Testing ... 1, 2, 3 ... Testing

June 23, 2003

8 Min Read
Testing ... 1, 2, 3 ... Testing

Testing ... 1, 2, 3 ... Testing

by Jim Lassiter

The proposed rule on dietary supplement good manufacturing practices (GMPs),published in the Federal Register on March 13, is a lengthy document.Despite the six-plus year lag on the part of the Food and Drug Administration(FDA) in getting a proposed rule published since the publication of the AdvancedNotice of Proposed Rulemaking (ANPR), there are a host of issues and ambiguitiesremaining within the document.

One of the main concerns with the proposed rule is that manufacturers ofdietary ingredients and dietary supplements must test, test and test in order toassure themselvesand ultimately the FDA inspectorsthe product meetsappropriate standards. Those standards are called specifications, and FDAsfocus on specifications relates to the identity, purity, quality, strengthand composition of the resultant product. Despite much discussion and verbalsleight of hand in the preamble to the proposed rule, the additional testingrequirements are far from simple. The real issue and the ambiguity concerningthis matter is that there is some likelihood the testing manufacturers perform(or, more correctly, are capable of performing) may not be enough.

Companies are only required to test the finished product (dietaryingredient or dietary supplement) to ensure it conforms to establishedspecifications for purity, quality, strength and composition. While thisseems simple enough, even FDA notes that testing at this late stage of theproduction cycle may not be such a good idea:

[We] would recommend, but would not require, that you also test materialsreceived for these same specifications ... so that you do not end up having todestroy an entire batch of finished product ...

[68 FR, 49:12198,emphasis added]

This provides some indication FDA has made headway in understanding businesspractices. Since companies are not allowed to rely on Certificates of Analysis(C of As) for incoming materials to gain assurance that they conform to theirspecifications, they are stuck with testing those materials. Using some ratherstrange logic, FDA businesses companies cannot rely on information from thesupplier but are not required to test the supplied material: Its onlya recommendation. FDA has focused much of its public discussion of testing tothe lack of validity of C of As rather than on the overall testingrequirements. At a public outreach meeting in Oakland, Calif., held in May, thefocus of the questions FDA addressed concerning testing was on this lack of C ofA acceptance. FDA representatives kept returning to the phrase identity,purity, quality, strength and composition when pressed on the matter of whichtesting requirements would exist. In mis-directing the discussion to FDAsdisallowance of C of As, the entire matter of the challenges with testing andthe amount of testing required was virtually made to disappear. Why do yousuppose that is?

The likely answers to this are found lurking in the preamble discussion ofthe proposed rule. On the one hand:

If ... a scientifically valid analytical method for testing each batch ofdietary ingredient or dietary supplement is not available for any one of thoserequired specifications, then you would be required ... to test incomingshipment lots of components, dietary ingredients, or dietary supplements ...andto test inprocess for any such specification ...

[Ibid, emphasis added]

In other words, if you cannot test the finished product, you must test theincoming materials and the inprocess materials. That is, if you have methods toperform such testing. What remains unanswered is how, if you cannot test thefinished product, you are able to test the incoming raw materials or inprocessmaterials? FDA addresses this as follows:

We recognize that certain tests ... for certain finished products may notbe available due to complex finished matrices that would make such testingimpracticable.

[68 FR, 49:12197]

FDA passes this concern off as a matter of challenges with the final productmatrix. In a real-world scenario, it is often necessary to develop methods basedon officially validated ones out of sources such as AOAC to accommodate thesematrices. However, the result is not an immediately validated method. Thus,since the requirement is for a scientifically valid method, companies are backto testing incoming materials and inprocess materials. But this stillbegs the ultimate question: What if there is currently no scientifically validmethod available? FDAs answer:

While there may not be an AOAC or FDA method available, we are not awareof a situation where an appropriate scientifically valid analytical method isnot available.

[68 FR, 49:12198]

Very interesting. If this is accurate, and there is little reason to suspectit is not since FDA has said it so plainly, perhaps the agency could enlightenthe industry as to the methods themselves, since there are a host of items thatcannot be currently tested with a scientifically valid method. This isowed to the criteria outlined by FDA:

Test method validation determines whether a newly developed or existingtest method is accurate, precise and specific for its intended purpose.

[68FR, 49:12208, emphasis added]

Now you have to scour through the preamble to determine what FDA means by accurateand precise.

The words, accurate and precise, ... mean that the instruments orcontrols must be accuratethe recorded measurements are equal to the truevalue of the thing being measuredand preciseindividual measurements shouldbe close to each other when made under the same conditions.

[68 FR,49:12190, emphasis added]

Granted, this discussion is about instruments, but FDA chose to apply themeanings of these words to test methods. The fact that it chose to define them insitu rather than turn to the dictionary of their choice as it did for theword model is another discussion. Nonetheless, there is a specificrequirement for what a companys scientifically valid test methods must looklike. The rest is up to interpretation of the inspectors who come to the door.The strongest likelihood is that companies will not have validated testmethods for a host of products in their line. The next strongest likelihood isthat no one has such methods for a host of materials at the incoming,inprocess or finished product level. At this point, the party is over.

Without having the ability to test in accordance with the requirements,companies can neither manufacture nor release products. This is a serious issue.FDA remains unconcerned because:

The proposed testing requirements provide flexibility for testing ...basedon the availability of scientifically valid testing methods to perform testingon each batch of dietary ingredients or dietary supplements.

[68 FR,49:12198]

This flexibility just potentially flexed a host of products out of themarket.

To close the loop on one last little absurdity concerning this area of theproposed rule, lets go back to FDAs dismissal of C of As. Whenmanufacturers purchase materials for inclusion in a product, they purchase,among other things, dietary ingredients. These dietary ingredients must bemanufactured in accordance with the same GMPs manufacturers are beholden to.Therefore, the testing they perform on the materials must conform to the samerequirements. Yet, somehow, between the time it leaves their facilities andarrives at the manufacturers location, something must have happened toinvalidate these test results. It must have occurred during the distributionprocess. Oops. This is covered by the same set of regulations. Then, thequestion is: Why the redundancy? FDAs answer:

Using a supplier certification, guarantee, or certification in lieu ofperforming testing on each shipment lot of components, dietary ingredients, ordietary supplements required in accordance with this section is not appropriatebecause it is possible that a suppliers certification or guarantee may notensure the identity, purity, quality, strength, or composition of a component,dietary ingredient or dietary supplement. [68 FR, 49:12198,emphasis added]

Applying this same rather circular logic, it is equally possible that youranalysis may not ensure the identity, purity, quality, strength orcomposition of your product. Then, what purpose does this section of theproposed rule serve?

The provisions would require manufacturers to evaluate the identity,purity, quality, strength, and composition of their dietary ingredients anddietary supplements.

[68 FR, No. 49 Page 12158]

This is the stated purpose of the GMPs. There is no product assuranceintended out of performing all of these steps, only that the steps wereperformed. The focus word here is evaluate. Ironically, using a dictionaryto determine the meaning of evaluate, it states in order to evaluatesomething you need to test it. That is, if there are scientifically validmethods. Thus, these proposed GMPs are intended to make manufacturers testthings, not to provide assurance to the consumer. This comes in spite of theexamples FDA cites as being eliminated (for example, product recalls going downto zero) with the implementation of these GMPs.

This piece of the proposed rule is but one point of concern. The dietarysupplement industry needs to engage in discussion with FDA with strength andpurpose on this matter. It is wholly appropriate to perform testing (or haveothers perform it for you). It is also wholly appropriate to have methodsdeveloped that apply appropriate and meaningful reference standards in order toensure product quality. What is inappropriate is a culling out of products basedon the current level of science. Many of those products have been in themarketplace for a long time. Eliminating them solely because there has not yetbeen a specific method developed makes no sense. Further, it makes even lesssense to accept the results you generate in order to conform GMP regulationswhile dismissing the results your supplier attained while operating under thesesame regulations.

The time is now to get involved and determine your status relative to theseproposed requirements and then act either directly or through your tradeassociations to let FDA know redundant or obtuse testing requirements offerlittle to the end userthe person both FDA and the industry are most concernedabout.

Jim Lassiter has more than 25 years of experience in regulatory affairs andis currently director of consulting and regulatory services at ChromaDex Inc., aglobal provider of high purity reference standards and research, regulatory andproduct consulting services. Lassiter can be contacted at [email protected].

Editors Notes: Additional commentary on this topic is available here.

INSIDER welcomes industry commentary and letters to the editor. Direct yourcomments to Heather Granato, group editor, 3300 N. Central Ave., Ste. 2500,Phoenix, AZ 85012, FAX (480) 990-0819, [email protected].Letters may be edited for content or clarity and do not necessarily representthe views of Virgo Publishing or


Subscribe and receive the latest insights on the health and nutrition industry.
Join 37,000+ members. Yes, it's completely free.

You May Also Like