Product Testing and the GMPs
June 2008 was the deadline for large companies with more than 500 employees to comply with the final good manufacturing practices (GMPs) for dietary supplements. As the June 2009 and June 2010 deadlines for midsize and small companies loom closer, some manufacturers are still struggling to fulfill the rule's product-testing requirements. The April 2008 recall of two dietary supplement brands found to contain 200 times the amount of selenium listed on their labels highlights the fact that not every company is ready, willing, or able to hit the ground running.
One way for smaller manufacturers that can't afford expensive analytical equipment to meet the rigorous testing requirements of the GMPs is through the use of contract laboratories. A key element of any successful QA/QC program, according to Elan Sudberg, COO of Alkemists Pharmaceuticals (Costa Mesa, CA), is identity testing. Aside from the fact that identity testing is required in the final GMP rule, there are several reasons why it's important, says Sudberg.
"Identity testing is paramount to purity, strength, and composition testing, as it doesn't make sense to spend any time or money investigating the latter without confirming the identity first," Sudberg says. "Also, it can be considered irresponsible not to test for identity if a company is truly offering a high-quality product that can guarantee consumer confidence in this competitive and demanding regulatory market."
Nearly everyone agrees on the importance of identity testing. Not everyone agrees, however, on how to carry out identity testing. Although FDA requires manufacturers to use analytical methods, it doesn't specify which ones.
"Current GMP requirements state that manufacturers are required to evaluate 100% of their ingredients via a scientifically valid method," Sudberg says, adding that the methods can be as simple as organoleptic testing or as complex as a full chromatographic profile. "Unfortunately, FDA has left the decision of which method to use to the industry, where choices may vary widely in suitability, allowing some to use inadequate methods and release substandard material into the marketplace."
FDA has indicated a preference for validated analytical methods developed by AOAC International (Gaithersburg, MD), which is also recognized by Codex Alimentarius (Rome), the International Organization for Standardization (Geneva), and the International Union of Pure and Applied Chemistry (Research Triangle Park, NC). According to Title 21 of the Code of Federal Regulations, FDA's enforcement policy is to use AOAC's methods of analysis "when available and applicable." However, methods aren't available for all ingredients, especially herbs. Moreover, different countries may have their own local requirements.
"When choosing which method to use, it's important to determine the specific regulations of the government overseeing the product's final destination," Sudberg says. "For example, Australia's Therapeutic Goods Administration does not allow Fourier transform infrared spectroscopy (FTIR) due to its questionable ability to identify botanicals."
Alkemists uses the philosphy of "fitness of purpose" in its testing program, exploiting the strengths and compensating for the weaknesses of each available method. Sudberg refers to the strategy as an "orthogonal" approach.
"Microscopy is a terrific and inexpensive technique for identity of raw botanicals," Sudberg explains. "However, it has no role in extracts or liquids. And while high-performance thin-layer chromatography (HPTLC) has proven to be the global method of choice in identity testing for raw materials, extracts, liquids, and finished products in a variety of matrices, it too has its shortcomings in detecting adulterants that are inert or are not chemically active. This is where the two methods work together."
Most people who have worked in the industry long enough have heard anecdotes about counterfeit ingredients that are passed off as the real thing. As the price of raw materials rises, the likelihood that manufacturers may find fake or watered down ingredients in their production line rises as well. However, negative test results could stem from factors other than fraud, such as simple confusion about botanical species, honest mistakes, or a combination of all of the above.
"We fail products on an infrequent basis," Sudberg says. "When a sample is rejected, it is more often than not an incorrect or closely related species that is submitted, as with arnica (Arnica montana), sarsaparilla (Smilax regelli), or licorice (Glycyrrhiza glabra), to name but a few. On occasion, a sample comes in that is a complete blank—as if it is simply diluted beyond recognition. While there may be an ingredient in the powder, it is not, for all intents and purposes, recognizable as the herb that it is purported to be."
Trendy ingredients that are in high demand and short supply may be likely candidates for a negative test. For instance, the African succulent Hoodia gordonii was a frequent victim of adulteration several years ago.
"Where there is the opportunity to make fast cash, you will find those entrepreneurs that are unfamiliar with the basic fundamental understanding of what quality means and to what extent it can be controlled," Sudberg says. "The hoodia industry as a whole was severely damaged as a result. There are now only a handful of companies that have managed to survive by responsible identity testing of each shipment to be certain that quality is maintained. Currently, hoodia is quickly making its way off our list of frequently failed botanicals."
Many manufacturers wonder what happens to ingredients that fail identity testing. Are they sent back to the original supplier and withdrawn from the market, or do they make the rounds until they eventually wind up in an unsuspecting consumer's dietary supplement? Due to the difficulty of tracing the migration of raw materials, that's a question that remains unanswered.
"When an ingredient fails our identity testing, I hope it is returned to the supplier and destroyed," Sudberg says. "Unfortunately, we have seen the same lot come back to us for a second round of identity testing, only to fail again." Sudberg adds that contract laboratories can't prevent a scenario in which a manufacturer shops around for a less-experienced or less-capable lab that ultimately accepts the sample. "To be honest, once we reject a sample and all of our due diligence is complete, we rarely have to deal with it in the future," he says.
While manufacturers and contract laboratories usually agree on the need for identity testing, they don't always see eye to eye on the results. Despite all of the hard work that goes into sample preparation and testing, disagreements aren't that uncommon, according to Sudberg. Fortunately, many of these conflicts can be resolved by carefully examining each step in the process.
"When a material does fail our analysis, we always go back and double-check our work and the data it produced and retest where appropriate to confirm our findings," Sudberg says. "Lack of clear communication between manufacturers and their customers can also lead to this outcome."
Sudberg says that these disputes often revolve around the issue of traceability. "Somewhere throughout the chain of custody, someone dropped the ball," he explains. "We always give the benefit of the doubt to all parties and work with all parties to help determine where the chain was broken."
When Alkemists works with crude botanicals, it tries to verify that it received the correct plant species and parts, using multiple accessions of voucher or authentic specimens whenever possible, Sudberg says. He adds that in the case of extracts, Alkemists asks the original supplier to provide a sample of the crude raw material source of the extract, so that it can perform microscopic and chromatographic retesting.
"This information will frequently tell us whether there was a potential problem with the original manufacturing process or with the way we prepared our reference samples," Sudberg says. "By the time all the investigative work has been completed, we usually find the source of the inconsistency and everyone is relieved."