On July 5, 2017, FDA (United States Food and Drug Administration) published a new document to the Federal Register titled “Pharmaceutical Quality/Chemistry Manufacturing and Controls (PQ/CMC) Data Elements and Technologies.” This is a draft document describing data structures that could take the place of much of the eCTD (Electronic Common Technical Document). Public comment is open for this document through September 11, 2017.
My first thought was that this might be FDA stepping closer to IDMP (Identification of Medicinal Products – the ISO standards that describe a medicinal drug product’s characteristics, composition and use). The current FDA Drug Listings and Establishment Registrations, while using the same Structured Product Language (SPL) are just a small fraction of what the full IDMP standards cover, so it seemed possible that this could be an expansion. On a similar note, the European Medicines Agency (EMA) has expressed interest in having IDMP act as the whole submission for many small registration variations – many of which are for CMC.
So I was a little surprised at the actual content of the data elements, as it was all the things I usually don’t think about: tests, specifications, analyses, and method validations.
I have to admit that of all the content that goes into a drug approval, the CMC section is the one I know the least about: over the years I’ve written specifications, coded and validated the systems, and edited word processing templates for pre-clinical, clinical, and regulatory portions of the filings… but little for Quality/CMC. While there were many templates for CMC in the systems I have created for authors (at companies before ACUTA), most of those templates only had a title and a sample table and figure, maybe a paragraph or two of boilerplate (stability study reports being a notable exception).
Interestingly, this isn’t the first time FDA has attempted to structure the CMC section. Back in the mid-1990s Norman Schmuff had proposed a structured version of the CMC information in a format called SGML, the ancestor to today’s XML. Similar projects were proposed by Health Canada, and the Swedish Medical Products Agency. Those initiatives, while not making a lot of headway (the tools just weren’t there yet for SGML or XML authoring), likely influenced the use of SGML, then XML for safety reporting.
The document describes the contents of fifteen tables, and the controlled vocabularies that support the data elements in those tables. The list below is grouped by function, and includes the table number in the document for cross reference:
- Description and Composition of Product (Table 9)
- Product Impurities (Table 14)
- Batch Formula (Table 10)
- Nomenclature & Structure of Drug Substance (Table 7)
- Drug Substance Impurities (Table 13)
- Drug Product – Control of Excipients (Table 12)
- Drug Substance – Control of Materials (Table 11)
- Specification (Table 1)
- Test (Table 2)
- Acceptance Criteria (Table 3)
- Drug Substance Characterization (Table 8)
- Batch/Lot info (Table 4)
- Batch Analysis of Drug Substance or Drug Product (Table 5)
- Stability Study (Table 6)
- Analytical Methods Validation (Table 15)
Only a few of these tables, and a few spots in other tables, overlap with IDMP: The Description and Composition of Product shares a lot with IDMP’s Medicinal Product Name (from ISO 11615) and Pharmaceutical Product (ISO 11616) structures. There are a few references to containers and closure systems, and I’d like to align those with IDMP and just reference the Package Identifiers, but because investigational products are still so far down the line for IDMP registration, this may not be practical. Some of the container info is on the bulk material batch lots – which isn’t covered by IDMP.
The Nomenclature & Structure of Drug Substance, the Drug Substance Control of Excipients (and of Materials), and the Product and Substance Impurities tables repeat much of the information in ISO 11238, which covers substances. I will be sending comments to FDA asking that these be made more explicitly an extension of those standards. Realistically, it shouldn’t be necessary to list the name, CAS Number, UNII, USAN, chemical structure etc., if the substances (including impurities) are all registered in the US (and eventually the global) Substance Registration System.
The bulk of the information in the document is related to testing and analysis: A specification is a set of tests; acceptance criteria are the goal values for those tests, the Analysis and Stability Study tables have the results of whether those tests met the criteria, and the Analytical Methods Validation is the justification and the documentation for the acceptance criteria. This makes this new FDA data submission very useful for the Quality review process. Compared to IDMP, though, it’s not as useful for the pharmacovigilance tasks of indexing and safety signal detection.
There are a few places where I’d do things differently (and will send to FDA as comments): when ingredients (active, excipient, diluent and impurities) are referenced, it’s with several identifiers – this should be simplified to just the US Substance Registration System number. There is one table that has information about the product as a whole that is repeated on multiple lines, as each ingredient is detailed – this should be normalized into separate tables, and align the structures closer to the IDMP Medicinal Product, Pharmaceutical Product and Manufactured Item. In another table, the stability parameters are expressed as a single coded field, when it represents about five different concepts (time, temperature range, package state, etc.) in the IDMP dataset. These changes would let this data structure be a ‘buddy’ to IDMP rather than an alternative – it should use IDMP’s data model wherever possible, and supplement beyond that.
The big question will be how this information will be managed. ACUTA’s ARIM system includes formulation management as of version 3.0. We will ensure that the product composition information in ARIM can be exported for use in the final, approved format PQ/CMC. However, the test and analysis tables are data that would come out of a Quality Management System, which is currently beyond ARIM’s scope (let us know if it’s a priority). Even in a QMS, though, the analysis details are likely part of reports, and not in the tabular format needed for this standard. That’s an area that ACUTA is investigating, as part of IRISS’s “DOCX” project to build structure into Microsoft Office Word documents. It’s not a product at this time, but we are at the forefront of the technology development.
Please contact ACUTA if you would like assistance in analyzing the impact of this new proposed standard on your company’s systems and processes, or would like to contribute comments to the FDA.