Feasibility of Cognitive-Sparing Stereotactic Radiosurgery for Brain Metastases: Initial Report of Phase II COG-SRS Trial

PURPOSE/OBJECTIVE(S)

Radiotherapy at high doses leads to white matter (WM) and hippocampal injury which can cause cognitive decline. We designed the first, to our knowledge, evidence-based cognitive sparing brain SRS trial for patients with limited brain metastases. In this report we analyze feasibility of cognitive-sparing SRS, utilizing dose constraints for eloquent WM tracts and bilateral hippocampi.

MATERIALS/METHODS

This is a single-institution phase II NCI-funded clinical trial (NCT04343157) of cognitive-sparing brain SRS; eloquent regions of interest (ROI) and associated cognitive domains are outlined in Table 1. Patients underwent high-resolution quantitative diffusion and volumetric MRI at baseline and post-SRS follow-up. Comprehensive neurocognitive assessment was performed at baseline and 3 months post-SRS by a neuropsychologist evaluating different neurocognitive domains (Table 1). Clinical processing/workflow was largely automated with robust, validated segmentation tools of eloquent WM tracts and the hippocampi. Single fraction dose constraints to WM tracts were 12 Gy Dmax and 8.4 Gy to hippocampi with 3 and 5 fraction equivalent doses, based on our prior NTCP studies. Patient accrual, follow up imaging and neurocognitive testing is still active.

RESULTS

To date, n = 59 patients and 114 lesions have been treated with cognitive-sparing Linac-based SRS on trial. Median age is 63 years. Lung cancer was the most common primary (35.6%), along with breast cancer (23.7%) and melanoma (15.3%). Patients had 1-4 brain metastases; median prescription dose was 24 Gy/1, 27 Gy/3, or 30 Gy/5 depending on target volume. Most patients were treated < 7 business days from MRI. Cognitive endpoints have been collected on 89.5% of patients. Cognitive sparing constraints were met in 79.3% of plans while maintaining standard clinical SRS plan indices and coverage benchmarks. 38.9% of the plans that failed to meet the Dmax constraint kept D0.03cc under constraints. Plans going over constraints had tumors overlying or within 1 mm of eloquent ROIs. Local control for treated lesions at 6 months was 97.9% and 95.7% at 12 months. Distant intracranial control was 68.1% at 6 and 12 months. Median OS was 18 months.

CONCLUSION

Cognitive-sparing SRS treatment planning was successfully implemented for the majority of treated lesions with excellent local control. Mitigation of damage to eloquent structures has potential to further reduce cognitive decline after SRS. Full neurocognitive outcomes will be reported after accrual and testing are complete.

Quantitative Analysis of BrainAge using Neuroimaging Data to Estimate Radiation-Induced Brain Aging and Associations with Cognitive Changes in Brain Metastases Patients

PURPOSE/OBJECTIVE(S)

Radiotherapy is critical in brain tumor management yet causes accelerated aging via irreparable damage to white matter, cortex, and subcortical areas. BrainAGE is an open-source machine learning tool that analyses morphology from raw quantitative structural MRI to generate a predictive 'BrainAge' value, enabling analyses of patients with neurodegenerative diseases. We assessed the utility of this algorithm in measuring brain age morphology in brain metastases patients on a clinical trial, analyzed differences in BrainAge and chronological age, and explored associations with cognitive performance.

MATERIALS/METHODS

We analyzed pre-SRS high-resolution T1-weighted noncontrast volumetric MRI in patients with brain metastases (n = 57, median age 63 years, range 20-87) treated on a prospective clinical trial of cognitive-sparing SRS. The BrainAGE machine learning tool processed images to derive a BrainAge; it reliably generates this value using a Gaussian Processes regression, with a robust neurotypical training cohort. We examined correlations between BrainAge and chronological age with regression. Brain-predicted age gap (Brain-Gap) was calculated, defined as the difference between a patient's BrainAge and chronological age. We assessed cognitive function at baseline pre-SRS with a comprehensive battery of validated tests across multiple domains performed by a neuropsychologist, including memory (HVLT total), executive function (COWAT category fluency), and language (COWAT letter fluency). Raw test scores were scaled to T-scores adjusting for age, sex, and education level per testing norms. We examined if Brain-Gap was associated with cognitive performance pre-SRS.

RESULTS

BrainAGE software was robust, estimating a BrainAge for all input MRI sequences. We found a strong correlation between BrainAge and chronological age (r-squared = 0.82). The Brain-Gap median value was 0.84 and the mean was -0.90, in years (range -18.30-11.36 years). Brain-Gap was not significantly associated with cognitive performance across the three tests at the pre-SRS timepoint. Among patients with impaired cognition (T<35), there was no significant difference in BrainAge relative to chronological age.

CONCLUSION

To our knowledge, this is the first study utilizing the BrainAGE model as a biomarker of neuroanatomical age for brain tumors patients. Existing publications had lesion-free cohorts. Our results show that BrainAge was reliably associated with chronologic age in brain metastases patients pre-SRS. The Brain-Gap difference was minimal at baseline, and we found no significant association with pre-SRS cognitive performance. Future studies will examine this biomarker after SRS to look for individual aging radiation effects and prediction for cognitive changes after therapy.

Identifying and Re-Planning Challenging Retrospective SRS Patients to Evaluate the Potential for a White Matter Tract Sparing Knowledge-Based Planning Model

PURPOSE/OBJECTIVE(S)

Diffusion tensor imaging (DTI) with automated tractography, along with auto-parcellation of volumetric MRI (vMRI), allows for reliable parcellation of critical white matter (WM) and hippocampal structures in the brain. Knowledge-based planning (KBP) may facilitate the optimization of new brain SRS plans that highly prioritize WM tracts and hippocampal sparing, allowing for cognitive preservation. However, building a high-performing model is challenging as the geometry of WM tracts and their proximity to planning target volumes (PTVs) can vary substantially. In this work we propose an iterative step for creating a KBP model where we used a preliminary model to identify challenging plans and worked on improving their dosimetry to refine the model predictions.

MATERIALS/METHODS

Tractography of eloquent deep WM was obtained from high resolution DTI, along with auto-segmentation of bilateral hippocampi from vMRI and an open-source software for processing and analyzing brain MRI images. These ROIs were imported to the treatment planning system as organs-at-risk (OARsDTI). A preliminary KBP model was built from an IRB-approved cohort of 30 consecutive patients (75 plans total) treated for brain metastasis with non-coplanar SRS with total dose from 20 to 30 Gy in one to five fractions using a LINAC. This preliminary model was used to identify the 7 plans with the highest doses to the OARsDTI. These plans were re-optimized with new objectives on the OARsDTI that were partially or totally overlapped with the isodose line representing 30% of the prescription dose (as a surrogate for proximity to the PTV). The OARsDTI that were overlapping with the PTV were excluded from this analysis. We evaluated the re-optimized plans (PlanNew) by comparing maximum and mean dose (Dmax and Dmean, respectively) of OARsDTI to the initial values from the preliminary plans (PlanPre). In order to verify that the PlanNew were not compromising other structures or treatment goals, we also evaluated conformity index (CI), PTV coverage and the dosimetry of OARs included in the PlanPre. Additionally, the PlanNew were normalized to our clinical standard used for the PlanPre (100% of prescription dose to 98% of the PTV).

RESULTS

In the 7 plans, there were 30 OARsDTI overlapping with the 30% isodose line. Refining the dosimetric objectives for the OARsDTI decreased the Dmean on average 31% when comparing PlanNew to Planpre (p<0.05, Wilcoxon rank-sum test). The CI of the PlanNew were not statistically different from the PlanPre. Although Dmax of PlanNew was 12% lower than Dmax of Planpre, the difference was not statistically different.

CONCLUSION

The addition of a plan refinement step was shown to improve dose to the OARsDTI in 7 challenging SRS plans. The generation of these plans are a necessary step to creating a refined KBP model capable of accounting for the high variability in WM tracts geometry and their proximity with PTV.

Evidence-Based Multivariate Normal Tissue Complication Probability (NTCP) Study of Domain-Specific Cognitive Decline after Partial Brain RT

PURPOSE/OBJECTIVE(S)

Beyond the hippocampus, there are no evidence-based dose constraints for eloquent brain structures which subserve memory and attention/processing speed. We performed a multivariate normal tissue complication probability analysis of post-RT neurocognitive decline, examining dosimetric predictors of eloquent brain regions.

MATERIALS/METHODS

Data were analyzed from a prospective longitudinal clinical trial. Patients (n = 78) with primary brain tumors receiving fractionated RT complete a comprehensive neurocognitive evaluation and high-resolution volumetric and diffusion MRI at baseline and 6 months post-RT. Image processing using robust, validated automated segmentation parcellated individual WM tracts, cortical regions, and hippocampi. Well-validated neurocognitive tests including Delis-Kaplan Executive Function System and Wechsler Adult Intelligence Scale-IV coding (attention/processing), Boston Naming Test (language) and Hopkins Verbal Learning Test and Brief Visuospatial Memory Test (verbal/visuospatial memory) were assessed. Reliable change indices adjusted for practice effects (RCI-PE) were calculated for each patient between baseline and 6 months; a negative RCI-PE was scored as decline. Univariate logistic regression was performed with mean and max dose to structures of interest as well as clinical variables. Multivariate model building was performed using automated bootstrapped logistic regression, LASSO and random forest modeling.

RESULTS

On univariate analysis mean and max dose to multiple regions of the corpus callosum (CC) were correlated with attention/processing speed decline; most significantly in WAIS coding, including Dmax to the anterior CC (p = 0.011) and central CC (p = 0.010), and Dmax and Dmean to the mid anterior CC (p = 0.006 and 0.010). Mean dose to the left fornix was associated with decline in memory (p = 0.023, cutoff 12.9 Gy), as were increasing age and both concurrent and adjuvant chemotherapy. On multivariate analysis for attention, automated bootstrapped logistic regression showed the most frequently selected variable was mean dose to the mid anterior CC. Performance at nested cross-validation by AUC was 0.80 (0.75-0.84); LASSO model performance by AUC was 0.76 (0.72-0.81) with Dmean to the mid anterior CC being the most frequent variable. The top five most important variables in the Random Forest as ranked by mean decrease in Gini coefficient were mean dose to mid anterior CC, all white matter, combined CC and max dose to CC and posterior CC. Model performance by AUC was 0.66 (0.60-0.71).

CONCLUSION

Here, we present the first, to our knowledge, NTCP model for decline in attention/processing speed, along with dosimetric predictors of memory decline beyond the hippocampus. We found that after partial brain RT, dose to several ROIs significantly correlated with post-RT impairment. These data can guide future cognitive-sparing strategies for brain RT.