扫一扫 添加小助手
服务热线
13818320332
扫一扫 关注我们
8.0 maintenance of Validated state
8.0 验证状态维护
A key part of the validation life cycle for any system is maintenance of the validated state. This section deals with activities after the cleaning process has been designed and developed and after the formal validation protocols have been successfully executed. This is critical for cleaning validation, because a lapse in the validated state has the potential to adversely impact the quality, safety and purity of subsequent batches of the same or different products. The main tools for ensuring the continued maintenance of the validated state are change control, risk-based periodic monitoring and data trending review. Additionally, training and retraining are important areas of control for manual cleaning processes, as they are the primary mechanisms for controlling the cleaning cycle. In each of these three areas, knowledge of the design space (see Section 3.8) should be applied.
验证状态维护是任何系统的验证生命周期的一个关键组成部分。这部分内容是指清洁过程已经设计和开发 并且正式验证方案已被成功实施后的活动。维护对清洁验证是至关重要的， 已验证状态的缺失可能会对随后生产的相同或不同产品的质量、安全和纯度带来不良影响。确保已验证状 态获得持续维护的主要手段是变更控制、基于风险的定期监控及数据趋势分析。另外，培训和再培训是人 工清洁过程的重要控制方式，因为它们是控制清洁周期的主要机制。以上三个方面，均应运用设计空间的 知识（参见 3.8 部分） 。
8.1 关键参数控制
8.1 Critical Parameter Control
In controlling a validated cleaning process, it is of utmost importance to understand the critical parameters used to control the cleaning process. Typically these include cleaning agent concentration, temperature, flow rate and times for all processing steps. During the design phase, an appropriate level of understanding of the process and its variability should be obtained to design a cleaning process capable of addressing this inherent variability. Once the process is well defined, there are a variety of control strategies that may be used. One control strategy is to set minimum and/or maximum values for each of the key parameters during a cleaning cycle. In this model, each of the steps of the cycle has a defined range that must be monitored and maintained during each execution of the cleaning cycle, and each parameter does not vary outside that range. This approach has an advantage in that that it is straightforward to implement and control.
在对一个已验证清洁程序的控制中，明白用于控制清洁过程的关键参数是至关重要的。通常，包括清洁剂 浓度、温度、流量及所有清洁步骤的时间。在设计阶段，应对清洁程序及其可变性有适当程度的理解，以 设计一个能够解决这个内在可变性的清洁程序。可能会使用各种各样的控制策略来制定清洁程序。控制策 略一，在一个清洁周期过程中为每个关键参数设置最小和/或最大值。在这种模式中，清洁周期的每个步骤 都有一个明确的范围，在每次清洁周期的实施过程中必须进行监控和维护，且每个参数的变化不能超出规 定的范围。这种做法的优势在于它可以直接进行实施和控制。
8.2 周期反馈控制
8.2 Control by Cycle Feedback
Another control strategy is to use analytical feedback to determine cycle step length. For example, the final rinse for a CIP cycle may be continued until the rinse conductivity indicates adequate completion of the rinsing step. This approach has elements of Process Analytical Technology (PAT) (see Section 11.8) to ensure the cleaning cycle is appropriately controlled. In the example given, other control parameters, such as temperature and cleaning agent concentration, are maintained in their appropriate ranges. Furthermore, it must be ensured in the design/development steps that conductivity is adequate to measure process step completion. Based on the initial validation, other analytical results (e.g., TOC) may be deemed more indicative of cycle step completion. However, since the cleaning of biotechnology products is accomplished by highly alkaline and/or acidic cleaning agents, conductivity is usually an appropriate indicator of completion of the rinsing step. If one ensures minimum and maximum values are set for other critical parameters and uses these values in concert with control of the rinse time based on analytical feedback, this approach will yield appropriate control of the cleaning cycle.
另一个控制策略是通过分析反馈确定周期步骤的时长。例如，一个 CIP（在线清洗）周期的最终淋洗应持续 至淋洗电导率显示淋洗步骤彻底完成。这种方法包含了 PAT（即过程分析技术，参见 11.8 部分）的内容， 可以确保清洁周期已被适当控制。在上面的例子中，其他控制参数如温度、清洁剂浓度在各自适当的设定 范围内被维护。此外，在设计/开发阶段，必须确保电导率足以测定过程步骤的完成。依据最初的验证，其 他分析结果（例如 TOC）可能会更好的指示周期步骤的完成。然而，因为生物制品的清洁使用了强酸/碱， 所以电导率通常是指示淋洗步骤完成的适当指标。如果可以确保其他关键参数已设置了最大和最小值，并 且根据分析反馈使用这些与淋洗时间相一致的值，那么这个方法将会对清洁周期产生适当控制。
8.3 过程报警
8.3 Process alarms
Another key component of applying design space to cleaning processes is alarming of critical parameters. In an automated CIP cycle, alarms may be based on a variety of parameters, such as temperature of the wash and rinse solutions, conductivity of the recirculating wash solution, pressure at the spray device, flow though various circuits, and conductivity of the final rinse. There are a variety of approaches to cleaning the equipment on which an alarm occurred. In all cases, the cause of the alarm should be investigated. One strategy is that on specified alarm conditions, the cleaning cycle may be restarted. For example, if inadequate cleaning agent concentration occurred (as indicated by an alarm on the wash cycle conductivity), the cleaning cycle can be restarted from the beginning after appropriate actions are taken to ensure the alarm does not reoccur. This is a conservative approach and ensures a complete cleaning cycle is performed, but care must be taken that alarms are noted and trended to ensure cycle performance is not trending towards being ineffective and to better correct repetitive problems. Alternately, the step in which the alarm occurs may be restarted. This approach strikes a balance between ensuring cycle performance and minimizing cleaning time, as the entire cycle does nothave to be repeated. Automated alarming is generally not done in manual cleaning operations. However, if cleaning agent dilution is confirmed by conductivity, or cleaning agent temperature is confirmed by temperature measurement, measurements outside the specified range can serve as an “alarm.” In addition, for all cleaning processes, visual inspection after cleaning can serve as an “alarm.” In all cases, it must be ensured that cycles performed during validation are not “best case” due to alarm conditions. For example, if equipment is soiled, and during the initial validation of the cleaning cycle alarms occur that result in multiple rinse steps being completed, this cycle is no longer representative or worst case, but best case.
设计空间应用于清洁程序的另一个重要方面是对关键参数进行报警。在自动 CIP 周期中，报警来自多个参 数，如清洗液和淋洗液的温度、再循环清洗液的电导率、喷雾装置的压力、多种环路的流动及最终淋洗的 电导率。有不同方法清洁出现报警的设备。无论在何种情况下，应调查报警出现的原因。一个策略是在预先报警的情况下，清洁周期可以被再次启动。例如，如果清洁剂浓度不足（清洗周期电导率报警提示） ， 这个清洁周期可在采取适当行动确保报警不再出现后开始重新启动。这是传统做法，可以确保进行一个完 整的清洁周期，但必须记录报警并作出趋势，以确保周期的执行是有效的，并能够更好的改正重复出现的 问题。或者说，出现报警的步骤可被再次启动。这种方法可以在确保周期性能和减少清洁时间之间找到平 衡。通常，人工清洁操作未使用自动报警。然而，如果清洁剂的稀释液由电导率确定，或者清洁剂温度由 温度测量确定，规定范围之外的测量可看作是一个“报警”。此外，对于所有清洁程序来说，清洁后目视检 查可看作是一个“报警”。不管在什么情况下，必须确保验证过程中因报警出现的多次循环不是“最好情况”。例如，如果设备有污渍，并且在清洁周期的最初验证阶段因报警出现了多个待完成的淋洗步骤，那么这个 循环就不是有代表性 的或是最差情况，而是最好情况。
8.4 变更控制
8.4 Change Control
A robust change control system is critical to ensuring maintenance of the validated state for cleaning processes. The change control system must cover all key parameters and components of the cleaning system to ensure that all changes with a potential to impact maintenance of the validated state are evaluated. This includes not only changes in the cleaning process, but also changes in equipment and changes in the manufacturing process (for example, a change in temperature in a manufacturing process) which might affect the performance of the validated cleaning process. Quality preapproval and robust tracking of changes are key requirements for this system.
The change control system should provide for a review of each change by an interdisciplinary team. This must include a review of current validation for the equipment being changed, and depending on the nature of the change, may result in laboratory, pilot scale and/or commercial scale evaluations. Significantly major changes may result in the decision that the new cleaning process requires separate validation as a new process. There are some important considerations for designing the test plan to verify changes; review of the design space will assist in this evaluation. First, control parameters must stay within their validated ranges or must be revalidated. For example, if the pump on a CIP skid is validated to deliver water between 5 and 10 liters per minute, and the desired change is to increase the flow rate to 12 liters per minute, new validation testing is required to verify that the pump is capable of delivering the desired flow before validation of the cleaning cycle can occur. Second, the acceptance criteria for analytical methods should remain unchanged from the previous validation unless there is a justified reason for the difference. This is to ensure that changes result in maintenance of the validated state rather than creation of a new state, which may require significant testing to ensure it is still validated. Finally, reduced sample sites and/or fewer analytical methods may be appropriate in many cases to confirm validation based on a change. For example, if the effect of the change is only on bioburden, then it may be appropriate to evaluate only bioburden in studies that evaluate the effects of the change. These differences must be justified in the testing plan/protocol.
一个强健的变更控制系统是确保清洁程序的验证状态得到维护的关键。变更控制系统须包含所有清洁系统 的关键参数和组件，以确保所有可能影响验证状态维护的变更得到评估。这不仅包括清洁程序的变更，也 包括设备变更及生产工艺的变更（例如，生产工艺中温度的变更） ，即可能影响验证清洁程序性能的变更。质量预批准和强健的变更追踪是变更系统的关键。
在变更控制系统中，每个变更须由一个跨学科专家团队进行评估。评估须包含对变更设备同步验证的评估， 依据变更的特性，可能会涉及关于实验室、中试规模和/或商业规模的评估。很明显重大变更可能会做出这样的计划行动，这个新的清洁程序作为一个新的程序需要进行独立验证。设 计清洁验证检测计划有一些重要的考虑因素；设计空间的评估将有助于变更的评估。首先，控制参数必须 在它们各自的验证范围内或必须进行再验证。例如， 如果 CIPskid 上的泵经验证可以以每分钟 5-10 公升的流量送水，并且理想的变更是使流量增加到每分钟 12 公升， 那么就需要新的验证测试验证，在清洁周 期验证前这个泵能够达到理想的流量。第二，分析方法的可接受标准应与之前验证的标准一致，除非有说 明这种差异的合理理由。这里要确保是变更产生了验证状态的维护，而非创建一个新的状态，这可能需要 重要的测试来确保验证的有效。最后，在多数情况下，减少取样点和/或简化分析方法可能适合依据变更确 认验证。例如，变更的影响仅是生物负载，那么在评估变更效果的研究中仅评估生物负载可能是合适的。这些差异必须在测试计划/方案中进行合理说明。
8.5 累计变更的评估
8.5 evaluation of Cumulative Changes
Equally important as a review of each individual change is the review of the cumulative impact of changes on a system. This review must provide evidence that the cleaning cycle meets prescribed requirements. It is possible that many minor changes (each deemed to have no impact on the validated state) could have an impact when considered in total. This review of cumulative changes should take two approaches. First, a documented analysis (i.e., review of the changes and the impact these changes will have on other parts of the process) of the changes should be undertaken on a regular basis. Second, process performance and alarms must be monitored (as discussed above) to ensure continued maintenance of the validated state and system performance.
另一个同样重要的是将每个单独变更的评估看做是所有单个变更对系统的累积影响的评估。评估必须能够 证明清洁周期符合规定要求。当将一些微小变更 （单个不会影响验证状态） 作为整体考虑时，很可能会 产生影响。累积变更的评估应采用两种方式。首先， 定期对所有变更进行文档化的分析 （即，变更评 估及这些变更对程序中其他内容的影响） ；其次，需监控程序性能及报警，正如上面讲述的，以确保对验 证状态和系统性能的持续维护。
8.6 Periodic monitoring 周期监控
Another tool for ensuring maintenance of the validated state is a risk-based periodic monitoring program. A periodic monitoring program may provide analytical data to be trended. In most cases involving automated processes, the data are provided by the CIP equipment itself. For example, data may be generated by the CIP skid on wash solution conductivity, final rinse conductivity, temperatures, times and pressure. In other cases, separate sampling may be established for data collection, such as rinse bioburden or TOC. Visual examination after each cleaning process is another type of periodic monitoring. For routine use, however, visual inspection typically does not involve disassembly of equipment solely for the purpose of that inspection.
另一种为确保验证状态维持的是一个基于风险的定期监控程序。定期监控程序可以提供分析数据形成趋势， 大部分情况下，数据是由 CIP 设备自己提供的。如，数据可能由 CIP 清洗液电导率也产生，最终淋洗水电导 率、温度、时间和压力。其他情况下，可以设立单独的样品数据收集，如清洗液的微生物负荷或 TOC。每次 清洁后的目视检查是另一种形式的定期监控。常规用法，目视检查通常不会为了仅为了检查而拆卸设备。
A documented risk-based approach should be used to optimize compliance in an efficient manner. This could include leveraging family approaches, reduced sample sites and reduced analytical methods. When defining these approaches, the inherent risk associated with a given cleaning process and historical experience/data should be considered. For example, when performing the initial validation on a bioreactor, TOC may be measured via a variety of swab and rinse samples. However, with the proper data analysis, it may be appropriate to measure only rinse TOC during periodic monitoring. Historically it was considered acceptable to perform periodic revalidation on cleaning processes in lieu of periodic monitoring. However, this approach yields a much less robust picture of the state of control of the cleaning process.
应该使用一个文件化的基于风险方法用于优化合规高效的方式。这可能利用家庭模式，减少取样点和减少 分析。当考虑这些方法时，这些固定的清洁过程风险应该基于工艺和历史经验/数据被考虑。如：当低生物 反应器进行初次验证时，TOC 可能通过擦拭或者淋洗水取样，然而，通过适当的数据分析，在日常周期监控 中，可能 TOC 仅仅用润洗水。过去的经验告诉我我可以执行周期性的再验证清洗过程代替定期监测。然而， 这种方法产生一个更具有说服力的清洁过程控制。
8.7 Trending 趋势
Trending of cleaning cycle performance, analytical data from routine monitoring, and alarms are another recommendation to ensure continued cleaning cycle performance. When trending any of these data sets, procedures must be in place to initiate an investigation when adverse trends are observed, even if ineffective cleaning cycles have not occurred. Trending of cleaning cycle performance data is important for identifying potential cleaning cycle issues before they result in ineffective cleaning cycles. For example, a slowly increasing trend in the final rinse conductivity may not be indicative of an ineffective cleaning process. However, such a trend should require an investigation of the cause. In the example given, it may be that the spray device is becoming clogged, in which case it should be cleaned, and appropriate steps should be taken to prevent clogging in the future. On the other hand, it may be a result of a fouled conductivity sensor. Alarm monitoring and trending will indicate cycle failure, though it will not proactively identify potential issues, as is desired. The incidents of all alarms should still be trended to determine if additional process controls are required to reduce the frequency of alarming.
清洗周期的趋势表现,分析数据从日常监视和警报是另一个建议,以确保持续的清洁周期性能。当趋势这些数 据集、程序必须到位立案调查发现不良趋势时,即使没有发生无效的清洗周期。趋势的清洗周期性能数据是 很重要的对于识别潜在的清洁周期问题才导致无效的清洗周期。例如,缓慢增加的趋势在最后漂洗电导率可 能不是无效的清洗过程的说明。然而,这种趋势应该需要调查原因。在给出的例子中,这可能是因为喷雾装置 被堵塞,在这种情况下,它应该清洗,应采取适当措施防止堵塞。另一方面,这可能是由于污染电导传感器。报 警监测和趋势将指示周期失败,虽然它不会主动识别潜在的问题,是理想的。事件的警报仍应趋势来决定是否 需要额外的过程控制来降低频率的担忧。
文章来源：允咨生物GMP学苑
本网站刊载的所有内容，包括文字、图片、音频、视频、软件等，如非标注为“原创”，则相关版权归原作者所有，如原作者不愿意在本网站刊登相关内容，请及时通知本站，我们将第一时间予以删除。
分享精彩资讯