Designing and implementing sample and data collection for an international genetics study: the Type 1 Diabetes Genetics Consortium (T1DGC).

BACKGROUND AND
PURPOSE: The Type 1 Diabetes Genetics Consortium (T1DGC) is an international project whose primary aims are to: (a) discover genes that modify type 1 diabetes risk; and (b) expand upon the existing genetic resources for type 1 diabetes research. The initial goal was to collect 2500 affected sibling pair (ASP) families worldwide.
METHODS: T1DGC was organized into four regional networks (Asia-Pacific, Europe, North America, and the United Kingdom) and a Coordinating Center. A Steering Committee, with representatives from each network, the Coordinating Center, and the funding organizations, was responsible for T1DGC operations. The Coordinating Center, with regional network representatives, developed study documents and data systems. Each network established laboratories for: DNA extraction and cell line production; human leukocyte antigen genotyping; and autoantibody measurement. Samples were tracked from the point of collection, processed at network laboratories and stored for deposit at National Institute for Diabetes and Digestive and Kidney Diseases (NIDDK) Central Repositories. Phenotypic data were collected and entered into the study database maintained by the Coordinating Center.
RESULTS: T1DGC achieved its original ASP recruitment goal. In response to research design changes, the T1DGC infrastructure also recruited trios, cases, and controls. Results of genetic analyses have identified many novel regions that affect susceptibility to type 1 diabetes. T1DGC created a resource of data and samples that is accessible to the research community. LIMITATIONS: Participation in T1DGC was declined by some countries due to study requirements for the processing of samples at network laboratories and/or final deposition of samples in NIDDK Central Repositories. Re-contact of participants was not included in informed consent templates, preventing collection of additional samples for functional studies.
CONCLUSIONS: T1DGC implemented a distributed, regional network structure to reach ASP recruitment targets. The infrastructure proved robust and flexible enough to accommodate additional recruitment. T1DGC has established significant resources that provide a basis for future discovery in the study of type 1 diabetes genetics.

Introduction

The importance of studying diverse groups of individuals and the need for increased sample sizes to answer specific disease questions have led to the conduct of international trials and consortia in the past decade [1-16]. While some publications regarding the challenges faced in conducting an international study are available, there is the need for more published information to define potential issues and solutions.

To pool data obtained from such efforts, it is critical to standardize the collection procedures across all sites worldwide. This may prove to be a formidable task, with a variety of issues not fully appreciated from the outset of such a project. The addition of sites worldwide adds complexity and considerable time to the planning and implementation processes.

The Type 1 Diabetes Genetics Consortium (T1DGC) is an international project sponsored by the National Institute for Diabetes and Digestive and Kidney Diseases (NIDDK) and the Juvenile Diabetes Research Foundation (JDRF) whose primary aims are to: (a) discover genes that modify the risk of type 1 diabetes; and (b) expand upon the existing genetic resources for type 1 diabetes. The initial Consortium goal was to collect 2500 affected sibling pair (ASP) families throughout the world. These families would provide medical history information as well as samples for immortalized cell lines, DNA, plasma, and serum. All samples eventually will be deposited in the NIDDK Central Repositories and made available to the scientific community.

Methods

Study organization

Defining the study organization is an important first step in developing the necessary infrastructure to undertake such a project. The T1DGC has its Project Office at NIDDK and includes a Steering Committee, an External Evaluation Committee (EEC), Network Centers, Network Laboratories, Standing Committees and a Coordinating Center as well as liaisons and program observers from various National Institutes of Health (NIH) agencies and studies. Figure 1 illustrates the size and complexity of this project.

Figure 1

T1DGC organization chart.

Steering Committee

The T1DGC Steering Committee was responsible for the overall T1DGC study. Steering Committee investigators participated in the design and execution of the project and collectively approved decisions for the Coordinating Center to execute. Members included representatives from each regional Network, the Coordinating Center, and program staff from the sponsoring organizations. Decisions were made by a majority vote of a quorum of the committee members. The Steering Committee met by conference call once a month and in face-to-face meetings twice per year.

External Evaluation Committee

NIDDK established an EEC that was responsible for ongoing evaluation of the study design and monitoring the progress of the T1DGC. EEC members included investigators with relevant scientific expertise, but who were not the members of the Consortium.

Network Centers

To facilitate participant recruitment, the Consortium was organized into four regional Networks: Asia-Pacific, European, North American, and United Kingdom. The Asia-Pacific Network Center was located at the Walter and Eliza Hall Institute of Medical Research in Melbourne, Australia, and had 20 clinics. The European Network Center was located at the Hagedorn Research Institute (formerly Steno Diabetes Center) in Gentofte, Denmark, with 84 clinics. The North American Network Center was located at Benaroya Research Institute in Seattle, WA, USA, and had 62 clinics. The United Kingdom Network Center was located at the University of Cambridge and included 48 clinics. A total of 214 clinics in 34 countries participated in recruitment for T1DGC.

Each network was responsible for coordinating and monitoring all the clinic activities within the region. Each of the four networks established a network infrastructure, developing the Network Center and regional organizations through contacts with investigators and clinicians with ASP families to contribute to the collection. A Network Coordinator was appointed for each region. Each network was given flexibility to develop its region as deemed necessary for the overall success of the Consortium.

To identify participating clinics, network meetings were held to outline the T1DGC collection requirements (i.e., data and samples required for inclusion) and to determine investigator interest and feasibility of participation. Following such meetings, the regional Network Coordinator would obtain detailed clinic information, such as the estimated number of available families, staff contacts, and local or national issues that might prevent participation in the Consortium.

Each regional Network Center was responsible for coordinating and monitoring study activities within the region. Network Centers worked with investigators at participating clinics to prepare materials for submission to Institutional Review Boards (IRBs) and Ethical Committees (ECs). Network Centers performed all data entry and maintained continuous interaction with network clinics and laboratories as well as the Coordinating Center.

Laboratories

Each of the four regional networks established three types of laboratories to perform activities integral to meeting the study goals: a DNA Repository to establish cell lines and extract DNA for genotyping projects; an Autoantibody and Storage Laboratory for measurement of autoantibodies and temporary storage of serum and plasma samples; and a Human Leukocyte Antigen (HLA) Genotyping Laboratory for HLA characterization. A quality assurance (QA) plan was established and implemented; assays were standardized across each type of laboratory. Internal quality control (QC) data or any existing comparisons between laboratories were submitted to and reviewed at the Coordinating Center. All of the T1DGC laboratories participated in annual comparisons and/or QC exercises.

Other laboratories were selected for specific genotyping projects. These included the Center for Inherited Disease Research (CIDR) (Johns Hopkins University, Baltimore, MD, USA; genotyping for linkage), The Wellcome Trust Sanger Institute (Hinxton, UK; fine mapping of the major histocompatibility complex (MHC) region), and The Broad Institute Center for Genotyping and Analysis (Cambridge, MA, USA; evaluation of candidate genes for type 1 diabetes). Data from these projects were sent to the Coordinating Center for additional QC checks prior to data distribution and analyses.

Coordinating Center

The T1DGC Coordinating Center (Division of Public Health Sciences, Wake Forest University Health Sciences, Winston-Salem, NC, USA) monitored and supported data collection activities within the four Network Centers. Since the regional Network Centers were charged with coordinating and monitoring all clinic activities within the region, the Coordinating Center interacted only with the Network Centers and not with individual clinics in a region.

The Coordinating Center established and maintained QA standards for all activities of the study and worked with the Network Centers and laboratories to implement decisions made by the T1DGC Steering Committee. In addition, the Coordinating Center was responsible for fiscal administration of the project.

Three specialized teams were established, each focusing on specific aspects of the study (i.e., Operations, Systems, and Statistics). The Operations Team, in collaboration with regional representatives, developed all study materials, including: a template for informed consent; the protocol; a manual of operations; and data collection forms for ASP, trio, and case–control collections. Revisions to each of these documents were implemented as needed. Included in the manual of operations were figures to provide visual references for key aspects of the data and sample collection (Figures 2 and 3). The Systems Team was responsible for data flow, architecture, and security. This team developed and finalized two study websites (T1DGC public site (www.t1dgc.org) and an internal T1DGC data entry site for certified Network Center and laboratory personnel) as well as other fully web-based applications, including a specimen tracking system and a HLA genotyping laboratory system. The Statistics Team was responsible for data management, QA/QC, data set creation and distribution, and initial analyses.

Figure 2

T1DGC data and sample collection flow.

Figure 3

T1DGC blood collection chart.

Standing Committees and Working Groups

Ten standing committees were established to implement Consortium activities and provide opportunities for T1DGC members to participate. Each committee included representatives from the four regional networks, the Coordinating Center, and the sponsoring organizations. T1DGC committees included: Access; Bioinformatics; Ethical, Legal, and Social Implications (ELSI); Network Coordinators; Phenotyping/Recruitment (including eligibility review and approval); Publications and Presentations; and four QC Committees (Autoantibody, DNA Repository, HLA Genotyping, and Forms Data). Monthly calls with each of the QC Committees were used to review QC reports and to discuss laboratory-specific issues. Other committees scheduled calls as required to deal with specific study issues. Face-to-face meetings of all T1DGC committees were held annually.

In addition to the Standing Committees, T1DGC established two Working Groups (MHC and Rapid Response) to analyze data associated with two genotyping projects. Each group comprised experts in the specific regions that were genotyped.

Training, certification, and pilot studies

Cultural and language differences made study-wide, central training sessions difficult. T1DGC used a ‘train the trainer’ model where Network Center staff members were trained at the Network Center by Coordinating Center staff. The Network Coordinator, in turn, was responsible for subsequent training of clinic staff, either centrally or individually. This model enabled networks to initiate data collection on a staggered timetable.

Following training, each participating clinic was required to conduct a pilot study before initiating T1DGC data collection. Data were reviewed by a Coordinating Center Project Manager who certified or provided final approval for the clinic to begin T1DGC participant recruitment. All data collection forms were data entered at the Network Center by staff trained and certified in the data entry system.

Quality control

The Coordinating Center established QA procedures and QC metrics for all Consortium activities. These activities included the data collection forms entry [17], sample assays for the Network Laboratories [18-20], and any genotyping performed on the samples [21,22]. T1DGC samples were deposited at the NIDDK Central Repositories and the Coordinating Center worked with NIH staff to assure that samples received were of high quality.

Results and lessons learned

Organization

The T1DGC Steering Committee was responsible for the overall T1DGC study and members actively participated in the design and execution of the project. From the outset, T1DGC decided that a distributed organization of regional networks was necessary to complete a worldwide recruitment of 2500 ASP families. Four regional Networks (in Asia-Pacific, Europe, North America, and the United Kingdom) were organized, with the aim to ensure standardized collection procedures across all sites worldwide. This proved to be a formidable task, with a variety of issues not fully identified from the outset of such a project.

An important component of the success of this consortium was the development of the T1DGC Consortium Agreement (Appendix 1) that incentivized investigator participation. This agreement clearly defined the activities of the Consortium and member rights and responsibilities. The agreement acknowledged contributing investigators and explicitly respected their research prerogatives. It also outlined a timeline for making T1DGC resources available to contributing investigators, to T1DGC members, and to the broader research community.

Recruitment

To facilitate worldwide recruitment, each network was given flexibility to develop its own Network Center and regional organization to meet the overall participant recruitment goals. While this resulted in very different approaches across the four networks, it led to the overall success of Consortium recruitment, as each network could deal with the unique social, cultural, ethical, and legal issues of different countries. This flexible approach proved to be an effective and successful strategy.

Network meetings were a key factor in facilitating interaction among participating investigators within networks. For example, the initial network meetings outlined the T1DGC collection requirements and determined investigator interest and feasibility of participating. Subsequent network meetings, generally on an annual basis, provided updates on the status of recruitment, activities of the Consortium, and new developments in type 1 diabetes genetics. At some subsequent network meetings, additional training was provided for areas of the study that required more emphasis and training for new aspects of the study was conducted.

The centralization of some activities combined with the delegation of other activities contributed to the smooth running of the Consortium and the success of recruitment. Initially, T1DGC collected only ASP families. Later, on the recommendation of the Steering Committee and the approval of the EEC, T1DGC included the recruitment of trios (father, mother, and a child with type 1 diabetes), as well as cases (with type 1 diabetes) and controls (no history of type 1 diabetes) from populations with a low prevalence of the disease. In Asia-Pacific, these included individuals from India, Thailand, Malaysia, Philippines, and Singapore. In Europe, Cameroon was included to provide trios, cases, and controls. In North America, Mexican-American, and African-American individuals were included. Table 1 provides a summary of the T1DGC recruitment and basic demographics for eligible participants as of July 4, 2009. Recruitment and data cleaning are ongoing.

Table 1

Demographics of completed affected sibling pair families, trios, cases, and controls by network, T1DGC, July 4, 2009

	Asia-Pacific	European	North American	United Kingdom	Overall
Affected sibling pair families
Number completed families	324	1215	1153	163	2855
Gender (percent)
Male	46.6	48.6	49.3	45.2	48.4
Female	53.4	51.4	50.7	54.8	51.6
Race (percent)
American Indian/Alaskan Native	0.0	0.0	0.2	0.0	0.1
Asian	6.7	0.0	0.7	1.3	1.2
Native Hawaiian or other Pacific Islander	1.3	0.0	0.2	0.0	0.2
Black or African American	1.6	0.2	2.2	0.7	1.2
White or Caucasian	90.3	99.8	96.7	97.9	97.3
For affected participants (mean ± SD[a])
Age at ascertainment (proband)	21.7 ± 12.4	25.6 ± 13.3	21.3 ± 12.9	16.7 ± 6.4	22.9 ± 13.0
Age at diagnosis (proband)	8.0 ± 6.3	9.7 ± 7.1	7.6 ± 5.7	6.2 ± 4.3	8.5 ± 6.4
Age at ascertainment (affected siblings)	20.0 ± 12.2	24.1 ± 13.2	19.7 ± 13.0	14.2 ± 5.8	21.3 ± 13.0
Age at diagnosis (affected siblings)	12.8 ± 8.4	14.8 ± 9.0	11.5 ± 7.9	8.8 ± 4.4	12.9 ± 8.5
Trio families
Number completed families	269	11	192	N/A[b]
Gender (percent)
Male	48.5	60.0	47.5	N/A	48.2
Female	51.5	40.0	52.5	N/A	51.8
Race (percent)
American Indian/Alaskan Native	0.0	0.0	0.0	N/A	0.0
Asian	100.0	0.0	0.0	N/A	57.5
Native Hawaiian or other Pacific Islander	0.0	0.0	0.0	N/A	0.0
Black or African American	0.0	100	48.0	N/A	21.2
White or Caucasian	0.0	0.0	52.0	N/A	21.3
For affected participant (mean ± SD)
Age at ascertainment (proband)	16.5 ± 7.4	14.7 ± 4.8	11.1 ± 4.6	N/A	14.3 ± 6.9
Age at diagnosis (proband)	10.2 ± 5.4	11.3 ± 4.2	7.2 ± 4.0	N/A	9.0 ± 5.0
Cases
Number completed	4	0	390	N/A	394
Gender (percent)
Male	100.0	0.0	46.0	N/A	46.5
Female	0.0	0.0	54.0	N/A	53.5
Race (percent)
American Indian/Alaskan Native	0.0	0.0	0.0	N/A	0.0
Asian	100.0	0.0	0.0	N/A	1.0
Native Hawaiian or other Pacific Islander	0.0	0.0	0.0	N/A	0.0
Black or African American	0.0	0.0	77.7	N/A	76.9
White or Caucasian	0.0	0.0	22.3	N/A	22.1
Age at ascertainment (mean ± SD)	21.8 ± 3.9	0.0	14.8 ± 7.7	N/A	14.9 ± 7.7
Age at diagnosis (mean ± SD)	14.2 ± 2.9	0.0	8.9 ± 5.4	N/A	8.9 ± 5.5
Controls
Number completed	2	0	527	N/A	529
Gender
Male	100.0	0.0	23.6	N/A	23.9
Female	0.0	0.0	76.4	N/A	76.1
Race
American Indian/Alaskan Native	0.0	0.0	0.0	N/A	0.0
Asian	100.0	0.0	0.0	N/A	0.4
Native Hawaiian or other Pacific Islander	0.0	0.0	0.0	N/A	0.0
Black or African American	0.0	0.0	81.0	N/A	80.7
White or Caucasian	0.0	0.0	19.0	N/A	18.9
Age at ascertainment (mean ± SD)	23.0 ± 1.4	0.0	32.8 ± 12.8	N/A	32.8 ± 12.8

SD – standard deviation. bN/A – not applicable.

The protocol, manual of operations, and data collection forms were developed centrally at the Coordinating Center, with input from network representatives. All study documents were made available on the T1DGC website (www.t1dgc.org).

T1DGC standardized supplies and services worldwide by establishing central billing accounts and using vendors that would permit clinics to order from a common catalog of supplies. Central billing accounts were created for: blood collection supplies to be used in the clinics (Sarstedt, Inc.); fetal bovine serum to be used in establishing cell lines in the Network DNA Repositories (Invitrogen, Inc.); and couriers to ship specimens from the clinics to laboratories and from DNA Repositories to genotyping facilities (Federal Express and World Courier). Locating vendors and establishing the master accounts took considerable time and effort, so the decision to pursue this type of arrangement should be made as early as possible in the planning process to avoid delays in data collection. For instance, when T1DGC realized that there could not be a single worldwide courier for shipping samples, an account with Federal Express was used for shipments within North America and another account with World Courier was used for shipments in the Asia-Pacific and European Networks. In the United Kingdom, no courier master account was required as the postal system was used for shipping cell line samples to Cambridge and a local van courier for frozen shipments to the laboratory in Bristol.

Regulatory issues

Every institution engaged in human subjects research supported or conducted by the US Department of Health and Human Services (DHHS) must obtain an assurance of compliance approved by the Office for Human Research Protections (OHRP). Some international institutions did not have an active Federal Wide Assurance (FWA) number and this was a primary cause of delayed recruitment in clinics. Some networks overcame this issue by using an Unaffiliated Investigator Agreement, where one institution agreed to serve as an umbrella for other collection sites. Clinic sites also were encouraged to register their IRB and apply for an FWA number online at the OHRP website. As with other studies, obtaining IRB or EC approval was the major source of delay in initiating recruitment, as each IRB or EC had its own set of requirements.

The T1DGC ELSI Committee dealt with the large number of issues related to informed consent [23]. This group finalized a set of templates (self consent, parental consent, teenage assent, and child assent) that was agreed to by all networks. Templates could be modified to comply with local IRB or EC requirements, as long as a defined set of specific elements required for the T1DGC collection were included in the final approved version.

Dealing with informed consent language is a time-consuming task in any study, but was particularly so in T1DGC, given the diverse requirements necessary to satisfy hundreds of IRBs or ECs. Particular to a genetics study, T1DGC had to be sensitive to specific cultural issues about the collection of genetic material and to reassure investigators from countries who felt that genetics collection was primarily an exploitive activity. T1DGC added language to the consent templates to specifically state that the Consortium would not claim any intellectual property rights, sell the DNA, or develop any commercial products.

In North America, several US IRBs required the T1DGC to apply for and obtain a Certificate of Confidentiality. To ensure compliance with the Health and Insurance Portability and Accountability Act (HIPAA), the Coordinating Center developed and executed Data Use Agreements for transfer of data between each of the Network Centers and the Coordinating Center.

Study communication

In general, T1DGC committees had monthly conference calls throughout the study. Email was the primary means of communication between the Coordinating Center and the Network Centers, especially in the intervals between conference calls. Face-to-face meetings of committee members occurred annually.

T1DGC greatly benefited from web-based communications. There were two study websites: a public site with login access for Consortium Members and an internal data entry website used for input and access to all study data that was accessible only to specified study personnel. Since data were available to all Network Coordinators, real-time monitoring of recruitment was possible and was used as an incentive to spur recruitment efforts. Network laboratories used the data entry site to report their results to the Coordinating Center. The T1DGC also developed a web-based application for the HLA Genotyping Laboratories to report their results [20].

The T1DGC website (www.t1dgc.org) was used to communicate with the general T1DGC membership and the public. Specific pages of the website were used for communications with different T1DGC committees and working groups, including the Steering Committee. The Consortium Agreement, access policies, and copies of data collection forms were all available on the T1DGC website.

The T1DGC Steering Committee appreciated that having standardized assays and/or methods across network laboratories would be critical for the success of the study. The T1DGC study data are primarily of two types: medical history information recorded on data collection forms and laboratory results. Since there were three types of laboratories within each network, the central QC Committee was comprised of four subcommittees: (1) Forms Data; (2) DNA Repositories; (3) Autoantibody Laboratories; and (4) HLA Genotyping Laboratories.

The QC Committee and the Coordinating Center developed QA procedures, including a central manual, for use in all networks. To implement QA, the QC Committee reviewed internal QC data or any existing comparison data among the network laboratories and also conducted annual comparisons of laboratories worldwide. This ensured consistency and allowed the study to monitor for assay drift.

The QC Committee conducted site visits to each regional Network Center and Network Laboratory to monitor adherence to the protocol under normal operating conditions. Site visits also were used to identify and resolve any data collection issues at individual clinics and/or any questions about sample shipments, handling, and analysis procedures at the laboratories. It proved challenging, but possible, to train and to impose rigorous QA procedures across networks. Again, insistence on uniform standards, with flexibility on specific details, was critical to implementing the established QA standards.

Access

One of the main goals of the T1DGC was to share its data, samples, and resources with the broader research community. This goal was prominently stated in the Consortium Agreement and was implemented by the Access Committee. The T1DGC access policy is available on the website (www.t1dgc.org) and included as Appendix 2. Importantly, there is a prominent banner on the website that highlights data and sample availability. There is also a list of all investigators who have been provided access to T1DGC resources (samples and/or data).

The T1DGC is depositing samples and data in all three NIDDK Central Repositories (Biosample, Genetics, and Data). The Central Repositories were established to expand the utility of NIDDK-supported studies by allowing the research community to continue to access these materials beyond the end of the study.

T1DGC conducted training workshops for HLA genotyping and for bioinformatics in a conscious effort to export technology, providing hands-on opportunities for T1DGC members. Since real T1DGC samples and data were used, these workshops also highlighted the availability of the resources.

Genotyping

All genotyping of T1DGC samples was performed centrally, although different laboratories were used for different aspects of the study. The Coordinating Center worked with the Network DNA Repositories to ship samples to the selected facility. Genotype data were returned to the Coordinating Center, where the Statistics Team was responsible for performing additional QC checks and initial analyses. The results of all analyses undertaken by T1DGC have been made available to the T1DGC membership and announced on the T1DGC website.

The initial T1DGC activity was a joint analysis of three historical genome-scan data sets (UK, US, and Scandinavia), combined with data from 254 T1DGC ASP families that had been genotyped by CIDR. These results were published as an Original Article in Diabetes [24]. This was followed by genotyping all T1DGC-collected ASP families, with genotyping performed by CIDR and the results published in Diabetes [25].

The T1DGC subsequently genotyped ∼10,000 samples at The Wellcome Trust Sanger Institute to generate a data set of single nucleotide polymorphisms (SNPs) and microsatellites within the 4 Mb classical MHC region. This information was combined with HLA class I and class II genotyping performed in the T1DGC HLA Genotyping Laboratories. This comprehensive, unique data resource was made available to multiple analytic working groups. Their results are presented in a supplement of Diabetes, Obesity and Metabolism [26].

The same ∼10,000 DNA samples have been used to investigate previously reported candidate genes for type 1 diabetes, to confirm the most highly associated SNPs reported by the Wellcome Trust Case Control Consortium (WTCCC) [27], to study recently reported genes that contribute to risk of type 2 diabetes and are implicated in β-cell function, and to interrogate recently reported genes from genome-wide association studies (GWAS) of other autoimmune diseases. Genotyping for these projects was performed at The Broad Institute and the results are presented in a supplement of Genes and Immunity [28].

The T1DGC has completed Stage 1 of a GWAS, using ∼500,000 SNPs in 4000 cases from the JDRF/Wellcome Trust British case collection and 2500 controls from the British 1958 Birth Cohort (B58C). These results were combined with existing data from the WTCCC study of type 1 diabetes [27] and from the Genetics of Kidneys in Diabetes (GoKinD) Study [29]. Genotyping was by contract to Illumina and Professor David Clayton (University of Cambridge, UK) led the analysis. The results have been published in Nature Genetics [30]. Follow-up studies to the GWAS Stage 1 currently are underway and are utilizing T1DGC samples.

Limitations

The original T1DGC research plan for a linkage study guided the decision to collect ASP families. From the outset, the Steering Committee recognized that the rapid development of genotyping technologies would make it necessary to anticipate modifications in the research design. Indeed, one rationale for establishing the EEC was to be able to obtain peer review of any proposed design changes and solicit agreement from the sponsoring organizations for implementation of these changes.

While the T1DGC infrastructure was established to collect ASPs, flexibility to accommodate design changes was built in, to the extent possible. This was especially true with respect to the language of study materials provided to the IRBs. The approval for recruitment of trios and case–control collections added complexity to the study. New data collection forms had to be designed and additional training had to be initiated. However, most IRBs considered the addition of new cohorts as a modification of the approved protocol and provided expedited review.

Although the T1DGC established four regional networks, the designation of networks was arbitrary (to some extent) and resulted from early participation of investigators from those regions in the planning of the project. Unfortunately, T1DGC could not accommodate investigators in regions or countries who wanted to have their own network. Each regional Network in turn designated a set of network laboratories. T1DGC insisted on rigorous QA procedures and was able to achieve and maintain quality performance across the different network laboratories. Investigators from some countries declined to participate in T1DGC, arguing that they had laboratories and technologies in their own countries and refusing to send samples to a central location. Finally, it was a condition for funding that samples collected for T1DGC must be exported and deposited in the NIDDK Central Repositories. These conditions prevented certain countries (e.g., China, Japan and Korea) from participating. Recruitment for a future worldwide genetics project might take these limitations into account and begin with pilot collections in several nations to increase confidence and accommodate differences.

As a genetics study (and not, for example, a clinical trial), T1DGC planned no sustained contact with the participants. In some networks, no re-contact was explicitly stated in the consent form. As data on T1DGC participants accumulates, it would have been useful to be able to re-contact participants for collection of additional samples to carry out functional studies.

Conclusions and recommendations

The T1DGC is a NIDDK- and JDRF-sponsored project whose primary aims are to: (a) discover genes that modify risk of type 1 diabetes and (b) expand upon existing genetic resources for type 1 diabetes research. T1DGC set an ambitious recruitment target of 2500 ASP families worldwide and established the organization and infrastructure that completed this recruitment and also collected additional trios, cases, and controls.

T1DGC’s four regional Networks (Asia-Pacific, European, North American, and United Kingdom) were responsible for coordinating recruitment activities within each region, but were given the flexibility to develop that region as deemed necessary for the overall success of the Consortium. This flexibility meant that each network could deal sensitively with the particular social, cultural, ethical, and legal issues of their different countries.

Standardized data collection was an overarching goal of the T1DGC – even with worldwide recruitment, sample handling, and analysis. The T1DGC Coordinating Center monitored and supported the activities within the four Network Centers. The protocol, manual of operations, and forms were developed centrally at the Coordinating Center, with input from network representatives. The Coordinating Center developed QA procedures and implemented quality monitoring for all networks, including the network laboratories. Good communication and expedient problem solving are key requirements for success.

The T1DGC data and sample collection includes ASP families, trios, cases, and controls. Results of genetic analyses have identified many novel new regions that affect susceptibility to type 1 diabetes [24-30]. T1DGC data and samples are accessible to the research community and should prove to be particularly rich resources well into the future.

Member	Institution	Location	Country
Beena Akolkar	NIDDK	Bethesda, MD	USA
Pat Concannon	University of Virginia	Charlottesville, VA	USA
Henry Erlich	Roche Molecular Systems/CHORI	Oakland, CA	USA
Cecile Julier	Centre National de Genotypage	Paris	France
Grant Morahan	Western Australian Institute for Medical Research	Perth	Australia
Jorn Nerup	Hagedorn Research Institute	Gentofte	Denmark
Flemming Pociot	Hagedorn Research Institute	Gentofte	Denmark
Stephen Rich (Chair)	University of Virginia	Charlottesville, VA	USA
John Todd	University of Cambridge	Cambridge	UK

Network	PI	Institution	Country
Asia-Pacific	Peter Colman	Walter & Eliza Hall Institute	Australia
	Grant Morahan
European	Jorn Nerup	Hagedorn Research Institute	Denmark
	Flemming Pociot	Hagedorn Research Institute	Denmark
North American	Carla Greenbaum	Benaroya Research Institute	USA
United Kingdom	John Todd	University of Cambridge	UK

Submission to CIDR	Data received from CIDR	Expected release of aggregate data analysis*
January 2003	October 2003	+6 months
December 2004	August 2005	+6 months
December 2005	August 2006	+6 months
December 2006	August 2007	+6 months

The Consortium expects to release the results of the aggregate analysis incorporating the samples from each submission according to the approximate schedule above.

TIMETABLE:

Samples and data may be requested (for each participant provided) by Contributing Investigators.

6 months later →

Access to samples/data may be requested by T1DGC members.

12 months later →

Access to samples/data may be requested by non-members.

Date of submission:

Resource Requested: (Mark all that apply.)

Renewable Resources:

□ DNA (5 mcg aliquot)

□ Data (Specify requested data) set[s]: _______________________________________

NOTE: Data set name is a required field. See list of “Available Data Sets and Samples” under “Access to T1DGC Data and Samples” link onwww.t1dgc.org.

Non-Renewable Resources:

□ Whole Genome Amplified DNA (5 mcg aliquot)

□ Serum (0.5 mL aliquot)

□ Plasma (0.5 mL aliquot)

Project title:

Corresponding investigator and full contact information: Name: Address: Telephone: FAX: E-mail:

Name(s), affiliation(s) and address(es) of major co-investigator(s) and/or collaborator(s):

Name(s), affiliation(s) and address(es) of project analyst(s):

Abstract (250 words or less):

Specific aims:

Previous peer review of the project. Indicate whether the genetic components of the project have undergone previous peer review, and by whom. Indicate the outcome of the review. If an application is currently pending at the NIH, provide details about the study section and institute assignment, if known.

Source(s) of funds for the project. If no new funds are required, this should be stated. If funded by the NIH, list the sponsoring institute and the dates of support. If approval is sought conditional on the applicant’s success in obtaining funding, a specific timeline for this must be provided.

Number of samples to be analyzed and the projected timeline for obtaining the samples from T1DGC. (For non-renewable resources, a formal justification for the requested number of samples must be provided, including power analyses. This is critical for access to the limited plasma and serum samples.)

Brief outline of the plan for the next phase of the project if linkage or association is found (if applicable). Include specific plans for isolating the locus (loci) and name the individuals responsible for each step. Attach letters of collaboration from these individuals.

Description of core data required from the T1DGC central databases, including process and phenotypic data.

Measures to ensure the security of specimens and data. In addition, plans for disposal of any unused specimens and data must be described.

Background information about the disease/trait including the rationale for carrying out this particular study. Describe any unique features about the disease/trait that would single out this project for special consideration.

Analysis strategy for the resulting information and choice of analytic methods and software. If collaborations are established for analytical services, include letters of collaboration.

Assurance that the project has been reviewed for human subject protection by an appropriate Institutional Review Board (IRB) or Ethics Committee (EC).

Description of any commercial aims and likely benefits ensuing from the project. Details of pending or granted patents relevant to the application must be provided.

If samples are requested, please provide shipping contact information.

Name of Contact:

Shipping Address:

E-mail Address:

Phone Number:

Have all of co-investigators and collaborators approved the final version of this application? □ YES □ NO

Is there a deadline for submission of this material to an external agency? □ YES □ NO

If yes, what is the deadline and when would you like comments back? __________/__________/___________

DD/MM/YYYY

I have read and agree to abide by the T1DGC Access Policy, the T1DGC Publications and Presentations Policy, and the consent guidelines conferred by study participants.

□ YES □ NO

I have signed and submitted the T1DGC Confidentiality Certification to the Network Center or the Coordinating Center.

□ YES □ NO

All individuals who will have access to the data and/or samples have submitted the T1DGC Confidentiality Certification to the Network Center or the Coordinating Center.

□ YES □ NO

I agree to submit all results from analysis of these samples to the NIDDK Central Repository for incorporation among the T1DGC data holdings (including information on quality control methods).

□ YES □ NO

I understand that some samples will be restricted to non-commercial use only.

□ YES □ NO

Applications for non-renewable resources must include submission of the following items:

• CV(s) for key personnel involved in the project (NIH format required)

• Letter(s) of support/commitment from major collaborator(s) and/or co-investigator(s)

• Essential reprints or preprints (no more than 3)

Submission of these items is optional for applications for renewable resources.