Investigating chemotherapy induced peripheral neuropathy (CIPN) and its treatment, using functional Magnetic Resonance Imaging (fMRI)
Background Chemotherapy Induced Peripheral Neuropathy (CIPN) is a debilitating neuropathy caused by commonly used chemotherapeutics. Clinically, the problem of CIPN is compounded by difficulties with diagnosis and limited treatment options. The pathophysiology of CIPN remains elusive, with current mechanistic postulates focused mainly on the peripheral nervous system. However, animal and human models of non-CIPN neuropathic conditions have shown the brain to be central to the development and maintenance of painful neuropathy. Moreover, evidence suggests that aberrant activity in key regions of the brain and brainstem could denote individual vulnerability for chronic pain states. The impact of the brain on CIPN development is unknown. Assessment of drug efficacy using brain imaging can provide sensitive readouts and is increasingly used in clinical trials. Aims Firstly, to prospectively explore the structure and function of the brain in cancer patients prior to chemotherapy administration, using functional magnetic resonance imaging (fMRI), in order to determine whether baseline differences exist between patients who progress to CIPN as compared to those who do not. Secondly, to develop a pilot study using fMRI to investigate a topical treatment for CIPN, in order to assess the feasibility of setting up a study with this kind of design. Methods To address the first aim of this thesis a prospective cohort study (the CIPN fMRI Study) was developed. Cancer patients scheduled to receive neurotoxic chemotherapy treatment including oxaliplatin, carboplatin, carbotaxol, or cisplatin, were recruited from three NHS trusts in Scotland, to undergo a high resolution (3 tesla) functional MRI scan, at a single time point prior to commencement of chemotherapy. During the scan structural, resting state and functional data were collected. Functional data involved the presentation of punctate stimuli (using a 256mN von Frey filament), above the patients’ right medial malleolus. While receiving the punctate stimuli, patients viewed images that had neutral or positive emotional content or a baseline coloured image with no content. Sample size was based on previously successful pain fMRI studies and pragmatic estimates. Acute CIPN was defined clinically by common toxicity criteria as necessitating a chemotherapy dose reduction or cessation. Data were analysed using FMRIB’s Software Library (FSL) version 5, 2015. Standard data pre-processing (brain extraction, registration, B0 unwarping, motion correction, and denosiing with FIX) was carried out. Structural analysis was conducted using FIRST. Resting state analysis utilised FSL’s MELODIC tool, and a non-parametric group comparison was made following a dual regression approach. FEAT was used for both first and second level functional analyses. Group comparisons were made using a mixed effects analysis (z threshold 2·3 and 2, regions considered significant at p<0·05, cluster corrected). The group was split by sex to explore known sex differences in pain processing. To address the second aim of this thesis, a pilot fMRI randomised controlled trial (MINT3 Study) was designed. Approvals from ethics and research and development were sought and obtained. Data collection forms were developed. An fMRI experiment was proposed and a single pilot scan was conducted and analysed. Results 30 patients were recruited for the CIPN fMRI study (mean age 60·4 years, [95% Confidence Interval: 57.4-63.4, 17 women). Two patients had lung cancer, nine had gynecological malignancies and 18 had colorectal cancer. 17 patients developed acute CIPN. Structural analysis showed that patients who developed CIPN had a smaller volume of the Nucleus Accumbens (NAc). Resting state analysis did not show clear differences between those who developed CIPN and those who did not. Finally, functional analysis showed that patients who did not develop CIPN had greater activation in the superior frontal gyrus when viewing positive emotional images as compared to those who did progress to CIPN. Region of interest analysis showed that female patients who developed CIPN had greater activity in their mesencephalic pontine reticular formation (MPRF). Male patients who progressed to CIPN had decreased activity in their thalamus. Feasability of the MINT3 study set up and fMRI paradigm was assessed. Interpretation Differences in brain structure and function are evident between patients who developed CIPN and those who did not. Crucially, the regions identified, in particular the NAc, have been postulated to denote a vulnerability for progression to pain states. Although the findings need further confirmation they suggest a paradigm shift in terms of CIPN as a clinical problem. Specifically, it appears that certain individuals can be considered as having increased risk of CIPN development prior to chemotherapy administration. This risk relates to the baseline structure, and function of their brains. Finally, the set up of the MINT3 fMRI study showed that this kind of study design is acceptable in terms of ethical and R&D approvals and a single healthy volunteer pilot.