Data Integrity: Identifying and Resolving the Issues. Gary Bird, Ph.D. Vice President, Regulatory Affairs and Quality

Transcription

1 Data Integrity: Identifying and Resolving the Issues Gary Bird, Ph.D. Vice President, Regulatory Affairs and Quality

2 Why is Data Integrity Important? Provides confidence that data supports Safety Identity Strength Quality Purity of the drug product

3 Real World Issues Causing Data Integrity Concerns Data integrity issues are not created equal. Organizational chart dependent, i.e., the culture at the top does not support absolute integrity Willingness of company management to ignore causes which lead to effects Poor training of personnel; e.g., don t care attitude directly related to culture of the company

4 What is data integrity? Assurance that data records are accurate, complete, intact and maintained within their original context, including their relationship to other data records. Applies to data recorded in electronic and paper formats or a hybrid of both. Data is protected from any form of modification (accidental or intentional falsification) or even deletion Data must be ALCOA.

5 Data are reliable Records/data have integrity if they are Accurate Legible/Long-Lasting Contemporaneous Original Attributable 2015 Cerulean Associates LLC 5

6 Legal framework FD&C Act 505(e): The Secretary shall, after due notice and opportunity for hearing to the applicant, withdraw approval of an application with respect to any drug under this section, if the Secretary finds (5) that the application contains any untrue statement of a material fact if the Secretary finds (1) that the applicant has failed to establish a system for maintaining required records, or has repeatedly or deliberately failed to maintain such records or to make required reports, in accordance with a regulation or order under subsection (k) or to comply with the notice requirements of section 510(k)(2), or the applicant has refused to permit access to, or copying or verification of, such records as required by paragraph (2) of such subsection; or (3) that on the basis of new information before him, evaluated together with the evidence before him when the application was approved, the labeling of such drug, based on a fair evaluation of all material facts, is false or misleading in any particular and was not corrected within a reasonable time after receipt of written notice from the Secretary specifying the matter complained of. Any order under this subsection shall state the findings upon which it is based.

7 Key regulatory references 21 CFR 11 ER/ES 21 CFR 58 Good Laboratory Practices 21 CFR 201 Labelling 21 CFR 202 Prescription Drug Advertising 21 CFR 203 Prescription Drug Marketing 21 CFR 210 cgmp in Manufacturing, Processing, Packing, or Holding of Drugs; General 21 CFR 211 cgmp for Finished Pharmaceuticals 21 CFR 600 Biological Products: General 21 CFR 601 Licensing 21 CFR 610 General Biological Products Standards 21 CFR 803 Medical Device Reporting (combo devices only) 21 CFR 806 Medical Devices: Reports of Corrections and Removal(combo devices only) 21 CFR 820 Quality System Regulations (Medical Devices) Application Integrity Policy (AIP) September 1991

8 FDA s Guidance on Data Integrity Published April 2016 Written in Q&A format Provides practical guidance for evaluating Data Integrity issues Answers common questions that trouble the industry Not the focus of this presentation

9 All Aspects of Data Collection and Reporting are Covered Clinical Studies (GCP) Manufacturing Activities (GMP) Non-Clinical Studies (GLP) Computer-Based Data Collection

10 Application Integrity Policy (1991) Four FDA employees and 11 companies engaged fraud in reviewing companies' ANDAs prior to The discovery of this extensive pattern of fraudulent data submissions prompted FDA to adopt the AIP and PAI Ensure validity of data submissions called into question by the agency's discovery of wrongful acts such as fraud, untrue statements of material fact, bribery, and illegal gratuities Sets forth the agency's general approach to applications that have been called into question by such wrongful acts and applications found to contain fraudulent data. Led to increased emphasis on data integrity Modified FDA understanding of what data integrity actually is Allow FDA to withdraw approval of, or refuse to approve, applications containing fraudulent data. Has been fundamental in the Agency s Regulatory Toolbox

11 Features of the AIP An administrative action that can affect all of the applications the company owns, both approved and unapproved Once AIP is invoked, FDA suspends review of the application or applications until the provisions of the AIP are met by the applicant holder Intended to assure the accuracy and reliability of data & information in applications submitted to FDA for scientific review and approval No statute of limitations

12 Validity Assessment According to the AIP FDA will investigate Scope Focus For pending applications, Defers scientific review of data to identify all instances of wrongful acts and to determine the extent to which the wrongful acts may be affected approved or pending applications. determined by the nature of the offense reliability of the applicant's research and manufacturing data. Will conduct validity assessments of those applications. Until the assessment is complete and questions regarding reliability of the data are resolved.

13 Validity Assessment, continued To approve an application Applicant is capable of producing a safe and, for some types of applications, an effective or functional product based on, among other things, testing and other data provided by the applicant and the adequacy of the applicant's manufacturing processes and controls. Basis for determination The reliability of data is of critical importance. If the agency determines that the criteria for approval cannot be met because of unresolved questions regarding reliability of data, the agency will not approve the application. With fraudulent data Refuse to approve the application withdraw approval, regardless of whether the applicant attempts to replace the unreliable data with a new submission in the form of an amendment or supplement. Actions Retesting may be required or other actions, including seizure, injunctions, civil penalties, and criminal prosecution

14 Corrective Actions, According to the AIP Determine the cause and scope of acts and assess the effects of the acts on the safety, effectiveness, or quality of products Cooperate with FDA and other Federal investigations Identify all individuals Involved in the wrongful acts remove on matters under the jurisdiction of FDA Senior management commits to develop and implement a corrective action operating plan to assure the safety, effectiveness, and quality of their products. Commit in writing Conduct a credible internal review by an outside consultant designed to identify all instances of wrongful acts associated with applications submitted to FDA

15 Data Integrity and Compliance with CGMP, On April 14, 2016, FDA published which addresses the role of data integrity in CGMP for drug manufacturing, finished pharmaceuticals, and positron emission tomography drugs, as required by 21 Code of Federal Regulations (CFR) parts 210, 211, and The guidance was issued in response to an increasing amount of data integrity violations found by the agency during CGMP inspections. According to FDA, data integrity CGMP violations have led to FDA warning letters, import alerts, and consent decrees. The new guidance answers data integrity questions in the hope of clarifying what FDA expects from manufacturers.

16 The Company: Nuts and Bolts of Data Integrity Must be culturally demanded Quality management system must not be up for debate Procedural and personnel controls must be in evidence Training is critical

17 Confirming Data integrity Frequently observed integrity issues No second person reviews Transcriptional errors Data falsification Data loss Data destruction or deletion Misrepresentation Inaccurate results (e.g., unvalidated methods) Discrepant - unexplainable results or recorded information Data degradation occasionally observed in computer based systems

18 Applies to all GXP Activities 21 CFR 312, IND 21 CFR 314, NDAs 21 CFR 600ff, Biologics Good Manufacturing Practices Good Laboratory Practices Good Clinical Practices

19 How Do You Find Integrity Issues? Auditor/Investigator Specific Audits Interact with personnel Note conditions in facilities, e.g., the Avalanche Zone Review raw data on problematic products Comparison of reported data versus other collected data Confirm the number of samples started with and compare to the actual number in storage (stability). Is the use rate acceptable? Look for second person verification Check for dates of performance versus entry Compare total number of analyses versus reported Confirm the specific data observed reflect the data reported General Approaches

20 How Do You Find Integrity Issues? Examine Periodic Product Reviews for data not identified Look at computer system audit trails Confirm all analysts have unique logons for computer-based systems Compare working hours of analysts to actual hours Compare the color of inks on a single page to confirm all entries were contemporaneous Review the names of individuals making entries in laboratory notebooks Confirm the number of analytical runs reported matches the number of runs documented in the LIMS computer-based systems Look for intentional modifications of calculations Do the data look too good? Good old fashioned, hard detective work General Approaches

21 Criminal Prosecution and Prison Sentence Clinical Investigator, MD - Ketek Study Applies to all GXP data, not just GMP OCI`s investigation determined that falsified CRFs were submitted to the sponsor along with falsified documentation to support the existence of a fictitious subject. Between November of 2001 and March of 2002, PI completed CRFs for 407 subjects enrolled in the study submitted by Aventis, the Sponsor, to reflect enrollment of the subjects in the study. OCI gathered evidence that over 200 subjects purportedly enrolled in this study had not, in fact, participated in it, and that one subject did not exist. FDA Notice of Opportunity for a Hearing, 21 December 2007

22 Bankruptcy GLP Cetero, , Ref 11-HFD Brought to FDA s attention by internal complainant The apparent manipulation of equilibration samples to meet predetermined acceptance criteria. Named in over 120 studies in ClinicalTrials.gov and affected over 1900 cases (individual activities in a study) Widespread falsification of dates and times in laboratory records for subject sample extractions Attempted to blame employees for company failures

23 Inspectional Findings (sadly, not uncommon) In the Laboratory No raw data for: Found at five (5) unrelated, company sites Standard preparation Sample weights Sample solution preparation and sample dilutions Without this information, assays cannot be calculated. Sample weights were made up and backdated, Stability data for 3 and 6 month stations tested on same day

24 Inspectional Finding, continued In the Laboratory Analyses were changed HPLC integration parameters were changed and re-run until passing results were obtained Stability samples were for a five year study were exhausted at the end of two years due to failing results Data transferred to a new computer and OOS data removed from hard drive and archive Quality control data: Test results for one batch were used to release at least 3 other batches at 3 unrelated firms

25 Inspectional Findings (sadly, not uncommon) In the Laboratory Untrained Personnel We didn t think the method had to be validated for this primary stability study. Multiple testing of samples to adjust elution times instead of using system suitability samples. Trial runs conducted to establish conditions and responses to allow adjustment of injection amounts to appear to be compliant.

26 GMP Inspectional Findings (sadly, not uncommon) In the Laboratory Data was not correct Data transferred to a new computer and OOS data removed from hard drive and archive Chemist recorded false data in the logbook under direction of a senior colleague Quality control data: Test results for one batch were used to release at least 3 other batches at 3 unrelated firms Data were transcribed on hands, scrap paper, sticky notes instead of being recorded in proper documentation form

27 Inspectional Findings (sadly, not uncommon) Computer Findings Computer not correctly utilized Multiple QC personnel accessing the same computer or analytical system using the same password e.g., a handful of passwords shared by 40 analysts. Personal computers were used to collect, analyze, and store data Audit trail function was disabled

28 Inspectional Findings (sadly, not uncommon) Spreadsheet Use Excel spreadsheets in analytical calculations Are neither controlled nor protected from modifications or deletion. Were not qualified. Must be retrospectively evaluated for the analytical values reported where such Excel spreadsheets have been used.

29 Inspectional Findings (sadly, not uncommon) In the Laboratory Analyses were changed HPLC integration parameters were changed and re-run until passing results were obtained Stability samples were for a five year study were exhausted at the end of two years due to failing results

30 Inspectional Findings (sadly, not uncommon) In the Laboratory Personnel Employee fabricating results to make it appear he was working overtime Chemist recorded false data in the logbook under direction of a senior colleague

31 WL: Torn and Discarded Records Ignored request An FDA investigator identified the presence of torn raw data records in the waste area and asked one of your firm s QA Officers to remove these torn raw data records for the investigator s review. This QA Officer presented the FDA investigator with approximately 20 paper records, none of which included raw data entries identified in the waste area earlier during the inspection. The FDA investigator asked three times if there were any more records found in the waste area, and the QA Officer responded to each question, "no, this is all of the records. The FDA investigator then re-visited the waste area and found that the raw data records had been removed and placed in a different holding bag. Unlabeled and partially labeled vials in the laboratory glassware washing area. When the investigator asked a QC Analyst to describe the contents of these vials, the QC Analyst immediately began dumping the contents of the vials into the drainage sink

32 WL: , continued Incomplete Records Your firm s laboratory records failed to include complete records of all stability testing performed. Unsecured Data Files Unacceptable Trial Runs and Consequences Quality control HPLC raw data files can be deleted from the hard drive using the common PC login used by all analysts. Deletion eliminates all records of sample injections and analyses. Your response indicates that this deletion function is only available on the software used for one of the sets of HPLC instruments. You also indicated that you have changed the access control privileges such that laboratory analysts in a user role cannot delete or rename files. You failed to provide the root cause for the unacceptable practice of performing undocumented trial runs at your facility, Failed to expand the scope of your investigations to include other instruments that use computerized electronic records both inside and outside the stability laboratory, and Failed to provide risk assessments on all the drugs where samples had been tested by these instruments. Your response failed to completely address how your firm will ensure the integrity and completeness of all analytical raw data.

33 Senior Management must be involved Show recent, example enforcement actions that have humiliated other firms Explain how data integrity simply asks Can the agency trust our data? Discuss how a data lifecycle-focus can help limit the scope and cost (and avoid mistakes of the past) Response must be governed by company Standard Operating Procedures Not exclusively used for data integrity issues Process should describe specific levels of escalation Formal documentation Description of issues Potential impact (Minor, Major, Critical) Review the costs of poor data integrity 3rd most common reason for a delayed or rejected submission untrustworthy product release records lead to public recall 5th most common warning letter with attendant loss of revenue 33

34 Summary Data integrity applies to more than just data Senior Management is personally responsible for compliance and data integrity. Every significant GXP area is subject to fraudulent behaviors Processes should address the potential of data integrity issues in conjunction with other corporate procedures such as senior management notification. The AIP is still a powerful tool in the hands of the FDA