PFS is the interval between the date of randomization and the date of progression, defined by Response Evaluation Criteria in Solid Tumors (RECIST), or death due to any cause. Per RECIST, for target lesions (TLs), disease progression (PD) is defined as >=20% increase in the sum of the longest diameters (LD) of TLs, taking as a reference, the smallest sum LD recorded since the treatment started or the appearance of >=1 new lesions. For non-target lesions (NTLs), PD is defined as the appearance of >=1 new lesions and/or unequivocal progression of existing NTLs. Participants (par.) who did not progress/die were censored at the date of last adequate assessment (LAA). Par. who started a new anti-cancer therapy (ACT) prior to radiological progression/death were censored at the date of LAA prior to the new ACT. Par. who progressed/died after an extended period (>=12 months) without adequate assessment (AA) were censored at the date of their last visit with AA prior to progression/death.

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

Overall surival is defined as the interval between the date of randomization and the date of death due to any cause. For participants who did not die, the time to death was censored at the time of last contact.

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

Overall surival is defined as the interval between the date of randomization and the date of death due to any cause. For participants who did not die, the time to death was censored at the time of last contact.

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

Progression-free survival by GCIG criteria is defined as the time from the date of randomization to the earliest date of disease progression per GCIG criteria or death due to any cause. Progression is defined according to RECIST but can also be based upon serum CA-125. Progression or recurrence based on serum CA-125 levels are defined on the basis of a progressive serial elevation of serum CA-125, according to the following criteria: (1) participants (par.) with elevated CA-125 pretreatment and normalization of CA-125 must show evidence of CA-125 >=2x the upper normal limit (UNL) on two occasions at least one week apart or; (2) par. with elevated CA-125 pretreatment, which never normalizes, must show evidence of CA-125 >=2x the nadir value on two occasions at least one week apart or; (3) par. with CA-125 in the normal range pretreatment must show evidence of CA-125 >=2x the UNL on two occasions at least one week apart.

Intent-to-Treat (ITT) Population: all randomized participants.

For participants who did not progress or die, progression-free survival was censored at the time of the last adequate disease assessment.

3-year progression-free survival is defined as the percentage of participants who are progression-free at 3 years from randomization. Progression-free survival is defined as the time from the date of randomization to the earliest date of disease progression (defined by RECIST) or death due to any cause. Per RECIST, for target lesions, disease progression (PD) is defined as at least a 20% increase in the sum of the longest diameters (LD) of target lesions, taking as a reference, the smallest sum LD recorded since the treatment started or the appearance of one or more new lesions. For non-target lesions, PD is defined as the appearance of one or more new lesions and/or unequivocal progression of existing non-target lesions.

Intent-to-Treat (ITT) Population: all randomized participants.

The EORTC QLQ-C30 is a self-reported, 30-item cancer-specific instrument that assesses 15 domains: 5 functional scales (physical, role, emotional, cognitive, and social functioning), 9 symptom scales (fatigue, nausea and vomiting, pain, dyspnea, insomnia, appetite loss, constipation, diarrhea, and financial difficulties), and a global health status, or quality of life. Global health status is assessed using a 7-item Likert scale, ranging from 1 to 7 ("poor" to "excellent"). Participants were asked to respond to the following questions using the 7-item Likert scale: "How would you rate your overall health during the past week"; "How would you rate your overall quality of life during the past week?" Data are transformed to a scale ranging from 0 to 100. Higher scores represent better functioning (better quality of life). Mean changes from Baseline were calculated via mixed model-repeated measures analysis of covariance (ANCOVA).

ITT Population. Only those participants available at the specified time points were analyzed.

The OV (ovarian)-28 module is a 28-item addition to the EORTC QLQ-C30 that focuses on issues specific to ovarian cancer. It assesses attitude to disease/treatment functional symptoms, among others. Participants were asked to indicate the extent to which they experienced attention to disease/treatment functional problems in the week prior to assessment. Participants responded on a scale of 1-4 (1=not at all, 2=a little, 3=quite a bit, 4=very much) to the following questions: How much has your disease been a burden to you?; How much has your treatment been a burden to you?; Were you worried about your future health? Data are transformed to a scale ranging from 0 to 100. Higher scores represent better functioning (better quality of life). Mean changes from Baseline were calculated via mixed model-repeated measures ANCOVA.

ITT Population. Only those participants available at the specified time points were analyzed.

The OV-28 module is a 28-item addition to the EORTC QLQ-C30 that focuses on issues specific to ovarian cancer. It assesses body image symptoms, among others. Participants were asked to indicate the extent to which they experienced body image problems in the week prior to assessment. Participants responded on a scale of 1-4 (1=not at all, 2=a little, 3=quite a bit, 4=very much) to the following questions: Have you felt physically less attractive as a result of your disease or treatment?; Have you been dissatisfied with your body? Data are transformed to a scale ranging from 0 to 100. Higher scores represent better functioning (better quality of life). Mean changes from Baseline were calculated via mixed model-repeated measures ANCOVA.

ITT Population. Only those participants available at the specified time points were analyzed.

The OV-28 module is a 28-item addition to the EORTC QLQ-C30 that focuses on issues specific to ovarian cancer. It assesses peripheral neuropathy symptoms, among others. Participants were asked to indicate the extent to which they experienced peripheral neuropathy symptoms or problems in the week prior to assessment. Participants responded on a scale of 1-4 (1=not at all, 2=a little, 3=quite a bit, 4=very much) to the following questions: Did you have tingling hands or feet?; Have you had numbness in your fingers or toes?; Have you felt weak in your arms or legs? Data are transformed to a scale from 0 to 100. Lower scores represent better health (fewer symptoms) for symptom scales. Mean changes from Baseline were calculated via mixed model-repeated measures ANCOVA.

ITT Population. Only those participants available at the specified time points were analyzed.

The OV-28 module is a 28-item addition to the EORTC QLQ-C30 that focuses on issues specific to ovarian cancer. It assesses AB/GI symptoms, among others. Participants were asked to indicate the extent to which they experienced AB/GI symptoms or problems in the week prior to assessment. Participants responded on a scale of 1-4 (1=not at all, 2=a little, 3=quite a bit, 4=very much) to the following questions: Did you have abdominal pain?; Did you have a bloated feeling in your abdomen/stomach?; Did you have problems with your clothes feeling too tight?; Did you experience any change in bowel habit as a result of your disease or treatment?; Were you troubled by passing wind/gas/flatulence?; Have you felt full too quickly after beginning to eat?; Have you had indigestion/heartburn? Data are transformed to a scale from 0 to 100. Lower scores represent better health (fewer symptoms) for symptom scales. Mean changes from Baseline were calculated via mixed model-repeated measures ANCOVA.

ITT Population. Only those participants available at the specified time points were analyzed.

The OV-28 module is a 28-item addition to the EORTC QLQ-C30 that focuses on issues specific to ovarian cancer. It assesses hormonal/menopausal symptoms, among others. Participants were asked to indicate the extent to which they experienced hormonal/menopausal symptoms or problems in the week prior to assessment. Participants responded on a scale of 1-4 (1=not at all, 2=a little, 3=quite a bit, 4=very much) to the following questions: Did you have hot flashes?; Did you have night sweats? Data are transformed to a scale from 0 to 100. Lower scores represent better health (fewer symptoms) for symptom scales. Mean changes from Baseline were calculated via mixed model-repeated measures ANCOVA.

ITT Population. Only those participants available at the specified time points were analyzed.

The OV-28 module is a 28-item addition to the EORTC QLQ-C30 that focuses on issues specific to ovarian cancer. It assesses sexual functioning symptoms, among others. Participants were asked to indicate the extent to which they experienced sexual functioning problems in the week prior to assessment. Participants responded on a scale of 1-4 (1=not at all, 2=a little, 3=quite a bit, 4=very much) to the following questions: To what extent were you interested in sex?; To what extent were you sexually active?; If sexually active, to what extent was sex enjoyable for you?; If sexually active, did you have a dry vagina during sexual activity? Higher scores represent better functioning (better quality of life). Mean changes from Baseline were calculated via mixed model-repeated measures ANCOVA. Data were not analyzed due to low compliance (<50% at Baseline).

Data were not analyzed due to low compliance (<50% at Baseline).

The OV-28 module is a 28-item addition to the EORTC QLQ-C30 that focuses on issues specific to ovarian cancer. It assesses other chemotherapy SE symptoms, among others. Participants were asked to indicate the extent to which they experienced other chemotherapy SE symptoms/problems in the week prior to assessment. Participants responded on a scale of 1-4 (1=not at all, 2=a little, 3=quite a bit, 4=very much) to the following questions: Have you lost any hair?; If yes, were you upset by the loss of your hair?; Did food/drink taste different from usual?; Did you have aches or pains in your muscles or joints?; Did you have problems with hearing?; Did you urinate frequently?; Have you had skin problems (e.g., itchy, dry)? Data are transformed to a scale from 0 to 100. Lower scores represent better health (fewer symptoms) for symptom scales. Mean changes from Baseline were calculated via mixed model-repeated measures ANCOVA. Data were not analyzed due to low compliance (<50% at Baseline).

Mean changes from Baseline were calculated via mixed model-repeated measures ANCOVA. Data were not analyzed due to low compliance (<50% at Baseline).

The EuroQol (EQ-5D) questionnaire is a 2-page, generic, preference-based quality of life measure comprised of a 5-item health status measure and a visual analogue scale (VAS) and is used to generate two scores: the utility score and the thermometer score The thermometer score is based on a vertical VAS. The VAS is designed like a thermometer scale on which the best health state the participant can imagine is referenced at 100, and the worst health state the participant can imagine is marked by 0. Based on how good or bad the current health state is, the participant is asked to draw a line across the thermometer scale. For example, a line drawn across 46 on the scale of 0 to 100 would be coded 46. A negative adjusted mean change from Baseline represents a worsening of quality of life. Mean changes from Baseline were calculated via mixed model-repeated measures ANCOVA.

Intent-to-Treat (ITT) Population: all randomized participants.

Only those participants available at the specified time points were analyzed.

The EQ-5D utility score captures health status across five dimensions: mobility, self-care, usual activities, pain/discomfort, and anxiety and/or depression. Participants indicated the level of perceived problems in each of the five dimensions on three levels: 1, no problems; 2, some problems; 3, an extreme problem. Unique health states were defined by combining response levels from each of the five dimensions. For example, state 11111 indicates no problem on any of the five dimensions, whereas state 11223 indicates no problems with mobility or self-care; some problems with performing usual activities, moderate pain/discomfort; and extreme anxiety/depression. Responses are typically converted into health utilities or valuations on a scale ranging from 0 (worst health) to 1 (perfect health). A negative adjusted mean change from Baseline represents a worsening of quality of life. Mean changes from Baseline were calculated via mixed model-repeated measures ANCOVA.

Intent-to-Treat (ITT) Population: all randomized participants.

Only those participants available at the specified time points were analyzed.

An AE is any untoward medical occurrence in a participant or clinical investigation participant, temporally associated with the use of a medicinal product, whether or not considered related to the medicinal product. AEs were graded according to the Common Terminiology Criteria for Adverse Events (CTCAE), Version 4.0. Grade refers to the severity of the AE. The CTCAE displays Grades 1 through 5 with unique clinical descriptions of severity for each AE based on this general guideline: Grade 1, mild; Grade 2, moderate; Grade 3, severe; Grade 4, life threatening; Grade 5, death.

All Treated Population

Hematology toxicities were graded according to the Common Terminiology Criteria for Adverse Events (CTCAE), Version 4.0. Grade refers to the severity of the toxicity. The CTCAE displays Grades 1 through 5 with unique clinical descriptions of severity for each toxicity based on this general guideline: Grade 1, mild; Grade 2, moderate; Grade 3, severe; Grade 4, life threatening; Grade 5, death. Participants with a missing Baseline grade were assumed to have a Baseline grade of 0. WBC=White blood cell.

All Treated Population. Only those participants contributing toxicity data were analyzed.

Hematology toxicities were graded according to the Common Terminiology Criteria for Adverse Events (CTCAE), Version 4.0. Grade refers to the severity of the toxicity. The CTCAE displays Grades 1 through 5 with unique clinical descriptions of severity for each toxicity based on this general guideline: Grade 1, mild; Grade 2, moderate; Grade 3, severe; Grade 4, life threatening; Grade 5, death. Participants with a missing Baseline grade were assumed to have a Baseline grade of 0.

All Treated Population. Only those participants contributing toxicity data were analyzed.

Participants were assessed for thyroid function abnormalities. Clinical hypothyroidism is defined as 5 \

All Treated Population. Only those participants with any TSH above 5 MU/L were analyzed.

On-treatment deaths were collected from first dose of study treatment up to 28 days after study drug discontinuation, for a maximum duration of 27 months. Deaths post treatment survival follow up were collected after the on treatment period, up to 96 months (8 years).

Treated set.

One randomized patient in each treatment arm was excluded from the All Treated population because they did not receive treatment. The pazopanib group included 6 patients who were randomized to placebo but took at least one dose of pazopanib.

Four patients (one in placebo, three in pazopanib) had missing death dates and therefore were treated as censored at the last contact date.