The Impact of Impurity Analysis on Future Regulations

Even at minute quantities, unwanted chemicals present in pharmaceutical ingredients may greatly influence a drug product's efficacy and safety. While researchers strive to eliminate or control impurities, they rely on fast analytical tools with high sensitivity and specificity to better detect, identify, quantitate, and characterize impurities. The technology for impurity analysis has improved beyond the traditional chromatographic (e.g., high-performance liquid chromatography [HPLC] and gas chromatography [GC] for volatile impurities) and spectroscopic methods (e.g, GC–mass spectrometry [GC–MS]). Scientists are gaining an ever-clearer picture of their materials' constituents. In some cases, technology is revealing never-before-seen impurities even in compounds thought to be well understood.

Along with improving analytical tools comes a growing concern about genotoxins—molecules that may bind or break DNA even in parts-per-million concentrations—which are only now practical to monitor.

Qualification of Impurities in ANDAs

Better information about pharmaceutical impurities is generating new questions. What is the best way to handle them? How should we report them? How do they really affect overall product quality? How should control procedures be put in place? Ingredient suppliers are working—and debating—with regulatory agencies to develop guidelines designed to eliminate redundant testing for APIs, explain the composition of excipients, and reasonably control the genotoxic impurities in both.

Improving analytical methods

Nongenotoxic impurities. LC and GC techniques with conventional detectors have been the traditional analytical methods for identifying and quantitating nongenotoxic impurities. According to a 2003 survey, HPLC accounted for approximately 53% of the reported analytical separations used in 1999–2001 (1). In these applications, specificity, sensitivity, and matrix interference remain the primary analytical challenges (2). These techniques may not necessarily measure impurity levels accurately. They assume that impurities are structurally related to the drug substance, and therefore have similar detector responses; this is not always the case.

Hyphenated techniques, including LC–MS, GC–MS (accounting for 8% and 2%, respectively, of the methods used in 1999–2001), and chromatography tandem mass-spectrometry (LC–MS–MS) are more sensitive and can provide better separation (see Figure 1) (1). In addition, MS–MS yields more accurate information about the structure of impurities.

Figure 1: Scientists are increasingly relying on hyphenated and tandem techniques to analyze their compounds, thereby enabling them to better detect and identify impurities (Waters Quattro Micro LC–MS–MS instrument with Waters Alliance 2695 HPLC).

"Incorporating this information with those obtained from NMR and IR techniques, one can identify the impurity with a fairly good degree of accuracy," says Liakatali Bodalbhai, group leader, Analytical Sciences R&D, DPT Laboratories (San Antonio, TX). "However, only by synthesizing the possible molecule and comparing its fragmentation pattern in MS and retention time by HPLC to the impurity of interest can one conclusively identify that impurity."

Bodalbhai points out that the development of multiple ionization modes in LC–MS and LC–MS–MS, namely electrospray ionization, atmospheric pressure chemical ionization, and atmospheric pressure photoionization, has expanded the capabilities of these techniques for a large class of compounds. Newer technologies include fast chromatography data treatment such as signal averaging, new LC–NMR platforms, thermogravimetric–MS for volatile impurities, and LC–MS–NMR (3).

"Incorporating NMR to LC–MS would make this technique one of the most powerful tools for an analytical chemist," says Bodalbhai. With this technology, scientists can fragment a molecule and obtain structural information as well as map atom-to-atom connectivity of the molecule, thereby generating three-dimensional structures of the isomers.

"On-line implementation of techniques is gaining popularity and the ability to conduct identification, characterization, and quantitative analysis at the same time will be one of the trends of the industry," predicts Qingxi Wang, director of Global Pharmaceutical Globalization, Merck Co., Inc. (Whitehouse Station, NJ). "For example, interfaced LC–MS–NMR could potentially be a widely used technique for the characterization of organic impurities."

Wet chemistry is currently the compendial method for inorganic impurity analysis. However, ICP–AES and ICP–MS are widely used and potentially methods of choice because of their higher sensitivity and selectivity. Other techniques for detection and characterization of inorganic impurities include laser induced bombardment spectroscopy, scanning electron microscopy-energy dispersive X-ray, and X-ray fluorescence (3).

Analysis of genotoxic impurities. In general, impurities should be quantitated at levels ≥0.03 or 0.05% by weight according to ICH guidelines (see Table I). Genotoxic impurities or potential genotoxic impurities must be controlled at levels significantly lower than the 0.03–0.05% levels that are typically reported by an HPLC impurity assay. Typically, developing limit tests (e.g., <50 ppm) for highly toxic impurities is readily achievable, however it can be difficult to develop a test to control a particular genotoxic impurity at 1 ppm (0.0001% w/w).

Table I: ICH Q3A (R2) thresholds for impurities in drug substances.

"There is a huge difference between the two," says Sandeep Modi, director of Quality Management at Bristol-Myers Squibb (New Brunswick, NJ). "For genotoxic impurities we need very sensitive and selective methods. One needs higher sensitivity to determine ppm-level impurities and selective methods to separate low levels of genotoxic impurities from base line noise and other organic impurities. The typical HPLC methods with a nonspecific detector (e.g., UV) that are used to measure organic impurities may not be appropriate to quantitate low ppm levels of genotoxic impurities."

The quantitation of low levels (in the range of ppms) of impurities is the challenging part, notes Modi, and using specific detectors such as MS or MS–MS with LC will significantly improve the method selectivity and the quantitation limit. The goal for scientists is to identify potential genotoxic impurities early in development, develop analytical methods to test for these impurities in the intermediates, and if possible, to demonstrate that the manufacturing process controls them before reaching the final drug substance.

"If you eliminate them early enough, then your actual active drug substance is pure, free of genotoxic impurities," says Modi.

A more flexible USP

Improvements in analytics as well as those in synthetic chemistry also have led to a greater number of synthetic routes discovered. Molecule synthesis has improved with better reagents, better techniques, better reaction types, and better control. As a result, some APIs now have as many as five possible synthetic routes. And APIs produced according to those synthetic routes may have different impurity profiles.

A USP monograph lists the minimum set of impurity tests that must be conducted. Currently, a company making an API must comply with all the tests listed in the published USP monograph, even if the company uses a synthetic route that is different from the original and therefore has an impurity profile that may or may not contain all the impurities listed in the monograph. Even if the company knows for certain (either because of the route of synthesis or some other scientifically justifiable means) that some monograph impurities will not be present, it must still test for every substance on the list. In addition, the company would still have to test for any other impurities that might be in its compound.

Flexible monographs were designed to eliminate redundant testing. They allow scientists to conduct only those tests that are relevant to their compound. If a company's API has an impurity profile that is different than the original, the company could conduct a full scientific evaluation of the material and file information with USP. Moreover, the company must notify FDA upfront that it plans to submit a different impurity profile, including the limits on each impurity as part of its drug application. The monograph would become effective upon FDA approval of the product.

Not surprisingly, generics companies stand to benefit from the flexible monographs program. Still, these companies face technical challenges in qualifying impurities (see sidebar, "Qualification of impurities in ANDAs").

"GPhA has not put out a unified public statement or position on flexible monographs," says Gordon Johnson, vice-president regulatory affairs for Generic Pharmaceutical Association (Arlington, VA). "However, we have been supportive of this effort in a number of stakeholder meetings with USP."

Excipient components

While API scientists aim to eliminate or control impurities, excipient makers are trying to move away from the impurity mindset and instead are advocating a better understanding of excipient composition and components, especially if these components are essential to some performance of the material or the drug product.

"Right now, there is a gross misunderstanding of this issue. Excipients come from a lot of different areas. They are from sources that are not synthetic chemistry, and by design they are not supposed to be pure. They are not in a drug product to be pure, they are there to be functional," says Dave Schoneker, director, Global Regulatory Affairs, Colorcon (West Point, PA) and Chair-Elect of the International Pharmaceutical Excipients Council of the Americas (IPEC Americas).

Sometimes the overall performance can be attributed to the presence of a minor component (USP refers to these as concomitant components; they have also been called essential minor components). "So 'pure' excipients are not necessarily good excipients, depending on performance," says Schoneker.

For example, microcrystalline cellulose contains several components, including pulp residues, sugar residues from the hydrolysis, and hemicelluloses. When too much of these "impurities" are removed, it can compromise performance of the excipient. Determining how much is too much, though, is not an easy task.

"It's a bit complicated because we don't have the analytical methods and understanding that we really need to be able to say 'These are the definitive limits for this component,'" says Chris Moreton, vice-president of Pharmaceutical Sciences at Idenix (Cambridge, MA). The only limitation, for safety reasons, is on toxic components, especially for parenteral preparations such as parenteral grade dextrose, mannitol, or povidone that also have restrictions on endotoxins.

"We don't have a good enough understanding of any excipient to be able to say, 'This is the item that needs to be controlled and it will be the same for every formulation,'" says Moreton.

"The big issue is the overall compositional profile and understanding how that relates to the use of the excipient," says Schoneker. "Users sometimes ask for an impurity profile when they should be asking for composition." This understanding is especially critical when users contemplate switching suppliers. Although both materials may meet all specifications, they may have different concomitant components, and so their performance during actual processing may differ.

Obtaining a useful compositional profile will again put the focus on good analytical methods.

"Some of the hyphenated techniques such as LC–MS may be able to reveal components that were never seen before. And if we are ever able to get a good solvent that will dissolve celluloses easily, we might be able to get more information. In order to go deeper into the composition of excipients we're going to need better applications or new analytical techniques to understand what we are doing," says Moreton.

Even if these techniques were available, the data generated might not be meaningful. "I would want to see data going over several years to see what the inherent variability of those components were in relation to time of year, harvest, and so forth. Without that, just introducing the technique is not going to help. In fact, it's going to make things worse."

Collaborating with the Japanese and European branches of TriPEC, IPEC-Americas is working on a composition guideline to help excipient suppliers communicate the overall composition of their products and why the components are present while retaining confidential proprietary information.

"We are trying to initiate a better dialogue between makers and users so that users understand what it is they are working with, using good science to determine how materials should be described," says Schoneker.

The IPEC guideline is the first step. Afterward, the biggest challenge will be getting excipient users to think of excipients in the correct framework.

"We want to educate the industry that they cannot think in traditional API terminology when they are dealing with excipients," says Moreton. "They simply don't work the same way."

Limiting genotoxic impurities

A major change in impurity testing involves a new regulatory guidance designed to tightly limit substances possessing potential for genotoxicity. The European Agency for the Evaluation of Medicinal Products (EMEA) guidance on genotoxic impurities, which became effective on Jan. 1, 2007, now applies a 1.5 μg daily exposure limit for such substances in most pharmaceuticals based on a precedent application of the threshold of toxicological concern (TTC) concept to food additives and food contact materials. Before the EMEA draft guidance, genotoxic impurities had been addressed only as a footnote in ICH Q3A (R2) "Impurities in New Drug Substances" (see Table I).

In 2004, the Pharmaceutical Research and Manufacturers of America (PhRMA) formed a task force to discuss genotoxic impurity limits. Concerned the 1.5 μg/day limit would be applied to drugs synthesized in the United States, even while these drugs were still in clinical development, the PhRMA group proposed a staged TTC approach that ties permissible impurity levels to the stage of development (see Table II) (4). Because clinical studies are conducted with limited duration of dosing, the group reasoned that total exposure is very low, and thus higher intake levels should be allowable during early clinical studies without a net increase in risk. PhRMA's staged TTC approach applies to all clinical routes and to compounds at all stages of development, for identified and predicted impurities. The TTC limits would not apply to already marketed products.

Table II: Staged TTC approach proposed by PhRMA.

Some questions surrounding both the EMEA and PhRMA positions could be heard, however, at last fall's annual meeting of the American Association of Pharmaceutical Scientists (AAPS).

"There's certainly a need to draw the line somewhere, but in my view it's gone overboard. It would appear that both the regulators and industry have become so focused on positions, that the bigger question of whether the new work processes created are actually benefiting the pharmaceutical consumer is being overlooked," commented one industry scientist (name withheld by request) .

Others point out that the original TTC design purposefully excluded consideration of the impact of the human body's natural DNA repair system, and the fact that the underpinning scientific studies were conducted in rodents, which have DNA repair systems that are about 10-fold less effective than those in humans. "There is an approximate two to three orders of magnitude of extra conservatism built into the original TTC limit that for some reason didn't seem to get much attention in translating the concept to pharmaceutical usage," another industry scientist points out.

Meanwhile, the industry is waiting for the release of an FDA draft guidance on genotoxic impurities, which should clarify the issue.

"Right now (the draft guidance) is going through some internal editing and clearing through various levels of CDER. It is still in draft form and hopefully in the not too distant future we will get it out for comment," said Tim McGovern, supervisory pharmacologist, Division of Pulmonary and Allergy products at FDA.

In preparing to issue its own draft guidance, FDA experts have studied both the EMEA and PhRMA proposal for limits on genotoxic impurities. "We've taken a look at both of those. In various meetings where we've presented on the issue, we've said that 1.5 μg/day appears to be a good threshold limit. We also thought that a staged approach, especially during clinical development, was a reasonable approach to take. So our plan right now is to incorporate a staged TTC approach into the document," says McGovern.

"We've dealt with issues regarding genotoxic impurities for a number of years. It got to a point where the issue was coming up on a regular-enough basis that we thought a guidance document would be helpful," says McGovern. "We would have developed a guidance regardless of whether EMEA had or not."

Scientists continue to challenge both the EMEA and PhRMA TTC proposals and raise concerns that they may add significant burden on the development of virtually all small-molecule projects in the pharmaceutical industry.

"Any compound that has structural alerts needs to be tested by an Ames test," says Modi. Structural alerts are functional groups associated with genotoxic or carcinogenic potential" says Modi. Based on reports, 20–25% of all intermediates can be expected to test Ames positive (see Figure 2) (5, 6). These intermediates are traditionally used in "standard" chemical reactions in the synthesis of new drug substances.

Figure 2: Structural alerts. Substances containing these structures should be tested for genotoxic or carcinogenic potential.

"Considering the fact that only 10% of compounds are successful in reaching the market, this means 90% of the efforts will not produce anything meaningful for patients," says Modi.

IPEC is also concerned that the EMEA limits will eventually apply to excipients. Although the document states the focus is specifically on APIs, there has been some discussion about its future extension into excipients.

"We know many people in the industry have heard comments from European regulators in different venues where they have said that their intention is to apply it to excipients," says Schoneker. "This could be a big problem because the levels (the EMEA) are discussing can be exceeded easily."

IPEC plans to address genotoxic impurities in its composition guideline. It also intends to write a position paper describing its concern about the possibility of having the EMEA limits for genotoxic impurities applied to excipients. In addition to other difficulties, this application could create conflicts in international regulation: "In our discussions with FDA, they have clearly said they have no intention of applying this to excipients," says Schoneker.

"There is a lot of common ground (among) the various approaches. We'll eventually get to a point where we'll at least be in a place where everybody is comfortable, (although) they may not agree to all the details," says McGovern.

References

1. N. Rao and V. Nagaraju, "An Overview of the Recent Trends in Development of HPLC Methods for Determination of Impurities in Drugs," J. Pharm. Biomed. Anal. 33, 335–377 (2003).

2. D. Elder, "To Boldly Go Where No Analyst Has Gone Before? New Technologies in Characterization and Control of Impurites" presented at the USP Annual Scientific Meeting, Denver, CO, Sept. 28, 2006.

3. Q. Wang et al., "A General Overview of Emerging Technologies for Pharmaceutical Impurity Detection and Characterization," presented at the USP Annual Scientific Meeting, Denver, CO, Sept. 28, 2006.

4. Müller et al., "A Rationale for Determining, Testing, and Controlling Specific Impurities in Pharmaceuticals that Possess Potential for Genotoxicity," Reg. Tox. Pharm. 44, 198–211 (2006).

5. R.W. Dugger, J.A. Ragan, and D.H. Brown, "Survey of GMP Bulk Reactions Run in a Research Facility Between 1985 and 2002," Org. Proc. Res. Dev. 9, 253–238 (2005).

6. J.S. Carey et. al, "Analysis of the Reactions Used for the Preparation of Drug Candidate Molecules," Org. Biomol. Chem. 4, 2337–2347 (2006). PT