Microglia activation in multiple sclerosis black holes predicts outcome in progressive patients: An in vivo [(11)C](R)-PK11195-PET pilot study

Highlights

• This study uses PK-PET to investigate microglial activity in MS black holes (BHs).

• PK binding (PKBP_ND) in MS BH demonstrates a wide heterogeneity.

• In relapsing MS patients PKBP_ND is significantly lower than in the progressive MS patients (p = 0.00003).

• The baseline PKBP_ND in progressive MS BHs correlates with disability (p < 0.05) and its accumulation 2 years later (p < 0.01).

• PK-PET shows that microglial activation in MS BH indicate different underlying processes in relapsing and progressive MS.

Abstract

The pathophysiological correlates and the contribution to persisting disability of hypointense T1-weighted MRI lesions, black holes (BH), in multiple sclerosis (MS) are still unclear. In order to study the in vivo functional correlates of this MRI finding, we used 11C-PK11195 PET (PK-PET) to investigate changes in microglial activity. Ten relapsing and 9 progressive MS subjects had a PK-PET scan and a MRI scan alongside a full clinical assessment, including the expanded disability status scale (EDSS) for evaluation of disability. We studied the PK binding potential of the specifically bound radioligand relative to the non-displaceable radioligand in tissue (BP_ND) in T1 BHs. Out of a total of 1242 BHs identified, 947 were PK enhancing. The PKBP_ND was correlated with the EDSS (r = 0.818; p < 0.05) only in the progressive group. In the relapsing patients there was an inverse correlation between PKBP_ND and BH total lesion volume in whole brain (r = − 0.781; p < 0.05). When progressive patients were grouped according to the disability outcome at 2 years from the PK-PET scan, the total PKBP_ND in BHs was found to be a significant outcome predictor of disability (p < 0.01). Our findings show that relapsing and progressive patients have heterogeneous patterns of PKBP_ND in T1 BHs and indicate that BHs are not just “holes” representing loss of axons and myelin, but display inflammatory activity in the form of activated microglia. The significant association between PKBP_ND, neurological impairment and outcome in progressive subjects supports a role for activated microglia in disability progression.

Introduction

Multiple sclerosis (MS) is characterised pathologically by focal areas of inflammatory demyelination and variable axonal loss in the central nervous system (CNS) (Reynolds et al., 2011). Magnetic resonance imaging (MRI) demonstrates a variety of changes that have different levels of sensitivity and specificity in MS. Although MRI assessed new/increased T2 lesions and Gd-enhancing T1 lesions are the current preferred outcomes for phase II trials, it is an imperfect aid in our understanding of current symptoms and disability and it has limited utility in predicting outcome (Sicotte, 2011). Black holes (BH) are those MS lesions appearing hypointense at T1-weighted MRI (Truyen et al., 1996). Unlike many other MRI-based measures, BHs in MS have been associated with irreversible disability (Paolillo et al., 1999, Simon et al., 2000, Truyen et al., 1996, van Walderveen et al., 1999b) although this is not a consistent finding (Giugni et al., 1997, Masek et al., 2008, O'Riordan et al., 1998). In keeping with this inconsistency, the evidence concerning pathological processes underlying BHs has been contradictory. BHs have been found to correlate with axonal damage (Brück et al., 1997, van Walderveen et al., 1998, van Walderveen et al., 1999a); the latter study performed in biopsied lesions reported axonal loss and extracellular oedema as the factors that mainly affect the degree of hypointensity on T1-weighted MRI sequences. At variance with these results, a correlation between BHs and axonal loss was not found in biopsies of active lesions from 14 patients with early MS (Bitsch et al., 2001). In the mouse model of MS, experimental autoimmune encephalomyelitis (EAE), T1-weighted hypointensity was correlated with increased cellular infiltrates and reduced myelin content in the lesion centre, with microglial and astroglial activity in the perilesional area (Nessler et al., 2007).

There is a large body of evidence supporting a major role for activated microglia in the pathology of MS, for review see Gao and Tsirka (2011). In the early stages of the disease activated microglia are involved in recruiting naïve T cells, whose activation is initiated by dendritic cells in the CNS acting as local antigen presenting cells (APCs) (McMahon et al., 2005). In turn microglia, as APCs re-stimulate tissue-invading memory T cells, which are essential to sustain the chronic inflammation in MS. In focal white matter (WM) lesions activated microglia are found in increased numbers, alongside blood derived macrophages (Breij et al., 2008). In the later progressive stages of MS chronically activated microglia are associated with neurodegeneration and areas of meningeal inflammation in the grey matter (GM) (Magliozzi et al., 2010), and diffuse activation and axon damage throughout the WM (Howell et al., 2010, Kutzelnigg et al., 2005).

Positron emission tomography (PET) ligand [(11)C](R)-PK11195 (PK) has a high affinity for the translocator protein (TSPO), which is localised to and expressed at high levels in mitochondria of activated microglia (Banati et al., 2000, Venneti et al., 2008, Vowinckel et al., 1997). This ligand has been used in several neurological conditions as a marker of inflammation and is also associated with neurodegeneration (Cagnin et al., 2006).

Microglia activation can be assayed in a sensitive and robust manner using PK with a reference tissue quantitative methodology that is extensively validated against the gold-standard plasma input function that has been used in different centres (Turkheimer et al., 2007, Yaqub et al., 2012). In MS it has been employed to visualise WM and GM abnormalities beyond WM lesions identified using MRI (Banati et al., 2000, Politis et al., 2012). Therefore, in this study we examined PK binding in MS BHs as an indicator of microglial activity and correlated the changes found with disability as measured using the EDSS.