Katherine Ulman is a Global Regulatory Compliance Manager at Dow Corning.

Katherine Ulman

Excipients are substances other than the API, which have been appropriately evaluated for safety and are intentionally included in drug delivery systems. Excipients are typically more complex than APIs, because they are frequently multi-component and less well-defined. In addition, excipient variability can present itself in various forms, compositionally, physically, or functionally.

Most excipients are not pure substances. Other residual and trace components present in excipients can impact key excipient functional characteristics when used in pharmaceutical formulations (1). The International Pharmaceutical Excipients Council (IPEC) refers to these other unavoidable substances in excipients as concomitant components (2). Performance of many excipients in a drug formulation may rely on the presence of such substances in the excipient. These excipient components are not considered impurities and could vary due to changes in raw materials, process variability, and from supplier to supplier. Excipients are not controlled at an individual parameter level. There are many degrees of freedom associated with polydispersity of particle size, molecular weight, and chemical composition, all of which may be process or supplier dependent. Particle size is usually controlled as an average or distribution, and dilute solution viscosities are often used for routine analysis to monitor batch-to-batch consistency of polymer molecular weight. The composition profile of a complex mixture may not be well defined, and the performance impact of the many components may not be well understood.

Silverstein contrasted the control of API quality versus that of excipients (3):

“API quality is improved by reducing the presence of all materials other than the desired chemical. […] Extraneous substances may be harmful to the patient in that they may lead to side effects, or they are inert, thus reducing API purity and thereby compromising efficacy.”

“Excipient quality is described quite differently. While one would again refer to compliance with the compendial monograph (if there is one) or the manufacturer’s specification, a higher assay is not always better. While this may seem counterintuitive, excipients are often complex mixtures that include constituents arising from raw materials, catalyst, solvent, initiator residue, or side reactions” (3).

Excipients are often mixtures of materials with major and minor components, and the minor components may contribute to the overall excipient performance in a particular application (4). For example, the compactibility of coarse grade dibasic calcium phosphate dihydrate is influenced by the presence of “impurities” in the crystal lattice, which cause dislocations and weakness, allowing brittle fracture to occur. Very pure coarse-grade dibasic calcium phosphate dihydrate does not compact as well. However, in general, the exact
relationship between excipient composition and excipient performance is not often well understood and will vary with each application (5).

Many excipients possess some inherent variability. Variability in the finished excipient attributes can arise from many sources such as, but not limited to those in 

Table I. Potential variability due to excipient source.

Excipient variability for some excipients can also be influenced by degradation, which also can be impacted by packaging and/or environmental conditions during storage and distribution.

As described in the IPEC quality-by-design (QbD) guide (6), “excipient manufacturers can control excipient variability only within the limits of their process capability and for which the processing equipment was designed and constructed.” Process capability is a general statistical concept based on the ratio of the specification limits to the process variability. If the variability is the same as the specification range the process capability is one, with no room for error. The higher the process capability the lower the risk of failure. Less than one means that not all batches will meet specification. Therefore, arbitrary narrowing of a specification by users may adversely impact sustainable excipient supply. Robust formulations are not critically impacted by typical excipient variabilities, not all of which may be reflected in the specification.

“In addition, the equipment train used in the process and the type of processing and unit operations included may all influence variability. Various excipients are continuously manufactured in many metric tons per annum. Results reported on a certificate of analysis (CoA) will often be a composite or average, potentially underestimating variability. Not all excipients will have the crystallization or precipitation steps typically used to purify fine chemicals. Users should discuss inherent variability with the supplier and their capability to supply product as it relates to the user’s needs (7). It is the responsibility of the user to identify potential critical material attributes (CMAs) for the excipients used in their formulations but the suppliers may be able to assist. A risk-based approach will probably be appropriate” (6).

Different excipient manufacturers are likely to use different raw material sources as well as different manufacturing processes. Therefore, variability introduced by alternate sources encompasses variability of both raw materials and processing.

Excipients from alternate sources (additional sites within a manufacturer and/or alternate supplier(s)) may not be compositionally, physically, or functionally identical, but may still demonstrate equivalent performance within the drug product depending on application. Excipient users should define functional performance requirements for alternate excipient sources (8).

When evaluating alternate excipient sources, users should not assume that all sources conforming with the same compendial requirements are interchangeable in all drug product formulations. Variability in excipient composition and other attributes among excipient sources could impact performance. To gain an understanding of the potential origin for various components that may be found in excipients from different sources, see 

Figure 1. The complexity in excipient composition. Figure courtesy of IPEC Americas.

Confirmation of equivalent performance is necessary to support interchangeability. The critical quality attributes (CQAs), including stability, of the drug product manufactured with the alternate source must be demonstrated.

Distributors should not substitute an alternate excipient source without prior notification to the excipient user to allow evaluation of the interchangeability of the alternate source in their specific drug applications.

In a 2008 article on “Comparative Tableting Properties of Sixteen Microcrystalline Celluloses” (9), Doelker 

noted that, “Great differences in packing and tableting properties and in sensitivity to the addition of a lubricant were generally observed between
products from the various manufacturers. In contrast, lot-to-lot variability was quite acceptable.”

Mitigating the impact of excipient variability

Moreton defined a robust formulation as being “able to accommodate the typical variability seen in API, excipients, and processes without compromising the manufacture, stability, performance, or any other attribute of the product critical to the patient’s care or well-being” (5).

Qualifying multiple sources should be considered when establishing the design space to reduce the time required to switch to an alternative supply after launch (10). If the impact of an excipient variability on finished product CQAs is unacceptable, then additional controls (CMAs) should be considered. Both limits and methods for CMAs must be agreed between user and excipient manufacturer.

It may be necessary for excipient manufacturers to make changes to their processes for various reasons that have the potential to impact excipient performance and thus finished product performance. Excipient manufacturers should consult the IPEC Significant Change Guide for Pharmaceutical Excipients (11) whenever a change is to be made to excipient raw materials, manufacturing process, and/or finished excipient testing and assess the need for customer notification. Based on a risk assessment, the excipient manufacturer should apply the same reasoning (as far as is practicable) to their own suppliers of starting materials, reagent, solvents, processing aids, and additives used in the manufacture of the excipient.

When investigating potential impacts of notified changes, it is strongly recommended that a risk assessment be undertaken by the excipient user. Based on the results of the risk assessment, the excipient user should determine their need for further action.

It may be necessary for the excipient user to build product inventory, thus requiring pre-change excipient supply, because regulatory filings may be subject to extended review times. This may, in turn, require coordination between excipient manufacturer and user.

It should also be remembered that not all changes to an excipient will be detrimental to the performance of the excipient or pharmaceutical finished product. All changes should be considered as significant unless it can be scientifically justified as not a significant change. Significant changes should be notified to the excipient user (11).

Factoring excipient variability into design of experiments, control strategy, and lifecycle management will increase finished product robustness (12),
which will improve the overall product quality and performance, leading to better patient outcomes.

Excipient variability should always be a concern unless proven otherwise. Relying only on the excipient sales specification or the compendial monograph is not enough to account for potential variation(s) because many CMAs may not be part of these requirements. Drug product manufacturers should have a robust dialogue with their excipient suppliers to understand the variability of material attributes.

PharmTech APIs, Excipients, and Manufacturing 

Composition Guide for Pharmaceutical Excipients

I. Silverstein, “Excipient Quality and Selection,” PharmTech Europe, 28 (2) 2016.

Incorporation of Pharmaceutical Excipient into Product Development using Quality-by-Design (QbD) Guide,

The IPEC-Americas Quality by Design (QbD) Sampling Guide, 

E. Doelker, D. Mordier, H. Iten & P. Humbert-Droz (1987) Comparative Tableting Properties of Sixteen Microcrystalline Celluloses, 

, 13:9-11, 1847-1875, DOI: 10.3109/03639048709068696

Qualification of Excipients for Use in Pharmaceuticals

Signifiant Change Guide for Pharmaceutical Excipients

CPhI Pharma 2021 Report: Post Pandemic Legacy Trends

Brian Carlin is owner Carlin Pharma Consulting LLC. George Collins is the Vice-President for Excipient Mineral Manufacturing, RT Vanderbilt Company. Chris Moreton is partner at FinnBrit Consulting. David R. Schoneker is the President/Owner of Black Diamond Regulatory Consulting, LLC. Jennifer Putnam is a Senior Supervisor, Analytical R&D at Perrigo.Joseph Zeleznik is Technical Director, North America for IMCD US Pharma. Katherine Ulman is the owner and Primary for KLU Consulting, LLC. Stacey Bremer, Director, Product Stewardship Medical at Celanese.


Vol. 46, No. 8
August 2022
Pages: 20–23

When referring to this article, please cite it as B. Carlin et al., “Mitigating the Risk from Excipient Variability,” 

Articles

This article reviews sources of excipient variability, including raw materials and processing, both of which may vary from supplier to supplier and from plant to plant for a single manufacturer.

Mitigating the Risk from Excipient Variability

	
	Peer reviewed:
	Submitted: Feb. 11, 2015. Accepted: April 7, 2015.

Applications for a drug product that is administered dermally must define acceptable limits for elemental impurities in the product. Current guidelines for exposure limits, however, are based on oral, parenteral, or inhalation exposure and do not adequately address other routes, such as dermal exposure. This article reviews the information available about dermal exposure to metals and considers how it differs from oral or parenteral exposure.

	The International Conference on Harmonization’s (ICH’s) 

 establishes acceptable exposure limits and defines these limits based on route of administration, but does not specifically address dermal limits (1). Thus, developers of dermally delivered drugs are challenged to establish and define those limits. Is it appropriate, for example, to transpose limits based on other routes of administration (i.e., to apply oral or even parenteral limits)? This article seeks to examine the available data relating to dermal exposure to metals and evaluate appropriate approaches to ensure that the limits established will adequately address actual risk.

	The skin is made up of several layers including 

, viable epidermis, and dermis (2). The largest organ in the body, the skin provides a formidable barrier to exogenous materials. Within the skin, the outer layer (the 

) provides the most significant barrier and is rate limiting for any substance applied to the skin, whether intentional, in the form of a deliberately applied ointment, or a contaminant to which the skin is inadvertently exposed (e.g., a metal impurity present in a dermal product). The 

is highly lipophilic and contains very little water. As a result, penetration of hydrophilic or charged molecules is particularly difficult, and such species are effectively unable to partition into the lipid layer and hence pass through the skin at significant levels.


	The skin provides a formidable barrier to any drug. Even when a formulation has been deliberately designed to facilitate transport across the dermal barrier, exposure levels are typically a fraction of those obtained when the drug is delivered by injection. Consider lidocane, for example. Using topical formulations containing lidocaine, the systemic levels that are reached are as follows (3):

		5 µg/mL, after the application of 2 g ointment to the endotracheal tube

		1.3 µg/mL, after the application of 10 g ointment to the oral mucosa for 30 minutes

		0.6 µg/mL, after an oral (ingested) dose of 10 g ointment (bioavailability 35%)

		2.5 µg/mL, after a rectal dose of 10 g ointment (bioavailability 60-70%)

		0.3 µg/mL, after the application of 10 g ointment to a partial thickness burn

		0.1 µg/mL after the application of 10 g ointment to intact skin.

	Systemic levels of lidocaine of 0.1 µg/mL were observed following the application of 10 g dermally (to intact skin). For the same dose administered orally, the level is 6 times higher.  Even when applied to burnt skin, the level (0.3 µg/mL) is still 50% lower than the level achieved through oral administration. When compared to endotracheal application, where there is no effective barrier, the level adjusted for dose is significantly below 5% of the systemic concentration level. These data clearly demonstrate the significant barrier the skin provides and indicate significant concerns for the simplistic approach of applying oral-based limits for elemental impurities.

	Similar data have been generated for polydimethylsiloxane and linear and cyclic siloxanes.

 Dimethicone, a polysiloxane polymer, is widely used in a range of cosmetics. 

human dermal absorption/penetration studies in which neat material was applied to abdominal skin showed that 0.2% and 0.1% of the applied dose of PDMS (10 cST fluid) or PDMS (350 cSt fluid), respectively, remained at the dosing site and could be considered absorbed. The percentage of applied dose that penetrated through the skin to the receptor fluid was ~0.0002% for the 10 cSt fluid and ~0.0003% for the 350 cSt fluid. Studies of its dermal absorption showed no evidence that it had penetrated the dermal barrier (4).

	Oral-dosage absorption is also challenged. In studies where an oral dose of 350-cSt 14C-PDMS was administered to rats at 1000 mg/kg body weight, more than 99% of the recovered dose was excreted by the animals (4). PDMS by the dermal or oral route is minimally absorbed.

 human dermal absorption data for cyclic siloxanes (octamethylcyclotetrasiloxane, D4; decamethylcyclopentasiloxane, D5; and dodecamethylcyclohexasiloxane, D6) again showed extremely low levels of absorption following dermal application (5,6). As the molecular weight and lipophilicity of the cyclic siloxane increased (from D4 through D5 and D6), the percutaneous absorption decreased. The percentages of applied dose absorbed following a neat application was 0.5%, 0.05%, and 0% for D4, D5, and D6, respectively. Penetration into the receptor fluid was 0.01%, 0.002%, and 0.003% for D4, D5, and D6, respectively.

	Absorption data are available in rats following a single oral administration to rats of D4 (300 mg/kg), D5 (1000 mg/kg), and D6 (1000 mg/kg). These mass balance studies indicate that 52%, 20%, and 12% of the orally administered dose was absorbed (7-9). In the case of cyclic siloxanes, a greater percentage of the administered dose was absorbed orally than dermally.


	There are several comprehensive reviews covering exposure to metals. These reviews are predominantly focused on environmental exposure from sources such as soils. In 1993, Hostynek et al. (10) sought to summarize data relevant to the qualitative and, where possible, quantitative evaluation of metal permeation through skin. They looked at assessments for 31 elements, although coverage of ICH Q3D Class 1 metals (mercury, arsenic, cadmium, and lead) is incomplete, as there is no assessment of mercury or lead.

	They concluded that dermal absorption of metals is a complex process affected by multiple factors including size, charge, and oxidation state. They did not, however, draw any definitive overall conclusions regarding generic estimates of absorption, nor does it seek to generically compare rates to other routes of administration.

	A review by the National Environmental Policy Institute, 

Assessing the Bioavailability of Metals in Soil for Use in Human Health Risk Assessments

 (11), sought to specifically examine bioavailability. This review provides a thorough assessment of the available toxicological data for soil samples containing six metals: arsenic, cadmium, chromium, lead, mercury, and nickel. This review focuses on evaluations performed using soluble salt forms of the metals and assesses main routes of administration, including dermal limits. The overriding conclusions are that there are little data available to allow for the specific calculation of dermal exposure; however, the data that are presented support the view that even when present in an aqueous soluble form, metals are poorly absorbed through the skin.

	The most comprehensive and most recent review of dermal exposure to metals is the HERAG 1 factsheet, 

Health Risk Assessment Guidance for Metals--Assessment of Occupational Dermal Exposure and Dermal Absorption for Metals and Inorganic Metal Compounds 

(12). This review, sponsored by the International Council on Mining and Metals and Eurometaux, critically examines existing models used to estimate levels of dermal exposure, and determines their value in assessing the absorption of inorganic metals. One of the most prevalent models used is the EASE (Estimation and Assessment of Substance Exposure) model, developed by the United Kingdom’s Health and Safety Executive. This model defines default dermal absorption rates of 100% or 10% depending on the properties of the substance in question.

	Without relevant experimental data, 10% dermal absorption is used when the molecular weight (MW) of the substance is > 500 and the log P

 [partition coefficient for the compound in question between octanol (o) and water (w)] is smaller than -1 or higher than 4, otherwise 100% dermal absorption is used. The major issue with such an approach is that it was developed for organic chemical compounds; this is not considered relevant for metals, for the following reasons:

 is a parameter that has no bearing whatsoever in the prediction of the properties of a metal or of an inorganic salt of a metal. Inorganic metal species do not permeate the skin by passive diffusion. Instead, the uptake of metals largely depends on the presence of specific transport systems that provide biological gateways for the metal to cross the membrane.

		The dissolution of an inorganic metal compound or the metal itself on the skin surface will intrinsically require dissociation and ultimately liberation of free metal cations. Thus, molecular weight, the second criterion for assigning a dermal absorption rate, is irrelevant for metals. Under no circumstances could any metal cation exceed the cut-off value of 500.

	The review repeatedly makes the point that such general approaches are not only scientifically flawed, but also grossly overestimate the actual levels of exposure.

	Crucially, the review cites recent studies that fundamentally challenge the EASE model, with data derived from studies performed on soluble and insoluble zinc compounds. These data show that penetration of the dermis by soluble zinc sulfate is low and that dermal penetration of insoluble zinc oxide is even lower. The review concludes that dissolution kinetics are the rate-limiting factor. The results from these and other studies are summarized in 

	The article concluded, based on these experimental data, that recent studies conducted to Organization for Economic Cooperation and Development (OECD) standards indicate dermal absorption rates to be at or below 0.3% (12). There is no clear correlation between absorption and such factors as speciation, valency, and/or water solubility. Crucially, the HERAG guidance document concluded that it should be feasible to establish default absorption factors, concluding that a default absorption rate of 1% was reasonable and adequately conservative when the skin is exposed to metals in wet or liquid media. For “dry” exposure to dusts, it proposes a default dermal absorption factor of 0.1% (12).


	The following summarizes literature references with specific data relating to ICH Class 1 metals.

 A white paper specifically focused on arsenic was developed by the New Jersey Department of Environmental Protection (13). The paper provides a useful summary of several studies pertaining to the dermal absorption of arsenic. The data presented show that significant levels of arsenic were found to penetrate the skin of mice; however, this contrasted markedly with the results obtained for human skin, for which levels were within the range of 2-6% of the dose applied. The differences were attributed to the significant intra-species variation in skin thickness, that of a human being substantively greater.

	The National Research Council (1999) (14) evaluated the available information on this subject and stated that “these results indicate a low degree of systemic absorption of arsenic via the skin.” The Agency for Toxic Substances and Disease Registry (15) concluded that “it is usually considered that dermal uptake of Arsenates and Arsenites is sufficiently low that this route is unlikely to be of health concern …, but studies to test the validity of this assumption would be valuable.” Arsenic does not act as a sensitizer upon casual skin contact due to poor skin-penetrating ability of its naturally occurring compounds.

	Further studies were reported by Wester et al. (16). These studies reported a low permeability coefficient of 2.71 x 10

 following topical application of water containing radio-active arsenic.

 Absorption of cadmium through the skin is reported to be extremely low (0.5%) (17). Specific 

 experiments reported by Hostynek et al. (10) were conducted using human skin; these involved exposure for 16 hours to 116 ppb cadmium chloride applied as 2.5 and 5 ul/cm

 solution. Only 0.1 to 0.6% was found in the receptor solution.

	The HERAG factsheet (12) cites the relative absorption rate for cadmium metal and cadmium oxide, concluding that percutaneous absorption is likely to be significantly less than 1%.

The US Occupational Safety and Health Organization (OSHA) (18) concluded that lead can be absorbed by inhalation (breathing) and ingestion (eating) lead (except for certain organic lead compounds not covered by the standard, such as tetraethyl lead). It was also concluded, however, that lead is not readily absorbed through skin, concluding that cutaneous absorption of lead is limited. The amount absorbed through the skin depends on the physical characteristics of the lead (i.e., organic vs. inorganic) and the integrity of the skin.

	The HERAG paper (12) presented data from a variety of studies concluding that dermal absorption rates greater than 0.01% were unfeasible.

 The Hostynek paper (10) concludes that mercury can penetrate the skin in all forms including its elemental form, outlining both intra and intercellular pathways. Data relating to studies performed using guinea-pig skin and limited data relating to human skin are described. The data are complex, showing an apparent non-linear dose response, which was thought to be related to significant skin retention caused by reaction with skin proteins. There are no specific conclusions as to the actual level (%) absorbed.


	The following summarizes literature references with specific data relating to ICH Class 2 metals.

 No specific studies were located regarding absorption in humans or animals after dermal exposure to vanadium, although absorption by this route is generally considered to be very low, according to the World Health Organization (19). Absorption through the skin is thought to be quite minimal due to its low lipid/water solubility.

 The dermal absorption of molybdenum (as the molybdate ion) is low to negligible at approximately 0.2% of the applied dose. These results were shown in an in-vitro percutaneous study by Roper of the absorption of sodium molybdate dehydrate, a highly soluble substance, in human skin (20).

 While the HERAG report (12) raises concerns over the methodology employed in the studies described, it reports that even in the highest example (one where data for level absorbed was combined with the level retained in the 

) the absorption of a soluble salt form was reported to be only 2%.

 There is some evidence of absorption of cobalt metal through skin. Scansetti et al. (21) found that an experimental study with four volunteers, whose skin was exposed to freshly mixed or waste powder containing 5-15% cobalt, measured a ten-fold increase of cobalt in urine. Results confirmed that absorption through skin is a potential route of entry. Later studies by Filon et al. (22) concluded that cobalt can pass through the skin only when it is oxidized to ionic cobalt (Co

 The US Environmental Protection Agency (EPA) has not developed actual toxicity values for metals based on dermal administration. Instead, the EPA has devised a process for making route-to-route (oral-to-dermal) extrapolations for systemic effects. The EPA developed 

Risk Assessment Guidance for Superfund: Volume I Human Health Evaluation Manual (Part E, Supplemental Guidance for Dermal Risk Assessment)

 (23) to address human health risk related to dermal exposures. Part E uses a consistent methodology for assessing the exposures from the dermal pathway for Superfund human health risk assessments.

	Primarily, Part E accounts for the fact that most oral reference doses (RfDs) and slope factors are expressed as the amount of substance administered per unit time and body weight, whereas exposure estimates for the dermal pathway are expressed as absorbed dose. The process uses the dose-response relationship obtained from oral administration studies and makes an adjustment for absorption efficiency to represent the toxicity factor in terms of absorbed dose. Crucially, this does not advocate direct correlation between oral and dermal exposures, which would be crude and simplistic, nor does it advocate the simple direct application of oral limits to dermal application.

European Union (EU) risk assessments, occupational dermal exposure.

 EU assessments have historically been based on the EASE model, which proposes a default absorption of 10%. This is now largely discredited (see HERAG review [12]).

Michigan Department of Environmental Quality (MDEQ)

. The review by the National Environmental Policy Institute (NEPI) (11) reports that MDEQ has had a 1% value in place as the default dermal absorption value for all inorganic compounds since 1995.

West Virginia Department of Environmental Protection (WVDEP).

 In addition, the NEPI report (11) indicates that WVDEP relies heavily on EPA in providing guidance for considering bioavailability in assessing dermal exposures. Default assumptions provided for dermal absorption of metals from soil are 3% for arsenic and 1% for cadmium and other metals.


	Although fragmentary in nature, existing data do nevertheless show that dermal exposure to metals is limited, often well below 10% of that observed when the same materials are administered orally. This conclusion is entirely consistent with the fundamental nature of the dermis and the obvious barrier it presents to charged materials in particular, such as metal ions. Reflecting on the variable approaches taken by regulatory bodies, it is clear than even the most conservative and largely discredited (in the context of applicability to charged metal species) EASE model differs substantially from assumed parity with parenteral or even oral exposure. Certainly the assumptions prescribed by MDEQ and WVDEP of exposure levels equivalent to <5% absorption correlate well with data available, even though it is limited.

	With respect to ICH Q3D and the establishment of dermal exposure limits, the available evidence and data challenge any attempt to correlate dermal limits directly with oral data.  The available evidence and data strongly support the need for a case-by-case approach that takes into consideration such factors as metal form and relevant specific data to establish scientifically justifiable limits. Ultimately, the overriding evidence to date supports the view that dermal exposure to low-level elemental impurities is unlikely to represent a substantive toxicological concern.

		R. Mehta, “Topical and Transdermal Drug Delivery: What a Pharmacist Needs to Know,” 

		Astra Zeneca internal report, “Clinical Documentation on Xylocaine Ointment 5% (50mg/mL),” 2003.	

Technical Report No. 55. Linear Polydimethylsiloxanes 

		Hostynek et al., Critical Reviews in Toxicology  23 (2) 171-235 (1993).	

		National Environmental Policy Institute, 

Assessing the Bioavailability of Metals in Soil for Use in Human Health Risk Assessments. Bioavailability Policy Project Phase II Metals Task Force Report

Health Risk Assessment Guidance For Metals--Assessment Of Occupational Dermal Exposure And Dermal Absorption For Metals And Inorganic Metal Compound

Environmental Assessment and Risk Analysis Element

 (National Academy Press, Washington, 1999).	

		US Dept. of Health and Human Services, Agency for Toxic Substances and Disease Registry, “

Toxicological Profile for Arsenic (Aug. 2007)

Vanadium. Environmental Health Criteria 8

		C. Roper, “Disodium Molybdate Dihydrate: The In Vitro Percutaneous Absorption Of Molybdenum Through Human Skin,” Unpublished study report by International Molybdenum Association (2009).	

Risk Assessment Guidance for Superfund (RAGS), Volume I: Human Health Evaluation Manual (Part E, Supplemental Guidance for Dermal Risk Assessment)


	Andrew Teasdale, PhD* is principal scientist, chair of AstraZeneca‘s Impurities Advisory Group, Hurdsfield Road, Macclesfield, Cheshire, England, SK10 2NX, andrew.teasdale@astrazeneca.com; Katherine Ulman is global regulatory compliance healthcare manager, and Jean Domoradzki is an EHS specialist, both at Dow Corning Corporation, Auburn, MI; and Phyllis Walsh is associate director, compendial affairs at Merck, Rahway, NJ.

	*To whom all correspondence should be addressed.

 When referring to this article, please cite it as A. Teasdale et al., “Establishing Limits for Dermal Absorption of Elemental Impurities,” 

Current guidance for absorption of elemental impurities does not address dermal exposure, resulting in a simplistic approach to limit setting.

Establishing Limits for Dermal Absorption of Elemental Impurities

New guidelines relating to elemental impurities from the International Conference on Harmonization (ICH), Q3D 

 (1) have presented the pharmaceutical industry with new challenges. These challenges include the complexity of introducing new analytical technology-specifically inductively coupled plasma (ICP)-based techniques replacing the wet chemical “heavy metals” limit test-along with new and specific limits for individual elements. Perhaps the most significant challenges, however, are related to the practical implementation of the guideline. 

	ICH Q3D advocates the use of a risk-based approach to assessing the potential presence of elemental impurities in drug products. While such assessments are common within other aspects of pharmaceutical development, application to elemental impurity assessment presents new challenges. Specific challenges include determining how to assess or quantify the risks associated with factors such as water, container-closure systems, and excipients. Defining where in the assessment process data may be required and identifying where risks can be determined to be negligible through a thorough scientific theoretical risk assessment also present significant questions. This article seeks to review these questions by looking at the various risk factors and, where possible, weighting the risk factors based on appropriate and relevant considerations to establish an effective framework for the systematic assessment of risk and final control strategy. 


	 The introduction of ICH Q3D (1) is one of the most complex changes in regulations pertaining to impurities seen by the pharmaceutical industry. While the guideline is ultimately intended to focus on final drug product quality, the actual risk assessment will touch all facets of the manufacture of a drug product. The guideline introduces toxicologically relevant permitted daily exposure (PDE) limits to individual elements replacing non-specific 19th century wet chemical “heavy metal” limit tests. ICH Q3D advocates the use of a risk-based approach to assessing the potential for presence of elemental impurities in drug products. The process of executing and documenting the risk assessment is a major challenge, primarily as a result of a limited global understanding about how to assess or quantify the risk associated with factors such as water, container-closure systems, and excipients.

	Defining where in the assessment process data may be required and identifying where risks can be determined to be negligible through a thorough scientific theoretical risk assessment also present a significant challenge. Where the risk assessment identifies the need for testing, the level of the PDEs for the element(s) of concern may also require the broader introduction of new, more sensitive, and specific analytical technology, adding still further to the complexity.

	This article specifically seeks to examine relevant risk factors and, where possible, the weighting of these risks based on appropriate and relevant considerations. It also seeks to specifically define where in the assessment process data may be required as well as seeking to identify where risks can be determined to be negligible simply through a thorough scientific theoretical risk assessment.

	The general principles outlined in this article are believed to address most scenarios or product types; however, ultimately any drug-product manufacturer needs to consider potential sources of elemental impurities appropriate for their specific product. 


	The evaluation of the potential risk posed by elemental impurities within a formulated drug product requires a holistic approach taking into account all potential sources of elemental impurities. 

 illustrates potential sources that should be considered in such an evaluation. 

: Sources of elemental impurities in finished drug products.

, the drug substance is a key component that can contribute elemental impurities to the finished drug product. The risk of inclusion of elemental impurities from a drug substance, therefore, needs to be considered when conducting a drug product risk assessment. Control of the elemental impurity content of a drug substance can be assured through a thorough understanding of the manufacturing process including equipment selection, equipment qualification, GMP processes, packaging components, and the selection and application of appropriate control strategies.

	A principal responsibility for any drug-substance manufacturer is to develop a strategy to ensure effective control of the levels of elemental impurities in the finished drug substance. An approach based on assessing and controlling potential sources of elemental impurities, coupled with focused, limited testing, is preferable to exhaustive testing on the finished drug substance. A scientific, risk-based approach combined with knowledge and control of the key sources of elemental impurities in the drug-substance manufacturing process such as catalysts, provides an efficient and comprehensive elemental impurity control strategy for finished drug substances. 

 shows potential sources of elemental impurities in the drug substance manufacturing process. Of the sources highlighted, the greatest risk comes from intentionally added metals (e.g., metal catalysts used in the process). Manufacturing equipment, processing aids, inorganic reagents, water, solvents, and other organic materials are less likely to serve as major contributors of elemental impurities in the finished drug substance, but do require consideration.

: Primary sources of elemental impurities in drug substances (DS).

 Metal catalysts such as palladium and platinum are often used in the drug-substance manufacturing process and can therefore be present at low levels in the finished drug substance. The synthetic route should be reviewed for intentionally added metals, and data from purging studies, including any supportive testing of appropriate isolated intermediates, should be used in the design of an appropriate control strategy.
	 The ability to remove the catalyst (purge capacity) will be influenced by catalyst loading and the nature of the catalyst used in the process (i.e., homogeneous vs. heterogeneous catalysts). Heterogeneous catalysts, such as palladium on carbon, are often easily removed from reaction mixtures by filtration, and therefore, the risk of carryover of elemental impurities into the drug substance is typically low. Even in cases where metal catalysts are used in the final stages of the process, good historical data and/or understanding of carry-over may permit reduced testing schemes. 

	Biotech products do not normally rely on the use of catalysts. As ICH Q3D points out, typical purification schemes in biotech drug substance manufacturing are well capable of clearing any elements introduced either intentionally or inadvertently “to negligible levels.” The principles outlined previously may nonetheless be relevant in some specific cases (e.g., chemically modified biotech drug substances).

	When considering the other potential sources highlighted in 

, it is recommended to focus primarily on the manufacturing steps that occur after the formation of the final intermediate. Washes, crystallizations, phase separations, chromatography, distillations, and processing aids/scavengers aid in purging of elemental impurities and, therefore, reduce the risk of carryover into the finished drug substance from stages earlier in the upstream process. Areas for further consideration include manufacturing equipment, processing aids/inorganic reagents, solvents, water, and packaging. 

 In general, GMPs, including equipment compatibility assessment and qualification, are sufficient to ensure that significant levels of elemental impurities are not leached from manufacturing equipment into the drug substance. Hastelloy, stainless steel, and glass are the most commonly used materials of construction for drug substance manufacturing equipment, due to their superior chemical resistance. Nickel, cobalt, vanadium, molybdenum, chromium, and copper are key elements in some Hastelloy and stainless-steel alloys. Under extreme/corrosive reaction conditions, such as high temperature and low/high pH, these elements could have the potential to leach from manufacturing equipment. In such cases, it may be necessary to supplement standard GMP equipment compatibility assessments with specific studies to assess the elemental impurity-leaching propensity from manufacturing equipment due to corrosive reaction conditions. 

	Other potential sources include high-energy processes such as milling/micronization equipment. These are also generally considered to be low risk, but should be addressed via appropriate GMP including cleaning records and visual inspection. Particle size reduction is discussed in the Drug Product Manufacture section.

 Processing aids such as charcoal, silica, celite, and darco, and inorganic reagents such as sodium chloride, magnesium sulfate, and sodium sulfate, are often used in drug-substance manufacturing processes and may be used in significant quantities. Depending on their specific composition, inorganic reagents should be considered within the risk assessment, especially when ICH Q3D elements are integral to the formula. The levels used should also be considered as some reagents may be employed in higher relative amounts (i.e., in stoichometric amounts). In such instances not only the elements (intentionally) present in the reagents, but also reagent purity (or lack thereof) need to be taken into account. 

	In general, the use of such reagents presents a low risk. In studies conducted within the organizations associated with this article, there is little evidence to support such materials being a significant risk. Therefore, the risk assessment should primarily focus on processing aids and inorganic reagents used late in the drug substance manufacturing process, and/or where aggressive reaction conditions exist (e.g., extreme pH/high temperatures for prolonged times).

 Most solvents used in the manufacture of drug substances, particularly those listed in ICH Q3C, Impurities: Guideline for Residual Solvents (2) Class 3, are unlikely to contribute elemental impurities to the finished drug substance. The majority of solvents are purified by distillation and few involve the direct use of metal catalysts in their manufacture; hence they are considered a low risk source of elemental impurities. In the event that solvents have not been purified by distillation, especially if a catalyst in used in their manufacture, further evaluation in the risk assessment should be considered.

 Refer to the environmental factors discussion is the Drug Product Manufacture section.

 Packaging is discussed in the section Container-closure Systems (CSS) as a Potential Source of Elemental Impurities in Finished Drug Product. 

 It must be recognized that, from a compliance perspective, the limits for elemental impurities in ICH Q3D apply only to the drug product. To ensure effective control of the level of elemental impurities in the drug substance a number of options are available:

		The ICH Q3D option 1 concentration limits assume a maximum daily drug product intake of 10 g/day. Drug substances that meet option 1 concentration limits can be used at any dose in the drug product.

		ICH Q3D option 2 concentration limits are calculated specifically based on the actual daily drug product intake (and composition) and may provide higher concentration limits than option 1 (if the maximum daily intake of drug product is <10g).

	The acceptable level of elemental impurities in a drug substance may be defined and agreed upon in a suitable quality agreement between the drug substance manufacturer and the drug product manufacturer.


	While drug substance manufacturing often involves a complex series of processes, some simple scientific principles can be applied to ensure that elemental impurity levels in the final drug substance are controlled to appropriate levels. The application of a risk-based control strategy, involving an understanding of the manufacturing process and key sources of elemental impurities, appropriate equipment selection/qualification, adoption of suitable GMP processes/procedures, and the selection and application of appropriate control options will typically result in the manufacture of drug substances with elemental impurity levels well below ICH Q3D option 1/option 2 concentration limits. This overall low risk status is supported by the emerging dataset from ICP-optical emission spectrometry (OES) and ICP-mass spectrometry (MS) screening of a wide range of drug substances plus the significant body of historical heavy-metals test data. 


	One of the greatest challenges to performing an elemental impurity risk assessment for a drug product is to understand the potential contribution of elemental impurities from excipients. 

	Elemental impurities of concern for excipients would typically be:

		Class 1 and Class 2a elements potentially present at trace levels in the excipient based on environmental factors

		Intentionally added catalysts or reagents for synthetic excipients

		Class 3 elements from excipients that are targeted for a specific route of administration (e.g., inhaled).

	Unlike drug substances, there may be less information available from excipient vendors with respect to the manufacturing equipment and processes used (e.g., any high energy steps, corrosive reagents, or added catalysts), which may potentially introduce elemental impurities. While many vendors will supply, on request, a compliance statement with the European Medicines Agency (EMA) metal catalyst guideline (3), the elemental impurity testing reported on the certificate of analysis is typically based on a non-specific, wet chemistry heavy-metals limit test with the result reported as less than the specification value in parts per million. When considering the risk of elemental impurities potentially being introduced into the drug product via excipients at levels greater than the PDE, the following points should be considered:

		Source of the excipient (mined, plant, animal, synthetic, etc.)

		Excipient level in the formulation (wt. %) (4)

 The origin of an excipient can have a significant impact on the degree of risk associated with elemental impurities. 

:  Potential sources of elemental impurities in excipients.

	Mined excipients may exhibit a natural variation depending on the location of the mine and the natural geology. The potential levels of elemental impurities from mined excipients may be more variable and therefore pose a higher risk than synthetic excipients manufactured using metal reagents and/or catalyst where the levels are less variable as a result of well-defined manufacturing processes and controls. 

	Excipients harvested from plants also pose a potential risk as a result of their uptake of metals from their environment. Synthetic excipients that are manufactured without the use of metal catalysts and/or reagents present the lowest risk of introducing elemental impurities to the drug product. 

 An essential consideration in determining the risk contribution for elemental impurities from an excipient is the proportion of the excipient used in the formulation. The risk contribution from a mined excipient used as a filler or diluent (typically >20 wt% of the blend), for example, will be greater than the risk contribution of a mined excipient used in a tablet film coat (typically <5 wt%).  

	Generally, low-weight percentage components that are not mined are regarded as low risk; however, the daily product dosing and route of administration also need to be considered. 

 Dose and dosing regimens should also be considered. A low, orally administered, daily dose clearly presents a lower risk than an inhaled product as shown by the lower PDE limits for inhalation vs. oral found in Table A.2.1 of the ICH Q3D guideline (1). 


	The above principles can be used as part of an excipient risk assessment and are a useful guide in the absence of data. In addition, manufacturers should consider the principles for controlling elemental impurity generation from manufacturing equipment, as described in the Manufacturing equipment section for Drug substances.

 (USP) <231> from the United States Pharmacopeial Convention, however, uncertainty exists as to whether excipient manufacturers and vendors will test for elemental impurity concentrations and, if so, for what elements, to what levels, and using what procedures/equipment.

	It must be recognized that, from a compliance perspective, the limits for elemental impurities in ICH Q3D only apply to the drug product. Excipient manufacturers may, however, be requested to assist in assuring compliance through awareness of the level of elemental impurities within the excipient itself. In addition, the same general principles of evaluation options for drug substances described previously apply to excipients.

	In some instances, it may be appropriate to define and agree on the acceptable level of elemental impurities in an excipient through a suitable quality agreement between the excipient manufacturer and the drug product manufacturer.

	Ultimately, it is the responsibility of pharmaceutical manufacturers to demonstrate, via risk assessment and/or data, that the drug product is compliant with ICH Q3D. To this end, the development of a common database for excipient elemental impurity profiles will be a useful activity to support risk assessments. The International Pharmaceutical Excipients Council of the Americas (IPEC-Americas) and FDA have performed an exercise examining elemental impurity levels from multiple excipients, which is planned for publication. This information indicates that the vast majority of excipients (analyzed by ICP-MS) contain elemental impurities at levels unlikely to cause concerns at typical usage levels in oral solid dosage forms. 


	The drug-product risk assessment needs to consider the potential elemental impurity contribution from the drug product manufacturing equipment/process. Manufacturing for solid products encompasses a large variety of processes, such as solid mixing (blending), granulation, tableting (compression), coating, and particle size reduction. For liquid product manufacture, dissolution or suspension of solid excipients and drug substance is often carried out in metallic equipment. In contemporary cGMP facilities, the likelihood of additional contribution of elemental impurities is negligible. However, there is a theoretical risk associated with Class 2A metals such as vanadium, nickel, etc., because these elements are commonly found in manufacturing equipment; for example, 316L stainless steel contains approximately 10% w/w nickel. Therefore, understanding the drug product equipment in terms of materials of construction will be a key factor in completing the risk assessment. The risk assessment should focus on any steps involving high kinetic energy (solids) or corrosive liquids that may facilitate the transfer of elements from the equipment into the product.

	A consideration of typical drug product manufacturing steps and conditions is provided in 

 Processing a liquid product containing a drug substance, excipients, and solvent (typically water-based buffer) in metallic vessels can potentially facilitate the transfer of elements into the liquid drug product, particularly at high/low pH. An assessment of material compatibility should be completed, taking into account factors such as ionic content, pH, temperature, hyrdophilicity/hydrophobicity, terminal sterilization conditions, and contact time. Testing should only be required if this assessment identifies a substantive risk.


	In general, the risk for elemental impurity contribution due to processing of solid drug product components in cGMP facilities is low, as stated explicitly in ICH Q3D. Although much equipment used to process drug product is metallic (e.g., stainless steel vessel), the majority of the processes used in drug product manufacture can be discounted as a source of risk. Even areas highlighted as needing consideration (

) are expected to only result in controls outside routine cGMP in extreme cases.

 As part of the holistic risk assessment of elemental impurities described in ICH Q3D, there is a need to consider the potential contribution resulting from environmental factors such as water and air. 

 Water used in the manufacture of both drug substances and formulated drug products is a potential source of elemental impurities. The level of risk, however, may be strongly related to the quality of water. This risk was examined in detail in a USP stimuli article on water for pharmaceutical use (5). The following position was articulated:

		The source water used in drug product manufacturing must meet the World Health Organization (WHO) standard for drinking water. When this source water is further purified in a contemporary plant to generate purified water (PW) and/or water-for-injection (WFI), the elemental impurity levels should be below acceptable concentrations allowed for drug roducts using option 1 control strategy defined in ICH Q3D.

		As part of standard GMP, water quality should be routinely monitored and the purification system and storage of the water should not re-introduce elemental impurities.

	Based on this position, the risk of elevated elemental impurity levels within aqueous-based formulations-even large-volume parenterals-is considered negligible. The risk associated with the use of water in the manufacture of the drug substance can also be effectively eliminated through the appropriate use of WHO-standard potable water combined with the use of USP purified water for the final stage in the manufacture of the drug substance, including its isolation. 

 Air is not likely to present a substantive risk; furthermore, air quality can also be managed through proper GMPs via use of HEPA filtered air, etc. No specific assessment is therefore generally required.

Container-closure systems as a potential source of elemental impurities in finished drug product


	One of the potential sources of elemental impurities is product packaging, often referred to as container-closure system (CCS). In determining the risk posed by the CCS, there are a number of factors that need to be taken into consideration including:

		Nature of formulation--mechanism for contamination

		Nature of risk: safety vs. quality risk

	In terms of the type of formulation, it is inconceivable that substantive or even trace-level contamination would occur where physical contact is limited to solid-to-solid contact. This is entirely consistent with the FDA Guidance for Industry, Container Closure Systems for Packaging Human Drugs & Biologics (6), which in relation to extractables and leachables considers solid-to-solid contact of low risk. It is therefore reasonable to conclude that any assessment of risk associated with CCSs should be limited to those associated with either liquid or semi-solid formulations.

	The second aspect of any risk assessment of the CCS involves an understanding of the potential levels of metals present within the material concerned. A major review of materials in manufacturing and packaging systems as sources of elemental impurities in packaged drug products was published in 2013 (7). The publication summarized literature data for a number of common packaging materials, including levels of elemental impurities within the component material (determined by digestion), as well as elemental impurities extracted from the component materials. The data, while fragmentary, are nevertheless comprehensive, and several key conclusions were drawn. While certain materials were found to contain elemental impurities, the presence of the elemental impurity was predominantly associated with deliberate use of metal catalysts, for example the use of antimony in the manufacture of polyethylene terephthalate (PET). In the case of PET, levels of antimony of approximately 50 ppm are typical (full digestion). Focusing particularly on those elements of high concern (Class 1 and Class 2), only fragmentary data exist to suggest even trace levels present within the component material, for example cadmium and lead levels up to 100 ppm were reported in polyvinyl chloride. Even when low levels of elemental impurities are present in the material itself, the effective “availability” of the elemental impurity needs to be considered. What is consistently clear from the extraction data presented is that extracted elemental impurity levels (under relevant conditions) are a minute fraction of the total elemental impurity levels present in the component materials, typically <0.1% of that observed following digestion. Therefore, even when trace levels of certain elements are found in the component material, the available elemental impurity concentration may represent an extremely low safety risk (

:  Illustration of risk factors associated with packaging.

	There may, however, be potential quality-related risks. Such risk may result from potential metal-catalyzed degradation of the product in question. While the interaction between packaging material and solid dosage forms is negligible (ICH Q3D), this may not always be the case for non-solid drug products. Reported examples include iron (8), nickel (9), and tungsten oxide. In the case of tungsten, this related to leaching from the syringe barrel, causing protein aggregation (10). As a consequence, many biopharmaceuticals include ethylenediaminetetraacetic acid to “mop up” metals. Assessments of such risks should be addressed on a case-by-case basis.

Assessing risk factors is key to implementing the new ICH Q3D guidelines for elemental impurities.