Inflammation and atrophy in multiple sclerosis: MRI associations with disease course

Abstract

Brain atrophy may be a useful surrogate marker of axonal loss and disease progression in multiple sclerosis (MS). Several studies have suggested that inflammatory disease activity is a risk factor for atrophy in the early stages of the disease, but may become less important later in the disease course. We aimed to investigate the relationships between atrophy and active inflammation at different stages of the disease course using brain volume measurements from magnetic resonance imaging (MRI) in patients with both relapsing–remitting (RR) (n=95) and secondary progressive (SP) (n=76) MS. Conventional dual echo and three-dimensional magnetization-prepared rapid-acquisition gradient echo imaging were performed in all patients and in 31 healthy controls. Supratentorial and infratentorial brain, and lateral ventricular volumes were determined using modern design stereology.

Patients with SP MS had smaller supratentorial (p=0.003) and infratentorial brain volumes (p=0.0003), and larger lateral ventricles (p=0.02) than patients with RR MS. RR MS patients with T₁-enhancing lesions had smaller supratentorial (p=0.02) and infratentorial (p=0.002) brain volumes and larger ventricles (p=0.002) than those without enhancing lesions. SP MS patients with enhancing lesions also had significantly larger lateral ventricles (p=0.03). Categorical analysis showed that more RR MS patients with enhancing lesions had smaller supratentorial brain (p=0.005), or larger lateral ventricular (p=0.028) volumes, and more SP MS patients with enhancing lesions had increased lateral ventricle volumes (p=0.013), than patients without enhancements. The number of enhancing lesions was significantly correlated with lateral ventricular volumes in both RR MS (r=0.39, p=0.0001) and SP MS (r=0.46, p<0.0001). Our data shows that the presence of active inflammation on a single MRI in the course of RR and SP MS, is associated with a higher risk and higher level of brain atrophy. These findings emphasise the important long-term relationship between inflammation and atrophy in MS and provide additional support for the strategy of early anti-inflammatory treatment to protect tissue integrity.

Introduction

Axonal loss is a well-documented pathological feature of multiple sclerosis (MS) [1], [2] and likely to be a critical factor in determining fixed neurological disability [2], [3]. The presence of brain or spinal cord atrophy in MS may be a simple, surrogate marker of axonal loss and indirectly, disease progression [4], [5], [6], [7], [8]. Several cross-sectional studies have shown that neurological disability correlates with spinal cord area or volume [8], [9], [10], atrophy of infratentorial structures [10], [11] and increased intracranial cerebrospinal fluid space [12].

It has been reported that axonal damage and transection occur early in the course of the disease and associate strongly with areas of active inflammation [2], [3]. These observations are consistent with reports that disease activity on magnetic resonance imaging (MRI) [4], [6], [13] and relapse rate [14], correlate with the rate of brain atrophy. Natural history [15] and MRI studies [16] have demonstrated that events in the first few years have a significant influence on disability levels later in the disease course. In addition, there is evidence that axonal loss and atrophy may be occurring at a faster rate in the first few years of the disease [14], [17]. On the other hand, relapses and MRI activity become less marked with time and the anti-inflammatory disease modifying drugs appear to lose effectiveness on disability progression in the secondary progressive phase [18], [19], [20]. Further, progressive disability appears not to be affected by relapses once a clinical threshold of irreversible disability has been reached [21]. Inflammation may be the pivotal event that causes demyelination and axonal transection, but axon degeneration may become the dominant pathological process in the secondary progressive phase [22]. The relationship between inflammation and atrophy may therefore differ at different stages of the disease course. To investigate this relationship further, we aimed to measure disease activity and structural brain volumes using three-dimensional (3-D) acquired MRI data in a large cohort of MS patients with both relapsing–remitting (RR) and secondary progressive (SP) disease.