Natural History of Adult-Onset Myotonic Dystrophy Type 1

Abstract

Background and Objectives

Adult-onset myotonic dystrophy type 1 (DM1) is a progressive, multisystemic disorder, characterized by distal muscle weakness and myotonia. As disease-modifying therapies emerge, robust and sensitive outcome measures are urgently needed. This study characterized the natural history of adult-onset DM1 over 2 years and evaluated quantitative MRI (qMRI) and clinical outcome measures for clinical trials.

Methods

This prospective, monocentric study assessed patients with genetically confirmed adult-onset DM1 and age-matched and sex-matched healthy controls at baseline and 12, 18, and 24 months. qMRI measured proton density fat fraction (PDFF, %) and T2_H2O (ms) of 18 proximal and 10 distal leg muscles using 3D segmentation. Clinical outcomes included 32-item motor function measure (MFM32), 6-minute walk distance (6MWD), 10-meter walk test, 30-second sit-to-stand, 9-hole peg test, hand opening time, MRC sum score, hand grip and pinch dynamometry, and peak cough flow. Patient-reported outcome measures included DM1-ActivC, Individualized Neuromuscular Quality of Life (INQoL), Brief Pain Inventory, and Fatigue and Daytime Sleepiness Scale.

Results

Thirty patients (median 38 years, 60% female, Muscular Impairment Rating Scale 3–4) and 30 matched controls were included. Baseline PDFF (%) was significantly higher in patients with distal (31.8% vs 5.9%, p < 0.001) and proximal (13.0% vs 7.6%, p < 0.001) leg muscles. T2_H2O was significantly elevated in the distal leg muscles (+3.3, p < 0.001). Over 24 months, PDFF (%) increased significantly in 9 of 10 distal and 15 of 18 proximal muscles. Distal leg PDFF (%) showed the largest progression (+3.5%, 95% CI 2.9–4.1, standardized response mean [SRM] = 1.6), with significant changes detectable within 6-month intervals (+1.0%, 95% CI 0.4–1.6, SRM = 1.1). T2_H2O remained stable. MFM32 (−3.2, −6.7, −7.3; SRM = 0.8, 1.9, 1.8) and MRC sum score (−5.0, −7.6, −9.6; SRM = 1.1, 1.4, 1.6) declined at 12, 18, and 24 months (p < 0.001). DM1-ActivC declined significantly at 12, 18, and 24 months (24 months: −10.3, 95% CI −14.0 to −6.7; SRM = 0.8). INQoL overall quality of life deteriorated significantly at 12 months (+5.4, 95% CI 0.8–10.0) and 18 months (+6.9, 95% CI 2.0–11.8), but not at 24 months. Distal leg PDFF (%) correlated strongly with 6MWD (ρ = −0.81) and MFM32 D1 (ρ = −0.75).

Discussion

PDFF (%) is a sensitive, objective outcome measure for adult-onset DM1, detecting change within 6-month intervals. MFM32, MRC sum score, DM1-ActivC, and INQoL demonstrate meaningful deterioration over 12–24 months, supporting their combined use with qMRI as outcomes in future trials for adult-onset DM1.

Introduction

Myotonic dystrophy type 1 (DM1) is an autosomal dominant muscle disorder, caused by an unstable cytosine-thymine-guanine (CTG) repeat expansion within the dystrophia myotonica protein kinase gene.^1,2 Larger expansions correlate with earlier disease onset and greater severity, supporting the classification into 4 main subtypes based on the age at onset: congenital, childhood, adult, and late-onset DM1.³ The adult-onset form is most prevalent and characterized by progressive distal muscle weakness and myotonia.^1,2,4

Currently, no disease-modifying therapies are available for patients with DM1, but promising clinical trials are emerging (eTable 1). To evaluate the efficacy of these new treatments, there is an urgent need for sensitive, reliable, noninvasive outcome measures that can detect small but meaningful changes relative to the natural disease progression.² Quantitative magnetic resonance imaging (qMRI) provides objective biomarkers, including proton density fat fraction (PDFF, %), which quantifies fat replacement,^5-7 and water T2 (T2_H2O), which measures disease activity.⁸ While the sensitivity of qMRI is established in some muscle disorders, its longitudinal application in DM1 is limited. Only 2 longitudinal qMRI studies have been conducted in DM1 cohorts.^9,10 Both studies included only a limited number of clinical outcome measures and were constrained by short follow-up durations (10 months), high dropout rates, heterogeneous populations, or restricted anatomical coverage. Moreover, analyses were limited to 2D rather than whole-muscle 3D muscle segmentations, which have been proven to be more sensitive.^6,7 Several studies have evaluated clinical outcome measures in adult-onset DM1.^11-15 Despite this growing evidence base, there is currently no consensus on a standardized outcome measure battery for adult-onset DM1, which is reflected by the heterogeneity across ongoing interventional trials (eTable 1). In this prospective, monocentric study, we evaluated a range of radiologic, clinical, and patient-reported outcome measures in patients with adult-onset DM1 over a 24-month period. Our aims were to identify robust and sensitive outcome measures of disease progression and to define the natural history of adult-onset DM1, thereby providing a reference framework for evaluating future therapeutic efficacy.

Methods

Patients and Study Design

Patients with genetically confirmed, adult-onset DM1 receiving care at the Neuromuscular Reference Centre, University Hospitals Leuven (Belgium), were invited to participate. Inclusion criteria were as follows: (1) age ≥18 years; (2) muscle weakness, defined as a Muscular Impairment Rating Scale (MIRS) score of 3 or 4; and (3) ambulatory status, with assistive devices or intermittent wheelchair use permitted. The main exclusion criteria were contraindications to MRI (e.g., pacemaker) or wheelchair dependency. Healthy age-matched (±3 years) and sex-matched adults were included to establish reference values.

Participants who met all eligibility criteria were assessed at baseline and 12, 18, and 24 months. The 18-month visit was included to examine whether significant changes could be detected over a shorter 6-month period (12–18 and 18–24 months). Radiologic, clinical, and patient-reported outcome measures (PROMs) were collected. All outcome measures used in this study have been validated in patients with DM1. The assessments were performed by the same examiner in a predefined order. Clinical measures involving the leg muscles were conducted after the MRI scan because physical activity can influence T2_H2O.¹⁶ Disease duration, defined as years between the patient-reported first symptoms and the baseline visit, was calculated for each patient. Clinical outcomes included MIRS,¹⁷ 32-item motor function measure (MFM32),¹⁸ 6-minute walk distance (6MWD),¹⁹ 10-meter walk test (10MWT),¹¹ 30-second sit-to-stand test (30SSS),¹¹ 9-hole peg test (9HPT),²⁰ and hand opening time (HOT).¹² HOT was defined as the time required to fully open the hand after a 5-second maximal voluntary contraction and was used to assess myotonia. HOT was assessed using a standard stopwatch method rather than video-based assessment (vHOT). Strength assessments comprised the 160-point Medical Research Council (MRC) sum score (shoulder abduction (seated); elbow flexion (seated) and extension (seated); wrist flexion and extension (seated); finger flexion, extension, and abduction (seated); hip abduction (side-lying), adduction (side-lying), and flexion (seated); knee flexion (prone) and extension (seated); ankle dorsiflexion (seated) and plantarflexion (prone); hallux extension (seated)), hand grip dynamometry (handheld Jamar dynamometer),²¹ and key and tip pinch strength.²¹ All MRC strength assessments were performed according to the standardized positions described by the Medical Research Council.²² Peak cough flow was obtained as a measure of expiratory respiratory muscle strength.²³ PROMs included the Rasch-built Myotonic Dystrophy Type 1 Activity and Participation Scale (DM1-ActivC),²⁴ the Brief Pain Inventory (BPI),²⁵ the Individualized Neuromuscular Quality of Life (INQoL) questionnaire,²⁶ and the Fatigue and Daytime Sleepiness Scale.²⁷ Finally, qMRI scans of the proximal and distal leg muscles were acquired at each visit.

MRI Acquisition Protocol

Imaging included T1-weighted, 6-point Dixon, and multislice multiecho (MSME) sequences obtained from the same 3-Tesla MRI scanner (Philips Achieva, Philips Medical Systems, The Netherlands). Participants were positioned supine, feet-first, with their legs stabilized using sandbags. Axial (TR 546 ms, TE 20 ms, 30 slices, 6-mm slice thickness, 1-mm interslice gap, FOV 455 × 301 mm², voxel size 0.8 × 0.8 mm²) and coronal (TR 562 ms, TE 20 ms, 33 slices, 6-mm slice thickness, 1-mm interslice gap, FOV 450 × 399 mm², voxel size 1 × 1 mm²) T1-weighted images, covering the entire legs, were acquired for anatomical reference. Muscle PDFF (%) was obtained using a 6-point 3D Dixon sequence (TR/TE₁/ΔTE = 10/1.5/1.0 ms, flip angle 3°, 130 slices, slice thickness 2 mm, FOV 450 × 394 × 260 mm³, matrix 348 × 304×130, voxel size 1.2 × 1.2 × 2 mm³), acquired in 5 stacks with a 25-slice overlap. Three proximal stacks covered the anterior superior iliac spine to the tibial plateau; 2 distal stacks covered the superior patellar border to the lateral malleolus. Muscle T2_H2O was quantified using an MSME sequence (TR = 3000 ms, echo train length 17, echo spacing 9.5 ms, voxel size 1.4 × 1.4 × 10 mm³, 11 slices, slice gap 25 mm, FOV 224 × 448 × 260 mm³), followed by a B1 map for radiofrequency inhomogeneity correction. To ensure reproducible positioning, the proximal stack was centered at one-third of the distance from the anterior superior iliac spine to the superior patellar border and the distal stack at one-third of the distance from the inferior patellar border to the lateral malleolus, measured on T1-weighted images. The total acquisition time was 60 minutes. Although a few patients reported some positional discomfort during the scan, no participant withdrew from the study because of MRI-related discomfort, and all completed the full scan protocol at each visit.

MRI Analysis

Dixon images were processed using an in-house convolutional neural network (CNN).²⁸ Fat, water, in-phase, and out-of-phase Dixon images were preprocessed to generate stitched fat-fraction (PDFF, %) maps for the proximal and distal legs. The CNN predicted 3D segmentations of 18 proximal and 10 distal leg muscles. All predicted segmentations were visually inspected and manually corrected as needed by the same assessor (L.I.) using ITK-snap. Subsequently, an in-house MATLAB script (MathWorks, Natick, MA) was applied to calculate the mean PDFF (%) for each individual muscle and the volume-weighted mean PDFF (%) for each muscle group (Figure 1). Muscle T2_H2O was computed using a triexponential model, which separates the muscle water and fat components in each voxel.²⁹ Study-specific fat parameters, derived from a manually delineated region in subcutaneous fat for each participant, were averaged and used as a fixed input for the T2_H2O analysis. Voxels were included if fits were reliable, fat was <65%, and B1 was 85%–130%.²⁹ Follow-up scans were co-registered to baseline to include only voxels present at all visits and meeting all reliability criteria. The 3D muscle segmentations were used to extract median T2_H2O for each muscle. Muscles were excluded if fewer than 10% of voxels met the reliability criteria. Finally, median T2_H2O was obtained automatically per muscle and muscle group using a MATLAB script (MathWorks). Owing to the reliability criteria and the MSME field of view, T2_H2O could not be obtained for the gluteus maximus, medius, and minimus; tensor fascia latae; pectineus; and flexor hallucis longus.

Figure 1. Baseline Muscle Fat Fraction in Patients With Adult-Onset DM1 Relative to Disease Duration.

Heatmap showing a general increase of PDFF (%) in the muscles by disease duration. Each row represents a single patient, ordered by ascending disease duration (years from symptom onset to baseline) as indicated in the left margin. Columns correspond to individual muscles and are categorized into distal (anterolateral and posterior) and proximal (quadriceps, adductors, and hamstrings) muscle groups. The color scale (right) indicates the severity of fat replacement, ranging from 0% (light green) to 100% (dark green). The bottom row provides the mean PDFF (%) for each muscle across the entire cohort at baseline. The dashed vertical line divides distal and proximal leg muscles. AB = adductor brevis; AL = adductor longus; AM = adductor magnus; BFB = biceps femoris brevis; BFL = biceps femoris longus; DM1 = myotonic dystrophy type 1; EDL = extensor digitorum longus; FDL = flexor digitorum longus; FHL = flexor hallucis longus; GL = gastrocnemius lateralis; GM = gastrocnemius medialis; Gma = gluteus maximus; Gme = gluteus medius; Gmi = gluteus minimus; GRA = gracilis; PDFF = proton density fat fraction; PEC = pectineus; PER = peroneus; POP = popliteus; RF = rectus femoris; SAR = sartorius; SM = semimembranosus; SOL = soleus; ST = semitendinosus; TA = tibialis anterior; TFL = tensor fascia latae; TP = tibialis posterior; VI = vastus intermedius; VL = vastus lateralis; VM = vastus medialis.

Statistical Analysis

Statistical analyses were performed in R (version 4.3.2) using RStudio. Longitudinal changes were assessed using linear mixed-effects models. The models included group (DM1 vs control), time (0, 12, 18, 24 months), and their interaction as fixed effects, with age and sex as covariates. Participant-specific random intercepts were included, with random slopes for time when supported by model fit. Estimated marginal means were obtained for each group-time combination. Standardized response means (SRMs) were calculated as mean change from baseline to 12, 18, and 24 months divided by the standard deviation of the change scores. Associations between PDFF (%) and clinical measures at baseline were assessed using Spearman or Pearson correlations. A significance level of α = 0.05 was applied. Given the large number of simultaneous tests performed, FDR correction was applied to reduce the risk of false-positive findings and enhance the robustness of the results.³⁰

Standard Protocol Approvals, Registrations, and Patient Consents

This study was conducted according with the Declaration of Helsinki and approved by the Ethics Committee Research UZ/KU Leuven (s67475). All study participants provided written informed consent.

Data Availability

The anonymized data sets are available from the corresponding author on reasonable request.

Results

Participants

Thirty-three patients with adult-onset DM1 and 33 age-matched and sex-matched healthy controls were enrolled. Three patients were excluded after baseline because of intervening medical events (pacemaker implantation, lower-limb fracture, stroke). The matched controls were excluded accordingly. One patient missed the 24-month follow-up visit for personal reasons but was retained in the final analysis. Demographic data are presented in Table 1.

Table 1.

Demographics of Patients With Adult-Onset DM1 and Healthy Controls

	Patients with DM1 (n = 30)	Controls (n = 30)
Female	60% (18)	60% (18)
Current BMI (kg/m2)	26.5 (16.5–34.2)	22.3 (18.4–29.4)
Current age (y)	38 (24–57)	38 (24–58)
CTG repeat expansion size (n)	430 (100–1,000)
Age at symptom onset (y)	25 (20–40)
Disease duration (y)	10 (1–28)
Ambulation status
Independent	80% (24)	100% (30)
Orthoses	10% (3)
Bilateral support	7% (2)
Intermittent wheelchair use	3% (1)
MIRS (score)
Distal weakness (3)	67% (20)
Mild proximal weakness (4)	33% (10)

Abbreviations: % = percentage; BMI = body mass index; CTG = cytosine-thymine-guanine; DM1 = myotonic dystrophy type 1; kg = kilograms; m = meter; MIRS = Muscular Impairment Rating Scale; n = number.

Data are shown as % (n) or as median [range]. No patients in our study were classified as MIRS 1, 2, or 5.

Baseline Characteristics of Muscle Involvement on MRI in Adult-Onset DM1

PDFF (%) increased with disease duration, showing a symmetrical pattern of fat replacement predominantly affecting the distal leg muscles (Figure 1).

Overall, 10% (n = 31) of the distal leg muscles and 1% (n = 5) of the proximal leg muscles exhibited a PDFF (%) greater than 65%. The posterior distal compartment (gastrocnemius medialis, soleus, and flexor hallucis longus) demonstrated most severe and early involvement, whereas the popliteus and tibialis posterior muscles displayed sparing over the disease course. Marked increases over time occurred in the anterolateral distal compartment (tibialis anterior, peroneus, and extensor digitorum longus). By contrast, proximal leg muscles showed lower baseline PDFF (%), with progressive increases mainly in the gluteus minimus, vastus intermedius > medialis > lateralis, and tensor fascia latae. Conversely, the rectus femoris, pectineus, adductor brevis and magnus, and gluteus medius remained largely spared (Figure 1). At baseline, patients exhibited significantly higher mean PDFF (%) across all distal and proximal muscles compared with controls, with the largest differences observed in distal muscles (distal leg: 31.8% vs 5.9%, p < 0.001; proximal leg: 13.0% vs 7.6%, p < 0.001; Figure 2A).

Figure 2. Analysis of Muscle PDFF (%) in Patients With DM1 and Controls at Baseline and Over 2 Years.

(A) Box plots representing the PDFF (%) at baseline for the distal and proximal leg muscle groups. Error bars show the interquartile range of the data, with individual points representing outliers. Significant differences between the patients and controls are indicated by asterisks (*). (B) Change in PDFF (%) over 24 months. Line graphs delineate the longitudinal progression of fat fraction for different muscle groups. Vertical bars show 95% CI. Numeric labels indicate the estimated marginal mean PDFF (%) increase at 12, 18, and 24-month intervals for patients with adult-onset DM1. DM1 = myotonic dystrophy type 1; PDFF = proton density fat fraction.

T2_H2O was significantly elevated in all distal leg muscles of patients compared with controls, except the flexor digitorum longus (Figure 3A). No significant differences were observed in most proximal leg muscles. Exceptions included the adductor longus and adductor magnus, which exhibited significantly lower T2_H2O in patients (−1.0 ms, p < 0.05 and −0.7 ms, p < 0.05, respectively), and the biceps femoris brevis, which showed a significantly higher T2_H2O (+1.5 ms, p < 0.01; Figure 3A).

Figure 3. Analysis of Muscle T2_H2O in Patients With DM1 and Controls at Baseline and Over 2 Years.

(A) Muscle T2_H2O relative to healthy controls at baseline. The Y-axis represents the individual patient difference from the estimated marginal mean of the healthy control group. Individual raw patient data points are shown as jittered blue circles to visualize cohort distribution. Central white circles and vertical black bars represent the group estimated marginal mean and 95% CI, respectively. Statistical significance is indicated by asterisks (*). The dashed vertical line divides distal and proximal leg muscles. (B) Change in T2_H2O over 24 months in distal and proximal leg muscles. Line graphs delineate the longitudinal progression of T2_H2O across specific muscle groups. Vertical bars show 95% CI. Numeric labels indicate the estimated marginal mean T2_H2O. Patients are represented in dark green and healthy controls in light green. AB = adductor brevis; AL = adductor longus; AM = adductor magnus; BFB = biceps femoris brevis; BFL = biceps femoris longus; DM1 = myotonic dystrophy type 1; EDL = extensor digitorum longus; FDL = flexor digitorum longus; FHL = flexor hallucis longus; GL = gastrocnemius lateralis; GM = gastrocnemius medialis; Gma = gluteus maximus; Gme = gluteus medius; Gmi = gluteus minimus; GRA = gracilis; PDFF = proton density fat fraction; PEC = pectineus; PER = peroneus; POP = popliteus; RF = rectus femoris; SAR = sartorius; SM = semimembranosus; SOL = soleus; ST = semitendinosus; T2_H2O = water T2; TA = tibialis anterior; TFL = tensor fascia latae; TP = tibialis posterior; VI = vastus intermedius; VL = vastus lateralis; VM = vastus medialis.

Longitudinal Changes in PDFF (%) and T2_H2O in the Distal and Proximal Leg Muscles in Adult-Onset DM1

PDFF (%) significantly increased after 12 months in 9 of 10 distal and 13 of 18 proximal leg muscles (Table 2). Significant increases were observed in the total distal leg (+1.7, p < 0.001) and total proximal leg (+0.8, p < 0.001) and in all muscle groups (Table 2, Figure 2B, eFigure 1). In the distal leg, 9 of 10 muscles continued to show significant increases at 18 and 24 months (+2.7% and +3.5%; both p < 0.001). A significant increase in PDFF (%) was found between 12 and 18 months in total distal, anterolateral, and posterior compartments (+1.0%, p < 0.01; +0.7%, p < 0.05; +1.2%, p < 0.001, eTable 2). From 18 to 24 months, increases persisted in the total distal and anterolateral compartments (+0.8%, p < 0.05 and +1.2%, p < 0.001; eTable 2). The largest 24-month effect size in distal muscles was observed in the flexor hallucis longus muscle (SRM = 1.7), followed by the total distal leg (SRM = 1.6; Table 2).

Table 2.

Change in Proton Density Fat Fraction (PDFF, %) of Proximal and Distal Leg Muscles Over 24 Months in Patients With DM1

	Baseline	12 mo–0 mo		18 mo–0 mo		24 mo–0 mo
	EMM (95% CI)	Δ (95% CI)	SRM	Δ (95% CI)	SRM	Δ (95% CI)	SRM
Tibialis anterior	25.7 (20.9 to 30.5)	1.7 (1.1 to 2.4)a	0.9a	2.3 (1.5 to 3.1)a	1.0a	3.3 (2.4 to 4.2)a	1.2a
Peroneus	23.6 (19.5 to 27.7)	1.2 (0.4 to 2.0)a	0.6	2.0 (1.0 to 3.0)a	0.7	3.3 (2.0 to 4.6)a	0.9a
Extensor digitorum longus	23.8 (19.8 to 27.9)	1.7 (0.9 to 2.5)a	0.7	2.6 (1.8 to 3.4)a	0.9a	3.7 (2.8 to 4.4)a	1.2a
Flexor digitorum longus	15.8 (12.1 to 19.6)	1.3 (0.7 to 2.0)a	0.9a	1.6 (0.8 to 2.4)a	0.7	2.0 (1.0 to 2.9)a	0.8a
Flexor hallucis longus	36.5 (30.6 to 42.4)	2.8 (1.6 to 4.0)a	0.8a	5.4 (4.2 to 6.6)a	1.5a	5.9 (4.7 to 7.1)a	1.7a
Soleus	43.6 (36.4 to 50.8)	1.7 (0.8 to 2.7)a	0.6	2.9 (1.6 to 4.1)a	0.8	3.5 (1.9 to 5.1)a	0.8a
Gastrocnemius medialis	50.2 (43.9 to 56.4)	1.1 (0.3 to 1.9)a	0.4	2.3 (1.2 to 3.4)a	0.8a	2.7 (1.4 to 4.1)a	0.7
Gastrocnemius lateralis	27.0 (20.8 to 33.2)	1.8 (0.9 to 2.6)a	0.7	3.1 (2.0 to 4.3)	0.9a	4.0 (2.6 to 5.4)a	1.1a
Tibialis posterior	11.5 (10.2 to 12.7)	1.0 (0.3 to 1.7)a	0.7	1.6 (0.7 to 2.6)a	0.6	2.1 (0.9 to 3.4)a	0.6
Popliteus	9.0 (8.3 to 9.8)	0.3 (−0.2 to 0.7)	0.2	0.4 (−0.1 to 0.9)	0.4	0.5 (0.0 to 1.0)	0.5
Distal leg	31.8 (26.8 to 36.7)	1.7 (1.1 to 2.2)a	1.0a	2.7 (2.1 to 3.3)a	1.2a	3.5 (2.9 to 4.1)a	1.6a
Anterolateral compartment	24.1 (20.3 to 27.9)	1.6 (1.0 to 2.2)a	0.9a	2.3 (1.5 to 3.1)a	1.0a	3.5 (2.5 to 4.5)a	1.2a
Posterior compartment	35.4 (29.9 to 40.9)	1.8 (1.1 to 2.4)a	0.9a	2.9 (2.1 to 3.7)a	1.2a	3.6 (2.6 to 4.5)a	1.4a
Rectus femoris	9.0 (8.1 to 9.9)	0.6 (0.1 to 1.1)a	0.5	0.9 (0.4 to 1.5)a	0.7	1.4 (0.8 to 2.1)a	0.8a
Vastus lateralis	14.0 (10.7 to 17.3)	0.8 (−0.3 to 2.0)	0.4	1.8 (0.6 to 3.0)a	0.5	2.6 (1.4 to 3.8)a	0.7
Vastus medialis	16.9 (12.1 to 21.8)	1.8 (0.9 to 2.7)a	0.7	2.3 (1.3 to 3.2)a	0.7	3.3 (2.4 to 4.2)a	0.9a
Vastus intermedius	19.9 (15.0 to 24.7)	1.5 (0.6 to 2.3)a	0.6	2.2 (1.0 to 3.4)a	0.7	3.1 (1.5 to 4.7a)	0.7
Sartorius	13.7 (12.0 to 15.4)	0.9 (0.4 to 1.4)	0.7	1.2 (0.6 to 1.8)a	0.8a	1.8 (1.0 to 2.6)a	0.8a
Adductor brevis	6.3 (5.8 to 6.8)	0.1 (−0.3 to 0.5)	0.2	0.3 (−0.1 to 0.7)	0.4	0.6 (0.2 to 1.0)a	0.7
Adductor longus	12.1 (8.6 to 15.7)	1.0 (0.3 to 1.7)a	0.6	1.6 (0.8 to 2.5)a	0.6	1.7 (0.8 to 3.0)a	0.7
Adductor magnus	8.5 (6.9 to 10.2)	0.6 (0.1 to 1.1)a	0.4	0.9 (0.4 to 1.4)a	0.5	0.9 (0.3 to 1.4)a	0.5
Gracilis	12.9 (10.3 to 15.4)	0.9 (−0.3 to 2.1)	0.5	2.6 (1.3 to 3.8)a	0.8a	3.3 (2.1 to 4.6)a	0.9a
Pectineus	7.1 (6.6 to 7.5)	0.3 (−0.2 to 0.8)	0.3	0.2 (−0.3 to 0.6)	0.2	0.4 (0.0 to 0.9)	0.5
Biceps femoris longus	11.1 (9.5 to 12.6)	0.6 (0.2 to 1.1)a	0.5	1.1 (0.6 to 1.5)a	1.0a	0.9 (0.4 to 1.3)a	0.7
Biceps femoris brevis	14.9 (12.6 to 17.2)	1.6 (0.8 to 2.4)a	0.6	2.3 (1.2 to 3.3)a	0.8a	2.7 (1.4 to 4.0)a	0.8a
Semitendinosus	11.9 (9.9 to 13.9)	0.8 (0.3 to 1.3)	0.7	1.3 (0.7 to 2.0)a	0.7	1.4 (0.6 to 2.3)a	0.7
Semimembranosus	15.8 (12.1 to 19.5)	0.9 (0.3 to 1.5)a	0.6	1.7 (1.0 to 2.3)a	1.0a	2.1 (1.3 to 2.9)a	1.0a
Tensor fascia latae	18.5 (16.3 to 20.7)	1.1 (0.2 to 1.9)a	0.4	1.4 (0.4 to 2.3)a	0.6	1.5 (0.5 to 2.6)a	0.7
Gluteus maximus	15.3 (13.5 to 17.2)	0.7 (0.1 to 1.2)a	0.4	1.3 (0.6 to 1.9)a	0.8a	0.9 (0.0 to 1.6)	0.5
Gluteus medius	10.1 (9.4 to 10.9)	0.4 (−0.2 to 1.0)	0.1	0.2 (−0.4 to 0.8)	0.0	0.0 (−0.7 to 0.6)	0.0
Gluteus minimus	24.7 (19.3 to 30.1)	1.3 (0.6 to 2.1)a	0.7	1.8 (0.9 to 2.8)a	0.8a	1.9 (0.7 to 2.9)a	0.8a
Proximal leg	13.0 (11.2 to 14.7)	0.8 (0.4 to 1.3)a	0.7	1.4 (0.8 to 1.9)a	0.8a	1.6 (0.9 to 2.3)a	0.9a
Quadriceps	14.9 (11.9 to 17.9)	1.2 (0.4 to 2.0)a	0.5	1.9 (0.8 to 3.0)a	0.6	2.8 (1.5 to 4.3)a	0.8a
Adductors	8.6 (7.3 to 9.8)	0.6 (0.1 to 1.0)a	0.5	0.9 (0.3 to 1.5)a	0.5	1.0 (0.3 to 1.8)a	0.6
Hamstrings	13.2 (10.9 to 15.5)	0.8 (0.4 to 1.3)a	0.7	1.4 (0.8 to 1.9)a	1.0a	1.6 (0.9 to 2.3)a	0.9a

Abbreviations: Δ = delta (absolute change); DM1 = myotonic dystrophy type 1; EMM = estimated marginal mean; SRM = standardized response mean.

Data are presented as estimated marginal mean (baseline or change from baseline) with 95% CI.

^aSignificant p values and SRM values with a large effect size (>0.8).

In the proximal leg, 15 of 18 individual muscles showed further significant increases at both 18 and 24 months (+1.4%; +1.5%; both p < 0.001; Table 2, Figure 2B, eFigure 1). Significant increases were also detected over the 6-month period from 12 to 18 months, in the total proximal leg muscles, quadriceps, and hamstrings (+0.5%, p < 0.01; +0.7%, p < 0.05; +0.6%, p < 0.05; eTable 2). From 18 to 24 months, a significant increase was observed only in the quadriceps (+0.9%, p < 0.001; eTable 3). Among the proximal muscles, the largest 24-month effect size was observed in the semimembranosus muscle (SRM = 1.0). In the control group, no significant changes over time were measured in any individual muscles or muscle groups (Figure 2B).

T2_H2O remained relatively stable over time (Figure 3B). Four distal leg muscles increased significantly at 18 months. However, this was not sustained at 24 months in 3 of these muscles (eTable 4). Accordingly, T2_H2O values for the total distal leg and anterolateral compartments were significantly increased at 18 months (+1.6 ms, p < 0.001 and +2.0 ms, p < 0.001, respectively). No significant longitudinal changes were observed in proximal leg muscles or in controls (Figure 3B; eTable 4).

Longitudinal Changes in Clinical and Patient-Reported Outcome Measures in Adult-Onset DM1

At baseline, patients scored significantly worse on all clinical and patient-reported outcome measures compared with controls. MFM32 total scores deteriorated significantly at 12, 18, and 24 months (−3.2, p < 0.001; −6.7, p < 0.001; −7.3, p < 0.001, respectively), with significant declines in D1 (standing and transfers) (−6.4, p < 0.001; −11.8, p < 0.001; −12.7, p < 0.001, respectively) and D2 (axial and proximal motor function) (−1.6, p < 0.01; −3.1, p < 0.001; −3.6, p < 0.001, respectively) subscores (Table 3). Moreover, a significant deterioration was observed between 12 and 18 months for these measures (total: −4.1, p < 0.001; D1: −6.3, p < 0.001; D2: −2.0, p < 0.05; eTable 5). The D3 (distal motor function) subscore declined significantly after 18 and 24 months (−3.1, p < 0.001; −3.8, p < 0.001, respectively) (Table 3) and showed a significant reduction between 12 and 18 months (−2.9, p < 0.001; eTable 5). A large effect size was already present after 12 months for the total score and D1 subscore (SRM = 0.8; Table 3). Muscle strength measured with the MRC sum score demonstrated significant deterioration at 12, 18, and 24 months (−5.0, p < 0.001; −7.6, p < 0.001; −9.6, p < 0.001, respectively), and between 12 and 18 months (−2.6, p < 0.01; Table 3; eTable 5). A large effect size was found after 12 months (SRM = 1.1; Table 3; eTable 5). Walking capacity declined over time, with a significant worsening in 10MWT after 18 months (+0.6 seconds, p < 0.05), but not after 24 months. Healthy controls showed no significant changes in any of the clinical outcome measures except for 30SSS, which significantly improved after 12, 18, and 24 months (baseline = 23; +3.0, p < 0.01; +4.3, p < 0.001; +5.1, p < 0.001, respectively), most likely due to a learning effect, although a genuine improvement in lower limb function cannot be excluded.

Table 3.

Change in Clinical and Patient-Reported Outcome Measures Over 24 Months in Patients With DM1

	Baseline	12 mo–0 mo		18 mo–0 mo		24 mo–0 mo
	EMM (95% CI)	Δ (95% CI)	SRM	Δ (95% CI	SRM	Δ (95% CI	SRM
6MWD (m)	426 (388 to 465)	−0.4 (−36.0 to 35.1)	0.0	−7.9 (−43.8 to 28.0)	0.1	−9.9 (−45.8 to 26.1)	0.4
10MWT (s)	7.7 (6.9 to 8.4)	0.4 (−0.1 to 1.0)	0.3	0.6 (0.1 to 1.2)a	0.3	0.5 (0.0 to 1.0)	0.3
MFM32 (total) (%)	84.3 (81.7 to 87.0)	−3.2 (−4.3 to −2.1)a	0.8a	−6.7 (−7.9 to −5.4)a	1.9a	−7.3 (−8.7 to −6.0)a	1.8a
MFM32 (D1) (%)	71.0 (65.8 to 76.1)	−6.4 (−8.7 to −4.1)a	0.8a	−11.8 (−14.2 to −9.5)a	1.8a	−12.7 (−15.0 to −10.3a	1.7a
MFM32 (D2) (%)	92.0 (90.4 to 93.5)	−1.6 (−2.7 to −0.4)a	0.4	−3.1 (−4.3 to −1.9)a	0.7	−3.6 (−4.8 to −2.4)a	0.7
MFM32 (D3) (%)	95.8 (94.1 to 97.5)	−0.2 (−1.7 to 1.3)	0.0	−3.1 (−4.7 to −1.5)a	0.6	−3.8 (−5.5 to −2.0)a	0.8a
Hand opening time (s)	3.2 (2.6 to 3.8)	0.4 (−0.2 to 1.1)	0.4	−0.3 (−1.0 to 0.3)	0.2	0.1 (−0.6 to 0.8)	0.0
9HPT (s)	22.9 (21.9 to 24.0)	1.0 (−0.2 to 2.3)	0.3	1.3 (0.0 to 2.7)	0.4	0.6 (−0.9 to 2.2)	0.2
30SSS (n)	13.2 (10.6 to 15.7)	−0.9 (−2.6 to 0.8)	0.5	−1.0 (−2.9 to 0.9)	0.4	−1.4 (−3.5 to 0.8)	0.4
PCF (L/min)	404 (373 to 435)	−6.8 (−31.3 to 17.7)	0.2	−4.4 (−29.3 to 20.5)	0.1	11.8 (−14.1 to 37.7)	0.3
Hand grip (ND) (kg)	12.0 (9.6 to 14.3)	−1.3 (−2.9 to 0.2)	0.3	0.2 (−1.8 to 1.4)	0.1	0.0 (−1.5 to 1.6)	0.2
Hand grip (D) (kg)	12.1 (9.6 to 14.5)	−1.4 (−3.0 to 0.2)	0.4	−0.4 (−2.1 to 1.2)	0.2	−0.5 (−2.2 to 1.1)	0.5
Key pinch (ND) (kg)	4.0 (3.4 to 4.6)	−0.5 (−0.9 to 0.0)	0.7	−0.4 (−0.9 to 0.1)	0.8a	−0.6 (−1.1 to 0.0)	0.7
Key pinch (D) (kg)	4.2 (3.6 to 4.8)	−0.4 (−0.9 to 0.1)	0.6	−0.6 (−1.0 to −0.1)	0.9a	−0.5 (−1.1 to 0.0)	0.5
Tip pinch (ND) (kg)	2.0 (1.7 to 2.4)	−0.1 (−0.5 to 0.2)	0.2	0.0 (−0.4 to 0.4)	0.0	−0.1 (−0.5 to 0.3)	0.1
Tip pinch (D) (kg)	2.2 (1.8 to 2.5)	−0.2 (−0.5 to 0.2)	0.3	−0.1 (−0.5 to 0.2)	0.2	−0.2 (−0.6 to 0.1)	0.3
MRC sum score (160)	145 (143 to 148)	−5.0 (−6.6 to −3.4)a	1.1a	−7.6 (−9.4 to −5.8)a	1.4a	−9.6 (−11.6 to −7.5)a	1.6a
DM1-ActivC	68.2 (63.9 to 72.5)	−5.6 (−9.2 to −2.0)a	0.5	−6.2 (−9.8 to −2.5)a	0.5	−10.3 (−14.0 to −6.7)a	0.8a
FDSS	42.8 (39.0 to 46.6)	−0.9 (−3.7 to 1.9)	0.1	1.5 (−1.4 to 4.3)	0.2	0.0 (−2.8 to 2.9)	0.0
INQoL
Symptom score (%)	38.5 (33.6 to 43.4)	1.2 (−2.9 to 5.4)	0.1	2.1 (−2.5 to 6.7)	0.1	3.6 (−1.5 to 8.7)	0.2
Weakness (%)	51.9 (46.2 to 57.7)	−0.6 (−6.1 to 5.0)	0.0	0.7 (−4.9 to 6.3)	0.1	−0.1 (−5.7 to 5.5)	0.0
Pain (%)	20.8 (14.0 to 27.6)	5.9 (−1.4 to 13.1)	0.3	11.7 (3.6 to 19.7)a	0.4	10.9 (1.9 to 20.0)a	0.4
Fatigue (%)	41.8 (35.1 to 46.5)	−2.3 (−8.2 to 3.6)	0.1	0.2 (−6.4 to 6.7)	0.1	2.9 (−4.4 to 10.2)	0.1
Myotonia (%)	39.5 (33.5 to 45.5)	2.0 (−4.9 to 8.8)	0.1	−4.1 (−11.1 to 2.8)	0.2	1.0 (−5.9 to 7.9)	0.0
Eyelids (%)	14.3 (8.6 to 20.1)	3.2 (−3.2 to 9.6)	0.2	−0.8 (−7.3 to 5.7)	0.1	−1.4 (−7.9 to 5.1)	0.1
Vision (%)	4.3 (1.0 to 7.7)	−1.2 (−5.9 to 3.5)	0.1	−2.5 (−7.2 to 2.3)	0.2	−3.0 (−7.8 to 1.7)	0.3
Swallow (%)	26.5 (19.0 to 34.0)	0.1 (−5.0 to 5.1)	0.0	1.7 (−3.4 to 6.8)	0.1	1.5 (−3.6 to 6.6)	0.1
Quality of life (%)	34.9 (29.5 to 40.3)	5.4 (0.8 to 10.0)a	0.4	6.9 (2.0 to 11.8)a	0.5	5.3 (0 to 10.7)	0.3
Activities (%)	33.1 (27.2 to 39.0)	8.5 (3.6 to 13.5)a	0.5	11.1 (6.1 to 16.1)a	0.7	11.6 (6.6 to 16.6)a	0.8a
Independence (%)	32.9 (29.8 to 39.0)	−4.1 (−9.4 to 1.1)	0.3	1.1 (−4.8 to 6.9)	0.0	3.1 (−3.5 to 9.8)	0.2
Relations (%)	18.9 (14.1 to 23.7)	4.8 (1.0 to 8.7)a	0.4	4.0 (0.2 to 7.9)	0.3	6.8 (2.9 to 10.7)a	0.6
Emotions (%)	28.7 (23.3 to 34.2)	−0.1 (−5.1 to 4.9)	0.0	2.2 (−3.3 to 7.6)	0.2	−0.7 (−6.7 to 5.3)	0.0
Body image (%)	31.0 (24.2 to 37.7)	5.6 (−0.8 to 11.9)	0.3	1.6 (−4.8 to 8.2)	0.1	0.0 (−6.7 to 6.8)	0.0
BPI
Severity index	2.7 (2.1 to 3.3)	0.2 (−0.4 to 0.8)	0.1	0.1 (−0.5 to 0.8)	0.0	0.2 (−0.4 to 0.8)	0.0
Interference index	1.9 (1.3 to 2.5)	0.4 (−0.4 to 1.1)	0.2	0.7 (0.0 to 1.5)	0.3	0.3 (−0.5 to 1.1)	0.2

Abbreviations: Δ = delta (absolute change); % = percentage; 6MWD = 6-minute walk distance; 9HPT = 9-hole peg test; 10MWT = 10-meter walk test; 30SSS = 30-second sit-to-stand; BPI = Brief Pain Inventory; D = dominant; DM1 = myotonic dystrophy type 1; EMM = estimated marginal mean; FDSS = fatigue and daytime sleepiness scale; kg = kilograms; INQoL = Individualized Neuromuscular Quality of Life; l = liter; m = meter; MFM32 = 32-item motor function measure (D1: standing and transfers, D2: axial and proximal motor function, D3: distal motor function); min = minute; MRC sum score = Medical Research Council sum score; n = number; ND = nondominant; PCF = peak cough flow; s = seconds; SRM = standardized response mean.

Data are presented as estimated marginal mean (baseline or change from baseline) with 95% confidence intervals (CI).

^aSignificant p values and SRM values with a large effect size (>0.8).

PROMs indicated significant worsening of DM1-ActivC and the INQoL-Activities subscore at 12, 18, and 24 months (DM1-ActivC: −5.6, p < 0.01; −6.2, p < 0.001; –10.3, p < 0.001; INQoL-Activities: 8.5, p < 0.001; 11.1, p < 0.001; 11.6, p < 0.001, respectively). In addition, the overall quality of life measured by the INQoL showed a significant deterioration after 12 and 18 months (5.4, p < 0.05; 6.9, p < 0.01) (eTable 6).

Correlations Between PDFF (%) and Clinical Outcome Measures in Adult-Onset DM1

Distal and total leg PDFF (%) showed strong significant correlations with 10MWT (ρ = 0.77, ρ = 0.78), 6MWD (ρ = −0.81, ρ = −0.81), MFM32 total score (ρ = −0.72, ρ = −0.75), and D1 subscore (ρ = −0.75, ρ = −0.78). Moderate, significant correlations were found for age (ρ = 0.46, ρ = 0.40), disease duration (ρ = 0.47, ρ = 0.52), MFM32 D2 subscore (ρ = −0.55, ρ = −0.52), 30SSS (ρ = −0.66, ρ = −0.68), and MRC sum score of the distal leg (ρ = −0.43, ρ = −0.45) (Figure 4, eFigures 2 and 3). For proximal leg PDFF (%), a similar, slightly weaker pattern of significant correlations was found (6MWD ρ = −0.75; 10MWT ρ = 0.67), except for the disease duration, MFM32 D2 subscore, and MRC sum score of distal leg muscles (eFigure 4). By contrast, both total leg and proximal leg PDFF (%) showed moderate, significant correlations with the total leg MRC sum score. PDFF (%) of the total, proximal, and distal legs did not correlate with CTG repeat length, HOT, and 9HPT (Figure 4). Correlations between clinical outcomes, MIRS scores, and CTG repeat size are shown in eFigure 5. MIRS score showed a strong correlation with the MFM32 total score (ρ = −0.70) and MFM32 D1 subscore (ρ = −0.73). Moderate, significant correlations were observed with MRC sum score (ρ = −0.65), 6MWD (ρ = −0.57), 10MWT (ρ = 0.51), and 30SSS (ρ = −0.53), whereas CTG repeat length showed no significant correlations with any clinical outcome measure. In addition, we evaluated the association between HOT and hand grip strength, but it did not reach statistical significance.

Figure 4. Correlation Heatmap of Muscle PDFF (%) and Clinical Outcome Measures at Baseline.

Spearman/Pearson correlation coefficients between PDFF (%) and clinical outcomes. The heatmap displays the strength and direction of the relationship between PDFF (%) measured in the distal leg, proximal leg, and total leg against demographic and clinical outcomes. The color gradient represents the correlation coefficient, ranging from −1.0 (strong negative correlation, dark green) to +1.0 (strong positive correlation, dark blue). The asterisks (*) denote statistically significant correlations (p < 0.05). 6MWD = 6-minute walk distance; 9HPT = 9-hole peg test; 10MWT = 10-meter walk test; 30SSS = 30-second sit-to-stand; CTG = cytosine-thymine-guanine; MFM32 = 32-item motor function measure (D1: standing and transfers, D2: axial and proximal motor function, D3: distal motor function); MRC sum score = Medical Research Council sum score; PDFF = proton density fat fraction; Prox. = proximal.

Discussion

In this 2-year longitudinal study, muscle PDFF (%) proved highly sensitive and objective as an outcome measure in patients with adult-onset DM1. By contrast, T2_H2O was significantly elevated only in distal leg muscles and revealed no significant change over the follow-up period. Furthermore, the clinical and patient-reported outcome measures MFM32, MRC sum score, 10MWT, DM1-ActivC, and INQoL showed significant deterioration over time.

Quantitative PDFF (%) analysis of 18 proximal and 10 distal leg muscles confirmed the characteristic pattern of predominant distal muscle involvement, described in previous semiquantitative MRI studies.^31-33 Despite the clinical importance, longitudinal quantitative muscle MRI studies in DM1 are limited to only 2 previous reports.^9,10 While these studies assessed progression over only 10-month or 12-month intervals, our study evaluated progression at 12, 18, and 24 months, enabling us to quantify changes over a longer period and as early as 6-month intervals. This is particularly relevant for clinical trials, where outcome measures must be sensitive enough to detect meaningful change within short time frames.² The 1.3% annual PDFF (%) increase in the total leg muscles is consistent with previous studies, which reported progression rates of 1.2% and 1.3%, respectively.^9,34 Comparison with one of these longitudinal studies.¹⁰ is challenging because the results are reported as standardized Z-scores and the study was limited by high dropout rates, with only 6 of 36 patients completing the 24-month visit. In contrast to the other longitudinal study, which reported the largest increase in the quadriceps (+1.7%) after 10 months,⁹ the greatest PDFF (%) increase in our study was observed in the posterior compartment of the distal leg muscles (+1.8%) after 12 months. This discrepancy may reflect our use of whole-muscle segmentation, which provides a more comprehensive assessment of PDFF (%) than the partial-slice methods, which can overlook heterogeneous fat distribution.⁹ Our study demonstrated that distal leg muscles had higher SRMs than proximal leg muscles at all time points. This suggests that while proximal leg muscles may be valuable for longer term assessments, distal leg muscles are particularly sensitive to short-term disease progression. Although a DM1-specific minimal clinically important difference (MCID) for muscle PDFF (%) has not been established, the annual increases in our cohort fall within the 1%–5% MCID range reported in Duchenne muscular dystrophy.³⁵ The lower progression rates are consistent with the slower course of adult-onset DM1. Defining a DM1-specific MCID will require larger, multicenter longitudinal studies. The clinical relevance of our qMRI findings is further supported by the significant correlations between PDFF (%) and most clinical outcomes. These findings confirm those of previous cross-sectional studies, which concluded that fat fraction is a strong predictor of muscle strength and functional capacity in DM1.^36,37 PDFF (%) correlated more strongly with the 10MWT and 6MWD than with the MRC sum score, a pattern similarly reported in Becker muscular dystrophy and Duchenne muscular dystrophy.^7,38 Walking-based measures capture the functional impact of fat replacement across multiple muscle groups, integrating strength, endurance, balance, and coordination, whereas the MRC sum score reflects muscle strength. Both should be considered complementary outcome measures in adult-onset DM1. While distal leg PDFF (%) demonstrated high SRMs across all time points (SRM = 1.0 at 12 months; SRM = 1.6 at 24 months), the MFM32 score (SRM = 1.9) and MRC sum score (SRM = 1.6) showed comparable responsiveness over longer intervals. PDFF (%) is particularly advantageous for shorter trials, given its sensitivity within 6 months and examiner-independent nature. A combination of both approaches provides the most comprehensive assessment of therapeutic efficacy.

In contrast to PDFF (%), T2_H2O was significantly elevated only in the distal leg muscles at baseline compared with controls, despite the proximal leg muscles also demonstrating significantly higher PDFF (%). This pattern likely reflects the different stages of disease involvement between the 2 regions: the distal leg muscles, with the most advanced fat replacement, are likely at a more active stage of ongoing fiber degeneration where edema and inflammation contribute to elevated T2_H2O, whereas the proximal muscles may not yet have reached this threshold.^29,39,40 Conversely, higher values in both proximal and distal leg muscles have previously been reported.⁴¹ Several factors may account for these differences. Methodologically, the previous study used a bicomponent extended-phase graph model and included a broader range of DM1 subtypes, encompassing patients with MIRS scores from 1 to 5, which may have influenced T2_H2O characteristics.⁴¹ Furthermore, the lower T2_H2O value observed in the adductor muscles in our cohort, although statistically significant, was numerically small (−0.7 ms, p < 0.05), suggesting limited clinical relevance. Our longitudinal analysis revealed that T2_H2O remained relatively stable over a 24-month period. The absence of a linear increase suggests that T2_H2O might not serve as a measure for disease progression in the same way as PDFF (%) but rather reflects disease activity–related changes. Whether T2H₂O may serve as a pharmacodynamic biomarker warrants further investigation.

Several other biomarkers have been explored for DM1, each offering unique insights into disease pathology. The Myotonic Dystrophy Splice Index, derived from muscle biopsies, provides a reliable measure of molecular pathophysiology.⁴² However, the invasive nature limits frequent longitudinal use. Second, blood-based markers, including serum protein panels,⁴³ plasma neurofilament light,⁴⁴ and p-tau181,⁴⁵ are minimally invasive and correlate with disease severity, but studies still need to evaluate their utility for longitudinal monitoring.⁴⁴ Finally, vHOT provides a noninvasive quantification of myotonia and is currently used as a primary end point in several DM1 clinical trials.⁴⁶ Together with qMRI, these biomarkers can provide complementary information from molecular mechanisms to macroscopic muscle changes.

Both the MFM32 score and MRC sum score showed significant deteriorations within a 6-month interval, with high SRM for the MFM32 total score and D1 subscore. By contrast, D2 and D3 subscores were less responsive (SRM = 0.4 and 0.0, respectively), likely reflecting the ceiling effect observed in the D2 and D3 domains during earlier disease stages.^18,47 However, these subscores may become more informative in later stages, as they capture the decline of remaining motor functions after loss of ambulation.¹⁸ Regarding the MRC sum score, a 4-year longitudinal study.¹² reported a significant reduction in the distal muscle groups but found no significant overall change over the follow-up period. The significant decrease in MRC sum score observed in our study likely reflects the inclusion of hallux extension, finger extension, and finger abduction, which are particularly sensitive to disease progression in DM1.¹⁷

Another key clinical feature and therapeutic target in DM1 is myotonia. Quantified using (video) HOT, it is often used as a primary end point in clinical trials (eTable 1). Despite a moderate test-retest reliability (ICC = 0.71), HOT remains the gold standard for myotonia assessment.⁴⁶ In this study, no significant longitudinal changes in HOT were observed. This contrasts with an earlier longitudinal study, which reported a significant decline in HOT over time and demonstrated an association between reduced HOT and declining hand grip strength.¹² Differences in follow-up duration, sample size, or included DM1 subtypes may explain these discrepancies. In our cohort, both HOT and hand grip strength remained relatively stable over time and were not significantly associated. A nonsignificant decrease in dominant hand grip strength was observed, consistent with natural history data (−0.24 kg/y).⁴⁸

Finally, 6MWD is often used as a clinical outcome measure in neuromuscular disorders. Consistent with a previous study¹³ in patients with DM1, the decline in 6MWD in our study did not reach statistical significance.

Several PROMs were included in this study to capture the patients' perspectives on disease progression. DM1-ActivC demonstrated significant declines at each visit, consistent with previous studies.^12,14 However, the rate of decline measured in our cohort was greater compared with those studies.^12,14 This might reflect the differences in DM1 subtypes included in the studies, as a previous study.¹² observed improvement over time in patients with childhood-onset DM1 and decreases in those with adult-onset and late-onset DM1. Similarly, the INQoL-Activities (at 12, 18, and 24 months), relationships (at 12 and 24 months), and pain (at 18 and 24 months) domains showed significant declines. The sensitivity of the INQoL to pain progression is particularly relevant, given the high prevalence of chronic pain in DM1.⁴⁹ While the INQoL detected significant changes in pain, the BPI did not, suggesting that the INQoL might be more sensitive to assess pain in adult-onset DM1. Finally, the deterioration across multiple INQoL domains led to a significant decline in the overall quality of life already after a 12-month period, highlighting the substantial multidimensional disease burden experienced by this population.⁵⁰

Although the sample size was relatively small, it is comparable to those of the few other quantitative MRI studies in adult-onset DM1.^9,10 This may limit the generalizability of the findings and reduce statistical power to detect smaller longitudinal changes in some outcome measures. In addition, we focused on patients with adult-onset DM1 who were still ambulatory, demonstrated distal muscle weakness, and had no MRI contraindications. Although these are often patients targeted for clinical trials, this introduces a selection bias and does not allow the natural history to be captured across all disease stadia and DM1 subtypes. Formal respiratory muscle strength testing was not included because of equipment constraints. Swallowing difficulty was assessed using the INQoL; however, this domain has not been validated against instrumental swallowing assessments, and a dedicated dysphagia-specific instrument was not included. In addition, myotonia was assessed using stopwatch-based HOT rather than vHOT, which offers superior reliability and is increasingly adopted as a primary end point in DM1 clinical trials. Inclusion of these measures, as well as presymptomatic or nonambulant cohorts, may help identify sensitive outcome measures across earlier and more advanced stages of DM1. A major strength of this study is the use of 3D, AI-assisted automated whole-muscle segmentations, rather than single-slice or partial muscle analyses, providing more accurate and sensitive assessment of MRI muscle changes. Another key strength is the comprehensive assessment of outcome measures, including a total of 28 individual leg muscles and 15 clinical and patient-reported outcome measures. This allowed us to compare the sensitivity of a broad range of measures in a single study in the same patients. The multiple follow-up time points over a total study period of 24 months enabled the definition and comparison of progression across different outcome measures, providing valuable insights into the design of future clinical trials. Overall, the methodological approach resulted in defining robust and sensitive outcome measures that are highly relevant for designing clinical trials in adult-onset DM1.

Muscle PDFF (%) is a sensitive and objective outcome measure for quantifying disease progression in patients with adult-onset DM1, with measurable changes observed within a 6-month interval. Changes of PDFF (%) progressed more rapidly in the distal leg muscles than in proximal leg muscles, highlighting their sensitivity to short-term disease progression. Furthermore, muscle PDFF (%) showed strong correlations with multiple clinical outcome measures, supporting its relevance for functional assessment. Over longer intervals, MFM32 score and MRC sum score demonstrate responsiveness comparable to distal leg PDFF (%), providing accessible alternatives. Combined with PROMs including DM1-ActivC and INQoL, all these measures provide a comprehensive assessment of disease progression. The choice of outcome measure should be guided by the duration and design of future clinical trials, with PDFF (%) being particularly advantageous for shorter trials and MFM32 score and MRC sum score offering comparable sensitivity over longer time frames.

Glossary

10MWT

10-meter walk test

30SSS

30-second sit-to-stand test

6MWD

6-minute walk distance

9HPT

9-hole peg test

BPI

Brief Pain Inventory

CNN

convolutional neural network

CTG

cytosine-thymine-guanine

DM1

myotonic dystrophy type 1

DM1‐ActivC

Rash‐built Myotonic Dystrophy Type 1 Activity and Participation Scale

HOT

hand opening time

INQoL

Individualized Neuromuscular Quality of Life

MCID

minimal clinically important difference

MFM32

32-item motor function measure

MIRS

Muscular Impairment Rating Scale

PDFF

proton density fat fraction

PROM

patient-reported outcome measure

qMRI

quantitative magnetic resonance imaging

SRM

standardized response mean

T2_H2O

water T2

Author Contributions

L. Iterbeke: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. L. Huysmans: drafting/revision of the manuscript for content, including medical writing for content; analysis or interpretation of data. K. Bamps: drafting/revision of the manuscript for content, including medical writing for content; analysis or interpretation of data. R. Peeters: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data. V. Goosens: drafting/revision of the manuscript for content, including medical writing for content; analysis or interpretation of data. F. Maes: drafting/revision of the manuscript for content, including medical writing for content; analysis or interpretation of data. P. Dupont: drafting/revision of the manuscript for content, including medical writing for content; analysis or interpretation of data. K.G. Claeys: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data.

Study Funding

Research funding provided by Vertex Pharmaceuticals and the Kan-GO! Fund of KU Leuven and was supported in part by the Flemish Government under the Onderzoeksprogramma Artificiële Intelligentie (AI) Vlaanderen program.

Disclosure

L. Iterbeke, L. Huysmans, K. Bamps, R. Peeters, V. Goosens, F. Maes, P. Dupont, and K. G. Claeys report no disclosures relevant to the manuscript. Go to Neurology.org/N for full disclosures.