Cleaning Validation MAC

• Swab sampling — defined-area swab (typical 25 – 100 cm²) using validated swab type (polyester, cotton, sponge); extraction into a defined solvent; HPLC / TOC / specific assay. Direct measurement at the worst-case surface. Recovery must be established per-substance per-surface per-swab.
• Rinse sampling — final rinse of the equipment collected and analysed; integrates over the entire wetted surface. Lower sensitivity than swab per area but covers inaccessible surfaces.
• Visual inspection — operator + supervisor visual confirmation; documented checklist; supplemented by white-light + UV-light inspection where appropriate; complementary to swab + rinse, never the only method.
• Recovery study — for each surface material (316L SS, glass, silicone) × each analyte: spike known mass per area, allow drying, swab, extract, analyse; calculate % recovery; correct MAC measurement by recovery factor. Recoveries < 50% are flagged for method improvement; < 30% usually unacceptable.

Most facilities manufacture multiple products on shared equipment. Validating cleaning for every product × every equipment combination is impractical. Instead: group products by similarity (potency, solubility, cleaning difficulty); identify the worst-case product in each group (lowest PDE, lowest solubility, most difficult to clean, highest active concentration); validate cleaning for the worst-case; bracket the rest. Equipment grouping similar — identify worst-case equipment (most complex geometry, hardest-to-clean surfaces); validate that one; bracket geometrically similar equipment.

• Dirty hold time (DHT) — maximum time between end of manufacture and start of cleaning. Beyond DHT, residue dries, hardens, or transforms (e.g. proteins denature) and the validated cleaning may no longer be effective. DHT is established by qualification studies — hold equipment with residue for the proposed time, then clean, then verify MAC.
• Clean hold time (CHT) — maximum time between end of cleaning and start of next manufacture. Beyond CHT, microbial regrowth may occur. CHT established by post-cleaning bioburden + endotoxin testing at the proposed time.
• Both DHT + CHT are documented in the cleaning procedure; deviation requires re-cleaning + (for CHT) re-disinfection; routine adherence tracked.

Cleaning validation requires three consecutive successful cleaning runs — each following the documented procedure, sampled per the protocol, evaluated against MAC + bioburden + visual criteria. Three is the industry default rooted in FDA's 1993 guide. The runs should span typical operational conditions (different operators, different shifts where applicable, equipment in typical condition). Three successes do not 'prove' the cleaning works forever — they confirm validation status at that point; ongoing verification (CPV-style) keeps the validation alive.

• Per-cleaning verification — at each commercial-batch cleaning event, visual inspection + (typically) periodic swab or rinse sample at a routine cadence; trend.
• Change triggers re-validation — new product (re-evaluate against the existing worst-case + bracketing), formulation change (re-assess solubility + PDE), equipment change, cleaning agent change, surface material change.
• Trend signals — periodic-verification swab drift, visual fails, manufacturing investigations attributing root cause to cross-contamination, microbial recovery trend.
• Periodic re-qualification — typical 3 – 5 year cycle even without change, especially for high-potency / cytotoxic.

• Carryover calculation using 10 ppm or 1/1000 only, no HBEL — inadequate for high-potency products; EMA / MHRA inspection cite.
• PDE established without qualified toxicologist sign-off — challenged at inspection.
• Recovery study done at one concentration only — non-linear recovery across the MAC range goes undetected.
• Visual-clean as the only criterion — limit of detection ~ 4 μg/cm²; many high-potency MAC limits are below visual detectability.
• Dirty hold time not qualified — operations hold dirty equipment > qualified time; cleaning then fails MAC.
• Worst-case product chosen on potency alone, ignoring solubility — least-soluble product may be harder to clean than the most-potent.
• Cleaning agent changed (e.g. from CIP-100 to alternative) without re-validation.
• Same cleaning procedure used across CIP + manual — actually a different cleaning, never validated separately for the manual operator-variability case.
• Disinfectant + cleaning agent treated as one programme — disinfectant log shows rotation but cleaning-agent effectiveness against soil never re-challenged.
• Periodic-verification trend showing drift toward alert ignored — fails at next regulatory inspection.

• Per-product PDE register: PDE μg/day + toxicologist + literature basis + review date; change-controlled.
• MAC calculator: per equipment × product-previous × product-next combination; HBEL, 10 ppm, 1/1000 calculations side-by-side; applied limit = most restrictive.
• Worst-case bracketing register: product groups + worst-case justification + bracket coverage statement; auto-update when a new product joins a group.
• Validation protocol + record: three-run protocol; per-run sample-point map + result entry + visual inspection sign-off; pass / fail per run; consecutive-success counter.
• Recovery study register: per analyte × surface × swab × diluent; recovery % + range; auto-applied to MAC measurements.
• DHT / CHT qualification + adherence: every cleaning event tagged with prior-batch end + cleaning start + next-batch start; out-of-window events trigger deviation.
• Per-cleaning verification: routine swab / visual schedule per equipment; per-event record; trend with alert / action.
• Change-control linkage: product change / equipment change / agent change auto-routes to cleaning-validation impact assessment.
• Equipment-status gate: equipment cannot be released to next product if cleaning validation is expired, in-investigation, or fails verification.
• Inspection pack: per-equipment cleaning history + validation protocols / reports + verification trend + DHT/CHT adherence — exports as one PDF.