Deficiency findings in biological evaluation documentation are one of the most frequent causes of delayed medical device submissions. Whether you are filing with the FDA or submitting to a European notified body under the MDR, the same categories of deficiency appear repeatedly across review cycles.
This article identifies the five most commonly cited deficiency areas in ISO 10993 biological evaluation documentation, based on published FDA guidance, NAMSA technical publications, and documented notified body expectations. For each area, we describe what reviewers expect, what they typically find instead, and what stronger documentation looks like.
Insufficient chemical characterization
Chemical characterization under ISO 10993-18 is one of the most scrutinized areas in any biological evaluation package. The FDA has specifically identified inconsistencies across laboratories and deficiencies in chemical analysis as a recurring concern in premarket submissions.1
The most common deficiency here is not the absence of chemical data, but rather incomplete or inadequately justified data. Reviewers expect to see extractable and leachable testing conducted under conditions that reflect or exceed clinical exposure, performed in triplicate, using solvents that represent the range of polarity relevant to the device's contact environment.2 What they often find instead is testing performed under a single extraction condition with limited analytical coverage, or a justification for omitting testing that lacks specificity about why the materials in question do not require it.
Analytical evaluation thresholds (AETs) are another frequent problem area. FDA reviewers have noted that submitted AET calculations frequently do not account for instrument variability inherent to semi-quantitative analytical techniques.3 When the threshold is set too high, potentially relevant extractables may go unreported, which undermines the toxicological risk assessment that depends on this data.
What to check: Does your chemical characterization section document extraction conditions, solvent selection rationale, AET calculations, and the number of replicates? If any of these are missing or vaguely described, expect a deficiency finding.
A Class IIb wound care device uses three polymeric components. The chemical characterization section includes extractable data for two of them but omits the third, citing its "established history of use in medical devices" without analytical data. The extraction was performed using a single polar solvent at one temperature, with no documented rationale for omitting non-polar extraction. The AET was set at 1.0 mcg/mL with no explanation of how instrument sensitivity or variability was factored in.
Likely reviewer response: Three separate deficiency questions: one requesting analytical data for the third material or a documented justification for its omission, one challenging the single-solvent extraction as insufficient to represent the full range of potential extractables, and one requesting the AET calculation methodology with instrument variability accounted for.
Weak or missing toxicological risk assessment
The toxicological risk assessment (TRA) under ISO 10993-17 is where chemical characterization data gets translated into a safety conclusion. This is also where many biological evaluation reports fall apart.
A TRA that concludes materials are "biocompatible" or "non-toxic" without identifying specific chemical constituents, citing tolerable exposure (TE) or tolerable intake (TI) values from recognized sources, and calculating margins of safety (MOS) for each identified substance is substantively insufficient. The FDA guidance explicitly states that risk assessments should include relevant health-based endpoints such as no observed adverse effect levels (NOAEL), tolerable intake, tolerable exposure, and margin of safety calculations.4
The 2025 revision of ISO 10993-1 further strengthens this requirement by embedding risk management terminology from ISO 14971 throughout the biological evaluation process.5 A TRA that does not explicitly connect its findings to the broader risk management framework will increasingly be viewed as incomplete.
Commonly, the deficiency is not that a TRA is absent, but that it exists in a general form that does not address the specific extractable profile identified for the device under review. Generic statements about material safety history do not substitute for substance-specific risk calculations.
A spinal implant manufacturer submits a BER with a TRA section that identifies six extractable compounds from the chemical characterization study. The TRA provides margin of safety calculations for four of them but addresses the remaining two with the statement: "These compounds are present at trace levels and are not expected to pose a toxicological concern." No tolerable exposure values or quantitative analysis is provided for either compound.
Likely reviewer response: A deficiency requesting quantitative risk assessment for the two unaddressed compounds, including identification of tolerable exposure values from recognized sources and documented MOS calculations. The phrase "not expected to pose a concern" is not a substitute for a documented risk assessment, regardless of the concentration.
Unsupported biological equivalence claims
Biological equivalence is one of the most challenged claims in biological evaluation documentation. The concept allows manufacturers to leverage existing data from a comparable device or material, potentially reducing or eliminating the need for new biocompatibility testing. In practice, reviewers scrutinize these claims heavily because the consequences of an unsupported equivalence argument can mean that critical safety data is missing from the evaluation.
ISO 10993-1:2025 significantly restructures the framework for biological equivalence. Where the 2018 version briefly addressed the concept in Figure 1, the 2025 revision provides a more detailed and structured approach. Critically, the standard now shifts focus from input comparisons (same material, same process, same use) to output comparisons: specifically, whether the biological risks to a patient are sufficiently similar.5
The most common deficiency is a claim of equivalence that documents similarity of materials and manufacturing but does not address whether the biological risk profile of the subject device is adequately represented by the comparator data. Reviewers also challenge equivalence claims where the comparator device data is not accessible to the manufacturer or is not sufficiently documented to support the comparison.
It is worth noting that the EU and US approach equivalence differently. The FDA's 510(k) pathway focuses on substantial equivalence to a predicate, and equivalence tables often stop at identifying same or similar characteristics. EU notified body reviewers apply a more rigorous three-phase assessment of equivalence (technical, biological, and clinical) as outlined in MDCG 2020-5, and they specifically scrutinize whether manufacturing-related differences could affect the release characteristics of substances from the device.6 A biological equivalence argument that satisfies FDA expectations may still receive deficiency findings from a European notified body if it does not address the MDR's more granular requirements.
Stronger documentation: A well-supported equivalence argument identifies the comparator device, documents the basis of comparison across material, manufacturing, physical properties, and clinical use, and explicitly addresses any differences. It then evaluates whether those differences could affect the biological risk profile.
A manufacturer of a Class IIa catheter claims biological equivalence with a predicate device from its own portfolio. The equivalence section documents that both devices use the same base polymer and have the same intended body contact. However, the subject device is manufactured at a different facility using a modified extrusion process, and this difference is not addressed. The predicate's biocompatibility data is referenced by test report number, but the test reports are not included in the submission package.
Likely reviewer response: Two findings: one requesting an assessment of whether the manufacturing process change could affect the extractable profile or material properties, and one requesting that the referenced test reports be provided or their key findings summarized with sufficient detail to support the equivalence conclusion.
Inadequate risk management integration
ISO 10993-1:2025 takes the integration of biological evaluation with ISO 14971 risk management to a new level. The standard now explicitly adopts the terminology, principles, and workflow of ISO 14971, framing the biological evaluation as a component of the overall risk management process.5
In practice, many biological evaluation reports still treat the risk assessment as a standalone exercise rather than an integrated part of the risk management file. The BER references "risk" in general terms but does not trace specific biological hazards through to severity estimation, probability assessment, or risk control measures documented in the risk management file.
The deficiency finding typically reads as a lack of traceability. Reviewers expect to see a clear connection between the biological hazards identified in the evaluation, the corresponding entries in the risk management file, and the risk control measures applied. If your BER concludes that a sensitization risk exists for a particular material constituent but your risk management file does not contain a corresponding hazard entry with a severity and probability assessment, you have a traceability gap.
The 2025 revision introduces the terminology of "biological hazard," "biologically hazardous situation," and "biological harm" to explicitly narrow the scope to biological aspects while maintaining alignment with the broader risk management vocabulary.5 Documentation that does not adopt this structure will increasingly be viewed as non-conforming.
A surgical instrument manufacturer's BER identifies a potential sensitization risk from nickel content in a stainless steel alloy. The conclusion states that the risk is "acceptable based on the material's established biocompatibility." The risk management file contains a general entry for "biocompatibility" but no specific hazard line item for nickel sensitization, no severity or probability assessment for that hazard, and no documented risk control measure.
Likely reviewer response: A finding noting that the biological evaluation identifies a specific hazard (nickel sensitization) that is not traceable to the risk management file. The reviewer requests a corresponding hazard entry with severity estimation, probability assessment, and documentation of how the risk is controlled or why it is acceptable without additional controls.
Incomplete or missing biological evaluation plan
The biological evaluation plan (BEP) is the foundational document that establishes the evaluation strategy. A deficient BEP undermines the credibility of the entire package because it suggests the evaluation was not systematically planned.
The most common deficiency is a BEP that does not adequately justify endpoint selection. ISO 10993-1:2025 replaces the single large Table A.1 from the 2018 version with four separate tables organized by contact type, and introduces new guidance sections on biological effects for evaluation, other biological effects, and other factors to be considered.5 A BEP that simply references "Table A.1" without addressing the restructured categorization framework will be flagged.
Other frequently cited deficiencies in the BEP include failure to justify endpoints that were excluded from evaluation, omission of foreseeable misuse considerations (a new requirement in 2025), and lack of a clear rationale for why existing data is sufficient versus where new testing is required.
The FDA guidance emphasizes that the biological evaluation should be conducted within the framework of the risk management process, including a device risk assessment that leads to identification of potential biological risks, a review of all relevant information, and a clear summary of conclusions with identified knowledge gaps.4
A startup submitting its first Class IIa device includes a BEP that lists the standard biological endpoints (cytotoxicity, sensitization, irritation) with a note that "testing will be performed per ISO 10993." The BEP does not justify why these specific endpoints were selected over others, does not explain why genotoxicity or chronic toxicity were excluded given the device's 30-day contact duration, and does not address foreseeable misuse. The device categorization references "Table A.1" without specifying which contact category applies under the 2025 framework.
Likely reviewer response: Multiple findings: one requesting justification for excluded endpoints with reference to the device's contact type and duration, one requesting foreseeable misuse to be addressed in the evaluation plan, and one requesting that device categorization be updated to reference the applicable 2025 contact category tables.
The pattern across all five findings: In every case, the deficiency is not about missing documents. It is about documents that exist but contain unsupported conclusions, missing rationale, or disconnected logic. The gap is in the quality of the argumentation, not in the presence of the paperwork.
What this means for your next submission
Each of these five deficiency areas shares a common thread: the documentation exists, but the reasoning within it is incomplete or insufficiently specific. Generic statements, unsubstantiated conclusions, and missing cross-references between documents are what trigger deficiency findings.
Before your next submission, review your package with these five areas in mind. For each section, ask whether your conclusions are supported by specific referenced data, whether your rationale addresses why you made the choices you made (not just what you did), and whether your documents reference each other consistently.
BioEvalPro evaluates your documentation against each of these deficiency patterns, scoring the strength of your justifications and flagging unsupported claims, conflicting statements, and missing sections with specific remediation guidance. request a gap analysis for early access, or get in touch to discuss how it fits your submission workflow.
1 FDA Draft Guidance, "Chemical Analysis for Biocompatibility Assessment of Medical Devices," 2025. Available at fda.gov/regulatory-information/search-fda-guidance-documents
2 ISO 10993-18:2020, "Biological evaluation of medical devices, Part 18: Chemical characterization of medical device materials within a risk management process."
3 OSMA Spring Meeting Summary (2023), as reported by BONEZONE. FDA representatives identified AET calculation deficiencies and extraction methodology inconsistencies as common findings in biocompatibility reviews. bonezonepub.com
4 FDA Guidance, "Use of International Standard ISO 10993-1: Biological evaluation of medical devices, Part 1: Evaluation and testing within a risk management process," updated 2023. fda.gov
5 NAMSA, "ISO 10993-1:2025 Updates: Top 10 Biological Evaluation Essentials in the Revision," October 2025. namsa.com
6 MDCG 2020-5, "Guidance on Clinical Evaluation: Equivalence." See also RQM+, "How To Approach Biological Equivalence Under The EU MDR," MedDevice Online. meddeviceonline.com
Every deficiency in this article is one BioEvalPro checks for.
Weak TRAs, unsupported equivalence claims, missing risk traceability, incomplete BEPs. Upload your package and see which ones are in yours.