Issue link: https://www.e-digitaleditions.com/i/1517802
30 Pharmaceutical Technology ® Quality and Regulatory Sourcebook eBook March 2024 PharmTech.com Qualit y by Design regarding process technical risks and mitigation plans. A parallel analytical risk assessment process links the analy tical target profile (ATP) (9) to ana- lytical procedure development with the identifica- tion of critical analytical procedure parameters (11). The outputs of both process QbD risk assessments and ana ly tica l met hod/testing strateg y informa- tion, along with product quality data obtained from cl i n ica l a nd/or com mercia l ma nu fac t u r i ng, a re used to document the overarching product controls strategy. The control strategy is then used as a key source of information for Module 3 of the common technical document (12) of the Biological Licensing Application (BLA) (or New Drug Application (NDA) for small molecules). Additional overlap and data f low between the regulatory filing and risk assess- ments is seen in the case where established condi- tions (ECs) are specified for a product per ICH Q12 (5). T here a re obv ious d rawback s to t h is sta nda rd data f low. First, even this generalized and simpli- f ied v iew is seen to be h igh ly complex. T he v iew in Figure 2 a lso on ly ref lects a single lifec ycle it- erat ion , a nd a s such , t h i s process wou ld repeat multiple times through process development, tech t ra n sfer, a nd m a nu f ac t u re at c l i n ica l a nd com- mercial scales. Such recurrent complexity creates inherent inefficiencies, and hence increased time and cost, in completing key chemistry, manufactur- ing, and controls (CMC) deliverables. Consequently, CMC project team members are forced to navigate t he web of intercon nected inputs and out puts to manually (and often painstakingly) search for the data needed to author the next deliverable in the se- quence. This is especially true in the case of QbD risk assessments, where leveraging platform and prior knowledge is key not only to ef ficiently complete the task, but also to applying the appropriate level of technical rigor by considering the entire body of data in a product portfolio. Too often this essential prior knowledge is housed in disparate databases and separate reports in suboptimal document man- agement systems. Additionally, each input from one data source to a spreadsheet or document represents a manual transcription, which in turn requires a check prior to approval and publication to ensure good pharmaceutical practices (GxP) control. The data transcription and checking process represents a major time commitment for CMC project teams and often corresponds with the end of a product life- cycle phase or major CMC activity (e.g., pre-process qualification, pre-submission) where resources are already constrained. The data f low associated with a digital risk assess- ment platform, especially when integrated within a CMC digital ecosystem (Figure 3), addresses these inefficiencies and limitations by streamlining the data f lows within a common platform. All product and process definition inputs, along with relevant platform development data and prior k nowledge, are fed into the platform and routed to the relevant assessments therein. Data f lows automatically be- tween assessments without the need for manual in- tervention. This not only enables CMC project team members to focus more directly on identification and discharging technical risks, but also represents a major time savings by removing the need for mul- tiple transcription checks. Enhanced understanding of intra- and inter-project risks can also be achieved using built-in risk visualization tools in digital sys- tems such as iRISK. Figure 4 presents a risk heat map and process risk mapping for an example bio- pharmaceutical drug substance "A-Mab". These vi- sualizations highlight areas of key focus for further development work and mitigation strategy. It is essential to note here that data standards that are necessary for use of a digital risk management platform will breakdown data silos, a well-known challenge not only with risk assessments but with ma ny processes i n t he pha r maceut ica l i ndust r y. More specif ically, common data standards are re- quired to enable the efficient querying of platform a nd pr ior k nowledge f rom a database, as wel l as cross-produc t su m m a r y a n a lyses of r i sk a ssess- ment outputs. In this sense, digital risk assessment platforms also help enforce and maintain data stan- dards by driving their structured implementation. This much needed structurization of data aligns di- rectly with FDA's quality assessment system known as Knowledge-aided Assessment and Structured Ap- plication (KASA) (6). In addition to risk assessments, digitalization of other QbD workf lows (1) using tools like iRISK can offer up similar benefits and drive improved inte- gration of activities for the development of a product and across different products. For example, QTPPs, Digital transformation of risk assessments can reduce the time and human errors possible from traditional risk management, technology transfer, and reporting.