You are here

Compounded medicines and good manufacturing practice (GMP)

Guide to the interpretation of the PIC/S Guide to GMP for compounded medicinal products

29 May 2017

Book pagination

Interpretation - quality control - chapter 6

Topic PIC/S clause Interpretation
Environmental Monitoring data 6.7, 6.9

As microbiological data is generally retrospective it is less useful for batch-specific actions for products with a very short shelf-life. Most products will have been released before the information is available. However, environmental monitoring data should be considered as part of batch release for those products with longer shelf-lives that permit the evaluation of data prior to release.

Multiple batches are compounded in a single session so any excursion cannot be readily traced or tied to a single batch but rather implicate a number of batches.

Generally, for sterile medicines, the use of closed systems should reduce the risk of microbial ingress to the product but only if all precautions have been taken (such as an effective decontamination program, use of good aseptic techniques including 'first air' principles, etc.)

If, when reviewing EM data, a potential problem with environmental control is identified, the basis on which the manufacturer is confident to continue with release should be documented for operations where the reporting of results is retrospective.

Sufficient information should be available from the EM program to identify any loss of control in a timely manner and to enable appropriate remedial actions.


Adequate microbiology expertise, either on or off site, is needed to support the provision of acceptable quality, (including sterility assurance and environmental monitoring programs).

For sterile products, the identification of all microorganisms in grade A areas should routinely be to species level. Staff performing identification test should be adequately trained and experienced.

Isolates from Grade B should be identified to at least genus level except when:

  • High individual counts are recovered
  • Negative trends indicating a deterioration in environmental control emerge
  • Recovery of potentially objectionable organisms

In these cases, additional identification of organisms (at least to species level) should be performed to aid in investigation and rectification of the event.

Typical local isolates should also form part of the validation for cleaning and EM programs.

Reading of any plates should be performed in a location and in a manner that does not present a risk to manufacturing operations.

For non-sterile products, appropriate microbiological testing of starting materials and products should be in place to demonstrate compliance with TGO 77 requirements. A risk based approach may be taken to testing based on the nature of the dosage form and the manufacturing process utilised, and the resultant risk of microbiological contamination of the product.

Controls in place for media should include supplier evaluation and the availability of a CofA. The suitability of each lot of prepared media should be verified before use, either by performing growth promotion testing of each delivery of each lot of media received, or alternatively, by validating the transport system usedused by qualified pre-prepared media suppliers to ensure that media deliveries are routinely transported under appropriately controlled conditions.

Testing of starting materials that are NOT finished therapeutic goods listed on the ARTG 6.17

Where products are compounded from 'starting materials' that are APIs and/or excipients, all GMP requirements should be met. For example:

  • An identity testing on all containers for both the API and excipients.
  • Full testing of the materials in accordance with the relevant pharmacopoeial monograph.
  •  A review of the manufacturer's/supplier's CoA against the company's specification.
Testing of starting materials that ARE finished therapeutic goods listed on the ARTG 6.17

Where sterile products are compounded from 'starting materials' that are sterile finished goods registered on the ARTG, the default requirements will be compliance with the attributes listed on the specification (item name, ARTG #, label, approved manufacturer/supplier, etc.)

There should also be a formal check to determine whether the goods show evidence of being falsified or counterfeit.

No additional testing of the starting material is expected.

Testing of finished products 6.3

The requirements for finished product testing should be commensurate with patient risk, taking into account the intended use of the product, and the methodology of manufacture.

Where manufacture involves the use of starting materials that ARE finished therapeutic goods listed on the ARTG the TGA does not expect routine chemical testing of the finished product. However, it is expected that the medicine meets any relevant label claim and pharmacopoeial standards if tested.

Where manufacture involves the use of starting materials that are NOT finished therapeutic goods listed on the ARTG e.g. APIs or excipients, there is an expectation that finished product testing will be performed; ID testing, assay and, in addition, for sterile products, the test for sterility and endotoxins.

Reference and Retention Samples 6.14 & Annex 19

Where products are compounded on a batch basis i.e. 10 or more of the same product in the same session, then retention samples need to be taken. The size of the retention sample kept should be determined following risk management principles and take into consideration potential need for retesting.

Reference samples are expected for products where manufacture involves a discrete bulk manufacturing step, i.e. products are produced using APIs or excipients as starting materials'.

Samples of Finished Product labels and any other printed items used are to be included as part of batch documentation.

Expiry date

Product expiry should be based on a scientific rationale, including test data. Laboratories used to generate this data should operate an appropriate quality system and be subject to the company's supplier approval system.

  • Test data may be obtained from literature searches, provided that the literature is relevant to the product formulation and container/closure system proposed.
  • Expert opinion on product shelf life must be supported with a documented rationale and test data if available.
  • Test data needs to cover physical, chemical and microbiological stability (i.e. remain sterile through the labelled shelf life).
  • For products manufactured from starting materials that are registered therapeutic goods the expiry of the resulting product should not be greater than the shelf life of the input product prior to reconstitution.
  • For products manufactured from starting materials that are API and excipients it is expected that all these are within the current shelf life at the time of manufacture.
  • For multiple use containers, the expiry date/time of the container starts when it is first opened. Data used to assign product expiry must be derived using stability indicating analytical methods, and be relevant to the proposed product formulation and container closure system. Preservative Efficacy Testing may be required to support in-use stability.
  • The assigned shelf life must include a margin of safety from the stability data available.
  • Special attention should be given to shelf lives assigned and the methodology used for biologically derived products such as MABs.
  • Stability summaries should include data relating to the suitability (including compatibility) of the container/closure system used. TGA is aware that not all devices (e.g. syringes) are intended to be used as closed container storage systems for medications. Reduced potency has been reported in some cases when medications are pre-filled and stored in syringes.
Ongoing stability 6.23-6.33

Where formulations are manufactured routinely and have an expiry greater than 24 hours then an ongoing stability programme should be conducted. Where justified, grouping of formulations for such a program is acceptable. Stability information should be based primarily on actual trials using the unique combinations of active, diluents and packaging components and secondly on available literature. The manufacturer should have a documented justification to its approach for each product.

There should be available to review at inspection evidence of both chemical and microbiological stability of the dosage units for the period of storage up to and including the labelled shelf life. That is, they need to comply with Part 3-1 of the Act which would mean compliance to the BP or USP monograph for Parenteral Preparations.

When using literature-based evidence the literature review needs to be based on known published journals. Manufacturers should verify the source material in terms of device, diluent etc.

Test for Sterility Annex 1: 125, 126, 127

A documented sterility test programme must be in place, which includes consideration of all process variables and risks.

The program should be designed to ensure that variables such as product and operators are adequately monitored and controlled. The TGA expects a higher rate of monitoring in newly established facilities or where historic test data is not available.

The frequency for sterility testing is determined by the nature of the starting materials used in manufacture:

For products manufactured from starting materials that are registered therapeutic goods,

  • The minimum expectation is one sterility sample per operational work station per week.
  • The requirement for sterility testing may be off-set by the use of a suitably designed "end of session media fill simulation" as part of an ongoing monitoring programme. The frequency of this should be in line with the minimum requirements for performing the test for sterility, i.e. minimum of one simulation per operational work station per week.
  • End of session media fills should evaluate all types of aseptic manipulation, and variables such as product and operators cyclically covered on a rolling basis. The processes to achieve this should be documented in the QMS.

For products manufactured from starting materials that are NOT registered therapeutic goods:

  • A formal test for sterility should be performed. In the case of product with an expiry exceeding one month, the results of the sterility test should be used for batch release. In shorter dated stock, the test for sterility results should be used for trending purposes.

Samples taken for sterility testing should be representative of 'worst case'. Refer Annex 1 Clause 127.

In performing the test for sterility, the use of a 'simulated product' may be accepted where scientifically justified and as long as it is processed using the same steps and conditions routinely used for production. Worst case scenarios should be replicated when manipulating simulated product. The testing of simulated product is not permitted for products manufactured from starting materials that are NOT registered therapeutic goods.

If antimicrobials are present in the product formulation (e.g. multi-use vials) then these need to be inactivated.

Book pagination