Clinical Study Received: April 4, 2013 Accepted after revision: July 19, 2013 Published online: January 15, 2014
Oncology 2014;86:79–85 DOI: 10.1159/000354641
Chemoradiation Therapy versus Chemotherapy Alone for Gastric Cancer after R0 Surgical Resection: A Meta-Analysis of Randomized Trials Christine Min Sripal Bangalore Sachin Jhawar Yu Guo Joseph Nicholson Silvia C. Formenti Lawrence P. Leichman Kevin L. Du New York University School of Medicine, New York, N.Y. USA
Abstract Objective: Current national guidelines include category 1 recommendations for perioperative chemotherapy or adjuvant chemoradiation with surgical resection for patients with stage IB–IIIB gastric cancer. We conducted a meta-analysis of randomized trials in which chemotherapy was prospectively tested against chemoradiation with surgical resection. Methods: We electronically searched PubMed and EMBASE for randomized, controlled clinical trials involving patients with gastric adenocarcinoma, status post-R0 resection. The interventions compared were adjuvant chemotherapy versus chemoradiation, with any chemotherapy regimen. The primary outcomes of interest were diseasefree survival and overall survival. The Mantel-Haenszel random-effects model was used to calculate effect sizes. Results: Six trials that included 1,171 patients were evaluated; 599 were randomized to adjuvant chemoradiation and 572 to chemotherapy alone. Chemoradiation was associated with a significant increase in disease-free survival (odds ratio 1.48, 95% confidence interval 1.08–2.03) when compared to chemotherapy alone. However, there was no significant difference in overall survival (odds ratio 1.27, 95% confidence
© 2014 S. Karger AG, Basel 0030–2414/14/0862–0079$39.50/0 E-Mail [email protected] www.karger.com/ocl
interval 0.95–1.71). Five trials found no statistically significant differences in toxicities between the two groups. Conclusion: In patients with gastric cancer status post-R0 resection, adjuvant chemoradiation was associated with higher disease-free survival when compared to chemotherapy alone. It remains appropriate to design trials testing new systemic agents with radiotherapy. © 2014 S. Karger AG, Basel
Introduction
Gastric cancer is the third most common cancer and the second leading cause of cancer-related deaths in the world. Approximately 21,000 new cases are diagnosed each year with nearly 11,000 deaths [1]. The incidence, etiologically linked to chronic Helicobacter pylori infection and to dietary intake of highly salted and smoked food, is highest in East Asian countries [2–3]. Although the incidence in the West has declined over the past 50 years, the notable exception is noncardia gastric cancers that frequently harbor a diffuse histology and are found in younger patients [4–5]. Partial or total gastrectomy is the mainstay of curative therapy for patients with gastric cancer. However, even with an aggressive D2 lymph node resection (removal of perigastric, celiac, splenic artery and splenic hilar lymph nodes), the risk of local and distant recurrence remains Christine Min, MD New York University School of Medicine 160 East 34th Street LL1 New York, NY 10016 (USA) E-Mail christinemin @ jhmi.edu
Downloaded by: Selçuk Universitesi 193.255.248.150 - 1/7/2015 1:33:23 PM
Key Words Chemoradiation · Gastric adenocarcinoma · Gastric cancer · Gastric carcinoma · Radiation therapy · Radiochemotherapy · Radiotherapy · Randomized trial
Methods Search Methods for Identification of Studies Electronic searches were conducted on PubMed and EMBASE with the following keywords and MeSH terms: ‘gastric adenocarcinoma’, ‘stomach adenocarcinoma’, ‘gastric cancer’, ‘stomach cancer’, ‘gastric malignancy’, ‘stomach malignancy’, ‘gastric neoplasm’ and ‘stomach neoplasm’. We also used ‘radiotherapy’, ‘adjuvant radiotherapy’, ‘radiation therapy’, ‘adjuvant radiation therapy’, ‘chemoradiotherapy’, ‘adjuvant chemoradiotherapy’, ‘chemoradiation therapy’, ‘adjuvant chemoradiation therapy’, ‘chemoradiation’ and ‘chemoirradiation’. They were limited to
80
Oncology 2014;86:79–85 DOI: 10.1159/000354641
only randomized, controlled clinical trials with human subjects. No language or date restrictions were applied. Criteria for Considering Studies for this Review Eligible studies had to fulfill specific criteria to be entered into this analysis. Only randomized, controlled clinical trials involving patients with histologically confirmed adenocarcinoma of the stomach status-post curative resection with negative surgical margins were included. We compared two postoperative interventions, adjuvant chemoradiation therapy and chemotherapy alone. Any chemotherapy regimen was allowed. We excluded studies that involved adjuvant radiotherapy versus chemotherapy, as well as studies in the unresectable or palliative settings. Outcome The primary outcomes of interest were disease-free survival, defined as survival without local or distant relapse, and overall survival. Data Collection and Analysis Studies were only selected if they were randomized, controlled clinical trials. For each study, data was extracted from the full manuscript based on number of events in each treatment group. Events were defined as disease-free and overall survival failures. Individual patient records were not reviewed. Quality Analysis Three authors (C.M., S.J., K.D.) independently assessed trial eligibility and bias risk, and the same authors extracted data. Disagreements were resolved by consensus. The bias risk of the trials was assessed based on the components recommended by the Cochrane Collaboration: sequence generation of allocation; allocation concealment; blinding of participants, personnel, and outcome assessors; incomplete outcome data; selective outcome reporting, and other sources of bias [14]. Of note, it was not possible to conceal allocation or blind participants in the trials we included. Trials with high or unclear risk for bias for any one of the first three components were considered as trials with high risk of bias. Otherwise, they were considered as low risk of bias trials. Statistical Analysis An intention-to-treat meta-analysis was performed in line with recommendations from the Cochrane Collaboration and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (fig. 1) [15]. Standard software was used (Review Manager, RevMan, version 5.2; The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, Denmark). Risk of bias was assessed and agreed upon by the authors and estimated visually by funnel plots (fig. 2). In general, larger studies were close to the average, while smaller studies spread on both sides of the average. Variation from this assumption may represent publication bias. Heterogeneity was assessed using I2 statistics, with I2 representing the proportion of total variation observed between studies attributable to differences between studies rather than sampling error. I2 < 25% was considered low, while I2 > 75% was considered high. The Mantel-Haenszel (M-H) random effects model was used to calculate effect sizes [16]. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.
Min/Bangalore/Jhawar/Guo/Nicholson/ Formenti/Leichman/Du
Downloaded by: Selçuk Universitesi 193.255.248.150 - 1/7/2015 1:33:23 PM
high, with reported rates greater than 50% [6]. Overall survival is poor with 5- to 10-year rates ranging from 10 to 25% [7]. With these outcomes, postoperative chemoradiation and perioperative chemotherapy have been successfully employed to increase cure rates. The landmark Intergroup 0116 study established the role of postoperative chemoradiation in the USA. In this trial, Macdonald et al. [8] demonstrated a benefit in overall survival with a regimen of adjuvant fluorouracil (5-FU), leucovorin and radiation (to a total dose of 45 Gy) when compared to surgery alone. A recently published update confirmed that these results persist at a median follow-up of more than 10 years [9]. The United Kingdom Medical Research Council Adjuvant Gastric Infusional Chemotherapy [10] trial evaluated epirubicin, cisplatin and 5-FU administered perioperatively. They found an improvement in both progression-free (hazard ratio for progression 0.66, 95% confidence interval 0.53–0.81, p < 0.001) and overall survival (hazard ratio for death 0.75, 95% confidence interval 0.60–0.93, p = 0.009) when compared to surgery alone [10]. In the adjuvant setting, two prospectively randomized phase III trials have shown efficacy for postoperative chemotherapy. One study used single agent S-1, an oral fluoropyrimidine that is not available in the USA, and the other the combination of capecitabine and oxaliplatin [11, 12]. Current practice guidelines for North America support the use of either chemotherapy alone or chemoradiation and recommend treatment choice on a case-bycase basis [13]. In the postoperative patient, combined chemoradiation may increase complexity, cost and toxicity when compared to either modality alone. Thus, we undertook this meta-analysis to investigate whether adjuvant chemoradiation therapy is associated with more favorable oncologic outcomes when compared to chemotherapy alone for patients who have undergone a curative resection for gastric cancer.
Results
Full abstracts assessed for eligibility (n = 498) Records excluded on the basis of abstract (n = 491) Full-text articles retrieved for detailed evaluation (n = 7) Trials excluded (n = 1; not randomized) Trials included in final metaanalysis (n = 6)
Fig. 1. Trial selection. RCT = Randomized, controlled trials.
0 0.2 0.4 0.6 0.8 1.0 0.01
0.1
1 Odds ratio
10
100
Fig. 2. Funnel plot: chemotherapy compared with chemoradiation.
SE = Standard error.
Effects of Interventions The total number of events, defined as disease-free survival failure, in the chemotherapy group was 275 (48.1%), while the total number of events in the chemoradiation group was 240 (40.1%). Adjuvant therapy with
chemoradiation was associated with a significant increase in disease-free survival when compared to chemotherapy alone (odds ratio 1.48, 95% confidence interval 1.08–2.03, p = 0.01). There was moderate heterogeneity (I2 = 30%) in the analysis (fig. 4). Overall survival was reported in five of the six studies assessed. The total number of events, defined as overall survival failure, in the chemotherapy group was 183 (53.2%), while the total number of events in the chemoradiation group was 175 (47.4%). There was no statistically significant difference in overall survival between the
A Meta-Analysis of Chemoradiation Therapy for Gastric Cancer
Oncology 2014;86:79–85 DOI: 10.1159/000354641
81
Downloaded by: Selçuk Universitesi 193.255.248.150 - 1/7/2015 1:33:23 PM
Risk of Bias in Included Studies All trials had high risk of bias because it was impossible to achieve allocation concealment or blinding of participants. On the other hand, there was likely a low risk of selection, attrition and reporting bias (fig. 3).
Records excluded on the basis of title (n = 2,302)
SE (log [OR])
Description of Studies The search resulted in approximately 2,800 titles, which were further narrowed to 498 abstracts and titles. Seven clinical trials comparing chemoradiation versus chemotherapy alone were retrieved. One of these trials did not specifically state whether the patients were randomized. Upon contacting the first author of this study he confirmed that it was not a randomized trial; therefore, this study was excluded. The other six were included and are described below (see table 1). Three of the trials were Korean, two were Chinese and the last was Greek. There were a total of 1,171 eligible patients, with 599 randomized to chemoradiation therapy and 572 to chemotherapy alone. All patients had histologically confirmed adenocarcinoma of the stomach and were statuspost curative resection, with negative surgical margins. Stages ranged from IIA to IIIC based on the 7th edition of the American Joint Committee on Cancer Staging System; patients with metastatic disease were excluded. On study entry, all patients were required to have an Eastern Cooperative Oncology Group performance status
Oncology 2014;86:79–85 DOI: 10.1159/000354641
Chemoradiation Therapy versus Chemotherapy Alone for Gastric Cancer after R0 Surgical Resection: A Meta-Analysis of Randomized Trials Christine Min Sripal Bangalore Sachin Jhawar Yu Guo Joseph Nicholson Silvia C. Formenti Lawrence P. Leichman Kevin L. Du New York University School of Medicine, New York, N.Y. USA
Abstract Objective: Current national guidelines include category 1 recommendations for perioperative chemotherapy or adjuvant chemoradiation with surgical resection for patients with stage IB–IIIB gastric cancer. We conducted a meta-analysis of randomized trials in which chemotherapy was prospectively tested against chemoradiation with surgical resection. Methods: We electronically searched PubMed and EMBASE for randomized, controlled clinical trials involving patients with gastric adenocarcinoma, status post-R0 resection. The interventions compared were adjuvant chemotherapy versus chemoradiation, with any chemotherapy regimen. The primary outcomes of interest were diseasefree survival and overall survival. The Mantel-Haenszel random-effects model was used to calculate effect sizes. Results: Six trials that included 1,171 patients were evaluated; 599 were randomized to adjuvant chemoradiation and 572 to chemotherapy alone. Chemoradiation was associated with a significant increase in disease-free survival (odds ratio 1.48, 95% confidence interval 1.08–2.03) when compared to chemotherapy alone. However, there was no significant difference in overall survival (odds ratio 1.27, 95% confidence
© 2014 S. Karger AG, Basel 0030–2414/14/0862–0079$39.50/0 E-Mail [email protected] www.karger.com/ocl
interval 0.95–1.71). Five trials found no statistically significant differences in toxicities between the two groups. Conclusion: In patients with gastric cancer status post-R0 resection, adjuvant chemoradiation was associated with higher disease-free survival when compared to chemotherapy alone. It remains appropriate to design trials testing new systemic agents with radiotherapy. © 2014 S. Karger AG, Basel
Introduction
Gastric cancer is the third most common cancer and the second leading cause of cancer-related deaths in the world. Approximately 21,000 new cases are diagnosed each year with nearly 11,000 deaths [1]. The incidence, etiologically linked to chronic Helicobacter pylori infection and to dietary intake of highly salted and smoked food, is highest in East Asian countries [2–3]. Although the incidence in the West has declined over the past 50 years, the notable exception is noncardia gastric cancers that frequently harbor a diffuse histology and are found in younger patients [4–5]. Partial or total gastrectomy is the mainstay of curative therapy for patients with gastric cancer. However, even with an aggressive D2 lymph node resection (removal of perigastric, celiac, splenic artery and splenic hilar lymph nodes), the risk of local and distant recurrence remains Christine Min, MD New York University School of Medicine 160 East 34th Street LL1 New York, NY 10016 (USA) E-Mail christinemin @ jhmi.edu
Downloaded by: Selçuk Universitesi 193.255.248.150 - 1/7/2015 1:33:23 PM
Key Words Chemoradiation · Gastric adenocarcinoma · Gastric cancer · Gastric carcinoma · Radiation therapy · Radiochemotherapy · Radiotherapy · Randomized trial
Methods Search Methods for Identification of Studies Electronic searches were conducted on PubMed and EMBASE with the following keywords and MeSH terms: ‘gastric adenocarcinoma’, ‘stomach adenocarcinoma’, ‘gastric cancer’, ‘stomach cancer’, ‘gastric malignancy’, ‘stomach malignancy’, ‘gastric neoplasm’ and ‘stomach neoplasm’. We also used ‘radiotherapy’, ‘adjuvant radiotherapy’, ‘radiation therapy’, ‘adjuvant radiation therapy’, ‘chemoradiotherapy’, ‘adjuvant chemoradiotherapy’, ‘chemoradiation therapy’, ‘adjuvant chemoradiation therapy’, ‘chemoradiation’ and ‘chemoirradiation’. They were limited to
80
Oncology 2014;86:79–85 DOI: 10.1159/000354641
only randomized, controlled clinical trials with human subjects. No language or date restrictions were applied. Criteria for Considering Studies for this Review Eligible studies had to fulfill specific criteria to be entered into this analysis. Only randomized, controlled clinical trials involving patients with histologically confirmed adenocarcinoma of the stomach status-post curative resection with negative surgical margins were included. We compared two postoperative interventions, adjuvant chemoradiation therapy and chemotherapy alone. Any chemotherapy regimen was allowed. We excluded studies that involved adjuvant radiotherapy versus chemotherapy, as well as studies in the unresectable or palliative settings. Outcome The primary outcomes of interest were disease-free survival, defined as survival without local or distant relapse, and overall survival. Data Collection and Analysis Studies were only selected if they were randomized, controlled clinical trials. For each study, data was extracted from the full manuscript based on number of events in each treatment group. Events were defined as disease-free and overall survival failures. Individual patient records were not reviewed. Quality Analysis Three authors (C.M., S.J., K.D.) independently assessed trial eligibility and bias risk, and the same authors extracted data. Disagreements were resolved by consensus. The bias risk of the trials was assessed based on the components recommended by the Cochrane Collaboration: sequence generation of allocation; allocation concealment; blinding of participants, personnel, and outcome assessors; incomplete outcome data; selective outcome reporting, and other sources of bias [14]. Of note, it was not possible to conceal allocation or blind participants in the trials we included. Trials with high or unclear risk for bias for any one of the first three components were considered as trials with high risk of bias. Otherwise, they were considered as low risk of bias trials. Statistical Analysis An intention-to-treat meta-analysis was performed in line with recommendations from the Cochrane Collaboration and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (fig. 1) [15]. Standard software was used (Review Manager, RevMan, version 5.2; The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, Denmark). Risk of bias was assessed and agreed upon by the authors and estimated visually by funnel plots (fig. 2). In general, larger studies were close to the average, while smaller studies spread on both sides of the average. Variation from this assumption may represent publication bias. Heterogeneity was assessed using I2 statistics, with I2 representing the proportion of total variation observed between studies attributable to differences between studies rather than sampling error. I2 < 25% was considered low, while I2 > 75% was considered high. The Mantel-Haenszel (M-H) random effects model was used to calculate effect sizes [16]. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.
Min/Bangalore/Jhawar/Guo/Nicholson/ Formenti/Leichman/Du
Downloaded by: Selçuk Universitesi 193.255.248.150 - 1/7/2015 1:33:23 PM
high, with reported rates greater than 50% [6]. Overall survival is poor with 5- to 10-year rates ranging from 10 to 25% [7]. With these outcomes, postoperative chemoradiation and perioperative chemotherapy have been successfully employed to increase cure rates. The landmark Intergroup 0116 study established the role of postoperative chemoradiation in the USA. In this trial, Macdonald et al. [8] demonstrated a benefit in overall survival with a regimen of adjuvant fluorouracil (5-FU), leucovorin and radiation (to a total dose of 45 Gy) when compared to surgery alone. A recently published update confirmed that these results persist at a median follow-up of more than 10 years [9]. The United Kingdom Medical Research Council Adjuvant Gastric Infusional Chemotherapy [10] trial evaluated epirubicin, cisplatin and 5-FU administered perioperatively. They found an improvement in both progression-free (hazard ratio for progression 0.66, 95% confidence interval 0.53–0.81, p < 0.001) and overall survival (hazard ratio for death 0.75, 95% confidence interval 0.60–0.93, p = 0.009) when compared to surgery alone [10]. In the adjuvant setting, two prospectively randomized phase III trials have shown efficacy for postoperative chemotherapy. One study used single agent S-1, an oral fluoropyrimidine that is not available in the USA, and the other the combination of capecitabine and oxaliplatin [11, 12]. Current practice guidelines for North America support the use of either chemotherapy alone or chemoradiation and recommend treatment choice on a case-bycase basis [13]. In the postoperative patient, combined chemoradiation may increase complexity, cost and toxicity when compared to either modality alone. Thus, we undertook this meta-analysis to investigate whether adjuvant chemoradiation therapy is associated with more favorable oncologic outcomes when compared to chemotherapy alone for patients who have undergone a curative resection for gastric cancer.
Results
Full abstracts assessed for eligibility (n = 498) Records excluded on the basis of abstract (n = 491) Full-text articles retrieved for detailed evaluation (n = 7) Trials excluded (n = 1; not randomized) Trials included in final metaanalysis (n = 6)
Fig. 1. Trial selection. RCT = Randomized, controlled trials.
0 0.2 0.4 0.6 0.8 1.0 0.01
0.1
1 Odds ratio
10
100
Fig. 2. Funnel plot: chemotherapy compared with chemoradiation.
SE = Standard error.
Effects of Interventions The total number of events, defined as disease-free survival failure, in the chemotherapy group was 275 (48.1%), while the total number of events in the chemoradiation group was 240 (40.1%). Adjuvant therapy with
chemoradiation was associated with a significant increase in disease-free survival when compared to chemotherapy alone (odds ratio 1.48, 95% confidence interval 1.08–2.03, p = 0.01). There was moderate heterogeneity (I2 = 30%) in the analysis (fig. 4). Overall survival was reported in five of the six studies assessed. The total number of events, defined as overall survival failure, in the chemotherapy group was 183 (53.2%), while the total number of events in the chemoradiation group was 175 (47.4%). There was no statistically significant difference in overall survival between the
A Meta-Analysis of Chemoradiation Therapy for Gastric Cancer
Oncology 2014;86:79–85 DOI: 10.1159/000354641
81
Downloaded by: Selçuk Universitesi 193.255.248.150 - 1/7/2015 1:33:23 PM
Risk of Bias in Included Studies All trials had high risk of bias because it was impossible to achieve allocation concealment or blinding of participants. On the other hand, there was likely a low risk of selection, attrition and reporting bias (fig. 3).
Records excluded on the basis of title (n = 2,302)
SE (log [OR])
Description of Studies The search resulted in approximately 2,800 titles, which were further narrowed to 498 abstracts and titles. Seven clinical trials comparing chemoradiation versus chemotherapy alone were retrieved. One of these trials did not specifically state whether the patients were randomized. Upon contacting the first author of this study he confirmed that it was not a randomized trial; therefore, this study was excluded. The other six were included and are described below (see table 1). Three of the trials were Korean, two were Chinese and the last was Greek. There were a total of 1,171 eligible patients, with 599 randomized to chemoradiation therapy and 572 to chemotherapy alone. All patients had histologically confirmed adenocarcinoma of the stomach and were statuspost curative resection, with negative surgical margins. Stages ranged from IIA to IIIC based on the 7th edition of the American Joint Committee on Cancer Staging System; patients with metastatic disease were excluded. On study entry, all patients were required to have an Eastern Cooperative Oncology Group performance status
