Thresholds of Toxicological Concern (TTC)

The biological evaluation of medical devices is based on a risk-management process conducted in accordance with the ISO 10993 standards. The toxicological risk assessment of a device’s constituents (ISO 10993-17) relies on data obtained either through chemical characterization (ISO 10993-18) or through analysis of available information (ISO 10993-1).

When specific information on an identified substance is unavailable or insufficient to establish its tolerable intake, the Threshold of Toxicological Concern (TTC) approach—a probabilistic tool—can be applied. The TTC concept is a public-health framework that sets a daily human exposure level, derived from extensive toxicological databases, below which the risk of adverse effects is considered negligible. It serves as a screening method when substance-specific toxicological data are lacking, providing a rapid orientation for risk assessment. TTC values can be used to evaluate risks related to genotoxicity, carcinogenicity, systemic toxicity, and reproductive and developmental toxicity. Note, however, that local risks such as irritation and sensitization are not covered by this approach and must be assessed through other means (biological tests, literature data, etc.).

For medical devices, the relevant TTC values are defined in Technical Specification ISO/TS 21726, which is based on the principles established in ICH M7(R2)—the guideline for assessing mutagenic impurities in pharmaceutical products. If the estimated maximum exposure dose is below the applicable TTC value, the substance is considered to pose an acceptable toxicological risk to the patient.

This article reviews the origin of TTC thresholds, the methods used to derive them, and their application in the toxicological assessment of healthcare products.

1. Non-carcinogenic substances

In 1978, Cramer et al. developed a decision-tree approach for predicting the toxicity of substances based on their chemical structure and physiological behavior. This model laid the groundwork for categorizing substances by toxicity level. Building on that work, Munro et al. (1996) analyzed 613 non-carcinogenic substances assigned to the Cramer classes and derived three class-specific human exposure thresholds:

• Class I (low toxicity): 1,800 µg/person/day
• Class II (moderate toxicity): 540 µg/person/day
• Class III (high toxicity): 90 µg/person/day

These thresholds were calculated from NOEL values (No Observed Effect Levels)—the highest tested concentration or dose that produced no measurable toxic or biological effect—in sub-chronic and chronic rodent studies.

Technical Specification ISO/TS 21726:2019 has adopted these thresholds, assigning non-carcinogenic substances to the corresponding Cramer structural classes according to their toxicity profiles.

Figure 1 indicates which Cramer threshold applies to the structural class of the substance being assessed.

This standard is currently under revision, and the thresholds are expected to change in the forthcoming edition. The new version is likely to introduce exposure-time–based TTC values for non-carcinogenic substances that are less conservative than the lifetime-protective limits presently derived from the Cramer classes.

TTC thresholds derived from the Cramer classification are based on studies involving oral exposure. When a healthcare product is administered parenterally, however, regulatory guidance is far less comprehensive. In this context, the Product Quality Research Institute (PQRI)—a U.S. organization founded by the Food and Drug Administration (FDA), the American Association of Pharmaceutical Scientists (AAPS), and the United States Pharmacopeia (USP)—recommends using a threshold of 50 µg/day to assess systemic toxicity risks from non-carcinogenic substances in parenteral and ophthalmic drug products (this value has not yet been codified and is referenced mainly in working-group presentations). A lower threshold of 5 µg/day is suggested when the substance may present a risk of irritation or sensitization.

The Extractables and Leachables Safety Information Exchange (ELSIE)—an international consortium of pharmaceutical, biotech, and materials-supplier companies that pools data and best practices on extractables and leachables—has evaluated 252 non-mutagenic compounds and, on that basis, has proposed additional parenteral TTC values:

• 35 µg/day for exposures longer than 10 years
• 110 µg/day for exposures of 1–10 years
• 180 µg/day for exposures shorter than 1 year

The draft ICH Q3E guideline, prepared by the European Medicines Agency (EMA), introduces Qualification Thresholds (QT) for impurities in products administered orally, parenterally, dermally, transdermally, or by inhalation. These thresholds—summarized above—were derived from a dataset of 330 substances. Because the draft does not detail the methodology used to calculate the limits, a thorough critical appraisal is challenging. It is nevertheless evident that the proposed parenteral, dermal, and inhalation QTs (for non-carcinogenic compounds) are often lower than the corresponding parenteral TTC values (aimed at mutagenic compounds and derived from ICH M7, discussed later).

Therefore, according to the work of Cramer and Munro, non-carcinogenic substances have been classified by toxicity level, with protective TTC thresholds based on oral exposure. ISO/TS 21726:2019 adopted these limits, but the forthcoming revision is expected to introduce less conservative values that take exposure duration into account. For parenteral, dermal, and inhalation routes, other bodies—such as PQRI, ELSIE, and the EMA—have proposed stricter, route- and duration-specific thresholds.

2. Carcinogenic substances

The exposure-threshold concept was incorporated into U.S. regulations for the first time in 1995, when the FDA applied it to chemicals that could migrate from food-contact materials. The agency established a regulatory limit called the Threshold of Regulation (TOR), derived from genotoxic and non-genotoxic carcinogenicity data obtained in rodent studies. The TOR was set at 0.5 µg/kg of food—calculated with mathematical modeling—which equates to a lifetime exposure of roughly 1.5 µg/person/day and corresponds to an estimated excess cancer risk of one in a million. The original derivation was based on a dataset of 492 substances. In 1999, Cheeseman et al. reinforced the robustness of the 0.5 µg/kg TOR after expanding the database to 709 substances.

In 2004, Kroes et al. re-examined the data set—now expanded to 730 substances—and revised the thresholds. For compounds with structural alerts for genotoxicity (i.e., a potential for genotoxic carcinogenicity), the regulatory limit was reduced from 1.5 µg/person/day to 0.15 µg/person/day. Even at this lower dose, however, certain “potent carcinogens” may still pose an unacceptably high cancer risk. These compounds, grouped in the so-called Cohort of Concern (CoC), include N-nitroso substances, azoxy compounds, aflatoxins, and steroids. Because of their heightened risk, the TTC approach does not apply to these chemicals.

In 2023, the EMA published ICH M7 (R2), which addresses mutagenic carcinogenic impurities in pharmaceutical products. The guideline endorses an exposure limit of 1.5 µg/person/day—corresponding to an excess lifetime cancer risk of 10⁻⁵—but only when such exposure can be justified by a therapeutic benefit to the patient.

According to ISO/TS 21726, and drawing on the principles of ICH M7, carcinogenic substances (through both genotoxic and non-genotoxic mechanisms) are addressed by the following thresholds, which vary with the medical-device category and the duration of patient contact:

3. Conclusion

The TTC concept has become a structured, centralized tool for risk assessment when specific toxicological data on medical-device constituents are lacking. It is, however, a method of last resort: whenever substance-specific toxicology data are available, the assessment must be carried out in accordance with ISO 10993-17, which takes precedence over the TTC approach.

Regulatory advances now provide clear TTC benchmarks for both non-carcinogenic and carcinogenic substances, tailored to the route and duration of exposure. Even so, TTC does not replace detailed knowledge of the chemicals involved, nor does it cover local effects such as irritation or sensitization. Those risks must still be evaluated through available literature or through biological testing—ISO 10993-23 for irritation and ISO 10993-10 for sensitization. Only by combining TTC with substance-specific data and appropriate local-effect testing can we ensure effective patient safety.

The ongoing evolution of standards—particularly the forthcoming revision of ISO/TS 21726 and the introduction of ICH Q3E—is expected to bring new thresholds, strengthen regulatory protections, and harmonize risk-assessment practices.

Need help?

Efor’s specialists can support you with toxicological analyses and the preparation of your Toxicological Risk Assessments (TRA). Our biocompatibility and toxicology experts can also help you:

• Shape an appropriate biological-evaluation strategy
• Draft your biological evaluation plans and reports
• Coordinate laboratory testing with the facility of your choice

Our team’s expertise is kept current through continuous monitoring of scientific advancements, new methodologies, and evolving standards.

To discuss your project, contact the technical expertise teams within our Solution & Project Delivery division at: solutionprojectdelivery@efor-group.com.