According to EU GMP guide part I: Basic requirements for medicinal products: Chapter 1: Pharmaceutical quality system, you are expected to conduct PQRs annually. They can be adjusted according to the manufacturing and campaign duration, providing one has written a workflow for this in an SOP.

Even if no manufacturing has occurred in the review period, the quality and regulatory review should be conducted as per section 1.10 and include stability results, returns, complaints, recalls, deviations (including those arising from qualification and validation activities) and regulatory background.

A PQR is a living document, and you should always consider previous PQRs when doing the next one.

Here are additional points to consider, suggested by various GMP experts:

Oliver Clem 

If you think about it from a Q9 perspective, it should ideally be a living document – new data on the process and product performance are always being added (Stability, batch manufactures, complaints…), which increase your knowledge and informs your risks. The assessments then take place more frequently, rather than the annual “feast” on smaller bites”. Overall this reduces your exposure if things change, reducing your compliance risk. A lot of effort to set up, and daunting when the complexity is understood by the wider team (not just those few who wrote them previously) but very worthwhile when running.

Bernd Bödecker

The key purpose of the PQR is to detect adverse trends, batches OOT, failures of the quality system related to the product under consideration (e.g. a new starting material supplier not subjected to prior qualification), cumulative adverse effects of various changes made, etc.,
and to check whether there is any need for action / improvement.

As rightfully stated by Oliver Clem, this is best achieved with a rolling PQR, but GMP guidelines also allow PQRs with fixed intervals.

Every company must decide on its own what is more advantageous. Much depends on the extent to which PQR data is anyway subject to continuous monitoring and trending (e.g. ongoing process verification; QC data, calibrations, environmental controls, stability data, …). Redundancies should, of course, be avoided.

Picking up an erroneous comment below, remember that the purpose of PQR is to look at a period of the product as a whole, NOT to reassess individual batches (this should have been done before their release …).

Accordingly, normally it does not make sense to perform PQRs with less than 3 batches – you simply don’t see any interpretable trends.

However, selected items (e.g. stability) may require annual review even if only a few batches are made.

Pradeep Ahire 

As per ICH Q7, a PQR review is expected annually. Review time frames can be appropriately adjusted based upon manufacturing and campaign duration with adequate justification. Even if no manufacturing has occurred in the review period, the quality review should be conducted per section ICH Q7, paragraph 2.50, and include stability, returns, complaints, and recalls. For example, a product quality review may encompass more or less than 12 months depending upon product campaign duration (ICH Q7, paragraph 2.50; ICH Q10, section III.F (2.6)).

An example of rolling a PQR can be done in such a manner that total products can be divided into groups such as Group 1 – January to April, Group 2 – May to August, and Group 3 – September to December. As per this example, the APQR of the Group 1 product needs to be completed from January to April every year.

While preparing the schedule, the priority can be decided on a risk-based approach. This priority can be decided with the help of parameters such as the number of batches marketed, customer complaints, product recall, stability failures, product deviations, out of trends, customer requirement, etc.