Docket Management
Docket: 00D-1418 - Q7a ICH Good Manufacturing Practice Guide for Active Pharmaceutical Ingredients
Comment Number: EC -10

Accepted - Volume 1

Comment Record
Commentor Mr. Gregory Bobrowicz Date/Time 2000-09-28 18:41:26
Organization FDA Ready Consulting
Category Individual

Recommended Changes / Reasons
Lines: 80-80
Section: 1.3
Change the Scope table so this document applies to APIs produced from recombinant plants at the "master" seed bank.
Recommended Change Type: General Comments The scope of the document has an internal conflict. There are APIs produced as recombinant plants, and presumably the guide would apply at the maintenance of the “master” seed bank.
Lines: 564-564
Section: 6.53
Materials failing to meet established specifications should be rejected. Rejected materials may be re-processed or reworked.
Recommended Change Type: Add new definition, section, or item. Nothing in the document specifically requires rejection of failing materials.
Lines: 179-179
Section: 2.50
Remove the word "normally"
Recommended Change Type: Revise, change, add or delete words or portions of sentences. Unless there are circumstances where the reviews would not be conducted, the word "normally" is confusing.
Lines: 262-262
Section: 4.16
Delete the word "normally."
Recommended Change Type: Revise, change, add or delete words or portions of sentences. Unless there are circumstances where laboratory areas can be contiguous with production areas (which should be stated, if so), the word "normally" introduces confusion.
Lines: 706-706
Section: 7.42
Delete the word "normally."
Recommended Change Type: Revise, change, add or delete words or portions of sentences. If there are circumstances were FIFO can be violated, they should be stated. The word "normally" causes confusion. The document should be prescriptive, not descriptive.
Lines: 986-986
Section: 11.21
Delete the word "normally."
Recommended Change Type: Revise, change, add or delete words or portions of sentences. Are there circumstances where the impurity profile would not depend on the process and origin of the API? The word "normally" causes confusion. The document should be prescriptive, not descriptive.
Lines: 986-988
Section: 11.21
Impurity profiles are not necessary for APIs that are not purified entities, such as crude extracts from herbal or animal tissue origin.
Recommended Change Type: Revise, change, add or delete words or portions of sentences. Impurity profiles would be expected for highly purified APIs that happen to derive from natural sources, such as animal hormones or Taxol.
Lines: 1726-1726
Section: 20
Critical A factor becomes critical if variability in the factor beyond predetermined values introduces significant variability or failure in a quality attribute of the API. Determination of whether a factor is critical is a matter of documented scientific judgement.
Recommended Change Type: Add new definition, section, or item. The word “critical” (or “criticality,” or “critically”) appears at line 181, 183, 362, 401, 407, 410, 416, 432, 441, 485, 561, 575, 578, 623, 626, 638, 641, 644, 692, 730, 734, 739, 742, 771, 806, 879, 905, 1083, 1086, 1087, 1089, 1091, 1097, 1147, 1171, 1270, 1498, 1539, 1542, 1587, referring to activities, materials, data, systems, equipment, steps, parameters, deviations, in-process controls, physical attributes, information, records, and quality attributes. In almost every case, the word is intended to limit the scope of some activity; specifically, if only the critical thing must be tended, then implicitly the non-critical thing can be ignored. The definition is unsatisfying because everything must be controlled within predetermined criteria to ensure that the API meets its specifications. There is no control in the pharmaceutical industry that can be flouted without some consequence. Even something as meaningless as rinse water for the floors is critical under the existing definition, because if thousands of gallons were used, it would flood the plant and the API would become contaminated!
Lines: 310-310
Section: 4.41
Dedicated production areas should also be employed when material of an....
Recommended Change Type: General Comments The semantic difference between employing dedicated facilities for penicillin (4.40) and considering dedicated facilities for other agents (4.41) is very unclear.
Lines: 1375-1375
Section: 16.11
... prior to signing the contract and before accepting any work products from the contract manufacturer.
Recommended Change Type: Add new definition, section, or item. The current wording, without any timing, would allow a contract manufacturer to remain unreviewed indefinitely provided that the contract giver had an unimplemented program or plan for performing audits.
Lines: 1514-1514
Section: 18.15
18.15 For biotech products, removal of media components, host cell proteins, other process-related impurities, product related impurities and contaminants must be validated.
Recommended Change Type: Revise, change, add or delete words or portions of sentences. The standard, "may be necessary" is ambiguous, and FDA has provided no guidance on when such validations would NOT be necessary.
Lines: 1539-1539
Section: 18.33
Delete the last two sentences of 18.33.
Recommended Change Type: Delete or omit entire sentence, paragraph, or item. There is no scientific basis for the statement (line 1543) that “for ‘classical fermentation’ certain parameters (cell viability, for example) may not need to be addressed.” First, "classical fermentation" is undefined. Second, viability is a critical parameter (as defined) for most biotech processes; specifically, if viability is not controlled within a preset range, no product will be produced, and the batch will therefore fail potency. All of section 18.33 appears to be redundant with section 8.30.
Lines: 1576-1579
Section: 18.43
If open systems are used, purification should be done under controlled environmental conditions appropriate for the preservation of product quality. Biotech products should be processed in areas with HEPA filtered air.
Recommended Change Type: Revise, change, add or delete words or portions of sentences. This document should be prescriptive, not descriptive. Words like "normally" and "may need" introduce confusion. If FDA intends flexibility, the document should include general guidelines of when HEPA filtered air would NOT be required.
Lines: 1600-1612
Section: 19.1
Unless specifically exempted or modified in this section, all requirements in the remainder of this document apply.
Recommended Change Type: Add new definition, section, or item. The document should contain more clarifying language regarding APIs for clinical trials. Is it true that only the specific requirements listed in chapter 19 apply, and all of the other requirements in the document do not apply? Or is it the reverse, that all of the requirements in the document apply with the exceptions in chapter 19?
Lines: 1659-1661
Section: 19.70
19.70 Although changes are expected during clinical development, every change in the production, specifications, or test procedures should be adequately recorded and reviewed. This requirement applies to the first and all subsequent batches intended for clinical trials.
Recommended Change Type: Revise, change, add or delete words or portions of sentences. The proposed section 19.70 implies that documentation of changes is not required before “knowledge is gained and the production is scaled up?” The document does not specify which knowledge and what scale is the trigger for the requirement for change control.
Lines: 1662-1664
Section: 19.80
Methods used for pivotal clinical trials should be fully validated according to ICH guidelines. Methods for early clinical trials should be demonstrated to be suitable for use according to the ICH guidelines, but it is not necessary to document validation including a protocol with pre-determined acceptance criteria. Instead, a qualification report describing the results of studies is sufficient, provided that the report is reviewed by the quality unit. It is only necessary to qualify a method for its intended use. For example, if a method is only performed in a single laboratory, then the lab-to-lab portion of intermediate precision is unnecessary.
Recommended Change Type: Add new definition, section, or item. Section 19.80 is useless because it is so broadly worded. What is the practical difference between “scientifically sound” and “fully validated?” Are “scientifically sound” methods proved to be appropriate through execution of a protocol with pre-established acceptance criteria? Must a method be demonstrated to be accurate, precise, and specific to be “scientifically sound?”
Lines: 1723-1725
Section: 20
Contract Manufacturer A person or business entity (including a laboratory or service provider, but excluding consultants) that is outside the immediate control of the API's owner, sponsor, or applicant.
Recommended Change Type: Add new definition, section, or item. The definition and requirements of “contract manufacturer” are circular. Line 1724 defines a contract manufacturer as “a company holding an agreement,” and section 16.12 requires a “written and approved contract.” Thus, if a company performs work for another company, but the nature of the work is unwritten, there is no “contract manufacturer” under the definition, and 16.12 would not apply.
Lines: 1522-1522
Section: 18.23
The usage period of the cell strain should be defined.
Recommended Change Type: Revise, change, add or delete words or portions of sentences. "Classical fermentation" is undefined and should be deleted. The word "usually" is descriptive, not prescriptive, and so introduces ambiguity on the regulatory intent.
Lines: 1521-1521
Section: 18.23
Stability of cell banks should be demonstrated according to ICH Q5D.
Recommended Change Type: Revise, change, add or delete words or portions of sentences. ICH Q5D adequately addresses stability requirements for recombinant cells. The proposed wording is unclear because it does not describe the nature of the testing. Is a cell bank demonstrably stable if it is regularly used in production? Does the proposed language require a regular evaluation of genetic stability? It is better to incorporate existing guidance by reference than to reopen this discussion.
Lines: 421-424
Section: 5.42
5.42 Commercially available software that has been used for a variety of applications (not limited to the pharmaceutical industry) for a significant period of time does not require validation of its core functions. A manufacturer should validate the specific application of those functions, included template spreadsheets, macros, queries, and databases. Where purchased software has not been used in a variety of applications for a significant time, a manufacturer may either perform full validation of the software or substitute comprehensive vendor audits. Audits may only substitute for aspects of validation that require a detailed knowledge of the source code or architecture of the purchased software. If an existing system was not validated at time of installation, a retrospective validation may be conducted if appropriate documentation is available.
Recommended Change Type: Add new definition, section, or item. The proposed section 5.42 provides a very weak standard; there are endless examples of “commercially available software” that is completely inadequate for any use because it is so bug-ridden.



EC -10