Myotonic dystrophy type 2 (DM2) is an autosomal dominant hereditary multisystem disease that is characterized by myotonia, muscle weakness, and early-onset cataracts (before the age of 50) (1). It is the most common adult-onset muscular dystrophy in Central and Northern Europe (Finland) with an estimated prevalence of about 9 in 100,000 (2). Patients with DM2 typically experience weakness in the proximal limb muscles, with a proximo-distal pattern of progression (3). Truncal (axial) muscles are also affected. Thus, DM2 can be classified as an axial myopathy with significant paraspinal involvement as part of a more widespread myopathy (4). Examination of the paraspinal muscles (both clinical and MRI) in patients with DM2 is often unjustly neglected, although proper functioning of these muscles is essential. These postural muscles stabilize the spine and play a crucial role in daily mobility and spinal health (5, 6). Significant dysfunction of the lumbar extensor muscles (LEM), which include the lumbar paraspinal muscles (LPM), in DM2 was shown in previous study. Reduced LEM strength was an independent risk factor of the frequent occurrence of chronic low back pain in these patients (7).

Magnetic resonance imaging (MRI) of muscles has become a very important tool in diagnosing and monitoring disease progression and treating neuromuscular disorders including myopathies (8, 9). In general, there has been an effort to identify MRI biological markers (biomarkers) that would be objective indicators of the muscle structure, reflect pathological processes in the muscle, and determine the severity of the disease. Currently, the percentage of intramuscular fat (fat fraction – FF), the muscle cross-sectional area or total muscle volume (TMV) and functional muscle volume (FMV), that is easily derived from TMV and FF (10), are considered as qMRI (quantitative MRI) muscle biomarkers, with FF and FMV as key biomarkers (11).

Few studies have evaluated muscle MRI findings in patients with DM2 (12–15) and only one study evaluated the qMRI parameters of lower limb muscles in DM2 patients (12). Previous reports of muscle MRI in DM2 suggest that the erector spinae, gluteus maximus muscle, and thigh muscles (slightly more pronounced in the posterior compartment) are primarily affected by increased fat replacement of muscle tissues (12–15). However, it has been documented that patients with DM2 frequently exhibit normal muscle MRI findings, with muscle fat infiltration only becoming evident in the late stages of the disease (13, 14). When assessing the morphology of the LPM, it is important to remember that it is influenced by several physiological variables, mostly by age, sex, and anthropometric parameters (16–21).

An unresolved issue is the correlation between morphological MRI parameters and muscle function. It has been reported that the performance of muscles is largely dependent on muscle mass and muscle composition (10).

The aims of this study were; (1) to evaluate qMRI parameters of the LPM in patients with DM2 and compare them to healthy volunteers (HV); (2) to explore the relation between qMRI parameters of LPM and functional LEM parameters in DM2; and (3) to evaluate whether dysfunction and morphological parameters of LPM in patients with DM2 additionally worsen with age compared to HV, suggesting progressive impairment in these muscles. The results of this study will help to determine whether qMRI parameters of LPM can be reliable biomarkers of LEM dysfunction and disease progression in DM2.

This was a single-center, observational, cross-sectional, comparative study. The study protocol (agreement number 05-090621/EK) was approved by the local institutional medical research ethics committee, and all participants provided written informed consent.

This study focuses on the evaluation of qMRI parameters of the LPM using an advanced MRI method (6-point Dixon gradient echo sequence) in patients with DM2, their correlation with functional examination, and the evolution of these parameters with aging. DM2 patients showed increased FF of LPM compared to matched HV. The FMV of LPM correlated significantly with strength (MILEMS), and the FF of LPM correlated significantly with endurance (LEME); however, these radiological parameters cannot be used as reliable biomarkers of LEM functional impairment in DM2 patients as the correlation coefficients and variability of MILEMS and LEME explained by radiological parameters were relatively low. Further, there was no significant additional worsening of the functional and morphological parameters of LPM in DM2 patients with increasing age compared to HV; therefore these parameters cannot be considered as biomarkers of disease progression.

To the best of our knowledge, qMRI parameters of LPM in DM2 have not yet been evaluated and published; only Kornblum et al. described frequent fatty degeneration of the ES in DM2, graded semi-quantitatively by the Mercuri scale on whole-body MRI (14). In our study, all the muscles evaluated showed higher FF, a typical finding that has been described in different types of muscular dystrophies, including DM2 (11–13, 30). The muscle volumes (TMV and FMV) of LPM were not significantly different in DM2 patients compared to HV, but the muscle volumes of PS were significantly smaller. This distinctive behavior of LPM volumes compared to PS volumes can be explained by the different representation of muscle fibers in individual muscles. A typical feature of LPM is the predominance of slow-twitch fibers (type 1); while PS has a predominance of fast-twitch fibers (type 2) (31, 32). In DM2 patients, mainly the atrophy of type 2 fibers is described (33–35). The literature describes that the fat infiltration of skeletal muscles often occurs without loss of muscle volume (30). It has also been shown that the onset and progression of changes in muscle volume and fatty infiltration are distinct, independent pathological processes differing between muscle groups in patients with muscular dystrophy (36).

Dahlquist et al. mentioned parameter muscle contractility (strength per fat-free cross-sectional area), that is usually reduced in muscular dystrophies and reflects the quality of the muscle composition (11). In our study, we used the parameter MILEMS to FMV ratio as FMV better reflects contractile muscle size than a fat-free (lean) cross-sectional area. We have shown that DM2 patients have a significantly lower MILEMS to FMV ratio in the LPM, which can be interpreted as 1 cm³ of fat-free LPM generates less strength in DM2 patients compared to HV. This indicates reduced muscle quality in DM2.

Dahlquist et al. described that muscle fat content negatively correlated with muscle function in neuromuscular diseases and observed that changes in fat content precede changes in function (11). Far fewer studies have evaluated the relationship between strength and muscle volume, most likely because muscle volume is more difficult to determine than FF, requiring manual muscle segmentation or a sophisticated automatic segmentation method using artificial intelligence. The literature suggests that muscle strength correlates more strongly with FMV than with FF (10). This is supported by our study in which MILEMS correlated with the FMV of LPM, whereas MILEMS showed only a trend to decrease with increasing FF, which was not statistically significant. In contrast, LEME correlated more strongly with FF compared to the FMV of LPM in DM2.

The relationship between functional and qMRI parameters is complicated and discordance was reported in DM2 patients (13, 14). There are numerous possible explanations. First, fatty degeneration may not always result in a loss of muscle strength or a decrease in functional status (8). Second, in addition to fat replacement in muscular dystrophies, muscle mass is replaced by connective tissue, directly impacting the muscle function. Fibrosis is often distributed around muscle fibers and its imaging is very challenging (10, 11). As fibrous tissue is composed mainly of collagen fibers, which appear as non-fat tissue on 6-point Dixon chemical shift imaging, making it indistinguishable from muscle, it may be included in the FMV and significantly influence FMV-strength correlations. Different imaging protocols have been suggested for its differentiation, particularly gadolinium contrast enhanced imaging, which is widely used in myocardial imaging. However, the need to apply a contrast agent is a disadvantage here; moreover, dystrophic muscle fibers in neuromuscular disorders could lead to the contrast agent penetrating into the muscle itself (10). Therefore, gadolinium contrast enhanced imaging is not suitable for fibrosis imaging in neuromuscular disorders. More advantageous seem to be methods using ultra-short echo times or magnetic resonance elastography (10, 30, 37). Last but not least, errors in muscle segmentation and/or MILEMS examination can occur, even though we proved these methods to be reliable in previous studies (21, 22). It should be noted that MILEMS and LEME are not the result of isolated LPM or LEM activity, but can be influenced by the involvement of other muscle groups.

Madrid et al. showed that in DM2 patients, the FF of the affected muscles of the lower extremities correlated with motor performance. The FF of the lower extremity muscles was thus considered as a potential biomarker of disease severity (12). In our study, despite the demonstrated statistically significant correlation between FMV and MILEMS and FF and LEME, we cannot consider qMRI parameters as reliable biomarkers of LPM impairment as the confidence intervals were relatively wide and correlation coefficients and variability of functional parameters explained by FF and FMV were relatively low. The discrepancy with the Madrid et al. study is probably due to the different selection of muscles analyzed (LPM vs. leg muscles) (12).

It is reported that symptoms such as muscle weakness and myotonia worsen progressively with increasing age in DM2 (38). Peric et al. (13) analyzed the FF of leg muscles in DM2 patients in relation to different disease durations; we do not consider this method accurate. The definition of disease duration in DM2 patients is questionable. Disease duration can refer to either the duration of symptoms, which is an anamnestic parameter that can be very imprecise, or the duration of the disease since genetic confirmation of the diagnosis, which is not reflective of the disease duration because there is often a significant diagnostic delay (Peric et al. reported diagnostic delays of 12.4 ± 10.9 years) (13). To assess the progression of LPM impairment in patients with DM2 over time, we compared the evolution of LPM parameters with aging between patients and HV. This approach was selected because muscle parameters change with age even in healthy people, with a significant increase in FF with aging, as shown in this study. We did not find a significant difference in the rate of deterioration in functional parameters (MILEMS, LEME) or MRI parameters (FF, FMV) of the LPM in patients with DM2 compared to HV. On the other hand, DM2 patients showed a faster increase in the FF of PS with age compared to HV. This suggests more pronounced radiological and probably also functional worsening of PS impairment in DM2 with age. It can be concluded that the evolution of LPM impairment in DM2 patients does not significantly differ from natural age-related structural and functional changes. Therefore, qMRI parameters of LPM and/or changes in MILEMS and LEME cannot be used as a biomarker of DM2 progression. On the other hand, the FF of PS showed an additional progressive increase with age in DM2 patients. This probably reflects disease progression, but it would require further analysis and correlation with functional muscle testing of PS, which was not the focus of this study. The different behavior of the LPM and PS during aging in DM2 probably arises again from different muscle fiber representation in these muscles and the atrophy of predominantly type 2 fibers in DM2 (33–35).

This study has limitations. Most importantly, as DM2 is an orphan disease, the number of patients is limited. This may affect the power of statistical tests and lead to greater sensitivity to individuals with outlier parameters. However, this study evaluated muscle MRI from the largest number of DM2 patients – the number of patients in previously reported studies with DM2 was only as high as 14 (12–15). Second, the exclusion criteria led to the recruitment of patients with only mild functional impairment. Also, the maximum age of our DM2 group subjects was 67 years, as older patients failed to meet these strict criteria. Therefore, conclusions cannot be generalized for the whole population of patients with DM2, but rather for a selected cohort of young to young-old patients with mild DM2.

This is the first study evaluating in detail the LPM in DM2 patients, which has been neglected and only marginally mentioned in the literature, despite the importance of these muscles in maintaining upright posture and spinal health. Further, functional and qMRI parameters of the LPM of DM2 patients were compared with a cohort of HV of the same sex and similar age, weight, and height. This approach is very important because the morphological and functional parameters of the muscles vary significantly with the physiological parameters of the individual. The contribution of this study also lies in the chosen methodology of the MRI examination. In order to obtain the most accurate qMRI parameters, we used a 6-point Dixon gradient echo sequence and performed segmentation of ES, MF, and PS over the entire muscle, although it was time consuming.

This study supports the classification of DM2 as an axial myopathy; this should be considered when managing patients. We believe that patients may benefit from physiotherapy focused on strengthening the LEM and improving the coordination of the deep stabilization system. A well-chosen exercise program may lead to a reduction in low back pain, for which LEM dysfunction is a risk factor (7). LPM dysfunction does not evince additional progressive worsening with aging in patients with DM2 compared to HV. Therefore, camptocormia and neuromuscular scoliosis will not be encountered in these patients or only very rarely (39); this is in line with common clinical practice. As DM2 is an under-diagnosed condition, we recommend keeping it in mind in the differential diagnosis of patients with more extensive LPM fatty infiltration on MRI than would be expected given the patient’s age and sex. This is particularly relevant in the evaluation of lumbar spine MRI in patients with chronic low back pain. Further studies using advanced MRI are needed to evaluate fibrosis as a potential biomarker of DM2 severity and progression.