT group (Fig 2). The overall pooled hazard ratio was 0.74, and its 95% CI was 0.61 to 0.88 (P = .001). There was no heterogeneity in survival among the six studies (P = .79; Fig 3, O/N, observed death/number of patients; O – E, observed minus expected number of deaths; V, variance of O – E). Even when survival analysis was performed in all randomly assigned patients, including ineligible patients; that is, ITT analysis, patients in the surgery plus UFT group also showed a significantly better survival than those in the surgery alone group (hazard ratio, 0.76, 95% CI, 0.64 to 0.90; P = .005).
Subset Analysis
Overall survival was favorable in the surgery plus UFT group, regardless of age (HR for 69 years or younger, 0.76, 95% CI, 0.61 to 0.94; HR for 70 years or older, 0.68, 95% CI, 0.49 to 0.94), sex (HR for male, 0.77, 95% CI, 0.62 to 0.96; HR for female, 0.67, 95% CI, 0.48 to 0.92), histologic type (HR for adenocarcinoma, 0.69, 95% CI, 0.56 to 0.85; HR for squamous cell carcinoma, 0.82; 95% CI, 0.57 to 1.19), and pT-factor (HR for T1, 0.73; 95% CI, 0.56 to 0.93; HR for T2, 0.78; 95% CI, 0.60 to 1.01). Among all subsets, the efficacy of postoperative UFT administration was similar (Fig 4, O/N, observed deaths/number of patients; O – E, observed minus expected number of deaths; V, variance of O – E).
Multivariate Analysis
Multivariate analysis was applied using Cox's proportional hazard model with six factors, including age ( 69 years, 70 years), sex, histologic type (adenocarcinoma, squamous cell plus others), pT-factor (T0 plus T1, T2-4), pN-factor (N0, N1-3), and treatment (surgery alone, surgery plus UFT), as covariates. The results confirmed the benefit of treatment after adjustment for these covariates (HR for surgery plus UFT, 0.75; 95% CI, 0.63 to 0.90; P = .002).
DISCUSSION
Before the 2003 Annual Meeting of the American Society of Clinical Oncology (ASCO), no large-scale randomized controlled trial in which more than 1,000 patients had been randomized had shown the efficacy of postoperative adjuvant chemotherapy for NSCLC. According to a previous meta-analysis conducted by the Non–Small-Cell Lung Cancer Group,5 long-term administration of alkylating agents significantly increased the risk of postoperative death (P = .005). In contrast, cisplatin-based chemotherapy as well as treatment with other drugs, mainly tegafur and UFT, slightly reduced the risk of postoperative death (P = .08 and HR, 0.87; 95% CI, 0.74 to 1.02, for cisplatin-based chemotherapy; P = .30 and HR, 0.89; 95% CI, 0.72 to 1.11, for other drugs including UFT). As compared with meta-analyses conducted by the Early Breast Cancer Trialists' Collaborative Group clearly demonstrating the efficacy of tamoxifen and polychemotherapy in the treatment of breast cancer,12,13 the number of patients included in the Non–Small-Cell Lung Cancer Group meta-analysis was too small to draw definite conclusions (alkylating agents, n = 2,145; cisplatin-based chemotherapy, n = 1,394; other drugs, n = 918). At the 2003 ASCO meeting, results of two large-scale randomized controlled trials of postoperative adjuvant chemotherapy, the IALT trial showing the efficacy of cisplatin-based chemotherapy6 and the JLCRG trial showing the efficacy of UFT,11 were reported. At the 2004 ASCO meeting, results of two additional randomized adjuvant trials showing the efficacy of platinum-based chemotherapy, the NCIC-JBR.10 trial (cisplatin and vinorelbine)7 and the CALGB-9633 trial (carboplatin and paclitaxel),8 were reported.
UFT does not show potent direct antitumor effects for advanced malignant tumors, as the response rate in advanced and recurrent NSCLC has been estimated to be only 7%.9 However, in the postoperative adjuvant setting, UFT was clinically effective in rectal cancer22,23 and breast cancer,24,25 which was shown in both randomized controlled studies and meta-analyses. In NSCLC, Wada et al were the first to report that long-term oral administration of UFT improved the postoperative survival of completely resected NSCLC.10 Considering the time-dependent antitumor effects of FU, long-term oral UFT administration may be the optimal schedule of FU administration. In the postoperative setting, residual lesions requiring control were micrometastases, not bulky tumors. Even if the direct antitumor effects of UFT are relatively low, it can be administered for a long time because of its mild toxicity, which results in controlling micrometastases and improving survival. This concept is supported by the fact that in vivo experiments in mouse metastasis models of breast cancer26 and lung cancer27 have shown that UFT suppresses development of micro-metastases. Recently, two meta-analyses assessing the effectiveness of postoperative adjuvant chemotherapy with UFT have been published.28,29 However, these meta-analyses were based on summary data shown in the article. Our meta-analysis based completely on individual patient data, is considered to be more statistically reliable and accurate, as well as more useful, because more detailed analysis as subset analysis, multivariate analysis, and ITT analysis can be performed.30
The present meta-analysis showed that UFT administration for 1 to 2 years after surgery improved survival of NSCLC patients, mainly p-stage I (T1-2, N0) patients. The efficacy of UFT administration was relatively greater on analysis of survival data at 7 years than that at 5 years. This finding suggests that long-term observation is required to demonstrate the efficacy of postoperative adjuvant therapy in early-stage patients, who have a lower risk of postoperative death. We cannot identify the causes of death in all patients included in the present meta-analysis because of the unavailability of data. For around half of the patients in the meta-analysis, all of whom participated in the JLCRG trial, treatment failure patterns in the control group and UFT group have been reported as follows: (1) intrathoracic recurrence, 11.9% and 11.1%, respectively; (2) extrathoracic recurrence, 8.6% and 6.1%; (3) second cancer, 4.1% and 5.1%.11 From these results documented in the JLCRG study, it may be speculated that UFT can improve postoperative survival through inhibiting extrathoracic recurrence.
There was no separation of the survival curves until after the 4-year mark in the meta-analysis (Fig 2). In contrast, the benefit from adjuvant chemotherapy in other studies such as the NCIC-JBR.10 trial and the CALGB-9633 trial has been seen considerably sooner, usually within the first year or so. We cannot explain the exact reasons why such a unique survival curve was obtained in the present UFT meta-analysis, but one reason may be that the majority of patients included in the present meta-analysis were p-stage I patients, especially, p-stage IA patients. In such very early-stage patients, postoperative recurrence and death may be later, which results in overlapping of the survival curves of the surgery-alone arm and the chemotherapy arm. In fact, the survival curves were separate within 1 year after surgery, and the efficacy of UFT appeared sooner in the WJSG-second trial where p-stage I-III patients were eligible.10
Although a meta-analysis is, generally speaking, the most reliable and effective tool to assess the efficacy of a therapy, attention should be paid to one issue in the present meta-analysis in that one large-scale trial (the JLCRG trial), where only p-stage I adenocarcinoma patients were eligible, contributed almost one half of the patients included in the present meta-analysis. As a result, more than 80% patients were adenocarcinoma patients, and most patients had p-stage I disease among whom two-thirds had p-stage IA disease. Thus, the results of the present meta-analysis cannot easily be compared with any other meta-analysis or other trials of adjuvant chemotherapy.
Subset analyses in the present study revealed no significant interaction between the efficacy of postoperative UFT administration and age, sex, histological type, or p-T factor. However, as p-stage I adenocarcinoma composed the vast majority of the patient populations included in the present meta-analysis, the present meta-analysis may suggest that postoperative UFT administration is effective especially in p-stage I adenocarcinoma of patients. Unfortunately, the present meta-analysis failed to show whether or not UFT was effective in advanced-stage diseases and/or squamous cell carcinoma, probably due to insufficient numbers of patients in these subsets.
In the IALT trial, where 36% of patients were stage I, 25% were stage II, and 39% were stage III, cisplatin-based chemotherapy was effective for p-stage I-III patients (HR, 0.86). In the NCIC-JBR.10 trial, where 45% of patients were stage IB and 55% were stage II, cisplatin in combination with vinorelbine was effective for p-stage IB-II patients (HR, 0.62). The CALGB trial, where all patients were stage IB, showed that carboplatin in combination with paclitaxel was effective for p-stage IB patients (HR, 0.71). These results show that platinum-based postoperative adjuvant chemotherapy may be active for a wide range of patient populations, including p-stage II-III patients. In contrast, the present meta-analysis showed that UFT was effective in a patient population composed primarily of stage I adenocarcinoma patients. In addition, the HR documented in the present meta-analysis for UFT administration (HR, 0.74) was inferior to those documented in the NCIC-JBR.10 trial and CALGB-9633 trial. However, these results indicate that UFT may be recommended for early-stage patients, and platinum-based chemotherapy should be employed for advanced-stage (ie, stage II-III) patients.
Anticancer therapy should be performed based on an individual assessment of the risk of recurrence and death, the balance between the toxicity and efficacy, and changes in quality of life caused by the therapy. According to a cross-over study on patients' choice of therapy (injection or oral therapy) showing the same efficacy, 85% of patients preferred oral therapy because it was more convenient.31 UFT is an oral drug that can be given for more than 1 year, and is preferable in the postoperative setting not only because of the mild toxicity and long-term effects but also because of the patients' convenience.
In conclusion, our meta-analysis shows that long-term treatment with UFT for 1 to 2 years was effective as postoperative adjuvant therapy in a Japanese patient population composed primarily of stage I adenocarcinoma patients. To assess the efficacy of UFT outside Japan, additional prospective randomized studies of adjuvant UFT in completely resected NSCLC should be conducted in North America and Europe.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
Acknowledgment
The authors are indebted to J. Patrick Barron of the International Medical Communications Center of Tokyo Medical University for his review of this article.
NOTES
Supported by Taiho Pharmaceutical Co, Tokyo, Japan (the six trials, including the meta-analysis). This company had no role in the meta-analysis study design, data collection, data analysis, data interpretation, or in writing the report.
Presented at the 12th European Cancer Conference meeting, Copenhagen, Denmark, 2003 and the 40th Annual Meeting of the American Society of Clinical Oncology, New Orleans, LA, June 5-8, 2004.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Mountain CF: Revisions in the international system for staging lung cancer. Chest 116:1710-1717, 1997
, Naruke T, Tsuchiya R, Kondo H, et al: Implication of staging in lung cancer. Chest 112:242s-248s, 1997
PORT Meta-analysis Trialists Group: Postoperative radiotherapy in non-small-cell lung cancer: Systematic review and meta-analysis of individual patient data from nine randomised controlled trials. Lancet 352:257-263, 1998
Scagliotti GV, Fossati R, Torri V, et al: Randomized study of adjuvant chemotherapy for completely resected stage I, II, or IIIA non-small-cell lung cancer. J Nat Cancer Inst 95:1453-1461, 2003
Non-small Cell Lung Cancer Collaborative Group: A meta-analysis using updated data on individual patients from 52 randomised clinical trials. BMJ 311:899-909, 1995
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 350:351-360, 2004
Winton TL, Livingston R, Johnson D, et al: A prospective randomized trial of adjuvant vinorelbine (VIN) and cisplatin (CIS) in completely resected stage IB and II non small cell lung cancer (NSCLC) Intergroup JBR 10. J Clin Oncol 22:621s, 2004 (abstr 7018)
Strauss GM, Maddaus MA, Johnstone DW, et al: Randomized clinical trial of adjuvant chemotherapy with paclitaxel and carboplatin following resection in stage IB non-small cell lung cancer (NSCLC): Report of Cancer and Leukemia Group B (CALGB) protocol 9633. J Clin Oncol 22:621s, 2004 (abstr 7019)
Taguchi T: Clinical application of biochemical modulation in cancer chemotherapy: Biochemical modulation for 5-FU. Oncology 54:12-18, 1997 (suppl 1)
Wada H, Hitomi S, Teramatsu T, et al: Adjuvant chemotherapy after complete resection in non-small-cell lung cancer. J Clin Oncol 14:1048-1054, 1996
Kato H, Ichinose Y, Ohta M, et al: A randomized trail of adjuvant chemotherapy with uracil-tegafur for adenocarcinoma of the lung. N Engl J Med 350:1713-1721, 2004
Early Breast Cancer Trialists' Collaborative Group: Tamoxifen for early breast cancer: An overview of the randomised trails. Lancet 351:1451-1467, 1998
Early Breast Cancer Trialists' Collaborative Group: Polychemotherapy for early breast cancer: An overview of the randomised trails. Lancet 352:930-942, 1998
Colorectal Cancer Collaborative Group: Adjuvant radiotherapy for rectal cancer: A systemic overview of 8,507 patients from 22 randomised trials. Lancet 358:1291-1304, 2001
Endo C, Saito Y, Iwanami H, et al: A randomized trial of postoperative UFT therapy in p stage I, II non-small cell lung cancer: North-east Japan study group for lung cancer surgery. Lung Cancer 40:181-186, 2003
Fukushima M, Tsushima T, Ito H, et al: A randomized controlled study of post-operative adjuvant therapy in non-small cell lung cancer. Jpn J Cancer Chemother 23:303-309, 1996
The Study Group of Adjuvant Chemotherapy for Lung Cancer (Chubu, Japan): A randomized trial of postoperative adjuvant chemotherapy in non-small cell lung cancer (the second cooperative study). Eur J Surg Oncol 21:69-77, 1995
Wada H, Miyahara R, Tanaka F, et al: Postoperative adjuvant chemotherapy with PVM (Cisplatin+Vindesine+Mitomycin C) and UFT (Uracil+Tegaful) a randomized clinical trial: West Japan Study Group for lung cancer surgery (WJSG). Eur J Cardiothorac Surg 15:438-443, 1999
Nakagawa M, Tanaka F, Tsubota N, et al: A randomized phase III trial of adjuvant chemotherapy with UFT for completely resected pathological stage I non-small-cell lung cancer: The west Japan study group for lung cancer surgery (WJSG), the 4th study. Ann Oncol 16:75-80, 2005
Tada H, Yasumitsu T, Iuchi K, et al: Randomized study of adjuvant chemotherapy for completely resected non-small cell lung cancer. Proc J Clin Oncol 21:313a, 2002 (abstr 1250)
Imaizumi M: A randomized trial of postoperative adjuvant chemotherapy for p-stage I non-small cell lung cancer (4th cooperative study). Lung Cancer 41:S2, 2003 (abstr O180)
Kato T, Ohashi Y, Nakazato H, et al: Efficacy of oral UFT as adjuvant chemotherapy to curative resection of colorectal cancer: Multicenter prospective randomized trial. Langenecks Arch Surg 386:575-581, 2001
Sakamoto J, Hamada C, Yasutomi M, et al: Efficacy of oral UFT as adjuvant chemotherapy after resection of rectal Cancer: Meta-analysis of 5-year results from four randomized trials in 1815 patients. Proc Am Soc Clin Oncol 22:270, 2003 (abstr 1082)
Ogita M, Uchino J, Asaishi K, et al: Efficacy of UFT plus tamoxifen for estrogen-receptor-positive breast cancer and tamoxifen plus UFT for estrogen-receptor-negative breast cancer. Clin Drug Invest 23:689-699, 2003
Kasumi F, Yoshimoto M, Uchino J, et al: Meta-analysis of five studies on tegafur plus uracil (UFT) as post-operative adjuvant chemotherapy for breast cancer. Oncology 64:146-153, 2003
Kurebayashi J, Nakatsuka M, Fujioka A, et al: Postsurgical oral administration of uracil and tegafur inhibits progression of micrometastasis of human breast cancer cells in nude mice. Clin Cancer Res 3:653-659, 1997
Miyahara R, Nakagawa T, Ishikawa S, et al: UFT inhibits lung metastases in spontaneous metastasis model of lung cancer. Thorac Cardiovasc Surg 53:118-121, 2005
Sedrakyan A, Der Meulen JV, O'Byrne K, et al: Postoperative chemotherapy for non-small cell lung cancer: A systematic review and meta-analysis. Gen Thorac Surg 128:414-419, 2004
Hotta K, Matsuo K, Ueoka H, et al: Role of adjuvant chemotherapy in patients with resected non-small-cell lung cancer: Reappraisal with a meta-analysis of randomized controlled trials. J Clin Oncol 22:3860-3867, 2004
Piedbois P, Mondor H, Buyse M: Meta-analyses based on abstracted data: A step in the right direction, but only s first step. J Clin Oncol 22:3839-3841, 2004
Borner MM, Schoffski P, de Wit R, et al: Patient preference and pharmacokinetics of oral modulated UFT versus intravenous fluorouracil and leucovorin: A randomised crossover trial inadvanced colorectal cancer. Eur J Cancer 38:349-358, 2002(Chikuma Hamada, Fumihiro )
The West Japan Study Group for Lung Cancer Surgery, Kyoto
The Japan Lung Cancer Research Group, Fukuoka
Northeast Japan Study Group for Lung Cancer Surgery, Sendai
Osaka Lung Cancer Study Group, Osaka
Adjuvant Chemotherapy for Lung Cancer Research Group, Nagoya
ABSTRACT
PURPOSE: Recent clinical trials have shown the efficacy of platinum-based adjuvant chemotherapy for completely resected non–small-cell lung cancer (NSCLC). In Japan, many clinical trials of adjuvant chemotherapy with tegafur-uracil (UFT) have been conducted, and some trials showed positive results while others showed negative results. Thus, we performed a meta-analysis to assess the efficacy of postoperative adjuvant chemotherapy with UFT in NSCLC.
METHODS: Among nine trials of postoperative adjuvant UFT-containing chemotherapy, six trials comparing surgery alone with surgery plus UFT were identified. Of six trials, two were three-arm trials including cisplatin-based chemotherapy followed by UFT, and data from that arm were not included in the meta-analysis.
RESULTS: Of 2,003 eligible patients, most (98.8%) had squamous cell carcinoma or adenocarcinoma, and most had stage I disease; the tumor classification was T1 in 1,308 (65.3%), T2 in 674 (33.6%), and the nodal status was N0 in 1,923 (96.0%). The two treatment groups did not differ significantly in major prognostic factors. The median duration of follow-up was 6.44 years. The survival rates at 5 and 7 years were significantly higher in the surgery plus UFT group (81.5% and 76.5%, respectively) than in the surgery alone group (77.2% and 69.5%, respectively; P = .011 and .001, respectively). The overall pooled hazard ratio was 0.74, and its 95% CI was 0.61 to 0.88 (P = .001).
CONCLUSION: This meta-analysis showed that postoperative adjuvant chemotherapy with UFT was associated with improved 5- and 7-year survival in a Japanese patient population composed primarily of stage I adenocarcinoma patients.
INTRODUCTION
Non–small-cell lung cancer (NSCLC) accounts for 75% to 85% of primary lung cancer, which is the leading cause of cancer deaths in most industrialized countries. Surgery is the most effective therapeutic modality for cure, but the postoperative survival of patients undergoing thoracotomy for NSCLC remains unsatisfactory, with 5-year survival rates less than 80% even in pathologic (p-) stage I disease.1,2 Thus, effective adjuvant therapy to improve postoperative survival (ie, 5-year overall survival rates exceeding 80% in early stage NSCLC) should be established.
Many clinical trials of postoperative adjuvant therapy have clearly indicated that postoperative radiation does not improve survival even when it can reduce the incidence of local recurrence.3 In addition, the efficacy of postoperative chemotherapy had not been demonstrated in most trials, such as the Adjuvant Lung Project Italy,4 whereas a meta-analysis conducted in 1995 showed that cisplatin-based chemotherapy might improve survival.5 Recently, however, a large-scale trial conducted by the International Adjuvant Lung Cancer Trial (IALT) group showing that cisplatin-based postoperative adjuvant chemotherapy can improve the survival of completely resected p-stage I-IIIA NSCLC patients (5-year overall survival rates, 44.5% for the chemotherapy group and 40.4%, respectively; P < .03) has been reported.6 More recently, two additional trials more clearly showing the efficacy of postoperative chemotherapy for completely resected NSCLC have been reported7,8; the JBR.10 trial conducted by the National Cancer Institute of Canada (NCIC) has shown that cisplatin in combination with vinorelbine can improve the survival of p-stage IB-II patients (5-year survival rates, 69% for chemotherapy group and 54% for surgery-alone group, respectively; hazard ratio [HR], 0.69; 95% CI, 0.52 to 0.92; P = .011),7 and the Cancer and Leukemia Group B (CALGB) 9633 trial has shown that carboplatin in combination with paclitaxel can improve the survival of p-stage IB patients (4-year survival rates, 71% and 59%, respectively; HR, 0.62; 95% CI, 0.41 to 0.95; P = .028).8 These results have clearly demonstrated the efficacy of platinum-based chemotherapy following complete resection for NSCLC.
In Japan, many randomized clinical trials of adjuvant chemotherapy with tegafur-uracil (UFT) in completely resected NSCLC have been conducted. UFT is an oral fluorouracil (FU) derivative chemotherapeutic drug composed of tegafur and uracil; tegafur is a prodrug of FU, and uracil is added to inhibit degradation of FU through inhibition of dihydropyrimidine dehydrogenase, which is the enzyme responsible for FU catabolism. Thus, oral UFT administration can achieve a certain plasma FU concentration for a long time. Direct antitumor effects of UFT are minimal in most malignant tumors, including NSCLC,9 but UFT is widely used in Japan in postoperative adjuvant therapy because of its mild toxicity characteristics. In NSCLC, some trials, including the West Japan Study Group for Lung Cancer Surgery (WJSG) second trial10 and the Japan Lung Cancer Research Group (JLCRG) trial,11 have shown that postoperative UFT administration can improve the survival of completely resected NSCLC patients, but others failed to show the efficacy. Since meta-analysis of randomized controlled trials can most accurately evaluate a therapy,5,12-14 we conducted a meta-analysis to gain insight into the efficacy of postoperative UFT administration for completely resected NSCLC.
METHODS
Inclusion Criteria of Trials
Trials were eligible for inclusion in the present meta-analysis if they were randomized controlled trials of patients with completely resected NSCLC. Patients in the control group had to receive surgery alone, and patients in the treated group had to receive UFT alone following surgery; trials comparing surgery alone with surgery plus intravenous chemotherapy and UFT were not included. Finally, trials were eligible only if the postoperative follow-up duration was at least 5 years.
Protocol for the Meta-Analysis
In August 2002, a protocol for the meta-analysis, describing its rationale, statistical methodology, and rules for publication was produced in Tokyo University of Science, and distributed to the principal investigators of the six trials. Investigators were asked to provide individual data for every randomly assigned patient, whether eligible or not, assessable or not, and properly followed up or not. Information items requested for every patient were as follows: patient identification, date of surgery, eligibility, treatment allocated by random assignment, age, sex, histological type, pathological tumor size, pathological lymph node metastases, date of death, or last visit. Survival was calculated from the date of surgery to the date of death, regardless of the cause of death. Toxicity data were not collected, because detailed analysis of adverse reactions can be found in the respective published reports of individual trials.10,11,15
Identification of Trials
Trials were identified by searches of MEDLINE (1966 onward). Experts in lung cancer were also asked to help to identify trials. As a result, a total of nine prospective randomized adjuvant trials of UFT-containing chemotherapy, all conducted in Japan, were identified.10,11,15-21 Among them, one trial (the Hirosaki University trial) was excluded because a surgery-alone arm was not employed.16 Two trials (the Chubu-Japan trial and the WJSG third trial) were excluded because the chemotherapy arm was not UFT alone but cisplatin-based chemotherapy followed by UFT17,18; cisplatin and doxorubicin followed by UFT was employed as adjuvant chemotherapy in the Chubu-Japan trial,17 and cisplatin, vindesine, and mitomycin followed by UFT was employed in the WJSG third trial.18
Finally, a total of six trials conducted by five different groups that met the inclusion criteria of the present meta-analysis were identified as follows (Table 1 and Fig 1): (1) The WJSG second trial, Wada et al10; (2) the WJSG fourth trial, Nakagawa et al19; (3) The Northeast Japan Study Group for Lung Cancer Surgery trial, Endo et al15; (4) The Osaka Lung Cancer Study Group trial, Tada et al20; (5) Adjuvant Chemotherapy for Lung Cancer Study Group (ACTLC) trial, Imaizumi21; and (6) The JLCRG trial, Kato et al.11 No trial was excluded due to inadequate follow-up. Among the six trials included in the present meta-analysis, the WJSG-second trial and the ACTLC trial were three-arm trials consisting of a surgery-alone arm, surgery-plus-UFT arm, and surgery plus cisplatin-based chemotherapy followed by UFT arm10,21; patient data in the third arm were not included in the present meta-analysis.
Data
Detailed trial protocols and data on individual patients of each trial were provided by the principal investigators of the trials. Eligibility criteria for inclusion common to all six trials included in the meta-analysis were as follows: (1) histologically confirmed NSCLC; (2) p-stage I to III disease; (3) complete resection achieved; (4) age less than 76 years; (5) no previous anticancer treatment (radiotherapy, chemotherapy, endocrine therapy, or immunotherapy); (6) no multiple cancer; (7) written or oral informed consent. In all six studies, confirmation of patients' eligibility and random assignment were performed by telephone or facsimile at a registration center.
Statistical Analysis
Two major approaches for meta-analysis are the fixed effect model and the random effect model. We performed analyses based on both approaches. However, since heterogeneity among the studies was not observed, these two models showed almost identical results. Therefore, we only describe the former statistical model here and do not show the result based on the random effect model.
The following analysis based on individual patient data, was performed using the Statistical Analysis System (SAS version 8.2), mainly the PHREG and LIFETEST procedures. In the present meta-analysis, the patients who were eligible in each trial were primarily analyzed in order to ensure consistency with the results of individual trials, but all randomized patients, that is, according to intent-to-treatment (ITT) principle, were also analyzed to confirm the robustness of the results. The primary end point of the present study was overall survival, defined as the interval from operation to death from any cause. Data on patients who were alive were censored on the date on which they were last known to be alive.
Comparability Between Two Groups
Comparability for major demographic factors (age, sex) and important prognostic factors (histologic type, pathologic T and N stage) between the surgery alone and the surgery plus UFT groups in all six trials was evaluated. The distributions of these factors were described using the contingency table, and an adequate statistical test, depending on the distribution form, was performed to detect differences of distribution between the two groups.
Methods for Meta-Analysis
Employing the following Cox's proportional hazard model (a), the hazard ratio between the surgery alone and surgery plus UFT group for each trial was calculated. Suffix k stands for each trial. h(t) and hk(t) show hazard function and the baseline hazard function in the k trial, respectively. Variable UFT is 0, 1 dummy variable, depending on whether a patient belonged to the surgery alone or surgery plus UFT group. ?k corresponds to the log hazard ratio of each trial. Heterogeneity for the effect of UFT among six trials was tested assuming the null hypothesis ?1 = ?2 = ?3 = ?4 = ?5 = ?6 (df = 5). Overall pooled effect was estimated based on the fixed effect model (b) and tested, setting the null hypothesis ? = 0 (df = 1). This test is essentially identical to a trial-stratified log-rank test.
Cumulative survival rates in all six trials were estimated according to the Kaplan-Meier method. Adjusted analysis was performed adding important prognostic factors to Cox's proportional hazard model (b). Subset analysis and test for interaction between UFT and other factors were also conducted, according to the major demographic variables of age (69 years or younger v 70 years or older), sex, histological type, and p-T stage (pT1 v pT2). All statistical tests were two-sided. P < .05 was considered to indicate statistically significant differences.
RESULTS
Trials Included and Numbers of Patients
Between December 1985 and March 1997, a total of 2,082 patients (1,046 patients in the surgery-alone group and 1,036 patients in the surgery-plus-UFT group) were enrolled in the six studies, of whom 1,002 patients (95.8%) in the surgery alone group and 1,001 patients (96.6%) in the surgery plus UFT group were eligible and were analyzed. It should be noted that one trial (the JLCRG trial) contributed almost one half of the patients included in the present meta-analysis and was restricted to stage I adenocarcinoma patients.
Clinical Characteristics of Patients
The most common histologic type was adenocarcinoma (1,679 patients, 83.8%), followed by squamous cell carcinoma (299 patients, 14.9%). Most patients had p-stage I disease: 1,308 patients (65.3%) had p-T1 disease, and 674 (33.6%) had p-T2 disease, and 1,923 (96.0%) had no nodal involvement (N0). There was no major difference between the surgery-alone group, and the surgery-plus-UFT group in any major prognostic factor including age, sex, histologic type, pT factor, and pN factor (Table 2).
Postoperative Survival
The median follow-up duration for all eligible patients was 6.44 years after operation. The 5-year overall survival rate in the surgery plus UFT group was 81.8%, which was 4.6% higher than that in the surgery-alone group (5-year survival rate, 77.2%). At 7 years, overall survival rates were 69.5% and 76.5% in the surgery-alone group and in the surgery plus UFT group, respectively, showing more significant survival benefit in the surgery plus UFT group (Fig 2). The overall pooled hazard ratio was 0.74, and its 95% CI was 0.61 to 0.88 (P = .001). There was no heterogeneity in survival among the six studies (P = .79; Fig 3, O/N, observed death/number of patients; O – E, observed minus expected number of deaths; V, variance of O – E). Even when survival analysis was performed in all randomly assigned patients, including ineligible patients; that is, ITT analysis, patients in the surgery plus UFT group also showed a significantly better survival than those in the surgery alone group (hazard ratio, 0.76, 95% CI, 0.64 to 0.90; P = .005).
Subset Analysis
Overall survival was favorable in the surgery plus UFT group, regardless of age (HR for 69 years or younger, 0.76, 95% CI, 0.61 to 0.94; HR for 70 years or older, 0.68, 95% CI, 0.49 to 0.94), sex (HR for male, 0.77, 95% CI, 0.62 to 0.96; HR for female, 0.67, 95% CI, 0.48 to 0.92), histologic type (HR for adenocarcinoma, 0.69, 95% CI, 0.56 to 0.85; HR for squamous cell carcinoma, 0.82; 95% CI, 0.57 to 1.19), and pT-factor (HR for T1, 0.73; 95% CI, 0.56 to 0.93; HR for T2, 0.78; 95% CI, 0.60 to 1.01). Among all subsets, the efficacy of postoperative UFT administration was similar (Fig 4, O/N, observed deaths/number of patients; O – E, observed minus expected number of deaths; V, variance of O – E).
Multivariate Analysis
Multivariate analysis was applied using Cox's proportional hazard model with six factors, including age ( 69 years, 70 years), sex, histologic type (adenocarcinoma, squamous cell plus others), pT-factor (T0 plus T1, T2-4), pN-factor (N0, N1-3), and treatment (surgery alone, surgery plus UFT), as covariates. The results confirmed the benefit of treatment after adjustment for these covariates (HR for surgery plus UFT, 0.75; 95% CI, 0.63 to 0.90; P = .002).
DISCUSSION
Before the 2003 Annual Meeting of the American Society of Clinical Oncology (ASCO), no large-scale randomized controlled trial in which more than 1,000 patients had been randomized had shown the efficacy of postoperative adjuvant chemotherapy for NSCLC. According to a previous meta-analysis conducted by the Non–Small-Cell Lung Cancer Group,5 long-term administration of alkylating agents significantly increased the risk of postoperative death (P = .005). In contrast, cisplatin-based chemotherapy as well as treatment with other drugs, mainly tegafur and UFT, slightly reduced the risk of postoperative death (P = .08 and HR, 0.87; 95% CI, 0.74 to 1.02, for cisplatin-based chemotherapy; P = .30 and HR, 0.89; 95% CI, 0.72 to 1.11, for other drugs including UFT). As compared with meta-analyses conducted by the Early Breast Cancer Trialists' Collaborative Group clearly demonstrating the efficacy of tamoxifen and polychemotherapy in the treatment of breast cancer,12,13 the number of patients included in the Non–Small-Cell Lung Cancer Group meta-analysis was too small to draw definite conclusions (alkylating agents, n = 2,145; cisplatin-based chemotherapy, n = 1,394; other drugs, n = 918). At the 2003 ASCO meeting, results of two large-scale randomized controlled trials of postoperative adjuvant chemotherapy, the IALT trial showing the efficacy of cisplatin-based chemotherapy6 and the JLCRG trial showing the efficacy of UFT,11 were reported. At the 2004 ASCO meeting, results of two additional randomized adjuvant trials showing the efficacy of platinum-based chemotherapy, the NCIC-JBR.10 trial (cisplatin and vinorelbine)7 and the CALGB-9633 trial (carboplatin and paclitaxel),8 were reported.
UFT does not show potent direct antitumor effects for advanced malignant tumors, as the response rate in advanced and recurrent NSCLC has been estimated to be only 7%.9 However, in the postoperative adjuvant setting, UFT was clinically effective in rectal cancer22,23 and breast cancer,24,25 which was shown in both randomized controlled studies and meta-analyses. In NSCLC, Wada et al were the first to report that long-term oral administration of UFT improved the postoperative survival of completely resected NSCLC.10 Considering the time-dependent antitumor effects of FU, long-term oral UFT administration may be the optimal schedule of FU administration. In the postoperative setting, residual lesions requiring control were micrometastases, not bulky tumors. Even if the direct antitumor effects of UFT are relatively low, it can be administered for a long time because of its mild toxicity, which results in controlling micrometastases and improving survival. This concept is supported by the fact that in vivo experiments in mouse metastasis models of breast cancer26 and lung cancer27 have shown that UFT suppresses development of micro-metastases. Recently, two meta-analyses assessing the effectiveness of postoperative adjuvant chemotherapy with UFT have been published.28,29 However, these meta-analyses were based on summary data shown in the article. Our meta-analysis based completely on individual patient data, is considered to be more statistically reliable and accurate, as well as more useful, because more detailed analysis as subset analysis, multivariate analysis, and ITT analysis can be performed.30
The present meta-analysis showed that UFT administration for 1 to 2 years after surgery improved survival of NSCLC patients, mainly p-stage I (T1-2, N0) patients. The efficacy of UFT administration was relatively greater on analysis of survival data at 7 years than that at 5 years. This finding suggests that long-term observation is required to demonstrate the efficacy of postoperative adjuvant therapy in early-stage patients, who have a lower risk of postoperative death. We cannot identify the causes of death in all patients included in the present meta-analysis because of the unavailability of data. For around half of the patients in the meta-analysis, all of whom participated in the JLCRG trial, treatment failure patterns in the control group and UFT group have been reported as follows: (1) intrathoracic recurrence, 11.9% and 11.1%, respectively; (2) extrathoracic recurrence, 8.6% and 6.1%; (3) second cancer, 4.1% and 5.1%.11 From these results documented in the JLCRG study, it may be speculated that UFT can improve postoperative survival through inhibiting extrathoracic recurrence.
There was no separation of the survival curves until after the 4-year mark in the meta-analysis (Fig 2). In contrast, the benefit from adjuvant chemotherapy in other studies such as the NCIC-JBR.10 trial and the CALGB-9633 trial has been seen considerably sooner, usually within the first year or so. We cannot explain the exact reasons why such a unique survival curve was obtained in the present UFT meta-analysis, but one reason may be that the majority of patients included in the present meta-analysis were p-stage I patients, especially, p-stage IA patients. In such very early-stage patients, postoperative recurrence and death may be later, which results in overlapping of the survival curves of the surgery-alone arm and the chemotherapy arm. In fact, the survival curves were separate within 1 year after surgery, and the efficacy of UFT appeared sooner in the WJSG-second trial where p-stage I-III patients were eligible.10
Although a meta-analysis is, generally speaking, the most reliable and effective tool to assess the efficacy of a therapy, attention should be paid to one issue in the present meta-analysis in that one large-scale trial (the JLCRG trial), where only p-stage I adenocarcinoma patients were eligible, contributed almost one half of the patients included in the present meta-analysis. As a result, more than 80% patients were adenocarcinoma patients, and most patients had p-stage I disease among whom two-thirds had p-stage IA disease. Thus, the results of the present meta-analysis cannot easily be compared with any other meta-analysis or other trials of adjuvant chemotherapy.
Subset analyses in the present study revealed no significant interaction between the efficacy of postoperative UFT administration and age, sex, histological type, or p-T factor. However, as p-stage I adenocarcinoma composed the vast majority of the patient populations included in the present meta-analysis, the present meta-analysis may suggest that postoperative UFT administration is effective especially in p-stage I adenocarcinoma of patients. Unfortunately, the present meta-analysis failed to show whether or not UFT was effective in advanced-stage diseases and/or squamous cell carcinoma, probably due to insufficient numbers of patients in these subsets.
In the IALT trial, where 36% of patients were stage I, 25% were stage II, and 39% were stage III, cisplatin-based chemotherapy was effective for p-stage I-III patients (HR, 0.86). In the NCIC-JBR.10 trial, where 45% of patients were stage IB and 55% were stage II, cisplatin in combination with vinorelbine was effective for p-stage IB-II patients (HR, 0.62). The CALGB trial, where all patients were stage IB, showed that carboplatin in combination with paclitaxel was effective for p-stage IB patients (HR, 0.71). These results show that platinum-based postoperative adjuvant chemotherapy may be active for a wide range of patient populations, including p-stage II-III patients. In contrast, the present meta-analysis showed that UFT was effective in a patient population composed primarily of stage I adenocarcinoma patients. In addition, the HR documented in the present meta-analysis for UFT administration (HR, 0.74) was inferior to those documented in the NCIC-JBR.10 trial and CALGB-9633 trial. However, these results indicate that UFT may be recommended for early-stage patients, and platinum-based chemotherapy should be employed for advanced-stage (ie, stage II-III) patients.
Anticancer therapy should be performed based on an individual assessment of the risk of recurrence and death, the balance between the toxicity and efficacy, and changes in quality of life caused by the therapy. According to a cross-over study on patients' choice of therapy (injection or oral therapy) showing the same efficacy, 85% of patients preferred oral therapy because it was more convenient.31 UFT is an oral drug that can be given for more than 1 year, and is preferable in the postoperative setting not only because of the mild toxicity and long-term effects but also because of the patients' convenience.
In conclusion, our meta-analysis shows that long-term treatment with UFT for 1 to 2 years was effective as postoperative adjuvant therapy in a Japanese patient population composed primarily of stage I adenocarcinoma patients. To assess the efficacy of UFT outside Japan, additional prospective randomized studies of adjuvant UFT in completely resected NSCLC should be conducted in North America and Europe.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
Acknowledgment
The authors are indebted to J. Patrick Barron of the International Medical Communications Center of Tokyo Medical University for his review of this article.
NOTES
Supported by Taiho Pharmaceutical Co, Tokyo, Japan (the six trials, including the meta-analysis). This company had no role in the meta-analysis study design, data collection, data analysis, data interpretation, or in writing the report.
Presented at the 12th European Cancer Conference meeting, Copenhagen, Denmark, 2003 and the 40th Annual Meeting of the American Society of Clinical Oncology, New Orleans, LA, June 5-8, 2004.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Mountain CF: Revisions in the international system for staging lung cancer. Chest 116:1710-1717, 1997
, Naruke T, Tsuchiya R, Kondo H, et al: Implication of staging in lung cancer. Chest 112:242s-248s, 1997
PORT Meta-analysis Trialists Group: Postoperative radiotherapy in non-small-cell lung cancer: Systematic review and meta-analysis of individual patient data from nine randomised controlled trials. Lancet 352:257-263, 1998
Scagliotti GV, Fossati R, Torri V, et al: Randomized study of adjuvant chemotherapy for completely resected stage I, II, or IIIA non-small-cell lung cancer. J Nat Cancer Inst 95:1453-1461, 2003
Non-small Cell Lung Cancer Collaborative Group: A meta-analysis using updated data on individual patients from 52 randomised clinical trials. BMJ 311:899-909, 1995
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 350:351-360, 2004
Winton TL, Livingston R, Johnson D, et al: A prospective randomized trial of adjuvant vinorelbine (VIN) and cisplatin (CIS) in completely resected stage IB and II non small cell lung cancer (NSCLC) Intergroup JBR 10. J Clin Oncol 22:621s, 2004 (abstr 7018)
Strauss GM, Maddaus MA, Johnstone DW, et al: Randomized clinical trial of adjuvant chemotherapy with paclitaxel and carboplatin following resection in stage IB non-small cell lung cancer (NSCLC): Report of Cancer and Leukemia Group B (CALGB) protocol 9633. J Clin Oncol 22:621s, 2004 (abstr 7019)
Taguchi T: Clinical application of biochemical modulation in cancer chemotherapy: Biochemical modulation for 5-FU. Oncology 54:12-18, 1997 (suppl 1)
Wada H, Hitomi S, Teramatsu T, et al: Adjuvant chemotherapy after complete resection in non-small-cell lung cancer. J Clin Oncol 14:1048-1054, 1996
Kato H, Ichinose Y, Ohta M, et al: A randomized trail of adjuvant chemotherapy with uracil-tegafur for adenocarcinoma of the lung. N Engl J Med 350:1713-1721, 2004
Early Breast Cancer Trialists' Collaborative Group: Tamoxifen for early breast cancer: An overview of the randomised trails. Lancet 351:1451-1467, 1998
Early Breast Cancer Trialists' Collaborative Group: Polychemotherapy for early breast cancer: An overview of the randomised trails. Lancet 352:930-942, 1998
Colorectal Cancer Collaborative Group: Adjuvant radiotherapy for rectal cancer: A systemic overview of 8,507 patients from 22 randomised trials. Lancet 358:1291-1304, 2001
Endo C, Saito Y, Iwanami H, et al: A randomized trial of postoperative UFT therapy in p stage I, II non-small cell lung cancer: North-east Japan study group for lung cancer surgery. Lung Cancer 40:181-186, 2003
Fukushima M, Tsushima T, Ito H, et al: A randomized controlled study of post-operative adjuvant therapy in non-small cell lung cancer. Jpn J Cancer Chemother 23:303-309, 1996
The Study Group of Adjuvant Chemotherapy for Lung Cancer (Chubu, Japan): A randomized trial of postoperative adjuvant chemotherapy in non-small cell lung cancer (the second cooperative study). Eur J Surg Oncol 21:69-77, 1995
Wada H, Miyahara R, Tanaka F, et al: Postoperative adjuvant chemotherapy with PVM (Cisplatin+Vindesine+Mitomycin C) and UFT (Uracil+Tegaful) a randomized clinical trial: West Japan Study Group for lung cancer surgery (WJSG). Eur J Cardiothorac Surg 15:438-443, 1999
Nakagawa M, Tanaka F, Tsubota N, et al: A randomized phase III trial of adjuvant chemotherapy with UFT for completely resected pathological stage I non-small-cell lung cancer: The west Japan study group for lung cancer surgery (WJSG), the 4th study. Ann Oncol 16:75-80, 2005
Tada H, Yasumitsu T, Iuchi K, et al: Randomized study of adjuvant chemotherapy for completely resected non-small cell lung cancer. Proc J Clin Oncol 21:313a, 2002 (abstr 1250)
Imaizumi M: A randomized trial of postoperative adjuvant chemotherapy for p-stage I non-small cell lung cancer (4th cooperative study). Lung Cancer 41:S2, 2003 (abstr O180)
Kato T, Ohashi Y, Nakazato H, et al: Efficacy of oral UFT as adjuvant chemotherapy to curative resection of colorectal cancer: Multicenter prospective randomized trial. Langenecks Arch Surg 386:575-581, 2001
Sakamoto J, Hamada C, Yasutomi M, et al: Efficacy of oral UFT as adjuvant chemotherapy after resection of rectal Cancer: Meta-analysis of 5-year results from four randomized trials in 1815 patients. Proc Am Soc Clin Oncol 22:270, 2003 (abstr 1082)
Ogita M, Uchino J, Asaishi K, et al: Efficacy of UFT plus tamoxifen for estrogen-receptor-positive breast cancer and tamoxifen plus UFT for estrogen-receptor-negative breast cancer. Clin Drug Invest 23:689-699, 2003
Kasumi F, Yoshimoto M, Uchino J, et al: Meta-analysis of five studies on tegafur plus uracil (UFT) as post-operative adjuvant chemotherapy for breast cancer. Oncology 64:146-153, 2003
Kurebayashi J, Nakatsuka M, Fujioka A, et al: Postsurgical oral administration of uracil and tegafur inhibits progression of micrometastasis of human breast cancer cells in nude mice. Clin Cancer Res 3:653-659, 1997
Miyahara R, Nakagawa T, Ishikawa S, et al: UFT inhibits lung metastases in spontaneous metastasis model of lung cancer. Thorac Cardiovasc Surg 53:118-121, 2005
Sedrakyan A, Der Meulen JV, O'Byrne K, et al: Postoperative chemotherapy for non-small cell lung cancer: A systematic review and meta-analysis. Gen Thorac Surg 128:414-419, 2004
Hotta K, Matsuo K, Ueoka H, et al: Role of adjuvant chemotherapy in patients with resected non-small-cell lung cancer: Reappraisal with a meta-analysis of randomized controlled trials. J Clin Oncol 22:3860-3867, 2004
Piedbois P, Mondor H, Buyse M: Meta-analyses based on abstracted data: A step in the right direction, but only s first step. J Clin Oncol 22:3839-3841, 2004
Borner MM, Schoffski P, de Wit R, et al: Patient preference and pharmacokinetics of oral modulated UFT versus intravenous fluorouracil and leucovorin: A randomised crossover trial inadvanced colorectal cancer. Eur J Cancer 38:349-358, 2002(Chikuma Hamada, Fumihiro )