The Cardiac Imaging Supplement to the TAUG-Duchenne Muscular Dystrophy v1.0 was developed using the CDISC standards development process with support from the Cardiac Working Group members on behalf of the Duchenne Regulatory Science Consortium at Critical Path Institute. This process includes collecting input from various stakeholders to ensure the standard is as thorough as possible. A number of volunteers and experts provided resources and input to support the development of this standard. The goal of this initiative is to identify a core set of clinical therapeutic area concepts and endpoints for targeted therapeutic areas and translate them into CDISC standards to improve semantic understanding, support data sharing, and facilitate global regulatory submission. As with all CDISC therapeutic area standards, the purpose of this standard is to describe how to use CDISC standards to represent data pertaining to a targeted therapeutic area.

The biomedical concepts covered in this guide were selected from concepts identified by one or more stakeholders as important, and which were not addressed or were not completely addressed by existing CDISC implementation guides. This user guide does not provide guidance on what data are needed for regulatory submission or approval; it only provides advice on how to represent data in a standard form.

To reiterate: It is important to note that the choice of biomedical concepts included in this user guide is not intended to influence sponsor decisions as to what data to collect. The examples included are intended to show how particular kinds of data can be represented using CDISC standards. This user guide emphasizes that examples are only examples and should not be overinterpreted. For guidance on the selection of biomedical concepts and endpoints, please refer to the appropriate clinical and regulatory authorities.

Clinical guidelines, articles, and other works consulted by the team during the creation of this document are referenced where appropriate, using the American Medical Association (AMA) style for citation. For the sources cited in this document and suggestions for further reading, see Appendix D, References.

1. First, read the foundational standards upon which this document is based—Study Data Tabulation Model (SDTM) v2.0, the SDTM Implementation Guides (SDTMIG v3.4 and the SDTMIG for Medical Devices (SDTMIG-MD) v1.1), Analysis Data Model (ADaM) v2.1, and the ADaM Implementation Guide (ADaMIG) v1.3—to gain some familiarity with data models and the basic rules for how they are implemented. These standards are available at http://www.cdisc.org/.
2. Next, read the Introduction to Therapeutic Area Standards and/or take CDISC's free training module TA001 - Overview of Therapeutic Area User Guides for an understanding of what to expect from this guide. 
3. Read this guide all the way through at least once.
4. Finally, revisit any sections of particular interest.

Some things to bear in mind while reading this document:

• This document does not replace or supersede the foundational CDISC standards or their implementation guides, and should not be used as a substitute for any other CDISC standard.
• This document generally does not repeat content already published in another CDISC standard.
• This document is not and does not try to be an exhaustive documentation of every possible kind of data that could be collected in relation to cardiac imaging for Duchenne Muscular Dystrophy.
• As this document ages, parts of it may become outdated. Please bear in mind the release date when contrasting advice and/or modeling in this guide against that in other CDISC standards.
• The examples in this document use CDISC Controlled Terminology where possible, but some values that seem to be controlled terminology may still be under development at the time of publication, or even especially plausible "best-guess" placeholder values. Do not rely on any source other than the CDISC value set in the National Cancer Institute Thesaurus for controlled terminology (available at http://www.cancer.gov/research/resources/terminology/cdisc).

All general caveats for therapeutic area standards given in the Introduction to Therapeutic Area Standards apply to this document.

Non-standard variables (NSVs)

This supplement follows the practices outlined in the "Alternative Representation of Non-Standard Variables" proposal (available in draft form at http://wiki.cdisc.org/x/Ui68AQ). Accordingly, SDTM-based examples containing sample data requiring the use of a variable outside the standard set of variables included in the SDTM are represented not with Supplemental Qualifier records, but with NSVs appended to the end of the parent domain, followed by sample value-level metadata for the NSVs. CDASH-based examples containing fields that collect data requiring the use of NSVs annotate their SDTMIG target as variables in the domain’s corresponding NS-- dataset, rather than as QNAM/QVAL pairs in the domain’s corresponding SUPP-- dataset (e.g., for the NSV "XXCCCC", as "NSXX.XXCCCC" rather than as "SUPPXX.QVAL where SUPPXX.QNAM = 'XXCCCC'"). In order to avoid confusion between standard variables and NSVs in the sample datasets, NSVs have been rendered visually distinct, as shown below, with white text on black in the header row, and separated from the standard variables by a small space. Note that NSVs should still be represented using SUPP-- when building SDTM datasets.

Metadata for NSVs, from the define.xml file that would accompany the submission, are tabulated below the example; only those attributes or elements that assist the example are included. (For more information on variable-level metadata in general, see Define-XML v2.1, Sections 4.3 and 5.3.2, available at https://www.cdisc.org/standards/data-exchange/define-xml). A list of all NSVs used in this document, and the variable-level metadata that might become normative for the NSVs should they be promoted to standard variables, is included in Appendix C, Non-standard Variables (NSVs).

Device Properties (DO) Domain

This supplement is based on SDTMIG-MD v1.1, which includes the DO domain. There are ongoing discussions to deprecate the DO domain in the next version of the SDTMIG-MD; however, this had not been finalized at the time of publication of this supplement.

Cardiovascular imaging tests may use different methods such as echocardiography, multigated acquisition scan, cardiac angiography, and cardiac magnetic resonance (CMR) imaging. The focus of this document is on tests done using cardiac magnetic resonance imaging.

The following concept map highlights the different components or aspects that can be represented for study purposes. The study requirements define what level of detail is required for each study. The examples illustrate how to represent the data if that level of detail is needed for the study.

Concept Map. Cardiac Imaging

The subsections listed below include examples illustrating the data representation using the SDTM for various cardiovascular imaging tests.

1. The Basic Systolic Function example includes parameters related to the overall basic cardiac function (e.g., ejection function, cardiac output, systolic and diastolic volumes).
2. The Late Gadolinium Enhancement example shows how to represent multiple CMR machines, tests related to LGE, and contrast administration.
3. The Parametric Mapping example includes information about the procedure, the CMR device used, and cardiovascular tests for parametric mapping (e.g., longitudinal relaxation time (T1 mapping) and transverse relaxation time (T2 mapping)) for different locations and segments. The example also shows how to represent imaging quality assessments and instances when a CMR test is not done. Additionally, this example shows how device information can be represented since device parameters may impact the results of these types of tests.
4. The Circumferential and Longitudinal Strain example includes the representation of an analysis method (e.g., "feature tracking").

This section illustrates how the results of several systolic function tests, and accompanying data, are represented using the SDTM.

LGE assessments can be done on each of the cardiovascular segments as shown in Figure 1.[6]

Figure 1. Left Ventricular Segments

1 = basal anterior; 2 = asal anteroseptal; 3 = basal inferoseptal; 4 = basal inferior; 5 = basal inferolateral; 6 = basal anterolateral; 7 = mid anterior; 8 = mid anteroseptal; 9 = mid inferoseptal; 10 = mid inferior; 11 = mid inferolateral; 12 = mid anterolateral; 13 = apical anterior; 14 = apical septal; 15 = apical inferior; 16 = apical lateral; 17 = apex.

The SDTM examples in this section show the findings of a CMR imaging procedure with late gadolinium enhancement. The level of granularity varies with the type of test used; research requirements will determine what level of granularity is needed.

For pediatric patients with DMD, the contrast administration procedure associated with LGE testing may be challenging. Non-contrast CMR methods, such as T1 mapping, exist and are useful for early myocardial remodeling detection. It has been found that T1 levels are typically elevated in patients with DMD, compared to healthy individuals, even prior to LGE.[8]

The following SDTM examples show T1, T2, and extracellular volume test results. This example includes device information (i.e., device manufacturer, software, version information, specific settings for each subject); however, study requirements will determine what data should be included.

Abnormal circumferential strain can be detected early in patients with DMD, despite the presence of normal ejection fraction.[9] Regular cardiac function evaluations can lead to timely treatment which may slow the progression of cardiac fibrosis.[10]

This section includes example ADaM datasets to support the analysis of statistical endpoints selected for this document. The examples in this section do not illustrate every variable that could or should be included in an analysis dataset. ADaM v2.1 (https://www.cdisc.org/standards/foundational/adam/adam-v2-1) and the ADaMIG v1.3 (https://www.cdisc.org/standards/foundational/adam/adamig-v1-3) should be referenced.

• Analysis Datasets for Ejection Fraction and NT-proBNP Endpoints
• Example Analysis Results Metadata (ARM) Tables

This section illustrates example analysis datasets for the following endpoints:

• Percent change in ejection fraction over time (1 year)
• Decline in ejection fraction over time (1 year) grouped by a decline greater than -5.0%
• Percent change in NT-proBNP over time (1 year)

Source Data

The SDTM examples used as the source data are from Section 2.1, Basic Systolic Function. The records in the CV dataset for which CVTESTCD = "LVEF_C" and "RVEF_C" were selected for analysis. The records in the LB dataset where LBTESTCD = "BNPPRONT" were used to compute the percent change over time and as a potential covariate. Both SDTM domains are used in one analysis dataset to demonstrate the flexibility of ADaM and to show that the various SDTM domains can be combined into one dataset both as rows and also as covariates.

Example Analysis Datasets

The following tables show the analysis dataset metadata and parameters used for this example.

-

ADSL

The ADSL contains subject characteristics and covariates that are important for analyses. Stratification variables can be created in the ADSL to subset the data. More than one SDTM dataset may be used as input to the ADSL. This is a simplified example ADSL dataset; the ADaMIG should be referenced for additional variables.

For this example:

• Analysis age (AAGE) was included to provide age with more precision.
• A flag indicating the use of ACE inhibitors (ACEINHFL) is shown with example derivation from the CM dataset.
• Body surface area at screening (BSASC) was derived from the VS dataset using height and weight at the screening visit. There are many possible calculations for BSA; for this example, the Du Bois method was used.[11] Note that the screening visit and visit 1 occurred at the same time in this example.

ADEFNTP

The ADEFNTP includes tests pertinent to analyses. The records from CV (CVTESTCD = "LVEF_C" and "RVEF_C") and the records from LB (LBTESTCD = "BNPPRONT") were selected. Many lab draws may occur throughout the year, but only those linked to a specific visit were included. By subsetting the LB dataset by LBLNKID not missing, laboratory data that aligns with CV data can be filtered out and sorted using USUBJID and VISIT. If LBLNKID is not available or the LB dataset does not contain VISIT, a windowing strategy could be used to select the laboratory draw closest to that visit using date of lab draw (LB.LBDTC.).

TAUGDMDCIS-8 - Getting issue details... STATUS

ADEFMRI

The ADEFMRI is an efficacy analysis dataset with only a few records for more complex modeling of changes in ejection fraction, plus the proB-type test results. It is an ADaM Basic Data Structure (BDS) dataset TAUGDMDCIS-15 - Getting issue details... STATUS , with additional variables added from ADEFNTP dataset.

The dataset can can be used for analyses with multiple covariates TAUGDMDCIS-14 - Getting issue details... STATUS by taking the lab tests of interest (PARAMCD = "BNPPRONT") or other analysis datasets, and merging variables of interest with the ADEFNTP dataset. This summary dataset is created from other ADaM datasets and only variables used in the analysis are kept.

Analysis results metadata tables may be generated for submissions to regulatory agencies as part of the Define XML. The examples here illustrate information that could be produced to describe primary and secondary endpoint outputs. They are presented as examples of the types of analyses that may be produced to explore the endpoints presented in the preceding section. These are not currently required by most regulatory agencies, although they are considered a "nice-to-have" feature. For more details on ARM, see the ADaMIG and the Analysis Results Metadata (ARM) v1.0 for Define-XML v2.0 (https://www.cdisc.org/standards/foundational/adam).

Table 1. A simple comparison of change by treatment groups over 1 year.

Table 2. This table demonstrates a model over multiple time points.

Table 3. This table demonstrates a model over multiple time points.

Table 4. This example uses the dataset ADEFMRI, with AGE, SEX, RACE, and BNPPRONT as covariates.

The following table lists the NSVs used in this document and gives their parent domain and variable-level metadata. Note that names of the NSVs are listed using the -- notation to represent the 2 letter domain name abbreviations. These NSVs may be used in other domains, if appropriate.