Behaviors and Mental Disorders

All Conditions

Neoplasms

Symptoms and General Pathology

Nervous System Diseases

Cognitive Dysfunction

Neoplasm Metastasis

Brain Neoplasms

Central Nervous System Neoplasms

Nervous System Neoplasms

Brain Diseases

Central Nervous System Diseases

Anti-Dyskinesia Agents

All Drugs and Chemicals

Cardiotonic Agents

Memantine

Antiparkinson Agents

Dopamine

Dopamine Agents

Neurotransmitter Agents

Excitatory Amino Acid Antagonists

Whole brain radiation therapy (WBRT) and memantine

Whole brain radiation therapy with hippocampal avoidance (HA-WBRT) using intensity modulated radiation therapy (IMRT) and memantine

Intensity modulated radiation therapy (IMRT) 30 Gy in 10 fractions once per day, 5 days per week for approximately two week; starting within 21 calendar days after randomization.

Whole brain radiation therapy with hippocampal avoidance

Given PO daily during and after radiation therapy for a total of 24 weeks. Week 1: 5 mg in the AM, none in the PM; Week 2: 5 mg in the AM, 5 mg in the PM; Week 3: 10 mg in the AM, 5 mg in the PM; Weeks 4-24: 10 mg in the AM, 10 mg in the PM. Should start the same day as radiation therapy, at latest before the fourth radiation treatment.

Whole brain radiation therapy (WBRT) 30 Gy in 10 fractions once per day, 5 days per week for approximately 2 weeks

Whole brain radiation therapy

Paul Brown

This randomized phase III trial compares memantine hydrochloride and whole-brain radiotherapy with or without hippocampal avoidance in reducing neurocognitive decline in patients with cancer that has spread from the primary site (place where it started) to the brain. Whole brain radiotherapy (WBRT) is the most common treatment for brain metastasis. Unfortunately, the majority of patients with brain metastases experience cognitive (such as learning and memory) deterioration after WBRT. Memantine hydrochloride may enhance cognitive function by binding to and inhibiting channels of receptors located in the central nervous system. Radiation therapy uses high energy x-rays to kill tumor cells and shrink tumors. Using radiation techniques, such as intensity modulated radiotherapy to avoid the hippocampal region during WBRT, may reduce the radiation dose to the hippocampus and help limit the radiation-induced cognitive decline. It is not yet known whether giving memantine hydrochloride and WBRT with or without hippocampal avoidance works better in reducing neurocognitive decline in patients with brain metastases.

PRIMARY OBJECTIVES:

I. Determine whether the addition of whole-brain radiotherapy with hippocampal avoidance (HA-WBRT) increases time to neurocognitive failure at months 2, 4, 6, and 12 as measured by neurocognitive decline on a battery of tests: the Hopkins Verbal Learning Test-Revised (HVLT-R) for Total Recall, Delayed Recall, and Delayed Recognition, Controlled Oral Word Association (COWA), and the Trail Making Test (TMT) Parts A and B.

SECONDARY OBJECTIVES:

I. Determine whether the addition of HA-WBRT preserves neurocognitive function at months 2, 4, 6, and 12 as separately measured by each test, the HVLT-R for Total Recall, Delayed Recall, and Delayed Recognition; COWA; and TMT Parts A and B.

II. Evaluate the potential benefit of HA-WBRT in symptom burden, as measured by the M. D. Anderson Symptom Inventory-Brain Tumor Module (MDASI-BT).

III. Assessment of quality adjusted survival and cost analysis using the five-level version of the EuroQol five-dimensional (EQ-5D-5L).

IV. Compare cumulative incidence of progression and overall survival after WBRT versus HA-WBRT.

V. Compare adverse events between the treatment arms according to the Common Terminology Criteria for Adverse Events (CTCAE) version (v)4.0 criteria.

TERTIARY OBJECTIVES:

I. Collect serum, plasma, and imaging studies for future translational research analyses.

II. Evaluate magnetic resonance (MR) imaging biomarkers of white matter injury and hippocampal volumetry at baseline and 6 months as potential predictors of neurocognitive decline and differential benefit from HA-WBRT as compared to WBRT.

III. Association of symptom burden and anxiety/depression with neurocognitive function.

IV. Evaluate the potential correlation between the prognostic scoring systems Radiation Therapy Oncology Group (RTOG) recursive partitioning analysis (RPA) and the diagnosis-specific graded prognostic assessment (DS-GPA) and neurocognitive function at baseline and overtime.

After completion of study treatment, patients are followed up at 12 months.

Memantine Hydrochloride and Whole-Brain Radiotherapy With or Without Hippocampal Avoidance in Reducing Neurocognitive Decline in Patients With Brain Metastases

A Randomized Phase III Trial of Memantine and Whole-Brain Radiotherapy With or Without Hippocampal Avoidance in Patients With Brain Metastases

An exploratory analysis, beginning with correlation coefficients, will be used to assess the association of symptom burden and anxiety/depression with neurocognitive function at each time point. The symptom burden items of interest are the "distressed (upset)", "sad", and "mood" items. From the EQ-5D-5L, the depression/anxiety item will be of interest.

Neurocognitive function, as measured by the HVLT-R, COWA, and TMT, will be correlated with both the RTOG RPA and the DS-GPA classification systems. Baseline neurocognitive function for each test will be compared between both RPA classes using either a t-test or Wilcoxon-Mann-Whitney test, depending on the normality of the data.

Evaluated through MRI scans using physician-contoured and auto-contoured scores. Concordance rates will be assessed using Kappa statistics. The auto-contoured scores will be used for the remaining analyses due to the number of physicians reviewing the scans. White matter injury is measured by FLAIR volume change and is a continuous variable. Hippocampal volume is measured as a continuous variable also and both will be covariates considered in the Cox proportional hazards model to assess the impact on time to neurocognitive failure and the longitudinal modeling of neurocognitive function.

The relationship between EQ-5D-5L and MDASI-BT mood variables and neurocognitive function will be assessed.

Neurocognitive failure is defined as the first failure, defined as a neurocognitive decline using the reliable change index (RCI) on at least one of the following assessments or parts of : Hopkins Verbal Learning Test - Revised (HVLT-R), Trail Making Test (TMT), or Controlled Oral Word Association (COWA). The HVLT-R has 3 parts that were analyzed separately for decline: Total Recall, Delayed Recall, and Delayed Recognition. The TMT has 2 parts that were analyzed separately: Part A and Part B. Neurocognitive failure rate is estimated using the cumulative incidence method. Analysis was planned to occur after 233 events were reported. The protocol specifies that the distributions of failure times be compared between the arms, which is reported in the statistical analysis results. Six-month rates are provided.Analysis was planned to occur after 233 events were reported.

The HVLT-R assesses verbal learning and memory. The test involves memorizing a list of 12 nouns for 3 consecutive trials (Total Recall), recalling the 12 targets after a 20-minute delay (Delayed Recall), and then identifying the 12 targets from a list of semantically related or unrelated items (delayed recognition). Raw scores are derived for total recall (sum of the number of targets correctly recalled), delayed recall (sum of the number of targets correctly recalled), and a delayed recognition discrimination index (sum of targets incorrectly identified subtracted from the sum of the number of targets correctly identified). The range of scores for total recall is 0 to 36, for delayed recall is 0 to 12, and -12 to 12 for recognition. A higher score indicates better functioning. Scores are standardized, adjusting for age, education, and gender as necessary, such that mean 0 and standard deviation is 1. Change is calculated as baseline score subtracted from post-baseline score.

The HVLT-R assesses verbal learning and memory. The test involves memorizing a list of 12 nouns for 3 consecutive trials (Total Recall), recalling the 12 targets after a 20-minute delay (Delayed Recall), and then identifying the 12 targets from a list of semantically related or unrelated items (delayed recognition). Raw scores are derived for total recall (sum of the number of targets correctly recalled), delayed recall (sum of the number of targets correctly recalled), and a delayed recognition discrimination index (sum of targets incorrectly identified subtracted from the sum of the number of targets correctly identified). The range of scores for total recall is 0 to 36, for delayed recall is 0 to 12, and -12 to 12 for recognition. A higher score indicates better functioning. Scores are standardized by expressing the deviation from the mean score of the group in units of standard deviation. Change is calculated as baseline score subtracted from post-baseline score.

The TMT is a neuropsychological test of visual attention and task switching that can provide information about visual search speed, scanning, speed of processing, mental flexibility, and executive functioning. Subject is instructed to connect a set of 25 dots as quickly as possible while still maintaining accuracy. There are two parts to the test: in the first (Part A), the targets are all numbers (1, 2, 3, etc.) and the test taker needs to connect them in sequential order; in the second part (Part B), the subject alternates between numbers and letters (1, A, 2, B, etc.). The score is the amount of time, in seconds, that it takes the patient to complete each maze. The range for Part A is 0 to 180 (3 minutes) and for Part B is 0 to 300 (5 minutes). Lower scores indicate better functioning. Scores are standardized, adjusting for age, education, gender as needed, so that mean is 0 and standard deviation is 1. Change is calculated as baseline score subtracted from post-baseline score.

The TMT is a neuropsychological test of visual attention and task switching that can provide information about visual search speed, scanning, speed of processing, mental flexibility, and executive functioning. Subject is instructed to connect a set of 25 dots as quickly as possible while still maintaining accuracy. There are two parts to the test: in the first (Part A), the targets are all numbers (1, 2, 3, etc.) and the test taker needs to connect them in sequential order; in the second part (Part B), the subject alternates between numbers and letters (1, A, 2, B, etc.). The score is the amount of time, in seconds, that it takes the patient to complete each maze. The range for Part A is 0 to 180 (3 minutes) and for Part B is 0 to 300 (5 minutes). A lower score indicates better functioning. Scores are standardized by expressing the deviation from the mean score of the group in units of standard deviation. Change is calculated as baseline score subtracted from post-baseline score.

The COWA is a verbal fluency test that measures spontaneous production of words belonging to the same category or beginning with some designated letter. Patients are given 1 minute to name as many words as possible beginning with the designated letter. The procedure is then repeated for the remaining two letters. Two alternate forms of the COWA are employed to minimize practice effects. The score is the sum of the correct responses with a range of 0 to infinity. A higher score indicates better functioning. Scores are standardized, adjusting for age, education, and gender as necessary, such that mean is 0 and standard deviation is 1. Change is calculated as baseline score subtracted from post-baseline score.

Clinical Trial Battery Composite score is the arithmetic mean of the HVLT-R (Free Recall, Delayed Recall, Delayed Recognition), TMTA, TMTB, and COWA scores, all of which are standardized, adjusting for age, education, and gender as necessary, such that mean is 0 and standard deviation is 1. A participant must have at least 5 of the 6 scores. A higher composite score indicates better neurocognitive function.Change is calculated as baseline score subtracted from post-baseline score.

The MD Anderson Symptom Inventory for brain tumor (MDASI-BT) is a 28-item multi-symptom patient-reported outcome measure assessing the severity of symptoms experienced by cancer patients and the interference with daily living caused by these symptoms, with 9 items specific to brain tumors. Each item ranges from 0 (best condition) to 10 (worst condition). A subscale score (Symptom Severity) is the average of the subscale items, given that a specified minimum numbers of items were completed.

The MD Anderson Symptom Inventory for brain tumor (MDASI-BT) is a 28-item multi-symptom patient-reported outcome measure assessing the severity of symptoms experienced by cancer patients and the interference with daily living caused by these symptoms, with 9 items specific to brain tumors. Each item ranges from 0 (best condition) to 10 (worst condition). A subscale score (Interference) is the average of the subscale items, given that a specified minimum numbers of items were completed.

The MD Anderson Symptom Inventory for brain tumor (MDASI-BT) is a 28-item multi-symptom patient-reported outcome measure assessing the severity of symptoms experienced by cancer patients and the interference with daily living caused by these symptoms, with 9 items specific to brain tumors. Each item ranges from 0 (best condition) to 10 (worst condition). A subscale score (Cognitive Factor) is the average of the subscale items, given that a specified minimum numbers of items were completed.

The MD Anderson Symptom Inventory for brain tumor (MDASI-BT) is a 28-item multi-symptom patient-reported outcome measure assessing the severity of symptoms experienced by cancer patients and the interference with daily living caused by these symptoms, with 9 items specific to brain tumors. Each item ranges from 0 (best condition) to 10 (worst condition). A subscale score (Neurologic Factor) is the average of the subscale items, given that a specified minimum numbers of items were completed.

The EQ-5D-5L is a 2-part self-assessment questionnaire. First part is 5 items (mobility, self care, usual activities, pain/discomfort, anxiety/depression) each with 5 problem levels (1-none to 5-extreme). The 5-item index score is transformed into a utility score between 0 (worst health state) and 1 (best health state). The 2nd part is a visual analogue scale (VAS) valuing current health state, measured on a 20-cm scale ranging from 0 for the worst imaginable health state to 100 for best imaginable health state, marked at 10-point intervals. The index score is reported here.

The EQ-5D-5L is a 2-part self-assessment questionnaire. First part is 5 items (mobility, self care, usual activities, pain/discomfort, anxiety/depression) each with 5 problem levels (1-none to 5-extreme). The 5-item index score is transformed into a utility score between 0 (worst health state) and 1 (best health state). The 2nd part is a visual analogue scale (VAS) valuing current health state, measured on a 20-cm scale ranging from 0 for the worst imaginable health state to 100 for best imaginable health state, marked at 10-point intervals. The VAS score is reported here.

Intracranial progression-free survival time is defined as time from registration/randomization to the date of progression in the brain or death from any cause. Intracranial progression-free survival rates are estimated by the Kaplan-Meier method. Patients last known to be alive are censored at the date of last contact. Analysis was planned to occur after 233 primary endpoint events (neurocognitive failure) were reported. The protocol specifies that the distributions of failure times be compared between the arms, which is reported in the statistical analysis results. Six-month rates are provided.

Overall survival time is defined as time from registration/randomization to the date of death from any cause. Overall survival rates are estimated by the Kaplan-Meier method. Patients last known to be alive are censored at the date of last contact. Analysis was planned to occur after 233 primary endpoint events (neurocognitive failure) were reported. The protocol specifies that the distributions of failure times be compared between the arms, which is reported in the statistical analysis results. Six-month rates are provided.

. Adverse events were graded using the Common Terminology Criteria for Adverse Events (CTCAE) v3.0. Grade refers to the severity of the AE. The CTCAE v3.0 assigns 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 or disabling, Grade 5 Death related to AE.

National Cancer Institute (NCI)

NRG Oncology

HA-WBRT + Memantine

Whole brain radiation therapy with hippocampal avoidance (HA-WBRT) and memantine

WBRT + Memantin

WBRT + Memantine

Total

Age, Continuous

Sex: Female, Male

Ethnicity (NIH/OMB)

Race (NIH/OMB)

Primary tumor

70 = Cares for self; unable to carry on normal activity or to do active work; 80 = Normal activity with effort; some signs or symptoms of disease; 90 = Able to carry on normal activity, minor signs or symptoms of disease; 100 = Normal no complaints; no evidence of disease.

Karnofsky Performance Status

NS = neurologic symptoms; FA = fully active at home/work

Neurologic Function Status

The Radiation Therapy Oncology Group (RTOG) previously developed three prognostic classes for patients with brain metastases using recursive partitioning analysis (RPA) of a large database. (This study only includes the first two classes.) Class I = Karnofsky Performance Status (KPS) ≥ 70, primary controlled, age \< 65, metastasis in brain only. Class II = KPS ≥ 70 and 1 of the following: primary uncontrolled; primary controlled, age ≥ 65; primary controlled, age \< 65, metastases in brain \& other sites.

Recursive Partitioning Analysis (RPA) Class

Prior Radiosurgery

Prior Surgical Resection

Metastases

Education

Wendy Seiferheld

Null hypothesis: the addition of HA-WBRT as compared to WBRT will increase time to neurocognitive failure from 53.8% in the WBRT arm to 42.8% in the HA-WBRT arm at 6 months. Treating death as a competing risk and using a Gray's test with two-sided α=0.05 to test for statistically significant difference in the distribution of neurocognitive failure times, it was calculated that 230 events over both arms would provide 90% statistical power.

Time to Neurocognitive Failure

A mixed effects model was run with HVLT-R Symptom Score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Treatment Arm (WBRT+Memantine vs. HA-WBRT Memantine) is reported here.

A mixed effects model was run with HVLT-R Symptom Score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Age (\>61 year vs. \<= 61 years) is reported here.

A mixed effects model was run with HVLT-R Symptom Score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Baseline Total Recall is reported here.

2 months

4 months

6 months

12 months

Participants with standardized score at baseline

Change From Baseline in the Hopkins Verbal Learning Test -Revised (HVLT-R) Total Recall Score (Neurocognitive Decline)

A mixed effects model was run with HVLT-R Delayed Recall Score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Treatment Arm (WBRT+Memantine vs. HA-WBRT Memantine) is reported here.

A mixed effects model was run with HVLT-R Delayed Recall Score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Age (\> 61 years vs. \<= 61 years) is reported here.

A mixed effects model was run with HVLT-R Delayed Recall Score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Baseline Delayed Recall is reported here.

Change From Baseline in the Hopkins Verbal Learning Test -Revised (HVLT-R) Delayed Recall Score (Neurocognitive Decline)

A mixed effects model was run with HVLT-R Delayed Recognition Score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Treatment Arm (WBRT+Memantine vs. HA-WBRT Memantine) is reported here.

A mixed effects model was run with HVLT-R Delayed Recognition Score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Age (\> 61 years vs. \<= 61 years) is reported here.

A mixed effects model was run with HVLT-R Delayed Recognition Score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Baseline Delayed Recognition is reported here.

Change From Baseline in the Hopkins Verbal Learning Test -Revised (HVLT-R) Delayed Recognition (Neurocognitive Decline)

A mixed effects model was run with TMT Part A (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Treatment Arm (WBRT+Memantine vs. HA-WBRT Memantine) is reported here.

A mixed effects model was run with TMT Part A (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Age (\> 61 years vs. \<= 61 years) is reported here.

A mixed effects model was run with TMT Part A (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Baseline TMT Part A is reported here.

Change From Baseline in the Trail Making Test (TMT) Part A (Neurocognitive Decline)

A mixed effects model was run with TMT Part B (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Treatment Arm (WBRT+Memantine vs. HA-WBRT Memantine) is reported here.

A mixed effects model was run with TMT Part B (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Age (\> 61 years vs. \<= 61 years) is reported here.

A mixed effects model was run with TMT Part B (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Baseline TMT Part B is reported here.

Change From Baseline in the Trail Making Test (TMT) Part B (Neurocognitive Decline)

A mixed effects model was run with COWA (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Baseline COWA is reported here.

A mixed effects model was run with COWA (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Treatment Arm (WBRT+Memantine vs. HA-WBRT Memantine) is reported here.

A mixed effects model was run with COWA (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Age (\> 61 years vs. \<= 61 years) is reported here. is reported here.

Change From Baseline in the Controlled Oral Word Association (COWA) Test (Neurocognitive Decline)

A mixed effects model was run with the neurocognitive composite score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Treatment Arm (WBRT+Memantine vs. HA-WBRT Memantine) is reported here.

A mixed effects model was run with the neurocognitive composite score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Age (\> 61 years vs. \<= 61 years) is reported here.

A mixed effects model was run with the neurocognitive composite score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Baseline composite score is reported here.

Change From Baseline in the Clinical Trial Battery Composite (CTB COMP) Score [Neurocognitive Decline]

A mixed effects model was run with MDASI-BT Symptom Score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Treatment Arm (WBRT+Memantine vs. HA-WBRT Memantine) is reported here.

A mixed effects model was run with MDASI-BT Symptom Severity score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Baseline symptom severity is reported here.

A two-sample t-test with a 2-sided type I error of 0.05 provides \>90% statistical power to detect a medium effect size of 0.5 for a comparison of the change from baseline to 6 months from the start of treatment. The comparison at six months was tested within a mixed effects model with covariates age, RPA class, prior radiosurgery, prior surgical resection, baseline score, treatment arm, and time was used.

Change in M. D. Anderson Symptom Inventory Brain Tumor (MDASI-BT) Symptom Severity Score

A mixed effects model was run with MDASI-BT Interference Score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Treatment Arm (WBRT+Memantine vs. HA-WBRT Memantine) is reported here.

A mixed effects model was run with MDASI-BT Interference Score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Baseline interference score is reported here.

Change in M. D. Anderson Symptom Inventory Brain Tumor (MDASI-BT) Interference Score

A mixed effects model was run with MDASI-BT Cognitive Factor Score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Treatment Arm (WBRT+Memantine vs. HA-WBRT Memantine) is reported here.

A mixed effects model was run with MDASI-BT Cognitive Factor Score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Baseline cognitive factor score is reported here.

A mixed effects model was run with MDASI-BT Cognitive Factor Score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Prior radiotherapy (yes vs. no) is reported here.

Change in M. D. Anderson Symptom Inventory Brain Tumor (MDASI-BT) Cognitive Factor Score

A mixed effects model was run with MDASI-BT Neurologic Factor Score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Treatment Arm (WBRT+Memantine vs. HA-WBRT Memantine) is reported here.

A mixed effects model was run with MDASI-BT Neurologic Factor Score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Baseline neurologic factor is reported here.

A mixed effects model was run with MDASI-BT Neurologic Factor Score (baseline, 2, 4, 6, and 12 months) as the outcome of interest. Age, RPA class, prior radiosurgery, prior surgical resection were the covariates considered in each model along with interaction terms treatment\*time, time\*time, time\*time\*time and remained if p\<0.05. Baseline score, treatment arm, and time were forced into the model. Age (\> 61 years vs. \<= 61 years) is reported here.

Change in M. D. Anderson Symptom Inventory Brain Tumor (MDASI-BT) Neurologic Factor Score

Participants with baseline and 2-month scores

Change in EQ-5D-5L Index Score at 2 Months

Participants with baseline and 4-month scores

Change in EQ-5D-5L Index Score at 4 Months

Change in EQ-5D-5L Index Score at 6 Months

Participants with baseline and 12-month scores

Change in EQ-5D-5L Index Score at 12 Months

Change in EQ-5D-5L VAS Score at 2 Months

Change in EQ-5D-5L VAS Score at 4 Months

Participants with baseline and 6-month scores

Change in EQ-5D-5L VAS Score at 6 Months

Change in EQ-5D-5L VAS Score at 12 Months

Intracranial Progression-Free Survival

Overall Survival

Randomized participants who started protocol treatment

Study title	NCT ID	Conditions	Interventions	Eligibility	Organization	Link
Memantine Hydrochloride and Whole-Brain Radiotherapy With or Without Hippocampal Avoidance in Reducing Neurocognitive Decline in Patients With Brain Metastases	NCT02360215	Cognitive Impai... Metastatic Mali... Solid Neoplasm	Whole brain rad... Memantine Whole brain rad...	18 Years -	NRG Oncology
Memantine Hydrochloride and Whole-Brain Radiotherapy With or Without Hippocampal Avoidance in Reducing Neurocognitive Decline in Patients With Brain Metastases	NCT02360215	Cognitive Impai... Metastatic Mali... Solid Neoplasm	Whole brain rad... Memantine Whole brain rad...	18 Years -	NRG Oncology

Spots Global Cancer Trial Database for whole brain radiation therapy with hippocampal avoidance