Introducing Reg Affairs Bench

1. Recommendation

Prior approval supplement (PAS) for the change as a whole. The proposed CBE-30 is not supportable, and the guidance line the change owner is relying on is the wrong row for a cellular therapy product.

The 2021 CMC changes guidance for certain biological products covers cellular and gene therapy products within its scope and contains two adjacent rows that resolve this directly:

• The CBE-30 row for "Addition of a duplicated cell culture process train with no change in process parameters specified in the approved BLA" carries the express note "This does not apply to cellular and gene therapy products."
• The matching row for this fact pattern is PAS: "Change in the production capacity for concurrent manufacturing processes by addition of identical unit processes (an increase in throughput)," with the note "This applies only to cellular and gene therapy products."

That second row is a near-verbatim description of CC-2026-0512: identical unit processes (same AX-200 model, software 4.2, tubing set TS-200-C, locked recipe TESS-R7) added to raise concurrent throughput from 8 to 16 lots — per BLA and equipment records. The generic EQUIPMENT row cited by the change owner ("Addition of duplicated process train or process step(s) ... with no change in the process parameters specified in the approved license," CBE-30) is displaced by the CGT-specific row; reading the EQUIPMENT row to cover a CGT throughput increase would make the CGT-only PAS row inoperative.

Statutory/regulatory basis: 21 CFR 601.12(a)(1)–(2) and 601.12(b)(1) (major change — substantial potential to adversely affect identity, strength, quality, purity or potency). Guidance basis: Chemistry, Manufacturing, and Controls Changes to an Approved Application: Certain Biological Products (June 2021), Appendix, 3.2.S.2.2 (Changes to the Upstream Steps of Drug Substance Manufacture through Harvesting) and Appendix, EQUIPMENT.

A secondary, independent PAS trigger exists on the sterility-assurance side (see element 3 below): 601.12(b)(2)(vi) lists as examples of major changes those "which may affect product sterility assurance, such as changes in product or component sterilization method(s), or an addition, deletion, or substitution of steps in an aseptic processing operation". On the present record — media fills run with no more than ten instruments in use, 16-lot concurrency assessed only by risk assessment RA-2B-2026-02 and never simulated, and an unchanged 6-operator cap — per aseptic records — the effect of doubled concurrency on interventions and operator traffic is unproven.

---

2. Element-by-element categorization

Roll-up rule: for multiple related changes, the submission takes the highest-risk category of its elements. The guidance states the applicant should ensure that the highest risk change determines the reporting category for the submission. Elements 1 and 3 are PAS, so CC-2026-0512 is filed as a single PAS; elements 4 and 6 are subsumed into it rather than reported separately.

CBE-0 is not available. CBE-0 under 601.12(c)(5) depends either on FDA having determined that this change type is routinely complete on receipt or on validation under an approved 601.12(e) protocol. There is no approved comparability protocol for this change, so neither path is open.

Consequence of the category call: product from bays 9–16 may not be distributed until FDA approves the supplement (601.12(b)(3)). If the company files CBE-30 and FDA converts it to a PAS within the 30-day window under 601.12(c)(4)(i), distribution must stop until compliance is achieved — a worse commercial outcome than filing PAS at the outset, and for an autologous product that means patient-specific lots already in process on the new bays. Flag this to the change owner explicitly.

---

3. Data package for the PAS

Structured to 601.12(b)(3)(i)–(vii).

Available and adequate

• Detailed change description; product and BLA; site and area affected (Building 2, Suite 2B, bays 9–16) — per change record
• IQ/OQ/PQ for all eight new instruments, reports AX200-IOPQ-09 to -16 — per equipment records
• ER-2026-07: three healthy-donor engineering runs (bays 9, 12, 15); all attributes met release criteria and fell within the historical commercial range of 212 lots — per validation records
• APS-2026-03: three media fills on bays 10, 13, 16 replicating full TESS-R7 including all sterile connections, sampling and closed transfer, at maximum 10-day duration with worst-case interventions; 0 positive units; EM within Grade C action limits — per aseptic records
• HVAC-2B-PQ-2022 qualification at simulated full 16-bay occupancy — per change record
• EM-2B-01 rev. 5 with added locations for bays 9–16 — per EM records
• Statement that TESS-R7 is locked and not editable at instrument level — per BLA record
• Relevant SOP reference list; CC-2026-0512 change record

Gaps to close before filing

1. Patient-lot comparability. ER-2026-07 states on its face that healthy donor cells typically expand better than patient cells and that the runs "demonstrate equipment performance, not patient-lot comparability" — per validation records. A PAS for a CGT throughput increase should not rest on donor runs alone. Propose a defined number of patient lots (or a prospective concurrent-lot comparability plan with pre-specified acceptance criteria against the n=212 commercial distribution) — TBD: number of patient lots and acceptance criteria.
2. Full-concurrency aseptic qualification. The 16-lot condition is the whole point of the change and has never been run: APS ran at no more than 10 instruments in use — per aseptic records — and engineering runs at no more than 8 — per validation records. Either execute an APS or a staffing/traffic simulation at 16-bay occupancy, or propose a phased build-up with defined concurrency tiers.
3. Operator capacity. The suite remains validated for a maximum of 6 operators — per change record — while lots in process double. Provide the time-and-motion or staffing analysis showing 16 lots can be run within the 6-operator cap without compressing critical interventions, or revalidate EM at a higher operator count.
4. RA-2B-2026-02 in full, not by reference — it is currently the only evidence covering full concurrency — per aseptic records.
5. Chain-of-identity / mix-up risk at doubled concurrency. Segregation relies on one lot per instrument plus patient-specific barcode labelling — per change record; provide the assessment specific to 16 simultaneous patient-specific lots.
6. Instrument-level APS coverage strategy. Current program is per-instrument at qualification then twice yearly across a rotating subset — per change record; state the revised rotation for 16 instruments.

---

4. Impacted CTD sections

---

5. Documented facts, assumptions, open questions

Documented facts — per change, BLA, aseptic/EM, and equipment records as stated

• One lot = one patient's leukapheresis = one patient dose; DS and DP made continuously with no stored intermediate, entirely inside one closed AX-200; one lot per instrument
• All steps occur in a single closed system; no open manipulations in the suite
• Suite 2B was designed and qualified at construction (2022) for 16 instrument bays; bays 9–16 were empty behind removable panels
• AHU-2B qualified for full 16-bay occupancy (HVAC-2B-PQ-2022)
• Maximum 6 operators in the suite, validated for EM
• New instruments: same model, software 4.2, tubing set TS-200-C, recipe TESS-R7; recipe not editable at instrument level
• IQ/OQ/PQ complete for bays 9–16; three healthy-donor engineering runs passed and fell within the n=212 historical range; three media fills passed with 0 positives
• 16 concurrent instruments have never been operated; APS ran at no more than 10 in use, engineering runs at no more than 8
• Full-concurrency effect assessed only by risk assessment RA-2B-2026-02, not simulated
• No approved comparability protocol exists for this change (user-stated)

Assumptions (to be confirmed by the change owner)

• Room classification, pressure regime, gowning, material and personnel flows are genuinely unchanged; panel removal involved no construction requiring requalification
• No change to release specifications, analytical methods, testing locations, container closure, freezing profile FP-03, or storage conditions
• No change to lentiviral vector supply, reagent suppliers or ancillary material lot-release arrangements at doubled demand
• Product from bays 9–16 is intended for US commercial distribution (not clinical/expanded access only)
• Suite 2B is single-product; no multiproduct introduction accompanies this change
• Corvane is not concurrently filing an EU or other regional variation whose data package must align

Open questions — internal

1. Does the approved BLA state a numerical capacity or instrument count as a licensed condition, or is 3.2.P.3.3 Section 4 descriptive? This affects whether the capacity statement itself is an approved condition of the license.
2. Was HVAC-2B-PQ-2022 executed with simulated heat load and particle counts at 16-bay occupancy including 6 operators present, or equipment load only? — per change record
3. What does RA-2B-2026-02 conclude on interventions per operator-hour at 16 lots, and what residual risk did it accept? — per aseptic records
4. What is the staffing model at 16 concurrent lots — shift structure, operators per lot, interventions per shift?
5. Is the twice-yearly APS rotation being revised so every instrument is covered on a defined cycle, and over what period? — per change record
6. APS-2026-03 records "four new EM locations ... for bays 9 to 16" — per EM records — while the baseline program had 22 locations set for eight equipped bays — per change record. Four new locations for eight new bays is disproportionate on its face — confirm the number is correct and justified, or correct the record.
7. Are there patient lots already manufactured or in process on bays 9–16? If so, their disposition needs an answer now, independent of the filing category.
8. Would a 601.12(e) comparability protocol be worth filing now (as a PAS) to reduce the reporting burden for future capacity expansions — for example a future Suite 2C build-out? An approved protocol can justify a reduced reporting category for subsequent changes of the same type.

Targeted questions for FDA (Type B/C or informal OTP interaction, before filing)

1. Does CBER agree that the Appendix 3.2.S.2.2 CGT row for increased concurrent production capacity by addition of identical unit processes controls this change and requires a PAS, notwithstanding the generic EQUIPMENT CBE-30 row?
2. Does CBER consider full 16-bay concurrent aseptic process simulation necessary before approval, or is a risk assessment plus a phased concurrency build-up with defined tiers acceptable?
3. How many patient lots from the new instruments, and against what acceptance criteria, would CBER expect for comparability, given that the recipe and parameters are unchanged and the instrument is closed?
4. Is a prospective concurrency-tier approach (approve at 16 bays, implement stepwise with EM and APS confirmation at each tier, reported in the annual report) acceptable in the PAS?
5. Would CBER entertain a 601.12(e) comparability protocol covering future addition of identical AX-200 instruments, and what would it need to contain to support a reduced reporting category?

---

Practical note on the CBE-30 proposal. The change owner's reading is defensible in isolation but fails on the CGT carve-outs, which exist precisely because throughput increases in autologous manufacturing raise mix-up, chain-of-identity and shared-environment risks that a duplicated bioreactor train in a monoclonal process does not. Filing CBE-30 risks a 601.12(c)(4) conversion mid-window with patient-specific lots in process. Recommend PAS, with pre-submission engagement on the comparability and full-concurrency questions so the data package is agreed before it is built.

This is regulatory analysis for internal use, not legal advice, and it is not an FDA determination.

Sources: Chemistry, Manufacturing, and Controls Changes to an Approved Application: Certain Biological Products; Guidance for Industry: https://www.gmp-compliance.org/files/guidemgr/Chemistry_Manufacturing_Controls_Changes_Application_Final_06-16-2021.pdf § 601.12 Changes to an approved application.: https://www.ecfr.gov/current/title-21/section-601.12

1. Executive recommendation

Why the bundle matters: FDA's 2021 guidance is explicit that a single change may be reportable in more than one category depending on its type and extent, and that when considering the correct reporting category for a submission, the applicant should ensure that the highest risk change determines the reporting category chosen, noting that several moderate-risk changes may accompany one major-risk change. The eight instruments cannot be de-coupled from what they are for (doubling concurrent operation) in the same room, with the same air handler, the same 6-operator limit and the same downstream support equipment.

---

2. Documented facts (source-traceable; no interpretation)

From the approved dossier (v2.0, 9 Oct 2025):

From the change package:

Regulatory landscape (verified 15 Sep 2026):

---

3. Assumptions (must be confirmed before the memo is used as the filing basis)

1. A1 — Appendix line quoted by the change owner. The memo assumes the entry "addition of duplicated process train or process step(s) … with no change in the process parameters" exists in the 2021 guidance Appendix in the moderate/CBE‑30 column. RA to pull the Appendix verbatim, record the row number and column, and check for any footnote or carve-out applicable to cellular/gene therapy products or to changes affecting facility capacity/segregation. Industry comment on the draft specifically flagged that some cellular, gene and cell-based gene therapy products are excluded as exceptions in certain examples in the appendix — so this check is not optional.
2. A2 — Scope. BLA 197388 is a 351(a) BLA for a cellular/gene therapy product and is within scope of the 2021 guidance (F23). Confirm.
3. A3 — CFR subparagraph citations. Tier framework cited as 601.12(b) (major/PAS), 601.12(c) (moderate/CBE‑30, including FDA's ability to require a PAS within 30 days of receipt), 601.12(d) (minor/annual report), 601.12(e) (protocols). Verify subparagraph designations against the current CFR text before filing.
4. A4 — Review timing. A manufacturing PAS is assumed to carry a 4-month user-fee goal, with the possibility of a pre-approval inspection of Suite 2B extending real timelines. Confirm current PDUFA commitment language and CBER OTP practice for CGT facility supplements.
5. A5 — Established conditions. The licensed established conditions include the instrument count and the "up to eight lots at the same time" statement (F3), the 16-bay design/qualification (F4/F5), the 6-operator limit (F6), the equipment list (F8) and the 22-location EM program (F9). Any change to these text elements is reportable under 601.12(a).
6. A6 — No other change. Materials, tubing set, vector lot strategy, release specifications, analytical methods, container closure, freezing profile FP‑03, storage conditions, shelf life and labeling are unchanged; no new suite, no new establishment, no new manufacturing address, no CMO.
7. A7 — Downstream capacity. QC/release testing throughput, LN₂ storage, cryoshipper fleet and the 2 controlled-rate freezers in Suite 2C are assumed not to require change. This is an assumption only — see OQ‑4.
8. A8 — No product-quality signal. No trend, deviation or EM excursion in Suite 2B in the last 24 months that would be aggravated by doubled occupancy. To be confirmed by the site quality metrics review.

---

4. Regulatory basis relied upon

---

5. Element-by-element reporting category

---

6. Assessment of the proposed CBE‑30

The Appendix citation is relevant but not sufficient. Three reasons:

1. It addresses the wrong element. "Addition of a duplicated process train … with no change in the process parameters" describes E1. It does not describe an increase in the number of patient-specific lots processed simultaneously in a single classified room, which is a change to the contamination control strategy and segregation regime, i.e. to the facility control elements described in 3.2.A.1. The Appendix is explicitly not an exhaustive list, and a change may be reportable even if not listed in the table.
2. The "worst case" being registered has never been executed. The applicant would be asking FDA to license a 16-lot concurrent state on the strength of (a) a 2022 HVAC qualification at simulated full occupancy, (b) three healthy-donor runs on three of eight instruments at ≤8 concurrency, (c) three media fills at ≤10 concurrency, and (d) a paper risk assessment (F5, F11, F13, F14, F16). For an autologous product where mix-up is a patient-fatal failure mode and where the Appendix categories are derived from FDA's assessment of risks to process performance and product quality, this is the kind of gap that CBER OTP reviewers and investigators pursue.
3. Conversion risk is asymmetric and clinically material. Under 601.12(c) FDA may inform the applicant within 30 days that a PAS is required, which bars distribution. If bays 9–16 are already loaded with patient lots when that notice arrives, Corvane is holding unreleasable autologous product for identified patients — a supply, compliance and ethics event, not merely a regulatory one. A PAS (or the capped CBE‑30 of Option 2) removes that exposure.

Counter-arguments to preserve in the file (they are good ones, and belong in the submission whichever route is chosen):

• Lot size is invariant — one leukapheresis, one dose. This is capacity multiplication, not scale-up; the classic PAS trigger of a changed scale of a unit operation is absent.
• The recipe is locked and not editable at instrument level (F2), so there is no instrument-to-instrument parameter drift pathway.
• The process is closed with no open manipulations (F7); the AHU was qualified at simulated 16-bay full occupancy (F5); and the suite was designed and qualified for 16 bays at construction — i.e. the 16-bay state was contemplated in the approved application.
• Segregation is achieved per-instrument and per-barcode, not by spatial or temporal separation, so adding instruments does not dilute the segregation logic.

---

7. Strategy options

Option 1 — PAS for the full change (recommended default). File one PAS covering E1–E8, with the data package in §8 completed (including a full-occupancy APS/EM/smoke study campaign and an APS on all eight new instruments). Pros: single approval, clean established-conditions update, no distribution risk, and the pre-approval inspection (if any) occurs before commercial use of bays 9–16. Cons: longest time to capacity; requires the full-occupancy simulation work.

Option 2 — CBE‑30 capped at the demonstrated state, then a second filing (recommended if capacity is needed in 2026). File a CBE‑30 covering E1, E3, E4, E5 that installs and qualifies all 16 instruments but registers a maximum concurrency of 10 — the state actually demonstrated in APS‑2026‑03 (F16). 3.2.P.3.3 §4 is amended to "up to ten lots in process at the same time," with an explicit statement that the remaining bays are installed, qualified and held at rest. The step from 10 → 16 is then filed separately (CBE‑30 or PAS per FDA feedback) once a full-occupancy APS campaign and EM/smoke-study qualification at 16 concurrent lots is complete. Pros: aligns the registered claim exactly with the executed evidence, which is the most defensible position in an inspection; delivers 25% capacity now. Cons: two filings; requires operational discipline (scheduling lock-out) to hold the cap, and that lock-out must be proceduralised and auditable.

Option 3 — Add a comparability protocol under 601.12(e) (recommended in either case). Include in the PAS/CBE‑30 a prospective CP defining the tests, studies and acceptance criteria for future additions of identical AX‑200 trains and concurrency increases (e.g. IQ/OQ/PQ, one qualification APS per instrument, N engineering or first-commercial lots per instrument with pre-set CQA acceptance criteria, EM qualification at the new occupancy). An approved CP would allow subsequent expansions — Suite 2C, a second site, or bays beyond the current envelope — to go in at a reduced category. Given the trajectory of this product, the CP is the highest-ROI item in this memo. There is no CP today (F17).

Not viable: CBE‑0 (no designated category or FDA agreement supporting immediate distribution here) and annual report for the change as a whole.

---

8. Data package to support the submission

(Content framed to satisfy FDA's expectation that any supplement include a detailed description of and rationale for the change, the methods and studies used to evaluate the effects of the change on product quality with the data, and a reference list of relevant SOPs.)

1. Change description and rationale, with a before/after established conditions table (instrument count; concurrent-lot maximum; equipment list; EM location count; APS strategy; operator limit) and an explicit statement of what is not changing.
2. Risk assessment RA‑2B‑2026‑02 in full (not a summary), with hazard-by-hazard mitigation and residual-risk conclusions, covering at minimum: chain-of-identity/mix-up at 16 concurrent lots; intervention and traffic load at 6 operators; gowning/airlock throughput; simultaneous alarm handling and deviation management; utility load (power, CO₂, exhaust, data) at 16 bays; waste and material flow; harvest-day collision (day 9, range 8–10) across up to 16 lots; and QC sampling/testing throughput.
3. Equipment qualification summaries AX200‑IOPQ‑09 to ‑16, with the traceability matrix showing identity of model, software 4.2 configuration and TS‑200‑C tubing set, plus 21 CFR Part 11 / data-integrity assessment for the eight new instrument control systems and their interface to the scheduling/COI system.
4. ER‑2026‑07 with an honest statement of its limits — three instruments, healthy donors, ≤8 concurrency (F11–F13) — and the per-attribute comparison against the n = 212 commercial baseline. Do not present it as comparability evidence; the report itself does not.
5. APS‑2026‑03 plus the gap closure: a qualification APS on each of the five new instruments not yet media-filled (F18), and — for a 16-lot claim — an APS/EM campaign at maximum occupancy with worst-case interventions and the maximum operator complement, supported by updated smoke/airflow visualisation studies at 16-bay occupancy and a re-verification of AHU‑2B performance with instruments and personnel actually present (the 2022 PQ was a simulation, F5).
6. EM program EM‑2B‑01 rev 5 with the location-selection rationale for bays 9–16 and a justification of proportionality against the original 22-location design (F9/F15); action/alert limits unchanged; trend data for bays 1–8 over 24 months as the baseline.
7. Segregation / chain-of-identity control package: barcode and two-person verification workflow at 16 concurrent lots, scheduling system controls preventing double-allocation of an instrument or a tubing set, label reconciliation, and the capacity/lock-out control that enforces the registered concurrency maximum (essential under Option 2).
8. Continued process verification (CPV) commitment, e.g. per-instrument monitoring of viability, %CAR+, VCN, IFN‑γ potency, total viable CAR+ cells and endotoxin for the first defined number of commercial lots on each new bay, with pre-specified statistical comparison to the n = 212 historical distribution, pre-defined action thresholds, and a commitment to report outcomes in the annual report (and to notify FDA of any signal). This is the mechanism by which patient-lot equivalence — which cannot be generated pre-implementation — is actually demonstrated, and is consistent with CBER's stated risk-based posture (F26) and with the acknowledgement in the May 2026 guidance that FDA may consider extending some CGT CMC flexibilities to post-licensure manufacturing supplements.
9. Support-capacity assessment for the 2 BSCs, 4 welders and 2 controlled-rate freezers, including day-9 harvest queuing analysis, hold-time justification between formulation and controlled-rate freezing, and LN₂ storage/shipper capacity.
10. Personnel qualification and training records/plan for the new bays; aseptic technique requalification; assessment of whether the 6-operator limit is sustainable at 16 lots.
11. Statements of no impact on specifications, analytical methods, materials (including vector), container closure, FP‑03 freezing profile, storage, shelf life/stability and labeling.
12. SOP reference list and the updated CTD sections (annotated, clean and redline).

---

9. Impacted CTD sections

---

10. Open questions

10.1 For the change owner / MSAT / site quality (must be closed before filing)

• OQ‑1. What is the business-required concurrency, and by when? If ≤10 lots meets 2026–27 demand, Option 2 is materially lower risk for no commercial cost.
• OQ‑2. Can 6 operators support 16 concurrent lots across the full day-0-to-day-10 cycle, including worst-case simultaneous harvests? Provide the staffing model and the interventions-per-shift calculation. If the answer is no, the operator limit is itself a change requiring EM/APS revalidation (E6).
• OQ‑3. Why were only 4 EM locations added for 8 new bays when 22 locations serve 8 existing bays? What is the documented sampling-point rationale (F9/F15)?
• OQ‑4. What is the day-9 harvest throughput limit given 2 BSCs, 4 tube welders and 2 controlled-rate freezers? What is the validated hold time between formulation and the start of FP‑03, and does a 16-lot schedule ever breach it?
• OQ‑5. Why were APS and engineering runs performed on different, non-overlapping instruments (bays 9/12/15 vs 10/13/16), leaving bays 11 and 14 with neither? What is the qualification rationale for releasing an instrument to commercial use with neither an engineering run nor an APS (F18)?
• OQ‑6. What is the plan and timeline for a full-occupancy APS/EM/smoke-study campaign at 16 concurrent lots with maximum operator load?
• OQ‑7. Does the scheduling/MES system have an enforceable, auditable control that caps concurrent lots at the registered maximum? (Prerequisite for Option 2.)
• OQ‑8. Does RA‑2B‑2026‑02 specifically address chain-of-identity mix-up as a hazard at 16 concurrent lots, with quantified detectability, or does it treat segregation as adequate by virtue of the closed-system/barcode design?
• OQ‑9. QC and release throughput: can the sterility, mycoplasma, VCN and potency workflows absorb doubled lot flow within the release timelines relied on for patient scheduling?
• OQ‑10. Any Suite 2B EM excursions, deviations, invalidated lots or out-of-trend CQA results in the last 24 months (A8)?
• OQ‑11. Are there parallel or planned changes (e.g. AX‑200 software upgrade, Suite 2C expansion, second site) that should be bundled or sequenced with CC‑2026‑0512 rather than filed serially?
• OQ‑12. Ex-US: EU/UK/JP/CA reporting categories for the same change, and whether the US strategy creates a divergent registered capacity across regions. (Out of scope for this memo; flag to International RA now, not after US filing.)

10.2 Proposed questions for FDA (CBER OTP) — written advice request or CMC meeting

1. Does the Agency agree that installation of eight additional, identical AX‑200 closed-system trains in the already-licensed and 16-bay-qualified Suite 2B, with no change to process parameters, materials, specifications or the locked TESS‑R7 recipe, is a moderate change reportable in a CBE‑30 under 21 CFR 601.12(c)?
2. Does the Agency consider the associated increase in maximum concurrent lots in process from 8 to 16 to be a separately assessable change and, if so, is it a major change requiring a PAS under 601.12(b)?
3. Would the Agency accept a staged approach in which a CBE‑30 registers a concurrency maximum equal to the demonstrated state (10 lots, per APS‑2026‑03), with the increase to 16 filed after a full-occupancy aseptic process simulation and EM qualification campaign?
4. For the 16-lot claim, is an APS campaign at maximum occupancy with the maximum operator complement required, or would the Agency accept the risk assessment (RA‑2B‑2026‑02) plus per-instrument APS supplemented by enhanced EM at the new occupancy?
5. Does the Agency agree that patient-lot comparability for the new instruments is appropriately demonstrated prospectively through a defined CPV plan with pre-specified CQA acceptance criteria against the n = 212 historical baseline, given that healthy-donor engineering runs cannot demonstrate patient-lot comparability and that autologous lots cannot be manufactured in duplicate?
6. Does the Agency have comments on the adequacy of the EM sampling-point expansion for bays 9–16?
7. Would the Agency review a comparability protocol under 601.12(e) covering future additions of identical AX‑200 process trains and defined concurrency increases, to support a reduced reporting category for subsequent expansions? What content would the Agency expect?
8. Should a PAS be filed, does the Agency anticipate a pre-approval inspection of Suite 2B, and are there sequencing preferences relative to the completion of the full-occupancy APS campaign?

Basis for the interaction: the 2021 guidance itself recommends early consultation with FDA for manufacturing changes not described in the Appendix table, or when considering an adjustment to the recommended reporting categories, and the CGT change/comparability guidance remains a draft (F24), so a documented FDA position is the only way to de-risk the category call.

---

11. Next steps

Bottom line: the Appendix line the change owner cites supports the instruments; it does not support the doubling of concurrent autologous lots in a single room that has never been run at that occupancy. Either file the PAS, or register only what has been demonstrated and earn the rest with a full-occupancy simulation campaign — and use this filing to put a 601.12(e) comparability protocol in place so the next expansion is a notification rather than a negotiation.