Muscular Dystrophy – PLOS Currents Muscular Dystrophy Wed, 17 Oct 2018 20:45:27 +0000 en-US hourly 1 Health Care Transition Experiences of Males with Childhood-onset Duchenne and Becker Muscular Dystrophy: Findings from the Muscular Dystrophy Surveillance Tracking and Research Network (MD STARnet) Health Care Transitions and Other Life Experiences Survey Tue, 21 Aug 2018 14:30:51 +0000 Introduction: As the proportion of males with Duchenne muscular dystrophy (DMD) surviving into adulthood increases, more information is needed regarding their health care transition planning, an essential process for adolescents and young adults with DMD. The objective of this study was to describe the health care transition experiences of a population of males living with Duchenne or Becker muscular dystrophy (DBMD).

Methods: The eligible participants, identified through the Muscular Dystrophy Surveillance Tracking and Research Network (MD STARnet) surveillance project, were 16–31 years old and lived in Arizona, Colorado, Georgia, Iowa, or western New York (n=258). The MD STARnet Health Care Transitions and Other Life Experiences Survey was conducted in 2013 and administered online or in a telephone interview. Sixty-five males (25%) completed the survey. Among non-ambulatory males, response differences were compared by age group. Statistical comparisons were conducted using Fisher’s exact test, or when appropriate, the Chisquare test.

Results: Twenty-one percent of non-ambulatory males aged 16–18 years, 28% of non-ambulatory males aged 19–23 years, 25% of non-ambulatory males aged 24–30 years, and 18 ambulatory males had a written transition plan. Nineteen percent of non-ambulatory males aged 24–30 years had delayed or gone without needed health care in the past 12 months. Among non-ambulatory males aged 24–30 years, 75% had cardiology providers and 69% had pulmonology providers involved in their care in the past 12 months. Twentyeight percent of non-ambulatory males aged 19–23 years and 25% of non-ambulatory males aged 24–30 years reported that they did not receive health care or other services at least once because they were unable to leave their home. Non-ambulatory males aged 16–18 years (29%) were less likely to have ever discussed how to obtain or keep health insurance as they get older compared to non-ambulatory males aged 24-30 years (69%) (p <0.01).

Discussion: This study identified potential barriers to the successful health care transition of males with DBMD. The results of this study may indicate a lack of targeted informational resources and education focused on supporting the transition of young men with DBMD as they age from adolescence into adulthood within the healthcare system. Future studies could determine the reasons for the potential barriers to health care and identify the optimal transition programs for males with DBMD. There are a few online resources on transition available to adolescents and young adults with special health care needs.



As a typical part of healthy development, individuals experience many different types of transitions across their lifespan. The transition from adolescence to adulthood is a process involving a variety of aspects and the health care transition is but one component. A successful health care transition should be a purposeful and systematic movement from a pediatric to an adult heath care system.1, 2 In recent years, emphasis has been placed on the health care transition experiences of youth with special health care needs. One of the objectives of Healthy People 2020 is to increase the percentage of youth with special health care needs whose health care provider has discussed transition planning with them.3

Duchenne muscular dystrophy (DMD) and Becker muscular dystrophy (BMD) are neuromuscular disorders caused by a mutation in the dystrophin gene on the X-chromosome, which leads to progressive muscle weakness.4, 5 Individuals diagnosed with DMD typically lose the ability to walk independently by age 13 years, and progress to cardiac and respiratory complications.4, 5 BMD is typically a milder phenotype with a variable progression.5, 6 In the recently published revisions to the Duchenne muscular dystrophy (DMD) care considerations, health care transition planning is emphasized.7

Positive health care transition experiences have been associated with better health outcomes. Satisfaction with transition of care has been associated with better social and emotional quality of life among adolescents with neuromuscular disorders.8 Healthcare transition programs have been associated with better disorder-specific health outcomes in young adults with diabetes9, 10 and those who have undergone renal transplantation11, 12.

The survival of individuals living with DMD has increased, with the median age of survival now in the mid-twenties13, 14, 15 and many individuals living into their third decade. It is essential for all adolescents living with Duchenne or Becker muscular dystrophy (DBMD) to have health care transition planning. In June of 2011, Parent Project Muscular Dystrophy, a nonprofit family-advocacy organization, convened an expert panel to discuss how to improve the transition experiences of young men with DMD.16 The major goal that emerged from this meeting was the need to change the perception that DMD is only a disease of childhood. A few qualitative studies have examined the transition experiences of young men with DMD living in England17, 18, 19, Canada20, 21, 22 and Japan23, however little information is available on the transition experience of adolescent and young men living with DMD in the United States. The objective of this manuscript is to describe the health care transition experiences of young males living with DBMD in certain areas of the United States who participated in the MD STARnet Health Care Transitions and Other Life Experiences Survey.


Starting in 2002, the Centers for Disease Control and Prevention funded a population-based surveillance system to determine the prevalence of childhood-onset DBMD and to collect information on clinical practices and health outcomes. An in-depth description of the MD STARnet methodology has been published.24 Beginning in 2004, MD STARnet retrospectively identified and longitudinally followed all individuals diagnosed with DBMD, who were born since January 1, 1982, who were symptomatic by age 21, and who resided in four US sites: Arizona, Colorado, Iowa, and 12 counties in western New York State. Georgia was added to the surveillance program in 2005 and Hawaii in 2008. Medical record data were abstracted from neuromuscular clinics, emergency departments, hospitals, and vital records. Annual medical record abstraction was conducted through December of 2011. For cases identified from September 2011 through December 2011, record abstraction was conducted through 2012.

A committee of clinical experts reviewed abstracted diagnostic data to assign each individual a case definition (definite, probable, possible, asymptomatic, affected female, or not DBMD).25 At the end of data collection, another group of experts reviewed mutation information from genetic testing, western-blot dystrophin levels from muscle biopsy, age when ambulation first ceased, steroid treatment, and age at onset of DBMD symptoms in order to classify males as Duchenne phenotype, Becker phenotype, or unable to determine.26

In 2013, from April through October, MD STARnet conducted a survey among males with DBMD in order to describe their transition experiences in health care, living arrangements, education and employment, and quality of life. The health care transition sections of the survey consisted of questions that assessed mobility and respiratory function, health care providers and access to health care facilities, health care utilization, health care coordination and management, health insurance, and barriers to care. A working group consisting of MD STARnet clinicians, researchers, and representatives from advocacy organizations developed the survey questions. Some survey questions regarding the health care transition process were taken from two national surveys: the National Survey of Children with Special Health Care Needs (NS-CSHCN)27, 28 and the Survey of Adult Transition and Health (SATH)29. Questions were also based on the 2012 version of the MDA transition survey30 and a survey of adolescents and young adults living with spina bifida31. Questions were piloted at neuromuscular clinics in Iowa and western New York among non-resident males with DBMD who were 16 years or older.

Eligibility criteria for the MD STARnet transition survey included being identified in the MD STARnet system by December 2011, living in the surveillance area at the time the survey was conducted, being at least 16 years old at the time of the survey, having an MD STARnet case definition of definite or probable, and being the oldest living male affected with DBMD in the household. The survey was not conducted in Hawaii because of incomplete case ascertainment. The survey was available in English and in Spanish. The transition survey was internet-based and was administered using the Qualtrics system housed at the University of Iowa

Ethical statement

In Colorado, Georgia, Iowa and western New York, public health authority permitted medical record abstraction for DBMD. In Arizona, institutional review board (IRB) approval was obtained from the University of Arizona, and when needed, from other health care facilities where data collection occurred.

IRB approval for the MD STARnet transition survey protocol was obtained at each site. Site differences in survey methodology were due to differences in site IRB requirements. Individuals who were eligible for the survey were sent a letter inviting them to participate. For males under age 18 years in Georgia, the letter was addressed to a parent or legal guardian and asked permission to allow the eligible male to participate. Invitations included an online survey link and login information. Participants were also given the option to request a telephone interview, by filling out their contact information and returning it in the postage-paid return envelope. A study coordinator then contacted the participant. Online participants (Arizona, Colorado, and Georgia) reviewed a consent page before completing the questionnaire; phone participants issued a verbal consent after a staff member read the online consent form verbatim. Implied consent, as indicated by completing the survey, was approved for the Iowa site.

Statistical analysis

Two analysts conducted statistical analyses independently at different sites using SAS versions 9.3 and 9.4 (Cary, North Carolina). Investigators used medical record data abstracted through December 2011 to compare eligible males who completed the survey to eligible males who did not complete the survey. Transition experiences were expected to differ not only by age, but also by ambulation status. Therefore, analysts grouped participants into one of two categories, those who used a wheelchair full-time (non-ambulatory) and those who either used a wheelchair part-time or who were fully ambulatory, according to their responses to the survey question on current mobility. One male who responded “does not apply” to the mobility question was included in the non-ambulatory group. Responses among the non-ambulatory group were compared for three different age groups: 16–18, 19–23, and 24–30 years. Given the small sample size, responses for the ambulatory group were not compared by age. Statistical comparisons were conducted using Fisher’s exact test, or when appropriate, the Chi-square test.


A total of 258 males affected with DBMD were eligible for the transition survey and 65 of them completed the survey (completion rate was 25%). Among sites, Iowa had the highest percentage of males who completed the survey (Table 1). Completion rate also differed by race/ethnicity; Hispanics and Non-Hispanic Blacks were significantly less likely to complete the survey. Completion rates did not differ significantly by age, Duchenne or Becker phenotype, use of respiratory devices by the end of 2011, or ambulation status at the end of 2011.


Table 1. Completion percent by study characteristics for the 258 males affected with Duchenne/Becker muscular dystrophy, aged 16–31 years, who were eligible for the 2013 MD STARnet Health Care Transitions and Other Life Experiences Survey

Among the 65 survey respondents, in 2013, 47 (72%) males indicated full-time current use of a wheelchair or scooter, 4 (6%) males indicated part-time use of a wheelchair or scooter, 13 (20%) indicated they walk without help. One male (2%) responded that the question does not apply and was added to the non-ambulatory group. Of the 48 males who were non-ambulatory, 43 (90%) were classified as Duchenne phenotype, one (2%) was classified as Becker phenotype, and for 4 (8%), phenotype could not be determined. Of the 17 males who were ambulatory, 14 (82%) were classified as Becker phenotype, 2 (12%) were classified as Duchenne phenotype, and for 1 (6%), phenotype could not be determined.

Table 2 displays results from questions regarding breathing and arm/hand ability. A higher percentage of non-ambulatory males aged 24-30 years (62%) used a ventilator with tracheostomy than males aged 19-23 years (6%) or 16-18 years (7%). While 44% non-ambulatory males aged 24-30 years reported that they could not use their hands, all of the non-ambulatory males aged 16-18 years and 19-23 years could use their hands.


Table 2. Breathing, arm, and hand abilities, by age and ambulation status, 2013 MD STARnet Health Care Transitions and Other Life Experiences Survey of males affected with Duchenne/Becker muscular dystrophy, aged 16-30 years

Table 3 displays results from questions on health care access and health care provider types. . A high percentage of males (79-84%) reported having a health care provider to call when they had a question, having a place to go when they are sick (63-79%), and having a place that they go for routine preventative care (56-88%). However, a low percentage of all male respondents (14-31%) had access to a satellite or outreach clinic. Seventy-five percent of non-ambulatory males aged 24-30 years had cardiology providers involved in their care in the last 12 months and 69% of non-ambulatory males age 24-30 years had pulmonology providers involved in their care in the last 12 months.


Table 3. Health care access and provider types by age and ambulatory status, 2013 MD STARnet Health Care Transitions and Other Life Experiences Survey of males affected with Duchenne/Becker muscular dystrophy, aged 16-30 years

Table 4 displays results from questions on the type of patients a provider treats and health care provider changes. Non-ambulatory adolescents were more likely to see a primary care/family medicine specialist, cardiologist, or pulmonologist who sees children only, while non-ambulatory males aged 24-30 years were more likely to see these specialist who saw adults only. A high percentage of non-ambulatory males aged 24-30 years (69%) and age 19-23 years (67%) reported ever changing a doctor or health care provider because the doctor was only treating children.


Table 4. Types of patients current provider treats and healthcare changes by age and ambulation status, 2013 MD STARnet Health Care Transitions and Other Life Experiences Survey of males affected with Duchenne/Becker muscular dystrophy, aged 16-30 years

Table 5 displays results from questions on discussions about transition health care needs with health care providers and care coordination. Among non-ambulatory males aged 16-18 years, 57% reported talking with their health care providers about eventually seeing different doctors. A high percentage (81-83%) of non-ambulatory males aged 19-30 reported they had help arranging or coordinating their care among different doctors or services. Parents and other family members (63-100%) were the ones who most frequently helped coordinate care, compared to the low percentage of primary care offices (13-22%) and neuromuscular offices (13-44%).


Table 5. Health care provider discussions and care coordination by age and ambulation status, 2013 MD STARnet Health Care Transitions and Other Life Experiences Survey of males affected with Duchenne/Becker muscular dystrophy, aged 16-30 years

Table 6 displays results from questions on health care transition planning, insurance, and barriers to care. Among all males, a low percentage (18-28%) reported having a written summary to assist in the transition from a pediatric to an adult health care provider. Among the 15 males who had a written transition plan, 5 (33%) indicated their parent helped develop the plan, while 9 (60%) indicated their pediatrician’s office, family medicine office, or neuromuscular office helped develop the plan; only one male helped develop his plan. Among non-ambulatory males, males aged 16-18 were less likely to have ever had discussions regarding health insurance as compared to older males. The oldest non-ambulatory males were also more likely to report having delayed or gone without needed health care in the past 12 months as compared to the youngest non-ambulatory males (19% vs. 0%).


Table 6. Health care transition plan, insurance, and barriers to care by age and ambulation status, 2013 MD Health Care Transitions and Other Life Experiences Survey of males affected with Duchenne/Becker muscular dystrophy, aged 16-30 years


The American Academy of Pediatrics, American Academy of Family Physicians, and the American College of Physicians have recommended that formal written transition plans be initiated for all adolescents starting at the age of 14 years.32 In this study only 1 in 4 males reported having a written summary to assist in the transition from a pediatric health care provider to an adult health care provider. Approximately 2 out of every 3 males aged 19-30 years reported ever changing a health care provider because their doctor only provided care for children, indicating that the majority of young men with DMD are making a health care transition to adult providers. The lack of written summary may be due to a lack of care coordination services. In this study, approximately 1 in 10 males had a primary care office and fewer than 1 in 3 males had a neuromuscular office help coordinate their care. In a recent national survey of pediatricians, those patients who had a care coordinator were more likely to develop written transition plans.33 Care coordination has also been associated with a lower likelihood of having an unmet specialty care need 34 and reduced functional disabilities 35 among children with special health care needs.

A higher percentage of males in our study (71%-89%) reported ever discussing how their health care needs might change as they get older with a health care provider, as compared to the 2007 SATH 29 (55%) and the 2009-2010 NS-CSHCN28 (59%) national surveys. Heath care providers of males with DBMD may be more likely to discuss future health care needs than health care providers of patients with other health care conditions.

In our study, 6 of 10 non-ambulatory males reported having ever discussed eventually seeing adult doctors with their pediatric health care providers, which was higher than the NSCSHCN (4 of 10 )28 However, for males who were ambulatory in our study, 1 in 4 reported having discussed eventually seeing adult providers with their pediatric health care providers.

In our study, 56-69% of the non-ambulatory males aged 24–30 years reported ever having a discussion with health care providers regarding health insurance as compared to only 29% of non-ambulatory males aged 16–18 years. The percent of participants who reported discussing health insurance with their providers was 35% in the NS-CSHCN28 and 53% in the SATH29. It is important that health insurance discussions occur before a transfer to adult health services. Approximately half of young adults with disabilities and those with special health care needs have been shown to experience gaps with insurance coverage.36, 37, 38 Young adults with no health insurance are more likely to experience a delayed transition of care than young adults with health insurance.39 Young adults with special health care needs who are uninsured have been shown to delay or forgo care and to have problems getting care.40 In our study, 4 in 10 non-ambulatory males aged 24-30 reported changing health insurance coverage in the past 12 months as compared to none of the non-ambulatory males aged 16-18 years. The reasons for the high proportion of non-ambulatory males aged 24–30 years who changed their health insurance coverage are unknown; possible explanations include men changing from their parents’ insurance to their own insurance, or qualifying for Medicare or Medicaid.

Lack of health care access may have been an issue for some of the non-ambulatory males aged 24–30 years in our study, with 2 of 10 males delaying or going without needed health care in the past 12 months. One reason for this barrier to care may be a lack of transportation. One in 4 males aged 19–30 years did not receive care because they were unable to leave the home.

The recent increase in survival among males with DMD has been attributed to advances in cardio-pulmonary therapies, including non-invasive ventilation and cough assistance.41, 42, 43, 44, 45 Among the non-ambulatory males aged 24–30 years, several reported not having a cardiologist or pulmonologist involved in their care in the past 12 months. A transition study of males with DMD living in England obtained similar results: 47% of parents reported their son had contact with a cardiac clinic and 39% of parents reported their son had contact with a respiratory clinic in the last six months.18 The 2010 Duchenne muscular dystrophy care considerations state that non-ambulatory males with DMD should have pulmonary assessments at least every 6 months and males with DMD over the age of 10 years should have cardiac assessments at least every year.46

Our study has several strengths. It is one of the largest studies in the United States to describe the transition experiences of males living with DBMD. Eligible participants were drawn from a surveillance system that identifies all individuals living with DBMD within specific geographic areas. Our study surveyed the males themselves, while many studies that examine the transition experiences among adolescents and young adults rely on responses from caregivers.

Our study also has some limitations. Whereas our study population was drawn from a population-based surveillance system, we had a survey completion rate of 25% and a small sample size. Males who completed the survey differed from males who did not complete the survey by both state of residence and race/ethnicity, and therefore the experiences of the participants may not reflect the experiences of all males with DBMD in the respective surveillance areas. Also, the results from this survey may not be generalized to all males living with DBMD in the United States. While no differences in functional status were observed at the end of the surveillance period between males who completed the survey and males who did not, it is possible that, differences in functional status may have existed two years later at the time of the transition survey. In our survey, we did not ask the males if a caregiver helped them complete the survey, which may have biased the responses.

Our study found that many adolescent and young adult males living with DBMD lack a written transition plan, and among those who did have a plan, only one male reported participating in its development. The American Academy of Pediatrics, American Academy of Family Physicians, and American College of Physicians-American Society of Internal Medicine recommends that health care providers, parents, and adolescents/young adults jointly participate in the development of transition plans.1, 32 The results of this study may indicate a lack of targeted informational resources and education focused on supporting the transition of young men with DBMD as they age from adolescence into adulthood within the healthcare system. Recently, updates to the 2010 DMD care considerations were published, including a new section devoted to healthcare transition planning.7 Future studies could try to determine the reasons for the potential barriers to health care described in this study as well as identify the optimal transition programs for adolescent and young adult males living with DBMD. Online resources are available to adolescents and young adults with special health care needs, their families, and health care providers, including Got Transition, a program of the national Alliance to Advance Adolescent Health which is supported through a cooperative agreement with the Maternal and Child Health Bureau of the Health Resources and Services Administration.47 There is also an online resource specific to adolescents and young adults with neuromuscular disorders, the Young Adult Programs from the MDA.30

Corresponding Author

Pangaja Paramsothy

Mailing Address: Division of Human Development and Disability, National Center on Birth Defects and Developmental Disabilities, Centers for Disease Control and Prevention, MS E88, 4770 Buford Hwy., Atlanta, GA 30341


Data Availability

Due to privacy concerns (detailed personal information was obtained from a small number of individuals living in a defined surveillance area), data from MD STARnet Health Care Transitions and Other Life Experiences Survey is not publicly available. We are willing to work with individuals who would like to verify the results.

Data from this analysis is kept at the Centers for Disease Control and Prevention. Researchers interested in utilizing MD STARnet data should contact: Natalie Street at the Division of Human Development and Disability, National Center on Birth Defects and Developmental Disabilities, Centers for Disease Control and Prevention, Mailing Address: Division of Human Development and Disability, National Center on Birth Defects and Developmental Disabilities, Centers for Disease Control and Prevention, MS E88, 4770 Buford Hwy., Atlanta, GA 30341. Email:

Competing Interests

The authors have declared that no competing interests exist.

]]> 0
A Pilot Survey Study of Adherence to Care Considerations for Duchenne Muscular Dystrophy Fri, 11 May 2018 10:09:31 +0000 Introduction Care Considerations supported by the Centers for Disease Control and Prevention for the management of Duchenne muscular dystrophy were published in 2010, but there has been limited study of implementation in the United States. Methods A questionnaire collecting information about standard care practices and perceived barriers was piloted by 9 clinic directors of facilities within the Muscular Dystrophy Surveillance, Tracking and Research network. Results Six clinic directors completed the questionnaire; 1 adult-only clinic was excluded. Over 80% adherence was found for 30 of 55 recommendations examined. Greatest variability was for initiation of corticosteroids, bone health monitoring, type of pulmonary function testing, and psychosocial management. Barriers included unclear guidelines, inadequate time and funding, family-specific barriers and lack of empirical support for some recommendations. Discussion This pilot study showed implementation of the 2010 Care Considerations, except for recommendations based largely on expert consensus. Complete adherence requires more studies and active promotion.


A pilot survey study of adherence to care considerations for Duchenne muscular dystrophy


Duchenne muscular dystrophy (DMD) is an X-linked genetic disorder of the dystrophin (DMD) gene and is well known as a complex, disabling disorder with shortened life expectancy. Advances in the clinical management of DMD have resulted in prolonged survival.1, 2, 3 In order to improve the consistency of treatment among providers, consensus-based Care Considerations for comprehensive, multi-disciplinary clinical management of patients with DMD were developed by an international panel of experts convened and supported by the C3enters for Disease Control and Prevention (CDC).4, 5

Survey studies on the provision of care to patients with DMD have been conducted using questionnaire responses from selected medical specialties6, 7 and patients or care providers.8, 9 In 2013, the CDC funded a pilot study to evaluate implementation of the 2010 Duchenne care recommendations in the United States. A questionnaire was designed to collect information about standard DMD care practices in clinics specializing in the treatment of children with DMD and the perceived barriers to provision of recommended care. The questionnaire was sent to clinic directors of Muscular Dystrophy Association (MDA) supported facilities located within the Muscular Dystrophy Surveillance, Tracking, and Research network (MD STARnet), a population-based surveillance program established in 2004 by the CDC.10, 11 In this paper, we report the findings from the pilot study, compare reported clinical practices to those delineated in the Care Considerations, and describe barriers to care as perceived by the directors.


Questionnaire Development

The pilot questionnaire was developed from existing data items in several programs, including the Canadian Paediatric Neuromuscular Physicians Survey,6 the Muscular Dystrophy Association (MDA) DMD registry items, Parent Project Muscular Dystrophy (PPMD) association’s metrics for certified Duchenne care clinics,12 and the Care-NMD Patient Survey ( After compiling existing data items, the research team developed additional questions or added response categories not already included. In addition, a literature review of challenges to translating clinical guidelines into practice was conducted to identify potential barriers that may influence implementation.13, 14 Based on the review, specific questions regarding perceived provider, patient/family and system barriers to implementation were added to the questionnaire.

The questionnaire was reviewed by representatives of two patient support organizations (JW for MDA and KK for PPMD) for content, breadth, and interpretability. It was piloted by a neuromuscular physician and a nurse practitioner who were involved in DMD care but not associated with the research project. The final questionnaire consisted of 30 multiple-choice questions and four open-ended questions asking about opinions regarding future recommendations, areas of disagreement with current recommendations, and suggestions to improve implementation.

For reference, the questionnaire can be found at the following link:

Questionnaire Implementation

A paper questionnaire was mailed to nine directors of MDA supported clinics located in participating MD STARnet sites (Arizona, Colorado, Iowa, western New York) in the fall of 2014. The questionnaire was expected to take 30 minutes to complete. Reminder emails and letters were sent at two and four weeks after the initial email. Responses were entered into a Research Electronic Data Capture (REDCap) database, a secure web application for managing online questionnaires and databases, housed at the University of Iowa. IRB approval was received by the New York State Department of Health Institutional Review Board (IRB number: 03-062/15-0643/15-0644).

Statistical Analyses

Percent adherence was calculated by combining response categories that met the minimum recommendation and dividing by the total number of responding clinic directors. SAS® software, Version 9.4 was used for analyses [Copyright (c) 2002-2012 by SAS Institute Inc., Cary, NC, USA.]. Responses to individual questions are available from the first author.


Questionnaires were completed by six clinic directors: two from pediatric clinics, three from mixed pediatric and adult clinics, and one adult only clinic. Questionnaire responses from the adult only clinic director were excluded since the responses pertained only to care of adult patients and the focus of our study was the care provided from the time of diagnosis. Percentages for adherence across all remaining clinics are presented in Table 1. Genetic testing and counseling services were reported as provided by all clinics (Table 1 – Diagnostic Assessment). Access to a multi-disciplinary team of medical providers was reported as available by all but one clinic director (Table 1 – Disease Management Team; for a list of specialties, see Clinic Director Questionnaire in Supplementary Materials). Specialties reported as not available by at least one clinic included: pulmonology; gastroenterology; psychiatry or psychology; physical, occupational or speech therapy; and palliative care clinic or services (data not shown). All clinic directors reported completing at least one recommended assessment every 6 months with variability across the types of assessments completed (Table 1 – Clinical Assessment). Recommended clinical management also varied; all clinic directors recommended stretching, bracing for positioning or standing, adaptive/assistive equipment and wheelchairs, but fewer recommended wrist/hand splints, bracing for walking and communication devices.

Table 1 picture

Fig. 1: Table 1. Clinic director reports of care provided and adherence to recommended care.

All clinic directors reported practice patterns consistent with the recommendations, including continued use of corticosteroids unless contraindicated (Table 1 – Corticosteroid Management and Side Effect Monitoring), but showed some variability in recommended age at initiation. Monitoring of side effects occurred at the recommended frequency. Variable adherence was found for the assessment of bone health (Table 1 – Bone Health Assessment and Management). Of those who assessed bone health, testing was completed at baseline and annually for high-risk patients (data not shown). Conversely, all directors reported offering bone health interventions consistent with recommended care (vitamin D, bisphosphonates) and consultations with endocrinologists to assess and treat osteopenia/osteoporosis. All clinic directors reported the use of radiographs to monitor scoliosis with variability in the indications for completing (Table 1 – Orthopedic Assessment).

Directors from all clinics reported recommending the use of electrocardiograms and echocardiograms for cardiac monitoring (Table 1 – Cardiac Assessment and Management); cardiac MRIs were not reported by any directors (data not shown). Adherence to frequency of monitoring was variable for patients under the age of 10, but consistent with the Care Considerations for those 10 and over. Holter monitoring was reported as recommended only if the child was symptomatic by all directors (data not shown). Medications were reported as used as treatment of cardiac dysfunction by all clinic directors. ACE inhibitor-angiotensin-converting-enzyme inhibitor was reported as the first medication recommended by all directors, but timing of use varied. Although all clinic directors reported referrals to a cardiac specialist, the timing of referrals varied (at diagnosis, at age 10, or abnormal cardiac test) (data not shown).

Annual monitoring of respiratory function of ambulatory boys ages 6 years and older was reported as recommended by all clinic directors (Table 1 – Pulmonary Assessment and Management). For non-ambulatory boys without respiratory symptoms or use of assisted ventilation, forced vital capacity and peak cough flow measures were reported as occurring at the recommended frequency by the majority (≥3) directors (Table 1). Of those not meeting the recommended frequency, pulmonary testing was reported as not being measured or tested annually (data not shown). The remaining pulmonary tests were reported less frequently as having been measured. Sleep studies were recommended with signs and symptoms and screening for awake or sleep hypoventilation every 6 months by all directors (data not shown). For the management of pulmonary dysfunction, cough assist and non-invasive ventilation were recommended by all clinic directors under multiple respiratory conditions (Table 1). Referrals to pulmonary specialists were not reported as recommended by all clinic directors, and the reasons for referral varied.

All clinic directors reported regular monitoring of gastrointestinal and urinary morbidities (Table 1 – Gastrointestinal and Urinary Assessments) and consultation with a nutritionist. The timing of the consultation typically occurred at initiation of steroids, after excessive weight gain, onset of chronic constipation, or when there was evidence of dysphagia (data not shown).

Only one clinic reported not completing a formal neuropsychological assessment before entering school (Table 1 – Neurocognitive Assessment); the remaining clinic directors reported either always completing an evaluation or completing if clinically indicated (data not shown). Evaluating for problems with coping, speech and language, learning disability, neurobehavioral disorders, emotional problems, or signs of social isolation at each visit were reported by most clinic directors. Psychotherapy, pharmacological and social interventions were infrequently reported as recommended.

Barriers to Implementation of DMD Care Considerations

Across all disease management areas, clinic directors agreed that not all health providers are familiar with each Care Consideration (data not shown). Specifically, primary care and emergency/urgent care providers were reported as having less familiarity or understanding of the Care Considerations. Regarding system barriers, clinic directors agreed there is inadequate time at clinic visits to complete all recommendations, inadequate funding to support ancillary staff and inadequate coordination of care across disciplines and services. Of the potential patient or family barriers queried, all clinic directors agreed that distance to neuromuscular centers, out-of-pocket expenses and access to services/resources in the patient’s community were significant barriers to implementation of care. There was little consistency between clinic directors on whether there is a lack of awareness or familiarity with the content of the Care Considerations by patients and families.


Directors of selected neuromuscular clinics located in the MD STARnet surveillance areas completed a pilot questionnaire describing their recommended clinical management of patients diagnosed with DMD. Their responses were compared to the Care Considerations published in 2010 to evaluate implementation across multiple disease management areas. Each director also provided a perspective about potential provider, system, and patient barriers to implementation of the Care Considerations. Overall, reported clinical practices related to the diagnostic process, access to a multi-disciplinary medical team, timeliness of follow-up by a medical professional specialized in the care of neuromuscular disorders, and recommendation of corticosteroids to all patients and monitoring their side effects were consistent with the Care Considerations and the recently released imperatives for DUCHENNE MD15 and the Transforming Duchenne Care Initiative.12 Rehabilitation interventions, such as stretching, orthotics for positioning, and standing devices were reported as recommended at various disease stages by all directors. In addition, implementation of cardiac monitoring and management showed high agreement with the Care Considerations for the use of echocardiograms in terms of initiation of studies, long-term follow-up studies, and referrals to a cardiac specialist. Finally, monitoring of orthopedic issues, specifically scoliosis, was consistent with recommended care.

There were several disease management areas for which broad adherence was observed, but variability in patterns of implementation were also evident. For example, although the frequency of respiratory function measurements for forced vital capacity and peak cough flow were consistent with the Care Considerations, other recommended measurements (e.g., blood gases, sleep studies, oxyhemoglobin saturation by pulse oximetry) were either not performed or only completed with signs and symptoms. A survey study of pediatric respirologists and Canadian Paediatric Neuromuscular Group (CPNG) members about respiratory consultations, frequency and reasons for assessing pulmonary function, and interventions showed similar findings.7 Responses were compared between provider types and showed differences in preferred tests for assessing pulmonary function. Like our data, the pediatric NMDs preferred cough peak flow, whereas pediatric respirologists preferred maximal inspiratory and expiratory pressures. Similarly, the use of non-invasive pulmonary interventions, such as cough assist and positive pressure, were reported as recommended by all directors in our study, but the initiation of use occurred under variable clinical signs. This variability may reflect, in part, the lack of empirical evidence for the optimal conditions under which non-invasive ventilation should be started.7, 16, 17 Management of bone health was another area that showed adherence with Care Considerations, but also demonstrated variability. The reported management of bone demineralization was consistent with recommendations, but the timing and frequency of monitoring bone health was not. Researchers have continued studying the use of DEXA scans to monitor bone health, which will inform on the accuracy with which it can be assessed and appropriate methods to do so in this patient population.18, 19

Research evaluating provider reports of adherence to the Care Considerations is limited. Prior to the release of the Care Considerations, but after the release of guidelines by the American Academy of Neurology, American Thoracic Society, and American Academy of Pediatrics, McMillan et al6 reported on findings from a questionnaire distributed to physicians who were members of the Canadian Pediatric Neuromuscular Group. Comparable to our findings, the reported clinical practice was largely consistent with recommendations from the existing guidelines. For those disease management areas that showed variability in adherence, the types of studies and levels of evidence associated with the recommendations were most often categorized as expert opinion with inadequate or conflicting directives given current knowledge (e.g., DEXA scans, respiratory assessments and interventions, cardiac assessments).6 Further, low rates of adherence for some interventions (e.g., pulmonary) in Canada were attributed to lack of coverage by insurance and balancing direct cost to families in the absence of efficacy data.7 Finally, the clinical management of neurocognitive and psychosocial issues showed poor adherence despite the high rates of these issues among patients diagnosed with DMD.20, 21, 22

Delays in clinical uptake of recommendations may be affected by multiple factors.13 Given the clinical complexity of DMD and required multidisciplinary interventions, barriers to comprehensive implementation of the recommended care practices may be intensified due to the need to disseminate and translate into practice across multiple disciplines. Further, barriers may differ by disease management areas. From our questions about barriers to care, most clinic directors reported that not all providers who manage patients with DMD are familiar with all areas of the Care Considerations, in particular, those providing primary care or treating DMD patients in the emergency room. Institutional limitations were found to be significant behavioral barriers due to inadequate space, time, and funding. Finally, access to treating facilities, community services/resources and out-of-pocket expenses were all viewed as significant barriers to provision of recommended care.

There are several limitations to the findings from this pilot questionnaire. This study relied on a non-random selection of clinic directors from medical facilities within the MD STARnet surveillance region. Because of the small sample size, we did not examine variability within clinics nor did we examine barriers as potential reasons for observed variability. Specialist practices were summarized by clinic directors and not provided by individual specialists who provided the care. Despite these limitations, this pilot study provides information on the clinical management of DMD patients several years following the release of the Care Considerations, but prior to the release of the updated considerations. In addition to reporting on adherence to specific recommendations, we described patterns of care as provided. This information is important for aligning current practice with further refinement of clinical recommendations. Finally, the questionnaire asked clinic directors about perceived barriers to implementing the Care Considerations, which may provide guidance on potential revisions to recommendations, promote scientific investigations to support the recommendations, and increase visibility of institutional challenges to providing the recommended care.


This study showed adherence to many of the assessments and interventions suggested in the Care Considerations. The areas showing less consistency could be classified as derived from expert opinion with little supporting data. Improvement in the implementation of all aspects of recommended care requires continued study into the health impact of the proposed recommendations and continued advocacy for coverage by agencies.

Corresponding Author

Kristin Conway, PhD, Department of Epidemiology, College of Public Health, The University of Iowa, Iowa City, Iowa 52240, USA; Telephone: 319-335-4641 Email:

Data Availability Statement

Data is available upon request.

Deborah Fox, MPH, Director, Congenital Malformations Registry Chief, Birth Defects Registry and Surveillance Section (BDRS) New York State Department of Health, Empire State Plaza, Corning Tower, Rm 1203 Albany, NY 12237, Phone 518-402-7950, Fax 518-402-7959,,

Ethics Statement

IRB approval was received by the New York State Department of Health Institutional Review Board (IRB number: 03-062/15-0643/15-0644). No identifiers were collected in the questionnaire. The New York State Department of Health de-identified the questionnaires and assigned random IDs for data entry by the Iowa collaborators. Participants were informed of the purpose of study, which is to inform on clinical practice and future care considerations, but posting of raw data was not disclosed.

Competing Interest Statement

The authors have declared that no competing interests exist.

]]> 0
Are Soy Products Effective in DMD? Tue, 27 Mar 2018 18:49:50 +0000 INTRODUCTION: 

In addition to their nutritional value, processed soy bean extracts contain several activities with potential therapeutic benefits. These include anti-oxidants, and tyrosine kinase and protease inhibitory activity. There are also anecdotal reports of health benefits of soy products in alleviating DMD symptoms.


Mdx mice were fed a control soy-free diet or the same diet containing either a proprietary soy preparation (Haelan 951), purified soy isoflavones, purified Bowman-Birk protease inhibitor or a combination of isoflavones and Bowman-Birk inhibitor. Mice were tested for their wire hanging ability at the start of the diet regimen and every 4 weeks until week 12 of treatment.


The diet containing Bowman-Birk inhibitor was the only one to show a significant and sustained improvement over the 12 weeks of the study. All other dietary additions; Haelan 951, isoflavones and isoflavones with Bowman-Birk inhibitor, were not significantly different from each other or from control. The effectiveness of Bowman-Birk inhibitor in mdx mice clearly warrants further study.



The health benefits of soy products have been recognised for centuries, with anecdotal and scientific evidence of beneficial effects in a wide range of disorders and conditions from diabetes and cancer to heart disease and inflammation1. Unprocessed soy beans contain high levels of phytic acid, an anti-nutrient that sequesters vitamins and minerals, and perturbs inositol lipid metabolism, which make it unsuitable for human consumption in large quantities. Soy is therefore most often consumed in fermented form, such as the familiar soy sauce, miso and tempeh or by processing a protein and fat extract of the beans, as in tofu or textured soy protein.

Nutraceutically soy is suggested to have various properties, but the precise chemical basis of many of these activities is poorly defined. The most recognised activities are those of the Bowman Birk inhibitor (BBI) and the isoflavones such as genistein. BBI is an 8 kDa peptide which is a non-competitive dual inhibitor of trypsin and chymotrypsin, it is acid and heat stable and can not only survive passage through the stomach but also passes the gut wall intact and can therefore have systemic effects2. BBI is capable of inhibiting many different cellular proteases, though it is thought that its main cellular activity is through inhibition of the chymotrypsin like activity of the proteasome3. Genistein, daidzein and glycitein are the principal isoflavones of soy, with well recognised and scientifically tested antioxidant, phytoestrogen and kinase inhibitory properties4, 5 Genistein is available cheaply and taken widely as a health supplement.

Amongst the DMD community there is significant use of a fermented soy product, Haelan 951, a commercially available health supplement In addition to its nutritional value – high in amino acids, essential fatty acids and vitamins, and low phytic acid content, it contains significant amounts of all 3 isoflavones and BBI. The benefits of Haelan 951 in DMD are entirely anecdotal, with no scientific study being conducted into its use, with only 2 published papers concerning its potential benefits. A case report of a single cancer patient6, and a study which suggested some effect in reducing the growth of pancreatic cancer cells in vitro7.

On the other hand, there is an abundance of published literature investigating the benefits of genistein 4, and BBI8, 9 usually given in the form of a Bowman Birk Inhibitor concentrate (BBIC), though only a handful with direct relevance to DMD. In one study in mdx mice BBIC was revealed to provide some benefit to mdx pathophysiology10 including benefits to muscle function. While investigation into the antioxidant properties of purified genistein did reveal some modest effects on muscle pathology and biochemical status in mdx mice, the mechanism of action was not elucidated precisely11, 12.


All mdx mice were fed standard mouse chow from weaning. At 12 weeks of age male mdx mice were assigned randomly to treatment groups (n=8 per treatment) and fed the test diets ad lib for a further 12 weeks. Diets comprised (i) the control AIN-93G diet alone or AIN-93G containing (ii) Haelan 951 (25mg/g;, or (iii) Bowman Birk Inhibitor (0.75mg/g; 95% pure T9777 Sigma Chemical Company), or (iv) the isoflavones genistein, daidzein and glycitein in the ratio 23:9:3 (17.5µg/g; LC Labs) or (v) a combination of BBI and the isoflavones. The amount of diet eaten and the mice were weighed every week from commencement of treatment and mice were subjected to wire hang test13 at 0, 4, 8 and 12 weeks of treatment.

Ethics Statement

Mdx mice were housed in the University of Sheffield animal facility according to national and international best practice guidelines for research using animals. All procedures were approved by the University of Sheffield Animal Welfare Committee and carried out under UK Home Office Project License (PPL 60/4453, Animals in Scientific Procedures Act, 1986).

Results and Discussion

There were no significant differences in diet consumption or in mouse weight gain between the control and experimental diets over the course of the 12 weeks of treatment, nor were there any observable adverse effects on the mice. Comparison of wire hang times over the 12 weeks of treatment using the maximum holding impulse (weight x time)13 relative to the start of treatment (100%) as a measure revealed a slight decline in wire hang times in animals on the control diet (Figure 1). Isoflavones and isoflavones with BBI showed a slight and sustained increase in wire hang times over the course of the treatment, but this was not significant from control, whereas Haelan 951was essentially the same as control except at 8 weeks where was a slight but non-significant increase (Figure 1). BBI alone however showed a sustained and significant increase in wire hang times at 4, 8 and 12 weeks of treatment reaching a peak at 8 weeks, followed by decline at 12 weeks (Figure). The reason for the decline is not clear from the wire hanging data alone. Given that the dose of BBI administered was based on a calculation of the theoretical amount of actual BBI administered in previous studies using BBIC10 the amount of pure BBI and its relative activity and absorption in its pure form, may have resulted in a different pharmacokinetic profile in the tissue leading to some sort of cytotoxic effect at this dose at longer time points. In order to investigate this a follow on dose ranging study would be required, along with pharmacokinetic analysis of the amounts of BBI actually present in the serum and tissues of the treated mice. Comparing the effect of the 4 diets to control at the 12 week treatment time point (Figure 2), demonstrates the spread of the data within in each treatment group but again shows the significant improvement obtained with the diet containing BBI only, whereas Halean 951, the isoflavones and isoflavones with BBI showed no significant difference as compared to control.

The diets tested were: (i) control AIN 93G diet alone, or 4 test diets of AIN 93G containing (ii) Haelan 951 (Haelan; 25mg/g), or (iii) Bowman Birk Inhibitor (BBI; 0.75mg/g), or (iv) isoflavones (17.5µg/g), or (v) a combination of BBI and the isoflavones (BBI+Isoflavones). Data plotted are mean from groups of 7-8 mice for the % change in maximum holding impulse (HI) for each individual mouse (longest wire hang in seconds/mouse mass in grams) taking the holding impulse at day 0 of treatment as 100%. One way repeated measures ANOVA were performed using GraphPad Prism software. * p<0.05, error bars have been omitted for clarity. Full data are avialble in Appendix 1.

Fig. 1: Longitudinal analysis of wire hanging ability in groups of 6-8 mice fed the indicated diets.

The diets tested were: (i) control AIN 93G diet alone, or 4 test diets of AIN 93G containing (ii) Haelan 951 (Haelan; 25mg/g), or (iii) Bowman Birk Inhibitor (BBI; 0.75mg/g), or (iv) isoflavones (17.5µg/g), or (v) a combination of BBI and the isoflavones (BBI+Isoflavones). Data plotted are mean from groups of 7-8 mice for the % change in maximum holding impulse (HI) for each individual mouse (longest wire hang in seconds/mouse mass in grams) taking the holding impulse at day 0 of treatment as 100%. One way repeated measures ANOVA were performed using GraphPad Prism software. * p<0.05, error bars have been omitted for clarity. Full data are available in the Data Availability Statement.

Box whisper plot representation of the data in Figure 1 at week 12 of treatment. All other parameters as for Figure 1. One way ANOVA and Kruskal Wallis test were performed using GraphPad Prism software. * p<0.05.

Fig. 2: Box whisper plot representation of the data in Figure 1 at week 12 of treatment.

Box whisper plot representation of the data in Figure 1 at week 12 of treatment. All other parameters as for Figure 1. One way ANOVA and Kruskal Wallis test were performed using GraphPad Prism software. * p<0.05.

A previous study examining the effects of 3 months of feeding BBIC on mdx mouse muscle pathophysiology demonstrated a significant improvement in several parameters of muscle function10. There was an approximately 25% and statistically significant improvement in EDL muscle mass, cross sectional area, tetanic force and force drop on repeated eccentric contraction. Total mouse mass was 10% higher (but not significant), relative EDL muscle mass was significantly higher, and whilst twitch force was also 25% higher the difference was not significant. There was no change in specific tension. Our findings therefore that purified BBI fed to mice at an equivalent dose to that administered in the Morris study10 resulted in a significant increase in wire hang times, is in line with their findings. However what is more difficult to rationalise is why the combination of BBI and isoflavones did not produce a significant improvement in wire hang times.

Both BBIC and Halean 951 contain significant amounts of BBI and isoflavones. Depending on the preparation method used, BBIC contains approximately 110mg/g BBI and 4.54 mg/g isoflavones14. Haelan 951 contains a 6-fold lower amount of isoflavones (0.69 mg/g) but in very similar ratios to BBIC. Genistin, diadzin and glycitin and their metabolites are present at 1 : 0.57 : 0.16 in BBIC14 and at 1 : 0.41 : 0.14 in Haelan 951 (, there is no measure of the amount of BBI in Haelan 951, though the manufacturers’ claim a ‘high rate of the Bowman Birk protease inhibitor compound’ ( Given that our rationale was to compare Haelan 951 with purified forms of its potential active ingredients, namely BBI and isoflavones, we used a dose of isoflavones in this study equivalent to the amount in Haelan 951.

A previous study investigating the effects of the predominant soy isoflavone genistein in mdx, used doses up to 2 mg/kg IP daily11, this is approximately twice the calculated daily dietary intake of total isoflavones (875 µg) of the mice used in this study. In the study by Messina and colleagues, they reported a siginificant increase in grip strength, and improvements in various parameters of muscle pathology including reduced serum creatine-kinase levels, markers of oxidative stress and muscle necrosis and enhanced regeneration, suggesting that genistein alone could be beneficial in mdx. We did not see any beneficial effect of combined isoflavones in our study, but then at a lower dosing regimen, 2 mg/kg IP, 3 times per week compared to 2 mg/kg IP daily, Messina and colleagues also saw no significant improvement11. This would suggest that the dose of isoflavones alone required to achieve a beneficial effect is critical, and/or that isoflavones other than genistein, such as daidzin and glycitin might have deleterious effects, either when used alone or in combination with BBI.


From the data obtained in this study, we saw no beneficial effect of Haelan 951 or isoflavones, and only Bowman-Birk Inhibitor alone produced any significant improvement in muscle function. Further studies into the dose-activity relationship of BBI and isoflavones in mdx mice are therefore clearly warranted.

Competing Interests Statement

G Marston declares that no competing interests exist. SJ Winder is a member of the Editorial Board of PLOS Currents Muscular Dystrophy. However he he took no part in the peer review or editorial decision making with respect to this manuscript.

Data Availability Statement

Metadata for the study can be found at

Corresponding Author

Steve J Winder is the Corresponding Author. They can be reached via the following email address:

]]> 0
Collective Statement Regarding Patient Access to Approved Therapies from the Center Directors of Parent Project Muscular Dystrophy’s Certified Duchenne Care Centers Thu, 15 Mar 2018 09:35:11 +0000 The dystrophinopathies (Duchenne [DMD] and Becker muscular dystrophy) are progressive diseases that until recently had no specific treatments. New FDA pathways to drug approval in rare diseases have resulted in a dramatic increase in the number of treatment trials for DMD and recently, two approved drugs. Health insurance policies for DMD products have been constructed with limited input from neuromuscular specialists directly involved in patient care and without patient input. These policies often reflect a lack of understanding of the disease, clinical population or the treatment. To ensure that policy determinations reflect best clinical practice, we recommend insurers work with neuromuscular specialists with expertise in care for patients with dystrophinopathy, as well as patients and families, and prominent advocacy organizations, such as Parent Project Muscular Dystrophy, in developing policies.



The FDA recently approved two drugs for treatment of Duchenne muscular dystrophy (DMD). There are many ongoing treatment trials for DMD and it is likely that the treatment options will continue to grow. Recent medical and pharmacy benefit policies from health insurers reviewing DMD products have been written with limited input from neuromuscular specialists directly involved in patient care, and without patient input. This has resulted in excessive variability between payors and a high rate of denials, even for patients who received initial approval and have begun treatment. This has been frustrating for patients, parents, and neuromuscular specialists, and these policies may delay treatment resulting in irreversible disease progression. We therefore urge healthcare insurers to work collaboratively with DMD clinical experts when developing policy determinations.

In 2014 Parent Project Muscular Dystrophy (PPMD) began an effort to identify and certify neuromuscular centers around the United States capable of providing comprehensive care to patients and families living with Duchenne muscular dystrophy (DMD). These centers provide care consistent with the published DMD Care Considerations developed with the support of the Centers for Disease Control (CDC). There are currently 17 of these Certified Duchenne Care Centers (CDCCs) throughout the USA. Each CDCC is directed by a neuromuscular specialist with extensive experience and expertise in the management of DMD. The following statement is intended to reflect the CDCC center directors’ collective position regarding the importance of ensuring that patients with DMD have access to FDA approved therapies when prescribed by their clinicians.

About Duchenne Muscular Dystrophy

DMD is the most common, lethal neuromuscular disease of childhood with an incidence of 1:5364 males in the US1. DMD leads to progressive muscle weakness, and eventual respiratory and cardiac failure. After early years of normal development, there is a noticeable lack of gross motor development, then deterioration of the ability to run and climb stairs. The mean age of diagnosis in the U.S. is 4 years old2,3 . Between the ages of nine and 14, the ability to walk is typically lost. In the late teenage years, respiratory and cardiac function deteriorates. Based upon improvements in care including corticosteroid use, the natural history of the disease has been improved, and many people with Duchenne are now expected to live into their mid to late twenties4 .

DMD is caused by mutations in the dystrophin gene on the X-chromosome. This is among the largest genes in the human genome, encompassing 79 exons5 . Different types of mutations are encountered including large deletions (60-65%), duplications (5-10%) and point mutations, small deletions, or point mutations/splice site mutations/intronic mutations (25-35%)6. These mutations result in a loss of the dystrophin protein. Dystrophin is a key component of the dystrophin-glycoprotein complex (DGC) that creates an essential link between the cytoskeleton and extracellular matrix, critical to maintaining muscle membrane stability and prevent muscle fiber breakdown. The loss of dystrophin results in breakdown of the muscle membrane, reduced resistance to contraction and ultimately muscle fiber death. In addition, non-mechanical roles of dystrophin and other components of the DGC are becoming apparent7. Disturbed signaling, as well as regenerative and fibrotic processes likely play a role in downstream pathophysiology and are partly responsible for phenotypic variability8. Until recently, no treatment was available to restore dystrophin production.

DMD is one of two diagnoses classified as “dystrophinopathies” (i.e., resulting in a deficiency or abnormality of dystrophin). There is a spectrum of clinical severity with DMD at the more severe end and Becker muscular dystrophy (BMD) demonstrating a milder phenotype. Different mutations in the dystrophin gene can result in complete absence of dystrophin protein, reduction in the amount of protein, or change in function of the protein. In general, patients with complete absence of dystrophin have DMD while those with residual dystrophin have milder phenotypes.

Since the gene was cloned and the dystrophin protein identified three decades ago9, a wealth of knowledge has accumulated about disease pathophysiology. Multiple drugs are at different stages of development addressing dystrophin restoration or different pathophysiologic processes involved.

The Path to Duchenne Therapies

Drug discovery for rare diseases is complicated by the high cost of drug development and small target patient populations. Since the 1980’s, several targeted legislative measures have been implemented to address these challenges in orphan drug development. As a result, the development of orphan drugs became a more sustainable business model for investors and pharmaceutical companies. In 2016 alone, 9 of the 22 novel drugs approved by FDA were to treat orphan diseases. The orphan product space was further seeded through the creation of innovation incentives such as ‘Fast Track’, ‘Break Through’, and Priority Review designations, and the expansion of the Accelerated Approval pathway. The Accelerated Approval pathway was initially available for serious and life-threatening diseases (i.e. HIV) and was expanded in the 2012 Food and Drug Administration Safety and Innovation Act (FDASIA) reauthorization to also include rare diseases. This has led to the Accelerated Approval of a number of new drugs based on surrogate endpoints. The FDA approval of a drug under the Accelerated Approval process is a full approval based on the fact that the surrogate endpoint is “reasonably likely” to predict clinical benefit. The FDA approval of a drug under the Accelerated Approval process mandates that the sponsor pursue clinical trials to further demonstrate the clinical efficacy and safety of their product. The FDA can withdraw approval for several reasons, including failure to demonstrate clinical benefit in follow on studies.

In September 2016, the DMD community celebrated the first FDA approved drug for DMD, Exondys51 (eteplirsen). This was followed by the approval of Emflaza (deflazacort), a corticosteroid that alters the course of DMD, in February 2017.

Access to Duchenne muscular dystrophy approved therapies

With the approval of these medications and anticipated approval of new therapies in the future, access to treatment has become more complex for the Duchenne community. We acknowledge the high price of these medications, while the US continues to struggle with rising healthcare costs. We also recognize that with the current global system of drug development, bringing an orphan drug to market is expensive. Perhaps as a result of these high costs, payors have developed policies for therapy initiation and continuation often based on incomplete understanding of the natural history and disease progression, potential benefits of these drugs and the risks of withholding therapies. In addition, several insurers have declined to cover these approved therapies, considering them investigational, not medically necessary, or instituting nearly insurmountable barriers to access or prior authorizations that may contradict provider guidance. We, as representatives of the PPMD Certified Duchenne Care Centers and the Duchenne community are extremely troubled by these policies.

It is our view that health insurance should not restrict access to care deemed appropriate by clinicians caring for those lives covered by those health insurers. The role of payors is not to diagnose, determine appropriateness of prescribed therapies, or to contradict the medical opinion of qualified expert medical providers, particularly when use of a drug is “on-label.” To that end, we assert the following positions:

  • Coverage of a drug must reflect the FDA approval status of a drug.
  • When the FDA determines approval of an investigational agent based on the Accelerated Approval pathway, it becomes an FDA-approved drug. With Accelerated Approval (or any other FDA approval), a therapy is approved and no longer considered an investigational product. The drug should, therefore, no longer be denied as an investigational drug.
  • Some insurers, as criteria for renewed approval, have proposed measurement of dystrophin in serial muscle biopsies as criteria for renewed approvals. A muscle biopsy is an invasive surgical procedure requiring sedation and differing degrees of intraoperative ventilatory support. Quantitative measurement of dystrophin protein in a muscle biopsy requires specialized expertise not available in a clinical setting. The risks of a muscle biopsy are carefully considered when constructing clinical trial protocols and may be acceptable for clinical trials, but are not appropriate in broad clinical practice. People living with Duchenne are at increased risk for developing rhabdomyolysis (massive muscle breakdown) and experiencing pulmonary and cardiac failure in the setting of surgical procedures involving general anesthesia. Therefore, this requirement is unreasonable, unethical and impractical.
  • Many insurers are utilizing clinical trial outcome measures in order to establish therapeutic efficacy, demonstrating a lack of understanding of this disease. As an example, the 6MWT (6 minute timed walk test) is a research tool often used in clinical trials but not suitable for assessment in clinical practice. Appropriate monitoring parameters for any drug are found in that drug’s United States product insert, which was constructed based on the parameters of the FDA approval.
  • While patients may receive a genetic report (genotype) that is consistent with Becker muscular dystrophy, they may exhibit the symptoms (phenotype) of a patient with Duchenne muscular dystrophy. Duchenne and Becker phenotypes are determined clinically as the genetic results are not 100% predictive and the genotype-phenotype correlation is often unpredictable.
  • We would therefore endorse that all patients with a genetically confirmed dystrophinopathy, whose neuromuscular specialist feels a therapy is appropriate, have access to that prescribed therapy, regardless of age, or sex.


Recent health insurance policies for DMD products have been constructed with limited input from neuromuscular specialists directly involved in patient care and without patient input. In order to ensure that policy determinations reflect best clinical practice, we implore insurers to work with neuromuscular specialists leading the PPMD Certified Duchenne Care Center teams, as well as patients and families, in developing policies. We represent a collective body of clinicians and clinical investigators, leading the world’s DMD care, registries, clinical trials, research, and natural history studies. We are committed to data-driven, evidence-based determinations and implementation of longitudinal tracking of outcomes of patients exposed to therapy. We also recommend that prominent advocacy organizations, such as PPMD, be involved in the development of policy determinations. Together, we are eager to engage with payors to develop policies that will improve health outcomes for patients with DMD.

Corresponding Author

Kathi Kinnett, Parent Project Muscular Dystrophy (

Competing Interests

Dr. Cristian Ionita and Ms. Kathi Kinnett have no conflict of interest. Dr. Katherine Mathews has read the journals policy and has the following conflicts: she has been a site investigator for Duchenne clinical trials conducted by the following companies: Sarepta Therapeutics, Pfizer, Santhera, PTC, BMS and Italfarmaco.

Data Availability Statement

All relevant data are within the paper.

]]> 0
Duchenne Regulatory Science Consortium Meeting on Disease Progression Modeling for Duchenne Muscular Dystrophy Thu, 12 Jan 2017 10:00:47 +0000 Introduction: The Duchenne Regulatory Science Consortium (D-RSC) was established to develop tools to accelerate drug development for DMD.  The resulting tools are anticipated to meet validity requirements outlined by qualification/endorsement pathways at both the U.S. Food and Drug Administration (FDA) and European Medicines Administration (EMA), and will be made available to the drug development community. The initial goals of the consortium include the development of a disease progression model, with the goal of creating a model that would be used to forecast changes in clinically meaningful endpoints, which would inform clinical trial protocol development and data analysis. 

Methods: In April of 2016 the consortium and other experts met to formulate plans for the development of the model. 

Conclusions: Here we report the results of the meeting, and discussion as to the form of the model that we plan to move forward to develop, after input from the regulatory authorities.



Recent advances in research for Duchenne Muscular Dystrophy (DMD) have resulted in a robust pipeline of drugs in development to treat the disease, with 23 interventional trials ongoing as of June 1, 2016 ( The prevalence of DMD is approximately 1.38 per 10,000 male individuals aged 5 to 24 years1, and given the resulting small numbers of patients with DMD, recruiting new trials with larger numbers of patients will be problematic. This issue is compounded by the fact that most trials to date have focused on endpoints that require the patient to be ambulatory, and that many test articles are directed toward prognostically enriched sub-populations of patients with specific genetic and functional characteristics, thus excluding large parts of the population. In light of the resulting need to design and conduct smaller, more targeted clinical trials, regulatory authorities are open to considering novel endpoints to look at efficacy of such drugs, so long as such endpoints are scientifically justified (FDA and EMA guidances)2,3.

The most advanced potential therapeutics are now completing late stage trials and are being proposed to the regulatory authorities for approval. However, to date only one late stage trial (Santhera’s Idebenone) has met its pre-specified primary endpoint. Different therapeutic mechanisms produce different expectations for the magnitude and time course of a treatment effect. Therapeutics that provide a transient improvement of function may be studied in for shorter durations. Therapeutics that stabilize disease progression require longer study durations to demonstrate slowing of disease progression, particularly in young patients who may be acquiring skills at a slower rate than typically developing unaffected children. Pre-specified and post-hoc subgroup analyses, comparisons to historical natural history data and analyses of specific endpoints most sensitive to a treatment effect in a short duration 12-month trial do, however, indicate the possibility of a drug effect in some cases. The interpretation of such analyses is an ongoing point of discussion throughout the community and with the regulatory authorities. It is clear that in order to conclusively demonstrate that drugs are effective, we need a clearer understanding of sources of variability in disease progression of DMD patients, so that appropriate endpoints can be investigated in appropriately selected patient subgroups. This should allow smaller, shorter trials to be informative.

The Duchenne Regulatory Science Consortium (D-RSC) was established to develop tools to accelerate drug development for DMD. The first tool that will be developed by the consortium is a disease progression model. The resulting tools are anticipated to meet validity requirements outlined by the fit-for-purpose pathway at the U.S. Food and Drug Administration (FDA) and the Qualification of Novel Methodologies for Medicine Development pathway at European Medicines Administration (EMA), and will be made available to the drug development community. The tools must meet the criteria of being clinically meaningful, useful to drug developers and acceptable to the regulatory authorities. To achieve this, the D-RSC is creating an aggregated clinical dataset from multiple industry and academic sources. The data will be used in the first instance to create a disease progression model that will describe how the disease progresses in subgroups of patients defined by clinical variables, with the initial goal of informing trial inclusion criteria and endpoint selection. If the model is formally endorsed by regulatory authorities as is anticipated, it is a further goal of D-RSC to make the model broadly available to the Duchenne research community. D-RSC members and clinical experts held an inaugural meeting in April of 2016 to discuss group practices and development of the initial model.

Summary of discussion of Duchenne disease progression:

Clinical experts lead the discussion with a summary of analyses from existing datasets, highlighting how different endpoints change over the disease course, how endpoints correlate with each other and how they could be used for modeling disease progression. Dr. Craig McDonald discussed data from the UC Davis / Cooperative International Neuromuscular Research Group (CINRG) Duchenne Natural History Study (DNHS)4 and from clinical trials, showing the changes in functional and respiratory endpoints over the course of the disease. He showed that, although progression of disease is variable between patients, it follows a predictable pattern of loss of specific functional milestones (ability to rise from the floor, ability to climb stairs, ability to walk, loss of upper limb function, and loss of respiratory function etc.)5. He presented data demonstrating that although measurements of the length of time it takes to complete tasks change over time, loss of functional milestones tends to happen precipitously, with limited change in timed functional tests prior to the sudden loss of the related milestone. Some of the precipitous decline in lower extremity endpoints is due to the onset of a critical threshold of lower extremity loss of muscle fibers (muscle substrate for the targeted therapeutic) in key muscle groups such as the knee extensors or proximal pelvic girdle. The loss of function may also correlate with an event, such as a fracture, that prevents the patient from being able to complete the test. However, the age or the timing of loss of each milestone could be predicted to some extent by the age or time of loss of previous milestones (unpublished data).

In addition, Dr. McDonald showed data demonstrating that functional endpoints are frequently predictive of each other, such that patients that lose a specific milestone function at a given time can be predicted to lose an additional milestone ability within a given time. For example, the age of loss of standing ability correlates with the age of loss of 4-stair climb and ambulation in individual patients, and a baseline time to stand from supine predicts a loss of standing ability, stair climbing ability and ambulation within a defined period of time6. Similarly, patients walking less than 300 m in a six minute walk test have more heterogeneity in their change in 6-minute walk distance (6MWD) are much more likely to lose the ability to complete the test within the next year than patients outside of that window7. Essentially no patients with baseline 6MWD above 400 meters lose ambulation in 12 months and these patients tend to show little decline in 6MWD over the course of a year. These correlations between milestones are seen throughout the disease, including non-ambulatory endpoints such as loss of ability to self-feed (bring the hand to mouth) and respiratory measures such as time to a forced vital capacity of 50% or 30%. Such correlations can be seen in multiple datasets, both in natural history data and in the placebo arms of trials. These correlations suggest that the pattern of progression of loss of these milestones is consistent, and may be able to be predicted by specific baseline characteristics of the patients. If that pattern of progression could be modeled, it might be possible to predict which patients are likely to lose specific milestone functions in a given period of time, and deviation from such a pattern might be indicative of a treatment effect.

The most commonly used composite endpoint scale that incorporates these milestone functional changes is the Northstar Ambulatory Assessment, which is increasingly being used in clinical trials. The non-linearized scale has each ambulatory milestone reduced to a score between 0 and 2 depending on the patient’s ability to complete the test8. Dr. McDonald showed data from one trial, where the overall linearized 100 point scale did not show a statistically significant difference between treated and placebo patients. However, the data shows that patients in the drug arm showed fewer relative losses of every function tested (shifts from a 2 or 1 score on the NSAA to 0) across all 17 functions tested using an odd’s ratio (a reduction in lost clinically meaningful milestones). Thus, considerations of losses of milestones or individual functions by calculation of an odd’s ratio appears to be more statistically robust and clinically meaningful than the change in the overall summated or linearized NSAA score. Dr. McDonald suggested that by summing the scale across all 17 functions or linearizing the scale there may be lost granularity or sensitivity of a scale that can potentially be used across patients across a wide spectrum of ambulatory disease progression.

Dr. McDonald showed data from the CINRG dataset relative to respiratory measures, specifically forced vital capacity (FVC). While absolute forced vital capacity increases with maturation and growth to a plateau phase during adolescence, percent predicted FVC does change in a more linear fashion with disease progression, slowly decreasing from the time patients are quite young. Differences in % predicted FVC can be seen between patients who have taken steroids and those who have not, indicating that the measure might be able to detect a treatment effect9. However, Dr. McDonald pointed out challenges with how the community calculates percentage predicted FVC and other spirometry values, based on the boy’s height, which is confounded by steroid treatment. Prediction equations for height based on ulnar length can be used to calculate percent predicted spirometry values. He also noted that it is unclear what a clinically meaningful change in FVC would be in earlier stages in disease. He suggested that passing certain thresholds of FVC, such as 50% predicted FVC (a time when mechanical cough assistance may be recommended) or 30% predicted FVC (when all patients are recommended to have started nocturnal ventilation), might be used as later stage functional milestones in disease, similar to the functional milestones he suggested in earlier stage disease. Alternatively, a threshold of less than 1 liter absolute forced vital capacity has been linked to 5-year survival in DMD and this could be another threshold for time to event analyses later in the disease course.

Dr. Erik Henricson discussed how the functional milestones discussed by Dr. McDonald correlate with patient reported functional health outcomes, and how such outcomes change over the course of disease. Correlation between patient reported outcomes and loss of functional endpoints may be useful to establish the clinical meaningfulness of such measures. He showed that the Pediatric Outcomes Data Collection Instrument (PODCI) mobility -oriented subdomain scales correlate with functional measures, and could be used across a significant age range of patients10. He showed data demonstrating that PODCI scores for the transfer and basic mobility subscores change linearly with 6 minute walk distance11. When patients in the CINRG natural history study were divided into groups relative to their loss of functional milestones, these groups of patients showed a distinct pattern of change in the PODCI indices10. Steroid treatment was associated with significant differences in the age of patients that fit into each milestone group, indicating a measurable treatment effect. This suggests that patient reported outcome scores in Duchenne do link closely to the loss of functional milestones.

The PODCI scale was developed to assess children with a variety of orthopedic limitations, and it describes DMD patients across most stages of disease. However, the instrument demonstrates ceiling effects in highly functional young children and floor effects in individuals with very advanced disease. To address these issues, Dr. Henricson described his team’s ongoing efforts to combine items from multiple scales into a single mobility construct-oriented PRO mobility assessment, the DMD Lifetime Mobility Scale (DMD-LMS) (Erik Henricson, unpublished data). The instrument will include a broadened list of “milestone” tasks that are meaningful to patients (e.g. activities of daily living), including tasks involved in walking and moving, changing and maintaining body position, and lifting and handling objects. Included Items are demonstrated to show responsiveness to differences in disease stage and steroid treatment effects, and are sensitive to 1-year changes in disease progression typical to today’s clinical trial designs. Once complete the DMD-LMS, developed using Item Response Theory techniques, will be a continuous scale instrument that describes mobility-related functional task ability from early ambulatory to late non-ambulatory levels of disease involvement.

Dr. William Rooney discussed data collected by the Imaging-DMD consortium (, which is looking at skeletal muscle magnetic resonance imaging (MRI) and magnetic resonance spectroscopy (MRS) as measures of disease progression across the course of the disease. The Imaging-DMD consortium was set up to investigate the hypothesis that magnetic resonance (MR) markers of muscle pathology are sensitive to disease progression in boys with DMD across a wide range of disease stages and predictive of loss of function12,13. Dr. Rooney shared data showing the visible changes in fat fraction over time in MR images, which start prior to any loss of function14. Dr. Rooney showed images and data that demonstrate that fat fraction (measured by either MRI and MRS) and MRI T2 increase markedly with DMD disease progression. Fat fraction measured by each technique provided different information, but the pattern of increase correlated between the measures14.

Imaging DMD has studied 133 patients with DMD and 50 controls over 5 years. Dr. Rooney showed that over time, the fat fraction of the muscle increased in all of the leg muscle groups studied, although the absolute amount of fat detected and the rate of change varied between muscles13,14. Substantial annual changes in fat fraction and MRI T2 values were observed from many muscles of the leg with standardized response means greater than 0.7514. He noted that the vastus lateralis muscle could serve as a “sentinel” muscle, showing effects earlier than most functional deficits can be detected, indeed even when functional abilities may be improving developmentally13,14. Other muscles showed later changes in fat fraction, but all muscles tested showed increases over the course of the study.14

Dr. Rooney showed data that demonstrated that the fat fraction of the vastus lateralis muscle changes in a Gaussian-type function15, which could be modeled statistically based on age and baseline fat fraction. Similar curves could be calculated for the other muscles. The curves mimicked the progression of the disease, suggesting that the model could be used to characterize muscle involvement across disease. The Imaging DMD consortium has additional data that demonstrates that functional endpoints can be mapped to changes in fat fraction in specific muscles based on disease progression modeling (William Rooney and Imaging-DMD Consortium, unpublished data).

Planning for a disease progression model:

The group discussed what a useful model of disease progression might look like. For the model to be useful in developing clinical trial protocols, it needs to be able to predict which patients are likely to change in specific endpoints in a statistically meaningful way over a period of less than a couple of years, so as to inform inclusion criteria and size and length of a trial. It needs to be able to predict clinically meaningful changes in the patients, in order to be of use in regulatory decisions. It also needs to be supported by high quality data.

The integrated dataset that D-RSC is building will be limited by the data that have been collected, and that the Duchenne community is willing to share with the consortium. It will integrate multiple datasets so as to include a variety of patients across the disease spectrum and located at different centers so as to represent patients as broadly as possible and account for differences between them. Over time, as more data become available, D-RSC envisages new versions of the model being developed to account for new data that becomes available.

The group agreed that modeling any individual functional endpoint would limit the utility of the model, as only a small proportion of patients are expected to change significantly in any given functional milestone over the period of one year due to disease alone. Measures that could be modeled over longer periods of disease include FVC, Northstar Ambulatory Assessment, patient reported outcome measures such as Dr. Henricson’s longitudinal mobility scale and fat fraction by MRI. Even these measurements do not cover the entire course of disease, but exclude the very young (MRI changes are detected as young as age 4, and patient reported outcomes do not currently cover those younger than 3-4). Late non-ambulatory patients may also not be covered as well by these measures.

The group discussed the use of muscle MR measures as a measurement that changes across disease course, and data supports its relationship with functional changes. Dr. Rooney’s data showed that different muscles showed different patterns of when fat fraction started to increase, and rates of increase, suggesting that fat fraction of different muscles might be best predictive of different functional changes. Dr. Wong noted that the the gluteal muscles are the first to be involved in DMD, and progression of weakness of the gluteal muscles (i.e. pelvic extension) would be expected to correlate with the timed rise from the floor test for early ambulatory DMD patients, while the quadriceps are not yet involved and see less fat infiltration in young DMD boys. Hence MRI of fatty changes of the gluteal muscles could be expected to be informative for disease progression in early stage patients, and might be expected to correlate with specific functional outcomes. Dr. McDonald suggested that creating a combined measurement of fat fraction from multiple muscles might provide a measure to predict multiple functional endpoints across disease.

The group noted that the curves for muscle fat fraction resembled those for manual muscle testing, and wondered how well the fat fraction correlates with muscle strength, both of which have previously been shown to correlate with functional tests16. This hypothesis warrants further testing. If the muscle fat fraction changes closely with strength of that muscle, it may be possible to create a relevant composite measure that describes the amount of remaining functional muscle, which might be expected to track closely with functional abilities.

Dr. Klaus Romero suggested that it might be possible to use a variable such as fat fraction, muscle strength or a composite of the two to model across disease stages in order to predict the loss of specific milestones. Loss of carefully selected functional milestones would likely be seen as clinically meaningful by the regulatory authorities, but this would need to be confirmed with FDA/EMA. He suggested that a time-to-event modeling approach could be developed, using the clinically-relevant milestones as the endpoints of interest, once these have been agreed-upon with regulators. This time-to-event model would, in turn, be driven by the parameters derived from non-linear-mixed-effects models for muscle MR (or muscle strength, or a combination), as well as optimized longitudinal outcome measure scales. These non-linear-mixed-effects models would, in turn, capture relevant sources of variability, which were discussed and are described in the following paragraphs. In this framework, the functional milestones would be the ultimate endpoints in a trial, but they would be linked and predicted by quantitative models that describe the progression of continuous optimized scales and biomarkers. The group agreed that this construct would be meaningful to them, and would be likely to be supported by the data.

The disease “milestones” most commonly referred to are functional outcome measures in the ambulatory patient group such as loss of ability to rise from the floor, loss of ability to climb stairs, loss of ability to walk, although upper extremity milestones such as loss of ability to reach overhead or get a hand to the mouth have also been described. There was discussion of additional milestones that could be included to mark changes later in the disease, and it was thought that meeting specific respiratory milestones that correlate with treatment changes, and some of the upper limb scores might be able to be used. It was noted that the definitions of such milestones will need to be carefully agreed-upon, based on a) what is clinically relevant, b) what has been captured in the available data sources, c) what milestones appropriately represent functions across the entire continuum of disease, and d) what makes sense from a drug development and regulatory perspective. This is critically important, as at different centers the loss of the milestones is interpreted differently. For example, definition of “loss of ambulation” can be based on a clinical evaluator or person-reported outcome of loss of ambulation or full-time wheelchair use, the inability to complete a 6 minute walk test, an inability to complete a 10m walk/run test, or an inability to stand or walk even a single step (on the NSAA). For the purposes of this model, we will need to clearly define what we mean by the milestone and ensure that the data in the database reflects those definitions.

The group noted that in considering the data we will need to describe other sources of variability in the population, such as differences in anthropomorphic characteristics (height, weight), variations in use of steroid regimes and other preventive clinical practices, and underlying differences in patient’s genetic characteristics. These factors, as well as baseline functional measures, are expected to affect the progression of the disease in individuals, and need to be incorporated into the proposed modeling approach, along with any biomarker data that can be accessed. The variables that will be used in the final model will be selected after stringent statistical analyses of the dataset, but expert opinion provided in this meeting allows D-RSC to ensure that datasets including variables that are thought to be of interest are included.


The group concluded that they were interested in pursuing such a modeling approach based on consideration of muscle fat fraction, timed function tests, muscle strength, as well as optimized scales of function in ambulant and non-ambulant patients, with additional consideration of time to event analyses of specific disease milestones. The variables that are finally included in the model will be dictated by the data – both what data is in the integrated dataset and what early stage analyses of that data tells us about what is most relevant to the final model. The context of use of the model would be to forecast changes in clinically meaningful outcome measures, which would inform clinical trial protocol development with respect to inclusion criteria, endpoints to include, and size and length of trials.

Data Availability Statement

This report is the proceedings of a meeting, and no new data was reported.

Competing Interest Statement

The authors have declared that no competing interests exist.

Corresponding Author

Jane Larkindale:

]]> 0
New Recommendation on Biological Materials Could Hamper Muscular Dystrophy Research Wed, 21 Dec 2016 10:00:06 +0000 The new ‘Recommendation of the Committee of Ministers to member States on research on biological materials of human origin’, adopted in Europe in May 2016 is confusing and lacks specificity on the research use of biomaterials taken from persons not able to consent. It is possible to interpret the relevant clauses in a restrictive manner and doing so would hamper biobank research, by requiring researchers or biobank curators to examine individual records in detail, to check they are adhering to the Recommendation. This would be particularly problematic for muscular dystrophy and other rare disease research, the progress of which relies increasingly on the sharing of biomaterials and data internationally, as it will add complexity to the logistics of biomaterials and data sharing and introduce barriers for researchers preparing biomaterials for sharing. Such barriers are contradictory to EC policies on promoting and funding rare disease research and removing barriers to better care and treatment. Such policies work in concert with international progress in rare disease research, in particular the NIH’s Rare Diseases Clinical Research Network and Genetic and Rare Diseases Information Centre. The rare disease community has in recent years worked to create a common framework of harmonised approaches to enable the responsible, voluntary, and secure sharing of biomaterials and data. These efforts are supported by the European Commission in such moves as FP7 funding to advance rare disease research and the introduction of National Plans for rare disease; and are bolstered by similar efforts in the USA via the Clinical and Translational Science Awards Program and the NIH/NCATS Patient Registry developments. Introducing Recommendations from the Committee of Ministers, containing clauses which are incompatible to the efforts to advance rare disease research, seems counter-productive.



The pre-amble to the new Recommendation of the Committee of Ministers to member States on research on biological materials of human origin, adopted in May 2016, announces that it is designed to take into account new developments in biobanking including the cross-border flow of such materials for research1. However, the recommendation contains a stipulation on materials from ‘persons not able to consent’ which is confusing and which has the potential to stifle global efforts in rare disease which aim to enable and encourage resource sharing among researchers.

Recommendation 12.1 (repeated at 22.5) states that for those persons not able to consent in law biomaterials ‘should only be obtained or stored for future research having the potential to produce, in the absence of direct benefit to the person concerned, benefit to other persons in the same age category or afflicted with the same disease or disorder or having the same condition, and if the aims of the research could not reasonably be achieved using biological materials from persons able to consent’ (ibid).

The ethical intention behind the recommendation is important because it recognises the potential vulnerability of incapacitated persons participating in research2. However the wording is problematic because does not indicate whether it applies to children and/or to adults who lack capacity. The inclusion of the term ‘same age category’ implies that the Recommendation is referring to children but is not explicit, which in itself in problematic as it means the Recommendation may be read as referring to children or not and therefore result in varying practices. The Recommendation does not provide a reasonable rationale for restricting the potential beneficiary groups to parity of age, nor is there any justification for limiting benefit to those with the same condition, a consideration which is particularly pertinent for rare disease research. In addition it does not distinguish sufficiently between the ethical concerns which arise at the point of entering a person who lacks capacity into a study in which tissue will be collected, from the ethical issues of using tissue already collected and stored. Most important here is that some consideration is given to the potential for risk and burden from the research procedure quite independent of whether the use of tissue has the potential to benefit others. Without clarification on these points the recommendation is likely to be interpreted in a restrictive way, for example by a researcher deciding that the Recommendation does refer to children’s samples in a biobank and that those samples can not therefore be shared with research studies which include adults with a different condition. Such a move has the potential to be highly detrimental to important areas of research collaboration, particularly in rare diseases, where there is general agreement that samples should have maximum utility.

Children or adults?

The wording is unclear as to whether the Recommendation is intended to apply to adults and children. Although there are similarities in the ethical issues raised there are important differences; there are also differences in the legal provisions for these different contexts. This latter point is important because in most jurisdictions there are provisions which allow a role for the proxy consent of parents or a person acting with parental responsibility. The person acting with parental responsibility can bring other safeguards into the decision including their knowledge of the child and their judgement regarding the child’s best interests. When it comes to adults who lack capacity and their involvement in research then the situation is more heterogeneous with far fewer jurisdictions providing specific legal provision for research. Many jurisdictions do not recognise a role for substituted judgement or proxy consent for adults which has prompted the development of specific legislation to deal with adults who lack capacity. England and Wales is one jurisdiction where there is a specific law dealing with the involvement of adults who lack capacity in research. The Mental Capacity Act (MCA 2005) of England and Wales is a good example where there is a specific legal framework for the recruitment of adults who lack capacity and sets out the criteria, which must be met in order for a research project to proceed.

Rare disease research

The scarcity and therefore elevated value of biological materials for research in rare diseases provides an argument for making use of all available materials and biobank legislation usually provides a robust legal framework for this. Of particular importance for rare disease research is the capacity for the international sharing of biological materials. This is an important consideration when recognising the right of people living with a rare disease to benefit from health care, prevention, and medical treatment.

This imperative is recognised internationally, as can be seen by the increasing profile of, and funding around, rare disease, which includes significant input from the European Commission and the NIH. In addition, worldwide collaborations are pushing forward with initiatives which are building infrastructures and standards to ease the sharing of data and biomaterials for rare disease. The International Rare Disease Research Consortium (IRDiRC), aims to accelerate rare disease diagnosis and therapies, and the Global Alliance for Genomics and Health (GA4GH) is working to create a common framework of harmonised approaches to enable the responsible, voluntary, and secure sharing of genomic and clinical data.

All these initiatives represent a significant expenditure of resources by public and private organisations around the world and epitomise an era of collaboration for rare disease which has never been seen before and which is already bearing fruit3. If organisations interpret the new Recommendation as being applicable to children, this could cause extensive problems for researchers using materials from rare disease biobanks.

We have indicated the ambiguity of recommendation 12.1 but in addition there are three significant reasons why they should not be applied to children’s tissues. First, because they conflict with the UN Convention on the Rights of the Child, the thrust of which emphasises the right of children to benefit from the ‘highest attainable standard of health and to facilities for the treatment of illness and rehabilitation of health’ (1997, p7), as outlined in Article 23.

Second because it conflicts with the express, wishes of rare disease patients, families and organisations to engage in collaborative, solidaristic actions to improve rare disease research through collective endeavour across different disorders and national boundaries. A child whose future is foreclosed by a progressive and incurable disease inspires a moral claim on human endeavour, to find a means of treating and ideally curing them, that few would dispute. The vulnerability of such children might be used to justify a highly precautionary attitude to research governance but Recommendation 12.1 is restrictive for no net gain in ethical safeguards. The application of recommendation 12.1 to children would repudiate the wishes of rare disease patients and their families and be contra to the decades of creative and collective self-organisation and education that rare disease patient organisations have initiated4,5. From the 1960s onwards, parents, on behalf of their children, developed organisations that are: major funders of research; effective political lobbyists; and founders of strategic alliances, and part of this has been a practical and intellectual challenge to the presumption of vulnerability6. The authors are connected to RD Connect, an EC FP7 funded project which is constructing an integrated global platform connecting databases, registries, biobanks and clinical bioinformatics for rare disease research. Our research shows that rare disease patients and patient organisations are highly solidaristic. They recognise that unity through common interests and objectives can be a powerful lever for action and also that they may benefit from research unrelated directly to their condition, in the future7,8. It can not be known where this potential crossover of benefit might arise and it is therefore important not to unduly restrict research to particular disease boundaries.

Third because the recommendation represents a narrowing of what is currently permitted in other ethical guidelines/national law and is therefore contradictory to current practices. The importance of emphasising potential benefit and negligible risk, which the recommendations do not mention, is that it provides considerable safeguard to the person who lacks capacity; especially significant in the context of taking tissue samples. Whereas the emphasis given in the recommendations to the potential beneficiary being of the ‘same age category’ is too vague and of questionable ethical importance. What is an age category? Is it a five or ten or twenty year range? Is it there to prevent the exploitation of children in favour of adults? On one interpretation the Recommendation seems unduly restrictive, as tissue taken from a healthy child who does not have the capacity to consent, may be an important control for research conducted on a specific children’s disease. The Recommendations do not seem to recognise that tissue obtained for a specific research purpose is often not exhausted for that purpose and often retained for future unspecified research, equally tissue may be obtained for an unspecified broad purpose, as is commonly done in the context of biobanks. The Recommendation seem particularly restrictive on this point and is therefore of serious concern to those conducting a broad programme of research, such as those undertaking research into genetic, childhood diseases.


If Recommendation 12.1 is applied to children, it would unduly restrict rare disease research by preventing the use of children’s biomaterials in research involving adults, or on anyone without the same condition and thereby constrict children’s right to benefit from the solidaristic actions of others. It would make efforts to improve data and biomaterials sharing for rare diseases more complex, time consuming and less efficient. Platforms such as RD Connect and BBMRI-ERIC which were designed to contribute to the efficacy and excellent of European research by easing access to resources, could be rendered less efficient by the need to introduce logistical complexity to isolate samples from children and to examine whether requests for samples are made in accordance with the proposed Recommendation. This is a crucial factor as 75% of rare diseases affect children and 30% of rare disease patients die before they are 5 years old.

Researchers and research organisations should be wary of a restrictive interpretation of Recommendation, lest this by default becomes soft policy. A clarification from the Council of Ministers as to whether the relevant paragraphs of the new Recommendation are meant to include biomaterials from children would be helpful and would recognise the potential ramifications of their decision for research on childhood conditions, and for the unprecedented progress in rare disease.

Data Availability Statement

All relevant data are within the article.

Competing Interest Statement

The authors declare that they have no competing interests.

Corresponding Author

Pauline McCormack:

Author Contributions

PMcC and SW contributed equally regarding ideas, drafting and editing of the manuscript.

]]> 0
Prednisone and Deflazacort in Duchenne Muscular Dystrophy: Do They Play a Different Role in Child Behavior and Perceived Quality of Life? Fri, 17 Jun 2016 09:32:14 +0000 The aim of this study was to determine whether prednisone and deflazacort play a different role in child behavior and perceived health related psychosocial quality of life in ambulant boys with Duchenne Muscular Dystrophy. As part of a prospective natural-history study, parents of sixty-seven ambulant boys with DMD (27 taking prednisone, 15 taking deflazacort, 25 were steroid naïve) completed the Child Behavior Checklist (CBCL) for assessment of behavioral, emotional and social problems and both parents and boys with DMD completed the PedsQL™4.0 generic core scale short form. Boys with DMD had higher rates of general behavioral problems than age-matched peers. No significant differences were found among the groups for any of the CBCL syndrome scales raw scores, including internalizing and externalizing behaviors; however, on average boys taking deflazacort demonstrated more withdrawn behaviors than those taking prednisone, while on average the boys taking prednisone demonstrated more aggressive behaviors than boys taking deflazacort. Age, internalizing and externalizing behaviors accounted for 39 and 48% of the variance in psychosocial quality of life for both parents and boys with DMD, respectively. Overall, the use of steroids was not associated with more behavioral problems in boys with DMD. As behavior played a significant role in psychosocial quality of life, comprehensive assessment and treatment of behavioral problems is crucial in this population.



Duchenne muscular dystrophy (DMD) is an X-linked recessive disease of muscle that affects 1 in 6000 live males.1 DMD is caused by complete absence of the dystrophin protein in skeletal muscle, myocardium, and brain2 and is characterized by a progressive loss of functional muscle mass leading to deterioration in motor milestones with age. Currently, corticosteroids, such as prednisone and prednisolone, are the only US Food and Drug Administration (FDA) approved medication available to stabilize muscle strength, extend ambulation and standing ability, and minimize the incidence of spinal deformity in individuals with DMD. Deflazacort, a similar corticosteroid is currently in clinical trials; however is not approved for use by the FDA. While the benefits of deflazacort are similar to that of prednisone, it is often prescribed because of a perceived reduction in side effects such as weight gain3 and undesired behaviors such as aggression.4 The decision to initiate the use of corticosteroids depends on functional state, age, and pre-existing risk factors for adverse side-effects;4 however, the precise timing of steroid initiation is typically an individual family decision with physician guidance.

There is emerging evidence indicating that the absence of dystrophin results in a disordered architecture of the central nervous system, which affects the brain.5 As a result, boys with DMD have an increased risk for behavioral problems, specifically attention deficit hyperactivity disorder (ADHD), autism spectrum disorder, obsessive-compulsive disorder, depression, anxiety, and social difficulties.6 In addition, boys with DMD are reported to have difficulties in social functioning that might be due to biologically based deficits in specific social cognitive skills such as social reciprocity, social judgment, and affective discrimination.4 These emotional and behavioral problems may lead to social isolation and withdrawal,4 with many boys with DMD, regardless of steroid use, becoming more withdrawn, depressed and isolated as they get older.7

Anecdotally, parents report an increase in negative behaviors, specifically mood swings, difficult behavior and aggression, with the initiation of corticosteroids; however, research evaluating the impact of corticosteroids on behavior and emotional difficulties remains equivocal.3,6,8 A recent study by Caspers Conway et al reported that corticosteroid and mobility device use were associated with behavior problems.9 Studies by both Hinton et al. and Hendriksen et al. have both found that steroids were not associated with the behavioral outcomes of the children in their studies.5,10 Hendriksen et al. reported that there was no significant difference in overall social adjustment between males with DMD taking corticosteroids and those who were not and that there was no indication that steroids played a major role in negatively modifying psychological adjustment.5 Similarly, Hinton et al. found that steroid use was not the primary factor contributing to an increase in social problems, reported by the parents.10 Despite the focus on the deterioration in motor capacity associated with DMD, parents report that emotional and behavioral problems are a significant issue, which can also negatively affect their son’s quality of life.11,12

To gain further insight into the impact that corticosteroids may have on behavioral and emotional problems and health related quality of life (HRQOL) in ambulant boys with DMD, the objective of this study was to determine whether prednisone and deflazacort play a different role in child behavior and perceived psychosocial quality of life.



Participants included boys with DMD, between the ages of 4 to 15 years who were ambulatory and were part of a multicenter (Shriners Hospitals for Children-Portland, Shriners Hospitals for Children-Northern California or The University of California-Los Angeles) natural history study focusing on the biomechanics of gait in boys with DMD. Eighty-five boys with DMD participated in the longitudinal study; however, at the time of implementation of a measure assessing emotional, behavioral and social functioning into the protocol only 67 boys remained ambulatory, were still participating in the study and completed this measure. In this paper, we investigate the initial behavioral assessments and health related quality of life using a cross-sectional design. Confirmation of Duchenne muscular dystrophy was based on typical clinical presentation of Duchenne muscular dystrophy and one or more of the following: documentation of disease-causing mutation in the dystrophin gene, elevated serum creating kinase levels, and a family history of an affected relative with either a disease-causing mutation in the dystrophic gene and/or complete dystrophin deficiency as shown by immunostaining on muscle biopsy. The Institutional Review Boards from all participating centers approved this study. All parents/guardians provided informed consent and boys provided assent when appropriate.


As part of the longitudinal study, parents completed the Child Behavior Checklist (CBCL) 1½-5 or 6-18 years.13 The CBCL is a 118-item questionnaire assessing behavioral, emotional, and social problems rated on a 3-point scale ranging from 0 (Not True) to 2 (Very True or Often True).13 In this study, we only used the syndrome scales of the CBCL. Syndromes are sets of concurrent problems that tend to co-occur together. Syndrome scales include anxious/depressed, withdrawn/depressed, somatic complaints, social problems, thought problems, attention problems, rule-breaking behavior and aggressive behavior. Syndrome scales are categorized into internalizing and externalizing behaviors. Internalizing behaviors are problems that are primarily within the individual and include anxious/depressed, withdrawn/depressed and somatic complaints, while externalizing behaviors are problems that mainly involve conflict with other people and their expectations for the child and include rule-breaking and aggressive behavior subscales.14 Higher scores on the CBCL indicate more problems. In addition to the completion of the CBCL, parents and boys with DMD completed the PedsQL™ 4.0 generic core scale short form (SF-15).15,16,17 For boys between the ages of 5 and 7 and in situations where the child was unable to read or complete the questionnaire independently because of cognitive impairment, the PedsQL™ was interview-administered by one of the researchers. The PedsQL asks “how much of a problem has your child had with…,” with a score of 0=never and a score of 4=almost always. The PedsQL is reversed scored so that a score of 100 means that the parent/child did not report any problems in this area.

Statistical Analysis

Boys were divided into three groups according to their parent-initiated use of corticosteroids: prednisone, deflazacort, and steroid naïve (had not started taking steroids at the time of the initial CBCL assessment). While power was not determined for the behavior assessment, power calculations for the primary outcomes (gait and energy cost) determined that a sample size of 68 subjects was needed to obtain 80% power to detect a correlation of 0.3 or greater between gait, energy and functional variables at p< 0.05; however, a sample size of 85 was recruited to account for potential subject drop out over the study period. To examine whether the behavior profiles for the syndrome scales differed among the three groups one-way ANOVAs of the raw CBCL scores were used. For the syndrome scales, T scores below 65 are considered within the normal range, while scores between 65 and 69 are considered borderline clinical behaviors and scores greater than 70 are considered clinical behaviors.14 Using 65 as the cutoff, boys were classified as having or not having borderline/clinical behaviors. Crosstabs were used to determine whether the proportion of boys falling into the borderline/clinical category differed by parent-selected corticosteroid group.

For the PedsQL, Pearson correlation coefficients and intraclass correlation coefficients (ICC 2,1) were used to determine agreement in quality of life domains between parents and their sons. ICC values were considered poor to fair if ≤.40, moderate if 0.40-0.60, good it 0.61-.80 and excellent if 0.81 to 1.0.8 A mean psychosocial health summary score was computed by summing the items from the emotional, social, and school functioning scales and dividing by the number of items.18 Due to the small number of subjects and the large number of subscales within the CBCL, only the summary scores from the internalizing and externalizing behaviors were used in the regression analysis. Hierarchical linear regressions were used to determine whether age and internalizing and externalizing behaviors predicted parent proxy and child reported psychosocial health quality of life. In order to account for the influence of age on behaviors, age was entered at step one, followed by internalizing and externalizing behaviors at step 2. Significance was set at p < 0.05.


Population Characteristics

All 67 boys were ambulant without assistive devices. 27 boys were taking prednisone (mean age 8 years, 5 months), 15 boys were taking deflazacort (mean age 9 years, 8 months), and 25 boys were not taking steroids or had not started taking steroids at the time of the evaluation (mean age 6 years, 5 months). The boys using steroids (prednisone and deflazacort) were significantly older than those who were steroid naïve, p=.01. No differences in age were found between the steroid groups (p=.38).

Behavior of boys with DMD and the influence of steroids

No significant differences were found among the groups for any of the CBCL syndrome scales raw scores, including internalizing and externalizing behaviors (Table 1). Mean T-scores for the syndrome scales (anxious/depressed, withdrawn/depressed, somatic complaints, social problems, thought problems, attention problems, rule-breaking behavior and aggressive behavior) fell within normal limits (<65) for all corticosteroid groups. There were no differences in the percentage of boys with T-scores in the borderline/clinical category within each parent-selected corticosteroid group for any of the behavioral subscales, including internalizing and externalizing. While not statistically significantly different, greater than 30% of boys on prednisone had T-scores in the borderline/clinical range for externalizing behaviors (rule breaking and aggressive behaviors), while greater than 30% of boys on deflazacort had T-scores in the borderline/clinical range on the withdrawn/depression syndrome subscale (Figure 1).

Table 1: CBCL Results Raw and T scores for the total group and by parent-selected corticosteroid group

Means (standard deviations)

Variable Total Group (n=67) No Steroid (n=25) Prednisone (n=27) Deflazacort (n=15) ANOVA
T Score Raw T Score Raw T Score Raw T Score Raw Significance
Internalizing 53 (11) 8 (6) 52 (12) 8 (6) 55 (10) 8 (5) 53 (10) 8 (6) p=.96
Externalizing 52 (11) 10 (8) 49 (14) 10 (10) 55 (9) 11 (7) 51 (10) 8 (8) p=.66
Anxiety/Depressed 55 (7) 10 (8) 55 (7) 3 (3) 55 (7) 4 (4) 55 (6) 3 (3) p=.88
Withdrawn/Depressed 57 (7) 2 (2) 56 (6) 2 (2) 58 (6) 2 (2) 57 (9) 2 (3) p=.80
Social Problems 59 (8) 5 (4) 61 (12) 6 (5) 60 (7) 5 (3) 57 (7) 3 (3) p=.28
Attention Problems 57 (7) 5 (4) 56 (7) 4 (5) 57 (7) 5 (4) 57 (7) 5 (4) p=.72
Aggressive Behavior 56 (8) 7 (7) 56 (10) 7 (8) 58 (7) 8 (5) 55 (7) 6 (6) p=.57

Percent of boys with T scores in the Borderline/Clinical Range (>65) for the total group and by parent-selected steroid group

Fig. 1: Bordeline/Clinical CBCL Results

Percent of boys with T scores in the Borderline/Clinical Range (>65) for the total group and by parent-selected steroid group

Agreement in perceived quality of life between boys with DMD and their parents

Agreement on the PedsQL between parent proxy and child self-report is shown in Table 2. Overall, boys with DMD self-reported lower quality of life than parent proxy in all domains, except school functioning. Moderate agreement between the boys and their parents was found for physical functioning (r=.49, ICC=.48) and social functioning (r=.42, ICC=.42) only. The agreement between boys and their parents for psychosocial health, which includes emotional, social and school functioning was considered poor to fair (r=.36, ICC=.36). Agreement between boys and their parents for emotional (r=.25, ICC=.24) and school functioning (r=.32, ICC=.36) were considered poor to fair.

Table 2: PedsQL Means and agreement indices for parent-son reports of quality of life

sd=standard deviation, ICC=intraclass correlation coefficient, CI=confidence interval **Correlation is significant at the 0.01 level (2-tailed) * Correlation is significant at the 0.05 level (2-tailed)

PedsQL Item Parent, Mean (sd) Child, Mean (sd) r ICC (95% CI)
Physical Health 43.2 (27.6) 42.9 (22.5) .49** .48 (.20-.69)
Psychosocial Health 66.2 (15.6) 64.2 (15.4) .36* .36 (.05-.61)
Emotional Functioning 66.5 (16.7) 63.2 (21.3) .25 .24 (-.07-.52)
Social Functioning 72.7 (22.8) 67.3 (23.8) .42** .42 (.13-.65)
School Functioning 59.7 (20.6) 61.7 (21.3) .32 .32 (.00-.58)

The influence of behavior problems on psychosocial quality of life

For parent-proxy reported psychosocial quality of life, the variables of age and internalizing and externalizing behaviors accounted for 39% of the variance in psychosocial quality of life F(3,63)=12.79, p<.001. At step 1 age accounted for a significant amount of variance (R2=.08), while the addition of internalizing and externalizing behaviors at step 2 resulted in a significant increase in the amount of explained variance (R2=.39, Finc (2, 60) =15.09, p<.001) Psychosocial quality of life was negatively associated with age (b=-.25, p=.018) and internalizing behaviors (b=-.36, p=.007). A similar pattern was also seen for psychosocial quality of life as self-reported by the boys with DMD. Age and internalizing and externalizing behaviors accounted for 48.5% of the variance in psychosocial quality of life F=(3, 33)=10.35, p<.001. At step 1, age accounted for a significant amount of variance (R2=.15), while the addition of internalizing and externalizing behaviors at step 2 resulted in a significant increase in the amount of explained variance (R2=.34, Finc (2,33)=10.87, p<.001). In contrast to the parents, self-reported psychosocial quality of life was positively associated with age (b=.31, p=.02). Similar to the parents, age was negatively associated with internalizing behaviors (b=-.45, p=.01).


Social and behavioral issues have been reported in boys with DMD, with parents reporting that these problems are exacerbated with the use of steroids. Similar to the findings reported in the literature,10,19 boys with DMD in this study had higher rates of behavioral problems than age-matched peers. Despite slightly higher rates of behavioral problems for the entire cohort, no statistically significant differences were found between steroid usage groups. These findings are similar to those of Hinton et al10 and Hendriksen et al5 who found that steroid usage was not significantly associated with behavioral outcomes. While not statistically significant, there were differences in the patterns of behavior seen among the three groups. Overall, the boys not taking steroids were less likely to be in the borderline/clinical range for any of the behavioral subscales. A greater percentage of boys taking prednisone had increased externalizing behaviors in the borderline/clinical range, which include rule-breaking and aggressive behaviors, while boys taking deflazacort demonstrated greater internalizing behavior problems associated with withdrawal and depression. It has been reported that boys with DMD who were using steroids showed less withdrawal than boys who were not using steroids;5,8 however, this is not consistent with the findings of this study. Solden et al20 reported that boys with DMD were more likely to have problems with anxiety-withdrawal than with aggression. In our cohort the behavioral tendencies, although not significant, differed by steroid use. The behavioral differences between the boys taking steroids and those who are not may be a function of age rather than medication, as the boys not on steroids were two to three years younger than boys taking steroids. While problems with behavior regulation, specifically oppositional-defiant and aggressive behaviors21 have been reported in boys with DMD, a greater percentage of boys with DMD demonstrated problems with internalizing behaviors rather than externalizing behaviors, which may be associated with having a chronic degenerative illness.

While self-report is considered the gold standard for measuring quality of life, parents are often asked to provide an assessment of their child’s quality of life. Proxy-reporting is problematic due to the inconsistencies in the level of agreement between the parent and child.22 Similar to the findings in the literature,8,23,24,25 boys with DMD and their parents showed poor to moderate agreement on perceived quality of life.24 Observable symptoms, such as physical functioning are highly consistent between the parent and child; however, internal domains such as school and emotional functioning yield poorer agreement.8 In previous studies, parents of boys with DMD tend to underestimate the health-related quality of life perceived by their sons.22,25 In contrast to Bray et al.,22 the boys with DMD in this study reported poorer quality of life than their parents for emotional and social functioning. The lack of agreement between the boys and their parents reflects the need for self-report questionnaires whenever possible in order to obtain the appropriate reference point.

Families of boys with DMD report that the quality of life and the mental health of their son are of major concern for them.26 Overall, both boys and their parents reported lower quality of life across all quality of life domains as compared to same-aged peers, similar to the results reported by Bendixen et al.11 In contrast to the literature, parents reported a decrease in psychosocial quality of life with increasing age. According to parental proxy data, decreased psychosocial quality of life and increased internalizing behaviors occurred as a function of age. This decrease in psychosocial quality of life may be related to reports in the literature that as boys with DMD get older they tend to become more withdrawn, depressed, and isolated.27 Additionally, as boys with DMD age, they may be physically unable to act out and demonstrate externalizing behaviors such as hitting. Similarly, increased internalizing behaviors such as anxiety, withdrawal and depressive symptoms were associated with decreased quality of life per child report. As emotional factors such as anxiety and depression have been found to contribute to peer problems for boys with DMD,5 this may further affect psychosocial quality of life. Despite the decreased psychosocial quality of life with increasing age reported by the parents, reports from boys with DMD were more consistent with the literature5,8 reporting an increase in psychosocial quality of life with increasing age. It is unclear from this study what factors influence the boys’ perceptions of improved psychosocial quality of life and how these differ from their parents.


The purpose of this study was to examine whether prednisone and deflazacort play a different role in parental reported behavior problems and psychosocial quality of life in boys with DMD, as measured by the CBCL and the PedsQL. The results from this study are based on both parent and child report and may represent some bias. In this study the assessment of emotional and behavior issues by the CBCL was completed by the caregiver only and did not incorporate input from the teacher and or a psychologist, which may bias the results. In situations where the child was too young to read or cognitive ability prevented them from completing the assessment independently, assistance with reading was provided by the researcher. Although the child answered the PedsQL independently by pointing to the appropriate face response, answers may be influenced by researcher intervention. While the subjects from this study were recruited from three different geographical centers, this sample may not be representative of boys with DMD as a group, due to the small number of subjects. Additionally, age may influence the commencement of behavioral problems; however, the sample size in this study prevented the stratification of results by both age and steroid regime. Additionally, the CBCL may not be the optimal tool to use with boys with DMD as it may be overly sensitive to living with a chronic condition and physical disabilities and may over-represent psychosocial maladjustment.5,28 Although steroid grouping was based on parent-selected choice, other medication factors such as compliance, dose, and duration of use were not controlled for in this study. Recently, the role of dystrophin and the disruption of GABAA receptors in the brain have received increasing attention with respect to their impact on behavioral and cognitive dysfunction.29 The higher rates of behavioral problems and the variability of the entire cohort may be a result of the lack of dystrophin. The interaction between the corticosteroids and the GABAA receptors is another factor that may affect the efficacy of these medications.29 In addition, the results of the behavioral assessment may be influenced by the use of concomitant medications. In contrast to the findings of Caspers Conway et al, who reported that 25% received neuropsychiatric medication and 27% received both counseling and medication9 to address behavioral issues, only 3 boys in our study were taking medication to address behavior issues. One boy in the steroid naïve group was taking clonazepam for panic attacks and methylphenidate for ADHD, which may impact the parent perceptions of behavioral problems. Similarly, two boys in the deflazacort group were taking fluoxetine, which may affect scores within the internalizing behavior domains.


Similar to the findings from Caspers Conway et al.9 , our study found that boys with DMD exhibit more behavioral problems than their peers and have a greater number of internalizing problems as compared to externalizing behaviors. As more information about the role of dystrophin in the brain becomes available,29,30 the etiology of neurobehavioral problems in boys with DMD may be better understood. Overall, the use of steroids was not associated with more behavioral problems in boys with DMD; however, slight differences in the types of undesired behaviors were evident between boys taking deflazacort and those taking prednisone; however, the longitudinal data from this natural history study may provide insight changes over time. As behavior played a significant role in psychosocial quality of life, comprehensive assessment and treatment of behavioral problems is crucial in this population.

]]> 0
Investigating Synthetic Oligonucleotide Targeting of Mir31 in Duchenne Muscular Dystrophy Thu, 16 Jun 2016 11:50:33 +0000 Exon-skipping via synthetic antisense oligonucleotides represents one of the most promising potential therapies for Duchenne muscular dystrophy (DMD), yet this approach is highly sequence-specific and thus each oligonucleotide is of benefit to only a subset of patients. The discovery that dystrophin mRNA is subject to translational suppression by the microRNA miR31, and that miR31 is elevated in the muscle of DMD patients, raises the possibility that the same oligonucleotide chemistries employed for exon skipping could be directed toward relieving this translational block. This approach would act synergistically with exon skipping where possible, but by targeting the 3’UTR it would further be of benefit to the many DMD patients who express low levels of in-frame transcript. We here present investigations into the feasibility of combining exon skipping with several different strategies for miR31-modulation, using both in vitro models and the mdx mouse (the classical animal model of DMD), and monitoring effects on dystrophin at the transcriptional and translational level. We show that despite promising results from our cell culture model, our in vivo data failed to demonstrate similarly reproducible enhancement of dystrophin translation, suggesting that miR31-modulation may not be practical under current oligonucleotide approaches. Possible explanations for this disappointing outcome are discussed, along with suggestions for future investigations.



Duchenne muscular dystrophy (DMD) is an X-linked muscle-wasting condition caused by low or absent expression of the muscle protein dystrophin, leaving muscle fibres exquisitely vulnerable to exercise-induced damage (particularly eccentric exercise). The condition is characterized by repeated cycles of muscle degeneration and regeneration, leading to progressive muscle wasting and accumulation of fibrosis and fatty deposits. DMD is invariably fatal, and no current cure exists: existing therapies are chiefly concerned with minimizing inflammatory damage (corticosteroid treatment regimes1), providing respiratory assistance with positive pressure ventilation2, and using drugs to treat the cardiomyopathy3, and do not address the primary defect (insufficient/absent dystrophin protein). While a number of additional avenues are being explored, including anti-fibrotic agents4, promotion of muscle hypertrophy5,6, and modulation of muscle metabolism7, the core focus of research remains the restoration of dystrophin expression.

The dystrophin gene is huge; at around 3Mb (and comprised of 79 exons), this gene occupies roughly 0.1% of the entire human genome. This gene is transcribed and spliced into an mRNA roughly 14,000 bases in length, and ultimately translated to a protein 427kDa in size. The dystrophin protein has a barbell-like structure, with the actin-binding N-terminus and the dystroglycan/nNOS binding C-terminus linked by an extended stretch of 24 spectrin-like repeats8,9. Becker’s muscular dystrophy (BMD), in most cases a considerably milder -and sometimes largely asymptomatic- dystrophic condition, typically results from mutations causing deletions of this internal repeat region: all crucially retaining the reading frame. While several functional elements are located within this linker region (including an additional actin-binding domain10 and an nNOS localisation motif11), as long as functional N and C terminal domains are expressed and remain linked, the full extent of the central linking domain is not absolutely critical to dystrophin function. This observation underpins many dystrophin-restoring therapies where restoring full-length dystrophin would otherwise be technically prohibitive: from microdystrophin therapies (plasmid or viral delivery of a substantially internally-truncated dystrophin)12,13,14 to exon-skipping approaches (using short synthetic antisense oligonucleotides to alter splicing patterns, ‘skipping’ exons to restore reading frame, generating an internally-deleted but functional dystrophin)15,16,17,18,19. Candidates from this latter category, using both 2-O’methyl phosphorothioate and phosphorodiamidate morpholino oligomers (2Ome and PMO, respectively) are currently at the late clinical trial stage, though the efficacy of this approach is limited by the extent of oligonucleotide delivery, and by the requirement for sequence-specific targeting. A wide range of mutations to the dystrophin gene can result in a DMD phenotype, necessitating skipping of a range of specific exons. Even the current generation of oligos, designed to restore the reading frame to multiple mutations in a “hot spot” on the dystrophin gene by excluding exon 51, are expected to be of use to only 13% of DMD patients20. Therapies capable of enhancing dystrophin expression in a less mutation-specific fashion would thus be of considerable advantage.

The discovery that the translation of dystrophin transcripts is subject to negative regulation by the microRNA miR31, and that this microRNA is dysregulated in dystrophic muscle21,22,23, suggested that inhibition of miR31 might constitute a therapeutic approach.

MicroRNAs (miRs) are produced as ca. 200 nucleotide “pri-microRNAs” which undergo several discrete processing steps to generate first “pre-microRNAs”, and ultimately mature microRNAs (see schematic: fig 1 A)24.


Fig. 1: miR processing, canonical mode of action and miR-modulation scheme

MicroRNAs are transcribed as pri-miRs, processed to pre-miRs in the nucleus and finally processed to mature miRs in the cytosol, remaining associated with several proteins in a silencing complex (A). miR-mediated translational suppression requires interaction of microRNA with the 3’UTR of transcripts (B), and can potentially be inhibited by sequestration of miR (Sponge, C), masking of the binding site on 3’UTR (protector, D) or by interfering with miR maturation (analogue, E).

MiR31 is highly expressed in satellite cells (the muscle stem cell population) where under healthy physiological conditions it is believed to suppress translation of transcripts for early myogenic factors such as myf525, allowing relatively high levels of mRNA to accumulate without eliciting a commitment to the myogenic program. Following satellite cell activation, levels of miR31 fall as myogenesis progresses, in a process believed to be nitrosylation-sensitive26: localisation of nNOS to the sarcolemma by dystrophin alters nuclear nitrosylation patterns, lowering miR31 expression and thus enhancing dystrophin levels, potentially forming a feed-forward loop generating a switch-like commitment to terminal myogenesis. In the absence of sufficient dystrophin no loop can be established and cells remain trapped at a late but incomplete stage of differentiation.

Blocking the action of miR31 would therefore be expected to increase translational competence of any given dystrophin transcript, giving effectively greater protein expression per mRNA molecule. Moreover, if the feed-forward hypothesis is correct, enhancement of translation might exhibit threshold effects: above a certain level of dystrophin the suppression of miR31 may become self-sustaining. A final appealing aspect of this strategy is that the binding site for miR31 resides in the 3’UTR of the dystrophin transcript and is not associated with reading frame or coding mutations. It should thus be common to all patients, and a miR31-based therapy would be synergistic with exon-skipping approaches regardless of the targeting of the skipping oligo.

Work by Cacchiarelli et al21 showed that modulation of miR31 activity, either by a “sponge” approach (using an mRNA carrying multiple partially-complementary miR31 binding sites to sequester active miR31 without being targeted for degradation) or by “protector” approach (using an antisense locked nucleic acid –LNA- to mask the miR31 binding sequence on the dystrophin transcript, preventing interaction of miR31 with this mRNA) enhanced expression of dystrophin at the protein level in dystrophic cells in culture, suggesting that this represents a viable therapeutic avenue.

In this study we therefore determined to investigate whether blockade of miR31 could be employed synergistically with exon-skipping, using PMO oligonucleotide chemistries already approved for use in clinical trials. Using an animal model of DMD, the mdx mouse (which carries a premature stop codon in dystrophin exon 23), and a dystrophic mouse cell culture line carrying the mdx mutation (H2KSF127), we combined the well-established mdx-specific skipping oligo (M23D)28 with PMOs representing “sponge”, “protector” and “miR31-analogue” sequences (schematic: Fig 1 B-E, sequences and hybridization scheme in supplementary fig S1).

As inhibition of miR31 would be expected to increase total “protein per transcript”, we measured (to high precision) the levels of both successfully-skipped dystrophin mRNA and of dystrophin protein, thus expressing efficacy of miR31 modulation as a protein:mRNA ratio

Materials and methods

Mdx mice were on a C57BL10ScSn background. Animal experiments were carried out under license from the Home Office (UK) in accordance with The Animals (Scientific Procedures) Act 1986 and were approved by the Royal Veterinary College ethics and welfare committee. Mice were housed in a minimal disease facility in cages of 3-5 individuals, with food and water ad libitum.

All reagents were purchased from Sigma/Fluka unless indicated otherwise.

Cell Culture

The H2KSF1 cell line (derived from an mdx-immortomouse cross) were grown in matrigel-coated flasks ( and maintained in a proliferative state by incubation at 33°C in proliferation medium: DMEM+glutamax (invitrogen) supplemented with 20% (v/v) heat inactivated-foetal bovine serum, 2% (v/v) chicken embryo extract (CEE, Sera laboratories international), 1% (v/v) Penicillin/Streptomycin (Sigma, final concentration penicillin, streptomycin), and 20 U/mL γ-IFN (Chemicon).

One day prior to differentiation, cells were seeded onto matrigel-coated 6-well plates at 5×105 cells.well-1; plating conditions established to allow optimal differentiation without under- or overcrowding cells. Differentiation was initiated by replacement of proliferation media with differentiation medium (DMEM+glutamax supplemented with 5% horse serum (PAA) and 1% pen/strep) and incubation at 37°C. Differentiation medium was partially replaced (50% of medium aspirated, replaced with fresh differentiation medium) after 2 days of differentiation.

For naked PMO/Vivo-Morpholino treatments, oligos (1.3uM final) were added directly to the culture medium 48 to 52 hours after initiation of differentiation, and gently swirled to mix. For staggered administrations additional oligos were added after a further 24 or 48 hours. No further media changes were performed after oligo addition.

For nucleofections, 1×106 proliferating cells were incubated in 100ul nucleofection solution supplemented with PMOs as indicated and nucleofected using protocol B32 (using a nucleofector 2B –Lonza), before dilution into 2ml proliferation media and seeding onto a 6-well plate. Cells were subsequently differentiated as described above.

Both protein and RNA were isolated from the same culture wells using the following protocol: culture media was aspirated gently to avoid detaching myotubes, and cells were lysed by addition of 100ul chilled RIPA buffer (50mM Tris pH8, 150mM NaCl, 1% NP40, 0.5% sodium deoxycholate, 0.1% SDS, supplemented with protease inhibitors –complete, Roche) then detached rapidly by agitation with a cell scraper before removal of a 50ul aliquot directly into 100ul of TRIzol reagent for RNA extraction. The remaining 50ul RIPA lysate was maintained on ice for 10min before centrifugation in a benchtop microfuge to remove insoluble myofibrillar material. Both RNA and protein isolates were stored at -80°C until needed.

In vivo studies

mdx mice were treated with morpholinos (in saline solution) as follows:

PMO: 5ug M23D skipping oligo either alone or supplemented with protector, 31 analogue (5ug) or sponge (20ug), in a final volume of 20ul, was delivered via a single direct intramuscular (IM) injection into the tibialis anterior (TA) muscle.

Vivo-Morpholino: M23D Vivo-Morpholino was delivered by a single intravenous (IV) injection into the tail vein at (ca. 200ug.mouse-1, 180-200ul final volume), or via a single direct IM injection to the TA muscle at 10ug.mouse-1 (20ul final volume), or via a single intraperitoneal (IP) injection ( M23D either alone or combined with protector or scrambled Vivo-Morpholino, 180-200ul final volume).

All IM injections were performed under hypnorm/hypnoval anaesthesia and IV injections under isoflurane. Anaesthetised mice were allowed to recover in a heated chamber. All mice treated intravenously with Vivo-Morpholinos died within minutes of injection (see results) thus this treatment was discontinued.

Two weeks post-injection mice were killed by cervical dislocation and tissues collected rapidly (as described below) for histology and RNA/protein analysis.

Vivo-Morpholino IP-treated mice: diaphragms and body wall sections were removed and halved: one half used for histology, the other for RNA/protein isolation. Histological samples were rolled up and mounted vertically on cork blocks, coated in cryogenic mounting medium (cryo-M-bed, bright) and frozen rapidly in liquid nitrogen-cooled isopentane to preserve tissue morphology. Tissues were cryosectioned into 10 micron sections and mounted on glass slides (Superfrost, fisher scientific), with serial sections collected and mounted every 500um through the sample, permitting assessment of the entire tissue bulk. For RNA/protein isolation, samples were flash frozen and pulverised under liquid nitrogen.

PMO IM-treated mice: injected TA muscles were mounted vertically on cork blocks free of cryogenic mounting and frozen under liquid nitrogen-cooled isopentane and serially cryosectioned as above. The omission of mounting medium allows intervening sections to be collected and used for RNA/protein analysis.


Muscle cryosections (10 micron thickness) were stained for dystrophin essentially as described by Wells et al19, using antibodies raised to the dystrophin C-terminal domain (rabbit polyclonal DysC3750, Cymbus Biotechnology, 1:500) with biotinylated secondary antibodies (biotinylated goat anti-rabbit, DAKO, 1:500). Signal was developed using the ABC kit system (VECTASTAIN) with diaminobenzedine (SIGMAFAST) for 3 mins per slide. For IM-injected muscle, dystrophin-positive fibres were counted, while for Vivo-Morpholino treated muscle samples were assessed by eye and densitometry (using ImageJ). In both cases samples were assessed blind to sample identity.

Western Blotting

RIPA-solubilized cell cultures and muscle sections (typically 30-40 10um sections) were used for western blotting essentially as described by Godfrey et al29, using hand-cast 4% acrylamide gels and slow wet-blotting onto PVDF membranes. Blots were probed with anti-dystrophin antibody 6D3 (millipore, 1/200 for tissue sections, 1/100 for cell culture) and anti-vinculin h-Vin1 (sigma, 1/100,000 for tissue sections, 1/250,000 for cell culture), using anti-mouse HRP (BioRad, 1/100,000) as secondary. Blots were developed using ECL prime (Amersham).

RNA extraction & cDNA synthesis

RNA extractions were performed as described previously30 using TRIzol. Tissue sections and frozen tissue powder were dissolved in an appropriate volume of TRIzol reagent, while cell culture samples were added to TRIzol as above. cDNA was prepared from 0.8-1ug of RNA, using the RTnanoscript kit (PrimerDesign) with oligo dT and random priming.

Quantitative PCR

Primers sets for dystrophin were designed using Primer3 ( PCR products were typically 80-200bp in length, and designed to span one or more introns (where possible) to prevent amplification of genomic DNA. Primers were designed to span exons 1-3 (total dystrophin) and exons 22-23 (unskipped dystrophin). Primers for skipped dystrophin were designed to bind the novel exon 22:24 junction, and exon 25.







Primers to Csnk2a2, Fbxw32, Ap3d1 and GAPDH form part of the geNormPlus kit, and are proprietary property of PrimerDesign.

qPCR runs were performed using Precision SYBR green mastermix (Primerdesign) in white hard-shell 384-well plates (BioRad) using 5-10ng of cDNA per well in a BioRad CFX384 (30ng for dystrophin in cell culture). Cq values were converted to relative quantities (or transcript numbers as described below) and normalized to the geometric mean of 2-3 suitable reference genes (determined by geNorm and Normfinder as described previously30); Cell culture: Csnk2a2, Fbxw32 and Ap3d1, Muscle tissue: Csnk2a2, Ap3d1 and GAPDH. Skipping percentages were derived by estimation of absolute transcript numbers, via qPCR analysis using constant amounts of sample cDNAs (or water) mixed with a standard dilution series (107-100 molecules.well-1) prepared from purified PCR product of known concentration. As cDNA is predominantly single-stranded an additional cycle is required for parity with PCR products, thus the concentration (in molecules.well-1 of standard) at which the dilution series (+cDNA) plateaus corresponds to approximately half the cDNA concentration


Intramuscular injection of naked PMO

IM injection of PMO into the tibialis anterior muscle is a well-established technique for establishing efficacy of targeted exon skipping oligos, and with the M23D PMO this approach routinely generates significant levels of skipped dystrophin mRNA and concomitant dystrophin protein, primarily localized around the injection site. We reasoned that IM injection of skipping oligo alone or in combination with miR31-modulating oligos (at either a 1:1 ratio for ‘protector’ and ‘miR31-analogue’, or 4-fold excess for ‘sponge’) would constitute an appropriate first-pass strategy for establishing the feasibility of a morpholino-mediated microRNA-targeting therapy for DMD.

Initial data encouragingly suggested that addition of ‘sponge’ oligo (skip+sponge) elicited enhanced dystrophin protein (compare fig 2A and B, fig 2C), moreover doing so despite also apparently lowering skipping frequency: exhibiting a pronounced protein/transcript ratio as a consequence (fig 2D). However individual animal-to-animal variation was substantial, thus this increase did not reach significance (indeed, the promising performance of the sponge oligo derived from only two of the six animals in the skip+sponge group). As mean levels of skipped transcript were lower in all combination groups (Fig 2D, top panel), and dramatically so in sponge-treated animals (where sponge oligo was present in a 4:1 ratio with skipping oligo) we reasoned that presence of additional (non-skipping) oligo might compete with skipping oligo for myofibre entry. Efficacy of exon skipping by IM injection is innately variable -being highly dependent on precise location of the needle- and by adding oligo competition we effectively compounded this variability. We thus decided to pursue further studies using our cell culture-based system, reasoning that these would allow us to optimize our treatment protocols more efficiently and cost-effectively.


Fig. 2: Intramuscular injection of exon skipping and miR31-modulating oligos

Left panels: Immunostaining for dystrophin restoration in tibialis anterior (TA) muscles injected with M23D skipping PMO alone (A) or with skipping PMO combined with a 4-fold excess of ‘sponge’ PMO (B). Right panel: quantification of dystrophin restoration in injected muscles. Typical western blot for dystrophin restoration in TA muscles treated as indicated (C), qPCR for skipped transcripts, densitometry of western blots for dystrophin protein, and resultant protein.mRNA-1 ratio for the treatment groups indicated (D). Means + SEM (N=6). AU: arbitrary units. Dystrophin level normalized to vinculin (protein) or to the geometric mean of three reference genes (mRNA); see methods.

Exon skipping in cell culture

The authors note that establishment of a working cell culture model for measurement of exon skipping at both the transcript and protein level was a non-trivial exercise: for the benefit of researchers wishing to employ a similar system, our methodologies (including unsuccessful nucleofection-based approaches) are below described in full.


Nucleofection of cells with M23D PMO resulted in detectable (but low) levels of skipped transcript (supplementary fig S2), though these levels were insufficient to generate detectable dystrophin at the protein level. Furthermore, co-nucleofection with miR31-modulating oligos reduced skipping levels, presumably again reflecting competition for oligo entry into cells. As nucleofection can only effectively be performed once, and cannot be employed in differentiated myotubes, this approach was abandoned.

Naked PMO

Treatment with naked PMO (by simple addition of unconjugated morpholino to the culture medium) in contrast produced skipped transcripts in a clear concentration-dependent fashion, achieving, at higher concentrations, sufficient skipped dystrophin mRNA to generate detectable protein (supplementary fig S2). We have previously shown that dystrophin expression in our cell culture model commences between 50 and 70 hours of differentiation in both dystrophic and wild type cells30, thus addition of skipping oligo at ca.50 hours post-differentiation ensures prompt generation of skipped transcripts. Dystrophin is readily detectable at the protein level in wild-type cultures after 90-100 hours (this delay presumably partly reflecting the lengthy 16 hour transcription of dystrophin31). In our mdx cell culture model dystrophin protein was not detected until at least 120 hours of differentiation, despite ready detection of skipped transcript at earlier timepoints. This suggests that detectable protein represents the steady accumulation of low levels of highly-stable protein rather than delayed translation. As myotube cultures exceeding 120-144 hours of differentiation demonstrate increasing levels of spontaneous contractile behaviour (resulting in myotube detachment from the culture substrate and thus constituting a necessary experimental endpoint) our data suggests a roughly 48-72 hour window (from dystrophin transcription/skipping to myotube detachment) over which miR31 modulation could be effective.

miR31-modulation in cell culture

This above assessment of the timing of dystrophin expression, taken together with the observation that oligonucleotides appear to compete for entry, was used to devise an experimental approach allowing investigation the effects of varying the ratio of miR31-modulating oligos to skipping oligos (either 1:1 or 4:1) and of varying time of delivery of miR31 oligos (either simultaneously with skipping oligo or 24/48 hours afterward: see schematic, supplementary fig S3): miR31 modulation would be expected to demonstrate obligate dependence on exon skipping, thus can potentially be delayed until sufficient skipped transcripts have been generated.

As shown in fig 3, only simultaneous addition of protector oligo at 1:1 produced significant increases in protein:mRNA ratio over skipping oligo alone (compare fig 3A with 3B and 3C), with all other treatments being equivalent to skip alone, or eliciting a reduction in ratio. Note however regardless of time of addition, presence of miR31-modulating oligos resulted in lower overall levels of skipped transcript in both protector and 31-analogue (with higher levels in sponge-treated samples likely stemming from increased dystrophin transcription –see below). Our data nevertheless suggests a biological role for all three modulating oligos, as treatment with sponge oligo consistently raised overall dystrophin transcription levels, without demonstrable effect on protein/mRNA ratio (supplementary fig S4), while miR31 oligo unexpectedly appears to lower dystrophin translation, leading to significant decrease in protein:mRNA ratio (fig 3C).


Fig. 3: miR31-modulation in a cell culture model of exon skipping

Skipped dystrophin transcripts (top), dystrophin protein (middle) and resultant protein.mRNA-1 ratio (bottom) in dystrophic myotube cultures carrying the mdx mutation (H2KSF1), either untreated (ntx), incubated with M23D skipping oligo alone (skip), or with M23D combined with miR31-modulating oligos (as indicated) added either simultaneously (T0), 24 hours after M23D addition (T24) or 48 hours after (T48). miR31-modulating oligos were added at 1:1 or 4:1 ratios with skipping oligo as indicated. Means + SEM (N=3) *=P<0.05, **=P<0.01. AU: arbitrary units. Dystrophin level normalized to vinculin (protein) or to the geometric mean of three reference genes (mRNA); see methods.

miR31-modulation via Vivo-Morpholino

The cell culture data above suggested that a simultaneous 1:1 skip+protector combination constituted a viable approach: as the most dystrophin-specific modulating oligo, protector-based strategies are also the most therapeutically attractive. It was therefore decided to take this approach forward to in vivo study in the mdx mouse. The high variability of our pilot in vivo investigation (intramuscular injection of naked PMO) demonstrated that this local delivery method is inappropriate, therefore for these further studies we decided to employ systemic delivery using Vivo-Morpholinos. Unexpectedly, intravenous delivery of M23D skipping Vivo-Morpholino proved to be universally fatal, with all treated mice apparently beginning to recover normally from isoflurane anaesthesia before abruptly becoming unresponsive (see discussion).

Our Vivo-Morpholino studies therefore utilized an intraperitoneal delivery route. Our preliminary trial using IP delivery of vivo-skipping oligo demonstrate relatively high levels of skipped transcript and protein in muscle tissues exposed to the peritoneal cavity, namely the diaphragm and the body wall musculature, therefore these tissues were selected for analysis. Skipped transcripts were detectable in other muscle tissues studied (quadriceps, TA, triceps, EDL, soleus) but were 1-2 orders of magnitude lower than levels in diaphragms and body walls, and did not result in demonstrable dystrophin protein (supplementary fig S5).

Mice were thus treated via IP injection with skipping oligo alone, or in 1:1 combination with protector oligo or a ‘scrambled’ protector sequence. As shown in fig 4A, greater quantities of skipped transcript were achieved (and achieved more consistently across treatment groups) in the diaphragm than in the body wall, though in contrast to our naked PMO data, the presence of additional non-skipping oligo -protector or scramble- unexpectedly appeared to potentiate uptake (albeit variably) in body wall. Expression of dystrophin is 30-50% higher in dystrophic diaphragm than body wall (supplementary fig S6) thus the higher levels of skipping achieved in diaphragms could conceivably be simply due to higher levels of transcripts available for skipping. However, even when expressed as a percentage of total dystrophin, diaphragm skipping percentages are consistently around 20% across treatment groups, while body walls treated with skipping oligo alone remain around 10% (as in our pilot study, supplementary fig S5), increasing to roughly 20% in the presence of additional non-skipping oligo (table 1). This phenomenon was also observed with combined Vivo-Morpholinos in cell culture (supplementary fig S7).


Fig. 4: Intraperitoneal delivery of exon skipping and miR31-modulating Vivo-Morpholinos to diaphragms and body wall musculature

Skipped dystrophin transcripts (A), dystrophin protein (B) and resultant protein.mRNA-1 ratio (C) in diaphragm (top) and body wall (bottom) musculature, treated with Vivo-Morpholinos (as indicated). NTX: no treatment; Skip only: M23D skipping oligo alone; Skip Pro: M23D combined with protector; Skip Scram: M23D combined with scrambled protector sequence. Means + SEM (N=6). AU: arbitrary units. Dystrophin level normalized to vinculin (protein) or to the geometric mean of three reference genes (mRNA); see methods.

Table 1: Skipping percentages following Vivo-Morpholino treatment

Skipped dystrophin transcripts as a percentage of total dystrophin transcripts in diaphragms and body walls from mdx mice treated with Vivo-Morpholinos (as indicated). NTX: no treatment; Skip only: M23D skipping oligo alone; skip+pro: M23D combined with protector; skip+scram: M23D combined with scrambled protector sequence. Means + SEM (N=6 –same sample groups as in figure 4)

Body wall (%) Diaphragm (%)
NTX 0.43±0.1 1.1±0.1
Skip Only 8.9±1.3 21.6±2.5
Skip + Pro 18.0±3.9 21.8±1.8
Skip + Scram 20.5±5.3 24.0±1.9

Note that, as expected, the generation of stable (skipped) mature dystrophin transcripts leads to a mild increase in total dystrophin transcripts in both tissues in treated animals (supplementary fig S6).

When examined at the histological level for presence of sarcolemmal dystrophin (figs 5 and 6), all three treatment groups were readily distinguished from untreated muscle, but could not be further separated into individual treatments: skipping oligo alone or combined with protector/scrambled oligo produced comparable quantities of widely-distributed dystrophin-positive fibres. While overall levels varied somewhat from animal to animal this was not associated with any particular treatment. Western blotting analysis revealed dystrophin levels (normalized to the muscle membrane marker vinculin) similarly varied substantially from sample to sample (fig 4B), and no significant differences were observed between treatment groups in diaphragm dystrophin levels. In body walls both scramble- and protector-treated tissue (mirroring the higher levels of successfully skipped mRNA described above) showed higher mean protein than “skip only” samples: an intriguing result but clearly not one restricted to (or dependent upon) “protector” sequence.


Fig. 5: Dystrophin expression in body walls following intraperitoneal delivery of exon skipping and miR31-modulating Vivo-Morpholinos

Immunostaining for dystrophin restoration in body wall cryosections treated as indicated. NTX: no treatment; SKIP ONLY: M23D skipping oligo alone; SKIP+PRO: M23D combined with protector; SKIP+SCRAM: M23D combined with scrambled protector sequence. Note that body wall architecture has two closely-associated muscle groups lying perpendicular to each other: Peritoneum-proximal tissue is transversely sectioned here, while distal tissue is longitudinal.


Fig. 6: Dystrophin expression in diaphragms following intraperitoneal delivery of exon skipping and miR31-modulating Vivo-Morpholinos

Immunostaining for dystrophin restoration in diaphragm cryosections treated as indicated. NTX: no treatment; SKIP ONLY: M23D skipping oligo alone; SKIP+PRO: M23D combined with protector; SKIP+SCRAM: M23D combined with scrambled protector sequence.

Ultimately (as shown in fig 4C) when assessed as a ratio of protein (as determined by western blot) to mRNA (as determined by qPCR), no significant differences were detected between treatments in either tissue studied.


A therapeutic agent capable of driving increased dystrophin translation would be a highly attractive discovery: acting at the post-transcriptional level, such an agent could theoretically act synergistically with all exon skipping approaches, regardless of the specific mutations in question (and indeed also be of benefit to patients producing viable but insufficient levels of dystrophin mRNA). If such an agent could be employed using existing oligonucleotide chemistry for which a substantial body of toxicology and pharmacokinetic data already exists, this would greatly expedite transition from a laboratory context to a therapeutic one. For these reasons we decided to build on the work of Cacchiarelli et al 21 and investigate the potential of modulating the microRNA miR31 using morpholino oligomers in a dystrophic context. As we here show (despite encouraging data from our cell culture model) under the conditions selected, modulation of miR31 via PMO appears to be of no significant benefit to our in vivo model, the mdx mouse. Were it not for our cell culture data, it would be tempting to take the parsimonious interpretation: that morpholino-based modulation of miR31 simply does not work. It is important to recognize, however, that our in vivo analysis is not exhaustive: due to practical restrictions, large scale studies similar to those used in cell culture (multiple doses, staggered administrations) were not carried out in our animal model, and the unexpected lethality of IV-administered Vivo-Morpholino precluded investigation of miR31 modulation in a more systemic context. This work also necessarily operated under several significant restrictions, which we discuss below.

The primary difficulty with this work is oligonucleotide entry: as suggested by both intramuscular injections and cell culture treatments with nucleofected PMO, oligo uptake appears to be both comparatively low and moreover, readily saturable: inclusion of additional oligonucleotides alongside the skipping oligo M23D is here repeatedly shown to lower overall levels of exon skipping. As levels of skipped transcripts otherwise correlate well with M23D concentration in cell culture (supplementary figure S2) the logical interpretation is that multiple oligonucleotides compete for cell entry, most likely via an endocytotic process32,33. As a consequence of this competition, levels of skipped transcript and subsequent dystrophin protein as measured here were frequently lower following co-treatment with miR31-modulating oligos, hence our assessment of miR31 modulation by a ratiometric (protein per transcript) rather than an absolute (total protein) approach. While the authors acknowledge that this competitive effect could itself be considered to present a significant barrier to therapies based on multiple oligos, such direct competition could presumably be readily circumvented by staggered oligo administration. Moreover, as much of the focus in oligonucleotide therapies is now moving toward enhancing cellular uptake of these molecules, saturation of oligo uptake (carrier-mediated or not) is unlikely to constitute a long-term impediment.

An additional complication with assessing the potential of oligonucleotide therapies is that as small uncharged (and unlabelled) molecules, oligo uptake is non-trivial to quantify. Uptake of skipping oligo can be assessed -at least in relative terms- by measuring levels of exon skipping; however within this study the extent of uptake of miR31-modulating oligos is essentially unknown. As demonstrated in our pilot IM injection studies, uptake is also inherently somewhat variable: the combination of two variable and mutually-competitive oligonucleotide uptakes, only one being (relatively) quantifiable, presents a significant barrier to data interpretation, and highlights the importance of sufficient sample numbers.

Application of naked PMO directly to differentiating myotubes in a cell culture system allows us to partially circumvent the issue of oligonucleotide competition: unlike in vivo scenarios where oligo not sequestered in cells is rapidly cleared via the circulatory system, here oligos remain trapped in the culture medium, increasing the effective dose delivered. If uptake is indeed carrier-mediated, this should allow intracellular oligonucleotide to reach equilibrium with the external media, regardless of total oligo concentration or number of different oligo species used. Under such a system, staggered administration of oligos would more accurately expected to represent a “head-start” for exon skipping. Unexpectedly, oligo combinations (both co-administered and delivered in a staggered fashion) consistently resulted in lower levels of skipped transcripts: either indicating that equilibrium had not been achieved even over the lengthy course of the experiment (implying a very slow rate of carrier-mediated uptake), or suggesting the presence of additional intracellular processes for which oligos compete (for example, subcellular trafficking –a non-trivial concern in the highly-structured intracellular environment of multinucleated myotubes). These observations accepted, our cell culture data suggests that all three miR31-modulating oligos are able to exert some biological effect (albeit not always a beneficial one). While having no apparent effect on protein:mRNA ratio, ‘Sponge’ oligo nevertheless appears to elicit a minor (but consistent) increase in dystrophin transcription (supplementary fig S4) –the most plausible explanation being that the ‘sponge’ sequence, as a direct miR31-chelating agent, influences all miR31-associated processes including those moderating commitment to, and progression through, myogenesis (such as translation of Myf525). In effect, myotubes from cultures treated with ‘sponge’ may proceed along the myogenic program at a moderately accelerated rate. While this could be considered to be of some small therapeutic use, miR31 is also implicated in the regulation of a large number of cellular processes throughout the body, including tumour suppression34: as a global miR31-targeting agent, a ‘sponge’ approach would be expected to carry significant side-effect burden, obviating use of this strategy outside of a commensurate level of therapeutic benefit.

Intriguingly, our miR31 analogue appears to actively reduce dystrophin expression at the protein level (fig 3C): we presently are unable to explain this phenomenon: a morpholino corresponding to the miR31 sequence should be incapable of forming a viable repressor complex with Dicer and Argonaute proteins, yet still be capable of interfering with miR processing (as outlined in Fig1). The reverse complement of miR31 (miR31*) may possess biological activity (though data supporting this is minimal35), under which scenario our morpholino might be expected to act as a miR31* ‘sponge’ (with the added complication of potentially sequestering degradation machinery due to full sequence complementarity). However, a reduction in dystrophin transcription as a consequence of miR31* blockade would imply diametrically opposed roles for miR31 and its complement, which seems biologically implausible.

Recently published work (as with Cacchiarelli et al21) from the group of Irene Bozzoni36 however suggests a more complex role for the miR31 locus, with this region generating both miR31 and a long non-coding RNA (lnc-mir31) via mutually exclusive pathways. While both miR31 and lnc-miR31 are proposed to play roles in control of myogenic commitment and progression, our miR31 analogue PMO might perturb the ratio of these two ncRNAs, with significantly reduced dystrophin translation as a consequence.

Finally, our cell-culture system revealed that ‘protector’ PMO exerted no effect on overall dystrophin transcription (as expected), instead slightly reducing levels of skipped transcript (as common to all three miR31-modulating oligos) but also reproducibly increasing the yield of protein per transcript when delivered simultaneously at a 1:1 ratio with skipping oligo. Failure to observe similar increases when using higher ratios (4:1) or when delivered 24 or 48 hours after skipping oligo are troubling, but may reflect the delicate balance of opposed factors involved: earlier addition (or higher concentration) of ‘protector’ oligo reduces total skipped transcripts, while later addition of ‘protector’ limits total exposure of skipped transcripts to this oligo before the necessary experimental endpoint (onset of spontaneous myotube contraction). Our data thus suggested that ‘protector’ based strategies might be effective in enhancing dystrophin translation, presenting a viable avenue to take forward to further investigation.

As described above, for our final in vivo study we elected to compare efficacy of skipping oligo alone versus skipping oligo combined with either protector oligo or a scrambled protector sequence (see supplementary fig S1). We reasoned that inclusion of a scrambled sequence should more readily allow identification of protector-specific effects even if overall skipping levels were lowered. We further reasoned that use of Vivo-Morpholinos might circumvent many of the oligo uptake concerns described above. These commercially-available morpholinos carry an octa-guanidine dendrimer, and this moiety confers independently cell-penetrant properties on the conjugated morpholino, circumventing saturation of slow carrier-mediated uptake processes and moreover permitting systemic delivery. As described here, however, systemic intravenous delivery of a Vivo-Morpholino of the M23D skipping sequence was invariably fatal within minutes, with a crude post-mortem suggesting systemic blood clotting as the cause. Why this should be the case is not clear: although sufficiently hepatotoxic to be inappropriate in a human therapeutic context, intravenous use of these cell-penetrant dendrimer-conjugated oligos as exon-skipping agents has previously been shown to drive systemic restoration of dystrophin in the mdx mouse37,38. Similar fatalities reported by Ferguson et al the same month our studies were performed39 led those authors to propose that oligo dimerisation may play a role, effectively doubling the local concentration of the dendrimer moiety, resulting in adverse interactions with clotting factors. The potential for self-dimer formation in M23D is limited however, and attempts to disrupt any possible multimers prior to injection (sonication, vortexing, heating and crash-cooling) had no effect on the toxicity of this treatment. Moreover, as noted the M23D sequence has previously been shown to be highly effective when employed as a Vivo-Morpholino, thus our experiences are puzzling. Given these findings, this approach was abandoned (it was judged unlikely that combination of M23D with miR31-modulating Vivo-Morpholinos would alter the lethality of intravenous delivery), and the authors would advise caution to any other investigators considering systemic use of Vivo-Morpholinos. Our alternative strategy (intraperitoneal administration) was well-tolerated however, resulting in widespread dystrophin restoration in the diaphragm muscle (fig 6), and adequate restoration in peritoneum-proximal body wall musculature (though as noted, our data suggests systemic penetrance of oligo was poor: dystrophin protein was effectively absent in limb muscle tissues, and limited even in distal body wall: fig 5).

Skipped transcripts and skipping percentages (skipped transcripts as a fraction of total dystrophin transcripts) were relatively consistent in diaphragms regardless of treatment (ca. 20%, see table 1), while levels in body walls were lower and more variable. Interestingly, co-administration of protector or scrambled sequence both resulted in higher mean levels of skipped transcripts in body walls, effectively the converse of our observations with naked PMO. Given mouse posture with respect to the peritoneal cavity, IP-administered morpholino might reasonably be expected to pool proximal to the body wall, thus duration of exposure to Vivo-Morpholino may be greater in this tissue. It is possible that the charged octa-guanidine conjugates act on membranes in a concentration-dependent fashion: under prolonged exposure to Vivo-Morpholino, presence of additional dendrimer moieties may facilitate uptake of PMO regardless of conjugated oligomer sequence. Supporting this hypothesis, incubation of cultured myotubes with Vivo-Morpholinos (analogous to proximal pooling of oligo) exhibited the same phenomenon: M23D combined with either protector or scrambled sequence at 500nM oligo consistently led to higher skipping levels than M23D alone (supplementary fig S7). The authors note that overall levels of skipping were lower and more variable in body walls, however, thus this in vivo observation may simply be an artefact. Unlike diaphragms, where the entire muscle bulk could be readily identified and collected, the body wall strips used for this analysis represent only a portion of the overall body wall musculature. While all effort was taken to collect a consistent quantity of tissue from the same region of the body wall, slight variations are essentially unavoidable.

As described above, the chief findings of our Vivo-Morpholino work were that (despite variable levels of skipping and protein expression) no significant differences in protein:mRNA ratios were observed in either muscle tissue in any of our treatment groups. These findings are disappointing: uptake of M23D skipping oligo appears to be reasonably high and unperturbed (indeed possibly even potentiated) by co-administration of protector or scrambled oligos, thus it seems reasonable to assume that both protector and scrambled oligos were taken up to a similar extent. In terms of minimizing the caveats listed herein this approach appears highly successful, yet reveals our protector-based strategy to be of no apparent benefit whatsoever under these conditions. It should be noted that while exon skipping works at the level of transcript splicing, a protector approach necessarily applies at the translational level. One molecule of skipping oligo can act repeatedly, splicing multiple transcripts; conversely, protector binds 1:1 with skipped transcripts thus one molecule of oligo would presumably protect only one transcript from miR31-mediated translational blockade at a time. Taken together with our cell culture data (where substantially lower levels of transcripts are involved), the failure of this approach at the in vivo level suggests that a level of protector oligo sufficient for efficacy in the mdx mouse was not achieved. As noted, due to time and resource restrictions (and the unexpected lethality of systemic administration) we were unable to perform more comprehensive investigations in our mouse model, such as a staggered administration approach. An attractive line for future studies would be use of the mdx3Cv mouse: this mouse exhibits a low (and consistent) level of endogenous skipping40,41, thus would allow study of miR31-blocking oligos in isolation, free from the additional variable of skipping efficiency.

In conclusion, it could reasonably be argued that if higher oligo doses are required for miR31-modulation, such doses might well be better reserved for exon skipping in the first place. While miR31-modulation potentially offers a universally-applicable therapeutic for DMD, the levels of oligo ostensibly required (as shown here) would unfortunately seem to render this approach impractical under current oligonucleotide chemistries. A recent study (published after the work described here) further showed that several microRNAs may jointly contribute to suppression of dystrophin translation in BMD, including miR146b and miR374a22. MiR31-modulation alone may thus be insufficient in vivo, necessitating PMO-mediated modulation of two or more additional miRs to relieve the translational block. Given that achieving effective delivery of even a single species of oligonucleotide is challenging, such a requirement would further argue strongly against the feasibility of this approach.


Sup Fig S1

Supplementary figure S1: Oligonucleotide sequences and hybridization scheme. Sequences for skipping oligo (M23D), miR31, ‘sponge’ and ‘protector’ oligos (miR31 analogue oligo is simply miR31 sequence, substituting T for U, as morpholinos do not use uracil)

Sup Fig S2

Supplementary figure S2: Exon skipping in cell culture Levels of skipping transcript detected in cells differentiated after nucleofection with M23D and miR31-modulating oligos (leftmost 5 columns), or treated with naked PMO added directly to culture wells at the concentrations indicated (rightmost 3 columns). NTX: no treatment; SKIP: M23D skipping oligo alone; SKIP+SPO: M23D combined with fourfold excess of sponge oligo; SKIP+PRO: M23D combined with equimolar protector; SKIP+31: M23D combined with equimolar miR31-analogue sequence. AU: arbitrary units. Dystrophin mRNA level normalized to the geometric mean of three reference genes; see methods.

Sup Fig S3

Supplementary figure S3: schematic for exon skipping in cell culture Proliferating H2KSF1 cells are transferred to differentiation medium (T0). M23D skipping oligo is added between 48 and 52 hours following a partial media change (50% replacement) and no further media changes occur. MiR31-modulating oligos are added either simultaneously, or after 24 or 48 hours. Beyond 120-144 hours of differentiation, onset of spontaneous contractile behaviour necessitated termination of culture and sample collection.

Sup Fig S4

Supplementary figure S4: Total dystrophin expression in cell culture following miR31-modulation. Total dystrophin transcripts in dystrophic myotube cultures carrying the mdx mutation (H2KSF1), either untreated (ntx), incubated with M23D skipping oligo alone (skip), or with M23D combined with miR31-modulating oligos (as indicated. Top: sponge. Middle: protector. Bottom: miR31-analogue) added either simultaneously (T0), 24 hours after M23D addition (T24) or 48 hours after (T48). miR31-modulating oligos were added at 1:1 or 4:1 ratios with skipping oligo as indicated. Means + SEM (N=3). AU: arbitrary units. Dystrophin mRNA level normalized to the geometric mean of three reference genes; see methods.

Sup Fig S5

Supplementary figure S5: Exon skipping in mice following IP administration of M23D Vivo-Morpholino. Unskipped (top) and skipped (bottom) dystrophin transcripts in 8 muscle groups (soleus, extensor digitorum longus, gastrocnemius, diaphragm, body wall, quadriceps, tibialis anterior and triceps, as indicated), 2 weeks after a single IP injection of M23D Vivo-Morpholino. Numbers in lower chart: percentage skipped transcripts (skipped transcripts as a percentage of total dystrophin transcripts). AU: arbitrary units. Dystrophin mRNA level normalized to the geometric mean of three reference genes; see methods.

Sup Fig S6

Supplementary figure S6: Intraperitoneal delivery of exon skipping and miR31-modulating Vivo-Morpholinos to diaphragms and body wall musculature. Total dystrophin transcripts (left panel), and unskipped dystrophin transcripts (right panel) in diaphragm (top) and body wall (bottom) musculature, treated with Vivo-Morpholinos (as indicated). NTX: no treatment; Skip only: M23D skipping oligo alone; Skip Pro: M23D combined with protector; Skip Scram: M23D combined with scrambled protector sequence. Means + SEM (N=6). AU: arbitrary units. Dystrophin mRNA level normalized to the geometric mean of three reference genes; see methods.

Sup Fig S7

Supplementary figure S7: Exon skipping via M23D and miR31-modulating Vivo-Morpholinos in cell culture. Skipped dystrophin transcript levels from dystrophic (H2KSF1) myotube cultures incubated with Vivo-Morpholinos at the concentrations indicated for 72 hours. Minimal skipping was detected at concentrations below 500nM; however at 500nM, presence of additional (non-skipping) oligo unexpectedly potentiates exon skipping. Skip only: M23D skipping oligo alone; Skip+Pro: M23D combined with protector; Skip_Scram: M23D combined with scrambled protector sequence. Means + SEM (N=3). AU: arbitrary units. Dystrophin mRNA level normalized to the geometric mean of three reference genes; see methods.


All data used to generate the figures shown is freely available under Creative Commons licence at the figshare link below.

Competing Interests

JCWH declares no competing interests exist.

DJW is on the Scientific Advisory Board of Akashi Therapeutics, a company developing treatments for Duchenne muscular dystrophy. DJW is also a member of the Treat-NMD Advisory Committee for Therapeutics that provides confidential guidance on the translation and development path of therapeutics programs in rare neuromuscular diseases. This does not alter the authors’ adherence to all PLOS policies on sharing data and materials.

]]> 0
Increased taurine in pre-weaned juvenile mdx mice greatly reduces the acute onset of myofibre necrosis and dystropathology and prevents inflammation Fri, 29 Apr 2016 14:46:04 +0000 Background: The mdx mouse model for the fatal muscle wasting disease Duchenne Muscular Dystrophy (DMD) shows a very mild pathology once growth has ceased, with low levels of myofibre necrosis in adults. However, from about 3 weeks of post-natal age, muscles of juvenile mdx mice undergo an acute bout of severe necrosis and inflammation: this subsequently decreases and stabilises to lower adult levels by about 6 weeks of age. Prior to the onset of this severe dystropathology, we have shown that mdx mice are deficient in the amino acid taurine (potentially due to weaning), and we propose that this exacerbates myofibre necrosis and inflammation in juvenile mdx mice. Objectives: The purpose of this study was to increase taurine availability to pre-weaned juvenile mdx mice (from 14 days of age), to evaluate the impact on levels of myofibre necrosis and inflammation (at 22 days) during the acute period of severe dystropathology. Results: Untreated 22 day old mdx muscle was not deficient in taurine, with similar levels to normal C57 control muscle. However taurine treatment, which increased the taurine content of young dystrophic muscle (by 40%), greatly reduced myofibre necrosis (by 75%) and prevented significant increases in 3 markers of inflammation. Conclusion: Taurine was very effective at preventing the acute phase of muscle damage that normally results in myofibre necrosis and inflammation in juvenile mdx mice, supporting continued research into the use of taurine as a therapeutic intervention for protecting growing muscles of young DMD boys



Duchenne Muscular Dystrophy (DMD) is a lethal, X-chromosome linked muscle disease affecting about 1 in 3500-6000 boys worldwide (Reviewed in 1,2). DMD is caused by the loss of functional dystrophin protein in muscle that results in increased necrosis and inflammation after muscle contraction34,5,6. Repeated cycles of widespread myofibre necrosis and progressive failure of regeneration (with replacement of myofibres by fatty and fibrous connective tissue) lead to the loss of muscle mass and function in DMD boys, with premature death often due to respiratory or cardiac failure (Reviewed in 1,7). There is no cure for DMD and the standard drug treatment for DMD, corticosteroids, are limited in their efficacy and are associated with severe side effects8. Consequently, there is considerable interest in pharmacological interventions and nutritional supplementation as potential therapies to reduce disease severity (reviewed in 9,10,11).

Much DMD research utilises the adult mdx mouse model to test potential therapies. Whilst the adult mdx mouse has a very mild pathology, the juvenile mdx mouse undergoes an acute onset of severe myofibre necrosis, associated with many inflammatory cells (and subsequent myogenesis and new muscle formation) between 3 and 4 weeks of age12,13. Therefore, the capacity of a therapy to prevent necrosis of juvenile mdx muscles would be a rigorous test of the efficacy of a potential clinical treatment, especially considering that DMD first manifests in young growing boys.

We and others have shown that the amino acid taurine decreases inflammation and improves muscle strength in adult mdx mice14,15,16,17. Taurine is found in many tissues and is considered important for the function of skeletal muscle, where it modulates ion channel function, membrane stability and calcium homeostasis, as well as having anti-inflammatory and antioxidant properties18,19,20,21,22,23. The effect of taurine treatment on severe myofibre necrosis in juvenile mdx mice has not been investigated. One reason for this is that juvenile pre-weaned mice are not routinely available from some commercial suppliers of mdx mice: thus many pre-clinical studies are limited to interventions using older mice.

We have shown that prior to the onset of pathology (18 days), mdx mice are deficient in taurine24. Mouse milk is very rich source of taurine, being the most abundant amino acid in mouse milk25. From about 10-17 days of age mouse pups begin eating solid food, and milk production of the mother dramatically drops between 16 and 21 days26,27. Since standard mouse chow is almost devoid of taurine, we propose that weaning (and subsequent associated drop in taurine ingestion) leads to a taurine deficiency in juvenile mdx mice, which initiates muscle necrosis.

To investigate the proposed contribution of taurine deficiency to susceptibility of young growing mdx muscles to necrosis, juvenile mdx mice were given access to taurine enriched chow from 14 days until sampling at 22 days (after the initiation of myofibre necrosis). Quadriceps muscles from untreated and taurine treated mdx mice, and untreated control normal C57Bl/10Scsn (C57) mice were analysed for taurine content, myofibre necrosis and markers of inflammation (neutrophil elastase, myeloperoxidase (MPO) and the pro-inflammatory cytokine tumour necrosis factor [TNF]). Taurine treatment increased the taurine content of mdx quadriceps muscles, and resulted in a striking decrease in myofibre necrosis and inflammation, providing further support for taurine as a potential intervention for growing DMD boys.


All reagents used were obtained from Sigma Aldrich unless otherwise specified.

Animal procedures All experiments were carried out on dystrophic mdx (C57Bl/10ScSnmdx/mdx) and non-dystrophic control C57 (C57Bl/10ScSn) mice (the parental strain for mdx). All mice were obtained from the Animal Resource Centre, Murdoch, Western Australia. Mice were maintained at the University of Western Australia on a 12-h light/dark cycle, under standard conditions, with free access to food and drinking water. All animal experiments were conducted in strict accordance with the guidelines of the National Health and Medical Research Council Code of practice for the care and use of animals for scientific purposes (2004), and the Animal Welfare act of Western Australia (2002), and were approved by the Animal Ethics committee at the University of Western Australia.

From 14 days of age, mice had access to soft chow, with taurine treated mice receiving 4% taurine in their chow. Each group contained n=8 pups, with approximately equal proportions of male and female pups (~50:50). There was no observable difference between male and female mice. All mice were sacrificed at 22 days by cervical dislocation while under terminal anesthesia (2%v/v Attane isoflurane Bomac Australia). Quadriceps muscles were collected and immediately snap frozen in liquid nitrogen for biochemical analysis, or prepared for histology by immersing in 4% paraformaldehyde before possessing for paraffin histology.

Taurine content of muscle Taurine in muscle was measured using reverse phase high performance liquid chromatography (HPLC) as previously described28. Frozen quadriceps muscles were crushed using a mortar and pestle under liquid nitrogen and homogenized in 25 times 5% TCA, and plasma samples were precipitated by addition of 10 times by weight of 5% TCA. After centrifugation, supernatants were removed and stored at -80°C before analysis. Analytes were separated using HPLC with fluorescent detection, with pre-column derivitisation with o-phthalaldehyde (OPA) and 2-mercaptoethanol (2ME). Supernatants were mixed with iodoacetamide, dissolved in 5% TCA, to a final concentration of 25 mM. An internal standard, o-phospho-dl-serine, dissolved in 5% TCA was added to a final concentration of 5 mM. Sodium borate was used to adjust the pH to 9. Samples were mixed on a sample loop with a derivatising solution containing 40 mM OPA and 160 mM 2ME in 100 mM sodium borate, pH 12, for 30 seconds before injection onto the column. Separation was achieved with a C18 column (5 µl, 4.6 x 150 mm, Phenomenex) using a Dionex Ultimate 3000 HPLC system. Mobile phase A consisted of 50 mM potassium phosphate buffer, methanol and tetrahydrofuran (94:3:3). Mobile phase B consisted of 90% methanol, with a gradient increase in B from 0 to 25%. Fluorescence was set at 360 nm and 455 nm for excitation and emission respectively. The protein content of muscle samples were quantified by solubilising the pellet in 0.5 M sodium hydroxide, before incubation at 80°C for 15 minutes. Once fully dissolved, protein concentrations of supernatants were quantified using a Bradford protein assay (Bio-Rad).

Myofibre necrosis Histological analysis was completed as per the TREAT-NMD recommended standard protocol “Histological measurements of dystrophic muscle – M.1.2_007”

Transverse muscle sections (5 μm) were cut through the mid-region of each quadriceps muscle on a Leica microtome, and sections were stained with Haematoxylin and Eosin (H&E) for morphological analysis.

Myofibre necrosis was identified as areas of myofibres with fragmented sarcoplasm and/or increased inflammatory cell infiltration, and was measured using non-overlapping tiled images of transverse muscle sections that provided a picture of the entire muscle cross section. Tiled digital images were captured at x10 magnification using a Nikon Eclipse Ti inverter microscope equipped with Nikon DS-Fi2 camera (Nikon Corporation). Analysis was performed blind, and areas of necrosis drawn manually by the researcher using Image Pro Plus 4.5.1 software.

Neutrophil elastase and TNF content of muscle Frozen quadriceps muscles were crushed using a mortar and pestle under liquid nitrogen and homogenized in ice-cold 1% NP40, 1 mM EDTA in phosphate buffered saline (PBS), supplemented with complete EDTA free protease inhibitor tablets and PhosSTOP phosphatase inhibitor tablets (Roche), and centrifuged. The protein concentration of supernatants was quantitated using the Detergent Compatible (DC) protein assay (Bio-Rad). Samples were resolved on 4-15% SDS-PAGE TGX gels (Bio-Rad) and transferred onto nitrocellulose membrane using a Trans Turbo Blot system (Bio-Rad). Immuno-blotting was performed on the same membrane with antibodies to neutrophil elastase (ab68672, Abcam), TNF (AB2148P, Chemicon), and glyceraldehyde 3-phosphate dehydrogenase (GAP, 14C10, Cell Signalling), all dissolved 1:1000 in 5% bovine serum albumin (BSA). HRP-conjugated secondary antibodies were from Thermo Fisher Scientific. Chemiluminescence signal was captured using the ChemiDoc MP Imaging System (Bio-Rad). Resultant images were quantified using ImageJ software29. Glyceraldehyde 3-phosphate dehydrogenase loading controls were immunoblotted on the same membrane as immunoblotted proteins, and signals for neutrophil elastase and TNF were standardised to this loading control.

Myeloperoxidase (MPO) content of muscle The enzyme MPO catalyses the production of hypochlorous acid from hydrogen peroxide and chloride30 and hypochlorous acid reacts with 2-[6-(4-aminophenoxy)-3-oxo-3H-xanthen-9-yl]benzoic acid (APF) to form the highly fluorescent compound fluorescein, that is measured in this method, as per [28]. Briefly, frozen quadriceps muscles were ground using a mortar and pestle under liquid nitrogen and homogenised in 0.5% hexadecyltrimethylammonium bromide in PBS. Samples were centrifuged and supernatants diluted in PBS. Human MPO was used as the standard for the assay (Cayman Chemical). Aliquots of each experimental sample or MPO standard was pipetted into a 384 well plate, before the addition of APF working solution (20 µM APF [Cayman Chemical] and 20 µM hydrogen peroxide in PBS) was added. The plate was incubated at room temperature (protected from light) for 30 minutes, with the fluorescence being measured every minute using excitation at 485 nm and emission at 515-530 nm. The rate of change of fluorescence for each sample was compared to that of the standards and results were expressed per mg of protein, quantified using the DC protein assay (Bio-Rad).

Statistics Data were analysed using GraphPad Prism software. One-way ANOVA tests with post-hoc (LSD) comparisons were used to identify significant differences between experimental groups. Statistical significance was accepted at p<0.05. All data are presented as mean ± SEM.


Muscle taurine content There was no significant difference between the taurine content of C57 and untreated mdx muscle at 22 days of age (Fig. 1). Taurine treatment of juvenile mdx mice for 8 days resulted in a 1.4 fold increase in muscle taurine content.

Fig. 1

Fig. 1: Taurine content of C57, untreated mdx and taurine treated mdx quadriceps muscles, from mice aged 22 days

Data are presented as mean ± SEM and n= 8 mice/group. Groups without a common letter are significantly (p<0.05) different.

Muscle necrosis Myofibre necrosis was minimal in normal C57 quadriceps muscle, whereas myofibre necrosis was conspicuous and represented about ~17% of the cross-sectional area of the quadriceps in untreated mdx mice aged 22 days (Fig. 2). Taurine treatment of mdx mice significantly reduced (by 4 fold) myofibre necrosis (to ~5%) (Fig. 2).

Fig. 2

Fig. 2: Myofibre necrosis in C57, untreated mdx and taurine treated mdx quadriceps muscle, from mice aged 22 days

(A) Histological quantification of myofibre necrosis. Data are presented as mean ± SEM of percentage of cross section area (CSA) and n= 8 mice/group. Groups without a common letter are significantly (p<0.05) different. Representative images of myofibre necrosis and histology of H&E stained muscle sections are shown for (B) untreated mdx (C) taurine treated mdx mice.

Muscle Inflammation Neutrophil accumulation is a hallmark of acute inflammation, and we assessed the incidence of neutrophils by western blotting for the protein neutrophil elastase and by measuring the activity of MPO, an enzyme secreted by inflammatory cells (primarily by neutrophils) that facilitates their antimicrobial activity. Neutrophil elastase and MPO activity were 6.7 and 4 fold higher (respectively) in mdx compared to C57 muscle (Fig. 3). Taurine treatment of mdx mice reduced neutrophil elastase protein and MPO activity by 2.3 and 2 fold, respectively (Fig. 3): these reduced levels were not significantly different to those in normal C57 muscles (Fig. 3).

Fig. 3

Fig. 3: Quantification of inflammation in C57, untreated mdx and taurine treated mdx quadriceps muscles, from mice aged 22 days

Measurements are of (A) Neutrophil elastase, (B) MPO and (C) TNF. Data are presented as mean ± SEM and n= 8 mice/group. Groups without a common letter are significantly (p<0.05) different. Representative blots are shown of neutrophil elastase, TNF and the loading control glyceraldehyde 3-phosphate dehydrogenase (GAP).

Protein levels of the pro-inflammatory cytokine TNF measured by western blotting (Fig. 3C) were 2.4 fold higher in mdx compared to C57 muscles. Taurine treatment of mdx mice resulted in a striking 2.4 fold reduction in muscle TNF content (compared with untreated mdx), to the same low levels as in C57 muscles (Fig. 3C).


Taurine administration to juvenile mdx mice from 14 days of age substantially increased muscle taurine content and greatly mitigated the severity of the acute onset of myofibre necrosis and prevented muscle inflammation at 22 days of age. These data are novel and provide strong support for the growing interest in taurine as a potential low cost clinical intervention to protect the muscles of growing DMD boys.

We have previously reported a taurine deficiency in young 18 day old mdx mice, prior to the acute onset of pathology, and this deficiency coincided with the time of weaning of pups from taurine rich milk to taurine poor chow: accordingly, we proposed that weaning (with subsequent drop in taurine ingestion) leads to a taurine deficiency in young mdx mice, which exacerbates muscle necrosis24. However, by 22 days taurine levels in mdx mice have recovered to normal control C57 levels. These data are unclear in defining the role of taurine levels in onset of mdx pathology: this may relate to precise timing of changing taurine levels in growing mdx mice between 18-22 days, since the initiation of myofibre necrosis, that occurs just before 22 days, may be intensified by persistent prior taurine deficiency. Many other cellular and molecular factors that change around 21 days in growing mice may also contribute to the timing of acute onset of myonecrosis in dystrophic muscles, and factors to consider also include the impact of growth, increased mechanical activity and loading of juvenile muscles3,31,32.

The early administration of taurine to juvenile mdx mice decreased (compared with untreated mdx mice) and normalised to C57 levels 3 measures of inflammation in muscles at 22 days: neutrophil elastase, MPO activity and TNF content. Neutrophils, key cells involved in acute inflammation, are phagocytes responsible for generation of various pro-inflammatory mediators33. After muscle injury, neutrophils rapidly invade the tissue to remove debris, however in doing so can exacerbate muscle damage34. A fundamental mechanism of neutrophil mediated damage to muscle is the secretion of MPO, a heme enzyme that oxidises chloride in the presence of hydrogen peroxide to form the potent and damaging oxidant hypochlorous acid (HOCl)35. The anti-inflammatory and antioxidant properties of taurine are attributed to its ability to react with hypochlorous acid to form the much less reactive molecule taurine chloramine which itself exerts anti-inflammatory effects such as inhibiting the production of pro-inflammatory cytokines, including TNF33.

Activated leucocytes such as neutrophils (as well as many other cell types including macrophages and muscle cells) produce TNF, and therefore TNF content of muscle is elevated after injury36. TNF plays several important roles in inflammation such as activation and chemotaxis of leucocytes, and can itself stimulate muscle injury via NF-κB mediated protein degradation36,37. In the current study, these anti-inflammatory properties of taurine may contribute to the decreased inflammation in mdx muscle observed after taurine treatment. In the current study, these anti-inflammatory properties of taurine may contribute to the decreased inflammation observed in mdx muscle after taurine treatment. However the reduced inflammation might simply be a consequence of the reduction in muscle necrosis by taurine (due to another effect such as membrane stabilisation or calcium homeostasis). More experimental research is required to understand the exact mechanism for the benefits of taurine on dystropathology and the protection of juvenile mdx muscle from necrosis and inflammation.

To summarise, taurine treatment was very effective in mitigating the severe bout of necrosis and preventing inflammation in dystrophic muscles of juvenile mdx mice. This is an important observation, since interventions that can protect the vulnerable growing dystrophic myofibres from necrosis could help preserve muscle mass and function in young DMD boys. These novel data support continued preclinical research into the use of taurine as a promising clinical intervention for DMD.

Competing interests

The authors have declared that no competing interests exist.


The corresponding author can be contacted at

]]> 0
Enhanced Reprogramming Efficiency and Kinetics of Induced Pluripotent Stem Cells Derived from Human Duchenne Muscular Dystrophy Thu, 03 Sep 2015 10:00:12 +0000 The generation of disease-specific induced pluripotent stem cells (iPSCs) holds a great promise for understanding disease mechanisms and for drug screening. Recently, patient-derived iPSCs, containing identical genetic anomalies of the patient, have offered a breakthrough approach to studying Duchenne muscular dystrophy (DMD), a fatal disease caused by the mutation in the dystrophin gene. However, development of scalable and high fidelity DMD-iPSCs is hampered by low reprogramming efficiency, the addition of expensive growth factors and slow kinetics of disease-specific fibroblasts. Here, we show an efficient generation of DMD-iPSCs on bFGF secreting human foreskin fibroblast feeders (I-HFF) by employing single polycistronic lentiviral vector for delivering of transcription factors to DMD patient-specific fibroblast cells. Using this method, DMD-iPSCs generated on I-HFF feeders displayed pluripotent characteristics and disease genotype with improved reprogramming efficiency and kinetics over to mouse feeders. Moreover, we were able to maintain disease-specific iPSCs without additional supplementation of bFGF on I-HFF feeders. Our findings offer improvements in the generation of DMD-iPSCs and will facilitate in understanding of pathological mechanisms and screening of safer drugs for clinical intervention. Key Words: Duchenne Muscular Dystrophy, Reprogramming, Induced pluripotent Stem Cells, Immortalized Human Feeder, Basic Fibroblast Growth Factor, Stem Cell Cassette



Disease-specific induced pluripotent stem cells (iPSCs) from patients with incurable diseases have come up with promising approach for research and clinical use in disease modeling including Duchenne muscular dystrophy (DMD), a disease caused by a mutation in the dystrophin gene 1,2,3. Innovations in iPSC technology have enabled us to generate an unlimited autologous number of cells for disease modeling with identical genetic anomalies with an endogenous regulatory system of disease 4,5. However, cellular changes during the reprogramming process, culture-induced differentiation, and differences in the genetic background limit their use in disease-specific phenotypes 6. Additionally, generation of disease-specific iPSCs is also surrounded by various impeding factors like inefficiency, slow kinetics, variability in protocols, and the high cost of the cell culture maintenance 7,8. Many approaches have been developed to overcome these limitations. The initial methods for reprogramming of DMD-iPSCs were based on the monocistronic approach which may lead to leaky expression of reprogramming factors and low reprogramming efficiency 9,1,10,11. Contrary to this, polycistronic lentiviral delivery offers the improvement over the monocistronic approach by synthesis of all four reprogramming factors from one mRNA to generate DMD-iPSCs 12,13. In addition to the transcription factors, suitable techniques of factor delivery are crucial in order to use of iPSCs for clinical applications 14.

Various viral delivery methods such as integrating viral methods i.e. retroviral monocistronic 15, lentiviral polycistronic with or without cre-lox mediated transgene excision 16,17; non-integrating viral methods i.e. adenovirus 18, Sendai virus 19; and non-viral methods i.e. direct protein delivery 20, defined media 21, plasmid transfections 22 have been employed to deliver reprogramming factors to generate iPSCs. Generally, non-integrative iPSCs derivation strategies are seemingly safe for cell therapy, yet associated with challenges to produce and purify required quantities of recombinant proteins in protein-based strategies 23 , stringent steps to manipulate sensitivity of viral RNA replicase in Sendai virus delivery method 19 and are often associated with low reprogramming efficiency 24. Messenger RNA (mRNA) offers an integration-free alternative method to make footprint free iPSCs with higher reprogramming efficiency, though it is relatively laborious due to serial delivery of multiple reprogramming molecules to target somatic cells to induce pluripotency 25,26. In contrast, lentiviral delivery methods have the advantage over non-viral methods by transducing non-proliferative cells and better reprogramming efficiency, though cells transduced with this method are not suitable for cell therapy due to transgene integration but could be helpful for in vitro disease modeling and screening of new drugs 27. Nevertheless, the use of non-integrating viral methods and non-viral delivery methods for the derivation and long-term and large-scale culture of human iPSCs is not cost-effective. In addition to reprogramming factor and delivery methods, use of feeder cells is also crucial to maintaining the pluripotency of iPSCs. Mostly, DMD-iPSCs have been generated using mouse feeders that can be a possible cause of xenogeneic contaminations, and variability in cells and culture conditions, limiting their use in clinical applications 9,10,28. Defined culture conditions by employing human feeders obtained during different stages of development (fetal, neonatal, and adult) from various tissue sources like skin, muscle, and placenta have also been shown to support the iPSCs culture in an undifferentiated state and to bypass the problem of xenogeneic contaminations 29,30. Nonetheless, the inconsistency and heterogeneity are still major concerns and can be overcome by the only use of a consistent source of the human feeders. Additionally, generation of DMD-iPSCs also require supplementation of growth factors and is not cost effective for large-scale use.

Thus, no single method adequately addresses all limitations and remains constrained by the complexity, low reprogramming efficiency, heterogeneity and higher cost associated with the method used 31,25. In this study, we demonstrated an efficient generation of DMD-iPSCs on immortalized human feeders with improved reprogramming efficiency, and kinetics.

Materials and Methods

Derivation of DMD primary human fibroblast cells (hFib)

Our study was conducted only after obtaining written informed consent from the parents or guardians of the participants and approval from the Institutional Committee for Stem Cell Research and Therapy Institutional (Ref 23/02/10-A5) and the Institute Ethics Committee All India Institute of Medical Science (Ref. IEC/NP-82/2010). Skin biopsies were taken from DMD patients ages 6-12 years old under properly administered local anesthesia were obtained from trained physician using a 6-mm punch biopsy needle. Biopsy specimens were transported to the laboratory under sterile conditions. Culture of hFib cells from skin biopsy was performed as previously described . Briefly, the dermis was separated out from the rest of the skin (epidermis, subcutaneous tissue, vascular structure) using scalpel and forceps.The dermis was cut to small pieces of explants of approximately 2mm x 2mm sizes. Few explants were then placed on 35-mm culture dish and covered with a sterile coverslip. The explants were maintained for 7-14 days in DMEM-high glucose (Gibco/Invitrogen, USA) supplemented with 10% FBS (Hyclone, USA), 2mM glutamine (Gibco/Invitrogen, USA), 1% Penicillin /streptomycin (Gibco/Invitrogen, USA). Cells from passage number 3-4 were used for reprogramming experiments. Human embryonic stem cells (KIND-1) and bFGF secreting immortalized human foreskin fibroblast (I-HFF) cells were kind gifts of Dr. Deepa Bhartiya (National Institute for Research in Reproductive Health, India) and Professor Anis Feki (Geneva University Hospital, Switzerland), respectively. These cells were cultured and maintained as described earlier 32,30. Primary mouse embryo fibroblast (PMEF) was commercially available and purchased from the company (Millipore, USA).

Preparation of Feeders and mitotic inactivation

Mouse and human feeders were plated onto 0.1% gelatin coated 6-well culture plates at a density of 4.75 x 105 cells. When cells reached 80% confluency, mitomycin C (10ug/ml) (Sigma, USA) was added to the medium and incubated for 3h at 37ºC and 5%CO2. After incubation, the medium was aspirated, and cells were washed with buffered saline to remove any traces of mitomycin C and used immediately or cryopreserved for future use.

Collection of Conditioned media

I-HFF feeder cells were plated on 25cm2 gelatin-coated tissue culture flask at a density of 1.25×106 in I-HFF expansion media. After 24h, I-HFF was mitotically inactivated by mitomycin C as described above and supplemented with basal iPSC media, composed of Knockout-DMEM (KO-DMEM) supplemented with 20% Knockout Serum Replacement (KO-SR), 1% nonessential amino acids, 2mM/ml L-glutamine, 1% penicillin and streptomycin (all media components from Gibco/Invitrogen, USA), 0.01mM β-mercaptoethanol (Sigma, USA). After 24h, conditioned medium (CM) was collected and stored after filtration at -20ºC until used for reprogramming and expansion medium of iPSCs.

Generation of DMD-iPSCs on I-HFF and mouse feeder

DMD-iPSCs were generated using human STEMCCA constitutive polycistronic (OKSM) lentivirus reprogramming kit (Millipore, USA) as per manufacturer’s instructions. Briefly, hFib cells were infected with the multiplicity of infection (MOI) of 1 in the presence of 6 μg/ml of polybrene (Sigma, USA). After 24h of infection, the medium was changed to fresh medium. The infected cells were maintained in hFib expansion media until day 6. After six days of infection, cells were cultured on the mitotically inactivated mouse as well as on human I-HFF feeders using standard hESC culture conditions. Cells grown on human feeders were also fed with conditioned medium from I-HFF and fresh iPSC basal medium in 1:1 ratio with 15ng/ml basic fibroblast growth factor (bFGF) (Peprotech, USA), as shown in figure (Figure1). DMD-iPSCs grown on human feeders were maintained in conditioned medium from I-HFF (I-HFF-CM).

Efficiency of iPSCs generation was calculated by the following formula:

Total number of iPSCs colonies obtained/0.1×105 fibroblasts exposed to the virus x 100.

DMD MS PIX revised-1

Fig. 1: Schematic representation of experimental approach for generating DMD-iPSCs.

Freshly isolated DMD derived hFib cells in fibroblast expansion medium were infected with lentivirus, and 6 days post infection, maintained under hESC culture conditions till colony picking; (A) medium cocktail used during reprogramming of DMD human Fib Cells with hSTEMCCA on mouse Feeders; (B) medium cocktail used during reprogramming of DMD human Fib Cells with hSTEMCCA on immortalized I-HFF feeders.

Passaging of DMD-iPSCs clones

Fully reprogrammed DMD-iPSC colonies were manually picked with the help of heat-pulled glass Pasteur pipettes (Sigma, USA) under stereomicroscope (Nikon, Japan) and disintegrated into small clumps mechanically without enzymatic digestion and transferred to 24-well plates on mitomycin C treated feeders. This stage was counted as passage 1.

Characterization of genetic defects in iPSCs cells

DMD-hFib and DMD-iPSCs were checked for the exons deletion in dystrophin gene using Chamberlain or Beggs multiplex primer sets and PCR 33,34.

Alkaline Phosphatase and Immunofluorescence staining

Alkaline phosphatase staining was performed using alkaline phosphatase detection kit as per manufacturer’s protocol (Millipore, USA). Live staining was performed as described previously 35. Briefly, TRA 1-60 antibody (Chemicon, USA) was diluted in KO-DMEM (1X) medium at 1:100 dilutions and added into reprogramming cell culture plates and incubated for 1 hour at 37ºC. Followed by washing of unbound primary antibody using KO-DMEM (1X) media. The reprogramming cell culture plates were again incubated with appropriate secondary antibody in 1:100 dilutions for 1 hour at 37ºC. After washing, the medium was changed to routine expansion medium. The colonies were identified under an inverted fluorescent microscope and picked up for further expansion and characterization. TRA-1-60 positive colonies were considered as fully reprogrammed clones and calculated as a fraction of positive TRA 1-60 colonies to total colonies obtained.

For intracellular staining, cells were fixed with 4% paraformaldehyde and permeabilized with 0.1% Triton X-100 followed by blocking in 2% BSA for 1 h. Cells were then probed for Oct4, Sox2, Cardiac Troponin I, α SMA, Tuj1, Nestin, AFP and GATA 4 primary antibodies (Abcam, USA). Afterward, cells were further incubated with Alexa Fluor 488- or 568 tagged secondary antibodies (Molecular Probes, USA) for 1 h. Images were captured on Nikon TE 80i (Nikon, Japan) fluorescent microscope and analyzed using NIS element software.

Gene expression analysis

Total RNA was isolated using Trizol (MRC Inc., USA) reagent and was reverse transcribed using cDNA synthesis kit (Applied Biosystems, USA), according to the kit’s instructions. Expression of various pluripotency and differentiation markers was analyzed using reverse transcriptase-polymerase chain reaction (RT-PCR). Quantitative analysis was carried out in Realplex4 real-time PCR system (Eppendorf, USA) using SYBR green chemistry. The reaction was performed in triplicates, and internal endogenous GAPDH gene expression was used for normalization. Relative quantification was calculated using the comparative Ct method/2-ΔΔCt method. The data are presented as mean fold change.

Analysis of pluripotency in vitro/Embryoid body (EB) formation

For embryoid body formation, undifferentiated DMD-iPSC colonies were manually picked up and dissociated into small clumps. These clumps were allowed to grow in suspension culture using non-adherent dishes in the hESC medium with 15% FBS without bFGF. The medium was changed every other day. After 14 days, cystic EBs was further cultured on gelatin-coated dishes for spontaneous differentiation into various cell lineages using same media. At day-14 and 21 EBs were harvested for gene and protein expression studies.

Statistical analysis

The data were analyzed and plotted using GraphPad Prism (GraphPad Software, Inc., USA) and Windows Excel (Microsoft, Redmond, USA). Statistical significance was determined using an unpaired Student t test. The data were presented as means±SD. The p-value < 0.05 was considered significant and > 0.05 was non-significant (ns). Single asterisk and double asterisk depict different p-values p<0.05 and p<0.01, respectively.


Human feeders improve kinetics and reprogramming efficiency of DMD-iPSCs

Reprogramming kinetics and efficiency to generate DMD-iPSCs were checked on both human I-HFF and mouse feeders by comparing temporal kinetics and fraction of total colonies to the number of transduced cells, respectively. DMD-specific hfib cells displayed improved reprogramming kinetics on human I-HFF feeder when compared to mouse feeders, which is evident by showing increase in percentage of TRA-1-60 positive (95.5%) live staining (Figure 2A left panel) and by early appearance (12-15 days) of colonies (Figure 2A right panel). We also observed higher reprogramming efficiency (0.99+0.20 %) on I-HFF feeders in comparison to mouse feeders (0.25+0.07 %) (Figure 2B). Colonies developed on both feeders displayed morphology similar to well-established human ESC line; however, colonies grown on human feeders were more defined and compact (Figure 2C).

(A) Percentage of TRA 1-60 positive DMD-iPSC colonies among total colonies generated by hSTEMCCA (left panel) and temporal kinetics of reprogramming (right panel) on both mouse and human feeders. The graph represents mean of three independent experiments (n=3). Error bars show mean ± S.D, **: p<0.01; (B) Percent reprogramming efficiency of DMD-iPSCs obtained on mouse and human feeders. Data are represented as mean ± S.D, **: p<0.01, (n=3); (C) Representative pictures of DMD-iPSCs colonies derived using skin fibroblast from DMD patient on both mouse (upper panel) and human feeders (lower panel) at 2x and 10x magnifications. Resembling typical hESC morphology with well-defined borders and high nuclear to cytoplasmic ratio.

Fig. 2: Enhanced reprogramming efficiency of hFib on immortalized human feeders.

(A) Percentage of TRA 1-60 positive DMD-iPSC colonies among total colonies generated by hSTEMCCA (left panel) and temporal kinetics of reprogramming (right panel) on both mouse and human feeders. The graph represents mean of three independent experiments (n=3). Error bars show mean ± S.D, **: p<0.01; (B) Percent reprogramming efficiency of DMD-iPSCs obtained on mouse and human feeders. Data are represented as mean ± S.D, **: p<0.01, (n=3); (C) Representative pictures of DMD-iPSCs colonies derived using skin fibroblast from DMD patient on both mouse (upper panel) and human feeders (lower panel) at 2x and 10x magnifications. Resembling typical hESC morphology with well-defined borders and high nuclear to cytoplasmic ratio.

Human feeders provide stable growth and pluripotency to DMD-iPSCs without bFGF supplementation

Stable DMD-iPSC clones (D-iPSC 1-3) obtained on I-HFF feeders were propagated in I-HFF-CM without supplementation of bFGF and checked for the pluripotency and human embryonic stem cells (hESC) like characteristics. These clones displayed expression of pluripotency markers such as TRA-1-60, OCT4, Sox2 using immunocytochemistry and alkaline phosphatase (AP) staining (Figure 3A; 3B and 3C). Pluripotency was also confirmed by gene expression analysis using RT-PCR and quantitative real-time PCR for pluripotency markers (Figure 3D and 3E, Table 1). These clones expressed endogenous OCT4, SOX2, KLF4, c-MYC and NANOG similar to hESCs.

Table 1

Primer Sequences for both RT and qPCR

Primer Sequence (for both RT and qPCR) Annealing Temp. Product Size
Pluripotency Markers
Differentiation Markers
i) Ectodermal
ii) Mesodermal
iii) Endodermal

(A) “Live” Tra-1-60staining (Red) and merged with bright-field appearance of indicated DMD-iPSC clones on mouse (upper panel) and human (lower panel) feeders. Red color indicates Texas Red labeled Tra-1-60 positive cells, I-HFF are tagged with green fluorescent protein (GFP); (B) Fluorescent microscopic images showing expression of hESC undifferentiated cell markers (Oct4, Sox2- Texas Red stained positive cells) nuclei- blue DAPI stained and I-HFF is green (GFP) stained, magnification 10X; (C) Representative image of DMD-iPSC showing alkaline phosphatase staining at 2x and 10x magnifications; (D) RT-PCR analysis and (E) qRT-PCR of endogenous gene expression of pluripotent markers in DMD-iPSC clones. Primers used for Oct3/4, Sox2, Klf4, and c-Myc specifically detect the transcripts from the endogenous genes, but not from the lentiviral transgenes. Negative control is without cDNA. D-iPSC-1, D-iPSC2 and D-iPSC3 represent three different clones from DMD-iPSCs. Data is presented relative to hESCs. Expression was normalized to GAPDH. Error bars represent standard deviation of replicates (n=5).

Fig. 3: Characterization of DMD-iPSCs without supplementation of bFGF.

(A) “Live” Tra-1-60staining (Red) and merged with bright-field appearance of indicated DMD-iPSC clones on mouse (upper panel) and human (lower panel) feeders. Red color indicates Texas Red labeled Tra-1-60 positive cells, I-HFF are tagged with green fluorescent protein (GFP); (B) Fluorescent microscopic images showing expression of hESC undifferentiated cell markers (Oct4, Sox2- Texas Red stained positive cells) nuclei- blue DAPI stained and I-HFF is green (GFP) stained, magnification 10X; (C) Representative image of DMD-iPSC showing alkaline phosphatase staining at 2x and 10x magnifications; (D) RT-PCR analysis and (E) qRT-PCR of endogenous gene expression of pluripotent markers in DMD-iPSC clones. Primers used for Oct3/4, Sox2, Klf4, and c-Myc specifically detect the transcripts from the endogenous genes, but not from the lentiviral transgenes. The negative control is without cDNA. D-iPSC-1, D-iPSC2 and D-iPSC3 represent three different clones from DMD-iPSCs. Data is presented relative to hESCs. Expression was normalized to GAPDH. Error bars represent standard deviation of replicates (n=5).

DMD-iPSCs show deletion in dystrophin gene

To further confirm the DMD disease state, we did multiplex PCR for exon-44 deletion. We observed deletion of exon-44 in both non-reprogrammed hFib cells and reprogrammed DMD-iPSCs (Figure 4).

DMD-hfib is fibroblasts from patients diagnosed with DMD. The control is genomic DNA from healthy individual. DMD-iPSCs (D-iPSCs) show deletion of exon 44 (268bp) similar to fibroblast derived from DMD patient.

Fig. 4: DMD-iPSC has a deletion over exon 44 (multiplex PCR for the dystrophin gene).

DMD-hfib is fibroblasts from patients diagnosed with DMD. The control is genomic DNA from a healthy individual. DMD-iPSCs (D-iPSCs) show deletion of exon 44 (268bp) similar to fibroblast derived from DMD patient.

Confirmation of DMD-iPSCs pluripotency by embryoid body formation

We confirmed the ability of all DMD-iPSCs clones to differentiate into three germ layers in vitro using the embryoid body (EB) assay. All iPSC clones displayed expression of all three lineages ectoderm, mesoderm, and endoderm markers, as evident by immunofluorescence and RT-PCR (Figure 5A; 5B and 5C). Embryoid bodies derived from DMD-iPSC clones displayed absence of pluripotency markers as quantified by qRT-PCR (Figure 5D).

(A) Representative images of cystic embryoid body (EB) in suspension culture (I) and their spontaneous differentiation (II) into three germ layers (magnification 10x); (B) Immunostaining of 21 days old EBs derived from D-iPSC revealed expression of ectodermal (Tuj 1 and Nestin), mesodermal (Cardiac troponin 1 and SMA), and endodermal (AFP and GATA4) marker proteins. Tuj 1, Nestin, Cardiac troponin 1, SMA and GATA4 are conjugated with fluorescein isothiocynate (FITC); AFP protein is conjugated with Taxas red. (Magnification 10x) (C) RT-PCR analysis of various differentiation markers for the three germ layers by EB mediated differentiation; D-iPSC is negative for expression. (D) Quantitative analysis of endogenous gene expression of pluripotent markers in D-iPSC EBs using qRT-PCR. Data is presented relative to D-iPSC clones. Expression is normalized to GAPDH. Error bars represent standard deviation of replicates (n=5). D-iPSC EB1, D-iPSC EB2 and D-iPSC EB3 represent embryoid bodies obtained from three different DMD-iPSCs clones.

Fig. 5: In-vitro differentiation of D-iPSC without supplementation of exogenous bFGF.

(A) Representative images of cystic embryoid body (EB) in suspension culture (I) and their spontaneous differentiation (II) into three germ layers (magnification 10x); (B) Immunostaining of 21 days old EBs derived from D-iPSC revealed expression of ectodermal (Tuj 1 and Nestin), mesodermal (Cardiac troponin 1 and SMA), and endodermal (AFP and GATA4) marker proteins. Tuj 1, Nestin, Cardiac troponin 1, SMA and GATA4 are conjugated with fluorescein isothiocyanate (FITC); AFP protein is conjugated with Texas red. (Magnification 10x) (C) RT-PCR analysis of various differentiation markers for the three germ layers by EB mediated differentiation; D-iPSC is negative for expression. (D) Quantitative analysis of endogenous gene expression of pluripotent markers in D-iPSC EBs using qRT-PCR. Data is presented relative to D-iPSC clones. The expression is normalized to GAPDH. Error bars represent standard deviation of replicates (n=5). D-iPSC EB1, D-iPSC EB2, and D-iPSC EB3 represent embryoid bodies obtained from three different DMD-iPSCs clones.


The field of iPSC is very promising when it comes to basic, pre-clinical and clinical aspects of regenerative medicine. However, the generation of DMD disease-specific iPSCs (DMD-iPSCs) is surrounded by various impeding factors like inefficiency, slow kinetics, safety issues, and the high cost of culture maintenance 9, 10 . Here we reprogrammed DMD skin fibroblast in less than three weeks, from infection of somatic cells to iPSCs colony expansion with enhanced reprogramming efficiency and kinetics. Thus, enables the generation of DMD disease-specific iPSCs in reliable, cost-effective, and timesaving manner. To examine the effect of reprogramming approach and delivery methods we checked reprogramming efficiency and appearance of colonies after transduction of DMD fibroblast using integrated viral monocistronic and polycistronic methods and integration-free monocistronic viral methods. Use of polycistronic lentiviral vector improved the appearance of reprogrammed colonies and reprogramming efficiency of DMD patient-derived fibroblasts.

Supplementary Table 1

Supplementary Table 1: Reprogramming efficiencies and kinetics of DMD derived fibroblast on both human and mouse feeders obtained by employing monocistronic and polycistronic integrating Lentiviral method and monocistronic integration-free Sendai virus approach. Data shown represent the mean ± SD of three independent experiments (n=3).

Experimentally, we were able to derive iPSCs clones with the multiplicity of infection (MOI) as low as one (Supplementary Table 2) as previously demonstrated by others 16.

Supplementary table 2: Depicting the variations at a glance adopted for standardization for reprogramming of hFib cells for the efficient generation of DMD-iPSCs using polycistronic approach.

Supplementary Table 2: Depicting the variations at a glance adopted for standardization for reprogramming of hFib cells for the efficient generation of DMD-iPSCs using polycistronic approach at the different multiplicity of infections (MOIs). Representative picture of iPSC colony obtained after infection of DMD patient-derived fibroblast with hSTEMCCA viral particles at MOI of 1.

We used bFGF secreting immortalized human feeder in our study that has been shown already to secrete stable amounts of bFGF to support growth and maintenance of induced pluripotent stem cells, though not has been used to generate DMD-iPSCs 30. Supplementation of bFGF has been shown dispensable for the self-renewal and survival of hESCs by direct activation of signaling pathways, indirectly stimulating autocrine effect and paracrine network 36,37. Many reports have demonstrated the cooperation of both TGF-β and IGF-II with the bFGF pathway in maintenance of hESC pluripotency by establishing a regulatory stem cell niche 38,39. In another study of ours (unpublished data), we have also detected the significant amount of both (TGF-β and IGF-II) the molecules in conditioned medium of I-HFF cells as compared to other primary human and mouse fibroblast feeders (data not shown). To further substantiate the possible role of secreted molecules we compared reprogramming efficiency and kinetics of DMD fibroblast on immortalized human feeders in basal hESC medium with or without I-HFF-CM. The addition of conditioned medium significantly improved the kinetics and reprogramming efficiency of DMD fibroblasts (Supplementary Table 3).

Supplementary Table-3

Supplementary table 3: Reprogramming efficiencies and kinetics of DMD derived fibroblast obtained on I-HFF feeders with or without the addition of conditioned medium (I-HFF-CM) by employing polycistronic lentiviral approach. Data shown represent the mean ± SD of three independent experiments (n=3).

Several types of feeder cells MEF 40, human placenta 41, human fetal muscle and adult fallopian tube fibroblasts 42, 43 , and human dermal fibroblast 44 have been used for reprogramming and maintenance of iPSCs. However, these feeders cannot be readily isolated, support only for few passages, and also bear a risk of infection, and contamination 44, 45,46. Moreover, hESC when grown in the presence of xenogeneic feeders, are contaminated with Neu5Gc, a non-human sialic acid that can induce an immune response in humans and limit their use in therapeutic applications 45. Therefore, we included immortalized human feeder to overcome these limitations of heterogeneity and inconsistency.

To the best of our knowledge, we for the first time demonstrated that bFGF secreting human I-HFF feeders together with conditioned medium (I-HFF-CM) promote the reprogramming efficiency and kinetics of DMD-iPSCs, appeared early in 17-20 days and developed into fully reprogrammed colonies earlier than colonies grown on mouse feeders, as confirmed by percent of TRA-1-60 positive DMD-iPSC colonies (Figure 2A), and calculated percent reprogramming efficiency (Figure 2B). Generated DMD-iPSCs developed on human feeders together with I-HFF-CM showed similar morphology to human ES cells (Figure 2C).

Conditioned medium (I-HFF-CM) alone was sufficient to maintain DMD-iPSCs in undifferentiated states without exogenous bFGF supplementation (Figure 3A and 3B). We confirmed the pluripotency of these cells using alkaline phosphatase (AP) staining (Figure 3C), pluripotency markers by RT-PCR (Figure 3D) and qRT-PCR (Figure 3E). DMD-iPSCs generated on human feeders was also confirmed the DMD-specific genotype of their parental fibroblast cells for exon deletion using multiplex PCR (Figure 4) as reported earlier 9. Similar to discussed clones from one patient, we also successfully generated and characterized DMD-iPSCs from different patients to rule out intra-patient variability (Supplementary Table 4).

Supplementary Table 4

Supplementary table 4: Characterization of established human DMD-iPSCs cell lines from various DMD donors.

These DMD specific iPSCs was also differentiated in to embryoid body that was confirmed by immunostaining (Figure 5A), RT-PCR (Figure 5B) and qRT-PCR (Figure 5C). Hence, the use of I-HFF-CM provided an insight in the role of supportive factors released by feeders in the successful and complete reprogramming.

These DMD-specific iPSCs was also differentiated into the embryoid body formation that was confirmed by immunostaining (Figure 5A), RT-PCR (Figure 5B) and qRT-PCR (Figure 5C). Hence, the use of I-HFF-CM provided an insight into the role of supportive factors released by feeders in the successful and complete reprogramming.

On the basis of above observations, it is confirmed that addition of I-HFF-CM could enhance the reprogramming and kinetics of DMD fibroblast, and pluripotent characteristics of DMD-iPSCs can be preserved using I-HFF-CM without supplementation of additional bFGF, making stem cell culture more economical, and less time-consuming, thereby reducing the time for feeder cell preparation. This study laid the preliminary findings that can be further extended to combine the ex-vivo gene and autologous cell therapy to exploit the therapeutic way to treat DMD that avoid immune rejection with sufficient engraftment ability 40,4. Also, the immaculate expertise in patient-derived iPSCs can accelerate the possibility of personalized medicine in the clinical arena.

Competing Interests

The authors indicate no potential conflicts of interest.

Author Information

Corresponding Author: Sujata Mohanty, Ph.D.

Complete Mailing Address: Stem Cell Facility, 1st Floor ORBO Complex,

All India Institute of Medical Sciences,

Ansari Nagar, New Delhi-110 029

Telephone and Fax Numbers: +91-11-2659 3085 (Telephone)

+91-11-2658 8663 (Fax)

Email Id:

Pooja Teotia:

Sujata Mohanty:

Madhulika Kabra:

Sheffali Gulati:

Balram Airan:

Short Title: i-PSCs for human Duchenne muscular dystrophy.

]]> 0