Peter J. Werth, President and Chief Executive Officer, ChemWerth, a US-headquartered full-service generic active pharmaceutical ingredient (API) development and supply company providing cGMP quality APIs to the regulated markets, is a proponent of removing data to a data-integrity (DI) file from data history, if it isn’t linked to product quality. According to Werth, FDA is unnecessarily obsessed with data integrity issues, and pharma companies have begun to fear FDA inspections. This needs to change, Werth tells PharmaCompass.
Excerpts from the interview.
Can you tell us about your journey in the world of
pharmaceuticals?
I
began by supplying generic cough and cold APIs by setting up a chemical
business back in 1975. There was a small group of companies (first generic companies)
that mainly sold cough and cold products. These companies wanted to sell patent-expired
generics and we developed select APIs for them. There was no real FDA (US Food
and Drug Administration) approval process involved.
By 1982, I realized the generic industry would be big business and API sourcing was key to success. Since I knew the API business, my wife and I formed ChemWerth – a company that supplied generic APIs.
By
1988, ChemWerth was working with 10 American and three Chinese manufacturers.
But the generic scandal changed everything. My American partner factories gave
up the API business when new
regulations came in. Chemical plants could no longer manufacture APIs. From my
previous 11 years of dealing with China, I knew that almost all of the dosage
form factories also produced APIs. This was especially important for the
injectable grade products. By the early 1990s, ChemWerth was developing APIs
for highly toxic oncology injection drugs and many antibiotics. We successfully
introduced doxorubicin, daunorubicin, mitomycin, bleomycin, etoposide, ifosfamide and several antibiotics such as lincomycin, clindamycin phosphate, amikacin sulfate, tobramycin and vancomycin (all from China).
Today, ChemWerth represents 25 Chinese factories for about 100 products. We still represent one American API supplier along with several Indian manufacturers. We recently expanded to supply veterinary drugs and partnered with a major supplier of polypeptides and a world-class supplier of heparin and related products.
You have been working with China for many years now. When
did you first come to China and how has the Chinese pharmaceutical industry
transformed over the years?
I first went to China in 1979. During that
trip, I realized I could successfully deal with the Chinese. Most of my friends
who went to China said it was impossible to deal with them. I always thought
this gave me an advantage in China. Over the next seven years, I kept in touch
with China, but did very little business.
In 1986, I was running ChemWerth and was a
consultant with a company to help its large pharmaceutical facility produce generic
APIs. We targeted clindamycin phosphate for development, since the product had
good sales, and was off patent. It also had no competition because Upjohn
(an erstwhile US-headquartered pharma company that got merged with Pharmacia in
1995, which in turn got bought over by Pfizer) controlled all the key starting material – lincomycin
hydrochloride. Using my contacts and my ability to deal with Chinese factory directors, I managed to break Upjohn’s stronghold in lincomycin. I convinced three factories to work with me so that I could file a DMF (Drug Master File). Clindamycin became a very good product for ChemWerth and gave me reason to visit China regularly to develop new business.
In those days, factories were generally old and needed upgradation. Documentation was always good, and in detail. However, it was mostly in codes and therefore useless to most others. It was extremely difficult to convince the factory to provide a true and detailed process to ChemWerth to file a DMF with the FDA. There is still a tendency (amid factories in China) to hide important facts. As a result, it takes ChemWerth at least three versions/iterations before we think we have got the right process details.
Today, the hardware in China’s pharmaceutical factories is excellent, and is usually purchased from the best overseas manufacturer. The software is good too and is likely to improve drastically as manufacturers realize software is equally important as hardware.
In view of the current GMP concerns being uncovered at
Chinese companies, are you still focused on China as a source of
pharmaceuticals?
In my mind, and in the ChemWerth representative factories, there are no
GMP concerns, no quality concerns, and no data history concerns. The FDA's
fascination with data integrity focuses on a problem that does not exist in more
than 90 percent of the pharmaceutical factories. There are at least 18 reasons
to remove data from data history. All of these 18 reasons are related to
analyst error and equipment failures (laboratory gross errors) and not related
to quality or GMP manufacturing. (Click here to read Werth’s 18 reasons to remove data).
I agree that gross errors/laboratory gross errors should not be deleted. I also agree that deleted gross errors/laboratory errors not related to quality do not deserve a warning letter.
However, the FDA should also know the main reason data is deleted is because the analyst (often) knows the result is not correct for reasons not related to quality (gross errors/laboratory gross errors). They do not want to conduct an expensive, time-consuming, and cumbersome OOS (out-of-specification) investigation. Therefore, they delete the data, weigh out new samples, fix the problem, analyze, and either receive acceptable results or rejection results.
Today, China has the most modern manufacturing facilities and the money to invest in software to manufacture the highest quality with the best traceable records. The Chinese need to trust the FDA to treat them fairly and not rely on two and three-year-old DI information. They will do the job correctly now and in future. But they cannot change the past. As stated earlier, all DI product-related observations have proven that the product quality and product history were acceptable.
China can only become a stronger API supplier. China’s volumes will increase and costs will decrease, because the domestic market is growing rapidly. The country will have cost/pricing pressures and will need to concentrate more on the difficult products that require expensive equipment. The standards set by the China FDA (CFDA) match the US FDA standards. And this will make it easier to operate under one quality and documentation system. Most importantly, they have the money to invest in the latest technology, best scientists, and equipment.
In your view, what can pharmaceutical regulators and the industry
do differently in order to communicate that the product quality is not at risk,
given the environment where data-integrity violations have become commonplace?
The FDA needs to understand that DI has no relationship with product quality. We have two recent examples. In the first case, DI issues applied to 70 batches. All 70 batches were retested and all were found to be in specification. The second case gave the same results – between two inspections over 100 batches were considered potentially fraudulent. All 100 batches were retested and were found to meet the specification. At both the factories, the data history showed all the batches are in specification – this implies 100 percent correlation to quality. The data integrity listed 100 batches and no batch failed specification – this shows 0 percent correlation to quality.
From a DI perspective, FDA inspections in factories that only produce APIs for export should be different from inspections in pharmaceutical companies that export APIs and dosage form. In reality, API manufacturers cannot cheat. They sell their product to dosage form manufacturers, which in turn are responsible for the quality that goes into the dosage form.
Off specification batches will be rejected and returned to the API manufacturer. While inspecting factories’ customer complaint file, it is rare to find batches being rejected and returned to the factory. The same does not apply to pharmaceutical companies that produce both API and finished dosage. They can hide information. Data integrity is more of an issue with companies that manufacture both API and dosage form.
What steps is ChemWerth taking with its partners in China
to improve compliance standards?
ChemWerth is addressing this issue with the FDA and ChemWerth representative factories. In broad terms, data history should be paramount, since it is related to product quality. Data integrity is a part of data history. ChemWerth has prepared a standard operating procedure (SOP) to handle electronic data.(Click here to read ChemWerth’s SOP). A simple version is that when one or both samples are out-of-specification, the analyst’s supervisor (who committed the error) reviews the data with the analysts. If they determine that one of the 18 reasons for analyst error or equipment failure is the cause of the failure, then this is noted. Three new samples are prepared, the error is corrected, and the samples are run and if all samples pass then these results are retained in data history and the product is passed. The previous bad result is documented and removed to the data integrity file. No lengthy, time-consuming, and expensive OOS report is conducted. This makes the procedure factory-friendly.
In the above example, if one of the three new samples fails, then all results are retained in the data history. The batch is rejected for rework or reprocessing.
I believe the present atmosphere between the
FDA and API manufacturers is at an all-time low. The industry feels the FDA
inspectors are looking at mostly irrelevant data to prove that the factory
operates fraudulently. Factories no longer welcome FDA inspections, out of fear
of DI failure. Factories feel the FDA is using DI to point out fraud where
there is none (excluding Hisun
and Ranbaxy).
In the past, we always welcomed FDA inspections as part of doing business. We
need to get back to that time and place.
Share