1. CONFIDENTIAL © 2012 | 1 Writing a Statistical Analysis Plan DIA Medical Writing SIAC July 12, 2012 Peter Riebling, MS, RAC Associate Director, Regulatory and Medical Writing PTC Therapeutics, Inc.

2. CONFIDENTIAL © 2012 | 2 Outline  Definition of a statistical analysis plan  Purpose and audience  When to write the SAP  Regulatory guidelines  Content of the SAP

3. CONFIDENTIAL © 2012 | 3 What is a Statistical Analysis Plan (SAP)?  ICH E9 definition:  “A statistical analysis plan is a document that contains a more technical and detailed elaboration of the principal features of the analysis described in the protocol, and includes detailed procedures for executing the statistical analysis of the primary and secondary variables and other data”  PSI definition:  “The statistical analysis plan is intended to be a comprehensive and detailed description of the methods and presentation of data analyses proposed for a clinical trial, in order to avoid post-hoc decisions that may affect the interpretation of the statistical analysis” (emphasis added)  PSI is a UK-based society of pharmaceutical industry statisticians  http://www.psiweb.org

4. CONFIDENTIAL © 2012 | 4  The SAP supplements and supersedes the statistics section of the protocol, pre-specifying the analyses to be conducted and the details regarding how the analyses will be conducted  Allows the statistics section of the protocol to be brief  “The extent to which the…primary analysis is planned a priori will contribute to the degree of confidence in the final results and conclusions of the trial” (ICH E9)  Sometimes, post-hoc analyses are justifiable and valid  Nonetheless, pre-specified analyses are generally viewed as more acceptable  The SAP provides instructions to the SAS programming team  May include an appendix showing all planned tables, listings, and figures (TLFs) and/or mock TLF examples Purpose(s) of a Statistical Analysis Plan

5. CONFIDENTIAL © 2012 | 5  Regulatory authorities  Regulatory authorities expect clarity with regard to the sponsor’s prospective plan for analyzing clinical trial data  FDA may provide comments on the SAP, potentially leading to revision  SAS programmers (internal or CRO)  The SAP provides guidance to be followed in the creation of TLFs  Internal personnel  Opportunity for cross-functional team members (eg, biostatistics, clinical, medical writing, regulatory) to reach consensus Audience(s) for a Statistical Analysis Plan

6. CONFIDENTIAL © 2012 | 6  After the protocol is final, but before the data are unblinded  Writing the SAP before or soon after first patient in has certain advantages:  Details of the protocol are fresh and may be forgotten later, especially for long-term studies  Writing the SAP involves careful consideration of issues such as handling of dropouts and missing data, which may prompt a desire to alter the study design if it is not too late  Writing the SAP can inform the development of the case report form  The SAP can be revised later if necessary  ICH E9 suggests that a blinded data review might lead to revisions  Time for submission to FDA, receipt of comments from FDA, and potential revision of the SAP should be considered also A Matter of Timing – When to Write the SAP

7. CONFIDENTIAL © 2012 | 7  E3 (Clinical Study Report)  Section 9.7.1 “Statistical and Analytical Plans”  “The statistical analyses planned in the protocol and any changes made before outcome results were available should be described.”  Does not directly address the writing of a SAP but provides some indication of statistical issues that regulatory authorities are particularly interested in and that should be addressed in a SAP  Whether the primary analyses includes adjustment for covariates  Whether it was planned to exclude patients from analysis  Whether it was planned to analyze a particular subgroup ICH guidelines

8. CONFIDENTIAL © 2012 | 8  E9 (Statistical Principles for Clinical Trials)  Section 5.1 “Pre-specification of the Analysis”  “The statistical analysis plan may be written as a separate document to be completed after finalising the protocol…The plan should be reviewed and possibly updated as a result of the blind review of the data (Section 7.1) and should be finalised before breaking the blind. Formal records should be kept of when the statistical analysis plan was finalised as well as when the blind was subsequently broken. If the blind review suggests changes to the principal features stated in the protocol, these should be documented in a protocol amendment. Otherwise, it will suffice to update the statistical analysis plan with the considerations suggested from the blind review.” ICH guidelines

9. CONFIDENTIAL © 2012 | 9  E9 (Statistical Principles for Clinical Trials)  Section 7.1 “Evaluation and Reporting”  “When the conduct of the trial is over and the data are assembled and available for preliminary inspection, it is valuable to carry out the blind review of the planned analysis…This pre-analysis review, blinded to treatment, should cover decisions concerning, for example, the exclusion of subjects or data from the analysis sets; possible transformations may also be checked, and outliers defined; important covariates identified in other recent research may be added to the model; the use of parametric or non-parametric methods may be reconsidered. Decisions made at this time should be described in the report, and should be distinguished from those made after the statistician has had access to the treatment codes, as blind decisions will generally introduce less potential for bias..” ICH guidelines

10. CONFIDENTIAL © 2012 | 10  E9 (Statistical Principles for Clinical Trials)  Section 7.1 “Evaluation and Reporting”  “Many of the more detailed aspects of presentation and tabulation should be finalised at or about the time of the blind review so that by the time of the actual analysis full plans exist for all its aspects including subject selection, data selection and modification, data summary and tabulation, estimation and hypothesis testing.” ICH guidelines

11. CONFIDENTIAL © 2012 | 11  Minimum required sections  Title page  Signature page  Table of contents  Section 1: List of abbreviations  Section 2: Introduction  Section 3: Study objectives Content of the SAP

12. CONFIDENTIAL © 2012 | 12  Minimum required sections (cont’d)  Section 4; Study design  Section 5: Planned sample size  Section 6: Study endpoints  Section 7: Analysis populations  Section 8: General statistical considerations  Section 9: Summaries of study population Content of the SAP

13. CONFIDENTIAL © 2012 | 13  Minimum required sections (cont’d)  Section 10: Efficacy analyses  Section 11: Safety analyses  Section 12: Changes from the protocol  Section 13: References  Don’t necessarily have to follow this outline precisely, but all of information should be included Content of the SAP

14. CONFIDENTIAL © 2012 | 14  Optional sections  Additional non-efficacy/safety analyses, such as:  PK/PD analyses  Quality of life analyses  Microbiology analyses  Separate section on handling of dropouts, eg, imputations  A section describing any planned interim and/or data monitoring committee analyses Content of the SAP

15. CONFIDENTIAL © 2012 | 15  Title page  Should at least include “Statistical Analysis Plan”, protocol number, study title, sponsor, and CRO if applicable  Signature page  Study biostatistician (required)  Study clinician (required)  Other study team members (optional) Content of the SAP

16. CONFIDENTIAL © 2012 | 16  Table of contents  Section 1: List of Abbreviations  Section 2: Introduction  States purpose of the SAP  States that the SAP supersedes the protocol in the event of conflicts  May briefly describe whether and how blinded data were used to refine the planned analysis in the protocol  May provide a brief version history  Section 3: Objectives  Primary objective  Secondary objectives Content of the SAP

17. CONFIDENTIAL © 2012 | 17  Section 4: Study design  Overview of study design, preferably including figure  Description of study population  Description of study randomization and blinding procedures  Description of timing of study assessments  Emphasis should be on issues that are pertinent to statistical analysis Content of the SAP

18. CONFIDENTIAL © 2012 | 18  Section 5: Sample size  Thorough description of the sample size calculation based on the primary endpiont, including:  Clinical rationale used in the assumptions of point estimates and variance  Details regarding how the calculations were derived (eg, software used)  Sample size calculations based on secondary endpoints, if any Content of the SAP

19. CONFIDENTIAL © 2012 | 19  Section 6: Study endpoints  Primary endpoint(s)  Secondary endpoints  Section 7: Analysis Populations  All analysis populations (eg, ITT, per-protocol, safety) should be defined Content of the SAP

20. CONFIDENTIAL © 2012 | 20  Section 8: General statistical considerations  Description of basic statistical considerations for the planned analyses, including:  What descriptive statistics will be reported for continuous (eg, mean, median, standard deviation) and discrete (eg, percentage) variables  General details of estimation and hypothesis testing  Formatting details (eg, number of decimal places), which provide guidance to SAS programming team Content of the SAP

21. CONFIDENTIAL © 2012 | 21  Section 9: Summaries of Study Population  Description of plans for summarization of subject disposition  Description of plans for summarization of demographic and baseline characteristics  Description of plans for summarization of other data that require only descriptive statistics (eg, medical history, concomitant medications, study drug compliance) Content of the SAP

22. CONFIDENTIAL © 2012 | 22  Section 10: Efficacy analyses  Details regarding the planned analyses for each efficacy endpoint, including:  The hypothesis or estimation assessing the endpoint  Statistical method to be used  Any multiplicity adjustment methods to be implemented, if applicable  Any departures from the general statistical considerations as stated in Section 8  Any planned subgroup analyses Content of the SAP

23. CONFIDENTIAL © 2012 | 23  Section 11: Safety analyses  Description of plans for summarization of the extent of exposure to study drug  Details regarding the planned analyses for each safety endpoint, including:  The hypothesis or estimation assessing the endpoint  Statistical method to be used  Any departures from the general statistical considerations as stated in Section 8  Any planned subgroup analyses Content of the SAP

24. CONFIDENTIAL © 2012 | 24  Section 12: Changes from the protocol  List of differences between the planned analyses in the protocol and those in the SAP, if any  Helpful later when writing Section 9.8.2 of the Clinical Study Report  Section 13: References  List of references Content of the SAP

25. CONFIDENTIAL © 2012 | 25  Mock TLFs are a supplemental living document that pre-specifies the following:  What data are to be summarized in tables  What data are to be listed in listings  What data are to be displayed in figures  Formatting of rows, row labels, columns, column labels, headers, footers, font, and font size Content of the SAP

26. CONFIDENTIAL © 2012 | 26  The SAP is an extension of the statistics section of the protocol  The SAP pre-specifies which analyses are to be conducted and provides details regarding how they will be conducted  The SAP must be finalized before breaking the blind Summary