CORDIS - EU research results

NORmalize MusculoskeletAL LOadings to Avoid bony Deformities in children with cerebral palsy

Periodic Reporting for period 1 - NORMAL-LOAD (NORmalize MusculoskeletAL LOadings to Avoid bony Deformities in children with cerebral palsy)

Reporting period: 2018-03-01 to 2020-02-29

Cerebral palsy is a collection of clinical syndromes caused by a non-progressive lesion in the immature brain. This lesion alters neural patterns of muscle activation leading to progressive deterioration of the musculoskeletal system. Muscle contractions in children with cerebral palsy cause abnormal musculoskeletal loading, which can lead to bony deformities. Femoral deformities, e.g. increased neck-shaft angle and increased anteversion angle, are very common in children with cerebral palsy and can lead to pathological gait pattern due to abnormal muscle-moment arms. These deformities are usually corrected by de-rotation osteotomies, which are severe invasive orthopedic intervention.
Growing bone is extremely responsive to mechanical loading. Hence, in children early clinical interventions, which aim to normalize the loading of the musculoskeletal system, could influence bone growth and prevent the development of bony deformities. Muscle and joint contact force cannot be measured in a non-invasive way but musculoskeletal simulations based on a person’s gait pattern can be used to calculate these forces. To investigate if clinical interventions have an impact on bone growth in children with cerebral palsy, a multi-scale modelling approach was developed and used to predict femoral growth trends before and after clinical interventions.

The overall objective of this project was to evaluate if early clinical interventions have the potential to alter musculoskeletal loading conditions and therefore prevent femoral deformities in children with cerebral palsy.

Importance for society
Cerebral palsy is the most common pediatric neurologic disorder with a prevalence of 2-3 cases per 1,000 live births in Europe and has estimated healthcare and socio-economic lifetime costs of €800,000 to €860,000 for each affected person. This project aimed to develop a workflow to calculate femoral loadings and predict femoral growth trends in children with cerebral palsy. The promising results from this project (described below) are likely to have an influence on clinical decision-making in children with cerebral palsy in the future. If early load-modifying clinical interventions can be used to reduce the number of surgical corrections, it will decrease the burden on the child as well as decrease the socio-economic costs related to the treatment of children with cerebral palsy.
In this project, previously collected motion capture data was analyzed and used to calculate musculoskeletal loadings. Following three participant groups were analyzed:
• 10 typically developing children
• 14 children with cerebral palsy before and after receiving Botulin Toxin-A injections
• 25 children with cerebral palsy before and after receiving a selective dorsal rhizotomy

Many children with cerebral palsy have an abnormal motor control. Considering that the standard musculoskeletal modelling workflow uses an approach to calculate muscle forces, which does not differentiate between normal and pathological motor control, we also investigated the impact of including electromyography data to account for the subject-specific motor control on the musculoskeletal simulation results.

Most generic musculoskeletal model are based on a bony geometry of an adult person. Hence, we also investigated how a modified child-specific femoral geometry in a musculoskeletal model influence hip joint contact forces, which have the biggest impact on femoral bone growth simulations.

During the development of the multi-scale bone growth prediction workflow we had to address several technical question. First, we developed a workflow to create a subject-specific finite element model of the femur of a child based on magnetic resonance images. This workflow, however, was too slow to create a patient-specific model for each participant. Hence, we developed a workflow in which we could morph the finite element model to different femoral geometries (Fig. 1). Using this workflow, we showed how the femoral geometry influence growth predictions. Furthermore, in a study based on simplified loading conditions, we compared two growth direction methods and showed that only the ‘average neck deflection’ method leads to reasonable results.
In the final study we showed that Botulinum-Toxin A injections have a minor impact on femoral bone growth. Furthermore, we showed that selective dorsal rhizotomy has a positive impact on femoral bone growth. Additionally, we investigated difference between children with cerebral palsy who have typically developing bone growth and who have pathological femoral growth. Our findings showed that the main difference is caused by a different pelvic and hip movement strategy during mid-stance of the gait cycle, which leads to a less posterior oriented hip joint contact force in children with pathological femoral growth.

Dissemination of results
The project led to five peer-reviewed full papers in top international journals (2 published, 2 under review, 1 in preparation), four published conference abstracts and six conference presentations. Furthermore, I was selected as a finalist for the ‘Clinical Biomechanist Award’ from the European Society of Biomechanics. Hence, in July 2020 I will present the findings of this project in a webinar (due to COVID-19) together with three more finalist.
Following unique key findings led to a progress in the field of clinical biomechanics beyond the state of the art:

We showed how the femoral geometry, i.e. neck-shaft angle and femoral anteversion angle, influences the magnitude and orientation of the hip joint contact force. Interestingly, the orientation of the hip joint contact force followed the orientation of the femoral neck, e.g. increased neck-shaft angle led to a more vertical oriented hip joint contact force. This finding is of high importance, especially when comparing hip joint contact forces between people with different femoral geometries.

We conducted femoral bone growth simulation in 6 typically developing children and 16 children with cerebral palsy and, therefore, overcame the limitations of previous studies, which were limited to small sample sizes (n between 1 and 4). We could show that the osteogenic index, which defines the growth rate and is based on principal stresses at the growth plate, is very consistent between typically developing children. Furthermore, in children with pathological bone growth, we found an altered osteogenic index, which was likely caused by a different pelvis and hip movement strategy during the mid-stance of the gait cycle.

No previous multi-scale modelling study investigated the impact of clinical interventions on bone growth. Hence, this project was the first in which the impact of two common clinical interventions on femoral growth was analysed.

Our femoral growth prediction were in agreement with experimental studies, in which femoral geometry was measured in a large sample of typically developing children and children with cerebral palsy. Hence, the developed workflow has the potential to influence clinical decision-making in children with cerebral palsy in the future. However, further validation of the workflow, based on medical images collected on two occasions of the same participants, is necessary before the workflow can be used in prospective clinical studies. I got a research grant to collect the necessary data for it and plan to conduct the required validation study in the near future.
Fig. 1. Multiscale modelling workflow used to predict femoral growth trends.