O15 Large-scale Proteomics Profiling of Peripheral Blood of DM1 Patients Identifies Biomarkers for Disease Severity and Physical Activity

Myotonic dystrophy type 1 (DM1) is a progressive multisystem disease with large heterogeneity in disease onset, symptom development, progression rates and severity. This poses significant challenges to the design of clinical trials. To overcome some of these challenges, this study aims to identify proteomic biomarkers that can serve as measures of therapy response. We profiled 450 peripheral blood samples from adult DM1 patients participating in the OPTIMISTIC trial using bottom-up mass spectrometry-based proteomic profiling with data independent acquisition (DIA). This resulted in the detection of 2702 unique peptides from 263 protein groups in at least 80% of the samples. Linear mixed effect models were used to study associations of protein groups with the effect of Cognitive Behavioural Therapy (CBT), the CTG-repeat length and 27 different outcome measures. A machine learning based algorithm (bootstrap enhanced Multivariate Elastic-Net regression) has been implemented to identify a combined minimal subset of protein groups that were associated with both the disease pathology (CTG-repeat) and (the capacity for) physical activity (Six-Minute-Walk-Test, 6MWT). We observed 165 protein groups significantly associated with CTG-repeat length and 73 with the 6MWT score, with an overlap of 47. Other aspects of DM1, like cognitive capacity and fatigue, were not reflected by the proteomic fingerprints. Remarkably, overall disease severity was mostly associated with hypogammaglobulinemia, reported earlier for DM1, whereas reduced physical activity was mostly associated with increased levels of complement factors. The multivariate Elastic-Net framework reduced the number of shared predictors to 10 non-redundant protein groups, which together explain up to 31% of the variance in CTG-repeat expansion and 27% in 6MWT scores. Although the OPTIMISTIC trial demonstrated positive clinical effects of CBT on social participation and the (capacity) for activity of DM1 patients, no strong proteomic profile was associated with the therapy response after 10 months. DM1 induces a broad disease fingerprint in the peripheral blood proteome, predominantly affecting proteins of the immune system. We suggest a panel of carefully selected proteins for monitoring disease progression and physical activity, which deserves validation in longitudinal studies spanning larger time frames than the OPTIMISTIC study. The ReCognitION project was funded through the European ERA-NET rare disease research program (grant N° 643578). Myotonic dystrophy type 1 (DM1) is a progressive multisystem disease with large heterogeneity in disease onset, symptom development, progression rates and severity. This poses significant challenges to the design of clinical trials. To overcome some of these challenges, this study aims to identify proteomic biomarkers that can serve as measures of therapy response. We profiled 450 peripheral blood samples from adult DM1 patients participating in the OPTIMISTIC trial using bottom-up mass spectrometry-based proteomic profiling with data independent acquisition (DIA). This resulted in the detection of 2702 unique peptides from 263 protein groups in at least 80% of the samples. Linear mixed effect models were used to study associations of protein groups with the effect of Cognitive Behavioural Therapy (CBT), the CTG-repeat length and 27 different outcome measures. A machine learning based algorithm (bootstrap enhanced Multivariate Elastic-Net regression) has been implemented to identify a combined minimal subset of protein groups that were associated with both the disease pathology (CTG-repeat) and (the capacity for) physical activity (Six-Minute-Walk-Test, 6MWT). We observed 165 protein groups significantly associated with CTG-repeat length and 73 with the 6MWT score, with an overlap of 47. Other aspects of DM1, like cognitive capacity and fatigue, were not reflected by the proteomic fingerprints. Remarkably, overall disease severity was mostly associated with hypogammaglobulinemia, reported earlier for DM1, whereas reduced physical activity was mostly associated with increased levels of complement factors. The multivariate Elastic-Net framework reduced the number of shared predictors to 10 non-redundant protein groups, which together explain up to 31% of the variance in CTG-repeat expansion and 27% in 6MWT scores. Although the OPTIMISTIC trial demonstrated positive clinical effects of CBT on social participation and the (capacity) for activity of DM1 patients, no strong proteomic profile was associated with the therapy response after 10 months. DM1 induces a broad disease fingerprint in the peripheral blood proteome, predominantly affecting proteins of the immune system. We suggest a panel of carefully selected proteins for monitoring disease progression and physical activity, which deserves validation in longitudinal studies spanning larger time frames than the OPTIMISTIC study. The ReCognitION project was funded through the European ERA-NET rare disease research program (grant N° 643578).