Kimberly Ehman, Ph.D., DABT, Director of Regulatory Toxicology, WuXi AppTec07.20.23
Regulatory guidance is always evolving. Regulators are frequently exposed to new data that can expand or alter the ways they apply existing rules. Innovation within medical device manufacturing exerts additional pressure on regulators to interpret existing standards in new and bold ways, too. Every so often, this dynamic landscape requires profound changes, and regulations need revising. Such is the case now with ISO 10993-17:2002.
The International Organization for Standardization (ISO) tries to review standards every five years to determine whether revisions are necessary. Published in 2002 and reviewed/confirmed most recently in 2016, revisions to 10993-17 have indeed become necessary. As the industry familiarizes itself with all the nuances of ISO/FDIS 10993-17—i.e., the final draft international standard—this column will detail the ways in which the proposed changes will help guide risk assessments.
The title change is significant for several reasons. ISO 10993-18:2020 established a need for medical device manufacturers to investigate their products’ chemical constituents more rigorously. It provides guidance on characterizing medical device materials and chemical substances used in the manufacturing process, including processing aids and residues, for the purpose of assessing the biological response to materials in the final product. The chemical characterization study provides an estimate of patient exposure to each reported chemical constituent and is the basis for the toxicological risk assessment (TRA).
In addition, the proposed title revision more accurately reflects the standard’s intent—estimating toxicological risk based on tolerable intake or tolerable contact level, and worst-case estimated exposure. It is worth noting the term “allowable limit” has been removed from the latest revision. ISO/FDIS 10993-17 also extends the previous version by clarifying when a toxicological risk assessment is recommended, how to calculate worst-case estimated exposure, and when potential harm to health should be addressed by other means.
Tolerable intake: TI estimates the daily exposure of an identified constituent over a specified time period (acute, subacute, subchronic, or chronic), on the basis of body weight, that is considered to be without appreciable harm to health. TI values are based on a toxicological review of the identified constituent, followed by selection of the most conservative point of departure (POD) for the intended use.
Once the appropriate POD is selected, uncertainty factors are applied to account for various forms of variability (inter- and intraspecies, dose route, duration of exposure). The draft version offers more detailed guidance about deriving a TI for both non-cancer and cancer endpoints, applying uncertainty factors, and considerations related to data quality and reliability. The TI is typically expressed as micrograms per kilogram of body weight per day (µg/kg/day).
Tolerable contact level: TCL estimates the surface-contact exposure to an identified constituent that is without appreciable irritation. TCLs are calculated from the non-irritating level (NIL), or minimally irritating level when a NIL is not available, and are considered in addition to tolerable intakes unless otherwise justified. The standard’s draft version offers guidance around application of uncertainty factors that must be considered when deriving a TSL. The TCL is represented in micrograms per centimeter squared of tissue at the contact site (µg/cm2).
Worst-case estimated exposure dose: EEDmax represents an exposure dose that is the maximum value for a specific intended clinical-use scenario. In the absence of release kinetics information, the EEDmax incorporates worst-case assumptions with respect to chemical release and device use. In most cases, the total quantity of each constituent identified in a chemical characterization study (following ISO 10993-18:2020) is assumed to be the daily patient exposure for the duration of intended use. The relevant patient population body weight and total number of devices used daily are included in the EEDmax calculation. The EEDmax is expressed in micrograms per kilogram body weight per day (µg/kg/day).
Margin of safety: MOS is unitless, as it is the ratio of the TI to the EEDmax. To characterize potential risk for each identified constituent, the EEDmax is compared to the derived TI or a Threshold of Toxicological Concern (TTC), when experimental data are unavailable to derive a TI.
In general, if the EEDmax is greater than the TI or TTC, the MOS will be less than one, indicating the exposure at that level could be harmful. A MOS value less than one requires additional risk analysis, risk evaluation, or risk control. If estimated exposure is less than the TI or TTC, the MOS will be greater than one, indicating negligible risk for that constituent, when contributing values to the MOS are demonstrated to be conservative (toxicological risk estimation is deliberately higher).
Expert judgement is warranted to determine if there are inherent uncertainties within the chemical characterization or risk assessment approach that may impact the MOS interpretation. The standard’s draft version discusses toxicological risk acceptance criteria as well as combining MOS values to address additivity of harm.
Identified constituent: An identified constituent is one for which molecular structure information is complete. Constituent identity can be obtained via non-targeted or targeted analytical approaches as described in ISO 10993-18. Examples of molecular structure information include molecular structure illustration or SMILES code, molecular formula, and Chemical Abstract Service Registry Number.
A chemistry report that contains too little structural information about a compound cannot support a TRA. In these cases, the chemical may be considered an “unknown.” Moreover, the TSL cannot be applied to constituents that are not adequately identified, as it would not be possible to exclude them as cohort of concern substances.
Total quantity: TQ is the amount of a constituent present in, or that can be extracted from, a medical device. The TQ of each constituent extracted shall account for the duration a medical device contacts the body, the quantity of devices extracted, and number of medical devices that contact the body. ISO/FDIS 10993-17 provides details about generating a TQ consistent with ISO 10993-18:2020 and documenting a TQ-TSL comparison to apply a screening limit. The TQ is expressed as micrograms (µg).
Toxicological screening limit: TSL is the cumulative exposure dose to an identified constituent over a specified period that will be without appreciable harm to health. Chemical characterization studies conducted in accordance with ISO 10993-18:2020 often lead to large data sets that must be evaluated in the TRA, and in many instances, the constituents are present at very low levels. The TSL provides an option for screening constituents when the TQ (which considers the number of devices that contact body) falls below a short-term (<30 day) or long-term (>30 day) TSL. The short-term TSL is 120 µg and considered protective for exposure <30 days, and the long-term TSL is 600 µg and considered protective for >30 days to lifetime.
The TSL can potentially streamline very large TRAs by reducing the number of chemicals that must go through the systematic approach to TRA. The standard’s draft version provides guidance about applying the TSL, including use limitations.
Release kinetics: Refers to the quantity of a constituent released from a medical device over time. Experimental release kinetics are usually generated in a leachables or simulated use chemical characterization study with multiple time points. The inclusion of multiple time points allows the risk assessor to refine the estimated daily exposure dose and generate a TRA that more appropriately captures potential risk.
Without experimental release kinetics, the standard’s draft version provides guidance about calculating assumed release kinetics based on a constituent’s TQ. To calculate assumed release kinetics, release days (Rd) for each category are considered—prolonged (Rd = 2), long-term up to one year (Rd = 31), and long-term > one year to lifetime (Rd = 366). Inclusion of assumed release kinetics can potentially reduce the need for additional chemistry and/or biological testing to mitigate concerns identified in exhaustive extractables studies.
ISO/FDIS 10993-17 provides extensive guidance around risk assessment practices for medical devices, which should improve consistency, and introduces additional tools to streamline the TRA approach and mitigate concerns without conducting additional testing—all of which will ultimately save time and money.
Kimberly Ehman, Ph.D., DABT is the Director of Regulatory Toxicology at WuXi AppTec Medical Device Testing. She has over 18 years of toxicology and medical device experience, with expertise in toxicological risk assessments for medical devices, food and beverage products, and electronic nicotine delivery systems. Prior to joining WuXi AppTec Medical Device Testing, she worked as a toxicologist for RTI International, Toxicology Regulatory Services, and Altria Client Services. She received her PhD in parasitology from McGill University and conducted postdoctoral research at the US EPA in developmental neurotoxicology. In her current position as Director of Regulatory Toxicology, Dr. Ehman provides medical device manufacturers and suppliers with technical and regulatory support for biocompatibility test programs and conducts quantitative toxicological risk assessments to support product safety and risk management decisions.
The International Organization for Standardization (ISO) tries to review standards every five years to determine whether revisions are necessary. Published in 2002 and reviewed/confirmed most recently in 2016, revisions to 10993-17 have indeed become necessary. As the industry familiarizes itself with all the nuances of ISO/FDIS 10993-17—i.e., the final draft international standard—this column will detail the ways in which the proposed changes will help guide risk assessments.
The Standard’s Name is Changing
Before its revision, the standard’s title was “ISO 10993-17: Biological evaluation of medical devices – Part 17: Methods for the establishment of allowable limits for leachable substances.” The current title of the draft is “ISO/FDIS 10993-17: Biological evaluation of medical devices – Part 17: Toxicological risk assessment of medical device constituents.”The title change is significant for several reasons. ISO 10993-18:2020 established a need for medical device manufacturers to investigate their products’ chemical constituents more rigorously. It provides guidance on characterizing medical device materials and chemical substances used in the manufacturing process, including processing aids and residues, for the purpose of assessing the biological response to materials in the final product. The chemical characterization study provides an estimate of patient exposure to each reported chemical constituent and is the basis for the toxicological risk assessment (TRA).
In addition, the proposed title revision more accurately reflects the standard’s intent—estimating toxicological risk based on tolerable intake or tolerable contact level, and worst-case estimated exposure. It is worth noting the term “allowable limit” has been removed from the latest revision. ISO/FDIS 10993-17 also extends the previous version by clarifying when a toxicological risk assessment is recommended, how to calculate worst-case estimated exposure, and when potential harm to health should be addressed by other means.
ISO/FDIS 10993-17’s Systematic Approach
The updated draft standard contains many new concepts and puts to rest several old ones. It describes a “systematic approach” to TRA that includes identifying hazards (a constituent that can cause harm relevant to the exposure scenario), and estimating risk based on three factors: tolerable intake (TI) or tolerable contact level (TCL); worst-case estimated exposure dose (EEDmax) and margin of safety (MOS).Tolerable intake: TI estimates the daily exposure of an identified constituent over a specified time period (acute, subacute, subchronic, or chronic), on the basis of body weight, that is considered to be without appreciable harm to health. TI values are based on a toxicological review of the identified constituent, followed by selection of the most conservative point of departure (POD) for the intended use.
Once the appropriate POD is selected, uncertainty factors are applied to account for various forms of variability (inter- and intraspecies, dose route, duration of exposure). The draft version offers more detailed guidance about deriving a TI for both non-cancer and cancer endpoints, applying uncertainty factors, and considerations related to data quality and reliability. The TI is typically expressed as micrograms per kilogram of body weight per day (µg/kg/day).
Tolerable contact level: TCL estimates the surface-contact exposure to an identified constituent that is without appreciable irritation. TCLs are calculated from the non-irritating level (NIL), or minimally irritating level when a NIL is not available, and are considered in addition to tolerable intakes unless otherwise justified. The standard’s draft version offers guidance around application of uncertainty factors that must be considered when deriving a TSL. The TCL is represented in micrograms per centimeter squared of tissue at the contact site (µg/cm2).
Worst-case estimated exposure dose: EEDmax represents an exposure dose that is the maximum value for a specific intended clinical-use scenario. In the absence of release kinetics information, the EEDmax incorporates worst-case assumptions with respect to chemical release and device use. In most cases, the total quantity of each constituent identified in a chemical characterization study (following ISO 10993-18:2020) is assumed to be the daily patient exposure for the duration of intended use. The relevant patient population body weight and total number of devices used daily are included in the EEDmax calculation. The EEDmax is expressed in micrograms per kilogram body weight per day (µg/kg/day).
Margin of safety: MOS is unitless, as it is the ratio of the TI to the EEDmax. To characterize potential risk for each identified constituent, the EEDmax is compared to the derived TI or a Threshold of Toxicological Concern (TTC), when experimental data are unavailable to derive a TI.
In general, if the EEDmax is greater than the TI or TTC, the MOS will be less than one, indicating the exposure at that level could be harmful. A MOS value less than one requires additional risk analysis, risk evaluation, or risk control. If estimated exposure is less than the TI or TTC, the MOS will be greater than one, indicating negligible risk for that constituent, when contributing values to the MOS are demonstrated to be conservative (toxicological risk estimation is deliberately higher).
Expert judgement is warranted to determine if there are inherent uncertainties within the chemical characterization or risk assessment approach that may impact the MOS interpretation. The standard’s draft version discusses toxicological risk acceptance criteria as well as combining MOS values to address additivity of harm.
Other Significant Updates
The entire list of additions can be found in the ISO/FDIS 10993-17 document when it’s published, but four more notable additions include the terms identified constituents, total quantity (TQ), toxicological screening limit (TSL), and release kinetics. These additions define the framework for establishing toxicological concern and underscore the importance of complete chemical characterization.Identified constituent: An identified constituent is one for which molecular structure information is complete. Constituent identity can be obtained via non-targeted or targeted analytical approaches as described in ISO 10993-18. Examples of molecular structure information include molecular structure illustration or SMILES code, molecular formula, and Chemical Abstract Service Registry Number.
A chemistry report that contains too little structural information about a compound cannot support a TRA. In these cases, the chemical may be considered an “unknown.” Moreover, the TSL cannot be applied to constituents that are not adequately identified, as it would not be possible to exclude them as cohort of concern substances.
Total quantity: TQ is the amount of a constituent present in, or that can be extracted from, a medical device. The TQ of each constituent extracted shall account for the duration a medical device contacts the body, the quantity of devices extracted, and number of medical devices that contact the body. ISO/FDIS 10993-17 provides details about generating a TQ consistent with ISO 10993-18:2020 and documenting a TQ-TSL comparison to apply a screening limit. The TQ is expressed as micrograms (µg).
Toxicological screening limit: TSL is the cumulative exposure dose to an identified constituent over a specified period that will be without appreciable harm to health. Chemical characterization studies conducted in accordance with ISO 10993-18:2020 often lead to large data sets that must be evaluated in the TRA, and in many instances, the constituents are present at very low levels. The TSL provides an option for screening constituents when the TQ (which considers the number of devices that contact body) falls below a short-term (<30 day) or long-term (>30 day) TSL. The short-term TSL is 120 µg and considered protective for exposure <30 days, and the long-term TSL is 600 µg and considered protective for >30 days to lifetime.
The TSL can potentially streamline very large TRAs by reducing the number of chemicals that must go through the systematic approach to TRA. The standard’s draft version provides guidance about applying the TSL, including use limitations.
Release kinetics: Refers to the quantity of a constituent released from a medical device over time. Experimental release kinetics are usually generated in a leachables or simulated use chemical characterization study with multiple time points. The inclusion of multiple time points allows the risk assessor to refine the estimated daily exposure dose and generate a TRA that more appropriately captures potential risk.
Without experimental release kinetics, the standard’s draft version provides guidance about calculating assumed release kinetics based on a constituent’s TQ. To calculate assumed release kinetics, release days (Rd) for each category are considered—prolonged (Rd = 2), long-term up to one year (Rd = 31), and long-term > one year to lifetime (Rd = 366). Inclusion of assumed release kinetics can potentially reduce the need for additional chemistry and/or biological testing to mitigate concerns identified in exhaustive extractables studies.
Subtractions From the Original ISO 10993-17
While not an exhaustive list, there are several old or outdated concepts that have been removed from the standard’s current version. Device manufacturers and lab testing partners familiar with ISO 10993-17:2002 will recognize these concepts and, if still in use, should update terminology and approaches, if necessary. Notable subtractions include:- Allowable limit
- Benefit factor
- Proportional exposure factor
- Tolerable exposure
- Utilization factor
A Final Word on ISO/FDIS 10993-17
It took a long time to get here. Whispers about the revised standard’s timing for publishing have swirled since early 2022. But the delay is understandable. Regulatory expectations around chemical characterization studies have led to large data sets for TRA evaluation, with the overwhelming majority concluding equivocal risk. Thus, the manufacturer is left to conduct additional chemistry or biocompatibility testing, which costs more money and creates submission delays.ISO/FDIS 10993-17 provides extensive guidance around risk assessment practices for medical devices, which should improve consistency, and introduces additional tools to streamline the TRA approach and mitigate concerns without conducting additional testing—all of which will ultimately save time and money.
Kimberly Ehman, Ph.D., DABT is the Director of Regulatory Toxicology at WuXi AppTec Medical Device Testing. She has over 18 years of toxicology and medical device experience, with expertise in toxicological risk assessments for medical devices, food and beverage products, and electronic nicotine delivery systems. Prior to joining WuXi AppTec Medical Device Testing, she worked as a toxicologist for RTI International, Toxicology Regulatory Services, and Altria Client Services. She received her PhD in parasitology from McGill University and conducted postdoctoral research at the US EPA in developmental neurotoxicology. In her current position as Director of Regulatory Toxicology, Dr. Ehman provides medical device manufacturers and suppliers with technical and regulatory support for biocompatibility test programs and conducts quantitative toxicological risk assessments to support product safety and risk management decisions.