化学表征数据怎么支撑毒理学风险评估

结论：表征是为了找出「有什么、有多少」，评估才回答「要不要紧」

化学表征的作用是识别产品可能释放到人体的化学物质，并给出量的信息。它本身不给出安全结论，结论要靠后续的毒理学评估。

这个分工要搞清楚，否则容易出现两种偏差：一种是做了表征就以为完成了评价；另一种是表征数据不足，评估缺乏输入而只能靠假设。

合理的关系是：表征提供「有什么、有多少」，毒理学评估回答「这个量在这个接触条件下要不要紧」。

表征的目的

化学表征在评价体系中主要服务于几个目的：

识别风险物质。 找出可能从产品中释放的化学物质。

为毒理学评估提供输入。 物质名称和量是评估的基础数据。

支持材料等同性论证。 两个材料的化学表征结果相近，是等同性的证据之一。

支持变更评估。 变更前后做表征对比，判断变更是否影响生物学安全。

替代部分动物试验。 在数据充分的前提下，可以减少某些终点的动物试验需求。

第五项是这项工作近年受重视的主要原因。 但要注意：替代是有条件的，条件是表征做得足够充分、评估做得足够严谨。表征做得粗糙，反而无法替代。

方法的选择

方法组合要覆盖可能存在的物质类型。 只做一类方法，会漏掉该方法覆盖不到的物质。比如只做气相色谱质谱，难挥发的物质就看不到。

方法选择的依据是材料信息。 知道材料里有什么类型的添加剂，才能选对方法。所以材料清单是方法选择的前提。

浸提条件的设定

浸提是表征中影响结果最大的环节，几个关键决定：

浸提介质。 要覆盖不同极性——通常包括极性和非极性介质，以模拟不同的体液环境。

浸提条件。 温度、时间、样品与介质的比例。条件越剧烈，提取出的物质越多。

浸提方式。 是模拟实际使用条件，还是采用加严条件以求覆盖。两种思路各有适用场合。

样品状态。 必须是最终成品，经过全部加工和灭菌。

浸提条件必须在报告中完整记录，因为数据的含义完全依赖于条件。脱离条件谈浓度，没有意义。

数据质量的要求

表征数据要能支撑评估，必须满足几个质量要求：

定性可靠。 物质的鉴定要有依据，不能只靠谱库匹配的相似度。

定量准确。 有标准品的用标准品定量；没有标准品的要说明半定量的方法和不确定度。

检出限足够低。 检出限要低于评估所需的判断限值，否则「未检出」说明不了问题。

空白对照。 排除来自容器、试剂、环境的干扰。

回收率。 说明方法的提取效率。

第三项最容易出问题。 报告写「未检出」，但检出限高于关注限值，这个「未检出」就没有意义。检出限必须与评估限值一起看。

数据怎么转化为评估输入

拿到表征数据之后，评估的基本步骤是：

第一，确定每种检出物质的日暴露量。 由浓度、浸提条件、产品接触方式和接触时间推算。

第二，查找该物质的毒理学数据。 允许限值或参考剂量。

第三，比较暴露量与限值。 得出该物质的风险判断。

第四，考虑组合作用。 多种物质同时存在时的综合考虑。

第五，对无毒理学数据的物质做处理。 可以用关注阈值的思路，或者进一步的数据检索。

第一步的推算是容易出错的环节。 浸提得到的浓度不能直接当成人体暴露量，要经过合理的换算，而换算的假设要写清楚。

常见问题

材料信息不足导致方法选错。 选了覆盖不到关键物质的方法。

浸提条件记录不全。 数据无法解释。

检出限不匹配。 未检出但限值更低。

未检物质未做说明。 谱图上有峰但未鉴定，报告中略过。

未做空白。 无法排除污染来源。

样品不是最终状态。 用未灭菌的样品，漏掉灭菌引入的物质。

最后一项在使用环氧乙烷或辐照灭菌的产品上尤其重要，因为灭菌过程本身可能产生或引入新的物质。

表征结果的呈现

表征数据量通常很大，呈现方式影响可用性。建议报告包含：

物质清单表。 检出物质、鉴定依据、浓度、检出限、不确定度。

谱图。 保留原始谱图作为附件。

未鉴定峰的说明。 有峰但未能鉴定的，说明其大致含量和后续处理方式。

空白与回收率数据。 作为质量控制证据。

条件汇总表。 浸提条件、方法参数一览。

「未鉴定峰的说明」这一项常被省略，但它对评估的完整性有影响——如果有相当含量的物质未被鉴定，评估就存在缺口，应当说明如何处理这个缺口。

与生物学试验的配合

化学表征与生物学试验是互补关系，不是替代关系：

化学表征的优势。 能识别具体物质、给出量的信息、可用于变更对比、不涉及动物。

化学表征的局限。 只能测到方法覆盖的物质；未鉴定的物质无法评估；不能反映材料的整体生物学效应。

生物学试验的优势。 反映整体效应，不依赖物质清单的完整性。

生物学试验的局限。 不能指出是哪种物质引起的效应，不利于改进。

合理的做法是两者配合使用，按缺口分析的结论确定各自的范围。

我们的做法

做化学表征时，我们会先要材料清单和加工信息，据此设计方法组合和浸提条件。没有材料信息就开测，方法可能覆盖不全，钱花了但数据不完整。

报告方面，我们会完整记录浸提条件、方法参数、检出限、空白结果和回收率。这些信息是数据可用性的前提 ——缺了它们，数据在后续评估中会被质疑。

如果你在准备化学表征，不确定方法怎么选、条件怎么定，可以把材料清单和产品信息发过来一起讨论，或者直接联系：132 4819 8029。能力范围见服务介绍，流程见检测流程，更多内容见知识库。

English version

Conclusion. Chemical characterisation identifies the chemical substances a product may release to the body and quantifies them. It does not itself produce a safety conclusion; that comes from the toxicological assessment that follows. Keeping the division clear avoids two errors: assuming that characterisation completes the evaluation, and generating characterisation data so thin that the assessment must fall back on assumption. Properly, characterisation supplies what is present and how much, and toxicological assessment answers whether that quantity matters under those contact conditions.

Purposes of characterisation. It identifies substances of concern that may be released. It provides input to toxicological assessment, since substance identity and quantity are the basic data. It supports material equivalence arguments, where similar characterisation results are one form of evidence. It supports change assessment, comparing characterisation before and after to judge whether biological safety is affected. And it can reduce the need for animal testing at certain endpoints where the data are sufficient. That last is the main reason the work has gained prominence, but the substitution is conditional: it requires characterisation thorough enough and assessment rigorous enough. Coarse characterisation substitutes for nothing.

Choosing methods. Gas chromatography with mass spectrometry suits volatile and semi-volatile organics. Liquid chromatography with mass spectrometry suits less volatile organics. Inductively coupled plasma techniques suit metallic elements. Ion chromatography suits inorganic ions. And total organic carbon serves as a screen for total organic content. The combination must cover the substance types that may be present, since a single technique misses whatever it cannot see; running gas chromatography alone leaves non-volatile substances invisible. Method selection rests on material information: knowing what additive types the material contains is what allows the right methods to be chosen, so the material list is a prerequisite.

Setting extraction conditions. Extraction affects results more than anything else, and several decisions matter. The extraction media must span polarities, usually including polar and non-polar media to represent different body fluid environments. Conditions of temperature, time and sample-to-medium ratio determine how much is extracted, with more aggressive conditions yielding more. The approach must be chosen between simulating actual use and applying exaggerated conditions for coverage, each suiting different situations. And the sample must be final product, fully processed and sterilised. Extraction conditions must be recorded completely in the report, because the meaning of the data depends entirely on them; a concentration divorced from its conditions says nothing.

Data quality requirements. Identification must be reliable, with a basis beyond library match similarity. Quantification must be accurate, using reference standards where available and, where not, stating the semi-quantitative approach and its uncertainty. Detection limits must be low enough, below the threshold the assessment needs, or a not-detected result proves nothing. Blanks must be run to exclude interference from containers, reagents and environment. And recovery must be stated to show extraction efficiency. The third causes the most trouble: a report stating not detected at a detection limit above the threshold of concern conveys no information. Detection limits must always be read alongside assessment thresholds.

Turning data into assessment input. Determine the daily exposure for each detected substance, derived from concentration, extraction conditions and the product's contact mode and duration. Find toxicological data for the substance, meaning a permitted limit or reference dose. Compare exposure against limit to judge that substance's risk. Consider combined effects where several substances are present. And handle substances lacking toxicological data, using a threshold-of-concern approach or further data searching. The first step is where errors arise: an extraction concentration is not a human exposure, and the conversion requires reasoned assumptions that must be written down.

Common problems. Insufficient material information leading to method selection that misses key substances. Incomplete recording of extraction conditions, leaving data uninterpretable. Mismatched detection limits, with not-detected results above the threshold of interest. Unidentified peaks passed over without comment. Absence of blanks, leaving contamination sources unexcluded. And samples that are not in final condition, so that substances introduced by sterilisation are missed. The last matters particularly for products sterilised by ethylene oxide or irradiation, since the process itself may generate or introduce substances.

How we handle it. We ask for the material list and processing information first and use them to design the method combination and extraction conditions. Starting without material information risks incomplete coverage, with money spent and data still incomplete. In reporting we record extraction conditions, method parameters, detection limits, blank results and recovery in full, because those are what make the data usable; without them the data will be challenged in the assessment that follows.

Send us the material list and product information and we will work out the methods and conditions. Phone or WeChat: +86 132 4819 8029.