Real-Time Release Testing

What Has to Be in Place Before RTRT Is Credible

RTRT is not simply a testing shortcut — Annex 17 and the underlying ICH Q8/Q10 framework expect a validated process understanding, a documented control strategy tying in-process measurements to the finished-product attribute they are substituting for, and process analytical technology capable of generating that data reliably in real time. Attempting RTRT without this foundation inverts the logic: it becomes testing less without actually knowing more.

Regulatory Basis

RTRT is most explicitly codified in EU GMP Annex 17, which sets out the assessment and validation expectations a manufacturer must satisfy before regulatory acceptance of an RTRT strategy for a given product and attribute.

The Mindset Shift RTRT Represents

SPEQ interpretation: RTRT deliberately moves away from a “test quality in” mindset, where a batch’s fate is decided by a sample tested after the fact, toward a “quality is demonstrated as it is built” mindset, where the process itself generates the evidence of conformance continuously. This is philosophically consistent with — and difficult to achieve without — the continuous manufacturing approaches described in ICH Q13.

What Annex 17 (Revision 1, 2018) Requires

The current Annex 17 came into operation on 26 December 2018 and extended a text that had covered only parametric release of terminally sterilised products. It applies RTRT to any stage of manufacture and any type of finished product or active substance, including intermediates, where control of critical parameters and relevant material attributes is authorised as an alternative to routine end-product testing (section 2.1). Its minimum criteria are that real-time measurements are accurate predictors of the corresponding finished-product attributes, that the combination of material attributes and process controls replacing a test is established with scientific evidence, and that the combined data form a robust basis for the release decision (3.2). The strategy is integrated into the quality system, with risk management, change control, a control strategy, specific training, qualification and validation of in-line and on-line methods, deviation/CAPA, a contingency procedure for sensor or equipment failure, and periodic review (3.3–3.7).

Two rules change day-to-day practice. Once RTRT is approved it should be used routinely for batch release, and if RTRT results fail or trend toward failure, end-product testing may not be substituted; the failure is investigated and considered in the release decision (3.10). Attributes controlled indirectly by approved RTRT still appear on the certificate of analysis, reported as “Complies if tested” with the footnote “Controlled by approved Real Time Release Testing” (3.11). Parametric release, by contrast, applies only to products sterilised in their final container by moist heat, dry heat or ionising radiation, and requires a history of acceptable GMP compliance and a documented sterility assurance programme (4.3–4.5).