PMDA's New Validation Rules Explained Q&A
- How to obtain P21 ID?
When downloading and installing Community 3.0 for the first time, you will be prompted to create your Pinnacle ID. If for some reason you skip this step, you can go to Help --> Preferences and set-up you Pinnacle ID anytime.
- With respect to ADaM Other, does the dataset name necessarily need to start with 'AD'?
According to ADaM standard, file names should start with prefix "AD"
- Can we submit the same CDISC package to PMDA that was submitted to FDA (assuming there are no 'Rejects' in the validation)?
It’s a complicated question. Formally, you cannot have one data package that would be good for both FDA and PMDA. Two agencies have different requirements. For example, FDA prefers conventional units, while PMDA requests SI units. PMDA also requests additional Japanese datasets. And the scope of analysis datasets is different for PMDA and FDA. For more details, please look at Useful Links page on our webpage.
- Are there any filename differences between FDA and PMDA dataset packages (i.e. aCRF.pdf versus blankcrf.pdf)?
Expected file names for annotated CRFs are the same for both PMDA and FDA. It should be “acrf.pdf” in lower case. There are also some differences. For example, SDTM reviewer's guide should be “csdrg.pdf” for FDA and “study-data-reviewers-guide.pdf” for PMDA. Please refer to PMDA technical conformance guide for details.
- Warning type issues will not necessarily require any explanation, but do we still need to include them in the reviewer's guide?
You do not need to report validation Warning issues to PMDA.
- You stated that Error issues "if violated without any prior explanation" would suspend review. Does "prior explanation" mean "explained in SDRG/ADRG accompanying the data transfer" or "explained in correspondence with PMDA prior to submission of data"?
You need to explain Error issues in Reviewer’s Guide and submit it with your data. Remember, that you submit your study data before the final submission of your NDA and discuss them later with PMDA during data consultation meeting that happens before your final NDA submission.
- Any news on when SDTM-IG 3.3 will be supported?
Pinnacle 21 implemented a draft version of validation for SDTM-IG 3.3 in Enterprise back in April. However, CDISC team is still working on official rules for SDTM-IG 3.3m which are currently in public review process. Also, SDTM-IG 3.3 is not yet supported by FDA and PMDA. We expect that CDISC will publish rules early next year. While FDA may adopt this standard quickly, I would expect additional time for PMDA adaption due to their rigorous review and testing process.
- Can I use the CDSIC CT files instead of CT files provided in Pinnacle website for community version? Will there be any difference?
No, you cannot use CDISC CT files. There is some difference in format and addition information needed for validation.
- Is Chinese agency coming up with their own rules as they plan to move towards CDISC data submission?
The China National Medical Products Administration (NMPA) has stated commitment to CDISC standards, but has not yet provided details. We will keep the industry posted of any changes.
- If study start date is before April 2020 and data will submit after April 2020, then which version is applicable?
Study start date is not applicable for PMDA submission. Applicable version is based on the date of submission instead. After April 1st, you can use PMDA validation rules version 2.0. While version 1.0 will still be optional until March 31st, 2021.
- What domains are effected most by PMDA validation rules 2.0?
Most new rules were added to TS domain in SDTM. However, in terms of potential impact, additional rules for validating value-level metadata vs. actual data and support for ADaM OTHER (which applies to OCCDS datasets in ADaM-IG 1.0) are most important.
- If submission includes PGx data, what is your recommendation?
Officially, there is no CDISC standard for Pharmacogenomics domains. CDISC PGx implementation guide was published in 2015 as "Provisional". This means that this document includes many new variables and special domains which are not supported by CDISC SDTM Model. The intention was to include these new variables and domains in the next release of SDTM Model. After that CDISC PGx IG would be promoted to official CDISC standard. Until then a "Provisional" version is still a draft version with many expected changes. The CDISC team was planning to add new PGx variables into SDTM Model but then pulled them back because it was decided that they were not well-defined. To be CDISC compliant, if you want to use new PGx variables you should add them as SUPPQUAL variables (note today there is no CDISC-compliant solution for use of RELSPEC dataset). However, we do not think that this is a good implementation in terms of data usability. Our advice is to contact FDA eData team and ask them and the reviewers assigned to your application.
- Did PMDA remove conformance guide from their website?
PMDA re-designed their website and some information was moved to other pages. You can look to Useful Links page on Pinnacle 21 website for updated links.
- Can you speak to Enterprise version requirements from PMDA? For example they mention version 4.0.2, but current version is 4.1.2
Since Enterprise 4.0, the validation process is controlled by separately published validation engines. This allows us to support different versions of rules within the same application. And it also allows users to take advantage of newer versions of the application while still running the same PMDA and FDA validation engines. So Enterprise 4.0.2 and Enterprise 4.1.2 will produce the same validation results as long as you use the same PMDA 1810.3 validation engine.
- There are several choices of validation engines within Enterprise version. Is there a choice that will do both latest FDA and latest PMDA?
You can find 4 different validation engines in Enterprise: FDA, PMDA, PMDA Legacy, and P21. The P21 engine represents the latest versions of all rules from FDA and PMDA, as well as additional rules to validate sponsor-specific standards. When using this P21 engine, please be aware that latest version of some rules may be different from PMDA version.
- For Enterprise version, do we need to include AE, DM and EX in the validation package?
Yes, please include AE, DM, and EX datasets every time you validate your ADaM data package.
- FDA uses JumpStart, DataFit and Janus warehouse to demand compliant datasets, why does PMDA require the same?
PMDA has a similar process. They build their own data warehouse for submission data and utilize standards-based reporting tools such as JMP Clinical.
- We downloaded v3.0.2 recently, but did not see engine 1810.3. What could be the reason?
Community 3.0.2 was released before engine 1810.3 was made available. If you have Pinnacle ID and stay online, the 1810.3 engine would have been downloaded automatically. If not, you'll have to manually download and reinstall Community to get the latest PMDA engine.
- You mentioned that Reject rules are limited to certain domains/variables. Does the P21E implementation of them reflects this already?
Yes, P21 Enterprise implements Reject rules according to PMDA published requirements.
- Where can I download Pinnacle 21 Community validator with new PMDA rules?
Actually, you do not need to do anything extra if you already use Community 3.0 or above and stay online with your Pinnacle ID during validation. When new validation engines are published they automatically became available and can be found in the engine drop-down. Engine files will be automatically downloaded to your computer during your next validation with the new engine. If you do not use Pinnacle ID, work off-line or still utilize old Community version 2.1.3, you will need to download the new version of Community from Pinnacle 21 download page. This installation package includes engines for both versions of PMDA validation rules
- Where I can download Pinnacle 21 Enterprise?
You cannot download Pinnacle 21 Enterprise version. It’s a hosted SaaS solution. If you are interested in more details, please visit Pinnacle 21 website to request a demo.
- How I can download validation configuration files for new PMDA rules?
As explained in answer to question #20, the latest validation engines will be downloaded automatically using your Pinnacle ID. Once downloaded, you'll be able to find the configuration files in My Documents --> Pinnacle 21 Community --> configs folder.
- Does PMDA accept ADaM-IG 1.1? We have recently submitted application to FDA and started preparing submission of the same drug to PMDA. Can we use existing analysis datasets in ADaM-IG 1.1 format?
No, ADaM-IG 1.1 is not included in PMDA data standards catalog as supported standard at this time. There are also no PMDA validation rules for ADaM-IG 1.1. This means that formally you cannot use ADaM-IG 1.1 for PMDA submissions. However, you can try validating your data package with PMDA's ADaM-IG 1.0 rules. If no Reject issues are found you can then explain the Error issues and submit your data package as ADaM 1.0. If you are thinking of using this approach, please ensure to communicate this clearly to PMDA during consultation meetings and get PMDA approval. Also, do not forget to modify your define.xml file with a reference to ADaM-IG 1.0 instead of 1.1.
- How often will PMDA update validation rules?
We don't have an answer to this question. The previous version of validation rules was release 3 years ago. However, we expect more frequent updates in the futures especially to add support for new versions of the standards, such as ADaM-IG 1.1 and SDTM-IG 3.3.
- When will PMDA start accepting new versions of standards like SDTM-IG 3.3, ADaM-IG 1.1 and 1.2, and Define-XML 2.1?
We know that PMDA are currently working on reviewing validation rules for ADaM-IG 1.1. So, we expect these rules to be published soon. However, we are not aware of PMDA plans about SDTM-IG 3.3, ADaM-IG 1.2 or Define-XML 2.1 at this time. Keep in mind that SDTM-IG 3.3 and ADaM-IG 1.2 rules are still not published by CDISC teams. Therefore, we do not expect any updates from PMDA in the near future.
- We plan our submission to PMDA in March 2020. What version of rules should we use?
Formally, in March 2020 the only acceptable version of PMDA validation rules is version 1.0. So, if you are confident about not missing your planned submission date, you must use version 1.0. Starting April 1st, you can use either version of validation rules for your submission.
- Why is PMDA’s validation report may not equal or be consistent to the one created by sponsor/CRO?
There could be many reasons. Here is simple example. You validated your data with MedDRA 17.0, while another team member created the define.xml file with incorrect MedDRA version 18.0. When validating at PMDA, the incorrect version of MedDRA from define.xml file will be utilized, producing different results. Other examples are issues with Community configuration like the deployment on unsupported operating systems, use of different system encoding, or use of custom terminology files. You can mitigate many of these problems by using Enterprise which is managed by Pinnacle 21 team who will ensure that your validation system is properly configured.
