The relation between macro- and microscopic intrinsic muscular alterations of the medial gastrocnemius in children with spastic cerebral palsy

Abstract

Children with spastic cerebral palsy (CP) are characterized by altered muscle growth, secondary to the pathological neural input to the muscular system, caused by the primary brain lesion. As a result, their medial gastrocnemius is commonly affected and is characterized by altered macro and microscopic muscular alterations. At the macroscopic level, the muscle volume (MV), anatomical cross-sectional area of the muscle belly (Belly-CSA), muscle belly length (ML), and the intrinsic muscle quality are reduced. At the microscopic level, the cross-sectional area of the fiber (Fiber-CSA) is characterized by an increased within-patient variability (coefficient of variation (CV)), the fiber type proportion is altered and capillarization is reduced. However, the relations between the muscular alterations at the macro- and microscopic level are not yet known. Therefore, this cross-sectional study integrated macro- and microscopic parameters of the medial gastrocnemius in one cohort of young ambulant children with CP and age-matched TD children, and explored how deficits in macroscopic muscle size are associated with alterations at the microscopic level. A group of 46 children with CP (median age 5.4 (3.3) years) and a control group of 34 TD children (median age 6.3 (3.4) years), who had data on microscopic muscular properties (defined through the histological analyses of muscle biopsies), as well as macroscopic muscle properties (defined by 3D freehand ultrasound) were included. We defined Pearson or Spearman correlations, depending on the data distribution. The macroscopic muscle size parameters (MV, Belly-CSA, ML) showed significant moderate correlations (0.504-0.592) with the microscopic average Fiber-CSA in TD and CP. To eliminate the common effect of anthropometric growth at the macro- as well as microscopic level, the data were expressed as deficits (i.e. z-scores from normative centile curves or means) or were normalized to body size parameters. A significant but low correlation was found between the z-scores of MV with the z-scores of the Fiber-CSA (r=0.420, p=0.006). The normalized muscle parameters also showed only low correlations between the macro- and microscopic muscle size parameters, namely between Belly-CSA and Fiber-CSA, both in the TD (r=0.408, p=0.023) and the CP (ρ=0.329, p=0.041) group. Explorations between macroscopic muscle parameters and other microscopic muscle parameters (capillary density, capillary to fiber ratio, and fiber type proportion) revealed no or only low correlations. These results highlighted the complexity of the interacting network of intrinsic muscle structures, with mainly low associations between the macro- and microstructural level, and it remains unclear how alterations in microscopic muscle structures contribute to the macroscopic muscle size deficits in children with CP.