Increased PK11195-PET binding in normal-appearing white matter in clinically isolated syndrome

Abstract

The most accurate predictor of the subsequent development of multiple sclerosis in clinically isolated syndrome is the presence of lesions at magnetic resonance imaging. We used in vivo positron emission tomography with (11)C-(R)-PK11195, a biomarker of activated microglia, to investigate the normal-appearing white matter and grey matter of subjects with clinically isolated syndrome to explore its role in the development of multiple sclerosis. Eighteen clinically isolated syndrome and eight healthy control subjects were recruited. Baseline assessment included: history, neurological examination, expanded disability status scale, magnetic resonance imaging and PK11195-positron emission tomography scans. All assessments except the PK11195-positron emission tomography scan were repeated over 2 years. SUPERPK methodology was used to measure the binding potential relative to the non-specific volume, BPND. We show a global increase of normal-appearing white matter PK11195 BPND in clinically isolated syndrome subjects compared with healthy controls (P = 0.014). Clinically isolated syndrome subjects with T2 magnetic resonance imaging lesions had higher PK11195 BPND in normal-appearing white matter (P = 0.009) and their normal-appearing white matter PK11195 BPND correlated with the Expanded Disability Status Scale (P = 0.007; r = 0.672). At 2 years those who developed dissemination in space or multiple sclerosis, had higher PK11195 BPND in normal-appearing white matter at baseline (P = 0.007 and P = 0.048, respectively). Central grey matter PK11195 BPND was increased in subjects with clinically isolated syndrome compared to healthy controls but no difference was found in cortical grey matter PK11195 BPND. Microglial activation in clinically isolated syndrome normal-appearing white matter is diffusely increased compared with healthy control subjects and is further increased in those who have magnetic resonance imaging lesions. Furthermore microglial activation in clinically isolated syndrome normal-appearing white matter is also higher in those subjects who developed multiple sclerosis at 2 years. Our finding, if replicated in a larger study, could be of prognostic value and aid early treatment decisions in clinically isolated syndrome.

Keywords: PK11195-PET; clinically isolated syndrome; microglia; multiple sclerosis; normal-appearing white matter.

Figure 1

Normal appearing white matter on PK11195-PET images co-registered and fused with MRI of three subjects of the study population. The first subject is a healthy control (A) with PK11195 BP_ND in normal-appearing white matter of 0.028; the second is a CIS subject without T₂ MRI lesions (B), an EDSS score of 3.5 and PK11195 BP_ND in normal-appearing white matter of 0.037; the third is a CIS subject with T₂ MRI lesions (C), EDSS = 2.0 and PK11195 BP_ND in normal-appearing white matter of 0.136. The colour scale bar represents the BP_ND of PK11195.

Figure 2

Global changes in normal-appearing white matter in subjects with CIS compared to healthy controls. (A) The average of PK11195 BP_ND for each subject represented by a black dot (healthy control PK11195 BP_ND average in normal-appearing white matter is 0.042, whereas it is 0.071 in CIS subjects). (B) Significant difference in normal-appearing white matter (NAWM) regions frequency distribution (y-axis, percentage) according to their PK11195 BP_ND (x-axis) between healthy control and CIS. There is a higher percentage of normal-appearing white matter regions with low PK11195 BP_ND in healthy control (left) than CIS (right). (C) Significant group effect studied using repeated measures with normal-appearing white matter Regions of interest as the repeated factor and Group (left, healthy control; right CIS) as a between subject factor.

Figure 3

Global changes in normal-appearing white matter in subjects with CIS without T₂ white matter lesions at MRI scan compared to CIS subjects with T₂ white matter lesions. (A) The average of PK11195 BP_ND for each subject represented by a black dot (CIS without white matter lesions PK11195 BP_ND average in normal-appearing white matter is 0.032, whereas it is 0.078 in CIS subjects with white matter lesions). (B) The significant group effect studied using repeated measures with normal-appearing white matter regions of interest as the repeated factor and Group (left, CIS without white matter lesions; right, CIS with white matter lesions) as a between subject factor. Inset C shows the significant correlation between PK11195 BP_ND (y-axis) and EDSS (x-axis) in CIS with white matter lesions (r = 0.673); dotted lines represent error bars. NAWM = normal-appearing white matter.

Figure 4

Baseline PK11195 BP_ND in normal-appearing white matter (NAWM) of healthy controls and CIS who remained CIS or were diagnosed with multiple sclerosis (MS) at the 2-year follow-up (FU). (A) The average of PK11195 BP_ND for each subject represented by a black dot (healthy control PK11195 BP_ND average in normal-appearing white matter is 0.042, CIS PK11195 BP_ND average in normal-appearing white matter is 0.045, whereas it is 0.080 in subjects with CIS who developed multiple sclerosis at 2-year follow-up). On the y-axis of the graph in B are listed the 16 subjects who completed the 2-year follow-up ordered according to their normal-appearing white matter PK11195 BP_ND value (x-axis) at the baseline PK11195-PET scan. The filled bars represent the dissemination in space (DIS, present in 13 subjects); in hatched, the dissemination in time (DIT, present in 13 subjects). Patients who developed both dissemination in space and dissemination in time, hence diagnosed with multiple sclerosis (12 subjects) according to the 2010 revision of McDonald criteria for multiple sclerosis diagnosis, are therefore multiple sclerosis labelled on the y-axis. The light grey bars represent the subjects who remained at the CIS stage and did not develop both dissemination in space and dissemination in time.

Figure 5

Coronal section of central structures on PK11195-PET images coregistered and fused with MRI of two subjects of the study population. The first subject is a healthy volunteer (A) with PK11195 BP_ND in grey matter central structures of 0.163; the second is a subject with CIS (B), EDSS = 4.0 and PK11195 BP_ND in grey matter central structures of 0.248. The colour scale bar represents the BP_ND of PK11195.

Figure 6

Baseline PK11195 BP_ND in central structure of healthy controls and CIS subjects. Each data point represents the average of PK11195 BP_ND for each subject (healthy control PK11195 BP_ND average in central structures is 0.167, whereas it is 0.197 in CIS subjects). Horizontal bars represent mean ± SD.