USFDA 483 to Torrent: Concerns on Handling OOTs and OOS

The firm had reported more than 70 valid out of trends (OOTs) between Oct 2022 & May 2024. But even for valid OOTs, no stability studies were initiated which was a requirement as per the Firm’s SOPs. The SOP for handling OOT was found to be deficient with no requirement for impact evaluation and market action on valid OOT batches in annual stability programme. OOT Batches in stability studies under long term were allowed to continue in the market. There were cases where valid OOT batches later turned out to be OOS. Several OOTs revealed unexpected increases in assay values during stability studies, with one instance showing an 11% increase. The cause of these increases remained unresolved, with one case attributing the issue to analytical method, which was not corrected for over 18 months. Meanwhile, batches continued to be analyzed and released using an unreliable analytical method. This raises the concern that many OOT batches could be truly OOS, raising risk of sub potent drugs in the market. There were several Adverse Events (ADE) reported for lack of efficacy, but these were not logged as product quality complaints (PQC) and evaluated. During the investigation of OOT events, historical trend evaluation considered only OOT results of batches which was expired and not OOTs from ongoing stability.

The inspection also discovered more fundamental lapse in handling of an OOS. A valid dissolution OOS was observed in a batch of particular bottle pack (counts) in 3^rd month long term stability. Even through the same bulk batch was used in different pack sizes of bottle packs, no market action was taken on these batches. The OOS SOP also was deficient in addressing handling of stability OOS, filing of FAR (Field Alert Reports) and batch recall. 

While regulatory guidelines are not specifically addressing handling of OOTs, the USFDA guidance Investigating Out-of Specification (OOS) (Foot Note 6) notes that the guidance may be useful in examining results that are out of trend. OOTs are atypical results, atypical because they deviate from expected variability or a time-dependent trend. For example, a batch analysis result for assay, impurities or dissolution may fall outside the expected variability or trend based on historical data. This is an Out of Expectation (OOE) result, also often termed as OOT. A stability analysis result of a specific parameter of a drug product or substance may fall outside an established or historical trend. For e.g. a sudden drop in a dissolution result or sudden surge of an impurity in a particular stability time point. Whether we designate these atypical results as OOE or OOT, there shall be a well defined procedure for handling these atypical results. (In the rest of the post the term OOT is used for sake of convenience). A validated process is expected to be in a state of statistical control with results of parameters for different batches falling within a range; similarly an analytical method will have an expected range of variability for the results. When results are falling outside these ranges, this need to be investigated to establish whether there is an emerging trend with an underlying cause (systematic error) which if left unattended can result in a future batch becoming out of specification. An OOT result in a stability analysis could be indicating there is a possibility for the batch to fall outside the specification (OOS) during shelflife. An anomaly in process or analysis may have caused an atypical results. An unusual impurity observed in a batch could be indicating contamination or it could be an analytical issue of sample degradation.

First and foremost in handling OOTs, is to define criteria for triggering an OOT alert. Define the criteria to trigger OOE / OOT judiciously. For batch analysis the limits for triggering an OOT may be defined applying statistical criteria based on historical trend of the product quality parameters– for e.g. OOT limits may be defined as Mean ±3σ or 2σ (Mean – Average or statistical mean value for a parameter from a number of batches, σ – standard deviation of the parameter). Where sufficient historical data is not available (in case of new products), limits for OOT may be defined as a percentage of the specification limits – for e.g. results shall be within 80% or 60% of the limits for assay, impurities, or an OOT shall be triggered. Give due to consideration to developmental data and stability data of the developmental batches while defining the criteria. Take care not to define the OOT limits not too tightly such that every other batch trigger an OOT investigation; at the same time don’t define it too loosely either or else aberrations are missed out. As historical data accumulates, the criteria should be reviewed periodically (e.g., annually) to reflect real-time trends in batch variability. While defining OOT limits for stability analysis, give consideration to how much the result differ from the specification limit as well as previous station results. For example in a stability assay analysis OOT may be triggered if results deviate beyond 5% of the specification limits or beyond 3% of previous station results. Once an OOT is triggered, the investigation itself shall be as rigorous as in OOS investigations – analytical investigation (Phase 1A, Phase 1B) and detailed investigation including manufacturing investigation (Phase 2). Based on the investigation conclude whether the OOT result observed is valid or invalid. In case of valid OOT, establish the corrective actions. 

OOT Impact Evaluation: Valid OOT results should also trigger impact evaluation. While technically the OOT batches are still within the specification, there should be an evaluation to assess whether the batch can go out of specification (OOS) during the shelflife. This is important if the OOT is on a stability indicating parameter which could keep changing during shelflife – assay, degradation impurities, dissolution and so on. Also the impact evaluation should consider whether other released batches could be impacted by the root cause for OOT. If there is a possibility for batch(es) to go out of specification during shelflife, such batches shall not be released to market and batch(es) already released may need to be recalled. In the case of valid stability OOT impact evaluation should consider not only the specific batch for which OOT was observed but all batches within valid expiry that has already been released. The form 483 specifically points that in spite of observing valid OOTs during annual stability, the batches were allowed to remain in the market. There shall be strong scientific justification if it is concluded that the OOT batch can be released or other distributed batches need not be recalled. Evaluating historical trend of OOTs can provide valuable insights into the possibility of a current OOT batch to become OOS during shelflife and even support a decision not to initiate a market action like recall. But this evaluation should consider OOTs observed in ongoing stability programme as well; or one could miss new trends, new parameters.

Loading OOT batches for stability:  In batch analysis, if the valid OOT results are on stability indicating parameters which can keep changing during shelflife (for e.g. assay, impurities, dissolution, water content etc.) the OOT batch shall be charged for stability studies. The frequency of testing may be also higher in the initial months to establish the trend (for e.g., every month for the first 3-6 months). If a batch with similar result is already under stability, or there is a strong case that the parameter will not fall outside the specifications during shelflife, not loading the batch for stability may be justified. But this should be documented with strong rationale.

OOS Impact Evaluation: Inadequate impact evaluation of OOS incidents and remediation actions is often cited issue in FDA Warning letters. An OOS incident in one particular packsize / bottle size is strong case for raising a flag on quality of other packing configurations /packsizes of product packed from same bulk batch. It should be comprehensively investigated, and unless it is established and documented with well justified rationale that the OOS incident is specific to only the specific bottle pack size, mitigation action including recall should be considered on all other packed batches.

Prompt Investigations: While handling OOTs and OOS, investigation related analytical activities, method evaluations etc. shall be accorded maximum priority. When reliability of method is doubted not only the OOS/ OOT batches, but even the results of the batches which comply is in question. Closing such investigations and concluding on the corrective actions should get priority even over other batch analysis for batch release.

Handling of Efficacy Adverse Events (ADE): When drug adverse events are reported which are in line with the adverse reactions detailed in the product label and label insert, this may be handled by the Pharmacovigilance (PV) or Medical Affairs teams. But when the adverse event is related to the drug quality performance (for e.g. lack of efficacy could be due to loss of potency or drop in assay), these should be treated as quality events. Product Quality Complaints (PQC) shall be logged and issue investigated with evaluation of reserve samples, market samples, historical trends etc. 

The 483 issued in June 2024 also highlighted repeat observations from a previous inspection in September 2022 (investigators Joseph E. Melendez and Pratik S. Upadhyay):

1.      Failure to Thoroughly Review OOS Incidents: The 2022 Form 483 had observed several OOS incidents for extraneous or unknown peaks results in batches which were attributed to glassware contamination without sufficient evidence. This imply that investigations into OOS results continue to be inadequate, impact evaluations inadequate, results are invalidated without scientifically justified root causes.

2.      Failure to Clean Manufacturing Equipment: Residues were found on supposedly cleaned surfaces (equipment, walls, floor) in 2022 as well, indicating the cleaning procedures are not improved despite previous warnings..

Concern over Repeat Observations: The recurrence of the same or similar issues suggests that Corrective and Preventive Actions (CAPAs) are ineffective. This raises questions about the robustness of the Quality System and its ability to implement lasting improvements. It also raises concern that product quality may have been compromised for an extended period since the last inspection. It is imperative that when deficiencies are discovered, investigations must be thorough. Actual root causes should be established, and CAPAs must be implemented that address these root causes effectively.

Thorough investigation and handling of OOT and OOS incidents is critical in Pharma Industry. There should be well defined procedures for handling OOT / OOS incidents which address:

•         Scientifically and statistically sound criteria for defining OOTs
•         Period review of the OOT limits as more batch data gets generated
•         Complete and comprehensive investigation to arrive at probable causes for OOT / OOS and corrective actions (Phase 1A, Phase 1B, Phase II investigations)
•         Impact evaluation of OOT incidents covering impact on the batch (potential OOS for the batch before expiry) and impact on other batches
•         Evaluation of need for stability loading for the OOT batch when OOT is on a stability indicating parameter
•         Detailed procedure for OOS  &  OOT investigations in line with regulatory guidance (e.g. USFDA OOS guidance; MHRA Guidance on OOS investigations) addressing impact evaluation and documentation
•         When adverse drug events (ADE) are reported which could be related to product quality, such issues shall be logged and investigated by the manufacturing site quality unit.

Ensure that deviation investigations are thorough, actual root causes are identified and effective corrective actions are implemented. Failure to do so can lead to serious regulatory implications.