Data-driven development of robust HPAPI processes
Lukas Brändli of Carbogen-Amics discusses criticality-based process data evaluation as a tool to navigate process data
The demand for cytotoxic drugs has led to an increase in the number of highly potent active APIs ( HPAPIs ) in the pharmaceutical development pipeline . The manufacturing of such compounds presents significant challenges related to limiting cross-contamination , and ensuring safety for workers and the environment . Comprehensive management systems , containment equipment , proper cleaning techniques and waste treatment systems are essential .
Due to the capital-intensive nature of such specialised capabilities , contract manufacturers play a crucial role , specifically for small and midsized players . Demanding regulatory requirements ( e . g . detailed data on the fate of impurities , tighter impurity specifications ) are driving the challenge for many CDMOs to supply customers with regulatory competence supported by relevant data .
One challenge in the field of HPAPI development is the combination of low material requirements , with the very often accelerated or fast-track status of the drug substance candidates . Small initial GMP quantities are sufficient to meet the usually low material demand of
the early clinical phases , which means that only a limited number of HPAPI batches are typically produced before a project enters the later clinical phases . Serious process optimisation and process characterisation activities are usually postponed to clinical Phase III , where process robustness in line with a proper control strategy finally needs to be established in preparation for process registration and validation . At this point , at the latest , extensive process and analytical datasets are generated to close the gaps in process understanding .
Figure 1 - Criticality plot combining process data with quality attributes normalised against their specification limits ( OOS > 1 )
Exploration of such multi-dimensional datasets is a significant bottleneck in the absence of proper data management tools that dynamically merge , transform , filter and visualise data from different sources in order to extract relevant information in a general or specific context . Process knowledge emerges from process data connected to the matching analytical data , evaluated against the corresponding specifications .
Triggered by a LEAN project in 2019 , Swiss CDMO Carbogen-Amcis officially initiated the development of a data
Criticality� 1 -+ in spec Criticality > 1 -+ out of spec ( 00S )
USL 0.1 % a / a
Product |
s1 |
Normalised specification range |
s1 |
Impurity |
LSL 97 % a / a
Figure 2 - Implementation of criticality concept for HPLC purity
16 SPECIALITY CHEMICALS MAGAZINE ESTABLISHED 1981