PFS was defined as the time from randomization to first documented radiographical progressive disease (PD), as determined by an independent review facility (IRF) using RECIST version 1.0, or death from any cause (within 18 weeks of last tumor assessment), whichever occurred first (Kaplan-Meier method). For target lesions, PD was defined as at least a 20% increase in the sum of the longest diameter (LD) of target lesions, taking as reference the smallest sum of the LD recorded since treatment started or the appearance of ≥1 new lesion. For non-target lesions, PD was defined as the appearance of ≥1 new lesion or unequivocal progression of existing lesions. Participants without IRF-determined PD or who had not died within 18 weeks of their last IRF-determined, progression-free tumor assessment were censored at the date of the last IRF-reviewed, evaluable tumor assessment; those with no post-baseline tumor assessment and who had not died within 18 weeks of baseline were censored at 1 day.

Intent-to-Treat (ITT) Population: All randomized participants

Overall survival (OS) was the time from randomization to death from any cause, using Kaplan-Meier methodology. Survival data was collected every 18 weeks during the post-treatment follow-up period until death, loss to follow-up, or withdrawal of consent. Those who were alive, lost to follow up, or withdrew consent were censored at the latest date they participated in the study; those without post-baseline data were censored at 1 day. OS analyses were planned to take place at the primary completion date (First Interim), after 385 deaths (Event-Driven Final), and at the end of study (End-of-Study). A second interim OS analysis was planned due to a formal request from the European Medicines Agency. Median [range] time in weeks on study at each OS analysis (Pertuzumab vs. Placebo): First: 77.1 [0.7-165.3] vs. 73.1 [0.4-165.3]; Second: 117.1 [0.7-207.9] vs. 105.9 [0.4-207.9]; Event-Driven Final: 189.9 [0.7-304.1] vs. 140.5 [0.4-301.6]; End-of-Study: 201.8 [0.7-520.0] vs. 138.0 [0.4-514.7].

ITT Population: All randomized participants

PFS was defined as the time from randomization to first documented radiographical progressive disease (PD), as determined by the investigator using RECIST version 1.0, or death from any cause (within 18 weeks of last tumor assessment), whichever occurred first (Kaplan-Meier method). For target lesions, PD was defined as at least a 20% increase in the sum of the longest diameter (LD) of target lesions, taking as reference the smallest sum of the LD recorded since treatment started or the appearance of ≥1 new lesion. For non-target lesions, PD was defined as the appearance of ≥1 new lesion or unequivocal progression of existing lesions. Participants without PD or who had not died within 18 weeks of their last investigator-determined, progression-free tumor assessment were censored at the date of the last investigator tumor assessment; those with no post-baseline tumor assessment and who had not died within 18 weeks of baseline were censored at 1 day.

Intent-to-treat population: All randomized patients.

An objective response was defined as the percentage of participants with confirmed best overall response of complete response (CR) or partial response (PR) determined by an independent review facility (IRF) using RECIST v1.0 on two consecutive occasions ≥4 weeks apart. For target lesions, CR: disappearance of all target lesions; PR: ≥30% decrease in the sum of the longest diameter (LD) of target lesions (baseline sum LD as reference); PD: ≥20% increase in the sum of the LD of target lesions (smallest sum of the LD recorded as reference) or appearance of ≥1 new lesion; SD: neither sufficient shrinkage to qualify for PR nor sufficient increase for PD. For non-target lesions, CR: disappearance of all non-target lesions; Incomplete/SD: persistence of ≥1 non-target lesions; PD: unequivocal progression of existing non-target lesions. 95% confidence intervals (CI) were calculated only for clinical responses using the Pearson-Clopper method.

ITT Population: All randomized participants; only participants with IRF-determined measurable disease at baseline (i.e., ≥1 target lesion) were included in the analysis.

Duration of objective response (estimated using the Kaplan-Meier method) was defined as the time from the initial confirmed complete response (CR) or partial response (PR), the date of tumor assessment at which the CR/PR was first detected by the independent review facility (IRF) using RECIST version 1.0, until the date of IRF-determined progressive disease (PD), death from any cause within 18 weeks of the last tumor assessment, or first administration of next line of anti-cancer therapy (whichever occurred first). If the visit when the initial CR or PR was observed spanned multiple dates, the latest date was used. Only participants in the ITT analysis population with an IRF-determined objective response (CR or PR), observed prior to IRF-assessed PD, death or next line of anti-cancer therapy, were included in the analysis. Participants who did not progress or die after they had a confirmed response were censored at the date of their last IRF-evaluable tumor measurement.

ITT Population: All randomized participants; only participants with IRF-determined measurable disease at baseline (i.e., ≥1 target lesion) that had an objective response were included in the analysis.

Time to symptom progression was defined as the time from randomization to the first symptom progression as measured by the Functional Assessment of Cancer Therapy-for patients with Breast Cancer (FACT-B) questionnaire with the Trial Outcomes Index-Physical/Functional/Breast (TOI-PFB) subscale. The FACT-B TOI-PFB subscale contains 24 items from 3 subsections of the FACT-B questionnaire: Physical well-being, functional well-being, and additional concerns for breast cancer patients (breast cancer subscale [BCS]). All items in the questionnaire were rated by the patient on a 5-point scale ranging from 0 ("not at all") to 4 ("very much"). The total score ranged from 0 to 96. A higher score indicates better perceived quality of life. A positive change score from baseline indicates improvement. Symptom progression was defined as a decrease from baseline of 5 points or more.

ITT population: All randomized participants; only female participants were included in the analysis.

Adverse event (AE) severity was assessed according to the National Cancer Institute Common Terminology Criteria for Adverse Events, Version 3.0 (NCI-CTCAE v3.0); if the AE was not specifically listed, the following grades of severity were used: Grade 1 = mild; Grade 2 = moderate; Grade 3 = severe; Grade 4 = life-threatening or disabling; and Grade 5 = death. Severe and serious are not synonymous. Severity refers to the intensity of an AE, whereas a serious AE must meet criteria set out in the protocol; both were independently assessed for each AE. Only the most severe intensity was counted for multiple occurrences of the same AE in one participant. AEs reported prior to first crossover treatment were included in the Placebo arm, and in the Crossover arm after that date, for participants who crossed over from placebo to pertuzumab. Median [range] time on study treatment per arm: Placebo: 49.3 [0.3-514.7] weeks; Pertuzumab: 75.7 [0.6-519.6] weeks; Crossover: 129.9 [0.3-322.3] weeks.

Safety Population: All participants who received at least one dose of any study medication.

Adverse event (AE) severity, including serious and non-serious AEs, was assessed according to the NCI-CTCAE version 3.0; if the AE was not specifically listed, the following grades of severity were used: Grade 1 is mild; Grade 2 is moderate; Grade 3 is severe; Grade 4 is life-threatening or disabling; and Grade 5 is death. Multiple occurrences of the same AE in 1 participant were counted multiple times. Only AEs that started during the overall study treatment period were included. The cutoff date for inclusion of events and for calculation of patient-years was the date of the most recent follow-up of the participant, defined as the last available date during the treatment period, excluding pre-treatment and safety follow-up data. Confidence intervals were calculated assuming the number of events followed a Poisson distribution. Data reported prior to the date of first crossover treatment were included under the Placebo arm for participants who crossed over from placebo to pertuzumab.

Safety Population: All participants who received at least one dose of any study medication.

Cardiac-related adverse events (AEs) to monitor during the study included investigator-assessed symptomatic left ventricular dysfunction (LVD), any LVD, or a serious adverse event (SAE) suggestive of congestive heart failure (CHF). All cardiac-related AEs were graded for severity according to NCI-CTCAE v3.0. Asymptomatic (Grades 1-2) and symptomatic (Grades 3-5) left ventricular systolic dysfunction (LVSD) both coded to the MedDRA preferred term LVD. Investigator-assessed events of symptomatic LVD were also graded for severity of symptoms according to Classes I (least severe) to IV (most severe) of the New York Heart Association (NYHA) Classification. SAEs suggestive of CHF were identified as serious events from the Standardized MedDRA Query (SMQ) (Wide) 'Cardiac Failure'. Median [range] time on study treatment per arm: Placebo: 49.3 [0.3-514.7] weeks; Pertuzumab: 75.7 [0.6-519.6] weeks; Crossover: 129.9 [0.3-322.3] weeks.

Safety Population: All participants who received at least one dose of any study medication.

The clinical diagnoses listed in this table, excluding cardiac safety (summarized separately), were also selected as adverse events (AEs) to monitor based on clinical and nonclinical data for pertuzumab and the safety profile established for trastuzumab, monoclonal antibodies in general, and potential effects associated with HER receptor inhibition. Search strategies were defined by single or aggregate MedDRA Preferred Terms (PT) through Standardized MedDRA Queries (SMQ), where possible, or based on Roche AE Group Terms (AEGT). Diarrhoea AEs: High-Level Term (HLT) 'Diarrhoea (excl. infective)' and PT 'Diarrhoea infectious'. Leukopenic and Febrile Neutropenic Infections: AEs from 'Infections & Infestations' with start ≤14 days after start date of Grade ≥3 AEs in SMQ(narrow) 'Leukopenia' or PT 'Febrile neutropenia', respectively. Median [range] time on study treatment per arm: Placebo: 49.3 [0.3-514.7] weeks; Pertuzumab: 75.7 [0.6-519.6] weeks; Crossover: 129.9 [0.3-322.3] weeks.

Safety Population: All participants who received at least one dose of any study medication.

Participants could continue study treatment with pertuzumab/placebo plus trastuzumab when docetaxel was discontinued due to an adverse event (AE). Discontinuation of pertuzumab/placebo or trastuzumab due to an AE led to discontinuation of all study medication. The number of participants who discontinued any study medication due to an AE includes those who discontinued all study medication and those who discontinued docetaxel only and then continued on targeted therapy (note: some of these participants may have subsequently discontinued all treatment due to a separate AE). Multiple occurrences of the same adverse event in 1 participant was counted only once. AEs reported prior to first crossover treatment were included in the Placebo arm, and after that date in the Crossover arm, for those who crossed over from placebo to pertuzumab. Median [range] time on study treatment per arm: Placebo: 49.3 [0.3-514.7] weeks; Pertuzumab: 75.7 [0.6-519.6] weeks; Crossover: 129.9 [0.3-322.3] weeks.

Safety Population: All participants who received at least one dose of any study medication.

Pertuzumab, trastuzumab, and docetaxel administration could have been delayed to assess or treat adverse events (AEs). Docetaxel dose reduction was allowed for myelosuppression, hepatic dysfunction, and other toxicities. No dose reduction was allowed for pertuzumab or trastuzumab. Multiple occurrences of the same adverse event in one participant was counted only once. AEs reported prior to first crossover treatment were included in the Placebo arm, and after that date in the Crossover arm, for participants who crossed over from placebo to pertuzumab. Median [range] time on study treatment per arm: Placebo: 49.3 [0.3-514.7] weeks; Pertuzumab: 75.7 [0.6-519.6] weeks; Crossover: 129.9 [0.3-322.3] weeks

Safety Population: All participants who received at least one dose of any study medication.

The post-treatment period was defined as the period following the treatment discontinuation visit. Only the following new adverse events (AEs) should have been reported during the post-treatment follow-up period: 1. Cardiac events (regardless of causality or seriousness) that started up to 1 year after the last dose, except for symptomatic left ventricular systolic dysfunction (regardless of causality) that started up to 3 years after the last dose; and 2. Treatment-related serious AEs, regardless of start date. AEs are listed by Medical Dictionary for Regulatory Activities, Version 21.1 (MedDRA v21.1) System Organ Class (SOC) and Preferred Term (PT); PTs fall under the SOC that is listed immediately above it in the table. Multiple occurrences of the same AE in one participant was counted only once. AEs reported prior to first crossover treatment were included in the Placebo arm, and after that date in the Crossover arm, for participants who crossed over from placebo to pertuzumab.

Safety Population: All participants who received at least one dose of any study medication.

All participants were required to have an left ventricular ejection fraction (LVEF) ≥50% at baseline, as measured by echocardiogram (preferred) or multiple-gated acquisition (MUGA) scan. The same method of LVEF assessment and the same institution/facility used at baseline was used throughout the study, to the extent possible. The baseline value was defined as the last valid value recorded during the pre-treatment period before or on study Day 1. The maximum absolute decrease in LVEF value was defined as the lowest post-baseline value up to the end of the overall study treatment period. Data reported prior to first crossover treatment were included in the Placebo arm, and after that date in the Crossover arm, for participants who crossed over from placebo to pertuzumab. Median [range] time on study treatment per arm: Placebo: 49.3 [0.3-514.7] weeks; Pertuzumab: 75.7 [0.6-519.6] weeks; Crossover: 129.9 [0.3-322.3] weeks.

Safety Population: All participants who received at least one dose of any study medication. Only participants with evaluable LVEF assessments during the overall study treatment period were included in the analysis.

All participants were required to have a left ventricular ejection fraction (LVEF) ≥50% at baseline, as measured by echocardiogram (preferred) or multiple-gated acquisition (MUGA) scan. The same method of LVEF assessment and the same institution/facility used at baseline was used throughout the study, to the extent possible. The baseline value was defined as the last valid value recorded during the pre-treatment period before or on study Day 1. The maximum absolute decrease in LVEF value was defined as the lowest post-baseline value up to the end of the overall study treatment period. Only data reported prior to the date of first crossover treatment were included for participants who crossed over from placebo to pertuzumab.

Safety Population: All participants who received at least one dose of any study medication. Only participants with evaluable LVEF assessments at baseline (BL) or at BL and post-BL (for change in LVEF from BL at max. decrease) were included in the analyses.

Clinical laboratory tests for blood biochemistry parameters were performed at local laboratories; any abnormal values (High or Low) were based on local laboratory normal ranges. Laboratory abnormalities are presented by the highest grade according to NCI-CTCAE v3.0. Not every abnormal laboratory value qualified as an adverse event, only if it met any of the following criteria: clinically significant (per investigator); accompanied by clinical symptoms; resulted in a change in study treatment; or required a change in concomitant therapy. Median [range] time on study treatment per arm: Placebo: 49.3 [0.3-514.7] weeks; Pertuzumab: 75.7 [0.6-519.6] weeks; Crossover: 129.9 [0.3-322.3] weeks. ALP = alkaline phosphatase; GGT = gamma-glutamyl transferase; SGOT = serum glutamic-oxaloacetic transaminase; SGPT = serum glutamic-pyruvic transaminase

Safety Population: All participants who received at least one dose of any study medication. Only participants with at least one valid laboratory value for any given parameter during the overall study treatment period were included in the analysis.

Clinical laboratory tests for hematology parameters were performed at local laboratories; any abnormal values (High or Low) were based on local laboratory normal ranges. Laboratory abnormalities are presented by the highest grade according to NCI-CTCAE v3.0. Not every abnormal laboratory value qualified as an adverse event, only if it met any of the following criteria: clinically significant (per investigator); accompanied by clinical symptoms; resulted in a change in study treatment; or required a change in concomitant therapy. Median [range] time on study treatment per arm: Placebo: 49.3 [0.3-514.7] weeks; Pertuzumab: 75.7 [0.6-519.6] weeks; Crossover: 129.9 [0.3-322.3] weeks. INR = International Normalized Ratio; PTT = partial thromboplastin time; WBC = white blood cell

Safety Population: All participants who received at least one dose of any study medication. Only participants with at least one valid laboratory value for any given parameter during the overall study treatment period were included in the analysis.