1. Is your clinical data ready for submission?
Presented by: Gil Harari, Eyal Wultz
April 2014

2. IATI ROUNDTABLE - AGENDA
Strategy and statistical considerations in clinical trials planning (Gil Harari, MediStat)
Clinical data submission – which way to go? (Eyal Wultz, Bioforum)
Q&A

3. CLINICAL DATA SUBMISSION
Bioforum Regulatory Services
Clinical Data Standardization – importance and priority
Legacy-based submission
CDISC-based submission
Which way to go?

4. BIOFORUM SERVICES

5. BIOFORUM REGULATORY SERVICES
Clinical Programming
Regulatory Operations
Pharmacovigilence Operations
Clinical programming – study support (tables, listings, figures)
Prepare data for submissions
Documents publishing
eCTD compilation
Lifecycle management
Adverse events processing and reporting
CIOMS, MedWatch,…
PSURs, DSURs,
Over 40 submissions to the FDA, EMA, UK, Germany, Spain
Over 100 studies to date

6. Clinical data deficiencies and ‘Data Presentation’ (~39%)
THE IMPORTANCE OF CLINICAL DATA STANDARDS
There has been a sharp increase in Refuse-to-File (RTFs) issued by the FDA from 2009 onwards1
Main reasons include2 –
Clinical data deficiencies and ‘Data Presentation’ (~39%)
CMC issues (~28%)
eCTD issues (25%)
Other (8%)
1 June 2011 issue of Nature Reviews: Drug Discovery, Volume 10, page 403
2 Based on publically-available RFTs.

7. THE IMPORTANCE OF CLINICAL DATAS STANDARDS
Sanofi, Bayer’s Lemtrada receives FDA refuse-to-file letter in relapsing MS
By Lianne Dane , Created 08/27/ :35
Sanofi’s Genzyme unit on Monday announced that it received a refuse-to-file letter from the FDA regarding its application for the approval of Lemtrada (alemtuzumab) as a treatment for relapsing multiple sclerosis. "We have had constructive dialogue with the FDA, and we are very confident in our ability to address the agency’s request and resubmit rapidly," remarked Genzyme CEO David Meeker.
Genzyme, which is developing the therapy with Bayer, said regulators requested that the company "modify the presentation of the data sets to enable the agency to better navigate the application." However, the agency did not request additional data or further studies, the company noted.
The drugmaker indicated that a filing submitted to the European Medicines Agency was accepted and the review process is underway. Last week, Sanofi announced that it was withdrawing a stronger dose of the therapy, which is sold under the brand name Campath, in the US and EU to prepare for the launch of Lemtrada.

8. Submission Data Package
PREPARING CLINICAL DATA FOR SUBMISSION
Legacy Package Conversion
CDISC Conversion
Prepare clinical data for Submission
Submission Data Package

9. LEGACY-BASED CONVERSION
Legacy Package Conversion
CRT Changes
Standardization changes
Case Report Tabulation
Change study data to comply with FDA Study Data Specification Guidance
Standardize common datasets across all studies

10. CRT REQUIRED CHANGES
Formats – remove variable formats. Split formatted variables into two variables.
Labels – add labels to datasets and to variables, confirm uniqueness of variable names and labels
Dates/time format – verify consistency throughout the studies
Change variables order – according to FDA guidance
Verify the key variables (subject ID, visit) in all datasets
Split variables with value greater than 200 characters
Split datasets greater than 1GB.
Convert datasets to SAS Transport format.

11. STUDY STANDARDIZATION
Standardize common datasets across all studies
Compare common domains (demographics, adverse events, labs, physical exams etc.)
Create a standard domain structure specification given the varies studies including
Standard variable names, labels, types, value codes, order
Approach for handling clinical significance and out of range
Standardize dataset names and labels
Change each study according to specification

12. ANALYSIS DATASETS CHANGES
Analysis datasets are a must!
Apply the same standardization and CRT changes to derived variables
Add core variables (as defined by the FDA) to all datasets (study, center/site, country, treatment, sex, age, race, etc.)
Revise the analysis programs according to the updated clinical data.
Verify existing outputs (tables, listings, figures) haven’t changed

13. Submission Data Package
PREPARING CLINICAL DATA FOR SUBMISSION
Legacy Package Conversion
CDISC Conversion
Prepare clinical data for Submission
Submission Data Package

14. CDISC and its Standards
CDISC – The Clinical Data Interchange Consortium, non-profitable organization, that defines the world-wide standards for representing clinical data, required by the regulatory authorities as part of submissions
SDTM – Study Data Tabulation Model, representing the collected clinical data in standardized structure and controlled terminology.
ADaM – Analysis Data Model, representing study analysis data in defined datasets structure.
CDASH - Clinical Data Acquisition Standards Harmonisation , describes the basic recommended data collection fields for 18 domains
SEND - Standard for Exchange of Nonclinical Data, defining the structure, and format of standard nonclinical tabulation datasets

15. FDA’S VIEWPOINT ON CDISC
SDTM and ADaM are the only two standards supported at FDA for submission of study data
“By supported, we mean that the listed FDA component has established processes and technology infrastructure to support the receipt, processing, review, and archive of study data using these standards. The submission of standardized study data using any standard not listed… should be discussed with the Agency in advance”
FDA Study Data Standard Catalog
Submitting Non-standard legacy data
“The submission of non-standardized datasets is not recommended… There are no further specifications for organizing legacy datasets”
FDA Study Data Specifications, 7/18/2012

16. FDA’S VIEWPOINT ON CDISC (cont.)
The FDA has published several new guidances on standardized study data for electronic submissions
The FDA has listed the CDISC standards as the only acceptable standards to be used for study data submission
The FDA will still accept legacy-based clinical data for studies that are starting within 24 months after the FDA register their new notice (planned within the next 6 months)
After that, only CDISC-based clinical data will be accepted

17. INDUSTRY PERSPECTIVE
More sponsors are now submitting CDISC-based clinical data
However, not all CDISC-based submissions are equal…
From CDER1
“CDER has received numerous “SDTM-like” applications over the past several years in which sponsors have not followed the SDTM Implementation Guide completely”
“aspects of particular sponsor implementations have actually resulted in increased review difficulty for CDER reviewers”
1CDER Common Data Standards Issues Document, Dec 2011

18. SDTM CONVERSION – EXAMPLE
Subj. id
Date
Albumin
Alkaline Phosphatase
Leukocytes
001
01/04/12
30 g/L
398 IU/L
5.9 10^9/L
Lab - external
Subj. id
Site
Visit
Visit date
Test Category
Status
Comment
001 3
01/04/12
Chemistry
Done
Subject wasn’t in 8 hours fast
Hematology
Two samples were taken
Lab - CRF
USUBJID
VISITNUM
LBDTC
LBTESTCD
LBSTRESC
LBSTRESU
LBORNRLO
LBORNRHI
Comment 3
ALB
3.0
g/dL
3.4
5.4
Subject wasn’t in 8 hours fast
ALP
398
IU/L 44
147
WBC
5.9
10^3/uL
4.0
11.0
Two samples were taken
LB SDTM

19. DATA VALIDATION WITH CDISC
Conversion with CDISC is ‘invasive’ -> validation is rigorous
For SDTM
‘Mirror’ programming
OpenCDISC®
Manual verification
For ADaM
ADaM-SDTM comparison
Double programming for imputed variables
Compare to CTR (if applicable)

20. STUDY DATA SUBMISSION PACKAGE
Standardized clinical Data
Standardized analysis data
Analysis Programs
Define.pdf/xml for clinical and analysis data
Annotated CRF
Reviewers guide
Programs Guide

21. CDISC VS. LEGACY – SUMMARY
Pros
Less invasive -> less effort in conversion and verification
Cons
Not recommended by FDA (but acceptable in the short term in certain cases)
CDISC
FDA’s favorite
Usually more invasive
More effort performing conversion
More validation required

22. TIMES ESTIMATES CDISC – ~3-5 months Legacy – 5-10 weeks
ADaM creation and verification
months
2 weeks
3 weeks
1 week
SDTM conversion and verification
Define xml
Define xml
PKG
2-3 weeks
Analysis data conversion & verification
Legacy – 5-10 weeks
2-3 weeks
1 week
1 week
1 week
Clinical data preparation & verification
Define. pdf
Define. pdf
PKG

23. SUBMISSION OPTIONS Legacy only SDTM & ADaM
SDTM & legacy analysis datasets
The analysis datasets will be revised to use the new SDTM domains
Hybrid
Important studies – SDTM + ADaM or SDTM + Legacy Analysis
Less important studies – Legacy

24. SUBMISSION OPTIONS (cont.)
Criteria
Legacy
SDTM+ Legacy Analysis
SDTM + ADaM
Standardization and FDA Acceptance √ √√
√√√
Effort & cost
Time Required
Sponsor required Involvement
Manipulation of
Clinical Data
As recommended by the agency, sponsor should consult with the designated project manager at the FDA prior to work inception

25. Thank you

26. Legacy-based conversion

27. LEGACY CONVERSION - WORKFLOW
Create Submission Conversion Spec
Revise study clinical datasets and annotated CRFs
Revise study analysis datasets and TLG programs
Create study define.pdfs and guides

28. LEGACY CONVERSION – WORKFLOW (cont.)
Studies materials
Submission Std. & CRT Spec
Approved Spec
Project Timelines
Submission Std. & CRT Spec
Study-specific spec
Revise clinical data
Revise annotated CRFs
Submission-ready clinical data
Submission-ready clinical data
Analysis data spec
Revise analysis data
Adjust TLG programs
Submission-ready analysis programs, analysis data
Comments for derived vars
Define.pdf1
Reviewer’s guide
Analysis programs guide
Study Submission documentation
1 Two Define.pdf files are required – one for clinical data and one for analysis data

29. Study submission package
PREPARING CLINICAL DATA FOR SUBMISSION
Legacy Package Conversion
CDISC Conversion
Prepare clinical data for Submission
Study submission package

30. CDISC-based conversion

31. LEGACY CONVERSION - WORKFLOW
Create Submission Conversion Spec
Creating SDTM domains from raw data
Preparing ADaM datasets from SDTM domains
Metadata (Define.XML) and study documentation

32. SDTM Conversion Domains Programming SDTM Domains and SUPP-- aCRF
Source – target mapping
Mapping practices
Controlled terminology
SDTM Domains and SUPP--
Conversion mapping spec
SDTM aCRF
Verification reports
aCRF
Raw datasets
Including external data
Protocol / CTR
Conversion Verification
‘Mirror’ Programming
Open CDISC
Manual verification

33. SDTM Domains and SUPP--
ADaM
ADaM Datasets
Analysis programs
Verification reports
SDTM Domains and SUPP--
SAP
CTR tables
Datasets Programming
Datasets Verification
ADaM-SDTM comparison
Double programming for imputed variables
Compare to CTR (if available)

34. DOCUMENTATION & PACKAGING
Raw/SDTM domains
Same process is applicable for Legacy and CDISC conversion
Analysis/ADaM domains
Study Submission metadata and documentation
Define.xml or Define.pdf1
Reviewer’s guide
Analysis programs guide
Comments for derived vars
SAP & CTR
1 Two Define.xml/pdf files are required – one for clinical data and one for analysis

35. Thank you