The Journal of International Medical Research 2008; 36: 875 – 882 [first published online as 36(5) 1]
Adjuvant Chemotherapy in Oesophageal Cancer: a Meta-analysis and Experience from the Shanghai Cancer Hospital J ZHANG, H-Q CHEN, Y-W ZHANG
AND
J-Q XIANG
Department of Oncology, Shanghai Medical College; Department of Thoracic Surgery, Shanghai Cancer Hospital, Fudan University, Shanghai, China
Whether adjuvant chemotherapy increases survival of oesophageal cancer patients has been widely debated. The present study used meta-analysis software to combine data from six studies up to July 2007 that were found and selected as suitable, comprising a total of 1001 oesophageal cancer patients. The results indicated that adjuvant chemotherapy did not significantly improve outcome in oesophageal cancer patients. A trend towards improved outcome from adjuvant chemotherapy was found in lymph nodeKEY WORDS: ADJUVANT
positive patients, but did not reach significance. In our own study including 270 oesophageal cancer patients, adjuvant chemotherapy did not improve overall patient survival, but did improve survival for patients with metastases in cervical and/or celiac lymph nodes (stage IVa). Although our study had the largest patient sample, more prospective clinical trials with large numbers of patients are necessary to confirm the value of adjuvant chemotherapy in stage IVa patients.
CHEMOTHERAPY;
OESOPHAGEAL META-ANALYSIS
Introduction Carcinoma of the oesophagus and gastrooesophageal junction is an aggressive disease with a poor prognosis. It is the second most common cause of cancer-related deaths in women and the fourth in men in China.1 Although surgical oesophagectomy remains the primary treatment for clinically localized thoracic oesophageal carcinoma, there is increasing enthusiasm for multidisciplinary treatment modality to improve outcome. In the 1990s, no survival benefit for postoperative chemotherapy in oesophageal cancer was shown in three small-sample
CANCER;
LYMPH
NODE-POSITIVE PATIENTS;
randomized clinical trials.2 – 4 However, some data showed that adjuvant chemotherapy might prolong disease-free survival (DFS) or improve treatment outcome in oesophageal cancer.5,6 In the 2000s, there has been increasing evidence. The Japan Clinical Oncology Group Study (JCOG 9204)7 showed that post-operative chemotherapy can significantly increase 5-year DFS, but not overall survival (OS) in oesophageal cancer patients; the effect was marked in the subgroup with lymph node metastasis. Lee et al.8 reported that the 3-year DFS rate was 47.6% in a group given adjuvant
875
J Zhang, H-Q Chen, Y-W Zhang et al. Adjuvant chemotherapy in oesophageal cancer chemotherapy compared with 35.6% in the control group (P = 0.049) for lymph nodepositive thoracic oesophageal squamous cell carcinoma. Another study from Japan showed long-term post-operative treatment with low-dose cisplatin (CDDP) plus 5fluorouracil (5-FU) was therapeutically beneficial and prolonged survival in patients with regional lymph node metastasis or lymphatic invasion, or stage III disease.9 To elucidate the role of adjuvant chemotherapy in oesophageal cancer and to guide future clinical practice, the current meta-analysis combined data from all relevant studies in English language publications up to July 2007. The survival rates of patients at the Shanghai Cancer Hospital who received adjuvant chemotherapy of cisplatin and 5-FU plus leucovorin were also compared with those who did not receive this regimen. The majority of patients underwent three-field lymphadenectomy.
Patients and methods META-ANALYSIS DATA GATHERING The resources used for the current metaanalysis were PubMed, American Society of Clinical Oncology (ASCO) abstracts and CancerLit. Key words used in the search were ‘oesophageal neoplasms’ and ‘adjuvant chemotherapy’. For the identified papers, related articles provided by PubMed and papers listed in the references were further reviewed to identify additional publications. The analysed data covered those reported in English language publications up to July 2007. When exact survival rates were not available we: (i) contacted the corresponding author to obtain the survival rates of subgroups8; or (ii) used survival curves to obtain survival rates at some time points, as previously reported10,11 (although we recognize this is a suboptimal method). Data
extraction from survival curves was performed by two researchers independently.
SHANGHAI CANCER HOSPITAL EXPERIENCE Patients All patients included in the study had been referred for surgical treatment at Shanghai Cancer Hospital, Fudan University between 1998 and 2004. They were categorized into two groups according to whether or not they received post-operative adjuvant chemotherapy. Patients received either transthoracic oesophagectomy with threefield lymphadenectomy or two-field lymph node dissection (including the lymph nodes along the recurrent nerve). The bilateral supraclavicular lymph nodes and those along the recurrent nerve were removed in three-field lympadenectomy. Oesophageal reconstruction was performed using the stomach, via the oesophageal bed or retrosternal route. Patient inclusion criteria were: (i) confirmation of oesophageal cancer by histology; (ii) pathological stages I, IIa, IIb, III or IVa due to distant node involvement (M1 lymph) only; (iii) baseline laboratory tests, i.e. white blood cell count > 4.0 × 109/l, platelet count > 100 × 109/l, adequate renal function (alanine aminotransferase and aspartate aminotransferase within normal levels); (iv) Karnofsky performance status > 70; (v) within 7 weeks post-operation; (vi) no prior antitumour therapy; and (vii) no serious medical conditions that would preclude safe administration of treatment. Patients or their legal representatives signed a patient consent form prior to treatment being given. Chemotherapy and follow-up Chemotherapy comprised cisplatin 25 mg/m2 daily from day 1 to 3, and 5-FU 375
876
J Zhang, H-Q Chen, Y-W Zhang et al. Adjuvant chemotherapy in oesophageal cancer mg/m2 and leucovorin 135 mg/m2 daily from day 1 to 5, every 3 weeks. Four to six cycles of chemotherapy were planned. Toxicity during chemotherapy was evaluated and managed by the oncologists. After treatment, patients were followed-up every 4 months for the first year, every 6 months for the next 2 years, and then annually thereafter. This was undertaken either as outpatients or by letter or telephone interview by a full-time senior nurse. The patients remaining alive were followed-up again prior to the start of this analysis.
DATA AND STATISTICAL ANALYSES The meta-analysis was conducted using Comprehensive Meta-analysis software downloaded from www.meta-analysis.com (Biostat Inc., Englewood, NJ, USA). A Q-test was performed to assess possible heterogeneity between the individual studies. Where heterogeneity existed, a random effects model was adopted; otherwise both fixed effects and random models were deemed appropriate. Patient survival curves were calculated according to the Kaplan–Meier method and
compared by the log-rank test. The OS rate was estimated using life tables. Frequencies were compared using crosstabs and the χ2 test for qualitative variables. The Cox proportional hazards model was used for multivariate analysis. A P-value < 0.05 was considered to be statistically significant.
Results META-ANALYSIS Up to July 2007 only six studies comprising a total of 1001 oesophageal cancer patients could be found and were selected for the current meta-analysis. The Q-test indicated heterogeneity between the individual studies (Q = 13.268, P = 0.021) and, therefore, the random effects model was adopted to calculate odd ratios (OR). The overall OR for adjuvant chemotherapy was 0.961 (95% confidence interval [CI] 0.741 – 1.246; not statistically significant); detailed results are shown in Table 1. The significance of adjuvant chemotherapy in lymph node-positive patients was further investigated and four of the original six studies were identified for analysis. There was no significant
TABLE 1: Meta-analysis of post-operative adjuvant chemotherapy on the survival of patients with oesophageal cancer (random effects model) No. of patients Study and year
OR (95% CI)
Treatment Control
Ando et al., 20037 Pouliquen et al., 19963 Ando et al., 19974 Lee et al., 20058 Heroor et al., 200312 Shiozaki et al., 20059 Overall
120 24 105 60 99 98 506
122 38 100 66 117 52 495
0.920 0.923 0.847 1.214 1.725 0.312 0.961
(0.543 (0.310 (0.487 (0.603 (1.005 (0.144 (0.741
– – – – – – –
Statistical significance
OR and 95% CI
1.560) NS 2.746) NS 1.471) NS 2.446) NS 2.962) P = 0.048 0.677) P = 0.003 1.246) NS 0.1 Favours adjuvant chemotherapy
OR, odds ratio; CI, confidence interval; NS, not statistically significant (P > 0.05).
877
1
10 Favours no adjuvant chemotherapy
J Zhang, H-Q Chen, Y-W Zhang et al. Adjuvant chemotherapy in oesophageal cancer heterogeneity among them (Q = 4.871). Using a fixed effects model, meta-analysis of these four studies showed a trend towards a better outcome from adjuvant chemotherapy in lymph node-positive patients, although this was not statistically significant (overall OR of 0.763 in favour of adjuvant chemotherapy; 95% CI 0.538 – 1.083; not statistically significant) (Table 2).
SHANGHAI CANCER CENTRE EXPERIENCE Patients Baseline clinical and pathological factors for the oesophageal cancer patients treated with (n = 90) or without (n = 180) post-operative adjuvant chemotherapy, including age, gender, tumour location, TNM (tumour, node, metastasis) stage and surgical methods were comparable between the two groups and are shown in Table 3. The ages of the patients ranged from 30 to 77 years, with a mean age of 53.81 years for the chemotherapy group and 56.06 years for the non-chemotherapy group. Four to six cycles of chemotherapy were intended and 66 (73.3%) patients received the planned schedule. The patients experienced mild-to-
moderate common acute and late toxicity to chemotherapy although, generally, the treatment was well tolerated and there was no treatment-induced mortality. In stage IVa, 14 out of 22 patients were of pathological stage T1 or T2, and 18 had lymph node metastasis limited to the cervical or celiac lymph node. DFS and OS by adjuvant treatment modality For the entire population studied, neither adjuvant chemotherapy nor radiation therapy improved DFS and OS. Similar treatment outcomes were found in terms of DFS and OS when the following adjuvant treatment modalities were compared: combination of chemotherapy and radiation therapy; chemotherapy alone; radiation therapy alone; and no adjuvant treatment (3-year DFS 45.0%, 34.4%, 50.0% and 46.3%, respectively, and 3-year OS 45.0%, 51.3%, 57.1% and 47.0%, respectively). DFS and OS by pathological stage In terms of pathological stage, both DFS and OS showed no significant differences for stages I, II and III between patients with and
TABLE 2: Meta-analysis of post-operative adjuvant chemotherapy on survival in lymph node-positive patients (fixed effects model) No. of patients Study and year Ando et al., 20037 Lee et al., 20058 Heroor et al., 200312 Shiozaki et al., 20059 Summary
OR (95% CI)
Treatment Control 97 60 76 68 301
101 66 60 26 253
0.567 1.214 0.996 0.388 0.763
(0.322 (0.603 (0.503 (0.133 (0.538
Statistical significance
OR and 95% CI
– 0.999) P = 0.05 – 2.446) NS – 1.974) NS – 1.13) NS –1.083) NS 0.1 Favours adjuvant chemotherapy
OR, odds ratio; CI, confidence interval; NS, not statistically significant (P > 0.05).
878
1
10 Favours no adjuvant chemotherapy
J Zhang, H-Q Chen, Y-W Zhang et al. Adjuvant chemotherapy in oesophageal cancer
TABLE 3: Baseline clinical and pathological factors for oesophageal cancer patients treated with or without post-operative adjuvant chemotherapy Parameter Mean age, years (range) Male gender, n (%) Tumour location n (%)a Proximal Middle Distal Tumour stage, n (%)a T1 T2 T3 – T4 Node, n (%) None 1–2 3–5 >6 TNM stage, n (%)a I IIa IIb III IVa Tumour length, n (%)a < 3 cm 3 – 7 cm > 7 cm Grade, n (%)a I (SCC) II III Adenocarcinoma, n (%) Surgery, n (%)a Three-field lymphadenectomy Two-field lymph node dissection
Without adjuvant chemotherapy (n = 180)
With adjuvant chemotherapy (n = 90)
56.06 (30 – 77) 136 (75.6)
53.81 (34 – 68) 70 (77.8)
6 (3.3) 106 (58.9) 66 (36.7)
8 (8.9) 36 (40.0) 42 (46.7)
16 (8.9) 66 (36.7) 86 (47.8)
10 (11.1) 24 (26.7) 56 (62.2)
76 66 26 10
(42.2) (36.7) (14.4) (5.6)
24 42 20 4
(26.7) (46.7) (22.2) (4.4)
10 62 32 42 26
(5.6) (34.4) (17.8) (23.3) (14.4)
4 20 8 36 22
(4.4) (22.2) (8.9) (40.0) (24.4)
22 (12.2) 100 (55.6) 36 (20.0)
16 (17.8) 36 (40.0) 16 (17.8)
24 108 24 8
28 42 12 4
(13.3) (60.0) (13.3) (4.4)
134 (74.4) 44 (24.4)
(31.1) (46.7) (13.3) (4.4)
66 (73.3) 20 (22.2)
aSome
data were missing for some assessments. SCC, squamous cell carcinoma; TNM, tumour, node, metastases. There were no significant differences between the two groups (P > 0.05).
without adjuvant chemotherapy (Table 4). On the other hand, patients who had metastasis in the cervical and/or celiac lymph nodes (stage IVa) and received postoperative chemotherapy had a much better outcome compared with patients in the other
stages (1-year DFS/OS 100%/100%, 3-year DFS/OS 75%/100%). The 1- and 3-year DFS and OS rates were significantly better for patients who received post-operative chemotherapy than for patients who did not receive it (DFS P = 0.038; OS P = 0.01; Table 4).
879
J Zhang, H-Q Chen, Y-W Zhang et al. Adjuvant chemotherapy in oesophageal cancer
TABLE 4: Disease free survival (DFS) and overall survival (OS) of patients with oesophageal cancer treated with or without post-operative adjuvant chemotherapy, stratified according to TNM (tumour, nodes, metastases) stage
DFS
OS
Without adjuvant chemotherapy
With adjuvant chemotherapy
3-year survival (%)
n
1-year survival (%)
75 58 38 24 75 58 40 25
4 28 36 22 4 28 36 22
TNM stage
na
1-year survival (%)
I II III IVa I II III IVa
10 94 42 26 10 94 42 26
75 79 44 68 75 79 46 75
100 64 76 100 100 60 76 100
3-year survival (%) 100 40 30 75 100 40 19 100
Statistical significance NS NS NS 0.038 NS NS NS 0.01
aSome
data were missing for some patients in the without adjuvant chemotherapy group. NS, not statistically significant (P > 0.05).
In the multivariate study using the Cox regression model, it was also found that adjuvant chemotherapy was an independent good prognostic factor in stage IVa patients.
Discussion In western countries, pre-operative chemotherapy and chemoradiation therapy can increase OS by 4.4% and 6.4%, respectively, however, treatment-related mortality increases by 1.7% with neoadjuvant chemotherapy and by 3.4% with chemoradiotherapy, compared with surgery alone.13 – 17 In eastern countries, including Japan and China, many physicians prefer to delay administration of chemotherapy or chemoradiation therapy until post-operation.18 Until now, no meta-analysis had exclusively studied the role of adjuvant chemotherapy in oesophageal cancer. The only related meta-analysis that had been performed was by the Gastrointestinal Cancer Disease Site Group of Cancer Care
Ontario’s Program in Evidence-based Care.2 They selected three studies, but one could not be included in the pooled analysis because the abstract did not report 3-year survival. No significant difference in the risk of mortality at 3 years for post-operative chemotherapy compared with surgery alone was found (relative risk 0.94; 95% CI 0.74 – 1.18; not statistically significant). Since this meta-analysis on adjuvant treatment in oesophageal cancer was published several years ago, there has been a renewed enthusiasm for the investigation of chemotherapy in oesophageal cancer and a considerable number of new studies have been reported. It is, therefore, now timely to undertake a major update and re-evaluation of adjuvant chemotherapy in oesophageal cancer. The present meta-analysis indicated that adjuvant chemotherapy did not improve the outcome of oesophageal cancer patients. A trend towards better outcome from adjuvant chemotherapy in lymph node-positive
880
J Zhang, H-Q Chen, Y-W Zhang et al. Adjuvant chemotherapy in oesophageal cancer patients was detected, but the difference did not reach significance. Considering the heterogeneities of the studies and the bias of the meta-analysis, with the studies spanning one decade and the type of surgery varying from one study to another, our conclusions are tempered at present. Comparisons with control groups were also varied: for example, Lee et al.8 used an historical control, whereas Ando et al.7 was a multicentre randomized controlled trial. Additionally, no individual patient data were available to enhance the meta-analysis and possible differences in statistical methods may also have influenced it. More clinical trials with larger sample sizes are, therefore, necessary to elucidate the value of adjuvant chemotherapy in lymph node metastasis patients. In the present study involving patients at Shanghai Cancer Hospital, the survival of patients with stages I, II and III disease who received adjuvant chemotherapy was similar to those did not receive chemotherapy. Adjuvant chemotherapy was most effective for patients who had metastasis in the cervical and/or celiac lymph nodes and significantly improved the treatment results of stage IVa patients. As this was a retrospective analysis, it had some advantages compared with other randomized clinical trials. First, three-field lymphadenectomy was performed on the majority of patients and the cervical–thoracic lymph nodes along the recurrent nerve and celiac lymph nodes were also dissected in a two-field oesophagectomy. Although threefield lymphadenectomy has not been the standard of care worldwide, it does provide improved staging. The patients in the present study could, therefore, be classified according to their cervical and/or celiac lymph node metastasis, thereby identifying those most likely to benefit from adjuvant chemotherapy. In the study by Lee et al.,8 only 10% of patients received three-field lymphadenectomy. In the
other studies selected for the meta-analysis, although three-field lymphadenectomy was one of the surgical procedures, the survival of patients with cervical and/or celiac lymph node-positive (stage IVa) was not calculated. Secondly, in our clinical practice setting, the chemotherapy schedule that we used worked well with Chinese patients: four to six cycles of chemotherapy were suggested for the oesophageal cancer patients and 73.3% of patients received the planned schedule. The dose intensity was similar to that of the Japanese and French studies3,7 mentioned above, but total dose was larger than in the Japanese study (two cycles) and smaller than in the French study (8 – 10 cycles). Thirdly, the current study had a larger patient sample compared with other studies published. It is important to remain cautious about the current finding that adjuvant chemotherapy can significantly improve the prognosis of stage IVa patients. In the present study, 14 out of 22 stage IVa patients were pathological T1 or T2, and 18 had lymph node metastasis limited to the cervical or celiac lymph node. It is likely that patients with metastatic disease confined only to the cervical or celiac lymph node can survive longer than expected, and achieve more benefit from adjuvant chemotherapy.
Acknowledgements This work was supported, in part, by Shanghai Cancer Hospital grant KY176, Chinese post-doc fellowship 20070410733 and Bio-medicine grant, Shanghai Natural Science Fund, Young Investigator grant from the Chinese Educational Department (J.Z.) and Shanghai Leading Academic Discipline Project (B115).
Conflicts of interest The authors had no conflicts of interest to declare in relation to this article.
881
J Zhang, H-Q Chen, Y-W Zhang et al. Adjuvant chemotherapy in oesophageal cancer • Received for publication 9 June 2008 • Accepted subject to revision 12 June 2008 • Revised accepted 18 July 2008 Copyright © 2008 Field House Publishing LLP
References 1 Liu JF, Wang QZ, Hou J: Surgical treatment for cancer of the oesophagus and gastric cardia in Hebei, China. Br J Surg 2004; 91: 90 – 98. 2 Malthaner RA, Wong RK, Rumble RB, et al and members of the Gastrointestinal Cancer Disease Site Group of Cancer Care Ontario’s Program in Evidence-based Care: Neoadjuvant or adjuvant therapy for resectable esophageal cancer: a systematic review and meta-analysis. BMC Med 2004; 2: 35. 3 Pouliquen X, Levard H, Hay JM, et al: 5Fluorouracil and cisplatin therapy after palliative surgical resection of squamous cell carcinoma of the esophagus. A multicenter randomized trial. French Associations for Surgical Research. Ann Surg 1996; 223: 127 – 133. 4 Ando N, Iizuka T, Kakegawa T, et al: A randomized trial of surgery with and without chemotherapy for localized squamous carcinoma of the thoracic esophagus: the Japan Clinical Oncology Group Study. J Thorac Cardiovasc Surg 1997; 114: 205 – 209. 5 Minsky BD: Carcinoma of the esophagus. Part 2: Adjuvant therapy. Oncology (Williston Park) 1999; 13: 1415 – 1427. 6 Liu HC, Hung SK, Huang CJ, et al: Esophagectomy for locally advanced esophageal cancer, followed by chemoradiotherapy and adjuvant chemotherapy. World J Gastroenterol 2005; 11: 5367 – 5372. 7 Ando N, Iizuka T, Ide H, et al: Surgery plus chemotherapy compared with surgery alone for localized squamous cell carcinoma of the thoracic esophagus: a Japan Clinical Oncology Group Study–JCOG9204. J Clin Oncol 2003; 21: 4592 – 4596. 8 Lee J, Lee KE, Im YH, et al: Adjuvant chemotherapy with 5-fluorouracil and cisplatin in lymph node-positive thoracic esophageal squamous cell carcinoma. Ann Thorac Surg 2005; 80: 1170 – 1175. 9 Shiozaki A, Yamagishi H, Itoi H, et al: Longterm administration of low-dose cisplatin plus
5-fluorouracil prolongs the postoperative survival of patients with esophageal cancer. Oncol Rep 2005; 13: 667 – 672. 10 Wang L, Shao ZM: Cyclin expression and prognosis in breast cancer patients: a metaanalysis of published studies. Cancer Invest 2006; 24: 581 – 587. 11 Li D, He L: Association study of the G-protein signaling 4 (RGS4) and proline dehydrogenase (PRODH) genes with schizophrenia: a metaanalysis. Eur J Hum Genet 2006; 14: 1130 – 1135. 12 Heroor A, Fujita H, Sueyoshi S, et al: Adjuvant chemotherapy after radical resection of squamous cell carcinoma in the thoracic esophagus: who benefits? Dig Surg 2003; 20: 229 – 237. 13 Kelsen DP, Ginsberg R, Pajak TF, et al: Chemotherapy followed by surgery compared with surgery alone for localized esophageal cancer. N Engl J Med 1998; 339: 1979 – 1984. 14 Iyer R, Wilkinson N, Demmy T, et al: Controversies in the multimodality management of locally advanced esophageal cancer: evidence-based review of surgery alone and combined-modality therapy. Ann Surg Oncol 2004; 11: 665 – 673. 15 Fiorica F, Di Bona D, Schepis F, et al: Preoperative chemoradiotherapy for oesophageal cancer: a systematic review and meta-analysis. Gut 2004; 53: 925 – 930. 16 Medical Research Council Oesophageal Cancer Working Group: Surgical resection with or without preoperative chemotherapy in oesophageal cancer: a randomised controlled trial. Lancet 2002; 359: 1727 – 1733. 17 Kaklamanos IG, Walker GR, Ferry K, et al: Neoadjuvant treatment for resectable cancer of the esophagus and the gastroesophageal junction: a meta-analysis of randomized clinical trials. Ann Surg Oncol 2003; 10: 754 – 761. 18 Chong G, Cunningham D: The role of preoperative chemotherapy for esophageal cancer: when is surgery alone feasible? Nat Clin Pract Oncol 2005; 2: 172 – 173.
Authors’ address for correspondence Dr J-Q Xiang or Dr H-Q Chen Department of Thoracic Surgery, Shanghai Cancer Hospital, Fudan University, Shanghai 200032, China E-mail:
[email protected]
882
