Background
Rare diseases, those affecting fewer than 5 in 10,000 EU citizens, pose a range of challenges to researchers[
1]. Identifying sufficient patients with a given condition, let alone a specific genotype, to conduct meaningful clinical trials, natural history studies or health care research is difficult in small populations. Limited patient cohorts have been a barrier to the development of new therapies for rare diseases. Orphan drugs legislation, designed to encourage investment in the field, has had a positive impact with an unprecedented number of clinical trials initiated for rare diseases. However, it has also exposed a lack of trial readiness in many rare disease communities, with an increasing need to identify both patients who might participate in studies and sites with sufficient experience to take part in trials.
The requirement to improve trial readiness for the neuromuscular field was addressed via infrastructures developed by the EU FP6-funded TREAT-NMD network of excellence. Within this network, a series of patient registries, frequently established and run in collaboration with patient organisations, have established a resource for the identification of patients with specific genotypes and diagnoses (Duchenne Muscular Dystrophy, Spinal Muscular Atrophy and other neuromuscular diseases)[
2]. Other disease areas have taken a similar approach with the most recent Orphanet Report “Disease Registries in Europe” cataloguing 588 registries[
3,
4].
Patient identification is merely part of the challenge of establishing clinical trials in rare diseases. Only a limited number of clinics exist with experience in caring for patients with rare diseases and conducting research in the field, and those for any specific disease are often few and far between. Such issues have spurred the development and activities of specialised professional networks for a variety of rare and more common conditions, including the Parkinson’s Study Group, the European Cystic Fibrosis Society, and the European Huntington’s Disease Network[
5‐
7].
Neuromuscular diseases form a group of conditions often treated by the same professionals. Specialist neuromuscular clinics caring for patients with Duchenne muscular dystrophy (DMD) will often also see patients with Spinal Muscular Atrophy (SMA), Congenital Muscular Dystrophies (CMDs) and sometimes other conditions such as Charcot-Marie-Tooth disease (CMT) and Limb-Girdle Muscular Dystrophies (LGMDs)[
8]. Those working in such clinics are thus likely to possess specialist skills and expertise with applicability to a range of neuromuscular conditions. However, identifying the locations of such sites and their capabilities has hitherto proved difficult.
We describe the development and utilisation of an online registry of care and trial sites for neuromuscular disorders that addresses these challenges, the TREAT-NMD Care and Trial Site Registry (CTSR), operational since December 2007.
Methods
The CTSR is an online self-registration database for neuromuscular centres hosted by the University Medical Center Freiburg, Germany. Its initial aim was to enable clinical trial feasibility enquiries, but it has since been used for a range of feasibility, trial, and research enquiries.
The CTSR technical platform comprises a Java web application running on a MySQL database accessed via a secure web server. This allows swift self-registration and update of information by any internet-connected centre, regardless of geographic location. The application is designed in such a way to permit future extension.
CTSR questions were chosen via expert consensus within the TREAT-NMD network, and are based on typical industry feasibility enquiries and EUCERD recommendations[
9]. They are wide-ranging across four broad categories: Patient Cohort, Care Settings, Research and Education, and Clinical Trial Infrastructure (Table
1). Further detail is available on the TREAT-NMD website[
2].
Table 1
Simplified description of data captured in TREAT-NMD CTSR
Patient cohort
| Patients stratified by disease and age range (currently 10 NMDs including subtypes e.g. SMA I, II, III). |
Diagnostic tools as most appropriate for each condition. |
Care settings
| Availability of specialists and services in-centre. |
Arrangements for transition care. |
Availability of particular pulmonary, cardiac, muscle and bone function tests in-centre. |
Availability of particular physiotherapy facilities and equipment in-centre. |
Availability of emergency care in-centre. |
Experience of centre in conducting skeletal muscle biopsies. |
Research and education
| Extent of use of centre data in research, research funding arrangements, and papers authored by staff at centre |
Extent to which staff at centre have been involved in providing training at national and international levels |
Clinical trial infrastructure
| Available personnel (e.g. Study Nurses, Physiotherapists, Pharmacists) |
Previous experience (e.g. details of past participation in Phase I, II, III, IV clinical trials) |
| Availability and details of equipment (e.g. refrigerators, IT support) |
Two additional sections capture project-specific data: a questionnaire on DMD care provision, used by the CARE-NMD project to investigate care for DMD across 7 European countries; and a series of qualitative questions for UK neuromuscular sites developed in conjunction with the Muscular Dystrophy Campaign[
10,
11].
Centres register for the CTSR via the TREAT-NMD website. All requests are manually curated to ensure the validity of new registrants. A user agreement must be signed and registrants can define the purposes for which their data may be used. Once registered, data is entered into online forms organised by the topic categories indicated above. Sites are recruited globally by a variety of means, particularly via professional networks, including TREAT-NMD and its associated communications infrastructure, such as the TREAT-NMD newsletter which is distributed monthly to 3,500 recipients. The CTSR is also frequently publicised at conferences and meetings, and through societies with a neuromuscular focus. In a country-specific approach, accredited neuromuscular centres were approached individually through key contacts where available.
As part of the user agreement, sites are asked to indicate whether their data might be forwarded to third parties[
12]. A typical enquiry is initiated by an approach of an academic institution or a private company to the TREAT-NMD Secretariat at Newcastle University. A preliminary form records broad enquiry parameters, including timelines, target disease and genotypic criteria. In some cases, the enquiry may utilise the TREAT-NMD Global Database (an international federated database of National Patient Registries) in addition to the CTSR. Once an enquiry goes ahead, the CTSR team generates the dataset defined by the terms of the enquiry, and creates a report containing the requested information from all sites that have agreed to forward their data. After each enquiry all sites are informed about the enquiry and the data that has been provided. The requesting party receives relevant information to answer their enquiry, but does not get uncontrolled access to the full CTSR database.
Discussion
To the best of our knowledge, the CTSR is the largest and most comprehensive database of neuromuscular centres in the world. It reduces the need for researchers to collate their own datasets on neuromuscular centres, in turn preventing duplication and wasted effort. A key strength lies in the independence and inclusiveness of the database, as registration is open to any centre and the data is held for the benefit of the neuromuscular community rather than being proprietary to one organisation. The data currently held in the CTSR is not exhaustive, and further work is being carried out to improve geographical and disease coverage. A sizable cohort of patients does not attend specialist centres (Kirschner, CARE-NMD unpublished data), while in some countries a significant proportion of neuromuscular centres have not yet registered in the CTSR. For example, the number of DMD patients accounted for in the CTSR (373) attending French neuromuscular centres is much lower than reported in the national mutational database (3,373)[
18]. This may be due in part to the low number of centres registered in the CTSR from France (compared, for example, to those reported in the Orphanet Expert Centres database). In contrast, the French UMD-DMD patient registry captures all patients with a mutation in the dystrophin gene detected by any accredited genetic laboratory in France. Therefore, patients with other forms of dystrophinopathies, patients not attending neuromuscular centres and even deceased patients are accounted for[
19,
20]. In other countries such as the UK, the number of DMD patients reported in the CTSR exceeds the self-report patient registry, and the coverage of sites in the CTSR is more exhaustive. CTSR data may therefore be supplemented by other sources, including national patient registries and other surveys. However, while acknowledging these limitations, we believe that CTSR is of significant value to researchers.
The structure of the database allows the selection of centres according the specific needs of any given project. When planning a clinical trial, inclusion and exclusion criteria (e.g. age, genetic mutation, steroid and/or ambulatory status) and willingness to take part restrict the number of patients that can be recruited at an individual site. For example, even the most common genetic mutation amenable to exon-skipping (skipping of exon 51) is only found in 13% of patients[
21]. Considering these factors, and that usually at least 5 recruited patients per site are sought, trial organisers may wish to screen sites to those that see a certain minimum number of patients – for instance 50 – with the disease of interest.
Furthermore, the CTSR is flexible and extensible to serve unforeseen needs, as demonstrated by the additional utility added to the core database by project-specific additions for CARE-NMD. This project provides an example of how the CTSR platform may be extended and used for innovative research, providing benefits to the neuromuscular field beyond those initially envisaged. It also demonstrates how national collaborative efforts with patient organisations and professional societies can help ensure record accuracy. The CTSR can help to determine the level of care available, in addition to feasibility or clinical trial readiness. It is also adaptable to different national and cultural contexts, for example by displaying different questions based on a site’s country of registration. This survey reveals several important factors about neuromuscular centres and care provision in different countries. The overall CTSR cohort appears to be broadly representative of that described in previous studies. Patients with DMD represent the largest group of those in the under 18 cohort, whilst those with Myotonic Dystrophy (DM1) form the largest group of adults. In 2009 Norwood
et al. studied prevalence of neuromuscular diseases in the Northern region in the UK, and found that the five largest categories of disease were DM1 (28.1%), B/DMD (~20% taken together), FSHD (10.7%), SMA(5.1%), and LGMD (6.15%)[
22]. The overall population reported in the CTSR does not match this (17.0% DM1, 37.4% B/DMD, FSHD 9.02%, SMA 14.3%, LGMD 15.3%), tending to show a higher proportion of childhood-onset conditions and lower of adult-onset.
We believe this is due to a better representation of paediatric patients in the CTSR. This could be because adult patients are more likely to be seen by their family doctor or a neurologist, rather than attending a specialist neuromuscular centre. The better representation of children in the CTSR (as compared to adults) may also reflect the relative severity of childhood-onset conditions that might stimulate significant interest in clinical trial participation and registration. By contrast, there may be an under-diagnosis of mild or late adult-onset conditions which is reflected in lower CTSR registrations. With the exception of LGMD, the adult cohort of the CTSR shows far closer correspondence with Norwood’s data (DM1 27.8%, B/DMD 22.2%, FSHD 15.4%, SMA 8.40%, and LGMD 22.4%), which is comprehensive in its coverage of the Northern region of the UK.
However, a recurrent theme from these data is that the demographics and structure of individual neuromuscular centres, as opposed to the overall population, can vary significantly between different countries. The heterogeneous nature of neuromuscular sites is evident even between countries that might otherwise be considered similar such as the UK, Germany, and France, while centre size can vary significantly between countries with similarly sized populations. Taking DMD as a case study in the three countries with the most centres, the size of centres that report seeing these patients varies dramatically: an average of 96 patients per site in the US, 52 in the UK, and 25 in Germany. Although the US (and Australia, with an average of 142 patients per site) are perhaps special geographic cases, the UK and Germany are not dissimilar and population density would not seem to explain the difference in average numbers of DMD patients seen between these two countries. This suggests that the characteristics of sites are influenced by factors other than obvious demographic and geographic considerations, such as the structure of national health care systems. As registration in the CTSR is open to any institution, it is possible that in some countries sites have registered that are not neuromuscular reference centres but still see a small number of neuromuscular patients. As accreditation of neuromuscular centres is very different from country to country this would require more in depth analysis on a national level.
A large number of sites might improve geographic accessibility to clinical and/or trial expertise. However, it is also known that the quality of such expertise for rare conditions depends upon regular exposure, in turn implying a minimally viable number of patients seen at an individual centre. Furthermore, the additional overheads associated with conducting clinical trials across many centres with small patient populations may be prohibitive, and prevent their participation in research. For these reasons, we suggest that care is taken when planning networks of neuromuscular centres to strike a careful balance between a minimally viable centre size and sufficient provision for the population to be served.
Regarding the size of centres and their relation to care, it may be worth examining the practices of rare disease centres for other conditions such as Cystic Fibrosis (CF). For example, the 2011 UK CF Registry Annual Data Report lists 32 paediatric and 28 adult specialist centres (with some centres serving both populations)[
19]. Although these too vary in size, it is immediately apparent that most sites are significantly larger than those serving neuromuscular conditions. Although this may be in part explained by the higher prevalence of CF, it provides an opportunity to investigate the link between centre size and care, for the primary purpose of the UK CF registry is to help drive up the standard of clinical care and a great deal of clinical data is collected in order to allow this. Whilst mindful of the differences between neuromuscular care requirements and those for CF, this may be a fruitful area for further investigation. Such an approach has recently been advocated for DMD patient registries by Scully
et al.[
23].
Challenges
As with any self-report database, a significant challenge is in ensuring data accuracy. Site visits by study monitors would be an excellent but expensive way to ensure data accuracy, and are not feasible in the current setting with limited resources. Errors can arise from mistakes in data entry, though this is minimised by validation at the input stage. As time passes, inaccuracies can occur naturally as information, particularly with regards to patient cohort and site facilities, becomes outdated. This issue is mitigated in the CTSR via an update process whereby sites are contacted at least annually and asked to update their records. In addition, secondary checks are carried out upon receipt of individual enquiries: those sites within the scope of the enquiry are contacted to confirm that the data held for them is accurate.
This curation of data over time requires time and effort on the part of the CTSR managers and the sites themselves. However, the benefits to the neuromuscular field of a central source of up-to-date data on cohort and capabilities of a large number of sites outweigh the costs associated with maintaining this infrastructure. Furthermore, motivation for sites to participate and keep their data accurate is provided through the potential for them to be considered for proposed trials, and in recognition of their collaboration internationally.
Future developments
The CTSR provides a flexible and extensible platform for the neuromuscular field, and a variety of further developments are possible. At the simplest these might be minor extensions with the creation of disease-, country- or study-specific modules to capture additional information. Similarly, more sophisticated access permissions could be created to enable specific categories of users (such as trial co-ordinators or national curators) with very specific abilities to view or edit particular types of data.
In the longer term, it might be possible to link existing patient registries to the CTSR, which would enable the correlation of health information with site information. This would provide valuable insights into links between health outcomes and the capabilities of neuromuscular sites. While there may be logistical and cultural obstacles to such linkage, it would enable the integration of patient and centre-level data which would potentially allow causal links to be identified.
The CTSR platform is also applicable to other disease areas. For example, within the EU-funded NeurOmics project the CTSR is currently being extended to cover a range of neurodegenerative diseases, which will allow the integration of a large number of new centres[
24].
In addition, the CTSR might be used to capture defined criteria for national or international centres of expertise or for European Reference Networks for neuromuscular diseases.
SR works at the TREAT-NMD Alliance Secretariat, Institute of Genetic Medicine, Newcastle University.
HL is Professor of Experimental Myology, Institute of Genetic Medicine, Newcastle University, and current chair of the TREAT-NMD Alliance Executive Committee.
AT and KK are IT Specialists and Consultants at University Medical Center, Freiburg.
KG is a Project Manager at University Medical Center, Freiburg.
KB is Action Research Chair of Neuromuscular Genetics, Institute of Genetic Medicine, Newcastle University.
VS is the Harold McMillan Professor of Medicine, Institute of Genetic Medicine, Newcastle University.
RK is Director, Neuropaediatrics and Muscular Disorders, University Medical Center, Freiburg.
JK is Co-ordinator of the CTSR, University Medical Center, Freiburg.
Competing interests
All authors and affiliated institutions are involved in the TREAT-NMD Alliance. KB and VS led the TREAT-NMD Network of Excellence 1997–2011, and HL has chaired the TREAT-NMD Alliance since that date. SR is based in the TREAT-NMD Alliance Secretariat. JK leads the TREAT-NMD Clinical Trials Co-ordination Centre (CTCC) at University Medical Center Freiburg and is member of the TREAT-NMD Alliance executive committee. AT, KG, and KK are involved with running the CTSR.
Authors’ contributions
JK, HL and KB conceived of the study. AT, KG, and KK provided data for the study from the CTSR. SR analysed the data, and drafted and edited the manuscript with guidance from HL, KB, VS and JK. All authors read and approved the final manuscript.