Geostatistical Analysis of White Matter Lesions in Multiple Sclerosis Identifies Gender Differences in Lesion Evolution.

Multiple sclerosis (MS) is a chronic inflammatory demyelinating disease of the central nervous system with presumed autoimmune origin. The development of lesions within the gray matter and white matter, which are highly variable with respect to number, total volume, morphology and spatial evolution and which only show a limited correlation with clinical disability, is a hallmark of the disease. Population-based studies indicate a distinct outcome depending on gender. Here, we studied gender-related differences in the evolution of white matter MS-lesions (MS-WML) in early MS by using geostatistical methods. Within a 3 years observation period, a female and a male MS patient group received disease modifying drugs and underwent standardized annual brain magnetic resonance imaging, accompanied by neurological examination. MS-WML were automatically extracted and the derived binary lesion masks were subject to geostatistical analysis, yielding quantitative spatial-statistics metrics on MS-WML pattern morphology and total lesion volume (TLV). Through the MS-lesion pattern discrimination plot, the following differences were disclosed: corresponding to gender and MS-WML pattern morphology at baseline, two female subgroups (F1, F2) and two male subgroups (M1, M2) are discerned that follow a distinct MS-WML pattern evolution in space and time. F1 and M1 start with medium-level MS-WML pattern smoothness and TLV, both behave longitudinally quasi-static. By contrast, F2 and M2 start with high-level MS-WML pattern smoothness and medium-level TLV. F2 and M2 longitudinal development is characterized by strongly diminishing MS-WML pattern smoothness and TLV, i.e., continued shrinking and break-up of MS-WML. As compared to the male subgroup M2, the female subgroup F2 shows continued, increased MS-WML pattern smoothness and TLV. Data from neurological examination suggest a correlation of MS-WML pattern morphology metrics and EDSS. Our results justify detailed studies on gender-related differences.