If you have an ACS member number, please enter it here so we can link this account to your membership. (optional)

ACS values your privacy. By submitting your information, you are gaining access to C&EN and subscribing to our weekly newsletter. We use the information you provide to make your reading experience better, and we will never sell your data to third party members.



Trace Metals Debate

Contentious new guidelines on pharmaceutical impurities will force the drug industry to change testing strategies

by Ann M. Thayer
August 19, 2013 | A version of this story appeared in Volume 91, Issue 33

Credit: Reading Scientific
A technician runs mass spectrometry analysis in Reading Scientific’s metal-testing lab.
Photo shows a RSSL scientist in new metals laboratory using Agilent ICP-MS.
Credit: Reading Scientific
A technician runs mass spectrometry analysis in Reading Scientific’s metal-testing lab.

In 1905, the U.S. Pharmacopeial Convention, the nonprofit standards-setting organization known as USP, introduced a method to check for heavy metals in U.S. pharmaceuticals. More than 100 years later, drug manufacturers still use the same chemistry to determine whether the level of 10 metal impurities is acceptable in their products.

Exposure to metals such as arsenic and mercury can have negative health effects; the substances can also weaken drug stability. But rather than targeting individual metals for their toxicity and setting limits accordingly, the old USP test, described in chapter 231 of the group’s standards, measures all of the metals that can react with a sulfide ion to form a visible precipitate. The aggregate threshold dictates the acceptable limit for all the metals combined.

Regulators and drug manufacturers agree on the need to modernize elemental impurity analysis. However, they disagree on the substance of a new limit-based standard and the timing of its implementation. Tensions have risen in the past three years as deadlines have come and gone. Coming to an agreement on the standard is important because almost every drug sold in the U.S. will have to comply.

After nearly two decades of discussions, postings, appeals, and revisions, two USP chapters replacing 231 became official in form on Feb. 1. Chapter 232 sets limits on 15 discrete elemental impurities in drug substances, excipients, and products. Chapter 233 describes procedures for analytical-instrument-based measuring of impurity levels. USP had intended to implement the new chapters on May 1, 2014, at which time products must comply, but rising dissension this spring—largely around the timeline for compliance—caused the organization to postpone the move.

The deferral gives time to rehash the content and timing for implementation of the chapters, but it doesn’t alter the sense that the changes are inevitable. Despite the uncertainties and unanswered questions, drugs manufacturers don’t want to risk being unprepared. Many companies are putting the needed analytical systems in place.

“I haven’t spoken to anyone who thinks this won’t happen, so they show no inclination of slowing down,” says Chuck Schneider, spectroscopy product planning leader at analytical instrument maker PerkinElmer. “But they are all approaching it differently.”

Instrumentation suppliers predict that drug firms initially will play it safe and test for all listed contaminants and later formulate risk-based compliance plans around actual suspected impurities. “They have to be on the side of caution, because they can’t afford to fail a Food & Drug Administration audit,” says Adrian Holley, director for elemental analysis at Thermo Fisher Scientific. Although USP sets standards, FDA is the agency that enforces compliance.

Credit: USP, ICH, EMA
Table shows the toxicologically proposed limits of three standards groups for daily exposure of elemental impurities.
Credit: USP, ICH, EMA

Heightening concern is that chapters 232 and 233 would apply to more than 1,200 monographs, explains Kahkashan Zaidi, principal scientific liaison at USP. A monograph describes a finished drug or drug ingredient in terms of its identity, packaging, labeling, and the testing and acceptance criteria required to ensure its strength, quality, and purity.

“We took extra measures to develop these new standards because of the high impact,” Zaidi says. “USP has been very flexible and has been listening to the industry and giving all the opportunities that are possible and within our reach to get feedback and take action wherever we can.”

Nonetheless, industry believes USP hasn’t adequately considered the ramifications of its plans. A two-year-old drug industry group, the Coalition for Rational Implementation of the USP Elemental Impurities Requirements, warned in a late-2012 position paper that “currently approved, and safe, drug products may be pushed out of compliance or forced to reformulate, and that drug shortages may result.”

More rigorous analytical tests could potentially reveal that drugs can’t meet the new standards. There’s no way to know on the basis of the old test, Zaidi says. Semiquantitative at best, the 231 method is not specific for individual metals and indicates the overall presence of mixed-metal impurities by comparison to a reference standard.


Plans To Change Testing Procedures For Heavy Metals Are A Decade In The Making

2004 U.S. Pharmacopeial Convention (USP) committee begins to look at updating metals testing including the possibility of using advanced sample preparation and analysis.

2005 USP identifies inductively coupled plasma-optical emission spectroscopy (ICP-OES) and ICP-mass spectrometry as analysis methods.

2010 After considering revisions of the existing USP chapter 231 chemistry test and new testing methods, USP drafts new limits and testing procedures.

May 2011 Final drafts of new USP chapter 232 on elemental impurity limits and 233 on test procedures are published. They are set to become official on Dec. 1, 2012.

November 2012 Industry groups petition USP to alter plans; USP postpones Dec. 1, 2012, release date.

February 2013 USP declares chapters 232 and 233 official and plans to put them into effect by May 1, 2014.

May 2013 Industry response leads USP to consider deferring May 1, 2014, implementation.

June 2013 USP forms an advisory group to consider the implementation plan.

The old method’s inability to identify which metals might exceed the limit isn’t its only flaw. The aggressive sample preparation sometimes required can alter or destroy what’s present, so the method doesn’t actually recover the full amounts of the different metals in a sample. Relying on false or misleading results, “the conclusion is that everything passes when you never know what actually is present,” Zaidi says.

Although USP once considered improving the old method, it concluded that better and more modern approaches to test for elemental impurities are available, Zaidi says. Chapter 233 describes using inductively coupled plasma optical emission spectroscopy (ICP–OES) and ICP-mass spectrometry (ICP-MS), as well as criteria for alternative analytical methods.

There’s no reason to think the drug industry has an impurity problem, observers note. FDA testing in 2012 did not identify “unsafe” products, according to Rx-360, a pharmaceutical supply-chain monitoring group. And FDA itself has reported that preliminary studies from 2007 revealed “nothing alarming.”

But good expectations aren’t enough, and drug manufacturers will need analytical systems to provide the proof. With installation under way, many large pharma companies will be ready when the time comes, instrument suppliers say. For others, including many excipient manufacturers, “this will be a burden, and there will be a steep learning curve,” PerkinElmer’s Schneider says. “Hundreds of small companies have been relying on USP 231.”

Even though big drug firms may be familiar with using ICP-OES or ICP-MS in R&D, many haven’t deployed these techniques in manufacturing and quality control. They must also adjust to using recommended sample preps on a range of dosage forms, or matrices, and may have to hire and train technical staff. The process from ordering an instrument through delivery, set up, method development, and validation can take months, Schneider cautions.

Depending on the number of products to be analyzed, companies may need to install multiple systems at multiple sites. “Large companies are talking about 40 to 50 installations globally,” Thermo Fisher’s Holley says. “China is already gearing up, and in India it’s predicted that there will be 50 to 60 ICP-MS and more than 100 ICP-OES systems.” An ICP-MS system costs about $150,000, and one for ICP-OES runs about $80,000 to $100,000, according to instrument makers.

Instrumentation providers have to be prepared as well. Because available technology is up to the task, no instrument modification was required, Schneider says, “but we had to get method development, operating procedures, standards, consumables, informatics, and validation services in place.”

Schneider has seen the need for testing on this scale only a few times, such as when contamination crises drove the toy and food industries to rapidly ramp up. “The good news for the pharma industry is that they are not responding to a scare,” he says. “The industry can be proactive and have good testing in place in a timely manner.”

Timing was an issue the industry coalition raised in its requests to postpone the May 2014 date. “Drug and component manufacturers do not have sufficient time to complete method development and validation for the many diverse matrices affected, nor for implementation and clinical testing of any consequent reformulations,” the group wrote in a March letter to USP.

It is a “misconception that several years have been allowed for this process,” the coalition argued. Instead, because the USP standards didn’t gel until February 2013, the coalition calculated that only 15 months remained to complete work by May 2014.

It’s unclear how much more time the deferral will provide to companies that need it. For a company to be ready by May 2014, it would have to make a decision to move ahead by this November, Thermo Fisher’s Holley points out. “Not everyone has got it in their heads that they need time to set things up.”

Although some firms are ahead of the game, others don’t want to invest until they have to, Holley adds. An alternative is to turn to an outside provider. Contract testing labs have been buying a lot of the necessary equipment to offer services for elemental impurities, suppliers say.

In June, Reading Scientific Services Ltd. opened a metals lab in England to prepare for USP standards as well as European ones coming into effect in September. Through a partnership with Agilent Technologies, it has ICP-MS and ICP-OES systems, as well as microwave plasma-atomic emission spectroscopy and atomic absorption spectrometry (AAS) systems.

After considering suspected impurities, and types and number of samples, Reading Scientific can advise customers on the choice of method, says Alan Cross, a metals specialist at the lab. Working with a contract lab removes the burden of training staff, qualifying an instrument, and maintaining it, he adds.

And because sample prep is critical and time-consuming, a lab that does the work routinely “gets a good idea how to treat different matrices and get around any problems,” Cross says. “Instrument manufacturers are trying to make things more intuitive, and the software is more powerful, but for us it becomes second nature knowing what works or can go wrong and how to fix things.”

The best available methods can recover and identify all the target metals within a couple of minutes in a single run. Despite a high price tag, most customers are opting for ICP-MS systems, instrumentation managers say. The technique can easily achieve the required detection limits across the widest range of concentrations and elements. It also requires much smaller samples than the 2 g called for by the 231 test.

“If you are a pharmaceutical company testing an active ingredient, you don’t want to waste your sample, and you want to make sure that you can achieve the lowest detection limits,” says Amir Liba, an applications chemist at Agilent. Another advantage of ICP-MS is the ability to identify different elemental species.

Using toxicological assessments, USP has set daily exposure and concentration limits that depend on whether a drug is administered orally, injected, or inhaled. Of the 15 elements on its list, the most toxic four—arsenic, cadmium, lead, and mercury—are environmental contaminants and have the lowest limits. “It is only mandatory to test for the big four,” Liba says.

The remaining ones might arise from catalysts, reagents, machinery, or equipment. “You test for the others if you think there is possible contamination,” Liba explains, “but the final product should be below the limit on all 15 analytes, regardless of whether you test or not.”

USP will accept results from any type of instrument as long as the user validates it for analyzing the elemental impurities. Like ICP-MS, ICP-OES is a rapid multi-element technique, generally sensitive enough for almost all the listed elements. Other techniques—such as graphite furnace AAS, which costs one-third as much as an ICP-MS system to buy and operate—work well but may analyze fewer elements and only one or a few at a time.

Sample preparation, such as extensive dilution, can increase the need for high sensitivity, Liba says. Although simple dilution or dissolution will work for some materials, others will require digestion. Because drug products are complex mixtures of ingredients, fillers, and coatings, suppliers expect that closed-vessel microwave digestion, which can cost up to $50,000 for a full system, is the best way to ensure complete sample breakdown.

Drug manufacturers must comply with the standards, but it is up to them to decide how. “Regulators have left a great degree of freedom in making it risk-based for manufacturers and not prescribed,” says Daniel Kutscher, an ICP-MS application specialist at Thermo Fisher. “But they’ll have to justify why they do their testing their way.”

Instrumentation company managers anticipate that, for peace of mind, most drug manufacturers will keep a tight grip on the situation by testing products themselves. “I expect there will be overtesting at first, whenever they start up a new batch,” Holley says. Eventually, manufacturers will determine the frequency of testing required after understanding the risk that specific impurities will appear.


Moreover, drug manufacturers are likely to ask excipient, raw material, and active ingredient suppliers to provide data on metals. However, not much is known about metal impurities in excipients, and those sourced from minerals and plants have suppliers the most concerned. It’s not known whether suppliers will be able to generate the data for pharma customers, says the International Pharmaceutical Excipients Council, a member of the coalition. But to facilitate information exchange, the group has created a standardized request letter and form templates.

“A manufacturer has to understand its product very, very well before determining what they need to test for, because there is no need to test for everything that is in chapter 232,” USP’s Zaidi says, provided the firm has the knowledge base to support a decision. “We advocate a risk-based strategy in which you evaluate your product, your supplier, your processes, and determine what you need to test and how often you need to test.”

In light of all the work required to comply with the new standards, the drug industry wants to have to do it only once. Right now, USP standards are not aligned with those under development since 2009 by the International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use, a group known as ICH that aims to align international regulatory requirements. Harmonization between ICH’s Q3D guideline and USP’s chapter 232 will “avoid significant problems throughout the industry and redundant non-value-added work,” the industry coalition wrote in its March letter.

ICH completed step 2b of its four-step process to create harmonized guidelines in June, and USP is reviewing the latest ICH version, which was published on Aug. 5. USP revised chapter 232 after previous interim ICH steps, Zaidi points out. ICH is scheduled to reach its final step in June 2014, after which regional implementation would occur in Europe, Japan, and the U.S. In 2008, the European Medicines Agency (EMA) published its own standards that go into effect next month.

The industry coalition says it sees “no scientific or safety-based justification” for USP to unilaterally implement the new chapters while ICH is still developing its guidelines. “Because the time is insufficient for properly preparing for compliance, a more rational approach would be to wait for ICH Q3D to be finalized so that 232 can be appropriately harmonized with this international standard,” the coalition says.

USP says new chapters are needed now because its become understood that currently used methods cannot differentiate between safe and unsafe products. “We have every intention to go along with the ICH whenever possible,” Zaidi says. “But one has to keep in mind that step two is still not the final document from ICH. One has to wait and see where it finally falls and then make the decision how much the alignment can happen.”

ICH guidelines address only the limits on elemental impurities, and not testing methods. USP, EMA, and ICH have different elements on their lists. Even where the lists do overlap, they have set widely different permitted daily exposure limits for different dosage forms and drug materials.

USP says the deferral announced in May will allow it to work closely with ICH to align activities. In early June, USP created a nine-member advisory group that includes people from FDA and from organizations within the industry coalition. USP says the group is expected to work quickly to provide advice on how to “advance a timely implementation” of chapters 232 and 233 while addressing special impacts on manufacturers.


This article has been sent to the following recipient:

Chemistry matters. Join us to get the news you need.