I was just watching “Holmes on Homes,” a program on HGTV where an expert contractor, Mike Holmes, visits houses and repairs and refinishes work that the original contractors did poorly—they took shortcuts, were inept, and, in some cases, defrauded the homeowners. At one house, Holmes found that massive rebuilding was needed because a contractor that had built a new addition did not take into account the state of the existing infrastructure. That is, did the electrical and water systems support the new rooms? Did the foundation and roof work with the new addition? Did the contractor understand what the process was, at all?
Relating this to the pharmaceutical industry, it brings to mind the expansion of the QC department at the company for whom I worked in the 1980s. We expanded quite rapidly (a “love note” from the FDA can do that), more than doubling our staff, expanding from four to 24 HPLCs, and so on. We quickly found that “necessity is a mother.” The analysts had very small areas in which to work, leading to what I referred to as “crowded rat syndrome.” People who had been close friends for years were suddenly fighting over small things (like six inches of additional space ) and, of course, the quality and amount of work suffered.
How does that translate into today’s situation? What an FDA or ICH Guidance says and what an individual FDA or other agency inspector does are oft en at odds. I was just made aware of a major company that used the best QbD techniques for the production of an API. Most of the inspecting team were happy with the results; however, one insisted that the company perform the (circa 1955) tests, as outlined in the USP. The resulting work doubled the time and cost of the approval. (That will certainly encourage other companies to use QbD.)
When I was involved with the NIR analysis of tablets, we managed to make dosage forms between 80 and 120% of label. Obviously, the NIR and HPLC analyses could only be performed on individual tablets, effectively a content uniformity analysis. This always gives a better picture of a process’ consistency, but, according to the letter of the Guidance (as per our QA Department), one needs to use tablets from 70 to 130% label claim for CU. What no one ever speaks about openly is that no one actually makes 70 to 130% tablets. Nor, for that matter, does anyone ever make 50 to 150% tablets for HPLC “recovery” validation. In truth, the analyst adds powdered API and excipients to a fl ask, fills it with solvent, shakes the flask, and ta-da, we have “recovery.”
To hold spectroscopic QbD/PAT analyses to these (wink-wink, nod-nod) rules is insane! There also seems to be a deferral to one or more Agency members of an inspection team to the other inspectors on that team; to point out potential non-conformity to a written Guidance or rule might embarrass them. It is far better to put companies and patients at risk than upset their colleagues.
How about GMP regulations? Since the 1960s, GMP regulations stated that a statistically significant number of samples need to be analyzed for each lot of product. What school teaches that 20 samples of 1,000,000 to 5,000,000 units are “statistically significant”? Nonetheless, NDA after NDA is approved with 20 tablets being assayed for product release. If a reviewer okays the application, another investigator certainly wouldn’t want to hurt the reviewer’s feelings by pointing out what the law says, would he/she? Better to just let it slide . . . like it has for the last 40+ years, I guess.
Since manufacturers resist doing more tests to show the actual “goodness” of the product and the Agency (supposed to be our advocate) doesn’t insist on best practices, we are in trouble as consumers. (Consumers don’t help either when they buy on-line drugs to save money.) Instead, we, as consumers and patients, should demand excellence from all parties involved in providing health care.
We as an industry need to develop better blueprints for our “houses,” and then build them to specification the first time. And we need inspectors who appreciate our handiwork. If not, we’re doomed to rework and repairs. To quote Pogo: “We have met the enemy and he is us.”
