Percentage of participants who died due to any cause was reported. As planned, ad hoc analysis was done for overall survival up to clinical cut-off date of 12 January 2012 (34 months), subsequent to the protocol-defined clinical cut-off date of 13 May 2011 (26 months). Overall survival was defined as the time between randomization and the date of death due to any cause.

ITT population

Overall survival was defined as the time between randomization and the date of death due to any cause. Overall survival was estimated using Kaplan Meier method. Reported data included censored observations. Participants for whom no death was captured on the clinical database were censored at the most recent date they were known to be alive. As planned, ad hoc analysis was done for overall survival up to clinical clinical cut-off date of 12 January 2012 (34 months), subsequent to the protocol-defined clinical cut-off date of 13 May 2011 (26 months).

ITT population

Progression of disease was defined as at least 20 percent (%) increase in the sum of longest diameters of target lesions compared to the smallest sum of longest diameters on-study and absolute increase of at least 5 millimeters (mm), progression of existing non-target lesions, or presence of new lesions. PFS was defined as the time between randomization and the date of first documented disease progression or death, whichever occurred first, according to Response Evaluation Criteria In Solid Tumors (RECIST).

ITT population

PFS was defined as the time between randomization and the date of first documented disease progression or death, whichever occurred first, according to RECIST. Progression of disease was defined as at least 20% increase in the sum of longest diameters of target lesions compared to the smallest sum of longest diameters on-study and absolute increase of at least 5 mm, progression of existing non-target lesions, or presence of new lesions. PFS was estimated using Kaplan Meier method. Reported data included censored observations. Participants who had neither progressed nor died at the time of study completion or who were lost to follow-up were censored at the date of the last tumor assessment or last follow-up for progression of disease. Participants for whom no post-baseline tumor assessments were available were censored at day of randomization.

ITT population

Progression of disease was defined as at least 20% increase in the sum of longest diameters of target lesions compared to the smallest sum of longest diameters on-study and absolute increase of at least 5 mm, progression of existing non-target lesions, or presence of new lesions. PFS during first-line therapy was defined as the time between randomization and the date of first documented disease progression or death, whichever occurred first and only if it occurred no later than 28 days after last confirmed intake of any study medication and before the start of non-study antineoplastic treatment, according to RECIST.

ITT population

PFS during first-line therapy was defined as time between randomization and date of first documented disease progression or death, whichever occurred first and only if it occurred no later than 28 days after last intake of any study medication and only if it occurred before start of non-study antineoplastic treatment, according to RECIST. Progression of disease was defined as at least 20% increase in sum of longest diameters of target lesions compared to smallest sum of longest diameters on-study & absolute increase of at least 5 mm, progression of existing non-target lesions, or presence of new lesions. PFS was estimated using Kaplan Meier method. Reported data included censored observations. Participants who neither progressed nor died in this interval, or who were lost to follow-up were censored at date of last tumor assessment/last follow-up within this time window. Participants for whom no post-baseline tumor assessments were available were censored at day of randomization.

ITT population.

Progression of disease was defined as at least 20% increase in the sum of longest diameters of target lesions compared to the smallest sum of longest diameters on-study and absolute increase of at least 5 mm, progression of existing non-target lesions, or presence of new lesions.

ITT population

Time to progression was defined as the time between randomization and the first occurrence of disease progression. Progression of disease was defined as at least 20% increase in the sum of longest diameters of target lesions compared to the smallest sum of longest diameters on-study and absolute increase of at least 5 mm, progression of existing non-target lesions, or presence of new lesions. Time to progression was estimated using Kaplan Meier method. Reported data included censored observations. Participants who had not progressed at the time of study completion (including participants who had died before disease progression) or who were lost to follow-up were censored at the date of the last tumor assessment or last follow-up for progression of disease. Participants for whom no post-baseline tumor assessments were available were censored at day of randomization.

ITT population

Best overall response during first-line therapy was defined as the occurrence of either a confirmed complete (CR) or a partial response (PR), as assessed by the RECIST criteria. CR: disappearance of all target and non-target lesions (TLs) and normalization of tumor markers. Pathological lymph nodes must have short axis measures less than (<) 10 mm. PR: at least a 30 % decrease in the sum of measures (longest diameter for tumor lesions and short axis measure for nodes) of TLs, taking as reference the baseline sum of longest diameters.

Measurable disease population included all randomized participants who had measurable disease at baseline according to RECIST.

Duration of response during first-line therapy was defined as the time from when response (CR or PR) was first documented to first documented disease progression or death (whichever occurs first) during first-line therapy. CR: disappearance of all target and non-TLs and normalization of tumor markers. Pathological lymph nodes must have short axis measures <10 mm. PR: at least a 30 % decrease in the sum of measures (longest diameter for tumor lesions and short axis measure for nodes) of TLs, taking as reference the baseline sum of longest diameters. Duration of response was estimated using Kaplan Meier method. Reported data included censored observations. Participants who did not progress or die after they had had a confirmed response were censored at the date of their last tumor measurement or last follow-up for progression of disease during first line therapy.

Measurable disease population

Disease control was defined as stable disease(SD) for 6 weeks or longer, CR plus PR as assessed by RECIST criteria for participants with measurable disease.CR:disappearance of all TLs & normalization of tumor markers. Pathological lymph nodes must have short axis measures<10 mm. PR:at least 30% decrease in sum of measures(longest diameter for tumor lesions and short axis measure for nodes)of TLs, taking as reference baseline sum of longest diameters.SD:neither sufficient shrinkage to qualify for PR nor sufficient increase to qualify for disease progression taking as reference smallest sum of longest diameters on study. For participants without measurable disease, clinical benefit rate was defined as no clinical disease progression for >/=6 weeks. Disease progression was defined as at least 20% increase in sum of diameters of TLs compared to smallest sum of diameters on-study and absolute increase of at least 5 mm, progression of existing non-target lesions, or presence of new lesions.

ITT population

EORTC QLQ-C30 included global health status (GHS)/quality of life (QOL), functional scales (physical, role, cognitive, emotional, and social), symptom scales (fatigue, pain, nausea/vomiting), and single items (dyspnea, appetite loss, insomnia, constipation, diarrhea, and financial difficulties). Most questions used a 4-point scale (1 'Not at All' to 4 'Very Much'); 2 questions used a 7-point scale (1 'Very Poor' to 7 'Excellent'). Scores were averaged and transformed to 0-100 scale; a higher score for Global Qol/functional scales=better level of QoL/functioning, or a higher score for symptom scale=greater degree of symptoms.

ITT population

The EORTC QLQ-STO22 is a gastric cancer quality of life questionnaire. There are 22 questions which comprise 5 scales (dysphagia, pain, reflux symptom, dietary restrictions, and anxiety) and 4 single items (dry mouth, hair loss, taste, body image). Most questions used 4-point scale (1 'Not at all' to 4 'Very much'; 1 question was a yes or no answer). A linear transformation was used to standardize all scores and single-items to a scale of 0 to 100; higher score=better level of functioning or greater degree of symptoms.

ITT population