Bites of Bitter PAI

Two PAIs in the face

"It was just after FDA had initiated the Pre-Approval Inspection process," reports our GMP Agent-in-Place. "We had put together a full PAI preparation process, including a policy and SOPs, all in a honking big binder along with pertinent checklists and FDA guidance, and all of which we sent to all of our US manufacturing sites. But management said that since our first PAI would be for 'just a topical product,' we didn't need to go through the full drill.

Control

"Well, the inspectors came, they saw, and they didn't concur. The product failed and the inspectors slapped a withhold approval on us."

"We worked like dogs to make the necessary changes. We were making progress, but the Quality Department didn't think the site was quite ready yet. Then management stepped in again: they wanted us to amend NDA and request re-inspection now, 'Why not try it,' they said. 'If it works, we're ahead of the game. If not, we can fix it and go on.'

"So we got a second PAI," groused our Agent. "QA was right: we got another withhold approval, but this inspection cited procedures used at another site as well. So not only did we delay the launch of the topical product, but FDA also fired up a 'for cause' inspection of both sites. That inspection produced a warning letter. The warning letter cited us for discarding some original, raw data after the 'one year beyond dating period' required by the GMPs. NDA rules, however, say we have to retain the raw data at least until NDA approval...which we didn't have because we failed the first PAI.

"We did get approval in the end...years later. We needed to repeat studies we'd already done, because we'd destroyed the data.

"So much for 'just a topical product.'"

PAI in the dustbin

"The product was not a high priority, and the NDA had been filed seven years before FDA started the PAI process," recalls the responsible GMP Agent-in-Place. "And for the intervening seven years, we'd gotten no NDA questions. It was like we and FDA had both forgotten about it.

"And then FDA finally came.

"It was our first PAI ever at this site. The inspection was going well, but FDA required original development data to back up the reports, graphs, and tables submitted in the NDA. Now, this product had been developed under contract by a small lab a thousand miles away. When we tried to track them down, we found that the lab had closed and the researcher had died! Now, it may seem heartless, but we were desperate: we tracked down the widow...and she told us that she had trashed all of his work papers. She hadn't seen any need to keep them around, and, after all, when the research was done, there was no FDA requirement to archive this old data.

"Luckily, though, it was one of the first PAIs for our FDA inspector, too. And since the filing was very old, and since I talked myself hoarse trying to persuade him—without any company recognition, I might add—he finally accepted this explanation and gave us a pass.

"By the way, after all the delay and all the effort, we made 40 batches of the product to roll it out to the market, and sold only one of them in the two years before folding it."

More failed PAIs

"Can you believe it?" our GMP Agent-in-Place sputters. "We failed two PAIs for this oral capsule product!"

"It was during the early days of the FDA's PAI program, and none of us knew what was coming, so we did the usual thing, filing the NDA based on research and small scale data.

"Obviously, one of the prime PAI criteria is that the site be capable of production. But the FDA showed up right after our filing went in, and we had torn up the manufacturing area to upgrade the utilities and install the necessary new equipment.

"So I meet the inspectors at the door, tell them we're not ready for an inspection, and turn them around.

"Of course, they issued a formal withhold approval. Later, during the second PAI, the inspector didn't like our justification for changing our granulation drying step from tray-in-oven to a fluid-bed dryer. We argued that this was a call for the reviewing chemist, but the inspector was adamant, saying 'You can get the chemist to overrule me then.' We knew that would never happen, and it didn't. So we received another withhold approval.

"The good news," sighs our hard-pressed Agent, "is that six months and one new bioavailability study later, the NDA was approved."

Pharmaceutical Technology's new monthly "Agent-in-Place" column distills true-life cautionary tales from the secret files of Control, a senior compliance officer. If you have story of clueless operators, oblivious management, inopportune lapses of judgment, or Murphy's Law in action, please send it to Control at AgentInPlace@advanstar.com We won't use any names, but if we do use your tale of disaster, courage, or just plain weirdness, Control will send you a coveted Pharmaceutical Technology t-shirt.