OS is defined as the time from the date of randomization until death from any cause. For participants not known to have died by the data-inclusion cutoff date, OS is censored at the last date they were known to be alive.

All randomized participants who received at least one dose of study drug (including the censored participants). Number of participants censored in Arm A: Galunisertib = 9; Arm B: Galunisertib + Lomustine = 8; and Arm C: Lomustine + Placebo = 6.

PFS was defined as the time from randomization to the date of the first observation of objective disease progression or death from any cause, whichever occurred first. Participants known to be alive and without disease progression were censored at the date of their last objective progression-free disease assessment prior to the initiation of any subsequent systemic anticancer therapy.

All randomized participants who received at least one dose of study drug (including the censored participants). Number of participants censored in Arm A: Galunisertib = 7; Arm B: Galunisertib + Lomustine = 8; and Arm C: Lomustine + Placebo = 4.

Tumour response was assessed using Response Assessment in Neuro-Oncology (RANO) criteria. Responses included Complete Response (CR), Partial Response (PR), Stable Disease (SD), and Progressive Disease (PD). CR required disappearance of all enhancing lesions, no new lesions, stable or improved non-enhancing lesions, and no corticosteroid use. PR was defined as ≥50% reduction in enhancing lesion size, no new lesions, stable or improved non-enhancing lesions, and stable or reduced corticosteroid use. SD indicated no significant change in lesion size or clinical status. PD was defined as ≥25% increase in lesion size, new lesions, or clinical deterioration. Percentage of participants with tumor response is defined as the percentage of participants who achieved these tumor responses based on RANO criteria.Participants whose tumor response could not be assessed due to inadequate imaging data were categorized as 'Unknown'.

All randomized participants who received at least one dose of study drug.

The absorption rate constant (Ka) of Galunisertib was estimated using PopPK modeling based on plasma concentration-time data collected during Cycle 1. A two-compartment model with first-order absorption was applied using nonlinear mixed-effects modeling. Samples were collected at the following time points: Cycle 1 Day 1: Predose, 0.5-2 hours (h), 3.5-5 h, and 48 h post-dose; Day 14: Predose, 0.5-2 h, 3.5-5 h, 24 h, and 48 h post last dose. Individual participant Ka values were derived from model-estimated parameters using all available PK timepoints.The reported outcome is the mean of these individual Ka estimates across both treatment arms (Arm A: Galunisertib; Arm B: Galunisertib + Lomustine).

All randomized participants who received at least one dose of Galunisertib (Arm A and Arm B) and had evaluable PK data were included in the analysis. As prespecified in the statistical analysis plan, PK analyses of Galunisertib exposure parameters were conducted using data combined from both arms.

The Vss at steady state of Galunisertib was estimated using PopPK modeling based on plasma concentration-time data collected during Cycle 1.A two-compartment model was applied using nonlinear mixed-effects modeling. Samples were collected at the following time points: Cycle 1 Day 1: Predose, 0.5-2 hours (h), 3.5-5 h, and 48 h post-dose; Day 14: Predose, 0.5-2 h, 3.5-5 h, 24 h, and 48 h post last dose. Individual participant Vss values were derived from model-estimated parameters using all available PK timepoints. The reported outcome is the mean of individual Vss estimates across both treatment arms (Arm A: Galunisertib; Arm B: Galunisertib + Lomustine).

All randomized participants who received at least one dose of Galunisertib (Arm A and Arm B) and had evaluable PK data were included in the analysis. As prespecified in the statistical analysis plan, PK analyses of Galunisertib exposure parameters were conducted using data combined from both arms.

The apparent clearance (CL/F) of Galunisertib was estimated using PopPK modeling based on plasma concentration-time data collected during Cycle 1. A two-compartment model was applied using nonlinear mixed-effects modeling. Samples were collected at the following time points: Cycle 1 Day 1: Predose, 0.5-2 hours (h), 3.5-5 h, and 48 h post-dose; Day 14: Predose, 0.5-2 h, 3.5-5 h, 24 h, and 48 h post last dose. Individual participant CL/F values were derived from model-estimated parameters using all available PK timepoints. The reported outcome is the mean of these individual CL/F estimates across both treatment arms (Arm A: Galunisertib; Arm B: Galunisertib + Lomustine).

All randomized participants who received at least one dose of Galunisertib (Arm A and Arm B) and had evaluable PK data were included in the analysis. As prespecified in the statistical analysis plan, PK analyses of Galunisertib exposure parameters were conducted using data combined from both arms.

The Hopkins Verbal Learning Test-Revised comprises:

• 3 Learning Trials: Participants were presented with a list of 12 words (from 3 semantic categories) and were asked to recall them to test verbal learning and memory.
• Delayed Recall Trial: Conducted 20-25 minutes after the 3rd learning trial to test memory retention.
• Delayed Recognition Trial: Participants identified previously presented words from a list that included 12 distractors to test recognition discrimination.

Scoring Components:

• Total Recall Score (0-36): Sum of correctly recalled words across the 3 learning trials.
• Delayed Recall Score (0-12): Number of correct words recalled after the delay.
• Recognition Discrimination Index Score (+12 to -12): Calculated as the number of true positives (correctly identified words) minus false positives (incorrectly identified words, i.e., distractors).

For each of the 3 reported scores, higher scores = better neurocognitive performance; lower scores = decline.

All randomized participants who received at least one dose of the study drug and had a baseline and at least one post-baseline measurement for this outcome.

The MDASI-BT assesses the severity of multiple brain tumor-related symptoms and the impact of these symptoms on daily functioning in the last 24 hours. It includes:

• 13 core symptoms measuring severity of pain, fatigue, nausea, disturbed sleep, distress, shortness of breath, memory problems, lack of appetite, drowsiness, dry mouth, sadness, vomiting, and numbness/tingling, rated 0-10, where 0 = "not present" and 10 = "as bad as you can imagine."
• 9 brain tumor-specific symptoms assess severity of difficulty speaking, weakness, seizures, difficulty understanding, vision changes, appearance changes, bowel pattern changes, concentration problems, and irritability, rated 0-10, where 0 = "not present" and 10 = "as bad as you can imagine."
• 6 interference items assess impact on general activity, mood, work, relations, walking, and enjoyment of life, rated 0-10, where 0 = "did not interfere" and 10 = "interfered completely."

All randomized participants who received at least one dose of the study drug and had a baseline and at least one post-baseline measurement for this outcome.