The primary objective of this study was to compare OS between everolimus + best supportive care (BSC) and placebo + BSC. OS, was defined as the time from date of randomization to the date of death due to any cause. If at the analysis cut-off date a patient was not known to have died, survival was censored at the date of the last contact. OS was analyzed using the Kaplan Meier estimates method.

The Full Analysis Set (FAS) consists of all randomized patients. Following the intent-to-treat principle, patients were analyzed according to the treatment and stratum that they were assigned to at randomization.

Progression free survival was defined as the time from the date of randomization to the date of the first documented disease progression or death due to any cause, where progression was based on Investigator assessment of baseline and post-baseline scans according to RECIST. Progression free survival was censored if no PFS event was observed before the first to occur out of (i) the cut-off date, or (ii) the date when a further anticancer therapy was started. The censoring date was the date of the last adequate tumor assessment before either of these two events occurred. If a PFS event was observed after two or more missing or non-evaluable tumor assessments, then the date of progression was censored at the date of the last adequate tumor assessment; for a PFS event observed after a single missing or non-evaluable tumor assessment, the actual date of disease progression was used. Anslsis was done using Kaplan-Meier estimates method.

The Full Analysis Set (FAS) consists of all randomized patients. Following the intent-to-treat principle, patients were analyzed according to the treatment and stratum that they were assigned to at randomization.

The EORTC QLQ-C30 global health status/quality of life sub-scale (QL) was pre-specified as the primary domain of interest, followed by physical functioning (PF), social functioning (SF) and emotional functioning (EF).The EORTC QLQ-C30 questionnaire, along with a module specific for gastric cancer patients (EORTC QLQ-STO22), was used to evaluate PRO. The QLQ-C30 has five function scales (physical, role, cognitive, emotional and social), three symptom scales (fatigue, pain and nausea/vomiting) and a global health status/quality of life scale. In addition, there are questions that assess specific symptoms. The QLQ-STO22 consists of 22 questions that make up five multi-item scales (dysphagia, pain, reflux, eating and anxiety) and four single-item scales (dry mouth, tasting, body image and hair loss).

The Full Analysis Set (FAS) consists of all randomized patients. Following the intent-to-treat principle, patients were analyzed according to the treatment and stratum that they were assigned to at randomization.

The ECOG PS scale was used to classify patients according to their functional impairment, with scores ranging from 0 (fully active) to 5 (dead). An analysis of the time to definitive deterioration of the ECOG PS by one category of the score from baseline was performed. Definitive deterioration was defined as a definitive increase by one category from baseline in ECOG PS, with no later improvements observed during the course of the study. A single measure reporting an increase in ECOG PS is sufficient to consider it as a definitive worsening only if it was the last one available for the patient. Kaplan-Meier method was used to estimate the distribution function of time to definitive worsening.

The Full Analysis Set (FAS) consists of all randomized patients. Following the intent-to-treat principle, patients were analyzed according to the treatment and stratum that they were assigned to at randomization.

ORR was defined as the proportion of patients with measurable disease in whom best overall response (OR) was either complete response (CR) or partial response (PR) according to RECIST criteria.

The Full Analysis Set (FAS) consists of all randomized patients. Following the intent-to-treat principle, patients were analyzed according to the treatment and stratum that they were assigned to at randomization.

Cmin is the minimum (trough) steady-state drug concentration in the blood during multiple dosing and Cmax is the maximum (peak) blood drug concentration after dose administration. Cmax is estimated as the maximum of C1h and C2H. C1h is 1 hour post-dose blood concentration and C2h is 2 hour post-dose blood concentration. Only valid pre-dose (Cmin), C1h, and C2h everolimus samples were included in the analysis. Valid pre-dose samples were confirmed blood samples collected at steady-state, collected immediately prior to dosing on the same study day, and collected at approximately 24 ± 4 hours after the previous dose and with no vomiting within the first 4 hours following the last dose. Valid C1h and C2h samples were confirmed blood samples collected at steady-state and within ± 1 hour window and with no vomiting within the first 4 hours following the current and previous dose.

PK analyses were based on the safety population in patients with evaluable samples. Only valid pre-dose (Cmin) & Cmax everolimus samples were included. For patients who were unable to tolerate the protocol-specified dosing schedule, dose adjustments were allowed to keep the patient on study drug. Some patients had dose reductions to 5mg daily.

Cmin is the minimum (trough) steady-state drug concentration in the blood during multiple dosing and Cmax is the maximum (peak) blood drug concentration after dose administration. Cmax is estimated as the maximum of C1h and C2H. C1h is 1 hour post-dose blood concentration and C2h is 2 hour post-dose blood concentration. Only valid pre-dose (Cmin), C1h, and C2h everolimus samples were included in the analysis. Valid pre-dose samples were confirmed blood samples collected at steady-state, collected immediately prior to dosing on the same study day, and collected at approximately 24 ± 4 hours after the previous dose and with no vomiting within the first 4 hours following the last dose. Valid C1h and C2h samples were confirmed blood samples collected at steady-state and within ± 1 hour window and with no vomiting within the first 4 hours following the current and previous dose.

PK analyses were based on the safety population in patients with evaluable samples. Only valid pre-dose (Cmin) & Cmax everolimus samples were included. For patients who were unable to tolerate the protocol-specified dosing schedule, dose adjustments were allowed to keep the patient on study drug. Some patients had dose reductions to 5mg daily.