美国药典 795 和 797 肽文档指南

什么 USP 795 和 797 肽文档实际上决定

各章节对准备工作进行分类, 不是分子. 美国药典 795 和 797 肽文档问题只需一个触发器: 成品制剂在给药时是否必须无菌. 那个单一条件, 不是肽的纯度或其价格等级, 决定您应持有的记录由哪一章管辖.

无菌路线位于下方 <797>. 注射, 静脉外加剂, 眼科, 鞘内制剂, 以及不与外界环境相通的体腔准备工作都落在了线的那一边. 非无菌剂型, 包括口服液, 胶囊, 面霜, 凝胶, 和栓剂, 落入 <795> (国家大剧院, USP 混合标准 <795> 和 <797> 推介会, 2026).

边界还有第二条边缘，令开发人员感到惊讶. 严格按照制造商批准的标签制备的常规生产的无菌产品被描述为超出了范围 <797>. 根据其标签为个别患者重新配制传统制造的非无菌产品, 而不存储它以供将来使用, 不复利 <795> (密西西比州药房委员会, <797> 常见问题解答文档, 2023).

对购买的错误分类导致常规质量问题升级. 在订购时弄错类别，每个下游文档请求都会按照错误的标准到达.

研究级电源, 开发支持, 和复配相关用途: 划清界限

肽所属的类别由其预期使用姿势和标签决定, 不是看色谱图看起来有多干净. 研究级肽供应链支持实验室和开发工作: 该材料被标记为仅供研究使用, 标签决定了它的地位. 通过 HPLC 和质谱法显示身份和纯度的大量分析证书记录了材料本身. 它不, 靠自己, 使该材料允许混合.

这种区别很重要，因为复合途径有其自己的进入条件, 两个框架之间存在差异.

根据第 503A 条, 仅当符合适用的 USP 或 NF 各论以及 USP 关于药物配制的章节时，才可以使用原料药, 或者 是 FDA 批准的药品的组成部分，但不存在专着, 或者 出现在 503A 散装列表中 (美国FDA, 根据第 503A 条用于配制的原料药, 当前内容截至 2026-05-14). 这些都是替代方案, 不是按顺序满足的清单. 分别地, 该物质必须带有有效的分析证书，并且必须由根据第 1 节在 FDA 注册的企业制造 510 (美国FDA, 当前内容截至 2026-05-14).

外包设施在较窄的规则下运作: 仅当原料药出现在 503B 原料药清单中或在配制时用于配制 FDA 药品短缺清单上的药物时，才可以使用原料药, 分配, 和配药 (美国FDA, 用于配制的原料药, 当前内容截至 2026-03-26).

经常出现的错误是将纯度数字视为品类升级. ≥98% HPLC 结果描述了该批次, 不是其监管姿态.

文档集: 辩护记录实际上包含什么

可辩护的记录不是证书文件夹. It is a set of artifacts where each one proves a specific obligation, and a reviewer can trace the chain from supplier lot to internal use lot without asking you a question.

这 FDA guidance to compounders on knowing bulk suppliers frames the obligations directly: source API from FDA-registered facilities, hold valid certificates of analysis, and know what testing the supplier actually performs. Those three obligations map onto distinct records.

The chain-of-custody handoff is the artifact most often missing. Without it, a supplier lot number and an internal use lot number sit in separate systems with no recorded link, and a reviewer cannot confirm that the material in the vial is the material the COA describes.

警告: The record that matters most is the one written when the numbers do not match. A discrepancy between in-house verification and the supplier COA, and the investigation that followed, is the document a quality reviewer will ask for first.

身份测试: 确认分子, 不仅仅是弥撒

Identity testing is a separate control from purity, and a single intact-mass figure does not establish it. A mass that matches the expected molecular weight tells you something is present at that mass. It does not tell you the sequence is correct, that the counterion is what the label claims, or that the material is free of the non-peptide components that sit outside the peptide-related impurity thresholds.

服务 这 EMA guideline on development and manufacture of synthetic peptides draws that boundary explicitly. Impurities in synthetic peptides are grouped as peptide-related (截断, 删除, 氧化, 脱酰胺化, sequence variants) and non-peptide (抗衡离子, residual solvents and reagents such as TFA). 抗衡离子的身份和含量, 残留三氟乙酸, and water content are separate control items, not part of the peptide-related impurity thresholds. A peptide identity testing package that reports only the peptide-related side leaves three control items unaddressed.

Method conditions are what make an identity result evaluable. 我Q2(R2) frames validation around specificity/selectivity, range, 准确性, and precision, with typical minimum experimental designs of 9 determinations across at least 3 concentration levels for accuracy and 9 determinations over 3 levels or 6 determinations at 100% of test concentration for precision. When assay and impurity share one test, the calibration range runs from the impurity reporting level up to 120% of the assay acceptance criterion.

So what does this mean for a supplier COA? An MS confirmation reported without method conditions cannot be evaluated. You cannot tell whether the method was specific to the sequence, what concentration levels were run, or whether the counterion was measured at all.

纯度评价: 读取百分比后面的数字

A purity percentage without method conditions, a chromatogram, and impurity context is not evaluable. “≥98%” from two suppliers can describe two different measurements, and peptide purity evaluation starts by asking what the number is measuring.

The threshold scale and the advertising scale are not the same scale. Under the Ph. 欧元. general monograph “Substances for Pharmaceutical Use” referenced by the EMA guideline on the development and manufacture of synthetic peptides, peptide-related impurities are reported above 0.1%, identified above 0.5%, and qualified above 1.0%. Those are actions a manufacturer must take as impurity levels rise. A ≥98% HPLC figure is a headline result from one run, on one column, under one set of method conditions.

Typical reported COA practice for lyophilized synthetic peptides clusters around HPLC purity tiers of ≥95%, ≥98%, and ≥99%, intact-mass identity within about ±0.5 Da (or under 5 ppm on high-resolution instruments), water content roughly ≤5–8% by Karl Fischer, and retest or expiry windows of 12–24 months at −20 °C, 如中所述 supplier COA-reading guides. These are industry-typical values that vary by supplier. They are not acceptance criteria, and treating them as such is the pitfall: a single percentage read as a purity specification.

供应商资质: 对证据进行评分, 不是声誉

Peptide supplier qualification is an evidence-scoring exercise, not a reputation check. The criteria are the same ones the earlier sections defined: 身份, 纯度, method conditions, lot linkage, 和变更控制. Score each supplier against those criteria and the choice becomes defensible on paper.

Three supplier-side errors recur. A COA arrives without method conditions, so the purity figure cannot be reproduced. A certificate of origin arrives without lot linkage, so the material cannot be traced to the site that made it. A change-notification clause is absent from the agreement, so a specification change reaches the buyer after the fact.

FDA’s guidance to compounders on knowing your bulk suppliers and its conditions for 503A bulk drug substances both point the same direction: the buyer carries the burden of establishing that the source is acceptable and reliable, a responsibility USP <795> assigns to the designated person. That is the standard to score against.

先进的: 当内部验证与供应商 COA 不一致时

If you already run in-house verification, this is the record that decides whether the disagreement is a supplier problem or a method problem. The distinction matters because the two failures route to different owners, and treating a method problem as a supplier problem burns a relationship you may need later.

Re-check method conditions and system suitability before you re-test the lot. Column chemistry, mobile phase composition, 坡度, and detector settings all shift results, and a system suitability failure invalidates the run regardless of what the sample contains. Only after your system passes should you re-inject the lot.

Then compare your method against the supplier’s stated method. The ICH Q2(R2) guideline frames validation around specificity, 准确性, 精确, and range, so if your method and the supplier’s differ on any of those characteristics, the two numbers are not measuring the same thing. 抗衡离子的身份和含量, 残留三氟乙酸, and water content are separate control items, and a gap in any of them can explain a purity delta without either party being wrong.

Document the discrepancy with both data sets attached, then route the outcome through your designated person.

The caution is a prerequisite, not a comment on your competence: this sequence only resolves a disagreement if your own method carries documented validation and a defined acceptance criterion. Without those two things, you have two uncalibrated numbers and no basis for deciding which one to trust.

入门: 前三步

The lowest-friction first action is not writing a new SOP. It is classifying every peptide lot you already hold, because the classification determines which documentation obligations apply at all. A lot that lands in the research-grade category carries a different evidence burden than one headed toward compounding-related use, and you cannot know which records are missing until each lot has a category.

步 1: Classify the inventory. Run every existing lot through the classification table and mark its category. This is a desk exercise. It needs no supplier contact, no legal review, and no new template.

步 2: Assemble what exists. For each lot in the compounding-related category, pull the certificate of analysis, the certificate of origin, and the lot linkage into one location, then mark what is absent. Missing items are findings, not failures. You are building the gap list that tells you where to spend effort.

步 3: Score your suppliers. Take your top two or three suppliers through the qualification matrix, and open a change-notification conversation with any that score weakly. Suppliers who cannot answer a notification question in writing are telling you something useful before a quality question escalates.

The hesitation worth naming directly: this is not a compliance project that requires legal sign-off before it can begin. Classification and evidence-gathering are internal, reversible, and produce no external commitment. Start with the inventory, and let the gaps tell you what USP 795 和 797 peptide documentation work actually remains.

常见问题解答

What does the difference between USP <795> 和 <797> mean for a peptide purchase?

It decides which documentation standard your record has to satisfy, not which supplier you buy from. 美国药典 <795> governs nonsterile compounded preparations and USP <797> governs sterile ones, so the same peptide can sit under either chapter depending on the physical form and handling route it is destined for. Buyers should confirm the intended use category first, because that choice determines the identity, 纯度, and handling evidence the record must carry. 多肽生产

COA ≥98% 的研究级肽可以用作复合成分吗?

不. A ≥98% purity figure describes the material’s assay result, not its suitability for a preparation, and research-grade supply is not manufactured or documented against compounding standards. The advertised percentage also says nothing about which impurities were detected, how they were identified, or whether they were qualified against a limit. Treat the research-grade label as a supply category, not as a compliance status.

有效的分析证书必须包含什么?

A defensible COA names the specific lot, the analytical method used for each test, the acceptance criterion, and the measured result, so a reviewer can trace the number back to the method that produced it. A document that reports a single purity percentage with no method conditions, no lot identifier, and no raw output is a summary, 不是证据. The record should also state who performed the testing and when.

身份测试和纯度测试是同一个测试吗?

不. Identity testing confirms the molecule is what the label says it is; purity testing measures how much of everything else is present. A sample can pass an identity check and still fail on impurity profile, which is why the two belong in separate rows of the documentation set rather than collapsed into one line.

供应商资格认证应多久重复一次?

Requalification should be triggered by change, not by the calendar alone. A new lot, a revised method, a relocated manufacturing site, or a change in the supplier’s own specification all warrant a fresh review of the evidence. A periodic re-check adds value mainly as a backstop for suppliers whose change notifications are unreliable.

仅靠 RUO 标签就能解决问题吗?

不. The research-use-only designation limits the claims a supplier can make about the material; it does not by itself establish or exclude anything about the buyer’s intended use. What settles the question is the documented chain from the material’s stated grade through to the use category the buyer has already identified.

杂质报告如何进行, 鉴别, 资格门槛与宣传的纯度等级不同?

They answer different questions. Reporting thresholds set the level at which an impurity must appear in the record at all; identification thresholds set the level at which its structure must be determined; qualification thresholds set the level at which it must be shown to be safe at the intended exposure. An advertised purity tier is a single headline number and carries none of that structure, which is why a high percentage and a clean impurity profile are not interchangeable claims.

结论

美国药典 795 和 797 peptide documentation is decided by the preparation and its intended use, not by the paperwork you assemble afterward: the category comes first, and every record you keep has to answer to it. That is the spine this guide has followed throughout. You draw the line between research-grade supply, development support, and compounding-related use. You build a record set that can survive a question. You confirm the molecule by identity testing rather than by mass alone, you read the purity percentage against the method that produced it, and you score supplier evidence on what it actually shows instead of on reputation. When in-house verification disagrees with a supplier COA, the disagreement itself is information, and it belongs in the file.

What moves is the enforcement posture, not the compendial requirements. Inspection attention, documentation expectations, and the questions buyers ask tend to shift faster than the standards they are built on, so a record set that is defensible today is the one that stays defensible when the scrutiny changes. Build for the requirement, not for the current mood.

This article is an educational overview. It is not legal, regulatory, or compliance advice, and it does not describe compounding services or compounded preparations. Consult qualified regulatory counsel or a licensed pharmacist for determinations about your own operations. MOL Changes operates in the peptide and life-sciences supply space, so treat this as a neutral commercial-interest disclosure rather than an independent assessment.

后续步骤

If you want to compare your current documentation set against the five-part framework above, request the analytical documentation package or speak with a technical specialist about the records that accompany your materials.