The past few weeks have brought more news of expensive pharmaceutical manufacturing failures. Not only Ranbaxy’s Consent Decree, which will cost over $500 million, but a mixup that drove Novartis to recall some bottles of over-the-counter medicines, which may have contained opioid prescription painkillers.
We also learned that a couple in Washington is suing Johnson & Johnson, alleging that their toddler died after taking Children’s Tylenol that contained 100 times the appropriate level of API. The product was part of a batch that had triggered a “phantom recall,” in which J&J hired people to buy back lots of questionable products. Is it possible that distraught parents accidentally overdosed the child? Administering acetaminophen in its various liquid forms, especially to young children, can be problematic, as FDA disclosed in an April report and safety communications throughout last year. But 100 times the API?
One hopes that the battery of QC tests routinely used today would detect superdosing of that level. Could contaminants have triggered allergic reactions? We may never know what really happened, but the story did make me wonder: How oft en do drug manufacturing problems result in patient reactions that we never hear about, that are attributed to something else?
You deal with huge levels of risk every day. There’s always a chance that ingredients might interact with one another, or that even a tiny change in their manufacture, or an issue with some piece of equipment and its operation and maintenance, at your or a supplier’s plant, might affect final product quality. Are you using the best approaches and technologies to handle all this risk?
One regulator minced no words at the IFPAC meeting in Baltimore last month: “It’s an indictment that this industry has not forced itself to form clinical connections with assays, and measurements such as dissolution, impurities, content uniformity . . . every company should be able to prove that every dose in every batch is as good as clinical testing batches.”
Such direct connections to the patient may remain years away, but at least a foundation of process understanding and risk assessment is being established to allow them to be made. More evidence? A double digit increase in attendance at IFPAC this year. At the conference, less was heard about three-letter acronyms and more about how to apply them, with PAT living on as smarter validation and QbD as process understanding and smart drug development.
More drug companies are using statistical process control (SPC) and process capability analysis in their everyday operations. Some are also applying Raman and NIR in new ways to ensure consistent ingredient and product safety and quality.
People within the industry are also openly discussing the need for new QC test sampling and quality standards, to bring everyday practice closer to the original spirit of cGMP mandates for “statistically significant” testing and process control.
More attention is being paid to out of trend data, as authors from Noven Pharmaceuticals outline in a recent article. All of these efforts are bringing science to bear on drug manufacturing uncertainty.
Driving the science is the voice of the customer. Patients are demanding low prices, improved dosage forms and convenience. But even before that, they and those who speak for them are demanding consistent safety.
The next time someone suggests that you cut a corner or use an outdated technique, even during these times when everyone is doing more with at least 10% less, listen for your customer’s voice and you may hear your own. Is that good enough? Would it be good enough for you or for your own child? Don’t you, and all of us, deserve better?
