VAL-100 Process Validation Sampling

All sampling will be supported by appropriate sampling plans detailed either in the manufacturing instructions or validation protocol.  Sampling will follow the general principles outlined below.

5.1 Sampling Plans

Sampling plans must be developed to consider the specific attributes being measured and the risks associated with accepting a defective lot. Sampling plans and methods must be predetermined and documented in a protocol. Colleagues involved must be trained.

The sampling procedures documented in the protocol should provide sample locations, sample size, sample frequency and should also include the tests or assessments required.

Samples must be representative of the population and the sampling plan must be designed to determine any segregation during blending, transport, or handling and any variability or other factors that might impact production consistency.

The need to show homogeneity of intermediates should be considered on a case-by-case basis depending on how the intermediate is used in subsequent processing. In general, studying the homogeneity of an intermediate is of less importance than that of a final product, especially if the intermediate will be further processed in the next or final step of the process. If homogeneity of the intermediate is critical to the quality of the final product prepared from it, demonstration of intermediate blend homogeneity should be considered.

Sample size, frequency and location for tests other than homogeneity will be determined on a test by test basis and documented in a validation protocol.  Section 5.2 to 5.10 is applicable to homogeneity sampling and testing only.

5.2 Blend Sample Size

Blend sample size should be as specified in LAB-125 Sampling of Raw Materials, In-process and Bulk Finished Product.

5.3 Sampling Locations

5.3.1 Blend Uniformity Sampling

For tumble blending, at least 10 locations in the blender should be sampled and must be chosen to represent potential areas of poor blending.  At minimum samples should be selected from at least 2 depths along the axis of the blender.

For ribbon blending, at least 20 locations in the blender should be sampled and must be chosen to adequately address potential dead spots such as corners and discharge areas.  Equipment geometry should be taken into consideration when establishing sampling locations.

Blend uniformity samples should be sampled directly from the blender immediately prior to discharge.  If it is not practical to take samples from the blender, samples may be taken from the discharge stream or drums/tote bins.  At least three replicate samples will be taken from each location with one samples tested per location.

Blend uniformity acceptance criteria: RSD =5.0% and all individuals are within the mean ± 10% (absolute).

5.3.2 Finished Product (Dosage Unit) Sampling

The number of samples taken for each validation study should be based on risk.  For example, the risk to a patient from a new formulation may be significantly higher than the risk associated with a change of excipient supplier.  The risk assessment must be documented in the protocol or in a separate risk assessment.

Throughout the compression operation for a new product or a change in formulation, at least 20 sample points should be identified and sampled.  Sampling locations must be chosen to represent significant events during the compression process and must include samples from the start, middle and end of the run.

At minimum ten samples will be taken from each location with three dosage units tested for content uniformity per location.  Tablets / capsules should be collected at the final point in this step of the process (that is, post metal detector).  Each time point sampled should be collected in a separate container.

For all other process / product changes the number of samples should be based on risk.

Acceptance Criteria: RSD of all individuals =6.0%.  Each location mean is within 90% – 110% of target potency and all individuals are within 75.0% and 125.0% of target potency.

5.4 Sample Handling and Containers

Samples should be handled and stored under appropriate conditions based on the individual product requirements and tested in a timely manner and in accordance to approved procedures. Representative samples should be collected and properly labeled in accordance with approved procedures.

5.5 Sample Thief Probes

Sampling should be performed using appropriate sample thieves as per the sampling procedure and sample plan.  The appropriate sample thief and specific product sample die(s) should be selected according to the product characteristics (static, lumps, fine powder, coarse granules, etc) and dosage unit weight.

5.6 Semi Solid Sampling

5.6.1Bulk Product Sampling

Following the completion of a given processing step, semi solid products may be sampled from the manufacturing vessel, during transfer or from the holding vessel, using a suitable sampling device.  Care must be taken to ensure that no further processing of the bulk takes place following sampling (e.g. due to further stirring).

Sample from the manufacturing vessel from the top, middle and bottom for homogeneity of the API by testing the samples for potency.  The number of samples to be taken will depend on the geometry of the vessel, available historical data on the process on the process and the level of change to the process.  At minimum 1 sample is to be taken from top, middle and bottom of the vessel.

5.6.2Packed Product Sampling

 Samples should be taken throughout the filling / packaging operation. Matrixing and bracketing can be applied and the worst case packaging size validated.  These samples should be taken and tested for potency and preservative content (if applicable) to prove product uniformity (of the API and preservatives) is maintained throughout the filling operation, and also that the filling operation does not have an adverse affect on the overall quality of the finished product. The following guideline should be considered at minimum for filling / packaging sampling:

a. Beginning of Filling (ideally first 10 packages kept as product for sale): 10 samples

b. Middle (taken from throughout the middle of the filling operation): 10 samples

c. End (ideally the very last containers packaged that would be kept as product for sale): 10 samples

In addition if there are critical occurrences during the packaging run which may affect the product, then sampling at start-up immediately following such an event should be considered. The rationale behind collection and/or testing of these samples should be outlined in the protocol.

5.7 Raw Material Change

A change in supplier or manufacturer of a raw material is assessed as per SOP VAL-105 Raw Material Evaluation Process. Where a change is made to an active or excipient in a formulation and where revalidation is required, sampling requirements will be based on risk and documented in the Raw Material Evaluation Report.

For those products where routine blend testing is not performed, an assessment of the impact on blend uniformity and the requirements of the blend specification must be performed. Additional monitoring and testing during or post validation batches may be required.  This will be documented either in the validation protocol or Raw Material Evaluation.

5.8 Process Change (Including Equipment)

Each process change should be address individually and the validation requirements documented in the protocol as per SOPVAL-115Process Validation for Liquids and Solid Dose Manufacture.  Sampling and testing should be based on risk.  If the number of samples taken is different to the number specified in this SOP, a documented justification must be included in the protocol.

For those products where routine blend testing is not performed, an assessment of the impact on blend uniformity and the requirements of the blend specification should be performed.  Additional monitoring and testing during or post validation batches may be required.  This will be documented either in the validation protocol or a risk assessment.  Once validated it is not considered necessary to continue blend testing.

5.9 QC Reductive NIR Sampling and Testing

Content uniformity testing of finished product validation samples may be performed by a combination of NIR and specification testing (QC Reductive Testing).  NIR testing may be used to qualitatively determine the content uniformity spread for the batch under validation.  Specification testing of the extremes of the spread will allow full characterization of the content uniformity.

Batch release testing will be performed as per specification test requirements and must be included in the validation report.

A validation batch using QC reductive testing must meet all QC reductive test requirements for the validation samples and specification requirements for the release for sale samples.

5.9.1 Sampling

QC reductive NIR sampling must meet the minimum requirements details in Section 5.3.1 to 5.3.8 above.  Additional samples may be taken.  Scanning in time series order across the run combined with significant process events is recommended as it enables analysis of trends and correlation of trends to actual process events.

At each sample point a minimum of 10 samples must be taken.  Additional samples may be needed as NIR and QC testing may be required.  The samples must be labelled as the time taken in order to correlate NIR trends with process events.

At each sample point scan 6 dosage units.  Each dosage unit scanned must be uniquely identified and stored.   Once scanning has been completed, identify the dosage units with the highest and lowest absorbance and send to the QC laboratory for potency determination by specification test methods.

5.9.2 Testing

The potency data from the dosage units with the highest and lowest absorbance combined with the spread of absorbance values within these potency values can be used to fully characterize the active content distribution.  If the potency of the dosage units with the highest and lowest absorbance is within the content uniformity specification as detailed in the specification, there is a high degree of confidence that all dosage units tested meet content uniformity requirements.

The absorbance data should also be used to calculate the RSD, which should be compared to the specification for RSD. If all validation samples are within specification, content uniformity and RSD specifications and all samples meet specifications, no further action is required.

If either of the highest or lowest samples tested for potency by QC is outside the  specification of 85% – 115%, test the next highest or lowest sample (depending on which of the original samples tested was outside the specification) to determine if one or more samples are outside this specification.

If only one samples is outside the 85% – 115% specification but within the 75% – 125% specification, 30 or more samples have been tested by NIR and the RSD is within the content uniformity specification then the data is considered acceptable and no further action is required.

If more than one sample is outside the 85% – 115% specification then further investigation is required.  If any one sample is outside the 75% – 125% specification, further action is required.

If the data from the NIR testing is inconclusive or the highest and lowest absorbance dosage units cannot be tested, content uniformity testing will be performed as per test methods.

5.10 Parallel NIR Sampling and Testing

Parallel NIR testing may also be performed.  Parallel NIR testing is for information only. Sampling and testing requirements will be detailed in individual validation protocols.