Abstract
Objective
To understand the natural disease upper limb progression over 3 years of ambulatory and non‐ambulatory patients with Duchenne muscular dystrophy (DMD) using functional assessments and quantitative magnetic resonance imaging (MRI) and to exploratively identify prognostic factors.
Methods
Forty boys with DMD (22 non‐ambulatory and 18 ambulatory) with deletions in dystrophin that make them eligible for exon 53‐skipping therapy were included. Clinical assessments, including Brooke score, motor function measure (MFM), hand grip and key pinch strength, and upper limb distal coordination and endurance (MoviPlate), were performed every 6 months and quantitative MRI of fat fraction (FF) and lean muscle cross sectional area (flexor and extensor muscles) were performed yearly.
Results
In the whole population, there were strong nonlinear correlations between outcome measures. In non‐ambulatory patients, annual changes over the course of 3 years were detected with high sensitivity standard response mean (|SRM| ≥0.8) for quantitative MRI‐based FF, hand grip and key pinch, and MFM. Boys who presented with a FF<20% and a grip strength >27% were able to bring a glass to their mouth and retained this ability in the following 3 years. Ambulatory patients with grip strength >35% of predicted value and FF <10% retained ambulation 3 years later.
Interpretation
We demonstrate that continuous decline in upper limb strength, function, and MRI measured muscle structure can be reliably measured in ambulatory and non‐ambulatory boys with DMD with high SRM and strong correlations between outcomes. Our results suggest that a combination of grip strength and FF can be used to predict important motor milestones.
Introduction
Duchenne muscular dystrophy (DMD) is an X‐linked neuromuscular disease caused by mutations in the dystrophin gene that affects approximately 1 in every 5000 boys. 1 Progressive muscle weakness starts before the age of 5 years and leads to death around the end of the third decade. 2 Treatment with glucocorticoids and advances in multidisciplinary care programs have delayed the loss of ambulation and considerably increased the life expectancy. 3 , 4
Over the last 7 years, several clinical trials have been conducted in non‐ambulatory patients with DMD (NCT02814019, NCT03406780, and NCT03603288). 5 No drug has yet demonstrated sufficient efficacy in a double‐blind placebo‐controlled study to result in a market approval, though five drugs have been FDA‐approved on the basis of retrospective comparison or surrogate biomarker. So far, important aspects for quality of life, like upper limb function and lung capacity were used as primary endpoints. Understanding upper limb involvement and weakness progression and the validation of outcomes sensitive to change and the ability to predict natural course of loss of various functional abilities are of primary importance in evaluating drug efficacy in patients at risk of losing ambulation or who are already non‐ambulatory.
We previously showed that quantitative magnetic resonance imaging (MRI) and precise strength measures could reliably measure disease progression over a 1‐year period. 6 , 7 The objective of the present study was to investigate upper limb evolution over a longer time frame using strength and function assessments, and quantitative MRI. In addition, we initiated an explorative approach about identifying predictive values of these measures on important clinical milestones such as loss of ambulation or loss of hand‐to‐mouth ability in boys with DMD.
Material and Methods
Participants and study design
The 7‐year natural history study included patients with DMD eligible for exon‐skipping therapy. Subjects were evaluated in two investigator centers (GOSH in London and I‐Motion in Paris). Patients were recruited directly from these two centers. In order to increase our sample, we also called on Belgian, Romanian, Polish, and German networks. Initially, ambulatory and non‐ambulatory patients were included, and glucocorticoid status was not an inclusion criterion. Ambulation was defined as being able to walk 10 meters without any kind of assistance. In August 2015, in anticipation of future clinical trials, an amendment to the initial study protocol (Amendment #3 in France, Amendment #1 in the United Kingdom) specified that henceforward only non‐ambulatory patients treated with glucocorticoids could be included. Throughout the whole study duration imaging and clinical raters were blinded to use of glucocorticoids. The study was approved by the local ERB (CPP Ile de France VI, ID RCB‐2010‐A01138‐31) and registered on clinicaltrials.gov (NCT01385917). All patients or legal guardians signed an informed consent. The study was strictly monitored according to established standard operating procedures (SOPs), ICH/GCPs, and current legislation for clinical trial regulation. The study details have been reported in more detail elsewhere. 6
Strength and functional assessments
Every 6 months patients underwent functional assessments with the MFM 32 (https://mfm‐nmd.org/?lang=en), 8 grip strength assessments with the MyoGrip and key pinch strength assessments with the MyoPinch (Ateliers Laumonier, Nesles‐La‐Vallée, France), and upper limb distal coordination and endurance assessments with the MoviPlate (Valotec, Villejuif, France). 9 The physiotherapists conducting the assessments were trained and certified for the MFM and the MyoTools (MyoGrip, MyoPinch, and Moviplate). They followed strict SOPs as previously described. 6 , 10 , 11 Grip and pinch strengths are reported either in absolute values (kg) or in percentage of predicted normal values for age (%pred).
MRI
All quantitative MRI data were acquired annually on either a 3‐T (France, Paris) or a 1.5‐T (United Kingdom, London) clinical system (Siemens Healthineers). The reproducibility of the examinations performed on the 1.5T and 3T scanners in London and Paris was assessed on four subjects, two healthy volunteers and two patients with an autoimmune disease. The inter‐site variability, as estimated by the mean standard deviation between sites, was 1% for lean cross‐sectional areas (lCSA), 1.2% for Fat fraction (FF) and 1.2 msec for water T2 (unpublished data from the BIOIMAGE‐NMD project, FP7 funded translational research program) for data processed by the same operator. These data indicate a low impact of the environment, i.e., scanner, magnetic field, imaging sequences, technologists, and on the results obtained in the two participating centers. Patients were placed in the supine head‐first position, and dominant and nondominant forearms were scanned in two subsequent sessions on the same day. Quantitative water‐fat imaging was performed using a 3D gradient echo (3‐point Dixon) sequence with three echo times (TE1, TE2, and TE3 values of 2.75, 3.95, and 5.15 msec, respectively) with the volume centered on the thickest part of the forearm. Total MRI acquisition time was approximately 20 min per forearm. Adequate repositioning of the image volume at follow‐up exams was performed using the images of the preceding year. Regions of interest were drawn manually, by the same person, in the flexor and the extensor muscle group of each forearm. FF values (expressed in absolute %) and lCSA (expressed in mm²) and defined as the lean muscle cross‐sectional area corresponding to the muscle fraction containing the contractile apparatus were computed as previously reported in both the flexors and the extensors. 6
Statistical analyses
Patients were classified according to their ambulation status at each visit. Descriptive statistics were computed at baseline for both groups of patients as mean and standard deviation (SD). Standard response means (SRMs) were computed to quantify the effect size of the changes between visits and were calculated as the mean change score divided by the SD of change scores. An SRM ≥0.8 in absolute value was considered to reflect a high responsiveness to change. Differences between dominant and nondominant sides and between visits were tested using a Wilcoxon signed‐rank test.
Differences between groups were tested using a Mann–Whitney test. Spearman’s rho correlation coefficients were computed to explore correlations between variables. We also performed a Mann–Whitney test in both ambulatory and non‐ambulatory populations to test the effect of glucocorticoids compared to treatment‐naïve patients over time on the various outcome measures. All analyses were performed using IBM SPSS v.22.0 statistical software. A p < 0.05 was considered as statistically significant and a Bonferroni correction was applied in case of multiple tests. SRM and changes from baseline are presented for the first three years of the seven years of follow‐up; loss of subjects to follow‐up precluded longer term analysis.
Data availability
Anonymized data can be made available to qualified investigators on request. All data requests will be reviewed by the study executive committee.
Results
Population description at baseline
Forty DMD patients presenting with mutations theoretically treatable by correction of skipping of dystrophin exon 53 were included in the study. Mean duration of follow‐up was 3.5 ± 1.8 years yielding 288 visits. The number of patients pursuing the study decreased with time (Fig. 1). Reasons for drop‐out varied. For example, five ambulatory patients entered therapeutic trials and six others became non‐ambulatory during the study. Four patients died during the 7 years of the study, at ages ranging from 14 to 19 years. The causes of death were respiratory complications (n = 2) and cardiac arrest (n = 2). At time of death, one patient was on glucocorticoids, one patient stopped 10 years before and two patients never initiated the treatment. These latter required nocturnal noninvasive ventilation until death. Characteristics of patients at baseline are presented in Table 1.
Figure 1.
CONSORT diagram. Reasons for premature withdrawal included: (A) loss of follow‐up, (B) withdrawal, (C) physician decision to withdraw, (D) enrollment into other clinical trials, (E) death, (F) missed visit, (G) last visit missed, (H) protocol completed, (
) loss of ambulation.
Table 1.
Baseline characteristics.
| A | NA | All | |
|---|---|---|---|
| N | 18 | 22 | 40 |
| Age (years) | 9.1 (2.5) | 13.8 (2.5) | 11.7 (3.4) |
| Weight (kg) | 32.7 (13.5) | 45.6 (15.7) | 39.6 (15.9) |
| Height (cm) | 125.8 (11.2) | 144.8 (11.9) | 136.1 (15.0) |
| BMI (kg/m²) | 19.3 (4.0) | 20.6 (6.9) | 20.0 (5.6) |
| Number of patients with right dominant side (%) | 16 (88.9%) | 21 (95.5%) | 37 (92.5%) |
| Age of onset of first symptoms (years) | 2.6 (1.3) | 3.3 (1.6) | 3.0 (1.5) |
| Age at diagnosis (years) | 4.0 (2.5) | 5.1 (2.1) | 4.6 (2.3) |
| Age at loss of ability to walk (years) | NA | 9.2 (2.0) | NA |
| Number of patients on glucocorticoids (%) | 15 (83.3%) | 14 (63.6%) | 29 (72.5%) |
| Brooke score | 1.3 (0.5) | 4.0 (1.4) | 2.8 (1.7) |
| Walton score | 3.9 (1.5) | 8.9 (0.8) | 6.7 (2.7) |
| Grip strength (kg) – dominant side | 6.9 (3.0) | 3.3 (2.4) | 4.9 (3.2) |
| Grip strength (%pred) – dominant side | 49.3 (16.3) | 13.4 (11.0) | 29.5 (22.6) |
| Grip strength (kg) – nondominant side | 6.9 (2.8) | 3.3 (2.7) | 4.9 (3.3) |
| Grip strength (%pred) – nondominant side | 49.8 (17.4) | 13.1 (11.2) | 29.6 (23.6) |
| Pinch strength (kg) – dominant side | 2.2 (0.7) | 1.1 (0.6) | 1.6 (0.8) |
| Pinch strength (%pred) – dominant side | 51.0 (13.8) | 18.0 (10.3) | 32.9 (20.4) |
| Pinch strength (kg) – nondominant side | 2.1 (0.6) | 1.1 (0.7) | 1.6 (0.8) |
| Pinch strength (%pred) – nondominant side | 49.4 (14.0) | 17.6 (11.2) | 31.9 (20.3) |
| MoviPlate score (#) – dominant side | 47.0 (15.3) | 42.9 (11.0) | 44.7 (13.1) |
| MoviPlate score (#) – nondominant side | 42.6 (13.4) | 42.5 (11.0) | 42.6 (12.0) |
| MFM total score (%) | 77.6 (11.8) | 34.1 (12.2) | 53.7 (24.9) |
| MFM D1 (%) | 56.7 (22.9) | 1.3 (2.6) | 26.2 (31.8) |
| MFM D2 (%) | 64.1 (7.0) | 48.2 (23.4) | 68.8 (29.2) |
| MFM D3 (%) | 88.1 (9.7) | 70.58 (15.1) | 78.6 (15.5) |
| FF extensors (%) – dominant side | 9.7 (5.6) | 32.1 (18.6) | 23.1 (18.4) |
| FF extensors (%) – nondominant side | 9.0 (5.6) | 35.2 (17.2) | 24.4 (18.8) |
| FF flexors (%) – dominant side | 11.3 (8.5) | 39.5 (19.6) | 27.8 (21.1) |
| FF flexors (%) – nondominant side | 11.2 (7.5) | 39.8 (16.7) | 28.6(19.7) |
| lCSA extensors (mm²) – dominant side | 417.6 (119.4) | 310.4 (79.1) | 348.5 (106.9) |
| lCSA extensors (mm²) – nondominant side | 420.1 (123.2) | 325.8 (103.3) | 361.6 (118.6) |
| lCSA flexors (mm²) – dominant side | 650.9 (206.5) | 444.3 (169.9) | 515.3 (205.7) |
| lCSA flexors (mm²) – nondominant side | 641.1 (165.6) | 414.5 (148.1) | 500.5 (188.8) |
Results are presented as means ± SD, except for number of patients with right dominant side and number of patients on glucocorticoids where results are presented in percentage. Abbreviations: A, ambulatory; NA, non‐ambulatory; BMI, body mass index; MFM, motor function measure; FF, fat fraction; lCSA, lean muscle cross‐sectional area.
With the exception of the MoviPlate scores, all the variables were significantly different between ambulatory and non‐ambulatory patients (all p values < 0.001). In individual patients, bilateral measurements were highly correlated (rho > 0.89, all p values <0.001). Strength measures, FF, and lCSA (flexors and extensors) were not statistically different between sides. MoviPlate scores were significantly better on the dominant side (p < 0.001).
Sensitivity to change
In addition to differences in relationships with time (e.g., exponential decrease for grip strength and sigmoidal increase for FF), individual trajectories of all outcome measures were heterogeneous (Fig. 2). Significant reductions relative to baseline were observed at all visits for grip and pinch strength in non‐ambulatory patients and at 12, 24, and 36 months in the overall population. Total MFM value was significantly decreased with high SRM values relative to baseline starting from 24 months in both non‐ambulatory and overall patient populations. FF increase in the flexor muscle group was significant (with high SRMs) from 12 months onwards in the non‐ambulatory population and from 24 months in the overall patient population. The lCSA (flexors and extensors) values tended to decrease with time but not always significantly (Table 2).
Figure 2.
Longitudinal evolution of clinical outcome measures. Outcomes (A) grip strength as percentage of predicted value, (B) pinch strength as percentage of predicted value, (C) MoviPlate score, (D) MFM total score expressed as percentage, (E) Brooke score (scale from 1 to 6), (F) Walton score, (G) FF expressed as percentage, (H) lCSA expressed as square millimeter of the flexors as a function of age in years for non‐ambulatory (red) and ambulatory (blue) subjects. Each line connects data points for an individual subject.
Table 2.
Standardized response means and differences from baseline (Δ) for the various outcome measures.
| 6 months | 12 months | 24 months | 36 months | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| A (n = 13) | NA (n = 20) | All (n = 33) | A (n = 13) | NA (n = 20) | All (n = 33) | A (n = 9) | NA (n = 19) | All (n = 28) | A (n = 4) | NA (n = 18) | All (n = 22) | ||
| Brooke score | Δ | 0.12 | 0.24 | 0.18 | 0.31 | 0.30 | 0.30 | 0.56 | 0.90 | 0.79 | 0.50 | 1.11 | 1.00 |
| SRM | 0.24 | 0.44 | 0.36 | 0.49 | 0.37 | 0.42 | 1.06 | 0.90 | 0.90 | 0.87 | 1.15 | 1.08 | |
| Walton score | Δ | 0.82 | 0.05 | 0.40 | 1.00 | 0.20 | 0.52 | 0.67 | 0.90 | 0.82 | 1.25 | 1.17 | 1.18 |
| SRM | 0.77 | 0.10 | 0.45 | 0.61 | 0.38 | 0.45 | 0.60 | 0.72 | 0.69 | 0.50 | 0.66 | 0.64 | |
| Grip strength (kg) | Δ | 0.28 | −0.15 | 0.05 | 0.45 | −0.26 | 0.02 | 0.78 | −0.55 | −0.13 | 2.73 | −1.12 | −0.42 |
| SRM | 0.31 | −0.28 | 0.06 | 0.24 | −0.50 | 0.01 | 0.43 | −0.57 | −0.09 | 1.79 | −0.90 | −0.21 | |
| Grip strength (%pred) | Δ | −0.96 | −1.40 | −1.21 | −1.36 | −3.11 | −2.42 | −6.09 | −6.83 | −6.59 | −2.81 | −10.72 | −9.28 |
| SRM | −0.11 | −0.72 | −0.20 | −0.09 | −0.96 | −0.25 | −0.48 | −1.04 | −0.75 | −0.27 | −1.09 | −0.91 | |
| Pinch strength (kg) | Δ | 0.14 | −0.08 | 0.02 | 0.14 | −0.13 | −0.02 | 0.13 | −0.14 | −0.05 | 0.62 | −0.34 | −0.17 |
| SRM | 0.30 | −0.58 | 0.06 | 0.39 | −0.56 | −0.08 | 0.53 | −0.52 | −0.18 | 5.30 | −0.99 | −0.34 | |
| Pinch strength (%pred) | Δ | −0.32 | −1.78 | −1.13 | −1.91 | −3.56 | −2.91 | −4.83 | −5.55 | −5.32 | −0.31 | −10.06 | −8.29 |
| SRM | −0.04 | −0.86 | −0.18 | −0.36 | −0.77 | −0.59 | −1.01 | −1.27 | −1.20 | −0.10 | −1.35 | −1.06 | |
| MoviPlate score (#) | Δ | 2.71 | 3.50 | 3.14 | 3.46 | 1.30 | 2.15 | 5.67 | 4.16 | 4.64 | 17.33 | 1.83 | 4.05 |
| SRM | 0.38 | 0.80 | 0.54 | 0.47 | 0.15 | 0.26 | 0.59 | 0.35 | 0.42 | 1.87 | 0.14 | 0.30 | |
| MFM total score (%) | Δ | −0.93 | −1.73 | −1.36 | −3.57 | −2.16 | −2.73 | −5.37 | −7.19 | −6.61 | −14.34 | −14.18 | −14.21 |
| SRM | −0.22 | −0.38 | −0.31 | −0.51 | −0.50 | −0.50 | −0.62 | −1.18 | −0.96 | −0.58 | −1.09 | −0.95 | |
| MFM D1 (%) | Δ | −3.64 | 0.01 | −1.67 | −0.27 | −7.89 | −3.36 | −6.21 | −8.54 | −6.96 | −9.68 | −19.89 | −11.54 |
| SRM | −0.41 | 0.00 | −0.26 | −0.53 | −0.23 | −0.34 | −0.49 | −0.51 | −0.50 | −0.74 | −0.57 | −0.61 | |
| MFM D2 (%) | Δ | −0.21 | −4.93 | −2.76 | −1.29 | −5.78 | −3.96 | −4.63 | −10.14 | −8.37 | −20.13 | −20.90 | −20.76 |
| SRM | −0.05 | −0.56 | −0.38 | −0.41 | −0.58 | −0.49 | −0.63 | −1.19 | −0.99 | −0.61 | −1.10 | −0.98 | |
| MFM D3 (%) | Δ | 2.60 | 0.47 | 1.45 | 0.72 | −2.02 | −0.90 | −0.52 | −4.02 | −2.90 | 5.98 | −11.11 | −8.00 |
| SRM | 0.33 | 0.05 | 0.17 | 0.09 | −0.23 | −0.11 | −0.05 | −0.37 | −0.27 | 0.56 | −0.99 | −0.63 | |
| FF (%) extensors | Δ | — | — | — | 0.09 | 2.95 | 1.83 | 1.72 | 10.70 | 8.25 | 3.85 | 15.36 | 13.83 |
| SRM | — | — | — | 0.04 | 0.39 | 0.30 | 0.32 | 1.12 | 0.87 | 0.79 | 1.60 | 1.40 | |
| FF (%) flexors | Δ | — | — | — | 0.49 | 4.42 | 2.82 | 1.25 | 11.66 | 8.82 | 6.15 | 15.59 | 14.33 |
| SRM | — | — | — | 0.14 | 0.81 | 0.54 | 0.68 | 1.38 | 0.98 | 0.91 | 2.21 | 1.66 | |
| lCSA (mm²) extensors | Δ | — | — | — | 6.43 | −28.77 | −14.37 | 31.88 | −35.66 | −16.36 | 34.70 | −51.05 | −39.62 |
| SRM | — | — | — | 0.15 | −1.17 | −0.39 | 0.53 | −0.65 | −0.26 | 0.51 | −1.05 | −0.70 | |
| lCSA (mm²) flexors | Δ | — | — | — | 8.06 | −58.15 | −32.24 | 42.93 | −75.50 | −43.20 | 64.82 | −86.05 | −65.93 |
| SRM | — | — | — | 0.11 | −0.91 | −0.43 | 0.67 | −0.78 | −0.42 | 0.83 | −1.00 | −0.67 | |
Significant negative changes are highlighted in gray. |SRM|≥0.8 are in bold (when in the expected direction). In case of bilateral measures, only the dominant side is shown. Abbreviations: A, ambulatory; NA, non‐ambulatory; MFM, motor function measure; FF, fat fraction; lCSA, lean muscle cross‐sectional area.
Patients treated with glucocorticoids were not evenly distributed over the various age groups. The large majority of those treated with glucocorticoids were young, ambulatory patients. Thus, these results present a critical bias and a limited interpretation. Non‐ambulatory patients on steroids at baseline were stronger than the treatment‐naive population but yearly changes were not statistically different. As only three treatment‐naive ambulatory patients were included, we found no significant differences, neither at baseline nor in follow‐up (Table S1).
Correlations between outcome measures
Strong significant correlations were observed between total MFM and relative grip and pinch strength scores. Equally strong correlations were found between FF (flexors and extensors) and grip and pinch relative values and between lCSA and grip and pinch absolute values. The strongest correlations were observed between FF and total MFM. Table 3 summarizes the correlation analyses. Relationships between variables were generally not linear (Fig. 3).
Table 3.
Correlations between variables.
| Brooke | Walton | Grip strength (kg) | Grip strength (%) | Pinch strength (kg) | Pinch strength (%) | MoviPlate (#) | MFM total score (%) | FF flexors (%) | FF extensors (%) | lCSA flexors (mm²) | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Walton | 0.906 | ||||||||||
| Grip strength (kg) | −0.691 | −0.713 | |||||||||
| Grip strength (%) | −0.772 | −0.771 | 0.909 | ||||||||
| Pinch strength (kg) | −0.747 | −0.735 | 0.903 | 0.898 | |||||||
| Pinch strength (%) | −0.793 | −0.773 | 0.855 | 0.950 | 0.960 | ||||||
| MoviPlate (#) | −0.430 | −0.388 | 0.552 | 0.399 | 0.580 | 0.450 | |||||
| MFM total score (%) | −0.826 | −0.842 | 0.751 | 0.860 | 0.836 | 0.877 | 0.444 | ||||
| FF flexors (%) | 0.888 | 0.870 | −0.776 | −0.883 | −0.849 | −0.892 | −0.305 | −0.921 | |||
| FF extensors (%) | 0.868 | 0.829 | −0.656 | −0.799 | −0.773 | −0.831 | −0.210 | −0.975 | 0.967 | ||
| lCSA flexors (mm²) | −0.613 | −0.677 | 0.865 | 0.800 | 0.820 | 0.781 | 0.396 | 0.742 | −0.757 | −0.646 | |
| lCSA extensors (mm²) | −0.609 | −0.620 | 0.716 | 0.800 | 0.746 | 0.773 | 0.176 | 0.721 | −0.772 | −0.745 | 0.841 |
Values are Spearman rho correlation coefficients. Significant correlations are highlighted in gray. When bilateral, only the dominant side was considered. Abbreviations: MFM, motor function measure; FF, fat fraction; lCSA, lean muscle cross‐sectional area.
Figure 3.
Relationship between grip strength and MFM total score, FF, and lCSA. Correlations of (A) percentage of predicted grip strength to MFM total score, (B) percentage of predicted grip strength to FF of the flexors, (C) grip strength absolute value (expressed in kg) to lCSA of the flexors (expressed in mm2). Each line connects data points for individual non‐ambulatory (red) or ambulatory (blue) subject.
Exploratory predictive approach
In order to identify factors predictive of clinically significant milestones, we assessed the correlation between different variables: the changes in grip strength at 6 or 12 months or flexor FF with the probability of losing ambulation or with the loss of 3 points on the MFM or 1 point on the Brooke score within 3 years. Some weakly significant correlations were detected, but the high interpatient variability precluded the identification of a strong predictive factor using a single variable (data not shown).
We then investigated whether the combination of strength and MRI‐based data were predictive of clinically significant milestones and abilities, such as the loss of ambulation or the loss of the ability to drink a glass of water independently (equivalent to Brooke score 3). The combination of low flexor FF and high grip strength was a 3‐year positive prognosis factor for maintaining ambulation and for the ability to raise a glass to the mouth independently. As indicated in plots of grip strength (%pred) versus flexor FF (% of baseline), patients with a FF of greater than 10% and a grip strength less than 35% at any point of the study had either lost ambulation or would lose it in the following 3 years (Fig. 4A–C). In contrast, patients with a flexor FF less than 10% and a grip strength greater than 35% at any point were ambulatory and did not lose ambulation within the next 3 years. The combination of grip strength and FF was also predictive of the Brooke score: boys who presented at any point with a FF less than 20% and a grip strength greater than 27% retained the ability to independently bring a glass of water to their mouth for the following 3 years (Fig. 4D–F).
Figure 4.
Correlation between flexor FF and grip strength with clinical parameters. Grip strength (%pred) versus FF in the flexors (% of baseline) for (A) subjects who were ambulatory (blue), non‐ambulatory (red), or changed between baseline and 3 years (blue/red) (n = 25), (B) subjects who were ambulatory (blue), non‐ambulatory (red), or changed between 12 months and 4 years (blue/red) (n = 15 from the cohort baseline 3 years), (C) subjects who were ambulatory (blue), non‐ambulatory (red), or changed between 2 and 5 years (blue/red) (n = 12 from the cohort 1–4 years), (D) subjects who had Brooke score ≤3 (black) or >3 (gray) or who changed (black/gray) between baseline and year 3 (n = 25), (E) subjects who had Brooke score ≤3 (black) or >3 (gray) or who changed (black/gray) between 12 months and year 4 (n = 15 from the cohort baseline 3 years), and (F) subjects who had Brooke score ≤3 (black) or >3 (gray) or who changed (black/gray) between year 2 and year 5 (n = 12 from the cohort 1–4 years).
Discussion
This study of DMD patients followed at least 3 years showed that high precision dynamometry and muscle MRI can reliably be used over time to assess upper limb performance evolution in ambulatory as well as non‐ambulatory boys who have mutations theoretically eligible for exon 53 skipping. The SRMs are indicative of the suitability of these techniques for use as outcome measures in a 1‐year clinical trial.
Measurements of motor function, strength, or daily activities such as the performance of upper limb (PUL), 12 , 13 the Brooke score, 14 the 9‐Hole Peg test, 14 MyoGrip/MyoPinch, 9 ActiMyo, 15 and the MFM 8 as well as questionnaires like the Egen Klassifikation 14 , 16 , 17 have been used to evaluate changes over time in non‐ambulatory DMD patients. Different blood and urine biomarkers responsive to treatment have also been proposed, but none are currently qualified as surrogate endpoints or as fully qualified biomarkers. 18 Quantitative MRI‐based evaluation of the progression of muscle fat replacement is a well‐established outcome in natural history studies in ambulatory DMD patients 19 and is also suitable for upper limb skeletal muscle evaluation, as previously shown in non‐ambulatory DMD patients. 6 , 7 , 20 Clinical trial readiness in this population requires a better understanding of the long‐term robustness and sensitivity with which various measures are able to detect a change over a given period of time.
The population in this study is comparable to those in other prospective natural history studies conducted in non‐ambulatory patients with DMD. In all but two studies that included older patients, 10 , 12 the overall mean ages at baseline were similar to that of this cohort. 14 , 21 , 22 Although only one prospective study has included exclusively glucocorticoid‐treated patients, 21 it is also common to include patients who have discontinued treatment once becoming wheelchair‐bound or even treatment‐naive patients. 7 , 10 , 12 , 14 , 22 In terms of functional status, the mean MFM total score in our study was 34%, which corresponds to the MFM score of a previous study that included only glucocorticoid‐treated patients. 21 Other studies included patients with a Brooke score ranging from 1 to 5; 10 , 14 the mean Brooke score of ambulatory patients in our cohort was 1.3 and that of non‐ambulatory subjects was 4.0.
Our population included only patients theoretically treatable with agents that alter splicing of dystrophin exon 53, who present with a more severe phenotype than other patients with DMD. 23 This study was initially planned to pave the way for a clinical trial in this population. 24 Given the approval of golodirsen, an antisense oligonucleotide that induces skipping of exon 53, 25 , 26 and the development of other therapies with similar modes of action, the present set of data offers the opportunity to benchmark treated patients with natural history of patients with the same genotype.
Although data were collected over a period of 78 months in some patients, we limited the SRM analysis to the first 36 months to avoid bias related to different follow‐up durations. This study constitutes, to our knowledge, the longest prospective natural history study which included non‐ambulatory patients at baseline. In previous studies, only ambulatory patients had been followed for this length of time. 2 , 3 , 27 In non‐ambulatory subjects, cross sectional studies 4 , 16 , 28 , 29 and prospective studies with 12‐month, 6 , 7 , 10 18‐month, 21 and 24‐month 12 , 13 , 14 , 22 follow‐up periods have been described.
We showed that in non‐ambulatory patients, changes were detected annually over the course of 3 years with high sensitivity (|SRM|≥0.8) for grip and pinch strength, function (MFM, Brooke), and composition of the forearm muscles (quantitative MRI‐based FF), illustrating the ability of these measures to quantify disease progression in individual patients accurately over 1 year and beyond. As in our previous study conducted in a smaller independent cohort, 10 we found a stronger correlation (rho = 0.86) between hand grip strength and function (MFM total score) than in another study. 28 We reproduced the strong correlations described between quantitative MRI‐based FF (both flexor and extensor) and function and strength measures previously demonstrated in cohorts of DMD 6 , 30 and spinal muscular atrophy 31 patients. The results of our study also confirm that when considering the strength data in absolute terms (i.e., in kg), the disease progression can be partially masked by the effect of the growth, especially in boys still in the ambulatory phase of the condition. This decreases the sensitivity to change of such data, making it necessary to express the strength relative to baseline values as recently suggested. 32
As a group, the DMD patients in our cohort presented with a clear disease progression; at an individual level, however, we observed a large heterogeneity in disease evolution on a clinical basis and from an imaging perspective. The difficulty of prediction of FF evolution at an individual level has already been highlighted based on studies of the lower limbs of ambulatory patients, although recent logistic modeling has shown promising improvements. 33 , 34 In our study, we observed a similar heterogeneity in the FF evolution in the forearm muscle groups with annual increases ranging from zero to 20%–25% depending on the individual; this range is similar to that previously reported for lower limb muscles. 20
Although our data indicate that FF (and lCSA), grip strength (expressed in %pred), and the MFM score are sensitive and reliable for use in the follow‐up of the natural evolution of the disease, correlating yearly changes with clinically meaningful endpoints remains challenging in view of this heterogeneity of trajectories. While the loss of ambulation is certainly the most commonly used milestone in DMD, 33 , 34 there is a need for other meaningful upper limb milestones for non‐ambulatory patients. In this regard, we investigated how the combined value of flexor FF and grip strength can predict the Brooke score evolution. This exploratory approach spotlighted the predictive nature of the combination of FF and grip strength of the 3‐year preservation of ambulation in ambulatory patients and the maintenance of the ability to carry a glass of water to the mouth in non‐ambulatory patients. Increased understanding of these relationships based on larger and independent cohorts will enable stratification of patients for future clinical trials. 6
At the time of preparing this natural history study, available and clinically relevant outcome measures were chosen that were sensitive to change in both ambulatory and non‐ambulatory patient groups. 8 , 9 , 35 , 36 On 1 January 2021, there are only three trials registered in clinicaltrial.gov that are currently recruiting non‐ambulatory patients (NCT03354039, NCT04371666, and NCT04004065), whereas five exclude non‐ambulatory patients. Non‐ambulatory subjects are likely excluded because these patients are not expected to respond as well as ambulatory subjects to treatment and because of decreased available muscle mass and increased retractions. As described in spinal muscular atrophy, disease duration is a constant predictive factor of treatment efficacy. 37 It has been thought that as non‐ambulatory patients constitute a more fragile population with higher comorbidity, changes in clinical condition would be more difficult to quantify than in ambulatory subjects. This study clearly demonstrates, however, that SRM of clinical and MRI endpoints in non‐ambulatory patients are comparable to those considered as acceptable in ambulatory patients. 38
This study has some limitations. First, the inclusion only of patients theoretically treatable with agents that promote exon 53 skipping may be perceived as a limitation, since this genotype results in the most severe symptoms, which could lead to an overestimation of the SRM. 23 Nevertheless, this restriction ensured a more homogeneous population, and this population is specifically targeted by clinical trials 25 (NCT02500381). Second, this study focused only on upper limbs; we did not evaluate global natural history such as cardiac and respiratory evolution. Furthermore, we had to limit the number of functional evaluation scales to reduce patient burden and to avoid fatigability induced by redundant measures. At the time the study was initiated, the PUL was not available. 13 , 39 This scale partially overlaps with the distal motor function domain of the MFM. Third, only two sites were involved in patient analysis; however, patients were included from different sites and countries that followed diverse guidelines of standard of care management. In fact, not all patients were on glucocorticoids at baseline and treatment initiation time, type, and posology varied, which may have influenced the disease course. 40 At the time of study initiation, glucocorticoids use in non‐ambulatory patients with DMD in France was very limited. Several studies have since then demonstrated the potential benefit of glucocorticoids including in the non‐ambulatory 29 and in the overall population, 41 and more and more non‐ambulatory patients are nowadays on glucocorticoids. Our study was not designed or powered to demonstrate glucocorticoids effect.
Conclusions
This study revealed a continuous decline as well as a strong correlation between upper limb muscle FF, function, and strength over a 3‐year period in non‐ambulatory and ambulatory DMD patients. Patients with DMD, even with a single genotype, constitute a heterogeneous population and some patients may present with an unpredictable clinical or MRI evolution. In addition, this study permitted an initiatory approach to predict patient evolution over a 3‐year period by using combined clinical and MRI evaluations. The added value of these longitudinal and correlative data of clinically relevant outcome measures will enable the scientific community and industry partners to better stratify patient populations for future clinical trials of the upper limb in non‐ambulatory patients with DMD.
Author Contributions
All the authors contributed to the study conception and design, analyzed and interpreted the data, commented on previous versions of the manuscript, and read and approved the final manuscript.
The natural history study was sponsored by Généthon and was registered on ClinicalTrials.gov under identification number NCT01385917. This project was supported by Association Française contre les Myopathies (AFM) and Advanced Diagnostics for New Therapeutic Approaches (ADNA), a program dedicated to personalized medicine, coordinated by Institut Mérieux, and supported by research and innovation aid from the French public agency OSEO. The funders were not involved in data collection or interpretation or preparation of the paper.
Conflict of Interest
CL has no direct disclosures to declare; she is an MFM, ActiMyo, and ATOM trainer. HR, AMS, TG, MA, VC, VD, IL, and JLL report no disclosures relevant to the manuscript. EG is president of BIOSSEC and was hired by Généthon for the statistical analyses. FM has no direct disclosures to declare; he consults for Pfizer, Sarepta, Santhera, and Dyne Therapeutics. JYH is a coinventor of the MyoGrip, MyoPinch, and MoviPlate. PGC receives support from the European Community and the Association Française contre les Myopathies. LS is a coinventor of the MoviPlate; he consults for Pfizer, Sarepta, Santhera, Catabasis, RegenexBio, Affinia, Biophytis, and Fibrogen.
Supporting information
Table S1. Mean and standard deviation (SD) in both ambulatory and non‐ambulatory populations to test the effect of glucocorticoids on the various outcome measures. The variables at baseline were not statistically different between glucocorticoid users and nonusers (p > 0.0038 with Bonferroni correction), except for MFM D2 and total score in non‐ambulatory patients. Their changes over time were not different. Note that the populations are not balanced in age and ambulatory status as explained in the main text.
Acknowledgments
The authors are grateful to the DMD patients and their families, to the North Star clinical network study, and to Muscular Dystrophy UK (MDUK) for the support of the North Star DMD Network. This work was also supported by the NIHR GOSH BRC in London.
The authors thank the members of the PreU7 Study Group (number in superscript refers to affiliation) (Clinical evaluation and follow‐up of patients: Ruxandra Cadras, 1 Laura Vanden Brande, 1 Silvana De Lucia, 1 Karolina Aragon‐Gawinska, 1 Mariacristina Scotto, 6 , 7 Valeria Ricotti, 6 , 7 Kate Maresh, 6 , 7 and Joana Domingos; 6 , 7 Coordination between Généthon and site: Sylvie Coulomb; Functional data acquisition: Allison Grangé, 1 Aurélie Canal, 1 Stéphanie Gilabert, 1 and Gwenn Gély‐Ollivier; 1 MRI data acquisition: Jean‐Marc Boisserie; 1 Organization of visits and data management: Nacera Reguiba 1 and Katie Groves 6 , 7 ) who were instrumental in the conduct of this study and collection of the data.
The authors also thank Simone Birnbaum and Jackie Wyatt for language and scientific editing.
Eric Guemas from BIOSSEC (mandated by the sponsor) and Jean‐Yves Hogrel, PhD, from Institute of Myology (academic) completed statistical analyses.
Funding Information
The authors are grateful to the DMD patients and their families, to the North Star clinical network study, and to Muscular Dystrophy UK (MDUK) for the support of the North Star DMD Network. This work was also supported by the NIHR GOSH BRC in London.
Funding Statement
This work was funded by Généthon; French public agency OSEO; Association Française contre les Myopathies (AFM) ; Advanced Diagnostics for New Therapeutic Approaches (Institut Mérieux); Muscular Dystrophy UK; NIHR GOSH BRC.
References
- 1. Crisafulli S, Sultana J, Fontana A, et al. Global epidemiology of Duchenne muscular dystrophy: an updated systematic review and meta‐analysis. Orphanet J Rare Dis 2020;15:141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Bello L, Gordish‐Dressman H, Morgenroth LP, et al. Prednisone/prednisolone and deflazacort regimens in the CINRG Duchenne Natural History Study. Neurology 2015;85:1048–1055. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Brogna C, Coratti G, Pane M, et al. Long‐term natural history data in Duchenne muscular dystrophy ambulant patients with mutations amenable to skip exons 44, 45, 51 and 53. PLoS One 2019;14:e0218683. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Connolly AM, Malkus EC, Mendell JR, et al. Outcome reliability in non‐ambulatory boys/men with Duchenne muscular dystrophy. Muscle Nerve 2015;51:522–532. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Mayer OH, Leinonen M, Rummey C, Meier T, Buyse GM; Group DS . Efficacy of idebenone to preserve respiratory function above clinically meaningful thresholds for forced vital capacity (FVC) in patients with Duchenne muscular dystrophy. J Neuromusc Dis 2017;4:189–198. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Hogrel J‐Y, Wary C, Moraux A, et al. Longitudinal functional and NMR assessment of upper limbs in Duchenne muscular dystrophy. Neurology 2016;86:1022–1030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Ricotti V, Evans MRB, Sinclair CDJ, et al. Upper limb evaluation in Duchenne muscular dystrophy: fat‐water quantification by MRI, muscle force and function define endpoints for clinical trials. PLoS One 2016;11:e0162542. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Vuillerot C, Girardot F, Payan C, et al. Monitoring changes and predicting loss of ambulation in Duchenne muscular dystrophy with the Motor Function Measure. Dev Med Child Neurol 2010;52:60–65. [DOI] [PubMed] [Google Scholar]
- 9. Servais L, Deconinck N, Moraux A, et al. Innovative methods to assess upper limb strength and function in non‐ambulant Duchenne patients. Neuromuscul Disord 2013;23:139–148. [DOI] [PubMed] [Google Scholar]
- 10. Seferian AM, Moraux A, Annoussamy M, et al. Upper limb strength and function changes during a one‐year follow‐up in non‐ambulant patients with Duchenne Muscular Dystrophy: an observational multicenter trial. PLoS One 2015;10:e0113999. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Seferian AM, Moraux A, Canal A, et al. Upper limb evaluation and one‐year follow up of non‐ambulant patients with spinal muscular atrophy: an observational multicenter trial. PLoS One 2015;10:e0121799. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Pane M, Coratti G, Brogna C, et al. Upper limb function in Duchenne muscular dystrophy: 24 month longitudinal data. PLoS One 2018;13:e0199223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Mayhew AG, Coratti G, Mazzone ES, et al. Performance of Upper Limb module for Duchenne muscular dystrophy. Dev Med Child Neurol 2020;62:633–639. [DOI] [PubMed] [Google Scholar]
- 14. Connolly AM, Florence JM, Zaidman CM, et al. Clinical trial readiness in non‐ambulatory boys and men with Duchenne muscular dystrophy: MDA‐DMD network follow‐up. Muscle Nerve 2016;54:681–689. [DOI] [PubMed] [Google Scholar]
- 15. Le Moing A‐G, Seferian AM, Moraux A, et al. A movement monitor based on magneto‐inertial sensors for non‐ambulant patients with Duchenne muscular dystrophy: a pilot study in controlled environment. PLoS One 2016;11:e0156696. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Lee HN, Sawnani H, Horn PS, et al. The Performance of the Upper Limb scores correlate with pulmonary function test measures and Egen Klassifikation scores in Duchenne muscular dystrophy. Neuromuscul Disord 2016;26:264–271. [DOI] [PubMed] [Google Scholar]
- 17. Steffensen B, Hyde S, Lyager S, Mattsson E. Validity of the EK scale: a functional assessment of non‐ambulatory individuals with Duchenne muscular dystrophy or spinal muscular atrophy. Physiother Res Int 2001;6:119–134. [DOI] [PubMed] [Google Scholar]
- 18. Szigyarto CA, Spitali P. Biomarkers of Duchenne muscular dystrophy: current findings. Degener Neurol Neuromuscul Dis 2018;8:1–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Ropars J, Gravot F, Ben Salem D, et al. Muscle MRI: a biomarker of disease severity in Duchenne muscular dystrophy? A systematic review. Neurology 2020;94:117–133. [DOI] [PubMed] [Google Scholar]
- 20. Willcocks RJ, Rooney WD, Triplett WT, et al. Multicenter prospective longitudinal study of magnetic resonance biomarkers in a large Duchenne muscular dystrophy cohort. Ann Neurol 2016;79:535–547. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Silva EC, Machado DL, Resende MB, et al. Motor function measure scale, steroid therapy and patients with Duchenne muscular dystrophy. Arq Neuropsiquiatr 2012;70:191–195. [DOI] [PubMed] [Google Scholar]
- 22. Killian M, Buchowski MS, Donnelly T, et al. Beyond ambulation: measuring physical activity in youth with Duchenne muscular dystrophy. Neuromuscul Disord 2020;30:277–282. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Servais L, Montus M, Guiner CL, et al. Non‐ambulant Duchenne patients theoretically treatable by exon 53 skipping have severe phenotype. J Neuromusc Dis 2015;2:269–279. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Le Guiner C, Montus M, Servais L, et al. Forelimb treatment in a large cohort of dystrophic dogs supports delivery of a recombinant AAV for exon skipping in Duchenne patients. Mol Ther 2014;22:1923–1935. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Frank DE, Schnell FJ, Akana C, et al. Increased dystrophin production with golodirsen in patients with Duchenne muscular dystrophy. Neurology 2020;94:e2270–e2282. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Scaglioni D, Catapano F, Ellis M, et al. The administration of antisense oligonucleotide golodirsen reduces pathological regeneration in patients with Duchenne muscular dystrophy. Acta Neuropathol Commun 2021;9:7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Arora H, Willcocks RJ, Lott DJ, et al. Longitudinal timed function tests in Duchenne muscular dystrophy: ImagingDMD cohort natural history. Muscle Nerve 2018;58:631–638. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Bulut N, Gurbuz I, Yilmaz O, et al. The association of hand grip strength with functional measures in non‐ambulatory children with Duchenne muscular dystrophy. Arq Neuropsiquiatr 2019;77:792–796. [DOI] [PubMed] [Google Scholar]
- 29. Demir G, Bulut N, Yılmaz Ö, et al. Manual ability and upper limb performance in nonambulatory stage of Duchenne muscular dystrophy. Arch Pediatr 2020;27:304–309. [DOI] [PubMed] [Google Scholar]
- 30. Forbes SC, Arora H, Willcocks RJ, et al. Upper and lower extremities in Duchenne muscular dystrophy evaluated with quantitative MRI and proton MR spectroscopy in a Multicenter Cohort. Radiology 2020;295:616–625. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Annoussamy M, Seferian AM, Daron A, et al. Natural history of Type 2 and 3 spinal muscular atrophy: 2‐year NatHis‐SMA study. Ann Clin Transl Neurol 2021;8:359–373. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Hogrel J‐Y, Decostre V, Ledoux I, et al. Normalized grip strength is a sensitive outcome measure through all stages of Duchenne muscular dystrophy. J Neurol 2020;267:2022–2028. [DOI] [PubMed] [Google Scholar]
- 33. Naarding KJ, Reyngoudt H, van Zwet EW, et al. MRI vastus lateralis fat fraction predicts loss of ambulation in Duchenne muscular dystrophy. Neurology 2020;94:e1386–e1394. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Rooney WD, Berlow YA, Triplett WT, et al. Modeling disease trajectory in Duchenne muscular dystrophy. Neurology 2020;94:e1622–e1633. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Brooke MH, Griggs RC, Mendell JR, et al. Clinical trial in Duchenne dystrophy. I. The design of the protocol. Muscle Nerve 1981;4:186–197. [DOI] [PubMed] [Google Scholar]
- 36. Mercuri E, McDonald C, Mayhew A, et al. International workshop on assessment of upper limb function in Duchenne Muscular Dystrophy: Rome, 15–16 February 2012. Neuromuscul Disord 2012;22:1025–1028. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Dangouloff T, Servais L. Clinical evidence supporting early treatment of patients with spinal muscular atrophy: current perspectives. Ther Clin Risk Manag 2019;15:1153–1161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Muntoni F, Domingos J, Manzur AY, et al. Categorising trajectories and individual item changes of the North Star Ambulatory Assessment in patients with Duchenne muscular dystrophy. PLoS One 2019;14:e0221097. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Mayhew A, Mazzone ES, Eagle M, et al. Development of the performance of the upper limb module for Duchenne muscular dystrophy. Dev Med Child Neurol 2013;55:1038–1045. [DOI] [PubMed] [Google Scholar]
- 40. Bylo M, Farewell R, Coppenrath VA, Yogaratnam D. A review of deflazacort for patients with Duchenne muscular dystrophy. Ann Pharmacother 2020;54:788–794. [DOI] [PubMed] [Google Scholar]
- 41. Schreiber A, Brochard S, Rippert P, et al. Corticosteroids in Duchenne muscular dystrophy: impact on the motor function measure sensitivity to change and implications for clinical trials. Dev Med Child Neurol 2018;60:185–191. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1. Mean and standard deviation (SD) in both ambulatory and non‐ambulatory populations to test the effect of glucocorticoids on the various outcome measures. The variables at baseline were not statistically different between glucocorticoid users and nonusers (p > 0.0038 with Bonferroni correction), except for MFM D2 and total score in non‐ambulatory patients. Their changes over time were not different. Note that the populations are not balanced in age and ambulatory status as explained in the main text.
Data Availability Statement
Anonymized data can be made available to qualified investigators on request. All data requests will be reviewed by the study executive committee.