USFDA 483 / IPCA / June 2023 /Failure of Quality Unit in taking adequate actions on Stability OOS, Rework of APIs.

Observation 2:

Quality unit failed to take adequate actions against failing batches sold in domestic and international markets. Two batches of API failed for impurities during stability testing at 24Months under long term conditions; the batches were consumed for domestic market. No market action of intimation to customers, recall or information to regulatory agencies. Firm only tested limited number of batches for investigation.

Observation 3:

Reworked batches of API released to market without adequate evaluation and review of stability data, the batches were not charged for stability

GMP application and compliance to GMP principles to ensure product integrity and patient safety are fundamental aspects of Pharma manufacturing and control. It cannot be applied selectively to select markets when a manufacturing site caters to multiple markets. Regulators look at the approach of the Company to Quality Management system implementation and GMPs.

• When the product is distributed with same specifications for different markets (stringent regulated markets, semi-regulated markets, domestic market), the market actions in case of quality events like OOS, Deviations etc. need to be same. Failure investigation should evaluate extent of impact, requirement for product recall, intimation to customers and agencies and take actions to ensure product integrity and patient safety are not compromised.
• If the specifications applicable to specific markets to which product is distributed is more relaxed (e.g. different limits for specified known impurities) a product recall / market intimation may not be necessary for these markets. But this conclusion should be supported with appropriate impact assessment & product safety risk assessment and documented.
• When OOS in stability test is confirmed, impact should be evaluated on all distributed batches of the product within valid expiry / retest period. Retesting of the reserve samples should be as per a defined protocol, and rationale for number batches to be tested to be scientifically justified, or else reserve samples of all batches need to be retested (For e.g. if a OOS is observed after a specific change, the batches before the change may be excluded from retesting; if the batches manufactured with specific lots of input materials caused OOS in stability, other batches may be excluded). The protocol should have scope for testing all batches if more failures are observed during the retest. 
• When products are reworked (rework means differences in the manufacturing process), impact of the changes should be evaluated and documented, before releasing the product to market. The evaluation should confirm the reworked batches are equivalent to regular batches in quality and safety at the time of batch release as well as through the shelf life. Stability data of the batches with the rework process should be evaluated. If there is no preexisting data, representative reworked batch(es) should be kept for stability at accelerated and long-term conditions and batches to be released only after evaluating stability data (e.g. 3 months  accelerated and long term data, and continue stability monitoring).

It will be prudent for the Quality units of the companies to review for similar events and take remediation actions if necessary:

• Review all stability OOS reported and evaluate whether adequate impact assessment and market actions are taken, irrespective of the markets distributed (regulated, semi regulated, domestic). Where gaps are observed perform the impact assessment and based on this initiate additional actions. If no further actions are required, document the same with scientific rationale.
• Review the OOS procedure / procedure for handling stability OOS, elaborating different scenarios and actions e.g. (but not limited to) – actions in case of products distributed to different markets with different specifications, products with single specification for all markets; Impact assessment should be compulsory for all such OOS.
• Review all products for reworks / rework methods and stability data. Where stability is not initiated on reworked products / batches initiate the same. Where stability data is not available, as an immediate action analyses the reworked batches on current date to evaluate whether the batches still confirm to specifications. Initiate market actions where failures are observed during the reanalysis. Review the procedure for handling of rework making it mandatory to review accelerated / long term data for defined period (e.g. 3 months) before releasing the reworked batches for distribution. This should be check point in the batch review and batch release checklists.
• Train and coach the Quality and Operational team on consistent application of GMP principles in handling Quality events and failures.