A Study of Trastuzumab Emtansine Versus Taxane in Participants With Human Epidermal Growth Factor Receptor 2 (HER2)-Positive Advanced Gastric Cancer

Overall survival was defined as the time between the date of randomization and date of death due to any cause. Kaplan-Meier estimates were used for analysis. Participants for whom no death was reported prior to an analysis cutoff (10 August 2013) was censored at the latest date before the cutoff in which they were known to be alive.

Time frame: Date of randomization until death (up to 1 year)

Population: Analysis population included all participants that had been enrolled in phase 2 (stage 1) up to a clinical cut-off date of 10 August 2013; participants grouped according to the therapy they were randomized to receive. Here, N (number of participants analyzed)=number of evaluable participants during phase 2 up to 10 August 2013.

Overall survival was defined as the time between the date of randomization and date of death due to any cause. Kaplan-Meier estimates were used for analysis. Participants for whom no death was reported prior to an analysis cutoff (30 June 2015) was censored at the latest date before the cutoff in which they were known to be alive. All data from the standard taxane therapy and trastuzumab emtansine 2.4 mg (selected treatment arm) from phase 2 and phase 3 (Stage 2) are combined into phase 3 data, and thus cumulative data are provided within the results presented for phase 3. The confirmatory analyses are restricted to comparisons between the taxane arm and the selected trastuzumab emtansine arm (2.4 mg).

Time frame: Date of randomization until death (up to 2 years 3 months)

Population: ITT population included all randomized participants; participants grouped according to the therapy they were randomized to receive.

AUCinf= Area under the plasma concentration versus time curve (AUC) from time zero (pre-dose) to extrapolated infinite time (0 - inf). It is obtained from AUC (0 - t) plus AUC (t - inf). Stage 1 consists of all participants recruited before the regimen selection decision. Regimen selection analysis was carried out after 12 weeks of randomization.

Time frame: D1C1 and D1C4, C1D2, C1D3, C1D4/C1D5, C1D8, C1D15, C2D1 (up to 12 weeks)

Population: Participants who had at least one PK parameter estimated were included for analysis.

DOR: time from the date when a clinical response \[CR or PR\] was first documented to the date of first documented progressive disease (PD) or death. CR:disappearance of all target lesions, non-target lesions, and normalization of tumor marker level. PR: \>= 30% decrease in sum of the LD of all target lesions taking as reference the screening sum LD. PD: could base on symptom deterioration or at least a 20% increase in the sum of diameters of target or non-target lesions and new lesions, taking as reference the smallest sum on study (nadir), including baseline. To be assigned a status of PR or CR, changes in tumor measurements had to be confirmed by repeat assessments that should have been performed no less than 4 weeks after the criteria for response were first met. Longer intervals as determined by the study protocol were also appropriate. Cumulative data (up to primary analysis cut-off date of 30-June-2015) are provided for both phase 2 and phase 3 within the results of this measure.

Time frame: Date of randomization until disease progression or death, whichever occurred first (assessed at baseline, every 6 weeks up to 2 years 3 months)

Population: ITT population included all randomized participants, participants grouped according to the therapy they were randomized to receive. Here, N=number of participants evaluable for this outcome measure.

Maximum observed plasma concentration of DM1 were reported. Stage 1 consists of all participants recruited before the regimen selection decision. Regimen selection analysis was carried out after 12 weeks of randomization.

Time frame: C1D1 and C1C4, C1D2, C1D3, C1D4/C1D5, C1D8, C1D15, C2D1 (up to 12 weeks)

Population: Participants who had at least one PK parameter estimated were included for analysis. Here, n=number of participants evaluable at specified timepoint.

Maximum observed plasma concentration of Trastuzumab Emtansine (T-DM1) and total trastuzumab were reported. Stage 1 consists of all participants recruited before the regimen selection, which was carried out after 12 weeks of randomization.

Time frame: Day 1 (D1) of Cycle 1 (C1) and C4, C1D2, C1D3, C1D4/C1D5, C1D8, C1D15, C2D1 (up to 12 weeks)

Population: Participants who had at least one PK parameter estimated were included for analysis. Here, n=number of participants evaluable at specified timepoint.

Stage 2 consists of all participants recruited after the regimen selection decision up to primary data cut-off date 30-June-2015.

Time frame: C1D1; C4D1

Population: Participants who had at least one PK parameter estimated were included for analysis. Here, n=number of participants evaluable at specified timepoint.

AGC symptomatic progression: a worsening of \>=10-points in any 1 of the abdominal discomfort, loss of appetite, weakness and fatigue, upper abdominal pain, change in bowel movement, and/or weight loss scales of the EORTC QLQ-C30 and QLQ-STO22. QLQ-STO22 supplements EORTC QLQ-C30 to assess symptoms and commonly reported treatment-related side effects. There are 22 questions comprise 5 scales (dysphagia, pain, reflux symptom, diet restrictions, anxiety), 4 single items (dry mouth, hair loss, taste, body image), which are related to the symptoms of the disease. Most questions used 4-point scale (1 'Not at all' to 4 'Very much'). All scores and single-items transformed to a scale of 0-100; higher score=better level of functioning or greater degree of symptoms. Cumulative data (up to primary analysis cut-off date of 30-June-2015) are provided for both phase 2 and phase 3 within the results of this measure.

Time frame: Day 1 of each treatment cycle, at the study drug completion visit, and thereafter at survival follow-up (up to 2 years 3 months)

Population: ITT population included all randomized participants, participants grouped according to the therapy they were randomized to receive. Here, N=number of participants evaluable for this measure.

The EORTC QLQ-C30 is a validated, cancer-specific 30-item patient-reported measure, and contains 14 domains to assess the impact of cancer treatment on 5 aspects of participants functioning (physical, role, cognitive, emotional, and social), 9 aspects of disease/treatment-related symptoms (fatigue, nausea and vomiting, pain, dyspnea, insomnia, loss of appetite, constipation, diarrhea) and a global QoL/overall health status scale. Questions used 4 point scale (1 'Not at all' to 4 'Very much'; with the exception of the QoL/health status scale which uses a 7-point scale (1 'very poor' to 7 'Excellent'). Each scale is transformed on a scale of 0-100; a higher score equals (=) a better level of functioning or greater degree of symptoms. Change of greater than or equal to (\>=) 10-points has been found to be clinically significant. Cumulative data (up to primary analysis cut-off date of 30-June-2015) are provided for both phase 2 and phase 3 within the results of this measure.

Time frame: Day 1 of each treatment cycle, at the study drug completion visit, and thereafter at follow-up (up to 2 years 3 months)

Population: ITT population included all randomized participants, participants grouped according to the therapy they were randomized to receive. Here, N=number of participants with baseline and at least one post-baseline valid score.

The Quality of Life Questionnaire Stomach Cancer Module 22 (QLQ-STO22) supplements the EORTC QLQ-C30 to assess symptoms and treatment-related side effects commonly reported in participants. 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 use 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. Change of \>=10 points has been found to be clinically significant. Cumulative data (up to primary analysis cut-off date of 30-June-2015) are provided for both phase 2 and phase 3 within the results of this measure.

Time frame: Day 1 of each treatment cycle, at the study drug completion visit, and thereafter at survival follow-up (up to 2 years 3 months)

Population: ITT population included all randomized participants, participants grouped according to the therapy they were randomized to receive. Here, N=number of participants with baseline and at least one post-baseline valid score.

Progressive disease could base on symptom deterioration or was defined as at least a 20 percent (%) increase in the sum of the longest diameter of target lesions, taking as reference the smallest sum longest diameter recorded since treatment started or the appearance of one or more new lesions and/or the unequivocal progression of existing non-target lesions. Tumor assessment was performed using modified RECIST v1.1. Cumulative data (up to primary analysis cut-off date of 30-June-2015) are provided for both phase 2 and phase 3 within the results of this measure.

Time frame: Date of randomization until disease progression or death, whichever occurred first (assessed at baseline, every 6 weeks up to 2 years 3 months)

Population: ITT population included all randomized participants; participants grouped according to the therapy they were randomized to receive.

Objective response referred to participants with complete response (CR) or partial response (PR). CR: disappearance of all target lesions, non-target lesions, and normalization of tumor marker level. PR: greater than or equal to (\>=) 30% decrease in sum of the longest diameter (LD) of all target lesions taking as reference the screening sum LD. To be assigned a status of PR or CR, changes in tumor measurements had to be confirmed by repeat assessments that should have been performed no less than 4 weeks after the criteria for response were first met. Longer intervals as determined by the study protocol were also appropriate. Cumulative data (up to primary analysis cut-off date of 30-June-2015) are provided for both phase 2 and phase 3 within the results of this measure.

Time frame: Date of randomization until disease progression or death, whichever occurred first (assessed at baseline, every 6 weeks up to 2 years 3 months)

Population: ITT population included all randomized participants, participants grouped according to the therapy they were randomized to receive. Here, N=number of participants with measurable disease were included in analysis of this outcome measure.

Plasma decay half-life is the time measured for the plasma concentration to decrease by one half. Stage 1 consists of all participants recruited before the regimen selection decision. Regimen selection analysis was carried out after 12 weeks of randomization.

Time frame: D1C1 and D1C4, C1D2, C1D3, C1D4/C1D5, C1D8, C1D15, C2D1 (up to 12 weeks)

Population: Participants who had at least one PK parameter estimated were included for analysis.

Progression-free survival was defined as the time between the date of randomization and the first date of documented progression or date of death due to any cause, whichever occurred first. Tumor assessment was performed using modified RECIST v1.1. Progressive disease could base on symptom deterioration or was defined as at least a 20% increase in the sum of the longest diameter of target lesions, taking as reference the smallest sum longest diameter recorded since treatment started or the appearance of one or more new lesions and/or the unequivocal progression of existing non-target lesions. Kaplan-Meier estimates were used for analysis. Cumulative data (up to primary analysis cut-off date of 30-June-2015) are provided for both phase 2 and phase 3 within the results of this measure. The confirmatory analyses are restricted to comparisons between the taxane arm and the selected trastuzumab emtansine arm (2.4 mg).

Time frame: Date of randomization until disease progression or death, whichever occurred first (assessed at baseline, every 6 weeks up to 2 years 3 months)

Population: ITT population included all randomized participants; participants grouped according to the therapy they were randomized to receive.

CL is a quantitative measure of the rate at which a drug substance is removed from the body. Stage 1 consists of all participants recruited before the regimen selection decision. Regimen selection analysis was carried out after 12 weeks of randomization.

Time frame: D1C1 and D1C4, C1D2, C1D3, C1D4/C1D5, C1D8, C1D15, C2D1 (up to 12 weeks)

Population: Participants who had at least one PK parameter estimated were included for analysis.

Time to AGC symptom were defined as the time from randomization to the first documentation of an increase in at least one of the pre-specified abdominal discomfort, loss of appetite, weakness and fatigue, upper abdominal pain, change in bowel movement, and weight loss subscales of the QLQ STO22 and EORTC QLQ-C30. Cumulative data (up to primary analysis cut-off date of 30-June-2015) are provided for both phase 2 and phase 3 within the results of this measure.

Time frame: Day 1 of each treatment cycle, at the study drug completion visit, and thereafter at survival follow-up (up to 2 years 3 months)

Population: ITT population included all randomized participants, participants grouped according to the therapy they were randomized to receive. Here, N=number of participants evaluable for this measure.

Volume of distribution is defined as the theoretical volume in which the total amount of drug would need to be uniformly distributed to produce the desired blood concentration of a drug. Steady state volume of distribution (Vss) is the apparent volume of distribution at steady-state. Stage 1 consists of all participants recruited before the regimen selection decision. Regimen selection analysis was carried out after 12 weeks of randomization.

Time frame: D1C1 and D1C4, C1D2, C1D3, C1D4/C1D5, C1D8, C1D15, C2D1 (up to 12 weeks)

Population: Participants who had at least one PK parameter estimated were included for analysis.