Protein A Resin Cycling Beyond Cleaning Verification Limit
After 95 cycles on a Protein A chromatography column, the eluate turbidity is 12 NTU against an action limit of 5 NTU, and the column pressure drop has risen from 0.4 to 0.8 bar at the same flow rate. The validated resin lifetime is 100 cycles. Cleaning verification after the 95th cycle shows TOC 0.8 ppm against a limit of 0.5 ppm. The next campaign starts in three days and requires 20 cycles. The column is 20 cm in diameter with a 15 cm bed height. A replacement resin lot is available but requires a 5-day equilibration and packing qualification. A backup column is not available. Historical eluate turbidity for this step is 2–4 NTU. The product is a marketed monoclonal antibody, and the deviation procedure requires a documented root cause and impact assessment before any extended use.
1. What is the most likely root cause of the rising pressure drop and turbidity, and what data would confirm it?
Read the worked answer
The most likely cause is resin fouling or aggregation in the bed, with a secondary possibility of ligand leaching or feed clarification failure. The pressure drop doubling at the same flow rate points to reduced bed permeability, not a flow-meter error. The turbitity increase from a historical 2–4 NTU to 12 NTU suggests particulate or aggregate carryover, and the failed TOC cleaning verification supports incomplete removal of residual protein or cleaning agent.
Confirming data should include:
- Pressure-flow curve compared with the baseline to distinguish bed compression from fouling.
- HETP and asymmetry measurements to check bed integrity and channeling.
- Resin inspection for discoloration, fines, or aggregates, plus particle counts in the eluate.
- Cleaning efficacy study with extended contact time or a different cleaning agent, measuring TOC and residual protein.
- Feed and load characterization to rule out a upstream clarification failure, including turbidity and particle counts.
- Ligand leaching assay if fouling alone does not explain the turbidity.
The investigation should not assume that because the resin is within its 100-cycle lifetime, the column is performing as validated.
2. Can the column be used for five more cycles to reach the 100-cycle validation limit? Explain.
Read the worked answer
No. The 100-cycle validation limit is not a guarantee that every column remains qualified through cycle 100; it is the maximum lifetime supported by the validation data. At cycle 95, the column is already outside expected performance: pressure drop is doubled, turbidity is more than twice the action limit, and cleaning verification fails. Using it for five more cycles would extend an existing out-of-trend condition and would exceed the validated cleaning performance, with no data to support product quality.
The correct decision is to stop using the column for product manufacture. The deviation should be opened before any further cycles, and the root cause should be investigated. If a replacement column cannot be qualified within three days, the campaign should be delayed rather than run on a fouled column. For a marketed monoclonal antibody, the risk of elevated turbidity, aggregate carryover, and cleaning failure is not acceptable, especially when the failure is already visible at cycle 95.
3. What is your disposition and requalification plan?
Read the worked answer
I would remove the column from production, replace the resin, and perform the full 5-day equilibration and packing qualification before the next campaign. The replacement column should meet the established acceptance criteria for bed height, HETP, asymmetry, pressure drop, and cleaning verification before use. The failed column should be preserved for inspection if possible.
The investigation should determine whether the root cause is feed-related, cleaning-related, or resin-related. If the fouling is linked to a feed change, the upstream process may need additional clarification. If cleaning is inadequate, the cleaning procedure should be revised and validated. If the resin itself is degrading earlier than expected, the validated lifetime should be reduced from 100 cycles to a lower number supported by data, and the change should go through change control. The deviation should include a product impact assessment covering turbidity, aggregates, residual Protein A, and cleaning carryover for the batches made near the end of the column life. Release of those batches should be held until the assessment is complete.