Feasibility of the APRICOT-trial: identifying MRI biomarkers for radiation-induced cognitive changes

Abstract

Purpose or Objective Radiotherapy (RT) is a standard treatment for patients with brain metastases (BMs), although it may lead to radiation-induced cognitive impairment. Since overall survival time is often limited, maintaining adequate quality of life during the remaining life span is pivotal and influences treatment shared decision making. It is currently not possible to predict the impact of RT on individuals as the pathogenesis of cognitive impairment is not fully understood. The main goal of the APRICOT-trial is to identify MRI-derived biomarkers for radiation-induced cognitive impairment three months after RT in BMs patients. The aim of the current pilot-study is to assess the feasibility of the APRICOT protocol. Materials and Methods In the APRICOT-trial a total of 69 patients will receive a 90-minute neurocognitive assessment (NCA) and series of MRI scans to derive vascular parameter maps. These include blood oxygen level-dependent (BOLD) MRI for cerebrovascular reactivity (CVR) during elevated CO2 breathing and blood volume/metabolic information during O2 breathing, multi-delay ASL for arterial arrival time and perfusion, resting-state for temporal signal to noise (tSNR) and low-frequency fluctuation data (fALFF), and finally T2FLAIR, T1w and susceptibility weighted imaging (SWI). Data was processed using in-house analysis pipelines to yield a series of parameter maps that were spatially co-localized with RT dose and contoured target volumes. Both the NCA and MRI are performed pre-RT and three months post-RT. The data of the first five patients with complete follow-up will be examined to assess feasibility of the protocol in BMs patients and to identify changes in brain tissue status. Results Four months after study start, six of ten eligible patients (56-81yrs; 3f) completed pre-RT measurements. Three patients also returned for post-RT measurements. All six patients completed the entire NCA. The CO2/O2 breathing challenges were well-tolerated by all but one patient. Parameter maps revealed lower CVR and longer hemodynamic lags, indicating vascular impairment in tumor and tumor-surrounding regions (see Fig1). Affected regions correlate with hyper-intensities in the T2Flair/tSNR and hypo-intense regions in the O2 maps; indicating low perfusion or blood volume or potential disease-related neuronal dysfunction (fALFF). Combining these MRI-derived parameters allows to identify subtle changes in brain tissue status, that are not visible on conventional MRI. Conclusion BMs patients pose a cognitively vulnerable population with limited survival. Nonetheless, this preliminary data indicates that performing a NCA and elaborate MRI pre- and three months post-RT is feasible and well-tolerated by the majority of patients. Moreover, the first MRI-derived parameters indicate that it is possible to identify subtle vascular changes both in the metastases and the rest of the brain pre-RT.