Process validationProcess Validation (FDA 2011 lifecycle + EU GMP Annex 15)

Process validation (PV) is the documented evidence that a commercial manufacturing process is capable of consistently delivering product meeting its critical quality attributes (CQAs) — and will continue to do so over its commercial life. The 2011 FDA guidance reframed PV as a lifecycle (replacing the older 'three-batch prospective validation' shorthand): Stage 1 Process Design produces the knowledge base; Stage 2 Process Performance Qualification (PPQ) confirms commercial-scale capability on a defined number of qualifying batches; Stage 3 Continued Process Verification (CPV) trends every commercial batch to assure the process remains in control. EU GMP Annex 15 is harmonised in intent if not in wording; ICH Q8/Q9/Q10/Q12 are the supporting scientific framework.

• QTPP and CQAs defined (ICH Q8) — what the product must deliver clinically and the attributes that prove it.
• CMA / CPP identification — multivariate DoE links input variability to output performance.
• Design Space (optional but powerful) — multivariate envelope inside which the process is known to deliver.
• Control Strategy (ICH Q10/Q11) — the planned controls on materials, process parameters, in-process tests, facility and finished spec that together assure CQAs.
• Risk assessment (ICH Q9) — FMEA / PHA / fishbone; drives the validation strategy and the sampling plan for Stage 2.
• Scale-up plan — laboratory → pilot → commercial; bracketing or matrixing rationale documented.

PPQ is the commercial-scale confirmation: the process, the facility, the equipment, the utilities and the trained personnel all come together on a defined number of qualifying batches (historically three; modern risk-based decisions can range from 3 to 10+ based on prior knowledge, process complexity and variability), with elevated sampling and analytical testing, against a pre-approved PPQ protocol. The PPQ report concludes whether the process is qualified and what the acceptable commercial operating envelope is.

EU GMP Annex 15 distinguishes Prospective, Concurrent and Retrospective validation — prospective (PPQ before commercial release) is the default; concurrent is rare and requires strong justification; retrospective is no longer accepted for new products. Modern Continuous Process Verification (CPV-style PV based on extensive in-process measurement) is also recognised under Annex 15 §5.

Stage 3 is the ongoing assurance that every commercial batch confirms the process is still in a state of control. Each lot's critical attributes feed into a Stage-3 SPC / trending programme — rolling means, Cpk/Ppk, Western Electric / Nelson out-of-trend signals. Adverse trends trigger investigation; positive trends inform process improvement (tightening of specs, recipe optimisation, design-space expansion). Stage-3 data is the principal feed into the APR / PQR and the management review.

Annex 15 covers Qualification (DQ, IQ, OQ, PQ — the equipment / utility spine), Validation (process, cleaning, analytical method, computerised system), Change Control, Re-Qualification / Re-Validation cycles, and the Validation Master Plan (VMP) that governs all of it. Process validation in Annex 15 sits inside this wider qualification framework — equipment must be qualified before PPQ can start; cleaning validation runs alongside; analytical methods supporting release testing must themselves be validated under ICH Q2(R2).

1. Manufacturing process (this entry) — Stage 1/2/3.
2. Cleaning process — separate cleaning validation with HBEL-derived MACO + recovery.
3. Analytical methods — ICH Q2(R2) accuracy/precision/specificity/LoD/LoQ/linearity/range/robustness.
4. Computerised systems — GAMP 5 + Annex 11 + Part 11 risk-based validation.
5. Sterilisation / sterilisation cycles — ISO 11135 (EO) / ISO 17665 (steam) / ISO 11137 (radiation).
6. Aseptic process — media fills (Annex 1 §9 / PIC/S PE 010).
7. Container-closure integrity (CCIT) — USP <1207> + Annex 1.
8. Packaging process — line clearance, label-count reconciliation, serialisation if applicable.
9. Shipping / cold chain — ISTA 7E + WHO TRS 961.

1. PPQ executed on three batches without a documented rationale why three is sufficient (post-2011 rationale required — historical 'three batches because regulation' is not acceptable).
2. Stage 1 design knowledge not summarised — process developed at the laboratory and pilot scale but the rationale connecting CMAs/CPPs to CQAs is in a dozen reports never tied together.
3. PPQ run at one site / one campaign with no consideration of inter-campaign variability (raw-material lot, operator shift, season).
4. Stage 3 CPV programme not implemented or implemented as a spreadsheet that hasn't been touched since launch.
5. Re-validation triggers not defined — when does a change to a raw-material supplier, a piece of equipment or a control parameter require a re-PPQ?
6. Acceptance criteria in the PPQ protocol set after the data is generated — the protocol must pre-define what 'pass' means.
7. PPQ deviations not assessed for impact on the validation conclusion — deviation closed, PPQ released anyway.

• Process master holds the QTPP, CQAs, CMAs, CPPs, design space (if filed), control strategy and the supporting Stage-1 development reports — one queryable record per product.
• PPQ workspace: protocol with pre-approved acceptance criteria, elevated-sampling plan, deviation tracking, raw and processed data attached, final PPQ report with QA sign-off via two-person e-signature. Cannot be backdated; cannot be released with open deviations un-assessed.
• Stage 3 CPV dashboards per product: rolling Cpk/Ppk per critical attribute, Western Electric / Nelson OOT signals, by-batch run charts with validated mean and ±3σ control limits, by-supplier / by-shift / by-season strata. Adverse trend opens a Stage-3 investigation.
• Re-validation triggers: every approved change in change control is assessed against the validated envelope; out-of-envelope changes auto-route to validation for a re-PPQ decision.
• VMP per site: the master plan holds every qualified piece of equipment, every validated method, every validated process — with the next re-qualification / re-validation due date. The validation programme cannot quietly fall behind.
• APR / PQR consumes the Stage-3 trends and the validation activity in the period — the annual review and the lifecycle data agree by construction.