Brain Metastases, Brain Tumor, Glioma of Brain, Metastatic Melanoma
Conditions
Brief summary
Dynamic contrast-enhanced (DCE) magnetic resonance imaging (MRI) is a potentially powerful diagnostic tool for the management of brain cancer and other conditions in which the blood-brain barrier is compromised. This trial studies how well precise DCE MRI works in diagnosing participants with high grade glioma that has come back or melanoma that has spread to the brain. The specially-tailored acquisition and reconstruction (STAR) DCE MRI could provide improved assessment of brain tumor status and response to therapy.
Detailed description
PRIMARY OBJECTIVES: I. To optimize and technically validate specially-tailored acquisition and reconstruction (STAR) DCE-MRI based on the accuracy and reproducibility of whole-brain tracer-kinetic (TK) parameter maps. SECONDARY OBJECTIVES: I. To develop a robust clinical implementation of STAR DCE-MRI. II. To clinically evaluate STAR DCE-MRI in patients with brain tumors. OUTLINE: Participants are assigned to 1 of 2 cohorts. COHORT I: Participants with recurrent high-grade glioma undergo STAR DCE-MRI every 2 months, and just prior to and 4-6 weeks after starting bevacizumab treatment. If there is concern for tumor progression (i.e. increased contrast enhancement), more frequent MRI scans will be scheduled. COHORT II: Participants with melanoma brain metastases undergo STAR DCE-MRI at baseline and 4-6 weeks after therapy. Participants may undergo more frequent MRI if there is concern for tumor progression.
Interventions
OTHERDynamic Contrast-Enhanced Magnetic Resonance Imaging
Undergo STAR DCE-MRI
OTHERBevacizumab Injection
Bevacizumab will be give to participants who have recurrent high-grade glioma as part of standard of care.
Sponsors
National Cancer Institute (NCI)
University of Southern California
Study design
Allocation
NON_RANDOMIZED
Intervention model
PARALLEL
Primary purpose
DIAGNOSTIC
Masking
NONE
Eligibility
Sex/Gender
ALL
Age
21 Years to No maximum
Healthy volunteers
No
Inclusion criteria
* COHORT I: Recurrent high-grade glioma (often with thin areas of enhancement) treated with bevacizumab. * COHORT I: We will include adult patients with histopathologically confirmed high-grade glioma with evidence of tumor progression at baseline MRI who will undergo treatment with an anti-angiogenic agent (bevacizumab) with or without concomitant chemotherapy, and Karnofsky Performance Score \> 60%. * COHORT I: At least 30 days should have elapsed since prior therapy including surgery and temozolomide chemoradiation. * COHORT I: Satisfactory renal, hepatic, and hematologic function is required. * COHORT II: Melanoma brain metastases (often small and spread throughout the brain) treated with immunotherapy. * COHORT II: We will include adult patients with a tissue-proven history of melanoma who have contrast enhancing brain masses who will undergo treatment with immunotherapy with an anti-CTLA-4 or anti-PD-1 approach (e.g. ipilimumab, pembrolizumab, or nivolumab), and Karnofsky Performance Score \> 60%. * COHORT II: At least 30 days should have elapsed since prior therapy including surgery, stereotactic brain irradiation, and corticosteroid use.
Exclusion criteria
* COHORT I:
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Volume transfer constant (Ktrans) | Up to 3 years | The raw data will be acquired at the voxel level. Then the analytic parameters will be extracted from voxel-wise data such as the mean, median, interquartile range, skewness and kurtosis. Receiver-operating characteristic curves (ROC) will be used to illustrate the univariate prediction accuracy for each parameter in predicting the clinically determined outcome. The pattern of change with different clinical response status will be visually illustrated using spaghetti plots or other graphical approaches. Classification and Regression Tree (CART) with 10-fold cross validation will be used for building the final prediction model and determine the diagnostic cut point(s). CART analysis will also include demographics, comorbidity information, and relevant biological variables including sex. The final model accuracy will be assessed using area under the curve (AUC) when fitting a ROC curve using predicted outcome against the actual outcome. |
| Fractional plasma volume (vp) | Up to 3 years | The raw data will be acquired at the voxel level. Then the analytic parameters will be extracted from voxel-wise data such as the mean, median, interquartile range, skewness and kurtosis. ROC will be used to illustrate the univariate prediction accuracy for each parameter in predicting the clinically determined outcome. The pattern of change with different clinical response status will be visually illustrated using spaghetti plots or other graphical approaches. CART with 10-fold cross validation will be used for building the final prediction model and determine the diagnostic cut point(s). CART analysis will also include demographics, comorbidity information, and relevant biological variables including sex. The final model accuracy will be assessed using AUC when fitting a ROC curve using predicted outcome against the actual outcome. |
| Fractional extravascular-extracellular space volume (ve) | Up to 3 years | The raw data will be acquired at the voxel level. Then the analytic parameters will be extracted from voxel-wise data such as the mean, median, interquartile range, skewness and kurtosis. ROC will be used to illustrate the univariate prediction accuracy for each parameter in predicting the clinically determined outcome. The pattern of change with different clinical response status will be visually illustrated using spaghetti plots or other graphical approaches. CART with 10-fold cross validation will be used for building the final prediction model and determine the diagnostic cut point(s). CART analysis will also include demographics, comorbidity information, and relevant biological variables including sex. The final model accuracy will be assessed using AUC when fitting a ROC curve using predicted outcome against the actual outcome. |
| Model-free initial area under the contrast agent concentration curve (iAUC) | Up to 3 years | The raw data will be acquired at the voxel level. Then the analytic parameters will be extracted from voxel-wise data such as the mean, median, interquartile range, skewness and kurtosis. ROC will be used to illustrate the univariate prediction accuracy for each parameter in predicting the clinically determined outcome. The pattern of change with different clinical response status will be visually illustrated using spaghetti plots or other graphical approaches. CART with 10-fold cross validation will be used for building the final prediction model and determine the diagnostic cut point(s). CART analysis will also include demographics, comorbidity information, and relevant biological variables including sex. The final model accuracy will be assessed using AUC when fitting a ROC curve using predicted outcome against the actual outcome. |
Countries
United States
Outcome results
None listed