Wednesday, February 9, 2011

Process Wednesday: current Good Manufacturing Practice

Something that I keep hearing about in the manufacture of pharmaceutical products is "cGMP", which stands for "current Good Manufacturing Practice". I'll be completely honest and say that I don't really quite understand what it really encompasses and what is different about a cGMP facility and a non-GMP facility (other than a lot less paperwork, I suppose.) It's worth considering the FDA's comments on GMP: "The regulations in this part [CJ: this is from the CFR] contain the minimum current good manufacturing practice for preparation of drug products for administration to humans or animals."

Wikipedia tells me that cGMP are "guidelines that outline the aspects of production and testing that can impact the quality of a product." It is my understanding that cGMP typically tells an drug manufacturer "what" must be done, as opposed to "how" it should be done. In the GMP Wikipedia article, there is also a list of the different components of a GMP program:
  • Manufacturing processes are clearly defined and controlled. All critical processes are validated to ensure consistency and compliance with specifications.
  • Manufacturing processes are controlled, and any changes to the process are evaluated. Changes that have an impact on the quality of the drug are validated as necessary.
  • Instructions and procedures are written in clear and unambiguous language.
  • Operators are trained to carry out and document procedures.
  • Records are made, manually or by instruments, during manufacture that demonstrate that all the steps required by the defined procedures and instructions were in fact taken and that the quantity and quality of the drug was as expected. Deviations are investigated and documented.
  • Records of manufacture (including distribution) that enable the complete history of a batch to be traced are retained in a comprehensible and accessible form.
  • The distribution of the drugs minimizes any risk to their quality.
  • A system is available for recalling any batch of drug from sale or supply.
  • Complaints about marketed drugs are examined, the causes of quality defects are investigated, and appropriate measures are taken with respect to the defective drugs and to prevent recurrence.
When I read this (mind-boggling) list of criteria, I begin to see the core of cGMP, which is that pharmaceutical manufacturing needs to be consistent and documented. When I think about how it might be applied to a single synthetic step, I think it means that chemistry needs to be well-documented, very, very robust and the product/impurity profile should not change over time.* Hard to imagine that it would be otherwise, but I suppose that there's a whole backstory of awful mistakes that led the industry and the government to this point.

Readers, what am I missing? If you were to explain cGMP to a 1st year graduate student, how would you do it?

*Dude, it's so clear I don't know what I'm talking about here.



  2. The pun is the lowest form of humor. :-)

  3. Should researchers learn GMP or GLP? And (since i don't know) what's the difference?

  4. If you're engaged in basic science (ie; the discovery side) then GLP is the more likely of the two to be useful.

    I've never been GMP trained, I imagine the difference is a matter of paperwork.

  5. This was one of the biggest challenges for me when I moved from research to regulatory. I'd worked in a lab for years, even doing a little bit of process chemistry, but I'd never worked in a GMP environment. It's definitely a different mindset. In the research lab, even if you've done a reaction dozens of times, you probably don't do it the same way every time. You adjust the conditions as you go along. In a GMP lab, you have to do it the same way every time. Or more accurately, you have to know what you can change and by how much. It's becoming more common to monitor reactions during the process and make adjustments but you can only make adjustments that have previously been tested and documented.

    One thing to remember about most manufacturing processes is there's very little purification. It's just not practical on a scale of hundreds or thousands of grams. You might have a washing step or possibly a recrystallization, but you're not going to run a column or a prep HPLC. So you can't fix the reaction after the fact - it has to work the first time. That doesn't necessarily mean there are no impurities, but you should understand what impurities are there and how they affect the product and its safety. That is completely different from working in research!

  6. Readers, what am I missing


  7. "If you were to explain cGMP to a 1st year graduate student, how would you do it?" I wouldn't. Given how much the 1st year is already suffering, it would qualify and cruel and unusual punishment!

    Given the small box that I have here, I would say GMP is 1) writing down what you are going to make and how you will make it, 2) make it, and 3) writing down what you made and how you made it.

    @Bad wolf: don't bother learning either one in school. Unless the prof has actual experience, he/she will get it wrong. Industry will not expect you to know it and will train you. It's part of the cost of doing business.

  8. By the way, if I was going to teach GMP to a 1st year graduate student, I would emphasize the importance of keeping a detailed and accurate lab notebook. It seems obvious, but a lot of people get in the habit of sloppy documentation. That's not all there is to cGMP, but if you understand how (and why) to keep a good lab notebook, you're 75% of the way there.

  9. cGMP is all the rage at my should see the extensive logbooks we have for EVERYTHING, from NMR use to solvent purchase to LCMS injection tables.

    Also, if you've never written down exact purity, CAS #, Lot #, and supplier of every chemical you use, well, sir, you're in for a treat!

  10. JNJ's smelly Tylenol is a fine example of cGMP.

  11. I would tell 1st year grad student GMP is about establishing high levels of control in order to maximize safety of the product: facility, materials, personnel, procedures, and testing with associated (and extensive) documentation all must meet and follow stringent (and largely predefined at every stage) standards. The c for "current" comes in as standards are always evolving requiring continuous monitoring and improvements.

    The difference between GMP and GLP is mostly applicable environments. GLP is focused on analytical (preclinical/tox/clinical samples and environmental/EPA) and animal testing. GMP covers manufacturing of drugs which includes testing. For the part the two are generally the same with what order of some steps, types of records and definition of certain roles being different.

    If one is in most basic research efforts/academia neither GXP is of great value and a real hindrance in practice. I have gone through much training in both GMP and GLP but no course/session is as good as being immersed in real world application (although reading 483s on FDA websites can be highly educational on how not to follow GMP).


  12. The easiest way I can explain the cGMP mindset is by explaining the manufacturing environment: The vast, vast majority of chemicals, regardless of how complicated or how delicate the process, are actually made by people who barely graduated high school. Many are high school dropouts with no education at all. Manufacturing staff may be extremely intelligent, very clever high school grads, who know an awful lot about their job and the equipment they run, but they are not MSc/PhDs. So you're designing and writing processes and SOPs for a high school grad who may not be able to perform basic algebra. You are not writing for peer reviewers who can sorta-kinda figure out your methods from the crap you scribbled on a paper towel, who will know how to back-calculate an extinction coefficient and the finer points of titration and so forth.

    The other thing is, everything is double-checked by someone else. All your work, and everyone else's work, must be more than accurate and precise, it must be right by someone else's standards. So if you've got a process that yielded, say, 100 mg/L when you ran it once, and someone else tries the same exact process that you scribbled on a paper towel and gets 10 mg/L 3/4 of the time, 50 mg/L twice and 100 mg/L only once, everyone will suddenly become amazingly humorless around you. In cGMP, the scale of the systems usually means that every screw-up on that order is costing millions of $$. Academics and the recently-graduated tend to have a real hard time with this notion, they think that peer review was hard enough.

    College grads in chemistry mostly get hired in cGMP environments as tech transfer, supervisors, troubleshooters, or someone who is designing a process for a cGMP environment--although generally they prefer process engineers who know a lot more about reactor design and process control software.

  13. The last post is the best description so far. I will say the only difference between how you would run a GMP batch versus a non-GMP batch is paperwork. If something is being produced under GMP, every little thing that happens must be written down.

    One of the main concepts in process is an operating window (hold the reaction for at least 3 hours between 75 and 80 °C, for instance). If something happens and the reaction is run outside of that window, you need to record it and hopefully have a backup plan in place that can rescue the batch. Running a little bit outside the window can have minimal impact because there are still reliable analytical methods run on the product to make sure the quality is fine, but the fact that the temperature reached 82 °C must be recorded nonetheless.

    The more crucial case is if something happens and an in-process control (IPC) fails. IPCs are analytical checkpoints, like an HPLC analysis and instruction that a reaction must go to x% completion before you move on to the next step. The procedure must be written this way in order to ensure that high-school educated operators can follow it. However, it's always a possibility that someone messed something up and the reaction isn't going to go to the same level of completion that it would if it were done correctly.

    When the process chemists are there for the first few GMP batches, they can usually figure out what to do if a reaction misbehaves or someone messes up. It's possible to rescue a GMP batch when things have gone wrong, but the paperwork to do it is hefty. Often, you have to back up your actions with data from smaller scale, non-GMP runs. Once you do this, however, and it's documented, it's easier for the engineers in the pilot plant to correct the same mistake in the plant.


looks like Blogger doesn't work with anonymous comments from Chrome browsers at the moment - works in Microsoft Edge, or from Chrome with a Blogger account - sorry! CJ 3/21/20