Resveratrol improves motor function in patients with muscular dystrophies: an open-label, single-arm, phase IIa study

Muscular dystrophies (MDs) are inherited disorders characterized by progressive muscle weakness. Previously, we have shown that resveratrol (3,5,4′-trihydroxy-trans-stilbene), an antioxidant and an activator of the protein deacetylase SIRT1, decreases muscular and cardiac oxidative damage and improves pathophysiological conditions in animal MD models. To determine whether resveratrol provides therapeutic benefits to patients with MDs, an open-label, single-arm, phase IIa trial of resveratrol was conducted in 11 patients with Duchenne, Becker or Fukuyama MD. The daily dose of resveratrol was 500 mg/day, which was increased every 8 weeks to 1000 and then 1500 mg/day. Primary outcomes were motor function, evaluated by a motor function measure (MFM) scale, muscular strength, monitored with quantitative muscle testing (QMT), and serum creatine kinase (CK) levels. Adverse effects and tolerability were evaluated as secondary outcomes. Despite the advanced medical conditions of the patients, the mean MFM scores increased significantly from 34.6 to 38.4 after 24 weeks of medication. A twofold increase was found in the mean QMT scores of scapula elevation and shoulder abduction. Mean CK levels decreased considerably by 34%. Diarrhoea and abdominal pain was noted in six and three patients, respectively. Resveratrol may provide some benefit to MD patients.

www.nature.com/scientificreports/ oxygen species (ROS), inhibiting inflammation, and promoting mitochondrial function and autophagy 6,7 . SIRT1 overexpression in the skeletal muscles of mdx mice, an animal model of DMD, decreases serum creatine kinase (CK) levels, tissue fibrosis, and myofibril damage and increases their ability to run long distances compared with control mdx mice 8 . Resveratrol (3,5,4′-trihydroxy-trans-stilbene), a natural polyphenol found in grapes and red wine, has been shown to allosterically activate SIRT1 6,7 . δ-Sarcoglycan is a component of the dystrophin glycoprotein complex, the defect of which leads to severe cardiomyopathy in hamsters. We found that SIRT1 induces antioxidative superoxide dismutase 2 (SOD2), and oral administration of resveratrol to δ-sarcoglycan-deficient TO-2 hamsters increases cardiac SOD2 levels, attenuates the deterioration of cardiac function and hypertrophy, and significantly extends the lifespan 9 . Resveratrol administration also improves the pathophysiological manifestations of dystrophin-deficient mdx mice. Resveratrol decreases oxidative stress levels in muscles, reduces cardiac and muscular fibrosis, increases the expression levels of myosin heavy chains and troponins, and inhibits cardiac hypertrophy [10][11][12] . Moreover, long-term administration of resveratrol improves cardiac and muscular function with a decrease in serum CK levels to one-third of those in untreated mdx mice 11,12 . Thus, resveratrol could be a useful medicine for treating MD patients. However, to date, there have been no clinical studies investigating this possibility. Because clinical studies show the relatively low toxicity of resveratrol 13 , we planned a pilot phase IIa study in MD patients to investigate its biological activity.
In the present study, we administered resveratrol to patients with DMD, BMD, or FCMD for 24 weeks. Since the diseases had progressed considerably in these patients, motor function, muscular strength, and CK levels were evaluated as primary outcomes. Adverse effects and tolerability were also evaluated as secondary outcomes.

Results
Patients and protocol. The mean ages of eleven DMD, BMD, and FCMD patients were 21.0 ± 9.7 years (range, 12-39 years), 33.5 ± 8.3 years (range, 23-46 years), and 14.5 ± 2.5 years (range, 12-17 years), respectively, and the mean age of all participants was 24.4 ± 11.1 years (range, 12-46 years) ( Table 1). One FCMD patient was female, and the remaining ten participants were male. One DMD patient and three patients with BMD were ambulatory. The other seven patients were unable to walk and needed a wheelchair; therefore, they were assigned a grade of 9 on the Vignos scale 14 . Resveratrol was orally administered daily to all participants at an initial dose of 500 mg/day for 8 weeks (Fig. 1). The dose was then increased every 8 weeks to 1000 and finally to 1500 mg/ day, similar to dosages of resveratrol administered to healthy subjects and patients with cancer or metabolic diseases 13 . All patients were assessed for motor function, muscle strength, and blood examinations. Mean scores of muscle function measure (MFM), quantitative muscle testing (QMT), and serum CK levels of participants before resveratrol medication are shown in Table 2.
Motor function. The MFM scale is a sensitive and reliable method for evaluating motor function in neuromuscular diseases, including muscular dystrophy 15 . Patients' motor functions were evaluated by the MFM scale scores on 32 items in three dimensions. The mean MFM total score of all participants before resveratrol administration was 34.6 ± 29.3 (range, 2-85). MFM scores after 16 and 24 weeks of medication were 38.5 ± 31.7 (range, 2-90) (p = 0.012) and 38.4 ± 31.3 (range, 2-90) (p = 0.017), respectively (Fig. 2a). Thus, long-term administration of resveratrol significantly increased MFM scores by 10%. Sub-score analyses of D1, D2, and D3 showed increasing tendencies, but these did not meet the threshold of statistical significance ( Fig. 2b-d), suggesting that resveratrol does not improve muscle function specifically but instead affects the whole body. The Brooke and Vignos scale grades did not change during this study. www.nature.com/scientificreports/

Muscle strength.
To determine whether muscle strength was affected by resveratrol, we evaluated the muscle strength of each patient. Strength was monitored with quantitative muscle testing (QMT), which is a reliable method for evaluating the muscle strength of ambulatory and non-ambulatory MD patients [16][17][18] . Nine different QMT scores were measured, but in some patients, certain items could not be scored due to their severe muscle weakness or contracture. For example, a 39-year-old patient with DMD could be evaluated only for pinch strength. The monitored activities and number of patients assessed were as follows: scapula elevation, 10; shoulder abduction, 5; pinch, 11; elbow extension, 5; elbow flexion, 6; hip flexion, 5; knee extension, 6; ankle dorsiflexion, 6; and grip, 7 ( Table 2, Fig. 3, Supplementary Fig. S1). Scapula elevation was measured in ten participants, including three non-ambulatory patients. The administration of resveratrol for 24 weeks increased the mean QMT score of scapula elevation from 4.0 ± 2.7 kg-force (kgf) (range, 0.9-8.6 kgf) to 8.3 ± 6.9 kgf (range, 1.5-24.9 kgf) (p = 0.001) (Fig. 3a), while the mean shoulder abduction strength measured in one DMD patient and four patients with BMD increased from 4.4 ± 1.1 kgf (range, 3.0-6.2 kgf) to 9.1 ± 4.3 kgf (range, 3.9-15.1 kgf) (p = 0.003) (Fig. 3b). QMT scores of elbow extension, elbow flexion, hip flexion, knee extension, and ankle dorsiflexion in some patients were markedly improved by resveratrol, but the differences were not significant (Supplementary Fig. S1a-e). Resveratrol had no effect on pinch strength between the thumb and forefinger ( Fig. 3c) or on grip strength ( Supplementary Fig. S1f).      www.nature.com/scientificreports/ Adverse events. Grade 1 diarrhoea was noted in four patients (Table 3). Grade 2 diarrhoea was found in two patients administered 1500 mg/day resveratrol. These two patients needed to reduce the dosage from 1500 to 1000 mg/day. One patient resumed a dose of 1500 mg/day after several days, whereas the other patient (FCMD 12 y) was maintained at the dose of 1000 mg/day. Grade 1 abdominal pain was observed in three patients. Two cases of upper respiratory tract infection and one case of pneumonia were observed during the study. Among them, two patients were temporarily hospitalized but left the hospital soon afterwards. During hospitalization, they continued to take resveratrol on schedule. Transient alopecia, erythema multiforme, and acneiform rash were observed in one patient each. All participants continued to take resveratrol for 24 weeks. Resveratrol did not affect blood markers of liver and renal function (Supplementary Table S1).

Discussion
Resveratrol has been used in humans, and previous data show that its toxicity is limited 13 . Since resveratrol ameliorated pathology and improved the function of MD animal models, we planned this study as a pilot study. Resveratrol improved motor function and muscle power in the proximal muscles of participants. However, the distal muscles of patients were unaffected (Fig. 3c, Supplementary Fig. S1f). One patient with BMD (BMD 23 y) experienced relief from muscle pain during resveratrol administration, which may have contributed to the www.nature.com/scientificreports/ improvement in motor function. The observed increase in patients' muscle strength following resveratrol treatment was consistent with findings in mdx mice 10,12 , in which approximately fivefold increases in slow-fibre-type myosin heavy chain and troponin gene transcripts were detected in the biceps femoris muscles 10 . Resveratrol binds the N-terminal region of the SIRT1 protein, resulting in the allosteric activation of SIRT1 activity 7 . SIRT1 deacetylates and activates peroxisome proliferator-activated receptor-γ coactivator 1α (PGC-1α), which switches muscle types from fast to slow fibres 6,7,19 . Indeed, transgenic mice overexpressing PGC-1α 19 or SIRT1 8 show muscle type conversion from fast-to slow-twitch fibres. PGC-1α transgenic mice also show increased resistance to muscular fatigue 19 . Thus, the beneficial effects of resveratrol on the muscles of MD patients may involve PGC-1α activation by SIRT1. In mdx mice, serum CK levels were decreased by the overexpression of SIRT1 in the muscles 8 and by the long-term administration of resveratrol 12 . Recently, we have shown that skeletal muscle-specific SIRT1-knockout mice have a fragile muscle cell plasma membrane and that SIRT1 is involved in membrane resealing after cell injury 20 . In the present study, mean CK levels were decreased by resveratrol, although the change was non-significant. We noted that patients with low serum CK levels, i.e., less than 2000 IU/L, showed little response to resveratrol treatment, which could be caused by the lack of sufficient muscle mass. Resveratrol administration at early disease stages may afford great benefits to MD patients. Oxidative stress plays a key role in the development of MD pathology 21 . Resveratrol decreases ROS levels by inducing the expression levels of the antioxidant enzyme superoxide dismutase 9 , suppressing superoxidegenerating NADPH oxidase 10 , and eliminating damaged mitochondria, a major cellular source of ROS 22 , via mitochondrial autophagy, known as mitophagy 11,12,23,24 . These antioxidative functions of resveratrol are mediated at least in part by the deacetylation and activation of Forkhead box O transcription factors and PGC-1α via SIRT1 activation 6,7 . www.nature.com/scientificreports/ Cardiomyopathy and subsequent heart failure is a major cause of death in MD patients. Resveratrol suppresses cardiac hypertrophy and fibrosis in TO-2 hamsters 9 and mdx mice 25 , which may be mediated by the deacetylation and ubiquitin-dependent degradation of transcriptional coactivator p300 via SIRT1 activation 25 . In the present study, we could not detect any improvement in echocardiographic parameters or levels of plasma brain natriuretic peptide (BNP), a marker of cardiomyopathy. This may be because all participants showed normal BNP levels (< 18.4 pg/ml), and only 2 patients had a reduced ejection fraction of less than 50% (Supplementary  Table S1). Longer-term administration of resveratrol may be necessary to reveal the cardiac function of resveratrol in MD patients.
Resveratrol is rapidly absorbed and yields peak serum concentration at 1 h post dose 13 . In the present study, the levels of serum resveratrol concentration in MD patients (Fig. 5b-d) were consistent with those in healthy  www.nature.com/scientificreports/ volunteers 13 . Six (54.5%) and three (27.3%) patients had diarrhoea and abdominal pain, respectively (Table 3). These symptoms were transient and tolerable in participants except for the female patient (FCMD 12 y), who suffered from severe diarrhoea at the dose of 1500 mg resveratrol. The weight of the patient was only 26.5 kg, and she could not continue the dose. Similar to our study, Brown et al. has reported mild to moderate gastrointestinal symptoms in healthy volunteers who ingested resveratrol at doses of 2500 and 5000 mg/day 26 . Two patients contracted transient grade 3 upper respiratory and lung infections, and one patient suffered from a grade 2 upper respiratory infection. We believe that these events were unrelated to resveratrol since all three patients had repeatedly suffered from respiratory infections before the clinical trial and recovered immediately while taking resveratrol. We found that administration of resveratrol to mdx mice with the highest dose (4 g/kg food) is more effective in increasing myofibre cross-section area than lower doses of 0.4 g and 0.04 g/kg food 12 . However, a moderate dose of resveratrol (0.4 g/kg food) also decreases the number of fibres with central nuclei, the number of fine fibres, and serum creatine kinase levels. Additionally, it significantly improves motor function in mdx mice 12 . Resveratrol at 0.4 g/kg food was estimated to be approximately 50 mg/kg body weight/day 10 . The mean resveratrol level at 1000 mg/day was 36.9 ± 44.7 ng/mL (range, < 1.0-215.0 ng/mL) (Fig. 5a), which is consistent with the data of Almeida et al. 27 . The administration of resveratrol to healthy adult subjects at 150 mg six times/day for 13 doses resulted in a mean plasma concentration of 63.8 ng/mL and a half-life of resveratrol of 2-5 h 27 . Resveratrol at 36.9 ng/mL is equal to 160 nM. Because 160 nM resveratrol was 1/68 of the concentration needed for twofold activation of SIRT1 enzymatic activity in vitro 28 , the activation of SIRT1 by resveratrol in the present study did not seem strong. However, other studies show the physiological effects of resveratrol even in small amounts 13,27 . The administration of resveratrol at 150 mg/day for 30 days significantly decreased plasma glucose, insulin, triglycerides and leptin in healthy obese men 29 . Using muscular biopsy, Timmers et al. showed that resveratrol activates AMPK activity, increases SIRT1 and PGC1α levels, and promotes mitochondrial respiration activity in the muscle 29 . Based on these findings, we suggest that resveratrol at a dose of 25-50 mg/kg body weight/day is sufficient to treat MD patients. The dosing period required to detect an effect of resveratrol seems to be more than 4 months for MD patients.
Glucocorticoids have been shown to prolong the survival of patients with DMD 5 . In the present study, glucocorticoids were treated in only one patient with DMD (DMD 12 y), who showed increased muscle strength and motor function without any adverse effects during resveratrol administration. Recently, novel methods to treat DMD have been developed. Ataluren is a stop codon readthrough drug for DMD based on nonsense mutations 3 , and is currently being evaluated in a post-approval safety study in Europe. Eteplirsen is a morpholino antisense oligomer that induces the skipping of exon 51 of the dystrophin gene 4 , and accelerated approval for this drug was obtained from the U.S. Food and Drug Administration in 2016. The use of these kinds of therapies is expected to increase, although they have some limitations 30,31 . Given its unique mode of action and relatively mild adverse effects, resveratrol could combine with glucocorticoids or recent novel medicines such as ataluren and eteplirsen.
A limitation of the present study was that it was not a randomized, double-blind, placebo-controlled trial. We could not analyse an untreated group because the number of patients was limited, and three types of MD patients with various clinical conditions were recruited. In the case of rare diseases, phase I studies for cancer and phase IIa studies for other diseases have been executed without any control [32][33][34][35][36] . These trials were planned because of problems in recruiting patients, non-uniformity in patients' disease conditions, the use of an investigational drug that has never been administered in humans, no prior information on dose levels, difficulty in preparing a control group or difficulty in obtaining the agreement of patients to use a placebo. Usually, if some possibility concerning effectiveness of the drug is found in a phase IIa study, a phase IIb clinical study is carried out to identify the effect and dose of the investigational drug by a randomized and double-blind study.
We repetitively measured the MFM and QMT scores of the participants. Some studies have reported that exercises increase the motor function and muscle strength of MD patients 37 . For example, the knee flexor maximum voluntary contraction torque of the limb-girdle in Becker and facioscapulohumeral MD patients increased by 13% after a 12-week resistance training programme 38 . However, two meta-analyses of exercise interventions on muscle strength in MD patients showed no evidence of improvement in muscle strength and in endurance 39 . Because the average time required for MFM has been reported to be 36 min (range 8-75 min) 15 , measuring MFM and QMT scores once every 8 weeks in the present study was unlikely to improve motor function of the patients. Since participants are aware of exercise interventions, the two meta-analyses 39 also suggest that placebo effects of interventions are minimal on MD patients. Moreover, average muscle strength in DMD patients treated with placebo shows similar decline with that of untreated patients during administration for 6 months 40 . Because mean QMT scores of scapula elevation and shoulder abduction increased by twofold 24 weeks after resveratrol administration (Fig. 3a,b), it was less likely that the efficacy of resveratrol was a placebo effect.
As mentioned above, our study was a phase IIa study to investigate the possibility of the use of resveratrol in MD patients. To determine whether resveratrol could provide benefit for MD patients and be used as a symptomatic medication for MDs, a phase IIb study of resveratrol in a multi-centre trial with randomized, double-blind, and parallel-group comparison design will be necessary. An evaluation of the long-term effects of resveratrol on MD patients in a uniform cohort, such as age-matched young DMD patients, would be benefical.

Methods
Clinical study design. This