SQA, Regulatory Review Committee, EPA Subcommittee Meeting
On May 9, 2000, at the invitation of the Committee's EPA Liaison's, Del Huntsinger and Diane D'Angelo, Ms. Francisca Liem of the EPA's Office of Enforcement and Compliance Assurance (OECA), Office of Compliance, Agricultural & Ecosystems Division, met with members of the EPA Subcommittee. Ms. Liem provided the Subcommittee with a brief overview of current events then answered questions that had been previously submitted by Subcommittee members. The group further discussed her answers. Afterwards, Ms. Liem fielded questions from the floor.
The following provides a summary of the meeting with Ms. Liem including the questions and answers given. Questions are bolded, the responses not. This summary is being provided to SQA members as a service by the RRC, EPA Subcommittee to help keep the membership informed and up-to-date on current GLPs activities and issues.
Please note that the responses are Ms. Liem's opinions and interpretations only. These opinions and interpretations are not to be considered official EPA policy. These responses may depend on individual circumstances and should be evaluated carefully to determine if they are applicable to a situation. In many cases, there is more than one way to accomplish compliance with GLP regulations.
The SQA membership greatly appreciates and thanks Ms. Liem for taking the time to meet with the RRC and provide candid and interactive dialogue on GLP matters.
Ms. Liem's Opening Remarks
The end of March 2000 marks the mid-year for the EPA's fiscal year. To date, 54 GLP inspections have been conducted with 169 data audits.
ELAB will not include GLP laboratories under the accreditation program. ELAB had submitted a final report to the EPA in 1998 (April 16); it recommended several options for enhancing the inspection program. A letter* responding to the options was signed on March 21, 2000 by Mike Stahl, Acting Director of the Office of Compliance, OECA of the EPA. Basically, the EPA agrees with the ELAB recommendations, but feels that improvements to the current system will be more efficient. The EPA did not specifically endorse any of the options provided by the ELAB. The latest Audit Report* (dated March 27, 2000, Report No. 00P00011) of the EPA's Office of Inspector General (OIG) indicated that they were pleased with what the EPA has been doing and is pretty much closed. (Report 00P00011 is a follow-up review to ensure that the issues requiring improvements in management processes from the OIG's report of March 1994 have been resolved. The OIG agrees that the actions taken and the actions planned and outlined in Report 00P00011 will adequately address their recommendations and that no further response is required.)
Regarding the EPA GLP consolidation document (FIFRA/TSCA GLP consolidated revisions); comments were due in by March 29, 2000. They are in the review process right now; it will take at least 6 months to review all comments. Most comments are somewhat similar.
OPP has rejected 7 studies this fiscal year for lack of GLP compliance. Reasons included no QAU; lack of raw data, missing raw data; and, recordings made in pencil. One of the rejected studies was also submitted in Europe. For any study rejected by the EPA, information is forwarded to European countries when requested. There were 21 studies rejected by OPP last year (FY99) due to GLP reasons. The rejected studies were a result of OECA inspections. The studies are being rejected based upon the compliance statement and inspection findings. In the future, Ms. Liem will post the reasons for study rejections on the GLP Forum list server in a generic format.
The EPA recently audited a study where the data was supplied by the sponsor to the contract laboratory in electronic format (on a CD). The audit was difficult because the laboratory did not have the appropriate hardware and software to read the data; computer "crashed" and, in some cases, two pages of data had been scanned to fit onto one page. Labs must have the resources on site and need to inform EPA in advance if the audit is to be conducted using electronic media.
Two additional comments in response to issues that have been discussed on the GLP Forum: first, statistical packages has to be validated before using them in a GLP environment, and, second, QA reports must be sent to the Study Director and his/her management regardless of their location or relationship to those audited.
Q&A Responses
1. Is there a "mechanism" to receive information on fines and enforcement actions levied by the EPA? (OECA Echo Newsletter? Or, OECA's "Enforcement@EPA.gov"?)
I am distributing information regarding GLP enforcement actions through the GLP Forum.
2. What can you tell us about the EPA's plans for publishing an electronic records and electronic signature rule. Also, will it be along the lines of the FDA's "Part 11?" When do you think the proposed rule will be published in the Federal Register?
EPA is undertaking a rule-making for Cross-Media Electronic Reporting and Recordkeeping Rule (CROMERRR) that will provide the legal framework for electronic reporting and record keeping under EPA's environmental regulations. The lead office for the rule making is the Office of Environmental Information, headed by Deputy Principle Assistant Administrator Margaret Schneider. Public meetings on the proposed rule will be held in Chicago (June 6th) and Washington, DC (July 11th). An individual from my (Ms. Liem) group is working on the rule.
As much as possible, EPA will be consistent with FDA's Part 11 rule.
EPA plans to publish the proposed rule in the fall of 2000.
3. What is the status of the EPA consolidated GLP rule? When do you think the final rule will be published?
EPA has closed the comment period and is currently reviewing all comments. It will take at least 6 months for the review. Once published in the Federal Register, the rule will become effective after 60 days of Congress in session.
4. What is the focus of the EPA GLP inspection program this year?
Antimicrobial efficacy laboratories Ð these labs are not in compliance. Also, organophosphate studies are being targeted due to impact to FQPA.
5. Can you comment on the case involving the New Jersey clinical lab which tested pesticides without informing the study subjects? Specifically, what are the sponsor's responsibilities in this situation? What are the lab's responsibilities? Has the EPA published any guidance in this area?
The sponsor is responsible for informing the testing facility of the nature of the pesticide. It is the lab's, and the sponsor's, responsibility is to tell the human subjects the nature and purpose of the test and of any physical and mental health consequences that are reasonably foreseeable therefrom. The sponsor has to inform EPA; FIFRA Section 12 (p).
EPA has no published guidance on the description of responsibilities of the laboratory and the sponsor.
6. Was the NJ lab the most recent enforcement initiative? If not can you update us on the latest initiative?
The NJ lab was not part of an enforcement initiative. There are no plans yet for a GLP enforcement initiative. EPA will review cases and decide on whether to take enforcement. Sentencing by the Department of Justice will be June or July.
7. Please provide an update on EPA's current position on human (pesticide) testing.
EPA has received 14 unsolicited human studies on 10 different pesticides. These studies raise difficult scientific and ethical questions that we are not yet able to answer, and EPA is deeply concerned about them. EPA has the authority to specify what tests are required and how they should be performed through test guidelines. EPA has never defined guidelines for testing pesticide effects in human subjects. We do not require such studies, we do not encourage them and we do not believe them to be necessary to do good risk assessments. EPA is developing a policy regarding human testing; the policy should be available in a couple of months.
EPA currently requires certain reentry/worker exposure studies under test guidelines and is developing guidelines for certain exposure/ repellent efficacy studies.
8. The GLPs require that study records must be archived at the close of the study, i.e., when the Study Director signs the final report. How would you suggest this be accomplished in cases where the archive is a different location than Study Director, e.g., another city?
In the situation where the archive is in a different city from the study director, the study director must ensure that the study records are transferred to the archive at the close of the study.
9. Is it the Study Director's responsibility to obtain the signature of the Sponsor on the compliance statement for the final report and must the test facility archive the Sponsor signature document?
It is the study director's responsibility to sign the final report. A GLP compliance statement is only required when the final report is submitted to EPA. In many cases, the study director is not aware when the final report will be submitted. So, it is the sponsor's responsibility to collect all signatures, i.e. study director, sponsor and submitter.
If the testing facility is a contract research organization, it does not need to archive the sponsor signature document.
10. Some facilities are entering data directly into a computer with no change control in place (e.g., EXCEL spreadsheets). Although this seems to be a violation of GLPs, it appears that EPA inspectors are not making an issue of it. Is EPA going to let this slide until an official electronic record rule is passed?
All original data entered directly into a computer must be in compliance with the GLP, including identification of the individual responsible for direct data input at the time of data input. Any change in entries must be made so as not to obscure the original entry, indicate the reason for change and the responsible individual who makes the change must be identified.
Ignoring the GLP requirement on data entry is a GLP violation, and EPA inspectors will write it up. This GLP requirement under EPA GLP Standards regulations has been in existence since 1984 for FIFRA studies and 1983 for TSCA studies.
11. Some companies use very complex systems such as GIS for mapping and gathering information used to set up studies. Obviously, GIS systems are not validated as per GLP. Do you have any guidance whether companies have to validate these systems and if so how would we go about it?
It is up to the companies to select an appropriate method for validating their systems. These systems have to be validated, whether the supplier or user performs the validation.
12. If a computerized monitoring system has the option of simultaneously printing out the monitoring data at the same time it is being electronically captured, which data set is considered the raw data Ð the printed log or the electronic reading?
The testing facility has the option to decide which is considered the original raw data. The decision has to be in a written SOP.
13. From a contract laboratory stand point, what documentation should be supplied by the sponsor to indicate appropriate GLP characterization of the test and reference substance(s)? If this information is not provided, should the compliance page of the final report address the issue (i.e., state that the test and or reference substance was not characterized or was not characterized under the realm of GLP)?
The sponsor should, at a minimum, provide the study director with a certificate of analysis of the characterization and whether it was performed under GLP. The study director and testing facility management are responsible to ensure that the test substance is characterized under the GLP.
The study director should persuade the sponsor to provide him/her with the certificate. The study director should show the sponsor the text of GLP regulations at 40 CFR 160.31(d), 40 CFR 160.33(e) and 40 CFR 160.105(a), where the following is stated:
- For each study, test facility management shall assure that test, control, and reference substances have been appropriately tested for identity, strength, purity, stability, and uniformity, as applicable [40CFR 160.31(d)].
- The study director has overall responsibility for the technical conduct of the study, as well as for the interpretation, analysis, documentation, and reporting of results, and represents the single point of study control [40 CFR 160.33].
- The study director shall assure that all applicable good laboratory practice regulations are followed [40 CFR 160.33(e)].
- The identity, strength, purity, and composition, or other characteristics which will appropriately define the test, control, or reference substance shall be determined for each batch and shall documented before its use in a study [40 CFR 160.105(a)].
If the study director is confident that the test substance was not characterized, or was not characterized under the GLP, he/she should state this fact on the GLP compliance statement. However, I do not recommend him/her to write a statement to the fact, that he/she is not aware of the extent of the characterization test, because the sponsor will write a contradicting statement, and this may trigger a data audit by EPA.
14. At company X, the QA group leaders, in some instances, also sign reports as management. Is this a GLP violation?
It is not a GLP requirement that management sign a final report. Only the study director needs to sign the final report. The QA cannot sign as testing facility management; this is a conflict of interest regarding QA independence.
15. The EPA GLPs state that a final report shall include "a description of all circumstances that may have affected the quality or integrity of the data." Should all protocol deviations be reported, even if the Study Director has determined that the deviations would have no adverse effects on the results of the study?
"A description of all circumstances that may have affected the quality or integrity of the data" means anything that was not expected. Any changes from the original protocol must be reported as required by the GLP at 40 CFR.160.185(a)(2).
16. In regards to statistical analysis, on occasion, one type of analysis is conducted as per protocol and then because of what is seen, it is determined that another analysis is needed. Does the agency still want to see the original analysis even if it is thoroughly explained in the report and driven by an amendment?
Yes, the Agency has to establish that the original analysis was indeed performed, which in turn triggered the need for another analysis that is supported by a protocol amendment and explained in the final report.
When asked about data that has been generated but rejected on sound scientific reasons, e.g., autoinjector error, Ms. Liem responsed that data should be reported also. Discussion ensued that EPA data reviewers had commented that they did not want this data reported in the final report. Ms. Liem noted that she would follow-up with the data reviewers on this issue.
17. In regards to statistical analysis, for some studies that are designed to answer a particular question, the correct analysis to use may not be known until looking at the data. Is it ok to run the data through Jump and other software packages to determine the best way to analyze and report the data and then not retain the information that was generated that doesn't support the analysis chosen to be reported?
If the laboratory uses "Jump" and/or other software packages, that data needs to be retained to support why one statistical analysis was used vs. another. This process needs to be explained in the study data and protocol amended once it is determined what will be used. EPA does not consider SAS to be a validated program that doesn't require end-user validation. End-user validation is required for all statistical programs.
18. Some of the product chemistry studies require specialized equipment that varies depending on the nature of the test substance. Specifically, flash point, flammability, surface tension and other physical property test requirements are different for liquids and solids, etc. It is difficult to find a GLP lab capable of conducting all of these experiments for every type of test substance. Would it be acceptable for a "GLP" study to be conducted under a protocol where the study director oversees the generation of data at a testing facility that may be certified to conduct an experiment under ASTM or ISO guidelines, but is not a totally compliant GLP facility? If the exception is described on the compliance page would the study be acceptable as a GLP study if the science was sound?
The GLP standards require that all studies submitted to EPA in support of a pesticide registration, or under a testing consent agreement, be conducted according to the GLP. EPA may reject studies that were not conducted according to the GLP, although the GLP compliance statement states truthfully this fact. So, sponsors are taking a chance by submitting non-GLP studies. The non-GLP study can't be accepted as a GLP study, although the exceptions to the GLP have been described on the compliance statement, and even if the study is scientifically sound.
19. For a metabolism study that has a radioactive test substance, what is the Agency's expectation in regards to documentation for the stability of the test substance? For example, if we have stability information on the cold test substance, is this sufficient information to cover the Agency's expectations for stability of the radiolabeled test substance as well? In these cases, the sponsor usually provides a contract laboratory with a radiopurity value for the radiolabeled test substance and then the contract laboratory performs another radiopurity check prior to the study being performed to assure the value compares with that received from the sponsor. If the protocol spells out all of the above information would this fulfill the Agency's requirements for stability of the test substance as defined in 40 CFR 160.105(b)?
Stability information on the cold test substance is not sufficient to cover the stability of the radiolabeled test substance.
A reevaluation of the radiolabeled test substances should be performed prior to the start of the study. Characterization of the radiolabeled test substance should include both qualitative and quantitative analyses.
20. Occasionally a long-term study will require a report to be written at the halfway point. This report will have the form of a final report, but by its nature cannot have final conclusions or a complete report of the data. Is an "interim" report just a "business communication of a study's progress" or does it come under the GLP requirements of archival, report amendments, and appropriate signature?
If the interim report is submitted to the Agency, all GLP elements apply.
Additional Comments From Ms. Liem
1. Status of PR 86-5 revision:
Revision not available; still working on it.
2. Can preprinted dates be used to fulfill GLP requirements of signing and dating?
No, the individuals must initial and date all entries.
3. MOU with Canada:
First two of three phases are complete. The first phase was the establishment of GLPs in Canada.The second phase (March 2000) was a Canadian visit to the US EPA to observe and compare GLP programs. The third phase is for the EPA to visit Canada to evaluate their program; a date for this is currently being worked on. Goal is to have the MOU ratified by the end of 2000.
Additional Responses From a Previous Q&A Session With Ms. Liem will be published in Septembers LCO Newsletter.
3. MOU with Canada:
First two of three phases are complete. The first phase was the establishment of GLPs in Canada.The second phase (March 2000) was a Canadian visit to the US EPA to observe and compare GLP programs. The third phase is for the EPA to visit Canada to evaluate their program; a date for this is currently being worked on. Goal is to have the MOU ratified by the end of 2000.
Additional Responses From a Previous Q&A Session With Ms. Liem
25411. Does the EPA expect all aspects of studies conducted for Biotechnology registrations to be conducted according to GLPs?
Yes, studies in support of Biotechnology product registrations must be conducted under GLPs.
25412. Will the EPA accept studies for Biotech registrations if they are not done under GLPs?
Acceptance of non-GLP Biotech studies by the EPA reviewers are on a case-by-case basis.
25413. A hypothetical laboratory of around sixty people has five or six departments. Three of the six departments conduct GLP work. A director heads each department. Most of the Study Directors report to these department directors. All department directors report to the Vice President of R&D. For GLP purposes (160.31), which level of this lab are considered management for the purposes of signing protocols, final reports, QAU audit reports etc.?
The reporting line of QAU at the test facility is the decision of management. It's up to management to decide who management is.
Is it appropriate to consider the department Directors as study management or is it the VP?
Again, these decisions are to be made by the facility management.
25414. Can management of a small laboratory serve as both Study Director and management? If so, what relationship of the person that serves in both capacities to the QAU? In other words, can a person that is study director also be the management person that signs the final reports?
Technically speaking, no. Management of a small lab cannot serve as Study Director at the same time, but EPA inspectors will evaluate each specific facility. The GLPs do not require "management" to sign final reports.
25415. When a GLP violation is corrected such as data inadvertently not signed or was omitted, does the violation still need to be addressed in the compliance statement or does the correction eliminate the need to report the corrected violation?
GLP deficiencies that could be corrected as part of the study review process do not have to be included as an exception in the compliance statement.
25416. Can QAU contractors be hired to audit reports, conduct inspections and sign the QAU statement in the final report that they audited?
Yes.
25417. 160.105(a) requires the identity, strength, purity and composition of test, control, and reference substances to be documented prior to its use in the study. Does this work need to be conducted under GLPs?
Yes.
If it is conducted under GLP what form of documentation is acceptable? Notebook entries? Final Reports? Does the characterization need to be reported to the Agency at the same time, or prior to, the subsequent studies that use the information to support their test control or reference substance characterization?
If the data are required to be submitted to the EPA, they are to be developed under GLPs.
25418. What actions can you recommend that a QAU take if they encounter facility management that is, a) resistant to complying with the overall GLP requirements of 40 CFR 160, and, b) not fulfilling test facility management obligations in 160.31(d),(e),(f) and (g) with respect to (g) QAU is communicating the deviation and compliance issues, but corrective actions many times are not taken or not adequately documented?
Document your recommendations in the periodic QAU inspection reports submitted to management. The Quality Assurance Unit statement in the final report contains only dates of inspection activities. The Study Director signs the GLP Compliance Statement that affirms the GLP compliance status of the study.
25419. Does the Agency require each testing facility to record all studies, GLP and Non-GLP into the Master Study Schedule?
The Master Schedule Sheet is required to include only GLP studies.
Do contracted GLP studies have to be included on the Master Schedule?
If the Study Director is at the Sponsor's site, contracted studies have to be on the Master Schedule and it could include when the contracted studies are being conducted.
25420. Our SOP notes that we document Sponsor approval of protocols and amendments through sponsor/monitor signature on those documents. The originals are retained and archived with the study records. The QAU is required by the GLPs to retain copies of protocols (and amendments) for studies, which have QAU involvement. Is there an expectation that the QAU retain fully signed copies of protocols and amendments for studies with QAU involvement or is it adequate that we retain the protocols and amendments in our files with just the Study Director signature (which is the condition upon internal distribution)? We do verify that fully signed documents exist at the time of report finalization.
The GLPs do not require Sponsors to sign protocols/amendments. There must be documented evidence that the Sponsor has approved the protocol. If your SOP requires that the Sponsor signs the protocol and all changes, then it becomes part of your facility requirements. QAU must not be involved in any GLP study. The GLPs require that the QAU records include copies of signed (by the Study Director) and approved (by the Sponsor) study protocols and amendments.
25421. Is there a requirement or expectation that Sponsor comments on draft protocols, draft reports, draft protocol and report amendments are retained and if so who is responsible for retaining these documents?
All documents related to interpretation and evaluation of the study data must be retained, i.e., Sponsor's comments on draft final reports and final report amendments. Draft protocols and Sponsor's comments on protocols do not need to be retained. During the conduct of the study, QA and Study Director must keep a copy of the study protocol and amendments. At the conclusion of the study, the Sponsor is responsible for the raw data, protocol and amendments, draft final report and Sponsor's comments, and the final report. The testing facility must maintain a copy of the final report.
25422. A study is finalized with the report indicating that data will be retained in the contract research organization (CRO) archives for 1 year and then may be transferred to a site specified by the Sponsor. The Sponsor decides to transfer the data to their archives and the CRO archive records reflect this transfer including where the data were sent. What is the expectation of the CRO's responsibility if the Sponsor decides to transfer the data again or the company is sold?
The Sponsor is responsible for retention of the necessary materials in the archives. The CRO must retain a copy of the final report. CRO has no further responsibility of maintaining the raw data, even if the Sponsor transfers the raw data to another company.
25423. What level of installation qualification and operation qualification (IQ/OQ) is required of refrigerators and freezers used in support of GLP studies?
The GLPs do not specify any IQ/OQ requirements for refrigerators or freezers; this is a FDA Good Manufacturing Practice (GMP) issue.
25424. How are companies handling electronic SOPs, especially keeping them secure?
EPA does not direct the regulated community how to keep them. The test facility should have control of all documents at all stages of issue.
25425. If a laboratory does not receive information on expiration dates for reagents or reference standards would it be acceptable to assign default expiration date defined by SOP (e.g., 3 years from date opened)?
Yes, this is a possibility if the material would still be acceptable at the expiration date.
25426. Is there any way that a laptop using an Excel spreadsheet can be used to capture/record raw data in a GLP-compliant manner?
If the system does not have an audit trail, the laptop cannot be used.
25427. For a given study, is it permissible to have the Test Substance and the analytical standard (reference substance) come from the same original batch or lot of material, provided they are separated into different assigned containers prior to study initiation and they each are appropriately characterized and handled for the purpose they are intended?
Yes.
25428. The GLPs require that each testing facility maintain a current summary of training and experience and a job description for each individual engaged in or supervising the conduct of a study. Many businesses retain and provide detailed training records, which include specific dates and identity of the individual performing the training. This appears to be more than the regulations require. Wouldn't summaries of the training be sufficient?
Only summaries are required by GLPs. But, it might be good business practice to maintain documents that support the summaries.
25429. According to the May 1992 EPA Question & Answer document, the EPA considers a study to be composed of all of the necessary elements of research performed to obtain the reported results of the study. If different phases are conducted at different laboratories they are still part of that study. The example given was one lab treating the test system and a second lab analyzing the samples. The agency has also made it clear that there should be one Study Director for each study. How is each testing site required to report their part of the study when:
- The first lab refuses to provide the second lab with the protocol then how does the second lab function under compliance, e.g., is it appropriate for the second lab to generate a protocol for their part of this study?
No.
- The Sponsor refuses to provide the second lab with the name of the Study Director should the second lab assign their own Study Director?
No.
- Would it be appropriate for the Sponsor to be named and serve as Study Director?
Yes, this is a possibility.
- What Experimental Start and Termination Dates should the second lab use?
The dates for the activities of the second lab only, footnoted to indicate the dates were for their portion only.
- How should this appear on the second lab's Master Schedule, e.g., which initiation and termination date should be used and which study director should be listed?
Again, the date of initiation for the Master Schedule would be for the second lab only and should indicate that fact. It is a GLP violation, if a laboratory does not know the Study Director. Where will QA reports go?
- How should reporting requirements be met by the second lab if there are more than one start date, Study Director, protocol, procedure, testing site, and management? Should the second lab report only their part of the study using their dates?
Yes, the report from the second lab should only report their part(s) of the study. This should be clearly stated in the report that the dates/time frame for the work done reflects only the second lab.
- Available from SQA Headquarters.