Capecitabine and cisplatin with or without cetuximab for patients with previously untreated advanced gastric cancer (EXPAND): a randomised, open-label phase 3 trial

Articles

Capecitabine and cisplatin with or without cetuximab for patients with previously untreated advanced gastric cancer (EXPAND): a randomised, open-label phase 3 trial Florian Lordick, Yoon-Koo Kang, Hyun-Cheol Chung, Pamela Salman, Sang Cheul Oh, György Bodoky, Galina Kurteva, Constantin Volovat, Vladimir M Moiseyenko, Vera Gorbunova, Joon Oh Park, Akira Sawaki, Ilhan Celik, Heiko Götte, Helena Melezínková, Markus Moehler, on behalf of the Arbeitsgemeinschaft Internistische Onkologie (AIO) and EXPAND Investigators*

Summary Lancet Oncol 2013; 14: 490–99 Published Online April 15, 2013 http://dx.doi.org/10.1016/ S1470-2045(13)70102-5 See Comment page 440 *Members listed in the appendix University Cancer Center Leipzig, University of Leipzig, Leipzig, Germany (Prof F Lordick MD); Asan Medical Center, Division Oncology Department, Seoul, Korea (Prof Y-K Kang MD); Yonsei Cancer Center, Yonsei University College of Medicine, Seoul, South Korea (Prof H-C Chung MD); Fundación Arturo López Pérez, Santiago, Chile (P Salman MD); Department of Oncology, Korea University Guro Hospital, Seoul, South Korea (Prof S C Oh MD); Department of Oncology, St László Hospital, Budapest, Hungary (Prof G Bodoky MD); Specialized Hospital for Active Treatment, Center of Oncology, Sofia, Bulgaria (G Kurteva MD); Centrul de Oncologie Medicala, Iasi, Romania (C Volovat MD); N N Petrov Research Institute of Oncology, St Petersburg, Russia (Prof V M Moiseyenko MD); Russian Cancer Centre, Moscow, Russia (Prof V Gorbunova MD); Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, South Korea (J O Park MD); Aichi Cancer Center Hospital, Nagoya, Japan (A Sawaki MD); Merck KGaA, Darmstadt, Germany (Prof I Celik MD, H Götte PhD, H Melezínková MD); and University Medical Center of the Johannes Gutenberg University Mainz, First Department of Internal Medicine, Mainz, Germany (Prof M Moehler MD)

490

Background Patients with advanced gastric cancer have a poor prognosis and few efficacious treatment options. We aimed to assess the addition of cetuximab to capecitabine-cisplatin chemotherapy in patients with advanced gastric or gastro-oesophageal junction cancer. Methods In our open-label, randomised phase 3 trial (EXPAND), we enrolled adults aged 18 years or older with histologically confirmed locally advanced unresectable (M0) or metastatic (M1) adenocarcinoma of the stomach or gastro-oesophageal junction. We enrolled patients at 164 sites (teaching hospitals and clinics) in 25 countries, and randomly assigned eligible participants (1:1) to receive first-line chemotherapy with or without cetuximab. Randomisation was done with a permuted block randomisation procedure (variable block size), stratified by disease stage (M0 vs M1), previous oesophagectomy or gastrectomy (yes vs no), and previous (neo)adjuvant (radio) chemotherapy (yes vs no). Treatment consisted of 3-week cycles of twice-daily capecitabine 1000 mg/m² (on days 1–14) and intravenous cisplatin 80 mg/m² (on day 1), with or without weekly cetuximab (400 mg/m² initial infusion on day 1 followed by 250 mg/m² per week thereafter). The primary endpoint was progression-free survival (PFS), assessed by a masked independent review committee in the intention-to-treat population. We assessed safety in all patients who received at least one dose of study drug. This study is registered at EudraCT, number 2007-004219-75. Findings Between June 30, 2008, and Dec 15, 2010, we enrolled 904 patients. Median PFS for 455 patients allocated capecitabine-cisplatin plus cetuximab was 4·4 months (95% CI 4·2–5·5) compared with 5·6 months (5·1–5·7) for 449 patients who were allocated to receive capecitabine-cisplatin alone (hazard ratio 1·09, 95% CI 0·92–1·29; p=0·32). 369 (83%) of 446 patients in the chemotherapy plus cetuximab group and 337 (77%) of 436 patients in the chemotherapy group had grade 3–4 adverse events, including grade 3–4 diarrhoea, hypokalaemia, hypomagnesaemia, rash, and hand-foot syndrome. Grade 3–4 neutropenia was more common in controls than in patients who received cetuximab. Incidence of grade 3–4 skin reactions and acne-like rash was substantially higher in the cetuximabcontaining regimen than in the control regimen. 239 (54%) of 446 in the cetuximab group and 194 (44%) of 436 in the control group had any grade of serious adverse event. Interpretation Addition of cetuximab to capecitabine-cisplatin provided no additional benefit to chemotherapy alone in the first-line treatment of advanced gastric cancer in our trial. Funding Merck KGaA.

Introduction In 2008, gastric cancer was the second most common cause of death related to cancer worldwide (9·7%).1 Although the overall incidence of gastric cancer is decreasing, the incidence of gastro-oesophageal junction tumours, which have an especially poor prognosis, is increasing. Complete surgical resection of early disease offers a chance for cure, although most patients (80–90%) in high-income non-Asian countries present with either advanced disease or develop a recurrence within 5 years of undergoing curative resection.2 Treatment with combination chemotherapy improves outcomes compared with single-drug chemotherapy or no chemotherapy in patients with advanced gastric cancer.3 With no international standard at present,

first-line chemotherapy regimens commonly used in treatment include a platinum compound (typically cisplatin) in combination with either infusional fluorouracil or an oral fluoropyrimidine (typically capecitabine) or S-1, which is a preferred treatment in Asia.4–6 Despite the development of new chemotherapy regimens, the 5-year survival of patients in this setting is less than 10%, and therefore a high and unmet need exists for efficacious treatment.2 Introduction of biological drugs into treatment strategies has shown promise in this setting. Trastuzumab, a HER2 antibody, in combination with first-line cisplatin and fluoropyrimidine-based chemotherapy significantly improved overall survival in patients with HER2-overexpressing tumours compared with those receiving www.thelancet.com/oncology Vol 14 May 2013

Articles

chemotherapy alone.7 EGFR, part of the family of receptor tyrosine kinases including HER2, is expressed in gastric and oesophageal tumours and is associated with poor prognosis.8,9 Addition of cetuximab (an EGFR antibody) to standard first-line chemotherapy regimens improved clinical outcome in patients with KRAS wild-type metastatic colorectal cancer,10,11 recurrent or metastatic squamous-cell carcinoma of the head and neck,12 and advanced non-small-cell lung cancer.13 Moreover, manageable and expected safety profiles with substantial activity (objective response rates of 41–65%) were reported in phase 2 studies of cetuximab plus various first-line chemotherapy regimens in patients with advanced gastric cancer.14–17 In the randomised international Erbitux (cetuximab) in combination with Xeloda (capecitabine) and cisplatin in advanced esophago-gastric cancer (EXPAND) study, we aimed to assess efficacy and safety of addition of cetuximab to first-line capecitabine-cisplatin chemotherapy in patients with unresectable advanced or metastatic gastric adenocarcinoma.

or C; chronic diarrhoea or short bowel syndrome; pregnancy or lactation; legal incapacity or limited legal capacity; and contraindications to study treatment. The study was done according to the principles of the Declaration of Helsinki and the International Conference on Harmonisation on Good Clinical Practice and was approved by local or national ethics committees or institutional review boards. All patients provided written informed consent. A separate data and safety monitoring board, with three independent members in Switzerland, Singapore, and Germany, reviewed study data.

In this open-label study, patients were randomly assigned (1:1) to receive capecitabine-cisplatin with or without cetuximab. Randomisation was done centrally with an interactive voice response system. We used a stratified, permuted, block randomisation procedure (variable block size) with the following strata: disease stage (M0 vs M1),

Study design and participants

www.thelancet.com/oncology Vol 14 May 2013

See Online for appendix

Randomisation and masking

Methods In this open-label, randomised, controlled, phase 3 study, we enrolled adults (aged ≥18 years) with histologically confirmed adenocarcinoma of the stomach or gastro-oesophageal junction with locally advanced unresectable (M0) or metastatic (M1) disease. Patients were enrolled at 164 sites (teaching hospitals and clinics) in 25 countries worldwide (appendix). Other inclusion criteria were availability of tumour material for EGFR expression assessment; at least one radiographically documented measurable lesion (≥2 cm in at least one dimension by conventional techniques or ≥1 cm by spiral CT) in a previously non-irradiated area according to Response Evaluation Criteria in Solid Tumors (RECIST; version 1);18 an Eastern Cooperative Oncology Group (ECOG) performance status of 0–1 with adequate organ function; no previous chemotherapy for metastatic or locally advanced unresectable gastric or gastrooesophageal junction cancer; adjuvant chemotherapy completed at least 1 year before randomisation and not more than 300 mg/m² cisplatin administered; no previous treatment with drugs targeting EGFR-related or VEGFR-related signalling pathways; and no clinically relevant coronary artery disease, congestive heart failure, cardiomyopathy, history of myocardial infarction in the last 12 months, or high risk of uncontrolled arrhythmia. Exclusion criteria were known brain metastasis or leptomeningeal disease; other malignant disease in the previous 5 years, apart from basal-cell cancer of the skin or pre-invasive cancer of the cervix uteri; radiotherapy (apart from limited volume radiotherapy for pain relief), major surgery, or any investigational drug within 30 days before the start of study treatment; concurrent chronic systemic immune or hormone therapy; active hepatitis B

Correspondence to: Prof Florian Lordick, University Clinic Leipzig, University Cancer Center Leipzig (UCCL), 04301 Leipzig, Germany fl[email protected]

Sex, male

Capecitabinecisplatin plus cetuximab group (n=455)

Capecitabinecisplatin group (n=449)

339 (75%)

334 (74%)

Age, years Median (range, IQR) ≥65

60 (23–84, 51–66)

59 (18–81, 53–66)

139 (31%)

141 (31%)

0

237 (52%)

228 (51%)

1

218 (48%)

220 (49%)

White

242 (53%)

248 (55%)

Asian

173 (38%)

166 (37%)

Other

40 (9%)

35 (8%)

376 (83%)

371 (83%)

71 (16%)

73 (16%)

8 (2%)

5 (1%)

ECOG performance status

Ethnic group

Primary site Stomach Gastro-oesophageal junction Unknown Histological type (Lauren) Intestinal

162 (36%)

149 (33%)

Diffuse

76 (17%)

94 (21%)

Mixed

24 (5%)

26 (6%)

193 (42%)

180 (40%)

113 (25%)

116 (26%)

Unknown Peritoneal metastasis Disease status* Locally advanced

16 (4%)

12 (3%)

439 (96%)

436 (97%)

Previous oesophagectomy or gastrectomy*

92 (20%)

90 (20%)

Previous (neo)adjuvant chemo(radio)therapy*

28 (6%)

27 (6%)

Metastatic

Data are n (%), unless otherwise stated. *Stratification factors.

Table 1: Baseline characteristics

491

Articles

previous oesophagectomy or gastrectomy (yes vs no), and previous (neo)adjuvant (radio)chemotherapy (yes vs no).

Procedures We treated patients with 3-week cycles of chemotherapy (oral capecitabine 1000 mg/m² twice daily from the evening of day 1 until the morning of day 15 plus intravenous cisplatin 80 mg/m² on day 1) with or without once-weekly cetuximab (400 mg/m² at the first infusion then 250 mg/m² every week), dependent on randomisation assignment. Treatment was continued until radiographically documented tumour progression, unacceptable toxicity, or withdrawal of consent by the patient. In the experimental group (capecitabine and cisplatin plus cetuximab), patients who discontinued chemotherapy could continue on cetuximab monotherapy and those who discontinued cetuximab could continue chemotherapy, until progressive disease, unacceptable toxicity, or withdrawal of consent. Dose reductions were specified in the protocol: 2-week therapy delay for toxic effects was allowed for all study drugs; for grade 3 skin reactions 1191 patients screened

297 ineligible* 229 violated inclusion or exclusion criteria 18 withdrew consent 2 adverse events 1 death 1 lost to follow-up 1 symptomatic deterioration 44 other† 1 reason missing 287 discontinued

904 randomly allocated treatment (intention-to-treat population)

455 allocated capecitabine-cisplatin plus cetuximab

449 allocated capecitabine-cisplatin

9 did not receive ≥1 dose and were excluded from the safety analysis

13 did not receive ≥1 dose and were excluded from the safety analysis

447 discontinued treatment 247 disease progression 74 adverse events 24 deaths 41 withdrawal of consent 31 symptomatic deterioration 4 non-compliance 2 lost to follow-up 24 other

444 discontinued treatment 223 disease progression 75 adverse events 24 deaths 51 withdrawal of consent 22 symptomatic deterioration 10 non-compliance 1 lost to follow-up 38 other

455 included in the intention-to-treat analyses 446 included in the safety analyses

449 included in the intention-to-treat analyses 436 included in the safety analyses

Figure 1: Trial profile *Ten of 297 ineligible patients were randomised to treatment. †Most patients in this category were ineligible because recruitment was put on hold.

492

cetuximab dose reductions to 200 mg/m² were permitted, with further reduction to 150 mg/m² for repeated grade 3 skin reactions; capecitabine dose reduction of up to 50% of the starting dose was allowed depending on severity and frequency of adverse drug reactions; cisplatin dose reduction to 25% of the starting dose was allowed for specific haematological toxic effects. We assessed response according to RECIST guidelines. We confirmed complete or partial responses at least 4 weeks after the initial response measurement. We did radiological imaging of the chest and abdomen at baseline and every 6 weeks from the date of randomisation until progressive disease occurrence. Date of progression and best overall response were assessed by masked review at an independent review committee (IRC) The IRC consisted of external experts from Germany, contracted by Perceptive Informatics. An oncologist at the IRC reviewed clinical documentation of progressive disease. We assessed safety in all patients who received at least one dose of any study treatment. We coded adverse events according to the Medical Dictionary for Regulatory Activities (MedDRA; version 14.1). We graded severity with the National Cancer Institute-Common Terminology Criteria for Adverse Events (NCI-CTCAE; version 3). We defined special adverse event categories of interest and composite categories of preferred terms according to MedDRA version 14.1; definitions of each special adverse event category are provided in the appendix. We determined tumour HER2 status according to a scoring system for gastric cancer,19 which included immunohistochemistry with a Dako HercepTest kit (Glostrup, Denmark) and fluorescence in-situ hybridisation (FISH) with a Dako HER2 FISH pharmDx kit. We retrospectively assessed tumour EGFR expression analysis by immunohistochemistry with a Dako EGFR pharmDx kit (Glostrup, Denmark). We used immunohistochemistry data to generate an EGFR immunohistochemistry score on a continuous scale of 0–300 as described previously (appendix).20

Statistical analysis Our main objective was to test the null hypothesis that IRC-adjudicated PFS—the primary endpoint—was equal between the experimental and capecitabine-cisplatin alone (control) groups. The primary analysis was done in the intention-to-treat population of all patients randomly allocated a study treatment. We used a stratified, twosided log-rank test to compare the treatment groups, accounting for the randomisation strata. We required 631 PFS events to ensure 80% power for rejection of the null hypothesis of equal treatment effects at a two-sided significance level of 5%, assuming a hazard ratio (HR) of 0·80 and 1:1 randomisation. We planned to enrol 870 patients to be randomised during an accrual period of 27 months and expected 631 PFS events after an additional 7 months of follow-up, assuming a median PFS time of 5·6 months with capecitabine-cisplatin alone.16 www.thelancet.com/oncology Vol 14 May 2013

Articles

Secondary objectives included assessment of overall survival, best overall tumour response (adjudicated by IRC), and safety. We compared secondary efficacy endpoints between the treatment groups with two-sided stratified tests, with stratification factors as for the primary endpoint. We compared best overall response between the treatment groups with a stratified Cochran-MantelHaenszel test. We defined PFS as the time from randomisation to either first observation of progressive disease or occurrence of death due to any cause within 60 days of the last tumour assessment or randomisation. Censoring rules for PFS are provided in the appendix. Overall survival was defined as the time from randomisation to death. To adjust for multiplicity, we applied a ranking of efficacy endpoints according to clinical relevance: PFS, followed by overall survival, followed by best overall response. With this hierarchical ordering, all p values can be tested at the 5% level. A hypothesis can only be rejected at the 5% level if the hypotheses with higher rank are rejected. Because of the hierarchical ordering, the familywise error rate is protected without formal adjustment of the p values for multiplicity. Ignoring the ranked hypotheses, at least 631 deaths were necessary to detect an HR of 0·8 with 80% power in terms of overall survival. Review of PFS projection data during the study suggested that the required number of IRC-confirmed PFS events (631 events) for the primary analysis might not be reached because patients went off study without radiologically confirmed progression. However, 631 deaths were likely to be reached. In response, we modified the timing of statistical analyses in a protocol amendment (from Oct 31, 2011) by fixing the cutoff date for statistical analyses to allow the PFS analysis to be done with fewer events, accepting some loss in power. Therefore final PFS and survival analyses were done in parallel on data obtained from the clinical cutoff date, determined when 631 PFS events were reported by the IRC, or March 31, 2012, which ever occurred first. We correlated tumour HER2 status (a predefined subgroup), EGFR immunohistochemistry score, and the presence of first-cycle grade 1–4 acne-like rash (exploratory landmark analysis,11 not prespecified in the protocol) with outcomes. We planned to generate a scatter plot showing PFS and overall survival time for each patient plotted against EGFR immunohistochemistry score. We assessed associations between clinical outcome and EGFR immunohistochemistry score with a subpopulation treatment-effect pattern-plot method.20,21 Statistical analyses were done with SAS version 9.1.3 or later and R version 2.14.1. This study is registered with EudraCT, number 2007-004219-75.

Role of the funding source The sponsor was responsible for data management and statistical analysis. The EXPAND clinical study was designed by Merck KGaA in collaboration with the www.thelancet.com/oncology Vol 14 May 2013

coordinating investigator (FL), the Arbeitsgemeinschaft Internistische Onkologie, and an international expert advisory panel (funded by Merck KGaA). This analysis and data interpretation was done by the sponsor in cooperation with the authors. The drafting of the manuscript was commissioned by the sponsor. All authors had full access to all study data and FL had the final responsibility for the decision to submit for publication.

Results Between June 30, 2008, and Dec 15, 2010, we enrolled 904 patients (table 1, figure 1). Recruitment was temporarily suspended in July 6, 2009, after 380 patients had been allocated treatment, because of an imbalance in cardiac events detected by the data and safety monitoring board. The study was restarted in Dec 11, 2009 after implementation of a cardiac monitoring programme (appendix). Baseline characteristics were generally well balanced (table 1). Only 16 (4%) of 455 patients in the experimental group and 12 (3%) of 449 patients in the control group had locally advanced unresectable disease, as assessed locally by the investigator.

Capecitabine-cisplatin plus cetuximab group (n=446)

Capecitabine-cisplatin group (n=436)

Cetuximab Duration of treatment, months Cumulative dose, mg/m²

3·4 (1·6–5·6) 3403 (1660–5654)

·· ··

Initial dose (400 mg/m²) Dose intensity, mg/m² per week

400 (397–401)

··

Relative dose intensity 80–<90% ≥90%

1 (<1%)

··

437 (98%)

··

236 (215–249)

··

Post-initial dose* (250 mg/m²) Dose intensity, mg/m² per week Relative dose intensity 80–<90% ≥90%

99 (22%)

··

267 (60%)

··

Cisplatin Duration of treatment, months Cumulative dose, mg/m² Dose intensity, mg/m² per 3 weeks

3·3 (1·6–4·8)

3·5 (1·5–5·0)

321 (160–479)

320 (160–476)

73 (66–79)

70 (62–79)

Relative dose intensity 80–<90%

126 (28%)

108 (25%)

≥90%

231 (52%)

193 (44%)

Capecitabine Duration of treatment, months Cumulative dose, mg/m² Dose intensity, mg/m² per 3 weeks

3·4 (1·7–5·4)

3·9 (1·6–6·0)

109 170 (56 247–167 705)

115 168 (55 186–180 488)

22 904 (19 178–26 373)

22 271 (18 598–25 790)

Relative dose intensity 80–<90%

101 (23%)

89 (20%)

≥90%

140 (31%)

122 (28%)

Data are median (IQR) or n (%) unless otherwise stated. *Initial cetuximab administration excluded.

Table 2: Treatment exposure

493

Articles

A 100 90

Capecitabine-cisplatin plus Capecitabine-cisplatin cetuximab group (n=455) group (n=449)

Capecitabine-cisplatin plus cetuximab (286 events; median progression-free survival 4·4 months [95% CI 4·2–5·5]) Capecitabine-cisplatin (269 events; median progression-free survival 5·6 months [95% CI 5·1–5·7])

2 (<1%)

2 (<1%)

Partial response

134 (29%)

129 (29%)

Stratified HR 1·09 (95% CI 0·92–1·29; p=0·32)

Stable disease

Complete response

Progression-free survival (%)

80 70 60 50 40 30

61 (14%)

Unknown*

58 (13%)

71 (16%)

Overall response†

136 (30%, 26–34)

131 (29%, 25–34)

Disease control

332 (73%, 69–77)

317 (71%, 66–75)

Table 3: Best overall response

10 0 0

3

6

9

Number at risk Capecitabine-cisplatin 455 233 94 44 plus cetuximab Capecitabine-cisplatin 449 244 116 50

12

15

18

21

24

27

30

33

36

39

42

30

20

14

8

4

4

3

3

1

1

0

29

14

10

4

4

4

4

2

0

0

0

B Capecitabine-cisplatin plus cetuximab (362 events; median overall survival 9·4 months [95% CI 8·3–10·6]) Capecitabine-cisplatin (351 events; median overall survival 10·7 months [95% CI 9·4–11·3])

100 90 80

Stratified HR 1·00 (95% CI 0·87–1·17; p=0·95)

70 Overall survival (%)

186 (41%)

65 (14%)

Data are n (%) or n (%, 95% CI). *The main reasons for unknown responses were no assessable target lesions at baseline, no assessable post-baseline scan (not done, partly missing, or quality issues), or pleural effusion or ascites. †Odds ratio 1·04 (95% CI 0·78–1·39, p=0·77).

20

60 50 40 30 20 10 0 0

3

6

9

12

15 18 21 24 27 30 33 Time from randomisation (months)

Number at risk Capecitabine-cisplatin 455 382 298 224 171 129 plus cetuximab Capecitabine-cisplatin 449 379 314 238 169 119

36

39

42

45

91

51

27

16

14

12

5

2

1

0

77

39

26

18

15

14

5

1

0

0

Figure 2: Kaplan-Meier estimates of progression-free survival (A) and overall survival (B) in the intention-totreat population Hazard ratios (95% CI) calculated with the stratified Cox’s proportional hazards model.

Data cutoff for the final analysis was on March 31, 2012, at which point 555 IRC-adjudicated PFS events had occurred. The median duration of cetuximab treatment was 3·4 months, and 366 (82%) of 446 patients received at least 80% of the planned cetuximab dose after the initial infusion. Relative dose intensity of cisplatincapecitabine was slightly lower in the control group than it was in the experimental group (table 2). Median PFS was 4·4 months (95% CI 4·2–5·5) for 455 patients in the experimental group compared with 494

196 (43%)

Progressive disease

5·6 months (5·1–5·7) for 449 controls (HR 1·09 [95% CI 0·92–1·29], p=0·32; figure 2). Median overall survival time was not substantially different between groups (figure 2). Median follow-up was 22·4 months (95% CI 21·3–24·0) in the experimental group and 21·0 months (20·0–24·9) in the control group. By data analysis (cutoff March 31, 2012), 362 (80%) patients in the experimental group had died, as had 351 (78%) in the control group. The number of patients who had an overall response or achieved disease control seemed similar between treatment groups (table 3). 241 (53%) of 455 patients in the experimental group and 238 (53%) of 449 controls received second-line treatment (appendix). In general, we noted few differences between treatment groups for either PFS or overall survival in an analysis of predefined subgroups (figure 3). Notably, patients aged 65 years or older might have benefited in terms of PFS after receipt of capecitabine-cisplatin compared with capecitabine-cisplatin plus cetuximab. Between treatment groups, PFS and overall survival were similar in patients with first-cycle acne-like rash (256 patients in the experimental group and eight controls) compared with those without first-cycle acne-like rash (147 patients and 419 patients; figure 3). In both treatment groups, patients with HER2positive tumours had longer median overall survival (figure 3) and better overall response than did patients with HER2-negative tumours. In the experimental group, 37 (51%, 95% CI 39–63) of 72 patients with HER2-positive tumours responded to treatment (ie, had an IRC adjudicated complete or partial response) compared with 76 (27%, 22–33) of 281 patients with HER2-negative tumours. In the control group, 27 (38%, 26–50) of 72 patients with HER2-positive tumours responded to treatment compared with 67 (26%, 21–32) of 254 patients with HER2-negative tumours. Tumour EGFR expression was assessable in 398 patients in the experimental group and 376 controls. The EGFR immunohistochemistry score was low; median score was 0. We noted no substantial differences between the www.thelancet.com/oncology Vol 14 May 2013

Articles

A

Patients

B

Median progression-free survival (months) Capecitabine- Capecitabinecisplatin plus cisplatin cetuximab

Hazard ratio* (95% CI)

904

4·4

5·6

1·10 (0·93–1·30)

<65 years

624

4·5

5·6

≥65 years

280

4·4

5·6

Male

673

5·5

Female

231

4·1

White

490

Asian

339

Other

Median overall survival (months) Capecitabine- Capecitabinecisplatin plus cisplatin cetuximab

Hazard ratio* (95% CI)

9·4

10·7

1·03 (0·89–1·19)

1·03 (0·84–1·26)

9·4

11·0

1·07 (0·90–1·28)

1·37 (1·01–1·86)

9·6

9·9

0·95 (0·73–1·24)

5·6

1·08 (0·89–1·31)

10·0

11·0

1·04 (0·88–1·23)

4·7

1·22 (0·87–1·71)

7·2

8·7

1·00 (0·75–1·34)

4·8

5·7

1·15 (0·92–1·45)

9·6

9·7

0·96 (0·79–1·17)

4·2

5·5

1·13 (0·86–1·48)

10·5

11·4

1·19 (0·93–1·52)

75

4·2

5·6

0·98 (0·54–1·78)

7·4

9·2

0·90 (0·54–1·50)

M0

28

6·0

6·9

1·03 (0·35–3·01)

11·4

13·3

1·29 (0·54–3·07)

M1

875

4·3

5·6

1·11 (0·94–1·32)

9·4

10·5

1·02 (0·88–1·18)

Yes

229

4·2

5·1

1·17 (0·83–1·65)

7·6

8·7

1·10 (0·82–1·47)

No

675

5·2

5·6

1·08 (0·89–1·31)

10·4

11·0

1·01 (0·85–1·20)

ITT Age

Sex

Ethnic origin

Disease stage

Peritoneal metastasis

Previous (neo)adjuvant chemo(radio)therapy Yes

55

5·6

6·4

1·34 (0·65–2·79)

11·1

14·6

1·47 (0·77–2·79)

No

849

4·4

5·6

1·10 (0·92–1·30)

9·4

10·5

1·01 (0·87–1·17)

Previous oesophagectomy/gastrectomy Yes

182

5·6

7·5

1·32 (0·88–1·97)

13·7

14·8

1·16 (0·82–1·64)

No

722

4·3

5·5

1·06 (0·88–1·27)

9·2

9·6

1·00 (0·85–1·17)

Gastro-oesophageal

144

5·6

5·6

1·12 (0·73–1·71)

13·4

11·0

0·80 (0·56–1·16)

Stomach

747

4·3

5·6

1·09 (0·91–1·31)

8·7

10·7

1·09 (0·93–1·29)

Primary site

Histological subclassification (Lauren) Intestinal type

311

5·5

5·5

0·98 (0·74–1·30)

11·9

9·4

0·90 (0·69–1·16)

Diffuse type

170

4·2

5·6

1·10 (0·74–1·65)

6·6

10·8

1·44 (1·01–2·03)

50

2·8

4·2

1·55 (0·75–3·22)

12·6

13·7

1·16 (0·63–2·16)

352

4·4

5·6

1·19 (0·91–1·55)

8·4

11·0

1·05 (0·83–1·32)

Mixed Unknown

ECOG performance status 0

465

5·6

5·8

1·06 (0·83–1·35)

11·7

12·0

1·01 (0·82–1·24)

1

438

4·1

4·5

1·18 (0·94–1·49)

6·7

8·6

1·10 (0·89–1·35)

Positive

144

5·6

6·9

1·33 (0·88–2·00)

13·3

14·0

1·04 (0·71–1·53)

Negative

535

4·6

5·5

0·99 (0·80–1·23)

9·2

9·7

0·98 (0·82–1·19)

Grade 0

566

4·3

5·6

1·20 (0·95–1·52)

8·3

10·8

1·14 (0·93–1·41)

Grade 1–4

264

5·5

6·8

1·86 (0·68–5·04)

10·9

12·2

1·09 (0·48–2·46)

HER2 status

First-cycle acne-like rash

0·5

1·0

Favours capecitabine-cisplatin plus cetuximab

10·0 Favours capecitabine-cisplatin

0·5

1·0

Favours capecitabine-cisplatin plus cetuximab

Favours capecitabine-cisplatin

Figure 3: Forest plots of progression-free survival (A) and overall survival (B) in a prespecified subgroup analysis of the intention-to-treat (ITT) population *Hazard ratio based on a univariable unstratified analysis of prespecified subgroups.

treatment groups for PFS or overall survival according to EGFR immunohistochemistry score (appendix). Compared with the control group, more patients in the capecitabine-cisplatin plus cetuximab group reported grade 3–4 adverse events, including grade 3–4 diarrhoea, hypokalaemia, hypomagnesaemia, rash, and hand-foot syndrome (table 4). Grade 3–4 neutropenia was more www.thelancet.com/oncology Vol 14 May 2013

common in the control group than in patients in the capecitabine-cisplatin plus cetuximab group. As expected, incidence of grade 3–4 skin reactions and acne-like rash was substantially higher in the cetuximabcontaining regimen than in the control regimen (table 4). Rates of grade 3–4 cardiac events were also slightly increased in the capecitabine-cisplatin plus cetuximab 495

Articles

Capecitabine-cisplatin plus cetuximab group (n=446)

Capecitabine-cisplatin group (n=436)

Grade 1–2

Grade 1–2

Grade 3

Grade 4

Grade 3

Grade 4

Adverse events Any

77 (17%)

245 (55%) 124 (28%)

95 (22%) 237 (54%)

100 (23%)

Neutropenia

98 (22%)

89 (20%)

11 (2%)

Hypokalaemia

32 (7%)

44 (10%)

12 (3%)

100 (23%) 119 (27%) 20 (5%)

35 (8%)

22 (5%)

Hypomagnesaemia

86 (19%)

33 (7%)

14 (3%)

55 (13%)

6 (1%)

Anaemia

91 (20%)

38 (9%)

1 (<1%)

113 (26%)

40 (9%)

7 (2%)

Fatigue

154 (35%)

34 (8%)

2 (<1%)

136 (31%)

25 (6%)

2 (<1%)

Diarrhoea

144 (32%)

33 (7%)

2 (<1%)

Nausea

247 (55%)

29 (7%)

0

5 (1%) 0

91 (21%)

15 (3%)

2 (<1%)

232 (53%)

38 (9%)

0 1 (<1%)

Decreased appetite

193 (43%)

32 (7%)

0

176 (40%)

24 (6%)

Vomiting

138 (31%)

29 (7%)

3 (1%)

165 (38%)

34 (8%)

Rash

165 (37%)

29 (7%)

0

23 (5%)

Hand-foot syndrome

131 (29%)

30 (7%)

1 (<1%)

11 (2%)

26 (6%)

5 (1%)

9 (2%)

4 (1%)

23 (5%)

Asthenia

72 (16%)

20 (4%)

2 (<1%)

74 (17%)

25 (6%)

2 (<1%)

Leucopenia

49 (11%)

14 (3%)

4 (1%)

75 (17%)

19 (4%)

5 (1%)

Thrombocytopenia

60 (13%)

15 (3%)

6 (1%)

68 (16%)

15 (3%)

7 (2%)

Hypocalcaemia

45 (10%)

14 (3%)

9 (2%)

26 (6%)

11 (3%)

1 (<1%)

Hyponatraemia Pulmonary embolism

1 (<1%)

0

0

88 (20%)

9 (2%)

0

11 (3%)

26 (6%)

7 (2%)

6 (1%)

0 10 (2%)

Special adverse event category Skin reactions

285 (64%)

56 (13%)

1 (<1%)

67 (15%)

0

Acne-like rash

286 (64%)

48 (11%)

1 (<1%)

63 (14%)

0

Cardiac events*

28 (6%)

14 (3%)

16 (4%)

22 (5%)

9 (2%)

11 (3%)

6 (1%)

7 (2%)

8 (2%)

5 (1%)

6 (1%)

3 (1%)

25 (6%)

6 (1%)

1 (<1%)

17 (4%)

2 (<1%)

1 (<1%)

1 (<1%)

4 (1%)

1 (<1%)

4 (1%)

Infarction/ischaemia Arrhythmia Arrest

0

0

0 0

Congestive heart failure

1 (<1%)

0

2 (<1%)

0

0

1 (<1%)

Sudden death

0

0

1 (<1%)

0

0

2 (<1%) 0

Infusion-related reactions Mucositis Septic events

25 (6%)

9 (2%)

3 (1%)

4 (1%)

1 (<1%)

146 (33%)

16 (4%)

1 (<1%)

67 (15%)

7 (2%)

1 (<1%)

1 (<1%)

5 (1%)

6 (1%)

2 (<1%)

0

0

Data are n (%). *Grade 3–4 cardiac events in the capecitabine-cisplatin plus cetuximab vs capecitabine-cisplatin group p=0·1696 (p value should be regarded as purely descriptive and is provided here only as additional information because cardiac imbalance was the reason for temporary clinical hold of the study).

Table 4: Adverse events in the safety population

group, mainly because of arrhythmia, infarction, and ischaemia. 239 (54%) of 446 patients in the capecitabinecisplatin plus cetuximab group had serious adverse events, as did 194 (44%) of 436 patients in the control group. Imbalances between the treatment groups included pulmonary embolism (26 [6%] in the experimental group vs 13 [3%] controls), decreased appetite (15 [3%] vs 3 [<1%]), and anaemia (ten [2%] vs 17 [4%]). Adverse events leading to discontinuation of all study treatments were reported in 72 (16%) of 446 patients in the cetuximab-containing regimen and 80 (18%) of 436 patients in the control group. Fewer patients in the experimental group (104 [23%] of 446) than the control group (243 (56%) of 436) had adverse events leading to delay of all study drugs. 41 (9%) of 446 patients in the chemotherapy plus 496

cetuximab group and 33 (8%) of 436 controls had an adverse event leading to death (appendix).

Discussion In this randomised phase 3 study of patients with advanced or metastatic gastric or gastro-oesophageal junction cancer, addition of cetuximab to capecitabine and cisplatin did not improve PFS compared with chemotherapy alone. Therefore, the primary study endpoint was not met. The safety profiles of the treatment regimens were as expected (panel). PFS in the experimental group (median 4·4 months) was not generally improved compared with reported data from randomised studies of doublet and triplet chemotherapy regimens in this setting (median 3·9–7·0 months).4,5,22,23 For gastric cancer, PFS might not be regarded as a universally acceptable choice of primary endpoint at present, whereas overall survival might now be more commonly used. However, when our study was designed, PFS was accepted and the approach in the EXPAND study was in line with European Medicines Agency guidelines.24 Furthermore, time-totumour progression was the primary endpoint in a phase 3 study that contributed to the approval of docetaxel by several regulatory authorities for use in first-line treatment of advanced gastric cancer.25 The choice and number of stratification factors in the EXPAND study was made by the scientific advisory board to the EXPAND investigators and was informed by the study from Chau and colleagues.26 One problem with the choice of randomisation strata in this setting is that no consensus exists about which are the most important factors for stage IV disease. Notably, however, other prognostic factors were reported to be balanced between the treatment groups, as expected due to the large number of patients who were randomly allocated. Furthermore, the unstratified HR for PFS supports the view that the choice of strata for the EXPAND study had no effect on the study results (figure 3). The recommendation to continue doublet chemotherapy until tumour progression and not to stipulate a specific number of cycles was also recommended by the scientific advisory boards on the basis of treatment preferences in most of the participating countries. Median overall survival (9·4 months) and the proportion of patients who had an overall response (136 [30%] of 455 patients) in the chemotherapy plus cetuximab group in our study were also much the same as data reported in other studies in this setting (median overall survival 8·6–13·0 months and response in 24·5–54% of patients).4–6,22,23,25,27 Furthermore, the proportion of patients who had a best response of complete or partial response was also disappointing when compared with data from phase 2 studies of cetuximab combined with FUFOX (oxaliplatin plus fluorouracil-folinic acid; 65%),16 irinotecan and fluorouracil-folinic acid (46%),15 FOLFOX6 (50%),14 cisplatin plus docetaxel (41%),28 and capecitabine plus oxaliplatin (51·2%).17 www.thelancet.com/oncology Vol 14 May 2013

Articles

In the REAL3 study29 of advanced oesophageal, gastrooesophageal junction, and gastric cancer, addition of panitumumab (an EGFR antibody) to epirubicin, oxaliplatin, and capecitabine was initially reported to be toxic. However, after reduction of the oxaliplatin and capecitabine doses in the experimental group, patients receiving panitumumab plus the modified chemotherapy triplet regimen reported worsened overall survival compared with those in the epirubicin-oxaliplatin-capecitabine group (8·8 months [95% CI 7·7–9·8 months] for the experimental group vs 11·3 months [9·6–13·0 months] for controls, HR 1·37 [1·07–1·76]; p=0·013).30 Data from the EXPAND and REAL3 studies suggest that addition of EGFR antibodies to chemotherapy does not convey additional benefit for patients with advanced gastric and gastro-oesophageal junction cancer. This absence of efficacy might be attributable to the choice of capecitabine and cisplatin as the chemotherapy backbone. This choice was initially justified because the reported efficacy of the capecitabine and cisplatin combination was much the same as reported in studies that used triplet chemotherapy regimens, and because the toxicity profiles suggested chemotherapy doublets might be preferable in combination with a targeted drug.22,25,27 We used capecitabine rather than infusional fluorouracil on the basis of data from randomised trials in European31 and Asian patients in which non-inferiority of capecitabine with cisplatin was reported compared with infusional fluorouracil with cisplatin, but with comparable safety profiles.5 Metaanalysis also suggested improved efficacy of capecitabinecontaining combinations compared with those that contained fluorouracil.32 However, data from recent randomised studies in patients with KRAS wild-type metastatic colorectal cancer suggest that cetuximab plus first-line chemotherapy regimens containing an oral fluoropyrimidine, or bolus fluorouracil, does not provide additional benefit in patients compared with chemotherapy alone.33,34 A meta analysis35 of four studies suggested that choice and schedule of fluoropyrimidine in combination with cetuximab might be important, with regimens that used infusional fluorouracil being most favourable. The reasons for this effect are unclear, and might be attributable to a negative pharmacokinetic interaction between capecitabine and cetuximab, or increased incidence of toxic effects associated with this treatment combination compared with cetuximab and infusional fluorouracil.35 In our study, despite controls receiving one fewer drug and generally experiencing fewer toxic side-effects, patients in the experimental group had a slightly higher dose intensity of capecitabine and cisplatin. We speculate that the reason for this effect might be attributable to a negative interaction between cetuximab and the chemotherapy backbone, leading to a lower rate of reported haematological toxicity in this group and to fewer dose reductions compared with the chemotherapy group. Notably, this outcome occurred despite more patients www.thelancet.com/oncology Vol 14 May 2013

receiving granulocyte colony-stimulating factor cotreatment in the chemotherapy only group than in those treated with additional cetuximab (77 [17%] of 449 controls vs 61 [13%] of 455 patients in the experimental group). Conversely, another reason for this effect might be that, as an open-label study, patients who received the experimental treatment might have been more likely from a psychological perspective to tolerate the toxicity associated with treatment. For patients with metastatic colorectal cancer, tumour KRAS mutations are negative predictive biomarkers of EGFR antibody efficacy.10,11,36 Because of the low expected frequency of KRAS mutations in gastric cancer (about 3%),15,16 the EXPAND study would not have been powered to detect KRAS mutation as a predictive biomarker. Data from single-group studies in this setting suggest that EGFR expression, EGFR gene copy number, and expression of other EGFR ligands (epiregulin and amphiregulin), or downstream components of the EGFR-signalling pathway might be candidate biomarkers for EGFR-antibody efficacy.14,15,37 Preliminary data presented from the REAL3 study did not identify predictive biomarkers from the first 200 patients investigated.30 In our trial, EGFR tumour expression was generally low, and use of subpopulation treatment-effect pattern-plot analyses suggested that the EGFR Panel: Research in context Systematic review At the time our study was designed, a high unmet clinical need existed for an efficacious treatment for patients with advanced or metastatic gastric or gastro-oesophageal junction cancer. The design of the phase 3 EXPAND study was influenced by an analysis of the literature about existing chemotherapy regimens for the treatment of patients in this setting, and the successful introduction of the EGFR antibody cetuximab in combination with first-line chemotherapy in patients with metastatic colorectal cancer, non-small-cell lung cancer, and squamous-cell carcinoma of the head and neck. We identified the existing evidence by searching PubMed with the search terms “advanced gastric cancer (GEJ cancer)”, “chemotherapy”, and “cetuximab” for reports in English, without date restriction, and used the search results to inform the choice of the chemotherapy backbone with which to combine cetuximab in the treatment of patients in this setting. Our EXPAND study aimed to characterise cetuximab as a potential addition to the treatment armamentarium for patients with this disease. Interpretation Our findings show that addition of cetuximab to capecitabine and cisplatin in the first-line treatment of unselected patients with advanced gastric or gastro-oesophageal junction cancer provides no benefit compared with chemotherapy alone, with an overall negative risk–benefit ratio for the experimental treatment. This treatment combination is not recommended in the absence of validated predictive biomarkers.

497

Articles

immunohistochemistry score was not associated with PFS or overall survival in either treatment group. In the CRYSTAL and EXTREME studies,38 EGFR expression level was also not reported to be a clinically useful predictive biomarker because the benefit from addition of cetuximab to first-line chemotherapy in patients with metastatic colorectal cancer and metastatic or recurrent squamous-cell carcinoma of the head and neck was reported irrespective of tumour EGFR expression levels.38 However, in advanced non-small-cell lung cancer, EGFR expression level was reported to be predictive of overall survival in patients treated with cetuximab and first-line chemotherapy.20 We chose to analyse tumour HER2 status because we speculated that, because HER2 and EGFR are members of the same family of receptor tyrosine kinases, an interaction might occur between these receptors that could affect the efficacy of cetuximab. This analysis also allowed us to investigate the controversial role of HER2 status as a prognostic marker in this setting, and to compare patients in the EXPAND study population with known HER2 statuses with those from the ToGA study.7 In the ToGA study in patients with HER2positive advanced gastric or gastro-oesophageal junction cancer, trastuzumab in combination with chemotherapy significantly prolonged overall survival compared with chemotherapy alone. However, any comparison between populations in different studies needs to be interpreted with caution. Tumour HER2 status did not seem to be predictive of treatment efficacy in patients in our study; however, patients with HER2-positive tumours generally had a better overall survival and tumour response than did patients with HER2-negative tumours. In a biomarker assessment of patients with advanced gastric cancer treated with firstline chemotherapy plus placebo or bevacizumab, tumour HER2 status was not predictive or prognostic.39 By contrast, HER2-positive tumours were associated with poor outcome in some studies,40,41 and the prognostic value of HER2 status in this setting is controversial and might be dependent on other factors including tumour stage, histology, and site. Unlike in metastatic colorectal cancer,11 first-cycle acnelike rash was not associated with prolonged PFS or overall survival in patients in our study. Addition of cetuximab to capecitabine and cisplatin did not produce any unexpected adverse events: our toxicity profile was much the same as in previous reports.14–16,28 Incidence of grade 3–4 skin reactions and acne-like rash was substantially higher in the cetuximab-containing regimen than in the chemotherapy-alone regimen. Grade 3–4 cardiac events were slightly more frequent in patients receiving cetuximab than in those who did not receive it, mainly from infarction or ischaemia and arrhythmia from electrolyte imbalances. Overall, addition of cetuximab to first-line capecitabine and cisplatin in unselected patients with advanced gastric 498

or gastro-oesophageal junction cancer provided no benefit compared with chemotherapy alone. These findings were generally consistent between subgroups. Further molecular classification of this heterogeneous disease, with respect to candidate biomarkers, might be required before significant improvements in treatment outcomes can be made. Contributors FL, Y-KK, MM, and IC contributed to study design. FL was involved with study organisation with the sponsor. FL, Y-KK, H-CC, PS, SCO, GB, GK, CV, VMM, VG, JOP, AS, and MM contributed to recruitment of patients or data collection. FL, Y-KK, H-CC, JOP, AS, IC, HG, HM, and MM contributed to data analysis or data interpretation. FL, Y-KK, HC-C, AS, IC, HG, HM, and MM contributed to report writing. A medical writer, Paul Hoban (Cancer Communications and Consultancy, Knutsford, UK), drafted and amended an initial outline under the guidance of FL. This draft was subsequently expanded by Paul Hoban into a full report draft after a series of comments and discussions from FL, Y-KK, H-CC, AS, IC, HG, HM, and MM. The final version of the report has been approved by all authors and the study sponsor. Conflicts of interest FL has held an advisory role for Amgen, Lilly, Fresenius Biotech, Roche, and Merck KGaA, and has received research support from Merck KGaA, Sanofi Aventis, GSK, and Fresenius Biotech. YKK has consulted for Merck KGaA and Roche, and has received honoraria and research funding from Roche. GB consulted or advised for Amgen. MM has consulted or advised for Merck KGaA, Roche, Medac, and Amgen, has received honoraria from Merck KGaA, Roche, and Medac, and has received other payments including travel grants from Merck KGaA, Roche, and Amgen. At the time of this report, HG, HM, and IC are salaried employees of Merck KGaA. All other authors declare that they have no conflicts of interest. Acknowledgments We thank all participating patients and their families, the study investigators (appendix), study nurses, study monitors, data managers, and all other members of the EXPAND study team. The sponsor of the EXPAND study was Merck, Darmstadt, Germany. Michael Schlichting (Merck KGaA) provided statistical support in the study design and planning, and Oliver Kisker (Merck KGaA) contributed towards the clinical study design and planning. For the biomarker analysis, Christopher Stroh (Merck KGaA) provided scientific input and Hans Jürgen Grote (Merck KGaA) provided technical support. Paul Hoban provided medical writing services, which were funded by the study sponsor. References 1 Ferlay J, Shin HR, Bray F, Forman D, Mathers C, Parkin DM. Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer 2010; 127: 2893–917. 2 Price TJ, Shapiro JD, Segelov E, et al. Management of advanced gastric cancer. Expert Rev Gastroenterol Hepatol 2012; 6: 199–208. 3 Wagner AD, Unverzagt S, Grothe W, et al. Chemotherapy for advanced gastric cancer. Cochrane Database Syst Rev 2010; 3: CD004064. 4 Koizumi W, Narahara H, Hara T, et al. S-1 plus cisplatin versus S-1 alone for first-line treatment of advanced gastric cancer (SPIRITS trial): a phase III trial. Lancet Oncol 2008; 9: 215–21. 5 Kang YK, Kang WK, Shin DB, et al. Capecitabine/cisplatin versus 5-fluorouracil/cisplatin as first-line therapy in patients with advanced gastric cancer: a randomised phase III noninferiority trial. Ann Oncol 2009; 20: 666–73. 6 Boku N, Yamamoto S, Fukuda H, et al, and the Gastrointestinal Oncology Study Group of the Japan Clinical Oncology Group. Fluorouracil versus combination of irinotecan plus cisplatin versus S-1 in metastatic gastric cancer: a randomised phase 3 study. Lancet Oncol 2009; 10: 1063–69. 7 Bang YJ, Van Cutsem E, Feyereislova A, et al, and the ToGA Trial Investigators. Trastuzumab in combination with chemotherapy versus chemotherapy alone for treatment of HER2-positive advanced gastric or gastro-oesophageal junction cancer (ToGA): a phase 3, open-label, randomised controlled trial. Lancet 2010; 376: 687–97. 8 Nicholson RI, Gee JM, Harper ME. EGFR and cancer prognosis. Eur J Cancer 2001; 37 (suppl 4): S9–15.

www.thelancet.com/oncology Vol 14 May 2013

Articles

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

Kim MA, Lee HS, Lee HE, Jeon YK, Yang HK, Kim WH. EGFR in gastric carcinomas: prognostic significance of protein overexpression and high gene copy number. Histopathology 2008; 52: 738–46. Bokemeyer C, Bondarenko I, Hartmann JT, et al. Efficacy according to biomarker status of cetuximab plus FOLFOX-4 as first-line treatment for metastatic colorectal cancer: the OPUS study. Ann Oncol 2011; 22: 1535–46. Van Cutsem E, Köhne CH, Láng I, et al. Cetuximab plus irinotecan, fluorouracil, and leucovorin as first-line treatment for metastatic colorectal cancer: updated analysis of overall survival according to tumor KRAS and BRAF mutation status. J Clin Oncol 2011; 29: 2011–19. Vermorken JB, Mesia R, Rivera F, et al. Platinum-based chemotherapy plus cetuximab in head and neck cancer. N Engl J Med 2008; 359: 1116–27. Pirker R, Pereira JR, Szczesna A, et al, and the FLEX Study Team. Cetuximab plus chemotherapy in patients with advanced non-small-cell lung cancer (FLEX): an open-label randomised phase III trial. Lancet 2009; 373: 1525–31. Han SW, Oh DY, Im SA, et al. Phase II study and biomarker analysis of cetuximab combined with modified FOLFOX6 in advanced gastric cancer. Br J Cancer 2009; 100: 298–304. Moehler M, Mueller A, Trarbach T, et al, and the German Arbeitsgemeinschaft Internistische Onkologie. Cetuximab with irinotecan, folinic acid and 5-fluorouracil as first-line treatment in advanced gastroesophageal cancer: a prospective multi-center biomarker-oriented phase II study. Ann Oncol 2011; 22: 1358–66. Lordick F, Luber B, Lorenzen S, et al. Cetuximab plus oxaliplatin/ leucovorin/5-fluorouracil in first-line metastatic gastric cancer: a phase II study of the Arbeitsgemeinschaft Internistische Onkologie (AIO). Br J Cancer 2010; 102: 500–05. Kim C, Lee JL, Ryu MH, et al. A prospective phase II study of cetuximab in combination with XELOX (capecitabine and oxaliplatin) in patients with metastatic and/or recurrent advanced gastric cancer. Invest New Drugs 2011; 29: 366–73. Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to evaluate the response to treatment in solid tumors. European Organization for Research and Treatment of Cancer, National Cancer Institute of the United States, National Cancer Institute of Canada. J Natl Cancer Inst 2000; 92: 205–16. Hofmann M, Stoss O, Shi D, et al. Assessment of a HER2 scoring system for gastric cancer: results from a validation study. Histopathology 2008; 52: 797–805. Pirker R, Pereira JR, von Pawel J, et al. EGFR expression as a predictor of survival for first-line chemotherapy plus cetuximab in patients with advanced non-small-cell lung cancer: analysis of data from the phase 3 FLEX study. Lancet Oncol 2012; 13: 33–42. Bonetti M, Gelber RD. A graphical method to assess treatment-covariate interactions using the Cox model on subsets of the data. Stat Med 2000; 19: 2595–609. Cunningham D, Starling N, Rao S, et al, and the Upper Gastrointestinal Clinical Studies Group of the National Cancer Research Institute of the United Kingdom. Capecitabine and oxaliplatin for advanced esophagogastric cancer. N Engl J Med 2008; 358: 36–46. Al-Batran SE, Hartmann JT, Probst S, et al, and the Arbeitsgemeinschaft Internistische Onkologie. Phase III trial in metastatic gastroesophageal adenocarcinoma with fluorouracil, leucovorin plus either oxaliplatin or cisplatin: a study of the Arbeitsgemeinschaft Internistische Onkologie. J Clin Oncol 2008; 26: 1435–42. European Medicines Agency. Committee for Medicinal Products for Human Use. Appendix 1 to the Guideline on the Evaluation of Anticancer Medicinal Products in Man (CHMP/EWP/205/95 REV. 3). http://www.emea.europa.eu/pdfs/human/ewp/26757506en.pdf (accessed Feb 4, 2013). Van Cutsem E, Moiseyenko VM, Tjulandin S, et al, and the V325 Study Group. Phase III study of docetaxel and cisplatin plus fluorouracil compared with cisplatin and fluorouracil as first-line therapy for advanced gastric cancer: a report of the V325 Study Group. J Clin Oncol 2006; 24: 4991–97. Chau I, Norman AR, Cunningham D, Waters JS, Oates J, Ross PJ. Multivariate prognostic factor analysis in locally advanced and metastatic esophago-gastric cancer—pooled analysis from three multicenter, randomized, controlled trials using individual patient data. J Clin Oncol 2004; 22: 2395–403.

www.thelancet.com/oncology Vol 14 May 2013

27

28

29

30

31

32

33

34

35

36

37

38

39

40 41

Dank M, Zaluski J, Barone C, et al. Randomized phase III study comparing irinotecan combined with 5-fluorouracil and folinic acid to cisplatin combined with 5-fluorouracil in chemotherapy naive patients with advanced adenocarcinoma of the stomach or esophagogastric junction. Ann Oncol 2008; 19: 1450–57. Pinto C, Di Fabio F, Barone C, et al. Phase II study of cetuximab in combination with cisplatin and docetaxel in patients with untreated advanced gastric or gastro-oesophageal junction adenocarcinoma (DOCETUX study). Br J Cancer 2009; 101: 1261–68. Okines AF, Ashley SE, Cunningham D, et al. Epirubicin, oxaliplatin, and capecitabine with or without panitumumab for advanced esophagogastric cancer: dose-finding study for the prospective multicenter, randomized, phase II/III REAL-3 trial. J Clin Oncol 2010; 28: 3945–50. Waddell T, Chau I, Cunningham D, et al. Epirubicin, oxaliplatin, and capecitabine with or without panitumumab for patients with previously untreated advanced oesophagogastric cancer (REAL3): a randomised, open-label phase 3 trial. Lancet Oncol 2013; published online April 15. http://dx.doi.org/10.1016/S1470-2045(13)700096-2. Moehler M, Kanzler S, Geissler M, et al, and the Arbeitsgemeinschaft Internistische Onkologie, Germany. A randomized multicenter phase II study comparing capecitabine with irinotecan or cisplatin in metastatic adenocarcinoma of the stomach or esophagogastric junction. Ann Oncol 2010; 21: 71–77. Okines AF, Norman AR, McCloud P, Kang YK, Cunningham D. Meta-analysis of the REAL-2 and ML17032 trials: evaluating capecitabine-based combination chemotherapy and infused 5-fluorouracil-based combination chemotherapy for the treatment of advanced oesophago-gastric cancer. Ann Oncol 2009; 20: 1529–34. Maughan TS, Adams RA, Smith CG, et al, and the MRC COIN Trial Investigators. Addition of cetuximab to oxaliplatin-based first-line combination chemotherapy for treatment of advanced colorectal cancer: results of the randomised phase 3 MRC COIN trial. Lancet 2011; 377: 2103–14. Tveit KM, Guren T, Glimelius B, et al. Phase III trial of cetuximab with continuous or intermittent fluorouracil, leucovorin, and oxaliplatin (Nordic FLOX) versus FLOX alone in first-line treatment of metastatic colorectal cancer: the NORDIC-VII study. J Clin Oncol 2012; 30: 1755–62. Ku GY, Haaland BA, de Lima Lopes G Jr. Cetuximab in the first-line treatment of K-ras wild-type metastatic colorectal cancer: the choice and schedule of fluoropyrimidine matters. Cancer Chemother Pharmacol 2012; 70: 231–38. Douillard JY, Siena S, Cassidy J, et al. Randomized, phase III trial of panitumumab with infusional fluorouracil, leucovorin, and oxaliplatin (FOLFOX4) versus FOLFOX4 alone as first-line treatment in patients with previously untreated metastatic colorectal cancer: the PRIME study. J Clin Oncol 2010; 28: 4697–705. Luber B, Deplazes J, Keller G, et al. Biomarker analysis of cetuximab plus oxaliplatin/leucovorin/5-fluorouracil in first-line metastatic gastric and oesophago-gastric junction cancer: results from a phase II trial of the Arbeitsgemeinschaft Internistische Onkologie (AIO). BMC Cancer 2011; 11: 509. Licitra L, Störkel S, Kerr KM, et al. Predictive value of EGFR expression for first-line chemotherapy plus cetuximab in patients with head and neck and colorectal cancer: analysis of data from the EXTREME and CRYSTAL studies. Eur J Cancer 2012; published online Dec 19. DOI:10.1016/j.ejca.2012.11.018. Van Cutsem E, de Haas S, Kang YK, et al. Bevacizumab in combination with chemotherapy as first-line therapy in advanced gastric cancer: a biomarker evaluation from the AVAGAST randomized phase III trial. J Clin Oncol 2012; 30: 2119–27. Gravalos C, Jimeno A. HER2 in gastric cancer: a new prognostic factor and a novel therapeutic target. Ann Oncol 2008; 19: 1523–29. Grabsch H, Sivakumar S, Gray S, Gabbert HE, Müller W. HER2 expression in gastric cancer: rare, heterogeneous and of no prognostic value—conclusions from 924 cases of two independent series. Cell Oncol 2010; 32: 57–65.

499