Assessed according to RECIST 1.1. PFS 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. If a patient had not progressed or died, on the date of the analysis cut-off or when he/she received any further anti-neoplastic therapy, PFS was censored on the date of last tumor assessment before the cutoff date or the anti-neoplastic therapy date. The distribution of PFS was estimated using the Kaplan-Meier method. The median PFS along with 95% confidence intervals was presented by treatment group.

Full Analysis Set (FAS) consisted of all randomized patients.

Overall survival (OS) was the key secondary endpoint and was defined as the time from date of randomization to the date of death due to any cause. If a patient was not known to have died, survival was censored on the date of last contact.

Full Analysis Set (FAS) consisted of all randomized patients.

PFS 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. The primary analysis for PFS (based on central review) was also to be repeated on FAS considering the Investigator assessments and using the same analytical conventions as the primary analysis.

Full Analysis Set (FAS) consisted of all randomized patients.

Overall response rate (ORR) was defined as the proportion of patients with best overall response of complete response (CR) or partial response (PR). Best overall esponse (BOR) for each patient was determined from the sequence of overall (lesion) responses according to the following rules: CR = at least two determinations of CR at least 4 weeks apart before progression where confirmation required or one determination of CR prior to progression where confirmation not required. CR = at least two determinations of CR at least 4 weeks apart before progression where confirmation required or one determination of CR prior to progression where confirmation not required. SD = at least one SD assessment (or better) > 6 weeks after randomization (and not qualifying for CR or PR). PD = progression ≤ 17 weeks after randomization (and not qualifying for CR, PR or SD).

Full Analysis Set (FAS) consisted of all randomized patients.

Time to definitive worsening of Karnofsky performance status (KPS) was defined as the time from date of randomization to the date of definitive worsening of KPS or to the date of death whichever occurred earlier. Definitive worsening was defined as a definitive decrease in performance status by at least one Karnofsky category (i.e. at least 10 points less) compared to Baseline. Worsening was considered definitive if no later increase above the defined threshold was observed within the course of the study. A single measure reporting a decrease in Karnofsky performance status was sufficient to consider it as definitive only if it was the last one available for this patient. Time to definitive worsening of KPS was analyzed at the time of the final analysis for PFS.

Full Analysis Set (FAS) consited of all randomized patients.

The Kidney Cancer Symptom Index - Disease Related Symptoms (FKSI-DRS) is a validated symptom scale used in studies of patients with kidney cancer. It includes 9-items that assess pain, bone pain, fatigue, lack of energy, shortness of breath, fevers, weight loss, coughing, and blood in urine and responses to each question are answered on a 5-point Likert-type scale ranging from 0 to 4 (e.g., 0 = not at all; 1 = a little bit; 2 = somewhat; 3 = quite a bit; 4 = very much). FKSI-DRS scores range from 0 to 36, where higher scores correspond to better outcomes (eg, fewer symptoms).

Full Analysis Set (FAS) consisted of all randomized patients.

The EORTC QLQ-C30 contains 30 items and is composed of both multi-item scales and single-item measures. These include five functional scales (physical, role, emotional, social and cognitive functioning), three symptom scales (fatigue, pain, nausea, and vomiting), a global health status/QoL scale, and six single items (dyspnea, diarrhea, constipation, anorexia, insomnia and financial impact). Each of the multiitem scales includes a different set of items - no item occurs in more than one scale. Each item in the EORTC QLQ-C30 has 4 response categories (1=Not at all, 2= A little, 3= Quite a bit, 4= Very much) with the higher number representing a worse outcome.

Full Analysis Set (FAS) consisted of all randomized patients.

The EORTC QLQ-C30 contains 30 items and is composed of both multi-item scales and single-item measures. These include five functional scales (physical, role, emotional, social and cognitive functioning), three symptom scales (fatigue, pain, nausea, and vomiting), a global health status/QoL scale, and six single items (dyspnea, diarrhea, constipation, anorexia, insomnia and financial impact). Each of the multiitem scales includes a different set of items - no item occurs in more than one scale. Each item in the EORTC QLQ-C30 has 4 response categories (1=Not at all, 2= A little, 3= Quite a bit, 4= Very much) with the higher number representing a worse outcome.

Full Analysis Set (FAS) consisted of all randomized patients.

Predose concentrations of dovitinib were summarized by visit using PAS. All concentration data was listed by patient and time point using FAS. Mean pre-dose concentrations along with standard deviation (SD) was plotted over time if appropriate.

Pharmacokinetic Analysis Set (PAS) consisted of all patients who received at least one dose of dovitinib and had at least one evaluable post-Baseline dovitinib concentration measurement.