201802 WHO制剂文件中常见缺陷

COMMON DEFICIENCIES IN

FINISHED PHARMACEUTICAL PRODUCT (FPP) DOSSIERS

制剂文件中常见缺陷

Additional guidance for manufacturers

给生产商的附加指南

This note identifies the most common quality related deficiencies in recently assessed dossiers in the Prequalification ofMedicines Programme (PQTm), and provides additional guidance to help manufacturers address these deficiencies infuture submissi** to PQTm. The discussi** are based on solid oral dosage forms, but may also be useful for otherdosage forms.

本注意事项写出了最近药品预确认项目（PQTm）文件审评中发现的最常见质量相关缺陷，提供更多指南帮助生产商解决未来给PQTm的申报中的这些缺陷。本文讨论是基于口服固体制剂，但可能对其它剂型也有用。

The Active Pharmaceutical Ingredient (API) supplier’s vs the FPP manufacturer’s API specificati**API供应商VS制剂生产商的API质量标准

Most of the reviewed dossiers had issues concerning the FPP manufacturer’s specification in relation to the APImanufacturer’s specification.

大部分审评的文件中有关于FPP生产商的质量标准与API生产商的质量标准问题。

In general, the FPP manufacturer’s specification for a given API should be in agreement with the API manufacturer’sspecification as accepted by PQTm (API-PQ or APIMF procedures) or EDQM (CEP).

一般来说，FPP生产商对某API的质量标准应与该API生产商的质量标准保持一致，这是PQTm（API预认证或APIMF程序）或EDQM（CEP）所接受的。

This means: 这意味着

-          All test parameters and corresponding acceptance criteria (same limits or tighter) included in the APImanufacturer’s specification should generally be in the FPP manufacturer’s specification (excepti** might includethe tests chosen for identity, difference in the type of method with different associated limits e.g. HPLC assayreplacing titration assay, etc.); any differences compared to the API supplier’s specificati** should be justified.

-          API生产商的质量标准中所包括的所有检测项目和对应的可接受标准（相同限度或更严格）应出现在FPP生产商的质量标准中（除外情况可以包括所选择的鉴定测试，由于方法类别不同而有不同的限度，例如，HPLC含量替代滴定含量等）；所有不与API供应商质量标准的差异均应进行论证。

-          Analytical methods for purity and assay should be the same as those validated by the API manufacturer or elseshould be fully validated and dem**trated as at least equivalent to the methods validated by the API manufacturer, orequivalent to the compendial method if compendial standard is claimed.

-          纯度和含量分析方法应与API生产商验证过的方法相同，或经过全面验证并证明至少等同于API生产商验证过的方法，或等同于药典方法（如声明使用药典标准）。

-          The FPP manufacturer’s specification should not include skip testing or periodic testing schedule except when sucha proposal has been accepted via APIMF assessment (i.e. is in the accepted API manufacturer’s specification). Insuch cases, a footnote identifying the skip tested parameters can be included. However, frequency of testing shouldnot normally be stated in the specification since this is to be established after prequalification based on trend data **everal batches tested by the FPP manufacturer, and based on the principles of vendor qualification.

-          FPP生产商的质量标准不应包括周期测试或定期测试计划，此方案已通过APIMF评估被接受者除外（即，在已被接受的API生产商的质量标准里）。在此情形下，应有脚注说明周期测试项目是可以包括在其中的。但是，测试频次一般不应在质量标准中声明，因为这要在预确认之后基于FPP生产商所测几批的趋势数据，以及供应商确定的原则来确定。

Additional user specific tests such as polymorph identity and particle size distribution (PSD), as appropriate, should also beincluded.

附加的用户特定测试如晶型鉴别和粒径分布（PSD）也应包括在其中（适当时）。

Control of polymorph identity and PSD 多晶型鉴别和PSD控制

In 35% of the reviewed dossiers missing or inadequate control of polymorph identity and/or PSD were noted.

在35%审核的文件中多晶型鉴别和/或PSD控制缺失或不充分。

Determination of BCS (The Biopharmaceutical Classification System) solubility is an integral part of the API data requiredin the dossier. All APIs of BCS low solubility[1] should be controlled for polymorph identity (when polymorphism is a factor)and should always include controls for PSD:

BCS（生物药品分类系统）溶解度的确定是文件中所需API数据不可分割的一部分。所有具有BCS低溶解度的API均应控制其晶型鉴别（如果多晶型是一个因素）并应总是包括PSD控制。

-        Unless otherwise justified using ICH Q6A decision tree # 4, a routine polymorph ID control should be included. Adiscussion should be included in section 3.2.S.3 of the dossier to justify exclusion of polymorph identity for BCS lowsolubility APIs.

-        除使用ICH Q6A决策树#4另有论证者外，应包括常规的晶型ID控制。应在文档的3.2.S.3部分包括一个讨论来论证BCS低溶解度API的晶型鉴别除外的情况。

-        The specification should include appropriate control for PSD. For this, the API lot used in the manufacturing of thebiobatch should be tested by laser diffraction and profiles established.

-        质量标准应包括适当的PSD控制。为此，用于生物批生产的API批次应采用激光衍射法测试并建立其概况。

Except when a limit for d90 of NMT 10?m is proposed, based on the biobatch API lot, the QC limit in all cases shouldinclude limits for d10, d50 (as a range) and d90, based on the results of the API lot(s) used in the biobatch.

除所拟限度为d90NMT10?m者外，基于生物批API批次，QC在所有情形下的限度均应包括d10、d50（作为一个范围）和d90的限度，基于生物批所用API批准结果确定。

In addition, APIs determined to be critically insoluble[2] should be monitored for change in polymorph identity, as required(see above) and PSD during stability studies. That is, the proposed retest period/shelf life for the API should also besupported by PSD and polymorph identity results.

另外，确定为极难溶的API应监测稳定性研究中PSD和晶型鉴别（如需要，见上）的变化。也就是说，所拟API复测期/货架期也应有PSD和晶型鉴别结果的支持。

Furthermore, these parameters should be included as retest parameters.

另外，这些项目均应包括作为复测项目。

Retest/shelf life parameters should be indicated in the API specification of the FPP manufacturer.

复测/货架期参数应包括在FPP生产商的API质量标准里。

Control of related substances in APIs and FPPs API和FPP中有关物质的控制

Other commonly identified issues (mostly with regard to FPP specificati** but also in relation to the FPP manufacturer’sAPI specificati**) were proposals for unacceptably wide limits for unspecified impurities. Some applicants incorrectlyassume that pharmacopoeial monograph limits for any unspecified impurities, which in some cases may be higher than theICH Q3A/Q3B identification thresholds, are also acceptable regulatory limits.

其它常见问题（通常是与FPP质量标准有关，但也与FPP生产商的API质量标准有关）是非特定杂质限度过宽不能接受。有些申报人错误地假定药典各论中任意非特定杂质的限度也是可接受的注册限度，这些标准有时可能会高于ICH Q3A/Q3B的鉴别限。

Except for some artemisinin derivative or fermentation APIs or FPPs containing such APIs, limits for any unspecifiedimpurities should correspond to the ICH Q3A/Q3B identification threshold. If this limit cannot be met for a particularunidentified impurity, the impurity should be identified and controlled as follows: All identified impurities should be controlledto the ICH Q3A/Q3B qualification threshold. For any impurity that cannot be controlled to this limit, a higher limit should bequalified according to the guidance provided in section 3.2.S.3.2 of the generic guide (Annex 4, TRS970).

除某些青蒿素衍生物或发酵类API，或含有此类API的FPP外，所有非特定杂质的限度均应符合ICH Q3A/Q3B的鉴别阈值。如果此限度不符合特殊的未鉴别杂质，则该杂质应进行鉴别并按以下要求受控：所有鉴别的杂质均应控制至ICH Q3A/Q3B确认限。所有不能控制至此限度的杂质，应依通用指南（TRS970B附录4）在3.2.S.3.2部分所提供的指南进行确认。

There should be a key below the API/FPP specificati** that identifies specified impurities, at a minimum by chemicalname. For any specified impurities that are also named in an officially recognized compendial monograph, this name shouldbe included, e.g.:

在API/FPP质量标准下应有线索识别特定杂质，至少通过化学名。所有亦被列名在官方认可的药典各论中的特定杂质，应包括此名称，如：

Impurity A: <chemical name> = USP RC B = Ph.Eur. impurity H

杂质A：<化学名>=USP RC B=欧洲药典杂质H

Granulation processes 制粒工艺

Inadequate or poorly defined end point for wet granulation process affected about 50% of the reviewed dossiers.

湿法制粒工艺终点不充分或定义不足影响了约50%审评的文件。

Regardless of the manufacturing process (direct compression, wet/dry granulation, etc.), the manufacturing process andcontrols proposed for production batches (detailed in the blank Batch Manufacturing Record, BMR) should be in agreementwith the process and controls used in the manufacturing of the pivotal clinical batch(es) or biobatch. Differences, if any,should not normally be beyond those that can be attributed to scale differences, for example, slight changes to blendingparameters.

无论采用何种生产工艺（直压、湿法/干法制粒等），为生产批次所拟的生产工艺和控制（在空白批生产记录BMR中详细载明）均应与关键临床批次或生物批次生产中所用的工艺和控制一致。如有任何差异，一般均不应超出规模差异所导致的差异，例如，混合参数有较小变更。

To this end, differences should be highlighted and discussed in order to avoid further questi** from the assessors. For thispurpose, the table provided in section 2.3.R.2.1 of the QOS-PD should be used. Furthermore, statements such as “stopgranulation when required c**istency is achieved” are not acceptable as a means of end-point determination. Asmentioned above, processes should be kept as close as possible to the process that was used for the clinical/biobatch.

由此，为了避免审评人员提出更多问题，应重点说明这些差异并进行讨论。为此，应使用QOS-PD的第2.3.R.2.1部分所提供的表格。还有，如“达到所需一致性时停止制粒”类的声明是不会被接受作为终点判定的方法的。如上所述，工艺应与临床/生物批所用工艺尽可能保持一致。

Manufacturers are advised to use the following approaches:

建议生产商使用以下方法：

1.       When the proposed production batch size is the same as the biobatch size当所拟生产批量与生物批批量相同时

The blank BMR for production batches should fully reflect the actual process parameters used for the biobatch, includingmixing/addition amounts/times, speeds, amperage targets/limits etc. For example, if additional water and/or mixing wereused for the biobatch, then the blank BMR should also require the same without statements such as “if required addadditional water and mix…”.

Similarly, if no additional water and/or mixing process were used then the blank BMR should not require the use of additionalliquid or mixing.

生产批次的空白BMR应全面反映生物批次所用的实际生产工艺参数，包括混合/加料数量/时长、速度、电流目标/限度。例如，如果生物批使用了额外的水和/或混合，则空白BMR也应有相同要求，不用声明如“需要时加额外的水和混合……”。类似的，如果没有使用额外的水和/或混合工艺，则空白BMR不应要求使用额外的液体或混合。

2.       When the proposed production batch size is larger than the biobatch size当所拟生产批量大于生物批批量时

The blank BMR instructi** for manufacture of the production batches should be kept as close as possible to the processused for the biobatch. It is acknowledged that slight changes or a provision for use of additional water or mixing may berequired to account for the scale difference. In this case, in addition to fully defining the parameters, manufacturers arerequested to include chopper/impeller ampere or torque readings as a means of control, with provisional limits defined.

生产批生产所用空白BMR指令应尽可能保持与生物批所用工艺接近。已知考虑到批量规模差异，可能会有轻微变更，或需要使用额外的水或混合操作。在此情形下，除了全面定义参数外，还要求生产商包括桨片/轮片电流或扭力读数作为控制方法，同时定义有条件的限度。

The provisional limits should as much as possible reflect the observati** recorded for the biobatch. The table provided insection 2.3.R.1.2 of the QOS-PD should be used for this purpose. Note that it is not acceptable to use the processvalidation of production batches as a means of establishing the parameters for scaled-up batches. Therefore, provisionalsettings and limits must be included in the blank BMR and it is understood that these will be validated and finalized duringthe process validation for production batches.

有条件的限度应尽可能反映生物批所记录的观察结果。此时应使用QOS-PD第2.3.R.1.2部分所提供的表格。注意，使用生产批次的工艺验证作为建立放大批次参数的方式是不会被接受的。因此，有条件地设置和限度必须包括在空白BMR中，要了解在生产批次工艺验证中需要对其进行验证并最终确定。

Similarly, compression machine speeds should be defined in the BMR as validated or to be validated with ranges reflectedin the process validation protocol. To this end, process validation protocols should also include a provision for machinespeed challenge runs.

类似地，压片机速度应在BMR中定义为已验证或待验证，配有工艺验证方案中所反映的范围。为此，工艺验证方案还应包括机器速度挑战轮次条款。

Hold times 存贮时长

Nearly all reviewed dossiers had one or more hold time related deficiencies.

几乎所有审评的文件都有一个或几个与存贮时长有关的缺陷。

Applicants should refer to the WHO guideline for General guidance on hold-time studies (Annex 3, TRS992). For purposesof prequalification, individual hold time of 30 days for intermediates such as final blend, core tablets and coated tablets canbe proposed without submission of supporting data. Otherwise, the dossier should include data on at least one batch of theproduct. In addition, a cumulative hold time of not more than 90 days from dispensing to primary packaging should be set.All hold times should be specified in BMRs and section 2.3.P.3.4 of the QIS.

申报人应参照WHO指南“保存时间研究通用指南”（RTS992附录3）。在预确认中，中间体比如最终混合物、片芯和包衣片单个保存时长30天，不需要提交支持数据即可拟定。否则，文件中应包括至少一批该产品的数据。此外，从配制至内包累积存贮时长应设定不超过90天。所有存贮时长均应在BMR和QIS的2.2.P.3.4部分中指定。

Proposals for a cumulative hold time of more than 90 days will need to be supported by accelerated and long-term stabilitydata on the final product as packaged with tablets that have been held for the proposed cumulative hold time period, or acommitment should be provided to put such a batch in stability studies when manufactured. The FPP release specificati**should be the measure for hold time studies, rather than the shelf life specificati**.

累积存贮时长超过90天的提议需要有最终成品与片剂包装后保存所拟累计存贮时长的加速和长期稳定性研究数据支持，或者提交一份承诺会在生产时将这样一个批次放入稳定性研究中。保存时长研究所用测试应该是FPP放行标准，而不是货架期质量标准。

Dissolution profiles and QC testing limits 溶出度概况和QC测试限度

Missing dissolution profiles and/or unacceptable limits affected nearly all the reviewed dossiers. Multimedia dissolution profile data on the biobatch is a critical reference data set that is used to support routine dissolution limits, successful processvalidation and future changes. Manufacturers are therefore requested to profile the biobatch in the standard three BCS pHmedia and in the routine              medium, if the latter is different from the standard media. Dissolution profiles should begenerated using the following dissolution conditi**.

几乎所有接受审评的文件都受至缺失溶出度概况和/或不可接受的限度的影响。生物批多介质溶出度概况数据是一个关键的对照数据，用以支持常规溶出限度、成功的工艺验证和未来的变更。因此生产商应获得生物批在标准的3个BCS pH介质和常规介质中的概况（如果后者与标准介质不同的话）。溶出度概况应使用以下溶出条件生成：

Media: pH 1.2, 4.5, 6.8 buffer (without surfactant), 900 ml or less, and in the routine medium (if different from the BCSmedia), 900 ml or less (unless justified).

介质：pH 1.2, 4.5, 6.8缓冲液（没有表面活性剂），900ml或更少，在常规介质中（如果不同于BCS介质），900ml或更少（另有论证者除外）。

Speed: Apparatus 1: 100rpm or less; Apparatus 2: 75rpm or less.

速度：仪器1:100rpm或更低；仪器2：2:75rpm或更低

Time points: 5, 10, 15, 20, 30 for very rapid release products; 10, 15, 20, 30 and 45 minutes for rapid release products;otherwise 10, 15, 30, 45 minutes, etc.

时间点：非常快速释放的药品5, 10, 15, 20, 30分钟，快速释放的10, 15, 20, 30 和45分钟，否则10, 15, 30, 45分钟等。

The applicants should note that the 5-minute time point is critical for very rapidly releasing products, since this time pointmay be necessary to meet the three time points required for calculation of f2 values for future comparis** (for example,when future batches manufactured to support certain changes fail to achieve a release of more than 85% in 15 minutes asachieved with the biobatch).

申报人应注意5分钟时间点对非常快速释放药品来说是至关重要的，因为该时间点可能需符合3个时间点，在未来比较中要计算f2值（例如，如果未来所生产的批准用于支持特定的变更，不能如生物批一样达到释放度在15分钟内超过85%）。

Appendix 1 of the main generic guideline (Annex 4, TRS 970) should be c**ulted.

应参考主要仿制药指南附录1（TRS970附录4）。

Dissolution limits for QC testing should be established based on the performance of the biobatch, irrespective of limitsspecified in a pharmacopoeial monograph. The approach described in the EMA’s reflection paper on the dissoluti**pecification for generic oral immediate release products may be used. The acceptance criterion set based on the biobatchbehaviour should then be used for stability studies. To this end, applicants are reminded to report stability results asindividual (at least the range of individual values) and average values. In the case where the QC limit established based onthe above approach could not be met during shelf life, stability samples can be dissolution profiled and similarity with thehistorical profile of the biobatch can be dem**trated. In this manner, an alternative QC limit based on the observed stabilityresults could be proposed.

QC测试的溶出度限度应基于生物批的表现来建立，而不管药典各论中所述的限度如何。可以使用EMA概念文件中所述的口服即释仿制药品溶出度质量标准。基于生物批表现所设定的可接受标准可用于稳定性试验。为此，提醒申报人将稳定性结果作为单值（至少是单值的范围）和平均值报告。如果在货架期内不能符合基于上述方法建立的QC限度，可使用稳定性样品测定溶出度概况，证明与生物批的历史概况具有相似性。采用此方法时，可以提议基于观察到的稳定性结果建立替代的QC限度。

For profiles that do not achieve a mean release of 75% or more in 60 minutes, a two-tier limit is appropriate. For example,first tier limit at about 40% mean (with RSD ≤ 10%) release point, and a second-tier limit at about 75% or more releasepoint, may be c**idered.

60分钟内未达到平均释放75%或以上的，2层限度较为适当。例如，可以考虑第1层限度为约40%平均（RSD ≤ 10%）释放点，第2层限度在75%或以上释放点。

Note: Routine release and shelf life limits for products supported by BCS-based biowaiver should be as follows:

注：受基于BCS豁免生物等效支持的药品的常规放行和货架期限度应如下：

-        FPPs containing BCS class 3 API (those that need to dem**trate very rapid release): NLT 80% (Q) in 15 minutes

-        含有BCS 3类API的FPP（需要证明非常快速释放的）：15分钟内NLT80%（Q）

-        FPPs containing BCS class 1 API (those that need to dem**trate at least rapid release): NLT 80% (Q) in 30 minutesor tighter depending on the performance of the biowaiver test batch.

-        含有BCS 1类API的FPP（需要证明至少为快速释放的）：30分钟内NLT80%或更严，取决于生物豁免测试批次的表现

Control of moisture content in final blend and the FPP for solid oral products固体口服药品中最终混合物和FPP的水分控制

Moisture content has several implicati** on manufacturability (e.g. smooth compression running) and on stability of theproduct. It should also be noted that the impact on product stability is not limited to facilitating potential degradation ormicrobial growth, but may also affect the product release characteristics. To this end, the following routine controls arerequired:

水份对于可制造性和药品稳定性有几层含义（例如，光滑压片运行）。还应注意其对药品稳定性的影响并不仅限于促进潜在的降解或微生物生长，也可能影响药品放行特性。因此，要求进行以下常规控制：

-          Final blend, regardless of the method of manufacture (wet, dry granulation or blend for direct compression), should becontrolled for optimal moisture content or loss on drying (LOD). The intermediate product specification for the finalblend should include a limit (as a range), set based on levels recorded for the primary and/or process validationbatches.

-          最终混合物，不论采用何种生产方法（湿法干法制粒或混合直压），均应控制在最佳水分含量或干燥失重（LOD）。最终混合物的中间体质量标准应包括一个限度（作为范围），基于最初和/或工艺验证批次所记录的水平进行设定。

-          The final product should also be tested for moisture or LOD at release and during stability studies. The test should beincluded in the release and shelf life specificati**. The limit proposed for the shelf life specification should be basedon the results observed during stability studies.

-          最终产品在放行时和稳定性研究中也应测试其水分或LOD。检测应包括在放行和货架期质量标准中。货架期质量标准所拟限度应基于稳定性研究中所观察到的结果。

Process validation 工艺验证

About 60% of the reviewed applicati** had protocols deficient with regard to one or more of the following:

所审评的申报资料中有约60%方案有以下一个或更多缺陷：

a)      The protocol or signed commitment did not specify the type of validation as prospective;

方案或签字的承诺函未说明验证类型是前验证；

b)      Missing provision for monitoring/evaluation of process parameters used for the validation batches;

缺失验证批次所用工艺参数监测/评估条款；

c)      Missing provision for compression machine speed range validation;

缺失压片机速度范围验证条款；

d)      Missing provision for dissolution profile comparison of each validation batch with the historical biobatch profileusing the QC medium.

缺失使用QC介质的每个验证批次与历史生物等效批次溶出度概况比较内容；

PQTm requires that process validation will be satisfactorily completed before any batch is released for supply (i.e.prospective validation). The process validation commitment and the proposed validation protocol should specify thatvalidation will be completed prospectively.

PQTm要求工艺验证在所批次放行销售之前圆满完成（即前验证）。工艺验证承诺和所拟的验证方案应说明验证将以前瞻方式完成。

Process validation protocols should also include a provision for monitoring process parameters for all critical steps,regardless of whether sampling occurs at each step. For example, although sampling at the wet granulation stage may notbe required, the granulation parameters applied for the validation batches should be summarized and discussed in theprotocol and the results should be captured in the validation report. Such monitoring may help in finalizing the granulationparameters for subsequent production batches.

工艺验证方法还应包括监测所有关键步骤中工艺参数的条款，无论是否在每步都有取样发生。例如，尽管在湿法制粒步骤可能并不要求取样，但验证批次所用的制粒参数应在方案里进行总结和讨论，结果则应录入验证报告。此类监测可能会有助于最终确定后续生产批次的制粒参数。

Validation of compression/encapsulation stages should include robustness studies for the proposed machine speed range.Depending on the level of optimization performed during development pharmaceutics, a provision for validation of theproposed range for press pressure (tablet hardness range) should also be included.

压片/胶囊灌装步骤的验证应包括所拟设备速度范围的耐受性研究。依开发药品过程中所做的优化程度不同，还应包括所拟压片压力范围（片剂硬度范围）的验证条款。

A critical component of the evaluation of process validation data is comparison of the dissolution profile generated on thevalidation batches against the historical profile established for the biobatch. The validation batches should be profiled using12 units in the routine medium (see also “Dissolution profiles and QC testing limits” above). The three validation batchesshould dem**trate similar profiles compared to the biobatch (i.e., meet similarity acceptance criteria) for the processvalidation to be deemed satisfactorily completed. Non-similar results should be investigated and immediately reported toWHO.

工艺验证数据评估的关键部分是将验证批次的溶出度概况与生物等效批次所建立的历史概况进行比较。验证批次应在常规介质中使用12剂测得其概况（也参见上述“溶出度概况与QC测试限度”）。3个验证批次应证明与生物等效批具有相似的概况（即符合相似度可接受标准），方可认为工艺验证圆满完成。非相似结果则应进行调查 ，并立即向WHO报告。

For questi** please refer to Dr M Stahl at stahlm@who.int

更多问题请通过上述邮箱咨询。

---

[1] BCS low solubility APIs dem**trate dose solubility volume > 250 mL at one or more points across the physiological pH range (pH 1.2, 4.5, 6.8, plus thepKa if within the range of pH 1.2 to 6.8). BCS低溶解度API证明剂溶解体积在生理pH范围中一个或多个点>250ml (pH 1.2, 4.5, 6.8, 加 pKa 如果在 pH 1.2 to 6.8范围内)。

[2] Critically insoluble APIs dem**trate BCS low solubility (dose solubility volume > 250 mL) across the physiological pH range and require use ofsurfactant for dissolution testing of the FPP, e.g. lumefantrine or efavirenz. 极难溶API证明在生理pH范围和FPP溶出度测试需使用表面活性剂 时BCS低溶解度（剂溶解体积<250ml），例如本芴醇或依法韦仑。